Re: [PATCH 02/18] vchecker: introduce the valid access checker

From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
To: Andi Kleen <ak@linux.intel.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
	Andrey Ryabinin <aryabinin@virtuozzo.com>,
	Alexander Potapenko <glider@google.com>,
	Dmitry Vyukov <dvyukov@google.com>,
	kasan-dev@googlegroups.com, linux-mm@kvack.org,
	linux-kernel@vger.kernel.org, Namhyung Kim <namhyung@kernel.org>,
	Wengang Wang <wen.gang.wang@oracle.com>
Subject: Re: [PATCH 02/18] vchecker: introduce the valid access checker
Date: Wed, 29 Nov 2017 14:36:37 +0900	[thread overview]
Message-ID: <20171129053637.GA8125@js1304-P5Q-DELUXE> (raw)
In-Reply-To: <87k1yajinf.fsf@linux.intel.com>

On Tue, Nov 28, 2017 at 11:41:08AM -0800, Andi Kleen wrote:
> js1304@gmail.com writes:
> 
> > From: Joonsoo Kim <iamjoonsoo.kim@lge.com>
> 
> Looks useful. Essentially unlimited hardware break points, combined
> with slab.

Thanks!!!

> 
> Didn't do a full review, but noticed some things below.
> > +
> > +	buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
> > +	if (!buf)
> > +		return -ENOMEM;
> > +
> > +	if (copy_from_user(buf, ubuf, cnt)) {
> > +		kfree(buf);
> > +		return -EFAULT;
> > +	}
> > +
> > +	if (isspace(buf[0]))
> > +		remove = true;
> 
> and that may be uninitialized.

I will add 'cnt == 0' check above.

> and the space changes the operation? That's a strange syntax.

Intention is to clear the all the previous configuration when user
input is '\n'. Will fix it by checking '\n' directly.

> 
> > +	buf[cnt - 1] = '\0';
> 
> That's an underflow of one byte if cnt is 0.

Will add 'cnt == 0' check above.

String parsing part in this patchset will not work properly when the
last input character is not '\n'. I will fix it on the next spin.

Thanks.