On 1/3/2011 7:52 PM, Ross Finlayson wrote:
If that's not it, then maybe timeNow isn't always a "sane" value here.
[snip]
could you try commenting out that statement, and let us know if it improves things (on the problematic OS)?
Yes, that returns performance to the pre-2010.12.05 state. Thanks!
this doesn't explain the differences between OSs that you are apparently seeing
[shrug] gettimeofday() is not a simple thing. I just re-read Chapter 6 in Bovet and Cesati, and in kernel 2.6, there are three different tick sources for this syscall, depending on available hardware and things like whether the CPU can throttle down to save power. Toss on top of that tick interpolation and the non-realtime nature of stock Linux... There's plenty of room for randomness.
Would more files help you to understand what's going on here? As I said, we have plenty to choose from.
_______________________________________________ live-devel mailing list live-devel@lists.live555.com http://lists.live555.com/mailman/listinfo/live-devel
