• Content Count

  • Joined

  • Last visited

About Flole

  • Rank
    Advanced Member

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. It does not work in the legacy branch, it just used to work on 4.13.x before I upgraded to focal (I think I was on bionic before but I need to check that again). I tried the legacy branch as I was hoping that it could be better, but unfortunately it isn't.
  2. Sure, I had to remove some sensitive information like IPv6 Addresses and NFS Paths though.
  3. This is what happened right before a crash, looks like it was adjusting the frequency quite often at that time. I would assume that many frequency adjustments are the cause of it and also that is the only thing I changed and since then it's running super stable. Not even my selfmade power supply that provides 5.25V 3A through the dc jack made a change, I still had the same crashes while the voltage that comes out of the USB ports is super stable (checked with an oscilloscope). Only after switching to the performance governor it started to run stable (at least until now, but's thats already lon
  4. I think I found the reason for my crashes: I used the ondemand power governor and apparently that caused the issues. Maybe the voltage regulator was set to an invalid level or something. Anyways, since I disabled it and used performance I now have had no crashes, that's an uptime of 8 days now. My absolute record was 14 days so I guess the next weeks will show if that was really the issue or not.
  5. I am running Armbian Focal with Linux 5.4.43-sunxi64 (and now downgraded to 4.19.125-sunxi64) on my Orange Pi PC 2. I use motion software to provide a "18ec:5555 Arkmicro Technologies Inc. USB2.0 PC CAMERA" to the network. I also use pjsip to provide SIP functionality for the same device as audio input. As soon as I start pjsip for some reason I get a very high load average until the kernel detectets stalled CPUs and finally the entire device crashes. This worked fine in a previous version and started after I upgraded to focal. I did a kernel backtrace and this is what I got: [ 401.002517
  6. That indeed disabled those messages. I have a few more, are those normal and necessary aswell or can I disable something to get rid of them:
  7. I assume there is no way to just disable the GPU completely? I don't need it at all, so if there is a way to disable it I would give that a try.
  8. Is anybody else experiencing stability issues? I have replaced the power supply now and double-checked with an oscilloscope to make sure the voltage is stable even under load. I have stable 5.25V at the USB Ports. I am using an USB HDD as rootfs, is anybody else using this with the system running stable? There are quite some regulator entries in the dmesg, but I think those are normal? I tried to enable the watchdog but that doesn't fix the lockups. My HDDs LED goes off when that lockup happens, the Ethernet LEDs stay on but it doesn't respond anymore at all, just as if the CPU
  9. But it is not "unsafe" to run it in this configuration, right? The regulators are all set now and there shouldn't be any need to change voltages now. Especially they shouldn't be set to an unsafe voltage, right? On the next reboot it'll become active again anyways but since last time I tried a reboot after unloading I got a complete lockup I would prefer to keep it running for 2 more days until I'm physically there again.
  10. I tried to unload i2c_mv64xxx using modprobe -r command and in dmesg I got [41723.377022] vcc-5v: Underflow of regulator enable count[41723.377036] vcc-5v: Underflow of regulator enable count[41723.377209] ------------[ cut here ]------------[41723.377223] WARNING: CPU: 1 PID: 23070 at drivers/regulator/core.c:5262 regulator_unregister+0xd0/0xd8[41723.377225] Modules linked in: nf_log_ipv4 nf_log_common xt_LOG xt_mac xt_conntrack iptable_filter xt_MASQUERADE xt_nat iptable_nat nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 tun bridge zram sun4i_i2s snd_soc_simple_card sunxi_cedrus(C) snd
  11. My Power Supply has a micro usb plug attached, so if I want to use the dc jack I would have to remove that one first Every connector adds additional resistance so I try to avoid it whenever it's possible.
  12. I think this board has quite some issues though, even when I go in with 5V on the Micro USB it's not enough to power a HDD on the USB Ports (and I'm not the only one with that issue), so I was wondering if this might be another issue this board has.... I guess I need a 5.5V Power Supply to account for whatever losses happen on the board. Is it better to go in through the DC Jack? I read about issues with that some time ago aswell and as I would have to cut my USB Cable I want to be absolutely sure that it's better before I try it.
  13. As you can read (some posts) above I am using 2 Power supplies and the voltage should be stable at the board. I'll try to run a stresstest though once I rebooted the board to see if that could be the issue. Just to confirm: For you this board is running stable?
  14. Unfortunately my Orange Pi 3 is still very unstable, even with the latest version. I get a maximum uptime of 4 days, and sometimes like now it won't even reboot automatically when it crashes. I'll try to see if it's possible to setup kdump tomorrow to see if this is caused by a panic (and what causes it) or the CPU simply goes into reset. I am running Armbian from a USB Drive, I need to check if maybe USB initialization fails and that causes the reboot to fail. Has anybody else experienced something like this?
  15. The air should be moving upwards by itself but anyways, now that I know that this is a known problem I prepared a 40mm fan with a USB plug that I can just mount on my enclosure and it will blow straight onto the heatsink. That should be enough cooling to keep it under 80°C. Another thing I noticed: I have a 5.2V 3A Power supply connected to the Micro USB connector. On the USB3 Ports I have a HDD connected and that HDD just gets 4.7V, so either some rails are a little too thin on the board or the Micro USB connector adds quite some resistance. Has someone seen that before? Is it bet