Messages in this thread | | | Subject | U2W SCSI problems | From | Turbo Fredriksson <> | Date | 30 Nov 1999 08:26:34 +0100 |
| |
[Please include me (turbo@nocrew.org) in any public replies]
I just bought a new SCSI card (AHA-2940U2W) to drive my two, new U2W SCSI disks that I bought over a couple of months. Previously I had the system on a 6.4Gb Narrow SCSI. Togheter with a Narrow CD-ROM, it was driven by a AHA-2940 card. With that setup I NEVER had a SCSI related problem, but now, with the new card/disks, I get SCSI hangs, constant SCSI resets etc...
I have tried a number of different kernels (2.3.5-acX, 2.3.23 and latest with the 2.2.14-pre9) and they all act the same.
I tried the stuff in the drivers/scsi/README.aic7xxx, but when booting with a rescue disk with the 2.2.14-pre9 kernel, I don't get those reset loops that I was expecting (read hoping :) for.
I have tried every different SCSI-BIOS option, disk jumper settings etc, but to no avail.
What happens is that if I boot with a properly shutdown system (ie, FS is clean), I get this:
----- Partition check: sda: sda1 sda2 < sda5 sda6 > sdb: sdb1 sdb2 sdb3 < sdb5 sdb6 sdb7 sdb8 sdb9 sdb10 > hdc: hdc1 <-- here I get about 10 sec delay --> SCSI : aborting command due to timeout : pid 60, scsi1, channel 0, id 0, lun 0 Read (6) 00 00 41 02 00 VFS: Mounted root (ext2 filesystem) readonly. -----
I _ALWAYS_ get the timeout, and the numbers are always the same (what are they by the way?). It usually comes online, but after a while (different timespan each time) it starts reseting the bus (looping).
If on the other hand I have been forced to cycle the power, it passes the hdc disk, then starts 'the SCSI reset loop', never comming online.
I've tried with only the U2W card, with one disk (which now have the system on it), the disk plugged in into different connectors on the cable, using the external LVD/SE terminator (at the very last connector of course :) or having the disk terminated (jumper setting, naturaly without the external terminator :).
I can not totaly rule out that it's the cables fault, but it's brand new (came with the previous U2W card, a Tekram, which I replaced because I thought that was the problem. I got the same thing with that card to).
To be able to give as much info as possible, I booted with the verbose and panic_on_abort kernel parameters, and have enclosed the output from that below.
This is my configuration: scsi0: AHA-2940 ID0: Quantum Fireball SE6.4S (Narrow disk, 6.4Gb) ID1: Toshiba XM-5701TA (CD-ROM) scsi1: AHA-2940U2W ID0: IBM DNES-309170D (Ultra2 LVD/SE disk, 9Gb) ID1: IBM DDRS-39130D (Ultra2 LVD/SE disk, 9Gb) ide1: Quantum Fireball EX10.2A (EIDE disk, 10Gb, /dev/hdc) OTI-SCYLLA (ATAPI-CDROM, /dev/hdd)
Booting from scsi0/0, with root on scsi1/0.
The reason for having the primary IDE channel disabled (in BIOS) is the boot priority problem, and the info on that is to be moved to the DDRS disk, totaly removing all IDE disks.
As stated above, I have tried to disconnect EVERYTHING (other expansion cards, IDE units etc) but the U2W card, and the DNES disk. The reason for two cards (the U2W card also have a Narrow channel), is that I suspected (read wanted to rule out) that it had problems on using both the Ultra2 LVD/SE channel AND the Narrow channel at the same time. Doesn't seem to be that problem, so I guess I don't need the Narrow card. No loss, could be sold :)
This is the output when booting with the verbose and panic_on_abort kernel options:
----- Partition check: sda: sda1 sda2 < sda5 sda6 > sdb: sdb1 sdb2 sdb3 < sdb5 sdb6 sdb7 sdb8 sdb9 sdb10 > hdc:<4>(scsi1:-1:-1:-1) Data Parity Error during PCI address or PCI write phase. aic7xxx driver version 5.1.21/2.2.14-pre9b Controller type: Adaptec AHA-294X Ultra2 SCSI host adapter p->flags=0x13804d01, p->chip=0x406, p->features=0x4f6, sequencer is paused Adaptec AHA-294X Ultra2 SCSI host adapter at PCI 12/0. PCI Dump: 04:07 05:00 06:90 07:02 08:00 0c:08 0d:40 0e:00 0f:80 10:01 11:74 12:00 13:00 14:04 15:10 16:00 17:e8 18:00 19:00 1a:00 1b:00 30:00 31:00 32:00 33:e7 34:dc 3c:09 3d:01 3e:27 3f:19 40:40 41:05 42:00 43:80 dc:01 dd:00 de:01 df:00 e0:00 e1:00 e2:00 e3:00 Card Dump: 00:12 01:00 02:37 03:00 04:00 05:00 08:00 09:00 0a:00 0b:00 0c:0a 0d:00 0e:00 0f:07 10:00 11:a4 18:87 19:00 1a:00 1b:00 1c:00 1d:fe 1e:97 1f:0b 60:10 62:09 63:00 64:00 65:00 66:e4 84:ab 85:00 86:00 87:06 88:00 89:00 8a:00 8b:00 8c:00 8d:00 8e:00 90:00 91:00 92:00 93:00 94:89 95:00 97:00 9a:00 9f:00 e0:6d e1:e4 e2:f9 e3:07 e4:00 e5:00 e6:00 e7:00 e8:00 e9:00 ea:81 eb:00 ec:00 ed:02 ee:00 ef:01 f0:00 f1:00 f4:3f f6:3f f8:15 fa:46 fb:77 fc:02 fe:02 ff:02 SEEPROM dump. 0x03bd 0x03bd 0x03bb 0x03bb 0x03bb 0x03bb 0x03bb 0x03bb 0x03bb 0x03bb 0x03bb 0x03bb 0x03bb 0x03bb 0x03bb 0x03bb 0x1886 ox1c5c 0x2807 0x0010 0xffff 0xffff 0xffff 0xffff 0xffff 0xffff 0xffff 0xffff 0xffff 0xffff 0xffff 0x98a2 Scratch RAM: 20:95 21:00 22:00 23:00 24:00 25:00 26:00 27:00 28:00 29:00 2a:00 2b:00 2c:00 2d:00 2e:00 2f:00 30:00 31:00 32:00 33:00 34:08 35:a8 36:00 37:00 38:00 39:00 3a:84 3b:f8 3c:07 3d:01 3e:ff 3f:ff 40:00 41:00 42:e0 43:f8 44:07 45:58 46:e3 47:f9 48:07 49:00 4a:00 4b:00 4c:00 4d:00 4e:3f 4f:00 50:00 51:00 52:01 53:80 54:7f 55:7f 56:a0 57:a2 58:a0 59:a0 5a:27 5b:07 5c:a0 5d:a0 5e:40 5f:a0 70:1f 71:00 72:00 73:00 74:00 75:00 76:00 77:00 78:00 79:00 7a:00 7b:00 7c:00 7d:00 7e:00 7f:00 -----
-- -- SDI supercomputer World Trade Center class struggle Uzi Ft. Bragg Panama DES cryptographic plutonium AK-47 Honduras arrangements North Korea smuggle
- 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/
| |