Lately Nikos Ntarmos <[EMAIL PROTECTED]> said:
> Ummm... What compiler version are you use? The pointer variant produces
> faster code for me on both NetBSD/alpha (gcc 3.3.3) and Linux/x86 (gcc
> 3.3.5), using both -O0 and -O2 (other compiler flags also tested with
> similar results). Perhaps somet
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi there.
On Sat, Jul 02, 2005 at 01:27:00PM +0300, Andriy Tkachuk wrote:
> here even more refereces to memory in your variant.
Ummm... What compiler version are you use? The pointer variant produces
faster code for me on both NetBSD/alpha (gcc 3.3.3
mov%edx,(%eax)
+22c1: 83 43 0c 04 addl $0x4,0xc(%ebx)
- Original Message -
From: "Nikos Ntarmos" <[EMAIL PROTECTED]>
To: "ant" <[EMAIL PROTECTED]>
Cc: <[EMAIL PROTECTED]>
Sent: Thursday, June 30, 2005 7:31 PM
Subject
> if one will decide to commit first optimization (about buckets),
> then there must some adjustments be made also
> regarding correct statistics gathering.
It seems that all is fine with statistics,
i mistook.
___
freebsd-hackers@freebsd.org mailing l
> On Thursday 30 June 2005 10:35 am, Andriy Tkachuk wrote:
> > I just checked the object code - you right, it almost the same:
> >
> > - bucket->ub_bucket[bucket->ub_cnt] = item;
> > -22b9: 0f bf 43 08 movswl 0x8(%ebx),%eax
> > -22bd: 8b 4d 0c mov0xc(%ebp),%ecx
> I ran ministat against your tests with 1000 sockets loop and there isn't a
> lot
> of difference in the user times:
it was not supposed to be (the difference in the user times)
___
freebsd-hackers@freebsd.org mailing list
http://lists.freebsd.org/m
0(%ebp)
>
>
> - Original Message -
> From: "Max Laier" <[EMAIL PROTECTED]>
> To:
> Cc: "ant" <[EMAIL PROTECTED]>
> Sent: Thursday, June 30, 2005 3:15 PM
> Subject: Re: hot path optimizations in uma_zalloc() & uma_zfree()
>
> > Ano
ant wrote this message on Thu, Jun 30, 2005 at 01:08 +0300:
> I just tryed to make buckets management in perCPU cache like in
> Solaris (see paper of Jeff Bonwick - Magazines and Vmem)
> and got perfomance gain around 10% in my test program.
> Then i made another minor code optimization and got ano
Hi there.
I wouldn't have gone into this if ant hadn't produced that 10% figure
for the speed improvement with simply reordering of increments and
dereferences (although jhb@ reported the speed-up he noticed was much
less than that).
I attach* a patch that: (i) incorporates ant's exchange of uc_f
;
To:
Cc: "ant" <[EMAIL PROTECTED]>
Sent: Thursday, June 30, 2005 3:15 PM
Subject: Re: hot path optimizations in uma_zalloc() & uma_zfree()
> Another optimization is very trivial, for example:
> - bucket->ub_cnt--;
> - item = bucket->ub_bucket[bucket->ub_cn
On Wednesday 29 June 2005 06:08 pm, ant wrote:
> Hi folks.
>
> I just tryed to make buckets management in perCPU cache like in
> Solaris (see paper of Jeff Bonwick - Magazines and Vmem)
> and got perfomance gain around 10% in my test program.
> Then i made another minor code optimization and got an
On Thursday 30 June 2005 14:15, Max Laier wrote:
> On Thursday 30 June 2005 00:08, ant wrote:
> > @@ -2263,8 +2262,7 @@
> >if (bucket->ub_cnt < bucket->ub_entries) {
> > KASSERT(bucket->ub_bucket[bucket->ub_cnt] == NULL,
> > ("uma_zfree: Freeing to non free bucket index."));
> > -
> Do you keep two buckets still? If so, this is something that I've always
> intended to do, but never got around to. I'm glad someone has taken the
> initiative. Will review the patch shortly.
yes, shure. Linux keeps only one array, that is 2 times large of bucket
and it copies half of this a
On Wed, 29 Jun 2005, Robert Watson wrote:
On Thu, 30 Jun 2005, ant wrote:
I just tryed to make buckets management in perCPU cache like in Solaris
(see paper of Jeff Bonwick - Magazines and Vmem) and got perfomance gain
around 10% in my test program. Then i made another minor code optimizatio
On Thursday 30 June 2005 00:08, ant wrote:
> I just tryed to make buckets management in perCPU cache like in
> Solaris (see paper of Jeff Bonwick - Magazines and Vmem)
> and got perfomance gain around 10% in my test program.
> Then i made another minor code optimization and got another 10%.
> The p
On Thu, 30 Jun 2005, ant wrote:
I just tryed to make buckets management in perCPU cache like in Solaris
(see paper of Jeff Bonwick - Magazines and Vmem) and got perfomance gain
around 10% in my test program. Then i made another minor code
optimization and got another 10%. The program just cre
16 matches
Mail list logo
