It's not rare that device drivers need delayed work.
It's not rare that this work needs driver's data.
Often this means that driver must ensure the work is not queued when
driver is detached. Often it is done by ensuring new work is not added and
then calling cancel_delayed_work_sync() at remove(). In many cases this
may also require cleanup at probe error path - which is easy to forget.
Also the "by ensuring new work is not added" has a gotcha.
It is not strange to see devm managed IRQs scheduling (delayed) work.
Mixing this with manua wq clean-up is hard to do correctly because the
devm is likely to free the IRQ only after the remove() is ran. So manual
wq cancellation and devm-based IRQ management do not mix well - there is
a short(?) time-window after the wq clean-up when IRQs are still not
freed and may schedule new work.
When both WQs and IRQs are managed by devm things are likely to just
work. WQs should be initialized before IRQs (when IRQs need to schedule
work) and devm unwinds things in "FILO" order.
This series implements delayed wq cancellation on top of devm and replaces
the obvious cases where only thing remove call-back in a driver does is
cancelling the work. There might be other cases where we could switch
more than just work cancellation to use managed version and thus get rid
of remove or mixed (manual and devm) resource management.
The series introduces include/linux/devm-helpers.h file which
hopefully works as a place where this kind of helpers can be inlined.
Please see previous discussion here:
RFC v1:
https://lore.kernel.org/lkml/[email protected]/
Changelog v3:
- Dropped RFC as advieced by Greg.
- No functional changes.
Changelog RFC v2 resend:
- rebased on 5.12-rc4
Changelog RFC v2:
- used correct terminology ("driver detach" instead of "exit, ...")
- inlined the devm_delayed_work_autocancel() in a header
- added Hans as a maintainer for the new header + myself as a reviewer
- used devm_add_action() instead of using plain devres_add()
---
Matti Vaittinen (8):
workqueue: Add resource managed version of delayed work init
MAINTAINERS: Add entry for devm helpers
extconn: Clean-up few drivers by using managed work init
hwmon: raspberry-pi: Clean-up few drivers by using managed work init
platform/x86: gpd pocket fan: Clean-up by using managed work init
power: supply: Clean-up few drivers by using managed work init
regulator: qcom_spmi-regulator: Clean-up by using managed work init
watchdog: retu_wdt: Clean-up by using managed work init
MAINTAINERS | 6 +++
drivers/extcon/extcon-gpio.c | 15 ++----
drivers/extcon/extcon-intel-int3496.c | 16 ++----
drivers/extcon/extcon-palmas.c | 17 +++----
drivers/extcon/extcon-qcom-spmi-misc.c | 17 +++----
drivers/hwmon/raspberrypi-hwmon.c | 17 +++----
drivers/platform/x86/gpd-pocket-fan.c | 17 +++----
drivers/power/supply/axp20x_usb_power.c | 15 ++----
drivers/power/supply/bq24735-charger.c | 18 +++----
drivers/power/supply/ltc2941-battery-gauge.c | 20 +++-----
drivers/power/supply/sbs-battery.c | 16 ++----
drivers/regulator/qcom_spmi-regulator.c | 34 +++----------
drivers/watchdog/retu_wdt.c | 22 +++-----
include/linux/devm-helpers.h | 53 ++++++++++++++++++++
14 files changed, 128 insertions(+), 155 deletions(-)
create mode 100644 include/linux/devm-helpers.h
base-commit: 0d02ec6b3136c73c09e7859f0d0e4e2c4c07b49b
--
2.25.4
--
Matti Vaittinen, Linux device drivers
ROHM Semiconductors, Finland SWDC
Kiviharjunlenkki 1E
90220 OULU
FINLAND
~~~ "I don't think so," said Rene Descartes. Just then he vanished ~~~
Simon says - in Latin please.
~~~ "non cogito me" dixit Rene Descarte, deinde evanescavit ~~~
Thanks to Simon Glass for the translation =]
Devm helper header containing small inline helpers was added.
Hans promised to maintain it.
Add Hans as maintainer and myself as designated reviewer.
Signed-off-by: Matti Vaittinen <[email protected]>
---
Changelog from RFCv2:
- RFC dropped. No functional changes.
MAINTAINERS | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 9e876927c60d..fa5ac3164678 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -5169,6 +5169,12 @@ M: Torben Mathiasen <[email protected]>
S: Maintained
W: http://lanana.org/docs/device-list/index.html
+DEVICE RESOURCE MANAGEMENT HELPERS
+M: Hans de Goede <[email protected]>
+R: Matti Vaittinen <[email protected]>
+S: Maintained
+F: include/linux/devm-helpers.h
+
DEVICE-MAPPER (LVM)
M: Alasdair Kergon <[email protected]>
M: Mike Snitzer <[email protected]>
--
2.25.4
--
Matti Vaittinen, Linux device drivers
ROHM Semiconductors, Finland SWDC
Kiviharjunlenkki 1E
90220 OULU
FINLAND
~~~ "I don't think so," said Rene Descartes. Just then he vanished ~~~
Simon says - in Latin please.
~~~ "non cogito me" dixit Rene Descarte, deinde evanescavit ~~~
Thanks to Simon Glass for the translation =]
Few drivers implement remove call-back only for ensuring a delayed
work gets cancelled prior driver removal. Clean-up these by switching
to use devm_delayed_work_autocancel() instead.
This change is compile-tested only. All testing is appreciated.
Signed-off-by: Matti Vaittinen <[email protected]>
Acked-by: Sebastian Reichel <[email protected]>
---
Changelog from RFCv2:
- RFC dropped. No functional changes.
drivers/power/supply/axp20x_usb_power.c | 15 +++++----------
drivers/power/supply/bq24735-charger.c | 18 ++++++------------
drivers/power/supply/ltc2941-battery-gauge.c | 20 +++++++-------------
drivers/power/supply/sbs-battery.c | 16 +++++-----------
4 files changed, 23 insertions(+), 46 deletions(-)
diff --git a/drivers/power/supply/axp20x_usb_power.c b/drivers/power/supply/axp20x_usb_power.c
index 8933ae26c3d6..4259709e3491 100644
--- a/drivers/power/supply/axp20x_usb_power.c
+++ b/drivers/power/supply/axp20x_usb_power.c
@@ -8,6 +8,7 @@
#include <linux/bitops.h>
#include <linux/device.h>
+#include <linux/devm-helpers.h>
#include <linux/init.h>
#include <linux/interrupt.h>
#include <linux/kernel.h>
@@ -646,21 +647,16 @@ static int axp20x_usb_power_probe(struct platform_device *pdev)
}
}
+ ret = devm_delayed_work_autocancel(&pdev->dev, &power->vbus_detect,
+ axp20x_usb_power_poll_vbus);
+ if (ret)
+ return ret;
if (axp20x_usb_vbus_needs_polling(power))
queue_delayed_work(system_power_efficient_wq, &power->vbus_detect, 0);
return 0;
}
-static int axp20x_usb_power_remove(struct platform_device *pdev)
-{
- struct axp20x_usb_power *power = platform_get_drvdata(pdev);
-
- cancel_delayed_work_sync(&power->vbus_detect);
-
- return 0;
-}
-
static const struct of_device_id axp20x_usb_power_match[] = {
{
.compatible = "x-powers,axp202-usb-power-supply",
@@ -680,7 +676,6 @@ MODULE_DEVICE_TABLE(of, axp20x_usb_power_match);
static struct platform_driver axp20x_usb_power_driver = {
.probe = axp20x_usb_power_probe,
- .remove = axp20x_usb_power_remove,
.driver = {
.name = DRVNAME,
.of_match_table = axp20x_usb_power_match,
diff --git a/drivers/power/supply/bq24735-charger.c b/drivers/power/supply/bq24735-charger.c
index ab2f4bf8f603..b5d619db79f6 100644
--- a/drivers/power/supply/bq24735-charger.c
+++ b/drivers/power/supply/bq24735-charger.c
@@ -17,6 +17,7 @@
* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
*/
+#include <linux/devm-helpers.h>
#include <linux/err.h>
#include <linux/i2c.h>
#include <linux/init.h>
@@ -473,7 +474,11 @@ static int bq24735_charger_probe(struct i2c_client *client,
if (!charger->poll_interval)
return 0;
- INIT_DELAYED_WORK(&charger->poll, bq24735_poll);
+ ret = devm_delayed_work_autocancel(&client->dev, &charger->poll,
+ bq24735_poll);
+ if (ret)
+ return ret;
+
schedule_delayed_work(&charger->poll,
msecs_to_jiffies(charger->poll_interval));
}
@@ -481,16 +486,6 @@ static int bq24735_charger_probe(struct i2c_client *client,
return 0;
}
-static int bq24735_charger_remove(struct i2c_client *client)
-{
- struct bq24735 *charger = i2c_get_clientdata(client);
-
- if (charger->poll_interval)
- cancel_delayed_work_sync(&charger->poll);
-
- return 0;
-}
-
static const struct i2c_device_id bq24735_charger_id[] = {
{ "bq24735-charger", 0 },
{}
@@ -509,7 +504,6 @@ static struct i2c_driver bq24735_charger_driver = {
.of_match_table = bq24735_match_ids,
},
.probe = bq24735_charger_probe,
- .remove = bq24735_charger_remove,
.id_table = bq24735_charger_id,
};
diff --git a/drivers/power/supply/ltc2941-battery-gauge.c b/drivers/power/supply/ltc2941-battery-gauge.c
index 10cd617516ec..09f3e78af4e0 100644
--- a/drivers/power/supply/ltc2941-battery-gauge.c
+++ b/drivers/power/supply/ltc2941-battery-gauge.c
@@ -8,6 +8,7 @@
* Author: Auryn Verwegen
* Author: Mike Looijmans
*/
+#include <linux/devm-helpers.h>
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/of_device.h>
@@ -445,15 +446,6 @@ static enum power_supply_property ltc294x_properties[] = {
POWER_SUPPLY_PROP_CURRENT_NOW,
};
-static int ltc294x_i2c_remove(struct i2c_client *client)
-{
- struct ltc294x_info *info = i2c_get_clientdata(client);
-
- cancel_delayed_work_sync(&info->work);
- power_supply_unregister(info->supply);
- return 0;
-}
-
static int ltc294x_i2c_probe(struct i2c_client *client,
const struct i2c_device_id *id)
{
@@ -547,7 +539,10 @@ static int ltc294x_i2c_probe(struct i2c_client *client,
psy_cfg.drv_data = info;
- INIT_DELAYED_WORK(&info->work, ltc294x_work);
+ ret = devm_delayed_work_autocancel(&client->dev, &info->work,
+ ltc294x_work);
+ if (ret)
+ return ret;
ret = ltc294x_reset(info, prescaler_exp);
if (ret < 0) {
@@ -555,8 +550,8 @@ static int ltc294x_i2c_probe(struct i2c_client *client,
return ret;
}
- info->supply = power_supply_register(&client->dev, &info->supply_desc,
- &psy_cfg);
+ info->supply = devm_power_supply_register(&client->dev,
+ &info->supply_desc, &psy_cfg);
if (IS_ERR(info->supply)) {
dev_err(&client->dev, "failed to register ltc2941\n");
return PTR_ERR(info->supply);
@@ -655,7 +650,6 @@ static struct i2c_driver ltc294x_driver = {
.pm = LTC294X_PM_OPS,
},
.probe = ltc294x_i2c_probe,
- .remove = ltc294x_i2c_remove,
.shutdown = ltc294x_i2c_shutdown,
.id_table = ltc294x_i2c_id,
};
diff --git a/drivers/power/supply/sbs-battery.c b/drivers/power/supply/sbs-battery.c
index b6a538ebb378..70ea404b2a36 100644
--- a/drivers/power/supply/sbs-battery.c
+++ b/drivers/power/supply/sbs-battery.c
@@ -7,6 +7,7 @@
#include <linux/bits.h>
#include <linux/delay.h>
+#include <linux/devm-helpers.h>
#include <linux/err.h>
#include <linux/gpio/consumer.h>
#include <linux/i2c.h>
@@ -1165,7 +1166,10 @@ static int sbs_probe(struct i2c_client *client)
}
}
- INIT_DELAYED_WORK(&chip->work, sbs_delayed_work);
+ rc = devm_delayed_work_autocancel(&client->dev, &chip->work,
+ sbs_delayed_work);
+ if (rc)
+ return rc;
chip->power_supply = devm_power_supply_register(&client->dev, sbs_desc,
&psy_cfg);
@@ -1185,15 +1189,6 @@ static int sbs_probe(struct i2c_client *client)
return rc;
}
-static int sbs_remove(struct i2c_client *client)
-{
- struct sbs_info *chip = i2c_get_clientdata(client);
-
- cancel_delayed_work_sync(&chip->work);
-
- return 0;
-}
-
#if defined CONFIG_PM_SLEEP
static int sbs_suspend(struct device *dev)
@@ -1248,7 +1243,6 @@ MODULE_DEVICE_TABLE(of, sbs_dt_ids);
static struct i2c_driver sbs_battery_driver = {
.probe_new = sbs_probe,
- .remove = sbs_remove,
.alert = sbs_alert,
.id_table = sbs_id,
.driver = {
--
2.25.4
--
Matti Vaittinen, Linux device drivers
ROHM Semiconductors, Finland SWDC
Kiviharjunlenkki 1E
90220 OULU
FINLAND
~~~ "I don't think so," said Rene Descartes. Just then he vanished ~~~
Simon says - in Latin please.
~~~ "non cogito me" dixit Rene Descarte, deinde evanescavit ~~~
Thanks to Simon Glass for the translation =]
Few drivers implement remove call-back only for ensuring a delayed
work gets cancelled prior driver removal. Clean-up these by switching
to use devm_delayed_work_autocancel() instead.
Additionally, this helps avoiding mixing devm and manual resource
management and cleans up a (theoretical?) bug from extconn-palmas.c
and extcon-qcom-spmi-misc.c where (devm managed)IRQ might schedule
new work item after wq was cleaned at remove().
This change is compile-tested only. All testing is appreciated.
Signed-off-by: Matti Vaittinen <[email protected]>
---
Changelog from RFCv2:
- RFC dropped. No functional changes.
drivers/extcon/extcon-gpio.c | 15 ++++-----------
drivers/extcon/extcon-intel-int3496.c | 16 ++++------------
drivers/extcon/extcon-palmas.c | 17 ++++++-----------
drivers/extcon/extcon-qcom-spmi-misc.c | 17 ++++++-----------
4 files changed, 20 insertions(+), 45 deletions(-)
diff --git a/drivers/extcon/extcon-gpio.c b/drivers/extcon/extcon-gpio.c
index c211222f5d0c..4105df74f2b0 100644
--- a/drivers/extcon/extcon-gpio.c
+++ b/drivers/extcon/extcon-gpio.c
@@ -9,6 +9,7 @@
* (originally switch class is supported)
*/
+#include <linux/devm-helpers.h>
#include <linux/extcon-provider.h>
#include <linux/gpio/consumer.h>
#include <linux/init.h>
@@ -112,7 +113,9 @@ static int gpio_extcon_probe(struct platform_device *pdev)
if (ret < 0)
return ret;
- INIT_DELAYED_WORK(&data->work, gpio_extcon_work);
+ ret = devm_delayed_work_autocancel(dev, &data->work, gpio_extcon_work);
+ if (ret)
+ return ret;
/*
* Request the interrupt of gpio to detect whether external connector
@@ -131,15 +134,6 @@ static int gpio_extcon_probe(struct platform_device *pdev)
return 0;
}
-static int gpio_extcon_remove(struct platform_device *pdev)
-{
- struct gpio_extcon_data *data = platform_get_drvdata(pdev);
-
- cancel_delayed_work_sync(&data->work);
-
- return 0;
-}
-
#ifdef CONFIG_PM_SLEEP
static int gpio_extcon_resume(struct device *dev)
{
@@ -158,7 +152,6 @@ static SIMPLE_DEV_PM_OPS(gpio_extcon_pm_ops, NULL, gpio_extcon_resume);
static struct platform_driver gpio_extcon_driver = {
.probe = gpio_extcon_probe,
- .remove = gpio_extcon_remove,
.driver = {
.name = "extcon-gpio",
.pm = &gpio_extcon_pm_ops,
diff --git a/drivers/extcon/extcon-intel-int3496.c b/drivers/extcon/extcon-intel-int3496.c
index 80c9abcc3f97..fb527c23639e 100644
--- a/drivers/extcon/extcon-intel-int3496.c
+++ b/drivers/extcon/extcon-intel-int3496.c
@@ -11,6 +11,7 @@
*/
#include <linux/acpi.h>
+#include <linux/devm-helpers.h>
#include <linux/extcon-provider.h>
#include <linux/gpio/consumer.h>
#include <linux/interrupt.h>
@@ -101,7 +102,9 @@ static int int3496_probe(struct platform_device *pdev)
return -ENOMEM;
data->dev = dev;
- INIT_DELAYED_WORK(&data->work, int3496_do_usb_id);
+ ret = devm_delayed_work_autocancel(dev, &data->work, int3496_do_usb_id);
+ if (ret)
+ return ret;
data->gpio_usb_id = devm_gpiod_get(dev, "id", GPIOD_IN);
if (IS_ERR(data->gpio_usb_id)) {
@@ -155,16 +158,6 @@ static int int3496_probe(struct platform_device *pdev)
return 0;
}
-static int int3496_remove(struct platform_device *pdev)
-{
- struct int3496_data *data = platform_get_drvdata(pdev);
-
- devm_free_irq(&pdev->dev, data->usb_id_irq, data);
- cancel_delayed_work_sync(&data->work);
-
- return 0;
-}
-
static const struct acpi_device_id int3496_acpi_match[] = {
{ "INT3496" },
{ }
@@ -177,7 +170,6 @@ static struct platform_driver int3496_driver = {
.acpi_match_table = int3496_acpi_match,
},
.probe = int3496_probe,
- .remove = int3496_remove,
};
module_platform_driver(int3496_driver);
diff --git a/drivers/extcon/extcon-palmas.c b/drivers/extcon/extcon-palmas.c
index a2852bcc5f0d..d2c1a8b89c08 100644
--- a/drivers/extcon/extcon-palmas.c
+++ b/drivers/extcon/extcon-palmas.c
@@ -9,6 +9,7 @@
* Author: Hema HK <[email protected]>
*/
+#include <linux/devm-helpers.h>
#include <linux/module.h>
#include <linux/interrupt.h>
#include <linux/platform_device.h>
@@ -237,7 +238,11 @@ static int palmas_usb_probe(struct platform_device *pdev)
palmas_usb->sw_debounce_jiffies = msecs_to_jiffies(debounce);
}
- INIT_DELAYED_WORK(&palmas_usb->wq_detectid, palmas_gpio_id_detect);
+ status = devm_delayed_work_autocancel(&pdev->dev,
+ &palmas_usb->wq_detectid,
+ palmas_gpio_id_detect);
+ if (status)
+ return status;
palmas->usb = palmas_usb;
palmas_usb->palmas = palmas;
@@ -359,15 +364,6 @@ static int palmas_usb_probe(struct platform_device *pdev)
return 0;
}
-static int palmas_usb_remove(struct platform_device *pdev)
-{
- struct palmas_usb *palmas_usb = platform_get_drvdata(pdev);
-
- cancel_delayed_work_sync(&palmas_usb->wq_detectid);
-
- return 0;
-}
-
#ifdef CONFIG_PM_SLEEP
static int palmas_usb_suspend(struct device *dev)
{
@@ -422,7 +418,6 @@ static const struct of_device_id of_palmas_match_tbl[] = {
static struct platform_driver palmas_usb_driver = {
.probe = palmas_usb_probe,
- .remove = palmas_usb_remove,
.driver = {
.name = "palmas-usb",
.of_match_table = of_palmas_match_tbl,
diff --git a/drivers/extcon/extcon-qcom-spmi-misc.c b/drivers/extcon/extcon-qcom-spmi-misc.c
index 6b836ae62176..74d57d951b68 100644
--- a/drivers/extcon/extcon-qcom-spmi-misc.c
+++ b/drivers/extcon/extcon-qcom-spmi-misc.c
@@ -7,6 +7,7 @@
* Stephen Boyd <[email protected]>
*/
+#include <linux/devm-helpers.h>
#include <linux/extcon-provider.h>
#include <linux/init.h>
#include <linux/interrupt.h>
@@ -80,7 +81,11 @@ static int qcom_usb_extcon_probe(struct platform_device *pdev)
}
info->debounce_jiffies = msecs_to_jiffies(USB_ID_DEBOUNCE_MS);
- INIT_DELAYED_WORK(&info->wq_detcable, qcom_usb_extcon_detect_cable);
+
+ ret = devm_delayed_work_autocancel(dev, &info->wq_detcable,
+ qcom_usb_extcon_detect_cable);
+ if (ret)
+ return ret;
info->irq = platform_get_irq_byname(pdev, "usb_id");
if (info->irq < 0)
@@ -105,15 +110,6 @@ static int qcom_usb_extcon_probe(struct platform_device *pdev)
return 0;
}
-static int qcom_usb_extcon_remove(struct platform_device *pdev)
-{
- struct qcom_usb_extcon_info *info = platform_get_drvdata(pdev);
-
- cancel_delayed_work_sync(&info->wq_detcable);
-
- return 0;
-}
-
#ifdef CONFIG_PM_SLEEP
static int qcom_usb_extcon_suspend(struct device *dev)
{
@@ -149,7 +145,6 @@ MODULE_DEVICE_TABLE(of, qcom_usb_extcon_dt_match);
static struct platform_driver qcom_usb_extcon_driver = {
.probe = qcom_usb_extcon_probe,
- .remove = qcom_usb_extcon_remove,
.driver = {
.name = "extcon-pm8941-misc",
.pm = &qcom_usb_extcon_pm_ops,
--
2.25.4
--
Matti Vaittinen, Linux device drivers
ROHM Semiconductors, Finland SWDC
Kiviharjunlenkki 1E
90220 OULU
FINLAND
~~~ "I don't think so," said Rene Descartes. Just then he vanished ~~~
Simon says - in Latin please.
~~~ "non cogito me" dixit Rene Descarte, deinde evanescavit ~~~
Thanks to Simon Glass for the translation =]
On Tue, Mar 23, 2021 at 02:58:28PM +0100, Hans de Goede wrote:
> Hi,
>
> On 3/23/21 2:56 PM, Matti Vaittinen wrote:
> > Devm helper header containing small inline helpers was added.
> > Hans promised to maintain it.
> >
> > Add Hans as maintainer and myself as designated reviewer.
> >
> > Signed-off-by: Matti Vaittinen <[email protected]>
>
> Yes I did promise that, didn't I? FWIW going this route is still
> fine by me, assuming that having someone else maintain this makes
> this easier on / more acceptable to Greg.
>
> Ultimately this is up to Greg though, so lets wait and see what
> Greg has to say about this.
Can we move some of the devm_* calls in include/device.h into here as
well so that you all can be in charge of them instead of me?
If so, I'm happy :)
Anyway, this looks sane, I'll queue it up and let's see what breaks in
linux-next...
thanks,
greg k-h
A few drivers which need a delayed work-queue must cancel work at driver
detach. Some of those implement remove() solely for this purpose. Help
drivers to avoid unnecessary remove and error-branch implementation by
adding managed verision of delayed work initialization. This will also
help drivers to avoid mixing manual and devm based unwinding when other
resources are handled by devm.
Signed-off-by: Matti Vaittinen <[email protected]>
---
Changelog from RFCv2:
- RFC dropped. No functional changes.
include/linux/devm-helpers.h | 53 ++++++++++++++++++++++++++++++++++++
1 file changed, 53 insertions(+)
create mode 100644 include/linux/devm-helpers.h
diff --git a/include/linux/devm-helpers.h b/include/linux/devm-helpers.h
new file mode 100644
index 000000000000..f64e0c9f3763
--- /dev/null
+++ b/include/linux/devm-helpers.h
@@ -0,0 +1,53 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+#ifndef __LINUX_DEVM_HELPERS_H
+#define __LINUX_DEVM_HELPERS_H
+
+/*
+ * Functions which do automatically cancel operations or release resources upon
+ * driver detach.
+ *
+ * These should be helpful to avoid mixing the manual and devm-based resource
+ * management which can be source of annoying, rarely occurring,
+ * hard-to-reproduce bugs.
+ *
+ * Please take into account that devm based cancellation may be performed some
+ * time after the remove() is ran.
+ *
+ * Thus mixing devm and manual resource management can easily cause problems
+ * when unwinding operations with dependencies. IRQ scheduling a work in a queue
+ * is typical example where IRQs are often devm-managed and WQs are manually
+ * cleaned at remove(). If IRQs are not manually freed at remove() (and this is
+ * often the case when we use devm for IRQs) we have a period of time after
+ * remove() - and before devm managed IRQs are freed - where new IRQ may fire
+ * and schedule a work item which won't be cancelled because remove() was
+ * already ran.
+ */
+
+#include <linux/device.h>
+#include <linux/workqueue.h>
+
+static inline void devm_delayed_work_drop(void *res)
+{
+ cancel_delayed_work_sync(res);
+}
+
+/**
+ * devm_delayed_work_autocancel - Resource-managed work allocation
+ * @dev: Device which lifetime work is bound to
+ * @pdata: work to be cancelled when driver is detached
+ *
+ * Initialize work which is automatically cancelled when driver is detached.
+ * A few drivers need delayed work which must be cancelled before driver
+ * is detached to avoid accessing removed resources.
+ * devm_delayed_work_autocancel() can be used to omit the explicit
+ * cancelleation when driver is detached.
+ */
+static inline int devm_delayed_work_autocancel(struct device *dev,
+ struct delayed_work *w,
+ work_func_t worker)
+{
+ INIT_DELAYED_WORK(w, worker);
+ return devm_add_action(dev, devm_delayed_work_drop, w);
+}
+
+#endif
--
2.25.4
--
Matti Vaittinen, Linux device drivers
ROHM Semiconductors, Finland SWDC
Kiviharjunlenkki 1E
90220 OULU
FINLAND
~~~ "I don't think so," said Rene Descartes. Just then he vanished ~~~
Simon says - in Latin please.
~~~ "non cogito me" dixit Rene Descarte, deinde evanescavit ~~~
Thanks to Simon Glass for the translation =]
Few drivers implement remove call-back only for ensuring a delayed
work gets cancelled prior driver removal. Clean-up these by switching
to use devm_delayed_work_autocancel() instead.
This change is compile-tested only. All testing is appreciated.
Signed-off-by: Matti Vaittinen <[email protected]>
Acked-by: Guenter Roeck <[email protected]>
---
Changelog from RFCv2:
- RFC dropped. No functional changes.
drivers/hwmon/raspberrypi-hwmon.c | 17 ++++++-----------
1 file changed, 6 insertions(+), 11 deletions(-)
diff --git a/drivers/hwmon/raspberrypi-hwmon.c b/drivers/hwmon/raspberrypi-hwmon.c
index d3a64a35f7a9..805d396aa81b 100644
--- a/drivers/hwmon/raspberrypi-hwmon.c
+++ b/drivers/hwmon/raspberrypi-hwmon.c
@@ -7,6 +7,7 @@
* Copyright (C) 2018 Stefan Wahren <[email protected]>
*/
#include <linux/device.h>
+#include <linux/devm-helpers.h>
#include <linux/err.h>
#include <linux/hwmon.h>
#include <linux/module.h>
@@ -106,6 +107,7 @@ static int rpi_hwmon_probe(struct platform_device *pdev)
{
struct device *dev = &pdev->dev;
struct rpi_hwmon_data *data;
+ int ret;
data = devm_kzalloc(dev, sizeof(*data), GFP_KERNEL);
if (!data)
@@ -119,7 +121,10 @@ static int rpi_hwmon_probe(struct platform_device *pdev)
&rpi_chip_info,
NULL);
- INIT_DELAYED_WORK(&data->get_values_poll_work, get_values_poll);
+ ret = devm_delayed_work_autocancel(dev, &data->get_values_poll_work,
+ get_values_poll);
+ if (ret)
+ return ret;
platform_set_drvdata(pdev, data);
if (!PTR_ERR_OR_ZERO(data->hwmon_dev))
@@ -128,18 +133,8 @@ static int rpi_hwmon_probe(struct platform_device *pdev)
return PTR_ERR_OR_ZERO(data->hwmon_dev);
}
-static int rpi_hwmon_remove(struct platform_device *pdev)
-{
- struct rpi_hwmon_data *data = platform_get_drvdata(pdev);
-
- cancel_delayed_work_sync(&data->get_values_poll_work);
-
- return 0;
-}
-
static struct platform_driver rpi_hwmon_driver = {
.probe = rpi_hwmon_probe,
- .remove = rpi_hwmon_remove,
.driver = {
.name = "raspberrypi-hwmon",
},
--
2.25.4
--
Matti Vaittinen, Linux device drivers
ROHM Semiconductors, Finland SWDC
Kiviharjunlenkki 1E
90220 OULU
FINLAND
~~~ "I don't think so," said Rene Descartes. Just then he vanished ~~~
Simon says - in Latin please.
~~~ "non cogito me" dixit Rene Descarte, deinde evanescavit ~~~
Thanks to Simon Glass for the translation =]
Few drivers implement remove call-back only for ensuring a delayed
work gets cancelled prior driver removal. Clean-up these by switching
to use devm_delayed_work_autocancel() instead.
This change is compile-tested only. All testing is appreciated.
Signed-off-by: Matti Vaittinen <[email protected]>
---
Changelog from RFCv2:
- RFC dropped. No functional changes.
drivers/platform/x86/gpd-pocket-fan.c | 17 ++++++-----------
1 file changed, 6 insertions(+), 11 deletions(-)
diff --git a/drivers/platform/x86/gpd-pocket-fan.c b/drivers/platform/x86/gpd-pocket-fan.c
index 5b516e4c2bfb..7a20f68ae206 100644
--- a/drivers/platform/x86/gpd-pocket-fan.c
+++ b/drivers/platform/x86/gpd-pocket-fan.c
@@ -6,6 +6,7 @@
*/
#include <linux/acpi.h>
+#include <linux/devm-helpers.h>
#include <linux/gpio/consumer.h>
#include <linux/module.h>
#include <linux/moduleparam.h>
@@ -124,7 +125,7 @@ static void gpd_pocket_fan_force_update(struct gpd_pocket_fan_data *fan)
static int gpd_pocket_fan_probe(struct platform_device *pdev)
{
struct gpd_pocket_fan_data *fan;
- int i;
+ int i, ret;
for (i = 0; i < ARRAY_SIZE(temp_limits); i++) {
if (temp_limits[i] < 20000 || temp_limits[i] > 90000) {
@@ -152,7 +153,10 @@ static int gpd_pocket_fan_probe(struct platform_device *pdev)
return -ENOMEM;
fan->dev = &pdev->dev;
- INIT_DELAYED_WORK(&fan->work, gpd_pocket_fan_worker);
+ ret = devm_delayed_work_autocancel(&pdev->dev, &fan->work,
+ gpd_pocket_fan_worker);
+ if (ret)
+ return ret;
/* Note this returns a "weak" reference which we don't need to free */
fan->dts0 = thermal_zone_get_zone_by_name("soc_dts0");
@@ -177,14 +181,6 @@ static int gpd_pocket_fan_probe(struct platform_device *pdev)
return 0;
}
-static int gpd_pocket_fan_remove(struct platform_device *pdev)
-{
- struct gpd_pocket_fan_data *fan = platform_get_drvdata(pdev);
-
- cancel_delayed_work_sync(&fan->work);
- return 0;
-}
-
#ifdef CONFIG_PM_SLEEP
static int gpd_pocket_fan_suspend(struct device *dev)
{
@@ -215,7 +211,6 @@ MODULE_DEVICE_TABLE(acpi, gpd_pocket_fan_acpi_match);
static struct platform_driver gpd_pocket_fan_driver = {
.probe = gpd_pocket_fan_probe,
- .remove = gpd_pocket_fan_remove,
.driver = {
.name = "gpd_pocket_fan",
.acpi_match_table = gpd_pocket_fan_acpi_match,
--
2.25.4
--
Matti Vaittinen, Linux device drivers
ROHM Semiconductors, Finland SWDC
Kiviharjunlenkki 1E
90220 OULU
FINLAND
~~~ "I don't think so," said Rene Descartes. Just then he vanished ~~~
Simon says - in Latin please.
~~~ "non cogito me" dixit Rene Descarte, deinde evanescavit ~~~
Thanks to Simon Glass for the translation =]
Few drivers implement remove call-back only for ensuring a delayed
work gets cancelled prior driver removal. Clean-up these by switching
to use devm_delayed_work_autocancel() instead.
Additionally, this helps avoiding mixing devm and manual resource
management and cleans up a (theoretical?) bug where devm managed
over-current IRQ might schedule a new work item after wq was cleaned
at remove().
This change is compile-tested only. All testing is appreciated.
Signed-off-by: Matti Vaittinen <[email protected]>
---
Changelog from RFCv2:
- RFC dropped. No functional changes.
drivers/regulator/qcom_spmi-regulator.c | 34 ++++++-------------------
1 file changed, 8 insertions(+), 26 deletions(-)
diff --git a/drivers/regulator/qcom_spmi-regulator.c b/drivers/regulator/qcom_spmi-regulator.c
index e62e1d72d943..c2442d7798ab 100644
--- a/drivers/regulator/qcom_spmi-regulator.c
+++ b/drivers/regulator/qcom_spmi-regulator.c
@@ -5,6 +5,7 @@
#include <linux/module.h>
#include <linux/delay.h>
+#include <linux/devm-helpers.h>
#include <linux/err.h>
#include <linux/kernel.h>
#include <linux/interrupt.h>
@@ -1842,7 +1843,10 @@ static int spmi_regulator_of_parse(struct device_node *node,
return ret;
}
- INIT_DELAYED_WORK(&vreg->ocp_work, spmi_regulator_vs_ocp_work);
+ ret = devm_delayed_work_autocancel(dev, &vreg->ocp_work,
+ spmi_regulator_vs_ocp_work);
+ if (ret)
+ return ret;
}
return 0;
@@ -2157,10 +2161,8 @@ static int qcom_spmi_regulator_probe(struct platform_device *pdev)
vreg->regmap = regmap;
if (reg->ocp) {
vreg->ocp_irq = platform_get_irq_byname(pdev, reg->ocp);
- if (vreg->ocp_irq < 0) {
- ret = vreg->ocp_irq;
- goto err;
- }
+ if (vreg->ocp_irq < 0)
+ return vreg->ocp_irq;
}
vreg->desc.id = -1;
vreg->desc.owner = THIS_MODULE;
@@ -2203,8 +2205,7 @@ static int qcom_spmi_regulator_probe(struct platform_device *pdev)
rdev = devm_regulator_register(dev, &vreg->desc, &config);
if (IS_ERR(rdev)) {
dev_err(dev, "failed to register %s\n", name);
- ret = PTR_ERR(rdev);
- goto err;
+ return PTR_ERR(rdev);
}
INIT_LIST_HEAD(&vreg->node);
@@ -2212,24 +2213,6 @@ static int qcom_spmi_regulator_probe(struct platform_device *pdev)
}
return 0;
-
-err:
- list_for_each_entry(vreg, vreg_list, node)
- if (vreg->ocp_irq)
- cancel_delayed_work_sync(&vreg->ocp_work);
- return ret;
-}
-
-static int qcom_spmi_regulator_remove(struct platform_device *pdev)
-{
- struct spmi_regulator *vreg;
- struct list_head *vreg_list = platform_get_drvdata(pdev);
-
- list_for_each_entry(vreg, vreg_list, node)
- if (vreg->ocp_irq)
- cancel_delayed_work_sync(&vreg->ocp_work);
-
- return 0;
}
static struct platform_driver qcom_spmi_regulator_driver = {
@@ -2238,7 +2221,6 @@ static struct platform_driver qcom_spmi_regulator_driver = {
.of_match_table = qcom_spmi_regulator_match,
},
.probe = qcom_spmi_regulator_probe,
- .remove = qcom_spmi_regulator_remove,
};
module_platform_driver(qcom_spmi_regulator_driver);
--
2.25.4
--
Matti Vaittinen, Linux device drivers
ROHM Semiconductors, Finland SWDC
Kiviharjunlenkki 1E
90220 OULU
FINLAND
~~~ "I don't think so," said Rene Descartes. Just then he vanished ~~~
Simon says - in Latin please.
~~~ "non cogito me" dixit Rene Descarte, deinde evanescavit ~~~
Thanks to Simon Glass for the translation =]
Hi,
On 3/23/21 2:56 PM, Matti Vaittinen wrote:
> Devm helper header containing small inline helpers was added.
> Hans promised to maintain it.
>
> Add Hans as maintainer and myself as designated reviewer.
>
> Signed-off-by: Matti Vaittinen <[email protected]>
Yes I did promise that, didn't I? FWIW going this route is still
fine by me, assuming that having someone else maintain this makes
this easier on / more acceptable to Greg.
Ultimately this is up to Greg though, so lets wait and see what
Greg has to say about this.
Regards,
Hans
> ---
> Changelog from RFCv2:
> - RFC dropped. No functional changes.
>
> MAINTAINERS | 6 ++++++
> 1 file changed, 6 insertions(+)
>
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 9e876927c60d..fa5ac3164678 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -5169,6 +5169,12 @@ M: Torben Mathiasen <[email protected]>
> S: Maintained
> W: http://lanana.org/docs/device-list/index.html
>
> +DEVICE RESOURCE MANAGEMENT HELPERS
> +M: Hans de Goede <[email protected]>
> +R: Matti Vaittinen <[email protected]>
> +S: Maintained
> +F: include/linux/devm-helpers.h
> +
> DEVICE-MAPPER (LVM)
> M: Alasdair Kergon <[email protected]>
> M: Mike Snitzer <[email protected]>
>
Few drivers implement remove call-back only for ensuring a delayed
work gets cancelled prior driver removal. Clean-up these by switching
to use devm_delayed_work_autocancel() instead.
This change is compile-tested only. All testing is appreciated.
Signed-off-by: Matti Vaittinen <[email protected]>
Acked-by: Guenter Roeck <[email protected]>
---
Changelog from RFCv2:
- RFC dropped. No functional changes.
drivers/watchdog/retu_wdt.c | 22 ++++++----------------
1 file changed, 6 insertions(+), 16 deletions(-)
diff --git a/drivers/watchdog/retu_wdt.c b/drivers/watchdog/retu_wdt.c
index 258dfcf9cbda..2b9017e1cd91 100644
--- a/drivers/watchdog/retu_wdt.c
+++ b/drivers/watchdog/retu_wdt.c
@@ -8,6 +8,7 @@
* Rewritten by Aaro Koskinen.
*/
+#include <linux/devm-helpers.h>
#include <linux/slab.h>
#include <linux/errno.h>
#include <linux/device.h>
@@ -127,9 +128,12 @@ static int retu_wdt_probe(struct platform_device *pdev)
wdev->rdev = rdev;
wdev->dev = &pdev->dev;
- INIT_DELAYED_WORK(&wdev->ping_work, retu_wdt_ping_work);
+ ret = devm_delayed_work_autocancel(&pdev->dev, &wdev->ping_work,
+ retu_wdt_ping_work);
+ if (ret)
+ return ret;
- ret = watchdog_register_device(retu_wdt);
+ ret = devm_watchdog_register_device(&pdev->dev, retu_wdt);
if (ret < 0)
return ret;
@@ -138,25 +142,11 @@ static int retu_wdt_probe(struct platform_device *pdev)
else
retu_wdt_ping_enable(wdev);
- platform_set_drvdata(pdev, retu_wdt);
-
- return 0;
-}
-
-static int retu_wdt_remove(struct platform_device *pdev)
-{
- struct watchdog_device *wdog = platform_get_drvdata(pdev);
- struct retu_wdt_dev *wdev = watchdog_get_drvdata(wdog);
-
- watchdog_unregister_device(wdog);
- cancel_delayed_work_sync(&wdev->ping_work);
-
return 0;
}
static struct platform_driver retu_wdt_driver = {
.probe = retu_wdt_probe,
- .remove = retu_wdt_remove,
.driver = {
.name = "retu-wdt",
},
--
2.25.4
--
Matti Vaittinen, Linux device drivers
ROHM Semiconductors, Finland SWDC
Kiviharjunlenkki 1E
90220 OULU
FINLAND
~~~ "I don't think so," said Rene Descartes. Just then he vanished ~~~
Simon says - in Latin please.
~~~ "non cogito me" dixit Rene Descarte, deinde evanescavit ~~~
Thanks to Simon Glass for the translation =]
Hi,
On 3/23/21 2:57 PM, Matti Vaittinen wrote:
> Few drivers implement remove call-back only for ensuring a delayed
> work gets cancelled prior driver removal. Clean-up these by switching
> to use devm_delayed_work_autocancel() instead.
>
> Additionally, this helps avoiding mixing devm and manual resource
> management and cleans up a (theoretical?) bug from extconn-palmas.c
> and extcon-qcom-spmi-misc.c where (devm managed)IRQ might schedule
> new work item after wq was cleaned at remove().
>
> This change is compile-tested only. All testing is appreciated.
>
> Signed-off-by: Matti Vaittinen <[email protected]>
Thanks, patch looks good to me:
Reviewed-by: Hans de Goede <[email protected]>
Note that this is not just a cleanup, it also actually fixes a
couple of driver-unbind races where the work was stopped before
the IRQ is free-ed, so there is a race where the work can be
restarted.
Regards,
Hans
> ---
> Changelog from RFCv2:
> - RFC dropped. No functional changes.
>
> drivers/extcon/extcon-gpio.c | 15 ++++-----------
> drivers/extcon/extcon-intel-int3496.c | 16 ++++------------
> drivers/extcon/extcon-palmas.c | 17 ++++++-----------
> drivers/extcon/extcon-qcom-spmi-misc.c | 17 ++++++-----------
> 4 files changed, 20 insertions(+), 45 deletions(-)
>
> diff --git a/drivers/extcon/extcon-gpio.c b/drivers/extcon/extcon-gpio.c
> index c211222f5d0c..4105df74f2b0 100644
> --- a/drivers/extcon/extcon-gpio.c
> +++ b/drivers/extcon/extcon-gpio.c
> @@ -9,6 +9,7 @@
> * (originally switch class is supported)
> */
>
> +#include <linux/devm-helpers.h>
> #include <linux/extcon-provider.h>
> #include <linux/gpio/consumer.h>
> #include <linux/init.h>
> @@ -112,7 +113,9 @@ static int gpio_extcon_probe(struct platform_device *pdev)
> if (ret < 0)
> return ret;
>
> - INIT_DELAYED_WORK(&data->work, gpio_extcon_work);
> + ret = devm_delayed_work_autocancel(dev, &data->work, gpio_extcon_work);
> + if (ret)
> + return ret;
>
> /*
> * Request the interrupt of gpio to detect whether external connector
> @@ -131,15 +134,6 @@ static int gpio_extcon_probe(struct platform_device *pdev)
> return 0;
> }
>
> -static int gpio_extcon_remove(struct platform_device *pdev)
> -{
> - struct gpio_extcon_data *data = platform_get_drvdata(pdev);
> -
> - cancel_delayed_work_sync(&data->work);
> -
> - return 0;
> -}
> -
> #ifdef CONFIG_PM_SLEEP
> static int gpio_extcon_resume(struct device *dev)
> {
> @@ -158,7 +152,6 @@ static SIMPLE_DEV_PM_OPS(gpio_extcon_pm_ops, NULL, gpio_extcon_resume);
>
> static struct platform_driver gpio_extcon_driver = {
> .probe = gpio_extcon_probe,
> - .remove = gpio_extcon_remove,
> .driver = {
> .name = "extcon-gpio",
> .pm = &gpio_extcon_pm_ops,
> diff --git a/drivers/extcon/extcon-intel-int3496.c b/drivers/extcon/extcon-intel-int3496.c
> index 80c9abcc3f97..fb527c23639e 100644
> --- a/drivers/extcon/extcon-intel-int3496.c
> +++ b/drivers/extcon/extcon-intel-int3496.c
> @@ -11,6 +11,7 @@
> */
>
> #include <linux/acpi.h>
> +#include <linux/devm-helpers.h>
> #include <linux/extcon-provider.h>
> #include <linux/gpio/consumer.h>
> #include <linux/interrupt.h>
> @@ -101,7 +102,9 @@ static int int3496_probe(struct platform_device *pdev)
> return -ENOMEM;
>
> data->dev = dev;
> - INIT_DELAYED_WORK(&data->work, int3496_do_usb_id);
> + ret = devm_delayed_work_autocancel(dev, &data->work, int3496_do_usb_id);
> + if (ret)
> + return ret;
>
> data->gpio_usb_id = devm_gpiod_get(dev, "id", GPIOD_IN);
> if (IS_ERR(data->gpio_usb_id)) {
> @@ -155,16 +158,6 @@ static int int3496_probe(struct platform_device *pdev)
> return 0;
> }
>
> -static int int3496_remove(struct platform_device *pdev)
> -{
> - struct int3496_data *data = platform_get_drvdata(pdev);
> -
> - devm_free_irq(&pdev->dev, data->usb_id_irq, data);
> - cancel_delayed_work_sync(&data->work);
> -
> - return 0;
> -}
> -
> static const struct acpi_device_id int3496_acpi_match[] = {
> { "INT3496" },
> { }
> @@ -177,7 +170,6 @@ static struct platform_driver int3496_driver = {
> .acpi_match_table = int3496_acpi_match,
> },
> .probe = int3496_probe,
> - .remove = int3496_remove,
> };
>
> module_platform_driver(int3496_driver);
> diff --git a/drivers/extcon/extcon-palmas.c b/drivers/extcon/extcon-palmas.c
> index a2852bcc5f0d..d2c1a8b89c08 100644
> --- a/drivers/extcon/extcon-palmas.c
> +++ b/drivers/extcon/extcon-palmas.c
> @@ -9,6 +9,7 @@
> * Author: Hema HK <[email protected]>
> */
>
> +#include <linux/devm-helpers.h>
> #include <linux/module.h>
> #include <linux/interrupt.h>
> #include <linux/platform_device.h>
> @@ -237,7 +238,11 @@ static int palmas_usb_probe(struct platform_device *pdev)
> palmas_usb->sw_debounce_jiffies = msecs_to_jiffies(debounce);
> }
>
> - INIT_DELAYED_WORK(&palmas_usb->wq_detectid, palmas_gpio_id_detect);
> + status = devm_delayed_work_autocancel(&pdev->dev,
> + &palmas_usb->wq_detectid,
> + palmas_gpio_id_detect);
> + if (status)
> + return status;
>
> palmas->usb = palmas_usb;
> palmas_usb->palmas = palmas;
> @@ -359,15 +364,6 @@ static int palmas_usb_probe(struct platform_device *pdev)
> return 0;
> }
>
> -static int palmas_usb_remove(struct platform_device *pdev)
> -{
> - struct palmas_usb *palmas_usb = platform_get_drvdata(pdev);
> -
> - cancel_delayed_work_sync(&palmas_usb->wq_detectid);
> -
> - return 0;
> -}
> -
> #ifdef CONFIG_PM_SLEEP
> static int palmas_usb_suspend(struct device *dev)
> {
> @@ -422,7 +418,6 @@ static const struct of_device_id of_palmas_match_tbl[] = {
>
> static struct platform_driver palmas_usb_driver = {
> .probe = palmas_usb_probe,
> - .remove = palmas_usb_remove,
> .driver = {
> .name = "palmas-usb",
> .of_match_table = of_palmas_match_tbl,
> diff --git a/drivers/extcon/extcon-qcom-spmi-misc.c b/drivers/extcon/extcon-qcom-spmi-misc.c
> index 6b836ae62176..74d57d951b68 100644
> --- a/drivers/extcon/extcon-qcom-spmi-misc.c
> +++ b/drivers/extcon/extcon-qcom-spmi-misc.c
> @@ -7,6 +7,7 @@
> * Stephen Boyd <[email protected]>
> */
>
> +#include <linux/devm-helpers.h>
> #include <linux/extcon-provider.h>
> #include <linux/init.h>
> #include <linux/interrupt.h>
> @@ -80,7 +81,11 @@ static int qcom_usb_extcon_probe(struct platform_device *pdev)
> }
>
> info->debounce_jiffies = msecs_to_jiffies(USB_ID_DEBOUNCE_MS);
> - INIT_DELAYED_WORK(&info->wq_detcable, qcom_usb_extcon_detect_cable);
> +
> + ret = devm_delayed_work_autocancel(dev, &info->wq_detcable,
> + qcom_usb_extcon_detect_cable);
> + if (ret)
> + return ret;
>
> info->irq = platform_get_irq_byname(pdev, "usb_id");
> if (info->irq < 0)
> @@ -105,15 +110,6 @@ static int qcom_usb_extcon_probe(struct platform_device *pdev)
> return 0;
> }
>
> -static int qcom_usb_extcon_remove(struct platform_device *pdev)
> -{
> - struct qcom_usb_extcon_info *info = platform_get_drvdata(pdev);
> -
> - cancel_delayed_work_sync(&info->wq_detcable);
> -
> - return 0;
> -}
> -
> #ifdef CONFIG_PM_SLEEP
> static int qcom_usb_extcon_suspend(struct device *dev)
> {
> @@ -149,7 +145,6 @@ MODULE_DEVICE_TABLE(of, qcom_usb_extcon_dt_match);
>
> static struct platform_driver qcom_usb_extcon_driver = {
> .probe = qcom_usb_extcon_probe,
> - .remove = qcom_usb_extcon_remove,
> .driver = {
> .name = "extcon-pm8941-misc",
> .pm = &qcom_usb_extcon_pm_ops,
>
Hi,
On Tue, Mar 23, 2021 at 9:58 PM Matti Vaittinen
<[email protected]> wrote:
>
> Few drivers implement remove call-back only for ensuring a delayed
> work gets cancelled prior driver removal. Clean-up these by switching
> to use devm_delayed_work_autocancel() instead.
>
> This change is compile-tested only. All testing is appreciated.
>
> Signed-off-by: Matti Vaittinen <[email protected]>
> Acked-by: Sebastian Reichel <[email protected]>
> ---
> Changelog from RFCv2:
> - RFC dropped. No functional changes.
>
> drivers/power/supply/axp20x_usb_power.c | 15 +++++----------
> drivers/power/supply/bq24735-charger.c | 18 ++++++------------
> drivers/power/supply/ltc2941-battery-gauge.c | 20 +++++++-------------
> drivers/power/supply/sbs-battery.c | 16 +++++-----------
> 4 files changed, 23 insertions(+), 46 deletions(-)
>
> diff --git a/drivers/power/supply/axp20x_usb_power.c b/drivers/power/supply/axp20x_usb_power.c
> index 8933ae26c3d6..4259709e3491 100644
> --- a/drivers/power/supply/axp20x_usb_power.c
> +++ b/drivers/power/supply/axp20x_usb_power.c
> @@ -8,6 +8,7 @@
>
> #include <linux/bitops.h>
> #include <linux/device.h>
> +#include <linux/devm-helpers.h>
> #include <linux/init.h>
> #include <linux/interrupt.h>
> #include <linux/kernel.h>
> @@ -646,21 +647,16 @@ static int axp20x_usb_power_probe(struct platform_device *pdev)
> }
> }
>
> + ret = devm_delayed_work_autocancel(&pdev->dev, &power->vbus_detect,
> + axp20x_usb_power_poll_vbus);
> + if (ret)
> + return ret;
This doesn't look right. The IRQ is requested before this, and the delayed_work
struct is initialized even earlier, so you'd be re-initializing the struct,
with the work item potentially running or queued up already.
ChenYu
> if (axp20x_usb_vbus_needs_polling(power))
> queue_delayed_work(system_power_efficient_wq, &power->vbus_detect, 0);
>
> return 0;
> }
>
> -static int axp20x_usb_power_remove(struct platform_device *pdev)
> -{
> - struct axp20x_usb_power *power = platform_get_drvdata(pdev);
> -
> - cancel_delayed_work_sync(&power->vbus_detect);
> -
> - return 0;
> -}
> -
> static const struct of_device_id axp20x_usb_power_match[] = {
> {
> .compatible = "x-powers,axp202-usb-power-supply",
> @@ -680,7 +676,6 @@ MODULE_DEVICE_TABLE(of, axp20x_usb_power_match);
>
> static struct platform_driver axp20x_usb_power_driver = {
> .probe = axp20x_usb_power_probe,
> - .remove = axp20x_usb_power_remove,
> .driver = {
> .name = DRVNAME,
> .of_match_table = axp20x_usb_power_match,
HI,
On 3/23/21 2:57 PM, Matti Vaittinen wrote:
> Few drivers implement remove call-back only for ensuring a delayed
> work gets cancelled prior driver removal. Clean-up these by switching
> to use devm_delayed_work_autocancel() instead.
>
> This change is compile-tested only. All testing is appreciated.
>
> Signed-off-by: Matti Vaittinen <[email protected]>
Thanks, patch looks good to me:
Reviewed-by: Hans de Goede <[email protected]>
Regards,
Hans
> ---
> Changelog from RFCv2:
> - RFC dropped. No functional changes.
>
> drivers/platform/x86/gpd-pocket-fan.c | 17 ++++++-----------
> 1 file changed, 6 insertions(+), 11 deletions(-)
>
> diff --git a/drivers/platform/x86/gpd-pocket-fan.c b/drivers/platform/x86/gpd-pocket-fan.c
> index 5b516e4c2bfb..7a20f68ae206 100644
> --- a/drivers/platform/x86/gpd-pocket-fan.c
> +++ b/drivers/platform/x86/gpd-pocket-fan.c
> @@ -6,6 +6,7 @@
> */
>
> #include <linux/acpi.h>
> +#include <linux/devm-helpers.h>
> #include <linux/gpio/consumer.h>
> #include <linux/module.h>
> #include <linux/moduleparam.h>
> @@ -124,7 +125,7 @@ static void gpd_pocket_fan_force_update(struct gpd_pocket_fan_data *fan)
> static int gpd_pocket_fan_probe(struct platform_device *pdev)
> {
> struct gpd_pocket_fan_data *fan;
> - int i;
> + int i, ret;
>
> for (i = 0; i < ARRAY_SIZE(temp_limits); i++) {
> if (temp_limits[i] < 20000 || temp_limits[i] > 90000) {
> @@ -152,7 +153,10 @@ static int gpd_pocket_fan_probe(struct platform_device *pdev)
> return -ENOMEM;
>
> fan->dev = &pdev->dev;
> - INIT_DELAYED_WORK(&fan->work, gpd_pocket_fan_worker);
> + ret = devm_delayed_work_autocancel(&pdev->dev, &fan->work,
> + gpd_pocket_fan_worker);
> + if (ret)
> + return ret;
>
> /* Note this returns a "weak" reference which we don't need to free */
> fan->dts0 = thermal_zone_get_zone_by_name("soc_dts0");
> @@ -177,14 +181,6 @@ static int gpd_pocket_fan_probe(struct platform_device *pdev)
> return 0;
> }
>
> -static int gpd_pocket_fan_remove(struct platform_device *pdev)
> -{
> - struct gpd_pocket_fan_data *fan = platform_get_drvdata(pdev);
> -
> - cancel_delayed_work_sync(&fan->work);
> - return 0;
> -}
> -
> #ifdef CONFIG_PM_SLEEP
> static int gpd_pocket_fan_suspend(struct device *dev)
> {
> @@ -215,7 +211,6 @@ MODULE_DEVICE_TABLE(acpi, gpd_pocket_fan_acpi_match);
>
> static struct platform_driver gpd_pocket_fan_driver = {
> .probe = gpd_pocket_fan_probe,
> - .remove = gpd_pocket_fan_remove,
> .driver = {
> .name = "gpd_pocket_fan",
> .acpi_match_table = gpd_pocket_fan_acpi_match,
>
Hi,
Need to fix the work as following:
s/extconn/extcon
And I'd like you to use the more correct patch title like the following example:
"extcon: Use resource-managed function for delayed work"
Thanks,
Chanwoo Choi
On 3/23/21 10:57 PM, Matti Vaittinen wrote:
> Few drivers implement remove call-back only for ensuring a delayed
> work gets cancelled prior driver removal. Clean-up these by switching
> to use devm_delayed_work_autocancel() instead.
>
> Additionally, this helps avoiding mixing devm and manual resource
> management and cleans up a (theoretical?) bug from extconn-palmas.c
> and extcon-qcom-spmi-misc.c where (devm managed)IRQ might schedule
> new work item after wq was cleaned at remove().
>
> This change is compile-tested only. All testing is appreciated.
>
> Signed-off-by: Matti Vaittinen <[email protected]>
> ---
> Changelog from RFCv2:
> - RFC dropped. No functional changes.
>
> drivers/extcon/extcon-gpio.c | 15 ++++-----------
> drivers/extcon/extcon-intel-int3496.c | 16 ++++------------
> drivers/extcon/extcon-palmas.c | 17 ++++++-----------
> drivers/extcon/extcon-qcom-spmi-misc.c | 17 ++++++-----------
> 4 files changed, 20 insertions(+), 45 deletions(-)
>
> diff --git a/drivers/extcon/extcon-gpio.c b/drivers/extcon/extcon-gpio.c
> index c211222f5d0c..4105df74f2b0 100644
> --- a/drivers/extcon/extcon-gpio.c
> +++ b/drivers/extcon/extcon-gpio.c
> @@ -9,6 +9,7 @@
> * (originally switch class is supported)
> */
>
> +#include <linux/devm-helpers.h>
> #include <linux/extcon-provider.h>
> #include <linux/gpio/consumer.h>
> #include <linux/init.h>
> @@ -112,7 +113,9 @@ static int gpio_extcon_probe(struct platform_device *pdev)
> if (ret < 0)
> return ret;
>
> - INIT_DELAYED_WORK(&data->work, gpio_extcon_work);
> + ret = devm_delayed_work_autocancel(dev, &data->work, gpio_extcon_work);
> + if (ret)
> + return ret;
Need to add the error log as following:
if (ret) {
dev_err(dev, "Failed to initialize delayed_work");
return ret;
}
>
> /*
> * Request the interrupt of gpio to detect whether external connector
> @@ -131,15 +134,6 @@ static int gpio_extcon_probe(struct platform_device *pdev)
> return 0;
> }
>
> -static int gpio_extcon_remove(struct platform_device *pdev)
> -{
> - struct gpio_extcon_data *data = platform_get_drvdata(pdev);
> -
> - cancel_delayed_work_sync(&data->work);
> -
> - return 0;
> -}
> -
> #ifdef CONFIG_PM_SLEEP
> static int gpio_extcon_resume(struct device *dev)
> {
> @@ -158,7 +152,6 @@ static SIMPLE_DEV_PM_OPS(gpio_extcon_pm_ops, NULL, gpio_extcon_resume);
>
> static struct platform_driver gpio_extcon_driver = {
> .probe = gpio_extcon_probe,
> - .remove = gpio_extcon_remove,
> .driver = {
> .name = "extcon-gpio",
> .pm = &gpio_extcon_pm_ops,
> diff --git a/drivers/extcon/extcon-intel-int3496.c b/drivers/extcon/extcon-intel-int3496.c
> index 80c9abcc3f97..fb527c23639e 100644
> --- a/drivers/extcon/extcon-intel-int3496.c
> +++ b/drivers/extcon/extcon-intel-int3496.c
> @@ -11,6 +11,7 @@
> */
>
> #include <linux/acpi.h>
> +#include <linux/devm-helpers.h>
> #include <linux/extcon-provider.h>
> #include <linux/gpio/consumer.h>
> #include <linux/interrupt.h>
> @@ -101,7 +102,9 @@ static int int3496_probe(struct platform_device *pdev)
> return -ENOMEM;
>
> data->dev = dev;
> - INIT_DELAYED_WORK(&data->work, int3496_do_usb_id);
> + ret = devm_delayed_work_autocancel(dev, &data->work, int3496_do_usb_id);
> + if (ret)
> + return ret;
>
> data->gpio_usb_id = devm_gpiod_get(dev, "id", GPIOD_IN);
> if (IS_ERR(data->gpio_usb_id)) {
> @@ -155,16 +158,6 @@ static int int3496_probe(struct platform_device *pdev)
> return 0;
> }
>
> -static int int3496_remove(struct platform_device *pdev)
> -{
> - struct int3496_data *data = platform_get_drvdata(pdev);
> -
> - devm_free_irq(&pdev->dev, data->usb_id_irq, data);
> - cancel_delayed_work_sync(&data->work);
> -
> - return 0;
> -}
> -
> static const struct acpi_device_id int3496_acpi_match[] = {
> { "INT3496" },
> { }
> @@ -177,7 +170,6 @@ static struct platform_driver int3496_driver = {
> .acpi_match_table = int3496_acpi_match,
> },
> .probe = int3496_probe,
> - .remove = int3496_remove,
> };
>
> module_platform_driver(int3496_driver);
> diff --git a/drivers/extcon/extcon-palmas.c b/drivers/extcon/extcon-palmas.c
> index a2852bcc5f0d..d2c1a8b89c08 100644
> --- a/drivers/extcon/extcon-palmas.c
> +++ b/drivers/extcon/extcon-palmas.c
> @@ -9,6 +9,7 @@
> * Author: Hema HK <[email protected]>
> */
>
> +#include <linux/devm-helpers.h>
> #include <linux/module.h>
> #include <linux/interrupt.h>
> #include <linux/platform_device.h>
> @@ -237,7 +238,11 @@ static int palmas_usb_probe(struct platform_device *pdev)
> palmas_usb->sw_debounce_jiffies = msecs_to_jiffies(debounce);
> }
>
> - INIT_DELAYED_WORK(&palmas_usb->wq_detectid, palmas_gpio_id_detect);
> + status = devm_delayed_work_autocancel(&pdev->dev,
> + &palmas_usb->wq_detectid,
> + palmas_gpio_id_detect);
> + if (status)
> + return status;
>
> palmas->usb = palmas_usb;
> palmas_usb->palmas = palmas;
> @@ -359,15 +364,6 @@ static int palmas_usb_probe(struct platform_device *pdev)
> return 0;
> }
>
> -static int palmas_usb_remove(struct platform_device *pdev)
> -{
> - struct palmas_usb *palmas_usb = platform_get_drvdata(pdev);
> -
> - cancel_delayed_work_sync(&palmas_usb->wq_detectid);
> -
> - return 0;
> -}
> -
> #ifdef CONFIG_PM_SLEEP
> static int palmas_usb_suspend(struct device *dev)
> {
> @@ -422,7 +418,6 @@ static const struct of_device_id of_palmas_match_tbl[] = {
>
> static struct platform_driver palmas_usb_driver = {
> .probe = palmas_usb_probe,
> - .remove = palmas_usb_remove,
> .driver = {
> .name = "palmas-usb",
> .of_match_table = of_palmas_match_tbl,
> diff --git a/drivers/extcon/extcon-qcom-spmi-misc.c b/drivers/extcon/extcon-qcom-spmi-misc.c
> index 6b836ae62176..74d57d951b68 100644
> --- a/drivers/extcon/extcon-qcom-spmi-misc.c
> +++ b/drivers/extcon/extcon-qcom-spmi-misc.c
> @@ -7,6 +7,7 @@
> * Stephen Boyd <[email protected]>
> */
>
> +#include <linux/devm-helpers.h>
> #include <linux/extcon-provider.h>
> #include <linux/init.h>
> #include <linux/interrupt.h>
> @@ -80,7 +81,11 @@ static int qcom_usb_extcon_probe(struct platform_device *pdev)
> }
>
> info->debounce_jiffies = msecs_to_jiffies(USB_ID_DEBOUNCE_MS);
> - INIT_DELAYED_WORK(&info->wq_detcable, qcom_usb_extcon_detect_cable);
> +
> + ret = devm_delayed_work_autocancel(dev, &info->wq_detcable,
> + qcom_usb_extcon_detect_cable);
> + if (ret)
> + return ret;
>
> info->irq = platform_get_irq_byname(pdev, "usb_id");
> if (info->irq < 0)
> @@ -105,15 +110,6 @@ static int qcom_usb_extcon_probe(struct platform_device *pdev)
> return 0;
> }
>
> -static int qcom_usb_extcon_remove(struct platform_device *pdev)
> -{
> - struct qcom_usb_extcon_info *info = platform_get_drvdata(pdev);
> -
> - cancel_delayed_work_sync(&info->wq_detcable);
> -
> - return 0;
> -}
> -
> #ifdef CONFIG_PM_SLEEP
> static int qcom_usb_extcon_suspend(struct device *dev)
> {
> @@ -149,7 +145,6 @@ MODULE_DEVICE_TABLE(of, qcom_usb_extcon_dt_match);
>
> static struct platform_driver qcom_usb_extcon_driver = {
> .probe = qcom_usb_extcon_probe,
> - .remove = qcom_usb_extcon_remove,
> .driver = {
> .name = "extcon-pm8941-misc",
> .pm = &qcom_usb_extcon_pm_ops,
>
Hello Chanwoo, Greg,
Thanks for the review.
On Wed, 2021-03-24 at 11:09 +0900, Chanwoo Choi wrote:
> Hi,
>
> Need to fix the work as following:
> s/extconn/extcon
>
> And I'd like you to use the more correct patch title like the
> following example:
> "extcon: Use resource-managed function for delayed work"
I think Greg merged this already. How should we handle this?
> @@ -112,7 +113,9 @@ static int gpio_extcon_probe(struct
> > platform_device *pdev)
> > if (ret < 0)
> > return ret;
> >
> > - INIT_DELAYED_WORK(&data->work, gpio_extcon_work);
> > + ret = devm_delayed_work_autocancel(dev, &data->work,
> > gpio_extcon_work);
> > + if (ret)
> > + return ret;
>
> Need to add the error log as following:
> if (ret) {
> dev_err(dev, "Failed to initialize delayed_work");
> return ret;
> }
I could send incremental patch to Greg for this but it does not change
the commit message.
Best Regards
Matti Vaittinen
Hi,
On 3/24/21 6:02 AM, Matti Vaittinen wrote:
> Hello Chanwoo, Greg,
>
> Thanks for the review.
>
> On Wed, 2021-03-24 at 11:09 +0900, Chanwoo Choi wrote:
>> Hi,
>>
>> Need to fix the work as following:
>> s/extconn/extcon
>>
>> And I'd like you to use the more correct patch title like the
>> following example:
>> "extcon: Use resource-managed function for delayed work"
>
> I think Greg merged this already. How should we handle this?
>
>> @@ -112,7 +113,9 @@ static int gpio_extcon_probe(struct
>>> platform_device *pdev)
>>> if (ret < 0)
>>> return ret;
>>>
>>> - INIT_DELAYED_WORK(&data->work, gpio_extcon_work);
>>> + ret = devm_delayed_work_autocancel(dev, &data->work,
>>> gpio_extcon_work);
>>> + if (ret)
>>> + return ret;
>>
>> Need to add the error log as following:
>> if (ret) {
>> dev_err(dev, "Failed to initialize delayed_work");
>> return ret;
>> }
>
> I could send incremental patch to Greg for this but it does not change
> the commit message.
We cannot do anything about the commit message anymore, but the ordering
issue which you introduced really needs to be fixed.
Please send an incremental patch fixing the wrong order and the double
init of the workqueue.
Regards,
Hans
Hello Chen-Yu, Hans, Greg,
On Tue, 2021-03-23 at 22:36 +0800, Chen-Yu Tsai wrote:
> Hi,
>
> On Tue, Mar 23, 2021 at 9:58 PM Matti Vaittinen
> <[email protected]> wrote:
> > Few drivers implement remove call-back only for ensuring a delayed
> > work gets cancelled prior driver removal. Clean-up these by
> > switching
> > to use devm_delayed_work_autocancel() instead.
> >
> > This change is compile-tested only. All testing is appreciated.
> >
> > Signed-off-by: Matti Vaittinen <[email protected]>
> > Acked-by: Sebastian Reichel <[email protected]>
> > ---
> > Changelog from RFCv2:
> > - RFC dropped. No functional changes.
> >
> > drivers/power/supply/axp20x_usb_power.c | 15 +++++----------
> > drivers/power/supply/bq24735-charger.c | 18 ++++++--------
> > ----
> > drivers/power/supply/ltc2941-battery-gauge.c | 20 +++++++---------
> > ----
> > drivers/power/supply/sbs-battery.c | 16 +++++-----------
> > 4 files changed, 23 insertions(+), 46 deletions(-)
> >
> > diff --git a/drivers/power/supply/axp20x_usb_power.c
> > b/drivers/power/supply/axp20x_usb_power.c
> > index 8933ae26c3d6..4259709e3491 100644
> > --- a/drivers/power/supply/axp20x_usb_power.c
> > +++ b/drivers/power/supply/axp20x_usb_power.c
> > @@ -8,6 +8,7 @@
> >
> > #include <linux/bitops.h>
> > #include <linux/device.h>
> > +#include <linux/devm-helpers.h>
> > #include <linux/init.h>
> > #include <linux/interrupt.h>
> > #include <linux/kernel.h>
> > @@ -646,21 +647,16 @@ static int axp20x_usb_power_probe(struct
> > platform_device *pdev)
> > }
> > }
> >
> > + ret = devm_delayed_work_autocancel(&pdev->dev, &power-
> > >vbus_detect,
> > + axp20x_usb_power_poll_vb
> > us);
> > + if (ret)
> > + return ret;
>
> This doesn't look right. The IRQ is requested before this, and the
> delayed_work
> struct is initialized even earlier, so you'd be re-initializing the
> struct,
> with the work item potentially running or queued up already.
Sigh. The company mail had redirected this to spam... :/
I will check this and send appropriate follow-up fix(es) to Greg. Big
thanks for the heads-up!
--Matti
On Tue, 2021-03-23 at 22:36 +0800, Chen-Yu Tsai wrote:
> Hi,
>
> On Tue, Mar 23, 2021 at 9:58 PM Matti Vaittinen
> <[email protected]> wrote:
> > Few drivers implement remove call-back only for ensuring a delayed
> > work gets cancelled prior driver removal. Clean-up these by
> > switching
> > to use devm_delayed_work_autocancel() instead.
> >
> > This change is compile-tested only. All testing is appreciated.
> >
> > Signed-off-by: Matti Vaittinen <[email protected]>
> > Acked-by: Sebastian Reichel <[email protected]>
> > ---
> > Changelog from RFCv2:
> > - RFC dropped. No functional changes.
> >
> > drivers/power/supply/axp20x_usb_power.c | 15 +++++----------
> > drivers/power/supply/bq24735-charger.c | 18 ++++++--------
> > ----
> > drivers/power/supply/ltc2941-battery-gauge.c | 20 +++++++---------
> > ----
> > drivers/power/supply/sbs-battery.c | 16 +++++-----------
> > 4 files changed, 23 insertions(+), 46 deletions(-)
> >
> > diff --git a/drivers/power/supply/axp20x_usb_power.c
> > b/drivers/power/supply/axp20x_usb_power.c
> > index 8933ae26c3d6..4259709e3491 100644
> > --- a/drivers/power/supply/axp20x_usb_power.c
> > +++ b/drivers/power/supply/axp20x_usb_power.c
> > @@ -8,6 +8,7 @@
> >
> > #include <linux/bitops.h>
> > #include <linux/device.h>
> > +#include <linux/devm-helpers.h>
> > #include <linux/init.h>
> > #include <linux/interrupt.h>
> > #include <linux/kernel.h>
> > @@ -646,21 +647,16 @@ static int axp20x_usb_power_probe(struct
> > platform_device *pdev)
> > }
> > }
> >
> > + ret = devm_delayed_work_autocancel(&pdev->dev, &power-
> > >vbus_detect,
> > + axp20x_usb_power_poll_vb
> > us);
> > + if (ret)
> > + return ret;
>
> This doesn't look right. The IRQ is requested before this, and the
> delayed_work
> struct is initialized even earlier, so you'd be re-initializing the
> struct,
> with the work item potentially running or queued up already.
I checked this and you are 100% correct.
b5e8642ed95ff6ecc20cc6038fe831affa9d098c
"power: supply: axp20x_usb_power: Init work before enabling IRQs" had
fixed the order between RFCv1 and the patch v3. This is what one gets
when not being careful with rebase. Thanks again for the heads-up! I'll
send follow-up fix still today.
Br,
Matti Vaittinen
On Tue, 2021-03-23 at 15:19 +0100, Greg KH wrote:
> On Tue, Mar 23, 2021 at 02:58:28PM +0100, Hans de Goede wrote:
> > Hi,
> >
> > On 3/23/21 2:56 PM, Matti Vaittinen wrote:
> > > Devm helper header containing small inline helpers was added.
> > > Hans promised to maintain it.
> > >
> > > Add Hans as maintainer and myself as designated reviewer.
> > >
> > Ultimately this is up to Greg though, so lets wait and see what
> > Greg has to say about this.
>
> Can we move some of the devm_* calls in include/device.h into here as
> well so that you all can be in charge of them instead of me?
Seems like this was left w/o answer. I guess the question was pointed
to Hans - but what comes to my (not always so humble) opinion - most of
the devm functions in device.h are tightly related to the device
interface or devres. Thus the device.h feels like appropriate place for
most of those. OTOH, the kmalloc/kfree related functions, strdub and
kmemdub might be candidates for move - those are not really "device
things".
But this is really not my call :)
Best Regards
Matti Vaittinen
--
Matti Vaittinen, Linux device drivers
ROHM Semiconductors, Finland SWDC
Kiviharjunlenkki 1E
90220 OULU
FINLAND
Hi,
On 4/21/21 9:51 AM, Matti Vaittinen wrote:
>
> On Tue, 2021-03-23 at 15:19 +0100, Greg KH wrote:
>> On Tue, Mar 23, 2021 at 02:58:28PM +0100, Hans de Goede wrote:
>>> Hi,
>>>
>>> On 3/23/21 2:56 PM, Matti Vaittinen wrote:
>>>> Devm helper header containing small inline helpers was added.
>>>> Hans promised to maintain it.
>>>>
>>>> Add Hans as maintainer and myself as designated reviewer.
>>>>
>>> Ultimately this is up to Greg though, so lets wait and see what
>>> Greg has to say about this.
>>
>> Can we move some of the devm_* calls in include/device.h into here as
>> well so that you all can be in charge of them instead of me?
>
> Seems like this was left w/o answer. I guess the question was pointed
> to Hans
I believe that Greg was (mostly) joking here. At least that is how
I interpreted Greg's reply,which is why I did not answer.
Also note that Greg merged this series, but not this patch,
so the new devm-helpers.h file will presumably be maintained by Greg.
Regards,
Hans
On Wed, 2021-04-21 at 13:58 +0200, Hans de Goede wrote:
> Hi,
>
> On 4/21/21 9:51 AM, Matti Vaittinen wrote:
> > On Tue, 2021-03-23 at 15:19 +0100, Greg KH wrote:
> > > On Tue, Mar 23, 2021 at 02:58:28PM +0100, Hans de Goede wrote:
> > > > Hi,
> > > >
> > > > On 3/23/21 2:56 PM, Matti Vaittinen wrote:
> > > > > Devm helper header containing small inline helpers was added.
> > > > > Hans promised to maintain it.
> > > > >
> > > > > Add Hans as maintainer and myself as designated reviewer.
> > > > >
> > > > Ultimately this is up to Greg though, so lets wait and see what
> > > > Greg has to say about this.
> > >
> > > Can we move some of the devm_* calls in include/device.h into
> > > here as
> > > well so that you all can be in charge of them instead of me?
> >
> > Seems like this was left w/o answer. I guess the question was
> > pointed
> > to Hans
>
> I believe that Greg was (mostly) joking here. At least that is how
> I interpreted Greg's reply,which is why I did not answer.
Ah. I missed the sarcastic tone of typing. I should've noted that by
the font :]
> Also note that Greg merged this series, but not this patch,
> so the new devm-helpers.h file will presumably be maintained by Greg.
Hmm. Are you sure?
https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core.git/commit/?h=driver-core-next&id=2077ca682169afb212d8a887c70057a660290df9
Best Regards
Matti Vaittinen
On Wed, Apr 21, 2021 at 01:58:29PM +0200, Hans de Goede wrote:
> Hi,
>
> On 4/21/21 9:51 AM, Matti Vaittinen wrote:
> >
> > On Tue, 2021-03-23 at 15:19 +0100, Greg KH wrote:
> >> On Tue, Mar 23, 2021 at 02:58:28PM +0100, Hans de Goede wrote:
> >>> Hi,
> >>>
> >>> On 3/23/21 2:56 PM, Matti Vaittinen wrote:
> >>>> Devm helper header containing small inline helpers was added.
> >>>> Hans promised to maintain it.
> >>>>
> >>>> Add Hans as maintainer and myself as designated reviewer.
> >>>>
> >>> Ultimately this is up to Greg though, so lets wait and see what
> >>> Greg has to say about this.
> >>
> >> Can we move some of the devm_* calls in include/device.h into here as
> >> well so that you all can be in charge of them instead of me?
> >
> > Seems like this was left w/o answer. I guess the question was pointed
> > to Hans
>
> I believe that Greg was (mostly) joking here. At least that is how
> I interpreted Greg's reply,which is why I did not answer.
I have no idea what this thread was about anymore, sorry :)
> Also note that Greg merged this series, but not this patch,
> so the new devm-helpers.h file will presumably be maintained by Greg.
What's one more file...
thanks,
greg k-h
Hi,
On 4/21/21 2:17 PM, Vaittinen, Matti wrote:
>
> On Wed, 2021-04-21 at 13:58 +0200, Hans de Goede wrote:
>> Hi,
>>
>> On 4/21/21 9:51 AM, Matti Vaittinen wrote:
>>> On Tue, 2021-03-23 at 15:19 +0100, Greg KH wrote:
>>>> On Tue, Mar 23, 2021 at 02:58:28PM +0100, Hans de Goede wrote:
>>>>> Hi,
>>>>>
>>>>> On 3/23/21 2:56 PM, Matti Vaittinen wrote:
>>>>>> Devm helper header containing small inline helpers was added.
>>>>>> Hans promised to maintain it.
>>>>>>
>>>>>> Add Hans as maintainer and myself as designated reviewer.
>>>>>>
>>>>> Ultimately this is up to Greg though, so lets wait and see what
>>>>> Greg has to say about this.
>>>>
>>>> Can we move some of the devm_* calls in include/device.h into
>>>> here as
>>>> well so that you all can be in charge of them instead of me?
>>>
>>> Seems like this was left w/o answer. I guess the question was
>>> pointed
>>> to Hans
>>
>> I believe that Greg was (mostly) joking here. At least that is how
>> I interpreted Greg's reply,which is why I did not answer.
>
> Ah. I missed the sarcastic tone of typing. I should've noted that by
> the font :]
>
>> Also note that Greg merged this series, but not this patch,
>> so the new devm-helpers.h file will presumably be maintained by Greg.
>
> Hmm. Are you sure?
> https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core.git/commit/?h=driver-core-next&id=2077ca682169afb212d8a887c70057a660290df9
Ah, you're right.
I was looking at the wrong branch, sorry about the confusion.
Ok, so I guess I do maintain the new devm-helpers.h file, that is fine.
Which makes your email from earlier today more relevant:
> but what comes to my (not always so humble) opinion - most of
> the devm functions in device.h are tightly related to the device
> interface or devres. Thus the device.h feels like appropriate place for
> most of those.
I agree with you that most devm_ functions in device.h are probably
left there. Moving them will also mean modifying all the drivers
which use them to include the new devm-helpers.h include file
which seems like needless churn.
> OTOH, the kmalloc/kfree related functions, strdub and
> kmemdub might be candidates for move - those are not really "device
> things".
I'm certainly open to moving some functions to devm-helpers.h, but
also see above about needless churn.
Regards,
Hans
On Wed, 2021-04-21 at 14:09 +0200, Greg KH wrote:
> On Wed, Apr 21, 2021 at 01:58:29PM +0200, Hans de Goede wrote:
> > Hi,
> >
> > On 4/21/21 9:51 AM, Matti Vaittinen wrote:
> > > On Tue, 2021-03-23 at 15:19 +0100, Greg KH wrote:
> > > > On Tue, Mar 23, 2021 at 02:58:28PM +0100, Hans de Goede wrote:
> > > > > Hi,
> > > > >
> > > > > On 3/23/21 2:56 PM, Matti Vaittinen wrote:
> > > > > > Devm helper header containing small inline helpers was
> > > > > > added.
> > > > > > Hans promised to maintain it.
> > > > > >
> > > > > > Add Hans as maintainer and myself as designated reviewer.
> > > > > >
> > > > > Ultimately this is up to Greg though, so lets wait and see
> > > > > what
> > > > > Greg has to say about this.
> > > >
> > > > Can we move some of the devm_* calls in include/device.h into
> > > > here as
> > > > well so that you all can be in charge of them instead of me?
> > >
> > > Seems like this was left w/o answer. I guess the question was
> > > pointed
> > > to Hans
> >
> > I believe that Greg was (mostly) joking here. At least that is how
> > I interpreted Greg's reply,which is why I did not answer.
>
> I have no idea what this thread was about anymore, sorry :)
Need more B12 ;) Can't remember a minor discussion just a few thousand
other patch mails ago? Huh, that's what the age does, right :p
Well,
this is not urgent in any way but here's some context:
> This series implements delayed wq cancellation on top of devm and
> replaces
> the obvious cases where only thing remove call-back in a driver does
> is
> cancelling the work. There might be other cases where we could switch
> more than just work cancellation to use managed version and thus get
> rid
> of remove or mixed (manual and devm) resource management.
>
> The series introduces include/linux/devm-helpers.h file which
> hopefully works as a place where this kind of helpers can be inlined.
and your reply:
> Can we move some of the devm_* calls in include/device.h into here
> as well ...
I thought you meant that literally, Hans assumed you were joking - and
you don't remember how it was :] So - if no further suggestions, then
we keep the devm stuff which is currently in device.h still in
device.h. Although the malloc/free stuff is not really strictly device
thing.
Best Regards
Matti Vaittinen
On Wed, 2021-04-21 at 14:26 +0200, Hans de Goede wrote:
> Hi,
>
> On 4/21/21 2:17 PM, Vaittinen, Matti wrote:
> > On Wed, 2021-04-21 at 13:58 +0200, Hans de Goede wrote:
> > > Hi,
> > >
> > > On 4/21/21 9:51 AM, Matti Vaittinen wrote:
> > > > On Tue, 2021-03-23 at 15:19 +0100, Greg KH wrote:
> > > > > On Tue, Mar 23, 2021 at 02:58:28PM +0100, Hans de Goede
> > > > > wrote:
> > > > > > Hi,
> > > > > >
> > > > > > On 3/23/21 2:56 PM, Matti Vaittinen wrote:
> > > > > > > Devm helper header containing small inline helpers was
> > > > > > > added.
> > > > > > > Hans promised to maintain it.
> > > > > > >
> > > > > > > Add Hans as maintainer and myself as designated reviewer.
> > > > > > >
> > > > > > Ultimately this is up to Greg though, so lets wait and see
> > > > > > what
> > > > > > Greg has to say about this.
> > > > >
> > > > > Can we move some of the devm_* calls in include/device.h into
> > > > > here as
> > > > > well so that you all can be in charge of them instead of me?
...
> > but what comes to my (not always so humble) opinion - most of
> > the devm functions in device.h are tightly related to the device
> > interface or devres. Thus the device.h feels like appropriate place
> > for
> > most of those.
>
> I agree with you that most devm_ functions in device.h are probably
> left there. Moving them will also mean modifying all the drivers
> which use them to include the new devm-helpers.h include file
> which seems like needless churn.
>
> > OTOH, the kmalloc/kfree related functions, strdub and
> > kmemdub might be candidates for move - those are not really "device
> > things".
>
> I'm certainly open to moving some functions to devm-helpers.h, but
> also see above about needless churn.
I agree. Whether the move is worth all the hassle depends on how much
Greg _really_ wishes to get rid of those devm functions. Especially the
allocs are used _a lot_.
Best Regards
--Matti