Re: [PATCH v3] platform/chrome: cros_ec_spi: Transfer messages at high priority

From: Enric Balletbo i Serra <enric.balletbo@collabora.com>
To: Matthias Kaehlcke <mka@chromium.org>,
	Doug Anderson <dianders@chromium.org>
Cc: "Benson Leung" <bleung@chromium.org>,
	"Alexandru M Stan" <amstan@chromium.org>,
	"open list:ARM/Rockchip SoC..."
	<linux-rockchip@lists.infradead.org>,
	"Simon Glass" <sjg@chromium.org>,
	"Brian Norris" <briannorris@chromium.org>,
	"Guenter Roeck" <groeck@chromium.org>,
	"Mark Brown" <broonie@kernel.org>,
	"Ryan Case" <ryandcase@chromium.org>,
	"Randall Spangler" <rspangler@chromium.org>,
	"Heiko Stübner" <heiko@sntech.de>,
	LKML <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v3] platform/chrome: cros_ec_spi: Transfer messages at high priority
Date: Thu, 4 Apr 2019 17:00:36 +0200	[thread overview]
Message-ID: <b54ef6c0-512a-19ff-e4f3-6963dfe9e8bd@collabora.com> (raw)
In-Reply-To: <20190403211943.GT112750@google.com>

Hi Doug,

Thanks for sending this patch upstream.

On 3/4/19 23:19, Matthias Kaehlcke wrote:
> On Wed, Apr 03, 2019 at 02:08:40PM -0700, Doug Anderson wrote:
>> Hi,
>>
>> On Wed, Apr 3, 2019 at 2:04 PM Matthias Kaehlcke <mka@chromium.org> wrote:
>>>> +static int cros_ec_xfer_high_pri(struct cros_ec_device *ec_dev,
>>>
>>> nit: the fact that a high priority workqueue is used is an
>>> implementation detail, since the driver has no function to perform a
>>> transfer with 'normal'/low priority there is no need to distinguish
>>> between the two cases. In this sense I'd be inclined to remove the
>>> 'high_pri' from the function names.
>>>
>>> Sorry for not mentioning this earlier, I focussed on other
>>> details, anyway it's just a nit.
>>
>> I still kinda like having the "high_pri" in there since the point of
>> this function is to transfer the work onto the high priority
>> workqueue.  It's not an exported function so having the implementation
>> detail leak into the name isn't a bad thing, is it?
> 
> IMO the long name with details mostly irrelevant to the caller (they
> want to do a 'normal' transfer, the function should do the right thing
> to get that done) is more distracting than helpful. But yeah, this is
> definitely 'nit/bikeshed' territory ;-)
> 
>> ...so unless someone else thinks the name should change or you feel
>> strongly about it I won't plan to change the name.
> 
> no strong feelings on my side, just wanted to mention it.
> 

Tested on veyron-jaq, veyron-minnie and peach-pi (well with peach-pi I didn't
reproduce the issue but at least I know doesn't breaks anything)

I'll add this patch to the chrome-platform for-next branch for the auto-builders
to play with and launch some few more automated tests. If all goes well I'll
queue this patch to chrome-platform-5.2.

Thanks,
 Enric