forked from OSchip/llvm-project
[AArch64] Refactor definition of EXT patterns to use a multiclass
Using a multiclass reduces duplication, and makes it easier to add new patterns later. This refactoring does add some new patterns, but as far as I can tell there's no IR that will end up triggering them so this is effectively NFC. Differential Revision: https://reviews.llvm.org/D53580 llvm-svn: 345271
This commit is contained in:
parent
49e61d90ca
commit
b8e7887f33
|
@ -4185,44 +4185,24 @@ def : Pat<(concat_vectors (v2i32 V64:$Rd),
|
|||
|
||||
defm EXT : SIMDBitwiseExtract<"ext">;
|
||||
|
||||
def : Pat<(v4i16 (AArch64ext V64:$Rn, V64:$Rm, (i32 imm:$imm))),
|
||||
(EXTv8i8 V64:$Rn, V64:$Rm, imm:$imm)>;
|
||||
def : Pat<(v8i16 (AArch64ext V128:$Rn, V128:$Rm, (i32 imm:$imm))),
|
||||
(EXTv16i8 V128:$Rn, V128:$Rm, imm:$imm)>;
|
||||
def : Pat<(v2i32 (AArch64ext V64:$Rn, V64:$Rm, (i32 imm:$imm))),
|
||||
(EXTv8i8 V64:$Rn, V64:$Rm, imm:$imm)>;
|
||||
def : Pat<(v2f32 (AArch64ext V64:$Rn, V64:$Rm, (i32 imm:$imm))),
|
||||
(EXTv8i8 V64:$Rn, V64:$Rm, imm:$imm)>;
|
||||
def : Pat<(v4i32 (AArch64ext V128:$Rn, V128:$Rm, (i32 imm:$imm))),
|
||||
(EXTv16i8 V128:$Rn, V128:$Rm, imm:$imm)>;
|
||||
def : Pat<(v4f32 (AArch64ext V128:$Rn, V128:$Rm, (i32 imm:$imm))),
|
||||
(EXTv16i8 V128:$Rn, V128:$Rm, imm:$imm)>;
|
||||
def : Pat<(v2i64 (AArch64ext V128:$Rn, V128:$Rm, (i32 imm:$imm))),
|
||||
(EXTv16i8 V128:$Rn, V128:$Rm, imm:$imm)>;
|
||||
def : Pat<(v2f64 (AArch64ext V128:$Rn, V128:$Rm, (i32 imm:$imm))),
|
||||
(EXTv16i8 V128:$Rn, V128:$Rm, imm:$imm)>;
|
||||
def : Pat<(v4f16 (AArch64ext V64:$Rn, V64:$Rm, (i32 imm:$imm))),
|
||||
(EXTv8i8 V64:$Rn, V64:$Rm, imm:$imm)>;
|
||||
def : Pat<(v8f16 (AArch64ext V128:$Rn, V128:$Rm, (i32 imm:$imm))),
|
||||
(EXTv16i8 V128:$Rn, V128:$Rm, imm:$imm)>;
|
||||
|
||||
// We use EXT to handle extract_subvector to copy the upper 64-bits of a
|
||||
// 128-bit vector.
|
||||
def : Pat<(v8i8 (extract_subvector V128:$Rn, (i64 8))),
|
||||
(EXTRACT_SUBREG (EXTv16i8 V128:$Rn, V128:$Rn, 8), dsub)>;
|
||||
def : Pat<(v4i16 (extract_subvector V128:$Rn, (i64 4))),
|
||||
(EXTRACT_SUBREG (EXTv16i8 V128:$Rn, V128:$Rn, 8), dsub)>;
|
||||
def : Pat<(v2i32 (extract_subvector V128:$Rn, (i64 2))),
|
||||
(EXTRACT_SUBREG (EXTv16i8 V128:$Rn, V128:$Rn, 8), dsub)>;
|
||||
def : Pat<(v1i64 (extract_subvector V128:$Rn, (i64 1))),
|
||||
(EXTRACT_SUBREG (EXTv16i8 V128:$Rn, V128:$Rn, 8), dsub)>;
|
||||
def : Pat<(v4f16 (extract_subvector V128:$Rn, (i64 4))),
|
||||
(EXTRACT_SUBREG (EXTv16i8 V128:$Rn, V128:$Rn, 8), dsub)>;
|
||||
def : Pat<(v2f32 (extract_subvector V128:$Rn, (i64 2))),
|
||||
(EXTRACT_SUBREG (EXTv16i8 V128:$Rn, V128:$Rn, 8), dsub)>;
|
||||
def : Pat<(v1f64 (extract_subvector V128:$Rn, (i64 1))),
|
||||
(EXTRACT_SUBREG (EXTv16i8 V128:$Rn, V128:$Rn, 8), dsub)>;
|
||||
multiclass ExtPat<ValueType VT64, ValueType VT128, int N> {
|
||||
def : Pat<(VT64 (AArch64ext V64:$Rn, V64:$Rm, (i32 imm:$imm))),
|
||||
(EXTv8i8 V64:$Rn, V64:$Rm, imm:$imm)>;
|
||||
def : Pat<(VT128 (AArch64ext V128:$Rn, V128:$Rm, (i32 imm:$imm))),
|
||||
(EXTv16i8 V128:$Rn, V128:$Rm, imm:$imm)>;
|
||||
// We use EXT to handle extract_subvector to copy the upper 64-bits of a
|
||||
// 128-bit vector.
|
||||
def : Pat<(VT64 (extract_subvector V128:$Rn, (i64 N))),
|
||||
(EXTRACT_SUBREG (EXTv16i8 V128:$Rn, V128:$Rn, 8), dsub)>;
|
||||
}
|
||||
|
||||
defm : ExtPat<v8i8, v16i8, 8>;
|
||||
defm : ExtPat<v4i16, v8i16, 4>;
|
||||
defm : ExtPat<v4f16, v8f16, 4>;
|
||||
defm : ExtPat<v2i32, v4i32, 2>;
|
||||
defm : ExtPat<v2f32, v4f32, 2>;
|
||||
defm : ExtPat<v1i64, v2i64, 1>;
|
||||
defm : ExtPat<v1f64, v2f64, 1>;
|
||||
|
||||
//----------------------------------------------------------------------------
|
||||
// AdvSIMD zip vector
|
||||
|
|
Loading…
Reference in New Issue