Re: Page allocator bottleneck

Tariq Toukan [off-list ref] writes:

Congestion in this case is very clear.
When monitored in perf top:
85.58% [kernel] [k] queued_spin_lock_slowpath

Please look at the callers. Spinlock profiles without callers
are usually useless because it's just blaming the messenger.

Most likely the PCP lists are too small for your extreme allocation
rate, so it goes back too often to the shared pool.

You can play with the vm.percpu_pagelist_fraction setting.

-Andi