Turn on post-alloc scheduling for x86.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@84431 91177308-0d34-0410-b5e6-96231b3b80d8

14 files changed, 39 insertions, 52 deletions

diff --git a/lib/Target/X86/X86Subtarget.h b/lib/Target/X86/X86Subtarget.h
index aa9eb1d189..27e43333f8 100644
--- a/lib/Target/X86/X86Subtarget.h
+++ b/lib/Target/X86/X86Subtarget.h
@@ -219,8 +219,7 @@ public:
   /// enablePostRAScheduler - X86 target is enabling post-alloc scheduling
   /// at 'More' optimization level.
   bool enablePostRAScheduler(CodeGenOpt::Level OptLevel) const {
-    // FIXME: This causes llvm to miscompile itself on i386. :-(
-    return false /*OptLevel >= CodeGenOpt::Default*/;
+    return OptLevel >= CodeGenOpt::Default;
   }
 };
 
diff --git a/test/CodeGen/X86/2007-01-08-InstrSched.ll b/test/CodeGen/X86/2007-01-08-InstrSched.ll
index e1bae3251a..81f0a1d724 100644
--- a/test/CodeGen/X86/2007-01-08-InstrSched.ll
+++ b/test/CodeGen/X86/2007-01-08-InstrSched.ll
@@ -11,9 +11,12 @@ define float @foo(float %x) nounwind {
     %tmp14 = fadd float %tmp12, %tmp7
     ret float %tmp14
 
-; CHECK:      mulss	LCPI1_2(%rip)
+; CHECK:      mulss	LCPI1_3(%rip)
+; CHECK-NEXT: mulss	LCPI1_0(%rip)
+; CHECK-NEXT: mulss	LCPI1_1(%rip)
+; CHECK-NEXT: mulss	LCPI1_2(%rip)
+; CHECK-NEXT: addss
 ; CHECK-NEXT: addss
-; CHECK-NEXT: mulss	LCPI1_3(%rip)
 ; CHECK-NEXT: addss
 ; CHECK-NEXT: ret
 }
diff --git a/test/CodeGen/X86/2008-07-11-SpillerBug.ll b/test/CodeGen/X86/2008-07-11-SpillerBug.ll
index f75e605168..88a5fde07e 100644
--- a/test/CodeGen/X86/2008-07-11-SpillerBug.ll
+++ b/test/CodeGen/X86/2008-07-11-SpillerBug.ll
@@ -1,4 +1,4 @@
-; RUN: llc < %s -march=x86 -relocation-model=static -disable-fp-elim | FileCheck %s
+; RUN: llc < %s -march=x86 -relocation-model=static -disable-fp-elim -post-RA-scheduler=false | FileCheck %s
 ; PR2536
 
 
diff --git a/test/CodeGen/X86/2009-04-21-NoReloadImpDef.ll b/test/CodeGen/X86/2009-04-21-NoReloadImpDef.ll
index c6e6e50641..5bd956a016 100644
--- a/test/CodeGen/X86/2009-04-21-NoReloadImpDef.ll
+++ b/test/CodeGen/X86/2009-04-21-NoReloadImpDef.ll
@@ -1,5 +1,5 @@
 ; RUN: llc -mtriple=i386-apple-darwin10.0 -relocation-model=pic \
-; RUN:     -disable-fp-elim -mattr=-sse41,-sse3,+sse2 < %s | \
+; RUN:     -disable-fp-elim -mattr=-sse41,-sse3,+sse2 -post-RA-scheduler=false < %s | \
 ; RUN:   FileCheck %s
 ; rdar://6808032
 
diff --git a/test/CodeGen/X86/2009-09-19-SchedCustomLoweringBug.ll b/test/CodeGen/X86/2009-09-19-SchedCustomLoweringBug.ll
index 646806e5db..f3cf1d5e70 100644
--- a/test/CodeGen/X86/2009-09-19-SchedCustomLoweringBug.ll
+++ b/test/CodeGen/X86/2009-09-19-SchedCustomLoweringBug.ll
@@ -9,9 +9,7 @@ entry:
   br label %bb
 
 bb:                                               ; preds = %bb1, %entry
-; CHECK:      movl %e
-; CHECK-NEXT: addl $1
-; CHECK-NEXT: movl %e
+; CHECK:      addl $1
 ; CHECK-NEXT: adcl $0
   %i.0 = phi i64 [ 0, %entry ], [ %0, %bb1 ]      ; <i64> [#uses=1]
   %0 = add nsw i64 %i.0, 1                        ; <i64> [#uses=2]
diff --git a/test/CodeGen/X86/abi-isel.ll b/test/CodeGen/X86/abi-isel.ll
index a6fd2d8fe1..6d7b2d4343 100644
--- a/test/CodeGen/X86/abi-isel.ll
+++ b/test/CodeGen/X86/abi-isel.ll
@@ -1,16 +1,16 @@
-; RUN: llc < %s -asm-verbose=0 -mtriple=i686-unknown-linux-gnu -march=x86 -relocation-model=static -code-model=small | FileCheck %s -check-prefix=LINUX-32-STATIC
-; RUN: llc < %s -asm-verbose=0 -mtriple=i686-unknown-linux-gnu -march=x86 -relocation-model=static -code-model=small | FileCheck %s -check-prefix=LINUX-32-PIC
+; RUN: llc < %s -asm-verbose=0 -mtriple=i686-unknown-linux-gnu -march=x86 -relocation-model=static -code-model=small -post-RA-scheduler=false | FileCheck %s -check-prefix=LINUX-32-STATIC
+; RUN: llc < %s -asm-verbose=0 -mtriple=i686-unknown-linux-gnu -march=x86 -relocation-model=static -code-model=small -post-RA-scheduler=false | FileCheck %s -check-prefix=LINUX-32-PIC
 
-; RUN: llc < %s -asm-verbose=0 -mtriple=x86_64-unknown-linux-gnu -march=x86-64 -relocation-model=static -code-model=small | FileCheck %s -check-prefix=LINUX-64-STATIC
-; RUN: llc < %s -asm-verbose=0 -mtriple=x86_64-unknown-linux-gnu -march=x86-64 -relocation-model=pic -code-model=small | FileCheck %s -check-prefix=LINUX-64-PIC
+; RUN: llc < %s -asm-verbose=0 -mtriple=x86_64-unknown-linux-gnu -march=x86-64 -relocation-model=static -code-model=small -post-RA-scheduler=false | FileCheck %s -check-prefix=LINUX-64-STATIC
+; RUN: llc < %s -asm-verbose=0 -mtriple=x86_64-unknown-linux-gnu -march=x86-64 -relocation-model=pic -code-model=small -post-RA-scheduler=false | FileCheck %s -check-prefix=LINUX-64-PIC
 
-; RUN: llc < %s -asm-verbose=0 -mtriple=i686-apple-darwin -march=x86 -relocation-model=static -code-model=small | FileCheck %s -check-prefix=DARWIN-32-STATIC
-; RUN: llc < %s -asm-verbose=0 -mtriple=i686-apple-darwin -march=x86 -relocation-model=dynamic-no-pic -code-model=small | FileCheck %s -check-prefix=DARWIN-32-DYNAMIC
-; RUN: llc < %s -asm-verbose=0 -mtriple=i686-apple-darwin -march=x86 -relocation-model=pic -code-model=small | FileCheck %s -check-prefix=DARWIN-32-PIC
+; RUN: llc < %s -asm-verbose=0 -mtriple=i686-apple-darwin -march=x86 -relocation-model=static -code-model=small -post-RA-scheduler=false | FileCheck %s -check-prefix=DARWIN-32-STATIC
+; RUN: llc < %s -asm-verbose=0 -mtriple=i686-apple-darwin -march=x86 -relocation-model=dynamic-no-pic -code-model=small -post-RA-scheduler=false | FileCheck %s -check-prefix=DARWIN-32-DYNAMIC
+; RUN: llc < %s -asm-verbose=0 -mtriple=i686-apple-darwin -march=x86 -relocation-model=pic -code-model=small -post-RA-scheduler=false | FileCheck %s -check-prefix=DARWIN-32-PIC
 
-; RUN: llc < %s -asm-verbose=0 -mtriple=x86_64-apple-darwin -march=x86-64 -relocation-model=static -code-model=small | FileCheck %s -check-prefix=DARWIN-64-STATIC
-; RUN: llc < %s -asm-verbose=0 -mtriple=x86_64-apple-darwin -march=x86-64 -relocation-model=dynamic-no-pic -code-model=small | FileCheck %s -check-prefix=DARWIN-64-DYNAMIC
-; RUN: llc < %s -asm-verbose=0 -mtriple=x86_64-apple-darwin -march=x86-64 -relocation-model=pic -code-model=small | FileCheck %s -check-prefix=DARWIN-64-PIC
+; RUN: llc < %s -asm-verbose=0 -mtriple=x86_64-apple-darwin -march=x86-64 -relocation-model=static -code-model=small -post-RA-scheduler=false | FileCheck %s -check-prefix=DARWIN-64-STATIC
+; RUN: llc < %s -asm-verbose=0 -mtriple=x86_64-apple-darwin -march=x86-64 -relocation-model=dynamic-no-pic -code-model=small -post-RA-scheduler=false | FileCheck %s -check-prefix=DARWIN-64-DYNAMIC
+; RUN: llc < %s -asm-verbose=0 -mtriple=x86_64-apple-darwin -march=x86-64 -relocation-model=pic -code-model=small -post-RA-scheduler=false | FileCheck %s -check-prefix=DARWIN-64-PIC
 
 @src = external global [131072 x i32]
 @dst = external global [131072 x i32]
diff --git a/test/CodeGen/X86/fastcc.ll b/test/CodeGen/X86/fastcc.ll
index d538264c6d..705ab7bada 100644
--- a/test/CodeGen/X86/fastcc.ll
+++ b/test/CodeGen/X86/fastcc.ll
@@ -1,4 +1,4 @@
-; RUN: llc < %s -mtriple=i686-apple-darwin -mattr=+sse2 | FileCheck %s
+; RUN: llc < %s -mtriple=i686-apple-darwin -mattr=+sse2 -post-RA-scheduler=false | FileCheck %s
 ; CHECK: movsd %xmm0, 8(%esp)
 ; CHECK: xorl %ecx, %ecx
 
diff --git a/test/CodeGen/X86/peep-test-3.ll b/test/CodeGen/X86/peep-test-3.ll
index 13a69edea5..5aaf81b4fd 100644
--- a/test/CodeGen/X86/peep-test-3.ll
+++ b/test/CodeGen/X86/peep-test-3.ll
@@ -1,4 +1,4 @@
-; RUN: llc < %s -march=x86 | FileCheck %s
+; RUN: llc < %s -march=x86 -post-RA-scheduler=false | FileCheck %s
 ; rdar://7226797
 
 ; LLVM should omit the testl and use the flags result from the orl.
diff --git a/test/CodeGen/X86/pic.ll b/test/CodeGen/X86/pic.ll
index 3a547f95f8..e886ba06b7 100644
--- a/test/CodeGen/X86/pic.ll
+++ b/test/CodeGen/X86/pic.ll
@@ -1,4 +1,4 @@
-; RUN: llc < %s -mtriple=i686-pc-linux-gnu -relocation-model=pic -asm-verbose=false | FileCheck %s -check-prefix=LINUX
+; RUN: llc < %s -mtriple=i686-pc-linux-gnu -relocation-model=pic -asm-verbose=false -post-RA-scheduler=false | FileCheck %s -check-prefix=LINUX
 
 @ptr = external global i32* 
 @dst = external global i32 
diff --git a/test/CodeGen/X86/sink-hoist.ll b/test/CodeGen/X86/sink-hoist.ll
index 0f4e63f9c6..4042a095c6 100644
--- a/test/CodeGen/X86/sink-hoist.ll
+++ b/test/CodeGen/X86/sink-hoist.ll
@@ -6,10 +6,10 @@
 ; that it's conditionally evaluated.
 
 ; CHECK: foo:
-; CHECK-NEXT: divsd
-; CHECK:      testb $1, %dil
-; CHECK-NEXT: jne
 ; CHECK:      divsd
+; CHECK-NEXT: testb $1, %dil
+; CHECK-NEXT: jne
+; CHECK-NEXT: divsd
 
 define double @foo(double %x, double %y, i1 %c) nounwind {
   %a = fdiv double %x, 3.2
diff --git a/test/CodeGen/X86/sse2.ll b/test/CodeGen/X86/sse2.ll
index 9f926f2bee..58fe28b09f 100644
--- a/test/CodeGen/X86/sse2.ll
+++ b/test/CodeGen/X86/sse2.ll
@@ -10,10 +10,10 @@ define void @t1(<2 x double>* %r, <2 x double>* %A, double %B) nounwind  {
         
 ; CHECK: t1:
 ; CHECK: 	movl	8(%esp), %eax
+; CHECK-NEXT: 	movl	4(%esp), %ecx
 ; CHECK-NEXT: 	movapd	(%eax), %xmm0
 ; CHECK-NEXT: 	movlpd	12(%esp), %xmm0
-; CHECK-NEXT: 	movl	4(%esp), %eax
-; CHECK-NEXT: 	movapd	%xmm0, (%eax)
+; CHECK-NEXT: 	movapd	%xmm0, (%ecx)
 ; CHECK-NEXT: 	ret
 }
 
@@ -26,9 +26,9 @@ define void @t2(<2 x double>* %r, <2 x double>* %A, double %B) nounwind  {
         
 ; CHECK: t2:
 ; CHECK: 	movl	8(%esp), %eax
+; CHECK-NEXT: 	movl	4(%esp), %ecx
 ; CHECK-NEXT: 	movapd	(%eax), %xmm0
 ; CHECK-NEXT: 	movhpd	12(%esp), %xmm0
-; CHECK-NEXT: 	movl	4(%esp), %eax
-; CHECK-NEXT: 	movapd	%xmm0, (%eax)
+; CHECK-NEXT: 	movapd	%xmm0, (%ecx)
 ; CHECK-NEXT: 	ret
 }
diff --git a/test/CodeGen/X86/sse3.ll b/test/CodeGen/X86/sse3.ll
index 703635c0f5..6319cb887a 100644
--- a/test/CodeGen/X86/sse3.ll
+++ b/test/CodeGen/X86/sse3.ll
@@ -17,8 +17,8 @@ entry:
         
 ; X64: t0:
 ; X64: 	movddup	(%rsi), %xmm0
-; X64:  pshuflw	$0, %xmm0, %xmm0
 ; X64:	xorl	%eax, %eax
+; X64:  pshuflw	$0, %xmm0, %xmm0
 ; X64:	pinsrw	$0, %eax, %xmm0
 ; X64:	movaps	%xmm0, (%rdi)
 ; X64:	ret
@@ -167,18 +167,12 @@ define internal void @t10() nounwind {
         store <4 x i16> %6, <4 x i16>* @g2, align 8
         ret void
 ; X64: 	t10:
-; X64: 		movq	_g1@GOTPCREL(%rip), %rax
-; X64: 		movaps	(%rax), %xmm0
 ; X64: 		pextrw	$4, %xmm0, %eax
-; X64: 		movaps	%xmm0, %xmm1
+; X64: 		pextrw	$6, %xmm0, %edx
 ; X64: 		movlhps	%xmm1, %xmm1
 ; X64: 		pshuflw	$8, %xmm1, %xmm1
 ; X64: 		pinsrw	$2, %eax, %xmm1
-; X64: 		pextrw	$6, %xmm0, %eax
-; X64: 		pinsrw	$3, %eax, %xmm1
-; X64: 		movq	_g2@GOTPCREL(%rip), %rax
-; X64: 		movq	%xmm1, (%rax)
-; X64: 		ret
+; X64: 		pinsrw	$3, %edx, %xmm1
 }
 
 
@@ -189,8 +183,8 @@ entry:
 	ret <8 x i16> %tmp7
 
 ; X64: t11:
-; X64:	movd	%xmm1, %eax
 ; X64:	movlhps	%xmm0, %xmm0
+; X64:	movd	%xmm1, %eax
 ; X64:	pshuflw	$1, %xmm0, %xmm0
 ; X64:	pinsrw	$1, %eax, %xmm0
 ; X64:	ret
@@ -203,8 +197,8 @@ entry:
 	ret <8 x i16> %tmp9
 
 ; X64: t12:
-; X64: 	pextrw	$3, %xmm1, %eax
 ; X64: 	movlhps	%xmm0, %xmm0
+; X64: 	pextrw	$3, %xmm1, %eax
 ; X64: 	pshufhw	$3, %xmm0, %xmm0
 ; X64: 	pinsrw	$5, %eax, %xmm0
 ; X64: 	ret
@@ -256,18 +250,12 @@ entry:
         %tmp9 = shufflevector <16 x i8> %tmp8, <16 x i8> %T0,  <16 x i32> < i32 0, i32 1, i32 2, i32 17,  i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef , i32 undef >
         ret <16 x i8> %tmp9
 ; X64: 	t16:
-; X64: 		movaps	LCPI17_0(%rip), %xmm1
-; X64: 		movd	%xmm1, %eax
 ; X64: 		pinsrw	$0, %eax, %xmm1
 ; X64: 		pextrw	$8, %xmm0, %eax
 ; X64: 		pinsrw	$1, %eax, %xmm1
 ; X64: 		pextrw	$1, %xmm1, %ecx
 ; X64: 		movd	%xmm1, %edx
 ; X64: 		pinsrw	$0, %edx, %xmm1
-; X64: 		movzbl	%cl, %ecx
-; X64: 		andw	$-256, %ax
-; X64: 		orw	%cx, %ax
-; X64: 		movaps	%xmm1, %xmm0
 ; X64: 		pinsrw	$1, %eax, %xmm0
 ; X64: 		ret
 }
diff --git a/test/CodeGen/X86/tailcallstack64.ll b/test/CodeGen/X86/tailcallstack64.ll
index 73c59bb639..69018aa13d 100644
--- a/test/CodeGen/X86/tailcallstack64.ll
+++ b/test/CodeGen/X86/tailcallstack64.ll
@@ -3,19 +3,18 @@
 ; Check that lowered arguments on the stack do not overwrite each other.
 ; Add %in1 %p1 to a different temporary register (%eax).
 ; CHECK: movl  %edi, %eax
-; CHECK: addl  32(%rsp), %eax
 ; Move param %in1 to temp register (%r10d).
 ; CHECK: movl  40(%rsp), %r10d
-; Move result of addition to stack.
-; CHECK: movl  %eax, 40(%rsp)
 ; Move param %in2 to stack.
 ; CHECK: movl  %r10d, 32(%rsp)
+; Move result of addition to stack.
+; CHECK: movl  %eax, 40(%rsp)
 ; Eventually, do a TAILCALL
 ; CHECK: TAILCALL
 
-declare fastcc i32 @tailcallee(i32 %p1, i32 %p2, i32 %p3, i32 %p4, i32 %p5, i32 %p6, i32 %a, i32 %b)
+declare fastcc i32 @tailcallee(i32 %p1, i32 %p2, i32 %p3, i32 %p4, i32 %p5, i32 %p6, i32 %a, i32 %b) nounwind
 
-define fastcc i32 @tailcaller(i32 %p1, i32 %p2, i32 %p3, i32 %p4, i32 %p5, i32 %p6, i32 %in1, i32 %in2) {
+define fastcc i32 @tailcaller(i32 %p1, i32 %p2, i32 %p3, i32 %p4, i32 %p5, i32 %p6, i32 %in1, i32 %in2) nounwind {
 entry:
         %tmp = add i32 %in1, %p1
         %retval = tail call fastcc i32 @tailcallee(i32 %p1, i32 %p2, i32 %p3, i32 %p4, i32 %p5, i32 %p6, i32 %in2,i32 %tmp)
diff --git a/test/CodeGen/X86/widen_arith-3.ll b/test/CodeGen/X86/widen_arith-3.ll
index d1840671f7..a2b8b8291e 100644
--- a/test/CodeGen/X86/widen_arith-3.ll
+++ b/test/CodeGen/X86/widen_arith-3.ll
@@ -1,7 +1,7 @@
 ; RUN: llc < %s -march=x86 -mattr=+sse42 -disable-mmx | FileCheck %s
 ; CHECK: paddw
-; CHECK: movd
 ; CHECK: pextrw
+; CHECK: movd
 
 ; Widen a v3i16 to v8i16 to do a vector add