[llvm-bugs] [Bug 34817] New: False positive on reading undefined value from streams

via llvm-bugs llvm-bugs at lists.llvm.org
Tue Oct 3 09:16:48 PDT 2017


https://bugs.llvm.org/show_bug.cgi?id=34817

            Bug ID: 34817
           Summary: False positive on reading undefined value from streams
           Product: clang
           Version: trunk
          Hardware: Macintosh
                OS: MacOS X
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: Static Analyzer
          Assignee: kremenek at apple.com
          Reporter: howard.hinnant at gmail.com
                CC: llvm-bugs at lists.llvm.org

This code gets an analyzer warning:

#include <cassert>
#include <sstream>

int
main()
{
    std::istringstream in{""};
    char c;
    in >> c;
    assert(!in.fail());
    return int{c};
}

scan-build: Using '/Users/howardhinnant/Development/llvm_build/bin/clang-6.0'
for static analysis
test.cpp:492:5: warning: Undefined or garbage value returned to caller
    return int{c};
    ^~~~~~~~~~~~~
1 warning generated.
scan-build: 1 bug found.
scan-build: Run 'scan-view
/var/folders/7d/bq37f_s53dbgw_z48smv9q9m0000gn/T/scan-build-2017-10-03-121247-59896-1'
to examine bug reports.

It would be nice if the analyzer could detect that c can't be garbage because
in.fail() is checked and if true never reaches the use of c.

Even better would be this:

#include <cassert>
#include <sstream>

int
main()
{
    std::istringstream in{""};
    in.exceptions(std::ios::failbit | std::ios::badbit);
    assert(((std::ios::failbit | std::ios::badbit) & in.exceptions()) ==
            (std::ios::failbit | std::ios::badbit));
    char c;
    in >> c;
    return int{c};
}

c can not be used with garbage, proven by the settings in in.exceptions().  So
no warning on the above either.

-- 
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/20171003/509f6df8/attachment.html>


More information about the llvm-bugs mailing list