2020-12-27 21:16:15

by Hans de Goede

[permalink] [raw]
Subject: [PATCH 06/14] extcon: arizona: Fix various races on driver unbind

We must free/disable all interrupts and cancel all pending works
before doing further cleanup.

Before this commit arizona_extcon_remove() was doing several
register writes to shut things down before disabling the IRQs
and it was cancelling only 1 of the 3 different works used.

Move all the register-writes shutting things down to after
the disabling of the IRQs and add the 2 missing
cancel_delayed_work_sync() calls.

This fixes various possible races on driver unbind. One of which
would always trigger on devices using the mic-clamp feature for
jack detection. The ARIZONA_MICD_CLAMP_MODE_MASK update was
done before disabling the IRQs, causing:
1. arizona_jackdet() to run
2. detect a jack being inserted (clamp disabled means jack inserted)
3. call arizona_start_mic() which:
3.1 Enables the MICVDD regulator
3.2 takes a pm_runtime_reference

And this was all happening after the ARIZONA_MICD_ENA bit clearing,
which would undo 3.1 and 3.2 because the ARIZONA_MICD_CLAMP_MODE_MASK
update was being done after the ARIZONA_MICD_ENA bit clearing.

So this means that arizona_extcon_remove() would exit with
1. MICVDD enabled and 2. The pm_runtime_reference being unbalanced.

MICVDD still being enabled caused the following oops when the
regulator is released by the devm framework:

[ 2850.745757] ------------[ cut here ]------------
[ 2850.745827] WARNING: CPU: 2 PID: 2098 at drivers/regulator/core.c:2123 _regulator_put.part.0+0x19f/0x1b0
[ 2850.745835] Modules linked in: extcon_arizona ...
...
[ 2850.746909] Call Trace:
[ 2850.746932] regulator_put+0x2d/0x40
[ 2850.746946] release_nodes+0x22a/0x260
[ 2850.746984] __device_release_driver+0x190/0x240
[ 2850.747002] driver_detach+0xd4/0x120
...
[ 2850.747337] ---[ end trace f455dfd7abd9781f ]---

Note this oops is just one of various theoretically possible races caused
by the wrong ordering inside arizona_extcon_remove(), this fixes the
ordering fixing all possible races, including the reported oops.

Signed-off-by: Hans de Goede <[email protected]>
---
drivers/extcon/extcon-arizona.c | 40 +++++++++++++++++----------------
1 file changed, 21 insertions(+), 19 deletions(-)

diff --git a/drivers/extcon/extcon-arizona.c b/drivers/extcon/extcon-arizona.c
index f7ef247de46a..76aacbac5869 100644
--- a/drivers/extcon/extcon-arizona.c
+++ b/drivers/extcon/extcon-arizona.c
@@ -1760,25 +1760,6 @@ static int arizona_extcon_remove(struct platform_device *pdev)
bool change;
int ret;

- ret = regmap_update_bits_check(arizona->regmap, ARIZONA_MIC_DETECT_1,
- ARIZONA_MICD_ENA, 0,
- &change);
- if (ret < 0) {
- dev_err(&pdev->dev, "Failed to disable micd on remove: %d\n",
- ret);
- } else if (change) {
- regulator_disable(info->micvdd);
- pm_runtime_put(info->dev);
- }
-
- gpiod_put(info->micd_pol_gpio);
-
- pm_runtime_disable(&pdev->dev);
-
- regmap_update_bits(arizona->regmap,
- ARIZONA_MICD_CLAMP_CONTROL,
- ARIZONA_MICD_CLAMP_MODE_MASK, 0);
-
if (info->micd_clamp) {
jack_irq_rise = ARIZONA_IRQ_MICD_CLAMP_RISE;
jack_irq_fall = ARIZONA_IRQ_MICD_CLAMP_FALL;
@@ -1794,10 +1775,31 @@ static int arizona_extcon_remove(struct platform_device *pdev)
arizona_free_irq(arizona, jack_irq_rise, info);
arizona_free_irq(arizona, jack_irq_fall, info);
cancel_delayed_work_sync(&info->hpdet_work);
+ cancel_delayed_work_sync(&info->micd_detect_work);
+ cancel_delayed_work_sync(&info->micd_timeout_work);
+
+ ret = regmap_update_bits_check(arizona->regmap, ARIZONA_MIC_DETECT_1,
+ ARIZONA_MICD_ENA, 0,
+ &change);
+ if (ret < 0) {
+ dev_err(&pdev->dev, "Failed to disable micd on remove: %d\n",
+ ret);
+ } else if (change) {
+ regulator_disable(info->micvdd);
+ pm_runtime_put(info->dev);
+ }
+
+ regmap_update_bits(arizona->regmap,
+ ARIZONA_MICD_CLAMP_CONTROL,
+ ARIZONA_MICD_CLAMP_MODE_MASK, 0);
regmap_update_bits(arizona->regmap, ARIZONA_JACK_DETECT_ANALOGUE,
ARIZONA_JD1_ENA, 0);
arizona_clk32k_disable(arizona);

+ gpiod_put(info->micd_pol_gpio);
+
+ pm_runtime_disable(&pdev->dev);
+
return 0;
}

--
2.28.0


2020-12-29 12:14:35

by Charles Keepax

[permalink] [raw]
Subject: Re: [PATCH 06/14] extcon: arizona: Fix various races on driver unbind

On Sun, Dec 27, 2020 at 10:12:24PM +0100, Hans de Goede wrote:
> We must free/disable all interrupts and cancel all pending works
> before doing further cleanup.
>
> Before this commit arizona_extcon_remove() was doing several
> register writes to shut things down before disabling the IRQs
> and it was cancelling only 1 of the 3 different works used.
>
> Move all the register-writes shutting things down to after
> the disabling of the IRQs and add the 2 missing
> cancel_delayed_work_sync() calls.
>
> This fixes various possible races on driver unbind. One of which
> would always trigger on devices using the mic-clamp feature for
> jack detection. The ARIZONA_MICD_CLAMP_MODE_MASK update was
> done before disabling the IRQs, causing:
> 1. arizona_jackdet() to run
> 2. detect a jack being inserted (clamp disabled means jack inserted)
> 3. call arizona_start_mic() which:
> 3.1 Enables the MICVDD regulator
> 3.2 takes a pm_runtime_reference
>
> And this was all happening after the ARIZONA_MICD_ENA bit clearing,
> which would undo 3.1 and 3.2 because the ARIZONA_MICD_CLAMP_MODE_MASK
> update was being done after the ARIZONA_MICD_ENA bit clearing.
>
> So this means that arizona_extcon_remove() would exit with
> 1. MICVDD enabled and 2. The pm_runtime_reference being unbalanced.
>
> MICVDD still being enabled caused the following oops when the
> regulator is released by the devm framework:
>
> [ 2850.745757] ------------[ cut here ]------------
> [ 2850.745827] WARNING: CPU: 2 PID: 2098 at drivers/regulator/core.c:2123 _regulator_put.part.0+0x19f/0x1b0
> [ 2850.745835] Modules linked in: extcon_arizona ...
> ...
> [ 2850.746909] Call Trace:
> [ 2850.746932] regulator_put+0x2d/0x40
> [ 2850.746946] release_nodes+0x22a/0x260
> [ 2850.746984] __device_release_driver+0x190/0x240
> [ 2850.747002] driver_detach+0xd4/0x120
> ...
> [ 2850.747337] ---[ end trace f455dfd7abd9781f ]---
>
> Note this oops is just one of various theoretically possible races caused
> by the wrong ordering inside arizona_extcon_remove(), this fixes the
> ordering fixing all possible races, including the reported oops.
>
> Signed-off-by: Hans de Goede <[email protected]>
> ---

Sorry yes there are a few rough corners on the extcon stuff, I
have clean up series I have been working on as part of
upstreaming the Madera extcon hopefully I will get that sent out
one day.

Acked-by: Charles Keepax <[email protected]>

Thanks,
Charles