Skip to content

Copy of Fix overflows in div_up for sub-maximal arguments#4712

Draft
asimonov1 wants to merge 8 commits intomainfrom
asimonov/fix_overflows_in_divup_for_submax_args_2
Draft

Copy of Fix overflows in div_up for sub-maximal arguments#4712
asimonov1 wants to merge 8 commits intomainfrom
asimonov/fix_overflows_in_divup_for_submax_args_2

Conversation

@asimonov1
Copy link
Contributor

@asimonov1 asimonov1 commented Feb 20, 2026

No description provided.

@github-actions github-actions bot added platform:cpu-x64 Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64 platform:cpu-aarch64 Codeowner: @oneapi-src/onednn-cpu-aarch64 platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel platform:cpu-rv64 RISC-V platform:gpu-generic Codeowner: @oneapi-src/onednn-gpu-generic component:common labels Feb 20, 2026
@asimonov1 asimonov1 force-pushed the asimonov/fix_overflows_in_divup_for_submax_args_2 branch from deff641 to 6721488 Compare February 20, 2026 13:37
@asimonov1
Copy link
Contributor Author

asimonov1 commented Feb 20, 2026

make test
set test_scope=NIGHTLY

@asimonov1 asimonov1 force-pushed the asimonov/fix_overflows_in_divup_for_submax_args_2 branch from 6721488 to fde9e4d Compare February 20, 2026 23:46
@asimonov1
Copy link
Contributor Author

asimonov1 commented Feb 21, 2026

make test
set test_scope=NIGHTLY

@asimonov1 asimonov1 force-pushed the asimonov/fix_overflows_in_divup_for_submax_args_2 branch from fde9e4d to cb5d923 Compare February 22, 2026 13:47
@asimonov1
Copy link
Contributor Author

asimonov1 commented Feb 22, 2026

make test
set test_scope=NIGHTLY

@asimonov1 asimonov1 force-pushed the asimonov/fix_overflows_in_divup_for_submax_args_2 branch from cb5d923 to 1a4c133 Compare February 22, 2026 18:40
@asimonov1
Copy link
Contributor Author

asimonov1 commented Feb 22, 2026

make test
set test_scope=NIGHTLY

@asimonov1 asimonov1 force-pushed the asimonov/fix_overflows_in_divup_for_submax_args_2 branch from 1a4c133 to f4d14a1 Compare February 23, 2026 21:24
@asimonov1
Copy link
Contributor Author

asimonov1 commented Feb 23, 2026

make test
set test_scope=NIGHTLY
enable arch_cpu_adl
enable arch_cpu_clx
enable arch_cpu_dmr
enable arch_cpu_gnr
enable arch_cpu_hsw
enable arch_cpu_nhm
enable arch_cpu_nvl
enable arch_cpu_skx
enable arch_cpu_snb
enable arch_cpu_spr
enable arch_cpu_srf
enable arch_gpu_xe-hpc
enable arch_gpu_xe-hpg-atsm
enable arch_gpu_xe-hpg-dg2
enable arch_gpu_xe-lp
enable arch_gpu_xe-lpg
enable arch_gpu_xe-lpg+
enable arch_gpu_xe2-hpg-bmg
enable arch_gpu_xe2-lpg
enable arch_gpu_xe3-lpg
enable arch_gpu_xe3p-fs1

@asimonov1 asimonov1 force-pushed the asimonov/fix_overflows_in_divup_for_submax_args_2 branch 7 times, most recently from 190d205 to 5abfbf0 Compare February 25, 2026 17:12
asimonov1 added 7 commits March 5, 2026 05:35
@asimonov1
common: fix overflows for sub-maximal values
61904ae
@asimonov1
common: use div_up for integral types only
5049f98 
Only integral arguments are allowed for div_up function.
The second type can also be an enum, since it is naturally converted
to integral types (mostly int).
div_up returns 0 if the 1st argument is 0 or negative.
@asimonov1
gpu: update div_up implementation
3a18e36
@asimonov1
common: add assert to div_up
642c9e5
@asimonov1
cpu: do not use div_up(a,b) with a<0
c6a6d7a
@asimonov1
cpu: matmul: do not use div_up(a,b) with a<0
2df39b3
@asimonov1
cpu: matmul: do not use div_up(a,b) with a<0
048f1fe
@asimonov1 asimonov1 force-pushed the asimonov/fix_overflows_in_divup_for_submax_args_2 branch from e389957 to 0ceda15 Compare March 5, 2026 21:09
@asimonov1 asimonov1 changed the title Do not call div_up(a,b) with a<0 Copy of Fix overflows in div_up for sub-maximal arguments Mar 5, 2026
@asimonov1
Copy link
Contributor Author

asimonov1 commented Mar 5, 2026

make test
enable build_mode_cpp20
enable compiler_icx
enable compiler_icx-previous
enable compiler_icx-oss
enable compiler_gnu8
enable compiler_gnu9
enable compiler_gnu10
enable compiler_gnu11
enable compiler_gnu12
enable compiler_gnu13
enable compiler_clang11
enable compiler_clang14
enable compiler_clang16
enable compiler_clang19
enable compiler_vs2022

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

component:common platform:cpu-aarch64 Codeowner: @oneapi-src/onednn-cpu-aarch64 platform:cpu-rv64 RISC-V platform:cpu-x64 Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64 platform:gpu-generic Codeowner: @oneapi-src/onednn-gpu-generic platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

1 participant

@asimonov1