Skip to contentSkip to conversation

membarrier: Provide register expedited private command

RFC PATCH Superseded Mathieu Desnoyers 3 participants 14 messages Sep 19, 2017 – Sep 25, 2017

Subsystems (8)
VFS PowerPC Kselftest MM core +4 more subsystems

v3 of 5 Superseded Latest is v5 →

lore
Share & tools
lore

Mathieu Desnoyers

author maintainer
#0 lore raw

RFC PATCH v3 1/2 membarrier: Provide register expedited private command

Provide a new command allowing processes to register their intent to use the private expedited command.

This allows PowerPC to skip the full memory barrier in switch_mm(), and only issue the barrier when scheduling into a task belonging to a process that has registered to use expedited private.

Processes are now required to register before using MEMBARRIER_CMD_PRIVATE_EXPEDITED, otherwise that command returns EPERM.

Changes since v1:
- Use test_ti_thread_flag(next, ...) instead of test_thread_flag() in powerpc membarrier_arch_sched_in(), given that we want to specifically check the next thread state.
- Add missing ARCH_HAS_MEMBARRIER_HOOKS in Kconfig.
- Use task_thread_info() to pass thread_info from task to *_ti_thread_flag().

Changes since v2:
- Move membarrier_arch_sched_in() call to finish_task_switch().
- Check for NULL t->mm in membarrier_arch_fork().
- Use membarrier_sched_in() in generic code, which invokes the arch-specific membarrier_arch_sched_in(). This fixes allnoconfig build on PowerPC.
- Move asm/membarrier.h include under CONFIG_MEMBARRIER, fixing allnoconfig build on PowerPC.
- Build and runtime tested on PowerPC.

CC Peter Zijlstra <[email protected]>
CC Paul E. McKenney <[email protected]>
CC Boqun Feng <[email protected]>
CC Andrew Hunter <[email protected]>
CC Maged Michael <[email protected]>
CC Avi Kivity <[email protected]>
CC Benjamin Herrenschmidt <[email protected]>
CC Paul Mackerras <[email protected]>
CC Michael Ellerman <[email protected]>
CC Dave Watson <[email protected]>
CC Alan Stern <[email protected]>
CC Will Deacon <[email protected]>
CC Andy Lutomirski <[email protected]>
---
 MAINTAINERS                            |  2 ++
 arch/powerpc/Kconfig                   |  1 +
 arch/powerpc/include/asm/membarrier.h  | 40 +++++++++++++++++++++++++
 arch/powerpc/include/asm/thread_info.h |  3 ++
 arch/powerpc/kernel/Makefile           |  2 ++
 arch/powerpc/kernel/membarrier.c       | 45 ++++++++++++++++++++++++++++
 fs/exec.c                              |  1 +
 include/linux/mm_types.h               |  3 ++
 include/linux/sched/mm.h               | 55 ++++++++++++++++++++++++++++++++++
 include/uapi/linux/membarrier.h        | 23 +++++++++-----
 init/Kconfig                           |  3 ++
 kernel/fork.c                          |  2 ++
 kernel/sched/core.c                    | 16 ++--------
 kernel/sched/membarrier.c              | 25 +++++++++++++---
 14 files changed, 197 insertions(+), 24 deletions(-)
 create mode 100644 arch/powerpc/include/asm/membarrier.h
 create mode 100644 arch/powerpc/kernel/membarrier.c
MAINTAINERS +2 -0
index 2281af4b41b6..fe6d34d8a2e0 100644
@@ -8805,6 +8805,8 @@ L:	[email protected]
 S:	Supported
 F:	kernel/sched/membarrier.c
 F:	include/uapi/linux/membarrier.h
+F:	arch/powerpc/kernel/membarrier.c
+F:	arch/powerpc/include/asm/membarrier.h
 
 MEMORY MANAGEMENT
 L:	[email protected]
arch/powerpc/Kconfig +1 -0
index 809c468edab1..6f44c5f74f71 100644
@@ -138,6 +138,7 @@ config PPC
 	select ARCH_HAS_ELF_RANDOMIZE
 	select ARCH_HAS_FORTIFY_SOURCE
 	select ARCH_HAS_GCOV_PROFILE_ALL
+	select ARCH_HAS_MEMBARRIER_HOOKS
 	select ARCH_HAS_SCALED_CPUTIME		if VIRT_CPU_ACCOUNTING_NATIVE
 	select ARCH_HAS_SG_CHAIN
 	select ARCH_HAS_TICK_BROADCAST		if GENERIC_CLOCKEVENTS_BROADCAST
arch/powerpc/include/asm/membarrier.h new file +40 -0
new file mode 100644
index 000000000000..ee208455674d
@@ -0,0 +1,40 @@
+#ifndef _ASM_POWERPC_MEMBARRIER_H
+#define _ASM_POWERPC_MEMBARRIER_H
+
+static inline void membarrier_arch_sched_in(struct task_struct *prev,
+		struct task_struct *next)
+{
+	/*
+	 * Only need the full barrier when switching between processes.
+	 */
+	if (likely(!test_ti_thread_flag(task_thread_info(next),
+			TIF_MEMBARRIER_PRIVATE_EXPEDITED)
+				|| prev->mm == next->mm))
+		return;
+
+	/*
+	 * The membarrier system call requires a full memory barrier
+	 * after storing to rq->curr, before going back to user-space.
+	 */
+	smp_mb();
+}
+static inline void membarrier_arch_fork(struct task_struct *t,
+		unsigned long clone_flags)
+{
+	/*
+	 * Coherence of TIF_MEMBARRIER_PRIVATE_EXPEDITED against thread
+	 * fork is protected by siglock. membarrier_arch_fork is called
+	 * with siglock held.
+	 */
+	if (t->mm && t->mm->membarrier_private_expedited)
+		set_ti_thread_flag(task_thread_info(t),
+				TIF_MEMBARRIER_PRIVATE_EXPEDITED);
+}
+static inline void membarrier_arch_execve(struct task_struct *t)
+{
+	clear_ti_thread_flag(task_thread_info(t),
+			TIF_MEMBARRIER_PRIVATE_EXPEDITED);
+}
+void membarrier_arch_register_private_expedited(struct task_struct *t);
+
+#endif /* _ASM_POWERPC_MEMBARRIER_H */
arch/powerpc/include/asm/thread_info.h +3 -0
index a941cc6fc3e9..2a208487724b 100644
@@ -100,6 +100,7 @@ static inline struct thread_info *current_thread_info(void)
 #if defined(CONFIG_PPC64)
 #define TIF_ELF2ABI		18	/* function descriptors must die! */
 #endif
+#define TIF_MEMBARRIER_PRIVATE_EXPEDITED	19	/* membarrier */
 
 /* as above, but as bit values */
 #define _TIF_SYSCALL_TRACE	(1<<TIF_SYSCALL_TRACE)
@@ -119,6 +120,8 @@ static inline struct thread_info *current_thread_info(void)
 #define _TIF_SYSCALL_TRACEPOINT	(1<<TIF_SYSCALL_TRACEPOINT)
 #define _TIF_EMULATE_STACK_STORE	(1<<TIF_EMULATE_STACK_STORE)
 #define _TIF_NOHZ		(1<<TIF_NOHZ)
+#define _TIF_MEMBARRIER_PRIVATE_EXPEDITED \
+				(1<<TIF_MEMBARRIER_PRIVATE_EXPEDITED)
 #define _TIF_SYSCALL_DOTRACE	(_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | \
 				 _TIF_SECCOMP | _TIF_SYSCALL_TRACEPOINT | \
 				 _TIF_NOHZ)
arch/powerpc/kernel/Makefile +2 -0
index 91960f83039c..2dd4b9e3313a 100644
@@ -135,6 +135,8 @@ endif
 obj-$(CONFIG_EPAPR_PARAVIRT)	+= epapr_paravirt.o epapr_hcalls.o
 obj-$(CONFIG_KVM_GUEST)		+= kvm.o kvm_emul.o
 
+obj-$(CONFIG_MEMBARRIER)	+= membarrier.o
+
 # Disable GCOV & sanitizers in odd or sensitive code
 GCOV_PROFILE_prom_init.o := n
 UBSAN_SANITIZE_prom_init.o := n
arch/powerpc/kernel/membarrier.c new file +45 -0

45 lines not on this page · raw message

fs/exec.c +1 -0
index ac34d9724684..b2448f2731b3 100644
@@ -1802,6 +1802,7 @@ static int do_execveat_common(int fd, struct filename *filename,
 	/* execve succeeded */
 	current->fs->in_exec = 0;
 	current->in_execve = 0;
+	membarrier_execve(current);
 	acct_update_integrals(current);
 	task_numa_free(current);
 	free_bprm(bprm);
include/linux/mm_types.h +3 -0
index 46f4ecf5479a..5e0fe8ce053b 100644
@@ -445,6 +445,9 @@ struct mm_struct {
 	unsigned long flags; /* Must use atomic bitops to access the bits */
 
 	struct core_state *core_state; /* coredumping support */
+#ifdef CONFIG_MEMBARRIER
+	int membarrier_private_expedited;
+#endif
 #ifdef CONFIG_AIO
 	spinlock_t			ioctx_lock;
 	struct kioctx_table __rcu	*ioctx_table;
include/linux/sched/mm.h +55 -0

59 lines not on this page · raw message

include/uapi/linux/membarrier.h +16 -7
index 6d47b3249d8a..4e01ad7ffe98 100644
@@ -52,21 +52,30 @@
  *                          (non-running threads are de facto in such a
  *                          state). This only covers threads from the
  *                          same processes as the caller thread. This
- *                          command returns 0. The "expedited" commands
- *                          complete faster than the non-expedited ones,
- *                          they never block, but have the downside of
- *                          causing extra overhead.
+ *                          command returns 0 on success. The
+ *                          "expedited" commands complete faster than
+ *                          the non-expedited ones, they never block,
+ *                          but have the downside of causing extra
+ *                          overhead. A process needs to register its
+ *                          intent to use the private expedited command
+ *                          prior to using it, otherwise this command
+ *                          returns -EPERM.
+ * @MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED:
+ *                          Register the process intent to use
+ *                          MEMBARRIER_CMD_PRIVATE_EXPEDITED. Always
+ *                          returns 0.
  *
  * Command to be passed to the membarrier system call. The commands need to
  * be a single bit each, except for MEMBARRIER_CMD_QUERY which is assigned to
  * the value 0.
  */
 enum membarrier_cmd {
-	MEMBARRIER_CMD_QUERY			= 0,
-	MEMBARRIER_CMD_SHARED			= (1 << 0),
+	MEMBARRIER_CMD_QUERY				= 0,
+	MEMBARRIER_CMD_SHARED				= (1 << 0),
 	/* reserved for MEMBARRIER_CMD_SHARED_EXPEDITED (1 << 1) */
 	/* reserved for MEMBARRIER_CMD_PRIVATE (1 << 2) */
-	MEMBARRIER_CMD_PRIVATE_EXPEDITED	= (1 << 3),
+	MEMBARRIER_CMD_PRIVATE_EXPEDITED		= (1 << 3),
+	MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED	= (1 << 4),
 };
 
 #endif /* _UAPI_LINUX_MEMBARRIER_H */
init/Kconfig +3 -0
index 78cb2461012e..a3dc6a66f0d1 100644
@@ -1400,6 +1400,9 @@ config MEMBARRIER
 
 	  If unsure, say Y.
 
+config ARCH_HAS_MEMBARRIER_HOOKS
+	bool
+
 config EMBEDDED
 	bool "Embedded system"
 	option allnoconfig_y
kernel/fork.c +2 -0
index 10646182440f..bd4a93915e08 100644
@@ -1840,6 +1840,8 @@ static __latent_entropy struct task_struct *copy_process(
 	 */
 	copy_seccomp(p);
 
+	membarrier_fork(p, clone_flags);
+
 	/*
 	 * Process group and session signals need to be delivered to just the
 	 * parent before the fork or both the parent and the child after the
kernel/sched/core.c +3 -13
index 7977b25acf54..08095bb1cfe6 100644
@@ -2643,17 +2643,8 @@ static struct rq *finish_task_switch(struct task_struct *prev)
 	 */
 	prev_state = prev->state;
 	vtime_task_switch(prev);
+	membarrier_sched_in(prev, current);
 	perf_event_task_sched_in(prev, current);
-	/*
-	 * The membarrier system call requires a full memory barrier
-	 * after storing to rq->curr, before going back to user-space.
-	 *
-	 * TODO: This smp_mb__after_unlock_lock can go away if PPC end
-	 * up adding a full barrier to switch_mm(), or we should figure
-	 * out if a smp_mb__after_unlock_lock is really the proper API
-	 * to use.
-	 */
-	smp_mb__after_unlock_lock();
 	finish_lock_switch(rq, prev);
 	finish_arch_post_lock_switch();
 
@@ -3363,9 +3354,8 @@ static void __sched notrace __schedule(bool preempt)
 		 * care of this barrier. For weakly ordered machines for
 		 * which spin_unlock() acts as a RELEASE barrier (only
 		 * arm64 and PowerPC), arm64 has a full barrier in
-		 * switch_to(), and PowerPC has
-		 * smp_mb__after_unlock_lock() before
-		 * finish_lock_switch().
+		 * switch_to(), and PowerPC has a full barrier in
+		 * membarrier_arch_sched_in().
 		 */
 		++*switch_count;
 
kernel/sched/membarrier.c +21 -4
index a92fddc22747..00a2618a36ba 100644
@@ -26,21 +26,25 @@
  * except MEMBARRIER_CMD_QUERY.
  */
 #define MEMBARRIER_CMD_BITMASK	\
-	(MEMBARRIER_CMD_SHARED | MEMBARRIER_CMD_PRIVATE_EXPEDITED)
+	(MEMBARRIER_CMD_SHARED | MEMBARRIER_CMD_PRIVATE_EXPEDITED	\
+	| MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED)
 
 static void ipi_mb(void *info)
 {
 	smp_mb();	/* IPIs should be serializing but paranoid. */
 }
 
-static void membarrier_private_expedited(void)
+static int membarrier_private_expedited(void)
 {
 	int cpu;
 	bool fallback = false;
 	cpumask_var_t tmpmask;
 
+	if (!current->mm->membarrier_private_expedited)
+		return -EPERM;
+
 	if (num_online_cpus() == 1)
-		return;
+		return 0;
 
 	/*
 	 * Matches memory barriers around rq->curr modification in
@@ -94,6 +98,17 @@ static void membarrier_private_expedited(void)
 	 * rq->curr modification in scheduler.
 	 */
 	smp_mb();	/* exit from system call is not a mb */
+	return 0;
+}
+
+static void membarrier_register_private_expedited(void)
+{
+	struct task_struct *p = current;
+
+	if (READ_ONCE(p->mm->membarrier_private_expedited))
+		return;
+	WRITE_ONCE(p->mm->membarrier_private_expedited, 1);
+	membarrier_arch_register_private_expedited(p);
 }
 
 /**
@@ -144,7 +159,9 @@ SYSCALL_DEFINE2(membarrier, int, cmd, int, flags)
 			synchronize_sched();
 		return 0;
 	case MEMBARRIER_CMD_PRIVATE_EXPEDITED:
-		membarrier_private_expedited();
+		return membarrier_private_expedited();
+	case MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED:
+		membarrier_register_private_expedited();
 		return 0;
 	default:
 		return -EINVAL;
Signature · 1 line
2.11.0
author maintainer
#1 lore raw

RFC PATCH 2/2 membarrier: selftest: Test private expedited cmd

Test the new MEMBARRIER_CMD_PRIVATE_EXPEDITED and MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED commands.

Add checks expecting specific error values on system calls expected to fail.

CC Peter Zijlstra <[email protected]>
CC Paul E. McKenney <[email protected]>
CC Boqun Feng <[email protected]>
CC Andrew Hunter <[email protected]>
CC Maged Michael <[email protected]>
CC Avi Kivity <[email protected]>
CC Benjamin Herrenschmidt <[email protected]>
CC Paul Mackerras <[email protected]>
CC Michael Ellerman <[email protected]>
CC Dave Watson <[email protected]>
CC Alan Stern <[email protected]>
CC Will Deacon <[email protected]>
CC Andy Lutomirski <[email protected]>
---
 .../testing/selftests/membarrier/membarrier_test.c | 109 ++++++++++++++++++---
 1 file changed, 94 insertions(+), 15 deletions(-)
tools/testing/selftests/membarrier/membarrier_test.c +94 -15
index 21399fcf1a59..f85657374b59 100644
@@ -15,49 +15,119 @@ static int sys_membarrier(int cmd, int flags)
 static int test_membarrier_cmd_fail(void)
 {
 	int cmd = -1, flags = 0;
+	const char *test_name = "sys membarrier invalid command";
 
 	if (sys_membarrier(cmd, flags) != -1) {
 		ksft_exit_fail_msg(
-			"sys membarrier invalid command test: command = %d, flags = %d. Should fail, but passed\n",
-			cmd, flags);
+			"%s test: command = %d, flags = %d. Should fail, but passed\n",
+			test_name, cmd, flags);
+	}
+	if (errno != EINVAL) {
+		ksft_exit_fail_msg(
+			"%s test: flags = %d. Should return (%d: \"%s\"), but returned (%d: \"%s\").\n",
+			test_name, flags, EINVAL, strerror(EINVAL),
+			errno, strerror(errno));
 	}
 
 	ksft_test_result_pass(
-		"sys membarrier invalid command test: command = %d, flags = %d. Failed as expected\n",
-		cmd, flags);
+		"%s test: command = %d, flags = %d, errno = %d. Failed as expected\n",
+		test_name, cmd, flags, errno);
 	return 0;
 }
 
 static int test_membarrier_flags_fail(void)
 {
 	int cmd = MEMBARRIER_CMD_QUERY, flags = 1;
+	const char *test_name = "sys membarrier MEMBARRIER_CMD_QUERY invalid flags";
 
 	if (sys_membarrier(cmd, flags) != -1) {
 		ksft_exit_fail_msg(
-			"sys membarrier MEMBARRIER_CMD_QUERY invalid flags test: flags = %d. Should fail, but passed\n",
-			flags);
+			"%s test: flags = %d. Should fail, but passed\n",
+			test_name, flags);
+	}
+	if (errno != EINVAL) {
+		ksft_exit_fail_msg(
+			"%s test: flags = %d. Should return (%d: \"%s\"), but returned (%d: \"%s\").\n",
+			test_name, flags, EINVAL, strerror(EINVAL),
+			errno, strerror(errno));
 	}
 
 	ksft_test_result_pass(
-		"sys membarrier MEMBARRIER_CMD_QUERY invalid flags test: flags = %d. Failed as expected\n",
-		flags);
+		"%s test: flags = %d, errno = %d. Failed as expected\n",
+		test_name, flags, errno);
 	return 0;
 }
 
-static int test_membarrier_success(void)
+static int test_membarrier_shared_success(void)
 {
 	int cmd = MEMBARRIER_CMD_SHARED, flags = 0;
-	const char *test_name = "sys membarrier MEMBARRIER_CMD_SHARED\n";
+	const char *test_name = "sys membarrier MEMBARRIER_CMD_SHARED";
+
+	if (sys_membarrier(cmd, flags) != 0) {
+		ksft_exit_fail_msg(
+			"%s test: flags = %d, errno = %d\n",
+			test_name, flags, errno);
+	}
+
+	ksft_test_result_pass(
+		"%s test: flags = %d\n", test_name, flags);
+	return 0;
+}
+
+static int test_membarrier_private_expedited_fail(void)
+{
+	int cmd = MEMBARRIER_CMD_PRIVATE_EXPEDITED, flags = 0;
+	const char *test_name = "sys membarrier MEMBARRIER_CMD_PRIVATE_EXPEDITED not registered failure";
+
+	if (sys_membarrier(cmd, flags) != -1) {
+		ksft_exit_fail_msg(
+			"%s test: flags = %d. Should fail, but passed\n",
+			test_name, flags);
+	}
+	if (errno != EPERM) {
+		ksft_exit_fail_msg(
+			"%s test: flags = %d. Should return (%d: \"%s\"), but returned (%d: \"%s\").\n",
+			test_name, flags, EPERM, strerror(EPERM),
+			errno, strerror(errno));
+	}
+
+	ksft_test_result_pass(
+		"%s test: flags = %d, errno = %d\n",
+		test_name, flags, errno);
+	return 0;
+}
+
+static int test_membarrier_register_private_expedited_success(void)
+{
+	int cmd = MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED, flags = 0;
+	const char *test_name = "sys membarrier MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED";
 
 	if (sys_membarrier(cmd, flags) != 0) {
 		ksft_exit_fail_msg(
-			"sys membarrier MEMBARRIER_CMD_SHARED test: flags = %d\n",
-			flags);
+			"%s test: flags = %d, errno = %d\n",
+			test_name, flags, errno);
 	}
 
 	ksft_test_result_pass(
-		"sys membarrier MEMBARRIER_CMD_SHARED test: flags = %d\n",
-		flags);
+		"%s test: flags = %d\n",
+		test_name, flags);
+	return 0;
+}
+
+static int test_membarrier_private_expedited_success(void)
+{
+	int cmd = MEMBARRIER_CMD_PRIVATE_EXPEDITED, flags = 0;
+	const char *test_name = "sys membarrier MEMBARRIER_CMD_PRIVATE_EXPEDITED";
+
+	if (sys_membarrier(cmd, flags) != 0) {
+		ksft_exit_fail_msg(
+			"%s test: flags = %d, errno = %d\n",
+			test_name, flags, errno);
+	}
+
+	ksft_test_result_pass(
+		"%s test: flags = %d\n",
+		test_name, flags);
 	return 0;
 }
 
@@ -71,7 +141,16 @@ static int test_membarrier(void)
 	status = test_membarrier_flags_fail();
 	if (status)
 		return status;
-	status = test_membarrier_success();
+	status = test_membarrier_shared_success();
+	if (status)
+		return status;
+	status = test_membarrier_private_expedited_fail();
+	if (status)
+		return status;
+	status = test_membarrier_register_private_expedited_success();
+	if (status)
+		return status;
+	status = test_membarrier_private_expedited_success();
 	if (status)
 		return status;
 	return 0;
Signature · 1 line
2.11.0
#2 lore raw
In reply to Mathieu Desnoyers

Hi Mathieu,

Quoted 10 lines · Mathieu Desnoyers

Provide a new command allowing processes to register their intent to use the private expedited command.

This allows PowerPC to skip the full memory barrier in switch_mm(), and only issue the barrier when scheduling into a task belonging to a process that has registered to use expedited private.

Processes are now required to register before using MEMBARRIER_CMD_PRIVATE_EXPEDITED, otherwise that command returns EPERM.

Sorry I'm late for the party, but I couldn't stop thinking whether we could avoid the register thing at all, because the registering makes sys_membarrier() more complex(both for the interface and the implementation). So how about we trade-off a little bit by taking some(not all) the rq->locks?

The idea is in membarrier_private_expedited(), we go through all ->curr on each CPU and

1) If it's a userspace task and its ->mm is matched, we send an ipi

2) If it's a kernel task, we skip

(Because there will be a smp_mb() implied by mmdrop(), when it switchs to userspace task).

3) If it's a userspace task and its ->mm is not matched, we take

	the corresponding rq->lock and check rq->curr again, if its ->mm
	matched, we send an ipi, otherwise we do nothing.

	(Because if we observe rq->curr is not matched with rq->lock
	held, when a task having matched ->mm schedules in, the rq->lock
	pairing along with the smp_mb__after_spinlock() will guarantee
	it observes all memory ops before sys_membarrir()).

membarrier_private_expedited() will look like this if we choose this way:

void membarrier_private_expedited()

{
	int cpu;
	bool fallback = false;
	cpumask_var_t tmpmask;
	struct rq_flags rf;


	if (num_online_cpus() == 1)
		return;

	smp_mb();

	if (!zalloc_cpumask_var(&tmpmask, GFP_NOWAIT)) {
		/* Fallback for OOM. */
		fallback = true;
	}

	cpus_read_lock();
	for_each_online_cpu(cpu) {
		struct task_struct *p;

		if (cpu == raw_smp_processor_id())
			continue;

		rcu_read_lock();
		p = task_rcu_dereference(&cpu_rq(cpu)->curr);

		if (!p) {
			rcu_read_unlock();
			continue;
		}

		if (p->mm == current->mm) {
			if (!fallback)
				__cpumask_set_cpu(cpu, tmpmask);
			else
				smp_call_function_single(cpu, ipi_mb, NULL, 1);
		}

		if (p->mm == current->mm || !p->mm) {
			rcu_read_unlock();
			continue;
		}

		rcu_read_unlock();
		
		/*
		 * This should be a arch-specific code, as we don't
		 * need it at else place other than some archs without
		 * a smp_mb() in switch_mm() (i.e. powerpc)
		 */
		rq_lock_irq(cpu_rq(cpu), &rf);
		if (p->mm == current->mm) {
			if (!fallback)
				__cpumask_set_cpu(cpu, tmpmask);
			else
				smp_call_function_single(cpu, ipi_mb, NULL, 1);
		}
		rq_unlock_irq(cpu_rq(cpu), &rf);
	}
	if (!fallback) {
		smp_call_function_many(tmpmask, ipi_mb, NULL, 1);
		free_cpumask_var(tmpmask);
	}
	cpus_read_unlock();

	smp_mb();
}

Thoughts?

Regards,
Boqun

Quoted 468 lines

Changes since v1:
- Use test_ti_thread_flag(next, ...) instead of test_thread_flag() in powerpc membarrier_arch_sched_in(), given that we want to specifically check the next thread state.
- Add missing ARCH_HAS_MEMBARRIER_HOOKS in Kconfig.
- Use task_thread_info() to pass thread_info from task to *_ti_thread_flag().

Changes since v2:
- Move membarrier_arch_sched_in() call to finish_task_switch().
- Check for NULL t->mm in membarrier_arch_fork().
- Use membarrier_sched_in() in generic code, which invokes the arch-specific membarrier_arch_sched_in(). This fixes allnoconfig build on PowerPC.
- Move asm/membarrier.h include under CONFIG_MEMBARRIER, fixing allnoconfig build on PowerPC.
- Build and runtime tested on PowerPC.

CC Peter Zijlstra <[email protected]>
CC Paul E. McKenney <[email protected]>
CC Boqun Feng <[email protected]>
CC Andrew Hunter <[email protected]>
CC Maged Michael <[email protected]>
CC Avi Kivity <[email protected]>
CC Benjamin Herrenschmidt <[email protected]>
CC Paul Mackerras <[email protected]>
CC Michael Ellerman <[email protected]>
CC Dave Watson <[email protected]>
CC Alan Stern <[email protected]>
CC Will Deacon <[email protected]>
CC Andy Lutomirski <[email protected]>
---
 MAINTAINERS                            |  2 ++
 arch/powerpc/Kconfig                   |  1 +
 arch/powerpc/include/asm/membarrier.h  | 40 +++++++++++++++++++++++++
 arch/powerpc/include/asm/thread_info.h |  3 ++
 arch/powerpc/kernel/Makefile           |  2 ++
 arch/powerpc/kernel/membarrier.c       | 45 ++++++++++++++++++++++++++++
 fs/exec.c                              |  1 +
 include/linux/mm_types.h               |  3 ++
 include/linux/sched/mm.h               | 55 ++++++++++++++++++++++++++++++++++
 include/uapi/linux/membarrier.h        | 23 +++++++++-----
 init/Kconfig                           |  3 ++
 kernel/fork.c                          |  2 ++
 kernel/sched/core.c                    | 16 ++--------
 kernel/sched/membarrier.c              | 25 +++++++++++++---
 14 files changed, 197 insertions(+), 24 deletions(-)
 create mode 100644 arch/powerpc/include/asm/membarrier.h
 create mode 100644 arch/powerpc/kernel/membarrier.c
MAINTAINERS +2 -0
index 2281af4b41b6..fe6d34d8a2e0 100644
@@ -8805,6 +8805,8 @@ L:	[email protected]
 S:	Supported
 F:	kernel/sched/membarrier.c
 F:	include/uapi/linux/membarrier.h
+F:	arch/powerpc/kernel/membarrier.c
+F:	arch/powerpc/include/asm/membarrier.h
 
 MEMORY MANAGEMENT
 L:	[email protected]
arch/powerpc/Kconfig +1 -0
index 809c468edab1..6f44c5f74f71 100644
@@ -138,6 +138,7 @@ config PPC
 	select ARCH_HAS_ELF_RANDOMIZE
 	select ARCH_HAS_FORTIFY_SOURCE
 	select ARCH_HAS_GCOV_PROFILE_ALL
+	select ARCH_HAS_MEMBARRIER_HOOKS
 	select ARCH_HAS_SCALED_CPUTIME		if VIRT_CPU_ACCOUNTING_NATIVE
 	select ARCH_HAS_SG_CHAIN
 	select ARCH_HAS_TICK_BROADCAST		if GENERIC_CLOCKEVENTS_BROADCAST
arch/powerpc/include/asm/membarrier.h new file +40 -0
new file mode 100644
index 000000000000..ee208455674d
@@ -0,0 +1,40 @@
+#ifndef _ASM_POWERPC_MEMBARRIER_H
+#define _ASM_POWERPC_MEMBARRIER_H
+
+static inline void membarrier_arch_sched_in(struct task_struct *prev,
+		struct task_struct *next)
+{
+	/*
+	 * Only need the full barrier when switching between processes.
+	 */
+	if (likely(!test_ti_thread_flag(task_thread_info(next),
+			TIF_MEMBARRIER_PRIVATE_EXPEDITED)
+				|| prev->mm == next->mm))
+		return;
+
+	/*
+	 * The membarrier system call requires a full memory barrier
+	 * after storing to rq->curr, before going back to user-space.
+	 */
+	smp_mb();
+}
+static inline void membarrier_arch_fork(struct task_struct *t,
+		unsigned long clone_flags)
+{
+	/*
+	 * Coherence of TIF_MEMBARRIER_PRIVATE_EXPEDITED against thread
+	 * fork is protected by siglock. membarrier_arch_fork is called
+	 * with siglock held.
+	 */
+	if (t->mm && t->mm->membarrier_private_expedited)
+		set_ti_thread_flag(task_thread_info(t),
+				TIF_MEMBARRIER_PRIVATE_EXPEDITED);
+}
+static inline void membarrier_arch_execve(struct task_struct *t)
+{
+	clear_ti_thread_flag(task_thread_info(t),
+			TIF_MEMBARRIER_PRIVATE_EXPEDITED);
+}
+void membarrier_arch_register_private_expedited(struct task_struct *t);
+
+#endif /* _ASM_POWERPC_MEMBARRIER_H */
arch/powerpc/include/asm/thread_info.h +3 -0
index a941cc6fc3e9..2a208487724b 100644
@@ -100,6 +100,7 @@ static inline struct thread_info *current_thread_info(void)
 #if defined(CONFIG_PPC64)
 #define TIF_ELF2ABI		18	/* function descriptors must die! */
 #endif
+#define TIF_MEMBARRIER_PRIVATE_EXPEDITED	19	/* membarrier */
 
 /* as above, but as bit values */
 #define _TIF_SYSCALL_TRACE	(1<<TIF_SYSCALL_TRACE)
@@ -119,6 +120,8 @@ static inline struct thread_info *current_thread_info(void)
 #define _TIF_SYSCALL_TRACEPOINT	(1<<TIF_SYSCALL_TRACEPOINT)
 #define _TIF_EMULATE_STACK_STORE	(1<<TIF_EMULATE_STACK_STORE)
 #define _TIF_NOHZ		(1<<TIF_NOHZ)
+#define _TIF_MEMBARRIER_PRIVATE_EXPEDITED \
+				(1<<TIF_MEMBARRIER_PRIVATE_EXPEDITED)
 #define _TIF_SYSCALL_DOTRACE	(_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | \
 				 _TIF_SECCOMP | _TIF_SYSCALL_TRACEPOINT | \
 				 _TIF_NOHZ)
arch/powerpc/kernel/Makefile +2 -0
index 91960f83039c..2dd4b9e3313a 100644
@@ -135,6 +135,8 @@ endif
 obj-$(CONFIG_EPAPR_PARAVIRT)	+= epapr_paravirt.o epapr_hcalls.o
 obj-$(CONFIG_KVM_GUEST)		+= kvm.o kvm_emul.o
 
+obj-$(CONFIG_MEMBARRIER)	+= membarrier.o
+
 # Disable GCOV & sanitizers in odd or sensitive code
 GCOV_PROFILE_prom_init.o := n
 UBSAN_SANITIZE_prom_init.o := n
arch/powerpc/kernel/membarrier.c new file +45 -0
new file mode 100644
index 000000000000..b0d79a5f5981
@@ -0,0 +1,45 @@
+/*
+ * Copyright (C) 2010-2017 Mathieu Desnoyers <[email protected]>
+ *
+ * membarrier system call - PowerPC architecture code
+ *
+ * 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.
+ */
+
+#include <linux/sched/mm.h>
+#include <linux/sched/signal.h>
+#include <linux/thread_info.h>
+#include <linux/spinlock.h>
+#include <linux/rcupdate.h>
+
+void membarrier_arch_register_private_expedited(struct task_struct *p)
+{
+	struct task_struct *t;
+
+	if (get_nr_threads(p) == 1) {
+		set_thread_flag(TIF_MEMBARRIER_PRIVATE_EXPEDITED);
+		return;
+	}
+	/*
+	 * Coherence of TIF_MEMBARRIER_PRIVATE_EXPEDITED against thread
+	 * fork is protected by siglock.
+	 */
+	spin_lock(&p->sighand->siglock);
+	for_each_thread(p, t)
+		set_ti_thread_flag(task_thread_info(t),
+				TIF_MEMBARRIER_PRIVATE_EXPEDITED);
+	spin_unlock(&p->sighand->siglock);
+	/*
+	 * Ensure all future scheduler executions will observe the new
+	 * thread flag state for this process.
+	 */
+	synchronize_sched();
+}
fs/exec.c +1 -0
index ac34d9724684..b2448f2731b3 100644
@@ -1802,6 +1802,7 @@ static int do_execveat_common(int fd, struct filename *filename,
 	/* execve succeeded */
 	current->fs->in_exec = 0;
 	current->in_execve = 0;
+	membarrier_execve(current);
 	acct_update_integrals(current);
 	task_numa_free(current);
 	free_bprm(bprm);
include/linux/mm_types.h +3 -0
index 46f4ecf5479a..5e0fe8ce053b 100644
@@ -445,6 +445,9 @@ struct mm_struct {
 	unsigned long flags; /* Must use atomic bitops to access the bits */
 
 	struct core_state *core_state; /* coredumping support */
+#ifdef CONFIG_MEMBARRIER
+	int membarrier_private_expedited;
+#endif
 #ifdef CONFIG_AIO
 	spinlock_t			ioctx_lock;
 	struct kioctx_table __rcu	*ioctx_table;
include/linux/sched/mm.h +55 -0
index 766cc47c4d7c..b88f85c8474e 100644
@@ -210,4 +210,59 @@ static inline void memalloc_noreclaim_restore(unsigned int flags)
 	current->flags = (current->flags & ~PF_MEMALLOC) | flags;
 }
 
+#ifdef CONFIG_MEMBARRIER
+
+#ifdef CONFIG_ARCH_HAS_MEMBARRIER_HOOKS
+#include <asm/membarrier.h>
+#else
+static inline void membarrier_arch_sched_in(struct task_struct *prev,
+		struct task_struct *next)
+{
+}
+static inline void membarrier_arch_fork(struct task_struct *t,
+		unsigned long clone_flags)
+{
+}
+static inline void membarrier_arch_execve(struct task_struct *t)
+{
+}
+static inline void membarrier_arch_register_private_expedited(
+		struct task_struct *p)
+{
+}
+#endif
+
+static inline void membarrier_sched_in(struct task_struct *prev,
+		struct task_struct *next)
+{
+	membarrier_arch_sched_in(prev, next);
+}
+static inline void membarrier_fork(struct task_struct *t,
+		unsigned long clone_flags)
+{
+	if (!current->mm || !t->mm)
+		return;
+	t->mm->membarrier_private_expedited =
+		current->mm->membarrier_private_expedited;
+	membarrier_arch_fork(t, clone_flags);
+}
+static inline void membarrier_execve(struct task_struct *t)
+{
+	t->mm->membarrier_private_expedited = 0;
+	membarrier_arch_execve(t);
+}
+#else
+static inline void membarrier_sched_in(struct task_struct *prev,
+		struct task_struct *next)
+{
+}
+static inline void membarrier_fork(struct task_struct *t,
+		unsigned long clone_flags)
+{
+}
+static inline void membarrier_execve(struct task_struct *t)
+{
+}
+#endif
+
 #endif /* _LINUX_SCHED_MM_H */
include/uapi/linux/membarrier.h +16 -7
index 6d47b3249d8a..4e01ad7ffe98 100644
@@ -52,21 +52,30 @@
  *                          (non-running threads are de facto in such a
  *                          state). This only covers threads from the
  *                          same processes as the caller thread. This
- *                          command returns 0. The "expedited" commands
- *                          complete faster than the non-expedited ones,
- *                          they never block, but have the downside of
- *                          causing extra overhead.
+ *                          command returns 0 on success. The
+ *                          "expedited" commands complete faster than
+ *                          the non-expedited ones, they never block,
+ *                          but have the downside of causing extra
+ *                          overhead. A process needs to register its
+ *                          intent to use the private expedited command
+ *                          prior to using it, otherwise this command
+ *                          returns -EPERM.
+ * @MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED:
+ *                          Register the process intent to use
+ *                          MEMBARRIER_CMD_PRIVATE_EXPEDITED. Always
+ *                          returns 0.
  *
  * Command to be passed to the membarrier system call. The commands need to
  * be a single bit each, except for MEMBARRIER_CMD_QUERY which is assigned to
  * the value 0.
  */
 enum membarrier_cmd {
-	MEMBARRIER_CMD_QUERY			= 0,
-	MEMBARRIER_CMD_SHARED			= (1 << 0),
+	MEMBARRIER_CMD_QUERY				= 0,
+	MEMBARRIER_CMD_SHARED				= (1 << 0),
 	/* reserved for MEMBARRIER_CMD_SHARED_EXPEDITED (1 << 1) */
 	/* reserved for MEMBARRIER_CMD_PRIVATE (1 << 2) */
-	MEMBARRIER_CMD_PRIVATE_EXPEDITED	= (1 << 3),
+	MEMBARRIER_CMD_PRIVATE_EXPEDITED		= (1 << 3),
+	MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED	= (1 << 4),
 };
 
 #endif /* _UAPI_LINUX_MEMBARRIER_H */
init/Kconfig +3 -0
index 78cb2461012e..a3dc6a66f0d1 100644
@@ -1400,6 +1400,9 @@ config MEMBARRIER
 
 	  If unsure, say Y.
 
+config ARCH_HAS_MEMBARRIER_HOOKS
+	bool
+
 config EMBEDDED
 	bool "Embedded system"
 	option allnoconfig_y
kernel/fork.c +2 -0
index 10646182440f..bd4a93915e08 100644
@@ -1840,6 +1840,8 @@ static __latent_entropy struct task_struct *copy_process(
 	 */
 	copy_seccomp(p);
 
+	membarrier_fork(p, clone_flags);
+
 	/*
 	 * Process group and session signals need to be delivered to just the
 	 * parent before the fork or both the parent and the child after the
kernel/sched/core.c +3 -13
index 7977b25acf54..08095bb1cfe6 100644
@@ -2643,17 +2643,8 @@ static struct rq *finish_task_switch(struct task_struct *prev)
 	 */
 	prev_state = prev->state;
 	vtime_task_switch(prev);
+	membarrier_sched_in(prev, current);
 	perf_event_task_sched_in(prev, current);
-	/*
-	 * The membarrier system call requires a full memory barrier
-	 * after storing to rq->curr, before going back to user-space.
-	 *
-	 * TODO: This smp_mb__after_unlock_lock can go away if PPC end
-	 * up adding a full barrier to switch_mm(), or we should figure
-	 * out if a smp_mb__after_unlock_lock is really the proper API
-	 * to use.
-	 */
-	smp_mb__after_unlock_lock();
 	finish_lock_switch(rq, prev);
 	finish_arch_post_lock_switch();
 
@@ -3363,9 +3354,8 @@ static void __sched notrace __schedule(bool preempt)
 		 * care of this barrier. For weakly ordered machines for
 		 * which spin_unlock() acts as a RELEASE barrier (only
 		 * arm64 and PowerPC), arm64 has a full barrier in
-		 * switch_to(), and PowerPC has
-		 * smp_mb__after_unlock_lock() before
-		 * finish_lock_switch().
+		 * switch_to(), and PowerPC has a full barrier in
+		 * membarrier_arch_sched_in().
 		 */
 		++*switch_count;
 
kernel/sched/membarrier.c +21 -4
index a92fddc22747..00a2618a36ba 100644
@@ -26,21 +26,25 @@
  * except MEMBARRIER_CMD_QUERY.
  */
 #define MEMBARRIER_CMD_BITMASK	\
-	(MEMBARRIER_CMD_SHARED | MEMBARRIER_CMD_PRIVATE_EXPEDITED)
+	(MEMBARRIER_CMD_SHARED | MEMBARRIER_CMD_PRIVATE_EXPEDITED	\
+	| MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED)
 
 static void ipi_mb(void *info)
 {
 	smp_mb();	/* IPIs should be serializing but paranoid. */
 }
 
-static void membarrier_private_expedited(void)
+static int membarrier_private_expedited(void)
 {
 	int cpu;
 	bool fallback = false;
 	cpumask_var_t tmpmask;
 
+	if (!current->mm->membarrier_private_expedited)
+		return -EPERM;
+
 	if (num_online_cpus() == 1)
-		return;
+		return 0;
 
 	/*
 	 * Matches memory barriers around rq->curr modification in
@@ -94,6 +98,17 @@ static void membarrier_private_expedited(void)
 	 * rq->curr modification in scheduler.
 	 */
 	smp_mb();	/* exit from system call is not a mb */
+	return 0;
+}
+
+static void membarrier_register_private_expedited(void)
+{
+	struct task_struct *p = current;
+
+	if (READ_ONCE(p->mm->membarrier_private_expedited))
+		return;
+	WRITE_ONCE(p->mm->membarrier_private_expedited, 1);
+	membarrier_arch_register_private_expedited(p);
 }
 
 /**
@@ -144,7 +159,9 @@ SYSCALL_DEFINE2(membarrier, int, cmd, int, flags)
 			synchronize_sched();
 		return 0;
 	case MEMBARRIER_CMD_PRIVATE_EXPEDITED:
-		membarrier_private_expedited();
+		return membarrier_private_expedited();
+	case MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED:
+		membarrier_register_private_expedited();
 		return 0;
 	default:
 		return -EINVAL;
Signature · 1 line
2.11.0
#3 lore raw
Quoted 89 lines · Boqun Feng

Hi Mathieu,

Quoted 10 lines · Mathieu Desnoyers

Provide a new command allowing processes to register their intent to use the private expedited command.

This allows PowerPC to skip the full memory barrier in switch_mm(), and only issue the barrier when scheduling into a task belonging to a process that has registered to use expedited private.

Processes are now required to register before using MEMBARRIER_CMD_PRIVATE_EXPEDITED, otherwise that command returns EPERM.

Sorry I'm late for the party, but I couldn't stop thinking whether we could avoid the register thing at all, because the registering makes sys_membarrier() more complex(both for the interface and the implementation). So how about we trade-off a little bit by taking some(not all) the rq->locks?

The idea is in membarrier_private_expedited(), we go through all ->curr on each CPU and

1) If it's a userspace task and its ->mm is matched, we send an ipi

2) If it's a kernel task, we skip

(Because there will be a smp_mb() implied by mmdrop(), when it switchs to userspace task).

3) If it's a userspace task and its ->mm is not matched, we take

	the corresponding rq->lock and check rq->curr again, if its ->mm
	matched, we send an ipi, otherwise we do nothing.

	(Because if we observe rq->curr is not matched with rq->lock
	held, when a task having matched ->mm schedules in, the rq->lock
	pairing along with the smp_mb__after_spinlock() will guarantee
	it observes all memory ops before sys_membarrir()).

membarrier_private_expedited() will look like this if we choose this way:

void membarrier_private_expedited()

{
	int cpu;
	bool fallback = false;
	cpumask_var_t tmpmask;
	struct rq_flags rf;


	if (num_online_cpus() == 1)
		return;

	smp_mb();

	if (!zalloc_cpumask_var(&tmpmask, GFP_NOWAIT)) {
		/* Fallback for OOM. */
		fallback = true;
	}

	cpus_read_lock();
	for_each_online_cpu(cpu) {
		struct task_struct *p;

		if (cpu == raw_smp_processor_id())
			continue;

		rcu_read_lock();
		p = task_rcu_dereference(&cpu_rq(cpu)->curr);

		if (!p) {
			rcu_read_unlock();
			continue;
		}

		if (p->mm == current->mm) {
			if (!fallback)
				__cpumask_set_cpu(cpu, tmpmask);
			else
				smp_call_function_single(cpu, ipi_mb, NULL, 1);
		}

		if (p->mm == current->mm || !p->mm) {
			rcu_read_unlock();
			continue;
		}

		rcu_read_unlock();
Quoted 7 lines
		/*
		 * This should be a arch-specific code, as we don't
		 * need it at else place other than some archs without
		 * a smp_mb() in switch_mm() (i.e. powerpc)
		 */
		rq_lock_irq(cpu_rq(cpu), &rf);
		if (p->mm == current->mm) {

Oops, this one should be

		if (cpu_curr(cpu)->mm == current->mm)
Quoted 4 lines
			if (!fallback)
				__cpumask_set_cpu(cpu, tmpmask);
			else
				smp_call_function_single(cpu, ipi_mb, NULL, 1);

, and this better be moved out of the lock rq->lock critical section.

Regards,
Boqun

Quoted 17 lines
		}
		rq_unlock_irq(cpu_rq(cpu), &rf);
	}
	if (!fallback) {
		smp_call_function_many(tmpmask, ipi_mb, NULL, 1);
		free_cpumask_var(tmpmask);
	}
	cpus_read_unlock();

	smp_mb();
}

Thoughts?

Regards,
Boqun

[...]

author maintainer
#4 lore raw
Quoted 107 lines · ----- On Sep 21, 2017, at 11:30 PM, Boqun Feng [email protected] wrote:
Quoted 89 lines · Boqun Feng

Hi Mathieu,

Quoted 10 lines · Mathieu Desnoyers

Provide a new command allowing processes to register their intent to use the private expedited command.

This allows PowerPC to skip the full memory barrier in switch_mm(), and only issue the barrier when scheduling into a task belonging to a process that has registered to use expedited private.

Processes are now required to register before using MEMBARRIER_CMD_PRIVATE_EXPEDITED, otherwise that command returns EPERM.

Sorry I'm late for the party, but I couldn't stop thinking whether we could avoid the register thing at all, because the registering makes sys_membarrier() more complex(both for the interface and the implementation). So how about we trade-off a little bit by taking some(not all) the rq->locks?

The idea is in membarrier_private_expedited(), we go through all ->curr on each CPU and

1) If it's a userspace task and its ->mm is matched, we send an ipi

2) If it's a kernel task, we skip

(Because there will be a smp_mb() implied by mmdrop(), when it switchs to userspace task).

3) If it's a userspace task and its ->mm is not matched, we take

	the corresponding rq->lock and check rq->curr again, if its ->mm
	matched, we send an ipi, otherwise we do nothing.

	(Because if we observe rq->curr is not matched with rq->lock
	held, when a task having matched ->mm schedules in, the rq->lock
	pairing along with the smp_mb__after_spinlock() will guarantee
	it observes all memory ops before sys_membarrir()).

membarrier_private_expedited() will look like this if we choose this way:

void membarrier_private_expedited()

{
	int cpu;
	bool fallback = false;
	cpumask_var_t tmpmask;
	struct rq_flags rf;


	if (num_online_cpus() == 1)
		return;

	smp_mb();

	if (!zalloc_cpumask_var(&tmpmask, GFP_NOWAIT)) {
		/* Fallback for OOM. */
		fallback = true;
	}

	cpus_read_lock();
	for_each_online_cpu(cpu) {
		struct task_struct *p;

		if (cpu == raw_smp_processor_id())
			continue;

		rcu_read_lock();
		p = task_rcu_dereference(&cpu_rq(cpu)->curr);

		if (!p) {
			rcu_read_unlock();
			continue;
		}

		if (p->mm == current->mm) {
			if (!fallback)
				__cpumask_set_cpu(cpu, tmpmask);
			else
				smp_call_function_single(cpu, ipi_mb, NULL, 1);
		}

		if (p->mm == current->mm || !p->mm) {
			rcu_read_unlock();
			continue;
		}

		rcu_read_unlock();
Quoted 7 lines
		/*
		 * This should be a arch-specific code, as we don't
		 * need it at else place other than some archs without
		 * a smp_mb() in switch_mm() (i.e. powerpc)
		 */
		rq_lock_irq(cpu_rq(cpu), &rf);
		if (p->mm == current->mm) {

Oops, this one should be

		if (cpu_curr(cpu)->mm == current->mm)
Quoted 4 lines
			if (!fallback)
				__cpumask_set_cpu(cpu, tmpmask);
			else
				smp_call_function_single(cpu, ipi_mb, NULL, 1);
Quoted 18 lines

, and this better be moved out of the lock rq->lock critical section.

Regards,
Boqun

Quoted 13 lines
		}
		rq_unlock_irq(cpu_rq(cpu), &rf);
	}
	if (!fallback) {
		smp_call_function_many(tmpmask, ipi_mb, NULL, 1);
		free_cpumask_var(tmpmask);
	}
	cpus_read_unlock();

	smp_mb();
}

Thoughts?

Hi Boqun,

The main concern Peter has with the runqueue locking approach is interference with the scheduler by hitting all CPU's runqueue locks repeatedly if someone performs membarrier system calls in a short loop.

Just reading the rq->curr pointer does not generate as much overhead as grabbing each rq lock.

Thanks,

Mathieu

Quoted 5 lines
Quoted 4 lines

Regards,
Boqun

[...]

Signature · 3 lines
Mathieu Desnoyers
EfficiOS Inc.
http://www.efficios.com
maintainer
#5 lore raw
In reply to Boqun Feng
Quoted 10 lines · Boqun Feng

The idea is in membarrier_private_expedited(), we go through all ->curr on each CPU and

1) If it's a userspace task and its ->mm is matched, we send an ipi

2) If it's a kernel task, we skip

(Because there will be a smp_mb() implied by mmdrop(), when it switchs to userspace task).

Quoted 8 lines

3) If it's a userspace task and its ->mm is not matched, we take

	the corresponding rq->lock and check rq->curr again, if its ->mm
	matched, we send an ipi, otherwise we do nothing.

	(Because if we observe rq->curr is not matched with rq->lock
	held, when a task having matched ->mm schedules in, the rq->lock
	pairing along with the smp_mb__after_spinlock() will guarantee
	it observes all memory ops before sys_membarrir()).

3) is an insta DoS.

#6 lore raw
Quoted 16 lines · Peter Zijlstra
Quoted 14 lines · Boqun Feng

The idea is in membarrier_private_expedited(), we go through all ->curr on each CPU and

1) If it's a userspace task and its ->mm is matched, we send an ipi

2) If it's a kernel task, we skip

(Because there will be a smp_mb() implied by mmdrop(), when it switchs to userspace task).

3) If it's a userspace task and its ->mm is not matched, we take

	the corresponding rq->lock and check rq->curr again, if its ->mm
	matched, we send an ipi, otherwise we do nothing.
Quoted 6 lines
Quoted 4 lines
	(Because if we observe rq->curr is not matched with rq->lock
	held, when a task having matched ->mm schedules in, the rq->lock
	pairing along with the smp_mb__after_spinlock() will guarantee
	it observes all memory ops before sys_membarrir()).

3) is an insta DoS.

OK, fair enough.

Regards,
Boqun

#7 lore raw
In reply to Mathieu Desnoyers

On Tue, Sep 19, 2017 at 06:13:41PM -0400, Mathieu Desnoyers wrote:
[...]

arch/powerpc/include/asm/membarrier.h line 12

Quoted 9 lines
arch/powerpc/include/asm/membarrier.h new file +9 -0
+static inline void membarrier_arch_sched_in(struct task_struct *prev,
+		struct task_struct *next)
+{
+	/*
+	 * Only need the full barrier when switching between processes.
+	 */
+	if (likely(!test_ti_thread_flag(task_thread_info(next),
+			TIF_MEMBARRIER_PRIVATE_EXPEDITED)
+				|| prev->mm == next->mm))

And we also don't need the smp_mb() if !prev->mm, because switching from kernel to user will have a smp_mb() implied by mmdrop()?

Quoted 8 lines
arch/powerpc/include/asm/membarrier.h new file +8 -0
+		return;
+
+	/*
+	 * The membarrier system call requires a full memory barrier
+	 * after storing to rq->curr, before going back to user-space.
+	 */
+	smp_mb();
+}

[...]

include/linux/sched/mm.h static inline void memalloc_noreclaim_restore(unsigned int flags) · line 246

Quoted 7 lines
include/linux/sched/mm.h +7 -0
+static inline void membarrier_fork(struct task_struct *t,
+		unsigned long clone_flags)
+{
+	if (!current->mm || !t->mm)
+		return;
+	t->mm->membarrier_private_expedited =
+		current->mm->membarrier_private_expedited;

Have we already done the copy of ->membarrier_private_expedited in copy_mm()?

Regards,
Boqun

Quoted 21 lines
include/linux/sched/mm.h +21 -0
+	membarrier_arch_fork(t, clone_flags);
+}
+static inline void membarrier_execve(struct task_struct *t)
+{
+	t->mm->membarrier_private_expedited = 0;
+	membarrier_arch_execve(t);
+}
+#else
+static inline void membarrier_sched_in(struct task_struct *prev,
+		struct task_struct *next)
+{
+}
+static inline void membarrier_fork(struct task_struct *t,
+		unsigned long clone_flags)
+{
+}
+static inline void membarrier_execve(struct task_struct *t)
+{
+}
+#endif
+

[...]

author maintainer
#8 lore raw
Quoted 14 lines · ----- On Sep 22, 2017, at 4:59 AM, Boqun Feng [email protected] wrote:

On Tue, Sep 19, 2017 at 06:13:41PM -0400, Mathieu Desnoyers wrote:
[...]

arch/powerpc/include/asm/membarrier.h line 12

Quoted 9 lines
arch/powerpc/include/asm/membarrier.h new file +9 -0
+static inline void membarrier_arch_sched_in(struct task_struct *prev,
+		struct task_struct *next)
+{
+	/*
+	 * Only need the full barrier when switching between processes.
+	 */
+	if (likely(!test_ti_thread_flag(task_thread_info(next),
+			TIF_MEMBARRIER_PRIVATE_EXPEDITED)
+				|| prev->mm == next->mm))

And we also don't need the smp_mb() if !prev->mm, because switching from kernel to user will have a smp_mb() implied by mmdrop()?

Right. And we also don't need it when switching from userspace to kernel thread neither. Something like this:

static inline void membarrier_arch_sched_in(struct task_struct *prev,
struct task_struct *next)
{
/*
* Only need the full barrier when switching between processes.
* Barrier when switching from kernel to userspace is not
* required here, given that it is implied by mmdrop(). Barrier
* when switching from userspace to kernel is not needed after
* store to rq->curr.
*/
if (likely(!test_ti_thread_flag(task_thread_info(next),
TIF_MEMBARRIER_PRIVATE_EXPEDITED)
|| !prev->mm || !next->mm || prev->mm == next->mm))
return;

/*
* The membarrier system call requires a full memory barrier
* after storing to rq->curr, before going back to user-space. */
smp_mb();
}

Quoted 12 lines
Quoted 9 lines
arch/powerpc/include/asm/membarrier.h new file +8 -0
+		return;
+
+	/*
+	 * The membarrier system call requires a full memory barrier
+	 * after storing to rq->curr, before going back to user-space.
+	 */
+	smp_mb();
+}

[...]

Quoted 10 lines

include/linux/sched/mm.h static inline void memalloc_noreclaim_restore(unsigned int flags) · line 246

Quoted 7 lines
include/linux/sched/mm.h +7 -0
+static inline void membarrier_fork(struct task_struct *t,
+		unsigned long clone_flags)
+{
+	if (!current->mm || !t->mm)
+		return;
+	t->mm->membarrier_private_expedited =
+		current->mm->membarrier_private_expedited;

Have we already done the copy of ->membarrier_private_expedited in copy_mm()?

copy_mm() is performed without holding current->sighand->siglock, so it appears to be racing with concurrent membarrier register cmd. However, given that it is a single flag updated with WRITE_ONCE() and read with READ_ONCE(), it might be OK to rely on copy_mm there. If userspace runs registration concurrently with fork, they should not expect the child to be specifically registered or unregistered.

So yes, I think you are right about removing this copy and relying on copy_mm() instead. I also think we can improve membarrier_arch_fork() on powerpc to test the current thread flag rather than using current->mm.

Which leads to those two changes:

static inline void membarrier_fork(struct task_struct *t,
unsigned long clone_flags)

{
        /*
         * Prior copy_mm() copies the membarrier_private_expedited field
         * from current->mm to t->mm.
         */
        membarrier_arch_fork(t, clone_flags);
}

And on PowerPC:

static inline void membarrier_arch_fork(struct task_struct *t,
unsigned long clone_flags)
{
/*
* Coherence of TIF_MEMBARRIER_PRIVATE_EXPEDITED against thread
* fork is protected by siglock. membarrier_arch_fork is called
* with siglock held.
*/
if (test_thread_flag(TIF_MEMBARRIER_PRIVATE_EXPEDITED))
set_ti_thread_flag(task_thread_info(t),
TIF_MEMBARRIER_PRIVATE_EXPEDITED);
}

Thanks,

Mathieu

Quoted 26 lines

Regards,
Boqun

Quoted 21 lines
include/linux/sched/mm.h +21 -0
+	membarrier_arch_fork(t, clone_flags);
+}
+static inline void membarrier_execve(struct task_struct *t)
+{
+	t->mm->membarrier_private_expedited = 0;
+	membarrier_arch_execve(t);
+}
+#else
+static inline void membarrier_sched_in(struct task_struct *prev,
+		struct task_struct *next)
+{
+}
+static inline void membarrier_fork(struct task_struct *t,
+		unsigned long clone_flags)
+{
+}
+static inline void membarrier_execve(struct task_struct *t)
+{
+}
+#endif
+

[...]

Signature · 3 lines
Mathieu Desnoyers
EfficiOS Inc.
http://www.efficios.com
#9 lore raw
Quoted 17 lines · Mathieu Desnoyers
Quoted 14 lines · ----- On Sep 22, 2017, at 4:59 AM, Boqun Feng [email protected] wrote:

On Tue, Sep 19, 2017 at 06:13:41PM -0400, Mathieu Desnoyers wrote:
[...]

arch/powerpc/include/asm/membarrier.h line 12

Quoted 9 lines
arch/powerpc/include/asm/membarrier.h new file +9 -0
+static inline void membarrier_arch_sched_in(struct task_struct *prev,
+		struct task_struct *next)
+{
+	/*
+	 * Only need the full barrier when switching between processes.
+	 */
+	if (likely(!test_ti_thread_flag(task_thread_info(next),
+			TIF_MEMBARRIER_PRIVATE_EXPEDITED)
+				|| prev->mm == next->mm))

And we also don't need the smp_mb() if !prev->mm, because switching from kernel to user will have a smp_mb() implied by mmdrop()?

Quoted 1 line

Right. And we also don't need it when switching from userspace to kernel

Yep, but this case is covered already, as I think we don't allow kernel thread to have TIF_MEMBARRIER_PRIVATE_EXPEDITED set, right?

Quoted 15 lines

thread neither. Something like this:

static inline void membarrier_arch_sched_in(struct task_struct *prev,
struct task_struct *next)
{
/*
* Only need the full barrier when switching between processes.
* Barrier when switching from kernel to userspace is not
* required here, given that it is implied by mmdrop(). Barrier
* when switching from userspace to kernel is not needed after
* store to rq->curr.
*/
if (likely(!test_ti_thread_flag(task_thread_info(next),
TIF_MEMBARRIER_PRIVATE_EXPEDITED)
|| !prev->mm || !next->mm || prev->mm == next->mm))

, so no need to test next->mm here.

Quoted 32 lines
                return;

        /*
         * The membarrier system call requires a full memory barrier
         * after storing to rq->curr, before going back to user-space.
         */
        smp_mb();
}
Quoted 12 lines
Quoted 9 lines
arch/powerpc/include/asm/membarrier.h new file +8 -0
+		return;
+
+	/*
+	 * The membarrier system call requires a full memory barrier
+	 * after storing to rq->curr, before going back to user-space.
+	 */
+	smp_mb();
+}

[...]

Quoted 10 lines

include/linux/sched/mm.h static inline void memalloc_noreclaim_restore(unsigned int flags) · line 246

Quoted 7 lines
include/linux/sched/mm.h +7 -0
+static inline void membarrier_fork(struct task_struct *t,
+		unsigned long clone_flags)
+{
+	if (!current->mm || !t->mm)
+		return;
+	t->mm->membarrier_private_expedited =
+		current->mm->membarrier_private_expedited;

Have we already done the copy of ->membarrier_private_expedited in copy_mm()?

Quoted 2 lines

copy_mm() is performed without holding current->sighand->siglock, so it appears to be racing with concurrent membarrier register cmd.

Speak of racing, I think we currently have a problem if we do a register_private_expedited in one thread and do a membarrer_private_expedited in another thread(sharing the same mm), as follow:

	{t1,t2,t3 sharing the same ->mm}
	CPU 0				CPU 1				CPU2
	====================		===================		============
	{in thread t1}							
	membarrier_register_private_expedited():
	  ...
	  WRITE_ONCE(->mm->membarrier_private_expedited, 1);
	  membarrier_arch_register_private_expedited():
	    ...
	    <haven't set the TIF for t3 yet>

	    				{in thread t2}
					membarrier_private_expedited():
					  READ_ONCE(->mm->membarrier_private_expedited); // == 1
					  ...
					  for_each_online_cpu()
					    ...
					    <p is cpu_rq(CPU2)->curr>
					    if (p && p->mm == current->mm) // false
					    <so no ipi sent to CPU2>
					    				
									{about to switch to t3}
									rq->curr = t3;
									....
									context_switch():
									  ...
									  finish_task_swtich():
									    membarrier_sched_in():
									    <TIF is not set>
									    // no smp_mb() here.

, and we will miss the smp_mb() on CPU2, right? And this could even happen if t2 has a membarrier_register_private_expedited() preceding the membarrier_private_expedited().

Am I missing something subtle here?

Regards,
Boqun

Quoted 72 lines

However, given that it is a single flag updated with WRITE_ONCE() and read with READ_ONCE(), it might be OK to rely on copy_mm there. If userspace runs registration concurrently with fork, they should not expect the child to be specifically registered or unregistered.

So yes, I think you are right about removing this copy and relying on copy_mm() instead. I also think we can improve membarrier_arch_fork() on powerpc to test the current thread flag rather than using current->mm.

Which leads to those two changes:

static inline void membarrier_fork(struct task_struct *t,
unsigned long clone_flags)

{
        /*
         * Prior copy_mm() copies the membarrier_private_expedited field
         * from current->mm to t->mm.
         */
        membarrier_arch_fork(t, clone_flags);
}

And on PowerPC:

static inline void membarrier_arch_fork(struct task_struct *t,
unsigned long clone_flags)
{
/*
* Coherence of TIF_MEMBARRIER_PRIVATE_EXPEDITED against thread
* fork is protected by siglock. membarrier_arch_fork is called
* with siglock held.
*/
if (test_thread_flag(TIF_MEMBARRIER_PRIVATE_EXPEDITED))
set_ti_thread_flag(task_thread_info(t),
TIF_MEMBARRIER_PRIVATE_EXPEDITED);
}

Thanks,

Mathieu

Quoted 26 lines

Regards,
Boqun

Quoted 21 lines
include/linux/sched/mm.h +21 -0
+	membarrier_arch_fork(t, clone_flags);
+}
+static inline void membarrier_execve(struct task_struct *t)
+{
+	t->mm->membarrier_private_expedited = 0;
+	membarrier_arch_execve(t);
+}
+#else
+static inline void membarrier_sched_in(struct task_struct *prev,
+		struct task_struct *next)
+{
+}
+static inline void membarrier_fork(struct task_struct *t,
+		unsigned long clone_flags)
+{
+}
+static inline void membarrier_execve(struct task_struct *t)
+{
+}
+#endif
+

[...]

Signature · 3 lines
Mathieu Desnoyers
EfficiOS Inc.
http://www.efficios.com
author maintainer
#10 lore raw
Quoted 18 lines · ----- On Sep 24, 2017, at 9:30 AM, Boqun Feng [email protected] wrote:
Quoted 17 lines · Mathieu Desnoyers
Quoted 14 lines · ----- On Sep 22, 2017, at 4:59 AM, Boqun Feng [email protected] wrote:

On Tue, Sep 19, 2017 at 06:13:41PM -0400, Mathieu Desnoyers wrote:
[...]

arch/powerpc/include/asm/membarrier.h line 12

Quoted 9 lines
arch/powerpc/include/asm/membarrier.h new file +9 -0
+static inline void membarrier_arch_sched_in(struct task_struct *prev,
+		struct task_struct *next)
+{
+	/*
+	 * Only need the full barrier when switching between processes.
+	 */
+	if (likely(!test_ti_thread_flag(task_thread_info(next),
+			TIF_MEMBARRIER_PRIVATE_EXPEDITED)
+				|| prev->mm == next->mm))

And we also don't need the smp_mb() if !prev->mm, because switching from kernel to user will have a smp_mb() implied by mmdrop()?

Quoted 4 lines
Quoted 1 line

Right. And we also don't need it when switching from userspace to kernel

Yep, but this case is covered already, as I think we don't allow kernel thread to have TIF_MEMBARRIER_PRIVATE_EXPEDITED set, right?

Good point.

Quoted 17 lines
Quoted 16 lines

thread neither. Something like this:

static inline void membarrier_arch_sched_in(struct task_struct *prev,
struct task_struct *next)
{
/*
* Only need the full barrier when switching between processes.
* Barrier when switching from kernel to userspace is not
* required here, given that it is implied by mmdrop(). Barrier
* when switching from userspace to kernel is not needed after
* store to rq->curr.
*/
if (likely(!test_ti_thread_flag(task_thread_info(next),
TIF_MEMBARRIER_PRIVATE_EXPEDITED)
|| !prev->mm || !next->mm || prev->mm == next->mm))

Quoted 2 lines

, so no need to test next->mm here.

Right, it's redundant wrt testing the thread flag.

Quoted 71 lines
Quoted 32 lines
                return;

        /*
         * The membarrier system call requires a full memory barrier
         * after storing to rq->curr, before going back to user-space.
         */
        smp_mb();
}
Quoted 12 lines
Quoted 9 lines
arch/powerpc/include/asm/membarrier.h new file +8 -0
+		return;
+
+	/*
+	 * The membarrier system call requires a full memory barrier
+	 * after storing to rq->curr, before going back to user-space.
+	 */
+	smp_mb();
+}

[...]

Quoted 10 lines

include/linux/sched/mm.h static inline void memalloc_noreclaim_restore(unsigned int flags) · line 246

Quoted 7 lines
include/linux/sched/mm.h +7 -0
+static inline void membarrier_fork(struct task_struct *t,
+		unsigned long clone_flags)
+{
+	if (!current->mm || !t->mm)
+		return;
+	t->mm->membarrier_private_expedited =
+		current->mm->membarrier_private_expedited;

Have we already done the copy of ->membarrier_private_expedited in copy_mm()?

Quoted 2 lines

copy_mm() is performed without holding current->sighand->siglock, so it appears to be racing with concurrent membarrier register cmd.

Speak of racing, I think we currently have a problem if we do a register_private_expedited in one thread and do a membarrer_private_expedited in another thread(sharing the same mm), as follow:

	{t1,t2,t3 sharing the same ->mm}
	CPU 0				CPU 1				CPU2
	====================		===================		============
	{in thread t1}
	membarrier_register_private_expedited():
	  ...
	  WRITE_ONCE(->mm->membarrier_private_expedited, 1);
	  membarrier_arch_register_private_expedited():
	    ...
	    <haven't set the TIF for t3 yet>

	    				{in thread t2}
					membarrier_private_expedited():
					  READ_ONCE(->mm->membarrier_private_expedited); // == 1
					  ...
					  for_each_online_cpu()
					    ...
					    <p is cpu_rq(CPU2)->curr>
					    if (p && p->mm == current->mm) // false
					    <so no ipi sent to CPU2>
					    				
									{about to switch to t3}
									rq->curr = t3;
									....
									context_switch():
									  ...
									  finish_task_swtich():
									    membarrier_sched_in():
									    <TIF is not set>
									    // no smp_mb() here.
Quoted 5 lines

, and we will miss the smp_mb() on CPU2, right? And this could even happen if t2 has a membarrier_register_private_expedited() preceding the membarrier_private_expedited().

Am I missing something subtle here?

I think the problem sits in this function:

static void membarrier_register_private_expedited(void)

{
        struct task_struct *p = current;

        if (READ_ONCE(p->mm->membarrier_private_expedited))
                return;
        WRITE_ONCE(p->mm->membarrier_private_expedited, 1);
        membarrier_arch_register_private_expedited(p);
}

I need to change the order between WRITE_ONCE() and invoking membarrier_arch_register_private_expedited. If I issue the WRITE_ONCE after the arch code (which sets the TIF flags), then concurrent membarrier priv exped commands will simply return an -EPERM error:

static void membarrier_register_private_expedited(void)

{
        struct task_struct *p = current;

        if (READ_ONCE(p->mm->membarrier_private_expedited))
                return;
        membarrier_arch_register_private_expedited(p);
        WRITE_ONCE(p->mm->membarrier_private_expedited, 1);
}

Do you agree that this would fix the race you identified ?

Thanks,

Mathieu

Quoted 77 lines

Regards,
Boqun

Quoted 72 lines

However, given that it is a single flag updated with WRITE_ONCE() and read with READ_ONCE(), it might be OK to rely on copy_mm there. If userspace runs registration concurrently with fork, they should not expect the child to be specifically registered or unregistered.

So yes, I think you are right about removing this copy and relying on copy_mm() instead. I also think we can improve membarrier_arch_fork() on powerpc to test the current thread flag rather than using current->mm.

Which leads to those two changes:

static inline void membarrier_fork(struct task_struct *t,
unsigned long clone_flags)

{
        /*
         * Prior copy_mm() copies the membarrier_private_expedited field
         * from current->mm to t->mm.
         */
        membarrier_arch_fork(t, clone_flags);
}

And on PowerPC:

static inline void membarrier_arch_fork(struct task_struct *t,
unsigned long clone_flags)
{
/*
* Coherence of TIF_MEMBARRIER_PRIVATE_EXPEDITED against thread
* fork is protected by siglock. membarrier_arch_fork is called
* with siglock held.
*/
if (test_thread_flag(TIF_MEMBARRIER_PRIVATE_EXPEDITED))
set_ti_thread_flag(task_thread_info(t),
TIF_MEMBARRIER_PRIVATE_EXPEDITED);
}

Thanks,

Mathieu

Quoted 18 lines

Regards,
Boqun

Quoted 14 lines
include/linux/sched/mm.h +14 -0
+	membarrier_arch_fork(t, clone_flags);
+}
+static inline void membarrier_execve(struct task_struct *t)
+{
+	t->mm->membarrier_private_expedited = 0;
+	membarrier_arch_execve(t);
+}
+#else
+static inline void membarrier_sched_in(struct task_struct *prev,
+		struct task_struct *next)
+{
+}
+static inline void membarrier_fork(struct task_struct *t,
+		unsigned long clone_flags)
Quoted 8 lines
Quoted 7 lines
include/linux/sched/mm.h +7 -0
+{
+}
+static inline void membarrier_execve(struct task_struct *t)
+{
+}
+#endif
+

[...]

--
Mathieu Desnoyers
EfficiOS Inc.
http://www.efficios.com

Signature · 3 lines
Mathieu Desnoyers
EfficiOS Inc.
http://www.efficios.com
#11 lore raw

On Sun, Sep 24, 2017 at 02:23:04PM +0000, Mathieu Desnoyers wrote:
[...]

Quoted 74 lines
Quoted 40 lines
Quoted 3 lines

copy_mm() is performed without holding current->sighand->siglock, so it appears to be racing with concurrent membarrier register cmd.

Speak of racing, I think we currently have a problem if we do a register_private_expedited in one thread and do a membarrer_private_expedited in another thread(sharing the same mm), as follow:

	{t1,t2,t3 sharing the same ->mm}
	CPU 0				CPU 1				CPU2
	====================		===================		============
	{in thread t1}
	membarrier_register_private_expedited():
	  ...
	  WRITE_ONCE(->mm->membarrier_private_expedited, 1);
	  membarrier_arch_register_private_expedited():
	    ...
	    <haven't set the TIF for t3 yet>

	    				{in thread t2}
					membarrier_private_expedited():
					  READ_ONCE(->mm->membarrier_private_expedited); // == 1
					  ...
					  for_each_online_cpu()
					    ...
					    <p is cpu_rq(CPU2)->curr>
					    if (p && p->mm == current->mm) // false
					    <so no ipi sent to CPU2>
					    				
									{about to switch to t3}
									rq->curr = t3;
									....
									context_switch():
									  ...
									  finish_task_swtich():
									    membarrier_sched_in():
									    <TIF is not set>
									    // no smp_mb() here.
Quoted 5 lines

, and we will miss the smp_mb() on CPU2, right? And this could even happen if t2 has a membarrier_register_private_expedited() preceding the membarrier_private_expedited().

Am I missing something subtle here?

I think the problem sits in this function:

static void membarrier_register_private_expedited(void)

{
        struct task_struct *p = current;

        if (READ_ONCE(p->mm->membarrier_private_expedited))
                return;
        WRITE_ONCE(p->mm->membarrier_private_expedited, 1);
        membarrier_arch_register_private_expedited(p);
}

I need to change the order between WRITE_ONCE() and invoking membarrier_arch_register_private_expedited. If I issue the WRITE_ONCE after the arch code (which sets the TIF flags), then concurrent membarrier priv exped commands will simply return an -EPERM error:

static void membarrier_register_private_expedited(void)

{
        struct task_struct *p = current;

        if (READ_ONCE(p->mm->membarrier_private_expedited))
                return;
        membarrier_arch_register_private_expedited(p);
        WRITE_ONCE(p->mm->membarrier_private_expedited, 1);
}
Quoted 2 lines

Do you agree that this would fix the race you identified ?

Yep, that will do the trick ;-)

Regards,
Boqun

Quoted 4 lines

Thanks,

Mathieu

maintainer
#12 lore raw
Quoted 7 lines · Boqun Feng
Quoted 7 lines

static void membarrier_register_private_expedited(void)

{
        struct task_struct *p = current;

        if (READ_ONCE(p->mm->membarrier_private_expedited))
                return;
        membarrier_arch_register_private_expedited(p);

Should we not then also do:

	    barrier();
Quoted 2 lines
Quoted 2 lines
        WRITE_ONCE(p->mm->membarrier_private_expedited, 1);
}

to avoid the compiler lifting that store?

author maintainer
#13 lore raw
Quoted 9 lines · ----- On Sep 25, 2017, at 8:25 AM, Peter Zijlstra [email protected] wrote:
Quoted 7 lines · Boqun Feng
Quoted 7 lines

static void membarrier_register_private_expedited(void)

{
        struct task_struct *p = current;

        if (READ_ONCE(p->mm->membarrier_private_expedited))
                return;
        membarrier_arch_register_private_expedited(p);
Quoted 8 lines

Should we not then also do:

	    barrier();
Quoted 2 lines
Quoted 2 lines
        WRITE_ONCE(p->mm->membarrier_private_expedited, 1);
}

to avoid the compiler lifting that store?

membarrier_arch_register_private_expedited() being a function call, I recall compilers cannot move load/stores across those. Moreover, even if that function would happen to be eventually inlined, synchronize_sched() is needed at the end of the function to ensure the scheduler will observe the thread flags before it returns. That too would then act as a compiler barrier if that function is ever inlined in the future.

So do you think we should still add the barrier() as documentation, or is having synchronize_sched() in the callee enough ?

By the way, I think I should add a READ_ONCE() in membarrier_private_expedited to pair with the WRITE_ONCE() in registration, such as:

        if (!READ_ONCE(current->mm->membarrier_private_expedited))
                return -EPERM;

Thanks!

Mathieu

Signature · 3 lines
Mathieu Desnoyers
EfficiOS Inc.
http://www.efficios.com