Found the bug myself! Change ST(1) in SSLeay.xs:2142 to ST(0), this is presumably the id of the $buffer parameter... SSLeay.xs:2142: - char * buf = SvPV( ST(1), len); + char * buf = SvPV( ST(0), len); > Hi all > > I've run into a problem with Net::SSLeay::RAND_seed($buf). > > I'm working on WinNT4 with Net::SSLeay 1.05 and OpenSSL 0.9.5. > (I'll be posting instructions how to get those two to work > together on NT as soon as the Net::SSLeay homepage is up again) > > Since OpenSSL 0.9.5 complains when the PRNG is not seeded I'm > using RAND_seed to provide some initial data. Because this > didn't seem to help while debugging my perl scripts I built > debug versions of Net::SSLeay and OpenSSL and tried to track it down. > > I found that <openssl>/crypto/rand/rand_lib.c/RAND_seed(const > void *buf, int num) ALWAYS gets called with a buffer pointing > to the value "*Net::SSLeay::RAND_seed" and num = 23!!! > (Except when run in the ActiveState debugger; then the buffer > always points to "10000" and num=5, which is definetely to > few bytes, so OpenSSL complains). > > There seems to be a problem with the interface from > Net::SSLeay::RAND_seed() to OpenSSL:RAND_seed(). Does anybody > know how to fix this? > > > Thanks Marcel > > (PS: Sampo, is there any Net::SSLeay 1.0.6 in the queue?) > ______________________________________________________________________ OpenSSL Project http://www.openssl.org User Support Mailing List [EMAIL PROTECTED] Automated List Manager [EMAIL PROTECTED]
