Re: [RFC] drm: implement generic firmware eviction

From: Hans de Goede <hdegoede@redhat.com>
To: Maxime Ripard <maxime.ripard@free-electrons.com>
Cc: devicetree <devicetree@vger.kernel.org>,
	Daniel Vetter <daniel.vetter@ffwll.ch>,
	Rob Herring <robh+dt@kernel.org>,
	"dri-devel@lists.freedesktop.org"
	<dri-devel@lists.freedesktop.org>
Subject: Re: [RFC] drm: implement generic firmware eviction
Date: Fri, 26 Aug 2016 11:02:17 +0200	[thread overview]
Message-ID: <04f9cb51-94f3-a26d-4966-385175340dd7@redhat.com> (raw)
In-Reply-To: <20160826085823.GB30441@lukather>

Hi,

On 26-08-16 10:58, Maxime Ripard wrote:
> Hi,
>
> On Fri, Aug 26, 2016 at 10:43:55AM +0200, Hans de Goede wrote:
>>>> I'm not sure we would want to remove the device at all, we
>>>> certainly should not be removing the dt_node from the devicetree
>>>> IMHO. Having that around to see how the bootloader set things up
>>>> is really useful for debugging and normally we should never modify
>>>> the devicetree as set up by the bootloader.
>>>>
>>>> Why not just unbind the driver from the platform device? That
>>>> should be enough.
>>>
>>> That will leave IORESOURCE_MEM around, causing conflicts if
>>> re-used/claimed by other devices/drivers. Furthermore, it is really
>>> fragile leaving the device around, without any control over
>>> possible future driver probing.
>>
>> Ah, good point. On ARM this currently typically is reserved by the bootloader
>> so never touched by the kernel at all, not even when the simplefb is no longer
>> used, actually returning this memory to the kernel after unbinding the simplefb /
>> destroying the simplefb platform-dev would be really good to do. We should
>> probably figure out how that should be done before getting rid of
>> remove_conflicting_framebuffers... (sorry).
>
> That would be rather easy to do. The firmware could generate a
> reserved-memory node instead of passing a smaller memory size to the
> kernel. That way, the kernel will know that it's actual ram that it
> can reclaim.

So when would the kernel reclaim the RAM then? We do not want it
to reclaim before loading the simplefb / simpledrm driver.

And the real drm driver will likely be a module, so if the kernel
reclaims just before executing /sbin/init that would be too
early, unless we can somehow manually make it do another reclaim
pass when the simplefb gets disabled ...

Regards,

Hans
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel