[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[lwip-users] Easiest way to get trace function call traces of lwip code
From: |
Richard Siegfried |
Subject: |
[lwip-users] Easiest way to get trace function call traces of lwip code |
Date: |
Sat, 10 Dec 2016 02:14:14 +0100 |
Hello,
## Background
I'm currently writing a research paper about
"A Comparison of different TCP Implementations"
comparing LWIP TCP with Linux TCP.
Besides
* a statical code analysis
* and a semantic analysis
(reading, understanding and explaining/comparing
the code)
I plan to include a trace log comparison. (And perhaps some other
dynamic analyses, if there are some other usefull/interesting ones).
Also the trace log could be a starting point for further exploration and
explaining of the code.
For linux this is easy, because I have some experience using ftrace.
## tl;dr: the question
My Question now is:
What do you think is the fastest/least cumbersome way, to get an
evironment running lwip and get some tracelog statistics.
I know there are some Operating systems using lwip (i know of Minix and
ReactOS), but at least for minix the tooling was not very comfortable
currently IMHO.
Links or pointers to good tutorials covering this topic are also highly
appreciated.
I know gcc and LLVM support trace instrumentation but I don't have
experience using them. I have some experience using tracing on perl or
python code and the linux kernel using ftrace. I also have some
experience using gdb.
## Etc.
If you have some ideas for good statistics or data points worth
including or considering, I'm always open. I'm still in a quite early an
flexible phase of this research work.
Thank you,
-- Richard Sailer
signature.asc
Description: OpenPGP digital signature
- [lwip-users] Easiest way to get trace function call traces of lwip code,
Richard Siegfried <=