[LLVMbugs] [Bug 23277] New: __builtin_object_size(null_pointer) returns the wrong result

bugzilla-daemon at llvm.org bugzilla-daemon at llvm.org
Sat Apr 18 09:37:31 PDT 2015


https://llvm.org/bugs/show_bug.cgi?id=23277

            Bug ID: 23277
           Summary: __builtin_object_size(null_pointer) returns the wrong
                    result
           Product: clang
           Version: unspecified
          Hardware: All
                OS: All
            Status: NEW
          Severity: normal
          Priority: P
         Component: -New Bugs
          Assignee: unassignedclangbugs at nondot.org
          Reporter: danielmicay at gmail.com
                CC: llvmbugs at cs.uiuc.edu
    Classification: Unclassified

This outputs (size_t)-1 in both cases with `gcc -O2` but clang gives 0 for the
second case:

  #include <stdio.h>

  int main(int argc, char **argv) {
    printf("%zu\n", __builtin_object_size(NULL, 0));
    void *p = NULL;
    printf("%zu\n", __builtin_object_size(p, 0));
    return 0;
  }

This is broken with glibc's _FORTIFY_SOURCE implementation for nullable
parameters. It assumes that it can do comparisons like `object_size(buffer) <
copy_size` after handling the unknown case, but this can result in a false
positive with Clang. It's not a glibc bug because they're using GNU C as it is
defined by GCC, which is where this feature comes from.

-- 
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/20150418/ada64f34/attachment.html>


More information about the llvm-bugs mailing list