Some modules show problem detected

Dennis Veatch dveatch at woh.rr.com
Wed Apr 20 14:35:07 UTC 2005


On Wednesday 20 April 2005 9:40 am, Auke Kok wrote:
> Dennis Veatch wrote:
> >I have noticed when lining some modules at the end, after the lin is
> > completed it says it detected a problem, like this one;
> >
> >make[1]: Leaving directory
> >`/usr/src/gcc-3.3.3/BUILD/i686-pc-linux-gnu/boehm-gc'
> >WARNING: BUILD does not call prepare_install or installs zero files!
> >+ Installing default documentation to: /usr/share/doc/gcc/README
> >+ Installing default documentation to: /usr/share/doc/gcc/README.SCO
> >+ Installing default documentation to: /usr/share/doc/gcc/INSTALL
> >+ Installing default documentation to: /usr/share/doc/gcc/FAQ
> >+ Installing default documentation to: /usr/share/doc/gcc/ChangeLog
> >Creating /var/log/lunar/compile/gcc-3.3.3.bz2
> >Creating /var/log/lunar/install/gcc-3.3.3
> >Creating /var/log/lunar/md5sum/gcc-3.3.3
> >+ updating lunar state files after module installation
> >+ module size is 176KB
> >+ destroying building dir "/usr/src/gcc-3.3.3"
> >+ removing installwatch file
> >+ destroying building dir "/usr/src/gcc-3.3.3"
> >! Problem detected
> >
> >
> >Is this a cosmetic thing?
>
> only for some modules, like the -profile modules. In the case of gcc I
> suspect something broke during your compile, or (????). I can't see this
> behaviour myself anywhere with gcc and the BUILD actually is conform
> requirements. (hint: 'prepare_install' is required to be called from
> BUILD files -> lvu BUILD gcc)
>
> sofar
>
> _______________________________________________


Well I don't see any show stopper errors in the gcc compile log and everything 
seems to be compiling fine. The "! Problem detected" does not show up in the 
compile log since it happens after the source dir is deleted. The activity 
log shows ;

20050419-18:18:30       lrm     gcc     3.3.3   success
20050419-18:21:28       lin     gcc     3.3.3   failed  Due to build errors 
during or after

In the other cases where I see this it appears to have a problem destroying 
the source dir. For example, j2sdk the compile and activity logs show the 
same things as with gcc and the j2sdk source was not removed.

So on a hunch I disabled "TMPFS" in the Option menu and relined j2sdk. This 
time there was no problem detected after destroying the source dir and it was 
actually removed this time.

I'm not sure what's going on here as I have the same settings in "/etc/fstab" 
for the tmpfs as always. Though I am not sure that really has any bearing on 
this.


More information about the Lunar mailing list