Page MenuHomePhabricator

Hardcode84 (Ivan Butygin)
User

Projects

User does not belong to any projects.

User Details

User Since
Dec 2 2020, 2:39 AM (96 w, 2 d)

Recent Activity

Yesterday

Hardcode84 updated the diff for D135323: [mlir][gpu] Add `subgroup_reduce` operation.

add uniform requirement

Thu, Oct 6, 1:59 AM · Restricted Project, Restricted Project
Hardcode84 added a comment to D135323: [mlir][gpu] Add `subgroup_reduce` operation.

are you planning to add some lowering for it? This can be useful to at least "enforce the semantic". For instance it should be easy to have a lowering of this shuffle ops.

Thu, Oct 6, 1:28 AM · Restricted Project, Restricted Project

Wed, Oct 5

Hardcode84 updated the diff for D135323: [mlir][gpu] Add `subgroup_reduce` operation.

fix typos

Wed, Oct 5, 3:05 PM · Restricted Project, Restricted Project
Hardcode84 requested review of D135323: [mlir][gpu] Add `subgroup_reduce` operation.
Wed, Oct 5, 3:02 PM · Restricted Project, Restricted Project
Hardcode84 committed rGa93ec06ae627: [mlir][gpu] Introduce `host_shared` flag to `gpu.alloc` (authored by Hardcode84).
[mlir][gpu] Introduce `host_shared` flag to `gpu.alloc`
Wed, Oct 5, 1:04 PM · Restricted Project, Restricted Project
Hardcode84 closed D133533: [mlir][gpu] Introduce `shared` flag to `gpu.alloc`.
Wed, Oct 5, 1:03 PM · Restricted Project, Restricted Project

Sun, Oct 2

Hardcode84 added a comment to D133533: [mlir][gpu] Introduce `shared` flag to `gpu.alloc`.

Can host_shared go along with any memory space or just the default memory space

Sun, Oct 2, 4:20 AM · Restricted Project, Restricted Project

Sat, Oct 1

Hardcode84 updated the diff for D133533: [mlir][gpu] Introduce `shared` flag to `gpu.alloc`.

typo

Sat, Oct 1, 5:15 AM · Restricted Project, Restricted Project
Hardcode84 updated the diff for D133533: [mlir][gpu] Introduce `shared` flag to `gpu.alloc`.

rebase, rename to host_shared, notifyMatchFailure

Sat, Oct 1, 5:13 AM · Restricted Project, Restricted Project

Sun, Sep 25

Hardcode84 added a comment to D133533: [mlir][gpu] Introduce `shared` flag to `gpu.alloc`.

We (MLIR codegen for Intel GPUs) are currently using such memory for transferring data between GPU and Host in our pipeline. And also, we have libraries, which are using shared memory under the hood. It is very effective for integrated GPUs (as they use same memory under the hood), it is less effective for our discrete GPUs, but still usable. We may replace some of these cases with explicit copies in the future, but some cases will definitely remain.

I understand this is not ideal naming choice, do you have better idea?

host_gpu_shared?

Sun, Sep 25, 3:24 AM · Restricted Project, Restricted Project

Wed, Sep 21

Hardcode84 committed rG54d81e49e3b7: [mlir] Allow negative strides and offset in StridedLayoutAttr (authored by Hardcode84).
[mlir] Allow negative strides and offset in StridedLayoutAttr
Wed, Sep 21, 4:54 AM · Restricted Project, Restricted Project
Hardcode84 closed D134147: [mlir] Allow negative strides and offset in StridedLayoutAttr.
Wed, Sep 21, 4:53 AM · Restricted Project, Restricted Project

Mon, Sep 19

Hardcode84 updated the diff for D134147: [mlir] Allow negative strides and offset in StridedLayoutAttr.

use consumeIf, basic llvm lowering test

Mon, Sep 19, 2:33 PM · Restricted Project, Restricted Project

Sun, Sep 18

Hardcode84 added a comment to D134147: [mlir] Allow negative strides and offset in StridedLayoutAttr.

Discussion https://discourse.llvm.org/t/status-of-negative-strides-support-in-memref-subview/65346

Sun, Sep 18, 2:52 PM · Restricted Project, Restricted Project
Hardcode84 added a reviewer for D134147: [mlir] Allow negative strides and offset in StridedLayoutAttr: ftynse.
Sun, Sep 18, 2:50 PM · Restricted Project, Restricted Project
Hardcode84 requested review of D134147: [mlir] Allow negative strides and offset in StridedLayoutAttr.
Sun, Sep 18, 2:50 PM · Restricted Project, Restricted Project

Tue, Sep 13

Hardcode84 added a comment to D133533: [mlir][gpu] Introduce `shared` flag to `gpu.alloc`.

We (MLIR codegen for Intel GPUs) are currently using such memory for transferring data between GPU in Host in our pipeline. And also, we have libraries, which are using shared memory under the hood. It is very effective for integrated GPUs (as they use same memory under the hood), it is less effective for our discrete GPUs, but still usable. We may replace some of these cases with explicit copies in the future, but some cases will definitely remain.

Tue, Sep 13, 2:21 AM · Restricted Project, Restricted Project
Hardcode84 added inline comments to D133533: [mlir][gpu] Introduce `shared` flag to `gpu.alloc`.
Tue, Sep 13, 1:12 AM · Restricted Project, Restricted Project

Mon, Sep 12

Hardcode84 updated the diff for D133533: [mlir][gpu] Introduce `shared` flag to `gpu.alloc`.

do not print space, add checck in llvm lowering

Mon, Sep 12, 1:15 PM · Restricted Project, Restricted Project

Sun, Sep 11

Hardcode84 added a comment to D133533: [mlir][gpu] Introduce `shared` flag to `gpu.alloc`.

We are not using upstream gpu-to-llvm lowering.

Sun, Sep 11, 1:47 AM · Restricted Project, Restricted Project

Thu, Sep 8

Hardcode84 requested review of D133533: [mlir][gpu] Introduce `shared` flag to `gpu.alloc`.
Thu, Sep 8, 3:05 PM · Restricted Project, Restricted Project

Wed, Sep 7

Hardcode84 added inline comments to D133424: [mlir][memref] Add realloc op..
Wed, Sep 7, 1:05 PM · Restricted Project, Restricted Project

Aug 25 2022

Hardcode84 abandoned D130849: [mlir][memref] Introduce memref.offset and memref.stride ops.

Abandoning in favor of https://reviews.llvm.org/D132243

Aug 25 2022, 3:33 PM · Restricted Project, Restricted Project

Aug 19 2022

Hardcode84 added a comment to D132243: [mlir][MemRef] Introduce a memref.extract_metadata op..

Op itself is LGTM but maybe we should also add some simple test in ops.mlir?

Aug 19 2022, 9:10 AM · Restricted Project, Restricted Project

Aug 5 2022

Hardcode84 updated the diff for D130849: [mlir][memref] Introduce memref.offset and memref.stride ops.

fix test func name

Aug 5 2022, 3:41 PM · Restricted Project, Restricted Project
Hardcode84 added inline comments to D130849: [mlir][memref] Introduce memref.offset and memref.stride ops.
Aug 5 2022, 3:39 PM · Restricted Project, Restricted Project
Hardcode84 updated the diff for D130849: [mlir][memref] Introduce memref.offset and memref.stride ops.

getStridesAndOffset

Aug 5 2022, 3:37 PM · Restricted Project, Restricted Project

Jul 31 2022

Hardcode84 updated the diff for D130849: [mlir][memref] Introduce memref.offset and memref.stride ops.

fix linux build

Jul 31 2022, 10:45 AM · Restricted Project, Restricted Project
Hardcode84 added a comment to D130849: [mlir][memref] Introduce memref.offset and memref.stride ops.

Discussion https://discourse.llvm.org/t/extracting-dynamic-offsets-strides-from-memref/64170

Jul 31 2022, 9:30 AM · Restricted Project, Restricted Project
Hardcode84 added reviewers for D130849: [mlir][memref] Introduce memref.offset and memref.stride ops: nicolasvasilache, chelini.
Jul 31 2022, 9:28 AM · Restricted Project, Restricted Project
Hardcode84 requested review of D130849: [mlir][memref] Introduce memref.offset and memref.stride ops.
Jul 31 2022, 9:26 AM · Restricted Project, Restricted Project

Jul 28 2022

Hardcode84 added inline comments to D130706: [mlir] Reuse the code between `getMixed*s()` funcs in ViewLikeInterface.cpp..
Jul 28 2022, 7:12 AM · Restricted Project, Restricted Project

Jul 22 2022

Hardcode84 committed rG917e4519bc2a: [mlir][arith] cmpi: move constant to the right side (authored by Hardcode84).
[mlir][arith] cmpi: move constant to the right side
Jul 22 2022, 3:42 AM · Restricted Project, Restricted Project, Restricted Project
Hardcode84 closed D129929: [mlir][arith] cmpi: move constant to the right side.
Jul 22 2022, 3:42 AM · Restricted Project, Restricted Project, Restricted Project
Hardcode84 committed rGf46744bd2a19: [mlir][linalg] Fix FoldTensorCastConsumerOp invalid folding (authored by Hardcode84).
[mlir][linalg] Fix FoldTensorCastConsumerOp invalid folding
Jul 22 2022, 2:40 AM · Restricted Project, Restricted Project
Hardcode84 closed D130314: [mlir][linalg] Fix FoldTensorCastConsumerOp invalid folding.
Jul 22 2022, 2:39 AM · Restricted Project, Restricted Project

Jul 21 2022

Hardcode84 added reviewers for D130314: [mlir][linalg] Fix FoldTensorCastConsumerOp invalid folding: gysit, mravishankar.
Jul 21 2022, 4:13 PM · Restricted Project, Restricted Project
Hardcode84 requested review of D130314: [mlir][linalg] Fix FoldTensorCastConsumerOp invalid folding.
Jul 21 2022, 4:11 PM · Restricted Project, Restricted Project
Hardcode84 updated the diff for D129929: [mlir][arith] cmpi: move constant to the right side.

rebase, fix flang tests

Jul 21 2022, 2:42 PM · Restricted Project, Restricted Project, Restricted Project
Hardcode84 committed rGd4217e6cc86b: [mlir][memref] Missing type conversion in memref.reshape llvm lowering (authored by Hardcode84).
[mlir][memref] Missing type conversion in memref.reshape llvm lowering
Jul 21 2022, 2:17 AM · Restricted Project, Restricted Project
Hardcode84 closed D129965: [mlir][memref] Missing type conversion in memref.reshape llvm lowering.
Jul 21 2022, 2:16 AM · Restricted Project, Restricted Project

Jul 20 2022

Hardcode84 updated the diff for D129965: [mlir][memref] Missing type conversion in memref.reshape llvm lowering.

use materializeTargetConversion directly

Jul 20 2022, 2:08 PM · Restricted Project, Restricted Project

Jul 18 2022

Hardcode84 added a comment to D129965: [mlir][memref] Missing type conversion in memref.reshape llvm lowering.

Trying to understand if it can be anything else?

Jul 18 2022, 12:11 AM · Restricted Project, Restricted Project

Jul 17 2022

Hardcode84 updated the diff for D129965: [mlir][memref] Missing type conversion in memref.reshape llvm lowering.

assert msg

Jul 17 2022, 1:04 PM · Restricted Project, Restricted Project
Hardcode84 requested review of D129965: [mlir][memref] Missing type conversion in memref.reshape llvm lowering.
Jul 17 2022, 9:51 AM · Restricted Project, Restricted Project

Jul 16 2022

Hardcode84 updated the summary of D129929: [mlir][arith] cmpi: move constant to the right side.
Jul 16 2022, 1:43 PM · Restricted Project, Restricted Project, Restricted Project
Hardcode84 updated the diff for D129929: [mlir][arith] cmpi: move constant to the right side.

fix comment

Jul 16 2022, 1:43 PM · Restricted Project, Restricted Project, Restricted Project
Hardcode84 updated the diff for D129929: [mlir][arith] cmpi: move constant to the right side.

typo

Jul 16 2022, 10:52 AM · Restricted Project, Restricted Project, Restricted Project
Hardcode84 updated the diff for D129929: [mlir][arith] cmpi: move constant to the right side.

review comments

Jul 16 2022, 10:48 AM · Restricted Project, Restricted Project, Restricted Project
Hardcode84 added a comment to D129929: [mlir][arith] cmpi: move constant to the right side.

Change in sccp-structured test is suspicious, I'm not sure what this test is supposed to check and why this change is needed (scf.while didn't get removed without it). Can someone help me with this?

Jul 16 2022, 3:38 AM · Restricted Project, Restricted Project, Restricted Project
Hardcode84 added reviewers for D129929: [mlir][arith] cmpi: move constant to the right side: mehdi_amini, ftynse, rriddle.
Jul 16 2022, 3:35 AM · Restricted Project, Restricted Project, Restricted Project
Hardcode84 requested review of D129929: [mlir][arith] cmpi: move constant to the right side.
Jul 16 2022, 3:33 AM · Restricted Project, Restricted Project, Restricted Project

May 25 2022

Hardcode84 added a comment to D102799: [mlir][SCF] Canonicalize nested ParallelOps's.

Isn't the whole point of canonicalization to convert semantically equivalent codes to the single form? So we should either always merge nested loops or always split them (which makes nested loops support in scf.parallel useless).

May 25 2022, 8:06 AM · Restricted Project, Restricted Project

Mar 29 2022

Hardcode84 committed rG6f792404303a: [mlir][spirv] Mark SPV_UndefOp NoSideEffect (authored by Hardcode84).
[mlir][spirv] Mark SPV_UndefOp NoSideEffect
Mar 29 2022, 7:55 AM · Restricted Project, Restricted Project
Hardcode84 closed D122561: [mlir][spirv] Mark SPV_UndefOp NoSideEffect.
Mar 29 2022, 7:55 AM · Restricted Project, Restricted Project

Mar 28 2022

Hardcode84 requested review of D122561: [mlir][spirv] Mark SPV_UndefOp NoSideEffect.
Mar 28 2022, 1:09 AM · Restricted Project, Restricted Project

Mar 15 2022

Hardcode84 abandoned D121622: [mlir] Make CSE available as utility function.

For downstream projects it can be more convenient to have this hook so they can experiment without pushing fundamental global changes upstream (like changing behavior of existing ops, which may or may not be a good idea), but I understand your point.
For now I will just copypaste this entire pass to my project and do necessary changes locally.

Mar 15 2022, 2:34 PM · Restricted Project, Restricted Project
Hardcode84 committed rG9f864a544750: [mlir][gpu] Introduce gpu.global_id op (authored by Hardcode84).
[mlir][gpu] Introduce gpu.global_id op
Mar 15 2022, 3:28 AM · Restricted Project
Hardcode84 closed D121548: [mlir][gpu] Introduce gpu.global_id op.
Mar 15 2022, 3:28 AM · Restricted Project, Restricted Project
Hardcode84 committed rG7b0e041df8c0: [mlir][spirv] Add AssumeTrueKHROp (authored by Hardcode84).
[mlir][spirv] Add AssumeTrueKHROp
Mar 15 2022, 3:25 AM · Restricted Project
Hardcode84 closed D121601: [mlir][spirv] Add AssumeTrueKHROp.
Mar 15 2022, 3:25 AM · Restricted Project, Restricted Project

Mar 14 2022

Hardcode84 added a comment to D121622: [mlir] Make CSE available as utility function.

Wouldn't it be better to standardize an attribute for this instead?

Mar 14 2022, 12:22 PM · Restricted Project, Restricted Project
Hardcode84 added a comment to D121622: [mlir] Make CSE available as utility function.
  1. I have a custom "mydialect.noSideEffect" attribute, which I can attach to any op (e.g. CallOp) and I want CSE to always simplify them
  2. I want to write a pass which CSE loads from constant data (which is not strictly MemoryEffectOpInterface::hasNoEffect)
Mar 14 2022, 10:40 AM · Restricted Project, Restricted Project
Hardcode84 added a comment to D121622: [mlir] Make CSE available as utility function.

It needed for CSE passes with some custom side-effect logic.

Mar 14 2022, 10:25 AM · Restricted Project, Restricted Project
Hardcode84 added a reviewer for D121622: [mlir] Make CSE available as utility function: jpienaar.
Mar 14 2022, 10:21 AM · Restricted Project, Restricted Project
Hardcode84 requested review of D121622: [mlir] Make CSE available as utility function.
Mar 14 2022, 10:17 AM · Restricted Project, Restricted Project
Hardcode84 requested review of D121601: [mlir][spirv] Add AssumeTrueKHROp.
Mar 14 2022, 7:20 AM · Restricted Project, Restricted Project
Hardcode84 updated the diff for D121548: [mlir][gpu] Introduce gpu.global_id op.

rebase, update op description

Mar 14 2022, 6:47 AM · Restricted Project, Restricted Project

Mar 13 2022

Hardcode84 updated the diff for D121548: [mlir][gpu] Introduce gpu.global_id op.

expand description

Mar 13 2022, 1:02 PM · Restricted Project, Restricted Project
Hardcode84 updated the diff for D121548: [mlir][gpu] Introduce gpu.global_id op.

fix op description

Mar 13 2022, 4:07 AM · Restricted Project, Restricted Project
Hardcode84 requested review of D121548: [mlir][gpu] Introduce gpu.global_id op.
Mar 13 2022, 4:03 AM · Restricted Project, Restricted Project

Mar 11 2022

Hardcode84 committed rG4df95441089a: [mlir][spirv] Make EntryPointABIAttr.local_size optional (authored by Hardcode84).
[mlir][spirv] Make EntryPointABIAttr.local_size optional
Mar 11 2022, 11:26 AM · Restricted Project
Hardcode84 closed D120399: [mlir][spirv] Make EntryPointABIAttr.local_size optional.
Mar 11 2022, 11:25 AM · Restricted Project, Restricted Project
Hardcode84 updated the diff for D120399: [mlir][spirv] Make EntryPointABIAttr.local_size optional.

rebase

Mar 11 2022, 10:51 AM · Restricted Project, Restricted Project
Hardcode84 committed rGf3676c3273b9: [mlir][memref] memref.reinterpret_cast folding (authored by Hardcode84).
[mlir][memref] memref.reinterpret_cast folding
Mar 11 2022, 10:23 AM · Restricted Project
Hardcode84 closed D120242: [mlir][memref] memref.reinterpret_cast folding.
Mar 11 2022, 10:23 AM · Restricted Project, Restricted Project
Hardcode84 updated the diff for D120242: [mlir][memref] memref.reinterpret_cast folding.

rebase, review comments

Mar 11 2022, 6:40 AM · Restricted Project, Restricted Project

Mar 8 2022

Hardcode84 added a comment to D120399: [mlir][spirv] Make EntryPointABIAttr.local_size optional.

@antiagainst ping

Mar 8 2022, 11:37 PM · Restricted Project, Restricted Project

Mar 3 2022

Hardcode84 added inline comments to D120617: [MLIR] Extend isLoopMemoryParallel to account for locally allocated memrefs.
Mar 3 2022, 6:17 AM · Restricted Project, Restricted Project

Mar 2 2022

Herald added a project to D120399: [mlir][spirv] Make EntryPointABIAttr.local_size optional: Restricted Project.

ping

Mar 2 2022, 6:10 AM · Restricted Project, Restricted Project
Herald added a project to D120242: [mlir][memref] memref.reinterpret_cast folding: Restricted Project.

ping

Mar 2 2022, 6:10 AM · Restricted Project, Restricted Project

Feb 23 2022

Hardcode84 requested review of D120399: [mlir][spirv] Make EntryPointABIAttr.local_size optional.
Feb 23 2022, 3:14 AM · Restricted Project, Restricted Project

Feb 21 2022

Hardcode84 added inline comments to D120242: [mlir][memref] memref.reinterpret_cast folding.
Feb 21 2022, 10:02 AM · Restricted Project, Restricted Project
Hardcode84 updated the diff for D120242: [mlir][memref] memref.reinterpret_cast folding.

add comments

Feb 21 2022, 9:58 AM · Restricted Project, Restricted Project
Hardcode84 updated the diff for D120242: [mlir][memref] memref.reinterpret_cast folding.

fix gcc build

Feb 21 2022, 5:41 AM · Restricted Project, Restricted Project
Hardcode84 added a reviewer for D120242: [mlir][memref] memref.reinterpret_cast folding: nicolasvasilache.
Feb 21 2022, 4:23 AM · Restricted Project, Restricted Project
Hardcode84 updated the diff for D120242: [mlir][memref] memref.reinterpret_cast folding.

fix style

Feb 21 2022, 4:22 AM · Restricted Project, Restricted Project
Hardcode84 requested review of D120242: [mlir][memref] memref.reinterpret_cast folding.
Feb 21 2022, 4:20 AM · Restricted Project, Restricted Project

Feb 17 2022

Hardcode84 added a comment to D119932: [mlir][gpu] Split ops sinking from gpu-kernel-outlining pass into separate pass.

mlir/tools/mlir-vulkan-runner/mlir-vulkan-runner.cpp did use the pass, was it intentional to change behavior, or should it now also have the new pass before outlining?

It should be safe to add this new pass before outlining to exactly preserve old behavior but I'm don't know if it really needed there.

Feb 17 2022, 12:48 AM · Restricted Project

Feb 16 2022

Hardcode84 committed rGd271fc04d5b9: [mlir][gpu] Split ops sinking from gpu-kernel-outlining pass into separate pass (authored by Hardcode84).
[mlir][gpu] Split ops sinking from gpu-kernel-outlining pass into separate pass
Feb 16 2022, 11:52 PM
Hardcode84 closed D119932: [mlir][gpu] Split ops sinking from gpu-kernel-outlining pass into separate pass.
Feb 16 2022, 11:51 PM · Restricted Project
Hardcode84 retitled D119932: [mlir][gpu] Split ops sinking from gpu-kernel-outlining pass into separate pass from [mlir][gpu] Split ops sinking from gpu-kernel-outlining pass ito separate pass to [mlir][gpu] Split ops sinking from gpu-kernel-outlining pass into separate pass.
Feb 16 2022, 7:34 AM · Restricted Project
Hardcode84 updated the diff for D119932: [mlir][gpu] Split ops sinking from gpu-kernel-outlining pass into separate pass.

rename pass

Feb 16 2022, 7:32 AM · Restricted Project
Hardcode84 updated the diff for D119932: [mlir][gpu] Split ops sinking from gpu-kernel-outlining pass into separate pass.

fix Passes.td

Feb 16 2022, 5:48 AM · Restricted Project
Hardcode84 updated the diff for D119932: [mlir][gpu] Split ops sinking from gpu-kernel-outlining pass into separate pass.

fix missing return

Feb 16 2022, 4:48 AM · Restricted Project
Hardcode84 requested review of D119932: [mlir][gpu] Split ops sinking from gpu-kernel-outlining pass into separate pass.
Feb 16 2022, 4:40 AM · Restricted Project

Feb 15 2022

Hardcode84 committed rGa2e2fbba17ac: [mlir][gpu] sinkOperationsIntoLaunchOp: Add user hook for isSinkingBeneficiary (authored by Hardcode84).
[mlir][gpu] sinkOperationsIntoLaunchOp: Add user hook for isSinkingBeneficiary
Feb 15 2022, 5:54 AM
Hardcode84 closed D119632: [mlir][gpu] sinkOperationsIntoLaunchOp: Add user hook for isSinkingBeneficiary.
Feb 15 2022, 5:53 AM · Restricted Project
Hardcode84 updated the diff for D119632: [mlir][gpu] sinkOperationsIntoLaunchOp: Add user hook for isSinkingBeneficiary.

rebase, remove unittest, rename default func

Feb 15 2022, 3:26 AM · Restricted Project
Hardcode84 committed rG32389d0c2e2d: [mlir][spirv] Add OpenCL fma op and lowering (authored by Hardcode84).
[mlir][spirv] Add OpenCL fma op and lowering
Feb 15 2022, 12:30 AM