* [PATCH net-next 1/2] bpf: skip sockopt hooks without CONFIG_NET
@ 2019-07-08 12:57 Arnd Bergmann
2019-07-08 12:57 ` [PATCH net-next 2/2] bpf: avoid unused variable warning in tcp_bpf_rtt() Arnd Bergmann
2019-07-08 15:06 ` [PATCH net-next 1/2] bpf: skip sockopt hooks without CONFIG_NET Yonghong Song
0 siblings, 2 replies; 5+ messages in thread
From: Arnd Bergmann @ 2019-07-08 12:57 UTC (permalink / raw)
To: Alexei Starovoitov, Daniel Borkmann
Cc: Arnd Bergmann, Andrii Nakryiko, Martin Lau, Stanislav Fomichev,
Song Liu, Yonghong Song, Mauricio Vasquez B, Roman Gushchin,
Matt Mullins, Willem de Bruijn, Andrey Ignatov, netdev, bpf,
linux-kernel
When CONFIG_NET is disabled, we get a link error:
kernel/bpf/cgroup.o: In function `__cgroup_bpf_run_filter_setsockopt':
cgroup.c:(.text+0x3010): undefined reference to `lock_sock_nested'
cgroup.c:(.text+0x3258): undefined reference to `release_sock'
kernel/bpf/cgroup.o: In function `__cgroup_bpf_run_filter_getsockopt':
cgroup.c:(.text+0x3568): undefined reference to `lock_sock_nested'
cgroup.c:(.text+0x3870): undefined reference to `release_sock'
kernel/bpf/cgroup.o: In function `cg_sockopt_func_proto':
cgroup.c:(.text+0x41d8): undefined reference to `bpf_sk_storage_delete_proto'
None of this code is useful in this configuration anyway, so we can
simply hide it in an appropriate #ifdef.
Fixes: 0d01da6afc54 ("bpf: implement getsockopt and setsockopt hooks")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
---
include/linux/bpf_types.h | 2 ++
kernel/bpf/cgroup.c | 6 ++++++
2 files changed, 8 insertions(+)
diff --git a/include/linux/bpf_types.h b/include/linux/bpf_types.h
index eec5aeeeaf92..3c7222b2db96 100644
--- a/include/linux/bpf_types.h
+++ b/include/linux/bpf_types.h
@@ -30,8 +30,10 @@ BPF_PROG_TYPE(BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE, raw_tracepoint_writable)
#ifdef CONFIG_CGROUP_BPF
BPF_PROG_TYPE(BPF_PROG_TYPE_CGROUP_DEVICE, cg_dev)
BPF_PROG_TYPE(BPF_PROG_TYPE_CGROUP_SYSCTL, cg_sysctl)
+#ifdef CONFIG_NET
BPF_PROG_TYPE(BPF_PROG_TYPE_CGROUP_SOCKOPT, cg_sockopt)
#endif
+#endif
#ifdef CONFIG_BPF_LIRC_MODE2
BPF_PROG_TYPE(BPF_PROG_TYPE_LIRC_MODE2, lirc_mode2)
#endif
diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c
index 76fa0076f20d..7be44460bd93 100644
--- a/kernel/bpf/cgroup.c
+++ b/kernel/bpf/cgroup.c
@@ -590,6 +590,7 @@ int cgroup_bpf_prog_query(const union bpf_attr *attr,
return ret;
}
+#ifdef CONFIG_NET
/**
* __cgroup_bpf_run_filter_skb() - Run a program for packet filtering
* @sk: The socket sending or receiving traffic
@@ -750,6 +751,7 @@ int __cgroup_bpf_run_filter_sock_ops(struct sock *sk,
return ret == 1 ? 0 : -EPERM;
}
EXPORT_SYMBOL(__cgroup_bpf_run_filter_sock_ops);
+#endif
int __cgroup_bpf_check_dev_permission(short dev_type, u32 major, u32 minor,
short access, enum bpf_attach_type type)
@@ -939,6 +941,7 @@ int __cgroup_bpf_run_filter_sysctl(struct ctl_table_header *head,
}
EXPORT_SYMBOL(__cgroup_bpf_run_filter_sysctl);
+#ifdef CONFIG_NET
static bool __cgroup_bpf_prog_array_is_empty(struct cgroup *cgrp,
enum bpf_attach_type attach_type)
{
@@ -1120,6 +1123,7 @@ int __cgroup_bpf_run_filter_getsockopt(struct sock *sk, int level,
return ret;
}
EXPORT_SYMBOL(__cgroup_bpf_run_filter_getsockopt);
+#endif
static ssize_t sysctl_cpy_dir(const struct ctl_dir *dir, char **bufp,
size_t *lenp)
@@ -1382,6 +1386,7 @@ const struct bpf_verifier_ops cg_sysctl_verifier_ops = {
const struct bpf_prog_ops cg_sysctl_prog_ops = {
};
+#ifdef CONFIG_NET
static const struct bpf_func_proto *
cg_sockopt_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog)
{
@@ -1531,3 +1536,4 @@ const struct bpf_verifier_ops cg_sockopt_verifier_ops = {
const struct bpf_prog_ops cg_sockopt_prog_ops = {
};
+#endif
--
2.20.0
^ permalink raw reply related [flat|nested] 5+ messages in thread* [PATCH net-next 2/2] bpf: avoid unused variable warning in tcp_bpf_rtt()
2019-07-08 12:57 [PATCH net-next 1/2] bpf: skip sockopt hooks without CONFIG_NET Arnd Bergmann
@ 2019-07-08 12:57 ` Arnd Bergmann
2019-07-08 13:55 ` Soheil Hassas Yeganeh
2019-07-08 15:06 ` [PATCH net-next 1/2] bpf: skip sockopt hooks without CONFIG_NET Yonghong Song
1 sibling, 1 reply; 5+ messages in thread
From: Arnd Bergmann @ 2019-07-08 12:57 UTC (permalink / raw)
To: Eric Dumazet, David S. Miller, Alexei Starovoitov, Daniel Borkmann
Cc: Arnd Bergmann, Priyaranjan Jha, Yuchung Cheng,
Soheil Hassas Yeganeh, Stanislav Fomichev, Martin KaFai Lau,
Song Liu, Yonghong Song, Neal Cardwell, Jason Baron, Yafang Shao,
Ard Biesheuvel, netdev, linux-kernel, bpf
When CONFIG_BPF is disabled, we get a warning for an unused
variable:
In file included from drivers/target/target_core_device.c:26:
include/net/tcp.h:2226:19: error: unused variable 'tp' [-Werror,-Wunused-variable]
struct tcp_sock *tp = tcp_sk(sk);
The variable is only used in one place, so it can be
replaced with its value there to avoid the warning.
Fixes: 23729ff23186 ("bpf: add BPF_CGROUP_SOCK_OPS callback that is executed on every RTT")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
---
include/net/tcp.h | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
diff --git a/include/net/tcp.h b/include/net/tcp.h
index e16d8a3fd3b4..cca3c59b98bf 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -2223,9 +2223,7 @@ static inline bool tcp_bpf_ca_needs_ecn(struct sock *sk)
static inline void tcp_bpf_rtt(struct sock *sk)
{
- struct tcp_sock *tp = tcp_sk(sk);
-
- if (BPF_SOCK_OPS_TEST_FLAG(tp, BPF_SOCK_OPS_RTT_CB_FLAG))
+ if (BPF_SOCK_OPS_TEST_FLAG(tcp_sk(sk), BPF_SOCK_OPS_RTT_CB_FLAG))
tcp_call_bpf(sk, BPF_SOCK_OPS_RTT_CB, 0, NULL);
}
--
2.20.0
^ permalink raw reply related [flat|nested] 5+ messages in thread* Re: [PATCH net-next 2/2] bpf: avoid unused variable warning in tcp_bpf_rtt()
2019-07-08 12:57 ` [PATCH net-next 2/2] bpf: avoid unused variable warning in tcp_bpf_rtt() Arnd Bergmann
@ 2019-07-08 13:55 ` Soheil Hassas Yeganeh
0 siblings, 0 replies; 5+ messages in thread
From: Soheil Hassas Yeganeh @ 2019-07-08 13:55 UTC (permalink / raw)
To: Arnd Bergmann
Cc: Eric Dumazet, David S. Miller, Alexei Starovoitov,
Daniel Borkmann, Priyaranjan Jha, Yuchung Cheng,
Stanislav Fomichev, Martin KaFai Lau, Song Liu, Yonghong Song,
Neal Cardwell, Jason Baron, Yafang Shao, Ard Biesheuvel, netdev,
linux-kernel, bpf
On Mon, Jul 8, 2019 at 8:57 AM Arnd Bergmann <arnd@arndb.de> wrote:
>
> When CONFIG_BPF is disabled, we get a warning for an unused
> variable:
>
> In file included from drivers/target/target_core_device.c:26:
> include/net/tcp.h:2226:19: error: unused variable 'tp' [-Werror,-Wunused-variable]
> struct tcp_sock *tp = tcp_sk(sk);
>
> The variable is only used in one place, so it can be
> replaced with its value there to avoid the warning.
>
> Fixes: 23729ff23186 ("bpf: add BPF_CGROUP_SOCK_OPS callback that is executed on every RTT")
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
> ---
> include/net/tcp.h | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/include/net/tcp.h b/include/net/tcp.h
> index e16d8a3fd3b4..cca3c59b98bf 100644
> --- a/include/net/tcp.h
> +++ b/include/net/tcp.h
> @@ -2223,9 +2223,7 @@ static inline bool tcp_bpf_ca_needs_ecn(struct sock *sk)
>
> static inline void tcp_bpf_rtt(struct sock *sk)
> {
> - struct tcp_sock *tp = tcp_sk(sk);
> -
> - if (BPF_SOCK_OPS_TEST_FLAG(tp, BPF_SOCK_OPS_RTT_CB_FLAG))
> + if (BPF_SOCK_OPS_TEST_FLAG(tcp_sk(sk), BPF_SOCK_OPS_RTT_CB_FLAG))
> tcp_call_bpf(sk, BPF_SOCK_OPS_RTT_CB, 0, NULL);
> }
>
> --
> 2.20.0
>
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH net-next 1/2] bpf: skip sockopt hooks without CONFIG_NET
2019-07-08 12:57 [PATCH net-next 1/2] bpf: skip sockopt hooks without CONFIG_NET Arnd Bergmann
2019-07-08 12:57 ` [PATCH net-next 2/2] bpf: avoid unused variable warning in tcp_bpf_rtt() Arnd Bergmann
@ 2019-07-08 15:06 ` Yonghong Song
2019-07-08 15:26 ` Daniel Borkmann
1 sibling, 1 reply; 5+ messages in thread
From: Yonghong Song @ 2019-07-08 15:06 UTC (permalink / raw)
To: Arnd Bergmann, Alexei Starovoitov, Daniel Borkmann
Cc: Andrii Nakryiko, Martin Lau, Stanislav Fomichev, Song Liu,
Mauricio Vasquez B, Roman Gushchin, Matt Mullins,
Willem de Bruijn, Andrey Ignatov, netdev, bpf, linux-kernel
On 7/8/19 5:57 AM, Arnd Bergmann wrote:
> When CONFIG_NET is disabled, we get a link error:
>
> kernel/bpf/cgroup.o: In function `__cgroup_bpf_run_filter_setsockopt':
> cgroup.c:(.text+0x3010): undefined reference to `lock_sock_nested'
> cgroup.c:(.text+0x3258): undefined reference to `release_sock'
> kernel/bpf/cgroup.o: In function `__cgroup_bpf_run_filter_getsockopt':
> cgroup.c:(.text+0x3568): undefined reference to `lock_sock_nested'
> cgroup.c:(.text+0x3870): undefined reference to `release_sock'
> kernel/bpf/cgroup.o: In function `cg_sockopt_func_proto':
> cgroup.c:(.text+0x41d8): undefined reference to `bpf_sk_storage_delete_proto'
>
> None of this code is useful in this configuration anyway, so we can
> simply hide it in an appropriate #ifdef.
>
> Fixes: 0d01da6afc54 ("bpf: implement getsockopt and setsockopt hooks")
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
FYI.
There is already a patch to fix the same issue,
https://lore.kernel.org/bpf/e9e489fe-feec-a211-82aa-5df0c6a308d1@huawei.com/T/#t
which has been acked and not merged yet.
> ---
> include/linux/bpf_types.h | 2 ++
> kernel/bpf/cgroup.c | 6 ++++++
> 2 files changed, 8 insertions(+)
>
> diff --git a/include/linux/bpf_types.h b/include/linux/bpf_types.h
> index eec5aeeeaf92..3c7222b2db96 100644
> --- a/include/linux/bpf_types.h
> +++ b/include/linux/bpf_types.h
> @@ -30,8 +30,10 @@ BPF_PROG_TYPE(BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE, raw_tracepoint_writable)
> #ifdef CONFIG_CGROUP_BPF
> BPF_PROG_TYPE(BPF_PROG_TYPE_CGROUP_DEVICE, cg_dev)
> BPF_PROG_TYPE(BPF_PROG_TYPE_CGROUP_SYSCTL, cg_sysctl)
> +#ifdef CONFIG_NET
> BPF_PROG_TYPE(BPF_PROG_TYPE_CGROUP_SOCKOPT, cg_sockopt)
> #endif
> +#endif
> #ifdef CONFIG_BPF_LIRC_MODE2
> BPF_PROG_TYPE(BPF_PROG_TYPE_LIRC_MODE2, lirc_mode2)
> #endif
> diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c
> index 76fa0076f20d..7be44460bd93 100644
> --- a/kernel/bpf/cgroup.c
> +++ b/kernel/bpf/cgroup.c
> @@ -590,6 +590,7 @@ int cgroup_bpf_prog_query(const union bpf_attr *attr,
> return ret;
> }
>
> +#ifdef CONFIG_NET
> /**
> * __cgroup_bpf_run_filter_skb() - Run a program for packet filtering
> * @sk: The socket sending or receiving traffic
> @@ -750,6 +751,7 @@ int __cgroup_bpf_run_filter_sock_ops(struct sock *sk,
> return ret == 1 ? 0 : -EPERM;
> }
> EXPORT_SYMBOL(__cgroup_bpf_run_filter_sock_ops);
> +#endif
>
> int __cgroup_bpf_check_dev_permission(short dev_type, u32 major, u32 minor,
> short access, enum bpf_attach_type type)
> @@ -939,6 +941,7 @@ int __cgroup_bpf_run_filter_sysctl(struct ctl_table_header *head,
> }
> EXPORT_SYMBOL(__cgroup_bpf_run_filter_sysctl);
>
> +#ifdef CONFIG_NET
> static bool __cgroup_bpf_prog_array_is_empty(struct cgroup *cgrp,
> enum bpf_attach_type attach_type)
> {
> @@ -1120,6 +1123,7 @@ int __cgroup_bpf_run_filter_getsockopt(struct sock *sk, int level,
> return ret;
> }
> EXPORT_SYMBOL(__cgroup_bpf_run_filter_getsockopt);
> +#endif
>
> static ssize_t sysctl_cpy_dir(const struct ctl_dir *dir, char **bufp,
> size_t *lenp)
> @@ -1382,6 +1386,7 @@ const struct bpf_verifier_ops cg_sysctl_verifier_ops = {
> const struct bpf_prog_ops cg_sysctl_prog_ops = {
> };
>
> +#ifdef CONFIG_NET
> static const struct bpf_func_proto *
> cg_sockopt_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog)
> {
> @@ -1531,3 +1536,4 @@ const struct bpf_verifier_ops cg_sockopt_verifier_ops = {
>
> const struct bpf_prog_ops cg_sockopt_prog_ops = {
> };
> +#endif
>
^ permalink raw reply [flat|nested] 5+ messages in thread* Re: [PATCH net-next 1/2] bpf: skip sockopt hooks without CONFIG_NET
2019-07-08 15:06 ` [PATCH net-next 1/2] bpf: skip sockopt hooks without CONFIG_NET Yonghong Song
@ 2019-07-08 15:26 ` Daniel Borkmann
0 siblings, 0 replies; 5+ messages in thread
From: Daniel Borkmann @ 2019-07-08 15:26 UTC (permalink / raw)
To: Yonghong Song, Arnd Bergmann, Alexei Starovoitov
Cc: Andrii Nakryiko, Martin Lau, Stanislav Fomichev, Song Liu,
Mauricio Vasquez B, Roman Gushchin, Matt Mullins,
Willem de Bruijn, Andrey Ignatov, netdev, bpf, linux-kernel
On 07/08/2019 05:06 PM, Yonghong Song wrote:
> On 7/8/19 5:57 AM, Arnd Bergmann wrote:
>> When CONFIG_NET is disabled, we get a link error:
>>
>> kernel/bpf/cgroup.o: In function `__cgroup_bpf_run_filter_setsockopt':
>> cgroup.c:(.text+0x3010): undefined reference to `lock_sock_nested'
>> cgroup.c:(.text+0x3258): undefined reference to `release_sock'
>> kernel/bpf/cgroup.o: In function `__cgroup_bpf_run_filter_getsockopt':
>> cgroup.c:(.text+0x3568): undefined reference to `lock_sock_nested'
>> cgroup.c:(.text+0x3870): undefined reference to `release_sock'
>> kernel/bpf/cgroup.o: In function `cg_sockopt_func_proto':
>> cgroup.c:(.text+0x41d8): undefined reference to `bpf_sk_storage_delete_proto'
>>
>> None of this code is useful in this configuration anyway, so we can
>> simply hide it in an appropriate #ifdef.
>>
>> Fixes: 0d01da6afc54 ("bpf: implement getsockopt and setsockopt hooks")
>> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
>
> FYI.
>
> There is already a patch to fix the same issue,
> https://lore.kernel.org/bpf/e9e489fe-feec-a211-82aa-5df0c6a308d1@huawei.com/T/#t
>
> which has been acked and not merged yet.
Done now, and I've also applied patch 2/2 from here, thanks.
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2019-07-08 15:41 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-07-08 12:57 [PATCH net-next 1/2] bpf: skip sockopt hooks without CONFIG_NET Arnd Bergmann
2019-07-08 12:57 ` [PATCH net-next 2/2] bpf: avoid unused variable warning in tcp_bpf_rtt() Arnd Bergmann
2019-07-08 13:55 ` Soheil Hassas Yeganeh
2019-07-08 15:06 ` [PATCH net-next 1/2] bpf: skip sockopt hooks without CONFIG_NET Yonghong Song
2019-07-08 15:26 ` Daniel Borkmann
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).