Page MenuHomePhabricator

kzhuravl (Konstantin Zhuravlyov)
User

Projects

User does not belong to any projects.

User Details

User Since
Feb 16 2016, 12:13 PM (258 w, 1 d)

Recent Activity

Tue, Jan 26

kzhuravl added a comment to D94585: [IndirectFunctions] Skip propagating attributes to address taken functions.

This change is causing infinite loop when compiling rocThrust and hipCUB. Can you take a look? Thanks.

FWIW, after testing rocThrust and hipCUB, it turned out that, it is not an infinite loop but a rise in compile-time which crossed timing threshold of the internal infra. The apps eventually compiled with 1-2% increment in the compile-time. I figured out the cause behind this. This patch makes two additional calls to Function::hasAddressTaken() and hasAddressTaken() is not optimal. Each time Function::hasAddressTaken() is called, it traverses over all uses of the function to deduce the required info. The calls made by this patch are itself in the loop which effectively made the suboptimality of Function::hasAddressTaken() conspicuous. In the new patch I am going to remove one call to Function::hasAddressTaken() which still preserves the intention of this patch. The new patch would basically be Diff3 of this patch.

We should think about optimizing Function::hasAddressTaken() by introducing a bool in Function class which would cache the result. In addition to this, Function::hasAddressTaken() can accept an additional bool parameter and based on its truthness, the function can return the cached value. The default value of parameter bool should be false to preserve today's behavior but client can set it to "true" if they want the cached value. Latter would be useful for this patch as the information is unlikely to change in a module (which is static naturally). This optimization would significantly improve the run-time of hasAddressTaken().

Tue, Jan 26, 10:37 PM · Restricted Project
kzhuravl accepted D85882: [AMDGPU] Update subtarget features for new target ID support.

LGTM

Tue, Jan 26, 11:24 AM · Restricted Project

Mon, Jan 25

kzhuravl requested review of D95391: AMDGPU: Add support for amdgpu-unsafe-fp-atomics attribute.
Mon, Jan 25, 1:47 PM · Restricted Project
kzhuravl added a comment to D94585: [IndirectFunctions] Skip propagating attributes to address taken functions.

This change is causing infinite loop when compiling rocThrust and hipCUB. Can you take a look? Thanks.

Mon, Jan 25, 1:13 PM · Restricted Project
kzhuravl added a reverting change for rGdd8ae42674b4: [IndirectFunctions] Skip propagating attributes to address taken functions: rG2cdb34efdac5: Revert "[IndirectFunctions] Skip propagating attributes to address taken….
Mon, Jan 25, 12:58 PM
kzhuravl committed rG2cdb34efdac5: Revert "[IndirectFunctions] Skip propagating attributes to address taken… (authored by kzhuravl).
Revert "[IndirectFunctions] Skip propagating attributes to address taken…
Mon, Jan 25, 12:58 PM
kzhuravl added a reverting change for D94585: [IndirectFunctions] Skip propagating attributes to address taken functions: rG2cdb34efdac5: Revert "[IndirectFunctions] Skip propagating attributes to address taken….
Mon, Jan 25, 12:58 PM · Restricted Project
kzhuravl closed D95389: AMDGPU: Revert "[IndirectFunctions] Skip propagating attributes to address taken functions".
Mon, Jan 25, 12:58 PM · Restricted Project
kzhuravl added a reverting change for rGdd8ae42674b4: [IndirectFunctions] Skip propagating attributes to address taken functions: D95389: AMDGPU: Revert "[IndirectFunctions] Skip propagating attributes to address taken functions".
Mon, Jan 25, 12:55 PM
kzhuravl requested review of D95389: AMDGPU: Revert "[IndirectFunctions] Skip propagating attributes to address taken functions".
Mon, Jan 25, 12:55 PM · Restricted Project
kzhuravl added a reverting change for D94585: [IndirectFunctions] Skip propagating attributes to address taken functions: D95389: AMDGPU: Revert "[IndirectFunctions] Skip propagating attributes to address taken functions".
Mon, Jan 25, 12:55 PM · Restricted Project

Wed, Jan 20

kzhuravl accepted D95018: [NFC][AMDGPU] Document target ID syntax for code object V2 to V3.

LGTM

Wed, Jan 20, 12:35 PM · Restricted Project

Mon, Jan 11

kzhuravl added a comment to D85882: [AMDGPU] Update subtarget features for new target ID support.

Any progress on this? Does it need another rebase?

Mon, Jan 11, 7:59 AM · Restricted Project

Dec 14 2020

kzhuravl accepted D93234: [NFC] Remove trailing whitespace in llvm/CMakeLists.txt.

lgtm

Dec 14 2020, 9:49 AM · Restricted Project

Dec 13 2020

kzhuravl accepted D93181: [NFC][AMDGPU] Reformat AMD GPU targets in cuda.cpp.

lgtm

Dec 13 2020, 9:49 AM · Restricted Project
kzhuravl accepted D93172: [NFC]{AMDGPU] Update AMDGPUUsage with AMD RDNA 2 reference.

lgtm

Dec 13 2020, 9:07 AM · Restricted Project

Dec 12 2020

kzhuravl accepted D93016: [NFC][AMDGPU] AMDGPUUsage updates.

lgtm

Dec 12 2020, 10:26 AM · Restricted Project
kzhuravl accepted D93017: [AMDGPU] Add missing targets to amdgpu-features.cl.

lgtm

Dec 12 2020, 10:22 AM · Restricted Project
kzhuravl accepted D93018: [AMDGPU] Add missing targets to target-invalid-cpu-note.c.

LGTM

Dec 12 2020, 10:21 AM · Restricted Project

Dec 4 2020

kzhuravl accepted D92708: [NFC][AMDGPU] AMDGPUUsage updates.

LGTM

Dec 4 2020, 5:18 PM · Restricted Project

Dec 1 2020

kzhuravl accepted D92441: Add CLangOffloadBundler documentation to Clang index.

LGTM

Dec 1 2020, 6:19 PM · Restricted Project
kzhuravl accepted D92434: [NFC][AMDGPU] AMDGPU code object V4 ABI documentation.

LGTM

Dec 1 2020, 3:36 PM · Restricted Project, Restricted Project

Nov 30 2020

kzhuravl accepted D91310: [AMDGPU] Add -mcode-object-version=n.

LGTM

Nov 30 2020, 8:49 AM · Restricted Project

Nov 10 2020

kzhuravl accepted D91202: [AMDGPU] Set default op_sel_hi on accvgpr read/write.

LGTM

Nov 10 2020, 12:49 PM · Restricted Project

Nov 5 2020

kzhuravl accepted D90886: [AMDGPU] Simplify amdgpu-macros.cl test. NFC..

LGTM

Nov 5 2020, 4:27 PM · Restricted Project
kzhuravl accepted D89168: [AMDGPU] Corrected declaration of VOPC instructions with SDWA addressing mode..

LGTM

Nov 5 2020, 9:14 AM · Restricted Project, Restricted Project

Oct 29 2020

kzhuravl accepted D90364: [AMDGPU] Update AMD GPU documentation.

LGTM

Oct 29 2020, 6:21 AM · Restricted Project, Restricted Project

Oct 21 2020

kzhuravl accepted D89880: [AMDGPU] Reorder SIMemoryLegalizer functions to be consistent.

LGTM

Oct 21 2020, 9:16 AM · Restricted Project

Oct 15 2020

kzhuravl accepted D89484: [AMDGPU][HIP] Switch default DWARF version to 5.

LGTM

Oct 15 2020, 12:41 PM · Restricted Project
kzhuravl accepted D89487: [AMDGPU] gfx1032 target.

LGTM

Oct 15 2020, 12:39 PM · Restricted Project, Restricted Project
kzhuravl committed rG67f189e93ce3: Make sure both cc1 and cc1as process -m[no-]code-object-v3 (authored by kzhuravl).
Make sure both cc1 and cc1as process -m[no-]code-object-v3
Oct 15 2020, 11:04 AM
kzhuravl closed D89478: AMDGPU: Make sure both cc1 and cc1as process -m[no-]code-object-v3.
Oct 15 2020, 11:03 AM · Restricted Project
kzhuravl updated the diff for D89478: AMDGPU: Make sure both cc1 and cc1as process -m[no-]code-object-v3.

Addressed review feedback

Oct 15 2020, 10:39 AM · Restricted Project
kzhuravl requested review of D89478: AMDGPU: Make sure both cc1 and cc1as process -m[no-]code-object-v3.
Oct 15 2020, 8:56 AM · Restricted Project

Oct 14 2020

kzhuravl committed rG3fdf3b153960: AMDGPU: Update AMDHSA code object version handling (authored by kzhuravl).
AMDGPU: Update AMDHSA code object version handling
Oct 14 2020, 10:05 AM
kzhuravl closed D89076: AMDGPU: Update AMDHSA code object version handling.
Oct 14 2020, 10:04 AM · Restricted Project, Restricted Project
kzhuravl committed rGefd84a5f5db2: Remove Combine.td.rej file (authored by kzhuravl).
Remove Combine.td.rej file
Oct 14 2020, 8:39 AM

Oct 13 2020

kzhuravl committed rGf218652a36bd: LLD/AMDGPU: Infer os abi based on input llvm bitcode (authored by kzhuravl).
LLD/AMDGPU: Infer os abi based on input llvm bitcode
Oct 13 2020, 9:21 AM
kzhuravl committed rGe2eaa914514c: AMDGPU: Remove -mamdgpu-debugger-abi option (authored by kzhuravl).
AMDGPU: Remove -mamdgpu-debugger-abi option
Oct 13 2020, 9:21 AM
kzhuravl closed D89042: LLD/AMDGPU: Infer os abi based on input llvm bitcode.
Oct 13 2020, 9:21 AM · Restricted Project
kzhuravl closed D89125: AMDGPU: Remove -mamdgpu-debugger-abi option.
Oct 13 2020, 9:20 AM · Restricted Project

Oct 9 2020

kzhuravl abandoned D80810: AMDGPU: Add URI definition to AMDGPUUsage.
Oct 9 2020, 3:04 PM · Restricted Project
kzhuravl requested review of D89125: AMDGPU: Remove -mamdgpu-debugger-abi option.
Oct 9 2020, 6:09 AM · Restricted Project
kzhuravl committed rGb581c5a42f05: Regenerate ClangCommandLineReference.rst (authored by kzhuravl).
Regenerate ClangCommandLineReference.rst
Oct 9 2020, 5:31 AM
kzhuravl closed D89091: Regenerate ClangCommandLineReference.rst.
Oct 9 2020, 5:30 AM · Restricted Project

Oct 8 2020

kzhuravl updated the diff for D89042: LLD/AMDGPU: Infer os abi based on input llvm bitcode.

Address review feedback:

  • Merge tests into one test file
Oct 8 2020, 8:02 PM · Restricted Project
kzhuravl added inline comments to D89076: AMDGPU: Update AMDHSA code object version handling.
Oct 8 2020, 8:00 PM · Restricted Project, Restricted Project
kzhuravl updated the diff for D89076: AMDGPU: Update AMDHSA code object version handling.

Address review feedback:

  • Regenerate docs in separate change
Oct 8 2020, 7:08 PM · Restricted Project, Restricted Project
kzhuravl requested review of D89091: Regenerate ClangCommandLineReference.rst.
Oct 8 2020, 7:01 PM · Restricted Project
kzhuravl added inline comments to D89076: AMDGPU: Update AMDHSA code object version handling.
Oct 8 2020, 5:38 PM · Restricted Project, Restricted Project
kzhuravl requested review of D89076: AMDGPU: Update AMDHSA code object version handling.
Oct 8 2020, 3:27 PM · Restricted Project, Restricted Project
kzhuravl requested review of D89042: LLD/AMDGPU: Infer os abi based on input llvm bitcode.
Oct 8 2020, 6:31 AM · Restricted Project

Sep 30 2020

kzhuravl accepted D88619: [AMDGPU] Reorganize VOP3P encoding.

LGTM

Sep 30 2020, 3:21 PM · Restricted Project

Sep 16 2020

kzhuravl accepted D87782: [AMDGPU] gfx1030 RT support.

LGTM

Sep 16 2020, 11:37 AM · Restricted Project

Sep 8 2020

kzhuravl accepted D80713: [AMDGPU] Support disassembly for AMDGPU kernel descriptors.

LGTM, thank you!

Sep 8 2020, 8:44 AM · Restricted Project

Sep 2 2020

kzhuravl added inline comments to D84822: Add documentation for target ID and ClangOffloadBundlerFormat.
Sep 2 2020, 11:31 AM · Restricted Project
kzhuravl added inline comments to D84822: Add documentation for target ID and ClangOffloadBundlerFormat.
Sep 2 2020, 11:21 AM · Restricted Project
kzhuravl requested changes to D84822: Add documentation for target ID and ClangOffloadBundlerFormat.
Sep 2 2020, 11:12 AM · Restricted Project

Aug 31 2020

kzhuravl requested changes to D86902: [AMDGPU] Correct documentation for default setting of sram-ecc.

sramecc is off by default for gfx906:
https://github.com/llvm/llvm-project/blob/master/llvm/test/CodeGen/AMDGPU/elf-header-flags-sram-ecc.ll#L5

Aug 31 2020, 7:34 PM · Restricted Project
kzhuravl accepted D86902: [AMDGPU] Correct documentation for default setting of sram-ecc.

Might want to mention that it is being corrected for gfx906 in the commit message

Aug 31 2020, 7:13 PM · Restricted Project

Aug 14 2020

kzhuravl accepted D85882: [AMDGPU] Update subtarget features for new target ID support.

LGTM

Aug 14 2020, 2:44 PM · Restricted Project
kzhuravl added inline comments to D85882: [AMDGPU] Update subtarget features for new target ID support.
Aug 14 2020, 7:34 AM · Restricted Project

Aug 11 2020

kzhuravl added inline comments to D81780: AMDGPU/AMDHSA: Implement new target ID support in AMDGPU backend.
Aug 11 2020, 8:11 AM · Restricted Project

Aug 6 2020

kzhuravl accepted D85476: [AMDGPU] Correct missing sram-ecc target feature for gfx906.

LGTM

Aug 6 2020, 3:11 PM · Restricted Project

Jul 30 2020

kzhuravl requested changes to D60146: AMDGPU: Don't use generation to determine encoding.

need to handle gfx10

Jul 30 2020, 9:57 AM

Jul 10 2020

kzhuravl accepted D82818: AMDGPU: Remove .value_type from kernel metadata.

lgtm

Jul 10 2020, 10:30 AM · Restricted Project
kzhuravl added inline comments to D82818: AMDGPU: Remove .value_type from kernel metadata.
Jul 10 2020, 7:58 AM · Restricted Project

Jul 9 2020

kzhuravl added a comment to D82818: AMDGPU: Remove .value_type from kernel metadata.

Do we also want to remove it from v2 metadata?

Jul 9 2020, 9:23 AM · Restricted Project

Jul 6 2020

Herald added a project to D83249: AMDGPU: Handle llvm.amdgcn.buffer.{load|store}.v2i16 intrinsics: Restricted Project.
Jul 6 2020, 12:10 PM · Restricted Project

Jun 20 2020

kzhuravl requested changes to D80713: [AMDGPU] Support disassembly for AMDGPU kernel descriptors.
Jun 20 2020, 3:08 AM · Restricted Project

Jun 18 2020

kzhuravl accepted D81974: [AMDGPU] Added new encoding to getMCOpcodeGen.

LGTM

Jun 18 2020, 9:45 AM · Restricted Project

Jun 16 2020

kzhuravl added a comment to D81780: AMDGPU/AMDHSA: Implement new target ID support in AMDGPU backend.

Changes to AMDGPUUsage are WIP.

Jun 16 2020, 12:05 PM · Restricted Project
kzhuravl added a comment to D81780: AMDGPU/AMDHSA: Implement new target ID support in AMDGPU backend.

lib/ObjectYAML/ELFYAML.cpp and tools/llvm-readobj/ELFDumper.cpp changes can be made to a separate patch. They are not tightly coupled with the MC/CodeGen changes. A group of binary format focused people will be interested to review that part.

Jun 16 2020, 12:05 PM · Restricted Project

Jun 15 2020

kzhuravl accepted D81886: [AMDGPU] Add gfx1030 target.

LGTM. This was reviewed downstream. Additional review feedback should be addressed in the follow up patch to simplify the merge downstream. Thank you.

Jun 15 2020, 4:35 PM · Restricted Project, Restricted Project

Jun 12 2020

kzhuravl added a comment to D81780: AMDGPU/AMDHSA: Implement new target ID support in AMDGPU backend.

Changes to AMDGPUUsage are WIP.

Jun 12 2020, 6:48 PM · Restricted Project
kzhuravl added a reviewer for D81780: AMDGPU/AMDHSA: Implement new target ID support in AMDGPU backend: kerbowa.
Jun 12 2020, 6:48 PM · Restricted Project
kzhuravl created D81780: AMDGPU/AMDHSA: Implement new target ID support in AMDGPU backend.
Jun 12 2020, 6:48 PM · Restricted Project

May 29 2020

kzhuravl accepted D80810: AMDGPU: Add URI definition to AMDGPUUsage.

lgtm

May 29 2020, 4:57 PM · Restricted Project
kzhuravl created D80810: AMDGPU: Add URI definition to AMDGPUUsage.
May 29 2020, 9:12 AM · Restricted Project

May 21 2020

kzhuravl accepted D80407: [AMDGPU] Correct formatting typos in documentation.

LGTM

May 21 2020, 3:11 PM · Restricted Project

Feb 27 2020

kzhuravl committed rG3e0e41508288: merge master into amd-stg-open (authored by kzhuravl).
merge master into amd-stg-open
Feb 27 2020, 9:44 AM
kzhuravl committed rG0923738d2420: merge master into amd-stg-open (authored by kzhuravl).
merge master into amd-stg-open
Feb 27 2020, 9:44 AM
kzhuravl committed rG3c23b6b76016: merge master into amd-stg-open (authored by kzhuravl).
merge master into amd-stg-open
Feb 27 2020, 9:40 AM
kzhuravl committed rG987b3e13b91b: merge master into amd-stg-open (authored by kzhuravl).
merge master into amd-stg-open
Feb 27 2020, 9:39 AM
kzhuravl committed rG8807382d40b1: merge master into amd-stg-open (authored by kzhuravl).
merge master into amd-stg-open
Feb 27 2020, 9:39 AM
kzhuravl committed rG587301f47f19: merge master into amd-stg-open (authored by kzhuravl).
merge master into amd-stg-open
Feb 27 2020, 9:39 AM
kzhuravl committed rGb69d67e9338c: merge master into amd-stg-open (authored by kzhuravl).
merge master into amd-stg-open
Feb 27 2020, 9:34 AM
kzhuravl committed rGc693d7697df2: merge master into amd-stg-open (authored by kzhuravl).
merge master into amd-stg-open
Feb 27 2020, 9:34 AM
kzhuravl committed rG0d36bc1ae8b1: merge master into amd-stg-open (authored by kzhuravl).
merge master into amd-stg-open
Feb 27 2020, 9:33 AM
kzhuravl committed rG39cad26ba9a8: merge master into amd-stg-open (authored by kzhuravl).
merge master into amd-stg-open
Feb 27 2020, 9:30 AM
kzhuravl committed rGfe1fe6ddb897: merge master into amd-stg-open (authored by kzhuravl).
merge master into amd-stg-open
Feb 27 2020, 9:30 AM
kzhuravl committed rGe77721bc1ae4: merge master into amd-stg-open (authored by kzhuravl).
merge master into amd-stg-open
Feb 27 2020, 9:29 AM
kzhuravl committed rG902023012f75: merge master into amd-stg-open (authored by kzhuravl).
merge master into amd-stg-open
Feb 27 2020, 9:28 AM
kzhuravl committed rG0f407f9ad830: merge master into amd-stg-open (authored by kzhuravl).
merge master into amd-stg-open
Feb 27 2020, 9:28 AM
kzhuravl committed rG2bab8f2a5a7b: AMDGCN: Skip -fgnuc-version version for amdgcn (temporary workaround) (authored by kzhuravl).
AMDGCN: Skip -fgnuc-version version for amdgcn (temporary workaround)
Feb 27 2020, 9:28 AM
kzhuravl committed rG6ca354baea3e: merge master into amd-stg-open (authored by msearles).
merge master into amd-stg-open
Feb 27 2020, 9:26 AM
kzhuravl committed rG0902071e00f7: SWDEV-179954 - OpenCL/LC - Merge branch amd-master into amd-common (authored by kzhuravl).
SWDEV-179954 - OpenCL/LC - Merge branch amd-master into amd-common
Feb 27 2020, 9:03 AM
kzhuravl committed rG4455f63233bf: SWDEV-179954 - OpenCL/LC - Merge branch amd-master into amd-common (authored by kzhuravl).
SWDEV-179954 - OpenCL/LC - Merge branch amd-master into amd-common
Feb 27 2020, 8:49 AM
kzhuravl committed rGcec93631936a: SWDEV-179954 - OpenCL/LC - Merge branch amd-master into amd-common (authored by kzhuravl).
SWDEV-179954 - OpenCL/LC - Merge branch amd-master into amd-common
Feb 27 2020, 8:42 AM
kzhuravl committed rGb576a306f06a: Revert "Revert "[AMDGPU] Use three- and five-dword result type in image ops"" (authored by kzhuravl).
Revert "Revert "[AMDGPU] Use three- and five-dword result type in image ops""
Feb 27 2020, 8:41 AM