}
+def FP32Denormals : Predicate<"Subtarget.hasFP32Denormals()">;
+def FP64Denormals : Predicate<"Subtarget.hasFP64Denormals()">;
+def UnsafeFPMath : Predicate<"TM.Options.UnsafeFPMath">;
+
def InstFlag : OperandWithDefaultOps <i32, (ops (i32 0))>;
def ADDRIndirect : ComplexPattern<iPTR, 2, "SelectADDRIndirect", [], []>;
+let OperandType = "OPERAND_IMMEDIATE" in {
+
def u32imm : Operand<i32> {
let PrintMethod = "printU32ImmOperand";
}
let PrintMethod = "printU8ImmOperand";
}
+} // End OperandType = "OPERAND_IMMEDIATE"
+
+//===--------------------------------------------------------------------===//
+// Custom Operands
+//===--------------------------------------------------------------------===//
+def brtarget : Operand<OtherVT>;
+
//===----------------------------------------------------------------------===//
// PatLeafs for floating-point comparisons
//===----------------------------------------------------------------------===//
def COND_NULL : PatLeaf <
(cond),
- [{return false;}]
+ [{(void)N; return false;}]
>;
//===----------------------------------------------------------------------===//
// Load/Store Pattern Fragments
//===----------------------------------------------------------------------===//
+class PrivateMemOp <dag ops, dag frag> : PatFrag <ops, frag, [{
+ return cast<MemSDNode>(N)->getAddressSpace() == AMDGPUAS::PRIVATE_ADDRESS;
+}]>;
+
+class PrivateLoad <SDPatternOperator op> : PrivateMemOp <
+ (ops node:$ptr), (op node:$ptr)
+>;
+
+class PrivateStore <SDPatternOperator op> : PrivateMemOp <
+ (ops node:$value, node:$ptr), (op node:$value, node:$ptr)
+>;
+
+def extloadi8_private : PrivateLoad <extloadi8>;
+def sextloadi8_private : PrivateLoad <sextloadi8>;
+def extloadi16_private : PrivateLoad <extloadi16>;
+def sextloadi16_private : PrivateLoad <sextloadi16>;
+def load_private : PrivateLoad <load>;
+
+def truncstorei8_private : PrivateStore <truncstorei8>;
+def truncstorei16_private : PrivateStore <truncstorei16>;
+def store_private : PrivateStore <store>;
+
+def global_store : PatFrag<(ops node:$val, node:$ptr),
+ (store node:$val, node:$ptr), [{
+ return isGlobalStore(dyn_cast<StoreSDNode>(N));
+}]>;
+
+// Global address space loads
+def global_load : PatFrag<(ops node:$ptr), (load node:$ptr), [{
+ return isGlobalLoad(dyn_cast<LoadSDNode>(N));
+}]>;
+
+// Constant address space loads
+def constant_load : PatFrag<(ops node:$ptr), (load node:$ptr), [{
+ return isConstantLoad(dyn_cast<LoadSDNode>(N), -1);
+}]>;
+
def az_extload : PatFrag<(ops node:$ptr), (unindexedload node:$ptr), [{
LoadSDNode *L = cast<LoadSDNode>(N);
return L->getExtensionType() == ISD::ZEXTLOAD ||
return isGlobalLoad(dyn_cast<LoadSDNode>(N));
}]>;
+def az_extloadi8_flat : PatFrag<(ops node:$ptr), (az_extloadi8 node:$ptr), [{
+ return isFlatLoad(dyn_cast<LoadSDNode>(N));
+}]>;
+
+def sextloadi8_flat : PatFrag<(ops node:$ptr), (sextloadi8 node:$ptr), [{
+ return isFlatLoad(dyn_cast<LoadSDNode>(N));
+}]>;
+
def az_extloadi8_constant : PatFrag<(ops node:$ptr), (az_extloadi8 node:$ptr), [{
return isConstantLoad(dyn_cast<LoadSDNode>(N), -1);
}]>;
return isGlobalLoad(dyn_cast<LoadSDNode>(N));
}]>;
+def az_extloadi16_flat : PatFrag<(ops node:$ptr), (az_extloadi16 node:$ptr), [{
+ return isFlatLoad(dyn_cast<LoadSDNode>(N));
+}]>;
+
+def sextloadi16_flat : PatFrag<(ops node:$ptr), (sextloadi16 node:$ptr), [{
+ return isFlatLoad(dyn_cast<LoadSDNode>(N));
+}]>;
+
def az_extloadi16_constant : PatFrag<(ops node:$ptr), (az_extloadi16 node:$ptr), [{
return isConstantLoad(dyn_cast<LoadSDNode>(N), -1);
}]>;
return isGlobalLoad(dyn_cast<LoadSDNode>(N));
}]>;
+def az_extloadi32_flat : PatFrag<(ops node:$ptr),
+ (az_extloadi32 node:$ptr), [{
+ return isFlatLoad(dyn_cast<LoadSDNode>(N));
+}]>;
+
def az_extloadi32_constant : PatFrag<(ops node:$ptr),
(az_extloadi32 node:$ptr), [{
return isConstantLoad(dyn_cast<LoadSDNode>(N), -1);
return isGlobalStore(dyn_cast<StoreSDNode>(N));
}]>;
+def truncstorei8_flat : PatFrag<(ops node:$val, node:$ptr),
+ (truncstorei8 node:$val, node:$ptr), [{
+ return isFlatStore(dyn_cast<StoreSDNode>(N));
+}]>;
+
+def truncstorei16_flat : PatFrag<(ops node:$val, node:$ptr),
+ (truncstorei16 node:$val, node:$ptr), [{
+ return isFlatStore(dyn_cast<StoreSDNode>(N));
+}]>;
+
def local_store : PatFrag<(ops node:$val, node:$ptr),
(store node:$val, node:$ptr), [{
return isLocalStore(dyn_cast<StoreSDNode>(N));
return isLocalLoad(dyn_cast<LoadSDNode>(N));
}]>;
-def atomic_load_add_local : PatFrag<(ops node:$ptr, node:$value),
- (atomic_load_add node:$ptr, node:$value), [{
- return dyn_cast<MemSDNode>(N)->getAddressSpace() == AMDGPUAS::LOCAL_ADDRESS;
+class Aligned8Bytes <dag ops, dag frag> : PatFrag <ops, frag, [{
+ return cast<MemSDNode>(N)->getAlignment() % 8 == 0;
}]>;
-def atomic_load_sub_local : PatFrag<(ops node:$ptr, node:$value),
- (atomic_load_sub node:$ptr, node:$value), [{
- return dyn_cast<MemSDNode>(N)->getAddressSpace() == AMDGPUAS::LOCAL_ADDRESS;
+def local_load_aligned8bytes : Aligned8Bytes <
+ (ops node:$ptr), (local_load node:$ptr)
+>;
+
+def local_store_aligned8bytes : Aligned8Bytes <
+ (ops node:$val, node:$ptr), (local_store node:$val, node:$ptr)
+>;
+
+class local_binary_atomic_op<SDNode atomic_op> :
+ PatFrag<(ops node:$ptr, node:$value),
+ (atomic_op node:$ptr, node:$value), [{
+ return cast<MemSDNode>(N)->getAddressSpace() == AMDGPUAS::LOCAL_ADDRESS;
}]>;
+
+def atomic_swap_local : local_binary_atomic_op<atomic_swap>;
+def atomic_load_add_local : local_binary_atomic_op<atomic_load_add>;
+def atomic_load_sub_local : local_binary_atomic_op<atomic_load_sub>;
+def atomic_load_and_local : local_binary_atomic_op<atomic_load_and>;
+def atomic_load_or_local : local_binary_atomic_op<atomic_load_or>;
+def atomic_load_xor_local : local_binary_atomic_op<atomic_load_xor>;
+def atomic_load_nand_local : local_binary_atomic_op<atomic_load_nand>;
+def atomic_load_min_local : local_binary_atomic_op<atomic_load_min>;
+def atomic_load_max_local : local_binary_atomic_op<atomic_load_max>;
+def atomic_load_umin_local : local_binary_atomic_op<atomic_load_umin>;
+def atomic_load_umax_local : local_binary_atomic_op<atomic_load_umax>;
+
def mskor_global : PatFrag<(ops node:$val, node:$ptr),
(AMDGPUstore_mskor node:$val, node:$ptr), [{
return dyn_cast<MemSDNode>(N)->getAddressSpace() == AMDGPUAS::GLOBAL_ADDRESS;
}]>;
+
+def atomic_cmp_swap_32_local :
+ PatFrag<(ops node:$ptr, node:$cmp, node:$swap),
+ (atomic_cmp_swap node:$ptr, node:$cmp, node:$swap), [{
+ AtomicSDNode *AN = cast<AtomicSDNode>(N);
+ return AN->getMemoryVT() == MVT::i32 &&
+ AN->getAddressSpace() == AMDGPUAS::LOCAL_ADDRESS;
+}]>;
+
+def atomic_cmp_swap_64_local :
+ PatFrag<(ops node:$ptr, node:$cmp, node:$swap),
+ (atomic_cmp_swap node:$ptr, node:$cmp, node:$swap), [{
+ AtomicSDNode *AN = cast<AtomicSDNode>(N);
+ return AN->getMemoryVT() == MVT::i64 &&
+ AN->getAddressSpace() == AMDGPUAS::LOCAL_ADDRESS;
+}]>;
+
+def flat_load : PatFrag<(ops node:$ptr), (load node:$ptr), [{
+ return isFlatLoad(dyn_cast<LoadSDNode>(N));
+}]>;
+
+def flat_store : PatFrag<(ops node:$val, node:$ptr),
+ (store node:$val, node:$ptr), [{
+ return isFlatStore(dyn_cast<StoreSDNode>(N));
+}]>;
+
+def mskor_flat : PatFrag<(ops node:$val, node:$ptr),
+ (AMDGPUstore_mskor node:$val, node:$ptr), [{
+ return dyn_cast<MemSDNode>(N)->getAddressSpace() == AMDGPUAS::FLAT_ADDRESS;
+}]>;
+
+class global_binary_atomic_op<SDNode atomic_op> : PatFrag<
+ (ops node:$ptr, node:$value),
+ (atomic_op node:$ptr, node:$value),
+ [{return cast<MemSDNode>(N)->getAddressSpace() == AMDGPUAS::GLOBAL_ADDRESS;}]
+>;
+
+def atomic_add_global : global_binary_atomic_op<atomic_load_add>;
+
+//===----------------------------------------------------------------------===//
+// Misc Pattern Fragments
+//===----------------------------------------------------------------------===//
+
+def fmad : PatFrag <
+ (ops node:$src0, node:$src1, node:$src2),
+ (fadd (fmul node:$src0, node:$src1), node:$src2)
+>;
+
class Constants {
int TWO_PI = 0x40c90fdb;
int PI = 0x40490fdb;
int TWO_PI_INV = 0x3e22f983;
int FP_UINT_MAX_PLUS_1 = 0x4f800000; // 1 << 32 in floating point encoding
+int FP32_NEG_ONE = 0xbf800000;
+int FP32_ONE = 0x3f800000;
}
def CONST : Constants;
(outs rc:$dst),
(ins rc:$src0),
"CLAMP $dst, $src0",
- [(set f32:$dst, (int_AMDIL_clamp f32:$src0, (f32 FP_ZERO), (f32 FP_ONE)))]
+ [(set f32:$dst, (AMDGPUclamp f32:$src0, (f32 FP_ZERO), (f32 FP_ONE)))]
>;
class FABS <RegisterClass rc> : AMDGPUShaderInst <
// BFI_INT patterns
-multiclass BFIPatterns <Instruction BFI_INT> {
+multiclass BFIPatterns <Instruction BFI_INT, Instruction LoadImm32> {
// Definition from ISA doc:
// (y & x) | (z & ~x)
(BFI_INT $x, $y, $z)
>;
+ def : Pat <
+ (fcopysign f32:$src0, f32:$src1),
+ (BFI_INT (LoadImm32 0x7fffffff), $src0, $src1)
+ >;
+
+ def : Pat <
+ (f64 (fcopysign f64:$src0, f64:$src1)),
+ (INSERT_SUBREG (INSERT_SUBREG (f64 (IMPLICIT_DEF)),
+ (i32 (EXTRACT_SUBREG $src0, sub0)), sub0),
+ (BFI_INT (LoadImm32 0x7fffffff),
+ (i32 (EXTRACT_SUBREG $src0, sub1)),
+ (i32 (EXTRACT_SUBREG $src1, sub1))), sub1)
+ >;
}
// SHA-256 Ma patterns
(Inst $src0, $src1, $src2)
>;
+multiclass Expand24IBitOps<Instruction MulInst, Instruction AddInst> {
+ def _expand_imad24 : Pat <
+ (AMDGPUmad_i24 i32:$src0, i32:$src1, i32:$src2),
+ (AddInst (MulInst $src0, $src1), $src2)
+ >;
+
+ def _expand_imul24 : Pat <
+ (AMDGPUmul_i24 i32:$src0, i32:$src1),
+ (MulInst $src0, $src1)
+ >;
+}
+
+multiclass Expand24UBitOps<Instruction MulInst, Instruction AddInst> {
+ def _expand_umad24 : Pat <
+ (AMDGPUmad_u24 i32:$src0, i32:$src1, i32:$src2),
+ (AddInst (MulInst $src0, $src1), $src2)
+ >;
+
+ def _expand_umul24 : Pat <
+ (AMDGPUmul_u24 i32:$src0, i32:$src1),
+ (MulInst $src0, $src1)
+ >;
+}
+
+class RcpPat<Instruction RcpInst, ValueType vt> : Pat <
+ (fdiv FP_ONE, vt:$src),
+ (RcpInst $src)
+>;
+
+multiclass RsqPat<Instruction RsqInst, ValueType vt> {
+ def : Pat <
+ (fdiv FP_ONE, (fsqrt vt:$src)),
+ (RsqInst $src)
+ >;
+
+ def : Pat <
+ (AMDGPUrcp (fsqrt vt:$src)),
+ (RsqInst $src)
+ >;
+}
include "R600Instructions.td"
include "R700Instructions.td"