Modify the the threads callback error to store the error

when the error comes from the measurement but not from some bug in the code. This was proposed by teor in #28868 (moved):

Modify result_putter_error() to store an error result to the queue.

changed milestone to %sbws: 1.0.x-final

Trac:
Parent Ticket: #28663 (moved)

added component::core tor/sbws milestone::sbws: 1.0.x-final parent::28663 priority::medium resolution::fixed severity::normal status::closed type::defect version::sbws 1.0.2 labels

This has been implemented as part of https://trac.torproject.org/projects/tor/ticket/28567#comment:5. Not creating a new branch for this ticket and cherry-picking the implementation because there'll be already many merge conflicts since the same parts of the code are being modified in different tickets. When it's confirmed sbws doesn't stall without implementing this, this ticket can be closed.

Change parent to #28663 (moved) so that #28868 (moved) can be closed.

Trac:
Parent: #28868 (moved) to #28663 (moved)

The error callback logs only exceptions when the callback could not be call, what means developer bugs So i think it's better to print the stack, which is not in the main process but traceback.print_stack takes from the thread. And this is already done in 8cbb73052240f685b409a89df910b915228c651c.

Trac:
Resolution: N/A to fixed
Status: new to closed

closed

moved to tpo/network-health/sbws#29157 (closed)