linux-kernel - Re: [PATCH] signal: Avoid undefined behaviour in kill_something

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: <20170605123744.GA9807@redhat.com>
Date:   Mon, 5 Jun 2017 14:37:45 +0200
From:   Oleg Nesterov <oleg@...hat.com>
To:     zhongjiang <zhongjiang@...wei.com>
Cc:     akpm@...ux-foundation.org, stsp@...t.ru, Waiman.Long@....com,
        mingo@...nel.org, mhocko@...nel.org, vbabka@...e.cz,
        linux-mm@...ck.org, linux-kernel@...r.kernel.org,
        qiuxishi@...wei.com
Subject: Re: [PATCH] signal: Avoid undefined behaviour in kill_something_info

On 06/05, zhongjiang wrote:
>
>  static int kill_something_info(int sig, struct siginfo *info, pid_t pid)
>  {
> -	int ret;
> +	int ret, vpid;
>  
>  	if (pid > 0) {
>  		rcu_read_lock();
> @@ -1395,8 +1395,12 @@ static int kill_something_info(int sig, struct siginfo *info, pid_t pid)
>  
>  	read_lock(&tasklist_lock);
>  	if (pid != -1) {
> +		if (pid == INT_MIN)
> +			vpid = INT_MAX;

Well, this probably needs a comment to explain that this is just "avoid ub".

And if we really want the fix, to me

	if (pid == INT_MIN)
		return -ESRCH;

at the start makes more sense...

Oleg.