Skip to contentSkip to conversation

Fix error handling path

PATCH In review Yuan Can 3 participants 5 messages Nov 14, 2022 – Nov 23, 2022

Subsystems (3)
lore
Share & tools
lore

Yuan Can

author
#0 lore raw

cover Fix error handling path

This series contains two missing error handling related bugfixes.

Yuan Can (2):
fm10k: Fix error handling in fm10k_init_module()
iavf: Fix error handling in iavf_init_module()

 drivers/net/ethernet/intel/fm10k/fm10k_main.c | 10 +++++++++-
 drivers/net/ethernet/intel/iavf/iavf_main.c   |  9 ++++++++-
 2 files changed, 17 insertions(+), 2 deletions(-)
Signature · 1 line
2.17.1
author
#1 lore raw

PATCH 1/2 fm10k: Fix error handling in fm10k_init_module()

A problem about modprobe fm10k failed is triggered with the following log given:

Intel(R) Ethernet Switch Host Interface Driver Copyright(c) 2013 - 2019 Intel Corporation.
debugfs: Directory 'fm10k' with parent '/' already present!

The reason is that fm10k_init_module() returns fm10k_register_pci_driver() directly without checking its return value, if fm10k_register_pci_driver() failed, it returns without removing debugfs and destroy workqueue, resulting the debugfs of fm10k can never be created later and leaks the workqueue.

fm10k_init_module()
alloc_workqueue()
fm10k_dbg_init() # create debugfs
fm10k_register_pci_driver()
pci_register_driver()
driver_register()
bus_add_driver()
priv = kzalloc(...) # OOM happened

   # return without remove debugfs and destroy workqueue

Fix by remove debugfs and destroy workqueue when
fm10k_register_pci_driver() returns error.

Fixes 7461fd913afe ("fm10k: Add support for debugfs")
Fixes b382bb1b3e2d ("fm10k: use separate workqueue for fm10k driver")
---
 drivers/net/ethernet/intel/fm10k/fm10k_main.c | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)
drivers/net/ethernet/intel/fm10k/fm10k_main.c +9 -1
index 4a6630586ec9..fc373472e4e1 100644
@@ -32,6 +32,8 @@ struct workqueue_struct *fm10k_workqueue;
  **/
 static int __init fm10k_init_module(void)
 {
+	int ret;
+
 	pr_info("%s\n", fm10k_driver_string);
 	pr_info("%s\n", fm10k_copyright);
 
@@ -43,7 +45,13 @@ static int __init fm10k_init_module(void)
 
 	fm10k_dbg_init();
 
-	return fm10k_register_pci_driver();
+	ret = fm10k_register_pci_driver();
+	if (ret) {
+		fm10k_dbg_exit();
+		destroy_workqueue(fm10k_workqueue);
+	}
+
+	return ret;
 }
 module_init(fm10k_init_module);
 
Signature · 1 line
2.17.1
Reviewed-by on v1 1/2
#2 lore raw

RE: [PATCH 1/2] fm10k: Fix error handling in fm10k_init_module()

Quoted 35 lines

-----Original Message-----
From: Yuan Can <[email protected]>
Sent: Monday, November 14, 2022 12:27 AM
To: Brandeburg, Jesse <[email protected]>; Nguyen, Anthony L <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; Vick, Matthew <[email protected]>; Keller, Jacob E <[email protected]>; [email protected]; [email protected]
Cc: [email protected]
Subject: [PATCH 1/2] fm10k: Fix error handling in fm10k_init_module()

A problem about modprobe fm10k failed is triggered with the following log given:

Intel(R) Ethernet Switch Host Interface Driver Copyright(c) 2013 - 2019 Intel Corporation.
debugfs: Directory 'fm10k' with parent '/' already present!

The reason is that fm10k_init_module() returns fm10k_register_pci_driver() directly without checking its return value, if fm10k_register_pci_driver() failed, it returns without removing debugfs and destroy workqueue, resulting the debugfs of fm10k can never be created later and leaks the workqueue.

fm10k_init_module()
alloc_workqueue()
fm10k_dbg_init() # create debugfs
fm10k_register_pci_driver()
pci_register_driver()
driver_register()
bus_add_driver()
priv = kzalloc(...) # OOM happened

Quoted 8 lines
   # return without remove debugfs and destroy workqueue

Fix by remove debugfs and destroy workqueue when
fm10k_register_pci_driver() returns error.

Fixes 7461fd913afe ("fm10k: Add support for debugfs")
Fixes b382bb1b3e2d ("fm10k: use separate workqueue for fm10k driver")

Makes sense. Thanks!

Quoted 34 lines
---
 drivers/net/ethernet/intel/fm10k/fm10k_main.c | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)
drivers/net/ethernet/intel/fm10k/fm10k_main.c +9 -2
index 4a6630586ec9..fc373472e4e1 100644
@@ -32,6 +32,8 @@ struct workqueue_struct *fm10k_workqueue;
  **/
 static int __init fm10k_init_module(void)
 {
+	int ret;
+
 	pr_info("%s\n", fm10k_driver_string);
 	pr_info("%s\n", fm10k_copyright);
 
@@ -43,7 +45,13 @@ static int __init fm10k_init_module(void)
 
 	fm10k_dbg_init();
 
-	return fm10k_register_pci_driver();
+	ret = fm10k_register_pci_driver();
+	if (ret) {
+		fm10k_dbg_exit();
+		destroy_workqueue(fm10k_workqueue);
+	}
+
+	return ret;
 }
 module_init(fm10k_init_module);
 
--
Quoted 1 line

2.17.1

_______________________________________________
Intel-wired-lan mailing list
[email protected]
https://lists.osuosl.org/mailman/listinfo/intel-wired-lan

author
#3 lore raw

PATCH 2/2 iavf: Fix error handling in iavf_init_module()

In reply to Yuan Can

The iavf_init_module() won't destroy workqueue when pci_register_driver() failed. Call destroy_workqueue() when pci_register_driver() failed to prevent the resource leak.

Similar to the handling of u132_hcd_init in commit f276e002793c ("usb: u132-hcd: fix resource leak")

Fixes 2803b16c10ea ("i40e/i40evf: Use private workqueue")
---
 drivers/net/ethernet/intel/iavf/iavf_main.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/net/ethernet/intel/iavf/iavf_main.c +8 -1
index 3fc572341781..69fded35e476 100644
@@ -5191,6 +5191,8 @@ static struct pci_driver iavf_driver = {
  **/
 static int __init iavf_init_module(void)
 {
+	int ret;
+
 	pr_info("iavf: %s\n", iavf_driver_string);
 
 	pr_info("%s\n", iavf_copyright);
@@ -5201,7 +5203,12 @@ static int __init iavf_init_module(void)
 		pr_err("%s: Failed to create workqueue\n", iavf_driver_name);
 		return -ENOMEM;
 	}
-	return pci_register_driver(&iavf_driver);
+
+	ret = pci_register_driver(&iavf_driver);
+	if (ret)
+		destroy_workqueue(iavf_wq);
+
+	return ret;
 }
 
 module_init(iavf_init_module);
Signature · 1 line
2.17.1
Tested-by on v1 2/2
#4 lore raw

RE: [Intel-wired-lan] [PATCH 2/2] iavf: Fix error handling in iavf_init_module()

Quoted 33 lines

-----Original Message-----
From: Intel-wired-lan <[email protected]> On Behalf Of Yuan Can
Sent: Monday, November 14, 2022 9:27 AM
To: Brandeburg, Jesse <[email protected]>; Nguyen, Anthony L <[email protected]>; [email protected];
[email protected]; [email protected]; [email protected]; [email protected]; [email protected]; Vick, Matthew <[email protected]>; Keller, Jacob E <[email protected]>; intel- [email protected]; [email protected]
Cc: [email protected]
Subject: [Intel-wired-lan] [PATCH 2/2] iavf: Fix error handling in iavf_init_module()

The iavf_init_module() won't destroy workqueue when pci_register_driver() failed. Call destroy_workqueue() when pci_register_driver() failed to prevent the resource leak.

Similar to the handling of u132_hcd_init in commit f276e002793c ("usb: u132-hcd: fix resource leak")

Fixes 2803b16c10ea ("i40e/i40evf: Use private workqueue")
---
 drivers/net/ethernet/intel/iavf/iavf_main.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/net/ethernet/intel/iavf/iavf_main.c +0 -0
index 3fc572341781..69fded35e476 100644
Quoted 1 line

https://lists.osuosl.org/mailman/listinfo/intel-wired-lan