02:33 <ndufresne> narmstrong, I'm been juggling with kernel branches, what do you use as a base these days, I tried multiple branches, the last was linux-amlogic branch integ, but it oops at boot in swapper ...

02:34 <ndufresne> before that I was on mjourdan 5.0 branch, which oopsed in kvm, that I disabled, but then it oopsed if I enable in-kernel drm driver ...

02:38 <ndufresne> hmm, same kernel, just rebuilt, now it crash in kvm again, let's disable that ...

03:30 <ndufresne> hmm, ok, so I tested the kernel on the potato, and that worked, while the h96pro, it mount the nfs successfully, try to start init and fails, panic

03:31 <ndufresne> could be a network issue

03:35 <chewitt> https://github.com/chewitt/LibreELEC.tv/commit/f2081728209a602e8795b81ff4787e90cf4ef6e9

03:35 <chewitt> ^ that's what I'm using for 5.0

03:36 <chewitt> https://github.com/chewitt/LibreELEC.tv/commits/amlogic

03:36 <chewitt> ^ using standard upstream 5.0 kernel and some patch-sets

03:36 <chewitt> look at commits with linux: prefix

03:37 <chewitt> no need for lima on a panfrost board

03:38 <chewitt> the narmnstrong patch-set is a general collection of essential bits that are en-route upstream

03:38 <chewitt> the mjourdan patch-set is all the vdec bits

03:38 <chewitt> the kwiboo patch-set has changes for 2.0 audio

03:39 <chewitt> the "temp patches" set is a collection of bits being validated that need to go upstream

04:14 <ndufresne> yeah, the kernel I have now should be fine (appart from having to remove kvm support)

04:15 <ndufresne> well, it'S fine for an S905X, maybe there is some patches as you say that, some cleanup

04:15 <ndufresne> chewitt, yeah, I'll give that a try tomorrow

04:27 <ndufresne> chewitt, right now my guess is that it's the drm driver fixes that blocks me

04:28 <ndufresne> thanks for the pointers