Andrew MacLeod wrote:
Those are the 4 actions/projects we left the summit with that I am aware of. With any luck at all, one or more of these will have a significant impact on our register allocator. Often projects like these proceed in virtual silence until they are mostly done. Perhaps I'll try to do a follow up in a few months to check the progress and results (if any) on the 4 projects, and post the results as a summary. Down the road, there may be some uniting of the work from the various projects, I can see some potential there...
Originally I objected to extracting part of code selection from reload (or its analog insn allocno allocator in YARA yara-insn.c). That was your cornerstone idea of RABLE proposal. I objected because solving two problems in integrated way might generate a better results and because of failure pre-reload pass with the new register allocator project. But it makes reload more complex too. Now I'd like to take some time to investigate again how your approach could help YARA (mostly how can it simplify RA). So I think there is a potential in uniting work of the different projects. There is already some intersection of yara and insn-select branch. What I read in Mike Matz's announcement about using only one register class for pseudos is already implemented in YARA. Vlad
