Re: [PATCH blktests v5 2/4] srp/rc: replace module removal with patient module removal

[Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com>]

On Dec 26, 2025 / 10:08, Bart Van Assche wrote:
> On 12/25/25 1:09 PM, Shin'ichiro Kawasaki wrote:
> > Suggested-by: Bart Van Assche <bvanassche@acm.org>
> 
> I can't remember that I suggested to use _patient_rmmod. Luis, can you
> provide a link that shows that I suggested to use _patient_rmmod for the
> SRP tests? If not, please leave out the Suggested-by tag.

Based on your another comment below, I think this patch can be dropped. If this
is the case, no need to worry about the Suggested-by tag.

> 
> > diff --git a/tests/srp/rc b/tests/srp/rc
> > index 47b9546..2d3d615 100755
> > --- a/tests/srp/rc
> > +++ b/tests/srp/rc
> > @@ -331,19 +331,10 @@ start_srp_ini() {
> >   # Unload the SRP initiator driver.
> >   stop_srp_ini() {
> > -	local i
> > -
> >   	log_out
> > -	for ((i=40;i>=0;i--)); do
> > -		remove_mpath_devs || return $?
> > -		_unload_module ib_srp >/dev/null 2>&1 && break
> > -		sleep 1
> > -	done
> > -	if [ -e /sys/module/ib_srp ]; then
> > -		echo "Error: unloading kernel module ib_srp failed"
> > -		return 1
> > -	fi
> > -	_unload_module scsi_transport_srp || return $?
> > +	remove_mpath_devs || return $?
> > +	_patient_rmmod ib_srp || return 1
> > +	_patient_rmmod scsi_transport_srp || return $?
> >   }
> 
> The loop should be around both remove_mpath_devs and _unload_module
> because multipathd may add new paths concurrently with the for-loop.
> Please back out the above changes from this patch.

I see. If we back out the changes above, this patch does not have much
meaning. Just replacing the _unload_module calls with _patient_rmmod calls. I
will drop this patch in the next spin. In stead, I will modify the first patch
to replace all of _unload_module calls in srp/rc with _patient_rmmod calls.