Using the syzkaller repro with reduced packet size it was discovered
that XDP_PACKET_HEADROOM is not checked in tun_can_build_skb(),
although pad may be incremented in tun_build_skb(). This may end up
with exceeding the PAGE_SIZE limit in tun_build_skb().
Fixes: 7df13219d757 ("tun: reserve extra headroom only when XDP is set")
Link: https://syzkaller.appspot.com/bug?extid=f817490f5bd20541b90a
Signed-off-by: Andrew Kanner <[email protected]>
---
Notes:
v3 -> v4:
* fall back to v1, fixing only missing XDP_PACKET_HEADROOM in pad and
removing bpf_xdp_adjust_tail() check for frame_sz.
* added rcu read lock, noted by Jason Wang <[email protected]> in v1
* I decided to leave the packet length check in tun_can_build_skb()
instead of moving to tun_build_skb() suggested by Jason Wang
<[email protected]>. Otherwise extra packets will be dropped
without falling back to tun_alloc_skb(). And in the discussion of v3
Jesper Dangaard Brouer <[email protected]> noticed that XDP is ok
with a higher order pages if it's a contiguous physical memory
allocation, so falling to tun_alloc_skb() -> do_xdp_generic() should
be ok.
v2 -> v3:
* attach the forgotten changelog
v1 -> v2:
* merged 2 patches in 1, fixing both issues: WARN_ON_ONCE with
syzkaller repro and missing XDP_PACKET_HEADROOM in pad
* changed the title and description of the execution path, suggested
by Jason Wang <[email protected]>
* move the limit check from tun_can_build_skb() to tun_build_skb() to
remove duplication and locking issue, and also drop the packet in
case of a failed check - noted by Jason Wang <[email protected]>
drivers/net/tun.c | 11 ++++++++++-
1 file changed, 10 insertions(+), 1 deletion(-)
diff --git a/drivers/net/tun.c b/drivers/net/tun.c
index d75456adc62a..a1d04bc9485f 100644
--- a/drivers/net/tun.c
+++ b/drivers/net/tun.c
@@ -1582,6 +1582,9 @@ static void tun_rx_batched(struct tun_struct *tun, struct tun_file *tfile,
static bool tun_can_build_skb(struct tun_struct *tun, struct tun_file *tfile,
int len, int noblock, bool zerocopy)
{
+ struct bpf_prog *xdp_prog;
+ int pad = TUN_RX_PAD;
+
if ((tun->flags & TUN_TYPE_MASK) != IFF_TAP)
return false;
@@ -1594,7 +1597,13 @@ static bool tun_can_build_skb(struct tun_struct *tun, struct tun_file *tfile,
if (zerocopy)
return false;
- if (SKB_DATA_ALIGN(len + TUN_RX_PAD) +
+ rcu_read_lock();
+ xdp_prog = rcu_dereference(tun->xdp_prog);
+ if (xdp_prog)
+ pad += XDP_PACKET_HEADROOM;
+ rcu_read_unlock();
+
+ if (SKB_DATA_ALIGN(len + pad) +
SKB_DATA_ALIGN(sizeof(struct skb_shared_info)) > PAGE_SIZE)
return false;
--
2.39.3
On 8/1/23 4:07 PM, Andrew Kanner wrote:
> @@ -1594,7 +1597,13 @@ static bool tun_can_build_skb(struct tun_struct *tun, struct tun_file *tfile,
> if (zerocopy)
> return false;
>
> - if (SKB_DATA_ALIGN(len + TUN_RX_PAD) +
> + rcu_read_lock();
> + xdp_prog = rcu_dereference(tun->xdp_prog);
> + if (xdp_prog)
> + pad += XDP_PACKET_HEADROOM;
> + rcu_read_unlock();
since you do not care about the actual xdp_prog (only that one is set) I
believe you can use rcu_access_pointer here.
On Tue, Aug 01, 2023 at 07:07:39PM -0600, David Ahern wrote:
> On 8/1/23 4:07 PM, Andrew Kanner wrote:
> > @@ -1594,7 +1597,13 @@ static bool tun_can_build_skb(struct tun_struct *tun, struct tun_file *tfile,
> > if (zerocopy)
> > return false;
> >
> > - if (SKB_DATA_ALIGN(len + TUN_RX_PAD) +
> > + rcu_read_lock();
> > + xdp_prog = rcu_dereference(tun->xdp_prog);
> > + if (xdp_prog)
> > + pad += XDP_PACKET_HEADROOM;
> > + rcu_read_unlock();
>
>
> since you do not care about the actual xdp_prog (only that one is set) I
> believe you can use rcu_access_pointer here.
Good point. Thanks, David.
I'll resend both as v5.
The correct cc-list for PATCH 2/2 is also needed. It fixes
net/core/filter.c instead of drivers/net/tun.c now.
pw-bot: changes-requested
--
Andrew Kanner
On 02/08/2023 00.07, Andrew Kanner wrote:
> Using the syzkaller repro with reduced packet size it was discovered
> that XDP_PACKET_HEADROOM is not checked in tun_can_build_skb(),
> although pad may be incremented in tun_build_skb(). This may end up
> with exceeding the PAGE_SIZE limit in tun_build_skb().
>
> Fixes: 7df13219d757 ("tun: reserve extra headroom only when XDP is set")
> Link: https://syzkaller.appspot.com/bug?extid=f817490f5bd20541b90a
> Signed-off-by: Andrew Kanner <[email protected]>
> ---
>
> Notes:
> v3 -> v4:
> * fall back to v1, fixing only missing XDP_PACKET_HEADROOM in pad and
> removing bpf_xdp_adjust_tail() check for frame_sz.
> * added rcu read lock, noted by Jason Wang <[email protected]> in v1
> * I decided to leave the packet length check in tun_can_build_skb()
> instead of moving to tun_build_skb() suggested by Jason Wang
> <[email protected]>. Otherwise extra packets will be dropped
> without falling back to tun_alloc_skb(). And in the discussion of v3
> Jesper Dangaard Brouer <[email protected]> noticed that XDP is ok
> with a higher order pages if it's a contiguous physical memory
> allocation, so falling to tun_alloc_skb() -> do_xdp_generic() should
> be ok.
>
> v2 -> v3:
> * attach the forgotten changelog
>
> v1 -> v2:
> * merged 2 patches in 1, fixing both issues: WARN_ON_ONCE with
> syzkaller repro and missing XDP_PACKET_HEADROOM in pad
> * changed the title and description of the execution path, suggested
> by Jason Wang <[email protected]>
> * move the limit check from tun_can_build_skb() to tun_build_skb() to
> remove duplication and locking issue, and also drop the packet in
> case of a failed check - noted by Jason Wang <[email protected]>
>
> drivers/net/tun.c | 11 ++++++++++-
> 1 file changed, 10 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/net/tun.c b/drivers/net/tun.c
> index d75456adc62a..a1d04bc9485f 100644
> --- a/drivers/net/tun.c
> +++ b/drivers/net/tun.c
> @@ -1582,6 +1582,9 @@ static void tun_rx_batched(struct tun_struct *tun, struct tun_file *tfile,
> static bool tun_can_build_skb(struct tun_struct *tun, struct tun_file *tfile,
> int len, int noblock, bool zerocopy)
> {
> + struct bpf_prog *xdp_prog;
> + int pad = TUN_RX_PAD;
> +
> if ((tun->flags & TUN_TYPE_MASK) != IFF_TAP)
> return false;
>
> @@ -1594,7 +1597,13 @@ static bool tun_can_build_skb(struct tun_struct *tun, struct tun_file *tfile,
> if (zerocopy)
> return false;
>
> - if (SKB_DATA_ALIGN(len + TUN_RX_PAD) +
> + rcu_read_lock();
> + xdp_prog = rcu_dereference(tun->xdp_prog);
> + if (xdp_prog)
> + pad += XDP_PACKET_HEADROOM;
> + rcu_read_unlock();
> +
Isolated seen, I guess, this is a correct fix to 7df13219d757.
> + if (SKB_DATA_ALIGN(len + pad) +
> SKB_DATA_ALIGN(sizeof(struct skb_shared_info)) > PAGE_SIZE)
> return false;
>
Question to Jason Wang:
Why fall back (to e.g. tun_alloc_skb()) when size is above PAGE_SIZE?
AFAIK tun_build_skb() *can* create get larger packets than PAGE_SIZE
from it's page_frag. Is there a reason for this limitation?
(To Andrew, I assume a change in this area is another patch).
--Jesper
On Wed, Aug 2, 2023 at 10:14 PM Jesper Dangaard Brouer <[email protected]> wrote:
>
>
>
> On 02/08/2023 00.07, Andrew Kanner wrote:
> > Using the syzkaller repro with reduced packet size it was discovered
> > that XDP_PACKET_HEADROOM is not checked in tun_can_build_skb(),
> > although pad may be incremented in tun_build_skb(). This may end up
> > with exceeding the PAGE_SIZE limit in tun_build_skb().
> >
> > Fixes: 7df13219d757 ("tun: reserve extra headroom only when XDP is set")
> > Link: https://syzkaller.appspot.com/bug?extid=f817490f5bd20541b90a
> > Signed-off-by: Andrew Kanner <[email protected]>
> > ---
> >
> > Notes:
> > v3 -> v4:
> > * fall back to v1, fixing only missing XDP_PACKET_HEADROOM in pad and
> > removing bpf_xdp_adjust_tail() check for frame_sz.
> > * added rcu read lock, noted by Jason Wang <[email protected]> in v1
> > * I decided to leave the packet length check in tun_can_build_skb()
> > instead of moving to tun_build_skb() suggested by Jason Wang
> > <[email protected]>. Otherwise extra packets will be dropped
> > without falling back to tun_alloc_skb(). And in the discussion of v3
> > Jesper Dangaard Brouer <[email protected]> noticed that XDP is ok
> > with a higher order pages if it's a contiguous physical memory
> > allocation, so falling to tun_alloc_skb() -> do_xdp_generic() should
> > be ok.
> >
> > v2 -> v3:
> > * attach the forgotten changelog
> >
> > v1 -> v2:
> > * merged 2 patches in 1, fixing both issues: WARN_ON_ONCE with
> > syzkaller repro and missing XDP_PACKET_HEADROOM in pad
> > * changed the title and description of the execution path, suggested
> > by Jason Wang <[email protected]>
> > * move the limit check from tun_can_build_skb() to tun_build_skb() to
> > remove duplication and locking issue, and also drop the packet in
> > case of a failed check - noted by Jason Wang <[email protected]>
> >
> > drivers/net/tun.c | 11 ++++++++++-
> > 1 file changed, 10 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/net/tun.c b/drivers/net/tun.c
> > index d75456adc62a..a1d04bc9485f 100644
> > --- a/drivers/net/tun.c
> > +++ b/drivers/net/tun.c
> > @@ -1582,6 +1582,9 @@ static void tun_rx_batched(struct tun_struct *tun, struct tun_file *tfile,
> > static bool tun_can_build_skb(struct tun_struct *tun, struct tun_file *tfile,
> > int len, int noblock, bool zerocopy)
> > {
> > + struct bpf_prog *xdp_prog;
> > + int pad = TUN_RX_PAD;
> > +
> > if ((tun->flags & TUN_TYPE_MASK) != IFF_TAP)
> > return false;
> >
> > @@ -1594,7 +1597,13 @@ static bool tun_can_build_skb(struct tun_struct *tun, struct tun_file *tfile,
> > if (zerocopy)
> > return false;
> >
> > - if (SKB_DATA_ALIGN(len + TUN_RX_PAD) +
> > + rcu_read_lock();
> > + xdp_prog = rcu_dereference(tun->xdp_prog);
> > + if (xdp_prog)
> > + pad += XDP_PACKET_HEADROOM;
> > + rcu_read_unlock();
> > +
>
> Isolated seen, I guess, this is a correct fix to 7df13219d757.
I think so.
Actually, I think we can probably always count XDP_PACKET_HEADROOM
here. Since there's a window that XDP program might be attached in the
middle of tun_can_build_skb() and tun_build_skb().
>
> > + if (SKB_DATA_ALIGN(len + pad) +
> > SKB_DATA_ALIGN(sizeof(struct skb_shared_info)) > PAGE_SIZE)
> > return false;
> >
>
> Question to Jason Wang:
> Why fall back (to e.g. tun_alloc_skb()) when size is above PAGE_SIZE?
>
> AFAIK tun_build_skb() *can* create get larger packets than PAGE_SIZE
> from it's page_frag. Is there a reason for this limitation?
I couldn't recall but I think we can relax.
Thanks
>
> (To Andrew, I assume a change in this area is another patch).
>
> --Jesper
>
>
>
On Thu, Aug 03, 2023 at 11:19:47AM +0800, Jason Wang wrote:
> > > @@ -1594,7 +1597,13 @@ static bool tun_can_build_skb(struct tun_struct *tun, struct tun_file *tfile,
> > > if (zerocopy)
> > > return false;
> > >
> > > - if (SKB_DATA_ALIGN(len + TUN_RX_PAD) +
> > > + rcu_read_lock();
> > > + xdp_prog = rcu_dereference(tun->xdp_prog);
> > > + if (xdp_prog)
> > > + pad += XDP_PACKET_HEADROOM;
> > > + rcu_read_unlock();
> > > +
> >
> > Isolated seen, I guess, this is a correct fix to 7df13219d757.
>
> I think so.
>
> Actually, I think we can probably always count XDP_PACKET_HEADROOM
> here. Since there's a window that XDP program might be attached in the
> middle of tun_can_build_skb() and tun_build_skb().
Thanks, that makes sense.
I will do it in v5.
> > Question to Jason Wang:
> > Why fall back (to e.g. tun_alloc_skb()) when size is above PAGE_SIZE?
> >
> > AFAIK tun_build_skb() *can* create get larger packets than PAGE_SIZE
> > from it's page_frag. Is there a reason for this limitation?
>
> I couldn't recall but I think we can relax.
Jesper already sent enough info for this idea in v2, I will use it for
the next patch/series.
Jesper, I will add this tag for this next patch/series if you don't
mind:
Suggested-by: Jesper Dangaard Brouer <[email protected]>
--
Andrew Kanner
On 03/08/2023 19.53, Andrew Kanner wrote:
>>> Question to Jason Wang:
>>> Why fall back (to e.g. tun_alloc_skb()) when size is above PAGE_SIZE?
>>>
>>> AFAIK tun_build_skb()*can* create get larger packets than PAGE_SIZE
>>> from it's page_frag. Is there a reason for this limitation?
>> >> I couldn't recall but I think we can relax.
> > Jesper already sent enough info for this idea in v2, I will use it for
> the next patch/series.
>
I have some more input and considerations when selecting the new
constant that replace PAGE_SIZE.
Lets see if Eric Dumazet or Alex Duyck disagree?
("inventors" of page_frag scheme)
The function tun_alloc_skb() uses a page_frag scheme for allocation.
The maximim size is 32768 bytes (Order-3), but using something that is
close to this max alloc size can cause memory waste and fragmentation.
My suggestion would be to use the constant SKB_MAX_ALLOC (16KiB).
Maybe Eric or Alex would recommend using something smaller? (e.g. 8192)
page_frag limit comes from:
#define SKB_FRAG_PAGE_ORDER get_order(32768)
> Jesper, I will add this tag for this next patch/series if you don't
> mind:
> Suggested-by: Jesper Dangaard Brouer <[email protected]>
ACK