Date: Mon, 15 May 2023 11:34:40 +0800
From: "D. Wythe" <alibuda@...ux.alibaba.com>
To: Yonghong Song <yhs@...a.com>, kgraul@...ux.ibm.com, wenjia@...ux.ibm.com,
 jaka@...ux.ibm.com, ast@...nel.org, daniel@...earbox.net, andrii@...nel.org,
 martin.lau@...ux.dev, pabeni@...hat.com, song@...nel.org, sdf@...gle.com,
 haoluo@...gle.com, yhs@...com, edumazet@...gle.com,
 john.fastabend@...il.com, kpsingh@...nel.org, jolsa@...nel.org,
 guwen@...ux.alibaba.com
Cc: kuba@...nel.org, davem@...emloft.net, netdev@...r.kernel.org,
 linux-s390@...r.kernel.org, linux-rdma@...r.kernel.org, bpf@...r.kernel.org
Subject: Re: [PATCH bpf-next v1 4/5] bpf: add smc negotiator support in BPF
 struct_ops



On 5/13/23 10:36 AM, Yonghong Song wrote:
>
>
> On 5/11/23 11:24 PM, D. Wythe wrote:
>> From: "D. Wythe" <alibuda@...ux.alibaba.com>
>>
>> This PATCH attempts to introduce BPF injection capability for SMC.
>> Considering that the SMC protocol is not suitable for all scenarios,
>> especially for short-lived. However, for most applications, they cannot
>> guarantee that there are no such scenarios at all. Therefore, apps
>> may need some specific strategies to decide shall we need to use SMC
>> or not, for example, apps can limit the scope of the SMC to a specific
>> IP address or port.
>>
>> Based on the consideration of transparent replacement, we hope that apps
>> can remain transparent even if they need to formulate some specific
>> strategies for SMC using. That is, do not need to recompile their code.
>>
>> On the other hand, we need to ensure the scalability of strategies
>> implementation. Although it is simple to use socket options or sysctl,
>> it will bring more complexity to subsequent expansion.
>>
>> Fortunately, BPF can solve these concerns very well, users can write
>> thire own strategies in eBPF to choose whether to use SMC or not.
>> And it's quite easy for them to modify their strategies in the future.
>>
>> This PATCH implement injection capability for SMC via struct_ops.
>> In that way, we can add new injection scenarios in the future.
>>
>> Signed-off-by: D. Wythe <alibuda@...ux.alibaba.com>
>> ---
>>   kernel/bpf/bpf_struct_ops_types.h |   4 +
>>   net/Makefile                      |   2 +-
>>   net/smc/bpf_smc.c                 | 171 
>> ++++++++++++++++++++++++++++++++++++++
>>   3 files changed, 176 insertions(+), 1 deletion(-)
>>   create mode 100644 net/smc/bpf_smc.c
>>
>> diff --git a/kernel/bpf/bpf_struct_ops_types.h 
>> b/kernel/bpf/bpf_struct_ops_types.h
>> index 5678a9d..d952b85 100644
>> --- a/kernel/bpf/bpf_struct_ops_types.h
>> +++ b/kernel/bpf/bpf_struct_ops_types.h
>> @@ -9,4 +9,8 @@
>>   #include <net/tcp.h>
>>   BPF_STRUCT_OPS_TYPE(tcp_congestion_ops)
>>   #endif
>> +#if IS_ENABLED(CONFIG_SMC_BPF)
>> +#include <net/smc.h>
>> +BPF_STRUCT_OPS_TYPE(smc_sock_negotiator_ops)
>> +#endif
>>   #endif
>> diff --git a/net/Makefile b/net/Makefile
>> index 222916a..2139fa4 100644
>> --- a/net/Makefile
>> +++ b/net/Makefile
>> @@ -52,7 +52,7 @@ obj-$(CONFIG_TIPC)        += tipc/
>>   obj-$(CONFIG_NETLABEL)        += netlabel/
>>   obj-$(CONFIG_IUCV)        += iucv/
>>   obj-$(CONFIG_SMC)        += smc/
>> -obj-$(CONFIG_SMC_BPF)        += smc/smc_negotiator.o
>> +obj-$(CONFIG_SMC_BPF)        += smc/smc_negotiator.o smc/bpf_smc.o
>>   obj-$(CONFIG_RFKILL)        += rfkill/
>>   obj-$(CONFIG_NET_9P)        += 9p/
>>   obj-$(CONFIG_CAIF)        += caif/
>> diff --git a/net/smc/bpf_smc.c b/net/smc/bpf_smc.c
>> new file mode 100644
>> index 0000000..ac9a9ae91
>> --- /dev/null
>> +++ b/net/smc/bpf_smc.c
>> @@ -0,0 +1,171 @@
>> +// SPDX-License-Identifier: GPL-2.0-only
>> +/*
>> + *  Support eBPF for Shared Memory Communications over RDMA (SMC-R) 
>> and RoCE
>> + *
>> + *  Copyright IBM Corp. 2016, 2018
>
> The above description and copyright sound very wierd.

Received, let me see how to modify it.

>
>> + *
>> + *  Author(s):  D. Wythe <alibuda@...ux.alibaba.com>
>
> One author, so just "Author: ...".

Got it. I will fix that.

>> + */
>> +
>> +#include <linux/bpf_verifier.h>
>> +#include <linux/btf_ids.h>
>> +#include <linux/kernel.h>
>> +#include <linux/bpf.h>
>> +#include <linux/btf.h>
>> +#include "smc_negotiator.h"
>> +
>> +extern struct bpf_struct_ops bpf_smc_sock_negotiator_ops;
>> +static u32 smc_sock_id, sock_id;
>> +
>> +static int bpf_smc_negotiator_init(struct btf *btf)
>> +{
>> +    s32 type_id;
>> +
>> +    type_id = btf_find_by_name_kind(btf, "sock", BTF_KIND_STRUCT);
>> +    if (type_id < 0)
>> +        return -EINVAL;
>> +    sock_id = type_id;
>> +
>> +    type_id = btf_find_by_name_kind(btf, "smc_sock", BTF_KIND_STRUCT);
>> +    if (type_id < 0)
>> +        return -EINVAL;
>> +    smc_sock_id = type_id;
>> +
>> +    return 0;
>> +}
>> +
>> +/* register ops */
>> +static int bpf_smc_negotiator_reg(void *kdata)
>> +{
>> +    return smc_sock_register_negotiator_ops(kdata);
>> +}
>> +
>> +/* unregister ops */
>> +static void bpf_smc_negotiator_unreg(void *kdata)
>> +{
>> +    smc_sock_unregister_negotiator_ops(kdata);
>> +}
>> +
>> +/* unregister ops */
>
> update ops?
> Also I think the above comments like
> 'register ops', 'unregister ops' and 'update ops' are not
> necessary. The code itself is self-explanary.
My mistake, thank you very much for your suggestion. The annotations here
are unnecessary indeed.
>
>> +static int bpf_smc_negotiator_update(void *kdata, void *old_kdata)
>> +{
>> +    return smc_sock_update_negotiator_ops(kdata, old_kdata);
>> +}
>> +
>> +static int bpf_smc_negotiator_validate(void *kdata)
>> +{
>> +    return smc_sock_validate_negotiator_ops(kdata);
>> +}
>> +
>> +static int bpf_smc_negotiator_check_member(const struct btf_type *t,
>> +                       const struct btf_member *member,
>> +                       const struct bpf_prog *prog)
>> +{
>> +    return 0;
>> +}
>> +
>> +static int bpf_smc_negotiator_init_member(const struct btf_type *t,
>> +                      const struct btf_member *member,
>> +                      void *kdata, const void *udata)
>> +{
>> +    const struct smc_sock_negotiator_ops *uops;
>> +    struct smc_sock_negotiator_ops *ops;
>> +    u32 moff;
>> +
>> +    uops = (const struct smc_sock_negotiator_ops *)udata;
>> +    ops = (struct smc_sock_negotiator_ops *)kdata;
>> +
>> +    moff = __btf_member_bit_offset(t, member) / 8;
>> +
>> +    /* init name */
>> +    if (moff ==  offsetof(struct smc_sock_negotiator_ops, name)) {
>> +        if (bpf_obj_name_cpy(ops->name, uops->name,
>> +                     sizeof(uops->name)) <= 0)
>> +            return -EINVAL;
>> +        return 1;
>> +    }
>> +
>> +    return 0;
>> +}
>> +
>> +BPF_CALL_1(bpf_smc_skc_to_tcp_sock, struct sock *, sk)
>> +{
>> +    if (sk && sk_fullsock(sk) && sk->sk_family == AF_SMC)
>> +        return (unsigned long)((struct smc_sock *)(sk))->clcsock->sk;
>> +
>> +    return (unsigned long)NULL;
>> +}
>> +
>> +static const struct bpf_func_proto bpf_smc_skc_to_tcp_sock_proto = {
>> +    .func            = bpf_smc_skc_to_tcp_sock,
>> +    .gpl_only        = false,
>> +    .ret_type        = RET_PTR_TO_BTF_ID_OR_NULL,
>> +    .arg1_type        = ARG_PTR_TO_BTF_ID_SOCK_COMMON,
>> +    .ret_btf_id        = &btf_sock_ids[BTF_SOCK_TYPE_TCP],
>> +};
>> +
>> +static const struct bpf_func_proto *
>> +smc_negotiator_prog_func_proto(enum bpf_func_id func_id, const 
>> struct bpf_prog *prog)
>> +{
>> +    const struct btf_member *m;
>> +    const struct btf_type *t;
>> +    u32 midx, moff;
>> +
>> +    midx = prog->expected_attach_type;
>> +    t = bpf_smc_sock_negotiator_ops.type;
>> +    m = &btf_type_member(t)[midx];
>> +
>> +    moff = __btf_member_bit_offset(t, m) / 8;
>> +
>> +    switch (func_id) {
>> +    case BPF_FUNC_setsockopt:
>> +        switch (moff) {
>> +        /* Avoid potential deadloop risk */
>> +        case offsetof(struct smc_sock_negotiator_ops, init):
>> +            fallthrough;
>
> I am not sure whether a 'fallthrough' is needed here or since the case
> itself does not have any code. Any warning will show up if
> 'fallthrough;' is removed?

Yes, if there is no code, fallthrough is unnecessary, I will fix it in 
the next version.

>
>> +        /* Avoid potential leak risk */
>
> I think more detailed explanation about 'deadloop risk' and 'leak risk'
> is necessary.

Got it, i will add more detailed explanation.
>
>> +        case offsetof(struct smc_sock_negotiator_ops, release):
>> +            return NULL;
>> +        }
>> +        return &bpf_sk_setsockopt_proto;
>> +    case BPF_FUNC_getsockopt:
>> +        return &bpf_sk_getsockopt_proto;
>> +    case BPF_FUNC_skc_to_tcp_sock:
>> +        return &bpf_smc_skc_to_tcp_sock_proto;
>> +    default:
>> +        return bpf_base_func_proto(func_id);
>> +    }
>> +}
>> +
>> +static bool smc_negotiator_prog_is_valid_access(int off, int size, 
>> enum bpf_access_type type,
>> +                        const struct bpf_prog *prog,
>> +                        struct bpf_insn_access_aux *info)
>> +{
>> +    if (!bpf_tracing_btf_ctx_access(off, size, type, prog, info))
>> +        return false;
>> +
>> +    /* promote it to smc_sock */
>> +    if (base_type(info->reg_type) == PTR_TO_BTF_ID &&
>> +        !bpf_type_has_unsafe_modifiers(info->reg_type) &&
>> +        info->btf_id == sock_id)
>> +        info->btf_id = smc_sock_id;
>> +
>> +    return true;
>> +}
>> +
>> +static const struct bpf_verifier_ops bpf_smc_negotiator_verifier_ops 
>> = {
>> +    .get_func_proto  = smc_negotiator_prog_func_proto,
>> +    .is_valid_access = smc_negotiator_prog_is_valid_access,
>> +};
>> +
>> +struct bpf_struct_ops bpf_smc_sock_negotiator_ops = {
>> +    .verifier_ops = &bpf_smc_negotiator_verifier_ops,
>> +    .init = bpf_smc_negotiator_init,
>> +    .check_member = bpf_smc_negotiator_check_member,
>> +    .init_member = bpf_smc_negotiator_init_member,
>> +    .reg = bpf_smc_negotiator_reg,
>> +    .update = bpf_smc_negotiator_update,
>> +    .unreg = bpf_smc_negotiator_unreg,
>> +    .validate = bpf_smc_negotiator_validate,
>> +    .name = "smc_sock_negotiator_ops",
>> +};
>> \ No newline at end of file
>
> Empty line at the end?

Will fix that, thanks.

Powered by blists - more mailing lists