The condition detecting whether somebody else has the device exclusively open in disk_scan_partitions() has a brownpaper bag bug. It triggers also when nobody has the device exclusively open and we are coming from BLKRRPART path. Interestingly this didn't have any adverse effects during testing because tools update kernel's notion of the partition table using ioctls and don't rely on BLKRRPART. Fix the bug before somebody trips over it.
Fixes: 8d67fc20caf8 ("block: Do not reread partition table on exclusively open device") CC: stable@vger.kernel.org Signed-off-by: Jan Kara jack@suse.cz --- block/genhd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/block/genhd.c b/block/genhd.c index 012529d36f5b..29fb2c98b401 100644 --- a/block/genhd.c +++ b/block/genhd.c @@ -367,7 +367,7 @@ int disk_scan_partitions(struct gendisk *disk, fmode_t mode, void *owner) if (disk->open_partitions) return -EBUSY; /* Someone else has bdev exclusively open? */ - if (disk->part0->bd_holder != owner) + if (disk->part0->bd_holder && disk->part0->bd_holder != owner) return -EBUSY;
set_bit(GD_NEED_PART_SCAN, &disk->state);
On 12/1/22 4:44 AM, Jan Kara wrote:
The condition detecting whether somebody else has the device exclusively open in disk_scan_partitions() has a brownpaper bag bug. It triggers also when nobody has the device exclusively open and we are coming from BLKRRPART path. Interestingly this didn't have any adverse effects during testing because tools update kernel's notion of the partition table using ioctls and don't rely on BLKRRPART. Fix the bug before somebody trips over it.
I folded this in. It's funny because I did consider this when reviewing it, but then convinced myself it was fine...
linux-stable-mirror@lists.linaro.org