linux-kernel - Re: Regression on cpufreq in v3.12-rc1

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:	Fri, 20 Sep 2013 18:34:34 +0200
From:	Linus Walleij <linus.walleij@...aro.org>
To:	"Srivatsa S. Bhat" <srivatsa.bhat@...ux.vnet.ibm.com>,
	Russell King - ARM Linux <linux@....linux.org.uk>
Cc:	Viresh Kumar <viresh.kumar@...aro.org>,
	"Rafael J. Wysocki" <rjw@...ysocki.net>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	Kristoffer Ericson <kristoffer.ericson@...il.com>
Subject: Re: Regression on cpufreq in v3.12-rc1

On Fri, Sep 20, 2013 at 5:40 PM, Linus Walleij <linus.walleij@...aro.org> wrote:
> On Fri, Sep 20, 2013 at 5:32 PM, Srivatsa S. Bhat
> <srivatsa.bhat@...ux.vnet.ibm.com> wrote:
>> On 09/20/2013 08:51 PM, Linus Walleij wrote:
>
>>> sa11x0_pcmcia_init() which starts this chain of events is called as
>>> an fs_initcall(), see drivers/pcmcia/sa1100_generic.c
>>
>> But fs_initcall() comes after arch_initcall() right? (looking at
>> include/linux/init.h). Since the corresponding cpufreq-driver (drivers/
>> cpufreq/sa1100-cpufreq.c) uses arch_initcall() to register itself, it
>> is all the more surprising how this could happen...
>
> Hm it does happen if the cpufreq driver does not probe *at all*
> right? I will look closer at this, maybe it just doesn't register
> properly for some SA1100 variants...

Aha yeah this is it. The h3600 has an SA1110 and uses
drivers/cpufreq/sa1110-cpufreq.c.

This driver needs to know the RAM mounted on the board,
if it is this or that type, to perform properly.

This is passed from the command line or a default is selected
from the machine type:

    if (!name[0]) {
        if (machine_is_assabet())
            name = "TC59SM716-CL3";
        if (machine_is_pt_system3())
            name = "K4S641632D";
        if (machine_is_h3100())
            name = "KM416S4030CT";
        if (machine_is_jornada720())
            name = "K4S281632B-1H";
        if (machine_is_nanoengine())
            name = "MT48LC8M16A2TG-75";
    }

In this case we have neither. So the cpufreq driver fails to
register. But the kernel used to survive in any case, as you
could call cpufreq_get() without a cpufreq driver registered.

I think your suggested patch fixes the issue.

As a proper fix I'd like to add a default RAM type for the h3600
but that is for later...

Yours,
Linus Walleij
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/