This is the mail archive of the
mailing list for the glibc project.
Re: [PING][PATCH] Fix strcpy_chk and stpcpy_chk performance.
- From: Joseph Myers <joseph at codesourcery dot com>
- To: Adhemerval Zanella <adhemerval dot zanella at linaro dot org>
- Cc: <libc-alpha at sourceware dot org>
- Date: Wed, 19 Aug 2015 14:53:00 +0000
- Subject: Re: [PING][PATCH] Fix strcpy_chk and stpcpy_chk performance.
- Authentication-results: sourceware.org; auth=none
- References: <20150812064213 dot GA6256 at domone> <20150819061522 dot GA19761 at domone> <55D4805D dot 5010001 at linaro dot org>
On Wed, 19 Aug 2015, Adhemerval Zanella wrote:
> Could you also get rid of the k&r function prototype (same for strcpy_chk) ?
I'm not really convinced it's good to mix such fixes with changes for
other things (at least it shouldn't be required).
OndÅej had some automation for fixing such old-style definitions, see e.g.
<https://sourceware.org/ml/libc-alpha/2013-06/msg00270.html>. I think it
would be good to try to get such changes in (eventually adding
-Wold-style-definition to the warning options, possibly with an
intermediate state of -Wold-style-definition
-Wno-error=old-style-definition while architecture maintainers clean up
their architectures) - starting with straightforward cases in files that
do not use assertions (so changes to line numbers don't affect comparisons
of generated code) and for which the changes do not affect the installed
stripped shared libraries at all, then moving to other cases for which
more careful review is needed.
Joseph S. Myers