>
> Sorry for this. Since I assume that you are referring to tests with the qib
> or hfi1 driver and since I do not have access to a setup with that hardware
> I will have to rely on you for further analysis.
I turned on some static tracing and there is a sequence of RNR naks on the wire from the target.
A cursory look at the normal ACK aeth doesn't show any issue with the flow control.
Again, still looking.
I don't know if qib fails, but I suspect it will.
We have nightly 12 hour tests on the lastest stuff that caught it.
>
> > I'm looking at why but...
> >
> > > Add new configfs attributes that allow to configure which mode to use
> > > (/sys/kernel/config/target/srpt/$GUID/$GUID/attrib/use_srq). Note:
> > > only the attribute for port 1 is relevant on multi-port adapters.
> >
> > It looks like the above work around cannot work.
> > The use_srq is tested and the SRQ is allocated when the device is added.
> > Setting a target relative attribute is too late in the sequence to enable
> > the old SRQ usage.
>
> Thanks for the report. I will post a patch series to address this.
>
Ok.
For now I can just manually disarm with a code change to experiment.
Mike
��.n��������+%������w��{.n�����{���fk��ܨ}���Ơz�j:+v�����w����ޙ��&�)ߡ�a����z�ޗ���ݢj��w�f
