On Thursday February 13 2003 12:09 am, Ryan Moe wrote:
> I have an Asus A7V133c m/b w/ an Athlon T-bird 1.3. It's not an
> OEM board. The other drives I have are /dev/hda some old shitty
> cd-rom, /dev/hdg a maxtor 30 gig running at ATA-100, /dev/hde a
> quantam 30 gig ATA-100. You think being less than 2 years old
> kernel support would have gone beyond my 12X drive?
Depends on the firmware. Have you flashed to the latest? Check the
upgrade first to see what bugfixes or improvements it provides.
There's no sense in flashing a to newer bios if all it provides is a
fix for PPC (Macintosh). MOF, in a case like that you could be worse
off.
I've gotten
> some pretty old, crappy hardware working w/ new kernels (IDE
> controller cards and whatnot). I wasn't using windows as a "hey it
> works in here why not linux", just that if it was something
> physically wrong (i.e. not a driver/config problem) it should show
> up every time I use the drive regardless of OS.
Windoze is fairly tolerant of marginal or problem hardware. Partly
cause it's just a sloppy OS, partly because fixes and work-arounds
are incorporated into drivers and dll's, vxd's, etc. M$ enjoys and
demands close support from hardware vendors. They do get the source
and specs, and often make changes in the hardware to suit their needs
(deficiencies).
here is the output
> of the hdparm command.
Hdparm results looked OK. The A7V133c should be a _good_ board for
Linux, specially with the kt133a VIA chipset. Particularly if it's a
revision 03. The fourth revision (0,1,2,3), which has the rumored ide
bug fixed (see the output from 'lspci' to get your revision number).
BUT with the exception of the faulty capacitors issue, and Asus
has been unfortunately involved in that for quite a while. Inspect
the caps carefully to see any evidence of burn (brown, yellowish)
markings, leaks (specially from the top, usually yellowish-orange
cream), or any swelling, bulging of the cylinder, or the disc on top.
Are they standin straight up (90º to the mobo), or do they appear to
have been bent over at some point in time (ie, damaged)?
I suspect tho the IDE controller, or the arrangement, order of
your drives may be the problem. What controller? If you've got two
CD drives, they should be on the same channel. Obviously you've got
more than two channels, ide0, ide1, ide2 .... Frankly IMO that
should be avoided. The old 33 mhz PCI bus can hardly handle two as it
is, specially with drives rated ata/66 or higher. But I think you
need to clarify where you've got each drive you have.
hda=
ide0 -- |
hdb=
hdc= Creative Burner
ide1 -- |
hdd=
.... and so on. IMO, your busy HDD's should be on ide0. I've had best
results with the cdrom as master, the burner as slave (on ide1, with
no more ide channels than those two). YMMV.
There's one other possible problem, tho I believe your ide
arrangement is much more likely. What PSU? brand, model#, and watts.
If you have lm_sensors installed, are your voltages on spec or
slightly over? are they rock steady? If they are your caps are most
likely OK.
So if your caps look OK, and your voltages are correct and steady
from a good PSU, then I believe civileme could probly offer more help
than me with you ide arrangement. As you can probly tell, I'm sort'a
biased against integrated or addon controller cards to overload the
PCI bus. What other PCI (including AGP) peripheals do you have on
the PCI bus? There's a tax to be paid with high bandwidth video and
sound cards, and it's paid by overall degradation of the PCI/IDE bus.
--
Tom Brinkman Corpus Christi, Texas
Want to buy your Pack or Services from MandrakeSoft?
Go to http://www.mandrakestore.com
