Test Suite summary

Umar Qureshey uq at ecs.fullerton.edu
Thu Oct 28 23:25:44 CEST 2004


Hi,

The second output is from running testsuite/kern/latency/.
When I run that, I get the output as stated in previous email.  And it
keeps looping with that output forever.  Other times, the system freezes
with no output.  One time I saw:

=========
=========
=========

in an infinite loop.

Should I try to turn off APIC support in the kernel?  Since I am
cross-compiling on an i686 for an i486 maybe this causing some issues
although I know the kernel detects the presence or absence of this part.
Perhaps, RTAI is getting confused?  I do not think my 486 has an APIC.

Paulo, I will run other tests in testsuite/ and post results before
tinkering with the kernel config.

Regards,

Umar Qureshey


On Thu, 28 Oct 2004, Paolo Mantegazza wrote:

> Umar Qureshey wrote:
> > Hi,
> >
> > I am running RTAI 3.1 w/ 2.6.8.1 kernel.  The hardware is an Intel
> > 80486-DX4 75 MHz laptop with 24 MB RAM.  Kernel has been compiled with
> > Processor Type -> 486.
> >
> > With kern/preempt:
> > ~
> > Adeos: Domain RTAI registered.
> > RTAI[hal]: 3.1 mounted over Adeos 2.6r7/x86.
> > RTAI[hal]: compiled with gcc version 3.3.2 (Mandrake Linux 10.0
> > 3.3.2-6mdk).
> > RTAI[malloc]: loaded (global heap size=131072 bytes).
> > RTAI[sched_up]: loaded.
> > RTAI[sched_up]: fpu=yes, timer=periodic.
> > RTAI[sched_up]: standard tick=1000 hz, CPU=1193180 hz.
> > RTAI[sched_up]: timer setup=1676 ns, resched latency=2514 ns.
> > Error opening /dev/rtf/0
> >
> >
> > CPU USE SUMMARY
> > # 0 -> 127305
> > END OF CPU USE SUMMARY
> >
> > RTAI[malloc]: unloaded.
> > RTAI[sched_up]: unloaded.
> > Adeos: Domain RTAI unregistered.
> > RTAI[hal]: unmounted.
> >
> > The program was unable to open /dev/rtf/0.  This could be because I am
> > compiling on a different machine and then running on a different one.
> > Which devices to create manually in /dev?  Anyone know?
> >
> >
> > With kern/latency:
> >
> > Adeos: Domain RTAI registered.
> > RTAI[hal]: 3.1 mounted over Adeos 2.6r7/x86.
> > RTAI[hal]: compiled with gcc version 3.3.2 (Mandrake Linux 10.0
> > 3.3.2-6mdk).
> > RTAI[malloc]: loaded (global heap size=131072 bytes).
> > RTAI[sched_up]: loaded.
> > RTAI[sched_up]: fpu=yes, timer=periodic.
> > RTAI[sched_up]: standard tick=1000 hz, CPU=1193180 hz.
> > RTAI[sched_up]: timer setup=1676 ns, resched latency=2514 ns.
> >
> > [system freezes for a while]
> >
> > Then I get a scrolling list of what looks like call traces...
> > [xxxxxxx] rcu_process_callbacks+0x0/0x130
> > [xxxxxxx] do_posix_clock_monotonic_settime+0x0/0x10
> > [xxxxxxx] do_posix_clock_monotonic_gettime+0x0/0x50
> > [xxxxxxx] __adeos_ack_common_irq+0x0/0x70
> > [xxxxxxx] irq_entries_start+0x0/0xe00
> >
> > (The above two entries repeat about 218 times with
> > irq_entries_start+0x10..0x20..0x30..0x40.........0xda0)
> >
> > [xxxxxxx] adeos_sync_console+0x0/0x60
> > [xxxxxxx] rtai_ssrq_trampoline+0x0/0x150  [rtai_hal]
> > [xxxxxxx] __adeos_set_root_ptd+0x0/0x20
> > [xxxxxxx] __adeos_get_root_ptd+0x0/0x20
> >
> > All this is scrolling at full speed on the screen so I could be not 100%
> > correct.  It's a good thing my full speed is only 75 Mhz allowing me to
> > read it somewhat at least.
> >
> > Does anyone have any clues??
>
> Can you run other examples? What happens with the latency test?
>
> Paolo.
>





More information about the Rtai mailing list