[RFA] The Integrated Register Allocator

[Vladimir Makarov]

Mark Mitchell wrote:
> Vladimir Makarov wrote:
>
>> The single requirement is to define macro IR_COVER_CLASSES in the 
>> machine description file.   There are a lot of variants for the 
>> choice and IRA is very sensitive to this decision (I tried long ago 
>> to make the choice automatically but failed to do it even for the 
>> targets it currently works for).  It means that the quality will be 
>> very strong dependent from the choice.  The code performance for a 
>> wrong choice may be even much worse than one for the old register 
>> allocator.
>
> Is there any way to compute this from the given macros, in a vaguely 
> sensible way?
>
As I wrote I've tried to do this and fail.  I think it is more easy for 
me to write these macros for all targets (i think it is a work for a 
week or two) than to design an algorithm for the task.