Обсуждение: BUG #4068: Failed to compile 8.3.1
The following bug has been logged online: Bug reference: 4068 Logged by: Ronald Kuczek Email address: kuczek@kuczek.pl PostgreSQL version: 8.3.1 Operating system: Open SuSE 10.3 Description: Failed to compile 8.3.1 Details: Compilation from sources failed with error message: make[3]: Entering directory `/usr/src/postgres/postgresql-8.3.1/src/backend/utils/mb/conversion_procs/ut f8_and_shift_jis_2004' make[3]: *** No rule to make target `utf8_and_shift_jis_2004.o', needed by `libutf8_and_shift_jis_2004.so.0.0'. Stop. make[3]: Leaving directory `/usr/src/postgres/postgresql-8.3.1/src/backend/utils/mb/conversion_procs/ut f8_and_shift_jis_2004' make[2]: *** [all] Error 2 make[2]: Leaving directory `/usr/src/postgres/postgresql-8.3.1/src/backend/utils/mb/conversion_procs' make[1]: *** [all] Error 2 make[1]: Leaving directory `/usr/src/postgres/postgresql-8.3.1/src' make: *** [all] Error 2 Best regards Rony
Ronald Kuczek napsal(a):
> The following bug has been logged online:
>
> Bug reference: 4068
> Logged by: Ronald Kuczek
> Email address: kuczek@kuczek.pl
> PostgreSQL version: 8.3.1
> Operating system: Open SuSE 10.3
> Description: Failed to compile 8.3.1
> Details:
>
> Compilation from sources failed with error message:
> make[3]: Entering directory
> `/usr/src/postgres/postgresql-8.3.1/src/backend/utils/mb/conversion_procs/ut
> f8_and_shift_jis_2004'
> make[3]: *** No rule to make target `utf8_and_shift_jis_2004.o', needed by
> `libutf8_and_shift_jis_2004.so.0.0'. Stop.
> make[3]: Leaving directory
> `/usr/src/postgres/postgresql-8.3.1/src/backend/utils/mb/conversion_procs/ut
> f8_and_shift_jis_2004'
> make[2]: *** [all] Error 2
> make[2]: Leaving directory
> `/usr/src/postgres/postgresql-8.3.1/src/backend/utils/mb/conversion_procs'
> make[1]: *** [all] Error 2
> make[1]: Leaving directory `/usr/src/postgres/postgresql-8.3.1/src'
> make: *** [all] Error 2
Please, could you send a configure setup and gcc version could be useful as well.
Zdenek
Zdenek Kotala wrote: > > Please, could you send a configure setup and gcc version could be > useful as well. > > Zdenek > Sure , see config.log file. Rony
Вложения
Zdenek Kotala <Zdenek.Kotala@Sun.COM> writes:
> Ronald Kuczek napsal(a):
>> make[3]: Entering directory
>> `/usr/src/postgres/postgresql-8.3.1/src/backend/utils/mb/conversion_procs/ut
>> f8_and_shift_jis_2004'
>> make[3]: *** No rule to make target `utf8_and_shift_jis_2004.o', needed by
>> `libutf8_and_shift_jis_2004.so.0.0'. Stop.
> Please, could you send a configure setup and gcc version could be useful as well.
I think the gmake version would be more relevant. Given that this works
for everyone else, I have to think the OP's copy of gmake is borked ...
regards, tom lane
Tom Lane wrote: > Zdenek Kotala <Zdenek.Kotala@Sun.COM> writes: > >> Ronald Kuczek napsal(a): >> >>> make[3]: Entering directory >>> `/usr/src/postgres/postgresql-8.3.1/src/backend/utils/mb/conversion_procs/ut >>> f8_and_shift_jis_2004' >>> make[3]: *** No rule to make target `utf8_and_shift_jis_2004.o', needed by >>> `libutf8_and_shift_jis_2004.so.0.0'. Stop. >>> > > >> Please, could you send a configure setup and gcc version could be useful as well. >> > > I think the gmake version would be more relevant. Given that this works > for everyone else, I have to think the OP's copy of gmake is borked ... > > regards, tom lane > Hi Zdenek, hi Tom, Thank you for the answer. Gmake: postgres@unix:/> gmake -v GNU Make 3.81 Copyright (C) 2006 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. This program built for i686-pc-linux-gnu BTW. I have compiled previous version (8.2.0) on the same machine without any problems. Regards Rony
Tom Lane wrote: > Zdenek Kotala <Zdenek.Kotala@Sun.COM> writes: > > Ronald Kuczek napsal(a): > >> make[3]: Entering directory > >> `/usr/src/postgres/postgresql-8.3.1/src/backend/utils/mb/conversion_procs/ut > >> f8_and_shift_jis_2004' > >> make[3]: *** No rule to make target `utf8_and_shift_jis_2004.o', needed by > >> `libutf8_and_shift_jis_2004.so.0.0'. Stop. > > > Please, could you send a configure setup and gcc version could be useful as well. > > I think the gmake version would be more relevant. Given that this works > for everyone else, I have to think the OP's copy of gmake is borked ... Actually I've seen this reported before: http://archives.postgresql.org/pgsql-es-ayuda/2008-03/msg00335.php This guy has always compiled from sources, so he was surprised that it didn't work with 8.3.1. He says he got the tarball from the main FTP site. He found out that the utf8_and_shift_jis_2004.c file was missing, so he copied it from the SVN repo: http://archives.postgresql.org/pgsql-es-ayuda/2008-03/msg00339.php What we haven't quite figured out is why the file wasn't present on the tarball he downloaded. Is it possible that a borked tarball was uploaded somewhere? Hmm, further down the thread, he says several people have seen this problem: http://archives.postgresql.org/pgsql-es-ayuda/2008-03/msg00353.php -- Alvaro Herrera http://www.CommandPrompt.com/ The PostgreSQL Company - Command Prompt, Inc.
Alvaro Herrera wrote: > Hmm, further down the thread, he says several people have seen this > problem: Yes, it is possible to find more reports about it: http://www.thescripts.com/forum/thread782006.html http://www.sql.ru/forum/actualthread.aspx?tid=535381 -- Alvaro Herrera Valdivia, Chile Geotag: -39,815 -73,257 "La rebeldía es la virtud original del hombre" (Arthur Schopenhauer)
Alvaro Herrera wrote: > Alvaro Herrera wrote: > > > Hmm, further down the thread, he says several people have seen this > > problem: > > Yes, it is possible to find more reports about it: > > http://www.thescripts.com/forum/thread782006.html > http://www.sql.ru/forum/actualthread.aspx?tid=535381 Yes, I saw this. As I remember that failured file had the longest path of all all file names, and in fact I have seen old tar complain about it. Did we have an original tarball that was missing that file too? I can't remember but I remember an aborted tarball with a missing file. -- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB http://enterprisedb.com + If your life is a hard drive, Christ can be your backup. +
Ronald Kuczek wrote: > Hi Alvaro, > Thanks for the answer. It seems to be something else: > postgres@unix:/usr/src/postgres/backend/utils/mb/conversion_procs/euc_jis_2004_and_shift_jis_2004> > ls -al > total 16 > drwxr-xr-x 2 postgres users 4096 Mar 27 19:58 . > drwxr-xr-x 3 postgres users 4096 Mar 27 19:58 .. > -rw-r--r-- 1 postgres users 6978 Jan 1 21:31 > euc_jis_2004_and_shift_jis_2004.c > > As you can see the file is present in 8.3.1 source tree or have I > overlook something ? Well, there should be a Makefile in there at the very least. I'll assume you'll just omitted pasting that part. What happens if you cd into that directory and run "make"? -- Alvaro Herrera http://www.CommandPrompt.com/ PostgreSQL Replication, Consulting, Custom Development, 24x7 support
Alvaro Herrera <alvherre@commandprompt.com> writes:
> What we haven't quite figured out is why the file wasn't present on the
> tarball he downloaded. Is it possible that a borked tarball was
> uploaded somewhere?
I seem to recall we found that some tar versions have a limit on total
pathname length? The tarball was actually OK but it got extracted
incorrectly. However the OP seems to be running a reasonably current
Linux so it's hard to believe he's got that problem.
regards, tom lane
Alvaro Herrera wrote: > Alvaro Herrera wrote: > > >> Hmm, further down the thread, he says several people have seen this >> problem: >> > > Yes, it is possible to find more reports about it: > > http://www.thescripts.com/forum/thread782006.html > http://www.sql.ru/forum/actualthread.aspx?tid=535381 > Hi Alvaro, Thanks for the answer. It seems to be something else: postgres@unix:/usr/src/postgres/backend/utils/mb/conversion_procs/euc_jis_2004_and_shift_jis_2004> ls -al total 16 drwxr-xr-x 2 postgres users 4096 Mar 27 19:58 . drwxr-xr-x 3 postgres users 4096 Mar 27 19:58 .. -rw-r--r-- 1 postgres users 6978 Jan 1 21:31 euc_jis_2004_and_shift_jis_2004.c As you can see the file is present in 8.3.1 source tree or have I overlook something ? Regards Rony
Alvaro Herrera wrote: > > Well, there should be a Makefile in there at the very least. I'll > assume you'll just omitted pasting that part. What happens if you cd > into that directory and run "make"? > > Makefile is not present and my unit says : make: *** No targets specified and no makefile found. Stop. As you can see error message differs with this from compilation in source dir it was: make[3]: *** No rule to make target `utf8_and_shift_jis_2004.o', needed by `libutf8_and_shift_jis_2004.so.0.0'. Stop. Gmake seems to be using some Makefile, but there are some entries missing. Any ideas ? Regards Rony
Tom Lane wrote: > > I seem to recall we found that some tar versions have a limit on total > pathname length? The tarball was actually OK but it got extracted > incorrectly. However the OP seems to be running a reasonably current > Linux so it's hard to believe he's got that problem. > > Hi Tom, Thank you for a hint. It was a Midnight Commander bug. MC uncompress sources into three separate directories. I have uncompressed the 8.3.1 archive with bzip -d and tar -xvf (one directory !) and compiled succesfully. It seems to be a Midnight Commander not PostgreSQL problem. Regards Rony
Ronald Kuczek <kuczek@kuczek.pl> writes:
> Thanks for the answer. It seems to be something else:
> postgres@unix:/usr/src/postgres/backend/utils/mb/conversion_procs/euc_jis_2004_and_shift_jis_2004>
Is that actually a correct transcription of the path? Because it's
bogus: it should be something ending in
src/backend/utils/mb/conversion_procs/euc_jis_2004_and_shift_jis_2004
If you've rearranged the top-level directory structure of the
distribution, that could explain a lot --- like makefiles failing to
locate Makefile.global.
regards, tom lane