Re: Use of disowned struct filename after 3c5499fa56f5?

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 06/11/2020 13:15, Dmitry Kadashev wrote:
> On Fri, Nov 06, 2020 at 12:49:05PM +0000, Pavel Begunkov wrote:
>> On 06/11/2020 10:08, Dmitry Kadashev wrote:
>>> On Thu, Nov 05, 2020 at 08:57:43PM +0000, Pavel Begunkov wrote:
>>> That's pretty much what do_unlinkat() does btw. Thanks Pavel for looking
>>> into this!
>>>
>>> Can I pick your brain some more? do_mkdirat() case is slightly
>>> different:
>>>
>>> static long do_mkdirat(int dfd, const char __user *pathname, umode_t mode)
>>> {
>>> 	struct dentry *dentry;
>>> 	struct path path;
>>> 	int error;
>>> 	unsigned int lookup_flags = LOOKUP_DIRECTORY;
>>>
>>> retry:
>>> 	dentry = user_path_create(dfd, pathname, &path, lookup_flags);
>>>
>>> If we just change @pathname to struct filename, then user_path_create
>>> can be swapped for filename_create(). But the same problem on retry
>>> arises. Is there some more or less "idiomatic" way to solve this?
>>
>> I don't think there is, fs guys may have a different opinion but
>> sometimes it's hard to get through them.
>>
>> I'd take a filename reference before "retry:"
> 
> How do I do that? Just `++name.refcnt` or is there a helper function /
> better way?

I don't know, take a look around if there is one. In the end, a review and
guys familiar with this code will hopefully suggest a better way (if any).

-- 
Pavel Begunkov



[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux