[torqueusers] problem with Altix compile

Garrick Staples garrick at usc.edu
Tue Jan 3 13:19:02 MST 2006


On Tue, Jan 03, 2006 at 02:13:10PM -0600, Kevin McMahon alleged:
>    From your error message, this looks to be a problem with the
> directory log_directory.
> 
>  >   pbs_log.c:114: storage size of `log_directory' isn't known
> 
> 
>    Perhaps there is an issue running as root on a crossmounted 
> file system?  Usually, root is not allowed to exercise root privileged
> file operations on such a file system.

If you look at that line, you'll see that log_directory is a char array:
  char log_directory[_POSIX_PATH_MAX/2];

The storage size isn't known because his system C headers are messed up.

Note below that /usr/include/linux/errno.h seems to be missing.

 
> > Jeroen,
> > 	Well so far I still haven't gotten this to work. I tried the other
> > example of switching the os_defines.h and I get an error right away:
> > 
> > Making all in Libattr
> > make[3]: Entering directory `/usr/local/src/Original/torque-2.0.0p4-
> > altix/src/lib/Libattr'make[4]: Entering directory
> > `/usr/local/src/Original/torque-2.0.0p4-altix/src/lib/Libattr'gcc -g -
> > I../../../src/include -I../../../src/include -DHAVE_CONFIG_H -c
> > uLTostr.c
> > In file included from /usr/include/errno.h:36,
> >                  from ../../../src/include/Long_.h:85,
> >                  from uLTostr.c:84:
> > /usr/include/bits/errno.h:25:26: linux/errno.h: No such file or
> > directory
> > uLTostr.c: In function `uLTostr':
> > uLTostr.c:116: `EDOM' undeclared (first use in this function)
> > uLTostr.c:116: (Each undeclared identifier is reported only once
> > uLTostr.c:116: for each function it appears in.)
> > make[4]: *** [uLTostr.o] Error 1
> > make[4]: Leaving directory `/usr/local/src/Original/torque-2.0.0p4-
> > altix/src/lib/Libattr'
> > make[3]: *** [all] Error 2
> > make[3]: Leaving directory `/usr/local/src/Original/torque-2.0.0p4-
> > altix/src/lib/Libattr'
> > make[2]: *** [all] Error 2
> > make[2]: Leaving directory `/usr/local/src/Original/torque-2.0.0p4-
> > altix/src/lib'
> > make[1]: *** [all] Error 2
> > make[1]: Leaving directory `/usr/local/src/Original/torque-2.0.0p4-
> > altix/src'
> > make: *** [all] Error 2
> > [root at kilburn torque-2.0.0p4-altix]#
> > 
> > 
> > When I look for the packages mentioned it looks like I do have them:
> > 
> > [root at kilburn torque-2.0.0p4-altix]# rpm -qa | grep gcc
> > gcc-ssa-3.5ssa-0.20030801.48
> > libgcc-3.2.3-47
> > libgcc-ssa-3.5ssa-0.20030801.48
> > libgcc-3.2.3-47
> > compat-gcc-g77-7.3-2.96.128
> > gcc-gnat-3.2.3-47
> > compat-gcc-c++-7.3-2.96.128
> > gcc-objc-3.2.3-47
> > gcc-c++-ssa-3.5ssa-0.20030801.48
> > compat-gcc-c++-7.3-2.96.128
> > gcc-g77-ssa-3.5ssa-0.20030801.48
> > gcc-java-ssa-3.5ssa-0.20030801.48
> > compat-gcc-objc-7.3-2.96.128
> > gcc-objc-ssa-3.5ssa-0.20030801.48
> > gcc-c++-3.2.3-47
> > gcc-java-3.2.3-47
> > gcc-3.2.3-47
> > compat-gcc-7.3-2.96.128
> > gcc-g77-3.2.3-47
> > [root at kilburn torque-2.0.0p4-altix]# rpm -qa | grep glibc
> > glibc-devel-2.3.2-95.33sgi306r1
> > glibc-profile-2.3.2-95.33sgi306r1
> > glibc-headers-2.3.2-95.33sgi306r1
> > glibc-common-2.3.2-95.33sgi306r1
> > glibc-2.3.2-95.33sgi306r1
> > glibc-utils-2.3.2-95.33sgi306r1
> > compat-glibc-7.x-2.2.4.32.6sgi306r1
> > glibc-2.3.2-95.30
> > glibc-debug-2.3.2-95.33sgi306r1
> > [root at kilburn torque-2.0.0p4-altix]#
> > 
> > [root at kilburn torque-2.0.0p4-altix]# uname -a
> > Linux kilburn 2.4.21-sgi306r6 #1 SMP Sat Jun 18 14:43:35 PDT 2005 ia64
> > ia64 ia64 GNU/Linux[root at kilburn torque-2.0.0p4-altix]#
> > 
> > 
> > So I'm at a loss at what I should look at now. The "talk to the SGI
> > people" suggestion will be my next step. The only problem is that I have
> > compiled all the other programs we are using here without issues. Since
> > I have support with SGI I will contact them and see what they can come
> > up with. Thanks everyone for the help. I'll post any results I come up
> > with. Sincerely,
> > 
> > -Steve
> > 
> > 
> > On Mon, 2006-01-02 at 21:18 +1100, Jeroen van den Muyzenberg wrote:
> > > Hi Steve,
> > > 
> > > If this is truly happening "out of the box", then you should be talking
> > > to SGI rather than this forum.
> > > 
> > > Jeroen
> > > 
> > > 
> > > 
> > > On Thu, 29 Dec 2005, Walter de Jong wrote:
> > > 
> > > > Hi,
> > > >
> > > > The short example program does compile on our Altix.
> > > > It runs ProPack 3 SP6, limits.h is in the package
> > > > glibc-headers-2.3.2-95.33sgi306r1
> > > >
> > > > So for PP4, look for a package named glibc-headers.
> > > >
> > > > (By the way, it gives _POSIX_PATH_MAX is 256, where I expected it to be
> > > > 1024 or more. Ah well.)
> > > >
> > > > Greets,
> > > >
> > > >    --Walter
> > > >
> > > >
> > > >
> > > > Steve Young wrote:
> > > >>  Hello, 	I hope all had a nice holiday. Anyhow, I still seem to have
> > > >>  this issue
> > > >>  with my Altix's. I have 3 of them out of the box from SGI. They all have
> > > >>  the same problem. I tried compiling the below program and it also does
> > > >>  the same thing. So it looks like I am missing some package in order to
> > > >>  compile this? Any idea's what I should look for? Thanks,
> > > >>
> > > >>  -Steve
> > > >>
> > > >>
> > > >>  On Thu, 2005-12-22 at 09:57 -0800, Garrick Staples wrote:
> > > >> 
> > > >> > On Wed, Dec 21, 2005 at 04:05:07PM -0500, Chemistry System Admin alleged:
> > > >> > 
> > > >> > > gcc -g -I../../../src/include -I../../../src/include -DHAVE_CONFIG_H -c
> > > >> > > pbs_log.c
> > > >> > > pbs_log.c:114: `_POSIX_PATH_MAX' undeclared here (not in a function)
> > > >> > > pbs_log.c: In function `log_open':
> > > >> > > pbs_log.c:181: `_POSIX_PATH_MAX' undeclared (first use in this 
> > > >> > > function)
> > > >> > > pbs_log.c:181: (Each undeclared identifier is reported only once
> > > >> > > pbs_log.c:181: for each function it appears in.)
> > > >> > > pbs_log.c: At top level:
> > > >> > > pbs_log.c:114: storage size of `log_directory' isn't known
> > > >> > > make[4]: *** [pbs_log.o] Error 1
> > > >> > 
> > > >> > So the following program won't build?  If not, something is broken with
> > > >> > your system's header files.
> > > >> > 
> > > >> > # include <stdio.h>
> > > >> > # include <limits.h>
> > > >> > 
> > > >> > int main() {
> > > >> >   printf("%d\n",_POSIX_PATH_MAX);
> > > >>> }
> > > >> > 
> > > >>> _______________________________________________
> > > _______________________________________________
> > > torqueusers mailing list
> > > torqueusers at supercluster.org
> > > http://www.supercluster.org/mailman/listinfo/torqueusers
> > 
> > _______________________________________________
> > torqueusers mailing list
> > torqueusers at supercluster.org
> > http://www.supercluster.org/mailman/listinfo/torqueusers
> > 
> 
> _______________________________________________
> torqueusers mailing list
> torqueusers at supercluster.org
> http://www.supercluster.org/mailman/listinfo/torqueusers

-- 
Garrick Staples, Linux/HPCC Administrator
University of Southern California
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://www.supercluster.org/pipermail/torqueusers/attachments/20060103/c41bb471/attachment.bin


More information about the torqueusers mailing list