From: Eric Dumazet <edumazet@xxxxxxxxxx>
[ Upstream commit 38ec4944b593fd90c5ef42aaaa53e66ae5769d04 ]
After commit 0f6925b3e8da ("virtio_net: Do not pull payload in skb->head")
Guenter Roeck reported one failure in his tests using sh architecture.
After much debugging, we have been able to spot silent unaligned accesses
in inet_gro_receive()
The issue at hand is that upper networking stacks assume their header
is word-aligned. Low level drivers are supposed to reserve NET_IP_ALIGN
bytes before the Ethernet header to make that happen.
This patch hardens skb_gro_reset_offset() to not allow frag0 fast-path
if the fragment is not properly aligned.
Some arches like x86, arm64 and powerpc do not care and define NET_IP_ALIGN
as 0, this extra check will be a NOP for them.
Note that if frag0 is not used, GRO will call pskb_may_pull()
as many times as needed to pull network and transport headers.
[ Backport note ]
A small conflict has been reported:
++<<<<<<< HEAD
+ if (skb_mac_header(skb) == skb_tail_pointer(skb) &&
+ pinfo->nr_frags &&
+ !PageHighMem(skb_frag_page(frag0))) {
++=======
+ if (!skb_headlen(skb) && pinfo->nr_frags &&
+ !PageHighMem(skb_frag_page(frag0)) &&
+ (!NET_IP_ALIGN || !(skb_frag_off(frag0) & 3))) {
++>>>>>>> 38ec4944b593 (gro: ensure frag0 meets IP header alignment)
This is expected because older kernels are missing
commit 8aef998df3979 ("net: core: allow fast GRO for skbs with Ethernet header in head").
This commit modifies the beginning of the 'if' statement.
The resolution was easy: the patch we want to backport here is
adding new conditions to the 'if' statement:
&& (!NET_IP_ALIGN || !(skb_frag_off(frag0) & 3))
We simply append these new conditions to it on older kernels.
Another issue had to be resolved: skb_frag_off() is used in this
patch we want to backport but this function is not defined in
kernels < 5.4. It has then been extracted and imported from
commit 7240b60c98d63 ("linux: Add skb_frag_t page_offset accessors").
Fixes: 0f6925b3e8da ("virtio_net: Do not pull payload in skb->head")
Fixes: 78a478d0efd9 ("gro: Inline skb_gro_header and cache frag0 virtual address")
Signed-off-by: Eric Dumazet <edumazet@xxxxxxxxxx>
Reported-by: Guenter Roeck <linux@xxxxxxxxxxxx>
Cc: Xuan Zhuo <xuanzhuo@xxxxxxxxxxxxxxxxx>
Cc: "Michael S. Tsirkin" <mst@xxxxxxxxxx>
Cc: Jason Wang <jasowang@xxxxxxxxxx>
Acked-by: Michael S. Tsirkin <mst@xxxxxxxxxx>
Tested-by: Guenter Roeck <linux@xxxxxxxxxxxx>
Signed-off-by: David S. Miller <davem@xxxxxxxxxxxxx>
Signed-off-by: Matthieu Baerts <matthieu.baerts@xxxxxxxxxxxx>
---
include/linux/skbuff.h | 9 +++++++++
net/core/dev.c | 3 ++-
2 files changed, 11 insertions(+), 1 deletion(-)
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 06176ef2a842..5f2e6451ece5 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -2788,6 +2788,15 @@ static inline void skb_propagate_pfmemalloc(struct page *page,
skb->pfmemalloc = true;
}
+/**
+ * skb_frag_off() - Returns the offset of a skb fragment
+ * @frag: the paged fragment
+ */
+static inline unsigned int skb_frag_off(const skb_frag_t *frag)
+{
+ return frag->page_offset;
+}
+
/**
* skb_frag_page - retrieve the page referred to by a paged fragment
* @frag: the paged fragment
diff --git a/net/core/dev.c b/net/core/dev.c
index 722ae0b57f3f..a6798117bb1a 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -5400,7 +5400,8 @@ static void skb_gro_reset_offset(struct sk_buff *skb)
if (skb_mac_header(skb) == skb_tail_pointer(skb) &&
pinfo->nr_frags &&
- !PageHighMem(skb_frag_page(frag0))) {
+ !PageHighMem(skb_frag_page(frag0)) &&
+ (!NET_IP_ALIGN || !(skb_frag_off(frag0) & 3))) {
NAPI_GRO_CB(skb)->frag0 = skb_frag_address(frag0);
NAPI_GRO_CB(skb)->frag0_len = min_t(unsigned int,
skb_frag_size(frag0),
--
2.31.1
