[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CALCETrV5mWJx5fhfkjdgk+FpqjvPhDvxyn0xbWDDkAMmW-9n+A@mail.gmail.com>
Date: Thu, 29 Sep 2016 10:30:02 -0700
From: Andy Lutomirski <luto@...capital.net>
To: Vineet Gupta <vgupta@...opsys.com>
Cc: Alexey Brodkin <Alexey.Brodkin@...opsys.com>,
arcml <linux-snps-arc@...ts.infradead.org>,
Vineet Gupta <Vineet.Gupta1@...opsys.com>,
lkml <linux-kernel@...r.kernel.org>,
Peter Zijlstra <peterz@...radead.org>
Subject: Re: NMI for ARC
On Sep 28, 2016 6:20 PM, "Vineet Gupta" <vgupta@...opsys.com> wrote:
>
> On 09/28/2016 03:26 PM, Andy Lutomirski wrote:
> >> Right, so what I think Vineet is asking is if we need to disable NMIs as
> >> > well, we cannot on x86 disable NMIs so no.
> >> >
> > The same argument works here, too: an NMI won't set TIF_NEED_RESCHED
> > without sending an IPI, so we can't miss a wakeup.
>
> But what exact wakeup miss are we taking about here. If intr were NOT disabled,
> how could this happen. Just trying to understand the need for "irqs-disabled" in
> resume_{user,kernel}_*
>
> The intr disabled before reg file restore makes complete sense though.
An interrupt after the TIF_NEED_RESCHED check could set
TIF_NEED_RESCHED (because it woke a waiting task), and the resume code
wouldn't notice. No IPI would be sent because TIF_NEED_RESCHED on the
current task is supposed to be acted on in a timely manner without
one.
More severe problems could happen. A signal could be queued or, on
x86, a return-to-usermode notifier could be set by KVM on a
preemptible kernel. The latter, if missed, will bring down the whole
machine.
Powered by blists - more mailing lists