From: "Li, Aubrey" <aubrey.li@linux.intel.com>
To: Joel Fernandes <joel@joelfernandes.org>
Cc: viremana@linux.microsoft.com,
"Nishanth Aravamudan" <naravamudan@digitalocean.com>,
"Julien Desfossez" <jdesfossez@digitalocean.com>,
"Peter Zijlstra" <peterz@infradead.org>,
"Tim Chen" <tim.c.chen@linux.intel.com>,
"Ingo Molnar" <mingo@kernel.org>,
"Thomas Glexiner" <tglx@linutronix.de>,
"Paul Turner" <pjt@google.com>,
"Linus Torvalds" <torvalds@linux-foundation.org>,
LKML <linux-kernel@vger.kernel.org>,
"Subhra Mazumdar" <subhra.mazumdar@oracle.com>,
"Frederic Weisbecker" <fweisbec@gmail.com>,
"Kees Cook" <keescook@chromium.org>,
"Greg Kerr" <kerrnel@google.com>, "Phil Auld" <pauld@redhat.com>,
"Aaron Lu" <aaron.lwe@gmail.com>,
"Aubrey Li" <aubrey.intel@gmail.com>,
"Valentin Schneider" <valentin.schneider@arm.com>,
"Mel Gorman" <mgorman@techsingularity.net>,
"Pawan Gupta" <pawan.kumar.gupta@linux.intel.com>,
"Paolo Bonzini" <pbonzini@redhat.com>,
"Joel Fernandes" <joelaf@google.com>,
"Vineeth Pillai" <vineethrp@gmail.com>,
"Chen Yu" <yu.c.chen@intel.com>,
"Christian Brauner" <christian.brauner@ubuntu.com>,
"Ning, Hongyu" <hongyu.ning@linux.intel.com>,
"benbjiang(蒋彪)" <benbjiang@tencent.com>
Subject: Re: [RFC PATCH 00/16] Core scheduling v6
Date: Wed, 5 Aug 2020 11:57:20 +0800 [thread overview]
Message-ID: <f986f5a9-5c97-10ed-1e44-84bbd929e605@linux.intel.com> (raw)
In-Reply-To: <CAEXW_YS6oW_AAkmOuXNMCj_N5763aG9SXEcWz_onPhQQU2TZ0g@mail.gmail.com>
On 2020/8/4 0:53, Joel Fernandes wrote:
> Hi Aubrey,
>
> On Mon, Aug 3, 2020 at 4:23 AM Li, Aubrey <aubrey.li@linux.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).
>
next prev parent reply other threads:[~2020-08-05 3:57 UTC|newest]
Thread overview: 81+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-06-30 21:32 [RFC PATCH 00/16] Core scheduling v6 Vineeth Remanan Pillai
2020-06-30 21:32 ` [RFC PATCH 01/16] sched: Wrap rq::lock access Vineeth Remanan Pillai
2020-06-30 21:32 ` [RFC PATCH 02/16] sched: Introduce sched_class::pick_task() Vineeth Remanan Pillai
2020-06-30 21:32 ` [RFC PATCH 03/16] sched: Core-wide rq->lock Vineeth Remanan Pillai
2020-06-30 21:32 ` [RFC PATCH 04/16] sched/fair: Add a few assertions Vineeth Remanan Pillai
2020-06-30 21:32 ` [RFC PATCH 05/16] sched: Basic tracking of matching tasks Vineeth Remanan Pillai
2020-07-21 14:02 ` [RFC PATCH 05/16] sched: Basic tracking of matching tasks(Internet mail) benbjiang(蒋彪)
2020-06-30 21:32 ` [RFC PATCH 06/16] sched: Add core wide task selection and scheduling Vineeth Remanan Pillai
2020-07-01 23:28 ` Joel Fernandes
2020-07-02 0:54 ` Tim Chen
2020-07-02 12:57 ` Joel Fernandes
2020-07-02 13:23 ` Joel Fernandes
2020-07-05 23:44 ` Tim Chen
2020-07-03 20:21 ` Vineeth Remanan Pillai
2020-07-06 14:09 ` Joel Fernandes
2020-07-06 14:38 ` Vineeth Remanan Pillai
2020-07-06 17:37 ` Joel Fernandes
2020-06-30 21:32 ` [RFC PATCH 07/16] sched/fair: Fix forced idle sibling starvation corner case Vineeth Remanan Pillai
2020-07-21 7:35 ` [RFC PATCH 07/16] sched/fair: Fix forced idle sibling starvation corner case(Internet mail) benbjiang(蒋彪)
2020-07-22 7:20 ` benbjiang(蒋彪)
2020-06-30 21:32 ` [RFC PATCH 08/16] sched/fair: wrapper for cfs_rq->min_vruntime Vineeth Remanan Pillai
2020-06-30 21:32 ` [RFC PATCH 09/16] sched/fair: core wide cfs task priority comparison Vineeth Remanan Pillai
2020-07-22 0:23 ` [RFC PATCH 09/16] sched/fair: core wide cfs task priority comparison(Internet mail) benbjiang(蒋彪)
2020-07-24 7:14 ` Aaron Lu
2020-07-24 12:08 ` Jiang Biao
2020-06-30 21:32 ` [RFC PATCH 10/16] sched: Trivial forced-newidle balancer Vineeth Remanan Pillai
2020-07-20 4:06 ` [RFC PATCH 10/16] sched: Trivial forced-newidle balancer(Internet mail) benbjiang(蒋彪)
2020-07-20 6:06 ` Li, Aubrey
[not found] ` <8082F052-2F52-42D3-B396-18A35A94F26F@tencent.com>
2020-07-20 8:03 ` Li, Aubrey
2020-07-20 8:22 ` benbjiang(蒋彪)
2020-07-20 14:34 ` benbjiang(蒋彪)
2020-06-30 21:32 ` [RFC PATCH 11/16] sched: migration changes for core scheduling Vineeth Remanan Pillai
2020-07-22 8:54 ` [RFC PATCH 11/16] sched: migration changes for core scheduling(Internet mail) benbjiang(蒋彪)
2020-07-22 12:13 ` Li, Aubrey
2020-07-22 14:32 ` benbjiang(蒋彪)
2020-07-23 1:57 ` Li, Aubrey
2020-07-23 2:42 ` benbjiang(蒋彪)
2020-07-23 3:35 ` Li, Aubrey
2020-07-23 4:23 ` benbjiang(蒋彪)
2020-07-23 5:39 ` Li, Aubrey
2020-07-23 7:47 ` benbjiang(蒋彪)
2020-07-23 8:06 ` Li, Aubrey
2020-07-23 8:28 ` benbjiang(蒋彪)
2020-07-23 23:43 ` Aubrey Li
2020-07-24 1:26 ` benbjiang(蒋彪)
2020-07-24 2:05 ` Li, Aubrey
2020-07-24 2:29 ` benbjiang(蒋彪)
2020-06-30 21:32 ` [RFC PATCH 12/16] sched: cgroup tagging interface for core scheduling Vineeth Remanan Pillai
2020-06-30 21:32 ` [RFC PATCH 13/16] sched: Fix pick_next_task() race condition in " Vineeth Remanan Pillai
2020-06-30 21:32 ` [RFC PATCH 14/16] irq: Add support for core-wide protection of IRQ and softirq Vineeth Remanan Pillai
2020-07-10 12:19 ` Li, Aubrey
2020-07-10 13:21 ` Joel Fernandes
2020-07-13 2:23 ` Li, Aubrey
2020-07-13 15:58 ` Joel Fernandes
2020-07-10 13:36 ` Vineeth Remanan Pillai
2020-07-11 1:33 ` Aubrey Li
2020-07-17 23:37 ` Thomas Gleixner
2020-07-18 17:05 ` Joel Fernandes
2020-07-17 23:36 ` Thomas Gleixner
2020-07-20 3:53 ` Joel Fernandes
2020-07-20 8:20 ` Thomas Gleixner
2020-07-20 11:09 ` Vineeth Pillai
2020-06-30 21:32 ` [RFC PATCH 15/16] Documentation: Add documentation on core scheduling Vineeth Remanan Pillai
2020-06-30 21:32 ` [RFC PATCH 16/16] sched: Debug bits Vineeth Remanan Pillai
2020-07-31 16:41 ` [RFC PATCH 00/16] Core scheduling v6 Vineeth Pillai
2020-08-03 8:23 ` Li, Aubrey
2020-08-03 16:53 ` Joel Fernandes
2020-08-05 3:57 ` Li, Aubrey [this message]
2020-08-05 6:16 ` [RFC PATCH 00/16] Core scheduling v6(Internet mail) benbjiang(蒋彪)
2020-08-09 16:44 ` [RFC PATCH 00/16] Core scheduling v6 Joel Fernandes
2020-08-12 2:01 ` Li, Aubrey
2020-08-12 23:08 ` Joel Fernandes
2020-08-13 4:28 ` Li, Aubrey
2020-08-14 0:26 ` [RFC PATCH 00/16] Core scheduling v6(Internet mail) benbjiang(蒋彪)
2020-08-14 1:36 ` Li, Aubrey
2020-08-14 4:04 ` benbjiang(蒋彪)
2020-08-14 5:18 ` Li, Aubrey
2020-08-14 7:54 ` benbjiang(蒋彪)
2020-08-20 22:37 ` [RFC PATCH 00/16] Core scheduling v6 Joel Fernandes
2020-08-27 0:30 ` Alexander Graf
2020-08-27 1:20 ` Vineeth Pillai
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=f986f5a9-5c97-10ed-1e44-84bbd929e605@linux.intel.com \
--to=aubrey.li@linux.intel.com \
--cc=aaron.lwe@gmail.com \
--cc=aubrey.intel@gmail.com \
--cc=benbjiang@tencent.com \
--cc=christian.brauner@ubuntu.com \
--cc=fweisbec@gmail.com \
--cc=hongyu.ning@linux.intel.com \
--cc=jdesfossez@digitalocean.com \
--cc=joel@joelfernandes.org \
--cc=joelaf@google.com \
--cc=keescook@chromium.org \
--cc=kerrnel@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mgorman@techsingularity.net \
--cc=mingo@kernel.org \
--cc=naravamudan@digitalocean.com \
--cc=pauld@redhat.com \
--cc=pawan.kumar.gupta@linux.intel.com \
--cc=pbonzini@redhat.com \
--cc=peterz@infradead.org \
--cc=pjt@google.com \
--cc=subhra.mazumdar@oracle.com \
--cc=tglx@linutronix.de \
--cc=tim.c.chen@linux.intel.com \
--cc=torvalds@linux-foundation.org \
--cc=valentin.schneider@arm.com \
--cc=vineethrp@gmail.com \
--cc=viremana@linux.microsoft.com \
--cc=yu.c.chen@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).