Add an example explaining the rationale that the limitation that old lockdep implies, can be relaxed. Signed-off-by: Byungchul Park <byungchul.park@xxxxxxx> --- Documentation/locking/crossrelease.txt | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) diff --git a/Documentation/locking/crossrelease.txt b/Documentation/locking/crossrelease.txt index bb449e8..dac56f4 100644 --- a/Documentation/locking/crossrelease.txt +++ b/Documentation/locking/crossrelease.txt @@ -281,6 +281,29 @@ causes a deadlock. The more lockdep adds dependencies, the more it thoroughly works. Thus, lockdep has to do its best to detect and add as many true dependencies to the graph as possible. +For example: + + CONTEXT X CONTEXT Y + --------- --------- + acquire A + acquire B /* A dependency 'A -> B' exists */ + release B + release A held by Y + + where A and B are different lock classes. + +In this case, a dependency 'A -> B' exists since: + + 1. A waiter for A and a waiter for B might exist when acquiring B. + 2. The only way to wake up each is to release what it waits for. + 3. Whether the waiter for A can be woken up depends on whether the + other can. In other words, CONTEXT X cannot release A if it fails + to acquire B. + +Considering only typical locks, lockdep builds nothing. However, +relaxing the limitation, a dependency 'A -> B' can be added, giving us +more chances to check circular dependencies. + However, it might suffer performance degradation since relaxing the limitation, with which design and implementation of lockdep can be efficient, might introduce inefficiency inevitably. So lockdep -- 1.9.1
