Skip to content

[WARP] Update warp LKG version to use 1.0.17 #671

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
bob80905 wants to merge 9 commits into
base: main
Choose a base branch
from
+6 −11
Open

[WARP] Update warp LKG version to use 1.0.17 #671

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
754776d
update LKG and other warp versions to use 1.0.17
bob80905 Jan 26, 2026
6dadf54
remove 2 warp xfails because test passes now and issues are fixed
bob80905 Jan 26, 2026
67b3a16
add Intel exclusion to xfail
bob80905 Jan 26, 2026
e3a330d
noop to run all tests
bob80905 Jan 27, 2026
b674333
filter on warp architecture
bob80905 Jan 27, 2026
d9f0646
remove debug print statement
bob80905 Jan 27, 2026
474b1ad
filter on arm64 instead of not x64
bob80905 Jan 27, 2026
414deac
Merge branch 'update_WARP_1.0.17' of https://github.com/bob80905/offl…
bob80905 Jan 27, 2026
95cf75f
add -Gis
bob80905 Jan 28, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion .github/workflows/windows-amd-clang-warp-preview-d3d12.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,4 +20,4 @@ jobs:
SKU: windows-amd
TestTarget: check-hlsl-clang-warp-d3d12
OffloadTest-branch: ${{ github.ref }}
LLVM-ExtraCMakeArgs: -DCMAKE_C_COMPILER=clang-cl -DCMAKE_CXX_COMPILER=clang-cl -DOFFLOADTEST_USE_CLANG_TIDY=On -DWARP_VERSION=1.0.15-preview
LLVM-ExtraCMakeArgs: -DCMAKE_C_COMPILER=clang-cl -DCMAKE_CXX_COMPILER=clang-cl -DOFFLOADTEST_USE_CLANG_TIDY=On -DWARP_VERSION=1.0.17
2 changes: 1 addition & 1 deletion .github/workflows/windows-amd-dxc-warp-preview-d3d12.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,4 +21,4 @@ jobs:
BuildType: Debug
TestTarget: check-hlsl-warp-d3d12
OffloadTest-branch: ${{ github.ref }}
LLVM-ExtraCMakeArgs: -DCMAKE_C_COMPILER=clang-cl -DCMAKE_CXX_COMPILER=clang-cl -DWARP_VERSION=1.0.15-preview
LLVM-ExtraCMakeArgs: -DCMAKE_C_COMPILER=clang-cl -DCMAKE_CXX_COMPILER=clang-cl -DWARP_VERSION=1.0.17
2 changes: 1 addition & 1 deletion cmake/modules/Warp.cmake
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,7 +24,7 @@ function(setup_warp version)
guess_nuget_arch(NUGET_ARCH)

if (version STREQUAL "LKG")
set(version "1.0.15")
set(version "1.0.17")
set(version_description "Latest Known Good for ${NUGET_ARCH} (${version})")
else ()
set(version_description "Custom (${version})")
Expand Down
5 changes: 3 additions & 2 deletions test/WaveOps/WaveActiveMax.test
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -66,7 +66,7 @@ DescriptorSets:
#--- end

# RUN: split-file %s %t
# RUN: %dxc_target -T cs_6_0 -Fo %t.o %t/source.hlsl
# RUN: %dxc_target -T cs_6_0 -Gis -Fo %t.o %t/source.hlsl
# RUN: %offloader %t/pipeline.yaml %t.o | FileCheck %s

# The behavior of this operation is consistent on Metal, so the test verifies that behavior.
Expand All @@ -87,7 +87,8 @@ DescriptorSets:
# DirectX driver implementations seem to match SPIR-V, except WARP, which does
# not treat -INF as an identity.

# XFAIL: DirectX && WARP
# XFAIL: arm64 && DirectX && WARP

Copy link
Member

@farzonl farzonl Jan 28, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Add an XFAIL for DirectX and Nvidia.
We are expecting [ inf, inf, inf, inf ] but Nvidia it is [ inf, 0, inf, inf ]. First try the GIS flag if that doesn't work we should look into filing a vendor bug.

Copy link
Contributor Author

@bob80905 bob80905 Jan 28, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this test is completely broken.
For one, this line Mix[TID.x % 8] = WaveActiveMax(Mix[TID.x % 8]); looks like it's accessing out of bounds memory.
Secondly, @llvm-beanz has had a PR to fix this specific test for a couple months now, here: #347

I think we should just update the relevant XFAIL, so as not to regress further, but leave this one out of it, so that it can be properly addressed separately.


# CHECK: Name: Nans
# CHECK-NEXT: Format: Float32
Expand Down
3 changes: 0 additions & 3 deletions test/WaveOps/WaveReadLaneAt.16.test
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -138,9 +138,6 @@ DescriptorSets:
# Bug https://github.com/llvm/offload-test-suite/issues/351
# XFAIL: Metal

# Bug tracked by https://github.com/llvm/offload-test-suite/issues/396
# XFAIL: DirectX && WARP

# Bug https://github.com/llvm/offload-test-suite/issues/532
# XFAIL: DirectX && QC

Expand Down
3 changes: 0 additions & 3 deletions test/WaveOps/WaveReadLaneFirst.fp16.test
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -313,9 +313,6 @@ DescriptorSets:
# Bug https://github.com/llvm/offload-test-suite/issues/627
# XFAIL: QC && DirectX

# Bug https://github.com/llvm/offload-test-suite/issues/433
# XFAIL: WARP

# RUN: split-file %s %t
# RUN: %dxc_target -enable-16bit-types -T cs_6_5 -Fo %t.o %t/source.hlsl
# RUN: %offloader %t/pipeline.yaml %t.o
Loading