[llvm-bugs] [Bug 27433] New: LVI misses opportunities to conclude overdefined

[via llvm-bugs]

https://llvm.org/bugs/show_bug.cgi?id=27433

            Bug ID: 27433
           Summary: LVI misses opportunities to conclude overdefined
           Product: new-bugs
           Version: trunk
          Hardware: PC
                OS: All
            Status: NEW
          Severity: normal
          Priority: P
         Component: new bugs
          Assignee: unassignedbugs at nondot.org
          Reporter: regehr at cs.utah.edu
                CC: listmail at philipreames.com, llvm-bugs at lists.llvm.org
    Classification: Unclassified

Following each icmp ne + branch, LVI properly concludes that x is not 10 and
20, respectively. However, it should also put these facts together and conclude
that x is overdefined at cont2. Instead of intersecting the incoming dataflow
fact with the new one, it seems to only overwrite it, concluding at cont2 that
x != 20.

define void @foo8(i32 %x) #0 {
entry:
  %cmp = icmp ne i32 %x, 10
  br i1 %cmp, label %if.end3, label %cont

cont:
  %cmp1 = icmp ne i32 %x, 20
  br i1 %cmp1, label %if.end3, label %cont2

cont2:
  ;; LVI should be able to tell we can't get here!
  br label %if.end3

if.end3:
  ret void
}

-- 
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20160420/488399e7/attachment.html>