; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt < %s -passes=instcombine -mtriple=x86_64-unknown-unknown -S | FileCheck %s
;; MASKED LOADS
; If the mask isn't constant, do nothing.
define <4 x float> @mload(ptr %f, <4 x i32> %mask) {
; CHECK-LABEL: @mload(
; CHECK-NEXT: [[LD:%.*]] = tail call <4 x float> @llvm.x86.avx.maskload.ps(ptr [[F:%.*]], <4 x i32> [[MASK:%.*]])
; CHECK-NEXT: ret <4 x float> [[LD]]
;
%ld = tail call <4 x float> @llvm.x86.avx.maskload.ps(ptr %f, <4 x i32> %mask)
ret <4 x float> %ld
}
; If the mask comes from a comparison, convert to an LLVM intrinsic. The backend should optimize further.
define <4 x float> @mload_v4f32_cmp(ptr %f, <4 x i32> %src) {
; CHECK-LABEL: @mload_v4f32_cmp(
; CHECK-NEXT: [[ICMP:%.*]] = icmp ne <4 x i32> [[SRC:%.*]], zeroinitializer
; CHECK-NEXT: [[LD:%.*]] = call <4 x float> @llvm.masked.load.v4f32.p0(ptr [[F:%.*]], i32 1, <4 x i1> [[ICMP]], <4 x float> zeroinitializer)
; CHECK-NEXT: ret <4 x float> [[LD]]
;
%icmp = icmp ne <4 x i32> %src, zeroinitializer
%mask = sext <4 x i1> %icmp to <4 x i32>
%ld = tail call <4 x float> @llvm.x86.avx.maskload.ps(ptr %f, <4 x i32> %mask)
ret <4 x float> %ld
}
; Zero mask returns a zero vector.
define <4 x float> @mload_zeros(ptr %f) {
; CHECK-LABEL: @mload_zeros(
; CHECK-NEXT: ret <4 x float> zeroinitializer
;
%ld = tail call <4 x float> @llvm.x86.avx.maskload.ps(ptr %f, <4 x i32> zeroinitializer)
ret <4 x float> %ld
}
; Only the sign bit matters.
define <4 x float> @mload_fake_ones(ptr %f) {
; CHECK-LABEL: @mload_fake_ones(
; CHECK-NEXT: ret <4 x float> zeroinitializer
;
%ld = tail call <4 x float> @llvm.x86.avx.maskload.ps(ptr %f, <4 x i32> <i32 1, i32 2, i32 3, i32 2147483647>)
ret <4 x float> %ld
}
; All mask bits are set, so this is just a vector load.
define <4 x float> @mload_real_ones(ptr %f) {
; CHECK-LABEL: @mload_real_ones(
; CHECK-NEXT: [[UNMASKEDLOAD:%.*]] = load <4 x float>, ptr [[F:%.*]], align 1
; CHECK-NEXT: ret <4 x float> [[UNMASKEDLOAD]]
;
%ld = tail call <4 x float> @llvm.x86.avx.maskload.ps(ptr %f, <4 x i32> <i32 -1, i32 -2, i32 -3, i32 2147483648>)
ret <4 x float> %ld
}
; It's a constant mask, so convert to an LLVM intrinsic. The backend should optimize further.
define <4 x float> @mload_one_one(ptr %f) {
; CHECK-LABEL: @mload_one_one(
; CHECK-NEXT: [[LD:%.*]] = call <4 x float> @llvm.masked.load.v4f32.p0(ptr [[F:%.*]], i32 1, <4 x i1> <i1 false, i1 false, i1 false, i1 true>, <4 x float> <float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float poison>)
; CHECK-NEXT: ret <4 x float> [[LD]]
;
%ld = tail call <4 x float> @llvm.x86.avx.maskload.ps(ptr %f, <4 x i32> <i32 0, i32 0, i32 0, i32 -1>)
ret <4 x float> %ld
}
; Try doubles.
define <2 x double> @mload_one_one_double(ptr %f) {
; CHECK-LABEL: @mload_one_one_double(
; CHECK-NEXT: [[LD:%.*]] = call <2 x double> @llvm.masked.load.v2f64.p0(ptr [[F:%.*]], i32 1, <2 x i1> <i1 true, i1 false>, <2 x double> <double poison, double 0.000000e+00>)
; CHECK-NEXT: ret <2 x double> [[LD]]
;
%ld = tail call <2 x double> @llvm.x86.avx.maskload.pd(ptr %f, <2 x i64> <i64 -1, i64 0>)
ret <2 x double> %ld
}
; Try 256-bit FP ops.
define <8 x float> @mload_v8f32(ptr %f) {
; CHECK-LABEL: @mload_v8f32(
; CHECK-NEXT: [[LD:%.*]] = call <8 x float> @llvm.masked.load.v8f32.p0(ptr [[F:%.*]], i32 1, <8 x i1> <i1 false, i1 false, i1 false, i1 true, i1 false, i1 false, i1 false, i1 false>, <8 x float> <float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float poison, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00>)
; CHECK-NEXT: ret <8 x float> [[LD]]
;
%ld = tail call <8 x float> @llvm.x86.avx.maskload.ps.256(ptr %f, <8 x i32> <i32 0, i32 0, i32 0, i32 -1, i32 0, i32 0, i32 0, i32 0>)
ret <8 x float> %ld
}
define <8 x float> @mload_v8f32_cmp(ptr %f, <8 x float> %src0, <8 x float> %src1) {
; CHECK-LABEL: @mload_v8f32_cmp(
; CHECK-NEXT: [[ICMP0:%.*]] = fcmp one <8 x float> [[SRC0:%.*]], zeroinitializer
; CHECK-NEXT: [[ICMP1:%.*]] = fcmp one <8 x float> [[SRC1:%.*]], zeroinitializer
; CHECK-NEXT: [[MASK1:%.*]] = and <8 x i1> [[ICMP0]], [[ICMP1]]
; CHECK-NEXT: [[LD:%.*]] = call <8 x float> @llvm.masked.load.v8f32.p0(ptr [[F:%.*]], i32 1, <8 x i1> [[MASK1]], <8 x float> zeroinitializer)
; CHECK-NEXT: ret <8 x float> [[LD]]
;
%icmp0 = fcmp one <8 x float> %src0, zeroinitializer
%icmp1 = fcmp one <8 x float> %src1, zeroinitializer
%ext0 = sext <8 x i1> %icmp0 to <8 x i32>
%ext1 = sext <8 x i1> %icmp1 to <8 x i32>
%mask = and <8 x i32> %ext0, %ext1
%ld = tail call <8 x float> @llvm.x86.avx.maskload.ps.256(ptr %f, <8 x i32> %mask)
ret <8 x float> %ld
}
define <4 x double> @mload_v4f64(ptr %f) {
; CHECK-LABEL: @mload_v4f64(
; CHECK-NEXT: [[LD:%.*]] = call <4 x double> @llvm.masked.load.v4f64.p0(ptr [[F:%.*]], i32 1, <4 x i1> <i1 true, i1 false, i1 false, i1 false>, <4 x double> <double poison, double 0.000000e+00, double 0.000000e+00, double 0.000000e+00>)
; CHECK-NEXT: ret <4 x double> [[LD]]
;
%ld = tail call <4 x double> @llvm.x86.avx.maskload.pd.256(ptr %f, <4 x i64> <i64 -1, i64 0, i64 0, i64 0>)
ret <4 x double> %ld
}
; Try the AVX2 variants.
define <4 x i32> @mload_v4i32(ptr %f) {
; CHECK-LABEL: @mload_v4i32(
; CHECK-NEXT: [[LD:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[F:%.*]], i32 1, <4 x i1> <i1 false, i1 false, i1 false, i1 true>, <4 x i32> <i32 0, i32 0, i32 0, i32 poison>)
; CHECK-NEXT: ret <4 x i32> [[LD]]
;
%ld = tail call <4 x i32> @llvm.x86.avx2.maskload.d(ptr %f, <4 x i32> <i32 0, i32 0, i32 0, i32 -1>)
ret <4 x i32> %ld
}
define <2 x i64> @mload_v2i64(ptr %f) {
; CHECK-LABEL: @mload_v2i64(
; CHECK-NEXT: [[LD:%.*]] = call <2 x i64> @llvm.masked.load.v2i64.p0(ptr [[F:%.*]], i32 1, <2 x i1> <i1 true, i1 false>, <2 x i64> <i64 poison, i64 0>)
; CHECK-NEXT: ret <2 x i64> [[LD]]
;
%ld = tail call <2 x i64> @llvm.x86.avx2.maskload.q(ptr %f, <2 x i64> <i64 -1, i64 0>)
ret <2 x i64> %ld
}
define <8 x i32> @mload_v8i32(ptr %f) {
; CHECK-LABEL: @mload_v8i32(
; CHECK-NEXT: [[LD:%.*]] = call <8 x i32> @llvm.masked.load.v8i32.p0(ptr [[F:%.*]], i32 1, <8 x i1> <i1 false, i1 false, i1 false, i1 true, i1 false, i1 false, i1 false, i1 false>, <8 x i32> <i32 0, i32 0, i32 0, i32 poison, i32 0, i32 0, i32 0, i32 0>)
; CHECK-NEXT: ret <8 x i32> [[LD]]
;
%ld = tail call <8 x i32> @llvm.x86.avx2.maskload.d.256(ptr %f, <8 x i32> <i32 0, i32 0, i32 0, i32 -1, i32 0, i32 0, i32 0, i32 0>)
ret <8 x i32> %ld
}
define <4 x i64> @mload_v4i64(ptr %f) {
; CHECK-LABEL: @mload_v4i64(
; CHECK-NEXT: [[LD:%.*]] = call <4 x i64> @llvm.masked.load.v4i64.p0(ptr [[F:%.*]], i32 1, <4 x i1> <i1 true, i1 false, i1 false, i1 false>, <4 x i64> <i64 poison, i64 0, i64 0, i64 0>)
; CHECK-NEXT: ret <4 x i64> [[LD]]
;
%ld = tail call <4 x i64> @llvm.x86.avx2.maskload.q.256(ptr %f, <4 x i64> <i64 -1, i64 0, i64 0, i64 0>)
ret <4 x i64> %ld
}
define <4 x i64> @mload_v4i64_cmp(ptr %f, <4 x i64> %src) {
; CHECK-LABEL: @mload_v4i64_cmp(
; CHECK-NEXT: [[ICMP:%.*]] = icmp sgt <4 x i64> [[SRC:%.*]], <i64 -1, i64 -1, i64 -1, i64 -1>
; CHECK-NEXT: [[LD:%.*]] = call <4 x i64> @llvm.masked.load.v4i64.p0(ptr [[F:%.*]], i32 1, <4 x i1> [[ICMP]], <4 x i64> zeroinitializer)
; CHECK-NEXT: ret <4 x i64> [[LD]]
;
%icmp = icmp sge <4 x i64> %src, zeroinitializer
%mask = sext <4 x i1> %icmp to <4 x i64>
%ld = tail call <4 x i64> @llvm.x86.avx2.maskload.q.256(ptr %f, <4 x i64> %mask)
ret <4 x i64> %ld
}
;; MASKED STORES
; If the mask isn't constant, do nothing.
define void @mstore(ptr %f, <4 x i32> %mask, <4 x float> %v) {
; CHECK-LABEL: @mstore(
; CHECK-NEXT: tail call void @llvm.x86.avx.maskstore.ps(ptr [[F:%.*]], <4 x i32> [[MASK:%.*]], <4 x float> [[V:%.*]])
; CHECK-NEXT: ret void
;
tail call void @llvm.x86.avx.maskstore.ps(ptr %f, <4 x i32> %mask, <4 x float> %v)
ret void
}
; If the mask comes from a comparison, convert to an LLVM intrinsic. The backend should optimize further.
define void @mstore_v4f32_cmp(ptr %f, <4 x i32> %src, <4 x float> %v) {
; CHECK-LABEL: @mstore_v4f32_cmp(
; CHECK-NEXT: [[ICMP:%.*]] = icmp eq <4 x i32> [[SRC:%.*]], zeroinitializer
; CHECK-NEXT: call void @llvm.masked.store.v4f32.p0(<4 x float> [[V:%.*]], ptr [[F:%.*]], i32 1, <4 x i1> [[ICMP]])
; CHECK-NEXT: ret void
;
%icmp = icmp eq <4 x i32> %src, zeroinitializer
%mask = sext <4 x i1> %icmp to <4 x i32>
tail call void @llvm.x86.avx.maskstore.ps(ptr %f, <4 x i32> %mask, <4 x float> %v)
ret void
}
; Zero mask is a nop.
define void @mstore_zeros(ptr %f, <4 x float> %v) {
; CHECK-LABEL: @mstore_zeros(
; CHECK-NEXT: ret void
;
tail call void @llvm.x86.avx.maskstore.ps(ptr %f, <4 x i32> zeroinitializer, <4 x float> %v)
ret void
}
; Only the sign bit matters.
define void @mstore_fake_ones(ptr %f, <4 x float> %v) {
; CHECK-LABEL: @mstore_fake_ones(
; CHECK-NEXT: ret void
;
tail call void @llvm.x86.avx.maskstore.ps(ptr %f, <4 x i32> <i32 1, i32 2, i32 3, i32 2147483647>, <4 x float> %v)
ret void
}
; All mask bits are set, so this is just a vector store.
define void @mstore_real_ones(ptr %f, <4 x float> %v) {
; CHECK-LABEL: @mstore_real_ones(
; CHECK-NEXT: store <4 x float> [[V:%.*]], ptr [[F:%.*]], align 1
; CHECK-NEXT: ret void
;
tail call void @llvm.x86.avx.maskstore.ps(ptr %f, <4 x i32> <i32 -1, i32 -2, i32 -3, i32 -2147483648>, <4 x float> %v)
ret void
}
; It's a constant mask, so convert to an LLVM intrinsic. The backend should optimize further.
define void @mstore_one_one(ptr %f, <4 x float> %v) {
; CHECK-LABEL: @mstore_one_one(
; CHECK-NEXT: call void @llvm.masked.store.v4f32.p0(<4 x float> [[V:%.*]], ptr [[F:%.*]], i32 1, <4 x i1> <i1 false, i1 false, i1 false, i1 true>)
; CHECK-NEXT: ret void
;
tail call void @llvm.x86.avx.maskstore.ps(ptr %f, <4 x i32> <i32 0, i32 0, i32 0, i32 -1>, <4 x float> %v)
ret void
}
; Try doubles.
define void @mstore_one_one_double(ptr %f, <2 x double> %v) {
; CHECK-LABEL: @mstore_one_one_double(
; CHECK-NEXT: call void @llvm.masked.store.v2f64.p0(<2 x double> [[V:%.*]], ptr [[F:%.*]], i32 1, <2 x i1> <i1 true, i1 false>)
; CHECK-NEXT: ret void
;
tail call void @llvm.x86.avx.maskstore.pd(ptr %f, <2 x i64> <i64 -1, i64 0>, <2 x double> %v)
ret void
}
; Try 256-bit FP ops.
define void @mstore_v8f32(ptr %f, <8 x float> %v) {
; CHECK-LABEL: @mstore_v8f32(
; CHECK-NEXT: call void @llvm.masked.store.v8f32.p0(<8 x float> [[V:%.*]], ptr [[F:%.*]], i32 1, <8 x i1> <i1 false, i1 false, i1 false, i1 false, i1 true, i1 true, i1 true, i1 true>)
; CHECK-NEXT: ret void
;
tail call void @llvm.x86.avx.maskstore.ps.256(ptr %f, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 -1, i32 -2, i32 -3, i32 -4>, <8 x float> %v)
ret void
}
define void @mstore_v4f64(ptr %f, <4 x double> %v) {
; CHECK-LABEL: @mstore_v4f64(
; CHECK-NEXT: call void @llvm.masked.store.v4f64.p0(<4 x double> [[V:%.*]], ptr [[F:%.*]], i32 1, <4 x i1> <i1 true, i1 false, i1 false, i1 false>)
; CHECK-NEXT: ret void
;
tail call void @llvm.x86.avx.maskstore.pd.256(ptr %f, <4 x i64> <i64 -1, i64 0, i64 1, i64 2>, <4 x double> %v)
ret void
}
define void @mstore_v4f64_cmp(ptr %f, <4 x i32> %src, <4 x double> %v) {
; CHECK-LABEL: @mstore_v4f64_cmp(
; CHECK-NEXT: [[ICMP:%.*]] = icmp sgt <4 x i32> [[SRC:%.*]], <i32 -1, i32 -1, i32 -1, i32 -1>
; CHECK-NEXT: call void @llvm.masked.store.v4f64.p0(<4 x double> [[V:%.*]], ptr [[F:%.*]], i32 1, <4 x i1> [[ICMP]])
; CHECK-NEXT: ret void
;
%icmp = icmp sge <4 x i32> %src, zeroinitializer
%mask = sext <4 x i1> %icmp to <4 x i64>
tail call void @llvm.x86.avx.maskstore.pd.256(ptr %f, <4 x i64> %mask, <4 x double> %v)
ret void
}
; Try the AVX2 variants.
define void @mstore_v4i32(ptr %f, <4 x i32> %v) {
; CHECK-LABEL: @mstore_v4i32(
; CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[V:%.*]], ptr [[F:%.*]], i32 1, <4 x i1> <i1 false, i1 false, i1 true, i1 true>)
; CHECK-NEXT: ret void
;
tail call void @llvm.x86.avx2.maskstore.d(ptr %f, <4 x i32> <i32 0, i32 1, i32 -1, i32 -2>, <4 x i32> %v)
ret void
}
define void @mstore_v2i64(ptr %f, <2 x i64> %v) {
; CHECK-LABEL: @mstore_v2i64(
; CHECK-NEXT: call void @llvm.masked.store.v2i64.p0(<2 x i64> [[V:%.*]], ptr [[F:%.*]], i32 1, <2 x i1> <i1 true, i1 false>)
; CHECK-NEXT: ret void
;
tail call void @llvm.x86.avx2.maskstore.q(ptr %f, <2 x i64> <i64 -1, i64 0>, <2 x i64> %v)
ret void
}
define void @mstore_v8i32(ptr %f, <8 x i32> %v) {
; CHECK-LABEL: @mstore_v8i32(
; CHECK-NEXT: call void @llvm.masked.store.v8i32.p0(<8 x i32> [[V:%.*]], ptr [[F:%.*]], i32 1, <8 x i1> <i1 false, i1 false, i1 false, i1 false, i1 true, i1 true, i1 true, i1 true>)
; CHECK-NEXT: ret void
;
tail call void @llvm.x86.avx2.maskstore.d.256(ptr %f, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 -1, i32 -2, i32 -3, i32 -4>, <8 x i32> %v)
ret void
}
define void @mstore_v4i64(ptr %f, <4 x i64> %v) {
; CHECK-LABEL: @mstore_v4i64(
; CHECK-NEXT: call void @llvm.masked.store.v4i64.p0(<4 x i64> [[V:%.*]], ptr [[F:%.*]], i32 1, <4 x i1> <i1 true, i1 false, i1 false, i1 false>)
; CHECK-NEXT: ret void
;
tail call void @llvm.x86.avx2.maskstore.q.256(ptr %f, <4 x i64> <i64 -1, i64 0, i64 1, i64 2>, <4 x i64> %v)
ret void
}
define void @mstore_v4i64_cmp(ptr %f, <4 x i64> %src0, <4 x i64> %src1, <4 x i64> %v) {
; CHECK-LABEL: @mstore_v4i64_cmp(
; CHECK-NEXT: [[ICMP0:%.*]] = icmp eq <4 x i64> [[SRC0:%.*]], zeroinitializer
; CHECK-NEXT: [[ICMP1:%.*]] = icmp ne <4 x i64> [[SRC1:%.*]], zeroinitializer
; CHECK-NEXT: [[MASK1:%.*]] = and <4 x i1> [[ICMP0]], [[ICMP1]]
; CHECK-NEXT: call void @llvm.masked.store.v4i64.p0(<4 x i64> [[V:%.*]], ptr [[F:%.*]], i32 1, <4 x i1> [[MASK1]])
; CHECK-NEXT: ret void
;
%icmp0 = icmp eq <4 x i64> %src0, zeroinitializer
%icmp1 = icmp ne <4 x i64> %src1, zeroinitializer
%ext0 = sext <4 x i1> %icmp0 to <4 x i64>
%ext1 = sext <4 x i1> %icmp1 to <4 x i64>
%mask = and <4 x i64> %ext0, %ext1
tail call void @llvm.x86.avx2.maskstore.q.256(ptr %f, <4 x i64> %mask, <4 x i64> %v)
ret void
}
; The original SSE2 masked store variant.
define void @mstore_v16i8_sse2_zeros(<16 x i8> %d, ptr %p) {
; CHECK-LABEL: @mstore_v16i8_sse2_zeros(
; CHECK-NEXT: ret void
;
tail call void @llvm.x86.sse2.maskmov.dqu(<16 x i8> %d, <16 x i8> zeroinitializer, ptr %p)
ret void
}
declare <4 x float> @llvm.x86.avx.maskload.ps(ptr, <4 x i32>)
declare <2 x double> @llvm.x86.avx.maskload.pd(ptr, <2 x i64>)
declare <8 x float> @llvm.x86.avx.maskload.ps.256(ptr, <8 x i32>)
declare <4 x double> @llvm.x86.avx.maskload.pd.256(ptr, <4 x i64>)
declare <4 x i32> @llvm.x86.avx2.maskload.d(ptr, <4 x i32>)
declare <2 x i64> @llvm.x86.avx2.maskload.q(ptr, <2 x i64>)
declare <8 x i32> @llvm.x86.avx2.maskload.d.256(ptr, <8 x i32>)
declare <4 x i64> @llvm.x86.avx2.maskload.q.256(ptr, <4 x i64>)
declare void @llvm.x86.avx.maskstore.ps(ptr, <4 x i32>, <4 x float>)
declare void @llvm.x86.avx.maskstore.pd(ptr, <2 x i64>, <2 x double>)
declare void @llvm.x86.avx.maskstore.ps.256(ptr, <8 x i32>, <8 x float>)
declare void @llvm.x86.avx.maskstore.pd.256(ptr, <4 x i64>, <4 x double>)
declare void @llvm.x86.avx2.maskstore.d(ptr, <4 x i32>, <4 x i32>)
declare void @llvm.x86.avx2.maskstore.q(ptr, <2 x i64>, <2 x i64>)
declare void @llvm.x86.avx2.maskstore.d.256(ptr, <8 x i32>, <8 x i32>)
declare void @llvm.x86.avx2.maskstore.q.256(ptr, <4 x i64>, <4 x i64>)
declare void @llvm.x86.sse2.maskmov.dqu(<16 x i8>, <16 x i8>, ptr)
Codebase Browser
llvm