[linux-yocto] [kernel v5.2/standard/xlnx-soc][PATCH 1/2] i2c: cadence: do not clear bus_hold_flag in mrecv

[Quanyang Wang]

Hi Shubhrajyoti,

On 10/23/19 5:23 PM, Shubhrajyoti Datta wrote:
>
>> -----Original Message-----
>> From: Shubhrajyoti Datta
>> Sent: Wednesday, October 23, 2019 2:47 PM
>> To: Quanyang Wang <quanyang.wang at windriver.com>; Michal Simek
>> <michals at xilinx.com>; Bruce <bruce.ashfield at gmail.com>
>> Cc: linux-yocto at yoctoproject.org; git <git at xilinx.com>
>> Subject: RE: [linux-yocto][kernel v5.2/standard/xlnx-soc][PATCH 1/2] i2c:
>> cadence: do not clear bus_hold_flag in mrecv
>>
>> Hi ,
>>
>>> -----Original Message-----
>>> From: Quanyang Wang <quanyang.wang at windriver.com>
>>> Sent: Wednesday, October 23, 2019 2:24 PM
>>> To: Michal Simek <michals at xilinx.com>; Bruce
>>> <bruce.ashfield at gmail.com>; Shubhrajyoti Datta <shubhraj at xilinx.com>
>>> Cc: linux-yocto at yoctoproject.org; git <git at xilinx.com>
>>> Subject: Re: [linux-yocto][kernel v5.2/standard/xlnx-soc][PATCH 1/2] i2c:
>>> cadence: do not clear bus_hold_flag in mrecv
>>>
>>> Hi,
>>>
>>> On 10/14/19 2:07 PM, Michal Simek wrote:
>>>> On 13. 10. 19 15:33, quanyang.wang at windriver.com wrote:
>>>>> From: Quanyang Wang <quanyang.wang at windriver.com>
>>>>>
>>>>> When using i2c_smbus_read_byte_data to read one byte from a slave
>>>>> device, because of the commit d358def70688
>>>>> ("i2c: cadence: Fix the hold bit setting"), the transaction becomes:
>>>>>
>>>>> S Addr Wr [A] Comm [A] P S Addr Rd [A] [Data] NA P
>>>>>                          ^
>>>>>                   CR_HOLD bit as 0
>>>>>
>>>>> This will result that the read operation fails and will read "0xff"
>>>>> from the slave device. In the SMBus protocol stipulates that it
>>>>> must follow that command with a repeated START condition as below:
>>>>>
>>>>> S Addr Wr [A] Comm [A] Sr Addr Rd [A] [Data] NA P
>>>>>
>>>>> So keep CR_HOLD bit to be 1 to make sure that the read operation
>>>>> begins with a Sr (repeated START) but not a STOP followed by START.
>>>>>
>>>>> Signed-off-by: Quanyang Wang <quanyang.wang at windriver.com>
>>>>> ---
>>>>>    drivers/i2c/busses/i2c-cadence.c | 2 --
>>>>>    1 file changed, 2 deletions(-)
>>>>>
>>>>> diff --git a/drivers/i2c/busses/i2c-cadence.c
>>>>> b/drivers/i2c/busses/i2c-cadence.c
>>>>> index 6b011931e090..d9493914be15 100644
>>>>> --- a/drivers/i2c/busses/i2c-cadence.c
>>>>> +++ b/drivers/i2c/busses/i2c-cadence.c
>>>>> @@ -626,8 +626,6 @@ static void cdns_i2c_mrecv(struct cdns_i2c *id)
>>>>>    	 */
>>>>>    	if ((id->recv_count > CDNS_I2C_FIFO_DEPTH)  || id->bus_hold_flag)
>>>>>    		ctrl_reg |= CDNS_I2C_CR_HOLD;
>>>>> -	else
>>>>> -		ctrl_reg = ctrl_reg & ~CDNS_I2C_CR_HOLD;
>>>>>
>>>>>    	cdns_i2c_writereg(ctrl_reg, CDNS_I2C_CR_OFFSET);
>>>>>
>>>>>
>>>> Shubhrajyoti: Can you please comment?
>>> Any comment?
>> I am not finding code segment in the tree I am checking
>>
> With the corrected link
> https://github.com/Xilinx/linux-xlnx/blob/master/drivers/i2c/busses/i2c-cadence.c

This issue is not at SDK, it's at mainline. The commit which crashes i2c 
dirver is at

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=d358def706880defa4c9e87381c5bf086a97d5f9


>> maybe you can try moving to the latest code.
>>>> M