[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <4DB7B0D5.5050504@st.com>
Date: Wed, 27 Apr 2011 11:29:49 +0530
From: viresh kumar <viresh.kumar@...com>
To: Russell King - ARM Linux <linux@....linux.org.uk>,
"vinod.koul@...el.com" <vinod.koul@...el.com>
Cc: "dan.j.williams@...el.com" <dan.j.williams@...el.com>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"linux-arm-kernel@...ts.infradead.org"
<linux-arm-kernel@...ts.infradead.org>,
Armando VISCONTI <armando.visconti@...com>,
Shiraz HASHIM <shiraz.hashim@...com>,
amitgoel <amit.goel@...com>,
"viresh.linux@...il.com" <viresh.linux@...il.com>,
"jamie@...ieiles.com" <jamie@...ieiles.com>
Subject: Re: [PATCH 1/6] dmaengine/dw_dmac: Replace spin_lock_bh with irqsave
variants
On 04/27/2011 02:03 AM, Russell King - ARM Linux wrote:
> On Tue, Apr 19, 2011 at 11:55:45AM +0530, viresh kumar wrote:
>> This must have been added in dw_dma_chan instead :(
>> Will resend it.
>
> Even that doesn't work. It has to be a local variable.
Hello,
There are two kinds of locks taken in dw_dmac: spin_lock_bh and spin_lock.
spin_lock_bh is taken from routines called from external drivers and spin_lock
taken inside tasklet.
For reasons mentioned earlier, spin_lock_bh has to be replaced with spin_lock_irqsave.
Now, Is simple spin_lock() sufficient inside tasklet?? As dma API's can be called
from interrupt context (which will try to take spin_lock_irqsave).
I think spin_lock must be replaced with irqsave variants, even in tasklet.
Now if we do that, do we need this tasklet at all??
--
viresh
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists