Re: [PATCH v4 2/2] Embedded USB Debugger (EUD) driver

[Bryan O'Donoghue <bryan.odonoghue@linaro.org>]

On 03/02/2020 19:35, Bjorn Andersson wrote:
> On Thu 30 Jan 20:43 PST 2020, Avaneesh Kumar Dwivedi wrote:

Hi Avaneesh.

> Please aim for keeping the sort order in this file (ignore QCOM_APR
> which obviously is in the wrong place)
> 
>> +       tristate "QTI Embedded USB Debugger (EUD)"
>> +       depends on ARCH_QCOM

If we persist with the model of EXTCON you should "select EXTCON" here.

>> +       help
>> +         The Embedded USB Debugger (EUD) driver is a driver for the
>> +         control peripheral which waits on events like USB attach/detach
>> +         and charger enable/disable. The control peripheral further helps
>> +         support the USB-based debug and trace capabilities.
>> +         This module enables support for Qualcomm Technologies, Inc.
>> +         Embedded USB Debugger (EUD).

Suggest.

This module enables support for Qualcomm Technologies, Inc.
Embedded USB Debugger (EUD).
The EUD is a control peripheral which reports VBUS attach/detach, 
charger enable/disable and USB-based debug and trace capabilities.


>> + * Copyright (c) 2016-2018, The Linux Foundation. All rights reserved.

2020

>> +
>> +static int enable_eud(struct eud_chip *priv)
>> +{
>> +	int ret;
>> +
>> +	/* write into CSR to enable EUD */
>> +	writel_relaxed(BIT(0), priv->eud_reg_base + EUD_REG_CSR_EUD_EN);
>> +	/* Enable vbus, chgr & safe mode warning interrupts */
>> +	writel_relaxed(EUD_INT_VBUS | EUD_INT_CHGR | EUD_INT_SAFE_MODE,
>> +			priv->eud_reg_base + EUD_REG_INT1_EN_MASK);
>> +
>> +	/* Ensure Register Writes Complete */

So... You are writing a register in an on-chip PMIC. The PMIC is 
responsible for detecting USB ID and supplying VBUS as appropriate.

You then get an interrupt to inform you of the state ?

>> +static ssize_t enable_store(struct device *dev,
>> +				struct device_attribute *attr,
>> +				const char *buf, size_t count)
>> +{
>> +	struct eud_chip *chip = dev_get_drvdata(dev);
>> +	int enable = 0;
> 
> You shouldn't need to initialize this as you're checking the return
> value of sscanf().
> 
>> +	int ret = 0;
>> +
>> +	if (sscanf(buf, "%du", &enable) != 1)
>> +		return -EINVAL;
>> +
>> +	if (enable == EUD_ENABLE_CMD)
>> +		ret = enable_eud(chip);
> 
> If ret is !0 you should probably return that, rather than count...
> 
>> +	else if (enable == EUD_DISABLE_CMD)
>> +		disable_eud(chip);
>> +	if (!ret)
> 
> ...and then you don't need this check, or initialize ret to 0 above.
> 
>> +		chip->enable = enable;
> 
> So if I write 42 to "enable" nothing will change in the hardware, but
> chip->enable will be 42...
> 
>> +	return count;
>> +}

I was just going to comment on usb_connector but, does the above code 
need a synchronization primitive to serialize with the worker and 
interrupt handler ?

>> +static int msm_eud_probe(struct platform_device *pdev)
>> +{
>> +	struct eud_chip *chip;
>> +	struct resource *res;
>> +	int ret;
>> +
>> +	chip = devm_kzalloc(&pdev->dev, sizeof(*chip), GFP_KERNEL);
>> +	if (!chip)
>> +		return -ENOMEM;
>> +
>> +	chip->dev = &pdev->dev;
>> +	platform_set_drvdata(pdev, chip);
>> +
>> +	chip->extcon = devm_extcon_dev_allocate(&pdev->dev, eud_extcon_cable);
> 
> Aren't we moving away from extcon in favor of the usb role switching
> thing?

Yes.

For the VBUS notification you could use

usb-role-switch and model the USB connector as a child-node of the 
dual-role controller.

See:
https://patchwork.kernel.org/cover/11346247/
https://patchwork.kernel.org/patch/11346295/
https://patchwork.kernel.org/patch/11346263/

Avaneesh do you have any kernel code that cares about the charger state ?

What we are suggesting here is dropping extcon and using role-switching 
but, if you have some other code that cares about EXTCON_CHG_USB_SDP 
you'd have to do additional work.

But, if I understood the implication of the code above where you write 
to the PMIC and let it handle VBUS/CHARGER on/off and you are just 
notified of the state change, you should be fine with usb-role-switching.

---
bod