Messages in this thread | | | From | Rasmus Villemoes <> | Subject | Re: [PATCH 1/4] lib/vsprintf.c: handle invalid format specifiers more robustly | Date | Mon, 28 Sep 2015 22:12:23 +0200 |
| |
On Mon, Sep 28 2015, Andy Shevchenko <andriy.shevchenko@linux.intel.com> wrote:
>> /* >> - * Since %n poses a greater security risk than >> utility, treat >> - * it as an invalid format specifier. Warn about its >> use so >> - * that new instances don't get added. >> + * Since %n poses a greater security risk than > > Any reason to wrap first string?
I just hit M-q in emacs and let that take care of somewhat sane wrapping. I don't play diff golf.
>> + /* >> + * Presumably the arguments passed gcc's >> type >> + * checking, but there is no safe or sane >> way >> + * for us to continue parsing the format and >> + * fetching from the va_list; the remaining >> + * specifiers and arguments would be out of >> + * sync. > > Could we use wider strings in the commentary here?
Ditto.
Rasmus
| |