Hi!
First of all: Thanks for your hints, but...
I'm using the same ethernet-controller (eepro100) and no frame-buffer
devices. The problem also occurs in runlevel 1 with no ethernet driver
loaded. There are no modules except the rtl-ones...
I'm running no X-windows!
Does anybody have any more ideas?
Thanks a lot for your help
Maik
> I had similar problem with ethernet controller (Intel eepro 100 chips), I
> was told that enabling the frame buffer can be disastrous as well.But since
> I don't use it, I can't have that problem ;-). Does the problem arise in
> single user mode with ethernet disabled. (ie after running init 1).
>
> If it's still the case, m you should check if third-party driver aren't
> playing it safe (i.e explicit use of cli/sti instruction instead of the
> macros ....).
>
> Running : objdump --disassemble nameofdriver.o | grep cli
>
> should give you a hint.
>
> Hope it helps
>
> Laurent
>
>
>
>
>
>
> Maik Hassel <[EMAIL PROTECTED]>@rtlinux.org on 07/02/2002 02:06:37 PM
>
> Please respond to [EMAIL PROTECTED]
>
> Sent by: [EMAIL PROTECTED]
>
>
> To: [EMAIL PROTECTED]
> cc:
> Subject: [rtl] Inexplicably realtime violations - need help
>
>
>
> On one of my test-systems I get pretty regular realtime-violations, which
> raise the strong suspicion of a hardware problem, but the System is a
> "normal" Intel Systems like many others that don't have this problem, and
> there are no other IO-Cards etc. except some standard on-borad IOs
> including
> an ethernet controller.
>
> The same routine runs fine on many other systems (it should - its
> simple!)...
>
> Of course, this problem affects my interrupt-handler written for an IO-card
> (which wasn't build-in during the timing-test!) in a way that the handler
> is
> called with a delay which relates to the relatime-violations...
>
> I tried play with bios-setting concerning the speed-step stuff or
> power-management, but there was no effect at all.
>
> Does anybody have any idea where these violations could come from?
> Any help ist appreciated. If possible, please send a copy of your answer to
> [EMAIL PROTECTED]
>
> I testet with the following routine:
>
>
> ------snip-----------------
>
> // Wait for next cycle
> pthread_wait_np();
>
>
>
> now = gethrtime();
> diff = now - erwartet ;
> erwartet = now + Thread->Periode;
> mittel += diff;
>
> if (diff > (int)(Thread->Periode*1100))
> PRINTL("Echtzeitverletzung in Thread Nr. %d
> Durchgang
> %d: %d us statt %d us\n", Thread->Nr, nCount, diff, Thread->Periode*1000);
>
> if (diff < zeit.min)
> zeit.min = diff;
> if (diff > zeit.max)
> zeit.max = diff;
>
> if (++nCount%1000 == 0)
> {
> zeit.mittel = mittel/1000;
> // tf_put(FIFO_TIMING, &(Thread->Nr),
> sizeof(pthread_t));
> // rtf_put(FIFO_TIMING, &zeit, sizeof(zeit));
>
> zeit.min = 1000000000LL;
> zeit.max = -1000000000LL;
> mittel = 0;
> }
> }
>
>
>
> ---------snap--------------------
>
>
> .... and get
>
> Jul 1 21:00:48 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 377: 3035916 us statt 500000 us
> Jul 1 21:00:48 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 1407: 2975948 us statt 500000 us
> Jul 1 21:00:48 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 2439: 2926956 us statt 500000 us
> Jul 1 21:00:48 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 3471: 2917196 us statt 500000 us
> Jul 1 21:00:49 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 4503: 2838892 us statt 500000 us
> Jul 1 21:00:49 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 5533: 3304716 us statt 500000 us
> Jul 1 21:00:49 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 6563: 3250028 us statt 500000 us
> Jul 1 21:00:50 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 7593: 3236876 us statt 500000 us
> Jul 1 21:00:50 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 8623: 3165132 us statt 500000 us
> Jul 1 21:00:50 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 9653: 3134412 us statt 500000 us
> Jul 1 21:00:50 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 10683: 3080396 us statt 500000 us
> Jul 1 21:00:51 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 11713: 3063916 us statt 500000 us
> Jul 1 21:00:51 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 12743: 2995276 us statt 500000 us
> Jul 1 21:00:51 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 13773: 2957964 us statt 500000 us
> Jul 1 21:00:51 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 14805: 2908460 us statt 500000 us
> Jul 1 21:00:52 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 15837: 2893708 us statt 500000 us
> Jul 1 21:00:52 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 16869: 2821676 us statt 500000 us
> Jul 1 21:00:52 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 17899: 3284908 us statt 500000 us
> Jul 1 21:00:52 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 18929: 3258604 us statt 500000 us
> Jul 1 21:00:53 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 19959: 3203340 us statt 500000 us
> Jul 1 21:00:53 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 20989: 3151084 us statt 500000 us
> Jul 1 21:00:53 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 22019: 3118828 us statt 500000 us
> Jul 1 21:00:53 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 23049: 3088268 us statt 500000 us
> Jul 1 21:00:54 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 24079: 3029964 us statt 500000 us
> Jul 1 21:00:54 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 25109: 2979436 us statt 500000 us
> Jul 1 21:00:54 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 26141: 2943244 us statt 500000 us
> Jul 1 21:00:54 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 27173: 2914316 us statt 500000 us
> Jul 1 21:00:55 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 28205: 2860268 us statt 500000 us
> Jul 1 21:00:55 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 29235: 3310348 us statt 500000 us
> Jul 1 21:00:55 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 30265: 3271692 us statt 500000 us
> Jul 1 21:00:55 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 31295: 3238860 us statt 500000 us
> Jul 1 21:00:56 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 32325: 3186796 us statt 500000 us
> Jul 1 21:00:56 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 33355: 3134508 us statt 500000 us
> Jul 1 21:00:56 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 34385: 3100972 us statt 500000 us
> Jul 1 21:00:57 siemens kernel: Echtzeitverletzung in Thread Nr. 5
> Durchgang
> 35415: 3072012 us statt 500000 us
>
>
>
>
>
>
>
>
>
>
> -- [rtl] ---
> To unsubscribe:
> echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
> echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
> --
> For more information on Real-Time Linux see:
> http://www.rtlinux.org/
-- [rtl] ---
To unsubscribe:
echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
--
For more information on Real-Time Linux see:
http://www.rtlinux.org/
