[LLVMbugs] [Bug 18606] New: A very small test resulting in extremely long compilation in "clang -O2" version 3.5, trunk 199677

[bugzilla-daemon at llvm.org]

http://llvm.org/bugs/show_bug.cgi?id=18606

            Bug ID: 18606
           Summary: A very small test resulting in extremely long
                    compilation in "clang  -O2" version 3.5, trunk 199677
           Product: tools
           Version: trunk
          Hardware: Macintosh
                OS: MacOS X
            Status: NEW
          Severity: normal
          Priority: P
         Component: opt
          Assignee: unassignedbugs at nondot.org
          Reporter: mohammad.r.haghighat at intel.com
                CC: llvmbugs at cs.uiuc.edu
    Classification: Unclassified

Created attachment 11934
  --> http://llvm.org/bugs/attachment.cgi?id=11934&action=edit
Sources of t27056 test including its source, input, and outputs

This is one of the tests generated by icFuzz in the attachment of comment
http://llvm.org/bugs/show_bug.cgi?id=16431#c30

the test resuls in an extremely long compilation in clang version 3.5, trunk
199677. Clang -O2 compilation seems to run until it goes out of memory. The
core of the test is very simple:

    for (kj = 1; kj < 2; ++kj) {
        for (jd = 1; jd < 32; ++jd) {
            zz *= zz;
        }
    }

I guess the problem is a result of unrolling and creating a sequence of
statements that result in analysis nodes that have exponential complexity. If
the 2nd loop upper bound (32) is changed to a larger number (e.g., 132) or a
much smaller number (e.g., 12), compilation terminates fast.

-- 
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/20140124/f8d286e7/attachment.html>