llvm
[SYCL] Specialize atomic fetch_add for floating point types
#2765

Merged

[SYCL] Specialize atomic fetch_add for floating point types #2765

bader merged 13 commits into intel:sycl from AGindinson:private/agindins/atomic-fadd-draft

AGindinson force pushed 5 years ago

AlexeySotkin requested a review from

AlexeySotkin 5 years ago

AlexeySotkin commented on 2020-11-26

AlexeySachkov requested a review from

Pennycook 5 years ago

AGindinson force pushed 5 years ago

Pennycook dismissed these changes on 2020-11-30

AGindinson dismissed their stale review 5 years ago

AGindinson force pushed 5 years ago

AGindinson changed the title ~~[DRAFT][SYCL] Support atomic float add in headers & SPIR-V~~ [SYCL] Specialize atomic fetch_add for floating point types 5 years ago

AGindinson force pushed 5 years ago

AlexeySotkin commented on 2020-12-09

AGindinson marked this pull request as ready for review 5 years ago

AGindinson requested a review 5 years ago

AGindinson requested a review from

v-klochkov 5 years ago

AGindinson force pushed 5 years ago

AGindinson requested a review from

Pennycook 5 years ago

AGindinson requested a review from

AlexeySachkov 5 years ago

AGindinson requested a review from

AlexeySotkin 5 years ago

Pennycook requested changes on 2021-01-13

AGindinson force pushed 5 years ago

Pennycook commented on 2021-01-13

AGindinson requested a review from

Pennycook 5 years ago

Pennycook dismissed these changes on 2021-01-13

v-klochkov commented on 2021-01-13

AGindinson dismissed their stale review 5 years ago

AGindinson requested a review from

v-klochkov 5 years ago

AGindinson requested a review from

Pennycook 5 years ago

AGindinson force pushed 5 years ago

[SYCL] Specialize atomic fetch_add for floating point types

7dce90a7

Provide a general macro for float atomics' emulation

28a3a1b6

[SYCL][Test] Configure atomic float tests for CUDA/CPU

d6a52081

Address review comments: fetch_sub in atomic.hpp, RUN lines' cleanup

3beb65a0

Fix the negation subject in the fetch_sub implementation

a9d8d88e

Remove the "emulated" atomics usage for the HOST tests

1e2ff470

Mark the tests unsupported on CUDA

73b5b5d4

Edit the UNSUPPORTED placement & content

da6edc50

Update tests to check the IR for spinlock emulation behavior

212aa4dd

Switch to using the emulation by default

0b178bc4

Revert changes in atomic.hpp - the new implementation can't be used t…

342a1796

Remove double underscore from the macro name

be711257

AGindinson force pushed to be711257 5 years ago

romanovvlad dismissed these changes on 2021-01-21

Switch to XFAIL for CUDA

906c4136

AGindinson dismissed their stale review via 906c4136 5 years ago

AGindinson requested a review from

romanovvlad 5 years ago

AlexeySachkov approved these changes on 2021-01-21

AlexeySotkin approved these changes on 2021-01-21

romanovvlad approved these changes on 2021-01-21

Pennycook approved these changes on 2021-01-21

bader merged 37a9a2a9 into sycl 5 years ago

AGindinson deleted the private/agindins/atomic-fadd-draft branch 4 years ago

Reviewers

Pennycook

romanovvlad

AlexeySotkin

AlexeySachkov

v-klochkov

Assignees

No one assigned

Labels

None yet

Milestone

No milestone

llvm [SYCL] Specialize atomic fetch_add for floating point types #2765 Merged

[SYCL] Specialize atomic fetch_add for floating point types #2765

llvm
[SYCL] Specialize atomic fetch_add for floating point types
#2765

Merged