Hi *,
we're experiencing a strange problem here and hope somebody can point us to
a possible solution:
Parallel makes get interrupted by messages like
gmake[1]: *** No rule to make target
`/opt/oracle/product/8.1.7/rdbms/demo/oci.h', needed by `EnumeratorInit.o'.
Stop.
But looking at the timestamp for the accused file it says
-rw-r--r-- 1 oracle dba 91457 Apr 25 2000
/opt/oracle/product/8.1.7/rdbms/demo/oci.h
so it really can't be. And sure enough just repeating the make-command is
good enough most of the times...given it doesn't complain about another file
in the same fashion :-(
We've had that with a number of library-header-dependencies read from
Makefile.dep - the more parallel, the more false errors. We haven't seen any
of it in non-parallel makes.
Systems are 4- and 8-way suns running solaris 8 - error seen in make 3.78
and 3.80 (3.79 not checked). Any ideas anybody?
thanks in advance, dave
_______________________________________________
Help-make mailing list
[EMAIL PROTECTED]
http://mail.gnu.org/mailman/listinfo/help-make
