Re: disk I/O slower with kernel 2.4.9

Steven Cole (elenstev@mesatop.com)
Sat, 18 Aug 2001 21:06:48 -0400


On Saturday 18 August 2001 09:19 pm, Luigi Genoni wrote:
> I still have to try it, tomorrow i will post the results....
>
> On Sat, 18 Aug 2001, Rik van Riel wrote:
> > On Sun, 19 Aug 2001, Luigi Genoni wrote:
> > > just making time make -j 2 bzImage with kernel source 2.4.9
> > > gives me:
> > >
> > > real 3m36.041s
> > > user 2m2.950s
> > > sys 0m9.740s
> > >
> > > while compiling the same sources running kernel 2.4.7 gives:
> > >
> > > real 2m28.350s
> > > user 1m56.150s
> > > sys 0m5.262s
> >
> > How does 2.4.8-ac7 do ?
> >
> > Rik

Rik, Luigi:

Excuse me for jumping in, but here are a few more data points,
using a slower system. I built 2.4.8-ac7 with the following kernels, using:

time make -j2 'MAKE = make -j2' bzImage

to produce the output. Yes, I know this is overkill for an UP system.
This version of time is GNU time 1.7. Each kernel build was performed
right after the boot and login, running KDE 2.2 and an xterm.

The system is UP, PIII 450, 384 MB, ATA-33, /usr/src mounted ReiserFS
on a WDC WD102AA, the rest of the system ReiserFS on a WDC WD136AA.

2.4.7
681.66user 80.16system 13:32.51elapsed 93%CPU (0avgtext+0avgdata
0maxresident)k

2.4.8
682.90user 83.25system 13:29.98elapsed 94%CPU (0avgtext+0avgdata
0maxresident)k

2.4.9
684.77user 80.55system 13:32.35elapsed 94%CPU (0avgtext+0avgdata
0maxresident)k

2.4.8-ac7
683.99user 80.54system 13:21.91elapsed 95%CPU (0avgtext+0avgdata
0maxresident)k

Steven
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/