Re: [PATCH v1] binfmt_misc: fix crash when load/unload module

[Murphy Zhou]

On Tue, Jan 25, 2022 at 10:33:22PM -0800, Tong Zhang wrote:
> On Tue, Jan 25, 2022 at 9:23 PM Tong Zhang <ztong0001@xxxxxxxxx> wrote:
> >
> > On Tue, Jan 25, 2022 at 9:04 PM Murphy Zhou <jencce.kernel@xxxxxxxxx> wrote:
> > >
> > > Still panic with this patch on Linux-next tree:
> > >
> > > [ 1128.275515] LTP: starting binfmt_misc02 (binfmt_misc02.sh)
> > > [ 1128.303975] CPU: 1 PID: 107182 Comm: modprobe Kdump: loaded
> > > Tainted: G        W         5.17.0-rc1-next-20220125+ #1
> > > [ 1128.305264] Hardware name: Red Hat KVM, BIOS 0.5.1 01/01/2011
> > > [ 1128.305992] Call Trace:
> > > [ 1128.306376]  <TASK>
> > > [ 1128.306682]  dump_stack_lvl+0x34/0x44
> > > [ 1128.307211]  __register_sysctl_table+0x2c7/0x4a0
> > > [ 1128.307846]  ? load_module+0xb37/0xbb0
> > > [ 1128.308339]  ? 0xffffffffc01b6000
> > > [ 1128.308762]  init_misc_binfmt+0x32/0x1000 [binfmt_misc]
> > > [ 1128.309402]  do_one_initcall+0x44/0x200
> > > [ 1128.309937]  ? kmem_cache_alloc_trace+0x163/0x2c0
> > > [ 1128.310535]  do_init_module+0x5c/0x260
> > > [ 1128.311045]  __do_sys_finit_module+0xb4/0x120
> > > [ 1128.311603]  do_syscall_64+0x3b/0x90
> > > [ 1128.312088]  entry_SYSCALL_64_after_hwframe+0x44/0xae
> > > [ 1128.312755] RIP: 0033:0x7f929ab85fbd
> > >
> > > Testing patch on Linus tree.
> >
> > Hi Murphy,
> > Did you apply this patch?
> > Link: https://lkml.kernel.org/r/20220124181812.1869535-2-ztong0001@xxxxxxxxx
> > I tested it on top of the current master branch and it works on my
> > setup using the reproducer I mentioned.
> > Could you share your test script?
> > Thanks,
> > - Tong
> 
> I can find binfmt_misc02.sh on github, and running the following
> command shows: failed 0.
> 
> ./runltp -s binfmt_misc
> Running tests.......
> <<<test_start>>>
> tag=binfmt_misc01 stime=1643178454
> cmdline="binfmt_misc01.sh"
> contacts=""
> analysis=exit
> <<<test_output>>>
> [   90.908282] LTP: starting binfmt_misc01 (binfmt_misc01.sh)
> binfmt_misc01 1 TINFO: timeout per run is 0h 5m 0s
> binfmt_misc01 1 TPASS: Failed to register a binary type
> binfmt_misc01 2 TPASS: Failed to register a binary type
> binfmt_misc01 3 TPASS: Failed to register a binary type
> binfmt_misc01 4 TPASS: Failed to register a binary type
> binfmt_misc01 5 TPASS: Failed to register a binary type
> binfmt_misc01 6 TPASS: Failed to register a binary type
> binfmt_misc01 7 TPASS: Failed to register a binary type
> binfmt_misc01 8 TPASS: Failed to register a binary type
> binfmt_misc01 9 TPASS: Failed to register a binary type
> 
> Summary:
> passed   9
> failed   0
> broken   0
> skipped  0
> warnings 0
> <<<execution_status>>>
> initiation_status="ok"
> duration=0 termination_type=exited termination_id=0 corefile=no
> cutime=2 cstime=17
> <<<test_end>>>
> <<<test_start>>>
> tag=binfmt_misc02 stime=1643178454
> cmdline="binfmt_misc02.sh"
> contacts=""
> analysis=exit
> <<<test_output>>>
> [   91.133399] LTP: starting binfmt_misc02 (binfmt_misc02.sh)
> incrementing stop
> binfmt_misc02 1 TINFO: timeout per run is 0h 5m 0s
> binfmt_misc02 1 TPASS: Recognise and unrecognise a binary type as expected
> binfmt_misc02 2 TPASS: Recognise and unrecognise a binary type as expected
> binfmt_misc02 3 TPASS: Recognise and unrecognise a binary type as expected
> binfmt_misc02 4 TPASS: Recognise and unrecognise a binary type as expected
> binfmt_misc02 5 TPASS: Fail to recognise a binary type
> binfmt_misc02 6 TPASS: Fail to recognise a binary type
> 
> Summary:
> passed   6
> failed   0
> broken   0
> skipped  0
> warnings 0
> <<<execution_status>>>
> initiation_status="ok"
> duration=0 termination_type=exited termination_id=0 corefile=no
> cutime=3 cstime=25
> <<<test_end>>>
> INFO: ltp-pan reported all tests PASS
> LTP Version: 20220121-9-g010e4f783
> 
>        ###############################################################
> 
>             Done executing testcases.
>             LTP Version:  20220121-9-g010e4f783
>        ###############################################################

Ya, looks like it's working. No panic on next-20220128 tree.

Thanks,
-- 
Murphy