[vdr] Re: caught signal 2

[hm@seneca.muc.de]

hm@seneca.muc.de wrote:
> hm@seneca.muc.de wrote:

> > OK, I asked fer it, so - here's a patch against 1.3.15, changing all
> > the signal() stuff to sigaction() in vdr.c. It appears to work as shown
> > (1455 is the master vdr thread):

> Hmmm - still observing what happens. Since I've been running this
> patch I had one signal 2 event, but strangely the sender PID was none
> of vdr's. Actually, the process or thread didn't even exist any more

It's all very strange. This is what happened last night. 

Dec 19 20:55:44 linvdr vdr[2004]: caught signal 2 from 2018

... and then there is no process 2018, but

Dec 19 20:55:44 linvdr logger:  2017 ?        S      0:00 ./vdr -w 60 -l 3 -t /dev/tty8 -p 2001 -s /usr/local/bin/vdrshutdown -r /usr/local/bin/utils-vdr.sh -c /video/linvdr -E /video/linvdr/epg.data -Premote -i /dev/input/event0 -Pmp3 -P lcdproc -Ptvonscreen -Posdpip -Posdteletext -n 7 -d /tmp/vtx -P streamdev-client
Dec 19 20:55:44 linvdr logger:  2019 tty1     S      0:00 sh tst.sh

where tst.sh is my test script evaluating /var/log/messages and dumping "ps
ax" to logger when a "caught signal" occurs. Apparently, 2018 is missing. A
little later: 

Dec 19 20:56:28 linvdr vdr[2570]: caught signal 2 from 2938

and no 2938 at all. 

Now I'm totally perplexed. There is no "SIGINT" in any of the vdr source
files. Does anybody have any idea what might be sending SIGINT, and
disappear immediately, i.e. within less than one second? 

:-( 

-- 
"He did decide, though, that with more time and a great deal of mental
effort, he could probably turn the activity into an acceptable
perversion."
		-- Mick Farren, "When Gravity Fails"