Chiebot-Mirror
/
FFmpeg
8
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Merge commit '38675229a879aa5258a8c71891fc8cbf74cf139f'

Browse Source 
* commit '38675229a879aa5258a8c71891fc8cbf74cf139f':
  dsputil_mmx: K&R formatting cosmetics

Conflicts:
	libavcodec/x86/dsputil_mmx.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>
pull/293/head
Michael Niedermayer 11 years ago
parent 55f53f6c29 38675229a8
commit d926c4b240
1 changed files with 50 additions and 57 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Unified View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 55
      libavcodec/x86/dsputil_mmx.c

55
libavcodec/x86/dsputil_mmx.c
Unescape View File

@ -86,7 +86,8 @@ void ff_put_pixels_clamped_mmx(const int16_t *block, uint8_t *pixels,
"movq %%mm2, (%0, %1) \n\t" "movq %%mm2, (%0, %1) \n\t"
"movq %%mm4, (%0, %1, 2) \n\t" "movq %%mm4, (%0, %1, 2) \n\t"
"movq %%mm6, (%0, %2) \n\t" "movq %%mm6, (%0, %2) \n\t"
:: "r"(pix), "r"((x86_reg)line_size), "r"((x86_reg)line_size * 3), "r"(p) :: "r" (pix), "r" ((x86_reg) line_size), "r" ((x86_reg) line_size * 3),
"r" (p)
: "memory"); : "memory");
} }
@ -182,8 +183,7 @@ void name(int16_t *blocks) \
"js 1b \n\t" \ "js 1b \n\t" \
:: "r"(((uint8_t *) blocks) + 128 * n), \ :: "r"(((uint8_t *) blocks) + 128 * n), \
"i"(-128 * n) \ "i"(-128 * n) \
: "%"REG_a \ : "%"REG_a); \
); \
} }
CLEAR_BLOCKS(ff_clear_blocks_mmx, 6) CLEAR_BLOCKS(ff_clear_blocks_mmx, 6)
CLEAR_BLOCKS(ff_clear_block_mmx, 1) CLEAR_BLOCKS(ff_clear_block_mmx, 1)
@ -201,8 +201,7 @@ void ff_clear_block_sse(int16_t *block)
"movaps %%xmm0, 96(%0) \n" "movaps %%xmm0, 96(%0) \n"
"movaps %%xmm0, 112(%0) \n" "movaps %%xmm0, 112(%0) \n"
:: "r" (block) :: "r" (block)
: "memory" : "memory");
);
} }
void ff_clear_blocks_sse(int16_t *blocks) void ff_clear_blocks_sse(int16_t *blocks)
@ -221,15 +220,14 @@ void ff_clear_blocks_sse(int16_t *blocks)
"movaps %%xmm0, 112(%0, %%"REG_a") \n" "movaps %%xmm0, 112(%0, %%"REG_a") \n"
"add $128, %%"REG_a" \n" "add $128, %%"REG_a" \n"
"js 1b \n" "js 1b \n"
:: "r"(((uint8_t *)blocks) + 128 * 6), :: "r"(((uint8_t *) blocks) + 128 * 6), "i"(-128 * 6)
"i"(-128 * 6) : "%"REG_a);
: "%"REG_a
);
} }
void ff_add_bytes_mmx(uint8_t *dst, uint8_t *src, int w) void ff_add_bytes_mmx(uint8_t *dst, uint8_t *src, int w)
{ {
x86_reg i = 0; x86_reg i = 0;
__asm__ volatile ( __asm__ volatile (
"jmp 2f \n\t" "jmp 2f \n\t"
"1: \n\t" "1: \n\t"
@ -246,8 +244,8 @@ void ff_add_bytes_mmx(uint8_t *dst, uint8_t *src, int w)
"cmp %3, %0 \n\t" "cmp %3, %0 \n\t"
"js 1b \n\t" "js 1b \n\t"
: "+r" (i) : "+r" (i)
: "r"(src), "r"(dst), "r"((x86_reg)w - 15) : "r" (src), "r" (dst), "r" ((x86_reg) w - 15));
);
for (; i < w; i++) for (; i < w; i++)
dst[i + 0] += src[i + 0]; dst[i + 0] += src[i + 0];
} }
@ -280,8 +278,8 @@ void ff_draw_edges_mmx(uint8_t *buf, int wrap, int width, int height,
"cmp %3, %0 \n\t" "cmp %3, %0 \n\t"
"jb 1b \n\t" "jb 1b \n\t"
: "+r" (ptr) : "+r" (ptr)
: "r"((x86_reg)wrap), "r"((x86_reg)width), "r"(ptr + wrap * height) : "r" ((x86_reg) wrap), "r" ((x86_reg) width),
); "r" (ptr + wrap * height));
} else if (w == 16) { } else if (w == 16) {
__asm__ volatile ( __asm__ volatile (
"1: \n\t" "1: \n\t"
@ -320,8 +318,8 @@ void ff_draw_edges_mmx(uint8_t *buf, int wrap, int width, int height,
"cmp %3, %0 \n\t" "cmp %3, %0 \n\t"
"jb 1b \n\t" "jb 1b \n\t"
: "+r" (ptr) : "+r" (ptr)
: "r"((x86_reg)wrap), "r"((x86_reg)width), "r"(ptr + wrap * height) : "r" ((x86_reg) wrap), "r" ((x86_reg) width),
); "r" (ptr + wrap * height));
} }
/* top and bottom (and hopefully also the corners) */ /* top and bottom (and hopefully also the corners) */
@ -339,9 +337,9 @@ void ff_draw_edges_mmx(uint8_t *buf, int wrap, int width, int height,
"cmp %4, %0 \n\t" "cmp %4, %0 \n\t"
"jb 1b \n\t" "jb 1b \n\t"
: "+r" (ptr) : "+r" (ptr)
: "r"((x86_reg)buf - (x86_reg)ptr - w), "r"((x86_reg) -wrap), : "r" ((x86_reg) buf - (x86_reg) ptr - w),
"r"((x86_reg) -wrap * 3), "r"(ptr + width + 2 * w) "r" ((x86_reg) - wrap), "r" ((x86_reg) - wrap * 3),
); "r" (ptr + width + 2 * w));
} }
} }
@ -361,8 +359,7 @@ void ff_draw_edges_mmx(uint8_t *buf, int wrap, int width, int height,
: "+r" (ptr) : "+r" (ptr)
: "r" ((x86_reg) last_line - (x86_reg) ptr - w), : "r" ((x86_reg) last_line - (x86_reg) ptr - w),
"r" ((x86_reg) wrap), "r" ((x86_reg) wrap * 3), "r" ((x86_reg) wrap), "r" ((x86_reg) wrap * 3),
"r"(ptr + width + 2 * w) "r" (ptr + width + 2 * w));
);
} }
} }
} }
@ -406,10 +403,10 @@ static av_always_inline void gmc(uint8_t *dst, uint8_t *src,
if ( // non-constant fullpel offset (3% of blocks) if ( // non-constant fullpel offset (3% of blocks)
((ox ^ (ox + dxw)) | (ox ^ (ox + dxh)) | (ox ^ (ox + dxw + dxh)) | ((ox ^ (ox + dxw)) | (ox ^ (ox + dxh)) | (ox ^ (ox + dxw + dxh)) |
(oy ^ (oy + dyw)) | (oy ^ (oy + dyh)) | (oy ^ (oy + dyw + dyh))) >> (16 + shift) (oy ^ (oy + dyw)) | (oy ^ (oy + dyh)) | (oy ^ (oy + dyw + dyh))) >> (16 + shift) ||
// uses more than 16 bits of subpel mv (only at huge resolution) // uses more than 16 bits of subpel mv (only at huge resolution)
|| (dxx | dxy | dyx | dyy) & 15 (dxx | dxy | dyx | dyy) & 15 ||
|| (need_emu && (h > MAX_H || stride > MAX_STRIDE))) { (need_emu && (h > MAX_H || stride > MAX_STRIDE))) {
// FIXME could still use mmx for some of the rows // FIXME could still use mmx for some of the rows
ff_gmc_c(dst, src, stride, h, ox, oy, dxx, dxy, dyx, dyy, ff_gmc_c(dst, src, stride, h, ox, oy, dxx, dxy, dyx, dyy,
shift, r, width, height); shift, r, width, height);
@ -427,8 +424,7 @@ static av_always_inline void gmc(uint8_t *dst, uint8_t *src,
"pxor %%mm7, %%mm7 \n\t" "pxor %%mm7, %%mm7 \n\t"
"punpcklwd %%mm6, %%mm6 \n\t" "punpcklwd %%mm6, %%mm6 \n\t"
"punpcklwd %%mm6, %%mm6 \n\t" "punpcklwd %%mm6, %%mm6 \n\t"
:: "r"(1<<shift) :: "r" (1 << shift));
);
for (x = 0; x < w; x += 4) { for (x = 0; x < w; x += 4) {
uint16_t dx4[4] = { oxs - dxys + dxxs * (x + 0), uint16_t dx4[4] = { oxs - dxys + dxxs * (x + 0),
@ -451,8 +447,7 @@ static av_always_inline void gmc(uint8_t *dst, uint8_t *src,
"psrlw $12, %%mm4 \n\t" "psrlw $12, %%mm4 \n\t"
"psrlw $12, %%mm5 \n\t" "psrlw $12, %%mm5 \n\t"
: "+m" (*dx4), "+m" (*dy4) : "+m" (*dx4), "+m" (*dy4)
: "m"(*dxy4), "m"(*dyy4) : "m" (*dxy4), "m" (*dyy4));
);
__asm__ volatile ( __asm__ volatile (
"movq %%mm6, %%mm2 \n\t" "movq %%mm6, %%mm2 \n\t"
@ -491,8 +486,7 @@ static av_always_inline void gmc(uint8_t *dst, uint8_t *src,
: "=m" (dst[x + y * stride]) : "=m" (dst[x + y * stride])
: "m" (src[0]), "m" (src[1]), : "m" (src[0]), "m" (src[1]),
"m" (src[stride]), "m" (src[stride + 1]), "m" (src[stride]), "m" (src[stride + 1]),
"m"(*r4), "m"(shift2) "m" (*r4), "m" (shift2));
);
src += stride; src += stride;
} }
src += 4 - h * stride; src += 4 - h * stride;
@ -631,8 +625,7 @@ void ff_vector_clipf_sse(float *dst, const float *src,
"jge 1b \n\t" "jge 1b \n\t"
: "+&r" (i) : "+&r" (i)
: "r" (dst), "r" (src), "m" (min), "m" (max) : "r" (dst), "r" (src), "m" (min), "m" (max)
: "memory" : "memory");
);
} }
#endif /* HAVE_INLINE_ASM */ #endif /* HAVE_INLINE_ASM */

Write Preview
Loading…
Cancel
Save