2021-11-01 15:14:52

by Andy Shevchenko

Subject: [PATCH v1 1/1] mfd: intel-lpss: Fix too early PM enablement in the ACPI ->probe()

The runtime PM callback may be called as soon as the runtime PM facility
is enabled and activated. It means that ->suspend() may be called before
we finish probing the device in the ACPI case. Hence, NULL pointer
dereference:

intel-lpss INT34BA:00: IRQ index 0 not found
BUG: kernel NULL pointer dereference, address: 0000000000000030
...
Workqueue: pm pm_runtime_work
RIP: 0010:intel_lpss_suspend+0xb/0x40 [intel_lpss]

To fix this, first try to register the device and only after that enable
runtime PM facility.

Fixes: 4b45efe85263 ("mfd: Add support for Intel Sunrisepoint LPSS devices")
Reported-by: Orlando Chamberlain <[email protected]>
Reported-by: Aditya Garg <[email protected]>
Signed-off-by: Andy Shevchenko <[email protected]>
---
drivers/mfd/intel-lpss-acpi.c | 7 ++++++-
1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/mfd/intel-lpss-acpi.c b/drivers/mfd/intel-lpss-acpi.c
index 3f1d976eb67c..be81507afb5e 100644
--- a/drivers/mfd/intel-lpss-acpi.c
+++ b/drivers/mfd/intel-lpss-acpi.c
@@ -136,6 +136,7 @@ static int intel_lpss_acpi_probe(struct platform_device *pdev)
{
struct intel_lpss_platform_info *info;
const struct acpi_device_id *id;
+ int ret;

id = acpi_match_device(intel_lpss_acpi_ids, &pdev->dev);
if (!id)
@@ -149,10 +150,14 @@ static int intel_lpss_acpi_probe(struct platform_device *pdev)
info->mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
info->irq = platform_get_irq(pdev, 0);

+ ret = intel_lpss_probe(&pdev->dev, info);
+ if (ret)
+ return ret;
+
pm_runtime_set_active(&pdev->dev);
pm_runtime_enable(&pdev->dev);

- return intel_lpss_probe(&pdev->dev, info);
+ return ret;
}

static int intel_lpss_acpi_remove(struct platform_device *pdev)
--
2.33.0

2021-11-01 18:24:21

by Aditya Garg

Subject: Re: [PATCH v1 1/1] mfd: intel-lpss: Fix too early PM enablement in the ACPI ->probe()

> On 01-Nov-2021, at 8:40 PM, Andy Shevchenko <[email protected]> wrote:
>
> The runtime PM callback may be called as soon as the runtime PM facility
> is enabled and activated. It means that ->suspend() may be called before
> we finish probing the device in the ACPI case. Hence, NULL pointer
> dereference:
>
> intel-lpss INT34BA:00: IRQ index 0 not found
> BUG: kernel NULL pointer dereference, address: 0000000000000030
> ...
> Workqueue: pm pm_runtime_work
> RIP: 0010:intel_lpss_suspend+0xb/0x40 [intel_lpss]
>
> To fix this, first try to register the device and only after that enable
> runtime PM facility.
>
> Fixes: 4b45efe85263 ("mfd: Add support for Intel Sunrisepoint LPSS devices")
> Reported-by: Orlando Chamberlain <[email protected]>
> Reported-by: Aditya Garg <[email protected]>
> Signed-off-by: Andy Shevchenko <[email protected]>
Following patch fixed the issue for me.
Tested-by: Aditya Garg <[email protected]>
> ---
> drivers/mfd/intel-lpss-acpi.c | 7 ++++++-
> 1 file changed, 6 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/mfd/intel-lpss-acpi.c b/drivers/mfd/intel-lpss-acpi.c
> index 3f1d976eb67c..be81507afb5e 100644
> --- a/drivers/mfd/intel-lpss-acpi.c
> +++ b/drivers/mfd/intel-lpss-acpi.c
> @@ -136,6 +136,7 @@ static int intel_lpss_acpi_probe(struct platform_device *pdev)
> {
> struct intel_lpss_platform_info *info;
> const struct acpi_device_id *id;
> + int ret;
>
> id = acpi_match_device(intel_lpss_acpi_ids, &pdev->dev);
> if (!id)
> @@ -149,10 +150,14 @@ static int intel_lpss_acpi_probe(struct platform_device *pdev)
> info->mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> info->irq = platform_get_irq(pdev, 0);
>
> + ret = intel_lpss_probe(&pdev->dev, info);
> + if (ret)
> + return ret;
> +
> pm_runtime_set_active(&pdev->dev);
> pm_runtime_enable(&pdev->dev);
>
> - return intel_lpss_probe(&pdev->dev, info);
> + return ret;
> }
>
> static int intel_lpss_acpi_remove(struct platform_device *pdev)
> --
> 2.33.0
>

2021-11-01 19:38:36

by Andy Shevchenko

Subject: Re: [PATCH v1 1/1] mfd: intel-lpss: Fix too early PM enablement in the ACPI ->probe()

On Mon, Nov 01, 2021 at 06:21:51PM +0000, Aditya Garg wrote:
> > On 01-Nov-2021, at 8:40 PM, Andy Shevchenko <[email protected]> wrote:
> >
> > The runtime PM callback may be called as soon as the runtime PM facility
> > is enabled and activated. It means that ->suspend() may be called before
> > we finish probing the device in the ACPI case. Hence, NULL pointer
> > dereference:
> >
> > intel-lpss INT34BA:00: IRQ index 0 not found
> > BUG: kernel NULL pointer dereference, address: 0000000000000030
> > ...
> > Workqueue: pm pm_runtime_work
> > RIP: 0010:intel_lpss_suspend+0xb/0x40 [intel_lpss]
> >
> > To fix this, first try to register the device and only after that enable
> > runtime PM facility.

> Following patch fixed the issue for me.
> Tested-by: Aditya Garg <[email protected]>

Thanks for testing!


--
With Best Regards,
Andy Shevchenko