Skip to content

Commit 14e5b26

Browse files
author
Wang, Jun
authored
Merge pull request llvm#395 from AMD-Lightning-Internal/amd/dev/junwang5/SWDEV-443122
[AMDGPU] Update AMDGPUUsage.rst to document two intrinsics (llvm#123816)
2 parents 37c96f1 + deb9eea commit 14e5b26

File tree

1 file changed

+13
-0
lines changed

1 file changed

+13
-0
lines changed

llvm/docs/AMDGPUUsage.rst

Lines changed: 13 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1414,6 +1414,19 @@ The AMDGPU backend implements the following LLVM IR intrinsics.
14141414
Returns a pair for the swapped registers. The first element of the return
14151415
corresponds to the swapped element of the first argument.
14161416

1417+
llvm.amdgcn.mov.dpp The llvm.amdgcn.mov.dpp.`<type>` intrinsic represents the mov.dpp operation in AMDGPU.
1418+
This operation is being deprecated and can be replaced with llvm.amdgcn.update.dpp.
1419+
1420+
llvm.amdgcn.update.dpp The llvm.amdgcn.update.dpp.`<type>` intrinsic represents the update.dpp operation in AMDGPU.
1421+
It takes an old value, a source operand, a DPP control operand, a row mask, a bank mask, and a bound control.
1422+
Various data types are supported, including, bf16, f16, f32, f64, i16, i32, i64, p0, p3, p5, v2f16, v2f32, v2i16, v2i32, v2p0, v3i32, v4i32, v8f16.
1423+
This operation is equivalent to a sequence of v_mov_b32 operations.
1424+
It is preferred over llvm.amdgcn.mov.dpp.`<type>` for future use.
1425+
`llvm.amdgcn.update.dpp.<type> <old> <src> <dpp_ctrl> <row_mask> <bank_mask> <bound_ctrl>`
1426+
Should be equivalent to:
1427+
- `v_mov_b32 <dest> <old>`
1428+
- `v_mov_b32 <dest> <src> <dpp_ctrl> <row_mask> <bank_mask> <bound_ctrl>`
1429+
14171430
============================================== ==========================================================
14181431

14191432
.. TODO::

0 commit comments

Comments
 (0)