Currently, during the LAN8814 PTP probe shared->phydev is only set if PTP
clock gets actually set, otherwise the function will return before setting
it.

This is an issue as shared->phydev is unconditionally being used when IRQ
is being handled, especially in lan8814_gpio_process_cap and since it was
not set it will cause a NULL pointer exception and crash the kernel.

So, simply always set shared->phydev to avoid the NULL pointer exception.

Fixes: b3f1a08fcf0d ("net: phy: micrel: Add support for PTP_PF_EXTTS for lan8814")
Signed-off-by: Robert Marko <robert.marko@sartura.hr>
---
 drivers/net/phy/micrel.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/phy/micrel.c b/drivers/net/phy/micrel.c
index 16855bf8c391..4929f9b81f54 100644
--- a/drivers/net/phy/micrel.c
+++ b/drivers/net/phy/micrel.c
@@ -4285,6 +4285,8 @@ static int __lan8814_ptp_probe_once(struct phy_device *phydev, char *pin_name,
 {
 	struct lan8814_shared_priv *shared = phy_package_get_priv(phydev);
 
+	shared->phydev = phydev;
+
 	/* Initialise shared lock for clock*/
 	mutex_init(&shared->shared_lock);
 
@@ -4340,8 +4342,6 @@ static int __lan8814_ptp_probe_once(struct phy_device *phydev, char *pin_name,
 
 	phydev_dbg(phydev, "successfully registered ptp clock\n");
 
-	shared->phydev = phydev;
-
 	/* The EP.4 is shared between all the PHYs in the package and also it
 	 * can be accessed by any of the PHYs
 	 */
-- 
2.51.0

On Tue, 21 Oct 2025 15:20:26 +0200 Robert Marko wrote:
> Currently, during the LAN8814 PTP probe shared->phydev is only set if PTP
> clock gets actually set, otherwise the function will return before setting
> it.
> 
> This is an issue as shared->phydev is unconditionally being used when IRQ
> is being handled, especially in lan8814_gpio_process_cap and since it was
> not set it will cause a NULL pointer exception and crash the kernel.
> 
> So, simply always set shared->phydev to avoid the NULL pointer exception.
> 
> Fixes: b3f1a08fcf0d ("net: phy: micrel: Add support for PTP_PF_EXTTS for lan8814")
> Signed-off-by: Robert Marko <robert.marko@sartura.hr>

Hopefully I'm not misunderstanding the situation, but since we're
considering taking the other patchset via net-next I'll apply
already this to prevent crashes in net..

On Thu, Oct 23, 2025 at 4:13 PM Jakub Kicinski <kuba@kernel.org> wrote:
>
> On Tue, 21 Oct 2025 15:20:26 +0200 Robert Marko wrote:
> > Currently, during the LAN8814 PTP probe shared->phydev is only set if PTP
> > clock gets actually set, otherwise the function will return before setting
> > it.
> >
> > This is an issue as shared->phydev is unconditionally being used when IRQ
> > is being handled, especially in lan8814_gpio_process_cap and since it was
> > not set it will cause a NULL pointer exception and crash the kernel.
> >
> > So, simply always set shared->phydev to avoid the NULL pointer exception.
> >
> > Fixes: b3f1a08fcf0d ("net: phy: micrel: Add support for PTP_PF_EXTTS for lan8814")
> > Signed-off-by: Robert Marko <robert.marko@sartura.hr>
>
> Hopefully I'm not misunderstanding the situation, but since we're
> considering taking the other patchset via net-next I'll apply
> already this to prevent crashes in net..


That would be great, as Horatiu explained, if PTP_1588_CLOCK is not
enabled in the kernel config
kernel will crash with a NULL pointer exception as soon as the first
interrupt is being handled.

It is not tied to the pending [PATCH net-next v2] net: phy: micrel:
Add support for non-PTP SKUs for lan8814
patch.

Regards,
Robert


-- 
Robert Marko
Staff Embedded Linux Engineer
Sartura d.d.
Lendavska ulica 16a
10000 Zagreb, Croatia
Email: robert.marko@sartura.hr
Web: www.sartura.hr

On Tue, Oct 21, 2025 at 03:20:26PM +0200, Robert Marko wrote:
> Currently, during the LAN8814 PTP probe shared->phydev is only set if PTP
> clock gets actually set, otherwise the function will return before setting
> it.
> 
> This is an issue as shared->phydev is unconditionally being used when IRQ
> is being handled, especially in lan8814_gpio_process_cap and since it was
> not set it will cause a NULL pointer exception and crash the kernel.
> 
> So, simply always set shared->phydev to avoid the NULL pointer exception.
> 
> Fixes: b3f1a08fcf0d ("net: phy: micrel: Add support for PTP_PF_EXTTS for lan8814")
> Signed-off-by: Robert Marko <robert.marko@sartura.hr>

Please could you look at how this patch and

[PATCH net-next v2] net: phy: micrel: Add support for non PTP SKUs for lan8814

work together. It might be this patch is not required because of
changes in that patch?

Thanks
	Andrew

The 10/21/2025 15:32, Andrew Lunn wrote:

Hi,

> 
> On Tue, Oct 21, 2025 at 03:20:26PM +0200, Robert Marko wrote:
> > Currently, during the LAN8814 PTP probe shared->phydev is only set if PTP
> > clock gets actually set, otherwise the function will return before setting
> > it.
> >
> > This is an issue as shared->phydev is unconditionally being used when IRQ
> > is being handled, especially in lan8814_gpio_process_cap and since it was
> > not set it will cause a NULL pointer exception and crash the kernel.
> >
> > So, simply always set shared->phydev to avoid the NULL pointer exception.
> >
> > Fixes: b3f1a08fcf0d ("net: phy: micrel: Add support for PTP_PF_EXTTS for lan8814")
> > Signed-off-by: Robert Marko <robert.marko@sartura.hr>
> 
> Please could you look at how this patch and
> 
> [PATCH net-next v2] net: phy: micrel: Add support for non PTP SKUs for lan8814
> 
> work together. It might be this patch is not required because of
> changes in that patch?

I managed to reproduce this issue also with the patch:

[PATCH net-next v2] net: phy: micrel: Add support for non PTP SKUs for lan8814

The way I reproduced the issue is by disabling the config PTP_1588_CLOCK
and then just set the port up to get an interrupt.
Then I try to apply this patch and then the issue was fixed.

Tested-by: Horatiu Vultur <horatiu.vultur@microchip.com>

> 
> Thanks
>         Andrew

-- 
/Horatiu