Page MenuHomePhabricator

emankov (Evgeny Mankov)
Engineering

Projects

User does not belong to any projects.

User Details

User Since
Nov 1 2016, 4:55 AM (232 w, 2 h)

Recent Activity

Oct 26 2020

emankov added a comment to D89832: [CUDA] Extract CUDA version from cuda.h if version.txt is not found.

D89832 eliminates 47332 on Windows as well. Tested against the same CUDA versions as for D89752.

Oct 26 2020, 8:27 AM · Restricted Project

Oct 23 2020

emankov added a comment to D89832: [CUDA] Extract CUDA version from cuda.h if version.txt is not found.

I confirm that D89752 eliminates 47332 on Windows. Tested against the following CUDA versions: 7.0, 7,5, 9,2, 10.0, 10,1, 10.2, 11.0 Update 1, 11.1.

Oct 23 2020, 6:24 AM · Restricted Project

Oct 22 2020

emankov accepted D89832: [CUDA] Extract CUDA version from cuda.h if version.txt is not found.
Oct 22 2020, 11:31 AM · Restricted Project
emankov requested changes to D89832: [CUDA] Extract CUDA version from cuda.h if version.txt is not found.
Oct 22 2020, 2:12 AM · Restricted Project

Oct 21 2020

emankov added inline comments to D89752: [CUDA] Improve clang's ability to detect recent CUDA versions..
Oct 21 2020, 12:09 PM · Restricted Project
emankov added inline comments to D89752: [CUDA] Improve clang's ability to detect recent CUDA versions..
Oct 21 2020, 11:24 AM · Restricted Project
emankov accepted D89752: [CUDA] Improve clang's ability to detect recent CUDA versions..

Thank you!
I've successfully applied your change in Cuda.cpp to the following clang versions: 10.0.0, 10.0.1, 11.0.0, and 12.0.0git; and have created the working patches (#206).

Oct 21 2020, 3:51 AM · Restricted Project

Oct 20 2020

emankov requested changes to D89752: [CUDA] Improve clang's ability to detect recent CUDA versions..
Oct 20 2020, 5:20 AM · Restricted Project
emankov added inline comments to D89752: [CUDA] Improve clang's ability to detect recent CUDA versions..
Oct 20 2020, 5:10 AM · Restricted Project

Oct 29 2019

emankov added a comment to D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3).
In D60220#1725633, @dim wrote:
$ cat check-isnan.cpp
#include <math.h>

int check_isnan(double d)
{
  return ::__isnan(d);
}
$ clang -c check-isnan.cpp
Why can't the regular `isnan` be used instead?  Or is this a CUDA-specific requirement?  (Apologies, but I know next to nothing about CUDA :) )
Oct 29 2019, 10:54 AM · Restricted Project

Oct 28 2019

emankov added a comment to D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3).

It looks like CUDA doesn't support double argument for device function __isnan on FreeBSD.

It's actually the opposite -- FreeBSD does not provide *host*-side __isnan(double) -- the error complains that it's the host code that tried to use __isnan and failed when overload resolution produced a device variant.

Oct 28 2019, 11:08 AM · Restricted Project
emankov added a comment to D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3).

I'm seeing quite similar errors on FreeBSD with Clang 8 and 9:
Any idea how to fix this?

Oct 28 2019, 8:17 AM · Restricted Project

Aug 27 2019

emankov committed rG2ed2e62498b7: [preprocessor] Add an opportunity to retain excluded conditional blocks (authored by emankov).
[preprocessor] Add an opportunity to retain excluded conditional blocks
Aug 27 2019, 3:16 PM
emankov committed rL370123: [preprocessor] Add an opportunity to retain excluded conditional blocks.
[preprocessor] Add an opportunity to retain excluded conditional blocks
Aug 27 2019, 3:16 PM
emankov closed D66597: [preprocessor] Add an opportunity to retain excluded conditional blocks.
Aug 27 2019, 3:16 PM · Restricted Project, Restricted Project

Aug 23 2019

emankov added a comment to D66597: [preprocessor] Add an opportunity to retain excluded conditional blocks.

I don't have a good idea of the intent of this patch. It seems to me that this could do some pretty nasty damage to C code, right?

Aug 23 2019, 9:20 AM · Restricted Project, Restricted Project
emankov added reviewers for D66597: [preprocessor] Add an opportunity to retain excluded conditional blocks: erichkeane, rsmith.
Aug 23 2019, 8:44 AM · Restricted Project, Restricted Project

Aug 22 2019

emankov created D66597: [preprocessor] Add an opportunity to retain excluded conditional blocks.
Aug 22 2019, 8:35 AM · Restricted Project, Restricted Project

Apr 18 2019

emankov committed rG88aa3d723768: [CUDA][Windows] Restrict long double device functions declarations to Windows (authored by emankov).
[CUDA][Windows] Restrict long double device functions declarations to Windows
Apr 18 2019, 3:08 AM
emankov committed rL358654: [CUDA][Windows] Restrict long double device functions declarations to Windows.
[CUDA][Windows] Restrict long double device functions declarations to Windows
Apr 18 2019, 3:07 AM
emankov committed rC358654: [CUDA][Windows] Restrict long double device functions declarations to Windows.
[CUDA][Windows] Restrict long double device functions declarations to Windows
Apr 18 2019, 3:07 AM
emankov closed D60818: [CUDA][Windows] restrict long double device functions declarations to Windows.
Apr 18 2019, 3:07 AM · Restricted Project

Apr 17 2019

emankov retitled D60818: [CUDA][Windows] restrict long double device functions declarations to Windows from [CUDA][Windows] restrict long double functions declarations to Windows to [CUDA][Windows] restrict long double device functions declarations to Windows.
Apr 17 2019, 6:26 AM · Restricted Project
emankov created D60818: [CUDA][Windows] restrict long double device functions declarations to Windows.
Apr 17 2019, 5:49 AM · Restricted Project

Apr 5 2019

emankov added a comment to D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3).
In D60220#1456449, @tra wrote:

It's not a big deal at the moment -- there are no long double users in CUDA on linux yet. You can clean up in another commit.
BTW, you may want to make commit description somewhat more concise than rC357779. Including all details of reproduction, etc is way too much info which is better suited for the bug report or review.

Apr 5 2019, 10:17 AM · Restricted Project
emankov added a comment to D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3).

Oooh, sorry, but I've just pushed the fix. But with the following words: "Add missing long double device functions' declarations. Provide only declarations to prevent any use of long double on the device side, because CUDA does not support long double on the device side."

Apr 5 2019, 9:52 AM · Restricted Project
emankov committed rG66a8b07cd991: [CUDA][Windows] Last fix for the clang Bug 38811 "Clang fails to compile with… (authored by emankov).
[CUDA][Windows] Last fix for the clang Bug 38811 "Clang fails to compile with…
Apr 5 2019, 9:52 AM
emankov committed rL357779: [CUDA][Windows] Last fix for the clang Bug 38811 "Clang fails to compile with….
[CUDA][Windows] Last fix for the clang Bug 38811 "Clang fails to compile with…
Apr 5 2019, 9:52 AM
emankov committed rC357779: [CUDA][Windows] Last fix for the clang Bug 38811 "Clang fails to compile with….
[CUDA][Windows] Last fix for the clang Bug 38811 "Clang fails to compile with…
Apr 5 2019, 9:52 AM
emankov closed D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3).
Apr 5 2019, 9:52 AM · Restricted Project
emankov updated the diff for D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3).

Provide only declarations for missing long double device functions to prevent any use of long double on the device side, because CUDA does not support long double on the device side.

Apr 5 2019, 8:11 AM · Restricted Project

Apr 3 2019

emankov created D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3).
Apr 3 2019, 11:05 AM · Restricted Project

Mar 15 2019

emankov committed rG177301f04824: [CUDA][Windows] Partial fix for bug 38811 (Step 2 of 3) (authored by emankov).
[CUDA][Windows] Partial fix for bug 38811 (Step 2 of 3)
Mar 15 2019, 12:06 PM
emankov committed rL356291: [CUDA][Windows] Partial fix for bug 38811 (Step 2 of 3).
[CUDA][Windows] Partial fix for bug 38811 (Step 2 of 3)
Mar 15 2019, 12:06 PM
emankov committed rC356291: [CUDA][Windows] Partial fix for bug 38811 (Step 2 of 3).
[CUDA][Windows] Partial fix for bug 38811 (Step 2 of 3)
Mar 15 2019, 12:06 PM
emankov closed D59423: [CUDA][Windows] Partial fix for bug 38811 (Step 2 of 3).
Mar 15 2019, 12:06 PM · Restricted Project
emankov added a comment to D59423: [CUDA][Windows] Partial fix for bug 38811 (Step 2 of 3).

The intent is to avoid unintentional clashes with the preprocessor macros the user may have defined.
https://reviews.llvm.org/rL260647

Mar 15 2019, 12:03 PM · Restricted Project
emankov added a comment to D59423: [CUDA][Windows] Partial fix for bug 38811 (Step 2 of 3).

Use of __ prefix is reserved for internal compiler use.

It is generally known. What is unknown, at least for me, is why in those clang's headers __ prefix is used so heavily?

Mar 15 2019, 11:51 AM · Restricted Project
emankov updated the diff for D59423: [CUDA][Windows] Partial fix for bug 38811 (Step 2 of 3).

As long as _sptr is also a keyword reserved by Microsoft, I agree, that float *__s and float *__c are a good choice. Moreover, there are also double *__b and int *__c presented as device functions' arguments.

Mar 15 2019, 11:37 AM · Restricted Project
emankov added a comment to D59423: [CUDA][Windows] Partial fix for bug 38811 (Step 2 of 3).

Perhaps for consistency sake it would be better to replace __sptr -> __s and __cptr -> __c.

Well, it came from NVIDIA code, you know, I mean all those double underscores. As long as _sptr is also a keyword reserved by Microsoft, I think __s and __c are a good choice. Moreover, there are also double *__b and int *__c presented as device functions' arguments.

Mar 15 2019, 11:31 AM · Restricted Project
emankov created D59423: [CUDA][Windows] Partial fix for bug 38811 (Step 2 of 3).
Mar 15 2019, 10:43 AM · Restricted Project
emankov committed rG04188fc0c6e6: [CUDA][Windows] Partial fix for bug #38811 (Step 1 of 3) (authored by emankov).
[CUDA][Windows] Partial fix for bug #38811 (Step 1 of 3)
Mar 15 2019, 5:07 AM
emankov committed rC356255: [CUDA][Windows] Partial fix for bug #38811 (Step 1 of 3).
[CUDA][Windows] Partial fix for bug #38811 (Step 1 of 3)
Mar 15 2019, 5:04 AM
emankov committed rL356255: [CUDA][Windows] Partial fix for bug #38811 (Step 1 of 3).
[CUDA][Windows] Partial fix for bug #38811 (Step 1 of 3)
Mar 15 2019, 5:04 AM
emankov closed D59361: [CUDA][Windows] Partial fix for bug 38811 (Step 1 of 3).
Mar 15 2019, 5:04 AM · Restricted Project

Mar 14 2019

emankov created D59361: [CUDA][Windows] Partial fix for bug 38811 (Step 1 of 3).
Mar 14 2019, 6:25 AM · Restricted Project

Sep 25 2017

emankov committed rL314108: [ELF][fix] Using statically linked lld leads to segmentation fault on linking….
[ELF][fix] Using statically linked lld leads to segmentation fault on linking…
Sep 25 2017, 7:44 AM
emankov closed D38131: [ELF] [fix] Using statically linked lld leads to segmentation fault on linking ELF by committing rL314108: [ELF][fix] Using statically linked lld leads to segmentation fault on linking….
Sep 25 2017, 7:43 AM · lld

Sep 21 2017

emankov created D38131: [ELF] [fix] Using statically linked lld leads to segmentation fault on linking ELF.
Sep 21 2017, 5:24 AM · lld

Sep 5 2017

emankov accepted D37386: [AMDGPU] Implement infrastructure to set options in AMDGPUToolChain.

LGTM

Sep 5 2017, 2:33 AM

Sep 1 2017

emankov added inline comments to D37386: [AMDGPU] Implement infrastructure to set options in AMDGPUToolChain.
Sep 1 2017, 10:50 AM

Aug 30 2017

emankov abandoned D37293: [lld][ELF] Fix for Bug 34378.
Aug 30 2017, 8:46 AM
emankov added a comment to D37293: [lld][ELF] Fix for Bug 34378.

Ooops, you right, you've already fixed it with:

Aug 30 2017, 8:46 AM
emankov created D37293: [lld][ELF] Fix for Bug 34378.
Aug 30 2017, 5:55 AM

Aug 28 2017

emankov committed rL311887: [Support][CommandLine] Add cl::Option::setDefault().
[Support][CommandLine] Add cl::Option::setDefault()
Aug 28 2017, 6:43 AM
emankov closed D36877: [Support][CommandLine] Add cl::Option::setDefault() by committing rL311887: [Support][CommandLine] Add cl::Option::setDefault().
Aug 28 2017, 6:43 AM

Aug 24 2017

emankov updated the diff for D36877: [Support][CommandLine] Add cl::Option::setDefault().

Test is added.

Aug 24 2017, 10:39 AM

Aug 23 2017

emankov added a comment to D36877: [Support][CommandLine] Add cl::Option::setDefault().
Aug 23 2017, 5:54 AM
emankov updated the diff for D36877: [Support][CommandLine] Add cl::Option::setDefault().

formatting

Aug 23 2017, 5:36 AM

Aug 18 2017

emankov created D36877: [Support][CommandLine] Add cl::Option::setDefault().
Aug 18 2017, 9:15 AM

Aug 16 2017

emankov committed rL311019: [AMDGPU] NFC: test commit.
[AMDGPU] NFC: test commit
Aug 16 2017, 9:48 AM