linux-kernel - Re: [PATCH 7/8] pipe: make account_pipe_buffers() return a value, and use it

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <57B6D303.2060301@oracle.com>
Date:   Fri, 19 Aug 2016 11:36:03 +0200
From:   Vegard Nossum <vegard.nossum@...cle.com>
To:     "Michael Kerrisk (man-pages)" <mtk.manpages@...il.com>,
        Andrew Morton <akpm@...ux-foundation.org>
Cc:     Willy Tarreau <w@....eu>, socketpair@...il.com,
        Tetsuo Handa <penguin-kernel@...ove.SAKURA.ne.jp>,
        Jens Axboe <axboe@...com>, Al Viro <viro@...iv.linux.org.uk>,
        linux-api@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 7/8] pipe: make account_pipe_buffers() return a value, and
 use it

On 08/19/2016 07:25 AM, Michael Kerrisk (man-pages) wrote:
> This is an optional patch, to provide a small performance improvement.
> Alter account_pipe_buffers() so that it returns the new value in
> user->pipe_bufs. This means that we can refactor too_many_pipe_buffers_soft()
> and too_many_pipe_buffers_hard() to avoid the costs of repeated use of
> atomic_long_read() to get the value user->pipe_bufs.
[...]
> @@ -627,17 +625,18 @@ struct pipe_inode_info *alloc_pipe_info(void)
>   	struct pipe_inode_info *pipe;
>   	unsigned long pipe_bufs = PIPE_DEF_BUFFERS;
>   	struct user_struct *user = get_current_user();
> +	unsigned long num_bufs;

Maybe user_bufs would be more descriptive since num_bufs is a bit
ambiguous without the context.

>
>   	pipe = kzalloc(sizeof(struct pipe_inode_info), GFP_KERNEL_ACCOUNT);
>   	if (pipe == NULL)
>   		goto out_free_uid;
>
> -	if (too_many_pipe_buffers_soft(user))
> +	if (too_many_pipe_buffers_soft(atomic_long_read(&user->pipe_bufs)))
>   		pipe_bufs = 1;
>
> -	account_pipe_buffers(user, 0, pipe_bufs);
> +	num_bufs = account_pipe_buffers(user, 0, pipe_bufs);
>
> -	if (too_many_pipe_buffers_hard(user))
> +	if (too_many_pipe_buffers_hard(num_bufs))
>   		goto out_revert_acct;
>
>   	pipe->bufs = kcalloc(pipe_bufs, sizeof(struct pipe_buffer),

Why not structure it like this?

num_bufs = account_pipe_buffers(user, 0, pipe_bufs);
if (too_many_pipe_buffers_soft(num_bufs)) {
	num_bufs = account_pipe_buffers(user, pipe_bufs, 1);
	pipe_bufs = 1;
}
if (too_many_pipe_buffers_hard(num_bufs))
	goto out_revert_acct;

Otherwise you still have the case that somebody makes it past
too_many_pipe_buffers_soft() before the accounting is done.


Vegard