Re: [PATCH v2 3/5] firmware: arm_scmi: Add message dump traces for bad and unexpected replies

[Sudeep Holla]

On Tue, Mar 26, 2024 at 11:57:23AM +0000, Cristian Marussi wrote:
> On Tue, Mar 26, 2024 at 11:24:38AM +0000, Sudeep Holla wrote:
> > On Mon, Mar 25, 2024 at 08:46:18PM +0000, Cristian Marussi wrote:
> > > Trace also late-timed-out, out-of-order and unexpected/spurious messages.
> > > 
> > > Signed-off-by: Cristian Marussi <cristian.marussi@xxxxxxx>
> > > ---
> > >  drivers/firmware/arm_scmi/driver.c  | 10 ++++++++++
> > >  drivers/firmware/arm_scmi/mailbox.c |  4 +++-
> > >  2 files changed, 13 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/drivers/firmware/arm_scmi/driver.c b/drivers/firmware/arm_scmi/driver.c
> > > index 7fc1c5b1a2a4..207ed1a52d69 100644
> > > --- a/drivers/firmware/arm_scmi/driver.c
> > > +++ b/drivers/firmware/arm_scmi/driver.c
> > > @@ -861,6 +861,9 @@ scmi_xfer_command_acquire(struct scmi_chan_info *cinfo, u32 msg_hdr)
> > >  			"Message for %d type %d is not expected!\n",
> > >  			xfer_id, msg_type);
> > >  		spin_unlock_irqrestore(&minfo->xfer_lock, flags);
> > > +
> > > +		scmi_bad_message_trace(cinfo, msg_hdr, MSG_UNEXPECTED);
> > > +
> > >  		return xfer;
> > >  	}
> > >  	refcount_inc(&xfer->users);
> > > @@ -885,6 +888,9 @@ scmi_xfer_command_acquire(struct scmi_chan_info *cinfo, u32 msg_hdr)
> > >  		dev_err(cinfo->dev,
> > >  			"Invalid message type:%d for %d - HDR:0x%X  state:%d\n",
> > >  			msg_type, xfer_id, msg_hdr, xfer->state);
> > > +
> > > +		scmi_bad_message_trace(cinfo, msg_hdr, MSG_INVALID);
> > > +
> > >  		/* On error the refcount incremented above has to be dropped */
> > >  		__scmi_xfer_put(minfo, xfer);
> > >  		xfer = ERR_PTR(-EINVAL);
> > > @@ -921,6 +927,9 @@ static void scmi_handle_notification(struct scmi_chan_info *cinfo,
> > >  	if (IS_ERR(xfer)) {
> > >  		dev_err(dev, "failed to get free message slot (%ld)\n",
> > >  			PTR_ERR(xfer));
> > > +
> > > +		scmi_bad_message_trace(cinfo, msg_hdr, MSG_NOMEM);
> > > +
> > >  		scmi_clear_channel(info, cinfo);
> > >  		return;
> > >  	}
> > > @@ -1040,6 +1049,7 @@ void scmi_rx_callback(struct scmi_chan_info *cinfo, u32 msg_hdr, void *priv)
> > >  		break;
> > >  	default:
> > >  		WARN_ONCE(1, "received unknown msg_type:%d\n", msg_type);
> > > +		scmi_bad_message_trace(cinfo, msg_hdr, MSG_UNKNOWN);
> > >  		break;
> > >  	}
> > >  }
> > > diff --git a/drivers/firmware/arm_scmi/mailbox.c b/drivers/firmware/arm_scmi/mailbox.c
> > > index b8d470417e8f..fb0824af7180 100644
> > > --- a/drivers/firmware/arm_scmi/mailbox.c
> > > +++ b/drivers/firmware/arm_scmi/mailbox.c
> > > @@ -56,7 +56,9 @@ static void rx_callback(struct mbox_client *cl, void *m)
> > >  	 */
> > >  	if (cl->knows_txdone && !shmem_channel_free(smbox->shmem)) {
> > >  		dev_warn(smbox->cinfo->dev, "Ignoring spurious A2P IRQ !\n");
> > > -		return;
> > > +		return scmi_bad_message_trace(smbox->cinfo,
> > > +				     shmem_read_header(smbox->shmem),
> > > +				     MSG_MBOX_SPURIOUS);
> > 
> > From previous patch, IIUC scmi_bad_message_trace is a void func and doesn't
> > return anything. Did you not get any build error/warning for this ?
> > 
> 
> No...I am building with W=1....but note that the caller itself here
> rx_callback() is supposed to return a void...
> 
> ...in V3 I may just split this into 2 lines and leave the return; alone on its
> own line to avoid any confusion...

Not need to respin unless I find something that needs reposting, can fix this
myself.

-- 
Regards,
Sudeep