> This patch looks good, and I'll commit it tomorrow morning unless somebody
> beats me to it. (I'm re-installing my laptop, so I don't have access to
> my apache trees right now). I agree 100% that we should THREAD_CPPFLAGS
> throughout apr_hints.m4, could you please provide a patch that does that?
> If not, I'll add that to the patch before I commit tomorrow.
Actually upon further review, CPPFLAGS includes THREADS_CPPFLAGS at
configure.in:578 - see Roy's comment - which is why I'm not mucking
around with apr_hints.m4.
APR_CHECK_SIGWAIT_ONE_ARG came BEFORE CPPFLAGS included THREAD_CPPFLAGS.
But, since it relies on THREAD_CPPFLAGS, it should only be called AFTER
CPPFLAGS includes THREAD_CPPFLAGS.
If threads are disallowed (i.e. --disable-threads or threads can't be
found), then apr_sigwait isn't included anyway. Therefore, it seems
that APR_CHECK_SIGWAIT_ONE_ARG should only be called when we have threads.
This patch moves the test of APR_CHECK_SIGWAIT_ONE_ARG to only be called
when we have threads.
APR_CHECK_SIGWAIT_ONE_ARG should just assume that CPPFLAGS is setup
correctly. So, I believe that this patch makes much more sense. You
can disregard my previous patch.
My $.02,
Justin Erenkrantz
[EMAIL PROTECTED]
Index: configure.in
===================================================================
RCS file: /home/cvspublic/apr/configure.in,v
retrieving revision 1.235
diff -u -r1.235 configure.in
--- configure.in 2001/02/21 18:51:07 1.235
+++ configure.in 2001/02/22 05:59:44
@@ -285,7 +285,6 @@
AC_SUBST(mmap)
AC_SUBST(have_memmove)
-APR_CHECK_SIGWAIT_ONE_ARG
dnl #----------------------------- Checks for Any required Headers
AC_HEADER_STDC
@@ -629,6 +628,7 @@
if test "$threads" = "1"; then
echo "APR will use threads"
AC_CHECK_LIB(c_r, readdir, AC_DEFINE(READDIR_IS_THREAD_SAFE))
+ APR_CHECK_SIGWAIT_ONE_ARG
else
echo "APR will be non-threaded"
fi
