rick@seismo.UUCP (03/05/87)
I have a Vax 11/780 running 4.3BSD with an Emulex TC13 controller (emulating a ts-11) connected to a Fuji-2444AC tape drive. Everything works fine except that I can not set the density on the tape drive from software. Emulex swears that if the driver sets the right bit in the characteristics packet, that the contoller will set the density on the drive. (It's always a driver problem to them.) A scope on the density select pin out of the controller shows that the line never goes high to indicate hi density. A printf in the ts driver (obviously modified for dual density) shows that the correct bit is being written into the controller. Has anyone tried to do a software density select with an Emulex TC13? I'm out of things to try. ---rick
ed@mtxinu.UUCP (Ed Gould) (03/06/87)
In article <43125@beno.seismo.CSS.GOV> rick@seismo.CSS.GOV (Rick Adams) writes: >I have a Vax 11/780 running 4.3BSD with an Emulex TC13 controller >(emulating a ts-11) connected to a Fuji-2444AC tape drive. > >Everything works fine except that I can not set the density on the >tape drive from software. Some years ago, I did a consulting job for Emulex to implement software density selection in the ts driver for a *modified* TC13 (they had a customer who wanted it). It worked, but did require special ROMs in the TC13. There was some problem with version mismatch when I sent them the driver initially - it didn't test correctly on the hardware they had in house, although it did work on my development system. After an hour or so of demonstrating the problem at their facility, the problem turned out to be with the particular transport they had - it worked correctly with the Cipher I used for debugging. Beyond that, I don't remember the details of the problem. Since that time, Emulex has been massively reorganized, and I've found them *very* unresponsive to Unix-related problems. -- Ed Gould mt Xinu, 2560 Ninth St., Berkeley, CA 94710 USA {ucbvax,decvax}!mtxinu!ed +1 415 644 0146 "A man of quality is not threatened by a woman of equality."
swa@COMET.LCS.MIT.EDU (Steven Augart) (03/06/87)
I don't have a good answer for this one, but if you'd let me know what happens with it, I'd appreciate finding out. Thanks, SWA
mangler@cit-vax.UUCP (03/12/87)
In article <43125@beno.seismo.CSS.GOV>, rick@seismo.CSS.GOV (Rick Adams) writes: > A scope on the density select pin out of the controller shows that the > line never goes high to indicate hi density. Some streamers, such as the CDC 92185, abuse the density select line to select high speed or low. The TC13 caters to this, asserting the line when the drive seems to be streaming well. (You didn't see it asserted because the program wasn't fast enough to stream the tape, probably). To use that line as a density select, you have to enable the "Remote Density Select" DIP switch (SW3-8), and I think you also have to set the transport type switch (SW1-4) to "Formatted" instead of "Streaming". When you try this with a CDC 92185 (see above), any time you try to select high density, the drive will go into high-speed streaming mode, rocking the tape back and forth. I never did figure out how to get the 92185 to interpret that signal as a density select instead of a speed select. (If anyone does, let me know...) Oh yes, the manual says this wasn't implemented until Rev. C; but Rev. B didn't work at all with the 4.2 BSD distribution tape, so I assume that's not your problem. (Only the low 8 bits of the cmd/msg buffer pointer would increment, and the generic kernel put it right across a page boundary. I gave Emulex a listing of VAX console examine/deposit commands to demonstrate it - if you want a hardware vendor to fix something, that really grabs them). Don Speck speck@vlsi.caltech.edu {seismo,rutgers,ames}!cit-vax!speck
rick@seismo.UUCP (03/14/87)
I finally got our TC13/Fuji244 to do remote density select. The major problem was that the TC13 ONLY looks at the density select bit IMMEDIATELY after the controller has been reset (e.g. with addr->ts_ssr =0). Once you figure this out (its not documented and Emulex doesn't tell you about it) "all" you have to do is change the ts11 driver to do a controller reset when setting the characteristics. Diffs for this will be in comp.bugs.4bsd in a few days (as soon as I verify that I havent' broken the driver on a real ts11) --rick