[PATCH] D158883: [Matrix] Try to emit fmuladd for both vector and matrix types

thegameg added a comment. In D158883#4635997 , @uweigand wrote: > The newly added test cases in ffp-model.c fail on SystemZ, making CI red: Should be fixed, thanks for the report and sorry for the delay. Repository: rG LLVM Github Monorepo CHANGES S

[PATCH] D158883: [Matrix] Try to emit fmuladd for both vector and matrix types

This revision was automatically updated to reflect the committed changes. Closed by commit rGc987f9d7fdc7: [Matrix] Try to emit fmuladd for both vector and matrix types (authored by thegameg). Herald added a project: clang. Herald added a subscriber: cfe-commits. Repository: rG LLVM Github Mono

[PATCH] D156515: [RemarkUtil] Refactor llvm-remarkutil to include size-diff

thegameg accepted this revision. thegameg added a comment. This revision is now accepted and ready to land. Clean, thank you! Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D156515/new/ https://reviews.llvm.org/D156515 __

[PATCH] D153746: [Remarks] Make sure -fdiagnostics-hotness-threshold implies -fdiagnostics-show-hotness

thegameg accepted this revision. thegameg added a comment. This revision is now accepted and ready to land. Makes sense, thanks. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D153746/new/ https://reviews.llvm.org/D153746 ___

[PATCH] D135488: [codegen] Add a remarks based Stack Layout Analysis pass

thegameg added a comment. In D135488#4049050 , @nickdesaulniers wrote: > In D135488#4049035 , @paulkirth > wrote: > >> Actually if we add >> >> if (!isFunctionInPrintList(MF.getName())) >>return false;

[PATCH] D135488: [codegen] Add a remarks based Stack Layout Analysis pass

thegameg added a comment. In D135488#4048854 , @paulkirth wrote: > In D135488#4048380 , > @nickdesaulniers wrote: > >> It would be really nice if we could limit this to a specific function >> somehow. > > I thin

[PATCH] D135488: [codegen] Add a remarks based Stack Layout Analysis pass

thegameg accepted this revision. thegameg added a comment. This revision is now accepted and ready to land. Looks great with the leftover minor changes, feel free to land this, thanks! I'll give this a try internally and provide feedback if any. Comment at: llvm/lib/CodeGen/St

[PATCH] D135488: [codegen] Add a remarks based Stack Layout Analysis pass

thegameg added a comment. In D135488#4044437 , @paulkirth wrote: > BTW, is there a way to nest some of the items? Ideally we'd be able to have a > `Slot` in the YAML that contains all the various data, similar to how > `DebugLoc` is a more complex objec

[PATCH] D135488: [codegen] Add a remarks based Stack Layout Analysis pass

thegameg added a comment. This looks great, thanks for updating this! A few more comments inline. Comment at: llvm/lib/CodeGen/StackFrameLayoutAnalysisPass.cpp:55 +Variable, // a Slot used to store a local data (could be a tmp) +Error // Its an error for

[PATCH] D135488: [codegen] Add a remarks based Stack Layout Analysis pass

thegameg added a comment. I don't think I understand why we can't achieve B with remarks? In C and D you generate one remark for each line, can't we generate a single multi-line remark instead? Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D135488/

[PATCH] D135488: [codegen] Add a remarks based Stack Layout Analysis pass

thegameg added a comment. In D135488#4024713 , @paulkirth wrote: > @arsenm @thegameg @nickdesaulniers @dblaikie @phosek Can we reach a consensus > here on how to output this kind of information? I feel like I've been told to > move towards remarks as th

[PATCH] D135488: [codegen] Add a remarks based Stack Layout Analysis pass

thegameg added inline comments. Comment at: llvm/lib/CodeGen/StackFrameLayoutAnalysisPass.cpp:92 + &MF.front()) + << "FunctionName: " << ore::NV("FunctionName", MF.getName()); +}); Why are we emitting

[PATCH] D135488: [codegen] Add a remarks based Stack Layout Analysis pass

thegameg added a comment. This is great! Any chance we can use `MachineFrameInfo::StackProtectorIdx` to annotate the slot that is reserved for the stack protector? Comment at: llvm/lib/CodeGen/StackFrameLayoutAnalysisPass.cpp:84 +ORE = &getAnalysis().getORE(); +if (!OR

[PATCH] D100901: [CMake][llvm] avoid conflict w/ (and use when available) new builtin check_linker_flag

This revision was automatically updated to reflect the committed changes. Closed by commit rG06d06f2f6403: [CMake][llvm] avoid conflict w/ (and use when available) new builtin… (authored by radford, committed by thegameg). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https:/

[PATCH] D97499: [PM] Show the pass argument in pre/post-pass IR dumps

thegameg accepted this revision. thegameg added a comment. This revision is now accepted and ready to land. Thank you! Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D97499/new/ https://reviews.llvm.org/D97499 ___

[PATCH] D91444: [InstCombine] Preserve !annotation metadata for memory combines.

thegameg accepted this revision. thegameg added a comment. This revision is now accepted and ready to land. LGTM, nice! Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D91444/new/ https://reviews.llvm.org/D91444 __

[PATCH] D91417: [IRGen] Add !annotation metadata for auto-init stores.

thegameg accepted this revision. thegameg added a comment. Looks good, thanks! Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D91417/new/ https://reviews.llvm.org/D91417 ___ cfe-commits mailing list cfe-co

[PATCH] D85810: [clang] Pass-through remarks options to linker

thegameg accepted this revision. thegameg added a comment. LGTM Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D85810/new/ https://reviews.llvm.org/D85810 ___ cfe-commits mailing list cfe-commits@lists.llv

[PATCH] D82213: [Remarks] Add callsite locations to inline remarks

thegameg added subscribers: fhahn, anemet, thegameg. thegameg added a comment. This sounds useful indeed. @fhahn, @anemet might want to take a look. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D82213/new/ https://reviews.llvm.org/D82213 ___

[PATCH] D76916: [Darwin] Respect -fno-unroll-loops during LTO.

thegameg accepted this revision. thegameg added a comment. This revision is now accepted and ready to land. LGTM, thanks! Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D76916/new/ https://reviews.llvm.org/D76916 __

[PATCH] D74094: [IRGen] Emit lifetime intrinsics around temporary aggregate argument allocas

thegameg added inline comments. Comment at: clang/lib/CodeGen/CGCall.cpp:3687 - args.add(EmitAnyExprToTemp(E), type); } Is there any other use of `EmitAnyExprToTemp` that can benefit from this? CHANGES SINCE LAST ACTION https://reviews.llvm.org/D74094/new

[PATCH] D73676: [Remarks] Extend the RemarkStreamer to support other emitters

This revision was automatically updated to reflect the committed changes. thegameg marked 2 inline comments as done. Closed by commit rG7531a5039fd7: [Remarks] Extend the RemarkStreamer to support other emitters (authored by thegameg). Changed prior to commit: https://reviews.llvm.org/D73676?vs

[PATCH] D73676: [Remarks] Extend the RemarkStreamer to support other emitters

thegameg created this revision. thegameg added reviewers: paquette, anemet, JDevlieghere, hfinkel, fhahn. thegameg added projects: LLVM, clang. Herald added subscribers: dang, dexonsmith, steven_wu, hiraditya, mgorny, mehdi_amini. This extends the RemarkStreamer to allow for other emitters (e.g.

[PATCH] D73495: [CodeGen] Attach no-builtin attributes to function definitions with no Decl

This revision was automatically updated to reflect the committed changes. Closed by commit rG4e799ada5860: [CodeGen] Attach no-builtin attributes to function definitions with no Decl (authored by thegameg). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.or

[PATCH] D73495: [CodeGen] Attach no-builtin attributes to function definitions with no Decl

thegameg updated this revision to Diff 240914. thegameg added a comment. Add the attribute to all TargetDecls. CHANGES SINCE LAST ACTION https://reviews.llvm.org/D73495/new/ https://reviews.llvm.org/D73495 Files: clang/lib/CodeGen/CGCall.cpp clang/test/CodeGenCXX/global-init.cpp Index:

[PATCH] D73495: [CodeGen] Attach no-builtin attributes to function definitions with no Decl

thegameg marked 2 inline comments as done. thegameg added inline comments. Comment at: clang/lib/CodeGen/CGCall.cpp:1917 const auto *NBA = Fn->getAttr(); -bool HasWildcard = NBA && llvm::is_contained(NBA->builtinNames(), "*"); -if (getLangOpts().NoBuiltin

[PATCH] D73495: [CodeGen] Attach no-builtin attributes to functions with no Decl

thegameg marked an inline comment as done. thegameg added inline comments. Comment at: clang/lib/CodeGen/CGCall.cpp:1917 const auto *NBA = Fn->getAttr(); -bool HasWildcard = NBA && llvm::is_contained(NBA->builtinNames(), "*"); -if (getLangOpts().NoBuiltin

[PATCH] D73495: [CodeGen] Attach no-builtin attributes to functions with no Decl

thegameg created this revision. thegameg added reviewers: gchatelet, tejohnson, aaron.ballman, efriedma. thegameg added a project: clang. When using -fno-builtin[-], we don't attach the IR attributes to functions with no Decl. This results in projects using -fno-builtin or -ffreestanding to star

[PATCH] D71301: [clang][IFS] Prevent Clang-IFS from Leaking symbols from inside a block.

thegameg added a comment. In D71301#1832801 , @plotfi wrote: > So, the culprit appears to be D69825 . The > way InterfaceStubs assembles the pipeline appears to trigger an asan bug. For > now I will alter the tests to get

[PATCH] D71301: [clang][IFS] Prevent Clang-IFS from Leaking symbols from inside a block.

thegameg added a comment. Hi @plotfi, this seems to cause failures with ASAN and UBSAN on green dragon: http://green.lab.llvm.org/green/job/clang-stage2-cmake-RgSan/6886/consoleFull The following tests are failing: Clang :: InterfaceStubs/driver-test.c Clang :: InterfaceStubs/driver-test2.c

[PATCH] D71675: [Remarks][Driver] Run dsymutil when remarks are enabled

This revision was automatically updated to reflect the committed changes. Closed by commit rGd79b11fefb8e: [Remarks][Driver] Run dsymutil when remarks are enabled (authored by thegameg). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D71675/new/ https

[PATCH] D71675: [Remarks][Driver] Run dsymutil when remarks are enabled

thegameg marked an inline comment as done. thegameg added inline comments. Comment at: clang/test/Driver/darwin-opt-record.c:21 +// -gline-tables-only and would need -fno-save-optimization-record to +// completely disable it. +// CHECK-DSYMUTIL-G0: "-cc1" The oth

[PATCH] D71675: [Remarks][Driver] Run dsymutil when remarks are enabled

thegameg created this revision. thegameg added reviewers: friss, JDevlieghere, aprantl. thegameg added projects: clang, debug-info. When clang is invoked with a source file without -c or -S, it creates a cc1 job, a linker job and if debug info is requested, a dsymutil job. In case of remarks, we

[PATCH] D71325: [Remarks][Driver] Ask for line tables when remarks are enabled

This revision was automatically updated to reflect the committed changes. Closed by commit rG60590b149b33: [Remarks][Driver] Ask for line tables when remarks are enabled (authored by thegameg). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D71325/new/

[PATCH] D71325: [Remarks][Driver] Ask for line tables when remarks are enabled

thegameg updated this revision to Diff 233459. CHANGES SINCE LAST ACTION https://reviews.llvm.org/D71325/new/ https://reviews.llvm.org/D71325 Files: clang/lib/Driver/ToolChains/Clang.cpp clang/test/Driver/debug-options.c Index: clang/test/Driver/debug-options.c ==

[PATCH] D71325: [Remarks][Driver] Ask for line tables when remarks are enabled

thegameg created this revision. thegameg added reviewers: JDevlieghere, friss, aprantl. Herald added a project: clang. Serialized remarks contain debug locations for each remark, by storing a file path, a line, and a column. Also, remarks support being embedded in a .dSYM bundle using a separate

[PATCH] D70748: [clang test] Do not assume default target

thegameg accepted this revision. thegameg added a comment. This revision is now accepted and ready to land. LGTM, thanks for fixing this. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D70748/new/ https://reviews.llvm.org/D70748 __

[PATCH] D68611: [IRGen] Emit lifetime markers for temporary struct allocas

This revision was automatically updated to reflect the committed changes. thegameg marked an inline comment as done. Closed by commit rG143f6b837790: [IRGen] Emit lifetime markers for temporary struct allocas (authored by thegameg). Herald added a project: clang. Herald added a subscriber: cfe-com

[PATCH] D67683: [Timers] Fix printing some `-ftime-report` sections twice. Fixes PR40328.

thegameg accepted this revision. thegameg added a comment. This revision is now accepted and ready to land. LGTM, thanks! CHANGES SINCE LAST ACTION https://reviews.llvm.org/D67683/new/ https://reviews.llvm.org/D67683 ___ cfe-commits mailing list

[PATCH] D65000: [ARM] Set default alignment to 64bits

thegameg added a comment. In D65000#1616368 , @dnsampaio wrote: > I have tested this in our MacOS and linux environments. @thakis @thegameg > @phosek, would it be possible for you to check if this works for you? I just built trunk with this patch applie

[PATCH] D65000: [ARM] Set default alignment to 64bits

thegameg added inline comments. Comment at: cfe/trunk/test/CodeGen/ARM/exception-alignment.cpp:9 +// A16-NEXT: store <2 x i64> , <2 x i64>* [[BC]], align 16 +#include + thakis wrote: > This fails on some bots: > > http://lab.llvm.org:8011/builders/clang-atom-d5

[PATCH] D63518: BitStream reader: propagate errors

thegameg added inline comments. Comment at: llvm/include/llvm/Bitcode/BitstreamReader.h:489 bool ReadBlockEnd() { if (BlockScope.empty()) return true; jfb wrote: > thegameg wrote: > > Any reason why this doesn't return `Error`? > I'm not sure it's reall

[PATCH] D63518: BitStream reader: propagate errors

thegameg added inline comments. Comment at: llvm/include/llvm/Bitcode/BitstreamReader.h:441 // If we found a sub-block, just skip over it and check the next entry. - if (SkipBlock()) -return BitstreamEntry::getError(); + if (llvm::Error Err = SkipBlock())

[PATCH] D62709: Fix -DBUILD_SHARED_LIBS=ON build after rL362160

thegameg accepted this revision. thegameg added a comment. This revision is now accepted and ready to land. Sorry about this! This LGTM, thanks for fixing it! Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D62709/new/ https://reviews.llvm.org/D62709

[PATCH] D58996: [Remarks] Refactor remark diagnostic emission in a RemarkStreamer

This revision was automatically updated to reflect the committed changes. Closed by commit rC355507: [Remarks] Refactor remark diagnostic emission in a RemarkStreamer (authored by thegameg, committed by ). Herald added a project: clang. Herald added a subscriber: cfe-commits. Changed prior to com

[PATCH] D55775: [Driver] Don't override '-march' when using '-arch x86_64h'

thegameg added a comment. In D55775#1338512 , @qcolombet wrote: > Should we emit an error if we request x86_64h with an arch older than haswell? Makes sense. I'll put up a patch soon. Repository: rL LLVM CHANGES SINCE LAST ACTION https://reviews.l

[PATCH] D55775: [Driver] Don't override '-march' when using '-arch x86_64h'

This revision was automatically updated to reflect the committed changes. Closed by commit rL349381: [Driver] Don't override '-march' when using '-arch x86_64h' (authored by thegameg, committed by ). Herald added a subscriber: llvm-commits. Changed prior to commit: https://reviews.llvm.org/D557

[PATCH] D55775: [Driver] Don't override '-march' when using '-arch x86_64h'

thegameg created this revision. thegameg added reviewers: vsk, dexonsmith, ab, t.p.northover, Gerolf. On Darwin, using '-arch x86_64h' would always override the option passed through '-march'. This patch allows users to use '-march' with x86_64h, while keeping the default to 'core-avx2' https

[PATCH] D44093: [BUILTINS] structure pretty printer

thegameg added a comment. Thanks for working on this! Few remarks in the comments. Comment at: lib/CodeGen/CGBuiltin.cpp:934 +static Value *dumpRecord(CodeGenFunction &CGF, QualType RType, + Value*& RecordPtr, CharUnits Align, `llvm::V

[PATCH] D26953: clang-format: handle formatting on constexpr if

thegameg added a comment. In https://reviews.llvm.org/D26953#783497, @jtbandes wrote: > Hm, I probably should've searched first — but I just re-implemented this in > https://reviews.llvm.org/D34330. Actually, I think my implementation solves > the `AllowShortIfStatementsOnASingleLine` issue yo

[PATCH] D26953: clang-format: handle formatting on constexpr if

thegameg added a comment. In https://reviews.llvm.org/D26953#751021, @thakis wrote: > This looks good to me, thanks. Sorry about the slow turnaround. Do you have > commit access? If not, I can land it for you – but it also looks like you've > contributed several patches by now, so you could als