Fraser Cormack 5c411b3c0b

[libclc] Use elementwise ctlz/cttz builtins for CLC clz/ctz (#154535) ...

Using the elementwise builtin optimizes the vector case; instead of
scalarizing we can compile directly to the vector intrinsics.

2025-08-21 09:32:03 +01:00

..

amdgcn

[libclc] Implement __clc_get_local_size/__clc_get_max_sub_group_size for amdgcn (#153785)

2025-08-19 07:51:17 +08:00

amdgpu

[NFC][libclc] add missing __CLC_ prefix all internal macros (#153523)

2025-08-18 07:21:04 +08:00

clspv

[NFC][libclc] add missing __CLC_ prefix all internal macros (#153523)

2025-08-18 07:21:04 +08:00

generic

[libclc] Use elementwise ctlz/cttz builtins for CLC clz/ctz (#154535)

2025-08-21 09:32:03 +01:00

ptx-nvidiacl

[NFC][libclc] add missing __CLC_ prefix all internal macros (#153523)

2025-08-18 07:21:04 +08:00

r600

[NFC][libclc] add missing __CLC_ prefix all internal macros (#153523)

2025-08-18 07:21:04 +08:00

spirv

[NFC][libclc] add missing __CLC_ prefix all internal macros (#153523)

2025-08-18 07:21:04 +08:00