From: Alan W. I. <ir...@be...> - 2004-03-29 20:43:34
|
On 2004-03-29 11:26-0800 Brian D. Wright wrote: > > Hi folks, > Trying to make a new RPM for the latest CVS tarball > and it appears the Java detection is broken. > case $system in > Linux* ) > JAVAMDINCDIR="$JAVAINCDIR/linux" > echo "$as_me:$LINENO: result: yes" >&5 > ;; > > ... > > and it appears that $system is not defined. Good catch. The problem extends to octave as well. I suspect everywhere $system was used before, we should be using some variant of $host instead. grep '$system' cf/*.ac cf/java.ac: case $system in cf/octave.ac:case "$system" in I have changed the above $system variables to $host_os and also the corresponding test strings to what is appropriate. For everybody's information here are examples of the $host variable variations for my system. grep host Makefile host_triplet = i686-pc-linux-gnu host = i686-pc-linux-gnu host_alias = host_cpu = i686 host_os = linux-gnu host_vendor = pc Here is what we have done elsewhere: grep '$host' cf/*.ac cf/c++.ac: case "$host_os" in cf/csiro.ac: case "$host_cpu" in cf/pthread.ac: case "$host_os" in cf/python.ac:AM_CONDITIONAL(darwin_os, [case "$host_os" in darwin*) true;; *) false; esac]) cf/summary.ac:system: $host I don't know much about it, but I presume the cf/csiro.ac logic using $host_cpu is appropriate because the mieee flags are compiler AND cpu dependent. In cf/summary.ac I have just changed the "system:" label to "host:" since system now seems to be a misnomer. Alan __________________________ Alan W. Irwin email: ir...@be... phone: 250-727-2902 Astronomical research affiliation with Department of Physics and Astronomy, University of Victoria (astrowww.phys.uvic.ca). Programming affiliations with the 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 __________________________ |