Re: [Xenomai-core] [PATCH-STACK] Synchronised timebases and more

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

[-- Attachment #1: Type: text/plain, Size: 2298 bytes --]

Gilles Chanteperdrix wrote:
> Jan Kiszka wrote:
>  > I was afraid you would insist on this support. ;)
> 
> Well, clock_settime is almost the only service missing in posix skin,
> and since I saw in Thomas Gleixner slides that one reason for not using
> Xenomai is that its posix support is incomplete, I am eager to implement

Well, those slides are from a time when Thomas already made up his plans
(I watched his first LibeRTOS presentation in 2003...).

> the missing services (and to remove the sentence "xenomai posix skin is
> a work in progress" from posix skin text file).

Anyway, your goal is valid.

> 
>  > 
>  > There are two ways to implement this:
>  > 
>  >  A) The poor man's variant
>  > 
>  >     On xntbase_adjust_time() (the code will change again, pay attention!
>  >     ;) ), iterate over all pending timers (or over all timers in the
>  >     base that POSIX uses?) and fix those which do not have the recently
>  >     introduced XNTIMER_MONOTONIC flag set. "Poor man" because it's
>  >     simple, but it scales poorly.
>  > 
>  >  B) The scalable but complex one
>  > 
>  >     Introduce a second time base for each existing one (or for the one
>  >     that POSIX uses?), put in all the adjustable (realtime) timers. We
>  >     then only need to play with the base's clock offset on adjustment,
>  >     but we would also have to include that offset into timeout
>  >     considerations inside the timer interrupt handler.
>  > 
>  > I wonder now if the number of use cases where people are playing with
>  > the wallclock all over the time while a significant amounts of timers
>  > are pending is actually worth the troubles of B)... What do you think?
> 
> I think one use of clock_settime would be to resync Xenomai clock with
> Linux one from time to time, but even if we implemented B, that would be
> a bad idea because of the effect on timers. So, A would be enough for me.

For the resync with any kind of external time source, I rather have a
scheme of one set-time during startup + continuous clock frequency
tuning in mind. As you say, permanently playing with the offset is _bad_.

So only the question remains if we should apply the timer adjustment on
all bases or only the POSIX-related one.

Jan


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