Received-SPF: pass (google.com: domain of linux-kernel+bounces-146674-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.199.223 as permitted sender) client-ip=147.75.199.223;
Message-ID: <564fa534b32f4a6e96da6752f531fc7447ec633d.camel@gmail.com>
Subject: Re: [PATCH v1] arm64: dts: imx8mm: fix missing pgc_vpu_* power
 domain parent
From: Vitor Soares <ivitro@gmail.com>
To: Lucas Stach <l.stach@pengutronix.de>, Rob Herring <robh@kernel.org>, 
 Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>, Conor Dooley
 <conor+dt@kernel.org>, Shawn Guo <shawnguo@kernel.org>,  Sascha Hauer
 <s.hauer@pengutronix.de>, Pengutronix Kernel Team <kernel@pengutronix.de>,
 Fabio Estevam <festevam@gmail.com>, Peng Fan <peng.fan@nxp.com>
Cc: Vitor Soares <vitor.soares@toradex.com>, devicetree@vger.kernel.org, 
	imx@lists.linux.dev, linux-arm-kernel@lists.infradead.org, 
	linux-kernel@vger.kernel.org, stable@vger.kernel.org
Date: Tue, 16 Apr 2024 11:53:26 +0100
In-Reply-To: <c76d98a300a9d65d236d334da62916a7d658ef27.camel@gmail.com>
References: <20240409085802.290439-1-ivitro@gmail.com>
	 <9ce35b9bb5a15891f6bd01bd54b7dc84b3ba4021.camel@pengutronix.de>
	 <e1552a3008a30ef7ed9097b4b80cda23ccb9e840.camel@gmail.com>
	 <fcd6acc268b8642371cf289149b2b1c3e90c7f45.camel@pengutronix.de>
	 <bd4d7198e58bd89b46a4c721546f6975b287a5fc.camel@gmail.com>
	 <c76d98a300a9d65d236d334da62916a7d658ef27.camel@gmail.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.44.4-0ubuntu2 
Precedence: bulk
MIME-Version: 1.0

++ Peng Fan <peng.fan@nxp.com>

Greetings,


On Wed, 2024-04-10 at 12:01 +0100, Vitor Soares wrote:
> Hi Lucas,
>=20
> On Tue, 2024-04-09 at 17:44 +0100, Vitor Soares wrote:
> > On Tue, 2024-04-09 at 16:36 +0200, Lucas Stach wrote:
> > > Am Dienstag, dem 09.04.2024 um 14:22 +0100 schrieb Vitor Soares:
> > > > Hi Lucas,
> > > >=20
> > > > Thanks for your feedback.
> > > >=20
> > > > On Tue, 2024-04-09 at 11:13 +0200, Lucas Stach wrote:
> > > > > Hi Vitor,
> > > > >=20
> > > > > Am Dienstag, dem 09.04.2024 um 09:58 +0100 schrieb Vitor
> > > > > Soares:
> > > > > > From: Vitor Soares <vitor.soares@toradex.com>
> > > > > >=20
> > > > > > The pgc_vpu_* nodes miss the reference to the power domain
> > > > > > parent,
> > > > > > leading the system to hang during the resume.
> > > > > >=20
> > > > > This change is not correct. The vpumix domain is controlled
> > > > > through
> > > > > the
> > > > > imx8mm-vpu-blk-ctrl and must not be directly triggered by the
> > > > > child
> > > > > domains in order to guarantee proper power sequencing.
> > > > >=20
> > > > > If the sequencing is incorrect for resume, it needs to be
> > > > > fixed
> > > > > in
> > > > > the
> > > > > blk-ctrl driver. I'll happily assist if you have any
> > > > > questions
> > > > > about
> > > > > this intricate mix between GPC and blk-ctrl hardware/drivers.
> > > > =C2=A0
> > > > I'm new into the topic, so I tried to follow same approach as
> > > > in
> > > > imx8mp
> > > > DT.
> > > >=20
> > > That's a good hint, the 8MP VPU GPC node additions missed my
> > > radar.
> > > The
> > > direct dependency there between the GPC domains is equally wrong.
> > >=20
> > > > I also checked the imx8mq DT and it only have one domain for
> > > > the
> > > > VPU in the GPC. It seem blk-ctrl also dependes on pgc_vpu_* to
> > > > work
> > > > properly.
> > > >=20
> > > > The blk-ctrl driver hangs on imx8m_blk_ctrl_power_on() when
> > > > access
> > > > the
> > > > ip registers for the soft reset. I tried to power-up the before
> > > > the
> > > > soft reset, but it didn't work.
> > > >=20
> > > The runtime_pm_get_sync() at the start of that function should
> > > ensure
> > > that bus GPC domain aka vpumix is powered up. Can you check if
> > > that
> > > is
> > > happening?
> >=20
> > I checked bc->bus_power_dev->power.runtime_status and it is
> > RPM_ACTIVE.
> >=20
> > Am I looking to on the right thing? It is RPM_ACTIVE event before
> > runtime_pm_get_sync().
>=20
> During the probe I can see that
> bus_power_dev->power.runtime_status =3D RPM_SUSPENDED and then vpumix
> is
> powered up on GPC driver.
>=20
> On resume routine I can't see this flow. bus_power_dev-
> > power.runtime_status =3D RPM_ACTIVE and vpumix end up not being
> > powered-
> up.
>=20
> I checked the suspend flow and the GPC tries to poweroff vpumix.
>=20
>=20

My understanding is that when resuming the 38310000.video-codec, the
vpumix isn't powered up. It happens because runtime_status and
runtime_last_status =3D RPM_ACTIVE.=20

I tried to change blk-ctrl suspend routine to force the runtime_status
=3D RPM_SUSPENDED, but the system ended up hanging on another device.

From the comment in blk-ctrl suspend, we rely on PM_SLEEP code that
iterates over dpm_list for suspend/resume.
I did look at the dpm_list, and it changes the order on every boot.=20

With all the tests, I also found that the system randomly hangs on
dispblk-lcdif suspend. I have confirmed this device is in a different
place in the dpm_list (not sure if it is the root cause).=20
I haven't understood how blk-ctrl ensures the correct order there yet.=20

Taking the following dpm_list excerpt:
idx - device
------------------------------
..                                                                  =20
191 - imx-pgc-domain.7                                               =20
192 - imx-pgc-domain.8                                               =20
193 - imx-pgc-domain.9                                               =20
194 - 38330000.blk-ctrl                                              =20
195 - 38310000.video-codec                                           =20
196 - 38300000.video-codec                                           =20
..
205 - genpd:0:38330000.blk-ctrl
206 - genpd:1:38330000.blk-ctrl
207 - genpd:2:38330000.blk-ctrl
208 - genpd:3:38330000.blk-ctrl
------------------------------

Shouldn't genpd devices be before 38330000.blk-ctrl?
As their power domain is GPC and the blk-ctrl power domain is genpd.

Best regards,
Vitor Soares

>=20
> >=20
> >=20
> > >=20
> > > Regards,
> > > Lucas
> > >=20
> > > > Do you have an idea how we can address this within blk-ctrl?
> > > >=20
> > > > Best regards,
> > > > Vitor
> >=20
>=20