fmax.
#95592

Merged

AMDGPU: Start selecting flat/global atomicrmw fmin/fmax. #95592

arsenm merged 3 commits into main from users/arsenm/amdgpu-select-flat-global-atomicrmw-fmin-fmax

arsenm added backend:AMDGPU

arsenm requested a review from

kosarev 1 year ago

arsenm requested a review from

mariusz-sikora-at-amd 1 year ago

arsenm requested a review from

piotrAMD 1 year ago

arsenm requested a review from

rampitec 1 year ago

arsenm requested a review from

rovka 1 year ago

arsenm requested a review from

shiltian 1 year ago

arsenm requested a review from

Sisyph 1 year ago

arsenm marked this pull request as ready for review 1 year ago

llvmbot added llvm:transforms

krzysz00 commented on 2024-06-14

arsenm force pushed from 7396f878 to 5e389eea 1 year ago

arsenm force pushed from 5e389eea to f661aa1d 1 year ago

arsenm changed the base branch from users/arsenm/amdgpu-flat-buffer-atomic-fmin-fmax-pseudos to users/arsenm/amdgpu-select-atomicrmw-fmin-fmax-local 1 year ago

arsenm force pushed from f661aa1d to b85a15ae 1 year ago

shiltian commented on 2024-06-18

shiltian approved these changes on 2024-06-18

Base automatically changed from users/arsenm/amdgpu-select-atomicrmw-fmin-fmax-local to main 1 year ago

arsenm force pushed from b85a15ae to 230ddd26 1 year ago

arsenm force pushed from 230ddd26 to 5853be59 1 year ago

arsenm force pushed from 5853be59 to 58b72f3a 1 year ago

llvmbot added clang

arsenm force pushed from 58b72f3a to d61d35a7 1 year ago

AMDGPU: Start selecting flat/global atomicrmw fmin/fmax.

549b5fc0

fix clang test, depends unsafe attr handling

67137490

Handle amdgpu-unsafe-fp-atomics attr

ff38a3c2

arsenm force pushed from d61d35a7 to ff38a3c2 1 year ago

arsenm merged a440a96e into main 1 year ago

arsenm deleted the users/arsenm/amdgpu-select-flat-global-atomicrmw-fmin-fmax branch 1 year ago

Reviewers

shiltian

krzysz00

kosarev

mariusz-sikora-at-amd

piotrAMD

rampitec

rovka

Sisyph

Assignees

No one assigned

Labels

clang backend:AMDGPU llvm:transforms

Milestone

No milestone

llvm-project AMDGPU: Start selecting flat/global atomicrmw fmin/fmax. #95592 Merged

AMDGPU: Start selecting flat/global atomicrmw fmin/fmax. #95592

llvm-project
AMDGPU: Start selecting flat/global atomicrmw fmin/fmax.
#95592

Merged