gem0 instability

[Jason Bacon]

On 2019-05-25 15:53, Michael Tuexen wrote:
>> On 25. May 2019, at 14:45, Jason Bacon <bacon4000 at gmail.com> wrote:
>>
>>
>> Is this a known issue with powermacs?
>>
>> May 24 14:09:06 powermacg5 kernel: gem0: device timeout
>> May 24 14:09:06 powermacg5 kernel: gem0: link state changed to DOWN
>> May 24 14:09:11 powermacg5 kernel: gem0: link state changed to UP
>> May 24 14:09:18 powermacg5 kernel: gem0: device timeout
>> May 24 14:09:18 powermacg5 kernel: gem0: link state changed to DOWN
>> May 24 14:09:22 powermacg5 kernel: gem0: link state changed to UP
>> May 24 14:09:30 powermacg5 kernel: gem0: device timeout
>> May 24 14:09:30 powermacg5 kernel: gem0: link state changed to DOWN
>> May 24 14:09:34 powermacg5 kernel: gem0: link state changed to UP
>>
>> The gem0 interface has a tendency to become unstable under heavy load.  Rebooting alleviates it.
>>
>> Launching keepassxc (buildable thanks to Mark M's qt5-core patch) from a remote display seems to be a reliable trigger.
>>
>> This machine has some visible physical damage, so I wouldn't be surprised if it's a local hardware issue.  It's reliable otherwise, though.
> I do see the same on a machine (G5) without physical damage. So I guess it is a software issue.
>
> Best regards
> Michael
>
Thanks for the feedback.  Is anyone still working on Mac G5 driver issues?

This is not a major problem, but I'd be willing to help out if someone 
qualified is interested in fixing it.  I'm not qualified to take the 
lead, although I do have some experience with device drivers (DOS, 
embedded systems).

-- 
Earth is a beta site.