|
a72ebeba
|
2018-06-28T11:16:58
|
|
Vulkan: Fix issue in DynamicBuffer
- We weren't keeping track of mSize correctly, causing some very
specific index buffer bindings to fail. The size returned when allocating
the buffer can be a bit different than the size requested, and that extra
space between mSize and the allocated size was not meant to be used,
causing errors in the validation layers.
Bug: angleproject:2580
Change-Id: I47eb7b8de6f4f657de14385b77ba6a459add599b
Reviewed-on: https://chromium-review.googlesource.com/1118607
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Luc Ferron <lucferron@chromium.org>
|
|
|
be30c4fb
|
2018-06-21T09:43:08
|
|
Vulkan: Framebuffer blit support for depth/stencil cases
Bug: angleproject:2643
Change-Id: Ib50e4051f5b3965c2a752cf2cd45d3470312cdcf
Reviewed-on: https://chromium-review.googlesource.com/1115370
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
|
|
8a4c49fb
|
2018-06-21T15:43:06
|
|
Vulkan: Refactor DynamicDescriptorPool.
Changes the set counting to use a free count by descriptor type. Also
changes the init to take a pool size count that more closely mirrors
Vulkan.
Will lead to a fix for descriptor set recreation.
Bug: angleproject:2678
Change-Id: I26816befec28ebb866fb0f3aaba27d2bc1d3be43
Reviewed-on: https://chromium-review.googlesource.com/1110661
Reviewed-by: Frank Henigman <fjhenigman@chromium.org>
Reviewed-by: Luc Ferron <lucferron@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>
|
|
|
50cf2be0
|
2018-06-15T09:46:57
|
|
Reformat some style in touched files.
This was triggered by running the code generation script.
Bug: angleproject:2665
Change-Id: Id639c78eb618182ee1859678590cf0f559b572c2
Reviewed-on: https://chromium-review.googlesource.com/1101564
Reviewed-by: Olli Etuaho <oetuaho@nvidia.com>
Reviewed-by: Luc Ferron <lucferron@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>
|
|
|
f6e160fa
|
2018-06-12T10:13:57
|
|
Vulkan: Don't use swizzle state when its not needed
Enables a bunch of dEQP tests in the functional.fbo.render.* namespace.
Bug: angleproject:2597
Change-Id: I1a06b335d5daf2987df52c460903081860887ce9
Reviewed-on: https://chromium-review.googlesource.com/1097596
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
|
|
|
4cc753e0
|
2018-06-13T13:25:33
|
|
Vulkan: Implement sampler arrays.
Bug: angleproject:2462
Change-Id: If9c9cb69624d6f9f0895f6883e1eed19a27e6cb4
Reviewed-on: https://chromium-review.googlesource.com/1089809
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Frank Henigman <fjhenigman@chromium.org>
|
|
|
c7918ceb
|
2018-06-13T13:25:31
|
|
Vulkan: Add PipelineLayout and DescriptorSetLayout caches.
This will be necessary to support more than one PipelineLayout per
instance of ANGLE. Sampler array handling requires different layouts
for different sampler uses.
Bug: angleproject:2462
Change-Id: I1d8b4919eed1a589002ad1898b05186f420061c7
Reviewed-on: https://chromium-review.googlesource.com/1089806
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Luc Ferron <lucferron@chromium.org>
|
|
|
3e313805
|
2018-06-11T14:44:33
|
|
Vulkan: Line loop edge case with ubyte indices
Bug: angleproject:2637
Change-Id: I2cec129ed1aa0e55e0dbf1d353caa65a51c528a3
Reviewed-on: https://chromium-review.googlesource.com/1096017
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
|
|
33f11fbc
|
2018-05-07T13:42:47
|
|
Initialize gl::Context state on the first call to MakeCurrent.
The constructor of gl::Context creates a bunch of zero and helper objects
but it is not current at this time. If the backend if not using virtualized
contexts this causes the created objects to be created on the wrong native
context.
BUG=angleproject:2464
Change-Id: I9718df87d0afeb08729920363d362d5f891061ed
Reviewed-on: https://chromium-review.googlesource.com/1048114
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Luc Ferron <lucferron@chromium.org>
Commit-Queue: Geoff Lang <geofflang@chromium.org>
|
|
|
c20b950e
|
2018-05-24T09:30:17
|
|
Vulkan: Fix clear of specific mip level was clearing all mip levels
Bug: angleproject:2502
Change-Id: Iffa012dce14584318c4dfd3d9b3a304291c9cebf
Reviewed-on: https://chromium-review.googlesource.com/1070666
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
|
|
|
574a6f25
|
2018-05-28T10:21:18
|
|
Vulkan: Fix edge case in DynamicBuffer causing wrong allocations
- on a very specific scenario when the next buffer to allocate was
fitting exactly in the allocated size. The condition was wrong and
needed to be >= instead of >.
Bug: angleproject:2562
Change-Id: Ide36bc3baadafdc1ffa40416b46998dc455caecf
Reviewed-on: https://chromium-review.googlesource.com/1075293
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
|
|
|
22695bf5
|
2018-05-22T15:52:08
|
|
Vulkan: Support cube mimaps generation
Bug: angleproject:2502
Change-Id: I953d99d04608cec04aad824b8b38f388ed1e4c2b
Reviewed-on: https://chromium-review.googlesource.com/1069544
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
|
|
|
316c6065
|
2018-05-29T10:49:45
|
|
Vulkan: Call GraphResource instead of GraphNode.
We don't need to use the CommandGraphNode class directly. This CL
consolidates our code so we never call the GraphNodes class directly.
Instead we call operations on GraphResource. This should simplify the
interaction with APIs from the various graph and dependency management
classes in the Vulkan back-end.
A new concept of 'starting' vs 'appending' commands is introduced.
Appending tries to avoid starting new command buffers when possible.
Should not change how the graphs are constructed, and mostly be a
refactoring change. There may be minor behaviour changes to some
commands.
Bug: angleproject:2539
Change-Id: Ia971e5cacb1164b9b3b22fa4a0a55b954d81f10e
Reviewed-on: https://chromium-review.googlesource.com/1052068
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Frank Henigman <fjhenigman@chromium.org>
|
|
|
64ac5d9e
|
2018-05-22T10:39:47
|
|
Vulkan: Quick rename of member variable in DynamicBuffer
The new name mNxtAllocationOffset instead of mNextWriteOffset makes more sense.
Also merged the 2 offset tracking vars since the user of DynamicBuffer
should know to flush or invalidate correctly.
Bug: angleproject:2480
Change-Id: If7a04b633794484ad88e73fd6f802fcfe0e3705a
Reviewed-on: https://chromium-review.googlesource.com/1069107
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Commit-Queue: Luc Ferron <lucferron@chromium.org>
|
|
|
534b00db
|
2018-05-18T08:16:53
|
|
Vulkan: readPixels improvement - use DynamicBuffer
Reintroduce the change that was previously reverted here:
https://chromium-review.googlesource.com/c/angle/angle/+/1064770
This includes a tentative fix the issue on Android that prompted the revert,
we need to call invalidate on the mapped memory range before we read it on
the host side.
Bug: angleproject:2480
Change-Id: Id637bafa2845628ae38483c6fc8e6d7f26ad2d3e
Reviewed-on: https://chromium-review.googlesource.com/1066229
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
|
|
|
a9ab0f37
|
2018-05-17T17:03:55
|
|
Vulkan: Fix DynamicBuffer usages, need to use specific alignment
The alignment used to allocate VkBuffers in the VkBuffer needs to be at
least the size of the nonCoherentAtomSize defined in the limits of the
VkDevicePhysicalProperties. The latest roll of the
vulkan-validation-layers added that check and caused a bunch of errors.
This is fixing them.
Bug: angleproject:2565
Change-Id: Ia2ad506dce7966adb6220c52ea891903922c47d0
Reviewed-on: https://chromium-review.googlesource.com/1064950
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
|
|
|
18e323ab
|
2018-05-11T16:54:17
|
|
D3D11: Fix out-of-range access with robust access.
When using a vertex buffer with DYNAMIC usage, with robust buffer
access enabled, we would sometimes read out-of-bounds when using very
large values for the index range. An unchecked signed addition would
overflow and lead to reading a negative offset.
Fix this problem by keeping the value size_t whenever possible. Also do
clamped casts when converting to a smaller values.
Also adds a regression test.
Bug: chromium:842028
Change-Id: Ie630ac857c6acfc0bace849a03eebfbaa2fbe89a
Reviewed-on: https://chromium-review.googlesource.com/1055928
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
|
|
|
90968360
|
2018-05-04T08:47:22
|
|
Vulkan: Incomplete textures support
2D multisample is out of scope since its ES 3 only.
Bug: angleproject:2499
Change-Id: Id5f81d713a2882ba2a91b7d3f281d71a3e9289f4
Reviewed-on: https://chromium-review.googlesource.com/1046786
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
|
|
|
61859817
|
2018-05-09T14:17:39
|
|
Vulkan: Fix use-after-free with DynamicBuffer.
The implementation of DynamicBuffer before my changes could have some
issues in the following use case:
- Allocate buffer 1 for Texture 1 (with size as big as the full buffer
size)
- Allocate buffer 2 for Texture 2 (triggers creation of a new underlying
BufferVk and releases the buffer 1 to the Renderer)
- Render with Texture 2 (texture 1 hasn't been flushed yet)
- swap buffers (causes garbage in the renderer to be cleaned up)
- Try rendering with Texture 1, and you'll get an error stating that
the buffer we're trying to copy is not valid (because its already been
freed).
This set of changes:
- Add a new test that specifically triggers this case.
- enables the texture.filtering.cube* tests in dEQP.
- Fixes the issue by adding a manual releasing pattern of the buffers
in DynamicBuffer.
Bug: angleproject:2505
Change-Id: I207ce4a694016766f008cca67d82b252f460e0df
Reviewed-on: https://chromium-review.googlesource.com/1052551
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Luc Ferron <lucferron@chromium.org>
|
|
|
a932b6b5
|
2018-05-01T07:53:17
|
|
Vulkan: Fix in DynamicBuffer, allocating too many buffers for no reason
-Enables more dEQP tests in functional.texture.filtering
-texture.filtering.cube.* tests don't work yet, another CL will be
targeted at fixing that after this one.
Bug: angleproject:2505
Change-Id: Ic634e0deafa54e8e8ebd0efbdae14263ffdbcef0
Reviewed-on: https://chromium-review.googlesource.com/1037183
Reviewed-by: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Luc Ferron <lucferron@chromium.org>
|
|
|
33e05bab
|
2018-04-23T15:12:34
|
|
Vulkan: Bugfix in TextureVk::setSubImage and DynamicBuffer
Bug 1) The offset wasn't plumbed through for setSubImage.
Bug 2) The DynamicBuffer allocation sometime allocates a bit more
than requested, but we were using the size requested as the next offset
instead of the actual allocated size. This could get us in a situation
in certain corner cases where the next allocation would be done on the
said buffer instead of using a new allocation as it should.
Also enables a bunch of new texture_specification_* tests that were
unable to run successfully without these 2 bug fixes.
Found a weird issue on WIN AMD only and suppressed these tests for now.
Will investigate part of the same bug number as a separate change.
Bug: angleproject:2495
Bug: angleproject:2492
Change-Id: I490b1bf2d1795b7a1033365e29eac12a8bc50bff
Reviewed-on: https://chromium-review.googlesource.com/1024380
Reviewed-by: Geoff Lang <geofflang@chromium.org>
Commit-Queue: Luc Ferron <lucferron@chromium.org>
|
|
|
66410530
|
2018-04-20T12:47:45
|
|
Vulkan: Textures mipmaps support
Note that generate mipmaps is not yet supported, but uploading mipmaps
is. This also enables 64 tests in dEQP gles2 to validate the changes.
While trying to enable tests in functional.texture.mipmap.*, I found
an issue where the graph ends up in an invalid state and triggers and
assert in onReadResource in ContextVk.cpp:188. It seems like an
unrelated bug to mip maps, I will investigate to enable these tests
separately in a following CL.
Bug: angleproject:2479
Change-Id: If51776d8ef3d994bee620d6a1cf982bb51838ff0
Reviewed-on: https://chromium-review.googlesource.com/1022232
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
|
|
|
0cb6dc4c
|
2018-04-16T10:36:39
|
|
Vulkan: Fix texture descriptor set alloc count.
We were reserving half the required descriptor sets for our pool. This
fixes the counting and ensures we won't regress by adding a test.
Also enables the cube map texture sample, and removes an UNIMPLEMENTED
warning that was spurious.
Bug: angleproject:2318
Change-Id: I371cd7c5b42e1ce980cce7bb0ef04885db72b614
Reviewed-on: https://chromium-review.googlesource.com/1014165
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
|
|
|
a4fa9c27
|
2018-04-13T07:00:56
|
|
Vulkan: drawElements with GL_LINE_LOOP and an offset
Also enables a test in LineLoopTest that validates this case.
Bug: angleproject:2473
Change-Id: Icb4c5735c11be40cdeceaa051f5a5cef33fd22c6
Reviewed-on: https://chromium-review.googlesource.com/1011669
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
|
|
|
20fa8d5c
|
2018-04-15T10:09:32
|
|
Vulkan: Implement cube map texture sampling.
This changes the TextureVk class to have a queue of staging
copies, instead of a single copy at a time. This will allow
us to upload multiple sub regions of a texture image at once
when we need to resolve the Image at sampling time.
Enables the remainder of the texture.format tests, and all
non-mipped tests in the texture.size tests. Also updates the
expectations with the bug number for mipmap support.
Bug: angleproject:2318
Change-Id: Ie55f8f2dc24d7b133ad735e37f0f78937f022b5b
Reviewed-on: https://chromium-review.googlesource.com/980775
Reviewed-by: Luc Ferron <lucferron@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>
|
|
|
a912046d
|
2018-04-12T13:11:03
|
|
Vulkan: DrawElements with line loops client side memory support
- Also enables 6 new tests in LineLoopTests.cpp in angle_end2end
Bug: angleproject:2458
Change-Id: I4aec12b0ac780e81e6811f1199a5acaf17d9b982
Reviewed-on: https://chromium-review.googlesource.com/1010411
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Luc Ferron <lucferron@chromium.org>
|
|
|
983c429f
|
2018-04-10T13:05:45
|
|
Vulkan: Lineloops edge base bugfix and new tests
The dynamic buffer we are using in the LineLoopHelper wasn't able
to support switching between different allocation sizes. Fix this by
simply using a min alignment of the maximum allocation size we can
reach.
Adds 2 new tests to validate these calls in StateChangeTest.cpp
Bug: angleproject:2458
Change-Id: I9d224e7dcfcd7627010832ca30dd9e1b9eceea4e
Reviewed-on: https://chromium-review.googlesource.com/1007335
Commit-Queue: Luc Ferron <lucferron@chromium.org>
Reviewed-by: Jamie Madill <jmadill@chromium.org>
|
|
|
22f12fed
|
2018-04-08T14:23:40
|
|
Vulkan: Rename vk::LineLoopHelper.
This more closely follows the general pattern laid out by the naming
in vk_helpers.h. It also changes the dynamic buffer that the helper
wraps to be stored by-value since the header include order problem is
fixed.
Bug: angleproject:2318
Change-Id: I1de9e1edee2125d3afd490b4f9c99cf70c61215c
Reviewed-on: https://chromium-review.googlesource.com/1001654
Reviewed-by: Yuly Novikov <ynovikov@chromium.org>
Reviewed-by: Luc Ferron <lucferron@chromium.org>
Commit-Queue: Jamie Madill <jmadill@chromium.org>
|
|
|
9cceac42
|
2018-03-31T14:19:16
|
|
Vulkan: Update resource dependency semantics.
This removes passing the Serial around to several methods, so that
dependency management is a bit more automatic.
This makes life a bit easier when dealing with state updates when
resources are in use by Vulkan.
The FramebuffeVk no longer stores an extra serial of the last draw,
instead it will trigger creation of a new writing node on a state
change update.
Bug: angleproject:2318
Change-Id: Ie58ec66e6e8644ba4d402c509255c3795d363dd3
Reviewed-on: https://chromium-review.googlesource.com/985201
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Frank Henigman <fjhenigman@chromium.org>
|
|
|
6c7ab7fe
|
2018-03-31T14:19:15
|
|
Vulkan: Reorganize helper classes.
This renames ResourceVk to vk::CommandGraphResource, which should help
clarify its usage. This also moves LineLoopHandler, ImageHelper, and
the DynamicBuffer and DynamicCommandPool classes into a new vk_helpers
module. This file contains helper classes that manage other resources.
Also this makes DynamicBuffer and DynamicDescriptorPool no longer
inherit from CommandGraphResource. In the future, only Impl objects
will be allowed to be graph resources.
Bug: angleproject:2318
Change-Id: I0fa23da2ac853d90f3c822547a4a314f247cc757
Reviewed-on: https://chromium-review.googlesource.com/985200
Commit-Queue: Jamie Madill <jmadill@chromium.org>
Reviewed-by: Frank Henigman <fjhenigman@chromium.org>
|