checkasm/yuv2yuv: add tests for semiplanar unscaled converters

4 files changed, 136 insertions, 1 deletions

diff --git a/tests/checkasm/Makefile b/tests/checkasm/Makefile
index 3a7670e24b..2d2e42e445 100644
--- a/tests/checkasm/Makefile
+++ b/tests/checkasm/Makefile
@@ -63,7 +63,7 @@ AVFILTEROBJS-$(CONFIG_SOBEL_FILTER)      += vf_convolution.o
 CHECKASMOBJS-$(CONFIG_AVFILTER) += $(AVFILTEROBJS-yes)
 
 # swscale tests
-SWSCALEOBJS                             += sw_gbrp.o sw_range_convert.o sw_rgb.o sw_scale.o sw_yuv2rgb.o
+SWSCALEOBJS                             += sw_gbrp.o sw_range_convert.o sw_rgb.o sw_scale.o sw_yuv2rgb.o sw_yuv2yuv.o
 
 CHECKASMOBJS-$(CONFIG_SWSCALE)  += $(SWSCALEOBJS)
 
diff --git a/tests/checkasm/checkasm.c b/tests/checkasm/checkasm.c
index 58597d3888..5c407de2ba 100644
--- a/tests/checkasm/checkasm.c
+++ b/tests/checkasm/checkasm.c
@@ -255,6 +255,7 @@ static const struct {
     { "sw_rgb", checkasm_check_sw_rgb },
     { "sw_scale", checkasm_check_sw_scale },
     { "sw_yuv2rgb", checkasm_check_sw_yuv2rgb },
+    { "sw_yuv2yuv", checkasm_check_sw_yuv2yuv },
 #endif
 #if CONFIG_AVUTIL
         { "fixed_dsp", checkasm_check_fixed_dsp },
diff --git a/tests/checkasm/checkasm.h b/tests/checkasm/checkasm.h
index 4d5f3e387e..3e73808739 100644
--- a/tests/checkasm/checkasm.h
+++ b/tests/checkasm/checkasm.h
@@ -123,6 +123,7 @@ void checkasm_check_sw_range_convert(void);
 void checkasm_check_sw_rgb(void);
 void checkasm_check_sw_scale(void);
 void checkasm_check_sw_yuv2rgb(void);
+void checkasm_check_sw_yuv2yuv(void);
 void checkasm_check_takdsp(void);
 void checkasm_check_utvideodsp(void);
 void checkasm_check_v210dec(void);
diff --git a/tests/checkasm/sw_yuv2yuv.c b/tests/checkasm/sw_yuv2yuv.c
new file mode 100644
index 0000000000..90a51601ed
--- /dev/null
+++ b/tests/checkasm/sw_yuv2yuv.c
@@ -0,0 +1,133 @@
+/*
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with FFmpeg; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+ */
+
+#include <string.h>
+
+#include "libavutil/common.h"
+#include "libavutil/intreadwrite.h"
+#include "libavutil/mem_internal.h"
+#include "libavutil/pixdesc.h"
+
+#include "libswscale/swscale.h"
+#include "libswscale/swscale_internal.h"
+
+#include "checkasm.h"
+
+#define randomize_buffers(buf, size)      \
+    do {                                  \
+        for (int j = 0; j < size; j += 4) \
+            AV_WN32(buf + j, rnd());      \
+    } while (0)
+
+static void check_semiplanar(int dst_pix_fmt)
+{
+    static const int src_fmts[] = {
+        AV_PIX_FMT_NV24,
+        AV_PIX_FMT_NV42,
+    };
+    const AVPixFmtDescriptor *dst_desc = av_pix_fmt_desc_get(dst_pix_fmt);
+#define NUM_LINES 4
+#define MAX_LINE_SIZE 1920
+    static const int input_sizes[] = {8, 128, 1080, MAX_LINE_SIZE};
+
+    declare_func_emms(AV_CPU_FLAG_MMX | AV_CPU_FLAG_MMXEXT,
+                      int, SwsContext *c, const uint8_t *src[],
+                           int srcStride[], int srcSliceY, int srcSliceH,
+                           uint8_t *dst[], int dstStride[]);
+
+    LOCAL_ALIGNED_8(uint8_t, src_y,  [MAX_LINE_SIZE * NUM_LINES]);
+    LOCAL_ALIGNED_8(uint8_t, src_uv, [MAX_LINE_SIZE * NUM_LINES * 2]);
+    const uint8_t *src[4] = { src_y, src_uv };
+
+    LOCAL_ALIGNED_8(uint8_t, dst0_y, [MAX_LINE_SIZE * NUM_LINES]);
+    LOCAL_ALIGNED_8(uint8_t, dst0_u, [MAX_LINE_SIZE * NUM_LINES / 2]);
+    LOCAL_ALIGNED_8(uint8_t, dst0_v, [MAX_LINE_SIZE * NUM_LINES / 2]);
+    uint8_t *dst0[4] = { dst0_y, dst0_u, dst0_v };
+
+    LOCAL_ALIGNED_8(uint8_t, dst1_y, [MAX_LINE_SIZE * NUM_LINES]);
+    LOCAL_ALIGNED_8(uint8_t, dst1_u, [MAX_LINE_SIZE * NUM_LINES / 2]);
+    LOCAL_ALIGNED_8(uint8_t, dst1_v, [MAX_LINE_SIZE * NUM_LINES / 2]);
+    uint8_t *dst1[4] = { dst1_y, dst1_u, dst1_v };
+
+    randomize_buffers(src_y,  MAX_LINE_SIZE * NUM_LINES);
+    randomize_buffers(src_uv, MAX_LINE_SIZE * NUM_LINES * 2);
+
+    for (int sfi = 0; sfi < FF_ARRAY_ELEMS(src_fmts); sfi++) {
+        int src_pix_fmt = src_fmts[sfi];
+        const AVPixFmtDescriptor *src_desc = av_pix_fmt_desc_get(src_pix_fmt);
+        for (int isi = 0; isi < FF_ARRAY_ELEMS(input_sizes); isi++) {
+            struct SwsContext *ctx;
+            int log_level;
+            int width = input_sizes[isi];
+            int srcSliceY = 0;
+            int srcSliceH = NUM_LINES;
+            int srcStride[4] = {
+                MAX_LINE_SIZE,
+                MAX_LINE_SIZE * 2,
+            };
+            int dstStride[4] = {
+                MAX_LINE_SIZE,
+                MAX_LINE_SIZE >> dst_desc->log2_chroma_w,
+                MAX_LINE_SIZE >> dst_desc->log2_chroma_w,
+            };
+
+            // override log level to prevent spamming of the message
+            // "No accelerated colorspace conversion found from %s to %s"
+            log_level = av_log_get_level();
+            av_log_set_level(AV_LOG_ERROR);
+            ctx = sws_getContext(width, srcSliceH, src_pix_fmt,
+                                 width, srcSliceH, dst_pix_fmt,
+                                 0, NULL, NULL, NULL);
+            av_log_set_level(log_level);
+            if (!ctx)
+                fail();
+
+            if (check_func(ctx->convert_unscaled, "%s_%s_%d", src_desc->name, dst_desc->name, width)) {
+                memset(dst0_y, 0xFF, MAX_LINE_SIZE * NUM_LINES);
+                memset(dst0_u, 0xFF, MAX_LINE_SIZE * NUM_LINES / 2);
+                memset(dst0_v, 0xFF, MAX_LINE_SIZE * NUM_LINES / 2);
+                memset(dst1_y, 0xFF, MAX_LINE_SIZE * NUM_LINES);
+                memset(dst1_u, 0xFF, MAX_LINE_SIZE * NUM_LINES / 2);
+                memset(dst1_v, 0xFF, MAX_LINE_SIZE * NUM_LINES / 2);
+
+                call_ref(ctx, src, srcStride, srcSliceY,
+                         srcSliceH, dst0, dstStride);
+                call_new(ctx, src, srcStride, srcSliceY,
+                         srcSliceH, dst1, dstStride);
+
+                if (memcmp(dst0_y, dst1_y, MAX_LINE_SIZE * NUM_LINES) ||
+                    memcmp(dst0_u, dst1_u, MAX_LINE_SIZE * NUM_LINES / 2) ||
+                    memcmp(dst0_v, dst1_v, MAX_LINE_SIZE * NUM_LINES / 2))
+                    fail();
+
+                bench_new(ctx, src, srcStride, srcSliceY,
+                          srcSliceH, dst0, dstStride);
+            }
+            sws_freeContext(ctx);
+        }
+    }
+}
+
+#undef NUM_LINES
+#undef MAX_LINE_SIZE
+
+void checkasm_check_sw_yuv2yuv(void)
+{
+    check_semiplanar(AV_PIX_FMT_YUV420P);
+    report("yuv420p");
+}