Re: [PATCH] mm: Use GFP_KERNEL allocation for the page cache in page_cache_read

From: Michal Hocko <mhocko@suse.cz>
To: Rik van Riel <riel@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
	Al Viro <viro@zeniv.linux.org.uk>,
	Johannes Weiner <hannes@cmpxchg.org>,
	Mel Gorman <mgorman@suse.de>, Neil Brown <neilb@suse.de>,
	Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>,
	Sage Weil <sage@inktank.com>, Mark Fasheh <mfasheh@suse.com>,
	linux-mm@kvack.org, LKML <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] mm: Use GFP_KERNEL allocation for the page cache in page_cache_read
Date: Wed, 18 Mar 2015 15:37:33 +0100	[thread overview]
Message-ID: <20150318143733.GI17241@dhcp22.suse.cz> (raw)
In-Reply-To: <55098C99.9040104@redhat.com>

On Wed 18-03-15 10:32:57, Rik van Riel wrote:
> On 03/18/2015 10:09 AM, Michal Hocko wrote:
> 
> > diff --git a/mm/filemap.c b/mm/filemap.c
> > index 968cd8e03d2e..26f62ba79f50 100644
> > --- a/mm/filemap.c
> > +++ b/mm/filemap.c
> > @@ -1752,7 +1752,7 @@ static int page_cache_read(struct file *file, pgoff_t offset)
> >  	int ret;
> >  
> >  	do {
> > -		page = page_cache_alloc_cold(mapping);
> > +		page = __page_cache_alloc(GFP_KERNEL|__GFP_COLD);
> >  		if (!page)
> >  			return -ENOMEM;
> 
> Won't this break on highmem systems, by failing to
> allocate the page cache from highmem, where previously
> it would?

It will! This is broken. I can see inode_init_always now. We need to add
GFP_HIGHUSER_MOVABLE here.

Thanks for pointing this out!

-- 
Michal Hocko
SUSE Labs