Re: [PATCH 5.15 000/251] 5.15.47-rc2 review

[Guenter Roeck]

On Tue, Jun 14, 2022 at 08:36:08AM -0700, Guenter Roeck wrote:
> On Mon, Jun 13, 2022 at 08:19:49PM +0200, Greg Kroah-Hartman wrote:
> > This is the start of the stable review cycle for the 5.15.47 release.
> > There are 251 patches in this series, all will be posted as a response
> > to this one.  If anyone has any issues with these being applied, please
> > let me know.
> > 
> > Responses should be made by Wed, 15 Jun 2022 18:18:03 +0000.
> > Anything received after that time might be too late.
> > 
> 
> Build results:
> 	total: 159 pass: 159 fail: 0
> Qemu test results:
> 	total: 488 pass: 488 fail: 0
> 

I spoke a bit too early. I see the following backtrace in some qemu arm
boot tests.

BUG: spinlock bad magic on CPU#0, kdevtmpfs/15
 lock: noop_backing_dev_info+0x6c/0x3b0, .magic: 00000000, .owner: <none>/-1, .owner_cpu: 0
CPU: 0 PID: 15 Comm: kdevtmpfs Not tainted 5.15.47-rc2-00252-g677f0128d0ed #1
Hardware name: ARM RealView Machine (Device Tree Support)
[<c01101d0>] (unwind_backtrace) from [<c010bc0c>] (show_stack+0x10/0x14)
[<c010bc0c>] (show_stack) from [<c0a10ae4>] (dump_stack_lvl+0x68/0x90)
[<c0a10ae4>] (dump_stack_lvl) from [<c0191250>] (do_raw_spin_lock+0xbc/0x124)
[<c0191250>] (do_raw_spin_lock) from [<c02eb578>] (__mark_inode_dirty+0x1cc/0x704)
[<c02eb578>] (__mark_inode_dirty) from [<c02e6a74>] (simple_setattr+0x44/0x5c)
[<c02e6a74>] (simple_setattr) from [<c02d7a18>] (notify_change+0x400/0x45c)
[<c02d7a18>] (notify_change) from [<c0a19ef8>] (devtmpfsd+0x1f8/0x2b8)
[<c0a19ef8>] (devtmpfsd) from [<c014cf3c>] (kthread+0x150/0x17c)
[<c014cf3c>] (kthread) from [<c0100120>] (ret_from_fork+0x14/0x34)
Exception stack(0xd00dbfb0 to 0xd00dbff8)
bfa0:                                     00000000 00000000 00000000 00000000
bfc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
bfe0: 00000000 00000000 00000000 00000000 00000013 00000000

This bisects to commit bc5d960d4e58 ("writeback: Fix inode->i_io_list not
be protected by inode->i_lock error"). The problem is also seen in the
mainline kernel. v5.15.y.queue and later are affected. Reverting the patch
here and in mainline fixes the problem.

Guenter