linux-kernel - Re: lightnvm: pblk: avoid ref warning on cache creation

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Fri, 23 Nov 2018 09:41:15 +0100
From:   Matias Bjørling <mb@...htnvm.io>
To:     javier@...xlabs.com
Cc:     javier@...igon.com, hans.holmberg@...xlabs.com,
        linux-block@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: lightnvm: pblk: avoid ref warning on cache creation

On 11/23/2018 09:31 AM, Javier Gonzalez wrote:
> 
> 
>> On 23 Nov 2018, at 09.24, Matias Bjørling <mb@...htnvm.io> wrote:
>>
>> On 11/23/2018 09:16 AM, Javier Gonzalez wrote:
>>>> On 23 Nov 2018, at 09.14, Matias Bjørling <mb@...htnvm.io> wrote:
>>>>
>>>>> On 11/22/2018 02:46 PM, Javier González wrote:
>>>>> Matias,
>>>>> Can you pick this up for 4.20? Even though it is not an error per se, it
>>>>> does trigger an ugly false positive warning when CONFIG_REFCOUNT_FULL is
>>>>> set.
>>>>> Thanks,
>>>>> Javier
>>>>> Javier González (1):
>>>>>    lightnvm: pblk: avoid ref warning on cache creation
>>>>>   drivers/lightnvm/pblk-init.c | 14 +++++---------
>>>>>   1 file changed, 5 insertions(+), 9 deletions(-)
>>>>
>>>> It is too late to get it for 4.20. I'll pull it in for 4.21.
>>> Well, if it is a fix it’s not too late AFAIK. It also applies to a patch picked up in this series. We should have catches this before.
>>> But I guess it’s your call.
>>
>> There is a likelihood that Linus would catch it, with the result of fire & fury. I don't feel like using one of my cat life's for this one.
> 
> I don’t get how a fix in rc3 will result in Jens or Linus complaining. But Ok, let’s release 4.20 with a false positive warning for use-after-free on pblk creation.
> 

It is a warning that would trigger only for kernel developers. The 
REFCOUNT_FULL is used when debugging code, and is usually not on in 
production kernels.

To my understanding, the rc's, are meant for errors that would affect 
end-users.

> 
>>>> We can put on a Fixes: 1864de94ec9d6 "lightnvm: pblk: stop recreating global caches" if you want Greg and friends to pick picked it up for stable kernels?
>>> I didn’t add it as it is not a bug in itself. But it is good to add it. Can you do it.
>>
>> I'll do it.
> 
> Thanks.
> 
> Javier.
>