linux-mips
[Top] [All Lists]

Re: WAIT vs. tickless kernel

To: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Subject: Re: WAIT vs. tickless kernel
From: Ralf Baechle <ralf@linux-mips.org>
Date: Wed, 31 Oct 2007 16:39:00 +0000
Cc: linux-mips@linux-mips.org
In-reply-to: <20071101.013124.108121433.anemo@mba.ocn.ne.jp>
Original-recipient: rfc822;linux-mips@linux-mips.org
References: <20071101.004906.106263529.anemo@mba.ocn.ne.jp> <20071031161333.GA22871@linux-mips.org> <20071101.013124.108121433.anemo@mba.ocn.ne.jp>
Sender: linux-mips-bounce@linux-mips.org
User-agent: Mutt/1.5.14 (2007-02-12)
On Thu, Nov 01, 2007 at 01:31:24AM +0900, Atsushi Nemoto wrote:

> On Wed, 31 Oct 2007 16:13:33 +0000, Ralf Baechle <ralf@linux-mips.org> wrote:
> > This one is definately playing with the fire.  Or alternatively requires
> > detailed knowledge of the pipeline and pipelines tend to change.  MIPS
> > Technologies does regular maintenance releases of its cores which also
> > add features and may change the pipelines in subtle way that may break
> > something like this.
> 
> Yes, I never think this is robust or guaranteed...
> 
> > The only safe but ugly workaround is to change the return from exception
> > code to detect if the EPC is in the range startin from the condition
> > check in the idle loop to including the WAIT instruction and if so to
> > patch the EPC to resume execution at the condition check or the
> > instruction following the WAIT.
> 
> I'm also thinking of this approach.  Still wondering if it is worth to
> implement.

The tickless kernel is very interesting for the low power fraction.  And
it's especially those users who would suffer most the loss of the ability
to use the WAIT instruction.  For a system running from two AAA cells the
tradeoff is clear ...  So I think it's become a must.

  Ralf

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