This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH] Convert old prototype.
- From: "Carlos O'Donell" <carlos at redhat dot com>
- To: OndÅej BÃlka <neleai at seznam dot cz>
- Cc: libc-alpha at sourceware dot org
- Date: Sat, 08 Jun 2013 10:40:34 -0400
- Subject: Re: [PATCH] Convert old prototype.
- References: <20130608104926 dot GA17305 at domone dot kolej dot mff dot cuni dot cz>
On 06/08/2013 06:49 AM, OndÅej BÃlka wrote:
> Continuing cleanup we now focus on converting old prototype to new.
>
> I now have patch that does formating rigth, see:
>
> http://kam.mff.cuni.cz/~ondra/convert_old_prototype.patch
I notice that __md5_crypt_r is wrapped, but unwrapped it would
only be 75 characters. What did your script wrap it? The same
goes for __sha256_crypt_r which is only 78 chars without using
two lines. Is there readability heuristic? I haven't gone through
your converter program yet.
> It could be applied as it is but it is not exhaustive as I needed
> to exclude several special cases like macro to modify function name.
It need not be exhaustive. A first pass that fixes 99% of these
cases and helps us move towards -Werror would be awesome. I dream
of the day when we can use -Werror :-)
> I could make this patch exhaustive if I could do
> make and make check with enabled
> -Wold-style-definition
> , my attempts ended in getting errors from -Werror.
I think we want to make incremental progress towards -Werror
witout needing to use any intermediate hacks before we enable it.
Does that make sense? Did I understand your comment correctly?
> then I could parse that output and run my script on these lines.
If it's only temporary, then you're free to do so yourself and
free to request the machine maintainers run with your special
patch and provide you their build logs?
Cheers,
Carlos.