Re: A recent patch increased GCC's memory consumption!

[Jan Hubicka <jh at suse dot cz>]

> > On Sat, 16 Jul 2005, Jan Hubicka wrote:
> > 
> > > Joseph,
> > > any idea why the patch increased memory by up to 18%?
> > 
> > It should if anything have reduced memory consumption by causing fewer 
> > types to be allocated - you'll need to give more details of what sort of 
> > objects there are more of being allocated and where they are being 
> > allocated.
> > 
> > One possibility might be that the optimizers are working better now they 
> > have fewer integer types with the same width, and this causes them to 
> > allocate more memory - but if so, the places where the optimizers care 
> > about which of multiple types with the same width is being used would be 
> > worth tracking down, as this could also apply with e.g. int and long, not 
> > just with bit-field types.
> 
> I am attaching logs from today and yesterday.  Looks like tree
> optimizers are actually working worse, at least more RTL instructions
> are produced so about every pass consume more memory.

Also note the difference in between O2 and O3 results.  O2 seems to
increase memory usage in non-GGC only that might be register allocator
or something, but O3 seems to just inline more.  I am having very little
time to look into it today, so just wild guess, but can your patch affect
inlining decisions?

Honza