linux-kernel - Re: x86, mce, Use user return notifier in mce

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]

Message-ID: <1326436352.8294.22.camel@yhuang-dev>
Date:	Fri, 13 Jan 2012 14:32:32 +0800
From:	Huang Ying <ying.huang@...el.com>
To:	Tony Luck <tony.luck@...el.com>
Cc:	linux-kernel@...r.kernel.org, Ingo Molnar <mingo@...e.hu>,
	Borislav Petkov <bp@...64.org>,
	Chen Gong <gong.chen@...ux.intel.com>,
	Hidetoshi Seto <seto.hidetoshi@...fujitsu.com>
Subject: Re: x86, mce, Use user return notifier in mce

On Thu, 2012-01-12 at 17:46 -0800, Tony Luck wrote:
> On Thu, Jan 12, 2012 at 4:36 PM, Huang Ying <ying.huang@...el.com> wrote:
> > Replace the home-made TIF_MCE_NOTIFY based code in MCE with user
> > return notifier.
> 
> When I looked at this before, I found that the existing user return notifier had
> the meaning "call a function before THIS CPU returns to user space". Use in KVM
> was to update some MSR that needed adjustment before a cpu ran another user
> process.
> 
> The MCE code wanted something slightly different: "call a function before THIS
> PROCESS returns to user space". So my prototype code from last year made a
> whole new set of interfaces - similar in style to the user return
> notifier, but with
> the MCE semantics.
> 
> At first glance it looks like you are just using the user return
> notifier code (perhaps
> I'm mis-reading the diff?). This won't work - it's possible for a
> context switch, and
> then the process that hit the MCE may get moved to another cpu, when it will
> be run. Meanwhile we'll execute our function in the context of some
> other process.

You are right.  User return notifier can not be used for SRAR.  I think
that may be useful for SRAO.  Where we need a way to do notify earlier
in case of the corresponding work_queue item is not executed in time.

Best Regards,
Huang Ying


--
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/