Re: [PATCH v4 2/5] x86, traps: Track entry into and exit from IST context

From: Andy Lutomirski <luto@amacapital.net>
To: Tony Luck <tony.luck@gmail.com>
Cc: "Borislav Petkov" <bp@alien8.de>,
	"Paul McKenney" <paulmck@linux.vnet.ibm.com>,
	"X86 ML" <x86@kernel.org>,
	"Linus Torvalds" <torvalds@linux-foundation.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"Peter Zijlstra" <peterz@infradead.org>,
	"Oleg Nesterov" <oleg@redhat.com>,
	"Andi Kleen" <andi@firstfloor.org>,
	"Josh Triplett" <josh@joshtriplett.org>,
	"Frédéric Weisbecker" <fweisbec@gmail.com>
Subject: Re: [PATCH v4 2/5] x86, traps: Track entry into and exit from IST context
Date: Wed, 10 Dec 2014 16:24:46 -0800	[thread overview]
Message-ID: <CALCETrWFBEvL4CSD+cZ0inMcvv09rLJi4Wpf4gLz5fvsGcV-wQ@mail.gmail.com> (raw)
In-Reply-To: <CA+8MBb+iw3KJz5EgX-DLBHGWANUSVFP2bQe2_U4kqCfM-wzH8Q@mail.gmail.com>

On Wed, Dec 10, 2014 at 4:22 PM, Tony Luck <tony.luck@gmail.com> wrote:
> On Tue, Nov 25, 2014 at 10:58 AM, Borislav Petkov <bp@alien8.de> wrote:
>> On Mon, Nov 24, 2014 at 03:52:10PM -0800, Andy Lutomirski wrote:
>>> Dunno. Tony and Borislav -- when do you want the IST stack switching
>>> stuff?
>>
>> I'd leave that up to Tony and his testbench. I mean, we can hammer on
>> it as much as we can and it can pass all testing locally but the real
>> fun starts once it hits upstream and for that it doesn't matter which
>> release... IMHO.
>
> So what was the net result of all the mode/RCU discussions?
>
> Do I need some extra magic incantations in the final version
> of do_machine_check() beyond what was in this patch:
>
>   https://lkml.kernel.org/r/546d169211835aded@agluck-desk.sc.intel.com
>
> to make everything happy?
>

I think you need ist_begin_non_atomic() before local_irq_enable() and
ist_end_non_atomic() after local_irq_disable().  Otherwise it should
be good.

--Andy

> -Tony

-- 
Andy Lutomirski
AMA Capital Management, LLC