From: Yonghong Song <yhs@fb.com>
To: Brian Vazquez <brianvv@google.com>
Cc: Brian Vazquez <brianvv.kernel@gmail.com>,
Alexei Starovoitov <ast@kernel.org>,
Daniel Borkmann <daniel@iogearbox.net>,
"David S . Miller" <davem@davemloft.net>,
Stanislav Fomichev <sdf@google.com>,
Petar Penkov <ppenkov@google.com>,
Willem de Bruijn <willemb@google.com>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
"netdev@vger.kernel.org" <netdev@vger.kernel.org>,
"bpf@vger.kernel.org" <bpf@vger.kernel.org>
Subject: Re: [PATCH v2 bpf-next 2/9] bpf: add generic support for lookup and lookup_and_delete batch ops
Date: Fri, 22 Nov 2019 00:34:15 +0000 [thread overview]
Message-ID: <e9cb66e4-16fd-8850-4755-3034f75788c5@fb.com> (raw)
In-Reply-To: <CAMzD94TfpQaFN=7cQR9kmHun0gZNF2oMwEJu7aZMYhsYhvgRDg@mail.gmail.com>
On 11/21/19 1:36 PM, Brian Vazquez wrote:
> Hi Yonghong,
> thanks for reviewing the patch, I will fix all the direct returns and
> small fixes in next version.
>
> On Thu, Nov 21, 2019 at 9:36 AM Yonghong Song <yhs@fb.com> wrote:
>>
>>
>>
>> On 11/19/19 11:30 AM, Brian Vazquez wrote:
>>> This commit introduces generic support for the bpf_map_lookup_batch and
>>> bpf_map_lookup_and_delete_batch ops. This implementation can be used by
>>> almost all the bpf maps since its core implementation is relying on the
>>> existing map_get_next_key, map_lookup_elem and map_delete_elem
>>> functions. The bpf syscall subcommands introduced are:
>>>
[...]
>>> + for (cp = 0; cp < max_count; cp++) {
>>> + if (cp || first_key) {
>>> + rcu_read_lock();
>>> + err = map->ops->map_get_next_key(map, prev_key, key);
>>> + rcu_read_unlock();
>>> + if (err)
>>> + break;
>>> + }
>>> + err = bpf_map_copy_value(map, key, value,
>>> + attr->batch.elem_flags, do_delete);
>>> +
>>> + if (err == -ENOENT) {
>>> + if (retry) {
>>> + retry--;
>>
>> What is the 'retry' semantics here? After 'continue', cp++ is executed.
>
> Good catch, I'll move cp++ to a proper place. retry is used to prevent
> the cases where the map is doing many concurrent additions and
> deletions, this could result in map_get_next_key succeeding but
> bpf_map_copy_value failing, in which case I think it'd be better to
> try and find a next elem, but we don't want to do this for more than 3
> times.
>
>>
>>> + continue;
>>> + }
>>> + err = -EINTR;
>>
>> Why returning -EINTR?
>
> I thought that this is the err more appropriate for the behaviour I
> describe above. Should I handle that case? WDYT?
I see. We do not want to use -ENOENT since get_next_key may return
-ENOENT. I think -EINTR is okay here to indicate we have key, but
key/value entry is gone right before the attempted access.
next prev parent reply other threads:[~2019-11-22 0:34 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-11-19 19:30 [PATCH v2 bpf-next 0/9] add bpf batch ops to process more than 1 elem Brian Vazquez
2019-11-19 19:30 ` [PATCH v2 bpf-next 1/9] bpf: add bpf_map_{value_size,update_value,map_copy_value} functions Brian Vazquez
2019-11-22 16:36 ` John Fastabend
2019-11-19 19:30 ` [PATCH v2 bpf-next 2/9] bpf: add generic support for lookup and lookup_and_delete batch ops Brian Vazquez
2019-11-21 17:36 ` Yonghong Song
2019-11-21 21:36 ` Brian Vazquez
2019-11-22 0:34 ` Yonghong Song [this message]
2019-11-22 17:25 ` John Fastabend
2019-11-19 19:30 ` [PATCH v2 bpf-next 3/9] bpf: add generic support for update and delete " Brian Vazquez
2019-11-21 18:00 ` Yonghong Song
2019-11-22 5:50 ` Brian Vazquez
2019-11-22 6:56 ` Yonghong Song
2019-11-19 19:30 ` [PATCH v2 bpf-next 4/9] bpf: add lookup and updated batch ops to arraymap Brian Vazquez
2019-11-19 19:30 ` [PATCH v2 bpf-next 5/9] bpf: add batch ops to all htab bpf map Brian Vazquez
2019-11-21 18:27 ` Yonghong Song
2019-11-21 21:27 ` Brian Vazquez
2019-11-19 19:30 ` [PATCH v2 bpf-next 6/9] tools/bpf: sync uapi header bpf.h Brian Vazquez
2019-11-19 19:30 ` [PATCH v2 bpf-next 7/9] libbpf: add libbpf support to batch ops Brian Vazquez
2019-11-21 18:30 ` Yonghong Song
2019-11-19 19:30 ` [PATCH v2 bpf-next 8/9] selftests/bpf: add batch ops testing for hmap and hmap_percpu Brian Vazquez
2019-11-21 18:36 ` Yonghong Song
2019-11-21 21:16 ` Brian Vazquez
2019-11-19 19:30 ` [PATCH v2 bpf-next 9/9] selftests/bpf: add batch ops testing to array bpf map Brian Vazquez
2019-11-21 18:43 ` Yonghong Song
2019-11-21 21:14 ` Brian Vazquez
2019-11-22 0:22 ` 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=e9cb66e4-16fd-8850-4755-3034f75788c5@fb.com \
--to=yhs@fb.com \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=brianvv.kernel@gmail.com \
--cc=brianvv@google.com \
--cc=daniel@iogearbox.net \
--cc=davem@davemloft.net \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=ppenkov@google.com \
--cc=sdf@google.com \
--cc=willemb@google.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;
as well as URLs for NNTP newsgroup(s).