Re: [PATCH 1/3] scsi/libfc: Remove get_cpu() semantics in fc_exch_em_alloc()

From: Sebastian Andrzej Siewior
Date: Wed Nov 17 2021 - 02:25:54 EST

On 2021-11-16 18:59:54 [-0800], Davidlohr Bueso wrote:
> The get_cpu() in fc_exch_em_alloc() was introduced in:
> f018b73af6db ([SCSI] libfc, libfcoe, fcoe: use smp_processor_id() only when preempt disabled)
> for no other reason than to simply use smp_processor_id()
> without getting a warning, because everything is done with
> the pool->lock held anyway. However, get_cpu(), by disabling
> preemption, does not play well with PREEMPT_RT, particularly
> when acquiring a regular (and thus sleepable) spinlock.
> Therefore remove the get_cpu() and just use the unstable value
> as we will have CPU locality guarantees next by taking the lock.
> The window of migration, as noted by Sebastian, is small and
> even if it happens the result is correct.
> Signed-off-by: Davidlohr Bueso <dbueso@xxxxxxx>
Acked-by: Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx>

> diff --git a/drivers/scsi/libfc/fc_exch.c b/drivers/scsi/libfc/fc_exch.c
> index 841000445b9a..be37bfb2814d 100644
> --- a/drivers/scsi/libfc/fc_exch.c
> +++ b/drivers/scsi/libfc/fc_exch.c
> @@ -825,10 +825,9 @@ static struct fc_exch *fc_exch_em_alloc(struct fc_lport *lport,
> }
> memset(ep, 0, sizeof(*ep));
> - cpu = get_cpu();
> + cpu = raw_smp_processor_id();
> pool = per_cpu_ptr(mp->pool, cpu);
> spin_lock_bh(&pool->lock);
> - put_cpu();

The `cpu' variable is later ORed into ep->oxid. I haven't figured out
why this is important/ required. The ep variable/ `fc_exch' is allocated
from a per-CPU memory pool and is later released to the same pool but
the pool's CPU number is not obtained from fc_exch::oxid but there is a
pointer to its pool stored in fc_exch::em.
So it remains a mystery to why the CPU number is stored here.

> /* peek cache of free slot */
> if (pool->left != FC_XID_UNKNOWN) {