AMDGPU: enable 128-bit for local addr space under an option
Author: Samuel Pitoiset
ds_read_b128 and ds_write_b128 have been recently enabled
under the amdgpu-ds128 option because the performance benefit
is unclear.
Though, using 128-bit loads/stores for the local address space
appears to introduce regressions in tessellation shaders. Not
sure what is broken, but as ds_read_b128/ds_write_b128 are not
enabled by default, just introduce a global option and enable
128-bit only if requested (until it's fixed/used correctly).
v2: - fix regressions in merge-stores.ll and multiple_tails.ll
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105464
llvm-svn: 329764
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h b/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h
index e3455d3..cd08026 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h
+++ b/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h
@@ -133,6 +133,7 @@
bool EnableLoadStoreOpt;
bool EnableUnsafeDSOffsetFolding;
bool EnableSIScheduler;
+ bool EnableDS128;
bool DumpCode;
// Subtarget statically properties set by tablegen
@@ -412,8 +413,8 @@
/// \returns If target supports ds_read/write_b128 and user enables generation
/// of ds_read/write_b128.
- bool useDS128(bool UserEnable) const {
- return CIInsts && UserEnable;
+ bool useDS128() const {
+ return CIInsts && EnableDS128;
}
/// \returns If MUBUF instructions always perform range checking, even for