mdc@spt.entity.com (Marty Connor) (03/08/89)
I have a question for someone intimate with the way SCO Xenix version 2.3 handles memory. I wish to run KCL (Kyoto Common Lisp) on my Xenix box, but have not been able to, because in the 2.2 and earlier there seem to be a problem with incremental compilation. Let me explain. As described to me: "There are two segments: TEXT and DATA, each with attributes. The TEXT segment (usually code) has attribute READ-ONLY, and the DATA segment has attribute "NON-EXECUTABLE". This makes using KCL hard if not impossible because it wants to load object files that it compiles into the text segment of the process. My question is, has anything changed in 2.3 that will allow this to happen more easily? I really would like to run KCL reasonably, and could not tell from 2.3 release notes if this was the shining moment. Question 2: Barring 2.3, is there ANY 386 Unix that you know of that will run KCL as it is intended? I know there is a 386 port, Though I am not sure for what versions. I suspect ISC. Any experiences or advice appreciated. -- ---------------- Marty Connor Director of Innovation, The Entity mdc@entity.com, ...{harvard|uunet}!mit-eddie!spt!mdc
mjs@cbnews.ATT.COM (martin.j.shannon) (03/10/89)
In article <201@spt.entity.com> mdc@spt.entity.com (Marty Connor) writes: >I wish to run KCL (Kyoto Common Lisp) on my Xenix box, but have not >been able to. As described to me: "There are two segments: TEXT and >DATA, each with attributes. The TEXT segment (usually code) has attribute >READ-ONLY, and the DATA segment has attribute "NON-EXECUTABLE". > >This makes using KCL hard if not impossible because it wants to load >object files that it compiles into the text segment of the process. Actually, it intends to execute in the DATA segment, but you'll see the same result: it doesn't work, and probably generates a core dump. >My question is, has anything changed in 2.3 that will allow this to >happen more easily? I don't know about Xenix, but.... >Barring 2.3, is there ANY 386 Unix that you know of that will run KCL >as it is intended? Yes, I have successfully run KCL & AKCL on an AT&T SVR3.1 running on an AT&T WGS6386E (20 MHz tower). In the next several weeks, I intend to verify that it continues to work on AT&T's SVR3.2, and that it builds & runs on ISC's 386/ix 2.0.1 (I have no reason to doubt that either will work flawlessly). >-- >Marty Connor >Director of Innovation, The Entity >mdc@entity.com, ...{harvard|uunet}!mit-eddie!spt!mdc -- Marty Shannon; AT&T Bell Labs; Liberty Corner, NJ (Affiliation given for identification only.)