From: Eric Biggers <ebiggers@kernel.org>
To: Guenter Roeck <linux@roeck-us.net>
Cc: Peter Zijlstra <peterz@infradead.org>,
tglx@linutronix.de, frederic@kernel.org,
linux-kernel@vger.kernel.org, x86@kernel.org, cai@lca.pw,
mgorman@techsingularity.net
Subject: Re: [RFC][PATCH 7/7] sched: Replace rq::wake_list
Date: Tue, 9 Jun 2020 13:21:34 -0700 [thread overview]
Message-ID: <20200609202134.GA1105@sol.localdomain> (raw)
In-Reply-To: <53318971-561c-b445-0408-530b3d3ba44e@roeck-us.net>
On Sat, Jun 06, 2020 at 04:13:33PM -0700, Guenter Roeck wrote:
> On 6/5/20 9:15 AM, Eric Biggers wrote:
> > On Fri, Jun 05, 2020 at 09:41:54AM +0200, Peter Zijlstra wrote:
> >> On Thu, Jun 04, 2020 at 05:24:33PM -0700, Eric Biggers wrote:
> >>> On Thu, Jun 04, 2020 at 07:18:37AM -0700, Guenter Roeck wrote:
> >>>> On Tue, May 26, 2020 at 06:11:04PM +0200, Peter Zijlstra wrote:
> >>>>> The recent commit: 90b5363acd47 ("sched: Clean up scheduler_ipi()")
> >>>>> got smp_call_function_single_async() subtly wrong. Even though it will
> >>>>> return -EBUSY when trying to re-use a csd, that condition is not
> >>>>> atomic and still requires external serialization.
> >>>>>
> >>>>> The change in ttwu_queue_remote() got this wrong.
> >>>>>
> >>>>> While on first reading ttwu_queue_remote() has an atomic test-and-set
> >>>>> that appears to serialize the use, the matching 'release' is not in
> >>>>> the right place to actually guarantee this serialization.
> >>>>>
> >>>>> The actual race is vs the sched_ttwu_pending() call in the idle loop;
> >>>>> that can run the wakeup-list without consuming the CSD.
> >>>>>
> >>>>> Instead of trying to chain the lists, merge them.
> >>>>>
> >>>>> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
> >>>>> ---
> >>>> ...
> >>>>> + /*
> >>>>> + * Assert the CSD_TYPE_TTWU layout is similar enough
> >>>>> + * for task_struct to be on the @call_single_queue.
> >>>>> + */
> >>>>> + BUILD_BUG_ON(offsetof(struct task_struct, wake_entry_type) - offsetof(struct task_struct, wake_entry) !=
> >>>>> + offsetof(struct __call_single_data, flags) - offsetof(struct __call_single_data, llist));
> >>>>> +
> >>>>
> >>>> There is no guarantee in C that
> >>>>
> >>>> type1 a;
> >>>> type2 b;
> >>>>
> >>>> in two different data structures means that offsetof(b) - offsetof(a)
> >>>> is the same in both data structures unless attributes such as
> >>>> __attribute__((__packed__)) are used.
> >>>>
> >>>> As result, this does and will cause a variety of build errors depending
> >>>> on the compiler version and compile flags.
> >>>>
> >>>> Guenter
> >>>
> >>> Yep, this breaks the build for me.
> >>
> >> -ENOCONFIG
> >
> > For me, the problem seems to be randstruct. To reproduce, you can use
> > (on x86_64):
> >
> > make defconfig
> > echo CONFIG_GCC_PLUGIN_RANDSTRUCT=y >> .config
> > make olddefconfig
> > make kernel/smp.o
> >
>
> I confirmed that disabling CONFIG_GCC_PLUGIN_RANDSTRUCT "fixes" the problem
> in my test builds. Maybe it would make sense to mark that configuration option
> for the time being as BROKEN.
>
Still occurring on Linus' tree. This needs to be fixed. (And not by removing
support for randstruct; that's not a "fix"...)
Shouldn't the kbuild test robot have caught this?
- Eric
next prev parent reply other threads:[~2020-06-09 20:21 UTC|newest]
Thread overview: 62+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-05-26 16:10 [RFC][PATCH 0/7] Fix the scheduler-IPI mess Peter Zijlstra
2020-05-26 16:10 ` [RFC][PATCH 1/7] sched: Fix smp_call_function_single_async() usage for ILB Peter Zijlstra
2020-05-26 23:56 ` Frederic Weisbecker
2020-05-27 10:23 ` Vincent Guittot
2020-05-27 11:28 ` Frederic Weisbecker
2020-05-27 12:07 ` Vincent Guittot
2020-05-29 15:26 ` Valentin Schneider
2020-06-01 9:52 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2020-06-01 11:40 ` Frederic Weisbecker
2020-05-26 16:10 ` [RFC][PATCH 2/7] smp: Optimize flush_smp_call_function_queue() Peter Zijlstra
2020-05-28 12:28 ` Frederic Weisbecker
2020-06-01 9:52 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2020-05-26 16:11 ` [RFC][PATCH 3/7] smp: Move irq_work_run() out of flush_smp_call_function_queue() Peter Zijlstra
2020-05-29 13:04 ` Frederic Weisbecker
2020-06-01 9:52 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2020-05-26 16:11 ` [RFC][PATCH 4/7] smp: Optimize send_call_function_single_ipi() Peter Zijlstra
2020-05-27 9:56 ` Peter Zijlstra
2020-05-27 10:15 ` Peter Zijlstra
2020-05-27 15:56 ` Paul E. McKenney
2020-05-27 16:35 ` Peter Zijlstra
2020-05-27 17:12 ` Peter Zijlstra
2020-05-27 19:39 ` Paul E. McKenney
2020-05-28 1:35 ` Joel Fernandes
2020-05-28 8:59 ` [tip: core/rcu] rcu: Allow for smp_call_function() running callbacks from idle tip-bot2 for Peter Zijlstra
2021-01-21 16:56 ` [RFC][PATCH 4/7] smp: Optimize send_call_function_single_ipi() Peter Zijlstra
2021-01-22 0:20 ` Paul E. McKenney
2021-01-22 8:31 ` Peter Zijlstra
2021-01-22 15:35 ` Paul E. McKenney
2020-05-29 13:01 ` Frederic Weisbecker
2020-06-01 9:52 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2020-05-26 16:11 ` [RFC][PATCH 5/7] irq_work, smp: Allow irq_work on call_single_queue Peter Zijlstra
2020-05-28 23:40 ` Frederic Weisbecker
2020-05-29 13:36 ` Peter Zijlstra
2020-06-05 9:37 ` Peter Zijlstra
2020-06-05 15:02 ` Frederic Weisbecker
2020-06-05 16:17 ` Peter Zijlstra
2020-06-05 15:24 ` Kees Cook
2020-06-10 13:24 ` Frederic Weisbecker
2020-06-01 9:52 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2020-05-26 16:11 ` [RFC][PATCH 6/7] sched: Add rq::ttwu_pending Peter Zijlstra
2020-06-01 9:52 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2020-05-26 16:11 ` [RFC][PATCH 7/7] sched: Replace rq::wake_list Peter Zijlstra
2020-05-29 15:10 ` Valdis Klētnieks
2020-06-01 9:52 ` [tip: sched/core] " tip-bot2 for Peter Zijlstra
2020-06-02 15:16 ` Frederic Weisbecker
2020-06-04 14:18 ` [RFC][PATCH 7/7] " Guenter Roeck
2020-06-05 0:24 ` Eric Biggers
2020-06-05 7:41 ` Peter Zijlstra
2020-06-05 16:15 ` Eric Biggers
2020-06-06 23:13 ` Guenter Roeck
2020-06-09 20:21 ` Eric Biggers [this message]
2020-06-09 21:25 ` Guenter Roeck
2020-06-09 21:38 ` Eric Biggers
2020-06-09 22:06 ` Peter Zijlstra
2020-06-09 23:03 ` Guenter Roeck
2020-06-10 9:09 ` Peter Zijlstra
2020-06-18 17:57 ` Steven Rostedt
2020-06-18 19:06 ` Guenter Roeck
2020-06-09 22:07 ` Peter Zijlstra
2020-06-05 8:10 ` Peter Zijlstra
2020-06-05 13:33 ` Guenter Roeck
2020-06-05 14:09 ` Peter Zijlstra
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=20200609202134.GA1105@sol.localdomain \
--to=ebiggers@kernel.org \
--cc=cai@lca.pw \
--cc=frederic@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux@roeck-us.net \
--cc=mgorman@techsingularity.net \
--cc=peterz@infradead.org \
--cc=tglx@linutronix.de \
--cc=x86@kernel.org \
/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).