RAND_screen() thread safety and RAND_poll() initialization

2009-07-23 Thread Tanguy Fautré
Hi, Using OpenSSL 0.9.8j via LibCurl 7.19.5, we've discovered some random crashes and deadlocks that, after further investigations, led us to a call to RAND_screen() from LibCurl. Two things puzzles me about the implementation of RAND_screen(), hinting that it could maybe be a thread safety i

Re: RAND_poll() and CreateToolhelp32Snapshot() stability

2009-03-11 Thread Tanguy Fautré
Hi Ger, I've tried the patch on 0.9.8j. All seems fine in theory and in practice. So I'm all for it. Cheers, Tanguy Ger Hobbelt wrote: Would the attached patch be agreeable to both of you? It contains a fix (okay, it's a hack) to ensure OpenSSL doesn't loop infinitely on crashing Heap32

Re: RAND_poll() and CreateToolhelp32Snapshot() stability

2009-03-04 Thread Tanguy Fautré
Hi Martin, Martin Simmons wrote: Because the data captured by CreateToolhelp32Snapshot is static and the system is dynamic, use try-excepts around the APIs that access this data. It could be genuinely different on WinCE. Making a snapshot might require lots of memory, which is typically a

RAND_poll() and CreateToolhelp32Snapshot() stability

2009-03-02 Thread Tanguy Fautré
Hi, We've been observing in our application several crashes on Windows related to RAND_poll(). We've been working on this issue for 3 days now, and came up with a possible explanation and fix. Bare with me on this rather lengthy email, as I'll try to document as best I can everything we've done