linux-kernel - Re: [PATCH v5 0/9] Add OPTPROBES feature on RISCV

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 04 Jan 2023 09:53:01 +0100
From:   Björn Töpel <bjorn@...nel.org>
To:     "liaochang (A)" <liaochang1@...wei.com>,
        Chen Guokai <chenguokai17@...ls.ucas.ac.cn>,
        paul.walmsley@...ive.com, palmer@...belt.com,
        aou@...s.berkeley.edu, rostedt@...dmis.org, mingo@...hat.com,
        sfr@...b.auug.org.au
Cc:     linux-riscv@...ts.infradead.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v5 0/9] Add OPTPROBES feature on RISCV

"liaochang (A)" <liaochang1@...wei.com> writes:

> Hi，Björn，appreciate for your review and testing about this feature.

Thank you for the hard work!

> 在 2023/1/3 2:02, Björn Töpel 写道:
>> Chen Guokai <chenguokai17@...ls.ucas.ac.cn> writes:
>> 
>>> Add jump optimization support for RISC-V.
>> 
>> Thank you for continuing to work on the series! I took the series for a
>> spin, and ran into a number of issues that makes me wonder how you test
>> the series, and how the testing is different from my runs.
>
> I have pick some kernel functions to test this series, which means all optprobe
> are install at entry of function, i guess the instruction pattern is not versatile
> enough for my testcases leads to some bugs are not discovered.
>
> Do you think it is good idea to test this feature via binary ftracetest and the
> kprobe related tc scripts in tools/testing/ftrace directory?

Definitely! Both running all tests in tools/testing/selftests/ftrace and
with the CONFIG_KPROBES_SANITY_TEST module.

[...]

>> 4. Lockdep splat

[...]

> Need to study this backtrace further, but at first glance, i guess CONFIG_DYNAMIC_FTRACE is enabled on your kernel, right?
> If so, all krpobe is installed via ftrace stub, then kprobe optimiztion occur in the ftrace trampoline code, and it also
> a corner case to current optprobe implementation.

Yes, CONFIG_DYNAMIC_FTRACE is on. My kernel config was simply:
  make ARCH=riscv CROSS_COMPILE=riscv64-linux-gnu- defconfig
  make ARCH=riscv CROSS_COMPILE=riscv64-linux-gnu- kselftest-merge


Thanks,
Björn