[llvm-commits] CVS: llvm/test/Regression/Transforms/LoopSimplify/2003-12-10-ExitBlocksProblem.ll
Chris Lattner
lattner at cs.uiuc.edu
Wed Dec 10 11:10:01 PST 2003
Changes in directory llvm/test/Regression/Transforms/LoopSimplify:
2003-12-10-ExitBlocksProblem.ll added (r1.1)
---
Log message:
New testcase which caused a bunch of tests to break last night with more
aggressive LICM
---
Diffs of the changes: (+42 -0)
Index: llvm/test/Regression/Transforms/LoopSimplify/2003-12-10-ExitBlocksProblem.ll
diff -c /dev/null llvm/test/Regression/Transforms/LoopSimplify/2003-12-10-ExitBlocksProblem.ll:1.1
*** /dev/null Wed Dec 10 11:09:45 2003
--- llvm/test/Regression/Transforms/LoopSimplify/2003-12-10-ExitBlocksProblem.ll Wed Dec 10 11:09:35 2003
***************
*** 0 ****
--- 1,42 ----
+ ; LoopSimplify is breaking LICM on this testcase because the exit blocks from
+ ; the loop are reachable from more than just the exit nodes: the exit blocks
+ ; have predecessors from outside of the loop!
+ ;
+ ; This is distilled from a monsterous crafty example.
+
+ ; RUN: llvm-as < %s | opt -licm | lli
+
+ %G = weak global int 0 ; <int*> [#uses=13]
+
+ implementation ; Functions:
+
+ int %main() {
+ entry:
+ store int 123, int* %G
+ br label %loopentry.i
+
+ loopentry.i: ; preds = %entry, %endif.1.i
+ %tmp.0.i = load int* %G ; <int> [#uses=1]
+ %tmp.1.i = seteq int %tmp.0.i, 123 ; <bool> [#uses=1]
+ br bool %tmp.1.i, label %Out.i, label %endif.0.i
+
+ endif.0.i: ; preds = %loopentry.i
+ %tmp.3.i = load int* %G ; <int> [#uses=1]
+ %tmp.4.i = seteq int %tmp.3.i, 126 ; <bool> [#uses=1]
+ br bool %tmp.4.i, label %ExitBlock.i, label %endif.1.i
+
+ endif.1.i: ; preds = %endif.0.i
+ %tmp.6.i = load int* %G ; <int> [#uses=1]
+ %inc.i = add int %tmp.6.i, 1 ; <int> [#uses=1]
+ store int %inc.i, int* %G
+ br label %loopentry.i
+
+ Out.i: ; preds = %loopentry.i
+ store int 0, int* %G
+ br label %ExitBlock.i
+
+ ExitBlock.i: ; preds = %endif.0.i, %Out.i
+ %tmp.7.i = load int* %G ; <int> [#uses=1]
+ ret int %tmp.7.i
+ }
+
More information about the llvm-commits
mailing list