Procedure inlining at the source level

I am looking for papers/technical reports/general pointers about the subject
of procedure inlining (also called integration, open compilation, open coding,
etc.) at the source level, and criteria for inlining that produce good
optimizations.


I have found a few papers (old) on the subject, but most are either very
sketchy or are written on the intermediate language transformation level.


I am interested in source-to-source transformations. If anyone has recent
papers on source-to-source transforms (implementation discussion) or recent
analyses of criteria for when to inline, please email.


Apologies if this raises a recent issue; I've just joined this newsgroup.


Thanks, Alan W. Yueh
(yueh@cs.unc.edu)
--

Return to the comp.compilers page.
Search the comp.compilers archives again.