SINGLESOCK - How much to strip away?
hmurray at megapathdsl.net
Wed May 30 18:41:32 UTC 2018
As much as I'd like to clean up this area, I think we should put it on the
back burner for now.
Maybe we should put a note in devel/ with a summary of what we have learned.
Actually, I'd like to see a class of notes there. We should use this as a
good example. I'm interested in things that are likely to be too big to just
toss into devel/TODO.
There is a slightly related area that needs cleaning up. There are no
externally visible impacts so there is no reason not to do it. I don't have
a good tag. Maybe RECVBUF.
The current code has a queue of free buffers. When a packet arrives, the IO
code gets a buffer, reads in the data, and puts it on the work queue.
Another loop takes buffers off the work queue and processes them. The
process-it code puts the buffer back on the free queue.
All that queuing stuff is wasted effort. It made sense back when the code
read the data is a signal handler to get better arrival timing. We now get
good timing from the kernel. Bouncing around on queues isn't doing anything
other than keeping the CPU warm.
There may be some rough edges with refclocks. They will be easy to fix.
These are my opinions. I hate spam.
More information about the devel