highperfocused/ffmpeg
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

swscale/swscale_unscaled : add X86_64 (SSE2 and AVX) for uyvyto422

Browse Source 
and checkasm test
This commit is contained in:
Martin Vignali
2018-04-22 19:14:23 +02:00
parent e6e4625862
commit 07a566e7d6
3 changed files with 219 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

50
tests/checkasm/sw_rgb.c
View File

@@ -35,8 +35,12 @@
} while (0)
static const uint8_t width[] = {12, 16, 20, 32, 36, 128};
static const struct {uint8_t w, h, s;} planes[] = {
{12,16,12}, {16,16,16}, {20,23,25}, {32,18,48}, {8,128,16}, {128,128,128}
};
#define MAX_STRIDE 128
#define MAX_HEIGHT 128
static void check_shuffle_bytes(void * func, const char * report)
{
@@ -64,6 +68,49 @@ static void check_shuffle_bytes(void * func, const char * report)
}
}
static void check_uyvy_to_422p()
{
int i;
LOCAL_ALIGNED_32(uint8_t, src0, [MAX_STRIDE * MAX_HEIGHT * 2]);
LOCAL_ALIGNED_32(uint8_t, src1, [MAX_STRIDE * MAX_HEIGHT * 2]);
LOCAL_ALIGNED_32(uint8_t, dst_y_0, [MAX_STRIDE * MAX_HEIGHT]);
LOCAL_ALIGNED_32(uint8_t, dst_y_1, [MAX_STRIDE * MAX_HEIGHT]);
LOCAL_ALIGNED_32(uint8_t, dst_u_0, [(MAX_STRIDE/2) * MAX_HEIGHT]);
LOCAL_ALIGNED_32(uint8_t, dst_u_1, [(MAX_STRIDE/2) * MAX_HEIGHT]);
LOCAL_ALIGNED_32(uint8_t, dst_v_0, [(MAX_STRIDE/2) * MAX_HEIGHT]);
LOCAL_ALIGNED_32(uint8_t, dst_v_1, [(MAX_STRIDE/2) * MAX_HEIGHT]);
declare_func_emms(AV_CPU_FLAG_MMX, void, uint8_t *ydst, uint8_t *udst, uint8_t *vdst,
const uint8_t *src, int width, int height,
int lumStride, int chromStride, int srcStride);
randomize_buffers(src0, MAX_STRIDE * MAX_HEIGHT * 2);
memcpy(src1, src0, MAX_STRIDE * MAX_HEIGHT * 2);
if (check_func(uyvytoyuv422, "uyvytoyuv422")) {
for (i = 0; i < 6; i ++) {
memset(dst_y_0, 0, MAX_STRIDE * MAX_HEIGHT);
memset(dst_y_1, 0, MAX_STRIDE * MAX_HEIGHT);
memset(dst_u_0, 0, (MAX_STRIDE/2) * MAX_HEIGHT);
memset(dst_u_1, 0, (MAX_STRIDE/2) * MAX_HEIGHT);
memset(dst_v_0, 0, (MAX_STRIDE/2) * MAX_HEIGHT);
memset(dst_v_1, 0, (MAX_STRIDE/2) * MAX_HEIGHT);
call_ref(dst_y_0, dst_u_0, dst_v_0, src0, planes[i].w, planes[i].h,
MAX_STRIDE, MAX_STRIDE / 2, planes[i].s);
call_new(dst_y_1, dst_u_1, dst_v_1, src1, planes[i].w, planes[i].h,
MAX_STRIDE, MAX_STRIDE / 2, planes[i].s);
if (memcmp(dst_y_0, dst_y_1, MAX_STRIDE * MAX_HEIGHT) ||
memcmp(dst_u_0, dst_u_1, (MAX_STRIDE/2) * MAX_HEIGHT) ||
memcmp(dst_v_0, dst_v_1, (MAX_STRIDE/2) * MAX_HEIGHT))
fail();
}
bench_new(dst_y_1, dst_u_1, dst_v_1, src1, planes[5].w, planes[5].h,
MAX_STRIDE, MAX_STRIDE / 2, planes[5].s);
}
}
void checkasm_check_sw_rgb(void)
{
ff_sws_rgb2rgb_init();
@@ -82,4 +129,7 @@ void checkasm_check_sw_rgb(void)
check_shuffle_bytes(shuffle_bytes_3210, "shuffle_bytes_3210");
report("shuffle_bytes_3210");
check_uyvy_to_422p();
report("uyvytoyuv422");
}