Page MenuHomePhabricator

artagnon (Ramkumar Ramachandra)
User

Projects

User does not belong to any projects.

User Details

User Since
Oct 28 2014, 11:11 AM (364 w, 6 d)

Recent Activity

Mar 10 2021

artagnon added a reviewer for D98217: [.mailmap] Add entry for Ramkumar Ramachandra: chandlerc.
Mar 10 2021, 3:41 AM · Restricted Project

Mar 8 2021

artagnon requested review of D98217: [.mailmap] Add entry for Ramkumar Ramachandra.
Mar 8 2021, 1:59 PM · Restricted Project

Aug 7 2016

artagnon added a comment to D23250: Fix jumptable with large code model.

The file was auto-formatted via clang-format on save, via Atom. Sorry about the noise.

Aug 7 2016, 4:34 PM
artagnon retitled D23250: Fix jumptable with large code model from to Fix jumptable with large code model.
Aug 7 2016, 4:25 PM

Nov 12 2015

artagnon committed rL252898: Use uniqueness of C++ fully-qualified names to resolve conflicts.
Use uniqueness of C++ fully-qualified names to resolve conflicts
Nov 12 2015, 6:47 AM
artagnon closed D14549: Use uniqueness of C++ fully-qualified names to resolve conflicts by committing rL252898: Use uniqueness of C++ fully-qualified names to resolve conflicts.
Nov 12 2015, 6:46 AM

Nov 10 2015

artagnon retitled D14549: Use uniqueness of C++ fully-qualified names to resolve conflicts from to Use uniqueness of C++ fully-qualified names to resolve conflicts.
Nov 10 2015, 11:59 AM
artagnon committed rL252628: AddLLVM: squelch CMP0007 by not adding empty elements to list.
AddLLVM: squelch CMP0007 by not adding empty elements to list
Nov 10 2015, 10:29 AM
artagnon closed D14339: AddLLVM: squelch CMP0007 by not adding empty elements to list by committing rL252628: AddLLVM: squelch CMP0007 by not adding empty elements to list.
Nov 10 2015, 10:29 AM
artagnon committed rL252624: LLDBStandalone: Report nice errors on missing vars.
LLDBStandalone: Report nice errors on missing vars
Nov 10 2015, 10:11 AM
artagnon closed D13819: LLDBStandalone: Report nice errors on missing vars by committing rL252624: LLDBStandalone: Report nice errors on missing vars.
Nov 10 2015, 10:11 AM

Nov 4 2015

artagnon added a comment to D14339: AddLLVM: squelch CMP0007 by not adding empty elements to list.

It's actually an extra newline character that my editor automatically stripped out. *nix has the requirement that all lines must end with a newline character, not that a file must end with an empty line.

Nov 4 2015, 11:08 AM
artagnon retitled D14339: AddLLVM: squelch CMP0007 by not adding empty elements to list from to AddLLVM: squelch CMP0007 by not adding empty elements to list.
Nov 4 2015, 10:55 AM
artagnon added a comment to D13819: LLDBStandalone: Report nice errors on missing vars.

Okay, so run this on the current HEAD:

$ cmake -GNinja -DCMAKE_BUILD_TYPE=Debug ../lldb
-- The C compiler identification is AppleClang 6.1.0.6020053
-- The CXX compiler identification is AppleClang 6.1.0.6020053
-- Check for working C compiler using: Ninja
-- Check for working C compiler using: Ninja -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler using: Ninja
-- Check for working CXX compiler using: Ninja -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
CMake Error at cmake/modules/LLDBStandalone.cmake:37 (get_filename_component):
  get_filename_component called with incorrect number of arguments
Call Stack (most recent call first):
  CMakeLists.txt:3 (include)
Nov 4 2015, 10:20 AM

Nov 3 2015

artagnon committed rL251979: Squelch a silly warning regarding an extra 'default' in 'case'.
Squelch a silly warning regarding an extra 'default' in 'case'
Nov 3 2015, 1:32 PM
artagnon closed D13816: Fix a some silly warnings by committing rL251979: Squelch a silly warning regarding an extra 'default' in 'case'.
Nov 3 2015, 1:32 PM
artagnon added a reviewer for D13819: LLDBStandalone: Report nice errors on missing vars: clayborg.
Nov 3 2015, 1:26 PM
artagnon accepted D14258: [CodegenPrepare] Do not rematerialize gc.relocates across different basic blocks.

Pretty silly mistake on my part. Thanks for the clear tests.

Nov 3 2015, 8:52 AM

Oct 16 2015

artagnon retitled D13819: LLDBStandalone: Report nice errors on missing vars from to LLDBStandalone: Report nice errors on missing vars.
Oct 16 2015, 11:15 AM
artagnon updated the diff for D13816: Fix a some silly warnings.

Er, silly mistake (build failure)

Oct 16 2015, 10:28 AM
artagnon retitled D13816: Fix a some silly warnings from to Fix a some silly warnings.
Oct 16 2015, 10:22 AM

Aug 27 2015

artagnon accepted D12216: Cleaning up LLVM IR mode for Emacs.

LGTM, with some minor doubts.

Aug 27 2015, 6:02 AM

Feb 25 2015

artagnon committed rL230591: PlaceSafepoints: use IRBuilder helpers.
PlaceSafepoints: use IRBuilder helpers
Feb 25 2015, 4:38 PM
artagnon closed D7518: PlaceSafepoints: use IRBuilder helpers by committing rL230591: PlaceSafepoints: use IRBuilder helpers.
Feb 25 2015, 4:38 PM
artagnon committed rL230582: MemDepPrinter: Fix some nits introduced in r228596.
MemDepPrinter: Fix some nits introduced in r228596
Feb 25 2015, 3:57 PM
artagnon closed D7644: MemDepPrinter: Fix some nits introduced in r228596 by committing rL230582: MemDepPrinter: Fix some nits introduced in r228596.
Feb 25 2015, 3:57 PM

Feb 14 2015

artagnon abandoned D7456: [Transforms] Write basic -place-safepoints test, fix bug.

I submitted the relevant test with 69a5c891.

Feb 14 2015, 2:13 PM
artagnon retitled D7644: MemDepPrinter: Fix some nits introduced in r228596 from to MemDepPrinter: Fix some nits introduced in r228596.
Feb 14 2015, 12:06 PM
artagnon committed rL229265: InstCombine: propagate deref via new addDereferenceableAttr.
InstCombine: propagate deref via new addDereferenceableAttr
Feb 14 2015, 11:40 AM
artagnon closed D7510: InstCombine: propagate deref via new addDereferenceableAttr by committing rL229265: InstCombine: propagate deref via new addDereferenceableAttr.
Feb 14 2015, 11:39 AM
artagnon added inline comments to D7535: Add invoke statepoint placement tests.
Feb 14 2015, 11:38 AM

Feb 9 2015

artagnon added a comment to D7510: InstCombine: propagate deref via new addDereferenceableAttr.

p.s. isDerefPtr is on Value, the cast is entirely unnecessary.

Feb 9 2015, 3:26 PM
artagnon added a comment to D7518: PlaceSafepoints: use IRBuilder helpers.

The relocate part of this change LGTM w/one change: The old code took the name of the replaced call for the result. This is useful for keeping the IR readable. Please add a test for this as well.

Feb 9 2015, 3:24 PM
artagnon committed rL228626: [Statepoint] Improve two asserts, fix some style (NFC).
[Statepoint] Improve two asserts, fix some style (NFC)
Feb 9 2015, 3:04 PM
artagnon closed D7517: [Statepoint] Improve two asserts, fix some style (NFC) by committing rL228626: [Statepoint] Improve two asserts, fix some style (NFC).
Feb 9 2015, 3:04 PM
artagnon committed rL228625: PlaceSafepoints: modernize gc.result.* -> gc.result.
PlaceSafepoints: modernize gc.result.* -> gc.result
Feb 9 2015, 3:02 PM
artagnon closed D7516: PlaceSafepoints: modernize gc.result.* -> gc.result by committing rL228625: PlaceSafepoints: modernize gc.result.* -> gc.result.
Feb 9 2015, 3:02 PM
artagnon updated the diff for D7510: InstCombine: propagate deref via new addDereferenceableAttr.

Don't assume that the derived pointer can always be cast to an
Argument; dyn_cast and check.

Feb 9 2015, 2:06 PM
artagnon added a comment to D7456: [Transforms] Write basic -place-safepoints test, fix bug.

Do try to rebase onto my changes; I did spend quite a bit of time painfully splitting the patches.

Feb 9 2015, 1:53 PM
artagnon committed rL228611: MemDerefPrinter: Require DataLayoutPass for higher accuracy.
MemDerefPrinter: Require DataLayoutPass for higher accuracy
Feb 9 2015, 1:52 PM
artagnon closed D7508: MemDerefPrinter: Require DataLayoutPass for higher accuracy by committing rL228611: MemDerefPrinter: Require DataLayoutPass for higher accuracy.
Feb 9 2015, 1:51 PM
artagnon retitled D7518: PlaceSafepoints: use IRBuilder helpers from to PlaceSafepoints: use IRBuilder helpers.
Feb 9 2015, 1:49 PM
artagnon retitled D7517: [Statepoint] Improve two asserts, fix some style (NFC) from to [Statepoint] Improve two asserts, fix some style (NFC).
Feb 9 2015, 1:38 PM
artagnon retitled D7516: PlaceSafepoints: modernize gc.result.* -> gc.result from to PlaceSafepoints: modernize gc.result.* -> gc.result.
Feb 9 2015, 1:28 PM
artagnon updated the diff for D7456: [Transforms] Write basic -place-safepoints test, fix bug.

As requested by reames, shrink the original submission to be just the
bugfix. Other changes will come as later patches.

Feb 9 2015, 1:25 PM
artagnon committed rL228606: isDereferenceablePointer: look through gc.relocate calls.
isDereferenceablePointer: look through gc.relocate calls
Feb 9 2015, 1:10 PM
artagnon closed D7454: isDereferenceablePointer: look through gc.relocate calls by committing rL228606: isDereferenceablePointer: look through gc.relocate calls.
Feb 9 2015, 1:10 PM
artagnon committed rL228596: MemDepPrinter: cleanup a few loops (NFC).
MemDepPrinter: cleanup a few loops (NFC)
Feb 9 2015, 11:51 AM
artagnon closed D7455: [Analysis] MemDepPrinter: cleanup a few loops (NFC) by committing rL228596: MemDepPrinter: cleanup a few loops (NFC).
Feb 9 2015, 11:51 AM
artagnon retitled D7510: InstCombine: propagate deref via new addDereferenceableAttr from to InstCombine: propagate deref via new addDereferenceableAttr.
Feb 9 2015, 9:54 AM
artagnon retitled D7508: MemDerefPrinter: Require DataLayoutPass for higher accuracy from to MemDerefPrinter: Require DataLayoutPass for higher accuracy.
Feb 9 2015, 9:12 AM

Feb 8 2015

artagnon committed rL228556: InstCombine: propagate nonNull through assume.
InstCombine: propagate nonNull through assume
Feb 8 2015, 5:15 PM
artagnon closed D7107: InstCombine: propagate nonNull through assume by committing rL228556: InstCombine: propagate nonNull through assume.
Feb 8 2015, 5:15 PM
artagnon committed rL228555: [emacs] Get llvm-mode to font-lock "personality".
[emacs] Get llvm-mode to font-lock "personality"
Feb 8 2015, 4:32 PM
artagnon closed D7494: [emacs] Get llvm-mode to font-lock "personality" by committing rL228555: [emacs] Get llvm-mode to font-lock "personality".
Feb 8 2015, 4:32 PM
artagnon updated the diff for D7107: InstCombine: propagate nonNull through assume.

Drop the deref propagation, because it requires more work. Redo the
nonNull propagation so that it is correct (thanks to Hal for poking me
to test it properly).

Feb 8 2015, 4:26 PM
artagnon retitled D7107: InstCombine: propagate nonNull through assume from InstCombine: propagate deref and nonNull return attributes to InstCombine: propagate nonNull through assume.
Feb 8 2015, 4:25 PM
artagnon retitled D7494: [emacs] Get llvm-mode to font-lock "personality" from to [emacs] Get llvm-mode to font-lock "personality".
Feb 8 2015, 2:37 PM
artagnon updated the diff for D7456: [Transforms] Write basic -place-safepoints test, fix bug.

A quick run through clang-format.

Feb 8 2015, 12:42 PM
artagnon updated the diff for D7456: [Transforms] Write basic -place-safepoints test, fix bug.

More work in the same direction.

Feb 8 2015, 12:39 PM

Feb 5 2015

artagnon retitled D7456: [Transforms] Write basic -place-safepoints test, fix bug from to [Transforms] Write basic -place-safepoints test, fix bug.
Feb 5 2015, 8:05 PM
artagnon added a comment to D7107: InstCombine: propagate nonNull through assume.

ping! (How) should I write a test for this?

Feb 5 2015, 6:23 PM
artagnon retitled D7455: [Analysis] MemDepPrinter: cleanup a few loops (NFC) from to [Analysis] MemDepPrinter: cleanup a few loops (NFC).
Feb 5 2015, 6:13 PM
artagnon retitled D7454: isDereferenceablePointer: look through gc.relocate calls from to isDereferenceablePointer: look through gc.relocate calls.
Feb 5 2015, 6:07 PM
artagnon committed rL228369: Introduce print-memderefs to test isDereferenceablePointer.
Introduce print-memderefs to test isDereferenceablePointer
Feb 5 2015, 5:48 PM
artagnon closed D7075: Introduce print-memderefs to test isDereferenceablePointer by committing rL228369: Introduce print-memderefs to test isDereferenceablePointer.
Feb 5 2015, 5:48 PM

Jan 27 2015

artagnon updated the diff for D7075: Introduce print-memderefs to test isDereferenceablePointer.

Split print-memderefs into its own change. Fix all-round sloppiness.

Jan 27 2015, 12:51 PM
artagnon retitled D7075: Introduce print-memderefs to test isDereferenceablePointer from isDereferenceablePointer: look through gc.relocates + test to Introduce print-memderefs to test isDereferenceablePointer.
Jan 27 2015, 12:48 PM
artagnon committed rL227229: overloaded-intrinsic-name: exercise anyptr on struct.
overloaded-intrinsic-name: exercise anyptr on struct
Jan 27 2015, 12:04 PM
artagnon closed D7037: overloaded-intrinsic-name: exercise anyptr on struct by committing rL227229: overloaded-intrinsic-name: exercise anyptr on struct.
Jan 27 2015, 12:04 PM

Jan 23 2015

artagnon committed rL226931: [emacs] llvm-mode: fix parens, font-lock i*.
[emacs] llvm-mode: fix parens, font-lock i*
Jan 23 2015, 11:47 AM
artagnon closed D7036: [emacs] fix parens, font-lock i* in llvm-mode by committing rL226931: [emacs] llvm-mode: fix parens, font-lock i*.
Jan 23 2015, 11:47 AM
artagnon abandoned D7037: overloaded-intrinsic-name: exercise anyptr on struct.

Abandoning: not important enough to wait for; this test file can be rewritten to exercise Any comprehensively, if anyone's interested.

Jan 23 2015, 8:31 AM
artagnon abandoned D6988: Make intrinsic signature iAny work with arrays.

Abandoning: it's not important enough; I have no dependent changes.

Jan 23 2015, 8:30 AM
artagnon added a comment to D7036: [emacs] fix parens, font-lock i* in llvm-mode.

I cannot reproduce the (( or [[ issue. What do I have to do for email to print out a (( instead of a (?

Jan 23 2015, 8:19 AM

Jan 22 2015

artagnon committed rL226861: [emacs] Use c-mode-common-hook, derive from "gnu".
[emacs] Use c-mode-common-hook, derive from "gnu"
Jan 22 2015, 12:57 PM
artagnon closed D7035: [emacs] Use c-mode-common-hook, derive from "gnu" by committing rL226861: [emacs] Use c-mode-common-hook, derive from "gnu".
Jan 22 2015, 12:57 PM
artagnon committed rL226857: Intrinsics: introduce llvm_any_ty aka ValueType Any.
Intrinsics: introduce llvm_any_ty aka ValueType Any
Jan 22 2015, 12:16 PM
artagnon closed D7020: Intrinsics: introduce llvm_any_ty aka ValueType Any by committing rL226857: Intrinsics: introduce llvm_any_ty aka ValueType Any.
Jan 22 2015, 12:16 PM
artagnon added a comment to D7035: [emacs] Use c-mode-common-hook, derive from "gnu".

This should be really easy to review and get out of the way; it's borderline non-functional.

Jan 22 2015, 11:28 AM
artagnon added a reviewer for D7035: [emacs] Use c-mode-common-hook, derive from "gnu": rafael.
Jan 22 2015, 11:27 AM
artagnon abandoned D6779: [OCaml] introduce link_gc_components.
Jan 22 2015, 11:19 AM

Jan 21 2015

artagnon added a comment to D7020: Intrinsics: introduce llvm_any_ty aka ValueType Any.

I have too many changes awaiting review to continue working :\

Jan 21 2015, 2:53 PM
artagnon retitled D7107: InstCombine: propagate nonNull through assume from to InstCombine: propagate deref and nonNull return attributes.
Jan 21 2015, 1:16 PM

Jan 20 2015

artagnon retitled D7075: Introduce print-memderefs to test isDereferenceablePointer from to isDereferenceablePointer: look through gc.relocates + test.
Jan 20 2015, 3:41 PM
artagnon committed rL226599: [GC] Verify-pass void vararg functions in gc.statepoint.
[GC] Verify-pass void vararg functions in gc.statepoint
Jan 20 2015, 11:44 AM
artagnon closed D7071: [GC] Verify-pass void vararg functions in gc.statepoint by committing rL226599: [GC] Verify-pass void vararg functions in gc.statepoint.
Jan 20 2015, 11:44 AM
artagnon retitled D7071: [GC] Verify-pass void vararg functions in gc.statepoint from to [GC] Verify-pass void vararg functions in gc.statepoint.
Jan 20 2015, 11:09 AM

Jan 18 2015

artagnon added a comment to D7020: Intrinsics: introduce llvm_any_ty aka ValueType Any.

Elena Demikhovsky wrote:

When you put "any", you probably mean any scalar of int, float, ptr. What happens if somebody specifies vector, i16, double or something else. May be you should put some code in the verifier?

Jan 18 2015, 10:20 AM

Jan 17 2015

artagnon retitled D7037: overloaded-intrinsic-name: exercise anyptr on struct from to overloaded-intrinsic-name: exercise anyptr on struct.
Jan 17 2015, 3:34 PM
artagnon retitled D7036: [emacs] fix parens, font-lock i* in llvm-mode from to [emacs] fix parens, font-lock i* in llvm-mode.
Jan 17 2015, 12:36 PM
artagnon retitled D7035: [emacs] Use c-mode-common-hook, derive from "gnu" from to [emacs] Use c-mode-common-hook, derive from "gnu".
Jan 17 2015, 11:59 AM

Jan 16 2015

artagnon added a reviewer for D7020: Intrinsics: introduce llvm_any_ty aka ValueType Any: craig.topper.
Jan 16 2015, 2:17 PM
artagnon updated the diff for D7020: Intrinsics: introduce llvm_any_ty aka ValueType Any.

Don't remove old intrinsics right away; just deprecate them.

Jan 16 2015, 2:16 PM
artagnon added a reviewer for D7020: Intrinsics: introduce llvm_any_ty aka ValueType Any: craig.topper.
Jan 16 2015, 2:15 PM
artagnon retitled D7020: Intrinsics: introduce llvm_any_ty aka ValueType Any from to Intrinsics: introduce llvm_any_ty aka ValueType Any.
Jan 16 2015, 1:01 PM

Jan 15 2015

artagnon updated the diff for D6988: Make intrinsic signature iAny work with arrays.

Rebase to include statepoint-example gc.

Jan 15 2015, 11:03 AM
artagnon abandoned D6989: statepoint-call-lowering: test relocate int and array.

After reading through StatepointLowering in greater detail, I realize that I should be testing the stackmaps, not the generic stores that are emitted; emitting those stores is not part of the statepoint code.

Jan 15 2015, 10:40 AM
artagnon added a comment to D7000: statepoint-call-lowering: use statepoint-example gc.

Philip Reames wrote:

As I said before, LGTM.

Jan 15 2015, 10:35 AM
artagnon closed D7000: statepoint-call-lowering: use statepoint-example gc.
Jan 15 2015, 10:24 AM