Date: Tue, 6 May 1997 09:42:52 +0200 From: j@uriah.heep.sax.de (J Wunsch) To: scsi@FreeBSD.ORG, hardware@FreeBSD.ORG Cc: gavin@ormond.unimelb.edu.au (Gavin Cameron) Subject: Re: Anyone using a CONNER CTT8000-S tape drive Message-ID: <19970506094252.DE24694@uriah.heep.sax.de> In-Reply-To: <199705060115.LAA20543@gateway.ormond.unimelb.edu.au>; from Gavin Cameron on May 6, 1997 11:15:35 %2B1000 References: <336E8151.19A13460@whistle.com> <199705060115.LAA20543@gateway.ormond.unimelb.edu.au>
next in thread | previous in thread | raw e-mail | index | archive | help
As Gavin Cameron wrote: > May 6 11:12:31 server /kernel: st0(ahc0:4:0): command: 5,0,0,0,0,0-[6 bytes] > May 6 11:12:31 server /kernel: ------------------------------ > May 6 11:12:31 server /kernel: 000: 00 00 00 00 00 00 > May 6 11:12:31 server /kernel: ------------------------------ > May 6 11:12:31 server /kernel: code70 valid0 seg0 key5 ili0 eom0 fmark0 > May 6 11:12:31 server /kernel: info: 0 0 0 0 followed by 10 extra bytes > May 6 11:12:31 server /kernel: extra: 0 ff d4 dd 42 0 0 0 0 0 > May 6 11:12:31 server /kernel: st0(ahc0:4:0): ILLEGAL REQUEST asc:24,0 Invalid field in CDB Julian, Justin? Wouldn't it be possible to silence this message? > May 6 11:12:31 server /kernel: st0(ahc0:4:0): command: 10,0,0,0,1,0-[0 bytes] > May 6 11:12:31 server /kernel: code70 valid0 seg0 key5 ili0 eom0 fmark0 > May 6 11:12:31 server /kernel: info: 0 0 0 0 followed by 10 extra bytes > May 6 11:12:31 server /kernel: extra: 0 ff b4 dd 42 0 0 0 0 0 That's a WRITE FILEMARKS command, requesting to write just one filemark. I can't seem to find what would be wrong with it. Maybe somebody can decipher the `extra' information and see which field in CDB they are complaining about. > PS Should this still be going to both scsi and hardware? Yes. As a warning for others to not buy that drive. :-/ Sorry to beat at you here, i couldn't resist. Conner is starting to producing the same crap as ATA (IDE) vendors are already so proud of. Either command 0x05 (READ BLOCK LIMITS) as well as 0x10 (WRITE FILEMARKS) are _mandatory_ for sequential-access devices, and the SCSI standard is fairly specific about the meaning of mandatory: the drive you've purchased might have a connector that makes it look like SCSI, but it's actually a forgery. It's not a SCSI device. If i were you, i would probably return it. The CDBs quoted above are not obviously wrong (you can't do much wrong with them at all, given that the CDB for command 0x10 just takes 5 0's as arguments, and the CDB for cmd 0x05 basically takes a `count' parameter). Sure, we could add yet another set of quirks for such i drive, and this will probably help you -- but mentally, i refuse to support vendors of non-SCSI devices. There is a standard, and goddam!, the vendors have to adhere to it, or they gotta stop advertising their crap as being SCSI. Sorry for the bitching. It isn't meant to be personally, of course. -- cheers, J"org joerg_wunsch@uriah.heep.sax.de -- http://www.sax.de/~joerg/ -- NIC: JW11-RIPE Never trust an operating system you don't have sources for. ;-)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19970506094252.DE24694>