>> 2. The patch description is incorrect -- the bug symptom is not >> SIGPIPE. SIGPIPE is globally ignored by Polipo. > > I had polipo segfaulting before applying this patch. With gdb I > thought to understand it was a SIGPIPE problem, but I can be wrong > because I am not a experienced developer.
Juliusz you are right. SIGPIPE was completely wrong. I went to study. I am very begineer gdb user and I did not have a good understanding of signals. The situation is the following: version 1.0.4: eventually segfaults, but I was not able yet to reproduce this bug in GDB version 1.0.4: you should ignore SIGPIPE signal in GDB if you want to debug version 1.0.4: debian applies this patch to the deb package on top of 1.0.4, so might have it also on OpenWRT because it should increase the overall stability git master: looks more stable, I have this version running now and it is not crashing. sorry for the confusion on the list. Making debug at late night after long days of work did not help :)) ciao :) Saverio _______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/mailman/listinfo/openwrt-devel
