diff mbox

[2/5] scsi/bnx2fc: Prevent recursive cpuhotplug locking

Message ID 20170724105341.384083666@linutronix.de (mailing list archive)
State Accepted, archived
Headers show

Commit Message

Thomas Gleixner July 24, 2017, 10:52 a.m. UTC 
The BNX2FC module init/exit code installs/removes the hotplug callbacks with
the cpu hotplug lock held. This worked with the old CPU locking
implementation which allowed recursive locking, but with the new percpu
rwsem based mechanism this is not longer allowed.

Use the _cpuslocked() variants to fix this.

Reported-by: kernel test robot <fengguang.wu@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 drivers/scsi/bnx2fc/bnx2fc_fcoe.c |   15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)
diff mbox

Patch

--- a/drivers/scsi/bnx2fc/bnx2fc_fcoe.c
+++ b/drivers/scsi/bnx2fc/bnx2fc_fcoe.c
@@ -2766,15 +2766,16 @@  static int __init bnx2fc_mod_init(void)
 	for_each_online_cpu(cpu)
 		bnx2fc_percpu_thread_create(cpu);
 
-	rc = cpuhp_setup_state_nocalls(CPUHP_AP_ONLINE_DYN,
-				       "scsi/bnx2fc:online",
-				       bnx2fc_cpu_online, NULL);
+	rc = cpuhp_setup_state_nocalls_cpuslocked(CPUHP_AP_ONLINE_DYN,
+						  "scsi/bnx2fc:online",
+						  bnx2fc_cpu_online, NULL);
 	if (rc < 0)
 		goto stop_threads;
 	bnx2fc_online_state = rc;
 
-	cpuhp_setup_state_nocalls(CPUHP_SCSI_BNX2FC_DEAD, "scsi/bnx2fc:dead",
-				  NULL, bnx2fc_cpu_dead);
+	cpuhp_setup_state_nocalls_cpuslocked(CPUHP_SCSI_BNX2FC_DEAD,
+					     "scsi/bnx2fc:dead",
+					     NULL, bnx2fc_cpu_dead);
 	put_online_cpus();
 
 	cnic_register_driver(CNIC_ULP_FCOE, &bnx2fc_cnic_cb);
@@ -2850,8 +2851,8 @@  static void __exit bnx2fc_mod_exit(void)
 		bnx2fc_percpu_thread_destroy(cpu);
 	}
 
-	cpuhp_remove_state_nocalls(bnx2fc_online_state);
-	cpuhp_remove_state_nocalls(CPUHP_SCSI_BNX2FC_DEAD);
+	cpuhp_remove_state_nocalls_cpuslocked(bnx2fc_online_state);
+	cpuhp_remove_state_nocalls_cpuslocked(CPUHP_SCSI_BNX2FC_DEAD);
 
 	put_online_cpus();