[AArch64][SVE2] Add pattern for BCAX #77159
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Bitwise clear and exclusive or
Add pattern for:
xor x, (and y, not(z)) -> bcax x, y, z
Change-Id: I42cb5e53eff59a12aef00b211400e59fd2112b54
|
@llvm/pr-subscribers-backend-aarch64 Author: Usman Nadeem (UsmanNadeem) ChangesBitwise clear and exclusive or Change-Id: I42cb5e53eff59a12aef00b211400e59fd2112b54 Full diff: https://github.com/llvm/llvm-project/pull/77159.diff 2 Files Affected:
diff --git a/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td b/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td
index 344a153890631e..ee10a7d1c706fc 100644
--- a/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td
+++ b/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td
@@ -453,6 +453,9 @@ def AArch64msb_m1 : PatFrags<(ops node:$pred, node:$op1, node:$op2, node:$op3),
def AArch64eor3 : PatFrags<(ops node:$op1, node:$op2, node:$op3),
[(int_aarch64_sve_eor3 node:$op1, node:$op2, node:$op3),
(xor node:$op1, (xor node:$op2, node:$op3))]>;
+def AArch64bcax : PatFrags<(ops node:$op1, node:$op2, node:$op3),
+ [(int_aarch64_sve_bcax node:$op1, node:$op2, node:$op3),
+ (xor node:$op1, (and node:$op2, (vnot node:$op3)))]>;
def AArch64fmla_m1 : PatFrags<(ops node:$pg, node:$za, node:$zn, node:$zm),
[(int_aarch64_sve_fmla node:$pg, node:$za, node:$zn, node:$zm),
@@ -3714,7 +3717,7 @@ let Predicates = [HasSVE2orSME] in {
// SVE2 bitwise ternary operations
defm EOR3_ZZZZ : sve2_int_bitwise_ternary_op<0b000, "eor3", AArch64eor3>;
- defm BCAX_ZZZZ : sve2_int_bitwise_ternary_op<0b010, "bcax", int_aarch64_sve_bcax>;
+ defm BCAX_ZZZZ : sve2_int_bitwise_ternary_op<0b010, "bcax", AArch64bcax>;
defm BSL_ZZZZ : sve2_int_bitwise_ternary_op<0b001, "bsl", int_aarch64_sve_bsl, AArch64bsp>;
defm BSL1N_ZZZZ : sve2_int_bitwise_ternary_op<0b011, "bsl1n", int_aarch64_sve_bsl1n>;
defm BSL2N_ZZZZ : sve2_int_bitwise_ternary_op<0b101, "bsl2n", int_aarch64_sve_bsl2n>;
diff --git a/llvm/test/CodeGen/AArch64/sve2-bcax.ll b/llvm/test/CodeGen/AArch64/sve2-bcax.ll
new file mode 100644
index 00000000000000..cadba288e8e7d9
--- /dev/null
+++ b/llvm/test/CodeGen/AArch64/sve2-bcax.ll
@@ -0,0 +1,143 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 4
+; RUN: llc -mtriple=aarch64 -mattr=+sve < %s -o - | FileCheck --check-prefix=SVE %s
+; RUN: llc -mtriple=aarch64 -mattr=+sve2 < %s -o - | FileCheck --check-prefix=SVE2 %s
+
+define <vscale x 2 x i64> @bcax_nxv2i64_1(<vscale x 2 x i64> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2) #0 {
+; SVE-LABEL: bcax_nxv2i64_1:
+; SVE: // %bb.0:
+; SVE-NEXT: bic z1.d, z2.d, z1.d
+; SVE-NEXT: eor z0.d, z1.d, z0.d
+; SVE-NEXT: ret
+;
+; SVE2-LABEL: bcax_nxv2i64_1:
+; SVE2: // %bb.0:
+; SVE2-NEXT: bcax z0.d, z0.d, z2.d, z1.d
+; SVE2-NEXT: ret
+ %4 = xor <vscale x 2 x i64> %1, shufflevector (<vscale x 2 x i64> insertelement (<vscale x 2 x i64> poison, i64 -1, i64 0), <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer)
+ %5 = and <vscale x 2 x i64> %4, %2
+ %6 = xor <vscale x 2 x i64> %5, %0
+ ret <vscale x 2 x i64> %6
+}
+
+define <vscale x 2 x i64> @bcax_nxv2i64_2(<vscale x 2 x i64> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2) #0 {
+; SVE-LABEL: bcax_nxv2i64_2:
+; SVE: // %bb.0:
+; SVE-NEXT: bic z0.d, z0.d, z1.d
+; SVE-NEXT: eor z0.d, z0.d, z2.d
+; SVE-NEXT: ret
+;
+; SVE2-LABEL: bcax_nxv2i64_2:
+; SVE2: // %bb.0:
+; SVE2-NEXT: bcax z2.d, z2.d, z0.d, z1.d
+; SVE2-NEXT: mov z0.d, z2.d
+; SVE2-NEXT: ret
+ %4 = xor <vscale x 2 x i64> %1, shufflevector (<vscale x 2 x i64> insertelement (<vscale x 2 x i64> poison, i64 -1, i64 0), <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer)
+ %5 = and <vscale x 2 x i64> %4, %0
+ %6 = xor <vscale x 2 x i64> %5, %2
+ ret <vscale x 2 x i64> %6
+}
+
+define <vscale x 4 x i32> @bcax_nxv4i32_1(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2) #0 {
+; SVE-LABEL: bcax_nxv4i32_1:
+; SVE: // %bb.0:
+; SVE-NEXT: bic z1.d, z2.d, z1.d
+; SVE-NEXT: eor z0.d, z1.d, z0.d
+; SVE-NEXT: ret
+;
+; SVE2-LABEL: bcax_nxv4i32_1:
+; SVE2: // %bb.0:
+; SVE2-NEXT: bcax z0.d, z0.d, z2.d, z1.d
+; SVE2-NEXT: ret
+ %4 = xor <vscale x 4 x i32> %1, shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 -1, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
+ %5 = and <vscale x 4 x i32> %4, %2
+ %6 = xor <vscale x 4 x i32> %5, %0
+ ret <vscale x 4 x i32> %6
+}
+
+define <vscale x 4 x i32> @bcax_nxv4i32_2(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2) #0 {
+; SVE-LABEL: bcax_nxv4i32_2:
+; SVE: // %bb.0:
+; SVE-NEXT: bic z0.d, z0.d, z1.d
+; SVE-NEXT: eor z0.d, z0.d, z2.d
+; SVE-NEXT: ret
+;
+; SVE2-LABEL: bcax_nxv4i32_2:
+; SVE2: // %bb.0:
+; SVE2-NEXT: bcax z2.d, z2.d, z0.d, z1.d
+; SVE2-NEXT: mov z0.d, z2.d
+; SVE2-NEXT: ret
+ %4 = xor <vscale x 4 x i32> %1, shufflevector (<vscale x 4 x i32> insertelement (<vscale x 4 x i32> poison, i32 -1, i64 0), <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer)
+ %5 = and <vscale x 4 x i32> %4, %0
+ %6 = xor <vscale x 4 x i32> %5, %2
+ ret <vscale x 4 x i32> %6
+}
+
+define <vscale x 8 x i16> @bcax_nxv8i16_1(<vscale x 8 x i16> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2) #0 {
+; SVE-LABEL: bcax_nxv8i16_1:
+; SVE: // %bb.0:
+; SVE-NEXT: bic z1.d, z2.d, z1.d
+; SVE-NEXT: eor z0.d, z1.d, z0.d
+; SVE-NEXT: ret
+;
+; SVE2-LABEL: bcax_nxv8i16_1:
+; SVE2: // %bb.0:
+; SVE2-NEXT: bcax z0.d, z0.d, z2.d, z1.d
+; SVE2-NEXT: ret
+ %4 = xor <vscale x 8 x i16> %1, shufflevector (<vscale x 8 x i16> insertelement (<vscale x 8 x i16> poison, i16 -1, i64 0), <vscale x 8 x i16> poison, <vscale x 8 x i32> zeroinitializer)
+ %5 = and <vscale x 8 x i16> %4, %2
+ %6 = xor <vscale x 8 x i16> %5, %0
+ ret <vscale x 8 x i16> %6
+}
+
+define <vscale x 8 x i16> @bcax_nxv8i16_2(<vscale x 8 x i16> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2) #0 {
+; SVE-LABEL: bcax_nxv8i16_2:
+; SVE: // %bb.0:
+; SVE-NEXT: bic z0.d, z0.d, z1.d
+; SVE-NEXT: eor z0.d, z0.d, z2.d
+; SVE-NEXT: ret
+;
+; SVE2-LABEL: bcax_nxv8i16_2:
+; SVE2: // %bb.0:
+; SVE2-NEXT: bcax z2.d, z2.d, z0.d, z1.d
+; SVE2-NEXT: mov z0.d, z2.d
+; SVE2-NEXT: ret
+ %4 = xor <vscale x 8 x i16> %1, shufflevector (<vscale x 8 x i16> insertelement (<vscale x 8 x i16> poison, i16 -1, i64 0), <vscale x 8 x i16> poison, <vscale x 8 x i32> zeroinitializer)
+ %5 = and <vscale x 8 x i16> %4, %0
+ %6 = xor <vscale x 8 x i16> %5, %2
+ ret <vscale x 8 x i16> %6
+}
+
+define <vscale x 16 x i8> @bcax_nxv16i8_1(<vscale x 16 x i8> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2) #0 {
+; SVE-LABEL: bcax_nxv16i8_1:
+; SVE: // %bb.0:
+; SVE-NEXT: bic z1.d, z2.d, z1.d
+; SVE-NEXT: eor z0.d, z1.d, z0.d
+; SVE-NEXT: ret
+;
+; SVE2-LABEL: bcax_nxv16i8_1:
+; SVE2: // %bb.0:
+; SVE2-NEXT: bcax z0.d, z0.d, z2.d, z1.d
+; SVE2-NEXT: ret
+ %4 = xor <vscale x 16 x i8> %1, shufflevector (<vscale x 16 x i8> insertelement (<vscale x 16 x i8> poison, i8 -1, i64 0), <vscale x 16 x i8> poison, <vscale x 16 x i32> zeroinitializer)
+ %5 = and <vscale x 16 x i8> %4, %2
+ %6 = xor <vscale x 16 x i8> %5, %0
+ ret <vscale x 16 x i8> %6
+}
+
+define <vscale x 16 x i8> @bcax_nxv16i8_2(<vscale x 16 x i8> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2) #0 {
+; SVE-LABEL: bcax_nxv16i8_2:
+; SVE: // %bb.0:
+; SVE-NEXT: bic z0.d, z0.d, z1.d
+; SVE-NEXT: eor z0.d, z0.d, z2.d
+; SVE-NEXT: ret
+;
+; SVE2-LABEL: bcax_nxv16i8_2:
+; SVE2: // %bb.0:
+; SVE2-NEXT: bcax z2.d, z2.d, z0.d, z1.d
+; SVE2-NEXT: mov z0.d, z2.d
+; SVE2-NEXT: ret
+ %4 = xor <vscale x 16 x i8> %1, shufflevector (<vscale x 16 x i8> insertelement (<vscale x 16 x i8> poison, i8 -1, i64 0), <vscale x 16 x i8> poison, <vscale x 16 x i32> zeroinitializer)
+ %5 = and <vscale x 16 x i8> %4, %0
+ %6 = xor <vscale x 16 x i8> %5, %2
+ ret <vscale x 16 x i8> %6
+}
|
UsmanNadeem
requested review from
davemgreen,
sdesmalen-arm,
david-arm and
paulwalker-arm
paulwalker-arm
approved these changes
Jan 6, 2024
| ; SVE2: // %bb.0: | ||
| ; SVE2-NEXT: bcax z0.d, z0.d, z2.d, z1.d | ||
| ; SVE2-NEXT: ret | ||
| %4 = xor <vscale x 2 x i64> %1, shufflevector (<vscale x 2 x i64> insertelement (<vscale x 2 x i64> poison, i64 -1, i64 0), <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer) |
There was a problem hiding this comment.
We're migrating away from ConstantExpr so do you mind using the new splat shorthand for these tests (i.e. splat ( i64 -1)). The created IR will be the same but the tests will at least be ready for the migration, plus easier to read.
justinfargnoli
pushed a commit
to justinfargnoli/llvm-project
that referenced
this pull request
Jan 28, 2024
Bitwise clear and exclusive or
Add pattern for:
xor x, (and y, not(z)) -> bcax x, y, z
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Bitwise clear and exclusive or
Add pattern for:
xor x, (and y, not(z)) -> bcax x, y, z
Change-Id: I42cb5e53eff59a12aef00b211400e59fd2112b54