>>
>> I've actually noticed this myself. Recently I compiled XFree86 4.0. Now
>> I've
>> seen people with pII/350's compile it in 20 minutes straight. But even with
>> make -j4 (4 processes at a time) my dual 500 took _40_ minutes to do the
>> same
>> job. If you ask me this motherboard is just plain dodgy.
>
> Well I'm using -j 8.
> (-j 4 the processor is still sleeping - also are you using DMA for disk
> transfers ?? (what is your disk transfer speed - is usualy this is
> the bottleneck of the system - especialy if you are compiling -
> the faster the harddrive the faster it will be....)
>
> As far as I know the kernel compiles 5 minutes with PIII500MHz
> - and just 3.5 with dual celeron.
>
> But this might be just my observation....
Well, I was talking about XFree86 though, not the kernel. Although, if you
want to talk about kernels, according to your stats a pIII 500 UP does it in 5
minutes, and this is my Dual Celeron 500 w/ BP6:
make -j8 bzImage 506.80s user 44.28s system 153% cpu 5:59.99 total
Almost 6 minutes. Shocking.
My disk transfer speed is as follows:
Password:
backdraft# hdparm -t -T /dev/sda
/dev/sda:
Timing buffer-cache reads: 128 MB in 1.77 seconds = 72.32 MB/sec
Timing buffered disk reads: 64 MB in 4.63 seconds = 13.82 MB/sec
My disk is a Seagate Barracuda UW and has no trouble with access time. It
doesn't seem to be under any stress at all while doing a make -j8.
Unfortunately we can't blame the HPT366 this time ;-).
Regards,
Patrick
--
=- To unsubscribe, email [EMAIL PROTECTED] with the -=
=- body of "unsubscribe linux-abit". -=
