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, 23 Jun 2021 15:14:55 +0200
From:   David Sterba <dsterba@suse.cz>
To:     Stephen Rothwell <sfr@canb.auug.org.au>
Cc:     "Gustavo A. R. Silva" <gustavo@embeddedor.com>,
        Kees Cook <keescook@chromium.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Linux Next Mailing List <linux-next@vger.kernel.org>
Subject: Re: linux-next: build warning after merge of the kspp-gustavo tree
Message-ID: <20210623131455.GM28158@suse.cz>
Reply-To: dsterba@suse.cz
Mail-Followup-To: dsterba@suse.cz, Stephen Rothwell <sfr@canb.auug.org.au>,
        "Gustavo A. R. Silva" <gustavo@embeddedor.com>,
        Kees Cook <keescook@chromium.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Linux Next Mailing List <linux-next@vger.kernel.org>
References: <20210623083901.1d49d19d@canb.auug.org.au>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20210623083901.1d49d19d@canb.auug.org.au>
User-Agent: Mutt/1.5.23.1-rc1 (2014-03-12)
Precedence: bulk

On Wed, Jun 23, 2021 at 08:39:01AM +1000, Stephen Rothwell wrote:
> Hi all,
> 
> After merging the kspp-gustavo tree, today's linux-next build (powerpc
> ppc64_defconfig) produced this warning:
> 
> In file included from fs/btrfs/ctree.h:9,
>                  from fs/btrfs/struct-funcs.c:8:
> fs/btrfs/struct-funcs.c: In function 'btrfs_get_token_16':
> fs/btrfs/struct-funcs.c:80:46: warning: array subscript 1 is above array bounds of 'struct page *[1]' [-Warray-bounds]
>    80 |  token->kaddr = page_address(token->eb->pages[idx + 1]);  \
>       |                              ~~~~~~~~~~~~~~~~^~~~~~~~~

The warning is correct, on powerpc and 64k pages the array has only 1
item, ie. only index 0 is valid. The overflow won't happen in practice
though because of previous branch that would happen in 100% cases. The
code handles when some bytes cross 2 pages but on 64k pages it's all
just one page.

To allow the warning to be enabled globally we'll fix it, I'll let
Gustavo know once it's done.