Re: x86 hal_*.c #if'ed out?

On 11/24/2015 05:37 AM, Sebastian Kuzminsky wrote:
> I'm working with a current checkout of Volcano.
>
> Why are the entire contents of base/arch/x86/hal/hal_32.c and hal_64.c
> disabled with "#if 0"?
Cal_32.c and  hal_64.c have been cleaned and unified in just one hal.c.
They are still in the making but, as you said, emptied in fact.
It is just a quick-dirty hack of mine to avoid editing the make 
structure, which needs seme cleaning.
In any case they do not create any trouble.
> The application i'm working on, LinuxCNC, uses some of the interfaces
> that hal_32.c used to export, specifically rt_free_timer().
>
> Is it intentional that those files are compiled, but effectively empty?
>   I see a ChangeLog entry saying the old timer request/free functions
> were removed.
>
> Should I switch to some other interface, maybe hal.c's rt_free_timers()?
RTAI-5 has a break in the way the pacth is used, no more oll legacy 
stuff is kept.
In such a view, without loosing any functionality and interface, it now 
exploits the hard timer setup available in the patches, which allows all 
of what was previously done in its hal without the need of recoding the 
same things.
So now there are just rt_request/free_timers.
Therefore reason you do not see the above mentioned function is that I 
did thought that the hard timers should have not been used at its level, 
as they were just for start/stop_rt_timer found in the scheduler.
In any case I would like to not break any established usage of interest. 
So, if you give mi a list of missing functions
I'll care to provide them again through an appropriate wrapping with the 
core supports.

Paolo


_______________________________________________
Rtai mailing list
[email protected]
https://mail.rtai.org/cgi-bin/mailman/listinfo/rtai