[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <00dd01d58ee6$a8ce43d0$fa6acb70$@samsung.com>
Date: Wed, 30 Oct 2019 11:25:20 +0530
From: "pankj.sharma" <pankj.sharma@...sung.com>
To: "'Marc Kleine-Budde'" <mkl@...gutronix.de>
Cc: <wg@...ndegger.com>, <davem@...emloft.net>,
<eugen.hristev@...rochip.com>, <ludovic.desroches@...rochip.com>,
<pankaj.dubey@...sung.com>, <rcsekar@...sung.com>,
<jhofstee@...tronenergy.com>, <simon.horman@...ronome.com>,
"'Sriram Dash'" <sriram.dash@...sung.com>,
<linux-can@...r.kernel.org>, <netdev@...r.kernel.org>,
<linux-kernel@...r.kernel.org>
Subject: RE: [PATCH v2] can: m_can: add support for handling arbitration
error
> From: Marc Kleine-Budde <mkl@...gutronix.de>
> Subject: Re: [PATCH v2] can: m_can: add support for handling arbitration error
>
> On 10/21/19 2:13 PM, Pankaj Sharma wrote:
> > The Bosch MCAN hardware (3.1.0 and above) supports interrupt flag to
> > detect Protocol error in arbitration phase.
> >
> > Transmit error statistics is currently not updated from the MCAN driver.
> > Protocol error in arbitration phase is a TX error and the network
> > statistics should be updated accordingly.
> >
> > The member "tx_error" of "struct net_device_stats" should be
> > incremented as arbitration is a transmit protocol error. Also
> > "arbitration_lost" of "struct can_device_stats" should be incremented
> > to report arbitration lost.
> >
> > Signed-off-by: Pankaj Sharma <pankj.sharma@...sung.com>
> > Signed-off-by: Sriram Dash <sriram.dash@...sung.com>
> > ---
> >
> > changes in v2:
> > - common m_can_ prefix for is_protocol_err function
> > - handling stats even if the allocation of the skb fails
> > - resolving build errors on net-next branch
> >
> > drivers/net/can/m_can/m_can.c | 37
> > +++++++++++++++++++++++++++++++++++
> > 1 file changed, 37 insertions(+)
> >
> > diff --git a/drivers/net/can/m_can/m_can.c
> > b/drivers/net/can/m_can/m_can.c index 75e7490c4299..a736297a875f
> > 100644
> > --- a/drivers/net/can/m_can/m_can.c
> > +++ b/drivers/net/can/m_can/m_can.c
> > @@ -778,6 +778,38 @@ static inline bool is_lec_err(u32 psr)
> > return psr && (psr != LEC_UNUSED);
> > }
> >
> > +static inline bool m_can_is_protocol_err(u32 irqstatus) {
> > + return irqstatus & IR_ERR_LEC_31X;
> > +}
> > +
> > +static int m_can_handle_protocol_error(struct net_device *dev, u32
> > +irqstatus) {
> > + struct net_device_stats *stats = &dev->stats;
> > + struct m_can_classdev *cdev = netdev_priv(dev);
> > + struct can_frame *cf;
> > + struct sk_buff *skb;
> > +
> > + /* propagate the error condition to the CAN stack */
> > + skb = alloc_can_err_skb(dev, &cf);
> > + if (unlikely(!skb)) {
> > + netdev_dbg(dev, "allocation of skb failed\n");
> > + stats->tx_errors++;
> > + return 0;
> > + }
> > + if (cdev->version >= 31 && (irqstatus & IR_PEA)) {
> > + netdev_dbg(dev, "Protocol error in Arbitration fail\n");
> > + stats->tx_errors++;
> > + cdev->can.can_stats.arbitration_lost++;
>
> If the skb allocation fails, you miss the stats here.
Alright. We shall handle the stats even when skb fails.
Shall post in upcoming revision.
>
> > + cf->can_id |= CAN_ERR_LOSTARB;
> > + cf->data[0] |= CAN_ERR_LOSTARB_UNSPEC;
> > + }
> > +
> > + netif_receive_skb(skb);
> > +
> > + return 1;
> > +}
> > +
> > static int m_can_handle_bus_errors(struct net_device *dev, u32 irqstatus,
> > u32 psr)
> > {
> > @@ -792,6 +824,11 @@ static int m_can_handle_bus_errors(struct
> net_device *dev, u32 irqstatus,
> > is_lec_err(psr))
> > work_done += m_can_handle_lec_err(dev, psr & LEC_UNUSED);
> >
> > + /* handle protocol errors in arbitration phase */
> > + if ((cdev->can.ctrlmode & CAN_CTRLMODE_BERR_REPORTING) &&
> > + m_can_is_protocol_err(irqstatus))
> > + work_done += m_can_handle_protocol_error(dev, irqstatus);
> > +
> > /* other unproccessed error interrupts */
> > m_can_handle_other_err(dev, irqstatus);
>
> Marc
>
> --
> Pengutronix e.K. | Marc Kleine-Budde |
> Industrial Linux Solutions | Phone: +49-231-2826-924 |
> Vertretung West/Dortmund | Fax: +49-5121-206917-5555 |
> Amtsgericht Hildesheim, HRA 2686 |
> https://protect2.fireeye.com/url?k=cb5c4e6130fb99cd.cb5dc52e-
> 7d616d604aa6cbcf&u=http://www.pengutronix.de |
Powered by blists - more mailing lists