Re: [LTP] [PATCH] syscalls/setsockopt09: Add another linux git

["xuyang2018.jy@fujitsu.com" <xuyang2018.jy@fujitsu.com>]

Hi Petr
>> Hi Petr
>>> Hi Xu,
>
>>>> On centos7.9ga, I still hit another crash problem because of use-after-free in
>>>> prb_retire_rx_blk_timer_expired(). Since we free it when timer expired, so this
>>>> case will crash after we print TPASS info.
>
>>> Reviewed-by: Petr Vorel<pvorel@suse.cz>
>
>>> LGTM.
>>> I tested two old kernels, the one with patch survives, the other got reboot.
>
>>> BTW funny enough the affected system manages to print "TPASS: Nothing bad
>>> happened, probably" before reboot :).
>> Yes, it crash when timer expired, so it will print TPASS before reboot.
>> Also, I try sleep 10ms, but it still print TPASS and reboot after serval
>> seconds.
>
>> I have wrote it in my commit message, so this should  be friendly for
>> user to know this situation.
>
> Sure, I didn't expect we would be able to fix this (I'm *not* voting for sleep
> 10+ s). It can just be a bit confusing when you read test logs if the framework
> does not clearly show that system got rebooted (I saw reset in dmesg but didn't
> believe it's caused by this test output due TPASS. Lesson learned :)).
Yes, it looks confused.

So, how about testing 100 times in runtest/syscall and runtest/cve like 
"cve-2021-22555 setsockopt08 -i 100" does.

ps: I tested it on old kernel and it works well.

Best Regards
Yang Xu
>
> Kind regards,
> Petr
>
>> Best Regards
>> Yang Xu
>
>>> Kind regards,
>>> Petr

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp