Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Tue, 18 Dec 2018 14:41:39 -0800
From:   Jaegeuk Kim <jaegeuk@kernel.org>
To:     Jens Axboe <axboe@kernel.dk>
Cc:     linux-kernel@vger.kernel.org, linux-block@vger.kernel.org
Subject: Re: [PATCH v3] loop: drop caches if offset or block_size are changed
Message-ID: <20181218224139.GB31062@jaegeuk-macbookpro.roam.corp.google.com>
References: <20181214203223.7063-1-jaegeuk@kernel.org>
 <20181217194236.GA50659@jaegeuk-macbookpro.roam.corp.google.com>
 <29369548-df14-a5a7-2bee-a40b3479df68@kernel.dk>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <29369548-df14-a5a7-2bee-a40b3479df68@kernel.dk>
User-Agent: Mutt/1.8.2 (2017-04-18)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

If we don't drop caches used in old offset or block_size, we can get old data
from new offset/block_size, which gives unexpected data to user.

For example, Martijn found a loopback bug in the below scenario.
1) LOOP_SET_FD loads first two pages on loop file
2) LOOP_SET_STATUS64 changes the offset on the loop file
3) mount is failed due to the cached pages having wrong superblock

Cc: Jens Axboe <axboe@kernel.dk>
Cc: linux-block@vger.kernel.org
Reported-by: Martijn Coenen <maco@google.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
---
v2 -> v3:
 - avoid to submit IOs on frozen mq

Jens, how about this?

Thanks,

 drivers/block/loop.c | 29 +++++++++++++++++++++++++++--
 1 file changed, 27 insertions(+), 2 deletions(-)

diff --git a/drivers/block/loop.c b/drivers/block/loop.c
index cb0cc8685076..6b03121d62aa 100644
--- a/drivers/block/loop.c
+++ b/drivers/block/loop.c
@@ -1126,6 +1126,12 @@ loop_set_status(struct loop_device *lo, const struct loop_info64 *info)
 	if ((unsigned int) info->lo_encrypt_key_size > LO_KEY_SIZE)
 		return -EINVAL;
 
+	if (lo->lo_offset != info->lo_offset ||
+	    lo->lo_sizelimit != info->lo_sizelimit) {
+		sync_blockdev(lo->lo_device);
+		kill_bdev(lo->lo_device);
+	}
+
 	/* I/O need to be drained during transfer transition */
 	blk_mq_freeze_queue(lo->lo_queue);
 
@@ -1154,6 +1160,11 @@ loop_set_status(struct loop_device *lo, const struct loop_info64 *info)
 
 	if (lo->lo_offset != info->lo_offset ||
 	    lo->lo_sizelimit != info->lo_sizelimit) {
+		/* kill_bdev should have truncated all the pages */
+		if (lo->lo_device->bd_inode->i_mapping->nrpages) {
+			err = -EAGAIN;
+			goto exit;
+		}
 		if (figure_loop_size(lo, info->lo_offset, info->lo_sizelimit)) {
 			err = -EFBIG;
 			goto exit;
@@ -1375,22 +1386,36 @@ static int loop_set_dio(struct loop_device *lo, unsigned long arg)
 
 static int loop_set_block_size(struct loop_device *lo, unsigned long arg)
 {
+	int err = 0;
+
 	if (lo->lo_state != Lo_bound)
 		return -ENXIO;
 
 	if (arg < 512 || arg > PAGE_SIZE || !is_power_of_2(arg))
 		return -EINVAL;
 
+	if (lo->lo_queue->limits.logical_block_size != arg) {
+		sync_blockdev(lo->lo_device);
+		kill_bdev(lo->lo_device);
+	}
+
 	blk_mq_freeze_queue(lo->lo_queue);
 
+	/* kill_bdev should have truncated all the pages */
+	if (lo->lo_queue->limits.logical_block_size != arg &&
+			lo->lo_device->bd_inode->i_mapping->nrpages) {
+		err = -EAGAIN;
+		goto out;
+	}
+
 	blk_queue_logical_block_size(lo->lo_queue, arg);
 	blk_queue_physical_block_size(lo->lo_queue, arg);
 	blk_queue_io_min(lo->lo_queue, arg);
 	loop_update_dio(lo);
-
+out:
 	blk_mq_unfreeze_queue(lo->lo_queue);
 
-	return 0;
+	return err;
 }
 
 static int lo_ioctl(struct block_device *bdev, fmode_t mode,
-- 
2.19.0.605.g01d371f741-goog