* [PATCH] arm64: dts: monaco: Fix CSI I2C controllers default bus frequency
@ 2026-07-23 11:48 Hans de Goede
2026-07-24 11:42 ` Konrad Dybcio
2026-07-31 3:24 ` Bjorn Andersson
0 siblings, 2 replies; 7+ messages in thread
From: Hans de Goede @ 2026-07-23 11:48 UTC (permalink / raw)
To: Bjorn Andersson, Konrad Dybcio, Krzysztof Kozlowski
Cc: Hans de Goede, Conor Dooley, Loic Poulain, Wenmeng Liu,
devicetree, linux-arm-msm
1 MHz is quite fast for an I2C bus and not necessarly reliable, change this
to 400 KHz which is a more reasonable default.
Signed-off-by: Hans de Goede <johannes.goede@oss.qualcomm.com>
---
arch/arm64/boot/dts/qcom/monaco.dtsi | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/arch/arm64/boot/dts/qcom/monaco.dtsi b/arch/arm64/boot/dts/qcom/monaco.dtsi
index a22be507ed9c..d3b59b9e84e6 100644
--- a/arch/arm64/boot/dts/qcom/monaco.dtsi
+++ b/arch/arm64/boot/dts/qcom/monaco.dtsi
@@ -5472,14 +5472,14 @@ cci0: cci@ac13000 {
cci0_i2c0: i2c-bus@0 {
reg = <0>;
- clock-frequency = <1000000>;
+ clock-frequency = <400000>;
#address-cells = <1>;
#size-cells = <0>;
};
cci0_i2c1: i2c-bus@1 {
reg = <1>;
- clock-frequency = <1000000>;
+ clock-frequency = <400000>;
#address-cells = <1>;
#size-cells = <0>;
};
@@ -5509,14 +5509,14 @@ cci1: cci@ac14000 {
cci1_i2c0: i2c-bus@0 {
reg = <0>;
- clock-frequency = <1000000>;
+ clock-frequency = <400000>;
#address-cells = <1>;
#size-cells = <0>;
};
cci1_i2c1: i2c-bus@1 {
reg = <1>;
- clock-frequency = <1000000>;
+ clock-frequency = <400000>;
#address-cells = <1>;
#size-cells = <0>;
};
@@ -5546,14 +5546,14 @@ cci2: cci@ac15000 {
cci2_i2c0: i2c-bus@0 {
reg = <0>;
- clock-frequency = <1000000>;
+ clock-frequency = <400000>;
#address-cells = <1>;
#size-cells = <0>;
};
cci2_i2c1: i2c-bus@1 {
reg = <1>;
- clock-frequency = <1000000>;
+ clock-frequency = <400000>;
#address-cells = <1>;
#size-cells = <0>;
};
--
2.55.0
^ permalink raw reply related [flat|nested] 7+ messages in thread
* Re: [PATCH] arm64: dts: monaco: Fix CSI I2C controllers default bus frequency
2026-07-23 11:48 [PATCH] arm64: dts: monaco: Fix CSI I2C controllers default bus frequency Hans de Goede
@ 2026-07-24 11:42 ` Konrad Dybcio
2026-07-24 14:22 ` Hans de Goede
2026-07-31 3:24 ` Bjorn Andersson
1 sibling, 1 reply; 7+ messages in thread
From: Konrad Dybcio @ 2026-07-24 11:42 UTC (permalink / raw)
To: Hans de Goede, Bjorn Andersson, Konrad Dybcio,
Krzysztof Kozlowski
Cc: Conor Dooley, Loic Poulain, Wenmeng Liu, devicetree,
linux-arm-msm
On 7/23/26 1:48 PM, Hans de Goede wrote:
> 1 MHz is quite fast for an I2C bus and not necessarly reliable, change this
> to 400 KHz which is a more reasonable default.
>
> Signed-off-by: Hans de Goede <johannes.goede@oss.qualcomm.com>
I guess this is a question of what's more common among client
devices. I assume you have some that aren't happy with 1 MHz?
Konrad
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] arm64: dts: monaco: Fix CSI I2C controllers default bus frequency
2026-07-24 11:42 ` Konrad Dybcio
@ 2026-07-24 14:22 ` Hans de Goede
2026-07-24 15:22 ` Konrad Dybcio
2026-07-31 6:41 ` Vladimir Zapolskiy
0 siblings, 2 replies; 7+ messages in thread
From: Hans de Goede @ 2026-07-24 14:22 UTC (permalink / raw)
To: Konrad Dybcio, Bjorn Andersson, Konrad Dybcio,
Krzysztof Kozlowski
Cc: Conor Dooley, Loic Poulain, Wenmeng Liu, devicetree,
linux-arm-msm
Hi,
On 24-Jul-26 1:42 PM, Konrad Dybcio wrote:
> On 7/23/26 1:48 PM, Hans de Goede wrote:
>> 1 MHz is quite fast for an I2C bus and not necessarly reliable, change this
>> to 400 KHz which is a more reasonable default.
>>
>> Signed-off-by: Hans de Goede <johannes.goede@oss.qualcomm.com>
>
>
> I guess this is a question of what's more common among client
> devices. I assume you have some that aren't happy with 1 MHz?
Looking at ACPI tables for many x86 devices 400 KHz is the standard,
with some devices going down to 100 KHz. 1 MHz is pretty much
unheard off.
Note the official I2C spec is limited to 100 KHz anything faster then
that is an extension.
My specific reason for submitting this patch is imx219 sensors on
Mona sometimes failing to initialize at 1MHz which is kinda to
be expected when going for 10x the I2C specification max value.
Regards,
Hans
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] arm64: dts: monaco: Fix CSI I2C controllers default bus frequency
2026-07-24 14:22 ` Hans de Goede
@ 2026-07-24 15:22 ` Konrad Dybcio
2026-07-31 6:41 ` Vladimir Zapolskiy
1 sibling, 0 replies; 7+ messages in thread
From: Konrad Dybcio @ 2026-07-24 15:22 UTC (permalink / raw)
To: Hans de Goede, Bjorn Andersson, Konrad Dybcio,
Krzysztof Kozlowski
Cc: Conor Dooley, Loic Poulain, Wenmeng Liu, devicetree,
linux-arm-msm
On 7/24/26 4:22 PM, Hans de Goede wrote:
> Hi,
>
> On 24-Jul-26 1:42 PM, Konrad Dybcio wrote:
>> On 7/23/26 1:48 PM, Hans de Goede wrote:
>>> 1 MHz is quite fast for an I2C bus and not necessarly reliable, change this
>>> to 400 KHz which is a more reasonable default.
>>>
>>> Signed-off-by: Hans de Goede <johannes.goede@oss.qualcomm.com>
>>
>>
>> I guess this is a question of what's more common among client
>> devices. I assume you have some that aren't happy with 1 MHz?
>
> Looking at ACPI tables for many x86 devices 400 KHz is the standard,
> with some devices going down to 100 KHz. 1 MHz is pretty much
> unheard off.
>
> Note the official I2C spec is limited to 100 KHz anything faster then
> that is an extension.
>
> My specific reason for submitting this patch is imx219 sensors on
> Mona sometimes failing to initialize at 1MHz which is kinda to
> be expected when going for 10x the I2C specification max value.
FWIW I don't mind this patch but wanted to know more background
Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
Konrad
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] arm64: dts: monaco: Fix CSI I2C controllers default bus frequency
2026-07-23 11:48 [PATCH] arm64: dts: monaco: Fix CSI I2C controllers default bus frequency Hans de Goede
2026-07-24 11:42 ` Konrad Dybcio
@ 2026-07-31 3:24 ` Bjorn Andersson
1 sibling, 0 replies; 7+ messages in thread
From: Bjorn Andersson @ 2026-07-31 3:24 UTC (permalink / raw)
To: Konrad Dybcio, Krzysztof Kozlowski, Hans de Goede
Cc: Conor Dooley, Loic Poulain, Wenmeng Liu, devicetree,
linux-arm-msm
On Thu, 23 Jul 2026 13:48:44 +0200, Hans de Goede wrote:
> 1 MHz is quite fast for an I2C bus and not necessarly reliable, change this
> to 400 KHz which is a more reasonable default.
>
>
Applied, thanks!
[1/1] arm64: dts: monaco: Fix CSI I2C controllers default bus frequency
commit: 1c2935b75f78fa74ef9cc37d95c4f734cbc0c15f
Best regards,
--
Bjorn Andersson <andersson@kernel.org>
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] arm64: dts: monaco: Fix CSI I2C controllers default bus frequency
2026-07-24 14:22 ` Hans de Goede
2026-07-24 15:22 ` Konrad Dybcio
@ 2026-07-31 6:41 ` Vladimir Zapolskiy
2026-08-20 10:11 ` Hans de Goede
1 sibling, 1 reply; 7+ messages in thread
From: Vladimir Zapolskiy @ 2026-07-31 6:41 UTC (permalink / raw)
To: Hans de Goede, Konrad Dybcio, Bjorn Andersson, Konrad Dybcio,
Krzysztof Kozlowski
Cc: Conor Dooley, Loic Poulain, Wenmeng Liu, devicetree,
linux-arm-msm
On 7/24/26 17:22, Hans de Goede wrote:
> Hi,
>
> On 24-Jul-26 1:42 PM, Konrad Dybcio wrote:
>> On 7/23/26 1:48 PM, Hans de Goede wrote:
>>> 1 MHz is quite fast for an I2C bus and not necessarly reliable, change this
>>> to 400 KHz which is a more reasonable default.
>>>
>>> Signed-off-by: Hans de Goede <johannes.goede@oss.qualcomm.com>
>>
>>
>> I guess this is a question of what's more common among client
>> devices. I assume you have some that aren't happy with 1 MHz?
>
> Looking at ACPI tables for many x86 devices 400 KHz is the standard,
> with some devices going down to 100 KHz. 1 MHz is pretty much
> unheard off.
>
> Note the official I2C spec is limited to 100 KHz anything faster then
> that is an extension.
>
> My specific reason for submitting this patch is imx219 sensors on
> Mona sometimes failing to initialize at 1MHz which is kinda to
> be expected when going for 10x the I2C specification max value.
>
1 MHz I2C bus speed is in the v3 of the I2C specification dated 2007,
its name is Fast-mode Plus (Fm+), thus the statement above is invalid.
As it's been said on the list recently likely the problem with sensor
controls on a number of platforms is due to invalid setting of I2C bus
speeds, e.g. out-of-spec 512 KHz I2C bus speed instead of wanted 1 MHz,
or 204.8 KHz I2C bus speed instead of 400 KHz, which is done by this
applied change.
The list of affected platform .dtsi files is:
* kodiak.dtsi
* lemans.dtsi
* milos.dtsi
* monaco.dtsi
* sc8280xp.dtsi
* sdm670.dtsi
* sm8250.dtsi
* sm8450.dtsi
* sm8550.dtsi
* sm8650.dtsi
* talos.dtsi
An example of a "fix" (it's not a fix, but a simple workaround of
the flaw in the driver) for SM8650 and other platforms can be like
the one recently published:
https://lore.kernel.org/linux-arm-msm/20260729112719.1151700-1-vladimir.zapolskiy@linaro.org
--
Best wishes,
Vladimir
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] arm64: dts: monaco: Fix CSI I2C controllers default bus frequency
2026-07-31 6:41 ` Vladimir Zapolskiy
@ 2026-08-20 10:11 ` Hans de Goede
0 siblings, 0 replies; 7+ messages in thread
From: Hans de Goede @ 2026-08-20 10:11 UTC (permalink / raw)
To: Vladimir Zapolskiy, Konrad Dybcio, Bjorn Andersson, Konrad Dybcio,
Krzysztof Kozlowski
Cc: Conor Dooley, Loic Poulain, Wenmeng Liu, devicetree,
linux-arm-msm
Hi,
On 31-Jul-26 08:41, Vladimir Zapolskiy wrote:
> On 7/24/26 17:22, Hans de Goede wrote:
>> Hi,
>>
>> On 24-Jul-26 1:42 PM, Konrad Dybcio wrote:
>>> On 7/23/26 1:48 PM, Hans de Goede wrote:
>>>> 1 MHz is quite fast for an I2C bus and not necessarly reliable, change this
>>>> to 400 KHz which is a more reasonable default.
>>>>
>>>> Signed-off-by: Hans de Goede <johannes.goede@oss.qualcomm.com>
>>>
>>>
>>> I guess this is a question of what's more common among client
>>> devices. I assume you have some that aren't happy with 1 MHz?
>>
>> Looking at ACPI tables for many x86 devices 400 KHz is the standard,
>> with some devices going down to 100 KHz. 1 MHz is pretty much
>> unheard off.
>>
>> Note the official I2C spec is limited to 100 KHz anything faster then
>> that is an extension.
>>
>> My specific reason for submitting this patch is imx219 sensors on
>> Mona sometimes failing to initialize at 1MHz which is kinda to
>> be expected when going for 10x the I2C specification max value.
>>
>
> 1 MHz I2C bus speed is in the v3 of the I2C specification dated 2007,
> its name is Fast-mode Plus (Fm+), thus the statement above is invalid.
Yes I'm aware of Fast-mode Plus being in the spec, but it is
an *extension* to the original spec.
The only speed which is guaranteed to work everywhere is 100 kHz.
Now a days 400 KHz usually will work, but e.g. the DSDT of
the Samsung Galaxy Go LTE NP345XLA Snapdragon laptop limits one of
the busses to 100 kHz because the TI BQ25710 cannot handle 400 kHz
(its an SMBus device and the SMBus 2.0 spec limits the bus to 100 KHz).
Or since this is about the CSI I2C controllers, picking a random
camera sensor I've a datasheet for here, the ov8110 lists the minimum
data low time as 1.3 ųs, which means it cannot run at 1 MHz either.
I've seen a ton of I2C devices and I2C device descriptions in
ACPI tables of x86_64 and them using 1MHz is something which I've
never seen.
So 1 MHz just is a really poor default independent of the issues
you mention where we end up with a different speed then requested
due to the I2C controller parent clock running at the wrong freq.
Regards,
Hans
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2026-08-20 10:11 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-07-23 11:48 [PATCH] arm64: dts: monaco: Fix CSI I2C controllers default bus frequency Hans de Goede
2026-07-24 11:42 ` Konrad Dybcio
2026-07-24 14:22 ` Hans de Goede
2026-07-24 15:22 ` Konrad Dybcio
2026-07-31 6:41 ` Vladimir Zapolskiy
2026-08-20 10:11 ` Hans de Goede
2026-07-31 3:24 ` Bjorn Andersson
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox