blob: f224a0dc21136e207e10602e4898ba14a1044ad8 [file] [log] [blame] [edit]
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 6
; RUN: llc < %s -mtriple=wasm32-unknown-unknown -mattr=+simd128,+relaxed-simd | FileCheck %s
; Test that fmaxnum and fmaximumnum get transformed to relaxed_max
target triple = "wasm32"
define <4 x float> @test_maxnum_f32x4(<4 x float> %a, <4 x float> %b) {
; CHECK-LABEL: test_maxnum_f32x4:
; CHECK: .functype test_maxnum_f32x4 (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f32x4.relaxed_max
; CHECK-NEXT: # fallthrough-return
%result = call <4 x float> @llvm.maxnum.v4f32(<4 x float> %a, <4 x float> %b)
ret <4 x float> %result
}
define <4 x float> @test_maximumnum_f32x4(<4 x float> %a, <4 x float> %b) {
; CHECK-LABEL: test_maximumnum_f32x4:
; CHECK: .functype test_maximumnum_f32x4 (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f32x4.relaxed_max
; CHECK-NEXT: # fallthrough-return
%result = call <4 x float> @llvm.maximumnum.v4f32(<4 x float> %a, <4 x float> %b)
ret <4 x float> %result
}
define <2 x double> @test_maxnum_f64x2(<2 x double> %a, <2 x double> %b) {
; CHECK-LABEL: test_maxnum_f64x2:
; CHECK: .functype test_maxnum_f64x2 (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f64x2.relaxed_max
; CHECK-NEXT: # fallthrough-return
%result = call <2 x double> @llvm.maxnum.v2f64(<2 x double> %a, <2 x double> %b)
ret <2 x double> %result
}
define <2 x double> @test_minimumnum_f64x2(<2 x double> %a, <2 x double> %b) {
; CHECK-LABEL: test_minimumnum_f64x2:
; CHECK: .functype test_minimumnum_f64x2 (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f64x2.relaxed_max
; CHECK-NEXT: # fallthrough-return
%result = call <2 x double> @llvm.maximumnum.v2f64(<2 x double> %a, <2 x double> %b)
ret <2 x double> %result
}
define <4 x float> @test_pmax_v4f32_olt(<4 x float> %x, <4 x float> %y) {
; CHECK-LABEL: test_pmax_v4f32_olt:
; CHECK: .functype test_pmax_v4f32_olt (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f32x4.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp olt <4 x float> %x, %y
%a = select <4 x i1> %c, <4 x float> %y, <4 x float> %x
ret <4 x float> %a
}
define <4 x float> @test_pmax_v4f32_ole(<4 x float> %x, <4 x float> %y) {
; CHECK-LABEL: test_pmax_v4f32_ole:
; CHECK: .functype test_pmax_v4f32_ole (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f32x4.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp ole <4 x float> %x, %y
%a = select <4 x i1> %c, <4 x float> %y, <4 x float> %x
ret <4 x float> %a
}
define <4 x float> @test_pmax_v4f32_ogt(<4 x float> %x, <4 x float> %y) {
; CHECK-LABEL: test_pmax_v4f32_ogt:
; CHECK: .functype test_pmax_v4f32_ogt (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f32x4.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp ogt <4 x float> %y, %x
%a = select <4 x i1> %c, <4 x float> %y, <4 x float> %x
ret <4 x float> %a
}
define <4 x float> @test_pmax_v4f32_oge(<4 x float> %x, <4 x float> %y) {
; CHECK-LABEL: test_pmax_v4f32_oge:
; CHECK: .functype test_pmax_v4f32_oge (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f32x4.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp oge <4 x float> %y, %x
%a = select <4 x i1> %c, <4 x float> %y, <4 x float> %x
ret <4 x float> %a
}
; For setlt
define <4 x float> @pmax_v4f32_fast_olt(<4 x float> %x, <4 x float> %y) {
; CHECK-LABEL: pmax_v4f32_fast_olt:
; CHECK: .functype pmax_v4f32_fast_olt (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f32x4.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp fast olt <4 x float> %x, %y
%a = select <4 x i1> %c, <4 x float> %y, <4 x float> %x
ret <4 x float> %a
}
; For setle
define <4 x float> @test_pmax_v4f32_fast_ole(<4 x float> %x, <4 x float> %y) {
; CHECK-LABEL: test_pmax_v4f32_fast_ole:
; CHECK: .functype test_pmax_v4f32_fast_ole (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f32x4.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp fast ole <4 x float> %x, %y
%a = select <4 x i1> %c, <4 x float> %y, <4 x float> %x
ret <4 x float> %a
}
; For setgt
define <4 x float> @test_pmax_v4f32_fast_ogt(<4 x float> %x, <4 x float> %y) {
; CHECK-LABEL: test_pmax_v4f32_fast_ogt:
; CHECK: .functype test_pmax_v4f32_fast_ogt (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f32x4.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp fast ogt <4 x float> %x, %y
%a = select <4 x i1> %c, <4 x float> %x, <4 x float> %y
ret <4 x float> %a
}
; For setge
define <4 x float> @test_pmax_v4f32_fast_oge(<4 x float> %x, <4 x float> %y) {
; CHECK-LABEL: test_pmax_v4f32_fast_oge:
; CHECK: .functype test_pmax_v4f32_fast_oge (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f32x4.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp fast oge <4 x float> %x, %y
%a = select <4 x i1> %c, <4 x float> %x, <4 x float> %y
ret <4 x float> %a
}
define <4 x i32> @test_pmax_int_v4f32(<4 x i32> %x, <4 x i32> %y) {
; CHECK-LABEL: test_pmax_int_v4f32:
; CHECK: .functype test_pmax_int_v4f32 (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 1
; CHECK-NEXT: local.get 0
; CHECK-NEXT: f32x4.relaxed_max
; CHECK-NEXT: # fallthrough-return
%fx = bitcast <4 x i32> %x to <4 x float>
%fy = bitcast <4 x i32> %y to <4 x float>
%c = fcmp olt <4 x float> %fy, %fx
%a = select <4 x i1> %c, <4 x i32> %x, <4 x i32> %y
ret <4 x i32> %a
}
define <2 x double> @test_pmax_v2f64_olt(<2 x double> %x, <2 x double> %y) {
; CHECK-LABEL: test_pmax_v2f64_olt:
; CHECK: .functype test_pmax_v2f64_olt (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f64x2.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp olt <2 x double> %x, %y
%a = select <2 x i1> %c, <2 x double> %y, <2 x double> %x
ret <2 x double> %a
}
define <2 x double> @test_pmax_v2f64_ole(<2 x double> %x, <2 x double> %y) {
; CHECK-LABEL: test_pmax_v2f64_ole:
; CHECK: .functype test_pmax_v2f64_ole (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f64x2.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp ole <2 x double> %x, %y
%a = select <2 x i1> %c, <2 x double> %y, <2 x double> %x
ret <2 x double> %a
}
define <2 x double> @test_pmax_v2f64_ogt(<2 x double> %x, <2 x double> %y) {
; CHECK-LABEL: test_pmax_v2f64_ogt:
; CHECK: .functype test_pmax_v2f64_ogt (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 1
; CHECK-NEXT: local.get 0
; CHECK-NEXT: f64x2.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp ogt <2 x double> %x, %y
%a = select <2 x i1> %c, <2 x double> %x, <2 x double> %y
ret <2 x double> %a
}
define <2 x double> @test_pmax_v2f64_oge(<2 x double> %x, <2 x double> %y) {
; CHECK-LABEL: test_pmax_v2f64_oge:
; CHECK: .functype test_pmax_v2f64_oge (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 1
; CHECK-NEXT: local.get 0
; CHECK-NEXT: f64x2.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp oge <2 x double> %x, %y
%a = select <2 x i1> %c, <2 x double> %x, <2 x double> %y
ret <2 x double> %a
}
; For setlt
define <2 x double> @pmax_v2f64_fast_olt(<2 x double> %x, <2 x double> %y) {
; CHECK-LABEL: pmax_v2f64_fast_olt:
; CHECK: .functype pmax_v2f64_fast_olt (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f64x2.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp fast olt <2 x double> %x, %y
%a = select <2 x i1> %c, <2 x double> %y, <2 x double> %x
ret <2 x double> %a
}
; For setle
define <2 x double> @test_pmax_v2f64_fast_ole(<2 x double> %x, <2 x double> %y) {
; CHECK-LABEL: test_pmax_v2f64_fast_ole:
; CHECK: .functype test_pmax_v2f64_fast_ole (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f64x2.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp fast ole <2 x double> %x, %y
%a = select <2 x i1> %c, <2 x double> %y, <2 x double> %x
ret <2 x double> %a
}
; For setgt
define <2 x double> @test_pmax_v2f64_fast_ogt(<2 x double> %x, <2 x double> %y) {
; CHECK-LABEL: test_pmax_v2f64_fast_ogt:
; CHECK: .functype test_pmax_v2f64_fast_ogt (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f64x2.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp fast ogt <2 x double> %x, %y
%a = select <2 x i1> %c, <2 x double> %x, <2 x double> %y
ret <2 x double> %a
}
; For setge
define <2 x double> @test_pmax_v2f64_fast_oge(<2 x double> %x, <2 x double> %y) {
; CHECK-LABEL: test_pmax_v2f64_fast_oge:
; CHECK: .functype test_pmax_v2f64_fast_oge (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 0
; CHECK-NEXT: local.get 1
; CHECK-NEXT: f64x2.relaxed_max
; CHECK-NEXT: # fallthrough-return
%c = fcmp fast oge <2 x double> %x, %y
%a = select <2 x i1> %c, <2 x double> %x, <2 x double> %y
ret <2 x double> %a
}
define <2 x i64> @test_pmax_int_v2f64(<2 x i64> %x, <2 x i64> %y) {
; CHECK-LABEL: test_pmax_int_v2f64:
; CHECK: .functype test_pmax_int_v2f64 (v128, v128) -> (v128)
; CHECK-NEXT: # %bb.0:
; CHECK-NEXT: local.get 1
; CHECK-NEXT: local.get 0
; CHECK-NEXT: f64x2.relaxed_max
; CHECK-NEXT: # fallthrough-return
%fx = bitcast <2 x i64> %x to <2 x double>
%fy = bitcast <2 x i64> %y to <2 x double>
%c = fcmp olt <2 x double> %fy, %fx
%a = select <2 x i1> %c, <2 x i64> %x, <2 x i64> %y
ret <2 x i64> %a
}
declare <4 x float> @llvm.maxnum.v4f32(<4 x float>, <4 x float>)
declare <4 x float> @llvm.maximumnum.v4f32(<4 x float>, <4 x float>)
declare <2 x double> @llvm.maxnum.v2f64(<2 x double>, <2 x double>)
declare <2 x double> @llvm.maximumnum.v2f64(<2 x double>, <2 x double>)