Date:   Tue, 14 Feb 2023 15:34:21 -0800
From:   Bjorn Andersson <quic_bjorande@quicinc.com>
To:     Chris Lew <quic_clew@quicinc.com>
CC:     Bjorn Andersson <andersson@kernel.org>,
        Mathieu Poirier <mathieu.poirier@linaro.org>,
        Sricharan Ramabadhran <quic_srichara@quicinc.com>,
        Arun Kumar Neelakantam <quic_aneelaka@quicinc.com>,
        <linux-arm-msm@vger.kernel.org>,
        <linux-remoteproc@vger.kernel.org>, <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] rpmsg: glink: Avoid infinite loop on intent for missing
 channel
Message-ID: <20230214233421.GB1332049@hu-bjorande-lv.qualcomm.com>
References: <20230109223801.1706213-1-quic_bjorande@quicinc.com>
 <393b7d5f-b565-6287-bbab-3bf51cc981f0@quicinc.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Disposition: inline
In-Reply-To: <393b7d5f-b565-6287-bbab-3bf51cc981f0@quicinc.com>
Precedence: bulk

On Tue, Feb 14, 2023 at 02:55:02PM -0800, Chris Lew wrote:
> On 1/9/2023 2:38 PM, Bjorn Andersson wrote:
> > In the event that an intent advertisement arrives on an unknown channel
> > the fifo is not advanced, resulting in the same message being handled
> > over and over.
> > 
> > Fixes: dacbb35e930f ("rpmsg: glink: Receive and store the remote intent buffers")
> > Signed-off-by: Bjorn Andersson <quic_bjorande@quicinc.com>
> > ---
> >   drivers/rpmsg/qcom_glink_native.c | 5 +++--
> >   1 file changed, 3 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/rpmsg/qcom_glink_native.c b/drivers/rpmsg/qcom_glink_native.c
> > index f36740cb6866..7b1320b1579e 100644
> > --- a/drivers/rpmsg/qcom_glink_native.c
> > +++ b/drivers/rpmsg/qcom_glink_native.c
> > @@ -946,12 +946,12 @@ static void qcom_glink_handle_intent(struct qcom_glink *glink,
> >   	spin_unlock_irqrestore(&glink->idr_lock, flags);
> >   	if (!channel) {
> >   		dev_err(glink->dev, "intents for non-existing channel\n");
> > -		return;
> > +		goto advance_rx;
> >   	}
> >   	msg = kmalloc(msglen, GFP_ATOMIC);
> >   	if (!msg)
> > -		return;
> > +		goto advance_rx;
> 
> 
> Should we be dropping the packet for this case? If we try again later more
> memory might be available to handle the command.
> 

You're right, we found a channel above, but we don't have enough memory
to handle the message right now. That seems like a message worth not
throwing away.

Thanks,
Bjorn

> >   	qcom_glink_rx_peak(glink, msg, 0, msglen);
> > @@ -973,6 +973,7 @@ static void qcom_glink_handle_intent(struct qcom_glink *glink,
> >   	}
> >   	kfree(msg);
> > +advance_rx:
> >   	qcom_glink_rx_advance(glink, ALIGN(msglen, 8));
> >   }
> >