I sent this to current-users 3 months ago, and I haven't seen any response.
I still see the problem with a kernel built from cvs today.  Is it just my
hardware, or is this really broken?

on 5/23/02 6:39 PM, Monroe Williams at monroe@pobox.com wrote:
> 
> I have this hardware:
> 
> pciide0 at pci0 dev 13 function 0: Acard ATP860-A Ultra66 IDE Controller
> (rev. 0x01)
> pciide0: bus-master DMA support present
> pciide0: primary channel wired to native-PCI mode
> pciide0: using irq 23 for native-PCI interrupt
> wd0 at pciide0 channel 0 drive 0: <WDC WD205BA>
> wd0: drive supports 16-sector PIO transfers, LBA addressing
> wd0: 19574 MB, 16383 cyl, 16 head, 63 sec, 512 bytes/sect x 40088160 sectors
> wd0: drive supports PIO mode 4, DMA mode 2, Ultra-DMA mode 4 (Ultra/66)
> wd0(pciide0:0:0): using PIO mode 4, DMA mode 2, Ultra-DMA mode 4 (Ultra/66)
> (using DMA data transfers)
> pciide0: secondary channel wired to native-PCI mode
> wd1 at pciide0 channel 1 drive 0: <WDC WD307AA>
> wd1: drive supports 16-sector PIO transfers, LBA addressing
> wd1: 29333 MB, 16383 cyl, 16 head, 63 sec, 512 bytes/sect x 60074784 sectors
> wd1: drive supports PIO mode 4, DMA mode 2, Ultra-DMA mode 4 (Ultra/66)
> wd1(pciide0:1:0): using PIO mode 4, DMA mode 2, Ultra-DMA mode 4 (Ultra/66)
> (using DMA data transfers)
> 
> in a macppc system running a relatively recent -current kernel (circa May
> 16).  (In case you can't immediately decode the dmesg output, that's a
> Sonnet Tempo ATA-66 Mac-compatible card with one drive on each channel.)
> 
> I was doing a cvs update of the netbsd 'src' module to one of the two
> drives, and I decided to do a simple throughput test ('cat /dev/zero >foo')
> on the other one at the same time.  The machine seemed to pause for a bit (I
> was accessing it remotely), and when it came back I had a couple dozen of
> these in my logs:
> 
> May 23 17:41:45 macppc /netbsd: pciide0:0:0: lost interrupt
> May 23 17:41:45 macppc /netbsd:         type: ata tc_bcount: 8192 tc_skip: 0
> May 23 17:41:45 macppc /netbsd: pciide0:1:0: lost interrupt
> May 23 17:41:45 macppc /netbsd:         type: ata tc_bcount: 65536 tc_skip:
> 0
> May 23 17:41:55 macppc /netbsd: pciide0:0:0: lost interrupt
> May 23 17:41:55 macppc /netbsd:         type: ata tc_bcount: 8192 tc_skip: 0
> May 23 17:41:55 macppc /netbsd: pciide0:1:0: lost interrupt
> May 23 17:41:55 macppc /netbsd:         type: ata tc_bcount: 65536 tc_skip:
> 0
> 
> The errors weren't fatal; the machine's still up.  I've unmounted all
> filesystems residing on one of the two drives for my own peace of mind.
> (The drive is new and has nothing on it yet.)
> 
> Any ideas what's going on here?  Is anyone out there using both channels of
> this model adapter under heavy load?

-- monroe
------------------------------------------------------------------------
Monroe Williams                                         monroe@pobox.com