Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Martin Kepplinger <martin.kepplinger@puri.sm>
To:     rafael@kernel.org, khilman@kernel.org, ulf.hansson@linaro.org,
        robh@kernel.org, krzysztof.kozlowski@linaro.org,
        shawnguo@kernel.org, s.hauer@pengutronix.de, festevam@gmail.com,
        pavel@ucw.cz
Cc:     kernel@puri.sm, linux-imx@nxp.com, broonie@kernel.org,
        l.stach@pengutronix.de, aford173@gmail.com,
        linux-pm@vger.kernel.org, devicetree@vger.kernel.org,
        linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org,
        Martin Kepplinger <martin.kepplinger@puri.sm>
Subject: [PATCH v6 0/2] power: domain: handle power supplies that need interrupts
Date:   Tue, 26 Jul 2022 10:32:55 +0200
Message-Id: <20220726083257.1730630-1-martin.kepplinger@puri.sm>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

hi Ulf, Lucas and all interested,

This makes available a new genpd flag GENPD_FLAG_IRQ_ON in a relatively
generic way:
genpd providers can set it when irqs are needed to manage power on/off.
Since the main goal here has been to fix systemd suspend/resume,
adjusting these callbacks is all that's being done when this flag gets set.

And since I'm working on imx8mq, the 2nd patch makes gpcv2 set this new
flag when a power domain has a power-supply described in DT.
For i.MX8M* platforms, this should be ok. For other platforms this might
be useful too but needs to be tested.


revision history
----------------
v6: (thank you Ulf and Lucas)
* drop the cleanup patch to use BIT()
* use of_property_read_bool() to find the power-supply
* clean up the callback assignment
* remove the wrong error printing
  (flag-check can be added later to *drivers* that have noirq callbacks)

v5: (thank you Lucas)
* simplify gpcv2 code: just set GENPD_FLAG_IRQ_ON when a power-supply is present
https://lore.kernel.org/linux-arm-kernel/20220721043608.1527686-1-martin.kepplinger@puri.sm/

v4: (thank you Ulf and Lucas)
* split up genpd core and gpcv2 changes
* set callbacks inside of pm_genpd_init()
* make flag name and description a bit more generic
* print an error in __genpd_dev_pm_attach() if there a "mismatch"
https://lore.kernel.org/linux-arm-kernel/20220720043444.1289952-1-martin.kepplinger@puri.sm/T/#t

v3: (thank you Ulf)
* move DT parsing to gpcv2 and create a genpd flag that gets set
https://lore.kernel.org/linux-arm-kernel/20220718210302.674897-1-martin.kepplinger@puri.sm/

v2: (thank you Krzysztof)
* rewrite: find possible regulators' interrupts property in parents
  instead of inventing a new property.
https://lore.kernel.org/linux-arm-kernel/20220712121832.3659769-1-martin.kepplinger@puri.sm/

v1: (initial idea)
https://lore.kernel.org/linux-arm-kernel/20220711094549.3445566-1-martin.kepplinger@puri.sm/T/#t


Martin Kepplinger (2):
  power: domain: handle genpd correctly when needing interrupts
  soc: imx: gpcv2: fix suspend/resume by setting GENPD_FLAG_IRQ_ON

 drivers/base/power/domain.c | 13 +++++++++++--
 drivers/soc/imx/gpcv2.c     |  3 +++
 include/linux/pm_domain.h   |  5 +++++
 3 files changed, 19 insertions(+), 2 deletions(-)

-- 
2.30.2