[Wifidog] Why WiFiDog Crashes

[Philippe April]

Turns out I was partially wrong.. Found out a SIGPIPE is sent when
write() returns "broken pipe", and WiFiDog would just quit. It's much
harder to debug when don't have all the tools to debug.

I don't know if it's the proper way to handle it, but I just used the
same handler that we have for SIGCHLD to trap SIGPIPE, which does ...
nothing about it. Works great!

It should be much more stable now, I'll let it run and see.

Whew! Glad I found that bug, it was so annoying.

Everybody have a great week-end,

-- 
Philippe April
papril777 at yahoo.com
GnuPG: http://key.philippeapril.com/

On Fri, May 14, 2004 at 09:55:05PM -0400, Philippe April wrote:
> Hi,
> 
> I have found out why WiFiDog crashes (you know, that one that I could
> not reproduce, I now can).
> 
> Basically, my significant other would call me and say "The internet's
> not working!" and I could never see it crash by myself!
> 
> I found out what she's probably doing... She's probably impatient and
> clicks reload or presses the sign-in button twice, and WiFiDog does not
> seem to be able to process the two requests properly.
> 
> Basically it's a call to write() on a socket that does not seem to exist
> anymore. I do not know if it's a bug in uClibc or if it exists in glibc
> too (I'll try to figure that one out), but if it's not we'll have to
> modify libhttpd and wifidog some more and figure out how to avoid
> this...
> 
> Just wanted to give an update,
> 
> -- 
> Philippe April
> papril777 at yahoo.com
> GnuPG: http://key.philippeapril.com/
> 

> _______________________________________________
> Wifidog mailing list
> Wifidog at isf.waglo.com
> http://isf.waglo.com/mailman/listinfo/wifidog_isf.waglo.com


-------------- next part --------------
_______________________________________________
Wifidog mailing list
Wifidog at isf.waglo.com
http://isf.waglo.com/mailman/listinfo/wifidog_isf.waglo.com