Re: Bug 1035537 - split -n k/N gives incorrect data on blocks after the first

[Sylvain Beucler <beuc@beuc.net>]

Hi,

On 19/05/2023 21:14, Chris Frey wrote:
> On Fri, May 19, 2023 at 08:45:23PM +0200, Sylvain Beucler wrote:
> > On 05/05/2023 05:14, Chris Frey wrote:
> > > 	https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1035537
> >
> > At first glance, it looks like this could lead to data corruption, and hence
> > warrant a 'grave' severity in the BTS.
>
> Thanks for the response!  Yes indeed, the split file results will be
> corrupt without the fix.
>
> Is the severity level something I control, or the maintainer?

Everybody can alter the severity in the BTS, check:
https://www.debian.org/Bugs/server-control#severity
https://www.debian.org/Bugs/Developer#severities
(of course the maintainer may set it to another level if he sees fit.)

> > However I believe both the Security Team (for a bullseye DSA or
> > point-update, probably initiated by the maintainer) and the LTS Team (for a
> > buster DLA, probably based on bullseye's) would rather get the coreutil's
> > package maintainers input on the subject first (right now the BTS entry has
> > no replies) :)
>
> I'm assuming the package maintainers have a closer relationship to the
> upstream maintainers than I do, but I can help if needed.

As you mentioned, upstream already fixed this issue, I would suggest 
linking the upstream commit (which comes with documentation and updated 
tests) for clarity:
https://git.savannah.gnu.org/gitweb/?p=coreutils.git;a=commitdiff;h=bb21daa125aeb4e32546309d370918ca47e612db

AFAICS the bug is here since stretch, what I'd recommend is to get 
Michael Stone's (package maintainer) opinion on the matter (does he 
thinks this warrants a fix in stable or not) :)

Cheers!
Sylvain Beucler
Debian LTS Team