linux-kernel - Re: WARNING in hrtimer

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-next>] [day] [month] [year] [list]

Message-ID: <87mu19kaup.fsf@nanos.tec.linutronix.de>
Date:   Mon, 28 Sep 2020 20:35:58 +0200
From:   Thomas Gleixner <tglx@...utronix.de>
To:     Hillf Danton <hdanton@...a.com>
Cc:     syzbot <syzbot+ca740b95a16399ceb9a5@...kaller.appspotmail.com>,
        davem@...emloft.net, hchunhui@...l.ustc.edu.cn, hdanton@...a.com,
        ja@....bg, jmorris@...ei.org, kuznet@....inr.ac.ru,
        linux-kernel@...r.kernel.org, netdev@...r.kernel.org,
        syzkaller-bugs@...glegroups.com, yoshfuji@...ux-ipv6.org,
        Johannes Berg <johannes.berg@...el.com>
Subject: Re: WARNING in hrtimer_forward

On Tue, Sep 29 2020 at 01:11, Hillf Danton wrote:
> On Mon, 28 Sep 2020 18:13:42 +0200 Thomas Gleixner wrote:
>> So the timer was armed at some point and then the expiry which does the
>> forward races with the ioctl which starts the timer. Lack of
>> serialization or such ...
>
> To make syzbot happy, s/hrtimer_is_queued/hrtimer_active/ can close
> that race but this warning looks benign.

Why only make sysbot happy? It's clearly an issue and the warning is not
benign simply because forwarding a queued timer is an absolute NONO.
timers (both timer_list and hrtimer) need external synchronization.

> --- a/drivers/net/wireless/mac80211_hwsim.c
> +++ b/drivers/net/wireless/mac80211_hwsim.c
> @@ -1698,7 +1698,7 @@ static int mac80211_hwsim_config(struct
>  
>  	if (!data->started || !data->beacon_int)
>  		hrtimer_cancel(&data->beacon_timer);
> -	else if (!hrtimer_is_queued(&data->beacon_timer)) {
> +	else if (!hrtimer_active(&data->beacon_timer)) {
>  		u64 tsf = mac80211_hwsim_get_tsf(hw, NULL);
>  		u32 bcn_int = data->beacon_int;
>  		u64 until_tbtt = bcn_int - do_div(tsf, bcn_int);
> @@ -1768,7 +1768,7 @@ static void mac80211_hwsim_bss_info_chan
>  			  info->enable_beacon, info->beacon_int);
>  		vp->bcn_en = info->enable_beacon;
>  		if (data->started &&
> -		    !hrtimer_is_queued(&data->beacon_timer) &&
> +		    !hrtimer_active(&data->beacon_timer) &&
>  		    info->enable_beacon) {
>  			u64 tsf, until_tbtt;
>  			u32 bcn_int;

Looks about right.

Thanks,

        tglx