[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <87613w1sf5.fsf@belgarion.home>
Date: Sun, 30 Aug 2015 21:34:54 +0200
From: Robert Jarzmik <robert.jarzmik@...e.fr>
To: Guennadi Liakhovetski <g.liakhovetski@....de>
Cc: Mauro Carvalho Chehab <mchehab@....samsung.com>,
Jiri Kosina <trivial@...nel.org>,
Linux Media Mailing List <linux-media@...r.kernel.org>,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3 4/4] media: pxa_camera: conversion to dmaengine
Guennadi Liakhovetski <g.liakhovetski@....de> writes:
>> + last_buf = list_entry(pcdev->capture.prev,
>> + struct pxa_buffer, vb.queue);
>
> You can use list_last_entry()
Ok.
>> + last_status = dma_async_is_tx_complete(pcdev->dma_chans[chan],
>> + last_buf->cookie[chan],
>> + NULL, &last_issued);
>> + if (camera_status & overrun &&
>> + last_status != DMA_COMPLETE) {
>> + dev_dbg(dev, "FIFO overrun! CISR: %x\n",
>> + camera_status);
>> + pxa_camera_stop_capture(pcdev);
>> + list_for_each_entry(buf, &pcdev->capture, vb.queue)
>> + pxa_dma_add_tail_buf(pcdev, buf);
>
> Why have you added this loop? Is it a bug in the current implementation or
> is it only needed with the switch to dmaengine?
It's a consequence of the switch.
With dmaengine, a dmaengine_terminate_all() removes all queued txs. It is
therefore necessary to requeue them. In the previous implementation, the
chaining was still good, and it was "enough" to just queue the first
videobuffer : the other buffers would follow by chaining.
Cheers.
--
Robert
--
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