diff mbox series

ARM: dts: am5729-beagleboneai: drop unused Broadcom WiFi properties

Message ID 20210505140015.60013-1-krzysztof.kozlowski@canonical.com (mailing list archive)
State New, archived
Headers show
Series ARM: dts: am5729-beagleboneai: drop unused Broadcom WiFi properties | expand

Commit Message

Krzysztof Kozlowski May 5, 2021, 2 p.m. UTC
The brcm,sd-head-align, brcm,sd_head_align and brcm,sd_sgentry_align are
not used by the driver and not present in dtschema:

  arch/arm/boot/dts/am5729-beagleboneai.dt.yaml: wifi@1:
    'brcm,sd-head-align', 'brcm,sd_head_align', 'brcm,sd_sgentry_align' do not match any of the regexes: 'pinctrl-[0-9]+'

Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
---
 arch/arm/boot/dts/am5729-beagleboneai.dts | 4 ----
 1 file changed, 4 deletions(-)

Comments

Robert Nelson May 5, 2021, 2:15 p.m. UTC | #1
On Wed, May 5, 2021 at 9:04 AM Krzysztof Kozlowski
<krzysztof.kozlowski@canonical.com> wrote:
>
> The brcm,sd-head-align, brcm,sd_head_align and brcm,sd_sgentry_align are
> not used by the driver and not present in dtschema:
>
>   arch/arm/boot/dts/am5729-beagleboneai.dt.yaml: wifi@1:
>     'brcm,sd-head-align', 'brcm,sd_head_align', 'brcm,sd_sgentry_align' do not match any of the regexes: 'pinctrl-[0-9]+'
>
> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
> ---
>  arch/arm/boot/dts/am5729-beagleboneai.dts | 4 ----
>  1 file changed, 4 deletions(-)
>
> diff --git a/arch/arm/boot/dts/am5729-beagleboneai.dts b/arch/arm/boot/dts/am5729-beagleboneai.dts
> index 149cfafb90bf..d3b2e87a6266 100644
> --- a/arch/arm/boot/dts/am5729-beagleboneai.dts
> +++ b/arch/arm/boot/dts/am5729-beagleboneai.dts
> @@ -582,10 +582,6 @@ brcmf: wifi@1 {
>                 reg = <1>;
>                 compatible = "brcm,bcm4329-fmac";
>
> -               brcm,sd-head-align = <4>;
> -               brcm,sd_head_align = <4>;
> -               brcm,sd_sgentry_align = <512>;
> -
>                 interrupt-parent = <&gpio3>;
>                 interrupts = <23 IRQ_TYPE_LEVEL_LOW>;
>                 interrupt-names = "host-wake";
> --
> 2.25.1
>

NAK, for what it is worth..

This is a plain old shitty situation, that everyone that mistakenly
chooses a BRCM based chipset in a design faces, then learns to NEVER
touch that chipset again..

These nodes have been posted before for mainline inclusion but always
get killed off:

https://lkml.org/lkml/2018/3/18/356

The TI AM57xx SDIO needs these flags for the brcmfmac driver to actually work.

I don't see the brcmfmac developers actually fixing it, as "out of
tree" work for brcmfmac is 100+ patches long:

https://www.cypress.com/documentation/software-and-drivers-archive/wifi-bt-linux-archive?source=search&cat=other

Regards,
Krzysztof Kozlowski May 5, 2021, 2:22 p.m. UTC | #2
On 05/05/2021 10:15, Robert Nelson wrote:
> On Wed, May 5, 2021 at 9:04 AM Krzysztof Kozlowski
> <krzysztof.kozlowski@canonical.com> wrote:
>>
>> The brcm,sd-head-align, brcm,sd_head_align and brcm,sd_sgentry_align are
>> not used by the driver and not present in dtschema:
>>
>>   arch/arm/boot/dts/am5729-beagleboneai.dt.yaml: wifi@1:
>>     'brcm,sd-head-align', 'brcm,sd_head_align', 'brcm,sd_sgentry_align' do not match any of the regexes: 'pinctrl-[0-9]+'
>>
>> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
>> ---
>>  arch/arm/boot/dts/am5729-beagleboneai.dts | 4 ----
>>  1 file changed, 4 deletions(-)
>>
>> diff --git a/arch/arm/boot/dts/am5729-beagleboneai.dts b/arch/arm/boot/dts/am5729-beagleboneai.dts
>> index 149cfafb90bf..d3b2e87a6266 100644
>> --- a/arch/arm/boot/dts/am5729-beagleboneai.dts
>> +++ b/arch/arm/boot/dts/am5729-beagleboneai.dts
>> @@ -582,10 +582,6 @@ brcmf: wifi@1 {
>>                 reg = <1>;
>>                 compatible = "brcm,bcm4329-fmac";
>>
>> -               brcm,sd-head-align = <4>;
>> -               brcm,sd_head_align = <4>;
>> -               brcm,sd_sgentry_align = <512>;
>> -
>>                 interrupt-parent = <&gpio3>;
>>                 interrupts = <23 IRQ_TYPE_LEVEL_LOW>;
>>                 interrupt-names = "host-wake";
>> --
>> 2.25.1
>>
> 
> NAK, for what it is worth..
> 
> This is a plain old shitty situation, that everyone that mistakenly
> chooses a BRCM based chipset in a design faces, then learns to NEVER
> touch that chipset again..
> 
> These nodes have been posted before for mainline inclusion but always
> get killed off:
> 
> https://lkml.org/lkml/2018/3/18/356
> 
> The TI AM57xx SDIO needs these flags for the brcmfmac driver to actually work.
> 
> I don't see the brcmfmac developers actually fixing it, as "out of
> tree" work for brcmfmac is 100+ patches long:


Thanks for explanation. I understand these are actually used by
out-of-tree driver? That's a valid reason to add them to the binding then...

> 
> https://www.cypress.com/documentation/software-and-drivers-archive/wifi-bt-linux-archive?source=search&cat=other

Requires login - I am not able to get the sources.

Best regards,
Krzysztof
Robert Nelson May 5, 2021, 3:10 p.m. UTC | #3
> > I don't see the brcmfmac developers actually fixing it, as "out of
> > tree" work for brcmfmac is 100+ patches long:
>
>
> Thanks for explanation. I understand these are actually used by
> out-of-tree driver? That's a valid reason to add them to the binding then...
>
> >
> > https://www.cypress.com/documentation/software-and-drivers-archive/wifi-bt-linux-archive?source=search&cat=other
>
> Requires login - I am not able to get the sources.

Sorry Krzysztof,

I forgot that a login is required, here is a quick mirror (with a
screenshot of the page.)

https://rcn-ee.net/workshops/cypress-fmac/

Screenshot: https://rcn-ee.net/workshops/cypress-fmac/brcm.png

There patch with that binding:

https://rcn-ee.net/workshops/cypress-fmac/cypress-fmac-v5.4.18-2021_0114/cypress-patch-v5.4.18-2021_0114/cypress-patch/0002-non-upstream-add-sg-parameters-dts-parsing.patch

Thank You!


Regards,

--
Robert Nelson
https://rcn-ee.com/
Krzysztof Kozlowski May 5, 2021, 3:12 p.m. UTC | #4
On 05/05/2021 11:10, Robert Nelson wrote:
>>> I don't see the brcmfmac developers actually fixing it, as "out of
>>> tree" work for brcmfmac is 100+ patches long:
>>
>>
>> Thanks for explanation. I understand these are actually used by
>> out-of-tree driver? That's a valid reason to add them to the binding then...
>>
>>>
>>> https://www.cypress.com/documentation/software-and-drivers-archive/wifi-bt-linux-archive?source=search&cat=other
>>
>> Requires login - I am not able to get the sources.
> 
> Sorry Krzysztof,
> 
> I forgot that a login is required, here is a quick mirror (with a
> screenshot of the page.)
> 
> https://rcn-ee.net/workshops/cypress-fmac/
> 
> Screenshot: https://rcn-ee.net/workshops/cypress-fmac/brcm.png
> 
> There patch with that binding:
> 
> https://rcn-ee.net/workshops/cypress-fmac/cypress-fmac-v5.4.18-2021_0114/cypress-patch-v5.4.18-2021_0114/cypress-patch/0002-non-upstream-add-sg-parameters-dts-parsing.patch

Thanks, I see it now!


Best regards,
Krzysztof
diff mbox series

Patch

diff --git a/arch/arm/boot/dts/am5729-beagleboneai.dts b/arch/arm/boot/dts/am5729-beagleboneai.dts
index 149cfafb90bf..d3b2e87a6266 100644
--- a/arch/arm/boot/dts/am5729-beagleboneai.dts
+++ b/arch/arm/boot/dts/am5729-beagleboneai.dts
@@ -582,10 +582,6 @@  brcmf: wifi@1 {
 		reg = <1>;
 		compatible = "brcm,bcm4329-fmac";
 
-		brcm,sd-head-align = <4>;
-		brcm,sd_head_align = <4>;
-		brcm,sd_sgentry_align = <512>;
-
 		interrupt-parent = <&gpio3>;
 		interrupts = <23 IRQ_TYPE_LEVEL_LOW>;
 		interrupt-names = "host-wake";