mariusz-sikora-at-amd wrote:
What is the plan for atomic_{flat/ds/global}_bf16 builtins ? Right now they are
accepting <2 x i16> instead of <2 x bfloat>. Do we want to create new builtins
or we want to override them to accept both <2 x i16> and <2 x bfloat> ?
https://github.com/llvm/llvm-proj
@@ -27,34 +27,23 @@ main_body:
ret float %out0
}
-define amdgpu_ps float @atomic_pk_add_bf16_1d_v2(<8 x i32> inreg %rsrc, <2 x
i16> %data, i32 %s) {
+define amdgpu_ps float @atomic_pk_add_bf16_1d_v2(<8 x i32> inreg %rsrc, <2 x
bfloat> %data, i32 %s) {
; GFX12-LABEL: atomi
@@ -362,24 +358,34 @@ define amdgpu_ps void
@struct_buffer_atomic_add_v2f16_noret(<2 x half> %val, <4
ret void
}
-define amdgpu_ps float @struct_buffer_atomic_add_v2bf16_ret(<2 x i16> %val, <4
x i32> inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
+define amd
@@ -0,0 +1,92 @@
+// RUN: %clang_cc1 -O0 -cl-std=CL2.0 -triple amdgcn-amd-amdhsa -target-cpu
gfx1200 \
+// RUN: %s -S -emit-llvm -o - | FileCheck %s
+
+// RUN: %clang_cc1 -O0 -cl-std=CL2.0 -triple amdgcn-amd-amdhsa -target-cpu
gfx1200 \
+// RUN: -S -o - %s | FileCheck -check
