linux-kernel - [PATCH 7/7] media: i2c: dw9719: Fix power on/off sequence

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: <20250817-dw9719-v1-7-426f46c69a5a@apitzsch.eu>
Date: Sun, 17 Aug 2025 19:09:26 +0200
From: André Apitzsch via B4 Relay <devnull+git.apitzsch.eu@...nel.org>
To: Mauro Carvalho Chehab <mchehab@...nel.org>, 
 Rob Herring <robh@...nel.org>, Krzysztof Kozlowski <krzk+dt@...nel.org>, 
 Conor Dooley <conor+dt@...nel.org>, devicetree@...r.kernel.org, 
 Sakari Ailus <sakari.ailus@...ux.intel.com>, 
 Daniel Scally <djrscally@...il.com>
Cc: ~postmarketos/upstreaming@...ts.sr.ht, phone-devel@...r.kernel.org, 
 linux-media@...r.kernel.org, linux-kernel@...r.kernel.org, 
 Val Packett <val@...kett.cool>, 
 André Apitzsch <git@...tzsch.eu>
Subject: [PATCH 7/7] media: i2c: dw9719: Fix power on/off sequence

From: Val Packett <val@...kett.cool>

The "jiggle" code was not actually expecting failure, which it should
because that's what actually happens when the device wasn't already woken
up by the regulator power-on (i.e. in the case of a shared regulator).

Also, do actually enter the internal suspend mode on shutdown, to save
power in the case of a shared regulator.

Also, wait a bit longer (2x tOPR) on waking up, 1x is not enough at least
on the DW9718S as found on the motorola-nora smartphone.

Signed-off-by: Val Packett <val@...kett.cool>
Signed-off-by: André Apitzsch <git@...tzsch.eu>
---
 drivers/media/i2c/dw9719.c | 23 ++++++++++++++++-------
 1 file changed, 16 insertions(+), 7 deletions(-)

diff --git a/drivers/media/i2c/dw9719.c b/drivers/media/i2c/dw9719.c
index 63c7fd4ab70a0e02518252b23b89c45df4ba273d..dd28a0223d6ac980084b1f661bd029ea6b0be503 100644
--- a/drivers/media/i2c/dw9719.c
+++ b/drivers/media/i2c/dw9719.c
@@ -95,12 +95,19 @@ struct dw9719_device {
 
 static int dw9719_power_down(struct dw9719_device *dw9719)
 {
+	u32 reg_pwr = (dw9719->model == DW9718S) ? DW9718S_PD : DW9719_CONTROL;
+
+	/*
+	 * Worth engaging the internal SHUTDOWN mode especially due to the
+	 * regulator being potentially shared with other devices.
+	 */
+	cci_write(dw9719->regmap, reg_pwr, DW9719_SHUTDOWN, NULL);
 	return regulator_disable(dw9719->regulator);
 }
 
 static int dw9719_power_up(struct dw9719_device *dw9719, bool detect)
 {
-	u32 reg_pwr;
+	u32 reg_pwr = (dw9719->model == DW9718S) ? DW9718S_PD : DW9719_CONTROL;
 	u64 val;
 	int ret;
 	int err;
@@ -109,13 +116,15 @@ static int dw9719_power_up(struct dw9719_device *dw9719, bool detect)
 	if (ret)
 		return ret;
 
-	/* Jiggle SCL pin to wake up device */
-	reg_pwr = (dw9719->model == DW9718S) ? DW9718S_PD : DW9719_CONTROL;
-	cci_write(dw9719->regmap, reg_pwr, DW9719_SHUTDOWN, &ret);
-	fsleep(100);
+	/*
+	 * Need 100us to transition from SHUTDOWN to STANDBY.
+	 * Jiggle the SCL pin to wake up the device (even when the regulator
+	 * is shared) and wait double the time to be sure, then retry the write.
+	 */
+	cci_write(dw9719->regmap, reg_pwr, DW9719_STANDBY, &ret);
+	ret = 0; /* the jiggle is expected to fail, don't even log that as error */
+	fsleep(200);
 	cci_write(dw9719->regmap, reg_pwr, DW9719_STANDBY, &ret);
-	/* Need 100us to transit from SHUTDOWN to STANDBY */
-	fsleep(100);
 
 	if (detect) {
 		/* This model does not have an INFO register */

-- 
2.50.1