Software reboot and WSOD
 

I have been keeping my 770 constantly on since I got it back from WSOD repair last November. I have now noticed that the bluetooth system won't work after several days of uptime. I tried to reboot with /sbin/reboot (as root) and it worked as expected. Is there any difference on making a warm reboot like this with respect to a risk of getting a WSOD?

Does any version of powerlaunch (http://powerlaunch.garage.maemo.org/)
work with OS2007HE?

I have the latest OS2007HE.

Julius

Re: Software reboot and WSOD
 

Theoretically software reboot shouldn't cause failure called WSOD since in n770 there is no change in voltages supplied to linear regulators. The only change is probably the state of V18 port (1.8V from "RETU") which starts power ON sequence. If there's no failure in power supply - everything should work as designed.
Keep in mind that power sequece based on SHUTDOWN inputs of LDO's is rather low quality solution for starting LCD's so probably "no one knows for sure".
There were posts about WSOD after powering ON from the switch (which in facts is the same process since power button is not a hardware solution).

Re: Software reboot and WSOD
 

AFAIK no, reboot does full power cycle. At least that's what I've been told when asking specifically about this. That's why I was working on 'soft' reboot - after shutdown linux kernel is not rebooted but init goes back to initfs. It needs hacking /sbin/init binary and some startup/shutdown scripts and linuxrc in initfs too. I got it mostly working but there are some slight issues due to linux kernel not being rebooted so some startup scripts do initializations which are already done and there are errors. Overall the system works after such reboot but some stuff (like bluetooth not working) cannot be solved in this way. It is slightly useful when rebooting to different rootfs (flash vs mmc) but quite complex to set up.

EDIT:
here is my question and igor's answer
http://www.internettablettalk.com/fo...ycle#post64269 my understanding from this is that reboot is not different from shutting device off via power button and then powering it on later, worse thing only is removing the battery

Re: Software reboot and WSOD
 

Thanks for your explanations, both fanoush and XTC.

"Was working"? Do you have any plans on continuing that work?

Is there any other way of restating the bluetooth system?

Re: Software reboot and WSOD
 

well, it was just an idea but it is relatively complex to set up and is not clean enough to have practical use as a regular reboot. Better would be to have kexec working that would shutdown whole linux kernel and load and switch to another one without power cycling and going through NOLO bootloader. That would be cleaner, solve more issues (=reasons why one wants to reboot) and would not neded modification of /sbin/init and various boot scripts. And it would still avoid triggering WSOD.

as for re-initializing bluetooth you may try to run (as root)
/etc/init.d/bluez-utils stop
/etc/init.d/bluez-utils start
does it help?

Re: Software reboot and WSOD
 

Unfortunately it doesn't. The commands output "Stopping bluez-utils: hcid rfcomm." and "Starting bluez-utils: hcid rfcomm." Still my gps test program reports Error 22 in gpsbt_start: Invalid argument (No rfcomm created).

Julius

Re: Software reboot and WSOD
 

you can also try "hciconfig hci0 reset" between stop and start but the "No rfcomm created" may be caused by something not directly related to bluetooth chip but the bluetooth software stack. Or maybe it is generic out of memory problem. Is there something interesting in kernel log (use 'dmesg') when bluetooth stops working or when you stop/start it?

Also maybe you somehow run out of rfcomm ports by allocating new ones without releasing old ones? Try rfcomm --help (bind, release command options)

Re: Software reboot and WSOD
 

It's been a while since I have had this problem, but now it happened again.

hciconfig hci0 reset did not help.

As far as I can see, no there isn't.

I can bind the GPS receiver to /dev/rfcomm0 and then cat -u /dev/rfcomm0 outputs the GPS sentences. I can even start gpsd by hand and it seems to work (/usr/sbin/gpsd -N -D 2 /dev/rfcomm0). Still the usage through gpsbt_start doesn't work. Really odd.