On 09/17/2018 11:22 PM, jbullock wrote:
The developers / maintainers are talking "race condition" in the bug notes. You gotta get pretty granular and on point to provide info to unwind one of those.
Anybody have an idea what factoids might help them run this down?
There isn't any "silver-bullet" factoid that will help the devs correct a race condition. It is a code screwup somewhere in 800,000 lines of multi-threaded (or otherwise parallel executed) code that leads to two treads attempting to access a shared bit of memory (or shared value) and there is some fault with the locking (or condition locks) so that a required thread is effectively prevented from executing the code it needs to run to complete whatever it is trying to do. If things are not moving quickly on the bug, it is because the code involved may be voluminous, what condition or which of the myriad of locks is the cause is anything but clear, and it will take many hours of tedious debugging to find out where in the daisy-chain of locking logic the problem is hiding. You can read more about what a "data race" is and what a "race condition" is here: What is a race condition? https://stackoverflow.com/questions/34510/what-is-a-race-condition -- David C. Rankin, J.D.,P.E. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org