AMDGPU: Fix folding SGPRs into madak/madmk src0

Because of the special immediate operand, the constant bus is already used so SGPRs are never useful. r263212 changed the name of the immediate operand, which broke the verifier check for the restriction. llvm-svn: 274564
llvm · Jul 5, 2016 · ffc8275 · ffc8275
1 parent a8d89f3
commit ffc8275
Show file tree

Hide file tree

Showing 5 changed files with 55 additions and 6 deletions.
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUInstructions.td b/llvm/lib/Target/AMDGPU/AMDGPUInstructions.td
@@ -49,6 +49,13 @@ def UnsafeFPMath : Predicate<"TM.Options.UnsafeFPMath">;
 def InstFlag : OperandWithDefaultOps <i32, (ops (i32 0))>;
 def ADDRIndirect : ComplexPattern<iPTR, 2, "SelectADDRIndirect", [], []>;
 
+// 32-bit VALU immediate operand that uses the constant bus.
+def u32kimm : Operand<i32> {
+  let OperandNamespace = "AMDGPU";
+  let OperandType = "OPERAND_KIMM32";
+  let PrintMethod = "printU32ImmOperand";
+}
+
 let OperandType = "OPERAND_IMMEDIATE" in {
 
 def u32imm : Operand<i32> {

diff --git a/llvm/lib/Target/AMDGPU/SIDefines.h b/llvm/lib/Target/AMDGPU/SIDefines.h
@@ -50,7 +50,12 @@ namespace AMDGPU {
     /// Operand with register or 32-bit immediate
     OPERAND_REG_IMM32 = MCOI::OPERAND_FIRST_TARGET,
     /// Operand with register or inline constant
-    OPERAND_REG_INLINE_C
+    OPERAND_REG_INLINE_C,
+
+    /// Operand with 32-bit immediate that uses the constant bus. The standard
+    /// OPERAND_IMMEDIATE should be used for special immediates such as source
+    /// modifiers.
+    OPERAND_KIMM32
   };
 }
 }

diff --git a/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp b/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
@@ -1695,6 +1695,7 @@ bool SIInstrInfo::verifyInstruction(const MachineInstr &MI,
       }
       break;
     case MCOI::OPERAND_IMMEDIATE:
+    case AMDGPU::OPERAND_KIMM32:
       // Check if this operand is an immediate.
       // FrameIndex operands will be replaced by immediates, so they are
       // allowed.
@@ -1731,6 +1732,10 @@ bool SIInstrInfo::verifyInstruction(const MachineInstr &MI,
     const int OpIndices[] = { Src0Idx, Src1Idx, Src2Idx };
 
     unsigned ConstantBusCount = 0;
+
+    if (AMDGPU::getNamedOperandIdx(Opcode, AMDGPU::OpName::imm) != -1)
+      ++ConstantBusCount;
+
     unsigned SGPRUsed = findImplicitSGPRRead(MI);
     if (SGPRUsed != AMDGPU::NoRegister)
       ++ConstantBusCount;
@@ -2020,9 +2025,12 @@ bool SIInstrInfo::isOperandLegal(const MachineInstr &MI, unsigned OpIdx,
       if (i == OpIdx)
         continue;
       const MachineOperand &Op = MI.getOperand(i);
-      if (Op.isReg() &&
-          (Op.getReg() != SGPRUsed.Reg || Op.getSubReg() != SGPRUsed.SubReg) &&
-          usesConstantBus(MRI, Op, getOpSize(MI, i))) {
+      if (Op.isReg()) {
+        if ((Op.getReg() != SGPRUsed.Reg || Op.getSubReg() != SGPRUsed.SubReg) &&
+            usesConstantBus(MRI, Op, getOpSize(MI, i))) {
+          return false;
+        }
+      } else if (InstDesc.OpInfo[i].OperandType == AMDGPU::OPERAND_KIMM32) {
         return false;
       }
     }

diff --git a/llvm/lib/Target/AMDGPU/SIInstrInfo.td b/llvm/lib/Target/AMDGPU/SIInstrInfo.td
@@ -1582,12 +1582,12 @@ def VOP_I64_I64_I64 : VOPProfile <[i64, i64, i64, untyped]>;
 
 def VOP_F32_F32_F32_F32 : VOPProfile <[f32, f32, f32, f32]>;
 def VOP_MADAK : VOPProfile <[f32, f32, f32, f32]> {
-  field dag Ins32 = (ins VCSrc_32:$src0, VGPR_32:$src1, u32imm:$imm);
+  field dag Ins32 = (ins VCSrc_32:$src0, VGPR_32:$src1, u32kimm:$imm);
   field string Asm32 = "$vdst, $src0, $src1, $imm";
   field bit HasExt = 0;
 }
 def VOP_MADMK : VOPProfile <[f32, f32, f32, f32]> {
-  field dag Ins32 = (ins VCSrc_32:$src0, u32imm:$imm, VGPR_32:$src1);
+  field dag Ins32 = (ins VCSrc_32:$src0, u32kimm:$imm, VGPR_32:$src1);
   field string Asm32 = "$vdst, $src0, $imm, $src1";
   field bit HasExt = 0;
 }

diff --git a/llvm/test/CodeGen/AMDGPU/madak.ll b/llvm/test/CodeGen/AMDGPU/madak.ll
@@ -191,3 +191,32 @@ define void @no_madak_src1_modifier_f32(float addrspace(1)* noalias %out, float
   store float %madak, float addrspace(1)* %out.gep, align 4
   ret void
 }
+
+; SIFoldOperands should not fold the SGPR copy into the instruction
+; because the implicit immediate already uses the constant bus.
+; GCN-LABEL: {{^}}madak_constant_bus_violation:
+; GCN: s_load_dword [[SGPR0:s[0-9]+]], s{{\[[0-9]+:[0-9]+\]}}, {{0xa|0x28}}
+; GCN: v_mov_b32_e32 [[SGPR0_VCOPY:v[0-9]+]], [[SGPR0]]
+; GCN: buffer_load_dword [[VGPR:v[0-9]+]]
+; GCN: v_madak_f32_e32 [[MADAK:v[0-9]+]], 0.5, [[SGPR0_VCOPY]], 0x42280000
+; GCN: v_mul_f32_e32 [[MUL:v[0-9]+]], [[VGPR]], [[MADAK]]
+; GCN: buffer_store_dword [[MUL]]
+define void @madak_constant_bus_violation(i32 %arg1, float %sgpr0, float %sgpr1) #0 {
+bb:
+  %tmp = icmp eq i32 %arg1, 0
+  br i1 %tmp, label %bb3, label %bb4
+
+bb3:
+  store volatile float 0.0, float addrspace(1)* undef
+  br label %bb4
+
+bb4:
+  %vgpr = load volatile float, float addrspace(1)* undef
+  %tmp0 = fmul float %sgpr0, 0.5
+  %tmp1 = fadd float %tmp0, 42.0
+  %tmp2 = fmul float %tmp1, %vgpr
+  store volatile float %tmp2, float addrspace(1)* undef, align 4
+  ret void
+}
+
+attributes #0 = { nounwind}