Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:7 as permitted sender) client-ip=2620:137:e000::3:7;
Date:   Wed, 01 Nov 2023 08:42:41 +0000
Message-ID: <87r0l96f0e.wl-maz@kernel.org>
From:   Marc Zyngier <maz@kernel.org>
To:     Thomas Richard <thomas.richard@bootlin.com>
Cc:     Aswath Govindraju <a-govindraju@ti.com>,
        Vignesh Raghavendra <vigneshr@ti.com>,
        Nishanth Menon <nm@ti.com>, Tero Kristo <kristo@kernel.org>,
        Santosh Shilimkar <ssantosh@kernel.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org,
        Gregory CLEMENT <gregory.clement@bootlin.com>
Subject: Re: [PATCH] irqchip/ti-sci-intr: Add support for system suspend/resume PM
In-Reply-To: <184d23fc-bb62-473c-8400-cda160aa44f3@bootlin.com>
References: <20220607061912.12222-1-a-govindraju@ti.com>
        <87sfohyma5.wl-maz@kernel.org>
        <0733f59a-497a-351b-0e97-26a2875a5352@ti.com>
        <871qvzr01u.wl-maz@kernel.org>
        <184d23fc-bb62-473c-8400-cda160aa44f3@bootlin.com>
User-Agent: Wanderlust/2.15.9 (Almost Unreal) SEMI-EPG/1.14.7 (Harue)
 FLIM-LB/1.14.9 (=?UTF-8?B?R29qxY0=?=) APEL-LB/10.8 EasyPG/1.0.0 Emacs/28.2
 (x86_64-pc-linux-gnu) MULE/6.0 (HANACHIRUSATO)
MIME-Version: 1.0 (generated by SEMI-EPG 1.14.7 - "Harue")
Content-Type: text/plain; charset=US-ASCII
Precedence: bulk

On Tue, 31 Oct 2023 16:00:59 +0000,
Thomas Richard <thomas.richard@bootlin.com> wrote:
> 
> On 6/8/22 11:12, Marc Zyngier wrote:
> > On Wed, 08 Jun 2022 09:48:20 +0100,
> > Aswath Govindraju <a-govindraju@ti.com> wrote:
> >>
> >> As, the resource allocator does not have enough memory to save and
> >> restore all the mappings corresponding various resources, this is being
> >> done on the requester or consumer side.
> > 
> > You're missing the point: the ti_sci_resource structure is managed by
> > this resource allocator, and it isn't exactly rocket science to add
> > the required context to it, and then get it to restore that context on
> > resume.
> 
> Hi Marc,
> 
> I'm interested to continue the work to upstream this patch.
> I read all the thread, but I think I'm also missing the point.
> 
> Do you mean the ti_sci_intr_irq_domain struct is not the right place to
> save the mappings, and it shall not be done in this irqchip driver?
> The context should be saved and restored from the ti_sci driver using
> the ti_sci_resource struct?

That's exactly my point. There is already a large body of code that is
aware of the resource allocation (the ti_sci driver), which is common
to a whole lot of subsystem that will need some form of save/restore.

It seems natural to put the save/restore aspects inside that driver,
unless there is some other reasons not to do so (which haven't so far
been explained).

Thanks,

	M.

-- 
Without deviation from the norm, progress is not possible.