linux-mips
[Top] [All Lists]

Re: [loongson-PATCH-v2 23/23] Hibernation Support in mips system

To: apatard@mandriva.com
Subject: Re: [loongson-PATCH-v2 23/23] Hibernation Support in mips system
From: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Date: Fri, 29 May 2009 00:32:49 +0900 (JST)
Cc: wuzhangjin@gmail.com, linux-mips@linux-mips.org, ralf@linux-mips.org, wuzj@lemote.com, yanh@lemote.com, philippe@cowpig.ca, r0bertz@gentoo.org, zhangfx@lemote.com, loongson-dev@googlegroups.com, der.herr@hofr.at, liujl@lemote.com, erwan@thiscow.com
In-reply-to: <m3my8yoovk.fsf@anduin.mandriva.com>
Original-recipient: rfc822;linux-mips@linux-mips.org
References: <cover.1243362545.git.wuzj@lemote.com> <1483a7cb0f587bd329ea3ca8d3af2881afadaf5e.1243362545.git.wuzj@lemote.com> <m3my8yoovk.fsf@anduin.mandriva.com>
Sender: linux-mips-bounce@linux-mips.org
On Wed, 27 May 2009 11:51:11 +0200, Arnaud Patard <apatard@mandriva.com> wrote:
> > +LEAF(swsusp_arch_resume)
...
> 
> you really need to flush cache/tlb here. If you don't do that you'll get
> some weird bugs.

I also wonder if we need to flush cache on swsusp_arch_suspend.  Maybe
kernel pages does not need to be flushed on here, but how about user
pages with dcache aliasing?  It seems swsusp_save() reads from source
page via kernel mapping without any care for coherency...

---
Atsushi Nemoto

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