MPlayer-cvslog

Download

mplayer-cvslog@mplayerhq.hu

July 2019

1 participants
2 discussions

r38150 - trunk/sub/spudec.c
by reimar 04 Jul '19

04 Jul '19

Author: reimar Date: Thu Jul 4 14:28:02 2019 New Revision: 38150 Log: spudec: Increase image data alignment to 16 bytes. Allows using SSE2 aligned loads on the data. Switch to av_malloc and av_freep as well, so it also works if native malloc has a smaller default alignment. Modified: trunk/sub/spudec.c Modified: trunk/sub/spudec.c ============================================================================== --- trunk/sub/spudec.c Thu Jul 4 14:28:01 2019 (r38149) +++ trunk/sub/spudec.c Thu Jul 4 14:28:02 2019 (r38150) @@ -148,7 +148,7 @@ static packet_t *spudec_dequeue_packet(s static void spudec_free_packet(packet_t *packet) { - free(packet->packet); + av_freep(&packet->packet); free(packet); } @@ -228,19 +228,19 @@ static int spudec_alloc_image(spudec_han this->height = height; if (this->image_size < this->stride * this->height) { if (this->image != NULL) { - free(this->image); + av_freep(&this->image); this->image = NULL; - free(this->pal_image); + av_freep(&this->pal_image); this->pal_image = NULL; this->image_size = 0; this->pal_width = this->pal_height = 0; } - this->image = malloc(2 * this->stride * this->height); + this->image = av_malloc(2 * this->stride * this->height); if (this->image) { this->image_size = this->stride * this->height; this->aimage = this->image + this->image_size; // use stride here as well to simplify reallocation checks - this->pal_image = malloc(this->stride * this->height); + this->pal_image = av_malloc(this->stride * this->height); } } return this->image != NULL; @@ -279,7 +279,7 @@ static int apply_palette_crop(spudec_han int i; uint8_t *src; uint16_t pal[4]; - unsigned stride = (crop_w + 7) & ~7; + unsigned stride = (crop_w + 15) & ~15; int ret = 1; if (crop_x > this->pal_width || crop_y > this->pal_height || crop_w > this->pal_width - crop_x || crop_h > this->pal_width - crop_y || @@ -530,7 +530,7 @@ static void spudec_process_control(spude start_col = a >> 12; end_col = a & 0xfff; width = (end_col < start_col) ? 0 : end_col - start_col + 1; - stride = (width + 7) & ~7; /* Kludge: draw_alpha needs width multiple of 8 */ + stride = (width + 15) & ~15; /* Kludge: draw_alpha needs width multiple of 16 */ start_row = b >> 12; end_row = b & 0xfff; height = (end_row < start_row) ? 0 : end_row - start_row /* + 1 */; @@ -577,7 +577,7 @@ static void spudec_process_control(spude packet->alpha[i] = this->alpha[i]; packet->palette[i] = this->palette[i]; } - packet->packet = malloc(this->packet_size); + packet->packet = av_malloc(this->packet_size); memcpy(packet->packet, this->packet, this->packet_size); spudec_queue_packet(this, packet); } @@ -698,7 +698,7 @@ void spudec_heartbeat(void *this, unsign spu->start_pts = packet->start_pts; spu->end_pts = packet->end_pts; if (packet->is_decoded) { - free(spu->image); + av_freep(&spu->image); spu->image_size = packet->data_len; spu->image = packet->packet; spu->aimage = packet->packet + packet->stride * packet->height; @@ -928,14 +928,14 @@ void spudec_draw_scaled(void *me, unsign spu->scaled_start_row = spu->start_row * scaley / 0x100; spu->scaled_width = spu->width * scalex / 0x100; spu->scaled_height = spu->height * scaley / 0x100; - /* Kludge: draw_alpha needs width multiple of 8 */ - spu->scaled_stride = (spu->scaled_width + 7) & ~7; + /* Kludge: draw_alpha needs width multiple of 16 */ + spu->scaled_stride = (spu->scaled_width + 15) & ~15; if (spu->scaled_image_size < spu->scaled_stride * spu->scaled_height) { if (spu->scaled_image) { - free(spu->scaled_image); + av_freep(&spu->scaled_image); spu->scaled_image_size = 0; } - spu->scaled_image = malloc(2 * spu->scaled_stride * spu->scaled_height); + spu->scaled_image = av_malloc(2 * spu->scaled_stride * spu->scaled_height); if (spu->scaled_image) { spu->scaled_image_size = spu->scaled_stride * spu->scaled_height; spu->scaled_aimage = spu->scaled_image + spu->scaled_image_size; @@ -1347,12 +1347,12 @@ void spudec_free(void *this) spudec_free_packet(spudec_dequeue_packet(spu)); free(spu->packet); spu->packet = NULL; - free(spu->scaled_image); + av_freep(&spu->scaled_image); spu->scaled_image = NULL; - free(spu->image); + av_freep(&spu->image); spu->image = NULL; spu->aimage = NULL; - free(spu->pal_image); + av_freep(&spu->pal_image); spu->pal_image = NULL; spu->image_size = 0; spu->pal_width = spu->pal_height = 0; @@ -1374,7 +1374,7 @@ void spudec_set_hw_spu(void *this, const packet_t *spudec_packet_create(int x, int y, int w, int h) { packet_t *packet; - int stride = (w + 7) & ~7; + int stride = (w + 15) & ~15; if ((unsigned)w >= 0x8000 || (unsigned)h > 0x4000) return NULL; packet = calloc(1, sizeof(packet_t)); @@ -1386,7 +1386,7 @@ packet_t *spudec_packet_create(int x, in packet->start_row = y; packet->data_len = 2 * stride * h; if (packet->data_len) { // size 0 is a special "clear" packet - packet->packet = malloc(packet->data_len); + packet->packet = av_malloc(packet->data_len); if (!packet->packet) { free(packet); packet = NULL;

1 0

r38149 - trunk/sub/osd_template.c
by reimar 04 Jul '19

04 Jul '19

Author: reimar Date: Thu Jul 4 14:28:01 2019 New Revision: 38149 Log: osd_template: fix possible subtitle corruption. Stride padding of alpha is not guaranteed to be zerod, which can result in some corruption on the right border. Change SSE2 code to fall back to C code for the last part if width is not a multiple of 16. Modified: trunk/sub/osd_template.c Modified: trunk/sub/osd_template.c ============================================================================== --- trunk/sub/osd_template.c Sat Jun 29 00:05:15 2019 (r38148) +++ trunk/sub/osd_template.c Thu Jul 4 14:28:01 2019 (r38149) @@ -147,7 +147,7 @@ static inline void RENAME(vo_draw_alpha_ } #elif HAVE_SSE2 __m128i zero = _mm_setzero_si128(); - for(x=0;x+8<w;x+=16){ + for(x=0;x+15<w;x+=16){ __m128i mmsrc, mmdst, res; __m128i mmsrca = _mm_load_si128((const __m128i *)(srca + x)); @@ -242,7 +242,7 @@ static inline void RENAME(vo_draw_alpha_ __m128i zero = _mm_setzero_si128(); __m128i ymask = _mm_set1_epi16(0xff); __m128i uvofs = _mm_set1_epi16(0x8000); - for(x=0;x+12<w;x+=16){ + for(x=0;x+15<w;x+=16){ __m128i mmsrc, mmsrcalo, mmsrcahi, mmdst, mmdst2, mmlow, mmhigh, mmy, mmuv, res; __m128i mmsrca = _mm_load_si128((const __m128i *)(srca + x)); int alpha = _mm_movemask_epi8(_mm_cmpeq_epi8(mmsrca, zero)); @@ -323,7 +323,7 @@ static inline void RENAME(vo_draw_alpha_ #if HAVE_SSE2 __m128i zero = _mm_setzero_si128(); __m128i uvofs = _mm_set1_epi16(0x80); - for(x=0;x+12<w;x+=16){ + for(x=0;x+15<w;x+=16){ __m128i mmsrc, mmsrcalo, mmsrcahi, mmdst, mmdst2, mmlow, mmhigh, mmy, mmuv, res; __m128i mmsrca = _mm_load_si128((const __m128i *)(srca + x)); int alpha = _mm_movemask_epi8(_mm_cmpeq_epi8(mmsrca, zero)); @@ -459,7 +459,7 @@ static inline void RENAME(vo_draw_alpha_ #elif HAVE_SSE2 __m128i one_in_three_mask = _mm_set_epi32(0xff0000ffu, 0x0000ff00u, 0x00ff0000u, 0xff0000ffu); __m128i zero = _mm_setzero_si128(); - for(x=0;x+14<w;x+=16){ + for(x=0;x+15<w;x+=16){ __m128i mmsrc, mmtmp, mmtmpa, mmdst, res; __m128i mmsrca = _mm_load_si128((const __m128i *)(srca + x)); int alpha = _mm_movemask_epi8(_mm_cmpeq_epi8(mmsrca, zero));

1 0