Re: 970/PowerMac G5 cpudep_ap_bootstrap slb-related hangup *solved* . . .
Mark Millard via freebsd-ppc <[email protected]>
| Newsgroups | gmane.os.freebsd.devel.ppc |
|---|---|
| Message-ID | <[email protected]> |
On 2019-May-11, at 14:26, Andreas Tobler <andreast-list at fgznet.ch> wrote:
> On 10.05.19 12:23, Mark Millard via freebsd-ppc wrote:
>
>> Of course, whitespace details, may not be well preserved
>> below. (The commenting out of the two #if/#endif lines
>> was unnecessary and is not done in the below.)
>> # svnlite diff /mnt/usr/src/sys/ | more
>> Index: /mnt/usr/src/sys/powerpc/aim/mp_cpudep.c
>> ===================================================================
>> --- /mnt/usr/src/sys/powerpc/aim/mp_cpudep.c (revision 347003)
>> +++ /mnt/usr/src/sys/powerpc/aim/mp_cpudep.c (working copy)
>> @@ -68,6 +68,46 @@
>> case IBM970:
>> case IBM970FX:
>> case IBM970MP:
>> + /* Set HIOR to 0 */
>> + __asm __volatile("mtspr 311,%0" :: "r"(0));
>> + powerpc_sync();
>
> Thank you for this incredible investigation! My quad is now under heavy load and behaves as expected.
You are welcome.
One long-term issue is still definitely active,
happening sometimes:
The thread(s) stuck-in-sleep problem can still happen.
This is worse than just the fans and the buf*deamon*
thread timeouts during shutdown: those same threads
were not doing there on-going jobs earlier, starting
when they first got stuck. The fans and shutdown
messages (and wait time) just make it rather obvious.
The command:
top -HIStazopid
will frequently show the threads in question when
they are not stuck. But it will also not show the
ones that are already stuck-sleeping.
===
Mark Millard
marklmi at yahoo.com
( dsl-only.net went
away in early 2018-Mar)
_______________________________________________
[email protected] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ppc
To unsubscribe, send any mail to "[email protected]"