net: phy: fix cleanup after probe failure
Subsystems (3)
v3 cover net: phy: fix cleanup after probe failure
From: Xuanqiang Luo <[email protected]>
phy_probe() initializes the PHY driver, ports, SFP upstream, and LEDs in stages. Its error paths do not always release only the resources acquired at each stage. It can also mark the PHY ready before all setup succeeds.
Port setup also leaves SFP cleanup split between phy_sfp_probe(), phy_setup_ports(), and phy_probe(), and default port setup ignores errors from attaching the port to the PHY driver.
This series makes each initialization layer own its cleanup and propagates setup failures to the caller.
Patch 1 splits the phy_probe() cleanup by initialization stage.
Patch 2 makes SFP and port setup unwind their resources in the required order.
Patch 3 sets PHY_READY only after LED setup succeeds.
Patch 4 calls the PHY driver remove callback after later probe failures.
Patch 5 propagates errors from default port setup.
---
Changes:
v3:
Patch 1:
- Rename cleanup labels to include verbs describing their actions. (Jakub Kicinski.)
Patch 3:
- Do not clear phydev->drv before device-core teardown completes; this can expose NULL dereferences in concurrent attach paths and devres callbacks. Set PHY_READY only after LED setup succeeds and update the Fixes tag. (Sashiko, Jakub Kicinski.)
v2: https://lore.kernel.org/all/[email protected]/
Patch 1:
- Limit this patch to splitting phy_probe() error paths, moving the SFP teardown fixes to Patch 2.
Patch 2 (new):
- makes SFP and port setup unwind their resources in the required order.
- Add phy_sfp_release() for complete SFP teardown instead of open-coding sfp_bus_del_upstream(). (Andrew Lunn, Maxime Chevallier.)
Patch 3 (new):
- Restore PHY_DOWN and clear phydev->drv after probe failure.
Patch 4:
- Move the former Patch 2 to Patch 4; no functional changes.
Patch 5 (new):
- Propagate errors from default port setup.
v1: https://lore.kernel.org/all/[email protected]/
Xuanqiang Luo (5):
net: phy: split phy_probe() error paths
net: phy: unregister SFP upstream before port cleanup
net: phy: set PHY_READY after LED setup
net: phy: call driver remove when core initialization fails
net: phy: propagate errors from default port setup
drivers/net/phy/phy_device.c | 80 ++++++++++++++++++++++++++---------- 1 file changed, 58 insertions(+), 22 deletions(-)
base-commit: f5bbbfec59b4e2fb7520a91de3df8a6174325d6a
Signature · 1 line
2.43.0
PATCH v3 1/5 net: phy: split phy_probe() error paths
From: Xuanqiang Luo <[email protected]>
phy_probe() uses one cleanup path for failures at every initialization stage. This runs cleanup for resources that have not been initialized.
Split the cleanup by initialization stage so each failure path unwinds only the resources that may have been initialized.
--- drivers/net/phy/phy_device.c | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-)
drivers/net/phy/phy_device.c +11 -9
@@ -3681,7 +3681,7 @@ static int phy_probe(struct device *dev)if (phydev->drv->probe) { err = phydev->drv->probe (phydev);if (err) - goto out; + goto out_reset; }phy_disable_interrupts (phydev);
@@ -3702,7 +3702,7 @@ static int phy_probe(struct device *dev) err =genphy_read_abilities (phydev);if (err) - goto out; + goto out_reset;if (!linkmode_test_bit (ETHTOOL_LINK_MODE_Autoneg_BIT, phydev->supported))
@@ -3719,7 +3719,7 @@ static int phy_probe(struct device *dev) err =phy_setup_ports (phydev);if (err) - goto out; + goto out_sfp_release;phy_advertise_supported (phydev);
@@ -3728,7 +3728,7 @@ static int phy_probe(struct device *dev)*/ err =genphy_c45_read_eee_adv (phydev, phydev->advertising_eee);if (err) - goto out; + goto out_sfp_release;/* Get the EEE modes we want to prohibit. */ of_set_phy_eee_broken (phydev);
@@ -3781,20 +3781,22 @@ static int phy_probe(struct device *dev)if (IS_ENABLED (CONFIG_PHYLIB_LEDS ) && !phy_driver_is_genphy (phydev)) { err =of_phy_leds (phydev);if (err) - goto out; + goto out_unreg_led_triggers; }return 0 ; -out: +out_unreg_led_triggers: +if (!phydev->is_on_sfp_module) +phy_led_triggers_unregister (phydev); + +out_sfp_release:sfp_bus_del_upstream (phydev->sfp_bus); phydev->sfp_bus =NULL ;phy_cleanup_ports (phydev); -if (!phydev->is_on_sfp_module) -phy_led_triggers_unregister (phydev); - +out_reset:/* Re-assert the reset signal on error */ phy_device_reset (phydev,1 );
Signature · 1 line
2.43.0
Andrew Lunn
maintainerRe: [PATCH net v3 1/5] net: phy: split phy_probe() error paths
Quoted 13 lines · Xuanqiang Luo
From: Xuanqiang Luo <[email protected]>
phy_probe()uses one cleanup path for failures at every initialization stage. This runs cleanup for resources that have not been initialized.Split the cleanup by initialization stage so each failure path unwinds only the resources that may have been initialized.
drivers/net/phy/phy_device.c +3
-1-out: +out_unreg_led_triggers: +if (!phydev->is_on_sfp_module) +phy_led_triggers_unregister (phydev);
drivers/net/phy/phy_device.c static int phy_probe(struct device *dev) · line 3797
Quoted 10 lines
drivers/net/phy/phy_device.c +2
-3+ +out_sfp_release:sfp_bus_del_upstream (phydev->sfp_bus); phydev->sfp_bus =NULL ;phy_cleanup_ports (phydev); -if (!phydev->is_on_sfp_module) -phy_led_triggers_unregister (phydev); -
It would be nice to mention this reorder in the commit message. The commit message is about "Why?". Explain why the order has been changed.
Andrew
---
pw-bot: cr
Re: [PATCH net v3 1/5] net: phy: split phy_probe() error paths
Quoted 12 lines · Andrew Lunn
Quoted 11 lines · Xuanqiang Luo
drivers/net/phy/phy_device.c +5
-1-out: +out_unreg_led_triggers: +if (!phydev->is_on_sfp_module) +phy_led_triggers_unregister (phydev); + +out_sfp_release:sfp_bus_del_upstream (phydev->sfp_bus); phydev->sfp_bus =NULL ;phy_cleanup_ports (phydev);
Quoted 8 lines
drivers/net/phy/phy_device.c static int phy_probe(struct device *dev) · line 3797
Quoted 3 lines
drivers/net/phy/phy_device.c +0
-3-if (!phydev->is_on_sfp_module) -phy_led_triggers_unregister (phydev); -It would be nice to mention this reorder in the commit message. The commit message is about "Why?". Explain why the order has been changed.
Andrew
Sure, I'll explain it in more detail in the next revision.
Thanks,
Xuanqiang
PATCH v3 2/5 net: phy: unregister SFP upstream before port cleanup
In reply to Xuanqiang LuoFrom: Xuanqiang Luo <[email protected]>
Commit 4497f5028675 ("net: phy: Clean the phy_ports after unregistering the downstream SFP bus") established that an SFP upstream must be unregistered before its phy_ports are destroyed because SFP callbacks may access these ports.
phy_setup_ports() does not follow this order when a later port setup step fails after phy_sfp_probe() succeeds. It destroys the SFP phy_port and leaves phy_probe() to unregister the upstream later, creating a race between port destruction and SFP upstream callbacks.
The error unwind is also split across three functions. If phy_setup_sfp_port() fails, phy_sfp_probe() leaves the upstream registered and relies on phy_probe() to remove it after phy_setup_ports() returns.
Make each layer unwind the resources it successfully set up. Unregister only the upstream in phy_sfp_probe() when SFP port setup fails, since the failed port has already been destroyed. Add phy_sfp_release() for a successful SFP probe, and make phy_setup_ports() use it before cleaning up the remaining ports. Once phy_setup_ports() has rolled back all port setup, make phy_probe() skip this cleanup.
--- drivers/net/phy/phy_device.c | 49 ++++++++++++++++++++++++++++-------- 1 file changed, 38 insertions(+), 11 deletions(-)
drivers/net/phy/phy_device.c +38 -11
@@ -1723,12 +1723,41 @@ static int phy_sfp_probe(struct phy_device *phydev) phydev->sfp_bus =NULL ; } -if (!ret && phydev->sfp_bus) +if (!ret && phydev->sfp_bus) { ret =phy_setup_sfp_port (phydev); +if (ret) { +sfp_bus_del_upstream (phydev->sfp_bus); + phydev->sfp_bus =NULL ; + } + }return ret; } +/** +* phy_sfp_release - release resources set up by phy_sfp_probe() +* @phydev: the PHY device +* +* Release the SFP resources set up by a successful phy_sfp_probe(). Unregister +* the upstream before destroying its phy_port, so SFP upstream callbacks cannot +* race with port destruction. +*/ +static void phy_sfp_release (struct phy_device *phydev) +{ +struct phy_port *port, *tmp; + +sfp_bus_del_upstream (phydev->sfp_bus); + phydev->sfp_bus =NULL ; + +list_for_each_entry_safe (port, tmp, &phydev->ports, head) { +if (!port->is_sfp) +continue ; + +phy_del_port (phydev, port); +phy_port_destroy (port); + } +} +static bool phy_drv_supports_irq (const struct phy_driver *phydrv) {return phydrv->config_intr && phydrv->handle_interrupt;
@@ -3522,13 +3551,13 @@ static int phy_setup_ports(struct phy_device *phydev)if (!phydev->is_genphy_driven) { ret =phy_sfp_probe (phydev);if (ret) - goto out; + goto err_ports; }if (phydev->n_ports < phydev->max_n_ports) { ret =phy_default_setup_single_port (phydev);if (ret) - goto out; + goto err_sfp; }linkmode_zero (ports_supported);
@@ -3555,7 +3584,9 @@ static int phy_setup_ports(struct phy_device *phydev)return 0 ; -out: +err_sfp: +phy_sfp_release (phydev); +err_ports:phy_cleanup_ports (phydev);return ret; }
@@ -3719,7 +3750,7 @@ static int phy_probe(struct device *dev) err =phy_setup_ports (phydev);if (err) - goto out_sfp_release; + goto out_reset;phy_advertise_supported (phydev);
@@ -3791,9 +3822,7 @@ static int phy_probe(struct device *dev)phy_led_triggers_unregister (phydev); out_sfp_release: -sfp_bus_del_upstream (phydev->sfp_bus); - phydev->sfp_bus =NULL ; - +phy_sfp_release (phydev);phy_cleanup_ports (phydev); out_reset:
@@ -3817,9 +3846,7 @@ static int phy_remove(struct device *dev) phydev->state =PHY_DOWN ; -sfp_bus_del_upstream (phydev->sfp_bus); - phydev->sfp_bus =NULL ; - +phy_sfp_release (phydev);phy_cleanup_ports (phydev);if (phydev->drv && phydev->drv->remove)
Signature · 1 line
2.43.0
Re: [PATCH net v3 2/5] net: phy: unregister SFP upstream before port cleanup
Quoted 24 lines · Xuanqiang Luo
From: Xuanqiang Luo <[email protected]>
Commit
4497f5028675("net: phy: Clean thephy_portsafter unregistering the downstream SFP bus") established that an SFP upstream must be unregistered before itsphy_portsare destroyed because SFP callbacks may access these ports.
phy_setup_ports()does not follow this order when a later port setup step fails afterphy_sfp_probe()succeeds. It destroys the SFPphy_portand leavesphy_probe()to unregister the upstream later, creating a race between port destruction and SFP upstream callbacks.The error unwind is also split across three functions. If
phy_setup_sfp_port()fails,phy_sfp_probe()leaves the upstream registered and relies onphy_probe()to remove it afterphy_setup_ports()returns.Make each layer unwind the resources it successfully set up. Unregister only the upstream in
phy_sfp_probe()when SFP port setup fails, since the failed port has already been destroyed. Addphy_sfp_release()for a successful SFP probe, and makephy_setup_ports()use it before cleaning up the remaining ports. Oncephy_setup_ports()has rolled back all port setup, makephy_probe()skip this cleanup.
Quoted 2 lines
Fixes 589e934d2735 ("net: phy: Introduce PHY ports representation")Signed-off-by Xuanqiang Luo <[email protected]>
Andrew
PATCH v3 3/5 net: phy: set PHY_READY after LED setup
In reply to Xuanqiang LuoFrom: Xuanqiang Luo <[email protected]>
phy_probe() sets PHY_READY before calling of_phy_leds(). If LED setup fails, the error path releases the initialized resources while the PHY state remains READY even though probing failed.
Set PHY_READY only after LED setup succeeds.
--- drivers/net/phy/phy_device.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/net/phy/phy_device.c +3 -3
@@ -3799,9 +3799,6 @@ static int phy_probe(struct device *dev) phydev->supported); } -/* Set the state to READY by default */ - phydev->state =PHY_READY ; -/* Register the PHY LED triggers */ if (!phydev->is_on_sfp_module)phy_led_triggers_register (phydev);
@@ -3815,6 +3812,9 @@ static int phy_probe(struct device *dev) goto out_unreg_led_triggers; } +/* Set the state to READY by default */ + phydev->state =PHY_READY ; +return 0 ; out_unreg_led_triggers:
Signature · 1 line
2.43.0
Re: [PATCH net v3 3/5] net: phy: set PHY_READY after LED setup
Quoted 10 lines · Xuanqiang Luo
From: Xuanqiang Luo <[email protected]>
phy_probe()setsPHY_READYbefore callingof_phy_leds(). If LED setup fails, the error path releases the initialized resources while the PHY state remains READY even though probing failed.Set
PHY_READYonly after LED setup succeeds.Fixes 01e5b728e9e4 ("net: phy: Add a binding for PHY LEDs")Signed-off-by Xuanqiang Luo <[email protected]>
Andrew
Re: [PATCH net v3 3/5] net: phy: set PHY_READY after LED setup
In reply to Xuanqiang LuoQuoted 11 lines · Xuanqiang Luo
From: Xuanqiang Luo <[email protected]>
phy_probe()setsPHY_READYbefore callingof_phy_leds(). If LED setup fails, the error path releases the initialized resources while the PHY state remains READY even though probing failed.Set
PHY_READYonly after LED setup succeeds.Fixes 01e5b728e9e4 ("net: phy: Add a binding for PHY LEDs")Signed-off-by Xuanqiang Luo <[email protected]>
I noticed that Sashiko found two new issues:
Quoted 24 lines
Quoted 5 lines
drivers/net/phy/phy_device.c +0
-0drivers/net/phy/phy_device.c static int phy_probe(struct device *dev) · line 3804
Quoted 9 lines
drivers/net/phy/phy_device.c +0
-3@@ -3799,9 +3799,6 @@ static int phy_probe(struct device *dev) phydev->supported); } - /* Set the state to READY by default */ - phydev->state = PHY_READY; - /* Register the PHY LED triggers */ if (!phydev->is_on_sfp_module) phy_led_triggers_register(phydev);[Severity: Critical]
This is a pre-existing issue, but by ignoring the return value ofphy_led_triggers_register()here, could we trigger a double-free later?In
drivers/net/phy/phy_led_triggers.c, ifphy_led_triggers_register()fails, it freesphy->phy_led_triggersbut leaves it as a dangling pointer and setsphy_num_led_triggers= 0. Since the return value is ignored here,phy_probe()continues execution.
Quoted 5 lines
If
of_phy_leds()then fails, the code jumps toout_unreg_led_triggersand callsphy_led_triggers_unregister(). Inside that function, the loop is skipped becausephy_num_led_triggersis 0, but it unconditionally callskfree(phy->phy_led_triggers). This results in a double-free of the dangling pointer.
Although this part was not introduced by the current patch, I think it still falls within the scope of probe failure handling, so I plan to add a separate patch in the next revision to fix it.
Quoted 17 lines
drivers/net/phy/phy_device.c static int phy_probe(struct device *dev) · line 3817
Quoted 7 lines
drivers/net/phy/phy_device.c +3
-0@@ -3815,6 +3812,9 @@ static int phy_probe(struct device *dev) goto out_unreg_led_triggers; } +/* Set the state to READY by default */ + phydev->state =PHY_READY ; +[Severity: High]
Does moving this lockless state assignment to the end ofphy_probe()create a race condition with concurrent MAC driver attachments?The driver core sets
dev->driverbefore invokingphy_probe(). Whilephy_probe()is still running and potentially blocked inof_phy_leds(), a concurrent network driver probe could find this PHY and callphy_attach_direct().
Quoted 9 lines
Because
dev->driveris already set,phy_attach_direct()proceeds without acquiring the PHY's device lock and sets the state toPHY_READY. The MAC driver then immediately callsphy_start(), which acquiresphydev->lockand transitions the state fromPHY_READYtoPHY_UP.If this lockless assignment completes after the MAC driver has already called
phy_start(), won't it silently clobber thePHY_UPstate back toPHY_READYand stall the PHY state machine?
This appears to be a pre-existing issue, since phy_attach_direct() and phy_probe() are not serialized.
This patch does widen the race window, but phydev->state should still be set to PHY_READY before phy_probe() returns.
Thanks,
Xuanqiang
Andrew Lunn
maintainerRe: [PATCH net v3 3/5] net: phy: set PHY_READY after LED setup
Quoted 43 lines · Xuanqiang Luo
Quoted 11 lines · Xuanqiang Luo
From: Xuanqiang Luo <[email protected]>
phy_probe()setsPHY_READYbefore callingof_phy_leds(). If LED setup fails, the error path releases the initialized resources while the PHY state remains READY even though probing failed.Set
PHY_READYonly after LED setup succeeds.Fixes 01e5b728e9e4 ("net: phy: Add a binding for PHY LEDs")Signed-off-by Xuanqiang Luo <[email protected]>I noticed that Sashiko found two new issues:
Quoted 22 lines
drivers/net/phy/phy_device.c static int phy_probe(struct device *dev) · line 3804
Quoted 13 lines
drivers/net/phy/phy_device.c +0
-2@@ -3799,9 +3799,6 @@ static int phy_probe(struct device *dev) phydev->supported); } > - /* Set the state to READY by default */ - phydev->state = PHY_READY; - /* Register the PHY LED triggers */ if (!phydev->is_on_sfp_module) phy_led_triggers_register(phydev);[Severity: Critical]
This is a pre-existing issue, but by ignoring the return value ofphy_led_triggers_register()here, could we trigger a double-free later?In
drivers/net/phy/phy_led_triggers.c, ifphy_led_triggers_register()fails, it freesphy->phy_led_triggersbut leaves it as a dangling pointer and setsphy_num_led_triggers= 0. Since the return value is ignored here,phy_probe()continues execution.Quoted 5 lines
If
of_phy_leds()then fails, the code jumps toout_unreg_led_triggersand callsphy_led_triggers_unregister(). Inside that function, the loop is skipped becausephy_num_led_triggersis 0, but it unconditionally callskfree(phy->phy_led_triggers). This results in a double-free of the dangling pointer.
Quoted 14 lines
Although this part was not introduced by the current patch, I think it still falls within the scope of probe failure handling, so I plan to add a separate patch in the next revision to fix it.
Quoted 10 lines
drivers/net/phy/phy_device.c static int phy_probe(struct device *dev) · line 3816
Quoted 6 lines
drivers/net/phy/phy_device.c +2
-0@@ -3815,6 +3812,9 @@ static int phy_probe(struct device *dev) goto out_unreg_led_triggers; } > + /* Set the state to READY by default */ + phydev->state =PHY_READY ; +[Severity: High]
Does moving this lockless state assignment to the end ofphy_probe()create a race condition with concurrent MAC driver attachments?
I'm not sure this is correct. If phy_probe() is still ongoing, can you even call phy_attach()?
In the normal case the MAC driver registers the MDIO bus during its probe. The PHYs will be scanned and probed synchronously. Once the MDIO is registered, the MAC driver will then continue. Generally the PHY is attached in open() but it could be later in probe. So think this serialisation is performed by general code structure.
It gets more interesting with a shared MDIO bus, which has two or more PHYs on it. The first MAC driver registers its MDIO bus as above, and later attaches its PHY. However in parallel a second MAC driver can probe, and use a phandle to point to the PHY on the MDIO bus, and try to attach to it. But i _guess_ following the phandle does not work until the PHY probe has completed?
Quoted 5 lines
This appears to be a pre-existing issue, since
phy_attach_direct()andphy_probe()are not serialized.This patch does widen the race window, but
phydev->stateshould still be set toPHY_READYbeforephy_probe()returns.
I would agree it is an old issues, if it is an issue. So i would say it is out of scope of this patchset. You can look deeper if you want, but it is not required.
Andrew
PATCH v3 4/5 net: phy: call driver remove when core initialization fails
In reply to Xuanqiang LuoFrom: Xuanqiang Luo <[email protected]>
phy_probe() may fail while querying features or completing other core initialization after the PHY driver probe callback has succeeded. The driver core does not run the remove path after a probe error, so resources that the PHY driver releases in its remove callback are leaked.
Call the PHY driver remove callback on these failures.
--- drivers/net/phy/phy_device.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
drivers/net/phy/phy_device.c +6 -2
@@ -3733,7 +3733,7 @@ static int phy_probe(struct device *dev) err =genphy_read_abilities (phydev);if (err) - goto out_reset; + goto out_remove;if (!linkmode_test_bit (ETHTOOL_LINK_MODE_Autoneg_BIT, phydev->supported))
@@ -3750,7 +3750,7 @@ static int phy_probe(struct device *dev) err =phy_setup_ports (phydev);if (err) - goto out_reset; + goto out_remove;phy_advertise_supported (phydev);
@@ -3825,6 +3825,10 @@ static int phy_probe(struct device *dev)phy_sfp_release (phydev);phy_cleanup_ports (phydev); +out_remove: +if (phydev->drv->remove) + phydev->drv->remove (phydev); + out_reset:/* Re-assert the reset signal on error */ phy_device_reset (phydev,1 );
Signature · 1 line
2.43.0
Re: [PATCH net v3 4/5] net: phy: call driver remove when core initialization fails
Quoted 8 lines · Xuanqiang Luo
From: Xuanqiang Luo <[email protected]>
phy_probe()may fail while querying features or completing other core initialization after the PHY driver probe callback has succeeded. The driver core does not run the remove path after a probe error, so resources that the PHY driver releases in its remove callback are leaked.
Quoted 4 lines
Call the PHY driver remove callback on these failures.
Fixes efbdfdc29bdd ("net: phy: Add support for asking the PHY its abilities")Signed-off-by Xuanqiang Luo <[email protected]>
Andrew
PATCH v3 5/5 net: phy: propagate errors from default port setup
In reply to Xuanqiang LuoFrom: Xuanqiang Luo <[email protected]>
phy_default_setup_single_port() ignores errors from phy_add_port() and always reports success. If a PHY driver attach_mdi_port() callback fails, the phy_port is leaked and PHY probing continues without the expected default port.
Destroy the port and return the error.
--- drivers/net/phy/phy_device.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)
drivers/net/phy/phy_device.c +5 -2
@@ -3458,6 +3458,7 @@ static int phy_default_setup_single_port(struct phy_device *phydev) {struct phy_port *port =phy_port_alloc ();unsigned long mode; +int ret;if (!port)return -ENOMEM ;
@@ -3484,9 +3485,11 @@ static int phy_default_setup_single_port(struct phy_device *phydev) port->pairs =max_t (int, port->pairs,ethtool_linkmode_n_pairs (mode)); -phy_add_port (phydev, port); + ret =phy_add_port (phydev, port); +if (ret) +phy_port_destroy (port); -return 0 ; +return ret; }static int of_phy_ports (struct phy_device *phydev)
Signature · 1 line
2.43.0
Re: [PATCH net v3 5/5] net: phy: propagate errors from default port setup
Hi,
Quoted 11 lines · Xuanqiang Luo
From: Xuanqiang Luo <[email protected]>
phy_default_setup_single_port()ignores errors fromphy_add_port()and always reports success. If a PHY driverattach_mdi_port()callback fails, thephy_portis leaked and PHY probing continues without the expected default port.Destroy the port and return the error.
Fixes 589e934d2735 ("net: phy: Introduce PHY ports representation")Signed-off-by Xuanqiang Luo <[email protected]>
Maxime
Quoted 30 lines
--- drivers/net/phy/phy_device.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)drivers/net/phy/phy_device.c +5
-2@@ -3458,6 +3458,7 @@ static int phy_default_setup_single_port(struct phy_device *phydev) {struct phy_port *port =phy_port_alloc ();unsigned long mode; +int ret;if (!port)return -ENOMEM ;@@ -3484,9 +3485,11 @@ static int phy_default_setup_single_port(struct phy_device *phydev) port->pairs =max_t (int, port->pairs,ethtool_linkmode_n_pairs (mode)); -phy_add_port (phydev, port); + ret =phy_add_port (phydev, port); +if (ret) +phy_port_destroy (port); -return 0 ; +return ret; }static int of_phy_ports (struct phy_device *phydev)
Andrew Lunn
maintainerQuoted 15 lines · Xuanqiang Luo
From: Xuanqiang Luo <[email protected]>
phy_probe()initializes the PHY driver, ports, SFP upstream, and LEDs in stages. Its error paths do not always release only the resources acquired at each stage. It can also mark the PHY ready before all setup succeeds.Port setup also leaves SFP cleanup split between
phy_sfp_probe(),phy_setup_ports(), andphy_probe(), and default port setup ignores errors from attaching the port to the PHY driver.This series makes each initialization layer own its cleanup and propagates setup failures to the caller.
Patch 1 splits the
phy_probe()cleanup by initialization stage.
Quoted 8 lines
Patch 2 makes SFP and port setup unwind their resources in the required order.
Patch 3 sets
PHY_READYonly after LED setup succeeds.Patch 4 calls the PHY driver remove callback after later probe failures.
Patch 5 propagates errors from default port setup.
The scope of these patches has increased quite a bit. It is now more like ongoing development work than a actual fix. Does this bother anybody?
Please submit for net-next, once it reopens.
Andrew
Jakub Kicinski
maintainerQuoted 1 line · On Wed, 19 Aug 2026 15:33:28 +0200 Andrew Lunn wrote:
Please submit for net-next, once it reopens.
FWIW I'd prefer to take these sort of error path fixes to net during the merge window. No point backlogging borderline stuff. Any particular reason you want this in net-next?
Andrew Lunn
maintainerQuoted 6 lines · Jakub Kicinski
Quoted 1 line · On Wed, 19 Aug 2026 15:33:28 +0200 Andrew Lunn wrote:
Please submit for net-next, once it reopens.
FWIW I'd prefer to take these sort of error path fixes to net during the merge window. No point backlogging borderline stuff. Any particular reason you want this in net-next?
I don't think it bothers anybody, so does not meet stable rules.
But i've nothing against it being merged now.
Andrew
Quoted 11 lines · Andrew Lunn
Quoted 5 lines · Jakub Kicinski
Quoted 1 line · On Wed, 19 Aug 2026 15:33:28 +0200 Andrew Lunn wrote:
Please submit for net-next, once it reopens.
FWIW I'd prefer to take these sort of error path fixes to net during the merge window. No point backlogging borderline stuff. Any particular reason you want this in net-next?
I don't think it bothers anybody, so does not meet stable rules.
But i've nothing against it being merged now.
Andrew
Thanks for the discussion.
Based on the feedback, I’ll continue targeting net and post a revised version in a follow-up.
Thanks,
Xuanqiang