Hi Richard, > Is the separate need_temps scan required for correctness? It looked > like we could test: > > if (reg_overlap_mentioned_p (dest, cond)) > ... > > on-the-fly during the main noce_convert_multiple_sets loop.
right, I didn't re-check it but after changes during interal patch rounds the check is now degenerated enought to allow this. Going to refactor that. > Would it work to set temporaries[i] to targets[i] whenever a temporary > isn't needed, and avoid temps_created? Yes, that should work. Already made that change locally. Regards Robin