On 2025-04-29 21:37:26 [-0700], Shakeel Butt wrote: > > > - if (gfpflags_allow_spinning(gfp_mask)) > > > - local_lock_irqsave(&memcg_stock.stock_lock, flags); > > > - else if (!local_trylock_irqsave(&memcg_stock.stock_lock, flags)) > > > + if (nr_pages > MEMCG_CHARGE_BATCH || > > > + !local_trylock_irqsave(&memcg_stock.stock_lock, flags)) > > > > I don't think it's a good idea. > > spin_trylock() will fail often enough in PREEMPT_RT. > > Even during normal boot I see preemption between tasks and they > > contend on the same cpu for the same local_lock==spin_lock. > > Making them take slow path is a significant behavior change > > that needs to be carefully considered. > > I didn't really think too much about PREEMPT_RT kernels as I assume > performance is not top priority but I think I get your point. Let me Not sure if this is performance nor simply failing to allocate memory. > explain and correct me if I am wrong. On PREEMPT_RT kernel, the local > lock is a spin lock which is actually a mutex but with priority > inheritance. A task having the local lock can still get context switched > (but will remain on same CPU run queue) and the newer task can try to > acquire the memcg stock local lock. If we just do trylock, it will > always go to the slow path but if we do local_lock() then it will sleeps > and possibly gives its priority to the task owning the lock and possibly > make that task to get the CPU. Later the task slept on memcg stock lock > will wake up and go through fast path. So far correct. On PREEMPT_RT a task with spinlock_t or local_lock_t can get preempted while owning the lock. The local_lock_t is a per-CPU lock. Sebastian
