Page MenuHomePhabricator

tpr (Tim Renouf)
User

Projects

User does not belong to any projects.

User Details

User Since
Jan 26 2017, 7:07 AM (141 w, 4 d)

Recent Activity

Mon, Sep 30

tpr updated the diff for D68231: [SLC] Allow llvm.pow(x,2.0) -> x*x etc even if no pow() lib func.

V2: Better fix that does not accidentally allow pow() transforms.

Mon, Sep 30, 11:43 AM · Restricted Project
tpr updated subscribers of D68231: [SLC] Allow llvm.pow(x,2.0) -> x*x etc even if no pow() lib func.

@foad pointed out that this fix is wrong. TLI saying pow() is not supported means if we find a call to a function called pow() then we don't know its semantics. So I will push a revised fix.

Mon, Sep 30, 11:06 AM · Restricted Project
tpr added reviewers for D68231: [SLC] Allow llvm.pow(x,2.0) -> x*x etc even if no pow() lib func: evandro, xbolva00.
Mon, Sep 30, 9:35 AM · Restricted Project
tpr created D68231: [SLC] Allow llvm.pow(x,2.0) -> x*x etc even if no pow() lib func.
Mon, Sep 30, 9:33 AM · Restricted Project

Mon, Sep 23

tpr committed rL372563: Request commit access for tpr.
Request commit access for tpr
Mon, Sep 23, 2:21 AM

Wed, Sep 18

tpr committed rG178611711122: [AMDGPU] Allow FP inline constant in v_madak_f16 and v_fmaak_f16 (authored by tpr).
[AMDGPU] Allow FP inline constant in v_madak_f16 and v_fmaak_f16
Wed, Sep 18, 2:31 AM
tpr committed rL372208: [AMDGPU] Allow FP inline constant in v_madak_f16 and v_fmaak_f16.
[AMDGPU] Allow FP inline constant in v_madak_f16 and v_fmaak_f16
Wed, Sep 18, 2:30 AM
tpr closed D67680: [AMDGPU] Allow FP inline constant in v_madak_f16 and v_fmaak_f16.
Wed, Sep 18, 2:30 AM · Restricted Project

Tue, Sep 17

tpr added a comment to D67680: [AMDGPU] Allow FP inline constant in v_madak_f16 and v_fmaak_f16.

LGTM. Can you also add a test for the mad case

Tue, Sep 17, 3:06 PM · Restricted Project
tpr added reviewers for D67680: [AMDGPU] Allow FP inline constant in v_madak_f16 and v_fmaak_f16: arsenm, kzhuravl, rampitec, vpykhtin.
Tue, Sep 17, 2:34 PM · Restricted Project
tpr created D67680: [AMDGPU] Allow FP inline constant in v_madak_f16 and v_fmaak_f16.
Tue, Sep 17, 2:27 PM · Restricted Project

Sep 11 2019

tpr committed rGc26b3940c320: [TLI][AMDGPU] AMDPAL does not have library functions (authored by tpr).
[TLI][AMDGPU] AMDPAL does not have library functions
Sep 11 2019, 12:30 AM
tpr committed rL371592: [TLI][AMDGPU] AMDPAL does not have library functions.
[TLI][AMDGPU] AMDPAL does not have library functions
Sep 11 2019, 12:26 AM
tpr closed D67406: [TLI][AMDGPU] AMDPAL does not have tan function.
Sep 11 2019, 12:26 AM · Restricted Project

Sep 10 2019

tpr updated the diff for D67406: [TLI][AMDGPU] AMDPAL does not have tan function.

V2: Disable all library functions, not just tan.

Sep 10 2019, 12:09 PM · Restricted Project
tpr added inline comments to D67406: [TLI][AMDGPU] AMDPAL does not have tan function.
Sep 10 2019, 9:34 AM · Restricted Project
tpr added reviewers for D67406: [TLI][AMDGPU] AMDPAL does not have tan function: dlj, nhaehnle.
Sep 10 2019, 9:19 AM · Restricted Project
tpr created D67406: [TLI][AMDGPU] AMDPAL does not have tan function.
Sep 10 2019, 9:13 AM · Restricted Project

Sep 2 2019

tpr added a comment to D67003: AMDGPU: Don't put constants in .text for Mesa.

I just noticed that this already came up in D65813 and it does the right thing, it's just waiting review.

Sep 2 2019, 4:19 AM · Restricted Project

Aug 13 2019

tpr committed rG10db641aabf0: [AMDGPU] Fix to 'Fold readlane from copy of SGPR or imm' (authored by tpr).
[AMDGPU] Fix to 'Fold readlane from copy of SGPR or imm'
Aug 13 2019, 12:02 PM
tpr committed rL368736: [AMDGPU] Fix to 'Fold readlane from copy of SGPR or imm'.
[AMDGPU] Fix to 'Fold readlane from copy of SGPR or imm'
Aug 13 2019, 11:57 AM
tpr closed D66133: [AMDGPU] Fix to 'Fold readlane from copy of SGPR or imm'.
Aug 13 2019, 11:57 AM · Restricted Project
tpr created D66133: [AMDGPU] Fix to 'Fold readlane from copy of SGPR or imm'.
Aug 13 2019, 3:52 AM · Restricted Project
tpr added a reviewer for D66133: [AMDGPU] Fix to 'Fold readlane from copy of SGPR or imm': arsenm.
Aug 13 2019, 3:52 AM · Restricted Project

Aug 6 2019

tpr committed rG5a0794327a67: [StructurizeCFG] Enable -structurizecfg-relaxed-uniform-regions by default (authored by tpr).
[StructurizeCFG] Enable -structurizecfg-relaxed-uniform-regions by default
Aug 6 2019, 7:31 AM
tpr committed rL368042: [StructurizeCFG] Enable -structurizecfg-relaxed-uniform-regions by default.
[StructurizeCFG] Enable -structurizecfg-relaxed-uniform-regions by default
Aug 6 2019, 7:29 AM
tpr closed D63198: [StructurizeCFG] Enable -structurizecfg-relaxed-uniform-regions by default.
Aug 6 2019, 7:29 AM · Restricted Project

Jul 4 2019

tpr committed rG5816889c748b: [AMDGPU] Custom lower INSERT_SUBVECTOR v3, v4, v5, v8 (authored by tpr).
[AMDGPU] Custom lower INSERT_SUBVECTOR v3, v4, v5, v8
Jul 4 2019, 10:39 AM
tpr committed rL365148: [AMDGPU] Custom lower INSERT_SUBVECTOR v3, v4, v5, v8.
[AMDGPU] Custom lower INSERT_SUBVECTOR v3, v4, v5, v8
Jul 4 2019, 10:38 AM
tpr closed D63160: [AMDGPU] Custom lower INSERT_SUBVECTOR v3, v4, v5, v8.
Jul 4 2019, 10:38 AM · Restricted Project

Jun 24 2019

tpr added a comment to D63712: [AMDGPU] Fix +DumpCode to print an entry label for the first function.

Possibly, but our majority use case is to not need the disassembly, and our thinking is that we don't want to change the tool flow between the majority case and the minority case. So can we get this fix in please?

Jun 24 2019, 2:31 PM · Restricted Project
tpr committed rGd2fdb956e044: [AMDGPU] Allow any value in unused src0 field in v_nop (authored by tpr).
[AMDGPU] Allow any value in unused src0 field in v_nop
Jun 24 2019, 10:39 AM
tpr committed rL364208: [AMDGPU] Allow any value in unused src0 field in v_nop.
[AMDGPU] Allow any value in unused src0 field in v_nop
Jun 24 2019, 10:36 AM
tpr closed D63724: [AMDGPU] Allow any value in unused src0 field in v_nop.
Jun 24 2019, 10:36 AM · Restricted Project
tpr added reviewers for D63724: [AMDGPU] Allow any value in unused src0 field in v_nop: rampitec, kzhuravl.
Jun 24 2019, 9:27 AM · Restricted Project
tpr created D63724: [AMDGPU] Allow any value in unused src0 field in v_nop.
Jun 24 2019, 9:27 AM · Restricted Project
tpr added a comment to D63712: [AMDGPU] Fix +DumpCode to print an entry label for the first function.

I don't think anyone is a fan of dumpcode. But we're still in the position that the proper disassembler does not support gfx6 or gfx7, and we need to get this particular problem fixed in the short term.

Jun 24 2019, 6:13 AM · Restricted Project

Jun 19 2019

tpr added a comment to D63510: [LiveInterval] Removed bogus empty subrange assert.

For this bug, whatever we do with a mir test, it is not going to be reliable in failing if the bug is present. Maybe there is a unit testing framework for LiveRangeCalc tests that I could add a test to.

Jun 19 2019, 12:04 PM · Restricted Project
tpr added inline comments to D63510: [LiveInterval] Removed bogus empty subrange assert.
Jun 19 2019, 12:01 PM · Restricted Project
tpr added inline comments to D63510: [LiveInterval] Removed bogus empty subrange assert.
Jun 19 2019, 2:30 AM · Restricted Project

Jun 18 2019

tpr added reviewers for D63510: [LiveInterval] Removed bogus empty subrange assert: MatzeB, qcolombet.
Jun 18 2019, 12:29 PM · Restricted Project
tpr created D63510: [LiveInterval] Removed bogus empty subrange assert.
Jun 18 2019, 12:29 PM · Restricted Project

Jun 13 2019

tpr added reviewers for D63198: [StructurizeCFG] Enable -structurizecfg-relaxed-uniform-regions by default: nhaehnle, arsenm, kzhuravl.
Jun 13 2019, 3:33 AM · Restricted Project

Jun 12 2019

tpr updated the diff for D63198: [StructurizeCFG] Enable -structurizecfg-relaxed-uniform-regions by default.

V2: Lit test fix.

Jun 12 2019, 12:26 PM · Restricted Project
tpr created D63198: [StructurizeCFG] Enable -structurizecfg-relaxed-uniform-regions by default.
Jun 12 2019, 6:17 AM · Restricted Project
tpr added inline comments to D63160: [AMDGPU] Custom lower INSERT_SUBVECTOR v3, v4, v5, v8.
Jun 12 2019, 3:32 AM · Restricted Project
tpr updated the diff for D63160: [AMDGPU] Custom lower INSERT_SUBVECTOR v3, v4, v5, v8.

V2: Addressed review comments re test.

Jun 12 2019, 3:30 AM · Restricted Project

Jun 11 2019

tpr created D63160: [AMDGPU] Custom lower INSERT_SUBVECTOR v3, v4, v5, v8.
Jun 11 2019, 1:10 PM · Restricted Project

May 30 2019

tpr committed rG7fecdf36cc5b: [AMDGPU] Added target-specific attribute amdgpu-max-memory-clause (authored by tpr).
[AMDGPU] Added target-specific attribute amdgpu-max-memory-clause
May 30 2019, 11:49 AM
tpr committed rL362127: [AMDGPU] Added target-specific attribute amdgpu-max-memory-clause.
[AMDGPU] Added target-specific attribute amdgpu-max-memory-clause
May 30 2019, 11:43 AM
tpr closed D62572: [AMDGPU] Added target feature +disable-form-clauses.
May 30 2019, 11:43 AM · Restricted Project
tpr updated the diff for D62572: [AMDGPU] Added target feature +disable-form-clauses.

V2: Target-specific attribute instead of target feature, as suggested by Stas.

May 30 2019, 8:25 AM · Restricted Project

May 29 2019

tpr added a reviewer for D62572: [AMDGPU] Added target feature +disable-form-clauses: rampitec.
May 29 2019, 1:14 AM · Restricted Project
tpr created D62572: [AMDGPU] Added target feature +disable-form-clauses.
May 29 2019, 1:12 AM · Restricted Project

May 21 2019

tpr added a comment to D60762: [SelectionDAG] Legalize vaargs that require vector splitting.

LGTM, but I don't think I know the legalization code well enough to approve this.

May 21 2019, 6:35 AM · Restricted Project

May 16 2019

tpr committed rGe3cbdaf1b5e7: [CodeGen] Fixed de-optimization of legalize subvector extract (authored by tpr).
[CodeGen] Fixed de-optimization of legalize subvector extract
May 16 2019, 2:47 PM
tpr committed rL360942: [CodeGen] Fixed de-optimization of legalize subvector extract.
[CodeGen] Fixed de-optimization of legalize subvector extract
May 16 2019, 2:46 PM
tpr closed D60457: [CodeGen] Fixed de-optimization of legalize subvector extract.
May 16 2019, 2:46 PM · Restricted Project
tpr added a comment to D60457: [CodeGen] Fixed de-optimization of legalize subvector extract.

Is someone now able to approve this? Eli?

May 16 2019, 10:22 AM · Restricted Project

May 14 2019

tpr committed rG33cb8f5b547c: [AMDGPU] Fixed +DumpCode (authored by tpr).
[AMDGPU] Fixed +DumpCode
May 14 2019, 9:17 AM
tpr committed rL360688: [AMDGPU] Fixed +DumpCode.
[AMDGPU] Fixed +DumpCode
May 14 2019, 9:15 AM
tpr closed D60682: [AMDGPU] Fixed +DumpCode.
May 14 2019, 9:14 AM · Restricted Project
tpr added inline comments to D60457: [CodeGen] Fixed de-optimization of legalize subvector extract.
May 14 2019, 8:09 AM · Restricted Project
tpr updated the diff for D60457: [CodeGen] Fixed de-optimization of legalize subvector extract.

V4: update_mir_test_checks the test.

May 14 2019, 8:06 AM · Restricted Project

May 8 2019

tpr added a comment to D60682: [AMDGPU] Fixed +DumpCode.

Ping: Could someone approve this please? Thanks.

May 8 2019, 2:36 AM · Restricted Project
tpr added a comment to D60457: [CodeGen] Fixed de-optimization of legalize subvector extract.

Hi Samuel. Sorry for the delay; I kind of lost track of this change.

May 8 2019, 2:31 AM · Restricted Project

Apr 17 2019

tpr committed rG7c55c8d8c347: [AMDGPU] Avoid DAG combining assert with fneg(fadd(A,0)) (authored by tpr).
[AMDGPU] Avoid DAG combining assert with fneg(fadd(A,0))
Apr 17 2019, 10:26 PM
tpr committed rL358640: [AMDGPU] Avoid DAG combining assert with fneg(fadd(A,0)).
[AMDGPU] Avoid DAG combining assert with fneg(fadd(A,0))
Apr 17 2019, 10:26 PM
tpr closed D60633: [AMDGPU] Avoid DAG combining assert with fneg(fadd(A,0)).
Apr 17 2019, 10:26 PM · Restricted Project
tpr committed rG59e8bd3093dc: [AMDGPU] Flag new raw/struct atomic ops as source of divergence (authored by tpr).
[AMDGPU] Flag new raw/struct atomic ops as source of divergence
Apr 17 2019, 7:03 AM
tpr committed rL358579: [AMDGPU] Flag new raw/struct atomic ops as source of divergence.
[AMDGPU] Flag new raw/struct atomic ops as source of divergence
Apr 17 2019, 7:03 AM
tpr closed D60731: [AMDGPU] Flag new raw/struct atomic ops as source of divergence.
Apr 17 2019, 7:03 AM · Restricted Project
tpr added inline comments to D60457: [CodeGen] Fixed de-optimization of legalize subvector extract.
Apr 17 2019, 3:12 AM · Restricted Project
tpr updated the diff for D60457: [CodeGen] Fixed de-optimization of legalize subvector extract.

V3: Further reduced test case.

Apr 17 2019, 3:12 AM · Restricted Project
tpr updated the diff for D60457: [CodeGen] Fixed de-optimization of legalize subvector extract.

V2: Addressed review comments.

Apr 17 2019, 3:05 AM · Restricted Project
tpr updated the diff for D60731: [AMDGPU] Flag new raw/struct atomic ops as source of divergence.

V2: Addressed review comment: moved tests.

Apr 17 2019, 2:16 AM · Restricted Project

Apr 15 2019

tpr added reviewers for D60600: [InstCombine] Fix a vector-of-pointers instcombine undef bug.: reames, spatel.
Apr 15 2019, 2:51 PM · Restricted Project
tpr added a reviewer for D60731: [AMDGPU] Flag new raw/struct atomic ops as source of divergence: nhaehnle.
Apr 15 2019, 2:30 PM · Restricted Project
tpr created D60731: [AMDGPU] Flag new raw/struct atomic ops as source of divergence.
Apr 15 2019, 2:30 PM · Restricted Project
tpr added a comment to D60633: [AMDGPU] Avoid DAG combining assert with fneg(fadd(A,0)).

Bug for adding remaining test cases is https://bugs.llvm.org/show_bug.cgi?id=41500

Apr 15 2019, 7:55 AM · Restricted Project
tpr updated the diff for D60633: [AMDGPU] Avoid DAG combining assert with fneg(fadd(A,0)).

V3: Addressed review comments.

Apr 15 2019, 7:55 AM · Restricted Project
tpr added a comment to D60682: [AMDGPU] Fixed +DumpCode.

PS I'm only speaking for LLPC. I don't know if Mesa uses it.

Apr 15 2019, 6:18 AM · Restricted Project
tpr added a comment to D60682: [AMDGPU] Fixed +DumpCode.

I think the only obstacle is getting round to doing the LLPC changes. So this is hopefully a short term fix until we can get around to it.

Apr 15 2019, 6:18 AM · Restricted Project
tpr committed rG842be38162ec: [AMDGPU] Fixed incorrect test in vcnd/vcmp optimization (authored by tpr).
[AMDGPU] Fixed incorrect test in vcnd/vcmp optimization
Apr 15 2019, 3:35 AM
tpr committed rL358392: [AMDGPU] Fixed incorrect test in vcnd/vcmp optimization.
[AMDGPU] Fixed incorrect test in vcnd/vcmp optimization
Apr 15 2019, 3:35 AM
tpr closed D60652: [AMDGPU] Fixed incorrect test in vcnd/vcmp optimization.
Apr 15 2019, 3:35 AM · Restricted Project
tpr added reviewers for D60682: [AMDGPU] Fixed +DumpCode: scott.linder, kzhuravl.
Apr 15 2019, 3:18 AM · Restricted Project
tpr created D60682: [AMDGPU] Fixed +DumpCode.
Apr 15 2019, 3:17 AM · Restricted Project
tpr added a comment to D60633: [AMDGPU] Avoid DAG combining assert with fneg(fadd(A,0)).

I have cut down the test a bit more and put it into fneg-combines.ll. I did not manage to repro any problems with the other cases that I added fixes for.

Apr 15 2019, 3:06 AM · Restricted Project
tpr updated the diff for D60633: [AMDGPU] Avoid DAG combining assert with fneg(fadd(A,0)).

V2: Cut down test a bit more and put it in fneg-combines.ll.

Apr 15 2019, 3:03 AM · Restricted Project

Apr 13 2019

tpr added reviewers for D60652: [AMDGPU] Fixed incorrect test in vcnd/vcmp optimization: vpykhtin, rampitec.

I did not add a test because I believe that we do not currently set the source modifiers in codegen, so this code never gets exercised.

Apr 13 2019, 9:36 AM · Restricted Project
tpr created D60652: [AMDGPU] Fixed incorrect test in vcnd/vcmp optimization.
Apr 13 2019, 9:32 AM · Restricted Project

Apr 12 2019

tpr added reviewers for D60633: [AMDGPU] Avoid DAG combining assert with fneg(fadd(A,0)): arsenm, rampitec.

That's the best bugpoint could do with the test.

Apr 12 2019, 1:22 PM · Restricted Project
tpr created D60633: [AMDGPU] Avoid DAG combining assert with fneg(fadd(A,0)).
Apr 12 2019, 1:16 PM · Restricted Project

Apr 9 2019

tpr added reviewers for D60457: [CodeGen] Fixed de-optimization of legalize subvector extract: hakzsam, arsenm, efriedma.
Apr 9 2019, 6:10 AM · Restricted Project
tpr added a comment to D58902: [AMDGPU] Support for v3i32/v3f32.

Hopefully fixed by D60457.

Apr 9 2019, 5:57 AM · Restricted Project
tpr created D60457: [CodeGen] Fixed de-optimization of legalize subvector extract.
Apr 9 2019, 5:50 AM · Restricted Project

Apr 1 2019

tpr added a comment to D58902: [AMDGPU] Support for v3i32/v3f32.

Hi Samuel

Apr 1 2019, 2:11 AM · Restricted Project

Mar 29 2019

tpr accepted D59295: [AMDGPU] Pre-allocate WWM registers to reduce VGPR pressure..

LGTM modulo the wrong license on the new file.

Mar 29 2019, 1:53 PM · Restricted Project, Restricted Project

Mar 27 2019

tpr abandoned D59871: [AMDGPU] Avoid mistaking a ds atomic for a load.

OK thanks Matt.

Mar 27 2019, 9:46 AM · Restricted Project