[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <Y4+r2c+d+gWi93og@casper.infradead.org>
Date: Tue, 6 Dec 2022 20:53:45 +0000
From: Matthew Wilcox <willy@...radead.org>
To: "Vishal Moola (Oracle)" <vishal.moola@...il.com>
Cc: linux-mm@...ck.org, tytso@....edu, linux-ext4@...r.kernel.org,
linux-kernel@...r.kernel.org, akpm@...ux-foundation.org
Subject: Re: [PATCH mm-unstable] ext4: Convert mext_page_double_lock() to
mext_folio_double_lock()
On Tue, Dec 06, 2022 at 08:50:02PM +0000, Matthew Wilcox wrote:
> > flags = memalloc_nofs_save();
> > - page[0] = grab_cache_page_write_begin(mapping[0], index1);
> > - if (!page[0]) {
> > + folio[0] = __filemap_get_folio(mapping[0], index1, fgp_flags,
> > + mapping_gfp_mask(mapping[0]));
>
> one
>
> > + if (!folio[0]) {
> > memalloc_nofs_restore(flags);
> > return -ENOMEM;
> > }
> >
> > - page[1] = grab_cache_page_write_begin(mapping[1], index2);
> > + folio[1] = __filemap_get_folio(mapping[1], index2, fgp_flags,
> > + mapping_gfp_mask(mapping[1]));
>
> two
*facepalm*. Those don't contain calls to compound_head(), they contain
calls to folio_file_page(), which is still a moderately-expensive
unnecessary conversion, but a conversion in the other direction.
Powered by blists - more mailing lists