[llvm-commits] [llvm] r145731 - in /llvm/trunk: lib/Transforms/Scalar/DeadStoreElimination.cpp test/Transforms/DeadStoreElimination/OverwriteStoreEnd.ll

[Pete Cooper]

Author: pete
Date: Fri Dec  2 18:04:30 2011
New Revision: 145731

URL: http://llvm.org/viewvc/llvm-project?rev=145731&view=rev
Log:
Fixed deadstoreelimination bug where negative indices were incorrectly causing the optimisation to occur

Turns out long long + unsigned long long is unsigned.  Doh!

Fixes http://llvm.org/bugs/show_bug.cgi?id=11455

Modified:
    llvm/trunk/lib/Transforms/Scalar/DeadStoreElimination.cpp
    llvm/trunk/test/Transforms/DeadStoreElimination/OverwriteStoreEnd.ll

Modified: llvm/trunk/lib/Transforms/Scalar/DeadStoreElimination.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Transforms/Scalar/DeadStoreElimination.cpp?rev=145731&r1=145730&r2=145731&view=diff
==============================================================================
--- llvm/trunk/lib/Transforms/Scalar/DeadStoreElimination.cpp (original)
+++ llvm/trunk/lib/Transforms/Scalar/DeadStoreElimination.cpp Fri Dec  2 18:04:30 2011
@@ -416,7 +416,7 @@
   // writes to addresses which will definitely be overwritten later
   if (LaterOff > EarlierOff &&
       LaterOff < int64_t(EarlierOff + Earlier.Size) &&
-      LaterOff + Later.Size >= EarlierOff + Earlier.Size)
+      int64_t(LaterOff + Later.Size) >= int64_t(EarlierOff + Earlier.Size))
     return OverwriteEnd;
 
   // Otherwise, they don't completely overlap.

Modified: llvm/trunk/test/Transforms/DeadStoreElimination/OverwriteStoreEnd.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Transforms/DeadStoreElimination/OverwriteStoreEnd.ll?rev=145731&r1=145730&r2=145731&view=diff
==============================================================================
--- llvm/trunk/test/Transforms/DeadStoreElimination/OverwriteStoreEnd.ll (original)
+++ llvm/trunk/test/Transforms/DeadStoreElimination/OverwriteStoreEnd.ll Fri Dec  2 18:04:30 2011
@@ -76,3 +76,20 @@
 
 declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
 declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
+
+%struct.trapframe = type { i64, i64, i64 }
+
+; bugzilla 11455 - make sure negative GEP's don't break this optimisation
+; CHECK: @cpu_lwp_fork
+define void @cpu_lwp_fork(%struct.trapframe* %md_regs, i64 %pcb_rsp0) nounwind uwtable noinline ssp {
+entry:
+  %0 = inttoptr i64 %pcb_rsp0 to %struct.trapframe*
+  %add.ptr = getelementptr inbounds %struct.trapframe* %0, i64 -1
+  %1 = bitcast %struct.trapframe* %add.ptr to i8*
+  %2 = bitcast %struct.trapframe* %md_regs to i8*
+; CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 24, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 24, i32 1, i1 false)
+  %tf_trapno = getelementptr inbounds %struct.trapframe* %0, i64 -1, i32 1
+  store i64 3, i64* %tf_trapno, align 8
+  ret void
+}