File: arm-neon-directed-rounding-constrained.c

package info (click to toggle)
llvm-toolchain-21 1%3A21.1.0-1
  • links: PTS, VCS
  • area: main
  • in suites: sid
  • size: 2,235,796 kB
  • sloc: cpp: 7,617,614; ansic: 1,433,901; asm: 1,058,726; python: 252,096; f90: 94,671; objc: 70,753; lisp: 42,813; pascal: 18,401; sh: 10,032; ml: 5,111; perl: 4,720; awk: 3,523; makefile: 3,401; javascript: 2,272; xml: 892; fortran: 770
file content (61 lines) | stat: -rw-r--r-- 3,661 bytes parent folder | download | duplicates (4)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
// RUN: %clang_cc1 -triple thumbv8-linux-gnueabihf -target-cpu cortex-a57 \
// RUN:     -ffreestanding -disable-O0-optnone -emit-llvm %s -o - | \
// RUN:     opt -S -passes=mem2reg,sroa | FileCheck -check-prefixes=COMMON,COMMONIR,UNCONSTRAINED %s
// RUN: %clang_cc1 -triple arm64-linux-gnueabihf -target-feature +neon \
// RUN:     -ffreestanding -disable-O0-optnone -emit-llvm %s -o - | \
// RUN:     opt -S -passes=mem2reg,sroa | FileCheck -check-prefixes=COMMON,COMMONIR,UNCONSTRAINED %s

// RUN: %clang_cc1 -triple thumbv8-linux-gnueabihf -target-cpu cortex-a57 \
// RUN:     -ffp-exception-behavior=strict \
// RUN:     -fexperimental-strict-floating-point \
// RUN:     -ffreestanding -disable-O0-optnone -emit-llvm %s -o - | \
// RUN:     opt -S -passes=mem2reg,sroa | FileCheck -check-prefixes=COMMON,COMMONIR,CONSTRAINED %s
// RUN: %clang_cc1 -triple arm64-linux-gnueabihf -target-feature +neon \
// RUN:     -ffp-exception-behavior=strict \
// RUN:     -ffreestanding -disable-O0-optnone -emit-llvm %s -o - | \
// RUN:     opt -S -passes=mem2reg,sroa | FileCheck -check-prefixes=COMMON,COMMONIR,CONSTRAINED %s

// RUN: %clang_cc1 -triple thumbv8-linux-gnueabihf -target-cpu cortex-a57 \
// RUN:     -ffreestanding -disable-O0-optnone -emit-llvm %s -o - | \
// RUN:     opt -S -passes=mem2reg,sroa | llc -o=- - | FileCheck -check-prefixes=COMMON,CHECK-ASM32 %s
// RUN: %clang_cc1 -triple arm64-linux-gnueabihf -target-feature +neon \
// RUN:     -ffreestanding -disable-O0-optnone -emit-llvm %s -o - | \
// RUN:     opt -S -passes=mem2reg,sroa | llc -o=- - | FileCheck -check-prefixes=COMMON,CHECK-ASM64 %s

// RUN: %clang_cc1 -triple thumbv8-linux-gnueabihf -target-cpu cortex-a57 \
// RUN:     -ffp-exception-behavior=strict \
// RUN:     -fexperimental-strict-floating-point \
// RUN:     -ffreestanding -disable-O0-optnone -emit-llvm %s -o - | \
// RUN:     opt -S -passes=mem2reg,sroa | llc -o=- - | FileCheck -check-prefixes=COMMON,CHECK-ASM32 %s
// RUN: %clang_cc1 -triple arm64-linux-gnueabihf -target-feature +neon \
// RUN:     -ffp-exception-behavior=strict \
// RUN:     -ffreestanding -disable-O0-optnone -emit-llvm %s -o - | \
// RUN:     opt -S -passes=mem2reg,sroa | llc -o=- - | FileCheck -check-prefixes=COMMON,CHECK-ASM64 %s

// REQUIRES: arm-registered-target,aarch64-registered-target

#include <arm_neon.h>

// COMMON-LABEL: test_vrndi_f32
// UNCONSTRAINED: [[VRNDI1_I:%.*]] = call <2 x float> @llvm.nearbyint.v2f32(<2 x float> [[VRNDI_I:%.*]])
// CONSTRAINED:   [[VRNDI1_I:%.*]] = call <2 x float> @llvm.experimental.constrained.nearbyint.v2f32(<2 x float> [[VRNDI_I:%.*]], metadata !"round.tonearest", metadata !"fpexcept.strict")
// CHECK-ASM32:   vrintr.f32 s{{[0-9]+}}, s{{[0-9]+}}
// CHECK-ASM32:   vrintr.f32 s{{[0-9]+}}, s{{[0-9]+}}
// CHECK-ASM64:   frinti v{{[0-9]+}}.2s, v{{[0-9]+}}.2s
// COMMONIR:      ret <2 x float> [[VRNDI1_I]]
float32x2_t test_vrndi_f32(float32x2_t a) {
  return vrndi_f32(a);
}

// COMMON-LABEL: test_vrndiq_f32
// UNCONSTRAINED: [[VRNDI1_I:%.*]] = call <4 x float> @llvm.nearbyint.v4f32(<4 x float> [[VRNDI_I:%.*]])
// CONSTRAINED:   [[VRNDI1_I:%.*]] = call <4 x float> @llvm.experimental.constrained.nearbyint.v4f32(<4 x float> [[VRNDI_I:%.*]], metadata !"round.tonearest", metadata !"fpexcept.strict")
// CHECK-ASM32:   vrintr.f32 s{{[0-9]+}}, s{{[0-9]+}}
// CHECK-ASM32:   vrintr.f32 s{{[0-9]+}}, s{{[0-9]+}}
// CHECK-ASM32:   vrintr.f32 s{{[0-9]+}}, s{{[0-9]+}}
// CHECK-ASM32:   vrintr.f32 s{{[0-9]+}}, s{{[0-9]+}}
// CHECK-ASM64:   frinti v{{[0-9]+}}.4s, v{{[0-9]+}}.4s
// COMMONIR:      ret <4 x float> [[VRNDI1_I]]
float32x4_t test_vrndiq_f32(float32x4_t a) {
  return vrndiq_f32(a);
}