* [PATCH 0/2 -rt] make locckdep selftest work on -rt
@ 2012-04-16 7:01 Yong Zhang
2012-04-16 7:01 ` [PATCH 1/2 -rt] lockdep: selftest: convert spinlock to raw spinlock Yong Zhang
` (2 more replies)
0 siblings, 3 replies; 4+ messages in thread
From: Yong Zhang @ 2012-04-16 7:01 UTC (permalink / raw)
To: linux-rt-users, linux-kernel; +Cc: tglx
On -rt,
1) There is no softirq context any more;
2) rwlock can not be used in harirq context;
Wrt that, convert spinlock to raw spinlock to make spinlock+hardirq
work, and discard the test for rwlock+hardirq and softirq.
BTW, I think patch#1 can go to upstream.
Thanks,
Yong
---
Yong Zhang (2):
lockdep: selftest: convert spinlock to raw spinlock
lockdep: selftest: Only do hardirq context test for raw spinlock
lib/locking-selftest.c | 57 +++++++++++++++++++++++++++++++++--------------
1 files changed, 40 insertions(+), 17 deletions(-)
--
1.7.5.4
^ permalink raw reply [flat|nested] 4+ messages in thread
* [PATCH 1/2 -rt] lockdep: selftest: convert spinlock to raw spinlock
2012-04-16 7:01 [PATCH 0/2 -rt] make locckdep selftest work on -rt Yong Zhang
@ 2012-04-16 7:01 ` Yong Zhang
2012-04-16 7:01 ` [PATCH 2/2 -rt] lockdep: selftest: Only do hardirq context test for " Yong Zhang
2012-04-17 7:23 ` [PATCH 3/3 -rt] lockdep: selftest: save/restore migrate_disable count Yong Zhang
2 siblings, 0 replies; 4+ messages in thread
From: Yong Zhang @ 2012-04-16 7:01 UTC (permalink / raw)
To: linux-rt-users, linux-kernel; +Cc: tglx, Yong Zhang
From: Yong Zhang <yong.zhang@windriver.com>
spinlock is sleepable on -rt and can not be used in
interrupt context.
Signed-off-by: Yong Zhang <yong.zhang0@gmail.com>
---
lib/locking-selftest.c | 34 +++++++++++++++++-----------------
1 files changed, 17 insertions(+), 17 deletions(-)
diff --git a/lib/locking-selftest.c b/lib/locking-selftest.c
index 7aae0f2..c3eb261 100644
--- a/lib/locking-selftest.c
+++ b/lib/locking-selftest.c
@@ -47,10 +47,10 @@ __setup("debug_locks_verbose=", setup_debug_locks_verbose);
* Normal standalone locks, for the circular and irq-context
* dependency tests:
*/
-static DEFINE_SPINLOCK(lock_A);
-static DEFINE_SPINLOCK(lock_B);
-static DEFINE_SPINLOCK(lock_C);
-static DEFINE_SPINLOCK(lock_D);
+static DEFINE_RAW_SPINLOCK(lock_A);
+static DEFINE_RAW_SPINLOCK(lock_B);
+static DEFINE_RAW_SPINLOCK(lock_C);
+static DEFINE_RAW_SPINLOCK(lock_D);
static DEFINE_RWLOCK(rwlock_A);
static DEFINE_RWLOCK(rwlock_B);
@@ -73,12 +73,12 @@ static DECLARE_RWSEM(rwsem_D);
* but X* and Y* are different classes. We do this so that
* we do not trigger a real lockup:
*/
-static DEFINE_SPINLOCK(lock_X1);
-static DEFINE_SPINLOCK(lock_X2);
-static DEFINE_SPINLOCK(lock_Y1);
-static DEFINE_SPINLOCK(lock_Y2);
-static DEFINE_SPINLOCK(lock_Z1);
-static DEFINE_SPINLOCK(lock_Z2);
+static DEFINE_RAW_SPINLOCK(lock_X1);
+static DEFINE_RAW_SPINLOCK(lock_X2);
+static DEFINE_RAW_SPINLOCK(lock_Y1);
+static DEFINE_RAW_SPINLOCK(lock_Y2);
+static DEFINE_RAW_SPINLOCK(lock_Z1);
+static DEFINE_RAW_SPINLOCK(lock_Z2);
static DEFINE_RWLOCK(rwlock_X1);
static DEFINE_RWLOCK(rwlock_X2);
@@ -107,10 +107,10 @@ static DECLARE_RWSEM(rwsem_Z2);
*/
#define INIT_CLASS_FUNC(class) \
static noinline void \
-init_class_##class(spinlock_t *lock, rwlock_t *rwlock, struct mutex *mutex, \
- struct rw_semaphore *rwsem) \
+init_class_##class(raw_spinlock_t *lock, rwlock_t *rwlock, \
+ struct mutex *mutex, struct rw_semaphore *rwsem)\
{ \
- spin_lock_init(lock); \
+ raw_spin_lock_init(lock); \
rwlock_init(rwlock); \
mutex_init(mutex); \
init_rwsem(rwsem); \
@@ -168,10 +168,10 @@ static void init_shared_classes(void)
* Shortcuts for lock/unlock API variants, to keep
* the testcases compact:
*/
-#define L(x) spin_lock(&lock_##x)
-#define U(x) spin_unlock(&lock_##x)
+#define L(x) raw_spin_lock(&lock_##x)
+#define U(x) raw_spin_unlock(&lock_##x)
#define LU(x) L(x); U(x)
-#define SI(x) spin_lock_init(&lock_##x)
+#define SI(x) raw_spin_lock_init(&lock_##x)
#define WL(x) write_lock(&rwlock_##x)
#define WU(x) write_unlock(&rwlock_##x)
@@ -911,7 +911,7 @@ GENERATE_PERMUTATIONS_3_EVENTS(irq_read_recursion_soft)
#define I2(x) \
do { \
- spin_lock_init(&lock_##x); \
+ raw_spin_lock_init(&lock_##x); \
rwlock_init(&rwlock_##x); \
mutex_init(&mutex_##x); \
init_rwsem(&rwsem_##x); \
--
1.7.5.4
^ permalink raw reply related [flat|nested] 4+ messages in thread
* [PATCH 2/2 -rt] lockdep: selftest: Only do hardirq context test for raw spinlock
2012-04-16 7:01 [PATCH 0/2 -rt] make locckdep selftest work on -rt Yong Zhang
2012-04-16 7:01 ` [PATCH 1/2 -rt] lockdep: selftest: convert spinlock to raw spinlock Yong Zhang
@ 2012-04-16 7:01 ` Yong Zhang
2012-04-17 7:23 ` [PATCH 3/3 -rt] lockdep: selftest: save/restore migrate_disable count Yong Zhang
2 siblings, 0 replies; 4+ messages in thread
From: Yong Zhang @ 2012-04-16 7:01 UTC (permalink / raw)
To: linux-rt-users, linux-kernel; +Cc: tglx, Yong Zhang
From: Yong Zhang <yong.zhang@windriver.com>
On -rt there is no softirq context any more and rwlock is sleepable,
disable softirq context test and rwlock+irq test.
Signed-off-by: Yong Zhang <yong.zhang0@gmail.com>
---
lib/locking-selftest.c | 23 +++++++++++++++++++++++
1 files changed, 23 insertions(+), 0 deletions(-)
diff --git a/lib/locking-selftest.c b/lib/locking-selftest.c
index c3eb261..23b8564 100644
--- a/lib/locking-selftest.c
+++ b/lib/locking-selftest.c
@@ -1175,6 +1175,7 @@ void locking_selftest(void)
printk(" --------------------------------------------------------------------------\n");
+#ifndef CONFIG_PREEMPT_RT_FULL
/*
* irq-context testcases:
*/
@@ -1187,6 +1188,28 @@ void locking_selftest(void)
DO_TESTCASE_6x2("irq read-recursion", irq_read_recursion);
// DO_TESTCASE_6x2B("irq read-recursion #2", irq_read_recursion2);
+#else
+ /* On -rt, we only do hardirq context test for raw spinlock */
+ DO_TESTCASE_1B("hard-irqs-on + irq-safe-A", irqsafe1_hard_spin, 12);
+ DO_TESTCASE_1B("hard-irqs-on + irq-safe-A", irqsafe1_hard_spin, 21);
+
+ DO_TESTCASE_1B("hard-safe-A + irqs-on", irqsafe2B_hard_spin, 12);
+ DO_TESTCASE_1B("hard-safe-A + irqs-on", irqsafe2B_hard_spin, 21);
+
+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 123);
+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 132);
+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 213);
+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 231);
+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 312);
+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #1", irqsafe3_hard_spin, 321);
+
+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 123);
+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 132);
+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 213);
+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 231);
+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 312);
+ DO_TESTCASE_1B("hard-safe-A + unsafe-B #2", irqsafe4_hard_spin, 321);
+#endif
if (unexpected_testcase_failures) {
printk("-----------------------------------------------------------------\n");
--
1.7.5.4
^ permalink raw reply related [flat|nested] 4+ messages in thread
* [PATCH 3/3 -rt] lockdep: selftest: save/restore migrate_disable count
2012-04-16 7:01 [PATCH 0/2 -rt] make locckdep selftest work on -rt Yong Zhang
2012-04-16 7:01 ` [PATCH 1/2 -rt] lockdep: selftest: convert spinlock to raw spinlock Yong Zhang
2012-04-16 7:01 ` [PATCH 2/2 -rt] lockdep: selftest: Only do hardirq context test for " Yong Zhang
@ 2012-04-17 7:23 ` Yong Zhang
2 siblings, 0 replies; 4+ messages in thread
From: Yong Zhang @ 2012-04-17 7:23 UTC (permalink / raw)
To: linux-rt-users, linux-kernel; +Cc: tglx, Yong Zhang
From: Yong Zhang <yong.zhang@windriver.com>
->migrate_disable_atomic will corrupt when doing some
case, such as double lock...
->migrate_disable will not be affected for now since
we have preempt disabled when doing seft test.
Just record it also for safe.
Signed-off-by: Yong Zhang <yong.zhang0@gmail.com>
---
Found it when enable CONFIG_SCHED_DEBUG.
lib/locking-selftest.c | 13 +++++++++++++
1 files changed, 13 insertions(+), 0 deletions(-)
diff --git a/lib/locking-selftest.c b/lib/locking-selftest.c
index 23b8564..a9cb9e4 100644
--- a/lib/locking-selftest.c
+++ b/lib/locking-selftest.c
@@ -939,6 +939,12 @@ static void dotest(void (*testcase_fn)(void), int expected, int lockclass_mask)
{
unsigned long saved_preempt_count = preempt_count();
int expected_failure = 0;
+#ifdef CONFIG_PREEMPT_RT_FULL
+ int saved_migrate_disable = current->migrate_disable;
+# ifdef CONFIG_SCHED_DEBUG
+ int saved_migrate_disable_atomic = current->migrate_disable_atomic;
+# endif
+#endif
WARN_ON(irqs_disabled());
@@ -980,6 +986,13 @@ static void dotest(void (*testcase_fn)(void), int expected, int lockclass_mask)
* count, so restore it:
*/
preempt_count() = saved_preempt_count;
+#ifdef CONFIG_PREEMPT_RT_FULL
+ current->migrate_disable = saved_migrate_disable;
+# ifdef CONFIG_SCHED_DEBUG
+ current->migrate_disable_atomic = saved_migrate_disable_atomic;
+# endif
+#endif
+
#ifdef CONFIG_TRACE_IRQFLAGS
if (softirq_count())
current->softirqs_enabled = 0;
--
1.7.5.4
^ permalink raw reply related [flat|nested] 4+ messages in thread
end of thread, other threads:[~2012-04-17 7:23 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2012-04-16 7:01 [PATCH 0/2 -rt] make locckdep selftest work on -rt Yong Zhang
2012-04-16 7:01 ` [PATCH 1/2 -rt] lockdep: selftest: convert spinlock to raw spinlock Yong Zhang
2012-04-16 7:01 ` [PATCH 2/2 -rt] lockdep: selftest: Only do hardirq context test for " Yong Zhang
2012-04-17 7:23 ` [PATCH 3/3 -rt] lockdep: selftest: save/restore migrate_disable count Yong Zhang
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).