This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: stormy16: limit SI reload regs
> In that case, maybe it'd be a good idea to just fix register allocation?
I thought about that, but I'm not that familiar with that code. I
suspect it would need to be something near one of the calls to
PREFERRED_RELOAD_CLASS(), needing to also call HARD_REGNO_NREGS. I'm
willing to look at it if a reload expert could give me a clue or two ;-)
With my luck, that would break other targets though.
I wonder if HARD_REGNO_MODE_OK could be used for this?