linux-kernel - Re: [PATCH] watchdog: hpwdt: Remove spinlock acquire and BIOS calls from NMI context

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:   Thu, 15 Feb 2018 20:02:24 +0100
From:   Ingo Molnar <mingo@...nel.org>
To:     Jerry Hoemann <jerry.hoemann@....com>
Cc:     Borislav Petkov <bp@...en8.de>,
        Tim Chen <tim.c.chen@...ux.intel.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Dave Hansen <dave@...1.net>, hpa@...or.com, tglx@...utronix.de,
        torvalds@...ux-foundation.org, linux-kernel@...r.kernel.org,
        dwmw@...zon.co.uk, linux-tip-commits@...r.kernel.org,
        Arjan van de Ven <arjan@...radead.org>,
        linux-watchdog@...r.kernel.org
Subject: Re: [PATCH] watchdog: hpwdt: Remove spinlock acquire and BIOS calls
 from NMI context


* Jerry Hoemann <jerry.hoemann@....com> wrote:

> On Thu, Feb 15, 2018 at 12:17:04AM +0100, Ingo Molnar wrote:
> > 
> > * Jerry Hoemann <jerry.hoemann@....com> wrote:
> > 
> > > 
> > > Ingo,
> > > 
> > > I have a patch set under review that brings hpwdt into compliance
> > > with the watchdog core.
> > > 
> > > One of the changes removes the callback into firmware in hpwdt_pretimeout
> > > and its associated spinlock.
> > > 
> > > https://lkml.org/lkml/2018/2/12/30
> > 
> >  drivers/watchdog/hpwdt.c | 490 +----------------------------------------------
> >  1 file changed, 8 insertions(+), 482 deletions(-)
> > 
> > Very nice, and this should solve all the NMI handling complications!
> > 
> > > I will add you to the CC list of the next version of the set.
> > 
> > Thanks!
> > 
> > 	Ingo
> 
> 
> Ingo,
> 
> Is your desire to remove of the firmware callback/spinlock in hpwdt_pretimeout
> related to David Woodhouse patch set:
> 
> 	https://lkml.org/lkml/2018/2/14/305  ?
> 
> Which I think are to mitigate performance issues resulting from the
> changes to address the specter/meltdown?

So the motivation was that we are trying to wrap BIOS/EFI calls into 
Spectre-disabling sections - and while doing that we realized that hpwdt calls the 
firmware from an NMI callback, which would have complicated the Spectre work..

But with that call removed from NMI context it's all perfect!

> Any there other changes to hpwdt needed related to this work?

No other changes needed I think!

Thanks,

	Ingo