Date:   Fri, 7 Oct 2022 17:58:59 +0200
From:   Vincent Guittot <vincent.guittot@...aro.org>
To:     Cristian Marussi <cristian.marussi@....com>
Cc:     linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
        sudeep.holla@....com, james.quinlan@...adcom.com,
        Jonathan.Cameron@...wei.com, f.fainelli@...il.com,
        etienne.carriere@...aro.org, souvik.chakravarty@....com,
        wleavitt@...vell.com, peter.hilber@...nsynergy.com,
        nicola.mazzucato@....com, tarek.el-sherbiny@....com
Subject: Re: [PATCH v3 0/9] Introduce a unified API for SCMI Server testing

On Fri, 7 Oct 2022 at 17:37, Cristian Marussi <cristian.marussi@....com> wrote:
>
> On Fri, Oct 07, 2022 at 04:23:33PM +0200, Vincent Guittot wrote:
> > Hi Cristian,
> >
>
> Hi Vincent
>
> thanks for give it a try !
>
> > On Sat, 3 Sept 2022 at 20:31, Cristian Marussi <cristian.marussi@....com> wrote:
> > >
> > > Hi all,
> > >
> > > This series aims to introduce a new SCMI unified userspace interface meant
> > > to ease testing an SCMI Server implementation for compliance, fuzzing etc.,
> > > from the perspective of the OSPM agent (non-secure world only ...)
> > >
> > > It is proposed as a testing/development facility, it is NOT meant to be a
> > > feature to use in production, but only enabled in Kconfig for test
> > > deployments.
> > >
> > > Currently an SCMI Compliance Suite like the one at [1] can only work by
> > > injecting SCMI messages at the SCMI transport layer using the mailbox test
> > > driver (CONFIG_MAILBOX_TEST) via its few debugfs entries and looking at
> > > the related replies from the SCMI backend Server.
> > >
> >
> > ...
> >
> > >
> > > In V2 the runtime enable/disable switching capability has been removed
> > > (for now) since still not deemed to be stable/reliable enough: as a
> > > consequence when SCMI Raw support is compiled in, the regular SCMI stack
> > > drivers are now inhibited permanently for that Kernel.
> > >
> > > A quick and trivial example from the shell...reading from a sensor
> > > injecting a properly crafted packet in raw mode:
> > >
> > >         # INJECT THE SENSOR_READING MESSAGE FOR SENSOR ID=1 (binary little endian)
> > >         root@...-buster-arm64:~# echo -e -n \\x06\\x54\\x00\\x00\\x01\\x00\\x00\\x00\\x00\\x00\\x00\\x00 > /sys/kernel/debug/scmi_raw/message
> >
> > I have tried your patchset with an SCMI server using an optee-os
> > transport channel but I have a timed out error when trying your
> > example above to read sensor1
> >
> > #  echo -e -n \\x06\\x54\\x00\\x00\\x01\\x00\\x00\\x00\\x00\\x00\\x00\\x00
> > > /sys/kernel/debug/scmi_raw/message
> > # [   93.306690] arm-scmi firmware:scmi0: timed out in RAW response -
> > HDR:00005406
> >
> > and there no response available when trying to read it with
> > # cat /sys/kernel/debug/scmi_raw/message
> >
>
> is there anything cat'ting /sys/kernel/debug/scmi_raw/errors ?

It was empty

>
> >
> > The sensor 1 can be successfully read in normal mode:
> > # cat /sys/class/hwmon/hwmon0/temp1_input
> > 25000
> > #
> >
> > In both case, the SCMI server received the requests and replied successfully
> >
> > Could it be that you process the answer differently in case of raw mode ?
> >
>
> Well, absolutely, when in raw mode the reply is picked up directly into
> the RX path and copied in a message queue to be read from asyncrhnously
> later via debugfs.
>
> ... mmm I think I found the problem...the reply is picked up on the RX *IRQ*
> path as of now...but in optee/SMC there is no interrupt (sometime there is in
> SMC) and the reply is instead read back straight away (transport is marked as
> sync_cmds_completed_on_ret=true in fact).... so this is the issue probably ...
> I have NOT tested on anything but mailbox and virtio till now...and I
> missed this possibility that this NO-irq reply was a thing even when NOT
> in polling mode (which I do not support)...my bad :<
>
> Ok, next week I'll rework the series to fix this big_BUG and some other minor
> things...in the meantime if you want to try this snippet down below...
>
> ... this won't definitely be the final patch but it could let you experiment
> more (only build tested though )

Thanks.
The patch below fixes my problem with optee transport layer

>
> Thanks for testing !
> Cristian
>
> --->8-------
>
> diff --git a/drivers/firmware/arm_scmi/raw_mode.c b/drivers/firmware/arm_scmi/raw_mode.c
> index 13eeebe4b7a8..b9fcb66a1b6a 100644
> --- a/drivers/firmware/arm_scmi/raw_mode.c
> +++ b/drivers/firmware/arm_scmi/raw_mode.c
> @@ -197,6 +197,8 @@ struct scmi_dbg_raw_data {
>         size_t rx_size;
>  };
>
> +void scmi_raw_message_report(void *r, struct scmi_xfer *xfer, unsigned int idx);
> +
>  static inline
>  struct scmi_raw_buffer *scmi_raw_buffer_get(struct scmi_raw_mode_info *raw,
>                                             unsigned int idx)
> @@ -389,22 +391,34 @@ static void scmi_xfer_raw_worker(struct work_struct *work)
>
>                 xfer = rw->xfer;
>
> -               /*
> -                * Waiters are queued by wait-deadline at the end, so some of
> -                * them could have been already expired when processed, BUT we
> -                * have to check the completion status anyway just in case a
> -                * virtually expired (aged) transaction was indeed completed
> -                * fine and we'll have to wait for the asynchronous part (if
> -                * any).
> -                */
> -               aging = jiffies - rw->start_jiffies;
> -               tmo = max_tmo > aging ? max_tmo - aging : 0;
> -
> -               if ((tmo && !wait_for_completion_timeout(&xfer->done, tmo)) ||
> -                   (!tmo && !try_wait_for_completion(&xfer->done))) {
> -                       dev_err(dev, "timed out in RAW response - HDR:%08X\n",
> -                               pack_scmi_header(&xfer->hdr));
> -                       ret = -ETIMEDOUT;
> +               if (!raw->desc->sync_cmds_completed_on_ret) {
> +                       /*
> +                        * Waiters are queued by wait-deadline at the end, so some of
> +                        * them could have been already expired when processed, BUT we
> +                        * have to check the completion status anyway just in case a
> +                        * virtually expired (aged) transaction was indeed completed
> +                        * fine and we'll have to wait for the asynchronous part (if
> +                        * any).
> +                        */
> +                       aging = jiffies - rw->start_jiffies;
> +                       tmo = max_tmo > aging ? max_tmo - aging : 0;
> +
> +                       if ((tmo &&
> +                            !wait_for_completion_timeout(&xfer->done, tmo)) ||
> +                            (!tmo && !try_wait_for_completion(&xfer->done))) {
> +                               dev_err(dev,
> +                                       "timed out in RAW response - HDR:%08X\n",
> +                                       pack_scmi_header(&xfer->hdr));
> +                               ret = -ETIMEDOUT;
> +                       }
> +               } else {
> +                       raw->desc->ops->fetch_response(rw->cinfo, xfer);
> +                       /* Trace polled replies. */
> +                       trace_scmi_msg_dump(xfer->hdr.protocol_id, xfer->hdr.id,
> +                                           "RESP",
> +                                           xfer->hdr.seq, xfer->hdr.status,
> +                                           xfer->rx.buf, xfer->rx.len);
> +                       scmi_raw_message_report(raw, xfer, SCMI_RAW_REPLY_QUEUE);
>                 }
>
>                 /* Avoid unneeded async waits */
>
>
> ---8<-------
>

Powered by blists - more mailing lists