[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <52DC8AF9.3040807@oracle.com>
Date: Mon, 20 Jan 2014 10:33:29 +0800
From: annie li <annie.li@...cle.com>
To: David Vrabel <david.vrabel@...rix.com>
CC: Wei Liu <wei.liu2@...rix.com>, ian.campbell@...rix.com,
netdev@...r.kernel.org, xen-devel@...ts.xen.org,
andrew.bennieston@...rix.com, davem@...emloft.net
Subject: Re: [Xen-devel] [PATCH net-next v2] xen-netfront: clean up code in
xennet_release_rx_bufs
On 2014/1/18 1:50, David Vrabel wrote:
> On 17/01/14 15:43, annie li wrote:
>> No, I am trying to implement 2 patches.
> I don't understand the need for two patches here, particularly when
> the first patch introduces a security issue.
This is basically connected with personal taste. I am thinking that my
original patch is removing unnecessary code for grant transfer and then
keep rx release consistent with tx path, the security issue you
mentioned exist in current tx too. The second one is to change
gnttab_end_foreign_access and netfront tx/rx, blkfront, etc. But if you
like to merge them together, I can do that.
Thanks
Annie
> You can fold the following
> (untested) patch into your v2 patch and give it a try?
>
> Thanks.
>
> David
>
> 8<----------------------
> xen-netfront: prevent unsafe reuse of rx buf pages after uninit
>
> ---
> drivers/net/xen-netfront.c | 21 +++++++++++++++++----
> 1 files changed, 17 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/net/xen-netfront.c b/drivers/net/xen-netfront.c
> index 692589e..47aa599 100644
> --- a/drivers/net/xen-netfront.c
> +++ b/drivers/net/xen-netfront.c
> @@ -1134,19 +1134,32 @@ static void xennet_release_tx_bufs(struct netfront_info *np)
>
> static void xennet_release_rx_bufs(struct netfront_info *np)
> {
> - struct sk_buff *skb;
> int id, ref;
>
> spin_lock_bh(&np->rx_lock);
>
> for (id = 0; id < NET_RX_RING_SIZE; id++) {
> + struct sk_buff *skb;
> + skb_frag_t *frag;
> + const struct page *page;
> +
> + skb = np->rx_skbs[id];
> + if (!skb)
> + continue;
> +
> ref = np->grant_rx_ref[id];
> if (ref == GRANT_INVALID_REF)
> continue;
>
> - skb = np->rx_skbs[id];
> - gnttab_end_foreign_access_ref(ref, 0);
> - gnttab_release_grant_reference(&np->gref_rx_head, ref);
> + frag = &skb_shinfo(skb)->frags[0];
> + page = skb_frag_page(frag);
> +
> + /* gnttab_end_foreign_access() needs a page ref until
> + * foreign access is ended (which may be deferred).
> + */
> + get_page(page);
> +
> + gnttab_end_foreign_access(ref, 0, page);
> np->grant_rx_ref[id] = GRANT_INVALID_REF;
>
> kfree_skb(skb);
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists