linux-kernel - Re: linux-next: build warning after merge of the rcu tree

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: <20191212060200.GW2889@paulmck-ThinkPad-P72>
Date:   Wed, 11 Dec 2019 22:02:00 -0800
From:   "Paul E. McKenney" <paulmck@...nel.org>
To:     Stephen Rothwell <sfr@...b.auug.org.au>
Cc:     Linux Next Mailing List <linux-next@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Eric Dumazet <edumazet@...gle.com>
Subject: Re: linux-next: build warning after merge of the rcu tree

On Thu, Dec 12, 2019 at 04:06:22PM +1100, Stephen Rothwell wrote:
> Hi all,
> 
> After merging the rcu (I think) tree, today's linux-next build (x86_64
> allnoconfig) produced this warning:
> 
> kernel/time/timer.c: In function 'schedule_timeout':
> kernel/time/timer.c:969:20: warning: 'timer.expires' may be used uninitialized in this function [-Wmaybe-uninitialized]
>   969 |   long diff = timer->expires - expires;
>       |               ~~~~~^~~~~~~~~
> 
> Introduced by (bisected to) commit
> 
>   c4127fce1d02 ("timer: Use hlist_unhashed_lockless() in timer_pending()")
> 
> x86_64-linux-gnu-gcc (Debian 9.2.1-21) 9.2.1 20191130

Well, if the timer is pending, then ->expires has to have been
initialized, but off where the compiler cannot see it, such as during a
previous call to __mod_timer().  And the change may have made it harder
for the compiler to see all of these relationships, but...

I don't see this warning with gcc version 7.4.0.  Just out of curiosity,
what are you running, Stephen?

Eric, any thoughts for properly educating the compiler on this one?

							Thanx, Paul