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  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]
Date:   Sat, 16 May 2020 21:45:15 +0530
From:   Subhashini Rao Beerisetty <subhashbeerisetty@...il.com>
To:     Randy Dunlap <rdunlap@...radead.org>
Cc:     Valdis Klētnieks <valdis.kletnieks@...edu>,
        kernelnewbies <kernelnewbies@...nelnewbies.org>,
        linux-kernel@...r.kernel.org
Subject: Re: general protection fault vs Oops

On Sat, May 16, 2020 at 9:29 PM Randy Dunlap <rdunlap@...radead.org> wrote:
>
> On 5/16/20 6:53 AM, Valdis Klētnieks wrote:
> > On Sat, 16 May 2020 18:05:07 +0530, Subhashini Rao Beerisetty said:
> >
> >> In the first attempt when I run that test case I landed into “general
> >> protection fault: 0000 [#1] SMP" .. Next I rebooted and ran the same
> >> test , but now it resulted the “Oops: 0002 [#1] SMP".
> >
> > And the 0002 is telling you that there's been 2 previous bug/oops since the
> > reboot, so you need to go back through your dmesg and find the *first* one.
> >
> >> In both cases the call trace looks exactly same and RIP points to
> >> “native_queued_spin_lock_slowpath+0xfe/0x170"..
> >
> > The first few entries in the call trace are the oops handler itself. So...
> >
> >
> >> May 16 12:06:17 test-pc kernel: [96934.567347] Call Trace:
> >> May 16 12:06:17 test-pc kernel: [96934.569475]  [<ffffffff8183c427>]__raw_spin_lock_irqsave+0x37/0x40
> >> May 16 12:06:17 test-pc kernel: [96934.571686]  [<ffffffffc0606812>] event_raise+0x22/0x60 [osa]
> >> May 16 12:06:17 test-pc kernel: [96934.573935]  [<ffffffffc06aa2a4>] multi_q_completed_one_buffer+0x34/0x40 [mcore]
> >
> > The above line is the one where you hit the wall.
> >
> >> May 16 12:59:22 test-pc kernel: [ 3011.405602] Call Trace:
> >> May 16 12:59:22 test-pc kernel: [ 3011.407892]  [<ffffffff8183c427>] _raw_spin_lock_irqsave+0x37/0x40
> >> May 16 12:59:22 test-pc kernel: [ 3011.410256]  [<ffffffffc0604812>] event_raise+0x22/0x60 [osa]
> >> May 16 12:59:22 test-pc kernel: [ 3011.412652]  [<ffffffffc06b72a4>] multi_q_completed_one_buffer+0x34/0x40 [mcore]
> >
> > And again.
> >
> > However,  given that it's a 4.4 kernel from 4 years ago, it's going to be
> > hard to find anybody who really cares.
>
> Right.
>
> > In fact. I'm wondering if this is from some out-of-tree or vendor patch,
> > because I'm not finding any sign of that function in either the 5.7 or 4.4
> > tree.  Not even a sign of ## catenation abuse - no relevant hits for
> > "completed_one_buffer" or "multi_q" either
>
> Modules linked in:
> dbg(OE) mcore(OE) osa(OE)
>
> Out-of-tree, unsigned modules loaded.
Yes, those are out-of-tree modules. Basically, my question is, in
general what is the difference between 'general protection fault' and
'Oops' failure in kernel mode.

> We don't know what those are or how to debug them.
>
> > I don't think anybody's going to be able to help unless somebody first
> > identifies where that function is....
> >
>
>
> --
> ~Randy
>

Powered by blists - more mailing lists