MIME-Version: 1.0
References: <20211213103243.33657-1-heikki.krogerus@linux.intel.com>
 <20211213103243.33657-2-heikki.krogerus@linux.intel.com> <CAJZ5v0jq5ve9+V3k5Vi=vJAheSyQbAsNmi6An9=oKxkyVed5bw@mail.gmail.com>
 <Ybnoy9e6VvFsJ4b1@kuha.fi.intel.com>
In-Reply-To: <Ybnoy9e6VvFsJ4b1@kuha.fi.intel.com>
From:   "Rafael J. Wysocki" <rafael@kernel.org>
Date:   Wed, 15 Dec 2021 15:52:12 +0100
Message-ID: <CAJZ5v0hiv18Ri5Vi4Y07bXRO9+Qf5bZ+TwmqwRG3spr9nzV0MA@mail.gmail.com>
Subject: Re: [PATCH v2 1/5] acpi: Store the Physical Location of Device (_PLD) information
To:     Heikki Krogerus <heikki.krogerus@linux.intel.com>
Cc:     "Rafael J. Wysocki" <rafael@kernel.org>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Andy Shevchenko <andriy.shevchenko@linux.intel.com>,
        Sakari Ailus <sakari.ailus@linux.intel.com>,
        Prashant Malani <pmalani@chromium.org>,
        ACPI Devel Maling List <linux-acpi@vger.kernel.org>,
        "open list:ULTRA-WIDEBAND (UWB) SUBSYSTEM:" 
        <linux-usb@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Wed, Dec 15, 2021 at 2:08 PM Heikki Krogerus
<heikki.krogerus@linux.intel.com> wrote:
>
> Hi Rafael,
>
> On Wed, Dec 15, 2021 at 12:43:48PM +0100, Rafael J. Wysocki wrote:
> > One overall problem I see here is that it potentially stores a bunch
> > of _PLD buffers that will never be used, which is a waste of memory.
> >
> > It may be better to make acpi_device_get_location() evaluate _PLD for
> > the target device if it cannot be found in the list and either update
> > one of the existing entries if the crc matches and return the pld data
> > from there, or add a new item to the list and return its pld data.
>
> The primary goal of this series is actually to see the other devices
> that share a fixed location with our device.

It would be good to mention that in the changelog then.

> The _PLD buffer cache
> would be nice, but not necessary. We don't have to store the _PLD
> buffer itself as part of the location structure at all if that's a
> problem.

It may be cached on the first invocation of acpi_device_get_location()
too, but I'm not sure how much overhead would be avoided by doing
that.  I guess it would make most sense in the shared _PLD case.

> I'll make the change.

OK