From mboxrd@z Thu Jan  1 00:00:00 1970
From: lyndon@orthanc.ca (Lyndon Nerenberg)
Date: Sun, 19 Nov 2017 14:38:53 -0800 (PST)
Subject: [TUHS] Determining what was on a tape back in the day
In-Reply-To: <alpine.BSF.2.21.1711200854090.780@aneurin.horsfall.org>
References: <20171119134109.59F3018C0F5@mercury.lcs.mit.edu>
 <EC9F5ECB-D02A-435B-8AB6-93F484908F99@ccc.com>
 <7AA20D14-571A-4AAD-96D0-45057F5B81DD@jctaylor.com>
 <CA4CC953-EA9D-48CB-95AE-881FC1577466@ronnatalie.com>
 <alpine.BSF.2.21.1711200854090.780@aneurin.horsfall.org>
Message-ID: <alpine.BSF.2.21.1711191434190.6984@vm.orthanc.ca>

> The early Sun boxes used to hang when rewinding the cartridge tape, because 
> the controller didn't release the Multibus; the stingy Lionel Singer used to 
> sell the 3/50 as cheap servers...

The Convergent Tech MiniFrame had a similar bug with the QIC tape drive. 
If you sent it an interrupt during a rewind (e.g. ^C a cpio after the 
extract was done, but before the rewind/close finished), it would block 
the tape device from any further use.

That was my first expedition into non-API UNIX kernel hacking, writing a 
program that would reach into the kernel to clear the 'busy' bit on the 
tape device.

--lyndon