Dmitry, We can't have a library disappear between real/complex builds. Is there a different library that can take in these symbols?
Note: not all petsc objects get a library of their own. For eg: IS goes into libpetscvec etc.. Satish On Mon, 21 Mar 2011, Jose E. Roman wrote: > src/characteristic/makefile has #requiresscalar real, so libcharacteristic.a > is not created with-scalar-type=complex. > 'make test' fails in this case. > > Also, with cmake builds: > Scanning dependencies of target petsccharacteristic > make[2]: *** No rule to make target > `CMakeFiles/petsccharacteristic.dir/build'. Stop. > make[1]: *** [CMakeFiles/petsccharacteristic.dir/all] Error 2 > make[1]: *** Waiting for unfinished jobs.... > > Jose > > > El 21/03/2011, a las 22:20, Satish Balay escribi?: > > > added > > > > satish > > > > On Mon, 21 Mar 2011, Barry Smith wrote: > > > >> > >> Maybe we just need to force Satish to have a nightly build that checks > >> the several library tests. > >> > >> Barry > >> > >> On Mar 21, 2011, at 4:10 PM, Jed Brown wrote: > >> > >>> On Mon, Mar 21, 2011 at 22:07, Barry Smith <bsmith at mcs.anl.gov> wrote: > >>> Why do you care if there are dependencies? Admittedly that may make PETSc > >>> harder to develop and maintain > >>> > >>> Exactly. > >>> > >>> but from the user point of view what is wrong with just having one > >>> big-assed blob? > >>> > >>> It doesn't matter for users. > >>> > >>> (But a circular dependency would make executables grow so its existence > >>> might matter to someone using only Mat+Vec in an extremely low-memory > >>> environment.) > >> > >> > >> > > > > >
