Page MenuHomePhabricator
Feed Advanced Search

Fri, Jun 11

ashi1 committed rGf2cc0427b13e: [HIP] Fix --hip-version flag with 0 as component (authored by ashi1).
[HIP] Fix --hip-version flag with 0 as component
Fri, Jun 11, 9:26 AM
ashi1 closed D104062: [HIP] Fix --hip-version flag with 0 as component.
Fri, Jun 11, 9:25 AM · Restricted Project
ashi1 updated the diff for D104062: [HIP] Fix --hip-version flag with 0 as component.

Fixed to yaxunl's comments, using ~0U , and allowing major only component.

Fri, Jun 11, 7:53 AM · Restricted Project

Thu, Jun 10

ashi1 added inline comments to D104062: [HIP] Fix --hip-version flag with 0 as component.
Thu, Jun 10, 2:29 PM · Restricted Project
ashi1 requested review of D104062: [HIP] Fix --hip-version flag with 0 as component.
Thu, Jun 10, 2:28 PM · Restricted Project

Wed, Jun 2

ashi1 abandoned D100404: Add Global support for #pragma clang attributes.

@palves was able to resolve his request following @aaron.ballman's proposal. Closing this patch.

Wed, Jun 2, 10:47 AM

Fri, May 21

ashi1 added inline comments to D100794: [HIP] Support overloaded math functions for hipRTC.
Fri, May 21, 9:09 AM · Restricted Project

May 14 2021

ashi1 added a comment to D102508: [HIP] Add test libstd_functional.

Looks good, let me verify on HIP builder.

It depends on https://reviews.llvm.org/D102507

May 14 2021, 9:30 AM
ashi1 added a comment to D102508: [HIP] Add test libstd_functional.

Looks good, let me verify on HIP builder.

May 14 2021, 9:26 AM

May 13 2021

ashi1 committed rGa249ffa42137: [HIP] Clean up llvm intrinsics using __asm (authored by ashi1).
[HIP] Clean up llvm intrinsics using __asm
May 13 2021, 11:56 AM
ashi1 closed D102427: [HIP] Clean up llvm intrinsics using __asm.
May 13 2021, 11:56 AM · Restricted Project
ashi1 requested review of D102427: [HIP] Clean up llvm intrinsics using __asm.
May 13 2021, 10:59 AM · Restricted Project
ashi1 committed rG6a67e05a26eb: [HIP] Add __builtin_amdgcn_groupstaticsize (authored by ashi1).
[HIP] Add __builtin_amdgcn_groupstaticsize
May 13 2021, 8:51 AM
ashi1 closed D102403: [HIP] Add __builtin_amdgcn_groupstaticsize.
May 13 2021, 8:51 AM · Restricted Project
ashi1 requested review of D102403: [HIP] Add __builtin_amdgcn_groupstaticsize.
May 13 2021, 7:45 AM · Restricted Project

Apr 30 2021

ashi1 added a comment to D100060: [zorg] Add HIP builder script.

@tra have you seen this exception in your CUDA buildbots before?
https://lab.llvm.org/staging/#/builders/152/builds/3

Sorry, I haven't seen it before and Can't tell what exactly is the bot unhappy about.
It says unsupported operand type(s) for %: 'WithProperties' , but it's not clear why.

Apr 30 2021, 10:15 AM

Apr 29 2021

ashi1 committed rZORGb766b0b8afb3: [zorg] Fix AnnotatedBuilder.py extra_args bug (authored by ashi1).
[zorg] Fix AnnotatedBuilder.py extra_args bug
Apr 29 2021, 3:26 PM
ashi1 closed D101575: [zorg] Fix AnnotatedBuilder.py extra_args bug.
Apr 29 2021, 3:26 PM
ashi1 requested review of D101575: [zorg] Fix AnnotatedBuilder.py extra_args bug.
Apr 29 2021, 2:45 PM
ashi1 added a comment to D100060: [zorg] Add HIP builder script.

@tra have you seen this exception in your CUDA buildbots before?
https://lab.llvm.org/staging/#/builders/152/builds/3

Apr 29 2021, 2:19 PM
ashi1 updated the diff for D100404: Add Global support for #pragma clang attributes.

Revised to @palves' comments.

Apr 29 2021, 8:40 AM

Apr 26 2021

ashi1 updated the diff for D100404: Add Global support for #pragma clang attributes.

Revised to Artem's comments.

Apr 26 2021, 12:02 PM

Apr 23 2021

ashi1 updated the diff for D100404: Add Global support for #pragma clang attributes.

Added documentation changes.

Apr 23 2021, 11:40 AM
ashi1 committed rZORGc531f51254be: [zorg] Add buildbot for HIP (authored by ashi1).
[zorg] Add buildbot for HIP
Apr 23 2021, 7:11 AM
ashi1 closed D99894: [zorg] Add buildbot for HIP.
Apr 23 2021, 7:11 AM

Apr 22 2021

ashi1 added a comment to D100404: Add Global support for #pragma clang attributes.

Please review again, I have changed the patch description, and revised to the comments.

This new patch will accept #pragma clang attribute (...) without push/pop, and store the filter details into a global attribute group.

Thank you! I really like the "#pragma diagnostics" parallel, makes it even easier to justify.

I've tested the patch against the GDB testsuite, using the global #pragma clang attribute (...) form, and it worked against the couple testcases I tried.

(Running the whole testsuite will take a while, but I don't expect problems, if one test works, others should too.)

Noobie question -- I didn't see any change to the documentation in the patch. Is that a separate repo?

Apr 22 2021, 12:57 PM
ashi1 updated the diff for D99894: [zorg] Add buildbot for HIP.

Revised to @gkistanova 's comments.

Apr 22 2021, 12:39 PM
ashi1 updated subscribers of D99894: [zorg] Add buildbot for HIP.
Apr 22 2021, 12:07 PM
ashi1 added a comment to D100404: Add Global support for #pragma clang attributes.

Please review again, I have changed the patch description, and revised to the comments.

Apr 22 2021, 9:00 AM
ashi1 updated the diff for D100060: [zorg] Add HIP builder script.

Moving this script to annotated/ directory. The AnnotatedBuilder.py now support running bash scripts.

Apr 22 2021, 8:10 AM
ashi1 updated the diff for D99894: [zorg] Add buildbot for HIP.

Hi @gkistanova, I've changed the buildbot to use a local relative path (in annotated/hip-build.sh dir).
I will keep it as a local uncommitted script during bring-up. D100666 was submitted, and script_interpreter="python" is disabled for this bash script.

Apr 22 2021, 8:00 AM
ashi1 committed rZORG9745babbea2f: [zorg] Support other relative scripts in AnnotatedBuilder (authored by ashi1).
[zorg] Support other relative scripts in AnnotatedBuilder
Apr 22 2021, 7:49 AM
ashi1 closed D100666: [zorg] Support other relative scripts in AnnotatedBuilder.
Apr 22 2021, 7:49 AM

Apr 21 2021

ashi1 updated the diff for D100404: Add Global support for #pragma clang attributes.
Apr 21 2021, 2:35 PM
ashi1 added a comment to D100404: Add Global support for #pragma clang attributes.

I agree that the no_pop variant may not be what we want, so removing it.

Apr 21 2021, 2:24 PM
ashi1 updated the summary of D100666: [zorg] Support other relative scripts in AnnotatedBuilder.
Apr 21 2021, 10:07 AM
ashi1 updated the diff for D100666: [zorg] Support other relative scripts in AnnotatedBuilder.

Revised to Artem's comments, moved --jobs to extra_args when are unspecified.

Apr 21 2021, 10:06 AM
ashi1 added inline comments to D100666: [zorg] Support other relative scripts in AnnotatedBuilder.
Apr 21 2021, 9:58 AM

Apr 20 2021

ashi1 updated the diff for D100666: [zorg] Support other relative scripts in AnnotatedBuilder.
Apr 20 2021, 1:20 PM
ashi1 added inline comments to D100666: [zorg] Support other relative scripts in AnnotatedBuilder.
Apr 20 2021, 1:20 PM
ashi1 added inline comments to D100666: [zorg] Support other relative scripts in AnnotatedBuilder.
Apr 20 2021, 10:29 AM
ashi1 added a comment to D100404: Add Global support for #pragma clang attributes.

If Clang supported a "global" scope, then you would be able to just use the "#pragma clang attribute (...)" syntax when you don't want to pop. Seems like the natural solution.

Hmm, the problem I see with that is that you have to know that nobody else push-ed a #pca scope that you're inadvertently adding to. I'd prefer having a syntax that always added to the "global scope" regardless of the #pca context, which is what no_pop is effectively doing. I think I'd also be okay with downgrading the EOF error into a warning (which you could suppress in the -include file), assuming that we still warned when the #pca comes from a macro declared in a system header.

Apr 20 2021, 10:19 AM
ashi1 updated the diff for D100404: Add Global support for #pragma clang attributes.

Fix unrelated pre-merge test failures. Something changed upstream, unrelated to this patch.

Apr 20 2021, 8:47 AM
ashi1 updated the diff for D100404: Add Global support for #pragma clang attributes.

Fixed formatting issues.

Apr 20 2021, 7:26 AM
ashi1 added inline comments to D100666: [zorg] Support other relative scripts in AnnotatedBuilder.
Apr 20 2021, 7:07 AM
ashi1 updated the diff for D100666: [zorg] Support other relative scripts in AnnotatedBuilder.
Apr 20 2021, 7:05 AM
ashi1 added a comment to D100060: [zorg] Add HIP builder script.

I may be moving this patch and combining with D99894 , since we plan to add this to the public repo.

Apr 20 2021, 6:56 AM
ashi1 added a comment to D100404: Add Global support for #pragma clang attributes.

If no additional opinions, I think this current patch with no_pop serves the purpose and can be committed, while we discuss if we want to add global scope. @palves, does that work for you?

Apr 20 2021, 6:46 AM

Apr 16 2021

ashi1 added a comment to D99894: [zorg] Add buildbot for HIP.

No. Cuda builders got that in the code base as a temporary solution while they are working on their annotated script. Exactly the same arguments as yours. And I see that being confusing, and don't think it worth it after all.

To get it working in the production you need to add the support for shell scripts anyway. With that in place you could put your script in the right place and keep working on that while your builder is staged. It seems the right thing to do.

Please feel free to ask if you have questions or will need help with adding the shell script support.

Apr 16 2021, 11:17 AM
ashi1 requested review of D100666: [zorg] Support other relative scripts in AnnotatedBuilder.
Apr 16 2021, 11:16 AM

Apr 15 2021

ashi1 updated subscribers of D100404: Add Global support for #pragma clang attributes.

Ping arphaman and aaron.ballman. Also adding @erik.pilkington, who introduced namespaces to clang attribute, for opinions.

Apr 15 2021, 2:05 PM

Apr 14 2021

ashi1 added a comment to D99894: [zorg] Add buildbot for HIP.

@gkistanova , is it okay to submit this with a full system path temporarily? /buildbot/hip-build.sh? Or do we have to switch to a python script?

Apr 14 2021, 11:14 AM

Apr 13 2021

ashi1 committed rTd590d0b156e4: Reapply [test-suite] Add HIP Tests to External (authored by ashi1).
Reapply [test-suite] Add HIP Tests to External
Apr 13 2021, 2:38 PM
ashi1 committed rT116d17093f74: [test-suite] Add HIP Tests to External (authored by ashi1).
[test-suite] Add HIP Tests to External
Apr 13 2021, 2:38 PM
ashi1 added a reverting change for D99997: [test-suite] Add HIP Tests to External: rT42674b53b826: Revert "[test-suite] Add HIP Tests to External".
Apr 13 2021, 2:38 PM
ashi1 committed rT42674b53b826: Revert "[test-suite] Add HIP Tests to External" (authored by ashi1).
Revert "[test-suite] Add HIP Tests to External"
Apr 13 2021, 2:38 PM
ashi1 closed D99997: [test-suite] Add HIP Tests to External.
Apr 13 2021, 2:38 PM
ashi1 updated the diff for D99997: [test-suite] Add HIP Tests to External.

Address Artem's comments.

Apr 13 2021, 2:01 PM
ashi1 added inline comments to D99997: [test-suite] Add HIP Tests to External.
Apr 13 2021, 1:49 PM
ashi1 abandoned D98201: [CUDA][HIP] Add #pragma clang force_cuda_device_globals {begin,end}.

Closing this revision, I have a patch to add no_pop variant of #pragma clang attribute push.
https://reviews.llvm.org/D100404
Alternatively, we could look into making no pop the default.

Apr 13 2021, 11:54 AM
ashi1 requested review of D100404: Add Global support for #pragma clang attributes.
Apr 13 2021, 11:53 AM

Apr 12 2021

ashi1 added a comment to D98201: [CUDA][HIP] Add #pragma clang force_cuda_device_globals {begin,end}.

Allowing pragma push to be unmatched, maybe with an explicit option to enable it, would probably be less controversial than adding a new pragma that duplicates existing functionality.

To be clear, the pragma that I mentioned allows unmatching -- #pragma clang force_cuda_host_device begin -- is a preexisting Clang pragma, not the one proposed by this review. It's highly inconsistent for one pragma to error out when unmatched, while the other doesn't. Would you suggest that the pre-existing #pragma clang force_cuda_host_device begin should error out when unmatched?

No, what I'm saying is that we can allow #pragma clang attribute push to be unbalanced if the user requests it. Injecting it with -include is a reasonable use case, IMO and you've correctly pointed out that there's no easy way to add a matching pop.

#pragma clang attribute appears to be a better and more generic mechanism for tinkering with attributes and I would prefer to use it instead of adding more pragmas that do about the same thing.

Apr 12 2021, 7:42 AM

Apr 9 2021

ashi1 added a comment to D100060: [zorg] Add HIP builder script.

Looks reasonable overall.

Few drive-by comments below for the pitfalls you may eventually run into later.

Apr 9 2021, 11:36 AM
ashi1 added inline comments to D99894: [zorg] Add buildbot for HIP.
Apr 9 2021, 11:25 AM
ashi1 added a reviewer for D99894: [zorg] Add buildbot for HIP: tra.
Apr 9 2021, 11:22 AM
ashi1 updated the diff for D99997: [test-suite] Add HIP Tests to External.

Revised to Artem's comments, and added empty.hip.

Apr 9 2021, 9:12 AM
ashi1 added a comment to D99997: [test-suite] Add HIP Tests to External.
In D99997#2672735, @tra wrote:

I've added @beanz as the reviewer who has more cmake know-how.

I'm glad to see HIP support being added to the testsuite. I don't know what your plans are for bringing up an LLVM build/test bot for HIP. If you want, I could probably run a compile-only HIP bot on one of the VMs that run CUDA bots.

Apr 9 2021, 9:11 AM

Apr 7 2021

ashi1 added a comment to D99894: [zorg] Add buildbot for HIP.

This HIP builder is dependent on two other patches:

Apr 7 2021, 11:58 AM
ashi1 planned changes to D100060: [zorg] Add HIP builder script.

Please note, this patch is under-development, and I've added it here, as open uncommitted review to allow HIP builder to use external script.
https://reviews.llvm.org/D99894

Apr 7 2021, 11:54 AM
ashi1 requested review of D100060: [zorg] Add HIP builder script.
Apr 7 2021, 11:53 AM
ashi1 committed rGdf59850038d8: [HIP] Fix rocm-detect.hip test path (authored by ashi1).
[HIP] Fix rocm-detect.hip test path
Apr 7 2021, 10:22 AM
ashi1 closed D100045: [HIP] Fix rocm-detect.hip test path.
Apr 7 2021, 10:21 AM · Restricted Project
ashi1 requested review of D100045: [HIP] Fix rocm-detect.hip test path.
Apr 7 2021, 9:19 AM · Restricted Project

Apr 6 2021

ashi1 requested review of D99997: [test-suite] Add HIP Tests to External.
Apr 6 2021, 3:36 PM

Apr 5 2021

ashi1 added inline comments to D99894: [zorg] Add buildbot for HIP.
Apr 5 2021, 11:44 AM
ashi1 requested review of D99894: [zorg] Add buildbot for HIP.
Apr 5 2021, 11:42 AM

Apr 1 2021

ashi1 accepted D99738: [HIP] remove overloaded abs in header.

LGTM, thank you.

Apr 1 2021, 9:00 AM · Restricted Project

Mar 31 2021

ashi1 updated the diff for D98201: [CUDA][HIP] Add #pragma clang force_cuda_device_globals {begin,end}.

Merged DiagnosticParseKinds as requested.

Mar 31 2021, 2:14 PM

Mar 18 2021

ashi1 added a comment to D98278: [test] Add ability to get error messages from CMake for errc substitution.

Add GetErrcMessages.cmake, which contains a cmake function to automatically get the error messages of various posix error codes needed by lit by running a small C++ program.
Currently ENOENT, EISDIR, EINVAL and EACCES are supplied.
These error messages are then currently supplied to clang, llvm and lld as the errc_messages config parameter.

Regarding Cross compiling: the function uses try_run which when cross compiling may use the CMAKE_CROSSCOMPILING_EMULATOR to run the code.

How does it behave if such a thing isn't hooked up? Ideally it'd fall back silently and these parts of tests would just fail, but not block things overall.

It will fall back to using Python's strerror, potentially failing if pythons strerror would not return the same strings (only the case for MSVC I believe).

Mar 18 2021, 10:52 AM · Restricted Project, Restricted Project

Mar 16 2021

ashi1 added a comment to D98201: [CUDA][HIP] Add #pragma clang force_cuda_device_globals {begin,end}.

Thank you for the review, please see latest test updates adding tests for static/constexpr local var combinations.

Mar 16 2021, 2:44 PM
ashi1 updated the diff for D98201: [CUDA][HIP] Add #pragma clang force_cuda_device_globals {begin,end}.
Mar 16 2021, 2:41 PM
ashi1 updated the diff for D98201: [CUDA][HIP] Add #pragma clang force_cuda_device_globals {begin,end}.
Mar 16 2021, 2:34 PM
ashi1 added a comment to D98201: [CUDA][HIP] Add #pragma clang force_cuda_device_globals {begin,end}.
In D98201#2617905, @tra wrote:

Interesting. Once the globals are forced to be __device__, what ends up using them? Is that just for the GDB itself to access them? Or are they used by some code? If so, how is the code forced into being __global__/__device__ functions?

I can see this patch being useful for the former case.

Mar 16 2021, 10:59 AM

Mar 8 2021

ashi1 updated subscribers of D98201: [CUDA][HIP] Add #pragma clang force_cuda_device_globals {begin,end}.
Mar 8 2021, 11:04 AM
ashi1 requested review of D98201: [CUDA][HIP] Add #pragma clang force_cuda_device_globals {begin,end}.
Mar 8 2021, 10:58 AM

Jan 14 2021

ashi1 committed rGbe40c12040a0: [HIP] Add signbit(long double) decl (authored by ashi1).
[HIP] Add signbit(long double) decl
Jan 14 2021, 10:29 AM
ashi1 closed D93062: [HIP] Add signbit(long double) decl.
Jan 14 2021, 10:29 AM · Restricted Project
ashi1 added a comment to D93062: [HIP] Add signbit(long double) decl.
In D93062#2496542, @tra wrote:

For CUDA we have tests in the test-suite (https://github.com/llvm/llvm-test-suite/tree/main/External/CUDA) and a handful of buildbots running them (e.g. http://lab.llvm.org:8011/#/builders/55).
AMD should probably set up some public build/test bots for HIP, too.
In this case, manual testing and rollback if something breaks is about all we can do.

Jan 14 2021, 8:43 AM · Restricted Project

Jan 13 2021

ashi1 added a comment to D93062: [HIP] Add signbit(long double) decl.

I was able to shrink down the testcase, but it still requires 2019 MSVC to be installed and Windows Kits 10 at a minimum.

Jan 13 2021, 12:35 PM · Restricted Project
ashi1 added a comment to D93062: [HIP] Add signbit(long double) decl.

Testcase?

Jan 13 2021, 11:55 AM · Restricted Project

Dec 10 2020

ashi1 requested review of D93062: [HIP] Add signbit(long double) decl.
Dec 10 2020, 12:54 PM · Restricted Project

Dec 3 2020

ashi1 retitled D92130: [HIP] cmath demote long double args to double from [HIP] cmath promote long double args to double to [HIP] cmath demote long double args to double.
Dec 3 2020, 3:01 PM · Restricted Project
ashi1 added a comment to D92130: [HIP] cmath demote long double args to double.
In D92130#2423801, @tra wrote:

LGTM in general. Will defer to Sam as it's HIP.

Nit:

long double, promote them to double, and use the fp64

I'd say long double -> double qualifies as a demotion.

Dec 3 2020, 3:01 PM · Restricted Project
ashi1 committed rGba2612ce01ea: [HIP] cmath demote long double args to double (authored by ashi1).
[HIP] cmath demote long double args to double
Dec 3 2020, 3:01 PM
ashi1 closed D92130: [HIP] cmath demote long double args to double.
Dec 3 2020, 3:01 PM · Restricted Project

Dec 2 2020

ashi1 accepted D92418: [hip] Fix host object creation from fatbin.
Dec 2 2020, 7:17 AM · Restricted Project

Dec 1 2020

ashi1 added a comment to D92418: [hip] Fix host object creation from fatbin.

Looks good to me.

Dec 1 2020, 12:33 PM · Restricted Project
ashi1 committed rGcd5897d55908: [HIP] Fix static-lib test CHECK bug (authored by ashi1).
[HIP] Fix static-lib test CHECK bug
Dec 1 2020, 7:50 AM
ashi1 added a comment to D92342: [HIP] Fix HIP test on windows due to lld suffix.

@ashi1 This is causing build failures - please can you take a look ? http://lab.llvm.org:8011/#/builders/14/builds/2514

<stdin>:17:66: note: possible intended match here
 "/b/1/clang-x86_64-debian-new-pass-manager-fast/llvm.obj/bin/llvm-ar" "rcsD" "a.out" "/tmp/a-88f827.o" "/tmp/b-11b5fe.o" "/tmp/a-72b91c.o"
Dec 1 2020, 7:36 AM · Restricted Project

Nov 30 2020

ashi1 committed rGf89e9c8201ea: [HIP] Fix HIP test on windows due to lld suffix (authored by ashi1).
[HIP] Fix HIP test on windows due to lld suffix
Nov 30 2020, 1:06 PM