linux-kernel - Re: Coverity: zram

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]

Date:   Thu, 10 Nov 2022 19:16:16 -0800
From:   Kees Cook <keescook@...omium.org>
To:     Sergey Senozhatsky <senozhatsky@...omium.org>
Cc:     Alexey Romanov <avromanov@...rdevices.ru>,
        linux-kernel@...r.kernel.org, Nick Terrell <terrelln@...com>,
        Minchan Kim <minchan@...nel.org>,
        Suleiman Souhlal <suleiman@...gle.com>,
        Nitin Gupta <ngupta@...are.org>, Jens Axboe <axboe@...nel.dk>,
        Nhat Pham <nphamcs@...il.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        linux-block@...r.kernel.org,
        "Gustavo A. R. Silva" <gustavo@...eddedor.com>,
        linux-next@...r.kernel.org, linux-hardening@...r.kernel.org
Subject: Re: Coverity: zram_recompress(): OVERRUN

On Fri, Nov 11, 2022 at 09:42:38AM +0900, Sergey Senozhatsky wrote:
> On (22/11/11 09:37), Sergey Senozhatsky wrote:
> > On (22/11/10 08:47), coverity-bot wrote:
> > [..]
> > > 1704     	class_index_old = zs_lookup_class_index(zram->mem_pool, comp_len_old);
> > > 1705     	/*
> > > 1706     	 * Iterate the secondary comp algorithms list (in order of priority)
> > > 1707     	 * and try to recompress the page.
> > > 1708     	 */
> > > 1709     	for (; prio < prio_max; prio++) {
> > > vvv     CID 1527270:    (OVERRUN)
> > > vvv     Overrunning array "zram->comps" of 4 8-byte elements at element index 4 (byte offset 39) using index "prio" (which evaluates to 4).
> > > 1710     		if (!zram->comps[prio])
> > > 1711     			continue;
> > > 1712
> > > 1713     		/*
> > > 1714     		 * Skip if the object is already re-compressed with a higher
> > > 1715     		 * priority algorithm (or same algorithm).
> > 
> > prio_max is always limited and max value it can have is 4 (ZRAM_MAX_COMPS).
> > Depending on use case we can limit prio_max even to lower values.
> > 
> > So we have
> > 
> > 	for (; prio < 4; prio++) {
> > 		foo = comps[prio];
> > 	}
> > 
> > I don't see how prio can be 4 inside of this loop.
> 
> Kees, if we do something like this will it make coverity happy?
> 
> ---
> 
> diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c
> index 9d33801e8ba8..e67a124f2e88 100644
> --- a/drivers/block/zram/zram_drv.c
> +++ b/drivers/block/zram/zram_drv.c
> @@ -1706,6 +1706,7 @@ static int zram_recompress(struct zram *zram, u32 index, struct page *page,
>  	 * Iterate the secondary comp algorithms list (in order of priority)
>  	 * and try to recompress the page.
>  	 */
> +	prio_max = min(prio_max, ZRAM_MAX_COMPS);
>  	for (; prio < prio_max; prio++) {
>  		if (!zram->comps[prio])
>  			continue;

It would, but given this is a clear false positive, don't feel the need
to add this just for Coverity's sake. It is a nice bit of added
robustness, but I leave that decision up to you! :)

-- 
Kees Cook