Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!gatech!psuvax1!psuvm!cunyvm!ndsuvm1.bitnet!nu013809
From: NU013809@NDSUVM1.BITNET (Greg Wettstein)
Newsgroups: comp.unix.xenix
Subject: Re: monitor diffs for xenix 386
Message-ID: <2586NU013809@NDSUVM1>
Date: 26 Jul 89 12:58:15 GMT
References: <1130@jpusa1.UUCP>
Organization: North Dakota Higher Education Computer Network, Fargo, ND
Lines: 31
Disclaimer: Author bears full responsibility for contents of this article.

I was wondering if anyone else had difficulties getting the monitor program
to work on XENIX-386 with the diffs supplied?

I was able to get the program compiled but when it is executed there is no
response, the program hangs but can be stopped with DEL.  I traced the problem
down to the first malloc call in the get_offset function in source file
monitor1.c.  This memory allocation call attempts to allocate 10080 bytes of
memory for a buffer to hold process information.  The program seems to simply
hang in the malloc and never returns.  I switched to an alternate screen and
the monitor program is consuming CPU time so it appears to be hung in a loop
somewhere.

Does anybody have any suggestions or has anyone had problem getting the
program to work?  Machine is an ALR 386/220 with 4 meg, Phoenix BIOS, 2.3.1
kernel and a 2.3 development system.  Comments or suggestions either to the
newsgroup or e-mail.  Once again thanks in advance for any assistance or
suggestions.

                                     As always,
                                     G.W. Wettstein
                                     NU013809@NDSUVM1

P.S.: Anybody else notice that the new lib.makefile has a problem creating
      the scrhelp.o file?  The previous file compiled help.c and renamed the
      object file scrhelp.o, the new makefile simply uses a standard .c.o
      implicit dependency which of course fails because there is no scrhelp.c
      source file.  I just rename the help.c file scrhelp.c and things seemed
      to go just fine.  Also compilation of docmd.c complains about a
      redefinition of signal.  I commented out the signal function declaration
      in docmd.c and allowed the compiler to use the signal declarations from
      sys/signal.h.....              G.W.