Soft reboot hangs on "Rebooting... reboot: Restarting system"
R
Rami Rosenbaum
started a topic
about 1 year ago
Hi,
I'm working on teh TS-7553-V2 (Rev. D), with the ts7553-V2-dec152017.tar.bz2 image.
I tried soft-rebooting the system with the rebootcommand.
System entered shut-down flow, but hanged in:
[ OK ] Stopped Replay Read-Ahead Data.
Stopping Collect Read-Ahead Data...
[ OK ] Stopped Collect Read-Ahead Data.
[ OK ] Reached target Shutdown.
[ 1680.164379] watchdog watchdog0: watchdog did not stop!
[ 1680.848623] reboot: Restarting system
I've seen some stuff in the internet (including imx6 fixes).
Best Answer
M
Michael Peters
said
10 months ago
In reviewing our forums I noticed this post did not update with the rest of Rami's story, which was concluded privately in his support ticket.
The problem, such as it were, turned out to be that his USB Cellular modem had some internal capacitors that held the power rails up while the power management chip was attempting to perform the reset function.
The solution in this case is to deliberately turn off USB power before issuing the reboot command. This causes power to be physically cut from the USB ports, shielding the internal power rail from the external device's pull.
Further information on the USB Power enable signal is found here:
I managed to constantly reproduce the 'hang', by connecting one of the cellular modems I'm using, Sierra Wireless GL6110.
It's an old modem we're still using, GPRS, full-speed, creating a /dev/ttyACM0 device.
A newer USB modem, Gemalto EHS6T, high-speed, creates /dev/ttyACM0 to /dev/ttyACM6, but reboots OK.
Michael Peters
said
10 months ago
Answer
In reviewing our forums I noticed this post did not update with the rest of Rami's story, which was concluded privately in his support ticket.
The problem, such as it were, turned out to be that his USB Cellular modem had some internal capacitors that held the power rails up while the power management chip was attempting to perform the reset function.
The solution in this case is to deliberately turn off USB power before issuing the reboot command. This causes power to be physically cut from the USB ports, shielding the internal power rail from the external device's pull.
Further information on the USB Power enable signal is found here:
Rami Rosenbaum
Hi,
I'm working on teh TS-7553-V2 (Rev. D), with the ts7553-V2-dec152017.tar.bz2 image.
I tried soft-rebooting the system with the reboot command.
System entered shut-down flow, but hanged in:
[ OK ] Stopped Replay Read-Ahead Data. Stopping Collect Read-Ahead Data... [ OK ] Stopped Collect Read-Ahead Data. [ OK ] Reached target Shutdown. [ 1680.164379] watchdog watchdog0: watchdog did not stop! [ 1680.848623] reboot: Restarting systemI've seen some stuff in the internet (including imx6 fixes).
In reviewing our forums I noticed this post did not update with the rest of Rami's story, which was concluded privately in his support ticket.
The problem, such as it were, turned out to be that his USB Cellular modem had some internal capacitors that held the power rails up while the power management chip was attempting to perform the reset function.
The solution in this case is to deliberately turn off USB power before issuing the reboot command. This causes power to be physically cut from the USB ports, shielding the internal power rail from the external device's pull.
Further information on the USB Power enable signal is found here:
https://wiki.embeddedarm.com/wiki/TS-7553-V2#Special_DIO
- Oldest First
- Popular
- Newest First
Sorted by Oldest FirstMichael Peters
Hi Rami,
Can you give us more detail? I have a TS-7553-V2 Rev. D right here, running the December 15 image, and it reboots consistently on command.
Rami Rosenbaum
Usually reboot command succeeds.
I tried looking for a scenario to reproduce... not much luck.
Maybe something to do with the USB-cellular modem connected and LAN cable (only peripherals I'm using).
I happened maybe ~5 times out of ~50 attempts.
I haven't dove into it so much, hoped it's a known issue (or this-or-that miss-use on my side)
Michael Peters
Good afternoon Rami,
One of the other engineers here pointed out that it might be interesting to see some diagnostic output. Could you run and post this for us?
tshwctl --info
uname -a
cat /root.version
This might shed some light on the trouble you're describing.
Rami Rosenbaum
Sorry for the late replay.
Rami Rosenbaum
I managed to constantly reproduce the 'hang', by connecting one of the cellular modems I'm using, Sierra Wireless GL6110.
It's an old modem we're still using, GPRS, full-speed, creating a /dev/ttyACM0 device.
A newer USB modem, Gemalto EHS6T, high-speed, creates /dev/ttyACM0 to /dev/ttyACM6, but reboots OK.
Michael Peters
In reviewing our forums I noticed this post did not update with the rest of Rami's story, which was concluded privately in his support ticket.
The problem, such as it were, turned out to be that his USB Cellular modem had some internal capacitors that held the power rails up while the power management chip was attempting to perform the reset function.
The solution in this case is to deliberately turn off USB power before issuing the reboot command. This causes power to be physically cut from the USB ports, shielding the internal power rail from the external device's pull.
Further information on the USB Power enable signal is found here:
https://wiki.embeddedarm.com/wiki/TS-7553-V2#Special_DIO