> -Original Message-
> From: Cliff Woolley [mailto:[EMAIL PROTECTED]]
> Sent: 25 November 2002 23:51
> On 25 Nov 2002, Brian Pane wrote:
>
> > > What should the max freelist size be? 64K? 32K? If somebody says "it
> > > should be configurable," I'm gonna shoot them. ;) hehe It probab
>>> [EMAIL PROTECTED] 11/25/02 03:00PM >>>On Mon,
2002-11-25 at 14:50, Cliff Woolley wrote:> Well, it just occurred to
me that most of the work for doing it> configurably is already
there. There's already a pools-max-free> directive, so if we just
have the pools and the bucket allocator sha
On Mon, 2002-11-25 at 14:50, Cliff Woolley wrote:
> Well, it just occurred to me that most of the work for doing it
> configurably is already there. There's already a pools-max-free
> directive, so if we just have the pools and the bucket allocator share the
> same apr_allocator_t, we should be g
On 25 Nov 2002, Brian Pane wrote:
> > What should the max freelist size be? 64K? 32K? If somebody says "it
> > should be configurable," I'm gonna shoot them. ;) hehe It probably
> > SHOULD be, but I don't have time to implement that right now.
>
> I think 16KB may be enough for most general u
On Mon, 2002-11-25 at 14:13, Cliff Woolley wrote:
> So thanks to prodding from Jean-Jacques Clar at Novell, I'm putting in an
> apr_bucket_alloc_max_free_set() that just calls
> apr_allocator_max_free_set() on the underlying allocator. Now we need to
> call that function each time we create an apr
So thanks to prodding from Jean-Jacques Clar at Novell, I'm putting in an
apr_bucket_alloc_max_free_set() that just calls
apr_allocator_max_free_set() on the underlying allocator. Now we need to
call that function each time we create an apr_bucket_alloc_t, which
happens in each MPM.
My only ques