Andre Hedrick wrote:
>
> On Tue, 28 Mar 2000, Alan Cox wrote:
>
> > > The OnStream DI-30 runs flawless in PIO or DMA mode.
> >
> > Then why are people seeing lost interrupts ?
> >
> > Right at the moment I suspect most of the lost interrupt stuff is PCI/APIC
> > stuff, certainly on 2.3.9x. You can't simply say 'it runs flawlessly' when
> > for someone it doesn't.
>
> If the chipset is tuned correctly and the drive is put in the correct
> transfer rate "it runs flawlessly". Even under PCI/APIC with a PIIX3
> 440FX PPro........this is with a broken MPTable, also........
>
> I can not make the came for VIA yet.........
>
> The question is how are they using the device?
>
> tar -cvvf /dev/ht0 -b 64 *
>
> Is the only way to call it with tar......you call the device with the
> wrong block size and the buffer blows out and would probably yield IRQ
> errors........
If I apply the latest IDE patches, my HP COLORADO 20GB refuses to restore
anything. It works fine during backup and booting to kernel without the
IDE patch allows the restore to work. The drive seems to work best in PIO
mode. If I set it to MWDMA2 it works but eventually something hickups and
ide resets, and it PIO time again. The error I see with failing restores
are:
Mar 26 08:33:44 oscar kernel: ide-tape: Reached idetape_chrdev_open
Mar 26 08:33:44 oscar kernel: ide-tape: ht0: I/O error, pc = 1e, key = 5, asc = 20, ascq = 0
Mar 26 08:33:44 oscar kernel: ide-tape: ht0: I/O error, pc = 8, key = 5, asc = 2c, ascq = 0
Mar 26 08:33:44 oscar kernel: ide-tape: ht0: I/O error, pc = 1e, key = 5, asc = 20, ascq = 0
There are also some odities in the ide-tape.c in that some of the items you
can tune without the patch no longer do anythings (setting max number of buffers).
seems that the patch sets the buffers depending on the drive speed. It does
not detect the speed correctly for the HP drive and defaults to 650 rather than
the 1000 or so it should use (the drives peaks higher than this).
Here is the drive info (this kernel has the default speed defaulting to 1100):
Mar 26 08:33:44 oscar kernel: ide-tape: Dumping ATAPI Identify Device tape parameters
Mar 26 08:33:44 oscar kernel: ide-tape: Protocol Type: <6>ATAPI
Mar 26 08:33:44 oscar kernel: ide-tape: Device Type: 1 - <6>Streaming Tape Device
Mar 26 08:33:44 oscar kernel: ide-tape: Removable: Yes
Mar 26 08:33:44 oscar kernel: ide-tape: Command Packet DRQ Type: <6>Interrupt DRQ
Mar 26 08:33:44 oscar kernel: ide-tape: Command Packet Size: <6>12 bytes
Mar 26 08:33:44 oscar kernel: ide-tape: Model: HP COLORADO 20GB
Mar 26 08:33:44 oscar kernel: ide-tape: Firmware Revision: 4.000026
Mar 26 08:33:44 oscar kernel: ide-tape: Serial Number: MX00577187
Mar 26 08:33:44 oscar kernel: ide-tape: Write buffer size: 0 bytes
Mar 26 08:33:44 oscar kernel: ide-tape: DMA: Yes
Mar 26 08:33:44 oscar kernel: ide-tape: LBA: Yes
Mar 26 08:33:44 oscar kernel: ide-tape: IORDY can be disabled: Yes
Mar 26 08:33:44 oscar kernel: ide-tape: IORDY supported: Yes
Mar 26 08:33:44 oscar kernel: ide-tape: ATAPI overlap supported: No
Mar 26 08:33:44 oscar kernel: ide-tape: PIO Cycle Timing Category: 2
Mar 26 08:33:44 oscar kernel: ide-tape: DMA Cycle Timing Category: 2
Mar 26 08:33:44 oscar kernel: ide-tape: Single Word DMA supported modes: <6>0 <6>1 <6>2 <6>
Mar 26 08:33:44 oscar kernel: ide-tape: Multi Word DMA supported modes: <6>0 <6>1 <6>2 <6>
Mar 26 08:33:44 oscar kernel: ide-tape: Enhanced PIO Modes: Mode 3
Mar 26 08:33:44 oscar kernel: ide-tape: Minimum Multi-word DMA cycle per word: <6>120 ns
Mar 26 08:33:44 oscar kernel: ide-tape: Manufacturer's Recommended Multi-word cycle: <6>120 ns
Mar 26 08:33:44 oscar kernel: ide-tape: Minimum PIO cycle without IORDY: <6>240 ns
Mar 26 08:33:44 oscar kernel: ide-tape: Minimum PIO cycle with IORDY: <6>120 ns
Mar 26 08:33:44 oscar kernel: ide-tape: hdd <-> ht0: HP COLORADO 20GB rev 4.00
Mar 26 08:33:44 oscar kernel: ide-tape: hdd: overriding capabilities->speed (assuming 1100KB/sec)
Mar 26 08:33:44 oscar kernel: ide-tape: hdd: overriding capabilities->max_speed (assuming
1100KB/sec)
Mar 26 08:33:44 oscar kernel: ide-tape: Dumping the results of the MODE SENSE packet command
Mar 26 08:33:44 oscar kernel: ide-tape: Mode Parameter Header:
Mar 26 08:33:44 oscar kernel: ide-tape: Mode Data Length - 23
Mar 26 08:33:44 oscar kernel: ide-tape: Medium Type - 133
Mar 26 08:33:44 oscar kernel: ide-tape: Device Specific Parameter - 16
Mar 26 08:33:44 oscar kernel: ide-tape: Block Descriptor Length - 0
Mar 26 08:33:44 oscar kernel: ide-tape: Capabilities and Mechanical Status Page:
Mar 26 08:33:44 oscar kernel: ide-tape: Page code - 42
Mar 26 08:33:44 oscar kernel: ide-tape: Page length - 18
Mar 26 08:33:44 oscar kernel: ide-tape: Read only - No
Mar 26 08:33:44 oscar kernel: ide-tape: Supports reverse space - Yes
Mar 26 08:33:44 oscar kernel: ide-tape: Supports erase initiated formatting - Yes
Mar 26 08:33:44 oscar kernel: ide-tape: Supports QFA two Partition format - Yes
Mar 26 08:33:44 oscar kernel: ide-tape: Supports locking the medium - No
Mar 26 08:33:44 oscar kernel: ide-tape: The volume is currently locked - No
Mar 26 08:33:44 oscar kernel: ide-tape: The device defaults in the prevent state - No
Mar 26 08:33:44 oscar kernel: ide-tape: Supports ejecting the medMar 26 08:33:44 oscar kernel:
ide-tape: Supports error correction - Yes
Mar 26 08:33:44 oscar kernel: ide-tape: Supports data compression - No
Mar 26 08:33:44 oscar kernel: ide-tape: Supports 512 bytes block size - Yes
Mar 26 08:33:44 oscar kernel: ide-tape: Supports 1024 bytes block size - No
Mar 26 08:33:44 oscar kernel: ide-tape: Supports 32768 bytes block size / Restricted byte count for
PIO transfers - No
Mar 26 08:33:44 oscar kernel: ide-tape: Maximum supported speed in KBps - 1100
Mar 26 08:33:44 oscar kernel: ide-tape: Continuous transfer limits in blocks - 64
Mar 26 08:33:44 oscar kernel: ide-tape: Current speed in KBps - 1100
Mar 26 08:33:44 oscar kernel: ide-tape: Buffer size - 442368
Mar 26 08:33:44 oscar kernel: ide-tape: hdd <-> ht0: 1100KBps, 13*32kB buffer, 10720kB pipeline,
50ms tDSC ium - No
If I can provide any more info please let me know.
Ed Tomlinson <tomlins@cam.org>
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Fri Mar 31 2000 - 21:00:26 EST