BPF List
 help / color / mirror / Atom feed
From: Eduard Zingerman <eddyz87@gmail.com>
To: Leon Hwang <leon.hwang@linux.dev>, bpf@vger.kernel.org
Cc: ast@kernel.org, daniel@iogearbox.net, andrii@kernel.org,
	toke@redhat.com,  martin.lau@kernel.org, yonghong.song@linux.dev,
	puranjay@kernel.org,  xukuohai@huaweicloud.com,
	iii@linux.ibm.com, kernel-patches-bot@fb.com
Subject: Re: [PATCH bpf-next v4 0/4] bpf: Fix tailcall infinite loop caused by freplace
Date: Tue, 01 Oct 2024 04:14:08 -0700	[thread overview]
Message-ID: <1ea99d1e31c3f52f8962b186a150dfb0ffd23e45.camel@gmail.com> (raw)
In-Reply-To: <20240929132757.79826-1-leon.hwang@linux.dev>

On Sun, 2024-09-29 at 21:27 +0800, Leon Hwang wrote:
> Previously, I fixed a tailcall infinite loop issue caused by trampoline[0].
> 
> At this time, I fix a tailcall infinite loop issue caused by tailcall and
> freplace combination by preventing updating extended prog to prog_array map
> and preventing extending tail callee prog with freplace:
> 
> 1. If a prog or its subprog has been extended by freplace prog, the prog
>    can not be updated to prog_array map.
> 2. If a prog has been updated to prog_array map, it or its subprog can not
>    be extended by freplace prog.

So, once this series is applied we essentially have:
- three variables:
  - tgt_prog->aux->is_extended
  - tgt_prog->aux->prog_array_member_cnt
  - trampoline->extension_prog
- four operations:
  - link/attach extension program 'prog' using trampoline 'tr'
  - unlink/detach extension program 'prog' using trampoline 'tr'
  - put program 'tgt_prog' into prog array
  - remove program 'tgt_prog' from prog array

And above four operations have the following pseudo-code with regards
to update of the variables:

- link/attach extension program 'prog' using trampoline 'tr':

    with lock(tgt_prog->ext_mutex):
      if tgt_prog->aux->prog_array_member_cnt:
         return error
      if tr->extension_prog:
         return error
      tr->extension_prog = prog
      tgt_prog->is_extended = true

- unlink/detach extension program 'prog' using trampoline 'tr':

    with lock(tgt_prog->ext_mutex):
      tr->extension_prog = NULL
      tgt_prog->is_extended = false

- put program 'tgt_prog' into prog array:

    with lock(tgt_prog->ext_mutex):
      if tgt_prog->aux->is_extended:
         return error
      tgt_prog->aux->prog_array_member_cnt++

- remove program 'tgt_prog' from prog array:

    with lock(tgt_prog->ext_mutex):
      tgt_prog->aux->prog_array_member_cnt--

I think this is correct, would be great if someone with more
concurrency related experience would take a look.

[...]


      parent reply	other threads:[~2024-10-01 11:14 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-09-29 13:27 [PATCH bpf-next v4 0/4] bpf: Fix tailcall infinite loop caused by freplace Leon Hwang
2024-09-29 13:27 ` [PATCH bpf-next v4 1/4] bpf: Prevent updating extended prog to prog_array map Leon Hwang
2024-10-01 11:13   ` Eduard Zingerman
2024-10-01 13:20     ` Leon Hwang
2024-10-01 16:54       ` Eduard Zingerman
2024-09-29 13:27 ` [PATCH bpf-next v4 2/4] bpf: Prevent extending tail callee prog with freplace prog Leon Hwang
2024-09-30  1:53   ` Leon Hwang
2024-10-04 19:33     ` Alexei Starovoitov
2024-10-04 20:37       ` Eduard Zingerman
2024-10-04 20:52       ` Eduard Zingerman
2024-10-04 23:30         ` Alexei Starovoitov
2024-09-29 13:27 ` [PATCH bpf-next v4 3/4] selftests/bpf: Add a test case to confirm a tailcall infinite loop issue has been prevented Leon Hwang
2024-09-29 13:27 ` [PATCH bpf-next v4 4/4] selftests/bpf: Add cases to test tailcall in freplace Leon Hwang
2024-10-01 11:14 ` Eduard Zingerman [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=1ea99d1e31c3f52f8962b186a150dfb0ffd23e45.camel@gmail.com \
    --to=eddyz87@gmail.com \
    --cc=andrii@kernel.org \
    --cc=ast@kernel.org \
    --cc=bpf@vger.kernel.org \
    --cc=daniel@iogearbox.net \
    --cc=iii@linux.ibm.com \
    --cc=kernel-patches-bot@fb.com \
    --cc=leon.hwang@linux.dev \
    --cc=martin.lau@kernel.org \
    --cc=puranjay@kernel.org \
    --cc=toke@redhat.com \
    --cc=xukuohai@huaweicloud.com \
    --cc=yonghong.song@linux.dev \
    /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