igt: chamelium: Test YUV buffers using the Chamelium
v5 cover igt: chamelium: Test YUV buffers using the Chamelium
Hi,
Here is a first attempt at expanding the current Chamelium formats test to YUV formats.
It builds on top of the current infrastructure for the YUV convertions, and the format suppport of the Chamelium. It reworks igt_fb quite significantly to add generic support for YUV formats, and adds a bunch of new formats along the way.
Let me know what you think,
Maxime
Changes from v4:
- Added Paul Reviewed-by's
- Removed the patch that moves calc_fb_size for everybody, and added some asserts
Changes from v3:
- Fixed build breakage
Changes from v2:
- Added Paul Reviewed-by's
- Fixed a stride issue on the buffer allocation
- Take into account the plane bpp so that we don't overallocate
- Rebased on master
Changes from v1:
- Support more YUV formats than just NV12
- Rework the YUV convertion functions to add generic routines
- Took the comments into account.
Maxime Ripard (13):
igt: fb: Add subsampling parameters to the formats
igt: fb: Reduce tile size alignment for non intel platforms
igt: fb: generic YUV convertion function
igt: fb: Move i915 YUV buffer clearing code to a function
igt: fb: Refactor dumb buffer allocation path
igt: fb: Account for all planes bpp
igt: fb: Don't pass the stride when allocating a dumb, multi-planar
buffer
igt: fb: Clear YUV dumb buffers
igt: fb: Rework YUV i915 allocation path
igt: fb: Add a bunch of new YUV formats
igt: tests: chamelium: Start to unify tests
igt: tests: chamelium: Convert VGA tests to do_test_display
igt: tests: chamelium: Add YUV formats tests
lib/igt_fb.c | 830 ++++++++++++++++++++---------------------- tests/kms_chamelium.c | 254 +++++++------ 2 files changed, 528 insertions(+), 556 deletions(-)
Signature · 6 lines
2.20.1 _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
PATCH v5 1/13 igt: fb: Add subsampling parameters to the formats
In order to improve the YUV support, let's add the horizontal and vertical subsampling parameters to the IGT formats.
Reviewed-by Paul Kocialkowski <[email protected]>
Signed-off-by Maxime Ripard <[email protected]>
--- lib/igt_fb.c | 35 +++++++++++++++++++++++++++++------ 1 file changed, 29 insertions(+), 6 deletions(-)
lib/igt_fb.c +29 -6
@@ -71,85 +71,104 @@ static const struct format_desc_struct {int depth;int num_planes;int plane_bpp[4 ]; +uint8_t hsub; +uint8_t vsub; } format_desc[] = { { .name = "ARGB1555", .depth = -1, .drm_id = DRM_FORMAT_ARGB1555, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_a1r5g5b5, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 1, .vsub = 1, }, { .name = "XRGB1555", .depth = -1, .drm_id = DRM_FORMAT_XRGB1555, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_x1r5g5b5, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 1, .vsub = 1, }, { .name = "RGB565", .depth = 16, .drm_id = DRM_FORMAT_RGB565, .cairo_id = CAIRO_FORMAT_RGB16_565, .pixman_id = PIXMAN_r5g6b5, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 1, .vsub = 1, }, { .name = "BGR565", .depth = -1, .drm_id = DRM_FORMAT_BGR565, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_b5g6r5, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 1, .vsub = 1, }, { .name = "BGR888", .depth = -1, .drm_id = DRM_FORMAT_BGR888, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_b8g8r8, .num_planes = 1, .plane_bpp = { 24, }, + .hsub = 1, .vsub = 1, }, { .name = "RGB888", .depth = -1, .drm_id = DRM_FORMAT_RGB888, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_r8g8b8, .num_planes = 1, .plane_bpp = { 24, }, + .hsub = 1, .vsub = 1, }, { .name = "XYUV8888", .depth = -1, .drm_id = DRM_FORMAT_XYUV8888, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "XRGB8888", .depth = 24, .drm_id = DRM_FORMAT_XRGB8888, .cairo_id = CAIRO_FORMAT_RGB24, .pixman_id = PIXMAN_x8r8g8b8, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "XBGR8888", .depth = -1, .drm_id = DRM_FORMAT_XBGR8888, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_x8b8g8r8, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "XRGB2101010", .depth = 30, .drm_id = DRM_FORMAT_XRGB2101010, .cairo_id = CAIRO_FORMAT_RGB30, .pixman_id = PIXMAN_x2r10g10b10, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "ARGB8888", .depth = 32, .drm_id = DRM_FORMAT_ARGB8888, .cairo_id = CAIRO_FORMAT_ARGB32, .pixman_id = PIXMAN_a8r8g8b8, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "ABGR8888", .depth = -1, .drm_id = DRM_FORMAT_ABGR8888, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_a8b8g8r8, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "NV12", .depth = -1, .drm_id = DRM_FORMAT_NV12, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 2, .plane_bpp = { 8, 16, }, + .hsub = 2, .vsub = 2, }, { .name = "YUYV", .depth = -1, .drm_id = DRM_FORMAT_YUYV, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 2, .vsub = 1, }, { .name = "YVYU", .depth = -1, .drm_id = DRM_FORMAT_YVYU, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 2, .vsub = 1, }, { .name = "UYVY", .depth = -1, .drm_id = DRM_FORMAT_UYVY, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 2, .vsub = 1, }, { .name = "VYUY", .depth = -1, .drm_id = DRM_FORMAT_VYUY, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 2, .vsub = 1, }, }; #define for_each_format(f) \
@@ -239,10 +258,12 @@ void igt_get_fb_tile_size(int fd, uint64_t tiling, int fb_bpp,static unsigned fb_plane_width (const struct igt_fb *fb,int plane) { -if (fb->drm_format ==DRM_FORMAT_NV12 && plane ==1 ) -return DIV_ROUND_UP (fb->width,2 ); +const struct format_desc_struct *format =lookup_drm_format (fb->drm_format); -return fb->width; +if (plane ==0 ) +return fb->width; + +return DIV_ROUND_UP (fb->width, format->hsub); }static unsigned fb_plane_bpp (const struct igt_fb *fb,int plane)
@@ -254,10 +275,12 @@ static unsigned fb_plane_bpp(const struct igt_fb *fb, int plane)static unsigned fb_plane_height (const struct igt_fb *fb,int plane) { -if (fb->drm_format ==DRM_FORMAT_NV12 && plane ==1 ) -return DIV_ROUND_UP (fb->height,2 ); +const struct format_desc_struct *format =lookup_drm_format (fb->drm_format); -return fb->height; +if (plane ==0 ) +return fb->height; + +return DIV_ROUND_UP (fb->height, format->vsub); }static int fb_num_planes (const struct igt_fb *fb)
Signature · 6 lines
2.20.1 _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 01/13] igt: fb: Add subsampling parameters to the formats
Reviewed-by Lyude Paul <[email protected]>
Quoted 155 lines · Maxime Ripard
In order to improve the YUV support, let's add the horizontal and vertical subsampling parameters to the IGT formats.
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 35 +++++++++++++++++++++++++++++------ 1 file changed, 29 insertions(+), 6 deletions(-)lib/igt_fb.c +19
-0@@ -71,85 +71,104 @@ static const struct format_desc_struct {int depth;int num_planes;int plane_bpp[4 ]; +uint8_t hsub; +uint8_t vsub; } format_desc[] = { { .name = "ARGB1555", .depth = -1, .drm_id = DRM_FORMAT_ARGB1555, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_a1r5g5b5, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 1, .vsub = 1, }, { .name = "XRGB1555", .depth = -1, .drm_id = DRM_FORMAT_XRGB1555, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_x1r5g5b5, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 1, .vsub = 1, }, { .name = "RGB565", .depth = 16, .drm_id = DRM_FORMAT_RGB565, .cairo_id = CAIRO_FORMAT_RGB16_565, .pixman_id = PIXMAN_r5g6b5, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 1, .vsub = 1, }, { .name = "BGR565", .depth = -1, .drm_id = DRM_FORMAT_BGR565, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_b5g6r5, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 1, .vsub = 1, }, { .name = "BGR888", .depth = -1, .drm_id = DRM_FORMAT_BGR888, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_b8g8r8, .num_planes = 1, .plane_bpp = { 24, }, + .hsub = 1, .vsub = 1, }, { .name = "RGB888", .depth = -1, .drm_id = DRM_FORMAT_RGB888, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_r8g8b8, .num_planes = 1, .plane_bpp = { 24, }, + .hsub = 1, .vsub = 1, }, { .name = "XYUV8888", .depth = -1, .drm_id = DRM_FORMAT_XYUV8888, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "XRGB8888", .depth = 24, .drm_id = DRM_FORMAT_XRGB8888, .cairo_id = CAIRO_FORMAT_RGB24, .pixman_id = PIXMAN_x8r8g8b8, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "XBGR8888", .depth = -1, .drm_id = DRM_FORMAT_XBGR8888, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_x8b8g8r8, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "XRGB2101010", .depth = 30, .drm_id = DRM_FORMAT_XRGB2101010, .cairo_id = CAIRO_FORMAT_RGB30, .pixman_id = PIXMAN_x2r10g10b10, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "ARGB8888", .depth = 32, .drm_id = DRM_FORMAT_ARGB8888, .cairo_id = CAIRO_FORMAT_ARGB32, .pixman_id = PIXMAN_a8r8g8b8, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "ABGR8888", .depth = -1, .drm_id = DRM_FORMAT_ABGR8888, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_a8b8g8r8, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "NV12", .depth = -1, .drm_id = DRM_FORMAT_NV12, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 2, .plane_bpp = { 8, 16, }, + .hsub = 2, .vsub = 2, }, { .name = "YUYV", .depth = -1, .drm_id = DRM_FORMAT_YUYV, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 2, .vsub = 1, }, { .name = "YVYU", .depth = -1, .drm_id = DRM_FORMAT_YVYU, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 2, .vsub = 1, }, { .name = "UYVY", .depth = -1, .drm_id = DRM_FORMAT_UYVY, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 2, .vsub = 1, }, { .name = "VYUY", .depth = -1, .drm_id = DRM_FORMAT_VYUY, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 2, .vsub = 1, }, };#define
for_each_format(f)\lib/igt_fb.c +5
-3@@ -239,10 +258,12 @@ void igt_get_fb_tile_size(int fd, uint64_t tiling, int fb_bpp, static unsigned fb_plane_width(const struct igt_fb *fb, int plane) { -if (fb->drm_format ==DRM_FORMAT_NV12 && plane ==1 ) -return DIV_ROUND_UP (fb->width,2 ); +const struct format_desc_struct *format =lookup_drm_format (fb- >drm_format); -return fb->width; +if (plane ==0 ) +return fb->width; + +return DIV_ROUND_UP (fb->width, format->hsub); }static unsigned
fb_plane_bpp(const struct igt_fb *fb, int plane)lib/igt_fb.c +5
-3@@ -254,10 +275,12 @@ static unsigned fb_plane_bpp(const struct igt_fb *fb, int plane)static unsigned fb_plane_height (const struct igt_fb *fb,int plane) { -if (fb->drm_format ==DRM_FORMAT_NV12 && plane ==1 ) -return DIV_ROUND_UP (fb->height,2 ); +const struct format_desc_struct *format =lookup_drm_format (fb- >drm_format); -return fb->height; +if (plane ==0 ) +return fb->height; + +return DIV_ROUND_UP (fb->height, format->vsub); }static int
fb_num_planes(const struct igt_fb *fb)
Signature · 7 lines
Cheers, Lyude Paul _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 01/13] igt: fb: Add subsampling parameters to the formats
In reply to Maxime RipardHi,
Quoted 2 lines · Maxime Ripard
In order to improve the YUV support, let's add the horizontal and vertical subsampling parameters to the IGT formats.
Pushed, thanks!
Cheers,
Paul
Quoted 147 lines
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 35 +++++++++++++++++++++++++++++------ 1 file changed, 29 insertions(+), 6 deletions(-)lib/igt_fb.c +29
-6@@ -71,85 +71,104 @@ static const struct format_desc_struct {int depth;int num_planes;int plane_bpp[4 ]; +uint8_t hsub; +uint8_t vsub; } format_desc[] = { { .name = "ARGB1555", .depth = -1, .drm_id = DRM_FORMAT_ARGB1555, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_a1r5g5b5, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 1, .vsub = 1, }, { .name = "XRGB1555", .depth = -1, .drm_id = DRM_FORMAT_XRGB1555, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_x1r5g5b5, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 1, .vsub = 1, }, { .name = "RGB565", .depth = 16, .drm_id = DRM_FORMAT_RGB565, .cairo_id = CAIRO_FORMAT_RGB16_565, .pixman_id = PIXMAN_r5g6b5, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 1, .vsub = 1, }, { .name = "BGR565", .depth = -1, .drm_id = DRM_FORMAT_BGR565, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_b5g6r5, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 1, .vsub = 1, }, { .name = "BGR888", .depth = -1, .drm_id = DRM_FORMAT_BGR888, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_b8g8r8, .num_planes = 1, .plane_bpp = { 24, }, + .hsub = 1, .vsub = 1, }, { .name = "RGB888", .depth = -1, .drm_id = DRM_FORMAT_RGB888, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_r8g8b8, .num_planes = 1, .plane_bpp = { 24, }, + .hsub = 1, .vsub = 1, }, { .name = "XYUV8888", .depth = -1, .drm_id = DRM_FORMAT_XYUV8888, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "XRGB8888", .depth = 24, .drm_id = DRM_FORMAT_XRGB8888, .cairo_id = CAIRO_FORMAT_RGB24, .pixman_id = PIXMAN_x8r8g8b8, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "XBGR8888", .depth = -1, .drm_id = DRM_FORMAT_XBGR8888, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_x8b8g8r8, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "XRGB2101010", .depth = 30, .drm_id = DRM_FORMAT_XRGB2101010, .cairo_id = CAIRO_FORMAT_RGB30, .pixman_id = PIXMAN_x2r10g10b10, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "ARGB8888", .depth = 32, .drm_id = DRM_FORMAT_ARGB8888, .cairo_id = CAIRO_FORMAT_ARGB32, .pixman_id = PIXMAN_a8r8g8b8, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "ABGR8888", .depth = -1, .drm_id = DRM_FORMAT_ABGR8888, .cairo_id = CAIRO_FORMAT_INVALID, .pixman_id = PIXMAN_a8b8g8r8, .num_planes = 1, .plane_bpp = { 32, }, + .hsub = 1, .vsub = 1, }, { .name = "NV12", .depth = -1, .drm_id = DRM_FORMAT_NV12, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 2, .plane_bpp = { 8, 16, }, + .hsub = 2, .vsub = 2, }, { .name = "YUYV", .depth = -1, .drm_id = DRM_FORMAT_YUYV, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 2, .vsub = 1, }, { .name = "YVYU", .depth = -1, .drm_id = DRM_FORMAT_YVYU, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 2, .vsub = 1, }, { .name = "UYVY", .depth = -1, .drm_id = DRM_FORMAT_UYVY, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 2, .vsub = 1, }, { .name = "VYUY", .depth = -1, .drm_id = DRM_FORMAT_VYUY, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 16, }, + .hsub = 2, .vsub = 1, }, }; #define for_each_format(f) \@@ -239,10 +258,12 @@ void igt_get_fb_tile_size(int fd, uint64_t tiling, int fb_bpp,static unsigned fb_plane_width (const struct igt_fb *fb,int plane) { -if (fb->drm_format ==DRM_FORMAT_NV12 && plane ==1 ) -return DIV_ROUND_UP (fb->width,2 ); +const struct format_desc_struct *format =lookup_drm_format (fb->drm_format); -return fb->width; +if (plane ==0 ) +return fb->width; + +return DIV_ROUND_UP (fb->width, format->hsub); }static unsigned fb_plane_bpp (const struct igt_fb *fb,int plane)@@ -254,10 +275,12 @@ static unsigned fb_plane_bpp(const struct igt_fb *fb, int plane)static unsigned fb_plane_height (const struct igt_fb *fb,int plane) { -if (fb->drm_format ==DRM_FORMAT_NV12 && plane ==1 ) -return DIV_ROUND_UP (fb->height,2 ); +const struct format_desc_struct *format =lookup_drm_format (fb->drm_format); -return fb->height; +if (plane ==0 ) +return fb->height; + +return DIV_ROUND_UP (fb->height, format->vsub); }static int fb_num_planes (const struct igt_fb *fb)
Signature · 8 lines
Paul Kocialkowski, Bootlin Embedded Linux and kernel engineering https://bootlin.com _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
PATCH v5 2/13 igt: fb: Reduce tile size alignment for non intel platforms
In reply to Maxime RipardAligning the width on 64 pixels only make sense on intel platforms, make sure to add a check against this.
Reviewed-by Paul Kocialkowski <[email protected]>
Signed-off-by Maxime Ripard <[email protected]>
--- lib/igt_fb.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
lib/igt_fb.c +5 -1
@@ -204,7 +204,11 @@ void igt_get_fb_tile_size(int fd, uint64_t tiling, int fb_bpp, {switch (tiling) {case LOCAL_DRM_FORMAT_MOD_NONE : - *width_ret =64 ; +if (is_i915_device (fd)) + *width_ret =64 ; +else + *width_ret =1 ; + *height_ret =1 ;break ;case LOCAL_I915_FORMAT_MOD_X_TILED :
Signature · 6 lines
2.20.1 _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 02/13] igt: fb: Reduce tile size alignment for non intel platforms
Reviewed-by Lyude Paul <[email protected]>
Quoted 27 lines · Maxime Ripard
Aligning the width on 64 pixels only make sense on intel platforms, make sure to add a check against this.
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)lib/igt_fb.c +5
-1@@ -204,7 +204,11 @@ void igt_get_fb_tile_size(int fd, uint64_t tiling, int fb_bpp, {switch (tiling) {case LOCAL_DRM_FORMAT_MOD_NONE : - *width_ret =64 ; +if (is_i915_device (fd)) + *width_ret =64 ; +else + *width_ret =1 ; + *height_ret =1 ;break ;case
LOCAL_I915_FORMAT_MOD_X_TILED:
Signature · 7 lines
Cheers, Lyude Paul _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 02/13] igt: fb: Reduce tile size alignment for non intel platforms
In reply to Maxime RipardHi,
Quoted 2 lines · Maxime Ripard
Aligning the width on 64 pixels only make sense on intel platforms, make sure to add a check against this.
Pushed, thanks!
Cheers,
Paul
Quoted 23 lines
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)lib/igt_fb.c +5
-1@@ -204,7 +204,11 @@ void igt_get_fb_tile_size(int fd, uint64_t tiling, int fb_bpp, {switch (tiling) {case LOCAL_DRM_FORMAT_MOD_NONE : - *width_ret =64 ; +if (is_i915_device (fd)) + *width_ret =64 ; +else + *width_ret =1 ; + *height_ret =1 ;break ;case LOCAL_I915_FORMAT_MOD_X_TILED :
Signature · 8 lines
Paul Kocialkowski, Bootlin Embedded Linux and kernel engineering https://bootlin.com _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
PATCH v5 3/13 igt: fb: generic YUV convertion function
In reply to Maxime RipardThe current way we work when we want to support a new YUV format to IGT, we also need to add new function to convert to and from XRGB8888.
This doesn't really scale however, and creates a lot of code to maintain. In order to work around this, create a generic function to convert to RGB and one to convert from RGB.
The only thing that is needed now is to add new parameters, and that's it.
Reviewed-by Paul Kocialkowski <[email protected]>
Signed-off-by Maxime Ripard <[email protected]>
--- lib/igt_fb.c | 537 ++++++++++++++++----------------------------------- 1 file changed, 171 insertions(+), 366 deletions(-)
lib/igt_fb.c +171 -366
@@ -1577,424 +1577,237 @@ static void convert_src_put(const struct fb_convert *cvt,free (src_buf); } -static void convert_nv12_to_rgb24 (struct fb_convert *cvt) +struct yuv_parameters { +unsigned y_inc; +unsigned uv_inc; +unsigned y_stride; +unsigned uv_stride; +unsigned y_offset; +unsigned u_offset; +unsigned v_offset; +}; + +static void get_yuv_parameters (struct igt_fb *fb,struct yuv_parameters *params) { -int i, j; -const uint8_t *y, *uv; -uint8_t *rgb24 = cvt->dst.ptr; -unsigned int rgb24_stride = cvt->dst.fb->strides[0 ]; -unsigned int planar_stride = cvt->src.fb->strides[0 ]; -struct igt_mat4 m =igt_ycbcr_to_rgb_matrix (cvt->src.fb->color_encoding, - cvt->src.fb->color_range); -uint8_t *buf; - -igt_assert (cvt->src.fb->drm_format ==DRM_FORMAT_NV12 && - cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 ); - - buf =convert_src_get (cvt); - y = buf + cvt->src.fb->offsets[0 ]; - uv = buf + cvt->src.fb->offsets[1 ]; - -for (i =0 ; i < cvt->dst.fb->height /2 ; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x2 pixel blocks */ -struct igt_vec4 yuv[4 ]; -struct igt_vec4 rgb[4 ]; - - yuv[0 ].d[0 ] = y[j *2 +0 ]; - yuv[1 ].d[0 ] = y[j *2 +1 ]; - yuv[2 ].d[0 ] = y[j *2 +0 + planar_stride]; - yuv[3 ].d[0 ] = y[j *2 +1 + planar_stride]; - - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = yuv[2 ].d[1 ] = yuv[3 ].d[1 ] = uv[j *2 +0 ]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = yuv[2 ].d[2 ] = yuv[3 ].d[2 ] = uv[j *2 +1 ]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] = yuv[2 ].d[3 ] = yuv[3 ].d[3 ] =1.0f ; - - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); - rgb[2 ] =igt_matrix_transform (&m, &yuv[2 ]); - rgb[3 ] =igt_matrix_transform (&m, &yuv[3 ]); - -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +4 ], &rgb[1 ]); -write_rgb (&rgb24[j *8 +0 + rgb24_stride], &rgb[2 ]); -write_rgb (&rgb24[j *8 +4 + rgb24_stride], &rgb[3 ]); - } - -if (cvt->dst.fb->width &1 ) { -/* Convert 1x2 pixel block */ -struct igt_vec4 yuv[2 ]; -struct igt_vec4 rgb[2 ]; - - yuv[0 ].d[0 ] = y[j *2 +0 ]; - yuv[1 ].d[0 ] = y[j *2 +0 + planar_stride]; +igt_assert (igt_format_is_yuv (fb->drm_format)); - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = uv[j *2 +0 ]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = uv[j *2 +1 ]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] =1.0f ; +switch (fb->drm_format) { +case DRM_FORMAT_NV12 : + params->y_inc =1 ; + params->uv_inc =2 ; +break ; - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); - -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +0 + rgb24_stride], &rgb[1 ]); - } - - rgb24 +=2 * rgb24_stride; - y +=2 * planar_stride; - uv += planar_stride; +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVYU : +case DRM_FORMAT_UYVY : +case DRM_FORMAT_VYUY : + params->y_inc =2 ; + params->uv_inc =4 ; +break ; + +case DRM_FORMAT_XYUV8888 : + params->y_inc =4 ; + params->uv_inc =4 ; +break ; } -if (cvt->dst.fb->height &1 ) { -/* Convert last row */ -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 yuv[2 ]; -struct igt_vec4 rgb[2 ]; - - yuv[0 ].d[0 ] = y[j *2 +0 ]; - yuv[1 ].d[0 ] = y[j *2 +1 ]; - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = uv[j *2 +0 ]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = uv[j *2 +1 ]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] =1.0f ; +switch (fb->drm_format) { +case DRM_FORMAT_NV12 : + params->y_stride = fb->strides[0 ]; + params->uv_stride = fb->strides[1 ]; +break ; - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVYU : +case DRM_FORMAT_UYVY : +case DRM_FORMAT_VYUY : +case DRM_FORMAT_XYUV8888 : + params->y_stride = fb->strides[0 ]; + params->uv_stride = fb->strides[0 ]; +break ; + } -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +4 ], &rgb[0 ]); - } +switch (fb->drm_format) { +case DRM_FORMAT_NV12 : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[1 ]; + params->v_offset = fb->offsets[1 ] +1 ; +break ; -if (cvt->dst.fb->width &1 ) { -/* Convert single pixel */ -struct igt_vec4 yuv; -struct igt_vec4 rgb; +case DRM_FORMAT_YUYV : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[0 ] +1 ; + params->v_offset = fb->offsets[0 ] +3 ; +break ; - yuv.d[0 ] = y[j *2 +0 ]; - yuv.d[1 ] = uv[j *2 +0 ]; - yuv.d[2 ] = uv[j *2 +1 ]; - yuv.d[3 ] =1.0f ; +case DRM_FORMAT_YVYU : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[0 ] +3 ; + params->v_offset = fb->offsets[0 ] +1 ; +break ; - rgb =igt_matrix_transform (&m, &yuv); +case DRM_FORMAT_UYVY : + params->y_offset = fb->offsets[0 ] +1 ; + params->u_offset = fb->offsets[0 ]; + params->v_offset = fb->offsets[0 ] +2 ; +break ; -write_rgb (&rgb24[j *8 +0 ], &rgb); - } +case DRM_FORMAT_VYUY : + params->y_offset = fb->offsets[0 ] +1 ; + params->u_offset = fb->offsets[0 ] +2 ; + params->v_offset = fb->offsets[0 ]; +break ; + +case DRM_FORMAT_XYUV8888 : + params->y_offset = fb->offsets[0 ] +1 ; + params->u_offset = fb->offsets[0 ] +2 ; + params->v_offset = fb->offsets[0 ] +3 ; +break ; } - -convert_src_put (cvt, buf); } -static void convert_yuv444_to_rgb24 (struct fb_convert *cvt) +static void convert_yuv_to_rgb24 (struct fb_convert *cvt) { +const struct format_desc_struct *src_fmt = +lookup_drm_format (cvt->src.fb->drm_format);int i, j; -uint8_t *yuv24; +uint8_t bpp =4 ; +uint8_t *y, *u, *v;uint8_t *rgb24 = cvt->dst.ptr; -unsigned rgb24_stride = cvt->dst.fb->strides[0 ], xyuv_stride = cvt->src.fb->strides[0 ]; -uint8_t *buf =malloc (cvt->src.fb->size); +unsigned int rgb24_stride = cvt->dst.fb->strides[0 ];struct igt_mat4 m =igt_ycbcr_to_rgb_matrix (cvt->src.fb->color_encoding, cvt->src.fb->color_range); +uint8_t *buf; +struct yuv_parameters params = { }; -/* -* Reading from the BO is awfully slow because of lack of read caching, -* it's faster to copy the whole BO to a temporary buffer and convert -* from there. -*/ -igt_memcpy_from_wc (buf, cvt->src.ptr + cvt->src.fb->offsets[0 ], cvt->src.fb->size); - yuv24 = buf; +igt_assert (cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 && +igt_format_is_yuv (cvt->src.fb->drm_format)); + + buf =convert_src_get (cvt); +get_yuv_parameters (cvt->src.fb, ¶ms); + y = buf + params.y_offset; + u = buf + params.u_offset; + v = buf + params.v_offset;for (i =0 ; i < cvt->dst.fb->height; i++) { +const uint8_t *y_tmp = y; +const uint8_t *u_tmp = u; +const uint8_t *v_tmp = v; +uint8_t *rgb_tmp = rgb24; +for (j =0 ; j < cvt->dst.fb->width; j++) { -float y, u, v; -struct igt_vec4 yuv; -struct igt_vec4 rgb; - - v = yuv24[i * xyuv_stride + j *4 ]; - u = yuv24[i * xyuv_stride + j *4 +1 ]; - y = yuv24[i * xyuv_stride + j *4 +2 ]; - yuv.d[0 ] = y; - yuv.d[1 ] = u; - yuv.d[2 ] = v; +struct igt_vec4 rgb, yuv; + + yuv.d[0 ] = *y_tmp; + yuv.d[1 ] = *u_tmp; + yuv.d[2 ] = *v_tmp; yuv.d[3 ] =1.0f ; rgb =igt_matrix_transform (&m, &yuv); +write_rgb (rgb_tmp, &rgb); -write_rgb (&rgb24[i * rgb24_stride + j *4 ], &rgb); - } - } - -free (buf); -} - - -static void convert_rgb24_to_yuv444 (struct fb_convert *cvt) -{ -int i, j; -uint8_t *rgb24; -uint8_t *yuv444 = cvt->dst.ptr + cvt->dst.fb->offsets[0 ]; -unsigned int rgb24_stride = cvt->src.fb->strides[0 ], xyuv_stride = cvt->dst.fb->strides[0 ]; -struct igt_mat4 m =igt_rgb_to_ycbcr_matrix (cvt->dst.fb->color_encoding, - cvt->dst.fb->color_range); - - rgb24 = cvt->src.ptr; + rgb_tmp += bpp; + y_tmp += params.y_inc; -igt_assert_f (cvt->dst.fb->drm_format ==DRM_FORMAT_XYUV8888 , -"Conversion not implemented for !XYUV packed formats\n" ); - -for (i =0 ; i < cvt->dst.fb->height; i++) { -for (j =0 ; j < cvt->dst.fb->width; j++) { -struct igt_vec4 rgb; -struct igt_vec4 yuv; - -read_rgb (&rgb, &rgb24[i * rgb24_stride + j *4 ]); +if ((src_fmt->hsub ==1 ) || (j % src_fmt->hsub)) { + u_tmp += params.uv_inc; + v_tmp += params.uv_inc; + } + } - yuv =igt_matrix_transform (&m, &rgb); + rgb24 += rgb24_stride; + y += params.y_stride; - yuv444[i * xyuv_stride + j *4 ] = yuv.d[2 ]; - yuv444[i * xyuv_stride + j *4 +1 ] = yuv.d[1 ]; - yuv444[i * xyuv_stride + j *4 +2 ] = yuv.d[0 ]; +if ((src_fmt->vsub ==1 ) || (i % src_fmt->vsub)) { + u += params.uv_stride; + v += params.uv_stride; } } + +convert_src_put (cvt, buf); } -static void convert_rgb24_to_nv12 (struct fb_convert *cvt) +static void convert_rgb24_to_yuv (struct fb_convert *cvt) { +const struct format_desc_struct *dst_fmt = +lookup_drm_format (cvt->dst.fb->drm_format);int i, j; -uint8_t *y = cvt->dst.ptr + cvt->dst.fb->offsets[0 ]; -uint8_t *uv = cvt->dst.ptr + cvt->dst.fb->offsets[1 ]; +uint8_t *y, *u, *v;const uint8_t *rgb24 = cvt->src.ptr; +uint8_t bpp =4 ;unsigned rgb24_stride = cvt->src.fb->strides[0 ]; -unsigned planar_stride = cvt->dst.fb->strides[0 ];struct igt_mat4 m =igt_rgb_to_ycbcr_matrix (cvt->dst.fb->color_encoding, cvt->dst.fb->color_range); +struct yuv_parameters params = { };igt_assert (cvt->src.fb->drm_format ==DRM_FORMAT_XRGB8888 && - cvt->dst.fb->drm_format ==DRM_FORMAT_NV12 ); +igt_format_is_yuv (cvt->dst.fb->drm_format)); -for (i =0 ; i < cvt->dst.fb->height /2 ; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x2 pixel blocks */ -struct igt_vec4 rgb[4 ]; -struct igt_vec4 yuv[4 ]; +get_yuv_parameters (cvt->dst.fb, ¶ms); + y = cvt->dst.ptr + params.y_offset; + u = cvt->dst.ptr + params.u_offset; + v = cvt->dst.ptr + params.v_offset; -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +4 ]); -read_rgb (&rgb[2 ], &rgb24[j *8 +0 + rgb24_stride]); -read_rgb (&rgb[3 ], &rgb24[j *8 +4 + rgb24_stride]); - - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); - yuv[2 ] =igt_matrix_transform (&m, &rgb[2 ]); - yuv[3 ] =igt_matrix_transform (&m, &rgb[3 ]); - - y[j *2 +0 ] = yuv[0 ].d[0 ]; - y[j *2 +1 ] = yuv[1 ].d[0 ]; - y[j *2 +0 + planar_stride] = yuv[2 ].d[0 ]; - y[j *2 +1 + planar_stride] = yuv[3 ].d[0 ]; +for (i =0 ; i < cvt->dst.fb->height; i++) { +const uint8_t *rgb_tmp = rgb24; +uint8_t *y_tmp = y; +uint8_t *u_tmp = u; +uint8_t *v_tmp = v; -/* -* We assume the MPEG2 chroma siting convention, where -* pixel center for Cb'Cr' is between the left top and -* bottom pixel in a 2x2 block, so take the average. -*/ - uv[j *2 +0 ] = (yuv[0 ].d[1 ] + yuv[2 ].d[1 ]) /2.0f ; - uv[j *2 +1 ] = (yuv[0 ].d[2 ] + yuv[2 ].d[2 ]) /2.0f ; - } +for (j =0 ; j < cvt->dst.fb->width; j++) { +const uint8_t *pair_rgb24 = rgb_tmp; +struct igt_vec4 pair_rgb, rgb; +struct igt_vec4 pair_yuv, yuv; -if (cvt->dst.fb->width &1 ) { -/* Convert 1x2 pixel block */ -struct igt_vec4 rgb[2 ]; -struct igt_vec4 yuv[2 ]; +read_rgb (&rgb, rgb_tmp); + yuv =igt_matrix_transform (&m, &rgb); -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +0 + rgb24_stride]); + rgb_tmp += bpp; - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); + *y_tmp = yuv.d[0 ]; + y_tmp += params.y_inc; - y[j *2 +0 ] = yuv[0 ].d[0 ]; - y[j *2 +0 + planar_stride] = yuv[1 ].d[0 ]; +if ((i % dst_fmt->vsub) || (j % dst_fmt->hsub)) +continue ;/* * We assume the MPEG2 chroma siting convention, where * pixel center for Cb'Cr' is between the left top and * bottom pixel in a 2x2 block, so take the average. +* +* Therefore, if we use subsampling, we only really care +* about two pixels all the time, either the two +* subsequent pixels horizontally, vertically, or the +* two corners in a 2x2 block. +* +* The only corner case is when we have an odd number of +* pixels, but this can be handled pretty easily by not +* incrementing the paired pixel pointer in the +* direction it's odd in. */ - uv[j *2 +0 ] = (yuv[0 ].d[1 ] + yuv[1 ].d[1 ]) /2.0f ; - uv[j *2 +1 ] = (yuv[0 ].d[2 ] + yuv[1 ].d[2 ]) /2.0f ; - } - - rgb24 +=2 * rgb24_stride; - y +=2 * planar_stride; - uv += planar_stride; - } - -/* Last row cannot be interpolated between 2 pixels, take the single value */ -if (cvt->dst.fb->height &1 ) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 rgb[2 ]; -struct igt_vec4 yuv[2 ]; - -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +4 ]); - - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); +if (j != (cvt->dst.fb->width -1 )) + pair_rgb24 += (dst_fmt->hsub -1 ) * bpp; - y[j *2 +0 ] = yuv[0 ].d[0 ]; - y[j *2 +1 ] = yuv[1 ].d[0 ]; - uv[j *2 +0 ] = yuv[0 ].d[1 ]; - uv[j *2 +1 ] = yuv[0 ].d[2 ]; - } - -if (cvt->dst.fb->width &1 ) { -/* Convert single pixel */ -struct igt_vec4 rgb; -struct igt_vec4 yuv; +if (i != (cvt->dst.fb->height -1 )) + pair_rgb24 += rgb24_stride * (dst_fmt->vsub -1 ); -read_rgb (&rgb, &rgb24[j *8 +0 ]); +read_rgb (&pair_rgb, pair_rgb24); + pair_yuv =igt_matrix_transform (&m, &pair_rgb); - yuv =igt_matrix_transform (&m, &rgb); + *u_tmp = (yuv.d[1 ] + pair_yuv.d[1 ]) /2.0f ; + *v_tmp = (yuv.d[2 ] + pair_yuv.d[2 ]) /2.0f ; - y[j *2 +0 ] = yuv.d[0 ]; - uv[j *2 +0 ] = yuv.d[1 ]; - uv[j *2 +1 ] = yuv.d[2 ]; - } - } -} - -/* { Y0, U, Y1, V } */ -static const unsigned char swizzle_yuyv[] = {0 ,1 ,2 ,3 }; -static const unsigned char swizzle_yvyu[] = {0 ,3 ,2 ,1 }; -static const unsigned char swizzle_uyvy[] = {1 ,0 ,3 ,2 }; -static const unsigned char swizzle_vyuy[] = {1 ,2 ,3 ,0 }; - -static const unsigned char *yuyv_swizzle (uint32_t format) -{ -switch (format) { -default : -case DRM_FORMAT_YUYV : -return swizzle_yuyv; -case DRM_FORMAT_YVYU : -return swizzle_yvyu; -case DRM_FORMAT_UYVY : -return swizzle_uyvy; -case DRM_FORMAT_VYUY : -return swizzle_vyuy; - } -} - -static void convert_yuyv_to_rgb24 (struct fb_convert *cvt) -{ -int i, j; -const uint8_t *yuyv; -uint8_t *rgb24 = cvt->dst.ptr; -unsigned int rgb24_stride = cvt->dst.fb->strides[0 ]; -unsigned int yuyv_stride = cvt->src.fb->strides[0 ]; -struct igt_mat4 m =igt_ycbcr_to_rgb_matrix (cvt->src.fb->color_encoding, - cvt->src.fb->color_range); -const unsigned char *swz =yuyv_swizzle (cvt->src.fb->drm_format); -uint8_t *buf; - -igt_assert ((cvt->src.fb->drm_format ==DRM_FORMAT_YUYV || - cvt->src.fb->drm_format ==DRM_FORMAT_UYVY || - cvt->src.fb->drm_format ==DRM_FORMAT_YVYU || - cvt->src.fb->drm_format ==DRM_FORMAT_VYUY ) && - cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 ); - - buf =convert_src_get (cvt); - yuyv = buf; - -for (i =0 ; i < cvt->dst.fb->height; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 yuv[2 ]; -struct igt_vec4 rgb[2 ]; - - yuv[0 ].d[0 ] = yuyv[j *4 + swz[0 ]]; - yuv[1 ].d[0 ] = yuyv[j *4 + swz[2 ]]; - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = yuyv[j *4 + swz[1 ]]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = yuyv[j *4 + swz[3 ]]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] =1.0f ; - - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); - -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +4 ], &rgb[1 ]); - } - -if (cvt->dst.fb->width &1 ) { -struct igt_vec4 yuv; -struct igt_vec4 rgb; - - yuv.d[0 ] = yuyv[j *4 + swz[0 ]]; - yuv.d[1 ] = yuyv[j *4 + swz[1 ]]; - yuv.d[2 ] = yuyv[j *4 + swz[3 ]]; - yuv.d[3 ] =1.0f ; - - rgb =igt_matrix_transform (&m, &yuv); - -write_rgb (&rgb24[j *8 +0 ], &rgb); + u_tmp += params.uv_inc; + v_tmp += params.uv_inc; } rgb24 += rgb24_stride; - yuyv += yuyv_stride; - } - -convert_src_put (cvt, buf); -} - -static void convert_rgb24_to_yuyv (struct fb_convert *cvt) -{ -int i, j; -uint8_t *yuyv = cvt->dst.ptr; -const uint8_t *rgb24 = cvt->src.ptr; -unsigned rgb24_stride = cvt->src.fb->strides[0 ]; -unsigned yuyv_stride = cvt->dst.fb->strides[0 ]; -struct igt_mat4 m =igt_rgb_to_ycbcr_matrix (cvt->dst.fb->color_encoding, - cvt->dst.fb->color_range); -const unsigned char *swz =yuyv_swizzle (cvt->dst.fb->drm_format); - -igt_assert (cvt->src.fb->drm_format ==DRM_FORMAT_XRGB8888 && - (cvt->dst.fb->drm_format ==DRM_FORMAT_YUYV || - cvt->dst.fb->drm_format ==DRM_FORMAT_UYVY || - cvt->dst.fb->drm_format ==DRM_FORMAT_YVYU || - cvt->dst.fb->drm_format ==DRM_FORMAT_VYUY )); - -for (i =0 ; i < cvt->dst.fb->height; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 rgb[2 ]; -struct igt_vec4 yuv[2 ]; - -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +4 ]); - - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); + y += params.y_stride; - yuyv[j *4 + swz[0 ]] = yuv[0 ].d[0 ]; - yuyv[j *4 + swz[2 ]] = yuv[1 ].d[0 ]; - yuyv[j *4 + swz[1 ]] = (yuv[0 ].d[1 ] + yuv[1 ].d[1 ]) /2.0f ; - yuyv[j *4 + swz[3 ]] = (yuv[0 ].d[2 ] + yuv[1 ].d[2 ]) /2.0f ; +if ((i % dst_fmt->vsub) == (dst_fmt->vsub -1 )) { + u += params.uv_stride; + v += params.uv_stride; } - -if (cvt->dst.fb->width &1 ) { -struct igt_vec4 rgb; -struct igt_vec4 yuv; - -read_rgb (&rgb, &rgb24[j *8 +0 ]); - - yuv =igt_matrix_transform (&m, &rgb); - - yuyv[j *4 + swz[0 ]] = yuv.d[0 ]; - yuyv[j *4 + swz[1 ]] = yuv.d[1 ]; - yuyv[j *4 + swz[3 ]] = yuv.d[2 ]; - } - - rgb24 += rgb24_stride; - yuyv += yuyv_stride; } }
@@ -2046,31 +1859,23 @@ static void fb_convert(struct fb_convert *cvt) }else if (cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 ) {switch (cvt->src.fb->drm_format) {case DRM_FORMAT_XYUV8888 : -convert_yuv444_to_rgb24 (cvt); -return ;case DRM_FORMAT_NV12 : -convert_nv12_to_rgb24 (cvt); -return ; -case DRM_FORMAT_YUYV : -case DRM_FORMAT_YVYU :case DRM_FORMAT_UYVY :case DRM_FORMAT_VYUY : -convert_yuyv_to_rgb24 (cvt); +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVYU : +convert_yuv_to_rgb24 (cvt);return ; } }else if (cvt->src.fb->drm_format ==DRM_FORMAT_XRGB8888 ) {switch (cvt->dst.fb->drm_format) {case DRM_FORMAT_XYUV8888 : -convert_rgb24_to_yuv444 (cvt); -return ;case DRM_FORMAT_NV12 : -convert_rgb24_to_nv12 (cvt); -return ;case DRM_FORMAT_YUYV :case DRM_FORMAT_YVYU :case DRM_FORMAT_UYVY :case DRM_FORMAT_VYUY : -convert_rgb24_to_yuyv (cvt); +convert_rgb24_to_yuv (cvt);return ; } }
Signature · 6 lines
2.20.1 _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 03/13] igt: fb: generic YUV convertion function
Acked-by Lyude Paul <[email protected]>
Quoted 669 lines · Maxime Ripard
The current way we work when we want to support a new YUV format to IGT, we also need to add new function to convert to and from XRGB8888.
This doesn't really scale however, and creates a lot of code to maintain. In order to work around this, create a generic function to convert to RGB and one to convert from RGB.
The only thing that is needed now is to add new parameters, and that's it.
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 537 ++++++++++++++++----------------------------------- 1 file changed, 171 insertions(+), 366 deletions(-)lib/igt_fb.c +145
-187@@ -1577,424 +1577,237 @@ static void convert_src_put(const structfb_convert *cvt,free (src_buf ); } -static void convert_nv12_to_rgb24 (struct fb_convert *cvt) +struct yuv_parameters { +unsigned y_inc; +unsigned uv_inc; +unsigned y_stride; +unsigned uv_stride; +unsigned y_offset; +unsigned u_offset; +unsigned v_offset; +}; + +static void get_yuv_parameters (struct igt_fb *fb,struct yuv_parameters *params) { -int i, j; -const uint8_t *y, *uv; -uint8_t *rgb24 = cvt->dst.ptr; -unsigned int rgb24_stride = cvt->dst.fb->strides[0 ]; -unsigned int planar_stride = cvt->src.fb->strides[0 ]; -struct igt_mat4 m =igt_ycbcr_to_rgb_matrix (cvt->src.fb- >color_encoding, - cvt->src.fb->color_range); -uint8_t *buf; - -igt_assert (cvt->src.fb->drm_format ==DRM_FORMAT_NV12 && - cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 ); - - buf =convert_src_get (cvt); - y = buf + cvt->src.fb->offsets[0 ]; - uv = buf + cvt->src.fb->offsets[1 ]; - -for (i =0 ; i < cvt->dst.fb->height /2 ; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x2 pixel blocks */ -struct igt_vec4 yuv[4 ]; -struct igt_vec4 rgb[4 ]; - - yuv[0 ].d[0 ] = y[j *2 +0 ]; - yuv[1 ].d[0 ] = y[j *2 +1 ]; - yuv[2 ].d[0 ] = y[j *2 +0 + planar_stride]; - yuv[3 ].d[0 ] = y[j *2 +1 + planar_stride]; - - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = yuv[2 ].d[1 ] = yuv[3 ].d[1 ] = uv[j * 2 + 0]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = yuv[2 ].d[2 ] = yuv[3 ].d[2 ] = uv[j * 2 + 1]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] = yuv[2 ].d[3 ] = yuv[3 ].d[3 ] = 1.0f; - - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); - rgb[2 ] =igt_matrix_transform (&m, &yuv[2 ]); - rgb[3 ] =igt_matrix_transform (&m, &yuv[3 ]); - -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +4 ], &rgb[1 ]); -write_rgb (&rgb24[j *8 +0 + rgb24_stride], &rgb[2 ]); -write_rgb (&rgb24[j *8 +4 + rgb24_stride], &rgb[3 ]); - } - -if (cvt->dst.fb->width &1 ) { -/* Convert 1x2 pixel block */ -struct igt_vec4 yuv[2 ]; -struct igt_vec4 rgb[2 ]; - - yuv[0 ].d[0 ] = y[j *2 +0 ]; - yuv[1 ].d[0 ] = y[j *2 +0 + planar_stride]; +igt_assert (igt_format_is_yuv (fb->drm_format)); - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = uv[j *2 +0 ]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = uv[j *2 +1 ]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] =1.0f ; +switch (fb->drm_format) { +case DRM_FORMAT_NV12 : + params->y_inc =1 ; + params->uv_inc =2 ; +break ; - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); - -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +0 + rgb24_stride], &rgb[1 ]); - } - - rgb24 +=2 * rgb24_stride; - y +=2 * planar_stride; - uv += planar_stride; +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVYU : +case DRM_FORMAT_UYVY : +case DRM_FORMAT_VYUY : + params->y_inc =2 ; + params->uv_inc =4 ; +break ; + +case DRM_FORMAT_XYUV8888 : + params->y_inc =4 ; + params->uv_inc =4 ; +break ; } -if (cvt->dst.fb->height &1 ) { -/* Convert last row */ -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 yuv[2 ]; -struct igt_vec4 rgb[2 ]; - - yuv[0 ].d[0 ] = y[j *2 +0 ]; - yuv[1 ].d[0 ] = y[j *2 +1 ]; - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = uv[j *2 +0 ]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = uv[j *2 +1 ]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] =1.0f ; +switch (fb->drm_format) { +case DRM_FORMAT_NV12 : + params->y_stride = fb->strides[0 ]; + params->uv_stride = fb->strides[1 ]; +break ; - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVYU : +case DRM_FORMAT_UYVY : +case DRM_FORMAT_VYUY : +case DRM_FORMAT_XYUV8888 : + params->y_stride = fb->strides[0 ]; + params->uv_stride = fb->strides[0 ]; +break ; + } -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +4 ], &rgb[0 ]); - } +switch (fb->drm_format) { +case DRM_FORMAT_NV12 : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[1 ]; + params->v_offset = fb->offsets[1 ] +1 ; +break ; -if (cvt->dst.fb->width &1 ) { -/* Convert single pixel */ -struct igt_vec4 yuv; -struct igt_vec4 rgb; +case DRM_FORMAT_YUYV : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[0 ] +1 ; + params->v_offset = fb->offsets[0 ] +3 ; +break ; - yuv.d[0 ] = y[j *2 +0 ]; - yuv.d[1 ] = uv[j *2 +0 ]; - yuv.d[2 ] = uv[j *2 +1 ]; - yuv.d[3 ] =1.0f ; +case DRM_FORMAT_YVYU : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[0 ] +3 ; + params->v_offset = fb->offsets[0 ] +1 ; +break ; - rgb =igt_matrix_transform (&m, &yuv); +case DRM_FORMAT_UYVY : + params->y_offset = fb->offsets[0 ] +1 ; + params->u_offset = fb->offsets[0 ]; + params->v_offset = fb->offsets[0 ] +2 ; +break ; -write_rgb (&rgb24[j *8 +0 ], &rgb); - } +case DRM_FORMAT_VYUY : + params->y_offset = fb->offsets[0 ] +1 ; + params->u_offset = fb->offsets[0 ] +2 ; + params->v_offset = fb->offsets[0 ]; +break ; + +case DRM_FORMAT_XYUV8888 : + params->y_offset = fb->offsets[0 ] +1 ; + params->u_offset = fb->offsets[0 ] +2 ; + params->v_offset = fb->offsets[0 ] +3 ; +break ; } - -convert_src_put (cvt, buf); } -static void convert_yuv444_to_rgb24 (struct fb_convert *cvt) +static void convert_yuv_to_rgb24 (struct fb_convert *cvt) { +const struct format_desc_struct *src_fmt = +lookup_drm_format (cvt->src.fb->drm_format);int i, j; -uint8_t *yuv24; +uint8_t bpp =4 ; +uint8_t *y, *u, *v;uint8_t *rgb24 = cvt->dst.ptr; -unsigned rgb24_stride = cvt->dst.fb->strides[0 ], xyuv_stride = cvt- >src.fb->strides[0]; -uint8_t *buf =malloc (cvt->src.fb->size); +unsigned int rgb24_stride = cvt->dst.fb->strides[0 ];struct igt_mat4 m =igt_ycbcr_to_rgb_matrix (cvt->src.fb- >color_encoding, cvt->src.fb->color_range); +uint8_t *buf; +struct yuv_parameters params = { }; -/* -* Reading from the BO is awfully slow because of lack of read caching, -* it's faster to copy the whole BO to a temporary buffer and convert -* from there. -*/ -igt_memcpy_from_wc (buf, cvt->src.ptr + cvt->src.fb->offsets[0 ], cvt- >src.fb->size); - yuv24 = buf; +igt_assert (cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 && +igt_format_is_yuv (cvt->src.fb->drm_format)); + + buf =convert_src_get (cvt); +get_yuv_parameters (cvt->src.fb, ¶ms); + y = buf + params.y_offset; + u = buf + params.u_offset; + v = buf + params.v_offset;for (i =0 ; i < cvt->dst.fb->height; i++) { +const uint8_t *y_tmp = y; +const uint8_t *u_tmp = u; +const uint8_t *v_tmp = v; +uint8_t *rgb_tmp = rgb24; +for (j =0 ; j < cvt->dst.fb->width; j++) { -float y, u, v; -struct igt_vec4 yuv; -struct igt_vec4 rgb; - - v = yuv24[i * xyuv_stride + j *4 ]; - u = yuv24[i * xyuv_stride + j *4 +1 ]; - y = yuv24[i * xyuv_stride + j *4 +2 ]; - yuv.d[0 ] = y; - yuv.d[1 ] = u; - yuv.d[2 ] = v; +struct igt_vec4 rgb, yuv; + + yuv.d[0 ] = *y_tmp; + yuv.d[1 ] = *u_tmp; + yuv.d[2 ] = *v_tmp; yuv.d[3 ] =1.0f ; rgb =igt_matrix_transform (&m, &yuv); +write_rgb (rgb_tmp, &rgb); -write_rgb (&rgb24[i * rgb24_stride + j *4 ], &rgb); - } - } - -free (buf); -} - - -static void convert_rgb24_to_yuv444 (struct fb_convert *cvt) -{ -int i, j; -uint8_t *rgb24; -uint8_t *yuv444 = cvt->dst.ptr + cvt->dst.fb->offsets[0 ]; -unsigned int rgb24_stride = cvt->src.fb->strides[0 ], xyuv_stride = cvt->dst.fb->strides[0 ]; -struct igt_mat4 m =igt_rgb_to_ycbcr_matrix (cvt->dst.fb- >color_encoding, - cvt->dst.fb->color_range); - - rgb24 = cvt->src.ptr; + rgb_tmp += bpp; + y_tmp += params.y_inc; -igt_assert_f (cvt->dst.fb->drm_format ==DRM_FORMAT_XYUV8888 , -"Conversion not implemented for !XYUV packed formats\n" ); - -for (i =0 ; i < cvt->dst.fb->height; i++) { -for (j =0 ; j < cvt->dst.fb->width; j++) { -struct igt_vec4 rgb; -struct igt_vec4 yuv; - -read_rgb (&rgb, &rgb24[i * rgb24_stride + j *4 ]); +if ((src_fmt->hsub ==1 ) || (j % src_fmt->hsub)) { + u_tmp += params.uv_inc; + v_tmp += params.uv_inc; + } + } - yuv =igt_matrix_transform (&m, &rgb); + rgb24 += rgb24_stride; + y += params.y_stride; - yuv444[i * xyuv_stride + j *4 ] = yuv.d[2 ]; - yuv444[i * xyuv_stride + j *4 +1 ] = yuv.d[1 ]; - yuv444[i * xyuv_stride + j *4 +2 ] = yuv.d[0 ]; +if ((src_fmt->vsub ==1 ) || (i % src_fmt->vsub)) { + u += params.uv_stride; + v += params.uv_stride; } } + +convert_src_put (cvt, buf); } -static void convert_rgb24_to_nv12 (struct fb_convert *cvt) +static void convert_rgb24_to_yuv (struct fb_convert *cvt) { +const struct format_desc_struct *dst_fmt = +lookup_drm_format (cvt->dst.fb->drm_format);int i, j; -uint8_t *y = cvt->dst.ptr + cvt->dst.fb->offsets[0 ]; -uint8_t *uv = cvt->dst.ptr + cvt->dst.fb->offsets[1 ]; +uint8_t *y, *u, *v;const uint8_t *rgb24 = cvt->src.ptr; +uint8_t bpp =4 ;unsigned rgb24_stride = cvt->src.fb->strides[0 ]; -unsigned planar_stride = cvt->dst.fb->strides[0 ];struct igt_mat4 m =igt_rgb_to_ycbcr_matrix (cvt->dst.fb- >color_encoding, cvt->dst.fb->color_range); +struct yuv_parameters params = { };igt_assert (cvt->src.fb->drm_format ==DRM_FORMAT_XRGB8888 && - cvt->dst.fb->drm_format ==DRM_FORMAT_NV12 ); +igt_format_is_yuv (cvt->dst.fb->drm_format)); -for (i =0 ; i < cvt->dst.fb->height /2 ; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x2 pixel blocks */ -struct igt_vec4 rgb[4 ]; -struct igt_vec4 yuv[4 ]; +get_yuv_parameters (cvt->dst.fb, ¶ms); + y = cvt->dst.ptr + params.y_offset; + u = cvt->dst.ptr + params.u_offset; + v = cvt->dst.ptr + params.v_offset; -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +4 ]); -read_rgb (&rgb[2 ], &rgb24[j *8 +0 + rgb24_stride]); -read_rgb (&rgb[3 ], &rgb24[j *8 +4 + rgb24_stride]); - - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); - yuv[2 ] =igt_matrix_transform (&m, &rgb[2 ]); - yuv[3 ] =igt_matrix_transform (&m, &rgb[3 ]); - - y[j *2 +0 ] = yuv[0 ].d[0 ]; - y[j *2 +1 ] = yuv[1 ].d[0 ]; - y[j *2 +0 + planar_stride] = yuv[2 ].d[0 ]; - y[j *2 +1 + planar_stride] = yuv[3 ].d[0 ]; +for (i =0 ; i < cvt->dst.fb->height; i++) { +const uint8_t *rgb_tmp = rgb24; +uint8_t *y_tmp = y; +uint8_t *u_tmp = u; +uint8_t *v_tmp = v; -/* -* We assume the MPEG2 chroma siting convention, where -* pixel center for Cb'Cr' is between the left top and -* bottom pixel in a 2x2 block, so take the average. -*/ - uv[j *2 +0 ] = (yuv[0 ].d[1 ] + yuv[2 ].d[1 ]) /2.0f ; - uv[j *2 +1 ] = (yuv[0 ].d[2 ] + yuv[2 ].d[2 ]) /2.0f ; - } +for (j =0 ; j < cvt->dst.fb->width; j++) { +const uint8_t *pair_rgb24 = rgb_tmp; +struct igt_vec4 pair_rgb, rgb; +struct igt_vec4 pair_yuv, yuv; -if (cvt->dst.fb->width &1 ) { -/* Convert 1x2 pixel block */ -struct igt_vec4 rgb[2 ]; -struct igt_vec4 yuv[2 ]; +read_rgb (&rgb, rgb_tmp); + yuv =igt_matrix_transform (&m, &rgb); -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +0 + rgb24_stride]); + rgb_tmp += bpp; - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); + *y_tmp = yuv.d[0 ]; + y_tmp += params.y_inc; - y[j *2 +0 ] = yuv[0 ].d[0 ]; - y[j *2 +0 + planar_stride] = yuv[1 ].d[0 ]; +if ((i % dst_fmt->vsub) || (j % dst_fmt->hsub)) +continue ;/* * We assume the MPEG2 chroma siting convention, where * pixel center for Cb'Cr' is between the left top and * bottom pixel in a 2x2 block, so take the average. +* +* Therefore, if we use subsampling, we only really care
lib/igt_fb.c +5
-0+* about two pixels all the time, either the two +* subsequent pixels horizontally, vertically, or the +* two corners in a 2x2 block. +* +* The only corner case is when we have an odd number of
lib/igt_fb.c +1
-0+* pixels, but this can be handled pretty easily by not
lib/igt_fb.c +2
-10+* incrementing the paired pixel pointer in the +* direction it's odd in. */ - uv[j *2 +0 ] = (yuv[0 ].d[1 ] + yuv[1 ].d[1 ]) /2.0f ; - uv[j *2 +1 ] = (yuv[0 ].d[2 ] + yuv[1 ].d[2 ]) /2.0f ; - } - - rgb24 +=2 * rgb24_stride; - y +=2 * planar_stride; - uv += planar_stride; - } - -/* Last row cannot be interpolated between 2 pixels, take the single value */
lib/igt_fb.c +4
-21-if (cvt->dst.fb->height &1 ) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 rgb[2 ]; -struct igt_vec4 yuv[2 ]; - -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +4 ]); - - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); +if (j != (cvt->dst.fb->width -1 )) + pair_rgb24 += (dst_fmt->hsub -1 ) * bpp; - y[j *2 +0 ] = yuv[0 ].d[0 ]; - y[j *2 +1 ] = yuv[1 ].d[0 ]; - uv[j *2 +0 ] = yuv[0 ].d[1 ]; - uv[j *2 +1 ] = yuv[0 ].d[2 ]; - } - -if (cvt->dst.fb->width &1 ) { -/* Convert single pixel */ -struct igt_vec4 rgb; -struct igt_vec4 yuv; +if (i != (cvt->dst.fb->height -1 )) + pair_rgb24 += rgb24_stride * (dst_fmt->vsub -1);
lib/igt_fb.c +4
-38-read_rgb (&rgb, &rgb24[j *8 +0 ]); +read_rgb (&pair_rgb, pair_rgb24); + pair_yuv =igt_matrix_transform (&m, &pair_rgb); - yuv =igt_matrix_transform (&m, &rgb); + *u_tmp = (yuv.d[1 ] + pair_yuv.d[1 ]) /2.0f ; + *v_tmp = (yuv.d[2 ] + pair_yuv.d[2 ]) /2.0f ; - y[j *2 +0 ] = yuv.d[0 ]; - uv[j *2 +0 ] = yuv.d[1 ]; - uv[j *2 +1 ] = yuv.d[2 ]; - } - } -} - -/* { Y0, U, Y1, V } */ -static const unsigned char swizzle_yuyv[] = {0 ,1 ,2 ,3 }; -static const unsigned char swizzle_yvyu[] = {0 ,3 ,2 ,1 }; -static const unsigned char swizzle_uyvy[] = {1 ,0 ,3 ,2 }; -static const unsigned char swizzle_vyuy[] = {1 ,2 ,3 ,0 }; - -static const unsigned char *yuyv_swizzle (uint32_t format) -{ -switch (format) { -default : -case DRM_FORMAT_YUYV : -return swizzle_yuyv; -case DRM_FORMAT_YVYU : -return swizzle_yvyu; -case DRM_FORMAT_UYVY : -return swizzle_uyvy; -case DRM_FORMAT_VYUY : -return swizzle_vyuy; - } -} - -static void convert_yuyv_to_rgb24 (struct fb_convert *cvt) -{ -int i, j; -const uint8_t *yuyv; -uint8_t *rgb24 = cvt->dst.ptr; -unsigned int rgb24_stride = cvt->dst.fb->strides[0 ]; -unsigned int yuyv_stride = cvt->src.fb->strides[0 ]; - struct igt_mat4 m = igt_ycbcr_to_rgb_matrix(cvt->src.fb-Quoted 1 line
color_encoding,lib/igt_fb.c +2
-58- cvt->src.fb->color_range); -const unsigned char *swz =yuyv_swizzle (cvt->src.fb->drm_format); -uint8_t *buf; - -igt_assert ((cvt->src.fb->drm_format ==DRM_FORMAT_YUYV || - cvt->src.fb->drm_format ==DRM_FORMAT_UYVY || - cvt->src.fb->drm_format ==DRM_FORMAT_YVYU || - cvt->src.fb->drm_format ==DRM_FORMAT_VYUY ) && - cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 ); - - buf =convert_src_get (cvt); - yuyv = buf; - -for (i =0 ; i < cvt->dst.fb->height; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 yuv[2 ]; -struct igt_vec4 rgb[2 ]; - - yuv[0 ].d[0 ] = yuyv[j *4 + swz[0 ]]; - yuv[1 ].d[0 ] = yuyv[j *4 + swz[2 ]]; - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = yuyv[j *4 + swz[1 ]]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = yuyv[j *4 + swz[3 ]]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] =1.0f ; - - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); - -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +4 ], &rgb[1 ]); - } - -if (cvt->dst.fb->width &1 ) { -struct igt_vec4 yuv; -struct igt_vec4 rgb; - - yuv.d[0 ] = yuyv[j *4 + swz[0 ]]; - yuv.d[1 ] = yuyv[j *4 + swz[1 ]]; - yuv.d[2 ] = yuyv[j *4 + swz[3 ]]; - yuv.d[3 ] =1.0f ; - - rgb =igt_matrix_transform (&m, &yuv); - -write_rgb (&rgb24[j *8 +0 ], &rgb); + u_tmp += params.uv_inc; + v_tmp += params.uv_inc; } rgb24 += rgb24_stride; - yuyv += yuyv_stride; - } - -convert_src_put (cvt, buf); -} - -static void convert_rgb24_to_yuyv (struct fb_convert *cvt) -{ -int i, j; -uint8_t *yuyv = cvt->dst.ptr; -const uint8_t *rgb24 = cvt->src.ptr; -unsigned rgb24_stride = cvt->src.fb->strides[0 ]; -unsigned yuyv_stride = cvt->dst.fb->strides[0 ]; - struct igt_mat4 m = igt_rgb_to_ycbcr_matrix(cvt->dst.fb-Quoted 1 line
color_encoding,lib/igt_fb.c +1
-23- cvt->dst.fb->color_range); -const unsigned char *swz =yuyv_swizzle (cvt->dst.fb->drm_format); - -igt_assert (cvt->src.fb->drm_format ==DRM_FORMAT_XRGB8888 && - (cvt->dst.fb->drm_format ==DRM_FORMAT_YUYV || - cvt->dst.fb->drm_format ==DRM_FORMAT_UYVY || - cvt->dst.fb->drm_format ==DRM_FORMAT_YVYU || - cvt->dst.fb->drm_format ==DRM_FORMAT_VYUY )); - -for (i =0 ; i < cvt->dst.fb->height; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 rgb[2 ]; -struct igt_vec4 yuv[2 ]; - -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +4 ]); - - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); + y += params.y_stride; - yuyv[j *4 + swz[0 ]] = yuv[0 ].d[0 ]; - yuyv[j *4 + swz[2 ]] = yuv[1 ].d[0 ]; - yuyv[j * 4 + swz[1]] = (yuv[0].d[1] + yuv[1].d[1]) /2.0f;
lib/igt_fb.c +0
-1- yuyv[j * 4 + swz[3]] = (yuv[0].d[2] + yuv[1].d[2]) /2.0f;
lib/igt_fb.c +7
-28+if ((i % dst_fmt->vsub) == (dst_fmt->vsub -1 )) { + u += params.uv_stride; + v += params.uv_stride; } - -if (cvt->dst.fb->width &1 ) { -struct igt_vec4 rgb; -struct igt_vec4 yuv; - -read_rgb (&rgb, &rgb24[j *8 +0 ]); - - yuv =igt_matrix_transform (&m, &rgb); - - yuyv[j *4 + swz[0 ]] = yuv.d[0 ]; - yuyv[j *4 + swz[1 ]] = yuv.d[1 ]; - yuyv[j *4 + swz[3 ]] = yuv.d[2 ]; - } - - rgb24 += rgb24_stride; - yuyv += yuyv_stride; } }@@ -2046,31 +1859,23 @@ static void fb_convert(struct fb_convert *cvt) }else if (cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 ) {switch (cvt->src.fb->drm_format) {case DRM_FORMAT_XYUV8888 : -convert_yuv444_to_rgb24 (cvt); -return ;case DRM_FORMAT_NV12 : -convert_nv12_to_rgb24 (cvt); -return ; -case DRM_FORMAT_YUYV : -case DRM_FORMAT_YVYU :case DRM_FORMAT_UYVY :case DRM_FORMAT_VYUY : -convert_yuyv_to_rgb24 (cvt); +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVYU : +convert_yuv_to_rgb24 (cvt);return ; } }else if (cvt->src.fb->drm_format ==DRM_FORMAT_XRGB8888 ) {switch (cvt->dst.fb->drm_format) {case DRM_FORMAT_XYUV8888 : -convert_rgb24_to_yuv444 (cvt); -return ;case DRM_FORMAT_NV12 : -convert_rgb24_to_nv12 (cvt); -return ;case DRM_FORMAT_YUYV :case DRM_FORMAT_YVYU :case DRM_FORMAT_UYVY :case DRM_FORMAT_VYUY : -convert_rgb24_to_yuyv (cvt); +convert_rgb24_to_yuv (cvt);return ; } }
Signature · 7 lines
Cheers, Lyude Paul _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 03/13] igt: fb: generic YUV convertion function
In reply to Maxime RipardQuoted 8 lines · Maxime Ripard
The current way we work when we want to support a new YUV format to IGT, we also need to add new function to convert to and from XRGB8888.
This doesn't really scale however, and creates a lot of code to maintain. In order to work around this, create a generic function to convert to RGB and one to convert from RGB.
The only thing that is needed now is to add new parameters, and that's it.
What's the performance impact of this?
Quoted 641 lines
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 537 ++++++++++++++++----------------------------------- 1 file changed, 171 insertions(+), 366 deletions(-)lib/igt_fb.c +171
-366@@ -1577,424 +1577,237 @@ static void convert_src_put(const struct fb_convert *cvt,free (src_buf); } -static void convert_nv12_to_rgb24 (struct fb_convert *cvt) +struct yuv_parameters { +unsigned y_inc; +unsigned uv_inc; +unsigned y_stride; +unsigned uv_stride; +unsigned y_offset; +unsigned u_offset; +unsigned v_offset; +}; + +static void get_yuv_parameters (struct igt_fb *fb,struct yuv_parameters *params) { -int i, j; -const uint8_t *y, *uv; -uint8_t *rgb24 = cvt->dst.ptr; -unsigned int rgb24_stride = cvt->dst.fb->strides[0 ]; -unsigned int planar_stride = cvt->src.fb->strides[0 ]; -struct igt_mat4 m =igt_ycbcr_to_rgb_matrix (cvt->src.fb->color_encoding, - cvt->src.fb->color_range); -uint8_t *buf; - -igt_assert (cvt->src.fb->drm_format ==DRM_FORMAT_NV12 && - cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 ); - - buf =convert_src_get (cvt); - y = buf + cvt->src.fb->offsets[0 ]; - uv = buf + cvt->src.fb->offsets[1 ]; - -for (i =0 ; i < cvt->dst.fb->height /2 ; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x2 pixel blocks */ -struct igt_vec4 yuv[4 ]; -struct igt_vec4 rgb[4 ]; - - yuv[0 ].d[0 ] = y[j *2 +0 ]; - yuv[1 ].d[0 ] = y[j *2 +1 ]; - yuv[2 ].d[0 ] = y[j *2 +0 + planar_stride]; - yuv[3 ].d[0 ] = y[j *2 +1 + planar_stride]; - - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = yuv[2 ].d[1 ] = yuv[3 ].d[1 ] = uv[j *2 +0 ]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = yuv[2 ].d[2 ] = yuv[3 ].d[2 ] = uv[j *2 +1 ]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] = yuv[2 ].d[3 ] = yuv[3 ].d[3 ] =1.0f ; - - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); - rgb[2 ] =igt_matrix_transform (&m, &yuv[2 ]); - rgb[3 ] =igt_matrix_transform (&m, &yuv[3 ]); - -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +4 ], &rgb[1 ]); -write_rgb (&rgb24[j *8 +0 + rgb24_stride], &rgb[2 ]); -write_rgb (&rgb24[j *8 +4 + rgb24_stride], &rgb[3 ]); - } - -if (cvt->dst.fb->width &1 ) { -/* Convert 1x2 pixel block */ -struct igt_vec4 yuv[2 ]; -struct igt_vec4 rgb[2 ]; - - yuv[0 ].d[0 ] = y[j *2 +0 ]; - yuv[1 ].d[0 ] = y[j *2 +0 + planar_stride]; +igt_assert (igt_format_is_yuv (fb->drm_format)); - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = uv[j *2 +0 ]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = uv[j *2 +1 ]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] =1.0f ; +switch (fb->drm_format) { +case DRM_FORMAT_NV12 : + params->y_inc =1 ; + params->uv_inc =2 ; +break ; - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); - -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +0 + rgb24_stride], &rgb[1 ]); - } - - rgb24 +=2 * rgb24_stride; - y +=2 * planar_stride; - uv += planar_stride; +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVYU : +case DRM_FORMAT_UYVY : +case DRM_FORMAT_VYUY : + params->y_inc =2 ; + params->uv_inc =4 ; +break ; + +case DRM_FORMAT_XYUV8888 : + params->y_inc =4 ; + params->uv_inc =4 ; +break ; } -if (cvt->dst.fb->height &1 ) { -/* Convert last row */ -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 yuv[2 ]; -struct igt_vec4 rgb[2 ]; - - yuv[0 ].d[0 ] = y[j *2 +0 ]; - yuv[1 ].d[0 ] = y[j *2 +1 ]; - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = uv[j *2 +0 ]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = uv[j *2 +1 ]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] =1.0f ; +switch (fb->drm_format) { +case DRM_FORMAT_NV12 : + params->y_stride = fb->strides[0 ]; + params->uv_stride = fb->strides[1 ]; +break ; - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVYU : +case DRM_FORMAT_UYVY : +case DRM_FORMAT_VYUY : +case DRM_FORMAT_XYUV8888 : + params->y_stride = fb->strides[0 ]; + params->uv_stride = fb->strides[0 ]; +break ; + } -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +4 ], &rgb[0 ]); - } +switch (fb->drm_format) { +case DRM_FORMAT_NV12 : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[1 ]; + params->v_offset = fb->offsets[1 ] +1 ; +break ; -if (cvt->dst.fb->width &1 ) { -/* Convert single pixel */ -struct igt_vec4 yuv; -struct igt_vec4 rgb; +case DRM_FORMAT_YUYV : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[0 ] +1 ; + params->v_offset = fb->offsets[0 ] +3 ; +break ; - yuv.d[0 ] = y[j *2 +0 ]; - yuv.d[1 ] = uv[j *2 +0 ]; - yuv.d[2 ] = uv[j *2 +1 ]; - yuv.d[3 ] =1.0f ; +case DRM_FORMAT_YVYU : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[0 ] +3 ; + params->v_offset = fb->offsets[0 ] +1 ; +break ; - rgb =igt_matrix_transform (&m, &yuv); +case DRM_FORMAT_UYVY : + params->y_offset = fb->offsets[0 ] +1 ; + params->u_offset = fb->offsets[0 ]; + params->v_offset = fb->offsets[0 ] +2 ; +break ; -write_rgb (&rgb24[j *8 +0 ], &rgb); - } +case DRM_FORMAT_VYUY : + params->y_offset = fb->offsets[0 ] +1 ; + params->u_offset = fb->offsets[0 ] +2 ; + params->v_offset = fb->offsets[0 ]; +break ; + +case DRM_FORMAT_XYUV8888 : + params->y_offset = fb->offsets[0 ] +1 ; + params->u_offset = fb->offsets[0 ] +2 ; + params->v_offset = fb->offsets[0 ] +3 ; +break ; } - -convert_src_put (cvt, buf); } -static void convert_yuv444_to_rgb24 (struct fb_convert *cvt) +static void convert_yuv_to_rgb24 (struct fb_convert *cvt) { +const struct format_desc_struct *src_fmt = +lookup_drm_format (cvt->src.fb->drm_format);int i, j; -uint8_t *yuv24; +uint8_t bpp =4 ; +uint8_t *y, *u, *v;uint8_t *rgb24 = cvt->dst.ptr; -unsigned rgb24_stride = cvt->dst.fb->strides[0 ], xyuv_stride = cvt->src.fb->strides[0 ]; -uint8_t *buf =malloc (cvt->src.fb->size); +unsigned int rgb24_stride = cvt->dst.fb->strides[0 ];struct igt_mat4 m =igt_ycbcr_to_rgb_matrix (cvt->src.fb->color_encoding, cvt->src.fb->color_range); +uint8_t *buf; +struct yuv_parameters params = { }; -/* -* Reading from the BO is awfully slow because of lack of read caching, -* it's faster to copy the whole BO to a temporary buffer and convert -* from there. -*/ -igt_memcpy_from_wc (buf, cvt->src.ptr + cvt->src.fb->offsets[0 ], cvt->src.fb->size); - yuv24 = buf; +igt_assert (cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 && +igt_format_is_yuv (cvt->src.fb->drm_format)); + + buf =convert_src_get (cvt); +get_yuv_parameters (cvt->src.fb, ¶ms); + y = buf + params.y_offset; + u = buf + params.u_offset; + v = buf + params.v_offset;for (i =0 ; i < cvt->dst.fb->height; i++) { +const uint8_t *y_tmp = y; +const uint8_t *u_tmp = u; +const uint8_t *v_tmp = v; +uint8_t *rgb_tmp = rgb24; +for (j =0 ; j < cvt->dst.fb->width; j++) { -float y, u, v; -struct igt_vec4 yuv; -struct igt_vec4 rgb; - - v = yuv24[i * xyuv_stride + j *4 ]; - u = yuv24[i * xyuv_stride + j *4 +1 ]; - y = yuv24[i * xyuv_stride + j *4 +2 ]; - yuv.d[0 ] = y; - yuv.d[1 ] = u; - yuv.d[2 ] = v; +struct igt_vec4 rgb, yuv; + + yuv.d[0 ] = *y_tmp; + yuv.d[1 ] = *u_tmp; + yuv.d[2 ] = *v_tmp; yuv.d[3 ] =1.0f ; rgb =igt_matrix_transform (&m, &yuv); +write_rgb (rgb_tmp, &rgb); -write_rgb (&rgb24[i * rgb24_stride + j *4 ], &rgb); - } - } - -free (buf); -} - - -static void convert_rgb24_to_yuv444 (struct fb_convert *cvt) -{ -int i, j; -uint8_t *rgb24; -uint8_t *yuv444 = cvt->dst.ptr + cvt->dst.fb->offsets[0 ]; -unsigned int rgb24_stride = cvt->src.fb->strides[0 ], xyuv_stride = cvt->dst.fb->strides[0 ]; -struct igt_mat4 m =igt_rgb_to_ycbcr_matrix (cvt->dst.fb->color_encoding, - cvt->dst.fb->color_range); - - rgb24 = cvt->src.ptr; + rgb_tmp += bpp; + y_tmp += params.y_inc; -igt_assert_f (cvt->dst.fb->drm_format ==DRM_FORMAT_XYUV8888 , -"Conversion not implemented for !XYUV packed formats\n" ); - -for (i =0 ; i < cvt->dst.fb->height; i++) { -for (j =0 ; j < cvt->dst.fb->width; j++) { -struct igt_vec4 rgb; -struct igt_vec4 yuv; - -read_rgb (&rgb, &rgb24[i * rgb24_stride + j *4 ]); +if ((src_fmt->hsub ==1 ) || (j % src_fmt->hsub)) { + u_tmp += params.uv_inc; + v_tmp += params.uv_inc; + } + } - yuv =igt_matrix_transform (&m, &rgb); + rgb24 += rgb24_stride; + y += params.y_stride; - yuv444[i * xyuv_stride + j *4 ] = yuv.d[2 ]; - yuv444[i * xyuv_stride + j *4 +1 ] = yuv.d[1 ]; - yuv444[i * xyuv_stride + j *4 +2 ] = yuv.d[0 ]; +if ((src_fmt->vsub ==1 ) || (i % src_fmt->vsub)) { + u += params.uv_stride; + v += params.uv_stride; } } + +convert_src_put (cvt, buf); } -static void convert_rgb24_to_nv12 (struct fb_convert *cvt) +static void convert_rgb24_to_yuv (struct fb_convert *cvt) { +const struct format_desc_struct *dst_fmt = +lookup_drm_format (cvt->dst.fb->drm_format);int i, j; -uint8_t *y = cvt->dst.ptr + cvt->dst.fb->offsets[0 ]; -uint8_t *uv = cvt->dst.ptr + cvt->dst.fb->offsets[1 ]; +uint8_t *y, *u, *v;const uint8_t *rgb24 = cvt->src.ptr; +uint8_t bpp =4 ;unsigned rgb24_stride = cvt->src.fb->strides[0 ]; -unsigned planar_stride = cvt->dst.fb->strides[0 ];struct igt_mat4 m =igt_rgb_to_ycbcr_matrix (cvt->dst.fb->color_encoding, cvt->dst.fb->color_range); +struct yuv_parameters params = { };igt_assert (cvt->src.fb->drm_format ==DRM_FORMAT_XRGB8888 && - cvt->dst.fb->drm_format ==DRM_FORMAT_NV12 ); +igt_format_is_yuv (cvt->dst.fb->drm_format)); -for (i =0 ; i < cvt->dst.fb->height /2 ; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x2 pixel blocks */ -struct igt_vec4 rgb[4 ]; -struct igt_vec4 yuv[4 ]; +get_yuv_parameters (cvt->dst.fb, ¶ms); + y = cvt->dst.ptr + params.y_offset; + u = cvt->dst.ptr + params.u_offset; + v = cvt->dst.ptr + params.v_offset; -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +4 ]); -read_rgb (&rgb[2 ], &rgb24[j *8 +0 + rgb24_stride]); -read_rgb (&rgb[3 ], &rgb24[j *8 +4 + rgb24_stride]); - - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); - yuv[2 ] =igt_matrix_transform (&m, &rgb[2 ]); - yuv[3 ] =igt_matrix_transform (&m, &rgb[3 ]); - - y[j *2 +0 ] = yuv[0 ].d[0 ]; - y[j *2 +1 ] = yuv[1 ].d[0 ]; - y[j *2 +0 + planar_stride] = yuv[2 ].d[0 ]; - y[j *2 +1 + planar_stride] = yuv[3 ].d[0 ]; +for (i =0 ; i < cvt->dst.fb->height; i++) { +const uint8_t *rgb_tmp = rgb24; +uint8_t *y_tmp = y; +uint8_t *u_tmp = u; +uint8_t *v_tmp = v; -/* -* We assume the MPEG2 chroma siting convention, where -* pixel center for Cb'Cr' is between the left top and -* bottom pixel in a 2x2 block, so take the average. -*/ - uv[j *2 +0 ] = (yuv[0 ].d[1 ] + yuv[2 ].d[1 ]) /2.0f ; - uv[j *2 +1 ] = (yuv[0 ].d[2 ] + yuv[2 ].d[2 ]) /2.0f ; - } +for (j =0 ; j < cvt->dst.fb->width; j++) { +const uint8_t *pair_rgb24 = rgb_tmp; +struct igt_vec4 pair_rgb, rgb; +struct igt_vec4 pair_yuv, yuv; -if (cvt->dst.fb->width &1 ) { -/* Convert 1x2 pixel block */ -struct igt_vec4 rgb[2 ]; -struct igt_vec4 yuv[2 ]; +read_rgb (&rgb, rgb_tmp); + yuv =igt_matrix_transform (&m, &rgb); -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +0 + rgb24_stride]); + rgb_tmp += bpp; - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); + *y_tmp = yuv.d[0 ]; + y_tmp += params.y_inc; - y[j *2 +0 ] = yuv[0 ].d[0 ]; - y[j *2 +0 + planar_stride] = yuv[1 ].d[0 ]; +if ((i % dst_fmt->vsub) || (j % dst_fmt->hsub)) +continue ;/* * We assume the MPEG2 chroma siting convention, where * pixel center for Cb'Cr' is between the left top and * bottom pixel in a 2x2 block, so take the average. +* +* Therefore, if we use subsampling, we only really care +* about two pixels all the time, either the two +* subsequent pixels horizontally, vertically, or the +* two corners in a 2x2 block. +* +* The only corner case is when we have an odd number of +* pixels, but this can be handled pretty easily by not +* incrementing the paired pixel pointer in the +* direction it's odd in. */ - uv[j *2 +0 ] = (yuv[0 ].d[1 ] + yuv[1 ].d[1 ]) /2.0f ; - uv[j *2 +1 ] = (yuv[0 ].d[2 ] + yuv[1 ].d[2 ]) /2.0f ; - } - - rgb24 +=2 * rgb24_stride; - y +=2 * planar_stride; - uv += planar_stride; - } - -/* Last row cannot be interpolated between 2 pixels, take the single value */ -if (cvt->dst.fb->height &1 ) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 rgb[2 ]; -struct igt_vec4 yuv[2 ]; - -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +4 ]); - - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); +if (j != (cvt->dst.fb->width -1 )) + pair_rgb24 += (dst_fmt->hsub -1 ) * bpp; - y[j *2 +0 ] = yuv[0 ].d[0 ]; - y[j *2 +1 ] = yuv[1 ].d[0 ]; - uv[j *2 +0 ] = yuv[0 ].d[1 ]; - uv[j *2 +1 ] = yuv[0 ].d[2 ]; - } - -if (cvt->dst.fb->width &1 ) { -/* Convert single pixel */ -struct igt_vec4 rgb; -struct igt_vec4 yuv; +if (i != (cvt->dst.fb->height -1 )) + pair_rgb24 += rgb24_stride * (dst_fmt->vsub -1 ); -read_rgb (&rgb, &rgb24[j *8 +0 ]); +read_rgb (&pair_rgb, pair_rgb24); + pair_yuv =igt_matrix_transform (&m, &pair_rgb); - yuv =igt_matrix_transform (&m, &rgb); + *u_tmp = (yuv.d[1 ] + pair_yuv.d[1 ]) /2.0f ; + *v_tmp = (yuv.d[2 ] + pair_yuv.d[2 ]) /2.0f ; - y[j *2 +0 ] = yuv.d[0 ]; - uv[j *2 +0 ] = yuv.d[1 ]; - uv[j *2 +1 ] = yuv.d[2 ]; - } - } -} - -/* { Y0, U, Y1, V } */ -static const unsigned char swizzle_yuyv[] = {0 ,1 ,2 ,3 }; -static const unsigned char swizzle_yvyu[] = {0 ,3 ,2 ,1 }; -static const unsigned char swizzle_uyvy[] = {1 ,0 ,3 ,2 }; -static const unsigned char swizzle_vyuy[] = {1 ,2 ,3 ,0 }; - -static const unsigned char *yuyv_swizzle (uint32_t format) -{ -switch (format) { -default : -case DRM_FORMAT_YUYV : -return swizzle_yuyv; -case DRM_FORMAT_YVYU : -return swizzle_yvyu; -case DRM_FORMAT_UYVY : -return swizzle_uyvy; -case DRM_FORMAT_VYUY : -return swizzle_vyuy; - } -} - -static void convert_yuyv_to_rgb24 (struct fb_convert *cvt) -{ -int i, j; -const uint8_t *yuyv; -uint8_t *rgb24 = cvt->dst.ptr; -unsigned int rgb24_stride = cvt->dst.fb->strides[0 ]; -unsigned int yuyv_stride = cvt->src.fb->strides[0 ]; -struct igt_mat4 m =igt_ycbcr_to_rgb_matrix (cvt->src.fb->color_encoding, - cvt->src.fb->color_range); -const unsigned char *swz =yuyv_swizzle (cvt->src.fb->drm_format); -uint8_t *buf; - -igt_assert ((cvt->src.fb->drm_format ==DRM_FORMAT_YUYV || - cvt->src.fb->drm_format ==DRM_FORMAT_UYVY || - cvt->src.fb->drm_format ==DRM_FORMAT_YVYU || - cvt->src.fb->drm_format ==DRM_FORMAT_VYUY ) && - cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 ); - - buf =convert_src_get (cvt); - yuyv = buf; - -for (i =0 ; i < cvt->dst.fb->height; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 yuv[2 ]; -struct igt_vec4 rgb[2 ]; - - yuv[0 ].d[0 ] = yuyv[j *4 + swz[0 ]]; - yuv[1 ].d[0 ] = yuyv[j *4 + swz[2 ]]; - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = yuyv[j *4 + swz[1 ]]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = yuyv[j *4 + swz[3 ]]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] =1.0f ; - - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); - -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +4 ], &rgb[1 ]); - } - -if (cvt->dst.fb->width &1 ) { -struct igt_vec4 yuv; -struct igt_vec4 rgb; - - yuv.d[0 ] = yuyv[j *4 + swz[0 ]]; - yuv.d[1 ] = yuyv[j *4 + swz[1 ]]; - yuv.d[2 ] = yuyv[j *4 + swz[3 ]]; - yuv.d[3 ] =1.0f ; - - rgb =igt_matrix_transform (&m, &yuv); - -write_rgb (&rgb24[j *8 +0 ], &rgb); + u_tmp += params.uv_inc; + v_tmp += params.uv_inc; } rgb24 += rgb24_stride; - yuyv += yuyv_stride; - } - -convert_src_put (cvt, buf); -} - -static void convert_rgb24_to_yuyv (struct fb_convert *cvt) -{ -int i, j; -uint8_t *yuyv = cvt->dst.ptr; -const uint8_t *rgb24 = cvt->src.ptr; -unsigned rgb24_stride = cvt->src.fb->strides[0 ]; -unsigned yuyv_stride = cvt->dst.fb->strides[0 ]; -struct igt_mat4 m =igt_rgb_to_ycbcr_matrix (cvt->dst.fb->color_encoding, - cvt->dst.fb->color_range); -const unsigned char *swz =yuyv_swizzle (cvt->dst.fb->drm_format); - -igt_assert (cvt->src.fb->drm_format ==DRM_FORMAT_XRGB8888 && - (cvt->dst.fb->drm_format ==DRM_FORMAT_YUYV || - cvt->dst.fb->drm_format ==DRM_FORMAT_UYVY || - cvt->dst.fb->drm_format ==DRM_FORMAT_YVYU || - cvt->dst.fb->drm_format ==DRM_FORMAT_VYUY )); - -for (i =0 ; i < cvt->dst.fb->height; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 rgb[2 ]; -struct igt_vec4 yuv[2 ]; - -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +4 ]); - - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); + y += params.y_stride; - yuyv[j *4 + swz[0 ]] = yuv[0 ].d[0 ]; - yuyv[j *4 + swz[2 ]] = yuv[1 ].d[0 ]; - yuyv[j *4 + swz[1 ]] = (yuv[0 ].d[1 ] + yuv[1 ].d[1 ]) /2.0f ; - yuyv[j *4 + swz[3 ]] = (yuv[0 ].d[2 ] + yuv[1 ].d[2 ]) /2.0f ; +if ((i % dst_fmt->vsub) == (dst_fmt->vsub -1 )) { + u += params.uv_stride; + v += params.uv_stride; } - -if (cvt->dst.fb->width &1 ) { -struct igt_vec4 rgb; -struct igt_vec4 yuv; - -read_rgb (&rgb, &rgb24[j *8 +0 ]); - - yuv =igt_matrix_transform (&m, &rgb); - - yuyv[j *4 + swz[0 ]] = yuv.d[0 ]; - yuyv[j *4 + swz[1 ]] = yuv.d[1 ]; - yuyv[j *4 + swz[3 ]] = yuv.d[2 ]; - } - - rgb24 += rgb24_stride; - yuyv += yuyv_stride; } }@@ -2046,31 +1859,23 @@ static void fb_convert(struct fb_convert *cvt) }else if (cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 ) {switch (cvt->src.fb->drm_format) {case DRM_FORMAT_XYUV8888 : -convert_yuv444_to_rgb24 (cvt); -return ;case DRM_FORMAT_NV12 : -convert_nv12_to_rgb24 (cvt); -return ; -case DRM_FORMAT_YUYV : -case DRM_FORMAT_YVYU :case DRM_FORMAT_UYVY :case DRM_FORMAT_VYUY : -convert_yuyv_to_rgb24 (cvt); +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVYU : +convert_yuv_to_rgb24 (cvt);return ; } }else if (cvt->src.fb->drm_format ==DRM_FORMAT_XRGB8888 ) {switch (cvt->dst.fb->drm_format) {case DRM_FORMAT_XYUV8888 : -convert_rgb24_to_yuv444 (cvt); -return ;case DRM_FORMAT_NV12 : -convert_rgb24_to_nv12 (cvt); -return ;case DRM_FORMAT_YUYV :case DRM_FORMAT_YVYU :case DRM_FORMAT_UYVY :case DRM_FORMAT_VYUY : -convert_rgb24_to_yuyv (cvt); +convert_rgb24_to_yuv (cvt);return ; } }Signature · 1 line
2.20.1
Signature · 6 lines
Ville Syrjälä Intel _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 03/13] igt: fb: generic YUV convertion function
Hi Ville,
Quoted 11 lines · Ville Syrjälä
Quoted 8 lines · Maxime Ripard
The current way we work when we want to support a new YUV format to IGT, we also need to add new function to convert to and from XRGB8888.
This doesn't really scale however, and creates a lot of code to maintain. In order to work around this, create a generic function to convert to RGB and one to convert from RGB.
The only thing that is needed now is to add new parameters, and that's it.
What's the performance impact of this?
I just measured one test on a old laptop we have here (with an i7 4600U):
The first run is with the current code (note that the message is wrong and its milliseconds, not seconds):
./build/tests/kms_plane --run-subtest pixel-format-pipe-a-planes
IGT-Version: 1.23-g1c1612bd (x86_64) (Linux: 4.18.0-1-amd64 x86_64) Starting subtest: pixel-format-pipe-A-planes
Testing format XR24 (0x34325258) on A.0
Testing format RG16 (0x36314752) on A.0
Testing format XR30 (0x30335258) on A.0
Testing format XR24 (0x34325258) on A.1
Testing format YUYV (0x56595559) on A.1
took 79.655000 seconds to execute
took 76.913000 seconds to execute
took 77.148000 seconds to execute
took 76.004000 seconds to execute
took 77.223000 seconds to execute
took 76.348000 seconds to execute
took 74.939000 seconds to execute
took 76.563000 seconds to execute
Testing format YVYU (0x55595659) on A.1
took 75.744000 seconds to execute
took 76.446000 seconds to execute
took 75.803000 seconds to execute
took 76.711000 seconds to execute
took 75.568000 seconds to execute
took 78.681000 seconds to execute
took 77.757000 seconds to execute
took 76.557000 seconds to execute
Testing format UYVY (0x59565955) on A.1
took 77.667000 seconds to execute
took 76.608000 seconds to execute
took 77.551000 seconds to execute
took 76.773000 seconds to execute
took 77.498000 seconds to execute
took 72.944000 seconds to execute
took 72.988000 seconds to execute
took 77.015000 seconds to execute
Testing format VYUY (0x59555956) on A.1
took 77.795000 seconds to execute
took 72.955000 seconds to execute
took 72.945000 seconds to execute
took 77.114000 seconds to execute
took 77.893000 seconds to execute
took 72.930000 seconds to execute
took 72.949000 seconds to execute
took 73.089000 seconds to execute
Testing format AR24 (0x34325241) on A.2
Subtest pixel-format-pipe-A-planes: SUCCESS (12.297s)
This is the same test, with the new conversion code:
./build/tests/kms_plane --run-subtest pixel-format-pipe-a-planes
IGT-Version: 1.23-gec936cad (x86_64) (Linux: 4.18.0-1-amd64 x86_64) Starting subtest: pixel-format-pipe-A-planes
Testing format XR24 (0x34325258) on A.0
Testing format RG16 (0x36314752) on A.0
Testing format XR30 (0x30335258) on A.0
Testing format XR24 (0x34325258) on A.1
Testing format YUYV (0x56595559) on A.1
took 79.515000 seconds to execute
took 79.751000 seconds to execute
took 79.271000 seconds to execute
took 75.816000 seconds to execute
took 75.774000 seconds to execute
took 79.381000 seconds to execute
took 78.029000 seconds to execute
took 78.134000 seconds to execute
Testing format YVYU (0x55595659) on A.1
took 78.705000 seconds to execute
took 78.814000 seconds to execute
took 78.498000 seconds to execute
took 78.636000 seconds to execute
took 78.821000 seconds to execute
took 78.706000 seconds to execute
took 78.662000 seconds to execute
took 78.800000 seconds to execute
Testing format UYVY (0x59565955) on A.1
took 78.731000 seconds to execute
took 78.949000 seconds to execute
took 79.031000 seconds to execute
took 79.433000 seconds to execute
took 77.495000 seconds to execute
took 79.217000 seconds to execute
took 78.766000 seconds to execute
took 78.944000 seconds to execute
Testing format VYUY (0x59555956) on A.1
took 74.723000 seconds to execute
took 79.081000 seconds to execute
took 78.154000 seconds to execute
took 78.163000 seconds to execute
took 78.282000 seconds to execute
took 78.346000 seconds to execute
took 78.326000 seconds to execute
took 78.272000 seconds to execute
Testing format AR24 (0x34325241) on A.2
Subtest pixel-format-pipe-A-planes: SUCCESS (13.881s)
So it looks like we're losing 2-3ms per conversion (on a 1920x1080 frame). It seems pretty negligible to me, but you're the jugde here :)
Maxime
Signature · 3 lines
Maxime Ripard, Bootlin Embedded Linux and Kernel engineering https://bootlin.com
_______________________________________________
igt-dev mailing list [email protected]
https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 03/13] igt: fb: generic YUV convertion function
In reply to Maxime RipardHi,
Quoted 8 lines · Maxime Ripard
The current way we work when we want to support a new YUV format to IGT, we also need to add new function to convert to and from XRGB8888.
This doesn't really scale however, and creates a lot of code to maintain. In order to work around this, create a generic function to convert to RGB and one to convert from RGB.
The only thing that is needed now is to add new parameters, and that's it.
Pushed, thanks!
Cheers,
Paul
Quoted 638 lines
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 537 ++++++++++++++++----------------------------------- 1 file changed, 171 insertions(+), 366 deletions(-)lib/igt_fb.c +171
-366@@ -1577,424 +1577,237 @@ static void convert_src_put(const struct fb_convert *cvt,free (src_buf); } -static void convert_nv12_to_rgb24 (struct fb_convert *cvt) +struct yuv_parameters { +unsigned y_inc; +unsigned uv_inc; +unsigned y_stride; +unsigned uv_stride; +unsigned y_offset; +unsigned u_offset; +unsigned v_offset; +}; + +static void get_yuv_parameters (struct igt_fb *fb,struct yuv_parameters *params) { -int i, j; -const uint8_t *y, *uv; -uint8_t *rgb24 = cvt->dst.ptr; -unsigned int rgb24_stride = cvt->dst.fb->strides[0 ]; -unsigned int planar_stride = cvt->src.fb->strides[0 ]; -struct igt_mat4 m =igt_ycbcr_to_rgb_matrix (cvt->src.fb->color_encoding, - cvt->src.fb->color_range); -uint8_t *buf; - -igt_assert (cvt->src.fb->drm_format ==DRM_FORMAT_NV12 && - cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 ); - - buf =convert_src_get (cvt); - y = buf + cvt->src.fb->offsets[0 ]; - uv = buf + cvt->src.fb->offsets[1 ]; - -for (i =0 ; i < cvt->dst.fb->height /2 ; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x2 pixel blocks */ -struct igt_vec4 yuv[4 ]; -struct igt_vec4 rgb[4 ]; - - yuv[0 ].d[0 ] = y[j *2 +0 ]; - yuv[1 ].d[0 ] = y[j *2 +1 ]; - yuv[2 ].d[0 ] = y[j *2 +0 + planar_stride]; - yuv[3 ].d[0 ] = y[j *2 +1 + planar_stride]; - - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = yuv[2 ].d[1 ] = yuv[3 ].d[1 ] = uv[j *2 +0 ]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = yuv[2 ].d[2 ] = yuv[3 ].d[2 ] = uv[j *2 +1 ]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] = yuv[2 ].d[3 ] = yuv[3 ].d[3 ] =1.0f ; - - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); - rgb[2 ] =igt_matrix_transform (&m, &yuv[2 ]); - rgb[3 ] =igt_matrix_transform (&m, &yuv[3 ]); - -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +4 ], &rgb[1 ]); -write_rgb (&rgb24[j *8 +0 + rgb24_stride], &rgb[2 ]); -write_rgb (&rgb24[j *8 +4 + rgb24_stride], &rgb[3 ]); - } - -if (cvt->dst.fb->width &1 ) { -/* Convert 1x2 pixel block */ -struct igt_vec4 yuv[2 ]; -struct igt_vec4 rgb[2 ]; - - yuv[0 ].d[0 ] = y[j *2 +0 ]; - yuv[1 ].d[0 ] = y[j *2 +0 + planar_stride]; +igt_assert (igt_format_is_yuv (fb->drm_format)); - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = uv[j *2 +0 ]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = uv[j *2 +1 ]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] =1.0f ; +switch (fb->drm_format) { +case DRM_FORMAT_NV12 : + params->y_inc =1 ; + params->uv_inc =2 ; +break ; - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); - -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +0 + rgb24_stride], &rgb[1 ]); - } - - rgb24 +=2 * rgb24_stride; - y +=2 * planar_stride; - uv += planar_stride; +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVYU : +case DRM_FORMAT_UYVY : +case DRM_FORMAT_VYUY : + params->y_inc =2 ; + params->uv_inc =4 ; +break ; + +case DRM_FORMAT_XYUV8888 : + params->y_inc =4 ; + params->uv_inc =4 ; +break ; } -if (cvt->dst.fb->height &1 ) { -/* Convert last row */ -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 yuv[2 ]; -struct igt_vec4 rgb[2 ]; - - yuv[0 ].d[0 ] = y[j *2 +0 ]; - yuv[1 ].d[0 ] = y[j *2 +1 ]; - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = uv[j *2 +0 ]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = uv[j *2 +1 ]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] =1.0f ; +switch (fb->drm_format) { +case DRM_FORMAT_NV12 : + params->y_stride = fb->strides[0 ]; + params->uv_stride = fb->strides[1 ]; +break ; - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVYU : +case DRM_FORMAT_UYVY : +case DRM_FORMAT_VYUY : +case DRM_FORMAT_XYUV8888 : + params->y_stride = fb->strides[0 ]; + params->uv_stride = fb->strides[0 ]; +break ; + } -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +4 ], &rgb[0 ]); - } +switch (fb->drm_format) { +case DRM_FORMAT_NV12 : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[1 ]; + params->v_offset = fb->offsets[1 ] +1 ; +break ; -if (cvt->dst.fb->width &1 ) { -/* Convert single pixel */ -struct igt_vec4 yuv; -struct igt_vec4 rgb; +case DRM_FORMAT_YUYV : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[0 ] +1 ; + params->v_offset = fb->offsets[0 ] +3 ; +break ; - yuv.d[0 ] = y[j *2 +0 ]; - yuv.d[1 ] = uv[j *2 +0 ]; - yuv.d[2 ] = uv[j *2 +1 ]; - yuv.d[3 ] =1.0f ; +case DRM_FORMAT_YVYU : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[0 ] +3 ; + params->v_offset = fb->offsets[0 ] +1 ; +break ; - rgb =igt_matrix_transform (&m, &yuv); +case DRM_FORMAT_UYVY : + params->y_offset = fb->offsets[0 ] +1 ; + params->u_offset = fb->offsets[0 ]; + params->v_offset = fb->offsets[0 ] +2 ; +break ; -write_rgb (&rgb24[j *8 +0 ], &rgb); - } +case DRM_FORMAT_VYUY : + params->y_offset = fb->offsets[0 ] +1 ; + params->u_offset = fb->offsets[0 ] +2 ; + params->v_offset = fb->offsets[0 ]; +break ; + +case DRM_FORMAT_XYUV8888 : + params->y_offset = fb->offsets[0 ] +1 ; + params->u_offset = fb->offsets[0 ] +2 ; + params->v_offset = fb->offsets[0 ] +3 ; +break ; } - -convert_src_put (cvt, buf); } -static void convert_yuv444_to_rgb24 (struct fb_convert *cvt) +static void convert_yuv_to_rgb24 (struct fb_convert *cvt) { +const struct format_desc_struct *src_fmt = +lookup_drm_format (cvt->src.fb->drm_format);int i, j; -uint8_t *yuv24; +uint8_t bpp =4 ; +uint8_t *y, *u, *v;uint8_t *rgb24 = cvt->dst.ptr; -unsigned rgb24_stride = cvt->dst.fb->strides[0 ], xyuv_stride = cvt->src.fb->strides[0 ]; -uint8_t *buf =malloc (cvt->src.fb->size); +unsigned int rgb24_stride = cvt->dst.fb->strides[0 ];struct igt_mat4 m =igt_ycbcr_to_rgb_matrix (cvt->src.fb->color_encoding, cvt->src.fb->color_range); +uint8_t *buf; +struct yuv_parameters params = { }; -/* -* Reading from the BO is awfully slow because of lack of read caching, -* it's faster to copy the whole BO to a temporary buffer and convert -* from there. -*/ -igt_memcpy_from_wc (buf, cvt->src.ptr + cvt->src.fb->offsets[0 ], cvt->src.fb->size); - yuv24 = buf; +igt_assert (cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 && +igt_format_is_yuv (cvt->src.fb->drm_format)); + + buf =convert_src_get (cvt); +get_yuv_parameters (cvt->src.fb, ¶ms); + y = buf + params.y_offset; + u = buf + params.u_offset; + v = buf + params.v_offset;for (i =0 ; i < cvt->dst.fb->height; i++) { +const uint8_t *y_tmp = y; +const uint8_t *u_tmp = u; +const uint8_t *v_tmp = v; +uint8_t *rgb_tmp = rgb24; +for (j =0 ; j < cvt->dst.fb->width; j++) { -float y, u, v; -struct igt_vec4 yuv; -struct igt_vec4 rgb; - - v = yuv24[i * xyuv_stride + j *4 ]; - u = yuv24[i * xyuv_stride + j *4 +1 ]; - y = yuv24[i * xyuv_stride + j *4 +2 ]; - yuv.d[0 ] = y; - yuv.d[1 ] = u; - yuv.d[2 ] = v; +struct igt_vec4 rgb, yuv; + + yuv.d[0 ] = *y_tmp; + yuv.d[1 ] = *u_tmp; + yuv.d[2 ] = *v_tmp; yuv.d[3 ] =1.0f ; rgb =igt_matrix_transform (&m, &yuv); +write_rgb (rgb_tmp, &rgb); -write_rgb (&rgb24[i * rgb24_stride + j *4 ], &rgb); - } - } - -free (buf); -} - - -static void convert_rgb24_to_yuv444 (struct fb_convert *cvt) -{ -int i, j; -uint8_t *rgb24; -uint8_t *yuv444 = cvt->dst.ptr + cvt->dst.fb->offsets[0 ]; -unsigned int rgb24_stride = cvt->src.fb->strides[0 ], xyuv_stride = cvt->dst.fb->strides[0 ]; -struct igt_mat4 m =igt_rgb_to_ycbcr_matrix (cvt->dst.fb->color_encoding, - cvt->dst.fb->color_range); - - rgb24 = cvt->src.ptr; + rgb_tmp += bpp; + y_tmp += params.y_inc; -igt_assert_f (cvt->dst.fb->drm_format ==DRM_FORMAT_XYUV8888 , -"Conversion not implemented for !XYUV packed formats\n" ); - -for (i =0 ; i < cvt->dst.fb->height; i++) { -for (j =0 ; j < cvt->dst.fb->width; j++) { -struct igt_vec4 rgb; -struct igt_vec4 yuv; - -read_rgb (&rgb, &rgb24[i * rgb24_stride + j *4 ]); +if ((src_fmt->hsub ==1 ) || (j % src_fmt->hsub)) { + u_tmp += params.uv_inc; + v_tmp += params.uv_inc; + } + } - yuv =igt_matrix_transform (&m, &rgb); + rgb24 += rgb24_stride; + y += params.y_stride; - yuv444[i * xyuv_stride + j *4 ] = yuv.d[2 ]; - yuv444[i * xyuv_stride + j *4 +1 ] = yuv.d[1 ]; - yuv444[i * xyuv_stride + j *4 +2 ] = yuv.d[0 ]; +if ((src_fmt->vsub ==1 ) || (i % src_fmt->vsub)) { + u += params.uv_stride; + v += params.uv_stride; } } + +convert_src_put (cvt, buf); } -static void convert_rgb24_to_nv12 (struct fb_convert *cvt) +static void convert_rgb24_to_yuv (struct fb_convert *cvt) { +const struct format_desc_struct *dst_fmt = +lookup_drm_format (cvt->dst.fb->drm_format);int i, j; -uint8_t *y = cvt->dst.ptr + cvt->dst.fb->offsets[0 ]; -uint8_t *uv = cvt->dst.ptr + cvt->dst.fb->offsets[1 ]; +uint8_t *y, *u, *v;const uint8_t *rgb24 = cvt->src.ptr; +uint8_t bpp =4 ;unsigned rgb24_stride = cvt->src.fb->strides[0 ]; -unsigned planar_stride = cvt->dst.fb->strides[0 ];struct igt_mat4 m =igt_rgb_to_ycbcr_matrix (cvt->dst.fb->color_encoding, cvt->dst.fb->color_range); +struct yuv_parameters params = { };igt_assert (cvt->src.fb->drm_format ==DRM_FORMAT_XRGB8888 && - cvt->dst.fb->drm_format ==DRM_FORMAT_NV12 ); +igt_format_is_yuv (cvt->dst.fb->drm_format)); -for (i =0 ; i < cvt->dst.fb->height /2 ; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x2 pixel blocks */ -struct igt_vec4 rgb[4 ]; -struct igt_vec4 yuv[4 ]; +get_yuv_parameters (cvt->dst.fb, ¶ms); + y = cvt->dst.ptr + params.y_offset; + u = cvt->dst.ptr + params.u_offset; + v = cvt->dst.ptr + params.v_offset; -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +4 ]); -read_rgb (&rgb[2 ], &rgb24[j *8 +0 + rgb24_stride]); -read_rgb (&rgb[3 ], &rgb24[j *8 +4 + rgb24_stride]); - - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); - yuv[2 ] =igt_matrix_transform (&m, &rgb[2 ]); - yuv[3 ] =igt_matrix_transform (&m, &rgb[3 ]); - - y[j *2 +0 ] = yuv[0 ].d[0 ]; - y[j *2 +1 ] = yuv[1 ].d[0 ]; - y[j *2 +0 + planar_stride] = yuv[2 ].d[0 ]; - y[j *2 +1 + planar_stride] = yuv[3 ].d[0 ]; +for (i =0 ; i < cvt->dst.fb->height; i++) { +const uint8_t *rgb_tmp = rgb24; +uint8_t *y_tmp = y; +uint8_t *u_tmp = u; +uint8_t *v_tmp = v; -/* -* We assume the MPEG2 chroma siting convention, where -* pixel center for Cb'Cr' is between the left top and -* bottom pixel in a 2x2 block, so take the average. -*/ - uv[j *2 +0 ] = (yuv[0 ].d[1 ] + yuv[2 ].d[1 ]) /2.0f ; - uv[j *2 +1 ] = (yuv[0 ].d[2 ] + yuv[2 ].d[2 ]) /2.0f ; - } +for (j =0 ; j < cvt->dst.fb->width; j++) { +const uint8_t *pair_rgb24 = rgb_tmp; +struct igt_vec4 pair_rgb, rgb; +struct igt_vec4 pair_yuv, yuv; -if (cvt->dst.fb->width &1 ) { -/* Convert 1x2 pixel block */ -struct igt_vec4 rgb[2 ]; -struct igt_vec4 yuv[2 ]; +read_rgb (&rgb, rgb_tmp); + yuv =igt_matrix_transform (&m, &rgb); -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +0 + rgb24_stride]); + rgb_tmp += bpp; - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); + *y_tmp = yuv.d[0 ]; + y_tmp += params.y_inc; - y[j *2 +0 ] = yuv[0 ].d[0 ]; - y[j *2 +0 + planar_stride] = yuv[1 ].d[0 ]; +if ((i % dst_fmt->vsub) || (j % dst_fmt->hsub)) +continue ;/* * We assume the MPEG2 chroma siting convention, where * pixel center for Cb'Cr' is between the left top and * bottom pixel in a 2x2 block, so take the average. +* +* Therefore, if we use subsampling, we only really care +* about two pixels all the time, either the two +* subsequent pixels horizontally, vertically, or the +* two corners in a 2x2 block. +* +* The only corner case is when we have an odd number of +* pixels, but this can be handled pretty easily by not +* incrementing the paired pixel pointer in the +* direction it's odd in. */ - uv[j *2 +0 ] = (yuv[0 ].d[1 ] + yuv[1 ].d[1 ]) /2.0f ; - uv[j *2 +1 ] = (yuv[0 ].d[2 ] + yuv[1 ].d[2 ]) /2.0f ; - } - - rgb24 +=2 * rgb24_stride; - y +=2 * planar_stride; - uv += planar_stride; - } - -/* Last row cannot be interpolated between 2 pixels, take the single value */ -if (cvt->dst.fb->height &1 ) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 rgb[2 ]; -struct igt_vec4 yuv[2 ]; - -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +4 ]); - - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); +if (j != (cvt->dst.fb->width -1 )) + pair_rgb24 += (dst_fmt->hsub -1 ) * bpp; - y[j *2 +0 ] = yuv[0 ].d[0 ]; - y[j *2 +1 ] = yuv[1 ].d[0 ]; - uv[j *2 +0 ] = yuv[0 ].d[1 ]; - uv[j *2 +1 ] = yuv[0 ].d[2 ]; - } - -if (cvt->dst.fb->width &1 ) { -/* Convert single pixel */ -struct igt_vec4 rgb; -struct igt_vec4 yuv; +if (i != (cvt->dst.fb->height -1 )) + pair_rgb24 += rgb24_stride * (dst_fmt->vsub -1 ); -read_rgb (&rgb, &rgb24[j *8 +0 ]); +read_rgb (&pair_rgb, pair_rgb24); + pair_yuv =igt_matrix_transform (&m, &pair_rgb); - yuv =igt_matrix_transform (&m, &rgb); + *u_tmp = (yuv.d[1 ] + pair_yuv.d[1 ]) /2.0f ; + *v_tmp = (yuv.d[2 ] + pair_yuv.d[2 ]) /2.0f ; - y[j *2 +0 ] = yuv.d[0 ]; - uv[j *2 +0 ] = yuv.d[1 ]; - uv[j *2 +1 ] = yuv.d[2 ]; - } - } -} - -/* { Y0, U, Y1, V } */ -static const unsigned char swizzle_yuyv[] = {0 ,1 ,2 ,3 }; -static const unsigned char swizzle_yvyu[] = {0 ,3 ,2 ,1 }; -static const unsigned char swizzle_uyvy[] = {1 ,0 ,3 ,2 }; -static const unsigned char swizzle_vyuy[] = {1 ,2 ,3 ,0 }; - -static const unsigned char *yuyv_swizzle (uint32_t format) -{ -switch (format) { -default : -case DRM_FORMAT_YUYV : -return swizzle_yuyv; -case DRM_FORMAT_YVYU : -return swizzle_yvyu; -case DRM_FORMAT_UYVY : -return swizzle_uyvy; -case DRM_FORMAT_VYUY : -return swizzle_vyuy; - } -} - -static void convert_yuyv_to_rgb24 (struct fb_convert *cvt) -{ -int i, j; -const uint8_t *yuyv; -uint8_t *rgb24 = cvt->dst.ptr; -unsigned int rgb24_stride = cvt->dst.fb->strides[0 ]; -unsigned int yuyv_stride = cvt->src.fb->strides[0 ]; -struct igt_mat4 m =igt_ycbcr_to_rgb_matrix (cvt->src.fb->color_encoding, - cvt->src.fb->color_range); -const unsigned char *swz =yuyv_swizzle (cvt->src.fb->drm_format); -uint8_t *buf; - -igt_assert ((cvt->src.fb->drm_format ==DRM_FORMAT_YUYV || - cvt->src.fb->drm_format ==DRM_FORMAT_UYVY || - cvt->src.fb->drm_format ==DRM_FORMAT_YVYU || - cvt->src.fb->drm_format ==DRM_FORMAT_VYUY ) && - cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 ); - - buf =convert_src_get (cvt); - yuyv = buf; - -for (i =0 ; i < cvt->dst.fb->height; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 yuv[2 ]; -struct igt_vec4 rgb[2 ]; - - yuv[0 ].d[0 ] = yuyv[j *4 + swz[0 ]]; - yuv[1 ].d[0 ] = yuyv[j *4 + swz[2 ]]; - yuv[0 ].d[1 ] = yuv[1 ].d[1 ] = yuyv[j *4 + swz[1 ]]; - yuv[0 ].d[2 ] = yuv[1 ].d[2 ] = yuyv[j *4 + swz[3 ]]; - yuv[0 ].d[3 ] = yuv[1 ].d[3 ] =1.0f ; - - rgb[0 ] =igt_matrix_transform (&m, &yuv[0 ]); - rgb[1 ] =igt_matrix_transform (&m, &yuv[1 ]); - -write_rgb (&rgb24[j *8 +0 ], &rgb[0 ]); -write_rgb (&rgb24[j *8 +4 ], &rgb[1 ]); - } - -if (cvt->dst.fb->width &1 ) { -struct igt_vec4 yuv; -struct igt_vec4 rgb; - - yuv.d[0 ] = yuyv[j *4 + swz[0 ]]; - yuv.d[1 ] = yuyv[j *4 + swz[1 ]]; - yuv.d[2 ] = yuyv[j *4 + swz[3 ]]; - yuv.d[3 ] =1.0f ; - - rgb =igt_matrix_transform (&m, &yuv); - -write_rgb (&rgb24[j *8 +0 ], &rgb); + u_tmp += params.uv_inc; + v_tmp += params.uv_inc; } rgb24 += rgb24_stride; - yuyv += yuyv_stride; - } - -convert_src_put (cvt, buf); -} - -static void convert_rgb24_to_yuyv (struct fb_convert *cvt) -{ -int i, j; -uint8_t *yuyv = cvt->dst.ptr; -const uint8_t *rgb24 = cvt->src.ptr; -unsigned rgb24_stride = cvt->src.fb->strides[0 ]; -unsigned yuyv_stride = cvt->dst.fb->strides[0 ]; -struct igt_mat4 m =igt_rgb_to_ycbcr_matrix (cvt->dst.fb->color_encoding, - cvt->dst.fb->color_range); -const unsigned char *swz =yuyv_swizzle (cvt->dst.fb->drm_format); - -igt_assert (cvt->src.fb->drm_format ==DRM_FORMAT_XRGB8888 && - (cvt->dst.fb->drm_format ==DRM_FORMAT_YUYV || - cvt->dst.fb->drm_format ==DRM_FORMAT_UYVY || - cvt->dst.fb->drm_format ==DRM_FORMAT_YVYU || - cvt->dst.fb->drm_format ==DRM_FORMAT_VYUY )); - -for (i =0 ; i < cvt->dst.fb->height; i++) { -for (j =0 ; j < cvt->dst.fb->width /2 ; j++) { -/* Convert 2x1 pixel blocks */ -struct igt_vec4 rgb[2 ]; -struct igt_vec4 yuv[2 ]; - -read_rgb (&rgb[0 ], &rgb24[j *8 +0 ]); -read_rgb (&rgb[1 ], &rgb24[j *8 +4 ]); - - yuv[0 ] =igt_matrix_transform (&m, &rgb[0 ]); - yuv[1 ] =igt_matrix_transform (&m, &rgb[1 ]); + y += params.y_stride; - yuyv[j *4 + swz[0 ]] = yuv[0 ].d[0 ]; - yuyv[j *4 + swz[2 ]] = yuv[1 ].d[0 ]; - yuyv[j *4 + swz[1 ]] = (yuv[0 ].d[1 ] + yuv[1 ].d[1 ]) /2.0f ; - yuyv[j *4 + swz[3 ]] = (yuv[0 ].d[2 ] + yuv[1 ].d[2 ]) /2.0f ; +if ((i % dst_fmt->vsub) == (dst_fmt->vsub -1 )) { + u += params.uv_stride; + v += params.uv_stride; } - -if (cvt->dst.fb->width &1 ) { -struct igt_vec4 rgb; -struct igt_vec4 yuv; - -read_rgb (&rgb, &rgb24[j *8 +0 ]); - - yuv =igt_matrix_transform (&m, &rgb); - - yuyv[j *4 + swz[0 ]] = yuv.d[0 ]; - yuyv[j *4 + swz[1 ]] = yuv.d[1 ]; - yuyv[j *4 + swz[3 ]] = yuv.d[2 ]; - } - - rgb24 += rgb24_stride; - yuyv += yuyv_stride; } }@@ -2046,31 +1859,23 @@ static void fb_convert(struct fb_convert *cvt) }else if (cvt->dst.fb->drm_format ==DRM_FORMAT_XRGB8888 ) {switch (cvt->src.fb->drm_format) {case DRM_FORMAT_XYUV8888 : -convert_yuv444_to_rgb24 (cvt); -return ;case DRM_FORMAT_NV12 : -convert_nv12_to_rgb24 (cvt); -return ; -case DRM_FORMAT_YUYV : -case DRM_FORMAT_YVYU :case DRM_FORMAT_UYVY :case DRM_FORMAT_VYUY : -convert_yuyv_to_rgb24 (cvt); +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVYU : +convert_yuv_to_rgb24 (cvt);return ; } }else if (cvt->src.fb->drm_format ==DRM_FORMAT_XRGB8888 ) {switch (cvt->dst.fb->drm_format) {case DRM_FORMAT_XYUV8888 : -convert_rgb24_to_yuv444 (cvt); -return ;case DRM_FORMAT_NV12 : -convert_rgb24_to_nv12 (cvt); -return ;case DRM_FORMAT_YUYV :case DRM_FORMAT_YVYU :case DRM_FORMAT_UYVY :case DRM_FORMAT_VYUY : -convert_rgb24_to_yuyv (cvt); +convert_rgb24_to_yuv (cvt);return ; } }
Signature · 8 lines
Paul Kocialkowski, Bootlin Embedded Linux and kernel engineering https://bootlin.com _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
PATCH v5 4/13 igt: fb: Move i915 YUV buffer clearing code to a function
In reply to Maxime RipardThe YUV buffer allocation path for the i915 driver also has some code to clear a YUV buffer.
As that is going to be useful for all the other drivers, let's move it to a separate function.
Reviewed-by Paul Kocialkowski <[email protected]>
Signed-off-by Maxime Ripard <[email protected]>
--- lib/igt_fb.c | 80 +++++++++++++++++++++++++++------------------------- 1 file changed, 42 insertions(+), 38 deletions(-)
lib/igt_fb.c +42 -38
@@ -484,6 +484,47 @@ uint64_t igt_fb_tiling_to_mod(uint64_t tiling) } } +static void clear_yuv_buffer (struct igt_fb *fb) +{ +bool full_range = fb->color_range ==IGT_COLOR_YCBCR_FULL_RANGE ; +void *ptr; + +igt_assert (igt_format_is_yuv (fb->drm_format)); + +/* Ensure the framebuffer is preallocated */ + ptr =igt_fb_map_buffer (fb->fd, fb); +igt_assert (*(uint32_t *)ptr ==0 ); + +switch (fb->drm_format) { +case DRM_FORMAT_NV12 : +memset (ptr + fb->offsets[0 ], + full_range ?0x00 :0x10 , + fb->strides[0 ] * fb->plane_height[0 ]); +memset (ptr + fb->offsets[1 ], +0x80 , + fb->strides[1 ] * fb->plane_height[1 ]); +break ; +case DRM_FORMAT_XYUV8888 : +wmemset (ptr + fb->offsets[0 ], full_range ?0x00008080 :0x00108080 , + fb->strides[0 ] * fb->plane_height[0 ] /sizeof (wchar_t)); +break ; +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVYU : +wmemset (ptr + fb->offsets[0 ], + full_range ?0x80008000 :0x80108010 , + fb->strides[0 ] * fb->plane_height[0 ] /sizeof (wchar_t)); +break ; +case DRM_FORMAT_UYVY : +case DRM_FORMAT_VYUY : +wmemset (ptr + fb->offsets[0 ], + full_range ?0x00800080 :0x10801080 , + fb->strides[0 ] * fb->plane_height[0 ] /sizeof (wchar_t)); +break ; + } + +igt_fb_unmap_buffer (fb, ptr); +} +/* helpers to create nice-looking framebuffers */ static int create_bo_for_fb (struct igt_fb *fb) {
@@ -501,50 +542,13 @@ static int create_bo_for_fb(struct igt_fb *fb) fb->is_dumb = false;if (is_i915_device (fd)) { -void *ptr; -bool full_range = fb->color_range ==IGT_COLOR_YCBCR_FULL_RANGE ; fb->gem_handle =gem_create (fd, fb->size);gem_set_tiling (fd, fb->gem_handle,igt_fb_mod_to_tiling (fb->tiling), fb->strides[0 ]); - -gem_set_domain (fd, fb->gem_handle, -I915_GEM_DOMAIN_GTT ,I915_GEM_DOMAIN_GTT ); - -/* Ensure the framebuffer is preallocated */ - ptr =gem_mmap__gtt (fd, fb->gem_handle, - fb->size,PROT_READ |PROT_WRITE ); -igt_assert (*(uint32_t *)ptr ==0 ); - -switch (fb->drm_format) { -case DRM_FORMAT_NV12 : -memset (ptr + fb->offsets[0 ], - full_range ?0x00 :0x10 , - fb->strides[0 ] * fb->plane_height[0 ]); -memset (ptr + fb->offsets[1 ], -0x80 , - fb->strides[1 ] * fb->plane_height[1 ]); -break ; -case DRM_FORMAT_XYUV8888 : -wmemset (ptr + fb->offsets[0 ], full_range ?0x00008080 :0x00108080 , - fb->strides[0 ] * fb->plane_height[0 ] /sizeof (wchar_t)); -break ; -case DRM_FORMAT_YUYV : -case DRM_FORMAT_YVYU : -wmemset (ptr + fb->offsets[0 ], - full_range ?0x80008000 :0x80108010 , - fb->strides[0 ] * fb->plane_height[0 ] /sizeof (wchar_t)); -break ; -case DRM_FORMAT_UYVY : -case DRM_FORMAT_VYUY : -wmemset (ptr + fb->offsets[0 ], - full_range ?0x00800080 :0x10801080 , - fb->strides[0 ] * fb->plane_height[0 ] /sizeof (wchar_t)); -break ; - } -gem_munmap (ptr, fb->size); +clear_yuv_buffer (fd);return fb->gem_handle; }else {
Signature · 6 lines
2.20.1 _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 04/13] igt: fb: Move i915 YUV buffer clearing code to a function
Reviewed-by Lyude Paul <[email protected]>
Quoted 126 lines · Maxime Ripard
The YUV buffer allocation path for the i915 driver also has some code to clear a YUV buffer.
As that is going to be useful for all the other drivers, let's move it to a separate function.
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 80 +++++++++++++++++++++++++++------------------------- 1 file changed, 42 insertions(+), 38 deletions(-)lib/igt_fb.c +35
-0@@ -484,6 +484,47 @@ uint64_t igt_fb_tiling_to_mod(uint64_t tiling) } } +static void clear_yuv_buffer (struct igt_fb *fb) +{ +bool full_range = fb->color_range ==IGT_COLOR_YCBCR_FULL_RANGE ; +void *ptr; + +igt_assert (igt_format_is_yuv (fb->drm_format)); + +/* Ensure the framebuffer is preallocated */ + ptr =igt_fb_map_buffer (fb->fd, fb); +igt_assert (*(uint32_t *)ptr ==0 ); + +switch (fb->drm_format) { +case DRM_FORMAT_NV12 : +memset (ptr + fb->offsets[0 ], + full_range ?0x00 :0x10 , + fb->strides[0 ] * fb->plane_height[0 ]); +memset (ptr + fb->offsets[1 ], +0x80 , + fb->strides[1 ] * fb->plane_height[1 ]); +break ; +case DRM_FORMAT_XYUV8888 : +wmemset (ptr + fb->offsets[0 ], full_range ?0x00008080 :0x00108080 , + fb->strides[0 ] * fb->plane_height[0 ] /sizeof (wchar_t)); +break ; +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVYU : +wmemset (ptr + fb->offsets[0 ], + full_range ?0x80008000 :0x80108010 , + fb->strides[0 ] * fb->plane_height[0 ] /sizeof (wchar_t)); +break ; +case DRM_FORMAT_UYVY : +case DRM_FORMAT_VYUY : + wmemset(ptr + fb->offsets[0], + full_range ? 0x00800080 : 0x10801080, + fb->strides[0] * fb->plane_height[0] /
sizeof(wchar_t));lib/igt_fb.c +6
-29+break ; + } + +igt_fb_unmap_buffer (fb, ptr); +} +/* helpers to create nice-looking framebuffers */ static int create_bo_for_fb (struct igt_fb *fb) {@@ -501,50 +542,13 @@ static int create_bo_for_fb(struct igt_fb *fb) fb->is_dumb = false;if (is_i915_device (fd)) { -void *ptr; -bool full_range = fb->color_range ==IGT_COLOR_YCBCR_FULL_RANGE ; fb->gem_handle =gem_create (fd, fb->size);gem_set_tiling (fd, fb->gem_handle,igt_fb_mod_to_tiling (fb->tiling), fb->strides[0 ]); - -gem_set_domain (fd, fb->gem_handle, -I915_GEM_DOMAIN_GTT , I915_GEM_DOMAIN_GTT); - -/* Ensure the framebuffer is preallocated */ - ptr =gem_mmap__gtt (fd, fb->gem_handle, - fb->size,PROT_READ |PROT_WRITE ); -igt_assert (*(uint32_t *)ptr ==0 ); - -switch (fb->drm_format) { -case DRM_FORMAT_NV12 : -memset (ptr + fb->offsets[0 ], - full_range ?0x00 :0x10 , - fb->strides[0 ] * fb->plane_height[0 ]); -memset (ptr + fb->offsets[1 ], -0x80 , - fb->strides[1 ] * fb->plane_height[1 ]); -break ; -case DRM_FORMAT_XYUV8888 : -wmemset (ptr + fb->offsets[0 ], full_range ? 0x00008080 : 0x00108080, - fb->strides[0 ] * fb->plane_height[0 ] / sizeof(wchar_t)); -break ; -case DRM_FORMAT_YUYV : -case DRM_FORMAT_YVYU : - wmemset(ptr + fb->offsets[0], - full_range ? 0x80008000 : 0x80108010, - fb->strides[0] * fb->plane_height[0] /
sizeof(wchar_t));lib/igt_fb.c +0
-6-break ; -case DRM_FORMAT_UYVY : -case DRM_FORMAT_VYUY : - wmemset(ptr + fb->offsets[0], - full_range ? 0x00800080 : 0x10801080, - fb->strides[0] * fb->plane_height[0] /
sizeof(wchar_t));lib/igt_fb.c +1
-3-break ; - } -gem_munmap (ptr, fb->size); +clear_yuv_buffer (fd);return fb->gem_handle; }else {
Signature · 7 lines
Cheers, Lyude Paul _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 04/13] igt: fb: Move i915 YUV buffer clearing code to a function
In reply to Maxime RipardHi,
Quoted 5 lines · Maxime Ripard
The YUV buffer allocation path for the i915 driver also has some code to clear a YUV buffer.
As that is going to be useful for all the other drivers, let's move it to a separate function.
Pushed, thanks!
Cheers,
Paul
Quoted 110 lines
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 80 +++++++++++++++++++++++++++------------------------- 1 file changed, 42 insertions(+), 38 deletions(-)lib/igt_fb.c +42
-38@@ -484,6 +484,47 @@ uint64_t igt_fb_tiling_to_mod(uint64_t tiling) } } +static void clear_yuv_buffer (struct igt_fb *fb) +{ +bool full_range = fb->color_range ==IGT_COLOR_YCBCR_FULL_RANGE ; +void *ptr; + +igt_assert (igt_format_is_yuv (fb->drm_format)); + +/* Ensure the framebuffer is preallocated */ + ptr =igt_fb_map_buffer (fb->fd, fb); +igt_assert (*(uint32_t *)ptr ==0 ); + +switch (fb->drm_format) { +case DRM_FORMAT_NV12 : +memset (ptr + fb->offsets[0 ], + full_range ?0x00 :0x10 , + fb->strides[0 ] * fb->plane_height[0 ]); +memset (ptr + fb->offsets[1 ], +0x80 , + fb->strides[1 ] * fb->plane_height[1 ]); +break ; +case DRM_FORMAT_XYUV8888 : +wmemset (ptr + fb->offsets[0 ], full_range ?0x00008080 :0x00108080 , + fb->strides[0 ] * fb->plane_height[0 ] /sizeof (wchar_t)); +break ; +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVYU : +wmemset (ptr + fb->offsets[0 ], + full_range ?0x80008000 :0x80108010 , + fb->strides[0 ] * fb->plane_height[0 ] /sizeof (wchar_t)); +break ; +case DRM_FORMAT_UYVY : +case DRM_FORMAT_VYUY : +wmemset (ptr + fb->offsets[0 ], + full_range ?0x00800080 :0x10801080 , + fb->strides[0 ] * fb->plane_height[0 ] /sizeof (wchar_t)); +break ; + } + +igt_fb_unmap_buffer (fb, ptr); +} +/* helpers to create nice-looking framebuffers */ static int create_bo_for_fb (struct igt_fb *fb) {@@ -501,50 +542,13 @@ static int create_bo_for_fb(struct igt_fb *fb) fb->is_dumb = false;if (is_i915_device (fd)) { -void *ptr; -bool full_range = fb->color_range ==IGT_COLOR_YCBCR_FULL_RANGE ; fb->gem_handle =gem_create (fd, fb->size);gem_set_tiling (fd, fb->gem_handle,igt_fb_mod_to_tiling (fb->tiling), fb->strides[0 ]); - -gem_set_domain (fd, fb->gem_handle, -I915_GEM_DOMAIN_GTT ,I915_GEM_DOMAIN_GTT ); - -/* Ensure the framebuffer is preallocated */ - ptr =gem_mmap__gtt (fd, fb->gem_handle, - fb->size,PROT_READ |PROT_WRITE ); -igt_assert (*(uint32_t *)ptr ==0 ); - -switch (fb->drm_format) { -case DRM_FORMAT_NV12 : -memset (ptr + fb->offsets[0 ], - full_range ?0x00 :0x10 , - fb->strides[0 ] * fb->plane_height[0 ]); -memset (ptr + fb->offsets[1 ], -0x80 , - fb->strides[1 ] * fb->plane_height[1 ]); -break ; -case DRM_FORMAT_XYUV8888 : -wmemset (ptr + fb->offsets[0 ], full_range ?0x00008080 :0x00108080 , - fb->strides[0 ] * fb->plane_height[0 ] /sizeof (wchar_t)); -break ; -case DRM_FORMAT_YUYV : -case DRM_FORMAT_YVYU : -wmemset (ptr + fb->offsets[0 ], - full_range ?0x80008000 :0x80108010 , - fb->strides[0 ] * fb->plane_height[0 ] /sizeof (wchar_t)); -break ; -case DRM_FORMAT_UYVY : -case DRM_FORMAT_VYUY : -wmemset (ptr + fb->offsets[0 ], - full_range ?0x00800080 :0x10801080 , - fb->strides[0 ] * fb->plane_height[0 ] /sizeof (wchar_t)); -break ; - } -gem_munmap (ptr, fb->size); +clear_yuv_buffer (fd);return fb->gem_handle; }else {
Signature · 8 lines
Paul Kocialkowski, Bootlin Embedded Linux and kernel engineering https://bootlin.com _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
PATCH v5 5/13 igt: fb: Refactor dumb buffer allocation path
In reply to Maxime RipardThe else condition is not needed, since all the other conditions return when they are done.
Move the KMS dumb buffer allocation outside of the outer else condition, this will also allow to ease later changes.
Reviewed-by Paul Kocialkowski <[email protected]>
Signed-off-by Maxime Ripard <[email protected]>
--- lib/igt_fb.c | 25 +++++++++++++++++-------- 1 file changed, 17 insertions(+), 8 deletions(-)
lib/igt_fb.c +17 -8
@@ -557,15 +557,24 @@ static int create_bo_for_fb(struct igt_fb *fb)igt_require (driver_has_gem_api);return -EINVAL ; } - }else { - fb->is_dumb = true; - - fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb->height, - fb->plane_bpp[0 ], - &fb->strides[0 ], &fb->size); - -return fb->gem_handle; } + +/* +* The current dumb buffer allocation API doesn't really allow to +* specify a custom size or stride. Yet the caller is free to specify +* them, so we need to make sure to error out in this case. +*/ +igt_assert (fb->size ==0 ); +igt_assert (fb->strides[0 ] ==0 ); + + fb->size =calc_fb_size (fb); + + fb->is_dumb = true; + fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb->height, + fb->plane_bpp[0 ], + &fb->strides[0 ], &fb->size); + +return fb->gem_handle; }/**
Signature · 6 lines
2.20.1 _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 05/13] igt: fb: Refactor dumb buffer allocation path
Quoted 39 lines · Maxime Ripard
The else condition is not needed, since all the other conditions return
when they are done.Move the KMS dumb buffer allocation outside of the outer else condition,
this will also allow to ease later changes.Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 25 +++++++++++++++++-------- 1 file changed, 17 insertions(+), 8 deletions(-)lib/igt_fb.c +3
-8@@ -557,15 +557,24 @@ static int create_bo_for_fb(struct igt_fb *fb)igt_require (driver_has_gem_api);return -EINVAL ; } - }else { - fb->is_dumb = true; - - fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb- >height, - fb->plane_bpp[0 ], - &fb->strides[0 ], &fb->size); - -return fb->gem_handle; } + +/* +* The current dumb buffer allocation API doesn't really allow to
lib/igt_fb.c +1
-0+* specify a custom size or stride. Yet the caller is free to specify
lib/igt_fb.c static int create_bo_for_fb(struct igt_fb *fb) · line 568
Quoted 4 lines
lib/igt_fb.c +4
-0+* them, so we need to make sure to error out in this case. +*/ +igt_assert (fb->size ==0 ); +igt_assert (fb->strides[0 ] ==0 );
Hi Maxime, thanks for doing this. I am not an expert but do you think this should be igt_skip instead? Tests will simply fail where they should really be skipping if driver don't support the feature.
lib/igt_fb.c static int create_bo_for_fb(struct igt_fb *fb) · line 580
Quoted 13 lines
lib/igt_fb.c +7
-0+ + fb->size =calc_fb_size (fb); + + fb->is_dumb = true; + fb->gem_handle = kmstest_dumb_create(fd, fb->width, fb->height, + fb->plane_bpp[0], + &fb->strides[0], &fb-Quoted 1 line
size);
lib/igt_fb.c +2
-0+ +return fb->gem_handle; }/**
_______________________________________________
igt-dev mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 05/13] igt: fb: Refactor dumb buffer allocation path
Hi!
Quoted 40 lines · Deepak Singh Rawat
Quoted 39 lines · Maxime Ripard
The else condition is not needed, since all the other conditions return
when they are done.Move the KMS dumb buffer allocation outside of the outer else condition,
this will also allow to ease later changes.Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 25 +++++++++++++++++-------- 1 file changed, 17 insertions(+), 8 deletions(-)lib/igt_fb.c +3
-8@@ -557,15 +557,24 @@ static int create_bo_for_fb(struct igt_fb *fb)igt_require (driver_has_gem_api);return -EINVAL ; } - }else { - fb->is_dumb = true; - - fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb- >height, - fb->plane_bpp[0 ], - &fb->strides[0 ], &fb->size); - -return fb->gem_handle; } + +/* +* The current dumb buffer allocation API doesn't really allow to
lib/igt_fb.c +1
-0+* specify a custom size or stride. Yet the caller is free to specify
Quoted 8 lines
lib/igt_fb.c static int create_bo_for_fb(struct igt_fb *fb) · line 568
Quoted 4 lines
lib/igt_fb.c +4
-0+* them, so we need to make sure to error out in this case. +*/ +igt_assert (fb->size ==0 ); +igt_assert (fb->strides[0 ] ==0 );Hi Maxime, thanks for doing this. I am not an expert but do you think this should be
igt_skipinstead? Tests will simply fail where they should really be skipping if driver don't support the feature.
That's a good suggestion :)
I'm not sure what the proper behaviour here is though. Both approaches make sense I guess, but skiping the test would hide that issue under the carpet which might or might not be an issue, I don't know.
Maxime
Signature · 3 lines
Maxime Ripard, Bootlin Embedded Linux and Kernel engineering https://bootlin.com
_______________________________________________
igt-dev mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 05/13] igt: fb: Refactor dumb buffer allocation path
In reply to Maxime RipardQuoted 31 lines · Maxime Ripard
The else condition is not needed, since all the other conditions return when they are done.
Move the KMS dumb buffer allocation outside of the outer else condition, this will also allow to ease later changes.
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 25 +++++++++++++++++-------- 1 file changed, 17 insertions(+), 8 deletions(-)lib/igt_fb.c +0
-8@@ -557,15 +557,24 @@ static int create_bo_for_fb(struct igt_fb *fb)igt_require (driver_has_gem_api);return -EINVAL ; } - }else { - fb->is_dumb = true; - - fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb- >height, - fb->plane_bpp[0 ], - &fb->strides[0 ], &fb- >size); - -return fb->gem_handle; }
lib/igt_fb.c static int create_bo_for_fb(struct igt_fb *fb) · line 572
Quoted 10 lines
lib/igt_fb.c +10
-0+ +/* +* The current dumb buffer allocation API doesn't really allow to +* specify a custom size or stride. Yet the caller is free to specify +* them, so we need to make sure to error out in this case. +*/ +igt_assert (fb->size ==0 ); +igt_assert (fb->strides[0 ] ==0 ); + + fb->size =calc_fb_size (fb);
^ Is this supposed to be part of a seperate patch?
Quoted 10 lines
lib/igt_fb.c +7
-0+ + fb->is_dumb = true; + fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb->height, + fb->plane_bpp[0 ], + &fb->strides[0 ], &fb->size); + +return fb->gem_handle; }/**
Signature · 7 lines
Cheers, Lyude Paul _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 05/13] igt: fb: Refactor dumb buffer allocation path
Hi!
Quoted 29 lines · Lyude Paul
Quoted 28 lines · Maxime Ripard
The else condition is not needed, since all the other conditions return when they are done.
Move the KMS dumb buffer allocation outside of the outer else condition, this will also allow to ease later changes.
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 25 +++++++++++++++++-------- 1 file changed, 17 insertions(+), 8 deletions(-)lib/igt_fb.c +0
-6@@ -557,15 +557,24 @@ static int create_bo_for_fb(struct igt_fb *fb)igt_require (driver_has_gem_api);return -EINVAL ; } - }else { - fb->is_dumb = true; - - fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb- >height, - fb->plane_bpp[0 ], - &fb->strides[0 ], &fb- >size);
Quoted 15 lines
lib/igt_fb.c static int create_bo_for_fb(struct igt_fb *fb) · line 570
Quoted 13 lines
lib/igt_fb.c +10
-2- -return fb->gem_handle; } + +/* +* The current dumb buffer allocation API doesn't really allow to +* specify a custom size or stride. Yet the caller is free to specify +* them, so we need to make sure to error out in this case. +*/ +igt_assert (fb->size ==0 ); +igt_assert (fb->strides[0 ] ==0 ); + + fb->size =calc_fb_size (fb);^ Is this supposed to be part of a seperate patch?
Yeah, definitely, I'll split it in another patch, thanks! Maxime
Signature · 3 lines
Maxime Ripard, Bootlin Embedded Linux and Kernel engineering https://bootlin.com
_______________________________________________
igt-dev mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/igt-dev
PATCH v5 6/13 igt: fb: Account for all planes bpp
In reply to Maxime RipardWhen allocating a dumb buffer for a format with multiple plane, we need to account for all plane's bpp in order to allocate the proper size.
Let's add all the planes bpp and use the result to allocate our buffer.
Reviewed-by Paul Kocialkowski <[email protected]>
Signed-off-by Maxime Ripard <[email protected]>
--- lib/igt_fb.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
lib/igt_fb.c +6 -2
@@ -528,6 +528,8 @@ static void clear_yuv_buffer(struct igt_fb *fb)/* helpers to create nice-looking framebuffers */ static int create_bo_for_fb (struct igt_fb *fb) { +const struct format_desc_struct *fmt =lookup_drm_format (fb->drm_format); +unsigned int plane, bpp;int fd = fb->fd;if (fb->tiling || fb->size || fb->strides[0 ] ||igt_format_is_yuv (fb->drm_format)) {
@@ -568,11 +570,13 @@ static int create_bo_for_fb(struct igt_fb *fb)igt_assert (fb->strides[0 ] ==0 ); fb->size =calc_fb_size (fb); +for (plane =0 ; plane < fb->num_planes; plane++) + bpp +=DIV_ROUND_UP (fb->plane_bpp[plane], + plane ? fmt->hsub * fmt->vsub :1 ); fb->is_dumb = true; fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb->height, - fb->plane_bpp[0 ], - &fb->strides[0 ], &fb->size); + bpp, &fb->strides[0 ], &fb->size);return fb->gem_handle; }
Signature · 6 lines
2.20.1 _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
PATCH v5 7/13 igt: fb: Don't pass the stride when allocating a dumb, multi-planar buffer
In reply to Maxime RipardThe dumb buffer allocation API only considers a single plane, and even though allocating multi-planar buffers through it is allowed, the stride it gives back is the the width times the bpp passed as an argument.
That doesn't work in our case, since the bpp is going to be the one we give as an argument, but split over three planes so the stride doesn't match anymore.
A proper fix for this would be to have a better dumb buffer allocation API, but for the time being, let's do it that way.
Reviewed-by Paul Kocialkowski <[email protected]>
Signed-off-by Maxime Ripard <[email protected]>
--- lib/igt_fb.c | 17 ++++++++++++++++- 1 file changed, 16 insertions(+), 1 deletion(-)
lib/igt_fb.c +16 -1
@@ -530,6 +530,7 @@ static int create_bo_for_fb(struct igt_fb *fb) {const struct format_desc_struct *fmt =lookup_drm_format (fb->drm_format);unsigned int plane, bpp; +unsigned *strides = &fb->strides[0 ];int fd = fb->fd;if (fb->tiling || fb->size || fb->strides[0 ] ||igt_format_is_yuv (fb->drm_format)) {
@@ -575,8 +576,22 @@ static int create_bo_for_fb(struct igt_fb *fb) plane ? fmt->hsub * fmt->vsub : 1); fb->is_dumb = true; + +/* +* We can't really pass the stride array here since the dumb +* buffer allocation is assuming that it operates on one +* plane, and therefore will calculate the stride as if each +* pixels were stored on a single plane. +* +* This might cause issues at some point on drivers that would +* change the stride of YUV buffers, but we haven't +* encountered any yet. +*/ +if (fb->num_planes >1 ) + strides =NULL ; + fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb->height, - bpp, &fb->strides[0 ], &fb->size); + bpp, strides, &fb->size);return fb->gem_handle; }
Signature · 6 lines
2.20.1 _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 07/13] igt: fb: Don't pass the stride when allocating a dumb, multi-planar buffer
Quoted 32 lines · Maxime Ripard
The dumb buffer allocation API only considers a single plane, and even though allocating multi-planar buffers through it is allowed, the stride it gives back is the the width times the bpp passed as an argument.
That doesn't work in our case, since the bpp is going to be the one we give as an argument, but split over three planes so the stride doesn't match anymore.
A proper fix for this would be to have a better dumb buffer allocation API, but for the time being, let's do it that way.
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 17 ++++++++++++++++- 1 file changed, 16 insertions(+), 1 deletion(-)lib/igt_fb.c +1
-0@@ -530,6 +530,7 @@ static int create_bo_for_fb(struct igt_fb *fb) {const struct format_desc_struct *fmt =lookup_drm_format (fb- >drm_format);unsigned int plane, bpp; +unsigned *strides = &fb->strides[0 ];int fd = fb->fd;if (
fb->tiling||fb->size|| fb->strides[0] || igt_format_is_yuv(fb-Quoted 1 line
drm_format)) {
lib/igt_fb.c static int create_bo_for_fb(struct igt_fb *fb) · line 584
Quoted 9 lines
lib/igt_fb.c +6
-0@@ -575,8 +576,22 @@ static int create_bo_for_fb(struct igt_fb *fb) plane ? fmt->hsub * fmt->vsub : 1); fb->is_dumb = true; + +/* +* We can't really pass the stride array here since the dumb +* buffer allocation is assuming that it operates on one +* plane, and therefore will calculate the stride as if each +* pixels were stored on a single plane.
nitpick: s/each pixels were/each pixel was/
lib/igt_fb.c static int create_bo_for_fb(struct igt_fb *fb) · line 589
Quoted 5 lines
lib/igt_fb.c +5
-0+* +* This might cause issues at some point on drivers that would +* change the stride of YUV buffers, but we haven't +* encountered any yet. +*/
Is it possible to add an igt_assert to check for this? Either way, with those changes:
Reviewed-by Lyude Paul <[email protected]>
Quoted 9 lines
lib/igt_fb.c +4
-1+if (fb->num_planes >1 ) + strides =NULL ; + fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb->height, - bpp, &fb->strides[0 ], &fb->size); + bpp, strides, &fb->size);return fb->gem_handle; }
Signature · 7 lines
Cheers, Lyude Paul _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 07/13] igt: fb: Don't pass the stride when allocating a dumb, multi-planar buffer
Hi!
Quoted 45 lines · Lyude Paul
Quoted 32 lines · Maxime Ripard
The dumb buffer allocation API only considers a single plane, and even though allocating multi-planar buffers through it is allowed, the stride it gives back is the the width times the bpp passed as an argument.
That doesn't work in our case, since the bpp is going to be the one we give as an argument, but split over three planes so the stride doesn't match anymore.
A proper fix for this would be to have a better dumb buffer allocation API, but for the time being, let's do it that way.
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 17 ++++++++++++++++- 1 file changed, 16 insertions(+), 1 deletion(-)lib/igt_fb.c +1
-0@@ -530,6 +530,7 @@ static int create_bo_for_fb(struct igt_fb *fb) {const struct format_desc_struct *fmt =lookup_drm_format (fb- >drm_format);unsigned int plane, bpp; +unsigned *strides = &fb->strides[0 ];int fd = fb->fd;if (
fb->tiling||fb->size|| fb->strides[0] || igt_format_is_yuv(fb-Quoted 1 line
drm_format)) {lib/igt_fb.c static int create_bo_for_fb(struct igt_fb *fb) · line 584
Quoted 9 lines
lib/igt_fb.c +6
-0@@ -575,8 +576,22 @@ static int create_bo_for_fb(struct igt_fb *fb) plane ? fmt->hsub * fmt->vsub : 1); fb->is_dumb = true; + +/* +* We can't really pass the stride array here since the dumb +* buffer allocation is assuming that it operates on one +* plane, and therefore will calculate the stride as if each +* pixels were stored on a single plane. nitpick: s/each pixels were/each pixel was/
Quoted 7 lines
lib/igt_fb.c static int create_bo_for_fb(struct igt_fb *fb) · line 589
Quoted 5 lines
lib/igt_fb.c +5
-0+* +* This might cause issues at some point on drivers that would +* change the stride of YUV buffers, but we haven't +* encountered any yet. +*/ Is it possible to add an
igt_assertto check for this? Either way, with those changes:
I'm not sure we can test this actually. If we don't pass any pointer, then we don't get a stride back, and if we do then we will always get a stride that is wrong for what we're trying to do :/
Maxime
Signature · 3 lines
Maxime Ripard, Bootlin Embedded Linux and Kernel engineering https://bootlin.com
_______________________________________________
igt-dev mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 07/13] igt: fb: Don't pass the stride when allocating a dumb, multi-planar buffer
Quoted 60 lines · Maxime Ripard
Hi!
Quoted 48 lines · Lyude Paul
Quoted 35 lines · Maxime Ripard
The dumb buffer allocation API only considers a single plane, and even though allocating multi-planar buffers through it is allowed, the stride it
gives back is the the width times the bpp passed as an argument.That doesn't work in our case, since the bpp is going to be the one we give
as an argument, but split over three planes so the stride doesn't match anymore.A proper fix for this would be to have a better dumb buffer allocation API,
but for the time being, let's do it that way.Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 17 ++++++++++++++++- 1 file changed, 16 insertions(+), 1 deletion(-)lib/igt_fb.c +1
-0@@ -530,6 +530,7 @@ static int create_bo_for_fb(struct igt_fb *fb) {const struct format_desc_struct *fmt =lookup_drm_format (fb- > drm_format);unsigned int plane, bpp; +unsigned *strides = &fb->strides[0 ];int fd = fb->fd;if (
fb->tiling||fb->size|| fb->strides[0] || igt_format_is_yuv(fb-Quoted 1 line
drm_format)) {lib/igt_fb.c static int create_bo_for_fb(struct igt_fb *fb) · line 584
Quoted 9 lines
lib/igt_fb.c +6
-0@@ -575,8 +576,22 @@ static int create_bo_for_fb(struct igt_fb *fb) plane ? fmt->hsub * fmt->vsub : 1); fb->is_dumb = true; + +/* +* We can't really pass the stride array here since the dumb +* buffer allocation is assuming that it operates on one +* plane, and therefore will calculate the stride as if each +* pixels were stored on a single plane. nitpick: s/each pixels were/each pixel was/
Quoted 8 lines
lib/igt_fb.c static int create_bo_for_fb(struct igt_fb *fb) · line 589
Quoted 5 lines
lib/igt_fb.c +5
-0+* +* This might cause issues at some point on drivers that would +* change the stride of YUV buffers, but we haven't +* encountered any yet. +*/ Is it possible to add an
igt_assertto check for this? Either way, with those
changes:
Quoted 3 lines
I'm not sure we can test this actually. If we don't pass any pointer, then we don't get a stride back, and if we do then we will always get a stride that is wrong for what we're trying to do :/
Ah, makes sense
Quoted 3 lines
Maxime
Signature · 7 lines
Cheers, Lyude Paul _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
PATCH v5 8/13 igt: fb: Clear YUV dumb buffers
In reply to Maxime RipardYUV dumb buffers, just like i915 GEM buffers also need to be cleared once allocated. Make sure it happens.
Reviewed-by Paul Kocialkowski <[email protected]>
Signed-off-by Maxime Ripard <[email protected]>
--- lib/igt_fb.c | 3 +++ 1 file changed, 3 insertions(+)
lib/igt_fb.c +3 -0
@@ -593,6 +593,9 @@ static int create_bo_for_fb(struct igt_fb *fb) fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb->height, bpp, strides, &fb->size); +if (igt_format_is_yuv (fb->drm_format)) +clear_yuv_buffer (fb); +return fb->gem_handle; }
Signature · 6 lines
2.20.1 _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 08/13] igt: fb: Clear YUV dumb buffers
Reviewed-by Lyude Paul <[email protected]>
Quoted 23 lines · Maxime Ripard
YUV dumb buffers, just like i915 GEM buffers also need to be cleared once allocated. Make sure it happens.
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 3 +++ 1 file changed, 3 insertions(+)lib/igt_fb.c +3
-0@@ -593,6 +593,9 @@ static int create_bo_for_fb(struct igt_fb *fb) fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb->height, bpp, strides, &fb->size); +if (igt_format_is_yuv (fb->drm_format)) +clear_yuv_buffer (fb); +return fb->gem_handle; }
Signature · 7 lines
Cheers, Lyude Paul _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
PATCH v5 9/13 igt: fb: Rework YUV i915 allocation path
In reply to Maxime RipardWe only need to allocate a buffer using the GEM API when we're in the condition that we currently test, and that we're running on i915.
All the other cases can be handled by a fallback to a dumb buffer allocation. Let's simplify the code a bit to reflect that.
Signed-off-by Maxime Ripard <[email protected]>
--- lib/igt_fb.c | 25 ++++++++----------------- 1 file changed, 8 insertions(+), 17 deletions(-)
lib/igt_fb.c +8 -17
@@ -533,7 +533,8 @@ static int create_bo_for_fb(struct igt_fb *fb)unsigned *strides = &fb->strides[0 ];int fd = fb->fd; -if (fb->tiling || fb->size || fb->strides[0 ] ||igt_format_is_yuv (fb->drm_format)) { +if (is_i915_device (fd) && + (fb->tiling || fb->size || fb->strides[0 ] ||igt_format_is_yuv (fb->drm_format))) {uint64_t size; size =calc_fb_size (fb);
@@ -543,23 +544,12 @@ static int create_bo_for_fb(struct igt_fb *fb) fb->size = size; fb->is_dumb = false; + fb->gem_handle =gem_create (fd, fb->size); +gem_set_tiling (fd, fb->gem_handle, +igt_fb_mod_to_tiling (fb->tiling), + fb->strides[0 ]); -if (is_i915_device (fd)) { - - fb->gem_handle =gem_create (fd, fb->size); - -gem_set_tiling (fd, fb->gem_handle, -igt_fb_mod_to_tiling (fb->tiling), - fb->strides[0 ]); -clear_yuv_buffer (fd); - -return fb->gem_handle; - }else { -bool driver_has_gem_api = false; - -igt_require (driver_has_gem_api); -return -EINVAL ; - } + goto out; }/*
@@ -593,6 +583,7 @@ static int create_bo_for_fb(struct igt_fb *fb) fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb->height, bpp, strides, &fb->size); +out:if (igt_format_is_yuv (fb->drm_format))clear_yuv_buffer (fb);
Signature · 6 lines
2.20.1 _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 09/13] igt: fb: Rework YUV i915 allocation path
Hi,
Quoted 7 lines · Maxime Ripard
We only need to allocate a buffer using the GEM API when we're in the condition that we currently test, and that we're running on i915.
All the other cases can be handled by a fallback to a dumb buffer allocation. Let's simplify the code a bit to reflect that.
Signed-off-by Maxime Ripard <[email protected]>
Reviewed-by Paul Kocialkowski <[email protected]>
Cheers,
Paul
Quoted 55 lines
--- lib/igt_fb.c | 25 ++++++++----------------- 1 file changed, 8 insertions(+), 17 deletions(-)lib/igt_fb.c +8
-17@@ -533,7 +533,8 @@ static int create_bo_for_fb(struct igt_fb *fb)unsigned *strides = &fb->strides[0 ];int fd = fb->fd; -if (fb->tiling || fb->size || fb->strides[0 ] ||igt_format_is_yuv (fb->drm_format)) { +if (is_i915_device (fd) && + (fb->tiling || fb->size || fb->strides[0 ] ||igt_format_is_yuv (fb->drm_format))) {uint64_t size; size =calc_fb_size (fb);@@ -543,23 +544,12 @@ static int create_bo_for_fb(struct igt_fb *fb) fb->size = size; fb->is_dumb = false; + fb->gem_handle =gem_create (fd, fb->size); +gem_set_tiling (fd, fb->gem_handle, +igt_fb_mod_to_tiling (fb->tiling), + fb->strides[0 ]); -if (is_i915_device (fd)) { - - fb->gem_handle =gem_create (fd, fb->size); - -gem_set_tiling (fd, fb->gem_handle, -igt_fb_mod_to_tiling (fb->tiling), - fb->strides[0 ]); -clear_yuv_buffer (fd); - -return fb->gem_handle; - }else { -bool driver_has_gem_api = false; - -igt_require (driver_has_gem_api); -return -EINVAL ; - } + goto out; }/* @@ -593,6 +583,7 @@ static int create_bo_for_fb(struct igt_fb *fb) fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb->height, bpp, strides, &fb->size); +out:if (igt_format_is_yuv (fb->drm_format))clear_yuv_buffer (fb);
Signature · 8 lines
Paul Kocialkowski, Bootlin Embedded Linux and kernel engineering https://bootlin.com _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 09/13] igt: fb: Rework YUV i915 allocation path
In reply to Maxime RipardReviewed-by Lyude Paul <[email protected]>
Quoted 64 lines · Maxime Ripard
We only need to allocate a buffer using the GEM API when we're in the condition that we currently test, and that we're running on i915.
All the other cases can be handled by a fallback to a dumb buffer allocation. Let's simplify the code a bit to reflect that.
Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 25 ++++++++----------------- 1 file changed, 8 insertions(+), 17 deletions(-)lib/igt_fb.c +2
-1@@ -533,7 +533,8 @@ static int create_bo_for_fb(struct igt_fb *fb)unsigned *strides = &fb->strides[0 ];int fd = fb->fd; -if (fb->tiling || fb->size || fb->strides[0 ] ||igt_format_is_yuv (fb- >drm_format)) { +if (is_i915_device (fd) && + (fb->tiling || fb->size || fb->strides[0 ] ||igt_format_is_yuv (fb- >drm_format))) {uint64_t size;lib/igt_fb.c +6
-16size =calc_fb_size (fb);@@ -543,23 +544,12 @@ static int create_bo_for_fb(struct igt_fb *fb) fb->size = size; fb->is_dumb = false; + fb->gem_handle =gem_create (fd, fb->size); +gem_set_tiling (fd, fb->gem_handle, +igt_fb_mod_to_tiling (fb->tiling), + fb->strides[0 ]); -if (is_i915_device (fd)) { - - fb->gem_handle =gem_create (fd, fb->size); - -gem_set_tiling (fd, fb->gem_handle, -igt_fb_mod_to_tiling (fb->tiling), - fb->strides[0 ]); -clear_yuv_buffer (fd); - -return fb->gem_handle; - }else { -bool driver_has_gem_api = false; - -igt_require (driver_has_gem_api); -return -EINVAL ; - } + goto out; }/* @@ -593,6 +583,7 @@ static int create_bo_for_fb(struct igt_fb *fb) fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb->height, bpp, strides, &fb->size); +out:if (igt_format_is_yuv (fb->drm_format))clear_yuv_buffer (fb);
Signature · 7 lines
Cheers, Lyude Paul _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 09/13] igt: fb: Rework YUV i915 allocation path
In reply to Maxime RipardHi,
Quoted 7 lines · Maxime Ripard
We only need to allocate a buffer using the GEM API when we're in the condition that we currently test, and that we're running on i915.
All the other cases can be handled by a fallback to a dumb buffer allocation. Let's simplify the code a bit to reflect that.
Signed-off-by Maxime Ripard <[email protected]>
Reviewed-by Paul Kocialkowski <[email protected]>
Cheers,
Paul
Quoted 55 lines
--- lib/igt_fb.c | 25 ++++++++----------------- 1 file changed, 8 insertions(+), 17 deletions(-)lib/igt_fb.c +8
-17@@ -533,7 +533,8 @@ static int create_bo_for_fb(struct igt_fb *fb)unsigned *strides = &fb->strides[0 ];int fd = fb->fd; -if (fb->tiling || fb->size || fb->strides[0 ] ||igt_format_is_yuv (fb->drm_format)) { +if (is_i915_device (fd) && + (fb->tiling || fb->size || fb->strides[0 ] ||igt_format_is_yuv (fb->drm_format))) {uint64_t size; size =calc_fb_size (fb);@@ -543,23 +544,12 @@ static int create_bo_for_fb(struct igt_fb *fb) fb->size = size; fb->is_dumb = false; + fb->gem_handle =gem_create (fd, fb->size); +gem_set_tiling (fd, fb->gem_handle, +igt_fb_mod_to_tiling (fb->tiling), + fb->strides[0 ]); -if (is_i915_device (fd)) { - - fb->gem_handle =gem_create (fd, fb->size); - -gem_set_tiling (fd, fb->gem_handle, -igt_fb_mod_to_tiling (fb->tiling), - fb->strides[0 ]); -clear_yuv_buffer (fd); - -return fb->gem_handle; - }else { -bool driver_has_gem_api = false; - -igt_require (driver_has_gem_api); -return -EINVAL ; - } + goto out; }/* @@ -593,6 +583,7 @@ static int create_bo_for_fb(struct igt_fb *fb) fb->gem_handle =kmstest_dumb_create (fd, fb->width, fb->height, bpp, strides, &fb->size); +out:if (igt_format_is_yuv (fb->drm_format))clear_yuv_buffer (fb);
Signature · 8 lines
Paul Kocialkowski, Bootlin Embedded Linux and kernel engineering https://bootlin.com _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
PATCH v5 10/13 igt: fb: Add a bunch of new YUV formats
In reply to Maxime RipardThanks to the previous reworks, we can now add new YUV formats pretty easily. This patch adds support for the NV12, NV16, NV21, NV61, YUV420, YVU420, YUV422 and YVU422 formats.
Reviewed-by Paul Kocialkowski <[email protected]>
Signed-off-by Maxime Ripard <[email protected]>
--- lib/igt_fb.c | 100 +++++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 98 insertions(+), 2 deletions(-)
lib/igt_fb.c +98 -2
@@ -150,6 +150,21 @@ static const struct format_desc_struct { .num_planes = 2, .plane_bpp = { 8, 16, }, .hsub = 2, .vsub = 2, }, + { .name = "NV16", .depth = -1, .drm_id = DRM_FORMAT_NV16, + .cairo_id = CAIRO_FORMAT_RGB24, + .num_planes = 2, .plane_bpp = { 8, 16, }, + .hsub = 2, .vsub = 1, + }, + { .name = "NV21", .depth = -1, .drm_id = DRM_FORMAT_NV21, + .cairo_id = CAIRO_FORMAT_RGB24, + .num_planes = 2, .plane_bpp = { 8, 16, }, + .hsub = 2, .vsub = 2, + }, + { .name = "NV61", .depth = -1, .drm_id = DRM_FORMAT_NV61, + .cairo_id = CAIRO_FORMAT_RGB24, + .num_planes = 2, .plane_bpp = { 8, 16, }, + .hsub = 2, .vsub = 1, + }, { .name = "YUYV", .depth = -1, .drm_id = DRM_FORMAT_YUYV, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 16, },
@@ -170,6 +185,26 @@ static const struct format_desc_struct { .num_planes = 1, .plane_bpp = { 16, }, .hsub = 2, .vsub = 1, }, + { .name = "YU12", .depth = -1, .drm_id = DRM_FORMAT_YUV420, + .cairo_id = CAIRO_FORMAT_RGB24, + .num_planes = 3, .plane_bpp = { 8, 8, 8, }, + .hsub = 2, .vsub = 2, + }, + { .name = "YU16", .depth = -1, .drm_id = DRM_FORMAT_YUV422, + .cairo_id = CAIRO_FORMAT_RGB24, + .num_planes = 3, .plane_bpp = { 8, 8, 8, }, + .hsub = 2, .vsub = 1, + }, + { .name = "YV12", .depth = -1, .drm_id = DRM_FORMAT_YVU420, + .cairo_id = CAIRO_FORMAT_RGB24, + .num_planes = 3, .plane_bpp = { 8, 8, 8, }, + .hsub = 2, .vsub = 2, + }, + { .name = "YV16", .depth = -1, .drm_id = DRM_FORMAT_YVU422, + .cairo_id = CAIRO_FORMAT_RGB24, + .num_planes = 3, .plane_bpp = { 8, 8, 8, }, + .hsub = 2, .vsub = 1, + }, };#define for_each_format (f) \ for (f = format_desc; f - format_desc < ARRAY_SIZE(format_desc); f++)
@@ -1619,10 +1654,21 @@ static void get_yuv_parameters(struct igt_fb *fb, struct yuv_parameters *params)switch (fb->drm_format) {case DRM_FORMAT_NV12 : +case DRM_FORMAT_NV16 : +case DRM_FORMAT_NV21 : +case DRM_FORMAT_NV61 : params->y_inc =1 ; params->uv_inc =2 ;break ; +case DRM_FORMAT_YUV420 : +case DRM_FORMAT_YUV422 : +case DRM_FORMAT_YVU420 : +case DRM_FORMAT_YVU422 : + params->y_inc =1 ; + params->uv_inc =1 ; +break ; +case DRM_FORMAT_YUYV :case DRM_FORMAT_YVYU :case DRM_FORMAT_UYVY :
@@ -1639,6 +1685,13 @@ static void get_yuv_parameters(struct igt_fb *fb, struct yuv_parameters *params)switch (fb->drm_format) {case DRM_FORMAT_NV12 : +case DRM_FORMAT_NV16 : +case DRM_FORMAT_NV21 : +case DRM_FORMAT_NV61 : +case DRM_FORMAT_YUV420 : +case DRM_FORMAT_YUV422 : +case DRM_FORMAT_YVU420 : +case DRM_FORMAT_YVU422 : params->y_stride = fb->strides[0 ]; params->uv_stride = fb->strides[1 ];break ;
@@ -1655,11 +1708,33 @@ static void get_yuv_parameters(struct igt_fb *fb, struct yuv_parameters *params)switch (fb->drm_format) {case DRM_FORMAT_NV12 : +case DRM_FORMAT_NV16 : params->y_offset = fb->offsets[0 ]; params->u_offset = fb->offsets[1 ]; params->v_offset = fb->offsets[1 ] +1 ;break ; +case DRM_FORMAT_NV21 : +case DRM_FORMAT_NV61 : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[1 ] +1 ; + params->v_offset = fb->offsets[1 ]; +break ; + +case DRM_FORMAT_YUV420 : +case DRM_FORMAT_YUV422 : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[1 ]; + params->v_offset = fb->offsets[2 ]; +break ; + +case DRM_FORMAT_YVU420 : +case DRM_FORMAT_YVU422 : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[2 ]; + params->v_offset = fb->offsets[1 ]; +break ; +case DRM_FORMAT_YUYV : params->y_offset = fb->offsets[0 ]; params->u_offset = fb->offsets[0 ] +1 ;
@@ -1886,9 +1961,16 @@ static void fb_convert(struct fb_convert *cvt)switch (cvt->src.fb->drm_format) {case DRM_FORMAT_XYUV8888 :case DRM_FORMAT_NV12 : +case DRM_FORMAT_NV16 : +case DRM_FORMAT_NV21 : +case DRM_FORMAT_NV61 :case DRM_FORMAT_UYVY :case DRM_FORMAT_VYUY : +case DRM_FORMAT_YUV420 : +case DRM_FORMAT_YUV422 :case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVU420 : +case DRM_FORMAT_YVU422 :case DRM_FORMAT_YVYU :convert_yuv_to_rgb24 (cvt);return ;
@@ -1897,10 +1979,17 @@ static void fb_convert(struct fb_convert *cvt)switch (cvt->dst.fb->drm_format) {case DRM_FORMAT_XYUV8888 :case DRM_FORMAT_NV12 : -case DRM_FORMAT_YUYV : -case DRM_FORMAT_YVYU : +case DRM_FORMAT_NV16 : +case DRM_FORMAT_NV21 : +case DRM_FORMAT_NV61 :case DRM_FORMAT_UYVY :case DRM_FORMAT_VYUY : +case DRM_FORMAT_YUV420 : +case DRM_FORMAT_YUV422 : +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVU420 : +case DRM_FORMAT_YVU422 : +case DRM_FORMAT_YVYU :convert_rgb24_to_yuv (cvt);return ; }
@@ -2247,6 +2336,13 @@ bool igt_format_is_yuv(uint32_t drm_format) {switch (drm_format) {case DRM_FORMAT_NV12 : +case DRM_FORMAT_NV16 : +case DRM_FORMAT_NV21 : +case DRM_FORMAT_NV61 : +case DRM_FORMAT_YUV420 : +case DRM_FORMAT_YUV422 : +case DRM_FORMAT_YVU420 : +case DRM_FORMAT_YVU422 :case DRM_FORMAT_YUYV :case DRM_FORMAT_YVYU :case DRM_FORMAT_UYVY :
Signature · 6 lines
2.20.1 _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 10/13] igt: fb: Add a bunch of new YUV formats
Reviewed-by Lyude Paul <[email protected]>
Quoted 187 lines · Maxime Ripard
Thanks to the previous reworks, we can now add new YUV formats pretty easily. This patch adds support for the NV12, NV16, NV21, NV61, YUV420, YVU420, YUV422 and YVU422 formats.
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- lib/igt_fb.c | 100 +++++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 98 insertions(+), 2 deletions(-)lib/igt_fb.c +46
-0@@ -150,6 +150,21 @@ static const struct format_desc_struct { .num_planes = 2, .plane_bpp = { 8, 16, }, .hsub = 2, .vsub = 2, }, + { .name = "NV16", .depth = -1, .drm_id = DRM_FORMAT_NV16, + .cairo_id = CAIRO_FORMAT_RGB24, + .num_planes = 2, .plane_bpp = { 8, 16, }, + .hsub = 2, .vsub = 1, + }, + { .name = "NV21", .depth = -1, .drm_id = DRM_FORMAT_NV21, + .cairo_id = CAIRO_FORMAT_RGB24, + .num_planes = 2, .plane_bpp = { 8, 16, }, + .hsub = 2, .vsub = 2, + }, + { .name = "NV61", .depth = -1, .drm_id = DRM_FORMAT_NV61, + .cairo_id = CAIRO_FORMAT_RGB24, + .num_planes = 2, .plane_bpp = { 8, 16, }, + .hsub = 2, .vsub = 1, + }, { .name = "YUYV", .depth = -1, .drm_id = DRM_FORMAT_YUYV, .cairo_id = CAIRO_FORMAT_RGB24, .num_planes = 1, .plane_bpp = { 16, },@@ -170,6 +185,26 @@ static const struct format_desc_struct { .num_planes = 1, .plane_bpp = { 16, }, .hsub = 2, .vsub = 1, }, + { .name = "YU12", .depth = -1, .drm_id = DRM_FORMAT_YUV420, + .cairo_id = CAIRO_FORMAT_RGB24, + .num_planes = 3, .plane_bpp = { 8, 8, 8, }, + .hsub = 2, .vsub = 2, + }, + { .name = "YU16", .depth = -1, .drm_id = DRM_FORMAT_YUV422, + .cairo_id = CAIRO_FORMAT_RGB24, + .num_planes = 3, .plane_bpp = { 8, 8, 8, }, + .hsub = 2, .vsub = 1, + }, + { .name = "YV12", .depth = -1, .drm_id = DRM_FORMAT_YVU420, + .cairo_id = CAIRO_FORMAT_RGB24, + .num_planes = 3, .plane_bpp = { 8, 8, 8, }, + .hsub = 2, .vsub = 2, + }, + { .name = "YV16", .depth = -1, .drm_id = DRM_FORMAT_YVU422, + .cairo_id = CAIRO_FORMAT_RGB24, + .num_planes = 3, .plane_bpp = { 8, 8, 8, }, + .hsub = 2, .vsub = 1, + }, };#define for_each_format (f) \ for (f = format_desc; f - format_desc < ARRAY_SIZE(format_desc); f++)@@ -1619,10 +1654,21 @@ static void get_yuv_parameters(struct igt_fb *fb, struct yuv_parameters *params)switch (fb->drm_format) {case DRM_FORMAT_NV12 : +case DRM_FORMAT_NV16 : +case DRM_FORMAT_NV21 : +case DRM_FORMAT_NV61 : params->y_inc =1 ; params->uv_inc =2 ;break ; +case DRM_FORMAT_YUV420 : +case DRM_FORMAT_YUV422 : +case DRM_FORMAT_YVU420 : +case DRM_FORMAT_YVU422 : + params->y_inc =1 ; + params->uv_inc =1 ; +break ; +case DRM_FORMAT_YUYV :case DRM_FORMAT_YVYU :lib/igt_fb.c +7
-0case DRM_FORMAT_UYVY :@@ -1639,6 +1685,13 @@ static void get_yuv_parameters(struct igt_fb *fb, struct yuv_parameters *params)switch (fb->drm_format) {case DRM_FORMAT_NV12 : +case DRM_FORMAT_NV16 : +case DRM_FORMAT_NV21 : +case DRM_FORMAT_NV61 : +case DRM_FORMAT_YUV420 : +case DRM_FORMAT_YUV422 : +case DRM_FORMAT_YVU420 : +case DRM_FORMAT_YVU422 : params->y_stride = fb->strides[0 ]; params->uv_stride = fb->strides[1 ];lib/igt_fb.c +22
-0break ;@@ -1655,11 +1708,33 @@ static void get_yuv_parameters(struct igt_fb *fb, struct yuv_parameters *params)switch (fb->drm_format) {case DRM_FORMAT_NV12 : +case DRM_FORMAT_NV16 : params->y_offset = fb->offsets[0 ]; params->u_offset = fb->offsets[1 ]; params->v_offset = fb->offsets[1 ] +1 ;break ; +case DRM_FORMAT_NV21 : +case DRM_FORMAT_NV61 : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[1 ] +1 ; + params->v_offset = fb->offsets[1 ]; +break ; + +case DRM_FORMAT_YUV420 : +case DRM_FORMAT_YUV422 : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[1 ]; + params->v_offset = fb->offsets[2 ]; +break ; + +case DRM_FORMAT_YVU420 : +case DRM_FORMAT_YVU422 : + params->y_offset = fb->offsets[0 ]; + params->u_offset = fb->offsets[2 ]; + params->v_offset = fb->offsets[1 ]; +break ; +case DRM_FORMAT_YUYV : params->y_offset = fb->offsets[0 ];lib/igt_fb.c +23
-2params->u_offset = fb->offsets[0 ] +1 ;@@ -1886,9 +1961,16 @@ static void fb_convert(struct fb_convert *cvt)switch (cvt->src.fb->drm_format) {case DRM_FORMAT_XYUV8888 :case DRM_FORMAT_NV12 : +case DRM_FORMAT_NV16 : +case DRM_FORMAT_NV21 : +case DRM_FORMAT_NV61 :case DRM_FORMAT_UYVY :case DRM_FORMAT_VYUY : +case DRM_FORMAT_YUV420 : +case DRM_FORMAT_YUV422 :case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVU420 : +case DRM_FORMAT_YVU422 :case DRM_FORMAT_YVYU :convert_yuv_to_rgb24 (cvt);return ;@@ -1897,10 +1979,17 @@ static void fb_convert(struct fb_convert *cvt)switch (cvt->dst.fb->drm_format) {case DRM_FORMAT_XYUV8888 :case DRM_FORMAT_NV12 : -case DRM_FORMAT_YUYV : -case DRM_FORMAT_YVYU : +case DRM_FORMAT_NV16 : +case DRM_FORMAT_NV21 : +case DRM_FORMAT_NV61 :case DRM_FORMAT_UYVY :case DRM_FORMAT_VYUY : +case DRM_FORMAT_YUV420 : +case DRM_FORMAT_YUV422 : +case DRM_FORMAT_YUYV : +case DRM_FORMAT_YVU420 : +case DRM_FORMAT_YVU422 : +case DRM_FORMAT_YVYU :convert_rgb24_to_yuv (cvt);return ; }@@ -2247,6 +2336,13 @@ bool igt_format_is_yuv(uint32_t drm_format) {switch (drm_format) {case DRM_FORMAT_NV12 : +case DRM_FORMAT_NV16 : +case DRM_FORMAT_NV21 : +case DRM_FORMAT_NV61 : +case DRM_FORMAT_YUV420 : +case DRM_FORMAT_YUV422 : +case DRM_FORMAT_YVU420 : +case DRM_FORMAT_YVU422 :case DRM_FORMAT_YUYV :case DRM_FORMAT_YVYU :case DRM_FORMAT_UYVY :
Signature · 7 lines
Cheers, Lyude Paul _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
PATCH v5 11/13 igt: tests: chamelium: Start to unify tests
In reply to Maxime RipardThe various tests in kms_chamelium are really variants of one another but share little code.
In addition to the duplication, this gets in the way of the introduction of more tests, or to be able to run all the tests on all the output, which isn't the case at the moment, with the HDMI and DP tests and the VGA tests being different.
Start by introducing a check parameter to the do_test_display function, that will tell which test method we want to use to compare the frames.
Reviewed-by Paul Kocialkowski <[email protected]>
Signed-off-by Maxime Ripard <[email protected]>
--- tests/kms_chamelium.c | 134 +++++++++++++++++++++++------------------- 1 file changed, 73 insertions(+), 61 deletions(-)
tests/kms_chamelium.c +73 -61
@@ -527,12 +527,15 @@ static int chamelium_get_pattern_fb(data_t *data, size_t width, size_t height,return fb_id; } -static void do_test_display_crc (data_t *data,struct chamelium_port *port, -igt_output_t *output,drmModeModeInfo *mode, -uint32_t fourcc,int count) +enum chamelium_check { +CHAMELIUM_CHECK_CRC , +}; + +static void do_test_display (data_t *data,struct chamelium_port *port, +igt_output_t *output,drmModeModeInfo *mode, +uint32_t fourcc,enum chamelium_check check, +int count) { -igt_crc_t *crc; -igt_crc_t *expected_crc;struct chamelium_fb_crc_async_data *fb_crc;struct igt_fb frame_fb, fb;int i, fb_id, captured_frame_count;
@@ -545,39 +548,46 @@ static void do_test_display_crc(data_t *data, struct chamelium_port *port, frame_id =igt_fb_convert (&frame_fb, &fb, fourcc);igt_assert (frame_id >0 ); - fb_crc =chamelium_calculate_fb_crc_async_start (data->drm_fd, - &fb); +if (check ==CHAMELIUM_CHECK_CRC ) + fb_crc =chamelium_calculate_fb_crc_async_start (data->drm_fd, + &fb);enable_output (data, port, output, mode, &frame_fb); -/* We want to keep the display running for a little bit, since -* there's always the potential the driver isn't able to keep -* the display running properly for very long -*/ -chamelium_capture (data->chamelium, port,0 ,0 ,0 ,0 , count); - crc =chamelium_read_captured_crcs (data->chamelium, - &captured_frame_count); +if (check ==CHAMELIUM_CHECK_CRC ) { +igt_crc_t *expected_crc; +igt_crc_t *crc; -igt_assert (captured_frame_count == count); +/* We want to keep the display running for a little bit, since +* there's always the potential the driver isn't able to keep +* the display running properly for very long +*/ +chamelium_capture (data->chamelium, port,0 ,0 ,0 ,0 , count); + crc =chamelium_read_captured_crcs (data->chamelium, + &captured_frame_count); -igt_debug ("Captured %d frames\n" , captured_frame_count); +igt_assert (captured_frame_count == count); - expected_crc =chamelium_calculate_fb_crc_async_finish (fb_crc); +igt_debug ("Captured %d frames\n" , captured_frame_count); -for (i =0 ; i < captured_frame_count; i++) -chamelium_assert_crc_eq_or_dump (data->chamelium, - expected_crc, &crc[i], - &fb, i); + expected_crc =chamelium_calculate_fb_crc_async_finish (fb_crc); -free (expected_crc); -free (crc); +for (i =0 ; i < captured_frame_count; i++) +chamelium_assert_crc_eq_or_dump (data->chamelium, + expected_crc, &crc[i], + &fb, i); + +free (expected_crc); +free (crc); + }igt_remove_fb (data->drm_fd, &frame_fb);igt_remove_fb (data->drm_fd, &fb); } -static void test_display_crc_one_mode (data_t *data,struct chamelium_port *port, -uint32_t fourcc,int count) +static void test_display_one_mode (data_t *data,struct chamelium_port *port, +uint32_t fourcc,enum chamelium_check check, +int count) {igt_output_t *output;drmModeConnector *connector;
@@ -590,13 +600,15 @@ static void test_display_crc_one_mode(data_t *data, struct chamelium_port *port, primary =igt_output_get_plane_type (output,DRM_PLANE_TYPE_PRIMARY );igt_assert (primary); -do_test_display_crc (data, port, output, &connector->modes[0 ], fourcc, count); +do_test_display (data, port, output, &connector->modes[0 ], fourcc, + check, count);drmModeFreeConnector (connector); } -static void test_display_crc_all_modes (data_t *data,struct chamelium_port *port, -uint32_t fourcc,int count) +static void test_display_all_modes (data_t *data,struct chamelium_port *port, +uint32_t fourcc,enum chamelium_check check, +int count) {igt_output_t *output;igt_plane_t *primary;
@@ -613,7 +625,7 @@ static void test_display_crc_all_modes(data_t *data, struct chamelium_port *portfor (i =0 ; i < connector->count_modes; i++) {drmModeModeInfo *mode = &connector->modes[i]; -do_test_display_crc (data, port, output, mode, fourcc, count); +do_test_display (data, port, output, mode, fourcc, check, count); }drmModeFreeConnector (connector);
@@ -874,16 +886,16 @@ igt_main edid_id, alt_edid_id); connector_subtest("dp-crc-single", DisplayPort) -test_display_crc_all_modes (&data, port, -DRM_FORMAT_XRGB8888 ,1 ); +test_display_all_modes (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("dp-crc-fast", DisplayPort) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_XRGB8888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("dp-crc-multiple", DisplayPort) -test_display_crc_all_modes (&data, port, -DRM_FORMAT_XRGB8888 ,3 ); +test_display_all_modes (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_CRC ,3 ); connector_subtest("dp-frame-dump", DisplayPort)test_display_frame_dump (&data, port);
@@ -941,56 +953,56 @@ igt_main edid_id, alt_edid_id); connector_subtest("hdmi-crc-single", HDMIA) -test_display_crc_all_modes (&data, port, -DRM_FORMAT_XRGB8888 ,1 ); +test_display_all_modes (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-fast", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_XRGB8888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-multiple", HDMIA) -test_display_crc_all_modes (&data, port, -DRM_FORMAT_XRGB8888 ,3 ); +test_display_all_modes (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_CRC ,3 ); connector_subtest("hdmi-crc-argb8888", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_ARGB8888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_ARGB8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-abgr8888", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_ABGR8888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_ABGR8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-xrgb8888", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_XRGB8888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-xbgr8888", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_XBGR8888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_XBGR8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-rgb888", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_RGB888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_RGB888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-bgr888", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_BGR888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_BGR888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-rgb565", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_RGB565 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_RGB565 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-bgr565", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_BGR565 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_BGR565 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-argb1555", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_ARGB1555 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_ARGB1555 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-xrgb1555", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_XRGB1555 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_XRGB1555 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-frame-dump", HDMIA)test_display_frame_dump (&data, port);
Signature · 6 lines
2.20.1 _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 11/13] igt: tests: chamelium: Start to unify tests
Reviewed-by Lyude Paul <[email protected]>
Quoted 268 lines · Maxime Ripard
The various tests in
kms_chameliumare really variants of one another but share little code.In addition to the duplication, this gets in the way of the introduction of more tests, or to be able to run all the tests on all the output, which isn't the case at the moment, with the HDMI and DP tests and the VGA tests being different.
Start by introducing a check parameter to the
do_test_displayfunction, that will tell which test method we want to use to compare the frames.Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- tests/kms_chamelium.c | 134 +++++++++++++++++++++++------------------- 1 file changed, 73 insertions(+), 61 deletions(-)tests/kms_chamelium.c +8
-5@@ -527,12 +527,15 @@ static int chamelium_get_pattern_fb(data_t *data, size_t width, size_t height, return fb_id; } -static void do_test_display_crc (data_t *data,struct chamelium_port *port, -igt_output_t *output,drmModeModeInfo *mode, -uint32_t fourcc,int count) +enum chamelium_check { +CHAMELIUM_CHECK_CRC , +}; + +static void do_test_display (data_t *data,struct chamelium_port *port, +igt_output_t *output,drmModeModeInfo *mode, +uint32_t fourcc,enum chamelium_check check, +int count) { -igt_crc_t *crc; -igt_crc_t *expected_crc;struct chamelium_fb_crc_async_data *fb_crc;struct igt_fb frame_fb, fb;tests/kms_chamelium.c +25
-20int i, fb_id, captured_frame_count;@@ -545,39 +548,46 @@ static void do_test_display_crc(data_t *data, structchamelium_port *port, frame_id =igt_fb_convert (&frame_fb, &fb, fourcc);igt_assert (frame_id >0 ); - fb_crc =chamelium_calculate_fb_crc_async_start (data->drm_fd, - &fb); +if (check ==CHAMELIUM_CHECK_CRC ) + fb_crc =chamelium_calculate_fb_crc_async_start (data->drm_fd, + &fb);enable_output (data, port, output, mode, &frame_fb); -/* We want to keep the display running for a little bit, since -* there's always the potential the driver isn't able to keep -* the display running properly for very long -*/ -chamelium_capture (data->chamelium, port,0 ,0 ,0 ,0 , count); - crc =chamelium_read_captured_crcs (data->chamelium, - &captured_frame_count); +if (check ==CHAMELIUM_CHECK_CRC ) { +igt_crc_t *expected_crc; +igt_crc_t *crc; -igt_assert (captured_frame_count == count); +/* We want to keep the display running for a little bit, since +* there's always the potential the driver isn't able to keep +* the display running properly for very long +*/ +chamelium_capture (data->chamelium, port,0 ,0 ,0 ,0 , count); + crc =chamelium_read_captured_crcs (data->chamelium, + &captured_frame_count); -igt_debug ("Captured %d frames\n" , captured_frame_count); +igt_assert (captured_frame_count == count); - expected_crc =chamelium_calculate_fb_crc_async_finish (fb_crc); +igt_debug ("Captured %d frames\n" , captured_frame_count); -for (i =0 ; i < captured_frame_count; i++) -chamelium_assert_crc_eq_or_dump (data->chamelium, - expected_crc, &crc[i], - &fb, i); + expected_crc =chamelium_calculate_fb_crc_async_finish (fb_crc); -free (expected_crc); -free (crc); +for (i =0 ; i < captured_frame_count; i++) +chamelium_assert_crc_eq_or_dump (data->chamelium, + expected_crc, &crc[i], + &fb, i); + +free (expected_crc); +free (crc); + }igt_remove_fb (data->drm_fd, &frame_fb);igt_remove_fb (data->drm_fd, &fb); } -static void test_display_crc_one_mode (data_t *data,struct chamelium_port *port, -uint32_t fourcc,int count) +static void test_display_one_mode(data_t *data, struct chamelium_port*port,
tests/kms_chamelium.c +5
-3+ uint32_t fourcc, enum chamelium_check check, + int count) { igt_output_t *output;drmModeConnector *connector;@@ -590,13 +600,15 @@ static void test_display_crc_one_mode(data_t *data,struct chamelium_port *port, primary =igt_output_get_plane_type (output,DRM_PLANE_TYPE_PRIMARY );igt_assert (primary); -do_test_display_crc (data, port, output, &connector->modes[0 ], fourcc, count); +do_test_display (data, port, output, &connector->modes[0 ], fourcc, + check, count);drmModeFreeConnector (connector); } -static void test_display_crc_all_modes (data_t *data,struct chamelium_port *port, -uint32_t fourcc,int count) +static void test_display_all_modes(data_t *data, struct chamelium_port*port,
tests/kms_chamelium.c +1
-0+ uint32_t fourcc, enum chamelium_checkcheck,
tests/kms_chamelium.c +1
-0+ int count) {igt_output_t *output;igt_plane_t *primary;@@ -613,7 +625,7 @@ static void test_display_crc_all_modes(data_t *data,struct
chamelium_port*porttests/kms_chamelium.c +1
-1for (i =0 ; i < connector->count_modes; i++) {drmModeModeInfo *mode = &connector->modes[i]; -do_test_display_crc (data, port, output, mode, fourcc, count); + do_test_display(data, port, output, mode, fourcc, check,count);
tests/kms_chamelium.c +6
-6}drmModeFreeConnector (connector);@@ -874,16 +886,16 @@ igt_main edid_id, alt_edid_id); connector_subtest("dp-crc-single", DisplayPort) -test_display_crc_all_modes (&data, port, -DRM_FORMAT_XRGB8888 ,1 ); +test_display_all_modes (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("dp-crc-fast", DisplayPort) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_XRGB8888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("dp-crc-multiple", DisplayPort) -test_display_crc_all_modes (&data, port, -DRM_FORMAT_XRGB8888 ,3 ); +test_display_all_modes (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_CRC ,3 );tests/kms_chamelium.c +23
-24connector_subtest("dp-frame-dump", DisplayPort)test_display_frame_dump (&data, port);@@ -941,56 +953,56 @@ igt_main edid_id, alt_edid_id); connector_subtest("hdmi-crc-single", HDMIA) -test_display_crc_all_modes (&data, port, -DRM_FORMAT_XRGB8888 ,1 ); +test_display_all_modes (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-fast", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_XRGB8888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-multiple", HDMIA) -test_display_crc_all_modes (&data, port, -DRM_FORMAT_XRGB8888 ,3 ); +test_display_all_modes (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_CRC ,3 ); connector_subtest("hdmi-crc-argb8888", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_ARGB8888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_ARGB8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-abgr8888", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_ABGR8888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_ABGR8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-xrgb8888", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_XRGB8888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-xbgr8888", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_XBGR8888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_XBGR8888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-rgb888", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_RGB888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_RGB888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-bgr888", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_BGR888 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_BGR888 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-rgb565", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_RGB565 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_RGB565 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-bgr565", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_BGR565 ,1 ); +test_display_one_mode (&data, port,DRM_FORMAT_BGR565 , +CHAMELIUM_CHECK_CRC ,1 ); connector_subtest("hdmi-crc-argb1555", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_ARGB1555 ,1 ); + test_display_one_mode(&data, port,
DRM_FORMAT_ARGB1555,tests/kms_chamelium.c +2
-2+ CHAMELIUM_CHECK_CRC, 1); connector_subtest("hdmi-crc-xrgb1555", HDMIA) -test_display_crc_one_mode (&data, port, -DRM_FORMAT_XRGB1555 ,1 ); + test_display_one_mode(&data, port,
DRM_FORMAT_XRGB1555,tests/kms_chamelium.c +1
-0+ CHAMELIUM_CHECK_CRC, 1); connector_subtest("hdmi-frame-dump", HDMIA)test_display_frame_dump (&data, port);
Signature · 7 lines
Cheers, Lyude Paul _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
PATCH v5 12/13 igt: tests: chamelium: Convert VGA tests to do_test_display
In reply to Maxime RipardThe VGA tests were run so far as part of a separate function. However, the only difference between that function and the do_test_display function was the kind of comparison we wanted to use. Indeed, VGA being an analog output, we can't rely on a pixel perfect image, and thus only compare the CRCs, but we need to make some more advanced comparison.
Let's add a new check method, add support for the analog check to do_test_display and remove the old version.
Reviewed-by Paul Kocialkowski <[email protected]>
Signed-off-by Maxime Ripard <[email protected]>
--- tests/kms_chamelium.c | 92 ++++++++++++++++--------------------------- 1 file changed, 33 insertions(+), 59 deletions(-)
tests/kms_chamelium.c +33 -59
@@ -528,6 +528,7 @@ static int chamelium_get_pattern_fb(data_t *data, size_t width, size_t height, }enum chamelium_check { +CHAMELIUM_CHECK_ANALOG ,CHAMELIUM_CHECK_CRC , };
@@ -579,6 +580,18 @@ static void do_test_display(data_t *data, struct chamelium_port *port,free (expected_crc);free (crc); + }else if (check ==CHAMELIUM_CHECK_ANALOG ) { +struct chamelium_frame_dump *dump; + +igt_assert (count ==1 ); + + dump =chamelium_port_dump_pixels (data->chamelium, port,0 ,0 , +0 ,0 ); +chamelium_crop_analog_frame (dump, mode->hdisplay, + mode->vdisplay); +chamelium_assert_analog_frame_match_or_dump (data->chamelium, + port, dump, &fb); +chamelium_destroy_frame_dump (dump); }igt_remove_fb (data->drm_fd, &frame_fb);
@@ -589,8 +602,9 @@ static void test_display_one_mode(data_t *data, struct chamelium_port *port, uint32_t fourcc, enum chamelium_check check, int count) { - igt_output_t *output; drmModeConnector *connector; +drmModeModeInfo *mode; +igt_output_t *output;igt_plane_t *primary;reset_state (data, port);
@@ -600,8 +614,14 @@ static void test_display_one_mode(data_t *data, struct chamelium_port *port, primary =igt_output_get_plane_type (output,DRM_PLANE_TYPE_PRIMARY );igt_assert (primary); -do_test_display (data, port, output, &connector->modes[0 ], fourcc, - check, count); + mode = &connector->modes[0 ]; +if (check ==CHAMELIUM_CHECK_ANALOG ) { +bool bridge =check_analog_bridge (data, port); + +igt_assert (!(bridge &&prune_vga_mode (data, mode))); + } + +do_test_display (data, port, output, mode, fourcc, check, count);drmModeFreeConnector (connector); }
@@ -613,6 +633,7 @@ static void test_display_all_modes(data_t *data, struct chamelium_port *port,igt_output_t *output;igt_plane_t *primary;drmModeConnector *connector; +bool bridge;int i;reset_state (data, port);
@@ -622,9 +643,16 @@ static void test_display_all_modes(data_t *data, struct chamelium_port *port, primary =igt_output_get_plane_type (output,DRM_PLANE_TYPE_PRIMARY );igt_assert (primary); +if (check ==CHAMELIUM_CHECK_ANALOG ) + bridge =check_analog_bridge (data, port); +for (i =0 ; i < connector->count_modes; i++) {drmModeModeInfo *mode = &connector->modes[i]; +if (check ==CHAMELIUM_CHECK_ANALOG && bridge && +prune_vga_mode (data, mode)) +continue ; +do_test_display (data, port, output, mode, fourcc, check, count); }
@@ -675,61 +703,6 @@ test_display_frame_dump(data_t *data, struct chamelium_port *port)drmModeFreeConnector (connector); } -static void -test_analog_frame_dump (data_t *data,struct chamelium_port *port) -{ -igt_output_t *output; -igt_plane_t *primary; -struct igt_fb fb; -struct chamelium_frame_dump *frame; -drmModeModeInfo *mode; -drmModeConnector *connector; -int fb_id, i; -bool bridge; - -reset_state (data, port); - - output =prepare_output (data, port); - connector =chamelium_port_get_connector (data->chamelium, port, false); - primary =igt_output_get_plane_type (output,DRM_PLANE_TYPE_PRIMARY ); -igt_assert (primary); - - bridge =check_analog_bridge (data, port); - -for (i =0 ; i < connector->count_modes; i++) { - mode = &connector->modes[i]; - -if (bridge &&prune_vga_mode (data, mode)) -continue ; - - fb_id =igt_create_color_pattern_fb (data->drm_fd, - mode->hdisplay, mode->vdisplay, -DRM_FORMAT_XRGB8888 , -LOCAL_DRM_FORMAT_MOD_NONE , -0 ,0 ,0 , &fb); -igt_assert (fb_id >0 ); - -enable_output (data, port, output, mode, &fb); - -igt_debug ("Reading frame dumps from Chamelium...\n" ); - - frame =chamelium_port_dump_pixels (data->chamelium, port,0 ,0 , -0 ,0 ); - -chamelium_crop_analog_frame (frame, mode->hdisplay, - mode->vdisplay); - -chamelium_assert_analog_frame_match_or_dump (data->chamelium, - port, frame, &fb); - -chamelium_destroy_frame_dump (frame); - -igt_remove_fb (data->drm_fd, &fb); - } - -drmModeFreeConnector (connector); -} -static void test_hpd_without_ddc (data_t *data,struct chamelium_port *port) {
@@ -1041,7 +1014,8 @@ igt_maintest_hpd_without_ddc (&data, port); connector_subtest("vga-frame-dump", VGA) -test_analog_frame_dump (&data, port); +test_display_all_modes (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_ANALOG ,1 ); } igt_subtest_group {
Signature · 6 lines
2.20.1 _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 12/13] igt: tests: chamelium: Convert VGA tests to do_test_display
Good thinking!
Reviewed-by Lyude Paul <[email protected]>
Quoted 184 lines · Maxime Ripard
The VGA tests were run so far as part of a separate function. However, the only difference between that function and the
do_test_displayfunction was the kind of comparison we wanted to use. Indeed, VGA being an analog output, we can't rely on a pixel perfect image, and thus only compare the CRCs, but we need to make some more advanced comparison.Let's add a new check method, add support for the analog check to
do_test_displayand remove the old version.Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- tests/kms_chamelium.c | 92 ++++++++++++++++--------------------------- 1 file changed, 33 insertions(+), 59 deletions(-)tests/kms_chamelium.c +13
-0@@ -528,6 +528,7 @@ static int chamelium_get_pattern_fb(data_t *data, size_t width, size_t height, }enum chamelium_check { +CHAMELIUM_CHECK_ANALOG ,CHAMELIUM_CHECK_CRC , };@@ -579,6 +580,18 @@ static void do_test_display(data_t *data, structchamelium_port *port,free (expected_crc );free (crc); + }else if (check ==CHAMELIUM_CHECK_ANALOG ) { +struct chamelium_frame_dump *dump; + +igt_assert (count ==1 ); + + dump =chamelium_port_dump_pixels (data->chamelium, port,0 ,0 , +0 ,0 ); +chamelium_crop_analog_frame (dump, mode->hdisplay, + mode->vdisplay); +chamelium_assert_analog_frame_match_or_dump (data->chamelium, + port, dump, &fb); +chamelium_destroy_frame_dump (dump); }tests/kms_chamelium.c +2
-1igt_remove_fb (data->drm_fd, &frame_fb);@@ -589,8 +602,9 @@ static void test_display_one_mode(data_t *data, structchamelium_port *port, uint32_t fourcc, enum chamelium_check check, int count) { - igt_output_t *output; drmModeConnector *connector; +drmModeModeInfo *mode; +igt_output_t *output;igt_plane_t *primary;tests/kms_chamelium.c +8
-2reset_state (data, port);@@ -600,8 +614,14 @@ static void test_display_one_mode(data_t *data, structchamelium_port *port, primary =igt_output_get_plane_type (output,DRM_PLANE_TYPE_PRIMARY );igt_assert (primary); -do_test_display (data, port, output, &connector->modes[0 ], fourcc, - check, count); + mode = &connector->modes[0 ]; +if (check ==CHAMELIUM_CHECK_ANALOG ) { +bool bridge =check_analog_bridge (data, port); + +igt_assert (!(bridge &&prune_vga_mode (data, mode))); + } + +do_test_display (data, port, output, mode, fourcc, check, count);drmModeFreeConnector (connector);}
tests/kms_chamelium.c +1
-0@@ -613,6 +633,7 @@ static void test_display_all_modes(data_t *data, struct chamelium_port *port, igt_output_t *output;igt_plane_t *primary;drmModeConnector *connector; +bool bridge;int i;tests/kms_chamelium.c +7
-0reset_state (data, port);@@ -622,9 +643,16 @@ static void test_display_all_modes(data_t *data, structchamelium_port *port, primary =igt_output_get_plane_type (output,DRM_PLANE_TYPE_PRIMARY );igt_assert (primary); +if (check ==CHAMELIUM_CHECK_ANALOG ) + bridge =check_analog_bridge (data, port); +for (i =0 ; i < connector->count_modes; i++) {drmModeModeInfo *mode = &connector->modes[i]; +if (check ==CHAMELIUM_CHECK_ANALOG && bridge && +prune_vga_mode (data, mode)) +continue ; +do_test_display (data, port, output, mode, fourcc, check, count);tests/kms_chamelium.c +0
-39}@@ -675,61 +703,6 @@ test_display_frame_dump(data_t *data, struct chamelium_port *port)drmModeFreeConnector (connector ); } -static void -test_analog_frame_dump (data_t *data,struct chamelium_port *port) -{ -igt_output_t *output; -igt_plane_t *primary; -struct igt_fb fb; -struct chamelium_frame_dump *frame; -drmModeModeInfo *mode; -drmModeConnector *connector; -int fb_id, i; -bool bridge; - -reset_state (data, port); - - output =prepare_output (data, port); - connector =chamelium_port_get_connector (data->chamelium, port, false); - primary =igt_output_get_plane_type (output,DRM_PLANE_TYPE_PRIMARY ); -igt_assert (primary); - - bridge =check_analog_bridge (data, port); - -for (i =0 ; i < connector->count_modes; i++) { - mode = &connector->modes[i]; - -if (bridge &&prune_vga_mode (data, mode)) -continue ; - - fb_id =igt_create_color_pattern_fb (data->drm_fd, - mode->hdisplay, mode- >vdisplay, -DRM_FORMAT_XRGB8888 , -LOCAL_DRM_FORMAT_MOD_NONE , -0 ,0 ,0 , &fb); -igt_assert (fb_id >0 ); - -enable_output (data, port, output, mode, &fb); - -igt_debug ("Reading frame dumps from Chamelium...\n" ); - - frame = chamelium_port_dump_pixels(data->chamelium, port, 0,0,
tests/kms_chamelium.c +2
-17- 0, 0); - -chamelium_crop_analog_frame (frame, mode->hdisplay, - mode->vdisplay); - -chamelium_assert_analog_frame_match_or_dump (data->chamelium, - port, frame, &fb); - -chamelium_destroy_frame_dump (frame); - -igt_remove_fb (data->drm_fd, &fb); - } - -drmModeFreeConnector (connector); -} -static void test_hpd_without_ddc (data_t *data,struct chamelium_port *port) {@@ -1041,7 +1014,8 @@ igt_maintest_hpd_without_ddc (&data, port); connector_subtest("vga-frame-dump", VGA) -test_analog_frame_dump (&data, port); +test_display_all_modes (&data, port,DRM_FORMAT_XRGB8888 , +CHAMELIUM_CHECK_ANALOG ,1 ); }
igt_subtest_group{
Signature · 7 lines
Cheers, Lyude Paul _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
PATCH v5 13/13 igt: tests: chamelium: Add YUV formats tests
In reply to Maxime RipardThe NV12, NV16, NV21, NV61, YUV420, YVU420, YUV422 and YVU422 are YUV formats that are currently supported in IGT.
We'll want to test those formats in addition to the RGB formats, so let's add some subtests. One thing worth noting is some hardware isn't able to output a pixel-perfect image, so we do the same kind of comparison than for VGA.
Reviewed-by Paul Kocialkowski <[email protected]>
Signed-off-by Maxime Ripard <[email protected]>
--- tests/kms_chamelium.c | 32 ++++++++++++++++++++++++++++++++ 1 file changed, 32 insertions(+)
tests/kms_chamelium.c +32 -0
@@ -977,6 +977,38 @@ igt_maintest_display_one_mode (&data, port,DRM_FORMAT_XRGB1555 ,CHAMELIUM_CHECK_CRC ,1 ); + connector_subtest("hdmi-cmp-nv12", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_NV12 , +CHAMELIUM_CHECK_ANALOG ,1 ); + + connector_subtest("hdmi-cmp-nv16", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_NV16 , +CHAMELIUM_CHECK_ANALOG ,1 ); + + connector_subtest("hdmi-cmp-nv21", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_NV21 , +CHAMELIUM_CHECK_ANALOG ,1 ); + + connector_subtest("hdmi-cmp-nv61", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_NV61 , +CHAMELIUM_CHECK_ANALOG ,1 ); + + connector_subtest("hdmi-cmp-yu12", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_YUV420 , +CHAMELIUM_CHECK_ANALOG ,1 ); + + connector_subtest("hdmi-cmp-yu16", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_YUV422 , +CHAMELIUM_CHECK_ANALOG ,1 ); + + connector_subtest("hdmi-cmp-yv12", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_YVU420 , +CHAMELIUM_CHECK_ANALOG ,1 ); + + connector_subtest("hdmi-cmp-yv16", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_YVU422 , +CHAMELIUM_CHECK_ANALOG ,1 ); + connector_subtest("hdmi-frame-dump", HDMIA)test_display_frame_dump (&data, port); }
Signature · 6 lines
2.20.1 _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Re: [igt-dev] [PATCH i-g-t v5 13/13] igt: tests: chamelium: Add YUV formats tests
Reviewed-by Lyude Paul <[email protected]>
Quoted 58 lines · Maxime Ripard
The NV12, NV16, NV21, NV61, YUV420, YVU420, YUV422 and YVU422 are YUV formats that are currently supported in IGT.
We'll want to test those formats in addition to the RGB formats, so let's add some subtests. One thing worth noting is some hardware isn't able to output a pixel-perfect image, so we do the same kind of comparison than for VGA.
Reviewed-by Paul Kocialkowski <[email protected]>Signed-off-by Maxime Ripard <[email protected]>--- tests/kms_chamelium.c | 32 ++++++++++++++++++++++++++++++++ 1 file changed, 32 insertions(+)tests/kms_chamelium.c +32
-0@@ -977,6 +977,38 @@ igt_maintest_display_one_mode (&data, port,DRM_FORMAT_XRGB1555 ,CHAMELIUM_CHECK_CRC ,1 ); + connector_subtest("hdmi-cmp-nv12", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_NV12 , +CHAMELIUM_CHECK_ANALOG ,1 ); + + connector_subtest("hdmi-cmp-nv16", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_NV16 , +CHAMELIUM_CHECK_ANALOG ,1 ); + + connector_subtest("hdmi-cmp-nv21", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_NV21 , +CHAMELIUM_CHECK_ANALOG ,1 ); + + connector_subtest("hdmi-cmp-nv61", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_NV61 , +CHAMELIUM_CHECK_ANALOG ,1 ); + + connector_subtest("hdmi-cmp-yu12", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_YUV420 , +CHAMELIUM_CHECK_ANALOG ,1 ); + + connector_subtest("hdmi-cmp-yu16", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_YUV422 , +CHAMELIUM_CHECK_ANALOG ,1 ); + + connector_subtest("hdmi-cmp-yv12", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_YVU420 , +CHAMELIUM_CHECK_ANALOG ,1 ); + + connector_subtest("hdmi-cmp-yv16", HDMIA) +test_display_one_mode (&data, port,DRM_FORMAT_YVU422 , +CHAMELIUM_CHECK_ANALOG ,1 ); + connector_subtest("hdmi-frame-dump", HDMIA)test_display_frame_dump (&data, port);}
Signature · 7 lines
Cheers, Lyude Paul _______________________________________________ igt-dev mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/igt-dev
Patchwork
CI[igt-dev] ✗ Fi.CI.BAT: failure for igt: chamelium: Test YUV buffers using the Chamelium (rev7)
In reply to Maxime Ripard== Series Details ==
== Series Details ==
Series: igt: chamelium: Test YUV buffers using the Chamelium (rev7)
URL : https://patchwork.freedesktop.org/series/53468/
State : failure
== Summary ==
CI Bug Log - changes from CI_DRM_5484 -> IGTPW_2294
====================================================
Summary
-------
**FAILURE**
Serious unknown changes coming with IGTPW_2294 absolutely need to be verified manually.
If you think the reported changes have nothing to do with the changes introduced in IGTPW_2294, please notify your bug team to allow them to document this new failure mode, which will reduce false positives in CI.
External URL: https://patchwork.freedesktop.org/api/1.0/series/53468/revisions/7/mbox/
Possible new issues
-------------------
Here are the unknown changes that may have been introduced in IGTPW_2294:
### IGT changes ###
#### Possible regressions ####
* igt@kms_addfb_basic@addfb25-y-tiled-small:
- fi-kbl-r: PASS -> FAIL +37
* igt@kms_addfb_basic@bo-too-small:
- fi-cfl-guc: PASS -> FAIL +31
- fi-skl-gvtdvm: PASS -> FAIL +33
- fi-kbl-7567u: PASS -> FAIL +3* igt@kms_addfb_basic@bo-too-small-due-to-tiling:
- fi-apl-guc: PASS -> FAIL +31
- fi-kbl-guc: PASS -> FAIL +3
- fi-kbl-8809g: PASS -> FAIL +3 * igt@kms_addfb_basic@too-high:
- fi-glk-j4005: PASS -> FAIL +36
- fi-gdg-551: PASS -> FAIL +24
* igt@kms_cursor_legacy@basic-busy-flip-before-cursor-legacy:
- fi-hsw-4770r: PASS -> FAIL +35
- fi-cfl-8700k: PASS -> FAIL +36
* igt@kms_cursor_legacy@basic-flip-after-cursor-varying-size:
- fi-skl-guc: PASS -> FAIL +32
- fi-whl-u: PASS -> FAIL +37
* igt@kms_cursor_legacy@basic-flip-before-cursor-atomic:
- fi-snb-2520m: PASS -> FAIL +28
* igt@kms_cursor_legacy@basic-flip-before-cursor-varying-size:
- fi-bsw-n3050: NOTRUN -> FAIL +23
- fi-skl-6700k2: PASS -> FAIL +37
- fi-hsw-4770: PASS -> FAIL +36* igt@kms_flip@basic-flip-vs-dpms:
- fi-ilk-650: PASS -> FAIL +28
- fi-pnv-d510: PASS -> FAIL +24
- fi-skl-6770hq: PASS -> FAIL +36 * igt@kms_flip@basic-flip-vs-modeset:
- fi-kbl-x1275: PASS -> FAIL +36
* igt@kms_flip@basic-flip-vs-wf_vblank:
- fi-hsw-peppy: PASS -> FAIL +36
- fi-skl-6260u: PASS -> FAIL +36
* igt@kms_pipe_crc_basic@hang-read-crc-pipe-b:
- fi-bxt-j4205: PASS -> FAIL +36
- fi-skl-6700hq: PASS -> FAIL +37
* igt@kms_pipe_crc_basic@nonblocking-crc-pipe-a:
- fi-ivb-3520m: PASS -> FAIL +34
* igt@kms_pipe_crc_basic@nonblocking-crc-pipe-a-frame-sequence:
- fi-elk-e7500: PASS -> FAIL +24
- fi-byt-n2820: PASS -> FAIL +28
* igt@kms_pipe_crc_basic@nonblocking-crc-pipe-b:
- fi-bdw-5557u: PASS -> FAIL +35
* igt@kms_pipe_crc_basic@nonblocking-crc-pipe-b-frame-sequence:
- fi-bdw-gvtdvm: PASS -> FAIL +32
* igt@kms_pipe_crc_basic@read-crc-pipe-a:
- fi-bwr-2160: PASS -> FAIL +24
* igt@kms_pipe_crc_basic@read-crc-pipe-b:
- fi-byt-clapper: PASS -> FAIL +29
- fi-cfl-8109u: PASS -> FAIL +37
* igt@pm_rpm@basic-pci-d3-state:
- fi-kbl-7560u: PASS -> FAIL +37
- fi-byt-j1900: PASS -> FAIL +29
* igt@pm_rpm@basic-rte:
- fi-skl-iommu: PASS -> FAIL +36
#### Warnings ####
* igt@pm_rpm@basic-pci-d3-state:
- fi-pnv-d510: {SKIP} [fdo#109271] -> FAIL +2
- fi-gdg-551: {SKIP} [fdo#109271] -> FAIL +2
* igt@pm_rpm@basic-rte:
- fi-skl-gvtdvm: {SKIP} [fdo#109271] -> FAIL +2
- fi-elk-e7500: {SKIP} [fdo#109271] -> FAIL +2
- fi-ilk-650: {SKIP} [fdo#109271] -> FAIL +2
- fi-bdw-gvtdvm: {SKIP} [fdo#109271] -> FAIL +2
- fi-snb-2520m: {SKIP} [fdo#109271] -> FAIL +2
- fi-bwr-2160: {SKIP} [fdo#109271] -> FAIL +2 * igt@pm_rpm@module-reload:
- fi-ivb-3520m: {SKIP} [fdo#109271] -> FAIL +2
#### Suppressed ####
The following results come from untrusted machines, tests, or statuses. They do not affect the overall result.
* igt@kms_cursor_legacy@basic-flip-after-cursor-atomic:
- {fi-icl-u3}: PASS -> FAIL +38
* igt@kms_pipe_crc_basic@hang-read-crc-pipe-a:
- {fi-icl-u2}: PASS -> FAIL +38
Known issues
------------
Here are the changes found in IGTPW_2294 that come from known issues:
### IGT changes ###
#### Issues hit ####
* igt@gem_exec_suspend@basic-s3:
- fi-blb-e6850: PASS -> INCOMPLETE [fdo#107718]
* igt@i915_module_load@reload:
- fi-hsw-peppy: PASS -> DMESG-WARN [fdo#102614]
* igt@kms_chamelium@hdmi-crc-fast:
- fi-skl-6700k2: PASS -> FAIL [fdo#108767]
- fi-kbl-7567u: PASS -> FAIL [fdo#108767]
* igt@pm_rpm@module-reload:
- fi-byt-clapper: PASS -> FAIL [fdo#108675]
- fi-byt-j1900: PASS -> FAIL [fdo#108675]
- fi-byt-n2820: PASS -> FAIL [fdo#108675]#### Possible fixes ####
* igt@kms_busy@basic-flip-b:
- fi-gdg-551: FAIL [fdo#103182] -> PASS
{name}: This element is suppressed. This means it is ignored when computing
the status of the difference (SUCCESS, WARNING, or FAILURE).
[fdo#102614]: https://bugs.freedesktop.org/show_bug.cgi?id=102614
[fdo#103182]: https://bugs.freedesktop.org/show_bug.cgi?id=103182
[fdo#107718]: https://bugs.freedesktop.org/show_bug.cgi?id=107718
[fdo#108675]: https://bugs.freedesktop.org/show_bug.cgi?id=108675
[fdo#108767]: https://bugs.freedesktop.org/show_bug.cgi?id=108767
[fdo#109271]: https://bugs.freedesktop.org/show_bug.cgi?id=109271
[fdo#109278]: https://bugs.freedesktop.org/show_bug.cgi?id=109278
Participating hosts (44 -> 40)
------------------------------
Additional (1): fi-bsw-n3050
Missing (5): fi-ilk-m540 fi-byt-squawks fi-bsw-cyan fi-kbl-7500u fi-ctg-p8600
Build changes
-------------
* IGT: IGT_4790 -> IGTPW_2294
CI_DRM_5484: 9f66ac94341eb12501097f9f8991c86aee70981c @ git://anongit.freedesktop.org/gfx-ci/linux
IGTPW_2294: https://intel-gfx-ci.01.org/tree/drm-tip/IGTPW_2294/
IGT_4790: dcdf4b04e16312f8f52ad389388d834f9d74b8f0 @ git://anongit.freedesktop.org/xorg/app/intel-gpu-tools
== Testlist changes ==
+igt@kms_chamelium@hdmi-cmp-nv12
+igt@kms_chamelium@hdmi-cmp-nv16
+igt@kms_chamelium@hdmi-cmp-nv21
+igt@kms_chamelium@hdmi-cmp-nv61
+igt@kms_chamelium@hdmi-cmp-yu12
+igt@kms_chamelium@hdmi-cmp-yu16
+igt@kms_chamelium@hdmi-cmp-yv12
+igt@kms_chamelium@hdmi-cmp-yv16
== Logs ==
For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/IGTPW_2294/
_______________________________________________
igt-dev mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/igt-dev