[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [microblaze-uclinux] Kernel BRAM usage

Hi,

John Williams wrote:
> Brettschneider Falk wrote:
> > John Williams wrote:
> >> [*] Allow placing code/data in BRAM
> >> [*]   Place interrupt entry path in BRAM
> > Interrupt latency from ISR to return from poll() in the 
> user app shortened to about 60% of the original time here. Excellent!
> 
> That's awesome, thanks for the feedback.  Low latency is what 
> we like to 
> see!
> 
> How's the jitter?  I'd expect to see it reduced as well since the IRQ 
> path is more deterministic (no cache misses).
216 measurings, min=417us, max=443us, avg=~429

CU, F@lk

P.S.: this measuring is based on JW's public patch plus an additional hack to privilege the user app waiting at poll() to be first scheduled after the IRQ happened. Together with SCHED_RR, this means real-time behaviour.

___________________________
microblaze-uclinux mailing list
microblaze-uclinux@xxxxxxxxxxxxxx
Project Home Page : http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
Mailing List Archive : http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/