Re: [PATCH 1/5] zonefs: pass GFP_KERNEL to blkdev_zone_mgmt() call

From: Dave Chinner <david@fromorbit.com>
To: Mikulas Patocka <mpatocka@redhat.com>
Cc: Johannes Thumshirn <johannes.thumshirn@wdc.com>,
	Damien Le Moal <dlemoal@kernel.org>,
	Naohiro Aota <naohiro.aota@wdc.com>,
	Mike Snitzer <snitzer@kernel.org>,
	dm-devel@lists.linux.dev, Chris Mason <clm@fb.com>,
	Josef Bacik <josef@toxicpanda.com>,
	David Sterba <dsterba@suse.com>, Jaegeuk Kim <jaegeuk@kernel.org>,
	Chao Yu <chao@kernel.org>, Jens Axboe <axboe@kernel.dk>,
	Christoph Hellwig <hch@lst.de>, Sagi Grimberg <sagi@grimberg.me>,
	Chaitanya Kulkarni <kch@nvidia.com>,
	linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
	linux-btrfs@vger.kernel.org,
	linux-f2fs-devel@lists.sourceforge.net,
	linux-block@vger.kernel.org, linux-nvme@lists.infradead.org,
	"Matthew Wilcox (Oracle)" <willy@infradead.org>
Subject: Re: [PATCH 1/5] zonefs: pass GFP_KERNEL to blkdev_zone_mgmt() call
Date: Wed, 24 Jan 2024 10:21:47 +1100	[thread overview]
Message-ID: <ZbBKC3U3/1yPvWDR@dread.disaster.area> (raw)
In-Reply-To: <31e0f796-1c5-b7f8-2f4b-d937770e8d5@redhat.com>

On Tue, Jan 23, 2024 at 09:39:02PM +0100, Mikulas Patocka wrote:
> 
> 
> On Tue, 23 Jan 2024, Johannes Thumshirn wrote:
> 
> > Pass GFP_KERNEL instead of GFP_NOFS to the blkdev_zone_mgmt() call in
> > zonefs_zone_mgmt().
> > 
> > As as zonefs_zone_mgmt() and zonefs_inode_zone_mgmt() are never called
> > from a place that can recurse back into the filesystem on memory reclaim,
> > it is save to call blkdev_zone_mgmt() with GFP_KERNEL.
> > 
> > Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
> > ---
> >  fs/zonefs/super.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/fs/zonefs/super.c b/fs/zonefs/super.c
> > index 93971742613a..63fbac018c04 100644
> > --- a/fs/zonefs/super.c
> > +++ b/fs/zonefs/super.c
> > @@ -113,7 +113,7 @@ static int zonefs_zone_mgmt(struct super_block *sb,
> >  
> >  	trace_zonefs_zone_mgmt(sb, z, op);
> >  	ret = blkdev_zone_mgmt(sb->s_bdev, op, z->z_sector,
> > -			       z->z_size >> SECTOR_SHIFT, GFP_NOFS);
> > +			       z->z_size >> SECTOR_SHIFT, GFP_KERNEL);
> >  	if (ret) {
> >  		zonefs_err(sb,
> >  			   "Zone management operation %s at %llu failed %d\n",
> > 
> > -- 
> > 2.43.0
> 
> zonefs_inode_zone_mgmt calls 
> lockdep_assert_held(&ZONEFS_I(inode)->i_truncate_mutex); - so, this 
> function is called with the mutex held - could it happen that the 
> GFP_KERNEL allocation recurses into the filesystem and attempts to take 
> i_truncate_mutex as well?
> 
> i.e. GFP_KERNEL -> iomap_do_writepage -> zonefs_write_map_blocks -> 
> zonefs_write_iomap_begin -> mutex_lock(&zi->i_truncate_mutex)

zonefs doesn't have a ->writepage method, so writeback can't be
called from memory reclaim like this.

-Dave.
-- 
Dave Chinner
david@fromorbit.com