* [PATCH 0/2] drivers/thermal/rcar_gen3_thermal: Fix device initialization @ 2023-02-07 17:10 Niklas Söderlund 2023-02-07 17:10 ` [PATCH 1/2] drivers/thermal/rcar_gen3_thermal: Create device local ops struct Niklas Söderlund 2023-02-07 17:10 ` [PATCH 2/2] drivers/thermal/rcar_gen3_thermal: Fix device initialization Niklas Söderlund 0 siblings, 2 replies; 9+ messages in thread From: Niklas Söderlund @ 2023-02-07 17:10 UTC (permalink / raw) To: Daniel Lezcano, Wolfram Sang, linux-pm Cc: linux-renesas-soc, Niklas Söderlund Hello, This small series fixes a window where incorrect values can be read from the driver before it is fully initialized. The root cause is that the thermal zone is register too early. Patch 1/2 prepares for the change while also fixing a theoretical issue where one thermal node described in DT would describe interrupts and another would not. Resulting in interrupt support being disabled for both of them. I'm not aware of any case where this configuration would be used, either the SoC supports interrupts, or it don't. While patch 2/2 fixes the real issue by fully initializing the device before registering the zone. Niklas Söderlund (2): drivers/thermal/rcar_gen3_thermal: Create device local ops struct drivers/thermal/rcar_gen3_thermal: Fix device initialization drivers/thermal/rcar_gen3_thermal.c | 34 ++++++++++++++++------------- 1 file changed, 19 insertions(+), 15 deletions(-) -- 2.39.1 ^ permalink raw reply [flat|nested] 9+ messages in thread
* [PATCH 1/2] drivers/thermal/rcar_gen3_thermal: Create device local ops struct 2023-02-07 17:10 [PATCH 0/2] drivers/thermal/rcar_gen3_thermal: Fix device initialization Niklas Söderlund @ 2023-02-07 17:10 ` Niklas Söderlund 2023-02-08 7:56 ` Wolfram Sang 2023-02-07 17:10 ` [PATCH 2/2] drivers/thermal/rcar_gen3_thermal: Fix device initialization Niklas Söderlund 1 sibling, 1 reply; 9+ messages in thread From: Niklas Söderlund @ 2023-02-07 17:10 UTC (permalink / raw) To: Daniel Lezcano, Wolfram Sang, linux-pm Cc: linux-renesas-soc, Niklas Söderlund The callback operations are modified on a driver global level. If one device tree description do not define interrupts, the set_trips() operation was disabled globally for all users of the driver. Fix this by creating a device local copy of the operations structure and modify the copy depending on what the device can do. Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se> --- drivers/thermal/rcar_gen3_thermal.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/thermal/rcar_gen3_thermal.c b/drivers/thermal/rcar_gen3_thermal.c index 4c1c6f89aa2f..bfa2ff20b945 100644 --- a/drivers/thermal/rcar_gen3_thermal.c +++ b/drivers/thermal/rcar_gen3_thermal.c @@ -87,6 +87,7 @@ struct rcar_gen3_thermal_tsc { struct rcar_gen3_thermal_priv { struct rcar_gen3_thermal_tsc *tscs[TSC_MAX_NUM]; + struct thermal_zone_device_ops ops; unsigned int num_tscs; void (*thermal_init)(struct rcar_gen3_thermal_tsc *tsc); int ptat[3]; @@ -225,7 +226,7 @@ static int rcar_gen3_thermal_set_trips(struct thermal_zone_device *tz, int low, return 0; } -static struct thermal_zone_device_ops rcar_gen3_tz_of_ops = { +static const struct thermal_zone_device_ops rcar_gen3_tz_of_ops = { .get_temp = rcar_gen3_thermal_get_temp, .set_trips = rcar_gen3_thermal_set_trips, }; @@ -466,6 +467,7 @@ static int rcar_gen3_thermal_probe(struct platform_device *pdev) if (!priv) return -ENOMEM; + priv->ops = rcar_gen3_tz_of_ops; priv->thermal_init = rcar_gen3_thermal_init; if (soc_device_match(r8a7795es1)) priv->thermal_init = rcar_gen3_thermal_init_r8a7795es1; @@ -473,7 +475,7 @@ static int rcar_gen3_thermal_probe(struct platform_device *pdev) platform_set_drvdata(pdev, priv); if (rcar_gen3_thermal_request_irqs(priv, pdev)) - rcar_gen3_tz_of_ops.set_trips = NULL; + priv->ops.set_trips = NULL; pm_runtime_enable(dev); pm_runtime_get_sync(dev); @@ -508,8 +510,7 @@ static int rcar_gen3_thermal_probe(struct platform_device *pdev) for (i = 0; i < priv->num_tscs; i++) { struct rcar_gen3_thermal_tsc *tsc = priv->tscs[i]; - zone = devm_thermal_of_zone_register(dev, i, tsc, - &rcar_gen3_tz_of_ops); + zone = devm_thermal_of_zone_register(dev, i, tsc, &priv->ops); if (IS_ERR(zone)) { dev_err(dev, "Sensor %u: Can't register thermal zone\n", i); ret = PTR_ERR(zone); -- 2.39.1 ^ permalink raw reply related [flat|nested] 9+ messages in thread
* Re: [PATCH 1/2] drivers/thermal/rcar_gen3_thermal: Create device local ops struct 2023-02-07 17:10 ` [PATCH 1/2] drivers/thermal/rcar_gen3_thermal: Create device local ops struct Niklas Söderlund @ 2023-02-08 7:56 ` Wolfram Sang 0 siblings, 0 replies; 9+ messages in thread From: Wolfram Sang @ 2023-02-08 7:56 UTC (permalink / raw) To: Niklas Söderlund; +Cc: Daniel Lezcano, linux-pm, linux-renesas-soc [-- Attachment #1: Type: text/plain, Size: 572 bytes --] On Tue, Feb 07, 2023 at 06:10:10PM +0100, Niklas Söderlund wrote: > The callback operations are modified on a driver global level. If one > device tree description do not define interrupts, the set_trips() > operation was disabled globally for all users of the driver. > > Fix this by creating a device local copy of the operations structure and > modify the copy depending on what the device can do. > > Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se> I like this! Reviewed-by: Wolfram Sang <wsa+renesas@sang-engineering.com> [-- Attachment #2: signature.asc --] [-- Type: application/pgp-signature, Size: 833 bytes --] ^ permalink raw reply [flat|nested] 9+ messages in thread
* [PATCH 2/2] drivers/thermal/rcar_gen3_thermal: Fix device initialization 2023-02-07 17:10 [PATCH 0/2] drivers/thermal/rcar_gen3_thermal: Fix device initialization Niklas Söderlund 2023-02-07 17:10 ` [PATCH 1/2] drivers/thermal/rcar_gen3_thermal: Create device local ops struct Niklas Söderlund @ 2023-02-07 17:10 ` Niklas Söderlund 2023-02-08 7:58 ` Wolfram Sang 2023-02-08 11:06 ` Daniel Lezcano 1 sibling, 2 replies; 9+ messages in thread From: Niklas Söderlund @ 2023-02-07 17:10 UTC (permalink / raw) To: Daniel Lezcano, Wolfram Sang, linux-pm Cc: linux-renesas-soc, Niklas Söderlund The thermal zone is registered before the device is register and the thermal coefficients are calculated, providing a window for very incorrect readings. The reason why the zone was register before the device was fully initialized was that the presence of the set_trips() callback is used to determine if the driver supports interrupt or not, as it is not defined if the device is incapable of interrupts. Fix this by using the operations structure in the private data instead of the zone to determine if interrupts are available or not, and initialize the device before registering the zone. Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se> --- drivers/thermal/rcar_gen3_thermal.c | 25 ++++++++++++++----------- 1 file changed, 14 insertions(+), 11 deletions(-) diff --git a/drivers/thermal/rcar_gen3_thermal.c b/drivers/thermal/rcar_gen3_thermal.c index bfa2ff20b945..1dedeece1a00 100644 --- a/drivers/thermal/rcar_gen3_thermal.c +++ b/drivers/thermal/rcar_gen3_thermal.c @@ -89,7 +89,8 @@ struct rcar_gen3_thermal_priv { struct rcar_gen3_thermal_tsc *tscs[TSC_MAX_NUM]; struct thermal_zone_device_ops ops; unsigned int num_tscs; - void (*thermal_init)(struct rcar_gen3_thermal_tsc *tsc); + void (*thermal_init)(struct rcar_gen3_thermal_priv *priv, + struct rcar_gen3_thermal_tsc *tsc); int ptat[3]; }; @@ -240,7 +241,7 @@ static irqreturn_t rcar_gen3_thermal_irq(int irq, void *data) for (i = 0; i < priv->num_tscs; i++) { status = rcar_gen3_thermal_read(priv->tscs[i], REG_GEN3_IRQSTR); rcar_gen3_thermal_write(priv->tscs[i], REG_GEN3_IRQSTR, 0); - if (status) + if (status && priv->tscs[i]->zone) thermal_zone_device_update(priv->tscs[i]->zone, THERMAL_EVENT_UNSPECIFIED); } @@ -311,7 +312,8 @@ static bool rcar_gen3_thermal_read_fuses(struct rcar_gen3_thermal_priv *priv) return true; } -static void rcar_gen3_thermal_init_r8a7795es1(struct rcar_gen3_thermal_tsc *tsc) +static void rcar_gen3_thermal_init_r8a7795es1(struct rcar_gen3_thermal_priv *priv, + struct rcar_gen3_thermal_tsc *tsc) { rcar_gen3_thermal_write(tsc, REG_GEN3_CTSR, CTSR_THBGR); rcar_gen3_thermal_write(tsc, REG_GEN3_CTSR, 0x0); @@ -322,7 +324,7 @@ static void rcar_gen3_thermal_init_r8a7795es1(struct rcar_gen3_thermal_tsc *tsc) rcar_gen3_thermal_write(tsc, REG_GEN3_IRQCTL, 0x3F); rcar_gen3_thermal_write(tsc, REG_GEN3_IRQMSK, 0); - if (tsc->zone->ops->set_trips) + if (priv->ops.set_trips) rcar_gen3_thermal_write(tsc, REG_GEN3_IRQEN, IRQ_TEMPD1 | IRQ_TEMP2); @@ -338,7 +340,8 @@ static void rcar_gen3_thermal_init_r8a7795es1(struct rcar_gen3_thermal_tsc *tsc) usleep_range(1000, 2000); } -static void rcar_gen3_thermal_init(struct rcar_gen3_thermal_tsc *tsc) +static void rcar_gen3_thermal_init(struct rcar_gen3_thermal_priv *priv, + struct rcar_gen3_thermal_tsc *tsc) { u32 reg_val; @@ -350,7 +353,7 @@ static void rcar_gen3_thermal_init(struct rcar_gen3_thermal_tsc *tsc) rcar_gen3_thermal_write(tsc, REG_GEN3_IRQCTL, 0); rcar_gen3_thermal_write(tsc, REG_GEN3_IRQMSK, 0); - if (tsc->zone->ops->set_trips) + if (priv->ops.set_trips) rcar_gen3_thermal_write(tsc, REG_GEN3_IRQEN, IRQ_TEMPD1 | IRQ_TEMP2); @@ -510,6 +513,9 @@ static int rcar_gen3_thermal_probe(struct platform_device *pdev) for (i = 0; i < priv->num_tscs; i++) { struct rcar_gen3_thermal_tsc *tsc = priv->tscs[i]; + priv->thermal_init(priv, tsc); + rcar_gen3_thermal_calc_coefs(priv, tsc, *ths_tj_1); + zone = devm_thermal_of_zone_register(dev, i, tsc, &priv->ops); if (IS_ERR(zone)) { dev_err(dev, "Sensor %u: Can't register thermal zone\n", i); @@ -518,9 +524,6 @@ static int rcar_gen3_thermal_probe(struct platform_device *pdev) } tsc->zone = zone; - priv->thermal_init(tsc); - rcar_gen3_thermal_calc_coefs(priv, tsc, *ths_tj_1); - tsc->zone->tzp->no_hwmon = false; ret = thermal_add_hwmon_sysfs(tsc->zone); if (ret) @@ -559,8 +562,8 @@ static int __maybe_unused rcar_gen3_thermal_resume(struct device *dev) struct rcar_gen3_thermal_tsc *tsc = priv->tscs[i]; struct thermal_zone_device *zone = tsc->zone; - priv->thermal_init(tsc); - if (zone->ops->set_trips) + priv->thermal_init(priv, tsc); + if (priv->ops.set_trips) rcar_gen3_thermal_set_trips(zone, zone->prev_low_trip, zone->prev_high_trip); } -- 2.39.1 ^ permalink raw reply related [flat|nested] 9+ messages in thread
* Re: [PATCH 2/2] drivers/thermal/rcar_gen3_thermal: Fix device initialization 2023-02-07 17:10 ` [PATCH 2/2] drivers/thermal/rcar_gen3_thermal: Fix device initialization Niklas Söderlund @ 2023-02-08 7:58 ` Wolfram Sang 2023-02-08 10:12 ` Niklas Söderlund 2023-02-08 11:06 ` Daniel Lezcano 1 sibling, 1 reply; 9+ messages in thread From: Wolfram Sang @ 2023-02-08 7:58 UTC (permalink / raw) To: Niklas Söderlund; +Cc: Daniel Lezcano, linux-pm, linux-renesas-soc [-- Attachment #1: Type: text/plain, Size: 1151 bytes --] On Tue, Feb 07, 2023 at 06:10:11PM +0100, Niklas Söderlund wrote: > The thermal zone is registered before the device is register and the > thermal coefficients are calculated, providing a window for very > incorrect readings. While I could never actually be in this race window, the patch makes a lot of sense to me. > The reason why the zone was register before the device was fully > initialized was that the presence of the set_trips() callback is used to > determine if the driver supports interrupt or not, as it is not defined > if the device is incapable of interrupts. > > Fix this by using the operations structure in the private data instead > of the zone to determine if interrupts are available or not, and > initialize the device before registering the zone. > > Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se> Reviewed-by: Wolfram Sang <wsa+renesas@sang-engineering.com> One minor question: > - if (status) > + if (status && priv->tscs[i]->zone) > thermal_zone_device_update(priv->tscs[i]->zone, > THERMAL_EVENT_UNSPECIFIED); Isn't this a seperate change? [-- Attachment #2: signature.asc --] [-- Type: application/pgp-signature, Size: 833 bytes --] ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH 2/2] drivers/thermal/rcar_gen3_thermal: Fix device initialization 2023-02-08 7:58 ` Wolfram Sang @ 2023-02-08 10:12 ` Niklas Söderlund 2023-02-08 14:14 ` Wolfram Sang 0 siblings, 1 reply; 9+ messages in thread From: Niklas Söderlund @ 2023-02-08 10:12 UTC (permalink / raw) To: Wolfram Sang, Daniel Lezcano, linux-pm, linux-renesas-soc Hi Wolfram, Thanks for your review. On 2023-02-08 08:58:33 +0100, Wolfram Sang wrote: > On Tue, Feb 07, 2023 at 06:10:11PM +0100, Niklas Söderlund wrote: > > The thermal zone is registered before the device is register and the > > thermal coefficients are calculated, providing a window for very > > incorrect readings. > > While I could never actually be in this race window, the patch makes a > lot of sense to me. > > > The reason why the zone was register before the device was fully > > initialized was that the presence of the set_trips() callback is used to > > determine if the driver supports interrupt or not, as it is not defined > > if the device is incapable of interrupts. > > > > Fix this by using the operations structure in the private data instead > > of the zone to determine if interrupts are available or not, and > > initialize the device before registering the zone. > > > > Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se> > > Reviewed-by: Wolfram Sang <wsa+renesas@sang-engineering.com> > > One minor question: > > > - if (status) > > + if (status && priv->tscs[i]->zone) > > thermal_zone_device_update(priv->tscs[i]->zone, > > THERMAL_EVENT_UNSPECIFIED); > > Isn't this a seperate change? Not really. This patch changes the driver to enable interrupts just before the zone is created. So this just guards from the very small window where an interrupt could be fired and the zone is created, which if it would happen would cause a null pointer dereference. While not updating a not yet created zone is fine. But this should never happen (tm) as the trip points have not been set when this window exists, but better safe then sorry. -- Kind Regards, Niklas Söderlund ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH 2/2] drivers/thermal/rcar_gen3_thermal: Fix device initialization 2023-02-08 10:12 ` Niklas Söderlund @ 2023-02-08 14:14 ` Wolfram Sang 0 siblings, 0 replies; 9+ messages in thread From: Wolfram Sang @ 2023-02-08 14:14 UTC (permalink / raw) To: Niklas Söderlund; +Cc: Daniel Lezcano, linux-pm, linux-renesas-soc [-- Attachment #1: Type: text/plain, Size: 228 bytes --] > While not updating a not yet created zone is fine. But this should never > happen (tm) as the trip points have not been set when this window > exists, but better safe then sorry. Full Ack! Thanks for the heads up. [-- Attachment #2: signature.asc --] [-- Type: application/pgp-signature, Size: 833 bytes --] ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH 2/2] drivers/thermal/rcar_gen3_thermal: Fix device initialization 2023-02-07 17:10 ` [PATCH 2/2] drivers/thermal/rcar_gen3_thermal: Fix device initialization Niklas Söderlund 2023-02-08 7:58 ` Wolfram Sang @ 2023-02-08 11:06 ` Daniel Lezcano 2023-02-08 18:44 ` Niklas Söderlund 1 sibling, 1 reply; 9+ messages in thread From: Daniel Lezcano @ 2023-02-08 11:06 UTC (permalink / raw) To: Niklas Söderlund, Wolfram Sang, linux-pm; +Cc: linux-renesas-soc On 07/02/2023 18:10, Niklas Söderlund wrote: > The thermal zone is registered before the device is register and the > thermal coefficients are calculated, providing a window for very > incorrect readings. > > The reason why the zone was register before the device was fully > initialized was that the presence of the set_trips() callback is used to > determine if the driver supports interrupt or not, as it is not defined > if the device is incapable of interrupts. > > Fix this by using the operations structure in the private data instead > of the zone to determine if interrupts are available or not, and > initialize the device before registering the zone. > > Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se> > --- > drivers/thermal/rcar_gen3_thermal.c | 25 ++++++++++++++----------- > 1 file changed, 14 insertions(+), 11 deletions(-) > > diff --git a/drivers/thermal/rcar_gen3_thermal.c b/drivers/thermal/rcar_gen3_thermal.c > index bfa2ff20b945..1dedeece1a00 100644 > --- a/drivers/thermal/rcar_gen3_thermal.c > +++ b/drivers/thermal/rcar_gen3_thermal.c > @@ -89,7 +89,8 @@ struct rcar_gen3_thermal_priv { > struct rcar_gen3_thermal_tsc *tscs[TSC_MAX_NUM]; > struct thermal_zone_device_ops ops; > unsigned int num_tscs; > - void (*thermal_init)(struct rcar_gen3_thermal_tsc *tsc); > + void (*thermal_init)(struct rcar_gen3_thermal_priv *priv, > + struct rcar_gen3_thermal_tsc *tsc); > int ptat[3]; > }; > > @@ -240,7 +241,7 @@ static irqreturn_t rcar_gen3_thermal_irq(int irq, void *data) > for (i = 0; i < priv->num_tscs; i++) { > status = rcar_gen3_thermal_read(priv->tscs[i], REG_GEN3_IRQSTR); > rcar_gen3_thermal_write(priv->tscs[i], REG_GEN3_IRQSTR, 0); > - if (status) > + if (status && priv->tscs[i]->zone) > thermal_zone_device_update(priv->tscs[i]->zone, > THERMAL_EVENT_UNSPECIFIED); > } > @@ -311,7 +312,8 @@ static bool rcar_gen3_thermal_read_fuses(struct rcar_gen3_thermal_priv *priv) > return true; > } > > -static void rcar_gen3_thermal_init_r8a7795es1(struct rcar_gen3_thermal_tsc *tsc) > +static void rcar_gen3_thermal_init_r8a7795es1(struct rcar_gen3_thermal_priv *priv, > + struct rcar_gen3_thermal_tsc *tsc) > { > rcar_gen3_thermal_write(tsc, REG_GEN3_CTSR, CTSR_THBGR); > rcar_gen3_thermal_write(tsc, REG_GEN3_CTSR, 0x0); > @@ -322,7 +324,7 @@ static void rcar_gen3_thermal_init_r8a7795es1(struct rcar_gen3_thermal_tsc *tsc) > > rcar_gen3_thermal_write(tsc, REG_GEN3_IRQCTL, 0x3F); > rcar_gen3_thermal_write(tsc, REG_GEN3_IRQMSK, 0); > - if (tsc->zone->ops->set_trips) > + if (priv->ops.set_trips) > rcar_gen3_thermal_write(tsc, REG_GEN3_IRQEN, > IRQ_TEMPD1 | IRQ_TEMP2); > > @@ -338,7 +340,8 @@ static void rcar_gen3_thermal_init_r8a7795es1(struct rcar_gen3_thermal_tsc *tsc) > usleep_range(1000, 2000); > } > > -static void rcar_gen3_thermal_init(struct rcar_gen3_thermal_tsc *tsc) > +static void rcar_gen3_thermal_init(struct rcar_gen3_thermal_priv *priv, > + struct rcar_gen3_thermal_tsc *tsc) > { > u32 reg_val; > > @@ -350,7 +353,7 @@ static void rcar_gen3_thermal_init(struct rcar_gen3_thermal_tsc *tsc) > > rcar_gen3_thermal_write(tsc, REG_GEN3_IRQCTL, 0); > rcar_gen3_thermal_write(tsc, REG_GEN3_IRQMSK, 0); > - if (tsc->zone->ops->set_trips) > + if (priv->ops.set_trips) > rcar_gen3_thermal_write(tsc, REG_GEN3_IRQEN, > IRQ_TEMPD1 | IRQ_TEMP2); > > @@ -510,6 +513,9 @@ static int rcar_gen3_thermal_probe(struct platform_device *pdev) > for (i = 0; i < priv->num_tscs; i++) { > struct rcar_gen3_thermal_tsc *tsc = priv->tscs[i]; > > + priv->thermal_init(priv, tsc); > + rcar_gen3_thermal_calc_coefs(priv, tsc, *ths_tj_1); > + > zone = devm_thermal_of_zone_register(dev, i, tsc, &priv->ops); > if (IS_ERR(zone)) { > dev_err(dev, "Sensor %u: Can't register thermal zone\n", i); > @@ -518,9 +524,6 @@ static int rcar_gen3_thermal_probe(struct platform_device *pdev) > } > tsc->zone = zone; > > - priv->thermal_init(tsc); > - rcar_gen3_thermal_calc_coefs(priv, tsc, *ths_tj_1); > - > tsc->zone->tzp->no_hwmon = false; > ret = thermal_add_hwmon_sysfs(tsc->zone); > if (ret) > @@ -559,8 +562,8 @@ static int __maybe_unused rcar_gen3_thermal_resume(struct device *dev) > struct rcar_gen3_thermal_tsc *tsc = priv->tscs[i]; > struct thermal_zone_device *zone = tsc->zone; > > - priv->thermal_init(tsc); > - if (zone->ops->set_trips) > + priv->thermal_init(priv, tsc); > + if (priv->ops.set_trips) > rcar_gen3_thermal_set_trips(zone, zone->prev_low_trip, > zone->prev_high_trip); This is not needed, at resume time, the thermal framework has a pm_notifier and calls thermal_zone_device_update() which in turn calls thermal_zone_set_trips(). If the ops is not set, it will continue. Actually, no call to set_trips should happen in the driver, just pass the ops the thermal framework and it will do the actions. The same happens when you call thermal_zone_device_register(), it calls thermal_zone_device_update(), then thermal_zone_set_trips(). -- <http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs Follow Linaro: <http://www.facebook.com/pages/Linaro> Facebook | <http://twitter.com/#!/linaroorg> Twitter | <http://www.linaro.org/linaro-blog/> Blog ^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH 2/2] drivers/thermal/rcar_gen3_thermal: Fix device initialization 2023-02-08 11:06 ` Daniel Lezcano @ 2023-02-08 18:44 ` Niklas Söderlund 0 siblings, 0 replies; 9+ messages in thread From: Niklas Söderlund @ 2023-02-08 18:44 UTC (permalink / raw) To: Daniel Lezcano; +Cc: Wolfram Sang, linux-pm, linux-renesas-soc Hi Daniel, Thanks for your feedback. On 2023-02-08 12:06:37 +0100, Daniel Lezcano wrote: > On 07/02/2023 18:10, Niklas Söderlund wrote: > > The thermal zone is registered before the device is register and the > > thermal coefficients are calculated, providing a window for very > > incorrect readings. > > > > The reason why the zone was register before the device was fully > > initialized was that the presence of the set_trips() callback is used to > > determine if the driver supports interrupt or not, as it is not defined > > if the device is incapable of interrupts. > > > > Fix this by using the operations structure in the private data instead > > of the zone to determine if interrupts are available or not, and > > initialize the device before registering the zone. > > > > Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se> > > --- > > drivers/thermal/rcar_gen3_thermal.c | 25 ++++++++++++++----------- > > 1 file changed, 14 insertions(+), 11 deletions(-) > > > > diff --git a/drivers/thermal/rcar_gen3_thermal.c b/drivers/thermal/rcar_gen3_thermal.c > > index bfa2ff20b945..1dedeece1a00 100644 > > --- a/drivers/thermal/rcar_gen3_thermal.c > > +++ b/drivers/thermal/rcar_gen3_thermal.c > > @@ -89,7 +89,8 @@ struct rcar_gen3_thermal_priv { > > struct rcar_gen3_thermal_tsc *tscs[TSC_MAX_NUM]; > > struct thermal_zone_device_ops ops; > > unsigned int num_tscs; > > - void (*thermal_init)(struct rcar_gen3_thermal_tsc *tsc); > > + void (*thermal_init)(struct rcar_gen3_thermal_priv *priv, > > + struct rcar_gen3_thermal_tsc *tsc); > > int ptat[3]; > > }; > > @@ -240,7 +241,7 @@ static irqreturn_t rcar_gen3_thermal_irq(int irq, void *data) > > for (i = 0; i < priv->num_tscs; i++) { > > status = rcar_gen3_thermal_read(priv->tscs[i], REG_GEN3_IRQSTR); > > rcar_gen3_thermal_write(priv->tscs[i], REG_GEN3_IRQSTR, 0); > > - if (status) > > + if (status && priv->tscs[i]->zone) > > thermal_zone_device_update(priv->tscs[i]->zone, > > THERMAL_EVENT_UNSPECIFIED); > > } > > @@ -311,7 +312,8 @@ static bool rcar_gen3_thermal_read_fuses(struct rcar_gen3_thermal_priv *priv) > > return true; > > } > > -static void rcar_gen3_thermal_init_r8a7795es1(struct rcar_gen3_thermal_tsc *tsc) > > +static void rcar_gen3_thermal_init_r8a7795es1(struct rcar_gen3_thermal_priv *priv, > > + struct rcar_gen3_thermal_tsc *tsc) > > { > > rcar_gen3_thermal_write(tsc, REG_GEN3_CTSR, CTSR_THBGR); > > rcar_gen3_thermal_write(tsc, REG_GEN3_CTSR, 0x0); > > @@ -322,7 +324,7 @@ static void rcar_gen3_thermal_init_r8a7795es1(struct rcar_gen3_thermal_tsc *tsc) > > rcar_gen3_thermal_write(tsc, REG_GEN3_IRQCTL, 0x3F); > > rcar_gen3_thermal_write(tsc, REG_GEN3_IRQMSK, 0); > > - if (tsc->zone->ops->set_trips) > > + if (priv->ops.set_trips) > > rcar_gen3_thermal_write(tsc, REG_GEN3_IRQEN, > > IRQ_TEMPD1 | IRQ_TEMP2); > > @@ -338,7 +340,8 @@ static void rcar_gen3_thermal_init_r8a7795es1(struct rcar_gen3_thermal_tsc *tsc) > > usleep_range(1000, 2000); > > } > > -static void rcar_gen3_thermal_init(struct rcar_gen3_thermal_tsc *tsc) > > +static void rcar_gen3_thermal_init(struct rcar_gen3_thermal_priv *priv, > > + struct rcar_gen3_thermal_tsc *tsc) > > { > > u32 reg_val; > > @@ -350,7 +353,7 @@ static void rcar_gen3_thermal_init(struct rcar_gen3_thermal_tsc *tsc) > > rcar_gen3_thermal_write(tsc, REG_GEN3_IRQCTL, 0); > > rcar_gen3_thermal_write(tsc, REG_GEN3_IRQMSK, 0); > > - if (tsc->zone->ops->set_trips) > > + if (priv->ops.set_trips) > > rcar_gen3_thermal_write(tsc, REG_GEN3_IRQEN, > > IRQ_TEMPD1 | IRQ_TEMP2); > > @@ -510,6 +513,9 @@ static int rcar_gen3_thermal_probe(struct platform_device *pdev) > > for (i = 0; i < priv->num_tscs; i++) { > > struct rcar_gen3_thermal_tsc *tsc = priv->tscs[i]; > > + priv->thermal_init(priv, tsc); > > + rcar_gen3_thermal_calc_coefs(priv, tsc, *ths_tj_1); > > + > > zone = devm_thermal_of_zone_register(dev, i, tsc, &priv->ops); > > if (IS_ERR(zone)) { > > dev_err(dev, "Sensor %u: Can't register thermal zone\n", i); > > @@ -518,9 +524,6 @@ static int rcar_gen3_thermal_probe(struct platform_device *pdev) > > } > > tsc->zone = zone; > > - priv->thermal_init(tsc); > > - rcar_gen3_thermal_calc_coefs(priv, tsc, *ths_tj_1); > > - > > tsc->zone->tzp->no_hwmon = false; > > ret = thermal_add_hwmon_sysfs(tsc->zone); > > if (ret) > > @@ -559,8 +562,8 @@ static int __maybe_unused rcar_gen3_thermal_resume(struct device *dev) > > struct rcar_gen3_thermal_tsc *tsc = priv->tscs[i]; > > struct thermal_zone_device *zone = tsc->zone; > > - priv->thermal_init(tsc); > > - if (zone->ops->set_trips) > > + priv->thermal_init(priv, tsc); > > + if (priv->ops.set_trips) > > rcar_gen3_thermal_set_trips(zone, zone->prev_low_trip, > > zone->prev_high_trip); > > This is not needed, at resume time, the thermal framework has a pm_notifier > and calls thermal_zone_device_update() which in turn calls > thermal_zone_set_trips(). If the ops is not set, it will continue. > > Actually, no call to set_trips should happen in the driver, just pass the > ops the thermal framework and it will do the actions. > > The same happens when you call thermal_zone_device_register(), it calls > thermal_zone_device_update(), then thermal_zone_set_trips(). I will send a v2 of this series addressing this before fixing the issue addressed in this patch. > > > -- > <http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs > > Follow Linaro: <http://www.facebook.com/pages/Linaro> Facebook | > <http://twitter.com/#!/linaroorg> Twitter | > <http://www.linaro.org/linaro-blog/> Blog > -- Kind Regards, Niklas Söderlund ^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2023-02-08 18:45 UTC | newest] Thread overview: 9+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2023-02-07 17:10 [PATCH 0/2] drivers/thermal/rcar_gen3_thermal: Fix device initialization Niklas Söderlund 2023-02-07 17:10 ` [PATCH 1/2] drivers/thermal/rcar_gen3_thermal: Create device local ops struct Niklas Söderlund 2023-02-08 7:56 ` Wolfram Sang 2023-02-07 17:10 ` [PATCH 2/2] drivers/thermal/rcar_gen3_thermal: Fix device initialization Niklas Söderlund 2023-02-08 7:58 ` Wolfram Sang 2023-02-08 10:12 ` Niklas Söderlund 2023-02-08 14:14 ` Wolfram Sang 2023-02-08 11:06 ` Daniel Lezcano 2023-02-08 18:44 ` Niklas Söderlund
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox