From: KP Singh <kpsingh@chromium.org>
To: Andrii Nakryiko <andrii.nakryiko@gmail.com>
Cc: open list <linux-kernel@vger.kernel.org>,
bpf <bpf@vger.kernel.org>, Alexei Starovoitov <ast@kernel.org>,
Daniel Borkmann <daniel@iogearbox.net>,
Paul Turner <pjt@google.com>,
Florent Revest <revest@chromium.org>,
Brendan Jackman <jackmanb@chromium.org>
Subject: Re: [PATCH bpf-next 3/7] bpf: Introduce BPF_MODIFY_RETURN
Date: Tue, 3 Mar 2020 23:51:24 +0100 [thread overview]
Message-ID: <20200303225124.GA12968@chromium.org> (raw)
In-Reply-To: <CAEf4BzZVV12WoHDnQSfOKpndr3qVLEAz8itMcdqnQq8Q4njc0w@mail.gmail.com>
On 03-Mär 14:37, Andrii Nakryiko wrote:
> On Tue, Mar 3, 2020 at 6:12 AM KP Singh <kpsingh@chromium.org> wrote:
> >
> > From: KP Singh <kpsingh@google.com>
> >
> > When multiple programs are attached, each program receives the return
> > value from the previous program on the stack and the last program
> > provides the return value to the attached function.
> >
> > The fmod_ret bpf programs are run after the fentry programs and before
> > the fexit programs. The original function is only called if all the
> > fmod_ret programs return 0 to avoid any unintended side-effects. The
> > success value, i.e. 0 is not currently configurable but can be made so
> > where user-space can specify it at load time.
> >
> > For example:
> >
> > int func_to_be_attached(int a, int b)
> > { <--- do_fentry
> >
> > do_fmod_ret:
> > <update ret by calling fmod_ret>
> > if (ret != 0)
> > goto do_fexit;
> >
> > original_function:
> >
> > <side_effects_happen_here>
> >
> > } <--- do_fexit
> >
> > The fmod_ret program attached to this function can be defined as:
> >
> > SEC("fmod_ret/func_to_be_attached")
> > BPF_PROG(func_name, int a, int b, int ret)
>
> same as on cover letter, return type is missing
Fixed. Thanks!
>
> > {
> > // This will skip the original function logic.
> > return 1;
> > }
> >
> > The first fmod_ret program is passed 0 in its return argument.
> >
> > Signed-off-by: KP Singh <kpsingh@google.com>
> > ---
> > arch/x86/net/bpf_jit_comp.c | 96 ++++++++++++++++++++++++++++++++--
> > include/linux/bpf.h | 1 +
> > include/uapi/linux/bpf.h | 1 +
> > kernel/bpf/btf.c | 3 +-
> > kernel/bpf/syscall.c | 1 +
> > kernel/bpf/trampoline.c | 5 +-
> > kernel/bpf/verifier.c | 1 +
> > tools/include/uapi/linux/bpf.h | 1 +
> > 8 files changed, 103 insertions(+), 6 deletions(-)
> >
>
> [...]
>
> >
> > + if (fmod_ret->nr_progs) {
> > + branches = kcalloc(fmod_ret->nr_progs, sizeof(u8 *),
> > + GFP_KERNEL);
> > + if (!branches)
> > + return -ENOMEM;
> > + if (invoke_bpf_mod_ret(m, &prog, fmod_ret, stack_size,
> > + branches))
>
> branches leaks here
Good catch, sloppy work here by me.
>
> > + return -EINVAL;
> > + }
> > +
> > if (flags & BPF_TRAMP_F_CALL_ORIG) {
> > - if (fentry->nr_progs)
> > + if (fentry->nr_progs || fmod_ret->nr_progs)
> > restore_regs(m, &prog, nr_args, stack_size);
> >
> > /* call original function */
> > @@ -1573,6 +1649,14 @@ int arch_prepare_bpf_trampoline(void *image, void *image_end,
>
> there is early return one line above here, you need to free branches
> in that case to not leak memory
>
> So I guess it's better to do goto cleanup approach at this point?
yeah, agreed, updated to doing a cleanup at the end.
- KP
>
> > emit_stx(&prog, BPF_DW, BPF_REG_FP, BPF_REG_0, -8);
> > }
> >
> > + if (fmod_ret->nr_progs) {
> > + align16_branch_target(&prog);
> > + for (i = 0; i < fmod_ret->nr_progs; i++)
> > + emit_cond_near_jump(&branches[i], prog, branches[i],
> > + X86_JNE);
> > + kfree(branches);
> > + }
> > +
>
> [...]
next prev parent reply other threads:[~2020-03-03 22:51 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-03-03 14:09 [PATCH bpf-next 0/7] Introduce BPF_MODIFY_RET tracing progs KP Singh
2020-03-03 14:09 ` [PATCH bpf-next 1/7] bpf: Refactor trampoline update code KP Singh
2020-03-03 22:12 ` Andrii Nakryiko
2020-03-03 22:24 ` KP Singh
2020-03-03 23:03 ` Andrii Nakryiko
2020-03-03 23:08 ` KP Singh
2020-03-03 14:09 ` [PATCH bpf-next 2/7] bpf: JIT helpers for fmod_ret progs KP Singh
2020-03-03 22:26 ` Andrii Nakryiko
2020-03-03 22:28 ` KP Singh
2020-03-03 23:56 ` Alexei Starovoitov
2020-03-04 1:26 ` KP Singh
2020-03-03 14:09 ` [PATCH bpf-next 3/7] bpf: Introduce BPF_MODIFY_RETURN KP Singh
2020-03-03 22:37 ` Andrii Nakryiko
2020-03-03 22:51 ` KP Singh [this message]
2020-03-03 14:09 ` [PATCH bpf-next 4/7] bpf: Attachment verification for BPF_MODIFY_RETURN KP Singh
2020-03-03 22:44 ` Andrii Nakryiko
2020-03-03 23:21 ` KP Singh
2020-03-04 0:03 ` Alexei Starovoitov
2020-03-04 1:06 ` KP Singh
2020-03-03 14:09 ` [PATCH bpf-next 5/7] tools/libbpf: Add support " KP Singh
2020-03-03 22:45 ` Andrii Nakryiko
2020-03-03 14:09 ` [PATCH bpf-next 6/7] bpf: Add test ops for BPF_PROG_TYPE_TRACING KP Singh
2020-03-03 22:51 ` Andrii Nakryiko
2020-03-03 22:57 ` KP Singh
2020-03-03 14:09 ` [PATCH bpf-next 7/7] bpf: Add selftests for BPF_MODIFY_RETURN KP Singh
2020-03-03 22:58 ` Andrii Nakryiko
2020-03-03 22:12 ` [PATCH bpf-next 0/7] Introduce BPF_MODIFY_RET tracing progs Andrii Nakryiko
2020-03-03 22:25 ` KP Singh
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=20200303225124.GA12968@chromium.org \
--to=kpsingh@chromium.org \
--cc=andrii.nakryiko@gmail.com \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=jackmanb@chromium.org \
--cc=linux-kernel@vger.kernel.org \
--cc=pjt@google.com \
--cc=revest@chromium.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;
as well as URLs for NNTP newsgroup(s).