On 2014-04-28 10:57:12 -0400, Tom Lane wrote:
> Andres Freund <andres@2ndquadrant.com> writes:
> > On 2014-04-28 10:03:58 -0400, Tom Lane wrote:
> >> What I find much more worrisome about Andres' proposals is that he
> >> seems to be thinking that there are *no* other changes to the buffer
> >> headers on the horizon.
>
> > Err. I am not thinking that at all. I am pretty sure I never made that
> > argument. The reason I want to limit the number of connections is it
> > allows *both*, shrinking the size of BufferDescs due to less alignment
> > padding *and* stuffing the refcount and flags into one integer.
>
> Weren't you saying you also wanted to stuff the usage count into that same
> integer? That's getting a little too tight for my taste, even if it would
> fit today.
That's a possible additional optimization that we could use. But it's
certainly not required. Would allow us to use fewer atomic operations...
Right now there'd be enough space for a more precise usagecount and more
flags. ATM there's 9 bits for flags and 3 bits of usagecount...
Greetings,
Andres Freund
-- Andres Freund http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training &
Services