Found the issue. Apparently build/aprenv.py needed to be modified.
Since autoconf doesn't normally rely on Python, this caught me a little
off guard.
In any case, I've updated the patch, and built it on Openbsd, Fedora,
Solaris.
https://issues.apache.org/bugzilla/show_bug.cgi?id=42848
Can someone please review this?
Thanks.
On 7/4/10 9:44 PM, Philip Prindeville wrote:
I've made the following changes, but for whatever reason
APR_HAVE_NETINET_IN_H isn't getting set.
% grep -i systm * */*
config.log:ac_cv_header_netinet_in_systm_h=yes
config.log:netinet_insystmh=''
config.log:#define HAVE_NETINET_IN_SYSTM_H 1
config.status:S["netinet_insystmh"]=""
config.status:D["HAVE_NETINET_IN_SYSTM_H"]=" 1"
configure:netinet_insystmh
configure: netinet/in_systm.h \
configure: netinet/in_systm.h \
configure:#ifdef HAVE_NETINET_IN_SYSTM_H
configure:#include<netinet/in_systm.h>
...
include/apr.h:#define APR_HAVE_NETINET_IN_SYSTM_H
...
%
These are the generated files. The modified repo files are below in
the diffs.
What did I miss here?
Not the last line has no right-hand side for the definition.
Why does netinet_insystmh not get set to 1?
Thanks.
-Philip
