; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=riscv32 -mattr=+m,+v -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,RV32
; RUN: llc -mtriple=riscv64 -mattr=+m,+v -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,RV64
define void @masked_load_v1i8(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v1i8:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 1, e8, mf8, ta, ma
; CHECK-NEXT: vle8.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle8.v v8, (a0), v0.t
; CHECK-NEXT: vse8.v v8, (a2)
; CHECK-NEXT: ret
%m = load <1 x i8>, ptr %m_ptr
%mask = icmp eq <1 x i8> %m, zeroinitializer
%load = call <1 x i8> @llvm.masked.load.v1i8(ptr %a, i32 8, <1 x i1> %mask, <1 x i8> undef)
store <1 x i8> %load, ptr %res_ptr
ret void
}
declare <1 x i8> @llvm.masked.load.v1i8(ptr, i32, <1 x i1>, <1 x i8>)
define void @masked_load_v1i16(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v1i16:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 1, e16, mf4, ta, ma
; CHECK-NEXT: vle16.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle16.v v8, (a0), v0.t
; CHECK-NEXT: vse16.v v8, (a2)
; CHECK-NEXT: ret
%m = load <1 x i16>, ptr %m_ptr
%mask = icmp eq <1 x i16> %m, zeroinitializer
%load = call <1 x i16> @llvm.masked.load.v1i16(ptr %a, i32 8, <1 x i1> %mask, <1 x i16> undef)
store <1 x i16> %load, ptr %res_ptr
ret void
}
declare <1 x i16> @llvm.masked.load.v1i16(ptr, i32, <1 x i1>, <1 x i16>)
define void @masked_load_v1i32(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v1i32:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
; CHECK-NEXT: vle32.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle32.v v8, (a0), v0.t
; CHECK-NEXT: vse32.v v8, (a2)
; CHECK-NEXT: ret
%m = load <1 x i32>, ptr %m_ptr
%mask = icmp eq <1 x i32> %m, zeroinitializer
%load = call <1 x i32> @llvm.masked.load.v1i32(ptr %a, i32 8, <1 x i1> %mask, <1 x i32> undef)
store <1 x i32> %load, ptr %res_ptr
ret void
}
declare <1 x i32> @llvm.masked.load.v1i32(ptr, i32, <1 x i1>, <1 x i32>)
define void @masked_load_v1i64(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v1i64:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 1, e64, m1, ta, ma
; CHECK-NEXT: vle64.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle64.v v8, (a0), v0.t
; CHECK-NEXT: vse64.v v8, (a2)
; CHECK-NEXT: ret
%m = load <1 x i64>, ptr %m_ptr
%mask = icmp eq <1 x i64> %m, zeroinitializer
%load = call <1 x i64> @llvm.masked.load.v1i64(ptr %a, i32 8, <1 x i1> %mask, <1 x i64> undef)
store <1 x i64> %load, ptr %res_ptr
ret void
}
declare <1 x i64> @llvm.masked.load.v1i64(ptr, i32, <1 x i1>, <1 x i64>)
define void @masked_load_v2i8(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v2i8:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 2, e8, mf8, ta, ma
; CHECK-NEXT: vle8.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle8.v v8, (a0), v0.t
; CHECK-NEXT: vse8.v v8, (a2)
; CHECK-NEXT: ret
%m = load <2 x i8>, ptr %m_ptr
%mask = icmp eq <2 x i8> %m, zeroinitializer
%load = call <2 x i8> @llvm.masked.load.v2i8(ptr %a, i32 8, <2 x i1> %mask, <2 x i8> undef)
store <2 x i8> %load, ptr %res_ptr
ret void
}
declare <2 x i8> @llvm.masked.load.v2i8(ptr, i32, <2 x i1>, <2 x i8>)
define void @masked_load_v2i16(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v2i16:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 2, e16, mf4, ta, ma
; CHECK-NEXT: vle16.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle16.v v8, (a0), v0.t
; CHECK-NEXT: vse16.v v8, (a2)
; CHECK-NEXT: ret
%m = load <2 x i16>, ptr %m_ptr
%mask = icmp eq <2 x i16> %m, zeroinitializer
%load = call <2 x i16> @llvm.masked.load.v2i16(ptr %a, i32 8, <2 x i1> %mask, <2 x i16> undef)
store <2 x i16> %load, ptr %res_ptr
ret void
}
declare <2 x i16> @llvm.masked.load.v2i16(ptr, i32, <2 x i1>, <2 x i16>)
define void @masked_load_v2i32(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v2i32:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
; CHECK-NEXT: vle32.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle32.v v8, (a0), v0.t
; CHECK-NEXT: vse32.v v8, (a2)
; CHECK-NEXT: ret
%m = load <2 x i32>, ptr %m_ptr
%mask = icmp eq <2 x i32> %m, zeroinitializer
%load = call <2 x i32> @llvm.masked.load.v2i32(ptr %a, i32 8, <2 x i1> %mask, <2 x i32> undef)
store <2 x i32> %load, ptr %res_ptr
ret void
}
declare <2 x i32> @llvm.masked.load.v2i32(ptr, i32, <2 x i1>, <2 x i32>)
define void @masked_load_v2i64(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v2i64:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 2, e64, m1, ta, ma
; CHECK-NEXT: vle64.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle64.v v8, (a0), v0.t
; CHECK-NEXT: vse64.v v8, (a2)
; CHECK-NEXT: ret
%m = load <2 x i64>, ptr %m_ptr
%mask = icmp eq <2 x i64> %m, zeroinitializer
%load = call <2 x i64> @llvm.masked.load.v2i64(ptr %a, i32 8, <2 x i1> %mask, <2 x i64> undef)
store <2 x i64> %load, ptr %res_ptr
ret void
}
declare <2 x i64> @llvm.masked.load.v2i64(ptr, i32, <2 x i1>, <2 x i64>)
define void @masked_load_v4i8(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v4i8:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma
; CHECK-NEXT: vle8.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle8.v v8, (a0), v0.t
; CHECK-NEXT: vse8.v v8, (a2)
; CHECK-NEXT: ret
%m = load <4 x i8>, ptr %m_ptr
%mask = icmp eq <4 x i8> %m, zeroinitializer
%load = call <4 x i8> @llvm.masked.load.v4i8(ptr %a, i32 8, <4 x i1> %mask, <4 x i8> undef)
store <4 x i8> %load, ptr %res_ptr
ret void
}
declare <4 x i8> @llvm.masked.load.v4i8(ptr, i32, <4 x i1>, <4 x i8>)
define void @masked_load_v4i16(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v4i16:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
; CHECK-NEXT: vle16.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle16.v v8, (a0), v0.t
; CHECK-NEXT: vse16.v v8, (a2)
; CHECK-NEXT: ret
%m = load <4 x i16>, ptr %m_ptr
%mask = icmp eq <4 x i16> %m, zeroinitializer
%load = call <4 x i16> @llvm.masked.load.v4i16(ptr %a, i32 8, <4 x i1> %mask, <4 x i16> undef)
store <4 x i16> %load, ptr %res_ptr
ret void
}
declare <4 x i16> @llvm.masked.load.v4i16(ptr, i32, <4 x i1>, <4 x i16>)
define void @masked_load_v4i32(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v4i32:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
; CHECK-NEXT: vle32.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle32.v v8, (a0), v0.t
; CHECK-NEXT: vse32.v v8, (a2)
; CHECK-NEXT: ret
%m = load <4 x i32>, ptr %m_ptr
%mask = icmp eq <4 x i32> %m, zeroinitializer
%load = call <4 x i32> @llvm.masked.load.v4i32(ptr %a, i32 8, <4 x i1> %mask, <4 x i32> undef)
store <4 x i32> %load, ptr %res_ptr
ret void
}
declare <4 x i32> @llvm.masked.load.v4i32(ptr, i32, <4 x i1>, <4 x i32>)
define void @masked_load_v4i64(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v4i64:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 4, e64, m2, ta, ma
; CHECK-NEXT: vle64.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle64.v v8, (a0), v0.t
; CHECK-NEXT: vse64.v v8, (a2)
; CHECK-NEXT: ret
%m = load <4 x i64>, ptr %m_ptr
%mask = icmp eq <4 x i64> %m, zeroinitializer
%load = call <4 x i64> @llvm.masked.load.v4i64(ptr %a, i32 8, <4 x i1> %mask, <4 x i64> undef)
store <4 x i64> %load, ptr %res_ptr
ret void
}
declare <4 x i64> @llvm.masked.load.v4i64(ptr, i32, <4 x i1>, <4 x i64>)
define void @masked_load_v8i8(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v8i8:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma
; CHECK-NEXT: vle8.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle8.v v8, (a0), v0.t
; CHECK-NEXT: vse8.v v8, (a2)
; CHECK-NEXT: ret
%m = load <8 x i8>, ptr %m_ptr
%mask = icmp eq <8 x i8> %m, zeroinitializer
%load = call <8 x i8> @llvm.masked.load.v8i8(ptr %a, i32 8, <8 x i1> %mask, <8 x i8> undef)
store <8 x i8> %load, ptr %res_ptr
ret void
}
declare <8 x i8> @llvm.masked.load.v8i8(ptr, i32, <8 x i1>, <8 x i8>)
define void @masked_load_v8i16(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v8i16:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma
; CHECK-NEXT: vle16.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle16.v v8, (a0), v0.t
; CHECK-NEXT: vse16.v v8, (a2)
; CHECK-NEXT: ret
%m = load <8 x i16>, ptr %m_ptr
%mask = icmp eq <8 x i16> %m, zeroinitializer
%load = call <8 x i16> @llvm.masked.load.v8i16(ptr %a, i32 8, <8 x i1> %mask, <8 x i16> undef)
store <8 x i16> %load, ptr %res_ptr
ret void
}
declare <8 x i16> @llvm.masked.load.v8i16(ptr, i32, <8 x i1>, <8 x i16>)
define void @masked_load_v8i32(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v8i32:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
; CHECK-NEXT: vle32.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle32.v v8, (a0), v0.t
; CHECK-NEXT: vse32.v v8, (a2)
; CHECK-NEXT: ret
%m = load <8 x i32>, ptr %m_ptr
%mask = icmp eq <8 x i32> %m, zeroinitializer
%load = call <8 x i32> @llvm.masked.load.v8i32(ptr %a, i32 8, <8 x i1> %mask, <8 x i32> undef)
store <8 x i32> %load, ptr %res_ptr
ret void
}
declare <8 x i32> @llvm.masked.load.v8i32(ptr, i32, <8 x i1>, <8 x i32>)
define void @masked_load_v8i64(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v8i64:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 8, e64, m4, ta, ma
; CHECK-NEXT: vle64.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle64.v v8, (a0), v0.t
; CHECK-NEXT: vse64.v v8, (a2)
; CHECK-NEXT: ret
%m = load <8 x i64>, ptr %m_ptr
%mask = icmp eq <8 x i64> %m, zeroinitializer
%load = call <8 x i64> @llvm.masked.load.v8i64(ptr %a, i32 8, <8 x i1> %mask, <8 x i64> undef)
store <8 x i64> %load, ptr %res_ptr
ret void
}
declare <8 x i64> @llvm.masked.load.v8i64(ptr, i32, <8 x i1>, <8 x i64>)
define void @masked_load_v16i8(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v16i8:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma
; CHECK-NEXT: vle8.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle8.v v8, (a0), v0.t
; CHECK-NEXT: vse8.v v8, (a2)
; CHECK-NEXT: ret
%m = load <16 x i8>, ptr %m_ptr
%mask = icmp eq <16 x i8> %m, zeroinitializer
%load = call <16 x i8> @llvm.masked.load.v16i8(ptr %a, i32 8, <16 x i1> %mask, <16 x i8> undef)
store <16 x i8> %load, ptr %res_ptr
ret void
}
declare <16 x i8> @llvm.masked.load.v16i8(ptr, i32, <16 x i1>, <16 x i8>)
define void @masked_load_v16i16(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v16i16:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 16, e16, m2, ta, ma
; CHECK-NEXT: vle16.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle16.v v8, (a0), v0.t
; CHECK-NEXT: vse16.v v8, (a2)
; CHECK-NEXT: ret
%m = load <16 x i16>, ptr %m_ptr
%mask = icmp eq <16 x i16> %m, zeroinitializer
%load = call <16 x i16> @llvm.masked.load.v16i16(ptr %a, i32 8, <16 x i1> %mask, <16 x i16> undef)
store <16 x i16> %load, ptr %res_ptr
ret void
}
declare <16 x i16> @llvm.masked.load.v16i16(ptr, i32, <16 x i1>, <16 x i16>)
define void @masked_load_v16i32(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v16i32:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
; CHECK-NEXT: vle32.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle32.v v8, (a0), v0.t
; CHECK-NEXT: vse32.v v8, (a2)
; CHECK-NEXT: ret
%m = load <16 x i32>, ptr %m_ptr
%mask = icmp eq <16 x i32> %m, zeroinitializer
%load = call <16 x i32> @llvm.masked.load.v16i32(ptr %a, i32 8, <16 x i1> %mask, <16 x i32> undef)
store <16 x i32> %load, ptr %res_ptr
ret void
}
declare <16 x i32> @llvm.masked.load.v16i32(ptr, i32, <16 x i1>, <16 x i32>)
define void @masked_load_v16i64(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v16i64:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; CHECK-NEXT: vle64.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle64.v v8, (a0), v0.t
; CHECK-NEXT: vse64.v v8, (a2)
; CHECK-NEXT: ret
%m = load <16 x i64>, ptr %m_ptr
%mask = icmp eq <16 x i64> %m, zeroinitializer
%load = call <16 x i64> @llvm.masked.load.v16i64(ptr %a, i32 8, <16 x i1> %mask, <16 x i64> undef)
store <16 x i64> %load, ptr %res_ptr
ret void
}
declare <16 x i64> @llvm.masked.load.v16i64(ptr, i32, <16 x i1>, <16 x i64>)
define void @masked_load_v32i8(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v32i8:
; CHECK: # %bb.0:
; CHECK-NEXT: li a3, 32
; CHECK-NEXT: vsetvli zero, a3, e8, m2, ta, ma
; CHECK-NEXT: vle8.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle8.v v8, (a0), v0.t
; CHECK-NEXT: vse8.v v8, (a2)
; CHECK-NEXT: ret
%m = load <32 x i8>, ptr %m_ptr
%mask = icmp eq <32 x i8> %m, zeroinitializer
%load = call <32 x i8> @llvm.masked.load.v32i8(ptr %a, i32 8, <32 x i1> %mask, <32 x i8> undef)
store <32 x i8> %load, ptr %res_ptr
ret void
}
declare <32 x i8> @llvm.masked.load.v32i8(ptr, i32, <32 x i1>, <32 x i8>)
define void @masked_load_v32i16(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v32i16:
; CHECK: # %bb.0:
; CHECK-NEXT: li a3, 32
; CHECK-NEXT: vsetvli zero, a3, e16, m4, ta, ma
; CHECK-NEXT: vle16.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle16.v v8, (a0), v0.t
; CHECK-NEXT: vse16.v v8, (a2)
; CHECK-NEXT: ret
%m = load <32 x i16>, ptr %m_ptr
%mask = icmp eq <32 x i16> %m, zeroinitializer
%load = call <32 x i16> @llvm.masked.load.v32i16(ptr %a, i32 8, <32 x i1> %mask, <32 x i16> undef)
store <32 x i16> %load, ptr %res_ptr
ret void
}
declare <32 x i16> @llvm.masked.load.v32i16(ptr, i32, <32 x i1>, <32 x i16>)
define void @masked_load_v32i32(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v32i32:
; CHECK: # %bb.0:
; CHECK-NEXT: li a3, 32
; CHECK-NEXT: vsetvli zero, a3, e32, m8, ta, ma
; CHECK-NEXT: vle32.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle32.v v8, (a0), v0.t
; CHECK-NEXT: vse32.v v8, (a2)
; CHECK-NEXT: ret
%m = load <32 x i32>, ptr %m_ptr
%mask = icmp eq <32 x i32> %m, zeroinitializer
%load = call <32 x i32> @llvm.masked.load.v32i32(ptr %a, i32 8, <32 x i1> %mask, <32 x i32> undef)
store <32 x i32> %load, ptr %res_ptr
ret void
}
declare <32 x i32> @llvm.masked.load.v32i32(ptr, i32, <32 x i1>, <32 x i32>)
define void @masked_load_v32i64(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; RV32-LABEL: masked_load_v32i64:
; RV32: # %bb.0:
; RV32-NEXT: addi a3, a1, 128
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vle64.v v0, (a1)
; RV32-NEXT: vle64.v v24, (a3)
; RV32-NEXT: li a1, 32
; RV32-NEXT: vsetvli zero, a1, e32, m8, ta, ma
; RV32-NEXT: vmv.v.i v16, 0
; RV32-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV32-NEXT: vmseq.vv v8, v0, v16
; RV32-NEXT: vmseq.vv v0, v24, v16
; RV32-NEXT: addi a1, a0, 128
; RV32-NEXT: vle64.v v16, (a1), v0.t
; RV32-NEXT: vmv1r.v v0, v8
; RV32-NEXT: vle64.v v8, (a0), v0.t
; RV32-NEXT: vse64.v v8, (a2)
; RV32-NEXT: addi a0, a2, 128
; RV32-NEXT: vse64.v v16, (a0)
; RV32-NEXT: ret
;
; RV64-LABEL: masked_load_v32i64:
; RV64: # %bb.0:
; RV64-NEXT: addi a3, a1, 128
; RV64-NEXT: vsetivli zero, 16, e64, m8, ta, ma
; RV64-NEXT: vle64.v v16, (a1)
; RV64-NEXT: vle64.v v24, (a3)
; RV64-NEXT: vmseq.vi v8, v16, 0
; RV64-NEXT: vmseq.vi v0, v24, 0
; RV64-NEXT: addi a1, a0, 128
; RV64-NEXT: vle64.v v16, (a1), v0.t
; RV64-NEXT: vmv1r.v v0, v8
; RV64-NEXT: vle64.v v8, (a0), v0.t
; RV64-NEXT: vse64.v v8, (a2)
; RV64-NEXT: addi a0, a2, 128
; RV64-NEXT: vse64.v v16, (a0)
; RV64-NEXT: ret
%m = load <32 x i64>, ptr %m_ptr
%mask = icmp eq <32 x i64> %m, zeroinitializer
%load = call <32 x i64> @llvm.masked.load.v32i64(ptr %a, i32 8, <32 x i1> %mask, <32 x i64> undef)
store <32 x i64> %load, ptr %res_ptr
ret void
}
declare <32 x i64> @llvm.masked.load.v32i64(ptr, i32, <32 x i1>, <32 x i64>)
define void @masked_load_v64i8(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v64i8:
; CHECK: # %bb.0:
; CHECK-NEXT: li a3, 64
; CHECK-NEXT: vsetvli zero, a3, e8, m4, ta, ma
; CHECK-NEXT: vle8.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle8.v v8, (a0), v0.t
; CHECK-NEXT: vse8.v v8, (a2)
; CHECK-NEXT: ret
%m = load <64 x i8>, ptr %m_ptr
%mask = icmp eq <64 x i8> %m, zeroinitializer
%load = call <64 x i8> @llvm.masked.load.v64i8(ptr %a, i32 8, <64 x i1> %mask, <64 x i8> undef)
store <64 x i8> %load, ptr %res_ptr
ret void
}
declare <64 x i8> @llvm.masked.load.v64i8(ptr, i32, <64 x i1>, <64 x i8>)
define void @masked_load_v64i16(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v64i16:
; CHECK: # %bb.0:
; CHECK-NEXT: li a3, 64
; CHECK-NEXT: vsetvli zero, a3, e16, m8, ta, ma
; CHECK-NEXT: vle16.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle16.v v8, (a0), v0.t
; CHECK-NEXT: vse16.v v8, (a2)
; CHECK-NEXT: ret
%m = load <64 x i16>, ptr %m_ptr
%mask = icmp eq <64 x i16> %m, zeroinitializer
%load = call <64 x i16> @llvm.masked.load.v64i16(ptr %a, i32 8, <64 x i1> %mask, <64 x i16> undef)
store <64 x i16> %load, ptr %res_ptr
ret void
}
declare <64 x i16> @llvm.masked.load.v64i16(ptr, i32, <64 x i1>, <64 x i16>)
define void @masked_load_v64i32(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v64i32:
; CHECK: # %bb.0:
; CHECK-NEXT: addi a3, a1, 128
; CHECK-NEXT: li a4, 32
; CHECK-NEXT: vsetvli zero, a4, e32, m8, ta, ma
; CHECK-NEXT: vle32.v v16, (a1)
; CHECK-NEXT: vle32.v v24, (a3)
; CHECK-NEXT: vmseq.vi v8, v16, 0
; CHECK-NEXT: vmseq.vi v0, v24, 0
; CHECK-NEXT: addi a1, a0, 128
; CHECK-NEXT: vle32.v v16, (a1), v0.t
; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vle32.v v8, (a0), v0.t
; CHECK-NEXT: vse32.v v8, (a2)
; CHECK-NEXT: addi a0, a2, 128
; CHECK-NEXT: vse32.v v16, (a0)
; CHECK-NEXT: ret
%m = load <64 x i32>, ptr %m_ptr
%mask = icmp eq <64 x i32> %m, zeroinitializer
%load = call <64 x i32> @llvm.masked.load.v64i32(ptr %a, i32 8, <64 x i1> %mask, <64 x i32> undef)
store <64 x i32> %load, ptr %res_ptr
ret void
}
declare <64 x i32> @llvm.masked.load.v64i32(ptr, i32, <64 x i1>, <64 x i32>)
define void @masked_load_v128i8(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v128i8:
; CHECK: # %bb.0:
; CHECK-NEXT: li a3, 128
; CHECK-NEXT: vsetvli zero, a3, e8, m8, ta, ma
; CHECK-NEXT: vle8.v v8, (a1)
; CHECK-NEXT: vmseq.vi v0, v8, 0
; CHECK-NEXT: vle8.v v8, (a0), v0.t
; CHECK-NEXT: vse8.v v8, (a2)
; CHECK-NEXT: ret
%m = load <128 x i8>, ptr %m_ptr
%mask = icmp eq <128 x i8> %m, zeroinitializer
%load = call <128 x i8> @llvm.masked.load.v128i8(ptr %a, i32 8, <128 x i1> %mask, <128 x i8> undef)
store <128 x i8> %load, ptr %res_ptr
ret void
}
declare <128 x i8> @llvm.masked.load.v128i8(ptr, i32, <128 x i1>, <128 x i8>)
define void @masked_load_v256i8(ptr %a, ptr %m_ptr, ptr %res_ptr) nounwind {
; CHECK-LABEL: masked_load_v256i8:
; CHECK: # %bb.0:
; CHECK-NEXT: addi a3, a1, 128
; CHECK-NEXT: li a4, 128
; CHECK-NEXT: vsetvli zero, a4, e8, m8, ta, ma
; CHECK-NEXT: vle8.v v16, (a1)
; CHECK-NEXT: vle8.v v24, (a3)
; CHECK-NEXT: vmseq.vi v8, v16, 0
; CHECK-NEXT: vmseq.vi v0, v24, 0
; CHECK-NEXT: addi a1, a0, 128
; CHECK-NEXT: vle8.v v16, (a1), v0.t
; CHECK-NEXT: vmv1r.v v0, v8
; CHECK-NEXT: vle8.v v8, (a0), v0.t
; CHECK-NEXT: vse8.v v8, (a2)
; CHECK-NEXT: addi a0, a2, 128
; CHECK-NEXT: vse8.v v16, (a0)
; CHECK-NEXT: ret
%m = load <256 x i8>, ptr %m_ptr
%mask = icmp eq <256 x i8> %m, zeroinitializer
%load = call <256 x i8> @llvm.masked.load.v256i8(ptr %a, i32 8, <256 x i1> %mask, <256 x i8> undef)
store <256 x i8> %load, ptr %res_ptr
ret void
}
declare <256 x i8> @llvm.masked.load.v256i8(ptr, i32, <256 x i1>, <256 x i8>)