Page MenuHomePhabricator
Feed Advanced Search

Today

tejohnson accepted D91687: [compiler-rt santizer] Use clock_gettime instead of timespec_get.

lgtm

Thu, Dec 3, 7:46 AM

Yesterday

tejohnson added a comment to D91474: [buildbot] Fix worker for ThinLTO whole program devirtualization.

Ping - @gkistanova can you take a look and also let me know if there is a way to test the LTO builder change?

Wed, Dec 2, 8:02 AM

Tue, Dec 1

tejohnson accepted D92335: [ThinLTO] Import symver directives for imported symbols (PR48214).

lgtm. Thanks for fixing this, and that's a nice cleanup on the existing handling! One request - could you submit the change to the current handling and associated test changes first, with the symver handling/test as a follow on?

Tue, Dec 1, 12:44 PM · Restricted Project
tejohnson added inline comments to D92335: [ThinLTO] Import symver directives for imported symbols (PR48214).
Tue, Dec 1, 11:22 AM · Restricted Project
tejohnson added a comment to D92335: [ThinLTO] Import symver directives for imported symbols (PR48214).

I just realized this makes Linker/ depend on Object/
I'm not sure whether that's a problem or not, but we could avoid that by storing the symvers in a metadata node again.

Tue, Dec 1, 8:38 AM · Restricted Project

Mon, Nov 30

tejohnson added a comment to D92335: [ThinLTO] Import symver directives for imported symbols (PR48214).

Thinking through this some more, I think I was making it too hard. I think we should just be able to collect the symvers out of the inline asm when we do the IR linking. However, not in the place you've modified below, as I note there. Instead, take a look at IRLinker::run(), where it appends the module inline asm strings together if !IsPerformingImport. I think here you could add a case where IsPerformingImport, and just invoke ModuleSymbolTable::CollectAsmSymvers here and directly add any symver you find to the DstM (probably do all this in a helper function). I was thinking that we might need to carry it via module metadata to avoid parsing it during IR linking, but I'm now not sure that's necessary - I don't think it should add significant overhead. Let me know if you run into issues with that approach.

Mon, Nov 30, 9:50 PM · Restricted Project
tejohnson accepted D85808: [Remarks][2/2] Expand remarks hotness threshold option support in more tools.

lgtm with a couple of minor nits noted below that you can fix before submitting

Mon, Nov 30, 8:20 PM · Restricted Project, Restricted Project
tejohnson accepted D85809: [Remarks][1/2] Expand remarks hotness threshold option support in more tools.

lgtm

Mon, Nov 30, 8:20 PM · Restricted Project

Tue, Nov 24

tejohnson committed rG07f234be1ccb: [lld] Add --no-lto-whole-program-visibility (authored by tejohnson).
[lld] Add --no-lto-whole-program-visibility
Tue, Nov 24, 4:46 PM
tejohnson closed D92060: [lld] Add --no-lto-whole-program-visibility.
Tue, Nov 24, 4:46 PM · Restricted Project
tejohnson updated the diff for D92060: [lld] Add --no-lto-whole-program-visibility.

Address comments

Tue, Nov 24, 3:12 PM · Restricted Project
tejohnson added inline comments to D92060: [lld] Add --no-lto-whole-program-visibility.
Tue, Nov 24, 3:07 PM · Restricted Project
tejohnson added a comment to D91583: [lld] Allow --export-dynamic to override --lto-whole-program-visibility.

While I investigate alternative mechanisms to handle --export-dynamic, I've sent D92060 to add a --no- version of the option, to simplify working around issues when --lto-whole-program-visibility is specified broadly.

Tue, Nov 24, 2:57 PM · Restricted Project
tejohnson requested review of D92060: [lld] Add --no-lto-whole-program-visibility.
Tue, Nov 24, 2:56 PM · Restricted Project
tejohnson committed rG0768b0576a93: Avoid redundant work when computing vtable vcall visibility (authored by tejohnson).
Avoid redundant work when computing vtable vcall visibility
Tue, Nov 24, 12:06 PM
tejohnson closed D91676: Avoid redundant work when computing vtable vcall visibility.
Tue, Nov 24, 12:06 PM · Restricted Project
tejohnson added inline comments to D91676: Avoid redundant work when computing vtable vcall visibility.
Tue, Nov 24, 10:16 AM · Restricted Project
tejohnson updated the diff for D91676: Avoid redundant work when computing vtable vcall visibility.

Improve comment

Tue, Nov 24, 10:14 AM · Restricted Project
tejohnson added inline comments to D91676: Avoid redundant work when computing vtable vcall visibility.
Tue, Nov 24, 9:38 AM · Restricted Project
tejohnson committed rG6e4c1cf29388: [ThinLTO/WPD] Enable -wholeprogramdevirt-skip in ThinLTO backends (authored by tejohnson).
[ThinLTO/WPD] Enable -wholeprogramdevirt-skip in ThinLTO backends
Tue, Nov 24, 9:35 AM
tejohnson closed D91812: [ThinLTO/WPD] Enable -wholeprogramdevirt-skip in ThinLTO backends.
Tue, Nov 24, 9:35 AM · Restricted Project, Restricted Project
tejohnson added a comment to D91676: Avoid redundant work when computing vtable vcall visibility.

Adding another reviewer - @wmi can you take a look? This is a straightforward compile time fix.

Tue, Nov 24, 8:26 AM · Restricted Project
tejohnson added a reviewer for D91676: Avoid redundant work when computing vtable vcall visibility: wmi.
Tue, Nov 24, 8:25 AM · Restricted Project

Mon, Nov 23

tejohnson added a comment to D91676: Avoid redundant work when computing vtable vcall visibility.

ping

Mon, Nov 23, 9:35 AM · Restricted Project

Fri, Nov 20

tejohnson added a comment to D91567: [llvm][inliner] Reuse the inliner pass to implement 'always inliner'.

From a ThinLTO perspective, no specific concerns as the buildModuleSimplificationPipeline is invoked in both the pre and post LTO link pipelines, so they both get an equivalent change. But there is an issue for regular LTO, noted below.

Fri, Nov 20, 9:20 AM · Restricted Project, Restricted Project

Thu, Nov 19

tejohnson accepted D91410: [llvm][clang][mlir] Add checks for the return values from Target::createXXX to prevent protential null deref.

LGTM

Thu, Nov 19, 8:16 PM · Restricted Project, Restricted Project, Restricted Project
tejohnson requested review of D91812: [ThinLTO/WPD] Enable -wholeprogramdevirt-skip in ThinLTO backends.
Thu, Nov 19, 11:47 AM · Restricted Project, Restricted Project
tejohnson accepted D91585: [NPM] Move more O0 pass building into PassBuilder.

lgtm

Thu, Nov 19, 10:37 AM · Restricted Project, Restricted Project
tejohnson added a comment to D91583: [lld] Allow --export-dynamic to override --lto-whole-program-visibility.
In D91583#2406005, @pcc wrote:
In D91583#2404519, @pcc wrote:

I've said before that I think that --lto-whole-program-visibility should relax visibility of vtable symbols etc to hidden. That way, --export-dynamic wouldn't actually allow you to make this kind of mistake.

That would presumably result in an error in some of the problematic cases, whereas here we want to simply suppress --lto-whole-program-visibility to avoid any issues automatically.

But isn't it the case that you don't even need for the vtable symbol itself to be exported in order to derive from the class and override its virtual methods?

That's true, but it's the same situation that we have now with --lto-whole-program-visibility and not passing --export-dynamic.

Thu, Nov 19, 10:27 AM · Restricted Project
tejohnson committed rGa75b2e87e6cb: [MemProf] Add interface to dump profile (authored by tejohnson).
[MemProf] Add interface to dump profile
Thu, Nov 19, 10:22 AM
tejohnson closed D91768: [MemProf] Add interface to dump profile.
Thu, Nov 19, 10:22 AM · Restricted Project
tejohnson added inline comments to D91585: [NPM] Move more O0 pass building into PassBuilder.
Thu, Nov 19, 10:13 AM · Restricted Project, Restricted Project
tejohnson committed rG8f778b283de5: [sanitizer_common] Add facility to get the full report path (authored by tejohnson).
[sanitizer_common] Add facility to get the full report path
Thu, Nov 19, 9:19 AM
tejohnson closed D91765: [sanitizer_common] Add facility to get the full report path.
Thu, Nov 19, 9:19 AM · Restricted Project
tejohnson updated the diff for D91765: [sanitizer_common] Add facility to get the full report path.

Beef up test as suggested to check the full path via an assert

Thu, Nov 19, 9:19 AM · Restricted Project
tejohnson added inline comments to D91585: [NPM] Move more O0 pass building into PassBuilder.
Thu, Nov 19, 9:18 AM · Restricted Project, Restricted Project
tejohnson added a comment to D91474: [buildbot] Fix worker for ThinLTO whole program devirtualization.

Ping - @gkistanova can you take a look and also let me know if there is a way to test the LTO builder change?

Thu, Nov 19, 8:55 AM

Wed, Nov 18

tejohnson added inline comments to D91765: [sanitizer_common] Add facility to get the full report path.
Wed, Nov 18, 11:44 PM · Restricted Project
tejohnson updated the diff for D91765: [sanitizer_common] Add facility to get the full report path.

Fix test to avoid using /tmp

Wed, Nov 18, 11:24 PM · Restricted Project
tejohnson added inline comments to D91765: [sanitizer_common] Add facility to get the full report path.
Wed, Nov 18, 11:24 PM · Restricted Project
tejohnson requested review of D91768: [MemProf] Add interface to dump profile.
Wed, Nov 18, 11:15 PM · Restricted Project
tejohnson added a comment to D91765: [sanitizer_common] Add facility to get the full report path.

It mentions the test, but it's not here

Wed, Nov 18, 10:11 PM · Restricted Project
tejohnson updated the diff for D91765: [sanitizer_common] Add facility to get the full report path.

Add the test

Wed, Nov 18, 10:10 PM · Restricted Project
tejohnson added a comment to D91583: [lld] Allow --export-dynamic to override --lto-whole-program-visibility.
In D91583#2404519, @pcc wrote:

I've said before that I think that --lto-whole-program-visibility should relax visibility of vtable symbols etc to hidden. That way, --export-dynamic wouldn't actually allow you to make this kind of mistake.

Wed, Nov 18, 10:09 PM · Restricted Project
tejohnson requested review of D91765: [sanitizer_common] Add facility to get the full report path.
Wed, Nov 18, 9:58 PM · Restricted Project
tejohnson added a comment to D91583: [lld] Allow --export-dynamic to override --lto-whole-program-visibility.

Thanks for the clarification.

Hi, I am still learning the feature and I've just played a bit with the test. I've a couple of questions:

  • The --export-dynamic usage in the test seems a bit confusing. Are the VisibleToRegularObj bits of these _ZTV* symbols the important matter and --export-dynamic is just an approach to make them true?

Correct. It was just a shorthand way of preventing these symbols from being eliminated.

  • If yes, I guess all these --export-dynamic can be replaced with -u _ZTV1B -u _ZTV1C -u _ZTV1D in all the RUN lines. --export-dynamic is preferred just due to its brevity. If that is the case, I think this deserves a comment considering its subtle interaction with noexportdynamic.

Right, that's why for one of the tests where I'm using the new noexportdynamic value I switched to the -u sequence instead. I can add a comment.

  • When is devirtualization invalid? For example, if _ZTV1D is exported to the dynamic symbol table and a shared object inherits from class D and overrides the method?

Correct. If it is both exported and then overridden.

If my last point is correct, I'd agree that we need something like a tri-state option. (I hope we can remove --lto-whole-program-visibility if possible)
on/noexportdynamic the value names do not capture the actual meaning. The on actually means: devirtualization is safe as long as the used _ZTV* symbols are not exported.

Suggestion on the name? It's basically an assertion, i.e. that there is LTO whole program visibility (no, when --export-dynamic not specified, and yes). 'on' means the user is asserting that the _ZTV* are not going to be exported and overridden.

I have a further question: is it realistic to add another bit along with VisibleToRegularObj to convey the information whether a symbol is includeInDynsym()?
This way virtual functions related to individual _ZTV* can be safely devirtualized, no matter how users specify --export-dynamic-symbol,--export-dynamic,--dynamic-list or add link-time shared objects to alter the includeInDynsym() state of _ZTV* symbols.

Wed, Nov 18, 6:14 PM · Restricted Project
tejohnson added inline comments to D91687: [compiler-rt santizer] Use clock_gettime instead of timespec_get.
Wed, Nov 18, 3:35 PM
tejohnson added a comment to D85808: [Remarks][2/2] Expand remarks hotness threshold option support in more tools.

Thanks for adding the Driver test. I was thinking of something to test the CompilerInvocation changes, similar to your test using opt, that ensures the option has the desired behavior when invoked via clang. Looks like there is an existing test clang/test/Frontend/optimization-remark-with-hotness.c that perhaps could be extended or leveraged?

Wed, Nov 18, 11:59 AM · Restricted Project, Restricted Project
tejohnson added a comment to D91410: [llvm][clang][mlir] Add checks for the return values from Target::createXXX to prevent protential null deref.

... the new checking is a mix of assert and fatal errors. Is that intended?

No. The added checks are based on the checks on other calls to the Target::createXXX functions in this file or other related files. If there are any fatal errors nearby (e.g. llvm/lib/LTO/ThinLTOCodeGenerator.cpp:581 vs 569), the check will be a fatal error; and if there are any assertions (e.g. llvm/lib/CodeGen/LLVMTargetMachine.cpp:43,45,52 vs 60) or the calls are never checked (e.g. llvm/lib/CodeGen/AsmPrinter/AsmPrinter.cpp:300), the added check will be an assertion.

If these are not likely due to user input issues, then perhaps they should all be assert so that they are compiled out in release compilers?

Since all these problems are reported by my static analyzer, I do not really know whether these checked pointers will actually be null when the code is executed. And I did not try to dynamically execute the program to check the problems either. But chances are that if the creator callbacks are not properly set or the called creator functions returns nullptr, the problem will happen. In my opinion, these problems may only happen during development. Therefore, I believe asserts can be sufficient to diagnose the problems.

If you think it would be better to use assertions instead of fatal errors, I will make an update on all llvm/lib/xxx files (or maybe all files). But before that, I'd prefer waiting for the replies from other reviewers on the remaining parts of the patch and making an update for all the suggestions.

Wed, Nov 18, 7:45 AM · Restricted Project, Restricted Project, Restricted Project

Tue, Nov 17

tejohnson requested review of D91676: Avoid redundant work when computing vtable vcall visibility.
Tue, Nov 17, 8:05 PM · Restricted Project
tejohnson added a comment to D91583: [lld] Allow --export-dynamic to override --lto-whole-program-visibility.

Hi, I am still learning the feature and I've just played a bit with the test. I've a couple of questions:

  • The --export-dynamic usage in the test seems a bit confusing. Are the VisibleToRegularObj bits of these _ZTV* symbols the important matter and --export-dynamic is just an approach to make them true?
Tue, Nov 17, 3:37 PM · Restricted Project
tejohnson added a comment to D91585: [NPM] Move more O0 pass building into PassBuilder.

Thanks, generally this seems to be good cleanup. Question on one of the changes below though.

Tue, Nov 17, 12:41 PM · Restricted Project, Restricted Project
tejohnson added a comment to D85808: [Remarks][2/2] Expand remarks hotness threshold option support in more tools.

Largely LGTM, except that the clang changes are missing a corresponding test.

Tue, Nov 17, 9:10 AM · Restricted Project, Restricted Project
tejohnson added a comment to D85809: [Remarks][1/2] Expand remarks hotness threshold option support in more tools.

LGTM, but please wait for @MaskRay since he had the most recent comments on this patch.

Tue, Nov 17, 9:09 AM · Restricted Project
tejohnson added a comment to D91410: [llvm][clang][mlir] Add checks for the return values from Target::createXXX to prevent protential null deref.

I was reviewing the llvm/lib/* changes and note that the new checking is a mix of assert (debug or assert enabled compiler only) and fatal errors (both debug and release compilers). Is that intended? If these are not likely due to user input issues, then perhaps they should all be assert so that they are compiled out in release compilers?

Tue, Nov 17, 7:59 AM · Restricted Project, Restricted Project, Restricted Project

Mon, Nov 16

tejohnson added a comment to D91583: [lld] Allow --export-dynamic to override --lto-whole-program-visibility.

I'll need to study what whole program visibility does (e.g. I need to read http://lists.llvm.org/pipermail/llvm-dev/2019-December/137543.html "RFC: Safe Whole Program Devirtualization Enablement").
My gut feeling is that: --export-dynamic is not sufficient to capture whether symbols are exported.
When linking an executable,

  • Symbols matched by a --dynamic-list pattern are exported to the dynamic symbol table
  • Symbols matched by a --export-dynamic-symbol pattern exported to the dynamic symbol table
  • Symbols defined in the executable which are referenced by a shared object are exported.

If the usage of --export-dynamic means whether the mode should apply to synthesized new symbols, I think having a mode referring to --export-dynamic makes sense.

Mon, Nov 16, 8:09 PM · Restricted Project
tejohnson requested review of D91583: [lld] Allow --export-dynamic to override --lto-whole-program-visibility.
Mon, Nov 16, 6:36 PM · Restricted Project

Fri, Nov 13

tejohnson requested review of D91474: [buildbot] Fix worker for ThinLTO whole program devirtualization.
Fri, Nov 13, 8:04 PM
tejohnson added a comment to D88433: [CGProfile] allows bitcast in metadata node storing function pointers.

LGTM again except for the patch title and summary which need an update. Go ahead and commit once those are updated.

Fri, Nov 13, 7:59 AM · Restricted Project
tejohnson added a comment to D79140: [ThinLTO] return error instead of crashing on invalid input.

Please address the earlier suggestions (needs a test case

The failure looks like a low-level profile ID mismatch when merging profiles obtained when running tests against slightly different sources. I'll need some help on how I can craft the equivalent using lit mechanism. Original 40MB compressed input to lld is at https://bugs.llvm.org/show_bug.cgi?id=45636#c1 if it's of any help.

Fri, Nov 13, 7:56 AM · Restricted Project

Thu, Nov 12

tejohnson added a comment to D79140: [ThinLTO] return error instead of crashing on invalid input.

Please address the earlier suggestions (needs a test case, and also agree it might be nice to have the message suggest stale objects as possible culprit).

Thu, Nov 12, 10:35 AM · Restricted Project
tejohnson accepted D88433: [CGProfile] allows bitcast in metadata node storing function pointers.

LGTM with a couple minor changes noted below, also the description needs an update as it still references fixing up the nodes.

Thu, Nov 12, 7:59 AM · Restricted Project

Wed, Nov 11

tejohnson committed rZORGdd1e514bf8aa: [llvm] New worker for ThinLTO whole program devirtualization (authored by tejohnson).
[llvm] New worker for ThinLTO whole program devirtualization
Wed, Nov 11, 3:02 PM
tejohnson closed D90578: [llvm] New worker for ThinLTO whole program devirtualization.
Wed, Nov 11, 3:02 PM

Tue, Nov 10

tejohnson added a comment to D90578: [llvm] New worker for ThinLTO whole program devirtualization.

@gkistanova did you want to look at this change? I'll go ahead and commit tomorrow since it has been accepted, unless I hear otherwise.

Tue, Nov 10, 8:22 AM
tejohnson added a comment to D91034: [Frontend] Treat CUDA header files as CUDA source files.

Not sure I'm the best person to review this either. But in any case, needs a test.

Tue, Nov 10, 8:21 AM · Restricted Project, Restricted Project

Nov 2 2020

tejohnson updated the diff for D90578: [llvm] New worker for ThinLTO whole program devirtualization.

Enable ccache

Nov 2 2020, 2:03 PM
tejohnson added inline comments to D90578: [llvm] New worker for ThinLTO whole program devirtualization.
Nov 2 2020, 2:03 PM
tejohnson added inline comments to D90578: [llvm] New worker for ThinLTO whole program devirtualization.
Nov 2 2020, 1:55 PM
tejohnson committed rGd1e02ea47a6a: [test] Fix unused FileCheck prefix in ThinLTO test (authored by tejohnson).
[test] Fix unused FileCheck prefix in ThinLTO test
Nov 2 2020, 9:07 AM
tejohnson committed rG7f32ddc99b30: [MemProf] Reenable test with fix for bot failures (authored by tejohnson).
[MemProf] Reenable test with fix for bot failures
Nov 2 2020, 9:00 AM
tejohnson closed D90466: [MemProf] Reenable test with fix for bot failures.
Nov 2 2020, 9:00 AM · Restricted Project

Nov 1 2020

tejohnson committed rG95824be18fcd: [MemProf] Fix test failure on windows (authored by tejohnson).
[MemProf] Fix test failure on windows
Nov 1 2020, 7:09 PM
tejohnson requested review of D90578: [llvm] New worker for ThinLTO whole program devirtualization.
Nov 1 2020, 6:05 PM
tejohnson committed rG0949f96dc652: [MemProf] Pass down memory profile name with optional path from clang (authored by tejohnson).
[MemProf] Pass down memory profile name with optional path from clang
Nov 1 2020, 5:39 PM
tejohnson closed D89087: [MemProf] Pass down memory profile name with optional path from clang.
Nov 1 2020, 5:39 PM · Restricted Project, Restricted Project, Restricted Project

Oct 31 2020

tejohnson updated the diff for D89087: [MemProf] Pass down memory profile name with optional path from clang.

Address comments

Oct 31 2020, 10:48 PM · Restricted Project, Restricted Project, Restricted Project
tejohnson added inline comments to D89087: [MemProf] Pass down memory profile name with optional path from clang.
Oct 31 2020, 10:47 PM · Restricted Project, Restricted Project, Restricted Project
tejohnson updated the diff for D89087: [MemProf] Pass down memory profile name with optional path from clang.

Rebase and fix bad merge

Oct 31 2020, 2:12 PM · Restricted Project, Restricted Project, Restricted Project

Oct 30 2020

tejohnson added a comment to D89087: [MemProf] Pass down memory profile name with optional path from clang.

There should be a related LLVM side of changes. Is it in a different patch?

Oct 30 2020, 5:00 PM · Restricted Project, Restricted Project, Restricted Project
tejohnson requested review of D90466: [MemProf] Reenable test with fix for bot failures.
Oct 30 2020, 8:14 AM · Restricted Project

Oct 29 2020

tejohnson added a comment to D90120: [sanitizer] Print errno for report file open failure.
Oct 29 2020, 3:28 PM · Restricted Project
tejohnson committed rGd124ac0c223a: [MemProf] Temporarily disable test failing on a couple bots (authored by tejohnson).
[MemProf] Temporarily disable test failing on a couple bots
Oct 29 2020, 11:27 AM
tejohnson committed rG240b4217382d: [MemProf] Augment test to debug avr bot failure (authored by tejohnson).
[MemProf] Augment test to debug avr bot failure
Oct 29 2020, 10:05 AM
tejohnson accepted D90366: [ThinLTO] Fix empty .llvmcmd sections.

LGTM

Oct 29 2020, 9:21 AM · Restricted Project
tejohnson committed rG81f7b96ed0a2: [sanitizer] Print errno for report file open failure (authored by tejohnson).
[sanitizer] Print errno for report file open failure
Oct 29 2020, 8:48 AM
tejohnson closed D90120: [sanitizer] Print errno for report file open failure.
Oct 29 2020, 8:47 AM · Restricted Project
tejohnson added a comment to D90366: [ThinLTO] Fix empty .llvmcmd sections.

The motivation and effects of the change are unclear to me. Does this mean that the .llvmcmd section is always emitted? Or always emitted whenever any bitcode embedding is requested?

Oct 29 2020, 7:56 AM · Restricted Project

Oct 28 2020

tejohnson added a comment to D84999: dynamic_cast optimization in LTO.

@pcc since some of the questions below related to type metadata and type intrinsics he added for WPD, and he's probably a good person to resolve some of the questions regarding where to do the CXX ABI handling.

Oct 28 2020, 6:40 PM · Restricted Project
tejohnson added a comment to D82745: [ThinLto] Fix Ifunc symbol usage.

Thanks for doing this and sorry for the slow reply! Some comments and suggested fixes below.

Oct 28 2020, 2:19 PM · Restricted Project
tejohnson added a comment to D88433: [CGProfile] allows bitcast in metadata node storing function pointers.

It seems a bit odd to have/need a custom fixup for one type of MDNode here. Can visitModuleFlagCGProfileEntry be fixed to handle the bitcast?

It is indeed a bit odd but that's close to where the artifact is produced. Debug metadata nodes involving globals/functions could potentially have this issue too (either their verification doesn't enforce similar logic as CGProfile node does or they just don't care I'm not sure).

Fixing visitModuleFlagCGProfileEntry works too although it addresses the consumer rather than the producer, which logically also seems odd. If this is the preferred method to address the issue, I have no strong opinion against it.

Oct 28 2020, 12:51 PM · Restricted Project
tejohnson accepted D90278: [ThinLTO] Fix .llvmcmd emission.

lgtm

Oct 28 2020, 12:48 PM · Restricted Project, Restricted Project
tejohnson accepted D90330: [NFC][ThinLTO] Change command line passing to EmbedBitcodeInModule.

lgtm except for the part that should go back into D90278 as noted below.

Oct 28 2020, 12:21 PM · Restricted Project, Restricted Project
tejohnson accepted D90278: [ThinLTO] Fix .llvmcmd emission.

LGTM but is there a functional reason why CmdArgs was changed to be passed by reference? If just generic cleanup might be better to split up that into a separate commit.

Oct 28 2020, 11:13 AM · Restricted Project, Restricted Project

Oct 26 2020

tejohnson committed rGba71a0746f31: [MemProf] Decouple memprof build from COMPILER_RT_BUILD_SANITIZERS (authored by tejohnson).
[MemProf] Decouple memprof build from COMPILER_RT_BUILD_SANITIZERS
Oct 26 2020, 1:53 PM
tejohnson closed D90190: [MemProf] Decouple memprof build from COMPILER_RT_BUILD_SANITIZERS.
Oct 26 2020, 1:53 PM · Restricted Project
tejohnson added inline comments to D90190: [MemProf] Decouple memprof build from COMPILER_RT_BUILD_SANITIZERS.
Oct 26 2020, 1:52 PM · Restricted Project
tejohnson requested review of D90190: [MemProf] Decouple memprof build from COMPILER_RT_BUILD_SANITIZERS.
Oct 26 2020, 1:44 PM · Restricted Project

Oct 25 2020

tejohnson requested review of D90120: [sanitizer] Print errno for report file open failure.
Oct 25 2020, 8:33 AM · Restricted Project

Oct 24 2020

tejohnson committed rG13c62ce99aab: [MemProf] Temporarily disable part of test (authored by tejohnson).
[MemProf] Temporarily disable part of test
Oct 24 2020, 11:09 PM