Re: hdparm, cdrom, and retaining params
On Sat, 28 Dec 2002, Alvin Oga wrote:
Drifiting to hardware issues...
I'm trying hard to get a better grasp of this stuff -- short of
reading ATA and PCI specs! I don't really understand the difference
between PIO modes and DMA settings, for one thing. I also have no idea
what all these settings do in real life -- especially for reasonably slow
CD drives. I suppose that can't all be explained in a few paragraphs,
though.
Ok, so I guess it makes sense to enable the fastest mode that's available
on *both* the IDE controller and the drive. And in this case the Intel
PIIXn chipset is UDMA/33 and so is the Plextor CD-R.
So I should set the -X66 flag to select UDMA mode 2.
Any idea how that setting works? Does hdparm send a flag to the kernel
IDE subsystem to enable udma mode 2, which knows to set a bit in the IDE
chipset (e.g. in this case UDMACTL at offset 48h bit 2 for the secondary
interface drive 0).
Or is it just telling the drive to go in that mode and we hope that the
IDE chipset is enabled for udma mode 2 as well? And if that's the case,
how does the IDE chipset get enabled for that mode?
BTW - I also notice that the hard drive on this machine is also not set
the best:
/dev/hda:
multcount = 16 (on)
I/O support = 0 (default 16-bit)
unmaskirq = 0 (off)
using_dma = 1 (on)
keepsettings = 0 (off)
nowerr = 0 (off)
readonly = 0 (off)
readahead = 8 (on)
geometry = 1662/255/63, sectors = 26712000, start = 0
busstate = 1 (on)
I wonder if the kernel IDE controller for the PIIXn chipset can't be a bit
more verbose on how it's configuring the drives. When I get a chance I'll
take your advice and disable everything not related in the kernel and see
if that fixes things.
I also notice this message:
ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
Is that the PCI bus speed to the IDE controller? How do I track down if
that can indeed be set differently?
Thanks,
--
Bill Moseley moseley@hank.org
Reply to: