Added support for ARM optimizations to configure.ac

ARM NEON optimizations need the use of 16-bit data types and
benefit from 16-byte alignemnt, so we want to have macro
WITH_SIMD defined in the code.

Also automated tests are updated to work correctly on ARM.
And the required stubs for ARM optimizations added (just
copied from 'jsimd_none.c').

4 files changed, 319 insertions, 1 deletions

diff --git a/Makefile.am b/Makefile.am
index adc4cb6..4e59a68 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -103,7 +103,7 @@ dist-hook:
 	rm -rf `find $(distdir) -name .svn`
 
 
-if WITH_SIMD
+if SIMD_TESTS
 
 test: testclean all
 	./jpegut
diff --git a/configure.ac b/configure.ac
index fb022f2..02e2778 100644
--- a/configure.ac
+++ b/configure.ac
@@ -17,6 +17,7 @@ SAVED_CXXFLAGS=${CXXFLAGS}
 AC_PROG_CPP
 AC_PROG_CC
 AC_PROG_CXX
+AM_PROG_AS
 AC_PROG_INSTALL
 AC_PROG_LIBTOOL
 AC_PROG_LN_S
@@ -218,11 +219,17 @@ if test "x${with_simd}" != "xno"; then
       AC_MSG_RESULT([yes (x86_64)])
       AC_PROG_NASM
       simd_arch=x86_64
+      simd_tests=yes
     ;;
     i*86 | x86 | ia32)
       AC_MSG_RESULT([yes (i386)])
       AC_PROG_NASM
       simd_arch=i386
+      simd_tests=yes
+    ;;
+    arm*)
+      AC_MSG_RESULT([yes (arm)])
+      simd_arch=arm
     ;;
     *)
       AC_MSG_RESULT([no ("$host_cpu")])
@@ -237,8 +244,10 @@ if test "x${with_simd}" != "xno"; then
 fi
 
 AM_CONDITIONAL([WITH_SIMD], [test "x$with_simd" != "xno"])
+AM_CONDITIONAL([SIMD_TESTS], [test "x$simd_tests" = "xyes"])
 AM_CONDITIONAL([SIMD_I386], [test "x$simd_arch" = "xi386"])
 AM_CONDITIONAL([SIMD_X86_64], [test "x$simd_arch" = "xx86_64"])
+AM_CONDITIONAL([SIMD_ARM], [test "x$simd_arch" = "xarm"])
 AM_CONDITIONAL([X86_64], [test "x$host_cpu" = "xx86_64" -o "x$host_cpu" = "xamd64"])
 
 case "$host_cpu" in
diff --git a/simd/Makefile.am b/simd/Makefile.am
index 81c23af..88bac93 100644
--- a/simd/Makefile.am
+++ b/simd/Makefile.am
@@ -49,6 +49,12 @@ jdmermmx.lo: jdmrgmmx.asm
 jdmerss2.lo: jdmrgss2.asm
 endif
 
+if SIMD_ARM
+
+libsimd_la_SOURCES = jsimd_arm.c
+
+endif
+
 AM_CPPFLAGS = -I$(top_srcdir) 
 
 .asm.lo:
diff --git a/simd/jsimd_arm.c b/simd/jsimd_arm.c
new file mode 100644
index 0000000..a03cd04
--- /dev/null
+++ b/simd/jsimd_arm.c
@@ -0,0 +1,303 @@
+/*
+ * jsimd_arm.c
+ *
+ * Copyright 2009 Pierre Ossman <ossman@cendio.se> for Cendio AB
+ * Copyright 2009 D. R. Commander
+ * 
+ * Based on the x86 SIMD extension for IJG JPEG library,
+ * Copyright (C) 1999-2006, MIYASAKA Masaru.
+ *
+ * This file contains the interface between the "normal" portions
+ * of the library and the SIMD implementations when running on
+ * ARM architecture.
+ *
+ * Based on the stubs from 'jsimd_none.c'
+ */
+
+#define JPEG_INTERNALS
+#include "../jinclude.h"
+#include "../jpeglib.h"
+#include "../jsimd.h"
+#include "../jdct.h"
+#include "../jsimddct.h"
+
+GLOBAL(int)
+jsimd_can_rgb_ycc (void)
+{
+  return 0;
+}
+
+GLOBAL(int)
+jsimd_can_ycc_rgb (void)
+{
+  return 0;
+}
+
+GLOBAL(void)
+jsimd_rgb_ycc_convert (j_compress_ptr cinfo,
+                       JSAMPARRAY input_buf, JSAMPIMAGE output_buf,
+                       JDIMENSION output_row, int num_rows)
+{
+}
+
+GLOBAL(void)
+jsimd_ycc_rgb_convert (j_decompress_ptr cinfo,
+                       JSAMPIMAGE input_buf, JDIMENSION input_row,
+                       JSAMPARRAY output_buf, int num_rows)
+{
+}
+
+GLOBAL(int)
+jsimd_can_h2v2_downsample (void)
+{
+  return 0;
+}
+
+GLOBAL(int)
+jsimd_can_h2v1_downsample (void)
+{
+  return 0;
+}
+
+GLOBAL(void)
+jsimd_h2v2_downsample (j_compress_ptr cinfo, jpeg_component_info * compptr,
+                       JSAMPARRAY input_data, JSAMPARRAY output_data)
+{
+}
+
+GLOBAL(void)
+jsimd_h2v1_downsample (j_compress_ptr cinfo, jpeg_component_info * compptr,
+                       JSAMPARRAY input_data, JSAMPARRAY output_data)
+{
+}
+
+GLOBAL(int)
+jsimd_can_h2v2_upsample (void)
+{
+  return 0;
+}
+
+GLOBAL(int)
+jsimd_can_h2v1_upsample (void)
+{
+  return 0;
+}
+
+GLOBAL(void)
+jsimd_h2v2_upsample (j_decompress_ptr cinfo,
+                     jpeg_component_info * compptr, 
+                     JSAMPARRAY input_data,
+                     JSAMPARRAY * output_data_ptr)
+{
+}
+
+GLOBAL(void)
+jsimd_h2v1_upsample (j_decompress_ptr cinfo,
+                     jpeg_component_info * compptr, 
+                     JSAMPARRAY input_data,
+                     JSAMPARRAY * output_data_ptr)
+{
+}
+
+GLOBAL(int)
+jsimd_can_h2v2_fancy_upsample (void)
+{
+  return 0;
+}
+
+GLOBAL(int)
+jsimd_can_h2v1_fancy_upsample (void)
+{
+  return 0;
+}
+
+GLOBAL(void)
+jsimd_h2v2_fancy_upsample (j_decompress_ptr cinfo,
+                           jpeg_component_info * compptr, 
+                           JSAMPARRAY input_data,
+                           JSAMPARRAY * output_data_ptr)
+{
+}
+
+GLOBAL(void)
+jsimd_h2v1_fancy_upsample (j_decompress_ptr cinfo,
+                           jpeg_component_info * compptr, 
+                           JSAMPARRAY input_data,
+                           JSAMPARRAY * output_data_ptr)
+{
+}
+
+GLOBAL(int)
+jsimd_can_h2v2_merged_upsample (void)
+{
+  return 0;
+}
+
+GLOBAL(int)
+jsimd_can_h2v1_merged_upsample (void)
+{
+  return 0;
+}
+
+GLOBAL(void)
+jsimd_h2v2_merged_upsample (j_decompress_ptr cinfo,
+                            JSAMPIMAGE input_buf,
+                            JDIMENSION in_row_group_ctr,
+                            JSAMPARRAY output_buf)
+{
+}
+
+GLOBAL(void)
+jsimd_h2v1_merged_upsample (j_decompress_ptr cinfo,
+                            JSAMPIMAGE input_buf,
+                            JDIMENSION in_row_group_ctr,
+                            JSAMPARRAY output_buf)
+{
+}
+
+GLOBAL(int)
+jsimd_can_convsamp (void)
+{
+  return 0;
+}
+
+GLOBAL(int)
+jsimd_can_convsamp_float (void)
+{
+  return 0;
+}
+
+GLOBAL(void)
+jsimd_convsamp (JSAMPARRAY sample_data, JDIMENSION start_col,
+                DCTELEM * workspace)
+{
+}
+
+GLOBAL(void)
+jsimd_convsamp_float (JSAMPARRAY sample_data, JDIMENSION start_col,
+                      FAST_FLOAT * workspace)
+{
+}
+
+GLOBAL(int)
+jsimd_can_fdct_islow (void)
+{
+  return 0;
+}
+
+GLOBAL(int)
+jsimd_can_fdct_ifast (void)
+{
+  return 0;
+}
+
+GLOBAL(int)
+jsimd_can_fdct_float (void)
+{
+  return 0;
+}
+
+GLOBAL(void)
+jsimd_fdct_islow (DCTELEM * data)
+{
+}
+
+GLOBAL(void)
+jsimd_fdct_ifast (DCTELEM * data)
+{
+}
+
+GLOBAL(void)
+jsimd_fdct_float (FAST_FLOAT * data)
+{
+}
+
+GLOBAL(int)
+jsimd_can_quantize (void)
+{
+  return 0;
+}
+
+GLOBAL(int)
+jsimd_can_quantize_float (void)
+{
+  return 0;
+}
+
+GLOBAL(void)
+jsimd_quantize (JCOEFPTR coef_block, DCTELEM * divisors,
+                DCTELEM * workspace)
+{
+}
+
+GLOBAL(void)
+jsimd_quantize_float (JCOEFPTR coef_block, FAST_FLOAT * divisors,
+                      FAST_FLOAT * workspace)
+{
+}
+
+GLOBAL(int)
+jsimd_can_idct_2x2 (void)
+{
+  return 0;
+}
+
+GLOBAL(int)
+jsimd_can_idct_4x4 (void)
+{
+  return 0;
+}
+
+GLOBAL(void)
+jsimd_idct_2x2 (j_decompress_ptr cinfo, jpeg_component_info * compptr,
+                JCOEFPTR coef_block, JSAMPARRAY output_buf,
+                JDIMENSION output_col)
+{
+}
+
+GLOBAL(void)
+jsimd_idct_4x4 (j_decompress_ptr cinfo, jpeg_component_info * compptr,
+                JCOEFPTR coef_block, JSAMPARRAY output_buf,
+                JDIMENSION output_col)
+{
+}
+
+GLOBAL(int)
+jsimd_can_idct_islow (void)
+{
+  return 0;
+}
+
+GLOBAL(int)
+jsimd_can_idct_ifast (void)
+{
+  return 0;
+}
+
+GLOBAL(int)
+jsimd_can_idct_float (void)
+{
+  return 0;
+}
+
+GLOBAL(void)
+jsimd_idct_islow (j_decompress_ptr cinfo, jpeg_component_info * compptr,
+                JCOEFPTR coef_block, JSAMPARRAY output_buf,
+                JDIMENSION output_col)
+{
+}
+
+GLOBAL(void)
+jsimd_idct_ifast (j_decompress_ptr cinfo, jpeg_component_info * compptr,
+                JCOEFPTR coef_block, JSAMPARRAY output_buf,
+                JDIMENSION output_col)
+{
+}
+
+GLOBAL(void)
+jsimd_idct_float (j_decompress_ptr cinfo, jpeg_component_info * compptr,
+                JCOEFPTR coef_block, JSAMPARRAY output_buf,
+                JDIMENSION output_col)
+{
+}
+