On 10/31/2011 09:25 AM, William A. Rowe Jr. wrote:
On 10/31/2011 2:37 AM, mt...@apache.org wrote:
--- tomcat/jk/trunk/native/iis/pcre/pcre.amd64 (original)
+++ tomcat/jk/trunk/native/iis/pcre/pcre.amd64 Mon Oct 31 07:37:07 2011
@@ -66,7 +66,7 @@ LIB32_OBJS= \
$(LIB32_FLAGS) $(DEF_FLAGS) $(LIB32_OBJS)
<<
-CPP_PROJ=-nologo -MD -W3 -O2 -Ob2 -Oy- -Zi -EHsc /D "_WIN32" /D "NDEBUG" /D "_WINDOWS" /D "_AMD64_=1" -DWIN64 /D
"_WIN64" /Wp64 /FIPRE64PRA.H /D "PCRE_STATIC" /Fo"$(INTDIR)\\" /Fd"$(INTDIR)\pcre_src" /FD /c
+CPP_PROJ=-nologo -MD -W3 -O2 -Ob2 -Oy- -Zi -EHsc /D "_WIN32" /D "NDEBUG" /D "_WINDOWS" /D "_AMD64_=1" -DWIN64 /D
"_WIN64" /D "PCRE_STATIC" /Fo"$(INTDIR)\\" /Fd"$(INTDIR)\pcre_src" /FD /c
You had informed me (correctly) that MS refers to 64 bit x86 as WIN64.
Why would _WIN32 remain?
Ask Microsoft :)
eg. <crtdefs.h>
#if !defined (_WIN32)
#error ERROR: Only Win32 target supported!
#endif /* !defined (_WIN32) */
And then you have in the <io.h>
#ifdef _WIN64
typedef unsigned __int64 size_t;
#else
typedef _W64 unsigned int size_t;
#endif
I suppose it has to do with backward compatibility.
I have no idea why they didn't just do:
#ifdef _WIN64
#define _WIN32
#endif
Regards
--
^TM
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org
