Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [PATCH] x86/xen: Remove use of VLAs
To:     David Brown <david.brown@linaro.org>
Cc:     Boris Ostrovsky <boris.ostrovsky@oracle.com>,
        Juergen Gross <jgross@suse.com>,
        Thomas Gleixner <tglx@linutronix.de>,
        Ingo Molnar <mingo@redhat.com>,
        "H. Peter Anvin" <hpa@zytor.com>, x86@kernel.org,
        xen-devel@lists.xenproject.org, linux-kernel@vger.kernel.org,
        kernel-hardening@lists.openwall.com
References: <20180413221146.28476-1-labbott@redhat.com>
 <20180414025553.GA32653@davidb.org>
From:   Laura Abbott <labbott@redhat.com>
Message-ID: <18e8c47d-55f7-795a-053a-f667650b43b7@redhat.com>
Date:   Fri, 13 Apr 2018 20:43:02 -0700
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101
 Thunderbird/52.7.0
MIME-Version: 1.0
In-Reply-To: <20180414025553.GA32653@davidb.org>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 04/13/2018 07:55 PM, David Brown wrote:
> On Fri, Apr 13, 2018 at 03:11:46PM -0700, Laura Abbott wrote:
> 
>> There's an ongoing effort to remove VLAs[1] from the kernel to eventually
>> turn on -Wvla. The few VLAs in use have an upper bound based on a size
>> of 64K. This doesn't produce an excessively large stack so just switch
>> the upper bound.
>>
>> [1] https://lkml.org/lkml/2018/3/7/621
> 
> This comment is more in regards to many of these patches, and not as
> much this one specifically.
> 
> How confident are we in the upper bounds we're setting, and how
> obvious is it in the resulting code so that something does later
> change to overflow these bounds.
> 
> The danger here is that we're converting something a little easier to
> detect (a stack overflow), with something harder to detect
> (overflowing an array on the stack).
> 

Several people have remarked on that and the solution has been to
put in some kind of WARN and/or error check to make it obvious something
needs to be adjusted.

> I guess the question is twofold: how did you determine that 64K was
> the largest 'size' value, and how should reviewers verify this as
> well.  Perhaps this should at least be in the commit text so someone
> tracking down something with this code can find it later.
> 

It's not in the patch context but there's a large comment below:

         /*
          * A GDT can be up to 64k in size, which corresponds to 8192
          * 8-byte entries, or 16 4k pages..
          */

         BUG_ON(size > 65536);


Given the frames was calculated based off the size, that seemed
sufficient.

> David

Thanks,
Laura