Re: WD 93c33 trouble
Michael <[email protected]>
| Newsgroups | gmane.os.netbsd.ports.sgimips,gmane.os.netbsd.devel.kernel |
|---|---|
| Message-ID | <[email protected]> |
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello, On Feb 17, 2009, at 6:50 AM, Izumi Tsutsui wrote: > [email protected] wrote: > >> The problem I'm seeing is this - the wdsc driver deadlocks itself >> when >> probing devices right after probing the first one. I ran into this >> when trying to put a 2nd disk into my Indy - ARCS reports both >> without >> trouble but NetBSD hangs after probing the first one. There's nothing >> wrong with the 2nd disk - if I disconnect the other one and netboot >> the machine NetBSD probes it just fine. Also, if I swap SCSI IDs so >> the new disk is probed first it's still hanging after probing the >> first device, so it's neither of the disks. >> >> Is anyone else seeing this? > > Works for me: ... > sd0 at scsibus0 target 1 lun 0: <IBM, DCAS-32160, S65A> disk fixed > sd0: 2063 MB, 8188 cyl, 3 head, 172 sec, 512 bytes/sect x 4226725 > sectors > sd0: sync (100.00ns offset 12), 8-bit (10.000MB/s) transfers, tagged > queueing > sd1 at scsibus0 target 2 lun 0: <MELCO, DSC-UGTV, 2.03> disk fixed > sd1: 76319 MB, 16383 cyl, 16 head, 596 sec, 512 bytes/sect x > 156301488 sectors > sd1: sync (100.00ns offset 12), 8-bit (10.000MB/s) transfers > cd0 at scsibus0 target 5 lun 0: <TEAC, CD-ROM CD-516S, 1.0D> cdrom > removable > cd0: sync (100.00ns offset 8), 8-bit (10.000MB/s) transfers I get this: scsibus0: waiting 2 seconds for devices to settle... wd33c93_intr: LCI asr:c0 csr:1f sd0 at scsibus0 target 1 lun 0: <SGI, IBM DORS-32160, WA6A> disk fixed sd0: 2049 MB, 6703 cyl, 5 head, 125 sec, 512 bytes/sect x 4197405 sectors sd0: sync (100.00ns offset 12), 8-bit (10.000MB/s) transfers, tagged queueing boot device: sd0 'LCI' as in 'Last Command Ignored' according to the wd93c33 header. And I have a ZIP drive with termination enabled on the external SCSI port. Removing it doesn't change a thing. > Note ID 0 is occupied by the host controller on SGI machines. Yeah, I found that out when setting the 2nd drive's ID to 0 - ARCS would complain loudly. have fun Michael -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (Darwin) iQEVAwUBSZruQspnzkX8Yg2nAQIO/wf+LqEZgfVJTZ9IIkpYEgQykQr6OqsUtQl8 Gh5rs0qsoQQLDo8w9kYt6oMEZTwkYhJMM4FNArKaqMorXkdHyeLjy4dQlJMsoxYs HbCTzQzji6/JDzhvc2ZDT84SQq92q+QjJwzCQk6rrOW91f4Uex5yCEZlVPUTZA0D UaTZRf1+27GgWowJ4sASqpi82tN806lq2ToY6cY89e/AmvOqOutO9FrqmRK/6zF6 GIE9SQ42QoQxoC4tgi4tpMEX3fw1FeUycLub0fMwU6LAu3ydoZP5VS8VLzh6uLIY dj/JFVDUZIUrnsroI4VnMI27d/GoHH5yW4dGdIaOYg7eg6QvEBp7xg== =RsoC -----END PGP SIGNATURE-----