diff options
Diffstat (limited to 'net/sunrpc')
-rw-r--r-- | net/sunrpc/auth_gss/gss_rpc_upcall.c | 29 | ||||
-rw-r--r-- | net/sunrpc/auth_gss/gss_rpc_xdr.c | 50 | ||||
-rw-r--r-- | net/sunrpc/auth_gss/gss_rpc_xdr.h | 5 | ||||
-rw-r--r-- | net/sunrpc/clnt.c | 6 | ||||
-rw-r--r-- | net/sunrpc/netns.h | 1 | ||||
-rw-r--r-- | net/sunrpc/rpcb_clnt.c | 48 | ||||
-rw-r--r-- | net/sunrpc/svcauth_unix.c | 2 | ||||
-rw-r--r-- | net/sunrpc/svcsock.c | 9 | ||||
-rw-r--r-- | net/sunrpc/xdr.c | 9 | ||||
-rw-r--r-- | net/sunrpc/xprtrdma/svc_rdma_marshal.c | 20 | ||||
-rw-r--r-- | net/sunrpc/xprtsock.c | 28 |
11 files changed, 145 insertions, 62 deletions
diff --git a/net/sunrpc/auth_gss/gss_rpc_upcall.c b/net/sunrpc/auth_gss/gss_rpc_upcall.c index d304f41260f..f1eb0d16666 100644 --- a/net/sunrpc/auth_gss/gss_rpc_upcall.c +++ b/net/sunrpc/auth_gss/gss_rpc_upcall.c @@ -120,7 +120,7 @@ static int gssp_rpc_create(struct net *net, struct rpc_clnt **_clnt) if (IS_ERR(clnt)) { dprintk("RPC: failed to create AF_LOCAL gssproxy " "client (errno %ld).\n", PTR_ERR(clnt)); - result = -PTR_ERR(clnt); + result = PTR_ERR(clnt); *_clnt = NULL; goto out; } @@ -213,6 +213,26 @@ static int gssp_call(struct net *net, struct rpc_message *msg) return status; } +static void gssp_free_receive_pages(struct gssx_arg_accept_sec_context *arg) +{ + int i; + + for (i = 0; i < arg->npages && arg->pages[i]; i++) + __free_page(arg->pages[i]); +} + +static int gssp_alloc_receive_pages(struct gssx_arg_accept_sec_context *arg) +{ + arg->npages = DIV_ROUND_UP(NGROUPS_MAX * 4, PAGE_SIZE); + arg->pages = kzalloc(arg->npages * sizeof(struct page *), GFP_KERNEL); + /* + * XXX: actual pages are allocated by xdr layer in + * xdr_partial_copy_from_skb. + */ + if (!arg->pages) + return -ENOMEM; + return 0; +} /* * Public functions @@ -261,10 +281,16 @@ int gssp_accept_sec_context_upcall(struct net *net, arg.context_handle = &ctxh; res.output_token->len = GSSX_max_output_token_sz; + ret = gssp_alloc_receive_pages(&arg); + if (ret) + return ret; + /* use nfs/ for targ_name ? */ ret = gssp_call(net, &msg); + gssp_free_receive_pages(&arg); + /* we need to fetch all data even in case of error so * that we can free special strctures is they have been allocated */ data->major_status = res.status.major_status; @@ -328,7 +354,6 @@ void gssp_free_upcall_data(struct gssp_upcall_data *data) kfree(data->in_handle.data); kfree(data->out_handle.data); kfree(data->out_token.data); - kfree(data->mech_oid.data); free_svc_cred(&data->creds); } diff --git a/net/sunrpc/auth_gss/gss_rpc_xdr.c b/net/sunrpc/auth_gss/gss_rpc_xdr.c index 357f613df7f..f0f78c5f1c7 100644 --- a/net/sunrpc/auth_gss/gss_rpc_xdr.c +++ b/net/sunrpc/auth_gss/gss_rpc_xdr.c @@ -166,14 +166,15 @@ static int dummy_dec_opt_array(struct xdr_stream *xdr, return 0; } -static int get_s32(void **p, void *max, s32 *res) +static int get_host_u32(struct xdr_stream *xdr, u32 *res) { - void *base = *p; - void *next = (void *)((char *)base + sizeof(s32)); - if (unlikely(next > max || next < base)) + __be32 *p; + + p = xdr_inline_decode(xdr, 4); + if (!p) return -EINVAL; - memcpy(res, base, sizeof(s32)); - *p = next; + /* Contents of linux creds are all host-endian: */ + memcpy(res, p, sizeof(u32)); return 0; } @@ -182,9 +183,9 @@ static int gssx_dec_linux_creds(struct xdr_stream *xdr, { u32 length; __be32 *p; - void *q, *end; - s32 tmp; - int N, i, err; + u32 tmp; + u32 N; + int i, err; p = xdr_inline_decode(xdr, 4); if (unlikely(p == NULL)) @@ -192,33 +193,28 @@ static int gssx_dec_linux_creds(struct xdr_stream *xdr, length = be32_to_cpup(p); - /* FIXME: we do not want to use the scratch buffer for this one - * may need to use functions that allows us to access an io vector - * directly */ - p = xdr_inline_decode(xdr, length); - if (unlikely(p == NULL)) + if (length > (3 + NGROUPS_MAX) * sizeof(u32)) return -ENOSPC; - q = p; - end = q + length; - /* uid */ - err = get_s32(&q, end, &tmp); + err = get_host_u32(xdr, &tmp); if (err) return err; creds->cr_uid = make_kuid(&init_user_ns, tmp); /* gid */ - err = get_s32(&q, end, &tmp); + err = get_host_u32(xdr, &tmp); if (err) return err; creds->cr_gid = make_kgid(&init_user_ns, tmp); /* number of additional gid's */ - err = get_s32(&q, end, &tmp); + err = get_host_u32(xdr, &tmp); if (err) return err; N = tmp; + if ((3 + N) * sizeof(u32) != length) + return -EINVAL; creds->cr_group_info = groups_alloc(N); if (creds->cr_group_info == NULL) return -ENOMEM; @@ -226,7 +222,7 @@ static int gssx_dec_linux_creds(struct xdr_stream *xdr, /* gid's */ for (i = 0; i < N; i++) { kgid_t kgid; - err = get_s32(&q, end, &tmp); + err = get_host_u32(xdr, &tmp); if (err) goto out_free_groups; err = -EINVAL; @@ -430,7 +426,7 @@ static int dummy_enc_nameattr_array(struct xdr_stream *xdr, static int dummy_dec_nameattr_array(struct xdr_stream *xdr, struct gssx_name_attr_array *naa) { - struct gssx_name_attr dummy; + struct gssx_name_attr dummy = { .attr = {.len = 0} }; u32 count, i; __be32 *p; @@ -493,12 +489,13 @@ static int gssx_enc_name(struct xdr_stream *xdr, return err; } + static int gssx_dec_name(struct xdr_stream *xdr, struct gssx_name *name) { - struct xdr_netobj dummy_netobj; - struct gssx_name_attr_array dummy_name_attr_array; - struct gssx_option_array dummy_option_array; + struct xdr_netobj dummy_netobj = { .len = 0 }; + struct gssx_name_attr_array dummy_name_attr_array = { .count = 0 }; + struct gssx_option_array dummy_option_array = { .count = 0 }; int err; /* name->display_name */ @@ -783,6 +780,9 @@ void gssx_enc_accept_sec_context(struct rpc_rqst *req, /* arg->options */ err = dummy_enc_opt_array(xdr, &arg->options); + xdr_inline_pages(&req->rq_rcv_buf, + PAGE_SIZE/2 /* pretty arbitrary */, + arg->pages, 0 /* page base */, arg->npages * PAGE_SIZE); done: if (err) dprintk("RPC: gssx_enc_accept_sec_context: %d\n", err); diff --git a/net/sunrpc/auth_gss/gss_rpc_xdr.h b/net/sunrpc/auth_gss/gss_rpc_xdr.h index 1c98b27d870..685a688f3d8 100644 --- a/net/sunrpc/auth_gss/gss_rpc_xdr.h +++ b/net/sunrpc/auth_gss/gss_rpc_xdr.h @@ -147,6 +147,8 @@ struct gssx_arg_accept_sec_context { struct gssx_cb *input_cb; u32 ret_deleg_cred; struct gssx_option_array options; + struct page **pages; + unsigned int npages; }; struct gssx_res_accept_sec_context { @@ -240,7 +242,8 @@ int gssx_dec_accept_sec_context(struct rpc_rqst *rqstp, 2 * GSSX_max_princ_sz + \ 8 + 8 + 4 + 4 + 4) #define GSSX_max_output_token_sz 1024 -#define GSSX_max_creds_sz (4 + 4 + 4 + NGROUPS_MAX * 4) +/* grouplist not included; we allocate separate pages for that: */ +#define GSSX_max_creds_sz (4 + 4 + 4 /* + NGROUPS_MAX*4 */) #define GSSX_RES_accept_sec_context_sz (GSSX_default_status_sz + \ GSSX_default_ctx_sz + \ GSSX_max_output_token_sz + \ diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 5a750b9c364..5b1bf7b530f 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1407,9 +1407,9 @@ call_refreshresult(struct rpc_task *task) return; case -ETIMEDOUT: rpc_delay(task, 3*HZ); - case -EKEYEXPIRED: case -EAGAIN: status = -EACCES; + case -EKEYEXPIRED: if (!task->tk_cred_retry) break; task->tk_cred_retry--; @@ -1644,6 +1644,10 @@ call_connect(struct rpc_task *task) task->tk_action = call_connect_status; if (task->tk_status < 0) return; + if (task->tk_flags & RPC_TASK_NOCONNECT) { + rpc_exit(task, -ENOTCONN); + return; + } xprt_connect(task); } } diff --git a/net/sunrpc/netns.h b/net/sunrpc/netns.h index 74d948f5d5a..779742cfc1f 100644 --- a/net/sunrpc/netns.h +++ b/net/sunrpc/netns.h @@ -23,6 +23,7 @@ struct sunrpc_net { struct rpc_clnt *rpcb_local_clnt4; spinlock_t rpcb_clnt_lock; unsigned int rpcb_users; + unsigned int rpcb_is_af_local : 1; struct mutex gssp_lock; wait_queue_head_t gssp_wq; diff --git a/net/sunrpc/rpcb_clnt.c b/net/sunrpc/rpcb_clnt.c index 3df764dc330..1891a1022c1 100644 --- a/net/sunrpc/rpcb_clnt.c +++ b/net/sunrpc/rpcb_clnt.c @@ -204,13 +204,15 @@ void rpcb_put_local(struct net *net) } static void rpcb_set_local(struct net *net, struct rpc_clnt *clnt, - struct rpc_clnt *clnt4) + struct rpc_clnt *clnt4, + bool is_af_local) { struct sunrpc_net *sn = net_generic(net, sunrpc_net_id); /* Protected by rpcb_create_local_mutex */ sn->rpcb_local_clnt = clnt; sn->rpcb_local_clnt4 = clnt4; + sn->rpcb_is_af_local = is_af_local ? 1 : 0; smp_wmb(); sn->rpcb_users = 1; dprintk("RPC: created new rpcb local clients (rpcb_local_clnt: " @@ -238,6 +240,14 @@ static int rpcb_create_local_unix(struct net *net) .program = &rpcb_program, .version = RPCBVERS_2, .authflavor = RPC_AUTH_NULL, + /* + * We turn off the idle timeout to prevent the kernel + * from automatically disconnecting the socket. + * Otherwise, we'd have to cache the mount namespace + * of the caller and somehow pass that to the socket + * reconnect code. + */ + .flags = RPC_CLNT_CREATE_NO_IDLE_TIMEOUT, }; struct rpc_clnt *clnt, *clnt4; int result = 0; @@ -263,7 +273,7 @@ static int rpcb_create_local_unix(struct net *net) clnt4 = NULL; } - rpcb_set_local(net, clnt, clnt4); + rpcb_set_local(net, clnt, clnt4, true); out: return result; @@ -315,7 +325,7 @@ static int rpcb_create_local_net(struct net *net) clnt4 = NULL; } - rpcb_set_local(net, clnt, clnt4); + rpcb_set_local(net, clnt, clnt4, false); out: return result; @@ -376,13 +386,16 @@ static struct rpc_clnt *rpcb_create(struct net *net, const char *hostname, return rpc_create(&args); } -static int rpcb_register_call(struct rpc_clnt *clnt, struct rpc_message *msg) +static int rpcb_register_call(struct sunrpc_net *sn, struct rpc_clnt *clnt, struct rpc_message *msg, bool is_set) { - int result, error = 0; + int flags = RPC_TASK_NOCONNECT; + int error, result = 0; + if (is_set || !sn->rpcb_is_af_local) + flags = RPC_TASK_SOFTCONN; msg->rpc_resp = &result; - error = rpc_call_sync(clnt, msg, RPC_TASK_SOFTCONN); + error = rpc_call_sync(clnt, msg, flags); if (error < 0) { dprintk("RPC: failed to contact local rpcbind " "server (errno %d).\n", -error); @@ -439,16 +452,19 @@ int rpcb_register(struct net *net, u32 prog, u32 vers, int prot, unsigned short .rpc_argp = &map, }; struct sunrpc_net *sn = net_generic(net, sunrpc_net_id); + bool is_set = false; dprintk("RPC: %sregistering (%u, %u, %d, %u) with local " "rpcbind\n", (port ? "" : "un"), prog, vers, prot, port); msg.rpc_proc = &rpcb_procedures2[RPCBPROC_UNSET]; - if (port) + if (port != 0) { msg.rpc_proc = &rpcb_procedures2[RPCBPROC_SET]; + is_set = true; + } - return rpcb_register_call(sn->rpcb_local_clnt, &msg); + return rpcb_register_call(sn, sn->rpcb_local_clnt, &msg, is_set); } /* @@ -461,6 +477,7 @@ static int rpcb_register_inet4(struct sunrpc_net *sn, const struct sockaddr_in *sin = (const struct sockaddr_in *)sap; struct rpcbind_args *map = msg->rpc_argp; unsigned short port = ntohs(sin->sin_port); + bool is_set = false; int result; map->r_addr = rpc_sockaddr2uaddr(sap, GFP_KERNEL); @@ -471,10 +488,12 @@ static int rpcb_register_inet4(struct sunrpc_net *sn, map->r_addr, map->r_netid); msg->rpc_proc = &rpcb_procedures4[RPCBPROC_UNSET]; - if (port) + if (port != 0) { msg->rpc_proc = &rpcb_procedures4[RPCBPROC_SET]; + is_set = true; + } - result = rpcb_register_call(sn->rpcb_local_clnt4, msg); + result = rpcb_register_call(sn, sn->rpcb_local_clnt4, msg, is_set); kfree(map->r_addr); return result; } @@ -489,6 +508,7 @@ static int rpcb_register_inet6(struct sunrpc_net *sn, const struct sockaddr_in6 *sin6 = (const struct sockaddr_in6 *)sap; struct rpcbind_args *map = msg->rpc_argp; unsigned short port = ntohs(sin6->sin6_port); + bool is_set = false; int result; map->r_addr = rpc_sockaddr2uaddr(sap, GFP_KERNEL); @@ -499,10 +519,12 @@ static int rpcb_register_inet6(struct sunrpc_net *sn, map->r_addr, map->r_netid); msg->rpc_proc = &rpcb_procedures4[RPCBPROC_UNSET]; - if (port) + if (port != 0) { msg->rpc_proc = &rpcb_procedures4[RPCBPROC_SET]; + is_set = true; + } - result = rpcb_register_call(sn->rpcb_local_clnt4, msg); + result = rpcb_register_call(sn, sn->rpcb_local_clnt4, msg, is_set); kfree(map->r_addr); return result; } @@ -519,7 +541,7 @@ static int rpcb_unregister_all_protofamilies(struct sunrpc_net *sn, map->r_addr = ""; msg->rpc_proc = &rpcb_procedures4[RPCBPROC_UNSET]; - return rpcb_register_call(sn->rpcb_local_clnt4, msg); + return rpcb_register_call(sn, sn->rpcb_local_clnt4, msg, false); } /** diff --git a/net/sunrpc/svcauth_unix.c b/net/sunrpc/svcauth_unix.c index 06bdf5a1082..1583c8a4eb7 100644 --- a/net/sunrpc/svcauth_unix.c +++ b/net/sunrpc/svcauth_unix.c @@ -493,8 +493,6 @@ static int unix_gid_parse(struct cache_detail *cd, if (rv) return -EINVAL; uid = make_kuid(&init_user_ns, id); - if (!uid_valid(uid)) - return -EINVAL; ug.uid = uid; expiry = get_expiry(&mesg); diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index 0f679df7d07..305374d4fb9 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -917,7 +917,10 @@ static void svc_tcp_clear_pages(struct svc_sock *svsk) len = svsk->sk_datalen; npages = (len + PAGE_SIZE - 1) >> PAGE_SHIFT; for (i = 0; i < npages; i++) { - BUG_ON(svsk->sk_pages[i] == NULL); + if (svsk->sk_pages[i] == NULL) { + WARN_ON_ONCE(1); + continue; + } put_page(svsk->sk_pages[i]); svsk->sk_pages[i] = NULL; } @@ -1092,8 +1095,10 @@ static int svc_tcp_recvfrom(struct svc_rqst *rqstp) goto err_noclose; } - if (svc_sock_reclen(svsk) < 8) + if (svsk->sk_datalen < 8) { + svsk->sk_datalen = 0; goto err_delete; /* client is nuts. */ + } rqstp->rq_arg.len = svsk->sk_datalen; rqstp->rq_arg.page_base = 0; diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c index 75edcfad6e2..1504bb11e4f 100644 --- a/net/sunrpc/xdr.c +++ b/net/sunrpc/xdr.c @@ -207,10 +207,13 @@ _shift_data_right_pages(struct page **pages, size_t pgto_base, pgfrom_base -= copy; vto = kmap_atomic(*pgto); - vfrom = kmap_atomic(*pgfrom); - memmove(vto + pgto_base, vfrom + pgfrom_base, copy); + if (*pgto != *pgfrom) { + vfrom = kmap_atomic(*pgfrom); + memcpy(vto + pgto_base, vfrom + pgfrom_base, copy); + kunmap_atomic(vfrom); + } else + memmove(vto + pgto_base, vto + pgfrom_base, copy); flush_dcache_page(*pgto); - kunmap_atomic(vfrom); kunmap_atomic(vto); } while ((len -= copy) != 0); diff --git a/net/sunrpc/xprtrdma/svc_rdma_marshal.c b/net/sunrpc/xprtrdma/svc_rdma_marshal.c index 8d2edddf48c..65b146297f5 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_marshal.c +++ b/net/sunrpc/xprtrdma/svc_rdma_marshal.c @@ -98,6 +98,7 @@ void svc_rdma_rcl_chunk_counts(struct rpcrdma_read_chunk *ch, */ static u32 *decode_write_list(u32 *va, u32 *vaend) { + unsigned long start, end; int nchunks; struct rpcrdma_write_array *ary = @@ -113,9 +114,12 @@ static u32 *decode_write_list(u32 *va, u32 *vaend) return NULL; } nchunks = ntohl(ary->wc_nchunks); - if (((unsigned long)&ary->wc_array[0] + - (sizeof(struct rpcrdma_write_chunk) * nchunks)) > - (unsigned long)vaend) { + + start = (unsigned long)&ary->wc_array[0]; + end = (unsigned long)vaend; + if (nchunks < 0 || + nchunks > (SIZE_MAX - start) / sizeof(struct rpcrdma_write_chunk) || + (start + (sizeof(struct rpcrdma_write_chunk) * nchunks)) > end) { dprintk("svcrdma: ary=%p, wc_nchunks=%d, vaend=%p\n", ary, nchunks, vaend); return NULL; @@ -129,6 +133,7 @@ static u32 *decode_write_list(u32 *va, u32 *vaend) static u32 *decode_reply_array(u32 *va, u32 *vaend) { + unsigned long start, end; int nchunks; struct rpcrdma_write_array *ary = (struct rpcrdma_write_array *)va; @@ -143,9 +148,12 @@ static u32 *decode_reply_array(u32 *va, u32 *vaend) return NULL; } nchunks = ntohl(ary->wc_nchunks); - if (((unsigned long)&ary->wc_array[0] + - (sizeof(struct rpcrdma_write_chunk) * nchunks)) > - (unsigned long)vaend) { + + start = (unsigned long)&ary->wc_array[0]; + end = (unsigned long)vaend; + if (nchunks < 0 || + nchunks > (SIZE_MAX - start) / sizeof(struct rpcrdma_write_chunk) || + (start + (sizeof(struct rpcrdma_write_chunk) * nchunks)) > end) { dprintk("svcrdma: ary=%p, wc_nchunks=%d, vaend=%p\n", ary, nchunks, vaend); return NULL; diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index ffd50348a50..8a0e04d0928 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -391,8 +391,10 @@ static int xs_send_kvec(struct socket *sock, struct sockaddr *addr, int addrlen, return kernel_sendmsg(sock, &msg, NULL, 0, 0); } -static int xs_send_pagedata(struct socket *sock, struct xdr_buf *xdr, unsigned int base, int more) +static int xs_send_pagedata(struct socket *sock, struct xdr_buf *xdr, unsigned int base, int more, bool zerocopy) { + ssize_t (*do_sendpage)(struct socket *sock, struct page *page, + int offset, size_t size, int flags); struct page **ppage; unsigned int remainder; int err, sent = 0; @@ -401,6 +403,9 @@ static int xs_send_pagedata(struct socket *sock, struct xdr_buf *xdr, unsigned i base += xdr->page_base; ppage = xdr->pages + (base >> PAGE_SHIFT); base &= ~PAGE_MASK; + do_sendpage = sock->ops->sendpage; + if (!zerocopy) + do_sendpage = sock_no_sendpage; for(;;) { unsigned int len = min_t(unsigned int, PAGE_SIZE - base, remainder); int flags = XS_SENDMSG_FLAGS; @@ -408,7 +413,7 @@ static int xs_send_pagedata(struct socket *sock, struct xdr_buf *xdr, unsigned i remainder -= len; if (remainder != 0 || more) flags |= MSG_MORE; - err = sock->ops->sendpage(sock, *ppage, base, len, flags); + err = do_sendpage(sock, *ppage, base, len, flags); if (remainder == 0 || err != len) break; sent += err; @@ -429,9 +434,10 @@ static int xs_send_pagedata(struct socket *sock, struct xdr_buf *xdr, unsigned i * @addrlen: UDP only -- length of destination address * @xdr: buffer containing this request * @base: starting position in the buffer + * @zerocopy: true if it is safe to use sendpage() * */ -static int xs_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, struct xdr_buf *xdr, unsigned int base) +static int xs_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, struct xdr_buf *xdr, unsigned int base, bool zerocopy) { unsigned int remainder = xdr->len - base; int err, sent = 0; @@ -459,7 +465,7 @@ static int xs_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, if (base < xdr->page_len) { unsigned int len = xdr->page_len - base; remainder -= len; - err = xs_send_pagedata(sock, xdr, base, remainder != 0); + err = xs_send_pagedata(sock, xdr, base, remainder != 0, zerocopy); if (remainder == 0 || err != len) goto out; sent += err; @@ -562,7 +568,7 @@ static int xs_local_send_request(struct rpc_task *task) req->rq_svec->iov_base, req->rq_svec->iov_len); status = xs_sendpages(transport->sock, NULL, 0, - xdr, req->rq_bytes_sent); + xdr, req->rq_bytes_sent, true); dprintk("RPC: %s(%u) = %d\n", __func__, xdr->len - req->rq_bytes_sent, status); if (likely(status >= 0)) { @@ -618,7 +624,7 @@ static int xs_udp_send_request(struct rpc_task *task) status = xs_sendpages(transport->sock, xs_addr(xprt), xprt->addrlen, xdr, - req->rq_bytes_sent); + req->rq_bytes_sent, true); dprintk("RPC: xs_udp_send_request(%u) = %d\n", xdr->len - req->rq_bytes_sent, status); @@ -689,6 +695,7 @@ static int xs_tcp_send_request(struct rpc_task *task) struct rpc_xprt *xprt = req->rq_xprt; struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt); struct xdr_buf *xdr = &req->rq_snd_buf; + bool zerocopy = true; int status; xs_encode_stream_record_marker(&req->rq_snd_buf); @@ -696,13 +703,20 @@ static int xs_tcp_send_request(struct rpc_task *task) xs_pktdump("packet data:", req->rq_svec->iov_base, req->rq_svec->iov_len); + /* Don't use zero copy if this is a resend. If the RPC call + * completes while the socket holds a reference to the pages, + * then we may end up resending corrupted data. + */ + if (task->tk_flags & RPC_TASK_SENT) + zerocopy = false; /* Continue transmitting the packet/record. We must be careful * to cope with writespace callbacks arriving _after_ we have * called sendmsg(). */ while (1) { status = xs_sendpages(transport->sock, - NULL, 0, xdr, req->rq_bytes_sent); + NULL, 0, xdr, req->rq_bytes_sent, + zerocopy); dprintk("RPC: xs_tcp_send_request(%u) = %d\n", xdr->len - req->rq_bytes_sent, status); |