ngfd was running for me:
And there’s nothing mentioning ngfd in journalctl -xe output.
Please add 4.3.0.23 into the list.
It just happened today that notifications and ringing stopped working. After rebooting all sounds are back again.
I had this on the Intex Aquafish at random, then on the Sailfish X with 3.3.0.16 it never happens. I will use the X on 3.3.0.16 for now. I need a reliably working daily driver.
This is another issue to piss me off. @Jolla do you hear me? please take it seriously - take time to get things stable
If you need volume in the number of complaints scribble me down as one seeing this on 4.3.0.12 / Xperia 10 Plus. Please fix.
4.4.0.58, still happening. It’s been over a year since I first noticed this bug, what’s going on?
Happened to me as well with no audible alarm, and also no vibration. At the same time my fingerprints where lost and had to be re-enabled through the usual sailfish-fpd procedure.
Don remember if I restarted or restarted homescreen to get back again.
Xperia 10 II w/ 4.4.0.58
I get the same symptoms from time to time. The fact that ngfd is still running is not all. It might have been deadlocked already.
Usually, when I miss notifications, or vibrations, ngfd is using 100% of cpu and some very long time after it crashes / gets restarted.
I can feel that 100% as the phone gets hotter.
Even after adding -vvv to the ngfd service command, I am not sure what is triggering this.
I have the exact same problem with my newly flashed Xperia 10 III.
For me ringing stops occasionally after using WhatsApp in Android. It is not 100% reproducible - means not every time I use WhatsApp ringing stops, but the opposite is true all the time ringing stops I happened to use WhatsApp.
Is whatsapp notifying you through the jolla homescreen?
I think so … if I understand your question correctly, then yes.
It pops up messages and plays the configured sound (usually) in the home screen and when phone is locked.
Yes, thats what I meant, thanks for confirming.
I also had the impression that some types of notifications are killing/locking ngfd.
what does locking have to do with the sound system? My wife uses still the Intex AquaFish and a lot of WhatsApp - never has the issue.
I was using the Xperia X and now the 10II … issue on both … but I started using whats app only on the 10II. I don’t know what was causing the same issue on the X even before fpd came into the game.
…and had it happen on Xperia 10 III yesterday.
It happens recurringly and I’m just dumping here some of the backtrace I got for later investigations.
When it happens, I attached a gdb
session to the running ngfd
. The threads are the followings:
(gdb) i threads
Id Target Id Frame
* 1 LWP 29473 "ngfd" 0x00000000075bd320 in ?? ()
2 LWP 29474 "threaded-ml" 0x00000075466aa780 in poll () from /lib64/libc.so.6
3 LWP 29476 "gmain" 0x00000075466aa780 in poll () from /lib64/libc.so.6
4 LWP 16279 "threaded-ml" 0x00000075466aa780 in poll () from /lib64/libc.so.6
5 LWP 16284 "typefind:sink" 0x00000075466afe60 in syscall ()
from /lib64/libc.so.6
6 LWP 16285 "mpegaudioparse1" 0x00000075466afe60 in syscall ()
from /lib64/libc.so.6
To be compared to the normal situation where:
(gdb) info threads
Id Target Id Frame
* 1 LWP 13935 "ngfd" 0x00000078d97b2780 in poll () from /lib64/libc.so.6
2 LWP 13937 "threaded-ml" 0x00000078d97b2780 in poll () from /lib64/libc.so.6
3 LWP 13939 "gmain" 0x00000078d97b2780 in poll () from /lib64/libc.so.6
The main thread is lost somewhere. But the most interesting one in my opinion is the mpegaudioparse1
with the following backtrace:
(gdb) t 6
[Switching to thread 6 (LWP 16285)]
#0 0x00000075466afe60 in syscall () from /lib64/libc.so.6
(gdb) bt
#0 0x00000075466afe60 in syscall () from /lib64/libc.so.6
#1 0x00000075468e75dc in g_cond_wait () from /usr/lib64/libglib-2.0.so.0
#2 0x000000753d4afc28 in gst_base_sink_wait_preroll ()
from /usr/lib64/libgstbase-1.0.so.0
#3 0x000000753d4aff30 in gst_base_sink_do_preroll ()
from /usr/lib64/libgstbase-1.0.so.0
#4 0x000000753d4b0710 in ?? () from /usr/lib64/libgstbase-1.0.so.0
#5 0x000000753d4b1bec in ?? () from /usr/lib64/libgstbase-1.0.so.0
#6 0x000000753d4b2f44 in ?? () from /usr/lib64/libgstbase-1.0.so.0
#7 0x0000007545b71df0 in ?? () from /usr/lib64/libgstreamer-1.0.so.0
#8 0x0000007545b7ac44 in gst_pad_push () from /usr/lib64/libgstreamer-1.0.so.0
#9 0x000000753d4bd784 in ?? () from /usr/lib64/libgstbase-1.0.so.0
#10 0x0000007545b71df0 in ?? () from /usr/lib64/libgstreamer-1.0.so.0
#11 0x0000007545b7ac44 in gst_pad_push () from /usr/lib64/libgstreamer-1.0.so.0
#12 0x000000753d4bd784 in ?? () from /usr/lib64/libgstbase-1.0.so.0
#13 0x0000007545b71df0 in ?? () from /usr/lib64/libgstreamer-1.0.so.0
#14 0x0000007545b7ac44 in gst_pad_push () from /usr/lib64/libgstreamer-1.0.so.0
#15 0x0000007545b5c67c in gst_proxy_pad_chain_default ()
from /usr/lib64/libgstreamer-1.0.so.0
#16 0x0000007545b71df0 in ?? () from /usr/lib64/libgstreamer-1.0.so.0
#17 0x0000007545b7ac44 in gst_pad_push () from /usr/lib64/libgstreamer-1.0.so.0
#18 0x000000753d37c02c in ?? () from /usr/lib64/libgstaudio-1.0.so.0
#19 0x000000753d37c380 in ?? () from /usr/lib64/libgstaudio-1.0.so.0
#20 0x000000753d382b3c in ?? () from /usr/lib64/libgstaudio-1.0.so.0
#21 0x000000753c5c2cdc in ?? () from /usr/lib64/gstreamer-1.0/libgstlibav.so
#22 0x000000753c5c3654 in ?? () from /usr/lib64/gstreamer-1.0/libgstlibav.so
#23 0x000000753d37e940 in ?? () from /usr/lib64/libgstaudio-1.0.so.0
#24 0x000000753d37ecf8 in ?? () from /usr/lib64/libgstaudio-1.0.so.0
#25 0x000000753d380464 in ?? () from /usr/lib64/libgstaudio-1.0.so.0
#26 0x0000007545b71df0 in ?? () from /usr/lib64/libgstreamer-1.0.so.0
#27 0x0000007545b7ac44 in gst_pad_push () from /usr/lib64/libgstreamer-1.0.so.0
#28 0x000000753d4a0b44 in gst_base_parse_push_frame ()
from /usr/lib64/libgstbase-1.0.so.0
#29 0x000000753d4a33c4 in gst_base_parse_finish_frame ()
from /usr/lib64/libgstbase-1.0.so.0
#30 0x000000753c60b070 in ?? () from /usr/lib64/gstreamer-1.0/libgstaudioparsers.so
#31 0x000000753d49c0bc in ?? () from /usr/lib64/libgstbase-1.0.so.0
#32 0x000000753d49c780 in ?? () from /usr/lib64/libgstbase-1.0.so.0
#33 0x000000753d49f6f4 in ?? () from /usr/lib64/libgstbase-1.0.so.0
#34 0x0000007545bab86c in ?? () from /usr/lib64/libgstreamer-1.0.so.0
#35 0x00000075468c0e60 in ?? () from /usr/lib64/libglib-2.0.so.0
#36 0x00000075468c0534 in ?? () from /usr/lib64/libglib-2.0.so.0
#37 0x0000007546765a4c in ?? () from /lib64/libpthread.so.0
#38 0x00000075466b48dc in ?? () from /lib64/libc.so.6
It seems that this thread is locked in a g_cond_wait()
call for ever. I need to investigate further but I don’t have time at the moment. Here is just a dump for these information not to be lost.
Thanks for the thread debugging you’ve done! I’m not sure if how to do this on an OBS port, but do you have an official device?
If so, can you download symbols for gst (gstreamer)?
Make sure you tick ”store debug symbols to home partition” in developer settings first, though.
Thanks!
Had it happen again. After killing ngfd (which took a bit because fingerterm stopped responding for about a minute the moment I touched the keyboard for the first time), the sounds returned.
Xperia 10 II - 4.4.0.64 Vahna Rauma
Happened this morning again, no sound, no vibration
My ngfd looked like the one in the scroot’s from @ender but with “defaultuser” instead of “nemo”. Several tries to be able to get su rights and also the Terminal response was very slow initially. Probably due to the fact that the Alarm keept poping up…
My fingerprints where also lost but they return after the new Utilities/Restart fingerprint. Mediaplayer worked so all sound where not lost, only notifications etc.
Did not know how to kill the ngfd so restarted Home screen from Utilities and are up running again.
Not an actual help with a bug, but I use qCommand GUI to restart ngfd when I notice this happens… Faster than using terminal which depends on haptic feedback. The command does not require root
systemctl --user restart ngfd
Xperia XA2plus in 4.4.0.68 version. I have realized that the sound functionality recovers when the power cord usb is plug to the phone. (You have to adjust volume because it is independent of the volume unppluged). Incoming calls, sound of radio apps, videos etc. After unppluging the power of usb the sound disappears.
Now, if I’m waiting for a important call, I try to connect to the charger the phone so I get back the loudsound again. Without cord, no sound only vibration works.
I have also verified to recover sound with CSD utility checking receiver playback. But the sound that you get is low level sound and after a call the sound is lost. With conection to the charger is working normal againg, but when you disconnect, the sound of radio apps or sound of incoming call, sms, alarms, etc, stop immediately as I have told.
Regards
UPDATE: After toggling fix/unfix sound connecting/disconnecting the usb power and after trying to use handsfree in a call and mutting/ unmutting microphone, I have got full functionality of my phone.Now it works properly, so I think that it’s possible to recover functionality in a no very complex way like creating processes to watch if a service is down and starting it again. It worked for me but I do’t know how or why.