xfs
[Top] [All Lists]

Re: Cant find sunit and swidth on luks encrypted raid6 device

To: nscott@xxxxxxxxxx
Subject: Re: Cant find sunit and swidth on luks encrypted raid6 device
From: Eric Sandeen <sandeen@xxxxxxxxxxx>
Date: Wed, 19 Sep 2007 17:47:31 -0500
Cc: "Jochen K." <jochen.f.k@xxxxxxxxxxxxxx>, xfs@xxxxxxxxxxx
In-reply-to: <33249.192.168.3.1.1190241583.squirrel@xxxxxxxxxxxxxxx>
References: <728b8e3b0709191129j27b7c319h2efd51ffc9cb64f3@xxxxxxxxxxxxxx> <46F17BBE.7060304@xxxxxxxxxxx> <46F1A3D7.5090805@xxxxxxxxxxx> <33249.192.168.3.1.1190241583.squirrel@xxxxxxxxxxxxxxx>
Sender: xfs-bounce@xxxxxxxxxxx
User-agent: Thunderbird 1.5.0.12 (X11/20070530)
nscott@xxxxxxxxxx wrote:
>> Eric Sandeen wrote:
>>
>> 2b2d4c133680: Badness in key lookup (length)
>> bp=(bno 13673620728, len 131072 bytes) key=(bno 13673620728, len 4096
>> bytes)
>>
>> If we take out the pwrite that does the end-of-device zeroing, we don't
>> get the error... but that's been there since forever, so again, I'm a
>> bit confused here.
> 
> libxfs has been changing alot - most likely the problem lies in the buffer
> caching
> code there - Barry will probably recognise the error (when he gets online).

--- xfs-cmds/xfsprogs/libxfs/rdwr.c     2007/02/21 14:37:52     1.34
+++ xfs-cmds/xfsprogs/libxfs/rdwr.c     2007/07/16 15:55:26     1.35
@@ -24,6 +24,8 @@
 #define BDSTRAT_SIZE   (256 * 1024)
 #define min(x, y)      ((x) < (y) ? (x) : (y))

+#define IO_BCOMPARE_CHECK
+

Yep.... :)

-Eric

> cheers.
> 
> --
> Nathan
> 
> 


<Prev in Thread] Current Thread [Next in Thread>