Re: device-mapper: fix TB stripe data corruption

[Alasdair G Kergon]

On Fri, Jan 21, 2005 at 03:57:38PM -0600, Kevin Corry wrote:
> On Friday 21 January 2005 3:20 pm, Roland Dreier wrote:
> > If I understand you correctly, do_div() (defined in <asm/div64.h>)

I went for the simplest and safest fix first as this is a data
corruption problem and I want assurance that this fixes device-mapper
striping.

I didn't want to change it to do_div() without first checking
it would not slow down the code on the main architectures: on the contrary
I would hope that use of an optimised library inline speeds it up, but 
want to be sure.  You don't need the 64-bit mod until you have hundreds
of TB in a single logical volume block device, filesystem...

So far, I've only seen two test reports, both of which say they
are still seeing data corruption in a filesystem on top of dm-stripe
after applying this patch.  But none of this information so far
is specific enough to say whether the remaining problem(s) is/are
in device-mapper.

Alasdair
-- 
agk@xxxxxxxxxx
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/