On Fri, Nov 23, 2001 at 05:53:32PM +0100, Reinoud Zandijk wrote:
> Hiya All,
> 
> on my PWS 500au i'm having the following problem and i just can't seem to 
> find what is causing this ... I added a 40 Gb harddisc as master on pciide0 
> and it got detected as :
> 
> pciide0 at pci0 dev 7 function 1: Cypress 82C693 IDE Controller (rev. 0x00)
> pciide0: bus-master DMA support present
> pciide0: primary channel wired to compatibility mode
> wd0 at pciide0 channel 0 drive 0: <\^P\M^?\^P?>
> wd0: drive supports 254-sector PIO transfers, chs addressing
> wd0: 16383 PB, 4096 cyl, 4159 head, 21569 sec, 512 bytes/sect x -1933578240 
> sectors
> wd0: 32-bit data port
> pciide0: primary channel interrupting at isa irq 14
> wd0(pciide0:0:0): using PIO mode 0
> 
> now i dont know how or what but this doesn't look good at all... esp. since 
> it does mention it is interrupting at isa irq 14 but when i run `atactl wd0 
> ident' it gives a lost interrupt on pciide0:0:0 and a stray irq 14 !
> 
> stray isa irq 14
> pciide0:0:0: lost interrupt
>         type: ata tc_bcount: 512 tc_skip: 0
> pciide0:0:0: geometry timed out
> wd0c: device timeout reading fsbn 0 (wd0 bn 0; cn 0 tn 0 sn 0), retrying
> stray isa irq 14
> .....
> 
> now that can be that the irq 14 doesn't come in and thus the kernel reads 
> garbage when detecting it .... i am puzzled. The firmware does detect it as 
> dba0.0.0.127.0 or so ... but prints a garbage ID string.
> 
> could it be that the isa irq 14 is not being handed over to the pciide?

No, I don't think it's the problem. interrupts are not used to probe drives.

I think there is a problem on the bus, and some lines a pulled up, so that
some bits always appears a '1' in the data register.
The "254-sector PIO transfers" doesn't looks good either; it's usually
16 or 32.
Check your harware, especially the cable and connectors.

--
Manuel Bouyer <bouyer@antioche.eu.org>
--