This is the mail archive of the gcc-patches@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: [tree-ssa] Put CFG into GGC memory


> On Tuesday 02 December 2003 16:27, Andrew MacLeod wrote:
> > It feels to me like we are running around like chickens with no heads at
> > the moment regarding memory allocation.
> 
> I love chickens, do you mind?  :-)
>  
> > Can we summarize exactly what we think we changing and why?
> >
> > I'll start.
> >
> > - I am changing the use/def/vuse/vdef operands in stmts to not be
> > varray's any more.
> 
> Good.  That was on my wish list.
> 
> > They'll either use their own little manager, or maybe
> > ggc if they can get their own page and/or a ggc_free maintained list.
> 
> They can get their own zone.  Is that good enough?
> 
> I want to use the zone collector to improve locality wrt. the (default) page 
> collector.  I want to put SSA_NAME and PHI_NODE nodes in a separate zone 
> because we know they will die after writing out of tree-SSA, and I want to 
> put all CFG objects in a separate zone because we know they are usually 
> referenced in groups.

This all sounds fine.  Looking forward for the CFG bits :)
> 
> I also would like us to be GGC safe everywhere.

How everywhere your definition of everywhere is?
We do have many and many places we do save produced expressions in local
variables and it does not seem to be very practical to make these
externally visible all the time...

Honza
> 
> Gr.
> Steven


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