linux-kernel - Re: [PATCH] Test for riscv fixes

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 for Android: free password hash cracker in your pocket

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

Message-ID: <CANp29Y79=Y76i72wHWfOJqU1tg8Z6vpbmwBRr=wcet7t5xdX5A@mail.gmail.com>
Date:   Mon, 2 Oct 2023 12:20:10 +0200
From:   Aleksandr Nogikh <nogikh@...gle.com>
To:     Conor Dooley <conor@...nel.org>
Cc:     Greg KH <gregkh@...uxfoundation.org>,
        Edward AD <twuufnxlz@...il.com>, alex@...ti.fr,
        alexghiti@...osinc.com, aou@...s.berkeley.edu, guoren@...nel.org,
        jirislaby@...nel.org, linux-kernel@...r.kernel.org,
        linux-riscv@...ts.infradead.org, linux-serial@...r.kernel.org,
        liushixin2@...wei.com, palmer@...belt.com,
        paul.walmsley@...ive.com,
        syzbot+8d2757d62d403b2d9275@...kaller.appspotmail.com,
        syzkaller-bugs@...glegroups.com
Subject: Re: [PATCH] Test for riscv fixes

On Sat, Sep 30, 2023 at 10:24 AM Conor Dooley <conor@...nel.org> wrote:
>
> On Sat, Sep 30, 2023 at 08:13:56AM +0200, Greg KH wrote:
> > On Sat, Sep 30, 2023 at 07:05:35AM +0800, Edward AD wrote:
> > > On Fri, 29 Sep 2023 08:04:57 +0200 Greg KH wrote:
> > > > Where are you getting your odd cc: list from?  This has nothing to do
> > > > with serial drivers...
> > > https://lore.kernel.org/all/0000000000000170df0605ccf91a@google.com/raw
> >
> > I do not understand this answer.
>
> AIUI, the original report from syzbot "blamed" the serial maintainers.
> Not too sure how it determined that though, given the contents.
>

blame is too strong a word for that auto-inferred hint :)

Yes, the actual problem was in a totally different place, but FWIW
here's how it happened:

Statistically, stacktrace.c and a number of other generic locations
are rather unlikely to actually contain the bug (they can, but the
chances are that it's deeper in the call stack), so syzbot, while
traversing the stack trace, skipped all frames until

[<ffffffff80fa0c7e>] fn_show_state+0x1a/0x22 drivers/tty/vt/keyboard.c:614
[<ffffffff80fa1152>] k_spec drivers/tty/vt/keyboard.c:667 [inline]
[<ffffffff80fa1152>] k_spec+0xce/0x102 drivers/tty/vt/keyboard.c:656
[<ffffffff80fa306c>] kbd_keycode drivers/tty/vt/keyboard.c:1524 [inline]
[<ffffffff80fa306c>] kbd_event+0x5fa/0xa5e drivers/tty/vt/keyboard.c:1543

Per MAINTAINERS, these locations belonged to "TTY LAYER AND SERIAL
DRIVERS". Therefore serial in Cc.

The automation actually attributes a lot of reports correctly, but,
unfortunately, there are also tricky cases like this.

-- 
Aleksandr