mirror of
https://github.com/c64scene-ar/llvm-6502.git
synced 2024-09-01 19:29:22 +00:00
This is a follow up to r227113. It is now required to use the amdgcn target for SI and newer GPUs. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@227316 91177308-0d34-0410-b5e6-96231b3b80d8
78 lines
2.8 KiB
LLVM
78 lines
2.8 KiB
LLVM
; RUN: llc -march=amdgcn -mcpu=SI < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s
|
|
|
|
declare i32 @llvm.r600.read.tidig.x() #1
|
|
|
|
; FUNC-LABEL: @test_fmin_legacy_f64
|
|
define void @test_fmin_legacy_f64(<4 x double> addrspace(1)* %out, <4 x double> inreg %reg0) #0 {
|
|
%r0 = extractelement <4 x double> %reg0, i32 0
|
|
%r1 = extractelement <4 x double> %reg0, i32 1
|
|
%r2 = fcmp uge double %r0, %r1
|
|
%r3 = select i1 %r2, double %r1, double %r0
|
|
%vec = insertelement <4 x double> undef, double %r3, i32 0
|
|
store <4 x double> %vec, <4 x double> addrspace(1)* %out, align 16
|
|
ret void
|
|
}
|
|
|
|
; FUNC-LABEL: @test_fmin_legacy_ule_f64
|
|
define void @test_fmin_legacy_ule_f64(double addrspace(1)* %out, double addrspace(1)* %in) #0 {
|
|
%tid = call i32 @llvm.r600.read.tidig.x() #1
|
|
%gep.0 = getelementptr double addrspace(1)* %in, i32 %tid
|
|
%gep.1 = getelementptr double addrspace(1)* %gep.0, i32 1
|
|
|
|
%a = load double addrspace(1)* %gep.0, align 8
|
|
%b = load double addrspace(1)* %gep.1, align 8
|
|
|
|
%cmp = fcmp ule double %a, %b
|
|
%val = select i1 %cmp, double %a, double %b
|
|
store double %val, double addrspace(1)* %out, align 8
|
|
ret void
|
|
}
|
|
|
|
; FUNC-LABEL: @test_fmin_legacy_ole_f64
|
|
define void @test_fmin_legacy_ole_f64(double addrspace(1)* %out, double addrspace(1)* %in) #0 {
|
|
%tid = call i32 @llvm.r600.read.tidig.x() #1
|
|
%gep.0 = getelementptr double addrspace(1)* %in, i32 %tid
|
|
%gep.1 = getelementptr double addrspace(1)* %gep.0, i32 1
|
|
|
|
%a = load double addrspace(1)* %gep.0, align 8
|
|
%b = load double addrspace(1)* %gep.1, align 8
|
|
|
|
%cmp = fcmp ole double %a, %b
|
|
%val = select i1 %cmp, double %a, double %b
|
|
store double %val, double addrspace(1)* %out, align 8
|
|
ret void
|
|
}
|
|
|
|
; FUNC-LABEL: @test_fmin_legacy_olt_f64
|
|
define void @test_fmin_legacy_olt_f64(double addrspace(1)* %out, double addrspace(1)* %in) #0 {
|
|
%tid = call i32 @llvm.r600.read.tidig.x() #1
|
|
%gep.0 = getelementptr double addrspace(1)* %in, i32 %tid
|
|
%gep.1 = getelementptr double addrspace(1)* %gep.0, i32 1
|
|
|
|
%a = load double addrspace(1)* %gep.0, align 8
|
|
%b = load double addrspace(1)* %gep.1, align 8
|
|
|
|
%cmp = fcmp olt double %a, %b
|
|
%val = select i1 %cmp, double %a, double %b
|
|
store double %val, double addrspace(1)* %out, align 8
|
|
ret void
|
|
}
|
|
|
|
; FUNC-LABEL: @test_fmin_legacy_ult_f64
|
|
define void @test_fmin_legacy_ult_f64(double addrspace(1)* %out, double addrspace(1)* %in) #0 {
|
|
%tid = call i32 @llvm.r600.read.tidig.x() #1
|
|
%gep.0 = getelementptr double addrspace(1)* %in, i32 %tid
|
|
%gep.1 = getelementptr double addrspace(1)* %gep.0, i32 1
|
|
|
|
%a = load double addrspace(1)* %gep.0, align 8
|
|
%b = load double addrspace(1)* %gep.1, align 8
|
|
|
|
%cmp = fcmp ult double %a, %b
|
|
%val = select i1 %cmp, double %a, double %b
|
|
store double %val, double addrspace(1)* %out, align 8
|
|
ret void
|
|
}
|
|
|
|
attributes #0 = { nounwind }
|
|
attributes #1 = { nounwind readnone }
|