| ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py |
| ; RUN: llc -mtriple=riscv32 -mattr=+v -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,RV32,RV32-ONLY |
| ; RUN: llc -mtriple=riscv32 -mattr=+v,+zba,+zbb -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,RV32,RV32VB |
| ; RUN: llc -mtriple=riscv32 -mattr=+v,+zba,+zbb,+zbkb -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,RV32,RV32VB-PACK |
| ; RUN: llc -mtriple=riscv64 -mattr=+v -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,RV64,RV64V,RV64V-ONLY |
| ; RUN: llc -mtriple=riscv64 -mattr=+v,+rva22u64 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,RV64,RV64V,RVA22U64 |
| ; RUN: llc -mtriple=riscv64 -mattr=+v,+rva22u64,+zbkb -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,RV64,RV64V,RVA22U64-PACK |
| ; RUN: llc -mtriple=riscv64 -mattr=+zve32x,+zvl128b -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,RV64,RV64ZVE32 |
| |
| define void @buildvec_vid_v16i8(ptr %x) { |
| ; CHECK-LABEL: buildvec_vid_v16i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vse8.v v8, (a0) |
| ; CHECK-NEXT: ret |
| store <16 x i8> <i8 0, i8 1, i8 2, i8 3, i8 4, i8 5, i8 6, i8 7, i8 8, i8 9, i8 10, i8 11, i8 12, i8 13, i8 14, i8 15>, ptr %x |
| ret void |
| } |
| |
| define void @buildvec_vid_undefelts_v16i8(ptr %x) { |
| ; CHECK-LABEL: buildvec_vid_undefelts_v16i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vse8.v v8, (a0) |
| ; CHECK-NEXT: ret |
| store <16 x i8> <i8 0, i8 1, i8 2, i8 poison, i8 4, i8 poison, i8 6, i8 7, i8 8, i8 9, i8 10, i8 11, i8 12, i8 13, i8 14, i8 15>, ptr %x |
| ret void |
| } |
| |
| ; TODO: Could do VID then insertelement on missing elements |
| define void @buildvec_notquite_vid_v16i8(ptr %x) { |
| ; CHECK-LABEL: buildvec_notquite_vid_v16i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: lui a1, %hi(.LCPI2_0) |
| ; CHECK-NEXT: addi a1, a1, %lo(.LCPI2_0) |
| ; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; CHECK-NEXT: vle8.v v8, (a1) |
| ; CHECK-NEXT: vse8.v v8, (a0) |
| ; CHECK-NEXT: ret |
| store <16 x i8> <i8 0, i8 1, i8 3, i8 3, i8 4, i8 5, i8 6, i8 7, i8 8, i8 9, i8 10, i8 11, i8 12, i8 13, i8 14, i8 15>, ptr %x |
| ret void |
| } |
| |
| define void @buildvec_vid_plus_imm_v16i8(ptr %x) { |
| ; CHECK-LABEL: buildvec_vid_plus_imm_v16i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vadd.vi v8, v8, 2 |
| ; CHECK-NEXT: vse8.v v8, (a0) |
| ; CHECK-NEXT: ret |
| store <16 x i8> <i8 2, i8 3, i8 4, i8 5, i8 6, i8 7, i8 8, i8 9, i8 10, i8 11, i8 12, i8 13, i8 14, i8 15, i8 16, i8 17>, ptr %x |
| ret void |
| } |
| |
| define void @buildvec_vid_plus_nonimm_v16i8(ptr %x) { |
| ; CHECK-LABEL: buildvec_vid_plus_nonimm_v16i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: li a1, 100 |
| ; CHECK-NEXT: vadd.vx v8, v8, a1 |
| ; CHECK-NEXT: vse8.v v8, (a0) |
| ; CHECK-NEXT: ret |
| store <16 x i8> <i8 100, i8 101, i8 102, i8 103, i8 104, i8 105, i8 106, i8 107, i8 108, i8 109, i8 110, i8 111, i8 112, i8 113, i8 114, i8 115>, ptr %x |
| ret void |
| } |
| |
| define void @buildvec_vid_mpy_imm_v16i8(ptr %x) { |
| ; CHECK-LABEL: buildvec_vid_mpy_imm_v16i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: li a1, 3 |
| ; CHECK-NEXT: vmul.vx v8, v8, a1 |
| ; CHECK-NEXT: vse8.v v8, (a0) |
| ; CHECK-NEXT: ret |
| store <16 x i8> <i8 0, i8 3, i8 6, i8 9, i8 12, i8 15, i8 18, i8 21, i8 24, i8 27, i8 30, i8 33, i8 36, i8 39, i8 42, i8 45>, ptr %x |
| ret void |
| } |
| |
| define <4 x i8> @buildvec_vid_step2_add0_v4i8() { |
| ; CHECK-LABEL: buildvec_vid_step2_add0_v4i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vadd.vv v8, v8, v8 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 0, i8 2, i8 4, i8 6> |
| } |
| |
| define <4 x i8> @buildvec_vid_step2_add0_v4i8_undef0() { |
| ; CHECK-LABEL: buildvec_vid_step2_add0_v4i8_undef0: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vadd.vv v8, v8, v8 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 poison, i8 2, i8 4, i8 6> |
| } |
| |
| define <4 x i8> @buildvec_vid_step2_add0_v4i8_undef1() { |
| ; CHECK-LABEL: buildvec_vid_step2_add0_v4i8_undef1: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vadd.vv v8, v8, v8 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 poison, i8 poison, i8 4, i8 6> |
| } |
| |
| define <4 x i8> @buildvec_vid_step2_add0_v4i8_undef2() { |
| ; CHECK-LABEL: buildvec_vid_step2_add0_v4i8_undef2: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vadd.vv v8, v8, v8 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 0, i8 poison, i8 poison, i8 6> |
| } |
| |
| define <4 x i8> @buildvec_vid_step2_add1_v4i8() { |
| ; CHECK-LABEL: buildvec_vid_step2_add1_v4i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vadd.vv v8, v8, v8 |
| ; CHECK-NEXT: vadd.vi v8, v8, 1 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 1, i8 3, i8 5, i8 7> |
| } |
| |
| define <4 x i8> @buildvec_vid_step2_add1_v4i8_undef0() { |
| ; CHECK-LABEL: buildvec_vid_step2_add1_v4i8_undef0: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vadd.vv v8, v8, v8 |
| ; CHECK-NEXT: vadd.vi v8, v8, 1 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 poison, i8 3, i8 5, i8 7> |
| } |
| |
| define <4 x i8> @buildvec_vid_step2_add1_v4i8_undef1() { |
| ; CHECK-LABEL: buildvec_vid_step2_add1_v4i8_undef1: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vadd.vv v8, v8, v8 |
| ; CHECK-NEXT: vadd.vi v8, v8, 1 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 poison, i8 poison, i8 5, i8 7> |
| } |
| |
| define <4 x i8> @buildvec_vid_step2_add1_v4i8_undef2() { |
| ; CHECK-LABEL: buildvec_vid_step2_add1_v4i8_undef2: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vadd.vv v8, v8, v8 |
| ; CHECK-NEXT: vadd.vi v8, v8, 1 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 1, i8 poison, i8 poison, i8 7> |
| } |
| |
| define <4 x i8> @buildvec_vid_stepn1_add0_v4i8() { |
| ; CHECK-LABEL: buildvec_vid_stepn1_add0_v4i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vrsub.vi v8, v8, 0 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 0, i8 -1, i8 -2, i8 -3> |
| } |
| |
| define <4 x i8> @buildvec_vid_stepn1_add0_v4i8_undef0() { |
| ; CHECK-LABEL: buildvec_vid_stepn1_add0_v4i8_undef0: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vrsub.vi v8, v8, 0 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 poison, i8 -1, i8 -2, i8 -3> |
| } |
| |
| define <4 x i8> @buildvec_vid_stepn1_add0_v4i8_undef1() { |
| ; CHECK-LABEL: buildvec_vid_stepn1_add0_v4i8_undef1: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vrsub.vi v8, v8, 0 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 poison, i8 poison, i8 -2, i8 -3> |
| } |
| |
| define <4 x i8> @buildvec_vid_stepn1_add0_v4i8_undef2() { |
| ; CHECK-LABEL: buildvec_vid_stepn1_add0_v4i8_undef2: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vrsub.vi v8, v8, 0 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 0, i8 poison, i8 poison, i8 -3> |
| } |
| |
| define <4 x i8> @buildvec_vid_stepn2_add0_v4i8() { |
| ; CHECK-LABEL: buildvec_vid_stepn2_add0_v4i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vadd.vv v8, v8, v8 |
| ; CHECK-NEXT: vrsub.vi v8, v8, 0 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 0, i8 -2, i8 -4, i8 -6> |
| } |
| |
| define <4 x i8> @buildvec_vid_stepn2_add0_v4i8_undef0() { |
| ; CHECK-LABEL: buildvec_vid_stepn2_add0_v4i8_undef0: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vadd.vv v8, v8, v8 |
| ; CHECK-NEXT: vrsub.vi v8, v8, 0 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 poison, i8 -2, i8 -4, i8 -6> |
| } |
| |
| define <4 x i8> @buildvec_vid_stepn2_add0_v4i8_undef1() { |
| ; CHECK-LABEL: buildvec_vid_stepn2_add0_v4i8_undef1: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vadd.vv v8, v8, v8 |
| ; CHECK-NEXT: vrsub.vi v8, v8, 0 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 poison, i8 poison, i8 -4, i8 -6> |
| } |
| |
| define <4 x i8> @buildvec_vid_stepn2_add0_v4i8_undef2() { |
| ; CHECK-LABEL: buildvec_vid_stepn2_add0_v4i8_undef2: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vmv.v.i v8, -6 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 poison, i8 poison, i8 poison, i8 -6> |
| } |
| |
| define <4 x i8> @buildvec_vid_stepn2_add3_v4i8() { |
| ; CHECK-LABEL: buildvec_vid_stepn2_add3_v4i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vadd.vv v8, v8, v8 |
| ; CHECK-NEXT: vrsub.vi v8, v8, 3 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 3, i8 1, i8 -1, i8 -3> |
| } |
| |
| define <4 x i8> @buildvec_vid_stepn3_add3_v4i8() { |
| ; CHECK-LABEL: buildvec_vid_stepn3_add3_v4i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vmv.v.i v9, 3 |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: li a0, -3 |
| ; CHECK-NEXT: vmadd.vx v8, a0, v9 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 3, i8 0, i8 -3, i8 -6> |
| } |
| |
| define void @buildvec_vid_stepn3_addn3_v4i32(ptr %z0, ptr %z1, ptr %z2, ptr %z3) { |
| ; CHECK-LABEL: buildvec_vid_stepn3_addn3_v4i32: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; CHECK-NEXT: vmv.v.i v8, -3 |
| ; CHECK-NEXT: vid.v v9 |
| ; CHECK-NEXT: li a4, -3 |
| ; CHECK-NEXT: vmadd.vx v9, a4, v8 |
| ; CHECK-NEXT: vse32.v v9, (a0) |
| ; CHECK-NEXT: vse32.v v9, (a1) |
| ; CHECK-NEXT: vse32.v v9, (a2) |
| ; CHECK-NEXT: vse32.v v9, (a3) |
| ; CHECK-NEXT: ret |
| store <4 x i32> <i32 -3, i32 -6, i32 -9, i32 -12>, ptr %z0 |
| store <4 x i32> <i32 poison, i32 -6, i32 -9, i32 -12>, ptr %z1 |
| store <4 x i32> <i32 poison, i32 poison, i32 -9, i32 -12>, ptr %z2 |
| store <4 x i32> <i32 -3, i32 poison, i32 poison, i32 -12>, ptr %z3 |
| ret void |
| } |
| |
| define <4 x i64> @buildvec_vid_step1_add0_v4i64() { |
| ; RV32-LABEL: buildvec_vid_step1_add0_v4i64: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: vsetivli zero, 4, e64, m2, ta, ma |
| ; RV32-NEXT: vid.v v8 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: buildvec_vid_step1_add0_v4i64: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: vsetivli zero, 4, e64, m2, ta, ma |
| ; RV64V-NEXT: vid.v v8 |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_vid_step1_add0_v4i64: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: li a1, 3 |
| ; RV64ZVE32-NEXT: li a2, 2 |
| ; RV64ZVE32-NEXT: li a3, 1 |
| ; RV64ZVE32-NEXT: sd zero, 0(a0) |
| ; RV64ZVE32-NEXT: sd a3, 8(a0) |
| ; RV64ZVE32-NEXT: sd a2, 16(a0) |
| ; RV64ZVE32-NEXT: sd a1, 24(a0) |
| ; RV64ZVE32-NEXT: ret |
| ret <4 x i64> <i64 0, i64 1, i64 2, i64 3> |
| } |
| |
| define <4 x i64> @buildvec_vid_step2_add0_v4i64() { |
| ; RV32-LABEL: buildvec_vid_step2_add0_v4i64: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: vsetivli zero, 4, e64, m2, ta, ma |
| ; RV32-NEXT: vid.v v8 |
| ; RV32-NEXT: vadd.vv v8, v8, v8 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: buildvec_vid_step2_add0_v4i64: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: vsetivli zero, 4, e64, m2, ta, ma |
| ; RV64V-NEXT: vid.v v8 |
| ; RV64V-NEXT: vadd.vv v8, v8, v8 |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_vid_step2_add0_v4i64: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: li a1, 6 |
| ; RV64ZVE32-NEXT: li a2, 4 |
| ; RV64ZVE32-NEXT: li a3, 2 |
| ; RV64ZVE32-NEXT: sd zero, 0(a0) |
| ; RV64ZVE32-NEXT: sd a3, 8(a0) |
| ; RV64ZVE32-NEXT: sd a2, 16(a0) |
| ; RV64ZVE32-NEXT: sd a1, 24(a0) |
| ; RV64ZVE32-NEXT: ret |
| ret <4 x i64> <i64 0, i64 2, i64 4, i64 6> |
| } |
| |
| define <4 x i8> @buildvec_no_vid_v4i8_0() { |
| ; CHECK-LABEL: buildvec_no_vid_v4i8_0: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: lui a0, 28768 |
| ; CHECK-NEXT: addi a0, a0, 769 |
| ; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; CHECK-NEXT: vmv.s.x v8, a0 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 1, i8 3, i8 6, i8 7> |
| } |
| |
| define <4 x i8> @buildvec_no_vid_v4i8_1() { |
| ; CHECK-LABEL: buildvec_no_vid_v4i8_1: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: lui a0, 28752 |
| ; CHECK-NEXT: addi a0, a0, 512 |
| ; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; CHECK-NEXT: vmv.s.x v8, a0 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 poison, i8 2, i8 5, i8 7> |
| } |
| |
| define <4 x i8> @buildvec_no_vid_v4i8_2() { |
| ; CHECK-LABEL: buildvec_no_vid_v4i8_2: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: lui a0, 32768 |
| ; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; CHECK-NEXT: vmv.s.x v8, a0 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 0, i8 poison, i8 poison, i8 8> |
| } |
| |
| define <4 x i8> @buildvec_no_vid_v4i8_3() { |
| ; CHECK-LABEL: buildvec_no_vid_v4i8_3: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: lui a0, 28672 |
| ; CHECK-NEXT: addi a0, a0, 255 |
| ; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; CHECK-NEXT: vmv.s.x v8, a0 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 -1, i8 poison, i8 poison, i8 7> |
| } |
| |
| define <4 x i8> @buildvec_no_vid_v4i8_4() { |
| ; CHECK-LABEL: buildvec_no_vid_v4i8_4: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vmv.v.i v8, -2 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 -2, i8 poison, i8 poison, i8 poison> |
| } |
| |
| define <4 x i8> @buildvec_no_vid_v4i8_5() { |
| ; CHECK-LABEL: buildvec_no_vid_v4i8_5: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: lui a0, 1032144 |
| ; CHECK-NEXT: addi a0, a0, -257 |
| ; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; CHECK-NEXT: vmv.s.x v8, a0 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 -1, i8 -2, i8 -4, i8 -5> |
| } |
| |
| define void @buildvec_dominant0_v8i16(ptr %x) { |
| ; CHECK-LABEL: buildvec_dominant0_v8i16: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma |
| ; CHECK-NEXT: vmv.s.x v8, zero |
| ; CHECK-NEXT: vmv.v.i v9, 8 |
| ; CHECK-NEXT: vsetivli zero, 4, e16, m1, tu, ma |
| ; CHECK-NEXT: vslideup.vi v9, v8, 3 |
| ; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma |
| ; CHECK-NEXT: vse16.v v9, (a0) |
| ; CHECK-NEXT: ret |
| store <8 x i16> <i16 8, i16 8, i16 poison, i16 0, i16 8, i16 poison, i16 8, i16 8>, ptr %x |
| ret void |
| } |
| |
| define void @buildvec_dominant0_v8i16_with_end_element(ptr %x) { |
| ; CHECK-LABEL: buildvec_dominant0_v8i16_with_end_element: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma |
| ; CHECK-NEXT: vmv.v.i v8, 8 |
| ; CHECK-NEXT: li a1, 3 |
| ; CHECK-NEXT: vslide1down.vx v8, v8, a1 |
| ; CHECK-NEXT: vse16.v v8, (a0) |
| ; CHECK-NEXT: ret |
| store <8 x i16> <i16 8, i16 8, i16 8, i16 8, i16 8, i16 8, i16 8, i16 3>, ptr %x |
| ret void |
| } |
| |
| define void @buildvec_dominant0_v8i16_with_tail(ptr %x) { |
| ; CHECK-LABEL: buildvec_dominant0_v8i16_with_tail: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: lui a1, %hi(.LCPI35_0) |
| ; CHECK-NEXT: addi a1, a1, %lo(.LCPI35_0) |
| ; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma |
| ; CHECK-NEXT: vle16.v v8, (a1) |
| ; CHECK-NEXT: vse16.v v8, (a0) |
| ; CHECK-NEXT: ret |
| store <8 x i16> <i16 8, i16 8, i16 8, i16 8, i16 8, i16 poison, i16 2, i16 3>, ptr %x |
| ret void |
| } |
| |
| |
| define void @buildvec_dominant1_v8i16(ptr %x) { |
| ; CHECK-LABEL: buildvec_dominant1_v8i16: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma |
| ; CHECK-NEXT: vmv.v.i v8, 8 |
| ; CHECK-NEXT: vse16.v v8, (a0) |
| ; CHECK-NEXT: ret |
| store <8 x i16> <i16 poison, i16 8, i16 poison, i16 poison, i16 poison, i16 poison, i16 poison, i16 poison>, ptr %x |
| ret void |
| } |
| |
| define <2 x i8> @buildvec_dominant0_v2i8() { |
| ; CHECK-LABEL: buildvec_dominant0_v2i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: ret |
| ret <2 x i8> <i8 poison, i8 poison> |
| } |
| |
| define <2 x i8> @buildvec_dominant1_v2i8() { |
| ; RV32-LABEL: buildvec_dominant1_v2i8: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: vsetivli zero, 2, e8, mf8, ta, ma |
| ; RV32-NEXT: vmv.v.i v8, -1 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: buildvec_dominant1_v2i8: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: vsetivli zero, 2, e8, mf8, ta, ma |
| ; RV64V-NEXT: vmv.v.i v8, -1 |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_dominant1_v2i8: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma |
| ; RV64ZVE32-NEXT: vmv.v.i v8, -1 |
| ; RV64ZVE32-NEXT: ret |
| ret <2 x i8> <i8 poison, i8 -1> |
| } |
| |
| define <2 x i8> @buildvec_dominant2_v2i8() { |
| ; RV32-LABEL: buildvec_dominant2_v2i8: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: vsetivli zero, 2, e8, mf8, ta, ma |
| ; RV32-NEXT: vid.v v8 |
| ; RV32-NEXT: vrsub.vi v8, v8, 0 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: buildvec_dominant2_v2i8: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: vsetivli zero, 2, e8, mf8, ta, ma |
| ; RV64V-NEXT: vid.v v8 |
| ; RV64V-NEXT: vrsub.vi v8, v8, 0 |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_dominant2_v2i8: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: vsetivli zero, 2, e8, mf4, ta, ma |
| ; RV64ZVE32-NEXT: vid.v v8 |
| ; RV64ZVE32-NEXT: vrsub.vi v8, v8, 0 |
| ; RV64ZVE32-NEXT: ret |
| ret <2 x i8> <i8 0, i8 -1> |
| } |
| |
| define void @buildvec_dominant0_v2i32(ptr %x) { |
| ; RV32-LABEL: buildvec_dominant0_v2i32: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: lui a1, %hi(.LCPI40_0) |
| ; RV32-NEXT: addi a1, a1, %lo(.LCPI40_0) |
| ; RV32-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32-NEXT: vle32.v v8, (a1) |
| ; RV32-NEXT: vse32.v v8, (a0) |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: buildvec_dominant0_v2i32: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: lui a1, %hi(.LCPI40_0) |
| ; RV64V-NEXT: ld a1, %lo(.LCPI40_0)(a1) |
| ; RV64V-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RV64V-NEXT: vmv.v.i v8, -1 |
| ; RV64V-NEXT: vsetvli zero, zero, e64, m1, tu, ma |
| ; RV64V-NEXT: vmv.s.x v8, a1 |
| ; RV64V-NEXT: vse64.v v8, (a0) |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_dominant0_v2i32: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: lui a1, %hi(.LCPI40_0) |
| ; RV64ZVE32-NEXT: ld a1, %lo(.LCPI40_0)(a1) |
| ; RV64ZVE32-NEXT: li a2, -1 |
| ; RV64ZVE32-NEXT: sd a1, 0(a0) |
| ; RV64ZVE32-NEXT: sd a2, 8(a0) |
| ; RV64ZVE32-NEXT: ret |
| store <2 x i64> <i64 2049638230412172402, i64 -1>, ptr %x |
| ret void |
| } |
| |
| define void @buildvec_dominant1_optsize_v2i32(ptr %x) optsize { |
| ; RV32-LABEL: buildvec_dominant1_optsize_v2i32: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: lui a1, %hi(.LCPI41_0) |
| ; RV32-NEXT: addi a1, a1, %lo(.LCPI41_0) |
| ; RV32-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32-NEXT: vle32.v v8, (a1) |
| ; RV32-NEXT: vse32.v v8, (a0) |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: buildvec_dominant1_optsize_v2i32: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: lui a1, %hi(.LCPI41_0) |
| ; RV64V-NEXT: addi a1, a1, %lo(.LCPI41_0) |
| ; RV64V-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RV64V-NEXT: vle64.v v8, (a1) |
| ; RV64V-NEXT: vse64.v v8, (a0) |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_dominant1_optsize_v2i32: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: lui a1, %hi(.LCPI41_0) |
| ; RV64ZVE32-NEXT: ld a1, %lo(.LCPI41_0)(a1) |
| ; RV64ZVE32-NEXT: li a2, -1 |
| ; RV64ZVE32-NEXT: sd a1, 0(a0) |
| ; RV64ZVE32-NEXT: sd a2, 8(a0) |
| ; RV64ZVE32-NEXT: ret |
| store <2 x i64> <i64 2049638230412172402, i64 -1>, ptr %x |
| ret void |
| } |
| |
| define void @buildvec_seq_v8i8_v4i16(ptr %x) { |
| ; CHECK-LABEL: buildvec_seq_v8i8_v4i16: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: li a1, 513 |
| ; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma |
| ; CHECK-NEXT: vmv.v.x v8, a1 |
| ; CHECK-NEXT: vse8.v v8, (a0) |
| ; CHECK-NEXT: ret |
| store <8 x i8> <i8 1, i8 2, i8 1, i8 2, i8 1, i8 2, i8 poison, i8 2>, ptr %x |
| ret void |
| } |
| |
| define void @buildvec_seq_v8i8_v2i32(ptr %x) { |
| ; RV32-LABEL: buildvec_seq_v8i8_v2i32: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: lui a1, 48 |
| ; RV32-NEXT: addi a1, a1, 513 |
| ; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma |
| ; RV32-NEXT: vmv.v.x v8, a1 |
| ; RV32-NEXT: vsetivli zero, 8, e8, mf2, ta, ma |
| ; RV32-NEXT: vse8.v v8, (a0) |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: buildvec_seq_v8i8_v2i32: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: lui a1, 48 |
| ; RV64V-NEXT: addi a1, a1, 513 |
| ; RV64V-NEXT: vsetivli zero, 2, e32, mf2, ta, ma |
| ; RV64V-NEXT: vmv.v.x v8, a1 |
| ; RV64V-NEXT: vsetivli zero, 8, e8, mf2, ta, ma |
| ; RV64V-NEXT: vse8.v v8, (a0) |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_seq_v8i8_v2i32: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: lui a1, 48 |
| ; RV64ZVE32-NEXT: addi a1, a1, 513 |
| ; RV64ZVE32-NEXT: vsetivli zero, 2, e32, m1, ta, ma |
| ; RV64ZVE32-NEXT: vmv.v.x v8, a1 |
| ; RV64ZVE32-NEXT: vsetivli zero, 8, e8, mf2, ta, ma |
| ; RV64ZVE32-NEXT: vse8.v v8, (a0) |
| ; RV64ZVE32-NEXT: ret |
| store <8 x i8> <i8 1, i8 2, i8 3, i8 poison, i8 1, i8 2, i8 3, i8 poison>, ptr %x |
| ret void |
| } |
| |
| define void @buildvec_seq_v16i8_v2i64(ptr %x) { |
| ; RV32-LABEL: buildvec_seq_v16i8_v2i64: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: lui a1, %hi(.LCPI44_0) |
| ; RV32-NEXT: addi a1, a1, %lo(.LCPI44_0) |
| ; RV32-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; RV32-NEXT: vle8.v v8, (a1) |
| ; RV32-NEXT: vse8.v v8, (a0) |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: buildvec_seq_v16i8_v2i64: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: lui a1, %hi(.LCPI44_0) |
| ; RV64V-NEXT: ld a1, %lo(.LCPI44_0)(a1) |
| ; RV64V-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RV64V-NEXT: vmv.v.x v8, a1 |
| ; RV64V-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; RV64V-NEXT: vse8.v v8, (a0) |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_seq_v16i8_v2i64: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: lui a1, %hi(.LCPI44_0) |
| ; RV64ZVE32-NEXT: addi a1, a1, %lo(.LCPI44_0) |
| ; RV64ZVE32-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; RV64ZVE32-NEXT: vle8.v v8, (a1) |
| ; RV64ZVE32-NEXT: vse8.v v8, (a0) |
| ; RV64ZVE32-NEXT: ret |
| store <16 x i8> <i8 1, i8 2, i8 3, i8 4, i8 5, i8 6, i8 7, i8 8, i8 1, i8 2, i8 3, i8 4, i8 5, i8 6, i8 7, i8 8>, ptr %x |
| ret void |
| } |
| |
| define void @buildvec_seq2_v16i8_v2i64(ptr %x) { |
| ; RV32-LABEL: buildvec_seq2_v16i8_v2i64: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: lui a1, 528432 |
| ; RV32-NEXT: addi a1, a1, 513 |
| ; RV32-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RV32-NEXT: vmv.v.x v8, a1 |
| ; RV32-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; RV32-NEXT: vse8.v v8, (a0) |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: buildvec_seq2_v16i8_v2i64: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: lui a1, 528432 |
| ; RV64V-NEXT: addi a1, a1, 513 |
| ; RV64V-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RV64V-NEXT: vmv.v.x v8, a1 |
| ; RV64V-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; RV64V-NEXT: vse8.v v8, (a0) |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_seq2_v16i8_v2i64: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: lui a1, %hi(.LCPI45_0) |
| ; RV64ZVE32-NEXT: addi a1, a1, %lo(.LCPI45_0) |
| ; RV64ZVE32-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; RV64ZVE32-NEXT: vle8.v v8, (a1) |
| ; RV64ZVE32-NEXT: vse8.v v8, (a0) |
| ; RV64ZVE32-NEXT: ret |
| store <16 x i8> <i8 1, i8 2, i8 3, i8 129, i8 -1, i8 -1, i8 -1, i8 -1, i8 1, i8 2, i8 3, i8 129, i8 -1, i8 -1, i8 -1, i8 -1>, ptr %x |
| ret void |
| } |
| |
| define void @buildvec_seq_v9i8(ptr %x) { |
| ; CHECK-LABEL: buildvec_seq_v9i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: li a1, 73 |
| ; CHECK-NEXT: vsetivli zero, 9, e8, m1, ta, ma |
| ; CHECK-NEXT: vmv.v.i v8, 3 |
| ; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma |
| ; CHECK-NEXT: vmv.s.x v0, a1 |
| ; CHECK-NEXT: li a1, 146 |
| ; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, ma |
| ; CHECK-NEXT: vmerge.vim v8, v8, 1, v0 |
| ; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, ma |
| ; CHECK-NEXT: vmv.s.x v0, a1 |
| ; CHECK-NEXT: vsetvli zero, zero, e8, m1, ta, ma |
| ; CHECK-NEXT: vmerge.vim v8, v8, 2, v0 |
| ; CHECK-NEXT: vse8.v v8, (a0) |
| ; CHECK-NEXT: ret |
| store <9 x i8> <i8 1, i8 2, i8 3, i8 1, i8 2, i8 3, i8 1, i8 2, i8 3>, ptr %x |
| ret void |
| } |
| |
| define void @buildvec_seq_v4i16_v2i32(ptr %x) { |
| ; CHECK-LABEL: buildvec_seq_v4i16_v2i32: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: li a1, -127 |
| ; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; CHECK-NEXT: vmv.v.x v8, a1 |
| ; CHECK-NEXT: vse16.v v8, (a0) |
| ; CHECK-NEXT: ret |
| store <4 x i16> <i16 -127, i16 -1, i16 -127, i16 -1>, ptr %x |
| ret void |
| } |
| |
| define void @buildvec_vid_step1o2_v4i32(ptr %z0, ptr %z1, ptr %z2, ptr %z3, ptr %z4, ptr %z5, ptr %z6) { |
| ; CHECK-LABEL: buildvec_vid_step1o2_v4i32: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; CHECK-NEXT: vmv.v.i v8, 1 |
| ; CHECK-NEXT: vmv.s.x v9, zero |
| ; CHECK-NEXT: vsetivli zero, 2, e32, m1, tu, ma |
| ; CHECK-NEXT: vslideup.vi v8, v9, 1 |
| ; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; CHECK-NEXT: vid.v v9 |
| ; CHECK-NEXT: vsrl.vi v9, v9, 1 |
| ; CHECK-NEXT: vse32.v v9, (a0) |
| ; CHECK-NEXT: vse32.v v9, (a1) |
| ; CHECK-NEXT: vse32.v v9, (a2) |
| ; CHECK-NEXT: vse32.v v9, (a3) |
| ; CHECK-NEXT: vse32.v v9, (a4) |
| ; CHECK-NEXT: vmv.v.i v9, 0 |
| ; CHECK-NEXT: li a0, 1 |
| ; CHECK-NEXT: vslide1down.vx v9, v9, a0 |
| ; CHECK-NEXT: vse32.v v8, (a5) |
| ; CHECK-NEXT: vse32.v v9, (a6) |
| ; CHECK-NEXT: ret |
| store <4 x i32> <i32 0, i32 0, i32 1, i32 1>, ptr %z0 |
| store <4 x i32> <i32 0, i32 0, i32 1, i32 poison>, ptr %z1 |
| store <4 x i32> <i32 0, i32 poison, i32 1, i32 1>, ptr %z2 |
| store <4 x i32> <i32 poison, i32 0, i32 poison, i32 1>, ptr %z3 |
| store <4 x i32> <i32 0, i32 poison, i32 1, i32 poison>, ptr %z4 |
| ; We don't catch this one |
| store <4 x i32> <i32 poison, i32 0, i32 1, i32 1>, ptr %z5 |
| ; We catch this one but as VID/3 rather than VID/2 |
| store <4 x i32> <i32 0, i32 0, i32 poison, i32 1>, ptr %z6 |
| ret void |
| } |
| |
| define void @buildvec_vid_step1o2_add3_v4i16(ptr %z0, ptr %z1, ptr %z2, ptr %z3, ptr %z4, ptr %z5, ptr %z6) { |
| ; CHECK-LABEL: buildvec_vid_step1o2_add3_v4i16: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vmv.v.i v9, 3 |
| ; CHECK-NEXT: vsrl.vi v8, v8, 1 |
| ; CHECK-NEXT: vadd.vi v8, v8, 3 |
| ; CHECK-NEXT: vse16.v v8, (a0) |
| ; CHECK-NEXT: vse16.v v8, (a1) |
| ; CHECK-NEXT: vse16.v v8, (a2) |
| ; CHECK-NEXT: vse16.v v8, (a3) |
| ; CHECK-NEXT: vse16.v v8, (a4) |
| ; CHECK-NEXT: vmv.v.i v8, 4 |
| ; CHECK-NEXT: li a0, 4 |
| ; CHECK-NEXT: vsetivli zero, 2, e16, mf2, tu, ma |
| ; CHECK-NEXT: vslideup.vi v8, v9, 1 |
| ; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma |
| ; CHECK-NEXT: vslide1down.vx v9, v9, a0 |
| ; CHECK-NEXT: vse16.v v8, (a5) |
| ; CHECK-NEXT: vse16.v v9, (a6) |
| ; CHECK-NEXT: ret |
| store <4 x i16> <i16 3, i16 3, i16 4, i16 4>, ptr %z0 |
| store <4 x i16> <i16 3, i16 3, i16 4, i16 poison>, ptr %z1 |
| store <4 x i16> <i16 3, i16 poison, i16 4, i16 4>, ptr %z2 |
| store <4 x i16> <i16 poison, i16 3, i16 poison, i16 4>, ptr %z3 |
| store <4 x i16> <i16 3, i16 poison, i16 4, i16 poison>, ptr %z4 |
| ; We don't catch this one |
| store <4 x i16> <i16 poison, i16 3, i16 4, i16 4>, ptr %z5 |
| ; We catch this one but as VID/3 rather than VID/2 |
| store <4 x i16> <i16 3, i16 3, i16 poison, i16 4>, ptr %z6 |
| ret void |
| } |
| |
| define void @buildvec_vid_stepn1o4_addn5_v8i8(ptr %z0) { |
| ; CHECK-LABEL: buildvec_vid_stepn1o4_addn5_v8i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 8, e8, mf2, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vsrl.vi v8, v8, 2 |
| ; CHECK-NEXT: vrsub.vi v8, v8, -5 |
| ; CHECK-NEXT: vse8.v v8, (a0) |
| ; CHECK-NEXT: ret |
| store <8 x i8> <i8 -5, i8 -5, i8 -5, i8 -5, i8 -6, i8 -6, i8 -6, i8 -6>, ptr %z0 |
| ret void |
| } |
| |
| define void @buildvec_vid_mpy_imm_v8i16(ptr %x) { |
| ; CHECK-LABEL: buildvec_vid_mpy_imm_v8i16: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: li a1, 17 |
| ; CHECK-NEXT: vmul.vx v8, v8, a1 |
| ; CHECK-NEXT: vse16.v v8, (a0) |
| ; CHECK-NEXT: ret |
| store <8 x i16> <i16 0, i16 17, i16 34, i16 51, i16 68, i16 85, i16 102, i16 119>, ptr %x |
| ret void |
| } |
| |
| define void @buildvec_vid_shl_imm_v8i16(ptr %x) { |
| ; CHECK-LABEL: buildvec_vid_shl_imm_v8i16: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: vsll.vi v8, v8, 9 |
| ; CHECK-NEXT: vse16.v v8, (a0) |
| ; CHECK-NEXT: ret |
| store <8 x i16> <i16 0, i16 512, i16 1024, i16 1536, i16 2048, i16 2560, i16 3072, i16 3584>, ptr %x |
| ret void |
| } |
| |
| define <4 x i32> @splat_c3_v4i32(<4 x i32> %v) { |
| ; CHECK-LABEL: splat_c3_v4i32: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; CHECK-NEXT: vrgather.vi v9, v8, 3 |
| ; CHECK-NEXT: vmv.v.v v8, v9 |
| ; CHECK-NEXT: ret |
| %x = extractelement <4 x i32> %v, i32 3 |
| %ins = insertelement <4 x i32> poison, i32 %x, i32 0 |
| %splat = shufflevector <4 x i32> %ins, <4 x i32> poison, <4 x i32> zeroinitializer |
| ret <4 x i32> %splat |
| } |
| |
| define <4 x i32> @splat_idx_v4i32(<4 x i32> %v, i64 %idx) { |
| ; CHECK-LABEL: splat_idx_v4i32: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; CHECK-NEXT: vrgather.vx v9, v8, a0 |
| ; CHECK-NEXT: vmv.v.v v8, v9 |
| ; CHECK-NEXT: ret |
| %x = extractelement <4 x i32> %v, i64 %idx |
| %ins = insertelement <4 x i32> poison, i32 %x, i32 0 |
| %splat = shufflevector <4 x i32> %ins, <4 x i32> poison, <4 x i32> zeroinitializer |
| ret <4 x i32> %splat |
| } |
| |
| define <8 x i16> @splat_c4_v8i16(<8 x i16> %v) { |
| ; CHECK-LABEL: splat_c4_v8i16: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma |
| ; CHECK-NEXT: vrgather.vi v9, v8, 4 |
| ; CHECK-NEXT: vmv.v.v v8, v9 |
| ; CHECK-NEXT: ret |
| %x = extractelement <8 x i16> %v, i32 4 |
| %ins = insertelement <8 x i16> poison, i16 %x, i32 0 |
| %splat = shufflevector <8 x i16> %ins, <8 x i16> poison, <8 x i32> zeroinitializer |
| ret <8 x i16> %splat |
| } |
| |
| define <8 x i16> @splat_idx_v8i16(<8 x i16> %v, i64 %idx) { |
| ; CHECK-LABEL: splat_idx_v8i16: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma |
| ; CHECK-NEXT: vrgather.vx v9, v8, a0 |
| ; CHECK-NEXT: vmv.v.v v8, v9 |
| ; CHECK-NEXT: ret |
| %x = extractelement <8 x i16> %v, i64 %idx |
| %ins = insertelement <8 x i16> poison, i16 %x, i32 0 |
| %splat = shufflevector <8 x i16> %ins, <8 x i16> poison, <8 x i32> zeroinitializer |
| ret <8 x i16> %splat |
| } |
| |
| define <4 x i8> @buildvec_not_vid_v4i8_1() { |
| ; CHECK-LABEL: buildvec_not_vid_v4i8_1: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: lui a0, 12320 |
| ; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; CHECK-NEXT: vmv.s.x v8, a0 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 0, i8 0, i8 2, i8 3> |
| } |
| |
| define <4 x i8> @buildvec_not_vid_v4i8_2() { |
| ; CHECK-LABEL: buildvec_not_vid_v4i8_2: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: lui a0, 16 |
| ; CHECK-NEXT: addi a0, a0, 771 |
| ; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; CHECK-NEXT: vmv.s.x v8, a0 |
| ; CHECK-NEXT: ret |
| ret <4 x i8> <i8 3, i8 3, i8 1, i8 0> |
| } |
| |
| ; We match this as a VID sequence (-3 / 8) + 5 but choose not to introduce |
| ; division to compute it. |
| define <16 x i8> @buildvec_not_vid_v16i8() { |
| ; CHECK-LABEL: buildvec_not_vid_v16i8: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 7, e8, m1, ta, ma |
| ; CHECK-NEXT: vmv.v.i v9, 3 |
| ; CHECK-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; CHECK-NEXT: vmv.v.i v8, 0 |
| ; CHECK-NEXT: vsetivli zero, 7, e8, m1, tu, ma |
| ; CHECK-NEXT: vslideup.vi v8, v9, 6 |
| ; CHECK-NEXT: ret |
| ret <16 x i8> <i8 poison, i8 poison, i8 poison, i8 poison, i8 poison, i8 poison, i8 3, i8 poison, i8 poison, i8 poison, i8 poison, i8 poison, i8 poison, i8 poison, i8 0, i8 0> |
| } |
| |
| define <512 x i8> @buildvec_vid_v512i8_indices_overflow() vscale_range(16, 1024) { |
| ; CHECK-LABEL: buildvec_vid_v512i8_indices_overflow: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: li a0, 512 |
| ; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma |
| ; CHECK-NEXT: vid.v v8 |
| ; CHECK-NEXT: ret |
| ret <512 x i8> <i8 0, i8 1, i8 2, i8 3, i8 4, i8 5, i8 6, i8 7, i8 8, i8 9, i8 10, i8 11, i8 12, i8 13, i8 14, i8 15, i8 16, i8 17, i8 18, i8 19, i8 20, i8 21, i8 22, i8 23, i8 24, i8 25, i8 26, i8 27, i8 28, i8 29, i8 30, i8 31, i8 32, i8 33, i8 34, i8 35, i8 36, i8 37, i8 38, i8 39, i8 40, i8 41, i8 42, i8 43, i8 44, i8 45, i8 46, i8 47, i8 48, i8 49, i8 50, i8 51, i8 52, i8 53, i8 54, i8 55, i8 56, i8 57, i8 58, i8 59, i8 60, i8 61, i8 62, i8 63, i8 64, i8 65, i8 66, i8 67, i8 68, i8 69, i8 70, i8 71, i8 72, i8 73, i8 74, i8 75, i8 76, i8 77, i8 78, i8 79, i8 80, i8 81, i8 82, i8 83, i8 84, i8 85, i8 86, i8 87, i8 88, i8 89, i8 90, i8 91, i8 92, i8 93, i8 94, i8 95, i8 96, i8 97, i8 98, i8 99, i8 100, i8 101, i8 102, i8 103, i8 104, i8 105, i8 106, i8 107, i8 108, i8 109, i8 110, i8 111, i8 112, i8 113, i8 114, i8 115, i8 116, i8 117, i8 118, i8 119, i8 120, i8 121, i8 122, i8 123, i8 124, i8 125, i8 126, i8 127, i8 128, i8 129, i8 130, i8 131, i8 132, i8 133, i8 134, i8 135, i8 136, i8 137, i8 138, i8 139, i8 140, i8 141, i8 142, i8 143, i8 144, i8 145, i8 146, i8 147, i8 148, i8 149, i8 150, i8 151, i8 152, i8 153, i8 154, i8 155, i8 156, i8 157, i8 158, i8 159, i8 160, i8 161, i8 162, i8 163, i8 164, i8 165, i8 166, i8 167, i8 168, i8 169, i8 170, i8 171, i8 172, i8 173, i8 174, i8 175, i8 176, i8 177, i8 178, i8 179, i8 180, i8 181, i8 182, i8 183, i8 184, i8 185, i8 186, i8 187, i8 188, i8 189, i8 190, i8 191, i8 192, i8 193, i8 194, i8 195, i8 196, i8 197, i8 198, i8 199, i8 200, i8 201, i8 202, i8 203, i8 204, i8 205, i8 206, i8 207, i8 208, i8 209, i8 210, i8 211, i8 212, i8 213, i8 214, i8 215, i8 216, i8 217, i8 218, i8 219, i8 220, i8 221, i8 222, i8 223, i8 224, i8 225, i8 226, i8 227, i8 228, i8 229, i8 230, i8 231, i8 232, i8 233, i8 234, i8 235, i8 236, i8 237, i8 238, i8 239, i8 240, i8 241, i8 242, i8 243, i8 244, i8 245, i8 246, i8 247, i8 248, i8 249, i8 250, i8 251, i8 252, i8 253, i8 254, i8 255, i8 0, i8 1, i8 2, i8 3, i8 4, i8 5, i8 6, i8 7, i8 8, i8 9, i8 10, i8 11, i8 12, i8 13, i8 14, i8 15, i8 16, i8 17, i8 18, i8 19, i8 20, i8 21, i8 22, i8 23, i8 24, i8 25, i8 26, i8 27, i8 28, i8 29, i8 30, i8 31, i8 32, i8 33, i8 34, i8 35, i8 36, i8 37, i8 38, i8 39, i8 40, i8 41, i8 42, i8 43, i8 44, i8 45, i8 46, i8 47, i8 48, i8 49, i8 50, i8 51, i8 52, i8 53, i8 54, i8 55, i8 56, i8 57, i8 58, i8 59, i8 60, i8 61, i8 62, i8 63, i8 64, i8 65, i8 66, i8 67, i8 68, i8 69, i8 70, i8 71, i8 72, i8 73, i8 74, i8 75, i8 76, i8 77, i8 78, i8 79, i8 80, i8 81, i8 82, i8 83, i8 84, i8 85, i8 86, i8 87, i8 88, i8 89, i8 90, i8 91, i8 92, i8 93, i8 94, i8 95, i8 96, i8 97, i8 98, i8 99, i8 100, i8 101, i8 102, i8 103, i8 104, i8 105, i8 106, i8 107, i8 108, i8 109, i8 110, i8 111, i8 112, i8 113, i8 114, i8 115, i8 116, i8 117, i8 118, i8 119, i8 120, i8 121, i8 122, i8 123, i8 124, i8 125, i8 126, i8 127, i8 128, i8 129, i8 130, i8 131, i8 132, i8 133, i8 134, i8 135, i8 136, i8 137, i8 138, i8 139, i8 140, i8 141, i8 142, i8 143, i8 144, i8 145, i8 146, i8 147, i8 148, i8 149, i8 150, i8 151, i8 152, i8 153, i8 154, i8 155, i8 156, i8 157, i8 158, i8 159, i8 160, i8 161, i8 162, i8 163, i8 164, i8 165, i8 166, i8 167, i8 168, i8 169, i8 170, i8 171, i8 172, i8 173, i8 174, i8 175, i8 176, i8 177, i8 178, i8 179, i8 180, i8 181, i8 182, i8 183, i8 184, i8 185, i8 186, i8 187, i8 188, i8 189, i8 190, i8 191, i8 192, i8 193, i8 194, i8 195, i8 196, i8 197, i8 198, i8 199, i8 200, i8 201, i8 202, i8 203, i8 204, i8 205, i8 206, i8 207, i8 208, i8 209, i8 210, i8 211, i8 212, i8 213, i8 214, i8 215, i8 216, i8 217, i8 218, i8 219, i8 220, i8 221, i8 222, i8 223, i8 224, i8 225, i8 226, i8 227, i8 228, i8 229, i8 230, i8 231, i8 232, i8 233, i8 234, i8 235, i8 236, i8 237, i8 238, i8 239, i8 240, i8 241, i8 242, i8 243, i8 244, i8 245, i8 246, i8 247, i8 248, i8 249, i8 250, i8 251, i8 252, i8 253, i8 254, i8 255> |
| } |
| |
| define <512 x i8> @buildvec_not_vid_v512i8_indices_overflow_1() vscale_range(16, 1024) { |
| ; RV32-LABEL: buildvec_not_vid_v512i8_indices_overflow_1: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: li a0, 512 |
| ; RV32-NEXT: vsetivli zero, 16, e32, mf2, ta, ma |
| ; RV32-NEXT: vid.v v8 |
| ; RV32-NEXT: vsrl.vi v8, v8, 3 |
| ; RV32-NEXT: vadd.vi v0, v8, -1 |
| ; RV32-NEXT: vsetvli zero, a0, e8, m4, ta, ma |
| ; RV32-NEXT: vmv.v.i v8, 1 |
| ; RV32-NEXT: vmerge.vim v8, v8, 0, v0 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: buildvec_not_vid_v512i8_indices_overflow_1: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: li a0, 512 |
| ; RV64V-NEXT: vsetivli zero, 8, e64, m1, ta, ma |
| ; RV64V-NEXT: vid.v v8 |
| ; RV64V-NEXT: vsrl.vi v8, v8, 2 |
| ; RV64V-NEXT: vadd.vi v0, v8, -1 |
| ; RV64V-NEXT: vsetvli zero, a0, e8, m4, ta, ma |
| ; RV64V-NEXT: vmv.v.i v8, 1 |
| ; RV64V-NEXT: vmerge.vim v8, v8, 0, v0 |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_not_vid_v512i8_indices_overflow_1: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: li a0, 512 |
| ; RV64ZVE32-NEXT: vsetivli zero, 16, e32, m1, ta, ma |
| ; RV64ZVE32-NEXT: vid.v v8 |
| ; RV64ZVE32-NEXT: vsrl.vi v8, v8, 3 |
| ; RV64ZVE32-NEXT: vadd.vi v0, v8, -1 |
| ; RV64ZVE32-NEXT: vsetvli zero, a0, e8, m4, ta, ma |
| ; RV64ZVE32-NEXT: vmv.v.i v8, 1 |
| ; RV64ZVE32-NEXT: vmerge.vim v8, v8, 0, v0 |
| ; RV64ZVE32-NEXT: ret |
| ret <512 x i8> <i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1> |
| } |
| |
| define <512 x i8> @buildvec_not_vid_v512i8_indices_overflow_2() vscale_range(16, 1024) { |
| ; RV32-LABEL: buildvec_not_vid_v512i8_indices_overflow_2: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: vsetivli zero, 16, e32, mf2, ta, ma |
| ; RV32-NEXT: vmv.v.i v0, 15 |
| ; RV32-NEXT: vmv.v.i v12, 0 |
| ; RV32-NEXT: li a0, 512 |
| ; RV32-NEXT: li a1, 240 |
| ; RV32-NEXT: vmerge.vim v13, v12, -1, v0 |
| ; RV32-NEXT: vsetvli zero, a0, e8, m4, ta, ma |
| ; RV32-NEXT: vmv.v.i v8, 3 |
| ; RV32-NEXT: vmv1r.v v0, v13 |
| ; RV32-NEXT: vmerge.vim v8, v8, 0, v0 |
| ; RV32-NEXT: vsetvli zero, zero, e16, m8, ta, ma |
| ; RV32-NEXT: vmv.s.x v0, a1 |
| ; RV32-NEXT: li a1, 15 |
| ; RV32-NEXT: vsetivli zero, 16, e32, mf2, ta, ma |
| ; RV32-NEXT: vmerge.vim v13, v12, -1, v0 |
| ; RV32-NEXT: slli a1, a1, 8 |
| ; RV32-NEXT: vmv1r.v v0, v13 |
| ; RV32-NEXT: vsetvli zero, a0, e8, m4, ta, ma |
| ; RV32-NEXT: vmerge.vim v8, v8, 1, v0 |
| ; RV32-NEXT: vsetvli zero, zero, e16, m8, ta, ma |
| ; RV32-NEXT: vmv.s.x v0, a1 |
| ; RV32-NEXT: vsetivli zero, 16, e32, mf2, ta, ma |
| ; RV32-NEXT: vmerge.vim v12, v12, -1, v0 |
| ; RV32-NEXT: vmv1r.v v0, v12 |
| ; RV32-NEXT: vsetvli zero, a0, e8, m4, ta, ma |
| ; RV32-NEXT: vmerge.vim v8, v8, 2, v0 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: buildvec_not_vid_v512i8_indices_overflow_2: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: vsetivli zero, 8, e64, m1, ta, ma |
| ; RV64V-NEXT: vmv.v.i v0, 3 |
| ; RV64V-NEXT: vmv.v.i v12, 0 |
| ; RV64V-NEXT: li a0, 512 |
| ; RV64V-NEXT: vmerge.vim v13, v12, -1, v0 |
| ; RV64V-NEXT: vsetvli zero, a0, e8, m4, ta, ma |
| ; RV64V-NEXT: vmv.v.i v8, 3 |
| ; RV64V-NEXT: vmv1r.v v0, v13 |
| ; RV64V-NEXT: vmerge.vim v8, v8, 0, v0 |
| ; RV64V-NEXT: vsetivli zero, 8, e64, m1, ta, ma |
| ; RV64V-NEXT: vmv.v.i v0, 12 |
| ; RV64V-NEXT: vmerge.vim v13, v12, -1, v0 |
| ; RV64V-NEXT: li a1, 48 |
| ; RV64V-NEXT: vmv.v.v v0, v13 |
| ; RV64V-NEXT: vsetvli zero, a0, e8, m4, ta, ma |
| ; RV64V-NEXT: vmerge.vim v8, v8, 1, v0 |
| ; RV64V-NEXT: vmv.s.x v0, a1 |
| ; RV64V-NEXT: vsetivli zero, 8, e64, m1, ta, ma |
| ; RV64V-NEXT: vmerge.vim v12, v12, -1, v0 |
| ; RV64V-NEXT: vmv.v.v v0, v12 |
| ; RV64V-NEXT: vsetvli zero, a0, e8, m4, ta, ma |
| ; RV64V-NEXT: vmerge.vim v8, v8, 2, v0 |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_not_vid_v512i8_indices_overflow_2: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: vsetivli zero, 16, e32, m1, ta, ma |
| ; RV64ZVE32-NEXT: vmv.v.i v0, 15 |
| ; RV64ZVE32-NEXT: vmv.v.i v12, 0 |
| ; RV64ZVE32-NEXT: li a0, 512 |
| ; RV64ZVE32-NEXT: li a1, 240 |
| ; RV64ZVE32-NEXT: vmerge.vim v13, v12, -1, v0 |
| ; RV64ZVE32-NEXT: vsetvli zero, a0, e8, m4, ta, ma |
| ; RV64ZVE32-NEXT: vmv.v.i v8, 3 |
| ; RV64ZVE32-NEXT: vmv1r.v v0, v13 |
| ; RV64ZVE32-NEXT: vmerge.vim v8, v8, 0, v0 |
| ; RV64ZVE32-NEXT: vsetvli zero, zero, e16, m8, ta, ma |
| ; RV64ZVE32-NEXT: vmv.s.x v0, a1 |
| ; RV64ZVE32-NEXT: li a1, 15 |
| ; RV64ZVE32-NEXT: vsetivli zero, 16, e32, m1, ta, ma |
| ; RV64ZVE32-NEXT: vmerge.vim v13, v12, -1, v0 |
| ; RV64ZVE32-NEXT: slli a1, a1, 8 |
| ; RV64ZVE32-NEXT: vmv.v.v v0, v13 |
| ; RV64ZVE32-NEXT: vsetvli zero, a0, e8, m4, ta, ma |
| ; RV64ZVE32-NEXT: vmerge.vim v8, v8, 1, v0 |
| ; RV64ZVE32-NEXT: vsetvli zero, zero, e16, m8, ta, ma |
| ; RV64ZVE32-NEXT: vmv.s.x v0, a1 |
| ; RV64ZVE32-NEXT: vsetivli zero, 16, e32, m1, ta, ma |
| ; RV64ZVE32-NEXT: vmerge.vim v12, v12, -1, v0 |
| ; RV64ZVE32-NEXT: vmv.v.v v0, v12 |
| ; RV64ZVE32-NEXT: vsetvli zero, a0, e8, m4, ta, ma |
| ; RV64ZVE32-NEXT: vmerge.vim v8, v8, 2, v0 |
| ; RV64ZVE32-NEXT: ret |
| ret <512 x i8> <i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3, i8 3> |
| } |
| |
| define <8 x i32> @prefix_overwrite(<8 x i32> %vin, i32 %a, i32 %b, i32 %c, i32 %d) { |
| ; CHECK-LABEL: prefix_overwrite: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 8, e32, m1, tu, ma |
| ; CHECK-NEXT: vmv.s.x v10, a1 |
| ; CHECK-NEXT: vmv.s.x v8, a0 |
| ; CHECK-NEXT: vsetivli zero, 2, e32, m1, tu, ma |
| ; CHECK-NEXT: vslideup.vi v8, v10, 1 |
| ; CHECK-NEXT: vmv.s.x v10, a2 |
| ; CHECK-NEXT: vsetivli zero, 3, e32, m1, tu, ma |
| ; CHECK-NEXT: vslideup.vi v8, v10, 2 |
| ; CHECK-NEXT: vmv.s.x v10, a3 |
| ; CHECK-NEXT: vsetivli zero, 4, e32, m1, tu, ma |
| ; CHECK-NEXT: vslideup.vi v8, v10, 3 |
| ; CHECK-NEXT: ret |
| %v0 = insertelement <8 x i32> %vin, i32 %a, i32 0 |
| %v1 = insertelement <8 x i32> %v0, i32 %b, i32 1 |
| %v2 = insertelement <8 x i32> %v1, i32 %c, i32 2 |
| %v3 = insertelement <8 x i32> %v2, i32 %d, i32 3 |
| ret <8 x i32> %v3 |
| } |
| |
| define <8 x i32> @suffix_overwrite(<8 x i32> %vin, i32 %a, i32 %b, i32 %c, i32 %d) { |
| ; CHECK-LABEL: suffix_overwrite: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 5, e32, m2, tu, ma |
| ; CHECK-NEXT: vmv.s.x v10, a0 |
| ; CHECK-NEXT: vslideup.vi v8, v10, 4 |
| ; CHECK-NEXT: vmv.s.x v10, a1 |
| ; CHECK-NEXT: vsetivli zero, 6, e32, m2, tu, ma |
| ; CHECK-NEXT: vslideup.vi v8, v10, 5 |
| ; CHECK-NEXT: vmv.s.x v10, a2 |
| ; CHECK-NEXT: vsetivli zero, 7, e32, m2, tu, ma |
| ; CHECK-NEXT: vslideup.vi v8, v10, 6 |
| ; CHECK-NEXT: vmv.s.x v10, a3 |
| ; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma |
| ; CHECK-NEXT: vslideup.vi v8, v10, 7 |
| ; CHECK-NEXT: ret |
| %v0 = insertelement <8 x i32> %vin, i32 %a, i32 4 |
| %v1 = insertelement <8 x i32> %v0, i32 %b, i32 5 |
| %v2 = insertelement <8 x i32> %v1, i32 %c, i32 6 |
| %v3 = insertelement <8 x i32> %v2, i32 %d, i32 7 |
| ret <8 x i32> %v3 |
| } |
| |
| define <4 x i64> @v4xi64_exact(i64 %a, i64 %b, i64 %c, i64 %d) vscale_range(2,2) { |
| ; RV32-LABEL: v4xi64_exact: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32-NEXT: vmv.v.x v8, a4 |
| ; RV32-NEXT: vmv.v.x v9, a0 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV32-NEXT: vslide1down.vx v10, v9, a1 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a6 |
| ; RV32-NEXT: vslide1down.vx v9, v8, a7 |
| ; RV32-NEXT: vslide1down.vx v8, v10, a2 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: v4xi64_exact: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RV64V-NEXT: vmv.v.x v8, a2 |
| ; RV64V-NEXT: vslide1down.vx v9, v8, a3 |
| ; RV64V-NEXT: vmv.v.x v8, a0 |
| ; RV64V-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: v4xi64_exact: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: sd a1, 0(a0) |
| ; RV64ZVE32-NEXT: sd a2, 8(a0) |
| ; RV64ZVE32-NEXT: sd a3, 16(a0) |
| ; RV64ZVE32-NEXT: sd a4, 24(a0) |
| ; RV64ZVE32-NEXT: ret |
| %v1 = insertelement <4 x i64> poison, i64 %a, i32 0 |
| %v2 = insertelement <4 x i64> %v1, i64 %b, i32 1 |
| %v3 = insertelement <4 x i64> %v2, i64 %c, i32 2 |
| %v4 = insertelement <4 x i64> %v3, i64 %d, i32 3 |
| ret <4 x i64> %v4 |
| } |
| |
| define <8 x i64> @v8xi64_exact(i64 %a, i64 %b, i64 %c, i64 %d, i64 %e, i64 %f, i64 %g, i64 %h) vscale_range(2,2) { |
| ; RV32-LABEL: v8xi64_exact: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: addi sp, sp, -16 |
| ; RV32-NEXT: .cfi_def_cfa_offset 16 |
| ; RV32-NEXT: sw s0, 12(sp) # 4-byte Folded Spill |
| ; RV32-NEXT: .cfi_offset s0, -4 |
| ; RV32-NEXT: lw t0, 44(sp) |
| ; RV32-NEXT: lw t1, 40(sp) |
| ; RV32-NEXT: lw t2, 36(sp) |
| ; RV32-NEXT: lw t3, 32(sp) |
| ; RV32-NEXT: lw t4, 28(sp) |
| ; RV32-NEXT: lw t5, 24(sp) |
| ; RV32-NEXT: lw t6, 20(sp) |
| ; RV32-NEXT: lw s0, 16(sp) |
| ; RV32-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32-NEXT: vmv.v.x v8, a4 |
| ; RV32-NEXT: vmv.v.x v9, a0 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV32-NEXT: vslide1down.vx v9, v9, a1 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a6 |
| ; RV32-NEXT: vslide1down.vx v10, v9, a2 |
| ; RV32-NEXT: vslide1down.vx v9, v8, a7 |
| ; RV32-NEXT: vslide1down.vx v8, v10, a3 |
| ; RV32-NEXT: vmv.v.x v10, s0 |
| ; RV32-NEXT: vslide1down.vx v10, v10, t6 |
| ; RV32-NEXT: vslide1down.vx v10, v10, t5 |
| ; RV32-NEXT: vslide1down.vx v10, v10, t4 |
| ; RV32-NEXT: vmv.v.x v11, t3 |
| ; RV32-NEXT: vslide1down.vx v11, v11, t2 |
| ; RV32-NEXT: vslide1down.vx v11, v11, t1 |
| ; RV32-NEXT: vslide1down.vx v11, v11, t0 |
| ; RV32-NEXT: lw s0, 12(sp) # 4-byte Folded Reload |
| ; RV32-NEXT: .cfi_restore s0 |
| ; RV32-NEXT: addi sp, sp, 16 |
| ; RV32-NEXT: .cfi_def_cfa_offset 0 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: v8xi64_exact: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RV64V-NEXT: vmv.v.x v8, a2 |
| ; RV64V-NEXT: vmv.v.x v10, a0 |
| ; RV64V-NEXT: vslide1down.vx v9, v8, a3 |
| ; RV64V-NEXT: vslide1down.vx v8, v10, a1 |
| ; RV64V-NEXT: vmv.v.x v10, a4 |
| ; RV64V-NEXT: vslide1down.vx v10, v10, a5 |
| ; RV64V-NEXT: vmv.v.x v11, a6 |
| ; RV64V-NEXT: vslide1down.vx v11, v11, a7 |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: v8xi64_exact: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: ld t0, 0(sp) |
| ; RV64ZVE32-NEXT: sd a5, 32(a0) |
| ; RV64ZVE32-NEXT: sd a6, 40(a0) |
| ; RV64ZVE32-NEXT: sd a7, 48(a0) |
| ; RV64ZVE32-NEXT: sd t0, 56(a0) |
| ; RV64ZVE32-NEXT: sd a1, 0(a0) |
| ; RV64ZVE32-NEXT: sd a2, 8(a0) |
| ; RV64ZVE32-NEXT: sd a3, 16(a0) |
| ; RV64ZVE32-NEXT: sd a4, 24(a0) |
| ; RV64ZVE32-NEXT: ret |
| %v1 = insertelement <8 x i64> poison, i64 %a, i32 0 |
| %v2 = insertelement <8 x i64> %v1, i64 %b, i32 1 |
| %v3 = insertelement <8 x i64> %v2, i64 %c, i32 2 |
| %v4 = insertelement <8 x i64> %v3, i64 %d, i32 3 |
| %v5 = insertelement <8 x i64> %v4, i64 %e, i32 4 |
| %v6 = insertelement <8 x i64> %v5, i64 %f, i32 5 |
| %v7 = insertelement <8 x i64> %v6, i64 %g, i32 6 |
| %v8 = insertelement <8 x i64> %v7, i64 %h, i32 7 |
| ret <8 x i64> %v8 |
| } |
| |
| define <8 x i64> @v8xi64_exact_equal_halves(i64 %a, i64 %b, i64 %c, i64 %d) vscale_range(2,2) { |
| ; RV32-LABEL: v8xi64_exact_equal_halves: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32-NEXT: vmv.v.x v8, a4 |
| ; RV32-NEXT: vmv.v.x v9, a0 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV32-NEXT: vslide1down.vx v10, v9, a1 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a6 |
| ; RV32-NEXT: vslide1down.vx v9, v8, a7 |
| ; RV32-NEXT: vslide1down.vx v8, v10, a2 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV32-NEXT: vmv.v.v v10, v8 |
| ; RV32-NEXT: vmv.v.v v11, v9 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: v8xi64_exact_equal_halves: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RV64V-NEXT: vmv.v.x v8, a2 |
| ; RV64V-NEXT: vslide1down.vx v9, v8, a3 |
| ; RV64V-NEXT: vmv.v.x v8, a0 |
| ; RV64V-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV64V-NEXT: vmv.v.v v10, v8 |
| ; RV64V-NEXT: vmv.v.v v11, v9 |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: v8xi64_exact_equal_halves: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: sd a1, 32(a0) |
| ; RV64ZVE32-NEXT: sd a2, 40(a0) |
| ; RV64ZVE32-NEXT: sd a3, 48(a0) |
| ; RV64ZVE32-NEXT: sd a4, 56(a0) |
| ; RV64ZVE32-NEXT: sd a1, 0(a0) |
| ; RV64ZVE32-NEXT: sd a2, 8(a0) |
| ; RV64ZVE32-NEXT: sd a3, 16(a0) |
| ; RV64ZVE32-NEXT: sd a4, 24(a0) |
| ; RV64ZVE32-NEXT: ret |
| %v1 = insertelement <8 x i64> poison, i64 %a, i32 0 |
| %v2 = insertelement <8 x i64> %v1, i64 %b, i32 1 |
| %v3 = insertelement <8 x i64> %v2, i64 %c, i32 2 |
| %v4 = insertelement <8 x i64> %v3, i64 %d, i32 3 |
| %v5 = insertelement <8 x i64> %v4, i64 %a, i32 4 |
| %v6 = insertelement <8 x i64> %v5, i64 %b, i32 5 |
| %v7 = insertelement <8 x i64> %v6, i64 %c, i32 6 |
| %v8 = insertelement <8 x i64> %v7, i64 %d, i32 7 |
| ret <8 x i64> %v8 |
| } |
| |
| define <8 x i64> @v8xi64_exact_undef_suffix(i64 %a, i64 %b, i64 %c, i64 %d) vscale_range(2,2) { |
| ; RV32-LABEL: v8xi64_exact_undef_suffix: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32-NEXT: vmv.v.x v8, a4 |
| ; RV32-NEXT: vmv.v.x v9, a0 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV32-NEXT: vslide1down.vx v10, v9, a1 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a6 |
| ; RV32-NEXT: vslide1down.vx v9, v8, a7 |
| ; RV32-NEXT: vslide1down.vx v8, v10, a2 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: v8xi64_exact_undef_suffix: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RV64V-NEXT: vmv.v.x v8, a2 |
| ; RV64V-NEXT: vslide1down.vx v9, v8, a3 |
| ; RV64V-NEXT: vmv.v.x v8, a0 |
| ; RV64V-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: v8xi64_exact_undef_suffix: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: sd a1, 0(a0) |
| ; RV64ZVE32-NEXT: sd a2, 8(a0) |
| ; RV64ZVE32-NEXT: sd a3, 16(a0) |
| ; RV64ZVE32-NEXT: sd a4, 24(a0) |
| ; RV64ZVE32-NEXT: ret |
| %v1 = insertelement <8 x i64> poison, i64 %a, i32 0 |
| %v2 = insertelement <8 x i64> %v1, i64 %b, i32 1 |
| %v3 = insertelement <8 x i64> %v2, i64 %c, i32 2 |
| %v4 = insertelement <8 x i64> %v3, i64 %d, i32 3 |
| ret <8 x i64> %v4 |
| } |
| |
| define <8 x i64> @v8xi64_exact_undef_prefix(i64 %a, i64 %b, i64 %c, i64 %d) vscale_range(2,2) { |
| ; RV32-LABEL: v8xi64_exact_undef_prefix: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32-NEXT: vmv.v.x v8, a4 |
| ; RV32-NEXT: vmv.v.x v9, a0 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV32-NEXT: vslide1down.vx v9, v9, a1 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a6 |
| ; RV32-NEXT: vslide1down.vx v11, v8, a7 |
| ; RV32-NEXT: vslide1down.vx v8, v9, a2 |
| ; RV32-NEXT: vslide1down.vx v10, v8, a3 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-LABEL: v8xi64_exact_undef_prefix: |
| ; RV64V: # %bb.0: |
| ; RV64V-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RV64V-NEXT: vmv.v.x v8, a2 |
| ; RV64V-NEXT: vslide1down.vx v11, v8, a3 |
| ; RV64V-NEXT: vmv.v.x v8, a0 |
| ; RV64V-NEXT: vslide1down.vx v10, v8, a1 |
| ; RV64V-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: v8xi64_exact_undef_prefix: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: sd a1, 32(a0) |
| ; RV64ZVE32-NEXT: sd a2, 40(a0) |
| ; RV64ZVE32-NEXT: sd a3, 48(a0) |
| ; RV64ZVE32-NEXT: sd a4, 56(a0) |
| ; RV64ZVE32-NEXT: ret |
| %v1 = insertelement <8 x i64> poison, i64 %a, i32 4 |
| %v2 = insertelement <8 x i64> %v1, i64 %b, i32 5 |
| %v3 = insertelement <8 x i64> %v2, i64 %c, i32 6 |
| %v4 = insertelement <8 x i64> %v3, i64 %d, i32 7 |
| ret <8 x i64> %v4 |
| } |
| |
| |
| define <16 x i8> @buildvec_v16i8_loads_contigous(ptr %p) { |
| ; RV32-ONLY-LABEL: buildvec_v16i8_loads_contigous: |
| ; RV32-ONLY: # %bb.0: |
| ; RV32-ONLY-NEXT: addi sp, sp, -16 |
| ; RV32-ONLY-NEXT: .cfi_def_cfa_offset 16 |
| ; RV32-ONLY-NEXT: sw s0, 12(sp) # 4-byte Folded Spill |
| ; RV32-ONLY-NEXT: .cfi_offset s0, -4 |
| ; RV32-ONLY-NEXT: lbu a1, 0(a0) |
| ; RV32-ONLY-NEXT: lbu a2, 1(a0) |
| ; RV32-ONLY-NEXT: lbu a3, 2(a0) |
| ; RV32-ONLY-NEXT: lbu a4, 3(a0) |
| ; RV32-ONLY-NEXT: lbu a5, 4(a0) |
| ; RV32-ONLY-NEXT: lbu a6, 5(a0) |
| ; RV32-ONLY-NEXT: lbu a7, 6(a0) |
| ; RV32-ONLY-NEXT: lbu t0, 7(a0) |
| ; RV32-ONLY-NEXT: lbu t1, 8(a0) |
| ; RV32-ONLY-NEXT: lbu t2, 9(a0) |
| ; RV32-ONLY-NEXT: lbu t3, 10(a0) |
| ; RV32-ONLY-NEXT: lbu t4, 11(a0) |
| ; RV32-ONLY-NEXT: li t5, 255 |
| ; RV32-ONLY-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV32-ONLY-NEXT: vmv.s.x v0, t5 |
| ; RV32-ONLY-NEXT: lbu t5, 12(a0) |
| ; RV32-ONLY-NEXT: lbu t6, 13(a0) |
| ; RV32-ONLY-NEXT: lbu s0, 14(a0) |
| ; RV32-ONLY-NEXT: lbu a0, 15(a0) |
| ; RV32-ONLY-NEXT: vsetivli zero, 16, e8, m1, ta, mu |
| ; RV32-ONLY-NEXT: vmv.v.x v8, t1 |
| ; RV32-ONLY-NEXT: vmv.v.x v9, a1 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, t2 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a2 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, t3 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a3 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, t4 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a4 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, t5 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a5 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, t6 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a6 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, s0 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a7 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, t0 |
| ; RV32-ONLY-NEXT: vslidedown.vi v8, v9, 8, v0.t |
| ; RV32-ONLY-NEXT: lw s0, 12(sp) # 4-byte Folded Reload |
| ; RV32-ONLY-NEXT: .cfi_restore s0 |
| ; RV32-ONLY-NEXT: addi sp, sp, 16 |
| ; RV32-ONLY-NEXT: .cfi_def_cfa_offset 0 |
| ; RV32-ONLY-NEXT: ret |
| ; |
| ; RV32VB-LABEL: buildvec_v16i8_loads_contigous: |
| ; RV32VB: # %bb.0: |
| ; RV32VB-NEXT: lbu a1, 0(a0) |
| ; RV32VB-NEXT: lbu a2, 1(a0) |
| ; RV32VB-NEXT: lbu a3, 2(a0) |
| ; RV32VB-NEXT: lbu a4, 3(a0) |
| ; RV32VB-NEXT: lbu a5, 4(a0) |
| ; RV32VB-NEXT: lbu a6, 5(a0) |
| ; RV32VB-NEXT: lbu a7, 6(a0) |
| ; RV32VB-NEXT: lbu t0, 7(a0) |
| ; RV32VB-NEXT: slli a2, a2, 8 |
| ; RV32VB-NEXT: slli a3, a3, 16 |
| ; RV32VB-NEXT: slli a4, a4, 24 |
| ; RV32VB-NEXT: slli a6, a6, 8 |
| ; RV32VB-NEXT: or a1, a1, a2 |
| ; RV32VB-NEXT: or a3, a4, a3 |
| ; RV32VB-NEXT: or a2, a5, a6 |
| ; RV32VB-NEXT: lbu a4, 8(a0) |
| ; RV32VB-NEXT: lbu a5, 9(a0) |
| ; RV32VB-NEXT: lbu a6, 10(a0) |
| ; RV32VB-NEXT: lbu t1, 11(a0) |
| ; RV32VB-NEXT: slli a7, a7, 16 |
| ; RV32VB-NEXT: slli t0, t0, 24 |
| ; RV32VB-NEXT: slli a5, a5, 8 |
| ; RV32VB-NEXT: slli a6, a6, 16 |
| ; RV32VB-NEXT: slli t1, t1, 24 |
| ; RV32VB-NEXT: or a7, t0, a7 |
| ; RV32VB-NEXT: or a4, a4, a5 |
| ; RV32VB-NEXT: or a5, t1, a6 |
| ; RV32VB-NEXT: lbu a6, 13(a0) |
| ; RV32VB-NEXT: lbu t0, 12(a0) |
| ; RV32VB-NEXT: lbu t1, 14(a0) |
| ; RV32VB-NEXT: lbu a0, 15(a0) |
| ; RV32VB-NEXT: slli a6, a6, 8 |
| ; RV32VB-NEXT: or a6, t0, a6 |
| ; RV32VB-NEXT: slli t1, t1, 16 |
| ; RV32VB-NEXT: slli a0, a0, 24 |
| ; RV32VB-NEXT: or a0, a0, t1 |
| ; RV32VB-NEXT: or a1, a1, a3 |
| ; RV32VB-NEXT: or a2, a2, a7 |
| ; RV32VB-NEXT: or a4, a4, a5 |
| ; RV32VB-NEXT: or a0, a6, a0 |
| ; RV32VB-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32VB-NEXT: vmv.v.x v8, a1 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a4 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-NEXT: ret |
| ; |
| ; RV32VB-PACK-LABEL: buildvec_v16i8_loads_contigous: |
| ; RV32VB-PACK: # %bb.0: |
| ; RV32VB-PACK-NEXT: lbu a1, 0(a0) |
| ; RV32VB-PACK-NEXT: lbu a2, 1(a0) |
| ; RV32VB-PACK-NEXT: lbu a3, 2(a0) |
| ; RV32VB-PACK-NEXT: lbu a4, 3(a0) |
| ; RV32VB-PACK-NEXT: lbu a5, 4(a0) |
| ; RV32VB-PACK-NEXT: lbu a6, 5(a0) |
| ; RV32VB-PACK-NEXT: lbu a7, 6(a0) |
| ; RV32VB-PACK-NEXT: lbu t0, 7(a0) |
| ; RV32VB-PACK-NEXT: packh a1, a1, a2 |
| ; RV32VB-PACK-NEXT: lbu a2, 8(a0) |
| ; RV32VB-PACK-NEXT: lbu t1, 9(a0) |
| ; RV32VB-PACK-NEXT: lbu t2, 10(a0) |
| ; RV32VB-PACK-NEXT: lbu t3, 11(a0) |
| ; RV32VB-PACK-NEXT: packh a3, a3, a4 |
| ; RV32VB-PACK-NEXT: packh a4, a5, a6 |
| ; RV32VB-PACK-NEXT: packh a5, a7, t0 |
| ; RV32VB-PACK-NEXT: lbu a6, 12(a0) |
| ; RV32VB-PACK-NEXT: lbu a7, 13(a0) |
| ; RV32VB-PACK-NEXT: lbu t0, 14(a0) |
| ; RV32VB-PACK-NEXT: lbu a0, 15(a0) |
| ; RV32VB-PACK-NEXT: packh a2, a2, t1 |
| ; RV32VB-PACK-NEXT: packh t1, t2, t3 |
| ; RV32VB-PACK-NEXT: packh a6, a6, a7 |
| ; RV32VB-PACK-NEXT: packh a0, t0, a0 |
| ; RV32VB-PACK-NEXT: pack a1, a1, a3 |
| ; RV32VB-PACK-NEXT: pack a3, a4, a5 |
| ; RV32VB-PACK-NEXT: pack a2, a2, t1 |
| ; RV32VB-PACK-NEXT: pack a0, a6, a0 |
| ; RV32VB-PACK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32VB-PACK-NEXT: vmv.v.x v8, a1 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-PACK-NEXT: ret |
| ; |
| ; RV64V-ONLY-LABEL: buildvec_v16i8_loads_contigous: |
| ; RV64V-ONLY: # %bb.0: |
| ; RV64V-ONLY-NEXT: addi sp, sp, -16 |
| ; RV64V-ONLY-NEXT: .cfi_def_cfa_offset 16 |
| ; RV64V-ONLY-NEXT: sd s0, 8(sp) # 8-byte Folded Spill |
| ; RV64V-ONLY-NEXT: .cfi_offset s0, -8 |
| ; RV64V-ONLY-NEXT: lbu a1, 0(a0) |
| ; RV64V-ONLY-NEXT: lbu a2, 1(a0) |
| ; RV64V-ONLY-NEXT: lbu a3, 2(a0) |
| ; RV64V-ONLY-NEXT: lbu a4, 3(a0) |
| ; RV64V-ONLY-NEXT: lbu a5, 4(a0) |
| ; RV64V-ONLY-NEXT: lbu a6, 5(a0) |
| ; RV64V-ONLY-NEXT: lbu a7, 6(a0) |
| ; RV64V-ONLY-NEXT: lbu t0, 7(a0) |
| ; RV64V-ONLY-NEXT: lbu t1, 8(a0) |
| ; RV64V-ONLY-NEXT: lbu t2, 9(a0) |
| ; RV64V-ONLY-NEXT: lbu t3, 10(a0) |
| ; RV64V-ONLY-NEXT: lbu t4, 11(a0) |
| ; RV64V-ONLY-NEXT: li t5, 255 |
| ; RV64V-ONLY-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.s.x v0, t5 |
| ; RV64V-ONLY-NEXT: lbu t5, 12(a0) |
| ; RV64V-ONLY-NEXT: lbu t6, 13(a0) |
| ; RV64V-ONLY-NEXT: lbu s0, 14(a0) |
| ; RV64V-ONLY-NEXT: lbu a0, 15(a0) |
| ; RV64V-ONLY-NEXT: vsetivli zero, 16, e8, m1, ta, mu |
| ; RV64V-ONLY-NEXT: vmv.v.x v8, t1 |
| ; RV64V-ONLY-NEXT: vmv.v.x v9, a1 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, t2 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a2 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, t3 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a3 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, t4 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a4 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, t5 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a5 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, t6 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a6 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, s0 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a7 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, t0 |
| ; RV64V-ONLY-NEXT: vslidedown.vi v8, v9, 8, v0.t |
| ; RV64V-ONLY-NEXT: ld s0, 8(sp) # 8-byte Folded Reload |
| ; RV64V-ONLY-NEXT: .cfi_restore s0 |
| ; RV64V-ONLY-NEXT: addi sp, sp, 16 |
| ; RV64V-ONLY-NEXT: .cfi_def_cfa_offset 0 |
| ; RV64V-ONLY-NEXT: ret |
| ; |
| ; RVA22U64-LABEL: buildvec_v16i8_loads_contigous: |
| ; RVA22U64: # %bb.0: |
| ; RVA22U64-NEXT: lbu a6, 0(a0) |
| ; RVA22U64-NEXT: lbu a2, 1(a0) |
| ; RVA22U64-NEXT: lbu a3, 2(a0) |
| ; RVA22U64-NEXT: lbu a4, 3(a0) |
| ; RVA22U64-NEXT: lbu a5, 4(a0) |
| ; RVA22U64-NEXT: lbu a1, 5(a0) |
| ; RVA22U64-NEXT: lbu a7, 6(a0) |
| ; RVA22U64-NEXT: lbu t0, 7(a0) |
| ; RVA22U64-NEXT: slli a2, a2, 8 |
| ; RVA22U64-NEXT: slli a3, a3, 16 |
| ; RVA22U64-NEXT: slli a4, a4, 24 |
| ; RVA22U64-NEXT: slli a5, a5, 32 |
| ; RVA22U64-NEXT: slli a1, a1, 40 |
| ; RVA22U64-NEXT: or a6, a6, a2 |
| ; RVA22U64-NEXT: or t2, a4, a3 |
| ; RVA22U64-NEXT: or t1, a1, a5 |
| ; RVA22U64-NEXT: lbu a4, 8(a0) |
| ; RVA22U64-NEXT: lbu a5, 9(a0) |
| ; RVA22U64-NEXT: lbu a2, 10(a0) |
| ; RVA22U64-NEXT: lbu a1, 11(a0) |
| ; RVA22U64-NEXT: slli a7, a7, 48 |
| ; RVA22U64-NEXT: slli t0, t0, 56 |
| ; RVA22U64-NEXT: slli a5, a5, 8 |
| ; RVA22U64-NEXT: slli a2, a2, 16 |
| ; RVA22U64-NEXT: slli a1, a1, 24 |
| ; RVA22U64-NEXT: or a7, t0, a7 |
| ; RVA22U64-NEXT: or a4, a4, a5 |
| ; RVA22U64-NEXT: or a1, a1, a2 |
| ; RVA22U64-NEXT: lbu a2, 12(a0) |
| ; RVA22U64-NEXT: lbu a5, 13(a0) |
| ; RVA22U64-NEXT: lbu a3, 14(a0) |
| ; RVA22U64-NEXT: lbu a0, 15(a0) |
| ; RVA22U64-NEXT: slli a2, a2, 32 |
| ; RVA22U64-NEXT: slli a5, a5, 40 |
| ; RVA22U64-NEXT: or a2, a2, a5 |
| ; RVA22U64-NEXT: slli a3, a3, 48 |
| ; RVA22U64-NEXT: slli a0, a0, 56 |
| ; RVA22U64-NEXT: or a0, a0, a3 |
| ; RVA22U64-NEXT: or a3, a6, t2 |
| ; RVA22U64-NEXT: or a5, a7, t1 |
| ; RVA22U64-NEXT: or a1, a1, a4 |
| ; RVA22U64-NEXT: or a0, a0, a2 |
| ; RVA22U64-NEXT: or a3, a3, a5 |
| ; RVA22U64-NEXT: or a0, a0, a1 |
| ; RVA22U64-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-NEXT: vmv.v.x v8, a3 |
| ; RVA22U64-NEXT: vslide1down.vx v8, v8, a0 |
| ; RVA22U64-NEXT: ret |
| ; |
| ; RVA22U64-PACK-LABEL: buildvec_v16i8_loads_contigous: |
| ; RVA22U64-PACK: # %bb.0: |
| ; RVA22U64-PACK-NEXT: lbu a1, 0(a0) |
| ; RVA22U64-PACK-NEXT: lbu a2, 1(a0) |
| ; RVA22U64-PACK-NEXT: lbu a6, 2(a0) |
| ; RVA22U64-PACK-NEXT: lbu a7, 3(a0) |
| ; RVA22U64-PACK-NEXT: lbu t0, 4(a0) |
| ; RVA22U64-PACK-NEXT: lbu a3, 5(a0) |
| ; RVA22U64-PACK-NEXT: lbu a4, 6(a0) |
| ; RVA22U64-PACK-NEXT: lbu a5, 7(a0) |
| ; RVA22U64-PACK-NEXT: packh t1, a1, a2 |
| ; RVA22U64-PACK-NEXT: lbu t2, 8(a0) |
| ; RVA22U64-PACK-NEXT: lbu t3, 9(a0) |
| ; RVA22U64-PACK-NEXT: lbu t4, 10(a0) |
| ; RVA22U64-PACK-NEXT: lbu a1, 11(a0) |
| ; RVA22U64-PACK-NEXT: packh a6, a6, a7 |
| ; RVA22U64-PACK-NEXT: packh a7, t0, a3 |
| ; RVA22U64-PACK-NEXT: packh t0, a4, a5 |
| ; RVA22U64-PACK-NEXT: lbu a5, 12(a0) |
| ; RVA22U64-PACK-NEXT: lbu a3, 13(a0) |
| ; RVA22U64-PACK-NEXT: lbu a2, 14(a0) |
| ; RVA22U64-PACK-NEXT: lbu a0, 15(a0) |
| ; RVA22U64-PACK-NEXT: packh a4, t2, t3 |
| ; RVA22U64-PACK-NEXT: packh a1, t4, a1 |
| ; RVA22U64-PACK-NEXT: packh a3, a5, a3 |
| ; RVA22U64-PACK-NEXT: packh a0, a2, a0 |
| ; RVA22U64-PACK-NEXT: packw a2, t1, a6 |
| ; RVA22U64-PACK-NEXT: packw a5, a7, t0 |
| ; RVA22U64-PACK-NEXT: packw a1, a4, a1 |
| ; RVA22U64-PACK-NEXT: packw a0, a3, a0 |
| ; RVA22U64-PACK-NEXT: pack a2, a2, a5 |
| ; RVA22U64-PACK-NEXT: pack a0, a1, a0 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.v.x v8, a2 |
| ; RVA22U64-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RVA22U64-PACK-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_v16i8_loads_contigous: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: addi sp, sp, -16 |
| ; RV64ZVE32-NEXT: .cfi_def_cfa_offset 16 |
| ; RV64ZVE32-NEXT: sd s0, 8(sp) # 8-byte Folded Spill |
| ; RV64ZVE32-NEXT: .cfi_offset s0, -8 |
| ; RV64ZVE32-NEXT: lbu a1, 0(a0) |
| ; RV64ZVE32-NEXT: lbu a2, 1(a0) |
| ; RV64ZVE32-NEXT: lbu a3, 2(a0) |
| ; RV64ZVE32-NEXT: lbu a4, 3(a0) |
| ; RV64ZVE32-NEXT: lbu a5, 4(a0) |
| ; RV64ZVE32-NEXT: lbu a6, 5(a0) |
| ; RV64ZVE32-NEXT: lbu a7, 6(a0) |
| ; RV64ZVE32-NEXT: lbu t0, 7(a0) |
| ; RV64ZVE32-NEXT: lbu t1, 8(a0) |
| ; RV64ZVE32-NEXT: lbu t2, 9(a0) |
| ; RV64ZVE32-NEXT: lbu t3, 10(a0) |
| ; RV64ZVE32-NEXT: lbu t4, 11(a0) |
| ; RV64ZVE32-NEXT: li t5, 255 |
| ; RV64ZVE32-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV64ZVE32-NEXT: vmv.s.x v0, t5 |
| ; RV64ZVE32-NEXT: lbu t5, 12(a0) |
| ; RV64ZVE32-NEXT: lbu t6, 13(a0) |
| ; RV64ZVE32-NEXT: lbu s0, 14(a0) |
| ; RV64ZVE32-NEXT: lbu a0, 15(a0) |
| ; RV64ZVE32-NEXT: vsetivli zero, 16, e8, m1, ta, mu |
| ; RV64ZVE32-NEXT: vmv.v.x v8, t1 |
| ; RV64ZVE32-NEXT: vmv.v.x v9, a1 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, t2 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a2 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, t3 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a3 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, t4 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a4 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, t5 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a5 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, t6 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a6 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, s0 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a7 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, t0 |
| ; RV64ZVE32-NEXT: vslidedown.vi v8, v9, 8, v0.t |
| ; RV64ZVE32-NEXT: ld s0, 8(sp) # 8-byte Folded Reload |
| ; RV64ZVE32-NEXT: .cfi_restore s0 |
| ; RV64ZVE32-NEXT: addi sp, sp, 16 |
| ; RV64ZVE32-NEXT: .cfi_def_cfa_offset 0 |
| ; RV64ZVE32-NEXT: ret |
| %p2 = getelementptr i8, ptr %p, i32 1 |
| %p3 = getelementptr i8, ptr %p, i32 2 |
| %p4 = getelementptr i8, ptr %p, i32 3 |
| %p5 = getelementptr i8, ptr %p, i32 4 |
| %p6 = getelementptr i8, ptr %p, i32 5 |
| %p7 = getelementptr i8, ptr %p, i32 6 |
| %p8 = getelementptr i8, ptr %p, i32 7 |
| %p9 = getelementptr i8, ptr %p, i32 8 |
| %p10 = getelementptr i8, ptr %p, i32 9 |
| %p11 = getelementptr i8, ptr %p, i32 10 |
| %p12 = getelementptr i8, ptr %p, i32 11 |
| %p13 = getelementptr i8, ptr %p, i32 12 |
| %p14 = getelementptr i8, ptr %p, i32 13 |
| %p15 = getelementptr i8, ptr %p, i32 14 |
| %p16 = getelementptr i8, ptr %p, i32 15 |
| |
| %ld1 = load i8, ptr %p |
| %ld2 = load i8, ptr %p2 |
| %ld3 = load i8, ptr %p3 |
| %ld4 = load i8, ptr %p4 |
| %ld5 = load i8, ptr %p5 |
| %ld6 = load i8, ptr %p6 |
| %ld7 = load i8, ptr %p7 |
| %ld8 = load i8, ptr %p8 |
| %ld9 = load i8, ptr %p9 |
| %ld10 = load i8, ptr %p10 |
| %ld11 = load i8, ptr %p11 |
| %ld12 = load i8, ptr %p12 |
| %ld13 = load i8, ptr %p13 |
| %ld14 = load i8, ptr %p14 |
| %ld15 = load i8, ptr %p15 |
| %ld16 = load i8, ptr %p16 |
| |
| %v1 = insertelement <16 x i8> poison, i8 %ld1, i32 0 |
| %v2 = insertelement <16 x i8> %v1, i8 %ld2, i32 1 |
| %v3 = insertelement <16 x i8> %v2, i8 %ld3, i32 2 |
| %v4 = insertelement <16 x i8> %v3, i8 %ld4, i32 3 |
| %v5 = insertelement <16 x i8> %v4, i8 %ld5, i32 4 |
| %v6 = insertelement <16 x i8> %v5, i8 %ld6, i32 5 |
| %v7 = insertelement <16 x i8> %v6, i8 %ld7, i32 6 |
| %v8 = insertelement <16 x i8> %v7, i8 %ld8, i32 7 |
| %v9 = insertelement <16 x i8> %v8, i8 %ld9, i32 8 |
| %v10 = insertelement <16 x i8> %v9, i8 %ld10, i32 9 |
| %v11 = insertelement <16 x i8> %v10, i8 %ld11, i32 10 |
| %v12 = insertelement <16 x i8> %v11, i8 %ld12, i32 11 |
| %v13 = insertelement <16 x i8> %v12, i8 %ld13, i32 12 |
| %v14 = insertelement <16 x i8> %v13, i8 %ld14, i32 13 |
| %v15 = insertelement <16 x i8> %v14, i8 %ld15, i32 14 |
| %v16 = insertelement <16 x i8> %v15, i8 %ld16, i32 15 |
| ret <16 x i8> %v16 |
| } |
| |
| |
| define <16 x i8> @buildvec_v16i8_loads_gather(ptr %p) { |
| ; RV32-ONLY-LABEL: buildvec_v16i8_loads_gather: |
| ; RV32-ONLY: # %bb.0: |
| ; RV32-ONLY-NEXT: addi sp, sp, -16 |
| ; RV32-ONLY-NEXT: .cfi_def_cfa_offset 16 |
| ; RV32-ONLY-NEXT: sw s0, 12(sp) # 4-byte Folded Spill |
| ; RV32-ONLY-NEXT: .cfi_offset s0, -4 |
| ; RV32-ONLY-NEXT: lbu a1, 0(a0) |
| ; RV32-ONLY-NEXT: lbu a2, 1(a0) |
| ; RV32-ONLY-NEXT: lbu a3, 22(a0) |
| ; RV32-ONLY-NEXT: lbu a4, 31(a0) |
| ; RV32-ONLY-NEXT: lbu a5, 623(a0) |
| ; RV32-ONLY-NEXT: lbu a6, 44(a0) |
| ; RV32-ONLY-NEXT: lbu a7, 55(a0) |
| ; RV32-ONLY-NEXT: lbu t0, 75(a0) |
| ; RV32-ONLY-NEXT: lbu t1, 82(a0) |
| ; RV32-ONLY-NEXT: lbu t2, 154(a0) |
| ; RV32-ONLY-NEXT: lbu t3, 161(a0) |
| ; RV32-ONLY-NEXT: lbu t4, 163(a0) |
| ; RV32-ONLY-NEXT: li t5, 255 |
| ; RV32-ONLY-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV32-ONLY-NEXT: vmv.s.x v0, t5 |
| ; RV32-ONLY-NEXT: lbu t5, 93(a0) |
| ; RV32-ONLY-NEXT: lbu t6, 105(a0) |
| ; RV32-ONLY-NEXT: lbu s0, 124(a0) |
| ; RV32-ONLY-NEXT: lbu a0, 144(a0) |
| ; RV32-ONLY-NEXT: vsetivli zero, 16, e8, m1, ta, mu |
| ; RV32-ONLY-NEXT: vmv.v.x v8, t1 |
| ; RV32-ONLY-NEXT: vmv.v.x v9, a1 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, t5 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a2 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, t6 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a3 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, t3 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a4 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, s0 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a6 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, t4 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a7 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a5 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, t2 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, t0 |
| ; RV32-ONLY-NEXT: vslidedown.vi v8, v9, 8, v0.t |
| ; RV32-ONLY-NEXT: lw s0, 12(sp) # 4-byte Folded Reload |
| ; RV32-ONLY-NEXT: .cfi_restore s0 |
| ; RV32-ONLY-NEXT: addi sp, sp, 16 |
| ; RV32-ONLY-NEXT: .cfi_def_cfa_offset 0 |
| ; RV32-ONLY-NEXT: ret |
| ; |
| ; RV32VB-LABEL: buildvec_v16i8_loads_gather: |
| ; RV32VB: # %bb.0: |
| ; RV32VB-NEXT: lbu a1, 0(a0) |
| ; RV32VB-NEXT: lbu a2, 1(a0) |
| ; RV32VB-NEXT: lbu a3, 22(a0) |
| ; RV32VB-NEXT: lbu a4, 31(a0) |
| ; RV32VB-NEXT: lbu a5, 623(a0) |
| ; RV32VB-NEXT: lbu a6, 44(a0) |
| ; RV32VB-NEXT: lbu a7, 55(a0) |
| ; RV32VB-NEXT: lbu t0, 75(a0) |
| ; RV32VB-NEXT: lbu t1, 82(a0) |
| ; RV32VB-NEXT: slli a2, a2, 8 |
| ; RV32VB-NEXT: slli a3, a3, 16 |
| ; RV32VB-NEXT: slli a4, a4, 24 |
| ; RV32VB-NEXT: slli a7, a7, 8 |
| ; RV32VB-NEXT: or a1, a1, a2 |
| ; RV32VB-NEXT: or a3, a4, a3 |
| ; RV32VB-NEXT: or a2, a6, a7 |
| ; RV32VB-NEXT: lbu a4, 93(a0) |
| ; RV32VB-NEXT: lbu a6, 105(a0) |
| ; RV32VB-NEXT: lbu a7, 124(a0) |
| ; RV32VB-NEXT: lbu t2, 144(a0) |
| ; RV32VB-NEXT: slli a5, a5, 16 |
| ; RV32VB-NEXT: slli t0, t0, 24 |
| ; RV32VB-NEXT: slli a4, a4, 8 |
| ; RV32VB-NEXT: or a5, t0, a5 |
| ; RV32VB-NEXT: or a4, t1, a4 |
| ; RV32VB-NEXT: lbu t0, 161(a0) |
| ; RV32VB-NEXT: lbu t1, 154(a0) |
| ; RV32VB-NEXT: lbu a0, 163(a0) |
| ; RV32VB-NEXT: slli a6, a6, 16 |
| ; RV32VB-NEXT: slli t0, t0, 24 |
| ; RV32VB-NEXT: or a6, t0, a6 |
| ; RV32VB-NEXT: slli a0, a0, 8 |
| ; RV32VB-NEXT: or a0, a7, a0 |
| ; RV32VB-NEXT: slli t2, t2, 16 |
| ; RV32VB-NEXT: slli t1, t1, 24 |
| ; RV32VB-NEXT: or a7, t1, t2 |
| ; RV32VB-NEXT: or a1, a1, a3 |
| ; RV32VB-NEXT: or a2, a2, a5 |
| ; RV32VB-NEXT: or a3, a4, a6 |
| ; RV32VB-NEXT: or a0, a0, a7 |
| ; RV32VB-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32VB-NEXT: vmv.v.x v8, a1 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-NEXT: ret |
| ; |
| ; RV32VB-PACK-LABEL: buildvec_v16i8_loads_gather: |
| ; RV32VB-PACK: # %bb.0: |
| ; RV32VB-PACK-NEXT: lbu a1, 0(a0) |
| ; RV32VB-PACK-NEXT: lbu a2, 1(a0) |
| ; RV32VB-PACK-NEXT: lbu a3, 22(a0) |
| ; RV32VB-PACK-NEXT: lbu a4, 31(a0) |
| ; RV32VB-PACK-NEXT: lbu a5, 623(a0) |
| ; RV32VB-PACK-NEXT: lbu a6, 44(a0) |
| ; RV32VB-PACK-NEXT: lbu a7, 55(a0) |
| ; RV32VB-PACK-NEXT: lbu t0, 75(a0) |
| ; RV32VB-PACK-NEXT: lbu t1, 82(a0) |
| ; RV32VB-PACK-NEXT: packh a1, a1, a2 |
| ; RV32VB-PACK-NEXT: lbu a2, 154(a0) |
| ; RV32VB-PACK-NEXT: lbu t2, 161(a0) |
| ; RV32VB-PACK-NEXT: lbu t3, 163(a0) |
| ; RV32VB-PACK-NEXT: packh a3, a3, a4 |
| ; RV32VB-PACK-NEXT: packh a4, a6, a7 |
| ; RV32VB-PACK-NEXT: packh a5, a5, t0 |
| ; RV32VB-PACK-NEXT: lbu a6, 93(a0) |
| ; RV32VB-PACK-NEXT: lbu a7, 105(a0) |
| ; RV32VB-PACK-NEXT: lbu t0, 124(a0) |
| ; RV32VB-PACK-NEXT: lbu a0, 144(a0) |
| ; RV32VB-PACK-NEXT: packh a6, t1, a6 |
| ; RV32VB-PACK-NEXT: packh a7, a7, t2 |
| ; RV32VB-PACK-NEXT: packh t0, t0, t3 |
| ; RV32VB-PACK-NEXT: packh a0, a0, a2 |
| ; RV32VB-PACK-NEXT: pack a1, a1, a3 |
| ; RV32VB-PACK-NEXT: pack a2, a4, a5 |
| ; RV32VB-PACK-NEXT: pack a3, a6, a7 |
| ; RV32VB-PACK-NEXT: pack a0, t0, a0 |
| ; RV32VB-PACK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32VB-PACK-NEXT: vmv.v.x v8, a1 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-PACK-NEXT: ret |
| ; |
| ; RV64V-ONLY-LABEL: buildvec_v16i8_loads_gather: |
| ; RV64V-ONLY: # %bb.0: |
| ; RV64V-ONLY-NEXT: addi sp, sp, -16 |
| ; RV64V-ONLY-NEXT: .cfi_def_cfa_offset 16 |
| ; RV64V-ONLY-NEXT: sd s0, 8(sp) # 8-byte Folded Spill |
| ; RV64V-ONLY-NEXT: .cfi_offset s0, -8 |
| ; RV64V-ONLY-NEXT: lbu a1, 0(a0) |
| ; RV64V-ONLY-NEXT: lbu a2, 1(a0) |
| ; RV64V-ONLY-NEXT: lbu a3, 22(a0) |
| ; RV64V-ONLY-NEXT: lbu a4, 31(a0) |
| ; RV64V-ONLY-NEXT: lbu a5, 623(a0) |
| ; RV64V-ONLY-NEXT: lbu a6, 44(a0) |
| ; RV64V-ONLY-NEXT: lbu a7, 55(a0) |
| ; RV64V-ONLY-NEXT: lbu t0, 75(a0) |
| ; RV64V-ONLY-NEXT: lbu t1, 82(a0) |
| ; RV64V-ONLY-NEXT: lbu t2, 154(a0) |
| ; RV64V-ONLY-NEXT: lbu t3, 161(a0) |
| ; RV64V-ONLY-NEXT: lbu t4, 163(a0) |
| ; RV64V-ONLY-NEXT: li t5, 255 |
| ; RV64V-ONLY-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.s.x v0, t5 |
| ; RV64V-ONLY-NEXT: lbu t5, 93(a0) |
| ; RV64V-ONLY-NEXT: lbu t6, 105(a0) |
| ; RV64V-ONLY-NEXT: lbu s0, 124(a0) |
| ; RV64V-ONLY-NEXT: lbu a0, 144(a0) |
| ; RV64V-ONLY-NEXT: vsetivli zero, 16, e8, m1, ta, mu |
| ; RV64V-ONLY-NEXT: vmv.v.x v8, t1 |
| ; RV64V-ONLY-NEXT: vmv.v.x v9, a1 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, t5 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a2 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, t6 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a3 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, t3 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a4 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, s0 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a6 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, t4 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a7 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a5 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, t2 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, t0 |
| ; RV64V-ONLY-NEXT: vslidedown.vi v8, v9, 8, v0.t |
| ; RV64V-ONLY-NEXT: ld s0, 8(sp) # 8-byte Folded Reload |
| ; RV64V-ONLY-NEXT: .cfi_restore s0 |
| ; RV64V-ONLY-NEXT: addi sp, sp, 16 |
| ; RV64V-ONLY-NEXT: .cfi_def_cfa_offset 0 |
| ; RV64V-ONLY-NEXT: ret |
| ; |
| ; RVA22U64-LABEL: buildvec_v16i8_loads_gather: |
| ; RVA22U64: # %bb.0: |
| ; RVA22U64-NEXT: lbu a7, 0(a0) |
| ; RVA22U64-NEXT: lbu a2, 1(a0) |
| ; RVA22U64-NEXT: lbu a3, 22(a0) |
| ; RVA22U64-NEXT: lbu a4, 31(a0) |
| ; RVA22U64-NEXT: lbu a6, 623(a0) |
| ; RVA22U64-NEXT: lbu a5, 44(a0) |
| ; RVA22U64-NEXT: lbu a1, 55(a0) |
| ; RVA22U64-NEXT: lbu t0, 75(a0) |
| ; RVA22U64-NEXT: lbu t1, 82(a0) |
| ; RVA22U64-NEXT: slli a2, a2, 8 |
| ; RVA22U64-NEXT: slli a3, a3, 16 |
| ; RVA22U64-NEXT: slli a4, a4, 24 |
| ; RVA22U64-NEXT: slli a5, a5, 32 |
| ; RVA22U64-NEXT: slli a1, a1, 40 |
| ; RVA22U64-NEXT: or a7, a7, a2 |
| ; RVA22U64-NEXT: or t3, a4, a3 |
| ; RVA22U64-NEXT: or t2, a1, a5 |
| ; RVA22U64-NEXT: lbu a4, 93(a0) |
| ; RVA22U64-NEXT: lbu t4, 105(a0) |
| ; RVA22U64-NEXT: lbu a2, 124(a0) |
| ; RVA22U64-NEXT: lbu t5, 144(a0) |
| ; RVA22U64-NEXT: slli a6, a6, 48 |
| ; RVA22U64-NEXT: slli t0, t0, 56 |
| ; RVA22U64-NEXT: slli a4, a4, 8 |
| ; RVA22U64-NEXT: or a3, t0, a6 |
| ; RVA22U64-NEXT: or a4, t1, a4 |
| ; RVA22U64-NEXT: lbu a5, 161(a0) |
| ; RVA22U64-NEXT: lbu a1, 154(a0) |
| ; RVA22U64-NEXT: lbu a0, 163(a0) |
| ; RVA22U64-NEXT: slli t4, t4, 16 |
| ; RVA22U64-NEXT: slli a5, a5, 24 |
| ; RVA22U64-NEXT: or a5, a5, t4 |
| ; RVA22U64-NEXT: slli a2, a2, 32 |
| ; RVA22U64-NEXT: slli a0, a0, 40 |
| ; RVA22U64-NEXT: or a0, a0, a2 |
| ; RVA22U64-NEXT: slli t5, t5, 48 |
| ; RVA22U64-NEXT: slli a1, a1, 56 |
| ; RVA22U64-NEXT: or a1, a1, t5 |
| ; RVA22U64-NEXT: or a2, a7, t3 |
| ; RVA22U64-NEXT: or a3, a3, t2 |
| ; RVA22U64-NEXT: or a4, a4, a5 |
| ; RVA22U64-NEXT: or a0, a0, a1 |
| ; RVA22U64-NEXT: or a2, a2, a3 |
| ; RVA22U64-NEXT: or a0, a0, a4 |
| ; RVA22U64-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-NEXT: vmv.v.x v8, a2 |
| ; RVA22U64-NEXT: vslide1down.vx v8, v8, a0 |
| ; RVA22U64-NEXT: ret |
| ; |
| ; RVA22U64-PACK-LABEL: buildvec_v16i8_loads_gather: |
| ; RVA22U64-PACK: # %bb.0: |
| ; RVA22U64-PACK-NEXT: addi sp, sp, -16 |
| ; RVA22U64-PACK-NEXT: .cfi_def_cfa_offset 16 |
| ; RVA22U64-PACK-NEXT: sd s0, 8(sp) # 8-byte Folded Spill |
| ; RVA22U64-PACK-NEXT: .cfi_offset s0, -8 |
| ; RVA22U64-PACK-NEXT: lbu a1, 0(a0) |
| ; RVA22U64-PACK-NEXT: lbu a2, 1(a0) |
| ; RVA22U64-PACK-NEXT: lbu a6, 22(a0) |
| ; RVA22U64-PACK-NEXT: lbu a7, 31(a0) |
| ; RVA22U64-PACK-NEXT: lbu t0, 623(a0) |
| ; RVA22U64-PACK-NEXT: lbu t3, 44(a0) |
| ; RVA22U64-PACK-NEXT: lbu t4, 55(a0) |
| ; RVA22U64-PACK-NEXT: lbu t5, 75(a0) |
| ; RVA22U64-PACK-NEXT: lbu t1, 82(a0) |
| ; RVA22U64-PACK-NEXT: packh t2, a1, a2 |
| ; RVA22U64-PACK-NEXT: lbu t6, 154(a0) |
| ; RVA22U64-PACK-NEXT: lbu s0, 161(a0) |
| ; RVA22U64-PACK-NEXT: lbu a3, 163(a0) |
| ; RVA22U64-PACK-NEXT: packh a6, a6, a7 |
| ; RVA22U64-PACK-NEXT: packh a7, t3, t4 |
| ; RVA22U64-PACK-NEXT: packh a2, t0, t5 |
| ; RVA22U64-PACK-NEXT: lbu a4, 93(a0) |
| ; RVA22U64-PACK-NEXT: lbu a5, 105(a0) |
| ; RVA22U64-PACK-NEXT: lbu a1, 124(a0) |
| ; RVA22U64-PACK-NEXT: lbu a0, 144(a0) |
| ; RVA22U64-PACK-NEXT: packh a4, t1, a4 |
| ; RVA22U64-PACK-NEXT: packh a5, a5, s0 |
| ; RVA22U64-PACK-NEXT: packh a1, a1, a3 |
| ; RVA22U64-PACK-NEXT: packh a0, a0, t6 |
| ; RVA22U64-PACK-NEXT: packw a3, t2, a6 |
| ; RVA22U64-PACK-NEXT: packw a2, a7, a2 |
| ; RVA22U64-PACK-NEXT: packw a4, a4, a5 |
| ; RVA22U64-PACK-NEXT: packw a0, a1, a0 |
| ; RVA22U64-PACK-NEXT: pack a1, a3, a2 |
| ; RVA22U64-PACK-NEXT: pack a0, a4, a0 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.v.x v8, a1 |
| ; RVA22U64-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RVA22U64-PACK-NEXT: ld s0, 8(sp) # 8-byte Folded Reload |
| ; RVA22U64-PACK-NEXT: .cfi_restore s0 |
| ; RVA22U64-PACK-NEXT: addi sp, sp, 16 |
| ; RVA22U64-PACK-NEXT: .cfi_def_cfa_offset 0 |
| ; RVA22U64-PACK-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_v16i8_loads_gather: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: addi sp, sp, -16 |
| ; RV64ZVE32-NEXT: .cfi_def_cfa_offset 16 |
| ; RV64ZVE32-NEXT: sd s0, 8(sp) # 8-byte Folded Spill |
| ; RV64ZVE32-NEXT: .cfi_offset s0, -8 |
| ; RV64ZVE32-NEXT: lbu a1, 0(a0) |
| ; RV64ZVE32-NEXT: lbu a2, 1(a0) |
| ; RV64ZVE32-NEXT: lbu a3, 22(a0) |
| ; RV64ZVE32-NEXT: lbu a4, 31(a0) |
| ; RV64ZVE32-NEXT: lbu a5, 623(a0) |
| ; RV64ZVE32-NEXT: lbu a6, 44(a0) |
| ; RV64ZVE32-NEXT: lbu a7, 55(a0) |
| ; RV64ZVE32-NEXT: lbu t0, 75(a0) |
| ; RV64ZVE32-NEXT: lbu t1, 82(a0) |
| ; RV64ZVE32-NEXT: lbu t2, 154(a0) |
| ; RV64ZVE32-NEXT: lbu t3, 161(a0) |
| ; RV64ZVE32-NEXT: lbu t4, 163(a0) |
| ; RV64ZVE32-NEXT: li t5, 255 |
| ; RV64ZVE32-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV64ZVE32-NEXT: vmv.s.x v0, t5 |
| ; RV64ZVE32-NEXT: lbu t5, 93(a0) |
| ; RV64ZVE32-NEXT: lbu t6, 105(a0) |
| ; RV64ZVE32-NEXT: lbu s0, 124(a0) |
| ; RV64ZVE32-NEXT: lbu a0, 144(a0) |
| ; RV64ZVE32-NEXT: vsetivli zero, 16, e8, m1, ta, mu |
| ; RV64ZVE32-NEXT: vmv.v.x v8, t1 |
| ; RV64ZVE32-NEXT: vmv.v.x v9, a1 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, t5 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a2 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, t6 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a3 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, t3 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a4 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, s0 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a6 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, t4 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a7 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a5 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, t2 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, t0 |
| ; RV64ZVE32-NEXT: vslidedown.vi v8, v9, 8, v0.t |
| ; RV64ZVE32-NEXT: ld s0, 8(sp) # 8-byte Folded Reload |
| ; RV64ZVE32-NEXT: .cfi_restore s0 |
| ; RV64ZVE32-NEXT: addi sp, sp, 16 |
| ; RV64ZVE32-NEXT: .cfi_def_cfa_offset 0 |
| ; RV64ZVE32-NEXT: ret |
| %p2 = getelementptr i8, ptr %p, i32 1 |
| %p3 = getelementptr i8, ptr %p, i32 22 |
| %p4 = getelementptr i8, ptr %p, i32 31 |
| %p5 = getelementptr i8, ptr %p, i32 44 |
| %p6 = getelementptr i8, ptr %p, i32 55 |
| %p7 = getelementptr i8, ptr %p, i32 623 |
| %p8 = getelementptr i8, ptr %p, i32 75 |
| %p9 = getelementptr i8, ptr %p, i32 82 |
| %p10 = getelementptr i8, ptr %p, i32 93 |
| %p11 = getelementptr i8, ptr %p, i32 105 |
| %p12 = getelementptr i8, ptr %p, i32 161 |
| %p13 = getelementptr i8, ptr %p, i32 124 |
| %p14 = getelementptr i8, ptr %p, i32 163 |
| %p15 = getelementptr i8, ptr %p, i32 144 |
| %p16 = getelementptr i8, ptr %p, i32 154 |
| |
| %ld1 = load i8, ptr %p |
| %ld2 = load i8, ptr %p2 |
| %ld3 = load i8, ptr %p3 |
| %ld4 = load i8, ptr %p4 |
| %ld5 = load i8, ptr %p5 |
| %ld6 = load i8, ptr %p6 |
| %ld7 = load i8, ptr %p7 |
| %ld8 = load i8, ptr %p8 |
| %ld9 = load i8, ptr %p9 |
| %ld10 = load i8, ptr %p10 |
| %ld11 = load i8, ptr %p11 |
| %ld12 = load i8, ptr %p12 |
| %ld13 = load i8, ptr %p13 |
| %ld14 = load i8, ptr %p14 |
| %ld15 = load i8, ptr %p15 |
| %ld16 = load i8, ptr %p16 |
| |
| %v1 = insertelement <16 x i8> poison, i8 %ld1, i32 0 |
| %v2 = insertelement <16 x i8> %v1, i8 %ld2, i32 1 |
| %v3 = insertelement <16 x i8> %v2, i8 %ld3, i32 2 |
| %v4 = insertelement <16 x i8> %v3, i8 %ld4, i32 3 |
| %v5 = insertelement <16 x i8> %v4, i8 %ld5, i32 4 |
| %v6 = insertelement <16 x i8> %v5, i8 %ld6, i32 5 |
| %v7 = insertelement <16 x i8> %v6, i8 %ld7, i32 6 |
| %v8 = insertelement <16 x i8> %v7, i8 %ld8, i32 7 |
| %v9 = insertelement <16 x i8> %v8, i8 %ld9, i32 8 |
| %v10 = insertelement <16 x i8> %v9, i8 %ld10, i32 9 |
| %v11 = insertelement <16 x i8> %v10, i8 %ld11, i32 10 |
| %v12 = insertelement <16 x i8> %v11, i8 %ld12, i32 11 |
| %v13 = insertelement <16 x i8> %v12, i8 %ld13, i32 12 |
| %v14 = insertelement <16 x i8> %v13, i8 %ld14, i32 13 |
| %v15 = insertelement <16 x i8> %v14, i8 %ld15, i32 14 |
| %v16 = insertelement <16 x i8> %v15, i8 %ld16, i32 15 |
| ret <16 x i8> %v16 |
| } |
| |
| define <16 x i8> @buildvec_v16i8_undef_low_half(ptr %p) { |
| ; RV32-ONLY-LABEL: buildvec_v16i8_undef_low_half: |
| ; RV32-ONLY: # %bb.0: |
| ; RV32-ONLY-NEXT: lbu a1, 82(a0) |
| ; RV32-ONLY-NEXT: lbu a2, 93(a0) |
| ; RV32-ONLY-NEXT: lbu a3, 105(a0) |
| ; RV32-ONLY-NEXT: lbu a4, 124(a0) |
| ; RV32-ONLY-NEXT: lbu a5, 144(a0) |
| ; RV32-ONLY-NEXT: lbu a6, 154(a0) |
| ; RV32-ONLY-NEXT: lbu a7, 161(a0) |
| ; RV32-ONLY-NEXT: lbu a0, 163(a0) |
| ; RV32-ONLY-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; RV32-ONLY-NEXT: vmv.v.x v8, a1 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a7 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a4 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a6 |
| ; RV32-ONLY-NEXT: ret |
| ; |
| ; RV32VB-LABEL: buildvec_v16i8_undef_low_half: |
| ; RV32VB: # %bb.0: |
| ; RV32VB-NEXT: lbu a1, 93(a0) |
| ; RV32VB-NEXT: lbu a2, 82(a0) |
| ; RV32VB-NEXT: lbu a3, 105(a0) |
| ; RV32VB-NEXT: lbu a4, 124(a0) |
| ; RV32VB-NEXT: slli a1, a1, 8 |
| ; RV32VB-NEXT: or a1, a2, a1 |
| ; RV32VB-NEXT: lbu a2, 161(a0) |
| ; RV32VB-NEXT: lbu a5, 144(a0) |
| ; RV32VB-NEXT: lbu a6, 154(a0) |
| ; RV32VB-NEXT: lbu a0, 163(a0) |
| ; RV32VB-NEXT: slli a3, a3, 16 |
| ; RV32VB-NEXT: slli a2, a2, 24 |
| ; RV32VB-NEXT: or a2, a2, a3 |
| ; RV32VB-NEXT: slli a0, a0, 8 |
| ; RV32VB-NEXT: or a0, a4, a0 |
| ; RV32VB-NEXT: slli a5, a5, 16 |
| ; RV32VB-NEXT: slli a6, a6, 24 |
| ; RV32VB-NEXT: or a3, a6, a5 |
| ; RV32VB-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32VB-NEXT: vmv.v.i v8, 0 |
| ; RV32VB-NEXT: or a1, a1, a2 |
| ; RV32VB-NEXT: or a0, a0, a3 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, zero |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-NEXT: ret |
| ; |
| ; RV32VB-PACK-LABEL: buildvec_v16i8_undef_low_half: |
| ; RV32VB-PACK: # %bb.0: |
| ; RV32VB-PACK-NEXT: lbu a1, 82(a0) |
| ; RV32VB-PACK-NEXT: lbu a2, 93(a0) |
| ; RV32VB-PACK-NEXT: lbu a3, 105(a0) |
| ; RV32VB-PACK-NEXT: lbu a4, 124(a0) |
| ; RV32VB-PACK-NEXT: lbu a5, 161(a0) |
| ; RV32VB-PACK-NEXT: lbu a6, 163(a0) |
| ; RV32VB-PACK-NEXT: lbu a7, 144(a0) |
| ; RV32VB-PACK-NEXT: lbu a0, 154(a0) |
| ; RV32VB-PACK-NEXT: packh a1, a1, a2 |
| ; RV32VB-PACK-NEXT: packh a2, a3, a5 |
| ; RV32VB-PACK-NEXT: packh a3, a4, a6 |
| ; RV32VB-PACK-NEXT: packh a0, a7, a0 |
| ; RV32VB-PACK-NEXT: pack a1, a1, a2 |
| ; RV32VB-PACK-NEXT: packh a2, a0, a0 |
| ; RV32VB-PACK-NEXT: pack a2, a2, a2 |
| ; RV32VB-PACK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32VB-PACK-NEXT: vmv.v.x v8, a2 |
| ; RV32VB-PACK-NEXT: pack a0, a3, a0 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-PACK-NEXT: ret |
| ; |
| ; RV64V-ONLY-LABEL: buildvec_v16i8_undef_low_half: |
| ; RV64V-ONLY: # %bb.0: |
| ; RV64V-ONLY-NEXT: lbu a1, 82(a0) |
| ; RV64V-ONLY-NEXT: lbu a2, 93(a0) |
| ; RV64V-ONLY-NEXT: lbu a3, 105(a0) |
| ; RV64V-ONLY-NEXT: lbu a4, 124(a0) |
| ; RV64V-ONLY-NEXT: lbu a5, 144(a0) |
| ; RV64V-ONLY-NEXT: lbu a6, 154(a0) |
| ; RV64V-ONLY-NEXT: lbu a7, 161(a0) |
| ; RV64V-ONLY-NEXT: lbu a0, 163(a0) |
| ; RV64V-ONLY-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.v.x v8, a1 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a7 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a4 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a6 |
| ; RV64V-ONLY-NEXT: ret |
| ; |
| ; RVA22U64-LABEL: buildvec_v16i8_undef_low_half: |
| ; RVA22U64: # %bb.0: |
| ; RVA22U64-NEXT: lbu a1, 93(a0) |
| ; RVA22U64-NEXT: lbu a2, 82(a0) |
| ; RVA22U64-NEXT: lbu a3, 105(a0) |
| ; RVA22U64-NEXT: lbu a4, 124(a0) |
| ; RVA22U64-NEXT: slli a1, a1, 8 |
| ; RVA22U64-NEXT: or a6, a2, a1 |
| ; RVA22U64-NEXT: lbu a2, 161(a0) |
| ; RVA22U64-NEXT: lbu a5, 144(a0) |
| ; RVA22U64-NEXT: lbu a1, 154(a0) |
| ; RVA22U64-NEXT: lbu a0, 163(a0) |
| ; RVA22U64-NEXT: slli a3, a3, 16 |
| ; RVA22U64-NEXT: slli a2, a2, 24 |
| ; RVA22U64-NEXT: or a2, a2, a3 |
| ; RVA22U64-NEXT: slli a4, a4, 32 |
| ; RVA22U64-NEXT: slli a0, a0, 40 |
| ; RVA22U64-NEXT: or a0, a0, a4 |
| ; RVA22U64-NEXT: slli a5, a5, 48 |
| ; RVA22U64-NEXT: slli a1, a1, 56 |
| ; RVA22U64-NEXT: or a1, a1, a5 |
| ; RVA22U64-NEXT: or a2, a6, a2 |
| ; RVA22U64-NEXT: or a0, a0, a1 |
| ; RVA22U64-NEXT: or a0, a0, a2 |
| ; RVA22U64-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-NEXT: vmv.v.i v8, 0 |
| ; RVA22U64-NEXT: vslide1down.vx v8, v8, a0 |
| ; RVA22U64-NEXT: ret |
| ; |
| ; RVA22U64-PACK-LABEL: buildvec_v16i8_undef_low_half: |
| ; RVA22U64-PACK: # %bb.0: |
| ; RVA22U64-PACK-NEXT: lbu a6, 82(a0) |
| ; RVA22U64-PACK-NEXT: lbu a7, 93(a0) |
| ; RVA22U64-PACK-NEXT: lbu t0, 105(a0) |
| ; RVA22U64-PACK-NEXT: lbu a4, 124(a0) |
| ; RVA22U64-PACK-NEXT: lbu a5, 161(a0) |
| ; RVA22U64-PACK-NEXT: lbu a1, 163(a0) |
| ; RVA22U64-PACK-NEXT: lbu a2, 144(a0) |
| ; RVA22U64-PACK-NEXT: lbu a0, 154(a0) |
| ; RVA22U64-PACK-NEXT: packh a3, a6, a7 |
| ; RVA22U64-PACK-NEXT: packh a5, t0, a5 |
| ; RVA22U64-PACK-NEXT: packh a1, a4, a1 |
| ; RVA22U64-PACK-NEXT: packh a0, a2, a0 |
| ; RVA22U64-PACK-NEXT: packw a2, a3, a5 |
| ; RVA22U64-PACK-NEXT: packh a3, a0, a0 |
| ; RVA22U64-PACK-NEXT: packw a3, a3, a3 |
| ; RVA22U64-PACK-NEXT: pack a3, a3, a3 |
| ; RVA22U64-PACK-NEXT: packw a0, a1, a0 |
| ; RVA22U64-PACK-NEXT: pack a0, a2, a0 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.v.x v8, a3 |
| ; RVA22U64-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RVA22U64-PACK-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_v16i8_undef_low_half: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: lbu a1, 82(a0) |
| ; RV64ZVE32-NEXT: lbu a2, 93(a0) |
| ; RV64ZVE32-NEXT: lbu a3, 105(a0) |
| ; RV64ZVE32-NEXT: lbu a4, 124(a0) |
| ; RV64ZVE32-NEXT: lbu a5, 144(a0) |
| ; RV64ZVE32-NEXT: lbu a6, 154(a0) |
| ; RV64ZVE32-NEXT: lbu a7, 161(a0) |
| ; RV64ZVE32-NEXT: lbu a0, 163(a0) |
| ; RV64ZVE32-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; RV64ZVE32-NEXT: vmv.v.x v8, a1 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a7 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a4 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a6 |
| ; RV64ZVE32-NEXT: ret |
| %p9 = getelementptr i8, ptr %p, i32 82 |
| %p10 = getelementptr i8, ptr %p, i32 93 |
| %p11 = getelementptr i8, ptr %p, i32 105 |
| %p12 = getelementptr i8, ptr %p, i32 161 |
| %p13 = getelementptr i8, ptr %p, i32 124 |
| %p14 = getelementptr i8, ptr %p, i32 163 |
| %p15 = getelementptr i8, ptr %p, i32 144 |
| %p16 = getelementptr i8, ptr %p, i32 154 |
| |
| %ld9 = load i8, ptr %p9 |
| %ld10 = load i8, ptr %p10 |
| %ld11 = load i8, ptr %p11 |
| %ld12 = load i8, ptr %p12 |
| %ld13 = load i8, ptr %p13 |
| %ld14 = load i8, ptr %p14 |
| %ld15 = load i8, ptr %p15 |
| %ld16 = load i8, ptr %p16 |
| |
| %v9 = insertelement <16 x i8> poison, i8 %ld9, i32 8 |
| %v10 = insertelement <16 x i8> %v9, i8 %ld10, i32 9 |
| %v11 = insertelement <16 x i8> %v10, i8 %ld11, i32 10 |
| %v12 = insertelement <16 x i8> %v11, i8 %ld12, i32 11 |
| %v13 = insertelement <16 x i8> %v12, i8 %ld13, i32 12 |
| %v14 = insertelement <16 x i8> %v13, i8 %ld14, i32 13 |
| %v15 = insertelement <16 x i8> %v14, i8 %ld15, i32 14 |
| %v16 = insertelement <16 x i8> %v15, i8 %ld16, i32 15 |
| ret <16 x i8> %v16 |
| } |
| |
| define <16 x i8> @buildvec_v16i8_undef_high_half(ptr %p) { |
| ; RV32-ONLY-LABEL: buildvec_v16i8_undef_high_half: |
| ; RV32-ONLY: # %bb.0: |
| ; RV32-ONLY-NEXT: lbu a1, 0(a0) |
| ; RV32-ONLY-NEXT: lbu a2, 1(a0) |
| ; RV32-ONLY-NEXT: lbu a3, 22(a0) |
| ; RV32-ONLY-NEXT: lbu a4, 31(a0) |
| ; RV32-ONLY-NEXT: lbu a5, 623(a0) |
| ; RV32-ONLY-NEXT: lbu a6, 44(a0) |
| ; RV32-ONLY-NEXT: lbu a7, 55(a0) |
| ; RV32-ONLY-NEXT: lbu a0, 75(a0) |
| ; RV32-ONLY-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; RV32-ONLY-NEXT: vmv.v.x v8, a1 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a4 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a6 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a7 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32-ONLY-NEXT: vslidedown.vi v8, v8, 8 |
| ; RV32-ONLY-NEXT: ret |
| ; |
| ; RV32VB-LABEL: buildvec_v16i8_undef_high_half: |
| ; RV32VB: # %bb.0: |
| ; RV32VB-NEXT: lbu a1, 1(a0) |
| ; RV32VB-NEXT: lbu a2, 0(a0) |
| ; RV32VB-NEXT: lbu a3, 22(a0) |
| ; RV32VB-NEXT: lbu a4, 31(a0) |
| ; RV32VB-NEXT: slli a1, a1, 8 |
| ; RV32VB-NEXT: or a1, a2, a1 |
| ; RV32VB-NEXT: lbu a2, 44(a0) |
| ; RV32VB-NEXT: lbu a5, 55(a0) |
| ; RV32VB-NEXT: slli a3, a3, 16 |
| ; RV32VB-NEXT: slli a4, a4, 24 |
| ; RV32VB-NEXT: or a3, a4, a3 |
| ; RV32VB-NEXT: lbu a4, 623(a0) |
| ; RV32VB-NEXT: lbu a0, 75(a0) |
| ; RV32VB-NEXT: slli a5, a5, 8 |
| ; RV32VB-NEXT: or a2, a2, a5 |
| ; RV32VB-NEXT: slli a4, a4, 16 |
| ; RV32VB-NEXT: slli a0, a0, 24 |
| ; RV32VB-NEXT: or a0, a0, a4 |
| ; RV32VB-NEXT: or a1, a1, a3 |
| ; RV32VB-NEXT: or a0, a2, a0 |
| ; RV32VB-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32VB-NEXT: vmv.v.x v8, a1 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, zero |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, zero |
| ; RV32VB-NEXT: ret |
| ; |
| ; RV32VB-PACK-LABEL: buildvec_v16i8_undef_high_half: |
| ; RV32VB-PACK: # %bb.0: |
| ; RV32VB-PACK-NEXT: lbu a1, 0(a0) |
| ; RV32VB-PACK-NEXT: lbu a2, 1(a0) |
| ; RV32VB-PACK-NEXT: lbu a3, 22(a0) |
| ; RV32VB-PACK-NEXT: lbu a4, 31(a0) |
| ; RV32VB-PACK-NEXT: lbu a5, 623(a0) |
| ; RV32VB-PACK-NEXT: lbu a6, 44(a0) |
| ; RV32VB-PACK-NEXT: lbu a7, 55(a0) |
| ; RV32VB-PACK-NEXT: lbu a0, 75(a0) |
| ; RV32VB-PACK-NEXT: packh a1, a1, a2 |
| ; RV32VB-PACK-NEXT: packh a2, a3, a4 |
| ; RV32VB-PACK-NEXT: packh a3, a6, a7 |
| ; RV32VB-PACK-NEXT: packh a0, a5, a0 |
| ; RV32VB-PACK-NEXT: pack a1, a1, a2 |
| ; RV32VB-PACK-NEXT: packh a2, a0, a0 |
| ; RV32VB-PACK-NEXT: pack a0, a3, a0 |
| ; RV32VB-PACK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32VB-PACK-NEXT: vmv.v.x v8, a1 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-PACK-NEXT: pack a0, a2, a2 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-PACK-NEXT: ret |
| ; |
| ; RV64V-ONLY-LABEL: buildvec_v16i8_undef_high_half: |
| ; RV64V-ONLY: # %bb.0: |
| ; RV64V-ONLY-NEXT: lbu a1, 0(a0) |
| ; RV64V-ONLY-NEXT: lbu a2, 1(a0) |
| ; RV64V-ONLY-NEXT: lbu a3, 22(a0) |
| ; RV64V-ONLY-NEXT: lbu a4, 31(a0) |
| ; RV64V-ONLY-NEXT: lbu a5, 623(a0) |
| ; RV64V-ONLY-NEXT: lbu a6, 44(a0) |
| ; RV64V-ONLY-NEXT: lbu a7, 55(a0) |
| ; RV64V-ONLY-NEXT: lbu a0, 75(a0) |
| ; RV64V-ONLY-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.v.x v8, a1 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a4 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a6 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a7 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64V-ONLY-NEXT: vslidedown.vi v8, v8, 8 |
| ; RV64V-ONLY-NEXT: ret |
| ; |
| ; RVA22U64-LABEL: buildvec_v16i8_undef_high_half: |
| ; RVA22U64: # %bb.0: |
| ; RVA22U64-NEXT: lbu a1, 1(a0) |
| ; RVA22U64-NEXT: lbu a2, 0(a0) |
| ; RVA22U64-NEXT: lbu a3, 22(a0) |
| ; RVA22U64-NEXT: lbu a4, 31(a0) |
| ; RVA22U64-NEXT: slli a1, a1, 8 |
| ; RVA22U64-NEXT: or a1, a1, a2 |
| ; RVA22U64-NEXT: lbu a2, 44(a0) |
| ; RVA22U64-NEXT: lbu a5, 55(a0) |
| ; RVA22U64-NEXT: slli a3, a3, 16 |
| ; RVA22U64-NEXT: slli a4, a4, 24 |
| ; RVA22U64-NEXT: or a3, a3, a4 |
| ; RVA22U64-NEXT: lbu a4, 623(a0) |
| ; RVA22U64-NEXT: lbu a0, 75(a0) |
| ; RVA22U64-NEXT: slli a2, a2, 32 |
| ; RVA22U64-NEXT: slli a5, a5, 40 |
| ; RVA22U64-NEXT: or a2, a2, a5 |
| ; RVA22U64-NEXT: slli a4, a4, 48 |
| ; RVA22U64-NEXT: slli a0, a0, 56 |
| ; RVA22U64-NEXT: or a0, a0, a4 |
| ; RVA22U64-NEXT: or a1, a1, a3 |
| ; RVA22U64-NEXT: or a0, a0, a2 |
| ; RVA22U64-NEXT: or a0, a0, a1 |
| ; RVA22U64-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-NEXT: vmv.v.x v8, a0 |
| ; RVA22U64-NEXT: vslide1down.vx v8, v8, zero |
| ; RVA22U64-NEXT: ret |
| ; |
| ; RVA22U64-PACK-LABEL: buildvec_v16i8_undef_high_half: |
| ; RVA22U64-PACK: # %bb.0: |
| ; RVA22U64-PACK-NEXT: lbu a6, 0(a0) |
| ; RVA22U64-PACK-NEXT: lbu a7, 1(a0) |
| ; RVA22U64-PACK-NEXT: lbu t0, 22(a0) |
| ; RVA22U64-PACK-NEXT: lbu a4, 31(a0) |
| ; RVA22U64-PACK-NEXT: lbu a5, 623(a0) |
| ; RVA22U64-PACK-NEXT: lbu a1, 44(a0) |
| ; RVA22U64-PACK-NEXT: lbu a2, 55(a0) |
| ; RVA22U64-PACK-NEXT: lbu a0, 75(a0) |
| ; RVA22U64-PACK-NEXT: packh a3, a6, a7 |
| ; RVA22U64-PACK-NEXT: packh a4, t0, a4 |
| ; RVA22U64-PACK-NEXT: packh a1, a1, a2 |
| ; RVA22U64-PACK-NEXT: packh a0, a5, a0 |
| ; RVA22U64-PACK-NEXT: packw a2, a3, a4 |
| ; RVA22U64-PACK-NEXT: packh a3, a0, a0 |
| ; RVA22U64-PACK-NEXT: packw a3, a3, a3 |
| ; RVA22U64-PACK-NEXT: packw a0, a1, a0 |
| ; RVA22U64-PACK-NEXT: pack a0, a2, a0 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.v.x v8, a0 |
| ; RVA22U64-PACK-NEXT: pack a0, a3, a3 |
| ; RVA22U64-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RVA22U64-PACK-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_v16i8_undef_high_half: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: lbu a1, 0(a0) |
| ; RV64ZVE32-NEXT: lbu a2, 1(a0) |
| ; RV64ZVE32-NEXT: lbu a3, 22(a0) |
| ; RV64ZVE32-NEXT: lbu a4, 31(a0) |
| ; RV64ZVE32-NEXT: lbu a5, 623(a0) |
| ; RV64ZVE32-NEXT: lbu a6, 44(a0) |
| ; RV64ZVE32-NEXT: lbu a7, 55(a0) |
| ; RV64ZVE32-NEXT: lbu a0, 75(a0) |
| ; RV64ZVE32-NEXT: vsetivli zero, 16, e8, m1, ta, ma |
| ; RV64ZVE32-NEXT: vmv.v.x v8, a1 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a4 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a6 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a7 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64ZVE32-NEXT: vslidedown.vi v8, v8, 8 |
| ; RV64ZVE32-NEXT: ret |
| %p2 = getelementptr i8, ptr %p, i32 1 |
| %p3 = getelementptr i8, ptr %p, i32 22 |
| %p4 = getelementptr i8, ptr %p, i32 31 |
| %p5 = getelementptr i8, ptr %p, i32 44 |
| %p6 = getelementptr i8, ptr %p, i32 55 |
| %p7 = getelementptr i8, ptr %p, i32 623 |
| %p8 = getelementptr i8, ptr %p, i32 75 |
| |
| %ld1 = load i8, ptr %p |
| %ld2 = load i8, ptr %p2 |
| %ld3 = load i8, ptr %p3 |
| %ld4 = load i8, ptr %p4 |
| %ld5 = load i8, ptr %p5 |
| %ld6 = load i8, ptr %p6 |
| %ld7 = load i8, ptr %p7 |
| %ld8 = load i8, ptr %p8 |
| |
| %v1 = insertelement <16 x i8> poison, i8 %ld1, i32 0 |
| %v2 = insertelement <16 x i8> %v1, i8 %ld2, i32 1 |
| %v3 = insertelement <16 x i8> %v2, i8 %ld3, i32 2 |
| %v4 = insertelement <16 x i8> %v3, i8 %ld4, i32 3 |
| %v5 = insertelement <16 x i8> %v4, i8 %ld5, i32 4 |
| %v6 = insertelement <16 x i8> %v5, i8 %ld6, i32 5 |
| %v7 = insertelement <16 x i8> %v6, i8 %ld7, i32 6 |
| %v8 = insertelement <16 x i8> %v7, i8 %ld8, i32 7 |
| ret <16 x i8> %v8 |
| } |
| |
| define <16 x i8> @buildvec_v16i8_undef_edges(ptr %p) { |
| ; RV32-ONLY-LABEL: buildvec_v16i8_undef_edges: |
| ; RV32-ONLY: # %bb.0: |
| ; RV32-ONLY-NEXT: lbu a1, 623(a0) |
| ; RV32-ONLY-NEXT: lbu a2, 31(a0) |
| ; RV32-ONLY-NEXT: lbu a3, 44(a0) |
| ; RV32-ONLY-NEXT: lbu a4, 55(a0) |
| ; RV32-ONLY-NEXT: lbu a5, 75(a0) |
| ; RV32-ONLY-NEXT: li a6, 255 |
| ; RV32-ONLY-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV32-ONLY-NEXT: vmv.s.x v0, a6 |
| ; RV32-ONLY-NEXT: lbu a6, 82(a0) |
| ; RV32-ONLY-NEXT: lbu a7, 93(a0) |
| ; RV32-ONLY-NEXT: lbu t0, 105(a0) |
| ; RV32-ONLY-NEXT: lbu a0, 161(a0) |
| ; RV32-ONLY-NEXT: vsetivli zero, 16, e8, m1, ta, mu |
| ; RV32-ONLY-NEXT: vmv.v.x v8, a6 |
| ; RV32-ONLY-NEXT: vmv.v.x v9, a2 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a7 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a3 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, t0 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a4 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a1 |
| ; RV32-ONLY-NEXT: vslidedown.vi v8, v8, 4 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a5 |
| ; RV32-ONLY-NEXT: vslidedown.vi v8, v9, 8, v0.t |
| ; RV32-ONLY-NEXT: ret |
| ; |
| ; RV32VB-LABEL: buildvec_v16i8_undef_edges: |
| ; RV32VB: # %bb.0: |
| ; RV32VB-NEXT: lbu a1, 623(a0) |
| ; RV32VB-NEXT: lbu a2, 55(a0) |
| ; RV32VB-NEXT: lbu a3, 31(a0) |
| ; RV32VB-NEXT: lbu a4, 75(a0) |
| ; RV32VB-NEXT: lbu a5, 44(a0) |
| ; RV32VB-NEXT: slli a2, a2, 8 |
| ; RV32VB-NEXT: slli a1, a1, 16 |
| ; RV32VB-NEXT: slli a4, a4, 24 |
| ; RV32VB-NEXT: or a2, a5, a2 |
| ; RV32VB-NEXT: or a1, a4, a1 |
| ; RV32VB-NEXT: lbu a4, 93(a0) |
| ; RV32VB-NEXT: lbu a5, 82(a0) |
| ; RV32VB-NEXT: lbu a6, 105(a0) |
| ; RV32VB-NEXT: lbu a0, 161(a0) |
| ; RV32VB-NEXT: slli a4, a4, 8 |
| ; RV32VB-NEXT: or a4, a5, a4 |
| ; RV32VB-NEXT: slli a6, a6, 16 |
| ; RV32VB-NEXT: slli a0, a0, 24 |
| ; RV32VB-NEXT: or a0, a0, a6 |
| ; RV32VB-NEXT: slli a3, a3, 24 |
| ; RV32VB-NEXT: or a1, a2, a1 |
| ; RV32VB-NEXT: or a0, a4, a0 |
| ; RV32VB-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32VB-NEXT: vmv.v.x v8, a3 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, zero |
| ; RV32VB-NEXT: ret |
| ; |
| ; RV32VB-PACK-LABEL: buildvec_v16i8_undef_edges: |
| ; RV32VB-PACK: # %bb.0: |
| ; RV32VB-PACK-NEXT: lbu a1, 623(a0) |
| ; RV32VB-PACK-NEXT: lbu a2, 31(a0) |
| ; RV32VB-PACK-NEXT: lbu a3, 44(a0) |
| ; RV32VB-PACK-NEXT: lbu a4, 55(a0) |
| ; RV32VB-PACK-NEXT: lbu a5, 75(a0) |
| ; RV32VB-PACK-NEXT: lbu a6, 82(a0) |
| ; RV32VB-PACK-NEXT: lbu a7, 93(a0) |
| ; RV32VB-PACK-NEXT: lbu t0, 105(a0) |
| ; RV32VB-PACK-NEXT: lbu a0, 161(a0) |
| ; RV32VB-PACK-NEXT: packh a3, a3, a4 |
| ; RV32VB-PACK-NEXT: packh a1, a1, a5 |
| ; RV32VB-PACK-NEXT: packh a4, a6, a7 |
| ; RV32VB-PACK-NEXT: packh a0, t0, a0 |
| ; RV32VB-PACK-NEXT: packh a5, a0, a0 |
| ; RV32VB-PACK-NEXT: packh a2, a0, a2 |
| ; RV32VB-PACK-NEXT: pack a2, a5, a2 |
| ; RV32VB-PACK-NEXT: pack a1, a3, a1 |
| ; RV32VB-PACK-NEXT: pack a0, a4, a0 |
| ; RV32VB-PACK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32VB-PACK-NEXT: vmv.v.x v8, a2 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-PACK-NEXT: pack a0, a5, a5 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-PACK-NEXT: ret |
| ; |
| ; RV64V-ONLY-LABEL: buildvec_v16i8_undef_edges: |
| ; RV64V-ONLY: # %bb.0: |
| ; RV64V-ONLY-NEXT: lbu a1, 623(a0) |
| ; RV64V-ONLY-NEXT: lbu a2, 31(a0) |
| ; RV64V-ONLY-NEXT: lbu a3, 44(a0) |
| ; RV64V-ONLY-NEXT: lbu a4, 55(a0) |
| ; RV64V-ONLY-NEXT: lbu a5, 75(a0) |
| ; RV64V-ONLY-NEXT: li a6, 255 |
| ; RV64V-ONLY-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.s.x v0, a6 |
| ; RV64V-ONLY-NEXT: lbu a6, 82(a0) |
| ; RV64V-ONLY-NEXT: lbu a7, 93(a0) |
| ; RV64V-ONLY-NEXT: lbu t0, 105(a0) |
| ; RV64V-ONLY-NEXT: lbu a0, 161(a0) |
| ; RV64V-ONLY-NEXT: vsetivli zero, 16, e8, m1, ta, mu |
| ; RV64V-ONLY-NEXT: vmv.v.x v8, a6 |
| ; RV64V-ONLY-NEXT: vmv.v.x v9, a2 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a7 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a3 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, t0 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a4 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a1 |
| ; RV64V-ONLY-NEXT: vslidedown.vi v8, v8, 4 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a5 |
| ; RV64V-ONLY-NEXT: vslidedown.vi v8, v9, 8, v0.t |
| ; RV64V-ONLY-NEXT: ret |
| ; |
| ; RVA22U64-LABEL: buildvec_v16i8_undef_edges: |
| ; RVA22U64: # %bb.0: |
| ; RVA22U64-NEXT: lbu a1, 623(a0) |
| ; RVA22U64-NEXT: lbu a2, 44(a0) |
| ; RVA22U64-NEXT: lbu a3, 55(a0) |
| ; RVA22U64-NEXT: lbu a6, 31(a0) |
| ; RVA22U64-NEXT: lbu a5, 75(a0) |
| ; RVA22U64-NEXT: slli a2, a2, 32 |
| ; RVA22U64-NEXT: slli a3, a3, 40 |
| ; RVA22U64-NEXT: slli a1, a1, 48 |
| ; RVA22U64-NEXT: slli a5, a5, 56 |
| ; RVA22U64-NEXT: or a2, a2, a3 |
| ; RVA22U64-NEXT: or a1, a1, a5 |
| ; RVA22U64-NEXT: lbu a3, 93(a0) |
| ; RVA22U64-NEXT: lbu a5, 82(a0) |
| ; RVA22U64-NEXT: lbu a4, 105(a0) |
| ; RVA22U64-NEXT: lbu a0, 161(a0) |
| ; RVA22U64-NEXT: slli a3, a3, 8 |
| ; RVA22U64-NEXT: or a3, a3, a5 |
| ; RVA22U64-NEXT: slli a4, a4, 16 |
| ; RVA22U64-NEXT: slli a0, a0, 24 |
| ; RVA22U64-NEXT: or a0, a0, a4 |
| ; RVA22U64-NEXT: slli a6, a6, 24 |
| ; RVA22U64-NEXT: or a1, a1, a2 |
| ; RVA22U64-NEXT: add.uw a1, a6, a1 |
| ; RVA22U64-NEXT: or a0, a0, a3 |
| ; RVA22U64-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-NEXT: vmv.v.x v8, a1 |
| ; RVA22U64-NEXT: vslide1down.vx v8, v8, a0 |
| ; RVA22U64-NEXT: ret |
| ; |
| ; RVA22U64-PACK-LABEL: buildvec_v16i8_undef_edges: |
| ; RVA22U64-PACK: # %bb.0: |
| ; RVA22U64-PACK-NEXT: lbu a7, 623(a0) |
| ; RVA22U64-PACK-NEXT: lbu a6, 31(a0) |
| ; RVA22U64-PACK-NEXT: lbu t0, 44(a0) |
| ; RVA22U64-PACK-NEXT: lbu a4, 55(a0) |
| ; RVA22U64-PACK-NEXT: lbu a5, 75(a0) |
| ; RVA22U64-PACK-NEXT: lbu a2, 82(a0) |
| ; RVA22U64-PACK-NEXT: lbu a1, 93(a0) |
| ; RVA22U64-PACK-NEXT: lbu a3, 105(a0) |
| ; RVA22U64-PACK-NEXT: lbu a0, 161(a0) |
| ; RVA22U64-PACK-NEXT: packh a4, t0, a4 |
| ; RVA22U64-PACK-NEXT: packh a5, a7, a5 |
| ; RVA22U64-PACK-NEXT: packh a1, a2, a1 |
| ; RVA22U64-PACK-NEXT: packh a0, a3, a0 |
| ; RVA22U64-PACK-NEXT: packh a2, a0, a0 |
| ; RVA22U64-PACK-NEXT: packh a3, a0, a6 |
| ; RVA22U64-PACK-NEXT: packw a3, a2, a3 |
| ; RVA22U64-PACK-NEXT: packw a2, a2, a2 |
| ; RVA22U64-PACK-NEXT: packw a4, a4, a5 |
| ; RVA22U64-PACK-NEXT: packw a0, a1, a0 |
| ; RVA22U64-PACK-NEXT: pack a1, a3, a4 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.v.x v8, a1 |
| ; RVA22U64-PACK-NEXT: pack a0, a0, a2 |
| ; RVA22U64-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RVA22U64-PACK-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_v16i8_undef_edges: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: lbu a1, 623(a0) |
| ; RV64ZVE32-NEXT: lbu a2, 31(a0) |
| ; RV64ZVE32-NEXT: lbu a3, 44(a0) |
| ; RV64ZVE32-NEXT: lbu a4, 55(a0) |
| ; RV64ZVE32-NEXT: lbu a5, 75(a0) |
| ; RV64ZVE32-NEXT: li a6, 255 |
| ; RV64ZVE32-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV64ZVE32-NEXT: vmv.s.x v0, a6 |
| ; RV64ZVE32-NEXT: lbu a6, 82(a0) |
| ; RV64ZVE32-NEXT: lbu a7, 93(a0) |
| ; RV64ZVE32-NEXT: lbu t0, 105(a0) |
| ; RV64ZVE32-NEXT: lbu a0, 161(a0) |
| ; RV64ZVE32-NEXT: vsetivli zero, 16, e8, m1, ta, mu |
| ; RV64ZVE32-NEXT: vmv.v.x v8, a6 |
| ; RV64ZVE32-NEXT: vmv.v.x v9, a2 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a7 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a3 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, t0 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a4 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a1 |
| ; RV64ZVE32-NEXT: vslidedown.vi v8, v8, 4 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a5 |
| ; RV64ZVE32-NEXT: vslidedown.vi v8, v9, 8, v0.t |
| ; RV64ZVE32-NEXT: ret |
| %p4 = getelementptr i8, ptr %p, i32 31 |
| %p5 = getelementptr i8, ptr %p, i32 44 |
| %p6 = getelementptr i8, ptr %p, i32 55 |
| %p7 = getelementptr i8, ptr %p, i32 623 |
| %p8 = getelementptr i8, ptr %p, i32 75 |
| %p9 = getelementptr i8, ptr %p, i32 82 |
| %p10 = getelementptr i8, ptr %p, i32 93 |
| %p11 = getelementptr i8, ptr %p, i32 105 |
| %p12 = getelementptr i8, ptr %p, i32 161 |
| |
| %ld4 = load i8, ptr %p4 |
| %ld5 = load i8, ptr %p5 |
| %ld6 = load i8, ptr %p6 |
| %ld7 = load i8, ptr %p7 |
| %ld8 = load i8, ptr %p8 |
| %ld9 = load i8, ptr %p9 |
| %ld10 = load i8, ptr %p10 |
| %ld11 = load i8, ptr %p11 |
| %ld12 = load i8, ptr %p12 |
| |
| %v4 = insertelement <16 x i8> poison, i8 %ld4, i32 3 |
| %v5 = insertelement <16 x i8> %v4, i8 %ld5, i32 4 |
| %v6 = insertelement <16 x i8> %v5, i8 %ld6, i32 5 |
| %v7 = insertelement <16 x i8> %v6, i8 %ld7, i32 6 |
| %v8 = insertelement <16 x i8> %v7, i8 %ld8, i32 7 |
| %v9 = insertelement <16 x i8> %v8, i8 %ld9, i32 8 |
| %v10 = insertelement <16 x i8> %v9, i8 %ld10, i32 9 |
| %v11 = insertelement <16 x i8> %v10, i8 %ld11, i32 10 |
| %v12 = insertelement <16 x i8> %v11, i8 %ld12, i32 11 |
| ret <16 x i8> %v12 |
| } |
| |
| define <16 x i8> @buildvec_v16i8_loads_undef_scattered(ptr %p) { |
| ; RV32-ONLY-LABEL: buildvec_v16i8_loads_undef_scattered: |
| ; RV32-ONLY: # %bb.0: |
| ; RV32-ONLY-NEXT: lbu a1, 0(a0) |
| ; RV32-ONLY-NEXT: lbu a2, 1(a0) |
| ; RV32-ONLY-NEXT: lbu a3, 44(a0) |
| ; RV32-ONLY-NEXT: lbu a4, 55(a0) |
| ; RV32-ONLY-NEXT: lbu a5, 75(a0) |
| ; RV32-ONLY-NEXT: lbu a6, 82(a0) |
| ; RV32-ONLY-NEXT: lbu a7, 93(a0) |
| ; RV32-ONLY-NEXT: lbu t0, 124(a0) |
| ; RV32-ONLY-NEXT: li t1, 255 |
| ; RV32-ONLY-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV32-ONLY-NEXT: vmv.s.x v0, t1 |
| ; RV32-ONLY-NEXT: lbu t1, 144(a0) |
| ; RV32-ONLY-NEXT: lbu a0, 154(a0) |
| ; RV32-ONLY-NEXT: vsetivli zero, 16, e8, m1, ta, mu |
| ; RV32-ONLY-NEXT: vmv.v.x v8, a6 |
| ; RV32-ONLY-NEXT: vmv.v.x v9, a1 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a7 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a2 |
| ; RV32-ONLY-NEXT: vslidedown.vi v8, v8, 2 |
| ; RV32-ONLY-NEXT: vslidedown.vi v9, v9, 2 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, t0 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a3 |
| ; RV32-ONLY-NEXT: vslidedown.vi v8, v8, 1 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a4 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, t1 |
| ; RV32-ONLY-NEXT: vslidedown.vi v9, v9, 1 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a5 |
| ; RV32-ONLY-NEXT: vslidedown.vi v8, v9, 8, v0.t |
| ; RV32-ONLY-NEXT: ret |
| ; |
| ; RV32VB-LABEL: buildvec_v16i8_loads_undef_scattered: |
| ; RV32VB: # %bb.0: |
| ; RV32VB-NEXT: lbu a1, 1(a0) |
| ; RV32VB-NEXT: lbu a2, 0(a0) |
| ; RV32VB-NEXT: lbu a3, 44(a0) |
| ; RV32VB-NEXT: lbu a4, 55(a0) |
| ; RV32VB-NEXT: slli a1, a1, 8 |
| ; RV32VB-NEXT: or a1, a2, a1 |
| ; RV32VB-NEXT: lbu a2, 75(a0) |
| ; RV32VB-NEXT: lbu a5, 82(a0) |
| ; RV32VB-NEXT: lbu a6, 93(a0) |
| ; RV32VB-NEXT: lbu a7, 124(a0) |
| ; RV32VB-NEXT: slli a4, a4, 8 |
| ; RV32VB-NEXT: or a3, a3, a4 |
| ; RV32VB-NEXT: lbu a4, 144(a0) |
| ; RV32VB-NEXT: lbu a0, 154(a0) |
| ; RV32VB-NEXT: slli a6, a6, 8 |
| ; RV32VB-NEXT: or a5, a5, a6 |
| ; RV32VB-NEXT: slli a4, a4, 16 |
| ; RV32VB-NEXT: slli a0, a0, 24 |
| ; RV32VB-NEXT: or a0, a0, a4 |
| ; RV32VB-NEXT: slli a2, a2, 24 |
| ; RV32VB-NEXT: or a2, a3, a2 |
| ; RV32VB-NEXT: or a0, a7, a0 |
| ; RV32VB-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32VB-NEXT: vmv.v.x v8, a1 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-NEXT: ret |
| ; |
| ; RV32VB-PACK-LABEL: buildvec_v16i8_loads_undef_scattered: |
| ; RV32VB-PACK: # %bb.0: |
| ; RV32VB-PACK-NEXT: lbu a1, 0(a0) |
| ; RV32VB-PACK-NEXT: lbu a2, 1(a0) |
| ; RV32VB-PACK-NEXT: lbu a3, 44(a0) |
| ; RV32VB-PACK-NEXT: lbu a4, 55(a0) |
| ; RV32VB-PACK-NEXT: lbu a5, 82(a0) |
| ; RV32VB-PACK-NEXT: lbu a6, 93(a0) |
| ; RV32VB-PACK-NEXT: packh a1, a1, a2 |
| ; RV32VB-PACK-NEXT: lbu a2, 144(a0) |
| ; RV32VB-PACK-NEXT: lbu a7, 154(a0) |
| ; RV32VB-PACK-NEXT: packh a3, a3, a4 |
| ; RV32VB-PACK-NEXT: lbu a4, 75(a0) |
| ; RV32VB-PACK-NEXT: lbu a0, 124(a0) |
| ; RV32VB-PACK-NEXT: packh a5, a5, a6 |
| ; RV32VB-PACK-NEXT: packh a2, a2, a7 |
| ; RV32VB-PACK-NEXT: packh a4, a0, a4 |
| ; RV32VB-PACK-NEXT: pack a3, a3, a4 |
| ; RV32VB-PACK-NEXT: packh a4, a0, a0 |
| ; RV32VB-PACK-NEXT: packh a0, a0, a0 |
| ; RV32VB-PACK-NEXT: pack a0, a0, a2 |
| ; RV32VB-PACK-NEXT: pack a1, a1, a4 |
| ; RV32VB-PACK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; RV32VB-PACK-NEXT: vmv.v.x v8, a1 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV32VB-PACK-NEXT: pack a1, a5, a4 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-PACK-NEXT: ret |
| ; |
| ; RV64V-ONLY-LABEL: buildvec_v16i8_loads_undef_scattered: |
| ; RV64V-ONLY: # %bb.0: |
| ; RV64V-ONLY-NEXT: lbu a1, 0(a0) |
| ; RV64V-ONLY-NEXT: lbu a2, 1(a0) |
| ; RV64V-ONLY-NEXT: lbu a3, 44(a0) |
| ; RV64V-ONLY-NEXT: lbu a4, 55(a0) |
| ; RV64V-ONLY-NEXT: lbu a5, 75(a0) |
| ; RV64V-ONLY-NEXT: lbu a6, 82(a0) |
| ; RV64V-ONLY-NEXT: lbu a7, 93(a0) |
| ; RV64V-ONLY-NEXT: lbu t0, 124(a0) |
| ; RV64V-ONLY-NEXT: li t1, 255 |
| ; RV64V-ONLY-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.s.x v0, t1 |
| ; RV64V-ONLY-NEXT: lbu t1, 144(a0) |
| ; RV64V-ONLY-NEXT: lbu a0, 154(a0) |
| ; RV64V-ONLY-NEXT: vsetivli zero, 16, e8, m1, ta, mu |
| ; RV64V-ONLY-NEXT: vmv.v.x v8, a6 |
| ; RV64V-ONLY-NEXT: vmv.v.x v9, a1 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a7 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a2 |
| ; RV64V-ONLY-NEXT: vslidedown.vi v8, v8, 2 |
| ; RV64V-ONLY-NEXT: vslidedown.vi v9, v9, 2 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, t0 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a3 |
| ; RV64V-ONLY-NEXT: vslidedown.vi v8, v8, 1 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a4 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, t1 |
| ; RV64V-ONLY-NEXT: vslidedown.vi v9, v9, 1 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a5 |
| ; RV64V-ONLY-NEXT: vslidedown.vi v8, v9, 8, v0.t |
| ; RV64V-ONLY-NEXT: ret |
| ; |
| ; RVA22U64-LABEL: buildvec_v16i8_loads_undef_scattered: |
| ; RVA22U64: # %bb.0: |
| ; RVA22U64-NEXT: lbu a1, 1(a0) |
| ; RVA22U64-NEXT: lbu a2, 0(a0) |
| ; RVA22U64-NEXT: lbu a3, 44(a0) |
| ; RVA22U64-NEXT: lbu a4, 55(a0) |
| ; RVA22U64-NEXT: slli a1, a1, 8 |
| ; RVA22U64-NEXT: or a6, a2, a1 |
| ; RVA22U64-NEXT: lbu a7, 75(a0) |
| ; RVA22U64-NEXT: lbu a5, 82(a0) |
| ; RVA22U64-NEXT: lbu a1, 93(a0) |
| ; RVA22U64-NEXT: lbu a2, 124(a0) |
| ; RVA22U64-NEXT: slli a3, a3, 32 |
| ; RVA22U64-NEXT: slli a4, a4, 40 |
| ; RVA22U64-NEXT: or a3, a3, a4 |
| ; RVA22U64-NEXT: lbu a4, 144(a0) |
| ; RVA22U64-NEXT: lbu a0, 154(a0) |
| ; RVA22U64-NEXT: slli a1, a1, 8 |
| ; RVA22U64-NEXT: or a1, a1, a5 |
| ; RVA22U64-NEXT: slli a4, a4, 48 |
| ; RVA22U64-NEXT: slli a0, a0, 56 |
| ; RVA22U64-NEXT: or a0, a0, a4 |
| ; RVA22U64-NEXT: slli a7, a7, 56 |
| ; RVA22U64-NEXT: or a3, a7, a3 |
| ; RVA22U64-NEXT: slli a2, a2, 32 |
| ; RVA22U64-NEXT: or a0, a0, a2 |
| ; RVA22U64-NEXT: or a2, a6, a3 |
| ; RVA22U64-NEXT: or a0, a0, a1 |
| ; RVA22U64-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-NEXT: vmv.v.x v8, a2 |
| ; RVA22U64-NEXT: vslide1down.vx v8, v8, a0 |
| ; RVA22U64-NEXT: ret |
| ; |
| ; RVA22U64-PACK-LABEL: buildvec_v16i8_loads_undef_scattered: |
| ; RVA22U64-PACK: # %bb.0: |
| ; RVA22U64-PACK-NEXT: lbu a1, 0(a0) |
| ; RVA22U64-PACK-NEXT: lbu a2, 1(a0) |
| ; RVA22U64-PACK-NEXT: lbu a6, 44(a0) |
| ; RVA22U64-PACK-NEXT: lbu a7, 55(a0) |
| ; RVA22U64-PACK-NEXT: lbu t1, 82(a0) |
| ; RVA22U64-PACK-NEXT: lbu a3, 93(a0) |
| ; RVA22U64-PACK-NEXT: packh t0, a1, a2 |
| ; RVA22U64-PACK-NEXT: lbu a2, 144(a0) |
| ; RVA22U64-PACK-NEXT: lbu a4, 154(a0) |
| ; RVA22U64-PACK-NEXT: packh a1, a6, a7 |
| ; RVA22U64-PACK-NEXT: lbu a5, 75(a0) |
| ; RVA22U64-PACK-NEXT: lbu a0, 124(a0) |
| ; RVA22U64-PACK-NEXT: packh a3, t1, a3 |
| ; RVA22U64-PACK-NEXT: packh a2, a2, a4 |
| ; RVA22U64-PACK-NEXT: packh a4, a0, a5 |
| ; RVA22U64-PACK-NEXT: packw a1, a1, a4 |
| ; RVA22U64-PACK-NEXT: packh a4, a0, a0 |
| ; RVA22U64-PACK-NEXT: packh a0, a0, a0 |
| ; RVA22U64-PACK-NEXT: packw a5, t0, a4 |
| ; RVA22U64-PACK-NEXT: packw a0, a0, a2 |
| ; RVA22U64-PACK-NEXT: packw a2, a3, a4 |
| ; RVA22U64-PACK-NEXT: pack a1, a5, a1 |
| ; RVA22U64-PACK-NEXT: pack a0, a2, a0 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.v.x v8, a1 |
| ; RVA22U64-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RVA22U64-PACK-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_v16i8_loads_undef_scattered: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: lbu a1, 0(a0) |
| ; RV64ZVE32-NEXT: lbu a2, 1(a0) |
| ; RV64ZVE32-NEXT: lbu a3, 44(a0) |
| ; RV64ZVE32-NEXT: lbu a4, 55(a0) |
| ; RV64ZVE32-NEXT: lbu a5, 75(a0) |
| ; RV64ZVE32-NEXT: lbu a6, 82(a0) |
| ; RV64ZVE32-NEXT: lbu a7, 93(a0) |
| ; RV64ZVE32-NEXT: lbu t0, 124(a0) |
| ; RV64ZVE32-NEXT: li t1, 255 |
| ; RV64ZVE32-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV64ZVE32-NEXT: vmv.s.x v0, t1 |
| ; RV64ZVE32-NEXT: lbu t1, 144(a0) |
| ; RV64ZVE32-NEXT: lbu a0, 154(a0) |
| ; RV64ZVE32-NEXT: vsetivli zero, 16, e8, m1, ta, mu |
| ; RV64ZVE32-NEXT: vmv.v.x v8, a6 |
| ; RV64ZVE32-NEXT: vmv.v.x v9, a1 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a7 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a2 |
| ; RV64ZVE32-NEXT: vslidedown.vi v8, v8, 2 |
| ; RV64ZVE32-NEXT: vslidedown.vi v9, v9, 2 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, t0 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a3 |
| ; RV64ZVE32-NEXT: vslidedown.vi v8, v8, 1 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a4 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, t1 |
| ; RV64ZVE32-NEXT: vslidedown.vi v9, v9, 1 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a5 |
| ; RV64ZVE32-NEXT: vslidedown.vi v8, v9, 8, v0.t |
| ; RV64ZVE32-NEXT: ret |
| %p2 = getelementptr i8, ptr %p, i32 1 |
| %p3 = getelementptr i8, ptr %p, i32 22 |
| %p4 = getelementptr i8, ptr %p, i32 31 |
| %p5 = getelementptr i8, ptr %p, i32 44 |
| %p6 = getelementptr i8, ptr %p, i32 55 |
| %p7 = getelementptr i8, ptr %p, i32 623 |
| %p8 = getelementptr i8, ptr %p, i32 75 |
| %p9 = getelementptr i8, ptr %p, i32 82 |
| %p10 = getelementptr i8, ptr %p, i32 93 |
| %p11 = getelementptr i8, ptr %p, i32 105 |
| %p12 = getelementptr i8, ptr %p, i32 161 |
| %p13 = getelementptr i8, ptr %p, i32 124 |
| %p14 = getelementptr i8, ptr %p, i32 163 |
| %p15 = getelementptr i8, ptr %p, i32 144 |
| %p16 = getelementptr i8, ptr %p, i32 154 |
| |
| %ld1 = load i8, ptr %p |
| %ld2 = load i8, ptr %p2 |
| %ld3 = load i8, ptr %p3 |
| %ld4 = load i8, ptr %p4 |
| %ld5 = load i8, ptr %p5 |
| %ld6 = load i8, ptr %p6 |
| %ld7 = load i8, ptr %p7 |
| %ld8 = load i8, ptr %p8 |
| %ld9 = load i8, ptr %p9 |
| %ld10 = load i8, ptr %p10 |
| %ld11 = load i8, ptr %p11 |
| %ld12 = load i8, ptr %p12 |
| %ld13 = load i8, ptr %p13 |
| %ld14 = load i8, ptr %p14 |
| %ld15 = load i8, ptr %p15 |
| %ld16 = load i8, ptr %p16 |
| |
| %v1 = insertelement <16 x i8> poison, i8 %ld1, i32 0 |
| %v2 = insertelement <16 x i8> %v1, i8 %ld2, i32 1 |
| %v3 = insertelement <16 x i8> %v2, i8 poison, i32 2 |
| %v4 = insertelement <16 x i8> %v3, i8 poison, i32 3 |
| %v5 = insertelement <16 x i8> %v4, i8 %ld5, i32 4 |
| %v6 = insertelement <16 x i8> %v5, i8 %ld6, i32 5 |
| %v7 = insertelement <16 x i8> %v6, i8 poison, i32 6 |
| %v8 = insertelement <16 x i8> %v7, i8 %ld8, i32 7 |
| %v9 = insertelement <16 x i8> %v8, i8 %ld9, i32 8 |
| %v10 = insertelement <16 x i8> %v9, i8 %ld10, i32 9 |
| %v11 = insertelement <16 x i8> %v10, i8 poison, i32 10 |
| %v12 = insertelement <16 x i8> %v11, i8 poison, i32 11 |
| %v13 = insertelement <16 x i8> %v12, i8 %ld13, i32 12 |
| %v14 = insertelement <16 x i8> %v13, i8 poison, i32 13 |
| %v15 = insertelement <16 x i8> %v14, i8 %ld15, i32 14 |
| %v16 = insertelement <16 x i8> %v15, i8 %ld16, i32 15 |
| ret <16 x i8> %v16 |
| } |
| |
| define <8 x i8> @buildvec_v8i8_pack(i8 %e1, i8 %e2, i8 %e3, i8 %e4, i8 %e5, i8 %e6, i8 %e7, i8 %e8) { |
| ; RV32-ONLY-LABEL: buildvec_v8i8_pack: |
| ; RV32-ONLY: # %bb.0: |
| ; RV32-ONLY-NEXT: vsetivli zero, 8, e8, mf2, ta, mu |
| ; RV32-ONLY-NEXT: vmv.v.x v8, a4 |
| ; RV32-ONLY-NEXT: vmv.v.x v9, a0 |
| ; RV32-ONLY-NEXT: vmv.v.i v0, 15 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a1 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a6 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a2 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a7 |
| ; RV32-ONLY-NEXT: vslide1down.vx v9, v9, a3 |
| ; RV32-ONLY-NEXT: vslidedown.vi v8, v9, 4, v0.t |
| ; RV32-ONLY-NEXT: ret |
| ; |
| ; RV32VB-LABEL: buildvec_v8i8_pack: |
| ; RV32VB: # %bb.0: |
| ; RV32VB-NEXT: slli a7, a7, 24 |
| ; RV32VB-NEXT: zext.b a6, a6 |
| ; RV32VB-NEXT: zext.b a4, a4 |
| ; RV32VB-NEXT: zext.b a5, a5 |
| ; RV32VB-NEXT: slli a3, a3, 24 |
| ; RV32VB-NEXT: zext.b a2, a2 |
| ; RV32VB-NEXT: zext.b a0, a0 |
| ; RV32VB-NEXT: zext.b a1, a1 |
| ; RV32VB-NEXT: slli a6, a6, 16 |
| ; RV32VB-NEXT: slli a5, a5, 8 |
| ; RV32VB-NEXT: slli a2, a2, 16 |
| ; RV32VB-NEXT: slli a1, a1, 8 |
| ; RV32VB-NEXT: or a6, a7, a6 |
| ; RV32VB-NEXT: or a4, a4, a5 |
| ; RV32VB-NEXT: or a2, a3, a2 |
| ; RV32VB-NEXT: or a0, a0, a1 |
| ; RV32VB-NEXT: or a1, a4, a6 |
| ; RV32VB-NEXT: or a0, a0, a2 |
| ; RV32VB-NEXT: vsetivli zero, 2, e32, mf2, ta, ma |
| ; RV32VB-NEXT: vmv.v.x v8, a0 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV32VB-NEXT: ret |
| ; |
| ; RV32VB-PACK-LABEL: buildvec_v8i8_pack: |
| ; RV32VB-PACK: # %bb.0: |
| ; RV32VB-PACK-NEXT: packh a6, a6, a7 |
| ; RV32VB-PACK-NEXT: packh a4, a4, a5 |
| ; RV32VB-PACK-NEXT: packh a2, a2, a3 |
| ; RV32VB-PACK-NEXT: packh a0, a0, a1 |
| ; RV32VB-PACK-NEXT: pack a1, a4, a6 |
| ; RV32VB-PACK-NEXT: pack a0, a0, a2 |
| ; RV32VB-PACK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma |
| ; RV32VB-PACK-NEXT: vmv.v.x v8, a0 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV32VB-PACK-NEXT: ret |
| ; |
| ; RV64V-ONLY-LABEL: buildvec_v8i8_pack: |
| ; RV64V-ONLY: # %bb.0: |
| ; RV64V-ONLY-NEXT: vsetivli zero, 8, e8, mf2, ta, mu |
| ; RV64V-ONLY-NEXT: vmv.v.x v8, a4 |
| ; RV64V-ONLY-NEXT: vmv.v.x v9, a0 |
| ; RV64V-ONLY-NEXT: vmv.v.i v0, 15 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a1 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a6 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a2 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a7 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v9, v9, a3 |
| ; RV64V-ONLY-NEXT: vslidedown.vi v8, v9, 4, v0.t |
| ; RV64V-ONLY-NEXT: ret |
| ; |
| ; RVA22U64-LABEL: buildvec_v8i8_pack: |
| ; RVA22U64: # %bb.0: |
| ; RVA22U64-NEXT: zext.b t0, a4 |
| ; RVA22U64-NEXT: zext.b a5, a5 |
| ; RVA22U64-NEXT: slli a7, a7, 56 |
| ; RVA22U64-NEXT: zext.b a4, a6 |
| ; RVA22U64-NEXT: zext.b a2, a2 |
| ; RVA22U64-NEXT: zext.b a3, a3 |
| ; RVA22U64-NEXT: zext.b a0, a0 |
| ; RVA22U64-NEXT: zext.b a1, a1 |
| ; RVA22U64-NEXT: slli t0, t0, 32 |
| ; RVA22U64-NEXT: slli a5, a5, 40 |
| ; RVA22U64-NEXT: slli a4, a4, 48 |
| ; RVA22U64-NEXT: slli a2, a2, 16 |
| ; RVA22U64-NEXT: slli a3, a3, 24 |
| ; RVA22U64-NEXT: slli a1, a1, 8 |
| ; RVA22U64-NEXT: or a5, a5, t0 |
| ; RVA22U64-NEXT: or a4, a7, a4 |
| ; RVA22U64-NEXT: or a2, a2, a3 |
| ; RVA22U64-NEXT: or a0, a0, a1 |
| ; RVA22U64-NEXT: or a4, a4, a5 |
| ; RVA22U64-NEXT: or a0, a0, a2 |
| ; RVA22U64-NEXT: or a0, a0, a4 |
| ; RVA22U64-NEXT: vsetivli zero, 1, e64, m1, ta, ma |
| ; RVA22U64-NEXT: vmv.s.x v8, a0 |
| ; RVA22U64-NEXT: ret |
| ; |
| ; RVA22U64-PACK-LABEL: buildvec_v8i8_pack: |
| ; RVA22U64-PACK: # %bb.0: |
| ; RVA22U64-PACK-NEXT: packh a6, a6, a7 |
| ; RVA22U64-PACK-NEXT: packh a4, a4, a5 |
| ; RVA22U64-PACK-NEXT: packh a2, a2, a3 |
| ; RVA22U64-PACK-NEXT: packh a0, a0, a1 |
| ; RVA22U64-PACK-NEXT: packw a1, a4, a6 |
| ; RVA22U64-PACK-NEXT: packw a0, a0, a2 |
| ; RVA22U64-PACK-NEXT: pack a0, a0, a1 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 1, e64, m1, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.s.x v8, a0 |
| ; RVA22U64-PACK-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_v8i8_pack: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: vsetivli zero, 8, e8, mf2, ta, mu |
| ; RV64ZVE32-NEXT: vmv.v.x v8, a4 |
| ; RV64ZVE32-NEXT: vmv.v.x v9, a0 |
| ; RV64ZVE32-NEXT: vmv.v.i v0, 15 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a1 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a6 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a2 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a7 |
| ; RV64ZVE32-NEXT: vslide1down.vx v9, v9, a3 |
| ; RV64ZVE32-NEXT: vslidedown.vi v8, v9, 4, v0.t |
| ; RV64ZVE32-NEXT: ret |
| %v1 = insertelement <8 x i8> poison, i8 %e1, i32 0 |
| %v2 = insertelement <8 x i8> %v1, i8 %e2, i32 1 |
| %v3 = insertelement <8 x i8> %v2, i8 %e3, i32 2 |
| %v4 = insertelement <8 x i8> %v3, i8 %e4, i32 3 |
| %v5 = insertelement <8 x i8> %v4, i8 %e5, i32 4 |
| %v6 = insertelement <8 x i8> %v5, i8 %e6, i32 5 |
| %v7 = insertelement <8 x i8> %v6, i8 %e7, i32 6 |
| %v8 = insertelement <8 x i8> %v7, i8 %e8, i32 7 |
| ret <8 x i8> %v8 |
| } |
| |
| define <6 x i8> @buildvec_v6i8_pack(i8 %e1, i8 %e2, i8 %e3, i8 %e4, i8 %e5, i8 %e6) { |
| ; RV32-ONLY-LABEL: buildvec_v6i8_pack: |
| ; RV32-ONLY: # %bb.0: |
| ; RV32-ONLY-NEXT: vsetivli zero, 8, e8, mf2, ta, ma |
| ; RV32-ONLY-NEXT: vmv.v.x v8, a0 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a4 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV32-ONLY-NEXT: vslidedown.vi v8, v8, 2 |
| ; RV32-ONLY-NEXT: ret |
| ; |
| ; RV32VB-LABEL: buildvec_v6i8_pack: |
| ; RV32VB: # %bb.0: |
| ; RV32VB-NEXT: slli a3, a3, 24 |
| ; RV32VB-NEXT: zext.b a2, a2 |
| ; RV32VB-NEXT: zext.b a0, a0 |
| ; RV32VB-NEXT: zext.b a1, a1 |
| ; RV32VB-NEXT: zext.b a4, a4 |
| ; RV32VB-NEXT: zext.b a5, a5 |
| ; RV32VB-NEXT: slli a2, a2, 16 |
| ; RV32VB-NEXT: slli a1, a1, 8 |
| ; RV32VB-NEXT: slli a5, a5, 8 |
| ; RV32VB-NEXT: or a2, a3, a2 |
| ; RV32VB-NEXT: or a0, a0, a1 |
| ; RV32VB-NEXT: or a0, a0, a2 |
| ; RV32VB-NEXT: or a4, a4, a5 |
| ; RV32VB-NEXT: vsetivli zero, 2, e32, mf2, ta, ma |
| ; RV32VB-NEXT: vmv.v.x v8, a0 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a4 |
| ; RV32VB-NEXT: ret |
| ; |
| ; RV32VB-PACK-LABEL: buildvec_v6i8_pack: |
| ; RV32VB-PACK: # %bb.0: |
| ; RV32VB-PACK-NEXT: packh a2, a2, a3 |
| ; RV32VB-PACK-NEXT: packh a0, a0, a1 |
| ; RV32VB-PACK-NEXT: packh a1, a4, a5 |
| ; RV32VB-PACK-NEXT: packh a3, a0, a0 |
| ; RV32VB-PACK-NEXT: pack a0, a0, a2 |
| ; RV32VB-PACK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma |
| ; RV32VB-PACK-NEXT: vmv.v.x v8, a0 |
| ; RV32VB-PACK-NEXT: pack a0, a1, a3 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32VB-PACK-NEXT: ret |
| ; |
| ; RV64V-ONLY-LABEL: buildvec_v6i8_pack: |
| ; RV64V-ONLY: # %bb.0: |
| ; RV64V-ONLY-NEXT: vsetivli zero, 8, e8, mf2, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.v.x v8, a0 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a4 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV64V-ONLY-NEXT: vslidedown.vi v8, v8, 2 |
| ; RV64V-ONLY-NEXT: ret |
| ; |
| ; RVA22U64-LABEL: buildvec_v6i8_pack: |
| ; RVA22U64: # %bb.0: |
| ; RVA22U64-NEXT: zext.b a2, a2 |
| ; RVA22U64-NEXT: zext.b a3, a3 |
| ; RVA22U64-NEXT: zext.b a0, a0 |
| ; RVA22U64-NEXT: zext.b a1, a1 |
| ; RVA22U64-NEXT: zext.b a4, a4 |
| ; RVA22U64-NEXT: zext.b a5, a5 |
| ; RVA22U64-NEXT: slli a2, a2, 16 |
| ; RVA22U64-NEXT: slli a3, a3, 24 |
| ; RVA22U64-NEXT: slli a1, a1, 8 |
| ; RVA22U64-NEXT: slli a4, a4, 32 |
| ; RVA22U64-NEXT: slli a5, a5, 40 |
| ; RVA22U64-NEXT: or a2, a2, a3 |
| ; RVA22U64-NEXT: or a0, a0, a1 |
| ; RVA22U64-NEXT: or a0, a0, a2 |
| ; RVA22U64-NEXT: or a4, a4, a5 |
| ; RVA22U64-NEXT: or a0, a0, a4 |
| ; RVA22U64-NEXT: vsetivli zero, 1, e64, m1, ta, ma |
| ; RVA22U64-NEXT: vmv.s.x v8, a0 |
| ; RVA22U64-NEXT: ret |
| ; |
| ; RVA22U64-PACK-LABEL: buildvec_v6i8_pack: |
| ; RVA22U64-PACK: # %bb.0: |
| ; RVA22U64-PACK-NEXT: packh a2, a2, a3 |
| ; RVA22U64-PACK-NEXT: packh a0, a0, a1 |
| ; RVA22U64-PACK-NEXT: packh a1, a4, a5 |
| ; RVA22U64-PACK-NEXT: packh a3, a0, a0 |
| ; RVA22U64-PACK-NEXT: packw a0, a0, a2 |
| ; RVA22U64-PACK-NEXT: packw a1, a1, a3 |
| ; RVA22U64-PACK-NEXT: pack a0, a0, a1 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 1, e64, m1, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.s.x v8, a0 |
| ; RVA22U64-PACK-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_v6i8_pack: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: vsetivli zero, 8, e8, mf2, ta, ma |
| ; RV64ZVE32-NEXT: vmv.v.x v8, a0 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a4 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a5 |
| ; RV64ZVE32-NEXT: vslidedown.vi v8, v8, 2 |
| ; RV64ZVE32-NEXT: ret |
| %v1 = insertelement <6 x i8> poison, i8 %e1, i32 0 |
| %v2 = insertelement <6 x i8> %v1, i8 %e2, i32 1 |
| %v3 = insertelement <6 x i8> %v2, i8 %e3, i32 2 |
| %v4 = insertelement <6 x i8> %v3, i8 %e4, i32 3 |
| %v5 = insertelement <6 x i8> %v4, i8 %e5, i32 4 |
| %v6 = insertelement <6 x i8> %v5, i8 %e6, i32 5 |
| ret <6 x i8> %v6 |
| } |
| |
| define <4 x i16> @buildvec_v4i16_pack(i16 %e1, i16 %e2, i16 %e3, i16 %e4) { |
| ; RV32-ONLY-LABEL: buildvec_v4i16_pack: |
| ; RV32-ONLY: # %bb.0: |
| ; RV32-ONLY-NEXT: vsetivli zero, 4, e16, mf2, ta, ma |
| ; RV32-ONLY-NEXT: vmv.v.x v8, a0 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV32-ONLY-NEXT: ret |
| ; |
| ; RV32VB-LABEL: buildvec_v4i16_pack: |
| ; RV32VB: # %bb.0: |
| ; RV32VB-NEXT: slli a3, a3, 16 |
| ; RV32VB-NEXT: zext.h a2, a2 |
| ; RV32VB-NEXT: slli a1, a1, 16 |
| ; RV32VB-NEXT: zext.h a0, a0 |
| ; RV32VB-NEXT: or a2, a2, a3 |
| ; RV32VB-NEXT: or a0, a0, a1 |
| ; RV32VB-NEXT: vsetivli zero, 2, e32, mf2, ta, ma |
| ; RV32VB-NEXT: vmv.v.x v8, a0 |
| ; RV32VB-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV32VB-NEXT: ret |
| ; |
| ; RV32VB-PACK-LABEL: buildvec_v4i16_pack: |
| ; RV32VB-PACK: # %bb.0: |
| ; RV32VB-PACK-NEXT: pack a2, a2, a3 |
| ; RV32VB-PACK-NEXT: pack a0, a0, a1 |
| ; RV32VB-PACK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma |
| ; RV32VB-PACK-NEXT: vmv.v.x v8, a0 |
| ; RV32VB-PACK-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV32VB-PACK-NEXT: ret |
| ; |
| ; RV64V-ONLY-LABEL: buildvec_v4i16_pack: |
| ; RV64V-ONLY: # %bb.0: |
| ; RV64V-ONLY-NEXT: vsetivli zero, 4, e16, mf2, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.v.x v8, a0 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV64V-ONLY-NEXT: ret |
| ; |
| ; RVA22U64-LABEL: buildvec_v4i16_pack: |
| ; RVA22U64: # %bb.0: |
| ; RVA22U64-NEXT: slli a3, a3, 48 |
| ; RVA22U64-NEXT: slli a2, a2, 48 |
| ; RVA22U64-NEXT: zext.h a0, a0 |
| ; RVA22U64-NEXT: slli a1, a1, 48 |
| ; RVA22U64-NEXT: srli a2, a2, 16 |
| ; RVA22U64-NEXT: srli a1, a1, 32 |
| ; RVA22U64-NEXT: or a2, a2, a3 |
| ; RVA22U64-NEXT: or a0, a0, a1 |
| ; RVA22U64-NEXT: or a0, a0, a2 |
| ; RVA22U64-NEXT: vsetivli zero, 1, e64, m1, ta, ma |
| ; RVA22U64-NEXT: vmv.s.x v8, a0 |
| ; RVA22U64-NEXT: ret |
| ; |
| ; RVA22U64-PACK-LABEL: buildvec_v4i16_pack: |
| ; RVA22U64-PACK: # %bb.0: |
| ; RVA22U64-PACK-NEXT: packw a2, a2, a3 |
| ; RVA22U64-PACK-NEXT: packw a0, a0, a1 |
| ; RVA22U64-PACK-NEXT: pack a0, a0, a2 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 1, e64, m1, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.s.x v8, a0 |
| ; RVA22U64-PACK-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_v4i16_pack: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: vsetivli zero, 4, e16, mf2, ta, ma |
| ; RV64ZVE32-NEXT: vmv.v.x v8, a0 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a3 |
| ; RV64ZVE32-NEXT: ret |
| %v1 = insertelement <4 x i16> poison, i16 %e1, i32 0 |
| %v2 = insertelement <4 x i16> %v1, i16 %e2, i32 1 |
| %v3 = insertelement <4 x i16> %v2, i16 %e3, i32 2 |
| %v4 = insertelement <4 x i16> %v3, i16 %e4, i32 3 |
| ret <4 x i16> %v4 |
| } |
| |
| define <2 x i32> @buildvec_v2i32_pack(i32 %e1, i32 %e2) { |
| ; RV32-LABEL: buildvec_v2i32_pack: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: vsetivli zero, 2, e32, mf2, ta, ma |
| ; RV32-NEXT: vmv.v.x v8, a0 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-ONLY-LABEL: buildvec_v2i32_pack: |
| ; RV64V-ONLY: # %bb.0: |
| ; RV64V-ONLY-NEXT: vsetivli zero, 2, e32, mf2, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.v.x v8, a0 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV64V-ONLY-NEXT: ret |
| ; |
| ; RVA22U64-LABEL: buildvec_v2i32_pack: |
| ; RVA22U64: # %bb.0: |
| ; RVA22U64-NEXT: slli a1, a1, 32 |
| ; RVA22U64-NEXT: add.uw a0, a0, a1 |
| ; RVA22U64-NEXT: vsetivli zero, 1, e64, m1, ta, ma |
| ; RVA22U64-NEXT: vmv.s.x v8, a0 |
| ; RVA22U64-NEXT: ret |
| ; |
| ; RVA22U64-PACK-LABEL: buildvec_v2i32_pack: |
| ; RVA22U64-PACK: # %bb.0: |
| ; RVA22U64-PACK-NEXT: pack a0, a0, a1 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 1, e64, m1, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.s.x v8, a0 |
| ; RVA22U64-PACK-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_v2i32_pack: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: vsetivli zero, 2, e32, m1, ta, ma |
| ; RV64ZVE32-NEXT: vmv.v.x v8, a0 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV64ZVE32-NEXT: ret |
| %v1 = insertelement <2 x i32> poison, i32 %e1, i32 0 |
| %v2 = insertelement <2 x i32> %v1, i32 %e2, i32 1 |
| ret <2 x i32> %v2 |
| } |
| |
| define <1 x i16> @buildvec_v1i16_pack(i16 %e1) { |
| ; CHECK-LABEL: buildvec_v1i16_pack: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; CHECK-NEXT: vmv.s.x v8, a0 |
| ; CHECK-NEXT: ret |
| %v1 = insertelement <1 x i16> poison, i16 %e1, i32 0 |
| ret <1 x i16> %v1 |
| } |
| |
| define <1 x i32> @buildvec_v1i32_pack(i32 %e1) { |
| ; CHECK-LABEL: buildvec_v1i32_pack: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 1, e32, m1, ta, ma |
| ; CHECK-NEXT: vmv.s.x v8, a0 |
| ; CHECK-NEXT: ret |
| %v1 = insertelement <1 x i32> poison, i32 %e1, i32 0 |
| ret <1 x i32> %v1 |
| } |
| |
| define <4 x i32> @buildvec_vslide1up(i32 %e1, i32 %e2) { |
| ; CHECK-LABEL: buildvec_vslide1up: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma |
| ; CHECK-NEXT: vmv.v.x v8, a0 |
| ; CHECK-NEXT: vsetvli zero, zero, e32, m1, tu, ma |
| ; CHECK-NEXT: vmv.s.x v8, a1 |
| ; CHECK-NEXT: ret |
| %v1 = insertelement <4 x i32> poison, i32 %e2, i32 0 |
| %v2 = insertelement <4 x i32> %v1, i32 %e1, i32 1 |
| %v3 = insertelement <4 x i32> %v2, i32 %e1, i32 2 |
| %v4 = insertelement <4 x i32> %v3, i32 %e1, i32 3 |
| ret <4 x i32> %v4 |
| } |
| |
| define <4 x i1> @buildvec_i1_splat(i1 %e1) { |
| ; CHECK-LABEL: buildvec_i1_splat: |
| ; CHECK: # %bb.0: |
| ; CHECK-NEXT: andi a0, a0, 1 |
| ; CHECK-NEXT: vsetivli zero, 4, e8, mf4, ta, ma |
| ; CHECK-NEXT: vmv.v.x v8, a0 |
| ; CHECK-NEXT: vmsne.vi v0, v8, 0 |
| ; CHECK-NEXT: ret |
| %v1 = insertelement <4 x i1> poison, i1 %e1, i32 0 |
| %v2 = insertelement <4 x i1> %v1, i1 %e1, i32 1 |
| %v3 = insertelement <4 x i1> %v2, i1 %e1, i32 2 |
| %v4 = insertelement <4 x i1> %v3, i1 %e1, i32 3 |
| ret <4 x i1> %v4 |
| } |
| |
| define <4 x i32> @buildvec_vredsum_slideup(<8 x i32> %arg0, <8 x i32> %arg1, <8 x i32> %arg2, <8 x i32> %arg3) nounwind { |
| ; RV32-LABEL: buildvec_vredsum_slideup: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma |
| ; RV32-NEXT: vmv.s.x v16, zero |
| ; RV32-NEXT: vredsum.vs v8, v8, v16 |
| ; RV32-NEXT: vredsum.vs v9, v10, v16 |
| ; RV32-NEXT: vredsum.vs v10, v12, v16 |
| ; RV32-NEXT: vredsum.vs v11, v14, v16 |
| ; RV32-NEXT: vsetivli zero, 4, e32, m1, tu, ma |
| ; RV32-NEXT: vslideup.vi v10, v11, 1 |
| ; RV32-NEXT: vslideup.vi v9, v10, 1 |
| ; RV32-NEXT: vslideup.vi v8, v9, 1 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-ONLY-LABEL: buildvec_vredsum_slideup: |
| ; RV64V-ONLY: # %bb.0: |
| ; RV64V-ONLY-NEXT: vsetivli zero, 8, e32, m2, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.s.x v16, zero |
| ; RV64V-ONLY-NEXT: vredsum.vs v8, v8, v16 |
| ; RV64V-ONLY-NEXT: vredsum.vs v9, v10, v16 |
| ; RV64V-ONLY-NEXT: vredsum.vs v10, v12, v16 |
| ; RV64V-ONLY-NEXT: vredsum.vs v11, v14, v16 |
| ; RV64V-ONLY-NEXT: vsetivli zero, 4, e32, m1, tu, ma |
| ; RV64V-ONLY-NEXT: vslideup.vi v10, v11, 1 |
| ; RV64V-ONLY-NEXT: vslideup.vi v9, v10, 1 |
| ; RV64V-ONLY-NEXT: vslideup.vi v8, v9, 1 |
| ; RV64V-ONLY-NEXT: ret |
| ; |
| ; RVA22U64-LABEL: buildvec_vredsum_slideup: |
| ; RVA22U64: # %bb.0: |
| ; RVA22U64-NEXT: vsetivli zero, 8, e32, m2, ta, ma |
| ; RVA22U64-NEXT: vmv.s.x v16, zero |
| ; RVA22U64-NEXT: vredsum.vs v8, v8, v16 |
| ; RVA22U64-NEXT: vredsum.vs v9, v10, v16 |
| ; RVA22U64-NEXT: vredsum.vs v10, v12, v16 |
| ; RVA22U64-NEXT: vredsum.vs v11, v14, v16 |
| ; RVA22U64-NEXT: vmv.x.s a0, v8 |
| ; RVA22U64-NEXT: vmv.x.s a1, v9 |
| ; RVA22U64-NEXT: vmv.x.s a2, v10 |
| ; RVA22U64-NEXT: slli a1, a1, 32 |
| ; RVA22U64-NEXT: add.uw a0, a0, a1 |
| ; RVA22U64-NEXT: vmv.x.s a1, v11 |
| ; RVA22U64-NEXT: slli a1, a1, 32 |
| ; RVA22U64-NEXT: add.uw a1, a2, a1 |
| ; RVA22U64-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-NEXT: vmv.v.x v8, a0 |
| ; RVA22U64-NEXT: vslide1down.vx v8, v8, a1 |
| ; RVA22U64-NEXT: ret |
| ; |
| ; RVA22U64-PACK-LABEL: buildvec_vredsum_slideup: |
| ; RVA22U64-PACK: # %bb.0: |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 8, e32, m2, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.s.x v16, zero |
| ; RVA22U64-PACK-NEXT: vredsum.vs v8, v8, v16 |
| ; RVA22U64-PACK-NEXT: vredsum.vs v9, v10, v16 |
| ; RVA22U64-PACK-NEXT: vredsum.vs v10, v12, v16 |
| ; RVA22U64-PACK-NEXT: vredsum.vs v11, v14, v16 |
| ; RVA22U64-PACK-NEXT: vmv.x.s a0, v8 |
| ; RVA22U64-PACK-NEXT: vmv.x.s a1, v9 |
| ; RVA22U64-PACK-NEXT: vmv.x.s a2, v10 |
| ; RVA22U64-PACK-NEXT: pack a0, a0, a1 |
| ; RVA22U64-PACK-NEXT: vmv.x.s a1, v11 |
| ; RVA22U64-PACK-NEXT: pack a1, a2, a1 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.v.x v8, a0 |
| ; RVA22U64-PACK-NEXT: vslide1down.vx v8, v8, a1 |
| ; RVA22U64-PACK-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_vredsum_slideup: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: vsetivli zero, 8, e32, m2, ta, ma |
| ; RV64ZVE32-NEXT: vmv.s.x v16, zero |
| ; RV64ZVE32-NEXT: vredsum.vs v8, v8, v16 |
| ; RV64ZVE32-NEXT: vredsum.vs v9, v10, v16 |
| ; RV64ZVE32-NEXT: vredsum.vs v10, v12, v16 |
| ; RV64ZVE32-NEXT: vredsum.vs v11, v14, v16 |
| ; RV64ZVE32-NEXT: vsetivli zero, 4, e32, m1, tu, ma |
| ; RV64ZVE32-NEXT: vslideup.vi v10, v11, 1 |
| ; RV64ZVE32-NEXT: vslideup.vi v9, v10, 1 |
| ; RV64ZVE32-NEXT: vslideup.vi v8, v9, 1 |
| ; RV64ZVE32-NEXT: ret |
| %247 = tail call i32 @llvm.vector.reduce.add.v8i32(<8 x i32> %arg0) |
| %248 = insertelement <4 x i32> poison, i32 %247, i64 0 |
| %250 = tail call i32 @llvm.vector.reduce.add.v8i32(<8 x i32> %arg1) |
| %251 = insertelement <4 x i32> %248, i32 %250, i64 1 |
| %252 = tail call i32 @llvm.vector.reduce.add.v8i32(<8 x i32> %arg2) |
| %253 = insertelement <4 x i32> %251, i32 %252, i64 2 |
| %254 = tail call i32 @llvm.vector.reduce.add.v8i32(<8 x i32> %arg3) |
| %255 = insertelement <4 x i32> %253, i32 %254, i64 3 |
| ret <4 x i32> %255 |
| } |
| |
| define <4 x i32> @buildvec_vredmax_slideup(<8 x i32> %arg0, <8 x i32> %arg1, <8 x i32> %arg2, <8 x i32> %arg3) nounwind { |
| ; RV32-LABEL: buildvec_vredmax_slideup: |
| ; RV32: # %bb.0: |
| ; RV32-NEXT: vsetivli zero, 8, e32, m2, ta, ma |
| ; RV32-NEXT: vredmaxu.vs v8, v8, v8 |
| ; RV32-NEXT: vredmaxu.vs v9, v10, v10 |
| ; RV32-NEXT: vredmaxu.vs v10, v12, v12 |
| ; RV32-NEXT: vredmaxu.vs v11, v14, v14 |
| ; RV32-NEXT: vsetivli zero, 4, e32, m1, tu, ma |
| ; RV32-NEXT: vslideup.vi v10, v11, 1 |
| ; RV32-NEXT: vslideup.vi v9, v10, 1 |
| ; RV32-NEXT: vslideup.vi v8, v9, 1 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-ONLY-LABEL: buildvec_vredmax_slideup: |
| ; RV64V-ONLY: # %bb.0: |
| ; RV64V-ONLY-NEXT: vsetivli zero, 8, e32, m2, ta, ma |
| ; RV64V-ONLY-NEXT: vredmaxu.vs v8, v8, v8 |
| ; RV64V-ONLY-NEXT: vredmaxu.vs v9, v10, v10 |
| ; RV64V-ONLY-NEXT: vredmaxu.vs v10, v12, v12 |
| ; RV64V-ONLY-NEXT: vredmaxu.vs v11, v14, v14 |
| ; RV64V-ONLY-NEXT: vsetivli zero, 4, e32, m1, tu, ma |
| ; RV64V-ONLY-NEXT: vslideup.vi v10, v11, 1 |
| ; RV64V-ONLY-NEXT: vslideup.vi v9, v10, 1 |
| ; RV64V-ONLY-NEXT: vslideup.vi v8, v9, 1 |
| ; RV64V-ONLY-NEXT: ret |
| ; |
| ; RVA22U64-LABEL: buildvec_vredmax_slideup: |
| ; RVA22U64: # %bb.0: |
| ; RVA22U64-NEXT: vsetivli zero, 8, e32, m2, ta, ma |
| ; RVA22U64-NEXT: vredmaxu.vs v8, v8, v8 |
| ; RVA22U64-NEXT: vredmaxu.vs v9, v10, v10 |
| ; RVA22U64-NEXT: vredmaxu.vs v10, v12, v12 |
| ; RVA22U64-NEXT: vredmaxu.vs v11, v14, v14 |
| ; RVA22U64-NEXT: vmv.x.s a0, v8 |
| ; RVA22U64-NEXT: vmv.x.s a1, v9 |
| ; RVA22U64-NEXT: vmv.x.s a2, v10 |
| ; RVA22U64-NEXT: slli a1, a1, 32 |
| ; RVA22U64-NEXT: add.uw a0, a0, a1 |
| ; RVA22U64-NEXT: vmv.x.s a1, v11 |
| ; RVA22U64-NEXT: slli a1, a1, 32 |
| ; RVA22U64-NEXT: add.uw a1, a2, a1 |
| ; RVA22U64-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-NEXT: vmv.v.x v8, a0 |
| ; RVA22U64-NEXT: vslide1down.vx v8, v8, a1 |
| ; RVA22U64-NEXT: ret |
| ; |
| ; RVA22U64-PACK-LABEL: buildvec_vredmax_slideup: |
| ; RVA22U64-PACK: # %bb.0: |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 8, e32, m2, ta, ma |
| ; RVA22U64-PACK-NEXT: vredmaxu.vs v8, v8, v8 |
| ; RVA22U64-PACK-NEXT: vredmaxu.vs v9, v10, v10 |
| ; RVA22U64-PACK-NEXT: vredmaxu.vs v10, v12, v12 |
| ; RVA22U64-PACK-NEXT: vredmaxu.vs v11, v14, v14 |
| ; RVA22U64-PACK-NEXT: vmv.x.s a0, v8 |
| ; RVA22U64-PACK-NEXT: vmv.x.s a1, v9 |
| ; RVA22U64-PACK-NEXT: vmv.x.s a2, v10 |
| ; RVA22U64-PACK-NEXT: pack a0, a0, a1 |
| ; RVA22U64-PACK-NEXT: vmv.x.s a1, v11 |
| ; RVA22U64-PACK-NEXT: pack a1, a2, a1 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 2, e64, m1, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.v.x v8, a0 |
| ; RVA22U64-PACK-NEXT: vslide1down.vx v8, v8, a1 |
| ; RVA22U64-PACK-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: buildvec_vredmax_slideup: |
| ; RV64ZVE32: # %bb.0: |
| ; RV64ZVE32-NEXT: vsetivli zero, 8, e32, m2, ta, ma |
| ; RV64ZVE32-NEXT: vredmaxu.vs v8, v8, v8 |
| ; RV64ZVE32-NEXT: vredmaxu.vs v9, v10, v10 |
| ; RV64ZVE32-NEXT: vredmaxu.vs v10, v12, v12 |
| ; RV64ZVE32-NEXT: vredmaxu.vs v11, v14, v14 |
| ; RV64ZVE32-NEXT: vsetivli zero, 4, e32, m1, tu, ma |
| ; RV64ZVE32-NEXT: vslideup.vi v10, v11, 1 |
| ; RV64ZVE32-NEXT: vslideup.vi v9, v10, 1 |
| ; RV64ZVE32-NEXT: vslideup.vi v8, v9, 1 |
| ; RV64ZVE32-NEXT: ret |
| %247 = tail call i32 @llvm.vector.reduce.umax.v8i32(<8 x i32> %arg0) |
| %248 = insertelement <4 x i32> poison, i32 %247, i64 0 |
| %250 = tail call i32 @llvm.vector.reduce.umax.v8i32(<8 x i32> %arg1) |
| %251 = insertelement <4 x i32> %248, i32 %250, i64 1 |
| %252 = tail call i32 @llvm.vector.reduce.umax.v8i32(<8 x i32> %arg2) |
| %253 = insertelement <4 x i32> %251, i32 %252, i64 2 |
| %254 = tail call i32 @llvm.vector.reduce.umax.v8i32(<8 x i32> %arg3) |
| %255 = insertelement <4 x i32> %253, i32 %254, i64 3 |
| ret <4 x i32> %255 |
| } |
| |
| define <16 x i16> @PR159294(<2 x i32> %a, <2 x i32> %b, <2 x i32> %c) { |
| ; RV32-ONLY-LABEL: PR159294: |
| ; RV32-ONLY: # %bb.0: # %entry |
| ; RV32-ONLY-NEXT: vsetivli zero, 1, e32, m1, ta, ma |
| ; RV32-ONLY-NEXT: vmv.x.s a0, v8 |
| ; RV32-ONLY-NEXT: vmv.x.s a1, v9 |
| ; RV32-ONLY-NEXT: vmv.x.s a2, v10 |
| ; RV32-ONLY-NEXT: vsetivli zero, 16, e16, m2, ta, ma |
| ; RV32-ONLY-NEXT: vmv.v.x v8, a2 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32-ONLY-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV32-ONLY-NEXT: vslidedown.vi v8, v8, 13 |
| ; RV32-ONLY-NEXT: ret |
| ; |
| ; RV32VB-LABEL: PR159294: |
| ; RV32VB: # %bb.0: # %entry |
| ; RV32VB-NEXT: vsetivli zero, 8, e32, m2, ta, ma |
| ; RV32VB-NEXT: vmv.x.s a0, v8 |
| ; RV32VB-NEXT: vmv.x.s a1, v10 |
| ; RV32VB-NEXT: slli a0, a0, 16 |
| ; RV32VB-NEXT: zext.h a1, a1 |
| ; RV32VB-NEXT: or a0, a1, a0 |
| ; RV32VB-NEXT: vmv.x.s a1, v9 |
| ; RV32VB-NEXT: vmv.v.i v8, 0 |
| ; RV32VB-NEXT: zext.h a1, a1 |
| ; RV32VB-NEXT: vsetvli zero, zero, e32, m2, tu, ma |
| ; RV32VB-NEXT: vmv.s.x v8, a0 |
| ; RV32VB-NEXT: vmv.s.x v10, a1 |
| ; RV32VB-NEXT: vsetivli zero, 2, e32, m1, tu, ma |
| ; RV32VB-NEXT: vslideup.vi v8, v10, 1 |
| ; RV32VB-NEXT: ret |
| ; |
| ; RV32VB-PACK-LABEL: PR159294: |
| ; RV32VB-PACK: # %bb.0: # %entry |
| ; RV32VB-PACK-NEXT: vsetivli zero, 8, e32, m2, ta, ma |
| ; RV32VB-PACK-NEXT: vmv.x.s a0, v8 |
| ; RV32VB-PACK-NEXT: vmv.x.s a1, v10 |
| ; RV32VB-PACK-NEXT: vmv.x.s a2, v9 |
| ; RV32VB-PACK-NEXT: pack a0, a1, a0 |
| ; RV32VB-PACK-NEXT: pack a1, a0, a0 |
| ; RV32VB-PACK-NEXT: vmv.v.x v8, a1 |
| ; RV32VB-PACK-NEXT: pack a1, a2, a0 |
| ; RV32VB-PACK-NEXT: vsetvli zero, zero, e32, m2, tu, ma |
| ; RV32VB-PACK-NEXT: vmv.s.x v8, a0 |
| ; RV32VB-PACK-NEXT: vmv.s.x v10, a1 |
| ; RV32VB-PACK-NEXT: vsetivli zero, 2, e32, m1, tu, ma |
| ; RV32VB-PACK-NEXT: vslideup.vi v8, v10, 1 |
| ; RV32VB-PACK-NEXT: ret |
| ; |
| ; RV64V-ONLY-LABEL: PR159294: |
| ; RV64V-ONLY: # %bb.0: # %entry |
| ; RV64V-ONLY-NEXT: vsetivli zero, 1, e32, m1, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.x.s a0, v8 |
| ; RV64V-ONLY-NEXT: vmv.x.s a1, v9 |
| ; RV64V-ONLY-NEXT: vmv.x.s a2, v10 |
| ; RV64V-ONLY-NEXT: vsetivli zero, 16, e16, m2, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.v.x v8, a2 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV64V-ONLY-NEXT: vslidedown.vi v8, v8, 13 |
| ; RV64V-ONLY-NEXT: ret |
| ; |
| ; RVA22U64-LABEL: PR159294: |
| ; RVA22U64: # %bb.0: # %entry |
| ; RVA22U64-NEXT: vsetivli zero, 8, e32, m2, ta, ma |
| ; RVA22U64-NEXT: vmv.x.s a0, v8 |
| ; RVA22U64-NEXT: vmv.x.s a1, v10 |
| ; RVA22U64-NEXT: slli a0, a0, 16 |
| ; RVA22U64-NEXT: zext.h a1, a1 |
| ; RVA22U64-NEXT: or a0, a0, a1 |
| ; RVA22U64-NEXT: vmv.x.s a1, v9 |
| ; RVA22U64-NEXT: vmv.v.i v8, 0 |
| ; RVA22U64-NEXT: zext.h a1, a1 |
| ; RVA22U64-NEXT: vsetvli zero, zero, e32, m2, tu, ma |
| ; RVA22U64-NEXT: vmv.s.x v8, a0 |
| ; RVA22U64-NEXT: vmv.s.x v10, a1 |
| ; RVA22U64-NEXT: vsetivli zero, 2, e32, m1, tu, ma |
| ; RVA22U64-NEXT: vslideup.vi v8, v10, 1 |
| ; RVA22U64-NEXT: ret |
| ; |
| ; RVA22U64-PACK-LABEL: PR159294: |
| ; RVA22U64-PACK: # %bb.0: # %entry |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 8, e32, m2, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.x.s a0, v8 |
| ; RVA22U64-PACK-NEXT: vmv.x.s a1, v10 |
| ; RVA22U64-PACK-NEXT: vmv.x.s a2, v9 |
| ; RVA22U64-PACK-NEXT: packw a0, a1, a0 |
| ; RVA22U64-PACK-NEXT: packw a1, a0, a0 |
| ; RVA22U64-PACK-NEXT: vmv.v.x v8, a1 |
| ; RVA22U64-PACK-NEXT: packw a1, a2, a0 |
| ; RVA22U64-PACK-NEXT: vsetvli zero, zero, e32, m2, tu, ma |
| ; RVA22U64-PACK-NEXT: vmv.s.x v8, a0 |
| ; RVA22U64-PACK-NEXT: vmv.s.x v10, a1 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 2, e32, m1, tu, ma |
| ; RVA22U64-PACK-NEXT: vslideup.vi v8, v10, 1 |
| ; RVA22U64-PACK-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: PR159294: |
| ; RV64ZVE32: # %bb.0: # %entry |
| ; RV64ZVE32-NEXT: vsetivli zero, 1, e32, m1, ta, ma |
| ; RV64ZVE32-NEXT: vmv.x.s a0, v8 |
| ; RV64ZVE32-NEXT: vmv.x.s a1, v9 |
| ; RV64ZVE32-NEXT: vmv.x.s a2, v10 |
| ; RV64ZVE32-NEXT: vsetivli zero, 16, e16, m2, ta, ma |
| ; RV64ZVE32-NEXT: vmv.v.x v8, a2 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV64ZVE32-NEXT: vslidedown.vi v8, v8, 13 |
| ; RV64ZVE32-NEXT: ret |
| entry: |
| %vecext3 = extractelement <2 x i32> %a, i32 0 |
| %conv4 = trunc i32 %vecext3 to i16 |
| %vecinit5 = insertelement <16 x i16> <i16 0, i16 poison, i16 poison, i16 poison, i16 poison, i16 poison, i16 poison, i16 poison, i16 poison, i16 poison, i16 poison, i16 poison, i16 poison, i16 poison, i16 poison, i16 poison>, i16 %conv4, i32 1 |
| %vecext7 = extractelement <2 x i32> %b, i32 0 |
| %conv8 = trunc i32 %vecext7 to i16 |
| %vecinit9 = insertelement <16 x i16> %vecinit5, i16 %conv8, i32 2 |
| %vecext59 = extractelement <2 x i32> %c, i32 0 |
| %conv60 = trunc i32 %vecext59 to i16 |
| %vecinit61 = insertelement <16 x i16> %vecinit9, i16 %conv60, i32 0 |
| ret <16 x i16> %vecinit61 |
| } |
| |
| define <16 x i32> @PR159294_zext(<2 x i16> %a, <2 x i16> %b, <2 x i16> %c) { |
| ; RV32-LABEL: PR159294_zext: |
| ; RV32: # %bb.0: # %entry |
| ; RV32-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV32-NEXT: vmv.x.s a0, v8 |
| ; RV32-NEXT: vmv.x.s a1, v9 |
| ; RV32-NEXT: vmv.x.s a2, v10 |
| ; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, ma |
| ; RV32-NEXT: vmv.v.x v8, a2 |
| ; RV32-NEXT: lui a2, 16 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV32-NEXT: vslidedown.vi v8, v8, 13 |
| ; RV32-NEXT: addi a2, a2, -1 |
| ; RV32-NEXT: vand.vx v8, v8, a2 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-ONLY-LABEL: PR159294_zext: |
| ; RV64V-ONLY: # %bb.0: # %entry |
| ; RV64V-ONLY-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.x.s a0, v8 |
| ; RV64V-ONLY-NEXT: lui a1, 16 |
| ; RV64V-ONLY-NEXT: vmv.x.s a2, v9 |
| ; RV64V-ONLY-NEXT: vmv.x.s a3, v10 |
| ; RV64V-ONLY-NEXT: addi a1, a1, -1 |
| ; RV64V-ONLY-NEXT: and a0, a0, a1 |
| ; RV64V-ONLY-NEXT: and a2, a2, a1 |
| ; RV64V-ONLY-NEXT: and a1, a3, a1 |
| ; RV64V-ONLY-NEXT: vsetivli zero, 16, e32, m4, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.v.x v8, a1 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV64V-ONLY-NEXT: vslidedown.vi v8, v8, 13 |
| ; RV64V-ONLY-NEXT: ret |
| ; |
| ; RVA22U64-LABEL: PR159294_zext: |
| ; RVA22U64: # %bb.0: # %entry |
| ; RVA22U64-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RVA22U64-NEXT: vmv.x.s a0, v8 |
| ; RVA22U64-NEXT: vmv.x.s a1, v10 |
| ; RVA22U64-NEXT: slli a0, a0, 48 |
| ; RVA22U64-NEXT: zext.h a1, a1 |
| ; RVA22U64-NEXT: srli a0, a0, 16 |
| ; RVA22U64-NEXT: or a0, a0, a1 |
| ; RVA22U64-NEXT: vmv.x.s a1, v9 |
| ; RVA22U64-NEXT: vsetivli zero, 8, e64, m4, ta, ma |
| ; RVA22U64-NEXT: vmv.v.i v8, 0 |
| ; RVA22U64-NEXT: zext.h a1, a1 |
| ; RVA22U64-NEXT: vsetvli zero, zero, e64, m4, tu, ma |
| ; RVA22U64-NEXT: vmv.s.x v8, a0 |
| ; RVA22U64-NEXT: vmv.s.x v12, a1 |
| ; RVA22U64-NEXT: vsetivli zero, 2, e64, m1, tu, ma |
| ; RVA22U64-NEXT: vslideup.vi v8, v12, 1 |
| ; RVA22U64-NEXT: ret |
| ; |
| ; RVA22U64-PACK-LABEL: PR159294_zext: |
| ; RVA22U64-PACK: # %bb.0: # %entry |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 1, e16, m2, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv1r.v v12, v9 |
| ; RVA22U64-PACK-NEXT: vmv.x.s a0, v8 |
| ; RVA22U64-PACK-NEXT: vmv.x.s a1, v10 |
| ; RVA22U64-PACK-NEXT: pack a2, a0, a0 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 8, e64, m4, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.v.x v8, a2 |
| ; RVA22U64-PACK-NEXT: vsetvli zero, zero, e16, m1, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.x.s a2, v12 |
| ; RVA22U64-PACK-NEXT: zext.h a0, a0 |
| ; RVA22U64-PACK-NEXT: zext.h a1, a1 |
| ; RVA22U64-PACK-NEXT: zext.h a2, a2 |
| ; RVA22U64-PACK-NEXT: pack a0, a1, a0 |
| ; RVA22U64-PACK-NEXT: pack a1, a2, a0 |
| ; RVA22U64-PACK-NEXT: vsetvli zero, zero, e64, m4, tu, ma |
| ; RVA22U64-PACK-NEXT: vmv.s.x v8, a0 |
| ; RVA22U64-PACK-NEXT: vmv.s.x v12, a1 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 2, e64, m1, tu, ma |
| ; RVA22U64-PACK-NEXT: vslideup.vi v8, v12, 1 |
| ; RVA22U64-PACK-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: PR159294_zext: |
| ; RV64ZVE32: # %bb.0: # %entry |
| ; RV64ZVE32-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV64ZVE32-NEXT: vmv.x.s a0, v8 |
| ; RV64ZVE32-NEXT: lui a1, 16 |
| ; RV64ZVE32-NEXT: vmv.x.s a2, v9 |
| ; RV64ZVE32-NEXT: vmv.x.s a3, v10 |
| ; RV64ZVE32-NEXT: addi a1, a1, -1 |
| ; RV64ZVE32-NEXT: and a0, a0, a1 |
| ; RV64ZVE32-NEXT: and a2, a2, a1 |
| ; RV64ZVE32-NEXT: and a1, a3, a1 |
| ; RV64ZVE32-NEXT: vsetivli zero, 16, e32, m4, ta, ma |
| ; RV64ZVE32-NEXT: vmv.v.x v8, a1 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a2 |
| ; RV64ZVE32-NEXT: vslidedown.vi v8, v8, 13 |
| ; RV64ZVE32-NEXT: ret |
| entry: |
| %vecext3 = extractelement <2 x i16> %a, i32 0 |
| %conv4 = zext i16 %vecext3 to i32 |
| %vecinit5 = insertelement <16 x i32> <i32 0, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>, i32 %conv4, i32 1 |
| %vecext7 = extractelement <2 x i16> %b, i32 0 |
| %conv8 = zext i16 %vecext7 to i32 |
| %vecinit9 = insertelement <16 x i32> %vecinit5, i32 %conv8, i32 2 |
| %vecext59 = extractelement <2 x i16> %c, i32 0 |
| %conv60 = zext i16 %vecext59 to i32 |
| %vecinit61 = insertelement <16 x i32> %vecinit9, i32 %conv60, i32 0 |
| ret <16 x i32> %vecinit61 |
| } |
| |
| define <16 x i32> @PR159294_sext(<2 x i16> %a, <2 x i16> %b, <2 x i16> %c) { |
| ; RV32-LABEL: PR159294_sext: |
| ; RV32: # %bb.0: # %entry |
| ; RV32-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV32-NEXT: vmv.x.s a0, v8 |
| ; RV32-NEXT: vmv.x.s a1, v9 |
| ; RV32-NEXT: vmv.x.s a2, v10 |
| ; RV32-NEXT: vsetivli zero, 16, e32, m4, ta, ma |
| ; RV32-NEXT: vmv.v.x v8, a2 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV32-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV32-NEXT: vslidedown.vi v8, v8, 13 |
| ; RV32-NEXT: ret |
| ; |
| ; RV64V-ONLY-LABEL: PR159294_sext: |
| ; RV64V-ONLY: # %bb.0: # %entry |
| ; RV64V-ONLY-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.x.s a0, v8 |
| ; RV64V-ONLY-NEXT: vmv.x.s a1, v9 |
| ; RV64V-ONLY-NEXT: vmv.x.s a2, v10 |
| ; RV64V-ONLY-NEXT: vsetivli zero, 16, e32, m4, ta, ma |
| ; RV64V-ONLY-NEXT: vmv.v.x v8, a2 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64V-ONLY-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV64V-ONLY-NEXT: vslidedown.vi v8, v8, 13 |
| ; RV64V-ONLY-NEXT: ret |
| ; |
| ; RVA22U64-LABEL: PR159294_sext: |
| ; RVA22U64: # %bb.0: # %entry |
| ; RVA22U64-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RVA22U64-NEXT: vmv.x.s a0, v8 |
| ; RVA22U64-NEXT: vmv.x.s a1, v10 |
| ; RVA22U64-NEXT: slli a0, a0, 32 |
| ; RVA22U64-NEXT: add.uw a0, a1, a0 |
| ; RVA22U64-NEXT: vmv.x.s a1, v9 |
| ; RVA22U64-NEXT: vsetivli zero, 8, e64, m4, ta, ma |
| ; RVA22U64-NEXT: vmv.v.i v8, 0 |
| ; RVA22U64-NEXT: zext.w a1, a1 |
| ; RVA22U64-NEXT: vsetvli zero, zero, e64, m4, tu, ma |
| ; RVA22U64-NEXT: vmv.s.x v8, a0 |
| ; RVA22U64-NEXT: vmv.s.x v12, a1 |
| ; RVA22U64-NEXT: vsetivli zero, 2, e64, m1, tu, ma |
| ; RVA22U64-NEXT: vslideup.vi v8, v12, 1 |
| ; RVA22U64-NEXT: ret |
| ; |
| ; RVA22U64-PACK-LABEL: PR159294_sext: |
| ; RVA22U64-PACK: # %bb.0: # %entry |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.x.s a0, v8 |
| ; RVA22U64-PACK-NEXT: vmv.x.s a1, v10 |
| ; RVA22U64-PACK-NEXT: vmv.x.s a2, v9 |
| ; RVA22U64-PACK-NEXT: pack a0, a1, a0 |
| ; RVA22U64-PACK-NEXT: pack a1, a0, a0 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 8, e64, m4, ta, ma |
| ; RVA22U64-PACK-NEXT: vmv.v.x v8, a1 |
| ; RVA22U64-PACK-NEXT: pack a1, a2, a0 |
| ; RVA22U64-PACK-NEXT: vsetvli zero, zero, e64, m4, tu, ma |
| ; RVA22U64-PACK-NEXT: vmv.s.x v8, a0 |
| ; RVA22U64-PACK-NEXT: vmv.s.x v12, a1 |
| ; RVA22U64-PACK-NEXT: vsetivli zero, 2, e64, m1, tu, ma |
| ; RVA22U64-PACK-NEXT: vslideup.vi v8, v12, 1 |
| ; RVA22U64-PACK-NEXT: ret |
| ; |
| ; RV64ZVE32-LABEL: PR159294_sext: |
| ; RV64ZVE32: # %bb.0: # %entry |
| ; RV64ZVE32-NEXT: vsetivli zero, 1, e16, m1, ta, ma |
| ; RV64ZVE32-NEXT: vmv.x.s a0, v8 |
| ; RV64ZVE32-NEXT: vmv.x.s a1, v9 |
| ; RV64ZVE32-NEXT: vmv.x.s a2, v10 |
| ; RV64ZVE32-NEXT: vsetivli zero, 16, e32, m4, ta, ma |
| ; RV64ZVE32-NEXT: vmv.v.x v8, a2 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a0 |
| ; RV64ZVE32-NEXT: vslide1down.vx v8, v8, a1 |
| ; RV64ZVE32-NEXT: vslidedown.vi v8, v8, 13 |
| ; RV64ZVE32-NEXT: ret |
| entry: |
| %vecext3 = extractelement <2 x i16> %a, i32 0 |
| %conv4 = sext i16 %vecext3 to i32 |
| %vecinit5 = insertelement <16 x i32> <i32 0, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>, i32 %conv4, i32 1 |
| %vecext7 = extractelement <2 x i16> %b, i32 0 |
| %conv8 = sext i16 %vecext7 to i32 |
| %vecinit9 = insertelement <16 x i32> %vecinit5, i32 %conv8, i32 2 |
| %vecext59 = extractelement <2 x i16> %c, i32 0 |
| %conv60 = sext i16 %vecext59 to i32 |
| %vecinit61 = insertelement <16 x i32> %vecinit9, i32 %conv60, i32 0 |
| ret <16 x i32> %vecinit61 |
| } |
| ;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line: |
| ; RV64: {{.*}} |
