From: Yonghong Song <yhs@fb.com>
To: Jiri Olsa <olsajiri@gmail.com>
Cc: bpf@vger.kernel.org, Alexei Starovoitov <ast@kernel.org>,
Andrii Nakryiko <andrii@kernel.org>,
Daniel Borkmann <daniel@iogearbox.net>,
kernel-team@fb.com
Subject: Re: [PATCH bpf-next v3 1/7] bpf: Allow struct argument in trampoline based programs
Date: Tue, 30 Aug 2022 10:03:03 -0700 [thread overview]
Message-ID: <0f254eba-e178-ecfb-c862-7bab26e8e87e@fb.com> (raw)
In-Reply-To: <Yw3+XcnFkT16Z3dx@krava>
On 8/30/22 5:11 AM, Jiri Olsa wrote:
> On Sat, Aug 27, 2022 at 07:54:43PM -0700, Yonghong Song wrote:
>
> SNIP
>
>> diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c
>> index 903719b89238..4a081bfb4c8a 100644
>> --- a/kernel/bpf/btf.c
>> +++ b/kernel/bpf/btf.c
>> @@ -5328,6 +5328,31 @@ static bool is_int_ptr(struct btf *btf, const struct btf_type *t)
>> return btf_type_is_int(t);
>> }
>>
>> +static u32 get_ctx_arg_idx(struct btf *btf, const struct btf_type *func_proto,
>> + int off)
>> +{
>> + const struct btf_param *args;
>> + const struct btf_type *t;
>> + u32 offset = 0, nr_args;
>> + int i;
>> +
>> + nr_args = btf_type_vlen(func_proto);
>> + args = (const struct btf_param *)(func_proto + 1);
>> + for (i = 0; i < nr_args; i++) {
>> + t = btf_type_skip_modifiers(btf, args[i].type, NULL);
>> + offset += btf_type_is_ptr(t) ? 8 : roundup(t->size, 8);
>> + if (off < offset)
>> + return i;
>> + }
>> +
>> + t = btf_type_skip_modifiers(btf, func_proto->type, NULL);
>> + offset += btf_type_is_ptr(t) ? 8 : roundup(t->size, 8);
>> + if (off < offset)
>> + return nr_args;
>> +
>> + return nr_args + 1;
>> +}
>> +
>> bool btf_ctx_access(int off, int size, enum bpf_access_type type,
>> const struct bpf_prog *prog,
>> struct bpf_insn_access_aux *info)
>> @@ -5347,7 +5372,7 @@ bool btf_ctx_access(int off, int size, enum bpf_access_type type,
>> tname, off);
>> return false;
>> }
>> - arg = off / 8;
>> + arg = t == NULL ? (off / 8) : get_ctx_arg_idx(btf, t, off);
>
> is the t == NULL check needed? we relied on t being defined later in the code
t is defined earlier in
const struct btf_type *t = prog->aux->attach_func_proto;
so we should be fine.
I guess I will fold
if (!t)
return off / 8;
inside get_ctx_arg_idx() to make interface cleaner.
Will do this in the next revision.
>
> jirka
>
>> args = (const struct btf_param *)(t + 1);
>> /* if (t == NULL) Fall back to default BPF prog with
>> * MAX_BPF_FUNC_REG_ARGS u64 arguments.
>> @@ -5417,7 +5442,7 @@ bool btf_ctx_access(int off, int size, enum bpf_access_type type,
>> /* skip modifiers */
>> while (btf_type_is_modifier(t))
>> t = btf_type_by_id(btf, t->type);
>> - if (btf_type_is_small_int(t) || btf_is_any_enum(t))
>> + if (btf_type_is_small_int(t) || btf_is_any_enum(t) || __btf_type_is_struct(t))
>> /* accessing a scalar */
>> return true;
>> if (!btf_type_is_ptr(t)) {
>> @@ -5881,7 +5906,7 @@ static int __get_type_size(struct btf *btf, u32 btf_id,
>> if (btf_type_is_ptr(t))
>> /* kernel size of pointer. Not BPF's size of pointer*/
>> return sizeof(void *);
>> - if (btf_type_is_int(t) || btf_is_any_enum(t))
>> + if (btf_type_is_int(t) || btf_is_any_enum(t) || __btf_type_is_struct(t))
>> return t->size;
>> return -EINVAL;
>
> SNIP
next prev parent reply other threads:[~2022-08-30 17:03 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-08-28 2:54 [PATCH bpf-next v3 0/7] bpf: Support struct argument for trampoline base progs Yonghong Song
2022-08-28 2:54 ` [PATCH bpf-next v3 1/7] bpf: Allow struct argument in trampoline based programs Yonghong Song
2022-08-30 12:11 ` Jiri Olsa
2022-08-30 17:03 ` Yonghong Song [this message]
2022-08-28 2:54 ` [PATCH bpf-next v3 2/7] bpf: x86: Support in-register struct arguments in trampoline programs Yonghong Song
2022-08-28 2:54 ` [PATCH bpf-next v3 3/7] bpf: Update descriptions for helpers bpf_get_func_arg[_cnt]() Yonghong Song
2022-08-28 2:54 ` [PATCH bpf-next v3 4/7] bpf: arm64: No support of struct argument in trampoline programs Yonghong Song
2022-08-28 2:55 ` [PATCH bpf-next v3 5/7] libbpf: Add new BPF_PROG2 macro Yonghong Song
2022-08-30 0:32 ` Kui-Feng Lee
2022-08-30 16:50 ` Yonghong Song
2022-08-28 2:55 ` [PATCH bpf-next v3 6/7] selftests/bpf: Add struct argument tests with fentry/fexit programs Yonghong Song
2022-08-29 22:12 ` Daniel Borkmann
2022-08-30 12:11 ` Jiri Olsa
2022-08-30 17:20 ` Yonghong Song
2022-08-30 16:36 ` Yonghong Song
2022-08-28 2:55 ` [PATCH bpf-next v3 7/7] selftests/bpf: Use BPF_PROG2 for some fentry programs without struct arguments Yonghong Song
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=0f254eba-e178-ecfb-c862-7bab26e8e87e@fb.com \
--to=yhs@fb.com \
--cc=andrii@kernel.org \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=kernel-team@fb.com \
--cc=olsajiri@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox