linux-kernel - [PATCH v2 2/2] firmware: arm_scmi: power_control: Set SCMI_SYSPOWER

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <20250704-scmi-pm-v2-2-9316cec2f9cc@nxp.com>
Date: Fri, 04 Jul 2025 11:09:36 +0800
From: Peng Fan <peng.fan@....com>
To: Sudeep Holla <sudeep.holla@....com>, 
 Cristian Marussi <cristian.marussi@....com>
Cc: arm-scmi@...r.kernel.org, linux-arm-kernel@...ts.infradead.org, 
 linux-kernel@...r.kernel.org, 
 Ranjani Vaidyanathan <ranjani.vaidyanathan@....com>, 
 Chuck Cannon <chuck.cannon@....com>, Peng Fan <peng.fan@....com>
Subject: [PATCH v2 2/2] firmware: arm_scmi: power_control: Set
 SCMI_SYSPOWER_IDLE in PM resume

When the Linux SCMI agent is suspended and another agent sends a second
suspend message, the Linux agent wakes up. The expected behavior is for
the Linux agent to wake up and then suspend again.

However, due to a race condition, the second suspend notification wakes
the system, but the SCMI state is not yet set to SCMI_SYSPOWER_IDLE.
This is because interrupts (e.g., i.MX95 Message Unit interrupts, which are
always enabled as wakeup sources) are active before thaw_processes() is
called. As a result, scmi_userspace_notifier() runs too early and
ignores the suspend request since the state hasn't been updated yet.

This patch addresses the race by setting the SCMI state to
SCMI_SYSPOWER_IDLE earlier-during the device resume phase, before
thaw_processes() is invoked. This ensures that when the notifier is
triggered, the state check passes and the system can suspend again as
intended.

On i.MX95, there are two scenarios where the Cortex-A cluster issues a
WFI signal:
- Linux CPU idle
- Linux suspend
The hardware state machine handles power-off sequences, with SCP involved
in one step. However, SCP cannot distinguish between a CPU idle request
and a full suspend request-it only sees a generic cluster power-off.

As a result, the SCP cannot filter out the second suspend message. By
setting SCMI_SYSPOWER_IDLE early, we ensure the Linux agent can correctly
handle the second suspend request and re-enter suspend.

Signed-off-by: Peng Fan <peng.fan@....com>
---
 drivers/firmware/arm_scmi/scmi_power_control.c | 24 +++++++++++++++++++-----
 1 file changed, 19 insertions(+), 5 deletions(-)

diff --git a/drivers/firmware/arm_scmi/scmi_power_control.c b/drivers/firmware/arm_scmi/scmi_power_control.c
index 21f467a92942883be66074c37c2cab08c3e8a5cc..d2cfd9d92e711f7247a13c7773c11c0a6e582876 100644
--- a/drivers/firmware/arm_scmi/scmi_power_control.c
+++ b/drivers/firmware/arm_scmi/scmi_power_control.c
@@ -46,6 +46,7 @@
 #include <linux/math.h>
 #include <linux/module.h>
 #include <linux/mutex.h>
+#include <linux/pm.h>
 #include <linux/printk.h>
 #include <linux/reboot.h>
 #include <linux/scmi_protocol.h>
@@ -324,12 +325,7 @@ static int scmi_userspace_notifier(struct notifier_block *nb,
 
 static void scmi_suspend_work_func(struct work_struct *work)
 {
-	struct scmi_syspower_conf *sc =
-		container_of(work, struct scmi_syspower_conf, suspend_work);
-
 	pm_suspend(PM_SUSPEND_MEM);
-
-	sc->state = SCMI_SYSPOWER_IDLE;
 }
 
 static int scmi_syspower_probe(struct scmi_device *sdev)
@@ -354,6 +350,7 @@ static int scmi_syspower_probe(struct scmi_device *sdev)
 	sc->required_transition = SCMI_SYSTEM_MAX;
 	sc->userspace_nb.notifier_call = &scmi_userspace_notifier;
 	sc->dev = &sdev->dev;
+	dev_set_drvdata(&sdev->dev, sc);
 
 	INIT_WORK(&sc->suspend_work, scmi_suspend_work_func);
 
@@ -363,6 +360,20 @@ static int scmi_syspower_probe(struct scmi_device *sdev)
 						       NULL, &sc->userspace_nb);
 }
 
+static int scmi_system_power_resume(struct device *dev)
+{
+
+	struct scmi_syspower_conf *sc = dev_get_drvdata(dev);
+
+	sc->state = SCMI_SYSPOWER_IDLE;
+
+	return 0;
+}
+
+static const struct dev_pm_ops scmi_system_power_pmops = {
+	SET_SYSTEM_SLEEP_PM_OPS(NULL, scmi_system_power_resume)
+};
+
 static const struct scmi_device_id scmi_id_table[] = {
 	{ SCMI_PROTOCOL_SYSTEM, "syspower" },
 	{ },
@@ -370,6 +381,9 @@ static const struct scmi_device_id scmi_id_table[] = {
 MODULE_DEVICE_TABLE(scmi, scmi_id_table);
 
 static struct scmi_driver scmi_system_power_driver = {
+	.driver	= {
+		.pm = &scmi_system_power_pmops,
+	},
 	.name = "scmi-system-power",
 	.probe = scmi_syspower_probe,
 	.id_table = scmi_id_table,

-- 
2.37.1