Re: [PATCH 1/2] hwspinlock/omap: add support for dt nodes

From: Kumar Gala
Date: Wed Sep 04 2013 - 13:51:16 EST



On Sep 4, 2013, at 12:03 PM, Suman Anna wrote:

>>>>>>>>
>>>>>>>>> HwSpinlock IP is present only on OMAP4 and other newer SoCs,
>>>>>>>>> which are all device-tree boot only. This patch adds the
>>>>>>>>> base support for parsing the DT nodes, and removes the code
>>>>>>>>> dealing with the traditional platform device instantiation.
>>>>>>>>>
>>>>>>>>> Signed-off-by: Suman Anna <s-anna@xxxxxx>
>>>>>>>>> ---
>>>>>>>>> .../devicetree/bindings/hwlock/omap-hwspinlock.txt | 28 ++++++++++
>>>>>>>>> arch/arm/mach-omap2/Makefile | 3 --
>>>>>>>>> arch/arm/mach-omap2/hwspinlock.c | 60 ----------------------
>>>>>>>>> drivers/hwspinlock/omap_hwspinlock.c | 21 ++++++--
>>>>>>>>> 4 files changed, 45 insertions(+), 67 deletions(-)
>>>>>>>>> create mode 100644 Documentation/devicetree/bindings/hwlock/omap-hwspinlock.txt
>>>>>>>>> delete mode 100644 arch/arm/mach-omap2/hwspinlock.c
>>>>>>>>>
>>>>>>>>> diff --git a/Documentation/devicetree/bindings/hwlock/omap-hwspinlock.txt b/Documentation/devicetree/bindings/hwlock/omap-hwspinlock.txt
>>>>>>>>> new file mode 100644
>>>>>>>>> index 0000000..adfb8ad
>>>>>>>>> --- /dev/null
>>>>>>>>> +++ b/Documentation/devicetree/bindings/hwlock/omap-hwspinlock.txt
>>>>>>>>> @@ -0,0 +1,28 @@
>>>>>>>>> +OMAP4+ HwSpinlock Driver
>>>>>>>>> +
>>>>>>>>> +Required properties:
>>>>>>>>> +- compatible: Currently supports only "ti,omap4-hwspinlock" for
>>>>>>>>> + OMAP44xx, OMAP54xx, AM33xx, AM43xx, DRA7xx SoCs
>>>>>>>>> +- reg: Contains the hwspinlock register address range (base
>>>>>>>>> + address and length)
>>>>>>>>> +- ti,hwmods: Name of the hwmod associated with the hwspinlock device
>>>>>>>>> +
>>>>>>>>> +Optional properties:
>>>>>>>>> +- base_id: Base Id for the locks for a particular hwspinlock
>>>>>>>>> + device. If not mentioned, a default value of 0 is used.
>>>>>>>>> + This property is mandatory ONLY if a SoC has several
>>>>>>>>> + hwspinlock devices. There are currently no such OMAP
>>>>>>>>> + SoCs.
>>>>>>>>
>>>>>>>> Should this be ti,base_id ? [ I know its kinda generic in its intent for any SoC w/multiple blocks ]
>>>>>>>
>>>>>>> I didn't add the "ti," prefix exactly for the same reason - it is
>>>>>>> generic w.r.t the hwspinlock core irrespective of the SoC family, and
>>>>>>> there is nothing ti or OMAP specific about it. I have added it to keep
>>>>>>> the DT node definition in sync with the driver code. If it is too
>>>>>>> generic a name, it can always be renamed as hwlock_base_id. This will be
>>>>>>> SoC agnostic property for the hwspinlock driver. What do you think?
>>>>>>
>>>>>> I'm wondering if we should use cell-index for this purpose.
>>>>>
>>>>> I didn't get you completely. Do you intend to compute the base_id using
>>>>> cell-index and number of locks (which may be a separate field altogether
>>>>> if this information cannot be read from the h/w)? My understanding is
>>>>> that cell-index is primarily used for identifying the h/w instance number.
>>>>
>>>> I was suggesting using cell-index instead of base_id. What we should probably due is have a devicetree/bindings/hwlock/hwlock.txt that would describe generic properties like this and just reference that in the omap binding spec.
>>>
>>> Common hwlock.txt sounds good. Will make the change.
>>>
>>>>
>>>> I'm thinking if we dont use cell-index, that it should probably be hwlock-base-id
>>>>
>>>
>>> I prefer to use hwlock-base-id. I think we should also be defining a
>>> common property name for number of locks, say hwlock-num-locks.
>>
>> I'm good with that, cell-index is always funny so might as well be explicit.
>>
>> I'm also good with hwlock-num-locks, I'll update the msm spinlock driver to use this.
>>
>> Can you also maybe add some helper functions into the hwspinlock core to return these values so we both don't duplicate code in drivers and maintain consistency.
>
> I am trying to understand what you would need these for. Your driver
> would already know the base_id and num_locks, since these are used in
> the registration function.


It would be something simple like:

static inline int of_get_hwlock_base_id (struct device_node *dn) {
#ifdef CONFIG_OF
u32 val;

if (of_property_read_u32(dn, "hwlock-base-id", &val)
return val;
#endif
return 0;
}

make sense?

- k

--
Employee of Qualcomm Innovation Center, Inc.
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation

--
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/