View previous topic :: View next topic |
Author |
Message |
mani001 Guru
Joined: 04 Dec 2004 Posts: 485 Location: Oleiros
|
Posted: Tue Jun 02, 2009 9:21 am Post subject: koffice 1.6.3 won't emerge any more |
|
|
is it just me or some recent update (not sure if it's gcc 4.3 or the new kernel 2.6.29) messed up the stable version of Koffice in portage?
I can't emerge neither libpqxx nor kword nor kexi any more
Greetings. |
|
Back to top |
|
|
hoacker Guru
Joined: 04 Aug 2007 Posts: 505 Location: Bürstadt, Germany
|
Posted: Tue Jun 02, 2009 10:21 am Post subject: |
|
|
Error messages would help to identify your problem. Anyway, have a look at this bug. |
|
Back to top |
|
|
mani001 Guru
Joined: 04 Dec 2004 Posts: 485 Location: Oleiros
|
Posted: Tue Jun 02, 2009 10:33 am Post subject: |
|
|
Thanks for the reply
I see your point It goes something like this (this is for libpqxx, the first package pulled in when you try to emerge kword)
Code: |
binarystring.cxx: In function 'std::string pqxx::escape_binary(const unsigned char*)':
binarystring.cxx:168: error: 'strlen' was not declared in this scope
binarystring.cxx: In function 'std::string pqxx::escape_binary(const char*)':
binarystring.cxx:178: error: 'strlen' was not declared in this scope
make[1]: *** [binarystring.lo] Error 1
make[1]: *** Waiting for unfinished jobs....
x86_64-pc-linux-gnu-g++ -DHAVE_CONFIG_H -I../include -I../include -I/usr/include/postgresql/libpq-4 -O2 -pipe -MT cachedresult.lo -MD -MP -MF .deps/cachedresult.Tpo -c cachedresult.cxx -o cachedresult.o >/dev/null 2>&1
connection_base.cxx: In member function 'void pqxx::connection_base::process_notice(const char*)':
connection_base.cxx:470: error: 'strlen' was not declared in this scope
connection_base.cxx:490: error: 'strcpy' was not declared in this scope
connection_base.cxx:494: error: 'memcpy' was not declared in this scope
connection_base.cxx:499: error: 'memcpy' was not declared in this scope
make[1]: *** [connection_base.lo] Error 1
make[1]: Leaving directory `/var/tmp/portage/dev-libs/libpqxx-2.6.8/work/libpqxx-2.6.8/src'
make: *** [all-recursive] Error 1
*
* ERROR: dev-libs/libpqxx-2.6.8 failed.
* Call stack:
* ebuild.sh, line 48: Called src_compile
* environment, line 2227: Called die
* The specific snippet of code:
* emake || die "emake failed"
* The die message:
* emake failed
*
* If you need support, post the topmost build error, and the call stack if relevant.
* A complete build log is located at '/var/tmp/portage/dev-libs/libpqxx-2.6.8/temp/build.log'.
* The ebuild environment file is located at '/var/tmp/portage/dev-libs/libpqxx-2.6.8/temp/environment'.
*
|
I'm guessing it's probably the same issue that causes the bug you mention...but I don't know for sure..
Greetings!! |
|
Back to top |
|
|
hoacker Guru
Joined: 04 Aug 2007 Posts: 505 Location: Bürstadt, Germany
|
Posted: Tue Jun 02, 2009 10:59 am Post subject: |
|
|
dev-libs/libpqxx-2.6.8 does not compile with gcc-4.3, see this bug. The bug is fixed in version 2.6.9 in the portage tree, but it seems devs are not going to patch 2.6.8. A patch for 2.6.8 is available in this bug.
I'm quite sure that the koffice errors are not related to libpqxx...
Last edited by hoacker on Tue Jun 02, 2009 4:01 pm; edited 1 time in total |
|
Back to top |
|
|
mani001 Guru
Joined: 04 Dec 2004 Posts: 485 Location: Oleiros
|
Posted: Tue Jun 02, 2009 11:26 am Post subject: |
|
|
Yes, you are right: switching to gcc 4.1 lipbqxx compiles just fine. Now the problem is kexi (which for some reason is pulled by kword):
Code: |
pqxxcursor.cpp: In member function 'virtual const char** KexiDB::pqxxSqlCursor::rowData() const':
pqxxcursor.cpp:262: error: 'malloc' was not declared in this scope
pqxxcursor.cpp:264: warning: comparison between signed and unsigned integer expressions
make[5]: *** [pqxxcursor.lo] Error 1
|
|
|
Back to top |
|
|
hoacker Guru
Joined: 04 Aug 2007 Posts: 505 Location: Bürstadt, Germany
|
Posted: Tue Jun 02, 2009 2:49 pm Post subject: |
|
|
Looks like this bug. |
|
Back to top |
|
|
rickj Guru
Joined: 06 Feb 2003 Posts: 427 Location: Calgary, Alberta, Canada
|
Posted: Tue Jun 02, 2009 3:18 pm Post subject: |
|
|
It's not dependant on KDE4: In KDE3.5.9 I get:
Code: | binarystring.cxx: In function 'std::string pqxx::escape_binary(const unsigned char*)':
binarystring.cxx:168: error: 'strlen' was not declared in this scope
binarystring.cxx: In function 'std::string pqxx::escape_binary(const char*)':
binarystring.cxx:178: error: 'strlen' was not declared in this scope
make[1]: *** [binarystring.lo] Error 1
make[1]: *** Waiting for unfinished jobs....
i686-pc-linux-gnu-g++ -DHAVE_CONFIG_H -I../include -I../include -I/usr/include/postgresql/libpq-4 -march=athlon -O2 -pipe -fomit-frame-pointer -MT cachedresult.lo -MD -MP -MF .deps/cachedresult.Tpo -c cachedresult.cxx -o cachedresult.o >/dev/null 2>&1
make[1]: Leaving directory `/var/tmp/portage/dev-libs/libpqxx-2.6.8/work/libpqxx-2.6.8/src'
make: *** [all-recursive] Error 1
*
* ERROR: dev-libs/libpqxx-2.6.8 failed.
* Call stack:
* ebuild.sh, line 48: Called src_compile
* environment, line 2184: Called die
* The specific snippet of code:
* emake || die "emake failed"
* The die message:
* emake failed
*
|
under gcc 4.3.2-r3, and my whole system is from the stable branch. |
|
Back to top |
|
|
hoacker Guru
Joined: 04 Aug 2007 Posts: 505 Location: Bürstadt, Germany
|
Posted: Tue Jun 02, 2009 4:00 pm Post subject: |
|
|
rickj wrote: | It's not dependant on KDE4: In KDE3.5.9 I get:
[...]
under gcc 4.3.2-r3, and my whole system is from the stable branch. |
Well, yes. This is because of the bug mentioned earlier in this thread. Do you need assistance solving this issue? |
|
Back to top |
|
|
rickj Guru
Joined: 06 Feb 2003 Posts: 427 Location: Calgary, Alberta, Canada
|
Posted: Tue Jun 02, 2009 4:34 pm Post subject: |
|
|
It seems that in my case, only showimg with postgresql needs the older version.
Code: | > equery depends libpqxx
[ Searching for packages depending on libpqxx... ]
media-gfx/showimg-0.9.5 (postgres? <dev-libs/libpqxx-2.6.9)
|
I'd expect that swapping back to gcc 4.1.2 would enable me to compile these. I'm not sure to what extent one can mix and match modules compiled under gcc 4.1.2 and 4.3.2 |
|
Back to top |
|
|
|