https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70162
--- Comment #5 from Oleg Endo <olegendo at gcc dot gnu.org> --- (In reply to Nick Clifton from comment #4) > Yes - I would much prefer to patch the assembler so that it can cope with > large > values, even when compiled for a 32-bit host. That way the assembler will be > able to cope even with source files that are not generated by gcc. > > A patch like the one attached here perhaps... :-) Of course GAS should be fixed, too. Adjusting GCC for the GAS bug will be a bit more robust, I think. > > I also happen to like seeing large constants as hex values, as 9 times out > of 10, the hex value will make more sense. (Well at least in my experience > anyway). Honestly, I don't care. Usually I don't have to read constants in asm code/dumps .. I assume they are correct ;) I was just a bit surprised to see 0xFFFFFFFFFFFFFFFF for -1 in the RX asm dumps, because I haven't seen that happening on SH. > > Would you like to close this PR and open a binutils one instead, or shall I > do it ? > Please go ahead.
