If we enter a hang through the normal timed loop of RunMonitorThread, and then
call NotifyWait, it will result in a hang with an indefinite duration being
reported once NotifyActivity is called again.
MozReview-Commit-ID: 4vUip65L0qo
When the BHR wants to report a tardy thread, it hands off the work of
getting a stacktrace to a helper thread. The helper threads used are in the
Stream Transport Service threadpool, which has a default limit of 25
threads.
This has a bad effect when we are in a severely compute-resource constrained
situation. Then, many threads will be late, and up to 25 STS worker threads
will be employed to do unwinding, a potentially expensive operation. This
further restricts the compute resources available to progress the rest of
the system.
Another effect is that the unwinder work will compete against the "real" STS
work for the 25 workers, potentially further slowing forward progress.
This patch replaces the use of the STS thread pool with a single nsThread
dedicated to unwinding/reporting hangs.
This patch was autogenerated by my decomponents.py
It covers almost every file with the extension js, jsm, html, py,
xhtml, or xul.
It removes blank lines after removed lines, when the removed lines are
preceded by either blank lines or the start of a new block. The "start
of a new block" is defined fairly hackily: either the line starts with
//, ends with */, ends with {, <
