Re: [PATCH V2 6/8] phy: tegra: xusb: Add support for charger detect

From: Thierry Reding <thierry.reding@gmail.com>
To: Nagarjuna Kristam <nkristam@nvidia.com>
Cc: balbi@kernel.org, gregkh@linuxfoundation.org,
	jonathanh@nvidia.com, mark.rutland@arm.com, robh+dt@kernel.org,
	kishon@ti.com, devicetree@vger.kernel.org,
	linux-tegra@vger.kernel.org, linux-usb@vger.kernel.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH V2 6/8] phy: tegra: xusb: Add support for charger detect
Date: Mon, 4 May 2020 17:50:29 +0200	[thread overview]
Message-ID: <20200504155029.GB614153@ulmo> (raw)
In-Reply-To: <ea0f5906-4681-8b84-a55a-e959ce40aece@nvidia.com>

[-- Attachment #1: Type: text/plain, Size: 2717 bytes --]

On Mon, May 04, 2020 at 02:32:51PM +0530, Nagarjuna Kristam wrote:
> >On 28-04-2020 16:25, Thierry Reding wrote:
> > > On Wed, Apr 15, 2020 at 01:55:06PM +0530, Nagarjuna Kristam wrote:
[...]
> > > diff --git a/drivers/phy/tegra/xusb-tegra-cd.c b/drivers/phy/tegra/xusb-tegra-cd.c
> > > +static void tegra_xusb_padctl_utmi_pad_dcd(struct tegra_xusb_padctl *padctl,
> > > +					      u32 index)
> > > +{
> > > +	u32 value;
> > > +	int dcd_timeout_ms = 0;
> > > +	bool ret = false;
> > > +
> > > +	/* Turn on IDP_SRC */
> > > +	value = padctl_readl(padctl, USB2_BATTERY_CHRG_OTGPADX_CTL0(index));
> > > +	value |= OP_I_SRC_EN;
> > > +	padctl_writel(padctl, value, USB2_BATTERY_CHRG_OTGPADX_CTL0(index));
> > > +
> > > +	/* Turn on D- pull-down resistor */
> > > +	value = padctl_readl(padctl, USB2_BATTERY_CHRG_OTGPADX_CTL1(index));
> > > +	value |= USBON_RPD_OVRD_VAL;
> > > +	padctl_writel(padctl, value, USB2_BATTERY_CHRG_OTGPADX_CTL1(index));
> > > +
> > > +	/* Wait for TDCD_DBNC */
> > > +	usleep_range(10000, 120000);
> >  From the comment this looks like we're waiting for some hardware
> > condition. Can we somehow obtain this rather than implementing a fixed
> > sleep? Especially since the range here is so large.
> > 
> As per data sheet we need to wait for 10 micro seconds as settle time.

Okay, so TDCD_DBNC is a value that comes from a timing diagram in a
datasheet? Seems fine to leave it as-is then. Perhaps add parentheses
and mention which exact datasheet that's from, and perhaps which figure
so that people can more easily reference it. Provided there is a
publicly available datasheet, of course.

Actually, one other thing: If the data sheet says to wait 10 us, why do
you use an upper range of 120 us? Shouldn't a range of 10-20 us be good
enough?

> > > +	/* Wait for TVDPSRC_ON */
> > > +	msleep(40);
> > Again, is this a hardware condition that we can wait on by polling a
> > register?
> > 
> It HW settle time before reading registers.

Again, perhaps link to the datasheet, or alternatively describe in the
comment what this is waiting for. That is, something like:

	/* wait for TVDPSRC_ON (wait for hardware to settle) */

or similar.

> > > +		if (tegra_xusb_padctl_utmi_pad_secondary_charger_detect(padctl,
> > > +									index))
> > > +			chrg_type = CDP_TYPE;
> > > +		else
> > > +			chrg_type = DCP_TYPE;
> > > +	} else {
> > > +		chrg_type = SDP_TYPE;
> > > +	}
> > > +
> > > +	dev_dbg(&port->dev, "charger detected of type %d", chrg_type);
> > Do we have a string representation of this?
> > 
> No String representation available. Shall i add one for wasy reading ?

Yeah, I think that'd be nice.

Thierry

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]