I am building perl 5.30 from source code. I successfully build perl.
Now i am trying to build DBD::Informix with my perl.
All the esql test compiled successfully but its failing when going for library.
perl Makefile.pl
The error which i am facing on console is
"
Determining which Informix libraries etc to use... MZÉ$#: not found
£èE°êE°δèM°Ç┴êM°δδ¡╢ò≥√ ╢äΦ· ïMïü╞╟à⌠√ ╕╟ñäG╟¿äGí╝⌐F╕╟î║G╟î║G╟î║G╟⌠çG╕çφïM∞╛QëUⁿïE⌠QïR ╨┌GtBïM∞╛QâΓâ·t3ïE∞╛Hâßâ∙t$ïU∞╛BâαthLâ∙:
not found jh₧% Φî╧ âδïUΦâΩ: not found"
The code which giving error is :
"my $esqlscript = (-f $esql) ? $esql : "$ID/bin/esql";
my $libs = `$Perl -pe 's/^\\s*exec/echo/' $esqlscript | sh -s -- -o pseudonym $esqlc_flags dbd::ix.o` || die "Couldn't execute $esqlscript to determine the ESQL/C libraries\n";
Note: I am using Informix client SDK 3.50."
Any help would be appreciable.Thanks
Related
im trying to relocate my latest ESP-IDF project from my harddrive into my local server, so I can work on it on multiple devices.
Sadly i was facing some issues building the project. So I tried to create the example project "Hello World" on my local server and build it. Exactly the same console ouput pops out:
Executing task: cmake -G Ninja -DPYTHON_DEPS_CHECKED=1 -DESP_PLATFORM=1 .. <
-- Found Git: C:/Users/user/.espressif/tools/idf-git/2.30.1/cmd/git.exe (found version "2.30.1.windows.1")
-- The C compiler identification is GNU 8.4.0
-- The CXX compiler identification is GNU 8.4.0
-- The ASM compiler identification is GNU
-- Found assembler: C:/Users/user/.espressif/tools/xtensa-esp32-elf/esp-2021r1-8.4.0/xtensa-esp32-elf/bin/xtensa-esp32-elf-gcc.exe
-- Check for working C compiler: C:/Users/user/.espressif/tools/xtensa-esp32-elf/esp-2021r1-8.4.0/xtensa-esp32-elf/bin/xtensa-esp32-elf-gcc.exe
-- Check for working C compiler: C:/Users/user/.espressif/tools/xtensa-esp32-elf/esp-2021r1-8.4.0/xtensa-esp32-elf/bin/xtensa-esp32-elf-gcc.exe -- broken
CMake Error at C:/Users/user/.espressif/tools/cmake/3.16.4/share/cmake-3.16/Modules/CMakeTestCCompiler.cmake:60 (message):
The C compiler
"C:/Users/user/.espressif/tools/xtensa-esp32-elf/esp-2021r1-8.4.0/xtensa-esp32-elf/bin/xtensa-esp32-elf-gcc.exe"
is not able to compile a simple test program.
It fails with the following output:
Change Dir: //192.168.0.10/projects/my_project/build/CMakeFiles/CMakeTmp
Run Build Command(s):C:/Users/user/.espressif/tools/ninja/1.10.2/ninja.exe cmTC_294a8 && [1/2] Building C object CMakeFiles/cmTC_294a8.dir/testCCompiler.c.obj
[2/2] Linking C executable cmTC_294a8
FAILED: cmTC_294a8
cmd.exe /C "cd . && C:\Users\user\.espressif\tools\xtensa-esp32-elf\esp-2021r1-8.4.0\xtensa-esp32-elf\bin\xtensa-esp32-elf-gcc.exe -mlongcalls -Wno-frame-address CMakeFiles/cmTC_294a8.dir/testCCompiler.c.obj -o cmTC_294a8 && cd ."
"\\192.168.0.10/projects/my_project\build\CMakeFiles\CMakeTmp"
CMD.EXE wurde mit dem oben angegebenen Pfad als aktuellem Verzeichnis gestartet.
UNC-Pfade werden nicht unterstützt.
Stattdessen wird das Windows-Verzeichnis als aktuelles Verzeichnis gesetzt.
xtensa-esp32-elf-gcc.exe: error: CMakeFiles/cmTC_294a8.dir/testCCompiler.c.obj: No such file or directory
xtensa-esp32-elf-gcc.exe: fatal error: no input files
compilation terminated.
ninja: build stopped: subcommand failed.
CMake will not be able to correctly generate this project.
Call Stack (most recent call first):
C:/Users/user/esp/esp-idf/tools/cmake/project.cmake:311 (__project)
CMakeLists.txt:6 (project)
-- Configuring incomplete, errors occurred!
See also "//192.168.0.10/projects/my_project/build/CMakeFiles/CMakeOutput.log".
See also "//192.168.0.10/projects/my_project/build/CMakeFiles/CMakeError.log".
The terminal process "C:\WINDOWS\System32\WindowsPowerShell\v1.0\powershell.exe -Command cmake -G Ninja -DPYTHON_DEPS_CHECKED=1 -DESP_PLATFORM=1 .." terminated with exit code: 1.
Searching for my problem lead me to similar threads, but sadly none of them were helping me.
Anyone experienced similar problems with a local server?
thanks in advance
Edit:
this might clearify my question even more. Here is my CMakeError.log:
Determining if the C compiler works failed with the following output:
Change Dir: //192.168.0.10/projects/my_project/build/CMakeFiles/CMakeTmp
Run Build Command(s):C:/Users/user/.espressif/tools/ninja/1.10.2/ninja.exe cmTC_294a8 && [1/2] Building C object CMakeFiles/cmTC_294a8.dir/testCCompiler.c.obj
[2/2] Linking C executable cmTC_294a8
FAILED: cmTC_294a8
cmd.exe /C "cd . && C:\Users\user\.espressif\tools\xtensa-esp32-elf\esp-2021r1-8.4.0\xtensa-esp32-elf\bin\xtensa-esp32-elf-gcc.exe -mlongcalls -Wno-frame-address CMakeFiles/cmTC_294a8.dir/testCCompiler.c.obj -o cmTC_294a8 && cd ."
"\\192.168.0.10\d\projects\my_project\build\CMakeFiles\CMakeTmp"
CMD.EXE wurde mit dem oben angegebenen Pfad als aktuellem Verzeichnis gestartet.
#CMD.Exe was started from current directory with above stated Path
UNC-Pfade werden nicht unterstützt.
# UNC Path is not supported
Stattdessen wird das Windows-Verzeichnis als aktuelles Verzeichnis gesetzt.
#instead windows-directory will be used as directory
xtensa-esp32-elf-gcc.exe: error: CMakeFiles/cmTC_294a8.dir/testCCompiler.c.obj: No such file or directory
xtensa-esp32-elf-gcc.exe: fatal error: no input files
compilation terminated.
ninja: build stopped: subcommand failed.
for my english-speaking fellas... the German part is roughly translated 1 line below.
You're trying to compile a project directly off a Windows network share? That's a marginal use case. I don't think there are many tools which support this, for the simple reason that nobody does such a thing. At the very least, mount the remote share as a drive in your computer.
I'd recommend creating a git repo, uploading your code there. That's a standard solution for sharing code.
I'm running this on msys2 64 bit MINGW65 shell.
I ran
$source/configure --host=x86_64-w64-mingw32 --prefix=$dist && make
and everything until the make runs properly. After getting the make issue and trying to tinker with the issue spot I keep running make for the same issue below.
Then when I run make on the folder.
it's running fine until it hits this part
make[2]: Leaving directory '/c/builds/build/REL_11_8/src/backend/utils'
make -C storage/lmgr lwlocknames.h lwlocknames.c
make[2]: Entering directory '/c/builds/build/REL_11_8/src/backend/storage/lmgr'
'/usr/bin/perl' /c/builds/source/src/backend/storage/lmgr/generate-lwlocknames.pl
/c/builds/source/src/backend/storage/lmgr/lwlocknames.txt
unable to parse lwlocknames.txt at /c/builds/source/src/backend/storage/lmgr/generate-
lwlocknames.pl line 36, <$lwlocknames> line 8.
make[2]: *** [Makefile:33: lwlocknames.h] Error 255
make[2]: Leaving directory '/c/builds/build/REL_11_8/src/backend/storage/lmgr'
make[1]: *** [Makefile:137: storage/lmgr/lwlocknames.h] Error 2
make[1]: Leaving directory '/c/builds/build/REL_11_8/src/backend'
make: *** [src/Makefile.global:372: submake-generated-headers] Error 2
In the generate-lwlocknames.pl
there is the following line that it's obviously triggering it.
But I don't know why it's triggering it or how to fix it.
I need to compile this build in order to be able to get the dll files for an extension for windows machine that is running postgres.
while (<$lwlocknames>)
{
chomp;
# Skip comments
next if /^#/;
next if /^\s*$/;
die "unable to parse lwlocknames.txt"
unless /^(\w+)\s+(\d+)$/;
(my $lockname, my $lockidx) = ($1, $2);
die "lwlocknames.txt not in order" if $lockidx < $lastlockidx;
die "lwlocknames.txt has duplicates" if $lockidx == $lastlockidx;
while ($lastlockidx < $lockidx - 1)
{
++$lastlockidx;
printf $c "%s \"<unassigned:%d>\"", $continue, $lastlockidx;
$continue = ",\n";
}
printf $c "%s \"%s\"", $continue, $lockname;
$lastlockidx = $lockidx;
$continue = ",\n";
print $h "#define $lockname (&MainLWLockArray[$lockidx].lock)\n";
}
After rerunning it a few more times it finished successfully. No changes were made to the files or methodology.
It looks like you are building with the Perl that comes with MinGW.
However, the PostgreSQL documentation is pretty clear on that point:
The following additional products are required to build PostgreSQL. Use the config.pl file to specify which directories the libraries are available in.
[...]
ActiveState Perl
ActiveState Perl is required to run the build generation scripts. MinGW or Cygwin Perl will not work. It must also be present in the PATH. Binaries can be downloaded from https://www.activestate.com (Note: version 5.8.3 or later is required, the free Standard Distribution is sufficient).
What should I change in order to install the module? php is 5.3.3 withyum install php-devel in place.
PHP.c: In function ‘PHP_set_php_input’:
PHP.c:818: warning: passing argument 2 of ‘Perl_sv_2pv_flags’ from incompatible pointer type
/home/mpapec/.plenv/versions/5.20.0/lib/perl5/5.20.0/x86_64-linux/CORE/proto.h:3931: note: expected ‘STRLEN * const’ but argument is of type ‘int *’
cc -c -I/usr/include/php -I/usr/include/php/main -I/usr/include/php/TSRM -I/usr/include/php/Zend -I/usr/include/php/ext -I/usr/include/php/ext/date/lib -Wall -O2 -DVERSION=\"0.15\" -DXS_VERSION=\"0.15\" -fPIC "-I/home/mpapec/.plenv/versions/5.20.0/lib/perl5/5.20.0/x86_64-linux/CORE" array.c
In file included from /usr/include/php/main/php.h:33,
from /usr/include/php/sapi/embed/php_embed.h:23,
from PHP.h:14,
from array.c:9:
/usr/include/php/main/php_config.h:2417:1: warning: "_GNU_SOURCE" redefined
In file included from /home/mpapec/.plenv/versions/5.20.0/lib/perl5/5.20.0/x86_64-linux/CORE/perl.h:28,
from PHP.h:9,
from array.c:9:
/home/mpapec/.plenv/versions/5.20.0/lib/perl5/5.20.0/x86_64-linux/CORE/config.h:1825:1: warning: this is the location of the previous definition
rm -f blib/arch/auto/PHP/PHP.so
cc -shared -O2 -L/usr/local/lib -fstack-protector PHP.o array.o -o blib/arch/auto/PHP/PHP.so \
\
chmod 755 blib/arch/auto/PHP/PHP.so
"/home/mpapec/.plenv/versions/5.20.0/bin/perl5.20.0" -MExtUtils::Command::MM -e 'cp_nonempty' -- PHP.bs blib/arch/auto/PHP/PHP.bs 644
Manifying 1 pod document
Running Mkbootstrap for PHP ()
chmod 644 "PHP.bs"
PERL_DL_NONLAZY=1 "/home/mpapec/.plenv/versions/5.20.0/bin/perl5.20.0" "-Iblib/lib" "-Iblib/arch" test.pl
1..79
not ok 1 - use_ok PHP
# Failed test 'use_ok PHP'
# at test.pl line 11.
not ok 2 - require PHP;
# Failed test 'require PHP;'
# at test.pl line 18.
# Tried to require 'PHP'.
# Error: Attempt to reload PHP.pm aborted.
# Compilation failed in require at (eval 6) line 2.
not ok 3 - eval
# Failed test 'eval'
# at test.pl line 49.
Module PHP failed to load at blib/lib/PHP.pm line 80.
Module PHP failed to load at blib/lib/PHP.pm line 80.
END failed--call queue aborted at test.pl line 50.
# Looks like you planned 79 tests but ran 3.
# Looks like you failed 3 tests of 3 run.
# Looks like your test exited with 22 just after 3.
make: *** [test_dynamic] Error 22
-> FAIL Installing PHP failed. See /home/mpapec/.cpanm/work/1440522239.12833/build.log for details. Retry with --force to force install it.
(
PHP is kind of fragile. It probably won't work out of the box with your system php installation, and may have trouble with 64-bit or multi-threaded versions of perl.
I have only ever gotten it to work on Linux. The latest version I have tried to use is 5.3.8 (back in 2013), though I remember things going smoothly from 5.2.x to 5.3.8.
I always build php from source, with this configuration:
./configure --enable-embed --with-zlib --with-openssl --with-mysql \
--with-libdir=lib/i386-linux-gnu
--enable-embed is absolutely required, as the pod mentions, to build a PHP interpreter with the SAPI extension, and which then allows perl to manipulate the PHP interpreter through XS code. The other extensions were for other requirements of my project; they may be optional, but I haven't experimented with building the PHP interpreter or the PHP module with any other configuration. The pod also says to never use the --with-apxs argument, which I was never tempted to do anyway.
The build process of the PHP module will look for and require a program called php-config. You may need to hack your $PATH, if only during the build process, so that the PHP module runs the correct php-config. After that the module will know where to look for the rest of your php installation.
I had fun working with this module for a while (writing a Catalyst and then a Mojolicious wrapper around WordPress), but it has fallen into disrepair and disrepute. Share whatever you learn trying to build it and we'll put it in the docs, making this module that much easier to use.
I have a Netgear ReadyNAS NV+ on which I'm trying to do some things using perl.
Since the perl installation from the vendor is compiled without the uselargefiles-flag I've tried to compile/install my own version to run in parallell with the system one.
When I first tried to compile perl on the machine itself I ran into trouble as the system is lacking alot of regular system tools which prevents the Configure-script from running properly.
# ./Configure -des -Dprefix=$HOME/perl-5.16.2
First let's make sure your kit is complete. Checking...
./Configure: line 1986: split: command not found
cat: /root/dload/perl-5.16.2/UU/x??: No such file or directory
Locating common programs...
I don't know where 'comm' is, and my life depends on it.
Go find a public domain implementation or fix your PATH setting!
To get around that I installed and used this cross-compiler to build perl on another machine.
I could not figure out how to cross-compile using the standard configure-script from perl, so I used this.
After that I could just copy my new perl-folder onto the NAS and have thing like the following work:
# ~/sparc-perl-5.16.2/bin/perl -version
This is perl 5, version 16, subversion 2 (v5.16.2) built for sparc-linux
Copyright 1987-2012, Larry Wall
Perl may be copied only under the terms of either the Artistic License or the
GNU General Public License, which may be found in the Perl 5 source kit.
Complete documentation for Perl, including FAQ lists, should be found on
this system using "man perl" or "perldoc perl". If you have access to the
Internet, point your browser at http://www.perl.org/, the Perl Home Page.
# ~/sparc-perl-5.16.2/bin/perl -e 'print "A string!\n";'
A string!
So in some sense I feel I've been successful in my quest to get "my" perl working on the machine.
I get in trouble when I try to install extra modules though.
I've tried installing Image::ExifTool using cpanp (among others), but the installation never finishes.
Today I resorted to trying the installation manually.
wget http://search.cpan.org/CPAN/authors/id/E/EX/EXIFTOOL/Image-ExifTool-9.12.tar.gz
tar xzf Image-ExifTool-9.12.tar.gz
cd Image-ExifTool-9.12
~/sparc-perl-5.16.2/bin/perl Makefile.pl
make
All work as they're supposed to.
when I get to make test however it hangs indefinitely like so:
# make test
PERL_DL_NONLAZY=1 /root/sparc-perl-5.16.2/bin/perl "-MExtUtils::Command::MM" "-e" "test_harness(0, 'blib/lib', 'blib/arch')" t/*.t t/AFCP.t ...........
Looking at the running processes I think these are the ones representing the command:
# ps aux | grep per[l]
root 26187 0.5 0.7 12000 7808 pts/0 S+ 10:31 0:07 /root/sparc-perl-5.16.2/bin/perl -MExtUtils::Command::MM -e test_harness(0, 'blib/lib', 'blib/arch') t/AFCP.t t/AIFF.t t/APE.t t/ASF.t t/BigTIFF.t t/BMP.t t/CanonRaw.t t/Canon.t t/CanonVRD.t t/Casio.t t/DICOM.t t/DjVu.t t/DNG.t t/DV.t t/EXE.t t/ExifTool.t t/FLAC.t t/FlashPix.t t/Flash.t t/Font.t t/FotoStation.t t/FujiFilm.t t/Geotag.t t/GeoTiff.t t/GE.t t/GIF.t t/GIMP.t t/GPS.t t/HTML.t t/InDesign.t t/IPTC.t t/ITC.t t/Jpeg2000.t t/JVC.t t/Kodak.t t/KyoceraRaw.t t/Lang.t t/LNK.t t/M2TS.t t/Matroska.t t/MIE.t t/MIFF.t t/Minolta.t t/MP3.t t/MWG.t t/MXF.t t/Nikon.t t/Olympus.t t/OpenEXR.t t/Panasonic.t t/PDF.t t/Pentax.t t/PGF.t t/PhotoCD.t t/PhotoMechanic.t t/Photoshop.t t/PICT.t t/PNG.t t/PostScript.t t/PPM.t t/PSP.t t/QuickTime.t t/Radiance.t t/Real.t t/Ricoh.t t/RIFF.t t/RTF.t t/Sanyo.t t/Sigma.t t/Sony.t t/Unknown.t t/Vorbis.t t/Writer.t t/XMP.t t/ZIP.t
root 26189 1.7 0.0 0 0 pts/0 Z+ 10:31 0:24 [perl] <defunct>
This is completely consistent with what happens when I try the same installation using cpanp.
I guess I could just run make install and hope that it won't be to much of a problem that the test is unable to run/complete, but that makes me feel a bit uneasy. I'd much rather figure out what is wrong.
I also find it odd that the test hangs the way it does, I could accept that it fails, but that it just keeps running in a non-productive manner feels weird.
Since it is the first time I've built perl from source and the first time I've used a cross-compiler I would not be surprised to learn I've done something very wrong, any help identifying what is very welcome.
What am I doing wrong?
EDIT - compiling on the mahcine:
Figured out how to get the Configure script working natively on the machine.
Will try to build there to see if that will work better. It'll take some time though, the machine is slow.
To be able to run the configure script I needed:
GCC
apt-get install libc6-dev gcc gdb libtag1-dev uuid-dev
coreutils
apt-get install coreutils
-Dcc=gcc
./Configure -des -Dprefix=$HOME/perl-5.16.2 -Dcc=gcc
EDIT2 - compile failed:
Most shortlived success ever:
# make
`sh cflags "optimize='-O2'" perlmini.o` -DPERL_IS_MINIPERL -DPERL_EXTERNAL_GLOB perlmini.c
CCCMD = gcc -DPERL_CORE -c -fno-strict-aliasing -pipe -I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -O2 -Wall
In file included from perl.c:33:
perl.h:699: error: conflicting types for `syscall'
/usr/include/unistd.h:939: error: previous declaration of `syscall'
ake: *** [perlmini.o] Error 1
EDIT3 - testdriving fork:
Tried the following to verify that fork works as it should. I believe that the test is successful, but I've never tried fork before, so please let me know if I read the output wrong.
# cat test_forkwait.px
#!/root/sparc-perl-5.16.2/bin/perl
$pid = fork();
die if $pid < 0;
$SIG{CHLD} = sub { warn "SIGCHLD\n"; };
if($pid) {
$status = waitpid($pid, 0);
warn "wait status: $status\n";
} else {
warn "Child starting\n";
sleep 1;
warn "Child terminating\n";
}
# ./test_forkwait.px
Child starting
Child terminating
SIGCHLD
wait status: 29913
I'm a bit late with this comment, and don't have any ideas about your "make test" hang, but I wanted to point out the Image::ExifTool is pure Perl, so it does not require compilation. The install step is as simple as copying the distribution files to the proper directories.
Phil
Gave up on figuring out why the test fails. Ended up using the module anyways. Seems to work OK with my files.
Okay i am trying to install BIND on solaris 11.
Upon downloading the tar file using wget, then i used extract it out.
After which, when i enter the command :
./configure
i face the following error
checking <<some stuff>>
..
..
..
checking <<some stuff>>
checking whether we are cross compiling... configure: error: in '/home/user/user/downloads/bind9.8.2
configure: error: cannot run C compiled programs.
If you meant to cross compile, use '--host'.
so i used --host in my option, but i cant get it to work,
For example, i entered the below command :
./configure --host=ulinux-uclibc
I received an error saying
checking host system type... Invalid configuration 'ulinux-uclibc': machine 'ulinux' not recognized
Thanks! :)
Why not just install the bind package from your configured solaris publisher?
# pkg install network/dns/bind service/network/dns/bind
should get you exactly what you need. That will get you ISC Bind v9.6.3 if you use pkg.oracle.com/solaris/release publisher.