tcpdump mailing list archives

Re: libpcap pluggable I/O branch, rebased


From: Ray Bellis via tcpdump-workers <tcpdump-workers () lists tcpdump org>
Date: Fri, 21 Aug 2020 21:09:50 +0100

--- Begin Message --- From: Ray Bellis <ray () bellis me uk>
Date: Fri, 21 Aug 2020 21:09:50 +0100


On 21/08/2020 20:26, Michael Richardson wrote:

Ray, I rebased raybellis-gzip-v2 upon libpcap head.
Do you have a github that I can add as as reviewer?

Do you mean my github a/c name?   That's just "raybellis".

Mostly I did this because the FreeBSD tests seemed to fail, and I wanted to
see them work.  I know this has taken forever.

There are some changes to use charset_fopen() that I think that I managed to
port across.  Do you have some tests that you could put this through?

Not any more, but I could likely devise some.

https://github.com/mcr/libpcap/tree/raybellis-gzip-v2

https://github.com/the-tcpdump-group/libpcap/pull/914/files

Thank you.
I hope to merge it next week, but I want to add:
  ./configure --enable-ioplugin

and the equivalent for CMake, but I really am CMake-challenged, so I don't
really understand how to do the equivalent there.

I am also CMake-challenged.   I loathe it as a build system :p

I've left a comment in the pull request above where we'd #ifdef out.

I also wonder if we should have a pcap_enableioplugin() option so that
applications have to opt-in.  I'm just really concerned about opening up
holes for some setuid() libpcap system.

I share that concern, but don't know how to address it.

Ray


--- End Message ---
_______________________________________________
tcpdump-workers mailing list
tcpdump-workers () lists tcpdump org
https://lists.sandelman.ca/mailman/listinfo/tcpdump-workers

Current thread: