On 15/02/2017 13:10, Saeed Mahameed wrote:
On Fri, Feb 10, 2017 at 2:27 PM, Eric Dumazet <eric.duma...@gmail.com> wrote:
From: Eric Dumazet <eduma...@google.com>
All rx and rx netdev interrupts are handled by respectively
by mlx4_en_rx_irq() and mlx4_en_tx_irq() which simply schedule a NAPI.
But mlx4_eq_int() also fires a tasklet to service all items that were
queued via mlx4_add_cq_to_tasklet(), but this handler was not called
unless user cqe was handled.
This is very confusing, as "mpstat -I SCPU ..." show huge number of
tasklet invocations.
This patch saves this overhead, by carefully firing the tasklet directly
from mlx4_add_cq_to_tasklet(), removing four atomic operations per IRQ.
So, in case of RDMA CQs, we add some per-CQE overhead of comparing the
list pointers and condition upon that. Maybe we could add an
invoke_tasklet boolean field on mlx4_cq and return its value from
mlx4_cq_completion.
That's way we could do invoke_tasklet |= mlx4_cq_completion(....);
Outside the while loop we could just
if (invoke_tasklet)
tasklet_schedule
Anyway, I guess that even with per-CQE overhead, the performance impact
here is pretty negligible - so I guess that's fine too :)
