Fix PR3149. If an early clobber def is a physical register and it is tied to an input operand, it effectively extends the live range of the physical register. Currently we do not have a good way to represent this.

172     %ECX<def> = MOV32rr %reg1039<kill>
180     INLINEASM <es:subl $5,$1
        sbbl $3,$0>, 10, %EAX<def>, 14, %ECX<earlyclobber,def>, 9, %EAX<kill>,
36, <fi#0>, 1, %reg0, 0, 9, %ECX<kill>, 36, <fi#1>, 1, %reg0, 0
188     %EAX<def> = MOV32rr %EAX<kill>
196     %ECX<def> = MOV32rr %ECX<kill>
204     %ECX<def> = MOV32rr %ECX<kill>
212     %EAX<def> = MOV32rr %EAX<kill>
220     %EAX<def> = MOV32rr %EAX
228     %reg1039<def> = MOV32rr %ECX<kill>

The early clobber operand ties ECX input to the ECX def.

The live interval of ECX is represented as this:
%reg20,inf = [46,47:1)[174,230:0)  0@174-(230) 1@46-(47)

The right way to represent this is something like
%reg20,inf = [46,47:2)[174,182:1)[181:230:0)  0@174-(182) 1@181-230 @2@46-(47)

Of course that won't work since that means overlapping live ranges defined by two val#.

The workaround for now is to add a bit to val# which says the val# is redefined by a early clobber def somewhere. This prevents the move at 228 from being optimized away by SimpleRegisterCoalescing::AdjustCopiesBackFrom.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@61259 91177308-0d34-0410-b5e6-96231b3b80d8

1 files changed, 32 insertions, 0 deletions

diff --git a/test/CodeGen/X86/2008-12-19-EarlyClobberBug.ll b/test/CodeGen/X86/2008-12-19-EarlyClobberBug.ll
new file mode 100644
index 0000000000..6442ac7c76
--- /dev/null
+++ b/test/CodeGen/X86/2008-12-19-EarlyClobberBug.ll
@@ -0,0 +1,32 @@
+; RUN: llvm-as < %s | llc -mtriple=i386-apple-darwin | %prcontext End 1 | grep {movl.*%ecx}
+; PR3149
+
+@"\01LC" = internal constant [7 x i8] c"n0=%d\0A\00"		; <[7 x i8]*> [#uses=1]
+@llvm.used = appending global [1 x i8*] [ i8* bitcast (i32 (i64, i64)* @umoddi3 to i8*) ], section "llvm.metadata"		; <[1 x i8*]*> [#uses=0]
+
+define i32 @umoddi3(i64 %u, i64 %v) nounwind noinline {
+entry:
+	%0 = trunc i64 %v to i32		; <i32> [#uses=2]
+	%1 = trunc i64 %u to i32		; <i32> [#uses=4]
+	%2 = lshr i64 %u, 32		; <i64> [#uses=1]
+	%3 = trunc i64 %2 to i32		; <i32> [#uses=2]
+	%4 = tail call i32 (i8*, ...)* @printf(i8* getelementptr ([7 x i8]* @"\01LC", i32 0, i32 0), i32 %1) nounwind		; <i32> [#uses=0]
+	%5 = icmp ult i32 %1, %0		; <i1> [#uses=1]
+	br i1 %5, label %bb2, label %bb
+
+bb:		; preds = %entry
+	%6 = lshr i64 %v, 32		; <i64> [#uses=1]
+	%7 = trunc i64 %6 to i32		; <i32> [#uses=1]
+	%asmtmp = tail call { i32, i32 } asm "subl $5,$1\0A\09sbbl $3,$0", "=r,=&r,0,imr,1,imr,~{dirflag},~{fpsr},~{flags}"(i32 %3, i32 %7, i32 %1, i32 %0) nounwind		; <{ i32, i32 }> [#uses=2]
+	%asmresult = extractvalue { i32, i32 } %asmtmp, 0		; <i32> [#uses=1]
+	%asmresult1 = extractvalue { i32, i32 } %asmtmp, 1		; <i32> [#uses=1]
+	br label %bb2
+
+bb2:		; preds = %bb, %entry
+	%n1.0 = phi i32 [ %asmresult, %bb ], [ %3, %entry ]		; <i32> [#uses=1]
+	%n0.0 = phi i32 [ %asmresult1, %bb ], [ %1, %entry ]		; <i32> [#uses=1]
+	%8 = add i32 %n0.0, %n1.0		; <i32> [#uses=1]
+	ret i32 %8
+}
+
+declare i32 @printf(i8*, ...) nounwind