linux-kernel - Re: [patch 14/45] posix-timers: Consolidate interval retrieval

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]

Date:   Fri, 30 Jun 2023 16:04:04 +0200
From:   Frederic Weisbecker <frederic@...nel.org>
To:     Thomas Gleixner <tglx@...utronix.de>
Cc:     LKML <linux-kernel@...r.kernel.org>,
        Anna-Maria Behnsen <anna-maria@...utronix.de>,
        John Stultz <jstultz@...gle.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Ingo Molnar <mingo@...nel.org>,
        Stephen Boyd <sboyd@...nel.org>,
        Eric Biederman <ebiederm@...ssion.com>,
        Oleg Nesterov <oleg@...hat.com>
Subject: Re: [patch 14/45] posix-timers: Consolidate interval retrieval

On Fri, Jun 30, 2023 at 03:07:17PM +0200, Thomas Gleixner wrote:
> On Fri, Jun 30 2023 at 13:25, Frederic Weisbecker wrote:
> > On Thu, Jun 29, 2023 at 08:47:30PM +0200, Thomas Gleixner wrote:
> >> On Wed, Jun 28 2023 at 15:08, Frederic Weisbecker wrote:
> >> 
> >> > Le Tue, Jun 06, 2023 at 04:37:40PM +0200, Thomas Gleixner a écrit :
> >> >> There is no point to collect the current interval in the posix clock
> >> >> specific settime() and gettime() callbacks. Just do it right in the common
> >> >> code.
> >> >> 
> >> >> Signed-off-by: Thomas Gleixner <tglx@...utronix.de>
> >> >
> >> > The only difference I see is that we now return the old interval
> >> > even if the target is reaped, which probably doesn't matter anyway.
> >> 
> >> But we don't return it to user space because ret != 0 in that case.
> >
> > In the case of ->set yes but in the case of ->get there is no error
> > handling.
> 
> SYSCALL_DEFINE2(timer_gettime, timer_t, timer_id,
> 		struct __kernel_itimerspec __user *, setting)
> {
> 	struct itimerspec64 cur_setting;
> 
> 	int ret = do_timer_gettime(timer_id, &cur_setting);
> 	if (!ret) {
> 		if (put_itimerspec64(&cur_setting, setting))
> 
> How exactly does this end up being copied to user space if ret != 0?

kc->timer_get() doesn't return any value.

So before the patch, interval is retrieved only if the target is not reaped:

timer_gettime() {
    do_timer_gettime() {
        posix_cpu_timer_get() {
            p = cpu_timer_task_rcu(timer);
            if (p)
                itp->interval = ....
        }
    }
}

After the patch it's retrieved unconditionally:

timer_gettime() {
    do_timer_gettime() {
        //unconditionally set
        itp->interval = ....
        posix_cpu_timer_get() {
            p = cpu_timer_task_rcu(timer);
            if (!p)
                //doesn't return any value so no failure reported
        }
    }
}