Pyry Haulos [Thu, 7 Jul 2016 18:50:35 +0000 (11:50 -0700)]
Fix warnings in shader derivate tests
Pyry Haulos [Thu, 7 Jul 2016 18:46:50 +0000 (11:46 -0700)]
Update glslang to d02dc5d
Pyry Haulos [Thu, 7 Jul 2016 18:03:51 +0000 (11:03 -0700)]
Merge vulkan-cts-1.0 into vulkan-cts-1.0-dev
scygan [Thu, 16 Jun 2016 12:53:06 +0000 (14:53 +0200)]
Update glslang to (
f5dcdf01) "Front-end: Don't lower precision of bit operations taking highp inputs."
Fixes #382
Mika Isojarvi [Mon, 27 Jun 2016 18:24:26 +0000 (18:24 +0000)]
Merge branch '324-shader-texture-function-tests' into 'vulkan-cts-1.0-dev'
Add shader texture function tests
Fixes #324
See merge request !471
Mika Isojarvi [Fri, 24 Jun 2016 17:28:13 +0000 (17:28 +0000)]
Merge branch '395-convert-assembly-fix-arraystride' into 'vulkan-cts-1.0-dev'
Fix missing ArrayStride decorator.
Fix the missing ArrayStride decorator in OpSConvert and OpUConvert tests.
See merge request !578
Akos Dirner [Thu, 23 Jun 2016 12:14:23 +0000 (14:14 +0200)]
ShaderRenderDerivateTests: fix threshold calculation.
Akos Dirner [Tue, 21 Jun 2016 11:47:26 +0000 (13:47 +0200)]
Add more texture function tests to the exclude list, regenerate mustpass list.
Akos Dirner [Mon, 20 Jun 2016 12:10:06 +0000 (14:10 +0200)]
Add more texture function tests to the exclude list, regenerate mustpass list.
Pyry Haulos [Fri, 17 Jun 2016 16:10:51 +0000 (16:10 +0000)]
Merge branch '392-split-tests' into 'vulkan-cts-1.0-dev'
Fix issues with undefined levels in tess primitive discard tests
In certain cases less than 1 inner tessellation levels result in
implementation-dependent number of vertices to be generated. Relax
verification in those cases, and add separate tests that exercise only
valid (>= 1) inner levels.
Fixes #392
See merge request !592
Akos Dirner [Fri, 17 Jun 2016 15:01:16 +0000 (17:01 +0200)]
Exclude some failing tests from mustpass.
Pyry Haulos [Thu, 16 Jun 2016 20:50:55 +0000 (13:50 -0700)]
Fix issues with undefined levels in tess primitive discard tests
In certain cases less than 1 inner tessellation levels result in
implementation-dependent number of vertices to be generated. Relax
verification in those cases, and add separate tests that exercise only
valid (>= 1) inner levels.
Fixes #392
Mika Isojärvi [Thu, 16 Jun 2016 17:58:09 +0000 (10:58 -0700)]
Update Khronos mustpass list
Mika Isojarvi [Thu, 16 Jun 2016 17:53:02 +0000 (17:53 +0000)]
Merge branch '342-opinboundsaccesschain-assembly-tests' into 'vulkan-cts-1.0-dev'
Add OpInBoundsAccessChain tests.
See merge request !549
Mika Isojarvi [Thu, 16 Jun 2016 17:51:43 +0000 (17:51 +0000)]
Merge branch '296-derivate-function-tests' into '324-shader-texture-function-tests'
Added Derivate function tests.
I set target to branch 360-texture-gather-tests because this code depends on some ShaderRendercase modifications from the unmerged branch.
See merge request !538
Maciej Jesionowski [Thu, 16 Jun 2016 12:02:45 +0000 (14:02 +0200)]
Explicitly declare gl_PerVertex block
This fixes SPIR-V validation error in shaders generated from GLSL (not ES)
due to missing ClipDistance capability.
(cherry picked from commit
0a6fe1448b303a6a4e4ab48712eaf1c0b16d75b1)
Mika Isojarvi [Thu, 16 Jun 2016 17:40:02 +0000 (17:40 +0000)]
Merge branch 'mobica_fix_spirv_validation_errors' into 'vulkan-cts-1.0-dev'
Explicitly declare gl_PerVertex block
This fixes SPIR-V validation error in shaders generated from GLSL (not ES) due to missing ClipDistance capability.
I'm adding "for discussion" label to highlight this. Also not sure if we really need to make these fixes, or should this be dealt with someplace else.
Affected tests (at least those I noticed):
- dEQP-VK.draw
- dEQP-VK.binding_model
- dEQP-VK.pipeline.early_fragment
- dEQP-VK.pipeline.spec_constant
See merge request !590
Mika Isojarvi [Thu, 16 Jun 2016 17:37:46 +0000 (17:37 +0000)]
Merge branch 'issue_408' into 'vulkan-cts-1.0-dev'
Fixes OpSConvert and OpUConvert output buffer
The OpSConvert and OpUConvert incorrectly bound the same buffer as input
and output. This can lead to issues when they are of different size.
Fixes #408
See merge request !588
Pyry Haulos [Thu, 16 Jun 2016 16:01:47 +0000 (09:01 -0700)]
Fix VkDeviceSize -> size_t warning
Pyry Haulos [Thu, 16 Jun 2016 16:01:29 +0000 (09:01 -0700)]
Re-build Vulkan CTS mustpass
Gabor Abraham [Wed, 8 Jun 2016 13:26:52 +0000 (15:26 +0200)]
Fix missing ArrayStride decorator.
Maciej Jesionowski [Thu, 16 Jun 2016 12:02:45 +0000 (14:02 +0200)]
Explicitly declare gl_PerVertex block
This fixes SPIR-V validation error in shaders generated from GLSL (not ES)
due to missing ClipDistance capability.
Lars-Ivar Hesselberg Simonsen [Thu, 16 Jun 2016 10:59:44 +0000 (12:59 +0200)]
Fixes OpSConvert and OpUConvert output buffer
The OpSConvert and OpUConvert incorrectly bound the same buffer as input
and output. This can lead to issues when they are of different size.
Pyry Haulos [Wed, 15 Jun 2016 16:04:16 +0000 (16:04 +0000)]
Merge branch 'buffer-overflow-in-deqp-vk-draw-indirect_draw-404' into 'vulkan-cts-1.0-dev'
Fix for buffer overflow in dEQP-VK.draw.indirect_draw.*
Closes #404
See merge request !584
Akos Dirner [Wed, 15 Jun 2016 13:35:06 +0000 (15:35 +0200)]
Disable textureproj* tests which are affected by glslang issue (319).
Pyry Haulos [Tue, 14 Jun 2016 15:45:41 +0000 (15:45 +0000)]
Merge branch 'git_log_process_update' into 'vulkan-cts-1.0'
See merge request !577
Pyry Haulos [Tue, 14 Jun 2016 15:44:27 +0000 (15:44 +0000)]
Merge branch '396-component-mapping-fix' into 'vulkan-cts-1.0-dev'
Component mapping fix for rasterization tests.
It fixes the improper use of the getFormatComponentMapping() in the rasterization tests. (vulkan/vulkancts#396)
See merge request !573
Pyry Haulos [Tue, 14 Jun 2016 15:44:14 +0000 (15:44 +0000)]
Merge branch 'scygan-fix_image_properties_tests' into 'vulkan-cts-1.0'
Fix dEQP-VK.api.info.image_format_properties - TRANSIENT usage flag used with DMA flags
Spec:
```
VK_IMAGE_USAGE_TRANSIENT_ATTACHMENT_BIT indicates that the ....
If this is set, then bits other than VK_IMAGE_USAGE_COLOR_ATTACHMENT_BIT,
VK_IMAGE_USAGE_DEPTH_STENCIL_ATTACHMENT_BIT, and VK_IMAGE_USAGE_INPUT_ATTACHMENT_BIT must not be set.
```
The test tries the forbidden combination of usage flags, for example VK_IMAGE_USAGE_TRANSIENT_ATTACHMENT_BIT + VK_IMAGE_USAGE_TRANSFER_SRC
See merge request !570
Pyry Haulos [Tue, 14 Jun 2016 15:43:26 +0000 (15:43 +0000)]
Merge branch 'validation_fix_api_command_buffers' into 'vulkan-cts-1.0-dev'
dEQP-VK.api.command_buffers: remove bad test
I think two test cases were wrong -- they allocated zero command buffers which shouldn't be allowed. Layers treat it as an error and I filed a spec bug because it's not stated explicitly. I'm removing these two tests.
See #357 for details on validation task.
### Vulkan spec issue
- [Issue 292](https://gitlab.khronos.org/vulkan/vulkan/issues/292) - VkCommandBufferAllocateInfo valid usage
See merge request !525
Akos Dirner [Fri, 10 Jun 2016 12:11:47 +0000 (14:11 +0200)]
Fix SPIR-V validation error (missing ClipDistance capability) by declaring gl_PerVertex explicitly.
Daniel Hegedus [Thu, 2 Jun 2016 11:27:32 +0000 (13:27 +0200)]
check if sample count actually supported on image format
Daniel Hegedus [Thu, 2 Jun 2016 10:56:56 +0000 (12:56 +0200)]
refactor vulnerable sample count handling
Daniel Hegedus [Fri, 22 Apr 2016 13:23:42 +0000 (15:23 +0200)]
Added Derivate function tests.
John Corbally [Mon, 13 Jun 2016 15:27:04 +0000 (16:27 +0100)]
Fix for buffer overflow in dEQP-VK.draw.indirect_draw.* CTS tests
Pyry Haulos [Mon, 13 Jun 2016 17:27:47 +0000 (10:27 -0700)]
Fix a few size_t -> deUint32 conversion warnings
Akos Dirner [Mon, 13 Jun 2016 14:24:28 +0000 (16:24 +0200)]
Add texture function and texture gather tests to the mustpass and exclude lists.
Akos Dirner [Tue, 26 Apr 2016 20:13:53 +0000 (22:13 +0200)]
Add texture gather tests. Fixes #360.
Akos Dirner [Thu, 2 Jun 2016 15:22:15 +0000 (17:22 +0200)]
Fix sampler definition used for texelFetch tests.
Akos Dirner [Fri, 11 Mar 2016 15:53:35 +0000 (16:53 +0100)]
Add ShaderRenderTextureFunctionTests. Fixes #324.
Akos Dirner [Fri, 11 Mar 2016 15:49:20 +0000 (16:49 +0100)]
ShaderRender:
* add support for 2darray, 3d and cube samplers.
* derived classes can configure more parameters (e.g. render size, color format of result image, etc.) and they can use custom rendering (even multiple times).
Related to #324.
Pyry Haulos [Mon, 13 Jun 2016 17:18:30 +0000 (10:18 -0700)]
Re-build Vulkan CTS 1.0.1 mustpass
Pyry Haulos [Mon, 13 Jun 2016 17:07:58 +0000 (10:07 -0700)]
Merge branch 381-new-synchronization-tests into vulkan-cts-1.0-dev
Pyry Haulos [Mon, 13 Jun 2016 17:07:18 +0000 (10:07 -0700)]
Merge branch 334-resolve-image-tests into vulkan-cts-1.0-dev
Gabor Abraham [Fri, 10 Jun 2016 12:27:11 +0000 (14:27 +0200)]
Add ArrayStride decorator in OpTypeArray cases.
Gabor Abraham [Fri, 10 Jun 2016 11:45:16 +0000 (13:45 +0200)]
Fix OpInBoundsAccessChain return type in matrix cases.
Alexander Galazin [Wed, 8 Jun 2016 09:25:47 +0000 (11:25 +0200)]
Update git log command
Csaba Osztrogonác [Tue, 17 May 2016 14:02:34 +0000 (16:02 +0200)]
Add OpAtomicIAdd, OpAtomicISub, OpAtomicIIncrement and OpAtomicIDecrement assembly tests
Peter Siket [Mon, 6 Jun 2016 07:56:42 +0000 (09:56 +0200)]
Component mapping fix.
Szilard Ledan [Fri, 3 Jun 2016 14:30:10 +0000 (16:30 +0200)]
Fix after review and validation layer errors
Szilard Ledan [Mon, 29 Feb 2016 09:32:23 +0000 (10:32 +0100)]
Add tests for resolve images
Pyry Haulos [Thu, 2 Jun 2016 18:37:24 +0000 (18:37 +0000)]
Merge branch 'jekstrand-spirv-assembly-fixes' into 'vulkan-cts-1.0-dev'
spirv_assembly: Don't use booleans in SSBOs
The bool type in spirv isn't allowed in external storage. From the SPIR-V
spec for OpTypeBool:
> Values of this type can only be either true or false. There is no
> physical size or bit pattern defined for these values. If they are
> stored (in conjunction with OpVariable), they can only be used with
> logical addressing operations, not physical, and only with
> non-externally visible shader Storage Classes: Workgroup,
> CrossWorkgroup, Private, and Function.
Instead, we should use an integer and use OpSelect to convert from a
boolean to the expected 0/1 value the test wants.
See merge request !569
Pyry Haulos [Thu, 2 Jun 2016 18:36:08 +0000 (18:36 +0000)]
Merge branch 'jekstrand-rasterization-fixes' into 'vulkan-cts-1.0-dev'
rasterization: Use LOAD_OP_DONT_CARE for the resolve attachment
It used to use LOAD_OP_CLEAR for the resolve attachment but only provided
one clear value which led to an out-of-bounds read.
See merge request !571
Pyry Haulos [Thu, 2 Jun 2016 18:35:46 +0000 (18:35 +0000)]
Merge branch 'jekstrand-clearing-fixes' into 'vulkan-cts-1.0-dev'
image_clearing: Don't set a component mapping for the view to be cleared
From the Vulkan 1.0.12 spec for VkComponentMapping:
> This remapping must be identity for storage image descriptors, input
> attachment descriptors, and framebuffer attachments.
This fixes 6 CTS tests on the Intel Linux driver.
See merge request !568
Pyry Haulos [Thu, 2 Jun 2016 18:34:51 +0000 (18:34 +0000)]
Merge branch 'scygan-update-glslang' into 'vulkan-cts-1.0-dev'
Update glslang to (5b3f755) to include struct member decorations fix
This fixes #384
CC @mjesionowski
See merge request !561 and !558.
Jason Ekstrand [Thu, 2 Jun 2016 18:25:00 +0000 (11:25 -0700)]
spirv_assembly: Use deInt32 instead of VkBool32 for OpFUnord* tests
Jason Ekstrand [Wed, 1 Jun 2016 21:04:26 +0000 (14:04 -0700)]
rasterization: Use LOAD_OP_DONT_CARE for the resolve attachment
It used to use LOAD_OP_CLEAR for the resolve attachment but only provided
one clear value which led to an out-of-bounds read.
scygan [Wed, 18 May 2016 10:59:59 +0000 (12:59 +0200)]
Update glslang to (
22e0d414) to include struct member and in/out struct decorations fixes
This unblocks MR !558
Jason Ekstrand [Tue, 31 May 2016 22:49:29 +0000 (15:49 -0700)]
spirv_assembly: Don't use booleans in SSBOs
The bool type in spirv isn't allowed in external storage. From the SPIR-V
spec for OpTypeBool:
"Values of this type can only be either true or false. There is no
physical size or bit pattern defined for these values. If they are
stored (in conjunction with OpVariable), they can only be used with
logical addressing operations, not physical, and only with
non-externally visible shader Storage Classes: Workgroup,
CrossWorkgroup, Private, and Function."
Instead, we should use an integer and use OpSelect to convert from a
boolean to the expected 0/1 value the test wants.
scygan [Wed, 1 Jun 2016 13:55:20 +0000 (15:55 +0200)]
Fix dEQP-VK.api.info.image_format_properties - TRANSIENT usage flag used with DMA flags
Conflicts:
intel/negativeList.txt
Arkadiusz Sarwa [Fri, 6 May 2016 14:50:31 +0000 (16:50 +0200)]
Added synchronization basic tests
Gabor Abraham [Thu, 28 Apr 2016 15:27:22 +0000 (17:27 +0200)]
Create OpInBoundsAccessChain tests using OpCompositeInsert utility functions.
Maciej Jesionowski [Mon, 9 May 2016 13:08:03 +0000 (15:08 +0200)]
Add synchronization module
- Moved existing synchronization tests to a subgroup
Jason Ekstrand [Tue, 31 May 2016 18:42:02 +0000 (11:42 -0700)]
image_clearing: Don't set a component mapping for the view to be cleared
From the Vulkan 1.0.12 spec for VkComponentMapping:
"This remapping must be identity for storage image descriptors, input
attachment descriptors, and framebuffer attachments."
This fixes 6 CTS tests on the Intel Linux driver.
Pyry Haulos [Tue, 31 May 2016 16:31:25 +0000 (09:31 -0700)]
Merge vulkan-cts-1.0 into vulkan-cts-1.0-dev
Pyry Haulos [Tue, 31 May 2016 16:29:45 +0000 (09:29 -0700)]
Merge vulkan-cts-1.0 into vulkan-cts-1.0-dev
Pyry Haulos [Tue, 31 May 2016 16:27:30 +0000 (16:27 +0000)]
Merge branch 'memory-release-fix' into 'vulkan-cts-1.0'
Use TCU_CHECK instead of DE_ASSERT
DE_ASSERT is optimized away in release build. In this case it resulted
in the 'modifyXor' function not being called, causing the reference
memory to go out of sync with the actual memory. Follow up to the fix
for #353.
See merge request !566
Sivert Berg [Tue, 31 May 2016 12:37:57 +0000 (14:37 +0200)]
Use TCU_CHECK instead of DE_ASSERT
DE_ASSERT is optimized away in release build. In this case it resulted
in the 'modifyXor' function not being called, causing the reference
memory to go out of sync with the actual memory.
Mika Isojrvi [Sat, 28 May 2016 00:37:59 +0000 (00:37 +0000)]
Fixes to random memory mapping tests am:
223b8f6911 am:
c277f941dd
am:
2996c7f493
* commit '
2996c7f4933ff1272451e5545fc47f43f0cf367d':
Change-Id: Id1462b89bb79e8e3895742e5c5b1b600f472109b
Mika Isojrvi [Sat, 28 May 2016 00:37:57 +0000 (00:37 +0000)]
Fixes to random memory mapping tests am:
223b8f6911 am:
1dd022e33d
am:
1c7af2b568
* commit '
1c7af2b568ab1ef38199a2a9587efad98a259b26':
Fixes to random memory mapping tests
Change-Id: If5e5ff9189d03f3d09a4594f566b2795ce400cf9
Mika Isojärvi [Sat, 28 May 2016 00:11:52 +0000 (00:11 +0000)]
Fixes to random memory mapping tests am:
223b8f6911
am:
c277f941dd
* commit '
c277f941dd00ab3f9cd0d7aff3028f5c18a5ebb5':
Change-Id: I38c50df26568d40cfb928ab88194deec03e97554
Mika Isojärvi [Sat, 28 May 2016 00:09:09 +0000 (00:09 +0000)]
Fixes to random memory mapping tests am:
223b8f6911
am:
1dd022e33d
* commit '
1dd022e33d0ead6fa4e3761a12be2b013a79f91b':
Fixes to random memory mapping tests
Change-Id: I7761467d2692a1fd4832047d677de011f5591f7a
Mika Isojrvi [Sat, 28 May 2016 00:04:04 +0000 (00:04 +0000)]
Fixes to random memory mapping tests
am:
223b8f6911
* commit '
223b8f6911719df9863119197c12106db5dc0abc':
Fixes to random memory mapping tests
Change-Id: I15e35517169216977eaa6f62b2ce4cbf9bdfc45e
Mika Isojrvi [Sat, 28 May 2016 00:04:03 +0000 (00:04 +0000)]
Fixes to random memory mapping tests
am:
223b8f6911
* commit '
223b8f6911719df9863119197c12106db5dc0abc':
Fixes to random memory mapping tests
Change-Id: Ia94be0fac2b283df0954596d5fa006b3a8416181
Pyry Haulos [Fri, 27 May 2016 23:57:38 +0000 (23:57 +0000)]
Merge "De-duplicate utility code between some Vulkan test groups"
Mika Isojrvi [Fri, 27 May 2016 23:17:05 +0000 (23:17 +0000)]
Flush changes before invalidate in memory mapping tests am:
e73ac75e3c am:
0257d9458a
am:
508ae6d9f1
* commit '
508ae6d9f1d37335b4c286b6d46a40e4d2f10b8a':
Change-Id: I0134a429f69897af6d70b73f2e472260d87cc495
Mika Isojrvi [Fri, 27 May 2016 23:17:04 +0000 (23:17 +0000)]
Flush changes before invalidate in memory mapping tests am:
e73ac75e3c am:
46e8192cb1
am:
4305224b1c
* commit '
4305224b1c8d49fce6b323b6dbc0c31fe6bae01f':
Flush changes before invalidate in memory mapping tests
Change-Id: I8a66f7d2426992d8148315ac351fc77afb76fc46
Pyry Haulos [Fri, 27 May 2016 23:03:46 +0000 (16:03 -0700)]
Fix several issues in dEQP-VK.memory.mapping tests
This is a very manual cherry-pick of memory mapping test fixes from
development branch into the release branch.
Fixes #353
Mika Isojärvi [Fri, 27 May 2016 22:55:39 +0000 (22:55 +0000)]
Flush changes before invalidate in memory mapping tests am:
e73ac75e3c
am:
0257d9458a
* commit '
0257d9458a73c0fc24bec89ba60b353b312c8de3':
Change-Id: I8850f380084125039d5ef0edd34ea650877bacd3
Mika Isojrvi [Fri, 27 May 2016 22:53:08 +0000 (22:53 +0000)]
Flush changes before invalidate in memory mapping tests
am:
e73ac75e3c
* commit '
e73ac75e3c0ee3068e2366bf69a83c77f3dba268':
Flush changes before invalidate in memory mapping tests
Change-Id: Ibe83e08a3a62a9253ff50b8732f3432976b92e7b
Mika Isojärvi [Fri, 27 May 2016 22:48:03 +0000 (22:48 +0000)]
Flush changes before invalidate in memory mapping tests am:
e73ac75e3c
am:
46e8192cb1
* commit '
46e8192cb1d3d11640778548b94e114119349219':
Flush changes before invalidate in memory mapping tests
Change-Id: I2a830a172b614b41408e917dcb5aa66475ae955a
Mika Isojrvi [Fri, 27 May 2016 22:43:01 +0000 (22:43 +0000)]
Flush changes before invalidate in memory mapping tests
am:
e73ac75e3c
* commit '
e73ac75e3c0ee3068e2366bf69a83c77f3dba268':
Flush changes before invalidate in memory mapping tests
Change-Id: I8f12f5627d626e6df5953d996f2ce42d6ced39f7
Mika Isojärvi [Fri, 27 May 2016 21:21:49 +0000 (14:21 -0700)]
Fixes to random memory mapping tests
- Take nonCoherentAtomicSize into account.
- Track and handle correctly invalidating memory that has not
been flushed.
Bug:
29006045
Bug:
28939771
Change-Id: Ib114822aff0fd4bf4df5d84a2e17b393368bdd3c
Mika Isojärvi [Fri, 27 May 2016 21:19:20 +0000 (21:19 +0000)]
Merge "Take nonCoherentAtomSize into account in memory mapping tests am:
81fe3f5a2c am:
0a2cce9dfb" into nyc-mr1-dev-plus-aosp
am:
b19f87d64e
* commit '
b19f87d64e804b0984d97b87844db3dd9fb6b68d':
Change-Id: Ic3fdaba128ceb6bf3768c39142ba6623fa9ab13d
Mika Isojrvi [Fri, 27 May 2016 21:19:19 +0000 (21:19 +0000)]
Take nonCoherentAtomSize into account in memory mapping tests am:
81fe3f5a2c am:
a2685d421f
am:
10f018138c
* commit '
10f018138c56bcb7a06d9e91163acf8ab9f40957':
Take nonCoherentAtomSize into account in memory mapping tests
Change-Id: I7ab6a2b281817e97e3cfbdccffcd00cdac8965aa
Pyry Haulos [Fri, 27 May 2016 20:48:48 +0000 (13:48 -0700)]
De-duplicate utility code between some Vulkan test groups
draw, query_pool, and dynamic_state groups had a few identical
utilities. Move them under draw/.
Change-Id: I4e05606d75da6ac98ff3734604d080fe71ee36e8
Mika Isojärvi [Fri, 27 May 2016 21:05:20 +0000 (14:05 -0700)]
Flush changes before invalidate in memory mapping tests
Bug:
29006045
Change-Id: I697c69838bc75d0dfc7ac232888d60e1a0d80ded
Android Build Merger (Role) [Fri, 27 May 2016 20:39:27 +0000 (20:39 +0000)]
Merge "Take nonCoherentAtomSize into account in memory mapping tests am:
81fe3f5a2c am:
0a2cce9dfb" into nyc-mr1-dev-plus-aosp
Mika Isojärvi [Fri, 27 May 2016 20:39:22 +0000 (20:39 +0000)]
Take nonCoherentAtomSize into account in memory mapping tests am:
81fe3f5a2c
am:
0a2cce9dfb
* commit '
0a2cce9dfb5fb64418815926cfa2996c6809cd18':
Take nonCoherentAtomSize into account in memory mapping tests
Change-Id: I02c90ddd5226906fa05a5675bd726c7ddaa1d244
Mika Isojärvi [Fri, 27 May 2016 20:39:13 +0000 (20:39 +0000)]
Take nonCoherentAtomSize into account in memory mapping tests am:
81fe3f5a2c
am:
a2685d421f
* commit '
a2685d421fe61eecc101827fd9fa037d85a96f26':
Take nonCoherentAtomSize into account in memory mapping tests
Change-Id: I297f620e4dd3201eb22e115996fc188025f84086
Mika Isojrvi [Fri, 27 May 2016 20:36:38 +0000 (20:36 +0000)]
Take nonCoherentAtomSize into account in memory mapping tests
am:
81fe3f5a2c
* commit '
81fe3f5a2ce6b1da0740f68e8dfd512db6d2dc11':
Take nonCoherentAtomSize into account in memory mapping tests
Change-Id: I97a0fce0fe584343719caa4ff7d4329ab95ef9dd
Mika Isojrvi [Fri, 27 May 2016 20:36:38 +0000 (20:36 +0000)]
Take nonCoherentAtomSize into account in memory mapping tests
am:
81fe3f5a2c
* commit '
81fe3f5a2ce6b1da0740f68e8dfd512db6d2dc11':
Take nonCoherentAtomSize into account in memory mapping tests
Change-Id: I845a3073047c365670011eef70b8ec413f8de822
TreeHugger Robot [Fri, 27 May 2016 20:19:43 +0000 (20:19 +0000)]
Merge "Fix several narrowing conversions inside struct initializers"
TreeHugger Robot [Fri, 27 May 2016 20:18:36 +0000 (20:18 +0000)]
Merge "Initialize m_context in EGL thread clean up test class"
TreeHugger Robot [Fri, 27 May 2016 20:18:23 +0000 (20:18 +0000)]
Merge "Do not throw exception from destructior in tcu::Either tests"
Collin Baker [Fri, 27 May 2016 19:52:12 +0000 (19:52 +0000)]
Merge "Add test for vkResetDescriptorPool memory leak"
Mika Isojärvi [Tue, 24 May 2016 22:29:36 +0000 (15:29 -0700)]
Take nonCoherentAtomSize into account in memory mapping tests
Bug:
28939771
Change-Id: I02831f2a36b30b8f5b1c7e6710822f7b6f81a950
TreeHugger Robot [Fri, 27 May 2016 18:28:44 +0000 (18:28 +0000)]
Merge "Add Visual Studio 2015 generator to build scripts"
Pyry Haulos [Fri, 27 May 2016 15:41:29 +0000 (08:41 -0700)]
Change unnamed enum VK_ constants to defines and update API
Converting enum to target type causes narrowing conversion warnings on
MSVC 2015.
Change-Id: I51c41f7be0964de674c7b954469fbc523fddac5f
Pyry Haulos [Fri, 27 May 2016 16:21:19 +0000 (09:21 -0700)]
Initialize m_context in EGL thread clean up test class
This causes a crash when generating test case lists on null driver if
m_context ends up being some non-0 value.
Change-Id: I377af85a636a71366f4da8f4b00fccde979e91f3
Pyry Haulos [Fri, 27 May 2016 16:20:49 +0000 (09:20 -0700)]
Do not throw exception from destructior in tcu::Either tests
Change-Id: I5b63d0c3722d3856ac4b7827f8a4b46da2796856