[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: parted/d-i bug : vanishing print output, bug traced to table.c code.

From: Sven Luther
Subject: Re: parted/d-i bug : vanishing print output, bug traced to table.c code.
Date: Wed, 8 Nov 2006 00:47:12 +0100
User-agent: Mutt/1.5.13 (2006-08-11)

On Tue, Nov 07, 2006 at 08:36:44PM +0100, address@hidden wrote:
> On Tue, Nov 07, 2006 at 06:09:01PM +0100, Sven Luther wrote:
> > On Thu, Nov 02, 2006 at 02:12:40PM +0100, Sven Luther wrote:
> >
> > > Hi, ...
> > >
> > > it seems this bug is also affecting i386 and s390.
> > >
> > > I CCed address@hidden, since it probably concerns upstream
> > > too. For info, the bug is that in some cases (triggered in
> > > my case by running inside the debain-installer) print shows
> > > no output, which is pretty much damaging for tools which
> > > parse the parted output. Here is the original bug report :
> > > http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=392767
> >
> > Issue was because of a debian-specific patch, sorry to have bothered
> > you about this.
> Are you sure?  I recently ran into print not showing anything with
> Parted 1.7.1 on INSERT rescue disk (this is Knoppix-based, I don't know
> whether they use Debian patches).  The issue has also been reported on
> FreeBSD, IIRC.
>   The table code is new, and although I tried hard to eliminate any bugs
> (also with Valgrind), it cannot be guaranteed.
>   Is the problem gone when using a non-patched version?

Yes, altough i can test more.

The problem seems related to the ENABLE_NLS stuff, and the use of wchars. I
think the code may not be entirely non-NLS clean or something, i spotted a few
cases where L" " was used, like for the padding, without encompasing it in
ENABLE_NLS defines.

I am not very familiar with this part of it. Did you wrote the table.c code
yourself ? 

Also, the problem was related to a part of the FreeBSD support patch which did
strange hacky things to table.c, so this patch may have been applied to both
your test cases (Knoppix being debian-based, and the patch went in in may, and
FreeBSD guys probably using the same patch, which was written for them). The
table.c code added some garbage due to using a memcpy after the realloc, and
the fact that it used wchars meant that we suddenly had an extraneough \0.


Sven Luther

reply via email to

[Prev in Thread] Current Thread [Next in Thread]