[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAKMK7uFTQx372YgDFJEu-hRBJdhA12dZYZDbfuLktNMjzvv9XA@mail.gmail.com>
Date: Wed, 22 Jun 2016 16:12:11 +0200
From: Daniel Vetter <daniel@...ll.ch>
To: Thierry Reding <thierry.reding@...il.com>
Cc: Tomeu Vizoso <tomeu.vizoso@...labora.com>,
Daniel Vetter <daniel.vetter@...el.com>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"dri-devel@...ts.freedesktop.org" <dri-devel@...ts.freedesktop.org>,
Emil Velikov <emil.velikov@...labora.com>
Subject: Re: [PATCH v1 2/3] drm: Add API for capturing frame CRCs
On Wed, Jun 22, 2016 at 3:32 PM, Thierry Reding
<thierry.reding@...il.com> wrote:
>> >> +const struct file_operations drm_crtc_crc_fops = {
>> >> + .owner = THIS_MODULE,
>> >> + .open = crtc_crc_open,
>> >> + .read = crtc_crc_read,
>> >> + .release = crtc_crc_release,
>> >> +};
>> >
>> > Do we want to support poll?
>>
>> Don't see the point of it, TBH.
>
> I have difficulty coming up with a concrete use-case, but given that we
> have most of the infrastructure to support it already, it might be nice
> to have. Could of course be added later on if there's a need.
O_NONBLOCK is definitely needed to be able to write testcases, so that
you can schedule flips and collect CRCs for them in parallel (e.g. to
validate atomic flips by making sure there's never an intermediate CRC
with garbage). I haven't seen a need yet for poll, and like you say
it's easy to add if we ever need it.
-Daniel
--
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch
Powered by blists - more mailing lists