[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAGnW=BaU__+B2uAK=qpV3QP60SqSUxx-kgJVs0t42K0fqTxL0g@mail.gmail.com>
Date: Thu, 20 Aug 2015 12:01:27 +0530
From: punnaiah choudary kalluri <punnaia@...inx.com>
To: Vinod Koul <vinod.koul@...el.com>
Cc: Punnaiah Choudary Kalluri <punnaiah.choudary.kalluri@...inx.com>,
"robh+dt@...nel.org" <robh+dt@...nel.org>,
"pawel.moll@....com" <pawel.moll@....com>,
"mark.rutland@....com" <mark.rutland@....com>,
"ijc+devicetree@...lion.org.uk" <ijc+devicetree@...lion.org.uk>,
Kumar Gala <galak@...eaurora.org>,
"michal.simek@...inx.com" <michal.simek@...inx.com>,
dan.j.williams@...el.com, dmaengine@...r.kernel.org,
"devicetree@...r.kernel.org" <devicetree@...r.kernel.org>,
linux-arm-kernel@...ts.infradead.org,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
Punnaiah Choudary <kpc528@...il.com>
Subject: Re: [PATCH v4 2/2] dma: Add Xilinx zynqmp dma engine driver support
On Thu, Aug 20, 2015 at 11:43 AM, Vinod Koul <vinod.koul@...el.com> wrote:
> On Thu, Aug 06, 2015 at 08:49:33AM +0530, Punnaiah Choudary Kalluri wrote:
>
>> + list_for_each_entry_safe(desc, next, &chan->done_list, node) {
>> + dma_async_tx_callback callback;
>> + void *callback_param;
>> +
>> + list_del(&desc->node);
>> +
>> + callback = desc->async_tx.callback;
>> + callback_param = desc->async_tx.callback_param;
>> + if (callback) {
>> + if (in_interrupt())
>> + spin_unlock_bh(&chan->lock);
>> + else
>> + spin_unlock(&chan->lock);
>
> This looks bad!
> Why would callback be called from different context. It should only be
> invoked from your tasklet
During the terminate call, driver need to clean up the existing BDs so that time
this function will be called from the thread or process context in
addition to the
tasklet context.
DO you have any suggestion here ?
>
>> +static int zynqmp_dma_device_terminate_all(struct dma_chan *dchan)
>> +{
>> + struct zynqmp_dma_chan *chan = to_chan(dchan);
>> +
>> + spin_lock_bh(&chan->lock);
>> + zynqmp_dma_reset(chan);
>> + spin_unlock_bh(&chan->lock);
>
> No descriptor cleanup
zynqmp_dma_reset will do the descriptor cleanup.
>
>> +static void zynqmp_dma_chan_remove(struct zynqmp_dma_chan *chan)
>> +{
>> + if (!chan)
>> + return;
>> +
>> + devm_free_irq(chan->zdev->dev, chan->irq, chan);
>> + tasklet_kill(&chan->tasklet);
>> + list_del(&chan->common.device_node);
>
> not deregistering with dmaengine?
This i am doing it in zynqmp_dma_remove function. Each channel will be
a standalone dma device for ZynqMP DMA case
>
>> + zdev->chan = chan;
>> + tasklet_init(&chan->tasklet, zynqmp_dma_do_tasklet, (ulong)chan);
>> + spin_lock_init(&chan->lock);
>> + INIT_LIST_HEAD(&chan->active_list);
>> + INIT_LIST_HEAD(&chan->pending_list);
>> + INIT_LIST_HEAD(&chan->done_list);
>> + INIT_LIST_HEAD(&chan->free_list);
>
> You can simmplify this by using vchan framework!
I got your point . but i have already said my reasons why i am
reluctant to use vchan framework in v3 review.
>
>> +MODULE_AUTHOR("Xilinx, Inc.");
>> +MODULE_DESCRIPTION("Xilinx ZynqMP DMA driver");
>> +MODULE_LICENSE("GPL");
> No alias, how did it get loaded?
I will fix this. thanks.
Regards,
Punnaiah
>
> --
> ~Vinod
>
--
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