This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: GVN-PRE sucks -- well, for SPECCFP2000 mgrid anyway


On Sun, 2004-08-15 at 21:08, Kenneth Zadeck wrote:
> 

> This is handled differently in different compilers.  There are basically 
> two schools of thought here:
> 
> 1) let pre be pre and have the register allocator "rematerialize" the 
> guys that have been pulled out of the loop.
> 
> 2) make each of the optimizations have a cost model that factors in the 
> way that several transformations effect "max live" i.e. the number of 
> items that will need to be kept in registers at any point in the program. 
> 
> I have always been in favor of the first school but there really are two 
> sides to this issue.  In the gcc world there are two parts of the 

Sorry, I've been away for a bit.

Not only would I promote (1), but I really despise (2) since it clutters
everything up.

When we were doing some of the early SSA implemention, we recognized
there would be a lot of pressure put on the register allocator which
would eventually need to be addressed. I do intend to take on some of
the register allocator issues shortly, I just have to finish up some
remaining SSA issues.

Of course, shortly doesnt mean anything will actually get fixed in the
near future, this is not a trivial issue, as everyone well knows! :-)


Andrew


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]