asm: FF_-prefix internal macros used in inline assembly

See merge commit '39d6d3618d48625decaff7d9bdbb45b44ef2a805'.

9 files changed, 243 insertions, 243 deletions

diff --git a/libavcodec/x86/cabac.h b/libavcodec/x86/cabac.h
index 4795f5bb07..cfd3b759c9 100644
--- a/libavcodec/x86/cabac.h
+++ b/libavcodec/x86/cabac.h
@@ -45,7 +45,7 @@
 #define END_CHECK(end) ""
 #else
 #define END_CHECK(end) \
-        "cmp    "end"       , %%"REG_c"                                 \n\t"\
+        "cmp    "end"       , %%"FF_REG_c"                              \n\t"\
         "jge    1f                                                      \n\t"
 #endif
 
@@ -92,11 +92,11 @@
         "mov    "tmpbyte"   , "statep"                                  \n\t"\
         "test   "lowword"   , "lowword"                                 \n\t"\
         "jnz    2f                                                      \n\t"\
-        "mov    "byte"      , %%"REG_c"                                 \n\t"\
+        "mov    "byte"      , %%"FF_REG_c"                              \n\t"\
         END_CHECK(end)\
-        "add"OPSIZE" $2     , "byte"                                    \n\t"\
+        "add"FF_OPSIZE" $2  , "byte"                                    \n\t"\
         "1:                                                             \n\t"\
-        "movzwl (%%"REG_c") , "tmp"                                     \n\t"\
+        "movzwl (%%"FF_REG_c") , "tmp"                                  \n\t"\
         "lea    -1("low")   , %%ecx                                     \n\t"\
         "xor    "low"       , %%ecx                                     \n\t"\
         "shr    $15         , %%ecx                                     \n\t"\
@@ -153,11 +153,11 @@
         "mov    "tmpbyte"   , "statep"                                  \n\t"\
         "test   "lowword"   , "lowword"                                 \n\t"\
         " jnz   2f                                                      \n\t"\
-        "mov    "byte"      , %%"REG_c"                                 \n\t"\
+        "mov    "byte"      , %%"FF_REG_c"                              \n\t"\
         END_CHECK(end)\
-        "add"OPSIZE" $2     , "byte"                                    \n\t"\
+        "add"FF_OPSIZE" $2  , "byte"                                    \n\t"\
         "1:                                                             \n\t"\
-        "movzwl (%%"REG_c")     , "tmp"                                 \n\t"\
+        "movzwl (%%"FF_REG_c") , "tmp"                                  \n\t"\
         "lea    -1("low")   , %%ecx                                     \n\t"\
         "xor    "low"       , %%ecx                                     \n\t"\
         "shr    $15         , %%ecx                                     \n\t"\
@@ -203,7 +203,7 @@ static av_always_inline int get_cabac_inline_x86(CABACContext *c,
           "i"(offsetof(CABACContext, bytestream_end))
           TABLES_ARG
           ,"1"(c->low), "2"(c->range)
-        : "%"REG_c, "memory"
+        : "%"FF_REG_c, "memory"
     );
     return bit & 1;
 }
@@ -240,7 +240,7 @@ static av_always_inline int get_cabac_bypass_sign_x86(CABACContext *c, int val)
         "addl          %%edx, %%eax     \n\t"
         "cmp         %c5(%2), %1        \n\t"
         "jge              1f            \n\t"
-        "add"OPSIZE"      $2, %c4(%2)   \n\t"
+        "add"FF_OPSIZE"   $2, %c4(%2)   \n\t"
 #endif
         "1:                             \n\t"
         "movl          %%eax, %c3(%2)   \n\t"
@@ -281,7 +281,7 @@ static av_always_inline int get_cabac_bypass_x86(CABACContext *c)
         "addl          %%ecx, %%eax     \n\t"
         "cmp         %c5(%2), %1        \n\t"
         "jge              1f            \n\t"
-        "add"OPSIZE"      $2, %c4(%2)   \n\t"
+        "add"FF_OPSIZE"   $2, %c4(%2)   \n\t"
         "1:                             \n\t"
         "movl          %%eax, %c3(%2)   \n\t"
 
diff --git a/libavcodec/x86/h264_i386.h b/libavcodec/x86/h264_i386.h
index 4dfbc30933..19cd128381 100644
--- a/libavcodec/x86/h264_i386.h
+++ b/libavcodec/x86/h264_i386.h
@@ -91,13 +91,13 @@ static int decode_significance_x86(CABACContext *c, int max_coeff,
         "sub  %10, %1                           \n\t"
         "mov  %2, %0                            \n\t"
         "movl %7, %%ecx                         \n\t"
-        "add  %1, %%"REG_c"                     \n\t"
+        "add  %1, %%"FF_REG_c"                  \n\t"
         "movl %%ecx, (%0)                       \n\t"
 
         "test $1, %4                            \n\t"
         " jnz 5f                                \n\t"
 
-        "add"OPSIZE"  $4, %2                    \n\t"
+        "add"FF_OPSIZE"  $4, %2                 \n\t"
 
         "4:                                     \n\t"
         "add  $1, %1                            \n\t"
@@ -105,7 +105,7 @@ static int decode_significance_x86(CABACContext *c, int max_coeff,
         " jb 3b                                 \n\t"
         "mov  %2, %0                            \n\t"
         "movl %7, %%ecx                         \n\t"
-        "add  %1, %%"REG_c"                     \n\t"
+        "add  %1, %%"FF_REG_c"                  \n\t"
         "movl %%ecx, (%0)                       \n\t"
         "5:                                     \n\t"
         "add  %9, %k0                           \n\t"
@@ -116,7 +116,7 @@ static int decode_significance_x86(CABACContext *c, int max_coeff,
           "i"(offsetof(CABACContext, bytestream)),
           "i"(offsetof(CABACContext, bytestream_end))
           TABLES_ARG
-        : "%"REG_c, "memory"
+        : "%"FF_REG_c, "memory"
     );
     return coeff_count;
 }
@@ -183,7 +183,7 @@ static int decode_significance_8x8_x86(CABACContext *c,
         "test $1, %4                            \n\t"
         " jnz 5f                                \n\t"
 
-        "add"OPSIZE"  $4, %2                    \n\t"
+        "add"FF_OPSIZE"  $4, %2                 \n\t"
 
         "4:                                     \n\t"
         "add $1, %6                             \n\t"
@@ -202,7 +202,7 @@ static int decode_significance_8x8_x86(CABACContext *c,
           "i"(offsetof(CABACContext, bytestream)),
           "i"(offsetof(CABACContext, bytestream_end)),
           "i"(H264_LAST_COEFF_FLAG_OFFSET_8x8_OFFSET) TABLES_ARG
-        : "%"REG_c, "memory"
+        : "%"FF_REG_c, "memory"
     );
     return coeff_count;
 }
diff --git a/libavcodec/x86/hpeldsp_rnd_template.c b/libavcodec/x86/hpeldsp_rnd_template.c
index e20d0658cd..2bff2d2766 100644
--- a/libavcodec/x86/hpeldsp_rnd_template.c
+++ b/libavcodec/x86/hpeldsp_rnd_template.c
@@ -32,7 +32,7 @@ av_unused static void DEF(put, pixels8_x2)(uint8_t *block, const uint8_t *pixels
 {
     MOVQ_BFE(mm6);
     __asm__ volatile(
-        "lea    (%3, %3), %%"REG_a"     \n\t"
+        "lea    (%3, %3), %%"FF_REG_a"  \n\t"
         ".p2align 3                     \n\t"
         "1:                             \n\t"
         "movq   (%1), %%mm0             \n\t"
@@ -42,8 +42,8 @@ av_unused static void DEF(put, pixels8_x2)(uint8_t *block, const uint8_t *pixels
         PAVGBP(%%mm0, %%mm1, %%mm4,   %%mm2, %%mm3, %%mm5)
         "movq   %%mm4, (%2)             \n\t"
         "movq   %%mm5, (%2, %3)         \n\t"
-        "add    %%"REG_a", %1           \n\t"
-        "add    %%"REG_a", %2           \n\t"
+        "add    %%"FF_REG_a", %1        \n\t"
+        "add    %%"FF_REG_a", %2        \n\t"
         "movq   (%1), %%mm0             \n\t"
         "movq   1(%1), %%mm1            \n\t"
         "movq   (%1, %3), %%mm2         \n\t"
@@ -51,20 +51,20 @@ av_unused static void DEF(put, pixels8_x2)(uint8_t *block, const uint8_t *pixels
         PAVGBP(%%mm0, %%mm1, %%mm4,   %%mm2, %%mm3, %%mm5)
         "movq   %%mm4, (%2)             \n\t"
         "movq   %%mm5, (%2, %3)         \n\t"
-        "add    %%"REG_a", %1           \n\t"
-        "add    %%"REG_a", %2           \n\t"
+        "add    %%"FF_REG_a", %1        \n\t"
+        "add    %%"FF_REG_a", %2        \n\t"
         "subl   $4, %0                  \n\t"
         "jnz    1b                      \n\t"
         :"+g"(h), "+S"(pixels), "+D"(block)
         :"r"((x86_reg)line_size)
-        :REG_a, "memory");
+        :FF_REG_a, "memory");
 }
 
 av_unused static void DEF(put, pixels16_x2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     MOVQ_BFE(mm6);
     __asm__ volatile(
-        "lea        (%3, %3), %%"REG_a" \n\t"
+        "lea    (%3, %3), %%"FF_REG_a"  \n\t"
         ".p2align 3                     \n\t"
         "1:                             \n\t"
         "movq   (%1), %%mm0             \n\t"
@@ -81,8 +81,8 @@ av_unused static void DEF(put, pixels16_x2)(uint8_t *block, const uint8_t *pixel
         PAVGBP(%%mm0, %%mm1, %%mm4,   %%mm2, %%mm3, %%mm5)
         "movq   %%mm4, 8(%2)            \n\t"
         "movq   %%mm5, 8(%2, %3)        \n\t"
-        "add    %%"REG_a", %1           \n\t"
-        "add    %%"REG_a", %2           \n\t"
+        "add    %%"FF_REG_a", %1        \n\t"
+        "add    %%"FF_REG_a", %2        \n\t"
         "movq   (%1), %%mm0             \n\t"
         "movq   1(%1), %%mm1            \n\t"
         "movq   (%1, %3), %%mm2         \n\t"
@@ -97,42 +97,42 @@ av_unused static void DEF(put, pixels16_x2)(uint8_t *block, const uint8_t *pixel
         PAVGBP(%%mm0, %%mm1, %%mm4,   %%mm2, %%mm3, %%mm5)
         "movq   %%mm4, 8(%2)            \n\t"
         "movq   %%mm5, 8(%2, %3)        \n\t"
-        "add    %%"REG_a", %1           \n\t"
-        "add    %%"REG_a", %2           \n\t"
+        "add    %%"FF_REG_a", %1        \n\t"
+        "add    %%"FF_REG_a", %2        \n\t"
         "subl   $4, %0                  \n\t"
         "jnz    1b                      \n\t"
         :"+g"(h), "+S"(pixels), "+D"(block)
         :"r"((x86_reg)line_size)
-        :REG_a, "memory");
+        :FF_REG_a, "memory");
 }
 
 av_unused static void DEF(put, pixels8_y2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
 {
     MOVQ_BFE(mm6);
     __asm__ volatile(
-        "lea (%3, %3), %%"REG_a"        \n\t"
+        "lea (%3, %3), %%"FF_REG_a"     \n\t"
         "movq (%1), %%mm0               \n\t"
         ".p2align 3                     \n\t"
         "1:                             \n\t"
         "movq   (%1, %3), %%mm1         \n\t"
-        "movq   (%1, %%"REG_a"),%%mm2   \n\t"
+        "movq   (%1, %%"FF_REG_a"),%%mm2\n\t"
         PAVGBP(%%mm1, %%mm0, %%mm4,   %%mm2, %%mm1, %%mm5)
         "movq   %%mm4, (%2)             \n\t"
         "movq   %%mm5, (%2, %3)         \n\t"
-        "add    %%"REG_a", %1           \n\t"
-        "add    %%"REG_a", %2           \n\t"
+        "add    %%"FF_REG_a", %1        \n\t"
+        "add    %%"FF_REG_a", %2        \n\t"
         "movq   (%1, %3), %%mm1         \n\t"
-        "movq   (%1, %%"REG_a"),%%mm0   \n\t"
+        "movq   (%1, %%"FF_REG_a"),%%mm0\n\t"
         PAVGBP(%%mm1, %%mm2, %%mm4,   %%mm0, %%mm1, %%mm5)
         "movq   %%mm4, (%2)             \n\t"
         "movq   %%mm5, (%2, %3)         \n\t"
-        "add    %%"REG_a", %1           \n\t"
-        "add    %%"REG_a", %2           \n\t"
+        "add    %%"FF_REG_a", %1        \n\t"
+        "add    %%"FF_REG_a", %2        \n\t"
         "subl   $4, %0                  \n\t"
         "jnz    1b                      \n\t"
         :"+g"(h), "+S"(pixels), "+D"(block)
         :"r"((x86_reg)line_size)
-        :REG_a, "memory");
+        :FF_REG_a, "memory");
 }
 
 av_unused static void DEF(avg, pixels16_x2)(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
@@ -166,12 +166,12 @@ av_unused static void DEF(avg, pixels8_y2)(uint8_t *block, const uint8_t *pixels
 {
     MOVQ_BFE(mm6);
     __asm__ volatile(
-        "lea    (%3, %3), %%"REG_a"     \n\t"
+        "lea    (%3, %3), %%"FF_REG_a"  \n\t"
         "movq   (%1), %%mm0             \n\t"
         ".p2align 3                     \n\t"
         "1:                             \n\t"
         "movq   (%1, %3), %%mm1         \n\t"
-        "movq   (%1, %%"REG_a"), %%mm2  \n\t"
+        "movq   (%1, %%"FF_REG_a"), %%mm2 \n\t"
         PAVGBP(%%mm1, %%mm0, %%mm4,   %%mm2, %%mm1, %%mm5)
         "movq   (%2), %%mm3             \n\t"
         PAVGB_MMX(%%mm3, %%mm4, %%mm0, %%mm6)
@@ -179,11 +179,11 @@ av_unused static void DEF(avg, pixels8_y2)(uint8_t *block, const uint8_t *pixels
         PAVGB_MMX(%%mm3, %%mm5, %%mm1, %%mm6)
         "movq   %%mm0, (%2)             \n\t"
         "movq   %%mm1, (%2, %3)         \n\t"
-        "add    %%"REG_a", %1           \n\t"
-        "add    %%"REG_a", %2           \n\t"
+        "add    %%"FF_REG_a", %1        \n\t"
+        "add    %%"FF_REG_a", %2        \n\t"
 
         "movq   (%1, %3), %%mm1         \n\t"
-        "movq   (%1, %%"REG_a"), %%mm0  \n\t"
+        "movq   (%1, %%"FF_REG_a"), %%mm0 \n\t"
         PAVGBP(%%mm1, %%mm2, %%mm4,   %%mm0, %%mm1, %%mm5)
         "movq   (%2), %%mm3             \n\t"
         PAVGB_MMX(%%mm3, %%mm4, %%mm2, %%mm6)
@@ -191,12 +191,12 @@ av_unused static void DEF(avg, pixels8_y2)(uint8_t *block, const uint8_t *pixels
         PAVGB_MMX(%%mm3, %%mm5, %%mm1, %%mm6)
         "movq   %%mm2, (%2)             \n\t"
         "movq   %%mm1, (%2, %3)         \n\t"
-        "add    %%"REG_a", %1           \n\t"
-        "add    %%"REG_a", %2           \n\t"
+        "add    %%"FF_REG_a", %1        \n\t"
+        "add    %%"FF_REG_a", %2        \n\t"
 
         "subl   $4, %0                  \n\t"
         "jnz    1b                      \n\t"
         :"+g"(h), "+S"(pixels), "+D"(block)
         :"r"((x86_reg)line_size)
-        :REG_a, "memory");
+        :FF_REG_a, "memory");
 }
diff --git a/libavcodec/x86/me_cmp_init.c b/libavcodec/x86/me_cmp_init.c
index 49f50d0eed..dc3e6f8668 100644
--- a/libavcodec/x86/me_cmp_init.c
+++ b/libavcodec/x86/me_cmp_init.c
@@ -283,15 +283,15 @@ static inline void sad8_1_mmx(uint8_t *blk1, uint8_t *blk2,
     __asm__ volatile (
         ".p2align 4                     \n\t"
         "1:                             \n\t"
-        "movq (%1, %%"REG_a"), %%mm0    \n\t"
-        "movq (%2, %%"REG_a"), %%mm2    \n\t"
-        "movq (%2, %%"REG_a"), %%mm4    \n\t"
-        "add %3, %%"REG_a"              \n\t"
+        "movq (%1, %%"FF_REG_a"), %%mm0 \n\t"
+        "movq (%2, %%"FF_REG_a"), %%mm2 \n\t"
+        "movq (%2, %%"FF_REG_a"), %%mm4 \n\t"
+        "add %3, %%"FF_REG_a"           \n\t"
         "psubusb %%mm0, %%mm2           \n\t"
         "psubusb %%mm4, %%mm0           \n\t"
-        "movq (%1, %%"REG_a"), %%mm1    \n\t"
-        "movq (%2, %%"REG_a"), %%mm3    \n\t"
-        "movq (%2, %%"REG_a"), %%mm5    \n\t"
+        "movq (%1, %%"FF_REG_a"), %%mm1 \n\t"
+        "movq (%2, %%"FF_REG_a"), %%mm3 \n\t"
+        "movq (%2, %%"FF_REG_a"), %%mm5 \n\t"
         "psubusb %%mm1, %%mm3           \n\t"
         "psubusb %%mm5, %%mm1           \n\t"
         "por %%mm2, %%mm0               \n\t"
@@ -306,7 +306,7 @@ static inline void sad8_1_mmx(uint8_t *blk1, uint8_t *blk2,
         "paddw %%mm3, %%mm2             \n\t"
         "paddw %%mm2, %%mm0             \n\t"
         "paddw %%mm0, %%mm6             \n\t"
-        "add %3, %%"REG_a"              \n\t"
+        "add %3, %%"FF_REG_a"           \n\t"
         " js 1b                         \n\t"
         : "+a" (len)
         : "r" (blk1 - len), "r" (blk2 - len), "r" (stride));
@@ -319,18 +319,18 @@ static inline void sad8_2_mmx(uint8_t *blk1a, uint8_t *blk1b, uint8_t *blk2,
     __asm__ volatile (
         ".p2align 4                     \n\t"
         "1:                             \n\t"
-        "movq (%1, %%"REG_a"), %%mm0    \n\t"
-        "movq (%2, %%"REG_a"), %%mm1    \n\t"
-        "movq (%1, %%"REG_a"), %%mm2    \n\t"
-        "movq (%2, %%"REG_a"), %%mm3    \n\t"
+        "movq (%1, %%"FF_REG_a"), %%mm0 \n\t"
+        "movq (%2, %%"FF_REG_a"), %%mm1 \n\t"
+        "movq (%1, %%"FF_REG_a"), %%mm2 \n\t"
+        "movq (%2, %%"FF_REG_a"), %%mm3 \n\t"
         "punpcklbw %%mm7, %%mm0         \n\t"
         "punpcklbw %%mm7, %%mm1         \n\t"
         "punpckhbw %%mm7, %%mm2         \n\t"
         "punpckhbw %%mm7, %%mm3         \n\t"
         "paddw %%mm0, %%mm1             \n\t"
         "paddw %%mm2, %%mm3             \n\t"
-        "movq (%3, %%"REG_a"), %%mm4    \n\t"
-        "movq (%3, %%"REG_a"), %%mm2    \n\t"
+        "movq (%3, %%"FF_REG_a"), %%mm4 \n\t"
+        "movq (%3, %%"FF_REG_a"), %%mm2 \n\t"
         "paddw %%mm5, %%mm1             \n\t"
         "paddw %%mm5, %%mm3             \n\t"
         "psrlw $1, %%mm1                \n\t"
@@ -344,7 +344,7 @@ static inline void sad8_2_mmx(uint8_t *blk1a, uint8_t *blk1b, uint8_t *blk2,
         "punpckhbw %%mm7, %%mm1         \n\t"
         "paddw %%mm1, %%mm0             \n\t"
         "paddw %%mm0, %%mm6             \n\t"
-        "add %4, %%"REG_a"              \n\t"
+        "add %4, %%"FF_REG_a"           \n\t"
         " js 1b                         \n\t"
         : "+a" (len)
         : "r" (blk1a - len), "r" (blk1b - len), "r" (blk2 - len),
@@ -356,8 +356,8 @@ static inline void sad8_4_mmx(uint8_t *blk1, uint8_t *blk2,
 {
     x86_reg len = -stride * h;
     __asm__ volatile (
-        "movq  (%1, %%"REG_a"), %%mm0   \n\t"
-        "movq 1(%1, %%"REG_a"), %%mm2   \n\t"
+        "movq  (%1, %%"FF_REG_a"), %%mm0\n\t"
+        "movq 1(%1, %%"FF_REG_a"), %%mm2\n\t"
         "movq %%mm0, %%mm1              \n\t"
         "movq %%mm2, %%mm3              \n\t"
         "punpcklbw %%mm7, %%mm0         \n\t"
@@ -368,8 +368,8 @@ static inline void sad8_4_mmx(uint8_t *blk1, uint8_t *blk2,
         "paddw %%mm3, %%mm1             \n\t"
         ".p2align 4                     \n\t"
         "1:                             \n\t"
-        "movq  (%2, %%"REG_a"), %%mm2   \n\t"
-        "movq 1(%2, %%"REG_a"), %%mm4   \n\t"
+        "movq  (%2, %%"FF_REG_a"), %%mm2\n\t"
+        "movq 1(%2, %%"FF_REG_a"), %%mm4\n\t"
         "movq %%mm2, %%mm3              \n\t"
         "movq %%mm4, %%mm5              \n\t"
         "punpcklbw %%mm7, %%mm2         \n\t"
@@ -383,8 +383,8 @@ static inline void sad8_4_mmx(uint8_t *blk1, uint8_t *blk2,
         "paddw %%mm3, %%mm1             \n\t"
         "paddw %%mm5, %%mm0             \n\t"
         "paddw %%mm5, %%mm1             \n\t"
-        "movq (%3, %%"REG_a"), %%mm4    \n\t"
-        "movq (%3, %%"REG_a"), %%mm5    \n\t"
+        "movq (%3, %%"FF_REG_a"), %%mm4 \n\t"
+        "movq (%3, %%"FF_REG_a"), %%mm5 \n\t"
         "psrlw $2, %%mm0                \n\t"
         "psrlw $2, %%mm1                \n\t"
         "packuswb %%mm1, %%mm0          \n\t"
@@ -398,7 +398,7 @@ static inline void sad8_4_mmx(uint8_t *blk1, uint8_t *blk2,
         "paddw %%mm4, %%mm6             \n\t"
         "movq  %%mm2, %%mm0             \n\t"
         "movq  %%mm3, %%mm1             \n\t"
-        "add %4, %%"REG_a"              \n\t"
+        "add %4, %%"FF_REG_a"           \n\t"
         " js 1b                         \n\t"
         : "+a" (len)
         : "r" (blk1 - len), "r" (blk1 - len + stride), "r" (blk2 - len),
diff --git a/libavcodec/x86/mpegvideo.c b/libavcodec/x86/mpegvideo.c
index 18113265ba..35a8264804 100644
--- a/libavcodec/x86/mpegvideo.c
+++ b/libavcodec/x86/mpegvideo.c
@@ -188,13 +188,13 @@ __asm__ volatile(
                 "movd %2, %%mm6                 \n\t"
                 "packssdw %%mm6, %%mm6          \n\t"
                 "packssdw %%mm6, %%mm6          \n\t"
-                "mov %3, %%"REG_a"              \n\t"
+                "mov %3, %%"FF_REG_a"           \n\t"
                 ".p2align 4                     \n\t"
                 "1:                             \n\t"
-                "movq (%0, %%"REG_a"), %%mm0    \n\t"
-                "movq 8(%0, %%"REG_a"), %%mm1   \n\t"
-                "movq (%1, %%"REG_a"), %%mm4    \n\t"
-                "movq 8(%1, %%"REG_a"), %%mm5   \n\t"
+                "movq (%0, %%"FF_REG_a"), %%mm0 \n\t"
+                "movq 8(%0, %%"FF_REG_a"), %%mm1\n\t"
+                "movq (%1, %%"FF_REG_a"), %%mm4 \n\t"
+                "movq 8(%1, %%"FF_REG_a"), %%mm5\n\t"
                 "pmullw %%mm6, %%mm4            \n\t" // q=qscale*quant_matrix[i]
                 "pmullw %%mm6, %%mm5            \n\t" // q=qscale*quant_matrix[i]
                 "pxor %%mm2, %%mm2              \n\t"
@@ -209,8 +209,8 @@ __asm__ volatile(
                 "pmullw %%mm5, %%mm1            \n\t" // abs(block[i])*q
                 "pxor %%mm4, %%mm4              \n\t"
                 "pxor %%mm5, %%mm5              \n\t" // FIXME slow
-                "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
-                "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
+                "pcmpeqw (%0, %%"FF_REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
+                "pcmpeqw 8(%0, %%"FF_REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
                 "psraw $3, %%mm0                \n\t"
                 "psraw $3, %%mm1                \n\t"
                 "psubw %%mm7, %%mm0             \n\t"
@@ -223,13 +223,13 @@ __asm__ volatile(
                 "psubw %%mm3, %%mm1             \n\t"
                 "pandn %%mm0, %%mm4             \n\t"
                 "pandn %%mm1, %%mm5             \n\t"
-                "movq %%mm4, (%0, %%"REG_a")    \n\t"
-                "movq %%mm5, 8(%0, %%"REG_a")   \n\t"
+                "movq %%mm4, (%0, %%"FF_REG_a") \n\t"
+                "movq %%mm5, 8(%0, %%"FF_REG_a")\n\t"
 
-                "add $16, %%"REG_a"             \n\t"
+                "add $16, %%"FF_REG_a"          \n\t"
                 "js 1b                          \n\t"
                 ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "rm" (qscale), "g" (-2*nCoeffs)
-                : "%"REG_a, "memory"
+                : "%"FF_REG_a, "memory"
         );
     block[0]= block0;
 }
@@ -251,13 +251,13 @@ __asm__ volatile(
                 "movd %2, %%mm6                 \n\t"
                 "packssdw %%mm6, %%mm6          \n\t"
                 "packssdw %%mm6, %%mm6          \n\t"
-                "mov %3, %%"REG_a"              \n\t"
+                "mov %3, %%"FF_REG_a"           \n\t"
                 ".p2align 4                     \n\t"
                 "1:                             \n\t"
-                "movq (%0, %%"REG_a"), %%mm0    \n\t"
-                "movq 8(%0, %%"REG_a"), %%mm1   \n\t"
-                "movq (%1, %%"REG_a"), %%mm4    \n\t"
-                "movq 8(%1, %%"REG_a"), %%mm5   \n\t"
+                "movq (%0, %%"FF_REG_a"), %%mm0 \n\t"
+                "movq 8(%0, %%"FF_REG_a"), %%mm1\n\t"
+                "movq (%1, %%"FF_REG_a"), %%mm4 \n\t"
+                "movq 8(%1, %%"FF_REG_a"), %%mm5\n\t"
                 "pmullw %%mm6, %%mm4            \n\t" // q=qscale*quant_matrix[i]
                 "pmullw %%mm6, %%mm5            \n\t" // q=qscale*quant_matrix[i]
                 "pxor %%mm2, %%mm2              \n\t"
@@ -276,8 +276,8 @@ __asm__ volatile(
                 "pmullw %%mm5, %%mm1            \n\t" // (abs(block[i])*2 + 1)*q
                 "pxor %%mm4, %%mm4              \n\t"
                 "pxor %%mm5, %%mm5              \n\t" // FIXME slow
-                "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
-                "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
+                "pcmpeqw (%0, %%"FF_REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
+                "pcmpeqw 8(%0, %%"FF_REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
                 "psraw $4, %%mm0                \n\t"
                 "psraw $4, %%mm1                \n\t"
                 "psubw %%mm7, %%mm0             \n\t"
@@ -290,13 +290,13 @@ __asm__ volatile(
                 "psubw %%mm3, %%mm1             \n\t"
                 "pandn %%mm0, %%mm4             \n\t"
                 "pandn %%mm1, %%mm5             \n\t"
-                "movq %%mm4, (%0, %%"REG_a")    \n\t"
-                "movq %%mm5, 8(%0, %%"REG_a")   \n\t"
+                "movq %%mm4, (%0, %%"FF_REG_a") \n\t"
+                "movq %%mm5, 8(%0, %%"FF_REG_a")\n\t"
 
-                "add $16, %%"REG_a"             \n\t"
+                "add $16, %%"FF_REG_a"          \n\t"
                 "js 1b                          \n\t"
                 ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "rm" (qscale), "g" (-2*nCoeffs)
-                : "%"REG_a, "memory"
+                : "%"FF_REG_a, "memory"
         );
 }
 
@@ -326,13 +326,13 @@ __asm__ volatile(
                 "movd %2, %%mm6                 \n\t"
                 "packssdw %%mm6, %%mm6          \n\t"
                 "packssdw %%mm6, %%mm6          \n\t"
-                "mov %3, %%"REG_a"              \n\t"
+                "mov %3, %%"FF_REG_a"           \n\t"
                 ".p2align 4                     \n\t"
                 "1:                             \n\t"
-                "movq (%0, %%"REG_a"), %%mm0    \n\t"
-                "movq 8(%0, %%"REG_a"), %%mm1   \n\t"
-                "movq (%1, %%"REG_a"), %%mm4    \n\t"
-                "movq 8(%1, %%"REG_a"), %%mm5   \n\t"
+                "movq (%0, %%"FF_REG_a"), %%mm0 \n\t"
+                "movq 8(%0, %%"FF_REG_a"), %%mm1\n\t"
+                "movq (%1, %%"FF_REG_a"), %%mm4 \n\t"
+                "movq 8(%1, %%"FF_REG_a"), %%mm5\n\t"
                 "pmullw %%mm6, %%mm4            \n\t" // q=qscale*quant_matrix[i]
                 "pmullw %%mm6, %%mm5            \n\t" // q=qscale*quant_matrix[i]
                 "pxor %%mm2, %%mm2              \n\t"
@@ -347,8 +347,8 @@ __asm__ volatile(
                 "pmullw %%mm5, %%mm1            \n\t" // abs(block[i])*q
                 "pxor %%mm4, %%mm4              \n\t"
                 "pxor %%mm5, %%mm5              \n\t" // FIXME slow
-                "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
-                "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
+                "pcmpeqw (%0, %%"FF_REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
+                "pcmpeqw 8(%0, %%"FF_REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
                 "psraw $4, %%mm0                \n\t"
                 "psraw $4, %%mm1                \n\t"
                 "pxor %%mm2, %%mm0              \n\t"
@@ -357,13 +357,13 @@ __asm__ volatile(
                 "psubw %%mm3, %%mm1             \n\t"
                 "pandn %%mm0, %%mm4             \n\t"
                 "pandn %%mm1, %%mm5             \n\t"
-                "movq %%mm4, (%0, %%"REG_a")    \n\t"
-                "movq %%mm5, 8(%0, %%"REG_a")   \n\t"
+                "movq %%mm4, (%0, %%"FF_REG_a") \n\t"
+                "movq %%mm5, 8(%0, %%"FF_REG_a")\n\t"
 
-                "add $16, %%"REG_a"             \n\t"
+                "add $16, %%"FF_REG_a"          \n\t"
                 "jng 1b                         \n\t"
                 ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "rm" (qscale), "g" (-2*nCoeffs)
-                : "%"REG_a, "memory"
+                : "%"FF_REG_a, "memory"
         );
     block[0]= block0;
         //Note, we do not do mismatch control for intra as errors cannot accumulate
@@ -390,13 +390,13 @@ __asm__ volatile(
                 "movd %2, %%mm6                 \n\t"
                 "packssdw %%mm6, %%mm6          \n\t"
                 "packssdw %%mm6, %%mm6          \n\t"
-                "mov %3, %%"REG_a"              \n\t"
+                "mov %3, %%"FF_REG_a"           \n\t"
                 ".p2align 4                     \n\t"
                 "1:                             \n\t"
-                "movq (%0, %%"REG_a"), %%mm0    \n\t"
-                "movq 8(%0, %%"REG_a"), %%mm1   \n\t"
-                "movq (%1, %%"REG_a"), %%mm4    \n\t"
-                "movq 8(%1, %%"REG_a"), %%mm5   \n\t"
+                "movq (%0, %%"FF_REG_a"), %%mm0 \n\t"
+                "movq 8(%0, %%"FF_REG_a"), %%mm1\n\t"
+                "movq (%1, %%"FF_REG_a"), %%mm4 \n\t"
+                "movq 8(%1, %%"FF_REG_a"), %%mm5\n\t"
                 "pmullw %%mm6, %%mm4            \n\t" // q=qscale*quant_matrix[i]
                 "pmullw %%mm6, %%mm5            \n\t" // q=qscale*quant_matrix[i]
                 "pxor %%mm2, %%mm2              \n\t"
@@ -415,8 +415,8 @@ __asm__ volatile(
                 "paddw %%mm5, %%mm1             \n\t" // (abs(block[i])*2 + 1)*q
                 "pxor %%mm4, %%mm4              \n\t"
                 "pxor %%mm5, %%mm5              \n\t" // FIXME slow
-                "pcmpeqw (%0, %%"REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
-                "pcmpeqw 8(%0, %%"REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
+                "pcmpeqw (%0, %%"FF_REG_a"), %%mm4 \n\t" // block[i] == 0 ? -1 : 0
+                "pcmpeqw 8(%0, %%"FF_REG_a"), %%mm5\n\t" // block[i] == 0 ? -1 : 0
                 "psrlw $5, %%mm0                \n\t"
                 "psrlw $5, %%mm1                \n\t"
                 "pxor %%mm2, %%mm0              \n\t"
@@ -427,10 +427,10 @@ __asm__ volatile(
                 "pandn %%mm1, %%mm5             \n\t"
                 "pxor %%mm4, %%mm7              \n\t"
                 "pxor %%mm5, %%mm7              \n\t"
-                "movq %%mm4, (%0, %%"REG_a")    \n\t"
-                "movq %%mm5, 8(%0, %%"REG_a")   \n\t"
+                "movq %%mm4, (%0, %%"FF_REG_a") \n\t"
+                "movq %%mm5, 8(%0, %%"FF_REG_a")\n\t"
 
-                "add $16, %%"REG_a"             \n\t"
+                "add $16, %%"FF_REG_a"          \n\t"
                 "jng 1b                         \n\t"
                 "movd 124(%0, %3), %%mm0        \n\t"
                 "movq %%mm7, %%mm6              \n\t"
@@ -445,7 +445,7 @@ __asm__ volatile(
                 "movd %%mm0, 124(%0, %3)        \n\t"
 
                 ::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "rm" (qscale), "r" (-2*nCoeffs)
-                : "%"REG_a, "memory"
+                : "%"FF_REG_a, "memory"
         );
 }
 
diff --git a/libavcodec/x86/mpegvideoenc_template.c b/libavcodec/x86/mpegvideoenc_template.c
index da76459cd6..b2512744ca 100644
--- a/libavcodec/x86/mpegvideoenc_template.c
+++ b/libavcodec/x86/mpegvideoenc_template.c
@@ -150,32 +150,32 @@ static int RENAME(dct_quantize)(MpegEncContext *s,
     if((s->out_format == FMT_H263 || s->out_format == FMT_H261) && s->mpeg_quant==0){
 
         __asm__ volatile(
-            "movd %%"REG_a", "MM"3              \n\t" // last_non_zero_p1
+            "movd %%"FF_REG_a", "MM"3           \n\t" // last_non_zero_p1
             SPREADW(MM"3")
             "pxor "MM"7, "MM"7                  \n\t" // 0
             "pxor "MM"4, "MM"4                  \n\t" // 0
             MOVQ" (%2), "MM"5                   \n\t" // qmat[0]
             "pxor "MM"6, "MM"6                  \n\t"
             "psubw (%3), "MM"6                  \n\t" // -bias[0]
-            "mov $-128, %%"REG_a"               \n\t"
+            "mov $-128, %%"FF_REG_a"            \n\t"
             ".p2align 4                         \n\t"
             "1:                                 \n\t"
-            MOVQ" (%1, %%"REG_a"), "MM"0        \n\t" // block[i]
+            MOVQ" (%1, %%"FF_REG_a"), "MM"0     \n\t" // block[i]
             SAVE_SIGN(MM"1", MM"0")                   // ABS(block[i])
             "psubusw "MM"6, "MM"0               \n\t" // ABS(block[i]) + bias[0]
             "pmulhw "MM"5, "MM"0                \n\t" // (ABS(block[i])*qmat[0] - bias[0]*qmat[0])>>16
             "por "MM"0, "MM"4                   \n\t"
             RESTORE_SIGN(MM"1", MM"0")                // out=((ABS(block[i])*qmat[0] - bias[0]*qmat[0])>>16)*sign(block[i])
-            MOVQ" "MM"0, (%5, %%"REG_a")        \n\t"
+            MOVQ" "MM"0, (%5, %%"FF_REG_a")     \n\t"
             "pcmpeqw "MM"7, "MM"0               \n\t" // out==0 ? 0xFF : 0x00
-            MOVQ" (%4, %%"REG_a"), "MM"1        \n\t"
-            MOVQ" "MM"7, (%1, %%"REG_a")        \n\t" // 0
+            MOVQ" (%4, %%"FF_REG_a"), "MM"1     \n\t"
+            MOVQ" "MM"7, (%1, %%"FF_REG_a")     \n\t" // 0
             "pandn "MM"1, "MM"0                 \n\t"
             PMAXW(MM"0", MM"3")
-            "add $"MMREG_WIDTH", %%"REG_a"      \n\t"
+            "add $"MMREG_WIDTH", %%"FF_REG_a"   \n\t"
             " js 1b                             \n\t"
             PMAX(MM"3", MM"0")
-            "movd "MM"3, %%"REG_a"              \n\t"
+            "movd "MM"3, %%"FF_REG_a"           \n\t"
             "movzbl %%al, %%eax                 \n\t" // last_non_zero_p1
             : "+a" (last_non_zero_p1)
             : "r" (block+64), "r" (qmat), "r" (bias),
@@ -185,31 +185,31 @@ static int RENAME(dct_quantize)(MpegEncContext *s,
         );
     }else{ // FMT_H263
         __asm__ volatile(
-            "movd %%"REG_a", "MM"3              \n\t" // last_non_zero_p1
+            "movd %%"FF_REG_a", "MM"3           \n\t" // last_non_zero_p1
             SPREADW(MM"3")
             "pxor "MM"7, "MM"7                  \n\t" // 0
             "pxor "MM"4, "MM"4                  \n\t" // 0
-            "mov $-128, %%"REG_a"               \n\t"
+            "mov $-128, %%"FF_REG_a"            \n\t"
             ".p2align 4                         \n\t"
             "1:                                 \n\t"
-            MOVQ" (%1, %%"REG_a"), "MM"0        \n\t" // block[i]
+            MOVQ" (%1, %%"FF_REG_a"), "MM"0     \n\t" // block[i]
             SAVE_SIGN(MM"1", MM"0")                   // ABS(block[i])
-            MOVQ" (%3, %%"REG_a"), "MM"6        \n\t" // bias[0]
+            MOVQ" (%3, %%"FF_REG_a"), "MM"6     \n\t" // bias[0]
             "paddusw "MM"6, "MM"0               \n\t" // ABS(block[i]) + bias[0]
-            MOVQ" (%2, %%"REG_a"), "MM"5        \n\t" // qmat[i]
+            MOVQ" (%2, %%"FF_REG_a"), "MM"5     \n\t" // qmat[i]
             "pmulhw "MM"5, "MM"0                \n\t" // (ABS(block[i])*qmat[0] + bias[0]*qmat[0])>>16
             "por "MM"0, "MM"4                   \n\t"
             RESTORE_SIGN(MM"1", MM"0")                // out=((ABS(block[i])*qmat[0] - bias[0]*qmat[0])>>16)*sign(block[i])
-            MOVQ" "MM"0, (%5, %%"REG_a")        \n\t"
+            MOVQ" "MM"0, (%5, %%"FF_REG_a")     \n\t"
             "pcmpeqw "MM"7, "MM"0               \n\t" // out==0 ? 0xFF : 0x00
-            MOVQ" (%4, %%"REG_a"), "MM"1        \n\t"
-            MOVQ" "MM"7, (%1, %%"REG_a")        \n\t" // 0
+            MOVQ" (%4, %%"FF_REG_a"), "MM"1     \n\t"
+            MOVQ" "MM"7, (%1, %%"FF_REG_a")     \n\t" // 0
             "pandn "MM"1, "MM"0                 \n\t"
             PMAXW(MM"0", MM"3")
-            "add $"MMREG_WIDTH", %%"REG_a"      \n\t"
+            "add $"MMREG_WIDTH", %%"FF_REG_a"   \n\t"
             " js 1b                             \n\t"
             PMAX(MM"3", MM"0")
-            "movd "MM"3, %%"REG_a"              \n\t"
+            "movd "MM"3, %%"FF_REG_a"           \n\t"
             "movzbl %%al, %%eax                 \n\t" // last_non_zero_p1
             : "+a" (last_non_zero_p1)
             : "r" (block+64), "r" (qmat+64), "r" (bias+64),
diff --git a/libavcodec/x86/rnd_template.c b/libavcodec/x86/rnd_template.c
index ddca4eb590..09946bd23f 100644
--- a/libavcodec/x86/rnd_template.c
+++ b/libavcodec/x86/rnd_template.c
@@ -46,12 +46,12 @@ av_unused STATIC void DEF(put, pixels8_xy2)(uint8_t *block, const uint8_t *pixel
         "punpckhbw %%mm7, %%mm5         \n\t"
         "paddusw %%mm0, %%mm4           \n\t"
         "paddusw %%mm1, %%mm5           \n\t"
-        "xor    %%"REG_a", %%"REG_a"    \n\t"
+        "xor    %%"FF_REG_a", %%"FF_REG_a" \n\t"
         "add    %3, %1                  \n\t"
         ".p2align 3                     \n\t"
         "1:                             \n\t"
-        "movq   (%1, %%"REG_a"), %%mm0  \n\t"
-        "movq   1(%1, %%"REG_a"), %%mm2 \n\t"
+        "movq   (%1, %%"FF_REG_a"), %%mm0  \n\t"
+        "movq   1(%1, %%"FF_REG_a"), %%mm2 \n\t"
         "movq   %%mm0, %%mm1            \n\t"
         "movq   %%mm2, %%mm3            \n\t"
         "punpcklbw %%mm7, %%mm0         \n\t"
@@ -67,11 +67,11 @@ av_unused STATIC void DEF(put, pixels8_xy2)(uint8_t *block, const uint8_t *pixel
         "psrlw  $2, %%mm4               \n\t"
         "psrlw  $2, %%mm5               \n\t"
         "packuswb  %%mm5, %%mm4         \n\t"
-        "movq   %%mm4, (%2, %%"REG_a")  \n\t"
-        "add    %3, %%"REG_a"           \n\t"
+        "movq   %%mm4, (%2, %%"FF_REG_a")  \n\t"
+        "add    %3, %%"FF_REG_a"           \n\t"
 
-        "movq   (%1, %%"REG_a"), %%mm2  \n\t" // 0 <-> 2   1 <-> 3
-        "movq   1(%1, %%"REG_a"), %%mm4 \n\t"
+        "movq   (%1, %%"FF_REG_a"), %%mm2  \n\t" // 0 <-> 2   1 <-> 3
+        "movq   1(%1, %%"FF_REG_a"), %%mm4 \n\t"
         "movq   %%mm2, %%mm3            \n\t"
         "movq   %%mm4, %%mm5            \n\t"
         "punpcklbw %%mm7, %%mm2         \n\t"
@@ -87,14 +87,14 @@ av_unused STATIC void DEF(put, pixels8_xy2)(uint8_t *block, const uint8_t *pixel
         "psrlw  $2, %%mm0               \n\t"
         "psrlw  $2, %%mm1               \n\t"
         "packuswb  %%mm1, %%mm0         \n\t"
-        "movq   %%mm0, (%2, %%"REG_a")  \n\t"
-        "add    %3, %%"REG_a"           \n\t"
+        "movq   %%mm0, (%2, %%"FF_REG_a")  \n\t"
+        "add    %3, %%"FF_REG_a"        \n\t"
 
         "subl   $2, %0                  \n\t"
         "jnz    1b                      \n\t"
         :"+g"(h), "+S"(pixels)
         :"D"(block), "r"((x86_reg)line_size)
-        :REG_a, "memory");
+        :FF_REG_a, "memory");
 }
 
 // avg_pixels
@@ -115,12 +115,12 @@ av_unused STATIC void DEF(avg, pixels8_xy2)(uint8_t *block, const uint8_t *pixel
         "punpckhbw %%mm7, %%mm5         \n\t"
         "paddusw %%mm0, %%mm4           \n\t"
         "paddusw %%mm1, %%mm5           \n\t"
-        "xor    %%"REG_a", %%"REG_a"    \n\t"
+        "xor    %%"FF_REG_a", %%"FF_REG_a" \n\t"
         "add    %3, %1                  \n\t"
         ".p2align 3                     \n\t"
         "1:                             \n\t"
-        "movq   (%1, %%"REG_a"), %%mm0  \n\t"
-        "movq   1(%1, %%"REG_a"), %%mm2 \n\t"
+        "movq   (%1, %%"FF_REG_a"), %%mm0  \n\t"
+        "movq   1(%1, %%"FF_REG_a"), %%mm2 \n\t"
         "movq   %%mm0, %%mm1            \n\t"
         "movq   %%mm2, %%mm3            \n\t"
         "punpcklbw %%mm7, %%mm0         \n\t"
@@ -135,16 +135,16 @@ av_unused STATIC void DEF(avg, pixels8_xy2)(uint8_t *block, const uint8_t *pixel
         "paddusw %%mm1, %%mm5           \n\t"
         "psrlw  $2, %%mm4               \n\t"
         "psrlw  $2, %%mm5               \n\t"
-                "movq   (%2, %%"REG_a"), %%mm3  \n\t"
+                "movq   (%2, %%"FF_REG_a"), %%mm3  \n\t"
         "packuswb  %%mm5, %%mm4         \n\t"
                 "pcmpeqd %%mm2, %%mm2   \n\t"
                 "paddb %%mm2, %%mm2     \n\t"
                 PAVGB_MMX(%%mm3, %%mm4, %%mm5, %%mm2)
-                "movq   %%mm5, (%2, %%"REG_a")  \n\t"
-        "add    %3, %%"REG_a"                \n\t"
+                "movq   %%mm5, (%2, %%"FF_REG_a")  \n\t"
+        "add    %3, %%"FF_REG_a"        \n\t"
 
-        "movq   (%1, %%"REG_a"), %%mm2  \n\t" // 0 <-> 2   1 <-> 3
-        "movq   1(%1, %%"REG_a"), %%mm4 \n\t"
+        "movq   (%1, %%"FF_REG_a"), %%mm2  \n\t" // 0 <-> 2   1 <-> 3
+        "movq   1(%1, %%"FF_REG_a"), %%mm4 \n\t"
         "movq   %%mm2, %%mm3            \n\t"
         "movq   %%mm4, %%mm5            \n\t"
         "punpcklbw %%mm7, %%mm2         \n\t"
@@ -159,17 +159,17 @@ av_unused STATIC void DEF(avg, pixels8_xy2)(uint8_t *block, const uint8_t *pixel
         "paddusw %%mm5, %%mm1           \n\t"
         "psrlw  $2, %%mm0               \n\t"
         "psrlw  $2, %%mm1               \n\t"
-                "movq   (%2, %%"REG_a"), %%mm3  \n\t"
+                "movq   (%2, %%"FF_REG_a"), %%mm3  \n\t"
         "packuswb  %%mm1, %%mm0         \n\t"
                 "pcmpeqd %%mm2, %%mm2   \n\t"
                 "paddb %%mm2, %%mm2     \n\t"
                 PAVGB_MMX(%%mm3, %%mm0, %%mm1, %%mm2)
-                "movq   %%mm1, (%2, %%"REG_a")  \n\t"
-        "add    %3, %%"REG_a"           \n\t"
+                "movq   %%mm1, (%2, %%"FF_REG_a")  \n\t"
+        "add    %3, %%"FF_REG_a"           \n\t"
 
         "subl   $2, %0                  \n\t"
         "jnz    1b                      \n\t"
         :"+g"(h), "+S"(pixels)
         :"D"(block), "r"((x86_reg)line_size)
-        :REG_a, "memory");
+        :FF_REG_a, "memory");
 }
diff --git a/libavcodec/x86/snowdsp.c b/libavcodec/x86/snowdsp.c
index e2ad511d0a..218e6864db 100644
--- a/libavcodec/x86/snowdsp.c
+++ b/libavcodec/x86/snowdsp.c
@@ -390,10 +390,10 @@ static void ff_snow_horizontal_compose97i_mmx(IDWTELEM *b, IDWTELEM *temp, int w
 
 #if HAVE_7REGS
 #define snow_vertical_compose_sse2_load_add(op,r,t0,t1,t2,t3)\
-        ""op" ("r",%%"REG_d"), %%"t0"      \n\t"\
-        ""op" 16("r",%%"REG_d"), %%"t1"    \n\t"\
-        ""op" 32("r",%%"REG_d"), %%"t2"    \n\t"\
-        ""op" 48("r",%%"REG_d"), %%"t3"    \n\t"
+        ""op" ("r",%%"FF_REG_d"), %%"t0"      \n\t"\
+        ""op" 16("r",%%"FF_REG_d"), %%"t1"    \n\t"\
+        ""op" 32("r",%%"FF_REG_d"), %%"t2"    \n\t"\
+        ""op" 48("r",%%"FF_REG_d"), %%"t3"    \n\t"
 
 #define snow_vertical_compose_sse2_load(r,t0,t1,t2,t3)\
         snow_vertical_compose_sse2_load_add("movdqa",r,t0,t1,t2,t3)
@@ -408,10 +408,10 @@ static void ff_snow_horizontal_compose97i_mmx(IDWTELEM *b, IDWTELEM *temp, int w
         "psubw %%"s3", %%"t3" \n\t"
 
 #define snow_vertical_compose_sse2_store(w,s0,s1,s2,s3)\
-        "movdqa %%"s0", ("w",%%"REG_d")      \n\t"\
-        "movdqa %%"s1", 16("w",%%"REG_d")    \n\t"\
-        "movdqa %%"s2", 32("w",%%"REG_d")    \n\t"\
-        "movdqa %%"s3", 48("w",%%"REG_d")    \n\t"
+        "movdqa %%"s0", ("w",%%"FF_REG_d")    \n\t"\
+        "movdqa %%"s1", 16("w",%%"FF_REG_d")  \n\t"\
+        "movdqa %%"s2", 32("w",%%"FF_REG_d")  \n\t"\
+        "movdqa %%"s3", 48("w",%%"FF_REG_d")  \n\t"
 
 #define snow_vertical_compose_sra(n,t0,t1,t2,t3)\
         "psraw $"n", %%"t0" \n\t"\
@@ -477,14 +477,14 @@ static void ff_snow_vertical_compose97i_sse2(IDWTELEM *b0, IDWTELEM *b1, IDWTELE
         "psrlw $13, %%xmm5                           \n\t"
         "paddw %%xmm7, %%xmm5                        \n\t"
         snow_vertical_compose_r2r_add("xmm5","xmm5","xmm5","xmm5","xmm0","xmm2","xmm4","xmm6")
-        "movq   (%2,%%"REG_d"), %%xmm1        \n\t"
-        "movq  8(%2,%%"REG_d"), %%xmm3        \n\t"
+        "movq   (%2,%%"FF_REG_d"), %%xmm1            \n\t"
+        "movq  8(%2,%%"FF_REG_d"), %%xmm3            \n\t"
         "paddw %%xmm7, %%xmm1                        \n\t"
         "paddw %%xmm7, %%xmm3                        \n\t"
         "pavgw %%xmm1, %%xmm0                        \n\t"
         "pavgw %%xmm3, %%xmm2                        \n\t"
-        "movq 16(%2,%%"REG_d"), %%xmm1        \n\t"
-        "movq 24(%2,%%"REG_d"), %%xmm3        \n\t"
+        "movq 16(%2,%%"FF_REG_d"), %%xmm1            \n\t"
+        "movq 24(%2,%%"FF_REG_d"), %%xmm3            \n\t"
         "paddw %%xmm7, %%xmm1                        \n\t"
         "paddw %%xmm7, %%xmm3                        \n\t"
         "pavgw %%xmm1, %%xmm4                        \n\t"
@@ -504,17 +504,17 @@ static void ff_snow_vertical_compose97i_sse2(IDWTELEM *b0, IDWTELEM *b1, IDWTELE
         snow_vertical_compose_sse2_store("%2","xmm0","xmm2","xmm4","xmm6")
 
         "2:                                          \n\t"
-        "sub $64, %%"REG_d"                          \n\t"
+        "sub $64, %%"FF_REG_d"                       \n\t"
         "jge 1b                                      \n\t"
         :"+d"(i)
         :"r"(b0),"r"(b1),"r"(b2),"r"(b3),"r"(b4),"r"(b5));
 }
 
 #define snow_vertical_compose_mmx_load_add(op,r,t0,t1,t2,t3)\
-        ""op" ("r",%%"REG_d"), %%"t0"   \n\t"\
-        ""op" 8("r",%%"REG_d"), %%"t1"  \n\t"\
-        ""op" 16("r",%%"REG_d"), %%"t2" \n\t"\
-        ""op" 24("r",%%"REG_d"), %%"t3" \n\t"
+        ""op" ("r",%%"FF_REG_d"), %%"t0"   \n\t"\
+        ""op" 8("r",%%"FF_REG_d"), %%"t1"  \n\t"\
+        ""op" 16("r",%%"FF_REG_d"), %%"t2" \n\t"\
+        ""op" 24("r",%%"FF_REG_d"), %%"t3" \n\t"
 
 #define snow_vertical_compose_mmx_load(r,t0,t1,t2,t3)\
         snow_vertical_compose_mmx_load_add("movq",r,t0,t1,t2,t3)
@@ -523,10 +523,10 @@ static void ff_snow_vertical_compose97i_sse2(IDWTELEM *b0, IDWTELEM *b1, IDWTELE
         snow_vertical_compose_mmx_load_add("paddw",r,t0,t1,t2,t3)
 
 #define snow_vertical_compose_mmx_store(w,s0,s1,s2,s3)\
-        "movq %%"s0", ("w",%%"REG_d")   \n\t"\
-        "movq %%"s1", 8("w",%%"REG_d")  \n\t"\
-        "movq %%"s2", 16("w",%%"REG_d") \n\t"\
-        "movq %%"s3", 24("w",%%"REG_d") \n\t"
+        "movq %%"s0", ("w",%%"FF_REG_d")   \n\t"\
+        "movq %%"s1", 8("w",%%"FF_REG_d")  \n\t"\
+        "movq %%"s2", 16("w",%%"FF_REG_d") \n\t"\
+        "movq %%"s3", 24("w",%%"FF_REG_d") \n\t"
 
 #define snow_vertical_compose_mmx_move(s0,s1,s2,s3,t0,t1,t2,t3)\
         "movq %%"s0", %%"t0" \n\t"\
@@ -571,14 +571,14 @@ static void ff_snow_vertical_compose97i_mmx(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM
         "psrlw $13, %%mm5                            \n\t"
         "paddw %%mm7, %%mm5                          \n\t"
         snow_vertical_compose_r2r_add("mm5","mm5","mm5","mm5","mm0","mm2","mm4","mm6")
-        "movq   (%2,%%"REG_d"), %%mm1         \n\t"
-        "movq  8(%2,%%"REG_d"), %%mm3         \n\t"
+        "movq   (%2,%%"FF_REG_d"), %%mm1             \n\t"
+        "movq  8(%2,%%"FF_REG_d"), %%mm3             \n\t"
         "paddw %%mm7, %%mm1                          \n\t"
         "paddw %%mm7, %%mm3                          \n\t"
         "pavgw %%mm1, %%mm0                          \n\t"
         "pavgw %%mm3, %%mm2                          \n\t"
-        "movq 16(%2,%%"REG_d"), %%mm1         \n\t"
-        "movq 24(%2,%%"REG_d"), %%mm3         \n\t"
+        "movq 16(%2,%%"FF_REG_d"), %%mm1             \n\t"
+        "movq 24(%2,%%"FF_REG_d"), %%mm3             \n\t"
         "paddw %%mm7, %%mm1                          \n\t"
         "paddw %%mm7, %%mm3                          \n\t"
         "pavgw %%mm1, %%mm4                          \n\t"
@@ -598,7 +598,7 @@ static void ff_snow_vertical_compose97i_mmx(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM
         snow_vertical_compose_mmx_store("%2","mm0","mm2","mm4","mm6")
 
         "2:                                          \n\t"
-        "sub $32, %%"REG_d"                          \n\t"
+        "sub $32, %%"FF_REG_d"                       \n\t"
         "jge 1b                                      \n\t"
         :"+d"(i)
         :"r"(b0),"r"(b1),"r"(b2),"r"(b3),"r"(b4),"r"(b5));
@@ -610,39 +610,39 @@ static void ff_snow_vertical_compose97i_mmx(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM
     IDWTELEM * * dst_array = sb->line + src_y;\
     x86_reg tmp;\
     __asm__ volatile(\
-             "mov  %7, %%"REG_c"             \n\t"\
+             "mov  %7, %%"FF_REG_c"          \n\t"\
              "mov  %6, %2                    \n\t"\
-             "mov  %4, %%"REG_S"             \n\t"\
+             "mov  %4, %%"FF_REG_S"          \n\t"\
              "pxor %%xmm7, %%xmm7            \n\t" /* 0 */\
              "pcmpeqd %%xmm3, %%xmm3         \n\t"\
              "psllw $15, %%xmm3              \n\t"\
              "psrlw $12, %%xmm3              \n\t" /* FRAC_BITS >> 1 */\
              "1:                             \n\t"\
-             "mov %1, %%"REG_D"              \n\t"\
-             "mov (%%"REG_D"), %%"REG_D"     \n\t"\
-             "add %3, %%"REG_D"              \n\t"
+             "mov %1, %%"FF_REG_D"           \n\t"\
+             "mov (%%"FF_REG_D"), %%"FF_REG_D" \n\t"\
+             "add %3, %%"FF_REG_D"           \n\t"
 
 #define snow_inner_add_yblock_sse2_start_8(out_reg1, out_reg2, ptr_offset, s_offset)\
-             "mov "PTR_SIZE"*"ptr_offset"(%%"REG_a"), %%"REG_d"; \n\t"\
-             "movq (%%"REG_d"), %%"out_reg1" \n\t"\
-             "movq (%%"REG_d", %%"REG_c"), %%"out_reg2" \n\t"\
+             "mov "FF_PTR_SIZE"*"ptr_offset"(%%"FF_REG_a"), %%"FF_REG_d"; \n\t"\
+             "movq (%%"FF_REG_d"), %%"out_reg1"                           \n\t"\
+             "movq (%%"FF_REG_d", %%"FF_REG_c"), %%"out_reg2"             \n\t"\
              "punpcklbw %%xmm7, %%"out_reg1" \n\t"\
              "punpcklbw %%xmm7, %%"out_reg2" \n\t"\
-             "movq "s_offset"(%%"REG_S"), %%xmm0 \n\t"\
-             "movq "s_offset"+16(%%"REG_S"), %%xmm4 \n\t"\
+             "movq "s_offset"(%%"FF_REG_S"), %%xmm0    \n\t"\
+             "movq "s_offset"+16(%%"FF_REG_S"), %%xmm4 \n\t"\
              "punpcklbw %%xmm7, %%xmm0       \n\t"\
              "punpcklbw %%xmm7, %%xmm4       \n\t"\
              "pmullw %%xmm0, %%"out_reg1"    \n\t"\
              "pmullw %%xmm4, %%"out_reg2"    \n\t"
 
 #define snow_inner_add_yblock_sse2_start_16(out_reg1, out_reg2, ptr_offset, s_offset)\
-             "mov "PTR_SIZE"*"ptr_offset"(%%"REG_a"), %%"REG_d"; \n\t"\
-             "movq (%%"REG_d"), %%"out_reg1" \n\t"\
-             "movq 8(%%"REG_d"), %%"out_reg2" \n\t"\
+             "mov "FF_PTR_SIZE"*"ptr_offset"(%%"FF_REG_a"), %%"FF_REG_d"; \n\t"\
+             "movq (%%"FF_REG_d"), %%"out_reg1"                           \n\t"\
+             "movq 8(%%"FF_REG_d"), %%"out_reg2"                          \n\t"\
              "punpcklbw %%xmm7, %%"out_reg1" \n\t"\
              "punpcklbw %%xmm7, %%"out_reg2" \n\t"\
-             "movq "s_offset"(%%"REG_S"), %%xmm0 \n\t"\
-             "movq "s_offset"+8(%%"REG_S"), %%xmm4 \n\t"\
+             "movq "s_offset"(%%"FF_REG_S"), %%xmm0   \n\t"\
+             "movq "s_offset"+8(%%"FF_REG_S"), %%xmm4 \n\t"\
              "punpcklbw %%xmm7, %%xmm0       \n\t"\
              "punpcklbw %%xmm7, %%xmm4       \n\t"\
              "pmullw %%xmm0, %%"out_reg1"    \n\t"\
@@ -659,12 +659,12 @@ static void ff_snow_vertical_compose97i_mmx(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM
              "paddusw %%xmm6, %%xmm5         \n\t"
 
 #define snow_inner_add_yblock_sse2_end_common1\
-             "add $32, %%"REG_S"             \n\t"\
-             "add %%"REG_c", %0              \n\t"\
-             "add %%"REG_c", "PTR_SIZE"*3(%%"REG_a");\n\t"\
-             "add %%"REG_c", "PTR_SIZE"*2(%%"REG_a");\n\t"\
-             "add %%"REG_c", "PTR_SIZE"*1(%%"REG_a");\n\t"\
-             "add %%"REG_c", (%%"REG_a")     \n\t"
+             "add $32, %%"FF_REG_S"                            \n\t"\
+             "add %%"FF_REG_c", %0                             \n\t"\
+             "add %%"FF_REG_c", "FF_PTR_SIZE"*3(%%"FF_REG_a"); \n\t"\
+             "add %%"FF_REG_c", "FF_PTR_SIZE"*2(%%"FF_REG_a"); \n\t"\
+             "add %%"FF_REG_c", "FF_PTR_SIZE"*1(%%"FF_REG_a"); \n\t"\
+             "add %%"FF_REG_c", (%%"FF_REG_a")                 \n\t"
 
 #define snow_inner_add_yblock_sse2_end_common2\
              "jnz 1b                         \n\t"\
@@ -672,18 +672,18 @@ static void ff_snow_vertical_compose97i_mmx(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM
              :\
              "rm"((x86_reg)(src_x<<1)),"m"(obmc),"a"(block),"m"(b_h),"m"(src_stride):\
              XMM_CLOBBERS("%xmm0", "%xmm1", "%xmm2", "%xmm3", "%xmm4", "%xmm5", "%xmm6", "%xmm7", )\
-             "%"REG_c"","%"REG_S"","%"REG_D"","%"REG_d"");
+             "%"FF_REG_c"","%"FF_REG_S"","%"FF_REG_D"","%"FF_REG_d"");
 
 #define snow_inner_add_yblock_sse2_end_8\
-             "sal $1, %%"REG_c"              \n\t"\
-             "add"OPSIZE" $"PTR_SIZE"*2, %1  \n\t"\
+             "sal $1, %%"FF_REG_c"                \n\t"\
+             "add"FF_OPSIZE" $"FF_PTR_SIZE"*2, %1 \n\t"\
              snow_inner_add_yblock_sse2_end_common1\
-             "sar $1, %%"REG_c"              \n\t"\
+             "sar $1, %%"FF_REG_c"           \n\t"\
              "sub $2, %2                     \n\t"\
              snow_inner_add_yblock_sse2_end_common2
 
 #define snow_inner_add_yblock_sse2_end_16\
-             "add"OPSIZE" $"PTR_SIZE"*1, %1  \n\t"\
+             "add"FF_OPSIZE" $"FF_PTR_SIZE"*1, %1 \n\t"\
              snow_inner_add_yblock_sse2_end_common1\
              "dec %2                         \n\t"\
              snow_inner_add_yblock_sse2_end_common2
@@ -696,28 +696,28 @@ snow_inner_add_yblock_sse2_accum_8("2", "8")
 snow_inner_add_yblock_sse2_accum_8("1", "128")
 snow_inner_add_yblock_sse2_accum_8("0", "136")
 
-             "mov %0, %%"REG_d"              \n\t"
-             "movdqa (%%"REG_D"), %%xmm0     \n\t"
+             "mov %0, %%"FF_REG_d"           \n\t"
+             "movdqa (%%"FF_REG_D"), %%xmm0  \n\t"
              "movdqa %%xmm1, %%xmm2          \n\t"
 
              "punpckhwd %%xmm7, %%xmm1       \n\t"
              "punpcklwd %%xmm7, %%xmm2       \n\t"
              "paddd %%xmm2, %%xmm0           \n\t"
-             "movdqa 16(%%"REG_D"), %%xmm2   \n\t"
+             "movdqa 16(%%"FF_REG_D"), %%xmm2\n\t"
              "paddd %%xmm1, %%xmm2           \n\t"
              "paddd %%xmm3, %%xmm0           \n\t"
              "paddd %%xmm3, %%xmm2           \n\t"
 
-             "mov %1, %%"REG_D"              \n\t"
-             "mov "PTR_SIZE"(%%"REG_D"), %%"REG_D";\n\t"
-             "add %3, %%"REG_D"              \n\t"
+             "mov %1, %%"FF_REG_D"           \n\t"
+             "mov "FF_PTR_SIZE"(%%"FF_REG_D"), %%"FF_REG_D"; \n\t"
+             "add %3, %%"FF_REG_D"           \n\t"
 
-             "movdqa (%%"REG_D"), %%xmm4     \n\t"
+             "movdqa (%%"FF_REG_D"), %%xmm4  \n\t"
              "movdqa %%xmm5, %%xmm6          \n\t"
              "punpckhwd %%xmm7, %%xmm5       \n\t"
              "punpcklwd %%xmm7, %%xmm6       \n\t"
              "paddd %%xmm6, %%xmm4           \n\t"
-             "movdqa 16(%%"REG_D"), %%xmm6   \n\t"
+             "movdqa 16(%%"FF_REG_D"), %%xmm6\n\t"
              "paddd %%xmm5, %%xmm6           \n\t"
              "paddd %%xmm3, %%xmm4           \n\t"
              "paddd %%xmm3, %%xmm6           \n\t"
@@ -726,13 +726,13 @@ snow_inner_add_yblock_sse2_accum_8("0", "136")
              "psrad $8, %%xmm2               \n\t" /* FRAC_BITS. */
              "packssdw %%xmm2, %%xmm0        \n\t"
              "packuswb %%xmm7, %%xmm0        \n\t"
-             "movq %%xmm0, (%%"REG_d")       \n\t"
+             "movq %%xmm0, (%%"FF_REG_d")    \n\t"
 
              "psrad $8, %%xmm4               \n\t" /* FRAC_BITS. */
              "psrad $8, %%xmm6               \n\t" /* FRAC_BITS. */
              "packssdw %%xmm6, %%xmm4        \n\t"
              "packuswb %%xmm7, %%xmm4        \n\t"
-             "movq %%xmm4, (%%"REG_d",%%"REG_c");\n\t"
+             "movq %%xmm4, (%%"FF_REG_d",%%"FF_REG_c"); \n\t"
 snow_inner_add_yblock_sse2_end_8
 }
 
@@ -744,18 +744,18 @@ snow_inner_add_yblock_sse2_accum_16("2", "16")
 snow_inner_add_yblock_sse2_accum_16("1", "512")
 snow_inner_add_yblock_sse2_accum_16("0", "528")
 
-             "mov %0, %%"REG_d"              \n\t"
+             "mov %0, %%"FF_REG_d"           \n\t"
              "psrlw $4, %%xmm1               \n\t"
              "psrlw $4, %%xmm5               \n\t"
-             "paddw   (%%"REG_D"), %%xmm1    \n\t"
-             "paddw 16(%%"REG_D"), %%xmm5    \n\t"
+             "paddw   (%%"FF_REG_D"), %%xmm1 \n\t"
+             "paddw 16(%%"FF_REG_D"), %%xmm5 \n\t"
              "paddw %%xmm3, %%xmm1           \n\t"
              "paddw %%xmm3, %%xmm5           \n\t"
              "psraw $4, %%xmm1               \n\t" /* FRAC_BITS. */
              "psraw $4, %%xmm5               \n\t" /* FRAC_BITS. */
              "packuswb %%xmm5, %%xmm1        \n\t"
 
-             "movdqu %%xmm1, (%%"REG_d")       \n\t"
+             "movdqu %%xmm1, (%%"FF_REG_d")  \n\t"
 
 snow_inner_add_yblock_sse2_end_16
 }
@@ -764,30 +764,30 @@ snow_inner_add_yblock_sse2_end_16
     IDWTELEM * * dst_array = sb->line + src_y;\
     x86_reg tmp;\
     __asm__ volatile(\
-             "mov  %7, %%"REG_c"             \n\t"\
+             "mov  %7, %%"FF_REG_c"          \n\t"\
              "mov  %6, %2                    \n\t"\
-             "mov  %4, %%"REG_S"             \n\t"\
+             "mov  %4, %%"FF_REG_S"          \n\t"\
              "pxor %%mm7, %%mm7              \n\t" /* 0 */\
              "pcmpeqd %%mm3, %%mm3           \n\t"\
              "psllw $15, %%mm3               \n\t"\
              "psrlw $12, %%mm3               \n\t" /* FRAC_BITS >> 1 */\
              "1:                             \n\t"\
-             "mov %1, %%"REG_D"              \n\t"\
-             "mov (%%"REG_D"), %%"REG_D"     \n\t"\
-             "add %3, %%"REG_D"              \n\t"
+             "mov %1, %%"FF_REG_D"           \n\t"\
+             "mov (%%"FF_REG_D"), %%"FF_REG_D" \n\t"\
+             "add %3, %%"FF_REG_D"           \n\t"
 
 #define snow_inner_add_yblock_mmx_start(out_reg1, out_reg2, ptr_offset, s_offset, d_offset)\
-             "mov "PTR_SIZE"*"ptr_offset"(%%"REG_a"), %%"REG_d"; \n\t"\
-             "movd "d_offset"(%%"REG_d"), %%"out_reg1" \n\t"\
-             "movd "d_offset"+4(%%"REG_d"), %%"out_reg2" \n\t"\
+             "mov "FF_PTR_SIZE"*"ptr_offset"(%%"FF_REG_a"), %%"FF_REG_d"; \n\t"\
+             "movd "d_offset"(%%"FF_REG_d"), %%"out_reg1"                 \n\t"\
+             "movd "d_offset"+4(%%"FF_REG_d"), %%"out_reg2"               \n\t"\
              "punpcklbw %%mm7, %%"out_reg1" \n\t"\
              "punpcklbw %%mm7, %%"out_reg2" \n\t"\
-             "movd "s_offset"(%%"REG_S"), %%mm0 \n\t"\
-             "movd "s_offset"+4(%%"REG_S"), %%mm4 \n\t"\
+             "movd "s_offset"(%%"FF_REG_S"), %%mm0   \n\t"\
+             "movd "s_offset"+4(%%"FF_REG_S"), %%mm4 \n\t"\
              "punpcklbw %%mm7, %%mm0       \n\t"\
              "punpcklbw %%mm7, %%mm4       \n\t"\
-             "pmullw %%mm0, %%"out_reg1"    \n\t"\
-             "pmullw %%mm4, %%"out_reg2"    \n\t"
+             "pmullw %%mm0, %%"out_reg1"   \n\t"\
+             "pmullw %%mm4, %%"out_reg2"   \n\t"
 
 #define snow_inner_add_yblock_mmx_accum(ptr_offset, s_offset, d_offset) \
              snow_inner_add_yblock_mmx_start("mm2", "mm6", ptr_offset, s_offset, d_offset)\
@@ -795,32 +795,32 @@ snow_inner_add_yblock_sse2_end_16
              "paddusw %%mm6, %%mm5         \n\t"
 
 #define snow_inner_add_yblock_mmx_mix(read_offset, write_offset)\
-             "mov %0, %%"REG_d"              \n\t"\
+             "mov %0, %%"FF_REG_d"           \n\t"\
              "psrlw $4, %%mm1                \n\t"\
              "psrlw $4, %%mm5                \n\t"\
-             "paddw "read_offset"(%%"REG_D"), %%mm1 \n\t"\
-             "paddw "read_offset"+8(%%"REG_D"), %%mm5 \n\t"\
+             "paddw "read_offset"(%%"FF_REG_D"), %%mm1   \n\t"\
+             "paddw "read_offset"+8(%%"FF_REG_D"), %%mm5 \n\t"\
              "paddw %%mm3, %%mm1             \n\t"\
              "paddw %%mm3, %%mm5             \n\t"\
              "psraw $4, %%mm1                \n\t"\
              "psraw $4, %%mm5                \n\t"\
              "packuswb %%mm5, %%mm1          \n\t"\
-             "movq %%mm1, "write_offset"(%%"REG_d") \n\t"
+             "movq %%mm1, "write_offset"(%%"FF_REG_d") \n\t"
 
 #define snow_inner_add_yblock_mmx_end(s_step)\
-             "add $"s_step", %%"REG_S"             \n\t"\
-             "add %%"REG_c", "PTR_SIZE"*3(%%"REG_a");\n\t"\
-             "add %%"REG_c", "PTR_SIZE"*2(%%"REG_a");\n\t"\
-             "add %%"REG_c", "PTR_SIZE"*1(%%"REG_a");\n\t"\
-             "add %%"REG_c", (%%"REG_a")     \n\t"\
-             "add"OPSIZE " $"PTR_SIZE"*1, %1 \n\t"\
-             "add %%"REG_c", %0              \n\t"\
+             "add $"s_step", %%"FF_REG_S"                      \n\t"\
+             "add %%"FF_REG_c", "FF_PTR_SIZE"*3(%%"FF_REG_a"); \n\t"\
+             "add %%"FF_REG_c", "FF_PTR_SIZE"*2(%%"FF_REG_a"); \n\t"\
+             "add %%"FF_REG_c", "FF_PTR_SIZE"*1(%%"FF_REG_a"); \n\t"\
+             "add %%"FF_REG_c", (%%"FF_REG_a")                 \n\t"\
+             "add"FF_OPSIZE " $"FF_PTR_SIZE"*1, %1             \n\t"\
+             "add %%"FF_REG_c", %0                             \n\t"\
              "dec %2                         \n\t"\
              "jnz 1b                         \n\t"\
              :"+m"(dst8),"+m"(dst_array),"=&r"(tmp)\
              :\
              "rm"((x86_reg)(src_x<<1)),"m"(obmc),"a"(block),"m"(b_h),"m"(src_stride):\
-             "%"REG_c"","%"REG_S"","%"REG_D"","%"REG_d"");
+             "%"FF_REG_c"","%"FF_REG_S"","%"FF_REG_D"","%"FF_REG_d"");
 
 static void inner_add_yblock_bw_8_obmc_16_mmx(const uint8_t *obmc, const x86_reg obmc_stride, uint8_t * * block, int b_w, x86_reg b_h,
                       int src_x, int src_y, x86_reg src_stride, slice_buffer * sb, int add, uint8_t * dst8){
diff --git a/libavcodec/x86/vc1dsp_mmx.c b/libavcodec/x86/vc1dsp_mmx.c
index da32a3ee34..45c8a68f29 100644
--- a/libavcodec/x86/vc1dsp_mmx.c
+++ b/libavcodec/x86/vc1dsp_mmx.c
@@ -84,7 +84,7 @@ static void OPNAME ## vc1_shift2_mmx(uint8_t *dst, const uint8_t *src,\
 {\
     rnd = 8-rnd;\
     __asm__ volatile(\
-        "mov       $8, %%"REG_c"           \n\t"\
+        "mov       $8, %%"FF_REG_c"        \n\t"\
         LOAD_ROUNDER_MMX("%5")\
         "movq      "MANGLE(ff_pw_9)", %%mm6\n\t"\
         "1:                                \n\t"\
@@ -119,13 +119,13 @@ static void OPNAME ## vc1_shift2_mmx(uint8_t *dst, const uint8_t *src,\
         "movq      %%mm3, (%1)             \n\t"\
         "add       %6, %0                  \n\t"\
         "add       %4, %1                  \n\t"\
-        "dec       %%"REG_c"               \n\t"\
+        "dec       %%"FF_REG_c"            \n\t"\
         "jnz 1b                            \n\t"\
         : "+r"(src),  "+r"(dst)\
         : "r"(offset), "r"(-2*offset), "g"(stride), "m"(rnd),\
           "g"(stride-offset)\
           NAMED_CONSTRAINTS_ADD(ff_pw_9)\
-        : "%"REG_c, "memory"\
+        : "%"FF_REG_c, "memory"\
     );\
 }