blankie
/
wlroots
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Remove support for DMA-BUF flags 

They are never used in practice, which makes all of our flag
handling effectively dead code. Also, APIs such as KMS don't
provide a good way to deal with the flags. Let's just fail the
DMA-BUF import when clients provide flags.
This commit is contained in:
Simon Ser 2021-11-16 22:51:06 +01:00 committed by Simon Zeni
parent 9a4e1095ca
commit a04cfca4da
15 changed files with 12 additions and 83 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
backend/drm/renderer.c
View File

@ -282,12 +282,6 @@ static struct wlr_drm_fb *drm_fb_create(struct wlr_drm_backend *drm,
goto error_get_dmabuf;
}
if (attribs.flags != 0) {
wlr_log(WLR_DEBUG, "Buffer with DMA-BUF flags 0x%"PRIX32" cannot be "
"scanned out", attribs.flags);
goto error_get_dmabuf;
}
if (formats && !wlr_drm_format_set_has(formats, attribs.format,
attribs.modifier)) {
// The format isn't supported by the plane. Try stripping the alpha

12
backend/wayland/output.c
View File

@ -166,18 +166,8 @@ static struct wl_buffer *import_dmabuf(struct wlr_wl_backend *wl,
dmabuf->offset[i], dmabuf->stride[i], modifier_hi, modifier_lo);
}
uint32_t flags = 0;
if (dmabuf->flags & WLR_DMABUF_ATTRIBUTES_FLAGS_Y_INVERT) {
flags |= ZWP_LINUX_BUFFER_PARAMS_V1_FLAGS_Y_INVERT;
}
if (dmabuf->flags & WLR_DMABUF_ATTRIBUTES_FLAGS_INTERLACED) {
flags |= ZWP_LINUX_BUFFER_PARAMS_V1_FLAGS_INTERLACED;
}
if (dmabuf->flags & WLR_DMABUF_ATTRIBUTES_FLAGS_BOTTOM_FIRST) {
flags |= ZWP_LINUX_BUFFER_PARAMS_V1_FLAGS_BOTTOM_FIRST;
}
struct wl_buffer *wl_buffer = zwp_linux_buffer_params_v1_create_immed(
params, dmabuf->width, dmabuf->height, dmabuf->format, flags);
params, dmabuf->width, dmabuf->height, dmabuf->format, 0);
// TODO: handle create() errors
return wl_buffer;
}

4
backend/x11/output.c
View File

@ -140,10 +140,6 @@ static xcb_pixmap_t import_dmabuf(struct wlr_x11_output *output,
return XCB_PIXMAP_NONE;
}
if (dmabuf->flags != 0) {
return XCB_PIXMAP_NONE;
}
// xcb closes the FDs after sending them, so we need to dup them here
struct wlr_dmabuf_attributes dup_attrs = {0};
if (!wlr_dmabuf_attributes_copy(&dup_attrs, dmabuf)) {

2
include/render/gles2.h
View File

@ -24,7 +24,6 @@ struct wlr_gles2_pixel_format {
struct wlr_gles2_tex_shader {
GLuint program;
GLint proj;
GLint invert_y;
GLint tex;
GLint alpha;
GLint pos_attrib;
@ -101,7 +100,6 @@ struct wlr_gles2_texture {
EGLImageKHR image;
bool inverted_y;
bool has_alpha;
// Only affects target == GL_TEXTURE_2D

1
include/render/vulkan.h
View File

@ -246,7 +246,6 @@ struct wlr_vk_texture {
bool dmabuf_imported;
bool owned; // if dmabuf_imported: whether we have ownership of the image
bool transitioned; // if dma_imported: whether we transitioned it away from preinit
bool invert_y; // if dma_imported: whether we must flip y
struct wl_list foreign_link;
struct wl_list destroy_link;
struct wl_list link; // wlr_gles2_renderer.textures

7
include/wlr/render/dmabuf.h
View File

@ -14,16 +14,9 @@
#define WLR_DMABUF_MAX_PLANES 4
enum wlr_dmabuf_attributes_flags {
WLR_DMABUF_ATTRIBUTES_FLAGS_Y_INVERT = 1 << 0,
WLR_DMABUF_ATTRIBUTES_FLAGS_INTERLACED = 1 << 1,
WLR_DMABUF_ATTRIBUTES_FLAGS_BOTTOM_FIRST = 1 << 2,
};
struct wlr_dmabuf_attributes {
int32_t width, height;
uint32_t format;
uint32_t flags; // enum wlr_dmabuf_attributes_flags
uint64_t modifier;
int n_planes;

1
include/wlr/render/gles2.h
View File

@ -30,7 +30,6 @@ struct wlr_gles2_texture_attribs {
GLenum target; /* either GL_TEXTURE_2D or GL_TEXTURE_EXTERNAL_OES */
GLuint tex;
bool inverted_y;
bool has_alpha;
};

4
render/gles2/renderer.c
View File

@ -309,7 +309,6 @@ static bool gles2_render_subtexture_with_matrix(
glUseProgram(shader->program);
glUniformMatrix3fv(shader->proj, 1, GL_FALSE, gl_matrix);
glUniform1i(shader->invert_y, texture->inverted_y);
glUniform1i(shader->tex, 0);
glUniform1f(shader->alpha, alpha);
@ -810,7 +809,6 @@ struct wlr_renderer *wlr_gles2_renderer_create(struct wlr_egl *egl) {
goto error;
}
renderer->shaders.tex_rgba.proj = glGetUniformLocation(prog, "proj");
renderer->shaders.tex_rgba.invert_y = glGetUniformLocation(prog, "invert_y");
renderer->shaders.tex_rgba.tex = glGetUniformLocation(prog, "tex");
renderer->shaders.tex_rgba.alpha = glGetUniformLocation(prog, "alpha");
renderer->shaders.tex_rgba.pos_attrib = glGetAttribLocation(prog, "pos");
@ -822,7 +820,6 @@ struct wlr_renderer *wlr_gles2_renderer_create(struct wlr_egl *egl) {
goto error;
}
renderer->shaders.tex_rgbx.proj = glGetUniformLocation(prog, "proj");
renderer->shaders.tex_rgbx.invert_y = glGetUniformLocation(prog, "invert_y");
renderer->shaders.tex_rgbx.tex = glGetUniformLocation(prog, "tex");
renderer->shaders.tex_rgbx.alpha = glGetUniformLocation(prog, "alpha");
renderer->shaders.tex_rgbx.pos_attrib = glGetAttribLocation(prog, "pos");
@ -835,7 +832,6 @@ struct wlr_renderer *wlr_gles2_renderer_create(struct wlr_egl *egl) {
goto error;
}
renderer->shaders.tex_ext.proj = glGetUniformLocation(prog, "proj");
renderer->shaders.tex_ext.invert_y = glGetUniformLocation(prog, "invert_y");
renderer->shaders.tex_ext.tex = glGetUniformLocation(prog, "tex");
renderer->shaders.tex_ext.alpha = glGetUniformLocation(prog, "alpha");
renderer->shaders.tex_ext.pos_attrib = glGetAttribLocation(prog, "pos");

7
render/gles2/shaders.c
View File

@ -28,18 +28,13 @@ const GLchar quad_fragment_src[] =
// Textured quads
const GLchar tex_vertex_src[] =
"uniform mat3 proj;\n"
"uniform bool invert_y;\n"
"attribute vec2 pos;\n"
"attribute vec2 texcoord;\n"
"varying vec2 v_texcoord;\n"
"\n"
"void main() {\n"
" gl_Position = vec4(proj * vec3(pos, 1.0), 1.0);\n"
" if (invert_y) {\n"
" v_texcoord = vec2(texcoord.x, 1.0 - texcoord.y);\n"
" } else {\n"
" v_texcoord = texcoord;\n"
" }\n"
" v_texcoord = texcoord;\n"
"}\n";
const GLchar tex_fragment_src_rgba[] =

3
render/gles2/texture.c
View File

@ -250,8 +250,6 @@ static struct wlr_texture *gles2_texture_from_dmabuf(
return NULL;
}
texture->drm_format = DRM_FORMAT_INVALID; // texture can't be written anyways
texture->inverted_y =
(attribs->flags & WLR_DMABUF_ATTRIBUTES_FLAGS_Y_INVERT) != 0;
const struct wlr_pixel_format_info *drm_fmt =
drm_get_pixel_format_info(attribs->format);
@ -363,6 +361,5 @@ void wlr_gles2_texture_get_attribs(struct wlr_texture *wlr_texture,
memset(attribs, 0, sizeof(*attribs));
attribs->target = texture->target;
attribs->tex = texture->tex;
attribs->inverted_y = texture->inverted_y;
attribs->has_alpha = texture->has_alpha;
}

14
render/vulkan/renderer.c
View File

@ -433,15 +433,6 @@ static struct wlr_vk_render_buffer *create_render_buffer(
wlr_log(WLR_DEBUG, "vulkan create_render_buffer: %.4s, %dx%d",
(const char*) &dmabuf.format, dmabuf.width, dmabuf.height);
// NOTE: we could at least support WLR_DMABUF_ATTRIBUTES_FLAGS_Y_INVERT
// if it is needed by anyone. Can be implemented using negative viewport
// height or flipping matrix.
if (dmabuf.flags != 0) {
wlr_log(WLR_ERROR, "dmabuf flags %x not supported/implemented on vulkan",
dmabuf.flags);
goto error_buffer;
}
buffer->image = vulkan_import_dmabuf(renderer, &dmabuf,
buffer->memories, &buffer->mem_count, true);
if (!buffer->image) {
@ -789,11 +780,6 @@ static bool vulkan_render_subtexture_with_matrix(struct wlr_renderer *wlr_render
vert_pcr_data.uv_size[0] = box->width / wlr_texture->width;
vert_pcr_data.uv_size[1] = box->height / wlr_texture->height;
if (texture->invert_y) {
vert_pcr_data.uv_off[1] += vert_pcr_data.uv_size[1];
vert_pcr_data.uv_size[1] = -vert_pcr_data.uv_size[1];
}
vkCmdPushConstants(cb, renderer->pipe_layout,
VK_SHADER_STAGE_VERTEX_BIT, 0, sizeof(vert_pcr_data), &vert_pcr_data);
vkCmdPushConstants(cb, renderer->pipe_layout,

13
render/vulkan/texture.c
View File

@ -605,19 +605,6 @@ static struct wlr_texture *vulkan_texture_from_dmabuf(struct wlr_renderer *wlr_r
goto error;
}
uint32_t flags = attribs->flags;
if (flags & WLR_DMABUF_ATTRIBUTES_FLAGS_Y_INVERT) {
texture->invert_y = true;
flags &= ~WLR_DMABUF_ATTRIBUTES_FLAGS_Y_INVERT;
}
if (flags != 0) {
wlr_log(WLR_ERROR, "dmabuf flags %x not supported/implemented on vulkan",
attribs->flags);
// NOTE: should probably make this a critical error in future
// return VK_NULL_HANDLE;
}
const struct wlr_pixel_format_info *format_info = drm_get_pixel_format_info(attribs->format);
assert(format_info);

2
types/wlr_export_dmabuf_v1.c
View File

@ -77,7 +77,7 @@ static void frame_output_handle_commit(struct wl_listener *listener,
uint32_t mod_high = attribs.modifier >> 32;
uint32_t mod_low = attribs.modifier & 0xFFFFFFFF;
zwlr_export_dmabuf_frame_v1_send_frame(frame->resource,
attribs.width, attribs.height, 0, 0, attribs.flags, frame_flags,
attribs.width, attribs.height, 0, 0, 0, frame_flags,
attribs.format, mod_high, mod_low, attribs.n_planes);
for (int i = 0; i < attribs.n_planes; ++i) {

6
types/wlr_linux_dmabuf_v1.c
View File

@ -215,10 +215,14 @@ static void params_create_common(struct wl_resource *params_resource,
goto err_out;
}
if (flags != 0) {
wlr_log(WLR_ERROR, "dmabuf flags aren't supported");
goto err_failed;
}
attribs.width = width;
attribs.height = height;
attribs.format = format;
attribs.flags = flags;
if (width < 1 || height < 1) {
wl_resource_post_error(params_resource,

13
types/wlr_screencopy_v1.c
View File

@ -254,8 +254,7 @@ error_src_tex:
return false;
}
static bool frame_dma_copy(struct wlr_screencopy_frame_v1 *frame,
uint32_t *flags) {
static bool frame_dma_copy(struct wlr_screencopy_frame_v1 *frame) {
struct wlr_dmabuf_v1_buffer *dma_buffer = frame->dma_buffer;
struct wlr_output *output = frame->output;
struct wlr_renderer *renderer = wlr_backend_get_renderer(output->backend);
@ -268,11 +267,7 @@ static bool frame_dma_copy(struct wlr_screencopy_frame_v1 *frame,
return false;
}
bool ok = blit_dmabuf(renderer, dma_buffer, output->front_buffer);
*flags = dma_buffer->attributes.flags & WLR_DMABUF_ATTRIBUTES_FLAGS_Y_INVERT ?
ZWLR_SCREENCOPY_FRAME_V1_FLAGS_Y_INVERT : 0;
return ok;
return blit_dmabuf(renderer, dma_buffer, output->front_buffer);
}
static void frame_handle_output_commit(struct wl_listener *listener,
@ -282,7 +277,6 @@ static void frame_handle_output_commit(struct wl_listener *listener,
struct wlr_output_event_commit *event = data;
struct wlr_output *output = frame->output;
struct wlr_renderer *renderer = wlr_backend_get_renderer(output->backend);
uint32_t flags;
assert(renderer);
if (!(event->committed & WLR_OUTPUT_STATE_BUFFER)) {
@ -304,8 +298,9 @@ static void frame_handle_output_commit(struct wl_listener *listener,
wl_list_remove(&frame->output_commit.link);
wl_list_init(&frame->output_commit.link);
uint32_t flags = 0;
bool ok = frame->shm_buffer ?
frame_shm_copy(frame, &flags) : frame_dma_copy(frame, &flags);
frame_shm_copy(frame, &flags) : frame_dma_copy(frame);
if (!ok) {
zwlr_screencopy_frame_v1_send_failed(frame->resource);
frame_destroy(frame);