Re: [Xenomai-core] First version for PCI framegrabber skeleton

[Jan Kiszka <jan.kiszka@domain.hid>]

[-- Attachment #1: Type: text/plain, Size: 2175 bytes --]

Rodrigo Rosenfeld Rosas wrote:
> First of all I would like to state I'm glad to write to this list once more :)
> 
> Secondly, I would like to apologize for the late that I'm sending you this first version of the
> old promised skeleton for the real-time PCI framegrabber I've developed last year...
> 
> Thirdly, I would like to apologize for it being not organized nor documented, but I'll explain the
> reasons for that.
> 
> Since I've finished my Master Thesis, I have been in a hurry and at this moment I'm working from
> 8h to 22h from monday to thursday and from 8h to 18h at friday in three different jobs... So, it
> is true... I do not have time at the moment for writing the skeleton the way I want. But since I
> received the message below yesterday (actually it was sent a week before but I didn't have time to
> check my other mail), I tried to help him.
> 
> I tried to get some free time at my job today for organizing a bit and mainly for removing the
> specific DataTranslation code from the skeleton because I do not have permission yet to send the
> full code. I cannot assure the attached code will even compile, since I work on a Windows box at
> work... Since I'm going to São Paulo tomorrow morning (6 am), in a business trip, I hope some of
> you could assist him for any doubts due to code desorganization and lack of documentation...
> 
> Hope I can help with this code. Sorry, but I cannot do more than that until monday, when I come
> back.

Your effort toward an RT framegrabbing driver interface is very welcome,
and I hope it can lay the ground for further work in this domain, also
via different media like FireWire or USB.

Just one issue remains as you already apologised for: this code needs a
/little/ bit cleanup. I tried to get an impression by browsing through
it, but it's fairly hard to read even for someone who thinks to know
what to look for. Too much experimental or dead code.

Maybe, once time permits, you can start the cleanup by reducing the
driver to a level that only leaves the removed HW access as markers,
killing everything that is not used right now in your real driver.

Jan


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 250 bytes --]