A new gimple pass (LRS: live range shrinking) to reduce register pressure

Andrew Pinski pinskia@gmail.com
Tue Dec 30 20:02:00 GMT 2008


On Tue, Dec 30, 2008 at 1:28 PM, Xinliang David Li <davidxl@google.com> wrote:
>>
>> All the normalize_gimple_stmt stuff looks weird - you basically copy
>> the whole IL
>> into yet another representation?  Note that all the uses/defs you collect are
>> readily and easily available by the SSA operand interface.
>
> Andrew mentioned that too. The only benefit of that is that the
> normalized lrs stmt is like a cached stmt -- it is slightly more
> efficient to walk through than using op_iter_next_use (via macro),
> especially when the traverse happens often.

op_iter_next_use is already a cache :).  It is a linked list though so
pointer chasing will happen.

Thanks,
Andrew Pinski



More information about the Gcc-patches mailing list