2018-10-19 12:14:11 +03:00
|
|
|
/*
|
|
|
|
* Copyright © 2018 Collabora, Ltd.
|
|
|
|
*
|
|
|
|
* Permission is hereby granted, free of charge, to any person obtaining
|
|
|
|
* a copy of this software and associated documentation files (the
|
|
|
|
* "Software"), to deal in the Software without restriction, including
|
|
|
|
* without limitation the rights to use, copy, modify, merge, publish,
|
|
|
|
* distribute, sublicense, and/or sell copies of the Software, and to
|
|
|
|
* permit persons to whom the Software is furnished to do so, subject to
|
|
|
|
* the following conditions:
|
|
|
|
*
|
|
|
|
* The above copyright notice and this permission notice (including the
|
|
|
|
* next paragraph) shall be included in all copies or substantial
|
|
|
|
* portions of the Software.
|
|
|
|
*
|
|
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
|
|
|
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
|
|
|
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
|
|
|
|
* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
|
|
|
|
* BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
|
|
|
|
* ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
|
|
|
|
* CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
|
|
|
* SOFTWARE.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include "config.h"
|
|
|
|
|
libweston: Support zwp_surface_synchronization_v1.set_acquire_fence
Implement the set_acquire_fence request of the
zwp_surface_synchronization_v1 interface.
The implementation uses the acquire fence in two ways:
1. If the associated buffer is used as GL render source, an
EGLSyncKHR is created from the fence and used to synchronize
access.
2. If the associated buffer is used as a plane framebuffer,
the acquire fence is treated as an in-fence for the atomic
commit operation. If in-fences are not supported and the buffer
has an acquire fence, we don't consider it for plane placement.
If the used compositor/renderer doesn't support explicit
synchronization, we don't advertise the protocol at all. Currently only
the DRM and X11 backends when using the GL renderer advertise the
protocol for production use.
Issues for discussion
---------------------
a. Currently, a server-side wait of EGLSyncKHR is performed before
using the EGLImage/texture during rendering. Unfortunately, it's not clear
from the specs whether this is generally safe to do, or we need to
sync before glEGLImageTargetTexture2DOES. The exception is
TEXTURE_EXTERNAL_OES where the spec mentions it's enough to sync
and then glBindTexture for any changes to take effect.
Changes in v5:
- Meson support.
- Make explicit sync server error reporting more generic, supporting
all explicit sync related interfaces not just
wp_linux_surface_synchronization.
- Fix typo in warning for missing EGL_KHR_wait_sync extension.
- Support minor version 2 of the explicit sync protocol (i.e., support
fences for opaque EGL buffers).
Changes in v4:
- Introduce and use fd_clear and and fd_move helpers.
- Don't check for a valid buffer when updating surface acquire fence fd
from state.
- Assert that pending state acquire fence fd is always clear
after a commit.
- Clarify that WESTON_CAP_EXPLICIT_SYNC applies to just the
renderer.
- Check for EGL_KHR_wait_sync before using eglWaitSyncKHR.
- Dup the acquire fence before passing to EGL.
Changes in v3:
- Keep acquire_fence_fd in surface instead of buffer.
- Clarify that WESTON_CAP_EXPLICIT_SYNC applies to both backend and
renderer.
- Move comment about non-ownership of in_fence_fd to struct
drm_plane_state definition.
- Assert that we don't try to use planes with in-fences when using the
legacy KMS API.
- Remove unnecessary info from wayland error messages.
- Handle acquire fence for subsurface commits.
- Guard against self-update in fd_update.
- Disconnect the client if acquire fence EGLSyncKHR creation or wait
fails.
- Use updated protocol interface names.
- User correct format specifier for resource ids.
- Advertise protocol for X11 backend with GL renderer.
Changes in v2:
- Remove sync file wait fallbacks.
- Raise UNSUPPORTED_BUFFER error at commit if we have an acquire
fence, but the committed buffer is not a valid linux_dmabuf.
- Don't put buffers with in-fences on planes that don't support
in-fences.
- Don't advertise explicit sync protocol if backend does not
support explicit sync.
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>
2018-10-19 12:14:11 +03:00
|
|
|
#include <assert.h>
|
2018-10-19 12:14:11 +03:00
|
|
|
#include <inttypes.h>
|
|
|
|
|
2019-03-28 17:28:47 +03:00
|
|
|
#include <libweston/libweston.h>
|
2018-10-19 12:14:11 +03:00
|
|
|
#include "linux-explicit-synchronization.h"
|
|
|
|
#include "linux-explicit-synchronization-unstable-v1-server-protocol.h"
|
libweston: Support zwp_surface_synchronization_v1.set_acquire_fence
Implement the set_acquire_fence request of the
zwp_surface_synchronization_v1 interface.
The implementation uses the acquire fence in two ways:
1. If the associated buffer is used as GL render source, an
EGLSyncKHR is created from the fence and used to synchronize
access.
2. If the associated buffer is used as a plane framebuffer,
the acquire fence is treated as an in-fence for the atomic
commit operation. If in-fences are not supported and the buffer
has an acquire fence, we don't consider it for plane placement.
If the used compositor/renderer doesn't support explicit
synchronization, we don't advertise the protocol at all. Currently only
the DRM and X11 backends when using the GL renderer advertise the
protocol for production use.
Issues for discussion
---------------------
a. Currently, a server-side wait of EGLSyncKHR is performed before
using the EGLImage/texture during rendering. Unfortunately, it's not clear
from the specs whether this is generally safe to do, or we need to
sync before glEGLImageTargetTexture2DOES. The exception is
TEXTURE_EXTERNAL_OES where the spec mentions it's enough to sync
and then glBindTexture for any changes to take effect.
Changes in v5:
- Meson support.
- Make explicit sync server error reporting more generic, supporting
all explicit sync related interfaces not just
wp_linux_surface_synchronization.
- Fix typo in warning for missing EGL_KHR_wait_sync extension.
- Support minor version 2 of the explicit sync protocol (i.e., support
fences for opaque EGL buffers).
Changes in v4:
- Introduce and use fd_clear and and fd_move helpers.
- Don't check for a valid buffer when updating surface acquire fence fd
from state.
- Assert that pending state acquire fence fd is always clear
after a commit.
- Clarify that WESTON_CAP_EXPLICIT_SYNC applies to just the
renderer.
- Check for EGL_KHR_wait_sync before using eglWaitSyncKHR.
- Dup the acquire fence before passing to EGL.
Changes in v3:
- Keep acquire_fence_fd in surface instead of buffer.
- Clarify that WESTON_CAP_EXPLICIT_SYNC applies to both backend and
renderer.
- Move comment about non-ownership of in_fence_fd to struct
drm_plane_state definition.
- Assert that we don't try to use planes with in-fences when using the
legacy KMS API.
- Remove unnecessary info from wayland error messages.
- Handle acquire fence for subsurface commits.
- Guard against self-update in fd_update.
- Disconnect the client if acquire fence EGLSyncKHR creation or wait
fails.
- Use updated protocol interface names.
- User correct format specifier for resource ids.
- Advertise protocol for X11 backend with GL renderer.
Changes in v2:
- Remove sync file wait fallbacks.
- Raise UNSUPPORTED_BUFFER error at commit if we have an acquire
fence, but the committed buffer is not a valid linux_dmabuf.
- Don't put buffers with in-fences on planes that don't support
in-fences.
- Don't advertise explicit sync protocol if backend does not
support explicit sync.
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>
2018-10-19 12:14:11 +03:00
|
|
|
#include "linux-sync-file.h"
|
|
|
|
#include "shared/fd-util.h"
|
2018-10-19 12:14:11 +03:00
|
|
|
|
2018-10-19 12:14:11 +03:00
|
|
|
static void
|
|
|
|
destroy_linux_buffer_release(struct wl_resource *resource)
|
|
|
|
{
|
|
|
|
struct weston_buffer_release *buffer_release =
|
|
|
|
wl_resource_get_user_data(resource);
|
|
|
|
|
|
|
|
fd_clear(&buffer_release->fence_fd);
|
|
|
|
free(buffer_release);
|
|
|
|
}
|
|
|
|
|
2018-10-19 12:14:11 +03:00
|
|
|
static void
|
|
|
|
destroy_linux_surface_synchronization(struct wl_resource *resource)
|
|
|
|
{
|
|
|
|
struct weston_surface *surface =
|
|
|
|
wl_resource_get_user_data(resource);
|
|
|
|
|
libweston: Support zwp_surface_synchronization_v1.set_acquire_fence
Implement the set_acquire_fence request of the
zwp_surface_synchronization_v1 interface.
The implementation uses the acquire fence in two ways:
1. If the associated buffer is used as GL render source, an
EGLSyncKHR is created from the fence and used to synchronize
access.
2. If the associated buffer is used as a plane framebuffer,
the acquire fence is treated as an in-fence for the atomic
commit operation. If in-fences are not supported and the buffer
has an acquire fence, we don't consider it for plane placement.
If the used compositor/renderer doesn't support explicit
synchronization, we don't advertise the protocol at all. Currently only
the DRM and X11 backends when using the GL renderer advertise the
protocol for production use.
Issues for discussion
---------------------
a. Currently, a server-side wait of EGLSyncKHR is performed before
using the EGLImage/texture during rendering. Unfortunately, it's not clear
from the specs whether this is generally safe to do, or we need to
sync before glEGLImageTargetTexture2DOES. The exception is
TEXTURE_EXTERNAL_OES where the spec mentions it's enough to sync
and then glBindTexture for any changes to take effect.
Changes in v5:
- Meson support.
- Make explicit sync server error reporting more generic, supporting
all explicit sync related interfaces not just
wp_linux_surface_synchronization.
- Fix typo in warning for missing EGL_KHR_wait_sync extension.
- Support minor version 2 of the explicit sync protocol (i.e., support
fences for opaque EGL buffers).
Changes in v4:
- Introduce and use fd_clear and and fd_move helpers.
- Don't check for a valid buffer when updating surface acquire fence fd
from state.
- Assert that pending state acquire fence fd is always clear
after a commit.
- Clarify that WESTON_CAP_EXPLICIT_SYNC applies to just the
renderer.
- Check for EGL_KHR_wait_sync before using eglWaitSyncKHR.
- Dup the acquire fence before passing to EGL.
Changes in v3:
- Keep acquire_fence_fd in surface instead of buffer.
- Clarify that WESTON_CAP_EXPLICIT_SYNC applies to both backend and
renderer.
- Move comment about non-ownership of in_fence_fd to struct
drm_plane_state definition.
- Assert that we don't try to use planes with in-fences when using the
legacy KMS API.
- Remove unnecessary info from wayland error messages.
- Handle acquire fence for subsurface commits.
- Guard against self-update in fd_update.
- Disconnect the client if acquire fence EGLSyncKHR creation or wait
fails.
- Use updated protocol interface names.
- User correct format specifier for resource ids.
- Advertise protocol for X11 backend with GL renderer.
Changes in v2:
- Remove sync file wait fallbacks.
- Raise UNSUPPORTED_BUFFER error at commit if we have an acquire
fence, but the committed buffer is not a valid linux_dmabuf.
- Don't put buffers with in-fences on planes that don't support
in-fences.
- Don't advertise explicit sync protocol if backend does not
support explicit sync.
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>
2018-10-19 12:14:11 +03:00
|
|
|
if (surface) {
|
|
|
|
fd_clear(&surface->pending.acquire_fence_fd);
|
2018-10-19 12:14:11 +03:00
|
|
|
surface->synchronization_resource = NULL;
|
libweston: Support zwp_surface_synchronization_v1.set_acquire_fence
Implement the set_acquire_fence request of the
zwp_surface_synchronization_v1 interface.
The implementation uses the acquire fence in two ways:
1. If the associated buffer is used as GL render source, an
EGLSyncKHR is created from the fence and used to synchronize
access.
2. If the associated buffer is used as a plane framebuffer,
the acquire fence is treated as an in-fence for the atomic
commit operation. If in-fences are not supported and the buffer
has an acquire fence, we don't consider it for plane placement.
If the used compositor/renderer doesn't support explicit
synchronization, we don't advertise the protocol at all. Currently only
the DRM and X11 backends when using the GL renderer advertise the
protocol for production use.
Issues for discussion
---------------------
a. Currently, a server-side wait of EGLSyncKHR is performed before
using the EGLImage/texture during rendering. Unfortunately, it's not clear
from the specs whether this is generally safe to do, or we need to
sync before glEGLImageTargetTexture2DOES. The exception is
TEXTURE_EXTERNAL_OES where the spec mentions it's enough to sync
and then glBindTexture for any changes to take effect.
Changes in v5:
- Meson support.
- Make explicit sync server error reporting more generic, supporting
all explicit sync related interfaces not just
wp_linux_surface_synchronization.
- Fix typo in warning for missing EGL_KHR_wait_sync extension.
- Support minor version 2 of the explicit sync protocol (i.e., support
fences for opaque EGL buffers).
Changes in v4:
- Introduce and use fd_clear and and fd_move helpers.
- Don't check for a valid buffer when updating surface acquire fence fd
from state.
- Assert that pending state acquire fence fd is always clear
after a commit.
- Clarify that WESTON_CAP_EXPLICIT_SYNC applies to just the
renderer.
- Check for EGL_KHR_wait_sync before using eglWaitSyncKHR.
- Dup the acquire fence before passing to EGL.
Changes in v3:
- Keep acquire_fence_fd in surface instead of buffer.
- Clarify that WESTON_CAP_EXPLICIT_SYNC applies to both backend and
renderer.
- Move comment about non-ownership of in_fence_fd to struct
drm_plane_state definition.
- Assert that we don't try to use planes with in-fences when using the
legacy KMS API.
- Remove unnecessary info from wayland error messages.
- Handle acquire fence for subsurface commits.
- Guard against self-update in fd_update.
- Disconnect the client if acquire fence EGLSyncKHR creation or wait
fails.
- Use updated protocol interface names.
- User correct format specifier for resource ids.
- Advertise protocol for X11 backend with GL renderer.
Changes in v2:
- Remove sync file wait fallbacks.
- Raise UNSUPPORTED_BUFFER error at commit if we have an acquire
fence, but the committed buffer is not a valid linux_dmabuf.
- Don't put buffers with in-fences on planes that don't support
in-fences.
- Don't advertise explicit sync protocol if backend does not
support explicit sync.
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>
2018-10-19 12:14:11 +03:00
|
|
|
}
|
2018-10-19 12:14:11 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
linux_surface_synchronization_destroy(struct wl_client *client,
|
|
|
|
struct wl_resource *resource)
|
|
|
|
{
|
|
|
|
wl_resource_destroy(resource);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
linux_surface_synchronization_set_acquire_fence(struct wl_client *client,
|
|
|
|
struct wl_resource *resource,
|
|
|
|
int32_t fd)
|
|
|
|
{
|
libweston: Support zwp_surface_synchronization_v1.set_acquire_fence
Implement the set_acquire_fence request of the
zwp_surface_synchronization_v1 interface.
The implementation uses the acquire fence in two ways:
1. If the associated buffer is used as GL render source, an
EGLSyncKHR is created from the fence and used to synchronize
access.
2. If the associated buffer is used as a plane framebuffer,
the acquire fence is treated as an in-fence for the atomic
commit operation. If in-fences are not supported and the buffer
has an acquire fence, we don't consider it for plane placement.
If the used compositor/renderer doesn't support explicit
synchronization, we don't advertise the protocol at all. Currently only
the DRM and X11 backends when using the GL renderer advertise the
protocol for production use.
Issues for discussion
---------------------
a. Currently, a server-side wait of EGLSyncKHR is performed before
using the EGLImage/texture during rendering. Unfortunately, it's not clear
from the specs whether this is generally safe to do, or we need to
sync before glEGLImageTargetTexture2DOES. The exception is
TEXTURE_EXTERNAL_OES where the spec mentions it's enough to sync
and then glBindTexture for any changes to take effect.
Changes in v5:
- Meson support.
- Make explicit sync server error reporting more generic, supporting
all explicit sync related interfaces not just
wp_linux_surface_synchronization.
- Fix typo in warning for missing EGL_KHR_wait_sync extension.
- Support minor version 2 of the explicit sync protocol (i.e., support
fences for opaque EGL buffers).
Changes in v4:
- Introduce and use fd_clear and and fd_move helpers.
- Don't check for a valid buffer when updating surface acquire fence fd
from state.
- Assert that pending state acquire fence fd is always clear
after a commit.
- Clarify that WESTON_CAP_EXPLICIT_SYNC applies to just the
renderer.
- Check for EGL_KHR_wait_sync before using eglWaitSyncKHR.
- Dup the acquire fence before passing to EGL.
Changes in v3:
- Keep acquire_fence_fd in surface instead of buffer.
- Clarify that WESTON_CAP_EXPLICIT_SYNC applies to both backend and
renderer.
- Move comment about non-ownership of in_fence_fd to struct
drm_plane_state definition.
- Assert that we don't try to use planes with in-fences when using the
legacy KMS API.
- Remove unnecessary info from wayland error messages.
- Handle acquire fence for subsurface commits.
- Guard against self-update in fd_update.
- Disconnect the client if acquire fence EGLSyncKHR creation or wait
fails.
- Use updated protocol interface names.
- User correct format specifier for resource ids.
- Advertise protocol for X11 backend with GL renderer.
Changes in v2:
- Remove sync file wait fallbacks.
- Raise UNSUPPORTED_BUFFER error at commit if we have an acquire
fence, but the committed buffer is not a valid linux_dmabuf.
- Don't put buffers with in-fences on planes that don't support
in-fences.
- Don't advertise explicit sync protocol if backend does not
support explicit sync.
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>
2018-10-19 12:14:11 +03:00
|
|
|
struct weston_surface *surface = wl_resource_get_user_data(resource);
|
|
|
|
|
|
|
|
if (!surface) {
|
|
|
|
wl_resource_post_error(
|
|
|
|
resource,
|
|
|
|
ZWP_LINUX_SURFACE_SYNCHRONIZATION_V1_ERROR_NO_SURFACE,
|
|
|
|
"surface no longer exists");
|
|
|
|
goto err;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!linux_sync_file_is_valid(fd)) {
|
|
|
|
wl_resource_post_error(
|
|
|
|
resource,
|
|
|
|
ZWP_LINUX_SURFACE_SYNCHRONIZATION_V1_ERROR_INVALID_FENCE,
|
|
|
|
"invalid fence fd");
|
|
|
|
goto err;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (surface->pending.acquire_fence_fd != -1) {
|
|
|
|
wl_resource_post_error(
|
|
|
|
resource,
|
|
|
|
ZWP_LINUX_SURFACE_SYNCHRONIZATION_V1_ERROR_DUPLICATE_FENCE,
|
|
|
|
"already have a fence fd");
|
|
|
|
goto err;
|
|
|
|
}
|
|
|
|
|
|
|
|
fd_update(&surface->pending.acquire_fence_fd, fd);
|
|
|
|
|
|
|
|
return;
|
|
|
|
|
|
|
|
err:
|
|
|
|
close(fd);
|
2018-10-19 12:14:11 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
linux_surface_synchronization_get_release(struct wl_client *client,
|
|
|
|
struct wl_resource *resource,
|
|
|
|
uint32_t id)
|
|
|
|
{
|
2018-10-19 12:14:11 +03:00
|
|
|
struct weston_surface *surface =
|
|
|
|
wl_resource_get_user_data(resource);
|
|
|
|
struct weston_buffer_release *buffer_release;
|
|
|
|
|
|
|
|
if (!surface) {
|
|
|
|
wl_resource_post_error(
|
|
|
|
resource,
|
|
|
|
ZWP_LINUX_SURFACE_SYNCHRONIZATION_V1_ERROR_NO_SURFACE,
|
|
|
|
"surface no longer exists");
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (surface->pending.buffer_release_ref.buffer_release) {
|
|
|
|
wl_resource_post_error(
|
|
|
|
resource,
|
|
|
|
ZWP_LINUX_SURFACE_SYNCHRONIZATION_V1_ERROR_DUPLICATE_RELEASE,
|
|
|
|
"already has a buffer release");
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
buffer_release = zalloc(sizeof *buffer_release);
|
|
|
|
if (buffer_release == NULL)
|
|
|
|
goto err_alloc;
|
|
|
|
|
|
|
|
buffer_release->fence_fd = -1;
|
|
|
|
buffer_release->resource =
|
|
|
|
wl_resource_create(client,
|
|
|
|
&zwp_linux_buffer_release_v1_interface,
|
|
|
|
wl_resource_get_version(resource), id);
|
|
|
|
if (!buffer_release->resource)
|
|
|
|
goto err_create;
|
|
|
|
|
|
|
|
wl_resource_set_implementation(buffer_release->resource, NULL,
|
|
|
|
buffer_release,
|
|
|
|
destroy_linux_buffer_release);
|
|
|
|
|
|
|
|
weston_buffer_release_reference(&surface->pending.buffer_release_ref,
|
|
|
|
buffer_release);
|
|
|
|
|
|
|
|
return;
|
|
|
|
|
|
|
|
err_create:
|
|
|
|
free(buffer_release);
|
|
|
|
|
|
|
|
err_alloc:
|
2018-10-19 12:14:11 +03:00
|
|
|
wl_client_post_no_memory(client);
|
2018-10-19 12:14:11 +03:00
|
|
|
|
2018-10-19 12:14:11 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
const struct zwp_linux_surface_synchronization_v1_interface
|
|
|
|
linux_surface_synchronization_implementation = {
|
|
|
|
linux_surface_synchronization_destroy,
|
|
|
|
linux_surface_synchronization_set_acquire_fence,
|
|
|
|
linux_surface_synchronization_get_release,
|
|
|
|
};
|
|
|
|
|
|
|
|
static void
|
|
|
|
linux_explicit_synchronization_destroy(struct wl_client *client,
|
|
|
|
struct wl_resource *resource)
|
|
|
|
{
|
|
|
|
wl_resource_destroy(resource);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
linux_explicit_synchronization_get_synchronization(struct wl_client *client,
|
|
|
|
struct wl_resource *resource,
|
|
|
|
uint32_t id,
|
|
|
|
struct wl_resource *surface_resource)
|
|
|
|
{
|
|
|
|
struct weston_surface *surface =
|
|
|
|
wl_resource_get_user_data(surface_resource);
|
|
|
|
|
|
|
|
if (surface->synchronization_resource) {
|
|
|
|
wl_resource_post_error(
|
|
|
|
resource,
|
|
|
|
ZWP_LINUX_EXPLICIT_SYNCHRONIZATION_V1_ERROR_SYNCHRONIZATION_EXISTS,
|
|
|
|
"wl_surface@%"PRIu32" already has a synchronization object",
|
|
|
|
wl_resource_get_id(surface_resource));
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
surface->synchronization_resource =
|
|
|
|
wl_resource_create(client,
|
|
|
|
&zwp_linux_surface_synchronization_v1_interface,
|
|
|
|
wl_resource_get_version(resource), id);
|
|
|
|
if (!surface->synchronization_resource) {
|
|
|
|
wl_client_post_no_memory(client);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
wl_resource_set_implementation(surface->synchronization_resource,
|
|
|
|
&linux_surface_synchronization_implementation,
|
|
|
|
surface,
|
|
|
|
destroy_linux_surface_synchronization);
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct zwp_linux_explicit_synchronization_v1_interface
|
|
|
|
linux_explicit_synchronization_implementation = {
|
|
|
|
linux_explicit_synchronization_destroy,
|
|
|
|
linux_explicit_synchronization_get_synchronization
|
|
|
|
};
|
|
|
|
|
|
|
|
static void
|
|
|
|
bind_linux_explicit_synchronization(struct wl_client *client,
|
|
|
|
void *data, uint32_t version,
|
|
|
|
uint32_t id)
|
|
|
|
{
|
|
|
|
struct weston_compositor *compositor = data;
|
|
|
|
struct wl_resource *resource;
|
|
|
|
|
|
|
|
resource = wl_resource_create(client,
|
|
|
|
&zwp_linux_explicit_synchronization_v1_interface,
|
|
|
|
version, id);
|
|
|
|
if (resource == NULL) {
|
|
|
|
wl_client_post_no_memory(client);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
wl_resource_set_implementation(resource,
|
|
|
|
&linux_explicit_synchronization_implementation,
|
|
|
|
compositor, NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/** Advertise linux_explicit_synchronization support
|
|
|
|
*
|
|
|
|
* Calling this initializes the zwp_linux_explicit_synchronization_v1
|
|
|
|
* protocol support, so that the interface will be advertised to clients.
|
|
|
|
* Essentially it creates a global. Do not call this function multiple times
|
|
|
|
* in the compositor's lifetime. There is no way to deinit explicitly, globals
|
|
|
|
* will be reaped when the wl_display gets destroyed.
|
|
|
|
*
|
|
|
|
* \param compositor The compositor to init for.
|
|
|
|
* \return Zero on success, -1 on failure.
|
|
|
|
*/
|
|
|
|
WL_EXPORT int
|
|
|
|
linux_explicit_synchronization_setup(struct weston_compositor *compositor)
|
|
|
|
{
|
|
|
|
/* TODO: Update to minor version 2 when the next version of
|
|
|
|
* wayland-protocols that contains it is released. */
|
|
|
|
if (!wl_global_create(compositor->wl_display,
|
|
|
|
&zwp_linux_explicit_synchronization_v1_interface,
|
|
|
|
1, compositor,
|
|
|
|
bind_linux_explicit_synchronization))
|
|
|
|
return -1;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
libweston: Support zwp_surface_synchronization_v1.set_acquire_fence
Implement the set_acquire_fence request of the
zwp_surface_synchronization_v1 interface.
The implementation uses the acquire fence in two ways:
1. If the associated buffer is used as GL render source, an
EGLSyncKHR is created from the fence and used to synchronize
access.
2. If the associated buffer is used as a plane framebuffer,
the acquire fence is treated as an in-fence for the atomic
commit operation. If in-fences are not supported and the buffer
has an acquire fence, we don't consider it for plane placement.
If the used compositor/renderer doesn't support explicit
synchronization, we don't advertise the protocol at all. Currently only
the DRM and X11 backends when using the GL renderer advertise the
protocol for production use.
Issues for discussion
---------------------
a. Currently, a server-side wait of EGLSyncKHR is performed before
using the EGLImage/texture during rendering. Unfortunately, it's not clear
from the specs whether this is generally safe to do, or we need to
sync before glEGLImageTargetTexture2DOES. The exception is
TEXTURE_EXTERNAL_OES where the spec mentions it's enough to sync
and then glBindTexture for any changes to take effect.
Changes in v5:
- Meson support.
- Make explicit sync server error reporting more generic, supporting
all explicit sync related interfaces not just
wp_linux_surface_synchronization.
- Fix typo in warning for missing EGL_KHR_wait_sync extension.
- Support minor version 2 of the explicit sync protocol (i.e., support
fences for opaque EGL buffers).
Changes in v4:
- Introduce and use fd_clear and and fd_move helpers.
- Don't check for a valid buffer when updating surface acquire fence fd
from state.
- Assert that pending state acquire fence fd is always clear
after a commit.
- Clarify that WESTON_CAP_EXPLICIT_SYNC applies to just the
renderer.
- Check for EGL_KHR_wait_sync before using eglWaitSyncKHR.
- Dup the acquire fence before passing to EGL.
Changes in v3:
- Keep acquire_fence_fd in surface instead of buffer.
- Clarify that WESTON_CAP_EXPLICIT_SYNC applies to both backend and
renderer.
- Move comment about non-ownership of in_fence_fd to struct
drm_plane_state definition.
- Assert that we don't try to use planes with in-fences when using the
legacy KMS API.
- Remove unnecessary info from wayland error messages.
- Handle acquire fence for subsurface commits.
- Guard against self-update in fd_update.
- Disconnect the client if acquire fence EGLSyncKHR creation or wait
fails.
- Use updated protocol interface names.
- User correct format specifier for resource ids.
- Advertise protocol for X11 backend with GL renderer.
Changes in v2:
- Remove sync file wait fallbacks.
- Raise UNSUPPORTED_BUFFER error at commit if we have an acquire
fence, but the committed buffer is not a valid linux_dmabuf.
- Don't put buffers with in-fences on planes that don't support
in-fences.
- Don't advertise explicit sync protocol if backend does not
support explicit sync.
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>
2018-10-19 12:14:11 +03:00
|
|
|
|
|
|
|
/** Resolve an internal compositor error by disconnecting the client.
|
|
|
|
*
|
|
|
|
* This function is used in cases when explicit synchronization
|
|
|
|
* turns out to be unusable and there is no fallback path.
|
|
|
|
*
|
|
|
|
* It is possible the fault is caused by a compositor bug, the underlying
|
|
|
|
* graphics stack bug or normal behaviour, or perhaps a client mistake.
|
|
|
|
* In any case, the options are to either composite garbage or nothing,
|
|
|
|
* or disconnect the client. This is a helper function for the latter.
|
|
|
|
*
|
|
|
|
* The error is sent as an INVALID_OBJECT error on the client's wl_display.
|
|
|
|
*
|
2019-06-12 16:05:44 +03:00
|
|
|
* \param resource The explicit synchronization related resource that is unusable.
|
libweston: Support zwp_surface_synchronization_v1.set_acquire_fence
Implement the set_acquire_fence request of the
zwp_surface_synchronization_v1 interface.
The implementation uses the acquire fence in two ways:
1. If the associated buffer is used as GL render source, an
EGLSyncKHR is created from the fence and used to synchronize
access.
2. If the associated buffer is used as a plane framebuffer,
the acquire fence is treated as an in-fence for the atomic
commit operation. If in-fences are not supported and the buffer
has an acquire fence, we don't consider it for plane placement.
If the used compositor/renderer doesn't support explicit
synchronization, we don't advertise the protocol at all. Currently only
the DRM and X11 backends when using the GL renderer advertise the
protocol for production use.
Issues for discussion
---------------------
a. Currently, a server-side wait of EGLSyncKHR is performed before
using the EGLImage/texture during rendering. Unfortunately, it's not clear
from the specs whether this is generally safe to do, or we need to
sync before glEGLImageTargetTexture2DOES. The exception is
TEXTURE_EXTERNAL_OES where the spec mentions it's enough to sync
and then glBindTexture for any changes to take effect.
Changes in v5:
- Meson support.
- Make explicit sync server error reporting more generic, supporting
all explicit sync related interfaces not just
wp_linux_surface_synchronization.
- Fix typo in warning for missing EGL_KHR_wait_sync extension.
- Support minor version 2 of the explicit sync protocol (i.e., support
fences for opaque EGL buffers).
Changes in v4:
- Introduce and use fd_clear and and fd_move helpers.
- Don't check for a valid buffer when updating surface acquire fence fd
from state.
- Assert that pending state acquire fence fd is always clear
after a commit.
- Clarify that WESTON_CAP_EXPLICIT_SYNC applies to just the
renderer.
- Check for EGL_KHR_wait_sync before using eglWaitSyncKHR.
- Dup the acquire fence before passing to EGL.
Changes in v3:
- Keep acquire_fence_fd in surface instead of buffer.
- Clarify that WESTON_CAP_EXPLICIT_SYNC applies to both backend and
renderer.
- Move comment about non-ownership of in_fence_fd to struct
drm_plane_state definition.
- Assert that we don't try to use planes with in-fences when using the
legacy KMS API.
- Remove unnecessary info from wayland error messages.
- Handle acquire fence for subsurface commits.
- Guard against self-update in fd_update.
- Disconnect the client if acquire fence EGLSyncKHR creation or wait
fails.
- Use updated protocol interface names.
- User correct format specifier for resource ids.
- Advertise protocol for X11 backend with GL renderer.
Changes in v2:
- Remove sync file wait fallbacks.
- Raise UNSUPPORTED_BUFFER error at commit if we have an acquire
fence, but the committed buffer is not a valid linux_dmabuf.
- Don't put buffers with in-fences on planes that don't support
in-fences.
- Don't advertise explicit sync protocol if backend does not
support explicit sync.
Signed-off-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>
2018-10-19 12:14:11 +03:00
|
|
|
* \param msg A custom error message attached to the protocol error.
|
|
|
|
*/
|
|
|
|
WL_EXPORT void
|
|
|
|
linux_explicit_synchronization_send_server_error(struct wl_resource *resource,
|
|
|
|
const char *msg)
|
|
|
|
{
|
|
|
|
uint32_t id = wl_resource_get_id(resource);
|
|
|
|
const char *class = wl_resource_get_class(resource);
|
|
|
|
struct wl_client *client = wl_resource_get_client(resource);
|
|
|
|
struct wl_resource *display_resource = wl_client_get_object(client, 1);
|
|
|
|
|
|
|
|
assert(display_resource);
|
|
|
|
wl_resource_post_error(display_resource,
|
|
|
|
WL_DISPLAY_ERROR_INVALID_OBJECT,
|
|
|
|
"linux_explicit_synchronization server error "
|
|
|
|
"with %s@%"PRIu32": %s",
|
|
|
|
class, id, msg);
|
|
|
|
}
|