Messages in this thread | | | Subject | Re: [PATCH] sparc: vio: use put_device() instead of kfree() | From | arvindY <> | Date | Wed, 25 Apr 2018 21:29:57 +0530 |
| |
On Wednesday 25 April 2018 09:14 PM, Shannon Nelson wrote: > On 4/25/2018 7:56 AM, Arvind Yadav wrote: >> Never directly free @dev after calling device_register(), even >> if it returned an error. Always use put_device() to give up the >> reference initialized. >> >> Signed-off-by: Arvind Yadav <arvind.yadav.cs@gmail.com> >> --- >> arch/sparc/kernel/vio.c | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/arch/sparc/kernel/vio.c b/arch/sparc/kernel/vio.c >> index 1a0fa10..32bae68 100644 >> --- a/arch/sparc/kernel/vio.c >> +++ b/arch/sparc/kernel/vio.c >> @@ -403,7 +403,7 @@ static struct vio_dev *vio_create_one(struct >> mdesc_handle *hp, u64 mp, >> if (err) { >> printk(KERN_ERR "VIO: Could not register device %s, err=%d\n", >> dev_name(&vdev->dev), err); >> - kfree(vdev); >> + put_device(&vdev->dev); > > Hmmm... I can see why the put_device() might be a good idea, but I > think we still need the kfree() so as to not leak the memory that was > kzalloc'd above for vdev. >
There is no need to call kfree() here. Because put_device() will decrement the last reference and then free the memory by calling dev->release(It'll call vio_dev_release()). Internally put_device() -> kobject_put() -> kobject_cleanup() which is responsible to call 'dev -> release' and also free other kobject resources. If we will call kfree() here, Then It'll be a redundant call.
~arvind
> sln > >> return NULL; >> } >> if (vdev->dp) >>
| |