rnk (Reid Kleckner)
User

Projects

User does not belong to any projects.

User Details

User Since
Jan 2 2013, 4:34 PM (246 w, 4 d)

Recent Activity

Thu, Sep 21

rnk accepted D37033: [llvm-rc] Add user-defined resources parsing ability. [8/8].

lgtm

Thu, Sep 21, 3:29 PM
rnk added inline comments to D37022: [llvm-rc] Add integer expressions parsing ability. [7/8].
Thu, Sep 21, 3:28 PM
rnk accepted D37022: [llvm-rc] Add integer expressions parsing ability. [7/8].

lgtm

Thu, Sep 21, 3:28 PM
rnk accepted D37283: [llvm-rc] Serialize HTML resources to .res files (serialization, pt 1)..

lgtm, thanks!

Thu, Sep 21, 2:11 PM
rnk committed rL313905: Re-land r313825: "[IR] Add llvm.dbg.addr, a control-dependent version of llvm..
Re-land r313825: "[IR] Add llvm.dbg.addr, a control-dependent version of llvm.
Thu, Sep 21, 12:53 PM
rnk added a comment to D37740: [SelectionDAG] Pick correct frame index in LowerArguments.

Seems ok to me.
@rnk: was your concern addressed?

Thu, Sep 21, 9:53 AM

Wed, Sep 20

rnk accepted D38109: [fixup][Sema] Allow in C to define tags inside enumerations..

lgtm

Wed, Sep 20, 4:31 PM
rnk committed rL313826: Commit local changes that missed llvm.dbg.addr.
Commit local changes that missed llvm.dbg.addr
Wed, Sep 20, 2:58 PM
rnk committed rL313825: [IR] Add llvm.dbg.addr, a control-dependent version of llvm.dbg.declare.
[IR] Add llvm.dbg.addr, a control-dependent version of llvm.dbg.declare
Wed, Sep 20, 2:54 PM
rnk closed D37768: [IR] Add llvm.dbg.addr, a control-dependent version of llvm.dbg.declare by committing rL313825: [IR] Add llvm.dbg.addr, a control-dependent version of llvm.dbg.declare.
Wed, Sep 20, 2:54 PM
rnk accepted D38092: [MS Compat]Allow __interfaces to have properties..

lgtm

Wed, Sep 20, 2:44 PM
rnk updated the diff for D37932: [DebugInfo] Correctly coalesce DBG_VALUEs that mix direct and indirect values.
  • rebase
Wed, Sep 20, 12:02 PM
rnk retitled D37932: [DebugInfo] Correctly coalesce DBG_VALUEs that mix direct and indirect values from [DebugInfo] Move the "indirect" bit to the location interval map to [DebugInfo] Correctly coalesce DBG_VALUEs that mix direct and indirect values.
Wed, Sep 20, 12:01 PM
rnk added a comment to D37957: [TableGen] Some simple optimizations to TableGen execution time.

lgtm, just nits

Wed, Sep 20, 11:29 AM
rnk committed rL313786: Re-land "[DebugInfo] Insert DW_OP_deref when spilling indirect DBG_VALUEs".
Re-land "[DebugInfo] Insert DW_OP_deref when spilling indirect DBG_VALUEs"
Wed, Sep 20, 11:20 AM
rnk committed rL313775: [DebugInfo] Use a MapVector to coalesce MachineOperand locations.
[DebugInfo] Use a MapVector to coalesce MachineOperand locations
Wed, Sep 20, 10:34 AM
rnk closed D38068: [DebugInfo] Use a DenseMap to coalesce MachineOperand locations by committing rL313775: [DebugInfo] Use a MapVector to coalesce MachineOperand locations.
Wed, Sep 20, 10:34 AM
rnk added inline comments to D38068: [DebugInfo] Use a DenseMap to coalesce MachineOperand locations.
Wed, Sep 20, 10:29 AM
rnk accepted D37900: [SROA] Really remove associated dbg.declare when removing dead alloca.

lgtm, I actually made this change as part of llvm.dbg.addr (D37768) and then reverted it to keep the diff minimal. Thanks!

Wed, Sep 20, 10:07 AM

Tue, Sep 19

rnk accepted D38058: [llvm-readobj] Teach readobj to output .res (WindowsResource)..

lgtm, thanks!

Tue, Sep 19, 6:09 PM
rnk created D38068: [DebugInfo] Use a DenseMap to coalesce MachineOperand locations.
Tue, Sep 19, 6:06 PM
rnk committed rL313679: Revert "[DebugInfo] Insert DW_OP_deref when spilling indirect DBG_VALUEs".
Revert "[DebugInfo] Insert DW_OP_deref when spilling indirect DBG_VALUEs"
Tue, Sep 19, 2:20 PM
rnk added a comment to D37768: [IR] Add llvm.dbg.addr, a control-dependent version of llvm.dbg.declare.

ping

Tue, Sep 19, 1:28 PM
rnk added inline comments to D37466: D37461: fixups for existing InlineAsm tests + adding new ones.
Tue, Sep 19, 12:00 PM
rnk accepted D37413: [X86][MS-InlineAsm] Extended support for variables / identifiers on memory / immediate expressions.

lgtm, thanks!

Tue, Sep 19, 11:59 AM
rnk committed rL313657: Re-land "Fix Bug 30978 by emitting cv file checksums.".
Re-land "Fix Bug 30978 by emitting cv file checksums."
Tue, Sep 19, 11:16 AM
rnk committed rL313640: Re-land r313400 "[DebugInfo] Insert DW_OP_deref when spilling indirect….
Re-land r313400 "[DebugInfo] Insert DW_OP_deref when spilling indirect…
Tue, Sep 19, 9:33 AM
rnk requested changes to D38016: [lit] Make lit stop writing .pyc files.

Are you sure you can't reproduce this behavior by checking out the lit subtree with SVN before you deleted that directory, importing some .py files from that tree to create .pycs, and then doing svn up and checking if it creates a conflict? I'd really prefer to have an explanatory comment in here that says why we're setting this mysterious configuration.

Tue, Sep 19, 8:48 AM

Mon, Sep 18

rnk committed rL313571: [gcov] Emit errors when opening the notes file fails.
[gcov] Emit errors when opening the notes file fails
Mon, Sep 18, 2:33 PM
rnk accepted D37865: [Sema] Fix a pair of crashes when generating exception specifiers with an error'ed field for a template class' default ctor..

Looks like the correct fix, thanks!

Mon, Sep 18, 2:02 PM
rnk added inline comments to D37461: [X86][AsmParser] re-introduce 'offset' operator.
Mon, Sep 18, 1:59 PM
rnk accepted D37986: [ubsan] Fix interface_symbols_windows test.

lgtm

Mon, Sep 18, 11:44 AM
rnk added a comment to D31951: TableGen support for parameterized register class information.

Please see https://bugs.llvm.org/show_bug.cgi?id=28222#c20 for a discussion of the performance implications of this patch. TL;DR - This has introduced a severe performance regression in tblgen (on the order of 4x slowdown) even for architectures such as X86 which according to the patch description should not be affected by this change.

Unless some algorithmic fix can be found relatively quickly, I'd like to open the possibility of reverting this patch until it can be re-worked to have more desirable performance characteristics, or at the very least be limited only to specific architectures.

Mon, Sep 18, 10:24 AM

Fri, Sep 15

rnk added inline comments to D37283: [llvm-rc] Serialize HTML resources to .res files (serialization, pt 1)..
Fri, Sep 15, 3:45 PM
rnk abandoned D37311: [DebugInfo] Lower dbg.declare to DBG_VALUE with DW_OP_deref.

Closing this, I now realize this was a step in the wrong direction.

Fri, Sep 15, 3:31 PM
rnk added a comment to D32049: [Preprocessor] Implement empty vs omitted __VA_ARGS__ string-expansion MSVC Extension.

Would this help with https://bugs.llvm.org/show_bug.cgi?id=32021 ?

Fri, Sep 15, 3:13 PM
rnk added a comment to D37911: [DebugInfo] Insert DW_OP_deref when spilling indirect DBG_VALUEs.

Thanks! Are you running the debuginfo-tests on your changes prior to committing?

Fri, Sep 15, 3:12 PM
rnk committed rL313405: Name the sentinel value used for the location number of the undefined register….
Name the sentinel value used for the location number of the undefined register…
Fri, Sep 15, 3:10 PM
rnk committed rL313401: [debuginfo-tests] Add string NRVO test for PR34513.
[debuginfo-tests] Add string NRVO test for PR34513
Fri, Sep 15, 3:01 PM
rnk committed rL313400: [DebugInfo] Insert DW_OP_deref when spilling indirect DBG_VALUEs.
[DebugInfo] Insert DW_OP_deref when spilling indirect DBG_VALUEs
Fri, Sep 15, 2:56 PM
rnk closed D37911: [DebugInfo] Insert DW_OP_deref when spilling indirect DBG_VALUEs by committing rL313400: [DebugInfo] Insert DW_OP_deref when spilling indirect DBG_VALUEs.
Fri, Sep 15, 2:56 PM
rnk committed rL313399: [DebugInfo] Add missing DW_OP_deref when an NRVO pointer is spilled.
[DebugInfo] Add missing DW_OP_deref when an NRVO pointer is spilled
Fri, Sep 15, 2:51 PM
rnk closed D37929: [DebugInfo] Add missing DW_OP_deref when an NRVO pointer is spilled by committing rL313399: [DebugInfo] Add missing DW_OP_deref when an NRVO pointer is spilled.
Fri, Sep 15, 2:51 PM
rnk added inline comments to D37929: [DebugInfo] Add missing DW_OP_deref when an NRVO pointer is spilled.
Fri, Sep 15, 2:46 PM
rnk added a comment to D37929: [DebugInfo] Add missing DW_OP_deref when an NRVO pointer is spilled.

Do you think you could craft a debuginfo-tests test from PR34513, so we can avoid regressing in the future?

Fri, Sep 15, 2:42 PM
rnk updated the diff for D37911: [DebugInfo] Insert DW_OP_deref when spilling indirect DBG_VALUEs.
  • Mostly s/MDNode/DILocalVariable/
Fri, Sep 15, 2:35 PM
rnk added inline comments to D37911: [DebugInfo] Insert DW_OP_deref when spilling indirect DBG_VALUEs.
Fri, Sep 15, 2:34 PM
rnk created D37932: [DebugInfo] Correctly coalesce DBG_VALUEs that mix direct and indirect values.
Fri, Sep 15, 2:30 PM
rnk accepted D37920: Resubmit (Fixed) "[lit] Force site configs to be run before source tree configs".

Looks good, sounds like a plan: run lit.py like we used to if we don't have or can't find an llvm-lit.in to configure.

Fri, Sep 15, 2:28 PM
rnk committed rL313397: Fix build for LLVM unittests.
Fix build for LLVM unittests
Fri, Sep 15, 2:13 PM
rnk added inline comments to D37911: [DebugInfo] Insert DW_OP_deref when spilling indirect DBG_VALUEs.
Fri, Sep 15, 2:04 PM
rnk created D37929: [DebugInfo] Add missing DW_OP_deref when an NRVO pointer is spilled.
Fri, Sep 15, 1:44 PM
rnk accepted D37882: ubsan: Unbreak ubsan_cxx runtime library on Windows..

lgtm

Fri, Sep 15, 12:55 PM
rnk added a comment to D37920: Resubmit (Fixed) "[lit] Force site configs to be run before source tree configs".

This is problematic because people already complain about how compiler-rt depends on the LLVM source tree for gtest sources: https://bugs.llvm.org/show_bug.cgi?id=33693

Fri, Sep 15, 12:52 PM
rnk accepted D37157: Fix Bug 30978 by emitting cv file checksums..

Thanks, looks good!

Fri, Sep 15, 11:10 AM
rnk created D37911: [DebugInfo] Insert DW_OP_deref when spilling indirect DBG_VALUEs.
Fri, Sep 15, 11:10 AM
rnk added inline comments to D37882: ubsan: Unbreak ubsan_cxx runtime library on Windows..
Fri, Sep 15, 10:00 AM
rnk accepted D37884: ubsan: Stop building the DLL version of the runtime library on Windows..

lgtm

Fri, Sep 15, 9:53 AM
rnk accepted D37883: cfi: Enable ThinLTO tests on Windows..

lgtm

Fri, Sep 15, 9:53 AM
rnk added inline comments to D37891: Driver: hoist the `wchar_t` handling to the driver.
Fri, Sep 15, 9:43 AM · Restricted Project
rnk added a reviewer for D37891: Driver: hoist the `wchar_t` handling to the driver: MatzeB.
Fri, Sep 15, 9:39 AM · Restricted Project
rnk added a comment to D37891: Driver: hoist the `wchar_t` handling to the driver.

I do remember recommending this approach over IRC, but I thought we concluded that we should leave all the defaults in lib/Basic/Targets/ and make the -cc1 -fwchar-type= and -f[no-]signed-wchar overrides that affected all targets equally. That would avoid the need for these test updates, anyway.

Fri, Sep 15, 9:20 AM · Restricted Project

Thu, Sep 14

rnk committed rL313332: Fix syntax in lldb lit.cfg.
Fix syntax in lldb lit.cfg
Thu, Sep 14, 6:32 PM
rnk committed rL313331: [cmake] Fix a variable shadowing bug.
[cmake] Fix a variable shadowing bug
Thu, Sep 14, 6:20 PM
rnk accepted D37308: Fix the __interface inheritence rules to work better with IUnknown and IDispatch.

lgtm

Thu, Sep 14, 6:09 PM
rnk committed rL313330: [COFF] Remove unused variable NFC.
[COFF] Remove unused variable NFC
Thu, Sep 14, 6:08 PM
rnk committed rL313329: [codeview] Use a type index of zero for static method "this" types.
[codeview] Use a type index of zero for static method "this" types
Thu, Sep 14, 6:00 PM
rnk added inline comments to D37157: Fix Bug 30978 by emitting cv file checksums..
Thu, Sep 14, 5:27 PM
rnk added inline comments to D37157: Fix Bug 30978 by emitting cv file checksums..
Thu, Sep 14, 4:12 PM
rnk added inline comments to D37308: Fix the __interface inheritence rules to work better with IUnknown and IDispatch.
Thu, Sep 14, 11:43 AM
rnk added inline comments to D37852: [dwarfdump] Make .eh_frame an alias for .debug_frame.
Thu, Sep 14, 11:20 AM

Wed, Sep 13

rnk added a comment to D37838: [lit] Allow lit config files to have a .py extension.
In D37838#870368, @rnk wrote:

lit.cfg.py still isn't a valid name for a python module. I had this idea that in the future we'd import the config module directly to simplify custom test formats defined in lit configuration files, which interact badly with multiprocessing.Pool. I guess that won't work well anyway since all the config files will have the same name, so we still have to move custom test formats out into well-named python modules.

Could we work around this by using __import__ or something or searching for a module loader with a given path, and just assigning a dynamically generated name to the module? Not trying to do anything like that with this patch by the way.

Wed, Sep 13, 6:14 PM
rnk added a comment to D37838: [lit] Allow lit config files to have a .py extension.

lit.cfg.py still isn't a valid name for a python module. I had this idea that in the future we'd import the config module directly to simplify custom test formats defined in lit configuration files, which interact badly with multiprocessing.Pool. I guess that won't work well anyway since all the config files will have the same name, so we still have to move custom test formats out into well-named python modules.

Wed, Sep 13, 5:32 PM
rnk added a comment to D37089: [Sema] Error out early for tags defined inside an enumeration..

I think this looks good, even without fixing the access control crash, this seems like a diagnostic improvement.

Wed, Sep 13, 4:46 PM
rnk committed rL313214: [codeview] Fold FIXME into comment, there's nothing to do. NFC.
[codeview] Fold FIXME into comment, there's nothing to do. NFC
Wed, Sep 13, 4:31 PM
rnk committed rL313203: [codeview] VLAs and unsized arrays should use a size of zero.
[codeview] VLAs and unsized arrays should use a size of zero
Wed, Sep 13, 2:55 PM
rnk updated the diff for D37768: [IR] Add llvm.dbg.addr, a control-dependent version of llvm.dbg.declare.
  • rebase
  • implement Paul's comments
Wed, Sep 13, 2:09 PM
rnk added inline comments to D37818: [lit] Update clang and lld to use the new shared LLVMConfig stuff.
Wed, Sep 13, 12:39 PM
rnk added inline comments to D37771: llvm-dwarfdump: automatically dump both regular and .dwo variant of sections.
Wed, Sep 13, 11:48 AM
rnk added a comment to D37715: Mark static member functions as static in CodeViewDebug.

This sounded to me like a Python 2/3 problem, so I reset my environment to Python 2.7, re-ran CMake, and tried again. But I got the same set of Python errors. What's the trick here?

Wed, Sep 13, 11:35 AM
rnk accepted D37787: Driver: Make -fwhole-program-vtables a core option so it can be used from clang-cl..

lgtm

Wed, Sep 13, 11:27 AM

Tue, Sep 12

rnk committed rL313110: Fix dwarfdump cmdline test on Windows.
Fix dwarfdump cmdline test on Windows
Tue, Sep 12, 6:52 PM
rnk committed rL313108: [InstCombine] Add a flag to disable LowerDbgDeclare.
[InstCombine] Add a flag to disable LowerDbgDeclare
Tue, Sep 12, 6:45 PM
rnk closed D36596: [InstCombine] Add a flag to disable LowerDbgDeclare by committing rL313108: [InstCombine] Add a flag to disable LowerDbgDeclare.
Tue, Sep 12, 6:44 PM
rnk accepted D37715: Mark static member functions as static in CodeViewDebug.

Looks good! Make sure to run check-clang and check-llvm. The CGDebugInfo.cpp change might affect clang codegen tests that generate debug info for DWARF.

Tue, Sep 12, 5:32 PM
rnk added a comment to D36596: [InstCombine] Add a flag to disable LowerDbgDeclare.

Does this look good as is? I'd like to use this to gather some user feedback.

Tue, Sep 12, 3:06 PM
rnk updated the diff for D37768: [IR] Add llvm.dbg.addr, a control-dependent version of llvm.dbg.declare.
  • update
Tue, Sep 12, 2:57 PM
rnk added inline comments to D37768: [IR] Add llvm.dbg.addr, a control-dependent version of llvm.dbg.declare.
Tue, Sep 12, 2:55 PM
rnk added a comment to D37756: [lit] Force site configs to be run before source-tree configs.

I think the main usage model this de-supports is putting tools on PATH and running lit.py directly on tests from the source directory like this:

PATH=$PATH:$build/bin llvm/utils/lit/lit.py llvm/test/Debug/X86/foo.ll
Tue, Sep 12, 2:13 PM
rnk added a comment to D37756: [lit] Force site configs to be run before source-tree configs.

This is awesome! phab doesn't tell you the diffstat, so I patched it in, and it is: 177 insertions(+), 788 deletions(-) Nobody understood that code anyway, so this is great.

Tue, Sep 12, 1:58 PM
rnk created D37768: [IR] Add llvm.dbg.addr, a control-dependent version of llvm.dbg.declare.
Tue, Sep 12, 1:48 PM
rnk added inline comments to D37413: [X86][MS-InlineAsm] Extended support for variables / identifiers on memory / immediate expressions.
Tue, Sep 12, 11:03 AM
rnk accepted D37727: Driver: Remove custom MinGW linker detection.

Nice! IIUC, now that the gnu ld driver works on Windows we can remove this logic. Thanks!

Tue, Sep 12, 10:58 AM
rnk added inline comments to D37740: [SelectionDAG] Pick correct frame index in LowerArguments.
Tue, Sep 12, 10:53 AM
rnk accepted D37745: [dwarfdump] Rename Brief to Verbose in DIDumpOptions.

Nice. :)

Tue, Sep 12, 10:41 AM

Mon, Sep 11

rnk added a comment to D37715: Mark static member functions as static in CodeViewDebug.

This should have an IRGen test. You can probably extend clang/test/CodeGenCXX/debug-info-ms-abi.cpp with a static method.

Mon, Sep 11, 4:46 PM
rnk added inline comments to D37717: llvm-dwarfdump: Make -brief the default and add -verbose instead..
Mon, Sep 11, 4:32 PM
rnk added a comment to D37714: llvm-dwarfdump: Replace -debug-dump=sect option with individual options..

+1 for dropping the old names

Mon, Sep 11, 4:29 PM
rnk added inline comments to D37529: [codeview] omit debug locations for nested exprs unless column info enabled.
Mon, Sep 11, 2:03 PM
rnk accepted D37412: [X86][MS-InlineAsm] Extended support for variables / identifiers on memory / immediate expressions.

Looks good

Mon, Sep 11, 1:34 PM