On Tue, Nov 12, 2019 at 09:22:51AM -0800, Evan Green wrote:
> Thanks for replying and taking a look Darrick. I didn't see your patch
> in Jens tree when I looked just before sending it, but maybe I missed
> it.
>
> On Mon, Nov 11, 2019 at 5:37 PM Darrick J. Wong <darrick.wong@xxxxxxxxxx> wrote:
> >
> > On Mon, Nov 11, 2019 at 10:50:30AM -0800, Evan Green wrote:
> > > If the backing device for a loop device is a block device,
> > > then mirror the "write zeroes" capabilities of the underlying
> > > block device into the loop device. Copy this capability into both
> > > max_write_zeroes_sectors and max_discard_sectors of the loop device.
> > >
> > > The reason for this is that REQ_OP_DISCARD on a loop device translates
> > > into blkdev_issue_zeroout(), rather than blkdev_issue_discard(). This
> > > presents a consistent interface for loop devices (that discarded data
> > > is zeroed), regardless of the backing device type of the loop device.
> > > There should be no behavior change for loop devices backed by regular
> > > files.
> > >
> > > While in there, differentiate between REQ_OP_DISCARD and
> > > REQ_OP_WRITE_ZEROES, which are different for block devices,
> > > but which the loop device had just been lumping together, since
> > > they're largely the same for files.
> > >
> > > This change fixes blktest block/003, and removes an extraneous
> > > error print in block/013 when testing on a loop device backed
> > > by a block device that does not support discard.
> > >
> > > Signed-off-by: Evan Green <evgreen@xxxxxxxxxxxx>
> > > Reviewed-by: Gwendal Grignou <gwendal@xxxxxxxxxxxx>
> > > Reviewed-by: Chaitanya Kulkarni <chaitanya.kulkarni@xxxxxxx>
> > > ---
> > >
> > > Changes in v6: None
> > > Changes in v5:
> > > - Don't mirror discard if lo_encrypt_key_size is non-zero (Gwendal)
> > >
> > > Changes in v4:
> > > - Mirror blkdev's write_zeroes into loopdev's discard_sectors.
> > >
> > > Changes in v3:
> > > - Updated commit description
> > >
> > > Changes in v2: None
> > >
> > > drivers/block/loop.c | 57 ++++++++++++++++++++++++++++----------------
> > > 1 file changed, 37 insertions(+), 20 deletions(-)
> > >
> > > diff --git a/drivers/block/loop.c b/drivers/block/loop.c
> > > index d749156a3d88..236f6deb0772 100644
> > > --- a/drivers/block/loop.c
> > > +++ b/drivers/block/loop.c
> > > @@ -417,19 +417,14 @@ static int lo_read_transfer(struct loop_device *lo, struct request *rq,
> > > return ret;
> > > }
> > >
> > > -static int lo_discard(struct loop_device *lo, struct request *rq, loff_t pos)
> > > +static int lo_discard(struct loop_device *lo, struct request *rq,
> > > + int mode, loff_t pos)
> > > {
> > > - /*
> > > - * We use punch hole to reclaim the free space used by the
> > > - * image a.k.a. discard. However we do not support discard if
> > > - * encryption is enabled, because it may give an attacker
> > > - * useful information.
> > > - */
> > > struct file *file = lo->lo_backing_file;
> > > - int mode = FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE;
> > > + struct request_queue *q = lo->lo_queue;
> > > int ret;
> > >
> > > - if ((!file->f_op->fallocate) || lo->lo_encrypt_key_size) {
> > > + if (!blk_queue_discard(q)) {
> > > ret = -EOPNOTSUPP;
> > > goto out;
> > > }
> > > @@ -599,8 +594,13 @@ static int do_req_filebacked(struct loop_device *lo, struct request *rq)
> > > case REQ_OP_FLUSH:
> > > return lo_req_flush(lo, rq);
> > > case REQ_OP_DISCARD:
> > > + return lo_discard(lo, rq,
> > > + FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE, pos);
> > > +
> > > case REQ_OP_WRITE_ZEROES:
> > > - return lo_discard(lo, rq, pos);
> > > + return lo_discard(lo, rq,
> > > + FALLOC_FL_ZERO_RANGE | FALLOC_FL_KEEP_SIZE, pos);
> >
> > Yes, this more or less reimplements what's already in -next...
>
> Agree, this part would disappear if I rebased on top of your patch.
> This series has been around for awhile, you see :)
Oh. Didn't quite realize that. :/
> > > +
> > > case REQ_OP_WRITE:
> > > if (lo->transfer)
> > > return lo_write_transfer(lo, rq, pos);
> > > @@ -854,6 +854,21 @@ static void loop_config_discard(struct loop_device *lo)
> > > struct file *file = lo->lo_backing_file;
> > > struct inode *inode = file->f_mapping->host;
> > > struct request_queue *q = lo->lo_queue;
> > > + struct request_queue *backingq;
> > > +
> > > + /*
> > > + * If the backing device is a block device, mirror its zeroing
> > > + * capability. REQ_OP_DISCARD translates to a zero-out even when backed
> > > + * by block devices to keep consistent behavior with file-backed loop
> > > + * devices.
> > > + */
> > > + if (S_ISBLK(inode->i_mode) && !lo->lo_encrypt_key_size) {
> > > + backingq = bdev_get_queue(inode->i_bdev);
> >
> > What happens if the inode is from a filesystem that can have multiple
> > backing devices (like btrfs)?
>
> Then I would expect S_ISBLK(inode->i_mode) would not be true. This is
> only for when you've created a loop device directly on top of a block
> device (ie you pointed the loop device at /dev/sda). We use this in
> our Chrome OS installer because it makes the logic simple whether
> you're installing to a real disk or a file image.
Heh, doh, that's right. :)
Sorry, for some reason I misread that as "If the backing device of the
filesystem from which the inode came is a block device..."
Might I suggest rewording the first sentence of the comment to read "If
the loop device's backing device is itself a block device" for oafs like
me? :)
--D
> >
> > > + blk_queue_max_discard_sectors(q,
> > > + backingq->limits.max_write_zeroes_sectors);
> > > +
> > > + blk_queue_max_write_zeroes_sectors(q,
> > > + backingq->limits.max_write_zeroes_sectors);
> >
> > Also, seeing as filesystems tend to implement PUNCH_HOLE and ZERO_RANGE
> > on their own independent of the hardware capabilities of the underlying
> > device, it doesn't make much sense to forward the blockdev limits to the
> > loop device.
> >
> > (Put another way, XFS's ZERO_RANGE implementation can zero hundreds of
> > gigabytes at a time even if the underlying device is a spinning rust.)
>
> Hopefully my comment above addresses this too (there is no file system
> in the scenario I'm coding for).
