* 2.5 SYSENTER context-switch overhead on P4 :-(
@ 2003-01-05 20:33 Mikael Pettersson
0 siblings, 0 replies; only message in thread
From: Mikael Pettersson @ 2003-01-05 20:33 UTC (permalink / raw)
To: torvalds; +Cc: linux-kernel
I've measured the cost of the two wrmsr()'s added to the
context-switch path to support SYSENTER in 2.5 kernels.
Here's the data (clock cycles, averaged):
CPU wrmsr only only
type both cs esp
---- ---- --- ---
P4 2.4GHz (Xeon) 1722 860 863
P4 1.7GHz (Willamette) 1642 816 817
P3 800MHz (Coppermine) 158 76 76
P3 450MHz (Katmai) 216 105 105
P2 233MHz (Deschutes) 176 88 88
P2 233MHz (Klamath) 150 75 75
The other implementation with keeps the MSRs constant and
uses a trampoline stack might be a better alternative.
/Mikael
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2003-01-05 20:25 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2003-01-05 20:33 2.5 SYSENTER context-switch overhead on P4 :-( Mikael Pettersson
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®