On 5/8/20 9:15 AM, Alex Bennée wrote:
>
> Richard Henderson <richard.hender...@linaro.org> writes:
>
>> V3 fixes the target/sparc regression during register allocation.
>
> Which patch has the fix in it? I couldn't see any per-patch rev notes.
Patch 3:
> + /*
> + * If the input is readonly, then it cannot also be an
> + * output and aliased to itself. If the input is not
> + * dead after the instruction, we must allocate a new
> + * register and move it.
> + */
> + if (temp_readonly(ts) || !IS_DEAD_ARG(i)) {
> + goto allocate_in_reg;
> + }
> Also I'm sure I reviewed a bunch of these, did they not count because I
> found a bug ;-)
Oops, you're right. I simply missed them, having gotten distracted by the bug.
r~
