And use it to print the driver version at the start of
dawn_end2end_tests. This will help when figuring out issues
happening on CQ but not necessarily locally.
Bug: None
Change-Id: Ibdb9ab8cab53cc1e1cf8a807da53edeca616bed9
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/29602
Reviewed-by: Stephen White <senorblanco@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Reland after a refactor of how the extension handling work in the Vulkan
backend.
The original author is David Turner <david.turner.dev@gmail.com>.
Certain Vulkan ICDs (Intel ones notably) will compile SPIR-V
shaders with an liberal, compiler-selected, subgroup size (i.e.
either 8, 16 or 32). For more context, see [1].
This can be a problem for compute, when one shader stores data
in device memory using a subgroup-size dependent layout, to be
consumed by a another shader. Problems arise when the compiler
decides to compile both shaders with different subgroup sizes.
To work-around this, the VK_EXT_subgroup_size_control device
extension was introduced recently: it allows the device to
report the min/max subgroup sizes it provides, and allows
the Vulkan program to control the subgroup size precisely
if it wants to.
This patch adds support to the Vulkan backend to report and
enable the extension if it is available. Note that:
- This changes the definition of VulkanDeviceKnobs to
make room for the required pNext-linked chains of
extensions.
- A helper class, PNextChainBuilder is also provided in
UtilsVulkan.h to make it easy to build pNext-linked
extension struct chains at runtime, as required when
probing device propertires/features, or when
creating a new VkDevice handle.
- This modifies VulkanDeviceInfo::GatherDeviceInfo() to
use PNextChainBuilder to query extension features and properties
in a single call to vkGetPhysicalDevice{Properties,Features}2.
Apart from that, there is no change in behaviour in this CL.
I.e. a later CL might force a specific subgroup size for
consistency, or introduce a new API to let Dawn clients
select a fixed subgroup size.
[1] https://bugs.freedesktop.org/show_bug.cgi?id=108875
Bug: dawn:464
Change-Id: I01e5c28e7dac66f0a57bf35532eb192912b254fa
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/23201
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Similarly to the CL for Instance extensions, it makes each device
extension linked to an enum class and a bitset. Logic surrounding device
extensions is changed to take advantage of this to be more programmatic
and less error prone when adding support for a new extension.
Bug: dawn:457
Change-Id: Iecf623c40b890b7e00ba972d5eac0712866692b5
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/22941
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
At the moment each instance extension needs special handling
in 6+ places when added to the Vulkan backend. This is very
error-prone and makes it difficult to do changes in how they
are extensions are handled.
This CL makes instance extensions linked with an enum class
and a bitset to know which are available (instead of individual
booleans). A table of known extensions with more information like
`versionPromoted` so that they can be handled programmatically.
Bug: dawn:457
Change-Id: I266deb730eb2b7f3ab0ee7ada1b06ff9748a60e4
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/22940
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
This reverts commit 0357eed7de15e901d14f6149e91bfe03d86958de
and reland commit bdc05c3d5fef780382c3fd5e5ebcb14bd929819a.
The Vulkan-Loader has a bug where if the instance is created
with Vulkan 1.1 and not the promoted extensions, it will skip
emulation and if the ICD doesn't support Vulkan 1.1 nor the
extensions. Enable the promoted extensions, even when creating
a Vulkan 1.1 instance.
Original change's description:
> Check FP16 support on vulkan backend
>
> This patch check FP16 support on vulkan backend, and introduces
> the shader_float16 extension.
>
> BUG=dawn:426
> TEST=dawn_end2end_tests
>
> Change-Id: Ie09568a416ce9eb2c11afeede3e7da520550d5fb
> Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/21901
> Commit-Queue: Xinghua Cao <xinghua.cao@intel.com>
> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
> Reviewed-by: Jiawei Shao <jiawei.shao@intel.com>
> Reviewed-by: Kai Ninomiya <kainino@chromium.org>
> Reviewed-by: Austin Eng <enga@chromium.org>
Bug: chromium:1087896, dawn:426
Change-Id: I2c4465fb2fe957966b44d3e5840112219481c639
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/22781
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
This patch check FP16 support on vulkan backend, and introduces
the shader_float16 extension.
BUG=dawn:426
TEST=dawn_end2end_tests
Change-Id: Ie09568a416ce9eb2c11afeede3e7da520550d5fb
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/21901
Commit-Queue: Xinghua Cao <xinghua.cao@intel.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Jiawei Shao <jiawei.shao@intel.com>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
This layer replaces VK_LAYER_LUNARG_standard_validation that is now
deprecated.
Bug: dawn:388
Change-Id: I488c175b997434a8e22deecfba3fc89495969cbc
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/19722
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
The effect to the user is the same, the Dawn device gets lost. However
we need to make the difference internally because when the backend
device is lost we can clean up immediately. On the contrary on internal
errors, the backend device is still alive and processing commands so we
need to gracefully shut it down.
Bug: dawn:269
Change-Id: Ie13b33a4f9ac2e1f5f98b3723d83cf1c6205c988
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/17965
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
All Vulkan implementations on macOS have EXT_metal_surface available
(MoltenVk, Swiftshader and gfx-rs), and it matches webgpu.h's surface
creation from a CAMetalLayer more closely.
Bug: dawn:269
Change-Id: I313cd366b3509fb8930552406d150f08fc8e3666
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/17962
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
This reverts commit 4ae315b0d11882f341c22147672e1661bcb3b7d7.
Reason for revert: crbug.com/1059205
Bug: chromium:1059205
Original change's description:
> Vulkan: Report and enable subgroup size control device extension.
>
> Certain Vulkan ICDs (Intel ones notably) will compile SPIR-V
> shaders with an liberal, compiler-selected, subgroup size (i.e.
> either 8, 16 or 32). For more context, see [1].
>
> This can be a problem for compute, when one shader stores data
> in device memory using a subgroup-size dependent layout, to be
> consumed by a another shader. Problems arise when the compiler
> decides to compile both shaders with different subgroup sizes.
>
> To work-around this, the VK_EXT_subgroup_size_control device
> extension was introduced recently: it allows the device to
> report the min/max subgroup sizes it provides, and allows
> the Vulkan program to control the subgroup size precisely
> if it wants to.
>
> This patch adds support to the Vulkan backend to report and
> enable the extension if it is available. Note that:
>
> - The corresponding VkStructureType enum values and
> struct types are not rolled to the third-party Vulkan
> headers used by Dawn yet, so vulkan_platform.h has been
> modified to define them if necessary. This can be
> removed in the future when the Vulkan-Headers are
> updated in a different patch.
>
> - This modifies VulkanDeviceInfo::GatherDeviceInfo() to
> use VkGetPhysicalDevice{Properties2,Features2} if the
> VK_KHR_get_device_properties2 instance extension is
> available. Otherwise, the Vulkan 1.0 APIs
> VkGetPhysicalDevice{Properties,Features} are used instead
> (and it is assumed that no subgroup size control is
> possible).
>
> - This changes the definition of VulkanDeviceKnobs to
> make room for the required pNext-linked chains of
> extensions.
>
> - A helper class, PNextChainBuilder is also provided in
> UtilsVulkan.h to make it easy to build pNext-linked
> extension struct chains at runtime, as required when
> probing device propertires/features, or when
> creating a new VkDevice handle.
>
> Apart from that, there is no change in behaviour in this CL.
> I.e. a later CL might force a specific subgroup size for
> consistency, or introduce a new API to let Dawn clients
> select a fixed subgroup size.
>
> [1] https://bugs.freedesktop.org/show_bug.cgi?id=108875
>
> Change-Id: I524af6ff3479f25b0a8bb139a062fe632c826893
> Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/16020
> Reviewed-by: Austin Eng <enga@chromium.org>
> Reviewed-by: Corentin Wallez <cwallez@chromium.org>
> Commit-Queue: Corentin Wallez <cwallez@chromium.org>
TBR=cwallez@google.com,cwallez@chromium.org,enga@chromium.org,enga@google.com,david.turner.dev@gmail.com
# Not skipping CQ checks because original CL landed > 1 day ago.
Change-Id: I893d771d7effdf83685dda3edac8a08f98d2f6e5
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/16522
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Certain Vulkan ICDs (Intel ones notably) will compile SPIR-V
shaders with an liberal, compiler-selected, subgroup size (i.e.
either 8, 16 or 32). For more context, see [1].
This can be a problem for compute, when one shader stores data
in device memory using a subgroup-size dependent layout, to be
consumed by a another shader. Problems arise when the compiler
decides to compile both shaders with different subgroup sizes.
To work-around this, the VK_EXT_subgroup_size_control device
extension was introduced recently: it allows the device to
report the min/max subgroup sizes it provides, and allows
the Vulkan program to control the subgroup size precisely
if it wants to.
This patch adds support to the Vulkan backend to report and
enable the extension if it is available. Note that:
- The corresponding VkStructureType enum values and
struct types are not rolled to the third-party Vulkan
headers used by Dawn yet, so vulkan_platform.h has been
modified to define them if necessary. This can be
removed in the future when the Vulkan-Headers are
updated in a different patch.
- This modifies VulkanDeviceInfo::GatherDeviceInfo() to
use VkGetPhysicalDevice{Properties2,Features2} if the
VK_KHR_get_device_properties2 instance extension is
available. Otherwise, the Vulkan 1.0 APIs
VkGetPhysicalDevice{Properties,Features} are used instead
(and it is assumed that no subgroup size control is
possible).
- This changes the definition of VulkanDeviceKnobs to
make room for the required pNext-linked chains of
extensions.
- A helper class, PNextChainBuilder is also provided in
UtilsVulkan.h to make it easy to build pNext-linked
extension struct chains at runtime, as required when
probing device propertires/features, or when
creating a new VkDevice handle.
Apart from that, there is no change in behaviour in this CL.
I.e. a later CL might force a specific subgroup size for
consistency, or introduce a new API to let Dawn clients
select a fixed subgroup size.
[1] https://bugs.freedesktop.org/show_bug.cgi?id=108875
Change-Id: I524af6ff3479f25b0a8bb139a062fe632c826893
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/16020
Reviewed-by: Austin Eng <enga@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Previously we would always assume that if the driver supported a Vulkan
version it would also support extensions that were promoted in that
version. This is not a spec requirement, so instead try to load the core
entrypoints, and only if the version is not available, load the
extension entrypoints.
Also renames VulkanFunction members that are from promoted extension to
not have a vendor prefix.
Also tag the promoted extensions that are the same in a core version as
available when that core version is available. This simplifies checking
for features in the Vulkan backend.
Bug:
Change-Id: I0817c01b8838ba26070858abb0cbed030e3291df
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/16040
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: David Turner <digit@google.com>
This will enable fuzzing the Vulkan backend with randomly injected
errors to help ensure the backend properly handles all errors. It also
redefines VkResult in the dawn_native::vulkan namespace such that a
VkResult cannot be used unless it is explicitly wrapped.
Bug: dawn:295
Change-Id: I3ab2f98702a67a61afe06315658a9ab76ed4ccc3
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/14520
Commit-Queue: Austin Eng <enga@chromium.org>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
This CL adds a few Vulkan extensions to be used for importing
dma-bufs as VkImages.
In particular, we need:
- VK_EXT_external_memory_dma_buf
- VK_EXT_image_drm_format_modifier
BUG=chromium:996470
Change-Id: I7a3dfd0184177c756b07613fbfe140506f54584c
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/13783
Commit-Queue: Brian Ho <hob@chromium.org>
Reviewed-by: Austin Eng <enga@chromium.org>
This CL adds Fuchsia-specific implementations to the following
classes:
dawn_native::vulkan::external_memory::Service
dawn_native::vulkan::external_semaphore::Service
The implementation is based on two Fuchsia Vulkan extensions
that are provides by the vulkan_fuchsia_extras.h header (i.e.
are not upstreamed to Khronos yet, but used/provided by the
Fuchsia platform):
VK_FUCHSIA_external_memory
VK_FUCHSIA_external_semaphore
Their details are similar to VK_KHR_external_XXXX_fd, but
uses Zircon handles instead of file decriptors.
BUG=dawn:221
Change-Id: I48238bcf3193433970cbe200a84b86a67103a2f2
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/10963
Reviewed-by: Austin Eng <enga@chromium.org>
Commit-Queue: David Turner <digit@google.com>
Enable the Vulkan backend when building Dawn with
the Chromium build system for Fuchsia. To make this
work properly the following is required:
- Modify VulkanInfo.cpp and BackendVk.cpp to correctly probe
the Fuchsia swapchain layer and its layer extension, as well
as enabling them when creating a new VkInstance.
- Modify VulkanFunctions.cpp to load the Fuchsia swapchain
related extension for this platform only.
- Provide a small mock GLFW library for Fuchsia under
src/utils/Glfw3Fuchsia.cpp, since the upstream project
does not support this platform at all. Its purpose is
only to allow the creation of the right VulkanBinding
instance, which depends on the creation of a display
surface for latter swapchain creation.
- Add //third_party/fuchsia-sdk:vulkan_base and
//third_party/fuchsia-sdk:vulkan_validation as
data_deps of the libdawn_native_sources target in
order to ensure that the Fuchsia package created by
the build system will include the correct Vulkan
libraries (loader and validation layers).
This builds correctly, and both dawn_unittests and
dawn_end2end_tests will run on a real Fuchsia device
or inside the Fuchsia emulator, using either GPU
virtualization or a software-based renderer.
Note: dawn_unittests will also run inside QEMU, but
not dawn_end2end_tests, since the latter requires
proper GPU emulation which is not available in this
environment.
NOTE: All end2end tests pass using a device with
an "Intel HD Graphics 615 (Kaby Lake GT2)"
adapter. However:
- For some reason, a single test takes up
to 129 seconds to pass
(BufferSetSubDataTests.ManySetSubData/Vulkan).
- The test process crashes inside VkDestroyInstance(),
apparently inside the Fuchsia-specific imagepipe
layer (which implements swapchain support).
This is likely a bug in the layer itself, and
not Dawn.
Also, may end2end tests will crash when run inside
the Fuchsia emulator (which uses GPU virtualization
to talk to the host GPU). The crashes happen inside
libvulkan-goldfish.so, the emulator-specific Vulkan
ICD on this sytem. Not a Dawn bug either.
Bug=dawn:221
Change-Id: Id3598b673e8c6393f24db728b8da49fdde3cac76
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/8963
Commit-Queue: David Turner <digit@google.com>
Reviewed-by: Austin Eng <enga@chromium.org>
DeviceLost is a better name because we don't have a "context" as in
OpenGL.
Bug: dawn:153
Change-Id: Idf7fcb731a0e138f46a81fb31f0d11ff58b4fa96
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/10541
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
Commit-Queue: Austin Eng <enga@chromium.org>
Also adds GetGlobalInfo, similar to GetDeviceInfo, but for code that wants to query instance extensions or API version.
Bug: chromium:976495
Change-Id: Ia7131aa1780737ec2691cde0007acb86a120d307
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/9620
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
Commit-Queue: Kai Ninomiya <kainino@chromium.org>
Introduces pushDebugGroup, popDebugGroup, and insertDebugMarker implementations
for Vulkan and Metal using VK_EXT_debug_marker and XCode, respectively.
Bug: dawn:44
Change-Id: I0ae56c4d67aa832123f27a1fcdddf65746261e57
Reviewed-on: https://dawn-review.googlesource.com/c/4241
Commit-Queue: Brandon Jones <brandon1.jones@intel.com>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
This also changes VulkanInfo to gather info on backends / adapters
instead of the device, because all the info gathering can happen before
the device is created.
BUG=dawn:29
Change-Id: I9dc4412f494428f1ae589544d3adf76fe8b9a3a3
Reviewed-on: https://dawn-review.googlesource.com/c/3941
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
The backend had a "requiredInstanceExtensions" parameter to device
creation so that we could request the right instance extensions to
support swapchains. When we have dawn_native::Instance we don't want to
require this information before we can use the Vulkan backend. Instead
we make the backend always require all VkSurfaceKHR extensions. This
should be safe because these extensions basically enable copying a
struct into an opaque VkSurfaceKHR object and not much else.
BUG=dawn:29
Change-Id: I7fc5426f5770b65bb35f02793a1319eb0653782c
Reviewed-on: https://dawn-review.googlesource.com/c/3662
Reviewed-by: Stephen White <senorblanco@chromium.org>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
This is both something we should strive for for robustness of the
backends and in this case helped debug what the issue was on the
linux-dawn-rel builder that was failing all Vulkan tests.
BUG=dawn:19
Change-Id: Ibe9f5ad76b7766575bb74fe4993625cf82cdf13f
Reviewed-on: https://dawn-review.googlesource.com/c/2702
Commit-Queue: Corentin Wallez <cwallez@chromium.org>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
