On Tue, 25 Jun 2013, Konstantin Belousov wrote: > > > Updates to the counter cannot be done from the interrupt context. > > > > This is fragile, however. It prevents using counters for things like > > counting interrupts. Most interrupt counting is now done directlyly > > and doesn't use PCPU_INC(). i386/i386 has just 1 use of PCPU_INC(). > > It is to count traps in machdep.c. Traps include nonmaskable > > interrupts. Even hard-disabling of interrupts doesn't prevent these. > > Otherwise, PCPU is used mainly for vm counters. E.g., for pagefaults. > > Now the trap is not an interrupt, so it shouldn't occur in the middle > > of the counter update and the PCPU_INC() can safely be replaced by > > a counter, but this is not clear. > Traps are not performance critical in the sense that there is no need to count > up to 1-10G traps per second. Anyway, as Gleb said, there is no point in > optimizing the i386 kernel.
Hmm, don't we count semi-embedded routers? Or, do we think they are either amd64 or arm/mips? [snip all the rest] -- Sincerely, D.Marck [DM5020, MCK-RIPE, DM3-RIPN] [ FreeBSD committer: ma...@freebsd.org ] ------------------------------------------------------------------------ *** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- ma...@rinet.ru *** ------------------------------------------------------------------------ _______________________________________________ svn-src-head@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"