-; RUN: llc < %s -march=x86-64 -asm-verbose=false -mtriple=x86_64-unknown-linux-gnu -post-RA-scheduler=true | FileCheck %s
+; RUN: llc < %s -march=x86-64 -asm-verbose=false -mtriple=x86_64-unknown-linux-gnu -mcpu=nehalem -post-RA-scheduler=true -schedmodel=false | FileCheck %s
; Currently, floating-point selects are lowered to CFG triangles.
; This means that one side of the select is always unconditionally
; evaluated, however with MachineSink we can sink the other side so
; that it's conditionally evaluated.
-; CHECK: foo:
-; CHECK: divsd
+; CHECK-LABEL: foo:
; CHECK-NEXT: testb $1, %dil
; CHECK-NEXT: jne
; CHECK-NEXT: divsd
+; CHECK-NEXT: movaps
+; CHECK-NEXT: ret
+; CHECK: divsd
define double @foo(double %x, double %y, i1 %c) nounwind {
%a = fdiv double %x, 3.2
ret double %z
}
+; Make sure the critical edge is broken so the divsd is sunken below
+; the conditional branch.
+; rdar://8454886
+
+; CHECK-LABEL: split:
+; CHECK-NEXT: testb $1, %dil
+; CHECK-NEXT: je
+; CHECK: divsd
+; CHECK: movaps
+; CHECK: ret
+define double @split(double %x, double %y, i1 %c) nounwind {
+ %a = fdiv double %x, 3.2
+ %z = select i1 %c, double %a, double %y
+ ret double %z
+}
+
+
; Hoist floating-point constant-pool loads out of loops.
-; CHECK: bar:
+; CHECK-LABEL: bar:
; CHECK: movsd
; CHECK: align
define void @bar(double* nocapture %p, i64 %n) nounwind {
; Sink instructions with dead EFLAGS defs.
-; CHECK: zzz:
-; CHECK: je
-; CHECK-NEXT: orb
+; FIXME: Unfail the zzz test if we can correctly mark pregs with the kill flag.
+;
+; See <rdar://problem/8030636>. This test isn't valid after we made machine
+; sinking more conservative about sinking instructions that define a preg into a
+; block when we don't know if the preg is killed within the current block.
-define zeroext i8 @zzz(i8 zeroext %a, i8 zeroext %b) nounwind readnone {
-entry:
- %tmp = zext i8 %a to i32 ; <i32> [#uses=1]
- %tmp2 = icmp eq i8 %a, 0 ; <i1> [#uses=1]
- %tmp3 = or i8 %b, -128 ; <i8> [#uses=1]
- %tmp4 = and i8 %b, 127 ; <i8> [#uses=1]
- %b_addr.0 = select i1 %tmp2, i8 %tmp4, i8 %tmp3 ; <i8> [#uses=1]
- ret i8 %b_addr.0
-}
+
+; FIXMEHECK: zzz:
+; FIXMEHECK: je
+; FIXMEHECK-NEXT: orb
+
+; define zeroext i8 @zzz(i8 zeroext %a, i8 zeroext %b) nounwind readnone {
+; entry:
+; %tmp = zext i8 %a to i32 ; <i32> [#uses=1]
+; %tmp2 = icmp eq i8 %a, 0 ; <i1> [#uses=1]
+; %tmp3 = or i8 %b, -128 ; <i8> [#uses=1]
+; %tmp4 = and i8 %b, 127 ; <i8> [#uses=1]
+; %b_addr.0 = select i1 %tmp2, i8 %tmp4, i8 %tmp3 ; <i8> [#uses=1]
+; ret i8 %b_addr.0
+; }
; Codegen should hoist and CSE these constants.
-; CHECK: vv:
-; CHECK: LCPI4_0(%rip), %xmm0
-; CHECK: LCPI4_1(%rip), %xmm1
-; CHECK: LCPI4_2(%rip), %xmm2
+; CHECK-LABEL: vv:
+; CHECK: LCPI3_0(%rip), %xmm0
+; CHECK: LCPI3_1(%rip), %xmm1
+; CHECK: LCPI3_2(%rip), %xmm2
; CHECK: align
; CHECK-NOT: LCPI
; CHECK: ret
br label %bb60
bb: ; preds = %bb60
+ %i.0 = phi i32 [ 0, %bb60 ] ; <i32> [#uses=2]
%0 = bitcast float* %x_addr.0 to <4 x float>* ; <<4 x float>*> [#uses=1]
%1 = load <4 x float>* %0, align 16 ; <<4 x float>> [#uses=4]
%tmp20 = bitcast <4 x float> %1 to <4 x i32> ; <<4 x i32>> [#uses=1]
%5 = getelementptr float* %x_addr.0, i64 4 ; <float*> [#uses=1]
%6 = getelementptr float* %y_addr.0, i64 4 ; <float*> [#uses=1]
%7 = add i32 %i.0, 4 ; <i32> [#uses=1]
- br label %bb60
+ %8 = load i32* %n, align 4 ; <i32> [#uses=1]
+ %9 = icmp sgt i32 %8, %7 ; <i1> [#uses=1]
+ br i1 %9, label %bb60, label %return
bb60: ; preds = %bb, %entry
- %i.0 = phi i32 [ 0, %entry ], [ %7, %bb ] ; <i32> [#uses=2]
%x_addr.0 = phi float* [ %x, %entry ], [ %5, %bb ] ; <float*> [#uses=2]
%y_addr.0 = phi float* [ %y, %entry ], [ %6, %bb ] ; <float*> [#uses=2]
- %8 = load i32* %n, align 4 ; <i32> [#uses=1]
- %9 = icmp sgt i32 %8, %i.0 ; <i1> [#uses=1]
- br i1 %9, label %bb, label %return
+ br label %bb
return: ; preds = %bb60
ret void
; CodeGen should use the correct register class when extracting
; a load from a zero-extending load for hoisting.
-; CHECK: default_get_pch_validity:
+; CHECK-LABEL: default_get_pch_validity:
; CHECK: movl cl_options_count(%rip), %ecx
@cl_options_count = external constant i32 ; <i32*> [#uses=2]