| ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py |
| ; RUN: llc < %s -mtriple=aarch64 | FileCheck %s --check-prefixes=CHECK,CHECK-CVT |
| ; RUN: llc < %s -mtriple=aarch64 -mattr=+fullfp16 | FileCheck %s --check-prefixes=CHECK,CHECK-FP16 |
| |
| ; i32 saturate |
| |
| define <2 x i32> @stest_f64i32(<2 x double> %x) { |
| ; CHECK-LABEL: stest_f64i32: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: mov d1, v0.d[1] |
| ; CHECK-NEXT: fcvtzs w8, d0 |
| ; CHECK-NEXT: fcvtzs w9, d1 |
| ; CHECK-NEXT: fmov s0, w8 |
| ; CHECK-NEXT: mov v0.s[1], w9 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 killed $q0 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x double> %x to <2 x i64> |
| %0 = icmp slt <2 x i64> %conv, <i64 2147483647, i64 2147483647> |
| %spec.store.select = select <2 x i1> %0, <2 x i64> %conv, <2 x i64> <i64 2147483647, i64 2147483647> |
| %1 = icmp sgt <2 x i64> %spec.store.select, <i64 -2147483648, i64 -2147483648> |
| %spec.store.select7 = select <2 x i1> %1, <2 x i64> %spec.store.select, <2 x i64> <i64 -2147483648, i64 -2147483648> |
| %conv6 = trunc <2 x i64> %spec.store.select7 to <2 x i32> |
| ret <2 x i32> %conv6 |
| } |
| |
| define <2 x i32> @utest_f64i32(<2 x double> %x) { |
| ; CHECK-LABEL: utest_f64i32: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: mov d1, v0.d[1] |
| ; CHECK-NEXT: fcvtzu w8, d0 |
| ; CHECK-NEXT: fcvtzu w9, d1 |
| ; CHECK-NEXT: fmov s0, w8 |
| ; CHECK-NEXT: mov v0.s[1], w9 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 killed $q0 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <2 x double> %x to <2 x i64> |
| %0 = icmp ult <2 x i64> %conv, <i64 4294967295, i64 4294967295> |
| %spec.store.select = select <2 x i1> %0, <2 x i64> %conv, <2 x i64> <i64 4294967295, i64 4294967295> |
| %conv6 = trunc <2 x i64> %spec.store.select to <2 x i32> |
| ret <2 x i32> %conv6 |
| } |
| |
| define <2 x i32> @ustest_f64i32(<2 x double> %x) { |
| ; CHECK-LABEL: ustest_f64i32: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: mov d1, v0.d[1] |
| ; CHECK-NEXT: fcvtzu w8, d0 |
| ; CHECK-NEXT: fcvtzu w9, d1 |
| ; CHECK-NEXT: fmov s0, w8 |
| ; CHECK-NEXT: mov v0.s[1], w9 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 killed $q0 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x double> %x to <2 x i64> |
| %0 = icmp slt <2 x i64> %conv, <i64 4294967295, i64 4294967295> |
| %spec.store.select = select <2 x i1> %0, <2 x i64> %conv, <2 x i64> <i64 4294967295, i64 4294967295> |
| %1 = icmp sgt <2 x i64> %spec.store.select, zeroinitializer |
| %spec.store.select7 = select <2 x i1> %1, <2 x i64> %spec.store.select, <2 x i64> zeroinitializer |
| %conv6 = trunc <2 x i64> %spec.store.select7 to <2 x i32> |
| ret <2 x i32> %conv6 |
| } |
| |
| define <4 x i32> @stest_f32i32(<4 x float> %x) { |
| ; CHECK-LABEL: stest_f32i32: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzs v0.4s, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <4 x float> %x to <4 x i64> |
| %0 = icmp slt <4 x i64> %conv, <i64 2147483647, i64 2147483647, i64 2147483647, i64 2147483647> |
| %spec.store.select = select <4 x i1> %0, <4 x i64> %conv, <4 x i64> <i64 2147483647, i64 2147483647, i64 2147483647, i64 2147483647> |
| %1 = icmp sgt <4 x i64> %spec.store.select, <i64 -2147483648, i64 -2147483648, i64 -2147483648, i64 -2147483648> |
| %spec.store.select7 = select <4 x i1> %1, <4 x i64> %spec.store.select, <4 x i64> <i64 -2147483648, i64 -2147483648, i64 -2147483648, i64 -2147483648> |
| %conv6 = trunc <4 x i64> %spec.store.select7 to <4 x i32> |
| ret <4 x i32> %conv6 |
| } |
| |
| define <4 x i32> @utest_f32i32(<4 x float> %x) { |
| ; CHECK-LABEL: utest_f32i32: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <4 x float> %x to <4 x i64> |
| %0 = icmp ult <4 x i64> %conv, <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295> |
| %spec.store.select = select <4 x i1> %0, <4 x i64> %conv, <4 x i64> <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295> |
| %conv6 = trunc <4 x i64> %spec.store.select to <4 x i32> |
| ret <4 x i32> %conv6 |
| } |
| |
| define <4 x i32> @ustest_f32i32(<4 x float> %x) { |
| ; CHECK-LABEL: ustest_f32i32: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <4 x float> %x to <4 x i64> |
| %0 = icmp slt <4 x i64> %conv, <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295> |
| %spec.store.select = select <4 x i1> %0, <4 x i64> %conv, <4 x i64> <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295> |
| %1 = icmp sgt <4 x i64> %spec.store.select, zeroinitializer |
| %spec.store.select7 = select <4 x i1> %1, <4 x i64> %spec.store.select, <4 x i64> zeroinitializer |
| %conv6 = trunc <4 x i64> %spec.store.select7 to <4 x i32> |
| ret <4 x i32> %conv6 |
| } |
| |
| define <4 x i32> @stest_f16i32(<4 x half> %x) { |
| ; CHECK-LABEL: stest_f16i32: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtl v0.4s, v0.4h |
| ; CHECK-NEXT: fcvtzs v0.4s, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <4 x half> %x to <4 x i64> |
| %0 = icmp slt <4 x i64> %conv, <i64 2147483647, i64 2147483647, i64 2147483647, i64 2147483647> |
| %spec.store.select = select <4 x i1> %0, <4 x i64> %conv, <4 x i64> <i64 2147483647, i64 2147483647, i64 2147483647, i64 2147483647> |
| %1 = icmp sgt <4 x i64> %spec.store.select, <i64 -2147483648, i64 -2147483648, i64 -2147483648, i64 -2147483648> |
| %spec.store.select7 = select <4 x i1> %1, <4 x i64> %spec.store.select, <4 x i64> <i64 -2147483648, i64 -2147483648, i64 -2147483648, i64 -2147483648> |
| %conv6 = trunc <4 x i64> %spec.store.select7 to <4 x i32> |
| ret <4 x i32> %conv6 |
| } |
| |
| define <4 x i32> @utesth_f16i32(<4 x half> %x) { |
| ; CHECK-LABEL: utesth_f16i32: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtl v0.4s, v0.4h |
| ; CHECK-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <4 x half> %x to <4 x i64> |
| %0 = icmp ult <4 x i64> %conv, <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295> |
| %spec.store.select = select <4 x i1> %0, <4 x i64> %conv, <4 x i64> <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295> |
| %conv6 = trunc <4 x i64> %spec.store.select to <4 x i32> |
| ret <4 x i32> %conv6 |
| } |
| |
| define <4 x i32> @ustest_f16i32(<4 x half> %x) { |
| ; CHECK-LABEL: ustest_f16i32: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtl v0.4s, v0.4h |
| ; CHECK-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <4 x half> %x to <4 x i64> |
| %0 = icmp slt <4 x i64> %conv, <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295> |
| %spec.store.select = select <4 x i1> %0, <4 x i64> %conv, <4 x i64> <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295> |
| %1 = icmp sgt <4 x i64> %spec.store.select, zeroinitializer |
| %spec.store.select7 = select <4 x i1> %1, <4 x i64> %spec.store.select, <4 x i64> zeroinitializer |
| %conv6 = trunc <4 x i64> %spec.store.select7 to <4 x i32> |
| ret <4 x i32> %conv6 |
| } |
| |
| ; i16 saturate |
| |
| define <2 x i16> @stest_f64i16(<2 x double> %x) { |
| ; CHECK-LABEL: stest_f64i16: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzs v0.2d, v0.2d |
| ; CHECK-NEXT: movi v1.2s, #127, msl #8 |
| ; CHECK-NEXT: xtn v0.2s, v0.2d |
| ; CHECK-NEXT: smin v0.2s, v0.2s, v1.2s |
| ; CHECK-NEXT: mvni v1.2s, #127, msl #8 |
| ; CHECK-NEXT: smax v0.2s, v0.2s, v1.2s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x double> %x to <2 x i32> |
| %0 = icmp slt <2 x i32> %conv, <i32 32767, i32 32767> |
| %spec.store.select = select <2 x i1> %0, <2 x i32> %conv, <2 x i32> <i32 32767, i32 32767> |
| %1 = icmp sgt <2 x i32> %spec.store.select, <i32 -32768, i32 -32768> |
| %spec.store.select7 = select <2 x i1> %1, <2 x i32> %spec.store.select, <2 x i32> <i32 -32768, i32 -32768> |
| %conv6 = trunc <2 x i32> %spec.store.select7 to <2 x i16> |
| ret <2 x i16> %conv6 |
| } |
| |
| define <2 x i16> @utest_f64i16(<2 x double> %x) { |
| ; CHECK-LABEL: utest_f64i16: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzu v0.2d, v0.2d |
| ; CHECK-NEXT: movi d1, #0x00ffff0000ffff |
| ; CHECK-NEXT: xtn v0.2s, v0.2d |
| ; CHECK-NEXT: umin v0.2s, v0.2s, v1.2s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <2 x double> %x to <2 x i32> |
| %0 = icmp ult <2 x i32> %conv, <i32 65535, i32 65535> |
| %spec.store.select = select <2 x i1> %0, <2 x i32> %conv, <2 x i32> <i32 65535, i32 65535> |
| %conv6 = trunc <2 x i32> %spec.store.select to <2 x i16> |
| ret <2 x i16> %conv6 |
| } |
| |
| define <2 x i16> @ustest_f64i16(<2 x double> %x) { |
| ; CHECK-LABEL: ustest_f64i16: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzs v0.2d, v0.2d |
| ; CHECK-NEXT: movi d1, #0x00ffff0000ffff |
| ; CHECK-NEXT: xtn v0.2s, v0.2d |
| ; CHECK-NEXT: smin v0.2s, v0.2s, v1.2s |
| ; CHECK-NEXT: movi v1.2d, #0000000000000000 |
| ; CHECK-NEXT: smax v0.2s, v0.2s, v1.2s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x double> %x to <2 x i32> |
| %0 = icmp slt <2 x i32> %conv, <i32 65535, i32 65535> |
| %spec.store.select = select <2 x i1> %0, <2 x i32> %conv, <2 x i32> <i32 65535, i32 65535> |
| %1 = icmp sgt <2 x i32> %spec.store.select, zeroinitializer |
| %spec.store.select7 = select <2 x i1> %1, <2 x i32> %spec.store.select, <2 x i32> zeroinitializer |
| %conv6 = trunc <2 x i32> %spec.store.select7 to <2 x i16> |
| ret <2 x i16> %conv6 |
| } |
| |
| define <4 x i16> @stest_f32i16(<4 x float> %x) { |
| ; CHECK-LABEL: stest_f32i16: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzs v0.4s, v0.4s |
| ; CHECK-NEXT: sqxtn v0.4h, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <4 x float> %x to <4 x i32> |
| %0 = icmp slt <4 x i32> %conv, <i32 32767, i32 32767, i32 32767, i32 32767> |
| %spec.store.select = select <4 x i1> %0, <4 x i32> %conv, <4 x i32> <i32 32767, i32 32767, i32 32767, i32 32767> |
| %1 = icmp sgt <4 x i32> %spec.store.select, <i32 -32768, i32 -32768, i32 -32768, i32 -32768> |
| %spec.store.select7 = select <4 x i1> %1, <4 x i32> %spec.store.select, <4 x i32> <i32 -32768, i32 -32768, i32 -32768, i32 -32768> |
| %conv6 = trunc <4 x i32> %spec.store.select7 to <4 x i16> |
| ret <4 x i16> %conv6 |
| } |
| |
| define <4 x i16> @utest_f32i16(<4 x float> %x) { |
| ; CHECK-LABEL: utest_f32i16: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-NEXT: uqxtn v0.4h, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <4 x float> %x to <4 x i32> |
| %0 = icmp ult <4 x i32> %conv, <i32 65535, i32 65535, i32 65535, i32 65535> |
| %spec.store.select = select <4 x i1> %0, <4 x i32> %conv, <4 x i32> <i32 65535, i32 65535, i32 65535, i32 65535> |
| %conv6 = trunc <4 x i32> %spec.store.select to <4 x i16> |
| ret <4 x i16> %conv6 |
| } |
| |
| define <4 x i16> @ustest_f32i16(<4 x float> %x) { |
| ; CHECK-LABEL: ustest_f32i16: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-NEXT: uqxtn v0.4h, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <4 x float> %x to <4 x i32> |
| %0 = icmp slt <4 x i32> %conv, <i32 65535, i32 65535, i32 65535, i32 65535> |
| %spec.store.select = select <4 x i1> %0, <4 x i32> %conv, <4 x i32> <i32 65535, i32 65535, i32 65535, i32 65535> |
| %1 = icmp sgt <4 x i32> %spec.store.select, zeroinitializer |
| %spec.store.select7 = select <4 x i1> %1, <4 x i32> %spec.store.select, <4 x i32> zeroinitializer |
| %conv6 = trunc <4 x i32> %spec.store.select7 to <4 x i16> |
| ret <4 x i16> %conv6 |
| } |
| |
| define <8 x i16> @stest_f16i16(<8 x half> %x) { |
| ; CHECK-CVT-LABEL: stest_f16i16: |
| ; CHECK-CVT: // %bb.0: // %entry |
| ; CHECK-CVT-NEXT: fcvtl v1.4s, v0.4h |
| ; CHECK-CVT-NEXT: fcvtl2 v0.4s, v0.8h |
| ; CHECK-CVT-NEXT: fcvtzs v1.4s, v1.4s |
| ; CHECK-CVT-NEXT: fcvtzs v2.4s, v0.4s |
| ; CHECK-CVT-NEXT: sqxtn v0.4h, v1.4s |
| ; CHECK-CVT-NEXT: sqxtn2 v0.8h, v2.4s |
| ; CHECK-CVT-NEXT: ret |
| ; |
| ; CHECK-FP16-LABEL: stest_f16i16: |
| ; CHECK-FP16: // %bb.0: // %entry |
| ; CHECK-FP16-NEXT: fcvtzs v0.8h, v0.8h |
| ; CHECK-FP16-NEXT: ret |
| entry: |
| %conv = fptosi <8 x half> %x to <8 x i32> |
| %0 = icmp slt <8 x i32> %conv, <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767> |
| %spec.store.select = select <8 x i1> %0, <8 x i32> %conv, <8 x i32> <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767> |
| %1 = icmp sgt <8 x i32> %spec.store.select, <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768> |
| %spec.store.select7 = select <8 x i1> %1, <8 x i32> %spec.store.select, <8 x i32> <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768> |
| %conv6 = trunc <8 x i32> %spec.store.select7 to <8 x i16> |
| ret <8 x i16> %conv6 |
| } |
| |
| define <8 x i16> @utesth_f16i16(<8 x half> %x) { |
| ; CHECK-CVT-LABEL: utesth_f16i16: |
| ; CHECK-CVT: // %bb.0: // %entry |
| ; CHECK-CVT-NEXT: fcvtl2 v2.4s, v0.8h |
| ; CHECK-CVT-NEXT: fcvtl v0.4s, v0.4h |
| ; CHECK-CVT-NEXT: movi v1.2d, #0x00ffff0000ffff |
| ; CHECK-CVT-NEXT: fcvtzu v2.4s, v2.4s |
| ; CHECK-CVT-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-CVT-NEXT: umin v2.4s, v2.4s, v1.4s |
| ; CHECK-CVT-NEXT: umin v0.4s, v0.4s, v1.4s |
| ; CHECK-CVT-NEXT: uzp1 v0.8h, v0.8h, v2.8h |
| ; CHECK-CVT-NEXT: ret |
| ; |
| ; CHECK-FP16-LABEL: utesth_f16i16: |
| ; CHECK-FP16: // %bb.0: // %entry |
| ; CHECK-FP16-NEXT: fcvtzu v0.8h, v0.8h |
| ; CHECK-FP16-NEXT: ret |
| entry: |
| %conv = fptoui <8 x half> %x to <8 x i32> |
| %0 = icmp ult <8 x i32> %conv, <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535> |
| %spec.store.select = select <8 x i1> %0, <8 x i32> %conv, <8 x i32> <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535> |
| %conv6 = trunc <8 x i32> %spec.store.select to <8 x i16> |
| ret <8 x i16> %conv6 |
| } |
| |
| define <8 x i16> @ustest_f16i16(<8 x half> %x) { |
| ; CHECK-CVT-LABEL: ustest_f16i16: |
| ; CHECK-CVT: // %bb.0: // %entry |
| ; CHECK-CVT-NEXT: fcvtl2 v2.4s, v0.8h |
| ; CHECK-CVT-NEXT: fcvtl v0.4s, v0.4h |
| ; CHECK-CVT-NEXT: movi v1.2d, #0x00ffff0000ffff |
| ; CHECK-CVT-NEXT: fcvtzu v2.4s, v2.4s |
| ; CHECK-CVT-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-CVT-NEXT: umin v2.4s, v2.4s, v1.4s |
| ; CHECK-CVT-NEXT: umin v0.4s, v0.4s, v1.4s |
| ; CHECK-CVT-NEXT: uzp1 v0.8h, v0.8h, v2.8h |
| ; CHECK-CVT-NEXT: ret |
| ; |
| ; CHECK-FP16-LABEL: ustest_f16i16: |
| ; CHECK-FP16: // %bb.0: // %entry |
| ; CHECK-FP16-NEXT: fcvtzu v0.8h, v0.8h |
| ; CHECK-FP16-NEXT: ret |
| entry: |
| %conv = fptosi <8 x half> %x to <8 x i32> |
| %0 = icmp slt <8 x i32> %conv, <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535> |
| %spec.store.select = select <8 x i1> %0, <8 x i32> %conv, <8 x i32> <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535> |
| %1 = icmp sgt <8 x i32> %spec.store.select, zeroinitializer |
| %spec.store.select7 = select <8 x i1> %1, <8 x i32> %spec.store.select, <8 x i32> zeroinitializer |
| %conv6 = trunc <8 x i32> %spec.store.select7 to <8 x i16> |
| ret <8 x i16> %conv6 |
| } |
| |
| ; i64 saturate |
| |
| define <2 x i64> @stest_f64i64(<2 x double> %x) { |
| ; CHECK-LABEL: stest_f64i64: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzs v0.2d, v0.2d |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x double> %x to <2 x i128> |
| %0 = icmp slt <2 x i128> %conv, <i128 9223372036854775807, i128 9223372036854775807> |
| %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 9223372036854775807, i128 9223372036854775807> |
| %1 = icmp sgt <2 x i128> %spec.store.select, <i128 -9223372036854775808, i128 -9223372036854775808> |
| %spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808> |
| %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @utest_f64i64(<2 x double> %x) { |
| ; CHECK-LABEL: utest_f64i64: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: sub sp, sp, #48 |
| ; CHECK-NEXT: str x30, [sp, #16] // 8-byte Folded Spill |
| ; CHECK-NEXT: stp x20, x19, [sp, #32] // 16-byte Folded Spill |
| ; CHECK-NEXT: .cfi_def_cfa_offset 48 |
| ; CHECK-NEXT: .cfi_offset w19, -8 |
| ; CHECK-NEXT: .cfi_offset w20, -16 |
| ; CHECK-NEXT: .cfi_offset w30, -32 |
| ; CHECK-NEXT: str q0, [sp] // 16-byte Folded Spill |
| ; CHECK-NEXT: mov d0, v0.d[1] |
| ; CHECK-NEXT: bl __fixunsdfti |
| ; CHECK-NEXT: ldr q0, [sp] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov x19, x0 |
| ; CHECK-NEXT: mov x20, x1 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 killed $q0 |
| ; CHECK-NEXT: bl __fixunsdfti |
| ; CHECK-NEXT: cmp x1, #0 |
| ; CHECK-NEXT: ldr x30, [sp, #16] // 8-byte Folded Reload |
| ; CHECK-NEXT: csel x8, x0, xzr, eq |
| ; CHECK-NEXT: cmp x20, #0 |
| ; CHECK-NEXT: csel x9, x19, xzr, eq |
| ; CHECK-NEXT: fmov d0, x8 |
| ; CHECK-NEXT: fmov d1, x9 |
| ; CHECK-NEXT: ldp x20, x19, [sp, #32] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov v0.d[1], v1.d[0] |
| ; CHECK-NEXT: add sp, sp, #48 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <2 x double> %x to <2 x i128> |
| %0 = icmp ult <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616> |
| %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616> |
| %conv6 = trunc <2 x i128> %spec.store.select to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @ustest_f64i64(<2 x double> %x) { |
| ; CHECK-LABEL: ustest_f64i64: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: sub sp, sp, #48 |
| ; CHECK-NEXT: str x30, [sp, #16] // 8-byte Folded Spill |
| ; CHECK-NEXT: stp x20, x19, [sp, #32] // 16-byte Folded Spill |
| ; CHECK-NEXT: .cfi_def_cfa_offset 48 |
| ; CHECK-NEXT: .cfi_offset w19, -8 |
| ; CHECK-NEXT: .cfi_offset w20, -16 |
| ; CHECK-NEXT: .cfi_offset w30, -32 |
| ; CHECK-NEXT: str q0, [sp] // 16-byte Folded Spill |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 killed $q0 |
| ; CHECK-NEXT: bl __fixdfti |
| ; CHECK-NEXT: ldr q0, [sp] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov x19, x0 |
| ; CHECK-NEXT: mov x20, x1 |
| ; CHECK-NEXT: mov d0, v0.d[1] |
| ; CHECK-NEXT: bl __fixdfti |
| ; CHECK-NEXT: cmp x1, #1 |
| ; CHECK-NEXT: ldr x30, [sp, #16] // 8-byte Folded Reload |
| ; CHECK-NEXT: csel x8, x0, xzr, lt |
| ; CHECK-NEXT: csinc x9, x1, xzr, lt |
| ; CHECK-NEXT: cmp x20, #1 |
| ; CHECK-NEXT: csel x10, x19, xzr, lt |
| ; CHECK-NEXT: csinc x11, x20, xzr, lt |
| ; CHECK-NEXT: cmp xzr, x10 |
| ; CHECK-NEXT: ldp x20, x19, [sp, #32] // 16-byte Folded Reload |
| ; CHECK-NEXT: ngcs xzr, x11 |
| ; CHECK-NEXT: csel x10, x10, xzr, lt |
| ; CHECK-NEXT: cmp xzr, x8 |
| ; CHECK-NEXT: ngcs xzr, x9 |
| ; CHECK-NEXT: fmov d0, x10 |
| ; CHECK-NEXT: csel x8, x8, xzr, lt |
| ; CHECK-NEXT: fmov d1, x8 |
| ; CHECK-NEXT: mov v0.d[1], v1.d[0] |
| ; CHECK-NEXT: add sp, sp, #48 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x double> %x to <2 x i128> |
| %0 = icmp slt <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616> |
| %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616> |
| %1 = icmp sgt <2 x i128> %spec.store.select, zeroinitializer |
| %spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> zeroinitializer |
| %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @stest_f32i64(<2 x float> %x) { |
| ; CHECK-LABEL: stest_f32i64: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 def $q0 |
| ; CHECK-NEXT: mov s1, v0.s[1] |
| ; CHECK-NEXT: fcvtzs x8, s0 |
| ; CHECK-NEXT: fcvtzs x9, s1 |
| ; CHECK-NEXT: fmov d0, x8 |
| ; CHECK-NEXT: mov v0.d[1], x9 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x float> %x to <2 x i128> |
| %0 = icmp slt <2 x i128> %conv, <i128 9223372036854775807, i128 9223372036854775807> |
| %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 9223372036854775807, i128 9223372036854775807> |
| %1 = icmp sgt <2 x i128> %spec.store.select, <i128 -9223372036854775808, i128 -9223372036854775808> |
| %spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808> |
| %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @utest_f32i64(<2 x float> %x) { |
| ; CHECK-LABEL: utest_f32i64: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: sub sp, sp, #48 |
| ; CHECK-NEXT: str x30, [sp, #16] // 8-byte Folded Spill |
| ; CHECK-NEXT: stp x20, x19, [sp, #32] // 16-byte Folded Spill |
| ; CHECK-NEXT: .cfi_def_cfa_offset 48 |
| ; CHECK-NEXT: .cfi_offset w19, -8 |
| ; CHECK-NEXT: .cfi_offset w20, -16 |
| ; CHECK-NEXT: .cfi_offset w30, -32 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 def $q0 |
| ; CHECK-NEXT: str q0, [sp] // 16-byte Folded Spill |
| ; CHECK-NEXT: mov s0, v0.s[1] |
| ; CHECK-NEXT: bl __fixunssfti |
| ; CHECK-NEXT: ldr q0, [sp] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov x19, x0 |
| ; CHECK-NEXT: mov x20, x1 |
| ; CHECK-NEXT: // kill: def $s0 killed $s0 killed $q0 |
| ; CHECK-NEXT: bl __fixunssfti |
| ; CHECK-NEXT: cmp x1, #0 |
| ; CHECK-NEXT: ldr x30, [sp, #16] // 8-byte Folded Reload |
| ; CHECK-NEXT: csel x8, x0, xzr, eq |
| ; CHECK-NEXT: cmp x20, #0 |
| ; CHECK-NEXT: csel x9, x19, xzr, eq |
| ; CHECK-NEXT: fmov d0, x8 |
| ; CHECK-NEXT: fmov d1, x9 |
| ; CHECK-NEXT: ldp x20, x19, [sp, #32] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov v0.d[1], v1.d[0] |
| ; CHECK-NEXT: add sp, sp, #48 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <2 x float> %x to <2 x i128> |
| %0 = icmp ult <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616> |
| %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616> |
| %conv6 = trunc <2 x i128> %spec.store.select to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @ustest_f32i64(<2 x float> %x) { |
| ; CHECK-LABEL: ustest_f32i64: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: sub sp, sp, #48 |
| ; CHECK-NEXT: str x30, [sp, #16] // 8-byte Folded Spill |
| ; CHECK-NEXT: stp x20, x19, [sp, #32] // 16-byte Folded Spill |
| ; CHECK-NEXT: .cfi_def_cfa_offset 48 |
| ; CHECK-NEXT: .cfi_offset w19, -8 |
| ; CHECK-NEXT: .cfi_offset w20, -16 |
| ; CHECK-NEXT: .cfi_offset w30, -32 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 def $q0 |
| ; CHECK-NEXT: str q0, [sp] // 16-byte Folded Spill |
| ; CHECK-NEXT: // kill: def $s0 killed $s0 killed $q0 |
| ; CHECK-NEXT: bl __fixsfti |
| ; CHECK-NEXT: ldr q0, [sp] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov x19, x0 |
| ; CHECK-NEXT: mov x20, x1 |
| ; CHECK-NEXT: mov s0, v0.s[1] |
| ; CHECK-NEXT: bl __fixsfti |
| ; CHECK-NEXT: cmp x1, #1 |
| ; CHECK-NEXT: ldr x30, [sp, #16] // 8-byte Folded Reload |
| ; CHECK-NEXT: csinc x8, x1, xzr, lt |
| ; CHECK-NEXT: csel x9, x0, xzr, lt |
| ; CHECK-NEXT: cmp x20, #1 |
| ; CHECK-NEXT: csel x10, x19, xzr, lt |
| ; CHECK-NEXT: csinc x11, x20, xzr, lt |
| ; CHECK-NEXT: cmp xzr, x10 |
| ; CHECK-NEXT: ldp x20, x19, [sp, #32] // 16-byte Folded Reload |
| ; CHECK-NEXT: ngcs xzr, x11 |
| ; CHECK-NEXT: csel x10, x10, xzr, lt |
| ; CHECK-NEXT: cmp xzr, x9 |
| ; CHECK-NEXT: ngcs xzr, x8 |
| ; CHECK-NEXT: fmov d0, x10 |
| ; CHECK-NEXT: csel x8, x9, xzr, lt |
| ; CHECK-NEXT: fmov d1, x8 |
| ; CHECK-NEXT: mov v0.d[1], v1.d[0] |
| ; CHECK-NEXT: add sp, sp, #48 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x float> %x to <2 x i128> |
| %0 = icmp slt <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616> |
| %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616> |
| %1 = icmp sgt <2 x i128> %spec.store.select, zeroinitializer |
| %spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> zeroinitializer |
| %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @stest_f16i64(<2 x half> %x) { |
| ; CHECK-CVT-LABEL: stest_f16i64: |
| ; CHECK-CVT: // %bb.0: // %entry |
| ; CHECK-CVT-NEXT: // kill: def $d0 killed $d0 def $q0 |
| ; CHECK-CVT-NEXT: mov h1, v0.h[1] |
| ; CHECK-CVT-NEXT: fcvt s0, h0 |
| ; CHECK-CVT-NEXT: fcvt s1, h1 |
| ; CHECK-CVT-NEXT: fcvtzs x8, s0 |
| ; CHECK-CVT-NEXT: fcvtzs x9, s1 |
| ; CHECK-CVT-NEXT: fmov d0, x8 |
| ; CHECK-CVT-NEXT: mov v0.d[1], x9 |
| ; CHECK-CVT-NEXT: ret |
| ; |
| ; CHECK-FP16-LABEL: stest_f16i64: |
| ; CHECK-FP16: // %bb.0: // %entry |
| ; CHECK-FP16-NEXT: // kill: def $d0 killed $d0 def $q0 |
| ; CHECK-FP16-NEXT: mov h1, v0.h[1] |
| ; CHECK-FP16-NEXT: fcvtzs x8, h0 |
| ; CHECK-FP16-NEXT: fcvtzs x9, h1 |
| ; CHECK-FP16-NEXT: fmov d0, x8 |
| ; CHECK-FP16-NEXT: mov v0.d[1], x9 |
| ; CHECK-FP16-NEXT: ret |
| entry: |
| %conv = fptosi <2 x half> %x to <2 x i128> |
| %0 = icmp slt <2 x i128> %conv, <i128 9223372036854775807, i128 9223372036854775807> |
| %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 9223372036854775807, i128 9223372036854775807> |
| %1 = icmp sgt <2 x i128> %spec.store.select, <i128 -9223372036854775808, i128 -9223372036854775808> |
| %spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808> |
| %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @utesth_f16i64(<2 x half> %x) { |
| ; CHECK-LABEL: utesth_f16i64: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: sub sp, sp, #48 |
| ; CHECK-NEXT: str x30, [sp, #16] // 8-byte Folded Spill |
| ; CHECK-NEXT: stp x20, x19, [sp, #32] // 16-byte Folded Spill |
| ; CHECK-NEXT: .cfi_def_cfa_offset 48 |
| ; CHECK-NEXT: .cfi_offset w19, -8 |
| ; CHECK-NEXT: .cfi_offset w20, -16 |
| ; CHECK-NEXT: .cfi_offset w30, -32 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 def $q0 |
| ; CHECK-NEXT: str q0, [sp] // 16-byte Folded Spill |
| ; CHECK-NEXT: mov h0, v0.h[1] |
| ; CHECK-NEXT: bl __fixunshfti |
| ; CHECK-NEXT: ldr q0, [sp] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov x19, x0 |
| ; CHECK-NEXT: mov x20, x1 |
| ; CHECK-NEXT: // kill: def $h0 killed $h0 killed $q0 |
| ; CHECK-NEXT: bl __fixunshfti |
| ; CHECK-NEXT: cmp x1, #0 |
| ; CHECK-NEXT: ldr x30, [sp, #16] // 8-byte Folded Reload |
| ; CHECK-NEXT: csel x8, x0, xzr, eq |
| ; CHECK-NEXT: cmp x20, #0 |
| ; CHECK-NEXT: csel x9, x19, xzr, eq |
| ; CHECK-NEXT: fmov d0, x8 |
| ; CHECK-NEXT: fmov d1, x9 |
| ; CHECK-NEXT: ldp x20, x19, [sp, #32] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov v0.d[1], v1.d[0] |
| ; CHECK-NEXT: add sp, sp, #48 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <2 x half> %x to <2 x i128> |
| %0 = icmp ult <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616> |
| %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616> |
| %conv6 = trunc <2 x i128> %spec.store.select to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @ustest_f16i64(<2 x half> %x) { |
| ; CHECK-LABEL: ustest_f16i64: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: sub sp, sp, #48 |
| ; CHECK-NEXT: str x30, [sp, #16] // 8-byte Folded Spill |
| ; CHECK-NEXT: stp x20, x19, [sp, #32] // 16-byte Folded Spill |
| ; CHECK-NEXT: .cfi_def_cfa_offset 48 |
| ; CHECK-NEXT: .cfi_offset w19, -8 |
| ; CHECK-NEXT: .cfi_offset w20, -16 |
| ; CHECK-NEXT: .cfi_offset w30, -32 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 def $q0 |
| ; CHECK-NEXT: str q0, [sp] // 16-byte Folded Spill |
| ; CHECK-NEXT: // kill: def $h0 killed $h0 killed $q0 |
| ; CHECK-NEXT: bl __fixhfti |
| ; CHECK-NEXT: ldr q0, [sp] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov x19, x0 |
| ; CHECK-NEXT: mov x20, x1 |
| ; CHECK-NEXT: mov h0, v0.h[1] |
| ; CHECK-NEXT: bl __fixhfti |
| ; CHECK-NEXT: cmp x1, #1 |
| ; CHECK-NEXT: ldr x30, [sp, #16] // 8-byte Folded Reload |
| ; CHECK-NEXT: csinc x8, x1, xzr, lt |
| ; CHECK-NEXT: csel x9, x0, xzr, lt |
| ; CHECK-NEXT: cmp x20, #1 |
| ; CHECK-NEXT: csel x10, x19, xzr, lt |
| ; CHECK-NEXT: csinc x11, x20, xzr, lt |
| ; CHECK-NEXT: cmp xzr, x10 |
| ; CHECK-NEXT: ldp x20, x19, [sp, #32] // 16-byte Folded Reload |
| ; CHECK-NEXT: ngcs xzr, x11 |
| ; CHECK-NEXT: csel x10, x10, xzr, lt |
| ; CHECK-NEXT: cmp xzr, x9 |
| ; CHECK-NEXT: ngcs xzr, x8 |
| ; CHECK-NEXT: fmov d0, x10 |
| ; CHECK-NEXT: csel x8, x9, xzr, lt |
| ; CHECK-NEXT: fmov d1, x8 |
| ; CHECK-NEXT: mov v0.d[1], v1.d[0] |
| ; CHECK-NEXT: add sp, sp, #48 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x half> %x to <2 x i128> |
| %0 = icmp slt <2 x i128> %conv, <i128 18446744073709551616, i128 18446744073709551616> |
| %spec.store.select = select <2 x i1> %0, <2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616> |
| %1 = icmp sgt <2 x i128> %spec.store.select, zeroinitializer |
| %spec.store.select7 = select <2 x i1> %1, <2 x i128> %spec.store.select, <2 x i128> zeroinitializer |
| %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| |
| |
| ; i32 saturate |
| |
| define <2 x i32> @stest_f64i32_mm(<2 x double> %x) { |
| ; CHECK-LABEL: stest_f64i32_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: mov d1, v0.d[1] |
| ; CHECK-NEXT: fcvtzs w8, d0 |
| ; CHECK-NEXT: fcvtzs w9, d1 |
| ; CHECK-NEXT: fmov s0, w8 |
| ; CHECK-NEXT: mov v0.s[1], w9 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 killed $q0 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x double> %x to <2 x i64> |
| %spec.store.select = call <2 x i64> @llvm.smin.v2i64(<2 x i64> %conv, <2 x i64> <i64 2147483647, i64 2147483647>) |
| %spec.store.select7 = call <2 x i64> @llvm.smax.v2i64(<2 x i64> %spec.store.select, <2 x i64> <i64 -2147483648, i64 -2147483648>) |
| %conv6 = trunc <2 x i64> %spec.store.select7 to <2 x i32> |
| ret <2 x i32> %conv6 |
| } |
| |
| define <2 x i32> @utest_f64i32_mm(<2 x double> %x) { |
| ; CHECK-LABEL: utest_f64i32_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: mov d1, v0.d[1] |
| ; CHECK-NEXT: fcvtzu w8, d0 |
| ; CHECK-NEXT: fcvtzu w9, d1 |
| ; CHECK-NEXT: fmov s0, w8 |
| ; CHECK-NEXT: mov v0.s[1], w9 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 killed $q0 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <2 x double> %x to <2 x i64> |
| %spec.store.select = call <2 x i64> @llvm.umin.v2i64(<2 x i64> %conv, <2 x i64> <i64 4294967295, i64 4294967295>) |
| %conv6 = trunc <2 x i64> %spec.store.select to <2 x i32> |
| ret <2 x i32> %conv6 |
| } |
| |
| define <2 x i32> @ustest_f64i32_mm(<2 x double> %x) { |
| ; CHECK-LABEL: ustest_f64i32_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: mov d1, v0.d[1] |
| ; CHECK-NEXT: fcvtzu w8, d0 |
| ; CHECK-NEXT: fcvtzu w9, d1 |
| ; CHECK-NEXT: fmov s0, w8 |
| ; CHECK-NEXT: mov v0.s[1], w9 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 killed $q0 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x double> %x to <2 x i64> |
| %spec.store.select = call <2 x i64> @llvm.smin.v2i64(<2 x i64> %conv, <2 x i64> <i64 4294967295, i64 4294967295>) |
| %spec.store.select7 = call <2 x i64> @llvm.smax.v2i64(<2 x i64> %spec.store.select, <2 x i64> zeroinitializer) |
| %conv6 = trunc <2 x i64> %spec.store.select7 to <2 x i32> |
| ret <2 x i32> %conv6 |
| } |
| |
| define <4 x i32> @stest_f32i32_mm(<4 x float> %x) { |
| ; CHECK-LABEL: stest_f32i32_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzs v0.4s, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <4 x float> %x to <4 x i64> |
| %spec.store.select = call <4 x i64> @llvm.smin.v4i64(<4 x i64> %conv, <4 x i64> <i64 2147483647, i64 2147483647, i64 2147483647, i64 2147483647>) |
| %spec.store.select7 = call <4 x i64> @llvm.smax.v4i64(<4 x i64> %spec.store.select, <4 x i64> <i64 -2147483648, i64 -2147483648, i64 -2147483648, i64 -2147483648>) |
| %conv6 = trunc <4 x i64> %spec.store.select7 to <4 x i32> |
| ret <4 x i32> %conv6 |
| } |
| |
| define <4 x i32> @utest_f32i32_mm(<4 x float> %x) { |
| ; CHECK-LABEL: utest_f32i32_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <4 x float> %x to <4 x i64> |
| %spec.store.select = call <4 x i64> @llvm.umin.v4i64(<4 x i64> %conv, <4 x i64> <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295>) |
| %conv6 = trunc <4 x i64> %spec.store.select to <4 x i32> |
| ret <4 x i32> %conv6 |
| } |
| |
| define <4 x i32> @ustest_f32i32_mm(<4 x float> %x) { |
| ; CHECK-LABEL: ustest_f32i32_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <4 x float> %x to <4 x i64> |
| %spec.store.select = call <4 x i64> @llvm.smin.v4i64(<4 x i64> %conv, <4 x i64> <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295>) |
| %spec.store.select7 = call <4 x i64> @llvm.smax.v4i64(<4 x i64> %spec.store.select, <4 x i64> zeroinitializer) |
| %conv6 = trunc <4 x i64> %spec.store.select7 to <4 x i32> |
| ret <4 x i32> %conv6 |
| } |
| |
| define <4 x i32> @stest_f16i32_mm(<4 x half> %x) { |
| ; CHECK-LABEL: stest_f16i32_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtl v0.4s, v0.4h |
| ; CHECK-NEXT: fcvtzs v0.4s, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <4 x half> %x to <4 x i64> |
| %spec.store.select = call <4 x i64> @llvm.smin.v4i64(<4 x i64> %conv, <4 x i64> <i64 2147483647, i64 2147483647, i64 2147483647, i64 2147483647>) |
| %spec.store.select7 = call <4 x i64> @llvm.smax.v4i64(<4 x i64> %spec.store.select, <4 x i64> <i64 -2147483648, i64 -2147483648, i64 -2147483648, i64 -2147483648>) |
| %conv6 = trunc <4 x i64> %spec.store.select7 to <4 x i32> |
| ret <4 x i32> %conv6 |
| } |
| |
| define <4 x i32> @utesth_f16i32_mm(<4 x half> %x) { |
| ; CHECK-LABEL: utesth_f16i32_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtl v0.4s, v0.4h |
| ; CHECK-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <4 x half> %x to <4 x i64> |
| %spec.store.select = call <4 x i64> @llvm.umin.v4i64(<4 x i64> %conv, <4 x i64> <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295>) |
| %conv6 = trunc <4 x i64> %spec.store.select to <4 x i32> |
| ret <4 x i32> %conv6 |
| } |
| |
| define <4 x i32> @ustest_f16i32_mm(<4 x half> %x) { |
| ; CHECK-LABEL: ustest_f16i32_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtl v0.4s, v0.4h |
| ; CHECK-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <4 x half> %x to <4 x i64> |
| %spec.store.select = call <4 x i64> @llvm.smin.v4i64(<4 x i64> %conv, <4 x i64> <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295>) |
| %spec.store.select7 = call <4 x i64> @llvm.smax.v4i64(<4 x i64> %spec.store.select, <4 x i64> zeroinitializer) |
| %conv6 = trunc <4 x i64> %spec.store.select7 to <4 x i32> |
| ret <4 x i32> %conv6 |
| } |
| |
| ; i16 saturate |
| |
| define <2 x i16> @stest_f64i16_mm(<2 x double> %x) { |
| ; CHECK-LABEL: stest_f64i16_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzs v0.2d, v0.2d |
| ; CHECK-NEXT: movi v1.2s, #127, msl #8 |
| ; CHECK-NEXT: xtn v0.2s, v0.2d |
| ; CHECK-NEXT: smin v0.2s, v0.2s, v1.2s |
| ; CHECK-NEXT: mvni v1.2s, #127, msl #8 |
| ; CHECK-NEXT: smax v0.2s, v0.2s, v1.2s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x double> %x to <2 x i32> |
| %spec.store.select = call <2 x i32> @llvm.smin.v2i32(<2 x i32> %conv, <2 x i32> <i32 32767, i32 32767>) |
| %spec.store.select7 = call <2 x i32> @llvm.smax.v2i32(<2 x i32> %spec.store.select, <2 x i32> <i32 -32768, i32 -32768>) |
| %conv6 = trunc <2 x i32> %spec.store.select7 to <2 x i16> |
| ret <2 x i16> %conv6 |
| } |
| |
| define <2 x i16> @utest_f64i16_mm(<2 x double> %x) { |
| ; CHECK-LABEL: utest_f64i16_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzu v0.2d, v0.2d |
| ; CHECK-NEXT: movi d1, #0x00ffff0000ffff |
| ; CHECK-NEXT: xtn v0.2s, v0.2d |
| ; CHECK-NEXT: umin v0.2s, v0.2s, v1.2s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <2 x double> %x to <2 x i32> |
| %spec.store.select = call <2 x i32> @llvm.umin.v2i32(<2 x i32> %conv, <2 x i32> <i32 65535, i32 65535>) |
| %conv6 = trunc <2 x i32> %spec.store.select to <2 x i16> |
| ret <2 x i16> %conv6 |
| } |
| |
| define <2 x i16> @ustest_f64i16_mm(<2 x double> %x) { |
| ; CHECK-LABEL: ustest_f64i16_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzs v0.2d, v0.2d |
| ; CHECK-NEXT: movi d1, #0x00ffff0000ffff |
| ; CHECK-NEXT: xtn v0.2s, v0.2d |
| ; CHECK-NEXT: smin v0.2s, v0.2s, v1.2s |
| ; CHECK-NEXT: movi v1.2d, #0000000000000000 |
| ; CHECK-NEXT: smax v0.2s, v0.2s, v1.2s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x double> %x to <2 x i32> |
| %spec.store.select = call <2 x i32> @llvm.smin.v2i32(<2 x i32> %conv, <2 x i32> <i32 65535, i32 65535>) |
| %spec.store.select7 = call <2 x i32> @llvm.smax.v2i32(<2 x i32> %spec.store.select, <2 x i32> zeroinitializer) |
| %conv6 = trunc <2 x i32> %spec.store.select7 to <2 x i16> |
| ret <2 x i16> %conv6 |
| } |
| |
| define <4 x i16> @stest_f32i16_mm(<4 x float> %x) { |
| ; CHECK-LABEL: stest_f32i16_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzs v0.4s, v0.4s |
| ; CHECK-NEXT: sqxtn v0.4h, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <4 x float> %x to <4 x i32> |
| %spec.store.select = call <4 x i32> @llvm.smin.v4i32(<4 x i32> %conv, <4 x i32> <i32 32767, i32 32767, i32 32767, i32 32767>) |
| %spec.store.select7 = call <4 x i32> @llvm.smax.v4i32(<4 x i32> %spec.store.select, <4 x i32> <i32 -32768, i32 -32768, i32 -32768, i32 -32768>) |
| %conv6 = trunc <4 x i32> %spec.store.select7 to <4 x i16> |
| ret <4 x i16> %conv6 |
| } |
| |
| define <4 x i16> @utest_f32i16_mm(<4 x float> %x) { |
| ; CHECK-LABEL: utest_f32i16_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-NEXT: uqxtn v0.4h, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <4 x float> %x to <4 x i32> |
| %spec.store.select = call <4 x i32> @llvm.umin.v4i32(<4 x i32> %conv, <4 x i32> <i32 65535, i32 65535, i32 65535, i32 65535>) |
| %conv6 = trunc <4 x i32> %spec.store.select to <4 x i16> |
| ret <4 x i16> %conv6 |
| } |
| |
| define <4 x i16> @ustest_f32i16_mm(<4 x float> %x) { |
| ; CHECK-LABEL: ustest_f32i16_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-NEXT: uqxtn v0.4h, v0.4s |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <4 x float> %x to <4 x i32> |
| %spec.store.select = call <4 x i32> @llvm.smin.v4i32(<4 x i32> %conv, <4 x i32> <i32 65535, i32 65535, i32 65535, i32 65535>) |
| %spec.store.select7 = call <4 x i32> @llvm.smax.v4i32(<4 x i32> %spec.store.select, <4 x i32> zeroinitializer) |
| %conv6 = trunc <4 x i32> %spec.store.select7 to <4 x i16> |
| ret <4 x i16> %conv6 |
| } |
| |
| define <8 x i16> @stest_f16i16_mm(<8 x half> %x) { |
| ; CHECK-CVT-LABEL: stest_f16i16_mm: |
| ; CHECK-CVT: // %bb.0: // %entry |
| ; CHECK-CVT-NEXT: fcvtl v1.4s, v0.4h |
| ; CHECK-CVT-NEXT: fcvtl2 v0.4s, v0.8h |
| ; CHECK-CVT-NEXT: fcvtzs v1.4s, v1.4s |
| ; CHECK-CVT-NEXT: fcvtzs v2.4s, v0.4s |
| ; CHECK-CVT-NEXT: sqxtn v0.4h, v1.4s |
| ; CHECK-CVT-NEXT: sqxtn2 v0.8h, v2.4s |
| ; CHECK-CVT-NEXT: ret |
| ; |
| ; CHECK-FP16-LABEL: stest_f16i16_mm: |
| ; CHECK-FP16: // %bb.0: // %entry |
| ; CHECK-FP16-NEXT: fcvtzs v0.8h, v0.8h |
| ; CHECK-FP16-NEXT: ret |
| entry: |
| %conv = fptosi <8 x half> %x to <8 x i32> |
| %spec.store.select = call <8 x i32> @llvm.smin.v8i32(<8 x i32> %conv, <8 x i32> <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>) |
| %spec.store.select7 = call <8 x i32> @llvm.smax.v8i32(<8 x i32> %spec.store.select, <8 x i32> <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>) |
| %conv6 = trunc <8 x i32> %spec.store.select7 to <8 x i16> |
| ret <8 x i16> %conv6 |
| } |
| |
| define <8 x i16> @utesth_f16i16_mm(<8 x half> %x) { |
| ; CHECK-CVT-LABEL: utesth_f16i16_mm: |
| ; CHECK-CVT: // %bb.0: // %entry |
| ; CHECK-CVT-NEXT: fcvtl2 v2.4s, v0.8h |
| ; CHECK-CVT-NEXT: fcvtl v0.4s, v0.4h |
| ; CHECK-CVT-NEXT: movi v1.2d, #0x00ffff0000ffff |
| ; CHECK-CVT-NEXT: fcvtzu v2.4s, v2.4s |
| ; CHECK-CVT-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-CVT-NEXT: umin v2.4s, v2.4s, v1.4s |
| ; CHECK-CVT-NEXT: umin v0.4s, v0.4s, v1.4s |
| ; CHECK-CVT-NEXT: uzp1 v0.8h, v0.8h, v2.8h |
| ; CHECK-CVT-NEXT: ret |
| ; |
| ; CHECK-FP16-LABEL: utesth_f16i16_mm: |
| ; CHECK-FP16: // %bb.0: // %entry |
| ; CHECK-FP16-NEXT: fcvtzu v0.8h, v0.8h |
| ; CHECK-FP16-NEXT: ret |
| entry: |
| %conv = fptoui <8 x half> %x to <8 x i32> |
| %spec.store.select = call <8 x i32> @llvm.umin.v8i32(<8 x i32> %conv, <8 x i32> <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>) |
| %conv6 = trunc <8 x i32> %spec.store.select to <8 x i16> |
| ret <8 x i16> %conv6 |
| } |
| |
| define <8 x i16> @ustest_f16i16_mm(<8 x half> %x) { |
| ; CHECK-CVT-LABEL: ustest_f16i16_mm: |
| ; CHECK-CVT: // %bb.0: // %entry |
| ; CHECK-CVT-NEXT: fcvtl2 v2.4s, v0.8h |
| ; CHECK-CVT-NEXT: fcvtl v0.4s, v0.4h |
| ; CHECK-CVT-NEXT: movi v1.2d, #0x00ffff0000ffff |
| ; CHECK-CVT-NEXT: fcvtzu v2.4s, v2.4s |
| ; CHECK-CVT-NEXT: fcvtzu v0.4s, v0.4s |
| ; CHECK-CVT-NEXT: umin v2.4s, v2.4s, v1.4s |
| ; CHECK-CVT-NEXT: umin v0.4s, v0.4s, v1.4s |
| ; CHECK-CVT-NEXT: uzp1 v0.8h, v0.8h, v2.8h |
| ; CHECK-CVT-NEXT: ret |
| ; |
| ; CHECK-FP16-LABEL: ustest_f16i16_mm: |
| ; CHECK-FP16: // %bb.0: // %entry |
| ; CHECK-FP16-NEXT: fcvtzu v0.8h, v0.8h |
| ; CHECK-FP16-NEXT: ret |
| entry: |
| %conv = fptosi <8 x half> %x to <8 x i32> |
| %spec.store.select = call <8 x i32> @llvm.smin.v8i32(<8 x i32> %conv, <8 x i32> <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>) |
| %spec.store.select7 = call <8 x i32> @llvm.smax.v8i32(<8 x i32> %spec.store.select, <8 x i32> zeroinitializer) |
| %conv6 = trunc <8 x i32> %spec.store.select7 to <8 x i16> |
| ret <8 x i16> %conv6 |
| } |
| |
| ; i64 saturate |
| |
| define <2 x i64> @stest_f64i64_mm(<2 x double> %x) { |
| ; CHECK-LABEL: stest_f64i64_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: fcvtzs v0.2d, v0.2d |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x double> %x to <2 x i128> |
| %spec.store.select = call <2 x i128> @llvm.smin.v2i128(<2 x i128> %conv, <2 x i128> <i128 9223372036854775807, i128 9223372036854775807>) |
| %spec.store.select7 = call <2 x i128> @llvm.smax.v2i128(<2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808>) |
| %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @utest_f64i64_mm(<2 x double> %x) { |
| ; CHECK-LABEL: utest_f64i64_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: sub sp, sp, #48 |
| ; CHECK-NEXT: str x30, [sp, #16] // 8-byte Folded Spill |
| ; CHECK-NEXT: stp x20, x19, [sp, #32] // 16-byte Folded Spill |
| ; CHECK-NEXT: .cfi_def_cfa_offset 48 |
| ; CHECK-NEXT: .cfi_offset w19, -8 |
| ; CHECK-NEXT: .cfi_offset w20, -16 |
| ; CHECK-NEXT: .cfi_offset w30, -32 |
| ; CHECK-NEXT: str q0, [sp] // 16-byte Folded Spill |
| ; CHECK-NEXT: mov d0, v0.d[1] |
| ; CHECK-NEXT: bl __fixunsdfti |
| ; CHECK-NEXT: ldr q0, [sp] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov x19, x0 |
| ; CHECK-NEXT: mov x20, x1 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 killed $q0 |
| ; CHECK-NEXT: bl __fixunsdfti |
| ; CHECK-NEXT: cmp x1, #0 |
| ; CHECK-NEXT: ldr x30, [sp, #16] // 8-byte Folded Reload |
| ; CHECK-NEXT: csel x8, x0, xzr, eq |
| ; CHECK-NEXT: cmp x20, #0 |
| ; CHECK-NEXT: csel x9, x19, xzr, eq |
| ; CHECK-NEXT: fmov d0, x8 |
| ; CHECK-NEXT: fmov d1, x9 |
| ; CHECK-NEXT: ldp x20, x19, [sp, #32] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov v0.d[1], v1.d[0] |
| ; CHECK-NEXT: add sp, sp, #48 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <2 x double> %x to <2 x i128> |
| %spec.store.select = call <2 x i128> @llvm.umin.v2i128(<2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>) |
| %conv6 = trunc <2 x i128> %spec.store.select to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @ustest_f64i64_mm(<2 x double> %x) { |
| ; CHECK-LABEL: ustest_f64i64_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: sub sp, sp, #48 |
| ; CHECK-NEXT: str x30, [sp, #16] // 8-byte Folded Spill |
| ; CHECK-NEXT: stp x20, x19, [sp, #32] // 16-byte Folded Spill |
| ; CHECK-NEXT: .cfi_def_cfa_offset 48 |
| ; CHECK-NEXT: .cfi_offset w19, -8 |
| ; CHECK-NEXT: .cfi_offset w20, -16 |
| ; CHECK-NEXT: .cfi_offset w30, -32 |
| ; CHECK-NEXT: str q0, [sp] // 16-byte Folded Spill |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 killed $q0 |
| ; CHECK-NEXT: bl __fixdfti |
| ; CHECK-NEXT: ldr q0, [sp] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov x19, x0 |
| ; CHECK-NEXT: mov x20, x1 |
| ; CHECK-NEXT: mov d0, v0.d[1] |
| ; CHECK-NEXT: bl __fixdfti |
| ; CHECK-NEXT: cmp x1, #1 |
| ; CHECK-NEXT: ldr x30, [sp, #16] // 8-byte Folded Reload |
| ; CHECK-NEXT: csel x8, x0, xzr, lt |
| ; CHECK-NEXT: csinc x9, x1, xzr, lt |
| ; CHECK-NEXT: cmp x20, #1 |
| ; CHECK-NEXT: csinc x10, x20, xzr, lt |
| ; CHECK-NEXT: csel x11, x19, xzr, lt |
| ; CHECK-NEXT: cmp x10, #0 |
| ; CHECK-NEXT: ldp x20, x19, [sp, #32] // 16-byte Folded Reload |
| ; CHECK-NEXT: csel x10, xzr, x11, lt |
| ; CHECK-NEXT: cmp x9, #0 |
| ; CHECK-NEXT: csel x8, xzr, x8, lt |
| ; CHECK-NEXT: fmov d0, x10 |
| ; CHECK-NEXT: fmov d1, x8 |
| ; CHECK-NEXT: mov v0.d[1], v1.d[0] |
| ; CHECK-NEXT: add sp, sp, #48 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x double> %x to <2 x i128> |
| %spec.store.select = call <2 x i128> @llvm.smin.v2i128(<2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>) |
| %spec.store.select7 = call <2 x i128> @llvm.smax.v2i128(<2 x i128> %spec.store.select, <2 x i128> zeroinitializer) |
| %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @stest_f32i64_mm(<2 x float> %x) { |
| ; CHECK-LABEL: stest_f32i64_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 def $q0 |
| ; CHECK-NEXT: mov s1, v0.s[1] |
| ; CHECK-NEXT: fcvtzs x8, s0 |
| ; CHECK-NEXT: fcvtzs x9, s1 |
| ; CHECK-NEXT: fmov d0, x8 |
| ; CHECK-NEXT: mov v0.d[1], x9 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x float> %x to <2 x i128> |
| %spec.store.select = call <2 x i128> @llvm.smin.v2i128(<2 x i128> %conv, <2 x i128> <i128 9223372036854775807, i128 9223372036854775807>) |
| %spec.store.select7 = call <2 x i128> @llvm.smax.v2i128(<2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808>) |
| %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @utest_f32i64_mm(<2 x float> %x) { |
| ; CHECK-LABEL: utest_f32i64_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: sub sp, sp, #48 |
| ; CHECK-NEXT: str x30, [sp, #16] // 8-byte Folded Spill |
| ; CHECK-NEXT: stp x20, x19, [sp, #32] // 16-byte Folded Spill |
| ; CHECK-NEXT: .cfi_def_cfa_offset 48 |
| ; CHECK-NEXT: .cfi_offset w19, -8 |
| ; CHECK-NEXT: .cfi_offset w20, -16 |
| ; CHECK-NEXT: .cfi_offset w30, -32 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 def $q0 |
| ; CHECK-NEXT: str q0, [sp] // 16-byte Folded Spill |
| ; CHECK-NEXT: mov s0, v0.s[1] |
| ; CHECK-NEXT: bl __fixunssfti |
| ; CHECK-NEXT: ldr q0, [sp] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov x19, x0 |
| ; CHECK-NEXT: mov x20, x1 |
| ; CHECK-NEXT: // kill: def $s0 killed $s0 killed $q0 |
| ; CHECK-NEXT: bl __fixunssfti |
| ; CHECK-NEXT: cmp x1, #0 |
| ; CHECK-NEXT: ldr x30, [sp, #16] // 8-byte Folded Reload |
| ; CHECK-NEXT: csel x8, x0, xzr, eq |
| ; CHECK-NEXT: cmp x20, #0 |
| ; CHECK-NEXT: csel x9, x19, xzr, eq |
| ; CHECK-NEXT: fmov d0, x8 |
| ; CHECK-NEXT: fmov d1, x9 |
| ; CHECK-NEXT: ldp x20, x19, [sp, #32] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov v0.d[1], v1.d[0] |
| ; CHECK-NEXT: add sp, sp, #48 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <2 x float> %x to <2 x i128> |
| %spec.store.select = call <2 x i128> @llvm.umin.v2i128(<2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>) |
| %conv6 = trunc <2 x i128> %spec.store.select to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @ustest_f32i64_mm(<2 x float> %x) { |
| ; CHECK-LABEL: ustest_f32i64_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: sub sp, sp, #48 |
| ; CHECK-NEXT: str x30, [sp, #16] // 8-byte Folded Spill |
| ; CHECK-NEXT: stp x20, x19, [sp, #32] // 16-byte Folded Spill |
| ; CHECK-NEXT: .cfi_def_cfa_offset 48 |
| ; CHECK-NEXT: .cfi_offset w19, -8 |
| ; CHECK-NEXT: .cfi_offset w20, -16 |
| ; CHECK-NEXT: .cfi_offset w30, -32 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 def $q0 |
| ; CHECK-NEXT: str q0, [sp] // 16-byte Folded Spill |
| ; CHECK-NEXT: // kill: def $s0 killed $s0 killed $q0 |
| ; CHECK-NEXT: bl __fixsfti |
| ; CHECK-NEXT: ldr q0, [sp] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov x19, x0 |
| ; CHECK-NEXT: mov x20, x1 |
| ; CHECK-NEXT: mov s0, v0.s[1] |
| ; CHECK-NEXT: bl __fixsfti |
| ; CHECK-NEXT: cmp x1, #1 |
| ; CHECK-NEXT: ldr x30, [sp, #16] // 8-byte Folded Reload |
| ; CHECK-NEXT: csel x8, x0, xzr, lt |
| ; CHECK-NEXT: csinc x9, x1, xzr, lt |
| ; CHECK-NEXT: cmp x20, #1 |
| ; CHECK-NEXT: csinc x10, x20, xzr, lt |
| ; CHECK-NEXT: csel x11, x19, xzr, lt |
| ; CHECK-NEXT: cmp x10, #0 |
| ; CHECK-NEXT: ldp x20, x19, [sp, #32] // 16-byte Folded Reload |
| ; CHECK-NEXT: csel x10, xzr, x11, lt |
| ; CHECK-NEXT: cmp x9, #0 |
| ; CHECK-NEXT: csel x8, xzr, x8, lt |
| ; CHECK-NEXT: fmov d0, x10 |
| ; CHECK-NEXT: fmov d1, x8 |
| ; CHECK-NEXT: mov v0.d[1], v1.d[0] |
| ; CHECK-NEXT: add sp, sp, #48 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x float> %x to <2 x i128> |
| %spec.store.select = call <2 x i128> @llvm.smin.v2i128(<2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>) |
| %spec.store.select7 = call <2 x i128> @llvm.smax.v2i128(<2 x i128> %spec.store.select, <2 x i128> zeroinitializer) |
| %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @stest_f16i64_mm(<2 x half> %x) { |
| ; CHECK-CVT-LABEL: stest_f16i64_mm: |
| ; CHECK-CVT: // %bb.0: // %entry |
| ; CHECK-CVT-NEXT: // kill: def $d0 killed $d0 def $q0 |
| ; CHECK-CVT-NEXT: mov h1, v0.h[1] |
| ; CHECK-CVT-NEXT: fcvt s0, h0 |
| ; CHECK-CVT-NEXT: fcvt s1, h1 |
| ; CHECK-CVT-NEXT: fcvtzs x8, s0 |
| ; CHECK-CVT-NEXT: fcvtzs x9, s1 |
| ; CHECK-CVT-NEXT: fmov d0, x8 |
| ; CHECK-CVT-NEXT: mov v0.d[1], x9 |
| ; CHECK-CVT-NEXT: ret |
| ; |
| ; CHECK-FP16-LABEL: stest_f16i64_mm: |
| ; CHECK-FP16: // %bb.0: // %entry |
| ; CHECK-FP16-NEXT: // kill: def $d0 killed $d0 def $q0 |
| ; CHECK-FP16-NEXT: mov h1, v0.h[1] |
| ; CHECK-FP16-NEXT: fcvtzs x8, h0 |
| ; CHECK-FP16-NEXT: fcvtzs x9, h1 |
| ; CHECK-FP16-NEXT: fmov d0, x8 |
| ; CHECK-FP16-NEXT: mov v0.d[1], x9 |
| ; CHECK-FP16-NEXT: ret |
| entry: |
| %conv = fptosi <2 x half> %x to <2 x i128> |
| %spec.store.select = call <2 x i128> @llvm.smin.v2i128(<2 x i128> %conv, <2 x i128> <i128 9223372036854775807, i128 9223372036854775807>) |
| %spec.store.select7 = call <2 x i128> @llvm.smax.v2i128(<2 x i128> %spec.store.select, <2 x i128> <i128 -9223372036854775808, i128 -9223372036854775808>) |
| %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @utesth_f16i64_mm(<2 x half> %x) { |
| ; CHECK-LABEL: utesth_f16i64_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: sub sp, sp, #48 |
| ; CHECK-NEXT: str x30, [sp, #16] // 8-byte Folded Spill |
| ; CHECK-NEXT: stp x20, x19, [sp, #32] // 16-byte Folded Spill |
| ; CHECK-NEXT: .cfi_def_cfa_offset 48 |
| ; CHECK-NEXT: .cfi_offset w19, -8 |
| ; CHECK-NEXT: .cfi_offset w20, -16 |
| ; CHECK-NEXT: .cfi_offset w30, -32 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 def $q0 |
| ; CHECK-NEXT: str q0, [sp] // 16-byte Folded Spill |
| ; CHECK-NEXT: mov h0, v0.h[1] |
| ; CHECK-NEXT: bl __fixunshfti |
| ; CHECK-NEXT: ldr q0, [sp] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov x19, x0 |
| ; CHECK-NEXT: mov x20, x1 |
| ; CHECK-NEXT: // kill: def $h0 killed $h0 killed $q0 |
| ; CHECK-NEXT: bl __fixunshfti |
| ; CHECK-NEXT: cmp x1, #0 |
| ; CHECK-NEXT: ldr x30, [sp, #16] // 8-byte Folded Reload |
| ; CHECK-NEXT: csel x8, x0, xzr, eq |
| ; CHECK-NEXT: cmp x20, #0 |
| ; CHECK-NEXT: csel x9, x19, xzr, eq |
| ; CHECK-NEXT: fmov d0, x8 |
| ; CHECK-NEXT: fmov d1, x9 |
| ; CHECK-NEXT: ldp x20, x19, [sp, #32] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov v0.d[1], v1.d[0] |
| ; CHECK-NEXT: add sp, sp, #48 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptoui <2 x half> %x to <2 x i128> |
| %spec.store.select = call <2 x i128> @llvm.umin.v2i128(<2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>) |
| %conv6 = trunc <2 x i128> %spec.store.select to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| define <2 x i64> @ustest_f16i64_mm(<2 x half> %x) { |
| ; CHECK-LABEL: ustest_f16i64_mm: |
| ; CHECK: // %bb.0: // %entry |
| ; CHECK-NEXT: sub sp, sp, #48 |
| ; CHECK-NEXT: str x30, [sp, #16] // 8-byte Folded Spill |
| ; CHECK-NEXT: stp x20, x19, [sp, #32] // 16-byte Folded Spill |
| ; CHECK-NEXT: .cfi_def_cfa_offset 48 |
| ; CHECK-NEXT: .cfi_offset w19, -8 |
| ; CHECK-NEXT: .cfi_offset w20, -16 |
| ; CHECK-NEXT: .cfi_offset w30, -32 |
| ; CHECK-NEXT: // kill: def $d0 killed $d0 def $q0 |
| ; CHECK-NEXT: str q0, [sp] // 16-byte Folded Spill |
| ; CHECK-NEXT: // kill: def $h0 killed $h0 killed $q0 |
| ; CHECK-NEXT: bl __fixhfti |
| ; CHECK-NEXT: ldr q0, [sp] // 16-byte Folded Reload |
| ; CHECK-NEXT: mov x19, x0 |
| ; CHECK-NEXT: mov x20, x1 |
| ; CHECK-NEXT: mov h0, v0.h[1] |
| ; CHECK-NEXT: bl __fixhfti |
| ; CHECK-NEXT: cmp x1, #1 |
| ; CHECK-NEXT: ldr x30, [sp, #16] // 8-byte Folded Reload |
| ; CHECK-NEXT: csel x8, x0, xzr, lt |
| ; CHECK-NEXT: csinc x9, x1, xzr, lt |
| ; CHECK-NEXT: cmp x20, #1 |
| ; CHECK-NEXT: csinc x10, x20, xzr, lt |
| ; CHECK-NEXT: csel x11, x19, xzr, lt |
| ; CHECK-NEXT: cmp x10, #0 |
| ; CHECK-NEXT: ldp x20, x19, [sp, #32] // 16-byte Folded Reload |
| ; CHECK-NEXT: csel x10, xzr, x11, lt |
| ; CHECK-NEXT: cmp x9, #0 |
| ; CHECK-NEXT: csel x8, xzr, x8, lt |
| ; CHECK-NEXT: fmov d0, x10 |
| ; CHECK-NEXT: fmov d1, x8 |
| ; CHECK-NEXT: mov v0.d[1], v1.d[0] |
| ; CHECK-NEXT: add sp, sp, #48 |
| ; CHECK-NEXT: ret |
| entry: |
| %conv = fptosi <2 x half> %x to <2 x i128> |
| %spec.store.select = call <2 x i128> @llvm.smin.v2i128(<2 x i128> %conv, <2 x i128> <i128 18446744073709551616, i128 18446744073709551616>) |
| %spec.store.select7 = call <2 x i128> @llvm.smax.v2i128(<2 x i128> %spec.store.select, <2 x i128> zeroinitializer) |
| %conv6 = trunc <2 x i128> %spec.store.select7 to <2 x i64> |
| ret <2 x i64> %conv6 |
| } |
| |
| declare <2 x i32> @llvm.smin.v2i32(<2 x i32>, <2 x i32>) |
| declare <2 x i32> @llvm.smax.v2i32(<2 x i32>, <2 x i32>) |
| declare <2 x i32> @llvm.umin.v2i32(<2 x i32>, <2 x i32>) |
| declare <4 x i32> @llvm.smin.v4i32(<4 x i32>, <4 x i32>) |
| declare <4 x i32> @llvm.smax.v4i32(<4 x i32>, <4 x i32>) |
| declare <4 x i32> @llvm.umin.v4i32(<4 x i32>, <4 x i32>) |
| declare <8 x i32> @llvm.smin.v8i32(<8 x i32>, <8 x i32>) |
| declare <8 x i32> @llvm.smax.v8i32(<8 x i32>, <8 x i32>) |
| declare <8 x i32> @llvm.umin.v8i32(<8 x i32>, <8 x i32>) |
| declare <2 x i64> @llvm.smin.v2i64(<2 x i64>, <2 x i64>) |
| declare <2 x i64> @llvm.smax.v2i64(<2 x i64>, <2 x i64>) |
| declare <2 x i64> @llvm.umin.v2i64(<2 x i64>, <2 x i64>) |
| declare <4 x i64> @llvm.smin.v4i64(<4 x i64>, <4 x i64>) |
| declare <4 x i64> @llvm.smax.v4i64(<4 x i64>, <4 x i64>) |
| declare <4 x i64> @llvm.umin.v4i64(<4 x i64>, <4 x i64>) |
| declare <2 x i128> @llvm.smin.v2i128(<2 x i128>, <2 x i128>) |
| declare <2 x i128> @llvm.smax.v2i128(<2 x i128>, <2 x i128>) |
| declare <2 x i128> @llvm.umin.v2i128(<2 x i128>, <2 x i128>) |