dmraid -r -E bus error
Heinz Mauelshagen
mauelshagen at redhat.com
Thu Oct 25 08:51:48 UTC 2007
On Thu, Oct 25, 2007 at 10:18:29AM +0800, Ian Kent wrote:
> On Wed, 2007-10-24 at 13:23 -0700, Hall, Eric R wrote:
> > >> http://www.redhat.com/archives/ataraid-list/2007-June/msg00037.html
> > >
> > >I've had a look at the patch and I admit that the source I'm working
> > >from doesn't contain this change and since I wasn't subscribed to the
> > >list at that time I don't know the history.
> > >
> > >It's not obvious, from that message (or messages in the thread index),
> > >what the patch actually does to prevent a SEGV and I don't see how this
> > >deals with the issue I'm trying to resolve, the SEGV when writing the
> > >meta data.
> > >
> > >Can someone explain the intent of the patch in the message above
> > please?
> >
> > The patch above applies to isw_read_extended(). The segfault occurs
> > when the metadata is smaller than 512 bytes when it tries to remove
> > metadata from the second disk in the set. The patch allocates an extra
> > sector even when the metadata is only one sector long.
>
> Yep, I get it now, thanks.
>
> The patch simply ensures that sufficient memory is allocated during the
> read for isw_write to function correctly, instead of not doing the copy
> up as I did. So it should also resolve the issue.
>
> Oh well, I guess whoever decides what goes in to the release tarball has
> a small decision to make.
Let's avoid the superfluous allocation and take Ians patch.
Thanks,
Heinz
>
> Ian
>
>
> _______________________________________________
> Ataraid-list mailing list
> Ataraid-list at redhat.com
> https://www.redhat.com/mailman/listinfo/ataraid-list
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Heinz Mauelshagen Red Hat GmbH
Consulting Development Engineer Am Sonnenhang 11
Storage Development 56242 Marienrachdorf
Germany
Mauelshagen at RedHat.com PHONE +49 171 7803392
FAX +49 2626 924446
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
More information about the Ataraid-list
mailing list