[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Response time
On Tuesday 22 October 2002 15.55, Arne Bockholdt wrote:
> Hi !
> could anybody on that list make a statement about the min/max/average
> response time for user programs and kernel drivers on the ETRAX LX system
I did some measurments a couple of days ago with some new features
in the gpio driver (setting "alarm" so you can do select() on the file
descriptor and wake up when bits have changed)
The test setup was something like this:
HW -> gpio driver-> select() app1 ->write() on AF_UNIX socket -> app2 doing
app2 uses the gpio driver to toggle an output pin and that pin is
connected to an input pin monitored by app1.
When detecting a high level on gpio port A (PA) that supports interupts
I get the latency 2ms from before app2 doing the ioctl() to toggle the pin
until it's select() returns (timing done with gettimeofday()).
The latency when detecting low level which is checked with the timer
interrupt I got 4-12 ms latency (timer only checks every 10 ms).
If I added a poll in cpu_idle() as well the latency dropped to 2 ms on
an otherwise fairly idle system. (webserver, ftp server etc. running but
not really any activity)
These 2ms latency includes 2 usermode processes and
the IPC between them - if you were to do something similar in
kernelspace you would get a lot less I guess.
> Is there any AXIS supported preemptive patch for the 2.4 kernel or a
> working version of the 2.5 kernel for the ETRAX platform ?
Not that I know of, although it has been discussed on this list before,
maybe someone has started working on it.
> Any other
> thoughts about how to perform time critical tasks on the ETRAX system ?
We have two Masters Theses workers that are looking at the different options,
and I believe they are doing some protoyp work with RTAI.
> Thanks in advance,
> Arne Bockholdt
> Dipl. Inform. Arne Bockholdt
> REA Elektronik GmbH
> Teichwiesenstr. 1
> 64367 Mühltal-Waschenbach
> Tel. 06154 / 638-115, Fax -195
Hope this helps