Re: pkg/40434: security/skey fails to build on Solaris

["OBATA Akio" <obache%netbsd.org@localhost>]

The following reply was made to PR pkg/40434; it has been noted by GNATS.

From: "OBATA Akio" <obache%netbsd.org@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc: 
Subject: Re: pkg/40434: security/skey fails to build on Solaris
Date: Tue, 20 Jan 2009 14:57:42 +0900

 On Mon, 19 Jan 2009 21:40:05 +0900, Jörn Clausen 
<joern%uni-bielefeld.de@localhost> wrote:
 
 >  Rats, now security/openssh is broken:
 
 Following patch shuold fixes it:
 
 Index: buildlink3.mk
 ===================================================================
 RCS file: /cvsroot/pkgsrc/security/skey/buildlink3.mk,v
 retrieving revision 1.7
 diff -u -r1.7 buildlink3.mk
 --- buildlink3.mk      8 Jul 2006 23:11:08 -0000       1.7
 +++ buildlink3.mk      20 Jan 2009 05:44:01 -0000
 @@ -18,4 +18,10 @@
  BUILDLINK_DEPMETHOD.skey?=    build
  .endif # SKEY_BUILDLINK3_MK
 
 +.include "../../mk/bsd.fast.prefs.mk"
 +
 +.if ${OPSYS} == "SunOS"
 +BUILDLINK_TRANSFORM+= l:skey:skey:md5
 +.endif
 +
  BUILDLINK_DEPTH:=     ${BUILDLINK_DEPTH:S/+$//}
 
 >  As mentioned before, ssh from openssh now pulls in libcrypto.so from
 >  openssl, and /usr/lib/libmd5.so via skey. Whose MD5 routines win? I
 >  think it would be much cleaner to build skey and openssh that solely
 >  rely on security/openssl and don't use any of the functions provided by
 >  Solaris. There is /usr/bin/ssh, if you insist on using the functions
 >  provided by the OS.
 
 OpenSSL define MD5_* functions, and skey using MD5* functions (underscore!).
 
 >  NB: This only makes the packages compile, I have no idea, if they work
 >  as intended. As far as I can tell, skey should now have no support for
 >  MD4 whatsoever. But
 > 
 >  $  skey -md4 ...
 > 
 >  works and returns a result. But I can't say, if it is correct.
 
 Did you want to say that MD4 should be disabled nowadays?
 
 -- 
 "Of course I love NetBSD":-)
 OBATA Akio / obache%NetBSD.org@localhost