[igt-dev] [PATCH i-g-t 4/7] lib/igt_fb: Use rendercopy for rendering into compressed buffers
Clinton Taylor
Clinton.A.Taylor at intel.com
Wed Mar 6 22:42:54 UTC 2019
Reviewed-by: Clint Taylor <Clinton.A.Taylor at intel.com>
-Clint
On 3/4/19 5:47 PM, Dhinakaran Pandiyan wrote:
> From: Ville Syrjälä <ville.syrjala at linux.intel.com>
>
> Plug in rendercopy to the cairo surface stuff so that we can
> generate compressed buffers with cairo.
>
> v2: s/modifier_has_ccs/is_ccs_modifier/
>
> v3: From DK
> Set buf.bpp when initializing and rebase.
> v4: From DK
> Fix Yf-tiled BGR formats
>
> Cc: Clinton Taylor <clinton.a.taylor at intel.com>
> Signed-off-by: Ville Syrjälä <ville.syrjala at linux.intel.com>
> Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandiyan at intel.com>
> ---
> lib/igt_fb.c | 155 +++++++++++++++++++++++++++++++++++++++++++++++----
> 1 file changed, 144 insertions(+), 11 deletions(-)
>
> diff --git a/lib/igt_fb.c b/lib/igt_fb.c
> index 9dca2a46..b480a0a1 100644
> --- a/lib/igt_fb.c
> +++ b/lib/igt_fb.c
> @@ -299,6 +299,7 @@ void igt_get_fb_tile_size(int fd, uint64_t tiling, int fb_bpp,
> }
> break;
> case LOCAL_I915_FORMAT_MOD_Y_TILED:
> + case LOCAL_I915_FORMAT_MOD_Y_TILED_CCS:
> igt_require_intel(fd);
> if (intel_gen(intel_get_drm_devid(fd)) == 2) {
> *width_ret = 128;
> @@ -312,6 +313,7 @@ void igt_get_fb_tile_size(int fd, uint64_t tiling, int fb_bpp,
> }
> break;
> case LOCAL_I915_FORMAT_MOD_Yf_TILED:
> + case LOCAL_I915_FORMAT_MOD_Yf_TILED_CCS:
> igt_require_intel(fd);
> switch (fb_bpp) {
> case 8:
> @@ -362,10 +364,19 @@ void igt_get_fb_tile_size(int fd, uint64_t tiling, int fb_bpp,
> }
> }
>
> +static bool is_ccs_modifier(uint64_t modifier)
> +{
> + return modifier == LOCAL_I915_FORMAT_MOD_Y_TILED_CCS ||
> + modifier == LOCAL_I915_FORMAT_MOD_Yf_TILED_CCS;
> +}
> +
> static unsigned fb_plane_width(const struct igt_fb *fb, int plane)
> {
> const struct format_desc_struct *format = lookup_drm_format(fb->drm_format);
>
> + if (is_ccs_modifier(fb->tiling) && plane == 1)
> + return DIV_ROUND_UP(fb->width, 1024) * 128;
> +
> if (plane == 0)
> return fb->width;
>
> @@ -376,13 +387,19 @@ static unsigned fb_plane_bpp(const struct igt_fb *fb, int plane)
> {
> const struct format_desc_struct *format = lookup_drm_format(fb->drm_format);
>
> - return format->plane_bpp[plane];
> + if (is_ccs_modifier(fb->tiling) && plane == 1)
> + return 8;
> + else
> + return format->plane_bpp[plane];
> }
>
> static unsigned fb_plane_height(const struct igt_fb *fb, int plane)
> {
> const struct format_desc_struct *format = lookup_drm_format(fb->drm_format);
>
> + if (is_ccs_modifier(fb->tiling) && plane == 1)
> + return DIV_ROUND_UP(fb->height, 512) * 32;
> +
> if (plane == 0)
> return fb->height;
>
> @@ -393,7 +410,10 @@ static int fb_num_planes(const struct igt_fb *fb)
> {
> const struct format_desc_struct *format = lookup_drm_format(fb->drm_format);
>
> - return format->num_planes;
> + if (is_ccs_modifier(fb->tiling))
> + return 2;
> + else
> + return format->num_planes;
> }
>
> static void fb_init(struct igt_fb *fb,
> @@ -561,8 +581,10 @@ uint64_t igt_fb_mod_to_tiling(uint64_t modifier)
> case LOCAL_I915_FORMAT_MOD_X_TILED:
> return I915_TILING_X;
> case LOCAL_I915_FORMAT_MOD_Y_TILED:
> + case LOCAL_I915_FORMAT_MOD_Y_TILED_CCS:
> return I915_TILING_Y;
> case LOCAL_I915_FORMAT_MOD_Yf_TILED:
> + case LOCAL_I915_FORMAT_MOD_Yf_TILED_CCS:
> return I915_TILING_Yf;
> default:
> igt_assert(0);
> @@ -1459,8 +1481,54 @@ struct fb_blit_upload {
> int fd;
> struct igt_fb *fb;
> struct fb_blit_linear linear;
> + drm_intel_bufmgr *bufmgr;
> + struct intel_batchbuffer *batch;
> };
>
> +static void init_buf(struct fb_blit_upload *blit,
> + struct igt_buf *buf,
> + const struct igt_fb *fb,
> + const char *name)
> +{
> + igt_assert_eq(fb->offsets[0], 0);
> +
> + buf->bo = gem_handle_to_libdrm_bo(blit->bufmgr, blit->fd,
> + name, fb->gem_handle);
> + buf->tiling = igt_fb_mod_to_tiling(fb->tiling);
> + buf->stride = fb->strides[0];
> + buf->bpp = fb->plane_bpp[0];
> + buf->size = fb->size;
> +
> + if (is_ccs_modifier(fb->tiling)) {
> + igt_assert_eq(fb->strides[0] & 127, 0);
> + igt_assert_eq(fb->strides[1] & 127, 0);
> +
> + buf->aux.offset = fb->offsets[1];
> + buf->aux.stride = fb->strides[1];
> + }
> +}
> +
> +static void rendercopy(struct fb_blit_upload *blit,
> + const struct igt_fb *dst_fb,
> + const struct igt_fb *src_fb)
> +{
> + struct igt_buf src = {}, dst = {};
> + igt_render_copyfunc_t render_copy =
> + igt_get_render_copyfunc(intel_get_drm_devid(blit->fd));
> +
> + igt_require(render_copy);
> +
> + igt_assert_eq(dst_fb->offsets[0], 0);
> + igt_assert_eq(src_fb->offsets[0], 0);
> +
> + init_buf(blit, &src, src_fb, "cairo rendercopy src");
> + init_buf(blit, &dst, dst_fb, "cairo rendercopy dst");
> +
> + render_copy(blit->batch, NULL,
> + &src, 0, 0, dst_fb->plane_width[0], dst_fb->plane_height[0],
> + &dst, 0, 0);
> +}
> +
> static void blitcopy(const struct igt_fb *dst_fb,
> const struct igt_fb *src_fb)
> {
> @@ -1498,7 +1566,10 @@ static void free_linear_mapping(struct fb_blit_upload *blit)
> gem_set_domain(fd, linear->fb.gem_handle,
> I915_GEM_DOMAIN_GTT, 0);
>
> - blitcopy(fb, &linear->fb);
> + if (blit->batch)
> + rendercopy(blit, fb, &linear->fb);
> + else
> + blitcopy(fb, &linear->fb);
>
> gem_sync(fd, linear->fb.gem_handle);
> gem_close(fd, linear->fb.gem_handle);
> @@ -1515,8 +1586,26 @@ static void destroy_cairo_surface__blit(void *arg)
> free(blit);
> }
>
> -static void setup_linear_mapping(int fd, struct igt_fb *fb, struct fb_blit_linear *linear)
> +static void destroy_cairo_surface__rendercopy(void *arg)
> +{
> + struct fb_blit_upload *blit = arg;
> +
> + blit->fb->cairo_surface = NULL;
> +
> + free_linear_mapping(blit);
> +
> + intel_batchbuffer_free(blit->batch);
> + drm_intel_bufmgr_destroy(blit->bufmgr);
> +
> + free(blit);
> +}
> +
> +static void setup_linear_mapping(struct fb_blit_upload *blit)
> {
> + int fd = blit->fd;
> + struct igt_fb *fb = blit->fb;
> + struct fb_blit_linear *linear = &blit->linear;
> +
> /*
> * We create a linear BO that we'll map for the CPU to write to (using
> * cairo). This linear bo will be then blitted to its final
> @@ -1535,7 +1624,10 @@ static void setup_linear_mapping(int fd, struct igt_fb *fb, struct fb_blit_linea
> gem_set_domain(fd, linear->fb.gem_handle,
> I915_GEM_DOMAIN_GTT, 0);
>
> - blitcopy(&linear->fb, fb);
> + if (blit->batch)
> + rendercopy(blit, &linear->fb, fb);
> + else
> + blitcopy(&linear->fb, fb);
>
> gem_sync(fd, linear->fb.gem_handle);
>
> @@ -1552,12 +1644,12 @@ static void create_cairo_surface__blit(int fd, struct igt_fb *fb)
> struct fb_blit_upload *blit;
> cairo_format_t cairo_format;
>
> - blit = malloc(sizeof(*blit));
> + blit = calloc(1, sizeof(*blit));
> igt_assert(blit);
>
> blit->fd = fd;
> blit->fb = fb;
> - setup_linear_mapping(fd, fb, &blit->linear);
> + setup_linear_mapping(blit);
>
> cairo_format = drm_format_to_cairo(fb->drm_format);
> fb->cairo_surface =
> @@ -1572,6 +1664,36 @@ static void create_cairo_surface__blit(int fd, struct igt_fb *fb)
> blit, destroy_cairo_surface__blit);
> }
>
> +static void create_cairo_surface__rendercopy(int fd, struct igt_fb *fb)
> +{
> + struct fb_blit_upload *blit;
> + cairo_format_t cairo_format;
> +
> + blit = calloc(1, sizeof(*blit));
> + igt_assert(blit);
> +
> + blit->fd = fd;
> + blit->fb = fb;
> +
> + blit->bufmgr = drm_intel_bufmgr_gem_init(fd, 4096);
> + blit->batch = intel_batchbuffer_alloc(blit->bufmgr,
> + intel_get_drm_devid(fd));
> +
> + setup_linear_mapping(blit);
> +
> + cairo_format = drm_format_to_cairo(fb->drm_format);
> + fb->cairo_surface =
> + cairo_image_surface_create_for_data(blit->linear.map,
> + cairo_format,
> + fb->width, fb->height,
> + blit->linear.fb.strides[0]);
> + fb->domain = I915_GEM_DOMAIN_GTT;
> +
> + cairo_surface_set_user_data(fb->cairo_surface,
> + (cairo_user_data_key_t *)create_cairo_surface__rendercopy,
> + blit, destroy_cairo_surface__rendercopy);
> +}
> +
> /**
> * igt_dirty_fb:
> * @fd: open drm file descriptor
> @@ -2337,7 +2459,7 @@ static void destroy_cairo_surface__convert(void *arg)
>
> static void create_cairo_surface__convert(int fd, struct igt_fb *fb)
> {
> - struct fb_convert_blit_upload *blit = malloc(sizeof(*blit));
> + struct fb_convert_blit_upload *blit = calloc(1, sizeof(*blit));
> struct fb_convert cvt = { };
> const struct format_desc_struct *f = lookup_drm_format(fb->drm_format);
> unsigned drm_format;
> @@ -2369,6 +2491,13 @@ static void create_cairo_surface__convert(int fd, struct igt_fb *fb)
>
> blit->base.fd = fd;
> blit->base.fb = fb;
> +
> + if (is_ccs_modifier(fb->tiling)) {
> + blit->base.bufmgr = drm_intel_bufmgr_gem_init(fd, 4096);
> + blit->base.batch = intel_batchbuffer_alloc(blit->base.bufmgr,
> + intel_get_drm_devid(fd));
> + }
> +
> blit->shadow_ptr = igt_fb_create_cairo_shadow_buffer(fd, drm_format,
> fb->width,
> fb->height,
> @@ -2376,8 +2505,10 @@ static void create_cairo_surface__convert(int fd, struct igt_fb *fb)
> igt_assert(blit->shadow_ptr);
>
> if (fb->tiling == LOCAL_I915_FORMAT_MOD_Y_TILED ||
> - fb->tiling == LOCAL_I915_FORMAT_MOD_Yf_TILED) {
> - setup_linear_mapping(fd, fb, &blit->base.linear);
> + fb->tiling == LOCAL_I915_FORMAT_MOD_Yf_TILED ||
> + fb->tiling == LOCAL_I915_FORMAT_MOD_Y_TILED_CCS ||
> + fb->tiling == LOCAL_I915_FORMAT_MOD_Yf_TILED_CCS) {
> + setup_linear_mapping(&blit->base);
> } else {
> blit->base.linear.fb = *fb;
> blit->base.linear.fb.gem_handle = 0;
> @@ -2456,8 +2587,10 @@ cairo_surface_t *igt_get_cairo_surface(int fd, struct igt_fb *fb)
> ((f->cairo_id == CAIRO_FORMAT_INVALID) &&
> (f->pixman_id != PIXMAN_invalid)))
> create_cairo_surface__convert(fd, fb);
> + else if (is_ccs_modifier(fb->tiling))
> + create_cairo_surface__rendercopy(fd, fb);
> else if (fb->tiling == LOCAL_I915_FORMAT_MOD_Y_TILED ||
> - fb->tiling == LOCAL_I915_FORMAT_MOD_Yf_TILED)
> + fb->tiling == LOCAL_I915_FORMAT_MOD_Yf_TILED)
> create_cairo_surface__blit(fd, fb);
> else
> create_cairo_surface__gtt(fd, fb);
More information about the igt-dev
mailing list