pixman: Branch 'master' - 6 commits
Søren Sandmann Pedersen
sandmann at kemper.freedesktop.org
Fri Nov 16 14:10:55 PST 2012
pixman/pixman-arm-simd.c | 2
pixman/pixman-fast-path.c | 4
pixman/pixman-mips-dspr2-asm.S | 424 +++++++++++++++++++++++++++++++++++++++++
pixman/pixman-mips-dspr2.c | 29 ++
pixman/pixman-mips-dspr2.h | 82 +++++++
pixman/pixman-mmx.c | 60 ++---
pixman/pixman-sse2.c | 120 +++++------
test/glyph-test.c | 4
test/utils.c | 6
9 files changed, 634 insertions(+), 97 deletions(-)
New commits:
commit 44dd746bb68625b2f6be77c3f80292b45defe9d7
Author: Stefan Weil <sw at weilnetz.de>
Date: Tue Nov 13 19:44:44 2012 +0100
test: Fix compiler warnings caused by unused code
Signed-off-by: Stefan Weil <sw at weilnetz.de>
diff --git a/test/utils.c b/test/utils.c
index ef0a171..c887a6d 100644
--- a/test/utils.c
+++ b/test/utils.c
@@ -694,6 +694,8 @@ get_random_seed (void)
return lcg_rand_u32 ();
}
+#ifdef HAVE_SIGACTION
+#ifdef HAVE_ALARM
static const char *global_msg;
static void
@@ -702,6 +704,8 @@ on_alarm (int signo)
printf ("%s\n", global_msg);
exit (1);
}
+#endif
+#endif
void
fail_after (int seconds, const char *msg)
commit 5f96022d3bca15050958512f1c15a0067d2225af
Author: Stefan Weil <sw at weilnetz.de>
Date: Tue Nov 13 19:38:32 2012 +0100
pixman: Use uintptr_t in type casts from pointer to integral value
These modifications fix lots of compiler warnings for systems where
sizeof(unsigned long) != sizeof(void *).
This is especially true for MinGW-w64 (64 bit Windows).
Signed-off-by: Stefan Weil <sw at weilnetz.de>
diff --git a/pixman/pixman-arm-simd.c b/pixman/pixman-arm-simd.c
index 3d19bfa..94f9a0c 100644
--- a/pixman/pixman-arm-simd.c
+++ b/pixman/pixman-arm-simd.c
@@ -56,7 +56,7 @@ pixman_composite_add_8_8_asm_armv6 (int32_t width,
/* ensure both src and dst are properly aligned before doing 32 bit reads
* we'll stay in this loop if src and dst have differing alignments
*/
- while (w && (((unsigned long)dst & 3) || ((unsigned long)src & 3)))
+ while (w && (((uintptr_t)dst & 3) || ((uintptr_t)src & 3)))
{
s = *src;
d = *dst;
diff --git a/pixman/pixman-fast-path.c b/pixman/pixman-fast-path.c
index d95cb4d..3429758 100644
--- a/pixman/pixman-fast-path.c
+++ b/pixman/pixman-fast-path.c
@@ -35,7 +35,7 @@
static force_inline uint32_t
fetch_24 (uint8_t *a)
{
- if (((unsigned long)a) & 1)
+ if (((uintptr_t)a) & 1)
{
#ifdef WORDS_BIGENDIAN
return (*a << 16) | (*(uint16_t *)(a + 1));
@@ -57,7 +57,7 @@ static force_inline void
store_24 (uint8_t *a,
uint32_t v)
{
- if (((unsigned long)a) & 1)
+ if (((uintptr_t)a) & 1)
{
#ifdef WORDS_BIGENDIAN
*a = (uint8_t) (v >> 16);
diff --git a/pixman/pixman-mmx.c b/pixman/pixman-mmx.c
index 76b6ced..aef468a 100644
--- a/pixman/pixman-mmx.c
+++ b/pixman/pixman-mmx.c
@@ -1402,7 +1402,7 @@ mmx_composite_over_n_8888 (pixman_implementation_t *imp,
CHECKPOINT ();
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
store8888 (dst, over (vsrc, vsrca, load8888 (dst)));
@@ -1468,7 +1468,7 @@ mmx_composite_over_n_0565 (pixman_implementation_t *imp,
CHECKPOINT ();
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
uint64_t d = *dst;
__m64 vdest = expand565 (to_m64 (d), 0);
@@ -1546,7 +1546,7 @@ mmx_composite_over_n_8888_8888_ca (pixman_implementation_t *imp,
uint32_t *p = (uint32_t *)mask_line;
uint32_t *q = (uint32_t *)dst_line;
- while (twidth && (unsigned long)q & 7)
+ while (twidth && (uintptr_t)q & 7)
{
uint32_t m = *(uint32_t *)p;
@@ -1637,7 +1637,7 @@ mmx_composite_over_8888_n_8888 (pixman_implementation_t *imp,
src_line += src_stride;
w = width;
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
__m64 s = load8888 (src);
__m64 d = load8888 (dst);
@@ -1707,7 +1707,7 @@ mmx_composite_over_x888_n_8888 (pixman_implementation_t *imp,
src_line += src_stride;
w = width;
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
uint32_t ssrc = *src | 0xff000000;
__m64 s = load8888 (&ssrc);
@@ -1881,7 +1881,7 @@ mmx_composite_over_8888_0565 (pixman_implementation_t *imp,
CHECKPOINT ();
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
__m64 vsrc = load8888 (src);
uint64_t d = *dst;
@@ -1984,7 +1984,7 @@ mmx_composite_over_n_8_8888 (pixman_implementation_t *imp,
CHECKPOINT ();
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
uint64_t m = *mask;
@@ -2127,21 +2127,21 @@ mmx_fill (pixman_implementation_t *imp,
byte_line += stride;
w = byte_width;
- if (w >= 1 && ((unsigned long)d & 1))
+ if (w >= 1 && ((uintptr_t)d & 1))
{
*(uint8_t *)d = (xor & 0xff);
w--;
d++;
}
- if (w >= 2 && ((unsigned long)d & 3))
+ if (w >= 2 && ((uintptr_t)d & 3))
{
*(uint16_t *)d = xor;
w -= 2;
d += 2;
}
- while (w >= 4 && ((unsigned long)d & 7))
+ while (w >= 4 && ((uintptr_t)d & 7))
{
*(uint32_t *)d = xor;
@@ -2227,7 +2227,7 @@ mmx_composite_src_x888_0565 (pixman_implementation_t *imp,
src_line += src_stride;
w = width;
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
s = *src++;
*dst = CONVERT_8888_TO_0565 (s);
@@ -2305,7 +2305,7 @@ mmx_composite_src_n_8_8888 (pixman_implementation_t *imp,
CHECKPOINT ();
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
uint64_t m = *mask;
@@ -2419,7 +2419,7 @@ mmx_composite_over_n_8_0565 (pixman_implementation_t *imp,
CHECKPOINT ();
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
uint64_t m = *mask;
@@ -2536,7 +2536,7 @@ mmx_composite_over_pixbuf_0565 (pixman_implementation_t *imp,
CHECKPOINT ();
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
__m64 vsrc = load8888 (src);
uint64_t d = *dst;
@@ -2651,7 +2651,7 @@ mmx_composite_over_pixbuf_8888 (pixman_implementation_t *imp,
src_line += src_stride;
w = width;
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
__m64 s = load8888 (src);
__m64 d = load8888 (dst);
@@ -2739,7 +2739,7 @@ mmx_composite_over_n_8888_0565_ca (pixman_implementation_t *imp,
uint32_t *p = (uint32_t *)mask_line;
uint16_t *q = (uint16_t *)dst_line;
- while (twidth && ((unsigned long)q & 7))
+ while (twidth && ((uintptr_t)q & 7))
{
uint32_t m = *(uint32_t *)p;
@@ -2840,7 +2840,7 @@ mmx_composite_in_n_8_8 (pixman_implementation_t *imp,
mask_line += mask_stride;
w = width;
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
uint16_t tmp;
uint8_t a;
@@ -2911,7 +2911,7 @@ mmx_composite_in_8_8 (pixman_implementation_t *imp,
src_line += src_stride;
w = width;
- while (w && (unsigned long)dst & 3)
+ while (w && (uintptr_t)dst & 3)
{
uint8_t s, d;
uint16_t tmp;
@@ -2990,7 +2990,7 @@ mmx_composite_add_n_8_8 (pixman_implementation_t *imp,
mask_line += mask_stride;
w = width;
- while (w && (unsigned long)dst & 3)
+ while (w && (uintptr_t)dst & 3)
{
uint16_t tmp;
uint16_t a;
@@ -3067,7 +3067,7 @@ mmx_composite_add_8_8 (pixman_implementation_t *imp,
src_line += src_stride;
w = width;
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
s = *src;
d = *dst;
@@ -3130,7 +3130,7 @@ mmx_composite_add_0565_0565 (pixman_implementation_t *imp,
src_line += src_stride;
w = width;
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
s = *src++;
if (s)
@@ -3212,7 +3212,7 @@ mmx_composite_add_8888_8888 (pixman_implementation_t *imp,
src_line += src_stride;
w = width;
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
store (dst, _mm_adds_pu8 (load ((const uint32_t *)src),
load ((const uint32_t *)dst)));
@@ -3296,7 +3296,7 @@ mmx_blt (pixman_implementation_t *imp,
dst_bytes += dst_stride;
w = byte_width;
- if (w >= 1 && ((unsigned long)d & 1))
+ if (w >= 1 && ((uintptr_t)d & 1))
{
*(uint8_t *)d = *(uint8_t *)s;
w -= 1;
@@ -3304,7 +3304,7 @@ mmx_blt (pixman_implementation_t *imp,
d += 1;
}
- if (w >= 2 && ((unsigned long)d & 3))
+ if (w >= 2 && ((uintptr_t)d & 3))
{
*(uint16_t *)d = *(uint16_t *)s;
w -= 2;
@@ -3312,7 +3312,7 @@ mmx_blt (pixman_implementation_t *imp,
d += 2;
}
- while (w >= 4 && ((unsigned long)d & 7))
+ while (w >= 4 && ((uintptr_t)d & 7))
{
*(uint32_t *)d = ldl_u ((uint32_t *)s);
@@ -3495,7 +3495,7 @@ mmx_composite_over_reverse_n_8888 (pixman_implementation_t *imp,
CHECKPOINT ();
- while (w && (unsigned long)dst & 7)
+ while (w && (uintptr_t)dst & 7)
{
__m64 vdest = load8888 (dst);
@@ -3778,7 +3778,7 @@ mmx_fetch_x8r8g8b8 (pixman_iter_t *iter, const uint32_t *mask)
iter->bits += iter->stride;
- while (w && ((unsigned long)dst) & 7)
+ while (w && ((uintptr_t)dst) & 7)
{
*dst++ = (*src++) | 0xff000000;
w--;
@@ -3820,7 +3820,7 @@ mmx_fetch_r5g6b5 (pixman_iter_t *iter, const uint32_t *mask)
iter->bits += iter->stride;
- while (w && ((unsigned long)dst) & 0x0f)
+ while (w && ((uintptr_t)dst) & 0x0f)
{
uint16_t s = *src++;
@@ -3864,7 +3864,7 @@ mmx_fetch_a8 (pixman_iter_t *iter, const uint32_t *mask)
iter->bits += iter->stride;
- while (w && (((unsigned long)dst) & 15))
+ while (w && (((uintptr_t)dst) & 15))
{
*dst++ = *(src++) << 24;
w--;
diff --git a/pixman/pixman-sse2.c b/pixman/pixman-sse2.c
index efed310..7e980c9 100644
--- a/pixman/pixman-sse2.c
+++ b/pixman/pixman-sse2.c
@@ -576,7 +576,7 @@ core_combine_over_u_sse2_mask (uint32_t * pd,
uint32_t s, d;
/* Align dst on a 16-byte boundary */
- while (w && ((unsigned long)pd & 15))
+ while (w && ((uintptr_t)pd & 15))
{
d = *pd;
s = combine1 (ps, pm);
@@ -661,7 +661,7 @@ core_combine_over_u_sse2_no_mask (uint32_t * pd,
uint32_t s, d;
/* Align dst on a 16-byte boundary */
- while (w && ((unsigned long)pd & 15))
+ while (w && ((uintptr_t)pd & 15))
{
d = *pd;
s = *ps;
@@ -753,7 +753,7 @@ sse2_combine_over_reverse_u (pixman_implementation_t *imp,
/* Align dst on a 16-byte boundary */
while (w &&
- ((unsigned long)pd & 15))
+ ((uintptr_t)pd & 15))
{
d = *pd;
s = combine1 (ps, pm);
@@ -840,7 +840,7 @@ sse2_combine_in_u (pixman_implementation_t *imp,
__m128i xmm_src_lo, xmm_src_hi;
__m128i xmm_dst_lo, xmm_dst_hi;
- while (w && ((unsigned long) pd & 15))
+ while (w && ((uintptr_t)pd & 15))
{
s = combine1 (ps, pm);
d = *pd;
@@ -901,7 +901,7 @@ sse2_combine_in_reverse_u (pixman_implementation_t *imp,
__m128i xmm_src_lo, xmm_src_hi;
__m128i xmm_dst_lo, xmm_dst_hi;
- while (w && ((unsigned long) pd & 15))
+ while (w && ((uintptr_t)pd & 15))
{
s = combine1 (ps, pm);
d = *pd;
@@ -957,7 +957,7 @@ sse2_combine_out_reverse_u (pixman_implementation_t *imp,
const uint32_t * pm,
int w)
{
- while (w && ((unsigned long) pd & 15))
+ while (w && ((uintptr_t)pd & 15))
{
uint32_t s = combine1 (ps, pm);
uint32_t d = *pd;
@@ -1026,7 +1026,7 @@ sse2_combine_out_u (pixman_implementation_t *imp,
const uint32_t * pm,
int w)
{
- while (w && ((unsigned long) pd & 15))
+ while (w && ((uintptr_t)pd & 15))
{
uint32_t s = combine1 (ps, pm);
uint32_t d = *pd;
@@ -1113,7 +1113,7 @@ sse2_combine_atop_u (pixman_implementation_t *imp,
__m128i xmm_alpha_src_lo, xmm_alpha_src_hi;
__m128i xmm_alpha_dst_lo, xmm_alpha_dst_hi;
- while (w && ((unsigned long) pd & 15))
+ while (w && ((uintptr_t)pd & 15))
{
s = combine1 (ps, pm);
d = *pd;
@@ -1197,7 +1197,7 @@ sse2_combine_atop_reverse_u (pixman_implementation_t *imp,
__m128i xmm_alpha_src_lo, xmm_alpha_src_hi;
__m128i xmm_alpha_dst_lo, xmm_alpha_dst_hi;
- while (w && ((unsigned long) pd & 15))
+ while (w && ((uintptr_t)pd & 15))
{
s = combine1 (ps, pm);
d = *pd;
@@ -1285,7 +1285,7 @@ sse2_combine_xor_u (pixman_implementation_t *imp,
__m128i xmm_alpha_src_lo, xmm_alpha_src_hi;
__m128i xmm_alpha_dst_lo, xmm_alpha_dst_hi;
- while (w && ((unsigned long) pd & 15))
+ while (w && ((uintptr_t)pd & 15))
{
s = combine1 (ps, pm);
d = *pd;
@@ -1357,7 +1357,7 @@ sse2_combine_add_u (pixman_implementation_t *imp,
const uint32_t* ps = src;
const uint32_t* pm = mask;
- while (w && (unsigned long)pd & 15)
+ while (w && (uintptr_t)pd & 15)
{
s = combine1 (ps, pm);
d = *pd;
@@ -1430,7 +1430,7 @@ sse2_combine_saturate_u (pixman_implementation_t *imp,
uint32_t pack_cmp;
__m128i xmm_src, xmm_dst;
- while (w && (unsigned long)pd & 15)
+ while (w && (uintptr_t)pd & 15)
{
s = combine1 (ps, pm);
d = *pd;
@@ -1518,7 +1518,7 @@ sse2_combine_src_ca (pixman_implementation_t *imp,
__m128i xmm_mask_lo, xmm_mask_hi;
__m128i xmm_dst_lo, xmm_dst_hi;
- while (w && (unsigned long)pd & 15)
+ while (w && (uintptr_t)pd & 15)
{
s = *ps++;
m = *pm++;
@@ -1586,7 +1586,7 @@ sse2_combine_over_ca (pixman_implementation_t *imp,
__m128i xmm_dst_lo, xmm_dst_hi;
__m128i xmm_mask_lo, xmm_mask_hi;
- while (w && (unsigned long)pd & 15)
+ while (w && (uintptr_t)pd & 15)
{
s = *ps++;
m = *pm++;
@@ -1662,7 +1662,7 @@ sse2_combine_over_reverse_ca (pixman_implementation_t *imp,
__m128i xmm_dst_lo, xmm_dst_hi;
__m128i xmm_mask_lo, xmm_mask_hi;
- while (w && (unsigned long)pd & 15)
+ while (w && (uintptr_t)pd & 15)
{
s = *ps++;
m = *pm++;
@@ -1727,7 +1727,7 @@ sse2_combine_in_ca (pixman_implementation_t *imp,
__m128i xmm_dst_lo, xmm_dst_hi;
__m128i xmm_mask_lo, xmm_mask_hi;
- while (w && (unsigned long)pd & 15)
+ while (w && (uintptr_t)pd & 15)
{
s = *ps++;
m = *pm++;
@@ -1802,7 +1802,7 @@ sse2_combine_in_reverse_ca (pixman_implementation_t *imp,
__m128i xmm_dst_lo, xmm_dst_hi;
__m128i xmm_mask_lo, xmm_mask_hi;
- while (w && (unsigned long)pd & 15)
+ while (w && (uintptr_t)pd & 15)
{
s = *ps++;
m = *pm++;
@@ -1875,7 +1875,7 @@ sse2_combine_out_ca (pixman_implementation_t *imp,
__m128i xmm_dst_lo, xmm_dst_hi;
__m128i xmm_mask_lo, xmm_mask_hi;
- while (w && (unsigned long)pd & 15)
+ while (w && (uintptr_t)pd & 15)
{
s = *ps++;
m = *pm++;
@@ -1951,7 +1951,7 @@ sse2_combine_out_reverse_ca (pixman_implementation_t *imp,
__m128i xmm_dst_lo, xmm_dst_hi;
__m128i xmm_mask_lo, xmm_mask_hi;
- while (w && (unsigned long)pd & 15)
+ while (w && (uintptr_t)pd & 15)
{
s = *ps++;
m = *pm++;
@@ -2048,7 +2048,7 @@ sse2_combine_atop_ca (pixman_implementation_t *imp,
__m128i xmm_alpha_dst_lo, xmm_alpha_dst_hi;
__m128i xmm_mask_lo, xmm_mask_hi;
- while (w && (unsigned long)pd & 15)
+ while (w && (uintptr_t)pd & 15)
{
s = *ps++;
m = *pm++;
@@ -2141,7 +2141,7 @@ sse2_combine_atop_reverse_ca (pixman_implementation_t *imp,
__m128i xmm_alpha_dst_lo, xmm_alpha_dst_hi;
__m128i xmm_mask_lo, xmm_mask_hi;
- while (w && (unsigned long)pd & 15)
+ while (w && (uintptr_t)pd & 15)
{
s = *ps++;
m = *pm++;
@@ -2237,7 +2237,7 @@ sse2_combine_xor_ca (pixman_implementation_t *imp,
__m128i xmm_alpha_dst_lo, xmm_alpha_dst_hi;
__m128i xmm_mask_lo, xmm_mask_hi;
- while (w && (unsigned long)pd & 15)
+ while (w && (uintptr_t)pd & 15)
{
s = *ps++;
m = *pm++;
@@ -2313,7 +2313,7 @@ sse2_combine_add_ca (pixman_implementation_t *imp,
__m128i xmm_dst_lo, xmm_dst_hi;
__m128i xmm_mask_lo, xmm_mask_hi;
- while (w && (unsigned long)pd & 15)
+ while (w && (uintptr_t)pd & 15)
{
s = *ps++;
m = *pm++;
@@ -2414,7 +2414,7 @@ sse2_composite_over_n_8888 (pixman_implementation_t *imp,
dst_line += dst_stride;
w = width;
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
d = *dst;
*dst++ = pack_1x128_32 (over_1x128 (xmm_src,
@@ -2483,7 +2483,7 @@ sse2_composite_over_n_0565 (pixman_implementation_t *imp,
dst_line += dst_stride;
w = width;
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
d = *dst;
@@ -2568,7 +2568,7 @@ sse2_composite_add_n_8888_8888_ca (pixman_implementation_t *imp,
dst_line += dst_stride;
mask_line += mask_stride;
- while (w && (unsigned long)pd & 15)
+ while (w && (uintptr_t)pd & 15)
{
m = *pm++;
@@ -2682,7 +2682,7 @@ sse2_composite_over_n_8888_8888_ca (pixman_implementation_t *imp,
dst_line += dst_stride;
mask_line += mask_stride;
- while (w && (unsigned long)pd & 15)
+ while (w && (uintptr_t)pd & 15)
{
m = *pm++;
@@ -2786,7 +2786,7 @@ sse2_composite_over_8888_n_8888 (pixman_implementation_t *imp,
src_line += src_stride;
w = width;
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
uint32_t s = *src++;
@@ -2878,7 +2878,7 @@ sse2_composite_src_x888_0565 (pixman_implementation_t *imp,
src_line += src_stride;
w = width;
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
s = *src++;
*dst = CONVERT_8888_TO_0565 (s);
@@ -2932,7 +2932,7 @@ sse2_composite_src_x888_8888 (pixman_implementation_t *imp,
src_line += src_stride;
w = width;
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
*dst++ = *src++ | 0xff000000;
w--;
@@ -2999,7 +2999,7 @@ sse2_composite_over_x888_n_8888 (pixman_implementation_t *imp,
src_line += src_stride;
w = width;
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
uint32_t s = (*src++) | 0xff000000;
uint32_t d = *dst;
@@ -3125,7 +3125,7 @@ sse2_composite_over_8888_0565 (pixman_implementation_t *imp,
/* Align dst on a 16-byte boundary */
while (w &&
- ((unsigned long)dst & 15))
+ ((uintptr_t)dst & 15))
{
s = *src++;
d = *dst;
@@ -3231,7 +3231,7 @@ sse2_composite_over_n_8_8888 (pixman_implementation_t *imp,
mask_line += mask_stride;
w = width;
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
uint8_t m = *mask++;
@@ -3372,21 +3372,21 @@ sse2_fill (pixman_implementation_t *imp,
byte_line += stride;
w = byte_width;
- if (w >= 1 && ((unsigned long)d & 1))
+ if (w >= 1 && ((uintptr_t)d & 1))
{
*(uint8_t *)d = xor;
w -= 1;
d += 1;
}
- while (w >= 2 && ((unsigned long)d & 3))
+ while (w >= 2 && ((uintptr_t)d & 3))
{
*(uint16_t *)d = xor;
w -= 2;
d += 2;
}
- while (w >= 4 && ((unsigned long)d & 15))
+ while (w >= 4 && ((uintptr_t)d & 15))
{
*(uint32_t *)d = xor;
@@ -3505,7 +3505,7 @@ sse2_composite_src_n_8_8888 (pixman_implementation_t *imp,
mask_line += mask_stride;
w = width;
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
uint8_t m = *mask++;
@@ -3621,7 +3621,7 @@ sse2_composite_over_n_8_0565 (pixman_implementation_t *imp,
mask_line += mask_stride;
w = width;
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
m = *mask++;
@@ -3745,7 +3745,7 @@ sse2_composite_over_pixbuf_0565 (pixman_implementation_t *imp,
src_line += src_stride;
w = width;
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
s = *src++;
d = *dst;
@@ -3854,7 +3854,7 @@ sse2_composite_over_pixbuf_8888 (pixman_implementation_t *imp,
src_line += src_stride;
w = width;
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
s = *src++;
d = *dst;
@@ -3957,7 +3957,7 @@ sse2_composite_over_n_8888_0565_ca (pixman_implementation_t *imp,
mask_line += mask_stride;
dst_line += dst_stride;
- while (w && ((unsigned long)dst & 15))
+ while (w && ((uintptr_t)dst & 15))
{
m = *(uint32_t *) mask;
@@ -4083,7 +4083,7 @@ sse2_composite_in_n_8_8 (pixman_implementation_t *imp,
mask_line += mask_stride;
w = width;
- while (w && ((unsigned long)dst & 15))
+ while (w && ((uintptr_t)dst & 15))
{
m = (uint32_t) *mask++;
d = (uint32_t) *dst;
@@ -4176,7 +4176,7 @@ sse2_composite_in_n_8 (pixman_implementation_t *imp,
dst_line += dst_stride;
w = width;
- while (w && ((unsigned long)dst & 15))
+ while (w && ((uintptr_t)dst & 15))
{
d = (uint32_t) *dst;
@@ -4245,7 +4245,7 @@ sse2_composite_in_8_8 (pixman_implementation_t *imp,
src_line += src_stride;
w = width;
- while (w && ((unsigned long)dst & 15))
+ while (w && ((uintptr_t)dst & 15))
{
s = (uint32_t) *src++;
d = (uint32_t) *dst;
@@ -4322,7 +4322,7 @@ sse2_composite_add_n_8_8 (pixman_implementation_t *imp,
mask_line += mask_stride;
w = width;
- while (w && ((unsigned long)dst & 15))
+ while (w && ((uintptr_t)dst & 15))
{
m = (uint32_t) *mask++;
d = (uint32_t) *dst;
@@ -4414,7 +4414,7 @@ sse2_composite_add_n_8 (pixman_implementation_t *imp,
dst_line += dst_stride;
w = width;
- while (w && ((unsigned long)dst & 15))
+ while (w && ((uintptr_t)dst & 15))
{
*dst = (uint8_t)_mm_cvtsi128_si32 (
_mm_adds_epu8 (
@@ -4474,7 +4474,7 @@ sse2_composite_add_8_8 (pixman_implementation_t *imp,
w = width;
/* Small head */
- while (w && (unsigned long)dst & 3)
+ while (w && (uintptr_t)dst & 3)
{
t = (*dst) + (*src++);
*dst++ = t | (0 - (t >> 8));
@@ -4582,7 +4582,7 @@ sse2_blt (pixman_implementation_t *imp,
dst_bytes += dst_stride;
w = byte_width;
- while (w >= 2 && ((unsigned long)d & 3))
+ while (w >= 2 && ((uintptr_t)d & 3))
{
*(uint16_t *)d = *(uint16_t *)s;
w -= 2;
@@ -4590,7 +4590,7 @@ sse2_blt (pixman_implementation_t *imp,
d += 2;
}
- while (w >= 4 && ((unsigned long)d & 15))
+ while (w >= 4 && ((uintptr_t)d & 15))
{
*(uint32_t *)d = *(uint32_t *)s;
@@ -4697,7 +4697,7 @@ sse2_composite_over_x888_8_8888 (pixman_implementation_t *imp,
w = width;
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
s = 0xff000000 | *src++;
m = (uint32_t) *mask++;
@@ -4821,7 +4821,7 @@ sse2_composite_over_8888_8_8888 (pixman_implementation_t *imp,
w = width;
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
uint32_t sa;
@@ -4960,7 +4960,7 @@ sse2_composite_over_reverse_n_8888 (pixman_implementation_t *imp,
dst_line += dst_stride;
w = width;
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
__m128i vd;
@@ -5045,7 +5045,7 @@ sse2_composite_over_8888_8888_8888 (pixman_implementation_t *imp,
w = width;
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
uint32_t sa;
@@ -5173,7 +5173,7 @@ scaled_nearest_scanline_sse2_8888_8888_OVER (uint32_t* pd,
return;
/* Align dst on a 16-byte boundary */
- while (w && ((unsigned long)pd & 15))
+ while (w && ((uintptr_t)pd & 15))
{
d = *pd;
s = combine1 (ps + pixman_fixed_to_int (vx), pm);
@@ -5291,7 +5291,7 @@ scaled_nearest_scanline_sse2_8888_n_8888_OVER (const uint32_t * mask,
xmm_mask = create_mask_16_128 (*mask >> 24);
- while (w && (unsigned long)dst & 15)
+ while (w && (uintptr_t)dst & 15)
{
uint32_t s = *(src + pixman_fixed_to_int (vx));
vx += unit_x;
@@ -5538,7 +5538,7 @@ scaled_bilinear_scanline_sse2_8888_8888_OVER (uint32_t * dst,
BILINEAR_DECLARE_VARIABLES;
uint32_t pix1, pix2, pix3, pix4;
- while (w && ((unsigned long)dst & 15))
+ while (w && ((uintptr_t)dst & 15))
{
BILINEAR_INTERPOLATE_ONE_PIXEL (pix1);
@@ -5639,7 +5639,7 @@ scaled_bilinear_scanline_sse2_8888_8_8888_OVER (uint32_t * dst,
uint32_t pix1, pix2, pix3, pix4;
uint32_t m;
- while (w && ((unsigned long)dst & 15))
+ while (w && ((uintptr_t)dst & 15))
{
uint32_t sa;
@@ -5930,7 +5930,7 @@ sse2_fetch_x8r8g8b8 (pixman_iter_t *iter, const uint32_t *mask)
iter->bits += iter->stride;
- while (w && ((unsigned long)dst) & 0x0f)
+ while (w && ((uintptr_t)dst) & 0x0f)
{
*dst++ = (*src++) | 0xff000000;
w--;
@@ -5966,7 +5966,7 @@ sse2_fetch_r5g6b5 (pixman_iter_t *iter, const uint32_t *mask)
iter->bits += iter->stride;
- while (w && ((unsigned long)dst) & 0x0f)
+ while (w && ((uintptr_t)dst) & 0x0f)
{
uint16_t s = *src++;
@@ -6012,7 +6012,7 @@ sse2_fetch_a8 (pixman_iter_t *iter, const uint32_t *mask)
iter->bits += iter->stride;
- while (w && (((unsigned long)dst) & 15))
+ while (w && (((uintptr_t)dst) & 15))
{
*dst++ = *(src++) << 24;
w--;
diff --git a/test/glyph-test.c b/test/glyph-test.c
index 9dd5b41..501cc2e 100644
--- a/test/glyph-test.c
+++ b/test/glyph-test.c
@@ -215,8 +215,8 @@ create_image (int max_size, const pixman_format_code_t *formats, uint32_t flags)
return image;
}
-#define KEY1(p) ((void *)(((unsigned long)p) ^ (0xa7e23dfaUL)))
-#define KEY2(p) ((void *)(((unsigned long)p) ^ (0xabcd9876UL)))
+#define KEY1(p) ((void *)(((uintptr_t)p) ^ (0xa7e23dfaUL)))
+#define KEY2(p) ((void *)(((uintptr_t)p) ^ (0xabcd9876UL)))
#define MAX_GLYPHS 32
diff --git a/test/utils.c b/test/utils.c
index 716bb75..ef0a171 100644
--- a/test/utils.c
+++ b/test/utils.c
@@ -377,7 +377,7 @@ fence_malloc (int64_t len)
return NULL;
}
- initial_page = (uint8_t *)(((unsigned long)addr + page_mask) & ~page_mask);
+ initial_page = (uint8_t *)(((uintptr_t)addr + page_mask) & ~page_mask);
leading_protected = initial_page + page_size;
payload = leading_protected + N_LEADING_PROTECTED * page_size;
trailing_protected = payload + n_payload_bytes;
commit a96efd02d68b726d6d140d0bd211bc7cc1be127a
Author: Stefan Weil <sw at weilnetz.de>
Date: Tue Nov 13 19:44:15 2012 +0100
Always use xmmintrin.h for 64 bit Windows
MinGW-w64 uses the GNU compiler and does not define _MSC_VER.
Nevertheless, it provides xmmintrin.h and must be handled
here like the MS compiler. Otherwise compilation fails due to
conflicting declarations.
Signed-off-by: Stefan Weil <sw at weilnetz.de>
diff --git a/pixman/pixman-mmx.c b/pixman/pixman-mmx.c
index 1e6dbe8..76b6ced 100644
--- a/pixman/pixman-mmx.c
+++ b/pixman/pixman-mmx.c
@@ -62,7 +62,7 @@ _mm_empty (void)
#endif
#ifdef USE_X86_MMX
-# if (defined(__SUNPRO_C) || defined(_MSC_VER))
+# if (defined(__SUNPRO_C) || defined(_MSC_VER) || defined(_WIN64))
# include <xmmintrin.h>
# else
/* We have to compile with -msse to use xmmintrin.h, but that causes SSE
commit 899e0d60524bcd2cff6cad6acb310181fb96b39a
Author: Nemanja Lukic <nemanja.lukic at rt-rk.com>
Date: Mon Nov 12 22:48:51 2012 +0100
MIPS: DSPr2: Added several nearest neighbor fast paths with a8 mask:
Performance numbers before/after on MIPS-74kc @ 1GHz:
lowlevel-blt-bench -n
Referent (before):
over_8888_8_0565 = L1: 9.62 L2: 8.85 M: 7.40 ( 39.27%) HT: 5.67 VT: 5.61 R: 5.45 RT: 2.98 ( 22Kops/s)
over_0565_8_0565 = L1: 7.90 L2: 7.49 M: 6.72 ( 26.75%) HT: 5.24 VT: 5.20 R: 5.06 RT: 2.90 ( 22Kops/s)
Optimized:
over_8888_8_0565 = L1: 18.51 L2: 16.82 M: 12.13 ( 64.43%) HT: 10.06 VT: 9.88 R: 9.54 RT: 5.63 ( 31Kops/s)
over_0565_8_0565 = L1: 14.82 L2: 13.94 M: 11.34 ( 45.20%) HT: 9.45 VT: 9.35 R: 9.03 RT: 5.50 ( 31Kops/s)
diff --git a/pixman/pixman-mips-dspr2-asm.S b/pixman/pixman-mips-dspr2-asm.S
index 791244a..ba22e62 100644
--- a/pixman/pixman-mips-dspr2-asm.S
+++ b/pixman/pixman-mips-dspr2-asm.S
@@ -2100,6 +2100,163 @@ LEAF_MIPS_DSPR2(pixman_composite_add_8888_8888_asm_mips)
END(pixman_composite_add_8888_8888_asm_mips)
+LEAF_MIPS_DSPR2(pixman_scaled_nearest_scanline_8888_8_0565_OVER_asm_mips)
+/*
+ * a0 - dst (r5g6b5)
+ * a1 - src (a8r8g8b8)
+ * a2 - mask (a8)
+ * a3 - w
+ * 16(sp) - vx
+ * 20(sp) - unit_x
+ */
+ beqz a3, 4f
+ nop
+
+ SAVE_REGS_ON_STACK 20, v0, v1, s0, s1, s2, s3, s4, s5
+ lw v0, 36(sp) /* v0 = vx */
+ lw v1, 40(sp) /* v1 = unit_x */
+ li t6, 0x00ff00ff
+ li t7, 0xf800f800
+ li t8, 0x07e007e0
+ li t9, 0x001F001F
+
+ addiu t1, a3, -1
+ beqz t1, 2f
+ nop
+1:
+ sra t0, v0, 16 /* t0 = vx >> 16 */
+ sll t0, t0, 2 /* t0 = t0 * 4 (a8r8g8b8) */
+ addu t0, a1, t0
+ lw t0, 0(t0) /* t0 = source (a8r8g8b8) */
+ addu v0, v0, v1 /* v0 = vx + unit_x */
+ sra t1, v0, 16 /* t1 = vx >> 16 */
+ sll t1, t1, 2 /* t1 = t1 * 4 (a8r8g8b8) */
+ addu t1, a1, t1
+ lw t1, 0(t1) /* t1 = source (a8r8g8b8) */
+ addu v0, v0, v1 /* v0 = vx + unit_x */
+ lbu t2, 0(a2) /* t2 = mask (a8) */
+ lbu t3, 1(a2) /* t3 = mask (a8) */
+ lhu t4, 0(a0) /* t4 = destination (r5g6b5) */
+ lhu t5, 2(a0) /* t5 = destination (r5g6b5) */
+ addiu a2, a2, 2
+
+ CONVERT_2x0565_TO_2x8888 t4, t5, s0, s1, t8, t9, s2, s3, s4, s5
+ OVER_2x8888_2x8_2x8888 t0, t1, \
+ t2, t3, \
+ s0, s1, \
+ t4, t5, \
+ t6, s2, s3, s4, s5, t2, t3
+ CONVERT_2x8888_TO_2x0565 t4, t5, s0, s1, t7, t8, t9, s2, s3
+
+ sh s0, 0(a0)
+ sh s1, 2(a0)
+ addiu a3, a3, -2
+ addiu t1, a3, -1
+ bgtz t1, 1b
+ addiu a0, a0, 4
+2:
+ beqz a3, 3f
+ nop
+ sra t0, v0, 16 /* t0 = vx >> 16 */
+ sll t0, t0, 2 /* t0 = t0 * 4 (a8r8g8b8) */
+ addu t0, a1, t0
+ lw t0, 0(t0) /* t0 = source (a8r8g8b8) */
+ lbu t1, 0(a2) /* t1 = mask (a8) */
+ lhu t2, 0(a0) /* t2 = destination (r5g6b5) */
+
+ CONVERT_1x0565_TO_1x8888 t2, t3, t4, t5
+ OVER_8888_8_8888 t0, t1, t3, t2, t6, t4, t5, t7, t8
+ CONVERT_1x8888_TO_1x0565 t2, t3, t4, t5
+
+ sh t3, 0(a0)
+3:
+ RESTORE_REGS_FROM_STACK 20, v0, v1, s0, s1, s2, s3, s4, s5
+4:
+ j ra
+ nop
+
+END(pixman_scaled_nearest_scanline_8888_8_0565_OVER_asm_mips)
+
+LEAF_MIPS_DSPR2(pixman_scaled_nearest_scanline_0565_8_0565_OVER_asm_mips)
+/*
+ * a0 - dst (r5g6b5)
+ * a1 - src (r5g6b5)
+ * a2 - mask (a8)
+ * a3 - w
+ * 16(sp) - vx
+ * 20(sp) - unit_x
+ */
+
+ beqz a3, 4f
+ nop
+ SAVE_REGS_ON_STACK 20, v0, v1, s0, s1, s2, s3, s4, s5
+ lw v0, 36(sp) /* v0 = vx */
+ lw v1, 40(sp) /* v1 = unit_x */
+ li t4, 0xf800f800
+ li t5, 0x07e007e0
+ li t6, 0x001F001F
+ li t7, 0x00ff00ff
+
+ addiu t1, a3, -1
+ beqz t1, 2f
+ nop
+1:
+ sra t0, v0, 16 /* t0 = vx >> 16 */
+ sll t0, t0, 1 /* t0 = t0 * 2 (r5g6b5) */
+ addu t0, a1, t0
+ lhu t0, 0(t0) /* t0 = source (r5g6b5) */
+ addu v0, v0, v1 /* v0 = vx + unit_x */
+ sra t1, v0, 16 /* t1 = vx >> 16 */
+ sll t1, t1, 1 /* t1 = t1 * 2 (r5g6b5) */
+ addu t1, a1, t1
+ lhu t1, 0(t1) /* t1 = source (r5g6b5) */
+ addu v0, v0, v1 /* v0 = vx + unit_x */
+ lbu t2, 0(a2) /* t2 = mask (a8) */
+ lbu t3, 1(a2) /* t3 = mask (a8) */
+ lhu t8, 0(a0) /* t8 = destination (r5g6b5) */
+ lhu t9, 2(a0) /* t9 = destination (r5g6b5) */
+ addiu a2, a2, 2
+
+ CONVERT_2x0565_TO_2x8888 t0, t1, s0, s1, t5, t6, s2, s3, s4, s5
+ CONVERT_2x0565_TO_2x8888 t8, t9, s2, s3, t5, t6, s4, s5, t0, t1
+ OVER_2x8888_2x8_2x8888 s0, s1, \
+ t2, t3, \
+ s2, s3, \
+ t0, t1, \
+ t7, t8, t9, s4, s5, s0, s1
+ CONVERT_2x8888_TO_2x0565 t0, t1, s0, s1, t4, t5, t6, s2, s3
+
+ sh s0, 0(a0)
+ sh s1, 2(a0)
+ addiu a3, a3, -2
+ addiu t1, a3, -1
+ bgtz t1, 1b
+ addiu a0, a0, 4
+2:
+ beqz a3, 3f
+ nop
+ sra t0, v0, 16 /* t0 = vx >> 16 */
+ sll t0, t0, 1 /* t0 = t0 * 2 (r5g6b5) */
+ addu t0, a1, t0
+
+ lhu t0, 0(t0) /* t0 = source (r5g6b5) */
+ lbu t1, 0(a2) /* t1 = mask (a8) */
+ lhu t2, 0(a0) /* t2 = destination (r5g6b5) */
+
+ CONVERT_1x0565_TO_1x8888 t0, t3, t4, t5
+ CONVERT_1x0565_TO_1x8888 t2, t4, t5, t6
+ OVER_8888_8_8888 t3, t1, t4, t0, t7, t2, t5, t6, t8
+ CONVERT_1x8888_TO_1x0565 t0, t3, t4, t5
+
+ sh t3, 0(a0)
+3:
+ RESTORE_REGS_FROM_STACK 20, v0, v1, s0, s1, s2, s3, s4, s5
+4:
+ j ra
+ nop
+
+END(pixman_scaled_nearest_scanline_0565_8_0565_OVER_asm_mips)
+
LEAF_MIPS_DSPR2(pixman_scaled_bilinear_scanline_8888_8888_SRC_asm_mips)
/*
* a0 - *dst
diff --git a/pixman/pixman-mips-dspr2.c b/pixman/pixman-mips-dspr2.c
index afe4ec6..11f1254 100644
--- a/pixman/pixman-mips-dspr2.c
+++ b/pixman/pixman-mips-dspr2.c
@@ -116,6 +116,11 @@ PIXMAN_MIPS_BIND_SCALED_BILINEAR_SRC_DST (SKIP_ZERO_SRC, 8888_8888, OVER,
PIXMAN_MIPS_BIND_SCALED_BILINEAR_SRC_DST (SKIP_ZERO_SRC, 8888_8888, ADD,
uint32_t, uint32_t)
+PIXMAN_MIPS_BIND_SCALED_NEAREST_SRC_A8_DST (SKIP_ZERO_SRC, 8888_8_0565,
+ OVER, uint32_t, uint16_t)
+PIXMAN_MIPS_BIND_SCALED_NEAREST_SRC_A8_DST (SKIP_ZERO_SRC, 0565_8_0565,
+ OVER, uint16_t, uint16_t)
+
PIXMAN_MIPS_BIND_SCALED_BILINEAR_SRC_A8_DST (0, 8888_8_8888, SRC,
uint32_t, uint32_t)
PIXMAN_MIPS_BIND_SCALED_BILINEAR_SRC_A8_DST (0, 8888_8_0565, SRC,
@@ -320,6 +325,12 @@ static const pixman_fast_path_t mips_dspr2_fast_paths[] =
PIXMAN_STD_FAST_PATH (ADD, a8r8g8b8, null, a8r8g8b8, mips_composite_add_8888_8888),
PIXMAN_STD_FAST_PATH (ADD, a8b8g8r8, null, a8b8g8r8, mips_composite_add_8888_8888),
+ PIXMAN_MIPS_SIMPLE_NEAREST_A8_MASK_FAST_PATH (OVER, a8r8g8b8, r5g6b5, mips_8888_8_0565),
+ PIXMAN_MIPS_SIMPLE_NEAREST_A8_MASK_FAST_PATH (OVER, a8b8g8r8, b5g6r5, mips_8888_8_0565),
+
+ PIXMAN_MIPS_SIMPLE_NEAREST_A8_MASK_FAST_PATH (OVER, r5g6b5, r5g6b5, mips_0565_8_0565),
+ PIXMAN_MIPS_SIMPLE_NEAREST_A8_MASK_FAST_PATH (OVER, b5g6r5, b5g6r5, mips_0565_8_0565),
+
SIMPLE_BILINEAR_FAST_PATH (SRC, a8r8g8b8, a8r8g8b8, mips_8888_8888),
SIMPLE_BILINEAR_FAST_PATH (SRC, a8r8g8b8, x8r8g8b8, mips_8888_8888),
SIMPLE_BILINEAR_FAST_PATH (SRC, x8r8g8b8, x8r8g8b8, mips_8888_8888),
diff --git a/pixman/pixman-mips-dspr2.h b/pixman/pixman-mips-dspr2.h
index 186a501..4ac9ff9 100644
--- a/pixman/pixman-mips-dspr2.h
+++ b/pixman/pixman-mips-dspr2.h
@@ -246,6 +246,52 @@ mips_composite_##name (pixman_implementation_t *imp, \
} \
}
+/*****************************************************************************/
+
+#define PIXMAN_MIPS_BIND_SCALED_NEAREST_SRC_A8_DST(flags, name, op, \
+ src_type, dst_type) \
+void \
+pixman_scaled_nearest_scanline_##name##_##op##_asm_mips ( \
+ dst_type * dst, \
+ const src_type * src, \
+ const uint8_t * mask, \
+ int32_t w, \
+ pixman_fixed_t vx, \
+ pixman_fixed_t unit_x); \
+ \
+static force_inline void \
+scaled_nearest_scanline_mips_##name##_##op (const uint8_t * mask, \
+ dst_type * pd, \
+ const src_type * ps, \
+ int32_t w, \
+ pixman_fixed_t vx, \
+ pixman_fixed_t unit_x, \
+ pixman_fixed_t max_vx, \
+ pixman_bool_t zero_src) \
+{ \
+ if ((flags & SKIP_ZERO_SRC) && zero_src) \
+ return; \
+ pixman_scaled_nearest_scanline_##name##_##op##_asm_mips (pd, ps, \
+ mask, w, \
+ vx, unit_x); \
+} \
+ \
+FAST_NEAREST_MAINLOOP_COMMON (mips_##name##_cover_##op, \
+ scaled_nearest_scanline_mips_##name##_##op, \
+ src_type, uint8_t, dst_type, COVER, TRUE, FALSE)\
+FAST_NEAREST_MAINLOOP_COMMON (mips_##name##_none_##op, \
+ scaled_nearest_scanline_mips_##name##_##op, \
+ src_type, uint8_t, dst_type, NONE, TRUE, FALSE) \
+FAST_NEAREST_MAINLOOP_COMMON (mips_##name##_pad_##op, \
+ scaled_nearest_scanline_mips_##name##_##op, \
+ src_type, uint8_t, dst_type, PAD, TRUE, FALSE)
+
+/* Provide entries for the fast path table */
+#define PIXMAN_MIPS_SIMPLE_NEAREST_A8_MASK_FAST_PATH(op,s,d,func) \
+ SIMPLE_NEAREST_A8_MASK_FAST_PATH_COVER (op,s,d,func), \
+ SIMPLE_NEAREST_A8_MASK_FAST_PATH_NONE (op,s,d,func), \
+ SIMPLE_NEAREST_A8_MASK_FAST_PATH_PAD (op,s,d,func)
+
/****************************************************************************/
#define PIXMAN_MIPS_BIND_SCALED_BILINEAR_SRC_DST(flags, name, op, \
commit a432bdce6637aa96060b9f1e25aae51c6fb95670
Author: Nemanja Lukic <nemanja.lukic at rt-rk.com>
Date: Mon Nov 12 22:48:53 2012 +0100
MIPS: DSPr2: Added more fast-paths for OVER operation:
Performance numbers before/after on MIPS-74kc @ 1GHz:
lowlevel-blt-bench results
Referent (before):
over_n_0565 = L1: 14.48 L2: 21.36 M: 17.57 ( 23.30%) HT: 6.95 VT: 6.44 R: 6.39 RT: 2.16 ( 22Kops/s)
over_n_8888 = L1: 92.60 L2: 86.13 M: 24.41 ( 64.74%) HT: 8.94 VT: 8.06 R: 8.00 RT: 2.53 ( 25Kops/s)
Optimized:
over_n_0565 = L1: 27.65 L2: 189.22 M: 58.19 ( 77.12%) HT: 52.80 VT: 49.88 R: 47.53 RT: 23.67 ( 72Kops/s)
over_n_8888 = L1: 235.99 L2: 230.86 M: 29.09 ( 77.11%) HT: 27.95 VT: 27.24 R: 26.58 RT: 18.10 ( 67Kops/s)
diff --git a/pixman/pixman-mips-dspr2-asm.S b/pixman/pixman-mips-dspr2-asm.S
index 02adb6d..791244a 100644
--- a/pixman/pixman-mips-dspr2-asm.S
+++ b/pixman/pixman-mips-dspr2-asm.S
@@ -1342,6 +1342,140 @@ LEAF_MIPS_DSPR2(pixman_composite_over_8888_8888_asm_mips)
END(pixman_composite_over_8888_8888_asm_mips)
+LEAF_MIPS_DSPR2(pixman_composite_over_n_0565_asm_mips)
+/*
+ * a0 - dst (r5g6b5)
+ * a1 - src (32bit constant)
+ * a2 - w
+ */
+
+ beqz a2, 5f
+ nop
+
+ not t0, a1
+ srl t0, t0, 24
+ bgtz t0, 1f
+ nop
+ CONVERT_1x8888_TO_1x0565 a1, t1, t2, t3
+0:
+ sh t1, 0(a0)
+ addiu a2, a2, -1
+ bgtz a2, 0b
+ addiu a0, a0, 2
+ j ra
+ nop
+
+1:
+ SAVE_REGS_ON_STACK 0, s0, s1, s2
+ li t4, 0x00ff00ff
+ li t5, 0xf800f800
+ li t6, 0x07e007e0
+ li t7, 0x001F001F
+ addiu t1, a2, -1
+ beqz t1, 3f
+ nop
+2:
+ lhu t1, 0(a0) /* t1 = destination (r5g6b5) */
+ lhu t2, 2(a0) /* t2 = destination (r5g6b5) */
+
+ CONVERT_2x0565_TO_2x8888 t1, t2, t3, t8, t6, t7, t9, s0, s1, s2
+ MIPS_2xUN8x4_MUL_2xUN8 t3, t8, t0, t0, t1, t2, t4, t9, s0, s1, s2, t3, t8
+ addu_s.qb t1, t1, a1
+ addu_s.qb t2, t2, a1
+ CONVERT_2x8888_TO_2x0565 t1, t2, t3, t8, t5, t6, t7, s0, s1
+
+ sh t3, 0(a0)
+ sh t8, 2(a0)
+
+ addiu a2, a2, -2
+ addiu t1, a2, -1
+ bgtz t1, 2b
+ addiu a0, a0, 4
+3:
+ beqz a2, 4f
+ nop
+
+ lhu t1, 0(a0) /* t1 = destination (r5g6b5) */
+
+ CONVERT_1x0565_TO_1x8888 t1, t2, s0, s1
+ MIPS_UN8x4_MUL_UN8 t2, t0, t1, t4, s0, s1, s2
+ addu_s.qb t1, t1, a1
+ CONVERT_1x8888_TO_1x0565 t1, t2, s0, s1
+
+ sh t2, 0(a0)
+
+4:
+ RESTORE_REGS_FROM_STACK 0, s0, s1, s2
+5:
+ j ra
+ nop
+
+END(pixman_composite_over_n_0565_asm_mips)
+
+LEAF_MIPS_DSPR2(pixman_composite_over_n_8888_asm_mips)
+/*
+ * a0 - dst (a8r8g8b8)
+ * a1 - src (32bit constant)
+ * a2 - w
+ */
+
+ beqz a2, 5f
+ nop
+
+ not t0, a1
+ srl t0, t0, 24
+ bgtz t0, 1f
+ nop
+0:
+ sw a1, 0(a0)
+ addiu a2, a2, -1
+ bgtz a2, 0b
+ addiu a0, a0, 4
+ j ra
+ nop
+
+1:
+ SAVE_REGS_ON_STACK 0, s0, s1, s2
+ li t4, 0x00ff00ff
+ addiu t1, a2, -1
+ beqz t1, 3f
+ nop
+2:
+ lw t2, 0(a0) /* t2 = destination (a8r8g8b8) */
+ lw t3, 4(a0) /* t3 = destination (a8r8g8b8) */
+
+ MIPS_2xUN8x4_MUL_2xUN8 t2, t3, t0, t0, t7, t8, t4, t9, s0, s1, s2, t2, t3
+
+ addu_s.qb t7, t7, a1
+ addu_s.qb t8, t8, a1
+
+ sw t7, 0(a0)
+ sw t8, 4(a0)
+
+ addiu a2, a2, -2
+ addiu t1, a2, -1
+ bgtz t1, 2b
+ addiu a0, a0, 8
+3:
+ beqz a2, 4f
+ nop
+
+ lw t1, 0(a0) /* t1 = destination (a8r8g8b8) */
+
+ MIPS_UN8x4_MUL_UN8 t1, t0, t3, t4, t5, t6, t7
+
+ addu_s.qb t3, t3, a1
+
+ sw t3, 0(a0)
+
+4:
+ RESTORE_REGS_FROM_STACK 0, s0, s1, s2
+5:
+ j ra
+ nop
+
+END(pixman_composite_over_n_8888_asm_mips)
+
LEAF_MIPS_DSPR2(pixman_composite_add_8_8_8_asm_mips)
/*
* a0 - dst (a8)
diff --git a/pixman/pixman-mips-dspr2.c b/pixman/pixman-mips-dspr2.c
index 44565e7..afe4ec6 100644
--- a/pixman/pixman-mips-dspr2.c
+++ b/pixman/pixman-mips-dspr2.c
@@ -81,6 +81,11 @@ PIXMAN_MIPS_BIND_FAST_PATH_SRC_N_DST (SKIP_ZERO_MASK, over_0565_n_0565,
PIXMAN_MIPS_BIND_FAST_PATH_SRC_N_DST (SKIP_ZERO_MASK, add_8888_n_8888,
uint32_t, 1, uint32_t, 1)
+PIXMAN_MIPS_BIND_FAST_PATH_N_DST (SKIP_ZERO_SRC, over_n_0565,
+ uint16_t, 1)
+PIXMAN_MIPS_BIND_FAST_PATH_N_DST (SKIP_ZERO_SRC, over_n_8888,
+ uint32_t, 1)
+
PIXMAN_MIPS_BIND_FAST_PATH_SRC_MASK_DST (add_8_8_8, uint8_t, 1,
uint8_t, 1, uint8_t, 1)
PIXMAN_MIPS_BIND_FAST_PATH_SRC_MASK_DST (add_8888_8_8888, uint32_t, 1,
@@ -278,7 +283,9 @@ static const pixman_fast_path_t mips_dspr2_fast_paths[] =
PIXMAN_STD_FAST_PATH (OVER, solid, a8, x8b8g8r8, mips_composite_over_n_8_8888),
PIXMAN_STD_FAST_PATH (OVER, solid, a8, r5g6b5, mips_composite_over_n_8_0565),
PIXMAN_STD_FAST_PATH (OVER, solid, a8, b5g6r5, mips_composite_over_n_8_0565),
-
+ PIXMAN_STD_FAST_PATH (OVER, solid, null, r5g6b5, mips_composite_over_n_0565),
+ PIXMAN_STD_FAST_PATH (OVER, solid, null, a8r8g8b8, mips_composite_over_n_8888),
+ PIXMAN_STD_FAST_PATH (OVER, solid, null, x8r8g8b8, mips_composite_over_n_8888),
PIXMAN_STD_FAST_PATH (OVER, a8r8g8b8, solid, a8r8g8b8, mips_composite_over_8888_n_8888),
PIXMAN_STD_FAST_PATH (OVER, a8r8g8b8, solid, x8r8g8b8, mips_composite_over_8888_n_8888),
PIXMAN_STD_FAST_PATH (OVER, a8r8g8b8, solid, r5g6b5, mips_composite_over_8888_n_0565),
diff --git a/pixman/pixman-mips-dspr2.h b/pixman/pixman-mips-dspr2.h
index bddcfd8..186a501 100644
--- a/pixman/pixman-mips-dspr2.h
+++ b/pixman/pixman-mips-dspr2.h
@@ -85,6 +85,42 @@ mips_composite_##name (pixman_implementation_t *imp, \
} \
}
+/****************************************************************/
+
+#define PIXMAN_MIPS_BIND_FAST_PATH_N_DST(flags, name, \
+ dst_type, dst_cnt) \
+void \
+pixman_composite_##name##_asm_mips (dst_type *dst, \
+ uint32_t src, \
+ int32_t w); \
+ \
+static void \
+mips_composite_##name (pixman_implementation_t *imp, \
+ pixman_composite_info_t *info) \
+{ \
+ PIXMAN_COMPOSITE_ARGS (info); \
+ dst_type *dst_line, *dst; \
+ int32_t dst_stride; \
+ uint32_t src; \
+ \
+ src = _pixman_image_get_solid ( \
+ imp, src_image, dest_image->bits.format); \
+ \
+ if ((flags & SKIP_ZERO_SRC) && src == 0) \
+ return; \
+ \
+ PIXMAN_IMAGE_GET_LINE (dest_image, dest_x, dest_y, dst_type, \
+ dst_stride, dst_line, dst_cnt); \
+ \
+ while (height--) \
+ { \
+ dst = dst_line; \
+ dst_line += dst_stride; \
+ \
+ pixman_composite_##name##_asm_mips (dst, src, width); \
+ } \
+}
+
/*******************************************************************/
#define PIXMAN_MIPS_BIND_FAST_PATH_N_MASK_DST(flags, name, \
commit e33e9d3f55590c369c532b0305f928045e0a46cb
Author: Nemanja Lukic <nemanja.lukic at rt-rk.com>
Date: Mon Nov 12 22:48:52 2012 +0100
MIPS: DSPr2: Added more fast-paths for SRC operation:
Performance numbers before/after on MIPS-74kc @ 1GHz:
lowlevel-blt-bench results
Referent (before):
src_n_8_8888 = L1: 13.79 L2: 22.47 M: 17.55 ( 58.28%) HT: 6.95 VT: 6.46 R: 6.34 RT: 2.07 ( 20Kops/s)
src_n_8_8 = L1: 20.22 L2: 20.21 M: 18.20 ( 24.17%) HT: 6.65 VT: 6.22 R: 6.11 RT: 2.03 ( 20Kops/s)
Optimized:
src_n_8_8888 = L1: 58.31 L2: 53.34 M: 25.69 ( 85.29%) HT: 22.55 VT: 21.44 R: 19.91 RT: 10.34 ( 48Kops/s)
src_n_8_8 = L1: 102.60 L2: 89.43 M: 65.01 ( 86.32%) HT: 37.87 VT: 37.02 R: 32.43 RT: 12.41 ( 51Kops/s)
diff --git a/pixman/pixman-mips-dspr2-asm.S b/pixman/pixman-mips-dspr2-asm.S
index b5cae16..02adb6d 100644
--- a/pixman/pixman-mips-dspr2-asm.S
+++ b/pixman/pixman-mips-dspr2-asm.S
@@ -310,6 +310,139 @@ LEAF_MIPS_DSPR2(pixman_composite_src_x888_8888_asm_mips)
END(pixman_composite_src_x888_8888_asm_mips)
+LEAF_MIPS_DSPR2(pixman_composite_src_n_8_8888_asm_mips)
+/*
+ * a0 - dst (a8r8g8b8)
+ * a1 - src (32bit constant)
+ * a2 - mask (a8)
+ * a3 - w
+ */
+
+
+ SAVE_REGS_ON_STACK 0, v0
+ li v0, 0x00ff00ff
+
+ beqz a3, 3f
+ nop
+ addiu t1, a3, -1
+ beqz t1, 2f
+ nop
+
+1:
+ /* a1 = source (32bit constant) */
+ lbu t0, 0(a2) /* t2 = mask (a8) */
+ lbu t1, 1(a2) /* t3 = mask (a8) */
+ addiu a2, a2, 2
+
+ MIPS_2xUN8x4_MUL_2xUN8 a1, a1, t0, t1, t2, t3, v0, t4, t5, t6, t7, t8, t9
+
+ sw t2, 0(a0)
+ sw t3, 4(a0)
+ addiu a3, a3, -2
+ addiu t2, a3, -1
+ bgtz t2, 1b
+ addiu a0, a0, 8
+
+ beqz a3, 3f
+ nop
+
+2:
+ lbu t0, 0(a2)
+ addiu a2, a2, 1
+
+ MIPS_UN8x4_MUL_UN8 a1, t0, t1, v0, t3, t4, t5
+
+ sw t1, 0(a0)
+ addiu a3, a3, -1
+ addiu a0, a0, 4
+
+3:
+ RESTORE_REGS_FROM_STACK 0, v0
+ j ra
+ nop
+
+END(pixman_composite_src_n_8_8888_asm_mips)
+
+LEAF_MIPS_DSPR2(pixman_composite_src_n_8_8_asm_mips)
+/*
+ * a0 - dst (a8)
+ * a1 - src (32bit constant)
+ * a2 - mask (a8)
+ * a3 - w
+ */
+
+ li t9, 0x00ff00ff
+ beqz a3, 3f
+ nop
+ srl t7, a3, 2 /* t7 = how many multiples of 4 dst pixels */
+ beqz t7, 1f /* branch if less than 4 src pixels */
+ nop
+
+ srl t8, a1, 24
+ replv.ph t8, t8
+
+0:
+ beqz t7, 1f
+ addiu t7, t7, -1
+ lbu t0, 0(a2)
+ lbu t1, 1(a2)
+ lbu t2, 2(a2)
+ lbu t3, 3(a2)
+
+ addiu a2, a2, 4
+
+ precr_sra.ph.w t1, t0, 0
+ precr_sra.ph.w t3, t2, 0
+ precr.qb.ph t0, t3, t1
+
+ muleu_s.ph.qbl t2, t0, t8
+ muleu_s.ph.qbr t3, t0, t8
+ shra_r.ph t4, t2, 8
+ shra_r.ph t5, t3, 8
+ and t4, t4, t9
+ and t5, t5, t9
+ addq.ph t2, t2, t4
+ addq.ph t3, t3, t5
+ shra_r.ph t2, t2, 8
+ shra_r.ph t3, t3, 8
+ precr.qb.ph t2, t2, t3
+
+ sb t2, 0(a0)
+ srl t2, t2, 8
+ sb t2, 1(a0)
+ srl t2, t2, 8
+ sb t2, 2(a0)
+ srl t2, t2, 8
+ sb t2, 3(a0)
+ addiu a3, a3, -4
+ b 0b
+ addiu a0, a0, 4
+
+1:
+ beqz a3, 3f
+ nop
+ srl t8, a1, 24
+2:
+ lbu t0, 0(a2)
+ addiu a2, a2, 1
+
+ mul t2, t0, t8
+ shra_r.ph t3, t2, 8
+ andi t3, t3, 0x00ff
+ addq.ph t2, t2, t3
+ shra_r.ph t2, t2, 8
+
+ sb t2, 0(a0)
+ addiu a3, a3, -1
+ bnez a3, 2b
+ addiu a0, a0, 1
+
+3:
+ j ra
+ nop
+
+END(pixman_composite_src_n_8_8_asm_mips)
+
LEAF_MIPS_DSPR2(pixman_composite_over_n_8888_8888_ca_asm_mips)
/*
* a0 - dst (a8r8g8b8)
diff --git a/pixman/pixman-mips-dspr2.c b/pixman/pixman-mips-dspr2.c
index 9da636d..44565e7 100644
--- a/pixman/pixman-mips-dspr2.c
+++ b/pixman/pixman-mips-dspr2.c
@@ -55,6 +55,10 @@ PIXMAN_MIPS_BIND_FAST_PATH_SRC_DST (0, add_8_8,
PIXMAN_MIPS_BIND_FAST_PATH_SRC_DST (0, add_8888_8888,
uint32_t, 1, uint32_t, 1)
+PIXMAN_MIPS_BIND_FAST_PATH_N_MASK_DST (0, src_n_8_8888,
+ uint8_t, 1, uint32_t, 1)
+PIXMAN_MIPS_BIND_FAST_PATH_N_MASK_DST (0, src_n_8_8,
+ uint8_t, 1, uint8_t, 1)
PIXMAN_MIPS_BIND_FAST_PATH_N_MASK_DST (SKIP_ZERO_SRC, over_n_8888_8888_ca,
uint32_t, 1, uint32_t, 1)
PIXMAN_MIPS_BIND_FAST_PATH_N_MASK_DST (SKIP_ZERO_SRC, over_n_8888_0565_ca,
@@ -256,6 +260,11 @@ static const pixman_fast_path_t mips_dspr2_fast_paths[] =
PIXMAN_STD_FAST_PATH (SRC, x8r8g8b8, null, a8r8g8b8, mips_composite_src_x888_8888),
PIXMAN_STD_FAST_PATH (SRC, x8b8g8r8, null, a8b8g8r8, mips_composite_src_x888_8888),
PIXMAN_STD_FAST_PATH (SRC, r8g8b8, null, r8g8b8, mips_composite_src_0888_0888),
+ PIXMAN_STD_FAST_PATH (SRC, solid, a8, a8r8g8b8, mips_composite_src_n_8_8888),
+ PIXMAN_STD_FAST_PATH (SRC, solid, a8, x8r8g8b8, mips_composite_src_n_8_8888),
+ PIXMAN_STD_FAST_PATH (SRC, solid, a8, a8b8g8r8, mips_composite_src_n_8_8888),
+ PIXMAN_STD_FAST_PATH (SRC, solid, a8, x8b8g8r8, mips_composite_src_n_8_8888),
+ PIXMAN_STD_FAST_PATH (SRC, solid, a8, a8, mips_composite_src_n_8_8),
PIXMAN_STD_FAST_PATH_CA (OVER, solid, a8r8g8b8, a8r8g8b8, mips_composite_over_n_8888_8888_ca),
PIXMAN_STD_FAST_PATH_CA (OVER, solid, a8r8g8b8, x8r8g8b8, mips_composite_over_n_8888_8888_ca),
More information about the xorg-commit
mailing list