diff options
Diffstat (limited to 'kernel')
| -rw-r--r-- | kernel/cpuset.c | 24 | ||||
| -rw-r--r-- | kernel/exit.c | 4 | ||||
| -rw-r--r-- | kernel/irq/handle.c | 2 | ||||
| -rw-r--r-- | kernel/module.c | 6 | ||||
| -rw-r--r-- | kernel/posix-timers.c | 1 | ||||
| -rw-r--r-- | kernel/sched.c | 7 | ||||
| -rw-r--r-- | kernel/signal.c | 11 | ||||
| -rw-r--r-- | kernel/spinlock.c | 8 |
8 files changed, 46 insertions, 17 deletions
diff --git a/kernel/cpuset.c b/kernel/cpuset.c index 961d74044deb..00e8f2575512 100644 --- a/kernel/cpuset.c +++ b/kernel/cpuset.c @@ -166,9 +166,8 @@ static struct super_block *cpuset_sb = NULL; * The hooks from fork and exit, cpuset_fork() and cpuset_exit(), don't * (usually) grab cpuset_sem. These are the two most performance * critical pieces of code here. The exception occurs on exit(), - * if the last task using a cpuset exits, and the cpuset was marked - * notify_on_release. In that case, the cpuset_sem is taken, the - * path to the released cpuset calculated, and a usermode call made + * when a task in a notify_on_release cpuset exits. Then cpuset_sem + * is taken, and if the cpuset count is zero, a usermode call made * to /sbin/cpuset_release_agent with the name of the cpuset (path * relative to the root of cpuset file system) as the argument. * @@ -1404,6 +1403,18 @@ void cpuset_fork(struct task_struct *tsk) * * Description: Detach cpuset from @tsk and release it. * + * Note that cpusets marked notify_on_release force every task + * in them to take the global cpuset_sem semaphore when exiting. + * This could impact scaling on very large systems. Be reluctant + * to use notify_on_release cpusets where very high task exit + * scaling is required on large systems. + * + * Don't even think about derefencing 'cs' after the cpuset use + * count goes to zero, except inside a critical section guarded + * by the cpuset_sem semaphore. If you don't hold cpuset_sem, + * then a zero cpuset use count is a license to any other task to + * nuke the cpuset immediately. + * **/ void cpuset_exit(struct task_struct *tsk) @@ -1415,10 +1426,13 @@ void cpuset_exit(struct task_struct *tsk) tsk->cpuset = NULL; task_unlock(tsk); - if (atomic_dec_and_test(&cs->count)) { + if (notify_on_release(cs)) { down(&cpuset_sem); - check_for_release(cs); + if (atomic_dec_and_test(&cs->count)) + check_for_release(cs); up(&cpuset_sem); + } else { + atomic_dec(&cs->count); } } diff --git a/kernel/exit.c b/kernel/exit.c index edaa50b5bbfa..2ef2ad540201 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -811,10 +811,8 @@ fastcall NORET_TYPE void do_exit(long code) acct_update_integrals(tsk); update_mem_hiwater(tsk); group_dead = atomic_dec_and_test(&tsk->signal->live); - if (group_dead) { - del_timer_sync(&tsk->signal->real_timer); + if (group_dead) acct_process(code); - } exit_mm(tsk); exit_sem(tsk); diff --git a/kernel/irq/handle.c b/kernel/irq/handle.c index 06b5a6323998..436c7d93c00a 100644 --- a/kernel/irq/handle.c +++ b/kernel/irq/handle.c @@ -119,8 +119,6 @@ fastcall unsigned int __do_IRQ(unsigned int irq, struct pt_regs *regs) */ desc->handler->ack(irq); action_ret = handle_IRQ_event(irq, regs, desc->action); - if (!noirqdebug) - note_interrupt(irq, desc, action_ret); desc->handler->end(irq); return 1; } diff --git a/kernel/module.c b/kernel/module.c index 5734ab09d3f9..83b3d376708c 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -1758,6 +1758,7 @@ sys_init_module(void __user *umod, const char __user *uargs) { struct module *mod; + mm_segment_t old_fs = get_fs(); int ret = 0; /* Must have permission */ @@ -1775,6 +1776,9 @@ sys_init_module(void __user *umod, return PTR_ERR(mod); } + /* flush the icache in correct context */ + set_fs(KERNEL_DS); + /* Flush the instruction cache, since we've played with text */ if (mod->module_init) flush_icache_range((unsigned long)mod->module_init, @@ -1783,6 +1787,8 @@ sys_init_module(void __user *umod, flush_icache_range((unsigned long)mod->module_core, (unsigned long)mod->module_core + mod->core_size); + set_fs(old_fs); + /* Now sew it into the lists. They won't access us, since strong_try_module_get() will fail. */ stop_machine_run(__link_module, mod, NR_CPUS); diff --git a/kernel/posix-timers.c b/kernel/posix-timers.c index fd316c272260..cabb63fc9e16 100644 --- a/kernel/posix-timers.c +++ b/kernel/posix-timers.c @@ -1197,6 +1197,7 @@ void exit_itimers(struct signal_struct *sig) tmr = list_entry(sig->posix_timers.next, struct k_itimer, list); itimer_delete(tmr); } + del_timer_sync(&sig->real_timer); } /* diff --git a/kernel/sched.c b/kernel/sched.c index 66b2ed784822..f12a0c8a7d98 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -3755,19 +3755,22 @@ EXPORT_SYMBOL(cond_resched); */ int cond_resched_lock(spinlock_t * lock) { + int ret = 0; + if (need_lockbreak(lock)) { spin_unlock(lock); cpu_relax(); + ret = 1; spin_lock(lock); } if (need_resched()) { _raw_spin_unlock(lock); preempt_enable_no_resched(); __cond_resched(); + ret = 1; spin_lock(lock); - return 1; } - return 0; + return ret; } EXPORT_SYMBOL(cond_resched_lock); diff --git a/kernel/signal.c b/kernel/signal.c index 8f3debc77c5b..b3c24c732c5a 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -522,7 +522,16 @@ static int __dequeue_signal(struct sigpending *pending, sigset_t *mask, { int sig = 0; - sig = next_signal(pending, mask); + /* SIGKILL must have priority, otherwise it is quite easy + * to create an unkillable process, sending sig < SIGKILL + * to self */ + if (unlikely(sigismember(&pending->signal, SIGKILL))) { + if (!sigismember(mask, SIGKILL)) + sig = SIGKILL; + } + + if (likely(!sig)) + sig = next_signal(pending, mask); if (sig) { if (current->notifier) { if (sigismember(current->notifier_mask, sig)) { diff --git a/kernel/spinlock.c b/kernel/spinlock.c index e15ed17863f1..0c3f9d8bbe17 100644 --- a/kernel/spinlock.c +++ b/kernel/spinlock.c @@ -294,7 +294,7 @@ EXPORT_SYMBOL(_spin_unlock_irq); void __lockfunc _spin_unlock_bh(spinlock_t *lock) { _raw_spin_unlock(lock); - preempt_enable(); + preempt_enable_no_resched(); local_bh_enable(); } EXPORT_SYMBOL(_spin_unlock_bh); @@ -318,7 +318,7 @@ EXPORT_SYMBOL(_read_unlock_irq); void __lockfunc _read_unlock_bh(rwlock_t *lock) { _raw_read_unlock(lock); - preempt_enable(); + preempt_enable_no_resched(); local_bh_enable(); } EXPORT_SYMBOL(_read_unlock_bh); @@ -342,7 +342,7 @@ EXPORT_SYMBOL(_write_unlock_irq); void __lockfunc _write_unlock_bh(rwlock_t *lock) { _raw_write_unlock(lock); - preempt_enable(); + preempt_enable_no_resched(); local_bh_enable(); } EXPORT_SYMBOL(_write_unlock_bh); @@ -354,7 +354,7 @@ int __lockfunc _spin_trylock_bh(spinlock_t *lock) if (_raw_spin_trylock(lock)) return 1; - preempt_enable(); + preempt_enable_no_resched(); local_bh_enable(); return 0; } |