On Wednesday September 4 2002 12:08 pm, Bjarne Thomsen wrote:
> Hi Tom,
>
> I found the following answer from Alan Cox to a question on
> the kernel mailing list. It looks as if he has a patch for the
> standard 2.4.19 kernel. What I would like to know is, if
> Mandrake soft is planning to make a similar modification for
> their kernel. Please, find below the output from the commands
> that you requested.
>
> -- Bjarne
Look below for my comments. You'll probly haft'a ask Mdk for an
answer to your question. Maybe Vincent D will reappear with what's
allowed ? or you can try on the cooker list and/or archive ;) I've
avoided Intel chipsets since they ceased the BX, IOW's i8<anything>, so
I can't be much help.
> On Thu, 2002-07-18 at 13:45, Andrew Halliwell wrote:
> > The P4B533 has the intel 801DB IDE controller (stated as supported
> > in
>
> rc1)
>
> > but in every 2.4 kernel I've seen so far, this appears in the
> > bootup.
> >
> > Uniform Multi-Platform E-IDE driver Revision: 6.31
> > ide: Assuming 33MHz system bus speed for PIO modes; override with
>
> idebus=xx
33 is correct, read 'ide.txt' for more info, it's at
/usr/src/linux-<kernel-version>/Documentation/ide.txt
> > PCI_IDE: unknown IDE controller on PCI bus 00 device f9, VID=8086,
>
> DID=24cb
>
> > PCI: Device 00:1f.1 not available because of resource collisions
> > PCI_IDE: (ide_setup_pci_device:) Could not enable device
>
> Blame your BIOS vendor
>
> The -ac tree has workarounds for the BIOS forgetting to set up the
> chip. Let me know if rc1-ac7 works for you.
>
>
> cat /proc/interrupts:
> CPU0
> 0: 11650099 XT-PIC timer
> 1: 100563 XT-PIC keyboard
> 2: 0 XT-PIC cascade
> 8: 1 XT-PIC rtc
> 9: 1348969 XT-PIC eth0, CMI8738-MC6
Whenever you see items on one line separated by a comma, they're
sharing the IRQ, potential conflict. Can you try moving your 3Com card
to a different slot? Caution, avoid the 1st pci slot (the one next to
the AGP slot) if possible, it already shares the AGP IRQ. I'm asuming
your Cmedia is integrated.
<snip most of lspci stuff>
> snd-cmipci : C-Media Electronics Inc|CM8738
> 3c59x : 3Com Corporation|3c905C-TX [Fast Etherlink]
>
> # hdparm -i /dev/hdb1
>
> /dev/hdb1:
> HDIO_GET_IDENTITY failed: Invalid argument
Not quite right, you're tryin a partition, use
'hdparm -i /dev/hdb' to access the drive's firmware ....
> # hdparm -v /dev/hdb1
>
> /dev/hdb1:
> multcount = 16 (on)
> IO_support = 0 (default 16-bit)
> unmaskirq = 0 (off)
> using_dma = 0 (off)
> keepsettings = 0 (off)
> readonly = 0 (off)
> readahead = 8 (on)
> geometry = 9732/255/63, sectors = 156344517, start = 63
.... Here again, a simple 'hparm -v /dev/hdb' is all that's required.
The '1' you added was ignored. It shows tho that neither 32bit, or DMA
are enabled.....
>
> On Wed, 2002-09-04 at 03:19, Tom Brinkman wrote:
> > On Monday September 2 2002 11:50 pm, Bjarne Thomsen wrote:
<snip again>
> > > on an Asus P4B533 MB?
> > > The speed of the harddisks is about 3 Mb/s, and
> > > kupdated sucs up CPU time. It is evidently in PIO mode.
... and yes it's in PIO mode. hdparm -t should be much better, at
least 10mb/sec for a 5400rpm ata/33 drive. ata/100, 7200rpm or better
should get ~40mb+/sec. You could try usin 'hdparm -c1 -d1 /dev/hdb'
-c1 turns on 32bit, -d1 turns on DMA. BUT, I suggest you do some
research first. I vaguely remember that there's issues with DMA and
i8xx chips, probly why the drive wasn't optimized to begin with. I
suspect it maybe a Mdk or kernel work-around. Sorry I can't be of more
help.
--
Tom Brinkman Corpus Christi, Texas
Want to buy your Pack or Services from MandrakeSoft?
Go to http://www.mandrakestore.com
