
Previously it was only being done if shouldAlignPointerArgs() returned true, which right now is only true for ARM targets. Updating the argument alignment attributes of memcpy/memset intrinsics if the underlying object has larger alignment can be beneficial even when CGP didn't increase alignment (as can be seen from the test changes), so invert the loop and if condition. Differential Revision: https://reviews.llvm.org/D134281
192 lines
5.2 KiB
LLVM
192 lines
5.2 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc < %s -mtriple=i686-pc-elfiamcu | FileCheck %s
|
|
|
|
%struct.st12_t = type { i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32 }
|
|
|
|
define i32 @test_ints(i32 %a, i32 %b, i32 %c, i32 %d) #0 {
|
|
; CHECK-LABEL: test_ints:
|
|
; CHECK: # %bb.0: # %entry
|
|
; CHECK-NEXT: addl %edx, %eax
|
|
; CHECK-NEXT: imull %ecx, %eax
|
|
; CHECK-NEXT: addl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: retl
|
|
entry:
|
|
%r1 = add i32 %b, %a
|
|
%r2 = mul i32 %c, %r1
|
|
%r3 = add i32 %d, %r2
|
|
ret i32 %r3
|
|
}
|
|
|
|
define i32 @test_floats(i32 %a, i32 %b, float %c, float %d) #0 {
|
|
; CHECK-LABEL: test_floats:
|
|
; CHECK: # %bb.0: # %entry
|
|
; CHECK-NEXT: addl %edx, %eax
|
|
; CHECK-NEXT: imull %ecx, %eax
|
|
; CHECK-NEXT: addl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: retl
|
|
entry:
|
|
%ci = bitcast float %c to i32
|
|
%di = bitcast float %d to i32
|
|
%r1 = add i32 %b, %a
|
|
%r2 = mul i32 %ci, %r1
|
|
%r3 = add i32 %di, %r2
|
|
ret i32 %r3
|
|
}
|
|
|
|
define double @test_doubles(double %d1, double %d2) #0 {
|
|
; CHECK-LABEL: test_doubles:
|
|
; CHECK: # %bb.0: # %entry
|
|
; CHECK-NEXT: addl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: adcl {{[0-9]+}}(%esp), %edx
|
|
; CHECK-NEXT: retl
|
|
entry:
|
|
%d1i = bitcast double %d1 to i64
|
|
%d2i = bitcast double %d2 to i64
|
|
%r = add i64 %d1i, %d2i
|
|
%rd = bitcast i64 %r to double
|
|
ret double %rd
|
|
}
|
|
|
|
define double @test_mixed_doubles(double %d2, i32 %i) #0 {
|
|
; CHECK-LABEL: test_mixed_doubles:
|
|
; CHECK: # %bb.0: # %entry
|
|
; CHECK-NEXT: addl %ecx, %eax
|
|
; CHECK-NEXT: adcl $0, %edx
|
|
; CHECK-NEXT: retl
|
|
entry:
|
|
%iext = zext i32 %i to i64
|
|
%d2i = bitcast double %d2 to i64
|
|
%r = add i64 %iext, %d2i
|
|
%rd = bitcast i64 %r to double
|
|
ret double %rd
|
|
}
|
|
|
|
define void @ret_large_struct(ptr noalias nocapture sret(%struct.st12_t) %agg.result, ptr byval(%struct.st12_t) nocapture readonly align 4 %r) #0 {
|
|
; CHECK-LABEL: ret_large_struct:
|
|
; CHECK: # %bb.0: # %entry
|
|
; CHECK-NEXT: pushl %edi
|
|
; CHECK-NEXT: pushl %esi
|
|
; CHECK-NEXT: leal {{[0-9]+}}(%esp), %esi
|
|
; CHECK-NEXT: movl $12, %ecx
|
|
; CHECK-NEXT: movl %eax, %edi
|
|
; CHECK-NEXT: rep;movsl (%esi), %es:(%edi)
|
|
; CHECK-NEXT: popl %esi
|
|
; CHECK-NEXT: popl %edi
|
|
; CHECK-NEXT: retl
|
|
entry:
|
|
call void @llvm.memcpy.p0.p0.i32(ptr %agg.result, ptr %r, i32 48, i1 false)
|
|
ret void
|
|
}
|
|
|
|
define i32 @var_args(i32 %i1, ...) #0 {
|
|
; CHECK-LABEL: var_args:
|
|
; CHECK: # %bb.0: # %entry
|
|
; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: retl
|
|
entry:
|
|
ret i32 %i1
|
|
}
|
|
|
|
%struct.S = type { i8 }
|
|
|
|
define i32 @test_lib_args(float %a, float %b) #0 {
|
|
; CHECK-LABEL: test_lib_args:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: movl %edx, %eax
|
|
; CHECK-NEXT: calll __fixsfsi
|
|
; CHECK-NEXT: retl
|
|
%ret = fptosi float %b to i32
|
|
ret i32 %ret
|
|
}
|
|
|
|
define i32 @test_fp128(ptr %ptr) #0 {
|
|
; CHECK-LABEL: test_fp128:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: pushl 12(%eax)
|
|
; CHECK-NEXT: pushl 8(%eax)
|
|
; CHECK-NEXT: pushl 4(%eax)
|
|
; CHECK-NEXT: pushl (%eax)
|
|
; CHECK-NEXT: calll __fixtfsi
|
|
; CHECK-NEXT: addl $16, %esp
|
|
; CHECK-NEXT: retl
|
|
%v = load fp128, ptr %ptr
|
|
%ret = fptosi fp128 %v to i32
|
|
ret i32 %ret
|
|
}
|
|
|
|
declare void @llvm.memcpy.p0.p0.i32(ptr nocapture, ptr nocapture readonly, i32, i1) #1
|
|
|
|
define void @test_alignment_d() #0 {
|
|
; CHECK-LABEL: test_alignment_d:
|
|
; CHECK: # %bb.0: # %entry
|
|
; CHECK-NEXT: subl $8, %esp
|
|
; CHECK-NEXT: movl $1073741824, {{[0-9]+}}(%esp) # imm = 0x40000000
|
|
; CHECK-NEXT: movl $0, (%esp)
|
|
; CHECK-NEXT: movl %esp, %eax
|
|
; CHECK-NEXT: calll food@PLT
|
|
; CHECK-NEXT: addl $8, %esp
|
|
; CHECK-NEXT: retl
|
|
entry:
|
|
%d = alloca double
|
|
store double 2.000000e+00, ptr %d
|
|
call void @food(ptr inreg %d)
|
|
ret void
|
|
}
|
|
|
|
define void @test_alignment_i() #0 {
|
|
; CHECK-LABEL: test_alignment_i:
|
|
; CHECK: # %bb.0: # %entry
|
|
; CHECK-NEXT: subl $8, %esp
|
|
; CHECK-NEXT: movl $0, {{[0-9]+}}(%esp)
|
|
; CHECK-NEXT: movl $2, (%esp)
|
|
; CHECK-NEXT: movl %esp, %eax
|
|
; CHECK-NEXT: calll fooi@PLT
|
|
; CHECK-NEXT: addl $8, %esp
|
|
; CHECK-NEXT: retl
|
|
entry:
|
|
%i = alloca i64
|
|
store i64 2, ptr %i
|
|
call void @fooi(ptr inreg %i)
|
|
ret void
|
|
}
|
|
|
|
define void @test_alignment_s() #0 {
|
|
; CHECK-LABEL: test_alignment_s:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: pushl %eax
|
|
; CHECK-NEXT: movl %esp, %eax
|
|
; CHECK-NEXT: calll foos@PLT
|
|
; CHECK-NEXT: popl %eax
|
|
; CHECK-NEXT: retl
|
|
%s = alloca %struct.S, align 4
|
|
call void @foos(ptr inreg %s)
|
|
ret void
|
|
}
|
|
|
|
define void @test_alignment_fp() #0 {
|
|
; CHECK-LABEL: test_alignment_fp:
|
|
; CHECK: # %bb.0: # %entry
|
|
; CHECK-NEXT: subl $16, %esp
|
|
; CHECK-NEXT: movl $1073741824, {{[0-9]+}}(%esp) # imm = 0x40000000
|
|
; CHECK-NEXT: movl $0, {{[0-9]+}}(%esp)
|
|
; CHECK-NEXT: movl $0, {{[0-9]+}}(%esp)
|
|
; CHECK-NEXT: movl $0, (%esp)
|
|
; CHECK-NEXT: movl %esp, %eax
|
|
; CHECK-NEXT: calll foofp@PLT
|
|
; CHECK-NEXT: addl $16, %esp
|
|
; CHECK-NEXT: retl
|
|
entry:
|
|
%f = alloca fp128
|
|
store fp128 0xL00000000000000004000000000000000, ptr %f
|
|
call void @foofp(ptr inreg %f)
|
|
ret void
|
|
}
|
|
|
|
declare void @food(ptr inreg)
|
|
declare void @fooi(ptr inreg)
|
|
declare void @foos(ptr inreg)
|
|
declare void @foofp(ptr inreg)
|
|
|
|
attributes #0 = { nounwind "use-soft-float"="true"}
|
|
attributes #1 = { nounwind argmemonly }
|