All of lore.kernel.org
 help / color / mirror / Atom feed
From: Greg KH <gregkh@linuxfoundation.org>
To: Krzysztof Opasiak <k.opasiak@samsung.com>
Cc: viro@zeniv.linux.org.uk, arnd@arndb.de,
	linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org,
	linux-arch@vger.kernel.org, k.lewandowsk@samsung.com,
	l.stelmach@samsung.com, p.szewczyk@samsung.com,
	b.zolnierkie@samsung.com, andrzej.p@samsung.com,
	kopasiak90@gmail.com
Subject: Re: [PATCH 2/4][PoC][RFC] Add rlimit-events framework
Date: Thu, 19 Oct 2017 09:41:18 +0200	[thread overview]
Message-ID: <20171019074118.GD20787@kroah.com> (raw)
In-Reply-To: <20171018203230.29871-3-k.opasiak@samsung.com>

Meta-comments on the code, I'm not commenting on the content, just
normal code review things that I always see in kernel code...

On Wed, Oct 18, 2017 at 10:32:28PM +0200, Krzysztof Opasiak wrote:
> diff --git a/include/linux/rlimit_noti_kern.h b/include/linux/rlimit_noti_kern.h
> new file mode 100644
> index 000000000000..e49fddaa21c0
> --- /dev/null
> +++ b/include/linux/rlimit_noti_kern.h
> @@ -0,0 +1,54 @@
> +/*
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License as published by
> + * the Free Software Foundation; either version 2 of the License, or
> + * (at your option) any later version.

I have to ask, do you really mean "any later version" for this, and the
other new files you created?

And, it is nice to use SPDX for new files to identify their license.
It's not that prevelant, but is getting there...

> --- a/include/uapi/linux/netlink.h
> +++ b/include/uapi/linux/netlink.h
> @@ -28,6 +28,7 @@
>  #define NETLINK_RDMA		20
>  #define NETLINK_CRYPTO		21	/* Crypto layer */
>  #define NETLINK_SMC		22	/* SMC monitoring */
> +#define NETLINK_RLIMIT_EVENTS   23      /* rlimit notification */

No tabs?

> --- /dev/null
> +++ b/include/uapi/linux/rlimit_noti.h
> @@ -0,0 +1,71 @@
> +/*
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License as published by
> + * the Free Software Foundation; either version 2 of the License, or
> + * (at your option) any later version.

GPLv2+ in a user api header file?  You are really brave :)

> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + */
> +
> +#ifndef _UAPI_LINUX_RLIMIT_NOTI_H_
> +#define _UAPI_LINUX_RLIMIT_NOTI_H_
> +
> +#ifdef __KERNEL__
> +#include <linux/types.h>
> +#include <linux/resource.h>
> +#else
> +#include <stdint.h>
> +#endif
> +
> +#define RLIMIT_GET_NOTI_FD 1000
> +
> +/* ioctl's */
> +#define RLIMIT_ADD_NOTI_LVL 1
> +#define RLIMIT_RM_NOTI_LVL 2
> +
> +#define RLIMIT_SET_NOTI_ALL 3
> +#define RLIMIT_CLEAR_NOTI_ALL 4

No tabs?

> +
> +/*
> + * For future (notify every 5, 10 units change):
> + * #define RLIMIT_SET_NOTI_STEP 5
> + */
> +
> +#define RLIMIT_GET_NOTI_LVLS 6
> +#define RLIMIT_GET_NOTI_LVL_COUNT 7
> +
> +/* Flags for ioctl's */
> +#define RLIMIT_FLAG_NO_INHERIT (1u << 0)
> +
> +/* Event types */
> +enum {
> +	RLIMIT_EVENT_TYPE_RES_CHANGED,
> +	RLIMIT_EVENT_TYPE_MAX
> +};
> +
> +/* TODO take care of padding (packed) */
> +struct rlimit_noti_subject {
> +	pid_t pid;
> +	uint32_t resource;
> +};

For structures that cross the user/kernel boundry, you have to use the
correct variable types.  And that is never "unit32_t" and such, use
"__u32" and the other "__" types.

And are you _sure_ that pid_t is able to be exported to userspace
correctly?

> +
> +struct rlimit_noti_level {
> +	struct rlimit_noti_subject subj;
> +	uint64_t value;

__u64

> +	uint32_t flags;

__u32

And so on for all others.

You don't seem to describe an ioctl here in the "normal" method, but
only use vague numbers up above, that's odd, why?

> diff --git a/init/Kconfig b/init/Kconfig
> index 1d3475fc9496..4bc44fa86640 100644
> --- a/init/Kconfig
> +++ b/init/Kconfig
> @@ -332,6 +332,12 @@ config AUDIT_TREE
>  	depends on AUDITSYSCALL
>  	select FSNOTIFY
>  
> +config RLIMIT_NOTIFICATION
> +       bool "Support fd notifications on given resource usage"
> +       depends on NET
> +       help
> +	Enable this to monitor process resource changes usage via fd.

Mix of tab and spaces :(

thanks,

greg k-h

  reply	other threads:[~2017-10-19  7:41 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <CGME20171018203324epcas2p310def24eea6171d0821f1068d055d3b7@epcas2p3.samsung.com>
2017-10-18 20:32 ` [PATCH 0/4][PoC][RFC] Add rlimit-resources change notification mechanism Krzysztof Opasiak
     [not found]   ` <CGME20171018203328epcas2p4bdcc3650d1c0f0add143488792243932@epcas2p4.samsung.com>
2017-10-18 20:32     ` [PATCH 1/4][PoC][RFC] sched: Allow to get() and put() signal struct Krzysztof Opasiak
2017-10-19  7:34       ` Greg KH
2017-10-19 17:33         ` Krzysztof Opasiak
     [not found]   ` <CGME20171018203333epcas1p3d8dd8f3a755cb6ee8e9dde63cb91851b@epcas1p3.samsung.com>
2017-10-18 20:32     ` [PATCH 2/4][PoC][RFC] Add rlimit-events framework Krzysztof Opasiak
2017-10-19  7:41       ` Greg KH [this message]
2017-10-19 18:17         ` Krzysztof Opasiak
     [not found]   ` <CGME20171018203337epcas2p22bc3a5ac063486c9c7906fe6e88277bb@epcas2p2.samsung.com>
2017-10-18 20:32     ` [PATCH 3/4][PoC][RFC] Connect rlimit-events with process life cycle Krzysztof Opasiak
2017-10-19  7:41       ` Greg KH
2017-10-19 18:19         ` Krzysztof Opasiak
     [not found]   ` <CGME20171018203342epcas1p23933a20a33807f92c716433374374397@epcas1p2.samsung.com>
2017-10-18 20:32     ` [PATCH 4/4][PoC][RFC] Allow to trace fd usage with rlimit-events Krzysztof Opasiak
2017-10-18 23:05       ` Al Viro
2017-10-19 17:28         ` Krzysztof Opasiak

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=20171019074118.GD20787@kroah.com \
    --to=gregkh@linuxfoundation.org \
    --cc=andrzej.p@samsung.com \
    --cc=arnd@arndb.de \
    --cc=b.zolnierkie@samsung.com \
    --cc=k.lewandowsk@samsung.com \
    --cc=k.opasiak@samsung.com \
    --cc=kopasiak90@gmail.com \
    --cc=l.stelmach@samsung.com \
    --cc=linux-arch@vger.kernel.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=p.szewczyk@samsung.com \
    --cc=viro@zeniv.linux.org.uk \
    /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.