Re: [PATCH 2.6.12-rc3 4/8] mm: manual page migration-rc2 -- add-sys_migrate_pages-rc2.patch

[Christoph Hellwig <hch@infradead.org>]

On Wed, May 18, 2005 at 02:07:21PM -0500, Ray Bryant wrote:
> Christoph Hellwig wrote:
> 
> >
> >>+	if (nr_busy > 0) {
> >>+		pass++;
> >>+		if (pass > 10)
> >>+			return -EAGAIN;
> >>+		/* wait until some I/O completes and try again */
> >>+		blk_congestion_wait(WRITE, HZ/10);
> >>+		goto retry;
> >
> >
> >this is a layering violation.  How to wait is up to the implementor
> >of the address_space
> >
> 
> Christoph,
> 
> I've done the other changes you suggested, but am a little confused
> by this one.  Is your suggestion that I should be calling:
> 
> vma->vm_file->f_mapping->a_ops->writepages()
> 
> (assuming this exists)
> 
> instead of doing the blk_congestion_wait()?  There is no "wait"
> function defined in the aops vector as near as I can tell.

I looked over the code again and most of the migration code isn't added
in the patchkit but expected to exist already, thus I'm not sure what's
going on at all.

address_space_operations are the wrong abstraction here, you're operating
on VMAs, thus any vectoring should happen at the vm_operations_struct
level.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"aart@kvack.org"> aart@kvack.org </a>