[llvm-dev] BoF on Sophisticated Program Analysis on LLVM IR

[John Criswell via llvm-dev]

Dear All,

I'm running a Birds-of-a-Feather (BoF) session on sophisticated program 
analysis for LLVM IR at the LLVM Developer's Meeting next week.  
Examples of sophisticated program analysis include (but are certainly 
not limited to) call graph analysis and whole-program points-to analysis.

While there has been consistent demand for sophisticated program 
analysis passes, there are no robust implementations for several such 
analyses, leaving a gap between what is needed and what is available.  
The goal of this BoF is to bring together users of sophisticated program 
analysis and LLVM developers to see what users need and whether there 
are common interests in creating, maintaining, and using these types of 
analyses.

As some community members may not be able to attend the meeting, I 
thought I'd post some of my tentative questions to the list and bring 
the results with me to the BoF.  Some discussion questions are:

o What program analyses do LLVM users and developers need that do no 
exist today?

o What obstacles are there to creating robust implementations of these 
analyses?

o Are there better methods of getting the information reported by these 
analyses (e.g., the way Type-Based Alias Analysis side-steps the need 
for Anderson’s alias analysis)?

o Would such analyses benefit existing LLVM optimizations while 
maintaining acceptable compile-time performance?

If you have an opinion on any of these, I'd love to hear from you!

Regards,

John Criswell

-- 
John Criswell
Assistant Professor
Department of Computer Science, University of Rochester
http://www.cs.rochester.edu/u/criswell