Pankaj,
On 8/30/23 02:28, Pankaj Pandey wrote:
Hi Guenter,
Thanks for the pointer. Actually during bmr454 probing during kernel boots-up it
called system reset/crash. For reference please find bmr454 probing log below:
First of all, please don't top-post.
[ 14.910630] pmbus 7-0010: bmr454 BMR status word : 0x2
[ 14.922945] pmbus 7-0010: bmr454 BMR status cml : 0x2
Bit 1 of the CML status register suggests that there is a communication
problem. Per PMBus standard:
Bit 1 - A communication fault other than the ones listed in this table has occurred
---> Immediately after above bmr454 probing it reset/crash the system
and jump to u-boot.
I don't think anything _called_ system reset/crash. There is no code
in the driver that would do that.
Using debug prints I observed system got resettted after reaching to
_pmbus_write_byte function. But
sometimes probing is fine, not observing any system reset/crash.
My best guess is that the i2c communication with the chip is unstable,
and that the write causes it to do something that causes the reset.
How it would do that - I don't know. Maybe one of the module pins is wired
to the system's reset pin (though I can not imagine why that would be
the case), or the miscommunication causes a chip reset. Either case,
maybe the problem is in the I2C bus speed. If you can, I would suggest
to lower it to 100 KHz to see if that makes a difference.
Guenter
Regards,
Pankaj Pandey
On Wed, Aug 30, 2023 at 7:38 AM Guenter Roeck <linux@xxxxxxxxxxxx> wrote:
On Wed, Aug 30, 2023 at 07:16:46AM +0530, Pankaj Pandey wrote:
---------- Forwarded message ---------
From: Pankaj Pandey <pankaj.embedded@xxxxxxxxx>
Date: Wed, Aug 30, 2023 at 4:42 AM
Subject: Regarding pmbus_core.c
To: <linux@xxxxxxxxxxxx>
Hi Guenter Roeck,
I am facing an issue at drivers/hwmon/pmbus/pmbus_core.c of linux-5.10 kernel.
I am using the bm454 driver from drivers/hwmon/pmbus/pmbus.c.
Observed sometimes my Powepc based target board called for reset from
inside int pmbus_write_byte_data(struct i2c_client *client, int page,
u8 reg, u8 value) function. It is not consistent, sometimes not
Sorry, I don't understand what you mean with "called for reset from
inside int pmbus_write_byte_data". Please clarify.
If I recall correctly, bmr454 is very picky when it comes to i2c bus
speed. If it acts up, you might have some success with picking a different
bus speed.
Thanks,
Guenter
observed any reset and bmr454 initialized fine. Please suggest any
pointers to debug this issue.
Regards,
Pankaj Pandey
