[PATCH 0/2] memory: brcmstb_memc: Support for BCM74165

[PATCH 0/2] memory: brcmstb_memc: Support for BCM74165

[Florian Fainelli]

This patch series documents and matches the compatible string used on
BCM74165, thanks!

Florian Fainelli (2):
  dt-bindings: memory-controller: Document rev c.1.5 compatible
  memory: brcmstb_memc: Match rev c.1.5 compatible string

 .../bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml    | 1 +
 drivers/memory/brcmstb_memc.c                                 | 4 ++++
 2 files changed, 5 insertions(+)

-- 
2.43.0

[PATCH 1/2] dt-bindings: memory-controller: Document rev c.1.5 compatible

[Florian Fainelli]

Document the revision c.1.5 compatible string that is present on newer
Broadcom STB memory controllers (74165 and onwards).

Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
---
 .../bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml       | 1 +
 1 file changed, 1 insertion(+)

diff --git a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
index 4b072c879b02..99d79ccd1036 100644
--- a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
+++ b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
@@ -29,6 +29,7 @@ properties:
           - brcm,brcmstb-memc-ddr-rev-c.1.2
           - brcm,brcmstb-memc-ddr-rev-c.1.3
           - brcm,brcmstb-memc-ddr-rev-c.1.4
+          - brcm,brcmstb-memc-ddr-rev-c.1.5
       - const: brcm,brcmstb-memc-ddr
 
   reg:
-- 
2.43.0

[PATCH 2/2] memory: brcmstb_memc: Match rev c.1.5 compatible string

[Florian Fainelli]

Match the revision c.1.5 compatible string and apply the v2.1x register
layout such that we can properly control the Self Refresh Power Down
timeouts with those controller version(s).

Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
---
 drivers/memory/brcmstb_memc.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/memory/brcmstb_memc.c b/drivers/memory/brcmstb_memc.c
index 4f17a93aa028..a2750142cef7 100644
--- a/drivers/memory/brcmstb_memc.c
+++ b/drivers/memory/brcmstb_memc.c
@@ -236,6 +236,10 @@ static const struct of_device_id brcmstb_memc_of_match[] = {
 		.compatible = "brcm,brcmstb-memc-ddr-rev-c.1.4",
 		.data = &brcmstb_memc_versions[BRCMSTB_MEMC_V21]
 	},
+	{
+		.compatible = "brcm,brcmstb-memc-ddr-rev-c.1.5",
+		.data = &brcmstb_memc_versions[BRCMSTB_MEMC_V21]
+	},
 	/* default to the original offset */
 	{
 		.compatible = "brcm,brcmstb-memc-ddr",
-- 
2.43.0

Re: [PATCH 1/2] dt-bindings: memory-controller: Document rev c.1.5 compatible

[Krzysztof Kozlowski]

On Tue, Dec 17, 2024 at 11:44:38AM -0800, Florian Fainelli wrote:
> Document the revision c.1.5 compatible string that is present on newer
> Broadcom STB memory controllers (74165 and onwards).
> 
> Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
> ---
>  .../bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml       | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
> index 4b072c879b02..99d79ccd1036 100644
> --- a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
> +++ b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
> @@ -29,6 +29,7 @@ properties:
>            - brcm,brcmstb-memc-ddr-rev-c.1.2
>            - brcm,brcmstb-memc-ddr-rev-c.1.3
>            - brcm,brcmstb-memc-ddr-rev-c.1.4
> +          - brcm,brcmstb-memc-ddr-rev-c.1.5

You should use v2.1 fallback and drop driver patch. Or explain in
commit briefly why different approach is suitable.

Best regards,
Krzysztof

Re: [PATCH 1/2] dt-bindings: memory-controller: Document rev c.1.5 compatible

[Florian Fainelli]

On 12/18/24 03:37, Krzysztof Kozlowski wrote:
> On Tue, Dec 17, 2024 at 11:44:38AM -0800, Florian Fainelli wrote:
>> Document the revision c.1.5 compatible string that is present on newer
>> Broadcom STB memory controllers (74165 and onwards).
>>
>> Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
>> ---
>>   .../bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml       | 1 +
>>   1 file changed, 1 insertion(+)
>>
>> diff --git a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>> index 4b072c879b02..99d79ccd1036 100644
>> --- a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>> +++ b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>> @@ -29,6 +29,7 @@ properties:
>>             - brcm,brcmstb-memc-ddr-rev-c.1.2
>>             - brcm,brcmstb-memc-ddr-rev-c.1.3
>>             - brcm,brcmstb-memc-ddr-rev-c.1.4
>> +          - brcm,brcmstb-memc-ddr-rev-c.1.5
> 
> You should use v2.1 fallback and drop driver patch. Or explain in
> commit briefly why different approach is suitable.

Are you suggesting that we should have fallback compatible strings, such 
that we have something like this:

compatible = "brcm,brcmstb-memc-ddr-rev-c.1.5", 
"brcm,brcmstb-memc-ddr-rev-c", "brcm,brcmstb-memc-ddr"

and the driver only needs to match on "brcm,brcmstb-memc-ddr-rev-c" and 
apply the adequate register offset table?

If so, that is not how the current binding, and therefore DTBs are being 
deployed, so that will introduce a breakage until we update all DTBs in 
the field...
-- 
Florian

Re: [PATCH 1/2] dt-bindings: memory-controller: Document rev c.1.5 compatible

[Krzysztof Kozlowski]

On Wed, Dec 18, 2024 at 09:15:08AM -0800, Florian Fainelli wrote:
> On 12/18/24 03:37, Krzysztof Kozlowski wrote:
> > On Tue, Dec 17, 2024 at 11:44:38AM -0800, Florian Fainelli wrote:
> > > Document the revision c.1.5 compatible string that is present on newer
> > > Broadcom STB memory controllers (74165 and onwards).
> > > 
> > > Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
> > > ---
> > >   .../bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml       | 1 +
> > >   1 file changed, 1 insertion(+)
> > > 
> > > diff --git a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
> > > index 4b072c879b02..99d79ccd1036 100644
> > > --- a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
> > > +++ b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
> > > @@ -29,6 +29,7 @@ properties:
> > >             - brcm,brcmstb-memc-ddr-rev-c.1.2
> > >             - brcm,brcmstb-memc-ddr-rev-c.1.3
> > >             - brcm,brcmstb-memc-ddr-rev-c.1.4
> > > +          - brcm,brcmstb-memc-ddr-rev-c.1.5
> > 
> > You should use v2.1 fallback and drop driver patch. Or explain in
> > commit briefly why different approach is suitable.
> 
> Are you suggesting that we should have fallback compatible strings, such
> that we have something like this:
> 
> compatible = "brcm,brcmstb-memc-ddr-rev-c.1.5",
> "brcm,brcmstb-memc-ddr-rev-c", "brcm,brcmstb-memc-ddr"
> 
> and the driver only needs to match on "brcm,brcmstb-memc-ddr-rev-c" and
> apply the adequate register offset table?

Almost, fallback should be brcm,brcmstb-memc-ddr-rev-b.2.1 or whatever
was in the driver first or whatever is the oldest known common
interface.

brcm,brcmstb-memc-ddr-rev-c is not a specific compatible.

> If so, that is not how the current binding, and therefore DTBs are being
> deployed, so that will introduce a breakage until we update all DTBs in the
> field...

No. First, I thought about new comaptible so the one you add here. No
breakage, it's new compatible. This saves you these pointless updates of
driver everytime you add new compatible.

Second, you can introduce fallbacks to older compatibles as well - there
will be no breakage, because you add one more compatible. The old
compatibles (covered by fallback) of course stays in the driver, so
there is no breakage at all. We did it multiple times for several
different bindings in Qualcomm. People were doing exactly the same:
adding compatible for new device to binding and driver, without
considering the compatibility at all.

Except being logically correct choice - using fallbacks - this really
has huge benefits when later upstreaming complete, big SoCs, like we do
for latest Qualcomm SoCs: several changes will be only bindings updates.

Best regards,
Krzysztof

Re: [PATCH 1/2] dt-bindings: memory-controller: Document rev c.1.5 compatible

[Florian Fainelli]

On 12/19/24 00:58, Krzysztof Kozlowski wrote:
> On Wed, Dec 18, 2024 at 09:15:08AM -0800, Florian Fainelli wrote:
>> On 12/18/24 03:37, Krzysztof Kozlowski wrote:
>>> On Tue, Dec 17, 2024 at 11:44:38AM -0800, Florian Fainelli wrote:
>>>> Document the revision c.1.5 compatible string that is present on newer
>>>> Broadcom STB memory controllers (74165 and onwards).
>>>>
>>>> Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
>>>> ---
>>>>    .../bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml       | 1 +
>>>>    1 file changed, 1 insertion(+)
>>>>
>>>> diff --git a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>>>> index 4b072c879b02..99d79ccd1036 100644
>>>> --- a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>>>> +++ b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>>>> @@ -29,6 +29,7 @@ properties:
>>>>              - brcm,brcmstb-memc-ddr-rev-c.1.2
>>>>              - brcm,brcmstb-memc-ddr-rev-c.1.3
>>>>              - brcm,brcmstb-memc-ddr-rev-c.1.4
>>>> +          - brcm,brcmstb-memc-ddr-rev-c.1.5
>>>
>>> You should use v2.1 fallback and drop driver patch. Or explain in
>>> commit briefly why different approach is suitable.
>>
>> Are you suggesting that we should have fallback compatible strings, such
>> that we have something like this:
>>
>> compatible = "brcm,brcmstb-memc-ddr-rev-c.1.5",
>> "brcm,brcmstb-memc-ddr-rev-c", "brcm,brcmstb-memc-ddr"
>>
>> and the driver only needs to match on "brcm,brcmstb-memc-ddr-rev-c" and
>> apply the adequate register offset table?
> 
> Almost, fallback should be brcm,brcmstb-memc-ddr-rev-b.2.1 or whatever
> was in the driver first or whatever is the oldest known common
> interface.
> 
> brcm,brcmstb-memc-ddr-rev-c is not a specific compatible.
> 
>> If so, that is not how the current binding, and therefore DTBs are being
>> deployed, so that will introduce a breakage until we update all DTBs in the
>> field...
> 
> No. First, I thought about new comaptible so the one you add here. No
> breakage, it's new compatible. This saves you these pointless updates of
> driver everytime you add new compatible.

Yes, but that is not how the binding has been defined until now, so all 
of the DTBs out there have:

compatible = "brcm,brcmstb-memc-ddr-rev-b.2.x", "brcm,brcmstb-memc-ddr"

(where X is in range [1..5])

and there is no fallback defined to "brcm,brcmstb-memc-ddr-rev-b.2.1", 
so it is not like we can retrofit that easily by adding one now.

 > > Second, you can introduce fallbacks to older compatibles as well - 
there
> will be no breakage, because you add one more compatible. The old
> compatibles (covered by fallback) of course stays in the driver, so
> there is no breakage at all. We did it multiple times for several
> different bindings in Qualcomm. People were doing exactly the same:
> adding compatible for new device to binding and driver, without
> considering the compatibility at all.
> 
> Except being logically correct choice - using fallbacks - this really
> has huge benefits when later upstreaming complete, big SoCs, like we do
> for latest Qualcomm SoCs: several changes will be only bindings updates.

Yes, there are advantages to using fallbacks and we (ab)use that 
whenever practical.

The driver only uses a very limited subset of registers (for now), the 
registers change between minor revisions as well in a way that using a 
fallback like "brcm,brcmstb-memc-ddr-rev-b.2.1" is not accurate enough 
not practical. In particular for some of the changes that I am thinking 
of adding later on, we would need the precise minor version because the 
behavior and/or register interface is subtly different that this matters.

Thanks
-- 
Florian

Re: [PATCH 1/2] dt-bindings: memory-controller: Document rev c.1.5 compatible

[Krzysztof Kozlowski]

On 20/12/2024 00:43, Florian Fainelli wrote:
> On 12/19/24 00:58, Krzysztof Kozlowski wrote:
>> On Wed, Dec 18, 2024 at 09:15:08AM -0800, Florian Fainelli wrote:
>>> On 12/18/24 03:37, Krzysztof Kozlowski wrote:
>>>> On Tue, Dec 17, 2024 at 11:44:38AM -0800, Florian Fainelli wrote:
>>>>> Document the revision c.1.5 compatible string that is present on newer
>>>>> Broadcom STB memory controllers (74165 and onwards).
>>>>>
>>>>> Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
>>>>> ---
>>>>>    .../bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml       | 1 +
>>>>>    1 file changed, 1 insertion(+)
>>>>>
>>>>> diff --git a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>>>>> index 4b072c879b02..99d79ccd1036 100644
>>>>> --- a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>>>>> +++ b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>>>>> @@ -29,6 +29,7 @@ properties:
>>>>>              - brcm,brcmstb-memc-ddr-rev-c.1.2
>>>>>              - brcm,brcmstb-memc-ddr-rev-c.1.3
>>>>>              - brcm,brcmstb-memc-ddr-rev-c.1.4
>>>>> +          - brcm,brcmstb-memc-ddr-rev-c.1.5
>>>>
>>>> You should use v2.1 fallback and drop driver patch. Or explain in
>>>> commit briefly why different approach is suitable.
>>>
>>> Are you suggesting that we should have fallback compatible strings, such
>>> that we have something like this:
>>>
>>> compatible = "brcm,brcmstb-memc-ddr-rev-c.1.5",
>>> "brcm,brcmstb-memc-ddr-rev-c", "brcm,brcmstb-memc-ddr"
>>>
>>> and the driver only needs to match on "brcm,brcmstb-memc-ddr-rev-c" and
>>> apply the adequate register offset table?
>>
>> Almost, fallback should be brcm,brcmstb-memc-ddr-rev-b.2.1 or whatever
>> was in the driver first or whatever is the oldest known common
>> interface.
>>
>> brcm,brcmstb-memc-ddr-rev-c is not a specific compatible.
>>
>>> If so, that is not how the current binding, and therefore DTBs are being
>>> deployed, so that will introduce a breakage until we update all DTBs in the
>>> field...
>>
>> No. First, I thought about new comaptible so the one you add here. No
>> breakage, it's new compatible. This saves you these pointless updates of
>> driver everytime you add new compatible.
> 
> Yes, but that is not how the binding has been defined until now, so all 
> of the DTBs out there have:
> 
> compatible = "brcm,brcmstb-memc-ddr-rev-b.2.x", "brcm,brcmstb-memc-ddr"

I don't understand the problem. We talk about new devices here, it does
not matter what existing/old devices have in binding in that matter.

> 
> (where X is in range [1..5])
> 
> and there is no fallback defined to "brcm,brcmstb-memc-ddr-rev-b.2.1", 
> so it is not like we can retrofit that easily by adding one now.
> 
>  > > Second, you can introduce fallbacks to older compatibles as well - 
> there
>> will be no breakage, because you add one more compatible. The old
>> compatibles (covered by fallback) of course stays in the driver, so
>> there is no breakage at all. We did it multiple times for several
>> different bindings in Qualcomm. People were doing exactly the same:
>> adding compatible for new device to binding and driver, without
>> considering the compatibility at all.
>>
>> Except being logically correct choice - using fallbacks - this really
>> has huge benefits when later upstreaming complete, big SoCs, like we do
>> for latest Qualcomm SoCs: several changes will be only bindings updates.
> 
> Yes, there are advantages to using fallbacks and we (ab)use that 
> whenever practical.
> 
> The driver only uses a very limited subset of registers (for now), the 
> registers change between minor revisions as well in a way that using a 
> fallback like "brcm,brcmstb-memc-ddr-rev-b.2.1" is not accurate enough 
> not practical. In particular for some of the changes that I am thinking 
> of adding later on, we would need the precise minor version because the 
> behavior and/or register interface is subtly different that this matters.

Devices work fine now with the same driver data, so they are compatible.
Just because you have some differences or new features does not
invalidate that this is exactly the point for what compatibility was
created.


Best regards,
Krzysztof

Re: [PATCH 1/2] dt-bindings: memory-controller: Document rev c.1.5 compatible

[Florian Fainelli]

On 12/21/2024 12:04 PM, Krzysztof Kozlowski wrote:
> On 20/12/2024 00:43, Florian Fainelli wrote:
>> On 12/19/24 00:58, Krzysztof Kozlowski wrote:
>>> On Wed, Dec 18, 2024 at 09:15:08AM -0800, Florian Fainelli wrote:
>>>> On 12/18/24 03:37, Krzysztof Kozlowski wrote:
>>>>> On Tue, Dec 17, 2024 at 11:44:38AM -0800, Florian Fainelli wrote:
>>>>>> Document the revision c.1.5 compatible string that is present on newer
>>>>>> Broadcom STB memory controllers (74165 and onwards).
>>>>>>
>>>>>> Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
>>>>>> ---
>>>>>>     .../bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml       | 1 +
>>>>>>     1 file changed, 1 insertion(+)
>>>>>>
>>>>>> diff --git a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>>>>>> index 4b072c879b02..99d79ccd1036 100644
>>>>>> --- a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>>>>>> +++ b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>>>>>> @@ -29,6 +29,7 @@ properties:
>>>>>>               - brcm,brcmstb-memc-ddr-rev-c.1.2
>>>>>>               - brcm,brcmstb-memc-ddr-rev-c.1.3
>>>>>>               - brcm,brcmstb-memc-ddr-rev-c.1.4
>>>>>> +          - brcm,brcmstb-memc-ddr-rev-c.1.5
>>>>>
>>>>> You should use v2.1 fallback and drop driver patch. Or explain in
>>>>> commit briefly why different approach is suitable.
>>>>
>>>> Are you suggesting that we should have fallback compatible strings, such
>>>> that we have something like this:
>>>>
>>>> compatible = "brcm,brcmstb-memc-ddr-rev-c.1.5",
>>>> "brcm,brcmstb-memc-ddr-rev-c", "brcm,brcmstb-memc-ddr"
>>>>
>>>> and the driver only needs to match on "brcm,brcmstb-memc-ddr-rev-c" and
>>>> apply the adequate register offset table?
>>>
>>> Almost, fallback should be brcm,brcmstb-memc-ddr-rev-b.2.1 or whatever
>>> was in the driver first or whatever is the oldest known common
>>> interface.
>>>
>>> brcm,brcmstb-memc-ddr-rev-c is not a specific compatible.
>>>
>>>> If so, that is not how the current binding, and therefore DTBs are being
>>>> deployed, so that will introduce a breakage until we update all DTBs in the
>>>> field...
>>>
>>> No. First, I thought about new comaptible so the one you add here. No
>>> breakage, it's new compatible. This saves you these pointless updates of
>>> driver everytime you add new compatible.
>>
>> Yes, but that is not how the binding has been defined until now, so all
>> of the DTBs out there have:
>>
>> compatible = "brcm,brcmstb-memc-ddr-rev-b.2.x", "brcm,brcmstb-memc-ddr"
> 
> I don't understand the problem. We talk about new devices here, it does
> not matter what existing/old devices have in binding in that matter.

It does matter, because the DTBs that contain this compatible string are 
already in the field, we cannot retrofit them overnight with an 
additional compatible string in order to provide a fallback. Because 
this is submitted now does not mean it is a new device, this was part of 
my backlog to get submitted earlier on.

> 
>>
>> (where X is in range [1..5])
>>
>> and there is no fallback defined to "brcm,brcmstb-memc-ddr-rev-b.2.1",
>> so it is not like we can retrofit that easily by adding one now.
>>
>>   > > Second, you can introduce fallbacks to older compatibles as well -
>> there
>>> will be no breakage, because you add one more compatible. The old
>>> compatibles (covered by fallback) of course stays in the driver, so
>>> there is no breakage at all. We did it multiple times for several
>>> different bindings in Qualcomm. People were doing exactly the same:
>>> adding compatible for new device to binding and driver, without
>>> considering the compatibility at all.
>>>
>>> Except being logically correct choice - using fallbacks - this really
>>> has huge benefits when later upstreaming complete, big SoCs, like we do
>>> for latest Qualcomm SoCs: several changes will be only bindings updates.
>>
>> Yes, there are advantages to using fallbacks and we (ab)use that
>> whenever practical.
>>
>> The driver only uses a very limited subset of registers (for now), the
>> registers change between minor revisions as well in a way that using a
>> fallback like "brcm,brcmstb-memc-ddr-rev-b.2.1" is not accurate enough
>> not practical. In particular for some of the changes that I am thinking
>> of adding later on, we would need the precise minor version because the
>> behavior and/or register interface is subtly different that this matters.
> 
> Devices work fine now with the same driver data, so they are compatible.

For now they do, but not for other features that will be submitted 
later, they will not be, and at that point we will need to know that 
this is a rev C.1.5 controller versus say a rev C.1.4 controller.

> Just because you have some differences or new features does not
> invalidate that this is exactly the point for what compatibility was
> created.
> 
> 
> Best regards,
> Krzysztof

-- 
Florian

Re: [PATCH 1/2] dt-bindings: memory-controller: Document rev c.1.5 compatible

[Krzysztof Kozlowski]

On 26/12/2024 19:50, Florian Fainelli wrote:
> 
> 
> On 12/21/2024 12:04 PM, Krzysztof Kozlowski wrote:
>> On 20/12/2024 00:43, Florian Fainelli wrote:
>>> On 12/19/24 00:58, Krzysztof Kozlowski wrote:
>>>> On Wed, Dec 18, 2024 at 09:15:08AM -0800, Florian Fainelli wrote:
>>>>> On 12/18/24 03:37, Krzysztof Kozlowski wrote:
>>>>>> On Tue, Dec 17, 2024 at 11:44:38AM -0800, Florian Fainelli wrote:
>>>>>>> Document the revision c.1.5 compatible string that is present on newer
>>>>>>> Broadcom STB memory controllers (74165 and onwards).
>>>>>>>
>>>>>>> Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
>>>>>>> ---
>>>>>>>     .../bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml       | 1 +
>>>>>>>     1 file changed, 1 insertion(+)
>>>>>>>
>>>>>>> diff --git a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>>>>>>> index 4b072c879b02..99d79ccd1036 100644
>>>>>>> --- a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>>>>>>> +++ b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>>>>>>> @@ -29,6 +29,7 @@ properties:
>>>>>>>               - brcm,brcmstb-memc-ddr-rev-c.1.2
>>>>>>>               - brcm,brcmstb-memc-ddr-rev-c.1.3
>>>>>>>               - brcm,brcmstb-memc-ddr-rev-c.1.4
>>>>>>> +          - brcm,brcmstb-memc-ddr-rev-c.1.5
>>>>>>
>>>>>> You should use v2.1 fallback and drop driver patch. Or explain in
>>>>>> commit briefly why different approach is suitable.
>>>>>
>>>>> Are you suggesting that we should have fallback compatible strings, such
>>>>> that we have something like this:
>>>>>
>>>>> compatible = "brcm,brcmstb-memc-ddr-rev-c.1.5",
>>>>> "brcm,brcmstb-memc-ddr-rev-c", "brcm,brcmstb-memc-ddr"
>>>>>
>>>>> and the driver only needs to match on "brcm,brcmstb-memc-ddr-rev-c" and
>>>>> apply the adequate register offset table?
>>>>
>>>> Almost, fallback should be brcm,brcmstb-memc-ddr-rev-b.2.1 or whatever
>>>> was in the driver first or whatever is the oldest known common
>>>> interface.
>>>>
>>>> brcm,brcmstb-memc-ddr-rev-c is not a specific compatible.
>>>>
>>>>> If so, that is not how the current binding, and therefore DTBs are being
>>>>> deployed, so that will introduce a breakage until we update all DTBs in the
>>>>> field...
>>>>
>>>> No. First, I thought about new comaptible so the one you add here. No
>>>> breakage, it's new compatible. This saves you these pointless updates of
>>>> driver everytime you add new compatible.
>>>
>>> Yes, but that is not how the binding has been defined until now, so all
>>> of the DTBs out there have:
>>>
>>> compatible = "brcm,brcmstb-memc-ddr-rev-b.2.x", "brcm,brcmstb-memc-ddr"
>>
>> I don't understand the problem. We talk about new devices here, it does
>> not matter what existing/old devices have in binding in that matter.
> 
> It does matter, because the DTBs that contain this compatible string are 
> already in the field, we cannot retrofit them overnight with an 
> additional compatible string in order to provide a fallback. Because 

Then they run downstream and it does not matter whether you change here
something or not. And anyway downstream choices do not shape choices
here. We never accept bindings just because downstream already shipped
something.

> this is submitted now does not mean it is a new device, this was part of 
> my backlog to get submitted earlier on.

Well, that's great yet is not an argument in the bindings.

> 
>>
>>>
>>> (where X is in range [1..5])
>>>
>>> and there is no fallback defined to "brcm,brcmstb-memc-ddr-rev-b.2.1",
>>> so it is not like we can retrofit that easily by adding one now.
>>>
>>>   > > Second, you can introduce fallbacks to older compatibles as well -
>>> there
>>>> will be no breakage, because you add one more compatible. The old
>>>> compatibles (covered by fallback) of course stays in the driver, so
>>>> there is no breakage at all. We did it multiple times for several
>>>> different bindings in Qualcomm. People were doing exactly the same:
>>>> adding compatible for new device to binding and driver, without
>>>> considering the compatibility at all.
>>>>
>>>> Except being logically correct choice - using fallbacks - this really
>>>> has huge benefits when later upstreaming complete, big SoCs, like we do
>>>> for latest Qualcomm SoCs: several changes will be only bindings updates.
>>>
>>> Yes, there are advantages to using fallbacks and we (ab)use that
>>> whenever practical.
>>>
>>> The driver only uses a very limited subset of registers (for now), the
>>> registers change between minor revisions as well in a way that using a
>>> fallback like "brcm,brcmstb-memc-ddr-rev-b.2.1" is not accurate enough
>>> not practical. In particular for some of the changes that I am thinking
>>> of adding later on, we would need the precise minor version because the
>>> behavior and/or register interface is subtly different that this matters.
>>
>> Devices work fine now with the same driver data, so they are compatible.
> 
> For now they do, but not for other features that will be submitted 
> later, they will not be, and at that point we will need to know that 
> this is a rev C.1.5 controller versus say a rev C.1.4 controller.

Which is exactly what compatibility is for.



Best regards,
Krzysztof

Re: [PATCH 1/2] dt-bindings: memory-controller: Document rev c.1.5 compatible

[Florian Fainelli]

On 12/26/2024 10:58 PM, Krzysztof Kozlowski wrote:
> On 26/12/2024 19:50, Florian Fainelli wrote:
>>
>>
>> On 12/21/2024 12:04 PM, Krzysztof Kozlowski wrote:
>>> On 20/12/2024 00:43, Florian Fainelli wrote:
>>>> On 12/19/24 00:58, Krzysztof Kozlowski wrote:
>>>>> On Wed, Dec 18, 2024 at 09:15:08AM -0800, Florian Fainelli wrote:
>>>>>> On 12/18/24 03:37, Krzysztof Kozlowski wrote:
>>>>>>> On Tue, Dec 17, 2024 at 11:44:38AM -0800, Florian Fainelli wrote:
>>>>>>>> Document the revision c.1.5 compatible string that is present on newer
>>>>>>>> Broadcom STB memory controllers (74165 and onwards).
>>>>>>>>
>>>>>>>> Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
>>>>>>>> ---
>>>>>>>>      .../bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml       | 1 +
>>>>>>>>      1 file changed, 1 insertion(+)
>>>>>>>>
>>>>>>>> diff --git a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>>>>>>>> index 4b072c879b02..99d79ccd1036 100644
>>>>>>>> --- a/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>>>>>>>> +++ b/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml
>>>>>>>> @@ -29,6 +29,7 @@ properties:
>>>>>>>>                - brcm,brcmstb-memc-ddr-rev-c.1.2
>>>>>>>>                - brcm,brcmstb-memc-ddr-rev-c.1.3
>>>>>>>>                - brcm,brcmstb-memc-ddr-rev-c.1.4
>>>>>>>> +          - brcm,brcmstb-memc-ddr-rev-c.1.5
>>>>>>>
>>>>>>> You should use v2.1 fallback and drop driver patch. Or explain in
>>>>>>> commit briefly why different approach is suitable.
>>>>>>
>>>>>> Are you suggesting that we should have fallback compatible strings, such
>>>>>> that we have something like this:
>>>>>>
>>>>>> compatible = "brcm,brcmstb-memc-ddr-rev-c.1.5",
>>>>>> "brcm,brcmstb-memc-ddr-rev-c", "brcm,brcmstb-memc-ddr"
>>>>>>
>>>>>> and the driver only needs to match on "brcm,brcmstb-memc-ddr-rev-c" and
>>>>>> apply the adequate register offset table?
>>>>>
>>>>> Almost, fallback should be brcm,brcmstb-memc-ddr-rev-b.2.1 or whatever
>>>>> was in the driver first or whatever is the oldest known common
>>>>> interface.
>>>>>
>>>>> brcm,brcmstb-memc-ddr-rev-c is not a specific compatible.
>>>>>
>>>>>> If so, that is not how the current binding, and therefore DTBs are being
>>>>>> deployed, so that will introduce a breakage until we update all DTBs in the
>>>>>> field...
>>>>>
>>>>> No. First, I thought about new comaptible so the one you add here. No
>>>>> breakage, it's new compatible. This saves you these pointless updates of
>>>>> driver everytime you add new compatible.
>>>>
>>>> Yes, but that is not how the binding has been defined until now, so all
>>>> of the DTBs out there have:
>>>>
>>>> compatible = "brcm,brcmstb-memc-ddr-rev-b.2.x", "brcm,brcmstb-memc-ddr"
>>>
>>> I don't understand the problem. We talk about new devices here, it does
>>> not matter what existing/old devices have in binding in that matter.
>>
>> It does matter, because the DTBs that contain this compatible string are
>> already in the field, we cannot retrofit them overnight with an
>> additional compatible string in order to provide a fallback. Because
> 
> Then they run downstream and it does not matter whether you change here
> something or not. And anyway downstream choices do not shape choices
> here. We never accept bindings just because downstream already shipped
> something.

Fair enough, but it does matter anyway, because as you can see from the 
original binding that you accepted [1] we have a precedent for having 
compatible strings of the form:

compatible = "brcm,brcmstb-memc-ddr-rev-c.1.x"

with each possible value of X being documented, this is just adding 
another way, and surprise, keeps things consistent.

Why did not you reject the original binding back then? Why the change of 
heart now?

All of this feels from my side very very arbitrary, and as such it is 
difficult to document for other people and build a correct mental model of.

[1]: 
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/Documentation/devicetree/bindings/memory-controllers/brcm,brcmstb-memc-ddr.yaml?id=fa0321ba51ddff78ebe3c7c945830a85c987e3ed

> 
>> this is submitted now does not mean it is a new device, this was part of
>> my backlog to get submitted earlier on.
> 
> Well, that's great yet is not an argument in the bindings.
 > >>
>>>
>>>>
>>>> (where X is in range [1..5])
>>>>
>>>> and there is no fallback defined to "brcm,brcmstb-memc-ddr-rev-b.2.1",
>>>> so it is not like we can retrofit that easily by adding one now.
>>>>
>>>>    > > Second, you can introduce fallbacks to older compatibles as well -
>>>> there
>>>>> will be no breakage, because you add one more compatible. The old
>>>>> compatibles (covered by fallback) of course stays in the driver, so
>>>>> there is no breakage at all. We did it multiple times for several
>>>>> different bindings in Qualcomm. People were doing exactly the same:
>>>>> adding compatible for new device to binding and driver, without
>>>>> considering the compatibility at all.
>>>>>
>>>>> Except being logically correct choice - using fallbacks - this really
>>>>> has huge benefits when later upstreaming complete, big SoCs, like we do
>>>>> for latest Qualcomm SoCs: several changes will be only bindings updates.
>>>>
>>>> Yes, there are advantages to using fallbacks and we (ab)use that
>>>> whenever practical.
>>>>
>>>> The driver only uses a very limited subset of registers (for now), the
>>>> registers change between minor revisions as well in a way that using a
>>>> fallback like "brcm,brcmstb-memc-ddr-rev-b.2.1" is not accurate enough
>>>> not practical. In particular for some of the changes that I am thinking
>>>> of adding later on, we would need the precise minor version because the
>>>> behavior and/or register interface is subtly different that this matters.
>>>
>>> Devices work fine now with the same driver data, so they are compatible.
>>
>> For now they do, but not for other features that will be submitted
>> later, they will not be, and at that point we will need to know that
>> this is a rev C.1.5 controller versus say a rev C.1.4 controller.
> 
> Which is exactly what compatibility is for.

Are you saying that the compatible strings must now all be updated to be 
of the form:

compatible = "brcm,brcmstb-memc-ddr-rev-c.1.5", 
"brcm,brcmstb-memc-ddr-rev-c.1.x", "brcm,brcmstb-memc-ddr"

or something different?
-- 
Florian