Ok, I have the same problems with my latest Maxtor drive in ide2 (doesn't matter which interconnect) and the issue revolves around the specific firmware that comes with the Maxtor HD. I have to turn off the dma for that specific settings because the irq and/or partitions fail (at least most of the time) and this solved the issue. Also, in my case, I have the HD on #2 interconnect on ide1 and this is mostly likely what the problem is. I really should change the settings in my fstab, the shutdown and change the interconnects to the onboard promise connects - but I digress. So, the problem isn't so much about setting things with hdparm as it is with the boot process trying to enable DMA at start up and failing. I have to use the boot parameter "ide=nodma" and then I set the dma in the kde CC/yast for that specific HD to off and the others to on. Though this is optimal or preferred it does solve my problem and I no longer suffer for lost IRQ's or bad patition readings. It's about the Maxtor drive and they way it is manufactured. I will probably stick with IBM or others, but there's never any real garauntee - the but the present Maxtor doesn't make be happy (it's a slim module and I suspect it may be along, if not actually, made for laptops - at least that was my impression with I really open the box and looked at the HD for the 1st time). HTH, Curtis On Saturday 04 January 2003 16:07, David Herman wrote:
Thanks to everyone for your efforts to help with my problem.
On Saturday 04 January 2003 12:56 pm, Curtis Rey wrote:
A udma cable has double the wires from the old/standard version. These are, to my understanding used to help with the signal. The extra wires are in between the data wires and are for the purpose of stopping signal leak/interferance between the wires (I think they are ground wires). Anyway, a udma66/100/133 cable can transmit the signal from a udma33 drive, it is backwards compatible, and of course a udma33 cable is incapable of utilizing the 66/100/133 Mb signal.
I still think that having the mixed drives will limit the spec to the lower transfer rate none the less.
Yes I agree with you, I remember reading that as well (now that you pointed it out). Seems that there are conflicting Ideas about what should work, but I agree that the drive speed isue is probably greater than the time needed to access swap... As far as I know my cables are UDMA100 capable but I may have to try changing them next. (soon I'll have more money in drive cables than I do in drives :-( )
I've read the numerous mails in this and the concurrent CD-R(W) threads but my problems continue.
Here's what I have tried.
Swapped my cdrw and dvd from their original positions, now I have: On the first IDE cable: 80gb IDE hard drive (udma100 capable) - linked to /dev/hda 80gb IDE hard drive (udma100 capable) - linked to /dev/hdb On the second IDE cable: ATAPI DVD-ROM - linked to /dev/hdc ATAPI CDRW - linked to /dev/sr1
Added to my lilo.conf: (as suggested by Joe Morris) append = "hdc=ide-cd hdd=ide-scsi"
/var/log/bootmsg shows: <4>Kernel command line: auto BOOT_IMAGE=win4lin ro root=345 hdc=ide-cd hdd=ide-scsi <4>ide_setup: hdc=ide-cd <4>ide_setup: hdd=ide-scsi ----------snip---------- <4>hda: ST380021A, ATA DISK drive <4>hdb: ST380021A, ATA DISK drive <4>hdc: HITACHI DVD-ROM GD-5000, ATAPI CD/DVD-ROM drive <4>hdd: LITE-ON LTR-48125W, ATAPI CD/DVD-ROM drive ------------snip----------- <4>VP_IDE: chipset revision 6 <4>VP_IDE: not 100%% native mode: will probe irqs later <4>VP_IDE: Unknown VIA SouthBridge, contact Vojtech Pavlik (my board is an msi KT3 Ultra2, via kt333 chipset) <4>hda: ST380021A, ATA DISK drive <4>hdb: ST380021A, ATA DISK drive <4>hdc: HITACHI DVD-ROM GD-5000, ATAPI CD/DVD-ROM drive <4>hdd: LITE-ON LTR-48125W, ATAPI CD/DVD-ROM drive <4>ide0 at 0x1f0-0x1f7,0x3f6 on irq 14 <4>ide0: probed IRQ 14 failed, using default. <4>ide1 at 0x170-0x177,0x376 on irq 15 <4>ide1: probed IRQ 15 failed, using default. (I don't know if these failures mean anything BUT it seems significant) (In a message from Carlos Robinson showing part of his boot.log DMA seemed to be set up at this point (snip follows) (<4>ICH2: not 100%% native mode: will probe irqs later <4> ide0: BM-DMA ... , BIOS settings: hda:DMA, hdb:DMA <4> ide1: BM-DMA at 0xf008-0xf00f, BIOS settings: hdc:DMA, hdd:DMA) ----------snip------------- <6>scsi1 : SCSI host adapter emulation for IDE ATAPI devices <4> Vendor: LITE-ON Model: LTR-48125W Rev: VS06 <4> Type: CD-ROM ANSI SCSI revision: 02 <7>sd_attach() ----------snip------------- Setting up IDE DMA mode
Force IDE DMA mode on: hda hdb done
So The drives are accessible but DMA mode (doesnt really seem to get set)
-- Billboard Writer vs. Literature = Microsoft vs. Computing,