From mboxrd@z Thu Jan  1 00:00:00 1970
From: Paul Mundt <lethal@linux-sh.org>
Date: Tue, 10 Jan 2012 06:01:02 +0000
Subject: Re: [PATCH v2 1/3] dmaengine: shdma: modify the DMAC Address Error registration
Message-Id: <20120110060102.GA5561@linux-sh.org>
List-Id: <linux-sh.vger.kernel.org>
References: <4F0BCA90.3050508@renesas.com>
In-Reply-To: <4F0BCA90.3050508@renesas.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: linux-sh@vger.kernel.org

On Tue, Jan 10, 2012 at 02:56:33PM +0900, Shimoda, Yoshihiro wrote:
> 2012/01/10 14:33, Paul Mundt wrote:
> > On Tue, Jan 10, 2012 at 02:20:16PM +0900, Shimoda, Yoshihiro wrote:
> >> +		if (chanirq_res = errirq_res ||
> >> +		    (errirq_res->flags & IORESOURCE_BITS) =
> >> +		    IORESOURCE_IRQ_SHAREABLE)
> >> +			irqflags = IRQF_SHARED;
> >> +
> > While not your fault, it seems that there are a few of these cases in
> > which the irqflags are clobbered rather than added to. Is there some
> > particular reason why we aren't doing |= IRQF_SHARED on top of the
> > default IRQF_DISABLED case?
> > 
> 
> I guess that old kernel (before the commit of 6932bf37 "genirq: Remove IRQF_DISABLED from
> core code") output some warning in request_irq() when irqflags is set to
> "IRQF_DISABLED | IRQF_SHARED".
> 
> By the way, according to the "feature-removal-schedule.txt", the IRQF_DISABLED will remove.
> So, we can remove the IRQF_DISABLED in the shdma driver, I think.
> 
Yes, it's deprecated now, so there's no need to keep it around. The
non-shared case can just be 0.