From: Jeroen Vreeken (ilea.hltil@siris.net)
Date: Fri Mar 12 2004 - 21:09:21 EET
On 2004.03.12 12:10:02 +0100 Jarmo wrote:
> As can be seen fom 2.2 to 2.6 HZ is multiplied by 10.
> And 6pack,what we use here
>
> /* default level 2 parameters */
>
> #define SIXP_TXDELAY (HZ/4) /* in 1 s */
>
> #define SIXP_PERSIST 50 /* in 256ths */
>
> #define SIXP_SLOTTIME (HZ/10) /* in 1 s */
>
> #define SIXP_INIT_RESYNC_TIMEOUT (3*HZ/2) /* in 1 s */
>
> #define SIXP_RESYNC_TIMEOUT 5*HZ /* in 1 s */
>
>
> Now if we take sixp_txdelay HZ/4 with this 2.6 kernel it gives
> value 2,5 seconds!
The kernel has an internal timer (called the jiffie counter) that counts in
a unit called a 'jiffie'.
HZ is the amount of jiffies that go in one second (and is thus the
frequency of the jiffie timer).
So HZ/4 should be 0.25 seconds.
If the 6pack driver does an 2.5 seconds wait it is either not due to the
HZ/4 counter or it has some pretty screwed up timing.
> So what we have done we have changed (HZ/4) to (HZ/40) and slottime
> value (HZ/10) into (HZ/100).
>
> A little bit more success to keep system up longer,but it crash still...
>
> And what we get into syslog
>
> Mar 5 16:46:58 oh1rbf kernel: [apm_do_busy+22/48] apm_do_busy+0x16/0x30
>
> Mar 5 16:46:58 oh1rbf kernel: [<c01148b6>] apm_do_busy+0x16/0x30
It might be coincedence, but could you try a kernel without apm?
>
>
> Question is.What we could do to get this mess working?
>
> McGyver help!
I have a pocket knife (and I have ductape lying within a few meters of me,
does that count? :)
Jeroen
-
To unsubscribe from this list: send the line "unsubscribe linux-hams" in
the body of a message to sil.fbyv@tortus.com
More majordomo info at http://vger.kernel.org/majordomo-info.html
This archive was generated by hypermail 2b30 : Fri Mar 12 2004 - 20:49:12 EET