linux-mips
[Top] [All Lists]

Re: [PATCH] rest of works for migration to GENERIC_TIME

To: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Subject: Re: [PATCH] rest of works for migration to GENERIC_TIME
From: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Date: Mon, 23 Oct 2006 19:39:22 +0400
Cc: linux-mips@linux-mips.org, ralf@linux-mips.org
In-reply-to: <20061024.003845.71086839.anemo@mba.ocn.ne.jp>
Organization: MontaVista Software Inc.
Original-recipient: rfc822;linux-mips@linux-mips.org
References: <20061023.033407.104640794.anemo@mba.ocn.ne.jp> <453CD3ED.8020005@ru.mvista.com> <20061024.003845.71086839.anemo@mba.ocn.ne.jp>
Sender: linux-mips-bounce@linux-mips.org
User-agent: Mozilla/5.0 (X11; U; Linux i686; rv:1.7.2) Gecko/20040803
Hello.

Atsushi Nemoto wrote:

* mips_hpt_mask variable to specify bitmask of hpt value.

   There's actually no need to introduce more variables. Just make
clocksource declaration public and override default mask if
necessary.
   Also, I don't see much sense in further existence of
mips_hpt_read() -- it only causes each clocksource read go thru a
double indirection which is really ugly. The same approach shouyld
be used here.

I agree with you that it would be a way to go.  For now exporting
clocksource_mips just to override the mask and keep using
mips_hpt_read looks somewhat inconsistent, so I just added
mips_hpt_mask variable.

Replacing mips_hpt_read involves changes for _all_ platform code so I

Only 3 files in the current arch/mips/ actually, excluding time.c and the header file (to make it visible).

think it would be better to do it on next -rc1 stage.

   AKA never. ;-)

WBR, Sergei

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