home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.sys.amiga.hardware:15135 comp.sys.amiga.misc:13708
- Path: sparky!uunet!europa.asd.contel.com!darwin.sura.net!spool.mu.edu!sol.ctr.columbia.edu!ira.uka.de!fhg!igd!john
- From: john@igd.fhg.de (Werner John)
- Newsgroups: comp.sys.amiga.hardware,comp.sys.amiga.misc
- Subject: A2090 and large drives : SOLVED !!
- Keywords: A2090 HD Drive large
- Message-ID: <1497@igd.fhg.de>
- Date: 10 Sep 92 14:21:11 GMT
- Sender: news@igd.fhg.de
- Followup-To: comp.sys.amiga.hardware
- Organization: IGD FHG Darmstadt, FRG
- Lines: 49
-
- Hello everyone,
-
- I've received many relpies with some nice ideas. The problem is solved.
- I'd like to thank Michael L. Hitch (cheers) from the Montana State
- University Bozeman, MT USA. He send me the solution and an explanation
- for it: (in short)
- The driver maintains 8 8-block buffers. When a read or write request for a
- single block is received, it checks to see if the block is in one of the
- buffers. It does a DIVU on the block number followed by a MULU to get the
- starting block number of the 8-block segment. With more than 0x80000 blocks,
- DIVU fails, because the quotient is more than 16 bits.
-
- Michael sent me a patch, I patched hddisk and now it works :-) Thanks
- to you, Michael.
-
- So, here are some other hints, that might be useful to other A2090 users.
- o Disable parity checking on your drive. Although it doesn't make
- a difference with my A2090. I had (and still have) parity enabled
- on my ST157N, ST1096N, LPS425 and it works. Perhaps parity is
- simply ignored.
- o Buy a new controller. New controllers have autoboot, FFS, and
- are faster. Currently, the transfer rate I got with my A2000
- rev. 4.4, Kick+WB 2.04, A2630 4MB, SetCPU FASTROM HEAD CACHE is
- somewhere between 1.0 and 1.1 MB per second with FFS. I used
- DiskPerf and DiskSpeed for testing.
- What about other controllers and system configurations. How fast
- are you?
-
- The maximum nuber of heads supported with the A2090 and hddisk is
- 32 (surfaces), 32768 Sectors and (2^^31)-1 Cylinders (or was it vice
- versa? I have to check). Of cause, the total number of blocks must not
- exceed the maximum 32-bit integer value (about 2,000,000,000 with signed
- integers) Anyway, it is large enough for the Amiga.
-
- By the way, I've formatted my primary partition with FFS too !! It
- works with my ST1096 since I got 2.04 and thats almost 8 month now.
- You have explicitly give the FFS option to system:format and you'll
- get 512 byte blocks. BUT: the buffers will be in CHIP-MEM, but this
- doesn't make much difference with the A2090.
-
- You got more questions, helps, ... I'll be with you.
-
- Thanks to all who replied, yours
- Johnny
-
- Werner John | Phone: ++49 6102 8750
- Zeppelinstr. 68 | Fax: ++49 69 885555
- Neu Isenburg W-6078 | I-net: john@igd.fhg.de
- West Germany |
-