[Lguest] [PATCH] lguest.c: truncate problems in handle_block_output

[Chris Malley]

On Wed, 2007-09-26 at 09:53 +1000, Rusty Russell wrote:
> On Tue, 2007-09-25 at 17:13 +0100, Chris Malley wrote:
> > Hi
> > 
> > I have been experiencing consistent problems with an lguest under moderate
> > I/O conditions (e.g. when updating the distribution with yum) where the guest
> > crashes with "Write past end 2965302784+4294967295", and I end up with a
> > zero-length disk image.
> > 
> > Turns out this is because (a) the offset is *not* 4294967295, but the failure
> > code -1 (so in fact it isn't a "Write past end" at all, just a write failure).
> > And (b) because my block device is 4G and ftruncate() can't handle it, the 
> > file is truncated to zero bytes.
> 
> Hi Chris,
> 
> 	Thanks for the patch!  But I can't figure out why you are getting a
> write failure to trigger this.  We open it with O_LARGEFILE: are you
> running out of disk space?

It's a possibility; that filesystem is getting pretty full up with guests :-)
That's why I put the errno reporting in there, to see if I could get a
decent error code.  If I find out I'll post a followup.

> The ftruncate vs ftrucate64 bug is nasty.  The block stuff is changing
> for 2.6.24, but of course I had the same bug there too...
>
> >  		len = readv(dev->fd, reply, reply_num);
> > -		*lenp = len;
> > +		if (len < 0) {
> > +			err(1, "Read failed with error %u", errno);
> > +			*lenp = 0;
> > +		} else
> > +			*lenp = len;
> >  	}
> 
> BTW, we don't need the *lenp = 0 after the err() here.

Good point!  (in an earlier version it didn't err() on failure).

> Thanks!
> Rusty.
> 

No sweat; thanks for the great work, lguest is indeed fun to hack on as
promised :-)

--
Chris