GNU bug report logs - #23234
unexpected results with charset handling in GNU grep 2.23

Previous Next

Full log

Message #49 received at 23234 <at> debbugs.gnu.org (full text, mbox):

From: Zev Weiss <zev <at> bewilderbeest.net>
To: Jim Meyering <jim <at> meyering.net>
Cc: Bjoern Jacke <bjoern <at> j3e.de>, 23234-done <at> debbugs.gnu.org,
 Paul Eggert <eggert <at> cs.ucla.edu>, 23234 <at> debbugs.gnu.org
Subject: Re: bug#23234: unexpected results with charset handling in GNU grep
 2.23
Date: Sun, 10 Apr 2016 16:59:08 -0500

On Sun, Apr 10, 2016 at 02:10:08PM -0700, Jim Meyering wrote:
>On Sun, Apr 10, 2016 at 1:43 AM, Paul Eggert <eggert <at> cs.ucla.edu> wrote:
>> Paul Eggert wrote:
>>>
>>> I plan to change GNU grep to use this new facility, and to add some grep
>>> test
>>> cases for this issue.
>>
>>
>> I did that by installing the attached patches into the grep master. This
>> fixes the bug for me, so I'm closing the bug report.
>>
>> These patches mostly just report the fix and add test cases. The actual fix
>> was in gnulib, here:
>>
>> http://git.savannah.gnu.org/cgit/gnulib.git/commit/?id=b7bc3c1a4e78add4cbad39ae1a0c4fb0747b483f
>>
>> This gnulib fix works around the underyling glibc facility which caused the
>> problem, for which I've filed a bug report here:
>>
>> https://sourceware.org/bugzilla/show_bug.cgi?id=19932
>>
>> It's not clear when the glibc bug will be fixed. Until it is, one should
>> expect similar problems to crop up in applications other than 'grep'.
>
>Thanks for the fine work, Paul.
>With this fix, I would like to make yet another grep release.
>Does anyone have any pending changes we should consider?
>
>Incidentally, looking at mbrtowc uses, I found an unused
>function and removed it with this patch:

Well, I still have my multithreading patch series 
(https://github.com/zevweiss/grep/) awaiting review, which I'd hope to 
get applied at some point, though I'd guess it's enough of a review task 
that delaying an impending release for it isn't likely (the 
mbtoupper()-removal patch made that series one patch shorter though, 
since one was to deal with that function's thread-unsafety).  I've been 
rebasing it periodically and running it on my own system in /usr/local 
without any problems for a while now, for what that's worth.

With current HEAD from savannah though, all check-very-expensive tests 
pass for me on Debian stretch with gcc 5.3, glibc 2.22, and Linux kernel 
4.3.


Zev Weiss

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.