Hello, I use the following compilers:
HP C V7.2-001 on OpenVMS IA64 V8.3 Compaq C V6.4-008 on OpenVMS Alpha V7.3-2 These were the changes that I applied - and rewarded me with clean compile and usable libraries. For sure you know better the code... and there might be better solutions. Please, include what you feel that is right. ... my only wish is to get a clean build on OpenVMS when I download the code next time. Thank you Richard. Regards, Z -----Original Message----- From: Richard Levitte [mailto:rich...@levitte.org] Sent: den 12 november 2009 22:56 To: openssl-dev@openssl.org; Arpadffy Zoltan Subject: Re: OpenSSL 1.0.0 beta4 release In message <839c820b5c926b4b89713b3a6ed68d2aae5...@sgstmail.scigames.at> on Thu, 12 Nov 2009 16:26:57 +0100, "Arpadffy Zoltan" <zoltan.arpad...@scientificgames.se> said: Zoltan.Arpadffy> Hello, Zoltan.Arpadffy> Zoltan.Arpadffy> Additionally to SMS's changes... here are the changes that are needed to Zoltan.Arpadffy> be added in order to get OpenVMS build correctly. Okay, since I'm obviously in an interactive mood (and it's been requested anyway ;-)), there are a couple of your patches that I don't understand: Zoltan.Arpadffy> TOR_ZAY $ gdiff -p Zoltan.Arpadffy> DSA104:<USERS.ZAY.WORK.OPENSSL-100-BETA4-ORIG.crypto.pqueue>pqueue.h Zoltan.Arpadffy> DSA104:<USERS.ZAY.WORK.OPENSSL-100-BETA4.CRYPTO.PQUEUE>PQUEUE.H Zoltan.Arpadffy> Zoltan.Arpadffy> *** dsa104:<users.zay.work.openssl-100-beta4-orig.crypto.pqueue>pqueue.h Sat May 16 18:17:46 2009 Zoltan.Arpadffy> --- dsa104:<users.zay.work.openssl-100-beta4.crypto.pqueue>pqueue.h Thu Nov 12 10:12:12 2009 Zoltan.Arpadffy> *************** Zoltan.Arpadffy> *** 64,69 **** Zoltan.Arpadffy> --- 64,74 ---- Zoltan.Arpadffy> #include <stdlib.h> Zoltan.Arpadffy> #include <string.h> Zoltan.Arpadffy> Zoltan.Arpadffy> + #ifdef OPENSSL_SYS_VMS Zoltan.Arpadffy> + #include <resource.h> Zoltan.Arpadffy> + #include <sys/timeb.h> Zoltan.Arpadffy> + #endif Zoltan.Arpadffy> + Zoltan.Arpadffy> typedef struct _pqueue *pqueue; Zoltan.Arpadffy> Zoltan.Arpadffy> typedef struct _pitem Why? I see nothing in that header file that would need anything from resource.h or sys/timeb.h... is it really something needed in one of the .c files? Then I think it's better to change there. Zoltan.Arpadffy> TOR_ZAY $ gdiff -p Zoltan.Arpadffy> DSA104:<USERS.ZAY.WORK.OPENSSL-100-BETA4-ORIG.apps>s_socket.c Zoltan.Arpadffy> DSA104:<USERS.ZAY.WORK.OPENSSL-100-BETA4.apps>s_socket.c Zoltan.Arpadffy> *** dsa104:<users.zay.work.openssl-100-beta4-orig.apps>s_socket.c Wed Aug 26 13:21:50 2009 Zoltan.Arpadffy> --- dsa104:<users.zay.work.openssl-100-beta4.apps>s_socket.c Thu Nov 12 10:47:18 2009 Zoltan.Arpadffy> *************** Zoltan.Arpadffy> *** 72,78 **** Zoltan.Arpadffy> recursive header file inclusion, resulting in the compiler complaining Zoltan.Arpadffy> that u_int isn't defined, but only if _POSIX_C_SOURCE is defined, which Zoltan.Arpadffy> is needed to have fileno() declared correctly... So let's define u_int */ Zoltan.Arpadffy> ! #if defined(OPENSSL_SYS_VMS_DECC) && !defined(__U_INT) Zoltan.Arpadffy> #define __U_INT Zoltan.Arpadffy> typedef unsigned int u_int; Zoltan.Arpadffy> #endif Zoltan.Arpadffy> --- 72,78 ---- Zoltan.Arpadffy> recursive header file inclusion, resulting in the compiler complaining Zoltan.Arpadffy> that u_int isn't defined, but only if _POSIX_C_SOURCE is defined, which Zoltan.Arpadffy> is needed to have fileno() declared correctly... So let's define u_int */ Zoltan.Arpadffy> ! #if (defined(VMS) || defined(__VMS)) && !defined(__U_INT) Zoltan.Arpadffy> #define __U_INT Zoltan.Arpadffy> typedef unsigned int u_int; Zoltan.Arpadffy> #endif Why? it includes e_os2.h, which defines OPENSSL_SYS_VMS_DECC if DEC C (Compaq C? HP C?) is used. Does this mean the definitions in e_os2.h need to be updated? Zoltan.Arpadffy> TOR_ZAY $ gdiff -p Zoltan.Arpadffy> DSA104:<USERS.ZAY.WORK.OPENSSL-100-BETA4-ORIG.apps>s_server.c Zoltan.Arpadffy> DSA104:<USERS.ZAY.WORK.OPENSSL-100-BETA4.apps>s_server.c Zoltan.Arpadffy> *** dsa104:<users.zay.work.openssl-100-beta4-orig.apps>s_server.c Wed Oct 28 18:49:38 2009 Zoltan.Arpadffy> --- dsa104:<users.zay.work.openssl-100-beta4.apps>s_server.c Thu Nov 12 10:47:57 2009 Zoltan.Arpadffy> *************** Zoltan.Arpadffy> *** 165,171 **** Zoltan.Arpadffy> recursive header file inclusion, resulting in the compiler complaining Zoltan.Arpadffy> that u_int isn't defined, but only if _POSIX_C_SOURCE is defined, which Zoltan.Arpadffy> is needed to have fileno() declared correctly... So let's define u_int */ Zoltan.Arpadffy> ! #if defined(OPENSSL_SYS_VMS_DECC) && !defined(__U_INT) Zoltan.Arpadffy> #define __U_INT Zoltan.Arpadffy> typedef unsigned int u_int; Zoltan.Arpadffy> #endif Zoltan.Arpadffy> --- 165,171 ---- Zoltan.Arpadffy> recursive header file inclusion, resulting in the compiler complaining Zoltan.Arpadffy> that u_int isn't defined, but only if _POSIX_C_SOURCE is defined, which Zoltan.Arpadffy> is needed to have fileno() declared correctly... So let's define u_int */ Zoltan.Arpadffy> ! #if (defined(VMS) || defined(__VMS)) && !defined(__U_INT) Zoltan.Arpadffy> #define __U_INT Zoltan.Arpadffy> typedef unsigned int u_int; Zoltan.Arpadffy> #endif Same thing here. Zoltan.Arpadffy> TOR_ZAY $ gdiff -p Zoltan.Arpadffy> DSA104:<USERS.ZAY.WORK.OPENSSL-100-BETA4-ORIG.test>tests.com Zoltan.Arpadffy> DSA104:<USERS.ZAY.WORK.OPENSSL-100-BETA4.test>tests.com Zoltan.Arpadffy> *** dsa104:<users.zay.work.openssl-100-beta4-orig.test>tests.com Tue Aug 25 09:17:14 2009 Zoltan.Arpadffy> --- dsa104:<users.zay.work.openssl-100-beta4.test>tests.com Thu Nov 12 15:46:44 2009 Zoltan.Arpadffy> *************** $ exe_dir := sys$disk:[-.'__arch'.exe.ap Zoltan.Arpadffy> *** 14,19 **** Zoltan.Arpadffy> --- 14,28 ---- Zoltan.Arpadffy> $ Zoltan.Arpadffy> $ set default '__here' Zoltan.Arpadffy> $ Zoltan.Arpadffy> + $ ROOT = F$PARSE(__here,"[]A.;0",,,"SYNTAX_ONLY,NO_CONCEAL") - Zoltan.Arpadffy> "A.;0" Zoltan.Arpadffy> + $ ROOT_DEV = F$PARSE(ROOT,,,"DEVICE","SYNTAX_ONLY") Zoltan.Arpadffy> + $ ROOT_DIR = F$PARSE(ROOT,,,"DIRECTORY","SYNTAX_ONLY") - Zoltan.Arpadffy> + - ".][000000" - "[000000." - "][" - "[" - "]" - Zoltan.Arpadffy> ".TEST" Zoltan.Arpadffy> + $ ROOT = ROOT_DEV + "[" + ROOT_DIR Zoltan.Arpadffy> + $ DEFINE/NOLOG SSLROOT 'ROOT'.APPS.] /TRANS=CONC Zoltan.Arpadffy> + $ Zoltan.Arpadffy> + $ openssl :== $'ROOT'.'__arch'.exe.apps]OPENSSL Zoltan.Arpadffy> + $ Zoltan.Arpadffy> $ on control_y then goto exit Zoltan.Arpadffy> $ on error then goto exit Zoltan.Arpadffy> $ Yeah, except if someone has a previous version of OpenSSL installed and runs that, he/she is in for a surprise (and perhaps not a good one). I think it's better to set the openssl variable properly on the specific .com files (I assume it's errors in those that provoked this change...). Zoltan.Arpadffy> *************** $! NOTE: This list reflects the list of Zoltan.Arpadffy> *** 25,38 **** Zoltan.Arpadffy> $! "alltests" target in Makefile. This should make it easy to see Zoltan.Arpadffy> $! if there's a difference that needs to be taken care of. Zoltan.Arpadffy> $ tests := - Zoltan.Arpadffy> ! test_des,test_idea,test_sha,test_md4,test_md5,test_hmac,- Zoltan.Arpadffy> test_md2,test_mdc2,test_wp,- Zoltan.Arpadffy> test_rmd,test_rc2,test_rc4,test_rc5,test_bf,test_cast,test_aes,- Zoltan.Arpadffy> test_rand,test_bn,test_ec,test_ecdsa,test_ecdh,- Zoltan.Arpadffy> test_enc,test_x509,test_rsa,test_crl,test_sid,- Zoltan.Arpadffy> test_gen,test_req,test_pkcs7,test_verify,test_dh,test_dsa,- Zoltan.Arpadffy> ! test_ss,test_ca,test_engine,test_evp,test_ssl,test_tsa,test_ige,- Zoltan.Arpadffy> ! test_jpake,test_cms Zoltan.Arpadffy> $ endif Zoltan.Arpadffy> $ tests = f$edit(tests,"COLLAPSE") Zoltan.Arpadffy> $ Zoltan.Arpadffy> --- 34,47 ---- Zoltan.Arpadffy> $! "alltests" target in Makefile. This should make it easy to see Zoltan.Arpadffy> $! if there's a difference that needs to be taken care of. Zoltan.Arpadffy> $ tests := - Zoltan.Arpadffy> ! test_tsa,test_des,test_idea,test_sha,test_md4,test_md5,test_hmac,- Zoltan.Arpadffy> test_md2,test_mdc2,test_wp,- Zoltan.Arpadffy> test_rmd,test_rc2,test_rc4,test_rc5,test_bf,test_cast,test_aes,- Zoltan.Arpadffy> test_rand,test_bn,test_ec,test_ecdsa,test_ecdh,- Zoltan.Arpadffy> test_enc,test_x509,test_rsa,test_crl,test_sid,- Zoltan.Arpadffy> test_gen,test_req,test_pkcs7,test_verify,test_dh,test_dsa,- Zoltan.Arpadffy> ! test_ss,test_ca,test_engine,test_evp,test_ssl,- Zoltan.Arpadffy> ! test_jpake,test_tsa,test_ige,test_cms Zoltan.Arpadffy> $ endif Zoltan.Arpadffy> $ tests = f$edit(tests,"COLLAPSE") Zoltan.Arpadffy> $ Hmmm, hold on, what source are you coming from? Those things are already added in beta4. Zoltan.Arpadffy> *************** $ test_jpake: Zoltan.Arpadffy> *** 284,289 **** Zoltan.Arpadffy> --- 293,302 ---- Zoltan.Arpadffy> $ write sys$output "Test JPAKE" Zoltan.Arpadffy> $ mcr 'texe_dir''jpaketest' Zoltan.Arpadffy> $ return Zoltan.Arpadffy> + $ test_cms: Zoltan.Arpadffy> + $ write sys$output "CMS consistency test" Zoltan.Arpadffy> + $ perl CMS-TEST.PL Zoltan.Arpadffy> + $ return Zoltan.Arpadffy> $ Zoltan.Arpadffy> $ Zoltan.Arpadffy> $ exit: Oh... ok, that will require everyone to have perl installed... back in time, there were quite a number of VMS folks that protested loudly at the idea... things change I guess ;-) Cheers, Richard -- Richard Levitte rich...@levitte.org http://richard.levitte.org/ "Life is a tremendous celebration - and I'm invited!" -- from a friend's blog, translated from Swedish ______________________________________________________________________ OpenSSL Project http://www.openssl.org Development Mailing List openssl-dev@openssl.org Automated List Manager majord...@openssl.org
