Re: Kernel 2.2.19 crashes, 2.4.4 bad performance
Hi,
[...]
> Not all UP2000s run with EV67 (or 68), there are some that have
> EV6 slotBs, yours is a 67 though.
I didn't know that. Mine is of course an EV67.
> If I remember correctly RTC in 2.2 is broken for SMP....did you have 2
> CPUs? Don't remember seeing that.
Oops, sorry. Yes, i do have two CPU's. If RTC is broken wrt SMP, i'll disable it.
Still remaining, however, is the question, if you encounter clock-skews without
CONFIG_RTC, as it has been in earlier 2.2.x kernels.
> There's two adaptec drivers in 2.4 right? One from Adaptec and then the
> original one (which is no longer maintained). Haven't even tried to
> compilethe adaptec version yet.....perhaps that could be one of the
> differences??
Like I wrote, i did test the drive performance in various ways and did not
encounter any significant differences. Thus, i'm fairly sure, the driver differences
do not account in that rather severe performance impact. Set aside the strange
behaviour of getrusage(2) in kernel 2.2.19, my key suspects are either the
new VM sussystem or the recent scheduyler tweaks.
--
Thomas Weyergraf kirk@colinet.de
My Favorite IA64 Opcode-guess ( see arch/ia64/lib/memset.S )
"br.ret.spnt.few" - got back from getting beer, did not spend a lot.
Reply to: