dsb@melpar.UUCP (Don Berman) (03/14/91)
I am attempting to port the MIT X window system X11R4 to the Motorola MVME 147SA-1 cpu running UNIX System V/68 R3V5 and I'm not getting very far. I made changes to imakemdep.h, imake.tmpl, and created an r3v5.cf file. When trying to make World (and setting BOOTSTRAPCFLAGS=-Dr3v5) I get a failure when trying to make the Makefile for /lib/Xau. The error indicates an invalid syntax for an include file. Help! Has anyone tried to do this port? If so, were you successful? And if successful, please tell me how you did this? Also, I plan to port MOTIF. Any help there? Thanks, Don Berman (uunet!melpar!dsb)
gildea@expo.lcs.mit.EDU (Stephen Gildea) (03/14/91)
When trying to make World (and setting BOOTSTRAPCFLAGS=-Dr3v5) I
get a failure when trying to make the Makefile for /lib/Xau. The error
indicates an invalid syntax for an include file. Help!
Remove the three lines from the top of the Imakefile that
deal with BandAidCompiler.
< Stephen
MIT X Consortiumrandy@erik.UUCP (Randy Brown) (03/15/91)
From: uunet!melpar!dsb (Don Berman) > I am attempting to port the MIT X window system X11R4 to the Motorola > MVME 147SA-1 cpu running UNIX System V/68 R3V5 and I'm not getting very > far. I made changes to imakemdep.h, imake.tmpl, and created an r3v5.cf > file. When trying to make World (and setting BOOTSTRAPCFLAGS=-Dr3v5) I > get a failure when trying to make the Makefile for /lib/Xau. The error > indicates an invalid syntax for an include file. Help! Has anyone tried > to do this port? If so, were you successful? And if successful, please > tell me how you did this? Also, I plan to port MOTIF. Any help there? > Thanks, Don Berman (uunet!melpar!dsb) I have done precisely this port (client side only, except for xmh [no mh, after all] and xterm). It was done on patch level 14, but I don't think any patches since have affected it. The base system I chose was the AT&T SystemV/386 Release 3.2 stuff in the standard release, which I first tweaked to actually make and run on the 386 (again, client-side only, xterm included). I used -DSTREAMSCONN, except that I used Streams pipes (which V/68 R3V5 has) instead of Streams pseudo-ttys. The lack of Streams pseudo-ttys is what held up the xterm port, since the AT&T code depends on them; it would be easy enough to go in and change the lines in xterm that say "#ifdef hpux" to "#if defined(hpux) || defined(MOTOROLA)", which is what I did in other places where AT&T's code either didn't work on Motorola or had never reached (such as xdm). I have sent you my patch file separately, not echoing it to xpert, since it's about 1000 lines. I would appreciate any feedback, especially an xterm patch or bug fixes. Certainly if you find the patch incomplete, let me know--I think I'm sending you the correct file, but it's been a while. If anyone else would like this, please write uunet!erik!rbrown, but notice again, this is CLIENT ONLY. Don't ask me for System V/386 or V/68 servers. Also, to forestall questions, it does not create shared libraries. Randy Brown (uunet!erik!rbrown)