1 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py 2 ; RUN: llc < %s -mattr=+sse2 | FileCheck %s 3 ; PR2656 4 5 target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128" 6 target triple = "i686-apple-darwin9.4.0" 7 8 %struct.anon = type <{ float, float }> 9 @.str = internal constant [17 x i8] c"pt: %.0f, %.0f\0A\00\00" ; <[17 x i8]*> [#uses=1] 10 11 ; We can not fold either stack load into an 'xor' instruction because that 12 ; would change what should be a 4-byte load into a 16-byte load. 13 ; We can fold the 16-byte constant load into either 'xor' instruction, 14 ; but we do not. It has more than one use, so it gets loaded into a register. 15 16 define void @foo(%struct.anon* byval %p) nounwind { 17 ; CHECK-LABEL: foo: 18 ; CHECK: ## %bb.0: ## %entry 19 ; CHECK-NEXT: subl $28, %esp 20 ; CHECK-NEXT: movss {{.*#+}} xmm0 = mem[0],zero,zero,zero 21 ; CHECK-NEXT: movss {{.*#+}} xmm1 = mem[0],zero,zero,zero 22 ; CHECK-NEXT: movaps {{.*#+}} xmm2 = [-0.000000e+00,-0.000000e+00,-0.000000e+00,-0.000000e+00] 23 ; CHECK-NEXT: xorps %xmm2, %xmm0 24 ; CHECK-NEXT: cvtss2sd %xmm0, %xmm0 25 ; CHECK-NEXT: xorps %xmm2, %xmm1 26 ; CHECK-NEXT: cvtss2sd %xmm1, %xmm1 27 ; CHECK-NEXT: movsd %xmm1, {{[0-9]+}}(%esp) 28 ; CHECK-NEXT: movsd %xmm0, {{[0-9]+}}(%esp) 29 ; CHECK-NEXT: movl $_.str, (%esp) 30 ; CHECK-NEXT: calll _printf 31 ; CHECK-NEXT: addl $28, %esp 32 ; CHECK-NEXT: retl 33 entry: 34 %tmp = getelementptr %struct.anon, %struct.anon* %p, i32 0, i32 0 ; <float*> [#uses=1] 35 %tmp1 = load float, float* %tmp ; <float> [#uses=1] 36 %tmp2 = getelementptr %struct.anon, %struct.anon* %p, i32 0, i32 1 ; <float*> [#uses=1] 37 %tmp3 = load float, float* %tmp2 ; <float> [#uses=1] 38 %neg = fsub float -0.000000e+00, %tmp1 ; <float> [#uses=1] 39 %conv = fpext float %neg to double ; <double> [#uses=1] 40 %neg4 = fsub float -0.000000e+00, %tmp3 ; <float> [#uses=1] 41 %conv5 = fpext float %neg4 to double ; <double> [#uses=1] 42 %call = call i32 (...) @printf( i8* getelementptr ([17 x i8], [17 x i8]* @.str, i32 0, i32 0), double %conv, double %conv5 ) ; <i32> [#uses=0] 43 ret void 44 } 45 46 declare i32 @printf(...) 47 48 ; We can not fold the load from the stack into the 'and' instruction because 49 ; that changes an 8-byte load into a 16-byte load (illegal memory access). 50 ; We can fold the load of the constant because it is a 16-byte vector constant. 51 52 define double @PR22371(double %x) { 53 ; CHECK-LABEL: PR22371: 54 ; CHECK: ## %bb.0: 55 ; CHECK-NEXT: subl $12, %esp 56 ; CHECK-NEXT: .cfi_def_cfa_offset 16 57 ; CHECK-NEXT: movsd {{.*#+}} xmm0 = mem[0],zero 58 ; CHECK-NEXT: andps LCPI1_0, %xmm0 59 ; CHECK-NEXT: movlps %xmm0, (%esp) 60 ; CHECK-NEXT: fldl (%esp) 61 ; CHECK-NEXT: addl $12, %esp 62 ; CHECK-NEXT: retl 63 %call = tail call double @fabs(double %x) #0 64 ret double %call 65 } 66 67 declare double @fabs(double) #0 68 attributes #0 = { readnone } 69
