* shared memory segment size alignment
@ 2004-01-16 18:31 umut aymakoglu
2004-01-16 19:18 ` David Mosberger
0 siblings, 1 reply; 2+ messages in thread
From: umut aymakoglu @ 2004-01-16 18:31 UTC (permalink / raw)
To: linux-ia64
Hi -
In some kernel versions the SHMLBA is defined as
1MB(1024*1024) and in some as 16K. What would be the
difference between aligning the shared memory segment
size on 1MB versus 16K before calling shmget()? Would
there be any performance gain/lose? We are aligning
the segment address on 1MB.
thanks,
Umut
__________________________________
Do you Yahoo!?
Yahoo! Hotjobs: Enter the "Signing Bonus" Sweepstakes
http://hotjobs.sweepstakes.yahoo.com/signingbonus
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: shared memory segment size alignment
2004-01-16 18:31 shared memory segment size alignment umut aymakoglu
@ 2004-01-16 19:18 ` David Mosberger
0 siblings, 0 replies; 2+ messages in thread
From: David Mosberger @ 2004-01-16 19:18 UTC (permalink / raw)
To: linux-ia64
>>>>> On Fri, 16 Jan 2004 10:31:11 -0800 (PST), umut aymakoglu <umutaymak@yahoo.com> said:
umut> Hi - In some kernel versions the SHMLBA is defined as
umut> 1MB(1024*1024) and in some as 16K. What would be the
umut> difference between aligning the shared memory segment size on
umut> 1MB versus 16K before calling shmget()? Would there be any
umut> performance gain/lose? We are aligning the segment address on
umut> 1MB.
The ia64 architecture allows CPUs to have a performance penalty when
there is virtual aliasing below a 1MB boundary. So if you don't align
to 1MB, it _could_ run slower. No existing processor models do that,
but future models might.
--david
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2004-01-16 19:18 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2004-01-16 18:31 shared memory segment size alignment umut aymakoglu
2004-01-16 19:18 ` David Mosberger
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox