RE: [Algorithms] Locking without atomic operations
Brought to you by:
vexxed72
From: Kelly B. <kb....@sb...> - 2006-05-31 08:29:18
|
Don't suppose you'd be willing to give up your testing code? :) > Ahem, no. Due to the following checks one or both threads > will fail the lock later. I really checked that for _every_ > execution order. Not just one in a million, for the 5-step > case, there are 17153136 legal permutations. All tested. |