powerpc (non-64) 11.0-CURRENT -r280598 vs. iMac G3 and PowerMac G5

[Mark Millard]

I built a powerpc (non-64) 11.0-CURRENT -r280598 SSD (via gcc 4.2.1, not 4.9.1) and it has worked okay in G4 PowerMac's that I've tried. (I have no prior 11.0-CURRENT non-64 powerpc G4/G3 history, but do have a little for for the G5's.) I'm counting here PowerMac's that 10.1-STABLE works fine for, as did older builds going back to 10.0-STABLE.

But the SSD does not boot in the iMac G3 that I tried. Nor in the PowerMac G5's. Likely different reasons, even based on the little information that I have so far.

It may be some time before I get to having much information but what I have now is:



The below are all for the same SSD plugged in different PowerMac's/iMac's and then used to boot, not changing the software between times.



PowerMac G4's:

No problems so far.

(Note: I've access to one PowerMac G4 that I'm ignoring here. While Mac OS X and Lubuntu have never had troubles with it, every FreeBSD build that I've tried from back in 10.0-STABLE times when I started to now hangs during the boot sequence, fairly early.)



iMac G3 (I've access to only one example G3):

After

> Time counters tick every 1.000 msec


it gets:

> [Thread pid 0 tid 100037]
> Stopped at pmap_activate+0x7c lwz r11,r1,0x0

which may be reporting the instruction after a indirect subroutine jump.



PowerMac G5's:

Unlike the past where things worked overall the build now gets to the point of just mostly reporting timeouts, retries, and such for ada0 (where the SSD is plugged in). This is fairly early in the boot sequence.

Notably some ata support has been removed from 11.0-CURRENT according to the UPDATING notes, supposedly replaced with better support. But many FreeBSD statements implicitly mean "tier 1" as the implicit context and have other implications elsewhere. I've no clue in this case yet.

So far powerpc64 11.0-CURRENT -r280598 has not had such problems when used with the same PowerMac G5's.


===
Mark Millard
markmi at dsl-only.net