Skip to content

[AArch64][NEON] Add eor3 patterns for V64 xors #165376

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Oct 29, 2025
Merged

[AArch64][NEON] Add eor3 patterns for V64 xors #165376

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
f2d0ef1
[AArch64][NEON] Add eor3 patterns for V64 xors
Lukacma Oct 28, 2025
71515c9
Apply suggestion from @jthackray
Lukacma Oct 29, 2025
9c8209c
Merge branch 'main' into neon-eor3
Lukacma Oct 29, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
24 changes: 16 additions & 8 deletions llvm/lib/Target/AArch64/AArch64InstrInfo.td
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1805,14 +1805,22 @@ def : SHA3_pattern<EOR3, int_aarch64_crypto_eor3u, v8i16>;
def : SHA3_pattern<EOR3, int_aarch64_crypto_eor3u, v4i32>;
def : SHA3_pattern<EOR3, int_aarch64_crypto_eor3u, v2i64>;

class EOR3_pattern<ValueType VecTy>
: Pat<(xor (xor (VecTy V128:$Vn), (VecTy V128:$Vm)), (VecTy V128:$Va)),
(EOR3 (VecTy V128:$Vn), (VecTy V128:$Vm), (VecTy V128:$Va))>;

def : EOR3_pattern<v16i8>;
def : EOR3_pattern<v8i16>;
def : EOR3_pattern<v4i32>;
def : EOR3_pattern<v2i64>;
multiclass EOR3_pattern<ValueType Vec128Ty, ValueType Vec64Ty>{
def: Pat<(xor (xor (Vec128Ty V128:$Vn), (Vec128Ty V128:$Vm)), (Vec128Ty V128:$Va)),
(EOR3 (Vec128Ty V128:$Vn), (Vec128Ty V128:$Vm), (Vec128Ty V128:$Va))>;
def : Pat<(xor (xor (Vec64Ty V64:$Vn), (Vec64Ty V64:$Vm)), (Vec64Ty V64:$Va)),
(EXTRACT_SUBREG
(EOR3
(INSERT_SUBREG (IMPLICIT_DEF), V64:$Vn, dsub),
(INSERT_SUBREG (IMPLICIT_DEF), V64:$Vm, dsub),
(INSERT_SUBREG (IMPLICIT_DEF), V64:$Va, dsub)),
dsub)>;
}

defm : EOR3_pattern<v16i8, v8i8>;
defm : EOR3_pattern<v8i16, v4i16>;
defm : EOR3_pattern<v4i32, v2i32>;
defm : EOR3_pattern<v2i64, v1i64>;

class BCAX_pattern<ValueType VecTy>
: Pat<(xor (VecTy V128:$Vn), (and (VecTy V128:$Vm), (vnot (VecTy V128:$Va)))),
Expand Down
151 changes: 151 additions & 0 deletions llvm/test/CodeGen/AArch64/eor3.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -277,3 +277,154 @@ define <2 x i64> @eor3_vnot(<2 x i64> %0, <2 x i64> %1) {
ret <2 x i64> %4
}

define <1 x i64> @eor3_1x64(<1 x i64> %0, <1 x i64> %1, <1 x i64> %2) {
; SHA3-LABEL: eor3_1x64:
; SHA3: // %bb.0:
; SHA3-NEXT: // kill: def $d0 killed $d0 def $q0
; SHA3-NEXT: // kill: def $d2 killed $d2 def $q2
; SHA3-NEXT: // kill: def $d1 killed $d1 def $q1
; SHA3-NEXT: eor3 v0.16b, v1.16b, v2.16b, v0.16b
; SHA3-NEXT: // kill: def $d0 killed $d0 killed $q0
; SHA3-NEXT: ret
;
; NOSHA3-LABEL: eor3_1x64:
; NOSHA3: // %bb.0:
; NOSHA3-NEXT: eor v1.8b, v1.8b, v2.8b
; NOSHA3-NEXT: eor v0.8b, v1.8b, v0.8b
; NOSHA3-NEXT: ret
;
; SVE2-LABEL: eor3_1x64:
; SVE2: // %bb.0:
; SVE2-NEXT: // kill: def $d1 killed $d1 def $z1
; SVE2-NEXT: // kill: def $d2 killed $d2 def $z2
; SVE2-NEXT: // kill: def $d0 killed $d0 def $z0
; SVE2-NEXT: eor3 z1.d, z1.d, z2.d, z0.d
; SVE2-NEXT: fmov d0, d1
; SVE2-NEXT: ret
;
; SHA3-SVE2-LABEL: eor3_1x64:
; SHA3-SVE2: // %bb.0:
; SHA3-SVE2-NEXT: // kill: def $d0 killed $d0 def $q0
; SHA3-SVE2-NEXT: // kill: def $d2 killed $d2 def $q2
; SHA3-SVE2-NEXT: // kill: def $d1 killed $d1 def $q1
; SHA3-SVE2-NEXT: eor3 v0.16b, v1.16b, v2.16b, v0.16b
; SHA3-SVE2-NEXT: // kill: def $d0 killed $d0 killed $q0
; SHA3-SVE2-NEXT: ret
%4 = xor <1 x i64> %1, %2
%5 = xor <1 x i64> %4, %0
ret <1 x i64> %5
}

define <2 x i32> @eor3_2x32(<2 x i32> %0, <2 x i32> %1, <2 x i32> %2) {
; SHA3-LABEL: eor3_2x32:
; SHA3: // %bb.0:
; SHA3-NEXT: // kill: def $d0 killed $d0 def $q0
; SHA3-NEXT: // kill: def $d2 killed $d2 def $q2
; SHA3-NEXT: // kill: def $d1 killed $d1 def $q1
; SHA3-NEXT: eor3 v0.16b, v1.16b, v2.16b, v0.16b
; SHA3-NEXT: // kill: def $d0 killed $d0 killed $q0
; SHA3-NEXT: ret
;
; NOSHA3-LABEL: eor3_2x32:
; NOSHA3: // %bb.0:
; NOSHA3-NEXT: eor v1.8b, v1.8b, v2.8b
; NOSHA3-NEXT: eor v0.8b, v1.8b, v0.8b
; NOSHA3-NEXT: ret
;
; SVE2-LABEL: eor3_2x32:
; SVE2: // %bb.0:
; SVE2-NEXT: // kill: def $d1 killed $d1 def $z1
; SVE2-NEXT: // kill: def $d2 killed $d2 def $z2
; SVE2-NEXT: // kill: def $d0 killed $d0 def $z0
; SVE2-NEXT: eor3 z1.d, z1.d, z2.d, z0.d
; SVE2-NEXT: fmov d0, d1
; SVE2-NEXT: ret
;
; SHA3-SVE2-LABEL: eor3_2x32:
; SHA3-SVE2: // %bb.0:
; SHA3-SVE2-NEXT: // kill: def $d0 killed $d0 def $q0
; SHA3-SVE2-NEXT: // kill: def $d2 killed $d2 def $q2
; SHA3-SVE2-NEXT: // kill: def $d1 killed $d1 def $q1
; SHA3-SVE2-NEXT: eor3 v0.16b, v1.16b, v2.16b, v0.16b
; SHA3-SVE2-NEXT: // kill: def $d0 killed $d0 killed $q0
; SHA3-SVE2-NEXT: ret
%4 = xor <2 x i32> %1, %2
%5 = xor <2 x i32> %4, %0
ret <2 x i32> %5
}

define <4 x i16> @eor3_4x16(<4 x i16> %0, <4 x i16> %1, <4 x i16> %2) {
; SHA3-LABEL: eor3_4x16:
; SHA3: // %bb.0:
; SHA3-NEXT: // kill: def $d0 killed $d0 def $q0
; SHA3-NEXT: // kill: def $d2 killed $d2 def $q2
; SHA3-NEXT: // kill: def $d1 killed $d1 def $q1
; SHA3-NEXT: eor3 v0.16b, v1.16b, v2.16b, v0.16b
; SHA3-NEXT: // kill: def $d0 killed $d0 killed $q0
; SHA3-NEXT: ret
;
; NOSHA3-LABEL: eor3_4x16:
; NOSHA3: // %bb.0:
; NOSHA3-NEXT: eor v1.8b, v1.8b, v2.8b
; NOSHA3-NEXT: eor v0.8b, v1.8b, v0.8b
; NOSHA3-NEXT: ret
;
; SVE2-LABEL: eor3_4x16:
; SVE2: // %bb.0:
; SVE2-NEXT: // kill: def $d1 killed $d1 def $z1
; SVE2-NEXT: // kill: def $d2 killed $d2 def $z2
; SVE2-NEXT: // kill: def $d0 killed $d0 def $z0
; SVE2-NEXT: eor3 z1.d, z1.d, z2.d, z0.d
; SVE2-NEXT: fmov d0, d1
; SVE2-NEXT: ret
;
; SHA3-SVE2-LABEL: eor3_4x16:
; SHA3-SVE2: // %bb.0:
; SHA3-SVE2-NEXT: // kill: def $d0 killed $d0 def $q0
; SHA3-SVE2-NEXT: // kill: def $d2 killed $d2 def $q2
; SHA3-SVE2-NEXT: // kill: def $d1 killed $d1 def $q1
; SHA3-SVE2-NEXT: eor3 v0.16b, v1.16b, v2.16b, v0.16b
; SHA3-SVE2-NEXT: // kill: def $d0 killed $d0 killed $q0
; SHA3-SVE2-NEXT: ret
%4 = xor <4 x i16> %1, %2
%5 = xor <4 x i16> %4, %0
ret <4 x i16> %5
}

define <8 x i8> @eor3_8x8(<8 x i8> %0, <8 x i8> %1, <8 x i8> %2) {
; SHA3-LABEL: eor3_8x8:
; SHA3: // %bb.0:
; SHA3-NEXT: // kill: def $d0 killed $d0 def $q0
; SHA3-NEXT: // kill: def $d2 killed $d2 def $q2
; SHA3-NEXT: // kill: def $d1 killed $d1 def $q1
; SHA3-NEXT: eor3 v0.16b, v1.16b, v2.16b, v0.16b
; SHA3-NEXT: // kill: def $d0 killed $d0 killed $q0
; SHA3-NEXT: ret
;
; NOSHA3-LABEL: eor3_8x8:
; NOSHA3: // %bb.0:
; NOSHA3-NEXT: eor v1.8b, v1.8b, v2.8b
; NOSHA3-NEXT: eor v0.8b, v1.8b, v0.8b
; NOSHA3-NEXT: ret
;
; SVE2-LABEL: eor3_8x8:
; SVE2: // %bb.0:
; SVE2-NEXT: // kill: def $d1 killed $d1 def $z1
; SVE2-NEXT: // kill: def $d2 killed $d2 def $z2
; SVE2-NEXT: // kill: def $d0 killed $d0 def $z0
; SVE2-NEXT: eor3 z1.d, z1.d, z2.d, z0.d
; SVE2-NEXT: fmov d0, d1
; SVE2-NEXT: ret
;
; SHA3-SVE2-LABEL: eor3_8x8:
; SHA3-SVE2: // %bb.0:
; SHA3-SVE2-NEXT: // kill: def $d0 killed $d0 def $q0
; SHA3-SVE2-NEXT: // kill: def $d2 killed $d2 def $q2
; SHA3-SVE2-NEXT: // kill: def $d1 killed $d1 def $q1
; SHA3-SVE2-NEXT: eor3 v0.16b, v1.16b, v2.16b, v0.16b
; SHA3-SVE2-NEXT: // kill: def $d0 killed $d0 killed $q0
; SHA3-SVE2-NEXT: ret
%4 = xor <8 x i8> %1, %2
%5 = xor <8 x i8> %4, %0
ret <8 x i8> %5
}