diff options
author | Paul E. McKenney <paulmck@kernel.org> | 2024-05-02 07:16:53 -0700 |
---|---|---|
committer | Paul E. McKenney <paulmck@kernel.org> | 2024-05-02 07:16:53 -0700 |
commit | 678e6cc3c0f18ca5ca361925e5696fcbf7a7c60f (patch) | |
tree | 8b88b2c9ecfafbeaff6bc26dc53477681bbf28e8 | |
parent | b2388ed23a901770d84dc3801dca2977fb7ebea3 (diff) | |
download | perfbook-678e6cc3c0f18ca5ca361925e5696fcbf7a7c60f.tar.gz |
locking: Cite Dice et al. for hierarchical lock
Reported-by: Neeraj Upadhyay <Neeraj.Upadhyay@amd.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
-rw-r--r-- | locking/locking.tex | 5 |
1 files changed, 5 insertions, 0 deletions
diff --git a/locking/locking.tex b/locking/locking.tex index c87e01f2..e4ebce51 100644 --- a/locking/locking.tex +++ b/locking/locking.tex @@ -2037,6 +2037,11 @@ steps to avoid starvation~\cite{McKenney02e,radovic03hierarchical,radovic02efficient,BenJackson02,McKenney02d}. Many of these can be thought of as analogous to the elevator algorithms traditionally used in scheduling disk I/O. +Dice et al.~discuss use of local and global locks in order to transform +any architecture-oblivious lock into a local/global locking scheme +that optimizes for the system's structure by providing (for example) +per-socket locks along with a global +lock~\cite{DavidDice2012NUMAlock1,DavidDice2012NUMAlock2}. Unfortunately, the same scheduling logic that improves the \IX{efficiency} of queued locks at high contention also increases their overhead at |