PR 31396 (suboptimal register allocation)

[Vladimir Makarov]

Jan Hubicka wrote:
> Hi,
>   
>> Thanks for the patch.  The patch is OK.  You can commit it.
>>
>> The analogous approach is used in IRA but on the base of IRA general 
>> cost model.
>>
>> I think that factor 4 in CALLER_SAVE_PROFITABLE might be bigger than 
>> necessary because  only store and load is necessary around the call (it 
>> means factor 2) and caller-saves.c frequently removes redundant 
>> load/store around calls when there are more one call in a basic block.
>>     
>
> I've commited the patch.  Quite interestingly tonight improvement in
> SPECfp especially with FDO bigger than I've expected
> http://www.suse.de/~gcctest/SPEC/CFP/sb-haydn-fdo-64-FDO/recent.html
> (it might be different patch comitted today too, but I don't see obvious
> candidate for it)
>
>   
My tester also found about 1% improvement on x86_64 on SPECFP2000.

http://vmakarov.fedorapeople.org/spec/spec2000.toolbox/gcc/global-run-ratio.html
> It seems that there is quite a bit of room for improvement here: the
> heuristics is definitly overestimating cost of calls for long lived
> pseudos.  Did you try on IRA to take into account 2)?  It looks like
> load/store placement for caller save/restore is LCM problem, perhaps it
> can be solved for all pseudo that might be allocated to caller saved
> registers and live across call to get realistic cost estimates
> beforehand?
>
>   
Yes, I did LCM for pseudos got caller-saved hard registers in IRA 
(former file ira-calls.c).   After moving the load/store, I improved the 
allocation by using a simple prioirity coloring algorithm to assign hard 
registers to pseudos got memory and conflicting with the moved 
pseudo-registers.

I have LCM in caller-saves.c on IRA-branch too.  Although doing LCM in 
IRA with reassigning after that generates a bit better code than one in 
caller-saves.c, I removed LCM in IRA a month ago because it takes to 
much time (mainly rebuilding conflicts).  So now LCM is done only 
caller-saves.c on IRA.

LCM is good to have but I don't think LCM is good to do to evaluate 
caller-saved hard register profitability.  It is too pricey with a small 
improvement.  In any case it is too late to be done for 4.3.  As for 
4.4, I am planing to submit IRA to the main line and hopefully it will 
be in gcc 4.4.  Besides mentioned LCM, IRA (and the reload: now actually 
after each reload pass IRA called to improve assignment because of 
register spilling) has many more improvements too and I don't think it 
is reasonable to work on the old register allocator anymore.  But of 
course you can do it if you want.


Vlad