Enable FeatureFlatForGlobal on Volcanic Islands
This switches to the workaround that HSA defaults to
for the mesa path.
This should be applied to the 4.0 branch.
Patch by Vedran Miletić <vedran@miletic.net>
llvm-svn: 292982
diff --git a/llvm/lib/Target/AMDGPU/AMDGPU.td b/llvm/lib/Target/AMDGPU/AMDGPU.td
index 0c5f9e7..371042d 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPU.td
+++ b/llvm/lib/Target/AMDGPU/AMDGPU.td
@@ -305,10 +305,23 @@
"Enable SI Machine Scheduler"
>;
+def FeatureNoAddr64 : SubtargetFeature<"mubuf-no-addr64",
+ "NoAddr64",
+ "true",
+ "MUBUF instructions have addr64 bit"
+>;
+
+// Unless +-flat-for-global is specified, turn on FlatForGlobal for
+// all OS-es on VI and newer hardware to avoid assertion failures due
+// to missing ADDR64 variants of MUBUF instructions.
+// FIXME: moveToVALU should be able to handle converting addr64 MUBUF
+// instructions.
+
def FeatureFlatForGlobal : SubtargetFeature<"flat-for-global",
"FlatForGlobal",
"true",
- "Force to generate flat instruction for global"
+ "Force to generate flat instruction for global",
+ [FeatureNoAddr64]
>;
// Dummy feature used to disable assembler instructions.
@@ -361,7 +374,7 @@
FeatureGCN3Encoding, FeatureCIInsts, Feature16BitInsts,
FeatureSMemRealTime, FeatureVGPRIndexMode, FeatureMovrel,
FeatureScalarStores, FeatureInv2PiInlineImm, FeatureSDWA,
- FeatureDPP
+ FeatureDPP, FeatureNoAddr64
]
>;
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.cpp b/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.cpp
index 4848b3b..5064d9e 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.cpp
@@ -82,6 +82,7 @@
FP64FP16Denormals(false),
FPExceptions(false),
FlatForGlobal(false),
+ NoAddr64(false),
UnalignedScratchAccess(false),
UnalignedBufferAccess(false),
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h b/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h
index c8414f9..bc43a68 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h
+++ b/llvm/lib/Target/AMDGPU/AMDGPUSubtarget.h
@@ -85,6 +85,7 @@
bool FP64FP16Denormals;
bool FPExceptions;
bool FlatForGlobal;
+ bool NoAddr64;
bool UnalignedScratchAccess;
bool UnalignedBufferAccess;
bool EnableXNACK;