diff options
Diffstat (limited to 'arch/x86')
204 files changed, 8247 insertions, 5182 deletions
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 5e8949953660..a283336bb5d5 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -93,10 +93,11 @@ config X86 select ARCH_USE_QUEUED_RWLOCKS select ARCH_USE_QUEUED_SPINLOCKS select ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH + select ARCH_WANT_DEFAULT_BPF_JIT if X86_64 select ARCH_WANTS_DYNAMIC_TASK_STRUCT select ARCH_WANT_HUGE_PMD_SHARE select ARCH_WANTS_THP_SWAP if X86_64 - select BUILDTIME_EXTABLE_SORT + select BUILDTIME_TABLE_SORT select CLKEVT_I8253 select CLOCKSOURCE_VALIDATE_LAST_CYCLE select CLOCKSOURCE_WATCHDOG @@ -124,6 +125,7 @@ config X86 select GENERIC_STRNLEN_USER select GENERIC_TIME_VSYSCALL select GENERIC_GETTIMEOFDAY + select GENERIC_VDSO_TIME_NS select GUP_GET_PTE_LOW_HIGH if X86_PAE select HARDLOCKUP_CHECK_TIMESTAMP if X86_64 select HAVE_ACPI_APEI if ACPI @@ -439,8 +441,8 @@ config X86_MPPARSE (esp with 64bit cpus) with acpi support, MADT and DSDT will override it config GOLDFISH - def_bool y - depends on X86_GOLDFISH + def_bool y + depends on X86_GOLDFISH config RETPOLINE bool "Avoid speculative indirect branches in kernel" @@ -456,6 +458,7 @@ config X86_CPU_RESCTRL bool "x86 CPU resource control support" depends on X86 && (CPU_SUP_INTEL || CPU_SUP_AMD) select KERNFS + select PROC_CPU_RESCTRL if PROC_FS help Enable x86 CPU resource control support. @@ -477,7 +480,7 @@ config X86_BIGSMP bool "Support for big SMP systems with more than 8 CPUs" depends on SMP ---help--- - This option is needed for the systems that have more than 8 CPUs + This option is needed for the systems that have more than 8 CPUs. config X86_EXTENDED_PLATFORM bool "Support for extended (non-PC) x86 platforms" @@ -561,9 +564,9 @@ config X86_UV # Please maintain the alphabetic order if and when there are additions config X86_GOLDFISH - bool "Goldfish (Virtual Platform)" - depends on X86_EXTENDED_PLATFORM - ---help--- + bool "Goldfish (Virtual Platform)" + depends on X86_EXTENDED_PLATFORM + ---help--- Enable support for the Goldfish virtual platform used primarily for Android development. Unless you are building for the Android Goldfish emulator say N here. @@ -806,9 +809,9 @@ config KVM_GUEST timing infrastructure such as time of day, and system time config ARCH_CPUIDLE_HALTPOLL - def_bool n - prompt "Disable host haltpoll when loading haltpoll driver" - help + def_bool n + prompt "Disable host haltpoll when loading haltpoll driver" + help If virtualized under KVM, disable host haltpoll. config PVH @@ -887,16 +890,16 @@ config HPET_EMULATE_RTC depends on HPET_TIMER && (RTC=y || RTC=m || RTC_DRV_CMOS=m || RTC_DRV_CMOS=y) config APB_TIMER - def_bool y if X86_INTEL_MID - prompt "Intel MID APB Timer Support" if X86_INTEL_MID - select DW_APB_TIMER - depends on X86_INTEL_MID && SFI - help - APB timer is the replacement for 8254, HPET on X86 MID platforms. - The APBT provides a stable time base on SMP - systems, unlike the TSC, but it is more expensive to access, - as it is off-chip. APB timers are always running regardless of CPU - C states, they are used as per CPU clockevent device when possible. + def_bool y if X86_INTEL_MID + prompt "Intel MID APB Timer Support" if X86_INTEL_MID + select DW_APB_TIMER + depends on X86_INTEL_MID && SFI + help + APB timer is the replacement for 8254, HPET on X86 MID platforms. + The APBT provides a stable time base on SMP + systems, unlike the TSC, but it is more expensive to access, + as it is off-chip. APB timers are always running regardless of CPU + C states, they are used as per CPU clockevent device when possible. # Mark as expert because too many people got it wrong. # The code disables itself when not needed. @@ -1035,8 +1038,8 @@ config SCHED_MC_PRIO If unsure say Y here. config UP_LATE_INIT - def_bool y - depends on !SMP && X86_LOCAL_APIC + def_bool y + depends on !SMP && X86_LOCAL_APIC config X86_UP_APIC bool "Local APIC support on uniprocessors" if !PCI_MSI @@ -1185,8 +1188,8 @@ config X86_LEGACY_VM86 If unsure, say N here. config VM86 - bool - default X86_LEGACY_VM86 + bool + default X86_LEGACY_VM86 config X86_16BIT bool "Enable support for 16-bit segments" if EXPERT @@ -1207,10 +1210,10 @@ config X86_ESPFIX64 depends on X86_16BIT && X86_64 config X86_VSYSCALL_EMULATION - bool "Enable vsyscall emulation" if EXPERT - default y - depends on X86_64 - ---help--- + bool "Enable vsyscall emulation" if EXPERT + default y + depends on X86_64 + ---help--- This enables emulation of the legacy vsyscall page. Disabling it is roughly equivalent to booting with vsyscall=none, except that it will also disable the helpful warning if a program @@ -1512,7 +1515,7 @@ config X86_CPA_STATISTICS bool "Enable statistic for Change Page Attribute" depends on DEBUG_FS ---help--- - Expose statistics about the Change Page Attribute mechanims, which + Expose statistics about the Change Page Attribute mechanism, which helps to determine the effectiveness of preserving large and huge page mappings when mapping protections are changed. @@ -1543,12 +1546,12 @@ config AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT # Common NUMA Features config NUMA - bool "Numa Memory Allocation and Scheduler Support" + bool "NUMA Memory Allocation and Scheduler Support" depends on SMP depends on X86_64 || (X86_32 && HIGHMEM64G && X86_BIGSMP) default y if X86_BIGSMP ---help--- - Enable NUMA (Non Uniform Memory Access) support. + Enable NUMA (Non-Uniform Memory Access) support. The kernel will try to allocate memory used by a CPU on the local memory controller of the CPU and add some more @@ -1648,9 +1651,9 @@ config ARCH_PROC_KCORE_TEXT depends on X86_64 && PROC_KCORE config ILLEGAL_POINTER_VALUE - hex - default 0 if X86_32 - default 0xdead000000000000 if X86_64 + hex + default 0 if X86_32 + default 0xdead000000000000 if X86_64 config X86_PMEM_LEGACY_DEVICE bool @@ -1991,11 +1994,12 @@ config EFI platforms. config EFI_STUB - bool "EFI stub support" - depends on EFI && !X86_USE_3DNOW - select RELOCATABLE - ---help--- - This kernel feature allows a bzImage to be loaded directly + bool "EFI stub support" + depends on EFI && !X86_USE_3DNOW + depends on $(cc-option,-mabi=ms) || X86_32 + select RELOCATABLE + ---help--- + This kernel feature allows a bzImage to be loaded directly by EFI firmware without the use of a bootloader. See Documentation/admin-guide/efi-stub.rst for more information. diff --git a/arch/x86/Kconfig.cpu b/arch/x86/Kconfig.cpu index af9c967782f6..bc3a497c029c 100644 --- a/arch/x86/Kconfig.cpu +++ b/arch/x86/Kconfig.cpu @@ -387,6 +387,14 @@ config X86_DEBUGCTLMSR def_bool y depends on !(MK6 || MWINCHIPC6 || MWINCHIP3D || MCYRIXIII || M586MMX || M586TSC || M586 || M486SX || M486) && !UML +config IA32_FEAT_CTL + def_bool y + depends on CPU_SUP_INTEL || CPU_SUP_CENTAUR || CPU_SUP_ZHAOXIN + +config X86_VMX_FEATURE_NAMES + def_bool y + depends on IA32_FEAT_CTL && X86_FEATURE_NAMES + menuconfig PROCESSOR_SELECT bool "Supported processor vendors" if EXPERT ---help--- diff --git a/arch/x86/boot/Makefile b/arch/x86/boot/Makefile index 95410d6ee2ff..748b6d28a91d 100644 --- a/arch/x86/boot/Makefile +++ b/arch/x86/boot/Makefile @@ -88,7 +88,7 @@ $(obj)/vmlinux.bin: $(obj)/compressed/vmlinux FORCE SETUP_OBJS = $(addprefix $(obj)/,$(setup-y)) -sed-zoffset := -e 's/^\([0-9a-fA-F]*\) [ABCDGRSTVW] \(startup_32\|startup_64\|efi32_stub_entry\|efi64_stub_entry\|efi_pe_entry\|input_data\|kernel_info\|_end\|_ehead\|_text\|z_.*\)$$/\#define ZO_\2 0x\1/p' +sed-zoffset := -e 's/^\([0-9a-fA-F]*\) [a-zA-Z] \(startup_32\|startup_64\|efi32_stub_entry\|efi64_stub_entry\|efi_pe_entry\|input_data\|kernel_info\|_end\|_ehead\|_text\|z_.*\)$$/\#define ZO_\2 0x\1/p' quiet_cmd_zoffset = ZOFFSET $@ cmd_zoffset = $(NM) $< | sed -n $(sed-zoffset) > $@ diff --git a/arch/x86/boot/compressed/Makefile b/arch/x86/boot/compressed/Makefile index 1dac210f7d44..56aa5fa0a66b 100644 --- a/arch/x86/boot/compressed/Makefile +++ b/arch/x86/boot/compressed/Makefile @@ -89,7 +89,7 @@ vmlinux-objs-$(CONFIG_ACPI) += $(obj)/acpi.o $(obj)/eboot.o: KBUILD_CFLAGS += -fshort-wchar -mno-red-zone -vmlinux-objs-$(CONFIG_EFI_STUB) += $(obj)/eboot.o $(obj)/efi_stub_$(BITS).o \ +vmlinux-objs-$(CONFIG_EFI_STUB) += $(obj)/eboot.o \ $(objtree)/drivers/firmware/efi/libstub/lib.a vmlinux-objs-$(CONFIG_EFI_MIXED) += $(obj)/efi_thunk_$(BITS).o diff --git a/arch/x86/boot/compressed/eboot.c b/arch/x86/boot/compressed/eboot.c index 72b08fde6de6..287393d725f0 100644 --- a/arch/x86/boot/compressed/eboot.c +++ b/arch/x86/boot/compressed/eboot.c @@ -6,6 +6,8 @@ * * ----------------------------------------------------------------------- */ +#pragma GCC visibility push(hidden) + #include <linux/efi.h> #include <linux/pci.h> @@ -19,32 +21,18 @@ #include "eboot.h" static efi_system_table_t *sys_table; +extern const bool efi_is64; -static struct efi_config *efi_early; - -__pure const struct efi_config *__efi_early(void) +__pure efi_system_table_t *efi_system_table(void) { - return efi_early; + return sys_table; } -#define BOOT_SERVICES(bits) \ -static void setup_boot_services##bits(struct efi_config *c) \ -{ \ - efi_system_table_##bits##_t *table; \ - \ - table = (typeof(table))sys_table; \ - \ - c->runtime_services = table->runtime; \ - c->boot_services = table->boottime; \ - c->text_output = table->con_out; \ -} -BOOT_SERVICES(32); -BOOT_SERVICES(64); - -void efi_char16_printk(efi_system_table_t *table, efi_char16_t *str) +__attribute_const__ bool efi_is_64bit(void) { - efi_call_proto(efi_simple_text_output_protocol, output_string, - efi_early->text_output, str); + if (IS_ENABLED(CONFIG_EFI_MIXED)) + return efi_is64; + return IS_ENABLED(CONFIG_X86_64); } static efi_status_t @@ -63,17 +51,17 @@ preserve_pci_rom_image(efi_pci_io_protocol_t *pci, struct pci_setup_rom **__rom) * large romsize. The UEFI spec limits the size of option ROMs to 16 * MiB so we reject any ROMs over 16 MiB in size to catch this. */ - romimage = (void *)(unsigned long)efi_table_attr(efi_pci_io_protocol, - romimage, pci); - romsize = efi_table_attr(efi_pci_io_protocol, romsize, pci); + romimage = efi_table_attr(pci, romimage); + romsize = efi_table_attr(pci, romsize); if (!romimage || !romsize || romsize > SZ_16M) return EFI_INVALID_PARAMETER; size = romsize + sizeof(*rom); - status = efi_call_early(allocate_pool, EFI_LOADER_DATA, size, &rom); + status = efi_bs_call(allocate_pool, EFI_LOADER_DATA, size, + (void **)&rom); if (status != EFI_SUCCESS) { - efi_printk(sys_table, "Failed to allocate memory for 'rom'\n"); + efi_printk("Failed to allocate memory for 'rom'\n"); return status; } @@ -85,27 +73,24 @@ preserve_pci_rom_image(efi_pci_io_protocol_t *pci, struct pci_setup_rom **__rom) rom->pcilen = pci->romsize; *__rom = rom; - status = efi_call_proto(efi_pci_io_protocol, pci.read, pci, - EfiPciIoWidthUint16, PCI_VENDOR_ID, 1, - &rom->vendor); + status = efi_call_proto(pci, pci.read, EfiPciIoWidthUint16, + PCI_VENDOR_ID, 1, &rom->vendor); if (status != EFI_SUCCESS) { - efi_printk(sys_table, "Failed to read rom->vendor\n"); + efi_printk("Failed to read rom->vendor\n"); goto free_struct; } - status = efi_call_proto(efi_pci_io_protocol, pci.read, pci, - EfiPciIoWidthUint16, PCI_DEVICE_ID, 1, - &rom->devid); + status = efi_call_proto(pci, pci.read, EfiPciIoWidthUint16, + PCI_DEVICE_ID, 1, &rom->devid); if (status != EFI_SUCCESS) { - efi_printk(sys_table, "Failed to read rom->devid\n"); + efi_printk("Failed to read rom->devid\n"); goto free_struct; } - status = efi_call_proto(efi_pci_io_protocol, get_location, pci, - &rom->segment, &rom->bus, &rom->device, - &rom->function); + status = efi_call_proto(pci, get_location, &rom->segment, &rom->bus, + &rom->device, &rom->function); if (status != EFI_SUCCESS) goto free_struct; @@ -114,7 +99,7 @@ preserve_pci_rom_image(efi_pci_io_protocol_t *pci, struct pci_setup_rom **__rom) return status; free_struct: - efi_call_early(free_pool, rom); + efi_bs_call(free_pool, rom); return status; } @@ -133,27 +118,24 @@ static void setup_efi_pci(struct boot_params *params) void **pci_handle = NULL; efi_guid_t pci_proto = EFI_PCI_IO_PROTOCOL_GUID; unsigned long size = 0; - unsigned long nr_pci; struct setup_data *data; + efi_handle_t h; int i; - status = efi_call_early(locate_handle, - EFI_LOCATE_BY_PROTOCOL, - &pci_proto, NULL, &size, pci_handle); + status = efi_bs_call(locate_handle, EFI_LOCATE_BY_PROTOCOL, + &pci_proto, NULL, &size, pci_handle); if (status == EFI_BUFFER_TOO_SMALL) { - status = efi_call_early(allocate_pool, - EFI_LOADER_DATA, - size, (void **)&pci_handle); + status = efi_bs_call(allocate_pool, EFI_LOADER_DATA, size, + (void **)&pci_handle); if (status != EFI_SUCCESS) { - efi_printk(sys_table, "Failed to allocate memory for 'pci_handle'\n"); + efi_printk("Failed to allocate memory for 'pci_handle'\n"); return; } - status = efi_call_early(locate_handle, - EFI_LOCATE_BY_PROTOCOL, &pci_proto, - NULL, &size, pci_handle); + status = efi_bs_call(locate_handle, EFI_LOCATE_BY_PROTOCOL, + &pci_proto, NULL, &size, pci_handle); } if (status != EFI_SUCCESS) @@ -164,15 +146,12 @@ static void setup_efi_pci(struct boot_params *params) while (data && data->next) data = (struct setup_data *)(unsigned long)data->next; - nr_pci = size / (efi_is_64bit() ? sizeof(u64) : sizeof(u32)); - for (i = 0; i < nr_pci; i++) { + for_each_efi_handle(h, pci_handle, size, i) { efi_pci_io_protocol_t *pci = NULL; struct pci_setup_rom *rom; - status = efi_call_early(handle_protocol, - efi_is_64bit() ? ((u64 *)pci_handle)[i] - : ((u32 *)pci_handle)[i], - &pci_proto, (void **)&pci); + status = efi_bs_call(handle_protocol, h, &pci_proto, + (void **)&pci); if (status != EFI_SUCCESS || !pci) continue; @@ -189,7 +168,7 @@ static void setup_efi_pci(struct boot_params *params) } free_handle: - efi_call_early(free_pool, pci_handle); + efi_bs_call(free_pool, pci_handle); } static void retrieve_apple_device_properties(struct boot_params *boot_params) @@ -198,34 +177,34 @@ static void retrieve_apple_device_properties(struct boot_params *boot_params) struct setup_data *data, *new; efi_status_t status; u32 size = 0; - void *p; + apple_properties_protocol_t *p; - status = efi_call_early(locate_protocol, &guid, NULL, &p); + status = efi_bs_call(locate_protocol, &guid, NULL, (void **)&p); if (status != EFI_SUCCESS) return; - if (efi_table_attr(apple_properties_protocol, version, p) != 0x10000) { - efi_printk(sys_table, "Unsupported properties proto version\n"); + if (efi_table_attr(p, version) != 0x10000) { + efi_printk("Unsupported properties proto version\n"); return; } - efi_call_proto(apple_properties_protocol, get_all, p, NULL, &size); + efi_call_proto(p, get_all, NULL, &size); if (!size) return; do { - status = efi_call_early(allocate_pool, EFI_LOADER_DATA, - size + sizeof(struct setup_data), &new); + status = efi_bs_call(allocate_pool, EFI_LOADER_DATA, + size + sizeof(struct setup_data), + (void **)&new); if (status != EFI_SUCCESS) { - efi_printk(sys_table, "Failed to allocate memory for 'properties'\n"); + efi_printk("Failed to allocate memory for 'properties'\n"); return; } - status = efi_call_proto(apple_properties_protocol, get_all, p, - new->data, &size); + status = efi_call_proto(p, get_all, new->data, &size); if (status == EFI_BUFFER_TOO_SMALL) - efi_call_early(free_pool, new); + efi_bs_call(free_pool, new); } while (status == EFI_BUFFER_TOO_SMALL); new->type = SETUP_APPLE_PROPERTIES; @@ -247,7 +226,7 @@ static const efi_char16_t apple[] = L"Apple"; static void setup_quirks(struct boot_params *boot_params) { efi_char16_t *fw_vendor = (efi_char16_t *)(unsigned long) - efi_table_attr(efi_system_table, fw_vendor, sys_table); + efi_table_attr(efi_system_table(), fw_vendor); if (!memcmp(fw_vendor, apple, sizeof(apple))) { if (IS_ENABLED(CONFIG_APPLE_PROPERTIES)) @@ -265,17 +244,16 @@ setup_uga(struct screen_info *si, efi_guid_t *uga_proto, unsigned long size) u32 width, height; void **uga_handle = NULL; efi_uga_draw_protocol_t *uga = NULL, *first_uga; - unsigned long nr_ugas; + efi_handle_t handle; int i; - status = efi_call_early(allocate_pool, EFI_LOADER_DATA, - size, (void **)&uga_handle); + status = efi_bs_call(allocate_pool, EFI_LOADER_DATA, size, + (void **)&uga_handle); if (status != EFI_SUCCESS) return status; - status = efi_call_early(locate_handle, - EFI_LOCATE_BY_PROTOCOL, - uga_proto, NULL, &size, uga_handle); + status = efi_bs_call(locate_handle, EFI_LOCATE_BY_PROTOCOL, + uga_proto, NULL, &size, uga_handle); if (status != EFI_SUCCESS) goto free_handle; @@ -283,24 +261,20 @@ setup_uga(struct screen_info *si, efi_guid_t *uga_proto, unsigned long size) width = 0; first_uga = NULL; - nr_ugas = size / (efi_is_64bit() ? sizeof(u64) : sizeof(u32)); - for (i = 0; i < nr_ugas; i++) { + for_each_efi_handle(handle, uga_handle, size, i) { efi_guid_t pciio_proto = EFI_PCI_IO_PROTOCOL_GUID; u32 w, h, depth, refresh; void *pciio; - unsigned long handle = efi_is_64bit() ? ((u64 *)uga_handle)[i] - : ((u32 *)uga_handle)[i]; - status = efi_call_early(handle_protocol, handle, - uga_proto, (void **)&uga); + status = efi_bs_call(handle_protocol, handle, uga_proto, + (void **)&uga); if (status != EFI_SUCCESS) continue; pciio = NULL; - efi_call_early(handle_protocol, handle, &pciio_proto, &pciio); + efi_bs_call(handle_protocol, handle, &pciio_proto, &pciio); - status = efi_call_proto(efi_uga_draw_protocol, get_mode, uga, - &w, &h, &depth, &refresh); + status = efi_call_proto(uga, get_mode, &w, &h, &depth, &refresh); if (status == EFI_SUCCESS && (!first_uga || pciio)) { width = w; height = h; @@ -336,7 +310,7 @@ setup_uga(struct screen_info *si, efi_guid_t *uga_proto, unsigned long size) si->rsvd_pos = 24; free_handle: - efi_call_early(free_pool, uga_handle); + efi_bs_call(free_pool, uga_handle); return status; } @@ -355,37 +329,38 @@ void setup_graphics(struct boot_params *boot_params) memset(si, 0, sizeof(*si)); size = 0; - status = efi_call_early(locate_handle, - EFI_LOCATE_BY_PROTOCOL, - &graphics_proto, NULL, &size, gop_handle); + status = efi_bs_call(locate_handle, EFI_LOCATE_BY_PROTOCOL, + &graphics_proto, NULL, &size, gop_handle); if (status == EFI_BUFFER_TOO_SMALL) - status = efi_setup_gop(NULL, si, &graphics_proto, size); + status = efi_setup_gop(si, &graphics_proto, size); if (status != EFI_SUCCESS) { size = 0; - status = efi_call_early(locate_handle, - EFI_LOCATE_BY_PROTOCOL, - &uga_proto, NULL, &size, uga_handle); + status = efi_bs_call(locate_handle, EFI_LOCATE_BY_PROTOCOL, + &uga_proto, NULL, &size, uga_handle); if (status == EFI_BUFFER_TOO_SMALL) setup_uga(si, &uga_proto, size); } } +void startup_32(struct boot_params *boot_params); + +void __noreturn efi_stub_entry(efi_handle_t handle, + efi_system_table_t *sys_table_arg, + struct boot_params *boot_params); + /* * Because the x86 boot code expects to be passed a boot_params we * need to create one ourselves (usually the bootloader would create * one for us). - * - * The caller is responsible for filling out ->code32_start in the - * returned boot_params. */ -struct boot_params *make_boot_params(struct efi_config *c) +efi_status_t __efiapi efi_pe_entry(efi_handle_t handle, + efi_system_table_t *sys_table_arg) { struct boot_params *boot_params; struct apm_bios_info *bi; struct setup_header *hdr; efi_loaded_image_t *image; - void *handle; efi_guid_t proto = LOADED_IMAGE_PROTOCOL_GUID; int options_size = 0; efi_status_t status; @@ -393,31 +368,22 @@ struct boot_params *make_boot_params(struct efi_config *c) unsigned long ramdisk_addr; unsigned long ramdisk_size; - efi_early = c; - sys_table = (efi_system_table_t *)(unsigned long)efi_early->table; - handle = (void *)(unsigned long)efi_early->image_handle; + sys_table = sys_table_arg; /* Check if we were booted by the EFI firmware */ if (sys_table->hdr.signature != EFI_SYSTEM_TABLE_SIGNATURE) - return NULL; - - if (efi_is_64bit()) - setup_boot_services64(efi_early); - else - setup_boot_services32(efi_early); + return EFI_INVALID_PARAMETER; - status = efi_call_early(handle_protocol, handle, - &proto, (void *)&image); + status = efi_bs_call(handle_protocol, handle, &proto, (void *)&image); if (status != EFI_SUCCESS) { - efi_printk(sys_table, "Failed to get handle for LOADED_IMAGE_PROTOCOL\n"); - return NULL; + efi_printk("Failed to get handle for LOADED_IMAGE_PROTOCOL\n"); + return status; } - status = efi_low_alloc(sys_table, 0x4000, 1, - (unsigned long *)&boot_params); + status = efi_low_alloc(0x4000, 1, (unsigned long *)&boot_params); if (status != EFI_SUCCESS) { - efi_printk(sys_table, "Failed to allocate lowmem for boot params\n"); - return NULL; + efi_printk("Failed to allocate lowmem for boot params\n"); + return status; } memset(boot_params, 0x0, 0x4000); @@ -439,7 +405,7 @@ struct boot_params *make_boot_params(struct efi_config *c) hdr->type_of_loader = 0x21; /* Convert unicode cmdline to ascii */ - cmdline_ptr = efi_convert_cmdline(sys_table, image, &options_size); + cmdline_ptr = efi_convert_cmdline(image, &options_size); if (!cmdline_ptr) goto fail; @@ -457,15 +423,15 @@ struct boot_params *make_boot_params(struct efi_config *c) if (status != EFI_SUCCESS) goto fail2; - status = handle_cmdline_files(sys_table, image, + status = handle_cmdline_files(image, (char *)(unsigned long)hdr->cmd_line_ptr, "initrd=", hdr->initrd_addr_max, &ramdisk_addr, &ramdisk_size); if (status != EFI_SUCCESS && hdr->xloadflags & XLF_CAN_BE_LOADED_ABOVE_4G) { - efi_printk(sys_table, "Trying to load files to higher address\n"); - status = handle_cmdline_files(sys_table, image, + efi_printk("Trying to load files to higher address\n"); + status = handle_cmdline_files(image, (char *)(unsigned long)hdr->cmd_line_ptr, "initrd=", -1UL, &ramdisk_addr, &ramdisk_size); @@ -478,14 +444,17 @@ struct boot_params *make_boot_params(struct efi_config *c) boot_params->ext_ramdisk_image = (u64)ramdisk_addr >> 32; boot_params->ext_ramdisk_size = (u64)ramdisk_size >> 32; - return boot_params; + hdr->code32_start = (u32)(unsigned long)startup_32; + + efi_stub_entry(handle, sys_table, boot_params); + /* not reached */ fail2: - efi_free(sys_table, options_size, hdr->cmd_line_ptr); + efi_free(options_size, hdr->cmd_line_ptr); fail: - efi_free(sys_table, 0x4000, (unsigned long)boot_params); + efi_free(0x4000, (unsigned long)boot_params); - return NULL; + return status; } static void add_e820ext(struct boot_params *params, @@ -620,13 +589,13 @@ static efi_status_t alloc_e820ext(u32 nr_desc, struct setup_data **e820ext, sizeof(struct e820_entry) * nr_desc; if (*e820ext) { - efi_call_early(free_pool, *e820ext); + efi_bs_call(free_pool, *e820ext); *e820ext = NULL; *e820ext_size = 0; } - status = efi_call_early(allocate_pool, EFI_LOADER_DATA, - size, (void **)e820ext); + status = efi_bs_call(allocate_pool, EFI_LOADER_DATA, size, + (void **)e820ext); if (status == EFI_SUCCESS) *e820ext_size = size; @@ -650,7 +619,7 @@ static efi_status_t allocate_e820(struct boot_params *params, boot_map.key_ptr = NULL; boot_map.buff_size = &buff_size; - status = efi_get_memory_map(sys_table, &boot_map); + status = efi_get_memory_map(&boot_map); if (status != EFI_SUCCESS) return status; @@ -672,8 +641,7 @@ struct exit_boot_struct { struct efi_info *efi; }; -static efi_status_t exit_boot_func(efi_system_table_t *sys_table_arg, - struct efi_boot_memmap *map, +static efi_status_t exit_boot_func(struct efi_boot_memmap *map, void *priv) { const char *signature; @@ -683,14 +651,14 @@ static efi_status_t exit_boot_func(efi_system_table_t *sys_table_arg, : EFI32_LOADER_SIGNATURE; memcpy(&p->efi->efi_loader_signature, signature, sizeof(__u32)); - p->efi->efi_systab = (unsigned long)sys_table_arg; + p->efi->efi_systab = (unsigned long)efi_system_table(); p->efi->efi_memdesc_size = *map->desc_size; p->efi->efi_memdesc_version = *map->desc_ver; p->efi->efi_memmap = (unsigned long)*map->map; p->efi->efi_memmap_size = *map->map_size; #ifdef CONFIG_X86_64 - p->efi->efi_systab_hi = (unsigned long)sys_table_arg >> 32; + p->efi->efi_systab_hi = (unsigned long)efi_system_table() >> 32; p->efi->efi_memmap_hi = (unsigned long)*map->map >> 32; #endif @@ -722,8 +690,7 @@ static efi_status_t exit_boot(struct boot_params *boot_params, void *handle) return status; /* Might as well exit boot services now */ - status = efi_exit_boot_services(sys_table, handle, &map, &priv, - exit_boot_func); + status = efi_exit_boot_services(handle, &map, &priv, exit_boot_func); if (status != EFI_SUCCESS) return status; @@ -741,33 +708,22 @@ static efi_status_t exit_boot(struct boot_params *boot_params, void *handle) * On success we return a pointer to a boot_params structure, and NULL * on failure. */ -struct boot_params * -efi_main(struct efi_config *c, struct boot_params *boot_params) +struct boot_params *efi_main(efi_handle_t handle, + efi_system_table_t *sys_table_arg, + struct boot_params *boot_params) { struct desc_ptr *gdt = NULL; struct setup_header *hdr = &boot_params->hdr; efi_status_t status; struct desc_struct *desc; - void *handle; - efi_system_table_t *_table; unsigned long cmdline_paddr; - efi_early = c; - - _table = (efi_system_table_t *)(unsigned long)efi_early->table; - handle = (void *)(unsigned long)efi_early->image_handle; - - sys_table = _table; + sys_table = sys_table_arg; /* Check if we were booted by the EFI firmware */ if (sys_table->hdr.signature != EFI_SYSTEM_TABLE_SIGNATURE) goto fail; - if (efi_is_64bit()) - setup_boot_services64(efi_early); - else - setup_boot_services32(efi_early); - /* * make_boot_params() may have been called before efi_main(), in which * case this is the second time we parse the cmdline. This is ok, @@ -782,14 +738,14 @@ efi_main(struct efi_config *c, struct boot_params *boot_params) * otherwise we ask the BIOS. */ if (boot_params->secure_boot == efi_secureboot_mode_unset) - boot_params->secure_boot = efi_get_secureboot(sys_table); + boot_params->secure_boot = efi_get_secureboot(); /* Ask the firmware to clear memory on unclean shutdown */ - efi_enable_reset_attack_mitigation(sys_table); + efi_enable_reset_attack_mitigation(); - efi_random_get_seed(sys_table); + efi_random_get_seed(); - efi_retrieve_tpm2_eventlog(sys_table); + efi_retrieve_tpm2_eventlog(); setup_graphics(boot_params); @@ -797,18 +753,17 @@ efi_main(struct efi_config *c, struct boot_params *boot_params) setup_quirks(boot_params); - status = efi_call_early(allocate_pool, EFI_LOADER_DATA, - sizeof(*gdt), (void **)&gdt); + status = efi_bs_call(allocate_pool, EFI_LOADER_DATA, sizeof(*gdt), + (void **)&gdt); if (status != EFI_SUCCESS) { - efi_printk(sys_table, "Failed to allocate memory for 'gdt' structure\n"); + efi_printk("Failed to allocate memory for 'gdt' structure\n"); goto fail; } gdt->size = 0x800; - status = efi_low_alloc(sys_table, gdt->size, 8, - (unsigned long *)&gdt->address); + status = efi_low_alloc(gdt->size, 8, (unsigned long *)&gdt->address); if (status != EFI_SUCCESS) { - efi_printk(sys_table, "Failed to allocate memory for 'gdt'\n"); + efi_printk("Failed to allocate memory for 'gdt'\n"); goto fail; } @@ -818,13 +773,13 @@ efi_main(struct efi_config *c, struct boot_params *boot_params) */ if (hdr->pref_address != hdr->code32_start) { unsigned long bzimage_addr = hdr->code32_start; - status = efi_relocate_kernel(sys_table, &bzimage_addr, + status = efi_relocate_kernel(&bzimage_addr, hdr->init_size, hdr->init_size, hdr->pref_address, hdr->kernel_alignment, LOAD_PHYSICAL_ADDR); if (status != EFI_SUCCESS) { - efi_printk(sys_table, "efi_relocate_kernel() failed!\n"); + efi_printk("efi_relocate_kernel() failed!\n"); goto fail; } @@ -834,7 +789,7 @@ efi_main(struct efi_config *c, struct boot_params *boot_params) status = exit_boot(boot_params, handle); if (status != EFI_SUCCESS) { - efi_printk(sys_table, "exit_boot() failed!\n"); + efi_printk("exit_boot() failed!\n"); goto fail; } @@ -927,7 +882,8 @@ efi_main(struct efi_config *c, struct boot_params *boot_params) return boot_params; fail: - efi_printk(sys_table, "efi_main() failed!\n"); + efi_printk("efi_main() failed!\n"); - return NULL; + for (;;) + asm("hlt"); } diff --git a/arch/x86/boot/compressed/eboot.h b/arch/x86/boot/compressed/eboot.h index 8297387c4676..99f35343d443 100644 --- a/arch/x86/boot/compressed/eboot.h +++ b/arch/x86/boot/compressed/eboot.h @@ -12,22 +12,20 @@ #define DESC_TYPE_CODE_DATA (1 << 0) -typedef struct { - u32 get_mode; - u32 set_mode; - u32 blt; -} efi_uga_draw_protocol_32_t; +typedef union efi_uga_draw_protocol efi_uga_draw_protocol_t; -typedef struct { - u64 get_mode; - u64 set_mode; - u64 blt; -} efi_uga_draw_protocol_64_t; - -typedef struct { - void *get_mode; - void *set_mode; - void *blt; -} efi_uga_draw_protocol_t; +union efi_uga_draw_protocol { + struct { + efi_status_t (__efiapi *get_mode)(efi_uga_draw_protocol_t *, + u32*, u32*, u32*, u32*); + void *set_mode; + void *blt; + }; + struct { + u32 get_mode; + u32 set_mode; + u32 blt; + } mixed_mode; +}; #endif /* BOOT_COMPRESSED_EBOOT_H */ diff --git a/arch/x86/boot/compressed/efi_stub_32.S b/arch/x86/boot/compressed/efi_stub_32.S deleted file mode 100644 index ed6c351d34ed..000000000000 --- a/arch/x86/boot/compressed/efi_stub_32.S +++ /dev/null @@ -1,87 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -/* - * EFI call stub for IA32. - * - * This stub allows us to make EFI calls in physical mode with interrupts - * turned off. Note that this implementation is different from the one in - * arch/x86/platform/efi/efi_stub_32.S because we're _already_ in physical - * mode at this point. - */ - -#include <linux/linkage.h> -#include <asm/page_types.h> - -/* - * efi_call_phys(void *, ...) is a function with variable parameters. - * All the callers of this function assure that all the parameters are 4-bytes. - */ - -/* - * In gcc calling convention, EBX, ESP, EBP, ESI and EDI are all callee save. - * So we'd better save all of them at the beginning of this function and restore - * at the end no matter how many we use, because we can not assure EFI runtime - * service functions will comply with gcc calling convention, too. - */ - -.text -SYM_FUNC_START(efi_call_phys) - /* - * 0. The function can only be called in Linux kernel. So CS has been - * set to 0x0010, DS and SS have been set to 0x0018. In EFI, I found - * the values of these registers are the same. And, the corresponding - * GDT entries are identical. So I will do nothing about segment reg - * and GDT, but change GDT base register in prelog and epilog. - */ - - /* - * 1. Because we haven't been relocated by this point we need to - * use relative addressing. - */ - call 1f -1: popl %edx - subl $1b, %edx - - /* - * 2. Now on the top of stack is the return - * address in the caller of efi_call_phys(), then parameter 1, - * parameter 2, ..., param n. To make things easy, we save the return - * address of efi_call_phys in a global variable. - */ - popl %ecx - movl %ecx, saved_return_addr(%edx) - /* get the function pointer into ECX*/ - popl %ecx - movl %ecx, efi_rt_function_ptr(%edx) - - /* - * 3. Call the physical function. - */ - call *%ecx - - /* - * 4. Balance the stack. And because EAX contain the return value, - * we'd better not clobber it. We need to calculate our address - * again because %ecx and %edx are not preserved across EFI function - * calls. - */ - call 1f -1: popl %edx - subl $1b, %edx - - movl efi_rt_function_ptr(%edx), %ecx - pushl %ecx - - /* - * 10. Push the saved return address onto the stack and return. - */ - movl saved_return_addr(%edx), %ecx - pushl %ecx - ret -SYM_FUNC_END(efi_call_phys) -.previous - -.data -saved_return_addr: - .long 0 -efi_rt_function_ptr: - .long 0 diff --git a/arch/x86/boot/compressed/efi_stub_64.S b/arch/x86/boot/compressed/efi_stub_64.S deleted file mode 100644 index 99494dff2113..000000000000 --- a/arch/x86/boot/compressed/efi_stub_64.S +++ /dev/null @@ -1,5 +0,0 @@ -#include <asm/segment.h> -#include <asm/msr.h> -#include <asm/processor-flags.h> - -#include "../../platform/efi/efi_stub_64.S" diff --git a/arch/x86/boot/compressed/efi_thunk_64.S b/arch/x86/boot/compressed/efi_thunk_64.S index 593913692d16..8fb7f6799c52 100644 --- a/arch/x86/boot/compressed/efi_thunk_64.S +++ b/arch/x86/boot/compressed/efi_thunk_64.S @@ -10,7 +10,7 @@ * needs to be able to service interrupts. * * On the plus side, we don't have to worry about mangling 64-bit - * addresses into 32-bits because we're executing with an identify + * addresses into 32-bits because we're executing with an identity * mapped pagetable and haven't transitioned to 64-bit virtual addresses * yet. */ @@ -23,16 +23,13 @@ .code64 .text -SYM_FUNC_START(efi64_thunk) +SYM_FUNC_START(__efi64_thunk) push %rbp push %rbx - subq $8, %rsp - leaq efi_exit32(%rip), %rax - movl %eax, 4(%rsp) - leaq efi_gdt64(%rip), %rax - movl %eax, (%rsp) - movl %eax, 2(%rax) /* Fixup the gdt base address */ + leaq 1f(%rip), %rbp + leaq efi_gdt64(%rip), %rbx + movl %ebx, 2(%rbx) /* Fixup the gdt base address */ movl %ds, %eax push %rax @@ -48,15 +45,10 @@ SYM_FUNC_START(efi64_thunk) movl %esi, 0x0(%rsp) movl %edx, 0x4(%rsp) movl %ecx, 0x8(%rsp) - movq %r8, %rsi - movl %esi, 0xc(%rsp) - movq %r9, %rsi - movl %esi, 0x10(%rsp) + movl %r8d, 0xc(%rsp) + movl %r9d, 0x10(%rsp) - sgdt save_gdt(%rip) - - leaq 1f(%rip), %rbx - movq %rbx, func_rt_ptr(%rip) + sgdt 0x14(%rsp) /* * Switch to gdt with 32-bit segments. This is the firmware GDT @@ -71,9 +63,9 @@ SYM_FUNC_START(efi64_thunk) pushq %rax lretq -1: addq $32, %rsp - - lgdt save_gdt(%rip) +1: lgdt 0x14(%rsp) + addq $32, %rsp + movq %rdi, %rax pop %rbx movl %ebx, %ss @@ -85,26 +77,13 @@ SYM_FUNC_START(efi64_thunk) /* * Convert 32-bit status code into 64-bit. */ - test %rax, %rax - jz 1f - movl %eax, %ecx - andl $0x0fffffff, %ecx - andl $0xf0000000, %eax - shl $32, %rax - or %rcx, %rax -1: - addq $8, %rsp + roll $1, %eax + rorq $1, %rax + pop %rbx pop %rbp ret -SYM_FUNC_END(efi64_thunk) - -SYM_FUNC_START_LOCAL(efi_exit32) - movq func_rt_ptr(%rip), %rax - push %rax - mov %rdi, %rax - ret -SYM_FUNC_END(efi_exit32) +SYM_FUNC_END(__efi64_thunk) .code32 /* @@ -144,9 +123,7 @@ SYM_FUNC_START_LOCAL(efi_enter32) */ cli - movl 56(%esp), %eax - movl %eax, 2(%eax) - lgdtl (%eax) + lgdtl (%ebx) movl %cr4, %eax btsl $(X86_CR4_PAE_BIT), %eax @@ -163,9 +140,8 @@ SYM_FUNC_START_LOCAL(efi_enter32) xorl %eax, %eax lldt %ax - movl 60(%esp), %eax pushl $__KERNEL_CS - pushl %eax + pushl %ebp /* Enable paging */ movl %cr0, %eax @@ -181,13 +157,6 @@ SYM_DATA_START(efi32_boot_gdt) .quad 0 SYM_DATA_END(efi32_boot_gdt) -SYM_DATA_START_LOCAL(save_gdt) - .word 0 - .quad 0 -SYM_DATA_END(save_gdt) - -SYM_DATA_LOCAL(func_rt_ptr, .quad 0) - SYM_DATA_START(efi_gdt64) .word efi_gdt64_end - efi_gdt64 .long 0 /* Filled out by user */ diff --git a/arch/x86/boot/compressed/head_32.S b/arch/x86/boot/compressed/head_32.S index f2dfd6d083ef..73f17d0544dd 100644 --- a/arch/x86/boot/compressed/head_32.S +++ b/arch/x86/boot/compressed/head_32.S @@ -145,67 +145,16 @@ SYM_FUNC_START(startup_32) SYM_FUNC_END(startup_32) #ifdef CONFIG_EFI_STUB -/* - * We don't need the return address, so set up the stack so efi_main() can find - * its arguments. - */ -SYM_FUNC_START(efi_pe_entry) - add $0x4, %esp - - call 1f -1: popl %esi - subl $1b, %esi - - popl %ecx - movl %ecx, efi32_config(%esi) /* Handle */ - popl %ecx - movl %ecx, efi32_config+8(%esi) /* EFI System table pointer */ - - /* Relocate efi_config->call() */ - leal efi32_config(%esi), %eax - add %esi, 40(%eax) - pushl %eax - - call make_boot_params - cmpl $0, %eax - je fail - movl %esi, BP_code32_start(%eax) - popl %ecx - pushl %eax - pushl %ecx - jmp 2f /* Skip efi_config initialization */ -SYM_FUNC_END(efi_pe_entry) - SYM_FUNC_START(efi32_stub_entry) +SYM_FUNC_START_ALIAS(efi_stub_entry) add $0x4, %esp - popl %ecx - popl %edx - - call 1f -1: popl %esi - subl $1b, %esi - - movl %ecx, efi32_config(%esi) /* Handle */ - movl %edx, efi32_config+8(%esi) /* EFI System table pointer */ - - /* Relocate efi_config->call() */ - leal efi32_config(%esi), %eax - add %esi, 40(%eax) - pushl %eax -2: call efi_main - cmpl $0, %eax movl %eax, %esi - jne 2f -fail: - /* EFI init failed, so hang. */ - hlt - jmp fail -2: movl BP_code32_start(%esi), %eax leal startup_32(%eax), %eax jmp *%eax SYM_FUNC_END(efi32_stub_entry) +SYM_FUNC_END_ALIAS(efi_stub_entry) #endif .text @@ -240,11 +189,9 @@ SYM_FUNC_START_LOCAL_NOALIGN(.Lrelocated) /* push arguments for extract_kernel: */ pushl $z_output_len /* decompressed length, end of relocs */ - movl BP_init_size(%esi), %eax - subl $_end, %eax - movl %ebx, %ebp - subl %eax, %ebp - pushl %ebp /* output address */ + leal _end(%ebx), %eax + subl BP_init_size(%esi), %eax + pushl %eax /* output address */ pushl $z_input_len /* input_len */ leal input_data(%ebx), %eax @@ -262,15 +209,6 @@ SYM_FUNC_START_LOCAL_NOALIGN(.Lrelocated) jmp *%eax SYM_FUNC_END(.Lrelocated) -#ifdef CONFIG_EFI_STUB - .data -efi32_config: - .fill 5,8,0 - .long efi_call_phys - .long 0 - .byte 0 -#endif - /* * Stack and heap for uncompression */ diff --git a/arch/x86/boot/compressed/head_64.S b/arch/x86/boot/compressed/head_64.S index ee60b81944a7..1f1f6c8139b3 100644 --- a/arch/x86/boot/compressed/head_64.S +++ b/arch/x86/boot/compressed/head_64.S @@ -208,10 +208,12 @@ SYM_FUNC_START(startup_32) pushl $__KERNEL_CS leal startup_64(%ebp), %eax #ifdef CONFIG_EFI_MIXED - movl efi32_config(%ebp), %ebx - cmp $0, %ebx + movl efi32_boot_args(%ebp), %edi + cmp $0, %edi jz 1f - leal handover_entry(%ebp), %eax + leal efi64_stub_entry(%ebp), %eax + movl %esi, %edx + movl efi32_boot_args+4(%ebp), %esi 1: #endif pushl %eax @@ -232,17 +234,14 @@ SYM_FUNC_START(efi32_stub_entry) popl %edx popl %esi - leal (BP_scratch+4)(%esi), %esp call 1f 1: pop %ebp subl $1b, %ebp - movl %ecx, efi32_config(%ebp) - movl %edx, efi32_config+8(%ebp) + movl %ecx, efi32_boot_args(%ebp) + movl %edx, efi32_boot_args+4(%ebp) sgdtl efi32_boot_gdt(%ebp) - - leal efi32_config(%ebp), %eax - movl %eax, efi_config(%ebp) + movb $0, efi_is64(%ebp) /* Disable paging */ movl %cr0, %eax @@ -450,70 +449,17 @@ trampoline_return: SYM_CODE_END(startup_64) #ifdef CONFIG_EFI_STUB - -/* The entry point for the PE/COFF executable is efi_pe_entry. */ -SYM_FUNC_START(efi_pe_entry) - movq %rcx, efi64_config(%rip) /* Handle */ - movq %rdx, efi64_config+8(%rip) /* EFI System table pointer */ - - leaq efi64_config(%rip), %rax - movq %rax, efi_config(%rip) - - call 1f -1: popq %rbp - subq $1b, %rbp - - /* - * Relocate efi_config->call(). - */ - addq %rbp, efi64_config+40(%rip) - - movq %rax, %rdi - call make_boot_params - cmpq $0,%rax - je fail - mov %rax, %rsi - leaq startup_32(%rip), %rax - movl %eax, BP_code32_start(%rsi) - jmp 2f /* Skip the relocation */ - -handover_entry: - call 1f -1: popq %rbp - subq $1b, %rbp - - /* - * Relocate efi_config->call(). - */ - movq efi_config(%rip), %rax - addq %rbp, 40(%rax) -2: - movq efi_config(%rip), %rdi + .org 0x390 +SYM_FUNC_START(efi64_stub_entry) +SYM_FUNC_START_ALIAS(efi_stub_entry) + and $~0xf, %rsp /* realign the stack */ call efi_main movq %rax,%rsi - cmpq $0,%rax - jne 2f -fail: - /* EFI init failed, so hang. */ - hlt - jmp fail -2: movl BP_code32_start(%esi), %eax leaq startup_64(%rax), %rax jmp *%rax -SYM_FUNC_END(efi_pe_entry) - - .org 0x390 -SYM_FUNC_START(efi64_stub_entry) - movq %rdi, efi64_config(%rip) /* Handle */ - movq %rsi, efi64_config+8(%rip) /* EFI System table pointer */ - - leaq efi64_config(%rip), %rax - movq %rax, efi_config(%rip) - - movq %rdx, %rsi - jmp handover_entry SYM_FUNC_END(efi64_stub_entry) +SYM_FUNC_END_ALIAS(efi_stub_entry) #endif .text @@ -682,24 +628,11 @@ SYM_DATA_START_LOCAL(gdt) .quad 0x0000000000000000 /* TS continued */ SYM_DATA_END_LABEL(gdt, SYM_L_LOCAL, gdt_end) -#ifdef CONFIG_EFI_STUB -SYM_DATA_LOCAL(efi_config, .quad 0) - #ifdef CONFIG_EFI_MIXED -SYM_DATA_START(efi32_config) - .fill 5,8,0 - .quad efi64_thunk - .byte 0 -SYM_DATA_END(efi32_config) +SYM_DATA_LOCAL(efi32_boot_args, .long 0, 0) +SYM_DATA(efi_is64, .byte 1) #endif -SYM_DATA_START(efi64_config) - .fill 5,8,0 - .quad efi_call - .byte 1 -SYM_DATA_END(efi64_config) -#endif /* CONFIG_EFI_STUB */ - /* * Stack and heap for uncompression */ diff --git a/arch/x86/boot/mkcpustr.c b/arch/x86/boot/mkcpustr.c index 9caa10e82217..da0ccc5de538 100644 --- a/arch/x86/boot/mkcpustr.c +++ b/arch/x86/boot/mkcpustr.c @@ -15,6 +15,7 @@ #include "../include/asm/required-features.h" #include "../include/asm/disabled-features.h" #include "../include/asm/cpufeatures.h" +#include "../include/asm/vmxfeatures.h" #include "../kernel/cpu/capflags.c" int main(void) diff --git a/arch/x86/boot/setup.ld b/arch/x86/boot/setup.ld index 0149e41d42c2..3da1c37c6dd5 100644 --- a/arch/x86/boot/setup.ld +++ b/arch/x86/boot/setup.ld @@ -51,7 +51,10 @@ SECTIONS . = ALIGN(16); _end = .; - /DISCARD/ : { *(.note*) } + /DISCARD/ : { + *(.eh_frame) + *(.note*) + } /* * The ASSERT() sink to . is intentional, for binutils 2.14 compatibility: diff --git a/arch/x86/crypto/.gitignore b/arch/x86/crypto/.gitignore new file mode 100644 index 000000000000..30be0400a439 --- /dev/null +++ b/arch/x86/crypto/.gitignore @@ -0,0 +1 @@ +poly1305-x86_64-cryptogams.S diff --git a/arch/x86/crypto/Makefile b/arch/x86/crypto/Makefile index 958440eae27e..b69e00bf20b8 100644 --- a/arch/x86/crypto/Makefile +++ b/arch/x86/crypto/Makefile @@ -73,6 +73,10 @@ aegis128-aesni-y := aegis128-aesni-asm.o aegis128-aesni-glue.o nhpoly1305-sse2-y := nh-sse2-x86_64.o nhpoly1305-sse2-glue.o blake2s-x86_64-y := blake2s-core.o blake2s-glue.o +poly1305-x86_64-y := poly1305-x86_64-cryptogams.o poly1305_glue.o +ifneq ($(CONFIG_CRYPTO_POLY1305_X86_64),) +targets += poly1305-x86_64-cryptogams.S +endif ifeq ($(avx_supported),yes) camellia-aesni-avx-x86_64-y := camellia-aesni-avx-asm_64.o \ @@ -101,10 +105,8 @@ aesni-intel-y := aesni-intel_asm.o aesni-intel_glue.o aesni-intel-$(CONFIG_64BIT) += aesni-intel_avx-x86_64.o aes_ctrby8_avx-x86_64.o ghash-clmulni-intel-y := ghash-clmulni-intel_asm.o ghash-clmulni-intel_glue.o sha1-ssse3-y := sha1_ssse3_asm.o sha1_ssse3_glue.o -poly1305-x86_64-y := poly1305-sse2-x86_64.o poly1305_glue.o ifeq ($(avx2_supported),yes) sha1-ssse3-y += sha1_avx2_x86_64_asm.o -poly1305-x86_64-y += poly1305-avx2-x86_64.o endif ifeq ($(sha1_ni_supported),yes) sha1-ssse3-y += sha1_ni_asm.o @@ -118,3 +120,8 @@ sha256-ssse3-y += sha256_ni_asm.o endif sha512-ssse3-y := sha512-ssse3-asm.o sha512-avx-asm.o sha512-avx2-asm.o sha512_ssse3_glue.o crct10dif-pclmul-y := crct10dif-pcl-asm_64.o crct10dif-pclmul_glue.o + +quiet_cmd_perlasm = PERLASM $@ + cmd_perlasm = $(PERL) $< > $@ +$(obj)/%.S: $(src)/%.pl FORCE + $(call if_changed,perlasm) diff --git a/arch/x86/crypto/aegis128-aesni-glue.c b/arch/x86/crypto/aegis128-aesni-glue.c index 46d227122643..4623189000d8 100644 --- a/arch/x86/crypto/aegis128-aesni-glue.c +++ b/arch/x86/crypto/aegis128-aesni-glue.c @@ -144,10 +144,8 @@ static int crypto_aegis128_aesni_setkey(struct crypto_aead *aead, const u8 *key, { struct aegis_ctx *ctx = crypto_aegis128_aesni_ctx(aead); - if (keylen != AEGIS128_KEY_SIZE) { - crypto_aead_set_flags(aead, CRYPTO_TFM_RES_BAD_KEY_LEN); + if (keylen != AEGIS128_KEY_SIZE) return -EINVAL; - } memcpy(ctx->key.bytes, key, AEGIS128_KEY_SIZE); diff --git a/arch/x86/crypto/aesni-intel_asm.S b/arch/x86/crypto/aesni-intel_asm.S index d28503f99f58..cad6e1bfa7d5 100644 --- a/arch/x86/crypto/aesni-intel_asm.S +++ b/arch/x86/crypto/aesni-intel_asm.S @@ -1942,7 +1942,7 @@ SYM_FUNC_START(aesni_set_key) SYM_FUNC_END(aesni_set_key) /* - * void aesni_enc(struct crypto_aes_ctx *ctx, u8 *dst, const u8 *src) + * void aesni_enc(const void *ctx, u8 *dst, const u8 *src) */ SYM_FUNC_START(aesni_enc) FRAME_BEGIN @@ -2131,7 +2131,7 @@ SYM_FUNC_START_LOCAL(_aesni_enc4) SYM_FUNC_END(_aesni_enc4) /* - * void aesni_dec (struct crypto_aes_ctx *ctx, u8 *dst, const u8 *src) + * void aesni_dec (const void *ctx, u8 *dst, const u8 *src) */ SYM_FUNC_START(aesni_dec) FRAME_BEGIN @@ -2716,8 +2716,8 @@ SYM_FUNC_END(aesni_ctr_enc) pxor CTR, IV; /* - * void aesni_xts_crypt8(struct crypto_aes_ctx *ctx, const u8 *dst, u8 *src, - * bool enc, u8 *iv) + * void aesni_xts_crypt8(const struct crypto_aes_ctx *ctx, u8 *dst, + * const u8 *src, bool enc, le128 *iv) */ SYM_FUNC_START(aesni_xts_crypt8) FRAME_BEGIN diff --git a/arch/x86/crypto/aesni-intel_glue.c b/arch/x86/crypto/aesni-intel_glue.c index 3e707e81afdb..bbbebbd35b5d 100644 --- a/arch/x86/crypto/aesni-intel_glue.c +++ b/arch/x86/crypto/aesni-intel_glue.c @@ -83,10 +83,8 @@ struct gcm_context_data { asmlinkage int aesni_set_key(struct crypto_aes_ctx *ctx, const u8 *in_key, unsigned int key_len); -asmlinkage void aesni_enc(struct crypto_aes_ctx *ctx, u8 *out, - const u8 *in); -asmlinkage void aesni_dec(struct crypto_aes_ctx *ctx, u8 *out, - const u8 *in); +asmlinkage void aesni_enc(const void *ctx, u8 *out, const u8 *in); +asmlinkage void aesni_dec(const void *ctx, u8 *out, const u8 *in); asmlinkage void aesni_ecb_enc(struct crypto_aes_ctx *ctx, u8 *out, const u8 *in, unsigned int len); asmlinkage void aesni_ecb_dec(struct crypto_aes_ctx *ctx, u8 *out, @@ -106,8 +104,8 @@ static void (*aesni_ctr_enc_tfm)(struct crypto_aes_ctx *ctx, u8 *out, asmlinkage void aesni_ctr_enc(struct crypto_aes_ctx *ctx, u8 *out, const u8 *in, unsigned int len, u8 *iv); -asmlinkage void aesni_xts_crypt8(struct crypto_aes_ctx *ctx, u8 *out, - const u8 *in, bool enc, u8 *iv); +asmlinkage void aesni_xts_crypt8(const struct crypto_aes_ctx *ctx, u8 *out, + const u8 *in, bool enc, le128 *iv); /* asmlinkage void aesni_gcm_enc() * void *ctx, AES Key schedule. Starts on a 16 byte boundary. @@ -318,14 +316,11 @@ static int aes_set_key_common(struct crypto_tfm *tfm, void *raw_ctx, const u8 *in_key, unsigned int key_len) { struct crypto_aes_ctx *ctx = aes_ctx(raw_ctx); - u32 *flags = &tfm->crt_flags; int err; if (key_len != AES_KEYSIZE_128 && key_len != AES_KEYSIZE_192 && - key_len != AES_KEYSIZE_256) { - *flags |= CRYPTO_TFM_RES_BAD_KEY_LEN; + key_len != AES_KEYSIZE_256) return -EINVAL; - } if (!crypto_simd_usable()) err = aes_expandkey(ctx, in_key, key_len); @@ -550,29 +545,24 @@ static int xts_aesni_setkey(struct crypto_skcipher *tfm, const u8 *key, } -static void aesni_xts_tweak(void *ctx, u8 *out, const u8 *in) -{ - aesni_enc(ctx, out, in); -} - -static void aesni_xts_enc(void *ctx, u128 *dst, const u128 *src, le128 *iv) +static void aesni_xts_enc(const void *ctx, u8 *dst, const u8 *src, le128 *iv) { - glue_xts_crypt_128bit_one(ctx, dst, src, iv, GLUE_FUNC_CAST(aesni_enc)); + glue_xts_crypt_128bit_one(ctx, dst, src, iv, aesni_enc); } -static void aesni_xts_dec(void *ctx, u128 *dst, const u128 *src, le128 *iv) +static void aesni_xts_dec(const void *ctx, u8 *dst, const u8 *src, le128 *iv) { - glue_xts_crypt_128bit_one(ctx, dst, src, iv, GLUE_FUNC_CAST(aesni_dec)); + glue_xts_crypt_128bit_one(ctx, dst, src, iv, aesni_dec); } -static void aesni_xts_enc8(void *ctx, u128 *dst, const u128 *src, le128 *iv) +static void aesni_xts_enc8(const void *ctx, u8 *dst, const u8 *src, le128 *iv) { - aesni_xts_crypt8(ctx, (u8 *)dst, (const u8 *)src, true, (u8 *)iv); + aesni_xts_crypt8(ctx, dst, src, true, iv); } -static void aesni_xts_dec8(void *ctx, u128 *dst, const u128 *src, le128 *iv) +static void aesni_xts_dec8(const void *ctx, u8 *dst, const u8 *src, le128 *iv) { - aesni_xts_crypt8(ctx, (u8 *)dst, (const u8 *)src, false, (u8 *)iv); + aesni_xts_crypt8(ctx, dst, src, false, iv); } static const struct common_glue_ctx aesni_enc_xts = { @@ -581,10 +571,10 @@ static const struct common_glue_ctx aesni_enc_xts = { .funcs = { { .num_blocks = 8, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(aesni_xts_enc8) } + .fn_u = { .xts = aesni_xts_enc8 } }, { .num_blocks = 1, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(aesni_xts_enc) } + .fn_u = { .xts = aesni_xts_enc } } } }; @@ -594,10 +584,10 @@ static const struct common_glue_ctx aesni_dec_xts = { .funcs = { { .num_blocks = 8, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(aesni_xts_dec8) } + .fn_u = { .xts = aesni_xts_dec8 } }, { .num_blocks = 1, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(aesni_xts_dec) } + .fn_u = { .xts = aesni_xts_dec } } } }; @@ -606,8 +596,7 @@ static int xts_encrypt(struct skcipher_request *req) struct crypto_skcipher *tfm = crypto_skcipher_reqtfm(req); struct aesni_xts_ctx *ctx = crypto_skcipher_ctx(tfm); - return glue_xts_req_128bit(&aesni_enc_xts, req, - XTS_TWEAK_CAST(aesni_xts_tweak), + return glue_xts_req_128bit(&aesni_enc_xts, req, aesni_enc, aes_ctx(ctx->raw_tweak_ctx), aes_ctx(ctx->raw_crypt_ctx), false); @@ -618,8 +607,7 @@ static int xts_decrypt(struct skcipher_request *req) struct crypto_skcipher *tfm = crypto_skcipher_reqtfm(req); struct aesni_xts_ctx *ctx = crypto_skcipher_ctx(tfm); - return glue_xts_req_128bit(&aesni_dec_xts, req, - XTS_TWEAK_CAST(aesni_xts_tweak), + return glue_xts_req_128bit(&aesni_dec_xts, req, aesni_enc, aes_ctx(ctx->raw_tweak_ctx), aes_ctx(ctx->raw_crypt_ctx), true); @@ -650,10 +638,9 @@ static int common_rfc4106_set_key(struct crypto_aead *aead, const u8 *key, { struct aesni_rfc4106_gcm_ctx *ctx = aesni_rfc4106_gcm_ctx_get(aead); - if (key_len < 4) { - crypto_aead_set_flags(aead, CRYPTO_TFM_RES_BAD_KEY_LEN); + if (key_len < 4) return -EINVAL; - } + /*Account for 4 byte nonce at the end.*/ key_len -= 4; diff --git a/arch/x86/crypto/blake2s-glue.c b/arch/x86/crypto/blake2s-glue.c index 1d9ff8a45e1f..06ef2d4a4701 100644 --- a/arch/x86/crypto/blake2s-glue.c +++ b/arch/x86/crypto/blake2s-glue.c @@ -64,10 +64,8 @@ static int crypto_blake2s_setkey(struct crypto_shash *tfm, const u8 *key, { struct blake2s_tfm_ctx *tctx = crypto_shash_ctx(tfm); - if (keylen == 0 || keylen > BLAKE2S_KEY_SIZE) { - crypto_shash_set_flags(tfm, CRYPTO_TFM_RES_BAD_KEY_LEN); + if (keylen == 0 || keylen > BLAKE2S_KEY_SIZE) return -EINVAL; - } memcpy(tctx->key, key, keylen); tctx->keylen = keylen; diff --git a/arch/x86/crypto/camellia_aesni_avx2_glue.c b/arch/x86/crypto/camellia_aesni_avx2_glue.c index a4f00128ea55..ccda647422d6 100644 --- a/arch/x86/crypto/camellia_aesni_avx2_glue.c +++ b/arch/x86/crypto/camellia_aesni_avx2_glue.c @@ -19,20 +19,17 @@ #define CAMELLIA_AESNI_AVX2_PARALLEL_BLOCKS 32 /* 32-way AVX2/AES-NI parallel cipher functions */ -asmlinkage void camellia_ecb_enc_32way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src); -asmlinkage void camellia_ecb_dec_32way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src); +asmlinkage void camellia_ecb_enc_32way(const void *ctx, u8 *dst, const u8 *src); +asmlinkage void camellia_ecb_dec_32way(const void *ctx, u8 *dst, const u8 *src); -asmlinkage void camellia_cbc_dec_32way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src); -asmlinkage void camellia_ctr_32way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); +asmlinkage void camellia_cbc_dec_32way(const void *ctx, u8 *dst, const u8 *src); +asmlinkage void camellia_ctr_32way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); -asmlinkage void camellia_xts_enc_32way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); -asmlinkage void camellia_xts_dec_32way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); +asmlinkage void camellia_xts_enc_32way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); +asmlinkage void camellia_xts_dec_32way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); static const struct common_glue_ctx camellia_enc = { .num_funcs = 4, @@ -40,16 +37,16 @@ static const struct common_glue_ctx camellia_enc = { .funcs = { { .num_blocks = CAMELLIA_AESNI_AVX2_PARALLEL_BLOCKS, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_ecb_enc_32way) } + .fn_u = { .ecb = camellia_ecb_enc_32way } }, { .num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_ecb_enc_16way) } + .fn_u = { .ecb = camellia_ecb_enc_16way } }, { .num_blocks = 2, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_enc_blk_2way) } + .fn_u = { .ecb = camellia_enc_blk_2way } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_enc_blk) } + .fn_u = { .ecb = camellia_enc_blk } } } }; @@ -59,16 +56,16 @@ static const struct common_glue_ctx camellia_ctr = { .funcs = { { .num_blocks = CAMELLIA_AESNI_AVX2_PARALLEL_BLOCKS, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(camellia_ctr_32way) } + .fn_u = { .ctr = camellia_ctr_32way } }, { .num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(camellia_ctr_16way) } + .fn_u = { .ctr = camellia_ctr_16way } }, { .num_blocks = 2, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(camellia_crypt_ctr_2way) } + .fn_u = { .ctr = camellia_crypt_ctr_2way } }, { .num_blocks = 1, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(camellia_crypt_ctr) } + .fn_u = { .ctr = camellia_crypt_ctr } } } }; @@ -78,13 +75,13 @@ static const struct common_glue_ctx camellia_enc_xts = { .funcs = { { .num_blocks = CAMELLIA_AESNI_AVX2_PARALLEL_BLOCKS, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(camellia_xts_enc_32way) } + .fn_u = { .xts = camellia_xts_enc_32way } }, { .num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(camellia_xts_enc_16way) } + .fn_u = { .xts = camellia_xts_enc_16way } }, { .num_blocks = 1, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(camellia_xts_enc) } + .fn_u = { .xts = camellia_xts_enc } } } }; @@ -94,16 +91,16 @@ static const struct common_glue_ctx camellia_dec = { .funcs = { { .num_blocks = CAMELLIA_AESNI_AVX2_PARALLEL_BLOCKS, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_ecb_dec_32way) } + .fn_u = { .ecb = camellia_ecb_dec_32way } }, { .num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_ecb_dec_16way) } + .fn_u = { .ecb = camellia_ecb_dec_16way } }, { .num_blocks = 2, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_dec_blk_2way) } + .fn_u = { .ecb = camellia_dec_blk_2way } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_dec_blk) } + .fn_u = { .ecb = camellia_dec_blk } } } }; @@ -113,16 +110,16 @@ static const struct common_glue_ctx camellia_dec_cbc = { .funcs = { { .num_blocks = CAMELLIA_AESNI_AVX2_PARALLEL_BLOCKS, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(camellia_cbc_dec_32way) } + .fn_u = { .cbc = camellia_cbc_dec_32way } }, { .num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(camellia_cbc_dec_16way) } + .fn_u = { .cbc = camellia_cbc_dec_16way } }, { .num_blocks = 2, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(camellia_decrypt_cbc_2way) } + .fn_u = { .cbc = camellia_decrypt_cbc_2way } }, { .num_blocks = 1, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(camellia_dec_blk) } + .fn_u = { .cbc = camellia_dec_blk } } } }; @@ -132,21 +129,20 @@ static const struct common_glue_ctx camellia_dec_xts = { .funcs = { { .num_blocks = CAMELLIA_AESNI_AVX2_PARALLEL_BLOCKS, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(camellia_xts_dec_32way) } + .fn_u = { .xts = camellia_xts_dec_32way } }, { .num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(camellia_xts_dec_16way) } + .fn_u = { .xts = camellia_xts_dec_16way } }, { .num_blocks = 1, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(camellia_xts_dec) } + .fn_u = { .xts = camellia_xts_dec } } } }; static int camellia_setkey(struct crypto_skcipher *tfm, const u8 *key, unsigned int keylen) { - return __camellia_setkey(crypto_skcipher_ctx(tfm), key, keylen, - &tfm->base.crt_flags); + return __camellia_setkey(crypto_skcipher_ctx(tfm), key, keylen); } static int ecb_encrypt(struct skcipher_request *req) @@ -161,8 +157,7 @@ static int ecb_decrypt(struct skcipher_request *req) static int cbc_encrypt(struct skcipher_request *req) { - return glue_cbc_encrypt_req_128bit(GLUE_FUNC_CAST(camellia_enc_blk), - req); + return glue_cbc_encrypt_req_128bit(camellia_enc_blk, req); } static int cbc_decrypt(struct skcipher_request *req) @@ -180,8 +175,7 @@ static int xts_encrypt(struct skcipher_request *req) struct crypto_skcipher *tfm = crypto_skcipher_reqtfm(req); struct camellia_xts_ctx *ctx = crypto_skcipher_ctx(tfm); - return glue_xts_req_128bit(&camellia_enc_xts, req, - XTS_TWEAK_CAST(camellia_enc_blk), + return glue_xts_req_128bit(&camellia_enc_xts, req, camellia_enc_blk, &ctx->tweak_ctx, &ctx->crypt_ctx, false); } @@ -190,8 +184,7 @@ static int xts_decrypt(struct skcipher_request *req) struct crypto_skcipher *tfm = crypto_skcipher_reqtfm(req); struct camellia_xts_ctx *ctx = crypto_skcipher_ctx(tfm); - return glue_xts_req_128bit(&camellia_dec_xts, req, - XTS_TWEAK_CAST(camellia_enc_blk), + return glue_xts_req_128bit(&camellia_dec_xts, req, camellia_enc_blk, &ctx->tweak_ctx, &ctx->crypt_ctx, true); } diff --git a/arch/x86/crypto/camellia_aesni_avx_glue.c b/arch/x86/crypto/camellia_aesni_avx_glue.c index f28d282779b8..4e5de6ef206e 100644 --- a/arch/x86/crypto/camellia_aesni_avx_glue.c +++ b/arch/x86/crypto/camellia_aesni_avx_glue.c @@ -18,41 +18,36 @@ #define CAMELLIA_AESNI_PARALLEL_BLOCKS 16 /* 16-way parallel cipher functions (avx/aes-ni) */ -asmlinkage void camellia_ecb_enc_16way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src); +asmlinkage void camellia_ecb_enc_16way(const void *ctx, u8 *dst, const u8 *src); EXPORT_SYMBOL_GPL(camellia_ecb_enc_16way); -asmlinkage void camellia_ecb_dec_16way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src); +asmlinkage void camellia_ecb_dec_16way(const void *ctx, u8 *dst, const u8 *src); EXPORT_SYMBOL_GPL(camellia_ecb_dec_16way); -asmlinkage void camellia_cbc_dec_16way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src); +asmlinkage void camellia_cbc_dec_16way(const void *ctx, u8 *dst, const u8 *src); EXPORT_SYMBOL_GPL(camellia_cbc_dec_16way); -asmlinkage void camellia_ctr_16way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); +asmlinkage void camellia_ctr_16way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); EXPORT_SYMBOL_GPL(camellia_ctr_16way); -asmlinkage void camellia_xts_enc_16way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); +asmlinkage void camellia_xts_enc_16way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); EXPORT_SYMBOL_GPL(camellia_xts_enc_16way); -asmlinkage void camellia_xts_dec_16way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); +asmlinkage void camellia_xts_dec_16way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); EXPORT_SYMBOL_GPL(camellia_xts_dec_16way); -void camellia_xts_enc(void *ctx, u128 *dst, const u128 *src, le128 *iv) +void camellia_xts_enc(const void *ctx, u8 *dst, const u8 *src, le128 *iv) { - glue_xts_crypt_128bit_one(ctx, dst, src, iv, - GLUE_FUNC_CAST(camellia_enc_blk)); + glue_xts_crypt_128bit_one(ctx, dst, src, iv, camellia_enc_blk); } EXPORT_SYMBOL_GPL(camellia_xts_enc); -void camellia_xts_dec(void *ctx, u128 *dst, const u128 *src, le128 *iv) +void camellia_xts_dec(const void *ctx, u8 *dst, const u8 *src, le128 *iv) { - glue_xts_crypt_128bit_one(ctx, dst, src, iv, - GLUE_FUNC_CAST(camellia_dec_blk)); + glue_xts_crypt_128bit_one(ctx, dst, src, iv, camellia_dec_blk); } EXPORT_SYMBOL_GPL(camellia_xts_dec); @@ -62,13 +57,13 @@ static const struct common_glue_ctx camellia_enc = { .funcs = { { .num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_ecb_enc_16way) } + .fn_u = { .ecb = camellia_ecb_enc_16way } }, { .num_blocks = 2, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_enc_blk_2way) } + .fn_u = { .ecb = camellia_enc_blk_2way } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_enc_blk) } + .fn_u = { .ecb = camellia_enc_blk } } } }; @@ -78,13 +73,13 @@ static const struct common_glue_ctx camellia_ctr = { .funcs = { { .num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(camellia_ctr_16way) } + .fn_u = { .ctr = camellia_ctr_16way } }, { .num_blocks = 2, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(camellia_crypt_ctr_2way) } + .fn_u = { .ctr = camellia_crypt_ctr_2way } }, { .num_blocks = 1, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(camellia_crypt_ctr) } + .fn_u = { .ctr = camellia_crypt_ctr } } } }; @@ -94,10 +89,10 @@ static const struct common_glue_ctx camellia_enc_xts = { .funcs = { { .num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(camellia_xts_enc_16way) } + .fn_u = { .xts = camellia_xts_enc_16way } }, { .num_blocks = 1, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(camellia_xts_enc) } + .fn_u = { .xts = camellia_xts_enc } } } }; @@ -107,13 +102,13 @@ static const struct common_glue_ctx camellia_dec = { .funcs = { { .num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_ecb_dec_16way) } + .fn_u = { .ecb = camellia_ecb_dec_16way } }, { .num_blocks = 2, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_dec_blk_2way) } + .fn_u = { .ecb = camellia_dec_blk_2way } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_dec_blk) } + .fn_u = { .ecb = camellia_dec_blk } } } }; @@ -123,13 +118,13 @@ static const struct common_glue_ctx camellia_dec_cbc = { .funcs = { { .num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(camellia_cbc_dec_16way) } + .fn_u = { .cbc = camellia_cbc_dec_16way } }, { .num_blocks = 2, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(camellia_decrypt_cbc_2way) } + .fn_u = { .cbc = camellia_decrypt_cbc_2way } }, { .num_blocks = 1, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(camellia_dec_blk) } + .fn_u = { .cbc = camellia_dec_blk } } } }; @@ -139,18 +134,17 @@ static const struct common_glue_ctx camellia_dec_xts = { .funcs = { { .num_blocks = CAMELLIA_AESNI_PARALLEL_BLOCKS, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(camellia_xts_dec_16way) } + .fn_u = { .xts = camellia_xts_dec_16way } }, { .num_blocks = 1, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(camellia_xts_dec) } + .fn_u = { .xts = camellia_xts_dec } } } }; static int camellia_setkey(struct crypto_skcipher *tfm, const u8 *key, unsigned int keylen) { - return __camellia_setkey(crypto_skcipher_ctx(tfm), key, keylen, - &tfm->base.crt_flags); + return __camellia_setkey(crypto_skcipher_ctx(tfm), key, keylen); } static int ecb_encrypt(struct skcipher_request *req) @@ -165,8 +159,7 @@ static int ecb_decrypt(struct skcipher_request *req) static int cbc_encrypt(struct skcipher_request *req) { - return glue_cbc_encrypt_req_128bit(GLUE_FUNC_CAST(camellia_enc_blk), - req); + return glue_cbc_encrypt_req_128bit(camellia_enc_blk, req); } static int cbc_decrypt(struct skcipher_request *req) @@ -183,7 +176,6 @@ int xts_camellia_setkey(struct crypto_skcipher *tfm, const u8 *key, unsigned int keylen) { struct camellia_xts_ctx *ctx = crypto_skcipher_ctx(tfm); - u32 *flags = &tfm->base.crt_flags; int err; err = xts_verify_key(tfm, key, keylen); @@ -191,13 +183,12 @@ int xts_camellia_setkey(struct crypto_skcipher *tfm, const u8 *key, return err; /* first half of xts-key is for crypt */ - err = __camellia_setkey(&ctx->crypt_ctx, key, keylen / 2, flags); + err = __camellia_setkey(&ctx->crypt_ctx, key, keylen / 2); if (err) return err; /* second half of xts-key is for tweak */ - return __camellia_setkey(&ctx->tweak_ctx, key + keylen / 2, keylen / 2, - flags); + return __camellia_setkey(&ctx->tweak_ctx, key + keylen / 2, keylen / 2); } EXPORT_SYMBOL_GPL(xts_camellia_setkey); @@ -206,8 +197,7 @@ static int xts_encrypt(struct skcipher_request *req) struct crypto_skcipher *tfm = crypto_skcipher_reqtfm(req); struct camellia_xts_ctx *ctx = crypto_skcipher_ctx(tfm); - return glue_xts_req_128bit(&camellia_enc_xts, req, - XTS_TWEAK_CAST(camellia_enc_blk), + return glue_xts_req_128bit(&camellia_enc_xts, req, camellia_enc_blk, &ctx->tweak_ctx, &ctx->crypt_ctx, false); } @@ -216,8 +206,7 @@ static int xts_decrypt(struct skcipher_request *req) struct crypto_skcipher *tfm = crypto_skcipher_reqtfm(req); struct camellia_xts_ctx *ctx = crypto_skcipher_ctx(tfm); - return glue_xts_req_128bit(&camellia_dec_xts, req, - XTS_TWEAK_CAST(camellia_enc_blk), + return glue_xts_req_128bit(&camellia_dec_xts, req, camellia_enc_blk, &ctx->tweak_ctx, &ctx->crypt_ctx, true); } diff --git a/arch/x86/crypto/camellia_glue.c b/arch/x86/crypto/camellia_glue.c index 7c62db56ffe1..242c056e5fa8 100644 --- a/arch/x86/crypto/camellia_glue.c +++ b/arch/x86/crypto/camellia_glue.c @@ -18,19 +18,17 @@ #include <asm/crypto/glue_helper.h> /* regular block cipher functions */ -asmlinkage void __camellia_enc_blk(struct camellia_ctx *ctx, u8 *dst, - const u8 *src, bool xor); +asmlinkage void __camellia_enc_blk(const void *ctx, u8 *dst, const u8 *src, + bool xor); EXPORT_SYMBOL_GPL(__camellia_enc_blk); -asmlinkage void camellia_dec_blk(struct camellia_ctx *ctx, u8 *dst, - const u8 *src); +asmlinkage void camellia_dec_blk(const void *ctx, u8 *dst, const u8 *src); EXPORT_SYMBOL_GPL(camellia_dec_blk); /* 2-way parallel cipher functions */ -asmlinkage void __camellia_enc_blk_2way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src, bool xor); +asmlinkage void __camellia_enc_blk_2way(const void *ctx, u8 *dst, const u8 *src, + bool xor); EXPORT_SYMBOL_GPL(__camellia_enc_blk_2way); -asmlinkage void camellia_dec_blk_2way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src); +asmlinkage void camellia_dec_blk_2way(const void *ctx, u8 *dst, const u8 *src); EXPORT_SYMBOL_GPL(camellia_dec_blk_2way); static void camellia_encrypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src) @@ -1229,12 +1227,10 @@ static void camellia_setup192(const unsigned char *key, u64 *subkey) } int __camellia_setkey(struct camellia_ctx *cctx, const unsigned char *key, - unsigned int key_len, u32 *flags) + unsigned int key_len) { - if (key_len != 16 && key_len != 24 && key_len != 32) { - *flags |= CRYPTO_TFM_RES_BAD_KEY_LEN; + if (key_len != 16 && key_len != 24 && key_len != 32) return -EINVAL; - } cctx->key_length = key_len; @@ -1257,8 +1253,7 @@ EXPORT_SYMBOL_GPL(__camellia_setkey); static int camellia_setkey(struct crypto_tfm *tfm, const u8 *key, unsigned int key_len) { - return __camellia_setkey(crypto_tfm_ctx(tfm), key, key_len, - &tfm->crt_flags); + return __camellia_setkey(crypto_tfm_ctx(tfm), key, key_len); } static int camellia_setkey_skcipher(struct crypto_skcipher *tfm, const u8 *key, @@ -1267,8 +1262,10 @@ static int camellia_setkey_skcipher(struct crypto_skcipher *tfm, const u8 *key, return camellia_setkey(&tfm->base, key, key_len); } -void camellia_decrypt_cbc_2way(void *ctx, u128 *dst, const u128 *src) +void camellia_decrypt_cbc_2way(const void *ctx, u8 *d, const u8 *s) { + u128 *dst = (u128 *)d; + const u128 *src = (const u128 *)s; u128 iv = *src; camellia_dec_blk_2way(ctx, (u8 *)dst, (u8 *)src); @@ -1277,9 +1274,11 @@ void camellia_decrypt_cbc_2way(void *ctx, u128 *dst, const u128 *src) } EXPORT_SYMBOL_GPL(camellia_decrypt_cbc_2way); -void camellia_crypt_ctr(void *ctx, u128 *dst, const u128 *src, le128 *iv) +void camellia_crypt_ctr(const void *ctx, u8 *d, const u8 *s, le128 *iv) { be128 ctrblk; + u128 *dst = (u128 *)d; + const u128 *src = (const u128 *)s; if (dst != src) *dst = *src; @@ -1291,9 +1290,11 @@ void camellia_crypt_ctr(void *ctx, u128 *dst, const u128 *src, le128 *iv) } EXPORT_SYMBOL_GPL(camellia_crypt_ctr); -void camellia_crypt_ctr_2way(void *ctx, u128 *dst, const u128 *src, le128 *iv) +void camellia_crypt_ctr_2way(const void *ctx, u8 *d, const u8 *s, le128 *iv) { be128 ctrblks[2]; + u128 *dst = (u128 *)d; + const u128 *src = (const u128 *)s; if (dst != src) { dst[0] = src[0]; @@ -1315,10 +1316,10 @@ static const struct common_glue_ctx camellia_enc = { .funcs = { { .num_blocks = 2, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_enc_blk_2way) } + .fn_u = { .ecb = camellia_enc_blk_2way } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_enc_blk) } + .fn_u = { .ecb = camellia_enc_blk } } } }; @@ -1328,10 +1329,10 @@ static const struct common_glue_ctx camellia_ctr = { .funcs = { { .num_blocks = 2, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(camellia_crypt_ctr_2way) } + .fn_u = { .ctr = camellia_crypt_ctr_2way } }, { .num_blocks = 1, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(camellia_crypt_ctr) } + .fn_u = { .ctr = camellia_crypt_ctr } } } }; @@ -1341,10 +1342,10 @@ static const struct common_glue_ctx camellia_dec = { .funcs = { { .num_blocks = 2, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_dec_blk_2way) } + .fn_u = { .ecb = camellia_dec_blk_2way } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(camellia_dec_blk) } + .fn_u = { .ecb = camellia_dec_blk } } } }; @@ -1354,10 +1355,10 @@ static const struct common_glue_ctx camellia_dec_cbc = { .funcs = { { .num_blocks = 2, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(camellia_decrypt_cbc_2way) } + .fn_u = { .cbc = camellia_decrypt_cbc_2way } }, { .num_blocks = 1, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(camellia_dec_blk) } + .fn_u = { .cbc = camellia_dec_blk } } } }; @@ -1373,8 +1374,7 @@ static int ecb_decrypt(struct skcipher_request *req) static int cbc_encrypt(struct skcipher_request *req) { - return glue_cbc_encrypt_req_128bit(GLUE_FUNC_CAST(camellia_enc_blk), - req); + return glue_cbc_encrypt_req_128bit(camellia_enc_blk, req); } static int cbc_decrypt(struct skcipher_request *req) diff --git a/arch/x86/crypto/cast6_avx_glue.c b/arch/x86/crypto/cast6_avx_glue.c index a8a38fffb4a9..48e0f37796fa 100644 --- a/arch/x86/crypto/cast6_avx_glue.c +++ b/arch/x86/crypto/cast6_avx_glue.c @@ -20,20 +20,17 @@ #define CAST6_PARALLEL_BLOCKS 8 -asmlinkage void cast6_ecb_enc_8way(struct cast6_ctx *ctx, u8 *dst, - const u8 *src); -asmlinkage void cast6_ecb_dec_8way(struct cast6_ctx *ctx, u8 *dst, - const u8 *src); - -asmlinkage void cast6_cbc_dec_8way(struct cast6_ctx *ctx, u8 *dst, - const u8 *src); -asmlinkage void cast6_ctr_8way(struct cast6_ctx *ctx, u8 *dst, const u8 *src, +asmlinkage void cast6_ecb_enc_8way(const void *ctx, u8 *dst, const u8 *src); +asmlinkage void cast6_ecb_dec_8way(const void *ctx, u8 *dst, const u8 *src); + +asmlinkage void cast6_cbc_dec_8way(const void *ctx, u8 *dst, const u8 *src); +asmlinkage void cast6_ctr_8way(const void *ctx, u8 *dst, const u8 *src, le128 *iv); -asmlinkage void cast6_xts_enc_8way(struct cast6_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); -asmlinkage void cast6_xts_dec_8way(struct cast6_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); +asmlinkage void cast6_xts_enc_8way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); +asmlinkage void cast6_xts_dec_8way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); static int cast6_setkey_skcipher(struct crypto_skcipher *tfm, const u8 *key, unsigned int keylen) @@ -41,21 +38,21 @@ static int cast6_setkey_skcipher(struct crypto_skcipher *tfm, return cast6_setkey(&tfm->base, key, keylen); } -static void cast6_xts_enc(void *ctx, u128 *dst, const u128 *src, le128 *iv) +static void cast6_xts_enc(const void *ctx, u8 *dst, const u8 *src, le128 *iv) { - glue_xts_crypt_128bit_one(ctx, dst, src, iv, - GLUE_FUNC_CAST(__cast6_encrypt)); + glue_xts_crypt_128bit_one(ctx, dst, src, iv, __cast6_encrypt); } -static void cast6_xts_dec(void *ctx, u128 *dst, const u128 *src, le128 *iv) +static void cast6_xts_dec(const void *ctx, u8 *dst, const u8 *src, le128 *iv) { - glue_xts_crypt_128bit_one(ctx, dst, src, iv, - GLUE_FUNC_CAST(__cast6_decrypt)); + glue_xts_crypt_128bit_one(ctx, dst, src, iv, __cast6_decrypt); } -static void cast6_crypt_ctr(void *ctx, u128 *dst, const u128 *src, le128 *iv) +static void cast6_crypt_ctr(const void *ctx, u8 *d, const u8 *s, le128 *iv) { be128 ctrblk; + u128 *dst = (u128 *)d; + const u128 *src = (const u128 *)s; le128_to_be128(&ctrblk, iv); le128_inc(iv); @@ -70,10 +67,10 @@ static const struct common_glue_ctx cast6_enc = { .funcs = { { .num_blocks = CAST6_PARALLEL_BLOCKS, - .fn_u = { .ecb = GLUE_FUNC_CAST(cast6_ecb_enc_8way) } + .fn_u = { .ecb = cast6_ecb_enc_8way } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(__cast6_encrypt) } + .fn_u = { .ecb = __cast6_encrypt } } } }; @@ -83,10 +80,10 @@ static const struct common_glue_ctx cast6_ctr = { .funcs = { { .num_blocks = CAST6_PARALLEL_BLOCKS, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(cast6_ctr_8way) } + .fn_u = { .ctr = cast6_ctr_8way } }, { .num_blocks = 1, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(cast6_crypt_ctr) } + .fn_u = { .ctr = cast6_crypt_ctr } } } }; @@ -96,10 +93,10 @@ static const struct common_glue_ctx cast6_enc_xts = { .funcs = { { .num_blocks = CAST6_PARALLEL_BLOCKS, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(cast6_xts_enc_8way) } + .fn_u = { .xts = cast6_xts_enc_8way } }, { .num_blocks = 1, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(cast6_xts_enc) } + .fn_u = { .xts = cast6_xts_enc } } } }; @@ -109,10 +106,10 @@ static const struct common_glue_ctx cast6_dec = { .funcs = { { .num_blocks = CAST6_PARALLEL_BLOCKS, - .fn_u = { .ecb = GLUE_FUNC_CAST(cast6_ecb_dec_8way) } + .fn_u = { .ecb = cast6_ecb_dec_8way } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(__cast6_decrypt) } + .fn_u = { .ecb = __cast6_decrypt } } } }; @@ -122,10 +119,10 @@ static const struct common_glue_ctx cast6_dec_cbc = { .funcs = { { .num_blocks = CAST6_PARALLEL_BLOCKS, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(cast6_cbc_dec_8way) } + .fn_u = { .cbc = cast6_cbc_dec_8way } }, { .num_blocks = 1, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(__cast6_decrypt) } + .fn_u = { .cbc = __cast6_decrypt } } } }; @@ -135,10 +132,10 @@ static const struct common_glue_ctx cast6_dec_xts = { .funcs = { { .num_blocks = CAST6_PARALLEL_BLOCKS, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(cast6_xts_dec_8way) } + .fn_u = { .xts = cast6_xts_dec_8way } }, { .num_blocks = 1, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(cast6_xts_dec) } + .fn_u = { .xts = cast6_xts_dec } } } }; @@ -154,8 +151,7 @@ static int ecb_decrypt(struct skcipher_request *req) static int cbc_encrypt(struct skcipher_request *req) { - return glue_cbc_encrypt_req_128bit(GLUE_FUNC_CAST(__cast6_encrypt), - req); + return glue_cbc_encrypt_req_128bit(__cast6_encrypt, req); } static int cbc_decrypt(struct skcipher_request *req) @@ -177,7 +173,6 @@ static int xts_cast6_setkey(struct crypto_skcipher *tfm, const u8 *key, unsigned int keylen) { struct cast6_xts_ctx *ctx = crypto_skcipher_ctx(tfm); - u32 *flags = &tfm->base.crt_flags; int err; err = xts_verify_key(tfm, key, keylen); @@ -185,13 +180,12 @@ static int xts_cast6_setkey(struct crypto_skcipher *tfm, const u8 *key, return err; /* first half of xts-key is for crypt */ - err = __cast6_setkey(&ctx->crypt_ctx, key, keylen / 2, flags); + err = __cast6_setkey(&ctx->crypt_ctx, key, keylen / 2); if (err) return err; /* second half of xts-key is for tweak */ - return __cast6_setkey(&ctx->tweak_ctx, key + keylen / 2, keylen / 2, - flags); + return __cast6_setkey(&ctx->tweak_ctx, key + keylen / 2, keylen / 2); } static int xts_encrypt(struct skcipher_request *req) @@ -199,8 +193,7 @@ static int xts_encrypt(struct skcipher_request *req) struct crypto_skcipher *tfm = crypto_skcipher_reqtfm(req); struct cast6_xts_ctx *ctx = crypto_skcipher_ctx(tfm); - return glue_xts_req_128bit(&cast6_enc_xts, req, - XTS_TWEAK_CAST(__cast6_encrypt), + return glue_xts_req_128bit(&cast6_enc_xts, req, __cast6_encrypt, &ctx->tweak_ctx, &ctx->crypt_ctx, false); } @@ -209,8 +202,7 @@ static int xts_decrypt(struct skcipher_request *req) struct crypto_skcipher *tfm = crypto_skcipher_reqtfm(req); struct cast6_xts_ctx *ctx = crypto_skcipher_ctx(tfm); - return glue_xts_req_128bit(&cast6_dec_xts, req, - XTS_TWEAK_CAST(__cast6_encrypt), + return glue_xts_req_128bit(&cast6_dec_xts, req, __cast6_encrypt, &ctx->tweak_ctx, &ctx->crypt_ctx, true); } diff --git a/arch/x86/crypto/crc32-pclmul_glue.c b/arch/x86/crypto/crc32-pclmul_glue.c index cb4ab6645106..418bd88acac8 100644 --- a/arch/x86/crypto/crc32-pclmul_glue.c +++ b/arch/x86/crypto/crc32-pclmul_glue.c @@ -94,10 +94,8 @@ static int crc32_pclmul_setkey(struct crypto_shash *hash, const u8 *key, { u32 *mctx = crypto_shash_ctx(hash); - if (keylen != sizeof(u32)) { - crypto_shash_set_flags(hash, CRYPTO_TFM_RES_BAD_KEY_LEN); + if (keylen != sizeof(u32)) return -EINVAL; - } *mctx = le32_to_cpup((__le32 *)key); return 0; } diff --git a/arch/x86/crypto/crc32c-intel_glue.c b/arch/x86/crypto/crc32c-intel_glue.c index eefa0862f309..c20d1b8a82c3 100644 --- a/arch/x86/crypto/crc32c-intel_glue.c +++ b/arch/x86/crypto/crc32c-intel_glue.c @@ -91,10 +91,8 @@ static int crc32c_intel_setkey(struct crypto_shash *hash, const u8 *key, { u32 *mctx = crypto_shash_ctx(hash); - if (keylen != sizeof(u32)) { - crypto_shash_set_flags(hash, CRYPTO_TFM_RES_BAD_KEY_LEN); + if (keylen != sizeof(u32)) return -EINVAL; - } *mctx = le32_to_cpup((__le32 *)key); return 0; } diff --git a/arch/x86/crypto/ghash-clmulni-intel_glue.c b/arch/x86/crypto/ghash-clmulni-intel_glue.c index 04d72a5a8ce9..a4b728518e28 100644 --- a/arch/x86/crypto/ghash-clmulni-intel_glue.c +++ b/arch/x86/crypto/ghash-clmulni-intel_glue.c @@ -57,10 +57,8 @@ static int ghash_setkey(struct crypto_shash *tfm, be128 *x = (be128 *)key; u64 a, b; - if (keylen != GHASH_BLOCK_SIZE) { - crypto_shash_set_flags(tfm, CRYPTO_TFM_RES_BAD_KEY_LEN); + if (keylen != GHASH_BLOCK_SIZE) return -EINVAL; - } /* perform multiplication by 'x' in GF(2^128) */ a = be64_to_cpu(x->a); @@ -257,16 +255,11 @@ static int ghash_async_setkey(struct crypto_ahash *tfm, const u8 *key, { struct ghash_async_ctx *ctx = crypto_ahash_ctx(tfm); struct crypto_ahash *child = &ctx->cryptd_tfm->base; - int err; crypto_ahash_clear_flags(child, CRYPTO_TFM_REQ_MASK); crypto_ahash_set_flags(child, crypto_ahash_get_flags(tfm) & CRYPTO_TFM_REQ_MASK); - err = crypto_ahash_setkey(child, key, keylen); - crypto_ahash_set_flags(tfm, crypto_ahash_get_flags(child) - & CRYPTO_TFM_RES_MASK); - - return err; + return crypto_ahash_setkey(child, key, keylen); } static int ghash_async_init_tfm(struct crypto_tfm *tfm) diff --git a/arch/x86/crypto/glue_helper.c b/arch/x86/crypto/glue_helper.c index d15b99397480..d3d91a0abf88 100644 --- a/arch/x86/crypto/glue_helper.c +++ b/arch/x86/crypto/glue_helper.c @@ -134,7 +134,8 @@ int glue_cbc_decrypt_req_128bit(const struct common_glue_ctx *gctx, src -= num_blocks - 1; dst -= num_blocks - 1; - gctx->funcs[i].fn_u.cbc(ctx, dst, src); + gctx->funcs[i].fn_u.cbc(ctx, (u8 *)dst, + (const u8 *)src); nbytes -= func_bytes; if (nbytes < bsize) @@ -188,7 +189,9 @@ int glue_ctr_req_128bit(const struct common_glue_ctx *gctx, /* Process multi-block batch */ do { - gctx->funcs[i].fn_u.ctr(ctx, dst, src, &ctrblk); + gctx->funcs[i].fn_u.ctr(ctx, (u8 *)dst, + (const u8 *)src, + &ctrblk); src += num_blocks; dst += num_blocks; nbytes -= func_bytes; @@ -210,7 +213,8 @@ int glue_ctr_req_128bit(const struct common_glue_ctx *gctx, be128_to_le128(&ctrblk, (be128 *)walk.iv); memcpy(&tmp, walk.src.virt.addr, nbytes); - gctx->funcs[gctx->num_funcs - 1].fn_u.ctr(ctx, &tmp, &tmp, + gctx->funcs[gctx->num_funcs - 1].fn_u.ctr(ctx, (u8 *)&tmp, + (const u8 *)&tmp, &ctrblk); memcpy(walk.dst.virt.addr, &tmp, nbytes); le128_to_be128((be128 *)walk.iv, &ctrblk); @@ -240,7 +244,8 @@ static unsigned int __glue_xts_req_128bit(const struct common_glue_ctx *gctx, if (nbytes >= func_bytes) { do { - gctx->funcs[i].fn_u.xts(ctx, dst, src, + gctx->funcs[i].fn_u.xts(ctx, (u8 *)dst, + (const u8 *)src, walk->iv); src += num_blocks; @@ -354,8 +359,8 @@ out: } EXPORT_SYMBOL_GPL(glue_xts_req_128bit); -void glue_xts_crypt_128bit_one(void *ctx, u128 *dst, const u128 *src, le128 *iv, - common_glue_func_t fn) +void glue_xts_crypt_128bit_one(const void *ctx, u8 *dst, const u8 *src, + le128 *iv, common_glue_func_t fn) { le128 ivblk = *iv; @@ -363,13 +368,13 @@ void glue_xts_crypt_128bit_one(void *ctx, u128 *dst, const u128 *src, le128 *iv, gf128mul_x_ble(iv, &ivblk); /* CC <- T xor C */ - u128_xor(dst, src, (u128 *)&ivblk); + u128_xor((u128 *)dst, (const u128 *)src, (u128 *)&ivblk); /* PP <- D(Key2,CC) */ - fn(ctx, (u8 *)dst, (u8 *)dst); + fn(ctx, dst, dst); /* P <- T xor PP */ - u128_xor(dst, dst, (u128 *)&ivblk); + u128_xor((u128 *)dst, (u128 *)dst, (u128 *)&ivblk); } EXPORT_SYMBOL_GPL(glue_xts_crypt_128bit_one); diff --git a/arch/x86/crypto/poly1305-avx2-x86_64.S b/arch/x86/crypto/poly1305-avx2-x86_64.S deleted file mode 100644 index d6063feda9da..000000000000 --- a/arch/x86/crypto/poly1305-avx2-x86_64.S +++ /dev/null @@ -1,390 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0-or-later */ -/* - * Poly1305 authenticator algorithm, RFC7539, x64 AVX2 functions - * - * Copyright (C) 2015 Martin Willi - */ - -#include <linux/linkage.h> - -.section .rodata.cst32.ANMASK, "aM", @progbits, 32 -.align 32 -ANMASK: .octa 0x0000000003ffffff0000000003ffffff - .octa 0x0000000003ffffff0000000003ffffff - -.section .rodata.cst32.ORMASK, "aM", @progbits, 32 -.align 32 -ORMASK: .octa 0x00000000010000000000000001000000 - .octa 0x00000000010000000000000001000000 - -.text - -#define h0 0x00(%rdi) -#define h1 0x04(%rdi) -#define h2 0x08(%rdi) -#define h3 0x0c(%rdi) -#define h4 0x10(%rdi) -#define r0 0x00(%rdx) -#define r1 0x04(%rdx) -#define r2 0x08(%rdx) -#define r3 0x0c(%rdx) -#define r4 0x10(%rdx) -#define u0 0x00(%r8) -#define u1 0x04(%r8) -#define u2 0x08(%r8) -#define u3 0x0c(%r8) -#define u4 0x10(%r8) -#define w0 0x14(%r8) -#define w1 0x18(%r8) -#define w2 0x1c(%r8) -#define w3 0x20(%r8) -#define w4 0x24(%r8) -#define y0 0x28(%r8) -#define y1 0x2c(%r8) -#define y2 0x30(%r8) -#define y3 0x34(%r8) -#define y4 0x38(%r8) -#define m %rsi -#define hc0 %ymm0 -#define hc1 %ymm1 -#define hc2 %ymm2 -#define hc3 %ymm3 -#define hc4 %ymm4 -#define hc0x %xmm0 -#define hc1x %xmm1 -#define hc2x %xmm2 -#define hc3x %xmm3 -#define hc4x %xmm4 -#define t1 %ymm5 -#define t2 %ymm6 -#define t1x %xmm5 -#define t2x %xmm6 -#define ruwy0 %ymm7 -#define ruwy1 %ymm8 -#define ruwy2 %ymm9 -#define ruwy3 %ymm10 -#define ruwy4 %ymm11 -#define ruwy0x %xmm7 -#define ruwy1x %xmm8 -#define ruwy2x %xmm9 -#define ruwy3x %xmm10 -#define ruwy4x %xmm11 -#define svxz1 %ymm12 -#define svxz2 %ymm13 -#define svxz3 %ymm14 -#define svxz4 %ymm15 -#define d0 %r9 -#define d1 %r10 -#define d2 %r11 -#define d3 %r12 -#define d4 %r13 - -SYM_FUNC_START(poly1305_4block_avx2) - # %rdi: Accumulator h[5] - # %rsi: 64 byte input block m - # %rdx: Poly1305 key r[5] - # %rcx: Quadblock count - # %r8: Poly1305 derived key r^2 u[5], r^3 w[5], r^4 y[5], - - # This four-block variant uses loop unrolled block processing. It - # requires 4 Poly1305 keys: r, r^2, r^3 and r^4: - # h = (h + m) * r => h = (h + m1) * r^4 + m2 * r^3 + m3 * r^2 + m4 * r - - vzeroupper - push %rbx - push %r12 - push %r13 - - # combine r0,u0,w0,y0 - vmovd y0,ruwy0x - vmovd w0,t1x - vpunpcklqdq t1,ruwy0,ruwy0 - vmovd u0,t1x - vmovd r0,t2x - vpunpcklqdq t2,t1,t1 - vperm2i128 $0x20,t1,ruwy0,ruwy0 - - # combine r1,u1,w1,y1 and s1=r1*5,v1=u1*5,x1=w1*5,z1=y1*5 - vmovd y1,ruwy1x - vmovd w1,t1x - vpunpcklqdq t1,ruwy1,ruwy1 - vmovd u1,t1x - vmovd r1,t2x - vpunpcklqdq t2,t1,t1 - vperm2i128 $0x20,t1,ruwy1,ruwy1 - vpslld $2,ruwy1,svxz1 - vpaddd ruwy1,svxz1,svxz1 - - # combine r2,u2,w2,y2 and s2=r2*5,v2=u2*5,x2=w2*5,z2=y2*5 - vmovd y2,ruwy2x - vmovd w2,t1x - vpunpcklqdq t1,ruwy2,ruwy2 - vmovd u2,t1x - vmovd r2,t2x - vpunpcklqdq t2,t1,t1 - vperm2i128 $0x20,t1,ruwy2,ruwy2 - vpslld $2,ruwy2,svxz2 - vpaddd ruwy2,svxz2,svxz2 - - # combine r3,u3,w3,y3 and s3=r3*5,v3=u3*5,x3=w3*5,z3=y3*5 - vmovd y3,ruwy3x - vmovd w3,t1x - vpunpcklqdq t1,ruwy3,ruwy3 - vmovd u3,t1x - vmovd r3,t2x - vpunpcklqdq t2,t1,t1 - vperm2i128 $0x20,t1,ruwy3,ruwy3 - vpslld $2,ruwy3,svxz3 - vpaddd ruwy3,svxz3,svxz3 - - # combine r4,u4,w4,y4 and s4=r4*5,v4=u4*5,x4=w4*5,z4=y4*5 - vmovd y4,ruwy4x - vmovd w4,t1x - vpunpcklqdq t1,ruwy4,ruwy4 - vmovd u4,t1x - vmovd r4,t2x - vpunpcklqdq t2,t1,t1 - vperm2i128 $0x20,t1,ruwy4,ruwy4 - vpslld $2,ruwy4,svxz4 - vpaddd ruwy4,svxz4,svxz4 - -.Ldoblock4: - # hc0 = [m[48-51] & 0x3ffffff, m[32-35] & 0x3ffffff, - # m[16-19] & 0x3ffffff, m[ 0- 3] & 0x3ffffff + h0] - vmovd 0x00(m),hc0x - vmovd 0x10(m),t1x - vpunpcklqdq t1,hc0,hc0 - vmovd 0x20(m),t1x - vmovd 0x30(m),t2x - vpunpcklqdq t2,t1,t1 - vperm2i128 $0x20,t1,hc0,hc0 - vpand ANMASK(%rip),hc0,hc0 - vmovd h0,t1x - vpaddd t1,hc0,hc0 - # hc1 = [(m[51-54] >> 2) & 0x3ffffff, (m[35-38] >> 2) & 0x3ffffff, - # (m[19-22] >> 2) & 0x3ffffff, (m[ 3- 6] >> 2) & 0x3ffffff + h1] - vmovd 0x03(m),hc1x - vmovd 0x13(m),t1x - vpunpcklqdq t1,hc1,hc1 - vmovd 0x23(m),t1x - vmovd 0x33(m),t2x - vpunpcklqdq t2,t1,t1 - vperm2i128 $0x20,t1,hc1,hc1 - vpsrld $2,hc1,hc1 - vpand ANMASK(%rip),hc1,hc1 - vmovd h1,t1x - vpaddd t1,hc1,hc1 - # hc2 = [(m[54-57] >> 4) & 0x3ffffff, (m[38-41] >> 4) & 0x3ffffff, - # (m[22-25] >> 4) & 0x3ffffff, (m[ 6- 9] >> 4) & 0x3ffffff + h2] - vmovd 0x06(m),hc2x - vmovd 0x16(m),t1x - vpunpcklqdq t1,hc2,hc2 - vmovd 0x26(m),t1x - vmovd 0x36(m),t2x - vpunpcklqdq t2,t1,t1 - vperm2i128 $0x20,t1,hc2,hc2 - vpsrld $4,hc2,hc2 - vpand ANMASK(%rip),hc2,hc2 - vmovd h2,t1x - vpaddd t1,hc2,hc2 - # hc3 = [(m[57-60] >> 6) & 0x3ffffff, (m[41-44] >> 6) & 0x3ffffff, - # (m[25-28] >> 6) & 0x3ffffff, (m[ 9-12] >> 6) & 0x3ffffff + h3] - vmovd 0x09(m),hc3x - vmovd 0x19(m),t1x - vpunpcklqdq t1,hc3,hc3 - vmovd 0x29(m),t1x - vmovd 0x39(m),t2x - vpunpcklqdq t2,t1,t1 - vperm2i128 $0x20,t1,hc3,hc3 - vpsrld $6,hc3,hc3 - vpand ANMASK(%rip),hc3,hc3 - vmovd h3,t1x - vpaddd t1,hc3,hc3 - # hc4 = [(m[60-63] >> 8) | (1<<24), (m[44-47] >> 8) | (1<<24), - # (m[28-31] >> 8) | (1<<24), (m[12-15] >> 8) | (1<<24) + h4] - vmovd 0x0c(m),hc4x - vmovd 0x1c(m),t1x - vpunpcklqdq t1,hc4,hc4 - vmovd 0x2c(m),t1x - vmovd 0x3c(m),t2x - vpunpcklqdq t2,t1,t1 - vperm2i128 $0x20,t1,hc4,hc4 - vpsrld $8,hc4,hc4 - vpor ORMASK(%rip),hc4,hc4 - vmovd h4,t1x - vpaddd t1,hc4,hc4 - - # t1 = [ hc0[3] * r0, hc0[2] * u0, hc0[1] * w0, hc0[0] * y0 ] - vpmuludq hc0,ruwy0,t1 - # t1 += [ hc1[3] * s4, hc1[2] * v4, hc1[1] * x4, hc1[0] * z4 ] - vpmuludq hc1,svxz4,t2 - vpaddq t2,t1,t1 - # t1 += [ hc2[3] * s3, hc2[2] * v3, hc2[1] * x3, hc2[0] * z3 ] - vpmuludq hc2,svxz3,t2 - vpaddq t2,t1,t1 - # t1 += [ hc3[3] * s2, hc3[2] * v2, hc3[1] * x2, hc3[0] * z2 ] - vpmuludq hc3,svxz2,t2 - vpaddq t2,t1,t1 - # t1 += [ hc4[3] * s1, hc4[2] * v1, hc4[1] * x1, hc4[0] * z1 ] - vpmuludq hc4,svxz1,t2 - vpaddq t2,t1,t1 - # d0 = t1[0] + t1[1] + t[2] + t[3] - vpermq $0xee,t1,t2 - vpaddq t2,t1,t1 - vpsrldq $8,t1,t2 - vpaddq t2,t1,t1 - vmovq t1x,d0 - - # t1 = [ hc0[3] * r1, hc0[2] * u1,hc0[1] * w1, hc0[0] * y1 ] - vpmuludq hc0,ruwy1,t1 - # t1 += [ hc1[3] * r0, hc1[2] * u0, hc1[1] * w0, hc1[0] * y0 ] - vpmuludq hc1,ruwy0,t2 - vpaddq t2,t1,t1 - # t1 += [ hc2[3] * s4, hc2[2] * v4, hc2[1] * x4, hc2[0] * z4 ] - vpmuludq hc2,svxz4,t2 - vpaddq t2,t1,t1 - # t1 += [ hc3[3] * s3, hc3[2] * v3, hc3[1] * x3, hc3[0] * z3 ] - vpmuludq hc3,svxz3,t2 - vpaddq t2,t1,t1 - # t1 += [ hc4[3] * s2, hc4[2] * v2, hc4[1] * x2, hc4[0] * z2 ] - vpmuludq hc4,svxz2,t2 - vpaddq t2,t1,t1 - # d1 = t1[0] + t1[1] + t1[3] + t1[4] - vpermq $0xee,t1,t2 - vpaddq t2,t1,t1 - vpsrldq $8,t1,t2 - vpaddq t2,t1,t1 - vmovq t1x,d1 - - # t1 = [ hc0[3] * r2, hc0[2] * u2, hc0[1] * w2, hc0[0] * y2 ] - vpmuludq hc0,ruwy2,t1 - # t1 += [ hc1[3] * r1, hc1[2] * u1, hc1[1] * w1, hc1[0] * y1 ] - vpmuludq hc1,ruwy1,t2 - vpaddq t2,t1,t1 - # t1 += [ hc2[3] * r0, hc2[2] * u0, hc2[1] * w0, hc2[0] * y0 ] - vpmuludq hc2,ruwy0,t2 - vpaddq t2,t1,t1 - # t1 += [ hc3[3] * s4, hc3[2] * v4, hc3[1] * x4, hc3[0] * z4 ] - vpmuludq hc3,svxz4,t2 - vpaddq t2,t1,t1 - # t1 += [ hc4[3] * s3, hc4[2] * v3, hc4[1] * x3, hc4[0] * z3 ] - vpmuludq hc4,svxz3,t2 - vpaddq t2,t1,t1 - # d2 = t1[0] + t1[1] + t1[2] + t1[3] - vpermq $0xee,t1,t2 - vpaddq t2,t1,t1 - vpsrldq $8,t1,t2 - vpaddq t2,t1,t1 - vmovq t1x,d2 - - # t1 = [ hc0[3] * r3, hc0[2] * u3, hc0[1] * w3, hc0[0] * y3 ] - vpmuludq hc0,ruwy3,t1 - # t1 += [ hc1[3] * r2, hc1[2] * u2, hc1[1] * w2, hc1[0] * y2 ] - vpmuludq hc1,ruwy2,t2 - vpaddq t2,t1,t1 - # t1 += [ hc2[3] * r1, hc2[2] * u1, hc2[1] * w1, hc2[0] * y1 ] - vpmuludq hc2,ruwy1,t2 - vpaddq t2,t1,t1 - # t1 += [ hc3[3] * r0, hc3[2] * u0, hc3[1] * w0, hc3[0] * y0 ] - vpmuludq hc3,ruwy0,t2 - vpaddq t2,t1,t1 - # t1 += [ hc4[3] * s4, hc4[2] * v4, hc4[1] * x4, hc4[0] * z4 ] - vpmuludq hc4,svxz4,t2 - vpaddq t2,t1,t1 - # d3 = t1[0] + t1[1] + t1[2] + t1[3] - vpermq $0xee,t1,t2 - vpaddq t2,t1,t1 - vpsrldq $8,t1,t2 - vpaddq t2,t1,t1 - vmovq t1x,d3 - - # t1 = [ hc0[3] * r4, hc0[2] * u4, hc0[1] * w4, hc0[0] * y4 ] - vpmuludq hc0,ruwy4,t1 - # t1 += [ hc1[3] * r3, hc1[2] * u3, hc1[1] * w3, hc1[0] * y3 ] - vpmuludq hc1,ruwy3,t2 - vpaddq t2,t1,t1 - # t1 += [ hc2[3] * r2, hc2[2] * u2, hc2[1] * w2, hc2[0] * y2 ] - vpmuludq hc2,ruwy2,t2 - vpaddq t2,t1,t1 - # t1 += [ hc3[3] * r1, hc3[2] * u1, hc3[1] * w1, hc3[0] * y1 ] - vpmuludq hc3,ruwy1,t2 - vpaddq t2,t1,t1 - # t1 += [ hc4[3] * r0, hc4[2] * u0, hc4[1] * w0, hc4[0] * y0 ] - vpmuludq hc4,ruwy0,t2 - vpaddq t2,t1,t1 - # d4 = t1[0] + t1[1] + t1[2] + t1[3] - vpermq $0xee,t1,t2 - vpaddq t2,t1,t1 - vpsrldq $8,t1,t2 - vpaddq t2,t1,t1 - vmovq t1x,d4 - - # Now do a partial reduction mod (2^130)-5, carrying h0 -> h1 -> h2 -> - # h3 -> h4 -> h0 -> h1 to get h0,h2,h3,h4 < 2^26 and h1 < 2^26 + a small - # amount. Careful: we must not assume the carry bits 'd0 >> 26', - # 'd1 >> 26', 'd2 >> 26', 'd3 >> 26', and '(d4 >> 26) * 5' fit in 32-bit - # integers. It's true in a single-block implementation, but not here. - - # d1 += d0 >> 26 - mov d0,%rax - shr $26,%rax - add %rax,d1 - # h0 = d0 & 0x3ffffff - mov d0,%rbx - and $0x3ffffff,%ebx - - # d2 += d1 >> 26 - mov d1,%rax - shr $26,%rax - add %rax,d2 - # h1 = d1 & 0x3ffffff - mov d1,%rax - and $0x3ffffff,%eax - mov %eax,h1 - - # d3 += d2 >> 26 - mov d2,%rax - shr $26,%rax - add %rax,d3 - # h2 = d2 & 0x3ffffff - mov d2,%rax - and $0x3ffffff,%eax - mov %eax,h2 - - # d4 += d3 >> 26 - mov d3,%rax - shr $26,%rax - add %rax,d4 - # h3 = d3 & 0x3ffffff - mov d3,%rax - and $0x3ffffff,%eax - mov %eax,h3 - - # h0 += (d4 >> 26) * 5 - mov d4,%rax - shr $26,%rax - lea (%rax,%rax,4),%rax - add %rax,%rbx - # h4 = d4 & 0x3ffffff - mov d4,%rax - and $0x3ffffff,%eax - mov %eax,h4 - - # h1 += h0 >> 26 - mov %rbx,%rax - shr $26,%rax - add %eax,h1 - # h0 = h0 & 0x3ffffff - andl $0x3ffffff,%ebx - mov %ebx,h0 - - add $0x40,m - dec %rcx - jnz .Ldoblock4 - - vzeroupper - pop %r13 - pop %r12 - pop %rbx - ret -SYM_FUNC_END(poly1305_4block_avx2) diff --git a/arch/x86/crypto/poly1305-sse2-x86_64.S b/arch/x86/crypto/poly1305-sse2-x86_64.S deleted file mode 100644 index d8ea29b96640..000000000000 --- a/arch/x86/crypto/poly1305-sse2-x86_64.S +++ /dev/null @@ -1,590 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0-or-later */ -/* - * Poly1305 authenticator algorithm, RFC7539, x64 SSE2 functions - * - * Copyright (C) 2015 Martin Willi - */ - -#include <linux/linkage.h> - -.section .rodata.cst16.ANMASK, "aM", @progbits, 16 -.align 16 -ANMASK: .octa 0x0000000003ffffff0000000003ffffff - -.section .rodata.cst16.ORMASK, "aM", @progbits, 16 -.align 16 -ORMASK: .octa 0x00000000010000000000000001000000 - -.text - -#define h0 0x00(%rdi) -#define h1 0x04(%rdi) -#define h2 0x08(%rdi) -#define h3 0x0c(%rdi) -#define h4 0x10(%rdi) -#define r0 0x00(%rdx) -#define r1 0x04(%rdx) -#define r2 0x08(%rdx) -#define r3 0x0c(%rdx) -#define r4 0x10(%rdx) -#define s1 0x00(%rsp) -#define s2 0x04(%rsp) -#define s3 0x08(%rsp) -#define s4 0x0c(%rsp) -#define m %rsi -#define h01 %xmm0 -#define h23 %xmm1 -#define h44 %xmm2 -#define t1 %xmm3 -#define t2 %xmm4 -#define t3 %xmm5 -#define t4 %xmm6 -#define mask %xmm7 -#define d0 %r8 -#define d1 %r9 -#define d2 %r10 -#define d3 %r11 -#define d4 %r12 - -SYM_FUNC_START(poly1305_block_sse2) - # %rdi: Accumulator h[5] - # %rsi: 16 byte input block m - # %rdx: Poly1305 key r[5] - # %rcx: Block count - - # This single block variant tries to improve performance by doing two - # multiplications in parallel using SSE instructions. There is quite - # some quardword packing involved, hence the speedup is marginal. - - push %rbx - push %r12 - sub $0x10,%rsp - - # s1..s4 = r1..r4 * 5 - mov r1,%eax - lea (%eax,%eax,4),%eax - mov %eax,s1 - mov r2,%eax - lea (%eax,%eax,4),%eax - mov %eax,s2 - mov r3,%eax - lea (%eax,%eax,4),%eax - mov %eax,s3 - mov r4,%eax - lea (%eax,%eax,4),%eax - mov %eax,s4 - - movdqa ANMASK(%rip),mask - -.Ldoblock: - # h01 = [0, h1, 0, h0] - # h23 = [0, h3, 0, h2] - # h44 = [0, h4, 0, h4] - movd h0,h01 - movd h1,t1 - movd h2,h23 - movd h3,t2 - movd h4,h44 - punpcklqdq t1,h01 - punpcklqdq t2,h23 - punpcklqdq h44,h44 - - # h01 += [ (m[3-6] >> 2) & 0x3ffffff, m[0-3] & 0x3ffffff ] - movd 0x00(m),t1 - movd 0x03(m),t2 - psrld $2,t2 - punpcklqdq t2,t1 - pand mask,t1 - paddd t1,h01 - # h23 += [ (m[9-12] >> 6) & 0x3ffffff, (m[6-9] >> 4) & 0x3ffffff ] - movd 0x06(m),t1 - movd 0x09(m),t2 - psrld $4,t1 - psrld $6,t2 - punpcklqdq t2,t1 - pand mask,t1 - paddd t1,h23 - # h44 += [ (m[12-15] >> 8) | (1 << 24), (m[12-15] >> 8) | (1 << 24) ] - mov 0x0c(m),%eax - shr $8,%eax - or $0x01000000,%eax - movd %eax,t1 - pshufd $0xc4,t1,t1 - paddd t1,h44 - - # t1[0] = h0 * r0 + h2 * s3 - # t1[1] = h1 * s4 + h3 * s2 - movd r0,t1 - movd s4,t2 - punpcklqdq t2,t1 - pmuludq h01,t1 - movd s3,t2 - movd s2,t3 - punpcklqdq t3,t2 - pmuludq h23,t2 - paddq t2,t1 - # t2[0] = h0 * r1 + h2 * s4 - # t2[1] = h1 * r0 + h3 * s3 - movd r1,t2 - movd r0,t3 - punpcklqdq t3,t2 - pmuludq h01,t2 - movd s4,t3 - movd s3,t4 - punpcklqdq t4,t3 - pmuludq h23,t3 - paddq t3,t2 - # t3[0] = h4 * s1 - # t3[1] = h4 * s2 - movd s1,t3 - movd s2,t4 - punpcklqdq t4,t3 - pmuludq h44,t3 - # d0 = t1[0] + t1[1] + t3[0] - # d1 = t2[0] + t2[1] + t3[1] - movdqa t1,t4 - punpcklqdq t2,t4 - punpckhqdq t2,t1 - paddq t4,t1 - paddq t3,t1 - movq t1,d0 - psrldq $8,t1 - movq t1,d1 - - # t1[0] = h0 * r2 + h2 * r0 - # t1[1] = h1 * r1 + h3 * s4 - movd r2,t1 - movd r1,t2 - punpcklqdq t2,t1 - pmuludq h01,t1 - movd r0,t2 - movd s4,t3 - punpcklqdq t3,t2 - pmuludq h23,t2 - paddq t2,t1 - # t2[0] = h0 * r3 + h2 * r1 - # t2[1] = h1 * r2 + h3 * r0 - movd r3,t2 - movd r2,t3 - punpcklqdq t3,t2 - pmuludq h01,t2 - movd r1,t3 - movd r0,t4 - punpcklqdq t4,t3 - pmuludq h23,t3 - paddq t3,t2 - # t3[0] = h4 * s3 - # t3[1] = h4 * s4 - movd s3,t3 - movd s4,t4 - punpcklqdq t4,t3 - pmuludq h44,t3 - # d2 = t1[0] + t1[1] + t3[0] - # d3 = t2[0] + t2[1] + t3[1] - movdqa t1,t4 - punpcklqdq t2,t4 - punpckhqdq t2,t1 - paddq t4,t1 - paddq t3,t1 - movq t1,d2 - psrldq $8,t1 - movq t1,d3 - - # t1[0] = h0 * r4 + h2 * r2 - # t1[1] = h1 * r3 + h3 * r1 - movd r4,t1 - movd r3,t2 - punpcklqdq t2,t1 - pmuludq h01,t1 - movd r2,t2 - movd r1,t3 - punpcklqdq t3,t2 - pmuludq h23,t2 - paddq t2,t1 - # t3[0] = h4 * r0 - movd r0,t3 - pmuludq h44,t3 - # d4 = t1[0] + t1[1] + t3[0] - movdqa t1,t4 - psrldq $8,t4 - paddq t4,t1 - paddq t3,t1 - movq t1,d4 - - # d1 += d0 >> 26 - mov d0,%rax - shr $26,%rax - add %rax,d1 - # h0 = d0 & 0x3ffffff - mov d0,%rbx - and $0x3ffffff,%ebx - - # d2 += d1 >> 26 - mov d1,%rax - shr $26,%rax - add %rax,d2 - # h1 = d1 & 0x3ffffff - mov d1,%rax - and $0x3ffffff,%eax - mov %eax,h1 - - # d3 += d2 >> 26 - mov d2,%rax - shr $26,%rax - add %rax,d3 - # h2 = d2 & 0x3ffffff - mov d2,%rax - and $0x3ffffff,%eax - mov %eax,h2 - - # d4 += d3 >> 26 - mov d3,%rax - shr $26,%rax - add %rax,d4 - # h3 = d3 & 0x3ffffff - mov d3,%rax - and $0x3ffffff,%eax - mov %eax,h3 - - # h0 += (d4 >> 26) * 5 - mov d4,%rax - shr $26,%rax - lea (%rax,%rax,4),%rax - add %rax,%rbx - # h4 = d4 & 0x3ffffff - mov d4,%rax - and $0x3ffffff,%eax - mov %eax,h4 - - # h1 += h0 >> 26 - mov %rbx,%rax - shr $26,%rax - add %eax,h1 - # h0 = h0 & 0x3ffffff - andl $0x3ffffff,%ebx - mov %ebx,h0 - - add $0x10,m - dec %rcx - jnz .Ldoblock - - # Zeroing of key material - mov %rcx,0x00(%rsp) - mov %rcx,0x08(%rsp) - - add $0x10,%rsp - pop %r12 - pop %rbx - ret -SYM_FUNC_END(poly1305_block_sse2) - - -#define u0 0x00(%r8) -#define u1 0x04(%r8) -#define u2 0x08(%r8) -#define u3 0x0c(%r8) -#define u4 0x10(%r8) -#define hc0 %xmm0 -#define hc1 %xmm1 -#define hc2 %xmm2 -#define hc3 %xmm5 -#define hc4 %xmm6 -#define ru0 %xmm7 -#define ru1 %xmm8 -#define ru2 %xmm9 -#define ru3 %xmm10 -#define ru4 %xmm11 -#define sv1 %xmm12 -#define sv2 %xmm13 -#define sv3 %xmm14 -#define sv4 %xmm15 -#undef d0 -#define d0 %r13 - -SYM_FUNC_START(poly1305_2block_sse2) - # %rdi: Accumulator h[5] - # %rsi: 16 byte input block m - # %rdx: Poly1305 key r[5] - # %rcx: Doubleblock count - # %r8: Poly1305 derived key r^2 u[5] - - # This two-block variant further improves performance by using loop - # unrolled block processing. This is more straight forward and does - # less byte shuffling, but requires a second Poly1305 key r^2: - # h = (h + m) * r => h = (h + m1) * r^2 + m2 * r - - push %rbx - push %r12 - push %r13 - - # combine r0,u0 - movd u0,ru0 - movd r0,t1 - punpcklqdq t1,ru0 - - # combine r1,u1 and s1=r1*5,v1=u1*5 - movd u1,ru1 - movd r1,t1 - punpcklqdq t1,ru1 - movdqa ru1,sv1 - pslld $2,sv1 - paddd ru1,sv1 - - # combine r2,u2 and s2=r2*5,v2=u2*5 - movd u2,ru2 - movd r2,t1 - punpcklqdq t1,ru2 - movdqa ru2,sv2 - pslld $2,sv2 - paddd ru2,sv2 - - # combine r3,u3 and s3=r3*5,v3=u3*5 - movd u3,ru3 - movd r3,t1 - punpcklqdq t1,ru3 - movdqa ru3,sv3 - pslld $2,sv3 - paddd ru3,sv3 - - # combine r4,u4 and s4=r4*5,v4=u4*5 - movd u4,ru4 - movd r4,t1 - punpcklqdq t1,ru4 - movdqa ru4,sv4 - pslld $2,sv4 - paddd ru4,sv4 - -.Ldoblock2: - # hc0 = [ m[16-19] & 0x3ffffff, h0 + m[0-3] & 0x3ffffff ] - movd 0x00(m),hc0 - movd 0x10(m),t1 - punpcklqdq t1,hc0 - pand ANMASK(%rip),hc0 - movd h0,t1 - paddd t1,hc0 - # hc1 = [ (m[19-22] >> 2) & 0x3ffffff, h1 + (m[3-6] >> 2) & 0x3ffffff ] - movd 0x03(m),hc1 - movd 0x13(m),t1 - punpcklqdq t1,hc1 - psrld $2,hc1 - pand ANMASK(%rip),hc1 - movd h1,t1 - paddd t1,hc1 - # hc2 = [ (m[22-25] >> 4) & 0x3ffffff, h2 + (m[6-9] >> 4) & 0x3ffffff ] - movd 0x06(m),hc2 - movd 0x16(m),t1 - punpcklqdq t1,hc2 - psrld $4,hc2 - pand ANMASK(%rip),hc2 - movd h2,t1 - paddd t1,hc2 - # hc3 = [ (m[25-28] >> 6) & 0x3ffffff, h3 + (m[9-12] >> 6) & 0x3ffffff ] - movd 0x09(m),hc3 - movd 0x19(m),t1 - punpcklqdq t1,hc3 - psrld $6,hc3 - pand ANMASK(%rip),hc3 - movd h3,t1 - paddd t1,hc3 - # hc4 = [ (m[28-31] >> 8) | (1<<24), h4 + (m[12-15] >> 8) | (1<<24) ] - movd 0x0c(m),hc4 - movd 0x1c(m),t1 - punpcklqdq t1,hc4 - psrld $8,hc4 - por ORMASK(%rip),hc4 - movd h4,t1 - paddd t1,hc4 - - # t1 = [ hc0[1] * r0, hc0[0] * u0 ] - movdqa ru0,t1 - pmuludq hc0,t1 - # t1 += [ hc1[1] * s4, hc1[0] * v4 ] - movdqa sv4,t2 - pmuludq hc1,t2 - paddq t2,t1 - # t1 += [ hc2[1] * s3, hc2[0] * v3 ] - movdqa sv3,t2 - pmuludq hc2,t2 - paddq t2,t1 - # t1 += [ hc3[1] * s2, hc3[0] * v2 ] - movdqa sv2,t2 - pmuludq hc3,t2 - paddq t2,t1 - # t1 += [ hc4[1] * s1, hc4[0] * v1 ] - movdqa sv1,t2 - pmuludq hc4,t2 - paddq t2,t1 - # d0 = t1[0] + t1[1] - movdqa t1,t2 - psrldq $8,t2 - paddq t2,t1 - movq t1,d0 - - # t1 = [ hc0[1] * r1, hc0[0] * u1 ] - movdqa ru1,t1 - pmuludq hc0,t1 - # t1 += [ hc1[1] * r0, hc1[0] * u0 ] - movdqa ru0,t2 - pmuludq hc1,t2 - paddq t2,t1 - # t1 += [ hc2[1] * s4, hc2[0] * v4 ] - movdqa sv4,t2 - pmuludq hc2,t2 - paddq t2,t1 - # t1 += [ hc3[1] * s3, hc3[0] * v3 ] - movdqa sv3,t2 - pmuludq hc3,t2 - paddq t2,t1 - # t1 += [ hc4[1] * s2, hc4[0] * v2 ] - movdqa sv2,t2 - pmuludq hc4,t2 - paddq t2,t1 - # d1 = t1[0] + t1[1] - movdqa t1,t2 - psrldq $8,t2 - paddq t2,t1 - movq t1,d1 - - # t1 = [ hc0[1] * r2, hc0[0] * u2 ] - movdqa ru2,t1 - pmuludq hc0,t1 - # t1 += [ hc1[1] * r1, hc1[0] * u1 ] - movdqa ru1,t2 - pmuludq hc1,t2 - paddq t2,t1 - # t1 += [ hc2[1] * r0, hc2[0] * u0 ] - movdqa ru0,t2 - pmuludq hc2,t2 - paddq t2,t1 - # t1 += [ hc3[1] * s4, hc3[0] * v4 ] - movdqa sv4,t2 - pmuludq hc3,t2 - paddq t2,t1 - # t1 += [ hc4[1] * s3, hc4[0] * v3 ] - movdqa sv3,t2 - pmuludq hc4,t2 - paddq t2,t1 - # d2 = t1[0] + t1[1] - movdqa t1,t2 - psrldq $8,t2 - paddq t2,t1 - movq t1,d2 - - # t1 = [ hc0[1] * r3, hc0[0] * u3 ] - movdqa ru3,t1 - pmuludq hc0,t1 - # t1 += [ hc1[1] * r2, hc1[0] * u2 ] - movdqa ru2,t2 - pmuludq hc1,t2 - paddq t2,t1 - # t1 += [ hc2[1] * r1, hc2[0] * u1 ] - movdqa ru1,t2 - pmuludq hc2,t2 - paddq t2,t1 - # t1 += [ hc3[1] * r0, hc3[0] * u0 ] - movdqa ru0,t2 - pmuludq hc3,t2 - paddq t2,t1 - # t1 += [ hc4[1] * s4, hc4[0] * v4 ] - movdqa sv4,t2 - pmuludq hc4,t2 - paddq t2,t1 - # d3 = t1[0] + t1[1] - movdqa t1,t2 - psrldq $8,t2 - paddq t2,t1 - movq t1,d3 - - # t1 = [ hc0[1] * r4, hc0[0] * u4 ] - movdqa ru4,t1 - pmuludq hc0,t1 - # t1 += [ hc1[1] * r3, hc1[0] * u3 ] - movdqa ru3,t2 - pmuludq hc1,t2 - paddq t2,t1 - # t1 += [ hc2[1] * r2, hc2[0] * u2 ] - movdqa ru2,t2 - pmuludq hc2,t2 - paddq t2,t1 - # t1 += [ hc3[1] * r1, hc3[0] * u1 ] - movdqa ru1,t2 - pmuludq hc3,t2 - paddq t2,t1 - # t1 += [ hc4[1] * r0, hc4[0] * u0 ] - movdqa ru0,t2 - pmuludq hc4,t2 - paddq t2,t1 - # d4 = t1[0] + t1[1] - movdqa t1,t2 - psrldq $8,t2 - paddq t2,t1 - movq t1,d4 - - # Now do a partial reduction mod (2^130)-5, carrying h0 -> h1 -> h2 -> - # h3 -> h4 -> h0 -> h1 to get h0,h2,h3,h4 < 2^26 and h1 < 2^26 + a small - # amount. Careful: we must not assume the carry bits 'd0 >> 26', - # 'd1 >> 26', 'd2 >> 26', 'd3 >> 26', and '(d4 >> 26) * 5' fit in 32-bit - # integers. It's true in a single-block implementation, but not here. - - # d1 += d0 >> 26 - mov d0,%rax - shr $26,%rax - add %rax,d1 - # h0 = d0 & 0x3ffffff - mov d0,%rbx - and $0x3ffffff,%ebx - - # d2 += d1 >> 26 - mov d1,%rax - shr $26,%rax - add %rax,d2 - # h1 = d1 & 0x3ffffff - mov d1,%rax - and $0x3ffffff,%eax - mov %eax,h1 - - # d3 += d2 >> 26 - mov d2,%rax - shr $26,%rax - add %rax,d3 - # h2 = d2 & 0x3ffffff - mov d2,%rax - and $0x3ffffff,%eax - mov %eax,h2 - - # d4 += d3 >> 26 - mov d3,%rax - shr $26,%rax - add %rax,d4 - # h3 = d3 & 0x3ffffff - mov d3,%rax - and $0x3ffffff,%eax - mov %eax,h3 - - # h0 += (d4 >> 26) * 5 - mov d4,%rax - shr $26,%rax - lea (%rax,%rax,4),%rax - add %rax,%rbx - # h4 = d4 & 0x3ffffff - mov d4,%rax - and $0x3ffffff,%eax - mov %eax,h4 - - # h1 += h0 >> 26 - mov %rbx,%rax - shr $26,%rax - add %eax,h1 - # h0 = h0 & 0x3ffffff - andl $0x3ffffff,%ebx - mov %ebx,h0 - - add $0x20,m - dec %rcx - jnz .Ldoblock2 - - pop %r13 - pop %r12 - pop %rbx - ret -SYM_FUNC_END(poly1305_2block_sse2) diff --git a/arch/x86/crypto/poly1305-x86_64-cryptogams.pl b/arch/x86/crypto/poly1305-x86_64-cryptogams.pl new file mode 100644 index 000000000000..7a6b5380a46f --- /dev/null +++ b/arch/x86/crypto/poly1305-x86_64-cryptogams.pl @@ -0,0 +1,4265 @@ +#!/usr/bin/env perl +# SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause +# +# Copyright (C) 2017-2018 Samuel Neves <sneves@dei.uc.pt>. All Rights Reserved. +# Copyright (C) 2017-2019 Jason A. Donenfeld <Jason@zx2c4.com>. All Rights Reserved. +# Copyright (C) 2006-2017 CRYPTOGAMS by <appro@openssl.org>. All Rights Reserved. +# +# This code is taken from the OpenSSL project but the author, Andy Polyakov, +# has relicensed it under the licenses specified in the SPDX header above. +# The original headers, including the original license headers, are +# included below for completeness. +# +# ==================================================================== +# Written by Andy Polyakov <appro@openssl.org> for the OpenSSL +# project. The module is, however, dual licensed under OpenSSL and +# CRYPTOGAMS licenses depending on where you obtain it. For further +# details see http://www.openssl.org/~appro/cryptogams/. +# ==================================================================== +# +# This module implements Poly1305 hash for x86_64. +# +# March 2015 +# +# Initial release. +# +# December 2016 +# +# Add AVX512F+VL+BW code path. +# +# November 2017 +# +# Convert AVX512F+VL+BW code path to pure AVX512F, so that it can be +# executed even on Knights Landing. Trigger for modification was +# observation that AVX512 code paths can negatively affect overall +# Skylake-X system performance. Since we are likely to suppress +# AVX512F capability flag [at least on Skylake-X], conversion serves +# as kind of "investment protection". Note that next *lake processor, +# Cannonlake, has AVX512IFMA code path to execute... +# +# Numbers are cycles per processed byte with poly1305_blocks alone, +# measured with rdtsc at fixed clock frequency. +# +# IALU/gcc-4.8(*) AVX(**) AVX2 AVX-512 +# P4 4.46/+120% - +# Core 2 2.41/+90% - +# Westmere 1.88/+120% - +# Sandy Bridge 1.39/+140% 1.10 +# Haswell 1.14/+175% 1.11 0.65 +# Skylake[-X] 1.13/+120% 0.96 0.51 [0.35] +# Silvermont 2.83/+95% - +# Knights L 3.60/? 1.65 1.10 0.41(***) +# Goldmont 1.70/+180% - +# VIA Nano 1.82/+150% - +# Sledgehammer 1.38/+160% - +# Bulldozer 2.30/+130% 0.97 +# Ryzen 1.15/+200% 1.08 1.18 +# +# (*) improvement coefficients relative to clang are more modest and +# are ~50% on most processors, in both cases we are comparing to +# __int128 code; +# (**) SSE2 implementation was attempted, but among non-AVX processors +# it was faster than integer-only code only on older Intel P4 and +# Core processors, 50-30%, less newer processor is, but slower on +# contemporary ones, for example almost 2x slower on Atom, and as +# former are naturally disappearing, SSE2 is deemed unnecessary; +# (***) strangely enough performance seems to vary from core to core, +# listed result is best case; + +$flavour = shift; +$output = shift; +if ($flavour =~ /\./) { $output = $flavour; undef $flavour; } + +$win64=0; $win64=1 if ($flavour =~ /[nm]asm|mingw64/ || $output =~ /\.asm$/); +$kernel=0; $kernel=1 if (!$flavour && !$output); + +if (!$kernel) { + $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1; + ( $xlate="${dir}x86_64-xlate.pl" and -f $xlate ) or + ( $xlate="${dir}../../perlasm/x86_64-xlate.pl" and -f $xlate) or + die "can't locate x86_64-xlate.pl"; + + open OUT,"| \"$^X\" \"$xlate\" $flavour \"$output\""; + *STDOUT=*OUT; + + if (`$ENV{CC} -Wa,-v -c -o /dev/null -x assembler /dev/null 2>&1` + =~ /GNU assembler version ([2-9]\.[0-9]+)/) { + $avx = ($1>=2.19) + ($1>=2.22) + ($1>=2.25); + } + + if (!$avx && $win64 && ($flavour =~ /nasm/ || $ENV{ASM} =~ /nasm/) && + `nasm -v 2>&1` =~ /NASM version ([2-9]\.[0-9]+)(?:\.([0-9]+))?/) { + $avx = ($1>=2.09) + ($1>=2.10) + ($1>=2.12); + $avx += 1 if ($1==2.11 && $2>=8); + } + + if (!$avx && $win64 && ($flavour =~ /masm/ || $ENV{ASM} =~ /ml64/) && + `ml64 2>&1` =~ /Version ([0-9]+)\./) { + $avx = ($1>=10) + ($1>=11); + } + + if (!$avx && `$ENV{CC} -v 2>&1` =~ /((?:^clang|LLVM) version|.*based on LLVM) ([3-9]\.[0-9]+)/) { + $avx = ($2>=3.0) + ($2>3.0); + } +} else { + $avx = 4; # The kernel uses ifdefs for this. +} + +sub declare_function() { + my ($name, $align, $nargs) = @_; + if($kernel) { + $code .= ".align $align\n"; + $code .= "SYM_FUNC_START($name)\n"; + $code .= ".L$name:\n"; + } else { + $code .= ".globl $name\n"; + $code .= ".type $name,\@function,$nargs\n"; + $code .= ".align $align\n"; + $code .= "$name:\n"; + } +} + +sub end_function() { + my ($name) = @_; + if($kernel) { + $code .= "SYM_FUNC_END($name)\n"; + } else { + $code .= ".size $name,.-$name\n"; + } +} + +$code.=<<___ if $kernel; +#include <linux/linkage.h> +___ + +if ($avx) { +$code.=<<___ if $kernel; +.section .rodata +___ +$code.=<<___; +.align 64 +.Lconst: +.Lmask24: +.long 0x0ffffff,0,0x0ffffff,0,0x0ffffff,0,0x0ffffff,0 +.L129: +.long `1<<24`,0,`1<<24`,0,`1<<24`,0,`1<<24`,0 +.Lmask26: +.long 0x3ffffff,0,0x3ffffff,0,0x3ffffff,0,0x3ffffff,0 +.Lpermd_avx2: +.long 2,2,2,3,2,0,2,1 +.Lpermd_avx512: +.long 0,0,0,1, 0,2,0,3, 0,4,0,5, 0,6,0,7 + +.L2_44_inp_permd: +.long 0,1,1,2,2,3,7,7 +.L2_44_inp_shift: +.quad 0,12,24,64 +.L2_44_mask: +.quad 0xfffffffffff,0xfffffffffff,0x3ffffffffff,0xffffffffffffffff +.L2_44_shift_rgt: +.quad 44,44,42,64 +.L2_44_shift_lft: +.quad 8,8,10,64 + +.align 64 +.Lx_mask44: +.quad 0xfffffffffff,0xfffffffffff,0xfffffffffff,0xfffffffffff +.quad 0xfffffffffff,0xfffffffffff,0xfffffffffff,0xfffffffffff +.Lx_mask42: +.quad 0x3ffffffffff,0x3ffffffffff,0x3ffffffffff,0x3ffffffffff +.quad 0x3ffffffffff,0x3ffffffffff,0x3ffffffffff,0x3ffffffffff +___ +} +$code.=<<___ if (!$kernel); +.asciz "Poly1305 for x86_64, CRYPTOGAMS by <appro\@openssl.org>" +.align 16 +___ + +my ($ctx,$inp,$len,$padbit)=("%rdi","%rsi","%rdx","%rcx"); +my ($mac,$nonce)=($inp,$len); # *_emit arguments +my ($d1,$d2,$d3, $r0,$r1,$s1)=("%r8","%r9","%rdi","%r11","%r12","%r13"); +my ($h0,$h1,$h2)=("%r14","%rbx","%r10"); + +sub poly1305_iteration { +# input: copy of $r1 in %rax, $h0-$h2, $r0-$r1 +# output: $h0-$h2 *= $r0-$r1 +$code.=<<___; + mulq $h0 # h0*r1 + mov %rax,$d2 + mov $r0,%rax + mov %rdx,$d3 + + mulq $h0 # h0*r0 + mov %rax,$h0 # future $h0 + mov $r0,%rax + mov %rdx,$d1 + + mulq $h1 # h1*r0 + add %rax,$d2 + mov $s1,%rax + adc %rdx,$d3 + + mulq $h1 # h1*s1 + mov $h2,$h1 # borrow $h1 + add %rax,$h0 + adc %rdx,$d1 + + imulq $s1,$h1 # h2*s1 + add $h1,$d2 + mov $d1,$h1 + adc \$0,$d3 + + imulq $r0,$h2 # h2*r0 + add $d2,$h1 + mov \$-4,%rax # mask value + adc $h2,$d3 + + and $d3,%rax # last reduction step + mov $d3,$h2 + shr \$2,$d3 + and \$3,$h2 + add $d3,%rax + add %rax,$h0 + adc \$0,$h1 + adc \$0,$h2 +___ +} + +######################################################################## +# Layout of opaque area is following. +# +# unsigned __int64 h[3]; # current hash value base 2^64 +# unsigned __int64 r[2]; # key value base 2^64 + +$code.=<<___; +.text +___ +$code.=<<___ if (!$kernel); +.extern OPENSSL_ia32cap_P + +.globl poly1305_init_x86_64 +.hidden poly1305_init_x86_64 +.globl poly1305_blocks_x86_64 +.hidden poly1305_blocks_x86_64 +.globl poly1305_emit_x86_64 +.hidden poly1305_emit_x86_64 +___ +&declare_function("poly1305_init_x86_64", 32, 3); +$code.=<<___; + xor %rax,%rax + mov %rax,0($ctx) # initialize hash value + mov %rax,8($ctx) + mov %rax,16($ctx) + + cmp \$0,$inp + je .Lno_key +___ +$code.=<<___ if (!$kernel); + lea poly1305_blocks_x86_64(%rip),%r10 + lea poly1305_emit_x86_64(%rip),%r11 +___ +$code.=<<___ if (!$kernel && $avx); + mov OPENSSL_ia32cap_P+4(%rip),%r9 + lea poly1305_blocks_avx(%rip),%rax + lea poly1305_emit_avx(%rip),%rcx + bt \$`60-32`,%r9 # AVX? + cmovc %rax,%r10 + cmovc %rcx,%r11 +___ +$code.=<<___ if (!$kernel && $avx>1); + lea poly1305_blocks_avx2(%rip),%rax + bt \$`5+32`,%r9 # AVX2? + cmovc %rax,%r10 +___ +$code.=<<___ if (!$kernel && $avx>3); + mov \$`(1<<31|1<<21|1<<16)`,%rax + shr \$32,%r9 + and %rax,%r9 + cmp %rax,%r9 + je .Linit_base2_44 +___ +$code.=<<___; + mov \$0x0ffffffc0fffffff,%rax + mov \$0x0ffffffc0ffffffc,%rcx + and 0($inp),%rax + and 8($inp),%rcx + mov %rax,24($ctx) + mov %rcx,32($ctx) +___ +$code.=<<___ if (!$kernel && $flavour !~ /elf32/); + mov %r10,0(%rdx) + mov %r11,8(%rdx) +___ +$code.=<<___ if (!$kernel && $flavour =~ /elf32/); + mov %r10d,0(%rdx) + mov %r11d,4(%rdx) +___ +$code.=<<___; + mov \$1,%eax +.Lno_key: + ret +___ +&end_function("poly1305_init_x86_64"); + +&declare_function("poly1305_blocks_x86_64", 32, 4); +$code.=<<___; +.cfi_startproc +.Lblocks: + shr \$4,$len + jz .Lno_data # too short + + push %rbx +.cfi_push %rbx + push %r12 +.cfi_push %r12 + push %r13 +.cfi_push %r13 + push %r14 +.cfi_push %r14 + push %r15 +.cfi_push %r15 + push $ctx +.cfi_push $ctx +.Lblocks_body: + + mov $len,%r15 # reassign $len + + mov 24($ctx),$r0 # load r + mov 32($ctx),$s1 + + mov 0($ctx),$h0 # load hash value + mov 8($ctx),$h1 + mov 16($ctx),$h2 + + mov $s1,$r1 + shr \$2,$s1 + mov $r1,%rax + add $r1,$s1 # s1 = r1 + (r1 >> 2) + jmp .Loop + +.align 32 +.Loop: + add 0($inp),$h0 # accumulate input + adc 8($inp),$h1 + lea 16($inp),$inp + adc $padbit,$h2 +___ + + &poly1305_iteration(); + +$code.=<<___; + mov $r1,%rax + dec %r15 # len-=16 + jnz .Loop + + mov 0(%rsp),$ctx +.cfi_restore $ctx + + mov $h0,0($ctx) # store hash value + mov $h1,8($ctx) + mov $h2,16($ctx) + + mov 8(%rsp),%r15 +.cfi_restore %r15 + mov 16(%rsp),%r14 +.cfi_restore %r14 + mov 24(%rsp),%r13 +.cfi_restore %r13 + mov 32(%rsp),%r12 +.cfi_restore %r12 + mov 40(%rsp),%rbx +.cfi_restore %rbx + lea 48(%rsp),%rsp +.cfi_adjust_cfa_offset -48 +.Lno_data: +.Lblocks_epilogue: + ret +.cfi_endproc +___ +&end_function("poly1305_blocks_x86_64"); + +&declare_function("poly1305_emit_x86_64", 32, 3); +$code.=<<___; +.Lemit: + mov 0($ctx),%r8 # load hash value + mov 8($ctx),%r9 + mov 16($ctx),%r10 + + mov %r8,%rax + add \$5,%r8 # compare to modulus + mov %r9,%rcx + adc \$0,%r9 + adc \$0,%r10 + shr \$2,%r10 # did 130-bit value overflow? + cmovnz %r8,%rax + cmovnz %r9,%rcx + + add 0($nonce),%rax # accumulate nonce + adc 8($nonce),%rcx + mov %rax,0($mac) # write result + mov %rcx,8($mac) + + ret +___ +&end_function("poly1305_emit_x86_64"); +if ($avx) { + +if($kernel) { + $code .= "#ifdef CONFIG_AS_AVX\n"; +} + +######################################################################## +# Layout of opaque area is following. +# +# unsigned __int32 h[5]; # current hash value base 2^26 +# unsigned __int32 is_base2_26; +# unsigned __int64 r[2]; # key value base 2^64 +# unsigned __int64 pad; +# struct { unsigned __int32 r^2, r^1, r^4, r^3; } r[9]; +# +# where r^n are base 2^26 digits of degrees of multiplier key. There are +# 5 digits, but last four are interleaved with multiples of 5, totalling +# in 9 elements: r0, r1, 5*r1, r2, 5*r2, r3, 5*r3, r4, 5*r4. + +my ($H0,$H1,$H2,$H3,$H4, $T0,$T1,$T2,$T3,$T4, $D0,$D1,$D2,$D3,$D4, $MASK) = + map("%xmm$_",(0..15)); + +$code.=<<___; +.type __poly1305_block,\@abi-omnipotent +.align 32 +__poly1305_block: + push $ctx +___ + &poly1305_iteration(); +$code.=<<___; + pop $ctx + ret +.size __poly1305_block,.-__poly1305_block + +.type __poly1305_init_avx,\@abi-omnipotent +.align 32 +__poly1305_init_avx: + push %rbp + mov %rsp,%rbp + mov $r0,$h0 + mov $r1,$h1 + xor $h2,$h2 + + lea 48+64($ctx),$ctx # size optimization + + mov $r1,%rax + call __poly1305_block # r^2 + + mov \$0x3ffffff,%eax # save interleaved r^2 and r base 2^26 + mov \$0x3ffffff,%edx + mov $h0,$d1 + and $h0#d,%eax + mov $r0,$d2 + and $r0#d,%edx + mov %eax,`16*0+0-64`($ctx) + shr \$26,$d1 + mov %edx,`16*0+4-64`($ctx) + shr \$26,$d2 + + mov \$0x3ffffff,%eax + mov \$0x3ffffff,%edx + and $d1#d,%eax + and $d2#d,%edx + mov %eax,`16*1+0-64`($ctx) + lea (%rax,%rax,4),%eax # *5 + mov %edx,`16*1+4-64`($ctx) + lea (%rdx,%rdx,4),%edx # *5 + mov %eax,`16*2+0-64`($ctx) + shr \$26,$d1 + mov %edx,`16*2+4-64`($ctx) + shr \$26,$d2 + + mov $h1,%rax + mov $r1,%rdx + shl \$12,%rax + shl \$12,%rdx + or $d1,%rax + or $d2,%rdx + and \$0x3ffffff,%eax + and \$0x3ffffff,%edx + mov %eax,`16*3+0-64`($ctx) + lea (%rax,%rax,4),%eax # *5 + mov %edx,`16*3+4-64`($ctx) + lea (%rdx,%rdx,4),%edx # *5 + mov %eax,`16*4+0-64`($ctx) + mov $h1,$d1 + mov %edx,`16*4+4-64`($ctx) + mov $r1,$d2 + + mov \$0x3ffffff,%eax + mov \$0x3ffffff,%edx + shr \$14,$d1 + shr \$14,$d2 + and $d1#d,%eax + and $d2#d,%edx + mov %eax,`16*5+0-64`($ctx) + lea (%rax,%rax,4),%eax # *5 + mov %edx,`16*5+4-64`($ctx) + lea (%rdx,%rdx,4),%edx # *5 + mov %eax,`16*6+0-64`($ctx) + shr \$26,$d1 + mov %edx,`16*6+4-64`($ctx) + shr \$26,$d2 + + mov $h2,%rax + shl \$24,%rax + or %rax,$d1 + mov $d1#d,`16*7+0-64`($ctx) + lea ($d1,$d1,4),$d1 # *5 + mov $d2#d,`16*7+4-64`($ctx) + lea ($d2,$d2,4),$d2 # *5 + mov $d1#d,`16*8+0-64`($ctx) + mov $d2#d,`16*8+4-64`($ctx) + + mov $r1,%rax + call __poly1305_block # r^3 + + mov \$0x3ffffff,%eax # save r^3 base 2^26 + mov $h0,$d1 + and $h0#d,%eax + shr \$26,$d1 + mov %eax,`16*0+12-64`($ctx) + + mov \$0x3ffffff,%edx + and $d1#d,%edx + mov %edx,`16*1+12-64`($ctx) + lea (%rdx,%rdx,4),%edx # *5 + shr \$26,$d1 + mov %edx,`16*2+12-64`($ctx) + + mov $h1,%rax + shl \$12,%rax + or $d1,%rax + and \$0x3ffffff,%eax + mov %eax,`16*3+12-64`($ctx) + lea (%rax,%rax,4),%eax # *5 + mov $h1,$d1 + mov %eax,`16*4+12-64`($ctx) + + mov \$0x3ffffff,%edx + shr \$14,$d1 + and $d1#d,%edx + mov %edx,`16*5+12-64`($ctx) + lea (%rdx,%rdx,4),%edx # *5 + shr \$26,$d1 + mov %edx,`16*6+12-64`($ctx) + + mov $h2,%rax + shl \$24,%rax + or %rax,$d1 + mov $d1#d,`16*7+12-64`($ctx) + lea ($d1,$d1,4),$d1 # *5 + mov $d1#d,`16*8+12-64`($ctx) + + mov $r1,%rax + call __poly1305_block # r^4 + + mov \$0x3ffffff,%eax # save r^4 base 2^26 + mov $h0,$d1 + and $h0#d,%eax + shr \$26,$d1 + mov %eax,`16*0+8-64`($ctx) + + mov \$0x3ffffff,%edx + and $d1#d,%edx + mov %edx,`16*1+8-64`($ctx) + lea (%rdx,%rdx,4),%edx # *5 + shr \$26,$d1 + mov %edx,`16*2+8-64`($ctx) + + mov $h1,%rax + shl \$12,%rax + or $d1,%rax + and \$0x3ffffff,%eax + mov %eax,`16*3+8-64`($ctx) + lea (%rax,%rax,4),%eax # *5 + mov $h1,$d1 + mov %eax,`16*4+8-64`($ctx) + + mov \$0x3ffffff,%edx + shr \$14,$d1 + and $d1#d,%edx + mov %edx,`16*5+8-64`($ctx) + lea (%rdx,%rdx,4),%edx # *5 + shr \$26,$d1 + mov %edx,`16*6+8-64`($ctx) + + mov $h2,%rax + shl \$24,%rax + or %rax,$d1 + mov $d1#d,`16*7+8-64`($ctx) + lea ($d1,$d1,4),$d1 # *5 + mov $d1#d,`16*8+8-64`($ctx) + + lea -48-64($ctx),$ctx # size [de-]optimization + pop %rbp + ret +.size __poly1305_init_avx,.-__poly1305_init_avx +___ + +&declare_function("poly1305_blocks_avx", 32, 4); +$code.=<<___; +.cfi_startproc + mov 20($ctx),%r8d # is_base2_26 + cmp \$128,$len + jae .Lblocks_avx + test %r8d,%r8d + jz .Lblocks + +.Lblocks_avx: + and \$-16,$len + jz .Lno_data_avx + + vzeroupper + + test %r8d,%r8d + jz .Lbase2_64_avx + + test \$31,$len + jz .Leven_avx + + push %rbp +.cfi_push %rbp + mov %rsp,%rbp + push %rbx +.cfi_push %rbx + push %r12 +.cfi_push %r12 + push %r13 +.cfi_push %r13 + push %r14 +.cfi_push %r14 + push %r15 +.cfi_push %r15 +.Lblocks_avx_body: + + mov $len,%r15 # reassign $len + + mov 0($ctx),$d1 # load hash value + mov 8($ctx),$d2 + mov 16($ctx),$h2#d + + mov 24($ctx),$r0 # load r + mov 32($ctx),$s1 + + ################################# base 2^26 -> base 2^64 + mov $d1#d,$h0#d + and \$`-1*(1<<31)`,$d1 + mov $d2,$r1 # borrow $r1 + mov $d2#d,$h1#d + and \$`-1*(1<<31)`,$d2 + + shr \$6,$d1 + shl \$52,$r1 + add $d1,$h0 + shr \$12,$h1 + shr \$18,$d2 + add $r1,$h0 + adc $d2,$h1 + + mov $h2,$d1 + shl \$40,$d1 + shr \$24,$h2 + add $d1,$h1 + adc \$0,$h2 # can be partially reduced... + + mov \$-4,$d2 # ... so reduce + mov $h2,$d1 + and $h2,$d2 + shr \$2,$d1 + and \$3,$h2 + add $d2,$d1 # =*5 + add $d1,$h0 + adc \$0,$h1 + adc \$0,$h2 + + mov $s1,$r1 + mov $s1,%rax + shr \$2,$s1 + add $r1,$s1 # s1 = r1 + (r1 >> 2) + + add 0($inp),$h0 # accumulate input + adc 8($inp),$h1 + lea 16($inp),$inp + adc $padbit,$h2 + + call __poly1305_block + + test $padbit,$padbit # if $padbit is zero, + jz .Lstore_base2_64_avx # store hash in base 2^64 format + + ################################# base 2^64 -> base 2^26 + mov $h0,%rax + mov $h0,%rdx + shr \$52,$h0 + mov $h1,$r0 + mov $h1,$r1 + shr \$26,%rdx + and \$0x3ffffff,%rax # h[0] + shl \$12,$r0 + and \$0x3ffffff,%rdx # h[1] + shr \$14,$h1 + or $r0,$h0 + shl \$24,$h2 + and \$0x3ffffff,$h0 # h[2] + shr \$40,$r1 + and \$0x3ffffff,$h1 # h[3] + or $r1,$h2 # h[4] + + sub \$16,%r15 + jz .Lstore_base2_26_avx + + vmovd %rax#d,$H0 + vmovd %rdx#d,$H1 + vmovd $h0#d,$H2 + vmovd $h1#d,$H3 + vmovd $h2#d,$H4 + jmp .Lproceed_avx + +.align 32 +.Lstore_base2_64_avx: + mov $h0,0($ctx) + mov $h1,8($ctx) + mov $h2,16($ctx) # note that is_base2_26 is zeroed + jmp .Ldone_avx + +.align 16 +.Lstore_base2_26_avx: + mov %rax#d,0($ctx) # store hash value base 2^26 + mov %rdx#d,4($ctx) + mov $h0#d,8($ctx) + mov $h1#d,12($ctx) + mov $h2#d,16($ctx) +.align 16 +.Ldone_avx: + pop %r15 +.cfi_restore %r15 + pop %r14 +.cfi_restore %r14 + pop %r13 +.cfi_restore %r13 + pop %r12 +.cfi_restore %r12 + pop %rbx +.cfi_restore %rbx + pop %rbp +.cfi_restore %rbp +.Lno_data_avx: +.Lblocks_avx_epilogue: + ret +.cfi_endproc + +.align 32 +.Lbase2_64_avx: +.cfi_startproc + push %rbp +.cfi_push %rbp + mov %rsp,%rbp + push %rbx +.cfi_push %rbx + push %r12 +.cfi_push %r12 + push %r13 +.cfi_push %r13 + push %r14 +.cfi_push %r14 + push %r15 +.cfi_push %r15 +.Lbase2_64_avx_body: + + mov $len,%r15 # reassign $len + + mov 24($ctx),$r0 # load r + mov 32($ctx),$s1 + + mov 0($ctx),$h0 # load hash value + mov 8($ctx),$h1 + mov 16($ctx),$h2#d + + mov $s1,$r1 + mov $s1,%rax + shr \$2,$s1 + add $r1,$s1 # s1 = r1 + (r1 >> 2) + + test \$31,$len + jz .Linit_avx + + add 0($inp),$h0 # accumulate input + adc 8($inp),$h1 + lea 16($inp),$inp + adc $padbit,$h2 + sub \$16,%r15 + + call __poly1305_block + +.Linit_avx: + ################################# base 2^64 -> base 2^26 + mov $h0,%rax + mov $h0,%rdx + shr \$52,$h0 + mov $h1,$d1 + mov $h1,$d2 + shr \$26,%rdx + and \$0x3ffffff,%rax # h[0] + shl \$12,$d1 + and \$0x3ffffff,%rdx # h[1] + shr \$14,$h1 + or $d1,$h0 + shl \$24,$h2 + and \$0x3ffffff,$h0 # h[2] + shr \$40,$d2 + and \$0x3ffffff,$h1 # h[3] + or $d2,$h2 # h[4] + + vmovd %rax#d,$H0 + vmovd %rdx#d,$H1 + vmovd $h0#d,$H2 + vmovd $h1#d,$H3 + vmovd $h2#d,$H4 + movl \$1,20($ctx) # set is_base2_26 + + call __poly1305_init_avx + +.Lproceed_avx: + mov %r15,$len + pop %r15 +.cfi_restore %r15 + pop %r14 +.cfi_restore %r14 + pop %r13 +.cfi_restore %r13 + pop %r12 +.cfi_restore %r12 + pop %rbx +.cfi_restore %rbx + pop %rbp +.cfi_restore %rbp +.Lbase2_64_avx_epilogue: + jmp .Ldo_avx +.cfi_endproc + +.align 32 +.Leven_avx: +.cfi_startproc + vmovd 4*0($ctx),$H0 # load hash value + vmovd 4*1($ctx),$H1 + vmovd 4*2($ctx),$H2 + vmovd 4*3($ctx),$H3 + vmovd 4*4($ctx),$H4 + +.Ldo_avx: +___ +$code.=<<___ if (!$win64); + lea 8(%rsp),%r10 +.cfi_def_cfa_register %r10 + and \$-32,%rsp + sub \$-8,%rsp + lea -0x58(%rsp),%r11 + sub \$0x178,%rsp +___ +$code.=<<___ if ($win64); + lea -0xf8(%rsp),%r11 + sub \$0x218,%rsp + vmovdqa %xmm6,0x50(%r11) + vmovdqa %xmm7,0x60(%r11) + vmovdqa %xmm8,0x70(%r11) + vmovdqa %xmm9,0x80(%r11) + vmovdqa %xmm10,0x90(%r11) + vmovdqa %xmm11,0xa0(%r11) + vmovdqa %xmm12,0xb0(%r11) + vmovdqa %xmm13,0xc0(%r11) + vmovdqa %xmm14,0xd0(%r11) + vmovdqa %xmm15,0xe0(%r11) +.Ldo_avx_body: +___ +$code.=<<___; + sub \$64,$len + lea -32($inp),%rax + cmovc %rax,$inp + + vmovdqu `16*3`($ctx),$D4 # preload r0^2 + lea `16*3+64`($ctx),$ctx # size optimization + lea .Lconst(%rip),%rcx + + ################################################################ + # load input + vmovdqu 16*2($inp),$T0 + vmovdqu 16*3($inp),$T1 + vmovdqa 64(%rcx),$MASK # .Lmask26 + + vpsrldq \$6,$T0,$T2 # splat input + vpsrldq \$6,$T1,$T3 + vpunpckhqdq $T1,$T0,$T4 # 4 + vpunpcklqdq $T1,$T0,$T0 # 0:1 + vpunpcklqdq $T3,$T2,$T3 # 2:3 + + vpsrlq \$40,$T4,$T4 # 4 + vpsrlq \$26,$T0,$T1 + vpand $MASK,$T0,$T0 # 0 + vpsrlq \$4,$T3,$T2 + vpand $MASK,$T1,$T1 # 1 + vpsrlq \$30,$T3,$T3 + vpand $MASK,$T2,$T2 # 2 + vpand $MASK,$T3,$T3 # 3 + vpor 32(%rcx),$T4,$T4 # padbit, yes, always + + jbe .Lskip_loop_avx + + # expand and copy pre-calculated table to stack + vmovdqu `16*1-64`($ctx),$D1 + vmovdqu `16*2-64`($ctx),$D2 + vpshufd \$0xEE,$D4,$D3 # 34xx -> 3434 + vpshufd \$0x44,$D4,$D0 # xx12 -> 1212 + vmovdqa $D3,-0x90(%r11) + vmovdqa $D0,0x00(%rsp) + vpshufd \$0xEE,$D1,$D4 + vmovdqu `16*3-64`($ctx),$D0 + vpshufd \$0x44,$D1,$D1 + vmovdqa $D4,-0x80(%r11) + vmovdqa $D1,0x10(%rsp) + vpshufd \$0xEE,$D2,$D3 + vmovdqu `16*4-64`($ctx),$D1 + vpshufd \$0x44,$D2,$D2 + vmovdqa $D3,-0x70(%r11) + vmovdqa $D2,0x20(%rsp) + vpshufd \$0xEE,$D0,$D4 + vmovdqu `16*5-64`($ctx),$D2 + vpshufd \$0x44,$D0,$D0 + vmovdqa $D4,-0x60(%r11) + vmovdqa $D0,0x30(%rsp) + vpshufd \$0xEE,$D1,$D3 + vmovdqu `16*6-64`($ctx),$D0 + vpshufd \$0x44,$D1,$D1 + vmovdqa $D3,-0x50(%r11) + vmovdqa $D1,0x40(%rsp) + vpshufd \$0xEE,$D2,$D4 + vmovdqu `16*7-64`($ctx),$D1 + vpshufd \$0x44,$D2,$D2 + vmovdqa $D4,-0x40(%r11) + vmovdqa $D2,0x50(%rsp) + vpshufd \$0xEE,$D0,$D3 + vmovdqu `16*8-64`($ctx),$D2 + vpshufd \$0x44,$D0,$D0 + vmovdqa $D3,-0x30(%r11) + vmovdqa $D0,0x60(%rsp) + vpshufd \$0xEE,$D1,$D4 + vpshufd \$0x44,$D1,$D1 + vmovdqa $D4,-0x20(%r11) + vmovdqa $D1,0x70(%rsp) + vpshufd \$0xEE,$D2,$D3 + vmovdqa 0x00(%rsp),$D4 # preload r0^2 + vpshufd \$0x44,$D2,$D2 + vmovdqa $D3,-0x10(%r11) + vmovdqa $D2,0x80(%rsp) + + jmp .Loop_avx + +.align 32 +.Loop_avx: + ################################################################ + # ((inp[0]*r^4+inp[2]*r^2+inp[4])*r^4+inp[6]*r^2 + # ((inp[1]*r^4+inp[3]*r^2+inp[5])*r^3+inp[7]*r + # \___________________/ + # ((inp[0]*r^4+inp[2]*r^2+inp[4])*r^4+inp[6]*r^2+inp[8])*r^2 + # ((inp[1]*r^4+inp[3]*r^2+inp[5])*r^4+inp[7]*r^2+inp[9])*r + # \___________________/ \____________________/ + # + # Note that we start with inp[2:3]*r^2. This is because it + # doesn't depend on reduction in previous iteration. + ################################################################ + # d4 = h4*r0 + h3*r1 + h2*r2 + h1*r3 + h0*r4 + # d3 = h3*r0 + h2*r1 + h1*r2 + h0*r3 + h4*5*r4 + # d2 = h2*r0 + h1*r1 + h0*r2 + h4*5*r3 + h3*5*r4 + # d1 = h1*r0 + h0*r1 + h4*5*r2 + h3*5*r3 + h2*5*r4 + # d0 = h0*r0 + h4*5*r1 + h3*5*r2 + h2*5*r3 + h1*5*r4 + # + # though note that $Tx and $Hx are "reversed" in this section, + # and $D4 is preloaded with r0^2... + + vpmuludq $T0,$D4,$D0 # d0 = h0*r0 + vpmuludq $T1,$D4,$D1 # d1 = h1*r0 + vmovdqa $H2,0x20(%r11) # offload hash + vpmuludq $T2,$D4,$D2 # d3 = h2*r0 + vmovdqa 0x10(%rsp),$H2 # r1^2 + vpmuludq $T3,$D4,$D3 # d3 = h3*r0 + vpmuludq $T4,$D4,$D4 # d4 = h4*r0 + + vmovdqa $H0,0x00(%r11) # + vpmuludq 0x20(%rsp),$T4,$H0 # h4*s1 + vmovdqa $H1,0x10(%r11) # + vpmuludq $T3,$H2,$H1 # h3*r1 + vpaddq $H0,$D0,$D0 # d0 += h4*s1 + vpaddq $H1,$D4,$D4 # d4 += h3*r1 + vmovdqa $H3,0x30(%r11) # + vpmuludq $T2,$H2,$H0 # h2*r1 + vpmuludq $T1,$H2,$H1 # h1*r1 + vpaddq $H0,$D3,$D3 # d3 += h2*r1 + vmovdqa 0x30(%rsp),$H3 # r2^2 + vpaddq $H1,$D2,$D2 # d2 += h1*r1 + vmovdqa $H4,0x40(%r11) # + vpmuludq $T0,$H2,$H2 # h0*r1 + vpmuludq $T2,$H3,$H0 # h2*r2 + vpaddq $H2,$D1,$D1 # d1 += h0*r1 + + vmovdqa 0x40(%rsp),$H4 # s2^2 + vpaddq $H0,$D4,$D4 # d4 += h2*r2 + vpmuludq $T1,$H3,$H1 # h1*r2 + vpmuludq $T0,$H3,$H3 # h0*r2 + vpaddq $H1,$D3,$D3 # d3 += h1*r2 + vmovdqa 0x50(%rsp),$H2 # r3^2 + vpaddq $H3,$D2,$D2 # d2 += h0*r2 + vpmuludq $T4,$H4,$H0 # h4*s2 + vpmuludq $T3,$H4,$H4 # h3*s2 + vpaddq $H0,$D1,$D1 # d1 += h4*s2 + vmovdqa 0x60(%rsp),$H3 # s3^2 + vpaddq $H4,$D0,$D0 # d0 += h3*s2 + + vmovdqa 0x80(%rsp),$H4 # s4^2 + vpmuludq $T1,$H2,$H1 # h1*r3 + vpmuludq $T0,$H2,$H2 # h0*r3 + vpaddq $H1,$D4,$D4 # d4 += h1*r3 + vpaddq $H2,$D3,$D3 # d3 += h0*r3 + vpmuludq $T4,$H3,$H0 # h4*s3 + vpmuludq $T3,$H3,$H1 # h3*s3 + vpaddq $H0,$D2,$D2 # d2 += h4*s3 + vmovdqu 16*0($inp),$H0 # load input + vpaddq $H1,$D1,$D1 # d1 += h3*s3 + vpmuludq $T2,$H3,$H3 # h2*s3 + vpmuludq $T2,$H4,$T2 # h2*s4 + vpaddq $H3,$D0,$D0 # d0 += h2*s3 + + vmovdqu 16*1($inp),$H1 # + vpaddq $T2,$D1,$D1 # d1 += h2*s4 + vpmuludq $T3,$H4,$T3 # h3*s4 + vpmuludq $T4,$H4,$T4 # h4*s4 + vpsrldq \$6,$H0,$H2 # splat input + vpaddq $T3,$D2,$D2 # d2 += h3*s4 + vpaddq $T4,$D3,$D3 # d3 += h4*s4 + vpsrldq \$6,$H1,$H3 # + vpmuludq 0x70(%rsp),$T0,$T4 # h0*r4 + vpmuludq $T1,$H4,$T0 # h1*s4 + vpunpckhqdq $H1,$H0,$H4 # 4 + vpaddq $T4,$D4,$D4 # d4 += h0*r4 + vmovdqa -0x90(%r11),$T4 # r0^4 + vpaddq $T0,$D0,$D0 # d0 += h1*s4 + + vpunpcklqdq $H1,$H0,$H0 # 0:1 + vpunpcklqdq $H3,$H2,$H3 # 2:3 + + #vpsrlq \$40,$H4,$H4 # 4 + vpsrldq \$`40/8`,$H4,$H4 # 4 + vpsrlq \$26,$H0,$H1 + vpand $MASK,$H0,$H0 # 0 + vpsrlq \$4,$H3,$H2 + vpand $MASK,$H1,$H1 # 1 + vpand 0(%rcx),$H4,$H4 # .Lmask24 + vpsrlq \$30,$H3,$H3 + vpand $MASK,$H2,$H2 # 2 + vpand $MASK,$H3,$H3 # 3 + vpor 32(%rcx),$H4,$H4 # padbit, yes, always + + vpaddq 0x00(%r11),$H0,$H0 # add hash value + vpaddq 0x10(%r11),$H1,$H1 + vpaddq 0x20(%r11),$H2,$H2 + vpaddq 0x30(%r11),$H3,$H3 + vpaddq 0x40(%r11),$H4,$H4 + + lea 16*2($inp),%rax + lea 16*4($inp),$inp + sub \$64,$len + cmovc %rax,$inp + + ################################################################ + # Now we accumulate (inp[0:1]+hash)*r^4 + ################################################################ + # d4 = h4*r0 + h3*r1 + h2*r2 + h1*r3 + h0*r4 + # d3 = h3*r0 + h2*r1 + h1*r2 + h0*r3 + h4*5*r4 + # d2 = h2*r0 + h1*r1 + h0*r2 + h4*5*r3 + h3*5*r4 + # d1 = h1*r0 + h0*r1 + h4*5*r2 + h3*5*r3 + h2*5*r4 + # d0 = h0*r0 + h4*5*r1 + h3*5*r2 + h2*5*r3 + h1*5*r4 + + vpmuludq $H0,$T4,$T0 # h0*r0 + vpmuludq $H1,$T4,$T1 # h1*r0 + vpaddq $T0,$D0,$D0 + vpaddq $T1,$D1,$D1 + vmovdqa -0x80(%r11),$T2 # r1^4 + vpmuludq $H2,$T4,$T0 # h2*r0 + vpmuludq $H3,$T4,$T1 # h3*r0 + vpaddq $T0,$D2,$D2 + vpaddq $T1,$D3,$D3 + vpmuludq $H4,$T4,$T4 # h4*r0 + vpmuludq -0x70(%r11),$H4,$T0 # h4*s1 + vpaddq $T4,$D4,$D4 + + vpaddq $T0,$D0,$D0 # d0 += h4*s1 + vpmuludq $H2,$T2,$T1 # h2*r1 + vpmuludq $H3,$T2,$T0 # h3*r1 + vpaddq $T1,$D3,$D3 # d3 += h2*r1 + vmovdqa -0x60(%r11),$T3 # r2^4 + vpaddq $T0,$D4,$D4 # d4 += h3*r1 + vpmuludq $H1,$T2,$T1 # h1*r1 + vpmuludq $H0,$T2,$T2 # h0*r1 + vpaddq $T1,$D2,$D2 # d2 += h1*r1 + vpaddq $T2,$D1,$D1 # d1 += h0*r1 + + vmovdqa -0x50(%r11),$T4 # s2^4 + vpmuludq $H2,$T3,$T0 # h2*r2 + vpmuludq $H1,$T3,$T1 # h1*r2 + vpaddq $T0,$D4,$D4 # d4 += h2*r2 + vpaddq $T1,$D3,$D3 # d3 += h1*r2 + vmovdqa -0x40(%r11),$T2 # r3^4 + vpmuludq $H0,$T3,$T3 # h0*r2 + vpmuludq $H4,$T4,$T0 # h4*s2 + vpaddq $T3,$D2,$D2 # d2 += h0*r2 + vpaddq $T0,$D1,$D1 # d1 += h4*s2 + vmovdqa -0x30(%r11),$T3 # s3^4 + vpmuludq $H3,$T4,$T4 # h3*s2 + vpmuludq $H1,$T2,$T1 # h1*r3 + vpaddq $T4,$D0,$D0 # d0 += h3*s2 + + vmovdqa -0x10(%r11),$T4 # s4^4 + vpaddq $T1,$D4,$D4 # d4 += h1*r3 + vpmuludq $H0,$T2,$T2 # h0*r3 + vpmuludq $H4,$T3,$T0 # h4*s3 + vpaddq $T2,$D3,$D3 # d3 += h0*r3 + vpaddq $T0,$D2,$D2 # d2 += h4*s3 + vmovdqu 16*2($inp),$T0 # load input + vpmuludq $H3,$T3,$T2 # h3*s3 + vpmuludq $H2,$T3,$T3 # h2*s3 + vpaddq $T2,$D1,$D1 # d1 += h3*s3 + vmovdqu 16*3($inp),$T1 # + vpaddq $T3,$D0,$D0 # d0 += h2*s3 + + vpmuludq $H2,$T4,$H2 # h2*s4 + vpmuludq $H3,$T4,$H3 # h3*s4 + vpsrldq \$6,$T0,$T2 # splat input + vpaddq $H2,$D1,$D1 # d1 += h2*s4 + vpmuludq $H4,$T4,$H4 # h4*s4 + vpsrldq \$6,$T1,$T3 # + vpaddq $H3,$D2,$H2 # h2 = d2 + h3*s4 + vpaddq $H4,$D3,$H3 # h3 = d3 + h4*s4 + vpmuludq -0x20(%r11),$H0,$H4 # h0*r4 + vpmuludq $H1,$T4,$H0 + vpunpckhqdq $T1,$T0,$T4 # 4 + vpaddq $H4,$D4,$H4 # h4 = d4 + h0*r4 + vpaddq $H0,$D0,$H0 # h0 = d0 + h1*s4 + + vpunpcklqdq $T1,$T0,$T0 # 0:1 + vpunpcklqdq $T3,$T2,$T3 # 2:3 + + #vpsrlq \$40,$T4,$T4 # 4 + vpsrldq \$`40/8`,$T4,$T4 # 4 + vpsrlq \$26,$T0,$T1 + vmovdqa 0x00(%rsp),$D4 # preload r0^2 + vpand $MASK,$T0,$T0 # 0 + vpsrlq \$4,$T3,$T2 + vpand $MASK,$T1,$T1 # 1 + vpand 0(%rcx),$T4,$T4 # .Lmask24 + vpsrlq \$30,$T3,$T3 + vpand $MASK,$T2,$T2 # 2 + vpand $MASK,$T3,$T3 # 3 + vpor 32(%rcx),$T4,$T4 # padbit, yes, always + + ################################################################ + # lazy reduction as discussed in "NEON crypto" by D.J. Bernstein + # and P. Schwabe + + vpsrlq \$26,$H3,$D3 + vpand $MASK,$H3,$H3 + vpaddq $D3,$H4,$H4 # h3 -> h4 + + vpsrlq \$26,$H0,$D0 + vpand $MASK,$H0,$H0 + vpaddq $D0,$D1,$H1 # h0 -> h1 + + vpsrlq \$26,$H4,$D0 + vpand $MASK,$H4,$H4 + + vpsrlq \$26,$H1,$D1 + vpand $MASK,$H1,$H1 + vpaddq $D1,$H2,$H2 # h1 -> h2 + + vpaddq $D0,$H0,$H0 + vpsllq \$2,$D0,$D0 + vpaddq $D0,$H0,$H0 # h4 -> h0 + + vpsrlq \$26,$H2,$D2 + vpand $MASK,$H2,$H2 + vpaddq $D2,$H3,$H3 # h2 -> h3 + + vpsrlq \$26,$H0,$D0 + vpand $MASK,$H0,$H0 + vpaddq $D0,$H1,$H1 # h0 -> h1 + + vpsrlq \$26,$H3,$D3 + vpand $MASK,$H3,$H3 + vpaddq $D3,$H4,$H4 # h3 -> h4 + + ja .Loop_avx + +.Lskip_loop_avx: + ################################################################ + # multiply (inp[0:1]+hash) or inp[2:3] by r^2:r^1 + + vpshufd \$0x10,$D4,$D4 # r0^n, xx12 -> x1x2 + add \$32,$len + jnz .Long_tail_avx + + vpaddq $H2,$T2,$T2 + vpaddq $H0,$T0,$T0 + vpaddq $H1,$T1,$T1 + vpaddq $H3,$T3,$T3 + vpaddq $H4,$T4,$T4 + +.Long_tail_avx: + vmovdqa $H2,0x20(%r11) + vmovdqa $H0,0x00(%r11) + vmovdqa $H1,0x10(%r11) + vmovdqa $H3,0x30(%r11) + vmovdqa $H4,0x40(%r11) + + # d4 = h4*r0 + h3*r1 + h2*r2 + h1*r3 + h0*r4 + # d3 = h3*r0 + h2*r1 + h1*r2 + h0*r3 + h4*5*r4 + # d2 = h2*r0 + h1*r1 + h0*r2 + h4*5*r3 + h3*5*r4 + # d1 = h1*r0 + h0*r1 + h4*5*r2 + h3*5*r3 + h2*5*r4 + # d0 = h0*r0 + h4*5*r1 + h3*5*r2 + h2*5*r3 + h1*5*r4 + + vpmuludq $T2,$D4,$D2 # d2 = h2*r0 + vpmuludq $T0,$D4,$D0 # d0 = h0*r0 + vpshufd \$0x10,`16*1-64`($ctx),$H2 # r1^n + vpmuludq $T1,$D4,$D1 # d1 = h1*r0 + vpmuludq $T3,$D4,$D3 # d3 = h3*r0 + vpmuludq $T4,$D4,$D4 # d4 = h4*r0 + + vpmuludq $T3,$H2,$H0 # h3*r1 + vpaddq $H0,$D4,$D4 # d4 += h3*r1 + vpshufd \$0x10,`16*2-64`($ctx),$H3 # s1^n + vpmuludq $T2,$H2,$H1 # h2*r1 + vpaddq $H1,$D3,$D3 # d3 += h2*r1 + vpshufd \$0x10,`16*3-64`($ctx),$H4 # r2^n + vpmuludq $T1,$H2,$H0 # h1*r1 + vpaddq $H0,$D2,$D2 # d2 += h1*r1 + vpmuludq $T0,$H2,$H2 # h0*r1 + vpaddq $H2,$D1,$D1 # d1 += h0*r1 + vpmuludq $T4,$H3,$H3 # h4*s1 + vpaddq $H3,$D0,$D0 # d0 += h4*s1 + + vpshufd \$0x10,`16*4-64`($ctx),$H2 # s2^n + vpmuludq $T2,$H4,$H1 # h2*r2 + vpaddq $H1,$D4,$D4 # d4 += h2*r2 + vpmuludq $T1,$H4,$H0 # h1*r2 + vpaddq $H0,$D3,$D3 # d3 += h1*r2 + vpshufd \$0x10,`16*5-64`($ctx),$H3 # r3^n + vpmuludq $T0,$H4,$H4 # h0*r2 + vpaddq $H4,$D2,$D2 # d2 += h0*r2 + vpmuludq $T4,$H2,$H1 # h4*s2 + vpaddq $H1,$D1,$D1 # d1 += h4*s2 + vpshufd \$0x10,`16*6-64`($ctx),$H4 # s3^n + vpmuludq $T3,$H2,$H2 # h3*s2 + vpaddq $H2,$D0,$D0 # d0 += h3*s2 + + vpmuludq $T1,$H3,$H0 # h1*r3 + vpaddq $H0,$D4,$D4 # d4 += h1*r3 + vpmuludq $T0,$H3,$H3 # h0*r3 + vpaddq $H3,$D3,$D3 # d3 += h0*r3 + vpshufd \$0x10,`16*7-64`($ctx),$H2 # r4^n + vpmuludq $T4,$H4,$H1 # h4*s3 + vpaddq $H1,$D2,$D2 # d2 += h4*s3 + vpshufd \$0x10,`16*8-64`($ctx),$H3 # s4^n + vpmuludq $T3,$H4,$H0 # h3*s3 + vpaddq $H0,$D1,$D1 # d1 += h3*s3 + vpmuludq $T2,$H4,$H4 # h2*s3 + vpaddq $H4,$D0,$D0 # d0 += h2*s3 + + vpmuludq $T0,$H2,$H2 # h0*r4 + vpaddq $H2,$D4,$D4 # h4 = d4 + h0*r4 + vpmuludq $T4,$H3,$H1 # h4*s4 + vpaddq $H1,$D3,$D3 # h3 = d3 + h4*s4 + vpmuludq $T3,$H3,$H0 # h3*s4 + vpaddq $H0,$D2,$D2 # h2 = d2 + h3*s4 + vpmuludq $T2,$H3,$H1 # h2*s4 + vpaddq $H1,$D1,$D1 # h1 = d1 + h2*s4 + vpmuludq $T1,$H3,$H3 # h1*s4 + vpaddq $H3,$D0,$D0 # h0 = d0 + h1*s4 + + jz .Lshort_tail_avx + + vmovdqu 16*0($inp),$H0 # load input + vmovdqu 16*1($inp),$H1 + + vpsrldq \$6,$H0,$H2 # splat input + vpsrldq \$6,$H1,$H3 + vpunpckhqdq $H1,$H0,$H4 # 4 + vpunpcklqdq $H1,$H0,$H0 # 0:1 + vpunpcklqdq $H3,$H2,$H3 # 2:3 + + vpsrlq \$40,$H4,$H4 # 4 + vpsrlq \$26,$H0,$H1 + vpand $MASK,$H0,$H0 # 0 + vpsrlq \$4,$H3,$H2 + vpand $MASK,$H1,$H1 # 1 + vpsrlq \$30,$H3,$H3 + vpand $MASK,$H2,$H2 # 2 + vpand $MASK,$H3,$H3 # 3 + vpor 32(%rcx),$H4,$H4 # padbit, yes, always + + vpshufd \$0x32,`16*0-64`($ctx),$T4 # r0^n, 34xx -> x3x4 + vpaddq 0x00(%r11),$H0,$H0 + vpaddq 0x10(%r11),$H1,$H1 + vpaddq 0x20(%r11),$H2,$H2 + vpaddq 0x30(%r11),$H3,$H3 + vpaddq 0x40(%r11),$H4,$H4 + + ################################################################ + # multiply (inp[0:1]+hash) by r^4:r^3 and accumulate + + vpmuludq $H0,$T4,$T0 # h0*r0 + vpaddq $T0,$D0,$D0 # d0 += h0*r0 + vpmuludq $H1,$T4,$T1 # h1*r0 + vpaddq $T1,$D1,$D1 # d1 += h1*r0 + vpmuludq $H2,$T4,$T0 # h2*r0 + vpaddq $T0,$D2,$D2 # d2 += h2*r0 + vpshufd \$0x32,`16*1-64`($ctx),$T2 # r1^n + vpmuludq $H3,$T4,$T1 # h3*r0 + vpaddq $T1,$D3,$D3 # d3 += h3*r0 + vpmuludq $H4,$T4,$T4 # h4*r0 + vpaddq $T4,$D4,$D4 # d4 += h4*r0 + + vpmuludq $H3,$T2,$T0 # h3*r1 + vpaddq $T0,$D4,$D4 # d4 += h3*r1 + vpshufd \$0x32,`16*2-64`($ctx),$T3 # s1 + vpmuludq $H2,$T2,$T1 # h2*r1 + vpaddq $T1,$D3,$D3 # d3 += h2*r1 + vpshufd \$0x32,`16*3-64`($ctx),$T4 # r2 + vpmuludq $H1,$T2,$T0 # h1*r1 + vpaddq $T0,$D2,$D2 # d2 += h1*r1 + vpmuludq $H0,$T2,$T2 # h0*r1 + vpaddq $T2,$D1,$D1 # d1 += h0*r1 + vpmuludq $H4,$T3,$T3 # h4*s1 + vpaddq $T3,$D0,$D0 # d0 += h4*s1 + + vpshufd \$0x32,`16*4-64`($ctx),$T2 # s2 + vpmuludq $H2,$T4,$T1 # h2*r2 + vpaddq $T1,$D4,$D4 # d4 += h2*r2 + vpmuludq $H1,$T4,$T0 # h1*r2 + vpaddq $T0,$D3,$D3 # d3 += h1*r2 + vpshufd \$0x32,`16*5-64`($ctx),$T3 # r3 + vpmuludq $H0,$T4,$T4 # h0*r2 + vpaddq $T4,$D2,$D2 # d2 += h0*r2 + vpmuludq $H4,$T2,$T1 # h4*s2 + vpaddq $T1,$D1,$D1 # d1 += h4*s2 + vpshufd \$0x32,`16*6-64`($ctx),$T4 # s3 + vpmuludq $H3,$T2,$T2 # h3*s2 + vpaddq $T2,$D0,$D0 # d0 += h3*s2 + + vpmuludq $H1,$T3,$T0 # h1*r3 + vpaddq $T0,$D4,$D4 # d4 += h1*r3 + vpmuludq $H0,$T3,$T3 # h0*r3 + vpaddq $T3,$D3,$D3 # d3 += h0*r3 + vpshufd \$0x32,`16*7-64`($ctx),$T2 # r4 + vpmuludq $H4,$T4,$T1 # h4*s3 + vpaddq $T1,$D2,$D2 # d2 += h4*s3 + vpshufd \$0x32,`16*8-64`($ctx),$T3 # s4 + vpmuludq $H3,$T4,$T0 # h3*s3 + vpaddq $T0,$D1,$D1 # d1 += h3*s3 + vpmuludq $H2,$T4,$T4 # h2*s3 + vpaddq $T4,$D0,$D0 # d0 += h2*s3 + + vpmuludq $H0,$T2,$T2 # h0*r4 + vpaddq $T2,$D4,$D4 # d4 += h0*r4 + vpmuludq $H4,$T3,$T1 # h4*s4 + vpaddq $T1,$D3,$D3 # d3 += h4*s4 + vpmuludq $H3,$T3,$T0 # h3*s4 + vpaddq $T0,$D2,$D2 # d2 += h3*s4 + vpmuludq $H2,$T3,$T1 # h2*s4 + vpaddq $T1,$D1,$D1 # d1 += h2*s4 + vpmuludq $H1,$T3,$T3 # h1*s4 + vpaddq $T3,$D0,$D0 # d0 += h1*s4 + +.Lshort_tail_avx: + ################################################################ + # horizontal addition + + vpsrldq \$8,$D4,$T4 + vpsrldq \$8,$D3,$T3 + vpsrldq \$8,$D1,$T1 + vpsrldq \$8,$D0,$T0 + vpsrldq \$8,$D2,$T2 + vpaddq $T3,$D3,$D3 + vpaddq $T4,$D4,$D4 + vpaddq $T0,$D0,$D0 + vpaddq $T1,$D1,$D1 + vpaddq $T2,$D2,$D2 + + ################################################################ + # lazy reduction + + vpsrlq \$26,$D3,$H3 + vpand $MASK,$D3,$D3 + vpaddq $H3,$D4,$D4 # h3 -> h4 + + vpsrlq \$26,$D0,$H0 + vpand $MASK,$D0,$D0 + vpaddq $H0,$D1,$D1 # h0 -> h1 + + vpsrlq \$26,$D4,$H4 + vpand $MASK,$D4,$D4 + + vpsrlq \$26,$D1,$H1 + vpand $MASK,$D1,$D1 + vpaddq $H1,$D2,$D2 # h1 -> h2 + + vpaddq $H4,$D0,$D0 + vpsllq \$2,$H4,$H4 + vpaddq $H4,$D0,$D0 # h4 -> h0 + + vpsrlq \$26,$D2,$H2 + vpand $MASK,$D2,$D2 + vpaddq $H2,$D3,$D3 # h2 -> h3 + + vpsrlq \$26,$D0,$H0 + vpand $MASK,$D0,$D0 + vpaddq $H0,$D1,$D1 # h0 -> h1 + + vpsrlq \$26,$D3,$H3 + vpand $MASK,$D3,$D3 + vpaddq $H3,$D4,$D4 # h3 -> h4 + + vmovd $D0,`4*0-48-64`($ctx) # save partially reduced + vmovd $D1,`4*1-48-64`($ctx) + vmovd $D2,`4*2-48-64`($ctx) + vmovd $D3,`4*3-48-64`($ctx) + vmovd $D4,`4*4-48-64`($ctx) +___ +$code.=<<___ if ($win64); + vmovdqa 0x50(%r11),%xmm6 + vmovdqa 0x60(%r11),%xmm7 + vmovdqa 0x70(%r11),%xmm8 + vmovdqa 0x80(%r11),%xmm9 + vmovdqa 0x90(%r11),%xmm10 + vmovdqa 0xa0(%r11),%xmm11 + vmovdqa 0xb0(%r11),%xmm12 + vmovdqa 0xc0(%r11),%xmm13 + vmovdqa 0xd0(%r11),%xmm14 + vmovdqa 0xe0(%r11),%xmm15 + lea 0xf8(%r11),%rsp +.Ldo_avx_epilogue: +___ +$code.=<<___ if (!$win64); + lea -8(%r10),%rsp +.cfi_def_cfa_register %rsp +___ +$code.=<<___; + vzeroupper + ret +.cfi_endproc +___ +&end_function("poly1305_blocks_avx"); + +&declare_function("poly1305_emit_avx", 32, 3); +$code.=<<___; + cmpl \$0,20($ctx) # is_base2_26? + je .Lemit + + mov 0($ctx),%eax # load hash value base 2^26 + mov 4($ctx),%ecx + mov 8($ctx),%r8d + mov 12($ctx),%r11d + mov 16($ctx),%r10d + + shl \$26,%rcx # base 2^26 -> base 2^64 + mov %r8,%r9 + shl \$52,%r8 + add %rcx,%rax + shr \$12,%r9 + add %rax,%r8 # h0 + adc \$0,%r9 + + shl \$14,%r11 + mov %r10,%rax + shr \$24,%r10 + add %r11,%r9 + shl \$40,%rax + add %rax,%r9 # h1 + adc \$0,%r10 # h2 + + mov %r10,%rax # could be partially reduced, so reduce + mov %r10,%rcx + and \$3,%r10 + shr \$2,%rax + and \$-4,%rcx + add %rcx,%rax + add %rax,%r8 + adc \$0,%r9 + adc \$0,%r10 + + mov %r8,%rax + add \$5,%r8 # compare to modulus + mov %r9,%rcx + adc \$0,%r9 + adc \$0,%r10 + shr \$2,%r10 # did 130-bit value overflow? + cmovnz %r8,%rax + cmovnz %r9,%rcx + + add 0($nonce),%rax # accumulate nonce + adc 8($nonce),%rcx + mov %rax,0($mac) # write result + mov %rcx,8($mac) + + ret +___ +&end_function("poly1305_emit_avx"); + +if ($kernel) { + $code .= "#endif\n"; +} + +if ($avx>1) { + +if ($kernel) { + $code .= "#ifdef CONFIG_AS_AVX2\n"; +} + +my ($H0,$H1,$H2,$H3,$H4, $MASK, $T4,$T0,$T1,$T2,$T3, $D0,$D1,$D2,$D3,$D4) = + map("%ymm$_",(0..15)); +my $S4=$MASK; + +sub poly1305_blocks_avxN { + my ($avx512) = @_; + my $suffix = $avx512 ? "_avx512" : ""; +$code.=<<___; +.cfi_startproc + mov 20($ctx),%r8d # is_base2_26 + cmp \$128,$len + jae .Lblocks_avx2$suffix + test %r8d,%r8d + jz .Lblocks + +.Lblocks_avx2$suffix: + and \$-16,$len + jz .Lno_data_avx2$suffix + + vzeroupper + + test %r8d,%r8d + jz .Lbase2_64_avx2$suffix + + test \$63,$len + jz .Leven_avx2$suffix + + push %rbp +.cfi_push %rbp + mov %rsp,%rbp + push %rbx +.cfi_push %rbx + push %r12 +.cfi_push %r12 + push %r13 +.cfi_push %r13 + push %r14 +.cfi_push %r14 + push %r15 +.cfi_push %r15 +.Lblocks_avx2_body$suffix: + + mov $len,%r15 # reassign $len + + mov 0($ctx),$d1 # load hash value + mov 8($ctx),$d2 + mov 16($ctx),$h2#d + + mov 24($ctx),$r0 # load r + mov 32($ctx),$s1 + + ################################# base 2^26 -> base 2^64 + mov $d1#d,$h0#d + and \$`-1*(1<<31)`,$d1 + mov $d2,$r1 # borrow $r1 + mov $d2#d,$h1#d + and \$`-1*(1<<31)`,$d2 + + shr \$6,$d1 + shl \$52,$r1 + add $d1,$h0 + shr \$12,$h1 + shr \$18,$d2 + add $r1,$h0 + adc $d2,$h1 + + mov $h2,$d1 + shl \$40,$d1 + shr \$24,$h2 + add $d1,$h1 + adc \$0,$h2 # can be partially reduced... + + mov \$-4,$d2 # ... so reduce + mov $h2,$d1 + and $h2,$d2 + shr \$2,$d1 + and \$3,$h2 + add $d2,$d1 # =*5 + add $d1,$h0 + adc \$0,$h1 + adc \$0,$h2 + + mov $s1,$r1 + mov $s1,%rax + shr \$2,$s1 + add $r1,$s1 # s1 = r1 + (r1 >> 2) + +.Lbase2_26_pre_avx2$suffix: + add 0($inp),$h0 # accumulate input + adc 8($inp),$h1 + lea 16($inp),$inp + adc $padbit,$h2 + sub \$16,%r15 + + call __poly1305_block + mov $r1,%rax + + test \$63,%r15 + jnz .Lbase2_26_pre_avx2$suffix + + test $padbit,$padbit # if $padbit is zero, + jz .Lstore_base2_64_avx2$suffix # store hash in base 2^64 format + + ################################# base 2^64 -> base 2^26 + mov $h0,%rax + mov $h0,%rdx + shr \$52,$h0 + mov $h1,$r0 + mov $h1,$r1 + shr \$26,%rdx + and \$0x3ffffff,%rax # h[0] + shl \$12,$r0 + and \$0x3ffffff,%rdx # h[1] + shr \$14,$h1 + or $r0,$h0 + shl \$24,$h2 + and \$0x3ffffff,$h0 # h[2] + shr \$40,$r1 + and \$0x3ffffff,$h1 # h[3] + or $r1,$h2 # h[4] + + test %r15,%r15 + jz .Lstore_base2_26_avx2$suffix + + vmovd %rax#d,%x#$H0 + vmovd %rdx#d,%x#$H1 + vmovd $h0#d,%x#$H2 + vmovd $h1#d,%x#$H3 + vmovd $h2#d,%x#$H4 + jmp .Lproceed_avx2$suffix + +.align 32 +.Lstore_base2_64_avx2$suffix: + mov $h0,0($ctx) + mov $h1,8($ctx) + mov $h2,16($ctx) # note that is_base2_26 is zeroed + jmp .Ldone_avx2$suffix + +.align 16 +.Lstore_base2_26_avx2$suffix: + mov %rax#d,0($ctx) # store hash value base 2^26 + mov %rdx#d,4($ctx) + mov $h0#d,8($ctx) + mov $h1#d,12($ctx) + mov $h2#d,16($ctx) +.align 16 +.Ldone_avx2$suffix: + pop %r15 +.cfi_restore %r15 + pop %r14 +.cfi_restore %r14 + pop %r13 +.cfi_restore %r13 + pop %r12 +.cfi_restore %r12 + pop %rbx +.cfi_restore %rbx + pop %rbp +.cfi_restore %rbp +.Lno_data_avx2$suffix: +.Lblocks_avx2_epilogue$suffix: + ret +.cfi_endproc + +.align 32 +.Lbase2_64_avx2$suffix: +.cfi_startproc + push %rbp +.cfi_push %rbp + mov %rsp,%rbp + push %rbx +.cfi_push %rbx + push %r12 +.cfi_push %r12 + push %r13 +.cfi_push %r13 + push %r14 +.cfi_push %r14 + push %r15 +.cfi_push %r15 +.Lbase2_64_avx2_body$suffix: + + mov $len,%r15 # reassign $len + + mov 24($ctx),$r0 # load r + mov 32($ctx),$s1 + + mov 0($ctx),$h0 # load hash value + mov 8($ctx),$h1 + mov 16($ctx),$h2#d + + mov $s1,$r1 + mov $s1,%rax + shr \$2,$s1 + add $r1,$s1 # s1 = r1 + (r1 >> 2) + + test \$63,$len + jz .Linit_avx2$suffix + +.Lbase2_64_pre_avx2$suffix: + add 0($inp),$h0 # accumulate input + adc 8($inp),$h1 + lea 16($inp),$inp + adc $padbit,$h2 + sub \$16,%r15 + + call __poly1305_block + mov $r1,%rax + + test \$63,%r15 + jnz .Lbase2_64_pre_avx2$suffix + +.Linit_avx2$suffix: + ################################# base 2^64 -> base 2^26 + mov $h0,%rax + mov $h0,%rdx + shr \$52,$h0 + mov $h1,$d1 + mov $h1,$d2 + shr \$26,%rdx + and \$0x3ffffff,%rax # h[0] + shl \$12,$d1 + and \$0x3ffffff,%rdx # h[1] + shr \$14,$h1 + or $d1,$h0 + shl \$24,$h2 + and \$0x3ffffff,$h0 # h[2] + shr \$40,$d2 + and \$0x3ffffff,$h1 # h[3] + or $d2,$h2 # h[4] + + vmovd %rax#d,%x#$H0 + vmovd %rdx#d,%x#$H1 + vmovd $h0#d,%x#$H2 + vmovd $h1#d,%x#$H3 + vmovd $h2#d,%x#$H4 + movl \$1,20($ctx) # set is_base2_26 + + call __poly1305_init_avx + +.Lproceed_avx2$suffix: + mov %r15,$len # restore $len +___ +$code.=<<___ if (!$kernel); + mov OPENSSL_ia32cap_P+8(%rip),%r9d + mov \$`(1<<31|1<<30|1<<16)`,%r11d +___ +$code.=<<___; + pop %r15 +.cfi_restore %r15 + pop %r14 +.cfi_restore %r14 + pop %r13 +.cfi_restore %r13 + pop %r12 +.cfi_restore %r12 + pop %rbx +.cfi_restore %rbx + pop %rbp +.cfi_restore %rbp +.Lbase2_64_avx2_epilogue$suffix: + jmp .Ldo_avx2$suffix +.cfi_endproc + +.align 32 +.Leven_avx2$suffix: +.cfi_startproc +___ +$code.=<<___ if (!$kernel); + mov OPENSSL_ia32cap_P+8(%rip),%r9d +___ +$code.=<<___; + vmovd 4*0($ctx),%x#$H0 # load hash value base 2^26 + vmovd 4*1($ctx),%x#$H1 + vmovd 4*2($ctx),%x#$H2 + vmovd 4*3($ctx),%x#$H3 + vmovd 4*4($ctx),%x#$H4 + +.Ldo_avx2$suffix: +___ +$code.=<<___ if (!$kernel && $avx>2); + cmp \$512,$len + jb .Lskip_avx512 + and %r11d,%r9d + test \$`1<<16`,%r9d # check for AVX512F + jnz .Lblocks_avx512 +.Lskip_avx512$suffix: +___ +$code.=<<___ if ($avx > 2 && $avx512 && $kernel); + cmp \$512,$len + jae .Lblocks_avx512 +___ +$code.=<<___ if (!$win64); + lea 8(%rsp),%r10 +.cfi_def_cfa_register %r10 + sub \$0x128,%rsp +___ +$code.=<<___ if ($win64); + lea 8(%rsp),%r10 + sub \$0x1c8,%rsp + vmovdqa %xmm6,-0xb0(%r10) + vmovdqa %xmm7,-0xa0(%r10) + vmovdqa %xmm8,-0x90(%r10) + vmovdqa %xmm9,-0x80(%r10) + vmovdqa %xmm10,-0x70(%r10) + vmovdqa %xmm11,-0x60(%r10) + vmovdqa %xmm12,-0x50(%r10) + vmovdqa %xmm13,-0x40(%r10) + vmovdqa %xmm14,-0x30(%r10) + vmovdqa %xmm15,-0x20(%r10) +.Ldo_avx2_body$suffix: +___ +$code.=<<___; + lea .Lconst(%rip),%rcx + lea 48+64($ctx),$ctx # size optimization + vmovdqa 96(%rcx),$T0 # .Lpermd_avx2 + + # expand and copy pre-calculated table to stack + vmovdqu `16*0-64`($ctx),%x#$T2 + and \$-512,%rsp + vmovdqu `16*1-64`($ctx),%x#$T3 + vmovdqu `16*2-64`($ctx),%x#$T4 + vmovdqu `16*3-64`($ctx),%x#$D0 + vmovdqu `16*4-64`($ctx),%x#$D1 + vmovdqu `16*5-64`($ctx),%x#$D2 + lea 0x90(%rsp),%rax # size optimization + vmovdqu `16*6-64`($ctx),%x#$D3 + vpermd $T2,$T0,$T2 # 00003412 -> 14243444 + vmovdqu `16*7-64`($ctx),%x#$D4 + vpermd $T3,$T0,$T3 + vmovdqu `16*8-64`($ctx),%x#$MASK + vpermd $T4,$T0,$T4 + vmovdqa $T2,0x00(%rsp) + vpermd $D0,$T0,$D0 + vmovdqa $T3,0x20-0x90(%rax) + vpermd $D1,$T0,$D1 + vmovdqa $T4,0x40-0x90(%rax) + vpermd $D2,$T0,$D2 + vmovdqa $D0,0x60-0x90(%rax) + vpermd $D3,$T0,$D3 + vmovdqa $D1,0x80-0x90(%rax) + vpermd $D4,$T0,$D4 + vmovdqa $D2,0xa0-0x90(%rax) + vpermd $MASK,$T0,$MASK + vmovdqa $D3,0xc0-0x90(%rax) + vmovdqa $D4,0xe0-0x90(%rax) + vmovdqa $MASK,0x100-0x90(%rax) + vmovdqa 64(%rcx),$MASK # .Lmask26 + + ################################################################ + # load input + vmovdqu 16*0($inp),%x#$T0 + vmovdqu 16*1($inp),%x#$T1 + vinserti128 \$1,16*2($inp),$T0,$T0 + vinserti128 \$1,16*3($inp),$T1,$T1 + lea 16*4($inp),$inp + + vpsrldq \$6,$T0,$T2 # splat input + vpsrldq \$6,$T1,$T3 + vpunpckhqdq $T1,$T0,$T4 # 4 + vpunpcklqdq $T3,$T2,$T2 # 2:3 + vpunpcklqdq $T1,$T0,$T0 # 0:1 + + vpsrlq \$30,$T2,$T3 + vpsrlq \$4,$T2,$T2 + vpsrlq \$26,$T0,$T1 + vpsrlq \$40,$T4,$T4 # 4 + vpand $MASK,$T2,$T2 # 2 + vpand $MASK,$T0,$T0 # 0 + vpand $MASK,$T1,$T1 # 1 + vpand $MASK,$T3,$T3 # 3 + vpor 32(%rcx),$T4,$T4 # padbit, yes, always + + vpaddq $H2,$T2,$H2 # accumulate input + sub \$64,$len + jz .Ltail_avx2$suffix + jmp .Loop_avx2$suffix + +.align 32 +.Loop_avx2$suffix: + ################################################################ + # ((inp[0]*r^4+inp[4])*r^4+inp[ 8])*r^4 + # ((inp[1]*r^4+inp[5])*r^4+inp[ 9])*r^3 + # ((inp[2]*r^4+inp[6])*r^4+inp[10])*r^2 + # ((inp[3]*r^4+inp[7])*r^4+inp[11])*r^1 + # \________/\__________/ + ################################################################ + #vpaddq $H2,$T2,$H2 # accumulate input + vpaddq $H0,$T0,$H0 + vmovdqa `32*0`(%rsp),$T0 # r0^4 + vpaddq $H1,$T1,$H1 + vmovdqa `32*1`(%rsp),$T1 # r1^4 + vpaddq $H3,$T3,$H3 + vmovdqa `32*3`(%rsp),$T2 # r2^4 + vpaddq $H4,$T4,$H4 + vmovdqa `32*6-0x90`(%rax),$T3 # s3^4 + vmovdqa `32*8-0x90`(%rax),$S4 # s4^4 + + # d4 = h4*r0 + h3*r1 + h2*r2 + h1*r3 + h0*r4 + # d3 = h3*r0 + h2*r1 + h1*r2 + h0*r3 + h4*5*r4 + # d2 = h2*r0 + h1*r1 + h0*r2 + h4*5*r3 + h3*5*r4 + # d1 = h1*r0 + h0*r1 + h4*5*r2 + h3*5*r3 + h2*5*r4 + # d0 = h0*r0 + h4*5*r1 + h3*5*r2 + h2*5*r3 + h1*5*r4 + # + # however, as h2 is "chronologically" first one available pull + # corresponding operations up, so it's + # + # d4 = h2*r2 + h4*r0 + h3*r1 + h1*r3 + h0*r4 + # d3 = h2*r1 + h3*r0 + h1*r2 + h0*r3 + h4*5*r4 + # d2 = h2*r0 + h1*r1 + h0*r2 + h4*5*r3 + h3*5*r4 + # d1 = h2*5*r4 + h1*r0 + h0*r1 + h4*5*r2 + h3*5*r3 + # d0 = h2*5*r3 + h0*r0 + h4*5*r1 + h3*5*r2 + h1*5*r4 + + vpmuludq $H2,$T0,$D2 # d2 = h2*r0 + vpmuludq $H2,$T1,$D3 # d3 = h2*r1 + vpmuludq $H2,$T2,$D4 # d4 = h2*r2 + vpmuludq $H2,$T3,$D0 # d0 = h2*s3 + vpmuludq $H2,$S4,$D1 # d1 = h2*s4 + + vpmuludq $H0,$T1,$T4 # h0*r1 + vpmuludq $H1,$T1,$H2 # h1*r1, borrow $H2 as temp + vpaddq $T4,$D1,$D1 # d1 += h0*r1 + vpaddq $H2,$D2,$D2 # d2 += h1*r1 + vpmuludq $H3,$T1,$T4 # h3*r1 + vpmuludq `32*2`(%rsp),$H4,$H2 # h4*s1 + vpaddq $T4,$D4,$D4 # d4 += h3*r1 + vpaddq $H2,$D0,$D0 # d0 += h4*s1 + vmovdqa `32*4-0x90`(%rax),$T1 # s2 + + vpmuludq $H0,$T0,$T4 # h0*r0 + vpmuludq $H1,$T0,$H2 # h1*r0 + vpaddq $T4,$D0,$D0 # d0 += h0*r0 + vpaddq $H2,$D1,$D1 # d1 += h1*r0 + vpmuludq $H3,$T0,$T4 # h3*r0 + vpmuludq $H4,$T0,$H2 # h4*r0 + vmovdqu 16*0($inp),%x#$T0 # load input + vpaddq $T4,$D3,$D3 # d3 += h3*r0 + vpaddq $H2,$D4,$D4 # d4 += h4*r0 + vinserti128 \$1,16*2($inp),$T0,$T0 + + vpmuludq $H3,$T1,$T4 # h3*s2 + vpmuludq $H4,$T1,$H2 # h4*s2 + vmovdqu 16*1($inp),%x#$T1 + vpaddq $T4,$D0,$D0 # d0 += h3*s2 + vpaddq $H2,$D1,$D1 # d1 += h4*s2 + vmovdqa `32*5-0x90`(%rax),$H2 # r3 + vpmuludq $H1,$T2,$T4 # h1*r2 + vpmuludq $H0,$T2,$T2 # h0*r2 + vpaddq $T4,$D3,$D3 # d3 += h1*r2 + vpaddq $T2,$D2,$D2 # d2 += h0*r2 + vinserti128 \$1,16*3($inp),$T1,$T1 + lea 16*4($inp),$inp + + vpmuludq $H1,$H2,$T4 # h1*r3 + vpmuludq $H0,$H2,$H2 # h0*r3 + vpsrldq \$6,$T0,$T2 # splat input + vpaddq $T4,$D4,$D4 # d4 += h1*r3 + vpaddq $H2,$D3,$D3 # d3 += h0*r3 + vpmuludq $H3,$T3,$T4 # h3*s3 + vpmuludq $H4,$T3,$H2 # h4*s3 + vpsrldq \$6,$T1,$T3 + vpaddq $T4,$D1,$D1 # d1 += h3*s3 + vpaddq $H2,$D2,$D2 # d2 += h4*s3 + vpunpckhqdq $T1,$T0,$T4 # 4 + + vpmuludq $H3,$S4,$H3 # h3*s4 + vpmuludq $H4,$S4,$H4 # h4*s4 + vpunpcklqdq $T1,$T0,$T0 # 0:1 + vpaddq $H3,$D2,$H2 # h2 = d2 + h3*r4 + vpaddq $H4,$D3,$H3 # h3 = d3 + h4*r4 + vpunpcklqdq $T3,$T2,$T3 # 2:3 + vpmuludq `32*7-0x90`(%rax),$H0,$H4 # h0*r4 + vpmuludq $H1,$S4,$H0 # h1*s4 + vmovdqa 64(%rcx),$MASK # .Lmask26 + vpaddq $H4,$D4,$H4 # h4 = d4 + h0*r4 + vpaddq $H0,$D0,$H0 # h0 = d0 + h1*s4 + + ################################################################ + # lazy reduction (interleaved with tail of input splat) + + vpsrlq \$26,$H3,$D3 + vpand $MASK,$H3,$H3 + vpaddq $D3,$H4,$H4 # h3 -> h4 + + vpsrlq \$26,$H0,$D0 + vpand $MASK,$H0,$H0 + vpaddq $D0,$D1,$H1 # h0 -> h1 + + vpsrlq \$26,$H4,$D4 + vpand $MASK,$H4,$H4 + + vpsrlq \$4,$T3,$T2 + + vpsrlq \$26,$H1,$D1 + vpand $MASK,$H1,$H1 + vpaddq $D1,$H2,$H2 # h1 -> h2 + + vpaddq $D4,$H0,$H0 + vpsllq \$2,$D4,$D4 + vpaddq $D4,$H0,$H0 # h4 -> h0 + + vpand $MASK,$T2,$T2 # 2 + vpsrlq \$26,$T0,$T1 + + vpsrlq \$26,$H2,$D2 + vpand $MASK,$H2,$H2 + vpaddq $D2,$H3,$H3 # h2 -> h3 + + vpaddq $T2,$H2,$H2 # modulo-scheduled + vpsrlq \$30,$T3,$T3 + + vpsrlq \$26,$H0,$D0 + vpand $MASK,$H0,$H0 + vpaddq $D0,$H1,$H1 # h0 -> h1 + + vpsrlq \$40,$T4,$T4 # 4 + + vpsrlq \$26,$H3,$D3 + vpand $MASK,$H3,$H3 + vpaddq $D3,$H4,$H4 # h3 -> h4 + + vpand $MASK,$T0,$T0 # 0 + vpand $MASK,$T1,$T1 # 1 + vpand $MASK,$T3,$T3 # 3 + vpor 32(%rcx),$T4,$T4 # padbit, yes, always + + sub \$64,$len + jnz .Loop_avx2$suffix + + .byte 0x66,0x90 +.Ltail_avx2$suffix: + ################################################################ + # while above multiplications were by r^4 in all lanes, in last + # iteration we multiply least significant lane by r^4 and most + # significant one by r, so copy of above except that references + # to the precomputed table are displaced by 4... + + #vpaddq $H2,$T2,$H2 # accumulate input + vpaddq $H0,$T0,$H0 + vmovdqu `32*0+4`(%rsp),$T0 # r0^4 + vpaddq $H1,$T1,$H1 + vmovdqu `32*1+4`(%rsp),$T1 # r1^4 + vpaddq $H3,$T3,$H3 + vmovdqu `32*3+4`(%rsp),$T2 # r2^4 + vpaddq $H4,$T4,$H4 + vmovdqu `32*6+4-0x90`(%rax),$T3 # s3^4 + vmovdqu `32*8+4-0x90`(%rax),$S4 # s4^4 + + vpmuludq $H2,$T0,$D2 # d2 = h2*r0 + vpmuludq $H2,$T1,$D3 # d3 = h2*r1 + vpmuludq $H2,$T2,$D4 # d4 = h2*r2 + vpmuludq $H2,$T3,$D0 # d0 = h2*s3 + vpmuludq $H2,$S4,$D1 # d1 = h2*s4 + + vpmuludq $H0,$T1,$T4 # h0*r1 + vpmuludq $H1,$T1,$H2 # h1*r1 + vpaddq $T4,$D1,$D1 # d1 += h0*r1 + vpaddq $H2,$D2,$D2 # d2 += h1*r1 + vpmuludq $H3,$T1,$T4 # h3*r1 + vpmuludq `32*2+4`(%rsp),$H4,$H2 # h4*s1 + vpaddq $T4,$D4,$D4 # d4 += h3*r1 + vpaddq $H2,$D0,$D0 # d0 += h4*s1 + + vpmuludq $H0,$T0,$T4 # h0*r0 + vpmuludq $H1,$T0,$H2 # h1*r0 + vpaddq $T4,$D0,$D0 # d0 += h0*r0 + vmovdqu `32*4+4-0x90`(%rax),$T1 # s2 + vpaddq $H2,$D1,$D1 # d1 += h1*r0 + vpmuludq $H3,$T0,$T4 # h3*r0 + vpmuludq $H4,$T0,$H2 # h4*r0 + vpaddq $T4,$D3,$D3 # d3 += h3*r0 + vpaddq $H2,$D4,$D4 # d4 += h4*r0 + + vpmuludq $H3,$T1,$T4 # h3*s2 + vpmuludq $H4,$T1,$H2 # h4*s2 + vpaddq $T4,$D0,$D0 # d0 += h3*s2 + vpaddq $H2,$D1,$D1 # d1 += h4*s2 + vmovdqu `32*5+4-0x90`(%rax),$H2 # r3 + vpmuludq $H1,$T2,$T4 # h1*r2 + vpmuludq $H0,$T2,$T2 # h0*r2 + vpaddq $T4,$D3,$D3 # d3 += h1*r2 + vpaddq $T2,$D2,$D2 # d2 += h0*r2 + + vpmuludq $H1,$H2,$T4 # h1*r3 + vpmuludq $H0,$H2,$H2 # h0*r3 + vpaddq $T4,$D4,$D4 # d4 += h1*r3 + vpaddq $H2,$D3,$D3 # d3 += h0*r3 + vpmuludq $H3,$T3,$T4 # h3*s3 + vpmuludq $H4,$T3,$H2 # h4*s3 + vpaddq $T4,$D1,$D1 # d1 += h3*s3 + vpaddq $H2,$D2,$D2 # d2 += h4*s3 + + vpmuludq $H3,$S4,$H3 # h3*s4 + vpmuludq $H4,$S4,$H4 # h4*s4 + vpaddq $H3,$D2,$H2 # h2 = d2 + h3*r4 + vpaddq $H4,$D3,$H3 # h3 = d3 + h4*r4 + vpmuludq `32*7+4-0x90`(%rax),$H0,$H4 # h0*r4 + vpmuludq $H1,$S4,$H0 # h1*s4 + vmovdqa 64(%rcx),$MASK # .Lmask26 + vpaddq $H4,$D4,$H4 # h4 = d4 + h0*r4 + vpaddq $H0,$D0,$H0 # h0 = d0 + h1*s4 + + ################################################################ + # horizontal addition + + vpsrldq \$8,$D1,$T1 + vpsrldq \$8,$H2,$T2 + vpsrldq \$8,$H3,$T3 + vpsrldq \$8,$H4,$T4 + vpsrldq \$8,$H0,$T0 + vpaddq $T1,$D1,$D1 + vpaddq $T2,$H2,$H2 + vpaddq $T3,$H3,$H3 + vpaddq $T4,$H4,$H4 + vpaddq $T0,$H0,$H0 + + vpermq \$0x2,$H3,$T3 + vpermq \$0x2,$H4,$T4 + vpermq \$0x2,$H0,$T0 + vpermq \$0x2,$D1,$T1 + vpermq \$0x2,$H2,$T2 + vpaddq $T3,$H3,$H3 + vpaddq $T4,$H4,$H4 + vpaddq $T0,$H0,$H0 + vpaddq $T1,$D1,$D1 + vpaddq $T2,$H2,$H2 + + ################################################################ + # lazy reduction + + vpsrlq \$26,$H3,$D3 + vpand $MASK,$H3,$H3 + vpaddq $D3,$H4,$H4 # h3 -> h4 + + vpsrlq \$26,$H0,$D0 + vpand $MASK,$H0,$H0 + vpaddq $D0,$D1,$H1 # h0 -> h1 + + vpsrlq \$26,$H4,$D4 + vpand $MASK,$H4,$H4 + + vpsrlq \$26,$H1,$D1 + vpand $MASK,$H1,$H1 + vpaddq $D1,$H2,$H2 # h1 -> h2 + + vpaddq $D4,$H0,$H0 + vpsllq \$2,$D4,$D4 + vpaddq $D4,$H0,$H0 # h4 -> h0 + + vpsrlq \$26,$H2,$D2 + vpand $MASK,$H2,$H2 + vpaddq $D2,$H3,$H3 # h2 -> h3 + + vpsrlq \$26,$H0,$D0 + vpand $MASK,$H0,$H0 + vpaddq $D0,$H1,$H1 # h0 -> h1 + + vpsrlq \$26,$H3,$D3 + vpand $MASK,$H3,$H3 + vpaddq $D3,$H4,$H4 # h3 -> h4 + + vmovd %x#$H0,`4*0-48-64`($ctx)# save partially reduced + vmovd %x#$H1,`4*1-48-64`($ctx) + vmovd %x#$H2,`4*2-48-64`($ctx) + vmovd %x#$H3,`4*3-48-64`($ctx) + vmovd %x#$H4,`4*4-48-64`($ctx) +___ +$code.=<<___ if ($win64); + vmovdqa -0xb0(%r10),%xmm6 + vmovdqa -0xa0(%r10),%xmm7 + vmovdqa -0x90(%r10),%xmm8 + vmovdqa -0x80(%r10),%xmm9 + vmovdqa -0x70(%r10),%xmm10 + vmovdqa -0x60(%r10),%xmm11 + vmovdqa -0x50(%r10),%xmm12 + vmovdqa -0x40(%r10),%xmm13 + vmovdqa -0x30(%r10),%xmm14 + vmovdqa -0x20(%r10),%xmm15 + lea -8(%r10),%rsp +.Ldo_avx2_epilogue$suffix: +___ +$code.=<<___ if (!$win64); + lea -8(%r10),%rsp +.cfi_def_cfa_register %rsp +___ +$code.=<<___; + vzeroupper + ret +.cfi_endproc +___ +if($avx > 2 && $avx512) { +my ($R0,$R1,$R2,$R3,$R4, $S1,$S2,$S3,$S4) = map("%zmm$_",(16..24)); +my ($M0,$M1,$M2,$M3,$M4) = map("%zmm$_",(25..29)); +my $PADBIT="%zmm30"; + +map(s/%y/%z/,($T4,$T0,$T1,$T2,$T3)); # switch to %zmm domain +map(s/%y/%z/,($D0,$D1,$D2,$D3,$D4)); +map(s/%y/%z/,($H0,$H1,$H2,$H3,$H4)); +map(s/%y/%z/,($MASK)); + +$code.=<<___; +.cfi_startproc +.Lblocks_avx512: + mov \$15,%eax + kmovw %eax,%k2 +___ +$code.=<<___ if (!$win64); + lea 8(%rsp),%r10 +.cfi_def_cfa_register %r10 + sub \$0x128,%rsp +___ +$code.=<<___ if ($win64); + lea 8(%rsp),%r10 + sub \$0x1c8,%rsp + vmovdqa %xmm6,-0xb0(%r10) + vmovdqa %xmm7,-0xa0(%r10) + vmovdqa %xmm8,-0x90(%r10) + vmovdqa %xmm9,-0x80(%r10) + vmovdqa %xmm10,-0x70(%r10) + vmovdqa %xmm11,-0x60(%r10) + vmovdqa %xmm12,-0x50(%r10) + vmovdqa %xmm13,-0x40(%r10) + vmovdqa %xmm14,-0x30(%r10) + vmovdqa %xmm15,-0x20(%r10) +.Ldo_avx512_body: +___ +$code.=<<___; + lea .Lconst(%rip),%rcx + lea 48+64($ctx),$ctx # size optimization + vmovdqa 96(%rcx),%y#$T2 # .Lpermd_avx2 + + # expand pre-calculated table + vmovdqu `16*0-64`($ctx),%x#$D0 # will become expanded ${R0} + and \$-512,%rsp + vmovdqu `16*1-64`($ctx),%x#$D1 # will become ... ${R1} + mov \$0x20,%rax + vmovdqu `16*2-64`($ctx),%x#$T0 # ... ${S1} + vmovdqu `16*3-64`($ctx),%x#$D2 # ... ${R2} + vmovdqu `16*4-64`($ctx),%x#$T1 # ... ${S2} + vmovdqu `16*5-64`($ctx),%x#$D3 # ... ${R3} + vmovdqu `16*6-64`($ctx),%x#$T3 # ... ${S3} + vmovdqu `16*7-64`($ctx),%x#$D4 # ... ${R4} + vmovdqu `16*8-64`($ctx),%x#$T4 # ... ${S4} + vpermd $D0,$T2,$R0 # 00003412 -> 14243444 + vpbroadcastq 64(%rcx),$MASK # .Lmask26 + vpermd $D1,$T2,$R1 + vpermd $T0,$T2,$S1 + vpermd $D2,$T2,$R2 + vmovdqa64 $R0,0x00(%rsp){%k2} # save in case $len%128 != 0 + vpsrlq \$32,$R0,$T0 # 14243444 -> 01020304 + vpermd $T1,$T2,$S2 + vmovdqu64 $R1,0x00(%rsp,%rax){%k2} + vpsrlq \$32,$R1,$T1 + vpermd $D3,$T2,$R3 + vmovdqa64 $S1,0x40(%rsp){%k2} + vpermd $T3,$T2,$S3 + vpermd $D4,$T2,$R4 + vmovdqu64 $R2,0x40(%rsp,%rax){%k2} + vpermd $T4,$T2,$S4 + vmovdqa64 $S2,0x80(%rsp){%k2} + vmovdqu64 $R3,0x80(%rsp,%rax){%k2} + vmovdqa64 $S3,0xc0(%rsp){%k2} + vmovdqu64 $R4,0xc0(%rsp,%rax){%k2} + vmovdqa64 $S4,0x100(%rsp){%k2} + + ################################################################ + # calculate 5th through 8th powers of the key + # + # d0 = r0'*r0 + r1'*5*r4 + r2'*5*r3 + r3'*5*r2 + r4'*5*r1 + # d1 = r0'*r1 + r1'*r0 + r2'*5*r4 + r3'*5*r3 + r4'*5*r2 + # d2 = r0'*r2 + r1'*r1 + r2'*r0 + r3'*5*r4 + r4'*5*r3 + # d3 = r0'*r3 + r1'*r2 + r2'*r1 + r3'*r0 + r4'*5*r4 + # d4 = r0'*r4 + r1'*r3 + r2'*r2 + r3'*r1 + r4'*r0 + + vpmuludq $T0,$R0,$D0 # d0 = r0'*r0 + vpmuludq $T0,$R1,$D1 # d1 = r0'*r1 + vpmuludq $T0,$R2,$D2 # d2 = r0'*r2 + vpmuludq $T0,$R3,$D3 # d3 = r0'*r3 + vpmuludq $T0,$R4,$D4 # d4 = r0'*r4 + vpsrlq \$32,$R2,$T2 + + vpmuludq $T1,$S4,$M0 + vpmuludq $T1,$R0,$M1 + vpmuludq $T1,$R1,$M2 + vpmuludq $T1,$R2,$M3 + vpmuludq $T1,$R3,$M4 + vpsrlq \$32,$R3,$T3 + vpaddq $M0,$D0,$D0 # d0 += r1'*5*r4 + vpaddq $M1,$D1,$D1 # d1 += r1'*r0 + vpaddq $M2,$D2,$D2 # d2 += r1'*r1 + vpaddq $M3,$D3,$D3 # d3 += r1'*r2 + vpaddq $M4,$D4,$D4 # d4 += r1'*r3 + + vpmuludq $T2,$S3,$M0 + vpmuludq $T2,$S4,$M1 + vpmuludq $T2,$R1,$M3 + vpmuludq $T2,$R2,$M4 + vpmuludq $T2,$R0,$M2 + vpsrlq \$32,$R4,$T4 + vpaddq $M0,$D0,$D0 # d0 += r2'*5*r3 + vpaddq $M1,$D1,$D1 # d1 += r2'*5*r4 + vpaddq $M3,$D3,$D3 # d3 += r2'*r1 + vpaddq $M4,$D4,$D4 # d4 += r2'*r2 + vpaddq $M2,$D2,$D2 # d2 += r2'*r0 + + vpmuludq $T3,$S2,$M0 + vpmuludq $T3,$R0,$M3 + vpmuludq $T3,$R1,$M4 + vpmuludq $T3,$S3,$M1 + vpmuludq $T3,$S4,$M2 + vpaddq $M0,$D0,$D0 # d0 += r3'*5*r2 + vpaddq $M3,$D3,$D3 # d3 += r3'*r0 + vpaddq $M4,$D4,$D4 # d4 += r3'*r1 + vpaddq $M1,$D1,$D1 # d1 += r3'*5*r3 + vpaddq $M2,$D2,$D2 # d2 += r3'*5*r4 + + vpmuludq $T4,$S4,$M3 + vpmuludq $T4,$R0,$M4 + vpmuludq $T4,$S1,$M0 + vpmuludq $T4,$S2,$M1 + vpmuludq $T4,$S3,$M2 + vpaddq $M3,$D3,$D3 # d3 += r2'*5*r4 + vpaddq $M4,$D4,$D4 # d4 += r2'*r0 + vpaddq $M0,$D0,$D0 # d0 += r2'*5*r1 + vpaddq $M1,$D1,$D1 # d1 += r2'*5*r2 + vpaddq $M2,$D2,$D2 # d2 += r2'*5*r3 + + ################################################################ + # load input + vmovdqu64 16*0($inp),%z#$T3 + vmovdqu64 16*4($inp),%z#$T4 + lea 16*8($inp),$inp + + ################################################################ + # lazy reduction + + vpsrlq \$26,$D3,$M3 + vpandq $MASK,$D3,$D3 + vpaddq $M3,$D4,$D4 # d3 -> d4 + + vpsrlq \$26,$D0,$M0 + vpandq $MASK,$D0,$D0 + vpaddq $M0,$D1,$D1 # d0 -> d1 + + vpsrlq \$26,$D4,$M4 + vpandq $MASK,$D4,$D4 + + vpsrlq \$26,$D1,$M1 + vpandq $MASK,$D1,$D1 + vpaddq $M1,$D2,$D2 # d1 -> d2 + + vpaddq $M4,$D0,$D0 + vpsllq \$2,$M4,$M4 + vpaddq $M4,$D0,$D0 # d4 -> d0 + + vpsrlq \$26,$D2,$M2 + vpandq $MASK,$D2,$D2 + vpaddq $M2,$D3,$D3 # d2 -> d3 + + vpsrlq \$26,$D0,$M0 + vpandq $MASK,$D0,$D0 + vpaddq $M0,$D1,$D1 # d0 -> d1 + + vpsrlq \$26,$D3,$M3 + vpandq $MASK,$D3,$D3 + vpaddq $M3,$D4,$D4 # d3 -> d4 + + ################################################################ + # at this point we have 14243444 in $R0-$S4 and 05060708 in + # $D0-$D4, ... + + vpunpcklqdq $T4,$T3,$T0 # transpose input + vpunpckhqdq $T4,$T3,$T4 + + # ... since input 64-bit lanes are ordered as 73625140, we could + # "vperm" it to 76543210 (here and in each loop iteration), *or* + # we could just flow along, hence the goal for $R0-$S4 is + # 1858286838784888 ... + + vmovdqa32 128(%rcx),$M0 # .Lpermd_avx512: + mov \$0x7777,%eax + kmovw %eax,%k1 + + vpermd $R0,$M0,$R0 # 14243444 -> 1---2---3---4--- + vpermd $R1,$M0,$R1 + vpermd $R2,$M0,$R2 + vpermd $R3,$M0,$R3 + vpermd $R4,$M0,$R4 + + vpermd $D0,$M0,${R0}{%k1} # 05060708 -> 1858286838784888 + vpermd $D1,$M0,${R1}{%k1} + vpermd $D2,$M0,${R2}{%k1} + vpermd $D3,$M0,${R3}{%k1} + vpermd $D4,$M0,${R4}{%k1} + + vpslld \$2,$R1,$S1 # *5 + vpslld \$2,$R2,$S2 + vpslld \$2,$R3,$S3 + vpslld \$2,$R4,$S4 + vpaddd $R1,$S1,$S1 + vpaddd $R2,$S2,$S2 + vpaddd $R3,$S3,$S3 + vpaddd $R4,$S4,$S4 + + vpbroadcastq 32(%rcx),$PADBIT # .L129 + + vpsrlq \$52,$T0,$T2 # splat input + vpsllq \$12,$T4,$T3 + vporq $T3,$T2,$T2 + vpsrlq \$26,$T0,$T1 + vpsrlq \$14,$T4,$T3 + vpsrlq \$40,$T4,$T4 # 4 + vpandq $MASK,$T2,$T2 # 2 + vpandq $MASK,$T0,$T0 # 0 + #vpandq $MASK,$T1,$T1 # 1 + #vpandq $MASK,$T3,$T3 # 3 + #vporq $PADBIT,$T4,$T4 # padbit, yes, always + + vpaddq $H2,$T2,$H2 # accumulate input + sub \$192,$len + jbe .Ltail_avx512 + jmp .Loop_avx512 + +.align 32 +.Loop_avx512: + ################################################################ + # ((inp[0]*r^8+inp[ 8])*r^8+inp[16])*r^8 + # ((inp[1]*r^8+inp[ 9])*r^8+inp[17])*r^7 + # ((inp[2]*r^8+inp[10])*r^8+inp[18])*r^6 + # ((inp[3]*r^8+inp[11])*r^8+inp[19])*r^5 + # ((inp[4]*r^8+inp[12])*r^8+inp[20])*r^4 + # ((inp[5]*r^8+inp[13])*r^8+inp[21])*r^3 + # ((inp[6]*r^8+inp[14])*r^8+inp[22])*r^2 + # ((inp[7]*r^8+inp[15])*r^8+inp[23])*r^1 + # \________/\___________/ + ################################################################ + #vpaddq $H2,$T2,$H2 # accumulate input + + # d4 = h4*r0 + h3*r1 + h2*r2 + h1*r3 + h0*r4 + # d3 = h3*r0 + h2*r1 + h1*r2 + h0*r3 + h4*5*r4 + # d2 = h2*r0 + h1*r1 + h0*r2 + h4*5*r3 + h3*5*r4 + # d1 = h1*r0 + h0*r1 + h4*5*r2 + h3*5*r3 + h2*5*r4 + # d0 = h0*r0 + h4*5*r1 + h3*5*r2 + h2*5*r3 + h1*5*r4 + # + # however, as h2 is "chronologically" first one available pull + # corresponding operations up, so it's + # + # d3 = h2*r1 + h0*r3 + h1*r2 + h3*r0 + h4*5*r4 + # d4 = h2*r2 + h0*r4 + h1*r3 + h3*r1 + h4*r0 + # d0 = h2*5*r3 + h0*r0 + h1*5*r4 + h3*5*r2 + h4*5*r1 + # d1 = h2*5*r4 + h0*r1 + h1*r0 + h3*5*r3 + h4*5*r2 + # d2 = h2*r0 + h0*r2 + h1*r1 + h3*5*r4 + h4*5*r3 + + vpmuludq $H2,$R1,$D3 # d3 = h2*r1 + vpaddq $H0,$T0,$H0 + vpmuludq $H2,$R2,$D4 # d4 = h2*r2 + vpandq $MASK,$T1,$T1 # 1 + vpmuludq $H2,$S3,$D0 # d0 = h2*s3 + vpandq $MASK,$T3,$T3 # 3 + vpmuludq $H2,$S4,$D1 # d1 = h2*s4 + vporq $PADBIT,$T4,$T4 # padbit, yes, always + vpmuludq $H2,$R0,$D2 # d2 = h2*r0 + vpaddq $H1,$T1,$H1 # accumulate input + vpaddq $H3,$T3,$H3 + vpaddq $H4,$T4,$H4 + + vmovdqu64 16*0($inp),$T3 # load input + vmovdqu64 16*4($inp),$T4 + lea 16*8($inp),$inp + vpmuludq $H0,$R3,$M3 + vpmuludq $H0,$R4,$M4 + vpmuludq $H0,$R0,$M0 + vpmuludq $H0,$R1,$M1 + vpaddq $M3,$D3,$D3 # d3 += h0*r3 + vpaddq $M4,$D4,$D4 # d4 += h0*r4 + vpaddq $M0,$D0,$D0 # d0 += h0*r0 + vpaddq $M1,$D1,$D1 # d1 += h0*r1 + + vpmuludq $H1,$R2,$M3 + vpmuludq $H1,$R3,$M4 + vpmuludq $H1,$S4,$M0 + vpmuludq $H0,$R2,$M2 + vpaddq $M3,$D3,$D3 # d3 += h1*r2 + vpaddq $M4,$D4,$D4 # d4 += h1*r3 + vpaddq $M0,$D0,$D0 # d0 += h1*s4 + vpaddq $M2,$D2,$D2 # d2 += h0*r2 + + vpunpcklqdq $T4,$T3,$T0 # transpose input + vpunpckhqdq $T4,$T3,$T4 + + vpmuludq $H3,$R0,$M3 + vpmuludq $H3,$R1,$M4 + vpmuludq $H1,$R0,$M1 + vpmuludq $H1,$R1,$M2 + vpaddq $M3,$D3,$D3 # d3 += h3*r0 + vpaddq $M4,$D4,$D4 # d4 += h3*r1 + vpaddq $M1,$D1,$D1 # d1 += h1*r0 + vpaddq $M2,$D2,$D2 # d2 += h1*r1 + + vpmuludq $H4,$S4,$M3 + vpmuludq $H4,$R0,$M4 + vpmuludq $H3,$S2,$M0 + vpmuludq $H3,$S3,$M1 + vpaddq $M3,$D3,$D3 # d3 += h4*s4 + vpmuludq $H3,$S4,$M2 + vpaddq $M4,$D4,$D4 # d4 += h4*r0 + vpaddq $M0,$D0,$D0 # d0 += h3*s2 + vpaddq $M1,$D1,$D1 # d1 += h3*s3 + vpaddq $M2,$D2,$D2 # d2 += h3*s4 + + vpmuludq $H4,$S1,$M0 + vpmuludq $H4,$S2,$M1 + vpmuludq $H4,$S3,$M2 + vpaddq $M0,$D0,$H0 # h0 = d0 + h4*s1 + vpaddq $M1,$D1,$H1 # h1 = d2 + h4*s2 + vpaddq $M2,$D2,$H2 # h2 = d3 + h4*s3 + + ################################################################ + # lazy reduction (interleaved with input splat) + + vpsrlq \$52,$T0,$T2 # splat input + vpsllq \$12,$T4,$T3 + + vpsrlq \$26,$D3,$H3 + vpandq $MASK,$D3,$D3 + vpaddq $H3,$D4,$H4 # h3 -> h4 + + vporq $T3,$T2,$T2 + + vpsrlq \$26,$H0,$D0 + vpandq $MASK,$H0,$H0 + vpaddq $D0,$H1,$H1 # h0 -> h1 + + vpandq $MASK,$T2,$T2 # 2 + + vpsrlq \$26,$H4,$D4 + vpandq $MASK,$H4,$H4 + + vpsrlq \$26,$H1,$D1 + vpandq $MASK,$H1,$H1 + vpaddq $D1,$H2,$H2 # h1 -> h2 + + vpaddq $D4,$H0,$H0 + vpsllq \$2,$D4,$D4 + vpaddq $D4,$H0,$H0 # h4 -> h0 + + vpaddq $T2,$H2,$H2 # modulo-scheduled + vpsrlq \$26,$T0,$T1 + + vpsrlq \$26,$H2,$D2 + vpandq $MASK,$H2,$H2 + vpaddq $D2,$D3,$H3 # h2 -> h3 + + vpsrlq \$14,$T4,$T3 + + vpsrlq \$26,$H0,$D0 + vpandq $MASK,$H0,$H0 + vpaddq $D0,$H1,$H1 # h0 -> h1 + + vpsrlq \$40,$T4,$T4 # 4 + + vpsrlq \$26,$H3,$D3 + vpandq $MASK,$H3,$H3 + vpaddq $D3,$H4,$H4 # h3 -> h4 + + vpandq $MASK,$T0,$T0 # 0 + #vpandq $MASK,$T1,$T1 # 1 + #vpandq $MASK,$T3,$T3 # 3 + #vporq $PADBIT,$T4,$T4 # padbit, yes, always + + sub \$128,$len + ja .Loop_avx512 + +.Ltail_avx512: + ################################################################ + # while above multiplications were by r^8 in all lanes, in last + # iteration we multiply least significant lane by r^8 and most + # significant one by r, that's why table gets shifted... + + vpsrlq \$32,$R0,$R0 # 0105020603070408 + vpsrlq \$32,$R1,$R1 + vpsrlq \$32,$R2,$R2 + vpsrlq \$32,$S3,$S3 + vpsrlq \$32,$S4,$S4 + vpsrlq \$32,$R3,$R3 + vpsrlq \$32,$R4,$R4 + vpsrlq \$32,$S1,$S1 + vpsrlq \$32,$S2,$S2 + + ################################################################ + # load either next or last 64 byte of input + lea ($inp,$len),$inp + + #vpaddq $H2,$T2,$H2 # accumulate input + vpaddq $H0,$T0,$H0 + + vpmuludq $H2,$R1,$D3 # d3 = h2*r1 + vpmuludq $H2,$R2,$D4 # d4 = h2*r2 + vpmuludq $H2,$S3,$D0 # d0 = h2*s3 + vpandq $MASK,$T1,$T1 # 1 + vpmuludq $H2,$S4,$D1 # d1 = h2*s4 + vpandq $MASK,$T3,$T3 # 3 + vpmuludq $H2,$R0,$D2 # d2 = h2*r0 + vporq $PADBIT,$T4,$T4 # padbit, yes, always + vpaddq $H1,$T1,$H1 # accumulate input + vpaddq $H3,$T3,$H3 + vpaddq $H4,$T4,$H4 + + vmovdqu 16*0($inp),%x#$T0 + vpmuludq $H0,$R3,$M3 + vpmuludq $H0,$R4,$M4 + vpmuludq $H0,$R0,$M0 + vpmuludq $H0,$R1,$M1 + vpaddq $M3,$D3,$D3 # d3 += h0*r3 + vpaddq $M4,$D4,$D4 # d4 += h0*r4 + vpaddq $M0,$D0,$D0 # d0 += h0*r0 + vpaddq $M1,$D1,$D1 # d1 += h0*r1 + + vmovdqu 16*1($inp),%x#$T1 + vpmuludq $H1,$R2,$M3 + vpmuludq $H1,$R3,$M4 + vpmuludq $H1,$S4,$M0 + vpmuludq $H0,$R2,$M2 + vpaddq $M3,$D3,$D3 # d3 += h1*r2 + vpaddq $M4,$D4,$D4 # d4 += h1*r3 + vpaddq $M0,$D0,$D0 # d0 += h1*s4 + vpaddq $M2,$D2,$D2 # d2 += h0*r2 + + vinserti128 \$1,16*2($inp),%y#$T0,%y#$T0 + vpmuludq $H3,$R0,$M3 + vpmuludq $H3,$R1,$M4 + vpmuludq $H1,$R0,$M1 + vpmuludq $H1,$R1,$M2 + vpaddq $M3,$D3,$D3 # d3 += h3*r0 + vpaddq $M4,$D4,$D4 # d4 += h3*r1 + vpaddq $M1,$D1,$D1 # d1 += h1*r0 + vpaddq $M2,$D2,$D2 # d2 += h1*r1 + + vinserti128 \$1,16*3($inp),%y#$T1,%y#$T1 + vpmuludq $H4,$S4,$M3 + vpmuludq $H4,$R0,$M4 + vpmuludq $H3,$S2,$M0 + vpmuludq $H3,$S3,$M1 + vpmuludq $H3,$S4,$M2 + vpaddq $M3,$D3,$H3 # h3 = d3 + h4*s4 + vpaddq $M4,$D4,$D4 # d4 += h4*r0 + vpaddq $M0,$D0,$D0 # d0 += h3*s2 + vpaddq $M1,$D1,$D1 # d1 += h3*s3 + vpaddq $M2,$D2,$D2 # d2 += h3*s4 + + vpmuludq $H4,$S1,$M0 + vpmuludq $H4,$S2,$M1 + vpmuludq $H4,$S3,$M2 + vpaddq $M0,$D0,$H0 # h0 = d0 + h4*s1 + vpaddq $M1,$D1,$H1 # h1 = d2 + h4*s2 + vpaddq $M2,$D2,$H2 # h2 = d3 + h4*s3 + + ################################################################ + # horizontal addition + + mov \$1,%eax + vpermq \$0xb1,$H3,$D3 + vpermq \$0xb1,$D4,$H4 + vpermq \$0xb1,$H0,$D0 + vpermq \$0xb1,$H1,$D1 + vpermq \$0xb1,$H2,$D2 + vpaddq $D3,$H3,$H3 + vpaddq $D4,$H4,$H4 + vpaddq $D0,$H0,$H0 + vpaddq $D1,$H1,$H1 + vpaddq $D2,$H2,$H2 + + kmovw %eax,%k3 + vpermq \$0x2,$H3,$D3 + vpermq \$0x2,$H4,$D4 + vpermq \$0x2,$H0,$D0 + vpermq \$0x2,$H1,$D1 + vpermq \$0x2,$H2,$D2 + vpaddq $D3,$H3,$H3 + vpaddq $D4,$H4,$H4 + vpaddq $D0,$H0,$H0 + vpaddq $D1,$H1,$H1 + vpaddq $D2,$H2,$H2 + + vextracti64x4 \$0x1,$H3,%y#$D3 + vextracti64x4 \$0x1,$H4,%y#$D4 + vextracti64x4 \$0x1,$H0,%y#$D0 + vextracti64x4 \$0x1,$H1,%y#$D1 + vextracti64x4 \$0x1,$H2,%y#$D2 + vpaddq $D3,$H3,${H3}{%k3}{z} # keep single qword in case + vpaddq $D4,$H4,${H4}{%k3}{z} # it's passed to .Ltail_avx2 + vpaddq $D0,$H0,${H0}{%k3}{z} + vpaddq $D1,$H1,${H1}{%k3}{z} + vpaddq $D2,$H2,${H2}{%k3}{z} +___ +map(s/%z/%y/,($T0,$T1,$T2,$T3,$T4, $PADBIT)); +map(s/%z/%y/,($H0,$H1,$H2,$H3,$H4, $D0,$D1,$D2,$D3,$D4, $MASK)); +$code.=<<___; + ################################################################ + # lazy reduction (interleaved with input splat) + + vpsrlq \$26,$H3,$D3 + vpand $MASK,$H3,$H3 + vpsrldq \$6,$T0,$T2 # splat input + vpsrldq \$6,$T1,$T3 + vpunpckhqdq $T1,$T0,$T4 # 4 + vpaddq $D3,$H4,$H4 # h3 -> h4 + + vpsrlq \$26,$H0,$D0 + vpand $MASK,$H0,$H0 + vpunpcklqdq $T3,$T2,$T2 # 2:3 + vpunpcklqdq $T1,$T0,$T0 # 0:1 + vpaddq $D0,$H1,$H1 # h0 -> h1 + + vpsrlq \$26,$H4,$D4 + vpand $MASK,$H4,$H4 + + vpsrlq \$26,$H1,$D1 + vpand $MASK,$H1,$H1 + vpsrlq \$30,$T2,$T3 + vpsrlq \$4,$T2,$T2 + vpaddq $D1,$H2,$H2 # h1 -> h2 + + vpaddq $D4,$H0,$H0 + vpsllq \$2,$D4,$D4 + vpsrlq \$26,$T0,$T1 + vpsrlq \$40,$T4,$T4 # 4 + vpaddq $D4,$H0,$H0 # h4 -> h0 + + vpsrlq \$26,$H2,$D2 + vpand $MASK,$H2,$H2 + vpand $MASK,$T2,$T2 # 2 + vpand $MASK,$T0,$T0 # 0 + vpaddq $D2,$H3,$H3 # h2 -> h3 + + vpsrlq \$26,$H0,$D0 + vpand $MASK,$H0,$H0 + vpaddq $H2,$T2,$H2 # accumulate input for .Ltail_avx2 + vpand $MASK,$T1,$T1 # 1 + vpaddq $D0,$H1,$H1 # h0 -> h1 + + vpsrlq \$26,$H3,$D3 + vpand $MASK,$H3,$H3 + vpand $MASK,$T3,$T3 # 3 + vpor 32(%rcx),$T4,$T4 # padbit, yes, always + vpaddq $D3,$H4,$H4 # h3 -> h4 + + lea 0x90(%rsp),%rax # size optimization for .Ltail_avx2 + add \$64,$len + jnz .Ltail_avx2$suffix + + vpsubq $T2,$H2,$H2 # undo input accumulation + vmovd %x#$H0,`4*0-48-64`($ctx)# save partially reduced + vmovd %x#$H1,`4*1-48-64`($ctx) + vmovd %x#$H2,`4*2-48-64`($ctx) + vmovd %x#$H3,`4*3-48-64`($ctx) + vmovd %x#$H4,`4*4-48-64`($ctx) + vzeroall +___ +$code.=<<___ if ($win64); + movdqa -0xb0(%r10),%xmm6 + movdqa -0xa0(%r10),%xmm7 + movdqa -0x90(%r10),%xmm8 + movdqa -0x80(%r10),%xmm9 + movdqa -0x70(%r10),%xmm10 + movdqa -0x60(%r10),%xmm11 + movdqa -0x50(%r10),%xmm12 + movdqa -0x40(%r10),%xmm13 + movdqa -0x30(%r10),%xmm14 + movdqa -0x20(%r10),%xmm15 + lea -8(%r10),%rsp +.Ldo_avx512_epilogue: +___ +$code.=<<___ if (!$win64); + lea -8(%r10),%rsp +.cfi_def_cfa_register %rsp +___ +$code.=<<___; + ret +.cfi_endproc +___ + +} + +} + +&declare_function("poly1305_blocks_avx2", 32, 4); +poly1305_blocks_avxN(0); +&end_function("poly1305_blocks_avx2"); + +if($kernel) { + $code .= "#endif\n"; +} + +####################################################################### +if ($avx>2) { +# On entry we have input length divisible by 64. But since inner loop +# processes 128 bytes per iteration, cases when length is not divisible +# by 128 are handled by passing tail 64 bytes to .Ltail_avx2. For this +# reason stack layout is kept identical to poly1305_blocks_avx2. If not +# for this tail, we wouldn't have to even allocate stack frame... + +if($kernel) { + $code .= "#ifdef CONFIG_AS_AVX512\n"; +} + +&declare_function("poly1305_blocks_avx512", 32, 4); +poly1305_blocks_avxN(1); +&end_function("poly1305_blocks_avx512"); + +if ($kernel) { + $code .= "#endif\n"; +} + +if (!$kernel && $avx>3) { +######################################################################## +# VPMADD52 version using 2^44 radix. +# +# One can argue that base 2^52 would be more natural. Well, even though +# some operations would be more natural, one has to recognize couple of +# things. Base 2^52 doesn't provide advantage over base 2^44 if you look +# at amount of multiply-n-accumulate operations. Secondly, it makes it +# impossible to pre-compute multiples of 5 [referred to as s[]/sN in +# reference implementations], which means that more such operations +# would have to be performed in inner loop, which in turn makes critical +# path longer. In other words, even though base 2^44 reduction might +# look less elegant, overall critical path is actually shorter... + +######################################################################## +# Layout of opaque area is following. +# +# unsigned __int64 h[3]; # current hash value base 2^44 +# unsigned __int64 s[2]; # key value*20 base 2^44 +# unsigned __int64 r[3]; # key value base 2^44 +# struct { unsigned __int64 r^1, r^3, r^2, r^4; } R[4]; +# # r^n positions reflect +# # placement in register, not +# # memory, R[3] is R[1]*20 + +$code.=<<___; +.type poly1305_init_base2_44,\@function,3 +.align 32 +poly1305_init_base2_44: + xor %rax,%rax + mov %rax,0($ctx) # initialize hash value + mov %rax,8($ctx) + mov %rax,16($ctx) + +.Linit_base2_44: + lea poly1305_blocks_vpmadd52(%rip),%r10 + lea poly1305_emit_base2_44(%rip),%r11 + + mov \$0x0ffffffc0fffffff,%rax + mov \$0x0ffffffc0ffffffc,%rcx + and 0($inp),%rax + mov \$0x00000fffffffffff,%r8 + and 8($inp),%rcx + mov \$0x00000fffffffffff,%r9 + and %rax,%r8 + shrd \$44,%rcx,%rax + mov %r8,40($ctx) # r0 + and %r9,%rax + shr \$24,%rcx + mov %rax,48($ctx) # r1 + lea (%rax,%rax,4),%rax # *5 + mov %rcx,56($ctx) # r2 + shl \$2,%rax # magic <<2 + lea (%rcx,%rcx,4),%rcx # *5 + shl \$2,%rcx # magic <<2 + mov %rax,24($ctx) # s1 + mov %rcx,32($ctx) # s2 + movq \$-1,64($ctx) # write impossible value +___ +$code.=<<___ if ($flavour !~ /elf32/); + mov %r10,0(%rdx) + mov %r11,8(%rdx) +___ +$code.=<<___ if ($flavour =~ /elf32/); + mov %r10d,0(%rdx) + mov %r11d,4(%rdx) +___ +$code.=<<___; + mov \$1,%eax + ret +.size poly1305_init_base2_44,.-poly1305_init_base2_44 +___ +{ +my ($H0,$H1,$H2,$r2r1r0,$r1r0s2,$r0s2s1,$Dlo,$Dhi) = map("%ymm$_",(0..5,16,17)); +my ($T0,$inp_permd,$inp_shift,$PAD) = map("%ymm$_",(18..21)); +my ($reduc_mask,$reduc_rght,$reduc_left) = map("%ymm$_",(22..25)); + +$code.=<<___; +.type poly1305_blocks_vpmadd52,\@function,4 +.align 32 +poly1305_blocks_vpmadd52: + shr \$4,$len + jz .Lno_data_vpmadd52 # too short + + shl \$40,$padbit + mov 64($ctx),%r8 # peek on power of the key + + # if powers of the key are not calculated yet, process up to 3 + # blocks with this single-block subroutine, otherwise ensure that + # length is divisible by 2 blocks and pass the rest down to next + # subroutine... + + mov \$3,%rax + mov \$1,%r10 + cmp \$4,$len # is input long + cmovae %r10,%rax + test %r8,%r8 # is power value impossible? + cmovns %r10,%rax + + and $len,%rax # is input of favourable length? + jz .Lblocks_vpmadd52_4x + + sub %rax,$len + mov \$7,%r10d + mov \$1,%r11d + kmovw %r10d,%k7 + lea .L2_44_inp_permd(%rip),%r10 + kmovw %r11d,%k1 + + vmovq $padbit,%x#$PAD + vmovdqa64 0(%r10),$inp_permd # .L2_44_inp_permd + vmovdqa64 32(%r10),$inp_shift # .L2_44_inp_shift + vpermq \$0xcf,$PAD,$PAD + vmovdqa64 64(%r10),$reduc_mask # .L2_44_mask + + vmovdqu64 0($ctx),${Dlo}{%k7}{z} # load hash value + vmovdqu64 40($ctx),${r2r1r0}{%k7}{z} # load keys + vmovdqu64 32($ctx),${r1r0s2}{%k7}{z} + vmovdqu64 24($ctx),${r0s2s1}{%k7}{z} + + vmovdqa64 96(%r10),$reduc_rght # .L2_44_shift_rgt + vmovdqa64 128(%r10),$reduc_left # .L2_44_shift_lft + + jmp .Loop_vpmadd52 + +.align 32 +.Loop_vpmadd52: + vmovdqu32 0($inp),%x#$T0 # load input as ----3210 + lea 16($inp),$inp + + vpermd $T0,$inp_permd,$T0 # ----3210 -> --322110 + vpsrlvq $inp_shift,$T0,$T0 + vpandq $reduc_mask,$T0,$T0 + vporq $PAD,$T0,$T0 + + vpaddq $T0,$Dlo,$Dlo # accumulate input + + vpermq \$0,$Dlo,${H0}{%k7}{z} # smash hash value + vpermq \$0b01010101,$Dlo,${H1}{%k7}{z} + vpermq \$0b10101010,$Dlo,${H2}{%k7}{z} + + vpxord $Dlo,$Dlo,$Dlo + vpxord $Dhi,$Dhi,$Dhi + + vpmadd52luq $r2r1r0,$H0,$Dlo + vpmadd52huq $r2r1r0,$H0,$Dhi + + vpmadd52luq $r1r0s2,$H1,$Dlo + vpmadd52huq $r1r0s2,$H1,$Dhi + + vpmadd52luq $r0s2s1,$H2,$Dlo + vpmadd52huq $r0s2s1,$H2,$Dhi + + vpsrlvq $reduc_rght,$Dlo,$T0 # 0 in topmost qword + vpsllvq $reduc_left,$Dhi,$Dhi # 0 in topmost qword + vpandq $reduc_mask,$Dlo,$Dlo + + vpaddq $T0,$Dhi,$Dhi + + vpermq \$0b10010011,$Dhi,$Dhi # 0 in lowest qword + + vpaddq $Dhi,$Dlo,$Dlo # note topmost qword :-) + + vpsrlvq $reduc_rght,$Dlo,$T0 # 0 in topmost word + vpandq $reduc_mask,$Dlo,$Dlo + + vpermq \$0b10010011,$T0,$T0 + + vpaddq $T0,$Dlo,$Dlo + + vpermq \$0b10010011,$Dlo,${T0}{%k1}{z} + + vpaddq $T0,$Dlo,$Dlo + vpsllq \$2,$T0,$T0 + + vpaddq $T0,$Dlo,$Dlo + + dec %rax # len-=16 + jnz .Loop_vpmadd52 + + vmovdqu64 $Dlo,0($ctx){%k7} # store hash value + + test $len,$len + jnz .Lblocks_vpmadd52_4x + +.Lno_data_vpmadd52: + ret +.size poly1305_blocks_vpmadd52,.-poly1305_blocks_vpmadd52 +___ +} +{ +######################################################################## +# As implied by its name 4x subroutine processes 4 blocks in parallel +# (but handles even 4*n+2 blocks lengths). It takes up to 4th key power +# and is handled in 256-bit %ymm registers. + +my ($H0,$H1,$H2,$R0,$R1,$R2,$S1,$S2) = map("%ymm$_",(0..5,16,17)); +my ($D0lo,$D0hi,$D1lo,$D1hi,$D2lo,$D2hi) = map("%ymm$_",(18..23)); +my ($T0,$T1,$T2,$T3,$mask44,$mask42,$tmp,$PAD) = map("%ymm$_",(24..31)); + +$code.=<<___; +.type poly1305_blocks_vpmadd52_4x,\@function,4 +.align 32 +poly1305_blocks_vpmadd52_4x: + shr \$4,$len + jz .Lno_data_vpmadd52_4x # too short + + shl \$40,$padbit + mov 64($ctx),%r8 # peek on power of the key + +.Lblocks_vpmadd52_4x: + vpbroadcastq $padbit,$PAD + + vmovdqa64 .Lx_mask44(%rip),$mask44 + mov \$5,%eax + vmovdqa64 .Lx_mask42(%rip),$mask42 + kmovw %eax,%k1 # used in 2x path + + test %r8,%r8 # is power value impossible? + js .Linit_vpmadd52 # if it is, then init R[4] + + vmovq 0($ctx),%x#$H0 # load current hash value + vmovq 8($ctx),%x#$H1 + vmovq 16($ctx),%x#$H2 + + test \$3,$len # is length 4*n+2? + jnz .Lblocks_vpmadd52_2x_do + +.Lblocks_vpmadd52_4x_do: + vpbroadcastq 64($ctx),$R0 # load 4th power of the key + vpbroadcastq 96($ctx),$R1 + vpbroadcastq 128($ctx),$R2 + vpbroadcastq 160($ctx),$S1 + +.Lblocks_vpmadd52_4x_key_loaded: + vpsllq \$2,$R2,$S2 # S2 = R2*5*4 + vpaddq $R2,$S2,$S2 + vpsllq \$2,$S2,$S2 + + test \$7,$len # is len 8*n? + jz .Lblocks_vpmadd52_8x + + vmovdqu64 16*0($inp),$T2 # load data + vmovdqu64 16*2($inp),$T3 + lea 16*4($inp),$inp + + vpunpcklqdq $T3,$T2,$T1 # transpose data + vpunpckhqdq $T3,$T2,$T3 + + # at this point 64-bit lanes are ordered as 3-1-2-0 + + vpsrlq \$24,$T3,$T2 # splat the data + vporq $PAD,$T2,$T2 + vpaddq $T2,$H2,$H2 # accumulate input + vpandq $mask44,$T1,$T0 + vpsrlq \$44,$T1,$T1 + vpsllq \$20,$T3,$T3 + vporq $T3,$T1,$T1 + vpandq $mask44,$T1,$T1 + + sub \$4,$len + jz .Ltail_vpmadd52_4x + jmp .Loop_vpmadd52_4x + ud2 + +.align 32 +.Linit_vpmadd52: + vmovq 24($ctx),%x#$S1 # load key + vmovq 56($ctx),%x#$H2 + vmovq 32($ctx),%x#$S2 + vmovq 40($ctx),%x#$R0 + vmovq 48($ctx),%x#$R1 + + vmovdqa $R0,$H0 + vmovdqa $R1,$H1 + vmovdqa $H2,$R2 + + mov \$2,%eax + +.Lmul_init_vpmadd52: + vpxorq $D0lo,$D0lo,$D0lo + vpmadd52luq $H2,$S1,$D0lo + vpxorq $D0hi,$D0hi,$D0hi + vpmadd52huq $H2,$S1,$D0hi + vpxorq $D1lo,$D1lo,$D1lo + vpmadd52luq $H2,$S2,$D1lo + vpxorq $D1hi,$D1hi,$D1hi + vpmadd52huq $H2,$S2,$D1hi + vpxorq $D2lo,$D2lo,$D2lo + vpmadd52luq $H2,$R0,$D2lo + vpxorq $D2hi,$D2hi,$D2hi + vpmadd52huq $H2,$R0,$D2hi + + vpmadd52luq $H0,$R0,$D0lo + vpmadd52huq $H0,$R0,$D0hi + vpmadd52luq $H0,$R1,$D1lo + vpmadd52huq $H0,$R1,$D1hi + vpmadd52luq $H0,$R2,$D2lo + vpmadd52huq $H0,$R2,$D2hi + + vpmadd52luq $H1,$S2,$D0lo + vpmadd52huq $H1,$S2,$D0hi + vpmadd52luq $H1,$R0,$D1lo + vpmadd52huq $H1,$R0,$D1hi + vpmadd52luq $H1,$R1,$D2lo + vpmadd52huq $H1,$R1,$D2hi + + ################################################################ + # partial reduction + vpsrlq \$44,$D0lo,$tmp + vpsllq \$8,$D0hi,$D0hi + vpandq $mask44,$D0lo,$H0 + vpaddq $tmp,$D0hi,$D0hi + + vpaddq $D0hi,$D1lo,$D1lo + + vpsrlq \$44,$D1lo,$tmp + vpsllq \$8,$D1hi,$D1hi + vpandq $mask44,$D1lo,$H1 + vpaddq $tmp,$D1hi,$D1hi + + vpaddq $D1hi,$D2lo,$D2lo + + vpsrlq \$42,$D2lo,$tmp + vpsllq \$10,$D2hi,$D2hi + vpandq $mask42,$D2lo,$H2 + vpaddq $tmp,$D2hi,$D2hi + + vpaddq $D2hi,$H0,$H0 + vpsllq \$2,$D2hi,$D2hi + + vpaddq $D2hi,$H0,$H0 + + vpsrlq \$44,$H0,$tmp # additional step + vpandq $mask44,$H0,$H0 + + vpaddq $tmp,$H1,$H1 + + dec %eax + jz .Ldone_init_vpmadd52 + + vpunpcklqdq $R1,$H1,$R1 # 1,2 + vpbroadcastq %x#$H1,%x#$H1 # 2,2 + vpunpcklqdq $R2,$H2,$R2 + vpbroadcastq %x#$H2,%x#$H2 + vpunpcklqdq $R0,$H0,$R0 + vpbroadcastq %x#$H0,%x#$H0 + + vpsllq \$2,$R1,$S1 # S1 = R1*5*4 + vpsllq \$2,$R2,$S2 # S2 = R2*5*4 + vpaddq $R1,$S1,$S1 + vpaddq $R2,$S2,$S2 + vpsllq \$2,$S1,$S1 + vpsllq \$2,$S2,$S2 + + jmp .Lmul_init_vpmadd52 + ud2 + +.align 32 +.Ldone_init_vpmadd52: + vinserti128 \$1,%x#$R1,$H1,$R1 # 1,2,3,4 + vinserti128 \$1,%x#$R2,$H2,$R2 + vinserti128 \$1,%x#$R0,$H0,$R0 + + vpermq \$0b11011000,$R1,$R1 # 1,3,2,4 + vpermq \$0b11011000,$R2,$R2 + vpermq \$0b11011000,$R0,$R0 + + vpsllq \$2,$R1,$S1 # S1 = R1*5*4 + vpaddq $R1,$S1,$S1 + vpsllq \$2,$S1,$S1 + + vmovq 0($ctx),%x#$H0 # load current hash value + vmovq 8($ctx),%x#$H1 + vmovq 16($ctx),%x#$H2 + + test \$3,$len # is length 4*n+2? + jnz .Ldone_init_vpmadd52_2x + + vmovdqu64 $R0,64($ctx) # save key powers + vpbroadcastq %x#$R0,$R0 # broadcast 4th power + vmovdqu64 $R1,96($ctx) + vpbroadcastq %x#$R1,$R1 + vmovdqu64 $R2,128($ctx) + vpbroadcastq %x#$R2,$R2 + vmovdqu64 $S1,160($ctx) + vpbroadcastq %x#$S1,$S1 + + jmp .Lblocks_vpmadd52_4x_key_loaded + ud2 + +.align 32 +.Ldone_init_vpmadd52_2x: + vmovdqu64 $R0,64($ctx) # save key powers + vpsrldq \$8,$R0,$R0 # 0-1-0-2 + vmovdqu64 $R1,96($ctx) + vpsrldq \$8,$R1,$R1 + vmovdqu64 $R2,128($ctx) + vpsrldq \$8,$R2,$R2 + vmovdqu64 $S1,160($ctx) + vpsrldq \$8,$S1,$S1 + jmp .Lblocks_vpmadd52_2x_key_loaded + ud2 + +.align 32 +.Lblocks_vpmadd52_2x_do: + vmovdqu64 128+8($ctx),${R2}{%k1}{z}# load 2nd and 1st key powers + vmovdqu64 160+8($ctx),${S1}{%k1}{z} + vmovdqu64 64+8($ctx),${R0}{%k1}{z} + vmovdqu64 96+8($ctx),${R1}{%k1}{z} + +.Lblocks_vpmadd52_2x_key_loaded: + vmovdqu64 16*0($inp),$T2 # load data + vpxorq $T3,$T3,$T3 + lea 16*2($inp),$inp + + vpunpcklqdq $T3,$T2,$T1 # transpose data + vpunpckhqdq $T3,$T2,$T3 + + # at this point 64-bit lanes are ordered as x-1-x-0 + + vpsrlq \$24,$T3,$T2 # splat the data + vporq $PAD,$T2,$T2 + vpaddq $T2,$H2,$H2 # accumulate input + vpandq $mask44,$T1,$T0 + vpsrlq \$44,$T1,$T1 + vpsllq \$20,$T3,$T3 + vporq $T3,$T1,$T1 + vpandq $mask44,$T1,$T1 + + jmp .Ltail_vpmadd52_2x + ud2 + +.align 32 +.Loop_vpmadd52_4x: + #vpaddq $T2,$H2,$H2 # accumulate input + vpaddq $T0,$H0,$H0 + vpaddq $T1,$H1,$H1 + + vpxorq $D0lo,$D0lo,$D0lo + vpmadd52luq $H2,$S1,$D0lo + vpxorq $D0hi,$D0hi,$D0hi + vpmadd52huq $H2,$S1,$D0hi + vpxorq $D1lo,$D1lo,$D1lo + vpmadd52luq $H2,$S2,$D1lo + vpxorq $D1hi,$D1hi,$D1hi + vpmadd52huq $H2,$S2,$D1hi + vpxorq $D2lo,$D2lo,$D2lo + vpmadd52luq $H2,$R0,$D2lo + vpxorq $D2hi,$D2hi,$D2hi + vpmadd52huq $H2,$R0,$D2hi + + vmovdqu64 16*0($inp),$T2 # load data + vmovdqu64 16*2($inp),$T3 + lea 16*4($inp),$inp + vpmadd52luq $H0,$R0,$D0lo + vpmadd52huq $H0,$R0,$D0hi + vpmadd52luq $H0,$R1,$D1lo + vpmadd52huq $H0,$R1,$D1hi + vpmadd52luq $H0,$R2,$D2lo + vpmadd52huq $H0,$R2,$D2hi + + vpunpcklqdq $T3,$T2,$T1 # transpose data + vpunpckhqdq $T3,$T2,$T3 + vpmadd52luq $H1,$S2,$D0lo + vpmadd52huq $H1,$S2,$D0hi + vpmadd52luq $H1,$R0,$D1lo + vpmadd52huq $H1,$R0,$D1hi + vpmadd52luq $H1,$R1,$D2lo + vpmadd52huq $H1,$R1,$D2hi + + ################################################################ + # partial reduction (interleaved with data splat) + vpsrlq \$44,$D0lo,$tmp + vpsllq \$8,$D0hi,$D0hi + vpandq $mask44,$D0lo,$H0 + vpaddq $tmp,$D0hi,$D0hi + + vpsrlq \$24,$T3,$T2 + vporq $PAD,$T2,$T2 + vpaddq $D0hi,$D1lo,$D1lo + + vpsrlq \$44,$D1lo,$tmp + vpsllq \$8,$D1hi,$D1hi + vpandq $mask44,$D1lo,$H1 + vpaddq $tmp,$D1hi,$D1hi + + vpandq $mask44,$T1,$T0 + vpsrlq \$44,$T1,$T1 + vpsllq \$20,$T3,$T3 + vpaddq $D1hi,$D2lo,$D2lo + + vpsrlq \$42,$D2lo,$tmp + vpsllq \$10,$D2hi,$D2hi + vpandq $mask42,$D2lo,$H2 + vpaddq $tmp,$D2hi,$D2hi + + vpaddq $T2,$H2,$H2 # accumulate input + vpaddq $D2hi,$H0,$H0 + vpsllq \$2,$D2hi,$D2hi + + vpaddq $D2hi,$H0,$H0 + vporq $T3,$T1,$T1 + vpandq $mask44,$T1,$T1 + + vpsrlq \$44,$H0,$tmp # additional step + vpandq $mask44,$H0,$H0 + + vpaddq $tmp,$H1,$H1 + + sub \$4,$len # len-=64 + jnz .Loop_vpmadd52_4x + +.Ltail_vpmadd52_4x: + vmovdqu64 128($ctx),$R2 # load all key powers + vmovdqu64 160($ctx),$S1 + vmovdqu64 64($ctx),$R0 + vmovdqu64 96($ctx),$R1 + +.Ltail_vpmadd52_2x: + vpsllq \$2,$R2,$S2 # S2 = R2*5*4 + vpaddq $R2,$S2,$S2 + vpsllq \$2,$S2,$S2 + + #vpaddq $T2,$H2,$H2 # accumulate input + vpaddq $T0,$H0,$H0 + vpaddq $T1,$H1,$H1 + + vpxorq $D0lo,$D0lo,$D0lo + vpmadd52luq $H2,$S1,$D0lo + vpxorq $D0hi,$D0hi,$D0hi + vpmadd52huq $H2,$S1,$D0hi + vpxorq $D1lo,$D1lo,$D1lo + vpmadd52luq $H2,$S2,$D1lo + vpxorq $D1hi,$D1hi,$D1hi + vpmadd52huq $H2,$S2,$D1hi + vpxorq $D2lo,$D2lo,$D2lo + vpmadd52luq $H2,$R0,$D2lo + vpxorq $D2hi,$D2hi,$D2hi + vpmadd52huq $H2,$R0,$D2hi + + vpmadd52luq $H0,$R0,$D0lo + vpmadd52huq $H0,$R0,$D0hi + vpmadd52luq $H0,$R1,$D1lo + vpmadd52huq $H0,$R1,$D1hi + vpmadd52luq $H0,$R2,$D2lo + vpmadd52huq $H0,$R2,$D2hi + + vpmadd52luq $H1,$S2,$D0lo + vpmadd52huq $H1,$S2,$D0hi + vpmadd52luq $H1,$R0,$D1lo + vpmadd52huq $H1,$R0,$D1hi + vpmadd52luq $H1,$R1,$D2lo + vpmadd52huq $H1,$R1,$D2hi + + ################################################################ + # horizontal addition + + mov \$1,%eax + kmovw %eax,%k1 + vpsrldq \$8,$D0lo,$T0 + vpsrldq \$8,$D0hi,$H0 + vpsrldq \$8,$D1lo,$T1 + vpsrldq \$8,$D1hi,$H1 + vpaddq $T0,$D0lo,$D0lo + vpaddq $H0,$D0hi,$D0hi + vpsrldq \$8,$D2lo,$T2 + vpsrldq \$8,$D2hi,$H2 + vpaddq $T1,$D1lo,$D1lo + vpaddq $H1,$D1hi,$D1hi + vpermq \$0x2,$D0lo,$T0 + vpermq \$0x2,$D0hi,$H0 + vpaddq $T2,$D2lo,$D2lo + vpaddq $H2,$D2hi,$D2hi + + vpermq \$0x2,$D1lo,$T1 + vpermq \$0x2,$D1hi,$H1 + vpaddq $T0,$D0lo,${D0lo}{%k1}{z} + vpaddq $H0,$D0hi,${D0hi}{%k1}{z} + vpermq \$0x2,$D2lo,$T2 + vpermq \$0x2,$D2hi,$H2 + vpaddq $T1,$D1lo,${D1lo}{%k1}{z} + vpaddq $H1,$D1hi,${D1hi}{%k1}{z} + vpaddq $T2,$D2lo,${D2lo}{%k1}{z} + vpaddq $H2,$D2hi,${D2hi}{%k1}{z} + + ################################################################ + # partial reduction + vpsrlq \$44,$D0lo,$tmp + vpsllq \$8,$D0hi,$D0hi + vpandq $mask44,$D0lo,$H0 + vpaddq $tmp,$D0hi,$D0hi + + vpaddq $D0hi,$D1lo,$D1lo + + vpsrlq \$44,$D1lo,$tmp + vpsllq \$8,$D1hi,$D1hi + vpandq $mask44,$D1lo,$H1 + vpaddq $tmp,$D1hi,$D1hi + + vpaddq $D1hi,$D2lo,$D2lo + + vpsrlq \$42,$D2lo,$tmp + vpsllq \$10,$D2hi,$D2hi + vpandq $mask42,$D2lo,$H2 + vpaddq $tmp,$D2hi,$D2hi + + vpaddq $D2hi,$H0,$H0 + vpsllq \$2,$D2hi,$D2hi + + vpaddq $D2hi,$H0,$H0 + + vpsrlq \$44,$H0,$tmp # additional step + vpandq $mask44,$H0,$H0 + + vpaddq $tmp,$H1,$H1 + # at this point $len is + # either 4*n+2 or 0... + sub \$2,$len # len-=32 + ja .Lblocks_vpmadd52_4x_do + + vmovq %x#$H0,0($ctx) + vmovq %x#$H1,8($ctx) + vmovq %x#$H2,16($ctx) + vzeroall + +.Lno_data_vpmadd52_4x: + ret +.size poly1305_blocks_vpmadd52_4x,.-poly1305_blocks_vpmadd52_4x +___ +} +{ +######################################################################## +# As implied by its name 8x subroutine processes 8 blocks in parallel... +# This is intermediate version, as it's used only in cases when input +# length is either 8*n, 8*n+1 or 8*n+2... + +my ($H0,$H1,$H2,$R0,$R1,$R2,$S1,$S2) = map("%ymm$_",(0..5,16,17)); +my ($D0lo,$D0hi,$D1lo,$D1hi,$D2lo,$D2hi) = map("%ymm$_",(18..23)); +my ($T0,$T1,$T2,$T3,$mask44,$mask42,$tmp,$PAD) = map("%ymm$_",(24..31)); +my ($RR0,$RR1,$RR2,$SS1,$SS2) = map("%ymm$_",(6..10)); + +$code.=<<___; +.type poly1305_blocks_vpmadd52_8x,\@function,4 +.align 32 +poly1305_blocks_vpmadd52_8x: + shr \$4,$len + jz .Lno_data_vpmadd52_8x # too short + + shl \$40,$padbit + mov 64($ctx),%r8 # peek on power of the key + + vmovdqa64 .Lx_mask44(%rip),$mask44 + vmovdqa64 .Lx_mask42(%rip),$mask42 + + test %r8,%r8 # is power value impossible? + js .Linit_vpmadd52 # if it is, then init R[4] + + vmovq 0($ctx),%x#$H0 # load current hash value + vmovq 8($ctx),%x#$H1 + vmovq 16($ctx),%x#$H2 + +.Lblocks_vpmadd52_8x: + ################################################################ + # fist we calculate more key powers + + vmovdqu64 128($ctx),$R2 # load 1-3-2-4 powers + vmovdqu64 160($ctx),$S1 + vmovdqu64 64($ctx),$R0 + vmovdqu64 96($ctx),$R1 + + vpsllq \$2,$R2,$S2 # S2 = R2*5*4 + vpaddq $R2,$S2,$S2 + vpsllq \$2,$S2,$S2 + + vpbroadcastq %x#$R2,$RR2 # broadcast 4th power + vpbroadcastq %x#$R0,$RR0 + vpbroadcastq %x#$R1,$RR1 + + vpxorq $D0lo,$D0lo,$D0lo + vpmadd52luq $RR2,$S1,$D0lo + vpxorq $D0hi,$D0hi,$D0hi + vpmadd52huq $RR2,$S1,$D0hi + vpxorq $D1lo,$D1lo,$D1lo + vpmadd52luq $RR2,$S2,$D1lo + vpxorq $D1hi,$D1hi,$D1hi + vpmadd52huq $RR2,$S2,$D1hi + vpxorq $D2lo,$D2lo,$D2lo + vpmadd52luq $RR2,$R0,$D2lo + vpxorq $D2hi,$D2hi,$D2hi + vpmadd52huq $RR2,$R0,$D2hi + + vpmadd52luq $RR0,$R0,$D0lo + vpmadd52huq $RR0,$R0,$D0hi + vpmadd52luq $RR0,$R1,$D1lo + vpmadd52huq $RR0,$R1,$D1hi + vpmadd52luq $RR0,$R2,$D2lo + vpmadd52huq $RR0,$R2,$D2hi + + vpmadd52luq $RR1,$S2,$D0lo + vpmadd52huq $RR1,$S2,$D0hi + vpmadd52luq $RR1,$R0,$D1lo + vpmadd52huq $RR1,$R0,$D1hi + vpmadd52luq $RR1,$R1,$D2lo + vpmadd52huq $RR1,$R1,$D2hi + + ################################################################ + # partial reduction + vpsrlq \$44,$D0lo,$tmp + vpsllq \$8,$D0hi,$D0hi + vpandq $mask44,$D0lo,$RR0 + vpaddq $tmp,$D0hi,$D0hi + + vpaddq $D0hi,$D1lo,$D1lo + + vpsrlq \$44,$D1lo,$tmp + vpsllq \$8,$D1hi,$D1hi + vpandq $mask44,$D1lo,$RR1 + vpaddq $tmp,$D1hi,$D1hi + + vpaddq $D1hi,$D2lo,$D2lo + + vpsrlq \$42,$D2lo,$tmp + vpsllq \$10,$D2hi,$D2hi + vpandq $mask42,$D2lo,$RR2 + vpaddq $tmp,$D2hi,$D2hi + + vpaddq $D2hi,$RR0,$RR0 + vpsllq \$2,$D2hi,$D2hi + + vpaddq $D2hi,$RR0,$RR0 + + vpsrlq \$44,$RR0,$tmp # additional step + vpandq $mask44,$RR0,$RR0 + + vpaddq $tmp,$RR1,$RR1 + + ################################################################ + # At this point Rx holds 1324 powers, RRx - 5768, and the goal + # is 15263748, which reflects how data is loaded... + + vpunpcklqdq $R2,$RR2,$T2 # 3748 + vpunpckhqdq $R2,$RR2,$R2 # 1526 + vpunpcklqdq $R0,$RR0,$T0 + vpunpckhqdq $R0,$RR0,$R0 + vpunpcklqdq $R1,$RR1,$T1 + vpunpckhqdq $R1,$RR1,$R1 +___ +######## switch to %zmm +map(s/%y/%z/, $H0,$H1,$H2,$R0,$R1,$R2,$S1,$S2); +map(s/%y/%z/, $D0lo,$D0hi,$D1lo,$D1hi,$D2lo,$D2hi); +map(s/%y/%z/, $T0,$T1,$T2,$T3,$mask44,$mask42,$tmp,$PAD); +map(s/%y/%z/, $RR0,$RR1,$RR2,$SS1,$SS2); + +$code.=<<___; + vshufi64x2 \$0x44,$R2,$T2,$RR2 # 15263748 + vshufi64x2 \$0x44,$R0,$T0,$RR0 + vshufi64x2 \$0x44,$R1,$T1,$RR1 + + vmovdqu64 16*0($inp),$T2 # load data + vmovdqu64 16*4($inp),$T3 + lea 16*8($inp),$inp + + vpsllq \$2,$RR2,$SS2 # S2 = R2*5*4 + vpsllq \$2,$RR1,$SS1 # S1 = R1*5*4 + vpaddq $RR2,$SS2,$SS2 + vpaddq $RR1,$SS1,$SS1 + vpsllq \$2,$SS2,$SS2 + vpsllq \$2,$SS1,$SS1 + + vpbroadcastq $padbit,$PAD + vpbroadcastq %x#$mask44,$mask44 + vpbroadcastq %x#$mask42,$mask42 + + vpbroadcastq %x#$SS1,$S1 # broadcast 8th power + vpbroadcastq %x#$SS2,$S2 + vpbroadcastq %x#$RR0,$R0 + vpbroadcastq %x#$RR1,$R1 + vpbroadcastq %x#$RR2,$R2 + + vpunpcklqdq $T3,$T2,$T1 # transpose data + vpunpckhqdq $T3,$T2,$T3 + + # at this point 64-bit lanes are ordered as 73625140 + + vpsrlq \$24,$T3,$T2 # splat the data + vporq $PAD,$T2,$T2 + vpaddq $T2,$H2,$H2 # accumulate input + vpandq $mask44,$T1,$T0 + vpsrlq \$44,$T1,$T1 + vpsllq \$20,$T3,$T3 + vporq $T3,$T1,$T1 + vpandq $mask44,$T1,$T1 + + sub \$8,$len + jz .Ltail_vpmadd52_8x + jmp .Loop_vpmadd52_8x + +.align 32 +.Loop_vpmadd52_8x: + #vpaddq $T2,$H2,$H2 # accumulate input + vpaddq $T0,$H0,$H0 + vpaddq $T1,$H1,$H1 + + vpxorq $D0lo,$D0lo,$D0lo + vpmadd52luq $H2,$S1,$D0lo + vpxorq $D0hi,$D0hi,$D0hi + vpmadd52huq $H2,$S1,$D0hi + vpxorq $D1lo,$D1lo,$D1lo + vpmadd52luq $H2,$S2,$D1lo + vpxorq $D1hi,$D1hi,$D1hi + vpmadd52huq $H2,$S2,$D1hi + vpxorq $D2lo,$D2lo,$D2lo + vpmadd52luq $H2,$R0,$D2lo + vpxorq $D2hi,$D2hi,$D2hi + vpmadd52huq $H2,$R0,$D2hi + + vmovdqu64 16*0($inp),$T2 # load data + vmovdqu64 16*4($inp),$T3 + lea 16*8($inp),$inp + vpmadd52luq $H0,$R0,$D0lo + vpmadd52huq $H0,$R0,$D0hi + vpmadd52luq $H0,$R1,$D1lo + vpmadd52huq $H0,$R1,$D1hi + vpmadd52luq $H0,$R2,$D2lo + vpmadd52huq $H0,$R2,$D2hi + + vpunpcklqdq $T3,$T2,$T1 # transpose data + vpunpckhqdq $T3,$T2,$T3 + vpmadd52luq $H1,$S2,$D0lo + vpmadd52huq $H1,$S2,$D0hi + vpmadd52luq $H1,$R0,$D1lo + vpmadd52huq $H1,$R0,$D1hi + vpmadd52luq $H1,$R1,$D2lo + vpmadd52huq $H1,$R1,$D2hi + + ################################################################ + # partial reduction (interleaved with data splat) + vpsrlq \$44,$D0lo,$tmp + vpsllq \$8,$D0hi,$D0hi + vpandq $mask44,$D0lo,$H0 + vpaddq $tmp,$D0hi,$D0hi + + vpsrlq \$24,$T3,$T2 + vporq $PAD,$T2,$T2 + vpaddq $D0hi,$D1lo,$D1lo + + vpsrlq \$44,$D1lo,$tmp + vpsllq \$8,$D1hi,$D1hi + vpandq $mask44,$D1lo,$H1 + vpaddq $tmp,$D1hi,$D1hi + + vpandq $mask44,$T1,$T0 + vpsrlq \$44,$T1,$T1 + vpsllq \$20,$T3,$T3 + vpaddq $D1hi,$D2lo,$D2lo + + vpsrlq \$42,$D2lo,$tmp + vpsllq \$10,$D2hi,$D2hi + vpandq $mask42,$D2lo,$H2 + vpaddq $tmp,$D2hi,$D2hi + + vpaddq $T2,$H2,$H2 # accumulate input + vpaddq $D2hi,$H0,$H0 + vpsllq \$2,$D2hi,$D2hi + + vpaddq $D2hi,$H0,$H0 + vporq $T3,$T1,$T1 + vpandq $mask44,$T1,$T1 + + vpsrlq \$44,$H0,$tmp # additional step + vpandq $mask44,$H0,$H0 + + vpaddq $tmp,$H1,$H1 + + sub \$8,$len # len-=128 + jnz .Loop_vpmadd52_8x + +.Ltail_vpmadd52_8x: + #vpaddq $T2,$H2,$H2 # accumulate input + vpaddq $T0,$H0,$H0 + vpaddq $T1,$H1,$H1 + + vpxorq $D0lo,$D0lo,$D0lo + vpmadd52luq $H2,$SS1,$D0lo + vpxorq $D0hi,$D0hi,$D0hi + vpmadd52huq $H2,$SS1,$D0hi + vpxorq $D1lo,$D1lo,$D1lo + vpmadd52luq $H2,$SS2,$D1lo + vpxorq $D1hi,$D1hi,$D1hi + vpmadd52huq $H2,$SS2,$D1hi + vpxorq $D2lo,$D2lo,$D2lo + vpmadd52luq $H2,$RR0,$D2lo + vpxorq $D2hi,$D2hi,$D2hi + vpmadd52huq $H2,$RR0,$D2hi + + vpmadd52luq $H0,$RR0,$D0lo + vpmadd52huq $H0,$RR0,$D0hi + vpmadd52luq $H0,$RR1,$D1lo + vpmadd52huq $H0,$RR1,$D1hi + vpmadd52luq $H0,$RR2,$D2lo + vpmadd52huq $H0,$RR2,$D2hi + + vpmadd52luq $H1,$SS2,$D0lo + vpmadd52huq $H1,$SS2,$D0hi + vpmadd52luq $H1,$RR0,$D1lo + vpmadd52huq $H1,$RR0,$D1hi + vpmadd52luq $H1,$RR1,$D2lo + vpmadd52huq $H1,$RR1,$D2hi + + ################################################################ + # horizontal addition + + mov \$1,%eax + kmovw %eax,%k1 + vpsrldq \$8,$D0lo,$T0 + vpsrldq \$8,$D0hi,$H0 + vpsrldq \$8,$D1lo,$T1 + vpsrldq \$8,$D1hi,$H1 + vpaddq $T0,$D0lo,$D0lo + vpaddq $H0,$D0hi,$D0hi + vpsrldq \$8,$D2lo,$T2 + vpsrldq \$8,$D2hi,$H2 + vpaddq $T1,$D1lo,$D1lo + vpaddq $H1,$D1hi,$D1hi + vpermq \$0x2,$D0lo,$T0 + vpermq \$0x2,$D0hi,$H0 + vpaddq $T2,$D2lo,$D2lo + vpaddq $H2,$D2hi,$D2hi + + vpermq \$0x2,$D1lo,$T1 + vpermq \$0x2,$D1hi,$H1 + vpaddq $T0,$D0lo,$D0lo + vpaddq $H0,$D0hi,$D0hi + vpermq \$0x2,$D2lo,$T2 + vpermq \$0x2,$D2hi,$H2 + vpaddq $T1,$D1lo,$D1lo + vpaddq $H1,$D1hi,$D1hi + vextracti64x4 \$1,$D0lo,%y#$T0 + vextracti64x4 \$1,$D0hi,%y#$H0 + vpaddq $T2,$D2lo,$D2lo + vpaddq $H2,$D2hi,$D2hi + + vextracti64x4 \$1,$D1lo,%y#$T1 + vextracti64x4 \$1,$D1hi,%y#$H1 + vextracti64x4 \$1,$D2lo,%y#$T2 + vextracti64x4 \$1,$D2hi,%y#$H2 +___ +######## switch back to %ymm +map(s/%z/%y/, $H0,$H1,$H2,$R0,$R1,$R2,$S1,$S2); +map(s/%z/%y/, $D0lo,$D0hi,$D1lo,$D1hi,$D2lo,$D2hi); +map(s/%z/%y/, $T0,$T1,$T2,$T3,$mask44,$mask42,$tmp,$PAD); + +$code.=<<___; + vpaddq $T0,$D0lo,${D0lo}{%k1}{z} + vpaddq $H0,$D0hi,${D0hi}{%k1}{z} + vpaddq $T1,$D1lo,${D1lo}{%k1}{z} + vpaddq $H1,$D1hi,${D1hi}{%k1}{z} + vpaddq $T2,$D2lo,${D2lo}{%k1}{z} + vpaddq $H2,$D2hi,${D2hi}{%k1}{z} + + ################################################################ + # partial reduction + vpsrlq \$44,$D0lo,$tmp + vpsllq \$8,$D0hi,$D0hi + vpandq $mask44,$D0lo,$H0 + vpaddq $tmp,$D0hi,$D0hi + + vpaddq $D0hi,$D1lo,$D1lo + + vpsrlq \$44,$D1lo,$tmp + vpsllq \$8,$D1hi,$D1hi + vpandq $mask44,$D1lo,$H1 + vpaddq $tmp,$D1hi,$D1hi + + vpaddq $D1hi,$D2lo,$D2lo + + vpsrlq \$42,$D2lo,$tmp + vpsllq \$10,$D2hi,$D2hi + vpandq $mask42,$D2lo,$H2 + vpaddq $tmp,$D2hi,$D2hi + + vpaddq $D2hi,$H0,$H0 + vpsllq \$2,$D2hi,$D2hi + + vpaddq $D2hi,$H0,$H0 + + vpsrlq \$44,$H0,$tmp # additional step + vpandq $mask44,$H0,$H0 + + vpaddq $tmp,$H1,$H1 + + ################################################################ + + vmovq %x#$H0,0($ctx) + vmovq %x#$H1,8($ctx) + vmovq %x#$H2,16($ctx) + vzeroall + +.Lno_data_vpmadd52_8x: + ret +.size poly1305_blocks_vpmadd52_8x,.-poly1305_blocks_vpmadd52_8x +___ +} +$code.=<<___; +.type poly1305_emit_base2_44,\@function,3 +.align 32 +poly1305_emit_base2_44: + mov 0($ctx),%r8 # load hash value + mov 8($ctx),%r9 + mov 16($ctx),%r10 + + mov %r9,%rax + shr \$20,%r9 + shl \$44,%rax + mov %r10,%rcx + shr \$40,%r10 + shl \$24,%rcx + + add %rax,%r8 + adc %rcx,%r9 + adc \$0,%r10 + + mov %r8,%rax + add \$5,%r8 # compare to modulus + mov %r9,%rcx + adc \$0,%r9 + adc \$0,%r10 + shr \$2,%r10 # did 130-bit value overflow? + cmovnz %r8,%rax + cmovnz %r9,%rcx + + add 0($nonce),%rax # accumulate nonce + adc 8($nonce),%rcx + mov %rax,0($mac) # write result + mov %rcx,8($mac) + + ret +.size poly1305_emit_base2_44,.-poly1305_emit_base2_44 +___ +} } } +} + +if (!$kernel) +{ # chacha20-poly1305 helpers +my ($out,$inp,$otp,$len)=$win64 ? ("%rcx","%rdx","%r8", "%r9") : # Win64 order + ("%rdi","%rsi","%rdx","%rcx"); # Unix order +$code.=<<___; +.globl xor128_encrypt_n_pad +.type xor128_encrypt_n_pad,\@abi-omnipotent +.align 16 +xor128_encrypt_n_pad: + sub $otp,$inp + sub $otp,$out + mov $len,%r10 # put len aside + shr \$4,$len # len / 16 + jz .Ltail_enc + nop +.Loop_enc_xmm: + movdqu ($inp,$otp),%xmm0 + pxor ($otp),%xmm0 + movdqu %xmm0,($out,$otp) + movdqa %xmm0,($otp) + lea 16($otp),$otp + dec $len + jnz .Loop_enc_xmm + + and \$15,%r10 # len % 16 + jz .Ldone_enc + +.Ltail_enc: + mov \$16,$len + sub %r10,$len + xor %eax,%eax +.Loop_enc_byte: + mov ($inp,$otp),%al + xor ($otp),%al + mov %al,($out,$otp) + mov %al,($otp) + lea 1($otp),$otp + dec %r10 + jnz .Loop_enc_byte + + xor %eax,%eax +.Loop_enc_pad: + mov %al,($otp) + lea 1($otp),$otp + dec $len + jnz .Loop_enc_pad + +.Ldone_enc: + mov $otp,%rax + ret +.size xor128_encrypt_n_pad,.-xor128_encrypt_n_pad + +.globl xor128_decrypt_n_pad +.type xor128_decrypt_n_pad,\@abi-omnipotent +.align 16 +xor128_decrypt_n_pad: + sub $otp,$inp + sub $otp,$out + mov $len,%r10 # put len aside + shr \$4,$len # len / 16 + jz .Ltail_dec + nop +.Loop_dec_xmm: + movdqu ($inp,$otp),%xmm0 + movdqa ($otp),%xmm1 + pxor %xmm0,%xmm1 + movdqu %xmm1,($out,$otp) + movdqa %xmm0,($otp) + lea 16($otp),$otp + dec $len + jnz .Loop_dec_xmm + + pxor %xmm1,%xmm1 + and \$15,%r10 # len % 16 + jz .Ldone_dec + +.Ltail_dec: + mov \$16,$len + sub %r10,$len + xor %eax,%eax + xor %r11,%r11 +.Loop_dec_byte: + mov ($inp,$otp),%r11b + mov ($otp),%al + xor %r11b,%al + mov %al,($out,$otp) + mov %r11b,($otp) + lea 1($otp),$otp + dec %r10 + jnz .Loop_dec_byte + + xor %eax,%eax +.Loop_dec_pad: + mov %al,($otp) + lea 1($otp),$otp + dec $len + jnz .Loop_dec_pad + +.Ldone_dec: + mov $otp,%rax + ret +.size xor128_decrypt_n_pad,.-xor128_decrypt_n_pad +___ +} + +# EXCEPTION_DISPOSITION handler (EXCEPTION_RECORD *rec,ULONG64 frame, +# CONTEXT *context,DISPATCHER_CONTEXT *disp) +if ($win64) { +$rec="%rcx"; +$frame="%rdx"; +$context="%r8"; +$disp="%r9"; + +$code.=<<___; +.extern __imp_RtlVirtualUnwind +.type se_handler,\@abi-omnipotent +.align 16 +se_handler: + push %rsi + push %rdi + push %rbx + push %rbp + push %r12 + push %r13 + push %r14 + push %r15 + pushfq + sub \$64,%rsp + + mov 120($context),%rax # pull context->Rax + mov 248($context),%rbx # pull context->Rip + + mov 8($disp),%rsi # disp->ImageBase + mov 56($disp),%r11 # disp->HandlerData + + mov 0(%r11),%r10d # HandlerData[0] + lea (%rsi,%r10),%r10 # prologue label + cmp %r10,%rbx # context->Rip<.Lprologue + jb .Lcommon_seh_tail + + mov 152($context),%rax # pull context->Rsp + + mov 4(%r11),%r10d # HandlerData[1] + lea (%rsi,%r10),%r10 # epilogue label + cmp %r10,%rbx # context->Rip>=.Lepilogue + jae .Lcommon_seh_tail + + lea 48(%rax),%rax + + mov -8(%rax),%rbx + mov -16(%rax),%rbp + mov -24(%rax),%r12 + mov -32(%rax),%r13 + mov -40(%rax),%r14 + mov -48(%rax),%r15 + mov %rbx,144($context) # restore context->Rbx + mov %rbp,160($context) # restore context->Rbp + mov %r12,216($context) # restore context->R12 + mov %r13,224($context) # restore context->R13 + mov %r14,232($context) # restore context->R14 + mov %r15,240($context) # restore context->R14 + + jmp .Lcommon_seh_tail +.size se_handler,.-se_handler + +.type avx_handler,\@abi-omnipotent +.align 16 +avx_handler: + push %rsi + push %rdi + push %rbx + push %rbp + push %r12 + push %r13 + push %r14 + push %r15 + pushfq + sub \$64,%rsp + + mov 120($context),%rax # pull context->Rax + mov 248($context),%rbx # pull context->Rip + + mov 8($disp),%rsi # disp->ImageBase + mov 56($disp),%r11 # disp->HandlerData + + mov 0(%r11),%r10d # HandlerData[0] + lea (%rsi,%r10),%r10 # prologue label + cmp %r10,%rbx # context->Rip<prologue label + jb .Lcommon_seh_tail + + mov 152($context),%rax # pull context->Rsp + + mov 4(%r11),%r10d # HandlerData[1] + lea (%rsi,%r10),%r10 # epilogue label + cmp %r10,%rbx # context->Rip>=epilogue label + jae .Lcommon_seh_tail + + mov 208($context),%rax # pull context->R11 + + lea 0x50(%rax),%rsi + lea 0xf8(%rax),%rax + lea 512($context),%rdi # &context.Xmm6 + mov \$20,%ecx + .long 0xa548f3fc # cld; rep movsq + +.Lcommon_seh_tail: + mov 8(%rax),%rdi + mov 16(%rax),%rsi + mov %rax,152($context) # restore context->Rsp + mov %rsi,168($context) # restore context->Rsi + mov %rdi,176($context) # restore context->Rdi + + mov 40($disp),%rdi # disp->ContextRecord + mov $context,%rsi # context + mov \$154,%ecx # sizeof(CONTEXT) + .long 0xa548f3fc # cld; rep movsq + + mov $disp,%rsi + xor %rcx,%rcx # arg1, UNW_FLAG_NHANDLER + mov 8(%rsi),%rdx # arg2, disp->ImageBase + mov 0(%rsi),%r8 # arg3, disp->ControlPc + mov 16(%rsi),%r9 # arg4, disp->FunctionEntry + mov 40(%rsi),%r10 # disp->ContextRecord + lea 56(%rsi),%r11 # &disp->HandlerData + lea 24(%rsi),%r12 # &disp->EstablisherFrame + mov %r10,32(%rsp) # arg5 + mov %r11,40(%rsp) # arg6 + mov %r12,48(%rsp) # arg7 + mov %rcx,56(%rsp) # arg8, (NULL) + call *__imp_RtlVirtualUnwind(%rip) + + mov \$1,%eax # ExceptionContinueSearch + add \$64,%rsp + popfq + pop %r15 + pop %r14 + pop %r13 + pop %r12 + pop %rbp + pop %rbx + pop %rdi + pop %rsi + ret +.size avx_handler,.-avx_handler + +.section .pdata +.align 4 + .rva .LSEH_begin_poly1305_init_x86_64 + .rva .LSEH_end_poly1305_init_x86_64 + .rva .LSEH_info_poly1305_init_x86_64 + + .rva .LSEH_begin_poly1305_blocks_x86_64 + .rva .LSEH_end_poly1305_blocks_x86_64 + .rva .LSEH_info_poly1305_blocks_x86_64 + + .rva .LSEH_begin_poly1305_emit_x86_64 + .rva .LSEH_end_poly1305_emit_x86_64 + .rva .LSEH_info_poly1305_emit_x86_64 +___ +$code.=<<___ if ($avx); + .rva .LSEH_begin_poly1305_blocks_avx + .rva .Lbase2_64_avx + .rva .LSEH_info_poly1305_blocks_avx_1 + + .rva .Lbase2_64_avx + .rva .Leven_avx + .rva .LSEH_info_poly1305_blocks_avx_2 + + .rva .Leven_avx + .rva .LSEH_end_poly1305_blocks_avx + .rva .LSEH_info_poly1305_blocks_avx_3 + + .rva .LSEH_begin_poly1305_emit_avx + .rva .LSEH_end_poly1305_emit_avx + .rva .LSEH_info_poly1305_emit_avx +___ +$code.=<<___ if ($avx>1); + .rva .LSEH_begin_poly1305_blocks_avx2 + .rva .Lbase2_64_avx2 + .rva .LSEH_info_poly1305_blocks_avx2_1 + + .rva .Lbase2_64_avx2 + .rva .Leven_avx2 + .rva .LSEH_info_poly1305_blocks_avx2_2 + + .rva .Leven_avx2 + .rva .LSEH_end_poly1305_blocks_avx2 + .rva .LSEH_info_poly1305_blocks_avx2_3 +___ +$code.=<<___ if ($avx>2); + .rva .LSEH_begin_poly1305_blocks_avx512 + .rva .LSEH_end_poly1305_blocks_avx512 + .rva .LSEH_info_poly1305_blocks_avx512 +___ +$code.=<<___; +.section .xdata +.align 8 +.LSEH_info_poly1305_init_x86_64: + .byte 9,0,0,0 + .rva se_handler + .rva .LSEH_begin_poly1305_init_x86_64,.LSEH_begin_poly1305_init_x86_64 + +.LSEH_info_poly1305_blocks_x86_64: + .byte 9,0,0,0 + .rva se_handler + .rva .Lblocks_body,.Lblocks_epilogue + +.LSEH_info_poly1305_emit_x86_64: + .byte 9,0,0,0 + .rva se_handler + .rva .LSEH_begin_poly1305_emit_x86_64,.LSEH_begin_poly1305_emit_x86_64 +___ +$code.=<<___ if ($avx); +.LSEH_info_poly1305_blocks_avx_1: + .byte 9,0,0,0 + .rva se_handler + .rva .Lblocks_avx_body,.Lblocks_avx_epilogue # HandlerData[] + +.LSEH_info_poly1305_blocks_avx_2: + .byte 9,0,0,0 + .rva se_handler + .rva .Lbase2_64_avx_body,.Lbase2_64_avx_epilogue # HandlerData[] + +.LSEH_info_poly1305_blocks_avx_3: + .byte 9,0,0,0 + .rva avx_handler + .rva .Ldo_avx_body,.Ldo_avx_epilogue # HandlerData[] + +.LSEH_info_poly1305_emit_avx: + .byte 9,0,0,0 + .rva se_handler + .rva .LSEH_begin_poly1305_emit_avx,.LSEH_begin_poly1305_emit_avx +___ +$code.=<<___ if ($avx>1); +.LSEH_info_poly1305_blocks_avx2_1: + .byte 9,0,0,0 + .rva se_handler + .rva .Lblocks_avx2_body,.Lblocks_avx2_epilogue # HandlerData[] + +.LSEH_info_poly1305_blocks_avx2_2: + .byte 9,0,0,0 + .rva se_handler + .rva .Lbase2_64_avx2_body,.Lbase2_64_avx2_epilogue # HandlerData[] + +.LSEH_info_poly1305_blocks_avx2_3: + .byte 9,0,0,0 + .rva avx_handler + .rva .Ldo_avx2_body,.Ldo_avx2_epilogue # HandlerData[] +___ +$code.=<<___ if ($avx>2); +.LSEH_info_poly1305_blocks_avx512: + .byte 9,0,0,0 + .rva avx_handler + .rva .Ldo_avx512_body,.Ldo_avx512_epilogue # HandlerData[] +___ +} + +open SELF,$0; +while(<SELF>) { + next if (/^#!/); + last if (!s/^#/\/\// and !/^$/); + print; +} +close SELF; + +foreach (split('\n',$code)) { + s/\`([^\`]*)\`/eval($1)/ge; + s/%r([a-z]+)#d/%e$1/g; + s/%r([0-9]+)#d/%r$1d/g; + s/%x#%[yz]/%x/g or s/%y#%z/%y/g or s/%z#%[yz]/%z/g; + + if ($kernel) { + s/(^\.type.*),[0-9]+$/\1/; + s/(^\.type.*),\@abi-omnipotent+$/\1,\@function/; + next if /^\.cfi.*/; + } + + print $_,"\n"; +} +close STDOUT; diff --git a/arch/x86/crypto/poly1305_glue.c b/arch/x86/crypto/poly1305_glue.c index 0cc4537e6617..79bb58737d52 100644 --- a/arch/x86/crypto/poly1305_glue.c +++ b/arch/x86/crypto/poly1305_glue.c @@ -1,8 +1,6 @@ -// SPDX-License-Identifier: GPL-2.0-or-later +// SPDX-License-Identifier: GPL-2.0 OR MIT /* - * Poly1305 authenticator algorithm, RFC7539, SIMD glue code - * - * Copyright (C) 2015 Martin Willi + * Copyright (C) 2015-2019 Jason A. Donenfeld <Jason@zx2c4.com>. All Rights Reserved. */ #include <crypto/algapi.h> @@ -13,108 +11,166 @@ #include <linux/jump_label.h> #include <linux/kernel.h> #include <linux/module.h> +#include <asm/intel-family.h> #include <asm/simd.h> -asmlinkage void poly1305_block_sse2(u32 *h, const u8 *src, - const u32 *r, unsigned int blocks); -asmlinkage void poly1305_2block_sse2(u32 *h, const u8 *src, const u32 *r, - unsigned int blocks, const u32 *u); -asmlinkage void poly1305_4block_avx2(u32 *h, const u8 *src, const u32 *r, - unsigned int blocks, const u32 *u); - -static __ro_after_init DEFINE_STATIC_KEY_FALSE(poly1305_use_simd); +asmlinkage void poly1305_init_x86_64(void *ctx, + const u8 key[POLY1305_KEY_SIZE]); +asmlinkage void poly1305_blocks_x86_64(void *ctx, const u8 *inp, + const size_t len, const u32 padbit); +asmlinkage void poly1305_emit_x86_64(void *ctx, u8 mac[POLY1305_DIGEST_SIZE], + const u32 nonce[4]); +asmlinkage void poly1305_emit_avx(void *ctx, u8 mac[POLY1305_DIGEST_SIZE], + const u32 nonce[4]); +asmlinkage void poly1305_blocks_avx(void *ctx, const u8 *inp, const size_t len, + const u32 padbit); +asmlinkage void poly1305_blocks_avx2(void *ctx, const u8 *inp, const size_t len, + const u32 padbit); +asmlinkage void poly1305_blocks_avx512(void *ctx, const u8 *inp, + const size_t len, const u32 padbit); + +static __ro_after_init DEFINE_STATIC_KEY_FALSE(poly1305_use_avx); static __ro_after_init DEFINE_STATIC_KEY_FALSE(poly1305_use_avx2); +static __ro_after_init DEFINE_STATIC_KEY_FALSE(poly1305_use_avx512); + +struct poly1305_arch_internal { + union { + struct { + u32 h[5]; + u32 is_base2_26; + }; + u64 hs[3]; + }; + u64 r[2]; + u64 pad; + struct { u32 r2, r1, r4, r3; } rn[9]; +}; -static void poly1305_simd_mult(u32 *a, const u32 *b) +/* The AVX code uses base 2^26, while the scalar code uses base 2^64. If we hit + * the unfortunate situation of using AVX and then having to go back to scalar + * -- because the user is silly and has called the update function from two + * separate contexts -- then we need to convert back to the original base before + * proceeding. It is possible to reason that the initial reduction below is + * sufficient given the implementation invariants. However, for an avoidance of + * doubt and because this is not performance critical, we do the full reduction + * anyway. Z3 proof of below function: https://xn--4db.cc/ltPtHCKN/py + */ +static void convert_to_base2_64(void *ctx) { - u8 m[POLY1305_BLOCK_SIZE]; - - memset(m, 0, sizeof(m)); - /* The poly1305 block function adds a hi-bit to the accumulator which - * we don't need for key multiplication; compensate for it. */ - a[4] -= 1 << 24; - poly1305_block_sse2(a, m, b, 1); + struct poly1305_arch_internal *state = ctx; + u32 cy; + + if (!state->is_base2_26) + return; + + cy = state->h[0] >> 26; state->h[0] &= 0x3ffffff; state->h[1] += cy; + cy = state->h[1] >> 26; state->h[1] &= 0x3ffffff; state->h[2] += cy; + cy = state->h[2] >> 26; state->h[2] &= 0x3ffffff; state->h[3] += cy; + cy = state->h[3] >> 26; state->h[3] &= 0x3ffffff; state->h[4] += cy; + state->hs[0] = ((u64)state->h[2] << 52) | ((u64)state->h[1] << 26) | state->h[0]; + state->hs[1] = ((u64)state->h[4] << 40) | ((u64)state->h[3] << 14) | (state->h[2] >> 12); + state->hs[2] = state->h[4] >> 24; +#define ULT(a, b) ((a ^ ((a ^ b) | ((a - b) ^ b))) >> (sizeof(a) * 8 - 1)) + cy = (state->hs[2] >> 2) + (state->hs[2] & ~3ULL); + state->hs[2] &= 3; + state->hs[0] += cy; + state->hs[1] += (cy = ULT(state->hs[0], cy)); + state->hs[2] += ULT(state->hs[1], cy); +#undef ULT + state->is_base2_26 = 0; } -static unsigned int poly1305_scalar_blocks(struct poly1305_desc_ctx *dctx, - const u8 *src, unsigned int srclen) +static void poly1305_simd_init(void *ctx, const u8 key[POLY1305_KEY_SIZE]) { - unsigned int datalen; - - if (unlikely(!dctx->sset)) { - datalen = crypto_poly1305_setdesckey(dctx, src, srclen); - src += srclen - datalen; - srclen = datalen; - } - if (srclen >= POLY1305_BLOCK_SIZE) { - poly1305_core_blocks(&dctx->h, dctx->r, src, - srclen / POLY1305_BLOCK_SIZE, 1); - srclen %= POLY1305_BLOCK_SIZE; - } - return srclen; + poly1305_init_x86_64(ctx, key); } -static unsigned int poly1305_simd_blocks(struct poly1305_desc_ctx *dctx, - const u8 *src, unsigned int srclen) +static void poly1305_simd_blocks(void *ctx, const u8 *inp, size_t len, + const u32 padbit) { - unsigned int blocks, datalen; - - if (unlikely(!dctx->sset)) { - datalen = crypto_poly1305_setdesckey(dctx, src, srclen); - src += srclen - datalen; - srclen = datalen; + struct poly1305_arch_internal *state = ctx; + + /* SIMD disables preemption, so relax after processing each page. */ + BUILD_BUG_ON(PAGE_SIZE < POLY1305_BLOCK_SIZE || + PAGE_SIZE % POLY1305_BLOCK_SIZE); + + if (!IS_ENABLED(CONFIG_AS_AVX) || !static_branch_likely(&poly1305_use_avx) || + (len < (POLY1305_BLOCK_SIZE * 18) && !state->is_base2_26) || + !crypto_simd_usable()) { + convert_to_base2_64(ctx); + poly1305_blocks_x86_64(ctx, inp, len, padbit); + return; } - if (IS_ENABLED(CONFIG_AS_AVX2) && - static_branch_likely(&poly1305_use_avx2) && - srclen >= POLY1305_BLOCK_SIZE * 4) { - if (unlikely(dctx->rset < 4)) { - if (dctx->rset < 2) { - dctx->r[1] = dctx->r[0]; - poly1305_simd_mult(dctx->r[1].r, dctx->r[0].r); - } - dctx->r[2] = dctx->r[1]; - poly1305_simd_mult(dctx->r[2].r, dctx->r[0].r); - dctx->r[3] = dctx->r[2]; - poly1305_simd_mult(dctx->r[3].r, dctx->r[0].r); - dctx->rset = 4; - } - blocks = srclen / (POLY1305_BLOCK_SIZE * 4); - poly1305_4block_avx2(dctx->h.h, src, dctx->r[0].r, blocks, - dctx->r[1].r); - src += POLY1305_BLOCK_SIZE * 4 * blocks; - srclen -= POLY1305_BLOCK_SIZE * 4 * blocks; + for (;;) { + const size_t bytes = min_t(size_t, len, PAGE_SIZE); + + kernel_fpu_begin(); + if (IS_ENABLED(CONFIG_AS_AVX512) && static_branch_likely(&poly1305_use_avx512)) + poly1305_blocks_avx512(ctx, inp, bytes, padbit); + else if (IS_ENABLED(CONFIG_AS_AVX2) && static_branch_likely(&poly1305_use_avx2)) + poly1305_blocks_avx2(ctx, inp, bytes, padbit); + else + poly1305_blocks_avx(ctx, inp, bytes, padbit); + kernel_fpu_end(); + len -= bytes; + if (!len) + break; + inp += bytes; } +} - if (likely(srclen >= POLY1305_BLOCK_SIZE * 2)) { - if (unlikely(dctx->rset < 2)) { - dctx->r[1] = dctx->r[0]; - poly1305_simd_mult(dctx->r[1].r, dctx->r[0].r); - dctx->rset = 2; - } - blocks = srclen / (POLY1305_BLOCK_SIZE * 2); - poly1305_2block_sse2(dctx->h.h, src, dctx->r[0].r, - blocks, dctx->r[1].r); - src += POLY1305_BLOCK_SIZE * 2 * blocks; - srclen -= POLY1305_BLOCK_SIZE * 2 * blocks; - } - if (srclen >= POLY1305_BLOCK_SIZE) { - poly1305_block_sse2(dctx->h.h, src, dctx->r[0].r, 1); - srclen -= POLY1305_BLOCK_SIZE; - } - return srclen; +static void poly1305_simd_emit(void *ctx, u8 mac[POLY1305_DIGEST_SIZE], + const u32 nonce[4]) +{ + if (!IS_ENABLED(CONFIG_AS_AVX) || !static_branch_likely(&poly1305_use_avx)) + poly1305_emit_x86_64(ctx, mac, nonce); + else + poly1305_emit_avx(ctx, mac, nonce); } -void poly1305_init_arch(struct poly1305_desc_ctx *desc, const u8 *key) +void poly1305_init_arch(struct poly1305_desc_ctx *dctx, const u8 *key) { - poly1305_init_generic(desc, key); + poly1305_simd_init(&dctx->h, key); + dctx->s[0] = get_unaligned_le32(&key[16]); + dctx->s[1] = get_unaligned_le32(&key[20]); + dctx->s[2] = get_unaligned_le32(&key[24]); + dctx->s[3] = get_unaligned_le32(&key[28]); + dctx->buflen = 0; + dctx->sset = true; } EXPORT_SYMBOL(poly1305_init_arch); +static unsigned int crypto_poly1305_setdctxkey(struct poly1305_desc_ctx *dctx, + const u8 *inp, unsigned int len) +{ + unsigned int acc = 0; + if (unlikely(!dctx->sset)) { + if (!dctx->rset && len >= POLY1305_BLOCK_SIZE) { + poly1305_simd_init(&dctx->h, inp); + inp += POLY1305_BLOCK_SIZE; + len -= POLY1305_BLOCK_SIZE; + acc += POLY1305_BLOCK_SIZE; + dctx->rset = 1; + } + if (len >= POLY1305_BLOCK_SIZE) { + dctx->s[0] = get_unaligned_le32(&inp[0]); + dctx->s[1] = get_unaligned_le32(&inp[4]); + dctx->s[2] = get_unaligned_le32(&inp[8]); + dctx->s[3] = get_unaligned_le32(&inp[12]); + inp += POLY1305_BLOCK_SIZE; + len -= POLY1305_BLOCK_SIZE; + acc += POLY1305_BLOCK_SIZE; + dctx->sset = true; + } + } + return acc; +} + void poly1305_update_arch(struct poly1305_desc_ctx *dctx, const u8 *src, unsigned int srclen) { - unsigned int bytes; + unsigned int bytes, used; if (unlikely(dctx->buflen)) { bytes = min(srclen, POLY1305_BLOCK_SIZE - dctx->buflen); @@ -124,31 +180,19 @@ void poly1305_update_arch(struct poly1305_desc_ctx *dctx, const u8 *src, dctx->buflen += bytes; if (dctx->buflen == POLY1305_BLOCK_SIZE) { - if (static_branch_likely(&poly1305_use_simd) && - likely(crypto_simd_usable())) { - kernel_fpu_begin(); - poly1305_simd_blocks(dctx, dctx->buf, - POLY1305_BLOCK_SIZE); - kernel_fpu_end(); - } else { - poly1305_scalar_blocks(dctx, dctx->buf, - POLY1305_BLOCK_SIZE); - } + if (likely(!crypto_poly1305_setdctxkey(dctx, dctx->buf, POLY1305_BLOCK_SIZE))) + poly1305_simd_blocks(&dctx->h, dctx->buf, POLY1305_BLOCK_SIZE, 1); dctx->buflen = 0; } } if (likely(srclen >= POLY1305_BLOCK_SIZE)) { - if (static_branch_likely(&poly1305_use_simd) && - likely(crypto_simd_usable())) { - kernel_fpu_begin(); - bytes = poly1305_simd_blocks(dctx, src, srclen); - kernel_fpu_end(); - } else { - bytes = poly1305_scalar_blocks(dctx, src, srclen); - } - src += srclen - bytes; - srclen = bytes; + bytes = round_down(srclen, POLY1305_BLOCK_SIZE); + srclen -= bytes; + used = crypto_poly1305_setdctxkey(dctx, src, bytes); + if (likely(bytes - used)) + poly1305_simd_blocks(&dctx->h, src + used, bytes - used, 1); + src += bytes; } if (unlikely(srclen)) { @@ -158,9 +202,17 @@ void poly1305_update_arch(struct poly1305_desc_ctx *dctx, const u8 *src, } EXPORT_SYMBOL(poly1305_update_arch); -void poly1305_final_arch(struct poly1305_desc_ctx *desc, u8 *digest) +void poly1305_final_arch(struct poly1305_desc_ctx *dctx, u8 *dst) { - poly1305_final_generic(desc, digest); + if (unlikely(dctx->buflen)) { + dctx->buf[dctx->buflen++] = 1; + memset(dctx->buf + dctx->buflen, 0, + POLY1305_BLOCK_SIZE - dctx->buflen); + poly1305_simd_blocks(&dctx->h, dctx->buf, POLY1305_BLOCK_SIZE, 0); + } + + poly1305_simd_emit(&dctx->h, dst, dctx->s); + *dctx = (struct poly1305_desc_ctx){}; } EXPORT_SYMBOL(poly1305_final_arch); @@ -168,38 +220,34 @@ static int crypto_poly1305_init(struct shash_desc *desc) { struct poly1305_desc_ctx *dctx = shash_desc_ctx(desc); - poly1305_core_init(&dctx->h); - dctx->buflen = 0; - dctx->rset = 0; - dctx->sset = false; - + *dctx = (struct poly1305_desc_ctx){}; return 0; } -static int crypto_poly1305_final(struct shash_desc *desc, u8 *dst) +static int crypto_poly1305_update(struct shash_desc *desc, + const u8 *src, unsigned int srclen) { struct poly1305_desc_ctx *dctx = shash_desc_ctx(desc); - if (unlikely(!dctx->sset)) - return -ENOKEY; - - poly1305_final_generic(dctx, dst); + poly1305_update_arch(dctx, src, srclen); return 0; } -static int poly1305_simd_update(struct shash_desc *desc, - const u8 *src, unsigned int srclen) +static int crypto_poly1305_final(struct shash_desc *desc, u8 *dst) { struct poly1305_desc_ctx *dctx = shash_desc_ctx(desc); - poly1305_update_arch(dctx, src, srclen); + if (unlikely(!dctx->sset)) + return -ENOKEY; + + poly1305_final_arch(dctx, dst); return 0; } static struct shash_alg alg = { .digestsize = POLY1305_DIGEST_SIZE, .init = crypto_poly1305_init, - .update = poly1305_simd_update, + .update = crypto_poly1305_update, .final = crypto_poly1305_final, .descsize = sizeof(struct poly1305_desc_ctx), .base = { @@ -213,17 +261,19 @@ static struct shash_alg alg = { static int __init poly1305_simd_mod_init(void) { - if (!boot_cpu_has(X86_FEATURE_XMM2)) - return 0; - - static_branch_enable(&poly1305_use_simd); - - if (IS_ENABLED(CONFIG_AS_AVX2) && - boot_cpu_has(X86_FEATURE_AVX) && + if (IS_ENABLED(CONFIG_AS_AVX) && boot_cpu_has(X86_FEATURE_AVX) && + cpu_has_xfeatures(XFEATURE_MASK_SSE | XFEATURE_MASK_YMM, NULL)) + static_branch_enable(&poly1305_use_avx); + if (IS_ENABLED(CONFIG_AS_AVX2) && boot_cpu_has(X86_FEATURE_AVX) && boot_cpu_has(X86_FEATURE_AVX2) && cpu_has_xfeatures(XFEATURE_MASK_SSE | XFEATURE_MASK_YMM, NULL)) static_branch_enable(&poly1305_use_avx2); - + if (IS_ENABLED(CONFIG_AS_AVX512) && boot_cpu_has(X86_FEATURE_AVX) && + boot_cpu_has(X86_FEATURE_AVX2) && boot_cpu_has(X86_FEATURE_AVX512F) && + cpu_has_xfeatures(XFEATURE_MASK_SSE | XFEATURE_MASK_YMM | XFEATURE_MASK_AVX512, NULL) && + /* Skylake downclocks unacceptably much when using zmm, but later generations are fast. */ + boot_cpu_data.x86_model != INTEL_FAM6_SKYLAKE_X) + static_branch_enable(&poly1305_use_avx512); return IS_REACHABLE(CONFIG_CRYPTO_HASH) ? crypto_register_shash(&alg) : 0; } @@ -237,7 +287,7 @@ module_init(poly1305_simd_mod_init); module_exit(poly1305_simd_mod_exit); MODULE_LICENSE("GPL"); -MODULE_AUTHOR("Martin Willi <martin@strongswan.org>"); +MODULE_AUTHOR("Jason A. Donenfeld <Jason@zx2c4.com>"); MODULE_DESCRIPTION("Poly1305 authenticator"); MODULE_ALIAS_CRYPTO("poly1305"); MODULE_ALIAS_CRYPTO("poly1305-simd"); diff --git a/arch/x86/crypto/serpent_avx2_glue.c b/arch/x86/crypto/serpent_avx2_glue.c index 13fd8d3d2da0..f973ace44ad3 100644 --- a/arch/x86/crypto/serpent_avx2_glue.c +++ b/arch/x86/crypto/serpent_avx2_glue.c @@ -19,18 +19,16 @@ #define SERPENT_AVX2_PARALLEL_BLOCKS 16 /* 16-way AVX2 parallel cipher functions */ -asmlinkage void serpent_ecb_enc_16way(struct serpent_ctx *ctx, u8 *dst, - const u8 *src); -asmlinkage void serpent_ecb_dec_16way(struct serpent_ctx *ctx, u8 *dst, - const u8 *src); -asmlinkage void serpent_cbc_dec_16way(void *ctx, u128 *dst, const u128 *src); +asmlinkage void serpent_ecb_enc_16way(const void *ctx, u8 *dst, const u8 *src); +asmlinkage void serpent_ecb_dec_16way(const void *ctx, u8 *dst, const u8 *src); +asmlinkage void serpent_cbc_dec_16way(const void *ctx, u8 *dst, const u8 *src); -asmlinkage void serpent_ctr_16way(void *ctx, u128 *dst, const u128 *src, +asmlinkage void serpent_ctr_16way(const void *ctx, u8 *dst, const u8 *src, le128 *iv); -asmlinkage void serpent_xts_enc_16way(struct serpent_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); -asmlinkage void serpent_xts_dec_16way(struct serpent_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); +asmlinkage void serpent_xts_enc_16way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); +asmlinkage void serpent_xts_dec_16way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); static int serpent_setkey_skcipher(struct crypto_skcipher *tfm, const u8 *key, unsigned int keylen) @@ -44,13 +42,13 @@ static const struct common_glue_ctx serpent_enc = { .funcs = { { .num_blocks = 16, - .fn_u = { .ecb = GLUE_FUNC_CAST(serpent_ecb_enc_16way) } + .fn_u = { .ecb = serpent_ecb_enc_16way } }, { .num_blocks = 8, - .fn_u = { .ecb = GLUE_FUNC_CAST(serpent_ecb_enc_8way_avx) } + .fn_u = { .ecb = serpent_ecb_enc_8way_avx } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(__serpent_encrypt) } + .fn_u = { .ecb = __serpent_encrypt } } } }; @@ -60,13 +58,13 @@ static const struct common_glue_ctx serpent_ctr = { .funcs = { { .num_blocks = 16, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(serpent_ctr_16way) } + .fn_u = { .ctr = serpent_ctr_16way } }, { .num_blocks = 8, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(serpent_ctr_8way_avx) } + .fn_u = { .ctr = serpent_ctr_8way_avx } }, { .num_blocks = 1, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(__serpent_crypt_ctr) } + .fn_u = { .ctr = __serpent_crypt_ctr } } } }; @@ -76,13 +74,13 @@ static const struct common_glue_ctx serpent_enc_xts = { .funcs = { { .num_blocks = 16, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(serpent_xts_enc_16way) } + .fn_u = { .xts = serpent_xts_enc_16way } }, { .num_blocks = 8, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(serpent_xts_enc_8way_avx) } + .fn_u = { .xts = serpent_xts_enc_8way_avx } }, { .num_blocks = 1, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(serpent_xts_enc) } + .fn_u = { .xts = serpent_xts_enc } } } }; @@ -92,13 +90,13 @@ static const struct common_glue_ctx serpent_dec = { .funcs = { { .num_blocks = 16, - .fn_u = { .ecb = GLUE_FUNC_CAST(serpent_ecb_dec_16way) } + .fn_u = { .ecb = serpent_ecb_dec_16way } }, { .num_blocks = 8, - .fn_u = { .ecb = GLUE_FUNC_CAST(serpent_ecb_dec_8way_avx) } + .fn_u = { .ecb = serpent_ecb_dec_8way_avx } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(__serpent_decrypt) } + .fn_u = { .ecb = __serpent_decrypt } } } }; @@ -108,13 +106,13 @@ static const struct common_glue_ctx serpent_dec_cbc = { .funcs = { { .num_blocks = 16, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(serpent_cbc_dec_16way) } + .fn_u = { .cbc = serpent_cbc_dec_16way } }, { .num_blocks = 8, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(serpent_cbc_dec_8way_avx) } + .fn_u = { .cbc = serpent_cbc_dec_8way_avx } }, { .num_blocks = 1, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(__serpent_decrypt) } + .fn_u = { .cbc = __serpent_decrypt } } } }; @@ -124,13 +122,13 @@ static const struct common_glue_ctx serpent_dec_xts = { .funcs = { { .num_blocks = 16, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(serpent_xts_dec_16way) } + .fn_u = { .xts = serpent_xts_dec_16way } }, { .num_blocks = 8, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(serpent_xts_dec_8way_avx) } + .fn_u = { .xts = serpent_xts_dec_8way_avx } }, { .num_blocks = 1, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(serpent_xts_dec) } + .fn_u = { .xts = serpent_xts_dec } } } }; @@ -146,8 +144,7 @@ static int ecb_decrypt(struct skcipher_request *req) static int cbc_encrypt(struct skcipher_request *req) { - return glue_cbc_encrypt_req_128bit(GLUE_FUNC_CAST(__serpent_encrypt), - req); + return glue_cbc_encrypt_req_128bit(__serpent_encrypt, req); } static int cbc_decrypt(struct skcipher_request *req) @@ -166,8 +163,8 @@ static int xts_encrypt(struct skcipher_request *req) struct serpent_xts_ctx *ctx = crypto_skcipher_ctx(tfm); return glue_xts_req_128bit(&serpent_enc_xts, req, - XTS_TWEAK_CAST(__serpent_encrypt), - &ctx->tweak_ctx, &ctx->crypt_ctx, false); + __serpent_encrypt, &ctx->tweak_ctx, + &ctx->crypt_ctx, false); } static int xts_decrypt(struct skcipher_request *req) @@ -176,8 +173,8 @@ static int xts_decrypt(struct skcipher_request *req) struct serpent_xts_ctx *ctx = crypto_skcipher_ctx(tfm); return glue_xts_req_128bit(&serpent_dec_xts, req, - XTS_TWEAK_CAST(__serpent_encrypt), - &ctx->tweak_ctx, &ctx->crypt_ctx, true); + __serpent_encrypt, &ctx->tweak_ctx, + &ctx->crypt_ctx, true); } static struct skcipher_alg serpent_algs[] = { diff --git a/arch/x86/crypto/serpent_avx_glue.c b/arch/x86/crypto/serpent_avx_glue.c index 7d3dca38a5a2..7806d1cbe854 100644 --- a/arch/x86/crypto/serpent_avx_glue.c +++ b/arch/x86/crypto/serpent_avx_glue.c @@ -20,33 +20,35 @@ #include <asm/crypto/serpent-avx.h> /* 8-way parallel cipher functions */ -asmlinkage void serpent_ecb_enc_8way_avx(struct serpent_ctx *ctx, u8 *dst, +asmlinkage void serpent_ecb_enc_8way_avx(const void *ctx, u8 *dst, const u8 *src); EXPORT_SYMBOL_GPL(serpent_ecb_enc_8way_avx); -asmlinkage void serpent_ecb_dec_8way_avx(struct serpent_ctx *ctx, u8 *dst, +asmlinkage void serpent_ecb_dec_8way_avx(const void *ctx, u8 *dst, const u8 *src); EXPORT_SYMBOL_GPL(serpent_ecb_dec_8way_avx); -asmlinkage void serpent_cbc_dec_8way_avx(struct serpent_ctx *ctx, u8 *dst, +asmlinkage void serpent_cbc_dec_8way_avx(const void *ctx, u8 *dst, const u8 *src); EXPORT_SYMBOL_GPL(serpent_cbc_dec_8way_avx); -asmlinkage void serpent_ctr_8way_avx(struct serpent_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); +asmlinkage void serpent_ctr_8way_avx(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); EXPORT_SYMBOL_GPL(serpent_ctr_8way_avx); -asmlinkage void serpent_xts_enc_8way_avx(struct serpent_ctx *ctx, u8 *dst, +asmlinkage void serpent_xts_enc_8way_avx(const void *ctx, u8 *dst, const u8 *src, le128 *iv); EXPORT_SYMBOL_GPL(serpent_xts_enc_8way_avx); -asmlinkage void serpent_xts_dec_8way_avx(struct serpent_ctx *ctx, u8 *dst, +asmlinkage void serpent_xts_dec_8way_avx(const void *ctx, u8 *dst, const u8 *src, le128 *iv); EXPORT_SYMBOL_GPL(serpent_xts_dec_8way_avx); -void __serpent_crypt_ctr(void *ctx, u128 *dst, const u128 *src, le128 *iv) +void __serpent_crypt_ctr(const void *ctx, u8 *d, const u8 *s, le128 *iv) { be128 ctrblk; + u128 *dst = (u128 *)d; + const u128 *src = (const u128 *)s; le128_to_be128(&ctrblk, iv); le128_inc(iv); @@ -56,17 +58,15 @@ void __serpent_crypt_ctr(void *ctx, u128 *dst, const u128 *src, le128 *iv) } EXPORT_SYMBOL_GPL(__serpent_crypt_ctr); -void serpent_xts_enc(void *ctx, u128 *dst, const u128 *src, le128 *iv) +void serpent_xts_enc(const void *ctx, u8 *dst, const u8 *src, le128 *iv) { - glue_xts_crypt_128bit_one(ctx, dst, src, iv, - GLUE_FUNC_CAST(__serpent_encrypt)); + glue_xts_crypt_128bit_one(ctx, dst, src, iv, __serpent_encrypt); } EXPORT_SYMBOL_GPL(serpent_xts_enc); -void serpent_xts_dec(void *ctx, u128 *dst, const u128 *src, le128 *iv) +void serpent_xts_dec(const void *ctx, u8 *dst, const u8 *src, le128 *iv) { - glue_xts_crypt_128bit_one(ctx, dst, src, iv, - GLUE_FUNC_CAST(__serpent_decrypt)); + glue_xts_crypt_128bit_one(ctx, dst, src, iv, __serpent_decrypt); } EXPORT_SYMBOL_GPL(serpent_xts_dec); @@ -102,10 +102,10 @@ static const struct common_glue_ctx serpent_enc = { .funcs = { { .num_blocks = SERPENT_PARALLEL_BLOCKS, - .fn_u = { .ecb = GLUE_FUNC_CAST(serpent_ecb_enc_8way_avx) } + .fn_u = { .ecb = serpent_ecb_enc_8way_avx } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(__serpent_encrypt) } + .fn_u = { .ecb = __serpent_encrypt } } } }; @@ -115,10 +115,10 @@ static const struct common_glue_ctx serpent_ctr = { .funcs = { { .num_blocks = SERPENT_PARALLEL_BLOCKS, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(serpent_ctr_8way_avx) } + .fn_u = { .ctr = serpent_ctr_8way_avx } }, { .num_blocks = 1, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(__serpent_crypt_ctr) } + .fn_u = { .ctr = __serpent_crypt_ctr } } } }; @@ -128,10 +128,10 @@ static const struct common_glue_ctx serpent_enc_xts = { .funcs = { { .num_blocks = SERPENT_PARALLEL_BLOCKS, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(serpent_xts_enc_8way_avx) } + .fn_u = { .xts = serpent_xts_enc_8way_avx } }, { .num_blocks = 1, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(serpent_xts_enc) } + .fn_u = { .xts = serpent_xts_enc } } } }; @@ -141,10 +141,10 @@ static const struct common_glue_ctx serpent_dec = { .funcs = { { .num_blocks = SERPENT_PARALLEL_BLOCKS, - .fn_u = { .ecb = GLUE_FUNC_CAST(serpent_ecb_dec_8way_avx) } + .fn_u = { .ecb = serpent_ecb_dec_8way_avx } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(__serpent_decrypt) } + .fn_u = { .ecb = __serpent_decrypt } } } }; @@ -154,10 +154,10 @@ static const struct common_glue_ctx serpent_dec_cbc = { .funcs = { { .num_blocks = SERPENT_PARALLEL_BLOCKS, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(serpent_cbc_dec_8way_avx) } + .fn_u = { .cbc = serpent_cbc_dec_8way_avx } }, { .num_blocks = 1, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(__serpent_decrypt) } + .fn_u = { .cbc = __serpent_decrypt } } } }; @@ -167,10 +167,10 @@ static const struct common_glue_ctx serpent_dec_xts = { .funcs = { { .num_blocks = SERPENT_PARALLEL_BLOCKS, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(serpent_xts_dec_8way_avx) } + .fn_u = { .xts = serpent_xts_dec_8way_avx } }, { .num_blocks = 1, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(serpent_xts_dec) } + .fn_u = { .xts = serpent_xts_dec } } } }; @@ -186,8 +186,7 @@ static int ecb_decrypt(struct skcipher_request *req) static int cbc_encrypt(struct skcipher_request *req) { - return glue_cbc_encrypt_req_128bit(GLUE_FUNC_CAST(__serpent_encrypt), - req); + return glue_cbc_encrypt_req_128bit(__serpent_encrypt, req); } static int cbc_decrypt(struct skcipher_request *req) @@ -206,8 +205,8 @@ static int xts_encrypt(struct skcipher_request *req) struct serpent_xts_ctx *ctx = crypto_skcipher_ctx(tfm); return glue_xts_req_128bit(&serpent_enc_xts, req, - XTS_TWEAK_CAST(__serpent_encrypt), - &ctx->tweak_ctx, &ctx->crypt_ctx, false); + __serpent_encrypt, &ctx->tweak_ctx, + &ctx->crypt_ctx, false); } static int xts_decrypt(struct skcipher_request *req) @@ -216,8 +215,8 @@ static int xts_decrypt(struct skcipher_request *req) struct serpent_xts_ctx *ctx = crypto_skcipher_ctx(tfm); return glue_xts_req_128bit(&serpent_dec_xts, req, - XTS_TWEAK_CAST(__serpent_encrypt), - &ctx->tweak_ctx, &ctx->crypt_ctx, true); + __serpent_encrypt, &ctx->tweak_ctx, + &ctx->crypt_ctx, true); } static struct skcipher_alg serpent_algs[] = { diff --git a/arch/x86/crypto/serpent_sse2_glue.c b/arch/x86/crypto/serpent_sse2_glue.c index 5fdf1931d069..4fed8d26b91a 100644 --- a/arch/x86/crypto/serpent_sse2_glue.c +++ b/arch/x86/crypto/serpent_sse2_glue.c @@ -31,9 +31,11 @@ static int serpent_setkey_skcipher(struct crypto_skcipher *tfm, return __serpent_setkey(crypto_skcipher_ctx(tfm), key, keylen); } -static void serpent_decrypt_cbc_xway(void *ctx, u128 *dst, const u128 *src) +static void serpent_decrypt_cbc_xway(const void *ctx, u8 *d, const u8 *s) { u128 ivs[SERPENT_PARALLEL_BLOCKS - 1]; + u128 *dst = (u128 *)d; + const u128 *src = (const u128 *)s; unsigned int j; for (j = 0; j < SERPENT_PARALLEL_BLOCKS - 1; j++) @@ -45,9 +47,11 @@ static void serpent_decrypt_cbc_xway(void *ctx, u128 *dst, const u128 *src) u128_xor(dst + (j + 1), dst + (j + 1), ivs + j); } -static void serpent_crypt_ctr(void *ctx, u128 *dst, const u128 *src, le128 *iv) +static void serpent_crypt_ctr(const void *ctx, u8 *d, const u8 *s, le128 *iv) { be128 ctrblk; + u128 *dst = (u128 *)d; + const u128 *src = (const u128 *)s; le128_to_be128(&ctrblk, iv); le128_inc(iv); @@ -56,10 +60,12 @@ static void serpent_crypt_ctr(void *ctx, u128 *dst, const u128 *src, le128 *iv) u128_xor(dst, src, (u128 *)&ctrblk); } -static void serpent_crypt_ctr_xway(void *ctx, u128 *dst, const u128 *src, +static void serpent_crypt_ctr_xway(const void *ctx, u8 *d, const u8 *s, le128 *iv) { be128 ctrblks[SERPENT_PARALLEL_BLOCKS]; + u128 *dst = (u128 *)d; + const u128 *src = (const u128 *)s; unsigned int i; for (i = 0; i < SERPENT_PARALLEL_BLOCKS; i++) { @@ -79,10 +85,10 @@ static const struct common_glue_ctx serpent_enc = { .funcs = { { .num_blocks = SERPENT_PARALLEL_BLOCKS, - .fn_u = { .ecb = GLUE_FUNC_CAST(serpent_enc_blk_xway) } + .fn_u = { .ecb = serpent_enc_blk_xway } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(__serpent_encrypt) } + .fn_u = { .ecb = __serpent_encrypt } } } }; @@ -92,10 +98,10 @@ static const struct common_glue_ctx serpent_ctr = { .funcs = { { .num_blocks = SERPENT_PARALLEL_BLOCKS, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(serpent_crypt_ctr_xway) } + .fn_u = { .ctr = serpent_crypt_ctr_xway } }, { .num_blocks = 1, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(serpent_crypt_ctr) } + .fn_u = { .ctr = serpent_crypt_ctr } } } }; @@ -105,10 +111,10 @@ static const struct common_glue_ctx serpent_dec = { .funcs = { { .num_blocks = SERPENT_PARALLEL_BLOCKS, - .fn_u = { .ecb = GLUE_FUNC_CAST(serpent_dec_blk_xway) } + .fn_u = { .ecb = serpent_dec_blk_xway } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(__serpent_decrypt) } + .fn_u = { .ecb = __serpent_decrypt } } } }; @@ -118,10 +124,10 @@ static const struct common_glue_ctx serpent_dec_cbc = { .funcs = { { .num_blocks = SERPENT_PARALLEL_BLOCKS, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(serpent_decrypt_cbc_xway) } + .fn_u = { .cbc = serpent_decrypt_cbc_xway } }, { .num_blocks = 1, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(__serpent_decrypt) } + .fn_u = { .cbc = __serpent_decrypt } } } }; @@ -137,7 +143,7 @@ static int ecb_decrypt(struct skcipher_request *req) static int cbc_encrypt(struct skcipher_request *req) { - return glue_cbc_encrypt_req_128bit(GLUE_FUNC_CAST(__serpent_encrypt), + return glue_cbc_encrypt_req_128bit(__serpent_encrypt, req); } diff --git a/arch/x86/crypto/sha1_avx2_x86_64_asm.S b/arch/x86/crypto/sha1_avx2_x86_64_asm.S index 6decc85ef7b7..1e594d60afa5 100644 --- a/arch/x86/crypto/sha1_avx2_x86_64_asm.S +++ b/arch/x86/crypto/sha1_avx2_x86_64_asm.S @@ -62,11 +62,11 @@ *Visit http://software.intel.com/en-us/articles/ *and refer to improving-the-performance-of-the-secure-hash-algorithm-1/ * - *Updates 20-byte SHA-1 record in 'hash' for even number of - *'num_blocks' consecutive 64-byte blocks + *Updates 20-byte SHA-1 record at start of 'state', from 'input', for + *even number of 'blocks' consecutive 64-byte blocks. * *extern "C" void sha1_transform_avx2( - * int *hash, const char* input, size_t num_blocks ); + * struct sha1_state *state, const u8* input, int blocks ); */ #include <linux/linkage.h> diff --git a/arch/x86/crypto/sha1_ssse3_asm.S b/arch/x86/crypto/sha1_ssse3_asm.S index 5d03c1173690..12e2d19d7402 100644 --- a/arch/x86/crypto/sha1_ssse3_asm.S +++ b/arch/x86/crypto/sha1_ssse3_asm.S @@ -457,9 +457,13 @@ W_PRECALC_SSSE3 movdqu \a,\b .endm -/* SSSE3 optimized implementation: - * extern "C" void sha1_transform_ssse3(u32 *digest, const char *data, u32 *ws, - * unsigned int rounds); +/* + * SSSE3 optimized implementation: + * + * extern "C" void sha1_transform_ssse3(struct sha1_state *state, + * const u8 *data, int blocks); + * + * Note that struct sha1_state is assumed to begin with u32 state[5]. */ SHA1_VECTOR_ASM sha1_transform_ssse3 @@ -545,8 +549,8 @@ W_PRECALC_AVX /* AVX optimized implementation: - * extern "C" void sha1_transform_avx(u32 *digest, const char *data, u32 *ws, - * unsigned int rounds); + * extern "C" void sha1_transform_avx(struct sha1_state *state, + * const u8 *data, int blocks); */ SHA1_VECTOR_ASM sha1_transform_avx diff --git a/arch/x86/crypto/sha1_ssse3_glue.c b/arch/x86/crypto/sha1_ssse3_glue.c index 639d4c2fd6a8..d70b40ad594c 100644 --- a/arch/x86/crypto/sha1_ssse3_glue.c +++ b/arch/x86/crypto/sha1_ssse3_glue.c @@ -27,11 +27,8 @@ #include <crypto/sha1_base.h> #include <asm/simd.h> -typedef void (sha1_transform_fn)(u32 *digest, const char *data, - unsigned int rounds); - static int sha1_update(struct shash_desc *desc, const u8 *data, - unsigned int len, sha1_transform_fn *sha1_xform) + unsigned int len, sha1_block_fn *sha1_xform) { struct sha1_state *sctx = shash_desc_ctx(desc); @@ -39,48 +36,47 @@ static int sha1_update(struct shash_desc *desc, const u8 *data, (sctx->count % SHA1_BLOCK_SIZE) + len < SHA1_BLOCK_SIZE) return crypto_sha1_update(desc, data, len); - /* make sure casting to sha1_block_fn() is safe */ + /* + * Make sure struct sha1_state begins directly with the SHA1 + * 160-bit internal state, as this is what the asm functions expect. + */ BUILD_BUG_ON(offsetof(struct sha1_state, state) != 0); kernel_fpu_begin(); - sha1_base_do_update(desc, data, len, - (sha1_block_fn *)sha1_xform); + sha1_base_do_update(desc, data, len, sha1_xform); kernel_fpu_end(); return 0; } static int sha1_finup(struct shash_desc *desc, const u8 *data, - unsigned int len, u8 *out, sha1_transform_fn *sha1_xform) + unsigned int len, u8 *out, sha1_block_fn *sha1_xform) { if (!crypto_simd_usable()) return crypto_sha1_finup(desc, data, len, out); kernel_fpu_begin(); if (len) - sha1_base_do_update(desc, data, len, - (sha1_block_fn *)sha1_xform); - sha1_base_do_finalize(desc, (sha1_block_fn *)sha1_xform); + sha1_base_do_update(desc, data, len, sha1_xform); + sha1_base_do_finalize(desc, sha1_xform); kernel_fpu_end(); return sha1_base_finish(desc, out); } -asmlinkage void sha1_transform_ssse3(u32 *digest, const char *data, - unsigned int rounds); +asmlinkage void sha1_transform_ssse3(struct sha1_state *state, + const u8 *data, int blocks); static int sha1_ssse3_update(struct shash_desc *desc, const u8 *data, unsigned int len) { - return sha1_update(desc, data, len, - (sha1_transform_fn *) sha1_transform_ssse3); + return sha1_update(desc, data, len, sha1_transform_ssse3); } static int sha1_ssse3_finup(struct shash_desc *desc, const u8 *data, unsigned int len, u8 *out) { - return sha1_finup(desc, data, len, out, - (sha1_transform_fn *) sha1_transform_ssse3); + return sha1_finup(desc, data, len, out, sha1_transform_ssse3); } /* Add padding and return the message digest. */ @@ -119,21 +115,19 @@ static void unregister_sha1_ssse3(void) } #ifdef CONFIG_AS_AVX -asmlinkage void sha1_transform_avx(u32 *digest, const char *data, - unsigned int rounds); +asmlinkage void sha1_transform_avx(struct sha1_state *state, + const u8 *data, int blocks); static int sha1_avx_update(struct shash_desc *desc, const u8 *data, unsigned int len) { - return sha1_update(desc, data, len, - (sha1_transform_fn *) sha1_transform_avx); + return sha1_update(desc, data, len, sha1_transform_avx); } static int sha1_avx_finup(struct shash_desc *desc, const u8 *data, unsigned int len, u8 *out) { - return sha1_finup(desc, data, len, out, - (sha1_transform_fn *) sha1_transform_avx); + return sha1_finup(desc, data, len, out, sha1_transform_avx); } static int sha1_avx_final(struct shash_desc *desc, u8 *out) @@ -190,8 +184,8 @@ static inline void unregister_sha1_avx(void) { } #if defined(CONFIG_AS_AVX2) && (CONFIG_AS_AVX) #define SHA1_AVX2_BLOCK_OPTSIZE 4 /* optimal 4*64 bytes of SHA1 blocks */ -asmlinkage void sha1_transform_avx2(u32 *digest, const char *data, - unsigned int rounds); +asmlinkage void sha1_transform_avx2(struct sha1_state *state, + const u8 *data, int blocks); static bool avx2_usable(void) { @@ -203,28 +197,26 @@ static bool avx2_usable(void) return false; } -static void sha1_apply_transform_avx2(u32 *digest, const char *data, - unsigned int rounds) +static void sha1_apply_transform_avx2(struct sha1_state *state, + const u8 *data, int blocks) { /* Select the optimal transform based on data block size */ - if (rounds >= SHA1_AVX2_BLOCK_OPTSIZE) - sha1_transform_avx2(digest, data, rounds); + if (blocks >= SHA1_AVX2_BLOCK_OPTSIZE) + sha1_transform_avx2(state, data, blocks); else - sha1_transform_avx(digest, data, rounds); + sha1_transform_avx(state, data, blocks); } static int sha1_avx2_update(struct shash_desc *desc, const u8 *data, unsigned int len) { - return sha1_update(desc, data, len, - (sha1_transform_fn *) sha1_apply_transform_avx2); + return sha1_update(desc, data, len, sha1_apply_transform_avx2); } static int sha1_avx2_finup(struct shash_desc *desc, const u8 *data, unsigned int len, u8 *out) { - return sha1_finup(desc, data, len, out, - (sha1_transform_fn *) sha1_apply_transform_avx2); + return sha1_finup(desc, data, len, out, sha1_apply_transform_avx2); } static int sha1_avx2_final(struct shash_desc *desc, u8 *out) @@ -267,21 +259,19 @@ static inline void unregister_sha1_avx2(void) { } #endif #ifdef CONFIG_AS_SHA1_NI -asmlinkage void sha1_ni_transform(u32 *digest, const char *data, - unsigned int rounds); +asmlinkage void sha1_ni_transform(struct sha1_state *digest, const u8 *data, + int rounds); static int sha1_ni_update(struct shash_desc *desc, const u8 *data, unsigned int len) { - return sha1_update(desc, data, len, - (sha1_transform_fn *) sha1_ni_transform); + return sha1_update(desc, data, len, sha1_ni_transform); } static int sha1_ni_finup(struct shash_desc *desc, const u8 *data, unsigned int len, u8 *out) { - return sha1_finup(desc, data, len, out, - (sha1_transform_fn *) sha1_ni_transform); + return sha1_finup(desc, data, len, out, sha1_ni_transform); } static int sha1_ni_final(struct shash_desc *desc, u8 *out) diff --git a/arch/x86/crypto/sha256-avx-asm.S b/arch/x86/crypto/sha256-avx-asm.S index 22e14c8dd2e4..fcbc30f58c38 100644 --- a/arch/x86/crypto/sha256-avx-asm.S +++ b/arch/x86/crypto/sha256-avx-asm.S @@ -341,8 +341,8 @@ a = TMP_ .endm ######################################################################## -## void sha256_transform_avx(void *input_data, UINT32 digest[8], UINT64 num_blks) -## arg 1 : pointer to digest +## void sha256_transform_avx(state sha256_state *state, const u8 *data, int blocks) +## arg 1 : pointer to state ## arg 2 : pointer to input data ## arg 3 : Num blocks ######################################################################## diff --git a/arch/x86/crypto/sha256-avx2-asm.S b/arch/x86/crypto/sha256-avx2-asm.S index 519b551ad576..499d9ec129de 100644 --- a/arch/x86/crypto/sha256-avx2-asm.S +++ b/arch/x86/crypto/sha256-avx2-asm.S @@ -520,8 +520,8 @@ STACK_SIZE = _RSP + _RSP_SIZE .endm ######################################################################## -## void sha256_transform_rorx(void *input_data, UINT32 digest[8], UINT64 num_blks) -## arg 1 : pointer to digest +## void sha256_transform_rorx(struct sha256_state *state, const u8 *data, int blocks) +## arg 1 : pointer to state ## arg 2 : pointer to input data ## arg 3 : Num blocks ######################################################################## diff --git a/arch/x86/crypto/sha256-ssse3-asm.S b/arch/x86/crypto/sha256-ssse3-asm.S index 69cc2f91dc4c..ddfa863b4ee3 100644 --- a/arch/x86/crypto/sha256-ssse3-asm.S +++ b/arch/x86/crypto/sha256-ssse3-asm.S @@ -347,8 +347,10 @@ a = TMP_ .endm ######################################################################## -## void sha256_transform_ssse3(void *input_data, UINT32 digest[8], UINT64 num_blks) -## arg 1 : pointer to digest +## void sha256_transform_ssse3(struct sha256_state *state, const u8 *data, +## int blocks); +## arg 1 : pointer to state +## (struct sha256_state is assumed to begin with u32 state[8]) ## arg 2 : pointer to input data ## arg 3 : Num blocks ######################################################################## diff --git a/arch/x86/crypto/sha256_ssse3_glue.c b/arch/x86/crypto/sha256_ssse3_glue.c index f9aff31fe59e..03ad657c04bd 100644 --- a/arch/x86/crypto/sha256_ssse3_glue.c +++ b/arch/x86/crypto/sha256_ssse3_glue.c @@ -41,12 +41,11 @@ #include <linux/string.h> #include <asm/simd.h> -asmlinkage void sha256_transform_ssse3(u32 *digest, const char *data, - u64 rounds); -typedef void (sha256_transform_fn)(u32 *digest, const char *data, u64 rounds); +asmlinkage void sha256_transform_ssse3(struct sha256_state *state, + const u8 *data, int blocks); static int _sha256_update(struct shash_desc *desc, const u8 *data, - unsigned int len, sha256_transform_fn *sha256_xform) + unsigned int len, sha256_block_fn *sha256_xform) { struct sha256_state *sctx = shash_desc_ctx(desc); @@ -54,28 +53,29 @@ static int _sha256_update(struct shash_desc *desc, const u8 *data, (sctx->count % SHA256_BLOCK_SIZE) + len < SHA256_BLOCK_SIZE) return crypto_sha256_update(desc, data, len); - /* make sure casting to sha256_block_fn() is safe */ + /* + * Make sure struct sha256_state begins directly with the SHA256 + * 256-bit internal state, as this is what the asm functions expect. + */ BUILD_BUG_ON(offsetof(struct sha256_state, state) != 0); kernel_fpu_begin(); - sha256_base_do_update(desc, data, len, - (sha256_block_fn *)sha256_xform); + sha256_base_do_update(desc, data, len, sha256_xform); kernel_fpu_end(); return 0; } static int sha256_finup(struct shash_desc *desc, const u8 *data, - unsigned int len, u8 *out, sha256_transform_fn *sha256_xform) + unsigned int len, u8 *out, sha256_block_fn *sha256_xform) { if (!crypto_simd_usable()) return crypto_sha256_finup(desc, data, len, out); kernel_fpu_begin(); if (len) - sha256_base_do_update(desc, data, len, - (sha256_block_fn *)sha256_xform); - sha256_base_do_finalize(desc, (sha256_block_fn *)sha256_xform); + sha256_base_do_update(desc, data, len, sha256_xform); + sha256_base_do_finalize(desc, sha256_xform); kernel_fpu_end(); return sha256_base_finish(desc, out); @@ -145,8 +145,8 @@ static void unregister_sha256_ssse3(void) } #ifdef CONFIG_AS_AVX -asmlinkage void sha256_transform_avx(u32 *digest, const char *data, - u64 rounds); +asmlinkage void sha256_transform_avx(struct sha256_state *state, + const u8 *data, int blocks); static int sha256_avx_update(struct shash_desc *desc, const u8 *data, unsigned int len) @@ -227,8 +227,8 @@ static inline void unregister_sha256_avx(void) { } #endif #if defined(CONFIG_AS_AVX2) && defined(CONFIG_AS_AVX) -asmlinkage void sha256_transform_rorx(u32 *digest, const char *data, - u64 rounds); +asmlinkage void sha256_transform_rorx(struct sha256_state *state, + const u8 *data, int blocks); static int sha256_avx2_update(struct shash_desc *desc, const u8 *data, unsigned int len) @@ -307,8 +307,8 @@ static inline void unregister_sha256_avx2(void) { } #endif #ifdef CONFIG_AS_SHA256_NI -asmlinkage void sha256_ni_transform(u32 *digest, const char *data, - u64 rounds); /*unsigned int rounds);*/ +asmlinkage void sha256_ni_transform(struct sha256_state *digest, + const u8 *data, int rounds); static int sha256_ni_update(struct shash_desc *desc, const u8 *data, unsigned int len) diff --git a/arch/x86/crypto/sha512-avx-asm.S b/arch/x86/crypto/sha512-avx-asm.S index 3704ddd7e5d5..90ea945ba5e6 100644 --- a/arch/x86/crypto/sha512-avx-asm.S +++ b/arch/x86/crypto/sha512-avx-asm.S @@ -271,11 +271,12 @@ frame_size = frame_GPRSAVE + GPRSAVE_SIZE .endm ######################################################################## -# void sha512_transform_avx(void* D, const void* M, u64 L) -# Purpose: Updates the SHA512 digest stored at D with the message stored in M. -# The size of the message pointed to by M must be an integer multiple of SHA512 -# message blocks. -# L is the message length in SHA512 blocks +# void sha512_transform_avx(sha512_state *state, const u8 *data, int blocks) +# Purpose: Updates the SHA512 digest stored at "state" with the message +# stored in "data". +# The size of the message pointed to by "data" must be an integer multiple +# of SHA512 message blocks. +# "blocks" is the message length in SHA512 blocks ######################################################################## SYM_FUNC_START(sha512_transform_avx) cmp $0, msglen diff --git a/arch/x86/crypto/sha512-avx2-asm.S b/arch/x86/crypto/sha512-avx2-asm.S index 80d830e7ee09..3dd886b14e7d 100644 --- a/arch/x86/crypto/sha512-avx2-asm.S +++ b/arch/x86/crypto/sha512-avx2-asm.S @@ -563,11 +563,12 @@ frame_size = frame_GPRSAVE + GPRSAVE_SIZE .endm ######################################################################## -# void sha512_transform_rorx(void* D, const void* M, uint64_t L)# -# Purpose: Updates the SHA512 digest stored at D with the message stored in M. -# The size of the message pointed to by M must be an integer multiple of SHA512 -# message blocks. -# L is the message length in SHA512 blocks +# void sha512_transform_rorx(sha512_state *state, const u8 *data, int blocks) +# Purpose: Updates the SHA512 digest stored at "state" with the message +# stored in "data". +# The size of the message pointed to by "data" must be an integer multiple +# of SHA512 message blocks. +# "blocks" is the message length in SHA512 blocks ######################################################################## SYM_FUNC_START(sha512_transform_rorx) # Allocate Stack Space diff --git a/arch/x86/crypto/sha512-ssse3-asm.S b/arch/x86/crypto/sha512-ssse3-asm.S index 838f984e95d9..7946a1bee85b 100644 --- a/arch/x86/crypto/sha512-ssse3-asm.S +++ b/arch/x86/crypto/sha512-ssse3-asm.S @@ -269,11 +269,14 @@ frame_size = frame_GPRSAVE + GPRSAVE_SIZE .endm ######################################################################## -# void sha512_transform_ssse3(void* D, const void* M, u64 L)# -# Purpose: Updates the SHA512 digest stored at D with the message stored in M. -# The size of the message pointed to by M must be an integer multiple of SHA512 -# message blocks. -# L is the message length in SHA512 blocks. +## void sha512_transform_ssse3(struct sha512_state *state, const u8 *data, +## int blocks); +# (struct sha512_state is assumed to begin with u64 state[8]) +# Purpose: Updates the SHA512 digest stored at "state" with the message +# stored in "data". +# The size of the message pointed to by "data" must be an integer multiple +# of SHA512 message blocks. +# "blocks" is the message length in SHA512 blocks. ######################################################################## SYM_FUNC_START(sha512_transform_ssse3) diff --git a/arch/x86/crypto/sha512_ssse3_glue.c b/arch/x86/crypto/sha512_ssse3_glue.c index 458356a3f124..1c444f41037c 100644 --- a/arch/x86/crypto/sha512_ssse3_glue.c +++ b/arch/x86/crypto/sha512_ssse3_glue.c @@ -39,13 +39,11 @@ #include <crypto/sha512_base.h> #include <asm/simd.h> -asmlinkage void sha512_transform_ssse3(u64 *digest, const char *data, - u64 rounds); - -typedef void (sha512_transform_fn)(u64 *digest, const char *data, u64 rounds); +asmlinkage void sha512_transform_ssse3(struct sha512_state *state, + const u8 *data, int blocks); static int sha512_update(struct shash_desc *desc, const u8 *data, - unsigned int len, sha512_transform_fn *sha512_xform) + unsigned int len, sha512_block_fn *sha512_xform) { struct sha512_state *sctx = shash_desc_ctx(desc); @@ -53,28 +51,29 @@ static int sha512_update(struct shash_desc *desc, const u8 *data, (sctx->count[0] % SHA512_BLOCK_SIZE) + len < SHA512_BLOCK_SIZE) return crypto_sha512_update(desc, data, len); - /* make sure casting to sha512_block_fn() is safe */ + /* + * Make sure struct sha512_state begins directly with the SHA512 + * 512-bit internal state, as this is what the asm functions expect. + */ BUILD_BUG_ON(offsetof(struct sha512_state, state) != 0); kernel_fpu_begin(); - sha512_base_do_update(desc, data, len, - (sha512_block_fn *)sha512_xform); + sha512_base_do_update(desc, data, len, sha512_xform); kernel_fpu_end(); return 0; } static int sha512_finup(struct shash_desc *desc, const u8 *data, - unsigned int len, u8 *out, sha512_transform_fn *sha512_xform) + unsigned int len, u8 *out, sha512_block_fn *sha512_xform) { if (!crypto_simd_usable()) return crypto_sha512_finup(desc, data, len, out); kernel_fpu_begin(); if (len) - sha512_base_do_update(desc, data, len, - (sha512_block_fn *)sha512_xform); - sha512_base_do_finalize(desc, (sha512_block_fn *)sha512_xform); + sha512_base_do_update(desc, data, len, sha512_xform); + sha512_base_do_finalize(desc, sha512_xform); kernel_fpu_end(); return sha512_base_finish(desc, out); @@ -144,8 +143,8 @@ static void unregister_sha512_ssse3(void) } #ifdef CONFIG_AS_AVX -asmlinkage void sha512_transform_avx(u64 *digest, const char *data, - u64 rounds); +asmlinkage void sha512_transform_avx(struct sha512_state *state, + const u8 *data, int blocks); static bool avx_usable(void) { if (!cpu_has_xfeatures(XFEATURE_MASK_SSE | XFEATURE_MASK_YMM, NULL)) { @@ -225,8 +224,8 @@ static inline void unregister_sha512_avx(void) { } #endif #if defined(CONFIG_AS_AVX2) && defined(CONFIG_AS_AVX) -asmlinkage void sha512_transform_rorx(u64 *digest, const char *data, - u64 rounds); +asmlinkage void sha512_transform_rorx(struct sha512_state *state, + const u8 *data, int blocks); static int sha512_avx2_update(struct shash_desc *desc, const u8 *data, unsigned int len) diff --git a/arch/x86/crypto/twofish_avx_glue.c b/arch/x86/crypto/twofish_avx_glue.c index d561c821788b..2dbc8ce3730e 100644 --- a/arch/x86/crypto/twofish_avx_glue.c +++ b/arch/x86/crypto/twofish_avx_glue.c @@ -22,20 +22,17 @@ #define TWOFISH_PARALLEL_BLOCKS 8 /* 8-way parallel cipher functions */ -asmlinkage void twofish_ecb_enc_8way(struct twofish_ctx *ctx, u8 *dst, - const u8 *src); -asmlinkage void twofish_ecb_dec_8way(struct twofish_ctx *ctx, u8 *dst, - const u8 *src); +asmlinkage void twofish_ecb_enc_8way(const void *ctx, u8 *dst, const u8 *src); +asmlinkage void twofish_ecb_dec_8way(const void *ctx, u8 *dst, const u8 *src); -asmlinkage void twofish_cbc_dec_8way(struct twofish_ctx *ctx, u8 *dst, - const u8 *src); -asmlinkage void twofish_ctr_8way(struct twofish_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); +asmlinkage void twofish_cbc_dec_8way(const void *ctx, u8 *dst, const u8 *src); +asmlinkage void twofish_ctr_8way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); -asmlinkage void twofish_xts_enc_8way(struct twofish_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); -asmlinkage void twofish_xts_dec_8way(struct twofish_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); +asmlinkage void twofish_xts_enc_8way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); +asmlinkage void twofish_xts_dec_8way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); static int twofish_setkey_skcipher(struct crypto_skcipher *tfm, const u8 *key, unsigned int keylen) @@ -43,22 +40,19 @@ static int twofish_setkey_skcipher(struct crypto_skcipher *tfm, return twofish_setkey(&tfm->base, key, keylen); } -static inline void twofish_enc_blk_3way(struct twofish_ctx *ctx, u8 *dst, - const u8 *src) +static inline void twofish_enc_blk_3way(const void *ctx, u8 *dst, const u8 *src) { __twofish_enc_blk_3way(ctx, dst, src, false); } -static void twofish_xts_enc(void *ctx, u128 *dst, const u128 *src, le128 *iv) +static void twofish_xts_enc(const void *ctx, u8 *dst, const u8 *src, le128 *iv) { - glue_xts_crypt_128bit_one(ctx, dst, src, iv, - GLUE_FUNC_CAST(twofish_enc_blk)); + glue_xts_crypt_128bit_one(ctx, dst, src, iv, twofish_enc_blk); } -static void twofish_xts_dec(void *ctx, u128 *dst, const u128 *src, le128 *iv) +static void twofish_xts_dec(const void *ctx, u8 *dst, const u8 *src, le128 *iv) { - glue_xts_crypt_128bit_one(ctx, dst, src, iv, - GLUE_FUNC_CAST(twofish_dec_blk)); + glue_xts_crypt_128bit_one(ctx, dst, src, iv, twofish_dec_blk); } struct twofish_xts_ctx { @@ -70,7 +64,6 @@ static int xts_twofish_setkey(struct crypto_skcipher *tfm, const u8 *key, unsigned int keylen) { struct twofish_xts_ctx *ctx = crypto_skcipher_ctx(tfm); - u32 *flags = &tfm->base.crt_flags; int err; err = xts_verify_key(tfm, key, keylen); @@ -78,13 +71,12 @@ static int xts_twofish_setkey(struct crypto_skcipher *tfm, const u8 *key, return err; /* first half of xts-key is for crypt */ - err = __twofish_setkey(&ctx->crypt_ctx, key, keylen / 2, flags); + err = __twofish_setkey(&ctx->crypt_ctx, key, keylen / 2); if (err) return err; /* second half of xts-key is for tweak */ - return __twofish_setkey(&ctx->tweak_ctx, key + keylen / 2, keylen / 2, - flags); + return __twofish_setkey(&ctx->tweak_ctx, key + keylen / 2, keylen / 2); } static const struct common_glue_ctx twofish_enc = { @@ -93,13 +85,13 @@ static const struct common_glue_ctx twofish_enc = { .funcs = { { .num_blocks = TWOFISH_PARALLEL_BLOCKS, - .fn_u = { .ecb = GLUE_FUNC_CAST(twofish_ecb_enc_8way) } + .fn_u = { .ecb = twofish_ecb_enc_8way } }, { .num_blocks = 3, - .fn_u = { .ecb = GLUE_FUNC_CAST(twofish_enc_blk_3way) } + .fn_u = { .ecb = twofish_enc_blk_3way } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(twofish_enc_blk) } + .fn_u = { .ecb = twofish_enc_blk } } } }; @@ -109,13 +101,13 @@ static const struct common_glue_ctx twofish_ctr = { .funcs = { { .num_blocks = TWOFISH_PARALLEL_BLOCKS, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(twofish_ctr_8way) } + .fn_u = { .ctr = twofish_ctr_8way } }, { .num_blocks = 3, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(twofish_enc_blk_ctr_3way) } + .fn_u = { .ctr = twofish_enc_blk_ctr_3way } }, { .num_blocks = 1, - .fn_u = { .ctr = GLUE_CTR_FUNC_CAST(twofish_enc_blk_ctr) } + .fn_u = { .ctr = twofish_enc_blk_ctr } } } }; @@ -125,10 +117,10 @@ static const struct common_glue_ctx twofish_enc_xts = { .funcs = { { .num_blocks = TWOFISH_PARALLEL_BLOCKS, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(twofish_xts_enc_8way) } + .fn_u = { .xts = twofish_xts_enc_8way } }, { .num_blocks = 1, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(twofish_xts_enc) } + .fn_u = { .xts = twofish_xts_enc } } } }; @@ -138,13 +130,13 @@ static const struct common_glue_ctx twofish_dec = { .funcs = { { .num_blocks = TWOFISH_PARALLEL_BLOCKS, - .fn_u = { .ecb = GLUE_FUNC_CAST(twofish_ecb_dec_8way) } + .fn_u = { .ecb = twofish_ecb_dec_8way } }, { .num_blocks = 3, - .fn_u = { .ecb = GLUE_FUNC_CAST(twofish_dec_blk_3way) } + .fn_u = { .ecb = twofish_dec_blk_3way } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(twofish_dec_blk) } + .fn_u = { .ecb = twofish_dec_blk } } } }; @@ -154,13 +146,13 @@ static const struct common_glue_ctx twofish_dec_cbc = { .funcs = { { .num_blocks = TWOFISH_PARALLEL_BLOCKS, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(twofish_cbc_dec_8way) } + .fn_u = { .cbc = twofish_cbc_dec_8way } }, { .num_blocks = 3, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(twofish_dec_blk_cbc_3way) } + .fn_u = { .cbc = twofish_dec_blk_cbc_3way } }, { .num_blocks = 1, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(twofish_dec_blk) } + .fn_u = { .cbc = twofish_dec_blk } } } }; @@ -170,10 +162,10 @@ static const struct common_glue_ctx twofish_dec_xts = { .funcs = { { .num_blocks = TWOFISH_PARALLEL_BLOCKS, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(twofish_xts_dec_8way) } + .fn_u = { .xts = twofish_xts_dec_8way } }, { .num_blocks = 1, - .fn_u = { .xts = GLUE_XTS_FUNC_CAST(twofish_xts_dec) } + .fn_u = { .xts = twofish_xts_dec } } } }; @@ -189,8 +181,7 @@ static int ecb_decrypt(struct skcipher_request *req) static int cbc_encrypt(struct skcipher_request *req) { - return glue_cbc_encrypt_req_128bit(GLUE_FUNC_CAST(twofish_enc_blk), - req); + return glue_cbc_encrypt_req_128bit(twofish_enc_blk, req); } static int cbc_decrypt(struct skcipher_request *req) @@ -208,8 +199,7 @@ static int xts_encrypt(struct skcipher_request *req) struct crypto_skcipher *tfm = crypto_skcipher_reqtfm(req); struct twofish_xts_ctx *ctx = crypto_skcipher_ctx(tfm); - return glue_xts_req_128bit(&twofish_enc_xts, req, - XTS_TWEAK_CAST(twofish_enc_blk), + return glue_xts_req_128bit(&twofish_enc_xts, req, twofish_enc_blk, &ctx->tweak_ctx, &ctx->crypt_ctx, false); } @@ -218,8 +208,7 @@ static int xts_decrypt(struct skcipher_request *req) struct crypto_skcipher *tfm = crypto_skcipher_reqtfm(req); struct twofish_xts_ctx *ctx = crypto_skcipher_ctx(tfm); - return glue_xts_req_128bit(&twofish_dec_xts, req, - XTS_TWEAK_CAST(twofish_enc_blk), + return glue_xts_req_128bit(&twofish_dec_xts, req, twofish_enc_blk, &ctx->tweak_ctx, &ctx->crypt_ctx, true); } diff --git a/arch/x86/crypto/twofish_glue_3way.c b/arch/x86/crypto/twofish_glue_3way.c index 1dc9e29f221e..768af6075479 100644 --- a/arch/x86/crypto/twofish_glue_3way.c +++ b/arch/x86/crypto/twofish_glue_3way.c @@ -25,21 +25,22 @@ static int twofish_setkey_skcipher(struct crypto_skcipher *tfm, return twofish_setkey(&tfm->base, key, keylen); } -static inline void twofish_enc_blk_3way(struct twofish_ctx *ctx, u8 *dst, - const u8 *src) +static inline void twofish_enc_blk_3way(const void *ctx, u8 *dst, const u8 *src) { __twofish_enc_blk_3way(ctx, dst, src, false); } -static inline void twofish_enc_blk_xor_3way(struct twofish_ctx *ctx, u8 *dst, +static inline void twofish_enc_blk_xor_3way(const void *ctx, u8 *dst, const u8 *src) { __twofish_enc_blk_3way(ctx, dst, src, true); } -void twofish_dec_blk_cbc_3way(void *ctx, u128 *dst, const u128 *src) +void twofish_dec_blk_cbc_3way(const void *ctx, u8 *d, const u8 *s) { u128 ivs[2]; + u128 *dst = (u128 *)d; + const u128 *src = (const u128 *)s; ivs[0] = src[0]; ivs[1] = src[1]; @@ -51,9 +52,11 @@ void twofish_dec_blk_cbc_3way(void *ctx, u128 *dst, const u128 *src) } EXPORT_SYMBOL_GPL(twofish_dec_blk_cbc_3way); -void twofish_enc_blk_ctr(void *ctx, u128 *dst, const u128 *src, le128 *iv) +void twofish_enc_blk_ctr(const void *ctx, u8 *d, const u8 *s, le128 *iv) { be128 ctrblk; + u128 *dst = (u128 *)d; + const u128 *src = (const u128 *)s; if (dst != src) *dst = *src; @@ -66,10 +69,11 @@ void twofish_enc_blk_ctr(void *ctx, u128 *dst, const u128 *src, le128 *iv) } EXPORT_SYMBOL_GPL(twofish_enc_blk_ctr); -void twofish_enc_blk_ctr_3way(void *ctx, u128 *dst, const u128 *src, - le128 *iv) +void twofish_enc_blk_ctr_3way(const void *ctx, u8 *d, const u8 *s, le128 *iv) { be128 ctrblks[3]; + u128 *dst = (u128 *)d; + const u128 *src = (const u128 *)s; if (dst != src) { dst[0] = src[0]; @@ -94,10 +98,10 @@ static const struct common_glue_ctx twofish_enc = { .funcs = { { .num_blocks = 3, - .fn_u = { .ecb = GLUE_FUNC_CAST(twofish_enc_blk_3way) } + .fn_u = { .ecb = twofish_enc_blk_3way } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(twofish_enc_blk) } + .fn_u = { .ecb = twofish_enc_blk } } } }; @@ -107,10 +111,10 @@ static const struct common_glue_ctx twofish_ctr = { .funcs = { { .num_blocks = 3, - .fn_u = { .ecb = GLUE_FUNC_CAST(twofish_enc_blk_ctr_3way) } + .fn_u = { .ctr = twofish_enc_blk_ctr_3way } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(twofish_enc_blk_ctr) } + .fn_u = { .ctr = twofish_enc_blk_ctr } } } }; @@ -120,10 +124,10 @@ static const struct common_glue_ctx twofish_dec = { .funcs = { { .num_blocks = 3, - .fn_u = { .ecb = GLUE_FUNC_CAST(twofish_dec_blk_3way) } + .fn_u = { .ecb = twofish_dec_blk_3way } }, { .num_blocks = 1, - .fn_u = { .ecb = GLUE_FUNC_CAST(twofish_dec_blk) } + .fn_u = { .ecb = twofish_dec_blk } } } }; @@ -133,10 +137,10 @@ static const struct common_glue_ctx twofish_dec_cbc = { .funcs = { { .num_blocks = 3, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(twofish_dec_blk_cbc_3way) } + .fn_u = { .cbc = twofish_dec_blk_cbc_3way } }, { .num_blocks = 1, - .fn_u = { .cbc = GLUE_CBC_FUNC_CAST(twofish_dec_blk) } + .fn_u = { .cbc = twofish_dec_blk } } } }; @@ -152,8 +156,7 @@ static int ecb_decrypt(struct skcipher_request *req) static int cbc_encrypt(struct skcipher_request *req) { - return glue_cbc_encrypt_req_128bit(GLUE_FUNC_CAST(twofish_enc_blk), - req); + return glue_cbc_encrypt_req_128bit(twofish_enc_blk, req); } static int cbc_decrypt(struct skcipher_request *req) diff --git a/arch/x86/entry/entry_64.S b/arch/x86/entry/entry_64.S index 76942cbd95a1..f2bb91e87877 100644 --- a/arch/x86/entry/entry_64.S +++ b/arch/x86/entry/entry_64.S @@ -1728,7 +1728,7 @@ SYM_CODE_END(nmi) SYM_CODE_START(ignore_sysret) UNWIND_HINT_EMPTY mov $-ENOSYS, %eax - sysret + sysretl SYM_CODE_END(ignore_sysret) #endif diff --git a/arch/x86/entry/vdso/vdso-layout.lds.S b/arch/x86/entry/vdso/vdso-layout.lds.S index 93c6dc7812d0..ea7e0155c604 100644 --- a/arch/x86/entry/vdso/vdso-layout.lds.S +++ b/arch/x86/entry/vdso/vdso-layout.lds.S @@ -16,18 +16,23 @@ SECTIONS * segment. */ - vvar_start = . - 3 * PAGE_SIZE; - vvar_page = vvar_start; + vvar_start = . - 4 * PAGE_SIZE; + vvar_page = vvar_start; /* Place all vvars at the offsets in asm/vvar.h. */ #define EMIT_VVAR(name, offset) vvar_ ## name = vvar_page + offset; -#define __VVAR_KERNEL_LDS #include <asm/vvar.h> -#undef __VVAR_KERNEL_LDS #undef EMIT_VVAR pvclock_page = vvar_start + PAGE_SIZE; hvclock_page = vvar_start + 2 * PAGE_SIZE; + timens_page = vvar_start + 3 * PAGE_SIZE; + +#undef _ASM_X86_VVAR_H + /* Place all vvars in timens too at the offsets in asm/vvar.h. */ +#define EMIT_VVAR(name, offset) timens_ ## name = timens_page + offset; +#include <asm/vvar.h> +#undef EMIT_VVAR . = SIZEOF_HEADERS; diff --git a/arch/x86/entry/vdso/vdso2c.c b/arch/x86/entry/vdso/vdso2c.c index 3a4d8d4d39f8..3842873b3ae3 100644 --- a/arch/x86/entry/vdso/vdso2c.c +++ b/arch/x86/entry/vdso/vdso2c.c @@ -75,12 +75,14 @@ enum { sym_vvar_page, sym_pvclock_page, sym_hvclock_page, + sym_timens_page, }; const int special_pages[] = { sym_vvar_page, sym_pvclock_page, sym_hvclock_page, + sym_timens_page, }; struct vdso_sym { @@ -93,6 +95,7 @@ struct vdso_sym required_syms[] = { [sym_vvar_page] = {"vvar_page", true}, [sym_pvclock_page] = {"pvclock_page", true}, [sym_hvclock_page] = {"hvclock_page", true}, + [sym_timens_page] = {"timens_page", true}, {"VDSO32_NOTE_MASK", true}, {"__kernel_vsyscall", true}, {"__kernel_sigreturn", true}, diff --git a/arch/x86/entry/vdso/vdso32-setup.c b/arch/x86/entry/vdso/vdso32-setup.c index 240626e7f55a..43842fade8fa 100644 --- a/arch/x86/entry/vdso/vdso32-setup.c +++ b/arch/x86/entry/vdso/vdso32-setup.c @@ -11,6 +11,7 @@ #include <linux/smp.h> #include <linux/kernel.h> #include <linux/mm_types.h> +#include <linux/elf.h> #include <asm/processor.h> #include <asm/vdso.h> diff --git a/arch/x86/entry/vdso/vma.c b/arch/x86/entry/vdso/vma.c index f5937742b290..c1b8496b5606 100644 --- a/arch/x86/entry/vdso/vma.c +++ b/arch/x86/entry/vdso/vma.c @@ -14,16 +14,30 @@ #include <linux/elf.h> #include <linux/cpu.h> #include <linux/ptrace.h> +#include <linux/time_namespace.h> + #include <asm/pvclock.h> #include <asm/vgtod.h> #include <asm/proto.h> #include <asm/vdso.h> #include <asm/vvar.h> +#include <asm/tlb.h> #include <asm/page.h> #include <asm/desc.h> #include <asm/cpufeature.h> #include <clocksource/hyperv_timer.h> +#undef _ASM_X86_VVAR_H +#define EMIT_VVAR(name, offset) \ + const size_t name ## _offset = offset; +#include <asm/vvar.h> + +struct vdso_data *arch_get_vdso_data(void *vvar_page) +{ + return (struct vdso_data *)(vvar_page + _vdso_data_offset); +} +#undef EMIT_VVAR + #if defined(CONFIG_X86_64) unsigned int __read_mostly vdso64_enabled = 1; #endif @@ -37,6 +51,7 @@ void __init init_vdso_image(const struct vdso_image *image) image->alt_len)); } +static const struct vm_special_mapping vvar_mapping; struct linux_binprm; static vm_fault_t vdso_fault(const struct vm_special_mapping *sm, @@ -84,10 +99,74 @@ static int vdso_mremap(const struct vm_special_mapping *sm, return 0; } +static int vvar_mremap(const struct vm_special_mapping *sm, + struct vm_area_struct *new_vma) +{ + const struct vdso_image *image = new_vma->vm_mm->context.vdso_image; + unsigned long new_size = new_vma->vm_end - new_vma->vm_start; + + if (new_size != -image->sym_vvar_start) + return -EINVAL; + + return 0; +} + +#ifdef CONFIG_TIME_NS +static struct page *find_timens_vvar_page(struct vm_area_struct *vma) +{ + if (likely(vma->vm_mm == current->mm)) + return current->nsproxy->time_ns->vvar_page; + + /* + * VM_PFNMAP | VM_IO protect .fault() handler from being called + * through interfaces like /proc/$pid/mem or + * process_vm_{readv,writev}() as long as there's no .access() + * in special_mapping_vmops(). + * For more details check_vma_flags() and __access_remote_vm() + */ + + WARN(1, "vvar_page accessed remotely"); + + return NULL; +} + +/* + * The vvar page layout depends on whether a task belongs to the root or + * non-root time namespace. Whenever a task changes its namespace, the VVAR + * page tables are cleared and then they will re-faulted with a + * corresponding layout. + * See also the comment near timens_setup_vdso_data() for details. + */ +int vdso_join_timens(struct task_struct *task, struct time_namespace *ns) +{ + struct mm_struct *mm = task->mm; + struct vm_area_struct *vma; + + if (down_write_killable(&mm->mmap_sem)) + return -EINTR; + + for (vma = mm->mmap; vma; vma = vma->vm_next) { + unsigned long size = vma->vm_end - vma->vm_start; + + if (vma_is_special_mapping(vma, &vvar_mapping)) + zap_page_range(vma, vma->vm_start, size); + } + + up_write(&mm->mmap_sem); + return 0; +} +#else +static inline struct page *find_timens_vvar_page(struct vm_area_struct *vma) +{ + return NULL; +} +#endif + static vm_fault_t vvar_fault(const struct vm_special_mapping *sm, struct vm_area_struct *vma, struct vm_fault *vmf) { const struct vdso_image *image = vma->vm_mm->context.vdso_image; + unsigned long pfn; long sym_offset; if (!image) @@ -107,8 +186,36 @@ static vm_fault_t vvar_fault(const struct vm_special_mapping *sm, return VM_FAULT_SIGBUS; if (sym_offset == image->sym_vvar_page) { - return vmf_insert_pfn(vma, vmf->address, - __pa_symbol(&__vvar_page) >> PAGE_SHIFT); + struct page *timens_page = find_timens_vvar_page(vma); + + pfn = __pa_symbol(&__vvar_page) >> PAGE_SHIFT; + + /* + * If a task belongs to a time namespace then a namespace + * specific VVAR is mapped with the sym_vvar_page offset and + * the real VVAR page is mapped with the sym_timens_page + * offset. + * See also the comment near timens_setup_vdso_data(). + */ + if (timens_page) { + unsigned long addr; + vm_fault_t err; + + /* + * Optimization: inside time namespace pre-fault + * VVAR page too. As on timens page there are only + * offsets for clocks on VVAR, it'll be faulted + * shortly by VDSO code. + */ + addr = vmf->address + (image->sym_timens_page - sym_offset); + err = vmf_insert_pfn(vma, addr, pfn); + if (unlikely(err & VM_FAULT_ERROR)) + return err; + + pfn = page_to_pfn(timens_page); + } + + return vmf_insert_pfn(vma, vmf->address, pfn); } else if (sym_offset == image->sym_pvclock_page) { struct pvclock_vsyscall_time_info *pvti = pvclock_get_pvti_cpu0_va(); @@ -123,6 +230,14 @@ static vm_fault_t vvar_fault(const struct vm_special_mapping *sm, if (tsc_pg && vclock_was_used(VCLOCK_HVCLOCK)) return vmf_insert_pfn(vma, vmf->address, virt_to_phys(tsc_pg) >> PAGE_SHIFT); + } else if (sym_offset == image->sym_timens_page) { + struct page *timens_page = find_timens_vvar_page(vma); + + if (!timens_page) + return VM_FAULT_SIGBUS; + + pfn = __pa_symbol(&__vvar_page) >> PAGE_SHIFT; + return vmf_insert_pfn(vma, vmf->address, pfn); } return VM_FAULT_SIGBUS; @@ -136,6 +251,7 @@ static const struct vm_special_mapping vdso_mapping = { static const struct vm_special_mapping vvar_mapping = { .name = "[vvar]", .fault = vvar_fault, + .mremap = vvar_mremap, }; /* diff --git a/arch/x86/events/amd/core.c b/arch/x86/events/amd/core.c index a7752cd78b89..1f22b6bbda68 100644 --- a/arch/x86/events/amd/core.c +++ b/arch/x86/events/amd/core.c @@ -14,6 +14,10 @@ static DEFINE_PER_CPU(unsigned long, perf_nmi_tstamp); static unsigned long perf_nmi_window; +/* AMD Event 0xFFF: Merge. Used with Large Increment per Cycle events */ +#define AMD_MERGE_EVENT ((0xFULL << 32) | 0xFFULL) +#define AMD_MERGE_EVENT_ENABLE (AMD_MERGE_EVENT | ARCH_PERFMON_EVENTSEL_ENABLE) + static __initconst const u64 amd_hw_cache_event_ids [PERF_COUNT_HW_CACHE_MAX] [PERF_COUNT_HW_CACHE_OP_MAX] @@ -301,6 +305,25 @@ static inline int amd_pmu_addr_offset(int index, bool eventsel) return offset; } +/* + * AMD64 events are detected based on their event codes. + */ +static inline unsigned int amd_get_event_code(struct hw_perf_event *hwc) +{ + return ((hwc->config >> 24) & 0x0f00) | (hwc->config & 0x00ff); +} + +static inline bool amd_is_pair_event_code(struct hw_perf_event *hwc) +{ + if (!(x86_pmu.flags & PMU_FL_PAIR)) + return false; + + switch (amd_get_event_code(hwc)) { + case 0x003: return true; /* Retired SSE/AVX FLOPs */ + default: return false; + } +} + static int amd_core_hw_config(struct perf_event *event) { if (event->attr.exclude_host && event->attr.exclude_guest) @@ -316,15 +339,10 @@ static int amd_core_hw_config(struct perf_event *event) else if (event->attr.exclude_guest) event->hw.config |= AMD64_EVENTSEL_HOSTONLY; - return 0; -} + if ((x86_pmu.flags & PMU_FL_PAIR) && amd_is_pair_event_code(&event->hw)) + event->hw.flags |= PERF_X86_EVENT_PAIR; -/* - * AMD64 events are detected based on their event codes. - */ -static inline unsigned int amd_get_event_code(struct hw_perf_event *hwc) -{ - return ((hwc->config >> 24) & 0x0f00) | (hwc->config & 0x00ff); + return 0; } static inline int amd_is_nb_event(struct hw_perf_event *hwc) @@ -855,6 +873,29 @@ amd_get_event_constraints_f15h(struct cpu_hw_events *cpuc, int idx, } } +static struct event_constraint pair_constraint; + +static struct event_constraint * +amd_get_event_constraints_f17h(struct cpu_hw_events *cpuc, int idx, + struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + + if (amd_is_pair_event_code(hwc)) + return &pair_constraint; + + return &unconstrained; +} + +static void amd_put_event_constraints_f17h(struct cpu_hw_events *cpuc, + struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + + if (is_counter_pair(hwc)) + --cpuc->n_pair; +} + static ssize_t amd_event_sysfs_show(char *page, u64 config) { u64 event = (config & ARCH_PERFMON_EVENTSEL_EVENT) | @@ -898,33 +939,15 @@ static __initconst const struct x86_pmu amd_pmu = { static int __init amd_core_pmu_init(void) { + u64 even_ctr_mask = 0ULL; + int i; + if (!boot_cpu_has(X86_FEATURE_PERFCTR_CORE)) return 0; - /* Avoid calulating the value each time in the NMI handler */ + /* Avoid calculating the value each time in the NMI handler */ perf_nmi_window = msecs_to_jiffies(100); - switch (boot_cpu_data.x86) { - case 0x15: - pr_cont("Fam15h "); - x86_pmu.get_event_constraints = amd_get_event_constraints_f15h; - break; - case 0x17: - pr_cont("Fam17h "); - /* - * In family 17h, there are no event constraints in the PMC hardware. - * We fallback to using default amd_get_event_constraints. - */ - break; - case 0x18: - pr_cont("Fam18h "); - /* Using default amd_get_event_constraints. */ - break; - default: - pr_err("core perfctr but no constraints; unknown hardware!\n"); - return -ENODEV; - } - /* * If core performance counter extensions exists, we must use * MSR_F15H_PERF_CTL/MSR_F15H_PERF_CTR msrs. See also @@ -939,6 +962,32 @@ static int __init amd_core_pmu_init(void) */ x86_pmu.amd_nb_constraints = 0; + if (boot_cpu_data.x86 == 0x15) { + pr_cont("Fam15h "); + x86_pmu.get_event_constraints = amd_get_event_constraints_f15h; + } + if (boot_cpu_data.x86 >= 0x17) { + pr_cont("Fam17h+ "); + /* + * Family 17h and compatibles have constraints for Large + * Increment per Cycle events: they may only be assigned an + * even numbered counter that has a consecutive adjacent odd + * numbered counter following it. + */ + for (i = 0; i < x86_pmu.num_counters - 1; i += 2) + even_ctr_mask |= 1 << i; + + pair_constraint = (struct event_constraint) + __EVENT_CONSTRAINT(0, even_ctr_mask, 0, + x86_pmu.num_counters / 2, 0, + PERF_X86_EVENT_PAIR); + + x86_pmu.get_event_constraints = amd_get_event_constraints_f17h; + x86_pmu.put_event_constraints = amd_put_event_constraints_f17h; + x86_pmu.perf_ctr_pair_en = AMD_MERGE_EVENT_ENABLE; + x86_pmu.flags |= PMU_FL_PAIR; + } + pr_cont("core perfctr, "); return 0; } diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c index f118af9f0718..3bb738f5a472 100644 --- a/arch/x86/events/core.c +++ b/arch/x86/events/core.c @@ -618,6 +618,7 @@ void x86_pmu_disable_all(void) int idx; for (idx = 0; idx < x86_pmu.num_counters; idx++) { + struct hw_perf_event *hwc = &cpuc->events[idx]->hw; u64 val; if (!test_bit(idx, cpuc->active_mask)) @@ -627,6 +628,8 @@ void x86_pmu_disable_all(void) continue; val &= ~ARCH_PERFMON_EVENTSEL_ENABLE; wrmsrl(x86_pmu_config_addr(idx), val); + if (is_counter_pair(hwc)) + wrmsrl(x86_pmu_config_addr(idx + 1), 0); } } @@ -699,7 +702,7 @@ struct sched_state { int counter; /* counter index */ int unassigned; /* number of events to be assigned left */ int nr_gp; /* number of GP counters used */ - unsigned long used[BITS_TO_LONGS(X86_PMC_IDX_MAX)]; + u64 used; }; /* Total max is X86_PMC_IDX_MAX, but we are O(n!) limited */ @@ -756,8 +759,12 @@ static bool perf_sched_restore_state(struct perf_sched *sched) sched->saved_states--; sched->state = sched->saved[sched->saved_states]; - /* continue with next counter: */ - clear_bit(sched->state.counter++, sched->state.used); + /* this assignment didn't work out */ + /* XXX broken vs EVENT_PAIR */ + sched->state.used &= ~BIT_ULL(sched->state.counter); + + /* try the next one */ + sched->state.counter++; return true; } @@ -782,20 +789,32 @@ static bool __perf_sched_find_counter(struct perf_sched *sched) if (c->idxmsk64 & (~0ULL << INTEL_PMC_IDX_FIXED)) { idx = INTEL_PMC_IDX_FIXED; for_each_set_bit_from(idx, c->idxmsk, X86_PMC_IDX_MAX) { - if (!__test_and_set_bit(idx, sched->state.used)) - goto done; + u64 mask = BIT_ULL(idx); + + if (sched->state.used & mask) + continue; + + sched->state.used |= mask; + goto done; } } /* Grab the first unused counter starting with idx */ idx = sched->state.counter; for_each_set_bit_from(idx, c->idxmsk, INTEL_PMC_IDX_FIXED) { - if (!__test_and_set_bit(idx, sched->state.used)) { - if (sched->state.nr_gp++ >= sched->max_gp) - return false; + u64 mask = BIT_ULL(idx); - goto done; - } + if (c->flags & PERF_X86_EVENT_PAIR) + mask |= mask << 1; + + if (sched->state.used & mask) + continue; + + if (sched->state.nr_gp++ >= sched->max_gp) + return false; + + sched->state.used |= mask; + goto done; } return false; @@ -872,12 +891,10 @@ EXPORT_SYMBOL_GPL(perf_assign_events); int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign) { struct event_constraint *c; - unsigned long used_mask[BITS_TO_LONGS(X86_PMC_IDX_MAX)]; struct perf_event *e; int n0, i, wmin, wmax, unsched = 0; struct hw_perf_event *hwc; - - bitmap_zero(used_mask, X86_PMC_IDX_MAX); + u64 used_mask = 0; /* * Compute the number of events already present; see x86_pmu_add(), @@ -920,6 +937,8 @@ int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign) * fastpath, try to reuse previous register */ for (i = 0; i < n; i++) { + u64 mask; + hwc = &cpuc->event_list[i]->hw; c = cpuc->event_constraint[i]; @@ -931,11 +950,16 @@ int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign) if (!test_bit(hwc->idx, c->idxmsk)) break; + mask = BIT_ULL(hwc->idx); + if (is_counter_pair(hwc)) + mask |= mask << 1; + /* not already used */ - if (test_bit(hwc->idx, used_mask)) + if (used_mask & mask) break; - __set_bit(hwc->idx, used_mask); + used_mask |= mask; + if (assign) assign[i] = hwc->idx; } @@ -958,6 +982,15 @@ int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign) READ_ONCE(cpuc->excl_cntrs->exclusive_present)) gpmax /= 2; + /* + * Reduce the amount of available counters to allow fitting + * the extra Merge events needed by large increment events. + */ + if (x86_pmu.flags & PMU_FL_PAIR) { + gpmax = x86_pmu.num_counters - cpuc->n_pair; + WARN_ON(gpmax <= 0); + } + unsched = perf_assign_events(cpuc->event_constraint, n, wmin, wmax, gpmax, assign); } @@ -1038,6 +1071,8 @@ static int collect_events(struct cpu_hw_events *cpuc, struct perf_event *leader, return -EINVAL; cpuc->event_list[n] = leader; n++; + if (is_counter_pair(&leader->hw)) + cpuc->n_pair++; } if (!dogrp) return n; @@ -1052,6 +1087,8 @@ static int collect_events(struct cpu_hw_events *cpuc, struct perf_event *leader, cpuc->event_list[n] = event; n++; + if (is_counter_pair(&event->hw)) + cpuc->n_pair++; } return n; } @@ -1238,6 +1275,13 @@ int x86_perf_event_set_period(struct perf_event *event) wrmsrl(hwc->event_base, (u64)(-left) & x86_pmu.cntval_mask); /* + * Clear the Merge event counter's upper 16 bits since + * we currently declare a 48-bit counter width + */ + if (is_counter_pair(hwc)) + wrmsrl(x86_pmu_event_addr(idx + 1), 0); + + /* * Due to erratum on certan cpu we need * a second write to be sure the register * is updated properly diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index ce83950036c5..4b94ae4ae369 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -7,6 +7,7 @@ #include <asm/perf_event.h> #include <asm/tlbflush.h> #include <asm/insn.h> +#include <asm/io.h> #include "../perf_event.h" diff --git a/arch/x86/events/intel/rapl.c b/arch/x86/events/intel/rapl.c index 5053a403e4ae..09913121e726 100644 --- a/arch/x86/events/intel/rapl.c +++ b/arch/x86/events/intel/rapl.c @@ -741,6 +741,8 @@ static const struct x86_cpu_id rapl_model_match[] __initconst = { X86_RAPL_MODEL_MATCH(INTEL_FAM6_ATOM_GOLDMONT_PLUS, model_hsw), X86_RAPL_MODEL_MATCH(INTEL_FAM6_ICELAKE_L, model_skl), X86_RAPL_MODEL_MATCH(INTEL_FAM6_ICELAKE, model_skl), + X86_RAPL_MODEL_MATCH(INTEL_FAM6_COMETLAKE_L, model_skl), + X86_RAPL_MODEL_MATCH(INTEL_FAM6_COMETLAKE, model_skl), {}, }; diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h index 930611db8f9a..f1cd1ca1a77b 100644 --- a/arch/x86/events/perf_event.h +++ b/arch/x86/events/perf_event.h @@ -77,6 +77,7 @@ static inline bool constraint_match(struct event_constraint *c, u64 ecode) #define PERF_X86_EVENT_AUTO_RELOAD 0x0200 /* use PEBS auto-reload */ #define PERF_X86_EVENT_LARGE_PEBS 0x0400 /* use large PEBS */ #define PERF_X86_EVENT_PEBS_VIA_PT 0x0800 /* use PT buffer for PEBS */ +#define PERF_X86_EVENT_PAIR 0x1000 /* Large Increment per Cycle */ struct amd_nb { int nb_id; /* NorthBridge id */ @@ -272,6 +273,7 @@ struct cpu_hw_events { struct amd_nb *amd_nb; /* Inverted mask of bits to clear in the perf_ctr ctrl registers */ u64 perf_ctr_virt_mask; + int n_pair; /* Large increment events */ void *kfree_on_online[X86_PERF_KFREE_MAX]; }; @@ -694,6 +696,7 @@ struct x86_pmu { * AMD bits */ unsigned int amd_nb_constraints : 1; + u64 perf_ctr_pair_en; /* * Extra registers for events @@ -743,6 +746,7 @@ do { \ #define PMU_FL_EXCL_ENABLED 0x8 /* exclusive counter active */ #define PMU_FL_PEBS_ALL 0x10 /* all events are valid PEBS events */ #define PMU_FL_TFA 0x20 /* deal with TSX force abort */ +#define PMU_FL_PAIR 0x40 /* merge counters for large incr. events */ #define EVENT_VAR(_id) event_attr_##_id #define EVENT_PTR(_id) &event_attr_##_id.attr.attr @@ -838,6 +842,11 @@ int x86_pmu_hw_config(struct perf_event *event); void x86_pmu_disable_all(void); +static inline bool is_counter_pair(struct hw_perf_event *hwc) +{ + return hwc->flags & PERF_X86_EVENT_PAIR; +} + static inline void __x86_pmu_enable_event(struct hw_perf_event *hwc, u64 enable_mask) { @@ -845,6 +854,14 @@ static inline void __x86_pmu_enable_event(struct hw_perf_event *hwc, if (hwc->extra_reg.reg) wrmsrl(hwc->extra_reg.reg, hwc->extra_reg.config); + + /* + * Add enabled Merge event on next counter + * if large increment event being enabled on this counter + */ + if (is_counter_pair(hwc)) + wrmsrl(x86_pmu_config_addr(hwc->idx + 1), x86_pmu.perf_ctr_pair_en); + wrmsrl(hwc->config_base, (hwc->config | enable_mask) & ~disable_mask); } @@ -861,6 +878,9 @@ static inline void x86_pmu_disable_event(struct perf_event *event) struct hw_perf_event *hwc = &event->hw; wrmsrl(hwc->config_base, hwc->config); + + if (is_counter_pair(hwc)) + wrmsrl(x86_pmu_config_addr(hwc->idx + 1), 0); } void x86_pmu_enable_event(struct perf_event *event); diff --git a/arch/x86/ia32/ia32_signal.c b/arch/x86/ia32/ia32_signal.c index 30416d7f19d4..a3aefe9b9401 100644 --- a/arch/x86/ia32/ia32_signal.c +++ b/arch/x86/ia32/ia32_signal.c @@ -114,8 +114,6 @@ static int ia32_restore_sigcontext(struct pt_regs *regs, err |= fpu__restore_sig(buf, 1); - force_iret(); - return err; } diff --git a/arch/x86/include/asm/acpi.h b/arch/x86/include/asm/acpi.h index bc9693c9107e..ca0976456a6b 100644 --- a/arch/x86/include/asm/acpi.h +++ b/arch/x86/include/asm/acpi.h @@ -13,7 +13,6 @@ #include <asm/processor.h> #include <asm/mmu.h> #include <asm/mpspec.h> -#include <asm/realmode.h> #include <asm/x86_init.h> #ifdef CONFIG_ACPI_APEI @@ -62,7 +61,7 @@ static inline void acpi_disable_pci(void) extern int (*acpi_suspend_lowlevel)(void); /* Physical address to resume after wakeup */ -#define acpi_wakeup_address ((unsigned long)(real_mode_header->wakeup_start)) +unsigned long acpi_get_wakeup_address(void); /* * Check if the CPU can handle C2 and deeper diff --git a/arch/x86/include/asm/cpu_entry_area.h b/arch/x86/include/asm/cpu_entry_area.h index 804734058c77..02c0078d3787 100644 --- a/arch/x86/include/asm/cpu_entry_area.h +++ b/arch/x86/include/asm/cpu_entry_area.h @@ -6,6 +6,7 @@ #include <linux/percpu-defs.h> #include <asm/processor.h> #include <asm/intel_ds.h> +#include <asm/pgtable_areas.h> #ifdef CONFIG_X86_64 @@ -134,15 +135,6 @@ DECLARE_PER_CPU(struct cea_exception_stacks *, cea_exception_stacks); extern void setup_cpu_entry_areas(void); extern void cea_set_pte(void *cea_vaddr, phys_addr_t pa, pgprot_t flags); -/* Single page reserved for the readonly IDT mapping: */ -#define CPU_ENTRY_AREA_RO_IDT CPU_ENTRY_AREA_BASE -#define CPU_ENTRY_AREA_PER_CPU (CPU_ENTRY_AREA_RO_IDT + PAGE_SIZE) - -#define CPU_ENTRY_AREA_RO_IDT_VADDR ((void *)CPU_ENTRY_AREA_RO_IDT) - -#define CPU_ENTRY_AREA_MAP_SIZE \ - (CPU_ENTRY_AREA_PER_CPU + CPU_ENTRY_AREA_ARRAY_SIZE - CPU_ENTRY_AREA_BASE) - extern struct cpu_entry_area *get_cpu_entry_area(int cpu); static inline struct entry_stack *cpu_entry_stack(int cpu) diff --git a/arch/x86/include/asm/cpufeatures.h b/arch/x86/include/asm/cpufeatures.h index e9b62498fe75..f3327cb56edf 100644 --- a/arch/x86/include/asm/cpufeatures.h +++ b/arch/x86/include/asm/cpufeatures.h @@ -220,6 +220,7 @@ #define X86_FEATURE_ZEN ( 7*32+28) /* "" CPU is AMD family 0x17 (Zen) */ #define X86_FEATURE_L1TF_PTEINV ( 7*32+29) /* "" L1TF workaround PTE inversion */ #define X86_FEATURE_IBRS_ENHANCED ( 7*32+30) /* Enhanced IBRS */ +#define X86_FEATURE_MSR_IA32_FEAT_CTL ( 7*32+31) /* "" MSR IA32_FEAT_CTL configured */ /* Virtualization flags: Linux defined, word 8 */ #define X86_FEATURE_TPR_SHADOW ( 8*32+ 0) /* Intel TPR Shadow */ @@ -357,6 +358,7 @@ /* Intel-defined CPU features, CPUID level 0x00000007:0 (EDX), word 18 */ #define X86_FEATURE_AVX512_4VNNIW (18*32+ 2) /* AVX-512 Neural Network Instructions */ #define X86_FEATURE_AVX512_4FMAPS (18*32+ 3) /* AVX-512 Multiply Accumulation Single precision */ +#define X86_FEATURE_FSRM (18*32+ 4) /* Fast Short Rep Mov */ #define X86_FEATURE_AVX512_VP2INTERSECT (18*32+ 8) /* AVX-512 Intersect for D/Q */ #define X86_FEATURE_MD_CLEAR (18*32+10) /* VERW clears CPU buffers */ #define X86_FEATURE_TSX_FORCE_ABORT (18*32+13) /* "" TSX_FORCE_ABORT */ diff --git a/arch/x86/include/asm/crypto/camellia.h b/arch/x86/include/asm/crypto/camellia.h index a5d86fc0593f..f6d91861cb14 100644 --- a/arch/x86/include/asm/crypto/camellia.h +++ b/arch/x86/include/asm/crypto/camellia.h @@ -26,71 +26,66 @@ struct camellia_xts_ctx { extern int __camellia_setkey(struct camellia_ctx *cctx, const unsigned char *key, - unsigned int key_len, u32 *flags); + unsigned int key_len); extern int xts_camellia_setkey(struct crypto_skcipher *tfm, const u8 *key, unsigned int keylen); /* regular block cipher functions */ -asmlinkage void __camellia_enc_blk(struct camellia_ctx *ctx, u8 *dst, - const u8 *src, bool xor); -asmlinkage void camellia_dec_blk(struct camellia_ctx *ctx, u8 *dst, - const u8 *src); +asmlinkage void __camellia_enc_blk(const void *ctx, u8 *dst, const u8 *src, + bool xor); +asmlinkage void camellia_dec_blk(const void *ctx, u8 *dst, const u8 *src); /* 2-way parallel cipher functions */ -asmlinkage void __camellia_enc_blk_2way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src, bool xor); -asmlinkage void camellia_dec_blk_2way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src); +asmlinkage void __camellia_enc_blk_2way(const void *ctx, u8 *dst, const u8 *src, + bool xor); +asmlinkage void camellia_dec_blk_2way(const void *ctx, u8 *dst, const u8 *src); /* 16-way parallel cipher functions (avx/aes-ni) */ -asmlinkage void camellia_ecb_enc_16way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src); -asmlinkage void camellia_ecb_dec_16way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src); - -asmlinkage void camellia_cbc_dec_16way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src); -asmlinkage void camellia_ctr_16way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); - -asmlinkage void camellia_xts_enc_16way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); -asmlinkage void camellia_xts_dec_16way(struct camellia_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); - -static inline void camellia_enc_blk(struct camellia_ctx *ctx, u8 *dst, - const u8 *src) +asmlinkage void camellia_ecb_enc_16way(const void *ctx, u8 *dst, const u8 *src); +asmlinkage void camellia_ecb_dec_16way(const void *ctx, u8 *dst, const u8 *src); + +asmlinkage void camellia_cbc_dec_16way(const void *ctx, u8 *dst, const u8 *src); +asmlinkage void camellia_ctr_16way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); + +asmlinkage void camellia_xts_enc_16way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); +asmlinkage void camellia_xts_dec_16way(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); + +static inline void camellia_enc_blk(const void *ctx, u8 *dst, const u8 *src) { __camellia_enc_blk(ctx, dst, src, false); } -static inline void camellia_enc_blk_xor(struct camellia_ctx *ctx, u8 *dst, - const u8 *src) +static inline void camellia_enc_blk_xor(const void *ctx, u8 *dst, const u8 *src) { __camellia_enc_blk(ctx, dst, src, true); } -static inline void camellia_enc_blk_2way(struct camellia_ctx *ctx, u8 *dst, +static inline void camellia_enc_blk_2way(const void *ctx, u8 *dst, const u8 *src) { __camellia_enc_blk_2way(ctx, dst, src, false); } -static inline void camellia_enc_blk_xor_2way(struct camellia_ctx *ctx, u8 *dst, +static inline void camellia_enc_blk_xor_2way(const void *ctx, u8 *dst, const u8 *src) { __camellia_enc_blk_2way(ctx, dst, src, true); } /* glue helpers */ -extern void camellia_decrypt_cbc_2way(void *ctx, u128 *dst, const u128 *src); -extern void camellia_crypt_ctr(void *ctx, u128 *dst, const u128 *src, +extern void camellia_decrypt_cbc_2way(const void *ctx, u8 *dst, const u8 *src); +extern void camellia_crypt_ctr(const void *ctx, u8 *dst, const u8 *src, le128 *iv); -extern void camellia_crypt_ctr_2way(void *ctx, u128 *dst, const u128 *src, +extern void camellia_crypt_ctr_2way(const void *ctx, u8 *dst, const u8 *src, le128 *iv); -extern void camellia_xts_enc(void *ctx, u128 *dst, const u128 *src, le128 *iv); -extern void camellia_xts_dec(void *ctx, u128 *dst, const u128 *src, le128 *iv); +extern void camellia_xts_enc(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); +extern void camellia_xts_dec(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); #endif /* ASM_X86_CAMELLIA_H */ diff --git a/arch/x86/include/asm/crypto/glue_helper.h b/arch/x86/include/asm/crypto/glue_helper.h index 8d4a8e1226ee..777c0f63418c 100644 --- a/arch/x86/include/asm/crypto/glue_helper.h +++ b/arch/x86/include/asm/crypto/glue_helper.h @@ -11,18 +11,13 @@ #include <asm/fpu/api.h> #include <crypto/b128ops.h> -typedef void (*common_glue_func_t)(void *ctx, u8 *dst, const u8 *src); -typedef void (*common_glue_cbc_func_t)(void *ctx, u128 *dst, const u128 *src); -typedef void (*common_glue_ctr_func_t)(void *ctx, u128 *dst, const u128 *src, +typedef void (*common_glue_func_t)(const void *ctx, u8 *dst, const u8 *src); +typedef void (*common_glue_cbc_func_t)(const void *ctx, u8 *dst, const u8 *src); +typedef void (*common_glue_ctr_func_t)(const void *ctx, u8 *dst, const u8 *src, le128 *iv); -typedef void (*common_glue_xts_func_t)(void *ctx, u128 *dst, const u128 *src, +typedef void (*common_glue_xts_func_t)(const void *ctx, u8 *dst, const u8 *src, le128 *iv); -#define GLUE_FUNC_CAST(fn) ((common_glue_func_t)(fn)) -#define GLUE_CBC_FUNC_CAST(fn) ((common_glue_cbc_func_t)(fn)) -#define GLUE_CTR_FUNC_CAST(fn) ((common_glue_ctr_func_t)(fn)) -#define GLUE_XTS_FUNC_CAST(fn) ((common_glue_xts_func_t)(fn)) - struct common_glue_func_entry { unsigned int num_blocks; /* number of blocks that @fn will process */ union { @@ -116,7 +111,8 @@ extern int glue_xts_req_128bit(const struct common_glue_ctx *gctx, common_glue_func_t tweak_fn, void *tweak_ctx, void *crypt_ctx, bool decrypt); -extern void glue_xts_crypt_128bit_one(void *ctx, u128 *dst, const u128 *src, - le128 *iv, common_glue_func_t fn); +extern void glue_xts_crypt_128bit_one(const void *ctx, u8 *dst, + const u8 *src, le128 *iv, + common_glue_func_t fn); #endif /* _CRYPTO_GLUE_HELPER_H */ diff --git a/arch/x86/include/asm/crypto/serpent-avx.h b/arch/x86/include/asm/crypto/serpent-avx.h index db7c9cc32234..251c2c89d7cf 100644 --- a/arch/x86/include/asm/crypto/serpent-avx.h +++ b/arch/x86/include/asm/crypto/serpent-avx.h @@ -15,26 +15,26 @@ struct serpent_xts_ctx { struct serpent_ctx crypt_ctx; }; -asmlinkage void serpent_ecb_enc_8way_avx(struct serpent_ctx *ctx, u8 *dst, +asmlinkage void serpent_ecb_enc_8way_avx(const void *ctx, u8 *dst, const u8 *src); -asmlinkage void serpent_ecb_dec_8way_avx(struct serpent_ctx *ctx, u8 *dst, +asmlinkage void serpent_ecb_dec_8way_avx(const void *ctx, u8 *dst, const u8 *src); -asmlinkage void serpent_cbc_dec_8way_avx(struct serpent_ctx *ctx, u8 *dst, +asmlinkage void serpent_cbc_dec_8way_avx(const void *ctx, u8 *dst, const u8 *src); -asmlinkage void serpent_ctr_8way_avx(struct serpent_ctx *ctx, u8 *dst, - const u8 *src, le128 *iv); +asmlinkage void serpent_ctr_8way_avx(const void *ctx, u8 *dst, const u8 *src, + le128 *iv); -asmlinkage void serpent_xts_enc_8way_avx(struct serpent_ctx *ctx, u8 *dst, +asmlinkage void serpent_xts_enc_8way_avx(const void *ctx, u8 *dst, const u8 *src, le128 *iv); -asmlinkage void serpent_xts_dec_8way_avx(struct serpent_ctx *ctx, u8 *dst, +asmlinkage void serpent_xts_dec_8way_avx(const void *ctx, u8 *dst, const u8 *src, le128 *iv); -extern void __serpent_crypt_ctr(void *ctx, u128 *dst, const u128 *src, +extern void __serpent_crypt_ctr(const void *ctx, u8 *dst, const u8 *src, le128 *iv); -extern void serpent_xts_enc(void *ctx, u128 *dst, const u128 *src, le128 *iv); -extern void serpent_xts_dec(void *ctx, u128 *dst, const u128 *src, le128 *iv); +extern void serpent_xts_enc(const void *ctx, u8 *dst, const u8 *src, le128 *iv); +extern void serpent_xts_dec(const void *ctx, u8 *dst, const u8 *src, le128 *iv); extern int xts_serpent_setkey(struct crypto_skcipher *tfm, const u8 *key, unsigned int keylen); diff --git a/arch/x86/include/asm/crypto/serpent-sse2.h b/arch/x86/include/asm/crypto/serpent-sse2.h index 1a345e8a7496..860ca248914b 100644 --- a/arch/x86/include/asm/crypto/serpent-sse2.h +++ b/arch/x86/include/asm/crypto/serpent-sse2.h @@ -9,25 +9,23 @@ #define SERPENT_PARALLEL_BLOCKS 4 -asmlinkage void __serpent_enc_blk_4way(struct serpent_ctx *ctx, u8 *dst, +asmlinkage void __serpent_enc_blk_4way(const struct serpent_ctx *ctx, u8 *dst, const u8 *src, bool xor); -asmlinkage void serpent_dec_blk_4way(struct serpent_ctx *ctx, u8 *dst, +asmlinkage void serpent_dec_blk_4way(const struct serpent_ctx *ctx, u8 *dst, const u8 *src); -static inline void serpent_enc_blk_xway(struct serpent_ctx *ctx, u8 *dst, - const u8 *src) +static inline void serpent_enc_blk_xway(const void *ctx, u8 *dst, const u8 *src) { __serpent_enc_blk_4way(ctx, dst, src, false); } -static inline void serpent_enc_blk_xway_xor(struct serpent_ctx *ctx, u8 *dst, - const u8 *src) +static inline void serpent_enc_blk_xway_xor(const struct serpent_ctx *ctx, + u8 *dst, const u8 *src) { __serpent_enc_blk_4way(ctx, dst, src, true); } -static inline void serpent_dec_blk_xway(struct serpent_ctx *ctx, u8 *dst, - const u8 *src) +static inline void serpent_dec_blk_xway(const void *ctx, u8 *dst, const u8 *src) { serpent_dec_blk_4way(ctx, dst, src); } @@ -36,25 +34,23 @@ static inline void serpent_dec_blk_xway(struct serpent_ctx *ctx, u8 *dst, #define SERPENT_PARALLEL_BLOCKS 8 -asmlinkage void __serpent_enc_blk_8way(struct serpent_ctx *ctx, u8 *dst, +asmlinkage void __serpent_enc_blk_8way(const struct serpent_ctx *ctx, u8 *dst, const u8 *src, bool xor); -asmlinkage void serpent_dec_blk_8way(struct serpent_ctx *ctx, u8 *dst, +asmlinkage void serpent_dec_blk_8way(const struct serpent_ctx *ctx, u8 *dst, const u8 *src); -static inline void serpent_enc_blk_xway(struct serpent_ctx *ctx, u8 *dst, - const u8 *src) +static inline void serpent_enc_blk_xway(const void *ctx, u8 *dst, const u8 *src) { __serpent_enc_blk_8way(ctx, dst, src, false); } -static inline void serpent_enc_blk_xway_xor(struct serpent_ctx *ctx, u8 *dst, - const u8 *src) +static inline void serpent_enc_blk_xway_xor(const struct serpent_ctx *ctx, + u8 *dst, const u8 *src) { __serpent_enc_blk_8way(ctx, dst, src, true); } -static inline void serpent_dec_blk_xway(struct serpent_ctx *ctx, u8 *dst, - const u8 *src) +static inline void serpent_dec_blk_xway(const void *ctx, u8 *dst, const u8 *src) { serpent_dec_blk_8way(ctx, dst, src); } diff --git a/arch/x86/include/asm/crypto/twofish.h b/arch/x86/include/asm/crypto/twofish.h index f618bf272b90..2c377a8042e1 100644 --- a/arch/x86/include/asm/crypto/twofish.h +++ b/arch/x86/include/asm/crypto/twofish.h @@ -7,22 +7,19 @@ #include <crypto/b128ops.h> /* regular block cipher functions from twofish_x86_64 module */ -asmlinkage void twofish_enc_blk(struct twofish_ctx *ctx, u8 *dst, - const u8 *src); -asmlinkage void twofish_dec_blk(struct twofish_ctx *ctx, u8 *dst, - const u8 *src); +asmlinkage void twofish_enc_blk(const void *ctx, u8 *dst, const u8 *src); +asmlinkage void twofish_dec_blk(const void *ctx, u8 *dst, const u8 *src); /* 3-way parallel cipher functions */ -asmlinkage void __twofish_enc_blk_3way(struct twofish_ctx *ctx, u8 *dst, - const u8 *src, bool xor); -asmlinkage void twofish_dec_blk_3way(struct twofish_ctx *ctx, u8 *dst, - const u8 *src); +asmlinkage void __twofish_enc_blk_3way(const void *ctx, u8 *dst, const u8 *src, + bool xor); +asmlinkage void twofish_dec_blk_3way(const void *ctx, u8 *dst, const u8 *src); /* helpers from twofish_x86_64-3way module */ -extern void twofish_dec_blk_cbc_3way(void *ctx, u128 *dst, const u128 *src); -extern void twofish_enc_blk_ctr(void *ctx, u128 *dst, const u128 *src, +extern void twofish_dec_blk_cbc_3way(const void *ctx, u8 *dst, const u8 *src); +extern void twofish_enc_blk_ctr(const void *ctx, u8 *dst, const u8 *src, le128 *iv); -extern void twofish_enc_blk_ctr_3way(void *ctx, u128 *dst, const u128 *src, +extern void twofish_enc_blk_ctr_3way(const void *ctx, u8 *dst, const u8 *src, le128 *iv); #endif /* ASM_X86_TWOFISH_H */ diff --git a/arch/x86/include/asm/efi.h b/arch/x86/include/asm/efi.h index d028e9acdf1c..86169a24b0d8 100644 --- a/arch/x86/include/asm/efi.h +++ b/arch/x86/include/asm/efi.h @@ -8,6 +8,7 @@ #include <asm/tlb.h> #include <asm/nospec-branch.h> #include <asm/mmu_context.h> +#include <linux/build_bug.h> /* * We map the EFI regions needed for runtime services non-contiguously, @@ -19,13 +20,16 @@ * This is the main reason why we're doing stable VA mappings for RT * services. * - * This flag is used in conjunction with a chicken bit called - * "efi=old_map" which can be used as a fallback to the old runtime - * services mapping method in case there's some b0rkage with a - * particular EFI implementation (haha, it is hard to hold up the - * sarcasm here...). + * SGI UV1 machines are known to be incompatible with this scheme, so we + * provide an opt-out for these machines via a DMI quirk that sets the + * attribute below. */ -#define EFI_OLD_MEMMAP EFI_ARCH_1 +#define EFI_UV1_MEMMAP EFI_ARCH_1 + +static inline bool efi_have_uv1_memmap(void) +{ + return IS_ENABLED(CONFIG_X86_UV) && efi_enabled(EFI_UV1_MEMMAP); +} #define EFI32_LOADER_SIGNATURE "EL32" #define EFI64_LOADER_SIGNATURE "EL64" @@ -34,10 +38,46 @@ #define ARCH_EFI_IRQ_FLAGS_MASK X86_EFLAGS_IF -#ifdef CONFIG_X86_32 +/* + * The EFI services are called through variadic functions in many cases. These + * functions are implemented in assembler and support only a fixed number of + * arguments. The macros below allows us to check at build time that we don't + * try to call them with too many arguments. + * + * __efi_nargs() will return the number of arguments if it is 7 or less, and + * cause a BUILD_BUG otherwise. The limitations of the C preprocessor make it + * impossible to calculate the exact number of arguments beyond some + * pre-defined limit. The maximum number of arguments currently supported by + * any of the thunks is 7, so this is good enough for now and can be extended + * in the obvious way if we ever need more. + */ -extern asmlinkage unsigned long efi_call_phys(void *, ...); +#define __efi_nargs(...) __efi_nargs_(__VA_ARGS__) +#define __efi_nargs_(...) __efi_nargs__(0, ##__VA_ARGS__, \ + __efi_arg_sentinel(7), __efi_arg_sentinel(6), \ + __efi_arg_sentinel(5), __efi_arg_sentinel(4), \ + __efi_arg_sentinel(3), __efi_arg_sentinel(2), \ + __efi_arg_sentinel(1), __efi_arg_sentinel(0)) +#define __efi_nargs__(_0, _1, _2, _3, _4, _5, _6, _7, n, ...) \ + __take_second_arg(n, \ + ({ BUILD_BUG_ON_MSG(1, "__efi_nargs limit exceeded"); 8; })) +#define __efi_arg_sentinel(n) , n +/* + * __efi_nargs_check(f, n, ...) will cause a BUILD_BUG if the ellipsis + * represents more than n arguments. + */ + +#define __efi_nargs_check(f, n, ...) \ + __efi_nargs_check_(f, __efi_nargs(__VA_ARGS__), n) +#define __efi_nargs_check_(f, p, n) __efi_nargs_check__(f, p, n) +#define __efi_nargs_check__(f, p, n) ({ \ + BUILD_BUG_ON_MSG( \ + (p) > (n), \ + #f " called with too many arguments (" #p ">" #n ")"); \ +}) + +#ifdef CONFIG_X86_32 #define arch_efi_call_virt_setup() \ ({ \ kernel_fpu_begin(); \ @@ -51,13 +91,7 @@ extern asmlinkage unsigned long efi_call_phys(void *, ...); }) -/* - * Wrap all the virtual calls in a way that forces the parameters on the stack. - */ -#define arch_efi_call_virt(p, f, args...) \ -({ \ - ((efi_##f##_t __attribute__((regparm(0)))*) p->f)(args); \ -}) +#define arch_efi_call_virt(p, f, args...) p->f(args) #define efi_ioremap(addr, size, type, attr) ioremap_cache(addr, size) @@ -65,9 +99,12 @@ extern asmlinkage unsigned long efi_call_phys(void *, ...); #define EFI_LOADER_SIGNATURE "EL64" -extern asmlinkage u64 efi_call(void *fp, ...); +extern asmlinkage u64 __efi_call(void *fp, ...); -#define efi_call_phys(f, args...) efi_call((f), args) +#define efi_call(...) ({ \ + __efi_nargs_check(efi_call, 7, __VA_ARGS__); \ + __efi_call(__VA_ARGS__); \ +}) /* * struct efi_scratch - Scratch space used while switching to/from efi_mm @@ -85,7 +122,7 @@ struct efi_scratch { kernel_fpu_begin(); \ firmware_restrict_branch_speculation_start(); \ \ - if (!efi_enabled(EFI_OLD_MEMMAP)) \ + if (!efi_have_uv1_memmap()) \ efi_switch_mm(&efi_mm); \ }) @@ -94,7 +131,7 @@ struct efi_scratch { #define arch_efi_call_virt_teardown() \ ({ \ - if (!efi_enabled(EFI_OLD_MEMMAP)) \ + if (!efi_have_uv1_memmap()) \ efi_switch_mm(efi_scratch.prev_mm); \ \ firmware_restrict_branch_speculation_end(); \ @@ -121,8 +158,6 @@ extern void __iomem *__init efi_ioremap(unsigned long addr, unsigned long size, extern struct efi_scratch efi_scratch; extern void __init efi_set_executable(efi_memory_desc_t *md, bool executable); extern int __init efi_memblock_x86_reserve_range(void); -extern pgd_t * __init efi_call_phys_prolog(void); -extern void __init efi_call_phys_epilog(pgd_t *save_pgd); extern void __init efi_print_memmap(void); extern void __init efi_memory_uc(u64 addr, unsigned long size); extern void __init efi_map_region(efi_memory_desc_t *md); @@ -140,6 +175,8 @@ extern void efi_delete_dummy_variable(void); extern void efi_switch_mm(struct mm_struct *mm); extern void efi_recover_from_page_fault(unsigned long phys_addr); extern void efi_free_boot_services(void); +extern pgd_t * __init efi_uv1_memmap_phys_prolog(void); +extern void __init efi_uv1_memmap_phys_epilog(pgd_t *save_pgd); struct efi_setup_data { u64 fw_vendor; @@ -152,93 +189,144 @@ struct efi_setup_data { extern u64 efi_setup; #ifdef CONFIG_EFI +extern efi_status_t __efi64_thunk(u32, ...); -static inline bool efi_is_native(void) +#define efi64_thunk(...) ({ \ + __efi_nargs_check(efi64_thunk, 6, __VA_ARGS__); \ + __efi64_thunk(__VA_ARGS__); \ +}) + +static inline bool efi_is_mixed(void) { - return IS_ENABLED(CONFIG_X86_64) == efi_enabled(EFI_64BIT); + if (!IS_ENABLED(CONFIG_EFI_MIXED)) + return false; + return IS_ENABLED(CONFIG_X86_64) && !efi_enabled(EFI_64BIT); } static inline bool efi_runtime_supported(void) { - if (efi_is_native()) - return true; - - if (IS_ENABLED(CONFIG_EFI_MIXED) && !efi_enabled(EFI_OLD_MEMMAP)) + if (IS_ENABLED(CONFIG_X86_64) == efi_enabled(EFI_64BIT)) return true; - return false; + return IS_ENABLED(CONFIG_EFI_MIXED); } extern void parse_efi_setup(u64 phys_addr, u32 data_len); extern void efifb_setup_from_dmi(struct screen_info *si, const char *opt); -#ifdef CONFIG_EFI_MIXED extern void efi_thunk_runtime_setup(void); -extern efi_status_t efi_thunk_set_virtual_address_map( - void *phys_set_virtual_address_map, - unsigned long memory_map_size, - unsigned long descriptor_size, - u32 descriptor_version, - efi_memory_desc_t *virtual_map); -#else -static inline void efi_thunk_runtime_setup(void) {} -static inline efi_status_t efi_thunk_set_virtual_address_map( - void *phys_set_virtual_address_map, - unsigned long memory_map_size, - unsigned long descriptor_size, - u32 descriptor_version, - efi_memory_desc_t *virtual_map) -{ - return EFI_SUCCESS; -} -#endif /* CONFIG_EFI_MIXED */ - +efi_status_t efi_set_virtual_address_map(unsigned long memory_map_size, + unsigned long descriptor_size, + u32 descriptor_version, + efi_memory_desc_t *virtual_map); /* arch specific definitions used by the stub code */ -struct efi_config { - u64 image_handle; - u64 table; - u64 runtime_services; - u64 boot_services; - u64 text_output; - efi_status_t (*call)(unsigned long, ...); - bool is64; -} __packed; - -__pure const struct efi_config *__efi_early(void); +__attribute_const__ bool efi_is_64bit(void); -static inline bool efi_is_64bit(void) +static inline bool efi_is_native(void) { if (!IS_ENABLED(CONFIG_X86_64)) - return false; - + return true; if (!IS_ENABLED(CONFIG_EFI_MIXED)) return true; + return efi_is_64bit(); +} + +#define efi_mixed_mode_cast(attr) \ + __builtin_choose_expr( \ + __builtin_types_compatible_p(u32, __typeof__(attr)), \ + (unsigned long)(attr), (attr)) - return __efi_early()->is64; +#define efi_table_attr(inst, attr) \ + (efi_is_native() \ + ? inst->attr \ + : (__typeof__(inst->attr)) \ + efi_mixed_mode_cast(inst->mixed_mode.attr)) + +/* + * The following macros allow translating arguments if necessary from native to + * mixed mode. The use case for this is to initialize the upper 32 bits of + * output parameters, and where the 32-bit method requires a 64-bit argument, + * which must be split up into two arguments to be thunked properly. + * + * As examples, the AllocatePool boot service returns the address of the + * allocation, but it will not set the high 32 bits of the address. To ensure + * that the full 64-bit address is initialized, we zero-init the address before + * calling the thunk. + * + * The FreePages boot service takes a 64-bit physical address even in 32-bit + * mode. For the thunk to work correctly, a native 64-bit call of + * free_pages(addr, size) + * must be translated to + * efi64_thunk(free_pages, addr & U32_MAX, addr >> 32, size) + * so that the two 32-bit halves of addr get pushed onto the stack separately. + */ + +static inline void *efi64_zero_upper(void *p) +{ + ((u32 *)p)[1] = 0; + return p; } -#define efi_table_attr(table, attr, instance) \ - (efi_is_64bit() ? \ - ((table##_64_t *)(unsigned long)instance)->attr : \ - ((table##_32_t *)(unsigned long)instance)->attr) +#define __efi64_argmap_free_pages(addr, size) \ + ((addr), 0, (size)) + +#define __efi64_argmap_get_memory_map(mm_size, mm, key, size, ver) \ + ((mm_size), (mm), efi64_zero_upper(key), efi64_zero_upper(size), (ver)) + +#define __efi64_argmap_allocate_pool(type, size, buffer) \ + ((type), (size), efi64_zero_upper(buffer)) -#define efi_call_proto(protocol, f, instance, ...) \ - __efi_early()->call(efi_table_attr(protocol, f, instance), \ - instance, ##__VA_ARGS__) +#define __efi64_argmap_handle_protocol(handle, protocol, interface) \ + ((handle), (protocol), efi64_zero_upper(interface)) -#define efi_call_early(f, ...) \ - __efi_early()->call(efi_table_attr(efi_boot_services, f, \ - __efi_early()->boot_services), __VA_ARGS__) +#define __efi64_argmap_locate_protocol(protocol, reg, interface) \ + ((protocol), (reg), efi64_zero_upper(interface)) -#define __efi_call_early(f, ...) \ - __efi_early()->call((unsigned long)f, __VA_ARGS__); +/* PCI I/O */ +#define __efi64_argmap_get_location(protocol, seg, bus, dev, func) \ + ((protocol), efi64_zero_upper(seg), efi64_zero_upper(bus), \ + efi64_zero_upper(dev), efi64_zero_upper(func)) + +/* + * The macros below handle the plumbing for the argument mapping. To add a + * mapping for a specific EFI method, simply define a macro + * __efi64_argmap_<method name>, following the examples above. + */ -#define efi_call_runtime(f, ...) \ - __efi_early()->call(efi_table_attr(efi_runtime_services, f, \ - __efi_early()->runtime_services), __VA_ARGS__) +#define __efi64_thunk_map(inst, func, ...) \ + efi64_thunk(inst->mixed_mode.func, \ + __efi64_argmap(__efi64_argmap_ ## func(__VA_ARGS__), \ + (__VA_ARGS__))) + +#define __efi64_argmap(mapped, args) \ + __PASTE(__efi64_argmap__, __efi_nargs(__efi_eat mapped))(mapped, args) +#define __efi64_argmap__0(mapped, args) __efi_eval mapped +#define __efi64_argmap__1(mapped, args) __efi_eval args + +#define __efi_eat(...) +#define __efi_eval(...) __VA_ARGS__ + +/* The three macros below handle dispatching via the thunk if needed */ + +#define efi_call_proto(inst, func, ...) \ + (efi_is_native() \ + ? inst->func(inst, ##__VA_ARGS__) \ + : __efi64_thunk_map(inst, func, inst, ##__VA_ARGS__)) + +#define efi_bs_call(func, ...) \ + (efi_is_native() \ + ? efi_system_table()->boottime->func(__VA_ARGS__) \ + : __efi64_thunk_map(efi_table_attr(efi_system_table(), \ + boottime), func, __VA_ARGS__)) + +#define efi_rt_call(func, ...) \ + (efi_is_native() \ + ? efi_system_table()->runtime->func(__VA_ARGS__) \ + : __efi64_thunk_map(efi_table_attr(efi_system_table(), \ + runtime), func, __VA_ARGS__)) extern bool efi_reboot_required(void); extern bool efi_is_table_address(unsigned long phys_addr); diff --git a/arch/x86/include/asm/ftrace.h b/arch/x86/include/asm/ftrace.h index c2a7458f912c..85be2f506272 100644 --- a/arch/x86/include/asm/ftrace.h +++ b/arch/x86/include/asm/ftrace.h @@ -47,8 +47,6 @@ struct dyn_arch_ftrace { /* No extra data needed for x86 */ }; -int ftrace_int3_handler(struct pt_regs *regs); - #define FTRACE_GRAPH_TRAMP_ADDR FTRACE_GRAPH_ADDR #endif /* CONFIG_DYNAMIC_FTRACE */ diff --git a/arch/x86/include/asm/intel-family.h b/arch/x86/include/asm/intel-family.h index c606c0b70738..4981c293f926 100644 --- a/arch/x86/include/asm/intel-family.h +++ b/arch/x86/include/asm/intel-family.h @@ -111,6 +111,7 @@ #define INTEL_FAM6_ATOM_TREMONT_D 0x86 /* Jacobsville */ #define INTEL_FAM6_ATOM_TREMONT 0x96 /* Elkhart Lake */ +#define INTEL_FAM6_ATOM_TREMONT_L 0x9C /* Jasper Lake */ /* Xeon Phi */ diff --git a/arch/x86/include/asm/intel_pmc_ipc.h b/arch/x86/include/asm/intel_pmc_ipc.h index 9e7adcdbe031..e6da1ce26256 100644 --- a/arch/x86/include/asm/intel_pmc_ipc.h +++ b/arch/x86/include/asm/intel_pmc_ipc.h @@ -31,30 +31,13 @@ #if IS_ENABLED(CONFIG_INTEL_PMC_IPC) -int intel_pmc_ipc_simple_command(int cmd, int sub); -int intel_pmc_ipc_raw_cmd(u32 cmd, u32 sub, u8 *in, u32 inlen, - u32 *out, u32 outlen, u32 dptr, u32 sptr); int intel_pmc_ipc_command(u32 cmd, u32 sub, u8 *in, u32 inlen, u32 *out, u32 outlen); int intel_pmc_s0ix_counter_read(u64 *data); -int intel_pmc_gcr_read(u32 offset, u32 *data); int intel_pmc_gcr_read64(u32 offset, u64 *data); -int intel_pmc_gcr_write(u32 offset, u32 data); -int intel_pmc_gcr_update(u32 offset, u32 mask, u32 val); #else -static inline int intel_pmc_ipc_simple_command(int cmd, int sub) -{ - return -EINVAL; -} - -static inline int intel_pmc_ipc_raw_cmd(u32 cmd, u32 sub, u8 *in, u32 inlen, - u32 *out, u32 outlen, u32 dptr, u32 sptr) -{ - return -EINVAL; -} - static inline int intel_pmc_ipc_command(u32 cmd, u32 sub, u8 *in, u32 inlen, u32 *out, u32 outlen) { @@ -66,26 +49,11 @@ static inline int intel_pmc_s0ix_counter_read(u64 *data) return -EINVAL; } -static inline int intel_pmc_gcr_read(u32 offset, u32 *data) -{ - return -EINVAL; -} - static inline int intel_pmc_gcr_read64(u32 offset, u64 *data) { return -EINVAL; } -static inline int intel_pmc_gcr_write(u32 offset, u32 data) -{ - return -EINVAL; -} - -static inline int intel_pmc_gcr_update(u32 offset, u32 mask, u32 val) -{ - return -EINVAL; -} - #endif /*CONFIG_INTEL_PMC_IPC*/ #endif diff --git a/arch/x86/include/asm/intel_scu_ipc.h b/arch/x86/include/asm/intel_scu_ipc.h index 4a8c6e817398..2a1442ba6e78 100644 --- a/arch/x86/include/asm/intel_scu_ipc.h +++ b/arch/x86/include/asm/intel_scu_ipc.h @@ -22,24 +22,12 @@ /* Read single register */ int intel_scu_ipc_ioread8(u16 addr, u8 *data); -/* Read two sequential registers */ -int intel_scu_ipc_ioread16(u16 addr, u16 *data); - -/* Read four sequential registers */ -int intel_scu_ipc_ioread32(u16 addr, u32 *data); - /* Read a vector */ int intel_scu_ipc_readv(u16 *addr, u8 *data, int len); /* Write single register */ int intel_scu_ipc_iowrite8(u16 addr, u8 data); -/* Write two sequential registers */ -int intel_scu_ipc_iowrite16(u16 addr, u16 data); - -/* Write four sequential registers */ -int intel_scu_ipc_iowrite32(u16 addr, u32 data); - /* Write a vector */ int intel_scu_ipc_writev(u16 *addr, u8 *data, int len); @@ -50,14 +38,6 @@ int intel_scu_ipc_update_register(u16 addr, u8 data, u8 mask); int intel_scu_ipc_simple_command(int cmd, int sub); int intel_scu_ipc_command(int cmd, int sub, u32 *in, int inlen, u32 *out, int outlen); -int intel_scu_ipc_raw_command(int cmd, int sub, u8 *in, int inlen, - u32 *out, int outlen, u32 dptr, u32 sptr); - -/* I2C control api */ -int intel_scu_ipc_i2c_cntrl(u32 addr, u32 *data); - -/* Update FW version */ -int intel_scu_ipc_fw_update(u8 *buffer, u32 length); extern struct blocking_notifier_head intel_scu_notifier; diff --git a/arch/x86/include/asm/intel_telemetry.h b/arch/x86/include/asm/intel_telemetry.h index 214394860632..2f77e31a1283 100644 --- a/arch/x86/include/asm/intel_telemetry.h +++ b/arch/x86/include/asm/intel_telemetry.h @@ -40,13 +40,10 @@ struct telemetry_evtmap { struct telemetry_unit_config { struct telemetry_evtmap *telem_evts; void __iomem *regmap; - u32 ssram_base_addr; u8 ssram_evts_used; u8 curr_period; u8 max_period; u8 min_period; - u32 ssram_size; - }; struct telemetry_plt_config { diff --git a/arch/x86/include/asm/io.h b/arch/x86/include/asm/io.h index 9997521fc5cd..e1aa17a468a8 100644 --- a/arch/x86/include/asm/io.h +++ b/arch/x86/include/asm/io.h @@ -399,4 +399,40 @@ extern bool arch_memremap_can_ram_remap(resource_size_t offset, extern bool phys_mem_access_encrypted(unsigned long phys_addr, unsigned long size); +/** + * iosubmit_cmds512 - copy data to single MMIO location, in 512-bit units + * @__dst: destination, in MMIO space (must be 512-bit aligned) + * @src: source + * @count: number of 512 bits quantities to submit + * + * Submit data from kernel space to MMIO space, in units of 512 bits at a + * time. Order of access is not guaranteed, nor is a memory barrier + * performed afterwards. + * + * Warning: Do not use this helper unless your driver has checked that the CPU + * instruction is supported on the platform. + */ +static inline void iosubmit_cmds512(void __iomem *__dst, const void *src, + size_t count) +{ + /* + * Note that this isn't an "on-stack copy", just definition of "dst" + * as a pointer to 64-bytes of stuff that is going to be overwritten. + * In the MOVDIR64B case that may be needed as you can use the + * MOVDIR64B instruction to copy arbitrary memory around. This trick + * lets the compiler know how much gets clobbered. + */ + volatile struct { char _[64]; } *dst = __dst; + const u8 *from = src; + const u8 *end = from + count * 64; + + while (from < end) { + /* MOVDIR64B [rdx], rax */ + asm volatile(".byte 0x66, 0x0f, 0x38, 0xf8, 0x02" + : "=m" (dst) + : "d" (from), "a" (dst)); + from += 64; + } +} + #endif /* _ASM_X86_IO_H */ diff --git a/arch/x86/include/asm/kdebug.h b/arch/x86/include/asm/kdebug.h index 75f1e35e7c15..247ab14c6309 100644 --- a/arch/x86/include/asm/kdebug.h +++ b/arch/x86/include/asm/kdebug.h @@ -33,6 +33,7 @@ enum show_regs_mode { }; extern void die(const char *, struct pt_regs *,long); +void die_addr(const char *str, struct pt_regs *regs, long err, long gp_addr); extern int __must_check __die(const char *, struct pt_regs *, long); extern void show_stack_regs(struct pt_regs *regs); extern void __show_regs(struct pt_regs *regs, enum show_regs_mode); diff --git a/arch/x86/include/asm/kprobes.h b/arch/x86/include/asm/kprobes.h index 5dc909d9ad81..95b1f053bd96 100644 --- a/arch/x86/include/asm/kprobes.h +++ b/arch/x86/include/asm/kprobes.h @@ -11,12 +11,11 @@ #include <asm-generic/kprobes.h> -#define BREAKPOINT_INSTRUCTION 0xcc - #ifdef CONFIG_KPROBES #include <linux/types.h> #include <linux/ptrace.h> #include <linux/percpu.h> +#include <asm/text-patching.h> #include <asm/insn.h> #define __ARCH_WANT_KPROBES_INSN_SLOT @@ -25,10 +24,7 @@ struct pt_regs; struct kprobe; typedef u8 kprobe_opcode_t; -#define RELATIVEJUMP_OPCODE 0xe9 -#define RELATIVEJUMP_SIZE 5 -#define RELATIVECALL_OPCODE 0xe8 -#define RELATIVE_ADDR_SIZE 4 + #define MAX_STACK_SIZE 64 #define CUR_STACK_SIZE(ADDR) \ (current_top_of_stack() - (unsigned long)(ADDR)) @@ -43,11 +39,11 @@ extern __visible kprobe_opcode_t optprobe_template_entry[]; extern __visible kprobe_opcode_t optprobe_template_val[]; extern __visible kprobe_opcode_t optprobe_template_call[]; extern __visible kprobe_opcode_t optprobe_template_end[]; -#define MAX_OPTIMIZED_LENGTH (MAX_INSN_SIZE + RELATIVE_ADDR_SIZE) +#define MAX_OPTIMIZED_LENGTH (MAX_INSN_SIZE + DISP32_SIZE) #define MAX_OPTINSN_SIZE \ (((unsigned long)optprobe_template_end - \ (unsigned long)optprobe_template_entry) + \ - MAX_OPTIMIZED_LENGTH + RELATIVEJUMP_SIZE) + MAX_OPTIMIZED_LENGTH + JMP32_INSN_SIZE) extern const int kretprobe_blacklist_size; @@ -73,7 +69,7 @@ struct arch_specific_insn { struct arch_optimized_insn { /* copy of the original instructions */ - kprobe_opcode_t copied_insn[RELATIVE_ADDR_SIZE]; + kprobe_opcode_t copied_insn[DISP32_SIZE]; /* detour code buffer */ kprobe_opcode_t *insn; /* the size of instructions copied to detour code buffer */ diff --git a/arch/x86/include/asm/mce.h b/arch/x86/include/asm/mce.h index dc2d4b206ab7..4359b955e0b7 100644 --- a/arch/x86/include/asm/mce.h +++ b/arch/x86/include/asm/mce.h @@ -144,7 +144,7 @@ struct mce_log_buffer { enum mce_notifier_prios { MCE_PRIO_FIRST = INT_MAX, - MCE_PRIO_SRAO = INT_MAX - 1, + MCE_PRIO_UC = INT_MAX - 1, MCE_PRIO_EXTLOG = INT_MAX - 2, MCE_PRIO_NFIT = INT_MAX - 3, MCE_PRIO_EDAC = INT_MAX - 4, @@ -290,6 +290,7 @@ extern void apei_mce_report_mem_error(int corrected, /* These may be used by multiple smca_hwid_mcatypes */ enum smca_bank_types { SMCA_LS = 0, /* Load Store */ + SMCA_LS_V2, /* Load Store */ SMCA_IF, /* Instruction Fetch */ SMCA_L2_CACHE, /* L2 Cache */ SMCA_DE, /* Decoder Unit */ diff --git a/arch/x86/include/asm/memtype.h b/arch/x86/include/asm/memtype.h new file mode 100644 index 000000000000..9c2447b3555d --- /dev/null +++ b/arch/x86/include/asm/memtype.h @@ -0,0 +1,27 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _ASM_X86_MEMTYPE_H +#define _ASM_X86_MEMTYPE_H + +#include <linux/types.h> +#include <asm/pgtable_types.h> + +extern bool pat_enabled(void); +extern void pat_disable(const char *reason); +extern void pat_init(void); +extern void init_cache_modes(void); + +extern int memtype_reserve(u64 start, u64 end, + enum page_cache_mode req_pcm, enum page_cache_mode *ret_pcm); +extern int memtype_free(u64 start, u64 end); + +extern int memtype_kernel_map_sync(u64 base, unsigned long size, + enum page_cache_mode pcm); + +extern int memtype_reserve_io(resource_size_t start, resource_size_t end, + enum page_cache_mode *pcm); + +extern void memtype_free_io(resource_size_t start, resource_size_t end); + +extern bool pat_pfn_immune_to_uc_mtrr(unsigned long pfn); + +#endif /* _ASM_X86_MEMTYPE_H */ diff --git a/arch/x86/include/asm/microcode_amd.h b/arch/x86/include/asm/microcode_amd.h index 209492849566..6685e1218959 100644 --- a/arch/x86/include/asm/microcode_amd.h +++ b/arch/x86/include/asm/microcode_amd.h @@ -53,6 +53,6 @@ static inline void __init load_ucode_amd_bsp(unsigned int family) {} static inline void load_ucode_amd_ap(unsigned int family) {} static inline int __init save_microcode_in_initrd_amd(unsigned int family) { return -EINVAL; } -void reload_ucode_amd(void) {} +static inline void reload_ucode_amd(void) {} #endif #endif /* _ASM_X86_MICROCODE_AMD_H */ diff --git a/arch/x86/include/asm/mmu_context.h b/arch/x86/include/asm/mmu_context.h index 5f33924e200f..b243234e90cb 100644 --- a/arch/x86/include/asm/mmu_context.h +++ b/arch/x86/include/asm/mmu_context.h @@ -69,14 +69,6 @@ struct ldt_struct { int slot; }; -/* This is a multiple of PAGE_SIZE. */ -#define LDT_SLOT_STRIDE (LDT_ENTRIES * LDT_ENTRY_SIZE) - -static inline void *ldt_slot_va(int slot) -{ - return (void *)(LDT_BASE_ADDR + LDT_SLOT_STRIDE * slot); -} - /* * Used for LDT copy/destruction. */ @@ -99,87 +91,21 @@ static inline void destroy_context_ldt(struct mm_struct *mm) { } static inline void ldt_arch_exit_mmap(struct mm_struct *mm) { } #endif -static inline void load_mm_ldt(struct mm_struct *mm) -{ #ifdef CONFIG_MODIFY_LDT_SYSCALL - struct ldt_struct *ldt; - - /* READ_ONCE synchronizes with smp_store_release */ - ldt = READ_ONCE(mm->context.ldt); - - /* - * Any change to mm->context.ldt is followed by an IPI to all - * CPUs with the mm active. The LDT will not be freed until - * after the IPI is handled by all such CPUs. This means that, - * if the ldt_struct changes before we return, the values we see - * will be safe, and the new values will be loaded before we run - * any user code. - * - * NB: don't try to convert this to use RCU without extreme care. - * We would still need IRQs off, because we don't want to change - * the local LDT after an IPI loaded a newer value than the one - * that we can see. - */ - - if (unlikely(ldt)) { - if (static_cpu_has(X86_FEATURE_PTI)) { - if (WARN_ON_ONCE((unsigned long)ldt->slot > 1)) { - /* - * Whoops -- either the new LDT isn't mapped - * (if slot == -1) or is mapped into a bogus - * slot (if slot > 1). - */ - clear_LDT(); - return; - } - - /* - * If page table isolation is enabled, ldt->entries - * will not be mapped in the userspace pagetables. - * Tell the CPU to access the LDT through the alias - * at ldt_slot_va(ldt->slot). - */ - set_ldt(ldt_slot_va(ldt->slot), ldt->nr_entries); - } else { - set_ldt(ldt->entries, ldt->nr_entries); - } - } else { - clear_LDT(); - } +extern void load_mm_ldt(struct mm_struct *mm); +extern void switch_ldt(struct mm_struct *prev, struct mm_struct *next); #else +static inline void load_mm_ldt(struct mm_struct *mm) +{ clear_LDT(); -#endif } - static inline void switch_ldt(struct mm_struct *prev, struct mm_struct *next) { -#ifdef CONFIG_MODIFY_LDT_SYSCALL - /* - * Load the LDT if either the old or new mm had an LDT. - * - * An mm will never go from having an LDT to not having an LDT. Two - * mms never share an LDT, so we don't gain anything by checking to - * see whether the LDT changed. There's also no guarantee that - * prev->context.ldt actually matches LDTR, but, if LDTR is non-NULL, - * then prev->context.ldt will also be non-NULL. - * - * If we really cared, we could optimize the case where prev == next - * and we're exiting lazy mode. Most of the time, if this happens, - * we don't actually need to reload LDTR, but modify_ldt() is mostly - * used by legacy code and emulators where we don't need this level of - * performance. - * - * This uses | instead of || because it generates better code. - */ - if (unlikely((unsigned long)prev->context.ldt | - (unsigned long)next->context.ldt)) - load_mm_ldt(next); -#endif - DEBUG_LOCKS_WARN_ON(preemptible()); } +#endif -void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk); +extern void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk); /* * Init a new mm. Used on mm copies, like at fork() diff --git a/arch/x86/include/asm/msr-index.h b/arch/x86/include/asm/msr-index.h index 084e98da04a7..ebe1685e92dd 100644 --- a/arch/x86/include/asm/msr-index.h +++ b/arch/x86/include/asm/msr-index.h @@ -558,7 +558,14 @@ #define MSR_IA32_EBL_CR_POWERON 0x0000002a #define MSR_EBC_FREQUENCY_ID 0x0000002c #define MSR_SMI_COUNT 0x00000034 -#define MSR_IA32_FEATURE_CONTROL 0x0000003a + +/* Referred to as IA32_FEATURE_CONTROL in Intel's SDM. */ +#define MSR_IA32_FEAT_CTL 0x0000003a +#define FEAT_CTL_LOCKED BIT(0) +#define FEAT_CTL_VMX_ENABLED_INSIDE_SMX BIT(1) +#define FEAT_CTL_VMX_ENABLED_OUTSIDE_SMX BIT(2) +#define FEAT_CTL_LMCE_ENABLED BIT(20) + #define MSR_IA32_TSC_ADJUST 0x0000003b #define MSR_IA32_BNDCFGS 0x00000d90 @@ -566,11 +573,6 @@ #define MSR_IA32_XSS 0x00000da0 -#define FEATURE_CONTROL_LOCKED (1<<0) -#define FEATURE_CONTROL_VMXON_ENABLED_INSIDE_SMX (1<<1) -#define FEATURE_CONTROL_VMXON_ENABLED_OUTSIDE_SMX (1<<2) -#define FEATURE_CONTROL_LMCE (1<<20) - #define MSR_IA32_APICBASE 0x0000001b #define MSR_IA32_APICBASE_BSP (1<<8) #define MSR_IA32_APICBASE_ENABLE (1<<11) diff --git a/arch/x86/include/asm/mtrr.h b/arch/x86/include/asm/mtrr.h index dbff1456d215..829df26fd7a3 100644 --- a/arch/x86/include/asm/mtrr.h +++ b/arch/x86/include/asm/mtrr.h @@ -24,7 +24,7 @@ #define _ASM_X86_MTRR_H #include <uapi/asm/mtrr.h> -#include <asm/pat.h> +#include <asm/memtype.h> /* @@ -86,7 +86,7 @@ static inline void mtrr_centaur_report_mcr(int mcr, u32 lo, u32 hi) } static inline void mtrr_bp_init(void) { - pat_disable("MTRRs disabled, skipping PAT initialization too."); + pat_disable("PAT support disabled because CONFIG_MTRR is disabled in the kernel."); } #define mtrr_ap_init() do {} while (0) diff --git a/arch/x86/include/asm/nmi.h b/arch/x86/include/asm/nmi.h index 75ded1d13d98..9d5d949e662e 100644 --- a/arch/x86/include/asm/nmi.h +++ b/arch/x86/include/asm/nmi.h @@ -41,7 +41,6 @@ struct nmiaction { struct list_head list; nmi_handler_t handler; u64 max_duration; - struct irq_work irq_work; unsigned long flags; const char *name; }; diff --git a/arch/x86/include/asm/nospec-branch.h b/arch/x86/include/asm/nospec-branch.h index 5c24a7b35166..07e95dcb40ad 100644 --- a/arch/x86/include/asm/nospec-branch.h +++ b/arch/x86/include/asm/nospec-branch.h @@ -37,7 +37,6 @@ */ #define RSB_CLEAR_LOOPS 32 /* To forcibly overwrite all entries */ -#define RSB_FILL_LOOPS 16 /* To avoid underflow */ /* * Google experimented with loop-unrolling and this turned out to be diff --git a/arch/x86/include/asm/pat.h b/arch/x86/include/asm/pat.h deleted file mode 100644 index 92015c65fa2a..000000000000 --- a/arch/x86/include/asm/pat.h +++ /dev/null @@ -1,27 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _ASM_X86_PAT_H -#define _ASM_X86_PAT_H - -#include <linux/types.h> -#include <asm/pgtable_types.h> - -bool pat_enabled(void); -void pat_disable(const char *reason); -extern void pat_init(void); -extern void init_cache_modes(void); - -extern int reserve_memtype(u64 start, u64 end, - enum page_cache_mode req_pcm, enum page_cache_mode *ret_pcm); -extern int free_memtype(u64 start, u64 end); - -extern int kernel_map_sync_memtype(u64 base, unsigned long size, - enum page_cache_mode pcm); - -int io_reserve_memtype(resource_size_t start, resource_size_t end, - enum page_cache_mode *pcm); - -void io_free_memtype(resource_size_t start, resource_size_t end); - -bool pat_pfn_immune_to_uc_mtrr(unsigned long pfn); - -#endif /* _ASM_X86_PAT_H */ diff --git a/arch/x86/include/asm/pci.h b/arch/x86/include/asm/pci.h index 90d0731fdcb6..c1fdd43fe187 100644 --- a/arch/x86/include/asm/pci.h +++ b/arch/x86/include/asm/pci.h @@ -9,7 +9,7 @@ #include <linux/scatterlist.h> #include <linux/numa.h> #include <asm/io.h> -#include <asm/pat.h> +#include <asm/memtype.h> #include <asm/x86_init.h> struct pci_sysdata { diff --git a/arch/x86/include/asm/perf_event.h b/arch/x86/include/asm/perf_event.h index ee26e9215f18..29964b0e1075 100644 --- a/arch/x86/include/asm/perf_event.h +++ b/arch/x86/include/asm/perf_event.h @@ -322,17 +322,10 @@ struct perf_guest_switch_msr { u64 host, guest; }; -extern struct perf_guest_switch_msr *perf_guest_get_msrs(int *nr); extern void perf_get_x86_pmu_capability(struct x86_pmu_capability *cap); extern void perf_check_microcode(void); extern int x86_perf_rdpmc_index(struct perf_event *event); #else -static inline struct perf_guest_switch_msr *perf_guest_get_msrs(int *nr) -{ - *nr = 0; - return NULL; -} - static inline void perf_get_x86_pmu_capability(struct x86_pmu_capability *cap) { memset(cap, 0, sizeof(*cap)); @@ -342,8 +335,23 @@ static inline void perf_events_lapic_init(void) { } static inline void perf_check_microcode(void) { } #endif +#if defined(CONFIG_PERF_EVENTS) && defined(CONFIG_CPU_SUP_INTEL) +extern struct perf_guest_switch_msr *perf_guest_get_msrs(int *nr); +#else +static inline struct perf_guest_switch_msr *perf_guest_get_msrs(int *nr) +{ + *nr = 0; + return NULL; +} +#endif + #ifdef CONFIG_CPU_SUP_INTEL extern void intel_pt_handle_vmx(int on); +#else +static inline void intel_pt_handle_vmx(int on) +{ + +} #endif #if defined(CONFIG_PERF_EVENTS) && defined(CONFIG_CPU_SUP_AMD) diff --git a/arch/x86/include/asm/pgtable_32_areas.h b/arch/x86/include/asm/pgtable_32_areas.h new file mode 100644 index 000000000000..b6355416a15a --- /dev/null +++ b/arch/x86/include/asm/pgtable_32_areas.h @@ -0,0 +1,53 @@ +#ifndef _ASM_X86_PGTABLE_32_AREAS_H +#define _ASM_X86_PGTABLE_32_AREAS_H + +#include <asm/cpu_entry_area.h> + +/* + * Just any arbitrary offset to the start of the vmalloc VM area: the + * current 8MB value just means that there will be a 8MB "hole" after the + * physical memory until the kernel virtual memory starts. That means that + * any out-of-bounds memory accesses will hopefully be caught. + * The vmalloc() routines leaves a hole of 4kB between each vmalloced + * area for the same reason. ;) + */ +#define VMALLOC_OFFSET (8 * 1024 * 1024) + +#ifndef __ASSEMBLY__ +extern bool __vmalloc_start_set; /* set once high_memory is set */ +#endif + +#define VMALLOC_START ((unsigned long)high_memory + VMALLOC_OFFSET) +#ifdef CONFIG_X86_PAE +#define LAST_PKMAP 512 +#else +#define LAST_PKMAP 1024 +#endif + +#define CPU_ENTRY_AREA_PAGES (NR_CPUS * DIV_ROUND_UP(sizeof(struct cpu_entry_area), PAGE_SIZE)) + +/* The +1 is for the readonly IDT page: */ +#define CPU_ENTRY_AREA_BASE \ + ((FIXADDR_TOT_START - PAGE_SIZE*(CPU_ENTRY_AREA_PAGES+1)) & PMD_MASK) + +#define LDT_BASE_ADDR \ + ((CPU_ENTRY_AREA_BASE - PAGE_SIZE) & PMD_MASK) + +#define LDT_END_ADDR (LDT_BASE_ADDR + PMD_SIZE) + +#define PKMAP_BASE \ + ((LDT_BASE_ADDR - PAGE_SIZE) & PMD_MASK) + +#ifdef CONFIG_HIGHMEM +# define VMALLOC_END (PKMAP_BASE - 2 * PAGE_SIZE) +#else +# define VMALLOC_END (LDT_BASE_ADDR - 2 * PAGE_SIZE) +#endif + +#define MODULES_VADDR VMALLOC_START +#define MODULES_END VMALLOC_END +#define MODULES_LEN (MODULES_VADDR - MODULES_END) + +#define MAXMEM (VMALLOC_END - PAGE_OFFSET - __VMALLOC_RESERVE) + +#endif /* _ASM_X86_PGTABLE_32_AREAS_H */ diff --git a/arch/x86/include/asm/pgtable_32_types.h b/arch/x86/include/asm/pgtable_32_types.h index 0416d42e5bdd..5356a46b0373 100644 --- a/arch/x86/include/asm/pgtable_32_types.h +++ b/arch/x86/include/asm/pgtable_32_types.h @@ -1,6 +1,6 @@ /* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _ASM_X86_PGTABLE_32_DEFS_H -#define _ASM_X86_PGTABLE_32_DEFS_H +#ifndef _ASM_X86_PGTABLE_32_TYPES_H +#define _ASM_X86_PGTABLE_32_TYPES_H /* * The Linux x86 paging architecture is 'compile-time dual-mode', it @@ -20,55 +20,4 @@ #define PGDIR_SIZE (1UL << PGDIR_SHIFT) #define PGDIR_MASK (~(PGDIR_SIZE - 1)) -/* Just any arbitrary offset to the start of the vmalloc VM area: the - * current 8MB value just means that there will be a 8MB "hole" after the - * physical memory until the kernel virtual memory starts. That means that - * any out-of-bounds memory accesses will hopefully be caught. - * The vmalloc() routines leaves a hole of 4kB between each vmalloced - * area for the same reason. ;) - */ -#define VMALLOC_OFFSET (8 * 1024 * 1024) - -#ifndef __ASSEMBLY__ -extern bool __vmalloc_start_set; /* set once high_memory is set */ -#endif - -#define VMALLOC_START ((unsigned long)high_memory + VMALLOC_OFFSET) -#ifdef CONFIG_X86_PAE -#define LAST_PKMAP 512 -#else -#define LAST_PKMAP 1024 -#endif - -/* - * This is an upper bound on sizeof(struct cpu_entry_area) / PAGE_SIZE. - * Define this here and validate with BUILD_BUG_ON() in cpu_entry_area.c - * to avoid include recursion hell. - */ -#define CPU_ENTRY_AREA_PAGES (NR_CPUS * 43) - -/* The +1 is for the readonly IDT page: */ -#define CPU_ENTRY_AREA_BASE \ - ((FIXADDR_TOT_START - PAGE_SIZE*(CPU_ENTRY_AREA_PAGES+1)) & PMD_MASK) - -#define LDT_BASE_ADDR \ - ((CPU_ENTRY_AREA_BASE - PAGE_SIZE) & PMD_MASK) - -#define LDT_END_ADDR (LDT_BASE_ADDR + PMD_SIZE) - -#define PKMAP_BASE \ - ((LDT_BASE_ADDR - PAGE_SIZE) & PMD_MASK) - -#ifdef CONFIG_HIGHMEM -# define VMALLOC_END (PKMAP_BASE - 2 * PAGE_SIZE) -#else -# define VMALLOC_END (LDT_BASE_ADDR - 2 * PAGE_SIZE) -#endif - -#define MODULES_VADDR VMALLOC_START -#define MODULES_END VMALLOC_END -#define MODULES_LEN (MODULES_VADDR - MODULES_END) - -#define MAXMEM (VMALLOC_END - PAGE_OFFSET - __VMALLOC_RESERVE) - -#endif /* _ASM_X86_PGTABLE_32_DEFS_H */ +#endif /* _ASM_X86_PGTABLE_32_TYPES_H */ diff --git a/arch/x86/include/asm/pgtable_areas.h b/arch/x86/include/asm/pgtable_areas.h new file mode 100644 index 000000000000..d34cce1b995c --- /dev/null +++ b/arch/x86/include/asm/pgtable_areas.h @@ -0,0 +1,16 @@ +#ifndef _ASM_X86_PGTABLE_AREAS_H +#define _ASM_X86_PGTABLE_AREAS_H + +#ifdef CONFIG_X86_32 +# include <asm/pgtable_32_areas.h> +#endif + +/* Single page reserved for the readonly IDT mapping: */ +#define CPU_ENTRY_AREA_RO_IDT CPU_ENTRY_AREA_BASE +#define CPU_ENTRY_AREA_PER_CPU (CPU_ENTRY_AREA_RO_IDT + PAGE_SIZE) + +#define CPU_ENTRY_AREA_RO_IDT_VADDR ((void *)CPU_ENTRY_AREA_RO_IDT) + +#define CPU_ENTRY_AREA_MAP_SIZE (CPU_ENTRY_AREA_PER_CPU + CPU_ENTRY_AREA_ARRAY_SIZE - CPU_ENTRY_AREA_BASE) + +#endif /* _ASM_X86_PGTABLE_AREAS_H */ diff --git a/arch/x86/include/asm/pgtable_types.h b/arch/x86/include/asm/pgtable_types.h index b5e49e6bac63..ea7400726d7a 100644 --- a/arch/x86/include/asm/pgtable_types.h +++ b/arch/x86/include/asm/pgtable_types.h @@ -110,11 +110,6 @@ #define _PAGE_PROTNONE (_AT(pteval_t, 1) << _PAGE_BIT_PROTNONE) -#define _PAGE_TABLE_NOENC (_PAGE_PRESENT | _PAGE_RW | _PAGE_USER |\ - _PAGE_ACCESSED | _PAGE_DIRTY) -#define _KERNPG_TABLE_NOENC (_PAGE_PRESENT | _PAGE_RW | \ - _PAGE_ACCESSED | _PAGE_DIRTY) - /* * Set of bits not changed in pte_modify. The pte's * protection key is treated like _PAGE_RW, for @@ -136,80 +131,93 @@ */ #ifndef __ASSEMBLY__ enum page_cache_mode { - _PAGE_CACHE_MODE_WB = 0, - _PAGE_CACHE_MODE_WC = 1, + _PAGE_CACHE_MODE_WB = 0, + _PAGE_CACHE_MODE_WC = 1, _PAGE_CACHE_MODE_UC_MINUS = 2, - _PAGE_CACHE_MODE_UC = 3, - _PAGE_CACHE_MODE_WT = 4, - _PAGE_CACHE_MODE_WP = 5, - _PAGE_CACHE_MODE_NUM = 8 + _PAGE_CACHE_MODE_UC = 3, + _PAGE_CACHE_MODE_WT = 4, + _PAGE_CACHE_MODE_WP = 5, + + _PAGE_CACHE_MODE_NUM = 8 }; #endif -#define _PAGE_CACHE_MASK (_PAGE_PAT | _PAGE_PCD | _PAGE_PWT) -#define _PAGE_NOCACHE (cachemode2protval(_PAGE_CACHE_MODE_UC)) -#define _PAGE_CACHE_WP (cachemode2protval(_PAGE_CACHE_MODE_WP)) +#define _PAGE_ENC (_AT(pteval_t, sme_me_mask)) -#define PAGE_NONE __pgprot(_PAGE_PROTNONE | _PAGE_ACCESSED) -#define PAGE_SHARED __pgprot(_PAGE_PRESENT | _PAGE_RW | _PAGE_USER | \ - _PAGE_ACCESSED | _PAGE_NX) - -#define PAGE_SHARED_EXEC __pgprot(_PAGE_PRESENT | _PAGE_RW | \ - _PAGE_USER | _PAGE_ACCESSED) -#define PAGE_COPY_NOEXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | \ - _PAGE_ACCESSED | _PAGE_NX) -#define PAGE_COPY_EXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | \ - _PAGE_ACCESSED) -#define PAGE_COPY PAGE_COPY_NOEXEC -#define PAGE_READONLY __pgprot(_PAGE_PRESENT | _PAGE_USER | \ - _PAGE_ACCESSED | _PAGE_NX) -#define PAGE_READONLY_EXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | \ - _PAGE_ACCESSED) - -#define __PAGE_KERNEL_EXEC \ - (_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_GLOBAL) -#define __PAGE_KERNEL (__PAGE_KERNEL_EXEC | _PAGE_NX) - -#define __PAGE_KERNEL_RO (__PAGE_KERNEL & ~_PAGE_RW) -#define __PAGE_KERNEL_RX (__PAGE_KERNEL_EXEC & ~_PAGE_RW) -#define __PAGE_KERNEL_NOCACHE (__PAGE_KERNEL | _PAGE_NOCACHE) -#define __PAGE_KERNEL_VVAR (__PAGE_KERNEL_RO | _PAGE_USER) -#define __PAGE_KERNEL_LARGE (__PAGE_KERNEL | _PAGE_PSE) -#define __PAGE_KERNEL_LARGE_EXEC (__PAGE_KERNEL_EXEC | _PAGE_PSE) -#define __PAGE_KERNEL_WP (__PAGE_KERNEL | _PAGE_CACHE_WP) - -#define __PAGE_KERNEL_IO (__PAGE_KERNEL) -#define __PAGE_KERNEL_IO_NOCACHE (__PAGE_KERNEL_NOCACHE) +#define _PAGE_CACHE_MASK (_PAGE_PWT | _PAGE_PCD | _PAGE_PAT) -#ifndef __ASSEMBLY__ +#define _PAGE_NOCACHE (cachemode2protval(_PAGE_CACHE_MODE_UC)) +#define _PAGE_CACHE_WP (cachemode2protval(_PAGE_CACHE_MODE_WP)) -#define _PAGE_ENC (_AT(pteval_t, sme_me_mask)) +#define __PP _PAGE_PRESENT +#define __RW _PAGE_RW +#define _USR _PAGE_USER +#define ___A _PAGE_ACCESSED +#define ___D _PAGE_DIRTY +#define ___G _PAGE_GLOBAL +#define __NX _PAGE_NX + +#define _ENC _PAGE_ENC +#define __WP _PAGE_CACHE_WP +#define __NC _PAGE_NOCACHE +#define _PSE _PAGE_PSE + +#define pgprot_val(x) ((x).pgprot) +#define __pgprot(x) ((pgprot_t) { (x) } ) +#define __pg(x) __pgprot(x) + +#define _PAGE_PAT_LARGE (_AT(pteval_t, 1) << _PAGE_BIT_PAT_LARGE) + +#define PAGE_NONE __pg( 0| 0| 0|___A| 0| 0| 0|___G) +#define PAGE_SHARED __pg(__PP|__RW|_USR|___A|__NX| 0| 0| 0) +#define PAGE_SHARED_EXEC __pg(__PP|__RW|_USR|___A| 0| 0| 0| 0) +#define PAGE_COPY_NOEXEC __pg(__PP| 0|_USR|___A|__NX| 0| 0| 0) +#define PAGE_COPY_EXEC __pg(__PP| 0|_USR|___A| 0| 0| 0| 0) +#define PAGE_COPY __pg(__PP| 0|_USR|___A|__NX| 0| 0| 0) +#define PAGE_READONLY __pg(__PP| 0|_USR|___A|__NX| 0| 0| 0) +#define PAGE_READONLY_EXEC __pg(__PP| 0|_USR|___A| 0| 0| 0| 0) + +#define __PAGE_KERNEL (__PP|__RW| 0|___A|__NX|___D| 0|___G) +#define __PAGE_KERNEL_EXEC (__PP|__RW| 0|___A| 0|___D| 0|___G) +#define _KERNPG_TABLE_NOENC (__PP|__RW| 0|___A| 0|___D| 0| 0) +#define _KERNPG_TABLE (__PP|__RW| 0|___A| 0|___D| 0| 0| _ENC) +#define _PAGE_TABLE_NOENC (__PP|__RW|_USR|___A| 0|___D| 0| 0) +#define _PAGE_TABLE (__PP|__RW|_USR|___A| 0|___D| 0| 0| _ENC) +#define __PAGE_KERNEL_RO (__PP| 0| 0|___A|__NX|___D| 0|___G) +#define __PAGE_KERNEL_RX (__PP| 0| 0|___A| 0|___D| 0|___G) +#define __PAGE_KERNEL_NOCACHE (__PP|__RW| 0|___A|__NX|___D| 0|___G| __NC) +#define __PAGE_KERNEL_VVAR (__PP| 0|_USR|___A|__NX|___D| 0|___G) +#define __PAGE_KERNEL_LARGE (__PP|__RW| 0|___A|__NX|___D|_PSE|___G) +#define __PAGE_KERNEL_LARGE_EXEC (__PP|__RW| 0|___A| 0|___D|_PSE|___G) +#define __PAGE_KERNEL_WP (__PP|__RW| 0|___A|__NX|___D| 0|___G| __WP) + + +#define __PAGE_KERNEL_IO __PAGE_KERNEL +#define __PAGE_KERNEL_IO_NOCACHE __PAGE_KERNEL_NOCACHE -#define _KERNPG_TABLE (_PAGE_PRESENT | _PAGE_RW | _PAGE_ACCESSED | \ - _PAGE_DIRTY | _PAGE_ENC) -#define _PAGE_TABLE (_KERNPG_TABLE | _PAGE_USER) -#define __PAGE_KERNEL_ENC (__PAGE_KERNEL | _PAGE_ENC) -#define __PAGE_KERNEL_ENC_WP (__PAGE_KERNEL_WP | _PAGE_ENC) +#ifndef __ASSEMBLY__ -#define __PAGE_KERNEL_NOENC (__PAGE_KERNEL) -#define __PAGE_KERNEL_NOENC_WP (__PAGE_KERNEL_WP) +#define __PAGE_KERNEL_ENC (__PAGE_KERNEL | _ENC) +#define __PAGE_KERNEL_ENC_WP (__PAGE_KERNEL_WP | _ENC) +#define __PAGE_KERNEL_NOENC (__PAGE_KERNEL | 0) +#define __PAGE_KERNEL_NOENC_WP (__PAGE_KERNEL_WP | 0) -#define default_pgprot(x) __pgprot((x) & __default_kernel_pte_mask) +#define __pgprot_mask(x) __pgprot((x) & __default_kernel_pte_mask) -#define PAGE_KERNEL default_pgprot(__PAGE_KERNEL | _PAGE_ENC) -#define PAGE_KERNEL_NOENC default_pgprot(__PAGE_KERNEL) -#define PAGE_KERNEL_RO default_pgprot(__PAGE_KERNEL_RO | _PAGE_ENC) -#define PAGE_KERNEL_EXEC default_pgprot(__PAGE_KERNEL_EXEC | _PAGE_ENC) -#define PAGE_KERNEL_EXEC_NOENC default_pgprot(__PAGE_KERNEL_EXEC) -#define PAGE_KERNEL_RX default_pgprot(__PAGE_KERNEL_RX | _PAGE_ENC) -#define PAGE_KERNEL_NOCACHE default_pgprot(__PAGE_KERNEL_NOCACHE | _PAGE_ENC) -#define PAGE_KERNEL_LARGE default_pgprot(__PAGE_KERNEL_LARGE | _PAGE_ENC) -#define PAGE_KERNEL_LARGE_EXEC default_pgprot(__PAGE_KERNEL_LARGE_EXEC | _PAGE_ENC) -#define PAGE_KERNEL_VVAR default_pgprot(__PAGE_KERNEL_VVAR | _PAGE_ENC) +#define PAGE_KERNEL __pgprot_mask(__PAGE_KERNEL | _ENC) +#define PAGE_KERNEL_NOENC __pgprot_mask(__PAGE_KERNEL | 0) +#define PAGE_KERNEL_RO __pgprot_mask(__PAGE_KERNEL_RO | _ENC) +#define PAGE_KERNEL_EXEC __pgprot_mask(__PAGE_KERNEL_EXEC | _ENC) +#define PAGE_KERNEL_EXEC_NOENC __pgprot_mask(__PAGE_KERNEL_EXEC | 0) +#define PAGE_KERNEL_RX __pgprot_mask(__PAGE_KERNEL_RX | _ENC) +#define PAGE_KERNEL_NOCACHE __pgprot_mask(__PAGE_KERNEL_NOCACHE | _ENC) +#define PAGE_KERNEL_LARGE __pgprot_mask(__PAGE_KERNEL_LARGE | _ENC) +#define PAGE_KERNEL_LARGE_EXEC __pgprot_mask(__PAGE_KERNEL_LARGE_EXEC | _ENC) +#define PAGE_KERNEL_VVAR __pgprot_mask(__PAGE_KERNEL_VVAR | _ENC) -#define PAGE_KERNEL_IO default_pgprot(__PAGE_KERNEL_IO) -#define PAGE_KERNEL_IO_NOCACHE default_pgprot(__PAGE_KERNEL_IO_NOCACHE) +#define PAGE_KERNEL_IO __pgprot_mask(__PAGE_KERNEL_IO) +#define PAGE_KERNEL_IO_NOCACHE __pgprot_mask(__PAGE_KERNEL_IO_NOCACHE) #endif /* __ASSEMBLY__ */ @@ -449,9 +457,6 @@ static inline pteval_t pte_flags(pte_t pte) return native_pte_val(pte) & PTE_FLAGS_MASK; } -#define pgprot_val(x) ((x).pgprot) -#define __pgprot(x) ((pgprot_t) { (x) } ) - extern uint16_t __cachemode2pte_tbl[_PAGE_CACHE_MODE_NUM]; extern uint8_t __pte2cachemode_tbl[8]; diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h index 0340aad3f2fc..6fb4870ed759 100644 --- a/arch/x86/include/asm/processor.h +++ b/arch/x86/include/asm/processor.h @@ -25,6 +25,7 @@ struct vm86; #include <asm/special_insns.h> #include <asm/fpu/types.h> #include <asm/unwind_hints.h> +#include <asm/vmxfeatures.h> #include <linux/personality.h> #include <linux/cache.h> @@ -85,6 +86,9 @@ struct cpuinfo_x86 { /* Number of 4K pages in DTLB/ITLB combined(in pages): */ int x86_tlbsize; #endif +#ifdef CONFIG_X86_VMX_FEATURE_NAMES + __u32 vmx_capability[NVMXINTS]; +#endif __u8 x86_virt_bits; __u8 x86_phys_bits; /* CPUID returned core id bits: */ @@ -1015,11 +1019,4 @@ enum mds_mitigations { MDS_MITIGATION_VMWERV, }; -enum taa_mitigations { - TAA_MITIGATION_OFF, - TAA_MITIGATION_UCODE_NEEDED, - TAA_MITIGATION_VERW, - TAA_MITIGATION_TSX_DISABLED, -}; - #endif /* _ASM_X86_PROCESSOR_H */ diff --git a/arch/x86/include/asm/ptrace.h b/arch/x86/include/asm/ptrace.h index 5057a8ed100b..6d6475fdd327 100644 --- a/arch/x86/include/asm/ptrace.h +++ b/arch/x86/include/asm/ptrace.h @@ -159,6 +159,19 @@ static inline bool user_64bit_mode(struct pt_regs *regs) #endif } +/* + * Determine whether the register set came from any context that is running in + * 64-bit mode. + */ +static inline bool any_64bit_mode(struct pt_regs *regs) +{ +#ifdef CONFIG_X86_64 + return !user_mode(regs) || user_64bit_mode(regs); +#else + return false; +#endif +} + #ifdef CONFIG_X86_64 #define current_user_stack_pointer() current_pt_regs()->sp #define compat_user_stack_pointer() current_pt_regs()->sp @@ -339,22 +352,6 @@ static inline unsigned long regs_get_kernel_argument(struct pt_regs *regs, #define ARCH_HAS_USER_SINGLE_STEP_REPORT -/* - * When hitting ptrace_stop(), we cannot return using SYSRET because - * that does not restore the full CPU state, only a minimal set. The - * ptracer can change arbitrary register values, which is usually okay - * because the usual ptrace stops run off the signal delivery path which - * forces IRET; however, ptrace_event() stops happen in arbitrary places - * in the kernel and don't force IRET path. - * - * So force IRET path after a ptrace stop. - */ -#define arch_ptrace_stop_needed(code, info) \ -({ \ - force_iret(); \ - false; \ -}) - struct user_desc; extern int do_get_thread_area(struct task_struct *p, int idx, struct user_desc __user *info); diff --git a/arch/x86/include/asm/realmode.h b/arch/x86/include/asm/realmode.h index 09ecc32f6524..b35030eeec36 100644 --- a/arch/x86/include/asm/realmode.h +++ b/arch/x86/include/asm/realmode.h @@ -14,7 +14,7 @@ #include <linux/types.h> #include <asm/io.h> -/* This must match data at realmode.S */ +/* This must match data at realmode/rm/header.S */ struct real_mode_header { u32 text_start; u32 ro_end; @@ -36,7 +36,7 @@ struct real_mode_header { #endif }; -/* This must match data at trampoline_32/64.S */ +/* This must match data at realmode/rm/trampoline_{32,64}.S */ struct trampoline_header { #ifdef CONFIG_X86_32 u32 start; diff --git a/arch/x86/include/asm/set_memory.h b/arch/x86/include/asm/set_memory.h index 2ee8e469dcf5..64c3dce374e5 100644 --- a/arch/x86/include/asm/set_memory.h +++ b/arch/x86/include/asm/set_memory.h @@ -81,8 +81,6 @@ int set_direct_map_invalid_noflush(struct page *page); int set_direct_map_default_noflush(struct page *page); extern int kernel_set_to_readonly; -void set_kernel_text_rw(void); -void set_kernel_text_ro(void); #ifdef CONFIG_X86_64 static inline int set_mce_nospec(unsigned long pfn) diff --git a/arch/x86/include/asm/text-patching.h b/arch/x86/include/asm/text-patching.h index 23c626a742e8..67315fa3956a 100644 --- a/arch/x86/include/asm/text-patching.h +++ b/arch/x86/include/asm/text-patching.h @@ -25,14 +25,6 @@ static inline void apply_paravirt(struct paravirt_patch_site *start, */ #define POKE_MAX_OPCODE_SIZE 5 -struct text_poke_loc { - void *addr; - int len; - s32 rel32; - u8 opcode; - const u8 text[POKE_MAX_OPCODE_SIZE]; -}; - extern void text_poke_early(void *addr, const void *opcode, size_t len); /* @@ -50,21 +42,13 @@ extern void text_poke_early(void *addr, const void *opcode, size_t len); * an inconsistent instruction while you patch. */ extern void *text_poke(void *addr, const void *opcode, size_t len); +extern void text_poke_sync(void); extern void *text_poke_kgdb(void *addr, const void *opcode, size_t len); extern int poke_int3_handler(struct pt_regs *regs); extern void text_poke_bp(void *addr, const void *opcode, size_t len, const void *emulate); -extern void text_poke_bp_batch(struct text_poke_loc *tp, unsigned int nr_entries); -extern void text_poke_loc_init(struct text_poke_loc *tp, void *addr, - const void *opcode, size_t len, const void *emulate); -extern int after_bootmem; -extern __ro_after_init struct mm_struct *poking_mm; -extern __ro_after_init unsigned long poking_addr; -#ifndef CONFIG_UML_X86 -static inline void int3_emulate_jmp(struct pt_regs *regs, unsigned long ip) -{ - regs->ip = ip; -} +extern void text_poke_queue(void *addr, const void *opcode, size_t len, const void *emulate); +extern void text_poke_finish(void); #define INT3_INSN_SIZE 1 #define INT3_INSN_OPCODE 0xCC @@ -78,6 +62,67 @@ static inline void int3_emulate_jmp(struct pt_regs *regs, unsigned long ip) #define JMP8_INSN_SIZE 2 #define JMP8_INSN_OPCODE 0xEB +#define DISP32_SIZE 4 + +static inline int text_opcode_size(u8 opcode) +{ + int size = 0; + +#define __CASE(insn) \ + case insn##_INSN_OPCODE: size = insn##_INSN_SIZE; break + + switch(opcode) { + __CASE(INT3); + __CASE(CALL); + __CASE(JMP32); + __CASE(JMP8); + } + +#undef __CASE + + return size; +} + +union text_poke_insn { + u8 text[POKE_MAX_OPCODE_SIZE]; + struct { + u8 opcode; + s32 disp; + } __attribute__((packed)); +}; + +static __always_inline +void *text_gen_insn(u8 opcode, const void *addr, const void *dest) +{ + static union text_poke_insn insn; /* per instance */ + int size = text_opcode_size(opcode); + + insn.opcode = opcode; + + if (size > 1) { + insn.disp = (long)dest - (long)(addr + size); + if (size == 2) { + /* + * Ensure that for JMP9 the displacement + * actually fits the signed byte. + */ + BUG_ON((insn.disp >> 31) != (insn.disp >> 7)); + } + } + + return &insn.text; +} + +extern int after_bootmem; +extern __ro_after_init struct mm_struct *poking_mm; +extern __ro_after_init unsigned long poking_addr; + +#ifndef CONFIG_UML_X86 +static inline void int3_emulate_jmp(struct pt_regs *regs, unsigned long ip) +{ + regs->ip = ip; +} + static inline void int3_emulate_push(struct pt_regs *regs, unsigned long val) { /* @@ -85,6 +130,9 @@ static inline void int3_emulate_push(struct pt_regs *regs, unsigned long val) * stack where the break point happened, and the saving of * pt_regs. We can extend the original stack because of * this gap. See the idtentry macro's create_gap option. + * + * Similarly entry_32.S will have a gap on the stack for (any) hardware + * exception and pt_regs; see FIXUP_FRAME. */ regs->sp -= sizeof(unsigned long); *(unsigned long *)regs->sp = val; diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h index d779366ce3f8..cf4327986e98 100644 --- a/arch/x86/include/asm/thread_info.h +++ b/arch/x86/include/asm/thread_info.h @@ -239,15 +239,6 @@ static inline int arch_within_stack_frames(const void * const stack, current_thread_info()->status & TS_COMPAT) #endif -/* - * Force syscall return via IRET by making it look as if there was - * some work pending. IRET is our most capable (but slowest) syscall - * return path, which is able to restore modified SS, CS and certain - * EFLAGS values that other (fast) syscall return instructions - * are not able to restore properly. - */ -#define force_iret() set_thread_flag(TIF_NOTIFY_RESUME) - extern void arch_task_cache_init(void); extern int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src); extern void arch_release_task_struct(struct task_struct *tsk); diff --git a/arch/x86/include/asm/vdso.h b/arch/x86/include/asm/vdso.h index 230474e2ddb5..bbcdc7b8f963 100644 --- a/arch/x86/include/asm/vdso.h +++ b/arch/x86/include/asm/vdso.h @@ -21,6 +21,7 @@ struct vdso_image { long sym_vvar_page; long sym_pvclock_page; long sym_hvclock_page; + long sym_timens_page; long sym_VDSO32_NOTE_MASK; long sym___kernel_sigreturn; long sym___kernel_rt_sigreturn; diff --git a/arch/x86/include/asm/vdso/gettimeofday.h b/arch/x86/include/asm/vdso/gettimeofday.h index e9ee139cf29e..6ee1f7dba34b 100644 --- a/arch/x86/include/asm/vdso/gettimeofday.h +++ b/arch/x86/include/asm/vdso/gettimeofday.h @@ -21,6 +21,7 @@ #include <clocksource/hyperv_timer.h> #define __vdso_data (VVAR(_vdso_data)) +#define __timens_vdso_data (TIMENS(_vdso_data)) #define VDSO_HAS_TIME 1 @@ -56,6 +57,13 @@ extern struct ms_hyperv_tsc_page hvclock_page __attribute__((visibility("hidden"))); #endif +#ifdef CONFIG_TIME_NS +static __always_inline const struct vdso_data *__arch_get_timens_vdso_data(void) +{ + return __timens_vdso_data; +} +#endif + #ifndef BUILD_VDSO32 static __always_inline @@ -96,8 +104,6 @@ long clock_getres_fallback(clockid_t _clkid, struct __kernel_timespec *_ts) #else -#define VDSO_HAS_32BIT_FALLBACK 1 - static __always_inline long clock_gettime_fallback(clockid_t _clkid, struct __kernel_timespec *_ts) { diff --git a/arch/x86/include/asm/vmalloc.h b/arch/x86/include/asm/vmalloc.h new file mode 100644 index 000000000000..29837740b520 --- /dev/null +++ b/arch/x86/include/asm/vmalloc.h @@ -0,0 +1,6 @@ +#ifndef _ASM_X86_VMALLOC_H +#define _ASM_X86_VMALLOC_H + +#include <asm/pgtable_areas.h> + +#endif /* _ASM_X86_VMALLOC_H */ diff --git a/arch/x86/include/asm/vmx.h b/arch/x86/include/asm/vmx.h index 1835767aa335..9fbba31be825 100644 --- a/arch/x86/include/asm/vmx.h +++ b/arch/x86/include/asm/vmx.h @@ -15,67 +15,70 @@ #include <linux/bitops.h> #include <linux/types.h> #include <uapi/asm/vmx.h> +#include <asm/vmxfeatures.h> + +#define VMCS_CONTROL_BIT(x) BIT(VMX_FEATURE_##x & 0x1f) /* * Definitions of Primary Processor-Based VM-Execution Controls. */ -#define CPU_BASED_VIRTUAL_INTR_PENDING 0x00000004 -#define CPU_BASED_USE_TSC_OFFSETING 0x00000008 -#define CPU_BASED_HLT_EXITING 0x00000080 -#define CPU_BASED_INVLPG_EXITING 0x00000200 -#define CPU_BASED_MWAIT_EXITING 0x00000400 -#define CPU_BASED_RDPMC_EXITING 0x00000800 -#define CPU_BASED_RDTSC_EXITING 0x00001000 -#define CPU_BASED_CR3_LOAD_EXITING 0x00008000 -#define CPU_BASED_CR3_STORE_EXITING 0x00010000 -#define CPU_BASED_CR8_LOAD_EXITING 0x00080000 -#define CPU_BASED_CR8_STORE_EXITING 0x00100000 -#define CPU_BASED_TPR_SHADOW 0x00200000 -#define CPU_BASED_VIRTUAL_NMI_PENDING 0x00400000 -#define CPU_BASED_MOV_DR_EXITING 0x00800000 -#define CPU_BASED_UNCOND_IO_EXITING 0x01000000 -#define CPU_BASED_USE_IO_BITMAPS 0x02000000 -#define CPU_BASED_MONITOR_TRAP_FLAG 0x08000000 -#define CPU_BASED_USE_MSR_BITMAPS 0x10000000 -#define CPU_BASED_MONITOR_EXITING 0x20000000 -#define CPU_BASED_PAUSE_EXITING 0x40000000 -#define CPU_BASED_ACTIVATE_SECONDARY_CONTROLS 0x80000000 +#define CPU_BASED_VIRTUAL_INTR_PENDING VMCS_CONTROL_BIT(VIRTUAL_INTR_PENDING) +#define CPU_BASED_USE_TSC_OFFSETING VMCS_CONTROL_BIT(TSC_OFFSETTING) +#define CPU_BASED_HLT_EXITING VMCS_CONTROL_BIT(HLT_EXITING) +#define CPU_BASED_INVLPG_EXITING VMCS_CONTROL_BIT(INVLPG_EXITING) +#define CPU_BASED_MWAIT_EXITING VMCS_CONTROL_BIT(MWAIT_EXITING) +#define CPU_BASED_RDPMC_EXITING VMCS_CONTROL_BIT(RDPMC_EXITING) +#define CPU_BASED_RDTSC_EXITING VMCS_CONTROL_BIT(RDTSC_EXITING) +#define CPU_BASED_CR3_LOAD_EXITING VMCS_CONTROL_BIT(CR3_LOAD_EXITING) +#define CPU_BASED_CR3_STORE_EXITING VMCS_CONTROL_BIT(CR3_STORE_EXITING) +#define CPU_BASED_CR8_LOAD_EXITING VMCS_CONTROL_BIT(CR8_LOAD_EXITING) +#define CPU_BASED_CR8_STORE_EXITING VMCS_CONTROL_BIT(CR8_STORE_EXITING) +#define CPU_BASED_TPR_SHADOW VMCS_CONTROL_BIT(VIRTUAL_TPR) +#define CPU_BASED_VIRTUAL_NMI_PENDING VMCS_CONTROL_BIT(VIRTUAL_NMI_PENDING) +#define CPU_BASED_MOV_DR_EXITING VMCS_CONTROL_BIT(MOV_DR_EXITING) +#define CPU_BASED_UNCOND_IO_EXITING VMCS_CONTROL_BIT(UNCOND_IO_EXITING) +#define CPU_BASED_USE_IO_BITMAPS VMCS_CONTROL_BIT(USE_IO_BITMAPS) +#define CPU_BASED_MONITOR_TRAP_FLAG VMCS_CONTROL_BIT(MONITOR_TRAP_FLAG) +#define CPU_BASED_USE_MSR_BITMAPS VMCS_CONTROL_BIT(USE_MSR_BITMAPS) +#define CPU_BASED_MONITOR_EXITING VMCS_CONTROL_BIT(MONITOR_EXITING) +#define CPU_BASED_PAUSE_EXITING VMCS_CONTROL_BIT(PAUSE_EXITING) +#define CPU_BASED_ACTIVATE_SECONDARY_CONTROLS VMCS_CONTROL_BIT(SEC_CONTROLS) #define CPU_BASED_ALWAYSON_WITHOUT_TRUE_MSR 0x0401e172 /* * Definitions of Secondary Processor-Based VM-Execution Controls. */ -#define SECONDARY_EXEC_VIRTUALIZE_APIC_ACCESSES 0x00000001 -#define SECONDARY_EXEC_ENABLE_EPT 0x00000002 -#define SECONDARY_EXEC_DESC 0x00000004 -#define SECONDARY_EXEC_RDTSCP 0x00000008 -#define SECONDARY_EXEC_VIRTUALIZE_X2APIC_MODE 0x00000010 -#define SECONDARY_EXEC_ENABLE_VPID 0x00000020 -#define SECONDARY_EXEC_WBINVD_EXITING 0x00000040 -#define SECONDARY_EXEC_UNRESTRICTED_GUEST 0x00000080 -#define SECONDARY_EXEC_APIC_REGISTER_VIRT 0x00000100 -#define SECONDARY_EXEC_VIRTUAL_INTR_DELIVERY 0x00000200 -#define SECONDARY_EXEC_PAUSE_LOOP_EXITING 0x00000400 -#define SECONDARY_EXEC_RDRAND_EXITING 0x00000800 -#define SECONDARY_EXEC_ENABLE_INVPCID 0x00001000 -#define SECONDARY_EXEC_ENABLE_VMFUNC 0x00002000 -#define SECONDARY_EXEC_SHADOW_VMCS 0x00004000 -#define SECONDARY_EXEC_ENCLS_EXITING 0x00008000 -#define SECONDARY_EXEC_RDSEED_EXITING 0x00010000 -#define SECONDARY_EXEC_ENABLE_PML 0x00020000 -#define SECONDARY_EXEC_PT_CONCEAL_VMX 0x00080000 -#define SECONDARY_EXEC_XSAVES 0x00100000 -#define SECONDARY_EXEC_PT_USE_GPA 0x01000000 -#define SECONDARY_EXEC_MODE_BASED_EPT_EXEC 0x00400000 -#define SECONDARY_EXEC_TSC_SCALING 0x02000000 +#define SECONDARY_EXEC_VIRTUALIZE_APIC_ACCESSES VMCS_CONTROL_BIT(VIRT_APIC_ACCESSES) +#define SECONDARY_EXEC_ENABLE_EPT VMCS_CONTROL_BIT(EPT) +#define SECONDARY_EXEC_DESC VMCS_CONTROL_BIT(DESC_EXITING) +#define SECONDARY_EXEC_RDTSCP VMCS_CONTROL_BIT(RDTSCP) +#define SECONDARY_EXEC_VIRTUALIZE_X2APIC_MODE VMCS_CONTROL_BIT(VIRTUAL_X2APIC) +#define SECONDARY_EXEC_ENABLE_VPID VMCS_CONTROL_BIT(VPID) +#define SECONDARY_EXEC_WBINVD_EXITING VMCS_CONTROL_BIT(WBINVD_EXITING) +#define SECONDARY_EXEC_UNRESTRICTED_GUEST VMCS_CONTROL_BIT(UNRESTRICTED_GUEST) +#define SECONDARY_EXEC_APIC_REGISTER_VIRT VMCS_CONTROL_BIT(APIC_REGISTER_VIRT) +#define SECONDARY_EXEC_VIRTUAL_INTR_DELIVERY VMCS_CONTROL_BIT(VIRT_INTR_DELIVERY) +#define SECONDARY_EXEC_PAUSE_LOOP_EXITING VMCS_CONTROL_BIT(PAUSE_LOOP_EXITING) +#define SECONDARY_EXEC_RDRAND_EXITING VMCS_CONTROL_BIT(RDRAND_EXITING) +#define SECONDARY_EXEC_ENABLE_INVPCID VMCS_CONTROL_BIT(INVPCID) +#define SECONDARY_EXEC_ENABLE_VMFUNC VMCS_CONTROL_BIT(VMFUNC) +#define SECONDARY_EXEC_SHADOW_VMCS VMCS_CONTROL_BIT(SHADOW_VMCS) +#define SECONDARY_EXEC_ENCLS_EXITING VMCS_CONTROL_BIT(ENCLS_EXITING) +#define SECONDARY_EXEC_RDSEED_EXITING VMCS_CONTROL_BIT(RDSEED_EXITING) +#define SECONDARY_EXEC_ENABLE_PML VMCS_CONTROL_BIT(PAGE_MOD_LOGGING) +#define SECONDARY_EXEC_PT_CONCEAL_VMX VMCS_CONTROL_BIT(PT_CONCEAL_VMX) +#define SECONDARY_EXEC_XSAVES VMCS_CONTROL_BIT(XSAVES) +#define SECONDARY_EXEC_MODE_BASED_EPT_EXEC VMCS_CONTROL_BIT(MODE_BASED_EPT_EXEC) +#define SECONDARY_EXEC_PT_USE_GPA VMCS_CONTROL_BIT(PT_USE_GPA) +#define SECONDARY_EXEC_TSC_SCALING VMCS_CONTROL_BIT(TSC_SCALING) #define SECONDARY_EXEC_ENABLE_USR_WAIT_PAUSE 0x04000000 -#define PIN_BASED_EXT_INTR_MASK 0x00000001 -#define PIN_BASED_NMI_EXITING 0x00000008 -#define PIN_BASED_VIRTUAL_NMIS 0x00000020 -#define PIN_BASED_VMX_PREEMPTION_TIMER 0x00000040 -#define PIN_BASED_POSTED_INTR 0x00000080 +#define PIN_BASED_EXT_INTR_MASK VMCS_CONTROL_BIT(INTR_EXITING) +#define PIN_BASED_NMI_EXITING VMCS_CONTROL_BIT(NMI_EXITING) +#define PIN_BASED_VIRTUAL_NMIS VMCS_CONTROL_BIT(VIRTUAL_NMIS) +#define PIN_BASED_VMX_PREEMPTION_TIMER VMCS_CONTROL_BIT(PREEMPTION_TIMER) +#define PIN_BASED_POSTED_INTR VMCS_CONTROL_BIT(POSTED_INTR) #define PIN_BASED_ALWAYSON_WITHOUT_TRUE_MSR 0x00000016 @@ -114,7 +117,9 @@ #define VMX_MISC_MSR_LIST_MULTIPLIER 512 /* VMFUNC functions */ -#define VMX_VMFUNC_EPTP_SWITCHING 0x00000001 +#define VMFUNC_CONTROL_BIT(x) BIT((VMX_FEATURE_##x & 0x1f) - 28) + +#define VMX_VMFUNC_EPTP_SWITCHING VMFUNC_CONTROL_BIT(EPTP_SWITCHING) #define VMFUNC_EPTP_ENTRIES 512 static inline u32 vmx_basic_vmcs_revision_id(u64 vmx_basic) diff --git a/arch/x86/include/asm/vmxfeatures.h b/arch/x86/include/asm/vmxfeatures.h new file mode 100644 index 000000000000..0d04d8bf15a5 --- /dev/null +++ b/arch/x86/include/asm/vmxfeatures.h @@ -0,0 +1,86 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _ASM_X86_VMXFEATURES_H +#define _ASM_X86_VMXFEATURES_H + +/* + * Defines VMX CPU feature bits + */ +#define NVMXINTS 3 /* N 32-bit words worth of info */ + +/* + * Note: If the comment begins with a quoted string, that string is used + * in /proc/cpuinfo instead of the macro name. If the string is "", + * this feature bit is not displayed in /proc/cpuinfo at all. + */ + +/* Pin-Based VM-Execution Controls, EPT/VPID, APIC and VM-Functions, word 0 */ +#define VMX_FEATURE_INTR_EXITING ( 0*32+ 0) /* "" VM-Exit on vectored interrupts */ +#define VMX_FEATURE_NMI_EXITING ( 0*32+ 3) /* "" VM-Exit on NMIs */ +#define VMX_FEATURE_VIRTUAL_NMIS ( 0*32+ 5) /* "vnmi" NMI virtualization */ +#define VMX_FEATURE_PREEMPTION_TIMER ( 0*32+ 6) /* VMX Preemption Timer */ +#define VMX_FEATURE_POSTED_INTR ( 0*32+ 7) /* Posted Interrupts */ + +/* EPT/VPID features, scattered to bits 16-23 */ +#define VMX_FEATURE_INVVPID ( 0*32+ 16) /* INVVPID is supported */ +#define VMX_FEATURE_EPT_EXECUTE_ONLY ( 0*32+ 17) /* "ept_x_only" EPT entries can be execute only */ +#define VMX_FEATURE_EPT_AD ( 0*32+ 18) /* EPT Accessed/Dirty bits */ +#define VMX_FEATURE_EPT_1GB ( 0*32+ 19) /* 1GB EPT pages */ + +/* Aggregated APIC features 24-27 */ +#define VMX_FEATURE_FLEXPRIORITY ( 0*32+ 24) /* TPR shadow + virt APIC */ +#define VMX_FEATURE_APICV ( 0*32+ 25) /* TPR shadow + APIC reg virt + virt intr delivery + posted interrupts */ + +/* VM-Functions, shifted to bits 28-31 */ +#define VMX_FEATURE_EPTP_SWITCHING ( 0*32+ 28) /* EPTP switching (in guest) */ + +/* Primary Processor-Based VM-Execution Controls, word 1 */ +#define VMX_FEATURE_VIRTUAL_INTR_PENDING ( 1*32+ 2) /* "" VM-Exit if INTRs are unblocked in guest */ +#define VMX_FEATURE_TSC_OFFSETTING ( 1*32+ 3) /* "tsc_offset" Offset hardware TSC when read in guest */ +#define VMX_FEATURE_HLT_EXITING ( 1*32+ 7) /* "" VM-Exit on HLT */ +#define VMX_FEATURE_INVLPG_EXITING ( 1*32+ 9) /* "" VM-Exit on INVLPG */ +#define VMX_FEATURE_MWAIT_EXITING ( 1*32+ 10) /* "" VM-Exit on MWAIT */ +#define VMX_FEATURE_RDPMC_EXITING ( 1*32+ 11) /* "" VM-Exit on RDPMC */ +#define VMX_FEATURE_RDTSC_EXITING ( 1*32+ 12) /* "" VM-Exit on RDTSC */ +#define VMX_FEATURE_CR3_LOAD_EXITING ( 1*32+ 15) /* "" VM-Exit on writes to CR3 */ +#define VMX_FEATURE_CR3_STORE_EXITING ( 1*32+ 16) /* "" VM-Exit on reads from CR3 */ +#define VMX_FEATURE_CR8_LOAD_EXITING ( 1*32+ 19) /* "" VM-Exit on writes to CR8 */ +#define VMX_FEATURE_CR8_STORE_EXITING ( 1*32+ 20) /* "" VM-Exit on reads from CR8 */ +#define VMX_FEATURE_VIRTUAL_TPR ( 1*32+ 21) /* "vtpr" TPR virtualization, a.k.a. TPR shadow */ +#define VMX_FEATURE_VIRTUAL_NMI_PENDING ( 1*32+ 22) /* "" VM-Exit if NMIs are unblocked in guest */ +#define VMX_FEATURE_MOV_DR_EXITING ( 1*32+ 23) /* "" VM-Exit on accesses to debug registers */ +#define VMX_FEATURE_UNCOND_IO_EXITING ( 1*32+ 24) /* "" VM-Exit on *all* IN{S} and OUT{S}*/ +#define VMX_FEATURE_USE_IO_BITMAPS ( 1*32+ 25) /* "" VM-Exit based on I/O port */ +#define VMX_FEATURE_MONITOR_TRAP_FLAG ( 1*32+ 27) /* "mtf" VMX single-step VM-Exits */ +#define VMX_FEATURE_USE_MSR_BITMAPS ( 1*32+ 28) /* "" VM-Exit based on MSR index */ +#define VMX_FEATURE_MONITOR_EXITING ( 1*32+ 29) /* "" VM-Exit on MONITOR (MWAIT's accomplice) */ +#define VMX_FEATURE_PAUSE_EXITING ( 1*32+ 30) /* "" VM-Exit on PAUSE (unconditionally) */ +#define VMX_FEATURE_SEC_CONTROLS ( 1*32+ 31) /* "" Enable Secondary VM-Execution Controls */ + +/* Secondary Processor-Based VM-Execution Controls, word 2 */ +#define VMX_FEATURE_VIRT_APIC_ACCESSES ( 2*32+ 0) /* "vapic" Virtualize memory mapped APIC accesses */ +#define VMX_FEATURE_EPT ( 2*32+ 1) /* Extended Page Tables, a.k.a. Two-Dimensional Paging */ +#define VMX_FEATURE_DESC_EXITING ( 2*32+ 2) /* "" VM-Exit on {S,L}*DT instructions */ +#define VMX_FEATURE_RDTSCP ( 2*32+ 3) /* "" Enable RDTSCP in guest */ +#define VMX_FEATURE_VIRTUAL_X2APIC ( 2*32+ 4) /* "" Virtualize X2APIC for the guest */ +#define VMX_FEATURE_VPID ( 2*32+ 5) /* Virtual Processor ID (TLB ASID modifier) */ +#define VMX_FEATURE_WBINVD_EXITING ( 2*32+ 6) /* "" VM-Exit on WBINVD */ +#define VMX_FEATURE_UNRESTRICTED_GUEST ( 2*32+ 7) /* Allow Big Real Mode and other "invalid" states */ +#define VMX_FEATURE_APIC_REGISTER_VIRT ( 2*32+ 8) /* "vapic_reg" Hardware emulation of reads to the virtual-APIC */ +#define VMX_FEATURE_VIRT_INTR_DELIVERY ( 2*32+ 9) /* "vid" Evaluation and delivery of pending virtual interrupts */ +#define VMX_FEATURE_PAUSE_LOOP_EXITING ( 2*32+ 10) /* "ple" Conditionally VM-Exit on PAUSE at CPL0 */ +#define VMX_FEATURE_RDRAND_EXITING ( 2*32+ 11) /* "" VM-Exit on RDRAND*/ +#define VMX_FEATURE_INVPCID ( 2*32+ 12) /* "" Enable INVPCID in guest */ +#define VMX_FEATURE_VMFUNC ( 2*32+ 13) /* "" Enable VM-Functions (leaf dependent) */ +#define VMX_FEATURE_SHADOW_VMCS ( 2*32+ 14) /* VMREAD/VMWRITE in guest can access shadow VMCS */ +#define VMX_FEATURE_ENCLS_EXITING ( 2*32+ 15) /* "" VM-Exit on ENCLS (leaf dependent) */ +#define VMX_FEATURE_RDSEED_EXITING ( 2*32+ 16) /* "" VM-Exit on RDSEED */ +#define VMX_FEATURE_PAGE_MOD_LOGGING ( 2*32+ 17) /* "pml" Log dirty pages into buffer */ +#define VMX_FEATURE_EPT_VIOLATION_VE ( 2*32+ 18) /* "" Conditionally reflect EPT violations as #VE exceptions */ +#define VMX_FEATURE_PT_CONCEAL_VMX ( 2*32+ 19) /* "" Suppress VMX indicators in Processor Trace */ +#define VMX_FEATURE_XSAVES ( 2*32+ 20) /* "" Enable XSAVES and XRSTORS in guest */ +#define VMX_FEATURE_MODE_BASED_EPT_EXEC ( 2*32+ 22) /* "ept_mode_based_exec" Enable separate EPT EXEC bits for supervisor vs. user */ +#define VMX_FEATURE_PT_USE_GPA ( 2*32+ 24) /* "" Processor Trace logs GPAs */ +#define VMX_FEATURE_TSC_SCALING ( 2*32+ 25) /* Scale hardware TSC when read in guest */ +#define VMX_FEATURE_ENCLV_EXITING ( 2*32+ 28) /* "" VM-Exit on ENCLV (leaf dependent) */ + +#endif /* _ASM_X86_VMXFEATURES_H */ diff --git a/arch/x86/include/asm/vvar.h b/arch/x86/include/asm/vvar.h index 32f5d9a0b90e..183e98e49ab9 100644 --- a/arch/x86/include/asm/vvar.h +++ b/arch/x86/include/asm/vvar.h @@ -19,10 +19,10 @@ #ifndef _ASM_X86_VVAR_H #define _ASM_X86_VVAR_H -#if defined(__VVAR_KERNEL_LDS) - -/* The kernel linker script defines its own magic to put vvars in the - * right place. +#ifdef EMIT_VVAR +/* + * EMIT_VVAR() is used by the kernel linker script to put vvars in the + * right place. Also, it's used by kernel code to import offsets values. */ #define DECLARE_VVAR(offset, type, name) \ EMIT_VVAR(name, offset) @@ -33,9 +33,12 @@ extern char __vvar_page; #define DECLARE_VVAR(offset, type, name) \ extern type vvar_ ## name[CS_BASES] \ - __attribute__((visibility("hidden"))); + __attribute__((visibility("hidden"))); \ + extern type timens_ ## name[CS_BASES] \ + __attribute__((visibility("hidden"))); \ #define VVAR(name) (vvar_ ## name) +#define TIMENS(name) (timens_ ## name) #define DEFINE_VVAR(type, name) \ type name[CS_BASES] \ diff --git a/arch/x86/kernel/acpi/sleep.c b/arch/x86/kernel/acpi/sleep.c index ca13851f0570..26b7256f590f 100644 --- a/arch/x86/kernel/acpi/sleep.c +++ b/arch/x86/kernel/acpi/sleep.c @@ -27,6 +27,17 @@ static char temp_stack[4096]; #endif /** + * acpi_get_wakeup_address - provide physical address for S3 wakeup + * + * Returns the physical address where the kernel should be resumed after the + * system awakes from S3, e.g. for programming into the firmware waking vector. + */ +unsigned long acpi_get_wakeup_address(void) +{ + return ((unsigned long)(real_mode_header->wakeup_start)); +} + +/** * x86_acpi_enter_sleep_state - enter sleep state * @state: Sleep state to enter. * diff --git a/arch/x86/kernel/acpi/sleep.h b/arch/x86/kernel/acpi/sleep.h index fbb60ca4255c..d06c2079b6c1 100644 --- a/arch/x86/kernel/acpi/sleep.h +++ b/arch/x86/kernel/acpi/sleep.h @@ -3,7 +3,7 @@ * Variables and functions used by the code in sleep.c */ -#include <asm/realmode.h> +#include <linux/linkage.h> extern unsigned long saved_video_mode; extern long saved_magic; diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c index 9ec463fe96f2..34360ca301a2 100644 --- a/arch/x86/kernel/alternative.c +++ b/arch/x86/kernel/alternative.c @@ -936,44 +936,81 @@ static void do_sync_core(void *info) sync_core(); } -static struct bp_patching_desc { +void text_poke_sync(void) +{ + on_each_cpu(do_sync_core, NULL, 1); +} + +struct text_poke_loc { + s32 rel_addr; /* addr := _stext + rel_addr */ + s32 rel32; + u8 opcode; + const u8 text[POKE_MAX_OPCODE_SIZE]; +}; + +struct bp_patching_desc { struct text_poke_loc *vec; int nr_entries; -} bp_patching; + atomic_t refs; +}; + +static struct bp_patching_desc *bp_desc; + +static inline struct bp_patching_desc *try_get_desc(struct bp_patching_desc **descp) +{ + struct bp_patching_desc *desc = READ_ONCE(*descp); /* rcu_dereference */ + + if (!desc || !atomic_inc_not_zero(&desc->refs)) + return NULL; + + return desc; +} + +static inline void put_desc(struct bp_patching_desc *desc) +{ + smp_mb__before_atomic(); + atomic_dec(&desc->refs); +} -static int patch_cmp(const void *key, const void *elt) +static inline void *text_poke_addr(struct text_poke_loc *tp) +{ + return _stext + tp->rel_addr; +} + +static int notrace patch_cmp(const void *key, const void *elt) { struct text_poke_loc *tp = (struct text_poke_loc *) elt; - if (key < tp->addr) + if (key < text_poke_addr(tp)) return -1; - if (key > tp->addr) + if (key > text_poke_addr(tp)) return 1; return 0; } NOKPROBE_SYMBOL(patch_cmp); -int poke_int3_handler(struct pt_regs *regs) +int notrace poke_int3_handler(struct pt_regs *regs) { + struct bp_patching_desc *desc; struct text_poke_loc *tp; + int len, ret = 0; void *ip; + if (user_mode(regs)) + return 0; + /* * Having observed our INT3 instruction, we now must observe - * bp_patching.nr_entries. + * bp_desc: * - * nr_entries != 0 INT3 + * bp_desc = desc INT3 * WMB RMB - * write INT3 if (nr_entries) - * - * Idem for other elements in bp_patching. + * write INT3 if (desc) */ smp_rmb(); - if (likely(!bp_patching.nr_entries)) - return 0; - - if (user_mode(regs)) + desc = try_get_desc(&bp_desc); + if (!desc) return 0; /* @@ -984,19 +1021,20 @@ int poke_int3_handler(struct pt_regs *regs) /* * Skip the binary search if there is a single member in the vector. */ - if (unlikely(bp_patching.nr_entries > 1)) { - tp = bsearch(ip, bp_patching.vec, bp_patching.nr_entries, + if (unlikely(desc->nr_entries > 1)) { + tp = bsearch(ip, desc->vec, desc->nr_entries, sizeof(struct text_poke_loc), patch_cmp); if (!tp) - return 0; + goto out_put; } else { - tp = bp_patching.vec; - if (tp->addr != ip) - return 0; + tp = desc->vec; + if (text_poke_addr(tp) != ip) + goto out_put; } - ip += tp->len; + len = text_opcode_size(tp->opcode); + ip += len; switch (tp->opcode) { case INT3_INSN_OPCODE: @@ -1004,7 +1042,7 @@ int poke_int3_handler(struct pt_regs *regs) * Someone poked an explicit INT3, they'll want to handle it, * do not consume. */ - return 0; + goto out_put; case CALL_INSN_OPCODE: int3_emulate_call(regs, (long)ip + tp->rel32); @@ -1019,10 +1057,18 @@ int poke_int3_handler(struct pt_regs *regs) BUG(); } - return 1; + ret = 1; + +out_put: + put_desc(desc); + return ret; } NOKPROBE_SYMBOL(poke_int3_handler); +#define TP_VEC_MAX (PAGE_SIZE / sizeof(struct text_poke_loc)) +static struct text_poke_loc tp_vec[TP_VEC_MAX]; +static int tp_vec_nr; + /** * text_poke_bp_batch() -- update instructions on live kernel on SMP * @tp: vector of instructions to patch @@ -1044,16 +1090,20 @@ NOKPROBE_SYMBOL(poke_int3_handler); * replacing opcode * - sync cores */ -void text_poke_bp_batch(struct text_poke_loc *tp, unsigned int nr_entries) +static void text_poke_bp_batch(struct text_poke_loc *tp, unsigned int nr_entries) { + struct bp_patching_desc desc = { + .vec = tp, + .nr_entries = nr_entries, + .refs = ATOMIC_INIT(1), + }; unsigned char int3 = INT3_INSN_OPCODE; unsigned int i; int do_sync; lockdep_assert_held(&text_mutex); - bp_patching.vec = tp; - bp_patching.nr_entries = nr_entries; + smp_store_release(&bp_desc, &desc); /* rcu_assign_pointer */ /* * Corresponding read barrier in int3 notifier for making sure the @@ -1065,18 +1115,20 @@ void text_poke_bp_batch(struct text_poke_loc *tp, unsigned int nr_entries) * First step: add a int3 trap to the address that will be patched. */ for (i = 0; i < nr_entries; i++) - text_poke(tp[i].addr, &int3, sizeof(int3)); + text_poke(text_poke_addr(&tp[i]), &int3, INT3_INSN_SIZE); - on_each_cpu(do_sync_core, NULL, 1); + text_poke_sync(); /* * Second step: update all but the first byte of the patched range. */ for (do_sync = 0, i = 0; i < nr_entries; i++) { - if (tp[i].len - sizeof(int3) > 0) { - text_poke((char *)tp[i].addr + sizeof(int3), - (const char *)tp[i].text + sizeof(int3), - tp[i].len - sizeof(int3)); + int len = text_opcode_size(tp[i].opcode); + + if (len - INT3_INSN_SIZE > 0) { + text_poke(text_poke_addr(&tp[i]) + INT3_INSN_SIZE, + (const char *)tp[i].text + INT3_INSN_SIZE, + len - INT3_INSN_SIZE); do_sync++; } } @@ -1087,7 +1139,7 @@ void text_poke_bp_batch(struct text_poke_loc *tp, unsigned int nr_entries) * not necessary and we'd be safe even without it. But * better safe than sorry (plus there's not only Intel). */ - on_each_cpu(do_sync_core, NULL, 1); + text_poke_sync(); } /* @@ -1098,19 +1150,20 @@ void text_poke_bp_batch(struct text_poke_loc *tp, unsigned int nr_entries) if (tp[i].text[0] == INT3_INSN_OPCODE) continue; - text_poke(tp[i].addr, tp[i].text, sizeof(int3)); + text_poke(text_poke_addr(&tp[i]), tp[i].text, INT3_INSN_SIZE); do_sync++; } if (do_sync) - on_each_cpu(do_sync_core, NULL, 1); + text_poke_sync(); /* - * sync_core() implies an smp_mb() and orders this store against - * the writing of the new instruction. + * Remove and synchronize_rcu(), except we have a very primitive + * refcount based completion. */ - bp_patching.vec = NULL; - bp_patching.nr_entries = 0; + WRITE_ONCE(bp_desc, NULL); /* RCU_INIT_POINTER */ + if (!atomic_dec_and_test(&desc.refs)) + atomic_cond_read_acquire(&desc.refs, !VAL); } void text_poke_loc_init(struct text_poke_loc *tp, void *addr, @@ -1118,11 +1171,7 @@ void text_poke_loc_init(struct text_poke_loc *tp, void *addr, { struct insn insn; - if (!opcode) - opcode = (void *)tp->text; - else - memcpy((void *)tp->text, opcode, len); - + memcpy((void *)tp->text, opcode, len); if (!emulate) emulate = opcode; @@ -1132,8 +1181,7 @@ void text_poke_loc_init(struct text_poke_loc *tp, void *addr, BUG_ON(!insn_complete(&insn)); BUG_ON(len != insn.length); - tp->addr = addr; - tp->len = len; + tp->rel_addr = addr - (void *)_stext; tp->opcode = insn.opcode.bytes[0]; switch (tp->opcode) { @@ -1167,6 +1215,55 @@ void text_poke_loc_init(struct text_poke_loc *tp, void *addr, } } +/* + * We hard rely on the tp_vec being ordered; ensure this is so by flushing + * early if needed. + */ +static bool tp_order_fail(void *addr) +{ + struct text_poke_loc *tp; + + if (!tp_vec_nr) + return false; + + if (!addr) /* force */ + return true; + + tp = &tp_vec[tp_vec_nr - 1]; + if ((unsigned long)text_poke_addr(tp) > (unsigned long)addr) + return true; + + return false; +} + +static void text_poke_flush(void *addr) +{ + if (tp_vec_nr == TP_VEC_MAX || tp_order_fail(addr)) { + text_poke_bp_batch(tp_vec, tp_vec_nr); + tp_vec_nr = 0; + } +} + +void text_poke_finish(void) +{ + text_poke_flush(NULL); +} + +void __ref text_poke_queue(void *addr, const void *opcode, size_t len, const void *emulate) +{ + struct text_poke_loc *tp; + + if (unlikely(system_state == SYSTEM_BOOTING)) { + text_poke_early(addr, opcode, len); + return; + } + + text_poke_flush(addr); + + tp = &tp_vec[tp_vec_nr++]; + text_poke_loc_init(tp, addr, opcode, len, emulate); +} + /** * text_poke_bp() -- update instructions on live kernel on SMP * @addr: address to patch @@ -1178,10 +1275,15 @@ void text_poke_loc_init(struct text_poke_loc *tp, void *addr, * dynamically allocated memory. This function should be used when it is * not possible to allocate memory. */ -void text_poke_bp(void *addr, const void *opcode, size_t len, const void *emulate) +void __ref text_poke_bp(void *addr, const void *opcode, size_t len, const void *emulate) { struct text_poke_loc tp; + if (unlikely(system_state == SYSTEM_BOOTING)) { + text_poke_early(addr, opcode, len); + return; + } + text_poke_loc_init(&tp, addr, opcode, len, emulate); text_poke_bp_batch(&tp, 1); } diff --git a/arch/x86/kernel/amd_nb.c b/arch/x86/kernel/amd_nb.c index 251c795b4eb3..69aed0ebbdfc 100644 --- a/arch/x86/kernel/amd_nb.c +++ b/arch/x86/kernel/amd_nb.c @@ -22,6 +22,7 @@ #define PCI_DEVICE_ID_AMD_17H_M10H_DF_F4 0x15ec #define PCI_DEVICE_ID_AMD_17H_M30H_DF_F4 0x1494 #define PCI_DEVICE_ID_AMD_17H_M70H_DF_F4 0x1444 +#define PCI_DEVICE_ID_AMD_19H_DF_F4 0x1654 /* Protect the PCI config register pairs used for SMN and DF indirect access. */ static DEFINE_MUTEX(smn_mutex); @@ -52,6 +53,7 @@ const struct pci_device_id amd_nb_misc_ids[] = { { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_17H_M30H_DF_F3) }, { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_CNB17H_F3) }, { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_17H_M70H_DF_F3) }, + { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_19H_DF_F3) }, {} }; EXPORT_SYMBOL_GPL(amd_nb_misc_ids); @@ -66,6 +68,7 @@ static const struct pci_device_id amd_nb_link_ids[] = { { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_17H_M10H_DF_F4) }, { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_17H_M30H_DF_F4) }, { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_17H_M70H_DF_F4) }, + { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_19H_DF_F4) }, { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_CNB17H_F4) }, {} }; diff --git a/arch/x86/kernel/apb_timer.c b/arch/x86/kernel/apb_timer.c index 5da106f84e84..fe698f96617c 100644 --- a/arch/x86/kernel/apb_timer.c +++ b/arch/x86/kernel/apb_timer.c @@ -95,7 +95,7 @@ static inline void apbt_set_mapping(void) printk(KERN_WARNING "No timer base from SFI, use default\n"); apbt_address = APBT_DEFAULT_BASE; } - apbt_virt_address = ioremap_nocache(apbt_address, APBT_MMAP_SIZE); + apbt_virt_address = ioremap(apbt_address, APBT_MMAP_SIZE); if (!apbt_virt_address) { pr_debug("Failed mapping APBT phy address at %lu\n",\ (unsigned long)apbt_address); diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c index d5b51a740524..ad53b2abc859 100644 --- a/arch/x86/kernel/apic/x2apic_uv_x.c +++ b/arch/x86/kernel/apic/x2apic_uv_x.c @@ -1493,65 +1493,34 @@ static void check_efi_reboot(void) } /* Setup user proc fs files */ -static int proc_hubbed_show(struct seq_file *file, void *data) +static int __maybe_unused proc_hubbed_show(struct seq_file *file, void *data) { seq_printf(file, "0x%x\n", uv_hubbed_system); return 0; } -static int proc_hubless_show(struct seq_file *file, void *data) +static int __maybe_unused proc_hubless_show(struct seq_file *file, void *data) { seq_printf(file, "0x%x\n", uv_hubless_system); return 0; } -static int proc_oemid_show(struct seq_file *file, void *data) +static int __maybe_unused proc_oemid_show(struct seq_file *file, void *data) { seq_printf(file, "%s/%s\n", oem_id, oem_table_id); return 0; } -static int proc_hubbed_open(struct inode *inode, struct file *file) -{ - return single_open(file, proc_hubbed_show, (void *)NULL); -} - -static int proc_hubless_open(struct inode *inode, struct file *file) -{ - return single_open(file, proc_hubless_show, (void *)NULL); -} - -static int proc_oemid_open(struct inode *inode, struct file *file) -{ - return single_open(file, proc_oemid_show, (void *)NULL); -} - -/* (struct is "non-const" as open function is set at runtime) */ -static struct file_operations proc_version_fops = { - .read = seq_read, - .llseek = seq_lseek, - .release = single_release, -}; - -static const struct file_operations proc_oemid_fops = { - .open = proc_oemid_open, - .read = seq_read, - .llseek = seq_lseek, - .release = single_release, -}; - static __init void uv_setup_proc_files(int hubless) { struct proc_dir_entry *pde; - char *name = hubless ? "hubless" : "hubbed"; pde = proc_mkdir(UV_PROC_NODE, NULL); - proc_create("oemid", 0, pde, &proc_oemid_fops); - proc_create(name, 0, pde, &proc_version_fops); + proc_create_single("oemid", 0, pde, proc_oemid_show); if (hubless) - proc_version_fops.open = proc_hubless_open; + proc_create_single("hubless", 0, pde, proc_hubless_show); else - proc_version_fops.open = proc_hubbed_open; + proc_create_single("hubbed", 0, pde, proc_hubbed_show); } /* Initialize UV hubless systems */ diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile index 890f60083eca..7dc4ad68eb41 100644 --- a/arch/x86/kernel/cpu/Makefile +++ b/arch/x86/kernel/cpu/Makefile @@ -29,6 +29,7 @@ obj-y += umwait.o obj-$(CONFIG_PROC_FS) += proc.o obj-$(CONFIG_X86_FEATURE_NAMES) += capflags.o powerflags.o +obj-$(CONFIG_IA32_FEAT_CTL) += feat_ctl.o ifdef CONFIG_CPU_SUP_INTEL obj-y += intel.o intel_pconfig.o tsx.o obj-$(CONFIG_PM) += intel_epb.o @@ -53,11 +54,12 @@ obj-$(CONFIG_ACRN_GUEST) += acrn.o ifdef CONFIG_X86_FEATURE_NAMES quiet_cmd_mkcapflags = MKCAP $@ - cmd_mkcapflags = $(CONFIG_SHELL) $(srctree)/$(src)/mkcapflags.sh $< $@ + cmd_mkcapflags = $(CONFIG_SHELL) $(srctree)/$(src)/mkcapflags.sh $@ $^ cpufeature = $(src)/../../include/asm/cpufeatures.h +vmxfeature = $(src)/../../include/asm/vmxfeatures.h -$(obj)/capflags.c: $(cpufeature) $(src)/mkcapflags.sh FORCE +$(obj)/capflags.c: $(cpufeature) $(vmxfeature) $(src)/mkcapflags.sh FORCE $(call if_changed,mkcapflags) endif targets += capflags.c diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index 62c30279be77..ac83a0fef628 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -319,13 +319,6 @@ static void legacy_fixup_core_id(struct cpuinfo_x86 *c) c->cpu_core_id %= cus_per_node; } - -static void amd_get_topology_early(struct cpuinfo_x86 *c) -{ - if (cpu_has(c, X86_FEATURE_TOPOEXT)) - smp_num_siblings = ((cpuid_ebx(0x8000001e) >> 8) & 0xff) + 1; -} - /* * Fixup core topology information for * (1) AMD multi-node processors @@ -717,7 +710,8 @@ static void early_init_amd(struct cpuinfo_x86 *c) } } - amd_get_topology_early(c); + if (cpu_has(c, X86_FEATURE_TOPOEXT)) + smp_num_siblings = ((cpuid_ebx(0x8000001e) >> 8) & 0xff) + 1; } static void init_amd_k8(struct cpuinfo_x86 *c) diff --git a/arch/x86/kernel/cpu/bugs.c b/arch/x86/kernel/cpu/bugs.c index 8bf64899f56a..ed54b3b21c39 100644 --- a/arch/x86/kernel/cpu/bugs.c +++ b/arch/x86/kernel/cpu/bugs.c @@ -286,6 +286,13 @@ early_param("mds", mds_cmdline); #undef pr_fmt #define pr_fmt(fmt) "TAA: " fmt +enum taa_mitigations { + TAA_MITIGATION_OFF, + TAA_MITIGATION_UCODE_NEEDED, + TAA_MITIGATION_VERW, + TAA_MITIGATION_TSX_DISABLED, +}; + /* Default mitigation for TAA-affected CPUs */ static enum taa_mitigations taa_mitigation __ro_after_init = TAA_MITIGATION_VERW; static bool taa_nosmt __ro_after_init; diff --git a/arch/x86/kernel/cpu/centaur.c b/arch/x86/kernel/cpu/centaur.c index 14433ff5b828..426792565d86 100644 --- a/arch/x86/kernel/cpu/centaur.c +++ b/arch/x86/kernel/cpu/centaur.c @@ -18,13 +18,6 @@ #define RNG_ENABLED (1 << 3) #define RNG_ENABLE (1 << 6) /* MSR_VIA_RNG */ -#define X86_VMX_FEATURE_PROC_CTLS_TPR_SHADOW 0x00200000 -#define X86_VMX_FEATURE_PROC_CTLS_VNMI 0x00400000 -#define X86_VMX_FEATURE_PROC_CTLS_2ND_CTLS 0x80000000 -#define X86_VMX_FEATURE_PROC_CTLS2_VIRT_APIC 0x00000001 -#define X86_VMX_FEATURE_PROC_CTLS2_EPT 0x00000002 -#define X86_VMX_FEATURE_PROC_CTLS2_VPID 0x00000020 - static void init_c3(struct cpuinfo_x86 *c) { u32 lo, hi; @@ -71,8 +64,6 @@ static void init_c3(struct cpuinfo_x86 *c) c->x86_cache_alignment = c->x86_clflush_size * 2; set_cpu_cap(c, X86_FEATURE_REP_GOOD); } - - cpu_detect_cache_sizes(c); } enum { @@ -119,31 +110,6 @@ static void early_init_centaur(struct cpuinfo_x86 *c) } } -static void centaur_detect_vmx_virtcap(struct cpuinfo_x86 *c) -{ - u32 vmx_msr_low, vmx_msr_high, msr_ctl, msr_ctl2; - - rdmsr(MSR_IA32_VMX_PROCBASED_CTLS, vmx_msr_low, vmx_msr_high); - msr_ctl = vmx_msr_high | vmx_msr_low; - - if (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_TPR_SHADOW) - set_cpu_cap(c, X86_FEATURE_TPR_SHADOW); - if (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_VNMI) - set_cpu_cap(c, X86_FEATURE_VNMI); - if (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_2ND_CTLS) { - rdmsr(MSR_IA32_VMX_PROCBASED_CTLS2, - vmx_msr_low, vmx_msr_high); - msr_ctl2 = vmx_msr_high | vmx_msr_low; - if ((msr_ctl2 & X86_VMX_FEATURE_PROC_CTLS2_VIRT_APIC) && - (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_TPR_SHADOW)) - set_cpu_cap(c, X86_FEATURE_FLEXPRIORITY); - if (msr_ctl2 & X86_VMX_FEATURE_PROC_CTLS2_EPT) - set_cpu_cap(c, X86_FEATURE_EPT); - if (msr_ctl2 & X86_VMX_FEATURE_PROC_CTLS2_VPID) - set_cpu_cap(c, X86_FEATURE_VPID); - } -} - static void init_centaur(struct cpuinfo_x86 *c) { #ifdef CONFIG_X86_32 @@ -250,8 +216,7 @@ static void init_centaur(struct cpuinfo_x86 *c) set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); #endif - if (cpu_has(c, X86_FEATURE_VMX)) - centaur_detect_vmx_virtcap(c); + init_ia32_feat_ctl(c); } #ifdef CONFIG_X86_32 diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 2e4d90294fe6..86b8241c8209 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -14,6 +14,7 @@ #include <linux/sched/mm.h> #include <linux/sched/clock.h> #include <linux/sched/task.h> +#include <linux/sched/smt.h> #include <linux/init.h> #include <linux/kprobes.h> #include <linux/kgdb.h> @@ -49,7 +50,7 @@ #include <asm/cpu.h> #include <asm/mce.h> #include <asm/msr.h> -#include <asm/pat.h> +#include <asm/memtype.h> #include <asm/microcode.h> #include <asm/microcode_intel.h> #include <asm/intel-family.h> @@ -1023,6 +1024,7 @@ static void identify_cpu_without_cpuid(struct cpuinfo_x86 *c) #define MSBDS_ONLY BIT(5) #define NO_SWAPGS BIT(6) #define NO_ITLB_MULTIHIT BIT(7) +#define NO_SPECTRE_V2 BIT(8) #define VULNWL(_vendor, _family, _model, _whitelist) \ { X86_VENDOR_##_vendor, _family, _model, X86_FEATURE_ANY, _whitelist } @@ -1084,6 +1086,10 @@ static const __initconst struct x86_cpu_id cpu_vuln_whitelist[] = { /* FAMILY_ANY must be last, otherwise 0x0f - 0x12 matches won't work */ VULNWL_AMD(X86_FAMILY_ANY, NO_MELTDOWN | NO_L1TF | NO_MDS | NO_SWAPGS | NO_ITLB_MULTIHIT), VULNWL_HYGON(X86_FAMILY_ANY, NO_MELTDOWN | NO_L1TF | NO_MDS | NO_SWAPGS | NO_ITLB_MULTIHIT), + + /* Zhaoxin Family 7 */ + VULNWL(CENTAUR, 7, X86_MODEL_ANY, NO_SPECTRE_V2 | NO_SWAPGS), + VULNWL(ZHAOXIN, 7, X86_MODEL_ANY, NO_SPECTRE_V2 | NO_SWAPGS), {} }; @@ -1116,7 +1122,9 @@ static void __init cpu_set_bug_bits(struct cpuinfo_x86 *c) return; setup_force_cpu_bug(X86_BUG_SPECTRE_V1); - setup_force_cpu_bug(X86_BUG_SPECTRE_V2); + + if (!cpu_matches(NO_SPECTRE_V2)) + setup_force_cpu_bug(X86_BUG_SPECTRE_V2); if (!cpu_matches(NO_SSB) && !(ia32_cap & ARCH_CAP_SSB_NO) && !cpu_has(c, X86_FEATURE_AMD_SSB_NO)) @@ -1449,6 +1457,9 @@ static void identify_cpu(struct cpuinfo_x86 *c) #endif c->x86_cache_alignment = c->x86_clflush_size; memset(&c->x86_capability, 0, sizeof(c->x86_capability)); +#ifdef CONFIG_X86_VMX_FEATURE_NAMES + memset(&c->vmx_capability, 0, sizeof(c->vmx_capability)); +#endif generic_identify(c); diff --git a/arch/x86/kernel/cpu/cpu.h b/arch/x86/kernel/cpu/cpu.h index 38ab6e115eac..37fdefd14f28 100644 --- a/arch/x86/kernel/cpu/cpu.h +++ b/arch/x86/kernel/cpu/cpu.h @@ -80,4 +80,8 @@ extern void x86_spec_ctrl_setup_ap(void); extern u64 x86_read_arch_cap_msr(void); +#ifdef CONFIG_IA32_FEAT_CTL +void init_ia32_feat_ctl(struct cpuinfo_x86 *c); +#endif + #endif /* ARCH_X86_CPU_H */ diff --git a/arch/x86/kernel/cpu/feat_ctl.c b/arch/x86/kernel/cpu/feat_ctl.c new file mode 100644 index 000000000000..0268185bef94 --- /dev/null +++ b/arch/x86/kernel/cpu/feat_ctl.c @@ -0,0 +1,145 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/tboot.h> + +#include <asm/cpufeature.h> +#include <asm/msr-index.h> +#include <asm/processor.h> +#include <asm/vmx.h> + +#undef pr_fmt +#define pr_fmt(fmt) "x86/cpu: " fmt + +#ifdef CONFIG_X86_VMX_FEATURE_NAMES +enum vmx_feature_leafs { + MISC_FEATURES = 0, + PRIMARY_CTLS, + SECONDARY_CTLS, + NR_VMX_FEATURE_WORDS, +}; + +#define VMX_F(x) BIT(VMX_FEATURE_##x & 0x1f) + +static void init_vmx_capabilities(struct cpuinfo_x86 *c) +{ + u32 supported, funcs, ept, vpid, ign; + + BUILD_BUG_ON(NVMXINTS != NR_VMX_FEATURE_WORDS); + + /* + * The high bits contain the allowed-1 settings, i.e. features that can + * be turned on. The low bits contain the allowed-0 settings, i.e. + * features that can be turned off. Ignore the allowed-0 settings, + * if a feature can be turned on then it's supported. + * + * Use raw rdmsr() for primary processor controls and pin controls MSRs + * as they exist on any CPU that supports VMX, i.e. we want the WARN if + * the RDMSR faults. + */ + rdmsr(MSR_IA32_VMX_PROCBASED_CTLS, ign, supported); + c->vmx_capability[PRIMARY_CTLS] = supported; + + rdmsr_safe(MSR_IA32_VMX_PROCBASED_CTLS2, &ign, &supported); + c->vmx_capability[SECONDARY_CTLS] = supported; + + rdmsr(MSR_IA32_VMX_PINBASED_CTLS, ign, supported); + rdmsr_safe(MSR_IA32_VMX_VMFUNC, &ign, &funcs); + + /* + * Except for EPT+VPID, which enumerates support for both in a single + * MSR, low for EPT, high for VPID. + */ + rdmsr_safe(MSR_IA32_VMX_EPT_VPID_CAP, &ept, &vpid); + + /* Pin, EPT, VPID and VM-Func are merged into a single word. */ + WARN_ON_ONCE(supported >> 16); + WARN_ON_ONCE(funcs >> 4); + c->vmx_capability[MISC_FEATURES] = (supported & 0xffff) | + ((vpid & 0x1) << 16) | + ((funcs & 0xf) << 28); + + /* EPT bits are full on scattered and must be manually handled. */ + if (ept & VMX_EPT_EXECUTE_ONLY_BIT) + c->vmx_capability[MISC_FEATURES] |= VMX_F(EPT_EXECUTE_ONLY); + if (ept & VMX_EPT_AD_BIT) + c->vmx_capability[MISC_FEATURES] |= VMX_F(EPT_AD); + if (ept & VMX_EPT_1GB_PAGE_BIT) + c->vmx_capability[MISC_FEATURES] |= VMX_F(EPT_1GB); + + /* Synthetic APIC features that are aggregates of multiple features. */ + if ((c->vmx_capability[PRIMARY_CTLS] & VMX_F(VIRTUAL_TPR)) && + (c->vmx_capability[SECONDARY_CTLS] & VMX_F(VIRT_APIC_ACCESSES))) + c->vmx_capability[MISC_FEATURES] |= VMX_F(FLEXPRIORITY); + + if ((c->vmx_capability[PRIMARY_CTLS] & VMX_F(VIRTUAL_TPR)) && + (c->vmx_capability[SECONDARY_CTLS] & VMX_F(APIC_REGISTER_VIRT)) && + (c->vmx_capability[SECONDARY_CTLS] & VMX_F(VIRT_INTR_DELIVERY)) && + (c->vmx_capability[MISC_FEATURES] & VMX_F(POSTED_INTR))) + c->vmx_capability[MISC_FEATURES] |= VMX_F(APICV); + + /* Set the synthetic cpufeatures to preserve /proc/cpuinfo's ABI. */ + if (c->vmx_capability[PRIMARY_CTLS] & VMX_F(VIRTUAL_TPR)) + set_cpu_cap(c, X86_FEATURE_TPR_SHADOW); + if (c->vmx_capability[MISC_FEATURES] & VMX_F(FLEXPRIORITY)) + set_cpu_cap(c, X86_FEATURE_FLEXPRIORITY); + if (c->vmx_capability[MISC_FEATURES] & VMX_F(VIRTUAL_NMIS)) + set_cpu_cap(c, X86_FEATURE_VNMI); + if (c->vmx_capability[SECONDARY_CTLS] & VMX_F(EPT)) + set_cpu_cap(c, X86_FEATURE_EPT); + if (c->vmx_capability[MISC_FEATURES] & VMX_F(EPT_AD)) + set_cpu_cap(c, X86_FEATURE_EPT_AD); + if (c->vmx_capability[MISC_FEATURES] & VMX_F(VPID)) + set_cpu_cap(c, X86_FEATURE_VPID); +} +#endif /* CONFIG_X86_VMX_FEATURE_NAMES */ + +void init_ia32_feat_ctl(struct cpuinfo_x86 *c) +{ + bool tboot = tboot_enabled(); + u64 msr; + + if (rdmsrl_safe(MSR_IA32_FEAT_CTL, &msr)) { + clear_cpu_cap(c, X86_FEATURE_VMX); + return; + } + + if (msr & FEAT_CTL_LOCKED) + goto update_caps; + + /* + * Ignore whatever value BIOS left in the MSR to avoid enabling random + * features or faulting on the WRMSR. + */ + msr = FEAT_CTL_LOCKED; + + /* + * Enable VMX if and only if the kernel may do VMXON at some point, + * i.e. KVM is enabled, to avoid unnecessarily adding an attack vector + * for the kernel, e.g. using VMX to hide malicious code. + */ + if (cpu_has(c, X86_FEATURE_VMX) && IS_ENABLED(CONFIG_KVM_INTEL)) { + msr |= FEAT_CTL_VMX_ENABLED_OUTSIDE_SMX; + + if (tboot) + msr |= FEAT_CTL_VMX_ENABLED_INSIDE_SMX; + } + + wrmsrl(MSR_IA32_FEAT_CTL, msr); + +update_caps: + set_cpu_cap(c, X86_FEATURE_MSR_IA32_FEAT_CTL); + + if (!cpu_has(c, X86_FEATURE_VMX)) + return; + + if ( (tboot && !(msr & FEAT_CTL_VMX_ENABLED_INSIDE_SMX)) || + (!tboot && !(msr & FEAT_CTL_VMX_ENABLED_OUTSIDE_SMX))) { + if (IS_ENABLED(CONFIG_KVM_INTEL)) + pr_err_once("VMX (%s TXT) disabled by BIOS\n", + tboot ? "inside" : "outside"); + clear_cpu_cap(c, X86_FEATURE_VMX); + } else { +#ifdef CONFIG_X86_VMX_FEATURE_NAMES + init_vmx_capabilities(c); +#endif + } +} diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index 4a900804a023..57473e2c0869 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -494,52 +494,6 @@ static void srat_detect_node(struct cpuinfo_x86 *c) #endif } -static void detect_vmx_virtcap(struct cpuinfo_x86 *c) -{ - /* Intel VMX MSR indicated features */ -#define X86_VMX_FEATURE_PROC_CTLS_TPR_SHADOW 0x00200000 -#define X86_VMX_FEATURE_PROC_CTLS_VNMI 0x00400000 -#define X86_VMX_FEATURE_PROC_CTLS_2ND_CTLS 0x80000000 -#define X86_VMX_FEATURE_PROC_CTLS2_VIRT_APIC 0x00000001 -#define X86_VMX_FEATURE_PROC_CTLS2_EPT 0x00000002 -#define X86_VMX_FEATURE_PROC_CTLS2_VPID 0x00000020 -#define x86_VMX_FEATURE_EPT_CAP_AD 0x00200000 - - u32 vmx_msr_low, vmx_msr_high, msr_ctl, msr_ctl2; - u32 msr_vpid_cap, msr_ept_cap; - - clear_cpu_cap(c, X86_FEATURE_TPR_SHADOW); - clear_cpu_cap(c, X86_FEATURE_VNMI); - clear_cpu_cap(c, X86_FEATURE_FLEXPRIORITY); - clear_cpu_cap(c, X86_FEATURE_EPT); - clear_cpu_cap(c, X86_FEATURE_VPID); - clear_cpu_cap(c, X86_FEATURE_EPT_AD); - - rdmsr(MSR_IA32_VMX_PROCBASED_CTLS, vmx_msr_low, vmx_msr_high); - msr_ctl = vmx_msr_high | vmx_msr_low; - if (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_TPR_SHADOW) - set_cpu_cap(c, X86_FEATURE_TPR_SHADOW); - if (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_VNMI) - set_cpu_cap(c, X86_FEATURE_VNMI); - if (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_2ND_CTLS) { - rdmsr(MSR_IA32_VMX_PROCBASED_CTLS2, - vmx_msr_low, vmx_msr_high); - msr_ctl2 = vmx_msr_high | vmx_msr_low; - if ((msr_ctl2 & X86_VMX_FEATURE_PROC_CTLS2_VIRT_APIC) && - (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_TPR_SHADOW)) - set_cpu_cap(c, X86_FEATURE_FLEXPRIORITY); - if (msr_ctl2 & X86_VMX_FEATURE_PROC_CTLS2_EPT) { - set_cpu_cap(c, X86_FEATURE_EPT); - rdmsr(MSR_IA32_VMX_EPT_VPID_CAP, - msr_ept_cap, msr_vpid_cap); - if (msr_ept_cap & x86_VMX_FEATURE_EPT_CAP_AD) - set_cpu_cap(c, X86_FEATURE_EPT_AD); - } - if (msr_ctl2 & X86_VMX_FEATURE_PROC_CTLS2_VPID) - set_cpu_cap(c, X86_FEATURE_VPID); - } -} - #define MSR_IA32_TME_ACTIVATE 0x982 /* Helpers to access TME_ACTIVATE MSR */ @@ -755,8 +709,7 @@ static void init_intel(struct cpuinfo_x86 *c) /* Work around errata */ srat_detect_node(c); - if (cpu_has(c, X86_FEATURE_VMX)) - detect_vmx_virtcap(c); + init_ia32_feat_ctl(c); if (cpu_has(c, X86_FEATURE_TME)) detect_tme(c); diff --git a/arch/x86/kernel/cpu/mce/amd.c b/arch/x86/kernel/cpu/mce/amd.c index d6cf5c18a7e0..b3a50d962851 100644 --- a/arch/x86/kernel/cpu/mce/amd.c +++ b/arch/x86/kernel/cpu/mce/amd.c @@ -78,6 +78,7 @@ struct smca_bank_name { static struct smca_bank_name smca_names[] = { [SMCA_LS] = { "load_store", "Load Store Unit" }, + [SMCA_LS_V2] = { "load_store", "Load Store Unit" }, [SMCA_IF] = { "insn_fetch", "Instruction Fetch Unit" }, [SMCA_L2_CACHE] = { "l2_cache", "L2 Cache" }, [SMCA_DE] = { "decode_unit", "Decode Unit" }, @@ -138,6 +139,7 @@ static struct smca_hwid smca_hwid_mcatypes[] = { /* ZN Core (HWID=0xB0) MCA types */ { SMCA_LS, HWID_MCATYPE(0xB0, 0x0), 0x1FFFFF }, + { SMCA_LS_V2, HWID_MCATYPE(0xB0, 0x10), 0xFFFFFF }, { SMCA_IF, HWID_MCATYPE(0xB0, 0x1), 0x3FFF }, { SMCA_L2_CACHE, HWID_MCATYPE(0xB0, 0x2), 0xF }, { SMCA_DE, HWID_MCATYPE(0xB0, 0x3), 0x1FF }, diff --git a/arch/x86/kernel/cpu/mce/core.c b/arch/x86/kernel/cpu/mce/core.c index 2e2a421c8528..2c4f949611e4 100644 --- a/arch/x86/kernel/cpu/mce/core.c +++ b/arch/x86/kernel/cpu/mce/core.c @@ -53,8 +53,6 @@ #include "internal.h" -static DEFINE_MUTEX(mce_log_mutex); - /* sysfs synchronization */ static DEFINE_MUTEX(mce_sysfs_mutex); @@ -156,19 +154,10 @@ void mce_log(struct mce *m) if (!mce_gen_pool_add(m)) irq_work_queue(&mce_irq_work); } - -void mce_inject_log(struct mce *m) -{ - mutex_lock(&mce_log_mutex); - mce_log(m); - mutex_unlock(&mce_log_mutex); -} -EXPORT_SYMBOL_GPL(mce_inject_log); - -static struct notifier_block mce_srao_nb; +EXPORT_SYMBOL_GPL(mce_log); /* - * We run the default notifier if we have only the SRAO, the first and the + * We run the default notifier if we have only the UC, the first and the * default notifier registered. I.e., the mandatory NUM_DEFAULT_NOTIFIERS * notifiers registered on the chain. */ @@ -594,26 +583,29 @@ static struct notifier_block first_nb = { .priority = MCE_PRIO_FIRST, }; -static int srao_decode_notifier(struct notifier_block *nb, unsigned long val, - void *data) +static int uc_decode_notifier(struct notifier_block *nb, unsigned long val, + void *data) { struct mce *mce = (struct mce *)data; unsigned long pfn; - if (!mce) + if (!mce || !mce_usable_address(mce)) return NOTIFY_DONE; - if (mce_usable_address(mce) && (mce->severity == MCE_AO_SEVERITY)) { - pfn = mce->addr >> PAGE_SHIFT; - if (!memory_failure(pfn, 0)) - set_mce_nospec(pfn); - } + if (mce->severity != MCE_AO_SEVERITY && + mce->severity != MCE_DEFERRED_SEVERITY) + return NOTIFY_DONE; + + pfn = mce->addr >> PAGE_SHIFT; + if (!memory_failure(pfn, 0)) + set_mce_nospec(pfn); return NOTIFY_OK; } -static struct notifier_block mce_srao_nb = { - .notifier_call = srao_decode_notifier, - .priority = MCE_PRIO_SRAO, + +static struct notifier_block mce_uc_nb = { + .notifier_call = uc_decode_notifier, + .priority = MCE_PRIO_UC, }; static int mce_default_notifier(struct notifier_block *nb, unsigned long val, @@ -763,26 +755,22 @@ bool machine_check_poll(enum mcp_flags flags, mce_banks_t *b) log_it: error_seen = true; - mce_read_aux(&m, i); + if (flags & MCP_DONTLOG) + goto clear_it; + mce_read_aux(&m, i); m.severity = mce_severity(&m, mca_cfg.tolerant, NULL, false); - /* * Don't get the IP here because it's unlikely to * have anything to do with the actual error location. */ - if (!(flags & MCP_DONTLOG) && !mca_cfg.dont_log_ce) - mce_log(&m); - else if (mce_usable_address(&m)) { - /* - * Although we skipped logging this, we still want - * to take action. Add to the pool so the registered - * notifiers will see it. - */ - if (!mce_gen_pool_add(&m)) - mce_schedule_work(); - } + if (mca_cfg.dont_log_ce && !mce_usable_address(&m)) + goto clear_it; + + mce_log(&m); + +clear_it: /* * Clear state for this bank. */ @@ -807,7 +795,7 @@ EXPORT_SYMBOL_GPL(machine_check_poll); static int mce_no_way_out(struct mce *m, char **msg, unsigned long *validp, struct pt_regs *regs) { - char *tmp; + char *tmp = *msg; int i; for (i = 0; i < this_cpu_read(mce_num_banks); i++) { @@ -1232,8 +1220,8 @@ void do_machine_check(struct pt_regs *regs, long error_code) DECLARE_BITMAP(toclear, MAX_NR_BANKS); struct mca_config *cfg = &mca_cfg; int cpu = smp_processor_id(); - char *msg = "Unknown"; struct mce m, *final; + char *msg = NULL; int worst = 0; /* @@ -1365,7 +1353,7 @@ void do_machine_check(struct pt_regs *regs, long error_code) ist_end_non_atomic(); } else { if (!fixup_exception(regs, X86_TRAP_MC, error_code, 0)) - mce_panic("Failed kernel mode recovery", &m, NULL); + mce_panic("Failed kernel mode recovery", &m, msg); } out_ist: @@ -2041,7 +2029,7 @@ int __init mcheck_init(void) { mcheck_intel_therm_init(); mce_register_decode_chain(&first_nb); - mce_register_decode_chain(&mce_srao_nb); + mce_register_decode_chain(&mce_uc_nb); mce_register_decode_chain(&mce_default_nb); mcheck_vendor_init_severity(); diff --git a/arch/x86/kernel/cpu/mce/inject.c b/arch/x86/kernel/cpu/mce/inject.c index 1f30117b24ba..3413b41b8d55 100644 --- a/arch/x86/kernel/cpu/mce/inject.c +++ b/arch/x86/kernel/cpu/mce/inject.c @@ -494,7 +494,7 @@ static void do_inject(void) i_mce.status |= MCI_STATUS_SYNDV; if (inj_type == SW_INJ) { - mce_inject_log(&i_mce); + mce_log(&i_mce); return; } diff --git a/arch/x86/kernel/cpu/mce/intel.c b/arch/x86/kernel/cpu/mce/intel.c index e270d0770134..5627b1091b85 100644 --- a/arch/x86/kernel/cpu/mce/intel.c +++ b/arch/x86/kernel/cpu/mce/intel.c @@ -115,15 +115,16 @@ static bool lmce_supported(void) /* * BIOS should indicate support for LMCE by setting bit 20 in - * IA32_FEATURE_CONTROL without which touching MCG_EXT_CTL will - * generate a #GP fault. + * IA32_FEAT_CTL without which touching MCG_EXT_CTL will generate a #GP + * fault. The MSR must also be locked for LMCE_ENABLED to take effect. + * WARN if the MSR isn't locked as init_ia32_feat_ctl() unconditionally + * locks the MSR in the event that it wasn't already locked by BIOS. */ - rdmsrl(MSR_IA32_FEATURE_CONTROL, tmp); - if ((tmp & (FEATURE_CONTROL_LOCKED | FEATURE_CONTROL_LMCE)) == - (FEATURE_CONTROL_LOCKED | FEATURE_CONTROL_LMCE)) - return true; + rdmsrl(MSR_IA32_FEAT_CTL, tmp); + if (WARN_ON_ONCE(!(tmp & FEAT_CTL_LOCKED))) + return false; - return false; + return tmp & FEAT_CTL_LMCE_ENABLED; } bool mce_intel_cmci_poll(void) diff --git a/arch/x86/kernel/cpu/mce/internal.h b/arch/x86/kernel/cpu/mce/internal.h index 842b273bce31..b785c0d0b590 100644 --- a/arch/x86/kernel/cpu/mce/internal.h +++ b/arch/x86/kernel/cpu/mce/internal.h @@ -84,8 +84,6 @@ static inline int apei_clear_mce(u64 record_id) } #endif -void mce_inject_log(struct mce *m); - /* * We consider records to be equivalent if bank+status+addr+misc all match. * This is only used when the system is going down because of a fatal error diff --git a/arch/x86/kernel/cpu/mce/therm_throt.c b/arch/x86/kernel/cpu/mce/therm_throt.c index 6c3e1c92f183..58b4ee3cda77 100644 --- a/arch/x86/kernel/cpu/mce/therm_throt.c +++ b/arch/x86/kernel/cpu/mce/therm_throt.c @@ -235,7 +235,7 @@ static void get_therm_status(int level, bool *proc_hot, u8 *temp) *temp = (msr_val >> 16) & 0x7F; } -static void throttle_active_work(struct work_struct *work) +static void __maybe_unused throttle_active_work(struct work_struct *work) { struct _thermal_state *state = container_of(to_delayed_work(work), struct _thermal_state, therm_work); diff --git a/arch/x86/kernel/cpu/mkcapflags.sh b/arch/x86/kernel/cpu/mkcapflags.sh index aed45b8895d5..1db560ed2ca3 100644 --- a/arch/x86/kernel/cpu/mkcapflags.sh +++ b/arch/x86/kernel/cpu/mkcapflags.sh @@ -6,8 +6,7 @@ set -e -IN=$1 -OUT=$2 +OUT=$1 dump_array() { @@ -15,6 +14,7 @@ dump_array() SIZE=$2 PFX=$3 POSTFIX=$4 + IN=$5 PFX_SZ=$(echo $PFX | wc -c) TABS="$(printf '\t\t\t\t\t')" @@ -57,11 +57,18 @@ trap 'rm "$OUT"' EXIT echo "#endif" echo "" - dump_array "x86_cap_flags" "NCAPINTS*32" "X86_FEATURE_" "" + dump_array "x86_cap_flags" "NCAPINTS*32" "X86_FEATURE_" "" $2 echo "" - dump_array "x86_bug_flags" "NBUGINTS*32" "X86_BUG_" "NCAPINTS*32" + dump_array "x86_bug_flags" "NBUGINTS*32" "X86_BUG_" "NCAPINTS*32" $2 + echo "" + echo "#ifdef CONFIG_X86_VMX_FEATURE_NAMES" + echo "#ifndef _ASM_X86_VMXFEATURES_H" + echo "#include <asm/vmxfeatures.h>" + echo "#endif" + dump_array "x86_vmx_flags" "NVMXINTS*32" "VMX_FEATURE_" "" $3 + echo "#endif /* CONFIG_X86_VMX_FEATURE_NAMES */" ) > $OUT trap - EXIT diff --git a/arch/x86/kernel/cpu/mtrr/generic.c b/arch/x86/kernel/cpu/mtrr/generic.c index aa5c064a6a22..51b9190c628b 100644 --- a/arch/x86/kernel/cpu/mtrr/generic.c +++ b/arch/x86/kernel/cpu/mtrr/generic.c @@ -15,7 +15,7 @@ #include <asm/tlbflush.h> #include <asm/mtrr.h> #include <asm/msr.h> -#include <asm/pat.h> +#include <asm/memtype.h> #include "mtrr.h" diff --git a/arch/x86/kernel/cpu/mtrr/if.c b/arch/x86/kernel/cpu/mtrr/if.c index 4d36dcc1cf87..da532f656a7b 100644 --- a/arch/x86/kernel/cpu/mtrr/if.c +++ b/arch/x86/kernel/cpu/mtrr/if.c @@ -101,9 +101,6 @@ mtrr_write(struct file *file, const char __user *buf, size_t len, loff_t * ppos) int length; size_t linelen; - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; - memset(line, 0, LINE_SIZE); len = min_t(size_t, len, LINE_SIZE - 1); @@ -226,8 +223,6 @@ mtrr_ioctl(struct file *file, unsigned int cmd, unsigned long __arg) #ifdef CONFIG_COMPAT case MTRRIOC32_ADD_ENTRY: #endif - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; err = mtrr_file_add(sentry.base, sentry.size, sentry.type, true, file, 0); @@ -236,24 +231,18 @@ mtrr_ioctl(struct file *file, unsigned int cmd, unsigned long __arg) #ifdef CONFIG_COMPAT case MTRRIOC32_SET_ENTRY: #endif - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; err = mtrr_add(sentry.base, sentry.size, sentry.type, false); break; case MTRRIOC_DEL_ENTRY: #ifdef CONFIG_COMPAT case MTRRIOC32_DEL_ENTRY: #endif - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; err = mtrr_file_del(sentry.base, sentry.size, file, 0); break; case MTRRIOC_KILL_ENTRY: #ifdef CONFIG_COMPAT case MTRRIOC32_KILL_ENTRY: #endif - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; err = mtrr_del(-1, sentry.base, sentry.size); break; case MTRRIOC_GET_ENTRY: @@ -279,8 +268,6 @@ mtrr_ioctl(struct file *file, unsigned int cmd, unsigned long __arg) #ifdef CONFIG_COMPAT case MTRRIOC32_ADD_PAGE_ENTRY: #endif - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; err = mtrr_file_add(sentry.base, sentry.size, sentry.type, true, file, 1); @@ -289,8 +276,6 @@ mtrr_ioctl(struct file *file, unsigned int cmd, unsigned long __arg) #ifdef CONFIG_COMPAT case MTRRIOC32_SET_PAGE_ENTRY: #endif - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; err = mtrr_add_page(sentry.base, sentry.size, sentry.type, false); break; @@ -298,16 +283,12 @@ mtrr_ioctl(struct file *file, unsigned int cmd, unsigned long __arg) #ifdef CONFIG_COMPAT case MTRRIOC32_DEL_PAGE_ENTRY: #endif - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; err = mtrr_file_del(sentry.base, sentry.size, file, 1); break; case MTRRIOC_KILL_PAGE_ENTRY: #ifdef CONFIG_COMPAT case MTRRIOC32_KILL_PAGE_ENTRY: #endif - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; err = mtrr_del_page(-1, sentry.base, sentry.size); break; case MTRRIOC_GET_PAGE_ENTRY: @@ -373,28 +354,6 @@ static int mtrr_close(struct inode *ino, struct file *file) return single_release(ino, file); } -static int mtrr_seq_show(struct seq_file *seq, void *offset); - -static int mtrr_open(struct inode *inode, struct file *file) -{ - if (!mtrr_if) - return -EIO; - if (!mtrr_if->get) - return -ENXIO; - return single_open(file, mtrr_seq_show, NULL); -} - -static const struct file_operations mtrr_fops = { - .owner = THIS_MODULE, - .open = mtrr_open, - .read = seq_read, - .llseek = seq_lseek, - .write = mtrr_write, - .unlocked_ioctl = mtrr_ioctl, - .compat_ioctl = mtrr_ioctl, - .release = mtrr_close, -}; - static int mtrr_seq_show(struct seq_file *seq, void *offset) { char factor; @@ -426,6 +385,28 @@ static int mtrr_seq_show(struct seq_file *seq, void *offset) return 0; } +static int mtrr_open(struct inode *inode, struct file *file) +{ + if (!mtrr_if) + return -EIO; + if (!mtrr_if->get) + return -ENXIO; + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + return single_open(file, mtrr_seq_show, NULL); +} + +static const struct file_operations mtrr_fops = { + .owner = THIS_MODULE, + .open = mtrr_open, + .read = seq_read, + .llseek = seq_lseek, + .write = mtrr_write, + .unlocked_ioctl = mtrr_ioctl, + .compat_ioctl = mtrr_ioctl, + .release = mtrr_close, +}; + static int __init mtrr_if_init(void) { struct cpuinfo_x86 *c = &boot_cpu_data; diff --git a/arch/x86/kernel/cpu/mtrr/mtrr.c b/arch/x86/kernel/cpu/mtrr/mtrr.c index 507039c20128..6a80f36b5d59 100644 --- a/arch/x86/kernel/cpu/mtrr/mtrr.c +++ b/arch/x86/kernel/cpu/mtrr/mtrr.c @@ -52,7 +52,7 @@ #include <asm/e820/api.h> #include <asm/mtrr.h> #include <asm/msr.h> -#include <asm/pat.h> +#include <asm/memtype.h> #include "mtrr.h" diff --git a/arch/x86/kernel/cpu/proc.c b/arch/x86/kernel/cpu/proc.c index cb2e49810d68..4eec8889b0ff 100644 --- a/arch/x86/kernel/cpu/proc.c +++ b/arch/x86/kernel/cpu/proc.c @@ -7,6 +7,10 @@ #include "cpu.h" +#ifdef CONFIG_X86_VMX_FEATURE_NAMES +extern const char * const x86_vmx_flags[NVMXINTS*32]; +#endif + /* * Get CPU information for use by the procfs. */ @@ -102,6 +106,17 @@ static int show_cpuinfo(struct seq_file *m, void *v) if (cpu_has(c, i) && x86_cap_flags[i] != NULL) seq_printf(m, " %s", x86_cap_flags[i]); +#ifdef CONFIG_X86_VMX_FEATURE_NAMES + if (cpu_has(c, X86_FEATURE_VMX) && c->vmx_capability[0]) { + seq_puts(m, "\nvmx flags\t:"); + for (i = 0; i < 32*NVMXINTS; i++) { + if (test_bit(i, (unsigned long *)c->vmx_capability) && + x86_vmx_flags[i] != NULL) + seq_printf(m, " %s", x86_vmx_flags[i]); + } + } +#endif + seq_puts(m, "\nbugs\t\t:"); for (i = 0; i < 32*NBUGINTS; i++) { unsigned int bug_bit = 32*NCAPINTS + i; diff --git a/arch/x86/kernel/cpu/resctrl/internal.h b/arch/x86/kernel/cpu/resctrl/internal.h index e49b77283924..181c992f448c 100644 --- a/arch/x86/kernel/cpu/resctrl/internal.h +++ b/arch/x86/kernel/cpu/resctrl/internal.h @@ -57,6 +57,7 @@ static inline struct rdt_fs_context *rdt_fc2context(struct fs_context *fc) } DECLARE_STATIC_KEY_FALSE(rdt_enable_key); +DECLARE_STATIC_KEY_FALSE(rdt_mon_enable_key); /** * struct mon_evt - Entry in the event list of a resource diff --git a/arch/x86/kernel/cpu/resctrl/monitor.c b/arch/x86/kernel/cpu/resctrl/monitor.c index 397206f23d14..773124b0e18a 100644 --- a/arch/x86/kernel/cpu/resctrl/monitor.c +++ b/arch/x86/kernel/cpu/resctrl/monitor.c @@ -514,7 +514,7 @@ void mbm_handle_overflow(struct work_struct *work) mutex_lock(&rdtgroup_mutex); - if (!static_branch_likely(&rdt_enable_key)) + if (!static_branch_likely(&rdt_mon_enable_key)) goto out_unlock; d = get_domain_from_cpu(cpu, &rdt_resources_all[RDT_RESOURCE_L3]); @@ -543,7 +543,7 @@ void mbm_setup_overflow_handler(struct rdt_domain *dom, unsigned long delay_ms) unsigned long delay = msecs_to_jiffies(delay_ms); int cpu; - if (!static_branch_likely(&rdt_enable_key)) + if (!static_branch_likely(&rdt_mon_enable_key)) return; cpu = cpumask_any(&dom->cpu_mask); dom->mbm_work_cpu = cpu; diff --git a/arch/x86/kernel/cpu/resctrl/rdtgroup.c b/arch/x86/kernel/cpu/resctrl/rdtgroup.c index dac7209a0708..1504bcabc63c 100644 --- a/arch/x86/kernel/cpu/resctrl/rdtgroup.c +++ b/arch/x86/kernel/cpu/resctrl/rdtgroup.c @@ -532,11 +532,15 @@ static void move_myself(struct callback_head *head) kfree(rdtgrp); } + if (unlikely(current->flags & PF_EXITING)) + goto out; + preempt_disable(); /* update PQR_ASSOC MSR to make resource group go into effect */ resctrl_sched_in(); preempt_enable(); +out: kfree(callback); } @@ -725,6 +729,92 @@ static int rdtgroup_tasks_show(struct kernfs_open_file *of, return ret; } +#ifdef CONFIG_PROC_CPU_RESCTRL + +/* + * A task can only be part of one resctrl control group and of one monitor + * group which is associated to that control group. + * + * 1) res: + * mon: + * + * resctrl is not available. + * + * 2) res:/ + * mon: + * + * Task is part of the root resctrl control group, and it is not associated + * to any monitor group. + * + * 3) res:/ + * mon:mon0 + * + * Task is part of the root resctrl control group and monitor group mon0. + * + * 4) res:group0 + * mon: + * + * Task is part of resctrl control group group0, and it is not associated + * to any monitor group. + * + * 5) res:group0 + * mon:mon1 + * + * Task is part of resctrl control group group0 and monitor group mon1. + */ +int proc_resctrl_show(struct seq_file *s, struct pid_namespace *ns, + struct pid *pid, struct task_struct *tsk) +{ + struct rdtgroup *rdtg; + int ret = 0; + + mutex_lock(&rdtgroup_mutex); + + /* Return empty if resctrl has not been mounted. */ + if (!static_branch_unlikely(&rdt_enable_key)) { + seq_puts(s, "res:\nmon:\n"); + goto unlock; + } + + list_for_each_entry(rdtg, &rdt_all_groups, rdtgroup_list) { + struct rdtgroup *crg; + + /* + * Task information is only relevant for shareable + * and exclusive groups. + */ + if (rdtg->mode != RDT_MODE_SHAREABLE && + rdtg->mode != RDT_MODE_EXCLUSIVE) + continue; + + if (rdtg->closid != tsk->closid) + continue; + + seq_printf(s, "res:%s%s\n", (rdtg == &rdtgroup_default) ? "/" : "", + rdtg->kn->name); + seq_puts(s, "mon:"); + list_for_each_entry(crg, &rdtg->mon.crdtgrp_list, + mon.crdtgrp_list) { + if (tsk->rmid != crg->mon.rmid) + continue; + seq_printf(s, "%s", crg->kn->name); + break; + } + seq_putc(s, '\n'); + goto unlock; + } + /* + * The above search should succeed. Otherwise return + * with an error. + */ + ret = -ENOENT; +unlock: + mutex_unlock(&rdtgroup_mutex); + + return ret; +} +#endif + static int rdt_last_cmd_status_show(struct kernfs_open_file *of, struct seq_file *seq, void *v) { diff --git a/arch/x86/kernel/cpu/scattered.c b/arch/x86/kernel/cpu/scattered.c index adf9b71386ef..62b137c3c97a 100644 --- a/arch/x86/kernel/cpu/scattered.c +++ b/arch/x86/kernel/cpu/scattered.c @@ -4,7 +4,7 @@ */ #include <linux/cpu.h> -#include <asm/pat.h> +#include <asm/memtype.h> #include <asm/apic.h> #include <asm/processor.h> diff --git a/arch/x86/kernel/cpu/topology.c b/arch/x86/kernel/cpu/topology.c index ee48c3fc8a65..d3a0791bc052 100644 --- a/arch/x86/kernel/cpu/topology.c +++ b/arch/x86/kernel/cpu/topology.c @@ -7,7 +7,7 @@ #include <linux/cpu.h> #include <asm/apic.h> -#include <asm/pat.h> +#include <asm/memtype.h> #include <asm/processor.h> #include "cpu.h" diff --git a/arch/x86/kernel/cpu/tsx.c b/arch/x86/kernel/cpu/tsx.c index 3e20d322bc98..e2ad30e474f8 100644 --- a/arch/x86/kernel/cpu/tsx.c +++ b/arch/x86/kernel/cpu/tsx.c @@ -14,6 +14,9 @@ #include "cpu.h" +#undef pr_fmt +#define pr_fmt(fmt) "tsx: " fmt + enum tsx_ctrl_states tsx_ctrl_state __ro_after_init = TSX_CTRL_NOT_SUPPORTED; void tsx_disable(void) @@ -99,7 +102,7 @@ void __init tsx_init(void) tsx_ctrl_state = x86_get_tsx_auto_mode(); } else { tsx_ctrl_state = TSX_CTRL_DISABLE; - pr_err("tsx: invalid option, defaulting to off\n"); + pr_err("invalid option, defaulting to off\n"); } } else { /* tsx= not provided */ @@ -115,11 +118,12 @@ void __init tsx_init(void) tsx_disable(); /* - * tsx_disable() will change the state of the - * RTM CPUID bit. Clear it here since it is now - * expected to be not set. + * tsx_disable() will change the state of the RTM and HLE CPUID + * bits. Clear them here since they are now expected to be not + * set. */ setup_clear_cpu_cap(X86_FEATURE_RTM); + setup_clear_cpu_cap(X86_FEATURE_HLE); } else if (tsx_ctrl_state == TSX_CTRL_ENABLE) { /* @@ -131,10 +135,10 @@ void __init tsx_init(void) tsx_enable(); /* - * tsx_enable() will change the state of the - * RTM CPUID bit. Force it here since it is now - * expected to be set. + * tsx_enable() will change the state of the RTM and HLE CPUID + * bits. Force them here since they are now expected to be set. */ setup_force_cpu_cap(X86_FEATURE_RTM); + setup_force_cpu_cap(X86_FEATURE_HLE); } } diff --git a/arch/x86/kernel/cpu/zhaoxin.c b/arch/x86/kernel/cpu/zhaoxin.c index 8e6f2f4b4afe..df1358ba622b 100644 --- a/arch/x86/kernel/cpu/zhaoxin.c +++ b/arch/x86/kernel/cpu/zhaoxin.c @@ -16,13 +16,6 @@ #define RNG_ENABLED (1 << 3) #define RNG_ENABLE (1 << 8) /* MSR_ZHAOXIN_RNG */ -#define X86_VMX_FEATURE_PROC_CTLS_TPR_SHADOW 0x00200000 -#define X86_VMX_FEATURE_PROC_CTLS_VNMI 0x00400000 -#define X86_VMX_FEATURE_PROC_CTLS_2ND_CTLS 0x80000000 -#define X86_VMX_FEATURE_PROC_CTLS2_VIRT_APIC 0x00000001 -#define X86_VMX_FEATURE_PROC_CTLS2_EPT 0x00000002 -#define X86_VMX_FEATURE_PROC_CTLS2_VPID 0x00000020 - static void init_zhaoxin_cap(struct cpuinfo_x86 *c) { u32 lo, hi; @@ -58,8 +51,6 @@ static void init_zhaoxin_cap(struct cpuinfo_x86 *c) if (c->x86 >= 0x6) set_cpu_cap(c, X86_FEATURE_REP_GOOD); - - cpu_detect_cache_sizes(c); } static void early_init_zhaoxin(struct cpuinfo_x86 *c) @@ -89,31 +80,6 @@ static void early_init_zhaoxin(struct cpuinfo_x86 *c) } -static void zhaoxin_detect_vmx_virtcap(struct cpuinfo_x86 *c) -{ - u32 vmx_msr_low, vmx_msr_high, msr_ctl, msr_ctl2; - - rdmsr(MSR_IA32_VMX_PROCBASED_CTLS, vmx_msr_low, vmx_msr_high); - msr_ctl = vmx_msr_high | vmx_msr_low; - - if (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_TPR_SHADOW) - set_cpu_cap(c, X86_FEATURE_TPR_SHADOW); - if (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_VNMI) - set_cpu_cap(c, X86_FEATURE_VNMI); - if (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_2ND_CTLS) { - rdmsr(MSR_IA32_VMX_PROCBASED_CTLS2, - vmx_msr_low, vmx_msr_high); - msr_ctl2 = vmx_msr_high | vmx_msr_low; - if ((msr_ctl2 & X86_VMX_FEATURE_PROC_CTLS2_VIRT_APIC) && - (msr_ctl & X86_VMX_FEATURE_PROC_CTLS_TPR_SHADOW)) - set_cpu_cap(c, X86_FEATURE_FLEXPRIORITY); - if (msr_ctl2 & X86_VMX_FEATURE_PROC_CTLS2_EPT) - set_cpu_cap(c, X86_FEATURE_EPT); - if (msr_ctl2 & X86_VMX_FEATURE_PROC_CTLS2_VPID) - set_cpu_cap(c, X86_FEATURE_VPID); - } -} - static void init_zhaoxin(struct cpuinfo_x86 *c) { early_init_zhaoxin(c); @@ -141,8 +107,7 @@ static void init_zhaoxin(struct cpuinfo_x86 *c) set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); #endif - if (cpu_has(c, X86_FEATURE_VMX)) - zhaoxin_detect_vmx_virtcap(c); + init_ia32_feat_ctl(c); } #ifdef CONFIG_X86_32 diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c index 00fc55ac7ffa..fd87b59452a3 100644 --- a/arch/x86/kernel/crash.c +++ b/arch/x86/kernel/crash.c @@ -370,7 +370,7 @@ int crash_setup_memmap_entries(struct kimage *image, struct boot_params *params) /* Add crashk_low_res region */ if (crashk_low_res.end) { ei.addr = crashk_low_res.start; - ei.size = crashk_low_res.end - crashk_low_res.start + 1; + ei.size = resource_size(&crashk_low_res); ei.type = E820_TYPE_RAM; add_e820_entry(params, &ei); } diff --git a/arch/x86/kernel/dumpstack.c b/arch/x86/kernel/dumpstack.c index e07424e19274..ae64ec7f752f 100644 --- a/arch/x86/kernel/dumpstack.c +++ b/arch/x86/kernel/dumpstack.c @@ -365,7 +365,7 @@ void oops_end(unsigned long flags, struct pt_regs *regs, int signr) } NOKPROBE_SYMBOL(oops_end); -int __die(const char *str, struct pt_regs *regs, long err) +static void __die_header(const char *str, struct pt_regs *regs, long err) { const char *pr = ""; @@ -384,7 +384,11 @@ int __die(const char *str, struct pt_regs *regs, long err) IS_ENABLED(CONFIG_KASAN) ? " KASAN" : "", IS_ENABLED(CONFIG_PAGE_TABLE_ISOLATION) ? (boot_cpu_has(X86_FEATURE_PTI) ? " PTI" : " NOPTI") : ""); +} +NOKPROBE_SYMBOL(__die_header); +static int __die_body(const char *str, struct pt_regs *regs, long err) +{ show_regs(regs); print_modules(); @@ -394,6 +398,13 @@ int __die(const char *str, struct pt_regs *regs, long err) return 0; } +NOKPROBE_SYMBOL(__die_body); + +int __die(const char *str, struct pt_regs *regs, long err) +{ + __die_header(str, regs, err); + return __die_body(str, regs, err); +} NOKPROBE_SYMBOL(__die); /* @@ -410,6 +421,19 @@ void die(const char *str, struct pt_regs *regs, long err) oops_end(flags, regs, sig); } +void die_addr(const char *str, struct pt_regs *regs, long err, long gp_addr) +{ + unsigned long flags = oops_begin(); + int sig = SIGSEGV; + + __die_header(str, regs, err); + if (gp_addr) + kasan_non_canonical_hook(gp_addr); + if (__die_body(str, regs, err)) + sig = 0; + oops_end(flags, regs, sig); +} + void show_regs(struct pt_regs *regs) { show_regs_print_info(KERN_DEFAULT); diff --git a/arch/x86/kernel/fpu/signal.c b/arch/x86/kernel/fpu/signal.c index 0071b794ed19..400a05e1c1c5 100644 --- a/arch/x86/kernel/fpu/signal.c +++ b/arch/x86/kernel/fpu/signal.c @@ -352,6 +352,7 @@ static int __fpu__restore_sig(void __user *buf, void __user *buf_fx, int size) fpregs_unlock(); return 0; } + fpregs_deactivate(fpu); fpregs_unlock(); } @@ -403,6 +404,8 @@ static int __fpu__restore_sig(void __user *buf, void __user *buf_fx, int size) } if (!ret) fpregs_mark_activate(); + else + fpregs_deactivate(fpu); fpregs_unlock(); err_out: diff --git a/arch/x86/kernel/fpu/xstate.c b/arch/x86/kernel/fpu/xstate.c index fa31470bbf24..a1806598aaa4 100644 --- a/arch/x86/kernel/fpu/xstate.c +++ b/arch/x86/kernel/fpu/xstate.c @@ -107,23 +107,20 @@ int cpu_has_xfeatures(u64 xfeatures_needed, const char **feature_name) } EXPORT_SYMBOL_GPL(cpu_has_xfeatures); -static int xfeature_is_supervisor(int xfeature_nr) +static bool xfeature_is_supervisor(int xfeature_nr) { /* - * We currently do not support supervisor states, but if - * we did, we could find out like this. - * - * SDM says: If state component 'i' is a user state component, - * ECX[0] return 0; if state component i is a supervisor - * state component, ECX[0] returns 1. + * Extended State Enumeration Sub-leaves (EAX = 0DH, ECX = n, n > 1) + * returns ECX[0] set to (1) for a supervisor state, and cleared (0) + * for a user state. */ u32 eax, ebx, ecx, edx; cpuid_count(XSTATE_CPUID, xfeature_nr, &eax, &ebx, &ecx, &edx); - return !!(ecx & 1); + return ecx & 1; } -static int xfeature_is_user(int xfeature_nr) +static bool xfeature_is_user(int xfeature_nr) { return !xfeature_is_supervisor(xfeature_nr); } @@ -419,7 +416,8 @@ static void __init setup_init_fpu_buf(void) print_xstate_features(); if (boot_cpu_has(X86_FEATURE_XSAVES)) - init_fpstate.xsave.header.xcomp_bv = (u64)1 << 63 | xfeatures_mask; + init_fpstate.xsave.header.xcomp_bv = XCOMP_BV_COMPACTED_FORMAT | + xfeatures_mask; /* * Init all the features state with header.xfeatures being 0x0 diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 024c3053dbba..37a0aeaf89e7 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -23,6 +23,7 @@ #include <linux/list.h> #include <linux/module.h> #include <linux/memory.h> +#include <linux/vmalloc.h> #include <trace/syscall.h> @@ -34,6 +35,8 @@ #ifdef CONFIG_DYNAMIC_FTRACE +static int ftrace_poke_late = 0; + int ftrace_arch_code_modify_prepare(void) __acquires(&text_mutex) { @@ -43,84 +46,37 @@ int ftrace_arch_code_modify_prepare(void) * ftrace has it set to "read/write". */ mutex_lock(&text_mutex); - set_kernel_text_rw(); - set_all_modules_text_rw(); + ftrace_poke_late = 1; return 0; } int ftrace_arch_code_modify_post_process(void) __releases(&text_mutex) { - set_all_modules_text_ro(); - set_kernel_text_ro(); + /* + * ftrace_make_{call,nop}() may be called during + * module load, and we need to finish the text_poke_queue() + * that they do, here. + */ + text_poke_finish(); + ftrace_poke_late = 0; mutex_unlock(&text_mutex); return 0; } -union ftrace_code_union { - char code[MCOUNT_INSN_SIZE]; - struct { - unsigned char op; - int offset; - } __attribute__((packed)); -}; - -static int ftrace_calc_offset(long ip, long addr) -{ - return (int)(addr - ip); -} - -static unsigned char * -ftrace_text_replace(unsigned char op, unsigned long ip, unsigned long addr) +static const char *ftrace_nop_replace(void) { - static union ftrace_code_union calc; - - calc.op = op; - calc.offset = ftrace_calc_offset(ip + MCOUNT_INSN_SIZE, addr); - - return calc.code; -} - -static unsigned char * -ftrace_call_replace(unsigned long ip, unsigned long addr) -{ - return ftrace_text_replace(0xe8, ip, addr); -} - -static inline int -within(unsigned long addr, unsigned long start, unsigned long end) -{ - return addr >= start && addr < end; + return ideal_nops[NOP_ATOMIC5]; } -static unsigned long text_ip_addr(unsigned long ip) +static const char *ftrace_call_replace(unsigned long ip, unsigned long addr) { - /* - * On x86_64, kernel text mappings are mapped read-only, so we use - * the kernel identity mapping instead of the kernel text mapping - * to modify the kernel text. - * - * For 32bit kernels, these mappings are same and we can use - * kernel identity mapping to modify code. - */ - if (within(ip, (unsigned long)_text, (unsigned long)_etext)) - ip = (unsigned long)__va(__pa_symbol(ip)); - - return ip; + return text_gen_insn(CALL_INSN_OPCODE, (void *)ip, (void *)addr); } -static const unsigned char *ftrace_nop_replace(void) +static int ftrace_verify_code(unsigned long ip, const char *old_code) { - return ideal_nops[NOP_ATOMIC5]; -} - -static int -ftrace_modify_code_direct(unsigned long ip, unsigned const char *old_code, - unsigned const char *new_code) -{ - unsigned char replaced[MCOUNT_INSN_SIZE]; - - ftrace_expected = old_code; + char cur_code[MCOUNT_INSN_SIZE]; /* * Note: @@ -129,31 +85,46 @@ ftrace_modify_code_direct(unsigned long ip, unsigned const char *old_code, * Carefully read and modify the code with probe_kernel_*(), and make * sure what we read is what we expected it to be before modifying it. */ - /* read the text we want to modify */ - if (probe_kernel_read(replaced, (void *)ip, MCOUNT_INSN_SIZE)) + if (probe_kernel_read(cur_code, (void *)ip, MCOUNT_INSN_SIZE)) { + WARN_ON(1); return -EFAULT; + } /* Make sure it is what we expect it to be */ - if (memcmp(replaced, old_code, MCOUNT_INSN_SIZE) != 0) + if (memcmp(cur_code, old_code, MCOUNT_INSN_SIZE) != 0) { + WARN_ON(1); return -EINVAL; + } - ip = text_ip_addr(ip); - - /* replace the text with the new text */ - if (probe_kernel_write((void *)ip, new_code, MCOUNT_INSN_SIZE)) - return -EPERM; + return 0; +} - sync_core(); +/* + * Marked __ref because it calls text_poke_early() which is .init.text. That is + * ok because that call will happen early, during boot, when .init sections are + * still present. + */ +static int __ref +ftrace_modify_code_direct(unsigned long ip, const char *old_code, + const char *new_code) +{ + int ret = ftrace_verify_code(ip, old_code); + if (ret) + return ret; + /* replace the text with the new text */ + if (ftrace_poke_late) + text_poke_queue((void *)ip, new_code, MCOUNT_INSN_SIZE, NULL); + else + text_poke_early((void *)ip, new_code, MCOUNT_INSN_SIZE); return 0; } -int ftrace_make_nop(struct module *mod, - struct dyn_ftrace *rec, unsigned long addr) +int ftrace_make_nop(struct module *mod, struct dyn_ftrace *rec, unsigned long addr) { - unsigned const char *new, *old; unsigned long ip = rec->ip; + const char *new, *old; old = ftrace_call_replace(ip, addr); new = ftrace_nop_replace(); @@ -167,19 +138,20 @@ int ftrace_make_nop(struct module *mod, * just modify the code directly. */ if (addr == MCOUNT_ADDR) - return ftrace_modify_code_direct(rec->ip, old, new); + return ftrace_modify_code_direct(ip, old, new); - ftrace_expected = NULL; - - /* Normal cases use add_brk_on_nop */ + /* + * x86 overrides ftrace_replace_code -- this function will never be used + * in this case. + */ WARN_ONCE(1, "invalid use of ftrace_make_nop"); return -EINVAL; } int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr) { - unsigned const char *new, *old; unsigned long ip = rec->ip; + const char *new, *old; old = ftrace_nop_replace(); new = ftrace_call_replace(ip, addr); @@ -189,43 +161,6 @@ int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr) } /* - * The modifying_ftrace_code is used to tell the breakpoint - * handler to call ftrace_int3_handler(). If it fails to - * call this handler for a breakpoint added by ftrace, then - * the kernel may crash. - * - * As atomic_writes on x86 do not need a barrier, we do not - * need to add smp_mb()s for this to work. It is also considered - * that we can not read the modifying_ftrace_code before - * executing the breakpoint. That would be quite remarkable if - * it could do that. Here's the flow that is required: - * - * CPU-0 CPU-1 - * - * atomic_inc(mfc); - * write int3s - * <trap-int3> // implicit (r)mb - * if (atomic_read(mfc)) - * call ftrace_int3_handler() - * - * Then when we are finished: - * - * atomic_dec(mfc); - * - * If we hit a breakpoint that was not set by ftrace, it does not - * matter if ftrace_int3_handler() is called or not. It will - * simply be ignored. But it is crucial that a ftrace nop/caller - * breakpoint is handled. No other user should ever place a - * breakpoint on an ftrace nop/caller location. It must only - * be done by this code. - */ -atomic_t modifying_ftrace_code __read_mostly; - -static int -ftrace_modify_code(unsigned long ip, unsigned const char *old_code, - unsigned const char *new_code); - -/* * Should never be called: * As it is only called by __ftrace_replace_code() which is called by * ftrace_replace_code() that x86 overrides, and by ftrace_update_code() @@ -237,452 +172,84 @@ int ftrace_modify_call(struct dyn_ftrace *rec, unsigned long old_addr, unsigned long addr) { WARN_ON(1); - ftrace_expected = NULL; return -EINVAL; } -static unsigned long ftrace_update_func; -static unsigned long ftrace_update_func_call; - -static int update_ftrace_func(unsigned long ip, void *new) -{ - unsigned char old[MCOUNT_INSN_SIZE]; - int ret; - - memcpy(old, (void *)ip, MCOUNT_INSN_SIZE); - - ftrace_update_func = ip; - /* Make sure the breakpoints see the ftrace_update_func update */ - smp_wmb(); - - /* See comment above by declaration of modifying_ftrace_code */ - atomic_inc(&modifying_ftrace_code); - - ret = ftrace_modify_code(ip, old, new); - - atomic_dec(&modifying_ftrace_code); - - return ret; -} - int ftrace_update_ftrace_func(ftrace_func_t func) { - unsigned long ip = (unsigned long)(&ftrace_call); - unsigned char *new; - int ret; - - ftrace_update_func_call = (unsigned long)func; - - new = ftrace_call_replace(ip, (unsigned long)func); - ret = update_ftrace_func(ip, new); - - /* Also update the regs callback function */ - if (!ret) { - ip = (unsigned long)(&ftrace_regs_call); - new = ftrace_call_replace(ip, (unsigned long)func); - ret = update_ftrace_func(ip, new); - } - - return ret; -} - -static nokprobe_inline int is_ftrace_caller(unsigned long ip) -{ - if (ip == ftrace_update_func) - return 1; - - return 0; -} - -/* - * A breakpoint was added to the code address we are about to - * modify, and this is the handle that will just skip over it. - * We are either changing a nop into a trace call, or a trace - * call to a nop. While the change is taking place, we treat - * it just like it was a nop. - */ -int ftrace_int3_handler(struct pt_regs *regs) -{ unsigned long ip; + const char *new; - if (WARN_ON_ONCE(!regs)) - return 0; - - ip = regs->ip - INT3_INSN_SIZE; - - if (ftrace_location(ip)) { - int3_emulate_call(regs, (unsigned long)ftrace_regs_caller); - return 1; - } else if (is_ftrace_caller(ip)) { - if (!ftrace_update_func_call) { - int3_emulate_jmp(regs, ip + CALL_INSN_SIZE); - return 1; - } - int3_emulate_call(regs, ftrace_update_func_call); - return 1; - } - - return 0; -} -NOKPROBE_SYMBOL(ftrace_int3_handler); - -static int ftrace_write(unsigned long ip, const char *val, int size) -{ - ip = text_ip_addr(ip); - - if (probe_kernel_write((void *)ip, val, size)) - return -EPERM; - - return 0; -} - -static int add_break(unsigned long ip, const char *old) -{ - unsigned char replaced[MCOUNT_INSN_SIZE]; - unsigned char brk = BREAKPOINT_INSTRUCTION; - - if (probe_kernel_read(replaced, (void *)ip, MCOUNT_INSN_SIZE)) - return -EFAULT; - - ftrace_expected = old; - - /* Make sure it is what we expect it to be */ - if (memcmp(replaced, old, MCOUNT_INSN_SIZE) != 0) - return -EINVAL; - - return ftrace_write(ip, &brk, 1); -} - -static int add_brk_on_call(struct dyn_ftrace *rec, unsigned long addr) -{ - unsigned const char *old; - unsigned long ip = rec->ip; - - old = ftrace_call_replace(ip, addr); - - return add_break(rec->ip, old); -} - - -static int add_brk_on_nop(struct dyn_ftrace *rec) -{ - unsigned const char *old; - - old = ftrace_nop_replace(); - - return add_break(rec->ip, old); -} - -static int add_breakpoints(struct dyn_ftrace *rec, bool enable) -{ - unsigned long ftrace_addr; - int ret; - - ftrace_addr = ftrace_get_addr_curr(rec); - - ret = ftrace_test_record(rec, enable); - - switch (ret) { - case FTRACE_UPDATE_IGNORE: - return 0; - - case FTRACE_UPDATE_MAKE_CALL: - /* converting nop to call */ - return add_brk_on_nop(rec); - - case FTRACE_UPDATE_MODIFY_CALL: - case FTRACE_UPDATE_MAKE_NOP: - /* converting a call to a nop */ - return add_brk_on_call(rec, ftrace_addr); - } - return 0; -} - -/* - * On error, we need to remove breakpoints. This needs to - * be done caefully. If the address does not currently have a - * breakpoint, we know we are done. Otherwise, we look at the - * remaining 4 bytes of the instruction. If it matches a nop - * we replace the breakpoint with the nop. Otherwise we replace - * it with the call instruction. - */ -static int remove_breakpoint(struct dyn_ftrace *rec) -{ - unsigned char ins[MCOUNT_INSN_SIZE]; - unsigned char brk = BREAKPOINT_INSTRUCTION; - const unsigned char *nop; - unsigned long ftrace_addr; - unsigned long ip = rec->ip; - - /* If we fail the read, just give up */ - if (probe_kernel_read(ins, (void *)ip, MCOUNT_INSN_SIZE)) - return -EFAULT; - - /* If this does not have a breakpoint, we are done */ - if (ins[0] != brk) - return 0; - - nop = ftrace_nop_replace(); - - /* - * If the last 4 bytes of the instruction do not match - * a nop, then we assume that this is a call to ftrace_addr. - */ - if (memcmp(&ins[1], &nop[1], MCOUNT_INSN_SIZE - 1) != 0) { - /* - * For extra paranoidism, we check if the breakpoint is on - * a call that would actually jump to the ftrace_addr. - * If not, don't touch the breakpoint, we make just create - * a disaster. - */ - ftrace_addr = ftrace_get_addr_new(rec); - nop = ftrace_call_replace(ip, ftrace_addr); - - if (memcmp(&ins[1], &nop[1], MCOUNT_INSN_SIZE - 1) == 0) - goto update; - - /* Check both ftrace_addr and ftrace_old_addr */ - ftrace_addr = ftrace_get_addr_curr(rec); - nop = ftrace_call_replace(ip, ftrace_addr); - - ftrace_expected = nop; - - if (memcmp(&ins[1], &nop[1], MCOUNT_INSN_SIZE - 1) != 0) - return -EINVAL; - } - - update: - return ftrace_write(ip, nop, 1); -} - -static int add_update_code(unsigned long ip, unsigned const char *new) -{ - /* skip breakpoint */ - ip++; - new++; - return ftrace_write(ip, new, MCOUNT_INSN_SIZE - 1); -} - -static int add_update_call(struct dyn_ftrace *rec, unsigned long addr) -{ - unsigned long ip = rec->ip; - unsigned const char *new; - - new = ftrace_call_replace(ip, addr); - return add_update_code(ip, new); -} - -static int add_update_nop(struct dyn_ftrace *rec) -{ - unsigned long ip = rec->ip; - unsigned const char *new; - - new = ftrace_nop_replace(); - return add_update_code(ip, new); -} - -static int add_update(struct dyn_ftrace *rec, bool enable) -{ - unsigned long ftrace_addr; - int ret; - - ret = ftrace_test_record(rec, enable); - - ftrace_addr = ftrace_get_addr_new(rec); - - switch (ret) { - case FTRACE_UPDATE_IGNORE: - return 0; - - case FTRACE_UPDATE_MODIFY_CALL: - case FTRACE_UPDATE_MAKE_CALL: - /* converting nop to call */ - return add_update_call(rec, ftrace_addr); - - case FTRACE_UPDATE_MAKE_NOP: - /* converting a call to a nop */ - return add_update_nop(rec); - } - - return 0; -} - -static int finish_update_call(struct dyn_ftrace *rec, unsigned long addr) -{ - unsigned long ip = rec->ip; - unsigned const char *new; - - new = ftrace_call_replace(ip, addr); - - return ftrace_write(ip, new, 1); -} - -static int finish_update_nop(struct dyn_ftrace *rec) -{ - unsigned long ip = rec->ip; - unsigned const char *new; - - new = ftrace_nop_replace(); - - return ftrace_write(ip, new, 1); -} - -static int finish_update(struct dyn_ftrace *rec, bool enable) -{ - unsigned long ftrace_addr; - int ret; - - ret = ftrace_update_record(rec, enable); - - ftrace_addr = ftrace_get_addr_new(rec); - - switch (ret) { - case FTRACE_UPDATE_IGNORE: - return 0; - - case FTRACE_UPDATE_MODIFY_CALL: - case FTRACE_UPDATE_MAKE_CALL: - /* converting nop to call */ - return finish_update_call(rec, ftrace_addr); + ip = (unsigned long)(&ftrace_call); + new = ftrace_call_replace(ip, (unsigned long)func); + text_poke_bp((void *)ip, new, MCOUNT_INSN_SIZE, NULL); - case FTRACE_UPDATE_MAKE_NOP: - /* converting a call to a nop */ - return finish_update_nop(rec); - } + ip = (unsigned long)(&ftrace_regs_call); + new = ftrace_call_replace(ip, (unsigned long)func); + text_poke_bp((void *)ip, new, MCOUNT_INSN_SIZE, NULL); return 0; } -static void do_sync_core(void *data) -{ - sync_core(); -} - -static void run_sync(void) -{ - int enable_irqs; - - /* No need to sync if there's only one CPU */ - if (num_online_cpus() == 1) - return; - - enable_irqs = irqs_disabled(); - - /* We may be called with interrupts disabled (on bootup). */ - if (enable_irqs) - local_irq_enable(); - on_each_cpu(do_sync_core, NULL, 1); - if (enable_irqs) - local_irq_disable(); -} - void ftrace_replace_code(int enable) { struct ftrace_rec_iter *iter; struct dyn_ftrace *rec; - const char *report = "adding breakpoints"; - int count = 0; + const char *new, *old; int ret; for_ftrace_rec_iter(iter) { rec = ftrace_rec_iter_record(iter); - ret = add_breakpoints(rec, enable); - if (ret) - goto remove_breakpoints; - count++; - } - - run_sync(); + switch (ftrace_test_record(rec, enable)) { + case FTRACE_UPDATE_IGNORE: + default: + continue; - report = "updating code"; - count = 0; + case FTRACE_UPDATE_MAKE_CALL: + old = ftrace_nop_replace(); + break; - for_ftrace_rec_iter(iter) { - rec = ftrace_rec_iter_record(iter); + case FTRACE_UPDATE_MODIFY_CALL: + case FTRACE_UPDATE_MAKE_NOP: + old = ftrace_call_replace(rec->ip, ftrace_get_addr_curr(rec)); + break; + } - ret = add_update(rec, enable); - if (ret) - goto remove_breakpoints; - count++; + ret = ftrace_verify_code(rec->ip, old); + if (ret) { + ftrace_bug(ret, rec); + return; + } } - run_sync(); - - report = "removing breakpoints"; - count = 0; - for_ftrace_rec_iter(iter) { rec = ftrace_rec_iter_record(iter); - ret = finish_update(rec, enable); - if (ret) - goto remove_breakpoints; - count++; - } + switch (ftrace_test_record(rec, enable)) { + case FTRACE_UPDATE_IGNORE: + default: + continue; - run_sync(); + case FTRACE_UPDATE_MAKE_CALL: + case FTRACE_UPDATE_MODIFY_CALL: + new = ftrace_call_replace(rec->ip, ftrace_get_addr_new(rec)); + break; - return; + case FTRACE_UPDATE_MAKE_NOP: + new = ftrace_nop_replace(); + break; + } - remove_breakpoints: - pr_warn("Failed on %s (%d):\n", report, count); - ftrace_bug(ret, rec); - for_ftrace_rec_iter(iter) { - rec = ftrace_rec_iter_record(iter); - /* - * Breakpoints are handled only when this function is in - * progress. The system could not work with them. - */ - if (remove_breakpoint(rec)) - BUG(); + text_poke_queue((void *)rec->ip, new, MCOUNT_INSN_SIZE, NULL); + ftrace_update_record(rec, enable); } - run_sync(); -} - -static int -ftrace_modify_code(unsigned long ip, unsigned const char *old_code, - unsigned const char *new_code) -{ - int ret; - - ret = add_break(ip, old_code); - if (ret) - goto out; - - run_sync(); - - ret = add_update_code(ip, new_code); - if (ret) - goto fail_update; - - run_sync(); - - ret = ftrace_write(ip, new_code, 1); - /* - * The breakpoint is handled only when this function is in progress. - * The system could not work if we could not remove it. - */ - BUG_ON(ret); - out: - run_sync(); - return ret; - - fail_update: - /* Also here the system could not work with the breakpoint */ - if (ftrace_write(ip, old_code, 1)) - BUG(); - goto out; + text_poke_finish(); } void arch_ftrace_update_code(int command) { - /* See comment above by declaration of modifying_ftrace_code */ - atomic_inc(&modifying_ftrace_code); - ftrace_modify_all_code(command); - - atomic_dec(&modifying_ftrace_code); } int __init ftrace_dyn_arch_init(void) @@ -747,6 +314,7 @@ create_trampoline(struct ftrace_ops *ops, unsigned int *tramp_size) unsigned long start_offset; unsigned long end_offset; unsigned long op_offset; + unsigned long call_offset; unsigned long offset; unsigned long npages; unsigned long size; @@ -763,10 +331,12 @@ create_trampoline(struct ftrace_ops *ops, unsigned int *tramp_size) start_offset = (unsigned long)ftrace_regs_caller; end_offset = (unsigned long)ftrace_regs_caller_end; op_offset = (unsigned long)ftrace_regs_caller_op_ptr; + call_offset = (unsigned long)ftrace_regs_call; } else { start_offset = (unsigned long)ftrace_caller; end_offset = (unsigned long)ftrace_epilogue; op_offset = (unsigned long)ftrace_caller_op_ptr; + call_offset = (unsigned long)ftrace_call; } size = end_offset - start_offset; @@ -823,16 +393,21 @@ create_trampoline(struct ftrace_ops *ops, unsigned int *tramp_size) /* put in the new offset to the ftrace_ops */ memcpy(trampoline + op_offset, &op_ptr, OP_REF_SIZE); + /* put in the call to the function */ + mutex_lock(&text_mutex); + call_offset -= start_offset; + memcpy(trampoline + call_offset, + text_gen_insn(CALL_INSN_OPCODE, + trampoline + call_offset, + ftrace_ops_get_func(ops)), CALL_INSN_SIZE); + mutex_unlock(&text_mutex); + /* ALLOC_TRAMP flags lets us know we created it */ ops->flags |= FTRACE_OPS_FL_ALLOC_TRAMP; set_vm_flush_reset_perms(trampoline); - /* - * Module allocation needs to be completed by making the page - * executable. The page is still writable, which is a security hazard, - * but anyhow ftrace breaks W^X completely. - */ + set_memory_ro((unsigned long)trampoline, npages); set_memory_x((unsigned long)trampoline, npages); return (unsigned long)trampoline; fail: @@ -859,62 +434,54 @@ static unsigned long calc_trampoline_call_offset(bool save_regs) void arch_ftrace_update_trampoline(struct ftrace_ops *ops) { ftrace_func_t func; - unsigned char *new; unsigned long offset; unsigned long ip; unsigned int size; - int ret, npages; + const char *new; - if (ops->trampoline) { - /* - * The ftrace_ops caller may set up its own trampoline. - * In such a case, this code must not modify it. - */ - if (!(ops->flags & FTRACE_OPS_FL_ALLOC_TRAMP)) - return; - npages = PAGE_ALIGN(ops->trampoline_size) >> PAGE_SHIFT; - set_memory_rw(ops->trampoline, npages); - } else { + if (!ops->trampoline) { ops->trampoline = create_trampoline(ops, &size); if (!ops->trampoline) return; ops->trampoline_size = size; - npages = PAGE_ALIGN(size) >> PAGE_SHIFT; + return; } + /* + * The ftrace_ops caller may set up its own trampoline. + * In such a case, this code must not modify it. + */ + if (!(ops->flags & FTRACE_OPS_FL_ALLOC_TRAMP)) + return; + offset = calc_trampoline_call_offset(ops->flags & FTRACE_OPS_FL_SAVE_REGS); ip = ops->trampoline + offset; - func = ftrace_ops_get_func(ops); - ftrace_update_func_call = (unsigned long)func; - + mutex_lock(&text_mutex); /* Do a safe modify in case the trampoline is executing */ new = ftrace_call_replace(ip, (unsigned long)func); - ret = update_ftrace_func(ip, new); - set_memory_ro(ops->trampoline, npages); - - /* The update should never fail */ - WARN_ON(ret); + text_poke_bp((void *)ip, new, MCOUNT_INSN_SIZE, NULL); + mutex_unlock(&text_mutex); } /* Return the address of the function the trampoline calls */ static void *addr_from_call(void *ptr) { - union ftrace_code_union calc; + union text_poke_insn call; int ret; - ret = probe_kernel_read(&calc, ptr, MCOUNT_INSN_SIZE); + ret = probe_kernel_read(&call, ptr, CALL_INSN_SIZE); if (WARN_ON_ONCE(ret < 0)) return NULL; /* Make sure this is a call */ - if (WARN_ON_ONCE(calc.op != 0xe8)) { - pr_warn("Expected e8, got %x\n", calc.op); + if (WARN_ON_ONCE(call.opcode != CALL_INSN_OPCODE)) { + pr_warn("Expected E8, got %x\n", call.opcode); return NULL; } - return ptr + MCOUNT_INSN_SIZE + calc.offset; + return ptr + CALL_INSN_SIZE + call.disp; } void prepare_ftrace_return(unsigned long self_addr, unsigned long *parent, @@ -981,19 +548,18 @@ void arch_ftrace_trampoline_free(struct ftrace_ops *ops) #ifdef CONFIG_DYNAMIC_FTRACE extern void ftrace_graph_call(void); -static unsigned char *ftrace_jmp_replace(unsigned long ip, unsigned long addr) +static const char *ftrace_jmp_replace(unsigned long ip, unsigned long addr) { - return ftrace_text_replace(0xe9, ip, addr); + return text_gen_insn(JMP32_INSN_OPCODE, (void *)ip, (void *)addr); } static int ftrace_mod_jmp(unsigned long ip, void *func) { - unsigned char *new; + const char *new; - ftrace_update_func_call = 0UL; new = ftrace_jmp_replace(ip, (unsigned long)func); - - return update_ftrace_func(ip, new); + text_poke_bp((void *)ip, new, MCOUNT_INSN_SIZE, NULL); + return 0; } int ftrace_enable_ftrace_graph_caller(void) @@ -1019,10 +585,9 @@ int ftrace_disable_ftrace_graph_caller(void) void prepare_ftrace_return(unsigned long self_addr, unsigned long *parent, unsigned long frame_pointer) { + unsigned long return_hooker = (unsigned long)&return_to_handler; unsigned long old; int faulted; - unsigned long return_hooker = (unsigned long) - &return_to_handler; /* * When resuming from suspend-to-ram, this function can be indirectly diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c index c6f791bc481e..7a50f0b62a70 100644 --- a/arch/x86/kernel/hpet.c +++ b/arch/x86/kernel/hpet.c @@ -84,7 +84,7 @@ static inline void hpet_writel(unsigned int d, unsigned int a) static inline void hpet_set_mapping(void) { - hpet_virt_address = ioremap_nocache(hpet_address, HPET_MMAP_SIZE); + hpet_virt_address = ioremap(hpet_address, HPET_MMAP_SIZE); } static inline void hpet_clear_mapping(void) diff --git a/arch/x86/kernel/jump_label.c b/arch/x86/kernel/jump_label.c index c1a8b9e71408..9c4498ea0b3c 100644 --- a/arch/x86/kernel/jump_label.c +++ b/arch/x86/kernel/jump_label.c @@ -16,15 +16,7 @@ #include <asm/alternative.h> #include <asm/text-patching.h> -union jump_code_union { - char code[JUMP_LABEL_NOP_SIZE]; - struct { - char jump; - int offset; - } __attribute__((packed)); -}; - -static void bug_at(unsigned char *ip, int line) +static void bug_at(const void *ip, int line) { /* * The location is not an op that we were expecting. @@ -35,42 +27,42 @@ static void bug_at(unsigned char *ip, int line) BUG(); } -static void __jump_label_set_jump_code(struct jump_entry *entry, - enum jump_label_type type, - union jump_code_union *code, - int init) +static const void * +__jump_label_set_jump_code(struct jump_entry *entry, enum jump_label_type type, int init) { const unsigned char default_nop[] = { STATIC_KEY_INIT_NOP }; const unsigned char *ideal_nop = ideal_nops[NOP_ATOMIC5]; - const void *expect; + const void *expect, *code; + const void *addr, *dest; int line; - code->jump = 0xe9; - code->offset = jump_entry_target(entry) - - (jump_entry_code(entry) + JUMP_LABEL_NOP_SIZE); + addr = (void *)jump_entry_code(entry); + dest = (void *)jump_entry_target(entry); + + code = text_gen_insn(JMP32_INSN_OPCODE, addr, dest); if (init) { expect = default_nop; line = __LINE__; } else if (type == JUMP_LABEL_JMP) { expect = ideal_nop; line = __LINE__; } else { - expect = code->code; line = __LINE__; + expect = code; line = __LINE__; } - if (memcmp((void *)jump_entry_code(entry), expect, JUMP_LABEL_NOP_SIZE)) - bug_at((void *)jump_entry_code(entry), line); + if (memcmp(addr, expect, JUMP_LABEL_NOP_SIZE)) + bug_at(addr, line); if (type == JUMP_LABEL_NOP) - memcpy(code, ideal_nop, JUMP_LABEL_NOP_SIZE); + code = ideal_nop; + + return code; } -static void __ref __jump_label_transform(struct jump_entry *entry, - enum jump_label_type type, - int init) +static void inline __jump_label_transform(struct jump_entry *entry, + enum jump_label_type type, + int init) { - union jump_code_union code; - - __jump_label_set_jump_code(entry, type, &code, init); + const void *opcode = __jump_label_set_jump_code(entry, type, init); /* * As long as only a single processor is running and the code is still @@ -84,31 +76,33 @@ static void __ref __jump_label_transform(struct jump_entry *entry, * always nop being the 'currently valid' instruction */ if (init || system_state == SYSTEM_BOOTING) { - text_poke_early((void *)jump_entry_code(entry), &code, + text_poke_early((void *)jump_entry_code(entry), opcode, JUMP_LABEL_NOP_SIZE); return; } - text_poke_bp((void *)jump_entry_code(entry), &code, JUMP_LABEL_NOP_SIZE, NULL); + text_poke_bp((void *)jump_entry_code(entry), opcode, JUMP_LABEL_NOP_SIZE, NULL); } -void arch_jump_label_transform(struct jump_entry *entry, - enum jump_label_type type) +static void __ref jump_label_transform(struct jump_entry *entry, + enum jump_label_type type, + int init) { mutex_lock(&text_mutex); - __jump_label_transform(entry, type, 0); + __jump_label_transform(entry, type, init); mutex_unlock(&text_mutex); } -#define TP_VEC_MAX (PAGE_SIZE / sizeof(struct text_poke_loc)) -static struct text_poke_loc tp_vec[TP_VEC_MAX]; -static int tp_vec_nr; +void arch_jump_label_transform(struct jump_entry *entry, + enum jump_label_type type) +{ + jump_label_transform(entry, type, 0); +} bool arch_jump_label_transform_queue(struct jump_entry *entry, enum jump_label_type type) { - struct text_poke_loc *tp; - void *entry_code; + const void *opcode; if (system_state == SYSTEM_BOOTING) { /* @@ -118,53 +112,19 @@ bool arch_jump_label_transform_queue(struct jump_entry *entry, return true; } - /* - * No more space in the vector, tell upper layer to apply - * the queue before continuing. - */ - if (tp_vec_nr == TP_VEC_MAX) - return false; - - tp = &tp_vec[tp_vec_nr]; - - entry_code = (void *)jump_entry_code(entry); - - /* - * The INT3 handler will do a bsearch in the queue, so we need entries - * to be sorted. We can survive an unsorted list by rejecting the entry, - * forcing the generic jump_label code to apply the queue. Warning once, - * to raise the attention to the case of an unsorted entry that is - * better not happen, because, in the worst case we will perform in the - * same way as we do without batching - with some more overhead. - */ - if (tp_vec_nr > 0) { - int prev = tp_vec_nr - 1; - struct text_poke_loc *prev_tp = &tp_vec[prev]; - - if (WARN_ON_ONCE(prev_tp->addr > entry_code)) - return false; - } - - __jump_label_set_jump_code(entry, type, - (union jump_code_union *)&tp->text, 0); - - text_poke_loc_init(tp, entry_code, NULL, JUMP_LABEL_NOP_SIZE, NULL); - - tp_vec_nr++; - + mutex_lock(&text_mutex); + opcode = __jump_label_set_jump_code(entry, type, 0); + text_poke_queue((void *)jump_entry_code(entry), + opcode, JUMP_LABEL_NOP_SIZE, NULL); + mutex_unlock(&text_mutex); return true; } void arch_jump_label_transform_apply(void) { - if (!tp_vec_nr) - return; - mutex_lock(&text_mutex); - text_poke_bp_batch(tp_vec, tp_vec_nr); + text_poke_finish(); mutex_unlock(&text_mutex); - - tp_vec_nr = 0; } static enum { @@ -193,5 +153,5 @@ __init_or_module void arch_jump_label_transform_static(struct jump_entry *entry, jlstate = JL_STATE_NO_UPDATE; } if (jlstate == JL_STATE_UPDATE) - __jump_label_transform(entry, type, 1); + jump_label_transform(entry, type, 1); } diff --git a/arch/x86/kernel/kexec-bzimage64.c b/arch/x86/kernel/kexec-bzimage64.c index d2f4e706a428..f293d872602a 100644 --- a/arch/x86/kernel/kexec-bzimage64.c +++ b/arch/x86/kernel/kexec-bzimage64.c @@ -177,7 +177,7 @@ setup_efi_state(struct boot_params *params, unsigned long params_load_addr, * acpi_rsdp=<addr> on kernel command line to make second kernel boot * without efi. */ - if (efi_enabled(EFI_OLD_MEMMAP)) + if (efi_have_uv1_memmap()) return 0; params->secure_boot = boot_params.secure_boot; diff --git a/arch/x86/kernel/kprobes/core.c b/arch/x86/kernel/kprobes/core.c index 4f13af7cbcdb..4d7022a740ab 100644 --- a/arch/x86/kernel/kprobes/core.c +++ b/arch/x86/kernel/kprobes/core.c @@ -40,6 +40,7 @@ #include <linux/frame.h> #include <linux/kasan.h> #include <linux/moduleloader.h> +#include <linux/vmalloc.h> #include <asm/text-patching.h> #include <asm/cacheflush.h> @@ -119,14 +120,14 @@ __synthesize_relative_insn(void *dest, void *from, void *to, u8 op) /* Insert a jump instruction at address 'from', which jumps to address 'to'.*/ void synthesize_reljump(void *dest, void *from, void *to) { - __synthesize_relative_insn(dest, from, to, RELATIVEJUMP_OPCODE); + __synthesize_relative_insn(dest, from, to, JMP32_INSN_OPCODE); } NOKPROBE_SYMBOL(synthesize_reljump); /* Insert a call instruction at address 'from', which calls address 'to'.*/ void synthesize_relcall(void *dest, void *from, void *to) { - __synthesize_relative_insn(dest, from, to, RELATIVECALL_OPCODE); + __synthesize_relative_insn(dest, from, to, CALL_INSN_OPCODE); } NOKPROBE_SYMBOL(synthesize_relcall); @@ -301,7 +302,7 @@ static int can_probe(unsigned long paddr) * Another debugging subsystem might insert this breakpoint. * In that case, we can't recover it. */ - if (insn.opcode.bytes[0] == BREAKPOINT_INSTRUCTION) + if (insn.opcode.bytes[0] == INT3_INSN_OPCODE) return 0; addr += insn.length; } @@ -356,7 +357,7 @@ int __copy_instruction(u8 *dest, u8 *src, u8 *real, struct insn *insn) return 0; /* Another subsystem puts a breakpoint, failed to recover */ - if (insn->opcode.bytes[0] == BREAKPOINT_INSTRUCTION) + if (insn->opcode.bytes[0] == INT3_INSN_OPCODE) return 0; /* We should not singlestep on the exception masking instructions */ @@ -400,14 +401,14 @@ static int prepare_boost(kprobe_opcode_t *buf, struct kprobe *p, int len = insn->length; if (can_boost(insn, p->addr) && - MAX_INSN_SIZE - len >= RELATIVEJUMP_SIZE) { + MAX_INSN_SIZE - len >= JMP32_INSN_SIZE) { /* * These instructions can be executed directly if it * jumps back to correct address. */ synthesize_reljump(buf + len, p->ainsn.insn + len, p->addr + insn->length); - len += RELATIVEJUMP_SIZE; + len += JMP32_INSN_SIZE; p->ainsn.boostable = true; } else { p->ainsn.boostable = false; @@ -501,12 +502,14 @@ int arch_prepare_kprobe(struct kprobe *p) void arch_arm_kprobe(struct kprobe *p) { - text_poke(p->addr, ((unsigned char []){BREAKPOINT_INSTRUCTION}), 1); + text_poke(p->addr, ((unsigned char []){INT3_INSN_OPCODE}), 1); + text_poke_sync(); } void arch_disarm_kprobe(struct kprobe *p) { text_poke(p->addr, &p->opcode, 1); + text_poke_sync(); } void arch_remove_kprobe(struct kprobe *p) @@ -609,7 +612,7 @@ static void setup_singlestep(struct kprobe *p, struct pt_regs *regs, regs->flags |= X86_EFLAGS_TF; regs->flags &= ~X86_EFLAGS_IF; /* single step inline if the instruction is an int3 */ - if (p->opcode == BREAKPOINT_INSTRUCTION) + if (p->opcode == INT3_INSN_OPCODE) regs->ip = (unsigned long)p->addr; else regs->ip = (unsigned long)p->ainsn.insn; @@ -695,7 +698,7 @@ int kprobe_int3_handler(struct pt_regs *regs) reset_current_kprobe(); return 1; } - } else if (*addr != BREAKPOINT_INSTRUCTION) { + } else if (*addr != INT3_INSN_OPCODE) { /* * The breakpoint instruction was removed right * after we hit it. Another cpu has removed diff --git a/arch/x86/kernel/kprobes/opt.c b/arch/x86/kernel/kprobes/opt.c index 8900329c28a7..3f45b5c43a71 100644 --- a/arch/x86/kernel/kprobes/opt.c +++ b/arch/x86/kernel/kprobes/opt.c @@ -38,7 +38,7 @@ unsigned long __recover_optprobed_insn(kprobe_opcode_t *buf, unsigned long addr) long offs; int i; - for (i = 0; i < RELATIVEJUMP_SIZE; i++) { + for (i = 0; i < JMP32_INSN_SIZE; i++) { kp = get_kprobe((void *)addr - i); /* This function only handles jump-optimized kprobe */ if (kp && kprobe_optimized(kp)) { @@ -62,10 +62,10 @@ found: if (addr == (unsigned long)kp->addr) { buf[0] = kp->opcode; - memcpy(buf + 1, op->optinsn.copied_insn, RELATIVE_ADDR_SIZE); + memcpy(buf + 1, op->optinsn.copied_insn, DISP32_SIZE); } else { offs = addr - (unsigned long)kp->addr - 1; - memcpy(buf, op->optinsn.copied_insn + offs, RELATIVE_ADDR_SIZE - offs); + memcpy(buf, op->optinsn.copied_insn + offs, DISP32_SIZE - offs); } return (unsigned long)buf; @@ -141,8 +141,6 @@ STACK_FRAME_NON_STANDARD(optprobe_template_func); #define TMPL_END_IDX \ ((long)optprobe_template_end - (long)optprobe_template_entry) -#define INT3_SIZE sizeof(kprobe_opcode_t) - /* Optimized kprobe call back function: called from optinsn */ static void optimized_callback(struct optimized_kprobe *op, struct pt_regs *regs) @@ -162,7 +160,7 @@ optimized_callback(struct optimized_kprobe *op, struct pt_regs *regs) regs->cs |= get_kernel_rpl(); regs->gs = 0; #endif - regs->ip = (unsigned long)op->kp.addr + INT3_SIZE; + regs->ip = (unsigned long)op->kp.addr + INT3_INSN_SIZE; regs->orig_ax = ~0UL; __this_cpu_write(current_kprobe, &op->kp); @@ -179,7 +177,7 @@ static int copy_optimized_instructions(u8 *dest, u8 *src, u8 *real) struct insn insn; int len = 0, ret; - while (len < RELATIVEJUMP_SIZE) { + while (len < JMP32_INSN_SIZE) { ret = __copy_instruction(dest + len, src + len, real + len, &insn); if (!ret || !can_boost(&insn, src + len)) return -EINVAL; @@ -271,7 +269,7 @@ static int can_optimize(unsigned long paddr) return 0; /* Check there is enough space for a relative jump. */ - if (size - offset < RELATIVEJUMP_SIZE) + if (size - offset < JMP32_INSN_SIZE) return 0; /* Decode instructions */ @@ -290,15 +288,15 @@ static int can_optimize(unsigned long paddr) kernel_insn_init(&insn, (void *)recovered_insn, MAX_INSN_SIZE); insn_get_length(&insn); /* Another subsystem puts a breakpoint */ - if (insn.opcode.bytes[0] == BREAKPOINT_INSTRUCTION) + if (insn.opcode.bytes[0] == INT3_INSN_OPCODE) return 0; /* Recover address */ insn.kaddr = (void *)addr; insn.next_byte = (void *)(addr + insn.length); /* Check any instructions don't jump into target */ if (insn_is_indirect_jump(&insn) || - insn_jump_into_range(&insn, paddr + INT3_SIZE, - RELATIVE_ADDR_SIZE)) + insn_jump_into_range(&insn, paddr + INT3_INSN_SIZE, + DISP32_SIZE)) return 0; addr += insn.length; } @@ -374,7 +372,7 @@ int arch_prepare_optimized_kprobe(struct optimized_kprobe *op, * Verify if the address gap is in 2GB range, because this uses * a relative jump. */ - rel = (long)slot - (long)op->kp.addr + RELATIVEJUMP_SIZE; + rel = (long)slot - (long)op->kp.addr + JMP32_INSN_SIZE; if (abs(rel) > 0x7fffffff) { ret = -ERANGE; goto err; @@ -401,7 +399,7 @@ int arch_prepare_optimized_kprobe(struct optimized_kprobe *op, /* Set returning jmp instruction at the tail of out-of-line buffer */ synthesize_reljump(buf + len, slot + len, (u8 *)op->kp.addr + op->optinsn.size); - len += RELATIVEJUMP_SIZE; + len += JMP32_INSN_SIZE; /* We have to use text_poke() for instruction buffer because it is RO */ text_poke(slot, buf, len); @@ -416,49 +414,50 @@ err: } /* - * Replace breakpoints (int3) with relative jumps. + * Replace breakpoints (INT3) with relative jumps (JMP.d32). * Caller must call with locking kprobe_mutex and text_mutex. + * + * The caller will have installed a regular kprobe and after that issued + * syncrhonize_rcu_tasks(), this ensures that the instruction(s) that live in + * the 4 bytes after the INT3 are unused and can now be overwritten. */ void arch_optimize_kprobes(struct list_head *oplist) { struct optimized_kprobe *op, *tmp; - u8 insn_buff[RELATIVEJUMP_SIZE]; + u8 insn_buff[JMP32_INSN_SIZE]; list_for_each_entry_safe(op, tmp, oplist, list) { s32 rel = (s32)((long)op->optinsn.insn - - ((long)op->kp.addr + RELATIVEJUMP_SIZE)); + ((long)op->kp.addr + JMP32_INSN_SIZE)); WARN_ON(kprobe_disabled(&op->kp)); /* Backup instructions which will be replaced by jump address */ - memcpy(op->optinsn.copied_insn, op->kp.addr + INT3_SIZE, - RELATIVE_ADDR_SIZE); + memcpy(op->optinsn.copied_insn, op->kp.addr + INT3_INSN_SIZE, + DISP32_SIZE); - insn_buff[0] = RELATIVEJUMP_OPCODE; + insn_buff[0] = JMP32_INSN_OPCODE; *(s32 *)(&insn_buff[1]) = rel; - text_poke_bp(op->kp.addr, insn_buff, RELATIVEJUMP_SIZE, NULL); + text_poke_bp(op->kp.addr, insn_buff, JMP32_INSN_SIZE, NULL); list_del_init(&op->list); } } -/* Replace a relative jump with a breakpoint (int3). */ +/* + * Replace a relative jump (JMP.d32) with a breakpoint (INT3). + * + * After that, we can restore the 4 bytes after the INT3 to undo what + * arch_optimize_kprobes() scribbled. This is safe since those bytes will be + * unused once the INT3 lands. + */ void arch_unoptimize_kprobe(struct optimized_kprobe *op) { - u8 insn_buff[RELATIVEJUMP_SIZE]; - u8 emulate_buff[RELATIVEJUMP_SIZE]; - - /* Set int3 to first byte for kprobes */ - insn_buff[0] = BREAKPOINT_INSTRUCTION; - memcpy(insn_buff + 1, op->optinsn.copied_insn, RELATIVE_ADDR_SIZE); - - emulate_buff[0] = RELATIVEJUMP_OPCODE; - *(s32 *)(&emulate_buff[1]) = (s32)((long)op->optinsn.insn - - ((long)op->kp.addr + RELATIVEJUMP_SIZE)); - - text_poke_bp(op->kp.addr, insn_buff, RELATIVEJUMP_SIZE, - emulate_buff); + arch_arm_kprobe(&op->kp); + text_poke(op->kp.addr + INT3_INSN_SIZE, + op->optinsn.copied_insn, DISP32_SIZE); + text_poke_sync(); } /* diff --git a/arch/x86/kernel/kvm.c b/arch/x86/kernel/kvm.c index 32ef1ee733b7..81045aabb6f4 100644 --- a/arch/x86/kernel/kvm.c +++ b/arch/x86/kernel/kvm.c @@ -245,17 +245,13 @@ NOKPROBE_SYMBOL(kvm_read_and_reset_pf_reason); dotraplinkage void do_async_page_fault(struct pt_regs *regs, unsigned long error_code, unsigned long address) { - enum ctx_state prev_state; - switch (kvm_read_and_reset_pf_reason()) { default: do_page_fault(regs, error_code, address); break; case KVM_PV_REASON_PAGE_NOT_PRESENT: /* page is swapped out by the host. */ - prev_state = exception_enter(); kvm_async_pf_task_wait((u32)address, !user_mode(regs)); - exception_exit(prev_state); break; case KVM_PV_REASON_PAGE_READY: rcu_irq_enter(); diff --git a/arch/x86/kernel/ldt.c b/arch/x86/kernel/ldt.c index b2463fcb20a8..c57e1ca70fd1 100644 --- a/arch/x86/kernel/ldt.c +++ b/arch/x86/kernel/ldt.c @@ -28,6 +28,89 @@ #include <asm/desc.h> #include <asm/mmu_context.h> #include <asm/syscalls.h> +#include <asm/pgtable_areas.h> + +/* This is a multiple of PAGE_SIZE. */ +#define LDT_SLOT_STRIDE (LDT_ENTRIES * LDT_ENTRY_SIZE) + +static inline void *ldt_slot_va(int slot) +{ + return (void *)(LDT_BASE_ADDR + LDT_SLOT_STRIDE * slot); +} + +void load_mm_ldt(struct mm_struct *mm) +{ + struct ldt_struct *ldt; + + /* READ_ONCE synchronizes with smp_store_release */ + ldt = READ_ONCE(mm->context.ldt); + + /* + * Any change to mm->context.ldt is followed by an IPI to all + * CPUs with the mm active. The LDT will not be freed until + * after the IPI is handled by all such CPUs. This means that, + * if the ldt_struct changes before we return, the values we see + * will be safe, and the new values will be loaded before we run + * any user code. + * + * NB: don't try to convert this to use RCU without extreme care. + * We would still need IRQs off, because we don't want to change + * the local LDT after an IPI loaded a newer value than the one + * that we can see. + */ + + if (unlikely(ldt)) { + if (static_cpu_has(X86_FEATURE_PTI)) { + if (WARN_ON_ONCE((unsigned long)ldt->slot > 1)) { + /* + * Whoops -- either the new LDT isn't mapped + * (if slot == -1) or is mapped into a bogus + * slot (if slot > 1). + */ + clear_LDT(); + return; + } + + /* + * If page table isolation is enabled, ldt->entries + * will not be mapped in the userspace pagetables. + * Tell the CPU to access the LDT through the alias + * at ldt_slot_va(ldt->slot). + */ + set_ldt(ldt_slot_va(ldt->slot), ldt->nr_entries); + } else { + set_ldt(ldt->entries, ldt->nr_entries); + } + } else { + clear_LDT(); + } +} + +void switch_ldt(struct mm_struct *prev, struct mm_struct *next) +{ + /* + * Load the LDT if either the old or new mm had an LDT. + * + * An mm will never go from having an LDT to not having an LDT. Two + * mms never share an LDT, so we don't gain anything by checking to + * see whether the LDT changed. There's also no guarantee that + * prev->context.ldt actually matches LDTR, but, if LDTR is non-NULL, + * then prev->context.ldt will also be non-NULL. + * + * If we really cared, we could optimize the case where prev == next + * and we're exiting lazy mode. Most of the time, if this happens, + * we don't actually need to reload LDTR, but modify_ldt() is mostly + * used by legacy code and emulators where we don't need this level of + * performance. + * + * This uses | instead of || because it generates better code. + */ + if (unlikely((unsigned long)prev->context.ldt | + (unsigned long)next->context.ldt)) + load_mm_ldt(next); + + DEBUG_LOCKS_WARN_ON(preemptible()); +} static void refresh_ldt_segments(void) { diff --git a/arch/x86/kernel/nmi.c b/arch/x86/kernel/nmi.c index e676a9916c49..54c21d6abd5a 100644 --- a/arch/x86/kernel/nmi.c +++ b/arch/x86/kernel/nmi.c @@ -104,18 +104,22 @@ static int __init nmi_warning_debugfs(void) } fs_initcall(nmi_warning_debugfs); -static void nmi_max_handler(struct irq_work *w) +static void nmi_check_duration(struct nmiaction *action, u64 duration) { - struct nmiaction *a = container_of(w, struct nmiaction, irq_work); + u64 whole_msecs = READ_ONCE(action->max_duration); int remainder_ns, decimal_msecs; - u64 whole_msecs = READ_ONCE(a->max_duration); + + if (duration < nmi_longest_ns || duration < action->max_duration) + return; + + action->max_duration = duration; remainder_ns = do_div(whole_msecs, (1000 * 1000)); decimal_msecs = remainder_ns / 1000; printk_ratelimited(KERN_INFO "INFO: NMI handler (%ps) took too long to run: %lld.%03d msecs\n", - a->handler, whole_msecs, decimal_msecs); + action->handler, whole_msecs, decimal_msecs); } static int nmi_handle(unsigned int type, struct pt_regs *regs) @@ -142,11 +146,7 @@ static int nmi_handle(unsigned int type, struct pt_regs *regs) delta = sched_clock() - delta; trace_nmi_handler(a->handler, (int)delta, thishandled); - if (delta < nmi_longest_ns || delta < a->max_duration) - continue; - - a->max_duration = delta; - irq_work_queue(&a->irq_work); + nmi_check_duration(a, delta); } rcu_read_unlock(); @@ -164,8 +164,6 @@ int __register_nmi_handler(unsigned int type, struct nmiaction *action) if (!action->handler) return -EINVAL; - init_irq_work(&action->irq_work, nmi_max_handler); - raw_spin_lock_irqsave(&desc->lock, flags); /* diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index 61e93a318983..839b5244e3b7 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -615,12 +615,8 @@ void speculation_ctrl_update_current(void) void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p) { - struct thread_struct *prev, *next; unsigned long tifp, tifn; - prev = &prev_p->thread; - next = &next_p->thread; - tifn = READ_ONCE(task_thread_info(next_p)->flags); tifp = READ_ONCE(task_thread_info(prev_p)->flags); diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c index 323499f48858..5052ced43373 100644 --- a/arch/x86/kernel/process_32.c +++ b/arch/x86/kernel/process_32.c @@ -124,7 +124,6 @@ start_thread(struct pt_regs *regs, unsigned long new_ip, unsigned long new_sp) regs->ip = new_ip; regs->sp = new_sp; regs->flags = X86_EFLAGS_IF; - force_iret(); } EXPORT_SYMBOL_GPL(start_thread); diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c index 506d66830d4d..ffd497804dbc 100644 --- a/arch/x86/kernel/process_64.c +++ b/arch/x86/kernel/process_64.c @@ -394,7 +394,6 @@ start_thread_common(struct pt_regs *regs, unsigned long new_ip, regs->cs = _cs; regs->ss = _ss; regs->flags = X86_EFLAGS_IF; - force_iret(); } void diff --git a/arch/x86/kernel/quirks.c b/arch/x86/kernel/quirks.c index 1daf8f2aa21f..896d74cb5081 100644 --- a/arch/x86/kernel/quirks.c +++ b/arch/x86/kernel/quirks.c @@ -110,7 +110,7 @@ static void ich_force_enable_hpet(struct pci_dev *dev) } /* use bits 31:14, 16 kB aligned */ - rcba_base = ioremap_nocache(rcba, 0x4000); + rcba_base = ioremap(rcba, 0x4000); if (rcba_base == NULL) { dev_printk(KERN_DEBUG, &dev->dev, "ioremap failed; " "cannot force enable HPET\n"); diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index cedfe2077a69..2441b64d061f 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -2,130 +2,54 @@ /* * Copyright (C) 1995 Linus Torvalds * - * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999 - * - * Memory region support - * David Parsons <orc@pell.chi.il.us>, July-August 1999 - * - * Added E820 sanitization routine (removes overlapping memory regions); - * Brian Moyle <bmoyle@mvista.com>, February 2001 - * - * Moved CPU detection code to cpu/${cpu}.c - * Patrick Mochel <mochel@osdl.org>, March 2002 - * - * Provisions for empty E820 memory regions (reported by certain BIOSes). - * Alex Achenbach <xela@slit.de>, December 2002. - * - */ - -/* - * This file handles the architecture-dependent parts of initialization + * This file contains the setup_arch() code, which handles the architecture-dependent + * parts of early kernel initialization. */ - -#include <linux/sched.h> -#include <linux/mm.h> -#include <linux/mmzone.h> -#include <linux/screen_info.h> -#include <linux/ioport.h> -#include <linux/acpi.h> -#include <linux/sfi.h> -#include <linux/apm_bios.h> -#include <linux/initrd.h> -#include <linux/memblock.h> -#include <linux/seq_file.h> #include <linux/console.h> -#include <linux/root_dev.h> -#include <linux/highmem.h> -#include <linux/export.h> +#include <linux/crash_dump.h> +#include <linux/dmi.h> #include <linux/efi.h> -#include <linux/init.h> -#include <linux/edd.h> +#include <linux/init_ohci1394_dma.h> +#include <linux/initrd.h> #include <linux/iscsi_ibft.h> -#include <linux/nodemask.h> -#include <linux/kexec.h> -#include <linux/dmi.h> -#include <linux/pfn.h> +#include <linux/memblock.h> #include <linux/pci.h> -#include <asm/pci-direct.h> -#include <linux/init_ohci1394_dma.h> -#include <linux/kvm_para.h> -#include <linux/dma-contiguous.h> -#include <xen/xen.h> -#include <uapi/linux/mount.h> - -#include <linux/errno.h> -#include <linux/kernel.h> -#include <linux/stddef.h> -#include <linux/unistd.h> -#include <linux/ptrace.h> -#include <linux/user.h> -#include <linux/delay.h> - -#include <linux/kallsyms.h> -#include <linux/cpufreq.h> -#include <linux/dma-mapping.h> -#include <linux/ctype.h> -#include <linux/uaccess.h> - -#include <linux/percpu.h> -#include <linux/crash_dump.h> +#include <linux/root_dev.h> +#include <linux/sfi.h> #include <linux/tboot.h> -#include <linux/jiffies.h> -#include <linux/mem_encrypt.h> -#include <linux/sizes.h> - #include <linux/usb/xhci-dbgp.h> -#include <video/edid.h> -#include <asm/mtrr.h> +#include <uapi/linux/mount.h> + +#include <xen/xen.h> + #include <asm/apic.h> -#include <asm/realmode.h> -#include <asm/e820/api.h> -#include <asm/mpspec.h> -#include <asm/setup.h> -#include <asm/efi.h> -#include <asm/timer.h> -#include <asm/i8259.h> -#include <asm/sections.h> -#include <asm/io_apic.h> -#include <asm/ist.h> -#include <asm/setup_arch.h> #include <asm/bios_ebda.h> -#include <asm/cacheflush.h> -#include <asm/processor.h> #include <asm/bugs.h> -#include <asm/kasan.h> - -#include <asm/vsyscall.h> #include <asm/cpu.h> -#include <asm/desc.h> -#include <asm/dma.h> -#include <asm/iommu.h> +#include <asm/efi.h> #include <asm/gart.h> -#include <asm/mmu_context.h> -#include <asm/proto.h> - -#include <asm/paravirt.h> #include <asm/hypervisor.h> -#include <asm/olpc_ofw.h> - -#include <asm/percpu.h> -#include <asm/topology.h> -#include <asm/apicdef.h> -#include <asm/amd_nb.h> +#include <asm/io_apic.h> +#include <asm/kasan.h> +#include <asm/kaslr.h> #include <asm/mce.h> -#include <asm/alternative.h> +#include <asm/mtrr.h> +#include <asm/realmode.h> +#include <asm/olpc_ofw.h> +#include <asm/pci-direct.h> #include <asm/prom.h> -#include <asm/microcode.h> -#include <asm/kaslr.h> +#include <asm/proto.h> #include <asm/unwind.h> +#include <asm/vsyscall.h> +#include <linux/vmalloc.h> /* - * max_low_pfn_mapped: highest direct mapped pfn under 4GB - * max_pfn_mapped: highest direct mapped pfn over 4GB + * max_low_pfn_mapped: highest directly mapped pfn < 4 GB + * max_pfn_mapped: highest directly mapped pfn > 4 GB * * The direct mapping only covers E820_TYPE_RAM regions, so the ranges and gaps are - * represented by pfn_mapped + * represented by pfn_mapped[]. */ unsigned long max_low_pfn_mapped; unsigned long max_pfn_mapped; @@ -135,14 +59,23 @@ RESERVE_BRK(dmi_alloc, 65536); #endif -static __initdata unsigned long _brk_start = (unsigned long)__brk_base; -unsigned long _brk_end = (unsigned long)__brk_base; +/* + * Range of the BSS area. The size of the BSS area is determined + * at link time, with RESERVE_BRK*() facility reserving additional + * chunks. + */ +static __initdata +unsigned long _brk_start = (unsigned long)__brk_base; +unsigned long _brk_end = (unsigned long)__brk_base; struct boot_params boot_params; /* - * Machine setup.. + * These are the four main kernel memory regions, we put them into + * the resource tree so that kdump tools and other debugging tools + * recover it: */ + static struct resource rodata_resource = { .name = "Kernel rodata", .start = 0, @@ -173,16 +106,16 @@ static struct resource bss_resource = { #ifdef CONFIG_X86_32 -/* cpu data as detected by the assembly code in head_32.S */ +/* CPU data as detected by the assembly code in head_32.S */ struct cpuinfo_x86 new_cpu_data; -/* common cpu data for all cpus */ +/* Common CPU data for all CPUs */ struct cpuinfo_x86 boot_cpu_data __read_mostly; EXPORT_SYMBOL(boot_cpu_data); unsigned int def_to_bigsmp; -/* for MCA, but anyone else can use it if they want */ +/* For MCA, but anyone else can use it if they want */ unsigned int machine_id; unsigned int machine_submodel_id; unsigned int BIOS_revision; @@ -468,15 +401,15 @@ static void __init memblock_x86_reserve_range_setup_data(void) /* * Keep the crash kernel below this limit. * - * On 32 bits earlier kernels would limit the kernel to the low 512 MiB + * Earlier 32-bits kernels would limit the kernel to the low 512 MB range * due to mapping restrictions. * - * On 64bit, kdump kernel need be restricted to be under 64TB, which is + * 64-bit kdump kernels need to be restricted to be under 64 TB, which is * the upper limit of system RAM in 4-level paging mode. Since the kdump - * jumping could be from 5-level to 4-level, the jumping will fail if - * kernel is put above 64TB, and there's no way to detect the paging mode - * of the kernel which will be loaded for dumping during the 1st kernel - * bootup. + * jump could be from 5-level paging to 4-level paging, the jump will fail if + * the kernel is put above 64 TB, and during the 1st kernel bootup there's + * no good way to detect the paging mode of the target kernel which will be + * loaded for dumping. */ #ifdef CONFIG_X86_32 # define CRASH_ADDR_LOW_MAX SZ_512M @@ -887,7 +820,7 @@ void __init setup_arch(char **cmdline_p) /* * Note: Quark X1000 CPUs advertise PGE incorrectly and require * a cr3 based tlb flush, so the following __flush_tlb_all() - * will not flush anything because the cpu quirk which clears + * will not flush anything because the CPU quirk which clears * X86_FEATURE_PGE has not been invoked yet. Though due to the * load_cr3() above the TLB has been flushed already. The * quirk is invoked before subsequent calls to __flush_tlb_all() diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c index 8eb7193e158d..8a29573851a3 100644 --- a/arch/x86/kernel/signal.c +++ b/arch/x86/kernel/signal.c @@ -151,8 +151,6 @@ static int restore_sigcontext(struct pt_regs *regs, err |= fpu__restore_sig(buf, IS_ENABLED(CONFIG_X86_32)); - force_iret(); - return err; } diff --git a/arch/x86/kernel/sysfb_simplefb.c b/arch/x86/kernel/sysfb_simplefb.c index 01f0e2263b86..298fc1edd9c9 100644 --- a/arch/x86/kernel/sysfb_simplefb.c +++ b/arch/x86/kernel/sysfb_simplefb.c @@ -90,11 +90,11 @@ __init int create_simplefb(const struct screen_info *si, if (si->orig_video_isVGA == VIDEO_TYPE_VLFB) size <<= 16; length = mode->height * mode->stride; - length = PAGE_ALIGN(length); if (length > size) { printk(KERN_WARNING "sysfb: VRAM smaller than advertised\n"); return -EINVAL; } + length = PAGE_ALIGN(length); /* setup IORESOURCE_MEM as framebuffer memory */ memset(&res, 0, sizeof(res)); diff --git a/arch/x86/kernel/tboot.c b/arch/x86/kernel/tboot.c index 4c61f0713832..b89f6ac6a0c0 100644 --- a/arch/x86/kernel/tboot.c +++ b/arch/x86/kernel/tboot.c @@ -354,7 +354,7 @@ static ssize_t tboot_log_read(struct file *file, char __user *user_buf, size_t c void *kbuf; int ret = -EFAULT; - log_base = ioremap_nocache(TBOOT_SERIAL_LOG_ADDR, TBOOT_SERIAL_LOG_SIZE); + log_base = ioremap(TBOOT_SERIAL_LOG_ADDR, TBOOT_SERIAL_LOG_SIZE); if (!log_base) return ret; diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 05da6b5b167b..9e6f822922a3 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -56,6 +56,8 @@ #include <asm/mpx.h> #include <asm/vm86.h> #include <asm/umip.h> +#include <asm/insn.h> +#include <asm/insn-eval.h> #ifdef CONFIG_X86_64 #include <asm/x86_init.h> @@ -518,11 +520,57 @@ exit_trap: do_trap(X86_TRAP_BR, SIGSEGV, "bounds", regs, error_code, 0, NULL); } -dotraplinkage void -do_general_protection(struct pt_regs *regs, long error_code) +enum kernel_gp_hint { + GP_NO_HINT, + GP_NON_CANONICAL, + GP_CANONICAL +}; + +/* + * When an uncaught #GP occurs, try to determine the memory address accessed by + * the instruction and return that address to the caller. Also, try to figure + * out whether any part of the access to that address was non-canonical. + */ +static enum kernel_gp_hint get_kernel_gp_address(struct pt_regs *regs, + unsigned long *addr) { - const char *desc = "general protection fault"; + u8 insn_buf[MAX_INSN_SIZE]; + struct insn insn; + + if (probe_kernel_read(insn_buf, (void *)regs->ip, MAX_INSN_SIZE)) + return GP_NO_HINT; + + kernel_insn_init(&insn, insn_buf, MAX_INSN_SIZE); + insn_get_modrm(&insn); + insn_get_sib(&insn); + + *addr = (unsigned long)insn_get_addr_ref(&insn, regs); + if (*addr == -1UL) + return GP_NO_HINT; + +#ifdef CONFIG_X86_64 + /* + * Check that: + * - the operand is not in the kernel half + * - the last byte of the operand is not in the user canonical half + */ + if (*addr < ~__VIRTUAL_MASK && + *addr + insn.opnd_bytes - 1 > __VIRTUAL_MASK) + return GP_NON_CANONICAL; +#endif + + return GP_CANONICAL; +} + +#define GPFSTR "general protection fault" + +dotraplinkage void do_general_protection(struct pt_regs *regs, long error_code) +{ + char desc[sizeof(GPFSTR) + 50 + 2*sizeof(unsigned long) + 1] = GPFSTR; + enum kernel_gp_hint hint = GP_NO_HINT; struct task_struct *tsk; + unsigned long gp_addr; + int ret; RCU_LOCKDEP_WARN(!rcu_is_watching(), "entry code didn't wake RCU"); cond_local_irq_enable(regs); @@ -539,48 +587,61 @@ do_general_protection(struct pt_regs *regs, long error_code) } tsk = current; - if (!user_mode(regs)) { - if (fixup_exception(regs, X86_TRAP_GP, error_code, 0)) - return; + if (user_mode(regs)) { tsk->thread.error_code = error_code; tsk->thread.trap_nr = X86_TRAP_GP; - /* - * To be potentially processing a kprobe fault and to - * trust the result from kprobe_running(), we have to - * be non-preemptible. - */ - if (!preemptible() && kprobe_running() && - kprobe_fault_handler(regs, X86_TRAP_GP)) - return; + show_signal(tsk, SIGSEGV, "", desc, regs, error_code); + force_sig(SIGSEGV); - if (notify_die(DIE_GPF, desc, regs, error_code, - X86_TRAP_GP, SIGSEGV) != NOTIFY_STOP) - die(desc, regs, error_code); return; } + if (fixup_exception(regs, X86_TRAP_GP, error_code, 0)) + return; + tsk->thread.error_code = error_code; tsk->thread.trap_nr = X86_TRAP_GP; - show_signal(tsk, SIGSEGV, "", desc, regs, error_code); + /* + * To be potentially processing a kprobe fault and to trust the result + * from kprobe_running(), we have to be non-preemptible. + */ + if (!preemptible() && + kprobe_running() && + kprobe_fault_handler(regs, X86_TRAP_GP)) + return; + + ret = notify_die(DIE_GPF, desc, regs, error_code, X86_TRAP_GP, SIGSEGV); + if (ret == NOTIFY_STOP) + return; + + if (error_code) + snprintf(desc, sizeof(desc), "segment-related " GPFSTR); + else + hint = get_kernel_gp_address(regs, &gp_addr); + + if (hint != GP_NO_HINT) + snprintf(desc, sizeof(desc), GPFSTR ", %s 0x%lx", + (hint == GP_NON_CANONICAL) ? "probably for non-canonical address" + : "maybe for address", + gp_addr); + + /* + * KASAN is interested only in the non-canonical case, clear it + * otherwise. + */ + if (hint != GP_NON_CANONICAL) + gp_addr = 0; + + die_addr(desc, regs, error_code, gp_addr); - force_sig(SIGSEGV); } NOKPROBE_SYMBOL(do_general_protection); dotraplinkage void notrace do_int3(struct pt_regs *regs, long error_code) { -#ifdef CONFIG_DYNAMIC_FTRACE - /* - * ftrace must be first, everything else may cause a recursive crash. - * See note by declaration of modifying_ftrace_code in ftrace.c - */ - if (unlikely(atomic_read(&modifying_ftrace_code)) && - ftrace_int3_handler(regs)) - return; -#endif if (poke_int3_handler(regs)) return; diff --git a/arch/x86/kernel/tsc_sync.c b/arch/x86/kernel/tsc_sync.c index b8acf639abd1..32a818764e03 100644 --- a/arch/x86/kernel/tsc_sync.c +++ b/arch/x86/kernel/tsc_sync.c @@ -233,7 +233,6 @@ static cycles_t check_tsc_warp(unsigned int timeout) * The measurement runs for 'timeout' msecs: */ end = start + (cycles_t) tsc_khz * timeout; - now = start; for (i = 0; ; i++) { /* diff --git a/arch/x86/kernel/unwind_orc.c b/arch/x86/kernel/unwind_orc.c index 332ae6530fa8..e9cc182aa97e 100644 --- a/arch/x86/kernel/unwind_orc.c +++ b/arch/x86/kernel/unwind_orc.c @@ -187,6 +187,8 @@ static struct orc_entry *orc_find(unsigned long ip) return orc_ftrace_find(ip); } +#ifdef CONFIG_MODULES + static void orc_sort_swap(void *_a, void *_b, int size) { struct orc_entry *orc_a, *orc_b; @@ -229,7 +231,6 @@ static int orc_sort_cmp(const void *_a, const void *_b) return orc_a->sp_reg == ORC_REG_UNDEFINED && !orc_a->end ? -1 : 1; } -#ifdef CONFIG_MODULES void unwind_module_init(struct module *mod, void *_orc_ip, size_t orc_ip_size, void *_orc, size_t orc_size) { @@ -273,9 +274,11 @@ void __init unwind_init(void) return; } - /* Sort the .orc_unwind and .orc_unwind_ip tables: */ - sort(__start_orc_unwind_ip, num_entries, sizeof(int), orc_sort_cmp, - orc_sort_swap); + /* + * Note, the orc_unwind and orc_unwind_ip tables were already + * sorted at build time via the 'sorttable' tool. + * It's ready for binary search straight away, no need to sort it. + */ /* Initialize the fast lookup table: */ lookup_num_blocks = orc_lookup_end - orc_lookup; diff --git a/arch/x86/kernel/vm86_32.c b/arch/x86/kernel/vm86_32.c index a76c12b38e92..91d55454e702 100644 --- a/arch/x86/kernel/vm86_32.c +++ b/arch/x86/kernel/vm86_32.c @@ -381,7 +381,6 @@ static long do_sys_vm86(struct vm86plus_struct __user *user_vm86, bool plus) mark_screen_rdonly(tsk->mm); memcpy((struct kernel_vm86_regs *)regs, &vm86regs, sizeof(vm86regs)); - force_iret(); return regs->ax; } diff --git a/arch/x86/kernel/vmlinux.lds.S b/arch/x86/kernel/vmlinux.lds.S index 3a1a819da137..e3296aa028fe 100644 --- a/arch/x86/kernel/vmlinux.lds.S +++ b/arch/x86/kernel/vmlinux.lds.S @@ -193,12 +193,10 @@ SECTIONS __vvar_beginning_hack = .; /* Place all vvars at the offsets in asm/vvar.h. */ -#define EMIT_VVAR(name, offset) \ +#define EMIT_VVAR(name, offset) \ . = __vvar_beginning_hack + offset; \ *(.vvar_ ## name) -#define __VVAR_KERNEL_LDS #include <asm/vvar.h> -#undef __VVAR_KERNEL_LDS #undef EMIT_VVAR /* diff --git a/arch/x86/kernel/x86_init.c b/arch/x86/kernel/x86_init.c index ce89430a7f80..23e25f3034c2 100644 --- a/arch/x86/kernel/x86_init.c +++ b/arch/x86/kernel/x86_init.c @@ -20,7 +20,7 @@ #include <asm/irq.h> #include <asm/io_apic.h> #include <asm/hpet.h> -#include <asm/pat.h> +#include <asm/memtype.h> #include <asm/tsc.h> #include <asm/iommu.h> #include <asm/mach_traps.h> diff --git a/arch/x86/kvm/Kconfig b/arch/x86/kvm/Kconfig index 840e12583b85..991019d5eee1 100644 --- a/arch/x86/kvm/Kconfig +++ b/arch/x86/kvm/Kconfig @@ -60,13 +60,11 @@ config KVM If unsure, say N. config KVM_INTEL - tristate "KVM for Intel processors support" - depends on KVM - # for perf_guest_get_msrs(): - depends on CPU_SUP_INTEL + tristate "KVM for Intel (and compatible) processors support" + depends on KVM && IA32_FEAT_CTL ---help--- - Provides support for KVM on Intel processors equipped with the VT - extensions. + Provides support for KVM on processors equipped with Intel's VT + extensions, a.k.a. Virtual Machine Extensions (VMX). To compile this as a module, choose M here: the module will be called kvm-intel. diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c index 6f92b40d798c..a32b847a8089 100644 --- a/arch/x86/kvm/mmu/mmu.c +++ b/arch/x86/kvm/mmu/mmu.c @@ -40,7 +40,7 @@ #include <linux/kthread.h> #include <asm/page.h> -#include <asm/pat.h> +#include <asm/memtype.h> #include <asm/cmpxchg.h> #include <asm/e820/api.h> #include <asm/io.h> diff --git a/arch/x86/kvm/vmx/nested.c b/arch/x86/kvm/vmx/nested.c index 4aea7d304beb..6879966b7648 100644 --- a/arch/x86/kvm/vmx/nested.c +++ b/arch/x86/kvm/vmx/nested.c @@ -4588,8 +4588,8 @@ static int handle_vmon(struct kvm_vcpu *vcpu) gpa_t vmptr; uint32_t revision; struct vcpu_vmx *vmx = to_vmx(vcpu); - const u64 VMXON_NEEDED_FEATURES = FEATURE_CONTROL_LOCKED - | FEATURE_CONTROL_VMXON_ENABLED_OUTSIDE_SMX; + const u64 VMXON_NEEDED_FEATURES = FEAT_CTL_LOCKED + | FEAT_CTL_VMX_ENABLED_OUTSIDE_SMX; /* * The Intel VMX Instruction Reference lists a bunch of bits that are diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c index e3394c839dea..cdb4bf50ee14 100644 --- a/arch/x86/kvm/vmx/vmx.c +++ b/arch/x86/kvm/vmx/vmx.c @@ -1839,11 +1839,11 @@ static int vmx_get_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info) case MSR_IA32_MCG_EXT_CTL: if (!msr_info->host_initiated && !(vmx->msr_ia32_feature_control & - FEATURE_CONTROL_LMCE)) + FEAT_CTL_LMCE_ENABLED)) return 1; msr_info->data = vcpu->arch.mcg_ext_ctl; break; - case MSR_IA32_FEATURE_CONTROL: + case MSR_IA32_FEAT_CTL: msr_info->data = vmx->msr_ia32_feature_control; break; case MSR_IA32_VMX_BASIC ... MSR_IA32_VMX_VMFUNC: @@ -2074,15 +2074,15 @@ static int vmx_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info) case MSR_IA32_MCG_EXT_CTL: if ((!msr_info->host_initiated && !(to_vmx(vcpu)->msr_ia32_feature_control & - FEATURE_CONTROL_LMCE)) || + FEAT_CTL_LMCE_ENABLED)) || (data & ~MCG_EXT_CTL_LMCE_EN)) return 1; vcpu->arch.mcg_ext_ctl = data; break; - case MSR_IA32_FEATURE_CONTROL: + case MSR_IA32_FEAT_CTL: if (!vmx_feature_control_msr_valid(vcpu, data) || (to_vmx(vcpu)->msr_ia32_feature_control & - FEATURE_CONTROL_LOCKED && !msr_info->host_initiated)) + FEAT_CTL_LOCKED && !msr_info->host_initiated)) return 1; vmx->msr_ia32_feature_control = data; if (msr_info->host_initiated && data == 0) @@ -2204,29 +2204,8 @@ static __init int cpu_has_kvm_support(void) static __init int vmx_disabled_by_bios(void) { - u64 msr; - - rdmsrl(MSR_IA32_FEATURE_CONTROL, msr); - if (msr & FEATURE_CONTROL_LOCKED) { - /* launched w/ TXT and VMX disabled */ - if (!(msr & FEATURE_CONTROL_VMXON_ENABLED_INSIDE_SMX) - && tboot_enabled()) - return 1; - /* launched w/o TXT and VMX only enabled w/ TXT */ - if (!(msr & FEATURE_CONTROL_VMXON_ENABLED_OUTSIDE_SMX) - && (msr & FEATURE_CONTROL_VMXON_ENABLED_INSIDE_SMX) - && !tboot_enabled()) { - printk(KERN_WARNING "kvm: disable TXT in the BIOS or " - "activate TXT before enabling KVM\n"); - return 1; - } - /* launched w/o TXT and VMX disabled */ - if (!(msr & FEATURE_CONTROL_VMXON_ENABLED_OUTSIDE_SMX) - && !tboot_enabled()) - return 1; - } - - return 0; + return !boot_cpu_has(X86_FEATURE_MSR_IA32_FEAT_CTL) || + !boot_cpu_has(X86_FEATURE_VMX); } static void kvm_cpu_vmxon(u64 addr) @@ -2241,7 +2220,6 @@ static int hardware_enable(void) { int cpu = raw_smp_processor_id(); u64 phys_addr = __pa(per_cpu(vmxarea, cpu)); - u64 old, test_bits; if (cr4_read_shadow() & X86_CR4_VMXE) return -EBUSY; @@ -2269,17 +2247,6 @@ static int hardware_enable(void) */ crash_enable_local_vmclear(cpu); - rdmsrl(MSR_IA32_FEATURE_CONTROL, old); - - test_bits = FEATURE_CONTROL_LOCKED; - test_bits |= FEATURE_CONTROL_VMXON_ENABLED_OUTSIDE_SMX; - if (tboot_enabled()) - test_bits |= FEATURE_CONTROL_VMXON_ENABLED_INSIDE_SMX; - - if ((old & test_bits) != test_bits) { - /* enable and lock */ - wrmsrl(MSR_IA32_FEATURE_CONTROL, old | test_bits); - } kvm_cpu_vmxon(phys_addr); if (enable_ept) ept_sync_global(); @@ -6801,7 +6768,7 @@ static struct kvm_vcpu *vmx_create_vcpu(struct kvm *kvm, unsigned int id) vmx->nested.posted_intr_nv = -1; vmx->nested.current_vmptr = -1ull; - vmx->msr_ia32_feature_control_valid_bits = FEATURE_CONTROL_LOCKED; + vmx->msr_ia32_feature_control_valid_bits = FEAT_CTL_LOCKED; /* * Enforce invariant: pi_desc.nv is always either POSTED_INTR_VECTOR @@ -6871,6 +6838,12 @@ static int __init vmx_check_processor_compat(void) struct vmcs_config vmcs_conf; struct vmx_capability vmx_cap; + if (!this_cpu_has(X86_FEATURE_MSR_IA32_FEAT_CTL) || + !this_cpu_has(X86_FEATURE_VMX)) { + pr_err("kvm: VMX is disabled on CPU %d\n", smp_processor_id()); + return -EIO; + } + if (setup_vmcs_config(&vmcs_conf, &vmx_cap) < 0) return -EIO; if (nested) @@ -7099,12 +7072,12 @@ static void vmx_cpuid_update(struct kvm_vcpu *vcpu) if (nested_vmx_allowed(vcpu)) to_vmx(vcpu)->msr_ia32_feature_control_valid_bits |= - FEATURE_CONTROL_VMXON_ENABLED_INSIDE_SMX | - FEATURE_CONTROL_VMXON_ENABLED_OUTSIDE_SMX; + FEAT_CTL_VMX_ENABLED_INSIDE_SMX | + FEAT_CTL_VMX_ENABLED_OUTSIDE_SMX; else to_vmx(vcpu)->msr_ia32_feature_control_valid_bits &= - ~(FEATURE_CONTROL_VMXON_ENABLED_INSIDE_SMX | - FEATURE_CONTROL_VMXON_ENABLED_OUTSIDE_SMX); + ~(FEAT_CTL_VMX_ENABLED_INSIDE_SMX | + FEAT_CTL_VMX_ENABLED_OUTSIDE_SMX); if (nested_vmx_allowed(vcpu)) { nested_vmx_cr_fixed1_bits_update(vcpu); @@ -7523,10 +7496,10 @@ static void vmx_setup_mce(struct kvm_vcpu *vcpu) { if (vcpu->arch.mcg_cap & MCG_LMCE_P) to_vmx(vcpu)->msr_ia32_feature_control_valid_bits |= - FEATURE_CONTROL_LMCE; + FEAT_CTL_LMCE_ENABLED; else to_vmx(vcpu)->msr_ia32_feature_control_valid_bits &= - ~FEATURE_CONTROL_LMCE; + ~FEAT_CTL_LMCE_ENABLED; } static int vmx_smi_allowed(struct kvm_vcpu *vcpu) diff --git a/arch/x86/kvm/vmx/vmx.h b/arch/x86/kvm/vmx/vmx.h index a4f7f737c5d4..7f42cf3dcd70 100644 --- a/arch/x86/kvm/vmx/vmx.h +++ b/arch/x86/kvm/vmx/vmx.h @@ -289,7 +289,7 @@ struct vcpu_vmx { /* * Only bits masked by msr_ia32_feature_control_valid_bits can be set in - * msr_ia32_feature_control. FEATURE_CONTROL_LOCKED is always included + * msr_ia32_feature_control. FEAT_CTL_LOCKED is always included * in msr_ia32_feature_control_valid_bits. */ u64 msr_ia32_feature_control; diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index cf917139de6b..740d3ee42455 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -1142,7 +1142,7 @@ static const u32 msrs_to_save_all[] = { MSR_CSTAR, MSR_KERNEL_GS_BASE, MSR_SYSCALL_MASK, MSR_LSTAR, #endif MSR_IA32_TSC, MSR_IA32_CR_PAT, MSR_VM_HSAVE_PA, - MSR_IA32_FEATURE_CONTROL, MSR_IA32_BNDCFGS, MSR_TSC_AUX, + MSR_IA32_FEAT_CTL, MSR_IA32_BNDCFGS, MSR_TSC_AUX, MSR_IA32_SPEC_CTRL, MSR_IA32_RTIT_CTL, MSR_IA32_RTIT_STATUS, MSR_IA32_RTIT_CR3_MATCH, MSR_IA32_RTIT_OUTPUT_BASE, MSR_IA32_RTIT_OUTPUT_MASK, diff --git a/arch/x86/lib/insn-eval.c b/arch/x86/lib/insn-eval.c index 306c3a0902ba..31600d851fd8 100644 --- a/arch/x86/lib/insn-eval.c +++ b/arch/x86/lib/insn-eval.c @@ -155,7 +155,7 @@ static bool check_seg_overrides(struct insn *insn, int regoff) */ static int resolve_default_seg(struct insn *insn, struct pt_regs *regs, int off) { - if (user_64bit_mode(regs)) + if (any_64bit_mode(regs)) return INAT_SEG_REG_IGNORE; /* * Resolve the default segment register as described in Section 3.7.4 @@ -266,7 +266,7 @@ static int resolve_seg_reg(struct insn *insn, struct pt_regs *regs, int regoff) * which may be invalid at this point. */ if (regoff == offsetof(struct pt_regs, ip)) { - if (user_64bit_mode(regs)) + if (any_64bit_mode(regs)) return INAT_SEG_REG_IGNORE; else return INAT_SEG_REG_CS; @@ -289,7 +289,7 @@ static int resolve_seg_reg(struct insn *insn, struct pt_regs *regs, int regoff) * In long mode, segment override prefixes are ignored, except for * overrides for FS and GS. */ - if (user_64bit_mode(regs)) { + if (any_64bit_mode(regs)) { if (idx != INAT_SEG_REG_FS && idx != INAT_SEG_REG_GS) idx = INAT_SEG_REG_IGNORE; @@ -646,23 +646,27 @@ unsigned long insn_get_seg_base(struct pt_regs *regs, int seg_reg_idx) */ return (unsigned long)(sel << 4); - if (user_64bit_mode(regs)) { + if (any_64bit_mode(regs)) { /* * Only FS or GS will have a base address, the rest of * the segments' bases are forced to 0. */ unsigned long base; - if (seg_reg_idx == INAT_SEG_REG_FS) + if (seg_reg_idx == INAT_SEG_REG_FS) { rdmsrl(MSR_FS_BASE, base); - else if (seg_reg_idx == INAT_SEG_REG_GS) + } else if (seg_reg_idx == INAT_SEG_REG_GS) { /* * swapgs was called at the kernel entry point. Thus, * MSR_KERNEL_GS_BASE will have the user-space GS base. */ - rdmsrl(MSR_KERNEL_GS_BASE, base); - else + if (user_mode(regs)) + rdmsrl(MSR_KERNEL_GS_BASE, base); + else + rdmsrl(MSR_GS_BASE, base); + } else { base = 0; + } return base; } @@ -703,7 +707,7 @@ static unsigned long get_seg_limit(struct pt_regs *regs, int seg_reg_idx) if (sel < 0) return 0; - if (user_64bit_mode(regs) || v8086_mode(regs)) + if (any_64bit_mode(regs) || v8086_mode(regs)) return -1L; if (!sel) @@ -948,7 +952,7 @@ static int get_eff_addr_modrm(struct insn *insn, struct pt_regs *regs, * following instruction. */ if (*regoff == -EDOM) { - if (user_64bit_mode(regs)) + if (any_64bit_mode(regs)) tmp = regs->ip + insn->length; else tmp = 0; @@ -1250,7 +1254,7 @@ static void __user *get_addr_ref_32(struct insn *insn, struct pt_regs *regs) * After computed, the effective address is treated as an unsigned * quantity. */ - if (!user_64bit_mode(regs) && ((unsigned int)eff_addr > seg_limit)) + if (!any_64bit_mode(regs) && ((unsigned int)eff_addr > seg_limit)) goto out; /* diff --git a/arch/x86/lib/memmove_64.S b/arch/x86/lib/memmove_64.S index 337830d7a59c..7ff00ea64e4f 100644 --- a/arch/x86/lib/memmove_64.S +++ b/arch/x86/lib/memmove_64.S @@ -29,10 +29,7 @@ SYM_FUNC_START_ALIAS(memmove) SYM_FUNC_START(__memmove) - /* Handle more 32 bytes in loop */ mov %rdi, %rax - cmp $0x20, %rdx - jb 1f /* Decide forward/backward copy mode */ cmp %rdi, %rsi @@ -42,7 +39,9 @@ SYM_FUNC_START(__memmove) cmp %rdi, %r8 jg 2f + /* FSRM implies ERMS => no length checks, do the copy directly */ .Lmemmove_begin_forward: + ALTERNATIVE "cmp $0x20, %rdx; jb 1f", "", X86_FEATURE_FSRM ALTERNATIVE "", "movq %rdx, %rcx; rep movsb; retq", X86_FEATURE_ERMS /* @@ -114,6 +113,8 @@ SYM_FUNC_START(__memmove) */ .p2align 4 2: + cmp $0x20, %rdx + jb 1f cmp $680, %rdx jb 6f cmp %dil, %sil diff --git a/arch/x86/mm/Makefile b/arch/x86/mm/Makefile index 3b89c201ac26..345848f270e3 100644 --- a/arch/x86/mm/Makefile +++ b/arch/x86/mm/Makefile @@ -12,8 +12,10 @@ CFLAGS_REMOVE_mem_encrypt.o = -pg CFLAGS_REMOVE_mem_encrypt_identity.o = -pg endif -obj-y := init.o init_$(BITS).o fault.o ioremap.o extable.o pageattr.o mmap.o \ - pat.o pgtable.o physaddr.o setup_nx.o tlb.o cpu_entry_area.o maccess.o +obj-y := init.o init_$(BITS).o fault.o ioremap.o extable.o mmap.o \ + pgtable.o physaddr.o setup_nx.o tlb.o cpu_entry_area.o maccess.o + +obj-y += pat/ # Make sure __phys_addr has no stackprotector nostackp := $(call cc-option, -fno-stack-protector) @@ -23,8 +25,6 @@ CFLAGS_mem_encrypt_identity.o := $(nostackp) CFLAGS_fault.o := -I $(srctree)/$(src)/../include/asm/trace -obj-$(CONFIG_X86_PAT) += pat_interval.o - obj-$(CONFIG_X86_32) += pgtable_32.o iomap_32.o obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c index 304d31d8cbbc..fa4ea09593ab 100644 --- a/arch/x86/mm/fault.c +++ b/arch/x86/mm/fault.c @@ -29,6 +29,7 @@ #include <asm/efi.h> /* efi_recover_from_page_fault()*/ #include <asm/desc.h> /* store_idt(), ... */ #include <asm/cpu_entry_area.h> /* exception stack */ +#include <asm/pgtable_areas.h> /* VMALLOC_START, ... */ #define CREATE_TRACE_POINTS #include <asm/trace/exceptions.h> @@ -1486,27 +1487,6 @@ good_area: } NOKPROBE_SYMBOL(do_user_addr_fault); -/* - * Explicitly marked noinline such that the function tracer sees this as the - * page_fault entry point. - */ -static noinline void -__do_page_fault(struct pt_regs *regs, unsigned long hw_error_code, - unsigned long address) -{ - prefetchw(¤t->mm->mmap_sem); - - if (unlikely(kmmio_fault(regs, address))) - return; - - /* Was the fault on kernel-controlled part of the address space? */ - if (unlikely(fault_in_kernel_space(address))) - do_kern_addr_fault(regs, hw_error_code, address); - else - do_user_addr_fault(regs, hw_error_code, address); -} -NOKPROBE_SYMBOL(__do_page_fault); - static __always_inline void trace_page_fault_entries(struct pt_regs *regs, unsigned long error_code, unsigned long address) @@ -1521,13 +1501,19 @@ trace_page_fault_entries(struct pt_regs *regs, unsigned long error_code, } dotraplinkage void -do_page_fault(struct pt_regs *regs, unsigned long error_code, unsigned long address) +do_page_fault(struct pt_regs *regs, unsigned long hw_error_code, + unsigned long address) { - enum ctx_state prev_state; + prefetchw(¤t->mm->mmap_sem); + trace_page_fault_entries(regs, hw_error_code, address); - prev_state = exception_enter(); - trace_page_fault_entries(regs, error_code, address); - __do_page_fault(regs, error_code, address); - exception_exit(prev_state); + if (unlikely(kmmio_fault(regs, address))) + return; + + /* Was the fault on kernel-controlled part of the address space? */ + if (unlikely(fault_in_kernel_space(address))) + do_kern_addr_fault(regs, hw_error_code, address); + else + do_user_addr_fault(regs, hw_error_code, address); } NOKPROBE_SYMBOL(do_page_fault); diff --git a/arch/x86/mm/init_32.c b/arch/x86/mm/init_32.c index 0a74407ef92e..23df4885bbed 100644 --- a/arch/x86/mm/init_32.c +++ b/arch/x86/mm/init_32.c @@ -52,6 +52,7 @@ #include <asm/page_types.h> #include <asm/cpu_entry_area.h> #include <asm/init.h> +#include <asm/pgtable_areas.h> #include "mm_internal.h" @@ -872,34 +873,6 @@ void arch_remove_memory(int nid, u64 start, u64 size, int kernel_set_to_readonly __read_mostly; -void set_kernel_text_rw(void) -{ - unsigned long start = PFN_ALIGN(_text); - unsigned long size = PFN_ALIGN(_etext) - start; - - if (!kernel_set_to_readonly) - return; - - pr_debug("Set kernel text: %lx - %lx for read write\n", - start, start+size); - - set_pages_rw(virt_to_page(start), size >> PAGE_SHIFT); -} - -void set_kernel_text_ro(void) -{ - unsigned long start = PFN_ALIGN(_text); - unsigned long size = PFN_ALIGN(_etext) - start; - - if (!kernel_set_to_readonly) - return; - - pr_debug("Set kernel text: %lx - %lx for read only\n", - start, start+size); - - set_pages_ro(virt_to_page(start), size >> PAGE_SHIFT); -} - static void mark_nxdata_nx(void) { /* diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index bcfede46fe02..abbdecb75fad 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -1258,42 +1258,6 @@ void __init mem_init(void) int kernel_set_to_readonly; -void set_kernel_text_rw(void) -{ - unsigned long start = PFN_ALIGN(_text); - unsigned long end = PFN_ALIGN(_etext); - - if (!kernel_set_to_readonly) - return; - - pr_debug("Set kernel text: %lx - %lx for read write\n", - start, end); - - /* - * Make the kernel identity mapping for text RW. Kernel text - * mapping will always be RO. Refer to the comment in - * static_protections() in pageattr.c - */ - set_memory_rw(start, (end - start) >> PAGE_SHIFT); -} - -void set_kernel_text_ro(void) -{ - unsigned long start = PFN_ALIGN(_text); - unsigned long end = PFN_ALIGN(_etext); - - if (!kernel_set_to_readonly) - return; - - pr_debug("Set kernel text: %lx - %lx for read only\n", - start, end); - - /* - * Set the kernel identity mapping for text RO. - */ - set_memory_ro(start, (end - start) >> PAGE_SHIFT); -} - void mark_rodata_ro(void) { unsigned long start = PFN_ALIGN(_text); diff --git a/arch/x86/mm/iomap_32.c b/arch/x86/mm/iomap_32.c index 6748b4c2baff..f60398aeb644 100644 --- a/arch/x86/mm/iomap_32.c +++ b/arch/x86/mm/iomap_32.c @@ -4,7 +4,7 @@ */ #include <asm/iomap.h> -#include <asm/pat.h> +#include <asm/memtype.h> #include <linux/export.h> #include <linux/highmem.h> @@ -26,7 +26,7 @@ int iomap_create_wc(resource_size_t base, unsigned long size, pgprot_t *prot) if (!is_io_mapping_possible(base, size)) return -EINVAL; - ret = io_reserve_memtype(base, base + size, &pcm); + ret = memtype_reserve_io(base, base + size, &pcm); if (ret) return ret; @@ -40,7 +40,7 @@ EXPORT_SYMBOL_GPL(iomap_create_wc); void iomap_free(resource_size_t base, unsigned long size) { - io_free_memtype(base, base + size); + memtype_free_io(base, base + size); } EXPORT_SYMBOL_GPL(iomap_free); diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c index b3a2936377b5..44e4beb4239f 100644 --- a/arch/x86/mm/ioremap.c +++ b/arch/x86/mm/ioremap.c @@ -24,7 +24,7 @@ #include <asm/pgtable.h> #include <asm/tlbflush.h> #include <asm/pgalloc.h> -#include <asm/pat.h> +#include <asm/memtype.h> #include <asm/setup.h> #include "physaddr.h" @@ -196,10 +196,10 @@ __ioremap_caller(resource_size_t phys_addr, unsigned long size, phys_addr &= PHYSICAL_PAGE_MASK; size = PAGE_ALIGN(last_addr+1) - phys_addr; - retval = reserve_memtype(phys_addr, (u64)phys_addr + size, + retval = memtype_reserve(phys_addr, (u64)phys_addr + size, pcm, &new_pcm); if (retval) { - printk(KERN_ERR "ioremap reserve_memtype failed %d\n", retval); + printk(KERN_ERR "ioremap memtype_reserve failed %d\n", retval); return NULL; } @@ -255,7 +255,7 @@ __ioremap_caller(resource_size_t phys_addr, unsigned long size, area->phys_addr = phys_addr; vaddr = (unsigned long) area->addr; - if (kernel_map_sync_memtype(phys_addr, size, pcm)) + if (memtype_kernel_map_sync(phys_addr, size, pcm)) goto err_free_area; if (ioremap_page_range(vaddr, vaddr + size, phys_addr, prot)) @@ -275,7 +275,7 @@ __ioremap_caller(resource_size_t phys_addr, unsigned long size, err_free_area: free_vm_area(area); err_free_memtype: - free_memtype(phys_addr, phys_addr + size); + memtype_free(phys_addr, phys_addr + size); return NULL; } @@ -451,7 +451,7 @@ void iounmap(volatile void __iomem *addr) return; } - free_memtype(p->phys_addr, p->phys_addr + get_vm_area_size(p)); + memtype_free(p->phys_addr, p->phys_addr + get_vm_area_size(p)); /* Finally remove it */ o = remove_vm_area((void __force *)addr); diff --git a/arch/x86/mm/kasan_init_64.c b/arch/x86/mm/kasan_init_64.c index cf5bc37c90ac..763e71abc0fe 100644 --- a/arch/x86/mm/kasan_init_64.c +++ b/arch/x86/mm/kasan_init_64.c @@ -288,23 +288,6 @@ static void __init kasan_shallow_populate_pgds(void *start, void *end) } while (pgd++, addr = next, addr != (unsigned long)end); } -#ifdef CONFIG_KASAN_INLINE -static int kasan_die_handler(struct notifier_block *self, - unsigned long val, - void *data) -{ - if (val == DIE_GPF) { - pr_emerg("CONFIG_KASAN_INLINE enabled\n"); - pr_emerg("GPF could be caused by NULL-ptr deref or user memory access\n"); - } - return NOTIFY_OK; -} - -static struct notifier_block kasan_die_notifier = { - .notifier_call = kasan_die_handler, -}; -#endif - void __init kasan_early_init(void) { int i; @@ -341,10 +324,6 @@ void __init kasan_init(void) int i; void *shadow_cpu_entry_begin, *shadow_cpu_entry_end; -#ifdef CONFIG_KASAN_INLINE - register_die_notifier(&kasan_die_notifier); -#endif - memcpy(early_top_pgt, init_top_pgt, sizeof(early_top_pgt)); /* diff --git a/arch/x86/mm/pat/Makefile b/arch/x86/mm/pat/Makefile new file mode 100644 index 000000000000..ea464c995161 --- /dev/null +++ b/arch/x86/mm/pat/Makefile @@ -0,0 +1,5 @@ +# SPDX-License-Identifier: GPL-2.0 + +obj-y := set_memory.o memtype.o + +obj-$(CONFIG_X86_PAT) += memtype_interval.o diff --git a/arch/x86/mm/pageattr-test.c b/arch/x86/mm/pat/cpa-test.c index facce271e8b9..facce271e8b9 100644 --- a/arch/x86/mm/pageattr-test.c +++ b/arch/x86/mm/pat/cpa-test.c diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat/memtype.c index 2d758e19ef22..394be8611748 100644 --- a/arch/x86/mm/pat.c +++ b/arch/x86/mm/pat/memtype.c @@ -1,11 +1,34 @@ // SPDX-License-Identifier: GPL-2.0-only /* - * Handle caching attributes in page tables (PAT) + * Page Attribute Table (PAT) support: handle memory caching attributes in page tables. * * Authors: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com> * Suresh B Siddha <suresh.b.siddha@intel.com> * * Loosely based on earlier PAT patchset from Eric Biederman and Andi Kleen. + * + * Basic principles: + * + * PAT is a CPU feature supported by all modern x86 CPUs, to allow the firmware and + * the kernel to set one of a handful of 'caching type' attributes for physical + * memory ranges: uncached, write-combining, write-through, write-protected, + * and the most commonly used and default attribute: write-back caching. + * + * PAT support supercedes and augments MTRR support in a compatible fashion: MTRR is + * a hardware interface to enumerate a limited number of physical memory ranges + * and set their caching attributes explicitly, programmed into the CPU via MSRs. + * Even modern CPUs have MTRRs enabled - but these are typically not touched + * by the kernel or by user-space (such as the X server), we rely on PAT for any + * additional cache attribute logic. + * + * PAT doesn't work via explicit memory ranges, but uses page table entries to add + * cache attribute information to the mapped memory range: there's 3 bits used, + * (_PAGE_PWT, _PAGE_PCD, _PAGE_PAT), with the 8 possible values mapped by the + * CPU to actual cache attributes via an MSR loaded into the CPU (MSR_IA32_CR_PAT). + * + * ( There's a metric ton of finer details, such as compatibility with CPU quirks + * that only support 4 types of PAT entries, and interaction with MTRRs, see + * below for details. ) */ #include <linux/seq_file.h> @@ -29,44 +52,48 @@ #include <asm/mtrr.h> #include <asm/page.h> #include <asm/msr.h> -#include <asm/pat.h> +#include <asm/memtype.h> #include <asm/io.h> -#include "pat_internal.h" -#include "mm_internal.h" +#include "memtype.h" +#include "../mm_internal.h" #undef pr_fmt #define pr_fmt(fmt) "" fmt -static bool __read_mostly boot_cpu_done; +static bool __read_mostly pat_bp_initialized; static bool __read_mostly pat_disabled = !IS_ENABLED(CONFIG_X86_PAT); -static bool __read_mostly pat_initialized; -static bool __read_mostly init_cm_done; +static bool __read_mostly pat_bp_enabled; +static bool __read_mostly pat_cm_initialized; -void pat_disable(const char *reason) +/* + * PAT support is enabled by default, but can be disabled for + * various user-requested or hardware-forced reasons: + */ +void pat_disable(const char *msg_reason) { if (pat_disabled) return; - if (boot_cpu_done) { + if (pat_bp_initialized) { WARN_ONCE(1, "x86/PAT: PAT cannot be disabled after initialization\n"); return; } pat_disabled = true; - pr_info("x86/PAT: %s\n", reason); + pr_info("x86/PAT: %s\n", msg_reason); } static int __init nopat(char *str) { - pat_disable("PAT support disabled."); + pat_disable("PAT support disabled via boot option."); return 0; } early_param("nopat", nopat); bool pat_enabled(void) { - return pat_initialized; + return pat_bp_enabled; } EXPORT_SYMBOL_GPL(pat_enabled); @@ -197,6 +224,8 @@ static void __init_cache_modes(u64 pat) char pat_msg[33]; int i; + WARN_ON_ONCE(pat_cm_initialized); + pat_msg[32] = 0; for (i = 7; i >= 0; i--) { cache = pat_get_cache_mode((pat >> (i * 8)) & 7, @@ -205,28 +234,28 @@ static void __init_cache_modes(u64 pat) } pr_info("x86/PAT: Configuration [0-7]: %s\n", pat_msg); - init_cm_done = true; + pat_cm_initialized = true; } #define PAT(x, y) ((u64)PAT_ ## y << ((x)*8)) -static void pat_bsp_init(u64 pat) +static void pat_bp_init(u64 pat) { u64 tmp_pat; if (!boot_cpu_has(X86_FEATURE_PAT)) { - pat_disable("PAT not supported by CPU."); + pat_disable("PAT not supported by the CPU."); return; } rdmsrl(MSR_IA32_CR_PAT, tmp_pat); if (!tmp_pat) { - pat_disable("PAT MSR is 0, disabled."); + pat_disable("PAT support disabled by the firmware."); return; } wrmsrl(MSR_IA32_CR_PAT, pat); - pat_initialized = true; + pat_bp_enabled = true; __init_cache_modes(pat); } @@ -248,7 +277,7 @@ void init_cache_modes(void) { u64 pat = 0; - if (init_cm_done) + if (pat_cm_initialized) return; if (boot_cpu_has(X86_FEATURE_PAT)) { @@ -291,7 +320,7 @@ void init_cache_modes(void) } /** - * pat_init - Initialize PAT MSR and PAT table + * pat_init - Initialize the PAT MSR and PAT table on the current CPU * * This function initializes PAT MSR and PAT table with an OS-defined value * to enable additional cache attributes, WC, WT and WP. @@ -305,6 +334,10 @@ void pat_init(void) u64 pat; struct cpuinfo_x86 *c = &boot_cpu_data; +#ifndef CONFIG_X86_PAT + pr_info_once("x86/PAT: PAT support disabled because CONFIG_X86_PAT is disabled in the kernel.\n"); +#endif + if (pat_disabled) return; @@ -364,9 +397,9 @@ void pat_init(void) PAT(4, WB) | PAT(5, WP) | PAT(6, UC_MINUS) | PAT(7, WT); } - if (!boot_cpu_done) { - pat_bsp_init(pat); - boot_cpu_done = true; + if (!pat_bp_initialized) { + pat_bp_init(pat); + pat_bp_initialized = true; } else { pat_ap_init(pat); } @@ -542,10 +575,10 @@ static u64 sanitize_phys(u64 address) * available type in new_type in case of no error. In case of any error * it will return a negative return value. */ -int reserve_memtype(u64 start, u64 end, enum page_cache_mode req_type, +int memtype_reserve(u64 start, u64 end, enum page_cache_mode req_type, enum page_cache_mode *new_type) { - struct memtype *new; + struct memtype *entry_new; enum page_cache_mode actual_type; int is_range_ram; int err = 0; @@ -593,22 +626,22 @@ int reserve_memtype(u64 start, u64 end, enum page_cache_mode req_type, return -EINVAL; } - new = kzalloc(sizeof(struct memtype), GFP_KERNEL); - if (!new) + entry_new = kzalloc(sizeof(struct memtype), GFP_KERNEL); + if (!entry_new) return -ENOMEM; - new->start = start; - new->end = end; - new->type = actual_type; + entry_new->start = start; + entry_new->end = end; + entry_new->type = actual_type; spin_lock(&memtype_lock); - err = memtype_check_insert(new, new_type); + err = memtype_check_insert(entry_new, new_type); if (err) { - pr_info("x86/PAT: reserve_memtype failed [mem %#010Lx-%#010Lx], track %s, req %s\n", + pr_info("x86/PAT: memtype_reserve failed [mem %#010Lx-%#010Lx], track %s, req %s\n", start, end - 1, - cattr_name(new->type), cattr_name(req_type)); - kfree(new); + cattr_name(entry_new->type), cattr_name(req_type)); + kfree(entry_new); spin_unlock(&memtype_lock); return err; @@ -616,18 +649,17 @@ int reserve_memtype(u64 start, u64 end, enum page_cache_mode req_type, spin_unlock(&memtype_lock); - dprintk("reserve_memtype added [mem %#010Lx-%#010Lx], track %s, req %s, ret %s\n", - start, end - 1, cattr_name(new->type), cattr_name(req_type), + dprintk("memtype_reserve added [mem %#010Lx-%#010Lx], track %s, req %s, ret %s\n", + start, end - 1, cattr_name(entry_new->type), cattr_name(req_type), new_type ? cattr_name(*new_type) : "-"); return err; } -int free_memtype(u64 start, u64 end) +int memtype_free(u64 start, u64 end) { - int err = -EINVAL; int is_range_ram; - struct memtype *entry; + struct memtype *entry_old; if (!pat_enabled()) return 0; @@ -640,28 +672,24 @@ int free_memtype(u64 start, u64 end) return 0; is_range_ram = pat_pagerange_is_ram(start, end); - if (is_range_ram == 1) { - - err = free_ram_pages_type(start, end); - - return err; - } else if (is_range_ram < 0) { + if (is_range_ram == 1) + return free_ram_pages_type(start, end); + if (is_range_ram < 0) return -EINVAL; - } spin_lock(&memtype_lock); - entry = memtype_erase(start, end); + entry_old = memtype_erase(start, end); spin_unlock(&memtype_lock); - if (IS_ERR(entry)) { + if (IS_ERR(entry_old)) { pr_info("x86/PAT: %s:%d freeing invalid memtype [mem %#010Lx-%#010Lx]\n", current->comm, current->pid, start, end - 1); return -EINVAL; } - kfree(entry); + kfree(entry_old); - dprintk("free_memtype request [mem %#010Lx-%#010Lx]\n", start, end - 1); + dprintk("memtype_free request [mem %#010Lx-%#010Lx]\n", start, end - 1); return 0; } @@ -700,6 +728,7 @@ static enum page_cache_mode lookup_memtype(u64 paddr) rettype = _PAGE_CACHE_MODE_UC_MINUS; spin_unlock(&memtype_lock); + return rettype; } @@ -723,7 +752,7 @@ bool pat_pfn_immune_to_uc_mtrr(unsigned long pfn) EXPORT_SYMBOL_GPL(pat_pfn_immune_to_uc_mtrr); /** - * io_reserve_memtype - Request a memory type mapping for a region of memory + * memtype_reserve_io - Request a memory type mapping for a region of memory * @start: start (physical address) of the region * @end: end (physical address) of the region * @type: A pointer to memtype, with requested type. On success, requested @@ -732,7 +761,7 @@ EXPORT_SYMBOL_GPL(pat_pfn_immune_to_uc_mtrr); * On success, returns 0 * On failure, returns non-zero */ -int io_reserve_memtype(resource_size_t start, resource_size_t end, +int memtype_reserve_io(resource_size_t start, resource_size_t end, enum page_cache_mode *type) { resource_size_t size = end - start; @@ -742,47 +771,47 @@ int io_reserve_memtype(resource_size_t start, resource_size_t end, WARN_ON_ONCE(iomem_map_sanity_check(start, size)); - ret = reserve_memtype(start, end, req_type, &new_type); + ret = memtype_reserve(start, end, req_type, &new_type); if (ret) goto out_err; if (!is_new_memtype_allowed(start, size, req_type, new_type)) goto out_free; - if (kernel_map_sync_memtype(start, size, new_type) < 0) + if (memtype_kernel_map_sync(start, size, new_type) < 0) goto out_free; *type = new_type; return 0; out_free: - free_memtype(start, end); + memtype_free(start, end); ret = -EBUSY; out_err: return ret; } /** - * io_free_memtype - Release a memory type mapping for a region of memory + * memtype_free_io - Release a memory type mapping for a region of memory * @start: start (physical address) of the region * @end: end (physical address) of the region */ -void io_free_memtype(resource_size_t start, resource_size_t end) +void memtype_free_io(resource_size_t start, resource_size_t end) { - free_memtype(start, end); + memtype_free(start, end); } int arch_io_reserve_memtype_wc(resource_size_t start, resource_size_t size) { enum page_cache_mode type = _PAGE_CACHE_MODE_WC; - return io_reserve_memtype(start, start + size, &type); + return memtype_reserve_io(start, start + size, &type); } EXPORT_SYMBOL(arch_io_reserve_memtype_wc); void arch_io_free_memtype_wc(resource_size_t start, resource_size_t size) { - io_free_memtype(start, start + size); + memtype_free_io(start, start + size); } EXPORT_SYMBOL(arch_io_free_memtype_wc); @@ -839,10 +868,10 @@ int phys_mem_access_prot_allowed(struct file *file, unsigned long pfn, } /* - * Change the memory type for the physial address range in kernel identity + * Change the memory type for the physical address range in kernel identity * mapping space if that range is a part of identity map. */ -int kernel_map_sync_memtype(u64 base, unsigned long size, +int memtype_kernel_map_sync(u64 base, unsigned long size, enum page_cache_mode pcm) { unsigned long id_sz; @@ -851,15 +880,14 @@ int kernel_map_sync_memtype(u64 base, unsigned long size, return 0; /* - * some areas in the middle of the kernel identity range - * are not mapped, like the PCI space. + * Some areas in the middle of the kernel identity range + * are not mapped, for example the PCI space. */ if (!page_is_ram(base >> PAGE_SHIFT)) return 0; id_sz = (__pa(high_memory-1) <= base + size) ? - __pa(high_memory) - base : - size; + __pa(high_memory) - base : size; if (ioremap_change_attr((unsigned long)__va(base), id_sz, pcm) < 0) { pr_info("x86/PAT: %s:%d ioremap_change_attr failed %s for [mem %#010Lx-%#010Lx]\n", @@ -873,7 +901,7 @@ int kernel_map_sync_memtype(u64 base, unsigned long size, /* * Internal interface to reserve a range of physical memory with prot. - * Reserved non RAM regions only and after successful reserve_memtype, + * Reserved non RAM regions only and after successful memtype_reserve, * this func also keeps identity mapping (if any) in sync with this new prot. */ static int reserve_pfn_range(u64 paddr, unsigned long size, pgprot_t *vma_prot, @@ -910,14 +938,14 @@ static int reserve_pfn_range(u64 paddr, unsigned long size, pgprot_t *vma_prot, return 0; } - ret = reserve_memtype(paddr, paddr + size, want_pcm, &pcm); + ret = memtype_reserve(paddr, paddr + size, want_pcm, &pcm); if (ret) return ret; if (pcm != want_pcm) { if (strict_prot || !is_new_memtype_allowed(paddr, size, want_pcm, pcm)) { - free_memtype(paddr, paddr + size); + memtype_free(paddr, paddr + size); pr_err("x86/PAT: %s:%d map pfn expected mapping type %s for [mem %#010Lx-%#010Lx], got %s\n", current->comm, current->pid, cattr_name(want_pcm), @@ -935,8 +963,8 @@ static int reserve_pfn_range(u64 paddr, unsigned long size, pgprot_t *vma_prot, cachemode2protval(pcm)); } - if (kernel_map_sync_memtype(paddr, size, pcm) < 0) { - free_memtype(paddr, paddr + size); + if (memtype_kernel_map_sync(paddr, size, pcm) < 0) { + memtype_free(paddr, paddr + size); return -EINVAL; } return 0; @@ -952,7 +980,7 @@ static void free_pfn_range(u64 paddr, unsigned long size) is_ram = pat_pagerange_is_ram(paddr, paddr + size); if (is_ram == 0) - free_memtype(paddr, paddr + size); + memtype_free(paddr, paddr + size); } /* @@ -1099,25 +1127,30 @@ EXPORT_SYMBOL_GPL(pgprot_writethrough); #if defined(CONFIG_DEBUG_FS) && defined(CONFIG_X86_PAT) +/* + * We are allocating a temporary printout-entry to be passed + * between seq_start()/next() and seq_show(): + */ static struct memtype *memtype_get_idx(loff_t pos) { - struct memtype *print_entry; + struct memtype *entry_print; int ret; - print_entry = kzalloc(sizeof(struct memtype), GFP_KERNEL); - if (!print_entry) + entry_print = kzalloc(sizeof(struct memtype), GFP_KERNEL); + if (!entry_print) return NULL; spin_lock(&memtype_lock); - ret = memtype_copy_nth_element(print_entry, pos); + ret = memtype_copy_nth_element(entry_print, pos); spin_unlock(&memtype_lock); - if (!ret) { - return print_entry; - } else { - kfree(print_entry); + /* Free it on error: */ + if (ret) { + kfree(entry_print); return NULL; } + + return entry_print; } static void *memtype_seq_start(struct seq_file *seq, loff_t *pos) @@ -1142,11 +1175,14 @@ static void memtype_seq_stop(struct seq_file *seq, void *v) static int memtype_seq_show(struct seq_file *seq, void *v) { - struct memtype *print_entry = (struct memtype *)v; + struct memtype *entry_print = (struct memtype *)v; + + seq_printf(seq, "PAT: [mem 0x%016Lx-0x%016Lx] %s\n", + entry_print->start, + entry_print->end, + cattr_name(entry_print->type)); - seq_printf(seq, "%s @ 0x%Lx-0x%Lx\n", cattr_name(print_entry->type), - print_entry->start, print_entry->end); - kfree(print_entry); + kfree(entry_print); return 0; } @@ -1178,7 +1214,6 @@ static int __init pat_memtype_list_init(void) } return 0; } - late_initcall(pat_memtype_list_init); #endif /* CONFIG_DEBUG_FS && CONFIG_X86_PAT */ diff --git a/arch/x86/mm/pat_internal.h b/arch/x86/mm/pat/memtype.h index 79a06684349e..cacecdbceb55 100644 --- a/arch/x86/mm/pat_internal.h +++ b/arch/x86/mm/pat/memtype.h @@ -1,6 +1,6 @@ /* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __PAT_INTERNAL_H_ -#define __PAT_INTERNAL_H_ +#ifndef __MEMTYPE_H_ +#define __MEMTYPE_H_ extern int pat_debug_enable; @@ -29,13 +29,13 @@ static inline char *cattr_name(enum page_cache_mode pcm) } #ifdef CONFIG_X86_PAT -extern int memtype_check_insert(struct memtype *new, +extern int memtype_check_insert(struct memtype *entry_new, enum page_cache_mode *new_type); extern struct memtype *memtype_erase(u64 start, u64 end); extern struct memtype *memtype_lookup(u64 addr); -extern int memtype_copy_nth_element(struct memtype *out, loff_t pos); +extern int memtype_copy_nth_element(struct memtype *entry_out, loff_t pos); #else -static inline int memtype_check_insert(struct memtype *new, +static inline int memtype_check_insert(struct memtype *entry_new, enum page_cache_mode *new_type) { return 0; } static inline struct memtype *memtype_erase(u64 start, u64 end) @@ -46,4 +46,4 @@ static inline int memtype_copy_nth_element(struct memtype *out, loff_t pos) { return 0; } #endif -#endif /* __PAT_INTERNAL_H_ */ +#endif /* __MEMTYPE_H_ */ diff --git a/arch/x86/mm/pat/memtype_interval.c b/arch/x86/mm/pat/memtype_interval.c new file mode 100644 index 000000000000..a07e4882bf36 --- /dev/null +++ b/arch/x86/mm/pat/memtype_interval.c @@ -0,0 +1,194 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Handle caching attributes in page tables (PAT) + * + * Authors: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com> + * Suresh B Siddha <suresh.b.siddha@intel.com> + * + * Interval tree used to store the PAT memory type reservations. + */ + +#include <linux/seq_file.h> +#include <linux/debugfs.h> +#include <linux/kernel.h> +#include <linux/interval_tree_generic.h> +#include <linux/sched.h> +#include <linux/gfp.h> + +#include <asm/pgtable.h> +#include <asm/memtype.h> + +#include "memtype.h" + +/* + * The memtype tree keeps track of memory type for specific + * physical memory areas. Without proper tracking, conflicting memory + * types in different mappings can cause CPU cache corruption. + * + * The tree is an interval tree (augmented rbtree) which tree is ordered + * by the starting address. The tree can contain multiple entries for + * different regions which overlap. All the aliases have the same + * cache attributes of course, as enforced by the PAT logic. + * + * memtype_lock protects the rbtree. + */ + +static inline u64 interval_start(struct memtype *entry) +{ + return entry->start; +} + +static inline u64 interval_end(struct memtype *entry) +{ + return entry->end - 1; +} + +INTERVAL_TREE_DEFINE(struct memtype, rb, u64, subtree_max_end, + interval_start, interval_end, + static, interval) + +static struct rb_root_cached memtype_rbroot = RB_ROOT_CACHED; + +enum { + MEMTYPE_EXACT_MATCH = 0, + MEMTYPE_END_MATCH = 1 +}; + +static struct memtype *memtype_match(u64 start, u64 end, int match_type) +{ + struct memtype *entry_match; + + entry_match = interval_iter_first(&memtype_rbroot, start, end-1); + + while (entry_match != NULL && entry_match->start < end) { + if ((match_type == MEMTYPE_EXACT_MATCH) && + (entry_match->start == start) && (entry_match->end == end)) + return entry_match; + + if ((match_type == MEMTYPE_END_MATCH) && + (entry_match->start < start) && (entry_match->end == end)) + return entry_match; + + entry_match = interval_iter_next(entry_match, start, end-1); + } + + return NULL; /* Returns NULL if there is no match */ +} + +static int memtype_check_conflict(u64 start, u64 end, + enum page_cache_mode reqtype, + enum page_cache_mode *newtype) +{ + struct memtype *entry_match; + enum page_cache_mode found_type = reqtype; + + entry_match = interval_iter_first(&memtype_rbroot, start, end-1); + if (entry_match == NULL) + goto success; + + if (entry_match->type != found_type && newtype == NULL) + goto failure; + + dprintk("Overlap at 0x%Lx-0x%Lx\n", entry_match->start, entry_match->end); + found_type = entry_match->type; + + entry_match = interval_iter_next(entry_match, start, end-1); + while (entry_match) { + if (entry_match->type != found_type) + goto failure; + + entry_match = interval_iter_next(entry_match, start, end-1); + } +success: + if (newtype) + *newtype = found_type; + + return 0; + +failure: + pr_info("x86/PAT: %s:%d conflicting memory types %Lx-%Lx %s<->%s\n", + current->comm, current->pid, start, end, + cattr_name(found_type), cattr_name(entry_match->type)); + + return -EBUSY; +} + +int memtype_check_insert(struct memtype *entry_new, enum page_cache_mode *ret_type) +{ + int err = 0; + + err = memtype_check_conflict(entry_new->start, entry_new->end, entry_new->type, ret_type); + if (err) + return err; + + if (ret_type) + entry_new->type = *ret_type; + + interval_insert(entry_new, &memtype_rbroot); + return 0; +} + +struct memtype *memtype_erase(u64 start, u64 end) +{ + struct memtype *entry_old; + + /* + * Since the memtype_rbroot tree allows overlapping ranges, + * memtype_erase() checks with EXACT_MATCH first, i.e. free + * a whole node for the munmap case. If no such entry is found, + * it then checks with END_MATCH, i.e. shrink the size of a node + * from the end for the mremap case. + */ + entry_old = memtype_match(start, end, MEMTYPE_EXACT_MATCH); + if (!entry_old) { + entry_old = memtype_match(start, end, MEMTYPE_END_MATCH); + if (!entry_old) + return ERR_PTR(-EINVAL); + } + + if (entry_old->start == start) { + /* munmap: erase this node */ + interval_remove(entry_old, &memtype_rbroot); + } else { + /* mremap: update the end value of this node */ + interval_remove(entry_old, &memtype_rbroot); + entry_old->end = start; + interval_insert(entry_old, &memtype_rbroot); + + return NULL; + } + + return entry_old; +} + +struct memtype *memtype_lookup(u64 addr) +{ + return interval_iter_first(&memtype_rbroot, addr, addr + PAGE_SIZE-1); +} + +/* + * Debugging helper, copy the Nth entry of the tree into a + * a copy for printout. This allows us to print out the tree + * via debugfs, without holding the memtype_lock too long: + */ +#ifdef CONFIG_DEBUG_FS +int memtype_copy_nth_element(struct memtype *entry_out, loff_t pos) +{ + struct memtype *entry_match; + int i = 1; + + entry_match = interval_iter_first(&memtype_rbroot, 0, ULONG_MAX); + + while (entry_match && pos != i) { + entry_match = interval_iter_next(entry_match, 0, ULONG_MAX); + i++; + } + + if (entry_match) { /* pos == i */ + *entry_out = *entry_match; + return 0; + } else { + return 1; + } +} +#endif diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pat/set_memory.c index 1b99ad05b117..62a8ebe72a52 100644 --- a/arch/x86/mm/pageattr.c +++ b/arch/x86/mm/pat/set_memory.c @@ -24,10 +24,10 @@ #include <linux/uaccess.h> #include <asm/pgalloc.h> #include <asm/proto.h> -#include <asm/pat.h> +#include <asm/memtype.h> #include <asm/set_memory.h> -#include "mm_internal.h" +#include "../mm_internal.h" /* * The current flushing context - we pass it instead of 5 arguments: @@ -331,7 +331,7 @@ static void cpa_flush_all(unsigned long cache) on_each_cpu(__cpa_flush_all, (void *) cache, 1); } -void __cpa_flush_tlb(void *data) +static void __cpa_flush_tlb(void *data) { struct cpa_data *cpa = data; unsigned int i; @@ -1801,7 +1801,7 @@ int set_memory_uc(unsigned long addr, int numpages) /* * for now UC MINUS. see comments in ioremap() */ - ret = reserve_memtype(__pa(addr), __pa(addr) + numpages * PAGE_SIZE, + ret = memtype_reserve(__pa(addr), __pa(addr) + numpages * PAGE_SIZE, _PAGE_CACHE_MODE_UC_MINUS, NULL); if (ret) goto out_err; @@ -1813,7 +1813,7 @@ int set_memory_uc(unsigned long addr, int numpages) return 0; out_free: - free_memtype(__pa(addr), __pa(addr) + numpages * PAGE_SIZE); + memtype_free(__pa(addr), __pa(addr) + numpages * PAGE_SIZE); out_err: return ret; } @@ -1839,14 +1839,14 @@ int set_memory_wc(unsigned long addr, int numpages) { int ret; - ret = reserve_memtype(__pa(addr), __pa(addr) + numpages * PAGE_SIZE, + ret = memtype_reserve(__pa(addr), __pa(addr) + numpages * PAGE_SIZE, _PAGE_CACHE_MODE_WC, NULL); if (ret) return ret; ret = _set_memory_wc(addr, numpages); if (ret) - free_memtype(__pa(addr), __pa(addr) + numpages * PAGE_SIZE); + memtype_free(__pa(addr), __pa(addr) + numpages * PAGE_SIZE); return ret; } @@ -1873,7 +1873,7 @@ int set_memory_wb(unsigned long addr, int numpages) if (ret) return ret; - free_memtype(__pa(addr), __pa(addr) + numpages * PAGE_SIZE); + memtype_free(__pa(addr), __pa(addr) + numpages * PAGE_SIZE); return 0; } EXPORT_SYMBOL(set_memory_wb); @@ -2014,7 +2014,7 @@ static int _set_pages_array(struct page **pages, int numpages, continue; start = page_to_pfn(pages[i]) << PAGE_SHIFT; end = start + PAGE_SIZE; - if (reserve_memtype(start, end, new_type, NULL)) + if (memtype_reserve(start, end, new_type, NULL)) goto err_out; } @@ -2040,7 +2040,7 @@ err_out: continue; start = page_to_pfn(pages[i]) << PAGE_SHIFT; end = start + PAGE_SIZE; - free_memtype(start, end); + memtype_free(start, end); } return -EINVAL; } @@ -2089,7 +2089,7 @@ int set_pages_array_wb(struct page **pages, int numpages) continue; start = page_to_pfn(pages[i]) << PAGE_SHIFT; end = start + PAGE_SIZE; - free_memtype(start, end); + memtype_free(start, end); } return 0; @@ -2215,7 +2215,7 @@ int __init kernel_map_pages_in_pgd(pgd_t *pgd, u64 pfn, unsigned long address, .pgd = pgd, .numpages = numpages, .mask_set = __pgprot(0), - .mask_clr = __pgprot(0), + .mask_clr = __pgprot(~page_flags & (_PAGE_NX|_PAGE_RW)), .flags = 0, }; @@ -2224,12 +2224,6 @@ int __init kernel_map_pages_in_pgd(pgd_t *pgd, u64 pfn, unsigned long address, if (!(__supported_pte_mask & _PAGE_NX)) goto out; - if (!(page_flags & _PAGE_NX)) - cpa.mask_clr = __pgprot(_PAGE_NX); - - if (!(page_flags & _PAGE_RW)) - cpa.mask_clr = __pgprot(_PAGE_RW); - if (!(page_flags & _PAGE_ENC)) cpa.mask_clr = pgprot_encrypted(cpa.mask_clr); @@ -2281,5 +2275,5 @@ int __init kernel_unmap_pages_in_pgd(pgd_t *pgd, unsigned long address, * be exposed to the rest of the kernel. Include these directly here. */ #ifdef CONFIG_CPA_DEBUG -#include "pageattr-test.c" +#include "cpa-test.c" #endif diff --git a/arch/x86/mm/pat_interval.c b/arch/x86/mm/pat_interval.c deleted file mode 100644 index 6855362eaf21..000000000000 --- a/arch/x86/mm/pat_interval.c +++ /dev/null @@ -1,185 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -/* - * Handle caching attributes in page tables (PAT) - * - * Authors: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com> - * Suresh B Siddha <suresh.b.siddha@intel.com> - * - * Interval tree used to store the PAT memory type reservations. - */ - -#include <linux/seq_file.h> -#include <linux/debugfs.h> -#include <linux/kernel.h> -#include <linux/interval_tree_generic.h> -#include <linux/sched.h> -#include <linux/gfp.h> - -#include <asm/pgtable.h> -#include <asm/pat.h> - -#include "pat_internal.h" - -/* - * The memtype tree keeps track of memory type for specific - * physical memory areas. Without proper tracking, conflicting memory - * types in different mappings can cause CPU cache corruption. - * - * The tree is an interval tree (augmented rbtree) with tree ordered - * on starting address. Tree can contain multiple entries for - * different regions which overlap. All the aliases have the same - * cache attributes of course. - * - * memtype_lock protects the rbtree. - */ -static inline u64 memtype_interval_start(struct memtype *memtype) -{ - return memtype->start; -} - -static inline u64 memtype_interval_end(struct memtype *memtype) -{ - return memtype->end - 1; -} -INTERVAL_TREE_DEFINE(struct memtype, rb, u64, subtree_max_end, - memtype_interval_start, memtype_interval_end, - static, memtype_interval) - -static struct rb_root_cached memtype_rbroot = RB_ROOT_CACHED; - -enum { - MEMTYPE_EXACT_MATCH = 0, - MEMTYPE_END_MATCH = 1 -}; - -static struct memtype *memtype_match(u64 start, u64 end, int match_type) -{ - struct memtype *match; - - match = memtype_interval_iter_first(&memtype_rbroot, start, end-1); - while (match != NULL && match->start < end) { - if ((match_type == MEMTYPE_EXACT_MATCH) && - (match->start == start) && (match->end == end)) - return match; - - if ((match_type == MEMTYPE_END_MATCH) && - (match->start < start) && (match->end == end)) - return match; - - match = memtype_interval_iter_next(match, start, end-1); - } - - return NULL; /* Returns NULL if there is no match */ -} - -static int memtype_check_conflict(u64 start, u64 end, - enum page_cache_mode reqtype, - enum page_cache_mode *newtype) -{ - struct memtype *match; - enum page_cache_mode found_type = reqtype; - - match = memtype_interval_iter_first(&memtype_rbroot, start, end-1); - if (match == NULL) - goto success; - - if (match->type != found_type && newtype == NULL) - goto failure; - - dprintk("Overlap at 0x%Lx-0x%Lx\n", match->start, match->end); - found_type = match->type; - - match = memtype_interval_iter_next(match, start, end-1); - while (match) { - if (match->type != found_type) - goto failure; - - match = memtype_interval_iter_next(match, start, end-1); - } -success: - if (newtype) - *newtype = found_type; - - return 0; - -failure: - pr_info("x86/PAT: %s:%d conflicting memory types %Lx-%Lx %s<->%s\n", - current->comm, current->pid, start, end, - cattr_name(found_type), cattr_name(match->type)); - return -EBUSY; -} - -int memtype_check_insert(struct memtype *new, - enum page_cache_mode *ret_type) -{ - int err = 0; - - err = memtype_check_conflict(new->start, new->end, new->type, ret_type); - if (err) - return err; - - if (ret_type) - new->type = *ret_type; - - memtype_interval_insert(new, &memtype_rbroot); - return 0; -} - -struct memtype *memtype_erase(u64 start, u64 end) -{ - struct memtype *data; - - /* - * Since the memtype_rbroot tree allows overlapping ranges, - * memtype_erase() checks with EXACT_MATCH first, i.e. free - * a whole node for the munmap case. If no such entry is found, - * it then checks with END_MATCH, i.e. shrink the size of a node - * from the end for the mremap case. - */ - data = memtype_match(start, end, MEMTYPE_EXACT_MATCH); - if (!data) { - data = memtype_match(start, end, MEMTYPE_END_MATCH); - if (!data) - return ERR_PTR(-EINVAL); - } - - if (data->start == start) { - /* munmap: erase this node */ - memtype_interval_remove(data, &memtype_rbroot); - } else { - /* mremap: update the end value of this node */ - memtype_interval_remove(data, &memtype_rbroot); - data->end = start; - memtype_interval_insert(data, &memtype_rbroot); - return NULL; - } - - return data; -} - -struct memtype *memtype_lookup(u64 addr) -{ - return memtype_interval_iter_first(&memtype_rbroot, addr, - addr + PAGE_SIZE-1); -} - -#if defined(CONFIG_DEBUG_FS) -int memtype_copy_nth_element(struct memtype *out, loff_t pos) -{ - struct memtype *match; - int i = 1; - - match = memtype_interval_iter_first(&memtype_rbroot, 0, ULONG_MAX); - while (match && pos != i) { - match = memtype_interval_iter_next(match, 0, ULONG_MAX); - i++; - } - - if (match) { /* pos == i */ - *out = *match; - return 0; - } else { - return 1; - } -} -#endif diff --git a/arch/x86/mm/pgtable_32.c b/arch/x86/mm/pgtable_32.c index 9bb7f0ab9fe6..0e6700eaa4f9 100644 --- a/arch/x86/mm/pgtable_32.c +++ b/arch/x86/mm/pgtable_32.c @@ -18,6 +18,7 @@ #include <asm/tlb.h> #include <asm/tlbflush.h> #include <asm/io.h> +#include <linux/vmalloc.h> unsigned int __VMALLOC_RESERVE = 128 << 20; diff --git a/arch/x86/mm/physaddr.c b/arch/x86/mm/physaddr.c index bdc98150d4db..fc3f3d3e2ef2 100644 --- a/arch/x86/mm/physaddr.c +++ b/arch/x86/mm/physaddr.c @@ -5,6 +5,7 @@ #include <linux/mm.h> #include <asm/page.h> +#include <linux/vmalloc.h> #include "physaddr.h" diff --git a/arch/x86/mm/testmmiotrace.c b/arch/x86/mm/testmmiotrace.c index 92153d054d6c..bda73cb7a044 100644 --- a/arch/x86/mm/testmmiotrace.c +++ b/arch/x86/mm/testmmiotrace.c @@ -79,7 +79,7 @@ static void do_read_far_test(void __iomem *p) static void do_test(unsigned long size) { - void __iomem *p = ioremap_nocache(mmio_address, size); + void __iomem *p = ioremap(mmio_address, size); if (!p) { pr_err("could not ioremap, aborting.\n"); return; @@ -104,7 +104,7 @@ static void do_test_bulk_ioremapping(void) int i; for (i = 0; i < 10; ++i) { - p = ioremap_nocache(mmio_address, PAGE_SIZE); + p = ioremap(mmio_address, PAGE_SIZE); if (p) iounmap(p); } diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c index e6a9edc5baaf..66f96f21a7b6 100644 --- a/arch/x86/mm/tlb.c +++ b/arch/x86/mm/tlb.c @@ -708,7 +708,7 @@ void native_flush_tlb_others(const struct cpumask *cpumask, (void *)info, 1); else on_each_cpu_cond_mask(tlb_is_not_lazy, flush_tlb_func_remote, - (void *)info, 1, GFP_ATOMIC, cpumask); + (void *)info, 1, cpumask); } /* diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c index b8be18427277..9ba08e9abc09 100644 --- a/arch/x86/net/bpf_jit_comp.c +++ b/arch/x86/net/bpf_jit_comp.c @@ -10,10 +10,12 @@ #include <linux/if_vlan.h> #include <linux/bpf.h> #include <linux/memory.h> +#include <linux/sort.h> #include <asm/extable.h> #include <asm/set_memory.h> #include <asm/nospec-branch.h> #include <asm/text-patching.h> +#include <asm/asm-prototypes.h> static u8 *emit_code(u8 *ptr, u32 bytes, unsigned int len) { @@ -1326,7 +1328,7 @@ emit_jmp: return proglen; } -static void save_regs(struct btf_func_model *m, u8 **prog, int nr_args, +static void save_regs(const struct btf_func_model *m, u8 **prog, int nr_args, int stack_size) { int i; @@ -1342,7 +1344,7 @@ static void save_regs(struct btf_func_model *m, u8 **prog, int nr_args, -(stack_size - i * 8)); } -static void restore_regs(struct btf_func_model *m, u8 **prog, int nr_args, +static void restore_regs(const struct btf_func_model *m, u8 **prog, int nr_args, int stack_size) { int i; @@ -1359,7 +1361,7 @@ static void restore_regs(struct btf_func_model *m, u8 **prog, int nr_args, -(stack_size - i * 8)); } -static int invoke_bpf(struct btf_func_model *m, u8 **pprog, +static int invoke_bpf(const struct btf_func_model *m, u8 **pprog, struct bpf_prog **progs, int prog_cnt, int stack_size) { u8 *prog = *pprog; @@ -1454,7 +1456,8 @@ static int invoke_bpf(struct btf_func_model *m, u8 **pprog, * add rsp, 8 // skip eth_type_trans's frame * ret // return to its caller */ -int arch_prepare_bpf_trampoline(void *image, struct btf_func_model *m, u32 flags, +int arch_prepare_bpf_trampoline(void *image, void *image_end, + const struct btf_func_model *m, u32 flags, struct bpf_prog **fentry_progs, int fentry_cnt, struct bpf_prog **fexit_progs, int fexit_cnt, void *orig_call) @@ -1521,15 +1524,160 @@ int arch_prepare_bpf_trampoline(void *image, struct btf_func_model *m, u32 flags /* skip our return address and return to parent */ EMIT4(0x48, 0x83, 0xC4, 8); /* add rsp, 8 */ EMIT1(0xC3); /* ret */ - /* One half of the page has active running trampoline. - * Another half is an area for next trampoline. - * Make sure the trampoline generation logic doesn't overflow. - */ - if (WARN_ON_ONCE(prog - (u8 *)image > PAGE_SIZE / 2 - BPF_INSN_SAFETY)) + /* Make sure the trampoline generation logic doesn't overflow */ + if (WARN_ON_ONCE(prog > (u8 *)image_end - BPF_INSN_SAFETY)) return -EFAULT; + return prog - (u8 *)image; +} + +static int emit_cond_near_jump(u8 **pprog, void *func, void *ip, u8 jmp_cond) +{ + u8 *prog = *pprog; + int cnt = 0; + s64 offset; + + offset = func - (ip + 2 + 4); + if (!is_simm32(offset)) { + pr_err("Target %p is out of range\n", func); + return -EINVAL; + } + EMIT2_off32(0x0F, jmp_cond + 0x10, offset); + *pprog = prog; + return 0; +} + +static void emit_nops(u8 **pprog, unsigned int len) +{ + unsigned int i, noplen; + u8 *prog = *pprog; + int cnt = 0; + + while (len > 0) { + noplen = len; + + if (noplen > ASM_NOP_MAX) + noplen = ASM_NOP_MAX; + + for (i = 0; i < noplen; i++) + EMIT1(ideal_nops[noplen][i]); + len -= noplen; + } + + *pprog = prog; +} + +static int emit_fallback_jump(u8 **pprog) +{ + u8 *prog = *pprog; + int err = 0; + +#ifdef CONFIG_RETPOLINE + /* Note that this assumes the the compiler uses external + * thunks for indirect calls. Both clang and GCC use the same + * naming convention for external thunks. + */ + err = emit_jump(&prog, __x86_indirect_thunk_rdx, prog); +#else + int cnt = 0; + + EMIT2(0xFF, 0xE2); /* jmp rdx */ +#endif + *pprog = prog; + return err; +} + +static int emit_bpf_dispatcher(u8 **pprog, int a, int b, s64 *progs) +{ + u8 *jg_reloc, *jg_target, *prog = *pprog; + int pivot, err, jg_bytes = 1, cnt = 0; + s64 jg_offset; + + if (a == b) { + /* Leaf node of recursion, i.e. not a range of indices + * anymore. + */ + EMIT1(add_1mod(0x48, BPF_REG_3)); /* cmp rdx,func */ + if (!is_simm32(progs[a])) + return -1; + EMIT2_off32(0x81, add_1reg(0xF8, BPF_REG_3), + progs[a]); + err = emit_cond_near_jump(&prog, /* je func */ + (void *)progs[a], prog, + X86_JE); + if (err) + return err; + + err = emit_fallback_jump(&prog); /* jmp thunk/indirect */ + if (err) + return err; + + *pprog = prog; + return 0; + } + + /* Not a leaf node, so we pivot, and recursively descend into + * the lower and upper ranges. + */ + pivot = (b - a) / 2; + EMIT1(add_1mod(0x48, BPF_REG_3)); /* cmp rdx,func */ + if (!is_simm32(progs[a + pivot])) + return -1; + EMIT2_off32(0x81, add_1reg(0xF8, BPF_REG_3), progs[a + pivot]); + + if (pivot > 2) { /* jg upper_part */ + /* Require near jump. */ + jg_bytes = 4; + EMIT2_off32(0x0F, X86_JG + 0x10, 0); + } else { + EMIT2(X86_JG, 0); + } + jg_reloc = prog; + + err = emit_bpf_dispatcher(&prog, a, a + pivot, /* emit lower_part */ + progs); + if (err) + return err; + + /* From Intel 64 and IA-32 Architectures Optimization + * Reference Manual, 3.4.1.4 Code Alignment, Assembly/Compiler + * Coding Rule 11: All branch targets should be 16-byte + * aligned. + */ + jg_target = PTR_ALIGN(prog, 16); + if (jg_target != prog) + emit_nops(&prog, jg_target - prog); + jg_offset = prog - jg_reloc; + emit_code(jg_reloc - jg_bytes, jg_offset, jg_bytes); + + err = emit_bpf_dispatcher(&prog, a + pivot + 1, /* emit upper_part */ + b, progs); + if (err) + return err; + + *pprog = prog; return 0; } +static int cmp_ips(const void *a, const void *b) +{ + const s64 *ipa = a; + const s64 *ipb = b; + + if (*ipa > *ipb) + return 1; + if (*ipa < *ipb) + return -1; + return 0; +} + +int arch_prepare_bpf_dispatcher(void *image, s64 *funcs, int num_funcs) +{ + u8 *prog = image; + + sort(funcs, num_funcs, sizeof(funcs[0]), cmp_ips, NULL); + return emit_bpf_dispatcher(&prog, 0, num_funcs - 1, funcs); +} + struct x64_jit_data { struct bpf_binary_header *header; int *addrs; diff --git a/arch/x86/pci/i386.c b/arch/x86/pci/i386.c index 9df652d3d927..fa855bbaebaf 100644 --- a/arch/x86/pci/i386.c +++ b/arch/x86/pci/i386.c @@ -34,7 +34,7 @@ #include <linux/errno.h> #include <linux/memblock.h> -#include <asm/pat.h> +#include <asm/memtype.h> #include <asm/e820/api.h> #include <asm/pci_x86.h> #include <asm/io_apic.h> diff --git a/arch/x86/pci/mmconfig_64.c b/arch/x86/pci/mmconfig_64.c index 887d181b769b..0c7b6e66c644 100644 --- a/arch/x86/pci/mmconfig_64.c +++ b/arch/x86/pci/mmconfig_64.c @@ -105,7 +105,7 @@ static void __iomem *mcfg_ioremap(struct pci_mmcfg_region *cfg) start = cfg->address + PCI_MMCFG_BUS_OFFSET(cfg->start_bus); num_buses = cfg->end_bus - cfg->start_bus + 1; size = PCI_MMCFG_BUS_OFFSET(num_buses); - addr = ioremap_nocache(start, size); + addr = ioremap(start, size); if (addr) addr -= PCI_MMCFG_BUS_OFFSET(cfg->start_bus); return addr; diff --git a/arch/x86/platform/efi/Makefile b/arch/x86/platform/efi/Makefile index fe29f3f5d384..84b09c230cbd 100644 --- a/arch/x86/platform/efi/Makefile +++ b/arch/x86/platform/efi/Makefile @@ -1,6 +1,7 @@ # SPDX-License-Identifier: GPL-2.0 OBJECT_FILES_NON_STANDARD_efi_thunk_$(BITS).o := y -OBJECT_FILES_NON_STANDARD_efi_stub_$(BITS).o := y +KASAN_SANITIZE := n +GCOV_PROFILE := n obj-$(CONFIG_EFI) += quirks.o efi.o efi_$(BITS).o efi_stub_$(BITS).o obj-$(CONFIG_EFI_MIXED) += efi_thunk_$(BITS).o diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c index 38d44f36d5ed..59f7f6d60cf6 100644 --- a/arch/x86/platform/efi/efi.c +++ b/arch/x86/platform/efi/efi.c @@ -54,8 +54,8 @@ #include <asm/x86_init.h> #include <asm/uv/uv.h> -static struct efi efi_phys __initdata; static efi_system_table_t efi_systab __initdata; +static u64 efi_systab_phys __initdata; static efi_config_table_type_t arch_tables[] __initdata = { #ifdef CONFIG_X86_UV @@ -97,32 +97,6 @@ static int __init setup_add_efi_memmap(char *arg) } early_param("add_efi_memmap", setup_add_efi_memmap); -static efi_status_t __init phys_efi_set_virtual_address_map( - unsigned long memory_map_size, - unsigned long descriptor_size, - u32 descriptor_version, - efi_memory_desc_t *virtual_map) -{ - efi_status_t status; - unsigned long flags; - pgd_t *save_pgd; - - save_pgd = efi_call_phys_prolog(); - if (!save_pgd) - return EFI_ABORTED; - - /* Disable interrupts around EFI calls: */ - local_irq_save(flags); - status = efi_call_phys(efi_phys.set_virtual_address_map, - memory_map_size, descriptor_size, - descriptor_version, virtual_map); - local_irq_restore(flags); - - efi_call_phys_epilog(save_pgd); - - return status; -} - void __init efi_find_mirror(void) { efi_memory_desc_t *md; @@ -330,10 +304,16 @@ static void __init efi_clean_memmap(void) } if (n_removal > 0) { - u64 size = efi.memmap.nr_map - n_removal; + struct efi_memory_map_data data = { + .phys_map = efi.memmap.phys_map, + .desc_version = efi.memmap.desc_version, + .desc_size = efi.memmap.desc_size, + .size = data.desc_size * (efi.memmap.nr_map - n_removal), + .flags = 0, + }; pr_warn("Removing %d invalid memory map entries.\n", n_removal); - efi_memmap_install(efi.memmap.phys_map, size); + efi_memmap_install(&data); } } @@ -353,89 +333,90 @@ void __init efi_print_memmap(void) } } -static int __init efi_systab_init(void *phys) +static int __init efi_systab_init(u64 phys) { + int size = efi_enabled(EFI_64BIT) ? sizeof(efi_system_table_64_t) + : sizeof(efi_system_table_32_t); + bool over4g = false; + void *p; + + p = early_memremap_ro(phys, size); + if (p == NULL) { + pr_err("Couldn't map the system table!\n"); + return -ENOMEM; + } + if (efi_enabled(EFI_64BIT)) { - efi_system_table_64_t *systab64; - struct efi_setup_data *data = NULL; - u64 tmp = 0; + const efi_system_table_64_t *systab64 = p; + + efi_systab.hdr = systab64->hdr; + efi_systab.fw_vendor = systab64->fw_vendor; + efi_systab.fw_revision = systab64->fw_revision; + efi_systab.con_in_handle = systab64->con_in_handle; + efi_systab.con_in = systab64->con_in; + efi_systab.con_out_handle = systab64->con_out_handle; + efi_systab.con_out = (void *)(unsigned long)systab64->con_out; + efi_systab.stderr_handle = systab64->stderr_handle; + efi_systab.stderr = systab64->stderr; + efi_systab.runtime = (void *)(unsigned long)systab64->runtime; + efi_systab.boottime = (void *)(unsigned long)systab64->boottime; + efi_systab.nr_tables = systab64->nr_tables; + efi_systab.tables = systab64->tables; + + over4g = systab64->con_in_handle > U32_MAX || + systab64->con_in > U32_MAX || + systab64->con_out_handle > U32_MAX || + systab64->con_out > U32_MAX || + systab64->stderr_handle > U32_MAX || + systab64->stderr > U32_MAX || + systab64->boottime > U32_MAX; if (efi_setup) { - data = early_memremap(efi_setup, sizeof(*data)); - if (!data) + struct efi_setup_data *data; + + data = early_memremap_ro(efi_setup, sizeof(*data)); + if (!data) { + early_memunmap(p, size); return -ENOMEM; - } - systab64 = early_memremap((unsigned long)phys, - sizeof(*systab64)); - if (systab64 == NULL) { - pr_err("Couldn't map the system table!\n"); - if (data) - early_memunmap(data, sizeof(*data)); - return -ENOMEM; - } + } + + efi_systab.fw_vendor = (unsigned long)data->fw_vendor; + efi_systab.runtime = (void *)(unsigned long)data->runtime; + efi_systab.tables = (unsigned long)data->tables; + + over4g |= data->fw_vendor > U32_MAX || + data->runtime > U32_MAX || + data->tables > U32_MAX; - efi_systab.hdr = systab64->hdr; - efi_systab.fw_vendor = data ? (unsigned long)data->fw_vendor : - systab64->fw_vendor; - tmp |= data ? data->fw_vendor : systab64->fw_vendor; - efi_systab.fw_revision = systab64->fw_revision; - efi_systab.con_in_handle = systab64->con_in_handle; - tmp |= systab64->con_in_handle; - efi_systab.con_in = systab64->con_in; - tmp |= systab64->con_in; - efi_systab.con_out_handle = systab64->con_out_handle; - tmp |= systab64->con_out_handle; - efi_systab.con_out = systab64->con_out; - tmp |= systab64->con_out; - efi_systab.stderr_handle = systab64->stderr_handle; - tmp |= systab64->stderr_handle; - efi_systab.stderr = systab64->stderr; - tmp |= systab64->stderr; - efi_systab.runtime = data ? - (void *)(unsigned long)data->runtime : - (void *)(unsigned long)systab64->runtime; - tmp |= data ? data->runtime : systab64->runtime; - efi_systab.boottime = (void *)(unsigned long)systab64->boottime; - tmp |= systab64->boottime; - efi_systab.nr_tables = systab64->nr_tables; - efi_systab.tables = data ? (unsigned long)data->tables : - systab64->tables; - tmp |= data ? data->tables : systab64->tables; - - early_memunmap(systab64, sizeof(*systab64)); - if (data) early_memunmap(data, sizeof(*data)); -#ifdef CONFIG_X86_32 - if (tmp >> 32) { - pr_err("EFI data located above 4GB, disabling EFI.\n"); - return -EINVAL; + } else { + over4g |= systab64->fw_vendor > U32_MAX || + systab64->runtime > U32_MAX || + systab64->tables > U32_MAX; } -#endif } else { - efi_system_table_32_t *systab32; - - systab32 = early_memremap((unsigned long)phys, - sizeof(*systab32)); - if (systab32 == NULL) { - pr_err("Couldn't map the system table!\n"); - return -ENOMEM; - } - - efi_systab.hdr = systab32->hdr; - efi_systab.fw_vendor = systab32->fw_vendor; - efi_systab.fw_revision = systab32->fw_revision; - efi_systab.con_in_handle = systab32->con_in_handle; - efi_systab.con_in = systab32->con_in; - efi_systab.con_out_handle = systab32->con_out_handle; - efi_systab.con_out = systab32->con_out; - efi_systab.stderr_handle = systab32->stderr_handle; - efi_systab.stderr = systab32->stderr; - efi_systab.runtime = (void *)(unsigned long)systab32->runtime; - efi_systab.boottime = (void *)(unsigned long)systab32->boottime; - efi_systab.nr_tables = systab32->nr_tables; - efi_systab.tables = systab32->tables; - - early_memunmap(systab32, sizeof(*systab32)); + const efi_system_table_32_t *systab32 = p; + + efi_systab.hdr = systab32->hdr; + efi_systab.fw_vendor = systab32->fw_vendor; + efi_systab.fw_revision = systab32->fw_revision; + efi_systab.con_in_handle = systab32->con_in_handle; + efi_systab.con_in = systab32->con_in; + efi_systab.con_out_handle = systab32->con_out_handle; + efi_systab.con_out = (void *)(unsigned long)systab32->con_out; + efi_systab.stderr_handle = systab32->stderr_handle; + efi_systab.stderr = systab32->stderr; + efi_systab.runtime = (void *)(unsigned long)systab32->runtime; + efi_systab.boottime = (void *)(unsigned long)systab32->boottime; + efi_systab.nr_tables = systab32->nr_tables; + efi_systab.tables = systab32->tables; + } + + early_memunmap(p, size); + + if (IS_ENABLED(CONFIG_X86_32) && over4g) { + pr_err("EFI data located above 4GB, disabling EFI.\n"); + return -EINVAL; } efi.systab = &efi_systab; @@ -455,108 +436,23 @@ static int __init efi_systab_init(void *phys) return 0; } -static int __init efi_runtime_init32(void) -{ - efi_runtime_services_32_t *runtime; - - runtime = early_memremap((unsigned long)efi.systab->runtime, - sizeof(efi_runtime_services_32_t)); - if (!runtime) { - pr_err("Could not map the runtime service table!\n"); - return -ENOMEM; - } - - /* - * We will only need *early* access to the SetVirtualAddressMap - * EFI runtime service. All other runtime services will be called - * via the virtual mapping. - */ - efi_phys.set_virtual_address_map = - (efi_set_virtual_address_map_t *) - (unsigned long)runtime->set_virtual_address_map; - early_memunmap(runtime, sizeof(efi_runtime_services_32_t)); - - return 0; -} - -static int __init efi_runtime_init64(void) -{ - efi_runtime_services_64_t *runtime; - - runtime = early_memremap((unsigned long)efi.systab->runtime, - sizeof(efi_runtime_services_64_t)); - if (!runtime) { - pr_err("Could not map the runtime service table!\n"); - return -ENOMEM; - } - - /* - * We will only need *early* access to the SetVirtualAddressMap - * EFI runtime service. All other runtime services will be called - * via the virtual mapping. - */ - efi_phys.set_virtual_address_map = - (efi_set_virtual_address_map_t *) - (unsigned long)runtime->set_virtual_address_map; - early_memunmap(runtime, sizeof(efi_runtime_services_64_t)); - - return 0; -} - -static int __init efi_runtime_init(void) -{ - int rv; - - /* - * Check out the runtime services table. We need to map - * the runtime services table so that we can grab the physical - * address of several of the EFI runtime functions, needed to - * set the firmware into virtual mode. - * - * When EFI_PARAVIRT is in force then we could not map runtime - * service memory region because we do not have direct access to it. - * However, runtime services are available through proxy functions - * (e.g. in case of Xen dom0 EFI implementation they call special - * hypercall which executes relevant EFI functions) and that is why - * they are always enabled. - */ - - if (!efi_enabled(EFI_PARAVIRT)) { - if (efi_enabled(EFI_64BIT)) - rv = efi_runtime_init64(); - else - rv = efi_runtime_init32(); - - if (rv) - return rv; - } - - set_bit(EFI_RUNTIME_SERVICES, &efi.flags); - - return 0; -} - void __init efi_init(void) { efi_char16_t *c16; char vendor[100] = "unknown"; int i = 0; - void *tmp; -#ifdef CONFIG_X86_32 - if (boot_params.efi_info.efi_systab_hi || - boot_params.efi_info.efi_memmap_hi) { + if (IS_ENABLED(CONFIG_X86_32) && + (boot_params.efi_info.efi_systab_hi || + boot_params.efi_info.efi_memmap_hi)) { pr_info("Table located above 4GB, disabling EFI.\n"); return; } - efi_phys.systab = (efi_system_table_t *)boot_params.efi_info.efi_systab; -#else - efi_phys.systab = (efi_system_table_t *) - (boot_params.efi_info.efi_systab | - ((__u64)boot_params.efi_info.efi_systab_hi<<32)); -#endif - if (efi_systab_init(efi_phys.systab)) + efi_systab_phys = boot_params.efi_info.efi_systab | + ((__u64)boot_params.efi_info.efi_systab_hi << 32); + + if (efi_systab_init(efi_systab_phys)) return; efi.config_table = (unsigned long)efi.systab->tables; @@ -566,14 +462,16 @@ void __init efi_init(void) /* * Show what we know for posterity */ - c16 = tmp = early_memremap(efi.systab->fw_vendor, 2); + c16 = early_memremap_ro(efi.systab->fw_vendor, + sizeof(vendor) * sizeof(efi_char16_t)); if (c16) { - for (i = 0; i < sizeof(vendor) - 1 && *c16; ++i) - vendor[i] = *c16++; + for (i = 0; i < sizeof(vendor) - 1 && c16[i]; ++i) + vendor[i] = c16[i]; vendor[i] = '\0'; - } else + early_memunmap(c16, sizeof(vendor) * sizeof(efi_char16_t)); + } else { pr_err("Could not map the firmware vendor!\n"); - early_memunmap(tmp, 2); + } pr_info("EFI v%u.%.02u by %s\n", efi.systab->hdr.revision >> 16, @@ -592,19 +490,21 @@ void __init efi_init(void) if (!efi_runtime_supported()) pr_info("No EFI runtime due to 32/64-bit mismatch with kernel\n"); - else { - if (efi_runtime_disabled() || efi_runtime_init()) { - efi_memmap_unmap(); - return; - } + + if (!efi_runtime_supported() || efi_runtime_disabled()) { + efi_memmap_unmap(); + return; } + set_bit(EFI_RUNTIME_SERVICES, &efi.flags); efi_clean_memmap(); if (efi_enabled(EFI_DBG)) efi_print_memmap(); } +#if defined(CONFIG_X86_32) || defined(CONFIG_X86_UV) + void __init efi_set_executable(efi_memory_desc_t *md, bool executable) { u64 addr, npages; @@ -669,6 +569,8 @@ void __init old_map_region(efi_memory_desc_t *md) (unsigned long long)md->phys_addr); } +#endif + /* Merge contiguous regions of the same type and attribute */ static void __init efi_merge_regions(void) { @@ -707,7 +609,7 @@ static void __init get_systab_virt_addr(efi_memory_desc_t *md) size = md->num_pages << EFI_PAGE_SHIFT; end = md->phys_addr + size; - systab = (u64)(unsigned long)efi_phys.systab; + systab = efi_systab_phys; if (md->phys_addr <= systab && systab < end) { systab += md->virt_addr - md->phys_addr; efi.systab = (efi_system_table_t *)(unsigned long)systab; @@ -767,7 +669,7 @@ static inline void *efi_map_next_entry_reverse(void *entry) */ static void *efi_map_next_entry(void *entry) { - if (!efi_enabled(EFI_OLD_MEMMAP) && efi_enabled(EFI_64BIT)) { + if (!efi_have_uv1_memmap() && efi_enabled(EFI_64BIT)) { /* * Starting in UEFI v2.5 the EFI_PROPERTIES_TABLE * config table feature requires us to map all entries @@ -828,7 +730,7 @@ static bool should_map_region(efi_memory_desc_t *md) * Map all of RAM so that we can access arguments in the 1:1 * mapping when making EFI runtime calls. */ - if (IS_ENABLED(CONFIG_EFI_MIXED) && !efi_is_native()) { + if (efi_is_mixed()) { if (md->type == EFI_CONVENTIONAL_MEMORY || md->type == EFI_LOADER_DATA || md->type == EFI_LOADER_CODE) @@ -899,11 +801,11 @@ static void __init kexec_enter_virtual_mode(void) /* * We don't do virtual mode, since we don't do runtime services, on - * non-native EFI. With efi=old_map, we don't do runtime services in + * non-native EFI. With the UV1 memmap, we don't do runtime services in * kexec kernel because in the initial boot something else might * have been mapped at these virtual addresses. */ - if (!efi_is_native() || efi_enabled(EFI_OLD_MEMMAP)) { + if (efi_is_mixed() || efi_have_uv1_memmap()) { efi_memmap_unmap(); clear_bit(EFI_RUNTIME_SERVICES, &efi.flags); return; @@ -958,11 +860,6 @@ static void __init kexec_enter_virtual_mode(void) efi.runtime_version = efi_systab.hdr.revision; efi_native_runtime_setup(); - - efi.set_virtual_address_map = NULL; - - if (efi_enabled(EFI_OLD_MEMMAP) && (__supported_pte_mask & _PAGE_NX)) - runtime_code_page_mkexec(); #endif } @@ -974,9 +871,9 @@ static void __init kexec_enter_virtual_mode(void) * * The old method which used to update that memory descriptor with the * virtual address obtained from ioremap() is still supported when the - * kernel is booted with efi=old_map on its command line. Same old - * method enabled the runtime services to be called without having to - * thunk back into physical mode for every invocation. + * kernel is booted on SG1 UV1 hardware. Same old method enabled the + * runtime services to be called without having to thunk back into + * physical mode for every invocation. * * The new method does a pagetable switch in a preemption-safe manner * so that we're in a different address space when calling a runtime @@ -999,16 +896,14 @@ static void __init __efi_enter_virtual_mode(void) if (efi_alloc_page_tables()) { pr_err("Failed to allocate EFI page tables\n"); - clear_bit(EFI_RUNTIME_SERVICES, &efi.flags); - return; + goto err; } efi_merge_regions(); new_memmap = efi_map_regions(&count, &pg_shift); if (!new_memmap) { pr_err("Error reallocating memory, EFI runtime non-functional!\n"); - clear_bit(EFI_RUNTIME_SERVICES, &efi.flags); - return; + goto err; } pa = __pa(new_memmap); @@ -1022,8 +917,7 @@ static void __init __efi_enter_virtual_mode(void) if (efi_memmap_init_late(pa, efi.memmap.desc_size * count)) { pr_err("Failed to remap late EFI memory map\n"); - clear_bit(EFI_RUNTIME_SERVICES, &efi.flags); - return; + goto err; } if (efi_enabled(EFI_DBG)) { @@ -1031,34 +925,22 @@ static void __init __efi_enter_virtual_mode(void) efi_print_memmap(); } - BUG_ON(!efi.systab); + if (WARN_ON(!efi.systab)) + goto err; - if (efi_setup_page_tables(pa, 1 << pg_shift)) { - clear_bit(EFI_RUNTIME_SERVICES, &efi.flags); - return; - } + if (efi_setup_page_tables(pa, 1 << pg_shift)) + goto err; efi_sync_low_kernel_mappings(); - if (efi_is_native()) { - status = phys_efi_set_virtual_address_map( - efi.memmap.desc_size * count, - efi.memmap.desc_size, - efi.memmap.desc_version, - (efi_memory_desc_t *)pa); - } else { - status = efi_thunk_set_virtual_address_map( - efi_phys.set_virtual_address_map, - efi.memmap.desc_size * count, - efi.memmap.desc_size, - efi.memmap.desc_version, - (efi_memory_desc_t *)pa); - } - + status = efi_set_virtual_address_map(efi.memmap.desc_size * count, + efi.memmap.desc_size, + efi.memmap.desc_version, + (efi_memory_desc_t *)pa); if (status != EFI_SUCCESS) { - pr_alert("Unable to switch EFI into virtual mode (status=%lx)!\n", - status); - panic("EFI call to SetVirtualAddressMap() failed!"); + pr_err("Unable to switch EFI into virtual mode (status=%lx)!\n", + status); + goto err; } efi_free_boot_services(); @@ -1071,13 +953,11 @@ static void __init __efi_enter_virtual_mode(void) */ efi.runtime_version = efi_systab.hdr.revision; - if (efi_is_native()) + if (!efi_is_mixed()) efi_native_runtime_setup(); else efi_thunk_runtime_setup(); - efi.set_virtual_address_map = NULL; - /* * Apply more restrictive page table mapping attributes now that * SVAM() has been called and the firmware has performed all @@ -1087,6 +967,10 @@ static void __init __efi_enter_virtual_mode(void) /* clean DUMMY object */ efi_delete_dummy_variable(); + return; + +err: + clear_bit(EFI_RUNTIME_SERVICES, &efi.flags); } void __init efi_enter_virtual_mode(void) @@ -1102,20 +986,6 @@ void __init efi_enter_virtual_mode(void) efi_dump_pagetable(); } -static int __init arch_parse_efi_cmdline(char *str) -{ - if (!str) { - pr_warn("need at least one option\n"); - return -EINVAL; - } - - if (parse_option_str(str, "old_map")) - set_bit(EFI_OLD_MEMMAP, &efi.flags); - - return 0; -} -early_param("efi", arch_parse_efi_cmdline); - bool efi_is_table_address(unsigned long phys_addr) { unsigned int i; diff --git a/arch/x86/platform/efi/efi_32.c b/arch/x86/platform/efi/efi_32.c index 9959657127f4..71dddd1620f9 100644 --- a/arch/x86/platform/efi/efi_32.c +++ b/arch/x86/platform/efi/efi_32.c @@ -66,9 +66,17 @@ void __init efi_map_region(efi_memory_desc_t *md) void __init efi_map_region_fixed(efi_memory_desc_t *md) {} void __init parse_efi_setup(u64 phys_addr, u32 data_len) {} -pgd_t * __init efi_call_phys_prolog(void) +efi_status_t efi_call_svam(efi_set_virtual_address_map_t *__efiapi *, + u32, u32, u32, void *); + +efi_status_t __init efi_set_virtual_address_map(unsigned long memory_map_size, + unsigned long descriptor_size, + u32 descriptor_version, + efi_memory_desc_t *virtual_map) { struct desc_ptr gdt_descr; + efi_status_t status; + unsigned long flags; pgd_t *save_pgd; /* Current pgd is swapper_pg_dir, we'll restore it later: */ @@ -80,14 +88,18 @@ pgd_t * __init efi_call_phys_prolog(void) gdt_descr.size = GDT_SIZE - 1; load_gdt(&gdt_descr); - return save_pgd; -} + /* Disable interrupts around EFI calls: */ + local_irq_save(flags); + status = efi_call_svam(&efi.systab->runtime->set_virtual_address_map, + memory_map_size, descriptor_size, + descriptor_version, virtual_map); + local_irq_restore(flags); -void __init efi_call_phys_epilog(pgd_t *save_pgd) -{ load_fixmap_gdt(0); load_cr3(save_pgd); __flush_tlb_all(); + + return status; } void __init efi_runtime_update_mappings(void) diff --git a/arch/x86/platform/efi/efi_64.c b/arch/x86/platform/efi/efi_64.c index 08ce8177c3af..e2accfe636bd 100644 --- a/arch/x86/platform/efi/efi_64.c +++ b/arch/x86/platform/efi/efi_64.c @@ -57,142 +57,6 @@ static u64 efi_va = EFI_VA_START; struct efi_scratch efi_scratch; -static void __init early_code_mapping_set_exec(int executable) -{ - efi_memory_desc_t *md; - - if (!(__supported_pte_mask & _PAGE_NX)) - return; - - /* Make EFI service code area executable */ - for_each_efi_memory_desc(md) { - if (md->type == EFI_RUNTIME_SERVICES_CODE || - md->type == EFI_BOOT_SERVICES_CODE) - efi_set_executable(md, executable); - } -} - -pgd_t * __init efi_call_phys_prolog(void) -{ - unsigned long vaddr, addr_pgd, addr_p4d, addr_pud; - pgd_t *save_pgd, *pgd_k, *pgd_efi; - p4d_t *p4d, *p4d_k, *p4d_efi; - pud_t *pud; - - int pgd; - int n_pgds, i, j; - - if (!efi_enabled(EFI_OLD_MEMMAP)) { - efi_switch_mm(&efi_mm); - return efi_mm.pgd; - } - - early_code_mapping_set_exec(1); - - n_pgds = DIV_ROUND_UP((max_pfn << PAGE_SHIFT), PGDIR_SIZE); - save_pgd = kmalloc_array(n_pgds, sizeof(*save_pgd), GFP_KERNEL); - if (!save_pgd) - return NULL; - - /* - * Build 1:1 identity mapping for efi=old_map usage. Note that - * PAGE_OFFSET is PGDIR_SIZE aligned when KASLR is disabled, while - * it is PUD_SIZE ALIGNED with KASLR enabled. So for a given physical - * address X, the pud_index(X) != pud_index(__va(X)), we can only copy - * PUD entry of __va(X) to fill in pud entry of X to build 1:1 mapping. - * This means here we can only reuse the PMD tables of the direct mapping. - */ - for (pgd = 0; pgd < n_pgds; pgd++) { - addr_pgd = (unsigned long)(pgd * PGDIR_SIZE); - vaddr = (unsigned long)__va(pgd * PGDIR_SIZE); - pgd_efi = pgd_offset_k(addr_pgd); - save_pgd[pgd] = *pgd_efi; - - p4d = p4d_alloc(&init_mm, pgd_efi, addr_pgd); - if (!p4d) { - pr_err("Failed to allocate p4d table!\n"); - goto out; - } - - for (i = 0; i < PTRS_PER_P4D; i++) { - addr_p4d = addr_pgd + i * P4D_SIZE; - p4d_efi = p4d + p4d_index(addr_p4d); - - pud = pud_alloc(&init_mm, p4d_efi, addr_p4d); - if (!pud) { - pr_err("Failed to allocate pud table!\n"); - goto out; - } - - for (j = 0; j < PTRS_PER_PUD; j++) { - addr_pud = addr_p4d + j * PUD_SIZE; - - if (addr_pud > (max_pfn << PAGE_SHIFT)) - break; - - vaddr = (unsigned long)__va(addr_pud); - - pgd_k = pgd_offset_k(vaddr); - p4d_k = p4d_offset(pgd_k, vaddr); - pud[j] = *pud_offset(p4d_k, vaddr); - } - } - pgd_offset_k(pgd * PGDIR_SIZE)->pgd &= ~_PAGE_NX; - } - - __flush_tlb_all(); - return save_pgd; -out: - efi_call_phys_epilog(save_pgd); - return NULL; -} - -void __init efi_call_phys_epilog(pgd_t *save_pgd) -{ - /* - * After the lock is released, the original page table is restored. - */ - int pgd_idx, i; - int nr_pgds; - pgd_t *pgd; - p4d_t *p4d; - pud_t *pud; - - if (!efi_enabled(EFI_OLD_MEMMAP)) { - efi_switch_mm(efi_scratch.prev_mm); - return; - } - - nr_pgds = DIV_ROUND_UP((max_pfn << PAGE_SHIFT) , PGDIR_SIZE); - - for (pgd_idx = 0; pgd_idx < nr_pgds; pgd_idx++) { - pgd = pgd_offset_k(pgd_idx * PGDIR_SIZE); - set_pgd(pgd_offset_k(pgd_idx * PGDIR_SIZE), save_pgd[pgd_idx]); - - if (!pgd_present(*pgd)) - continue; - - for (i = 0; i < PTRS_PER_P4D; i++) { - p4d = p4d_offset(pgd, - pgd_idx * PGDIR_SIZE + i * P4D_SIZE); - - if (!p4d_present(*p4d)) - continue; - - pud = (pud_t *)p4d_page_vaddr(*p4d); - pud_free(&init_mm, pud); - } - - p4d = (p4d_t *)pgd_page_vaddr(*pgd); - p4d_free(&init_mm, p4d); - } - - kfree(save_pgd); - - __flush_tlb_all(); - early_code_mapping_set_exec(0); -} - EXPORT_SYMBOL_GPL(efi_mm); /* @@ -211,7 +75,7 @@ int __init efi_alloc_page_tables(void) pud_t *pud; gfp_t gfp_mask; - if (efi_enabled(EFI_OLD_MEMMAP)) + if (efi_have_uv1_memmap()) return 0; gfp_mask = GFP_KERNEL | __GFP_ZERO; @@ -252,7 +116,7 @@ void efi_sync_low_kernel_mappings(void) pud_t *pud_k, *pud_efi; pgd_t *efi_pgd = efi_mm.pgd; - if (efi_enabled(EFI_OLD_MEMMAP)) + if (efi_have_uv1_memmap()) return; /* @@ -346,7 +210,7 @@ int __init efi_setup_page_tables(unsigned long pa_memmap, unsigned num_pages) unsigned npages; pgd_t *pgd = efi_mm.pgd; - if (efi_enabled(EFI_OLD_MEMMAP)) + if (efi_have_uv1_memmap()) return 0; /* @@ -373,10 +237,6 @@ int __init efi_setup_page_tables(unsigned long pa_memmap, unsigned num_pages) * as trim_bios_range() will reserve the first page and isolate it away * from memory allocators anyway. */ - pf = _PAGE_RW; - if (sev_active()) - pf |= _PAGE_ENC; - if (kernel_map_pages_in_pgd(pgd, 0x0, 0x0, 1, pf)) { pr_err("Failed to create 1:1 mapping for the first page!\n"); return 1; @@ -388,21 +248,22 @@ int __init efi_setup_page_tables(unsigned long pa_memmap, unsigned num_pages) * text and allocate a new stack because we can't rely on the * stack pointer being < 4GB. */ - if (!IS_ENABLED(CONFIG_EFI_MIXED) || efi_is_native()) + if (!efi_is_mixed()) return 0; page = alloc_page(GFP_KERNEL|__GFP_DMA32); - if (!page) - panic("Unable to allocate EFI runtime stack < 4GB\n"); + if (!page) { + pr_err("Unable to allocate EFI runtime stack < 4GB\n"); + return 1; + } - efi_scratch.phys_stack = virt_to_phys(page_address(page)); - efi_scratch.phys_stack += PAGE_SIZE; /* stack grows down */ + efi_scratch.phys_stack = page_to_phys(page + 1); /* stack grows down */ - npages = (_etext - _text) >> PAGE_SHIFT; + npages = (__end_rodata_aligned - _text) >> PAGE_SHIFT; text = __pa(_text); pfn = text >> PAGE_SHIFT; - pf = _PAGE_RW | _PAGE_ENC; + pf = _PAGE_ENC; if (kernel_map_pages_in_pgd(pgd, pfn, text, npages, pf)) { pr_err("Failed to map kernel text 1:1\n"); return 1; @@ -417,6 +278,22 @@ static void __init __map_region(efi_memory_desc_t *md, u64 va) unsigned long pfn; pgd_t *pgd = efi_mm.pgd; + /* + * EFI_RUNTIME_SERVICES_CODE regions typically cover PE/COFF + * executable images in memory that consist of both R-X and + * RW- sections, so we cannot apply read-only or non-exec + * permissions just yet. However, modern EFI systems provide + * a memory attributes table that describes those sections + * with the appropriate restricted permissions, which are + * applied in efi_runtime_update_mappings() below. All other + * regions can be mapped non-executable at this point, with + * the exception of boot services code regions, but those will + * be unmapped again entirely in efi_free_boot_services(). + */ + if (md->type != EFI_BOOT_SERVICES_CODE && + md->type != EFI_RUNTIME_SERVICES_CODE) + flags |= _PAGE_NX; + if (!(md->attribute & EFI_MEMORY_WB)) flags |= _PAGE_PCD; @@ -434,7 +311,7 @@ void __init efi_map_region(efi_memory_desc_t *md) unsigned long size = md->num_pages << PAGE_SHIFT; u64 pa = md->phys_addr; - if (efi_enabled(EFI_OLD_MEMMAP)) + if (efi_have_uv1_memmap()) return old_map_region(md); /* @@ -449,7 +326,7 @@ void __init efi_map_region(efi_memory_desc_t *md) * booting in EFI mixed mode, because even though we may be * running a 64-bit kernel, the firmware may only be 32-bit. */ - if (!efi_is_native () && IS_ENABLED(CONFIG_EFI_MIXED)) { + if (efi_is_mixed()) { md->virt_addr = md->phys_addr; return; } @@ -491,26 +368,6 @@ void __init efi_map_region_fixed(efi_memory_desc_t *md) __map_region(md, md->virt_addr); } -void __iomem *__init efi_ioremap(unsigned long phys_addr, unsigned long size, - u32 type, u64 attribute) -{ - unsigned long last_map_pfn; - - if (type == EFI_MEMORY_MAPPED_IO) - return ioremap(phys_addr, size); - - last_map_pfn = init_memory_mapping(phys_addr, phys_addr + size); - if ((last_map_pfn << PAGE_SHIFT) < phys_addr + size) { - unsigned long top = last_map_pfn << PAGE_SHIFT; - efi_ioremap(top, size - (top - phys_addr), type, attribute); - } - - if (!(attribute & EFI_MEMORY_WB)) - efi_memory_uc((u64)(unsigned long)__va(phys_addr), size); - - return (void __iomem *)__va(phys_addr); -} - void __init parse_efi_setup(u64 phys_addr, u32 data_len) { efi_setup = phys_addr + sizeof(struct setup_data); @@ -559,7 +416,7 @@ void __init efi_runtime_update_mappings(void) { efi_memory_desc_t *md; - if (efi_enabled(EFI_OLD_MEMMAP)) { + if (efi_have_uv1_memmap()) { if (__supported_pte_mask & _PAGE_NX) runtime_code_page_mkexec(); return; @@ -613,7 +470,7 @@ void __init efi_runtime_update_mappings(void) void __init efi_dump_pagetable(void) { #ifdef CONFIG_EFI_PGT_DUMP - if (efi_enabled(EFI_OLD_MEMMAP)) + if (efi_have_uv1_memmap()) ptdump_walk_pgd_level(NULL, swapper_pg_dir); else ptdump_walk_pgd_level(NULL, efi_mm.pgd); @@ -634,63 +491,74 @@ void efi_switch_mm(struct mm_struct *mm) switch_mm(efi_scratch.prev_mm, mm, NULL); } -#ifdef CONFIG_EFI_MIXED -extern efi_status_t efi64_thunk(u32, ...); - static DEFINE_SPINLOCK(efi_runtime_lock); -#define runtime_service32(func) \ -({ \ - u32 table = (u32)(unsigned long)efi.systab; \ - u32 *rt, *___f; \ - \ - rt = (u32 *)(table + offsetof(efi_system_table_32_t, runtime)); \ - ___f = (u32 *)(*rt + offsetof(efi_runtime_services_32_t, func)); \ - *___f; \ +/* + * DS and ES contain user values. We need to save them. + * The 32-bit EFI code needs a valid DS, ES, and SS. There's no + * need to save the old SS: __KERNEL_DS is always acceptable. + */ +#define __efi_thunk(func, ...) \ +({ \ + efi_runtime_services_32_t *__rt; \ + unsigned short __ds, __es; \ + efi_status_t ____s; \ + \ + __rt = (void *)(unsigned long)efi.systab->mixed_mode.runtime; \ + \ + savesegment(ds, __ds); \ + savesegment(es, __es); \ + \ + loadsegment(ss, __KERNEL_DS); \ + loadsegment(ds, __KERNEL_DS); \ + loadsegment(es, __KERNEL_DS); \ + \ + ____s = efi64_thunk(__rt->func, __VA_ARGS__); \ + \ + loadsegment(ds, __ds); \ + loadsegment(es, __es); \ + \ + ____s ^= (____s & BIT(31)) | (____s & BIT_ULL(31)) << 32; \ + ____s; \ }) /* * Switch to the EFI page tables early so that we can access the 1:1 * runtime services mappings which are not mapped in any other page - * tables. This function must be called before runtime_service32(). + * tables. * * Also, disable interrupts because the IDT points to 64-bit handlers, * which aren't going to function correctly when we switch to 32-bit. */ -#define efi_thunk(f, ...) \ +#define efi_thunk(func...) \ ({ \ efi_status_t __s; \ - u32 __func; \ \ arch_efi_call_virt_setup(); \ \ - __func = runtime_service32(f); \ - __s = efi64_thunk(__func, __VA_ARGS__); \ + __s = __efi_thunk(func); \ \ arch_efi_call_virt_teardown(); \ \ __s; \ }) -efi_status_t efi_thunk_set_virtual_address_map( - void *phys_set_virtual_address_map, - unsigned long memory_map_size, - unsigned long descriptor_size, - u32 descriptor_version, - efi_memory_desc_t *virtual_map) +static efi_status_t __init __no_sanitize_address +efi_thunk_set_virtual_address_map(unsigned long memory_map_size, + unsigned long descriptor_size, + u32 descriptor_version, + efi_memory_desc_t *virtual_map) { efi_status_t status; unsigned long flags; - u32 func; efi_sync_low_kernel_mappings(); local_irq_save(flags); efi_switch_mm(&efi_mm); - func = (u32)(unsigned long)phys_set_virtual_address_map; - status = efi64_thunk(func, memory_map_size, descriptor_size, - descriptor_version, virtual_map); + status = __efi_thunk(set_virtual_address_map, memory_map_size, + descriptor_size, descriptor_version, virtual_map); efi_switch_mm(efi_scratch.prev_mm); local_irq_restore(flags); @@ -993,8 +861,11 @@ efi_thunk_query_capsule_caps(efi_capsule_header_t **capsules, return EFI_UNSUPPORTED; } -void efi_thunk_runtime_setup(void) +void __init efi_thunk_runtime_setup(void) { + if (!IS_ENABLED(CONFIG_EFI_MIXED)) + return; + efi.get_time = efi_thunk_get_time; efi.set_time = efi_thunk_set_time; efi.get_wakeup_time = efi_thunk_get_wakeup_time; @@ -1010,4 +881,46 @@ void efi_thunk_runtime_setup(void) efi.update_capsule = efi_thunk_update_capsule; efi.query_capsule_caps = efi_thunk_query_capsule_caps; } -#endif /* CONFIG_EFI_MIXED */ + +efi_status_t __init __no_sanitize_address +efi_set_virtual_address_map(unsigned long memory_map_size, + unsigned long descriptor_size, + u32 descriptor_version, + efi_memory_desc_t *virtual_map) +{ + efi_status_t status; + unsigned long flags; + pgd_t *save_pgd = NULL; + + if (efi_is_mixed()) + return efi_thunk_set_virtual_address_map(memory_map_size, + descriptor_size, + descriptor_version, + virtual_map); + + if (efi_have_uv1_memmap()) { + save_pgd = efi_uv1_memmap_phys_prolog(); + if (!save_pgd) + return EFI_ABORTED; + } else { + efi_switch_mm(&efi_mm); + } + + kernel_fpu_begin(); + + /* Disable interrupts around EFI calls: */ + local_irq_save(flags); + status = efi_call(efi.systab->runtime->set_virtual_address_map, + memory_map_size, descriptor_size, + descriptor_version, virtual_map); + local_irq_restore(flags); + + kernel_fpu_end(); + + if (save_pgd) + efi_uv1_memmap_phys_epilog(save_pgd); + else + efi_switch_mm(efi_scratch.prev_mm); + + return status; +} diff --git a/arch/x86/platform/efi/efi_stub_32.S b/arch/x86/platform/efi/efi_stub_32.S index eed8b5b441f8..75c46e7a809f 100644 --- a/arch/x86/platform/efi/efi_stub_32.S +++ b/arch/x86/platform/efi/efi_stub_32.S @@ -7,118 +7,43 @@ */ #include <linux/linkage.h> +#include <linux/init.h> #include <asm/page_types.h> -/* - * efi_call_phys(void *, ...) is a function with variable parameters. - * All the callers of this function assure that all the parameters are 4-bytes. - */ - -/* - * In gcc calling convention, EBX, ESP, EBP, ESI and EDI are all callee save. - * So we'd better save all of them at the beginning of this function and restore - * at the end no matter how many we use, because we can not assure EFI runtime - * service functions will comply with gcc calling convention, too. - */ + __INIT +SYM_FUNC_START(efi_call_svam) + push 8(%esp) + push 8(%esp) + push %ecx + push %edx -.text -SYM_FUNC_START(efi_call_phys) /* - * 0. The function can only be called in Linux kernel. So CS has been - * set to 0x0010, DS and SS have been set to 0x0018. In EFI, I found - * the values of these registers are the same. And, the corresponding - * GDT entries are identical. So I will do nothing about segment reg - * and GDT, but change GDT base register in prolog and epilog. - */ - - /* - * 1. Now I am running with EIP = <physical address> + PAGE_OFFSET. - * But to make it smoothly switch from virtual mode to flat mode. - * The mapping of lower virtual memory has been created in prolog and - * epilog. + * Switch to the flat mapped alias of this routine, by jumping to the + * address of label '1' after subtracting PAGE_OFFSET from it. */ movl $1f, %edx subl $__PAGE_OFFSET, %edx jmp *%edx 1: - /* - * 2. Now on the top of stack is the return - * address in the caller of efi_call_phys(), then parameter 1, - * parameter 2, ..., param n. To make things easy, we save the return - * address of efi_call_phys in a global variable. - */ - popl %edx - movl %edx, saved_return_addr - /* get the function pointer into ECX*/ - popl %ecx - movl %ecx, efi_rt_function_ptr - movl $2f, %edx - subl $__PAGE_OFFSET, %edx - pushl %edx - - /* - * 3. Clear PG bit in %CR0. - */ + /* disable paging */ movl %cr0, %edx andl $0x7fffffff, %edx movl %edx, %cr0 - jmp 1f -1: - /* - * 4. Adjust stack pointer. - */ + /* convert the stack pointer to a flat mapped address */ subl $__PAGE_OFFSET, %esp - /* - * 5. Call the physical function. - */ - jmp *%ecx + /* call the EFI routine */ + call *(%eax) -2: - /* - * 6. After EFI runtime service returns, control will return to - * following instruction. We'd better readjust stack pointer first. - */ - addl $__PAGE_OFFSET, %esp + /* convert ESP back to a kernel VA, and pop the outgoing args */ + addl $__PAGE_OFFSET + 16, %esp - /* - * 7. Restore PG bit - */ + /* re-enable paging */ movl %cr0, %edx orl $0x80000000, %edx movl %edx, %cr0 - jmp 1f -1: - /* - * 8. Now restore the virtual mode from flat mode by - * adding EIP with PAGE_OFFSET. - */ - movl $1f, %edx - jmp *%edx -1: - - /* - * 9. Balance the stack. And because EAX contain the return value, - * we'd better not clobber it. - */ - leal efi_rt_function_ptr, %edx - movl (%edx), %ecx - pushl %ecx - /* - * 10. Push the saved return address onto the stack and return. - */ - leal saved_return_addr, %edx - movl (%edx), %ecx - pushl %ecx ret -SYM_FUNC_END(efi_call_phys) -.previous - -.data -saved_return_addr: - .long 0 -efi_rt_function_ptr: - .long 0 +SYM_FUNC_END(efi_call_svam) diff --git a/arch/x86/platform/efi/efi_stub_64.S b/arch/x86/platform/efi/efi_stub_64.S index b1d2313fe3bf..15da118f04f0 100644 --- a/arch/x86/platform/efi/efi_stub_64.S +++ b/arch/x86/platform/efi/efi_stub_64.S @@ -8,41 +8,12 @@ */ #include <linux/linkage.h> -#include <asm/segment.h> -#include <asm/msr.h> -#include <asm/processor-flags.h> -#include <asm/page_types.h> +#include <asm/nospec-branch.h> -#define SAVE_XMM \ - mov %rsp, %rax; \ - subq $0x70, %rsp; \ - and $~0xf, %rsp; \ - mov %rax, (%rsp); \ - mov %cr0, %rax; \ - clts; \ - mov %rax, 0x8(%rsp); \ - movaps %xmm0, 0x60(%rsp); \ - movaps %xmm1, 0x50(%rsp); \ - movaps %xmm2, 0x40(%rsp); \ - movaps %xmm3, 0x30(%rsp); \ - movaps %xmm4, 0x20(%rsp); \ - movaps %xmm5, 0x10(%rsp) - -#define RESTORE_XMM \ - movaps 0x60(%rsp), %xmm0; \ - movaps 0x50(%rsp), %xmm1; \ - movaps 0x40(%rsp), %xmm2; \ - movaps 0x30(%rsp), %xmm3; \ - movaps 0x20(%rsp), %xmm4; \ - movaps 0x10(%rsp), %xmm5; \ - mov 0x8(%rsp), %rsi; \ - mov %rsi, %cr0; \ - mov (%rsp), %rsp - -SYM_FUNC_START(efi_call) +SYM_FUNC_START(__efi_call) pushq %rbp movq %rsp, %rbp - SAVE_XMM + and $~0xf, %rsp mov 16(%rbp), %rax subq $48, %rsp mov %r9, 32(%rsp) @@ -50,9 +21,7 @@ SYM_FUNC_START(efi_call) mov %r8, %r9 mov %rcx, %r8 mov %rsi, %rcx - call *%rdi - addq $48, %rsp - RESTORE_XMM - popq %rbp + CALL_NOSPEC %rdi + leave ret -SYM_FUNC_END(efi_call) +SYM_FUNC_END(__efi_call) diff --git a/arch/x86/platform/efi/efi_thunk_64.S b/arch/x86/platform/efi/efi_thunk_64.S index 3189f1394701..26f0da238c1c 100644 --- a/arch/x86/platform/efi/efi_thunk_64.S +++ b/arch/x86/platform/efi/efi_thunk_64.S @@ -25,15 +25,16 @@ .text .code64 -SYM_FUNC_START(efi64_thunk) +SYM_CODE_START(__efi64_thunk) push %rbp push %rbx /* * Switch to 1:1 mapped 32-bit stack pointer. */ - movq %rsp, efi_saved_sp(%rip) + movq %rsp, %rax movq efi_scratch(%rip), %rsp + push %rax /* * Calculate the physical address of the kernel text. @@ -41,113 +42,31 @@ SYM_FUNC_START(efi64_thunk) movq $__START_KERNEL_map, %rax subq phys_base(%rip), %rax - /* - * Push some physical addresses onto the stack. This is easier - * to do now in a code64 section while the assembler can address - * 64-bit values. Note that all the addresses on the stack are - * 32-bit. - */ - subq $16, %rsp - leaq efi_exit32(%rip), %rbx - subq %rax, %rbx - movl %ebx, 8(%rsp) - - leaq __efi64_thunk(%rip), %rbx + leaq 1f(%rip), %rbp + leaq 2f(%rip), %rbx + subq %rax, %rbp subq %rax, %rbx - call *%rbx - - movq efi_saved_sp(%rip), %rsp - pop %rbx - pop %rbp - retq -SYM_FUNC_END(efi64_thunk) -/* - * We run this function from the 1:1 mapping. - * - * This function must be invoked with a 1:1 mapped stack. - */ -SYM_FUNC_START_LOCAL(__efi64_thunk) - movl %ds, %eax - push %rax - movl %es, %eax - push %rax - movl %ss, %eax - push %rax - - subq $32, %rsp - movl %esi, 0x0(%rsp) - movl %edx, 0x4(%rsp) - movl %ecx, 0x8(%rsp) - movq %r8, %rsi - movl %esi, 0xc(%rsp) - movq %r9, %rsi - movl %esi, 0x10(%rsp) - - leaq 1f(%rip), %rbx - movq %rbx, func_rt_ptr(%rip) + subq $28, %rsp + movl %ebx, 0x0(%rsp) /* return address */ + movl %esi, 0x4(%rsp) + movl %edx, 0x8(%rsp) + movl %ecx, 0xc(%rsp) + movl %r8d, 0x10(%rsp) + movl %r9d, 0x14(%rsp) /* Switch to 32-bit descriptor */ pushq $__KERNEL32_CS - leaq efi_enter32(%rip), %rax - pushq %rax + pushq %rdi /* EFI runtime service address */ lretq -1: addq $32, %rsp - +1: movq 24(%rsp), %rsp pop %rbx - movl %ebx, %ss - pop %rbx - movl %ebx, %es - pop %rbx - movl %ebx, %ds - - /* - * Convert 32-bit status code into 64-bit. - */ - test %rax, %rax - jz 1f - movl %eax, %ecx - andl $0x0fffffff, %ecx - andl $0xf0000000, %eax - shl $32, %rax - or %rcx, %rax -1: - ret -SYM_FUNC_END(__efi64_thunk) - -SYM_FUNC_START_LOCAL(efi_exit32) - movq func_rt_ptr(%rip), %rax - push %rax - mov %rdi, %rax - ret -SYM_FUNC_END(efi_exit32) + pop %rbp + retq .code32 -/* - * EFI service pointer must be in %edi. - * - * The stack should represent the 32-bit calling convention. - */ -SYM_FUNC_START_LOCAL(efi_enter32) - movl $__KERNEL_DS, %eax - movl %eax, %ds - movl %eax, %es - movl %eax, %ss - - call *%edi - - /* We must preserve return value */ - movl %eax, %edi - - movl 72(%esp), %eax - pushl $__KERNEL_CS - pushl %eax - +2: pushl $__KERNEL_CS + pushl %ebp lret -SYM_FUNC_END(efi_enter32) - - .data - .balign 8 -func_rt_ptr: .quad 0 -efi_saved_sp: .quad 0 +SYM_CODE_END(__efi64_thunk) diff --git a/arch/x86/platform/efi/quirks.c b/arch/x86/platform/efi/quirks.c index f8f0220b6a66..88d32c06cffa 100644 --- a/arch/x86/platform/efi/quirks.c +++ b/arch/x86/platform/efi/quirks.c @@ -16,6 +16,7 @@ #include <asm/efi.h> #include <asm/uv/uv.h> #include <asm/cpu_device_id.h> +#include <asm/realmode.h> #include <asm/reboot.h> #define EFI_MIN_RESERVE 5120 @@ -243,7 +244,7 @@ EXPORT_SYMBOL_GPL(efi_query_variable_store); */ void __init efi_arch_mem_reserve(phys_addr_t addr, u64 size) { - phys_addr_t new_phys, new_size; + struct efi_memory_map_data data = { 0 }; struct efi_mem_range mr; efi_memory_desc_t md; int num_entries; @@ -271,24 +272,21 @@ void __init efi_arch_mem_reserve(phys_addr_t addr, u64 size) num_entries = efi_memmap_split_count(&md, &mr.range); num_entries += efi.memmap.nr_map; - new_size = efi.memmap.desc_size * num_entries; - - new_phys = efi_memmap_alloc(num_entries); - if (!new_phys) { + if (efi_memmap_alloc(num_entries, &data) != 0) { pr_err("Could not allocate boot services memmap\n"); return; } - new = early_memremap(new_phys, new_size); + new = early_memremap(data.phys_map, data.size); if (!new) { pr_err("Failed to map new boot services memmap\n"); return; } efi_memmap_insert(&efi.memmap, new, &mr); - early_memunmap(new, new_size); + early_memunmap(new, data.size); - efi_memmap_install(new_phys, num_entries); + efi_memmap_install(&data); e820__range_update(addr, size, E820_TYPE_RAM, E820_TYPE_RESERVED); e820__update_table(e820_table); } @@ -384,10 +382,10 @@ static void __init efi_unmap_pages(efi_memory_desc_t *md) /* * To Do: Remove this check after adding functionality to unmap EFI boot - * services code/data regions from direct mapping area because - * "efi=old_map" maps EFI regions in swapper_pg_dir. + * services code/data regions from direct mapping area because the UV1 + * memory map maps EFI regions in swapper_pg_dir. */ - if (efi_enabled(EFI_OLD_MEMMAP)) + if (efi_have_uv1_memmap()) return; /* @@ -395,7 +393,7 @@ static void __init efi_unmap_pages(efi_memory_desc_t *md) * EFI runtime calls, hence don't unmap EFI boot services code/data * regions. */ - if (!efi_is_native()) + if (efi_is_mixed()) return; if (kernel_unmap_pages_in_pgd(pgd, pa, md->num_pages)) @@ -407,7 +405,7 @@ static void __init efi_unmap_pages(efi_memory_desc_t *md) void __init efi_free_boot_services(void) { - phys_addr_t new_phys, new_size; + struct efi_memory_map_data data = { 0 }; efi_memory_desc_t *md; int num_entries = 0; void *new, *new_md; @@ -462,14 +460,12 @@ void __init efi_free_boot_services(void) if (!num_entries) return; - new_size = efi.memmap.desc_size * num_entries; - new_phys = efi_memmap_alloc(num_entries); - if (!new_phys) { + if (efi_memmap_alloc(num_entries, &data) != 0) { pr_err("Failed to allocate new EFI memmap\n"); return; } - new = memremap(new_phys, new_size, MEMREMAP_WB); + new = memremap(data.phys_map, data.size, MEMREMAP_WB); if (!new) { pr_err("Failed to map new EFI memmap\n"); return; @@ -493,7 +489,7 @@ void __init efi_free_boot_services(void) memunmap(new); - if (efi_memmap_install(new_phys, num_entries)) { + if (efi_memmap_install(&data) != 0) { pr_err("Could not install new EFI memmap\n"); return; } @@ -558,7 +554,7 @@ out: return ret; } -static const struct dmi_system_id sgi_uv1_dmi[] = { +static const struct dmi_system_id sgi_uv1_dmi[] __initconst = { { NULL, "SGI UV1", { DMI_MATCH(DMI_PRODUCT_NAME, "Stoutland Platform"), DMI_MATCH(DMI_PRODUCT_VERSION, "1.0"), @@ -581,8 +577,15 @@ void __init efi_apply_memmap_quirks(void) } /* UV2+ BIOS has a fix for this issue. UV1 still needs the quirk. */ - if (dmi_check_system(sgi_uv1_dmi)) - set_bit(EFI_OLD_MEMMAP, &efi.flags); + if (dmi_check_system(sgi_uv1_dmi)) { + if (IS_ENABLED(CONFIG_X86_UV)) { + set_bit(EFI_UV1_MEMMAP, &efi.flags); + } else { + pr_warn("EFI runtime disabled, needs CONFIG_X86_UV=y on UV1\n"); + clear_bit(EFI_RUNTIME_SERVICES, &efi.flags); + efi_memmap_unmap(); + } + } } /* @@ -720,7 +723,7 @@ void efi_recover_from_page_fault(unsigned long phys_addr) /* * Make sure that an efi runtime service caused the page fault. * "efi_mm" cannot be used to check if the page fault had occurred - * in the firmware context because efi=old_map doesn't use efi_pgd. + * in the firmware context because the UV1 memmap doesn't use efi_pgd. */ if (efi_rts_work.efi_rts_id == EFI_NONE) return; diff --git a/arch/x86/platform/intel-quark/imr.c b/arch/x86/platform/intel-quark/imr.c index 6dd25dc5f027..e9d97d52475e 100644 --- a/arch/x86/platform/intel-quark/imr.c +++ b/arch/x86/platform/intel-quark/imr.c @@ -29,6 +29,8 @@ #include <asm/cpu_device_id.h> #include <asm/imr.h> #include <asm/iosf_mbi.h> +#include <asm/io.h> + #include <linux/debugfs.h> #include <linux/init.h> #include <linux/mm.h> diff --git a/arch/x86/platform/intel-quark/imr_selftest.c b/arch/x86/platform/intel-quark/imr_selftest.c index 42f879b75f9b..4307830e1b6f 100644 --- a/arch/x86/platform/intel-quark/imr_selftest.c +++ b/arch/x86/platform/intel-quark/imr_selftest.c @@ -14,6 +14,8 @@ #include <asm-generic/sections.h> #include <asm/cpu_device_id.h> #include <asm/imr.h> +#include <asm/io.h> + #include <linux/init.h> #include <linux/mm.h> #include <linux/types.h> diff --git a/arch/x86/platform/uv/bios_uv.c b/arch/x86/platform/uv/bios_uv.c index ece9cb9c1189..607f58147311 100644 --- a/arch/x86/platform/uv/bios_uv.c +++ b/arch/x86/platform/uv/bios_uv.c @@ -31,13 +31,16 @@ static s64 __uv_bios_call(enum uv_bios_cmd which, u64 a1, u64 a2, u64 a3, return BIOS_STATUS_UNIMPLEMENTED; /* - * If EFI_OLD_MEMMAP is set, we need to fall back to using our old EFI + * If EFI_UV1_MEMMAP is set, we need to fall back to using our old EFI * callback method, which uses efi_call() directly, with the kernel page tables: */ - if (unlikely(efi_enabled(EFI_OLD_MEMMAP))) + if (unlikely(efi_enabled(EFI_UV1_MEMMAP))) { + kernel_fpu_begin(); ret = efi_call((void *)__va(tab->function), (u64)which, a1, a2, a3, a4, a5); - else + kernel_fpu_end(); + } else { ret = efi_call_virt_pointer(tab, function, (u64)which, a1, a2, a3, a4, a5); + } return ret; } @@ -214,3 +217,163 @@ int uv_bios_init(void) pr_info("UV: UVsystab: Revision:%x\n", uv_systab->revision); return 0; } + +static void __init early_code_mapping_set_exec(int executable) +{ + efi_memory_desc_t *md; + + if (!(__supported_pte_mask & _PAGE_NX)) + return; + + /* Make EFI service code area executable */ + for_each_efi_memory_desc(md) { + if (md->type == EFI_RUNTIME_SERVICES_CODE || + md->type == EFI_BOOT_SERVICES_CODE) + efi_set_executable(md, executable); + } +} + +void __init efi_uv1_memmap_phys_epilog(pgd_t *save_pgd) +{ + /* + * After the lock is released, the original page table is restored. + */ + int pgd_idx, i; + int nr_pgds; + pgd_t *pgd; + p4d_t *p4d; + pud_t *pud; + + nr_pgds = DIV_ROUND_UP((max_pfn << PAGE_SHIFT) , PGDIR_SIZE); + + for (pgd_idx = 0; pgd_idx < nr_pgds; pgd_idx++) { + pgd = pgd_offset_k(pgd_idx * PGDIR_SIZE); + set_pgd(pgd_offset_k(pgd_idx * PGDIR_SIZE), save_pgd[pgd_idx]); + + if (!pgd_present(*pgd)) + continue; + + for (i = 0; i < PTRS_PER_P4D; i++) { + p4d = p4d_offset(pgd, + pgd_idx * PGDIR_SIZE + i * P4D_SIZE); + + if (!p4d_present(*p4d)) + continue; + + pud = (pud_t *)p4d_page_vaddr(*p4d); + pud_free(&init_mm, pud); + } + + p4d = (p4d_t *)pgd_page_vaddr(*pgd); + p4d_free(&init_mm, p4d); + } + + kfree(save_pgd); + + __flush_tlb_all(); + early_code_mapping_set_exec(0); +} + +pgd_t * __init efi_uv1_memmap_phys_prolog(void) +{ + unsigned long vaddr, addr_pgd, addr_p4d, addr_pud; + pgd_t *save_pgd, *pgd_k, *pgd_efi; + p4d_t *p4d, *p4d_k, *p4d_efi; + pud_t *pud; + + int pgd; + int n_pgds, i, j; + + early_code_mapping_set_exec(1); + + n_pgds = DIV_ROUND_UP((max_pfn << PAGE_SHIFT), PGDIR_SIZE); + save_pgd = kmalloc_array(n_pgds, sizeof(*save_pgd), GFP_KERNEL); + if (!save_pgd) + return NULL; + + /* + * Build 1:1 identity mapping for UV1 memmap usage. Note that + * PAGE_OFFSET is PGDIR_SIZE aligned when KASLR is disabled, while + * it is PUD_SIZE ALIGNED with KASLR enabled. So for a given physical + * address X, the pud_index(X) != pud_index(__va(X)), we can only copy + * PUD entry of __va(X) to fill in pud entry of X to build 1:1 mapping. + * This means here we can only reuse the PMD tables of the direct mapping. + */ + for (pgd = 0; pgd < n_pgds; pgd++) { + addr_pgd = (unsigned long)(pgd * PGDIR_SIZE); + vaddr = (unsigned long)__va(pgd * PGDIR_SIZE); + pgd_efi = pgd_offset_k(addr_pgd); + save_pgd[pgd] = *pgd_efi; + + p4d = p4d_alloc(&init_mm, pgd_efi, addr_pgd); + if (!p4d) { + pr_err("Failed to allocate p4d table!\n"); + goto out; + } + + for (i = 0; i < PTRS_PER_P4D; i++) { + addr_p4d = addr_pgd + i * P4D_SIZE; + p4d_efi = p4d + p4d_index(addr_p4d); + + pud = pud_alloc(&init_mm, p4d_efi, addr_p4d); + if (!pud) { + pr_err("Failed to allocate pud table!\n"); + goto out; + } + + for (j = 0; j < PTRS_PER_PUD; j++) { + addr_pud = addr_p4d + j * PUD_SIZE; + + if (addr_pud > (max_pfn << PAGE_SHIFT)) + break; + + vaddr = (unsigned long)__va(addr_pud); + + pgd_k = pgd_offset_k(vaddr); + p4d_k = p4d_offset(pgd_k, vaddr); + pud[j] = *pud_offset(p4d_k, vaddr); + } + } + pgd_offset_k(pgd * PGDIR_SIZE)->pgd &= ~_PAGE_NX; + } + + __flush_tlb_all(); + return save_pgd; +out: + efi_uv1_memmap_phys_epilog(save_pgd); + return NULL; +} + +void __iomem *__init efi_ioremap(unsigned long phys_addr, unsigned long size, + u32 type, u64 attribute) +{ + unsigned long last_map_pfn; + + if (type == EFI_MEMORY_MAPPED_IO) + return ioremap(phys_addr, size); + + last_map_pfn = init_memory_mapping(phys_addr, phys_addr + size); + if ((last_map_pfn << PAGE_SHIFT) < phys_addr + size) { + unsigned long top = last_map_pfn << PAGE_SHIFT; + efi_ioremap(top, size - (top - phys_addr), type, attribute); + } + + if (!(attribute & EFI_MEMORY_WB)) + efi_memory_uc((u64)(unsigned long)__va(phys_addr), size); + + return (void __iomem *)__va(phys_addr); +} + +static int __init arch_parse_efi_cmdline(char *str) +{ + if (!str) { + pr_warn("need at least one option\n"); + return -EINVAL; + } + + if (!efi_is_mixed() && parse_option_str(str, "old_map")) + set_bit(EFI_UV1_MEMMAP, &efi.flags); + + return 0; +} +early_param("efi", arch_parse_efi_cmdline); diff --git a/arch/x86/xen/Kconfig b/arch/x86/xen/Kconfig index ba5a41828e9d..1aded63a95cb 100644 --- a/arch/x86/xen/Kconfig +++ b/arch/x86/xen/Kconfig @@ -62,10 +62,10 @@ config XEN_512GB boot parameter "xen_512gb_limit". config XEN_SAVE_RESTORE - bool - depends on XEN - select HIBERNATE_CALLBACKS - default y + bool + depends on XEN + select HIBERNATE_CALLBACKS + default y config XEN_DEBUG_FS bool "Enable Xen debug and tuning parameters in debugfs" diff --git a/arch/x86/xen/efi.c b/arch/x86/xen/efi.c index a04551ee5568..1abe455d926a 100644 --- a/arch/x86/xen/efi.c +++ b/arch/x86/xen/efi.c @@ -31,7 +31,7 @@ static efi_system_table_t efi_systab_xen __initdata = { .con_in_handle = EFI_INVALID_TABLE_ADDR, /* Not used under Xen. */ .con_in = EFI_INVALID_TABLE_ADDR, /* Not used under Xen. */ .con_out_handle = EFI_INVALID_TABLE_ADDR, /* Not used under Xen. */ - .con_out = EFI_INVALID_TABLE_ADDR, /* Not used under Xen. */ + .con_out = NULL, /* Not used under Xen. */ .stderr_handle = EFI_INVALID_TABLE_ADDR, /* Not used under Xen. */ .stderr = EFI_INVALID_TABLE_ADDR, /* Not used under Xen. */ .runtime = (efi_runtime_services_t *)EFI_INVALID_TABLE_ADDR, diff --git a/arch/x86/xen/mmu_pv.c b/arch/x86/xen/mmu_pv.c index c8dbee62ec2a..bbba8b17829a 100644 --- a/arch/x86/xen/mmu_pv.c +++ b/arch/x86/xen/mmu_pv.c @@ -67,7 +67,7 @@ #include <asm/linkage.h> #include <asm/page.h> #include <asm/init.h> -#include <asm/pat.h> +#include <asm/memtype.h> #include <asm/smp.h> #include <asm/tlb.h> |