https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90744
--- Comment #6 from Thomas Koenig <tkoenig at gcc dot gnu.org> --- (In reply to Tomáš Trnka from comment #5) > (In reply to Thomas Koenig from comment #4) > > Good analysis, and this is indeed the correct fix. > > OK. I thought that perhaps get_formal_from_actual_arglist() should be done > already before processing the arguments for the first call so that all calls > go through the same processing (just in case there were more bugs lurking in > the current approach). Not sure if that makes any sense, though. During resolution? Might also make sense, but the order of resolving things is quite fragile, and easy to get wrong... I will check.
