From: noreply@...
Date: 2005-07-15T05:04:18+09:00
Subject: [ ruby-Bugs-2105 ] ruby (v182 or CVS-HEAD) fails @ 'make' w/ OpenSSL v>= 0.9

Bugs item #2105, was opened at 2005-07-14 12:58
You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=1698&aid=2105&group_id=426

Category: None
Group: None
Status: Open
Resolution: None
Priority: 3
Submitted By: Richard _ (openmacnews)
Assigned to: Nobody (None)
Summary: ruby (v182 or CVS-HEAD) fails @ 'make' w/ OpenSSL v>= 0.9

Initial Comment:
hi all,

bldg EITHER ruby v182 or CVS-HEAD on OSX 10.4.2 w/ external OpenSSL in /usr/local

	OpenSSL 0.9.8 05 Jul 2005


this error exists for either/both:

	gcc_select -force 3.3
	gcc_select -force 4.0
 


after a successful configure with:


	./configure \
	--prefix=/usr/local/ruby \
	--disable-fat-binary \
	--enable-frame-address \
	--enable-largefile \
	--enable-pthread \
	--enable-shared \
	--enable-install-doc

'make' subsequently fails @:

	...
	compiling digest/md5
	mkdir -p ../../../.ext/powerpc-darwin8.2.0/digest
	gcc -fno-common -g -O2  -fno-common -pipe -fno-common -DHAVE_CONFIG_H -I../../.././ext/digest/md5/..  -I. -I../../.. -I../../../. -I../../.././ext/digest/md5 -DHAVE_OPENSSL_MD5_H -DHAVE_SYS_CDEFS_H -DHAVE_INTTYPES_H -DHAVE_UNISTD_H  -c md5init.c
	gcc -fno-common -g -O2  -fno-common -pipe -fno-common -DHAVE_CONFIG_H -I../../.././ext/digest/md5/..  -I. -I../../.. -I../../../. -I../../.././ext/digest/md5 -DHAVE_OPENSSL_MD5_H -DHAVE_SYS_CDEFS_H -DHAVE_INTTYPES_H -DHAVE_UNISTD_H  -c md5ossl.c
	In file included from md5ossl.h:6,
					 from md5ossl.c:3:
	/usr/local/include/openssl/md5.h:108: error: parse error before 'size_t'
	/usr/local/include/openssl/md5.h:110: error: parse error before 'size_t'
	md5ossl.c: In function 'MD5_End':
	md5ossl.c:17: warning: pointer targets in passing argument 1 of 'sprintf' differ in signedness
	make[1]: *** [md5ossl.o] Error 1
	make: *** [all] Error 1

this looks like a problem with changes in OpenSSL v<0.9.7 --> v>=0.9.8 (which is ow a 'release' version ...).

specifically, comparing Apple's bundled ssl:

	% /usr/bin/openssl version
		OpenSSL 0.9.7b 10 Apr 2003
	
	% grep size_t /usr/include/openssl/md5.h
	%
	
with LATEST ssl:

	% /usr/local/bin/openssl version
		OpenSSL 0.9.8 05 Jul 2005

	% grep size_t /usr/local/include/openssl/md5.h
		int MD5_Update(MD5_CTX *c, const void *data, size_t len);
		unsigned char *MD5(const unsigned char *d, size_t n, unsigned char *md);"

i suspect that ruby's redefining size_t somehow/where ...

cheers,

richard

----------------------------------------------------------------------

You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=1698&aid=2105&group_id=426