Re: [PATCH -next] fs: Fix memory leaks in do_renameat2() error paths

[Jens Axboe]

On 10/30/20 9:52 AM, Qian Cai wrote:
> On Fri, 2020-10-30 at 09:27 -0600, Jens Axboe wrote:
>> On 10/30/20 9:24 AM, Qian Cai wrote:
>>> We will need to call putname() before do_renameat2() returning -EINVAL
>>> to avoid memory leaks.
>>
>> Thanks, should mention that this isn't final by any stretch (which is
>> why it hasn't been posted yet), just pushed out for some exposure.
> 
> I don't know what other people think about this, but I do find a bit
> discouraging in testing those half-baked patches in linux-next where it does not
> even ready to post for a review.

I don't disagree with that and this doesn't normally happen. I don't
want to get into the reasonings, but things had to be shuffled which is
why they ended up in -next before being posted this week. They will go
out shortly.

-- 
Jens Axboe