[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <d1668ff4-0aac-4727-9886-2f04c5f1104f@redhat.com>
Date: Mon, 26 Aug 2024 07:53:43 +0800
From: Xiubo Li <xiubli@...hat.com>
To: Luis Henriques <luis.henriques@...ux.dev>
Cc: Ilya Dryomov <idryomov@...il.com>, ceph-devel@...r.kernel.org,
linux-kernel@...r.kernel.org
Subject: Re: [RFC PATCH] ceph: fix out-of-bound array access when doing a file
read
On 8/23/24 15:25, Luis Henriques wrote:
> On Fri, Aug 23 2024, Xiubo Li wrote:
>
>> On 8/22/24 23:01, Luis Henriques (SUSE) wrote:
>>> If, while doing a read, the inode is updated and the size is set to zero,
>>> __ceph_sync_read() may not be able to handle it. It is thus easy to hit a
>>> NULL pointer dereferrence by continuously reading a file while, on another
>>> client, we keep truncating and writing new data into it.
>>>
>>> This patch fixes the issue by adding extra checks to avoid integer overflows
>>> for the case of a zero size inode. This will prevent the loop doing page
>>> copies from running and thus accessing the pages[] array beyond num_pages.
>>>
>>> Link: https://tracker.ceph.com/issues/67524
>>> Signed-off-by: Luis Henriques (SUSE) <luis.henriques@...ux.dev>
>>> ---
>>> Hi!
>>>
>>> Please note that this patch is only lightly tested and, to be honest, I'm
>>> not sure if this is the correct way to fix this bug. For example, if the
>>> inode size is 0, then maybe ceph_osdc_wait_request() should have returned
>>> 0 and the problem would be solved. However, it seems to be returning the
>>> size of the reply message and that's not something easy to change. Or maybe
>>> I'm just reading it wrong. Anyway, this is just an RFC to see if there's
>>> other ideas.
>>>
>>> Also, the tracker contains a simple testcase for crashing the client.
>>>
>>> fs/ceph/file.c | 7 ++++---
>>> 1 file changed, 4 insertions(+), 3 deletions(-)
>>>
>>> diff --git a/fs/ceph/file.c b/fs/ceph/file.c
>>> index 4b8d59ebda00..dc23d5e5b11e 100644
>>> --- a/fs/ceph/file.c
>>> +++ b/fs/ceph/file.c
>>> @@ -1200,9 +1200,9 @@ ssize_t __ceph_sync_read(struct inode *inode, loff_t *ki_pos,
>>> }
>>> idx = 0;
>>> - if (ret <= 0)
>>> + if ((ret <= 0) || (i_size == 0))
>> Hi Luis,
>>
>> This change looks incorrect to me.
>>
>> As I mentioned before when the 'IFILE' lock is in MIX state the 'Frw' caps could
>> be issued to multiple clients at the same time. Which means the file could be
>> updated by another client and the local 'i_size' may haven't been changed in
>> time. So in this case the 'ret' will be larger than '0' and the 'i_size' could
>> be '0'.
>>
>>
>>> left = 0;
>>> - else if (off + ret > i_size)
>>> + else if ((i_size >= off) && (off + ret > i_size))
>> And the 'off' also could equal to little than the 'i_size'.
>>
>> BTW, could you reproduce the crash issue ?
> Yes, 100% reproducible :-)
>
> See https://tracker.ceph.com/issues/67524
Okay, Let me have a try about this.
Thanks
>
> Cheers,
Powered by blists - more mailing lists