Date: Sat, 22 Feb 2003 02:35:27 EST-10EDT,10,-1,0,7200,3,-1,0,7200,3600 Subject: [Ux2bs_Archive] No. 85 ************************************************** Friday 21 February 2003 Number 85 ************************************************** Subjects for today 1 Re: Perl build on JFS : Henry Sobotka 2 Re: Perl build on JFS : Henry Sobotka 3 Re: Perl build on JFS : Hakan" 4 Re: New Posix/2 build : Stefan.Neis at t-online.de 5 UX2BS Mail List Moving! : Ted Sikora 6 Re: New Posix/2 build : John Poltorak 7 Re: Perl build on JFS : Henry Sobotka 8 Re: Perl build on JFS : John Poltorak 9 Re: Perl build on JFS : Henry Sobotka 10 Re: New Posix/2 build : brpms at earthlink.net 11 Re: UX2BS Mail List Moving! : Ted Sikora 12 Re: Perl build on JFS : Hakan" 13 dlfcn.h : John Poltorak 14 Perl: ../lib/Net/t/hostname.t : Maynard" 15 Re: Perl build on JFS : Henry Sobotka 16 Re: Perl build on JFS : Maynard" 17 Re: Perl build on JFS : Henry Sobotka 18 Re: Perl build on JFS : Hakan" 19 Re: Perl build on JFS : Henry Sobotka 20 Re: Perl build on JFS : Maynard" 21 Re: Perl: ../lib/Net/t/hostname.t : Maynard" 22 Re: Perl build on JFS : Hakan" 23 Perl build with Posix/2 : John Poltorak 24 Re: dlfcn.h : Stefan.Neis at t-online.de 25 Re: Perl build with Posix/2 : Stefan.Neis at t-online.de 26 Re: Perl build on JFS : Henry Sobotka 27 Re: UX2BS Mail List Moving! : John Poltorak 28 Re: Perl build on JFS : John Poltorak 29 Re: Perl build on JFS : John Poltorak 30 Re: UX2BS Mail List Moving! : Ted Sikora 31 Re: Perl build on JFS : John Poltorak 32 Re: UX2BS Mail List Moving! : John Poltorak 33 Re: Perl build on JFS : John Poltorak 34 Re: UX2BS Mail List Moving! : Andrew Belov" 35 Re: Perl build on JFS : John Poltorak 36 Re: Perl: ../lib/Net/t/hostname.t : John Poltorak 37 Re: New Posix/2 build : John Poltorak 38 Re: Perl build on JFS : John Poltorak 39 Re: ux2_bootstrap : John Poltorak 40 Re: ux2_bootstrap : Hartmut Krafft" 41 Perl & Posix/2 : Stefan.Neis at t-online.de **= Email 1 ==========================** Date: Sat, 22 Feb 2003 01:13:27 -0500 From: Henry Sobotka Subject: Re: Perl build on JFS Hakan wrote: > > I am disappointed and annoyed that we are getting different results, > different errors etc. when building the very same application, perl. > The different errors and test results are telling us that we are ending > up with versions of perl which are different in some respects, however > small, something that quite clearly can/will influence our ability to > run various perl scripts and could result in various insidious errors. First of all, playing build roulette is not a particularly good way to get a production-quality Perl if that's what you're after. Nor at this stage is that what the UX2 build is intended for, but rather as a test for the build system. Moreover, people are working on it so it's changing and perhaps you're also tinkering at your end, so you simply can't expect consistent identical results. As for the test failures, the important ones are the common denominator, in this case the five that everyone on JFS and without Object REXX has run into; and those turn out to all be problems with the tests in four cases and Configure in the fifth, and not with Perl itself. The others are local problems. If you look at HiRes.t, for instance, you'll see that the two tests your Perl fails are skipped unless usleep is defined; usleep isn't in EMX so you're obviously linking with some other library that has an implementation, and the Perl test failure may be exposing a bug in it, or the test may have to be tweaked to accommodate some quirk of the usleep you're linking with, or it may be something completely different. Similarly, the hostname.t failures reported don't indicate a flaw in perl.exe, but rather a missing envar the test expects. Perhaps the test should be skipped altogether if $ENV{'HOSTNAME'} is undefined. In other words, the way to avoid the insidious errors you mention is to put the test failures under a microscope and try to understand what's causing them; especially those that might only be occurring on your system because all anyone else who can't reproduce them can do is speculate. h~ _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 2 ==========================** Date: Sat, 22 Feb 2003 01:40:29 -0500 From: Henry Sobotka Subject: Re: Perl build on JFS John Poltorak wrote: > > If so, I'll set HOSTNAME to localhost in the standard build environment. Is > that likely to cause any problems? It might, as localhost is generally attached to 127.0.0.1, and the box could have a real name associated with a static or dynamic IP address even though HOSTNAME isn't set; if a get_hostname function returns "localhost" from another machine, "ping localhost" will ping yourself instead of it, assuming you have loopback enabled. hostname.exe spits out "standalone" if HOSTNAME isn't set, which might be a safer solution. h~ _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 3 ==========================** Date: Sat, 22 Feb 2003 11:22:22 -0500 (EST) From: "Hakan" Subject: Re: Perl build on JFS Henry, Two quick comments, HOSTNAME is indeed defined on this system (just like it was the first time I built a few weeks ago) and no other Unix-related software has been installed inbetween (which might explain the HiRes failure.) The only difference is that the first time I built on a local HPFS-formatted drive, this time I built on a network JFS-drive, otherwise no changes. Hakan On Sat, 22 Feb 2003 01:13:27 -0500, Henry Sobotka wrote: >Hakan wrote: >> >> I am disappointed and annoyed that we are getting different results, >> different errors etc. when building the very same application, perl. >> The different errors and test results are telling us that we are ending >> up with versions of perl which are different in some respects, however >> small, something that quite clearly can/will influence our ability to >> run various perl scripts and could result in various insidious errors. > >First of all, playing build roulette is not a particularly good way to >get a production-quality Perl if that's what you're after. Nor at this >stage is that what the UX2 build is intended for, but rather as a test >for the build system. Moreover, people are working on it so it's >changing and perhaps you're also tinkering at your end, so you simply >can't expect consistent identical results. As for the test failures, the >important ones are the common denominator, in this case the five that >everyone on JFS and without Object REXX has run into; and those turn out >to all be problems with the tests in four cases and Configure in the >fifth, and not with Perl itself. The others are local problems. If you >look at HiRes.t, for instance, you'll see that the two tests your Perl >fails are skipped unless usleep is defined; usleep isn't in EMX so >you're obviously linking with some other library that has an >implementation, and the Perl test failure may be exposing a bug in it, >or the test may have to be tweaked to accommodate some quirk of the >usleep you're linking with, or it may be something completely different. >Similarly, the hostname.t failures reported don't indicate a flaw in >perl.exe, but rather a missing envar the test expects. Perhaps the test >should be skipped altogether if $ENV{'HOSTNAME'} is undefined. In other >words, the way to avoid the insidious errors you mention is to put the >test failures under a microscope and try to understand what's causing >them; especially those that might only be occurring on your system >because all anyone else who can't reproduce them can do is speculate. > >h~ > >_______________________________________________ >UX2BS mailing list >UX2BS at powerusersbbs.net >http://powerusersbbs.net/mailman/listinfo/ux2bs > > _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 4 ==========================** Date: Sat, 22 Feb 2003 12:08:01 +0100 From: Stefan.Neis at t-online.de Subject: Re: New Posix/2 build > > Hi, > > >I uploaded a new build of Posix/2 to os2ports.com/incoming. > >This one should fix the known issues of the previous one > >(Note: new p2-gcc.exe, new include/sys/unistdx.h, new > > cExt.{a,lib}). > > Don't find it there. Well, instead of describing all the exciting new files contained in the package, maybe I should rather have mentioned that it is named 'p2beta0_1_1_bin.zip'. ;-) Does that help in finding it? Regards, Stefan _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 5 ==========================** Date: Sat, 22 Feb 2003 12:16:49 -0500 From: Ted Sikora Subject: UX2BS Mail List Moving! Sometime Monday. The need for archiving has prompted the move. All list members will be migrated and an announcement will be posted when it is operational. http://os2ports.com/mailman/listinfo/ux2bs Zope-OS2 and Mailman-OS2 will remain where they are. -- Ted Sikora tsikora at ntplx.net _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 6 ==========================** Date: Sat, 22 Feb 2003 12:28:36 +0000 From: John Poltorak Subject: Re: New Posix/2 build On Fri, Feb 21, 2003 at 05:19:18PM -0700, brpms at earthlink.net wrote: > In <18mIme-0TQVV2C at fmrl05.sul.t-online.com>, on 02/21/03 > at 08:24 PM, Stefan.Neis at t-online.de said: > > > Hi, > > >I uploaded a new build of Posix/2 to os2ports.com/incoming. > >This one should fix the known issues of the previous one > >(Note: new p2-gcc.exe, new include/sys/unistdx.h, new > > cExt.{a,lib}). > > >Ted, is that you, who is maintaining the server? > >I think it would be a good idea to delete the two older > >versions (p2alpha*.zip and p2beta0_1_0.zip) from incoming, > >now that a newer one is available. > > > Regards, > > Stefan > > Don't find it there. You should get it automatically by running ux2_bootstrap.cmd, assuming Ted's RSYNC server has been refreshed. The Posix/2 archive is part of the Baseline toolset and should get refreshed every time there is an update. The latest release allows me to build a number of apps using Posix/2 two which would not build with them before. Anyone who is feeling adventurous may want to try out a new UX2BS which does incorporate Posix/2 as much as possible at this stage. You should be able to get this test release with a new ux2_bootstrap.cmd which you can get using:- wget ftp://unixos2: at 213.152.37.92/pub/unixos2/build_system.test/lib/ux2_bootstrap.cmd Please not that the directory structure has been amended slightly and that all the baseline tools are now under \unixos2. This means that everything which you now see under \usr\bin is something that has been freshly built. This means that starting from scratch is much easier since you can easily use the installed toolset but delete all the apps you have built by deleting \usr. If anyone tries this and has any problems, then please let me know. > Paul Schwartz > > -- > ----------------------------------------------------------- > brpms at earthlink.net > ----------------------------------------------------------- -- John _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 7 ==========================** Date: Sat, 22 Feb 2003 12:37:21 -0500 From: Henry Sobotka Subject: Re: Perl build on JFS John Poltorak wrote: > > Whilst I don't disagree with what you say, I don't see how it could be > linking with some other library since the variables are explicityly set > and I don't think that other libraries can be picked up arbitrarily. usleep is in Posix/2. A Configure run with Stefan's gcc would normally find and define it. > Is there any way at all that hardware, including video drivers and > screen resolution can play any part in test results? I'm thinking in > particular about the tests which cause lots of windows to appear all over > the screen... Given that the "lots" is really only four (it looks like more because they're minimized and maximized several times) and they're VIO windows, I would think any hardware/video driver problems would turn up before trying to build Perl. > Another problem I'm trying to get to the bottom of is 'too many files > open' This happens on one of my systems, although it didn't used to. It's > cured by adding emxopt=-h1024 to the environment, the value is an > arbitrary one which just happens to work. I'm sure this will also catch a > number of other people, but I'm not sure what the ideal solution is... Basically the value should be high enough to accommodate the largest number of files opened by the preprocessor reading #include's. Here it's set to 2048 (IIRC 1024 turned out to be too low for Mozilla). Since the ceiling for OS/2 is 65,536 there's lots of room for play. If you determine, for instance, that 1024 is enough for UX2, the build script could read EMXOPT from the environment and, if -h is 1024 or higher, continue; otherwise set it to 1024. h~ _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 8 ==========================** Date: Sat, 22 Feb 2003 12:57:22 +0000 From: John Poltorak Subject: Re: Perl build on JFS On Sat, Feb 22, 2003 at 01:13:27AM -0500, Henry Sobotka wrote: > Hakan wrote: > > > > I am disappointed and annoyed that we are getting different results, > > different errors etc. when building the very same application, perl. > > The different errors and test results are telling us that we are ending > > up with versions of perl which are different in some respects, however > > small, something that quite clearly can/will influence our ability to > > run various perl scripts and could result in various insidious errors. > > First of all, playing build roulette is not a particularly good way to > get a production-quality Perl if that's what you're after. Nor at this > stage is that what the UX2 build is intended for, but rather as a test > for the build system. Moreover, people are working on it so it's > changing and perhaps you're also tinkering at your end, so you simply > can't expect consistent identical results. As for the test failures, the > important ones are the common denominator, in this case the five that > everyone on JFS and without Object REXX has run into; and those turn out > to all be problems with the tests in four cases and Configure in the > fifth, and not with Perl itself. The others are local problems. If you > look at HiRes.t, for instance, you'll see that the two tests your Perl > fails are skipped unless usleep is defined; usleep isn't in EMX so > you're obviously linking with some other library that has an > implementation, Whilst I don't disagree with what you say, I don't see how it could be linking with some other library since the variables are explicityly set and I don't think that other libraries can be picked up arbitrarily. > and the Perl test failure may be exposing a bug in it, > or the test may have to be tweaked to accommodate some quirk of the > usleep you're linking with, or it may be something completely different. > Similarly, the hostname.t failures reported don't indicate a flaw in > perl.exe, but rather a missing envar the test expects. Perhaps the test > should be skipped altogether if $ENV{'HOSTNAME'} is undefined. In other > words, the way to avoid the insidious errors you mention is to put the > test failures under a microscope and try to understand what's causing > them; especially those that might only be occurring on your system > because all anyone else who can't reproduce them can do is speculate. Is there any way at all that hardware, including video drivers and screen resolution can play any part in test results? I'm thinking in particular about the tests which cause lots of windows to appear all over the screen... Another problem I'm trying to get to the bottom of is 'too many files open' This happens on one of my systems, although it didn't used to. It's cured by adding emxopt=-h1024 to the environment, the value is an arbitrary one which just happens to work. I'm sure this will also catch a number of other people, but I'm not sure what the ideal solution is... > h~ -- John _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 9 ==========================** Date: Sat, 22 Feb 2003 13:01:50 -0500 From: Henry Sobotka Subject: Re: Perl build on JFS Hakan wrote: > > The only difference is that the first time I built on a local HPFS-formatted > drive, this time I built on a network JFS-drive, otherwise no changes. I vaguely recall John determining that building on network drives results in the stat.t failure. It might be worthwhile if the people encountering the uncommon failures tried to pinpoint the cause by running those tests individually, and a page or section of a page listed them and what we know about them, e.g. Perl Test Failures stat.t Occurs when building on a network drive hostname.t HOSTNAME not set Shell.t ??? HiRes.t #7-8; usleep defined (using POSIX/2?) basic.t Configure bug; fixed since 5.8.0 os2_ea.t Occurs on JFS drives because of illegal EA name os2_process.t (#174 and #209) possible problem with test (Others) ??? os2_process_kid.t Failure goes away if os2_process.t passes rx_cmprt.t Object REXX supports Unix line-endings, vanilla REXX doesn't and sooner or later the blanks might get filled in. h~ _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 10 ==========================** Date: Sat, 22 Feb 2003 13:24:49 -0700 From: brpms at earthlink.net Subject: Re: New Posix/2 build In <18mXVi-0gCBIuC at fmrl05.sul.t-online.com>, on 02/22/03 at 12:08 PM, Stefan.Neis at t-online.de said: >> > Hi, >> >> >I uploaded a new build of Posix/2 to os2ports.com/incoming. >> >This one should fix the known issues of the previous one >> >(Note: new p2-gcc.exe, new include/sys/unistdx.h, new >> > cExt.{a,lib}). >> >> Don't find it there. >Well, instead of describing all the exciting new files contained in the >package, maybe I should rather have mentioned that it is named >'p2beta0_1_1_bin.zip'. ;-) >Does that help in finding it? Yes! That one I see. > Regards, > Stefan and rsync got it, but i don;t think anything is using it yet. I'm still trying to figure out why i can't reproduce that good perl build. Paul Schwartz -- ----------------------------------------------------------- brpms at earthlink.net ----------------------------------------------------------- _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 11 ==========================** Date: Sat, 22 Feb 2003 14:14:53 -0500 From: Ted Sikora Subject: Re: UX2BS Mail List Moving! Andrew Belov wrote: > On Sat, 22 Feb 2003 12:16:49 -0500, Ted Sikora wrote: > > >>Sometime Monday. The need for archiving has prompted the move. >>All list members will be migrated and an announcement will be posted >>when it is operational. > > > By the way, what happened to the original UnixOS2 mailing list? There is no mail, no > response from there, and even my own posts are not echoed - no signs of any > activity since Feb 16th. Have I missed something? > os2-unix at manninghammills.org ? It's alive and very busy. Have John check and see if your still subscribed? His server was acting up around that time I think. -- Ted Sikora tsikora at ntplx.net _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 12 ==========================** Date: Sat, 22 Feb 2003 14:26:16 -0500 (EST) From: "Hakan" Subject: Re: Perl build on JFS John, On Sat, 22 Feb 2003 19:15:28 +0000, John Poltorak wrote: >On Sat, Feb 22, 2003 at 01:01:50PM -0500, Henry Sobotka wrote: > >> Perl Test Failures >> >> HiRes.t #7-8; usleep defined (using POSIX/2?) > >This could not happen using the current standard build environment because >the Posix/2 package is simply extracted to %uxrt%\posix2 but not >referenced anywhere within the environment. > >Of course, it could have been due to an outdated script which tried to >incorporated Posix/2 by extracting it to \usr\lib and \usr\include. > >I guess the date of baseline_inst.cmd would tell you how uptodate it was. baseline_inst.cmd is dated 2/3/03 2.42PM. Again, please also note that I do have hostname defined in config.sys which would not explain the failure in hostname.t. Hakan > >> h~ > > >-- >John > > > >_______________________________________________ >UX2BS mailing list >UX2BS at powerusersbbs.net >http://powerusersbbs.net/mailman/listinfo/ux2bs > > _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 13 ==========================** Date: Sat, 22 Feb 2003 14:43:13 +0000 From: John Poltorak Subject: dlfcn.h Until yesterday I had never heard of dlfcn.h, but now I know it is a header provided by Posix/2. It is also exists, albeit a completely different version, in Perl, in the os2 subdir. In fact os2.c contains:- #include "dlfcn.h" Should this be removed if using Posix/2? -- John _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 14 ==========================** Date: Sat, 22 Feb 2003 14:51:14 -0600 (CST) From: "Maynard" Subject: Perl: ../lib/Net/t/hostname.t ../lib/Net/t/hostname.t passes if there is a DOMAIN envar set. as I reported on January 19 and 25 and has been posted since then at http://warped.mentabolism.org/UnixOS2/perl.html Please reserve your disbelief until after you've tested it yourself, and/or examined the error report and the source file. EMX returns a hostname of 'standalone' if none other is set elsewhere. Prove that to yourself by typing 'hostname' at a command prompt and searching your files for instance of 'standalone'. Later, -- Maynard _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 15 ==========================** Date: Sat, 22 Feb 2003 15:01:50 -0500 From: Henry Sobotka Subject: Re: Perl build on JFS Hakan wrote: > > Again, please also note that I do have hostname defined in config.sys which > would not explain the failure in hostname.t. I don't see hostname.t in your last list of failures, but I notice io_sock.t which should go into the ??? category (especially as I don't think you're the first to hit it). If hostname.t failed the first time for you and you had HOSTNAME set, then we know there at least two possible causes for it failing and the second remains unknown. h~ _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 16 ==========================** Date: Sat, 22 Feb 2003 15:08:09 -0600 (CST) From: "Maynard" Subject: Re: Perl build on JFS Hakan, >What is the rationale for "unsetting" hostname even though it may exist >in config.sys? Many months ago we determined to construct the ux2bs "build environment baseline" with a minimum configuration, particularly in ENVARs and on PATH and LIBPATH. You can modify a ux2bs environment for production use, rather than ux2bs development and testing, by modifying \UnixOS2\lib\ux2_env.cmd, and running that to start your command session. It's pretty easy to support and maintain several different environments in this manner. Cheers/ux/2, -- Maynard _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 17 ==========================** Date: Sat, 22 Feb 2003 15:13:02 -0500 From: Henry Sobotka Subject: Re: Perl build on JFS John Poltorak wrote: > > What you have in config.sys is largely irrelevant. If you look at the end > of any build log, then will see the settings of the the environment variables > in use by the build. They are set fairly strictly by \unixos2\ux2_env.cmd. Unless you wipe out the environment created by config.sys, all Perl has to do is go $ENV{'WHATEVER_VARIABLE'} to read and use it. Similarly, if you look at the output of 'make -p', you'll find your whole config.sys environment there. At best ux2_env.cmd can override some variables and supply missing requirements. h~ _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 18 ==========================** Date: Sat, 22 Feb 2003 15:32:36 -0500 (EST) From: "Hakan" Subject: Re: Perl build on JFS What is the rationale for "unsetting" hostname even though it may exist in config.sys? On Sat, 22 Feb 2003 20:11:06 +0000, John Poltorak wrote: >On Sat, Feb 22, 2003 at 03:01:50PM -0500, Henry Sobotka wrote: >> Hakan wrote: >> > >> > Again, please also note that I do have hostname defined in config.sys which >> > would not explain the failure in hostname.t. >> >> I don't see hostname.t in your last list of failures, but I notice >> io_sock.t which should go into the ??? category (especially as I don't >> think you're the first to hit it). If hostname.t failed the first time >> for you and you had HOSTNAME set, then we know there at least two >> possible causes for it failing and the second remains unknown. > >As it stands just at the moment *everyone* should fail the hostname test, >because any HOSTNAME variable in config.sys is unset by the build >environment. But I don't think everyone is failing it, and have a >suspicion that Perl determines it via some host name resolution through >RESOLV. Is that a possibility? > >> h~ > >-- >John > > > >_______________________________________________ >UX2BS mailing list >UX2BS at powerusersbbs.net >http://powerusersbbs.net/mailman/listinfo/ux2bs > > _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 19 ==========================** Date: Sat, 22 Feb 2003 16:22:07 -0500 From: Henry Sobotka Subject: Re: Perl build on JFS John Poltorak wrote: > > As it stands just at the moment *everyone* should fail the hostname test, > because any HOSTNAME variable in config.sys is unset by the build > environment. But I don't think everyone is failing it, and have a > suspicion that Perl determines it via some host name resolution through > RESOLV. Is that a possibility? hostname.t uses Net::Domain, at the top of which is a comment: # Try every conceivable way to get hostname. which includes if(open(RES,"/etc/resolv.conf")) { h~ _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 20 ==========================** Date: Sat, 22 Feb 2003 16:28:30 -0600 (CST) From: "Maynard" Subject: Re: Perl build on JFS Hakan, >If hostname is needed, why not keep the setting already present in >config.sys? at risk of repeating myself again, HOSTNAME is not needed in config.sys by ux2bs because it is supplied by emx via hostname.exe -- Maynard _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 21 ==========================** Date: Sat, 22 Feb 2003 16:32:53 -0600 (CST) From: "Maynard" Subject: Re: Perl: ../lib/Net/t/hostname.t John, >Do we know whether Perl can establish the local hostame via DNS using >RESOLV if the correct entry exists on the server? RESOLV does not provide hostnames The perl test fails for resolv and resolv2 in %etc%, and passes when %etc%\resolv.conf (thanks Henry!) contains a 'domain' line typical of os2's resolv2 -- Maynard _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 22 ==========================** Date: Sat, 22 Feb 2003 16:47:48 -0500 (EST) From: "Hakan" Subject: Re: Perl build on JFS If hostname is needed, why not keep the setting already present in config.sys? On Sat, 22 Feb 2003 20:47:03 +0000, John Poltorak wrote: >On Sat, Feb 22, 2003 at 03:32:36PM -0500, Hakan wrote: >> What is the rationale for "unsetting" hostname even though it may exist >> in config.sys? > >It isn't a matter of unsetting HOSTNAME in particular, it's a matter of >clearing everything and then adding things back when it is known there is >a need for particular variables. > >On one of my systems I have 84 variables set, and I doubt whether I have >two systems set the same. It's impossible to have a standard environment >without setting it to known values. Of course setting these values only >affects the session in which a build takes place and has no effect on your >normal settings since no changes are made to config.sys. > > >-- >John > > > >_______________________________________________ >UX2BS mailing list >UX2BS at powerusersbbs.net >http://powerusersbbs.net/mailman/listinfo/ux2bs > > _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 23 ==========================** Date: Sat, 22 Feb 2003 16:53:38 +0000 From: John Poltorak Subject: Perl build with Posix/2 My latest attempt at building Perl using Posix/2 including supplied patches ends during Make with errors such as:- gcc -Zexe -Zomf -Zmt -Zcrtdll -Zstack 32000 -Zlinker /e:2 -o miniperl miniperlmain.obj perl.obj `echo malloc.obj gv.obj toke.obj perly.obj op.obj regcomp.obj dump.obj util.obj mg.obj reentr.obj hv.obj av.obj run.obj pp_hot.obj sv.obj pp.obj scope.obj pp_ctl.obj pp_sys.obj doop.obj doio.obj regexec.obj utf8.obj taint.obj deb.obj universal.obj xsutils.obj globals.obj perlio.obj perlapi.obj numeric.obj locale.obj pp_pack.obj pp_sort.obj os2.obj dl_os2.obj|sed -e 's/\bop\./opmini./g'` -lsocket -lm -lbsd -Zmap -Zlinker /map/PM:VIO pp_sys.obj(pp_sys.obj) : error L2029: 'getpriority' : unresolved external pp_sys.obj(pp_sys.obj) : error L2029: 'setpriority' : unresolved external pp_sys.obj(pp_sys.obj) : error L2029: 'endgrent' : unresolved external pp_sys.obj(pp_sys.obj) : error L2029: 'getgrent' : unresolved external pp_sys.obj(pp_sys.obj) : error L2029: 'setgrent' : unresolved external There were 5 errors detected make: *** [miniperl] Error 1 Does this mean anything to anyone? -- John _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 24 ==========================** Date: Sat, 22 Feb 2003 18:47:29 +0100 From: Stefan.Neis at t-online.de Subject: Re: dlfcn.h Hi, > Until yesterday I had never heard of dlfcn.h, but now I know it is a > header provided by Posix/2. It is also exists, albeit a completely > different version, in Perl, in the os2 subdir. Actually, it's sufficiently similar, so it should not cause problems... > In fact os2.c contains:- > > #include "dlfcn.h" > > Should this be removed if using Posix/2? Well, having two versions (and especially two implementations, see dl_os2.c) is certainly unnecessary, so those could be removed, but there's no pressing need to get rid of them ... ;-) Regards, Stefan _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 25 ==========================** Date: Sat, 22 Feb 2003 18:47:30 +0100 From: Stefan.Neis at t-online.de Subject: Re: Perl build with Posix/2 ** Reply to note from John Poltorak Sat, 22 Feb 2003 16:53:38 +0000 > > > My latest attempt at building Perl using Posix/2 including supplied > patches ends during Make with errors such as:- > > > gcc (snipp) > pp_sort.obj os2.obj dl_os2.obj|sed -e 's/\bop\./opmini./g'` -lsocket -lm > -lbsd -Zmap -Zlinker /map/PM:VIO > > > pp_sys.obj(pp_sys.obj) : error L2029: 'getpriority' : unresolved external (snipp) > > Does this mean anything to anyone? You're apparently not linking against the posix2-libraries (which probably also implies that Configure didn't do to well :-( ). One more patch I forgot to mention again, sorry. :-( Either "explain" the build system to link with "-lcExt -lstcExt" by modifying the "libs=..." line in hints/os2.sh (line 216?) to read libs='-lcExt -lstcExt -lsocket -lm -lbsd' as suggested by Henry some time ago (that's what I'm doing), or "explain" the build system that you want to use p2-gcc, _not_ gcc (or temporarily rename p2-gcc to gcc), presumably by changing the "cc=..." and the "ld=..." lines in the same file (hints/os2.sh, lines 24 and 225 ?) (untested!). Looks like hints/os2.sh is overriding any settings from the environment. :-( Regards, Stefan _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 26 ==========================** Date: Sat, 22 Feb 2003 18:50:29 -0500 From: Henry Sobotka Subject: Re: Perl build on JFS John Poltorak wrote: > > That is not much use on OS/2. > > Should we get this changed to %ETC%\RESOLV(2) ? Something like: my $resolv = "/etc/resolv.conf"; $resolv = $ENV{"ETC"}."/resolv" if $^O eq "os2"; if (open (RES, $resolv)) { would be more acceptable than simply changing it. But here "domain" is only in resolv; resolv2 just lists nameservers; and there's no resolv(2). So trying "every conceivable way" would require something like: if ($^O eq "os2") { my at resolv = ("/resolv", "/resolv2", "/resolv(2)"); foreach my $r ( at resolv) { $domain = read_res($ENV{"ETC"}.$r); next unless defined $domain; } } else { $domain = read_res("/etc/resolv.conf"); } with the read process: sub read_res { if (open (RES, shift)) { etc. returning $domain if defined. h~ _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 27 ==========================** Date: Sat, 22 Feb 2003 19:03:21 +0000 From: John Poltorak Subject: Re: UX2BS Mail List Moving! On Sat, Feb 22, 2003 at 08:31:54PM +0300, Andrew Belov wrote: > On Sat, 22 Feb 2003 12:16:49 -0500, Ted Sikora wrote: > > >Sometime Monday. The need for archiving has prompted the move. > >All list members will be migrated and an announcement will be posted > >when it is operational. > > By the way, what happened to the original UnixOS2 mailing list? There is no mail, no > response from there, and even my own posts are not echoed - no signs of any > activity since Feb 16th. Have I missed something? I just checked my logs and see this:- 5770 C> RCPT To: 5770 C< 553 Mail blocked. Please contact abuse at nm.ru (213.152.37.93) 5770 C> RSET 5770 C< 250 flushed 5770 C> QUIT 5770 C< 221 grif.newmail.ru I don't know whether this is due to majordomo sending out mail with the wrong domain name. I hadn't changed a setting since my domain name was changed, but I wasn't aware of any problems. Maybe your ISP attempts to verify that the domain name is correct. I'll try sending you a test email to see if it gets through... -- John _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 28 ==========================** Date: Sat, 22 Feb 2003 19:15:28 +0000 From: John Poltorak Subject: Re: Perl build on JFS On Sat, Feb 22, 2003 at 01:01:50PM -0500, Henry Sobotka wrote: > Perl Test Failures > > HiRes.t #7-8; usleep defined (using POSIX/2?) This could not happen using the current standard build environment because the Posix/2 package is simply extracted to %uxrt%\posix2 but not referenced anywhere within the environment. Of course, it could have been due to an outdated script which tried to incorporated Posix/2 by extracting it to \usr\lib and \usr\include. I guess the date of baseline_inst.cmd would tell you how uptodate it was. > h~ -- John _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 29 ==========================** Date: Sat, 22 Feb 2003 19:37:54 +0000 From: John Poltorak Subject: Re: Perl build on JFS On Sat, Feb 22, 2003 at 02:26:16PM -0500, Hakan wrote: > John, > > On Sat, 22 Feb 2003 19:15:28 +0000, John Poltorak wrote: > > >On Sat, Feb 22, 2003 at 01:01:50PM -0500, Henry Sobotka wrote: > > > >> Perl Test Failures > >> > >> HiRes.t #7-8; usleep defined (using POSIX/2?) > > > >This could not happen using the current standard build environment because > >the Posix/2 package is simply extracted to %uxrt%\posix2 but not > >referenced anywhere within the environment. > > > >Of course, it could have been due to an outdated script which tried to > >incorporated Posix/2 by extracting it to \usr\lib and \usr\include. > > > >I guess the date of baseline_inst.cmd would tell you how uptodate it was. > > baseline_inst.cmd is dated 2/3/03 2.42PM. That is almost three weeks ago. I suggest you do a refresh. > Again, please also note that > I do have hostname defined in config.sys which would not explain the > failure in hostname.t. What you have in config.sys is largely irrelevant. If you look at the end of any build log, then will see the settings of the the environment variables in use by the build. They are set fairly strictly by \unixos2\ux2_env.cmd. > Hakan -- John _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 30 ==========================** Date: Sat, 22 Feb 2003 19:48:56 -0500 From: Ted Sikora Subject: Re: UX2BS Mail List Moving! Is it mail.ru? I had a problem with them a month ago all messages were staying in the queue and couldn't be sent. John Poltorak wrote: > On Sat, Feb 22, 2003 at 08:31:54PM +0300, Andrew Belov wrote: > >>On Sat, 22 Feb 2003 12:16:49 -0500, Ted Sikora wrote: >> >> >>>Sometime Monday. The need for archiving has prompted the move. >>>All list members will be migrated and an announcement will be posted >>>when it is operational. >> >>By the way, what happened to the original UnixOS2 mailing list? There is no mail, no >>response from there, and even my own posts are not echoed - no signs of any >>activity since Feb 16th. Have I missed something? > > > I just received your email, but can't reply, because it won't get through > to you. > > It seems that ISP's in Russia are very quick to block mail and it is > difficult to know how this situation arose. The same happened with > Andrew Zabolotny who had been on the list and then suddenly got blocked. > It is obviously impossible for me to notify anyone directly when mail is > getting blocked. > > -- -- Ted Sikora tsikora at ntplx.net _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 31 ==========================** Date: Sat, 22 Feb 2003 20:11:06 +0000 From: John Poltorak Subject: Re: Perl build on JFS On Sat, Feb 22, 2003 at 03:01:50PM -0500, Henry Sobotka wrote: > Hakan wrote: > > > > Again, please also note that I do have hostname defined in config.sys which > > would not explain the failure in hostname.t. > > I don't see hostname.t in your last list of failures, but I notice > io_sock.t which should go into the ??? category (especially as I don't > think you're the first to hit it). If hostname.t failed the first time > for you and you had HOSTNAME set, then we know there at least two > possible causes for it failing and the second remains unknown. As it stands just at the moment *everyone* should fail the hostname test, because any HOSTNAME variable in config.sys is unset by the build environment. But I don't think everyone is failing it, and have a suspicion that Perl determines it via some host name resolution through RESOLV. Is that a possibility? > h~ -- John _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 32 ==========================** Date: Sat, 22 Feb 2003 20:21:42 +0000 From: John Poltorak Subject: Re: UX2BS Mail List Moving! On Sat, Feb 22, 2003 at 08:31:54PM +0300, Andrew Belov wrote: > On Sat, 22 Feb 2003 12:16:49 -0500, Ted Sikora wrote: > > >Sometime Monday. The need for archiving has prompted the move. > >All list members will be migrated and an announcement will be posted > >when it is operational. > > By the way, what happened to the original UnixOS2 mailing list? There is no mail, no > response from there, and even my own posts are not echoed - no signs of any > activity since Feb 16th. Have I missed something? I just received your email, but can't reply, because it won't get through to you. It seems that ISP's in Russia are very quick to block mail and it is difficult to know how this situation arose. The same happened with Andrew Zabolotny who had been on the list and then suddenly got blocked. It is obviously impossible for me to notify anyone directly when mail is getting blocked. -- John _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 33 ==========================** Date: Sat, 22 Feb 2003 20:28:23 +0000 From: John Poltorak Subject: Re: Perl build on JFS On Sat, Feb 22, 2003 at 03:13:02PM -0500, Henry Sobotka wrote: > John Poltorak wrote: > > > > What you have in config.sys is largely irrelevant. If you look at the end > > of any build log, then will see the settings of the the environment variables > > in use by the build. They are set fairly strictly by \unixos2\ux2_env.cmd. > > Unless you wipe out the environment created by config.sys, That is exactly what happens. Then a specific number are set to known values to eliminate the possibility of any differences to the standard environment being introduced locally. Just have a look through a few of the scripts. If you run any build you will see a ux2_env.cmd which is the program that sets up all the environment variables availble in this shell. > h~ -- John _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 34 ==========================** Date: Sat, 22 Feb 2003 20:31:54 +0300 (MSK) From: "Andrew Belov" Subject: Re: UX2BS Mail List Moving! On Sat, 22 Feb 2003 12:16:49 -0500, Ted Sikora wrote: >Sometime Monday. The need for archiving has prompted the move. >All list members will be migrated and an announcement will be posted >when it is operational. By the way, what happened to the original UnixOS2 mailing list? There is no mail, no response from there, and even my own posts are not echoed - no signs of any activity since Feb 16th. Have I missed something? _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 35 ==========================** Date: Sat, 22 Feb 2003 20:47:03 +0000 From: John Poltorak Subject: Re: Perl build on JFS On Sat, Feb 22, 2003 at 03:32:36PM -0500, Hakan wrote: > What is the rationale for "unsetting" hostname even though it may exist > in config.sys? It isn't a matter of unsetting HOSTNAME in particular, it's a matter of clearing everything and then adding things back when it is known there is a need for particular variables. On one of my systems I have 84 variables set, and I doubt whether I have two systems set the same. It's impossible to have a standard environment without setting it to known values. Of course setting these values only affects the session in which a build takes place and has no effect on your normal settings since no changes are made to config.sys. -- John _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 36 ==========================** Date: Sat, 22 Feb 2003 21:08:41 +0000 From: John Poltorak Subject: Re: Perl: ../lib/Net/t/hostname.t On Sat, Feb 22, 2003 at 02:51:14PM -0600, Maynard wrote: > ../lib/Net/t/hostname.t passes if there is a DOMAIN envar set. > > as I reported on January 19 and 25 and has been posted since then at > > http://warped.mentabolism.org/UnixOS2/perl.html > > Please reserve your disbelief until after you've tested it yourself, > and/or examined the error report and the source file. > > EMX returns a hostname of 'standalone' if none other is set elsewhere. > Prove that to yourself by typing 'hostname' at a command prompt and > searching your files for instance of 'standalone'. Do we know whether Perl can establish the local hostame via DNS using RESOLV if the correct entry exists on the server? > Later, > > -- Maynard -- John _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 37 ==========================** Date: Sat, 22 Feb 2003 21:36:44 +0000 From: John Poltorak Subject: Re: New Posix/2 build On Sat, Feb 22, 2003 at 01:24:49PM -0700, brpms at earthlink.net wrote: > In <18mXVi-0gCBIuC at fmrl05.sul.t-online.com>, on 02/22/03 > at 12:08 PM, Stefan.Neis at t-online.de said: > > >> > Hi, > >> > >> >I uploaded a new build of Posix/2 to os2ports.com/incoming. > >> >This one should fix the known issues of the previous one > >> >(Note: new p2-gcc.exe, new include/sys/unistdx.h, new > >> > cExt.{a,lib}). > >> > >> Don't find it there. > > >Well, instead of describing all the exciting new files contained in the > >package, maybe I should rather have mentioned that it is named > >'p2beta0_1_1_bin.zip'. ;-) > >Does that help in finding it? > > Yes! That one I see. > > > Regards, > > Stefan > > and rsync got it, but i don;t think anything is using it yet. No, it isn't used in the current environment, just provided as a convenience for those people who want to use it. It is an integral part of the new build environment though... > I'm still > trying to figure out why i can't reproduce that good perl build. You should delete \unixos2\workdir\perl-5.8.0 and any occurances of perl* under \usr\bin or \usr\lib. You are probably pulling in some part of a previous build somewhere. I use a RAMFS disk for building things on and can be certain of a clean start for every build. If you get stuck, compare your perl.log with that provided by Maynard on the webpages he mentioned. > Paul Schwartz > > -- > ----------------------------------------------------------- > brpms at earthlink.net > ----------------------------------------------------------- > -- John _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 38 ==========================** Date: Sat, 22 Feb 2003 21:41:01 +0000 From: John Poltorak Subject: Re: Perl build on JFS On Sat, Feb 22, 2003 at 04:22:07PM -0500, Henry Sobotka wrote: > John Poltorak wrote: > > > > As it stands just at the moment *everyone* should fail the hostname test, > > because any HOSTNAME variable in config.sys is unset by the build > > environment. But I don't think everyone is failing it, and have a > > suspicion that Perl determines it via some host name resolution through > > RESOLV. Is that a possibility? > > hostname.t uses Net::Domain, at the top of which is a comment: > > # Try every conceivable way to get hostname. > > which includes > > if(open(RES,"/etc/resolv.conf")) { That is not much use on OS/2. Should we get this changed to %ETC%\RESOLV(2) ? > h~ -- John _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 39 ==========================** Date: Sat, 22 Feb 2003 21:57:22 +0000 From: John Poltorak Subject: Re: ux2_bootstrap On Sat, Feb 22, 2003 at 10:23:41PM +0100, Hartmut Krafft wrote: > On Thu, 20 Feb 2003 12:00:29 +0000, John Poltorak wrote: > > >Basically, I'd like to know about any existing probelms as I'd like to > >release some updated scripts which now incorporate Posix/2, as well > >slightly restructuring the build environment. If there are any outstanding > >problems, I can address them at the same time. > > Well, I've not yet met a success till now. > > I had no problems WGETting stuff. > First I tried on my main box which is still at W4/FP10, and the process failed quite early > (which is due to extensive use of BEGINLIBPATH, as I found out). > > Then, I tried again on my eCS test box, added link386.exe, too. Here, the build fails for some > reason at the point where Perl is to be made. '*** No rule to make target `miniperl.exe''. > > Also, I get 'Too many open files in shell' during the 'make depend' which may be related. > I tried adding "SET EMXOPT=-h1024" to Config.sys, but did not see a change. Now, I'm out > of ideas. Try setting it to -h2048 and see if it gets any further... Remember to save the old log so you have something to compare it with. Also delete \unixos2\workdir\perl-5.8.0 so you start with a clean working directory. -- John _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 40 ==========================** Date: Sat, 22 Feb 2003 22:23:41 +0100 (MEZ) From: "Hartmut Krafft" Subject: Re: ux2_bootstrap On Thu, 20 Feb 2003 12:00:29 +0000, John Poltorak wrote: >Basically, I'd like to know about any existing probelms as I'd like to >release some updated scripts which now incorporate Posix/2, as well >slightly restructuring the build environment. If there are any outstanding >problems, I can address them at the same time. Well, I've not yet met a success till now. I had no problems WGETting stuff. First I tried on my main box which is still at W4/FP10, and the process failed quite early (which is due to extensive use of BEGINLIBPATH, as I found out). Then, I tried again on my eCS test box, added link386.exe, too. Here, the build fails for some reason at the point where Perl is to be made. '*** No rule to make target `miniperl.exe''. Also, I get 'Too many open files in shell' during the 'make depend' which may be related. I tried adding "SET EMXOPT=-h1024" to Config.sys, but did not see a change. Now, I'm out of ideas. Any hints? Below, Im attaching the perl.log - please snip when replying! Greetings, Hartmut --- perl.log --- Sat Feb 22 18:54:45 GMT 2003 There is no LINK386.EXE on the path, which could cause problems Press ENTER to continue DIR URL CFLAGS LDFLAGS CFGPARMS MAKEPARM SRC http://www.cpan.org/src/perl-5.8.0.tar.gz . URL http://www.cpan.org/src/perl-5.8.0.tar.gz CFLAGS LDFLAGS CFGPARMS . MAKEPARM SRC perl-5.8.0.tar.gz already retrieved e:/unixos2/workdir gzip: stdout: Broken pipe tar: "gzip" exit status 1 e:/unixos2/workdir/perl-5.8.0 using perl specific Build script [E:\unixos2\workdir\perl-5.8.0]set C_INCLUDE_PATH=e:/emx/include [E:\unixos2\workdir\perl-5.8.0]set LIBRARY_PATH=e:/emx/lib [E:\unixos2\workdir\perl-5.8.0]set PATH=e:/emx/bin;e:\usr\bin;e:\emx\bin;e:\usr\local\bin;c:\os2; [E:\unixos2\workdir\perl-5.8.0]sed "s~c:\/~c:\/~" Configure 1>Configure.tmp && mv Configure.tmp Configure [E:\unixos2\workdir\perl-5.8.0]sh Configure -des -Dprefix=/usr (I see you are using the Korn shell. Some ksh's blow up on Configure, mainly on older exotic systems. If yours does, try the Bourne shell instead.) First let's make sure your kit is complete. Checking... Locating common programs... Checking compatibility between e:/usr/bin/echo.exe and builtin echo (if any)... Symbolic links are NOT supported. No symbolic links, so not testing for their testing... Good, your tr supports [:lower:] and [:upper:] to convert case. Using [:upper:] and [:lower:] to convert case. 3b1 dynix isc_2 opus super-ux aix dynixptx linux os2 svr4 altos486 epix lynxos os390 svr5 amigaos esix4 machten posix-bc ti1500 apollo fps machten_2 powerux titanos atheos freebsd mint qnx ultrix_4 aux_3 genix mips rhapsody umips beos gnu mpc sco unicos bsdos greenhills mpeix sco_2_3_0 unicosmk convexos hpux ncr_tower sco_2_3_1 unisysdynix cxux i386 netbsd sco_2_3_2 utekv cygwin irix_4 newsos4 sco_2_3_3 uts darwin irix_5 next_3 sco_2_3_4 uwin dcosx irix_6 next_3_0 solaris_2 vmesa dec_osf irix_6_0 next_4 stellar vos dgux irix_6_1 nonstopux sunos_4_0 dos_djgpp isc openbsd sunos_4_1 Which of these apply, if any? [os2] !!! Apparently there is no need to patch Configure. Operating system name? [os2] Operating system version? [2] Build Perl for SOCKS? [n] Use the PerlIO abstraction layer? [y] Build a threading Perl? [n] Build Perl for multiplicity? [n] Use which C compiler? [gcc] Checking for GNU cc in disguise and/or its version number... Now, how can we feed standard input to your C preprocessor... Directories to use for library searches? [e:/emx/lib e:/emx/lib/mt] What is the file extension used for shared libraries? [dll] Checking for optional libraries... What libraries to use? [-lsocket -lm -lbsd] What optimizer/debugger flag should be used? [-O2 -fomit-frame-pointer -malign-loops=2 -malign-jumps=2 -malign-functions=2 -s] Any additional cc flags? [-Zomf -Zmt -DDOSISH -DOS2=2 -DEMBED -I. -D_EMX_CRT_REV_=64] Let me guess what the preprocessor flags are... Any additional ld flags (NOT including libraries)? [-Zexe -Zomf -Zmt -Zcrtdll -Zstack 32000 -Zlinker /e:2] Checking your choice of C compiler and flags for coherency... Checking to see how big your integers are... Checking to see if you have long long... Checking to see how big your long longs are... Computing filename position in cpp output for #include directives... NOT found. Checking to see if you have int64_t... Checking which 64-bit integer type we could use... We could use 'long long' for 64-bit integers. Try to use 64-bit integers, if available? [n] Try to use maximal 64-bit support, if available? [n] Checking for GNU C Library... Shall I use e:/emx/bin/nm to extract C symbols from the libraries? [y] Where is your C library? [e:/emx/lib/mt/c_import.lib] Extracting names from the following files for later perusal: e:/emx/lib/bsd.lib e:/emx/lib/c_alias.lib e:/emx/lib/m.lib e:/emx/lib/mt/c_import.lib e:/emx/lib/socket.lib This may take a while..................e:/emx/bin/nm didn't seem to work right. Trying emxomfar instead... Ok. sqrtl() found. Checking to see how big your double precision numbers are... Checking to see if you have long double... Checking to see how big your long doubles are... Checking how to print long doubles... modfl() found. Your modfl() seems okay for large values. Try to use long doubles if available? [n] What is your architecture name [os2] Perlio selected. Installation prefix to use? (~name ok) [/usr] AFS does not seem to be running... What installation prefix should I use for installing files? (~name ok) [/usr] Getting the current patchlevel... Pathname where the private library files will reside? (~name ok) [/usr/lib/perl5/5.8.0] Where do you want to put the public architecture-dependent libraries? (~name ok) [/usr/lib/perl5/5.8.0/os2] I don't think setuid scripts are secure (no /dev/fd directory). Do you want to do setuid/setgid emulation? [n] found. found. Checking to see how well your C compiler groks the void type... Good. It appears to support void to the level perl5 wants. Checking to see how big your pointers are... Do you wish to attempt to use the malloc that comes with perl5? [y] Your system wants malloc to return 'void *', it would seem. Your system uses void free(), it would seem. Installation prefix to use for add-on modules and utilities? (~name ok) [/usr] Pathname for the site-specific library files? (~name ok) [/usr/lib/perl5/site_perl/5.8.0] Pathname for the site-specific architecture-dependent library files? (~name ok) [/usr/lib/perl5/site_perl/5.8.0/os2] Do you want to configure vendor-specific add-on directories? [n] Colon-separated list of additional directories for perl to search? [none] Checking out function prototypes... Pathname where the public executables will reside? (~name ok) [/usr/bin] Install any extra modules (y or n)? [n] List of earlier versions to include in at INC? [none] NOT found. dlopen() NOT found. Do you wish to use dynamic loading? [y] Source file to use for dynamic loading [ext/DynaLoader/dl_dlopen.xs] Any special flags to pass to gcc -c to compile shared library modules? [-Zdll] What command should be used to create dynamic libraries? [gcc] Any special flags to pass to gcc to create a dynamically loaded library? [-Zdll -Zomf -Zmt -Zcrtdll -Zlinker /e:2] Any special flags to pass to gcc to use dynamic linking? [none] Build a shared libperl.dll (y/n) [y] What name do you want to give to the shared libperl? [libperl.lib] Could not find manual pages in source form. Where do the main Perl5 manual pages (source) go? (~name ok) [/usr/man/man1] What suffix should be used for the main Perl5 man pages? [1] You can have filenames longer than 14 characters. Where do the perl5 library man pages (source) go? (~name ok) [/usr/man/man3] What suffix should be used for the perl5 library man pages? [3] Figuring out host name... Your host name appears to be "standalone". Right? [y] What is your domain name? [.nonet] What is your e-mail address? [root at standalone.nonet] Perl administrator e-mail address [root at standalone.nonet] Do you want to install only the version-specific parts of perl? [n] What shall I put after the #! to start up perl ("none" to not use #!)? [/usr/bin/perl] Where do you keep publicly executable scripts? (~name ok) [/usr/bin] Pathname where the add-on public executables should be installed? (~name ok) [/usr/bin] off_t found. Checking to see how big your file offsets are... fpos_t found. Checking the size of fpos_t... Try to understand large files, if available? [y] qgcvt() NOT found. fwalk() NOT found. access() found. defines the *_OK access constants. accessx() NOT found. alarm() found. NOT found. found. found. Testing to see if we should include , or both. We'll include . Checking to see if your struct tm has tm_zone field... Checking to see if your struct tm has tm_gmtoff field... asctime_r() NOT found. atolf() NOT found. atoll() NOT found. Checking whether your compiler can handle __attribute__ ... bcmp() found. bcopy() found. found. getpgrp() found. You have to use getpgrp() instead of getpgrp(pid). setpgrp() NOT found. bzero() found. You have void (*signal())(). Checking whether your C compiler can cast large floats to int32. Checking whether your C compiler can cast negative float to unsigned. vprintf() found. Your vsprintf() returns (int). chown() NOT found. chroot() NOT found. chsize() found. class() NOT found. Hmm... Looks like you have Berkeley networking support. socketpair() NOT found. Checking the availability of certain socket constants... found. Checking to see if your system supports struct cmsghdr... Checking to see if your C compiler knows about "const"... crypt() NOT found. crypt() NOT found. NOT found. crypt_r() NOT found. ctermid_r() NOT found. ctime_r() NOT found. cuserid() found. found. found. DBL_DIG found. dbmclose() NOT found. difftime() found. found. :7: warning: #warning requires Your directory entries are struct dirent. :7: warning: #warning requires Good, your directory entry keeps length information in d_namlen. found. NOT found. dirfd() found. dlerror() NOT found. NOT found. *** WHOA THERE!!! *** The recommended value for $i_dlfcn on this machine was "define"! Keep the recommended value? [y] What is the extension of dynamically loaded modules [dll] Checking whether your dlsym() needs a leading underscore ... I can't compile and run the test program. I'm guessing that dlsym doesn't need a leading underscore. drand48_r() NOT found. dup2() found. eaccess() NOT found. endgrent() NOT found. found. endgrent_r() NOT found. endhostent() found. found. endhostent_r() NOT found. endnetent() found. endnetent_r() NOT found. endprotoent() found. endprotoent_r() NOT found. endpwent() found. found. endpwent_r() NOT found. endservent() found. endservent_r() NOT found. defines the O_* constants... and you have the 3 argument form of open(). Using instead of . found. We'll be including . found. We don't need to include if we include . Figuring out the flag used by open() for non-blocking I/O... Let's see what value errno gets from read() on a O_NONBLOCK file... fchdir() NOT found. fchmod() NOT found. fchown() NOT found. fcntl() found. Well, your system knows about the normal fd_set typedef... and you have the normal fd_set macros (just as I'd expect). fgetpos() found. finite() NOT found. finitel() NOT found. flock() found. fork() found. fp_class() NOT found. pathconf() found. fpathconf() found. fpclass() NOT found. fpclassify() NOT found. fpclassl() NOT found. Checking to see if you have fpos64_t... frexpl() found. found. NOT found. Checking to see if your system supports struct fs_data... fseeko() NOT found. fsetpos() found. fstatfs() NOT found. statvfs() NOT found. fstatvfs() NOT found. fsync() found. ftello() NOT found. getcwd() found. getespwnam() NOT found. getfsstat() NOT found. getgrent() NOT found. getgrent_r() NOT found. getgrgid_r() NOT found. getgrnam_r() NOT found. gethostbyaddr() found. gethostbyname() found. gethostent() found. gethostname() found. uname() found. Shall I ignore gethostname() from now on? [n] gethostbyaddr_r() NOT found. gethostbyname_r() NOT found. gethostent_r() NOT found. getitimer() NOT found. getlogin() found. getlogin_r() NOT found. getmnt() NOT found. getmntent() NOT found. getnetbyaddr() found. getnetbyname() found. getnetent() found. getnetbyaddr_r() NOT found. getnetbyname_r() NOT found. getnetent_r() NOT found. getpagesize() found. getprotobyname() found. getprotobynumber() found. getprotoent() found. getpgid() NOT found. getpgrp2() NOT found. getppid() found. getpriority() NOT found. *** WHOA THERE!!! *** The recommended value for $d_getprior on this machine was "define"! Keep the recommended value? [y] getprotobyname_r() NOT found. getprotobynumber_r() NOT found. getprotoent_r() NOT found. getprpwnam() NOT found. getpwent() found. getpwent_r() NOT found. getpwnam_r() NOT found. getpwuid_r() NOT found. getservbyname() found. getservbyport() found. getservent() found. getservbyname_r() NOT found. getservbyport_r() NOT found. getservent_r() NOT found. getspnam() NOT found. NOT found. getspnam_r() NOT found. gettimeofday() found. gmtime_r() NOT found. hasmntopt() NOT found. found. found. htonl() NOT found. But it seems to be defined as a macro. strchr() found. inet_aton() NOT found. isascii() found. isfinite() NOT found. isinf() NOT found. isnan() NOT found. isnanl() NOT found. killpg() NOT found. lchown() NOT found. LDBL_DIG found. link() NOT found. localtime_r() NOT found. localeconv() found. lockf() NOT found. lstat() NOT found. madvise() NOT found. mblen() found. mbstowcs() found. mbtowc() found. memchr() found. memcmp() found. memcpy() found. memmove() found. memset() found. mkdir() found. mkdtemp() NOT found. mkfifo() NOT found. mkstemp() found. mkstemps() NOT found. mktime() found. NOT found. mmap() NOT found. mprotect() NOT found. msgctl() NOT found. msgget() NOT found. msgsnd() NOT found. msgrcv() NOT found. You don't have the full msg*(2) library. Checking to see if your system supports struct msghdr... msync() NOT found. munmap() NOT found. nice() NOT found. NOT found. nl_langinfo() NOT found. Checking to see how big your characters are (hey, you never know)... What is the size of a character (in bytes)? [1] Checking to see if your C compiler knows about "volatile"... Choosing the C types to be used for Perl's internal types... Checking how many bits of your UVs your NVs can preserve... Checking to see if you have off64_t... pause() found. pipe() found. poll() NOT found. readlink() NOT found. random_r() NOT found. readdir() found. seekdir() found. telldir() found. rewinddir() found. readdir64_r() NOT found. readdir_r() NOT found. readv() found. recvmsg() NOT found. rename() found. rmdir() found. found. Checking if your memcmp() can compare relative magnitude... select() found. semctl() NOT found. semget() NOT found. semop() NOT found. You don't have the full sem*(2) library. You do not have union semun in . sendmsg() NOT found. setegid() NOT found. seteuid() NOT found. setgrent() NOT found. setgrent_r() NOT found. sethostent() found. sethostent_r() NOT found. setitimer() NOT found. setlinebuf() NOT found. setlocale() found. found. setlocale_r() NOT found. setnetent() found. setnetent_r() NOT found. setprotoent() found. setpgid() found. setpgrp2() NOT found. setpriority() NOT found. *** WHOA THERE!!! *** The recommended value for $d_setprior on this machine was "define"! Keep the recommended value? [y] setproctitle() NOT found. setprotoent_r() NOT found. setpwent() found. setpwent_r() NOT found. setregid() NOT found. setresgid() NOT found. setreuid() NOT found. setresuid() NOT found. setrgid() NOT found. setruid() NOT found. setservent() found. setservent_r() NOT found. setsid() NOT found. setvbuf() found. NOT found. shmctl() NOT found. shmget() NOT found. shmat() NOT found. shmdt() NOT found. You don't have the full shm*(2) library. sigaction() found. sigprocmask() found. POSIX sigsetjmp found. sockatmark() NOT found. socks5_init() NOT found. srand48_r() NOT found. srandom_r() NOT found. found. Checking to see if your struct stat has st_blocks field... NOT found. found. Checking to see if your system supports struct statfs... Checking to see if your struct statfs has f_flags field... Checking how std your stdio is... Checking to see what happens if we set the stdio ptr... Increasing ptr in your stdio leaves cnt unchanged. Good. Checking how to access stdio streams by file descriptor number... I can't figure out how to access stdio streams by file descriptor number. strcoll() found. Checking to see if your C compiler can copy structs... strerror() found. strerror_r() NOT found. strftime() found. strtod() found. strtol() found. strtold() found. strtoll() NOT found. *** WHOA THERE!!! *** The recommended value for $d_strtoll on this machine was "define"! Keep the recommended value? [y] strtoq() NOT found. strtoul() found. strtoull() NOT found. *** WHOA THERE!!! *** The recommended value for $d_strtoull on this machine was "define"! Keep the recommended value? [y] strtouq() NOT found. strxfrm() found. symlink() NOT found. syscall() NOT found. sysconf() found. system() found. tcgetpgrp() found. tcsetpgrp() found. time() found. time_t found. found. times() found. clock_t found. tmpnam_r() NOT found. truncate() found. ttyname_r() NOT found. tzname[] found. (Testing for character data alignment may crash the test. That's okay.) You can access character data pretty unalignedly. ualarm() NOT found. umask() found. unordered() NOT found. usleep() NOT found. ustat() NOT found. vfork() NOT found. closedir() found. Checking whether closedir() returns a status... wait4() NOT found. waitpid() found. wcstombs() found. wctomb() found. writev() found. Checking alignment constraints... Doubles must be aligned on a how-many-byte boundary? [4] Checking to see how your cpp does stuff like catenate tokens... Oh! Smells like ANSI's been here. found. Checking Berkeley DB version ... Looks OK. Checking return type needed for hash for Berkeley DB ... Checking return type needed for prefix for Berkeley DB ... Looking for a random number function... OK, found random(). Use which function to generate random numbers? [random] How many bits does your random() function produce? [31] Determining whether or not we are on an EBCDIC system... Nope, no EBCDIC, probably ASCII or some ISO Latin. Or UTF-8. Checking how to flush all pending stdio output... Checking the size of gid_t... Checking the sign of gid_t... Checking how to print 64-bit integers... Checking the format strings to be used for Perl's internal types... Checking the format string to be used for gids... getgroups() found. setgroups() NOT found. What type pointer is the second argument to getgroups() and setgroups()? [gid_t] Checking if your e:/usr/bin/make program sets $(MAKE)... mode_t found. found. found. We'll include to get va_dcl definition. It seems that you don't need va_copy(). size_t found. What pager is used on your system? [cmd /c more] pid_t found. Checking how to generate random libraries on your machine... Your select() operates on 256 bits at a time. Generating a list of signal names and numbers... Checking the size of size_t... Checking to see if you have socklen_t... NOT found. I'll be using ssize_t for functions returning a byte count. Your stdio uses signed chars. Checking the size of uid_t... Checking the sign of uid_t... Checking the format string to be used for uids... Which compiler compiler (yacc) shall I use? [yacc] NOT found. NOT found. NOT found. NOT found. found. NOT found. NOT found. NOT found. found. NOT found. NOT found. Guessing which symbols your C compiler and preprocessor define... tcsetattr() found. You have POSIX termios.h... good! found. NOT found. NOT found. NOT found. found. No socket ioctls found in . NOT found. NOT found. found. NOT found. NOT found. found. found. found. NOT found. found. NOT found. NOT found. Looking for extensions... What extensions do you wish to load dynamically? [B ByteLoader Cwd DB_File Data/Dumper Devel/DProf Devel/PPPort Devel/Peek Digest/MD5 Encode Fcntl File/Glob Filter/Util/Call IO List/Util MIME/Base64 OS2/ExtAttr OS2/PrfDB OS2/Process OS2/REXX Opcode POSIX PerlIO/encoding PerlIO/scalar PerlIO/via SDBM_File Socket Storable Sys/Hostname Sys/Syslog Time/HiRes Unicode/Normalize XS/APItest XS/Typemap attrs re threads threads/shared] What extensions do you wish to load statically? [none] Stripping down executable paths... Creating config.sh... Doing variable substitutions on .SH files... Extracting cflags (with variable substitutions) Extracting config.h (with variable substitutions) Extracting makeaperl (with variable substitutions) Extracting makedepend (with variable substitutions) Extracting makedir (with variable substitutions) Extracting Makefile (with variable substitutions) Configure: os2/Makefile.SHs[14]: too many files open in shell Configure: ./Makefile.SH[24]: too many files open in shell Configure: UU/myread[16]: too many files open in shell [E:\unixos2\workdir\perl-5.8.0]make makefile:737: warning: overriding commands for target `TAGS' makefile:729: warning: ignoring old commands for target `TAGS' make depend MAKEDEPEND= make[1]: Entering directory `/unixos2/workdir/perl-5.8.0' makefile:737: warning: overriding commands for target `TAGS' makefile:729: warning: ignoring old commands for target `TAGS' sh ./makedepend MAKE=make make[2]: Entering directory `/unixos2/workdir/perl-5.8.0' GNUmakefile:737: warning: overriding commands for target `TAGS' GNUmakefile:729: warning: ignoring old commands for target `TAGS' sh writemain lib/auto/DynaLoader/DynaLoader.lib > writemain.tmp sh mv-if-diff writemain.tmp perlmain.c echo malloc.c av.c scope.c op.c doop.c doio.c dump.c hv.c mg.c reentr.c perl.c perly.c pp.c pp_hot.c pp_ctl.c pp_sys.c regcomp.c regexec.c utf8.c gv.c sv.c taint.c toke.c util.c deb.c run.c universal.c xsutils.c globals.c perlio.c perlapi.c numeric.c locale.c pp_pack.c pp_sort.c miniperlmain.c perlmain.c | tr ' ' '\n\r' >.clist make[2]: Leaving directory `/unixos2/workdir/perl-5.8.0' Finding dependencies for malloc.obj. Finding dependencies for av.obj. Finding dependencies for scope.obj. Finding dependencies for op.obj. Finding dependencies for doop.obj. Finding dependencies for doio.obj. Finding dependencies for dump.obj. Finding dependencies for hv.obj. Finding dependencies for mg.obj. Finding dependencies for reentr.obj. Finding dependencies for perl.obj. Finding dependencies for perly.obj. Finding dependencies for pp.obj. Finding dependencies for pp_hot.obj. Finding dependencies for pp_ctl.obj. Finding dependencies for pp_sys.obj. Finding dependencies for regcomp.obj. Finding dependencies for regexec.obj. Finding dependencies for utf8.obj. Finding dependencies for gv.obj. Finding dependencies for sv.obj. Finding dependencies for taint.obj. Finding dependencies for toke.obj. Finding dependencies for util.obj. Finding dependencies for deb.obj. Finding dependencies for run.obj. Finding dependencies for universal.obj. Finding dependencies for xsutils.obj. Finding dependencies for globals.obj. Finding dependencies for perlio.obj. Finding dependencies for perlapi.obj. Finding dependencies for numeric.obj. Finding dependencies for locale.obj. Finding dependencies for pp_pack.obj. Finding dependencies for pp_sort.obj. Finding dependencies for miniperlmain.obj. Finding dependencies for perlmain.obj. make[2]: Entering directory `/unixos2/workdir/perl-5.8.0' GNUmakefile:737: warning: overriding commands for target `TAGS' GNUmakefile:729: warning: ignoring old commands for target `TAGS' echo Makefile.SH cflags.SH config_h.SH makeaperl.SH makedepend.SH makedir.SH myconfig.SH writemain.SH pod/Makefile.SH | tr ' ' '\n\r' >.shlist make[2]: Leaving directory `/unixos2/workdir/perl-5.8.0' Updating GNUmakefile... test -s perlmain.c && touch perlmain.c cd x2p; make depend make[2]: Entering directory `/unixos2/workdir/perl-5.8.0/x2p' sh ../makedepend MAKE=make make[3]: Entering directory `/unixos2/workdir/perl-5.8.0/x2p' echo hash.c str.c util.c walk.c | tr ' ' '\n\r' >.clist make[3]: Leaving directory `/unixos2/workdir/perl-5.8.0/x2p' Finding dependencies for hash.obj. Finding dependencies for str.obj. Finding dependencies for util.obj. Finding dependencies for walk.obj. make[3]: Entering directory `/unixos2/workdir/perl-5.8.0/x2p' echo Makefile.SH cflags.SH | tr ' ' '\n\r' >.shlist make[3]: Leaving directory `/unixos2/workdir/perl-5.8.0/x2p' Updating GNUmakefile... make[2]: Leaving directory `/unixos2/workdir/perl-5.8.0/x2p' make[1]: Leaving directory `/unixos2/workdir/perl-5.8.0' make: *** No rule to make target `miniperl.exe', needed by `all'. Stop. [E:\unixos2\workdir\perl-5.8.0]make test GNUmakefile:737: warning: overriding commands for target `TAGS' GNUmakefile:729: warning: ignoring old commands for target `TAGS' make: *** No rule to make target `miniperl.exe', needed by `preplibrary'. Stop. [E:\unixos2\workdir\perl-5.8.0]make install GNUmakefile:737: warning: overriding commands for target `TAGS' GNUmakefile:729: warning: ignoring old commands for target `TAGS' make install.perl install.man STRIPFLAGS= make[1]: Entering directory `/unixos2/workdir/perl-5.8.0' GNUmakefile:737: warning: overriding commands for target `TAGS' GNUmakefile:729: warning: ignoring old commands for target `TAGS' make[1]: *** No rule to make target `miniperl.exe', needed by `all'. Stop. make[1]: Leaving directory `/unixos2/workdir/perl-5.8.0' make: *** [install] Error 2 [E:\unixos2\workdir\perl-5.8.0]cd t [E:\unixos2\workdir\perl-5.8.0\t]perl harness SYS1041: perl ist weder ein interner oder externer Befehl, noch ein ausführbares Programm oder eine Stapeldatei. [E:\unixos2\workdir\perl-5.8.0\t]cp perl.exe e:/usr/bin cp: perl.exe: No such file or directory [E:\unixos2\workdir\perl-5.8.0\t]cp perl*.dll e:/usr/lib cp: perl*.dll: No such file or directory UXRT=e: TMP=e:\tmp PATH=e:\usr\bin;e:\emx\bin;e:\usr\local\bin;c:\os2; DPATH=c:\os2\system BLD_HOME=e:/unixos2 CFLAGS= ETC=e:\etc CONFIG_SITE= CPLUS_INCLUDE_PATH=e:/usr/include/cpp;e:/usr/include;e:/emx/include/cpp;e:/emx/in clude C_INCLUDE_PATH=e:/usr/include;e:/emx/include LIBRARY_PATH=e:/usr/lib;e:/emx/lib HOME=e:/home/root COMSPEC=c:\os2\cmd.exe REPOSITORY=e:/unixos2/archives/source BUILDLOGS=e:/unixos2/logs LDFLAGS= BLDRT=e: OS2_SHELL=c:\os2\cmd.exe TERM=ansi-color-3 OSRT=c: WORKDIR=e:/unixos2/workdir OS2PATH=e:\usr\bin;e:\emx\bin;e:\usr\local\bin;c:\os2; _emx_sig=000033d2:00000000 Version 4.50 des Operating System/2 wird verwendet Überarbeitungsversion 14.072 Sat Feb 22 19:25:55 GMT 2003 elapsed time: 1870 secs end --- _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs **= Email 41 ==========================** Date: Sat, 22 Feb 2003 23:55:01 +0100 From: Stefan.Neis at t-online.de Subject: Perl & Posix/2 Hi, Just to give everyone something to compare too, here are the final statistics of my Perl build with Posix/2 enabled. Not quite as good as before, so this gives some hints for debugging/improving Posix/2. Regards, Stefan ---------------------------------- Failed 15/726 test scripts, 97.93% okay. 433/68803 subtests failed, 99.37% okay. Failed Test Stat Wstat Total Fail Failed List of Failed ------------------------------------------------------------------------------- ./ext/DB_File/t/db-btree.t 22 5632 163 145 88.96% 19-163 ./ext/DB_File/t/db-hash.t 22 5632 117 103 88.03% 15-117 ./ext/DB_File/t/db-recno.t 2 512 160 144 90.00% 17-160 ./ext/IO/lib/IO/t/io_poll.t 10 1 10.00% 3 ./ext/Time/HiRes/HiRes.t 3 768 25 6 24.00% 20-25 ./lib/ExtUtils/t/basic.t 1 256 17 1 5.88% 14 ./lib/File/Temp/t/posix.t 255 65280 7 5 71.43% 3-7 ./lib/File/Temp/t/tempfile.t 255 65280 20 11 55.00% 10-20 ./lib/Memoize/t/errors.t 11 1 9.09% 4 ./lib/Memoize/t/tie.t 22 5632 4 4 100.00% 1-4 lib/os2_process.t 3 768 227 3 1.32% 90 174 209 lib/os2_process_kid.t 227 3 1.32% 90 174 209 lib/rx_cmprt.t 255 65280 18 3 16.67% 16-18 op/stat.t 22 5632 73 2 2.74% 72-73 run/fresh_perl.t 97 1 1.03% 91 60 tests and 554 subtests skipped. _______________________________________________ UX2BS mailing list UX2BS at powerusersbbs.net http://powerusersbbs.net/mailman/listinfo/ux2bs