Some driver uses tasklet_disable in device remove/release process,
tasklet_disable will inc tasklet->count and return. If the tasklet
is not handled yet under some softirq pressure, the tasklet will be
placed on the tasklet_vec, never have a chance to be excuted. This might
lead to a heavy loaded ksoftirqd, wakeup with pending_softirq, but
tasklet is disabled. tasklet_kill should be used in this case.
Signed-off-by: Xiaotian Feng <dannyfeng@xxxxxxxxxxx>
Cc: Rupesh Gujare <rgujare@xxxxxxxxxxxxxxx>
Cc: Chris Kelly <ckelly@xxxxxxxxxxxxxxx>
Cc: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
Cc: devel@xxxxxxxxxxxxxxxxxxxx
---
drivers/staging/ozwpan/ozhcd.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/staging/ozwpan/ozhcd.c b/drivers/staging/ozwpan/ozhcd.c
index 2e087ac..33c0009 100644
--- a/drivers/staging/ozwpan/ozhcd.c
+++ b/drivers/staging/ozwpan/ozhcd.c
@@ -2304,8 +2304,8 @@ error:
*/
void oz_hcd_term(void)
{
- tasklet_disable(&g_urb_process_tasklet);
- tasklet_disable(&g_urb_cancel_tasklet);
+ tasklet_kill(&g_urb_process_tasklet);
+ tasklet_kill(&g_urb_cancel_tasklet);
platform_device_unregister(g_plat_dev);
platform_driver_unregister(&g_oz_plat_drv);
oz_trace("Pending urbs:%d\n", atomic_read(&g_pending_urbs));
--
1.7.9.5
_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/devel
