IRC log of #cubox of Fri 21 Apr 2017. All times are in CEST < Back to index

13:37 topi` Hit any key to stop autoboot: 0
13:37 topi` ## Executing script at 17f00000
13:37 topi` Wrong image format for "source" command
13:38 topi` what kind of error is this? My HB still boots successfully...
13:38 topi` autoboots, even
13:54 vpeter Seems junk at that address?
14:44 topi` what should there be in that address?
14:45 topi` btw there's something odd or buggy in uboot's ext2 support... I have symlinks from zImage.abc -> zImage and if I try to ext2load that zImage (via symlink) uboot says "File not found zImage"
14:45 topi` if I mount it under linux, the symlink is there and works fine
14:46 topi` no, that zimage was mis-linked, but this still applies to my imx6q-hummingboard2-som-v15.dtb -> ... symlinks
14:46 topi` maybe there is file name length limit in uboot :)
14:51 vpeter at that address should be loaded bootscript.
15:00 topi` and if it fails to run that bootscript, it runs the "builtin" bootscript that was bolted in at uboot compile time?
15:26 topi` does anyone have a working coda module? if I try "modprobe coda", I end up with "firmware request failed"
15:27 topi` even though I have installed firmware-imx_5.2+pkg1 from repo.solid-build.xyz
15:27 topi` Direct firmware load for v4l-coda960-imx6q.bin failed with error -2
15:28 topi` ah, that filename simply does not exist in /lib/firmware
15:43 jnettlet topi`, start using my imx6-next branch for u-boot. That was a commit that was demanded by a customer for their custom image that had to go to production.
15:44 jnettlet I veto'd it but due to their bad decisions there were no alternatives.
15:46 jnettlet actually hold off one second. git submodules just pushed something they shoouldn't have
15:47 jnettlet nevermind. that is just in my local repo. phew. just use imx6-next branch
16:22 topi` :)
16:22 topi` we use whatever uboot just happens to float our boat
16:22 topi` as long as it can load a kernel ;)
16:23 topi` I'm currently working on getting proper watchdog support so that if any mission critical tasks happen to die or get stuck, then we'd have the watchdog biting the hell out of the system
16:23 topi` systemd seems to support using hardware watchdogs and I can decide per-service whether or not the force-reboot is the right action
16:24 topi` unfortunately my colleagues are not great fans of systemd
16:36 topi` jnettlet: next thing I try to achieve, is using coda driver to accelerate h.264 decode. I got some 2.1.4 firmware and the driver thinks it is unsupported
16:37 topi` hacked the driver a bit to stop complaining, but now I need gstreamer1 to test the coda stuff out. Installing gstreamer would bring in 700+ megabytes of crap, which makes me shudder