GNU bug report logs - #55891
[PATCH] gnu: Add iec16022

Previous Next

Full log

View this message in rfc822 format

From: Remco van 't Veer <remco <at> remworks.net>
To: Maxime Devos <maximedevos <at> telenet.be>
Cc: 55891 <at> debbugs.gnu.org
Subject: [bug#55891] [PATCH] gnu: Add iec16022
Date: Sat, 25 Jun 2022 16:53:19 +0200

2022/06/10 23:47, Maxime Devos:

> I think it's ok, _if_ the bugs are eventually at least reported
> upstream if we (i.e., you) cannot fix them timely (needs some
> contrived circumstances to trigger the bug).

The maintainer's stance on not checking malloc response (from
https://github.com/rdoeffinger/iec16022/pull/17#issuecomment-1152892833):

> > I'll open an issue for the unchecked [mc]alloc and write calls
> > instead if you do and maybe have another stab at fixing these issues
> > when my C-programming skills have improved.
>
> I agree on the unchecked writes. For the unchecked [mc]alloc I am more
> tempted towards removing the existing checks, since I am not convinced
> checking won't always do more harm than good for those.

He elaborated on this at:

  https://github.com/rdoeffinger/iec16022/pull/17#issuecomment-1153103749

Boiling down to: when memory is that low exiting with a proper error
messages and doing cleanup probably won't work anyway and the OOM killer
will have swept in before the situation exists.  Which makes sense,
IMHO, but I am no expert.

I've created an issue for the unchecked writes:

  https://github.com/rdoeffinger/iec16022/issues/18

HTH,
Remco

Previous Next

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