Re: 0.9.6 incompatible with 0.9.5a on Win32

[Jeffrey Altman]

> 
> jaltman> What my code does is call RAND_status() to determine whether
> jaltman> or not random data needs to be computed.  Does RAND_status()
> jaltman> no longer simply provide a test, but instead performs entropy
> jaltman> gathering?
> 
> Not "instead".  "Additionally".  And it has done so for a long time
> (it does in 0.9.5a, but we didn't have the new RAND_poll() in place
> then).
> 

Ok.  Built a debug version.  My code first calls RAND_screen()
immediately after loading the OpenSSL DLLs.  This function calls
RAND_poll().  The call to RAN_poll() takes 6 seconds.

Then I call RAND_status() which in turn calls RAND_poll().  This
second time the call takes over 60 seconds.  This appears to be caused
by the walking of the heap.  Between the time the RAND_screen() is
called  and RAND_status() is called a large amount of memory is
allocated (most of it uninitialized at this point.)   Do you really
want to search through 14MB of data each time this routine is called?



                  Jeffrey Altman * Sr.Software Designer
                 The Kermit Project * Columbia University
               612 West 115th St * New York, NY * 10025 * USA
     http://www.kermit-project.org/ * [EMAIL PROTECTED]


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