linux-kernel - Re: [GIT PULL] hotfixes for 6.2

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] [day] [month] [year] [list]

Message-ID: <dd2ba3b9-8fab-ce3b-4f01-eadbb527f3bc@linux.alibaba.com>
Date:   Tue, 14 Feb 2023 09:26:03 +0800
From:   Baolin Wang <baolin.wang@...ux.alibaba.com>
To:     Linus Torvalds <torvalds@...ux-foundation.org>,
        Andrew Morton <akpm@...ux-foundation.org>
Cc:     linux-mm@...ck.org, mm-commits@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [GIT PULL] hotfixes for 6.2



On 2/14/2023 6:19 AM, Linus Torvalds wrote:
> On Mon, Feb 13, 2023 at 2:08 PM Andrew Morton <akpm@...ux-foundation.org> wrote:
>>
>> Kuan-Ying Lee (1):
>>        mm/gup: add folio to list when folio_isolate_lru() succeed
> 
> Ugh. I really hate fixes like this.
> 
> The problem came from mis-understanding the return value of
> folio_isolate_lru(), and thinking that it was a boolean
> success/failure thing.
> 
> It wasn't, it was an integer "success/errno" thing, and the sense of
> the test was wrong. So the patch is
> 
> -       if (!folio_isolate_lru(folio))
> +       if (folio_isolate_lru(folio))
>                  continue;
> 
> but at no point was the code *clarified*.
> 
> Wouldn't it have been much better to write the new code to be
> 
>          if (folio_isolate_lru(folio) < 0)
>                  continue;
> 
> to actually make it clear that this is a "negative error return check".
> 
> I've pulled this, but I really think that when somebody notices that
> we had a silly bug because of a misunderstanding like this, it's not
> just that the bug should be fixed, the code should also be *clarified*
> at the same time.

Yes, agree, I need to check the return value of folio_isolate_lru() 
every time when looking at the code. I can help to create a patch to 
make it clear for all users.