Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Wed, 16 Jun 2021 11:27:26 -0700
From:   Rustam Kovhaev <rkovhaev@gmail.com>
To:     David Hildenbrand <david@redhat.com>
Cc:     Catalin Marinas <catalin.marinas@arm.com>,
        Andrew Morton <akpm@linux-foundation.org>, linux-mm@kvack.org,
        linux-kernel@vger.kernel.org, dvyukov@google.com,
        gregkh@linuxfoundation.org
Subject: Re: kmemleak memory scanning
Message-ID: <YMpCjmCrhPwpCiid@nuc10>
References: <YMe8ktUsdtwFKHuF@nuc10>
 <f60e2a4f-2eb8-1097-60b5-eee998478068@redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <f60e2a4f-2eb8-1097-60b5-eee998478068@redhat.com>
Precedence: bulk

On Tue, Jun 15, 2021 at 10:12:06AM +0200, David Hildenbrand wrote:
> On 14.06.21 22:31, Rustam Kovhaev wrote:
> > hello Catalin, Andrew!
> > 
> > while troubleshooting a false positive syzbot kmemleak report i have
> > noticed an interesting behavior in kmemleak and i wonder whether it is
> > behavior by design and should be documented, or maybe something to
> > improve.
> 
> Hi,
> 
> See below regarding documentation.
> 
> > apologies if some of the questions do not make sense, i am still going
> > through kmemleak code..
> > 
> > a) kmemleak scans struct page (kmemleak.c:1462), but it does not scan
> > the actual contents (page_address(page)) of the page.
> > if we allocate an object with kmalloc(), then allocate page with
> > alloc_page(), and if we put kmalloc pointer somewhere inside that page,
> > kmemleak will report kmalloc pointer as a false positive.
> > should we improve kmemleak and make it scan page contents?
> > or will this bring too many false negatives?
> 
> I looked into this a while ago to see which parts of the kernel end up
> reading random physical page content and was happy to see that kmemleak does
> *not* scan random physical memory :)
> 
> We have to be very careful when reading random physical page content,
> especially in virt environments this is really undesired, or when dealing
> with memory holes, memory with problematic semantics like gart memory ...

i see, makes sense, thank you for the info!

> 
> The doc (Documentation/dev-tools/kmemleak.rst) states "Page allocations and
> ioremap are not tracked.", which includes the alloc_page() example you gave
> I think.

i see it now, ty!