In esp_remove_trailer(), to avoid an unexpected result returned by
pskb_trim, we should check the return value of pskb_trim().
Signed-off-by: Ma Ke <[email protected]>
---
net/ipv6/esp6.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/net/ipv6/esp6.c b/net/ipv6/esp6.c
index fddd0cbdede1..e023d29e919c 100644
--- a/net/ipv6/esp6.c
+++ b/net/ipv6/esp6.c
@@ -770,7 +770,9 @@ static inline int esp_remove_trailer(struct sk_buff *skb)
skb->csum = csum_block_sub(skb->csum, csumdiff,
skb->len - trimlen);
}
- pskb_trim(skb, skb->len - trimlen);
+ ret = pskb_trim(skb, skb->len - trimlen);
+ if (unlikely(ret))
+ return ret;
ret = nexthdr[1];
--
2.37.2
On Sat, Oct 07, 2023 at 08:59:53AM +0800, Ma Ke wrote:
> In esp_remove_trailer(), to avoid an unexpected result returned by
> pskb_trim, we should check the return value of pskb_trim().
>
> Signed-off-by: Ma Ke <[email protected]>
> ---
> net/ipv6/esp6.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
We need the same change for net/ipv4/esp4.c?
Thanks.
On Sun, Oct 08, 2023 at 10:07:30AM -0700, Cong Wang wrote:
> On Sat, Oct 07, 2023 at 08:59:53AM +0800, Ma Ke wrote:
> > In esp_remove_trailer(), to avoid an unexpected result returned by
> > pskb_trim, we should check the return value of pskb_trim().
> >
> > Signed-off-by: Ma Ke <[email protected]>
> > ---
> > net/ipv6/esp6.c | 4 +++-
> > 1 file changed, 3 insertions(+), 1 deletion(-)
>
> We need the same change for net/ipv4/esp4.c?
Please verify that it can actually fail first. Note that I'm
busy right now so I haven't looked at it at all.
Thanks,
--
Email: Herbert Xu <[email protected]>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
On Tue, Oct 10, 2023 at 05:03:48PM +0800, Herbert Xu wrote:
> On Sun, Oct 08, 2023 at 10:07:30AM -0700, Cong Wang wrote:
> > On Sat, Oct 07, 2023 at 08:59:53AM +0800, Ma Ke wrote:
> > > In esp_remove_trailer(), to avoid an unexpected result returned by
> > > pskb_trim, we should check the return value of pskb_trim().
> > >
> > > Signed-off-by: Ma Ke <[email protected]>
> > > ---
> > > net/ipv6/esp6.c | 4 +++-
> > > 1 file changed, 3 insertions(+), 1 deletion(-)
> >
> > We need the same change for net/ipv4/esp4.c?
>
> Please verify that it can actually fail first. Note that I'm
> busy right now so I haven't looked at it at all.
It can, we don't linearize the packte data anymore.
On Sat, Oct 07, 2023 at 08:59:53AM +0800, Ma Ke wrote:
> In esp_remove_trailer(), to avoid an unexpected result returned by
> pskb_trim, we should check the return value of pskb_trim().
>
> Signed-off-by: Ma Ke <[email protected]>
Applied, thanks a lot!