On Sun, Sep 2, 2012 at 10:20 AM, Michael Haberler <mai...@mah.priv.at> wrote:
> I am just toying with an idea
>
> "simulator mode" LinuxCNC is kind of a misnomer - it really means 'compile 
> the whole thing on a standard linux kernel (=non-RTAI) and dont compile the 
> driver components"; the fine thing with "simulator mode" is - as long as you 
> have a linux system which can fulfill the package prerequisites, it should be 
> easy to compile
>
> however, I dont see a good reason why there cannot be a hal_parport component 
> which works in "simulator" mode, and I'm just integrating a 'parapin' library 
> into hal_parport.* to create such a driver; parapin being created by a 
> certain Mr John Elson. Not done, but havent hit a roadblock yet.
>
> here comes your "Linuxcnc on a Raspberry pi" ;) (provided somebody hacks a 
> driver for the GPIO pins there) - appalling RT  performance but a great 
> sprinkler system; fine as long as folks understand what they are getting into 
> in terms of latency
>
> the other reason: the whole exercise needs to be done for RT-Preempt support 
> anyway because these drivers will be pretty much userland processes (with 
> some hormon-treatment system calls sprinkled in, but nothing earth-shaking).

Excellent idea---FWIW, we used a non-RT system for quite ambitious
data acquisition (DAQ) system (300kHz peak event rate). The system had
a modest amount of hardware FIFO, and we developed it using regular
Linux kernel with RT scheduling for the DAQ task, planning to switch
to the RT kernel for production. We realized that performance with the
RT scheduling was adequate, so we never went to the custom kernel.

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to