Brian - question on splitnslu
I was playing with building new flash image, and also with enabling telnet on boot up. This
leads to two questions I have for you.
1. I noted that the file SysConf created by splitnslu is always of zero bytes in length. In
looking through your code, you are using the first ten bytes of the SysConf segment in the
flash image as the count, pulling the count out of the first 4 bytes of that segment and
then skipping the remaining ten. I'm not so sure that's right, though, because I tried
pulling out those ten bytes from the original Linksys firmware image, and they all seem to
dd if=NSLU2_V23R25.bin skip=0x40000 bs=1 count=10 of=foo
and foo has ten bytes of zero in it!
Do you think the lengths are off (unlikely from what we know) or that maybe this isn't how
the length is encoded? Maybe one of the startup scripts builds the sysconf segment in
flash the first time the nslu2 is booted IF it sees that it is empty, so the flash image
actually has nothign in it? Any thoughts? (In fact, the first 500 bytes of the flash image
starting at ofset 0x40000 are zero).
2. I cannot get telnet to be enabled by default on boot, and I recall you saying that you
did. I have flashed a new /etc/rc.local which calls my /share/hdd/conf/etc/rc.d/rc.local at
startup my own rc.local symlinks /etc/inetd to my version which has the enabling line for
telnet. I know my rc.local is running, because I added three beeps (/usr/bin/Set_Led
beep1) to the end of it and that works. but I still cannot telnet in until I enable telnet via
the web interface - and of course doing that overwrites my symlink so I cannot tell it it
had worked or not! When I run my rc.local "by hand" it does what it should do, so I know
the script isn't coded wrong.
Did you discover that you also have to set the config variable for telnet_enable=yes for
telnet to work? If so, that's odd unless the version of telnet on the unit is hacked to check
for that or something...Did you find a way to modify that variable in the running system?