Date: Wed, 7 Jan 2004 00:04:07 EST-10EDT,10,-1,0,7200,3,-1,0,7200,3600 Subject: [UnixOS2_Archive] No. 285 ************************************************** Tuesday 06 January 2004 Number 285 ************************************************** Subjects for today 1 Re: NetPBM : Stefan Neis 2 Re: NetPBM : John Poltorak 3 RE: cannot find command processor in the path : Dave Webster 4 Re: NetPBM : Stefan Neis 5 Re: GNU ed : Dave and Natalie" 6 Re: NetPBM : Franz Bakan" 7 Re: Make bugs? : Franz Bakan" 8 Re: Make bugs? : Dave and Natalie" 9 Re: New ZLIB : Dave and Natalie" 10 Re: NetPBM : Stefan Neis 11 Re: NetPBM : Andreas Buening **= Email 1 ==========================** Date: Mon, 5 Jan 2004 13:48:33 +0100 (CET) From: Stefan Neis Subject: Re: NetPBM On Mon, 5 Jan 2004, Franz Bakan wrote: > >and even more confusing > >there is some error about "spawn"... > > try an older make version Well, I intentionally tried a newer one, because the version I "normally" use complained about getcwd or something similar... Anyway, all I wanted to say is, that it doesn't work OOTB. > checking whether setvbuf arguments are reversed... configure: error: can not run > test program while cross compiling > ... Now the interesting question is why does configure think that you are using a cross-compiler. What autoconf version was used to generate this "configure"? Regards, Stefan -- Micro$oft is not an answer. It is a question. The answer is 'no'. **= Email 2 ==========================** Date: Mon, 5 Jan 2004 12:53:18 +0000 From: John Poltorak Subject: Re: NetPBM On Mon, Jan 05, 2004 at 01:48:33PM +0100, Stefan Neis wrote: > > checking whether setvbuf arguments are reversed... configure: error: can not run > > test program while cross compiling > > ... > > Now the interesting question is why does configure think that you are > using a cross-compiler. What autoconf version was used to generate > this "configure"? Are you referring to GOCR or NetPBM? > Regards, > Stefan > -- > Micro$oft is not an answer. It is a question. The answer is 'no'. > -- John **= Email 3 ==========================** Date: Mon, 5 Jan 2004 08:42:47 -0600 From: Dave Webster Subject: RE: cannot find command processor in the path This can also be a result from a syntactical makefile bug. In particular look for any extra line continuation characters or missing ones (the "\" character). -----Original Message----- From: Andreas Buening [mailto:andreas.buening at nexgo.de] Sent: Saturday, January 03, 2004 8:51 AM To: os2-unix at mail.warpix.org Subject: Re: cannot find command processor in the path John Poltorak wrote: > > On Tue, Dec 30, 2003 at 02:59:44PM +0000, John Poltorak wrote: > > Why do I get an error msg about the command processor not being found in > > the path when it clearly is? > > This appears to be problem related to Make.... > > It does not occur when using an old version of Make (v3.76.1) but it's a > definite no-go with v3.79.1. Should I be using some additional envar such > as MAKESHELL or somesuch to get round the problem? I think I fixed such a bug some time ago. Can you try a newer version? Bye, Andreas **= Email 4 ==========================** Date: Mon, 5 Jan 2004 17:51:24 +0100 (CET) From: Stefan Neis Subject: Re: NetPBM On Mon, 5 Jan 2004, John Poltorak wrote: > > > checking whether setvbuf arguments are reversed... configure: error: can not run > > > test program while cross compiling > > > ... > > > > Now the interesting question is why does configure think that you are > > using a cross-compiler. What autoconf version was used to generate > > this "configure"? > > Are you referring to GOCR or NetPBM? I don't really no which configure script generated said error ... I suppose it's probably GOCR (since NetPBM uses some perl script that's apparently not related to autoconf), but I'm just guessing... Regards, Stefan **= Email 5 ==========================** Date: Mon, 05 Jan 2004 10:43:51 -0800 From: "Dave and Natalie" Subject: Re: GNU ed On Sun, 4 Jan 2004 21:11:47 +0000, John Poltorak wrote: >> While ed seems to build ok make check seems totally broken on OS/2. > >Could it be anything to do with line termination?... > >> Unluckily I can't see an easy fix > >Do you have a Linux system available? It would be interesting to see how >Make check on OS/2 differred. Well right now I don't have UX2BS working (moved /usr aside temporarilly and it won't let me move it back, open file?) so I tried it with my build enviroment. Problems, path is hard coded. It doesn't find expr, cat, or chmod. Then it tried to create a directory X:, due to pwd reporting the drive and directory. After this I just get errors like *** The script /usr/src/ed-0.2/ed-test/x.ed exited abnormally *** I don't know enough to go any further here. Dave New Email Address - please update your Address book dave_yeo at paralynx.com **= Email 6 ==========================** Date: Mon, 05 Jan 2004 22:15:56 +0100 (CET) From: "Franz Bakan" Subject: Re: NetPBM On Mon, 5 Jan 2004 13:48:33 +0100 (CET), Stefan Neis wrote: >On Mon, 5 Jan 2004, Franz Bakan wrote: > >> >and even more confusing >> >there is some error about "spawn"... >> >> try an older make version just tried myself (gnu) and found the same 'ln' and spawn-messages. It has nothing to do with make-version but with missing prgs (or prgs that do not run). By copying manually instead of 'ln -s' and patchting the scripts I can get so far to get the .c in /lib compiled to .o but not further ( ld fails ... ;-) ) Perhaps someone with more knowledge about ld and friends has more luck... >> checking whether setvbuf arguments are reversed... configure: error: can not run >> test program while cross compiling >> ... > >Now the interesting question is why does configure think that you are >using a cross-compiler. What autoconf version was used to generate >this "configure"? configure has this line: # Generated automatically using autoconf version 2.13 AC_FUNC_SETVBUF_REVERSED in configure.in is probably the cause of the trouble. But I have no idea what it does. Bye Franz **= Email 7 ==========================** Date: Tue, 06 Jan 2004 00:16:51 +0100 (CET) From: "Franz Bakan" Subject: Re: Make bugs? On Wed, 17 Dec 2003 16:16:13 -0800, Dave and Natalie wrote: >>> can see some sed errors and I even had gcc not find foo.RES but renaming to foo.res worked. >> >>Confirmed and (hopefully) fixed: >>http://unix.os2site.com/sw/pub/binary/make/make-3.81rc1-r3-bin.zip > >Make still is having problems with this sed command >X:\usr\src\gnuplot\src >make -f ../config/makefile.os2 >Building allterm.h >echo T=..\term\ | tr \\ / >4alltrm.sh >grep "(T)" makefile.all | \ > sed "s/CORETERM = /CORETERM=%/;s/xlib.trm/xlib.trm%/" | \ > sed "s/%/\""/g;s/(/{/g;s/)/}/g" >>4alltrm.sh >sed: -e expression #1, char 11: Unknown option to `s' > >Process terminated by SIGPIPE >make: *** [..\docs\allterm.h] Error 1 > >Whereas using make 3.75 gives > > X:\usr\src\gnuplot\src >make3_75 -f ../config/makefile.os2 >"Building allterm.h" >echo T=..\term\ | tr \\ / >4alltrm.sh >grep "(T)" makefile.all | \ > sed "s/CORETERM = /CORETERM=%/;s/xlib.trm/xlib.trm%/" | \ > sed "s/%/\""/g;s/(/{/g;s/)/}/g" >>4alltrm.sh >bash ./4alltrm.sh >..\docs\allterm.h >... > >It appears to be the equals sign that make is stumbling over. No, probably the second sed-line is the culprit. Perhaps it's the ';' I get the same error with make rc1-r3 but it's still 'char 11' after changing sed "s/CORETERM = /CORETERM=%/;s/xlib.trm/xlib.trm%/" | \ to sed "s/CORETER = /CORETERM=%/;s/xlib.trm/xlib.trm%/" | \ in makefile.os2 Bye Franz **= Email 8 ==========================** Date: Mon, 05 Jan 2004 17:36:10 -0800 From: "Dave and Natalie" Subject: Re: Make bugs? On Tue, 06 Jan 2004 00:16:51 +0100 (CET), Franz Bakan wrote: >No, probably the second sed-line is the culprit. Perhaps it's the ';' > >I get the same error with make rc1-r3 > >but it's still 'char 11' after > >changing > >sed "s/CORETERM = /CORETERM=%/;s/xlib.trm/xlib.trm%/" | \ > >to > >sed "s/CORETER = /CORETERM=%/;s/xlib.trm/xlib.trm%/" | \ > >in makefile.os2 Interestingly I tried using 4OS2 as the shell, then the error changed to char #10. Other then that I've given up for now Dave New Email Address - please update your Address book dave_yeo at paralynx.com **= Email 9 ==========================** Date: Mon, 05 Jan 2004 21:01:37 -0800 From: "Dave and Natalie" Subject: Re: New ZLIB On Sat, 03 Jan 2004 15:52:12 +0100, Andreas Buening wrote: >As far as I know the simplest way is to write a .def file which >looks the following way: > > >LIBRARY 'intl' INITINSTANCE TERMINSTANCE >DESCRIPTION 'intl 0.11.5' >DATA MULTIPLE NONSHARED >STUB NONE > >EXPORTS > gettext at 1 > newname_for_gettext at 4 >... Thats what I would think. Unluckily I get LINK386 : error L2022: ÿ (alias ÿ) : export undefined Which is sopposed to be a nonfatal error and I do get a DLL but I get [X:\usr\src\zlib]example SYS0192: The operating system cannot run Z. Ok, I did miss an import. Link386 sure doesn't give clear messages . Dave New Email Address - please update your Address book dave_yeo at paralynx.com **= Email 10 ==========================** Date: Tue, 6 Jan 2004 12:06:29 +0100 (CET) From: Stefan Neis Subject: Re: NetPBM On Mon, 5 Jan 2004, Franz Bakan wrote: > configure has this line: > # Generated automatically using autoconf version 2.13 Since "normal" autoconf-2.13 doesn't know anything about OS/2, that probably explains why configure believes it is cross-compiling. At least, you'll need to rerun autoconf... > AC_FUNC_SETVBUF_REVERSED > > in configure.in is probably the cause of the trouble. But I have no > idea what it does. It's apparently tries to compile and run some tes program to determine some information about the functions "setvbuf". Since autoconf "knows" that it is cross-compiling, it also "knows" that it won't be able to run any generated test program anyway, so it just stops. So the problem really is to make configure realize that you actually are _not_ cross-compiling for some different platform. Regards, Stefan -- Micro$oft is not an answer. It is a question. The answer is 'no'. **= Email 11 ==========================** Date: Tue, 06 Jan 2004 12:32:08 +0100 From: Andreas Buening Subject: Re: NetPBM Franz Bakan wrote: > configure has this line: > # Generated automatically using autoconf version 2.13 > > AC_FUNC_SETVBUF_REVERSED > > in configure.in is probably the cause of the trouble. But I have no > idea what it does. From the autoconf manual: - Macro: AC_FUNC_SETVBUF_REVERSED If `setvbuf' takes the buffering type as its second argument and the buffer pointer as the third, instead of the other way around, define `SETVBUF_REVERSED'. Bye, Andreas