This is the mail archive of the
mailing list for the eCos project.
Re: [ECOS] JFFS2 Garbage Collection Error? (NO! rbtree.c bug)
- From: Gary Thomas <gary at chez-thomas dot org>
- To: Scott Wilkinson <scott at alliantnetworks dot com>
- Cc: David Woodhouse <dwmw2 at infradead dot org>,"Gary D. Thomas" <gary at mind dot be>,eCos Discussion <ecos-discuss at sources dot redhat dot com>,eCos patches <ecos-patches at sources dot redhat dot com>
- Date: 14 Jul 2003 11:16:40 -0600
- Subject: Re: [ECOS] JFFS2 Garbage Collection Error? (NO! rbtree.c bug)
- References: <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com>
On Mon, 2003-07-14 at 11:05, Scott Wilkinson wrote:
> On Sun, 2003-07-13 at 10:55, David Woodhouse wrote:
> > On Sat, 2003-07-12 at 23:55, Scott Wilkinson wrote:
> > > If, for testing purposes, I compile eCos using the linux rbtree.[ch]
> > > instead of the eCos/OpenBSD port, my JFFS2 stuff works (and very well so
> > > far I might add!).
> > I note the trees dumped before the call to rb_erase() and the crash are
> > identical in the Linux and eCos versions -- it doesn't seem to be an
> > error in _building_ the tree, only removing stuff from it.
> > I note further that your test case also works if I take the Linux
> > rb_erase() function and drop it into the BSD/eCos rbtree.c, adjusting
> > name and order of rb_remove_color() and its arguments respectively.
> I believe I fixed the bug in the eCos rbtree.c that was causing JFFS2 to
> fail during certain garbage collection events. It was a simple problem
> of copying pointers to structures instead of copying the entire
> structure itself, as is necessary.
> I am continuing to test, but the attached patch seems to do the trick!
That's great! Are there any other [similar] issues like this in that
Gary Thomas <firstname.lastname@example.org>