Re: [PATCH] ovl: use a dedicated semaphore for dir upperfile caching

[Icenowy Zheng]

于 2021年1月4日 GMT+08:00 下午4:35:20, Amir Goldstein <amir73il@xxxxxxxxx> 写到:
>On Mon, Jan 4, 2021 at 9:28 AM Icenowy Zheng <icenowy@xxxxxxx> wrote:
>>
>> 在 2021-01-03星期日的 16:10 +0200，Amir Goldstein写道：
>> > On Fri, Jan 1, 2021 at 10:12 PM Icenowy Zheng <icenowy@xxxxxxx>
>> > wrote:
>> > >
>> > > The function ovl_dir_real_file() currently uses the semaphore of
>> > > the
>> > > inode to synchronize write to the upperfile cache field.
>> > >
>> > > However, this function will get called by ovl_ioctl_set_flags(),
>> > > which
>> > > utilizes the inode semaphore too. In this case
>ovl_dir_real_file()
>> > > will
>> > > try to claim a lock that is owned by a function in its call
>stack,
>> > > which
>> > > won't get released before ovl_dir_real_file() returns.
>> >
>> > oops. I wondered why I didn't see any warnings on this from
>lockdep.
>> > Ah! because the xfstest that exercises ovl_ioctl_set_flags() on
>> > directory,
>> > generic/079, starts with an already upper dir.
>> >
>> > And the xfstest that checks chattr+i on lower/upper files,
>> > overlay/040,
>> > does not check chattr on dirs (ioctl on overlay dirs wasn't
>supported
>> > at
>> > the time the test was written).
>> >
>> > Would you be able to create a variant of test overlay/040 that also
>> > tests
>> > chattr +i on lower/upper dirs to test your patch and confirm that
>the
>> > test
>> > fails on master with the appropriate Kconfig debug options.
>>
>> https://gist.github.com/Icenowy/c7d8decb6812d6e5064d143c57281ad3
>>
>> Here's a test that would break on master (I used linux-next/master
>for
>> test).
>
>Thanks.
>I am working on another test to improve overlay/030 that may also
>cover this bug, so maybe no need for both tests. I'll let you know when
>I'm done.
>If you like, I can post your test for you with your Signed-of-by if I
>think
>it is also needed.
>
>>
>> [  246.521880] INFO: task chattr:715 blocked for more than 122
>seconds.
>> [  246.525659]       Not tainted 5.11.0-rc1-next-20210104+ #20
>> [  246.528498] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs"
>> disables this message.
>> [  246.535076] task:chattr          state:D stack:13736 pid:  715
>ppid:
>> 529 flags:0x00000000
>> [  246.538923] Call Trace:
>> [  246.540241]  __schedule+0x2a9/0x820
>> [  246.541986]  schedule+0x56/0xc0
>> [  246.543616]  rwsem_down_write_slowpath+0x375/0x630
>> [  246.545565]  ovl_dir_real_file+0xc1/0x120
>> [  246.547512]  ovl_real_fdget+0x35/0x80
>> [  246.549303]  ovl_real_ioctl+0x26/0x90
>> [  246.551050]  ? mnt_drop_write+0x2c/0x70
>> [  246.553068]  ovl_ioctl_set_flags+0x93/0x110
>> [  246.555407]  __x64_sys_ioctl+0x7e/0xb0
>> [  246.557175]  do_syscall_64+0x33/0x40
>> [  246.558869]  entry_SYSCALL_64_after_hwframe+0x44/0xa9
>> [  246.561057] RIP: 0033:0x7fe4a3830b67
>> [  246.565799] RSP: 002b:00007ffe7ad504f8 EFLAGS: 00000246 ORIG_RAX:
>> 0000000000000010
>> [  246.569438] RAX: ffffffffffffffda RBX: 0000000000000001 RCX:
>> 00007fe4a3830b67
>> [  246.572061] RDX: 00007ffe7ad5050c RSI: 0000000040086602 RDI:
>> 0000000000000003
>> [  246.575509] RBP: 0000000000000003 R08: 0000000000000001 R09:
>> 0000000000000000
>> [  246.578932] R10: 0000000000000000 R11: 0000000000000246 R12:
>> 0000000000000010
>> [  246.581014] R13: 00007ffe7ad50810 R14: 0000000000000002 R15:
>> 0000000000000001
>> [  246.582818]
>> [  246.582818] Showing all locks held in the system:
>> [  246.584741] 1 lock held by khungtaskd/18:
>> [  246.586085]  #0: ffffffff9e951540 (rcu_read_lock){....}-{1:2}, at:
>> debug_show_all_locks+0x15/0x100
>> [  246.589775] 3 locks held by chattr/715:
>> [  246.591364]  #0: ffff96a74b92c450 (sb_writers#11){....}-{0:0}, at:
>> ovl_ioctl_set_flags+0x2f/0x110
>> [  246.597182]  #1: ffff96a7489c3500
>> (&ovl_i_mutex_dir_key[depth]){....}-{3:3}, at:
>> ovl_ioctl_set_flags+0x54/0x110
>> [  246.601325]  #2: ffff96a7489c3500
>> (&ovl_i_mutex_dir_key[depth]){....}-{3:3}, at:
>> ovl_dir_real_file+0xc1/0x120
>>
>> >
>> > >
>> > > Define a dedicated semaphore for the upperfile cache, so that the
>> > > deadlock won't happen.
>> > >
>> > > Fixes: 61536bed2149 ("ovl: support [S|G]ETFLAGS and
>FS[S|G]ETXATTR
>> > > ioctls for directories")
>> > > Cc: stable@xxxxxxxxxxxxxxx # v5.10
>> > > Signed-off-by: Icenowy Zheng <icenowy@xxxxxxx>
>> > > ---
>> > >  fs/overlayfs/readdir.c | 6 ++++--
>> > >  1 file changed, 4 insertions(+), 2 deletions(-)
>> > >
>> > > diff --git a/fs/overlayfs/readdir.c b/fs/overlayfs/readdir.c
>> > > index 01620ebae1bd..f10701aabb71 100644
>> > > --- a/fs/overlayfs/readdir.c
>> > > +++ b/fs/overlayfs/readdir.c
>> > > @@ -56,6 +56,7 @@ struct ovl_dir_file {
>> > >         struct list_head *cursor;
>> > >         struct file *realfile;
>> > >         struct file *upperfile;
>> > > +       struct semaphore upperfile_sem;
>> >
>> > mutex please
>> >
>
>You missed this comment.
>semaphore is discouraged as a locking primitive.
>Please use struct mutex.

Okay, sorry.

I will check it out.

>
>Thanks,
>Amir.