On 2015-6-7 00:32, Dave Chinner wrote:
On Fri, Jun 05, 2015 at 06:03:53PM -0500, Eric Sandeen wrote:
On 6/5/15 5:22 PM, Dave Chinner wrote:
On Fri, Jun 05, 2015 at 03:28:45PM +0200, Ales Novak wrote:
However, this rule is only mentioned in the documentation, while it
could be checked during the mount.
Where in the documentation is that mentioned?
Documentation/filesystems/xfs.txt:
logbsize=value
Set the size of each in-memory log buffer. The size may be
specified in bytes, or in kilobytes with a "k" suffix.
Valid sizes for version 1 and version 2 logs are 16384 (16k)
and 32768 (32k). Valid sizes for version 2 logs also
include 65536 (64k), 131072 (128k) and 262144 (256k). The
logbsize must be an integer multiple of the log
stripe unit configured at mkfs time.
Ah, ok. I'll need to look at the history of that, because I think I
can see what it is intended to mean, but the "power-of-two" sizes
that are enforced will also enforce the "integer multiple" part,
too. I think it was more intended for people using wierd stripe
units (e.g. 96k) to say the equivalent on 2x96k is a valid log
buffer size. I suspect we need to revisit both the code and the
documentation here....
So, how will we proceed here? I understand that my patch only somehow
blindly implements the restriction mentioned in the documentation and that
proper solution would mean completely reworking these restrictions;
however I don't like the state where setting seemingly harmless
option leads to kernel crash (and possibly fs corruption).
While I understand that only playful root would do that.
--
Ales Novak
_______________________________________________
xfs mailing list
xfs@xxxxxxxxxxx
http://oss.sgi.com/mailman/listinfo/xfs
