An upgrade from core 159 failed to reboot. I’ve tried two fresh installs of core 160, from nightly build, dated 20210920, to Sandisk uSD cards, with same result.
I then tried Friendlywrt, as a check of hardware - it does boot normally
U-Boot or the kernel is unchanged between core159 and core160 so normally this should not change the boot process,
Have you any debug output?
Also i not have a core160 from this date.
The latest is Index of /master/latest/armv6l
I could not get any display on Putty during the boot, although red LED on the nanopi R1 flashed blinking then solid, as it does during boot. Then back to blinking, with no green LED, all of which is the outcome of failed boots.
My nanopi R1 is from the second variant and has been very unreliable booting IPFire, but had no problem with openWRT. I often have to unplug & replug the power lead twice, before it boots sucessfully. Its WiFi was not found by the first core that supported nanopi R1.
My bad on the date - I referred to the date on the folder, when it should have been date on the file, which is 20210919, I double-checked by the md5sum that I had noted down.
A flashing RED led means that the kernel has bootet. The green leds are initialised after all other initskripts so this may need long time if the internet connection could not established.
If you use the external serial port the output from u-boot will not displayed. Maybee the external port is broken. There is a second user with a non working external port. This user has switched to the internal port for debugging. (u-boot always use the internal port) and IPFire can configured by adding:
console=ttyS1,115200n8
to uENV.txt on the first partition.
I have tested core160 on R1 and R1S H3 without problems.
1 Like
Thanks for the additional info. It looks like the kernel usually boots on my R1, but the load does not complete. I’ll try the second external port and see if that produces anything useful.
I first moved the nanopi R1 from its shelf, high in a corner, to a desk, where I could better follow the LEDs.
Setting the second external serial port made no difference. I then moved the PL2303 USB-serial dongle back to a bannapi M1, where it also did not produce any output either. Have yet to further investigate that cable. Meanwhile, raided yet another PL2303 from my CubieTruck. It produced no output from core 160 either.
A fresh uSD card having nightly build of 19 sep 2021 also produced no output to serial. It did get to flashing red LED on front plus solid green LED on LAN socket and blinking amber on WAN, but without serial I could not configure it nor activate DHCP on LAN.
I then used friendlywrt (May 2021) as a “rescue system”. All functions worked, leading me to believe that the nanopi R1 had not developed a fault. The nanopi had a working core 159 testing on emmc, that had failed to reboot, following upgrade to core 160 testing. I did the following:
- rsynced mmcblk0p1 from the uSD card to mmcblk2p1 on emmc (reports as mmcblk1p1 in friendlywrt, if anyone needs a rescue system)
- reflashed u-boot from /usr/share/u-boot on uSD to mmcblk2
- created a hybrid, booting from uSD, but with root=UUID set to that of mmcblk2p3
Neither worked.
I’ve now done a fresh install of core 159, stable to a uSD card. It booted and produced installation menu to serial. Have configured that with RED+GREEN, which will suffice, for the time being.
Reiterating that I have the second variant of nanopi R1, although, as I recall, there was no identifiable difference on the PCB