Re: [PATCH v3] locking/osq_lock: Ensure proper locking semantics for osq_lock/osq_unlock()

[Peter Zijlstra]

On Tue, Sep 15, 2026 at 01:53:05PM -0400, Waiman Long wrote:
> The main reason for this patch is for addressing the locking test failure
> reported by Håkon due to missing barrier.

What crackpot test is that? osq isn't meant to be used as a stand alone
lock.