Richard,
Our configure shouldn't define USHORT if there is an existing definition.
I have seen this sort of thing happen when configure "finds" cpp as the C
preprocessor and runs configure tests with cpp. Various defines (e.g.
-Dsgi) that occur by default with cc are not set. System include files
testing e.g. "sgi" will give different results than when make invokes cc to
compile.
Bottom line is, if this is the case, setting
CPPFLAGS=-Dsgi
prior to configure will cure the problem.
for what it is worth, my sgi configure environment is:
setenv CPPFLAGS " -Dsgi -I /usr/local/include "
setenv CFLAGS " -I /usr/local/include -n32 "
setenv LDFLAGS " -L /usr/local/lib -L /usr/lib32 "
setenv CXXFLAGS " -n32 "
setenv SO_FLAGS " -n32 "
alias cpconfigure "./configure
--with-cosmojar-path=/homes/ppc4/daniel/classes/npcosmop211.jar
--with-java40jar_path=/homes/ppc4/daniel/classes/java40.jar"
hope this helps.
Pete
Richard Gillilan wrote:
> David L. Thompson wrote:
>
> > Sounds like your configure goof'd. Look at config.cache for ushort
> > and tell us if it says yes or no.
>
> ac_cv_type_ushort ${ac_cv_type_ushort = no}
>
> > Are you saying that your /usr/include/sys/bsd_types.h has it declared
> > twice?
>
> nope, the bsd_types.h definition conflicts with dxconfig.h.
