diff mbox series

[FFmpeg-devel,v2,2/2] libswscale: add AVBSwapDSPContext and use

Message ID 20221205221641.1215-2-mindmark@gmail.com
State New
Headers show
Series [FFmpeg-devel,v2,1/2] avutil: move bswapdsp from avcodec to avutil | expand

Checks

Context Check Description
yinshiyou/make_loongarch64 success Make finished
yinshiyou/make_fate_loongarch64 success Make fate finished
andriy/make_x86 success Make finished
andriy/make_fate_x86 success Make fate finished

Commit Message

Mark Reid Dec. 5, 2022, 10:16 p.m. UTC
From: Mark Reid <mindmark@gmail.com>

There are some places in input.c that could use it too
but they aren't currently being pass the SwsContext
---
 libswscale/output.c           | 36 +++++++++++++++--------------------
 libswscale/swscale_internal.h |  3 +++
 libswscale/swscale_unscaled.c | 26 +++++++++----------------
 libswscale/utils.c            |  2 ++
 4 files changed, 29 insertions(+), 38 deletions(-)

Comments

Michael Niedermayer Dec. 6, 2022, 6:16 p.m. UTC | #1
On Mon, Dec 05, 2022 at 02:16:41PM -0800, mindmark@gmail.com wrote:
> From: Mark Reid <mindmark@gmail.com>
> 
> There are some places in input.c that could use it too
> but they aren't currently being pass the SwsContext
> ---
>  libswscale/output.c           | 36 +++++++++++++++--------------------
>  libswscale/swscale_internal.h |  3 +++
>  libswscale/swscale_unscaled.c | 26 +++++++++----------------
>  libswscale/utils.c            |  2 ++
>  4 files changed, 29 insertions(+), 38 deletions(-)

LGTM

thx

[...]
diff mbox series

Patch

diff --git a/libswscale/output.c b/libswscale/output.c
index 5c85bff971..cd44081e3d 100644
--- a/libswscale/output.c
+++ b/libswscale/output.c
@@ -2313,13 +2313,11 @@  yuv2gbrp_full_X_c(SwsContext *c, const int16_t *lumFilter,
         }
     }
     if (SH != 22 && (!isBE(c->dstFormat)) != (!HAVE_BIGENDIAN)) {
-        for (i = 0; i < dstW; i++) {
-            dest16[0][i] = av_bswap16(dest16[0][i]);
-            dest16[1][i] = av_bswap16(dest16[1][i]);
-            dest16[2][i] = av_bswap16(dest16[2][i]);
-            if (hasAlpha)
-                dest16[3][i] = av_bswap16(dest16[3][i]);
-        }
+        c->bsdsp.bswap16_buf(dest16[0], dest16[0], dstW);
+        c->bsdsp.bswap16_buf(dest16[1], dest16[1], dstW);
+        c->bsdsp.bswap16_buf(dest16[2], dest16[2], dstW);
+        if (hasAlpha)
+            c->bsdsp.bswap16_buf(dest16[3], dest16[3], dstW);
     }
 }
 
@@ -2385,13 +2383,11 @@  yuv2gbrp16_full_X_c(SwsContext *c, const int16_t *lumFilter,
             dest16[3][i] = av_clip_uintp2(A, 30) >> 14;
     }
     if ((!isBE(c->dstFormat)) != (!HAVE_BIGENDIAN)) {
-        for (i = 0; i < dstW; i++) {
-            dest16[0][i] = av_bswap16(dest16[0][i]);
-            dest16[1][i] = av_bswap16(dest16[1][i]);
-            dest16[2][i] = av_bswap16(dest16[2][i]);
-            if (hasAlpha)
-                dest16[3][i] = av_bswap16(dest16[3][i]);
-        }
+        c->bsdsp.bswap16_buf(dest16[0], dest16[0], dstW);
+        c->bsdsp.bswap16_buf(dest16[1], dest16[1], dstW);
+        c->bsdsp.bswap16_buf(dest16[2], dest16[2], dstW);
+        if (hasAlpha)
+            c->bsdsp.bswap16_buf(dest16[3], dest16[3], dstW);
     }
 }
 
@@ -2461,13 +2457,11 @@  yuv2gbrpf32_full_X_c(SwsContext *c, const int16_t *lumFilter,
             dest32[3][i] = av_float2int(float_mult * (float)(av_clip_uintp2(A, 30) >> 14));
     }
     if ((!isBE(c->dstFormat)) != (!HAVE_BIGENDIAN)) {
-        for (i = 0; i < dstW; i++) {
-            dest32[0][i] = av_bswap32(dest32[0][i]);
-            dest32[1][i] = av_bswap32(dest32[1][i]);
-            dest32[2][i] = av_bswap32(dest32[2][i]);
-            if (hasAlpha)
-                dest32[3][i] = av_bswap32(dest32[3][i]);
-        }
+        c->bsdsp.bswap32_buf(dest32[0], dest32[0], dstW);
+        c->bsdsp.bswap32_buf(dest32[1], dest32[1], dstW);
+        c->bsdsp.bswap32_buf(dest32[2], dest32[2], dstW);
+        if (hasAlpha)
+            c->bsdsp.bswap32_buf(dest32[3], dest32[3], dstW);
     }
 }
 
diff --git a/libswscale/swscale_internal.h b/libswscale/swscale_internal.h
index abeebbb002..400f0bc8ed 100644
--- a/libswscale/swscale_internal.h
+++ b/libswscale/swscale_internal.h
@@ -26,6 +26,7 @@ 
 #include "config.h"
 
 #include "libavutil/avassert.h"
+#include "libavutil/bswapdsp.h"
 #include "libavutil/common.h"
 #include "libavutil/frame.h"
 #include "libavutil/intreadwrite.h"
@@ -682,6 +683,8 @@  typedef struct SwsContext {
     atomic_int   data_unaligned_warned;
 
     Half2FloatTables *h2f_tables;
+
+    AVBSwapDSPContext bsdsp;
 } SwsContext;
 //FIXME check init (where 0)
 
diff --git a/libswscale/swscale_unscaled.c b/libswscale/swscale_unscaled.c
index 9af2e7ecc3..0010ab24d1 100644
--- a/libswscale/swscale_unscaled.c
+++ b/libswscale/swscale_unscaled.c
@@ -468,7 +468,7 @@  static int bswap_16bpc(SwsContext *c, const uint8_t *src[],
                               int srcStride[], int srcSliceY, int srcSliceH,
                               uint8_t *dst[], int dstStride[])
 {
-    int i, j, p;
+    int i, p;
 
     for (p = 0; p < 4; p++) {
         int srcstr = srcStride[p] / 2;
@@ -480,9 +480,7 @@  static int bswap_16bpc(SwsContext *c, const uint8_t *src[],
             continue;
         dstPtr += (srcSliceY >> c->chrDstVSubSample) * dststr;
         for (i = 0; i < (srcSliceH >> c->chrDstVSubSample); i++) {
-            for (j = 0; j < min_stride; j++) {
-                dstPtr[j] = av_bswap16(srcPtr[j]);
-            }
+            c->bsdsp.bswap16_buf(dstPtr, srcPtr, min_stride);
             srcPtr += srcstr;
             dstPtr += dststr;
         }
@@ -495,7 +493,7 @@  static int bswap_32bpc(SwsContext *c, const uint8_t *src[],
                               int srcStride[], int srcSliceY, int srcSliceH,
                               uint8_t *dst[], int dstStride[])
 {
-    int i, j, p;
+    int i, p;
 
     for (p = 0; p < 4; p++) {
         int srcstr = srcStride[p] / 4;
@@ -507,9 +505,7 @@  static int bswap_32bpc(SwsContext *c, const uint8_t *src[],
             continue;
         dstPtr += (srcSliceY >> c->chrDstVSubSample) * dststr;
         for (i = 0; i < (srcSliceH >> c->chrDstVSubSample); i++) {
-            for (j = 0; j < min_stride; j++) {
-                dstPtr[j] = av_bswap32(srcPtr[j]);
-            }
+            c->bsdsp.bswap32_buf(dstPtr, srcPtr, min_stride);
             srcPtr += srcstr;
             dstPtr += dststr;
         }
@@ -1616,19 +1612,17 @@  static int rgbToRgbWrapper(SwsContext *c, const uint8_t *src[], int srcStride[],
             conv(srcPtr, dstPtr + dstStride[0] * srcSliceY,
                  (srcSliceH - 1) * srcStride[0] + c->srcW * srcBpp);
         else {
-            int i, j;
+            int i;
             dstPtr += dstStride[0] * srcSliceY;
 
             for (i = 0; i < srcSliceH; i++) {
                 if(src_bswap) {
-                    for(j=0; j<c->srcW; j++)
-                        ((uint16_t*)c->formatConvBuffer)[j] = av_bswap16(((uint16_t*)srcPtr)[j]);
+                    c->bsdsp.bswap16_buf((uint16_t*)c->formatConvBuffer, (uint16_t*)srcPtr, c->srcW);
                     conv(c->formatConvBuffer, dstPtr, c->srcW * srcBpp);
                 }else
                     conv(srcPtr, dstPtr, c->srcW * srcBpp);
                 if(dst_bswap)
-                    for(j=0; j<c->srcW; j++)
-                        ((uint16_t*)dstPtr)[j] = av_bswap16(((uint16_t*)dstPtr)[j]);
+                    c->bsdsp.bswap16_buf((uint16_t*)dstPtr, (uint16_t*)dstPtr, c->srcW);
                 srcPtr += srcStride[0];
                 dstPtr += dstStride[0];
             }
@@ -1932,16 +1926,14 @@  static int planarCopyWrapper(SwsContext *c, const uint8_t *src[],
                       isBE(c->srcFormat) != isBE(c->dstFormat)) {
 
                 for (i = 0; i < height; i++) {
-                    for (j = 0; j < length; j++)
-                        ((uint16_t *) dstPtr)[j] = av_bswap16(((const uint16_t *) srcPtr)[j]);
+                    c->bsdsp.bswap16_buf((uint16_t *)dstPtr, (const uint16_t *)srcPtr, length);
                     srcPtr += srcStride[plane];
                     dstPtr += dstStride[plane];
                 }
             } else if (isFloat(c->srcFormat) && isFloat(c->dstFormat) &&
                        isBE(c->srcFormat) != isBE(c->dstFormat)) { /* swap float plane */
                 for (i = 0; i < height; i++) {
-                    for (j = 0; j < length; j++)
-                        ((uint32_t *) dstPtr)[j] = av_bswap32(((const uint32_t *) srcPtr)[j]);
+                    c->bsdsp.bswap32_buf((uint32_t *)dstPtr, (const uint32_t *)srcPtr, length);
                     srcPtr += srcStride[plane];
                     dstPtr += dstStride[plane];
                 }
diff --git a/libswscale/utils.c b/libswscale/utils.c
index 90734f66ef..0514062d85 100644
--- a/libswscale/utils.c
+++ b/libswscale/utils.c
@@ -1921,6 +1921,8 @@  static av_cold int sws_init_single_context(SwsContext *c, SwsFilter *srcFilter,
         return 0;
     }
 
+    av_bswapdsp_init(&c->bsdsp);
+
     /* unscaled special cases */
     if (unscaled && !usesHFilter && !usesVFilter &&
         (c->srcRange == c->dstRange || isAnyRGB(dstFormat) ||