i’ve been getting a fully blank screen and a boot loop, so it sounds like you got further than i did?
Hmm interesting. I found a dracut issue in the config file I ship in pocket-reform-base. I’m testing it now before I push but if you want to try it out running
dracut -H --force --fstab
fixed my issue. Give it a shot!
Still having problems with missing drivers in the kernel package. Going to try cross-compiling from x86 Void, but at this point it’s hard not to feel like this thing is conspiring against me.
Still no luck. : (
Cross compiling from x86 also gives me the same issue — no drivers, no device tree.
I’m battling a bunch of dracut and extlinux demons right now. I apologize for how long this is taking me to resolve
1 Like
All good, there’s at least some solace knowing I’m not the only one struggling through it.
Alright the kernel definitely compiles correctly now from my latest commit. I need to fix the WiFi drivers next but the kernel is booting from a fresh compile now
1 Like
Okay, coming back to this after a two-month break.
Started over fresh, built a rootfs and installed to SD card, and still having similar issues. Dracut doesn’t throw any errors when building the ramfs anymore, but I still get a black screen on boot. Out of the ones in the dracut conf, I can’t seem to find the drivers pwm_imx27, mxsfb, and imx8mq-interconnect either in the ramfs or on-disk. From the patched modules, I can’t find any of the patched drivers for mmc/core.
I actually have been working on this recently.
I updated my pocket-reform-kernel package in my void-packages repo
Be sure to fixup extlinux after you install that package. That is the kernel I’m currently running on my pocket reform and it’s working great, so that should kill any issues with the kernel stopping you from booting
1 Like
Howdy yall, While I’ve been waiting for my mnt pocket from crowd supply. I’ve done a little research into the reform-tools to see if any of them would be compatible without too much work. @joe-albanese does the pocket suspend properly without reform-standy? I’ve taken a look into the script for suspending bin/reform-standby · main · Reform / MNT Reform Tools · GitLab while void doesn’t have systemd, it still has execute on suspend/resume folders in /etc/zzz.d which can be used to execute reform-standby properly
as far as i’m aware, the pocket does not appear to support ACPI sleep states, so zzz(8) does not work out of the box. Even suspending to disk (S4 sleep, ZZZ) does not appear to work on mine; the system just boots as normal.
hello @TwoWheeledIdiot and welcome to the forum!
The question is not “does the pocket suspend” but “does SoM XY suspend” which again is also not technically the right question, a much better one is “does SoM XY not only suspend but also resume”? 
The answer is: right now, only imx8mq can successfully suspend and resume. minute has on their todo list to investigate and make suspend and resume work for rk3588 for the reform next. At that point, it is very likely that the issue will also be solved for rk3588 in the classic reform as well as pocket reform.
We tried suspend/resume with rk3588 during 39c3 with gordon and while it does suspend successfully, resuming it fails due to problems with pci wake.
A311D also suspends successfully but doesn’t resume and while minute does not plan to work on the issue with A311D, user f_ in IRC said that being able to resume “should” be a simple fix of the wake-up source of ATF.
1 Like
Ah thanks for letting me know, my mnt pocket arrived a few hours ago, does booting from nvme work, or having boot file on an sd card and pointing to the nvme for root/home work
I seem to be having an issue with cloning your repo @joe-albanese, I’ve tried several times trying to clone it and your packages are missing every time I attempt cloning it, I’ve manually downloaded the kernel, kernel-headers, and pocket-reform-base and the kernel, kernel-headers keep failing because kernel-headers template file doesn’t exist
feels like a silly question but i forget it every time i clone, are you on the pocket-reform-imx8mp branch on the repo when you’re trying to package them?
ymmv since i don’t have an NVMe drive in mine, but i’ve found all the pocket reform needs to boot from a particular drive is an appropriately set up extlinux/extlinux.conf in its first partition, and at least one drive in the system with a valid boot.scr in the root of its first partition.
That is not entirely correct. u-boot works with either extlinux.conf or boot.scr and it will prefer the former over the latter but it does require both. It needs either in a partition on the drives it checks and that partition does not need to be the first. The boot order is: sd-card, nvme, emmc, usb.
1 Like
I figured that out at 1 am today, pulled the right fork and ran xbps-src for everything needed. Ill be installing everything to the chroot and setting up ext.conf and kernel cmdline for the encrypted root
1 Like
First test boot attempted, I was able to get a uboot screen to pop up when I connected a micro hdmi to the microhdmi port but no further behavior
I have display output, display is rotated the wrong way and I have initramfs failing to find my xfs partition