[PATCH] i2c: qcom-geni: fix I2C frequency table to achieve accurate bus rates

[PATCH] i2c: qcom-geni: fix I2C frequency table to achieve accurate bus rates

[Kathiravan Thirumoorthy]

Update the I2C frequency table to match the recommended values
specified in the I2C hardware programming guide. In the current IPQ5424
configuration where 32MHz is the source clock, the I2C bus frequencies do
not meet expectations—for instance, 363KHz is achieved instead of the
expected 400KHz.

Cc: stable@kernel.org
Fixes: 506bb2ab0075 ("i2c: qcom-geni: Support systems with 32MHz serial engine clock")
Signed-off-by: Kathiravan Thirumoorthy <kathiravan.thirumoorthy@oss.qualcomm.com>
---
 drivers/i2c/busses/i2c-qcom-geni.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/i2c/busses/i2c-qcom-geni.c b/drivers/i2c/busses/i2c-qcom-geni.c
index ccea575fb7838db864ca4a2b21ebb3835b2567b2..2fec7b44bfc1baec68e321a9f57de4156120919b 100644
--- a/drivers/i2c/busses/i2c-qcom-geni.c
+++ b/drivers/i2c/busses/i2c-qcom-geni.c
@@ -155,9 +155,9 @@ static const struct geni_i2c_clk_fld geni_i2c_clk_map_19p2mhz[] = {
 
 /* source_clock = 32 MHz */
 static const struct geni_i2c_clk_fld geni_i2c_clk_map_32mhz[] = {
-	{ I2C_MAX_STANDARD_MODE_FREQ, 8, 14, 18, 40 },
-	{ I2C_MAX_FAST_MODE_FREQ, 4,  3, 11, 20 },
-	{ I2C_MAX_FAST_MODE_PLUS_FREQ, 2, 3,  6, 15 },
+	{ I2C_MAX_STANDARD_MODE_FREQ, 8, 14, 18, 38 },
+	{ I2C_MAX_FAST_MODE_FREQ, 4,  3, 9, 19 },
+	{ I2C_MAX_FAST_MODE_PLUS_FREQ, 2, 3, 5, 15 },
 	{}
 };
 

---
base-commit: edef457004774e598fc4c1b7d1d4f0bcd9d0bb30
change-id: 20250513-i2c-bus-freq-ac46343869a4

Best regards,
-- 
Kathiravan Thirumoorthy <kathiravan.thirumoorthy@oss.qualcomm.com>

Re: [PATCH] i2c: qcom-geni: fix I2C frequency table to achieve accurate bus rates

[Andi Shyti]

Hi,

On Tue, May 13, 2025 at 04:38:33PM +0530, Kathiravan Thirumoorthy wrote:
> Update the I2C frequency table to match the recommended values
> specified in the I2C hardware programming guide. In the current IPQ5424
> configuration where 32MHz is the source clock, the I2C bus frequencies do
> not meet expectations—for instance, 363KHz is achieved instead of the
> expected 400KHz.
> 
> Cc: stable@kernel.org
> Fixes: 506bb2ab0075 ("i2c: qcom-geni: Support systems with 32MHz serial engine clock")
> Signed-off-by: Kathiravan Thirumoorthy <kathiravan.thirumoorthy@oss.qualcomm.com>
> ---
>  drivers/i2c/busses/i2c-qcom-geni.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/i2c/busses/i2c-qcom-geni.c b/drivers/i2c/busses/i2c-qcom-geni.c
> index ccea575fb7838db864ca4a2b21ebb3835b2567b2..2fec7b44bfc1baec68e321a9f57de4156120919b 100644
> --- a/drivers/i2c/busses/i2c-qcom-geni.c
> +++ b/drivers/i2c/busses/i2c-qcom-geni.c
> @@ -155,9 +155,9 @@ static const struct geni_i2c_clk_fld geni_i2c_clk_map_19p2mhz[] = {
>  
>  /* source_clock = 32 MHz */
>  static const struct geni_i2c_clk_fld geni_i2c_clk_map_32mhz[] = {
> -	{ I2C_MAX_STANDARD_MODE_FREQ, 8, 14, 18, 40 },
> -	{ I2C_MAX_FAST_MODE_FREQ, 4,  3, 11, 20 },
> -	{ I2C_MAX_FAST_MODE_PLUS_FREQ, 2, 3,  6, 15 },
> +	{ I2C_MAX_STANDARD_MODE_FREQ, 8, 14, 18, 38 },
> +	{ I2C_MAX_FAST_MODE_FREQ, 4,  3, 9, 19 },
> +	{ I2C_MAX_FAST_MODE_PLUS_FREQ, 2, 3, 5, 15 },

argh!

Can someone from Qualcomm look at this fix? Mukesh, Viken?

Thanks,
Andi

>  	{}
>  };
>  
> 
> ---
> base-commit: edef457004774e598fc4c1b7d1d4f0bcd9d0bb30
> change-id: 20250513-i2c-bus-freq-ac46343869a4
> 
> Best regards,
> -- 
> Kathiravan Thirumoorthy <kathiravan.thirumoorthy@oss.qualcomm.com>
> 

Re: [PATCH] i2c: qcom-geni: fix I2C frequency table to achieve accurate bus rates

[Mukesh Kumar Savaliya]

Thanks Andi !

On 5/14/2025 9:04 PM, Andi Shyti wrote:

[...]
>>   /* source_clock = 32 MHz */
>>   static const struct geni_i2c_clk_fld geni_i2c_clk_map_32mhz[] = {
>> -	{ I2C_MAX_STANDARD_MODE_FREQ, 8, 14, 18, 40 },
>> -	{ I2C_MAX_FAST_MODE_FREQ, 4,  3, 11, 20 },
>> -	{ I2C_MAX_FAST_MODE_PLUS_FREQ, 2, 3,  6, 15 },
>> +	{ I2C_MAX_STANDARD_MODE_FREQ, 8, 14, 18, 38 },
>> +	{ I2C_MAX_FAST_MODE_FREQ, 4,  3, 9, 19 },
>> +	{ I2C_MAX_FAST_MODE_PLUS_FREQ, 2, 3, 5, 15 },
> 
> argh!
> 
> Can someone from Qualcomm look at this fix? Mukesh, Viken?
Yes, I am reviewing internally and verifying the same with latest 
Guidance and updates. Let me update on this once reviewed.
> 
> Thanks,
> Andi
> 
>>   	{}
>>   };
>>   
>>
>> ---
>> base-commit: edef457004774e598fc4c1b7d1d4f0bcd9d0bb30
>> change-id: 20250513-i2c-bus-freq-ac46343869a4
>>
>> Best regards,
>> -- 
>> Kathiravan Thirumoorthy <kathiravan.thirumoorthy@oss.qualcomm.com>
>>

Re: [PATCH] i2c: qcom-geni: fix I2C frequency table to achieve accurate bus rates

[Andi Shyti]

Hi Mukesh,

On Fri, May 16, 2025 at 12:23:46AM +0530, Mukesh Kumar Savaliya wrote:
> > >   /* source_clock = 32 MHz */
> > >   static const struct geni_i2c_clk_fld geni_i2c_clk_map_32mhz[] = {
> > > -	{ I2C_MAX_STANDARD_MODE_FREQ, 8, 14, 18, 40 },
> > > -	{ I2C_MAX_FAST_MODE_FREQ, 4,  3, 11, 20 },
> > > -	{ I2C_MAX_FAST_MODE_PLUS_FREQ, 2, 3,  6, 15 },
> > > +	{ I2C_MAX_STANDARD_MODE_FREQ, 8, 14, 18, 38 },
> > > +	{ I2C_MAX_FAST_MODE_FREQ, 4,  3, 9, 19 },
> > > +	{ I2C_MAX_FAST_MODE_PLUS_FREQ, 2, 3, 5, 15 },
> > 
> > argh!
> > 
> > Can someone from Qualcomm look at this fix? Mukesh, Viken?
> Yes, I am reviewing internally and verifying the same with latest Guidance
> and updates. Let me update on this once reviewed.

thanks a lot! It's very much appreciated!

Andi

Re: [PATCH] i2c: qcom-geni: fix I2C frequency table to achieve accurate bus rates

[Mukesh Kumar Savaliya]

Looks good to me, Thanks Kathiravan, Andi !

On 5/16/2025 2:16 PM, Andi Shyti wrote:
> Hi Mukesh,
> 
> On Fri, May 16, 2025 at 12:23:46AM +0530, Mukesh Kumar Savaliya wrote:
>>>>    /* source_clock = 32 MHz */
>>>>    static const struct geni_i2c_clk_fld geni_i2c_clk_map_32mhz[] = {
>>>> -	{ I2C_MAX_STANDARD_MODE_FREQ, 8, 14, 18, 40 },
>>>> -	{ I2C_MAX_FAST_MODE_FREQ, 4,  3, 11, 20 },
>>>> -	{ I2C_MAX_FAST_MODE_PLUS_FREQ, 2, 3,  6, 15 },
>>>> +	{ I2C_MAX_STANDARD_MODE_FREQ, 8, 14, 18, 38 },
>>>> +	{ I2C_MAX_FAST_MODE_FREQ, 4,  3, 9, 19 },
>>>> +	{ I2C_MAX_FAST_MODE_PLUS_FREQ, 2, 3, 5, 15 },
>>>
>>> argh!
>>>
>>> Can someone from Qualcomm look at this fix? Mukesh, Viken?
>> Yes, I am reviewing internally and verifying the same with latest Guidance
>> and updates. Let me update on this once reviewed.
> 
> thanks a lot! It's very much appreciated!
> 
Reviewed-by: Mukesh Kumar Savaliya <quic_msavaliy@quicinc.com>
> Andi

Re: [PATCH] i2c: qcom-geni: fix I2C frequency table to achieve accurate bus rates

[Andi Shyti]

Hi Kathiravan,

On Tue, May 13, 2025 at 04:38:33PM +0530, Kathiravan Thirumoorthy wrote:
> Update the I2C frequency table to match the recommended values
> specified in the I2C hardware programming guide. In the current IPQ5424
> configuration where 32MHz is the source clock, the I2C bus frequencies do
> not meet expectations—for instance, 363KHz is achieved instead of the
> expected 400KHz.
> 
> Cc: stable@kernel.org
> Fixes: 506bb2ab0075 ("i2c: qcom-geni: Support systems with 32MHz serial engine clock")
> Signed-off-by: Kathiravan Thirumoorthy <kathiravan.thirumoorthy@oss.qualcomm.com>

I was sure I did apply this patch, but for some reason it has
completely slipped from my sight.

Now it's applied in i2c/i2c-host.

Thanks,
Andi

Re: [PATCH] i2c: qcom-geni: fix I2C frequency table to achieve accurate bus rates

[Kathiravan Thirumoorthy]

On 7/30/2025 4:12 AM, Andi Shyti wrote:
> Hi Kathiravan,
>
> On Tue, May 13, 2025 at 04:38:33PM +0530, Kathiravan Thirumoorthy wrote:
>> Update the I2C frequency table to match the recommended values
>> specified in the I2C hardware programming guide. In the current IPQ5424
>> configuration where 32MHz is the source clock, the I2C bus frequencies do
>> not meet expectations—for instance, 363KHz is achieved instead of the
>> expected 400KHz.
>>
>> Cc: stable@kernel.org
>> Fixes: 506bb2ab0075 ("i2c: qcom-geni: Support systems with 32MHz serial engine clock")
>> Signed-off-by: Kathiravan Thirumoorthy <kathiravan.thirumoorthy@oss.qualcomm.com>
> I was sure I did apply this patch, but for some reason it has
> completely slipped from my sight.


Thanks Andi! I was about to remind you on this today :)


>
> Now it's applied in i2c/i2c-host.


Hope it will be part of 6.17 PR.


>
> Thanks,
> Andi