Comment by mwcampbell
3 years ago
> So calling another process can be used almost like a subroutine call.
Any numbers on the overhead of a message send/reply round trip compared to a subroutine call? I always assumed it was just axiomatic that the difference in latency between those two options would be orders of magnitude.
I don't have numbers, but the key ingredient is really minimizing context switches. This also results in priority inheritance while message passing which means that a high priority process gets the answer with it's own high priority, even if sending a message to a low priority process. Very smart.
Teading the old open sourced sources can be quite interesting...
https://github.com/vocho/openqnx/tree/master/trunk/lib/qnx43...
Here are some numbers from QNX. These are old.[1]
[1] http://support.qnx.com/developers/docs/6.5.0/index.jsp?topic...