Hello all, I'm looking to (possibly) hack on dx for my Master's thesis. (Not sure if this is the right mailing list for this sort of thing, but we'll see.) I'm thinking of converting from JVM bytecode to dex through the use of a variant of CFA2 adapted from continuation-passing style to analyze the JVM stack. However, I need to be sure of A) how much effort it would require to put this into dx (or do it alone) and B) of the relative utility -- in terms of the resultant dex's efficiency -- of such a program (given that I don't know how "optimal" dx's current conversion is).
In order to be sure of these things, I need to understand what dx does now. Unfortunately, documentation is hard to come by (really just the dexopt document) and the current code is expansive. I see that it's doing SSA and all manner of middle-end compiler stuff... which is confusing because I thought it really just did a bit of abstract interpretation and conversion, not some abstraction followed by running a plethora of standard compiler passes, then compiling back into dex. Can anyone shed some descriptive light as to what's going on? If not, I guess I'll just read through all the code and trace its execution... :/ Thanks! Sincerely, Nicholas (Alex) Marquez -- You received this message because you are subscribed to the Google Groups "Android Developers" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/android-developers?hl=en
