[cfe-commits] [Patch] -Wformat: better handling of qualifiers on pointer arguments

Hans Wennborg hans at chromium.org
Tue Jul 31 09:39:41 PDT 2012


Thanks! Landed r161052.

On Tue, Jul 31, 2012 at 5:19 PM, Jordan Rose <jordan_rose at apple.com> wrote:
> LGTM!
>
>
> On Jul 31, 2012, at 8:57 , Hans Wennborg <hans at chromium.org> wrote:
>
>> Thanks, Jordan!
>>
>> I've added some tests for the "const volatile" case. Yes, the types
>> have already been canonicalized, but I've added a couple of tests for
>> that anyway.
>>
>> New patch attached. OK to commit?
>>
>> - Hans
>>
>> On Tue, Jul 31, 2012 at 4:23 PM, Jordan Rose <jordan_rose at apple.com> wrote:
>>> Wow, nice idea. I would add "const volatile" to your test cases as well, even though I'm pretty sure "removeLocalVolatile" does the right thing.
>>>
>>> (Also, typedefs? Or have we already canonicalized the type by this point?)
>>>
>>>
>>> On Jul 31, 2012, at 3:25 AM, Hans Wennborg <hans at chromium.org> wrote:
>>>
>>>> Hi all,
>>>>
>>>> The attached patch makes Clang warn about passing in pointers to
>>>> const-qualified types to scanf, such as passing in "const int*" for a
>>>> "%d" format string.
>>>>
>>>> It also ignores the volatile-qualifier, allowing to pass in a
>>>> "volatile int*" pointer for a "%n" format string.
>>>>
>>>> Please take a look!
>>>>
>>>> Thanks,
>>>> Hans
>> <wformat_type_qualifiers2.patch>
>



More information about the cfe-commits mailing list