From: Jon Doron <arilou@gmail.com>
To: Song Liu <song@kernel.org>
Cc: bpf <bpf@vger.kernel.org>, Alexei Starovoitov <ast@kernel.org>,
Andrii Nakryiko <andrii@kernel.org>,
Daniel Borkmann <daniel@iogearbox.net>, Jon Doron <jond@wiz.io>
Subject: Re: [PATCH v1 1/1] libbpf: perfbuf: allow raw access to buffers
Date: Fri, 8 Jul 2022 09:08:12 +0300 [thread overview]
Message-ID: <YsfJzASkXAFSFCAy@jondnuc> (raw)
In-Reply-To: <CAPhsuW7k3ExOjjonL=-=sg_rx0kyrgF=k7jxb+W0ZpxgxnoeLA@mail.gmail.com>
On 07/07/2022, Song Liu wrote:
Hi Song Liu, thank you for the fast reply, I'll add comments on top
of your original email.
I will not be available through out next week, so sorry if I'll have a
late reply for follow ups.
Thanks in advance,
-- Jon.
>On Thu, Jul 7, 2022 at 12:14 AM Jon Doron <arilou@gmail.com> wrote:
>
>Please prefix the patch with the target tree (bpf or bpf-next). For example,
>this patch should go via bpf-next.
Done
>>
>> From: Jon Doron <jond@wiz.io>
>>
>> Add API for perfbuf to support writing a custom event reader.
>
>This is too brief for such change. It is ok to duplicate text from the cover
>letter.
>
Done
>Please also update libbpf.map.
>
Done
>Also, we should add a selftest to use these new APIs.
>
I'm not sure I'm following here what were you had in mind, in practice I
could just change **bpf_perf_event_read_simple** implementation to use
these new APIs to initialize it's variables (data_head, data_tail,
base, mmap_size) and it would act exactly the same.
>>
>> Signed-off-by: Jon Doron <jond@wiz.io>
>> ---
>> tools/lib/bpf/libbpf.c | 40 ++++++++++++++++++++++++++++++++++++++++
>> tools/lib/bpf/libbpf.h | 6 ++++++
>> 2 files changed, 46 insertions(+)
>>
>> diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
>> index e89cc9c885b3..37299aa05185 100644
>> --- a/tools/lib/bpf/libbpf.c
>> +++ b/tools/lib/bpf/libbpf.c
>> @@ -12433,6 +12433,46 @@ static int perf_buffer__process_records(struct perf_buffer *pb,
>> return 0;
>> }
>>
>> +int perf_buffer__raw_ring_buf(const struct perf_buffer *pb, size_t buf_idx,
>> + void **base, size_t *buf_size, __u64 *head,
>> + __u64 *tail)
>
>Please add comments to each API function.
Done
>> +{
>> + struct perf_cpu_buf *cpu_buf;
>> + struct perf_event_mmap_page *header;
>> +
>> + if (buf_idx >= pb->cpu_cnt)
>> + return libbpf_err(-EINVAL);
>> +
>> + cpu_buf = pb->cpu_bufs[buf_idx];
>> + if (!cpu_buf)
>> + return libbpf_err(-ENOENT);
>> +
>> + header = cpu_buf->base;
>> + *head = ring_buffer_read_head(header);
>> + *tail = header->data_tail;
>> + *base = ((__u8 *)header) + pb->page_size;
>> + *buf_size = pb->mmap_size;
>> + return 0;
>> +}
>> +
>> +int perf_buffer__set_ring_buf_tail(const struct perf_buffer *pb, size_t buf_idx,
>> + __u64 tail)
>> +{
>> + struct perf_cpu_buf *cpu_buf;
>> + struct perf_event_mmap_page *header;
>> +
>> + if (buf_idx >= pb->cpu_cnt)
>> + return libbpf_err(-EINVAL);
>> +
>> + cpu_buf = pb->cpu_bufs[buf_idx];
>> + if (!cpu_buf)
>> + return libbpf_err(-ENOENT);
>> +
>> + header = cpu_buf->base;
>> + ring_buffer_write_tail(header, tail);
>> + return 0;
>> +}
>> +
>> int perf_buffer__epoll_fd(const struct perf_buffer *pb)
>> {
>> return pb->epoll_fd;
>> diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h
>> index 9e9a3fd3edd8..b6f6b6a12d70 100644
>> --- a/tools/lib/bpf/libbpf.h
>> +++ b/tools/lib/bpf/libbpf.h
>> @@ -1381,6 +1381,12 @@ LIBBPF_API int perf_buffer__consume(struct perf_buffer *pb);
>> LIBBPF_API int perf_buffer__consume_buffer(struct perf_buffer *pb, size_t buf_idx);
>> LIBBPF_API size_t perf_buffer__buffer_cnt(const struct perf_buffer *pb);
>> LIBBPF_API int perf_buffer__buffer_fd(const struct perf_buffer *pb, size_t buf_idx);
>> +LIBBPF_API int perf_buffer__raw_ring_buf(const struct perf_buffer *pb,
>> + size_t buf_idx, void **base,
>> + size_t *buf_size, __u64 *head,
>> + __u64 *tail);
>> +LIBBPF_API int perf_buffer__set_ring_buf_tail(const struct perf_buffer *pb,
>> + size_t buf_idx, __u64 tail);
>>
>> typedef enum bpf_perf_event_ret
>> (*bpf_perf_event_print_t)(struct perf_event_header *hdr,
>> --
>> 2.36.1
>>
prev parent reply other threads:[~2022-07-08 6:08 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-07-07 7:13 [PATCH v1 0/1] libbpf: perfbuf custom event reader Jon Doron
2022-07-07 7:13 ` [PATCH v1 1/1] libbpf: perfbuf: allow raw access to buffers Jon Doron
2022-07-08 5:26 ` Song Liu
2022-07-08 6:08 ` Jon Doron [this message]
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=YsfJzASkXAFSFCAy@jondnuc \
--to=arilou@gmail.com \
--cc=andrii@kernel.org \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=jond@wiz.io \
--cc=song@kernel.org \
/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