Re: ZRAM 3.10.6 Buffer I/O error

[Greg Kroah-Hartman <gregkh@linuxfoundation.org>]

On Tue, Aug 13, 2013 at 02:18:29PM +0900, Minchan Kim wrote:
> On Tue, Aug 13, 2013 at 08:01:11AM +0300, Thomas Backlund wrote:
> > Minchan Kim skrev 13.8.2013 07:33:
> > >Hello Greg,
> > >
> > >On Mon, Aug 12, 2013 at 08:58:02PM -0700, Greg Kroah-Hartman wrote:
> > >>On Tue, Aug 13, 2013 at 10:39:00AM +0900, Minchan Kim wrote:
> > >>>Hello,
> > >>>
> > >>>On Mon, Aug 12, 2013 at 07:39:46PM +0300, Thomas Backlund wrote:
> > >>>>12.08.2013 17:27, Jordi Pujol skrev:
> > >>>>>Hello,
> > >>>>>
> > >>>>>zram shows an error when mounting a swap partition,
> > >>>>>current version Linux kernel 3.10.6,
> > >>>>>previous versions worked, I suppose that latest zram patches have some
> > >>>>>problem,
> > >>>>>
> > >>>>>machine is an AMD64 dual core, 4GB RAM
> > >>>>>
> > >>>>>+ modprobe -qb zram num_devices=2
> > >>>>>+ echo 104857600 > /sys/block/zram0/disksize
> > >>>>>
> > >>>>># mkswap /dev/zram0
> > >>>>>Setting up swapspace version 1, size = 102396 KiB
> > >>>>>no label, UUID=6c249930-2ba0-46cf-a8c6-766481942b7d
> > >>>>>
> > >>>>># pager /var/log/dmesg
> > >>>>>(no more errors shown)
> > >>>>>
> > >>>>># swapon /dev/zram0
> > >>>>>
> > >>>>># pager /var/log/dmesg
> > >>>>>...
> > >>>>>[  309.300479] Buffer I/O error on device zram0, logical block 25599
> > >>>>>[  309.300491] Buffer I/O error on device zram0, logical block 25599
> > >>>>>[  309.300514] Buffer I/O error on device zram0, logical block 25599
> > >>>>>[  345.205887] Adding 102396k swap on /dev/zram0.  Priority:-2 extents:1
> > >>>>>across:102396k SSFS
> > >>>>>
> > >>>>>full log files and the kernel source are stored in the address:
> > >>>>>
> > >>>>>http://livenet.selfip.com/ftp/debian/zram_3.10.6_IO_error/
> > >>>>>
> > >>>>
> > >>>>
> > >>>>I think this one should fix it and belongs in 3.10 stable too:
> > >>>
> > >>>Very true. Let's Cc Greg, explicitely.
> > >>
> > >>-ENOCONTEXT
> > >
> > >https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/commit/drivers/staging/zram?id=75c7caf5a052ffd8db3312fa7864ee2d142890c4
> > >
> > >It seems you already merge "zram:allow request end to conincide with disksize" to stable.
> > >So there is no concern any more.
> > 
> > Nope...
> > 
> > thats master branch wich tracks upstream tree...
> > 
> > it's not in the 3.10 stable as of 3.10.6 (or current stable queue)
> > 
> > https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/log/drivers/staging/zram?h=linux-3.10.y
> 
> Okay, Then, we should explain it to Greg.
> 
> While Jiang was fixing a BUG A with [1], [1] made another BUG B.
> So, Sergey fixed BUG B with [2]. But unfortunately, [1] was in stable
> now but [2] wasn't so [2] should go to the stable to fix BUG B.
> 
> [1] 12a7ad3b8, zram: avoid access beyond the zram device
> [2] 75c7caf5a, zram: allow request end to coincide with disksize

Ok, but that's what Thomas's original email to stable@vger.kernel.org
asked for, to apply the 2nd patch above to the stable tree, which I've
already done...

Emailing me directly, with half-of-a-thread causes me nothing but
confusion, as I'm sure you can imagine.

thanks,

greg k-h