From mboxrd@z Thu Jan  1 00:00:00 1970
From: maximlevitsky@gmail.com (Maxim Levitsky)
Date: Fri, 22 Oct 2010 02:25:11 +0200
Subject: [PATCH] sdio: fix suspend/resume regression
In-Reply-To: <alpine.LFD.2.00.1010211952110.2764@xanadu.home>
References: <1286788315.3101.11.camel@sven>
	<1286955116-22793-1-git-send-email-ohad@wizery.com>
	<1287704855.12734.14.camel@maxim-laptop>
	<alpine.LFD.2.00.1010211952110.2764@xanadu.home>
Message-ID: <1287707111.12734.30.camel@maxim-laptop>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

On Thu, 2010-10-21 at 19:55 -0400, Nicolas Pitre wrote:
> On Fri, 22 Oct 2010, Maxim Levitsky wrote:
> 
> > So what happens if you set .suspend to NULL? instead of -ENOSYS return?
> 
> At run time, the driver may decide to handle the suspend request and 
> return 0, or decide to simply return -ENOSYS if it doesn't want to 
> bother with the suspend meaning that the card should be removed, or 
> return another error code meaning that the suspend request must be 
> aborted.

I am afraid that this is very hard to support.
If .suspend would be called from pm notifier, it would create problems
last time I checked.
Or maybe it won't?
I will test that again, as it might be very nice solution.
I could even just call the mmc_suspend_host from the pm notifier.

Best regards,
	Maxim Levitsky