[LLVMbugs] [Bug 16431] New: [icFuzz] Incorrect results at -O2 optimization and infinite loop in the Clang optimizer

bugzilla-daemon at llvm.org bugzilla-daemon at llvm.org
Mon Jun 24 14:18:31 PDT 2013


            Bug ID: 16431
           Summary: [icFuzz] Incorrect results at -O2 optimization and
                    infinite loop in the Clang optimizer
           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 10750
  --> http://llvm.org/bugs/attachment.cgi?id=10750&action=edit
File contains 107 small tests failing with the latest LLVM/clang 3.4 main trunk
(184563). Sources, compilation commands, test input file, and results at -O2
and -O0 are included.

Attached are a number of tests (107 of them) that currently fail on the clang
3.4 main trunk (184563). These tests have been automatically generated by an
internal tool at Intel, the Intel Compiler fuzzer -- icFuzz,. 

The tests are in two categories 
- diff: those that have different runtime outputs when compiled at -O0 and -O2 
- InfLoopInClang: those that cause infinite loops in the Clang optimizer 

Many of these failing tests could be due to the same bug. The tests are small.
Their main computation is between the statement that reads the random input
(using fscanf) and the statement that computes the final result (res = ...).
The tests have a nice property that each statement between these two boundaries
can be deleted, resulting in another correct running test. This would help test

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/20130624/07b6b5b4/attachment.html>

More information about the llvm-bugs mailing list