Commit graph

635 commits

Author SHA1 Message Date
Simon Ser
6b7d1d732a render/pixman: fix scaling in render_pass_add_texture()
Closes: https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/3642
Fixes: 6830bfc17f ("render/pixman: implement render pass API")
2023-05-02 15:49:17 +00:00
Simon Ser
f5a5712a02 render/vulkan: use VK_KHR_synchronization2
This simplifies the vkQueueSubmit call, removing the need to pass
timeline semaphore point values as separate arrays.
2023-04-28 15:36:39 +00:00
Simon Ser
6830bfc17f render/pixman: implement render pass API 2023-04-25 17:26:35 +02:00
Simon Ser
6ce371a317 render/pixman: add begin_pixman_data_ptr_access()
This logic is used twice.
2023-04-25 17:26:35 +02:00
Simon Ser
471249ae4f render/pass: add legacy render pass implementation
Add a render pass implementation which leverages the legacy
rendering API. This is helpful during the migration.
2023-04-25 17:25:10 +02:00
Simon Ser
4a1ad32534 render: add render pass helpers 2023-04-25 17:25:10 +02:00
Simon Ser
756dedae20 Add a new renderer API
Goals:

- Extensibility: we need to be able to add new params to the calls
  to render a texture/rect. For instance we'll need to add fences to
  the render texture operation for explicit sync purposes.
- No implicit state: no more bind_buffer, begin, end.
- No matrices: these hurt Pixman and we don't need them.
- Clip regions for optimized damage repainting.

Closes: https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/3188
2023-04-25 17:25:10 +02:00
Simon Ser
d795824346 render/vulkan: fix signedness of strcmp() return value
strcmp() returns a signed integer.
2023-04-19 15:29:53 +02:00
Simon Ser
62e60ba889 render/vulkan: use initializer in vulkan_change_layout_queue() 2023-04-19 15:29:26 +02:00
Simon Ser
74e8fd8a9a render/vulkan: remove extra newline 2023-04-19 15:29:11 +02:00
Simon Ser
92b726ef67 render/vulkan: use full names for extension command fields
This makes grepping much easier.
2023-04-18 15:58:55 +00:00
Manuel Stoeckl
40dde59475 render/gles2: use byte-aligned rows for glReadPixels
Setting the GLESv2 parameter GL_PACK_ALIGNMENT to 1 ensures that the
stride of the glReadPixels output matches the value computed in
`pack_stride`. Since the default value of GL_PACK_ALIGNMENT is 4, this
does not make a difference under normal use; but without this patch
the stride can be incorrect; for example, with RGB565 buffers and
screenshots of regions with odd width.
2023-04-12 14:04:25 +00:00
Simon Ser
59d2743c0c render/vulkan: avoid double-close on sync_file FD after semaphore import
Closes: https://gitlab.freedesktop.org/wlroots/wlroots/-/issues/3610
2023-03-26 07:19:01 +00:00
Brett Ernst
7d9938c957 drm_format_set: fix realloc size 2023-03-16 18:13:11 -07:00
Simon Ser
40117e1e0d render/vulkan: prefer render node for wlr_renderer_get_drm_fd()
Instead of returning a primary node from wlr_renderer_get_drm_fd(),
prefer to return a render node if any.
2023-02-28 14:25:21 +00:00
Jan Beich
59acc69737 vulkan: silence -Wint-conversion on 32-bit architectures
render/vulkan/renderer.c:388:70: error: incompatible pointer to integer conversion passing 'void *' to parameter of type 'VkFence' (aka 'unsigned long long') [-Werror,-Wint-conversion]
        VkResult res = vkQueueSubmit(renderer->dev->queue, 1, &submit_info, NULL);
                                                                            ^~~~
render/vulkan/renderer.c:1141:81: error: incompatible pointer to integer conversion passing 'void *' to parameter of type 'VkFence' (aka 'unsigned long long') [-Werror,-Wint-conversion]
        VkResult res = vkQueueSubmit(renderer->dev->queue, submit_count, submit_infos, NULL);
                                                                                       ^~~~
/usr/include/sys/_null.h:34:14: note: expanded from macro 'NULL'
 #define NULL    ((void *)0)
                 ^~~~~~~~~~~
/usr/local/include/vulkan/vulkan_core.h:4054:49: note: passing argument to parameter 'fence' here
    VkFence                                     fence);
                                                ^

Fixes: a8a194d695 ("render/vulkan: switch to timeline semaphores")
2023-02-24 16:41:40 +00:00
Simon Ser
8619ffab21 swapchain: remove allocator listener on destroy
Past that point the wlr_allocator is gone and the listeners are
invalidated.
2023-02-24 00:07:41 +01:00
Simon Ser
b33ab26fe7 render/swapchain: make public
We've had this struct for a while. It'd be useful for compositors
if they want to manage the swap chains themselves instead of being
forced to use wlr_output's. Some compositors might also want to use
a swapchain without an output.
2023-02-21 17:14:31 +01:00
Simon Zeni
782b5e6565 Revert "render/egl: skip incompatible EGL devices"
Reverts commit c73e20628a.

This caused a regression in the GLES2 renderer because `egl->exts.EXT_device_drm` is set in
`egl_init_display()`, which is invoked after `get_egl_device_from_drm_fd()`. So the function will
always return `EGL_NO_DEVICE_EXT`.
2023-02-18 11:49:51 -05:00
Eric Engestrom
c73e20628a render/egl: skip incompatible EGL devices
Without EGL_EXT_device_drm, eglQueryDeviceStringEXT(EGL_DRM_DEVICE_FILE_EXT) further below is invalid.
2023-02-16 19:43:06 +00:00
Simon Ser
677a3f2f88 render/gles2: default to highp if available
The spec [1] says that the maximum value for a mediump float
is at least 2¹⁴ in section 4.5.2. However, when using a 4k
resolution texture coordinates will exceed this value. This causes
issues on drivers which implement mediump as a 16-bit [2].

Switch to highp. There's a twist: on GLES2, support for highp is
optional. So we need to guard it with cute GL_FRAGMENT_PRECISION_HIGH
ifdefs.

[1]: https://registry.khronos.org/OpenGL/specs/es/2.0/GLSL_ES_Specification_1.00.pdf
[2]: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21082
2023-02-02 22:54:13 +01:00
Simon Ser
0534d12b28 render/gles2: use correct type for shader type
Doesn't matter a lot, but let's try to be consistent with the
GL headers.
2023-02-02 22:42:42 +01:00
Simon Ser
377668aaf6 render/vulkan: check that requested size is smaller than max
Fail with a clearer error when the requested size is too large.
Without this, we allocate a buffer with a size smaller than what
was requested.

References: https://gitlab.freedesktop.org/wlroots/wlroots/-/merge_requests/3975
2023-01-19 18:50:59 +00:00
Brett Ernst
1ba322e9ff vulkan: increase max stage size to support large buffers 2023-01-18 19:45:26 -08:00
llyyr
10e16db738 render/vulkan: use correct source offset in read_pixels 2023-01-18 06:56:49 +05:30
Kirill Primak
faa31d123f render/vulkan: always finish buffer addon on texture destroy 2022-12-22 12:42:08 +03:00
Kirill Primak
0702eb9219 render/vulkan: remove NULL renderer check on texture destroy 2022-12-22 11:45:12 +03:00
Simon Ser
f12cdc53f3 render/egl: stop advertising render support for external-only formats
Some formats like YUV are only supported by Mesa for sampling, not
for rendering. However we always unconditionally added the INVALID
modifier to the list of render formats.

Check whether all modifiers are external-only, in that case, don't
add INVALID to the list of render formats.
2022-12-13 16:43:22 +00:00
Simon Ser
80074d95fb egl: consistently use EGLint for DMA-BUF format
EGL_EXT_image_dma_buf_import_modifiers [1] uses EGLint to carry
formats. We were casting it to int (which may not be the same
width) and uint32_t (which may change the value).

Instead, use EGLint everywhere.

[1]: https://registry.khronos.org/EGL/extensions/EXT/EGL_EXT_image_dma_buf_import_modifiers.txt
2022-12-12 23:08:48 +01:00
Simon Ser
32fc23a383 render/pixman: skip mask for opaque texture rendering 2022-12-08 16:53:14 +01:00
Simon Ser
86fc2199f8 build: unify naming for HAVE_* defines
We sometimes used HAS_, sometimes polluted the LIBINPUT_ namespace,
etc.
2022-12-06 22:39:45 +00:00
Simon Ser
cb6b1193c8 render/allocator/gbm: use internal_config
Removes a project argument.
2022-12-06 22:39:45 +00:00
Simon Ser
1bd0ba3949 render: simplify renderer_autocreate_with_drm_fd()
Use early returns to remove the !renderer checks.
2022-12-06 14:58:30 +00:00
Simon Ser
8456ac6fa9 render/vulkan: wait for DMA-BUF fences
The Vulkan spec doesn't guarantee that the driver will wait for
implicitly synchronized client buffers before texturing from them.
radv happens to perform the wait, but anv doesn't.

Fix this by extracting implicit fences from DMA-BUFs, importing
them into Vulkan as a VkSemaphore objects, and make the render pass
wait on these VkSemaphores.
2022-12-06 14:54:09 +00:00
Simon Ser
30219cf76b render/dmabuf: add dmabuf_export_sync_file() 2022-12-06 14:54:09 +00:00
Simon Ser
a37a22e344 render/vulkan: use initializers for VkSubmitInfo 2022-12-03 09:53:32 +01:00
Simon Ser
666c45bb29 render/vulkan: remove stage_cb conditional
We always have a stage_cb at this point.
2022-12-03 09:39:12 +01:00
Simon Ser
dd5be430b4 render/vulkan: remove pre_cb in vulkan_end()
pre_cb was an alias for stage_cb->vk. Let's just use a single
variable instead.

Additionally, early return when vulkan_record_stage_cb() fails.
We were crashing in vkCmdPipelineBarrier() if that happened.
2022-12-03 09:36:18 +01:00
Simon Ser
27e057319a render/vulkan: fix vkCmdClearAttachments validation error
Skip clears with an empty scissor.

Fixes the following validation error:

    00:00:09.734 [wlr] [render/vulkan/vulkan.c:61] Validation Error: [ VUID-vkCmdClearAttachments-rect-02682 ] Object 0: handle = 0x62600001b100, type = VK_OBJECT_TYPE_COMMAND_BUFFER; | MessageID = 0xadbd476f | CmdClearAttachments(): pRects[0].rect.extent.width is zero. The Vulkan spec states: The rect member of each element of pRects must have an extent.width greater than 0 (https://www.khronos.org/registry/vulkan/specs/1.3-extensions/html/vkspec.html#VUID-vkCmdClearAttachments-rect-02682) (VUID-vkCmdClearAttachments-rect-02682)
2022-12-03 09:30:30 +01:00
Simon Ser
d6ff20268b render/vulkan: fix comma 2022-12-02 23:26:34 +01:00
Simon Ser
56326c7998 render/vulkan: use initializers for VkImageMemoryBarrier 2022-12-02 23:26:01 +01:00
Simon Ser
79aea58834 render/vulkan: fix missing pSignalSemaphores for stage CB
We were filling VkTimelineSemaphoreSubmitInfoKHR.pSignalSemaphoreValues,
but we were missing VkSubmitInfo.pSignalSemaphores.

This was causing VkTimelineSemaphoreSubmitInfoKHR.pSignalSemaphoreValues
to be ignored. By chance, the render command buffer was using the
next timeline point, so we were waiting for that instead.
2022-12-02 19:28:13 +00:00
Simon Ser
47a038c90e render/vulkan: move VkDescriptorSetAllocateInfo down
Move it down, right before it's used.
2022-12-02 18:28:55 +00:00
Simon Zeni
4b3bbb0c4f render/drm_format_set: fill dst in wlr_drm_format_set_copy 2022-12-02 11:47:47 -05:00
Simon Ser
337ef33edc render/vulkan: make shm/dmabuf split clearer in wlr_vk_format_props
struct wlr_vk_format_props contains a mix of properties for shm
and dmabuf, and it's not immediately clear which fields are for
which kind of buffer. Use a nested struct to group the fields.
2022-12-02 15:18:59 +00:00
Simon Ser
db9d277614 render/vulkan: make vulkan_format_props_find_modifier() return value const 2022-12-02 15:18:59 +00:00
Simon Ser
b4ad4671dc render/vulkan: check for barrier array alloc failure 2022-12-02 15:14:27 +00:00
Simon Ser
bc7f8de842 render/vulkan: improve message on format prop error
Use a more appropriate message on
vkGetPhysicalDeviceImageFormatProperties2() error.
2022-12-02 16:10:42 +01:00
Simon Ser
338a9616bd render/vulkan: simplify vulkan_format_props_query() 2022-12-02 16:10:41 +01:00
Simon Ser
a3874cac6c render/vulkan: extract DMA-BUF format query to separate function 2022-12-02 16:09:58 +01:00