xfs
[Top] [All Lists]

Re: [PATCH v2 1/3] xfs: use b_maps[] for discontiguous buffers

To: Dave Chinner <david@xxxxxxxxxxxxx>
Subject: Re: [PATCH v2 1/3] xfs: use b_maps[] for discontiguous buffers
From: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Date: Fri, 30 Nov 2012 18:09:38 -0500
Cc: Mark Tinguely <tinguely@xxxxxxx>, Christoph Hellwig <hch@xxxxxxxxxxxxx>, xfs@xxxxxxxxxxx
In-reply-to: <20121130222110.GB12955@dastard>
References: <20121128222309.109033307@.sgi.com> <20121128222622.688093599@xxxxxxx> <20121130160920.GA9873@xxxxxxxxxxxxx> <50B8F2FC.6060501@xxxxxxx> <20121130222110.GB12955@dastard>
User-agent: Mutt/1.5.21 (2010-09-15)
On Sat, Dec 01, 2012 at 09:21:10AM +1100, Dave Chinner wrote:
> > eeek. yep, I will loop through and count the pages needed in each
> > segment.
> 
> That's wrong - the pages must be mappable as a contiguous memory
> range. That's how this code avoids copying the data from
> discontiguous page ranges into a contiguous mapped memory range.
> If you separate each IO into to it's own set of pages, then you have
> to completely rewrite _xfs_buf_ioapply(), which AFAICT works just
> fine with the page allocation that is done right now....

Yeah, you're right.

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