[<prev] [next>] [day] [month] [year] [list]
Message-ID: <0a3780a5073324772b873b8c474358c5@codeaurora.org>
Date: Fri, 02 Sep 2016 16:58:12 +0530
From: vivek.gautam@...eaurora.org
To: Stephen Boyd <stephen.boyd@...aro.org>
Cc: Linux USB Mailing List <linux-usb@...r.kernel.org>,
linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org,
linux-arm-msm@...r.kernel.org, Andy Gross <andy.gross@...aro.org>,
Bjorn Andersson <bjorn.andersson@...aro.org>,
Neil Armstrong <narmstrong@...libre.com>,
Arnd Bergmann <arnd@...db.de>, Felipe Balbi <balbi@...nel.org>,
Peter Chen <peter.chen@....com>,
Kishon Vijay Abraham I <kishon@...com>,
devicetree@...r.kernel.org
Subject: Re: [PATCH v3 21/22] phy: Add support for Qualcomm's USB HSIC phy
On 2016-09-02 03:30, Stephen Boyd wrote:
> (Please trim replies)
sorry, will take care from next time.
>
> Quoting Vivek Gautam (2016-08-31 23:17:55)
>> On Thu, Sep 1, 2016 at 6:10 AM, Stephen Boyd <stephen.boyd@...aro.org>
>> wrote:
>> > +
>> > + uphy->cal_sleep_clk = clk = devm_clk_get(&ulpi->dev, "cal_sleep");
>> > + if (IS_ERR(clk))
>> > + return PTR_ERR(clk);
>> > +
>> > + uphy->phy = devm_phy_create(&ulpi->dev, ulpi->dev.of_node,
>> > + &qcom_usb_hsic_phy_ops);
>>
>> There's a ulpi_phy library available in drivers/phy/. Do we want to
>> use that ?
>> That also creates a phy-lookup of this PHY so that the ulpi device's
>> parent
>> can request the PHY.
>
> I don't have any interest in using those two functions (does two
> functions constitute a library?).
Not really.
> There's no devm as you say,
I meant to say that you may need to change the existing APIs to the
devm_* APIs.
> and it seems to be specific to the ULPI hardware for dwc3 (the only
> user) where
> the phy is called "usb2-phy".
This was used with TI's USB 2.0 PHY, that has ULPI interface.
> This is a phy for the ChipIdea controller
> which only has one phy and it's called "usb-phy" in that case.
In a way this is also a USB 2.0 phy, isn't it ?
>
>>
>> You may want to modify the APIs available in ulpi_phy library to use
>> the
>> devm_* APIs.
>>
>
> The lookup isn't necessary because we use DT to find the lookup. I seem
> to recall the phy framework requiring a DT lookup too.
The lookup created in this ulpi_phy.h was used to help getting the PHY
in the
driver for parent device (a core wrapper) of the controller (platform
glue layer) that requests this PHY.
I am not certain at this point about how PHY has to be handled in case
of Chipidea.
Was just throwing in ideas. :-)
Thanks
Vivek
Powered by blists - more mailing lists