[PATCH] raw/ntb: add check for DB intr handler registering

[Guo, Junfeng]

> -----Original Message-----
> From: Thomas Monjalon <thomas at monjalon.net>
> Sent: Tuesday, March 8, 2022 01:07
> To: Wu, Jingjing <jingjing.wu at intel.com>; Guo, Junfeng
> <junfeng.guo at intel.com>
> Cc: dev at dpdk.org; stable at dpdk.org
> Subject: Re: [PATCH] raw/ntb: add check for DB intr handler registering
> 
> 10/02/2022 11:43, Guo, Junfeng:
> >
> > > -----Original Message-----
> > > From: Guo, Junfeng
> > > Sent: Thursday, February 10, 2022 15:18
> > > To: Wu, Jingjing <jingjing.wu at intel.com>
> > > Cc: dev at dpdk.org; stable at dpdk.org
> > > Subject: RE: [PATCH] raw/ntb: add check for DB intr handler
> registering
> > >
> > >
> > >
> > > > -----Original Message-----
> > > > From: Wu, Jingjing <jingjing.wu at intel.com>
> > > > Sent: Thursday, February 10, 2022 15:05
> > > > To: Guo, Junfeng <junfeng.guo at intel.com>
> > > > Cc: dev at dpdk.org; stable at dpdk.org
> > > > Subject: RE: [PATCH] raw/ntb: add check for DB intr handler
> registering
> > > >
> > > >
> > > >
> > > > > -----Original Message-----
> > > > > From: Guo, Junfeng <junfeng.guo at intel.com>
> > > > > Sent: Thursday, February 10, 2022 2:29 PM
> > > > > To: Wu, Jingjing <jingjing.wu at intel.com>
> > > > > Cc: dev at dpdk.org; stable at dpdk.org; Guo, Junfeng
> > > > <junfeng.guo at intel.com>
> > > > > Subject: [PATCH] raw/ntb: add check for DB intr handler
> registering
> > > > >
> > > > > The callback registering of doorbell interrupt handler should be
> > > > > finished before enabling the interrupt event fd. Thus add the return
> > > > > value check for this callback registering.
> > > > >
> > > > > Fixes: 62012a76811e ("raw/ntb: add handshake process")
> > > > > Cc: stable at dpdk.org
> > > > >
> > > > > Signed-off-by: Junfeng Guo <junfeng.guo at intel.com>
> > > > > ---
> > > > >  drivers/raw/ntb/ntb.c | 8 ++++++--
> > > > >  1 file changed, 6 insertions(+), 2 deletions(-)
> > > > >
> > > > > diff --git a/drivers/raw/ntb/ntb.c b/drivers/raw/ntb/ntb.c
> > > > > index cc611dfbb9..0801e6d1ae 100644
> > > > > --- a/drivers/raw/ntb/ntb.c
> > > > > +++ b/drivers/raw/ntb/ntb.c
> > > > > @@ -1403,8 +1403,12 @@ ntb_init_hw(struct rte_rawdev *dev,
> > > struct
> > > > rte_pci_device
> > > > > *pci_dev)
> > > > >
> > > > >  	intr_handle = pci_dev->intr_handle;
> > > > >  	/* Register callback func to eal lib */
> > > > > -	rte_intr_callback_register(intr_handle,
> > > > > -				   ntb_dev_intr_handler, dev);
> > > > > +	ret = rte_intr_callback_register(intr_handle,
> > > > > +					 ntb_dev_intr_handler,
> dev);
> > > > > +	if (ret) {
> > > > > +		NTB_LOG(ERR, "Unable to register doorbell intr
> > > > handler.");
> > > > > +		return ret;
> > > > > +	}
> > > > When will this register failure happen? Have you checked what is the
> > > root
> > > > cause?
> > >
> > > When bind with vfio-pci, the DB interrupt callback function often
> cannot
> > > work as expected.
> > > I think this is due to that, the intr callback registering not finished
> before
> > > event fd enabled.
> > > Add the check of the return value here can make sure about correct
> > > processing sequence.
> >
> > I think some compiler optimization would lead to this condition.
> > So add return value check can prevent this (i.e., registering not finished).
> 
> There is no formal ack. What is the status of this patch?
> 

Thanks for the reminding!
Currently, we agreed that this fix is not for the root cause of no response of doorbell
interrupt when building the connection between two NTB ports. So we decided to
change the status to be superseded and do further investigation for NTB.
Thanks for the reviewing!

> 
>