[AMDGPU] Disable sub-dword scralar loads IR widening

These will be widened in the DAG. In the meanwhile early
widening prevents otherwise possible vectorization of
such loads.

Differential Revision: https://reviews.llvm.org/D77835

diff --git a/llvm/lib/Target/AMDGPU/AMDGPUCodeGenPrepare.cpp b/llvm/lib/Target/AMDGPU/AMDGPUCodeGenPrepare.cpp
index da17d94f3265b5df88a1f17f0edf9c791558bcfb..e8c32cb346e8992d6e4c286b78a41f1f28f57191 100644 (file)
--- a/llvm/lib/Target/AMDGPU/AMDGPUCodeGenPrepare.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPUCodeGenPrepare.cpp
@@ -57,7 +57,7 @@ static cl::opt<bool> WidenLoads(
   "amdgpu-codegenprepare-widen-constant-loads",
   cl::desc("Widen sub-dword constant address space loads in AMDGPUCodeGenPrepare"),
   cl::ReallyHidden,
-  cl::init(true));
+  cl::init(false));
 
 static cl::opt<bool> UseMul24Intrin(
   "amdgpu-codegenprepare-mul24",

diff --git a/llvm/test/CodeGen/AMDGPU/vectorize-loads.ll b/llvm/test/CodeGen/AMDGPU/vectorize-loads.ll
new file mode 100644 (file)
index 0000000..e3126d1
--- /dev/null
+++ b/llvm/test/CodeGen/AMDGPU/vectorize-loads.ll
@@ -0,0 +1,27 @@
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s
+
+; GCN-LABEL: {{^}}load_idx_idy:
+; GCN-NOT: global_load
+; GCN: s_load_dword [[ID_XY:s[0-9]+]], s[4:5], 0x4
+; GCN-NOT: global_load
+; GCN: s_lshr_b32 [[ID_Y:s[0-9]+]], [[ID_XY]], 16
+; GCN: s_add_i32 [[ID_SUM:s[0-9]+]], [[ID_Y]], [[ID_XY]]
+; GCN: s_and_b32 s{{[0-9]+}}, [[ID_SUM]], 0xffff
+define protected amdgpu_kernel void @load_idx_idy(i32 addrspace(1)* %out) {
+entry:
+  %disp = tail call align 4 dereferenceable(64) i8 addrspace(4)* @llvm.amdgcn.dispatch.ptr()
+  %gep_x = getelementptr i8, i8 addrspace(4)* %disp, i64 4
+  %gep_x.cast = bitcast i8 addrspace(4)* %gep_x to i16 addrspace(4)*
+  %id_x = load i16, i16 addrspace(4)* %gep_x.cast, align 4, !invariant.load !0 ; load workgroup size x
+  %gep_y = getelementptr i8, i8 addrspace(4)* %disp, i64 6
+  %gep_y.cast = bitcast i8 addrspace(4)* %gep_y to i16 addrspace(4)*
+  %id_y = load i16, i16 addrspace(4)* %gep_y.cast, align 2, !invariant.load !0 ; load workgroup size y
+  %add = add nuw nsw i16 %id_y, %id_x
+  %conv = zext i16 %add to i32
+  store i32 %conv, i32 addrspace(1)* %out, align 4
+  ret void
+}
+
+declare i8 addrspace(4)* @llvm.amdgcn.dispatch.ptr()
+
+!0 = !{!0}

diff --git a/llvm/test/CodeGen/AMDGPU/widen_extending_scalar_loads.ll b/llvm/test/CodeGen/AMDGPU/widen_extending_scalar_loads.ll
index 6c1f9239f2abc3061227a104da2bde60b2d72531..1d6388bc1d2114052d2320a49e3d76fbdd9f9efc 100644 (file)
--- a/llvm/test/CodeGen/AMDGPU/widen_extending_scalar_loads.ll
+++ b/llvm/test/CodeGen/AMDGPU/widen_extending_scalar_loads.ll
@@ -1,4 +1,4 @@
-; RUN: opt -S -mtriple=amdgcn-- -amdgpu-codegenprepare < %s | FileCheck -check-prefix=OPT %s
+; RUN: opt -S -mtriple=amdgcn-- -amdgpu-codegenprepare -amdgpu-codegenprepare-widen-constant-loads < %s | FileCheck -check-prefix=OPT %s
 
 declare i8 addrspace(4)* @llvm.amdgcn.dispatch.ptr() #0