Re: [PATCH v2 5/8] c: Do not ignore some forms of -Wimplicit-int in system headers

[Jason Merrill <jason@redhat.com>]

On 11/15/23 00:13, Sam James wrote:
> Florian Weimer <fweimer@redhat.com> writes:
>> * Sam James:
>>> Florian Weimer <fweimer@redhat.com> writes:
>>>
>>>> Most -Wimplicit-int warnings were unconditionally disabled for system
>>>> headers.  Only missing types for parameters in old-style function
>>>> definitions resulted in warnings.  This is inconsistent with the
>>>> treatment of other permerrors, which are active in system headers.
>>>
>>> The situation with system headers is kind of a mess still. I went
>>> looking for a bug for the -Wimplicit-int behaviour but I only found
>>> PR78000 for -Wimplicit-function-declaration. But in the bug, Joseph
>>> makes the same observation.
>>>
>>> I don't suppose he'll want to block on that at this late point though.
>>>
>>> Do you know offhand what Clang's behaviour is wrt warnings in system
>>> headers?
>>
>> Clang ignores these new errors in system headers by default.  I don't
>> know if that's deliberate or a bug.  Our permerrors are deliberately
>> active in system headers.  As the test shows, -Wimplicit-int really was
>> the outlier here because of that check outside the permerror machinery.
> 
> Thanks - my assumption was that it was more widespread because of a few
> lingering bugs I've seen a few projects complain about, but maybe they
> were using old versions or similar.

The permerror behavior was chosen for C++, which was the only front-end 
that used it before; we might want to reconsider if this patch set runs 
into problems with system C headers.

Jason