Comment 5 for bug 2018504

Revision history for this message
James Henstridge (jamesh) wrote :

I was probably stracing that thread #0. It looks like the timeout in question can be added with a zero second delay, which would match up with the poll calls with a zero timeout:

https://github.com/OpenPrinting/cups-browsed/blob/1921fa792c01278068f5e19535c2479959187b09/daemon/cups-browsed.c#L9650-L9691

The update_cups_queues function ends by calling recheck_timer() again, likely scheduling another zero second timeout right away.