Promise ide controller in xp1000
Spam Me Please
I tried installing a promise (non-raid) PDC 20269 ATA 133 in a XP1000.
I got the following error messages:
ide: Assuming 66MHz system bus speed for PIO modes
PDC20269: IDE controller at PCI slot 02:09.0
PDC20269: chipset revision 2
PDC20269: not 100% native mode: will probe irqs later
PDC20269: ROM enabled at 0x209110000
ide0: BM-DMA at 0x200008400-0x200008407, BIOS settings: hda:pio, hdb:pio
ide1: BM-DMA at 0x200008408-0x20000840f, BIOS settings: hdc:pio, hdd:pio
hda: Maxtor 6Y160P0, ATA DISK drive
blk: queue fffffc000063c4f0, no I/O memory limit
hdc: Maxtor 6Y160P0, ATA DISK drive
blk: queue fffffc000063cc28, no I/O memory limit
ide0 at 0x200008410-0x200008417,0x200008422 on irq 28
ide1 at 0x200008418-0x20000841f,0x200008426 on irq 28
hda: attached ide-disk driver.
hda: host protected area => 1
hda: 320173056 sectors (163929 MB) w/7936KiB Cache, CHS=19929/255/63, UDMA(133)
hdc: attached ide-disk driver.
hdc: host protected area => 1
hdc: 320173056 sectors (163929 MB) w/7936KiB Cache, CHS=19929/255/63, UDMA(133)
Partition check:
/dev/ide/host0/bus0/target0/lun0:<4>hda: dma_timer_expiry: dma status == 0x21
hda: error waiting for DMA
hda: error waiting for DMA
hda: dma timeout retry: status=0x51 { DriveReady SeekComplete Error }
hda: dma timeout retry: error=0x84 { DriveStatusError BadCRC }
blk: queue fffffc000063c4f0, no I/O memory limit
p1
/dev/ide/host0/bus1/target0/lun0:<4>hdc: dma_timer_expiry: dma status == 0x21
hdc: error waiting for DMA
hdc: dma timeout retry: status=0x51 { DriveReady SeekComplete Error }
hdc: dma timeout retry: error=0x84 { DriveStatusError BadCRC }
blk: queue fffffc000063cc28, no I/O memory limit
p1
*****
Has anyone seen this error and are there an solutions?
Thanks,
Alan
Måns Rullgård
> I tried installing a promise (non-raid) PDC 20269 ATA 133 in a XP1000.
> I got the following error messages:
>
> hdc: 320173056 sectors (163929 MB) w/7936KiB Cache, CHS=19929/255/63, UDMA(133)
> Partition check:
> /dev/ide/host0/bus0/target0/lun0:<4>hda: dma_timer_expiry: dma status == 0x21
> hda: error waiting for DMA
> hda: error waiting for DMA
> hda: dma timeout retry: status=0x51 { DriveReady SeekComplete Error }
> hda: dma timeout retry: error=0x84 { DriveStatusError BadCRC }
>
> Has anyone seen this error and are there an solutions?
I've seen the similar things with a 20268 in a Miata and SX164. Make
sure your cables are no longer than 4 inches, or use shielded cables.
I suspect that there is too much noise generated by something in alpha
systems. Serial ATA works just fine, though.
--
Måns Rullgård
m...@users.sf.net
Spam Me Please
Any recommended ata serial controller? I want to avoid this problem
again if possible.
Thanks
>>>>> "Måns" == Måns Rullgård <m...@users.sourceforge.net> writes:
Måns> Spam Me Please <spa...@edge.net> writes:
>> I tried installing a promise (non-raid) PDC 20269 ATA 133 in a
>> XP1000. I got the following error messages:
>>
>> hdc: 320173056 sectors (163929 MB) w/7936KiB Cache,
>> CHS=19929/255/63, UDMA(133) Partition check:
>> /dev/ide/host0/bus0/target0/lun0:<4>hda: dma_timer_expiry: dma
>> status == 0x21 hda: error waiting for DMA hda: error waiting for
>> DMA hda: dma timeout retry: status=0x51 { DriveReady SeekComplete
>> Error } hda: dma timeout retry: error=0x84 { DriveStatusError
>> BadCRC }
>>
>> Has anyone seen this error and are there an solutions?
Måns> I've seen the similar things with a 20268 in a Miata and SX164.
Måns> Make sure your cables are no longer than 4 inches, or use
Måns> shielded cables. I suspect that there is too much noise
Måns> generated by something in alpha systems. Serial ATA works just
Måns> fine, though.
Måns> -- Måns Rullgård m...@users.sf.net
Måns Rullgård
> Any recommended ata serial controller? I want to avoid this problem
> again if possible.
I'm using a Highpoint Rocketraid 1540. I'd stay away from the Promise
ones.
--
Måns Rullgård
m...@users.sf.net
Spam Me Please
Thanks!
>>>>> "Måns" == Måns Rullgård <m...@users.sourceforge.net> writes:
Måns> Spam Me Please <spa...@edge.net> writes:
>> Any recommended ata serial controller? I want to avoid this
>> problem again if possible.
Måns> I'm using a Highpoint Rocketraid 1540. I'd stay away from the
Måns> Promise ones.
Alan P. Kennedy
Thanks for the help on this matter. The store took the promise
controller back, but refused to take the harddrives back because I
turned in the rebate. That was fair. Anyway I was stuck with two 160g
drives but nothing to put them in.
I tried the drives in an amd computer as a raid setup and the old via
kt133 based motherboard had massive data corruption and the system
crashed. This is a known problem with old via motherboards, but gave
it a try anyway. For the record the built-in ide controller on the
motherboard was a promise controller also. I suspect with another
motherboard things would have worked fine, but I did not want to buy
another motherboard.
I ordered a cheap silicon image raid controller on ebay, and installed
the card in the alpha xp1000.
lspci -v shows this
02:09.0 RAID bus controller: CMD Technology Inc PCI0680 (rev 02)
Subsystem: CMD Technology Inc: Unknown device 3680
Flags: bus master, medium devsel, latency 32, IRQ 28
On boot the following is displayed:
ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
SiI680: IDE controller at PCI slot 02:09.0
SiI680: chipset revision 2
SiI680: not 100% native mode: will probe irqs later
SiI680: BASE CLOCK == 133
ide0: MMIO-DMA , BIOS settings: hda:pio, hdb:pio
ide1: MMIO-DMA , BIOS settings: hdc:pio, hdd:pio
hda: Maxtor 6Y160P0, ATA DISK drive
blk: queue fffffc0000654630, no I/O memory limit
hdc: Maxtor 6Y160P0, ATA DISK drive
blk: queue fffffc0000654d68, no I/O memory limit
ide0 at 0xfffffd0209001080-0xfffffd0209001087,0xfffffd020900108a on irq 28
ide1 at 0xfffffd02090010c0-0xfffffd02090010c7,0xfffffd02090010ca on irq 28
hda: attached ide-disk driver.
hda: host protected area => 1
hda: 320173056 sectors (163929 MB) w/7936KiB Cache, CHS=19929/255/63, UDMA(133)
hdc: attached ide-disk driver.
hdc: host protected area => 1
hdc: 320173056 sectors (163929 MB) w/7936KiB Cache, CHS=19929/255/63, UDMA(133)
Partition check:
/dev/ide/host0/bus0/target0/lun0: p1
/dev/ide/host0/bus1/target0/lun0: p1
I setup up the two 160g as raid 1, and at first it didn't appear to
work. I had some setup problems, but after getting the setup correct
everything is working like a charm for the last three days.
cat /proc/mdstat
Personalities : [raid0] [raid1] [raid5]
read_ahead 1024 sectors
md0 : active raid1 ide/host0/bus1/target0/lun0/part1[1] ide/host0/bus0/target0/lun0/part1[0]
160079552 blocks [2/2] [UU]
Setup LVM on top of the raid to partition the drive. The old scsi
drive is used to boot the system.
Hope this helps someone that wants a cheap big ide raid in a xp1000.
I guess this means that the promise controller does not work well with
alphas at least the one I had. Silicon image raid controller works
prefectly. Note one cannot use the built in raid on the card. I'm
using the raid in the linux kernel.
If it crashes and burns I will post an update, but so far so good.
The nasty dma timeout errors are completely gone!
Later
>>>>> "Måns" == Måns Rullgård <m...@users.sourceforge.net> writes:
Måns> Spam Me Please <spa...@edge.net> writes:
>> I tried installing a promise (non-raid) PDC 20269 ATA 133 in a
>> XP1000. I got the following error messages:
>>
>> hdc: 320173056 sectors (163929 MB) w/7936KiB Cache,
>> CHS=19929/255/63, UDMA(133) Partition check:
>> /dev/ide/host0/bus0/target0/lun0:<4>hda: dma_timer_expiry: dma
>> status == 0x21 hda: error waiting for DMA hda: error waiting for
>> DMA hda: dma timeout retry: status=0x51 { DriveReady SeekComplete
>> Error } hda: dma timeout retry: error=0x84 { DriveStatusError
>> BadCRC }
>>
>> Has anyone seen this error and are there an solutions?
Måns> I've seen the similar things with a 20268 in a Miata and SX164.
Måns> Make sure your cables are no longer than 4 inches, or use
Måns> shielded cables. I suspect that there is too much noise
Måns> generated by something in alpha systems. Serial ATA works just
Måns> fine, though.