Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Thu, 28 Mar 2019 10:11:33 -0400 (EDT)
From:   Alan Stern <stern@rowland.harvard.edu>
To:     Sekhar Nori <nsekhar@ti.com>
cc:     Bartosz Golaszewski <brgl@bgdev.pl>,
        Kevin Hilman <khilman@kernel.org>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Linux ARM <linux-arm-kernel@lists.infradead.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        <linux-usb@vger.kernel.org>,
        Bartosz Golaszewski <bgolaszewski@baylibre.com>
Subject: Re: [PATCH 0/3] ARM: davinci: ohci-da8xx: model the vbus GPIO as a
 fixed regulator
In-Reply-To: <3d7f7814-6184-96c6-8087-6373a489ebc0@ti.com>
Message-ID: <Pine.LNX.4.44L0.1903281009430.1448-100000@iolanthe.rowland.org>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Thu, 28 Mar 2019, Sekhar Nori wrote:

> >> Can you document why the current solution is not optimal? Is it to make
> >> future device-tree conversion for these boards easier? Or?
> >>
> > 
> > It's sub-optimal from the HW modeling in SW PoV - it is in fact a
> > regulator enabled/disabled by a GPIO. Also: it's code duplication as
> > currently we check if the vbus GPIO exists and then use it or check if
> > the regulator exists and use this as the second choice. The third
> > patch actually shrinks the driver.
> 
> I see now that the driver supports controlling the VBUS gpio as
> regulator already. Something I should have caught in review last time
> around.
> 
> I agree this patch is an improvement. Lets see what Alan feels.

I'm not an expert on this stuff, but the patch looks reasonable.  
However, I do wish that in the devm_request_threaded_irq() call, the 
indentation of the continuation lines was left unchanged.

Alan Stern

> Also, reg_enabled member of da8xx_ohci_hcd structure seems to be pretty
> useless considering regulator API already has use counting. Can you take
> a look and remove that too as an added bonus.
> 
> Thanks,
> Sekhar