Date: Thu, 12 Dec 1996 11:16:34 -0800 From: Amancio Hasty <hasty@rah.star-gate.com> To: Luigi Rizzo <luigi@labinfo.iet.unipi.it> Cc: CSP1DWD@MVS.OAC.UCLA.EDU, multimedia@FREEBSD.ORG Subject: Re: Intel's Smart Video Recorder III Message-ID: <199612121916.LAA01118@rah.star-gate.com> In-Reply-To: Your message of "Thu, 12 Dec 1996 19:12:06 %2B0100." <199612121812.TAA02155@labinfo.iet.unipi.it>
next in thread | previous in thread | raw e-mail | index | archive | help
>From The Desk Of Luigi Rizzo :
> > the driver.
>
> It would be a good idea, if the hardware permits, to keep (a subset)
> the same ioctls of the meteor, so that one can reuse the same
> software. I find annoying that there isn't a standard set of ioctls for
> video acquisition boards. Maybe those in the meteor driver are not
> better than others, but they are relatively flexible, and the
> meteor/omnimedia are the only boards which are well supported under
> FreeBSD, so it might be wise to settle on that as a standard.
>
> > capture chipset very similar to the Matrox Meteor however its
> > dma capability is far superior to the Meteor because the
> > Intel card's dma controller accepts "risc instructions" that is
> > the dma controller is programmable.
>
> can this be used to do some preprocessing of the data or it is just a
> flexible dma ?
>
My current test program for my device driver uses the same ioctls
as the Matrox Meteor. in fact , my test program is "tv" with no
modifications except for the name of the device which I use
the access my card.
The dma controller is just a flexible dma. In a nutshell, it has
the following instructions:
o write, write bytes to host memory
o sync, wait for a captured field
o skip, skip number of bytes
o jump, go to a particular host memory. Basically it is used to
to establish a loop.
The pre-processing that it can do is scaling, yuv to rgb , and dithering.
Cheers,
Amancio
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199612121916.LAA01118>