[PATCH v3 0/7] OMAP System Control Module

[PATCH v3 0/7] OMAP System Control Module

[Konstantin Baydarov]

 Hello.

This is a next version of series of patches(based on Eduardo Valentin's patch set) adding a basic support for system control module, on OMAP4+ context. It is a working in progress.

Main changes since previous patch set version:
- Bandgap and usb phy: drivers are now independent from control module driver, they use their own functions to acess scm registers.
- omap-control-core: resources aren't hardcoded, they are specified in dts file.
- omap-control-core: Control module is a built-in driver - added control module select to ARCH_HAS_CONTROL_MODULE and ARCH_OMAP4.
Probably, no configuration option is required!
- omap-control-core: Added early init call that ioremaps control module IOMEM window, this allows access of SCM registers very early, for example from omap_type()
- omap-control-core: Removed device pointer from omap-control-core API arguments, becuase there can be only one instance control
module device.
- omap-control-core: removed omap_control_get, omap_control_readl, omap_control_writel
- omap-control-core: added omap_control_status_read that is used early in omap_type
- Bandgap and usb phy: Added private spinlocks for bandgap and usb drivers.
- Bandgap: Check the type of bandgap dynamically in bandgap driver probe function by reading
omap core control module revision register CONTROL_GEN_CORE_REVISION.
- Bandgap and usb phy: Parent SCM platform device IOMEM resources is used to get the base address of SCM window.
- Bandgap masks defines were moved to drivers/thermal/omap-bandgap.c.

TODO list for bandgap driver:
- Reserve omap-control-core IOMEM window.
- Improve thermal zone definition for OMAP4
- Introduce the thermal zones for OMAP5

Re: [PATCH v3 0/7] OMAP System Control Module

[Eduardo Valentin]

Hello,

On Wed, Jun 27, 2012 at 10:04:32PM +0400, Konstantin Baydarov wrote:
>  Hello.
> 
> This is a next version of series of patches(based on Eduardo Valentin's patch set) adding a basic support for system control module, on OMAP4+ context. It is a working in progress.
> 
> Main changes since previous patch set version:
> - Bandgap and usb phy: drivers are now independent from control module driver, they use their own functions to acess scm registers.

Well, I believe the idea of having them with their own io resource was to avoid
children drivers accessing each other io areas. Is this now working?

> - omap-control-core: resources aren't hardcoded, they are specified in dts file.
> - omap-control-core: Control module is a built-in driver - added control module select to ARCH_HAS_CONTROL_MODULE and ARCH_OMAP4.
> Probably, no configuration option is required!

Why is this? I suppose the idea is to have the arch config selecting that flag.

> - omap-control-core: Added early init call that ioremaps control module IOMEM window, this allows access of SCM registers very early, for example from omap_type()

Is this going to be reserved as well? if yes, how children are going to have
their own access functions?

> - omap-control-core: Removed device pointer from omap-control-core API arguments, becuase there can be only one instance control
> module device.
> - omap-control-core: removed omap_control_get, omap_control_readl, omap_control_writel

fine, assuming the io split works...

> - omap-control-core: added omap_control_status_read that is used early in omap_type
> - Bandgap and usb phy: Added private spinlocks for bandgap and usb drivers.
> - Bandgap: Check the type of bandgap dynamically in bandgap driver probe function by reading
> omap core control module revision register CONTROL_GEN_CORE_REVISION.

this has some issue... I will post comment on the patch

> - Bandgap and usb phy: Parent SCM platform device IOMEM resources is used to get the base address of SCM window.

Ohh.. Then what is the point of having them using their own io access functions,
nothing is again protected as you have only 1 io resource which is shared.
And now is even worse because you have several io access function..
I think this is moving backwards.

> - Bandgap masks defines were moved to drivers/thermal/omap-bandgap.c.
> 
> TODO list for bandgap driver:
> - Reserve omap-control-core IOMEM window.
> - Improve thermal zone definition for OMAP4
> - Introduce the thermal zones for OMAP5

Based on the review comments on RFC patch series, there are more
things to be done.

I have the O4 and O5 zone definition ready. But that work depends
on generic CPU cooling patches. I have also looked the driver support
for 4430. But the probing and io resource split is still based on
previous design. I have also a patch which does the remapping in the
bandgap driver. But really, for this to work it would require to have
several entries in the reg property. And that is going to change from
BG version to BG version.

I can prepare some patches for this.

Re: [PATCH v3 0/7] OMAP System Control Module

[Valentin, Eduardo]

Hello Konstantin,

On Thu, Jun 28, 2012 at 7:43 AM, Eduardo Valentin
<eduardo.valentin@ti.com> wrote:
> Hello,
>
> On Wed, Jun 27, 2012 at 10:04:32PM +0400, Konstantin Baydarov wrote:
>>  Hello.
>>
>> This is a next version of series of patches(based on Eduardo Valentin's patch set) adding a basic support for system control module, on OMAP4+ context. It is a working in progress.
>>
>> Main changes since previous patch set version:
>> - Bandgap and usb phy: drivers are now independent from control module driver, they use their own functions to acess scm registers.
>
> Well, I believe the idea of having them with their own io resource was to avoid
> children drivers accessing each other io areas. Is this now working?
>
>> - omap-control-core: resources aren't hardcoded, they are specified in dts file.
>> - omap-control-core: Control module is a built-in driver - added control module select to ARCH_HAS_CONTROL_MODULE and ARCH_OMAP4.
>> Probably, no configuration option is required!
>
> Why is this? I suppose the idea is to have the arch config selecting that flag.
>
>> - omap-control-core: Added early init call that ioremaps control module IOMEM window, this allows access of SCM registers very early, for example from omap_type()
>
> Is this going to be reserved as well? if yes, how children are going to have
> their own access functions?
>
>> - omap-control-core: Removed device pointer from omap-control-core API arguments, becuase there can be only one instance control
>> module device.
>> - omap-control-core: removed omap_control_get, omap_control_readl, omap_control_writel
>
> fine, assuming the io split works...
>
>> - omap-control-core: added omap_control_status_read that is used early in omap_type
>> - Bandgap and usb phy: Added private spinlocks for bandgap and usb drivers.
>> - Bandgap: Check the type of bandgap dynamically in bandgap driver probe function by reading
>> omap core control module revision register CONTROL_GEN_CORE_REVISION.
>
> this has some issue... I will post comment on the patch
>
>> - Bandgap and usb phy: Parent SCM platform device IOMEM resources is used to get the base address of SCM window.
>
> Ohh.. Then what is the point of having them using their own io access functions,
> nothing is again protected as you have only 1 io resource which is shared.
> And now is even worse because you have several io access function..
> I think this is moving backwards.
>
>> - Bandgap masks defines were moved to drivers/thermal/omap-bandgap.c.
>>
>> TODO list for bandgap driver:
>> - Reserve omap-control-core IOMEM window.
>> - Improve thermal zone definition for OMAP4
>> - Introduce the thermal zones for OMAP5
>
> Based on the review comments on RFC patch series, there are more
> things to be done.
>
> I have the O4 and O5 zone definition ready. But that work depends
> on generic CPU cooling patches. I have also looked the driver support
> for 4430. But the probing and io resource split is still based on
> previous design. I have also a patch which does the remapping in the
> bandgap driver. But really, for this to work it would require to have
> several entries in the reg property. And that is going to change from
> BG version to BG version.
>
> I can prepare some patches for this.

You can find the updated patches here:
git@gitorious.org:thermal-framework/thermal-framework.git
thermal_work/omap/bandgap_clean

But, as I said they are based on the generic cpu cooling. And the zone
registration follows the API as in linux-next.


-- 

Eduardo Valentin
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v3 0/7] OMAP System Control Module

[Valentin, Eduardo]

Hello again,

On Thu, Jun 28, 2012 at 9:29 AM, Valentin, Eduardo
<eduardo.valentin@ti.com> wrote:
> Hello Konstantin,
>
> On Thu, Jun 28, 2012 at 7:43 AM, Eduardo Valentin
> <eduardo.valentin@ti.com> wrote:
>> Hello,
>>
>> On Wed, Jun 27, 2012 at 10:04:32PM +0400, Konstantin Baydarov wrote:
>>>  Hello.
>>>
>>> This is a next version of series of patches(based on Eduardo Valentin's patch set) adding a basic support for system control module, on OMAP4+ context. It is a working in progress.
>>>
>>> Main changes since previous patch set version:
>>> - Bandgap and usb phy: drivers are now independent from control module driver, they use their own functions to acess scm registers.
>>
>> Well, I believe the idea of having them with their own io resource was to avoid
>> children drivers accessing each other io areas. Is this now working?
>>
>>> - omap-control-core: resources aren't hardcoded, they are specified in dts file.
>>> - omap-control-core: Control module is a built-in driver - added control module select to ARCH_HAS_CONTROL_MODULE and ARCH_OMAP4.
>>> Probably, no configuration option is required!
>>
>> Why is this? I suppose the idea is to have the arch config selecting that flag.
>>
>>> - omap-control-core: Added early init call that ioremaps control module IOMEM window, this allows access of SCM registers very early, for example from omap_type()
>>
>> Is this going to be reserved as well? if yes, how children are going to have
>> their own access functions?
>>
>>> - omap-control-core: Removed device pointer from omap-control-core API arguments, becuase there can be only one instance control
>>> module device.
>>> - omap-control-core: removed omap_control_get, omap_control_readl, omap_control_writel
>>
>> fine, assuming the io split works...
>>
>>> - omap-control-core: added omap_control_status_read that is used early in omap_type
>>> - Bandgap and usb phy: Added private spinlocks for bandgap and usb drivers.
>>> - Bandgap: Check the type of bandgap dynamically in bandgap driver probe function by reading
>>> omap core control module revision register CONTROL_GEN_CORE_REVISION.
>>
>> this has some issue... I will post comment on the patch
>>
>>> - Bandgap and usb phy: Parent SCM platform device IOMEM resources is used to get the base address of SCM window.
>>
>> Ohh.. Then what is the point of having them using their own io access functions,
>> nothing is again protected as you have only 1 io resource which is shared.
>> And now is even worse because you have several io access function..
>> I think this is moving backwards.
>>
>>> - Bandgap masks defines were moved to drivers/thermal/omap-bandgap.c.
>>>
>>> TODO list for bandgap driver:
>>> - Reserve omap-control-core IOMEM window.
>>> - Improve thermal zone definition for OMAP4
>>> - Introduce the thermal zones for OMAP5
>>
>> Based on the review comments on RFC patch series, there are more
>> things to be done.
>>
>> I have the O4 and O5 zone definition ready. But that work depends
>> on generic CPU cooling patches. I have also looked the driver support
>> for 4430. But the probing and io resource split is still based on
>> previous design. I have also a patch which does the remapping in the
>> bandgap driver. But really, for this to work it would require to have
>> several entries in the reg property. And that is going to change from
>> BG version to BG version.
>>
>> I can prepare some patches for this.
>
> You can find the updated patches here:
> git@gitorious.org:thermal-framework/thermal-framework.git
> thermal_work/omap/bandgap_clean
>
> But, as I said they are based on the generic cpu cooling. And the zone
> registration follows the API as in linux-next.

FYI, I just updated the branches on that repo. The branch
thermal_work/omap/pu has all required changes in order to get BG
driver working on 4430/4460/4470 and O5.

It relies on Rui's patch set sent to linux-pm, which reorgs the
thermal fw a bit,  combined with the generic cpu cooling device by
Amit. I also pushed in the 'pu' branch the APIs changes mentioned by
Durga on other thread.

While checking the pu branch you should be able to get a thermal zone
with cpu cooling for OMAP on the mentioned omap versions.

Once there is an agreement about the APIs between SCM core driver and
its children, I can send out the BG driver patches and this initial
omap thermal support using the generic thermal layer.

There are some changes in the core thermal layer in the
thermal_work/eduardo_thermal_upgrade. But there could be probably some
extra changes in the core layer to finalize the omap thermal support.
That's a work in progress :-)

>
>
> --
>
> Eduardo Valentin



-- 

Eduardo Valentin
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html