Re: [PATCH] arch: um: kernel: skas: mmu: remove pmd_free() and pud_free() for failure processing in init_stub_pte()

[Richard Weinberger <richard@nod.at>]

Am 14.11.2013 08:33, schrieb Chen Gang:
> On 11/14/2013 02:48 PM, Chen Gang wrote:
>>> >From the look of it, if an error did occur in init_stub_pte(),
>>>> then the special mapping of STUB_CODE and STUB_DATA would not
>>>> be installed, so this area would be invisible to munmap and exit,
>>>> and with your patch then the pages allocated likely to be leaked.
>>>>
>> It sounds reasonable to me: "although 'pgd' related with 'mm', but they
>> are not installed". But just like you said originally: "better get ACK
>> from some mm guys".
>>
>>
>> Hmm... is it another issue: "after STUB_CODE succeeds, but STUB_DATA
>> fails, the STUB_CODE will be leaked".
>>
>>
>>>> Which is not to say that the existing code is actually correct:
>>>> you're probably right that it's technically wrong.  But it would
>>>> be very hard to get init_stub_pte() to fail, and has anyone
>>>> reported a problem with it?  My guess is not, and my own
>>>> inclination to dabble here is zero.
>>>>
>> Yeah.
>>
> 
> If we can not get ACK from any mm guys, and we have no enough time
> resource to read related source code, for me, I still recommend to
> remove p?d_free() in failure processing.

It's rather easy, does your commit fix a real problem you are facing?
If the answer is "yes" we can talk.

Chen, If you really want to help us, please investigate into existing/real problems.
Toralf does a very good job in finding strange issues using trinity.
You could help him resolving the issue described in that thread:
"[uml-devel] fuzz tested 32 bit user mode linux image hangs in radix_tree_next_chunk()"

Thanks,
//richard