Page MenuHomePhabricator

abinavpp (Abinav Puthan Purayil)
User

Projects

User does not belong to any projects.

User Details

User Since
Feb 14 2021, 4:15 AM (84 w, 4 d)

Recent Activity

Yesterday

abinavpp committed rG3759398b4bf2: [AMDGPU] Report minimum scratch size in code object v5 and later by default (authored by abinavpp).
[AMDGPU] Report minimum scratch size in code object v5 and later by default
Wed, Sep 28, 9:24 PM · Restricted Project, Restricted Project
abinavpp closed D128346: [AMDGPU] Report minimum scratch size in code object v5 and later by default.
Wed, Sep 28, 9:23 PM · Restricted Project, Restricted Project

Tue, Sep 27

abinavpp updated the diff for D134714: [AMDGPU] Make the uses_dynamic_stack field in the kernel descriptor and the metadata map specific to code object v5 and later.

Address review comment

Tue, Sep 27, 10:40 PM · Restricted Project, Restricted Project
abinavpp requested review of D134714: [AMDGPU] Make the uses_dynamic_stack field in the kernel descriptor and the metadata map specific to code object v5 and later.
Tue, Sep 27, 1:02 AM · Restricted Project, Restricted Project

Mon, Sep 26

abinavpp updated the diff for D128346: [AMDGPU] Report minimum scratch size in code object v5 and later by default.

Moving the doc change to amdgpu-amdhsa-kernel-descriptor-v3-table; I think we
should remove the "V3" bit from that section since the kernel descriptor applies
for later code object versions

Mon, Sep 26, 9:35 AM · Restricted Project, Restricted Project
abinavpp added inline comments to D128346: [AMDGPU] Report minimum scratch size in code object v5 and later by default.
Mon, Sep 26, 8:01 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D128346: [AMDGPU] Report minimum scratch size in code object v5 and later by default.
  • Get the new scratch reporting specific to v5+ as per internal discussion
  • Not removing the -amdgpu-assume-* options now; Might need this when we run into problems
Mon, Sep 26, 4:16 AM · Restricted Project, Restricted Project

Wed, Sep 21

abinavpp updated the diff for D134183: [RegCoalescer] Replace undef subreg copies used in early-clobber def insts with IMPLICIT_DEFs.

Rebase on f09e3ad88a19

Wed, Sep 21, 3:22 AM · Restricted Project, Restricted Project
abinavpp added a reviewer for D134183: [RegCoalescer] Replace undef subreg copies used in early-clobber def insts with IMPLICIT_DEFs: dmgreen.
Wed, Sep 21, 3:07 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D134183: [RegCoalescer] Replace undef subreg copies used in early-clobber def insts with IMPLICIT_DEFs.

Rebase; Move TwoAddrInst changes to D134343

Wed, Sep 21, 3:03 AM · Restricted Project, Restricted Project
abinavpp requested review of D134343: [TwoAddrInst] Emit COPYs for undef REG_SEQUENCE operands if used in inst with early-clobber defs.
Wed, Sep 21, 2:57 AM · Restricted Project, Restricted Project

Tue, Sep 20

abinavpp added inline comments to D134183: [RegCoalescer] Replace undef subreg copies used in early-clobber def insts with IMPLICIT_DEFs.
Tue, Sep 20, 3:45 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D134183: [RegCoalescer] Replace undef subreg copies used in early-clobber def insts with IMPLICIT_DEFs.
  • Replaced the subreg copy with an implicit_def; This eliminates the new copies in the Thumb2 tests
  • Reword a comment
Tue, Sep 20, 2:27 AM · Restricted Project, Restricted Project

Mon, Sep 19

abinavpp added a comment to D134183: [RegCoalescer] Replace undef subreg copies used in early-clobber def insts with IMPLICIT_DEFs.

Haven't looked into all the test changes, especially the ones in Thumb2; Not sure if there's a better way to handle this

Mon, Sep 19, 7:42 AM · Restricted Project, Restricted Project
abinavpp requested review of D134183: [RegCoalescer] Replace undef subreg copies used in early-clobber def insts with IMPLICIT_DEFs.
Mon, Sep 19, 7:38 AM · Restricted Project, Restricted Project

Jul 27 2022

abinavpp added inline comments to D130579: AMDGPU: Use tablegen patterns for buffer global and flat atomic fadd.
Jul 27 2022, 1:26 AM · Restricted Project, Restricted Project

Jul 19 2022

abinavpp committed rG9fa425c1ab2f: [AMDGPU] Set amdgpu-memory-bound if a basic block has dense global memory access (authored by abinavpp).
[AMDGPU] Set amdgpu-memory-bound if a basic block has dense global memory access
Jul 19 2022, 2:48 AM · Restricted Project, Restricted Project
abinavpp closed D129759: [AMDGPU] Set amdgpu-memory-bound if a basic block has dense global memory access.
Jul 19 2022, 2:48 AM · Restricted Project, Restricted Project
abinavpp committed rG4baf8f092b47: [AMDGPU] Pre-commit tests for D129759 (authored by abinavpp).
[AMDGPU] Pre-commit tests for D129759
Jul 19 2022, 2:47 AM · Restricted Project, Restricted Project
abinavpp closed D129760: [AMDGPU] Pre-commit tests for D129759.
Jul 19 2022, 2:47 AM · Restricted Project, Restricted Project
abinavpp added inline comments to D129759: [AMDGPU] Set amdgpu-memory-bound if a basic block has dense global memory access.
Jul 19 2022, 2:45 AM · Restricted Project, Restricted Project

Jul 18 2022

abinavpp added inline comments to D129759: [AMDGPU] Set amdgpu-memory-bound if a basic block has dense global memory access.
Jul 18 2022, 12:35 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D129759: [AMDGPU] Set amdgpu-memory-bound if a basic block has dense global memory access.
  • Rebase
  • Don't compute GlobalMemAccPercentage if HasDenseGlobalMemAcc is already set for the function
  • Replace ternary expression with if
Jul 18 2022, 12:33 AM · Restricted Project, Restricted Project

Jul 17 2022

abinavpp committed rGd96361d714ee: [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor and the… (authored by abinavpp).
[AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor and the…
Jul 17 2022, 9:38 PM · Restricted Project, Restricted Project
abinavpp closed D128344: [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor and the kernel metadata map.
Jul 17 2022, 9:37 PM · Restricted Project, Restricted Project

Jul 15 2022

abinavpp added a comment to D129759: [AMDGPU] Set amdgpu-memory-bound if a basic block has dense global memory access.

I guess it comes down to the question if that block is hot or cold. There is no perfect heuristic out there, but chances are higher a block is hot if it is in a loop.

Jul 15 2022, 10:51 AM · Restricted Project, Restricted Project

Jul 14 2022

abinavpp requested review of D129760: [AMDGPU] Pre-commit tests for D129759.
Jul 14 2022, 5:15 AM · Restricted Project, Restricted Project
abinavpp requested review of D129759: [AMDGPU] Set amdgpu-memory-bound if a basic block has dense global memory access.
Jul 14 2022, 5:14 AM · Restricted Project, Restricted Project
abinavpp added a comment to D128544: [Orc][JITLink] Add a shared memory based implementation of MemoryMapper.

Hi! I am not very familiar with powerpc but it seems like POSIX shared memory APIs should be available.
From this page I suspect shm_ family of functions could be in librt instead of libc.

Jul 14 2022, 12:26 AM · Restricted Project, Restricted Project

Jul 7 2022

abinavpp committed rG17a81ecf857f: [AMDGPU] Use the HasNoUse predicate for no-ret atomic op selection (authored by abinavpp).
[AMDGPU] Use the HasNoUse predicate for no-ret atomic op selection
Jul 7 2022, 9:23 PM · Restricted Project, Restricted Project
abinavpp closed D125213: [AMDGPU] Use the HasNoUse predicate for no-ret atomic op selection.
Jul 7 2022, 9:22 PM · Restricted Project, Restricted Project
abinavpp committed rGc42fe5bd7a33: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate (authored by abinavpp).
[GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate
Jul 7 2022, 9:22 PM · Restricted Project, Restricted Project
abinavpp closed D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate.
Jul 7 2022, 9:22 PM · Restricted Project, Restricted Project
abinavpp committed rG7504c7a8772b: [AMDGPU] Use AddedComplexity for ret and noret atomic ops selection (authored by abinavpp).
[AMDGPU] Use AddedComplexity for ret and noret atomic ops selection
Jul 7 2022, 9:18 PM · Restricted Project, Restricted Project
abinavpp closed D128241: [AMDGPU] Use AddedComplexity for ret and noret atomic ops selection.
Jul 7 2022, 9:18 PM · Restricted Project, Restricted Project

Jul 6 2022

abinavpp added a comment to D128241: [AMDGPU] Use AddedComplexity for ret and noret atomic ops selection.

it seems like the default complexity logic could be smarter

Jul 6 2022, 10:28 AM · Restricted Project, Restricted Project
abinavpp added a comment to D128241: [AMDGPU] Use AddedComplexity for ret and noret atomic ops selection.

I don't think I understood the question. I set the AddedComplexity in such a way that no-return matches are always done first since we're only adding predicates for those. By "default complexity", are we talking about getting PatternToMatch::getPatternComplexity() in CodeGenDagPatterns to give higher complexity to patterns with HasNoUse predicates?

Yes. I thought predicates boosted the complexity, such that this would work out by default:

// If this node has some predicate function that must match, it adds to the
// complexity of this node.
if (!P->getPredicateCalls().empty())
  ++Size;
Jul 6 2022, 10:24 AM · Restricted Project, Restricted Project
abinavpp added a comment to D128241: [AMDGPU] Use AddedComplexity for ret and noret atomic ops selection.

Did the default complexity calculation account for the additional no use predicate in the opposite direction of what we want?

Jul 6 2022, 6:59 AM · Restricted Project, Restricted Project
abinavpp added a comment to D128344: [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor and the kernel metadata map.

Could use some assembler tests demonstrating the default behavior

Jul 6 2022, 6:39 AM · Restricted Project, Restricted Project

Jul 4 2022

abinavpp retitled D125213: [AMDGPU] Use the HasNoUse predicate for no-ret atomic op selection from [AMDGPU][GlobalISel] Enable no-ret atomic selection to [AMDGPU] Use the HasNoUse predicate for no-ret atomic op selection.
Jul 4 2022, 3:08 AM · Restricted Project, Restricted Project
abinavpp retitled D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate from [GlobalISel] Implement the HasNoUse builtin predicate to [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate.
Jul 4 2022, 3:08 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D125213: [AMDGPU] Use the HasNoUse predicate for no-ret atomic op selection.
  • Rebase
  • Replace both the C++ predicates with HasNoUse
Jul 4 2022, 3:05 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate.
  • Rebase
  • Add the SelectionDAG implementation; Had to add this in the TreePredicateFn::hasPredCode() since the predicate was not added for intrinsics without it (due to the absence of IsAtomic?).
Jul 4 2022, 3:04 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D128241: [AMDGPU] Use AddedComplexity for ret and noret atomic ops selection.

Rebase

Jul 4 2022, 3:01 AM · Restricted Project, Restricted Project
abinavpp added inline comments to D128346: [AMDGPU] Report minimum scratch size in code object v5 and later by default.
Jul 4 2022, 2:19 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D128346: [AMDGPU] Report minimum scratch size in code object v5 and later by default.
  • Rebase
  • Updated the tests (GlobalISel/)non-entry-alloca.ll, recursion.ll, hsa-metadata-resource-usage-function-ordering.ll to check the dynamic stack
Jul 4 2022, 2:17 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D128344: [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor and the kernel metadata map.
  • Rebase
  • Make the field optional
Jul 4 2022, 2:16 AM · Restricted Project, Restricted Project

Jun 29 2022

abinavpp retitled D128344: [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor and the kernel metadata map from [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor to [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor and the kernel metadata map.
Jun 29 2022, 9:43 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D128344: [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor and the kernel metadata map.
  • Rebase
  • Add the field to the kernel metadata map
Jun 29 2022, 9:41 AM · Restricted Project, Restricted Project

Jun 28 2022

abinavpp added a comment to D128344: [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor and the kernel metadata map.

I see the kernel descriptor changes reflected in the test updates, but I don't see the metadata entries tested

Jun 28 2022, 10:35 AM · Restricted Project, Restricted Project

Jun 24 2022

abinavpp added inline comments to D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate.
Jun 24 2022, 11:39 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D128346: [AMDGPU] Report minimum scratch size in code object v5 and later by default.

Rebase

Jun 24 2022, 10:22 AM · Restricted Project, Restricted Project
abinavpp retitled D128344: [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor and the kernel metadata map from [AMDGPU] Add the is_dynamic_stack field to the kernel descriptor to [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor.
Jun 24 2022, 10:21 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D128344: [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor and the kernel metadata map.

Rename the field to uses_dynamic_stack

Jun 24 2022, 10:20 AM · Restricted Project, Restricted Project
abinavpp added a comment to D128346: [AMDGPU] Report minimum scratch size in code object v5 and later by default.

Do we need to worry about non HSA platforms here? Like the behavior of AMDGPUPALMetadata::setFunctionScratchSize(). Also, the calculation of SIProgramInfo::ScratchBlocks?

Jun 24 2022, 3:48 AM · Restricted Project, Restricted Project
abinavpp retitled D128346: [AMDGPU] Report minimum scratch size in code object v5 and later by default from [AMDGPU] Set scratch size to 0 if is_dynamic_callstack is set to [AMDGPU] Disable the private segment size fallback to 16k.
Jun 24 2022, 3:39 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D128346: [AMDGPU] Report minimum scratch size in code object v5 and later by default.

Rebase

Jun 24 2022, 3:30 AM · Restricted Project, Restricted Project
abinavpp retitled D128344: [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor and the kernel metadata map from [AMDGPU] Add the "is_dynamic_callstack" of amd_kernel_code_t to the kernel descriptor to [AMDGPU] Add the is_dynamic_stack field to the kernel descriptor.
Jun 24 2022, 3:29 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D128344: [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor and the kernel metadata map.
  • As per internal discussion, use the 11th bit
  • Rename the field to is_dynamic_stack
Jun 24 2022, 3:25 AM · Restricted Project, Restricted Project

Jun 23 2022

abinavpp added inline comments to D128346: [AMDGPU] Report minimum scratch size in code object v5 and later by default.
Jun 23 2022, 11:17 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D128346: [AMDGPU] Report minimum scratch size in code object v5 and later by default.

Stop falling back to 16k (WIP)

Jun 23 2022, 11:14 AM · Restricted Project, Restricted Project
abinavpp added inline comments to D128346: [AMDGPU] Report minimum scratch size in code object v5 and later by default.
Jun 23 2022, 5:04 AM · Restricted Project, Restricted Project

Jun 22 2022

abinavpp added a comment to D128344: [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor and the kernel metadata map.

Do we want to tie this bit to a particular code object version? Code object v4 and up? Why did you pick code object v3 and up?

Jun 22 2022, 11:43 PM · Restricted Project, Restricted Project
abinavpp requested review of D128346: [AMDGPU] Report minimum scratch size in code object v5 and later by default.
Jun 22 2022, 8:08 AM · Restricted Project, Restricted Project
abinavpp requested review of D128344: [AMDGPU] Add the uses_dynamic_stack field to the kernel descriptor and the kernel metadata map.
Jun 22 2022, 8:07 AM · Restricted Project, Restricted Project

Jun 21 2022

abinavpp updated the diff for D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate.

Address review comment

Jun 21 2022, 5:03 AM · Restricted Project, Restricted Project
abinavpp added inline comments to D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate.
Jun 21 2022, 4:23 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate.

Address review comments

Jun 21 2022, 4:16 AM · Restricted Project, Restricted Project
abinavpp added a comment to D128241: [AMDGPU] Use AddedComplexity for ret and noret atomic ops selection.

This likely needs some tests?

Jun 21 2022, 2:44 AM · Restricted Project, Restricted Project

Jun 20 2022

abinavpp updated the diff for D125213: [AMDGPU] Use the HasNoUse predicate for no-ret atomic op selection.

Use only the HasNoUse predicate

Jun 20 2022, 9:57 PM · Restricted Project, Restricted Project
abinavpp retitled D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate from [GlobalISel] Implement the Has{No}Use builtin predicates to [GlobalISel] Implement the HasNoUse builtin predicate.
Jun 20 2022, 9:55 PM · Restricted Project, Restricted Project
abinavpp added inline comments to D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate.
Jun 20 2022, 9:54 PM · Restricted Project, Restricted Project
abinavpp added inline comments to D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate.
Jun 20 2022, 9:52 PM · Restricted Project, Restricted Project
abinavpp updated the diff for D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate.

Address review comments

Jun 20 2022, 9:50 PM · Restricted Project, Restricted Project
abinavpp requested review of D128241: [AMDGPU] Use AddedComplexity for ret and noret atomic ops selection.
Jun 20 2022, 9:48 PM · Restricted Project, Restricted Project

May 20 2022

abinavpp updated the diff for D125213: [AMDGPU] Use the HasNoUse predicate for no-ret atomic op selection.

Replace GISelPredicateCode with Has{No}Use predicates

May 20 2022, 7:54 AM · Restricted Project, Restricted Project
abinavpp retitled D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate from [GlobalISel] Allow destination patterns having empty outs to [GlobalISel] Implement the Has{No}Use builtin predicates.
May 20 2022, 7:52 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate.

Change the SDNPHasChain approach to the Has{No}Use predicates approach

May 20 2022, 7:51 AM · Restricted Project, Restricted Project

May 17 2022

abinavpp added a comment to D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate.

I do think there should be a builtin hasOneUse pattern predicate the emitter could use to verify

If we have a bit HasNoUse builtin predicate in PatFrags we can get the matcher
table to check the empty usage of the result without the generic C++ predicates.
The problem I'm seeing here is that the default behaviour should be to not check
for the use count since most of the selection doesn't care about it. Having a
ternary HasNoUse builtin is fine since we could set it to false for selection
that needs at least 1 use (e.g.: AMDGPU's return atomic ops), true for selection
that needs no use (e.g.: AMDGPU's no return atomic ops), "don't care" for the
majority of the other selections. I don't think we can rely on the '?'
(uninitialized) value since TreePredicateFn::isPredefinedPredicateEqualTo()
return false for uninitialized fields.

How should we do this?

Having a false/unset builtin HasOneUse predicate doesn't imply applying the negated predicate. It would just be nothing. If we wanted a hasOneOrMultiple predicate, it would be a separate PatFrag bit

May 17 2022, 4:14 AM · Restricted Project, Restricted Project

May 16 2022

abinavpp committed rG485dd0b752cd: [GlobalISel] Handle constant splat in funnel shift combine (authored by abinavpp).
[GlobalISel] Handle constant splat in funnel shift combine
May 16 2022, 3:38 AM · Restricted Project, Restricted Project
abinavpp closed D125516: [GlobalISel] Handle constant splat in funnel shift combine.
May 16 2022, 3:38 AM · Restricted Project, Restricted Project
abinavpp committed rGf96d20450c00: [AMDGPU][GlobalISel] Pre-commit tests for D125516 (authored by abinavpp).
[AMDGPU][GlobalISel] Pre-commit tests for D125516
May 16 2022, 3:38 AM · Restricted Project, Restricted Project
abinavpp closed D125539: [AMDGPU][GlobalISel] Pre-commit tests for D125516.
May 16 2022, 3:37 AM · Restricted Project, Restricted Project
abinavpp updated the diff for D125516: [GlobalISel] Handle constant splat in funnel shift combine.

Rebased; Addressed review comment

May 16 2022, 3:32 AM · Restricted Project, Restricted Project

May 13 2022

abinavpp updated the diff for D125516: [GlobalISel] Handle constant splat in funnel shift combine.

Pre-committed tests

May 13 2022, 4:29 AM · Restricted Project, Restricted Project
abinavpp requested review of D125539: [AMDGPU][GlobalISel] Pre-commit tests for D125516.
May 13 2022, 4:28 AM · Restricted Project, Restricted Project

May 12 2022

abinavpp added a comment to D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate.

I do think there should be a builtin hasOneUse pattern predicate the emitter could use to verify

May 12 2022, 10:26 PM · Restricted Project, Restricted Project
abinavpp updated the diff for D125516: [GlobalISel] Handle constant splat in funnel shift combine.

Moved matchConstantSplat() closer to struct ICstOrSplatMatch

May 12 2022, 8:49 PM · Restricted Project, Restricted Project
abinavpp requested review of D125516: [GlobalISel] Handle constant splat in funnel shift combine.
May 12 2022, 8:37 PM · Restricted Project, Restricted Project

May 9 2022

abinavpp added a comment to D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate.

Would it be helpful if the generated matcher actually checked (or maybe asserted) that the discarded defs are unused?

May 9 2022, 2:41 AM · Restricted Project, Restricted Project
abinavpp requested review of D125213: [AMDGPU] Use the HasNoUse predicate for no-ret atomic op selection.
May 9 2022, 2:00 AM · Restricted Project, Restricted Project
abinavpp requested review of D125212: [GlobalISel][SelectionDAG] Implement the HasNoUse builtin predicate.
May 9 2022, 1:56 AM · Restricted Project, Restricted Project
abinavpp committed rG7f6489d0e3cc: [AMDGPU] Regenerate checks in a mir test (authored by abinavpp).
[AMDGPU] Regenerate checks in a mir test
May 9 2022, 1:27 AM · Restricted Project, Restricted Project

Apr 22 2022

abinavpp committed rG561af89fed10: [AMDGPU] Use a wrapper multiclass for buffer atomic intrinsic patterns. NFC (authored by abinavpp).
[AMDGPU] Use a wrapper multiclass for buffer atomic intrinsic patterns. NFC
Apr 22 2022, 1:30 AM · Restricted Project, Restricted Project

Apr 21 2022

abinavpp committed rG272a8768042a: [AMDGPU] Rename the FlatSignedIntrPat multiclass to FlatSignedAtomicIntrPat. NFC (authored by abinavpp).
[AMDGPU] Rename the FlatSignedIntrPat multiclass to FlatSignedAtomicIntrPat. NFC
Apr 21 2022, 11:25 PM · Restricted Project, Restricted Project
abinavpp committed rG2147b6c89d44: [AMDGPU] Remove no-ret atomic ops selection in the post-isel hook (authored by abinavpp).
[AMDGPU] Remove no-ret atomic ops selection in the post-isel hook
Apr 21 2022, 9:19 PM · Restricted Project, Restricted Project
abinavpp closed D124086: [AMDGPU] Remove no-ret atomic ops selection in the post-isel hook.
Apr 21 2022, 9:19 PM · Restricted Project, Restricted Project
abinavpp committed rG165ae7276c27: [AMDGPU] Remove atomic pattern args in FLAT_[Global_]Atomic_Pseudo defs (authored by abinavpp).
[AMDGPU] Remove atomic pattern args in FLAT_[Global_]Atomic_Pseudo defs
Apr 21 2022, 9:18 PM · Restricted Project, Restricted Project
abinavpp closed D124084: [AMDGPU] Remove atomic pattern args in FLAT_[Global_]Atomic_Pseudo defs.
Apr 21 2022, 9:18 PM · Restricted Project, Restricted Project
abinavpp committed rGf935908d7b62: [AMDGPU] Select no-return DS_PK_ADD_F16 in tblgen (authored by abinavpp).
[AMDGPU] Select no-return DS_PK_ADD_F16 in tblgen
Apr 21 2022, 9:18 PM · Restricted Project, Restricted Project