linux-stable.git/drivers/mtd/ubi, branch v5.4.232

ubi: ubi_create_volume: Fix use-after-free when volume creation failed

2022-06-14T16:11:55+00:00

[ Upstream commit 8c03a1c21d72210f81cb369cc528e3fde4b45411 ]

There is an use-after-free problem for 'eba_tbl' in ubi_create_volume()'s
error handling path:

  ubi_eba_replace_table(vol, eba_tbl)
    vol->eba_tbl = tbl
out_mapping:
  ubi_eba_destroy_table(eba_tbl)   // Free 'eba_tbl'
out_unlock:
  put_device(&vol->dev)
    vol_release
      kfree(tbl->entries)	  // UAF

Fix it by removing redundant 'eba_tbl' releasing.
Fetch a reproducer in [Link].

Fixes: 493cfaeaa0c9b ("mtd: utilize new cdev_device_add helper function")
Link: https://bugzilla.kernel.org/show_bug.cgi?id=215965
Signed-off-by: Zhihao Cheng 
Signed-off-by: Richard Weinberger 
Signed-off-by: Sasha Levin

ubi: fastmap: Return error code if memory allocation fails in add_aeb()

2022-04-15T12:18:30+00:00

commit c3c07fc25f37c157fde041b3a0c3dfcb1590cbce upstream.

Abort fastmap scanning and return error code if memory allocation fails
in add_aeb(). Otherwise ubi will get wrong peb statistics information
after scanning.

Fixes: dbb7d2a88d2a7b ("UBI: Add fastmap core")
Signed-off-by: Zhihao Cheng 
Signed-off-by: Richard Weinberger 
Signed-off-by: Greg Kroah-Hartman

ubi: Fix race condition between ctrl_cdev_ioctl and ubi_cdev_ioctl

2022-04-15T12:18:29+00:00

commit 3cbf0e392f173ba0ce425968c8374a6aa3e90f2e upstream.

Hulk Robot reported a KASAN report about use-after-free:
 ==================================================================
 BUG: KASAN: use-after-free in __list_del_entry_valid+0x13d/0x160
 Read of size 8 at addr ffff888035e37d98 by task ubiattach/1385
 [...]
 Call Trace:
  klist_dec_and_del+0xa7/0x4a0
  klist_put+0xc7/0x1a0
  device_del+0x4d4/0xed0
  cdev_device_del+0x1a/0x80
  ubi_attach_mtd_dev+0x2951/0x34b0 [ubi]
  ctrl_cdev_ioctl+0x286/0x2f0 [ubi]

 Allocated by task 1414:
  device_add+0x60a/0x18b0
  cdev_device_add+0x103/0x170
  ubi_create_volume+0x1118/0x1a10 [ubi]
  ubi_cdev_ioctl+0xb7f/0x1ba0 [ubi]

 Freed by task 1385:
  cdev_device_del+0x1a/0x80
  ubi_remove_volume+0x438/0x6c0 [ubi]
  ubi_cdev_ioctl+0xbf4/0x1ba0 [ubi]
 [...]
 ==================================================================

The lock held by ctrl_cdev_ioctl is ubi_devices_mutex, but the lock held
by ubi_cdev_ioctl is ubi->device_mutex. Therefore, the two locks can be
concurrent.

ctrl_cdev_ioctl contains two operations: ubi_attach and ubi_detach.
ubi_detach is bug-free because it uses reference counting to prevent
concurrency. However, uif_init and uif_close in ubi_attach may race with
ubi_cdev_ioctl.

uif_init will race with ubi_cdev_ioctl as in the following stack.
           cpu1                   cpu2                  cpu3
_______________________|________________________|______________________
ctrl_cdev_ioctl
 ubi_attach_mtd_dev
  uif_init
                           ubi_cdev_ioctl
                            ubi_create_volume
                             cdev_device_add
   ubi_add_volume
   // sysfs exist
   kill_volumes
                                                    ubi_cdev_ioctl
                                                     ubi_remove_volume
                                                      cdev_device_del
                                                       // first free
    ubi_free_volume
     cdev_del
     // double free
   cdev_device_del

And uif_close will race with ubi_cdev_ioctl as in the following stack.
           cpu1                   cpu2                  cpu3
_______________________|________________________|______________________
ctrl_cdev_ioctl
 ubi_attach_mtd_dev
  uif_init
                           ubi_cdev_ioctl
                            ubi_create_volume
                             cdev_device_add
  ubi_debugfs_init_dev
  //error goto out_uif;
  uif_close
   kill_volumes
                                                    ubi_cdev_ioctl
                                                     ubi_remove_volume
                                                      cdev_device_del
                                                       // first free
    ubi_free_volume
    // double free

The cause of this problem is that commit 714fb87e8bc0 make device
"available" before it becomes accessible via sysfs. Therefore, we
roll back the modification. We will fix the race condition between
ubi device creation and udev by removing ubi_get_device in
vol_attribute_show and dev_attribute_show.This avoids accessing
uninitialized ubi_devices[ubi_num].

ubi_get_device is used to prevent devices from being deleted during
sysfs execution. However, now kernfs ensures that devices will not
be deleted before all reference counting are released.
The key process is shown in the following stack.

device_del
  device_remove_attrs
    device_remove_groups
      sysfs_remove_groups
        sysfs_remove_group
          remove_files
            kernfs_remove_by_name
              kernfs_remove_by_name_ns
                __kernfs_remove
                  kernfs_drain

Fixes: 714fb87e8bc0 ("ubi: Fix race condition between ubi device creation and udev")
Reported-by: Hulk Robot 
Signed-off-by: Baokun Li 
Signed-off-by: Richard Weinberger 
Signed-off-by: Greg Kroah-Hartman

ubi: check kthread_should_stop() after the setting of task state

2020-11-05T10:43:33+00:00

commit d005f8c6588efcfbe88099b6edafc6f58c84a9c1 upstream.

A detach hung is possible when a race occurs between the detach process
and the ubi background thread. The following sequences outline the race:

  ubi thread: if (list_empty(&ubi->works)...

  ubi detach: set_bit(KTHREAD_SHOULD_STOP, &kthread->flags)
              => by kthread_stop()
              wake_up_process()
              => ubi thread is still running, so 0 is returned

  ubi thread: set_current_state(TASK_INTERRUPTIBLE)
              schedule()
              => ubi thread will never be scheduled again

  ubi detach: wait_for_completion()
              => hung task!

To fix that, we need to check kthread_should_stop() after we set the
task state, so the ubi thread will either see the stop bit and exit or
the task state is reset to runnable such that it isn't scheduled out
indefinitely.

Signed-off-by: Zhihao Cheng 
Cc: 
Fixes: 801c135ce73d5df1ca ("UBI: Unsorted Block Images")
Reported-by: syzbot+853639d0cb16c31c7a14@syzkaller.appspotmail.com
Signed-off-by: Richard Weinberger 
Signed-off-by: Greg Kroah-Hartman

ubi: fastmap: Free unused fastmap anchor peb during detach

2020-10-01T11:18:11+00:00

[ Upstream commit c16f39d14a7e0ec59881fbdb22ae494907534384 ]

When CONFIG_MTD_UBI_FASTMAP is enabled, fm_anchor will be assigned
a free PEB during ubi_wl_init() or ubi_update_fastmap(). However
if fastmap is not used or disabled on the MTD device, ubi_wl_entry
related with the PEB will not be freed during detach.

So Fix it by freeing the unused fastmap anchor during detach.

Fixes: f9c34bb52997 ("ubi: Fix producing anchor PEBs")
Reported-by: syzbot+f317896aae32eb281a58@syzkaller.appspotmail.com
Reviewed-by: Sascha Hauer 
Signed-off-by: Hou Tao 
Signed-off-by: Richard Weinberger 
Signed-off-by: Sasha Levin

ubi: Fix producing anchor PEBs

2020-10-01T11:17:19+00:00

[ Upstream commit f9c34bb529975fe9f85b870a80c53a83a3c5a182 ]

When a new fastmap is about to be written UBI must make sure it has a
free block for a fastmap anchor available. For this ubi_update_fastmap()
calls ubi_ensure_anchor_pebs(). This stopped working with 2e8f08deabbc
("ubi: Fix races around ubi_refill_pools()"), with this commit the wear
leveling code is blocked and can no longer produce free PEBs. UBI then
more often than not falls back to write the new fastmap anchor to the
same block it was already on which means the same erase block gets
erased during each fastmap write and wears out quite fast.

As the locking prevents us from producing the anchor PEB when we
actually need it, this patch changes the strategy for creating the
anchor PEB. We no longer create it on demand right before we want to
write a fastmap, but instead we create an anchor PEB right after we have
written a fastmap. This gives us enough time to produce a new anchor PEB
before it is needed. To make sure we have an anchor PEB for the very
first fastmap write we call ubi_ensure_anchor_pebs() during
initialisation as well.

Fixes: 2e8f08deabbc ("ubi: Fix races around ubi_refill_pools()")
Signed-off-by: Sascha Hauer 
Signed-off-by: Richard Weinberger 
Signed-off-by: Sasha Levin

ubi: Fix seq_file usage in detailed_erase_block_info debugfs file

2020-05-27T15:46:19+00:00

[ Upstream commit 0e7572cffe442290c347e779bf8bd4306bb0aa7c ]

3bfa7e141b0b ("fs/seq_file.c: seq_read(): add info message about buggy .next functions")
showed that we don't use seq_file correctly.
So make sure that our ->next function always updates the position.

Fixes: 7bccd12d27b7 ("ubi: Add debugfs file for tracking PEB state")
Signed-off-by: Richard Weinberger 
Signed-off-by: Sasha Levin

ubi: Fix an error pointer dereference in error handling code

2020-02-11T12:35:47+00:00

commit 5d3805af279c93ef49a64701f35254676d709622 upstream.

If "seen_pebs = init_seen(ubi);" fails then "seen_pebs" is an error pointer
and we try to kfree() it which results in an Oops.

This patch re-arranges the error handling so now it only frees things
which have been allocated successfully.

Fixes: daef3dd1f0ae ("UBI: Fastmap: Add self check to detect absent PEBs")
Signed-off-by: Dan Carpenter 
Signed-off-by: Richard Weinberger 
Signed-off-by: Greg Kroah-Hartman

ubi: fastmap: Fix inverted logic in seen selfcheck

2020-02-11T12:35:47+00:00

commit ef5aafb6e4e9942a28cd300bdcda21ce6cbaf045 upstream.

set_seen() sets the bit corresponding to the PEB number in the bitmap,
so when self_check_seen() wants to find PEBs that haven't been seen we
have to print the PEBs that have their bit cleared, not the ones which
have it set.

Fixes: 5d71afb00840 ("ubi: Use bitmaps in Fastmap self-check code")
Signed-off-by: Sascha Hauer 
Signed-off-by: Richard Weinberger 
Signed-off-by: Greg Kroah-Hartman

ubi: block: Warn if volume size is not multiple of 512

2019-09-15T20:12:49+00:00

If volume size is not a multiple of 512, ubi block cuts
off the last bytes of an volume since the block layer works
on 512 byte sectors.
This can happen especially on NOR flash with minimal io
size of 1.

To avoid unpleasant surprises, print a warning.

Signed-off-by: Richard Weinberger