05:16 <t> is there a reason the camera and audio have ??? in compatibility. Is it that m1n1 is only logging calls to the GPU or is it just due to so many other issues would it be possible for me to work on something like camera driver?

06:29 <marcan> t: someone already got started on the camera, but it hasn't been looked at in a while I think

16:42 <eiln> jannau: fixed to handle sub pds

16:42 <eiln> updated t8103 dt bindings @ https://github.com/eiln/linux.git on "ane" branch

17:00 <maz> how long will it take to build defconfig on 4 M2 e-cores in a L2 guest? I'm taking bets.

17:01 <j`ey> 22 minutes

17:10 <null> I won't bet on the absolute time, but I will bet at least one RCU stall in the L2 guest

17:13 <maz> null: you can't change the rules, even today! ;-)

17:20 <maz> real 22m56.589s

17:20 <maz> user 82m0.547s

17:20 <maz> sys 8m57.296s

17:20 <maz> j`ey: you have been cheating.

17:21 <sven> lol :D

17:21 <maz> null: no stalls!

17:26 <null> hahaha

17:26 <null> congrats

17:28 <maz> now building bare-metal to get an element of comparison...

17:35 <j`ey> maz: :D

17:35 <j`ey> maz: do I win the m2?

17:36 <maz> j`ey: I wouldn't mind, it's not even mine!

17:37 <j`ey> ;-)

17:44 <maz> bare metal:

17:44 <maz> real18m36.761s

17:44 <maz> user70m53.127s

17:44 <maz> sys3m5.665s

17:46 <maz> less than 25% overhead. given that the L2 is a 4kB guest, it doesn't look too bad.

18:53 <maz> sven: you definitely should be *given* one.

18:56 <maz> 16k guest:

18:56 <maz> real 21m19.267s

18:56 <maz> user 79m51.019s

18:56 <maz> sys 4m29.950s

18:56 <maz> but I'm hitting the swap... time for a bigger VM!

19:24 <maz> real 19m46.287s

19:24 <maz> user 73m49.162s

19:24 <maz> sys 4m33.061s

19:24 <maz> 6% overhead when not swapping. maybe this shit works after all...

23:41 <t> is it in a staging branch or something?

23:48 <j`ey> t: if you mean the camera, there's proxyclient/hv/trace_isp.py

23:49 <j`ey> and stuff here https://github.com/kellermanrivero/m1n1/tree/isp