signal=SEGV

[sphh]

signal=SEGV

[sphh]

I now compiled cornrow on my Raspberry Pi 1 (it took about 8 hours!), uninstalled the old one and installed the freshly compiled one. Same result.

If I start cornrow from the command line, I get

$ /usr/bin/cornrowd
time         [ thread name/id ]                   file:line     v| 

Loguru caught a signal: SIGSEGV
Stack trace:
0   0xb4f01130 __default_rt_sa_restorer + 0
10:10:40.457 [main thread     ]                       :0     FATL| Signal: SIGSEGV
Segmentation fault

[mincequi]

Hi,
Thanks for checking this out. Can you try to run cornrowd from command line without init script?
cornrowd

It should print some log messages...

[mincequi]

Ok, sorry. I just saw you were trying this.

Well, in this case you can try to start it using gdb.
There might be a valid stack trace.

[sphh]

Here we go:

$ sudo apt install ./cornrowd_0.7.0_armhf.deb ./cornrowd-dbgsym_0.7.0_armhf.deb
$ gdb -q /usr/bin/cornrowd
Reading symbols from /usr/bin/cornrowd...Reading symbols from /usr/lib/debug/.build-id/9e/9a06b41fade2059cc4d496eab3771a3a22740b.debug...done.
done.
(gdb) run
Starting program: /usr/bin/cornrowd 
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/arm-linux-gnueabihf/libthread_db.so.1".
time         [ thread name/id ]                   file:line     v| 
[New Thread 0xb05b30f0 (LWP 30877)]

Thread 1 "cornrowd" received signal SIGSEGV, Segmentation fault.
BluezQt::Adapter::media (this=0x0) at ./thirdparty/bluez-qt/src/adapter.cpp:161
161	./thirdparty/bluez-qt/src/adapter.cpp: No such file or directory.
(gdb) backtrace
#0  BluezQt::Adapter::media (this=0x0)
    at ./thirdparty/bluez-qt/src/adapter.cpp:161
#1  0x00025c80 in bluetooth::Controller::Controller (this=0x1dae728, 
    parent=<optimized out>)
    at /usr/include/arm-linux-gnueabihf/qt5/QtCore/qsharedpointer_impl.h:312
#2  0x0001eee0 in Controller::Controller (this=0x1db5ed0, 
    parent=<optimized out>) at ./cornrowd/src/Controller.cpp:11
#3  0x0001ceac in main (argc=<optimized out>, argv=<optimized out>)
    at ./cornrowd/src/main.cpp:81
(gdb) 

Does that mean anything to you?

[mincequi]

Do you have a recent version of Bluez installed?
Can you check if you find the media interface for your bluetooth adapter?

sudo dbus-send --system --print-reply --type=method_call --dest='org.bluez' '/' org.freedesktop.DBus.ObjectManager.GetManagedObjects

Search for org.bluez.Media1

[sphh]

The bluez package installed has the version 5.50-1.2~deb10u1+rpt2.

$ sudo dbus-send --system --print-reply --type=method_call --dest='org.bluez' '/' org.freedesktop.DBus.ObjectManager.GetManagedObjects
method return time=1600879826.907833 sender=:1.6 -> destination=:1.13 serial=16 reply_serial=2
   array [
      dict entry(
         object path "/org/bluez"
         array [
            dict entry(
               string "org.freedesktop.DBus.Introspectable"
               array [
               ]
            )
            dict entry(
               string "org.bluez.AgentManager1"
               array [
               ]
            )
            dict entry(
               string "org.bluez.ProfileManager1"
               array [
               ]
            )
            dict entry(
               string "org.bluez.HealthManager1"
               array [
               ]
            )
         ]
      )
   ]

so no Media anywhere in sight.

Oh, btw, do I have to pair a device before I call cornrowd?

[mincequi]

Thats looks strange. Is bluetooth actually enabled on your raspi?
I cannot see any bluetooth adapter.

It should look something like this:

dict entry(
         object path "/org/bluez"
         array [
            dict entry(
               string "org.freedesktop.DBus.Introspectable"
               array [
               ]
            )
            dict entry(
               string "org.bluez.AgentManager1"
               array [
               ]
            )
            dict entry(
               string "org.bluez.ProfileManager1"
               array [
               ]
            )
            dict entry(
               string "org.bluez.HealthManager1"
               array [
               ]
            )
         ]
      )
      dict entry(
         object path "/org/bluez/hci0"
         array [
            dict entry(
               string "org.freedesktop.DBus.Introspectable"
               array [
               ]
            )
            dict entry(
               string "org.bluez.Adapter1"
               array [
                  dict entry(
                     string "Address"
                     variant                         string "B8:27:EB:4B:58:63"
                  )
...

There is no need to pair a device before starting cornrowd.

[sphh]

Looks like my Bluetooth dongle is not recognized (https://www.raspberrypi.org/forums/viewtopic.php?t=138145):

$ systemctl status bluetooth
● bluetooth.service - Bluetooth service
   Loaded: loaded (/lib/systemd/system/bluetooth.service; enabled; vendor preset
   Active: active (running) since Thu 2020-09-24 21:15:41 CEST; 50s ago
     Docs: man:bluetoothd(8)
 Main PID: 8344 (bluetoothd)
   Status: "Running"
   CGroup: /system.slice/bluetooth.service
           └─8344 /usr/lib/bluetooth/bluetoothd

So that looks ok. Also

$ dmesg | grep Blue
[   21.102716] Bluetooth: Core ver 2.22
[   21.102928] Bluetooth: HCI device and connection manager initialized
[   21.102975] Bluetooth: HCI socket layer initialized
[   21.103005] Bluetooth: L2CAP socket layer initialized
[   21.103071] Bluetooth: SCO socket layer initialized
[   42.519478] Bluetooth: BNEP (Ethernet Emulation) ver 1.3
[   42.519505] Bluetooth: BNEP filters: protocol multicast
[   42.519551] Bluetooth: BNEP socket layer initialized

But

$ hcitool dev
Devices:

draws a blank. Also

$ sudo bluetoothctl
Agent registered
[bluetooth]# devices
No default controller available

I guess my dongle is too modern for the Raspbian Buster. :-( I got it working without any fuss on Linux Mint 19.3, so I assumed that it will also work on the RPi …

[mincequi]

Phew, no idea.

Anyhow, i could add some logging to cornrowd, whether an adapter is present...

[sphh]

Yes, such a log message would be a good idea.