Re: uvm_map improvements

2010-02-25 Thread Anton Maksimenkov
2010/2/26 Otto Moerbeek : >> 2. ... >> I think that we may introduce some variable - sysctl variable or >> malloc.conf flag - which will prevent sys_mmap() or uvm_map_hint() > If fragmentation really is a problem for you, my first reaction is: > use a machine with a bigger adress spacce. No deal.

Re: uvm_map improvements

2010-02-25 Thread Theo de Raadt
> There are cases when we need memory rather than randomization. > Dedicated database server is such case. Postgres can give more > performance if it can use as much memory as userspace can allocate. > Many big random holes quickly leads to fails of allocations. So then > we must dramatically decre

Re: uvm_map improvements

2010-02-25 Thread Otto Moerbeek
On Fri, Feb 26, 2010 at 10:58:09AM +0500, Anton Maksimenkov wrote: > 2010/2/20 Anton Maksimenkov : > > it was started discussion about uvm_map.c improvements. I prepared the > > diff and trying to explain it. > > In addition to that diff I want to discuss two problems. > > 1. > AFAIK, the usersp

Re: uvm_map improvements

2010-02-25 Thread Anton Maksimenkov
2010/2/20 Anton Maksimenkov : > it was started discussion about uvm_map.c improvements. I prepared the > diff and trying to explain it. In addition to that diff I want to discuss two problems. 1. AFAIK, the userspace malloc() uses mmap (sys_mmap) to do it's job. As we can see, sys_mmap() uses uvm

Notebooks en cuotas!!!

2010-02-25 Thread Rodrigo Ferreri
[IMAGE] [IMAGE] [IMAGE] [IMAGE] [IMAGE] [IMAGE] [IMAGE] [IMAGE] Para visualizar este correo en el explorador, clic aqum Este correo ha sido enviado a tech@openbsd.org | Remover suscripcisn [IMAGE]

e-shop.gr: Νέα HP Notebooks σε μοναδικές τιμές!

2010-02-25 Thread members
Episjeuhe_te tgm die}humsg http://www.e-shop.gr/newsletter/mail-100220.html cia ma de_te tir pqosvoq]r lar TGKEVYMIJES PAQACCEKIES 9:00-20:00 STO 211 5000500 Oi til]r isw}oum ap| 20/02/10 l]wqi 06/03/10, ]yr enamtk^seyr tym apohel\tym jai l|mo cia ta l]kg tou e-shop.gr Am h]kete ma diacqave_t

Još 3 dana - popusti hit proizvoda!

2010-02-25 Thread E-topshop
PoE!tovani, Super ponude za mesec Ljubavi traju do kraja februara! PodseDamo Vas da iskoristite ponudu za zaljubljene koja Vam je na raspolaganju joE! samo 3 dana! Najbolje popuste na najpopularnije proizvode moE>ete videti ako kliknete ovde! Ne gubite vreme, prava je prilika da iznenadite onog

[patch] core.5 - core dumps for setuid processes

2010-02-25 Thread Rogier Krieger
Seeing a thread on misc@ [1] earlier, I was wondering if the attached patch would be an improvement to show a user where to look if they do want a core dump from a setuid process. I changed the wording slightly since my earlier posting. Regards, Rogier References 1. MARC.info - 'Core dumps from

Re: Poor performance with gem(4)?

2010-02-25 Thread Stuart Henderson
On 2010/02/25 00:18, Brian Keefer wrote: > For the record, I've just patched -current with Brad's flow control diff from > last year and now I'm getting 90Mbps with tcpbench on this card. I'm not sure > if it was that diff or some other one in the last three years, but 800Kbps -> > 90Mbps is quite

Re: Poor performance with gem(4)?

2010-02-25 Thread Brian Keefer
For the record, I've just patched -current with Brad's flow control diff from last year and now I'm getting 90Mbps with tcpbench on this card. I'm not sure if it was that diff or some other one in the last three years, but 800Kbps -> 90Mbps is quite a nice improvement. Pity I didn't attach ifconf