linux-kernel - Re: [PATCH] firmware: arm_scmi: Make scmi core independent of transport type

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20191203120002.GB4171@bogus>
Date:   Tue, 3 Dec 2019 12:00:02 +0000
From:   Sudeep Holla <sudeep.holla@....com>
To:     Viresh Kumar <viresh.kumar@...aro.org>
Cc:     Vincent Guittot <vincent.guittot@...aro.org>,
        Arnd Bergmann <arnd@...db.de>,
        Jassi Brar <jassisinghbrar@...il.com>,
        linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
        Sudeep Holla <sudeep.holla@....com>
Subject: Re: [PATCH] firmware: arm_scmi: Make scmi core independent of
 transport type

On Fri, Nov 29, 2019 at 03:01:39PM +0530, Viresh Kumar wrote:
> The SCMI specification is fairly independent of the transport protocol,
> which can be a simple mailbox (already implemented) or anything else.
> The current Linux implementation however is very much dependent of the
> mailbox transport layer.
>
> This patch makes the SCMI core code (driver.c) independent of the
> mailbox transport layer and moves all mailbox related code to a new
> file: mailbox.c.
>

The implementation looks fine to me.

> We can now implement more transport protocols to transport SCMI
> messages.
>

I am more interested in this part. As I am aware the only 2 other
transport being discussed is SMC/HVC and new/yet conceptual SPCI(built
on top of SMC/HVC). There are already discussions on the list to make
former as mailbox[1]. While I see both pros and cons with that approach,
there's a need to converge. One main advantage I see with SMC/HVC mailbox
is that it can be used with any other client and not just SCMI. Equally,
the queuing in the mailbox may not be needed with fast SMC/HVC but may
be needed for new SPCI(not yet fully analysed).

> The transport protocols just need to provide struct scmi_transport_ops,
> with its version of the callbacks to enable exchange of SCMI messages.
>

As I mentioned I am fine with implementation in this patch. But I would
like to hear especially from Arnd and Jassi as the abstraction look more
like mailbox APIs themselves and may look like duplication. I don't
want people to realise late that this is not good idea for whatever
reasons. If we have valid and enough reasons to do so, we can take
this approach. I really need some feedback here.

--
Regards,
Sudeep

[1] https://lore.kernel.org/lkml/1575281525-1549-1-git-send-email-peng.fan@nxp.com