¿ªÔÆÌåÓý

ctrl + shift + ? for shortcuts
© 2025 Groups.io
  1. Bpq32
  2. Topics

BPQ Shutting Down

 

Upgraded Pilinbpq to 6.0.24.59 and after about 5 minutes BPQ closes down with the following message.

SIGSEGV Received

Went back to .45 which is working fine.

Anyone else seeing this?

--
--
Apologies for any grammar?or spelling errors!
--
73 de Glenn N3MEL
(TPRFN) The Packet Radio Forwarding Network
Emergency Coordinator
Chester County ARES/RACES
 

Hi Glenn,

I do not use rPi's but .59 has been stable here on x64. I've had it running on 10m in the "lab" for about 2-3 days now. RF connects are only from me as I move around the house but include ax25, il2p, VARA, and ARDOP. I've staged it for use everywhere here after restarts.

73 de Rich WA3WLH
 

?
Hey Glenn
?
Yes, I had this happen yesterday after I downloaded .59 and started it up. It ran for?
5 or 10 mins and then and then crash. I'm running pilinbpq on a RPI4b with Bullseye OS.
I'm running linbpq as a service, so it would start back up and then 5 mins later, it would?
crash again. I ran linbpq under debug to see what was going on before I went back to .45.
Here's the results I got back from it:
?
Thread 151 "linbpq" received signal SIGPIPE, Broken pipe.
[Switching to Thread 0xf14ff440 (LWP 21143)]
0xf7eca3b4 in __libc_send (fd=35, buf=0xf14c19d0, len=69434, flags=0)
? ? at ../sysdeps/unix/sysv/linux/send.c:26
26 ? ?../sysdeps/unix/sysv/linux/send.c: No such file or directory.
(gdb) bt
#0 ?0xf7eca3b4 in __libc_send (fd=35, buf=0xf14c19d0, len=69434, flags=0)
? ? at ../sysdeps/unix/sysv/linux/send.c:26
#1 ?0x0014bcd8 in ProcessWebmailWebSockThread (conn=0x9ad3e0)
? ? at /mnt/Source/bpq32/CommonSource/HTTPcode.c:4735
#2 ?0xf7ebd310 in start_thread (arg=0xf14ff440) at pthread_create.c:477
#3 ?0xf7d5c5c8 in ?? () at ../sysdeps/unix/sysv/linux/arm/clone.S:73
? ?from /lib/arm-linux-gnueabihf/libc.so.6
Backtrace stopped: previous frame identical to this frame (corrupt stack?)
?
--------------------------------------------------------------------------------------------------------------
Now, just for squirts n' giggles, I fired up .59 today and it's been running for several
hours with no issues so far. It's running under debug right now. I'm keeping my eye
on it.
?
73 de ww6q
Roy
?
?
 

I had what might have been the same problem a while back when upgrading from the .53 to the .56 version. Crashes seemed to correlate with incoming VARA connections, but not every incoming connection would trigger a crash.

73,
Lee K5DAT
 

I noticed some other issues building and using .59.

Cheaders.h seems to replace cheaders.h but I dont think it was reflected fully or something in Makefile.

Another issue with .59 is try clawsmail out for smtp/nntp feeds from bpq. Mine would abort going through some of the nntp lists.

I reverted to .56, but I dont have the 'vara' issue others have on .56. I have previously discussed this with K5DAT in private mail.

Here my build/packet system is deb 10.
 

Still no crashes here with .59 after nearly a week running on x86 and x64.

In case the errors described have nothing to do with rPi's or ARM processors I should mention that I always build linbpq from source and I do not include support for I2C or MQTT.

I thought this worth a mention since the gdb backtrace provided by Roy looks like there is a problem with threading and / or libc but that could be misleading without digging deeper with gdb or something like strace.

73 de Rich WA3WLH

1 - 6 of 6