Skip to content

[SYCL CI][Linux GPU RT Uplift] Uplift GPU RT version for Linux CI Process #2504

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

Jump to bottom
Merged
merged 16 commits into from
Sep 29, 2020
Merged

[SYCL CI][Linux GPU RT Uplift] Uplift GPU RT version for Linux CI Process #2504

merged 16 commits into from
Sep 29, 2020

Conversation

bb-sycl
Copy link
Contributor

@bb-sycl bb-sycl commented Sep 21, 2020

Uplift GPU RT version for Linux to 20.37.17906

@bb-sycl bb-sycl requested a review from bader as a code owner September 21, 2020 12:24
@DoyleLi
Copy link
Contributor

DoyleLi commented Sep 22, 2020

/summary:run

@vladimirlaz
Copy link
Contributor

@yinyangsx could you please remove XFAIL for passing tests on specific configurations and xfail SYCL::matrix_transpose_glb.cpp test (please create ticket to track fix of the test)?

@yinyangsx
Copy link

@yinyangsx could you please remove XFAIL for passing tests on specific configurations and xfail SYCL::matrix_transpose_glb.cpp test (please create ticket to track fix of the test)?

Ok, got it

@bb-sycl bb-sycl requested a review from a team as a code owner September 22, 2020 06:47
@yinyangsx yinyangsx force-pushed the private/bb-sycl/Lin_GPU_RT_Uplift_20.37.17906 branch from 15e85d4 to 295a39a Compare September 22, 2020 06:54
vladimirlaz
vladimirlaz previously approved these changes Sep 22, 2020
View reviewed changes
bader
bader requested changes Sep 22, 2020
View reviewed changes
@bader
Copy link
Contributor

bader commented Sep 22, 2020

This PR conflicts with #2486, which probably should go first.

@yinyangsx
Copy link

Hello @vladimirlaz and @yanfeng3721 , From the log of sycl-ubu-x64-pr, the Linux LIT test hangs machine. Could you please take a look?

@yinyangsx
Copy link

Hello @vladimirlaz and @yanfeng3721 , From the log of sycl-ubu-x64-pr, the Linux LIT test hangs machine. Could you please take a look?
Found SYCL :: usm/smemll.cpp hangs with this GPU version 20.37.17906 on Linux platform, already marked it as UNSUPPORTED.

@bader
Copy link
Contributor

bader commented Sep 23, 2020

Found SYCL :: usm/smemll.cpp hangs with this GPU version 20.37.17906 on Linux platform, already marked it as UNSUPPORTED.

This is not the best way to handle "hangs".
Marking the test "UNSUPPORTED" on linux, we loose coverage on other devices, which are expected to work.
Please, communicate this issue to GPU driver team ASAP and disable test execution on GPU device only by removing corresponding RUN command.

bader
bader requested changes Sep 23, 2020
View reviewed changes
@vladimirlaz
Copy link
Contributor

@yinyangsx coudl you please merge the latest changes from sycl branch to you PR to resolve conflict.

@yinyangsx
Copy link

@yinyangsx coudl you please merge the latest changes from sycl branch to you PR to resolve conflict.

@vladimirlaz , Yanfeng reverted his change due to no public windows GPU version, no conflict now.

@yinyangsx
Copy link

Pending this GPU version available in share folder.

bader
bader requested changes Sep 25, 2020
View reviewed changes
@bader
Copy link
Contributor

bader commented Sep 28, 2020

@bb-sycl, please, fix CI failures and address code review comments.

@bader
Copy link
Contributor

bader commented Sep 28, 2020

/summary:run

@yanfeng3721
Update warnings.cpp
4fba63e 
restore warnings.cpp
@yanfeng3721
Update smemll.cpp
e7a5d5c 
Restore smemll.cpp
@yanfeng3721
Update smemll.cpp
b1adbac 
restore smemll.cpp
@yanfeng3721
Update smem_varied.cpp
7ac1cf1 
restore smem_varied.cpp
@yanfeng3721
Update warnings.cpp
2dc719f 
Restore warnings.cpp
@yanfeng3721
Copy link
Contributor

/summary:run

bader
bader reviewed Sep 29, 2020
View reviewed changes
sycl/test/esimd/on-device/matrix_transpose_glb.cpp
@@ -10,6 +10,8 @@
// REQUIRES: gpu
// RUN: %clangxx-esimd -fsycl %s -o %t.out
// RUN: %ESIMD_RUN_PLACEHOLDER %t.out
// XFAIL: linux
Copy link
Contributor

Choose a reason for hiding this comment

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

@bb-sycl, please, file a tracker for this failure. I don't think it's expected.
@kbobrovs, FYI.

Copy link
Contributor

Choose a reason for hiding this comment

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

Thanks, @bader . Yes, this looks like a regression.

Copy link
Contributor

Choose a reason for hiding this comment

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

@bader , I have file a tracker.

@bader bader merged commit 0c8d46e into intel:sycl Sep 29, 2020
This was referenced Sep 29, 2020
Closed
Merged
bader added a commit to bader/llvm that referenced this pull request Sep 30, 2020
@bader
[NOT-FOR-MERGE]Revert "[BuildBot] Uplift GPU RT version to 20.37.17906 (
46b41d9 
intel#2504)"

This reverts commit 0c8d46e.

Just to check if new regressions are caused by the driver update.
bader added a commit to bader/llvm that referenced this pull request Oct 1, 2020
@bader
Revert "[NOT-FOR-MERGE]Revert "[BuildBot] Uplift GPU RT version to 20…
2179514 
….37.17906 (intel#2504)""

This reverts commit 46b41d9.
iclsrc pushed a commit that referenced this pull request Jun 6, 2024
@sys-ce-bb
Round up #elts of TypeVector when calculating memory size (#2504)
a109500 
Round up number of elements in a Vector to a power of 2 when calculating memory size. Memory size will be calculated as BaseType * bit_ceil(ComponentCount). The previous calculation already rounded 3 elements to 4 elements.

Signed-off-by: Lu, John <[email protected]>

Original commit:
KhronosGroup/SPIRV-LLVM-Translator@a6398f45674bb5d
jsji pushed a commit that referenced this pull request Jun 27, 2024
@sys-ce-bb
Round up #elts of TypeVector when calculating memory size (#2504)
6edbfe1 
Round up number of elements in a Vector to a power of 2 when calculating memory size. Memory size will be calculated as BaseType * bit_ceil(ComponentCount). The previous calculation already rounded 3 elements to 4 elements.

Signed-off-by: Lu, John <[email protected]>

Original commit:
KhronosGroup/SPIRV-LLVM-Translator@a6398f45674bb5d
kbenzie added a commit to kbenzie/intel-llvm that referenced this pull request Feb 17, 2025
@kbenzie
Merge pull request intel#2504 from Bensuo/neil-fix
45e3405 
[L0] Remove handle Translation in MCL if loader extension used
Chenyang-L pushed a commit that referenced this pull request Feb 18, 2025
@kbenzie
Merge pull request #2504 from Bensuo/neil-fix
a97789b 
[L0] Remove handle Translation in MCL if loader extension used
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@kbobrovs kbobrovs kbobrovs left review comments

@bader bader bader approved these changes

@vladimirlaz vladimirlaz vladimirlaz approved these changes

@DoyleLi DoyleLi Awaiting requested review from DoyleLi

@yanfeng3721 yanfeng3721 Awaiting requested review from yanfeng3721

@yinyangsx yinyangsx Awaiting requested review from yinyangsx

@dm-vodopyanov dm-vodopyanov Awaiting requested review from dm-vodopyanov

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
7 participants
@bb-sycl @DoyleLi @vladimirlaz @yinyangsx @bader @yanfeng3721 @kbobrovs