All of lore.kernel.org
 help / color / mirror / Atom feed
From: Dmitrii Banshchikov <me@ubique.spb.ru>
To: Song Liu <song@kernel.org>
Cc: bpf <bpf@vger.kernel.org>, Alexei Starovoitov <ast@kernel.org>,
	"David S . Miller" <davem@davemloft.net>,
	Daniel Borkmann <daniel@iogearbox.net>,
	Andrii Nakryiko <andrii@kernel.org>,
	Martin KaFai Lau <kafai@fb.com>, Song Liu <songliubraving@fb.com>,
	Yonghong Song <yhs@fb.com>,
	John Fastabend <john.fastabend@gmail.com>,
	KP Singh <kpsingh@kernel.org>,
	Networking <netdev@vger.kernel.org>, Andrey Ignatov <rdna@fb.com>
Subject: Re: [PATCH bpf-next 02/11] bpfilter: Add logging facility
Date: Thu, 20 May 2021 11:08:07 +0400	[thread overview]
Message-ID: <20210520070807.cpmloff4urdsifuy@amnesia> (raw)
In-Reply-To: <CAPhsuW4osuNOagPRwUB30tk3V=ECANktt9jzb+NK1mqOamouSQ@mail.gmail.com>

On Wed, May 19, 2021 at 10:32:25AM -0700, Song Liu wrote:
> On Tue, May 18, 2021 at 11:05 PM Dmitrii Banshchikov <me@ubique.spb.ru> wrote:
> >
> > There are three logging levels for messages: FATAL, NOTICE and DEBUG.
> > When a message is logged with FATAL level it results in bpfilter
> > usermode helper termination.
> 
> Could you please explain why we choose to have 3 levels? Will we need
> more levels,
> like WARNING, ERROR, etc.?


I found that I need one level for development - to trace what
goes rignt and wrong. At the same time as those messages go to
dmesg this level is too verbose to be used under normal
circumstances. That is why another level is introduced. And the
last one exists to verify invariants or error condintions from
which there is no right way to recover and they result in
bpfilter termination.

Probably we may have just two levels - DEBUG and NOTICE and some
analogue of BUG_ON/WARN_ON/runtime assert that results in a
message on NOTICE level and program termination if the checked
condition is false.

I don't think that we will need more levels - until we decide to
utilize syslog facility. Even in that case I don't know how to
differntiate between e.g. NOTICE and INFO messages.

> 
> >
> > Introduce struct context to avoid use of global objects and store there
> > the logging parameters: log level and log sink.
> >
> > Signed-off-by: Dmitrii Banshchikov <me@ubique.spb.ru>
> > ---
> >  net/bpfilter/Makefile  |  2 +-
> >  net/bpfilter/bflog.c   | 29 +++++++++++++++++++++++++++++
> >  net/bpfilter/bflog.h   | 24 ++++++++++++++++++++++++
> >  net/bpfilter/context.h | 16 ++++++++++++++++
> 
> Maybe combine bflog.h and context.h into one file? And bflog() can
> probably fit in
> that file too.


Sure.

> 
> Thanks,
> Song
> 
> >  4 files changed, 70 insertions(+), 1 deletion(-)
> >  create mode 100644 net/bpfilter/bflog.c
> >  create mode 100644 net/bpfilter/bflog.h
> >  create mode 100644 net/bpfilter/context.h
> >
> > diff --git a/net/bpfilter/Makefile b/net/bpfilter/Makefile
> > index cdac82b8c53a..874d5ef6237d 100644
> > --- a/net/bpfilter/Makefile
> > +++ b/net/bpfilter/Makefile
> > @@ -4,7 +4,7 @@
> >  #
> >
> >  userprogs := bpfilter_umh
> > -bpfilter_umh-objs := main.o
> > +bpfilter_umh-objs := main.o bflog.o
> >  userccflags += -I $(srctree)/tools/include/ -I $(srctree)/tools/include/uapi
> >
> >  ifeq ($(CONFIG_BPFILTER_UMH), y)
> > diff --git a/net/bpfilter/bflog.c b/net/bpfilter/bflog.c
> > new file mode 100644
> > index 000000000000..2752e39060e4
> > --- /dev/null
> > +++ b/net/bpfilter/bflog.c
> > @@ -0,0 +1,29 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +/*
> > + * Copyright (c) 2021 Telegram FZ-LLC
> > + */
> > +
> > +#define _GNU_SOURCE
> > +
> > +#include "bflog.h"
> > +
> > +#include <stdarg.h>
> > +#include <stdio.h>
> > +#include <stdlib.h>
> > +
> > +#include "context.h"
> > +
> > +void bflog(struct context *ctx, int level, const char *fmt, ...)
> > +{
> > +       if (ctx->log_file &&
> > +           (level == BFLOG_LEVEL_FATAL || (level & ctx->log_level))) {
> > +               va_list va;
> > +
> > +               va_start(va, fmt);
> > +               vfprintf(ctx->log_file, fmt, va);
> > +               va_end(va);
> > +       }
> > +
> > +       if (level == BFLOG_LEVEL_FATAL)
> > +               exit(EXIT_FAILURE);
> > +}
> > diff --git a/net/bpfilter/bflog.h b/net/bpfilter/bflog.h
> > new file mode 100644
> > index 000000000000..4ed12791cfa1
> > --- /dev/null
> > +++ b/net/bpfilter/bflog.h
> > @@ -0,0 +1,24 @@
> > +/* SPDX-License-Identifier: GPL-2.0 */
> > +/*
> > + * Copyright (c) 2021 Telegram FZ-LLC
> > + */
> > +
> > +#ifndef NET_BPFILTER_BFLOG_H
> > +#define NET_BPFILTER_BFLOG_H
> > +
> > +struct context;
> > +
> > +#define BFLOG_IMPL(ctx, level, fmt, ...) bflog(ctx, level, "bpfilter: " fmt, ##__VA_ARGS__)
> > +
> > +#define BFLOG_LEVEL_FATAL (0)
> > +#define BFLOG_LEVEL_NOTICE (1)
> > +#define BFLOG_LEVEL_DEBUG (2)
> > +
> > +#define BFLOG_FATAL(ctx, fmt, ...)                                                                 \
> > +       BFLOG_IMPL(ctx, BFLOG_LEVEL_FATAL, "fatal error: " fmt, ##__VA_ARGS__)
> > +#define BFLOG_NOTICE(ctx, fmt, ...) BFLOG_IMPL(ctx, BFLOG_LEVEL_NOTICE, fmt, ##__VA_ARGS__)
> > +#define BFLOG_DEBUG(ctx, fmt, ...) BFLOG_IMPL(ctx, BFLOG_LEVEL_DEBUG, fmt, ##__VA_ARGS__)
> > +
> > +void bflog(struct context *ctx, int level, const char *fmt, ...);
> > +
> > +#endif // NET_BPFILTER_BFLOG_H
> > diff --git a/net/bpfilter/context.h b/net/bpfilter/context.h
> > new file mode 100644
> > index 000000000000..e85c97c3d010
> > --- /dev/null
> > +++ b/net/bpfilter/context.h
> > @@ -0,0 +1,16 @@
> > +/* SPDX-License-Identifier: GPL-2.0 */
> > +/*
> > + * Copyright (c) 2021 Telegram FZ-LLC
> > + */
> > +
> > +#ifndef NET_BPFILTER_CONTEXT_H
> > +#define NET_BPFILTER_CONTEXT_H
> > +
> > +#include <stdio.h>
> > +
> > +struct context {
> > +       FILE *log_file;
> > +       int log_level;
> > +};
> > +
> > +#endif // NET_BPFILTER_CONTEXT_H
> > --
> > 2.25.1
> >

-- 

Dmitrii Banshchikov

  reply	other threads:[~2021-05-20  7:08 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-17 22:52 [PATCH bpf-next 00/11] bpfilter Dmitrii Banshchikov
2021-05-17 22:52 ` [PATCH bpf-next 01/11] bpfilter: Add types for usermode helper Dmitrii Banshchikov
2021-05-17 22:52 ` [PATCH bpf-next 02/11] bpfilter: Add logging facility Dmitrii Banshchikov
2021-05-19 17:32   ` Song Liu
2021-05-20  7:08     ` Dmitrii Banshchikov [this message]
2021-05-20 16:35       ` Song Liu
2021-05-21  6:46         ` Dmitrii Banshchikov
2021-05-17 22:53 ` [PATCH bpf-next 03/11] bpfilter: Add IO functions Dmitrii Banshchikov
2021-05-19 18:47   ` Song Liu
2021-05-17 22:53 ` [PATCH bpf-next 04/11] tools: Add bpfilter usermode helper header Dmitrii Banshchikov
2021-05-17 22:53 ` [PATCH bpf-next 05/11] bpfilter: Add map container Dmitrii Banshchikov
2021-05-17 22:53 ` [PATCH bpf-next 06/11] bpfilter: Add struct match Dmitrii Banshchikov
2021-05-20  4:26   ` Song Liu
2021-05-20  7:31     ` Dmitrii Banshchikov
2021-05-20 17:44       ` Song Liu
2021-05-17 22:53 ` [PATCH bpf-next 07/11] bpfilter: Add struct target Dmitrii Banshchikov
2021-05-20  4:36   ` Song Liu
2021-05-20  7:44     ` Dmitrii Banshchikov
2021-05-17 22:53 ` [PATCH bpf-next 08/11] bpfilter: Add struct rule Dmitrii Banshchikov
2021-05-17 22:53 ` [PATCH bpf-next 09/11] bpfilter: Add struct table Dmitrii Banshchikov
2021-05-20 18:07   ` Song Liu
2021-05-17 22:53 ` [PATCH bpf-next 10/11] bpfilter: Add handling of setsockopt() calls Dmitrii Banshchikov
2021-05-17 22:53 ` [PATCH bpf-next 11/11] bpfilter: Handle setsockopts Dmitrii Banshchikov
2021-05-20  4:54 ` [PATCH bpf-next 00/11] bpfilter Song Liu
2021-05-20  7:53   ` Dmitrii Banshchikov
2021-05-20 16:55     ` Alexei Starovoitov
2021-05-20 17:56       ` Song Liu
2021-05-21  6:00         ` Dmitrii Banshchikov

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=20210520070807.cpmloff4urdsifuy@amnesia \
    --to=me@ubique.spb.ru \
    --cc=andrii@kernel.org \
    --cc=ast@kernel.org \
    --cc=bpf@vger.kernel.org \
    --cc=daniel@iogearbox.net \
    --cc=davem@davemloft.net \
    --cc=john.fastabend@gmail.com \
    --cc=kafai@fb.com \
    --cc=kpsingh@kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=rdna@fb.com \
    --cc=song@kernel.org \
    --cc=songliubraving@fb.com \
    --cc=yhs@fb.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.