Re: 9.0-CURRENT r220692 cc1: internal compiler error: Segmentation fault: 11
El día Tuesday, April 26, 2011 a las 08:43:54PM -0400, Arnaud Lacombe escribió: Hi, On Tue, Apr 26, 2011 at 5:19 PM, Matthias Apitz g...@unixarea.de wrote: El día Tuesday, April 26, 2011 a las 01:51:55PM -0400, Arnaud Lacombe escribió: Then you can go to gcc.gnu.org/bugzilla and file a PR. Thanks for your hints concerning the GCC problem; I will collect the data and file a bug report; If it is the gcc in the FreeBSD tree, please do not bug upstream. gcc 4.2.1 is no longer supported/maintained. The least you can do is to try a newer, maintained, gcc and see if the issue persist. Hi, It is the normal gcc of the FreeBSD used to compile ports: # cc -v Using built-in specs. Target: i386-undermydesk-freebsd Configured with: FreeBSD/i386 system compiler Thread model: posix gcc version 4.2.1 20070719 [FreeBSD] In /usr/ports/lang there are gcc46 and gcc47 ports, should I build one of them and use it to compile ports/mail/evolution-exchange? Which one would be the best option in HEAD? Thanks 4.6.0 is the latest stable release, might still have regression, but at least, you can report them :). I installed /usr/ports/lang/gcc46 and compiled evolution-exchange-2.32.3 without SIGSEGV; we can close this thread; thx matthias -- Matthias Apitz t +49-89-61308 351 - f +49-89-61308 399 - m +49-170-4527211 e g...@unixarea.de - w http://www.unixarea.de/ ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: 9.0-CURRENT r220692 cc1: internal compiler error: Segmentation fault: 11
Hello.
Don't know is this related.
I've got rather fresh 9.0-CURRENT (checked out few days ago) built with clang.
And I use clang as the system compiler, but ruby fails to build with clang. So
I've tried gcc. But with gcc I've got this:
..
configure:3211: checking whether the C compiler works
configure:3233: cc -I/usr/include -O2 -pipe -march=native -fno-strict-aliasing
-I/usr/include -rpath=/usr/lib:/usr/local/lib -pthread conft
est.c -L/usr/lib -rpath=/usr/lib:/usr/local/lib -pthread 5
Segmentation fault (core dumped)
configure:3237: $? = 139
configure:3275: result: no
configure: failed program was:
| /* confdefs.h */
| #define PACKAGE_NAME
| #define PACKAGE_TARNAME
| #define PACKAGE_VERSION
| #define PACKAGE_STRING
| #define PACKAGE_BUGREPORT
| #define PACKAGE_URL
| /* end confdefs.h. */
|
| int
| main ()
| {
|
| ;
| return 0;
| }
configure:3280: error: in
`/mnt/portworkdir/usr/ports/lang/ruby18/work/ruby-1.8.7-p302':
configure:3283: error: C compiler cannot create executables
..
As far as I remeber, all was ok when I had base gcc build by gcc not clang. But
this could be unrelated.
Regards.
26.04.2011, 12:04, Matthias Apitz g...@unixarea.de:
Hello,
I'm trying to compile /usr/ports/mail/evolution-exchange/ and the gcc
crashes with:
[root@vm-9Current /usr/ports/mail/evolution-exchange]# LANG=C make
=== Building for evolution-exchange-2.32.1_1
gmake all-recursive
gmake[1]: Entering directory
`/usr/ports/mail/evolution-exchange/work/evolution-exchange-2.32.1'
Making all in server
gmake[2]: Entering directory
`/usr/ports/mail/evolution-exchange/work/evolution-exchange-2.32.1/server'
Making all in xntlm
gmake[3]: Entering directory
`/usr/ports/mail/evolution-exchange/work/evolution-exchange-2.32.1/server/xntlm'
CC libxntlm_la-xntlm.lo
cc1: internal compiler error: Segmentation fault: 11
Some notes about this:
- the system runs in a VMworkstation 7.x
- it has already compliled kernel, userland and ~1000 ports without any
crash, i.e. it is *not* the typical hardware related crash;
- the above mentioned version evolution-exchange-2.32.1_1 is a fake, in
real it is compiling the original evolution-exchange-2.32.3 sources;
- it is fully reproduceable
What next?
(David, should it be posted to evolut...@gnome.org as well?)
matthias
--
Matthias Apitz
t +49-89-61308 351 - f +49-89-61308 399 - m +49-170-4527211
e g...@unixarea.de; - w http://www.unixarea.de/
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
___
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: 9.0-CURRENT r220692 cc1: internal compiler error: Segmentation fault: 11
El día Tuesday, April 26, 2011 a las 10:22:21AM +0100, David Woodhouse escribió: crash, i.e. it is *not* the typical hardware related crash; - the above mentioned version evolution-exchange-2.32.1_1 is a fake, in real it is compiling the original evolution-exchange-2.32.3 sources; - it is fully reproduceable What next? (David, should it be posted to evolut...@gnome.org as well?) Looks like a compiler bug, so probably not. What version of GCC? You probably want to file a PR at gcc.gnu.org. Then you can go to gcc.gnu.org/bugzilla and file a PR. Thanks for your hints concerning the GCC problem; I will collect the data and file a bug report; In the meantime, however, if you don't need evolution-exchange then you can just ignore this problem and get on with the other things you were trying to test? I need the evolution-exchange connector; can I live with 2.32.1 while evolution and its dataserver are 2.32.3? If not, I must go back fully to 2.32.1 after the debugging of the calendar stuff; matthias -- Matthias Apitz t +49-89-61308 351 - f +49-89-61308 399 - m +49-170-4527211 e g...@unixarea.de - w http://www.unixarea.de/ ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: 9.0-CURRENT r220692 cc1: internal compiler error: Segmentation fault: 11
On Tue, 2011-04-26 at 10:04 +0200, Matthias Apitz wrote: Hello, I'm trying to compile /usr/ports/mail/evolution-exchange/ and the gcc crashes with: [root@vm-9Current /usr/ports/mail/evolution-exchange]# LANG=C make === Building for evolution-exchange-2.32.1_1 gmake all-recursive gmake[1]: Entering directory `/usr/ports/mail/evolution-exchange/work/evolution-exchange-2.32.1' Making all in server gmake[2]: Entering directory `/usr/ports/mail/evolution-exchange/work/evolution-exchange-2.32.1/server' Making all in xntlm gmake[3]: Entering directory `/usr/ports/mail/evolution-exchange/work/evolution-exchange-2.32.1/server/xntlm' CC libxntlm_la-xntlm.lo cc1: internal compiler error: Segmentation fault: 11 Some notes about this: - the system runs in a VMworkstation 7.x - it has already compliled kernel, userland and ~1000 ports without any crash, i.e. it is *not* the typical hardware related crash; - the above mentioned version evolution-exchange-2.32.1_1 is a fake, in real it is compiling the original evolution-exchange-2.32.3 sources; - it is fully reproduceable What next? (David, should it be posted to evolut...@gnome.org as well?) Looks like a compiler bug, so probably not. What version of GCC? You probably want to file a PR at gcc.gnu.org. Run 'make V=1' so you can see the compiler command line being used, and then you can run it manually from the command line. Then try to cut the test case down as much as possible. Your task is to delete as *much* code as you can from xntlm.c, and still reproduce the problem. Try removing (or comment out the body of) one function at a time, working backwards from the end of the file. Ideally you'll be left with just a *single* function that triggers the compiler bug — and then you can remove lines of code from *that* to make it simpler, too. It doesn't matter that it will no longer *work*, as long as it still makes the compiler barf (and is still valid C code). The smaller your test case is, the better. Finally, run the compiler with '-dD -save-temps' added to the command line. It'll save its intermediate files, including a '.i' file that is completely preprocessed source. It contains everything it needed from the header files on your system, so it can be compiled *anywhere*. Then you can go to gcc.gnu.org/bugzilla and file a PR. In the meantime, however, if you don't need evolution-exchange then you can just ignore this problem and get on with the other things you were trying to test? -- David WoodhouseOpen Source Technology Centre david.woodho...@intel.com Intel Corporation ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: 9.0-CURRENT r220692 cc1: internal compiler error: Segmentation fault: 11
On Tue, 2011-04-26 at 11:51 +0200, Matthias Apitz wrote: I need the evolution-exchange connector; That's for Exchange 2003, right? If you were on a newer version of Exchange perhaps you'd be able to use the new Evolution-EWS connector. can I live with 2.32.1 while evolution and its dataserver are 2.32.3? That should work. It may have entirely *artificial* dependencies on a newer version of e-d-s, but I don't think there's any good reason for such dependencies so just ignoring them and forcing it to install/run anyway (by changing a '3' to a '1' in configure.ac) should be fine. But this code hasn't changed, so if you failed to build evo-exchange 2.32.3 I strongly suspect that the same compiler bug will prevent you from building 2.32.1. You could try building evolution-exchange (or just xntlm.c) with -O0; that often helps to avoid triggering compiler bugs. If not, I must go back fully to 2.32.1 after the debugging of the calendar stuff; Well, the FreeBSD ports should be updated to 2.32.3 anyway. -- David WoodhouseOpen Source Technology Centre david.woodho...@intel.com Intel Corporation ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: 9.0-CURRENT r220692 cc1: internal compiler error: Segmentation fault: 11
Hi, On Tue, Apr 26, 2011 at 5:51 AM, Matthias Apitz g...@unixarea.de wrote: El día Tuesday, April 26, 2011 a las 10:22:21AM +0100, David Woodhouse escribió: crash, i.e. it is *not* the typical hardware related crash; - the above mentioned version evolution-exchange-2.32.1_1 is a fake, in real it is compiling the original evolution-exchange-2.32.3 sources; - it is fully reproduceable What next? (David, should it be posted to evolut...@gnome.org as well?) Looks like a compiler bug, so probably not. What version of GCC? You probably want to file a PR at gcc.gnu.org. Then you can go to gcc.gnu.org/bugzilla and file a PR. Thanks for your hints concerning the GCC problem; I will collect the data and file a bug report; If it is the gcc in the FreeBSD tree, please do not bug upstream. gcc 4.2.1 is no longer supported/maintained. The least you can do is to try a newer, maintained, gcc and see if the issue persist. - Arnaud In the meantime, however, if you don't need evolution-exchange then you can just ignore this problem and get on with the other things you were trying to test? I need the evolution-exchange connector; can I live with 2.32.1 while evolution and its dataserver are 2.32.3? If not, I must go back fully to 2.32.1 after the debugging of the calendar stuff; matthias -- Matthias Apitz t +49-89-61308 351 - f +49-89-61308 399 - m +49-170-4527211 e g...@unixarea.de - w http://www.unixarea.de/ ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: 9.0-CURRENT r220692 cc1: internal compiler error: Segmentation fault: 11
El día Tuesday, April 26, 2011 a las 01:51:55PM -0400, Arnaud Lacombe escribió: Then you can go to gcc.gnu.org/bugzilla and file a PR. Thanks for your hints concerning the GCC problem; I will collect the data and file a bug report; If it is the gcc in the FreeBSD tree, please do not bug upstream. gcc 4.2.1 is no longer supported/maintained. The least you can do is to try a newer, maintained, gcc and see if the issue persist. Hi, It is the normal gcc of the FreeBSD used to compile ports: # cc -v Using built-in specs. Target: i386-undermydesk-freebsd Configured with: FreeBSD/i386 system compiler Thread model: posix gcc version 4.2.1 20070719 [FreeBSD] In /usr/ports/lang there are gcc46 and gcc47 ports, should I build one of them and use it to compile ports/mail/evolution-exchange? Which one would be the best option in HEAD? Thanks matthias -- Matthias Apitz t +49-89-61308 351 - f +49-89-61308 399 - m +49-170-4527211 e g...@unixarea.de - w http://www.unixarea.de/ ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: 9.0-CURRENT r220692 cc1: internal compiler error: Segmentation fault: 11
Hi, On Tue, Apr 26, 2011 at 5:19 PM, Matthias Apitz g...@unixarea.de wrote: El día Tuesday, April 26, 2011 a las 01:51:55PM -0400, Arnaud Lacombe escribió: Then you can go to gcc.gnu.org/bugzilla and file a PR. Thanks for your hints concerning the GCC problem; I will collect the data and file a bug report; If it is the gcc in the FreeBSD tree, please do not bug upstream. gcc 4.2.1 is no longer supported/maintained. The least you can do is to try a newer, maintained, gcc and see if the issue persist. Hi, It is the normal gcc of the FreeBSD used to compile ports: # cc -v Using built-in specs. Target: i386-undermydesk-freebsd Configured with: FreeBSD/i386 system compiler Thread model: posix gcc version 4.2.1 20070719 [FreeBSD] In /usr/ports/lang there are gcc46 and gcc47 ports, should I build one of them and use it to compile ports/mail/evolution-exchange? Which one would be the best option in HEAD? Thanks 4.6.0 is the latest stable release, might still have regression, but at least, you can report them :). - Arnaud matthias -- Matthias Apitz t +49-89-61308 351 - f +49-89-61308 399 - m +49-170-4527211 e g...@unixarea.de - w http://www.unixarea.de/ ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
