Maybe I'll post a patch, but the OS's it crashes on, has mmap() defined as a function, which collides with the global mmap in 32v..
Hey, I saw a few people actually download this.. Ive updated the source to be slightly more build friendly. I may try to update this again. If there is anyone interested post in the forums, good, bad or indefferent!
My foot is killing me, so I didn't go running in the morning.. instead I ripped apart the RK disk driver from the pdp11 version, and it's starting to make some sense.. Anyways, I am now starting to see how the IO goes. I have to find a slightly faster machine to do pdp11 emulation on to test my ideas though... I think I will need to chase down UNIX_mbainit() and see why it freaks out..
ok, so I'm dead wrong on how it works. I've just dl'd the interdata port of v7, and simh, and I'm trying to remember how to use ed to look at the strategy part of the block drivers.. (I dont see why its first call has the active int with the value of 512, not 0). Anyways I hope to post a new version in a couple of days.
I fleshed out some more of the missing.c stuff. Now I'm trying to verify the filesystem. It does look screwed up though...
I found some docs on the Device Driver thing. so this is a 1st pass at a block driver.. does it work? I doubt it very much, but I'll post what I have.
Ok, it compiles using Open Watcom (The debugger is cool, with Watcom, because you can go *BACKWARDS*) Visual C++ (using C, I haven't tried managed C++ yet!) and GCC (cygwin, Linux & OSX).
This release sucks, it doesnt work correctly there are a buch of important things removed to make it 'work' but it goes into the scheduler! (I really need block drivers now.. I need to figure out the idea of strategy devices!)