Message-ID: <f986f5a9-5c97-10ed-1e44-84bbd929e605@linux.intel.com>
Date:   Wed, 5 Aug 2020 11:57:20 +0800
From:   "Li, Aubrey" <aubrey.li@...ux.intel.com>
To:     Joel Fernandes <joel@...lfernandes.org>
Cc:     viremana@...ux.microsoft.com,
        Nishanth Aravamudan <naravamudan@...italocean.com>,
        Julien Desfossez <jdesfossez@...italocean.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Tim Chen <tim.c.chen@...ux.intel.com>,
        Ingo Molnar <mingo@...nel.org>,
        Thomas Glexiner <tglx@...utronix.de>,
        Paul Turner <pjt@...gle.com>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        LKML <linux-kernel@...r.kernel.org>,
        Subhra Mazumdar <subhra.mazumdar@...cle.com>,
        Frederic Weisbecker <fweisbec@...il.com>,
        Kees Cook <keescook@...omium.org>,
        Greg Kerr <kerrnel@...gle.com>, Phil Auld <pauld@...hat.com>,
        Aaron Lu <aaron.lwe@...il.com>,
        Aubrey Li <aubrey.intel@...il.com>,
        Valentin Schneider <valentin.schneider@....com>,
        Mel Gorman <mgorman@...hsingularity.net>,
        Pawan Gupta <pawan.kumar.gupta@...ux.intel.com>,
        Paolo Bonzini <pbonzini@...hat.com>,
        Joel Fernandes <joelaf@...gle.com>,
        Vineeth Pillai <vineethrp@...il.com>,
        Chen Yu <yu.c.chen@...el.com>,
        Christian Brauner <christian.brauner@...ntu.com>,
        "Ning, Hongyu" <hongyu.ning@...ux.intel.com>,
        benbjiang(蒋彪) <benbjiang@...cent.com>
Subject: Re: [RFC PATCH 00/16] Core scheduling v6

On 2020/8/4 0:53, Joel Fernandes wrote:
> Hi Aubrey,
> 
> On Mon, Aug 3, 2020 at 4:23 AM Li, Aubrey <aubrey.li@...ux.intel.com> wrote:
>>
>> On 2020/7/1 5:32, Vineeth Remanan Pillai wrote:
>>> Sixth iteration of the Core-Scheduling feature.
>>>
>>> Core scheduling is a feature that allows only trusted tasks to run
>>> concurrently on cpus sharing compute resources (eg: hyperthreads on a
>>> core). The goal is to mitigate the core-level side-channel attacks
>>> without requiring to disable SMT (which has a significant impact on
>>> performance in some situations). Core scheduling (as of v6) mitigates
>>> user-space to user-space attacks and user to kernel attack when one of
>>> the siblings enters the kernel via interrupts. It is still possible to
>>> have a task attack the sibling thread when it enters the kernel via
>>> syscalls.
>>>
>>> By default, the feature doesn't change any of the current scheduler
>>> behavior. The user decides which tasks can run simultaneously on the
>>> same core (for now by having them in the same tagged cgroup). When a
>>> tag is enabled in a cgroup and a task from that cgroup is running on a
>>> hardware thread, the scheduler ensures that only idle or trusted tasks
>>> run on the other sibling(s). Besides security concerns, this feature
>>> can also be beneficial for RT and performance applications where we
>>> want to control how tasks make use of SMT dynamically.
>>>
>>> This iteration is mostly a cleanup of v5 except for a major feature of
>>> pausing sibling when a cpu enters kernel via nmi/irq/softirq. Also
>>> introducing documentation and includes minor crash fixes.
>>>
>>> One major cleanup was removing the hotplug support and related code.
>>> The hotplug related crashes were not documented and the fixes piled up
>>> over time leading to complex code. We were not able to reproduce the
>>> crashes in the limited testing done. But if they are reroducable, we
>>> don't want to hide them. We should document them and design better
>>> fixes if any.
>>>
>>> In terms of performance, the results in this release are similar to
>>> v5. On a x86 system with N hardware threads:
>>> - if only N/2 hardware threads are busy, the performance is similar
>>>   between baseline, corescheduling and nosmt
>>> - if N hardware threads are busy with N different corescheduling
>>>   groups, the impact of corescheduling is similar to nosmt
>>> - if N hardware threads are busy and multiple active threads share the
>>>   same corescheduling cookie, they gain a performance improvement over
>>>   nosmt.
>>>   The specific performance impact depends on the workload, but for a
>>>   really busy database 12-vcpu VM (1 coresched tag) running on a 36
>>>   hardware threads NUMA node with 96 mostly idle neighbor VMs (each in
>>>   their own coresched tag), the performance drops by 54% with
>>>   corescheduling and drops by 90% with nosmt.
>>>
>>
>> We found uperf(in cgroup) throughput drops by ~50% with corescheduling.
>>
>> The problem is, uperf triggered a lot of softirq and offloaded softirq
>> service to *ksoftirqd* thread.
>>
>> - default, ksoftirqd thread can run with uperf on the same core, we saw
>>   100% CPU utilization.
>> - coresched enabled, ksoftirqd's core cookie is different from uperf, so
>>   they can't run concurrently on the same core, we saw ~15% forced idle.
>>
>> I guess this kind of performance drop can be replicated by other similar
>> (a lot of softirq activities) workloads.
>>
>> Currently core scheduler picks cookie-match tasks for all SMT siblings, does
>> it make sense we add a policy to allow cookie-compatible task running together?
>> For example, if a task is trusted(set by admin), it can work with kernel thread.
>> The difference from corescheduling disabled is that we still have user to user
>> isolation.
> 
> In ChromeOS we are considering all cookie-0 tasks as trusted.
> Basically if you don't trust a task, then that is when you assign the
> task a tag. We do this for the sandboxed processes.

I have a proposal of this, by changing cpu.tag to cpu.coresched_policy,
something like the following:

+/*
+ * Core scheduling policy:
+ * - CORE_SCHED_DISABLED: core scheduling is disabled.
+ * - CORE_COOKIE_MATCH: tasks with same cookie can run
+ *                     on the same core concurrently.
+ * - CORE_COOKIE_TRUST: trusted task can run with kernel
			thread on the same core concurrently. 
+ * - CORE_COOKIE_LONELY: tasks with cookie can run only
+ *                     with idle thread on the same core.
+ */
+enum coresched_policy {
+       CORE_SCHED_DISABLED,
+       CORE_SCHED_COOKIE_MATCH,
+	CORE_SCHED_COOKIE_TRUST,
+       CORE_SCHED_COOKIE_LONELY,
+};

We can set policy to CORE_COOKIE_TRUST of uperf cgroup and fix this kind
of performance regression. Not sure if this sounds attractive?

> 
> Is the uperf throughput worse with SMT+core-scheduling versus no-SMT ?

This is a good question, from the data we measured by uperf,
SMT+core-scheduling is 28.2% worse than no-SMT, :(

Thanks,
-Aubrey

> 
> thanks,
> 
>  - Joel
> PS: I am planning to write a patch behind a CONFIG option that tags
> all processes (default untrusted) so everything gets a cookie which
> some folks said was how they wanted (have a whitelist instead of
> blacklist).
> 

Powered by blists - more mailing lists