[RFC,nf-next,1/2] netfilter: bpf: support prog update
Commit Message
From: "D. Wythe" <alibuda@linux.alibaba.com>
To support the prog update, we need to ensure that the prog seen
within the hook is always valid. Considering that hooks are always
protected by rcu_read_lock(), which provide us the ability to use a
new RCU-protected context to access the prog.
Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
---
net/netfilter/nf_bpf_link.c | 124 +++++++++++++++++++++++++++++++++++++++-----
1 file changed, 111 insertions(+), 13 deletions(-)
Comments
D. Wythe <alibuda@linux.alibaba.com> wrote:
> From: "D. Wythe" <alibuda@linux.alibaba.com>
>
> To support the prog update, we need to ensure that the prog seen
> within the hook is always valid. Considering that hooks are always
> protected by rcu_read_lock(), which provide us the ability to use a
> new RCU-protected context to access the prog.
>
> Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
> ---
> net/netfilter/nf_bpf_link.c | 124 +++++++++++++++++++++++++++++++++++++++-----
> 1 file changed, 111 insertions(+), 13 deletions(-)
>
> diff --git a/net/netfilter/nf_bpf_link.c b/net/netfilter/nf_bpf_link.c
> index e502ec0..918c470 100644
> --- a/net/netfilter/nf_bpf_link.c
> +++ b/net/netfilter/nf_bpf_link.c
> @@ -8,17 +8,11 @@
> #include <net/netfilter/nf_bpf_link.h>
> #include <uapi/linux/netfilter_ipv4.h>
>
> -static unsigned int nf_hook_run_bpf(void *bpf_prog, struct sk_buff *skb,
> - const struct nf_hook_state *s)
> +struct bpf_nf_hook_ctx
> {
> - const struct bpf_prog *prog = bpf_prog;
> - struct bpf_nf_ctx ctx = {
> - .state = s,
> - .skb = skb,
> - };
> -
> - return bpf_prog_run(prog, &ctx);
> -}
> + struct bpf_prog *prog;
> + struct rcu_head rcu;
> +};
I don't understand the need for this structure. AFAICS bpf_prog_put()
will always release the program via call_rcu()?
If it doesn't, we are probably already in trouble as-is without this
patch, I don't think anything that prevents us from ending up calling already
released bpf prog, or releasing it while another cpu is still running it
if bpf_prog_put releases the actual underlying prog instantly.
A BPF expert could confirm bpf-prog-put-is-call-rcu.
> struct bpf_nf_link {
> struct bpf_link link;
> @@ -26,8 +20,59 @@ struct bpf_nf_link {
> struct net *net;
> u32 dead;
> const struct nf_defrag_hook *defrag_hook;
> + /* protect link update in parallel */
> + struct mutex update_lock;
> + struct bpf_nf_hook_ctx __rcu *hook_ctx;
What kind of replacements-per-second rate are you aiming for?
I think
static DEFINE_MUTEX(bpf_nf_mutex);
is enough.
Then bpf_nf_link gains
struct bpf_prog __rcu *prog
and possibly a trailing struct rcu_head, see below.
> +static void bpf_nf_hook_ctx_free_rcu(struct bpf_nf_hook_ctx *hook_ctx)
> +{
> + call_rcu(&hook_ctx->rcu, __bpf_nf_hook_ctx_free_rcu);
> +}
Don't understand the need for call_rcu either, see below.
> +static unsigned int nf_hook_run_bpf(void *bpf_link, struct sk_buff *skb,
> + const struct nf_hook_state *s)
> +{
> + const struct bpf_nf_link *link = bpf_link;
> + struct bpf_nf_hook_ctx *hook_ctx;
> + struct bpf_nf_ctx ctx = {
> + .state = s,
> + .skb = skb,
> + };
> +
> + hook_ctx = rcu_dereference(link->hook_ctx);
This could then just rcu_deref link->prog.
> + return bpf_prog_run(hook_ctx->prog, &ctx);
> +}
> +
> #if IS_ENABLED(CONFIG_NF_DEFRAG_IPV4) || IS_ENABLED(CONFIG_NF_DEFRAG_IPV6)
> static const struct nf_defrag_hook *
> get_proto_defrag_hook(struct bpf_nf_link *link,
> @@ -120,6 +165,10 @@ static void bpf_nf_link_release(struct bpf_link *link)
> if (!cmpxchg(&nf_link->dead, 0, 1)) {
> nf_unregister_net_hook(nf_link->net, &nf_link->hook_ops);
> bpf_nf_disable_defrag(nf_link);
> + /* Wait for outstanding hook to complete before the
> + * link gets released.
> + */
> + synchronize_rcu();
> }
Could you convert bpf_nf_link_dealloc to release via kfree_rcu instead?
> @@ -162,7 +212,42 @@ static int bpf_nf_link_fill_link_info(const struct bpf_link *link,
> static int bpf_nf_link_update(struct bpf_link *link, struct bpf_prog *new_prog,
> struct bpf_prog *old_prog)
> {
> - return -EOPNOTSUPP;
> + struct bpf_nf_link *nf_link = container_of(link, struct bpf_nf_link, link);
> + struct bpf_nf_hook_ctx *hook_ctx;
> + int err = 0;
> +
> + mutex_lock(&nf_link->update_lock);
> +
I think you need to check link->dead here too.
> + /* bpf_nf_link_release() ensures that after its execution, there will be
> + * no ongoing or upcoming execution of nf_hook_run_bpf() within any context.
> + * Therefore, within nf_hook_run_bpf(), the link remains valid at all times."
> + */
> + link->hook_ops.priv = link;
ATM we only need to make sure the bpf prog itself stays alive until after
all concurrent rcu critical sections have completed.
After this change, struct bpf_link gets passed instead, so we need to
keep that alive too.
Which works with synchronize_rcu, sure, but that seems a bit overkill here.
On Wed, Dec 13, 2023 at 2:24 PM Florian Westphal <fw@strlen.de> wrote:
>
> D. Wythe <alibuda@linux.alibaba.com> wrote:
> > From: "D. Wythe" <alibuda@linux.alibaba.com>
> >
> > To support the prog update, we need to ensure that the prog seen
> > within the hook is always valid. Considering that hooks are always
> > protected by rcu_read_lock(), which provide us the ability to use a
> > new RCU-protected context to access the prog.
> >
> > Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
> > ---
> > net/netfilter/nf_bpf_link.c | 124 +++++++++++++++++++++++++++++++++++++++-----
> > 1 file changed, 111 insertions(+), 13 deletions(-)
> >
> > diff --git a/net/netfilter/nf_bpf_link.c b/net/netfilter/nf_bpf_link.c
> > index e502ec0..918c470 100644
> > --- a/net/netfilter/nf_bpf_link.c
> > +++ b/net/netfilter/nf_bpf_link.c
> > @@ -8,17 +8,11 @@
> > #include <net/netfilter/nf_bpf_link.h>
> > #include <uapi/linux/netfilter_ipv4.h>
> >
> > -static unsigned int nf_hook_run_bpf(void *bpf_prog, struct sk_buff *skb,
> > - const struct nf_hook_state *s)
> > +struct bpf_nf_hook_ctx
> > {
> > - const struct bpf_prog *prog = bpf_prog;
> > - struct bpf_nf_ctx ctx = {
> > - .state = s,
> > - .skb = skb,
> > - };
> > -
> > - return bpf_prog_run(prog, &ctx);
> > -}
> > + struct bpf_prog *prog;
> > + struct rcu_head rcu;
> > +};
>
> I don't understand the need for this structure. AFAICS bpf_prog_put()
> will always release the program via call_rcu()?
>
> If it doesn't, we are probably already in trouble as-is without this
> patch, I don't think anything that prevents us from ending up calling already
> released bpf prog, or releasing it while another cpu is still running it
> if bpf_prog_put releases the actual underlying prog instantly.
>
> A BPF expert could confirm bpf-prog-put-is-call-rcu.
+1
These patches look unnecessary.
It seems that they accidently fix something else.
On 12/14/23 6:24 AM, Florian Westphal wrote:
> D. Wythe <alibuda@linux.alibaba.com> wrote:
>> From: "D. Wythe" <alibuda@linux.alibaba.com>
>>
>> To support the prog update, we need to ensure that the prog seen
>> within the hook is always valid. Considering that hooks are always
>> protected by rcu_read_lock(), which provide us the ability to use a
>> new RCU-protected context to access the prog.
>>
>> Signed-off-by: D. Wythe <alibuda@linux.alibaba.com>
>> ---
>> net/netfilter/nf_bpf_link.c | 124 +++++++++++++++++++++++++++++++++++++++-----
>> 1 file changed, 111 insertions(+), 13 deletions(-)
>>
>> diff --git a/net/netfilter/nf_bpf_link.c b/net/netfilter/nf_bpf_link.c
>> index e502ec0..918c470 100644
>> --- a/net/netfilter/nf_bpf_link.c
>> +++ b/net/netfilter/nf_bpf_link.c
>> @@ -8,17 +8,11 @@
>> #include <net/netfilter/nf_bpf_link.h>
>> #include <uapi/linux/netfilter_ipv4.h>
>>
>> -static unsigned int nf_hook_run_bpf(void *bpf_prog, struct sk_buff *skb,
>> - const struct nf_hook_state *s)
>> +struct bpf_nf_hook_ctx
>> {
>> - const struct bpf_prog *prog = bpf_prog;
>> - struct bpf_nf_ctx ctx = {
>> - .state = s,
>> - .skb = skb,
>> - };
>> -
>> - return bpf_prog_run(prog, &ctx);
>> -}
>> + struct bpf_prog *prog;
>> + struct rcu_head rcu;
>> +};
> I don't understand the need for this structure. AFAICS bpf_prog_put()
> will always release the program via call_rcu()?
>
> If it doesn't, we are probably already in trouble as-is without this
> patch, I don't think anything that prevents us from ending up calling already
> released bpf prog, or releasing it while another cpu is still running it
> if bpf_prog_put releases the actual underlying prog instantly.
>
> A BPF expert could confirm bpf-prog-put-is-call-rcu.
Hi Florian,
I must admit that I did not realize that bpf_prog is released
under RCU ...
>> struct bpf_nf_link {
>> struct bpf_link link;
>> @@ -26,8 +20,59 @@ struct bpf_nf_link {
>> struct net *net;
>> u32 dead;
>> const struct nf_defrag_hook *defrag_hook;
>> + /* protect link update in parallel */
>> + struct mutex update_lock;
>> + struct bpf_nf_hook_ctx __rcu *hook_ctx;
> What kind of replacements-per-second rate are you aiming for?
> I think
>
> static DEFINE_MUTEX(bpf_nf_mutex);
>
> is enough.
I'm okay with that.
>
> Then bpf_nf_link gains
>
> struct bpf_prog __rcu *prog
>
> and possibly a trailing struct rcu_head, see below.
Yes, that's what we need.
>> +static void bpf_nf_hook_ctx_free_rcu(struct bpf_nf_hook_ctx *hook_ctx)
>> +{
>> + call_rcu(&hook_ctx->rcu, __bpf_nf_hook_ctx_free_rcu);
>> +}
> Don't understand the need for call_rcu either, see below.
>
>> +static unsigned int nf_hook_run_bpf(void *bpf_link, struct sk_buff *skb,
>> + const struct nf_hook_state *s)
>> +{
>> + const struct bpf_nf_link *link = bpf_link;
>> + struct bpf_nf_hook_ctx *hook_ctx;
>> + struct bpf_nf_ctx ctx = {
>> + .state = s,
>> + .skb = skb,
>> + };
>> +
>> + hook_ctx = rcu_dereference(link->hook_ctx);
> This could then just rcu_deref link->prog.
>
>> + return bpf_prog_run(hook_ctx->prog, &ctx);
>> +}
>> +
>> #if IS_ENABLED(CONFIG_NF_DEFRAG_IPV4) || IS_ENABLED(CONFIG_NF_DEFRAG_IPV6)
>> static const struct nf_defrag_hook *
>> get_proto_defrag_hook(struct bpf_nf_link *link,
>> @@ -120,6 +165,10 @@ static void bpf_nf_link_release(struct bpf_link *link)
>> if (!cmpxchg(&nf_link->dead, 0, 1)) {
>> nf_unregister_net_hook(nf_link->net, &nf_link->hook_ops);
>> bpf_nf_disable_defrag(nf_link);
>> + /* Wait for outstanding hook to complete before the
>> + * link gets released.
>> + */
>> + synchronize_rcu();
>> }
> Could you convert bpf_nf_link_dealloc to release via kfree_rcu instead?
>
Got it.
>> @@ -162,7 +212,42 @@ static int bpf_nf_link_fill_link_info(const struct bpf_link *link,
>> static int bpf_nf_link_update(struct bpf_link *link, struct bpf_prog *new_prog,
>> struct bpf_prog *old_prog)
>> {
>> - return -EOPNOTSUPP;
>> + struct bpf_nf_link *nf_link = container_of(link, struct bpf_nf_link, link);
>> + struct bpf_nf_hook_ctx *hook_ctx;
>> + int err = 0;
>> +
>> + mutex_lock(&nf_link->update_lock);
>> +
> I think you need to check link->dead here too.
Got that.
>
>> + /* bpf_nf_link_release() ensures that after its execution, there will be
>> + * no ongoing or upcoming execution of nf_hook_run_bpf() within any context.
>> + * Therefore, within nf_hook_run_bpf(), the link remains valid at all times."
>> + */
>> + link->hook_ops.priv = link;
> ATM we only need to make sure the bpf prog itself stays alive until after
> all concurrent rcu critical sections have completed.
>
> After this change, struct bpf_link gets passed instead, so we need to
> keep that alive too.
>
> Which works with synchronize_rcu, sure, but that seems a bit overkill here.
Got it! Thank you very much for your suggestion.
I will address those issues you mentioned in the next version.
Best wishes,
D. Wythe
On Wed, Dec 13, 2023 at 9:31 PM D. Wythe <alibuda@linux.alibaba.com> wrote:
>
> I will address those issues you mentioned in the next version.
Don't. There is no need for the next version.
None of these changes are necessary.
On 12/14/23 1:50 PM, Alexei Starovoitov wrote:
> On Wed, Dec 13, 2023 at 9:31 PM D. Wythe <alibuda@linux.alibaba.com> wrote:
>> I will address those issues you mentioned in the next version.
> Don't. There is no need for the next version.
> None of these changes are necessary.
Can I know the reason ? Updating prog for active link is kind of
important feature
for real application..
Best wishes,
D. Wythe
On Thu, Dec 14, 2023 at 12:57 AM D. Wythe <alibuda@linux.alibaba.com> wrote:
>
>
>
> On 12/14/23 1:50 PM, Alexei Starovoitov wrote:
> > On Wed, Dec 13, 2023 at 9:31 PM D. Wythe <alibuda@linux.alibaba.com> wrote:
> >> I will address those issues you mentioned in the next version.
> > Don't. There is no need for the next version.
> > None of these changes are necessary.
>
> Can I know the reason ? Updating prog for active link is kind of
> important feature
> for real application..
yes. it's and it's working as expected. Do you see an issue?
On 12/14/23 9:37 PM, Alexei Starovoitov wrote:
> yes. it's and it's working as expected. Do you see an issue?
Hi Alexei,
I see the issue here is that bpf_nf_link has not yet implemented
prog_update,
which just simply returned -EOPNOTSUPP right now.
Do you mean that it is already implemented in the latest tree or
the not-supported was expected?
Thanks,
D. Wythe
On Thu, Dec 14, 2023 at 7:56 AM D. Wythe <alibuda@linux.alibaba.com> wrote:
>
>
>
> On 12/14/23 9:37 PM, Alexei Starovoitov wrote:
> > yes. it's and it's working as expected. Do you see an issue?
>
> Hi Alexei,
>
> I see the issue here is that bpf_nf_link has not yet implemented
> prog_update,
> which just simply returned -EOPNOTSUPP right now.
I see. The commit log didn't make it clear.
Yes. That would be good to support.
On 12/15/23 12:02 AM, Alexei Starovoitov wrote:
> I see. The commit log didn't make it clear.
> Yes. That would be good to support.
That's my bad. I will make the commit log more clear in the next version.
In any case, thanks very much for your feedback.
Besh wishes,
D. Wythe
@@ -8,17 +8,11 @@
#include <net/netfilter/nf_bpf_link.h>
#include <uapi/linux/netfilter_ipv4.h>
-static unsigned int nf_hook_run_bpf(void *bpf_prog, struct sk_buff *skb,
- const struct nf_hook_state *s)
+struct bpf_nf_hook_ctx
{
- const struct bpf_prog *prog = bpf_prog;
- struct bpf_nf_ctx ctx = {
- .state = s,
- .skb = skb,
- };
-
- return bpf_prog_run(prog, &ctx);
-}
+ struct bpf_prog *prog;
+ struct rcu_head rcu;
+};
struct bpf_nf_link {
struct bpf_link link;
@@ -26,8 +20,59 @@ struct bpf_nf_link {
struct net *net;
u32 dead;
const struct nf_defrag_hook *defrag_hook;
+ /* protect link update in parallel */
+ struct mutex update_lock;
+ struct bpf_nf_hook_ctx __rcu *hook_ctx;
};
+static struct bpf_nf_hook_ctx *
+bpf_nf_hook_ctx_from_prog(struct bpf_prog *prog, gfp_t flags)
+{
+ struct bpf_nf_hook_ctx *hook_ctx;
+
+ hook_ctx = kmalloc(sizeof(*hook_ctx), flags);
+ if (hook_ctx) {
+ hook_ctx->prog = prog;
+ bpf_prog_inc(prog);
+ }
+ return hook_ctx;
+}
+
+static void bpf_nf_hook_ctx_free(struct bpf_nf_hook_ctx *hook_ctx)
+{
+ if (!hook_ctx)
+ return;
+ if (hook_ctx->prog)
+ bpf_prog_put(hook_ctx->prog);
+ kfree(hook_ctx);
+}
+
+static void __bpf_nf_hook_ctx_free_rcu(struct rcu_head *head)
+{
+ struct bpf_nf_hook_ctx *hook_ctx = container_of(head, struct bpf_nf_hook_ctx, rcu);
+
+ bpf_nf_hook_ctx_free(hook_ctx);
+}
+
+static void bpf_nf_hook_ctx_free_rcu(struct bpf_nf_hook_ctx *hook_ctx)
+{
+ call_rcu(&hook_ctx->rcu, __bpf_nf_hook_ctx_free_rcu);
+}
+
+static unsigned int nf_hook_run_bpf(void *bpf_link, struct sk_buff *skb,
+ const struct nf_hook_state *s)
+{
+ const struct bpf_nf_link *link = bpf_link;
+ struct bpf_nf_hook_ctx *hook_ctx;
+ struct bpf_nf_ctx ctx = {
+ .state = s,
+ .skb = skb,
+ };
+
+ hook_ctx = rcu_dereference(link->hook_ctx);
+ return bpf_prog_run(hook_ctx->prog, &ctx);
+}
+
#if IS_ENABLED(CONFIG_NF_DEFRAG_IPV4) || IS_ENABLED(CONFIG_NF_DEFRAG_IPV6)
static const struct nf_defrag_hook *
get_proto_defrag_hook(struct bpf_nf_link *link,
@@ -120,6 +165,10 @@ static void bpf_nf_link_release(struct bpf_link *link)
if (!cmpxchg(&nf_link->dead, 0, 1)) {
nf_unregister_net_hook(nf_link->net, &nf_link->hook_ops);
bpf_nf_disable_defrag(nf_link);
+ /* Wait for outstanding hook to complete before the
+ * link gets released.
+ */
+ synchronize_rcu();
}
}
@@ -127,6 +176,7 @@ static void bpf_nf_link_dealloc(struct bpf_link *link)
{
struct bpf_nf_link *nf_link = container_of(link, struct bpf_nf_link, link);
+ bpf_nf_hook_ctx_free(nf_link->hook_ctx);
kfree(nf_link);
}
@@ -162,7 +212,42 @@ static int bpf_nf_link_fill_link_info(const struct bpf_link *link,
static int bpf_nf_link_update(struct bpf_link *link, struct bpf_prog *new_prog,
struct bpf_prog *old_prog)
{
- return -EOPNOTSUPP;
+ struct bpf_nf_link *nf_link = container_of(link, struct bpf_nf_link, link);
+ struct bpf_nf_hook_ctx *hook_ctx;
+ int err = 0;
+
+ mutex_lock(&nf_link->update_lock);
+
+ /* target old_prog mismatch */
+ if (old_prog && link->prog != old_prog) {
+ err = -EPERM;
+ goto out;
+ }
+
+ old_prog = link->prog;
+ if (old_prog == new_prog) {
+ /* don't need update */
+ bpf_prog_put(new_prog);
+ goto out;
+ }
+
+ hook_ctx = bpf_nf_hook_ctx_from_prog(new_prog, GFP_USER);
+ if (!hook_ctx) {
+ err = -ENOMEM;
+ goto out;
+ }
+
+ /* replace and get the old one */
+ hook_ctx = rcu_replace_pointer(nf_link->hook_ctx, hook_ctx,
+ lockdep_is_held(&nf_link->update_lock));
+ /* free old hook_ctx */
+ bpf_nf_hook_ctx_free_rcu(hook_ctx);
+
+ old_prog = xchg(&link->prog, new_prog);
+ bpf_prog_put(old_prog);
+out:
+ mutex_unlock(&nf_link->update_lock);
+ return err;
}
static const struct bpf_link_ops bpf_nf_link_lops = {
@@ -222,11 +307,22 @@ int bpf_nf_link_attach(const union bpf_attr *attr, struct bpf_prog *prog)
if (!link)
return -ENOMEM;
+ link->hook_ctx = bpf_nf_hook_ctx_from_prog(prog, GFP_USER);
+ if (!link->hook_ctx) {
+ kfree(link);
+ return -ENOMEM;
+ }
+
bpf_link_init(&link->link, BPF_LINK_TYPE_NETFILTER, &bpf_nf_link_lops, prog);
link->hook_ops.hook = nf_hook_run_bpf;
link->hook_ops.hook_ops_type = NF_HOOK_OP_BPF;
- link->hook_ops.priv = prog;
+
+ /* bpf_nf_link_release() ensures that after its execution, there will be
+ * no ongoing or upcoming execution of nf_hook_run_bpf() within any context.
+ * Therefore, within nf_hook_run_bpf(), the link remains valid at all times."
+ */
+ link->hook_ops.priv = link;
link->hook_ops.pf = attr->link_create.netfilter.pf;
link->hook_ops.priority = attr->link_create.netfilter.priority;
@@ -236,9 +332,11 @@ int bpf_nf_link_attach(const union bpf_attr *attr, struct bpf_prog *prog)
link->dead = false;
link->defrag_hook = NULL;
+ mutex_init(&link->update_lock);
+
err = bpf_link_prime(&link->link, &link_primer);
if (err) {
- kfree(link);
+ bpf_nf_link_dealloc(&link->link);
return err;
}