Observation of a memory leak with commit e98e49b2bbf7 ("io_uring: extend task put optimisations")

Observation of a memory leak with commit e98e49b2bbf7 ("io_uring: extend task put optimisations")

[Lukas Bulwahn]

Dear Pavel, dear Jens,

In our syzkaller instance running on linux-next,
https://elisa-builder-00.iol.unh.edu/syzkaller-next/, we have been
observing a memory leak in copy_process for quite some time.

It is reproducible on v5.15-rc1, v5.15, v5.16-rc8 and next-20220106:

https://elisa-builder-00.iol.unh.edu/syzkaller-next/crash?id=1169da08a3e72457301987b70bcce62f0f49bdbb

So, it is in mainline, was released and has not been fixed in linux-next yet.

As syzkaller also provides a reproducer, we bisected this memory leak
to be introduced with commit e98e49b2bbf7 ("io_uring: extend task put
optimisations").

Could you please have a look how your commit introduces this memory
leak? We will gladly support testing your fix in case help is needed.


Best regards,

Lukas

Re: Observation of a memory leak with commit e98e49b2bbf7 ("io_uring: extend task put optimisations")

[Pavel Begunkov]

On 1/7/22 15:00, Lukas Bulwahn wrote:
> Dear Pavel, dear Jens,
> 
> In our syzkaller instance running on linux-next,
> https://elisa-builder-00.iol.unh.edu/syzkaller-next/, we have been
> observing a memory leak in copy_process for quite some time.
> 
> It is reproducible on v5.15-rc1, v5.15, v5.16-rc8 and next-20220106:
> 
> https://elisa-builder-00.iol.unh.edu/syzkaller-next/crash?id=1169da08a3e72457301987b70bcce62f0f49bdbb
> 
> So, it is in mainline, was released and has not been fixed in linux-next yet.
> 
> As syzkaller also provides a reproducer, we bisected this memory leak
> to be introduced with commit e98e49b2bbf7 ("io_uring: extend task put
> optimisations").
> 
> Could you please have a look how your commit introduces this memory
> leak? We will gladly support testing your fix in case help is needed.

Thanks for letting know
I think I know what it is, will get back later.

-- 
Pavel Begunkov

Re: Observation of a memory leak with commit e98e49b2bbf7 ("io_uring: extend task put optimisations")

[Pavel Begunkov]

On 1/7/22 16:22, Pavel Begunkov wrote:
> On 1/7/22 15:00, Lukas Bulwahn wrote:
>> Dear Pavel, dear Jens,
>>
>> In our syzkaller instance running on linux-next,
>> https://elisa-builder-00.iol.unh.edu/syzkaller-next/, we have been
>> observing a memory leak in copy_process for quite some time.
>>
>> It is reproducible on v5.15-rc1, v5.15, v5.16-rc8 and next-20220106:
>>
>> https://elisa-builder-00.iol.unh.edu/syzkaller-next/crash?id=1169da08a3e72457301987b70bcce62f0f49bdbb
>>
>> So, it is in mainline, was released and has not been fixed in linux-next yet.
>>
>> As syzkaller also provides a reproducer, we bisected this memory leak
>> to be introduced with commit e98e49b2bbf7 ("io_uring: extend task put
>> optimisations").
>>
>> Could you please have a look how your commit introduces this memory
>> leak? We will gladly support testing your fix in case help is needed.
> 
> Thanks for letting know
> I think I know what it is, will get back later.

https://lore.kernel.org/io-uring/69f226b35fbdb996ab799a8bbc1c06bf634ccec1.1641688805.git.asml.silence@gmail.com

Should fix it, tested with the repro

-- 
Pavel Begunkov