Page MenuHomePhabricator

nickdesaulniers (Nick Desaulniers)
Google

Projects

User does not belong to any projects.

User Details

User Since
Apr 13 2018, 4:23 PM (61 w, 2 d)

Recent Activity

Today

nickdesaulniers accepted D63369: [AST] Fixed extraneous warnings for binary conditional operator.

Thanks for the patch and following up on the review comments. Let's work on getting you commit access now.

Mon, Jun 17, 11:28 AM · Restricted Project
nickdesaulniers added a comment to D63432: [ELF] Allow placing SHF_MERGE sections with different alignments into the same MergeSyntheticSection.

Thanks for taking a look @MaskRay ! From this thread from our mailing list, I think that the different entity size and alignment may be a bug: https://groups.google.com/forum/#!topic/clang-built-linux/0V6APpBcl3k

Mon, Jun 17, 11:14 AM · Restricted Project
nickdesaulniers added inline comments to D63369: [AST] Fixed extraneous warnings for binary conditional operator.
Mon, Jun 17, 11:12 AM · Restricted Project
nickdesaulniers added inline comments to D63369: [AST] Fixed extraneous warnings for binary conditional operator.
Mon, Jun 17, 9:49 AM · Restricted Project

Fri, Jun 14

nickdesaulniers added inline comments to D63369: [AST] Fixed extraneous warnings for binary conditional operator.
Fri, Jun 14, 4:17 PM · Restricted Project
nickdesaulniers added a comment to D63105: Add --print-supported-cpus flag for clang..

@pirama and @kongyi pointed out to me that -emit-llvm tests don't require backend support, and that the undocumented (https://llvm.org/docs/CommandGuide/FileCheck.html) REQUIRES: tag would fix this.

Fri, Jun 14, 4:15 PM · Restricted Project
nickdesaulniers added a comment to D63105: Add --print-supported-cpus flag for clang..

https://llvm.org/docs/TestingGuide.html mentions .cfg files, but tests I've looked at don't seem to have a .cfg in their directory.

Fri, Jun 14, 3:55 PM · Restricted Project
nickdesaulniers added a comment to D63105: Add --print-supported-cpus flag for clang..

There are lots of tests under clang/test/ that target arm-- as a -triple. How is it that they don't fail in the same way? %clang_cc1 -triple arm-- is somehow different?

Fri, Jun 14, 3:54 PM · Restricted Project

Thu, Jun 13

nickdesaulniers added inline comments to D63260: [Attr] Support _attribute__ ((fallthrough)).
Thu, Jun 13, 2:54 PM · Restricted Project
nickdesaulniers added inline comments to D63260: [Attr] Support _attribute__ ((fallthrough)).
Thu, Jun 13, 11:40 AM · Restricted Project
nickdesaulniers added inline comments to D63260: [Attr] Support _attribute__ ((fallthrough)).
Thu, Jun 13, 11:26 AM · Restricted Project
nickdesaulniers added a comment to D63260: [Attr] Support _attribute__ ((fallthrough)).

but as part of this commit

Thu, Jun 13, 11:18 AM · Restricted Project
nickdesaulniers added inline comments to D63260: [Attr] Support _attribute__ ((fallthrough)).
Thu, Jun 13, 11:17 AM · Restricted Project
nickdesaulniers added a comment to D63260: [Attr] Support _attribute__ ((fallthrough)).

Thanks for the patch, I look forward to this feature!

Thu, Jun 13, 9:49 AM · Restricted Project

Tue, Jun 11

nickdesaulniers added a comment to D59963: [clang-tidy] Add a module for the Linux kernel..

See also: https://github.com/ClangBuiltLinux/linux/issues/520

Tue, Jun 11, 2:47 PM · Restricted Project, Restricted Project

Fri, Jun 7

nickdesaulniers abandoned D63017: [Android] add toplevel Android.mk to stop breakage.

Looks like this is not the correct solution. See also: https://android-review.googlesource.com/c/toolchain/llvm-project/+/979791

Fri, Jun 7, 1:12 PM · Restricted Project
nickdesaulniers committed rG7ddd694d3684: [APFloat] APFloat::Storage::Storage - refix use after move (authored by nickdesaulniers).
[APFloat] APFloat::Storage::Storage - refix use after move
Fri, Jun 7, 12:50 PM
nickdesaulniers committed rL362833: [APFloat] APFloat::Storage::Storage - refix use after move.
[APFloat] APFloat::Storage::Storage - refix use after move
Fri, Jun 7, 12:50 PM
nickdesaulniers closed D62767: [APFloat] APFloat::Storage::Storage - refix use after move.
Fri, Jun 7, 12:50 PM · Restricted Project
nickdesaulniers created D63017: [Android] add toplevel Android.mk to stop breakage.
Fri, Jun 7, 12:04 PM · Restricted Project
nickdesaulniers added inline comments to D59963: [clang-tidy] Add a module for the Linux kernel..
Fri, Jun 7, 11:39 AM · Restricted Project, Restricted Project

Wed, Jun 5

nickdesaulniers added a comment to D56571: [RFC prototype] Implementation of asm-goto support in clang.

For future travelers, we're tracking 2 related bugs:

Wed, Jun 5, 3:10 PM

Tue, Jun 4

nickdesaulniers updated subscribers of D62767: [APFloat] APFloat::Storage::Storage - refix use after move.

parting thoughts @eli.friedman ?

Tue, Jun 4, 6:29 PM · Restricted Project
nickdesaulniers committed rG2e05045e2601: [TargetTransformInfo] assert on nullptr (authored by nickdesaulniers).
[TargetTransformInfo] assert on nullptr
Tue, Jun 4, 6:29 PM
nickdesaulniers committed rL362567: [TargetTransformInfo] assert on nullptr.
[TargetTransformInfo] assert on nullptr
Tue, Jun 4, 6:28 PM
nickdesaulniers closed D62229: [TargetTransformInfo] assert on nullptr.
Tue, Jun 4, 6:28 PM · Restricted Project

Sun, Jun 2

nickdesaulniers added a comment to D62229: [TargetTransformInfo] assert on nullptr.
Sun, Jun 2, 7:23 PM · Restricted Project
nickdesaulniers updated the diff for D62229: [TargetTransformInfo] assert on nullptr.
Sun, Jun 2, 7:23 PM · Restricted Project

Fri, May 31

nickdesaulniers committed rGb380846a125d: [RuntimeDyld] fix too-small-bitmask error (authored by nickdesaulniers).
[RuntimeDyld] fix too-small-bitmask error
Fri, May 31, 9:50 PM
nickdesaulniers committed rL362286: [RuntimeDyld] fix too-small-bitmask error.
[RuntimeDyld] fix too-small-bitmask error
Fri, May 31, 9:49 PM
nickdesaulniers closed D62227: [RuntimeDyld] fix too-small-bitmask error.
Fri, May 31, 9:49 PM · Restricted Project
nickdesaulniers updated the diff for D62229: [TargetTransformInfo] assert on nullptr.
  • git-clang-format HEAD~
Fri, May 31, 9:46 PM · Restricted Project
nickdesaulniers retitled D62229: [TargetTransformInfo] assert on nullptr from [TargetTransformInfo] early return on nullptr to [TargetTransformInfo] assert on nullptr.
Fri, May 31, 9:45 PM · Restricted Project
nickdesaulniers updated the diff for D62229: [TargetTransformInfo] assert on nullptr.
  • prefer assertion to release build checks
Fri, May 31, 9:45 PM · Restricted Project
nickdesaulniers updated subscribers of D62227: [RuntimeDyld] fix too-small-bitmask error.
Fri, May 31, 9:28 PM · Restricted Project
nickdesaulniers added a reviewer for D62227: [RuntimeDyld] fix too-small-bitmask error: MaskRay.
Fri, May 31, 9:28 PM · Restricted Project
nickdesaulniers updated the diff for D62227: [RuntimeDyld] fix too-small-bitmask error.
Fri, May 31, 9:28 PM · Restricted Project
nickdesaulniers created D62767: [APFloat] APFloat::Storage::Storage - refix use after move.
Fri, May 31, 8:24 PM · Restricted Project
nickdesaulniers committed rG7fcad2f17111: [PowerPC] check for INLINEASM_BR along w/ INLINEASM (authored by nickdesaulniers).
[PowerPC] check for INLINEASM_BR along w/ INLINEASM
Fri, May 31, 4:03 PM
nickdesaulniers committed rL362278: [PowerPC] check for INLINEASM_BR along w/ INLINEASM.
[PowerPC] check for INLINEASM_BR along w/ INLINEASM
Fri, May 31, 4:02 PM
nickdesaulniers closed D62403: [PowerPC] check for INLINEASM_BR along w/ INLINEASM.
Fri, May 31, 4:02 PM · Restricted Project
nickdesaulniers committed rG8b1f64f63d77: [Bugpoint] fix another use-after-move. NFC (authored by nickdesaulniers).
[Bugpoint] fix another use-after-move. NFC
Fri, May 31, 2:34 PM
nickdesaulniers committed rL362267: [Bugpoint] fix another use-after-move. NFC.
[Bugpoint] fix another use-after-move. NFC
Fri, May 31, 2:33 PM
nickdesaulniers closed D62114: [Bugpoint] fix another use-after-move. NFC.
Fri, May 31, 2:33 PM · Restricted Project
nickdesaulniers committed rG103bd108a71c: [RegisterCoalescer] fix potential use of undef value. NFC (authored by nickdesaulniers).
[RegisterCoalescer] fix potential use of undef value. NFC
Fri, May 31, 2:19 PM
nickdesaulniers committed rL362265: [RegisterCoalescer] fix potential use of undef value. NFC.
[RegisterCoalescer] fix potential use of undef value. NFC
Fri, May 31, 2:18 PM
nickdesaulniers closed D62109: [RegisterCoalescer] fix potential use of undef value. NFC.
Fri, May 31, 2:18 PM · Restricted Project
nickdesaulniers added a comment to D62403: [PowerPC] check for INLINEASM_BR along w/ INLINEASM.

Also, I'm fine with these as by-inspection changes, but if there are reasonably-easy ways to write test cases (or update existing ones), please do.

Fri, May 31, 1:52 PM · Restricted Project
nickdesaulniers added a comment to D62749: [X86] remove derived method w/ same impl as base.

Adding @hfinkel for review because I think we can go further. The only other override of TargetInstrInfo::isUnpredicatedTerminator is PPCInstrInfo::isUnpredicatedTerminator and it almost matches, modulo a check for whether the MachineInstr isPredicable. I'm not sure if that's intentional, or bug, but adding return TargetInstrInfo::isUnpredicatedTerminator(MI); as the first statement in PPCInstrInfo::isUnpredicatedTerminator doesn't cause any failures in existing tests.

Fri, May 31, 1:52 PM · Restricted Project
nickdesaulniers created D62749: [X86] remove derived method w/ same impl as base.
Fri, May 31, 1:50 PM · Restricted Project
nickdesaulniers updated the diff for D62555: [TailDuplicator] prevent tail duplication for INLINEASM_BR.
  • add fixme
Fri, May 31, 11:05 AM · Restricted Project

Thu, May 30

nickdesaulniers updated the diff for D62555: [TailDuplicator] prevent tail duplication for INLINEASM_BR.
  • fix typo
Thu, May 30, 5:49 PM · Restricted Project
nickdesaulniers added a comment to D62555: [TailDuplicator] prevent tail duplication for INLINEASM_BR.

It's possible the transform isn't profitable if we're optimizing for size, but that's a completely different issue.

Thu, May 30, 3:26 PM · Restricted Project
nickdesaulniers added a reviewer for D62555: [TailDuplicator] prevent tail duplication for INLINEASM_BR: eli.friedman.
Thu, May 30, 2:49 PM · Restricted Project
nickdesaulniers added inline comments to D62555: [TailDuplicator] prevent tail duplication for INLINEASM_BR.
Thu, May 30, 2:49 PM · Restricted Project
nickdesaulniers updated the diff for D62555: [TailDuplicator] prevent tail duplication for INLINEASM_BR.
  • get rid of some debug info
  • hoist INLINEASM_BR check into shouldTailDuplicate
Thu, May 30, 2:49 PM · Restricted Project
nickdesaulniers added a comment to D62555: [TailDuplicator] prevent tail duplication for INLINEASM_BR.

so probably a bug in if conversion. Maybe the same issue as https://bugs.llvm.org/show_bug.cgi?id=41121 .

Thu, May 30, 1:19 PM · Restricted Project
nickdesaulniers added a comment to D56571: [RFC prototype] Implementation of asm-goto support in clang.

Was this committed accidentally?

Thu, May 30, 12:55 PM

Wed, May 29

nickdesaulniers added a comment to D62555: [TailDuplicator] prevent tail duplication for INLINEASM_BR.

Otherwise, conservatively do not allow INLINEASM_BR to be predicable. (I'm leaning more towards this, since such machinery would be rewriting the inline asm the user specified).

Wed, May 29, 3:39 PM · Restricted Project
nickdesaulniers added a comment to D62555: [TailDuplicator] prevent tail duplication for INLINEASM_BR.

Could this be fixed in the if conversion pass IfConverter::ScanInstructions() by having TargetInstrInfo::isPredicable(MachineInstr&) reject TargetOpcode::INLINEASM_BR?

Wed, May 29, 10:29 AM · Restricted Project

Tue, May 28

nickdesaulniers updated the summary of D62555: [TailDuplicator] prevent tail duplication for INLINEASM_BR.
Tue, May 28, 4:33 PM · Restricted Project
nickdesaulniers updated the diff for D62555: [TailDuplicator] prevent tail duplication for INLINEASM_BR.
  • get rid of some debug info
Tue, May 28, 4:17 PM · Restricted Project
nickdesaulniers retitled D62555: [TailDuplicator] prevent tail duplication for INLINEASM_BR from prevent tail duplication for INLINEASM_BR to [TailDuplicator] prevent tail duplication for INLINEASM_BR.
Tue, May 28, 4:03 PM · Restricted Project
nickdesaulniers created D62555: [TailDuplicator] prevent tail duplication for INLINEASM_BR.
Tue, May 28, 4:01 PM · Restricted Project

Fri, May 24

nickdesaulniers committed rG33bc64202b32: [AArch64] check for INLINEASM_BR along w/ INLINEASM (authored by nickdesaulniers).
[AArch64] check for INLINEASM_BR along w/ INLINEASM
Fri, May 24, 11:58 AM
nickdesaulniers committed rL361661: [AArch64] check for INLINEASM_BR along w/ INLINEASM.
[AArch64] check for INLINEASM_BR along w/ INLINEASM
Fri, May 24, 11:58 AM
nickdesaulniers closed D62402: [AArch64] check for INLINEASM_BR along w/ INLINEASM.
Fri, May 24, 11:58 AM · Restricted Project
nickdesaulniers accepted D56571: [RFC prototype] Implementation of asm-goto support in clang.

Ok, from the Linux kernel's perspective, I believe we have worked out all underlying issues in LLVM wrt callbr that would prevent us from using asm goto successfully. This patch now has my blessing; thanks for all the hard work that went into this large feature. Please wait for a final LGTM from @rsmith .

Fri, May 24, 11:58 AM
nickdesaulniers committed rG9f7bd71cf581: [ARM] additionally check for ARM::INLINEASM_BR w/ ARM::INLINEASM (authored by nickdesaulniers).
[ARM] additionally check for ARM::INLINEASM_BR w/ ARM::INLINEASM
Fri, May 24, 11:56 AM
nickdesaulniers committed rL361659: [ARM] additionally check for ARM::INLINEASM_BR w/ ARM::INLINEASM.
[ARM] additionally check for ARM::INLINEASM_BR w/ ARM::INLINEASM
Fri, May 24, 11:56 AM
nickdesaulniers closed D62400: [ARM] additionally check for ARM::INLINEASM_BR w/ ARM::INLINEASM.
Fri, May 24, 11:55 AM · Restricted Project
nickdesaulniers added a comment to D62400: [ARM] additionally check for ARM::INLINEASM_BR w/ ARM::INLINEASM.

Landing for now to unblock asm goto https://reviews.llvm.org/D56571. I will try to get a test in for this later.

Fri, May 24, 11:44 AM · Restricted Project
nickdesaulniers added a comment to D62400: [ARM] additionally check for ARM::INLINEASM_BR w/ ARM::INLINEASM.
Fri, May 24, 10:34 AM · Restricted Project
nickdesaulniers added a comment to D62400: [ARM] additionally check for ARM::INLINEASM_BR w/ ARM::INLINEASM.
Fri, May 24, 10:29 AM · Restricted Project
nickdesaulniers added a comment to D62400: [ARM] additionally check for ARM::INLINEASM_BR w/ ARM::INLINEASM.

constant pool goes at the end by default

Fri, May 24, 10:23 AM · Restricted Project
nickdesaulniers retitled D62403: [PowerPC] check for INLINEASM_BR along w/ INLINEASM from check for INLINEASM_BR along w/ INLINEASM to [PowerPC] check for INLINEASM_BR along w/ INLINEASM.
Fri, May 24, 9:34 AM · Restricted Project
nickdesaulniers created D62403: [PowerPC] check for INLINEASM_BR along w/ INLINEASM.
Fri, May 24, 9:33 AM · Restricted Project
nickdesaulniers updated subscribers of D62402: [AArch64] check for INLINEASM_BR along w/ INLINEASM.
Fri, May 24, 9:28 AM · Restricted Project
nickdesaulniers created D62402: [AArch64] check for INLINEASM_BR along w/ INLINEASM.
Fri, May 24, 9:25 AM · Restricted Project
nickdesaulniers added reviewers for D62400: [ARM] additionally check for ARM::INLINEASM_BR w/ ARM::INLINEASM: rengolin, t.p.northover.
Fri, May 24, 9:23 AM · Restricted Project
nickdesaulniers updated subscribers of D62400: [ARM] additionally check for ARM::INLINEASM_BR w/ ARM::INLINEASM.
Fri, May 24, 9:06 AM · Restricted Project
nickdesaulniers created D62400: [ARM] additionally check for ARM::INLINEASM_BR w/ ARM::INLINEASM.
Fri, May 24, 9:02 AM · Restricted Project

Tue, May 21

nickdesaulniers updated subscribers of D62232: [Clang][Driver] recheck for nullptr.
Tue, May 21, 8:36 PM · Restricted Project
nickdesaulniers created D62232: [Clang][Driver] recheck for nullptr.
Tue, May 21, 8:34 PM · Restricted Project
nickdesaulniers created D62231: [DebugInfoMetadata] return early on nullptr arg.
Tue, May 21, 8:22 PM · Restricted Project
nickdesaulniers created D62230: [CGDebugInfo] return early on failed dyn_cast.
Tue, May 21, 8:05 PM · Restricted Project
nickdesaulniers added a comment to D62230: [CGDebugInfo] return early on failed dyn_cast.

Also of note is that no existing test covers this case.

Tue, May 21, 8:05 PM · Restricted Project
nickdesaulniers created D62229: [TargetTransformInfo] assert on nullptr.
Tue, May 21, 7:11 PM · Restricted Project
nickdesaulniers created D62227: [RuntimeDyld] fix too-small-bitmask error.
Tue, May 21, 6:36 PM · Restricted Project
nickdesaulniers added inline comments to rL361329: [c++20] P1330R0: permit simple-assignments that change the active member.
Tue, May 21, 6:13 PM
nickdesaulniers added inline comments to D62116: [Sema] raise nullptr check to cover additional uses.
Tue, May 21, 4:46 PM · Restricted Project
nickdesaulniers added a comment to D62110: [LV] prevent potential divide by zero. NFC.

Can expectedCost ever return zero?

Tue, May 21, 4:31 PM · Restricted Project
nickdesaulniers added a comment to D62112: [ORC] fix use-after-move. NFC.

because is it a string reference? (that might be invalidated by the move), a return by value (that then means it's reference extending, which is best avoided unless really needed),

Tue, May 21, 3:24 PM · Restricted Project
nickdesaulniers added a comment to D62120: [ORC] fix unreachable code bug.

Wouldn't just deleting L142 from the existing sources be simpler? (or moving L146-148 into the for loop at the end (L142)?

Tue, May 21, 3:22 PM · Restricted Project
nickdesaulniers added a comment to D62114: [Bugpoint] fix another use-after-move. NFC.

Correction: -DLLVM_BUILD_LLVM_DYLIB="ON" is required to run ninja check-llvm-bugpoint.

Tue, May 21, 3:19 PM · Restricted Project
nickdesaulniers abandoned D62218: [docs][cmake] document LLVM_ENABLE_PLUGINS.

Is there a point to adding this if it isn't a user selectable option? LLVM_ENABLE_PLUGINS doesn't have an option in any cmake files, it is only ever set.

Tue, May 21, 3:15 PM · Restricted Project
nickdesaulniers updated subscribers of D62218: [docs][cmake] document LLVM_ENABLE_PLUGINS.
Tue, May 21, 2:28 PM · Restricted Project
nickdesaulniers updated the diff for D62218: [docs][cmake] document LLVM_ENABLE_PLUGINS.
  • capitalize LLVM
Tue, May 21, 2:28 PM · Restricted Project
nickdesaulniers committed rG17694af98087: [Driver] Verify GCCInstallation is valid (authored by nickdesaulniers).
[Driver] Verify GCCInstallation is valid
Tue, May 21, 2:19 PM
nickdesaulniers committed rL361314: [Driver] Verify GCCInstallation is valid.
[Driver] Verify GCCInstallation is valid
Tue, May 21, 2:19 PM