linux-mips
[Top] [All Lists]

Re: [PATCH 6/7] MIPS: Fix CP0 COUNTER clockevent race

To: Kevin Cernekee <cernekee@gmail.com>
Subject: Re: [PATCH 6/7] MIPS: Fix CP0 COUNTER clockevent race
From: Ralf Baechle <ralf@linux-mips.org>
Date: Wed, 24 Nov 2010 11:34:05 +0000
Cc: linux-mips@linux-mips.org, linux-kernel@vger.kernel.org
In-reply-to: <444ef6c4bbb47d55c700452d8cd23229@localhost>
Original-recipient: rfc822;linux-mips@linux-mips.org
References: <8a8eee995454c8b271cceb440e31699a@localhost> <444ef6c4bbb47d55c700452d8cd23229@localhost>
Sender: linux-mips-bounce@linux-mips.org
User-agent: Mutt/1.5.21 (2010-09-15)
On Tue, Nov 23, 2010 at 10:26:44AM -0800, Kevin Cernekee wrote:

> write_c0_compare(read_c0_count());
> 
> Even if the counter doesn't increment during execution, this might not
> generate an interrupt until the counter wraps around.  The CPU may
> perform the comparison each time CP0 COUNT increments, not when CP0
> COMPARE is written.
> 
> If mips_next_event() is called with a very small delta, and CP0 COUNT
> increments during the calculation of "cnt += delta", it is possible
> that CP0 COMPARE will be written with the current value of CP0 COUNT.
> If this is detected, the function should return -ETIME, to indicate
> that the interrupt might not have actually gotten scheduled.

Good catch - though on real hardware it should be theoretical as the
minimum timer interval is 300ns.  So it should only be trigerable on
a very slow system like a hardware emulator or maybe if a software
emulator like qemu gets rescheduled between the update and the read-back.

Applied,

  Ralf

<Prev in Thread] Current Thread [Next in Thread>