2009-03-29 20:31:35

by Paul E. McKenney

[permalink] [raw]
Subject: [PATCH] v3 RCU: the bloatwatch edition

This patch is a version of RCU designed for (!SMP && EMBEDDED)
provided as a proof of concept of a small-footprint RCU implementation.
In particular, the implementation of synchronize_rcu() is extremely
lightweight and high performance. It passes rcutorture testing in each
of the four relevant configurations (combinations of NO_HZ and PREEMPT)
on x86. This saves about 900 bytes compared to Classic RCU, and a
couple kilobytes compared to Hierarchical RCU:

CONFIG_CLASSIC_RCU:

text data bss dec hex filename
363 12 24 399 18f kernel/rcupdate.o
1237 64 124 1425 591 kernel/rcuclassic.o
1824 Total

CONFIG_TREE_RCU:

text data bss dec hex filename
363 12 24 399 18f kernel/rcupdate.o
2344 240 184 2768 ad0 kernel/rcutree.o
3167 Total

CONFIG_TINY_RCU:

text data bss dec hex filename
294 12 24 330 14a kernel/rcupdate.o
563 36 0 599 257 kernel/rcutiny.o
929 Total


Changes from v2 (http://lkml.org/lkml/2009/2/3/333) include:

o Fix whitespace issues.

o Change short-circuit "||" operator to instead be "+" in order to
fix performance bug noted by "kraai" on LWN.

(http://lwn.net/Articles/324348/)

Changes from v1 (http://lkml.org/lkml/2009/1/13/440) include:

o This version depends on EMBEDDED as well as !SMP, as suggested
by Ingo.

o Updated rcu_needs_cpu() to unconditionally return zero,
permitting the CPU to enter dynticks-idle mode at any time.
This works because callbacks can be invoked upon entry to
dynticks-idle mode.

o I am now OK with this being included, based on a poll at the
Kernel Miniconf at linux.conf.au, where about ten people said
that they cared about saving 900 bytes on single-CPU systems.

o Applies to both mainline and tip/core/rcu.

Signed-off-by: Paul E. McKenney <[email protected]>
---

include/linux/rcupdate.h | 2
include/linux/rcutiny.h | 68 +++++++++++
init/Kconfig | 7 +
kernel/Makefile | 1
kernel/rcupdate.c | 4
kernel/rcutiny.c | 288 +++++++++++++++++++++++++++++++++++++++++++++++
6 files changed, 370 insertions(+)

diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h
index 528343e..19966aa 100644
--- a/include/linux/rcupdate.h
+++ b/include/linux/rcupdate.h
@@ -61,6 +61,8 @@ extern int rcu_scheduler_active;
#include <linux/rcutree.h>
#elif defined(CONFIG_PREEMPT_RCU)
#include <linux/rcupreempt.h>
+#elif CONFIG_TINY_RCU
+#include <linux/rcutiny.h>
#else
#error "Unknown RCU implementation specified to kernel configuration"
#endif /* #else #if defined(CONFIG_CLASSIC_RCU) */
diff --git a/include/linux/rcutiny.h b/include/linux/rcutiny.h
new file mode 100644
index 0000000..f007cfa
--- /dev/null
+++ b/include/linux/rcutiny.h
@@ -0,0 +1,68 @@
+/*
+ * Read-Copy Update mechanism for mutual exclusion, the Bloatwatch edition.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ * Copyright IBM Corporation, 2008
+ *
+ * Author: Paul E. McKenney <[email protected]>
+ *
+ * For detailed explanation of Read-Copy Update mechanism see -
+ * Documentation/RCU
+ */
+
+#ifndef __LINUX_TINY_H
+#define __LINUX_TINY_H
+
+#include <linux/cache.h>
+
+/* Global control variables for rcupdate callback mechanism. */
+struct rcu_ctrlblk {
+ long completed; /* Number of last completed batch. */
+ struct rcu_head *rcucblist; /* List of pending callbacks (CBs). */
+ struct rcu_head **donetail; /* ->next pointer of last "done" CB. */
+ struct rcu_head **curtail; /* ->next pointer of last CB. */
+};
+
+void rcu_qsctr_inc(int cpu);
+void rcu_bh_qsctr_inc(int cpu);
+extern int rcu_needs_cpu(int cpu);
+
+#define __rcu_read_lock() preempt_disable()
+#define __rcu_read_unlock() preempt_enable()
+#define __rcu_read_lock_bh() local_bh_disable()
+#define __rcu_read_unlock_bh() local_bh_enable()
+#define __synchronize_sched synchronize_rcu
+#define call_rcu_sched call_rcu
+
+extern void __rcu_init(void);
+#define rcu_init_sched() do { } while (0)
+extern void rcu_check_callbacks(int cpu, int user);
+extern void rcu_restart_cpu(int cpu);
+
+extern long rcu_batches_completed(void);
+extern long rcu_batches_completed_bh(void);
+
+#define rcu_pending(cpu) 1
+
+#ifdef CONFIG_NO_HZ
+void rcu_enter_nohz(void);
+void rcu_exit_nohz(void);
+#else /* #ifdef CONFIG_NO_HZ */
+#define rcu_enter_nohz() do { } while (0)
+#define rcu_exit_nohz() do { } while (0)
+#endif /* #else #ifdef CONFIG_NO_HZ */
+
+#endif /* __LINUX_RCUTINY_H */
diff --git a/init/Kconfig b/init/Kconfig
index f068071..9404637 100644
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -271,6 +271,13 @@ config PREEMPT_RCU
now-naive assumptions about each RCU read-side critical section
remaining on a given CPU through its execution.

+config TINY_RCU
+ bool "Tiny-Memory RCU"
+ depends on !SMP && EMBEDDED
+ help
+ This option greatly reduces the memory footprint of RCU,
+ but is usable only on UP systems.
+
endchoice

config RCU_TRACE
diff --git a/kernel/Makefile b/kernel/Makefile
index e4791b3..7a8c7c5 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -80,6 +80,7 @@ obj-$(CONFIG_RCU_TORTURE_TEST) += rcutorture.o
obj-$(CONFIG_CLASSIC_RCU) += rcuclassic.o
obj-$(CONFIG_TREE_RCU) += rcutree.o
obj-$(CONFIG_PREEMPT_RCU) += rcupreempt.o
+obj-$(CONFIG_TINY_RCU) += rcutiny.o
obj-$(CONFIG_TREE_RCU_TRACE) += rcutree_trace.o
obj-$(CONFIG_PREEMPT_RCU_TRACE) += rcupreempt_trace.o
obj-$(CONFIG_RELAY) += relay.o
diff --git a/kernel/rcupdate.c b/kernel/rcupdate.c
index cae8a05..3a5bd43 100644
--- a/kernel/rcupdate.c
+++ b/kernel/rcupdate.c
@@ -70,6 +70,8 @@ void wakeme_after_rcu(struct rcu_head *head)
complete(&rcu->completion);
}

+#ifndef CONFIG_TINY_RCU
+
/**
* synchronize_rcu - wait until a grace period has elapsed.
*
@@ -94,6 +96,8 @@ void synchronize_rcu(void)
}
EXPORT_SYMBOL_GPL(synchronize_rcu);

+#endif /* #ifndef CONFIG_TINY_RCU */
+
static void rcu_barrier_callback(struct rcu_head *notused)
{
if (atomic_dec_and_test(&rcu_barrier_cpu_count))
diff --git a/kernel/rcutiny.c b/kernel/rcutiny.c
new file mode 100644
index 0000000..3d36583
--- /dev/null
+++ b/kernel/rcutiny.c
@@ -0,0 +1,288 @@
+/*
+ * Read-Copy Update mechanism for mutual exclusion, the Bloatwatch edition.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ * Copyright IBM Corporation, 2008
+ *
+ * Author: Paul E. McKenney <[email protected]>
+ *
+ * For detailed explanation of Read-Copy Update mechanism see -
+ * Documentation/RCU
+ */
+
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/rcupdate.h>
+#include <linux/interrupt.h>
+#include <linux/sched.h>
+#include <linux/module.h>
+#include <linux/completion.h>
+#include <linux/moduleparam.h>
+#include <linux/notifier.h>
+#include <linux/cpu.h>
+#include <linux/mutex.h>
+#include <linux/time.h>
+
+/* Definition for rcupdate control block. */
+static struct rcu_ctrlblk rcu_ctrlblk = {
+ .completed = -300,
+ .rcucblist = NULL,
+ .donetail = &rcu_ctrlblk.rcucblist,
+ .curtail = &rcu_ctrlblk.rcucblist,
+};
+static struct rcu_ctrlblk rcu_bh_ctrlblk = {
+ .completed = -300,
+ .rcucblist = NULL,
+ .donetail = &rcu_bh_ctrlblk.rcucblist,
+ .curtail = &rcu_bh_ctrlblk.rcucblist,
+};
+
+#ifdef CONFIG_NO_HZ
+static long dynticks_nesting = 1;
+#endif /* #ifdef CONFIG_NO_HZ */
+
+/*
+ * Helper function for rcu_qsctr_inc() and rcu_bh_qsctr_inc().
+ */
+static int rcu_qsctr_help(struct rcu_ctrlblk *rcp)
+{
+ if (rcp->rcucblist != NULL &&
+ rcp->donetail != rcp->curtail) {
+ rcp->donetail = rcp->curtail;
+ return 1;
+ }
+ return 0;
+}
+
+/*
+ * Record an rcu quiescent state. And an rcu_bh quiescent state while we
+ * are at it, given that any rcu quiescent state is also an rcu_bh
+ * quiescent state. Use "+" instead of "||" to defeat short circuiting.
+ */
+void rcu_qsctr_inc(int cpu)
+{
+ if (rcu_qsctr_help(&rcu_ctrlblk) + rcu_qsctr_help(&rcu_bh_ctrlblk))
+ raise_softirq(RCU_SOFTIRQ);
+}
+
+/*
+ * Record an rcu_bh quiescent state.
+ */
+void rcu_bh_qsctr_inc(int cpu)
+{
+ if (rcu_qsctr_help(&rcu_bh_ctrlblk))
+ raise_softirq(RCU_SOFTIRQ);
+}
+
+/*
+ * Return non-zero if there is RCU work remaining to be done.
+ */
+int rcu_needs_cpu(int cpu)
+{
+ return 0;
+}
+
+/*
+ * Check to see if the scheduling-clock interrupt came from an extended
+ * quiescent state, and, if so, tell RCU about it.
+ */
+void rcu_check_callbacks(int cpu, int user)
+{
+ if (!rcu_needs_cpu(0))
+ return; /* RCU doesn't need anything to be done. */
+ if (user ||
+ (idle_cpu(cpu) &&
+ !in_softirq() &&
+ hardirq_count() <= (1 << HARDIRQ_SHIFT)))
+ rcu_qsctr_inc(cpu);
+ else if (!in_softirq())
+ rcu_bh_qsctr_inc(cpu);
+}
+
+#ifdef CONFIG_NO_HZ
+
+/*
+ * Enter dynticks-idle mode, which is an extended quiescent state
+ * if we have fully entered that mode (i.e., if the new value of
+ * dynticks_nesting is zero).
+ */
+void rcu_enter_nohz(void)
+{
+ if (--dynticks_nesting == 0)
+ rcu_qsctr_inc(0); /* implies rcu_bh_qsctr_inc(0) */
+}
+
+/*
+ * Exit dynticks-idle mode, so that we are no longer in an extended
+ * quiescent state.
+ */
+void rcu_exit_nohz(void)
+{
+ dynticks_nesting++;
+}
+
+/*
+ * Entering an interrupt handler exits nohz mode.
+ */
+void rcu_irq_enter(void)
+{
+ rcu_exit_nohz();
+}
+
+/*
+ * Exiting an interrupt handler enters nohz mode.
+ */
+void rcu_irq_exit(void)
+{
+ rcu_enter_nohz();
+}
+
+void rcu_nmi_enter(void)
+{
+}
+
+void rcu_nmi_exit(void)
+{
+}
+
+#endif /* #ifdef CONFIG_NO_HZ */
+
+/*
+ * Helper function for rcu_process_callbacks() that operates on the
+ * specified rcu_ctrlkblk structure.
+ */
+static void __rcu_process_callbacks(struct rcu_ctrlblk *rcp)
+{
+ unsigned long flags;
+ struct rcu_head *next, *list;
+
+ /* If no RCU callbacks ready to invoke, just return. */
+ if (&rcp->rcucblist == rcp->donetail)
+ return;
+
+ /* Move the ready-to-invoke callbacks to a local list. */
+ local_irq_save(flags);
+ rcp->completed++;
+ list = rcp->rcucblist;
+ rcp->rcucblist = *rcp->donetail;
+ *rcp->donetail = NULL;
+ if (rcp->curtail == rcp->donetail)
+ rcp->curtail = &rcp->rcucblist;
+ rcp->donetail = &rcp->rcucblist;
+ local_irq_restore(flags);
+
+ /* Invoke the callbacks on the local list. */
+ while (list) {
+ next = list->next;
+ prefetch(next);
+ list->func(list);
+ list = next;
+ }
+}
+
+/*
+ * Invoke any callbacks whose grace period has completed.
+ */
+static void rcu_process_callbacks(struct softirq_action *unused)
+{
+ __rcu_process_callbacks(&rcu_ctrlblk);
+ __rcu_process_callbacks(&rcu_bh_ctrlblk);
+}
+
+/*
+ * Wait for a grace period to elapse. But it is illegal to invoke
+ * synchronize_rcu() from within an RCU read-side critical section.
+ * Therefore, any legal call to synchronize_rcu() is a quiescent
+ * state, and so on a UP system, synchronize_rcu() need do nothing.
+ *
+ * Cool, huh? (Due to Josh Triplett.)
+ *
+ * However, we do update the grace-period counter to prevent rcutorture
+ * from hammering us.
+ */
+void synchronize_rcu(void)
+{
+ unsigned long flags;
+
+ local_irq_save(flags);
+ rcu_ctrlblk.completed++;
+ local_irq_restore(flags);
+}
+EXPORT_SYMBOL(synchronize_rcu);
+
+/*
+ * Helper function for call_rcu() and call_rcu_bh().
+ */
+static void __call_rcu(struct rcu_head *head,
+ void (*func)(struct rcu_head *rcu),
+ struct rcu_ctrlblk *rcp)
+{
+ unsigned long flags;
+
+ head->func = func;
+ head->next = NULL;
+ local_irq_save(flags);
+ *rcp->curtail = head;
+ rcp->curtail = &head->next;
+ local_irq_restore(flags);
+}
+
+/*
+ * Post an RCU callback to be invoked after the end of an RCU grace
+ * period. But since we have but one CPU, that would be after any
+ * quiescent state.
+ */
+void call_rcu(struct rcu_head *head,
+ void (*func)(struct rcu_head *rcu))
+{
+ __call_rcu(head, func, &rcu_ctrlblk);
+}
+EXPORT_SYMBOL(call_rcu);
+
+/*
+ * Post an RCU bottom-half callback to be invoked after any subsequent
+ * quiescent state.
+ */
+void call_rcu_bh(struct rcu_head *head,
+ void (*func)(struct rcu_head *rcu))
+{
+ __call_rcu(head, func, &rcu_bh_ctrlblk);
+}
+EXPORT_SYMBOL(call_rcu_bh);
+
+/*
+ * Return the number of grace periods.
+ */
+long rcu_batches_completed(void)
+{
+ return rcu_ctrlblk.completed;
+}
+EXPORT_SYMBOL(rcu_batches_completed);
+
+/*
+ * Return the number of bottom-half grace periods.
+ */
+long rcu_batches_completed_bh(void)
+{
+ return rcu_bh_ctrlblk.completed;
+}
+EXPORT_SYMBOL(rcu_batches_completed_bh);
+
+void __init __rcu_init(void)
+{
+ open_softirq(RCU_SOFTIRQ, rcu_process_callbacks);
+}


2009-04-02 22:37:07

by Ingo Molnar

[permalink] [raw]
Subject: Re: [PATCH] v3 RCU: the bloatwatch edition


* Paul E. McKenney <[email protected]> wrote:

> This patch is a version of RCU designed for (!SMP && EMBEDDED)
> provided as a proof of concept of a small-footprint RCU
> implementation. In particular, the implementation of
> synchronize_rcu() is extremely lightweight and high performance.
> It passes rcutorture testing in each of the four relevant
> configurations (combinations of NO_HZ and PREEMPT) on x86. This
> saves about 900 bytes compared to Classic RCU, and a couple
> kilobytes compared to Hierarchical RCU:
>
> CONFIG_CLASSIC_RCU:
>
> text data bss dec hex filename
> 363 12 24 399 18f kernel/rcupdate.o
> 1237 64 124 1425 591 kernel/rcuclassic.o
> 1824 Total
>
> CONFIG_TREE_RCU:
>
> text data bss dec hex filename
> 363 12 24 399 18f kernel/rcupdate.o
> 2344 240 184 2768 ad0 kernel/rcutree.o
> 3167 Total
>
> CONFIG_TINY_RCU:
>
> text data bss dec hex filename
> 294 12 24 330 14a kernel/rcupdate.o
> 563 36 0 599 257 kernel/rcutiny.o
> 929 Total
>
>
> Changes from v2 (http://lkml.org/lkml/2009/2/3/333) include:
>
> o Fix whitespace issues.
>
> o Change short-circuit "||" operator to instead be "+" in order to
> fix performance bug noted by "kraai" on LWN.
>
> (http://lwn.net/Articles/324348/)
>
> Changes from v1 (http://lkml.org/lkml/2009/1/13/440) include:
>
> o This version depends on EMBEDDED as well as !SMP, as suggested
> by Ingo.
>
> o Updated rcu_needs_cpu() to unconditionally return zero,
> permitting the CPU to enter dynticks-idle mode at any time.
> This works because callbacks can be invoked upon entry to
> dynticks-idle mode.
>
> o I am now OK with this being included, based on a poll at the
> Kernel Miniconf at linux.conf.au, where about ten people said
> that they cared about saving 900 bytes on single-CPU systems.
>
> o Applies to both mainline and tip/core/rcu.

Andrew, what do you think?

A worry is yet another RCU variant - we already have 3.

A trick we could use would be to put it into Documentation/rcu/,
linked in via some clever Makefile magic and only usable if a
ultra-embedded developer does a build with something like
CONFIG_RCU_TINY=y. That way there's no real maintenance and testing
overhead.

It _does_ have documentation value beyond the ~900 bytes: it's the
simplest and smallest possible still-working UP RCU implementation
so it would be easy to teach RCU concepts via that, gradually.

Ingo

2009-04-02 22:49:47

by Paul Mundt

[permalink] [raw]
Subject: Re: [PATCH] v3 RCU: the bloatwatch edition

On Fri, Apr 03, 2009 at 12:36:05AM +0200, Ingo Molnar wrote:
> * Paul E. McKenney <[email protected]> wrote:
> > This patch is a version of RCU designed for (!SMP && EMBEDDED)
> > provided as a proof of concept of a small-footprint RCU
> > implementation. In particular, the implementation of
> > synchronize_rcu() is extremely lightweight and high performance.
> > It passes rcutorture testing in each of the four relevant
> > configurations (combinations of NO_HZ and PREEMPT) on x86. This
> > saves about 900 bytes compared to Classic RCU, and a couple
> > kilobytes compared to Hierarchical RCU:
>
> Andrew, what do you think?
>
> A worry is yet another RCU variant - we already have 3.
>
> A trick we could use would be to put it into Documentation/rcu/,
> linked in via some clever Makefile magic and only usable if a
> ultra-embedded developer does a build with something like
> CONFIG_RCU_TINY=y. That way there's no real maintenance and testing
> overhead.
>
> It _does_ have documentation value beyond the ~900 bytes: it's the
> simplest and smallest possible still-working UP RCU implementation
> so it would be easy to teach RCU concepts via that, gradually.
>
A similar argument could have been used for tiny-shmem when it was first
integrated. As this is hiding behind CONFIG_EMBEDDED, most users are not
going to run in to it, so the confusion of 1 more RCU variant is not
likely to be a problem for those that aren't actively seeking it out.

So, personally I think it is a good idea, and I have no reservations
about default enabling it for a number of more constrained SH platforms.

2009-04-02 22:51:01

by Andrew Morton

[permalink] [raw]
Subject: Re: [PATCH] v3 RCU: the bloatwatch edition

On Fri, 3 Apr 2009 00:36:05 +0200
Ingo Molnar <[email protected]> wrote:

>
> Andrew, what do you think?

I'm really struggling to see how the 900-odd bytes saved justifies
creating (yet another) variant of core kernel machinery.

> A worry is yet another RCU variant - we already have 3.

That would make four?

I wonder if that was sane of us.

> A trick we could use would be to put it into Documentation/rcu/,
> linked in via some clever Makefile magic and only usable if a
> ultra-embedded developer does a build with something like
> CONFIG_RCU_TINY=y. That way there's no real maintenance and testing
> overhead.
>
> It _does_ have documentation value beyond the ~900 bytes: it's the
> simplest and smallest possible still-working UP RCU implementation
> so it would be easy to teach RCU concepts via that, gradually.
>

hm.

2009-04-03 00:03:24

by Paul E. McKenney

[permalink] [raw]
Subject: Re: [PATCH] v3 RCU: the bloatwatch edition

On Thu, Apr 02, 2009 at 03:44:44PM -0700, Andrew Morton wrote:
> On Fri, 3 Apr 2009 00:36:05 +0200
> Ingo Molnar <[email protected]> wrote:
>
> >
> > Andrew, what do you think?
>
> I'm really struggling to see how the 900-odd bytes saved justifies
> creating (yet another) variant of core kernel machinery.
>
> > A worry is yet another RCU variant - we already have 3.
>
> That would make four?

Three, once treercu is deemed stable enough to permit dropping Classic RCU.

Thanx, Paul

2009-04-03 06:52:58

by Andi Kleen

[permalink] [raw]
Subject: Re: [PATCH] v3 RCU: the bloatwatch edition

Andrew Morton <[email protected]> writes:

> On Fri, 3 Apr 2009 00:36:05 +0200
> Ingo Molnar <[email protected]> wrote:
>
>>
>> Andrew, what do you think?
>
> I'm really struggling to see how the 900-odd bytes saved justifies
> creating (yet another) variant of core kernel machinery.

Also it's unclear if anything special cased !SMP is worth it for the
future. After all multi core or SMT is becoming more and more common even in
the embedded world.

-Andi

--
[email protected] -- Speaking for myself only.

2009-04-03 10:45:58

by Lennert Buytenhek

[permalink] [raw]
Subject: Re: [PATCH] v3 RCU: the bloatwatch edition

On Fri, Apr 03, 2009 at 08:52:41AM +0200, Andi Kleen wrote:

> >> Andrew, what do you think?
> >
> > I'm really struggling to see how the 900-odd bytes saved justifies
> > creating (yet another) variant of core kernel machinery.
>
> Also it's unclear if anything special cased !SMP is worth it for the
> future. After all multi core or SMT is becoming more and more common
> even in the embedded world.

At least for ARM, >99% of the dual core CPUs out there are non-SMP.
(With one core typically running Linux and another core typically not
running an OS at all: packet processing, DSP things, baseband stack,
etc. That second CPU core often doesn't even have an MMU.) I don't
think any SMP ARMs have been widely deployed yet.

Even if SMP ARM chips become widely available: the die area size of
a chip alone pretty much tells you the cost of that chip, and in a
world where every sub-mm^2 reduction in area matters a _lot_, I doubt
ARM CPU manufacturers will start including second CPU cores in chips
meant for things like cell phones, wireless access points and broadband
routers (ARM CPUs tend to be highly application-specific) just because
they can.

2009-04-08 16:39:34

by Ingo Molnar

[permalink] [raw]
Subject: Re: [PATCH] v3 RCU: the bloatwatch edition


* Paul Mundt <[email protected]> wrote:

> On Fri, Apr 03, 2009 at 12:36:05AM +0200, Ingo Molnar wrote:
> > * Paul E. McKenney <[email protected]> wrote:
> > > This patch is a version of RCU designed for (!SMP && EMBEDDED)
> > > provided as a proof of concept of a small-footprint RCU
> > > implementation. In particular, the implementation of
> > > synchronize_rcu() is extremely lightweight and high performance.
> > > It passes rcutorture testing in each of the four relevant
> > > configurations (combinations of NO_HZ and PREEMPT) on x86. This
> > > saves about 900 bytes compared to Classic RCU, and a couple
> > > kilobytes compared to Hierarchical RCU:
> >
> > Andrew, what do you think?
> >
> > A worry is yet another RCU variant - we already have 3.
> >
> > A trick we could use would be to put it into Documentation/rcu/,
> > linked in via some clever Makefile magic and only usable if a
> > ultra-embedded developer does a build with something like
> > CONFIG_RCU_TINY=y. That way there's no real maintenance and testing
> > overhead.
> >
> > It _does_ have documentation value beyond the ~900 bytes: it's the
> > simplest and smallest possible still-working UP RCU implementation
> > so it would be easy to teach RCU concepts via that, gradually.
>
> A similar argument could have been used for tiny-shmem when it was
> first integrated. As this is hiding behind CONFIG_EMBEDDED, most
> users are not going to run in to it, so the confusion of 1 more
> RCU variant is not likely to be a problem for those that aren't
> actively seeking it out.
>
> So, personally I think it is a good idea, and I have no
> reservations about default enabling it for a number of more
> constrained SH platforms.

but at least tiny-shmem is now nicely hidden in mm/shmem.c, in an
unintrusive !CONFIG_SHMEM branch. There's no CONFIG_TINY_SHMEM
option anymore - it's all done in the !CONFIG_SHMEM case.

Is tiny-RCU in the same category?

Ingo

2009-04-08 17:04:47

by Paul Mundt

[permalink] [raw]
Subject: Re: [PATCH] v3 RCU: the bloatwatch edition

On Wed, Apr 08, 2009 at 06:38:38PM +0200, Ingo Molnar wrote:
>
> * Paul Mundt <[email protected]> wrote:
>
> > On Fri, Apr 03, 2009 at 12:36:05AM +0200, Ingo Molnar wrote:
> > > * Paul E. McKenney <[email protected]> wrote:
> > > > This patch is a version of RCU designed for (!SMP && EMBEDDED)
> > > > provided as a proof of concept of a small-footprint RCU
> > > > implementation. In particular, the implementation of
> > > > synchronize_rcu() is extremely lightweight and high performance.
> > > > It passes rcutorture testing in each of the four relevant
> > > > configurations (combinations of NO_HZ and PREEMPT) on x86. This
> > > > saves about 900 bytes compared to Classic RCU, and a couple
> > > > kilobytes compared to Hierarchical RCU:
> > >
> > > Andrew, what do you think?
> > >
> > > A worry is yet another RCU variant - we already have 3.
> > >
> > > A trick we could use would be to put it into Documentation/rcu/,
> > > linked in via some clever Makefile magic and only usable if a
> > > ultra-embedded developer does a build with something like
> > > CONFIG_RCU_TINY=y. That way there's no real maintenance and testing
> > > overhead.
> > >
> > > It _does_ have documentation value beyond the ~900 bytes: it's the
> > > simplest and smallest possible still-working UP RCU implementation
> > > so it would be easy to teach RCU concepts via that, gradually.
> >
> > A similar argument could have been used for tiny-shmem when it was
> > first integrated. As this is hiding behind CONFIG_EMBEDDED, most
> > users are not going to run in to it, so the confusion of 1 more
> > RCU variant is not likely to be a problem for those that aren't
> > actively seeking it out.
> >
> > So, personally I think it is a good idea, and I have no
> > reservations about default enabling it for a number of more
> > constrained SH platforms.
>
> but at least tiny-shmem is now nicely hidden in mm/shmem.c, in an
> unintrusive !CONFIG_SHMEM branch. There's no CONFIG_TINY_SHMEM
> option anymore - it's all done in the !CONFIG_SHMEM case.
>
Now it is, yes, but it was not originally, and it was still useful when
it was split out. If we are going to tolerate multiple RCU
implementations in the kernel, then I see no reason to not include
tiny-RCU in the same category. Even in the case where some of the other
RCU variants go away, tiny-RCU remains a viable option for simple
platforms that are more concerned about memory than anything else, so
it's always a valid alternative.

If in the future things are more consolidated and the config option goes
away then great, but that hardly seems like a sane prerequisite for
merging it. CONFIG_EMBEDDED handles this just fine. You don't need to
enable it if you don't wish to, but it's certainly measurable enough to
be useful for those of us that have no problems enabling it ;-)

2009-04-08 18:44:42

by Paul E. McKenney

[permalink] [raw]
Subject: Re: [PATCH] v3 RCU: the bloatwatch edition

On Thu, Apr 09, 2009 at 01:55:29AM +0900, Paul Mundt wrote:
> On Wed, Apr 08, 2009 at 06:38:38PM +0200, Ingo Molnar wrote:
> >
> > * Paul Mundt <[email protected]> wrote:
> >
> > > On Fri, Apr 03, 2009 at 12:36:05AM +0200, Ingo Molnar wrote:
> > > > * Paul E. McKenney <[email protected]> wrote:
> > > > > This patch is a version of RCU designed for (!SMP && EMBEDDED)
> > > > > provided as a proof of concept of a small-footprint RCU
> > > > > implementation. In particular, the implementation of
> > > > > synchronize_rcu() is extremely lightweight and high performance.
> > > > > It passes rcutorture testing in each of the four relevant
> > > > > configurations (combinations of NO_HZ and PREEMPT) on x86. This
> > > > > saves about 900 bytes compared to Classic RCU, and a couple
> > > > > kilobytes compared to Hierarchical RCU:
> > > >
> > > > Andrew, what do you think?
> > > >
> > > > A worry is yet another RCU variant - we already have 3.
> > > >
> > > > A trick we could use would be to put it into Documentation/rcu/,
> > > > linked in via some clever Makefile magic and only usable if a
> > > > ultra-embedded developer does a build with something like
> > > > CONFIG_RCU_TINY=y. That way there's no real maintenance and testing
> > > > overhead.
> > > >
> > > > It _does_ have documentation value beyond the ~900 bytes: it's the
> > > > simplest and smallest possible still-working UP RCU implementation
> > > > so it would be easy to teach RCU concepts via that, gradually.
> > >
> > > A similar argument could have been used for tiny-shmem when it was
> > > first integrated. As this is hiding behind CONFIG_EMBEDDED, most
> > > users are not going to run in to it, so the confusion of 1 more
> > > RCU variant is not likely to be a problem for those that aren't
> > > actively seeking it out.
> > >
> > > So, personally I think it is a good idea, and I have no
> > > reservations about default enabling it for a number of more
> > > constrained SH platforms.
> >
> > but at least tiny-shmem is now nicely hidden in mm/shmem.c, in an
> > unintrusive !CONFIG_SHMEM branch. There's no CONFIG_TINY_SHMEM
> > option anymore - it's all done in the !CONFIG_SHMEM case.
> >
> Now it is, yes, but it was not originally, and it was still useful when
> it was split out. If we are going to tolerate multiple RCU
> implementations in the kernel, then I see no reason to not include
> tiny-RCU in the same category. Even in the case where some of the other
> RCU variants go away, tiny-RCU remains a viable option for simple
> platforms that are more concerned about memory than anything else, so
> it's always a valid alternative.
>
> If in the future things are more consolidated and the config option goes
> away then great, but that hardly seems like a sane prerequisite for
> merging it. CONFIG_EMBEDDED handles this just fine. You don't need to
> enable it if you don't wish to, but it's certainly measurable enough to
> be useful for those of us that have no problems enabling it ;-)

>From a kernel-size viewpoint:

788 kernel/rcuclassic.c
190 include/linux/rcuclassic.h
978 total

288 kernel/rcutiny.c
68 include/linux/rcutiny.h
356 total

Almost a 3x decrease in lines of code. So, Seems to me that dropping
rcuclassic (as rcutree proves itself) and taking up rcutiny instead is
a good step forward. ;-)

Thanx, Paul

2009-04-28 14:25:31

by David Howells

[permalink] [raw]
Subject: Re: [PATCH] v3 RCU: the bloatwatch edition

Paul E. McKenney <[email protected]> wrote:

> This patch is a version of RCU designed for (!SMP && EMBEDDED)
> provided as a proof of concept of a small-footprint RCU implementation.
> In particular, the implementation of synchronize_rcu() is extremely
> lightweight and high performance. It passes rcutorture testing in each
> of the four relevant configurations (combinations of NO_HZ and PREEMPT)
> on x86. This saves about 900 bytes compared to Classic RCU, and a
> couple kilobytes compared to Hierarchical RCU:

On FRV, CLASSIC_RCU:

text data bss dec hex filename
2616 184 0 2800 af0 kernel/rcuclassic.o
884 32 20 936 3a8 kernel/rcupdate.o

TREE_RCU:

3940 328 0 4268 10ac kernel/rcutree.o
884 32 20 936 3a8 kernel/rcupdate.o

TINY_RCU:

1152 32 0 1184 4a0 kernel/rcutiny.o
836 32 20 888 378 kernel/rcupdate.o

It works on my FRV board.

Possibly TINY_RCU could be shrunk a bit more by a judicious bit of inlining of
some of the very small functions.

Acked-by: David Howells <[email protected]>

2009-04-28 19:45:52

by Paul E. McKenney

[permalink] [raw]
Subject: Re: [PATCH] v3 RCU: the bloatwatch edition

On Tue, Apr 28, 2009 at 03:24:14PM +0100, David Howells wrote:
> Paul E. McKenney <[email protected]> wrote:
>
> > This patch is a version of RCU designed for (!SMP && EMBEDDED)
> > provided as a proof of concept of a small-footprint RCU implementation.
> > In particular, the implementation of synchronize_rcu() is extremely
> > lightweight and high performance. It passes rcutorture testing in each
> > of the four relevant configurations (combinations of NO_HZ and PREEMPT)
> > on x86. This saves about 900 bytes compared to Classic RCU, and a
> > couple kilobytes compared to Hierarchical RCU:
>
> On FRV, CLASSIC_RCU:
>
> text data bss dec hex filename
> 2616 184 0 2800 af0 kernel/rcuclassic.o
> 884 32 20 936 3a8 kernel/rcupdate.o
>
> TREE_RCU:
>
> 3940 328 0 4268 10ac kernel/rcutree.o
> 884 32 20 936 3a8 kernel/rcupdate.o
>
> TINY_RCU:
>
> 1152 32 0 1184 4a0 kernel/rcutiny.o
> 836 32 20 888 378 kernel/rcupdate.o
>
> It works on my FRV board.
>
> Possibly TINY_RCU could be shrunk a bit more by a judicious bit of inlining of
> some of the very small functions.
>
> Acked-by: David Howells <[email protected]>

Thank you for looking this over!

Your thought is that some of the functions could be moved to tinyrcu.h?
Indeed, some of them would be smaller if inlined than even the call
sequence. For example, rcu_needs_cpu() should remove code from the
dynticks implementation given that it always returns zero.

Thanx, Paul

2009-04-28 21:41:35

by David Howells

[permalink] [raw]
Subject: Re: [PATCH] v3 RCU: the bloatwatch edition

Paul E. McKenney <[email protected]> wrote:

> Your thought is that some of the functions could be moved to tinyrcu.h?
> Indeed, some of them would be smaller if inlined than even the call
> sequence. For example, rcu_needs_cpu() should remove code from the
> dynticks implementation given that it always returns zero.

tinyrcu.h is probably not a bad idea. Some of the functions are trivial, and
the code to do a function call is bigger than the body of the function itself.

rcu_exit_nohz(), rcu_nmi_enter/exit(), rcu_batches_completed[_bh](), for
example. Even call_rcu() and call_rcu_bh() might perhaps benefit from
inlining.

David

2009-04-29 00:57:27

by Paul E. McKenney

[permalink] [raw]
Subject: Re: [PATCH] v3 RCU: the bloatwatch edition

On Tue, Apr 28, 2009 at 10:39:59PM +0100, David Howells wrote:
> Paul E. McKenney <[email protected]> wrote:
>
> > Your thought is that some of the functions could be moved to tinyrcu.h?
> > Indeed, some of them would be smaller if inlined than even the call
> > sequence. For example, rcu_needs_cpu() should remove code from the
> > dynticks implementation given that it always returns zero.
>
> tinyrcu.h is probably not a bad idea. Some of the functions are trivial, and
> the code to do a function call is bigger than the body of the function itself.
>
> rcu_exit_nohz(), rcu_nmi_enter/exit(), rcu_batches_completed[_bh](), for
> example. Even call_rcu() and call_rcu_bh() might perhaps benefit from
> inlining.

Well, here is something that should knock off an additional 100 bytes or
so. Untested, probably does not compile. I skipped putting call_rcu()
in tinyrcu.h because it is called so many times that the extra argument
would probably bite harder than the current code + export.

Signed-off-by: Paul E. McKenney <[email protected]>
---

include/linux/hardirq.h | 19 ++--
include/linux/rcupdate.h | 2
include/linux/rcutiny.h | 112 ++++++++++++++++++++++++
init/Kconfig | 7 +
kernel/Makefile | 1
kernel/rcupdate.c | 4
kernel/rcutiny.c | 214 +++++++++++++++++++++++++++++++++++++++++++++++
7 files changed, 352 insertions(+), 7 deletions(-)

diff -urpNa -X dontdiff linux-2.6.30-rc2-rcu/include/linux/hardirq.h linux-2.6.29-tinyrcu/include/linux/hardirq.h
--- linux-2.6.30-rc2-rcu/include/linux/hardirq.h 2009-03-23 16:12:14.000000000 -0700
+++ linux-2.6.29-tinyrcu/include/linux/hardirq.h 2009-04-28 17:05:11.000000000 -0700
@@ -119,17 +119,22 @@ static inline void account_system_vtime(
}
#endif

-#if defined(CONFIG_NO_HZ) && !defined(CONFIG_CLASSIC_RCU)
+#if !defined(CONFIG_NO_HZ) || defined(CONFIG_CLASSIC_RCU)
+# define rcu_irq_enter() do { } while (0)
+# define rcu_irq_exit() do { } while (0)
+# define rcu_nmi_enter() do { } while (0)
+# define rcu_nmi_exit() do { } while (0)
+#elif defined(CONFIG_RCU_TINY)
+# define rcu_irq_enter rcu_exit_nohz
+# define rcu_irq_exit rcu_enter_nohz
+# define rcu_nmi_enter() do { } while (0)
+# define rcu_nmi_exit() do { } while (0)
+#else
extern void rcu_irq_enter(void);
extern void rcu_irq_exit(void);
extern void rcu_nmi_enter(void);
extern void rcu_nmi_exit(void);
-#else
-# define rcu_irq_enter() do { } while (0)
-# define rcu_irq_exit() do { } while (0)
-# define rcu_nmi_enter() do { } while (0)
-# define rcu_nmi_exit() do { } while (0)
-#endif /* #if defined(CONFIG_NO_HZ) && !defined(CONFIG_CLASSIC_RCU) */
+#endif

/*
* It is safe to do non-atomic ops on ->hardirq_context,
diff -urpNa -X dontdiff linux-2.6.30-rc2-rcu/include/linux/rcupdate.h linux-2.6.29-tinyrcu/include/linux/rcupdate.h
--- linux-2.6.30-rc2-rcu/include/linux/rcupdate.h 2009-04-28 17:20:36.000000000 -0700
+++ linux-2.6.29-tinyrcu/include/linux/rcupdate.h 2009-04-28 16:40:37.000000000 -0700
@@ -60,6 +60,8 @@ extern int rcu_scheduler_active;
#include <linux/rcutree.h>
#elif defined(CONFIG_PREEMPT_RCU)
#include <linux/rcupreempt.h>
+#elif CONFIG_TINY_RCU
+#include <linux/rcutiny.h>
#else
#error "Unknown RCU implementation specified to kernel configuration"
#endif /* #else #if defined(CONFIG_CLASSIC_RCU) */
diff -urpNa -X dontdiff linux-2.6.30-rc2-rcu/include/linux/rcutiny.h linux-2.6.29-tinyrcu/include/linux/rcutiny.h
--- linux-2.6.30-rc2-rcu/include/linux/rcutiny.h 1969-12-31 16:00:00.000000000 -0800
+++ linux-2.6.29-tinyrcu/include/linux/rcutiny.h 2009-04-28 17:17:01.000000000 -0700
@@ -0,0 +1,112 @@
+/*
+ * Read-Copy Update mechanism for mutual exclusion, the Bloatwatch edition.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ * Copyright IBM Corporation, 2008
+ *
+ * Author: Paul E. McKenney <[email protected]>
+ *
+ * For detailed explanation of Read-Copy Update mechanism see -
+ * Documentation/RCU
+ */
+
+#ifndef __LINUX_TINY_H
+#define __LINUX_TINY_H
+
+#include <linux/cache.h>
+
+/* Global control variables for rcupdate callback mechanism. */
+struct rcu_ctrlblk {
+ long completed; /* Number of last completed batch. */
+ struct rcu_head *rcucblist; /* List of pending callbacks (CBs). */
+ struct rcu_head **donetail; /* ->next pointer of last "done" CB. */
+ struct rcu_head **curtail; /* ->next pointer of last CB. */
+};
+
+extern struct rcu_ctrlblk rcu_ctrlblk;
+extern struct rcu_ctrlblk rcu_bh_ctrlblk;
+
+/*
+ * Return non-zero if there is RCU work remaining to be done.
+ */
+static inline int rcu_needs_cpu(int cpu)
+{
+ return 0;
+}
+
+void rcu_qsctr_inc(int cpu);
+void rcu_bh_qsctr_inc(int cpu);
+
+#define __rcu_read_lock() preempt_disable()
+#define __rcu_read_unlock() preempt_enable()
+#define __rcu_read_lock_bh() local_bh_disable()
+#define __rcu_read_unlock_bh() local_bh_enable()
+#define __synchronize_sched synchronize_rcu
+#define call_rcu_sched call_rcu
+
+#define rcu_init_sched() do { } while (0)
+extern void rcu_check_callbacks(int cpu, int user);
+extern void __rcu_init(void);
+/* extern void rcu_restart_cpu(int cpu); */
+
+/*
+ * Return the number of grace periods.
+ */
+static inline long rcu_batches_completed(void)
+{
+ return rcu_ctrlblk.completed;
+}
+
+/*
+ * Return the number of bottom-half grace periods.
+ */
+static inline long rcu_batches_completed_bh(void)
+{
+ return rcu_bh_ctrlblk.completed;
+}
+
+#define rcu_pending(cpu) 1
+
+#ifdef CONFIG_NO_HZ
+
+extern long rcu_dynticks_nesting;
+
+/*
+ * Enter dynticks-idle mode, which is an extended quiescent state
+ * if we have fully entered that mode (i.e., if the new value of
+ * dynticks_nesting is zero).
+ */
+static inline void rcu_enter_nohz(void)
+{
+ if (--rcu_dynticks_nesting == 0)
+ rcu_qsctr_inc(0); /* implies rcu_bh_qsctr_inc(0) */
+}
+
+/*
+ * Exit dynticks-idle mode, so that we are no longer in an extended
+ * quiescent state.
+ */
+static inline void rcu_exit_nohz(void)
+{
+ rcu_dynticks_nesting++;
+}
+
+#else /* #ifdef CONFIG_NO_HZ */
+#define rcu_enter_nohz() do { } while (0)
+#define rcu_exit_nohz() do { } while (0)
+#endif /* #else #ifdef CONFIG_NO_HZ */
+
+#endif /* __LINUX_RCUTINY_H */
diff -urpNa -X dontdiff linux-2.6.30-rc2-rcu/init/Kconfig linux-2.6.29-tinyrcu/init/Kconfig
--- linux-2.6.30-rc2-rcu/init/Kconfig 2009-03-23 16:12:14.000000000 -0700
+++ linux-2.6.29-tinyrcu/init/Kconfig 2009-04-28 16:40:37.000000000 -0700
@@ -271,6 +271,13 @@ config PREEMPT_RCU
now-naive assumptions about each RCU read-side critical section
remaining on a given CPU through its execution.

+config TINY_RCU
+ bool "Tiny-Memory RCU"
+ depends on !SMP && EMBEDDED
+ help
+ This option greatly reduces the memory footprint of RCU,
+ but is usable only on UP systems.
+
endchoice

config RCU_TRACE
diff -urpNa -X dontdiff linux-2.6.30-rc2-rcu/kernel/Makefile linux-2.6.29-tinyrcu/kernel/Makefile
--- linux-2.6.30-rc2-rcu/kernel/Makefile 2009-03-23 16:12:14.000000000 -0700
+++ linux-2.6.29-tinyrcu/kernel/Makefile 2009-04-28 16:40:37.000000000 -0700
@@ -80,6 +80,7 @@ obj-$(CONFIG_RCU_TORTURE_TEST) += rcutor
obj-$(CONFIG_CLASSIC_RCU) += rcuclassic.o
obj-$(CONFIG_TREE_RCU) += rcutree.o
obj-$(CONFIG_PREEMPT_RCU) += rcupreempt.o
+obj-$(CONFIG_TINY_RCU) += rcutiny.o
obj-$(CONFIG_TREE_RCU_TRACE) += rcutree_trace.o
obj-$(CONFIG_PREEMPT_RCU_TRACE) += rcupreempt_trace.o
obj-$(CONFIG_RELAY) += relay.o
diff -urpNa -X dontdiff linux-2.6.30-rc2-rcu/kernel/rcupdate.c linux-2.6.29-tinyrcu/kernel/rcupdate.c
--- linux-2.6.30-rc2-rcu/kernel/rcupdate.c 2009-04-28 17:20:36.000000000 -0700
+++ linux-2.6.29-tinyrcu/kernel/rcupdate.c 2009-04-28 16:40:37.000000000 -0700
@@ -74,6 +74,8 @@ void wakeme_after_rcu(struct rcu_head *
complete(&rcu->completion);
}

+#ifndef CONFIG_TINY_RCU
+
/**
* synchronize_rcu - wait until a grace period has elapsed.
*
@@ -98,6 +100,8 @@ void synchronize_rcu(void)
}
EXPORT_SYMBOL_GPL(synchronize_rcu);

+#endif /* #ifndef CONFIG_TINY_RCU */
+
static void rcu_barrier_callback(struct rcu_head *notused)
{
if (atomic_dec_and_test(&rcu_barrier_cpu_count))
diff -urpNa -X dontdiff linux-2.6.30-rc2-rcu/kernel/rcutiny.c linux-2.6.29-tinyrcu/kernel/rcutiny.c
--- linux-2.6.30-rc2-rcu/kernel/rcutiny.c 1969-12-31 16:00:00.000000000 -0800
+++ linux-2.6.29-tinyrcu/kernel/rcutiny.c 2009-04-28 17:15:40.000000000 -0700
@@ -0,0 +1,214 @@
+/*
+ * Read-Copy Update mechanism for mutual exclusion, the Bloatwatch edition.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ * Copyright IBM Corporation, 2008
+ *
+ * Author: Paul E. McKenney <[email protected]>
+ *
+ * For detailed explanation of Read-Copy Update mechanism see -
+ * Documentation/RCU
+ */
+
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/rcupdate.h>
+#include <linux/interrupt.h>
+#include <linux/sched.h>
+#include <linux/module.h>
+#include <linux/completion.h>
+#include <linux/moduleparam.h>
+#include <linux/notifier.h>
+#include <linux/cpu.h>
+#include <linux/mutex.h>
+#include <linux/time.h>
+
+/* Definition for rcupdate control block. */
+struct rcu_ctrlblk rcu_ctrlblk = {
+ .completed = -300,
+ .rcucblist = NULL,
+ .donetail = &rcu_ctrlblk.rcucblist,
+ .curtail = &rcu_ctrlblk.rcucblist,
+};
+struct rcu_ctrlblk rcu_bh_ctrlblk = {
+ .completed = -300,
+ .rcucblist = NULL,
+ .donetail = &rcu_bh_ctrlblk.rcucblist,
+ .curtail = &rcu_bh_ctrlblk.rcucblist,
+};
+
+#ifdef CONFIG_NO_HZ
+long rcu_dynticks_nesting = 1;
+#endif /* #ifdef CONFIG_NO_HZ */
+
+/*
+ * Helper function for rcu_qsctr_inc() and rcu_bh_qsctr_inc().
+ */
+static int rcu_qsctr_help(struct rcu_ctrlblk *rcp)
+{
+ if (rcp->rcucblist != NULL &&
+ rcp->donetail != rcp->curtail) {
+ rcp->donetail = rcp->curtail;
+ return 1;
+ }
+ return 0;
+}
+
+/*
+ * Record an rcu quiescent state. And an rcu_bh quiescent state while we
+ * are at it, given that any rcu quiescent state is also an rcu_bh
+ * quiescent state. Use "+" instead of "||" to defeat short circuiting.
+ */
+void rcu_qsctr_inc(int cpu)
+{
+ if (rcu_qsctr_help(&rcu_ctrlblk) + rcu_qsctr_help(&rcu_bh_ctrlblk))
+ raise_softirq(RCU_SOFTIRQ);
+}
+
+/*
+ * Record an rcu_bh quiescent state.
+ */
+void rcu_bh_qsctr_inc(int cpu)
+{
+ if (rcu_qsctr_help(&rcu_bh_ctrlblk))
+ raise_softirq(RCU_SOFTIRQ);
+}
+
+/*
+ * Check to see if the scheduling-clock interrupt came from an extended
+ * quiescent state, and, if so, tell RCU about it.
+ */
+void rcu_check_callbacks(int cpu, int user)
+{
+ if (!rcu_needs_cpu(0))
+ return; /* RCU doesn't need anything to be done. */
+ if (user ||
+ (idle_cpu(cpu) &&
+ !in_softirq() &&
+ hardirq_count() <= (1 << HARDIRQ_SHIFT)))
+ rcu_qsctr_inc(cpu);
+ else if (!in_softirq())
+ rcu_bh_qsctr_inc(cpu);
+}
+
+/*
+ * Helper function for rcu_process_callbacks() that operates on the
+ * specified rcu_ctrlkblk structure.
+ */
+static void __rcu_process_callbacks(struct rcu_ctrlblk *rcp)
+{
+ unsigned long flags;
+ struct rcu_head *next, *list;
+
+ /* If no RCU callbacks ready to invoke, just return. */
+ if (&rcp->rcucblist == rcp->donetail)
+ return;
+
+ /* Move the ready-to-invoke callbacks to a local list. */
+ local_irq_save(flags);
+ rcp->completed++;
+ list = rcp->rcucblist;
+ rcp->rcucblist = *rcp->donetail;
+ *rcp->donetail = NULL;
+ if (rcp->curtail == rcp->donetail)
+ rcp->curtail = &rcp->rcucblist;
+ rcp->donetail = &rcp->rcucblist;
+ local_irq_restore(flags);
+
+ /* Invoke the callbacks on the local list. */
+ while (list) {
+ next = list->next;
+ prefetch(next);
+ list->func(list);
+ list = next;
+ }
+}
+
+/*
+ * Invoke any callbacks whose grace period has completed.
+ */
+static void rcu_process_callbacks(struct softirq_action *unused)
+{
+ __rcu_process_callbacks(&rcu_ctrlblk);
+ __rcu_process_callbacks(&rcu_bh_ctrlblk);
+}
+
+/*
+ * Wait for a grace period to elapse. But it is illegal to invoke
+ * synchronize_rcu() from within an RCU read-side critical section.
+ * Therefore, any legal call to synchronize_rcu() is a quiescent
+ * state, and so on a UP system, synchronize_rcu() need do nothing.
+ *
+ * Cool, huh? (Due to Josh Triplett.)
+ *
+ * However, we do update the grace-period counter to prevent rcutorture
+ * from hammering us.
+ */
+void synchronize_rcu(void)
+{
+ unsigned long flags;
+
+ local_irq_save(flags);
+ rcu_ctrlblk.completed++;
+ local_irq_restore(flags);
+}
+EXPORT_SYMBOL(synchronize_rcu);
+
+/*
+ * Helper function for call_rcu() and call_rcu_bh().
+ */
+static void __call_rcu(struct rcu_head *head,
+ void (*func)(struct rcu_head *rcu),
+ struct rcu_ctrlblk *rcp)
+{
+ unsigned long flags;
+
+ head->func = func;
+ head->next = NULL;
+ local_irq_save(flags);
+ *rcp->curtail = head;
+ rcp->curtail = &head->next;
+ local_irq_restore(flags);
+}
+
+/*
+ * Post an RCU callback to be invoked after the end of an RCU grace
+ * period. But since we have but one CPU, that would be after any
+ * quiescent state.
+ */
+void call_rcu(struct rcu_head *head,
+ void (*func)(struct rcu_head *rcu))
+{
+ __call_rcu(head, func, &rcu_ctrlblk);
+}
+EXPORT_SYMBOL(call_rcu);
+
+/*
+ * Post an RCU bottom-half callback to be invoked after any subsequent
+ * quiescent state.
+ */
+void call_rcu_bh(struct rcu_head *head,
+ void (*func)(struct rcu_head *rcu))
+{
+ __call_rcu(head, func, &rcu_bh_ctrlblk);
+}
+EXPORT_SYMBOL(call_rcu_bh);
+
+void __rcu_init(void)
+{
+ open_softirq(RCU_SOFTIRQ, rcu_process_callbacks);
+}