Received-SPF: pass (google.com: domain of linux-kernel+bounces-57398-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
Message-ID: <defcaf0b-21a9-454f-b44a-c91820f541c8@linux.dev>
Date: Thu, 8 Feb 2024 10:41:49 +0800
Precedence: bulk
MIME-Version: 1.0
Subject: Re: [PATCH v4] mm/zswap: invalidate old entry when store fail or
 !zswap_enabled
Content-Language: en-US
To: Andrew Morton <akpm@linux-foundation.org>
Cc: hannes@cmpxchg.org, yosryahmed@google.com, nphamcs@gmail.com,
 linux-mm@kvack.org, linux-kernel@vger.kernel.org,
 Chengming Zhou <zhouchengming@bytedance.com>, stable@vger.kernel.org,
 Chris Li <chrisl@kernel.org>
References: <20240207115406.3865746-1-chengming.zhou@linux.dev>
 <20240207154308.bc275f3e72ec1c1fd06cf5a2@linux-foundation.org>
From: Chengming Zhou <chengming.zhou@linux.dev>
In-Reply-To: <20240207154308.bc275f3e72ec1c1fd06cf5a2@linux-foundation.org>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit

On 2024/2/8 07:43, Andrew Morton wrote:
> On Wed,  7 Feb 2024 11:54:06 +0000 chengming.zhou@linux.dev wrote:
> 
>> From: Chengming Zhou <zhouchengming@bytedance.com>
>>
>> We may encounter duplicate entry in the zswap_store():
>>
>> 1. swap slot that freed to per-cpu swap cache, doesn't invalidate
>>    the zswap entry, then got reused. This has been fixed.
>>
>> 2. !exclusive load mode, swapin folio will leave its zswap entry
>>    on the tree, then swapout again. This has been removed.
>>
>> 3. one folio can be dirtied again after zswap_store(), so need to
>>    zswap_store() again. This should be handled correctly.
>>
>> So we must invalidate the old duplicate entry before insert the
>> new one, which actually doesn't have to be done at the beginning
>> of zswap_store(). And this is a normal situation, we shouldn't
>> WARN_ON(1) in this case, so delete it. (The WARN_ON(1) seems want
>> to detect swap entry UAF problem? But not very necessary here.)
>>
>> The good point is that we don't need to lock tree twice in the
>> store success path.
>>
>> Note we still need to invalidate the old duplicate entry in the
>> store failure path, otherwise the new data in swapfile could be
>> overwrite by the old data in zswap pool when lru writeback.
>>
>> We have to do this even when !zswap_enabled since zswap can be
>> disabled anytime. If the folio store success before, then got
>> dirtied again but zswap disabled, we won't invalidate the old
>> duplicate entry in the zswap_store(). So later lru writeback
>> may overwrite the new data in swapfile.
>>
>> Fixes: 42c06a0e8ebe ("mm: kill frontswap")
>> Cc: <stable@vger.kernel.org>
> 
> We have a patch ordering issue.
> 
> As a cc:stable hotfix, this should be merged into 6.8-rcX and later
> backported into -stable trees.  So it will go
> mm-hotfixes-unstable->mm-hotfixes-stable->mainline.  So someone has to
> make this patch merge and work against latest mm-hotfixes-unstable.

Ah, right. I just sent a fix based on mm-hotfixes-unstable [1], which
is split from this patch to only include bugfix, so easy to backport.

This patch actually include two parts: bugfix and a little optimization
for the zswap_store() normal case.

Should I split this patch into two small patches and resend based on
mm-unstable?

[1] https://lore.kernel.org/all/20240208023254.3873823-1-chengming.zhou@linux.dev/

> 
> The patch you sent appears to be based on linux-next, so it has
> dependencies upon mm-unstable patches which won't be merged into
> mainline until the next merge window.
> 
> So can you please redo and retest this against mm.git's
> mm-hotfixes-unstable branch?  Then I'll try to figure out how to merge
> the gigentic pile of mm-unstable zswap changes on top of that.
> 
> Thanks.