[llvm] r284397 - AMDGPU/SI: Fix LowerParameter() for i16 arguments

[Tom Stellard via llvm-commits]

Author: tstellar
Date: Mon Oct 17 11:21:45 2016
New Revision: 284397

URL: http://llvm.org/viewvc/llvm-project?rev=284397&view=rev
Log:
AMDGPU/SI: Fix LowerParameter() for i16 arguments

Summary:
If we are loading an i16 value from a 32-bit memory location, then
we need to be able to truncate the loaded value to i16.

Reviewers: arsenm

Subscribers: kzhuravl, wdng, nhaehnle, yaxunl, tony-tye, llvm-commits

Differential Revision: https://reviews.llvm.org/D25198

Modified:
    llvm/trunk/lib/Target/AMDGPU/SIISelLowering.cpp
    llvm/trunk/test/CodeGen/AMDGPU/merge-stores.ll

Modified: llvm/trunk/lib/Target/AMDGPU/SIISelLowering.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AMDGPU/SIISelLowering.cpp?rev=284397&r1=284396&r2=284397&view=diff
==============================================================================
--- llvm/trunk/lib/Target/AMDGPU/SIISelLowering.cpp (original)
+++ llvm/trunk/lib/Target/AMDGPU/SIISelLowering.cpp Mon Oct 17 11:21:45 2016
@@ -587,23 +587,31 @@ SDValue SITargetLowering::LowerParameter
                                          unsigned Offset, bool Signed) const {
   const DataLayout &DL = DAG.getDataLayout();
   Type *Ty = VT.getTypeForEVT(*DAG.getContext());
-  MVT PtrVT = getPointerTy(DL, AMDGPUAS::CONSTANT_ADDRESS);
   PointerType *PtrTy = PointerType::get(Ty, AMDGPUAS::CONSTANT_ADDRESS);
-  SDValue PtrOffset = DAG.getUNDEF(PtrVT);
   MachinePointerInfo PtrInfo(UndefValue::get(PtrTy));
 
   unsigned Align = DL.getABITypeAlignment(Ty);
 
-  ISD::LoadExtType ExtTy = Signed ? ISD::SEXTLOAD : ISD::ZEXTLOAD;
+  SDValue Ptr = LowerParameterPtr(DAG, SL, Chain, Offset);
+  SDValue Load = DAG.getLoad(MemVT, SL, Chain, Ptr, PtrInfo, Align,
+                             MachineMemOperand::MONonTemporal |
+                             MachineMemOperand::MODereferenceable |
+                             MachineMemOperand::MOInvariant);
+
+  SDValue Val;
   if (MemVT.isFloatingPoint())
-    ExtTy = ISD::EXTLOAD;
+    Val = DAG.getNode(ISD::FP_EXTEND, SL, VT, Load);
+  else if (Signed)
+    Val = DAG.getSExtOrTrunc(Load, SL, VT);
+  else
+    Val = DAG.getZExtOrTrunc(Load, SL, VT);
 
-  SDValue Ptr = LowerParameterPtr(DAG, SL, Chain, Offset);
-  return DAG.getLoad(ISD::UNINDEXED, ExtTy, VT, SL, Chain, Ptr, PtrOffset,
-                     PtrInfo, MemVT, Align,
-                     MachineMemOperand::MONonTemporal |
-                         MachineMemOperand::MODereferenceable |
-                         MachineMemOperand::MOInvariant);
+  SDValue Ops[] = {
+    Val,
+    Load.getValue(1)
+  };
+
+  return DAG.getMergeValues(Ops, SL);
 }
 
 SDValue SITargetLowering::LowerFormalArguments(

Modified: llvm/trunk/test/CodeGen/AMDGPU/merge-stores.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/AMDGPU/merge-stores.ll?rev=284397&r1=284396&r2=284397&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/merge-stores.ll (original)
+++ llvm/trunk/test/CodeGen/AMDGPU/merge-stores.ll Mon Oct 17 11:21:45 2016
@@ -149,12 +149,8 @@ define void @merge_global_store_4_consta
   ret void
 }
 
-; FIXME: Should be able to merge this
 ; GCN-LABEL: {{^}}merge_global_store_4_constants_mixed_i32_f32:
-; GCN-NOAA: buffer_store_dword v
-; GCN-NOAA: buffer_store_dword v
-; GCN-NOAA: buffer_store_dword v
-; GCN-NOAA: buffer_store_dword v
+; GCN-NOAA: buffer_store_dwordx4 v
 
 ; GCN-AA: buffer_store_dwordx2
 ; GCN-AA: buffer_store_dword v