linux-kernel - Re: [PATCH RFC v2] tpm: tpm_vtpm_proxy: do not reference kernel memory as user memory

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: <edca8d02-01d9-c328-3d3e-015d0ed61a0b@linux.ibm.com>
Date:   Wed, 31 May 2023 13:20:47 -0400
From:   Stefan Berger <stefanb@...ux.ibm.com>
To:     Jarkko Sakkinen <jarkko@...nel.org>,
        linux-integrity@...r.kernel.org
Cc:     Jason Gunthorpe <jgg@...dia.com>,
        Alejandro Cabrera <alejandro.cabreraaldaya@...i.fi>,
        Jarkko Sakkinen <jarkko.sakkinen@...i.fi>,
        stable@...r.kernel.org, Stefan Berger <stefanb@...ux.vnet.ibm.com>,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH RFC v2] tpm: tpm_vtpm_proxy: do not reference kernel
 memory as user memory



On 5/31/23 13:17, Jarkko Sakkinen wrote:
> On Tue, 2023-05-30 at 13:45 -0400, Stefan Berger wrote:
>>
>> On 5/29/23 22:01, Jarkko Sakkinen wrote:
>>> From: Jarkko Sakkinen <jarkko.sakkinen@...i.fi>
>>>
>>
>>> -	rc = copy_to_user(buf, proxy_dev->buffer, len);
>>> +	if (buf)
>>> +		rc = copy_to_user(buf, proxy_dev->buffer, len);
>>> +
>>
>> Looking through other drivers it seems buf is always expected to be a valid non-NULL pointer on file_operations.read().
>>
>>
>> https://elixir.bootlin.com/linux/latest/source/arch/x86/mm/tlb.c#L1279   simple_read_from_buffer will pass the pointer to the user buffer along and it ('to') ends up in copy_to_user(to, ...);
>>
>>
>> Same here: https://elixir.bootlin.com/linux/latest/source/security/integrity/ima/ima_fs.c#L41
> 
> It is good to mention here that IMA uses __user tagged pointers
> correctly, and it does not really compare to the vtpm driver code
> by any possible means. So let's not add illegit comparison points.

Yes, sir. Did you read David' response?
> 
> BR, Jarkko