Steve, Tom, Eric (and others)!!!,
Thank you so much for your suggestions. I just changed the names of the 
functions and used  #defines in the necessary .h files and things are 
working fine.... I really do appreciate your help!!

BTW, just a last query, while changing the signature of the MD5_Update() 
and the SHA1_Update(), I also looked for the actual implementation of 
these functions. While their signatures exist in md5.h and sha.h 
respectively, I could not find any .c file within  the source that 
contained the function definition. According to my understanding, I 
should also have the source so that my changes take effect after 
compilation/installation , right??? So, did I miss something somewhere?? 
BTW, this is not a critical issue (now that it's OK!!), but I just want 
to get to know it.....

- Ashwin



Dr. Stephen Henson wrote:

>On Wed, Aug 28, 2002, Ashwin C Uthappa wrote:
>
>>Hi All,
>>Just  some updates on my iPlanet problems with SSL. For those who came 
>>in late, a quick recap!!
>>
>>I intend to carry out SSL communication (NOT web transactions) between a 
>>communication application and a engine. In the actual production 
>>environment, I am supposed to create a shared object to be plugged into 
>>iPlanet. When I start the iPlanet web server, during the Init Stage, I 
>>fork a child process  within which I invoke the  communication methods 
>>to  establish communication with the engine.
>>
>>When I code the communication layer as an independent application ( 
>>started from command line), it works perfectly. But when trying to fork 
>>and call the functions from the iPlanet web server, it fails and the 
>>child process crashes.  The error I see at my engine side is
>>        1305:error:140790E5:SSL routines:SSL23_WRITE:ssl handshake 
>>failure:s23_lib.c:226:
>>
>>On snooping around the OpenSSL code, I found that the RAND_add() and 
>>RAND_pseudo_bytes() functions (among other functions) fail at my 
>>client's (communication layer) side. They do not return and they cause 
>>the program to crash. Why this happens with iplanet and not with a 
>>standalone application - that's my question?? Could there be something 
>>in iPlanet's environment/settings/context that could be causing this 
>>problem?
>>
>
>This is an indication of the name clash which has already been mentioned.
>
>There are two (at least) function with the same name in iPlanet and
>OpenSSL MD5_Update and SHA1_Update. The structures they expect are
>different and so when OpenSSL calls the iPlanet one it crashes.
>
>You may be able to use some compilation options to get your application
>to use the OpenSSL versions of the functions, or alternatively you
>can change the names of the offending functions, e.g. with a few
>#defines in the header files.
>
>Steve.
>--
>Dr. Stephen Henson      [EMAIL PROTECTED]            
>OpenSSL Project         http://www.openssl.org/~steve/
>______________________________________________________________________
>OpenSSL Project                                 http://www.openssl.org
>User Support Mailing List                    [EMAIL PROTECTED]
>Automated List Manager                           [EMAIL PROTECTED]
>
>
>
>


______________________________________________________________________
OpenSSL Project                                 http://www.openssl.org
User Support Mailing List                    [EMAIL PROTECTED]
Automated List Manager                           [EMAIL PROTECTED]

Reply via email to