Re: RFC: [PATCH-2.6] Add helper function to lock multiple page cache pages.

[Anton Altaparmakov <aia21@cam.ac.uk>]

On Thu, 3 Feb 2005, Andrew Morton wrote:
> I did a patch which switched loop to use the file_operations.read/write
> about a year ago.  Forget what happened to it.  It always seemed the right
> thing to do..

How did you implement the write?  At the moment the loop driver gets hold 
of both source and destination pages (the latter via grab_cache_page() and 
aops->prepare_write()) and copies/transforms directly from the source to 
the destination page (and then calls commit_write() on the destination 
page).  Did you allocate a buffer for each request, copy/transform to the 
buffer and then submit the buffer via file_operations->write?  That would 
clearly be not very efficient but given fops->write() is not atomic I 
don't see how that could be optimised further...

Perhaps the loop driver should work as is when 
aops->{prepare,commit}_write() are not NULL and should fall back to 
a buffered fops->write() otherwise?

Or have I missed some way in which the fops->write() case can be 
optimized?

Best regards,

	Anton
-- 
Anton Altaparmakov <aia21 at cam.ac.uk> (replace at with @)
Unix Support, Computing Service, University of Cambridge, CB2 3QH, UK
Linux NTFS maintainer / IRC: #ntfs on irc.freenode.net
WWW: http://linux-ntfs.sf.net/ & http://www-stu.christs.cam.ac.uk/~aia21/