06:21 <hanetzer> eballetbo[m]: present the emmc as mass storage. I don't know how the kmod maps to a particular port/etc

06:22 <hanetzer> also, anyone here can confirm if kexec is in a workable state for arm64/aarch64

08:40 <eballetbo[m]> @ma

08:40 <eballetbo[m]> @maz ^ (maybe he can answer your question)

09:16 <maz> hanetzer: kexec has now regressed due to the reversal of the VOP shutdown patch.

09:17 <maz> in general, kexec works pretty well, but there is a number of platform-specific caveats.

09:35 <hanetzer> maz: huh, can you point me at it?

09:37 <maz> hanetzer: at what exactly? the patch?

09:37 <hanetzer> maz: yeah, I wanna see the depth of changes that would be required. I'm a bit miffed that my work to get petitboot working as a bootloader for my chromebook was for naught :P

09:39 <maz> hanetzer: well, there is much more to it than just that. the lack of UEFI means that we can't preserve the GICv3 LPI tables across kexec, for example. which means you will loose all MSIs, and thus the wifi.

09:39 <hanetzer> ah

09:40 <maz> unless you boot with the magic option that prevents the use of LPIs in the primary kernel.

09:41 <hanetzer> huh. well, I get black screen and nothing particularly noticable after pb does its thing so its prolly in a funky state anyways

09:41 <maz> hanetzer: https://patchwork.kernel.org/patch/10556135/

09:42 <maz> hanetzer: note that this caused regressions on ChromeOS, so it was reverted.

09:44 <hanetzer> hm. well, perhaps I could just rig up a /etc/kernel/postinst.d script to generate the fit image/vbutil_kernel for now

09:45 <hanetzer> well, if nothing else this has shown me how to craft an initramfs for use with petitboot, so I can head forward with my intent to make this machine (ryzen 1700) multiboot via petitboot :)

13:12 <corkie> hey all, anyone here running linux on a RK3229 device? i have this old androidTV box i would love to be able to ssh into and use it for something else than collecting dust heh

13:24 <corkie> or could push me in the right direction on maybe compiling my own stuff, no idea what iam getting into here but i see Rockchip kernel 4.4 support RK322x , and i see custom_boards overs at the rockchip github has the rk3229 in there, so was hoping there was a tiiiny chance it would be possible to make something run on this box, just thinking out loud i guess x)

13:25 <DuClare> Hmm damn, hsic still not properly working after warm boot. Someone with a clue: is there a usual/common cause to things working after cold boot but not after reboot?

13:29 <Pix> usually the bootloader/driver that does not check the state of the chip before attempting to init it

13:30 <Pix> and usually chips does not enjoy being init 2 times in a row :)

13:38 <DuClare> Aha..

13:40 <DuClare> Somehow I kinda expected that getting init would amount to a reset, more or less. And that getting reset would be better than being in whatever state the chip was left in prior to boot

13:40 <DuClare> :)

13:40 <DuClare> Hmh