Michael wrote:
If I use /dev/random in my patch the unit tests take more than three minutes without any user interaction. As to be expected, it runs faster if I move the mouse around wildly to generate entropy =>
You're right, this seems unreasonable for 32 bytes of random data, especially in this context. Alexandre has the final word anyway, so submitting to wine-patches was a good idea. Again, sorry I didn't see all this up front, but I'd like to suggest the following improvement:
If OpenSSL is available, go ahead and use RAND_bytes, but call RAND_add first with the input buffer. I don't have a good estimate for the number of bits of entropy to expect, since we don't know whether the caller will actually have provided a seed. Assuming it's unlikely, maybe RAND_add(pbBuffer, dwLen, dwLen / 16.0f)?
And go ahead and fall back to /dev/urandom if OpenSSL isn't available.
Thanks, --Juan
__________________________________ Do you Yahoo!? Yahoo! Mail Address AutoComplete - You start. We finish. http://promotions.yahoo.com/new_mail