Page MenuHomePhabricator
Feed Advanced Search

Tue, Apr 20

ldrumm added a comment to D99334: [TransformUtils] Don't generate invalid llvm.dbg.cu .

Why would code handling llvm.dbg.cu need to add an if statement? I'd expect it to walk through whatever's in llvm.dbg.cu and if there's nothing it'd be fine with that?

Because 3 lines later we have auto *NMD = NewModule->getOrInsertNamedMetadata("llvm.dbg.cu");
In the case that the source module has no llvm.dbu.cu the target module ends up with on without operands. This is both illegal and a pain to deal with in user code.

I don't understand why it would be either illegal or a pain to deal with. (& I'm not sure it's illegal either - I modified an example IR file to have an empty llvm.dbg.cu list and it seemed to compile file (compared to other violations of IR debug info verification, which did result in an error/warning like this:

inlinable function call in a function with debug info must have a !dbg location
  call void @_Z2f1v()
warning: ignoring invalid debug info in test.ll
```)
Tue, Apr 20, 10:56 AM · Restricted Project
ldrumm added a comment to D99334: [TransformUtils] Don't generate invalid llvm.dbg.cu .

Why would code handling llvm.dbg.cu need to add an if statement? I'd expect it to walk through whatever's in llvm.dbg.cu and if there's nothing it'd be fine with that?

Tue, Apr 20, 8:19 AM · Restricted Project

Apr 1 2021

ldrumm added a comment to D99334: [TransformUtils] Don't generate invalid llvm.dbg.cu .

ping

Apr 1 2021, 9:27 AM · Restricted Project

Mar 31 2021

ldrumm added a comment to D99334: [TransformUtils] Don't generate invalid llvm.dbg.cu .

what was invalid about the metadata? The llvm.dbg.cu list was empty? I wonder if we should just say that's valid. It seems pretty benign/reasonable - makes it easier to write code like this, or code that, say, stripped a CU out wouldn't have to worry about removing a potentially zero-length cu list. @aprantl @JDevlieghere ?

Mar 31 2021, 4:08 AM · Restricted Project

Mar 30 2021

ldrumm added a reviewer for D84877: Support for soft fp16 to fp64 IEEE conversions: kito-cheng.

kito-cheng would you mind looking at the testing here and see whether the correctness testing of NaNs is sufficient?

Mar 30 2021, 7:42 AM · Restricted Project, Restricted Project
ldrumm added a comment to D98670: [RISCV] Pass 'half' in the lower 16 bits of an f32 value when F extension is enabled, but Zfh is not..

I just found an issue is unrelated this patch, but related to fp16, RISC-V GCC using the traditional libgcc function name scheme like __extendhfdf2(__<op><srcT><dstT><N_OP>) rather than __gnu_f2h_ieee (__gnu_*2*_ieee).

Function used in GCC:

  • __extendhfsf2 for half -> float
  • __truncsfhf2 for float -> half
  • __extendhfdf2 for double -> half
  • __truncdfhf2 for half -> double
Mar 30 2021, 3:11 AM · Restricted Project

Mar 25 2021

ldrumm updated the diff for D99334: [TransformUtils] Don't generate invalid llvm.dbg.cu .
Mar 25 2021, 6:07 AM · Restricted Project
ldrumm updated the diff for D99334: [TransformUtils] Don't generate invalid llvm.dbg.cu .
Mar 25 2021, 6:01 AM · Restricted Project
ldrumm updated the diff for D99334: [TransformUtils] Don't generate invalid llvm.dbg.cu .
Mar 25 2021, 6:01 AM · Restricted Project
ldrumm requested review of D99334: [TransformUtils] Don't generate invalid llvm.dbg.cu .
Mar 25 2021, 5:55 AM · Restricted Project

Mar 23 2021

ldrumm committed rG520f70e94d9c: [NFC] clang-format llvm/lib/Transforms/Utils/CloneFunction.cpp (authored by ldrumm).
[NFC] clang-format llvm/lib/Transforms/Utils/CloneFunction.cpp
Mar 23 2021, 5:54 AM
ldrumm committed rGab44ec1b22b7: [NFC] Minor refactor (authored by ldrumm).
[NFC] Minor refactor
Mar 23 2021, 5:54 AM
ldrumm committed rG0448ddd169ef: [NFCI] cleanup CloneFunctionInto (authored by ldrumm).
[NFCI] cleanup CloneFunctionInto
Mar 23 2021, 5:54 AM
ldrumm closed D98957: [NFCI] [TransformUtils] cleanup `CloneFunctionInto`.
Mar 23 2021, 5:53 AM · Restricted Project
ldrumm added a comment to D98957: [NFCI] [TransformUtils] cleanup `CloneFunctionInto`.

Thanks for cleaning this up!

Mostly LGTM, just please split this up into three separate commits (no need for new reviews; any order seems fine, as they seem fully independent):

Good idea. Will do

  • Unrelated whitespace changes (BTW, if you just got these incidentally, you can use git-clang-format in the future to just change the lines your commit is touching...).

Seeing as there is now a whitespace only commit, I've just opted to format the whole file as there are plently of other formatting issues picked up in this file

Mar 23 2021, 5:43 AM · Restricted Project

Mar 19 2021

ldrumm requested review of D98957: [NFCI] [TransformUtils] cleanup `CloneFunctionInto`.
Mar 19 2021, 8:49 AM · Restricted Project

Mar 16 2021

ldrumm committed rG440f6bdf34f4: [OpenCL][NFCI] Prefer CodeGenFunction::EmitRuntimeCall (authored by ldrumm).
[OpenCL][NFCI] Prefer CodeGenFunction::EmitRuntimeCall
Mar 16 2021, 9:23 AM
ldrumm closed D98705: [OpenCL] Prefer CodeGenFunction::EmitRuntimeCall.
Mar 16 2021, 9:23 AM · Restricted Project
ldrumm requested review of D98705: [OpenCL] Prefer CodeGenFunction::EmitRuntimeCall.
Mar 16 2021, 6:34 AM · Restricted Project
ldrumm added inline comments to D98411: [OpenCL] Respect calling convention for builtin .
Mar 16 2021, 5:20 AM · Restricted Project
ldrumm added a comment to D98411: [OpenCL] Respect calling convention for builtin .

Given this fixes a runtime crash bug, should it be backported to the relevant release branch?

Sure, thanks!

Mar 16 2021, 4:49 AM · Restricted Project

Mar 15 2021

ldrumm added a comment to D98411: [OpenCL] Respect calling convention for builtin .

Given this fixes a runtime crash bug, should it be backported to the relevant release branch?

Mar 15 2021, 10:32 AM · Restricted Project
ldrumm committed rGfcfd3fda7190: [OpenCL] Respect calling convention for builtin (authored by ldrumm).
[OpenCL] Respect calling convention for builtin
Mar 15 2021, 10:27 AM
ldrumm closed D98411: [OpenCL] Respect calling convention for builtin .
Mar 15 2021, 10:27 AM · Restricted Project

Mar 12 2021

ldrumm updated subscribers of D98411: [OpenCL] Respect calling convention for builtin .

@Anastasia @svenvh is this already covered by https://reviews.llvm.org/D98039 ?

Mar 12 2021, 3:38 AM · Restricted Project

Mar 11 2021

ldrumm requested review of D98411: [OpenCL] Respect calling convention for builtin .
Mar 11 2021, 4:00 AM · Restricted Project

Mar 4 2021

ldrumm added a comment to D97071: [lit] Add a TAP-formatted output mode to lit.

Ping

Mar 4 2021, 8:21 AM · Restricted Project

Feb 25 2021

ldrumm added reviewers for D97071: [lit] Add a TAP-formatted output mode to lit: jhenderson, thopre.
Feb 25 2021, 9:18 AM · Restricted Project

Feb 19 2021

ldrumm requested review of D97071: [lit] Add a TAP-formatted output mode to lit.
Feb 19 2021, 11:33 AM · Restricted Project

Nov 17 2020

ldrumm committed rG537cbd90c43d: Escape command line arguments in backtraces (authored by ldrumm).
Escape command line arguments in backtraces
Nov 17 2020, 4:16 AM
ldrumm closed D90759: Escape command line arguments in backtraces.
Nov 17 2020, 4:16 AM · Restricted Project

Nov 16 2020

ldrumm updated the diff for D90759: Escape command line arguments in backtraces.

Patch updated to not insert trailing blank

Nov 16 2020, 10:30 AM · Restricted Project

Nov 13 2020

ldrumm added a comment to D90759: Escape command line arguments in backtraces.

I absent mindedly responded to this review via email, without actually checking whether Phabricator supports this. Apologies for the delay

Nov 13 2020, 9:38 AM · Restricted Project

Nov 4 2020

ldrumm updated the diff for D90759: Escape command line arguments in backtraces.

included full diff context

Nov 4 2020, 5:45 AM · Restricted Project
ldrumm added a comment to D90759: Escape command line arguments in backtraces.

An example of the output:

Nov 4 2020, 4:51 AM · Restricted Project
ldrumm requested review of D90759: Escape command line arguments in backtraces.
Nov 4 2020, 4:51 AM · Restricted Project

Aug 25 2020

ldrumm added inline comments to D84877: Support for soft fp16 to fp64 IEEE conversions.
Aug 25 2020, 7:06 AM · Restricted Project, Restricted Project

Aug 21 2020

ldrumm updated the diff for D84877: Support for soft fp16 to fp64 IEEE conversions.
Aug 21 2020, 5:36 AM · Restricted Project, Restricted Project
ldrumm added inline comments to D84877: Support for soft fp16 to fp64 IEEE conversions.
Aug 21 2020, 4:56 AM · Restricted Project, Restricted Project

Aug 20 2020

ldrumm updated the diff for D84877: Support for soft fp16 to fp64 IEEE conversions.
Aug 20 2020, 10:14 AM · Restricted Project, Restricted Project
ldrumm added inline comments to D84877: Support for soft fp16 to fp64 IEEE conversions.
Aug 20 2020, 9:38 AM · Restricted Project, Restricted Project

Jul 29 2020

ldrumm requested review of D84877: Support for soft fp16 to fp64 IEEE conversions.
Jul 29 2020, 11:07 AM · Restricted Project, Restricted Project

Mar 12 2020

ldrumm committed rG0fa3320931e9: [lldb] reject `.debug_arange` sections with nonzero segment size (authored by ldrumm).
[lldb] reject `.debug_arange` sections with nonzero segment size
Mar 12 2020, 5:54 AM
ldrumm closed D75925: [lldb] reject `.debug_arange` sections with nonzero segment size.
Mar 12 2020, 5:54 AM · Restricted Project

Mar 11 2020

ldrumm updated the diff for D75925: [lldb] reject `.debug_arange` sections with nonzero segment size.

I found similar unittests for other DWARF entries which allow me to check more about the parser state, so I went with Pavel's suggestion for the testcase

Mar 11 2020, 10:09 AM · Restricted Project

Mar 10 2020

ldrumm created D75925: [lldb] reject `.debug_arange` sections with nonzero segment size.
Mar 10 2020, 8:39 AM · Restricted Project

Jan 27 2020

ldrumm committed rG482e890d1f94: [tablegen] Emit string literals instead of char arrays (authored by ldrumm).
[tablegen] Emit string literals instead of char arrays
Jan 27 2020, 10:29 AM
ldrumm closed D73044: {tablegen] Emit string literals instead of char arrays.
Jan 27 2020, 10:28 AM · Restricted Project
ldrumm reopened D73044: {tablegen] Emit string literals instead of char arrays.
Jan 27 2020, 7:43 AM · Restricted Project
ldrumm updated the diff for D73044: {tablegen] Emit string literals instead of char arrays.

Now emit long-string disabling GCC diagnostic pragmas around string definitions to silence gcc's -Woverlength-strings on a case-by-case basis

Jan 27 2020, 7:43 AM · Restricted Project

Jan 24 2020

ldrumm added inline comments to D73044: {tablegen] Emit string literals instead of char arrays.
Jan 24 2020, 4:08 AM · Restricted Project
ldrumm added inline comments to D73044: {tablegen] Emit string literals instead of char arrays.
Jan 24 2020, 4:08 AM · Restricted Project

Jan 23 2020

ldrumm updated the diff for D73044: {tablegen] Emit string literals instead of char arrays.
  • only pass the --long-string-literals flags to *LLVM* tablegen (fixes clang-tablegen problem)
  • suppress overlenth-string-literals warning seen in the wild: this behaviour is intended and shouldn't be warned against
Jan 23 2020, 9:40 AM · Restricted Project
ldrumm added a comment to D73044: {tablegen] Emit string literals instead of char arrays.

Hello @ldrumm,

Sorry, I had to revert this commit in e464b31c1565204e3be114d043bcbf4de61fe2e9, because it broke some Windows builds:
http://lab.llvm.org:8011/builders/clang-x64-windows-msvc/builds/13870

Jan 23 2020, 7:52 AM · Restricted Project
ldrumm committed rGce23515f5ab0: [tablegen] Emit string literals instead of char arrays (authored by ldrumm).
[tablegen] Emit string literals instead of char arrays
Jan 23 2020, 6:03 AM
ldrumm closed D73044: {tablegen] Emit string literals instead of char arrays.
Jan 23 2020, 6:03 AM · Restricted Project

Jan 21 2020

ldrumm updated the diff for D73044: {tablegen] Emit string literals instead of char arrays.

Remove compile-time detection. Make this a runtime only flag

Jan 21 2020, 11:37 AM · Restricted Project
ldrumm added a comment to D73044: {tablegen] Emit string literals instead of char arrays.

First char of the title: { -> [.

Thanks. Phabricator doesn't seem to track the git commit message, but I've fixed this is on my tree

Jan 21 2020, 8:36 AM · Restricted Project
mstojanovic awarded D73044: {tablegen] Emit string literals instead of char arrays a Love token.
Jan 21 2020, 3:46 AM · Restricted Project
ldrumm updated the diff for D73044: {tablegen] Emit string literals instead of char arrays.

Change the cmake flag check to if(MSVC) rather than if(CMAKE_GENERATOR MATCHES "Visual Studio"). CMAKE_GENERATOR is not the compiler

Jan 21 2020, 3:41 AM · Restricted Project

Jan 20 2020

ldrumm updated the diff for D73044: {tablegen] Emit string literals instead of char arrays.
Jan 20 2020, 10:07 AM · Restricted Project
ldrumm updated the diff for D73044: {tablegen] Emit string literals instead of char arrays.

Ensure elaborate cross-compilation setups are supported

Jan 20 2020, 10:07 AM · Restricted Project
ldrumm created D73044: {tablegen] Emit string literals instead of char arrays.
Jan 20 2020, 7:03 AM · Restricted Project

Jan 15 2020

ldrumm abandoned D22060: [expression evaluation] Prevent invalid function declarations ending up in clang AST.
Jan 15 2020, 6:48 AM

Jan 4 2017

ldrumm committed rL290957: Add a debuginfo version check for RenderScript modules.
Add a debuginfo version check for RenderScript modules
Jan 4 2017, 4:21 AM

Dec 22 2016

ldrumm committed rL290359: Fix a couple of incorrect format string warnings.
Fix a couple of incorrect format string warnings
Dec 22 2016, 11:25 AM
ldrumm closed D28028: Fix a couple of incorrect format strings by committing rL290359: Fix a couple of incorrect format string warnings.
Dec 22 2016, 11:25 AM
ldrumm updated the diff for D28028: Fix a couple of incorrect format strings.

switched to the new llvm::Format API

Dec 22 2016, 5:22 AM

Dec 21 2016

ldrumm retitled D28028: Fix a couple of incorrect format strings from to Fix a couple of incorrect format strings.
Dec 21 2016, 12:30 PM
ldrumm updated the diff for D17719: Track expression language from one place in ClangExpressionParser.

This patch was committed back in March https://reviews.llvm.org/rL263099 and then reverted 2 hours later in https://reviews.llvm.org/rL263107 due to a testing failure on greendragon related to ObjC. Unfortunately, I no longer have access to Apple hardware (and at the time the testsuite passed on my local mac), so I'm unable to find out exactly what went wrong.

Dec 21 2016, 11:50 AM · Restricted Project
ldrumm reopened D17719: Track expression language from one place in ClangExpressionParser.

This patch was committed back in March https://reviews.llvm.org/rL263099 and then reverted 2 hours later in https://reviews.llvm.org/rL263107 due to a testing failure on greendragon related to ObjC. Unfortunately, I no longer have access to Apple hardware (and at the time the testsuite passed on my local mac), so I'm unable to find out exactly what went wrong.

Dec 21 2016, 11:49 AM · Restricted Project

Dec 19 2016

ldrumm abandoned D20825: Don't force Python2.7.
Dec 19 2016, 10:37 AM
ldrumm abandoned D22370: Respect LLVM_HOST_TRIPLE when manually specified.
Dec 19 2016, 10:36 AM
ldrumm committed rL290117: Expression evaluation for overloaded C functions (redux).
Expression evaluation for overloaded C functions (redux)
Dec 19 2016, 9:33 AM
ldrumm closed D27223: Expression evaluation for overloaded C functions (redux) by committing rL290117: Expression evaluation for overloaded C functions (redux).
Dec 19 2016, 9:33 AM

Dec 16 2016

ldrumm committed rL289956: Fix broken escaping of commands in the build.
Fix broken escaping of commands in the build
Dec 16 2016, 8:49 AM
ldrumm closed D26757: Fix broken escaping of commands in the build by committing rL289956: Fix broken escaping of commands in the build.
Dec 16 2016, 8:49 AM
ldrumm added inline comments to D27223: Expression evaluation for overloaded C functions (redux).
Dec 16 2016, 7:06 AM
ldrumm updated the diff for D27223: Expression evaluation for overloaded C functions (redux).

updated to use StringRef-based parser rather than char pointers

Dec 16 2016, 7:06 AM

Dec 15 2016

ldrumm added a comment to D26757: Fix broken escaping of commands in the build.

ping

Dec 15 2016, 9:46 AM

Nov 29 2016

ldrumm added a comment to D17957: Expression evaluation for overloaded C functions.

New candidate is here

Nov 29 2016, 12:18 PM
ldrumm retitled D27223: Expression evaluation for overloaded C functions (redux) from to Expression evaluation for overloaded C functions (redux).
Nov 29 2016, 12:16 PM
ldrumm updated the diff for D26757: Fix broken escaping of commands in the build.

Addressed some unwanted whitespace changes as suggested by @bryant; use os.dir.curdir as instead of .

Nov 29 2016, 8:24 AM

Nov 18 2016

ldrumm added a comment to D20825: Don't force Python2.7.

@chandlerc any more thoughts on this?

Nov 18 2016, 8:22 AM

Nov 16 2016

ldrumm added inline comments to D26757: Fix broken escaping of commands in the build.
Nov 16 2016, 12:31 PM
ldrumm added inline comments to D26757: Fix broken escaping of commands in the build.
Nov 16 2016, 12:11 PM
ldrumm added inline comments to D26757: Fix broken escaping of commands in the build.
Nov 16 2016, 12:09 PM
ldrumm added inline comments to D26757: Fix broken escaping of commands in the build.
Nov 16 2016, 12:06 PM
ldrumm added inline comments to D26757: Fix broken escaping of commands in the build.
Nov 16 2016, 12:04 PM
ldrumm added inline comments to D26757: Fix broken escaping of commands in the build.
Nov 16 2016, 11:43 AM
ldrumm retitled D26757: Fix broken escaping of commands in the build from to Fix broken escaping of commands in the build.
Nov 16 2016, 11:37 AM

Nov 9 2016

ldrumm committed rL286387: [RenderScript] Fix reduction argument printing.
[RenderScript] Fix reduction argument printing
Nov 9 2016, 11:30 AM

Oct 13 2016

ldrumm added inline comments to D25396: Fix Python binding generation build step on Windows.
Oct 13 2016, 7:30 AM · Restricted Project

Oct 5 2016

ldrumm committed rL283362: Add the ability to set breakpoints on named RenderScript reductions.
Add the ability to set breakpoints on named RenderScript reductions
Oct 5 2016, 12:19 PM
ldrumm committed rL283335: [RenderScript] reflow/reword some comments and normalize names.
[RenderScript] reflow/reword some comments and normalize names
Oct 5 2016, 9:36 AM
ldrumm committed rL283320: cleanup RSCoordinate handling and factor out coordinate parser.
cleanup RSCoordinate handling and factor out coordinate parser
Oct 5 2016, 7:43 AM
ldrumm committed rL283313: Delete unused global in ClangExpressionVariable.cpp.
Delete unused global in ClangExpressionVariable.cpp
Oct 5 2016, 5:50 AM
ldrumm closed D24793: Delete an (apparently unused) global in ClangExpressionVaraible.cpp by committing rL283313: Delete unused global in ClangExpressionVariable.cpp.
Oct 5 2016, 5:49 AM · Restricted Project

Sep 21 2016

ldrumm committed rL282066: Remove an invalid doxygen `@return` docstring on a void function.
Remove an invalid doxygen `@return` docstring on a void function
Sep 21 2016, 4:21 AM
ldrumm retitled D24793: Delete an (apparently unused) global in ClangExpressionVaraible.cpp from to Delete an (apparently unused) global in ClangExpressionVaraible.cpp.
Sep 21 2016, 4:06 AM · Restricted Project