linux-kernel - Re: [PATCH] locking/osq: Drop the overload of osq lock

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: <20160625142447.GK30154@twins.programming.kicks-ass.net>
Date:	Sat, 25 Jun 2016 16:24:47 +0200
From:	Peter Zijlstra <peterz@...radead.org>
To:	Pan Xinhui <xinhui.pan@...ux.vnet.ibm.com>
Cc:	linux-kernel@...r.kernel.org, mingo@...hat.com, dave@...olabs.net,
	will.deacon@....com, Waiman.Long@....com, benh@...nel.crashing.org
Subject: Re: [PATCH] locking/osq: Drop the overload of osq lock

On Sat, Jun 25, 2016 at 01:42:03PM -0400, Pan Xinhui wrote:
> An over-committed guest with more vCPUs than pCPUs has a heavy overload
> in osq_lock().
> 
> This is because vCPU A hold the osq lock and yield out, vCPU B wait
> per_cpu node->locked to be set. IOW, vCPU B wait vCPU A to run and
> unlock the osq lock. Even there is need_resched(), it did not help on
> such scenario.
> 
> To fix such bad issue, add a threshold in one while-loop of osq_lock().
> The value of threshold is somehow equal to SPIN_THRESHOLD.

Blergh, virt ...

So yes, lock holder preemption sucks. You would also want to limit the
immediate spin on owner.

Also; I really hate these random number spin-loop thresholds.

Is it at all possible to get feedback from your LPAR stuff that the vcpu
was preempted? Because at that point we can add do something like:


	int vpc = vcpu_preempt_count();

	...

	for (;;) {

		/* the big spin loop */

		if (need_resched() || vpc != vcpu_preempt_count())
			/* bail */

	}


With a default implementation like:

static inline int vcpu_preempt_count(void)
{
	return 0;
}

So the compiler can make it all go away.


But on virt muck it would stop spinning the moment the vcpu gets
preempted, which is the right moment I'm thinking.