12 replies to this topic

[rumensat]

periodicaly zgemma h7s freezing - some time every night some time after 2-3 days.

here is crash log

compiledate=Mar 23 2020

skin=PLi-FullNightHD/skin.xml

sourcedate=2020-03-21

branch=release-7.2

rev=85ed4d2

component=enigma2, signal

 

stbmodel=h7

stbmodel=h7

kernelcmdline=root=/dev/mmcblk0p7 rootsubdir=linuxrootfs2 kernel=/dev/mmcblk0p4 rw rootwait h7_4.boxmode=1 brcm_cma=440M@328M brcm_cma=192M@768M console=ttyS0,115200

nimsockets=NIM Socket 0:

imageissue=openpli 7.2-release %h

 

 

REMOVE Peer h7

REMOVE Peer h7

Backtrace:

/usr/bin/enigma2(_Z17handleFatalSignaliP9siginfo_tPv) [0x72D0C]

/lib/libc.so.6(__default_rt_sa_restorer) [0xB6154C60]

/lib/libc.so.6(gsignal) [0xB615394C]

/lib/libc.so.6(abort) [0xB6154DC4]

-------FATAL SIGNAL

 

[rumensat]

one more - after that crash have to sw. off - on power to restart, nothing else help

[WanWizard]

We are aware of this problem, it has to do with concurrent access to the EPG cache. Using EPGRefresh, EPGimport, or both?

 

We hope to have this fixed in 7.3.

[rumensat]

i am not using anything with EPG

[rumensat]

thanks for info

i will sw off all epg adds and tell for result after 5 days test

[betacentauri]

Does the box wake up from deep standby when it happens or do you use standby?

[babsy98]

many boxes show this  here a log from other box with more infos

 

04:20:47.1181 [ ] base/e2avahi.cpp:279 avahi_client_callback [Avahi] client state: 2
04:20:47.1189 [ ] base/e2avahi.cpp:360 avahi_timeout_new [Avahi] avahi_timeout_new
04:20:47.1240 [ ] base/e2avahi.cpp:380 avahi_timeout_free [Avahi] avahi_timeout_free
04:20:47.1243 [ ] base/e2avahi.cpp:360 avahi_timeout_new [Avahi] avahi_timeout_new
04:20:47.1266 [ ] base/e2avahi.cpp:380 avahi_timeout_free [Avahi] avahi_timeout_free
04:20:47.1271 [ ] base/e2avahi.cpp:360 avahi_timeout_new [Avahi] avahi_timeout_new
04:20:47.1286 [ ] base/e2avahi.cpp:380 avahi_timeout_free [Avahi] avahi_timeout_free
04:20:47.1296 [ ] base/e2avahi.cpp:360 avahi_timeout_new [Avahi] avahi_timeout_new
04:20:47.1314 [ ] base/e2avahi.cpp:380 avahi_timeout_free [Avahi] avahi_timeout_free
04:20:47.1323 [ ] base/e2avahi.cpp:360 avahi_timeout_new [Avahi] avahi_timeout_new
04:20:47.1345 [ ] base/e2avahi.cpp:380 avahi_timeout_free [Avahi] avahi_timeout_free
04:20:47.1346 [ ] base/e2avahi.cpp:372 avahi_timeout_update [Avahi] avahi_timeout_update
04:20:47.1347 [ ] base/e2avahi.cpp:149 avahi_service_try_register [Avahi] Registered vuduo2-2 (_e2stream._tcp) on vuduo2-2:8001

04:20:47.1358 [ ] base/e2avahi.cpp:360 avahi_timeout_new [Avahi] avahi_timeout_new
04:20:47.1369 [ ] base/e2avahi.cpp:380 avahi_timeout_free [Avahi] avahi_timeout_free
04:20:47.1372 [ ] base/e2avahi.cpp:360 avahi_timeout_new [Avahi] avahi_timeout_new
04:20:47.1382 [ ] base/e2avahi.cpp:380 avahi_timeout_free [Avahi] avahi_timeout_free
04:20:47.1696 [ ] bsod.cpp:379 oops PC: 762e2470
04:20:47.1696 [ ] bsod.cpp:383 oops 00000000 fffffffe 00000000 7fd232b0
04:20:47.1697 [ ] bsod.cpp:383 oops 00000003 7fd23330 00000000 00000000
04:20:47.1698 [ ] bsod.cpp:383 oops 00000002 000440c4 762b5efc 6e697375
04:20:47.1698 [ ] bsod.cpp:383 oops 7fd23428 7641a520 75ed9044 75f115ac
04:20:47.1699 [ ] bsod.cpp:383 oops 00000000 7fd23330 76413000 7fd2375c
04:20:47.1699 [ ] bsod.cpp:383 oops 7fd236dc 00000061 00000073 00000073
04:20:47.1699 [ ] bsod.cpp:383 oops 00000000 7fd23210 00000010 00000000
04:20:47.1700 [ ] bsod.cpp:383 oops 7641a520 7fd23298 fffffff7 762e23e0
04:20:47.1807 [ ] bsod.cpp:402 print_backtrace Backtrace:
04:20:47.1818 [ ] bsod.cpp:410 print_backtrace /usr/bin/enigma2(_Z17handleFatalSignaliP9siginfo_tPv) [0x47C7A0]
04:20:47.1820 [ ] bsod.cpp:410 print_backtrace /lib/libc.so.6(gsignal) [0x762E2472]
04:20:47.1821 [ ] bsod.cpp:422 handleFatalSignal -------FATAL SIGNAL

[betacentauri]

So it might be avahi or EPG.

 

If it's avahi, athoik is analyzing it and maybe has a fix for it.

[ims]

Why epg ? Why avahi ?

I have h7s and this error I've never seen it before, then I flash last owif change (then it is on all my boxes)... but this user using: 

 

sourcedate=2020-03-21

branch=release-7.2

[littlesat]

Epg due to eEPGCache.cpp has a bug forever... with a corrupted epg.dat a failure is easy inserted... plus it will made a corrupted epg.dat forever... Maybe via console init 4, remove epg.dat, init 4 might help... especially with client box that imports epg.dat from fallback tuner will fail. Maybe consider disable import epg from fallback tuner (see the release notes). Pre 7.2 the import itself in a lot of cases did actually not work and now they work and the issue is triggered.
Avahi because it was mentioned with a timeout multiple times in the log.
Image date also 23-mar-2020....
Mag also update or even reflash (via ui) and try again !

Edited by littlesat, 16 May 2020 - 13:33.

[athoik]

many boxes show this  here a log from other box with more infos

 

04:20:47.1181 [ ] base/e2avahi.cpp:279 avahi_client_callback [Avahi] client state: 2
04:20:47.1189 [ ] base/e2avahi.cpp:360 avahi_timeout_new [Avahi] avahi_timeout_new
04:20:47.1240 [ ] base/e2avahi.cpp:380 avahi_timeout_free [Avahi] avahi_timeout_free
04:20:47.1243 [ ] base/e2avahi.cpp:360 avahi_timeout_new [Avahi] avahi_timeout_new
04:20:47.1266 [ ] base/e2avahi.cpp:380 avahi_timeout_free [Avahi] avahi_timeout_free
04:20:47.1271 [ ] base/e2avahi.cpp:360 avahi_timeout_new [Avahi] avahi_timeout_new
04:20:47.1286 [ ] base/e2avahi.cpp:380 avahi_timeout_free [Avahi] avahi_timeout_free
04:20:47.1296 [ ] base/e2avahi.cpp:360 avahi_timeout_new [Avahi] avahi_timeout_new
04:20:47.1314 [ ] base/e2avahi.cpp:380 avahi_timeout_free [Avahi] avahi_timeout_free
04:20:47.1323 [ ] base/e2avahi.cpp:360 avahi_timeout_new [Avahi] avahi_timeout_new
04:20:47.1345 [ ] base/e2avahi.cpp:380 avahi_timeout_free [Avahi] avahi_timeout_free
04:20:47.1346 [ ] base/e2avahi.cpp:372 avahi_timeout_update [Avahi] avahi_timeout_update
04:20:47.1347 [ ] base/e2avahi.cpp:149 avahi_service_try_register [Avahi] Registered vuduo2-2 (_e2stream._tcp) on vuduo2-2:8001

04:20:47.1358 [ ] base/e2avahi.cpp:360 avahi_timeout_new [Avahi] avahi_timeout_new
04:20:47.1369 [ ] base/e2avahi.cpp:380 avahi_timeout_free [Avahi] avahi_timeout_free
04:20:47.1372 [ ] base/e2avahi.cpp:360 avahi_timeout_new [Avahi] avahi_timeout_new
04:20:47.1382 [ ] base/e2avahi.cpp:380 avahi_timeout_free [Avahi] avahi_timeout_free
04:20:47.1696 [ ] bsod.cpp:379 oops PC: 762e2470
04:20:47.1696 [ ] bsod.cpp:383 oops 00000000 fffffffe 00000000 7fd232b0
04:20:47.1697 [ ] bsod.cpp:383 oops 00000003 7fd23330 00000000 00000000
04:20:47.1698 [ ] bsod.cpp:383 oops 00000002 000440c4 762b5efc 6e697375
04:20:47.1698 [ ] bsod.cpp:383 oops 7fd23428 7641a520 75ed9044 75f115ac
04:20:47.1699 [ ] bsod.cpp:383 oops 00000000 7fd23330 76413000 7fd2375c
04:20:47.1699 [ ] bsod.cpp:383 oops 7fd236dc 00000061 00000073 00000073
04:20:47.1699 [ ] bsod.cpp:383 oops 00000000 7fd23210 00000010 00000000
04:20:47.1700 [ ] bsod.cpp:383 oops 7641a520 7fd23298 fffffff7 762e23e0
04:20:47.1807 [ ] bsod.cpp:402 print_backtrace Backtrace:
04:20:47.1818 [ ] bsod.cpp:410 print_backtrace /usr/bin/enigma2(_Z17handleFatalSignaliP9siginfo_tPv) [0x47C7A0]
04:20:47.1820 [ ] bsod.cpp:410 print_backtrace /lib/libc.so.6(gsignal) [0x762E2472]
04:20:47.1821 [ ] bsod.cpp:422 handleFatalSignal -------FATAL SIGNAL

The problem with Avahi is that we are trying to keep a const char* from something that is not const...

 

 

We are calling avaya from OWIF using the following command:

 

# BJ
def BJregisterService(protocol, port):
        try:
                from Plugins.Extensions.Bonjour.Bonjour import bonjour
                service = bonjour.buildService(protocol, port, 'OpenWebif')
                bonjour.registerService(service, True)
        except: # nosec
                pass
        try:
                servicetype = '_' + protocol + '._tcp'
                enigma.e2avahi_announce(None, servicetype, port)
        except: # nosec
                pass

 

The servicetype variable is just a local variable that python garbage collect after some time.

 

Before we never had the chance to see that error, simply because we where starting OWIF toooooooooo late.

 

Now OWIF starts tooooooooo fast.

 

My solution is to use malloc to store the references on Avahi, it works great more than a month ...

[WanWizard]

If it works, implement it.

[athoik]

Pushed here: https://github.com/O...f2d968fc2960bcd

 

Although I believe that reporter issue has nothing to do with avahi issue seen after switching on netifaces parsing of interfaces.