[Top] [All Lists]

Re: [PATCH 4/7][TAKE5] support new modes in fallocate

To: David Chinner <dgc@xxxxxxx>, Andreas Dilger <adilger@xxxxxxxxxxxxx>
Subject: Re: [PATCH 4/7][TAKE5] support new modes in fallocate
From: Nathan Scott <nscott@xxxxxxxxxx>
Date: Thu, 28 Jun 2007 09:28:36 +1000
Cc: xfs-oss <xfs@xxxxxxxxxxx>, "Amit K. Arora" <aarora@xxxxxxxxxxxxxxxxxx>, linux-fsdevel@xxxxxxxxxxxxxxx, linux-ext4@xxxxxxxxxxxxxxx, suparna@xxxxxxxxxx, cmm@xxxxxxxxxx
In-reply-to: <20070627133657.GQ989688@xxxxxxx>
Organization: Aconex
References: <20070613235217.GS86004887@xxxxxxx> <20070614091458.GH5181@xxxxxxxxxxxxxxxxxxxx> <20070614120413.GD86004887@xxxxxxx> <20070614193347.GN5181@xxxxxxxxxxxxxxxxxxxx> <20070625132810.GA1951@xxxxxxxxxxxxxxxxxxxx> <20070625134500.GE1951@xxxxxxxxxxxxxxxxxxxx> <20070625150320.GA8686@xxxxxxxxxxxxxxxxxxxx> <20070625214626.GJ5181@xxxxxxxxxxxxxxxxxxxx> <20070626231431.GO31489@xxxxxxx> <20070627034915.GR6652@xxxxxxxxxxxxxxxxxxxx> <20070627133657.GQ989688@xxxxxxx>
Reply-to: nscott@xxxxxxxxxx
Sender: xfs-bounce@xxxxxxxxxxx
On Wed, 2007-06-27 at 23:36 +1000, David Chinner wrote:
> .... Allows setup_swap_extents() to use preallocated files on XFS
> filesystems for swap files without ever needing to convert them.

Using unwritten extents (as opposed to the MKSWAP flag mentioned
earlier) has the unfortunate down side of requiring transactions,
possibly additional IO, and memory allocation during swap.  (but,
this patch should probably go in regardless, as teaching generic
code about unwritten extents is not a bad idea).



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