Re: [PATCHv5 2/4] mailbox: Introduce framework for mailbox

From: Sudeep Holla
Date: Tue Jun 03 2014 - 11:06:12 EST


Hi Jassi,

On 03/06/14 11:21, Jassi Brar wrote:
On 3 June 2014 15:05, Sudeep Holla <sudeep.holla@xxxxxxx> wrote:
Hi Jassi,

On Mon, Jun 2, 2014 at 6:11 PM, Jassi Brar <jassisinghbrar@xxxxxxxxx> wrote:
On Mon, Jun 2, 2014 at 8:44 PM, Matt Porter <mporter@xxxxxxxxxx> wrote:
On Fri, May 30, 2014 at 11:01:55AM +0530, Jassi Brar wrote:

Being more specific to your platform, I think you need some server
code (mailbox's client) that every driver (like clock, pmu, pinmux
etc) registers with to send messages to remote and receive commands
from remote ... perhaps by registering some filter to sort out
messages for each driver.

Right, and here's where you hit on the problem. This server you mention
is not a piece of hardware, it would be a software construct. As such, it
doesn't fit into the DT binding as it exists. It's probably best to
illustrate in DT syntax.

If I were to represent the hardware relationship in the DT binding now
it would look like this:

---
cpm: mailbox@deadbeef {
compatible = "brcm,bcm-cpm-mailbox";
reg = <...>;
#mbox-cells <1>;
interrupts = <...>;
};

/* clock complex */
ccu {
compatible = "brcm,bcm-foo-ccu";
mbox = <&cpm CPM_SYSTEM_CHANNEL>;
mbox-names = "system";
/* leaving out other mailboxes for brevity */
#clock-cells <1>;
clock-output-names = "bar",
"baz";
};

pmu {
compatible = "brcm,bcm-foo-pmu"
mbox = <&cpm CPM_SYSTEM_CHANNEL>;
mbox-names = "system";
};

pinmux {
compatible = "brcm,bcm-foo-pinctrl";
mbox = <&cpm CPM_SYSTEM_CHANNEL>;
mbox-names = "system";
};
---
Yeah, I too don't think its a good idea.


What we would need to do is completely ignore this information in each
of the of the client drivers associated with the clock, pmu, and pinmux
devices. This IPC server would need to be instantiated and get the
mailbox information from some source. mbox_request_channel() only works
when the client has an of_node with the mbox-names property present.
Let's say we follow this model and represent it in DT:

--
cpm: mailbox@deadbeef {
compatible = "brcm,bcm-cpm-mailbox";
reg = <...>;
#mbox-cells <1>;
interrupts = <...>;
};

cpm_ipc {
compatible = "brcm,bcm-cpm-ipc";
mbox = <&cpm CPM_SYSTEM_CHANNEL>;
mbox-names = "system";
/* leaving out other mailboxes for brevity */
};

I am assuming this is what you referring below: cpm(which is the controller)
and cpm_ipc(a virtual node referring to the client). If yes that's fine, but is
this virtual node acceptable ?

---

This would allow an ipc driver to exclusively own this system channel,
but now we've invented a binding that doesn't reflect the hardware at
all. It's describing software so I don't believe the DT maintainers will
allow this type of construct.

Must the server node specify MMIO and an IRQ, to be acceptable? Like ...

cpm_ipc : cpm@deadbeef {
compatible = "brcm,bcm-cpm-ipc";
/* reg = <0xdeadbeef 0x100>; */
/* interrupts = <0 123 4>; */
mbox = <&cpm CPM_SYSTEM_CHANNEL>;
mbox-names = "system";
};


Sorry got carried away with this example, assumed it to be accidental comment
and the proposal was to unify both controller and client. Now its clear, ignore
my noise earlier.

cpm_ipc already specifies a hardware resource (mbox) that its driver
needs, I think that should be enough reason. If it were some purely
soft property for the driver like
mode = "poll"; //or "irq"
then the node wouldn't be justified because that is the job of a
build-time config or run-time module option.


Like Matt, I am also in similar situation where there's a lot of common
code necessary to construct/parse IPCs for each of the drivers using the
mailbox.

As per your suggestion if we have single DT node to specify both the
controller and the client, we might still have to pollute this node with
software specific compatibles.

I am afraid you misunderstood me. I don't suggest single node for
mailbox controller and client, and IIUC, neither did Matt. Please note
the controller is cpm and client is cpm_ipc.

BTW, here we at least have a hardware resource to specify in the DT
node, there are examples in kernel where the DT nodes are purely
virtual. For ex, grep for "linux,spdif-dit". So I think we should be
ok.


Yes while I agree with you, but will it be acceptable for a generic mailbox
binding ? I think this is what even Matt mentioned earlier.

One possible scenario I can think of is that if the mailbox controller is
a standard primecell like PL320 used in multiple SoCs, each SoC vendor
will develop their own protocol implemented in their firmware. This is true
even with single SoC vendor having same IP but changing the protocol to
talk to their firmware.

Yeah, people have noted that in previous threads.


Ok that's good. I tried to follow previous discussions, but couldn't. I followed
only your last postings(v4) after I started using this driver.

We will need a way to identify that protocol mechanism.
Does it make sense to add that ? Is that something acceptable ?

IMO we can't help it more than _trying_ to write the controller
driver as versatile as possible. And still some protocol
version/peculiarity could make reuse of the controller driver worse
than write a new for the protocol version. Any minor change in
behavior could be flagged to controller and client in platform
specific way.

Yes I understand that. So I was worried about unifying both controller and
protocol in one node. The controller driver must just handle raw data and never
interpret it. Now we just need DT maintainers' view on this virtual DT node.

Regards,
Sudeep

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/