netdev - Re: [PATCH v3 9/9] vhost: add WARNING if log

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]

Message-ID: <e00d882e-9ce7-48b0-bc2f-bf937ff6b9c3@oracle.com>
Date: Mon, 14 Apr 2025 09:52:04 -0700
From: Dongli Zhang <dongli.zhang@...cle.com>
To: "Michael S. Tsirkin" <mst@...hat.com>
Cc: virtualization@...ts.linux.dev, kvm@...r.kernel.org,
        netdev@...r.kernel.org, jasowang@...hat.com,
        michael.christie@...cle.com, pbonzini@...hat.com, stefanha@...hat.com,
        eperezma@...hat.com, joao.m.martins@...cle.com, joe.jin@...cle.com,
        si-wei.liu@...cle.com, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3 9/9] vhost: add WARNING if log_num is more than limit

Hi Michael,

On 4/14/25 9:32 AM, Michael S. Tsirkin wrote:
> On Wed, Apr 02, 2025 at 11:29:54PM -0700, Dongli Zhang wrote:
>> Since long time ago, the only user of vq->log is vhost-net. The concern is
>> to add support for more devices (i.e. vhost-scsi or vsock) may reveals
>> unknown issue in the vhost API. Add a WARNING.
>>
>> Suggested-by: Joao Martins <joao.m.martins@...cle.com>
>> Signed-off-by: Dongli Zhang <dongli.zhang@...cle.com>
> 
> 
> Userspace can trigger this I think, this is a problem since
> people run with reboot on warn.

I think it will be a severe kernel bug (page fault) if userspace can trigger this.

If (*log_num >= vq->dev->iov_limit), the next line will lead to an out-of-bound
memory access:

    log[*log_num].addr = vhost64_to_cpu(vq, desc.addr);

I could not propose a case to trigger the WARNING from userspace. Would you mind
helping explain if that can happen?

> Pls grammar issues in comments... I don't think so.

I did an analysis of code and so far I could not identify any case to trigger
(*log_num >= vq->dev->iov_limit).

The objective of the patch is to add a WARNING to double confirm the case won't
happen.

Regarding "I don't think so", would you mean we don't need this patch/WARNING
because the code is robust enough?

Thank you very much!

Dongli Zhang

> 
>> ---
>>  drivers/vhost/vhost.c | 18 ++++++++++++++++++
>>  1 file changed, 18 insertions(+)
>>
>> diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c
>> index 494b3da5423a..b7d51d569646 100644
>> --- a/drivers/vhost/vhost.c
>> +++ b/drivers/vhost/vhost.c
>> @@ -2559,6 +2559,15 @@ static int get_indirect(struct vhost_virtqueue *vq,
>>  		if (access == VHOST_ACCESS_WO) {
>>  			*in_num += ret;
>>  			if (unlikely(log && ret)) {
>> +				/*
>> +				 * Since long time ago, the only user of
>> +				 * vq->log is vhost-net. The concern is to
>> +				 * add support for more devices (i.e.
>> +				 * vhost-scsi or vsock) may reveals unknown
>> +				 * issue in the vhost API. Add a WARNING.
>> +				 */
>> +				WARN_ON_ONCE(*log_num >= vq->dev->iov_limit);
>> +
>>  				log[*log_num].addr = vhost64_to_cpu(vq, desc.addr);
>>  				log[*log_num].len = vhost32_to_cpu(vq, desc.len);
>>  				++*log_num;
>> @@ -2679,6 +2688,15 @@ int vhost_get_vq_desc(struct vhost_virtqueue *vq,
>>  			 * increment that count. */
>>  			*in_num += ret;
>>  			if (unlikely(log && ret)) {
>> +				/*
>> +				 * Since long time ago, the only user of
>> +				 * vq->log is vhost-net. The concern is to
>> +				 * add support for more devices (i.e.
>> +				 * vhost-scsi or vsock) may reveals unknown
>> +				 * issue in the vhost API. Add a WARNING.
>> +				 */
>> +				WARN_ON_ONCE(*log_num >= vq->dev->iov_limit);
>> +
>>  				log[*log_num].addr = vhost64_to_cpu(vq, desc.addr);
>>  				log[*log_num].len = vhost32_to_cpu(vq, desc.len);
>>  				++*log_num;
>> -- 
>> 2.39.3
>