Page MenuHomePhabricator

kzhuravl (Konstantin Zhuravlyov)
User

Projects

User does not belong to any projects.

User Details

User Since
Feb 16 2016, 12:13 PM (147 w, 6 d)

Recent Activity

Tue, Dec 11

kzhuravl accepted D48179: [AMDGPU] Emit MessagePack HSA Metadata for v3 code object.

LGTM

Tue, Dec 11, 9:27 PM
kzhuravl accepted D55267: [AMDGPU] Set metadata access for explicit section.

LGTM

Tue, Dec 11, 5:23 AM

Mon, Nov 19

kzhuravl added a comment to D54545: AMDGPU: Fix V_FMA_F16 selection on GFX9.

Is this the one that zeros the high bits vs. leaves them intact?

Mon, Nov 19, 1:14 PM
kzhuravl committed rL347265: AMDGPU: Fix V_FMA_F16 selection on GFX9.
AMDGPU: Fix V_FMA_F16 selection on GFX9
Mon, Nov 19, 1:13 PM
kzhuravl closed D54545: AMDGPU: Fix V_FMA_F16 selection on GFX9.
Mon, Nov 19, 1:12 PM
kzhuravl accepted D54718: Restored selection of scalar_to_vector (v2x16).

LGTM

Mon, Nov 19, 11:59 AM

Nov 16 2018

kzhuravl accepted D54301: [AMDGPU] Derive GCNSubtarget from MF to get overridden target features.

LGTM

Nov 16 2018, 10:06 AM

Nov 15 2018

kzhuravl committed rL347001: AMDHSA: More code object v3 fixes:.
AMDHSA: More code object v3 fixes:
Nov 15 2018, 3:16 PM
kzhuravl added a comment to D48179: [AMDGPU] Emit MessagePack HSA Metadata for v3 code object.

what is the rationale for putting the metadata verifier in binary format instead of support?

Nov 15 2018, 12:46 PM

Nov 14 2018

kzhuravl committed rL346925: AMDGPU: Fix check lines in fdot2 test:.
AMDGPU: Fix check lines in fdot2 test:
Nov 14 2018, 6:45 PM
kzhuravl committed rL346923: AMDGPU: Enable code object v3 for AMDHSA only.
AMDGPU: Enable code object v3 for AMDHSA only
Nov 14 2018, 6:35 PM
kzhuravl closed D54186: AMDGPU: Enable code object v3 for AMDHSA only.
Nov 14 2018, 6:35 PM
kzhuravl created D54545: AMDGPU: Fix V_FMA_F16 selection on GFX9.
Nov 14 2018, 2:01 PM

Nov 8 2018

kzhuravl requested changes to D48179: [AMDGPU] Emit MessagePack HSA Metadata for v3 code object.

Can you also rebase your change? I have done some changes to your patch in amd-common that you need to incorporate here.

Nov 8 2018, 12:43 PM

Nov 7 2018

kzhuravl committed rL346361: AMDGPU/NFC: Split FLAT_Global_Atomic_Pseudo into RTN/NO_RTN multiclasses.
AMDGPU/NFC: Split FLAT_Global_Atomic_Pseudo into RTN/NO_RTN multiclasses
Nov 7 2018, 1:44 PM
kzhuravl committed rL346357: AMDGPU/NFC: Split MUBUF_Pseudo_Atomics into RTN/NO_RTN multiclasses.
AMDGPU/NFC: Split MUBUF_Pseudo_Atomics into RTN/NO_RTN multiclasses
Nov 7 2018, 1:24 PM
kzhuravl committed rL346354: AMDGPU/Docs: Add product names for Vega20.
AMDGPU/Docs: Add product names for Vega20
Nov 7 2018, 12:56 PM
kzhuravl closed D54178: AMDGPU/Docs: Add product names for Vega20.
Nov 7 2018, 12:56 PM

Nov 6 2018

kzhuravl created D54186: AMDGPU: Enable code object v3 for AMDHSA only.
Nov 6 2018, 3:06 PM
kzhuravl created D54178: AMDGPU/Docs: Add product names for Vega20.
Nov 6 2018, 12:50 PM
kzhuravl committed rL346263: AMDGPU/Docs: Fix the processor table.
AMDGPU/Docs: Fix the processor table
Nov 6 2018, 12:26 PM

Nov 5 2018

kzhuravl committed rL346178: AMDGPU: Add sram-ecc feature options.
AMDGPU: Add sram-ecc feature options
Nov 5 2018, 2:47 PM
kzhuravl committed rC346178: AMDGPU: Add sram-ecc feature options.
AMDGPU: Add sram-ecc feature options
Nov 5 2018, 2:47 PM
kzhuravl closed D53223: AMDGPU: Add sram-ecc feature options.
Nov 5 2018, 2:47 PM
kzhuravl closed D53223: AMDGPU: Add sram-ecc feature options.
Nov 5 2018, 2:47 PM
kzhuravl committed rL346177: AMDGPU: Add sram-ecc feature.
AMDGPU: Add sram-ecc feature
Nov 5 2018, 2:47 PM
kzhuravl closed D53222: AMDGPU: Add sram-ecc feature.
Nov 5 2018, 2:46 PM
kzhuravl added a comment to D53222: AMDGPU: Add sram-ecc feature.

Does the assembler need any support for this? For example, does the .amdgcn_target directive need to accept the +sramecc and ensure the e_flag is set accordingly?

Nov 5 2018, 10:59 AM
kzhuravl added a comment to D53223: AMDGPU: Add sram-ecc feature options.

ping

Nov 5 2018, 8:42 AM
kzhuravl added a comment to D53222: AMDGPU: Add sram-ecc feature.

ping

Nov 5 2018, 8:42 AM

Oct 30 2018

kzhuravl committed rL345662: Revert r345542: AMDGPU: Enable code object v3 by default.
Revert r345542: AMDGPU: Enable code object v3 by default
Oct 30 2018, 3:05 PM

Oct 29 2018

kzhuravl committed rL345542: AMDGPU: Enable code object v3 by default.
AMDGPU: Enable code object v3 by default
Oct 29 2018, 2:10 PM
kzhuravl closed D53525: AMDGPU: Enable code object v3 by default.
Oct 29 2018, 2:10 PM
kzhuravl committed rLLD345530: AMDGPU: Switch some lld tests to v2.
AMDGPU: Switch some lld tests to v2
Oct 29 2018, 1:13 PM
kzhuravl committed rL345530: AMDGPU: Switch some lld tests to v2.
AMDGPU: Switch some lld tests to v2
Oct 29 2018, 1:01 PM
kzhuravl closed D53526: AMDGPU: Switch some lld tests to v2.
Oct 29 2018, 1:01 PM

Oct 24 2018

kzhuravl added a comment to D53223: AMDGPU: Add sram-ecc feature options.

ping

Oct 24 2018, 12:59 PM
kzhuravl added a comment to D53222: AMDGPU: Add sram-ecc feature.

ping

Oct 24 2018, 12:59 PM
kzhuravl added a comment to D53558: Add gfx909 to GPU Arch.

Are changes for https://github.com/llvm-mirror/clang/blob/master/lib/Basic/Targets/AMDGPU.cpp#L137 in a separate review? (I could not locate it).

Oct 24 2018, 12:14 PM
kzhuravl committed rL345181: AMDGPU: Handle gfx909 in AMDGPUTargetInfo::initFeatureMap.
AMDGPU: Handle gfx909 in AMDGPUTargetInfo::initFeatureMap
Oct 24 2018, 12:10 PM
kzhuravl committed rC345181: AMDGPU: Handle gfx909 in AMDGPUTargetInfo::initFeatureMap.
AMDGPU: Handle gfx909 in AMDGPUTargetInfo::initFeatureMap
Oct 24 2018, 12:10 PM

Oct 23 2018

kzhuravl accepted D53418: [AMDGPU] Defined gfx909 Raven Ridge 2.

LGTM

Oct 23 2018, 9:42 AM
kzhuravl added a comment to D53558: Add gfx909 to GPU Arch.

Are changes for https://github.com/llvm-mirror/clang/blob/master/lib/Basic/Targets/AMDGPU.cpp#L137 in a separate review? (I could not locate it).

Oct 23 2018, 9:33 AM

Oct 22 2018

kzhuravl added a comment to D53418: [AMDGPU] Defined gfx909 Raven Ridge 2.

This also requires a clang change.

Oct 22 2018, 8:01 PM
kzhuravl updated the diff for D53222: AMDGPU: Add sram-ecc feature.

Fix elf flag value in docs.

Oct 22 2018, 1:36 PM
kzhuravl added a parent revision for D53526: AMDGPU: Switch some lld tests to v2: D53525: AMDGPU: Enable code object v3 by default.
Oct 22 2018, 1:33 PM
kzhuravl added a child revision for D53525: AMDGPU: Enable code object v3 by default: D53526: AMDGPU: Switch some lld tests to v2.
Oct 22 2018, 1:33 PM
kzhuravl created D53526: AMDGPU: Switch some lld tests to v2.
Oct 22 2018, 1:33 PM
kzhuravl added a comment to D53525: AMDGPU: Enable code object v3 by default.

I have currently switched some tests back to v2 and will update those gradually.

Oct 22 2018, 1:32 PM
kzhuravl created D53525: AMDGPU: Enable code object v3 by default.
Oct 22 2018, 1:31 PM
kzhuravl added a comment to D53223: AMDGPU: Add sram-ecc feature options.

ping

Oct 22 2018, 12:49 PM
kzhuravl added a comment to D53222: AMDGPU: Add sram-ecc feature.

ping

Oct 22 2018, 12:49 PM
kzhuravl added inline comments to D53445: [AMDGPU] Update code object metadata format documentation.
Oct 22 2018, 12:49 PM
kzhuravl added a comment to D53445: [AMDGPU] Update code object metadata format documentation.

In v2 asm, we had the following:

.option.machine_version_major
.option.machine_version_minor
.option.machine_version_stepping
Oct 22 2018, 12:44 PM

Oct 17 2018

kzhuravl committed rC344711: AMDGPU: Add options to enable/disable code object v3.
AMDGPU: Add options to enable/disable code object v3
Oct 17 2018, 2:41 PM
kzhuravl committed rL344711: AMDGPU: Add options to enable/disable code object v3.
AMDGPU: Add options to enable/disable code object v3
Oct 17 2018, 2:41 PM
kzhuravl closed D53386: AMDGPU: Add options to enable/disable code object v3.
Oct 17 2018, 2:41 PM
kzhuravl added a comment to D53386: AMDGPU: Add options to enable/disable code object v3.

Would it work with "clang -x assembler..." as well?

Oct 17 2018, 2:17 PM
kzhuravl created D53386: AMDGPU: Add options to enable/disable code object v3.
Oct 17 2018, 1:58 PM
kzhuravl abandoned D53351: AMDGPU/Docs: Correct note ids for metadata.

@scott.linder is working on a complete set of doc changes.

Oct 17 2018, 1:56 PM

Oct 16 2018

kzhuravl updated subscribers of D53351: AMDGPU/Docs: Correct note ids for metadata.
Oct 16 2018, 6:25 PM
kzhuravl created D53351: AMDGPU/Docs: Correct note ids for metadata.
Oct 16 2018, 6:25 PM
kzhuravl updated the diff for D53222: AMDGPU: Add sram-ecc feature.

Change to 0x200. Minor text updates as suggested by Nicolai.

Oct 16 2018, 6:24 PM
kzhuravl accepted D53325: Disable code object version 3 for HIP toolchain.

LGTM

Oct 16 2018, 9:37 AM

Oct 15 2018

kzhuravl committed rL344552: AMDGPU: Generate .amdgcn_target for object code v3.
AMDGPU: Generate .amdgcn_target for object code v3
Oct 15 2018, 1:39 PM
kzhuravl closed D53221: AMDGPU: Generate .amdgcn_target for object code v3.
Oct 15 2018, 1:39 PM

Oct 12 2018

kzhuravl added a child revision for D53222: AMDGPU: Add sram-ecc feature: D53223: AMDGPU: Add sram-ecc feature options.
Oct 12 2018, 2:50 PM
kzhuravl added a parent revision for D53223: AMDGPU: Add sram-ecc feature options: D53222: AMDGPU: Add sram-ecc feature.
Oct 12 2018, 2:50 PM
kzhuravl added a child revision for D53221: AMDGPU: Generate .amdgcn_target for object code v3: D53222: AMDGPU: Add sram-ecc feature.
Oct 12 2018, 2:50 PM
kzhuravl added a parent revision for D53222: AMDGPU: Add sram-ecc feature: D53221: AMDGPU: Generate .amdgcn_target for object code v3.
Oct 12 2018, 2:50 PM
kzhuravl created D53223: AMDGPU: Add sram-ecc feature options.
Oct 12 2018, 2:49 PM
kzhuravl created D53222: AMDGPU: Add sram-ecc feature.
Oct 12 2018, 2:49 PM
kzhuravl created D53221: AMDGPU: Generate .amdgcn_target for object code v3.
Oct 12 2018, 2:48 PM

Oct 4 2018

kzhuravl committed rL343813: AMDGPU: Rename isAmdCodeObjectV2 -> isAmdHsaOrMesa.
AMDGPU: Rename isAmdCodeObjectV2 -> isAmdHsaOrMesa
Oct 4 2018, 2:04 PM
kzhuravl closed D52897: AMDGPU: Rename isAmdCodeObjectV2 -> isAmdHsaOrMesa.
Oct 4 2018, 2:04 PM
kzhuravl created D52897: AMDGPU: Rename isAmdCodeObjectV2 -> isAmdHsaOrMesa.
Oct 4 2018, 11:26 AM

Oct 1 2018

kzhuravl accepted D52736: [AMDGPU] Fixed SIInstrInfo::getOpSize to handle subregs.

LGTM

Oct 1 2018, 10:56 AM

Sep 27 2018

kzhuravl committed rL343264: AMDGPU: Split HasExt into HasExtDPP/SDWA/SDWA9.
AMDGPU: Split HasExt into HasExtDPP/SDWA/SDWA9
Sep 27 2018, 1:50 PM
kzhuravl committed rL343259: AMDGPU: Split VOP2Inst into VOP2Inst_e32/e64/sdwa.
AMDGPU: Split VOP2Inst into VOP2Inst_e32/e64/sdwa
Sep 27 2018, 12:48 PM
kzhuravl committed rL343254: AMDGPU/NFC: Simplify VOP_MAC_F16/F32.
AMDGPU/NFC: Simplify VOP_MAC_F16/F32
Sep 27 2018, 12:25 PM
kzhuravl accepted D52577: [AMDGPU] Fold copy (copy vgpr).

LGTM

Sep 27 2018, 11:28 AM

Sep 25 2018

kzhuravl accepted D52522: [AMDGPU] Fix ds combine with subregs.

LGTM

Sep 25 2018, 4:14 PM

Sep 17 2018

kzhuravl accepted D52123: [AMDGPU] Initialize instruction itinerary from GCNSubtarget.

LGTM

Sep 17 2018, 7:58 AM

Sep 14 2018

kzhuravl committed rL342270: AMDGPU: Clear the bits before they are being set in program resource registers.
AMDGPU: Clear the bits before they are being set in program resource registers
Sep 14 2018, 1:03 PM

Sep 12 2018

kzhuravl committed rL342077: AMDGPU: Print all kernel descriptor directives (including the ones with default….
AMDGPU: Print all kernel descriptor directives (including the ones with default…
Sep 12 2018, 1:27 PM
kzhuravl closed D51954: AMDGPU: Print all kernel descriptor directives (including the ones with default values).
Sep 12 2018, 1:27 PM
kzhuravl committed rL342069: AMDGPU: Re-apply r341982 after fixing the layering issue.
AMDGPU: Re-apply r341982 after fixing the layering issue
Sep 12 2018, 11:52 AM

Sep 11 2018

kzhuravl created D51954: AMDGPU: Print all kernel descriptor directives (including the ones with default values).
Sep 11 2018, 3:27 PM
kzhuravl committed rL341982: AMDGPU: Move isa version and EF_AMDGPU_MACH_* determination.
AMDGPU: Move isa version and EF_AMDGPU_MACH_* determination
Sep 11 2018, 12:00 PM
kzhuravl closed D51890: AMDGPU: Move isa version and EF_AMDGPU_MACH_* determination into TargetParser..
Sep 11 2018, 12:00 PM

Sep 10 2018

kzhuravl created D51890: AMDGPU: Move isa version and EF_AMDGPU_MACH_* determination into TargetParser..
Sep 10 2018, 3:01 PM

Sep 7 2018

kzhuravl accepted D51790: AMDGPU: Don't abort on unknown addrspace argument.

LGTM

Sep 7 2018, 9:06 AM

Aug 16 2018

kzhuravl accepted D50843: AMDGPU: Correct errors in device table.

LGTM

Aug 16 2018, 1:04 PM

Aug 9 2018

kzhuravl accepted D50493: AMDGPU: Add another missing builtin.

LGTM

Aug 9 2018, 10:15 AM

Aug 8 2018

kzhuravl accepted D50376: AMDGPU: Fix enabling denormals by default on pre-VI targets.

LGTM

Aug 8 2018, 6:05 AM

Aug 6 2018

kzhuravl accepted D50320: AMDGPU: Add builtin for s_dcache_inv_vol.

LGTM

Aug 6 2018, 10:44 AM
kzhuravl accepted D50321: AMDGPU: Add builtin for s_dcache_wb.

LGTM

Aug 6 2018, 10:43 AM

Aug 3 2018

kzhuravl accepted D50265: AMDGPU: Add feature vi-insts.

LGTM

Aug 3 2018, 11:05 AM

Jul 31 2018

kzhuravl committed rC338471: AMDGPU: Add clamp bit to dot builtins.
AMDGPU: Add clamp bit to dot builtins
Jul 31 2018, 6:32 PM
kzhuravl committed rL338471: AMDGPU: Add clamp bit to dot builtins.
AMDGPU: Add clamp bit to dot builtins
Jul 31 2018, 6:32 PM