I think the issue is the re-arranging from "select shortest type" to "select longest"... Since we treat OS X special now, I'll revert that change.
On Dec 20, 2010, at 12:47 PM, Ruediger Pluem wrote: > > > On 12/18/2010 03:56 PM, [email protected] wrote: >> Author: jim >> Date: Sat Dec 18 14:56:47 2010 >> New Revision: 1050644 >> >> URL: http://svn.apache.org/viewvc?rev=1050644&view=rev >> Log: >> (Try to) support Darwin's "universal" builds by allowing for >> 2 separate bit sizes for some types in the same binary/library >> (which is what a universal build is). As such, the C header files >> need to know how the *current* codebase is being built (which >> links to apr) and choose the correct bit size depending on >> the compiler flags. >> >> Modified: >> apr/apr/trunk/build/apr_hints.m4 >> apr/apr/trunk/configure.in >> apr/apr/trunk/include/apr.h.in > > Not sure what exactly causes this but since then I see a lot of > the following warnings when building httpd on 64 Bit Linux with APR trunk: > > ssl_engine_io.c:1611: warning: long long int format, apr_off_t arg (arg 9) > > > Regards > > Rüdiger >
