@@ -1110,15 +1110,19 @@ static int loop_configure(struct loop_device *lo, fmode_t mode,
clear_bit(GD_SUPPRESS_PART_SCAN, &lo->lo_disk->state);
loop_global_unlock(lo, is_loop);
- if (partscan)
- loop_reread_partitions(lo);
if (!(mode & FMODE_EXCL))
bd_abort_claiming(bdev, loop_configure);
+ /*
+ * Now that we are done, reread the partitions with uevent
+ * re-enabled to let userspace know about the changes.
+ */
+ dev_set_uevent_suppress(disk_to_dev(lo->lo_disk), 0);
+ if (partscan)
+ loop_reread_partitions(lo);
+
error = 0;
done:
- /* enable and uncork uevent now that we are done */
- dev_set_uevent_suppress(disk_to_dev(lo->lo_disk), 0);
return error;
out_unlock:
@@ -1130,6 +1134,7 @@ static int loop_configure(struct loop_device *lo, fmode_t mode,
fput(file);
/* This is safe: open() is still holding a reference. */
module_put(THIS_MODULE);
+ dev_set_uevent_suppress(disk_to_dev(lo->lo_disk), 0);
goto done;
}
LOOP_CONFIGURE is, as I understand it, supposed to be a way to combine LOOP_SET_FD and LOOP_SET_STATUS64 into a single syscall. When using LOOP_SET_FD+LOOP_SET_STATUS64, a single uevent would be sent for each partition found on the loop device during LOOP_SET_STATUS64, but when using LOOP_CONFIGURE, no such uevent was being sent. In the old setup, uevents are disabled for LOOP_SET_FD, but not for LOOP_SET_STATUS64. This makes sense, as it prevents uevents being sent for a partially configured device during LOOP_SET_FD — they're only sent at the end of LOOP_SET_STATUS64. But for LOOP_CONFIGURE, uevents were disabled for the entire operation, so that final notification was never issued. To fix this, I've moved the loop_reread_partitions() call, which causes the uevents to be issued, to after uevents are re-enabled, matching the behaviour of the LOOP_SET_FD+LOOP_SET_STATUS64 combination. I noticed this because Busybox's losetup program recently changed from using LOOP_SET_FD+LOOP_SET_STATUS64 to LOOP_CONFIGURE, and this broke my setup, for which I want a notification from the kernel any time a new partition becomes available. Signed-off-by: Alyssa Ross <hi@alyssa.is> Fixes: 3448914e8cc5 ("loop: Add LOOP_CONFIGURE ioctl") --- I've marked this as an RFC because there's still a problem with this patch that I'd be grateful for advice on how to solve: this change accidentally makes LOOP_SET_FD emit uevents as well if max_part is configured. There are a few ways I could imagine resolving this: - Have loop_configure distinguish between LOOP_SET_FD and LOOP_CONFIGURE somehow. - Have loop_configure take a bool argument specifying whether uevents should be reenabled before or after the loop_reread_partitions() call. - Move re-enabling the uevent and calling loop_reread_partitions() out of loop_configure(). All of these have drawbacks for the understandability of the code though, so I wanted to ask what the best way to proceed would be. drivers/block/loop.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-)