Home
Reading
Searching
Subscribe
Sponsors
Statistics
Posting
Contact
Spam
Lists
Links
About
Hosting
Filtering
Features Download
Marketing
Archives
FAQ
Blog
 
Gmane
From: Junio Cezar <juniocezar.dcc <at> gmail.com>
Subject: Probing the complexity of the LLVM's optimizations
Newsgroups: gmane.comp.compilers.llvm.devel
Date: Sunday 12th October 2014 15:43:21 UTC (over 2 years ago)
Hello guys,

I am a student from UFMG, Brazil, and I am doing a research project in
LLVM. I am running some experiments to estimate the complexity of the
LLVM's optimizations. To do this, I run the optimizations over
hundreds of different bytecode files, and collect the time spent by
each optimization.

To get a large number of samples, I use CSmith, the random C program
generator of Utah (from John Regehr's group). So, I can get hundreds
of C files with different sizes. I also use the benchmarks available
in the LLVM test suite.

The results of these experiments are available in my page:
http://homepages.dcc.ufmg.br/~juniocezar/llvm/.
You will find plots
for all the optimizations there. They are mostly linear, but there are
a few weird ones. You are all welcome to give me suggestions or point
out mistakes in my report.

Regards
 
CD: 3ms