Re: "cat /proc/lockdep" after "rmmod <some module>" when !debug_locks will crash the system

[Bart Van Assche <bvanassche@acm.org>]

On Tue, 2019-03-26 at 10:44 -0700, Bart Van Assche wrote:
> On Tue, 2019-03-26 at 08:35 +0800, shenghui wrote:
> > My test steps:
> > --------------
> > 1) bootup the system, and check the calltrace in dmesg. Just warning and ignore it.
> > 2) cat /proc/lockdep   # everything is well
> > 3) rmmod some module which provides lock_class in lockdep
> >    In my system, module bcache is used:  ('grep bcache /proc/lockdep' prints something)
> >         * stop bcache set
> >         * rmmod bcache
> >    I have tried other module, e.g: rmmod iwldvm
> > 4) cat /proc/lockdep   # system crash
> 
> Hi shenghui,
> 
> Thank you for having shared your test steps. I ran a slightly different test
> myself:
> 
> while true; do cat /proc/lockdep >/dev/null; done &
> (cd blktests && while ./check -q; do :; done)

It seems like I hit "send" too quickly. That test just triggered the following:

BUG: unable to handle kernel paging request at fffffbfff40ca448
#PF error: [normal kernel read fault]
PGD 13bfde067 P4D 13bfde067 PUD 13bf7a067 PMD 1167d3067 PTE 0
Oops: 0000 [#1] PREEMPT SMP KASAN
CPU: 4 PID: 4529 Comm: cat Tainted: G    B   W  O      5.1.0-rc1-dbg+ #4
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1 04/01/2014
RIP: 0010:__asan_load1+0x28/0x50
Call Trace:
 string+0xac/0x180
 vsnprintf+0x23e/0x820
 seq_vprintf+0x82/0xc0
 seq_printf+0x92/0xb0
 print_name+0x34/0xb0
 l_show+0x184/0x200
 seq_read+0x59e/0x6c0
 proc_reg_read+0x11f/0x170
 __vfs_read+0x4d/0x90
 vfs_read+0xc5/0x1f0
 ksys_read+0xab/0x130
 __x64_sys_read+0x43/0x50
 do_syscall_64+0x71/0x210
 entry_SYSCALL_64_after_hwframe+0x49/0xbe

I will have a closer look at this.

Bart.