remove alignment correction of the destination pointers in luma_16x6

interpolations, since they are always 16-bytes aligned in practice.
Add asserts to ease narrowing down potential image corructions on exotic plateforms

Based on a patch by Mauricio Alvarez % lokifo A gmail P com %
Original thread:
Date: Jun 26, 2007 1:07 PM
Subject: Re: [FFmpeg-devel] [PATCH] h264 luma interpolation 8x8 for altivec

Originally committed as revision 10238 to svn://svn.ffmpeg.org/ffmpeg/trunk
This commit is contained in:
Guillaume Poirier 2007-08-26 21:47:58 +00:00
parent 7bcc1d5b66
commit 16dae5173b
1 changed files with 19 additions and 42 deletions

View File

@ -18,6 +18,13 @@
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
//#define DEBUG_ALIGNMENT
#ifdef DEBUG_ALIGNMENT
#define ASSERT_ALIGNED(ptr) assert(((unsigned long)ptr&0x0000000F));
#else
#define ASSERT_ALIGNED(ptr) ;
#endif
/* this code assume that stride % 16 == 0 */
void PREFIX_h264_chroma_mc8_altivec(uint8_t * dst, uint8_t * src, int stride, int h, int x, int y) {
POWERPC_PERF_DECLARE(PREFIX_h264_chroma_mc8_num, 1);
@ -165,9 +172,6 @@ static void PREFIX_h264_qpel16_h_lowpass_altivec(uint8_t * dst, uint8_t * src, i
const vec_u16_t v5us = vec_splat_u16(5);
const vec_s16_t v20ss = vec_sl(vec_splat_s16(5),vec_splat_u16(2));
const vec_s16_t v16ss = vec_sl(vec_splat_s16(1),vec_splat_u16(4));
const vec_u8_t dstperm = vec_lvsr(0, dst);
const vec_u8_t neg1 = (const vec_u8_t) vec_splat_s8(-1);
const vec_u8_t dstmask = vec_perm(zero_u8v, neg1, dstperm);
vec_u8_t srcM2, srcM1, srcP0, srcP1, srcP2, srcP3;
@ -180,7 +184,7 @@ static void PREFIX_h264_qpel16_h_lowpass_altivec(uint8_t * dst, uint8_t * src, i
pp1A, pp1B, pp2A, pp2B, pp3A, pp3B,
psumA, psumB, sumA, sumB;
vec_u8_t sum, dst1, dst2, vdst, fsum, rsum, fdst1, fdst2;
vec_u8_t sum, vdst, fsum;
POWERPC_PERF_START_COUNT(PREFIX_h264_qpel16_h_lowpass_num, 1);
@ -282,18 +286,12 @@ static void PREFIX_h264_qpel16_h_lowpass_altivec(uint8_t * dst, uint8_t * src, i
sum = vec_packsu(sumA, sumB);
dst1 = vec_ld(0, dst);
dst2 = vec_ld(16, dst);
vdst = vec_perm(dst1, dst2, vec_lvsl(0, dst));
ASSERT_ALIGNED(dst);
vdst = vec_ld(0, dst);
OP_U8_ALTIVEC(fsum, sum, vdst);
rsum = vec_perm(fsum, fsum, dstperm);
fdst1 = vec_sel(dst1, rsum, dstmask);
fdst2 = vec_sel(rsum, dst2, dstmask);
vec_st(fdst1, 0, dst);
vec_st(fdst2, 16, dst);
vec_st(fsum, 0, dst);
src += srcStride;
dst += dstStride;
@ -313,9 +311,6 @@ static void PREFIX_h264_qpel16_v_lowpass_altivec(uint8_t * dst, uint8_t * src, i
const vec_u16_t v5us = vec_splat_u16(5);
const vec_s16_t v5ss = vec_splat_s16(5);
const vec_s16_t v16ss = vec_sl(vec_splat_s16(1),vec_splat_u16(4));
const vec_u8_t dstperm = vec_lvsr(0, dst);
const vec_u8_t neg1 = (const vec_u8_t)vec_splat_s8(-1);
const vec_u8_t dstmask = vec_perm(zero_u8v, neg1, dstperm);
uint8_t *srcbis = src - (srcStride * 2);
@ -356,7 +351,7 @@ static void PREFIX_h264_qpel16_v_lowpass_altivec(uint8_t * dst, uint8_t * src, i
srcP3ssA, srcP3ssB,
sum1A, sum1B, sum2A, sum2B, sum3A, sum3B;
vec_u8_t sum, dst1, dst2, vdst, fsum, rsum, fdst1, fdst2, srcP3a, srcP3b, srcP3;
vec_u8_t sum, vdst, fsum, srcP3a, srcP3b, srcP3;
POWERPC_PERF_START_COUNT(PREFIX_h264_qpel16_v_lowpass_num, 1);
@ -403,18 +398,12 @@ static void PREFIX_h264_qpel16_v_lowpass_altivec(uint8_t * dst, uint8_t * src, i
sum = vec_packsu(sumA, sumB);
dst1 = vec_ld(0, dst);
dst2 = vec_ld(16, dst);
vdst = vec_perm(dst1, dst2, vec_lvsl(0, dst));
ASSERT_ALIGNED(dst);
vdst = vec_ld(0, dst);
OP_U8_ALTIVEC(fsum, sum, vdst);
rsum = vec_perm(fsum, fsum, dstperm);
fdst1 = vec_sel(dst1, rsum, dstmask);
fdst2 = vec_sel(rsum, dst2, dstmask);
vec_st(fdst1, 0, dst);
vec_st(fdst2, 16, dst);
vec_st(fsum, 0, dst);
dst += dstStride;
}
@ -441,18 +430,12 @@ static void PREFIX_h264_qpel16_hv_lowpass_altivec(uint8_t * dst, int16_t * tmp,
register int align = ((((unsigned long)src) - 2) % 16);
const vec_u8_t neg1 = (const vec_u8_t) vec_splat_s8(-1);
vec_s16_t srcP0A, srcP0B, srcP1A, srcP1B,
srcP2A, srcP2B, srcP3A, srcP3B,
srcM1A, srcM1B, srcM2A, srcM2B,
sum1A, sum1B, sum2A, sum2B, sum3A, sum3B,
pp1A, pp1B, pp2A, pp2B, psumA, psumB;
const vec_u8_t dstperm = vec_lvsr(0, dst);
const vec_u8_t dstmask = vec_perm(zero_u8v, neg1, dstperm);
const vec_u8_t mperm = (const vec_u8_t)
AVV(0x00, 0x08, 0x01, 0x09, 0x02, 0x0A, 0x03, 0x0B,
0x04, 0x0C, 0x05, 0x0D, 0x06, 0x0E, 0x07, 0x0F);
@ -466,7 +449,7 @@ static void PREFIX_h264_qpel16_hv_lowpass_altivec(uint8_t * dst, int16_t * tmp,
pp3Ae, pp3Ao, pp3Be, pp3Bo, pp1cAe, pp1cAo, pp1cBe, pp1cBo,
pp32Ae, pp32Ao, pp32Be, pp32Bo, sumAe, sumAo, sumBe, sumBo,
ssumAe, ssumAo, ssumBe, ssumBo;
vec_u8_t fsum, sumv, sum, dst1, dst2, vdst, rsum, fdst1, fdst2;
vec_u8_t fsum, sumv, sum, vdst;
vec_s16_t ssume, ssumo;
POWERPC_PERF_START_COUNT(PREFIX_h264_qpel16_hv_lowpass_num, 1);
@ -650,18 +633,12 @@ static void PREFIX_h264_qpel16_hv_lowpass_altivec(uint8_t * dst, int16_t * tmp,
sumv = vec_packsu(ssume, ssumo);
sum = vec_perm(sumv, sumv, mperm);
dst1 = vec_ld(0, dst);
dst2 = vec_ld(16, dst);
vdst = vec_perm(dst1, dst2, vec_lvsl(0, dst));
ASSERT_ALIGNED(dst);
vdst = vec_ld(0, dst);
OP_U8_ALTIVEC(fsum, sum, vdst);
rsum = vec_perm(fsum, fsum, dstperm);
fdst1 = vec_sel(dst1, rsum, dstmask);
fdst2 = vec_sel(rsum, dst2, dstmask);
vec_st(fdst1, 0, dst);
vec_st(fdst2, 16, dst);
vec_st(fsum, 0, dst);
dst += dstStride;
}