I am not sure locks are the answer here.
If two distinct entities want to control the thread count, there will be
issues no matter what, regardless of whether changing the thread count is
atomic. And I don't really see a scenario where two cooperative entities
control the thread count cooperatively; why not just let one of them do the
changing?
On Fri, 29 Apr 2022, Henry Rich wrote:
Agreed we will need locks.
Henry Rich
On 4/29/2022 2:27 PM, Raul Miller wrote:
I should also point out that a bounded 0 T. '' mechanism -- one which
creates threads up to a desired limit -- would need a reliable locking
mechanism which would stall all other uses of T. for counting or
creating threads until the desired number of threads has been created.
Without that, we must cope with creating a number of threads based on
an inaccurate thread count.
FYI,
--
This email has been checked for viruses by AVG.
https://www.avg.com
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
