From: Alan W. I. <ir...@be...> - 2006-10-03 17:04:47
|
On 2006-10-02 23:44-0400 Hazen Babcock wrote: > >> it doesn't reprocess (using the neat new NOT DEFINED logic pointed out to >> me >> on the CMake mailing list). Please try the result to make sure >> octave.cmake >> now works properly on Mac OS X. > > It is still behaving as before. > > I added: > message(STATUS "TRIMMED = ${OCTAVE_INCLUDE_PATH_TRIMMED}") > > before: > if(NOT DEFINED OCTAVE_INCLUDE_PATH_TRIMMED) > > in octave.cmake & it looks like the value of OCTAVE_INCLUDE_PATH_TRIMMED is > being forgotten between invocations of octave.cmake Thanks for helping me finally track this one down. Yes, my previous non-working fix depended on OCTAVE_INCLUDE_PATH_TRIMMED which is uncached and therefore by definition undefined for each cmake invocation. Normally, I clean out the build tree before every cmake invocation so I never saw the bug. The breakthrough in understanding came when I realized I could generate the bad 2^n result you discovered by simply repeating n cmake commands without cleaning out the build directory. The new logic in octave.cmake (which I just committed) depends on the cached variable OCTAVE_INCLUDE_PATH, and that new logic works for me. Please try it for yourself. Alan __________________________ Alan W. Irwin Astronomical research affiliation with Department of Physics and Astronomy, University of Victoria (astrowww.phys.uvic.ca). Programming affiliations with the FreeEOS equation-of-state implementation for stellar interiors (freeeos.sf.net); PLplot scientific plotting software package (plplot.org); the Yorick front-end to PLplot (yplot.sf.net); the Loads of Linux Links project (loll.sf.net); and the Linux Brochure Project (lbproject.sf.net). __________________________ Linux-powered Science __________________________ |