Ryan Bloom wrote:
This will work, and should be portable. Just a question, how big a performance improvement is this? Have we hit the point where we are optimizing code just to optimize code?
Based on the profile data, 64-bit divisions *are* a problem, but I think there's a better solution for this one...
I would not support this change, simply because I don't see it making a huge difference. We are better off fixing the apr_time_t implementation and then looking for things like this.
I agree. Once we go to the binary microsecond implementation, what we'll be able to do within poll is: - extract the microseconds with a bitwise AND (encapsulated within apr_time_usec()) - cast to 32 bits if the macro doesn't do so automatically - then do a 32-bit division by 1000 to get milliseconds (which is a lot cheaper than the current 64-bit division on 32-bit platforms)
--Brian
