Ken Nakata wrote...

>With the diffs Allen gave to me, I've built a new Xserver which works
>fine with SE/30.  The cause of the problem appears to be grfmmap()
>routine failing to map SE/30's frame buffer correctly.  I changed
>mmap() back to ioctl() and the problem went away.  I'm not sure as to
>why mmap() doesn't work with SE/30's internal video, but I've been
>investigating.  It looks like mmap() maps the declaration ROM instead
>of the frame buffer into process' virtual address space.
>
>Anyway, the working server binary is placed at:
>
>ftp://puma.bevd.blacksburg.va.us/private/kenn/Xmacbsd-960127.tar.gz
>
>Untar it in /usr/X11R6/bin directory.  Don't forget to mv your old
>server to, for instance, Xmacbsd.old before untarring the new one,
>just in case it doesn't work for you.
>
>Note: the server file may be moved to a different place later.
>
>Have fun!
>
>ken

It does not work for me...  Xmacbsd-960127 still looks for /dev/grf2 :-(
My machine is pretty similar to yours, Ken: SE/30 20 MB RAM, Micron XCEED
30 with 15" portrait monitor (640x870).

Your remark about mmap() is interesting. Some weeks ago I tried to get mSQL
to work (a tiny SQL engine). Though the doc claimed that mSQLs use of
mmap() works on *BSD systems it dumped core after some inserts. After I
disabled the mmap() option the test suites ran fine. --  ???



	hauke

---
"It's never straight up and down"     (DEVO)