The error seems to have been thrown by /etc/init.d/cgminer.sh
I can't seem to reproduce it right now. Instead I am getting "-sh: etc/init.d/cgminer: not found" however "less /etc/init.d/cgminer.sh" gives:
It seems you've missed the leading slash before etc, no?
just execute the command:
/etc/init.d/cgminer.sh restart
It seems you've missed the leading slash before etc, YES? Stupid error
Results:
root@Jupiter-3D0:~# /etc/init.d/cgminer.sh restart
Restarting Cgminer daemon: killall: cgminer: no process killed
Error: Write failed
root@Jupiter-3D0:~# /etc/init.d/cgminer.sh start
Starting Cgminer daemon: Error: Write failed
root@Jupiter-3D0:~# /etc/init.d/cgminer.sh stop
Stopping Cgminer daemon: killall: cgminer: no process killed
cgminer.
TaggedYa
strange.
since there're a lot of things going on when you issue a
cgminer.sh start and a lot of those are related with the initialization of the asic boards I'm afraid that there's something wrong going on in this phase.
this are all the tasks performed:
root@lucas:~# fgrep \# /etc/init.d/cgminer.sh
#!/bin/sh
# Stop SPI poller
# CLear faults in megadlynx's
# Re-enable PLL
# re-enable all cores
# Disable PLL
# disable all cores
# Disable direct SPI
# Enable SPI poller
all of them involved some sort of
i2cset command executions.
If I were you I would join #kncminer channel on freenode and try to ask hno (an OrSoC eng) if he have an idea of what's going on.