[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <Zdyk4ErGQHc1q-1W@casper.infradead.org>
Date: Mon, 26 Feb 2024 14:49:04 +0000
From: Matthew Wilcox <willy@...radead.org>
To: "Pankaj Raghav (Samsung)" <kernel@...kajraghav.com>
Cc: linux-xfs@...r.kernel.org, linux-fsdevel@...r.kernel.org,
linux-kernel@...r.kernel.org, david@...morbit.com,
chandan.babu@...cle.com, akpm@...ux-foundation.org,
mcgrof@...nel.org, ziy@...dia.com, hare@...e.de, djwong@...nel.org,
gost.dev@...sung.com, linux-mm@...ck.org,
Pankaj Raghav <p.raghav@...sung.com>
Subject: Re: [PATCH 05/13] readahead: set file_ra_state->ra_pages to be at
least mapping_min_order
On Mon, Feb 26, 2024 at 10:49:28AM +0100, Pankaj Raghav (Samsung) wrote:
> From: Luis Chamberlain <mcgrof@...nel.org>
>
> Set the file_ra_state->ra_pages in file_ra_state_init() to be at least
> mapping_min_order of pages if the bdi->ra_pages is less than that.
Don't we rather want to round up to a multiple of mapping_min_nrpages?
> file_ra_state_init(struct file_ra_state *ra, struct address_space *mapping)
> {
> + unsigned int min_nrpages = mapping_min_folio_nrpages(mapping);
> +
> ra->ra_pages = inode_to_bdi(mapping->host)->ra_pages;
> + if (ra->ra_pages < min_nrpages)
> + ra->ra_pages = min_nrpages;
> ra->prev_pos = -1;
Powered by blists - more mailing lists