linux-kernel - Re: 4.10-rc1: thinkpad x60: who ate my cpu?

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]

Date:   Mon, 13 Feb 2017 09:02:15 +0100
From:   Pavel Machek <pavel@....cz>
To:     Woody Suwalski <terraluna977@...il.com>
Cc:     "Rafael J. Wysocki" <rjw@...ysocki.net>,
        kernel list <linux-kernel@...r.kernel.org>, tglx@...utronix.de,
        mingo@...hat.com, hpa@...or.com
Subject: Re: 4.10-rc1: thinkpad x60: who ate my cpu?

Hi!

> >>>And now it is:
> >>>[692517.868523] ACPI: Preparing to enter system sleep state S3
> >>>[692518.172074] ACPI : EC: EC stopped
> >>>[692518.172076] PM: Saving platform NVS memory
> >>>[692518.172269] Disabling non-boot CPUs ...
> >>>[692518.172269] ACPI: Low-level resume complete
> >>>[692518.172269] ACPI : EC: EC started
> >>>[692518.172269] PM: Restoring platform NVS memory
> >>>[692518.172269] ACPI: Waking up from system sleep state S3
> >>>
> >>>Is there any test I could do on the CPU wakeup while in that state?
> >>>
> >>Is there a way to kick the offline-CPU into operation from /sys level?
> >echo 0 > /sys/devices/system/cpu/cpu1/online
> >
> >should work. And... good thinking :-).
> >									Pavel
> Did not work,
>     echo 0 > /sys/devices/system/cpu/cpu1/online
>     -su: echo: write error: Device or resource busy
> 
> However
>     echo 1 > /sys/devices/system/cpu/cpu1/online
> did not return an error ( but still only CPU0 seen)
> 
> Interesting experiment: I have hibernated and then woke up - and still only
> CPU0. I was expecting that after the power cycle the hotplug will bring CPU1
> up...

Which is interesting indeed. Hardware was powercycled, so what is
broken is probably some kernel state...

[Evil: /sys/devices/cpu/ -- some perf stuff.
       /sys/devices/system/cpu -- that's where real CPU stuff is
       hiding.]

cd /sys/devices/system/cpu/cpu1
while true; do echo 0 > online; echo 1 > online; done

...crashes x60 with 4.10-rc in few minutes. [Which is bad -- it should
not die, but also good -- this is easier to reproduce then running 100
suspend cycles.]

									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Download attachment "signature.asc" of type "application/pgp-signature" (182 bytes)