Last old PowerMac G5 that I have access to may be starting to fail: iichb0: I2C error

[Mark Millard]

On 2021-Apr-16, at 15:33, Kevin Bowling <kevin.bowling at kev009.com> wrote:

> I think there have been generic i2c changes lately too, Mark if you run an
> older release does it have the same symptom?

An issue is that failure is not quick. It might have to operate
for, say, 2+ weeks to be clearly well beyond what I've been
seeing lately. Not exactly a fast test --making additional
testing (such as of multiple versions) painful as well.

Any clue what timeframe or version is appropriate? Randomly
picking older releases would not seem appropriate.

For reference, it is running based on 7381bbee29df from
back on 2021-03-12, but, as usual for me, it is running
with my workaround patches for PowerMacs as well. Also:
a non-debug kernel/world (with symbols, however).

# ~/fbsd-based-on-what-freebsd-main.sh 
FreeBSD FBSDG5L2 14.0-CURRENT FreeBSD 14.0-CURRENT mm-src-n245445-def0058cc690 GENERIC64vtsc-NODBG-dcons  powerpc powerpc64 1400005 1400005
def0058cc690 (HEAD -> mm-src) mm-src snapshot for mm's patched build in git context.
merge-base: 7381bbee29df959e88ec59866cf2878263e7f3b2
merge-base: CommitDate: 2021-03-12 20:29:42 +0000
7381bbee29df (freebsd/main, freebsd/HEAD, pure-src, main) cam: Run all XPT_ASYNC ccbs in a dedicated thread
n245444 (--first-parent --count for merge-base)

> On Fri, Apr 16, 2021 at 1:42 PM Brandon Bergren <bdragon at freebsd.org> wrote:
> 
>> Do you have a dmesg? There's some bits in the thermal drivers I've been
>> looking into recently where I have certain machines where the thermal chip
>> is left in automatic mode but also the system attempts to drive it in
>> manual mode, which causes things like the fans jerking on and off and
>> spurious 103.0C readings on the temperature probe the chip is internally
>> sampling. I half wonder if there's a similar issue causing us to
>> accidentally crash the chip or something.
>> 
>> On Thu, Apr 15, 2021, at 10:53 PM, Mark Millard via freebsd-ppc wrote:
>>> The 2-socket/1-core-each PowerMac G5 has had several
>>> sudden crashes recently, mostly while idle, usually
>>> with no messages. No response but to power controls
>>> once it fails.
>>> 
>>> But I got one example (so far) where it reported
>>> something as it failed (and the fans then started
>>> to speed up):
>>> 
>>> iichb0: I2C error
>>> 
>>> The message was even recorded in /var/log/messages .
>>> 
>>> The kernel and world have not been updated since
>>> back around 2021-Mar-12 or so. So the behavior
>>> appears to be new, without a system software change
>>> in a similar time frame.
>> 

===
Mark Millard
marklmi at yahoo.com
( dsl-only.net went
away in early 2018-Mar)