xfs
[Top] [All Lists]

Re: [PATCH 5/5] xfs: remove shared supberlock feature checking

To: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject: Re: [PATCH 5/5] xfs: remove shared supberlock feature checking
From: Dave Chinner <david@xxxxxxxxxxxxx>
Date: Tue, 6 May 2014 18:28:57 +1000
Cc: xfs@xxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <20140506081820.GD24143@xxxxxxxxxxxxx>
References: <1399348559-19889-1-git-send-email-david@xxxxxxxxxxxxx> <1399348559-19889-6-git-send-email-david@xxxxxxxxxxxxx> <20140506081820.GD24143@xxxxxxxxxxxxx>
User-agent: Mutt/1.5.21 (2010-09-15)
On Tue, May 06, 2014 at 01:18:20AM -0700, Christoph Hellwig wrote:
> On Tue, May 06, 2014 at 01:55:59PM +1000, Dave Chinner wrote:
> > 
> > We reject any filesystem that is mounted with this feature bit set,
> > so we don't need to check for it anywhere else. Remove the function
> > for checking if the feature bit is set and any code that uses it.
> > 
> > Signed-off-by: Dave Chinner <dchinner@xxxxxxxxxx>
> 
> Looks good,
> 
> Reviewed-by: Christoph Hellwig <hch@xxxxxx>
> 
> If someone at SGI has some spare time for a BugWorks search I'd really
> love to figure out of the story behind the shared superblock feature..

http://oss.sgi.com/cgi-bin/gitweb.cgi?p=archive/xfs-import.git;a=commitdiff;h=9ea7d56d2db84bd45183c4ab11903b97c5b9a8e9

tells us that the PV to start looking at is 522678. It's something
to do with shared read-only filesystems - I suspect the sb_shared_vn
field counts the number of read-only mounts - and my guess is it was
for Cellular Irix. i.e. SGI's aborted attempt to make a distributed
operating system that left useless crap all through the Irix 6.5
codebase...

Cheers,

Dve.
-- 
Dave Chinner
david@xxxxxxxxxxxxx

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