2012/2/26 Måns Rullgård <m...@mansr.com>:
> Felipe Contreras <felipe.contre...@gmail.com> writes:
>> This patch series is a first try of implementing support for TI's OMAP3 DSP
>> algorithms. These algorithms are privided officially by TI and distributed in
>> many products like the Nokia N900, and also publicly for non-commercial
>> purposes[1]. The interface to access then is through tidspbridge driver is is
>> linux's staging area[2].
>
> I still object to naming this "tidsp".

The library is named libtidsp, that's outside of the purview of libav,
and it's my call how to name it. It's named after the tidspbridge
driver. If you don't like the name of the driver, complain on the
staging mailing list.

> Furthermore, it is quite clear
> that the tidspbridge interface is dead, used only on abandoned products
> like the N900.

First of all, the tidspbridge driver is the *only* DSP-related driver
that lives on the vanilla Linux kernel (drivers/staging/tidspbridge),
and it's the *only* one that has the chance to be merged to the main
tree.

Secondly, it's used in all kinds of phones, including Android phones.
They are old because they use OMAP3.

Thirdly, the N900 is not abandoned, I, and many other people, still
have and use their N900, and there's an active community working on
it.

Not to mention the myriad of other OMAP3 boards out there, and even
some DaVinci boards that have the same DSP.

The only other DSP-related interface that has a chance of making it to
upstream, is remoteproc, and as you can see, it doesn't have TI, or
OMAP, or DSP in it's name, so it can't possibly interfere with tidsp.

Possibly in the future tidspbridge would be converted to remoteproc,
and probably there would be a v4l2 interface to it. But there's a long
way to go for that to happen.

In the meantime, libtidsp is here, and it works, and there's no other
interface that is in the vanilla Linux kernel today, nor in the planed
near, or far future. There's only tidspbridge, and remoteproc, that's
it.

Cheers.

-- 
Felipe Contreras
_______________________________________________
libav-devel mailing list
libav-devel@libav.org
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to