From 59121003721a8fad11ee72e646fd9d3076b5679c Mon Sep 17 00:00:00 2001 From: Christoph Lameter Date: Thu, 23 Jun 2005 00:08:25 -0700 Subject: [PATCH] i386: Selectable Frequency of the Timer Interrupt Make the timer frequency selectable. The timer interrupt may cause bus and memory contention in large NUMA systems since the interrupt occurs on each processor HZ times per second. Signed-off-by: Christoph Lameter Signed-off-by: Shai Fultheim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/Kconfig.hz | 46 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 46 insertions(+) create mode 100644 kernel/Kconfig.hz (limited to 'kernel') diff --git a/kernel/Kconfig.hz b/kernel/Kconfig.hz new file mode 100644 index 00000000000..248e1c396f8 --- /dev/null +++ b/kernel/Kconfig.hz @@ -0,0 +1,46 @@ +# +# Timer Interrupt Frequency Configuration +# + +choice + prompt "Timer frequency" + default HZ_250 + help + Allows the configuration of the timer frequency. It is customary + to have the timer interrupt run at 1000 HZ but 100 HZ may be more + beneficial for servers and NUMA systems that do not need to have + a fast response for user interaction and that may experience bus + contention and cacheline bounces as a result of timer interrupts. + Note that the timer interrupt occurs on each processor in an SMP + environment leading to NR_CPUS * HZ number of timer interrupts + per second. + + + config HZ_100 + bool "100 HZ" + help + 100 HZ is a typical choice for servers, SMP and NUMA systems + with lots of processors that may show reduced performance if + too many timer interrupts are occurring. + + config HZ_250 + bool "250 HZ" + help + 250 HZ is a good compromise choice allowing server performance + while also showing good interactive responsiveness even + on SMP and NUMA systems. + + config HZ_1000 + bool "1000 HZ" + help + 1000 HZ is the preferred choice for desktop systems and other + systems requiring fast interactive responses to events. + +endchoice + +config HZ + int + default 100 if HZ_100 + default 250 if HZ_250 + default 1000 if HZ_1000 + -- cgit v1.2.3-70-g09d2 From 55c888d6d09a0df236adfaf8ccf06ff5d0646775 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Thu, 23 Jun 2005 00:08:56 -0700 Subject: [PATCH] timers fixes/improvements This patch tries to solve following problems: 1. del_timer_sync() is racy. The timer can be fired again after del_timer_sync have checked all cpus and before it will recheck timer_pending(). 2. It has scalability problems. All cpus are scanned to determine if the timer is running on that cpu. With this patch del_timer_sync is O(1) and no slower than plain del_timer(pending_timer), unless it has to actually wait for completion of the currently running timer. The only restriction is that the recurring timer should not use add_timer_on(). 3. The timers are not serialized wrt to itself. If CPU_0 does mod_timer(jiffies+1) while the timer is currently running on CPU 1, it is quite possible that local interrupt on CPU_0 will start that timer before it finished on CPU_1. 4. The timers locking is suboptimal. __mod_timer() takes 3 locks at once and still requires wmb() in del_timer/run_timers. The new implementation takes 2 locks sequentially and does not need memory barriers. Currently ->base != NULL means that the timer is pending. In that case ->base.lock is used to lock the timer. __mod_timer also takes timer->lock because ->base can be == NULL. This patch uses timer->entry.next != NULL as indication that the timer is pending. So it does __list_del(), entry->next = NULL instead of list_del() when the timer is deleted. The ->base field is used for hashed locking only, it is initialized in init_timer() which sets ->base = per_cpu(tvec_bases). When the tvec_bases.lock is locked, it means that all timers which are tied to this base via timer->base are locked, and the base itself is locked too. So __run_timers/migrate_timers can safely modify all timers which could be found on ->tvX lists (pending timers). When the timer's base is locked, and the timer removed from ->entry list (which means that _run_timers/migrate_timers can't see this timer), it is possible to set timer->base = NULL and drop the lock: the timer remains locked. This patch adds lock_timer_base() helper, which waits for ->base != NULL, locks the ->base, and checks it is still the same. __mod_timer() schedules the timer on the local CPU and changes it's base. However, it does not lock both old and new bases at once. It locks the timer via lock_timer_base(), deletes the timer, sets ->base = NULL, and unlocks old base. Then __mod_timer() locks new_base, sets ->base = new_base, and adds this timer. This simplifies the code, because AB-BA deadlock is not possible. __mod_timer() also ensures that the timer's base is not changed while the timer's handler is running on the old base. __run_timers(), del_timer() do not change ->base anymore, they only clear pending flag. So del_timer_sync() can test timer->base->running_timer == timer to detect whether it is running or not. We don't need timer_list->lock anymore, this patch kills it. We also don't need barriers. del_timer() and __run_timers() used smp_wmb() before clearing timer's pending flag. It was needed because __mod_timer() did not lock old_base if the timer is not pending, so __mod_timer()->list_add() could race with del_timer()->list_del(). With this patch these functions are serialized through base->lock. One problem. TIMER_INITIALIZER can't use per_cpu(tvec_bases). So this patch adds global struct timer_base_s { spinlock_t lock; struct timer_list *running_timer; } __init_timer_base; which is used by TIMER_INITIALIZER. The corresponding fields in tvec_t_base_s struct are replaced by struct timer_base_s t_base. It is indeed ugly. But this can't have scalability problems. The global __init_timer_base.lock is used only when __mod_timer() is called for the first time AND the timer was compile time initialized. After that the timer migrates to the local CPU. Signed-off-by: Oleg Nesterov Acked-by: Ingo Molnar Signed-off-by: Renaud Lienhart Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/timer.h | 30 ++--- kernel/timer.c | 328 ++++++++++++++++++++++++-------------------------- 2 files changed, 166 insertions(+), 192 deletions(-) (limited to 'kernel') diff --git a/include/linux/timer.h b/include/linux/timer.h index 90db1cc62dd..2e78fedfc06 100644 --- a/include/linux/timer.h +++ b/include/linux/timer.h @@ -6,45 +6,33 @@ #include #include -struct tvec_t_base_s; +struct timer_base_s; struct timer_list { struct list_head entry; unsigned long expires; - spinlock_t lock; unsigned long magic; void (*function)(unsigned long); unsigned long data; - struct tvec_t_base_s *base; + struct timer_base_s *base; }; #define TIMER_MAGIC 0x4b87ad6e +extern struct timer_base_s __init_timer_base; + #define TIMER_INITIALIZER(_function, _expires, _data) { \ .function = (_function), \ .expires = (_expires), \ .data = (_data), \ - .base = NULL, \ + .base = &__init_timer_base, \ .magic = TIMER_MAGIC, \ - .lock = SPIN_LOCK_UNLOCKED, \ } -/*** - * init_timer - initialize a timer. - * @timer: the timer to be initialized - * - * init_timer() must be done to a timer prior calling *any* of the - * other timer functions. - */ -static inline void init_timer(struct timer_list * timer) -{ - timer->base = NULL; - timer->magic = TIMER_MAGIC; - spin_lock_init(&timer->lock); -} +void fastcall init_timer(struct timer_list * timer); /*** * timer_pending - is a timer pending? @@ -58,7 +46,7 @@ static inline void init_timer(struct timer_list * timer) */ static inline int timer_pending(const struct timer_list * timer) { - return timer->base != NULL; + return timer->entry.next != NULL; } extern void add_timer_on(struct timer_list *timer, int cpu); @@ -89,12 +77,12 @@ static inline void add_timer(struct timer_list * timer) #ifdef CONFIG_SMP extern int del_timer_sync(struct timer_list *timer); - extern int del_singleshot_timer_sync(struct timer_list *timer); #else # define del_timer_sync(t) del_timer(t) -# define del_singleshot_timer_sync(t) del_timer(t) #endif +#define del_singleshot_timer_sync(t) del_timer_sync(t) + extern void init_timers(void); extern void run_local_timers(void); extern void it_real_fn(unsigned long); diff --git a/kernel/timer.c b/kernel/timer.c index 207aa4f0aa1..8aadc62efd6 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -57,6 +57,11 @@ static void time_interpolator_update(long delta_nsec); #define TVN_MASK (TVN_SIZE - 1) #define TVR_MASK (TVR_SIZE - 1) +struct timer_base_s { + spinlock_t lock; + struct timer_list *running_timer; +}; + typedef struct tvec_s { struct list_head vec[TVN_SIZE]; } tvec_t; @@ -66,9 +71,8 @@ typedef struct tvec_root_s { } tvec_root_t; struct tvec_t_base_s { - spinlock_t lock; + struct timer_base_s t_base; unsigned long timer_jiffies; - struct timer_list *running_timer; tvec_root_t tv1; tvec_t tv2; tvec_t tv3; @@ -77,18 +81,16 @@ struct tvec_t_base_s { } ____cacheline_aligned_in_smp; typedef struct tvec_t_base_s tvec_base_t; +static DEFINE_PER_CPU(tvec_base_t, tvec_bases); static inline void set_running_timer(tvec_base_t *base, struct timer_list *timer) { #ifdef CONFIG_SMP - base->running_timer = timer; + base->t_base.running_timer = timer; #endif } -/* Fake initialization */ -static DEFINE_PER_CPU(tvec_base_t, tvec_bases) = { SPIN_LOCK_UNLOCKED }; - static void check_timer_failed(struct timer_list *timer) { static int whine_count; @@ -103,7 +105,6 @@ static void check_timer_failed(struct timer_list *timer) /* * Now fix it up */ - spin_lock_init(&timer->lock); timer->magic = TIMER_MAGIC; } @@ -156,65 +157,113 @@ static void internal_add_timer(tvec_base_t *base, struct timer_list *timer) list_add_tail(&timer->entry, vec); } +typedef struct timer_base_s timer_base_t; +/* + * Used by TIMER_INITIALIZER, we can't use per_cpu(tvec_bases) + * at compile time, and we need timer->base to lock the timer. + */ +timer_base_t __init_timer_base + ____cacheline_aligned_in_smp = { .lock = SPIN_LOCK_UNLOCKED }; +EXPORT_SYMBOL(__init_timer_base); + +/*** + * init_timer - initialize a timer. + * @timer: the timer to be initialized + * + * init_timer() must be done to a timer prior calling *any* of the + * other timer functions. + */ +void fastcall init_timer(struct timer_list *timer) +{ + timer->entry.next = NULL; + timer->base = &per_cpu(tvec_bases, raw_smp_processor_id()).t_base; + timer->magic = TIMER_MAGIC; +} +EXPORT_SYMBOL(init_timer); + +static inline void detach_timer(struct timer_list *timer, + int clear_pending) +{ + struct list_head *entry = &timer->entry; + + __list_del(entry->prev, entry->next); + if (clear_pending) + entry->next = NULL; + entry->prev = LIST_POISON2; +} + +/* + * We are using hashed locking: holding per_cpu(tvec_bases).t_base.lock + * means that all timers which are tied to this base via timer->base are + * locked, and the base itself is locked too. + * + * So __run_timers/migrate_timers can safely modify all timers which could + * be found on ->tvX lists. + * + * When the timer's base is locked, and the timer removed from list, it is + * possible to set timer->base = NULL and drop the lock: the timer remains + * locked. + */ +static timer_base_t *lock_timer_base(struct timer_list *timer, + unsigned long *flags) +{ + timer_base_t *base; + + for (;;) { + base = timer->base; + if (likely(base != NULL)) { + spin_lock_irqsave(&base->lock, *flags); + if (likely(base == timer->base)) + return base; + /* The timer has migrated to another CPU */ + spin_unlock_irqrestore(&base->lock, *flags); + } + cpu_relax(); + } +} + int __mod_timer(struct timer_list *timer, unsigned long expires) { - tvec_base_t *old_base, *new_base; + timer_base_t *base; + tvec_base_t *new_base; unsigned long flags; int ret = 0; BUG_ON(!timer->function); - check_timer(timer); - spin_lock_irqsave(&timer->lock, flags); + base = lock_timer_base(timer, &flags); + + if (timer_pending(timer)) { + detach_timer(timer, 0); + ret = 1; + } + new_base = &__get_cpu_var(tvec_bases); -repeat: - old_base = timer->base; - /* - * Prevent deadlocks via ordering by old_base < new_base. - */ - if (old_base && (new_base != old_base)) { - if (old_base < new_base) { - spin_lock(&new_base->lock); - spin_lock(&old_base->lock); - } else { - spin_lock(&old_base->lock); - spin_lock(&new_base->lock); - } + if (base != &new_base->t_base) { /* - * The timer base might have been cancelled while we were - * trying to take the lock(s): + * We are trying to schedule the timer on the local CPU. + * However we can't change timer's base while it is running, + * otherwise del_timer_sync() can't detect that the timer's + * handler yet has not finished. This also guarantees that + * the timer is serialized wrt itself. */ - if (timer->base != old_base) { - spin_unlock(&new_base->lock); - spin_unlock(&old_base->lock); - goto repeat; - } - } else { - spin_lock(&new_base->lock); - if (timer->base != old_base) { - spin_unlock(&new_base->lock); - goto repeat; + if (unlikely(base->running_timer == timer)) { + /* The timer remains on a former base */ + new_base = container_of(base, tvec_base_t, t_base); + } else { + /* See the comment in lock_timer_base() */ + timer->base = NULL; + spin_unlock(&base->lock); + spin_lock(&new_base->t_base.lock); + timer->base = &new_base->t_base; } } - /* - * Delete the previous timeout (if there was any), and install - * the new one: - */ - if (old_base) { - list_del(&timer->entry); - ret = 1; - } timer->expires = expires; internal_add_timer(new_base, timer); - timer->base = new_base; - - if (old_base && (new_base != old_base)) - spin_unlock(&old_base->lock); - spin_unlock(&new_base->lock); - spin_unlock_irqrestore(&timer->lock, flags); + spin_unlock_irqrestore(&new_base->t_base.lock, flags); return ret; } @@ -232,15 +281,15 @@ void add_timer_on(struct timer_list *timer, int cpu) { tvec_base_t *base = &per_cpu(tvec_bases, cpu); unsigned long flags; - + BUG_ON(timer_pending(timer) || !timer->function); check_timer(timer); - spin_lock_irqsave(&base->lock, flags); + spin_lock_irqsave(&base->t_base.lock, flags); + timer->base = &base->t_base; internal_add_timer(base, timer); - timer->base = base; - spin_unlock_irqrestore(&base->lock, flags); + spin_unlock_irqrestore(&base->t_base.lock, flags); } @@ -295,27 +344,22 @@ EXPORT_SYMBOL(mod_timer); */ int del_timer(struct timer_list *timer) { + timer_base_t *base; unsigned long flags; - tvec_base_t *base; + int ret = 0; check_timer(timer); -repeat: - base = timer->base; - if (!base) - return 0; - spin_lock_irqsave(&base->lock, flags); - if (base != timer->base) { + if (timer_pending(timer)) { + base = lock_timer_base(timer, &flags); + if (timer_pending(timer)) { + detach_timer(timer, 1); + ret = 1; + } spin_unlock_irqrestore(&base->lock, flags); - goto repeat; } - list_del(&timer->entry); - /* Need to make sure that anybody who sees a NULL base also sees the list ops */ - smp_wmb(); - timer->base = NULL; - spin_unlock_irqrestore(&base->lock, flags); - return 1; + return ret; } EXPORT_SYMBOL(del_timer); @@ -332,72 +376,39 @@ EXPORT_SYMBOL(del_timer); * Synchronization rules: callers must prevent restarting of the timer, * otherwise this function is meaningless. It must not be called from * interrupt contexts. The caller must not hold locks which would prevent - * completion of the timer's handler. Upon exit the timer is not queued and - * the handler is not running on any CPU. + * completion of the timer's handler. The timer's handler must not call + * add_timer_on(). Upon exit the timer is not queued and the handler is + * not running on any CPU. * * The function returns whether it has deactivated a pending timer or not. - * - * del_timer_sync() is slow and complicated because it copes with timer - * handlers which re-arm the timer (periodic timers). If the timer handler - * is known to not do this (a single shot timer) then use - * del_singleshot_timer_sync() instead. */ int del_timer_sync(struct timer_list *timer) { - tvec_base_t *base; - int i, ret = 0; + timer_base_t *base; + unsigned long flags; + int ret = -1; check_timer(timer); -del_again: - ret += del_timer(timer); + do { + base = lock_timer_base(timer, &flags); - for_each_online_cpu(i) { - base = &per_cpu(tvec_bases, i); - if (base->running_timer == timer) { - while (base->running_timer == timer) { - cpu_relax(); - preempt_check_resched(); - } - break; + if (base->running_timer == timer) + goto unlock; + + ret = 0; + if (timer_pending(timer)) { + detach_timer(timer, 1); + ret = 1; } - } - smp_rmb(); - if (timer_pending(timer)) - goto del_again; +unlock: + spin_unlock_irqrestore(&base->lock, flags); + } while (ret < 0); return ret; } -EXPORT_SYMBOL(del_timer_sync); -/*** - * del_singleshot_timer_sync - deactivate a non-recursive timer - * @timer: the timer to be deactivated - * - * This function is an optimization of del_timer_sync for the case where the - * caller can guarantee the timer does not reschedule itself in its timer - * function. - * - * Synchronization rules: callers must prevent restarting of the timer, - * otherwise this function is meaningless. It must not be called from - * interrupt contexts. The caller must not hold locks which wold prevent - * completion of the timer's handler. Upon exit the timer is not queued and - * the handler is not running on any CPU. - * - * The function returns whether it has deactivated a pending timer or not. - */ -int del_singleshot_timer_sync(struct timer_list *timer) -{ - int ret = del_timer(timer); - - if (!ret) { - ret = del_timer_sync(timer); - BUG_ON(ret); - } - - return ret; -} -EXPORT_SYMBOL(del_singleshot_timer_sync); +EXPORT_SYMBOL(del_timer_sync); #endif static int cascade(tvec_base_t *base, tvec_t *tv, int index) @@ -415,7 +426,7 @@ static int cascade(tvec_base_t *base, tvec_t *tv, int index) struct timer_list *tmp; tmp = list_entry(curr, struct timer_list, entry); - BUG_ON(tmp->base != base); + BUG_ON(tmp->base != &base->t_base); curr = curr->next; internal_add_timer(base, tmp); } @@ -437,7 +448,7 @@ static inline void __run_timers(tvec_base_t *base) { struct timer_list *timer; - spin_lock_irq(&base->lock); + spin_lock_irq(&base->t_base.lock); while (time_after_eq(jiffies, base->timer_jiffies)) { struct list_head work_list = LIST_HEAD_INIT(work_list); struct list_head *head = &work_list; @@ -453,8 +464,7 @@ static inline void __run_timers(tvec_base_t *base) cascade(base, &base->tv5, INDEX(3)); ++base->timer_jiffies; list_splice_init(base->tv1.vec + index, &work_list); -repeat: - if (!list_empty(head)) { + while (!list_empty(head)) { void (*fn)(unsigned long); unsigned long data; @@ -462,11 +472,9 @@ repeat: fn = timer->function; data = timer->data; - list_del(&timer->entry); set_running_timer(base, timer); - smp_wmb(); - timer->base = NULL; - spin_unlock_irq(&base->lock); + detach_timer(timer, 1); + spin_unlock_irq(&base->t_base.lock); { u32 preempt_count = preempt_count(); fn(data); @@ -475,12 +483,11 @@ repeat: BUG(); } } - spin_lock_irq(&base->lock); - goto repeat; + spin_lock_irq(&base->t_base.lock); } } set_running_timer(base, NULL); - spin_unlock_irq(&base->lock); + spin_unlock_irq(&base->t_base.lock); } #ifdef CONFIG_NO_IDLE_HZ @@ -499,7 +506,7 @@ unsigned long next_timer_interrupt(void) int i, j; base = &__get_cpu_var(tvec_bases); - spin_lock(&base->lock); + spin_lock(&base->t_base.lock); expires = base->timer_jiffies + (LONG_MAX >> 1); list = 0; @@ -547,7 +554,7 @@ found: expires = nte->expires; } } - spin_unlock(&base->lock); + spin_unlock(&base->t_base.lock); return expires; } #endif @@ -1286,9 +1293,9 @@ static void __devinit init_timers_cpu(int cpu) { int j; tvec_base_t *base; - + base = &per_cpu(tvec_bases, cpu); - spin_lock_init(&base->lock); + spin_lock_init(&base->t_base.lock); for (j = 0; j < TVN_SIZE; j++) { INIT_LIST_HEAD(base->tv5.vec + j); INIT_LIST_HEAD(base->tv4.vec + j); @@ -1302,22 +1309,16 @@ static void __devinit init_timers_cpu(int cpu) } #ifdef CONFIG_HOTPLUG_CPU -static int migrate_timer_list(tvec_base_t *new_base, struct list_head *head) +static void migrate_timer_list(tvec_base_t *new_base, struct list_head *head) { struct timer_list *timer; while (!list_empty(head)) { timer = list_entry(head->next, struct timer_list, entry); - /* We're locking backwards from __mod_timer order here, - beware deadlock. */ - if (!spin_trylock(&timer->lock)) - return 0; - list_del(&timer->entry); + detach_timer(timer, 0); + timer->base = &new_base->t_base; internal_add_timer(new_base, timer); - timer->base = new_base; - spin_unlock(&timer->lock); } - return 1; } static void __devinit migrate_timers(int cpu) @@ -1331,39 +1332,24 @@ static void __devinit migrate_timers(int cpu) new_base = &get_cpu_var(tvec_bases); local_irq_disable(); -again: - /* Prevent deadlocks via ordering by old_base < new_base. */ - if (old_base < new_base) { - spin_lock(&new_base->lock); - spin_lock(&old_base->lock); - } else { - spin_lock(&old_base->lock); - spin_lock(&new_base->lock); - } + spin_lock(&new_base->t_base.lock); + spin_lock(&old_base->t_base.lock); - if (old_base->running_timer) + if (old_base->t_base.running_timer) BUG(); for (i = 0; i < TVR_SIZE; i++) - if (!migrate_timer_list(new_base, old_base->tv1.vec + i)) - goto unlock_again; - for (i = 0; i < TVN_SIZE; i++) - if (!migrate_timer_list(new_base, old_base->tv2.vec + i) - || !migrate_timer_list(new_base, old_base->tv3.vec + i) - || !migrate_timer_list(new_base, old_base->tv4.vec + i) - || !migrate_timer_list(new_base, old_base->tv5.vec + i)) - goto unlock_again; - spin_unlock(&old_base->lock); - spin_unlock(&new_base->lock); + migrate_timer_list(new_base, old_base->tv1.vec + i); + for (i = 0; i < TVN_SIZE; i++) { + migrate_timer_list(new_base, old_base->tv2.vec + i); + migrate_timer_list(new_base, old_base->tv3.vec + i); + migrate_timer_list(new_base, old_base->tv4.vec + i); + migrate_timer_list(new_base, old_base->tv5.vec + i); + } + + spin_unlock(&old_base->t_base.lock); + spin_unlock(&new_base->t_base.lock); local_irq_enable(); put_cpu_var(tvec_bases); - return; - -unlock_again: - /* Avoid deadlock with __mod_timer, by backing off. */ - spin_unlock(&old_base->lock); - spin_unlock(&new_base->lock); - cpu_relax(); - goto again; } #endif /* CONFIG_HOTPLUG_CPU */ -- cgit v1.2.3-70-g09d2 From fd450b7318b75343fd76b3d95416853e34e72c95 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Thu, 23 Jun 2005 00:08:59 -0700 Subject: [PATCH] timers: introduce try_to_del_timer_sync() This patch splits del_timer_sync() into 2 functions. The new one, try_to_del_timer_sync(), returns -1 when it hits executing timer. It can be used in interrupt context, or when the caller hold locks which can prevent completion of the timer's handler. NOTE. Currently it can't be used in interrupt context in UP case, because ->running_timer is used only with CONFIG_SMP. Should the need arise, it is possible to kill #ifdef CONFIG_SMP in set_running_timer(), it is cheap. Signed-off-by: Oleg Nesterov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/timer.h | 4 +++- kernel/timer.c | 53 ++++++++++++++++++++++++++++++++------------------- 2 files changed, 36 insertions(+), 21 deletions(-) (limited to 'kernel') diff --git a/include/linux/timer.h b/include/linux/timer.h index 2e78fedfc06..221f81ac200 100644 --- a/include/linux/timer.h +++ b/include/linux/timer.h @@ -76,9 +76,11 @@ static inline void add_timer(struct timer_list * timer) } #ifdef CONFIG_SMP + extern int try_to_del_timer_sync(struct timer_list *timer); extern int del_timer_sync(struct timer_list *timer); #else -# define del_timer_sync(t) del_timer(t) +# define try_to_del_timer_sync(t) del_timer(t) +# define del_timer_sync(t) del_timer(t) #endif #define del_singleshot_timer_sync(t) del_timer_sync(t) diff --git a/kernel/timer.c b/kernel/timer.c index 8aadc62efd6..1f986c16d89 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -365,6 +365,34 @@ int del_timer(struct timer_list *timer) EXPORT_SYMBOL(del_timer); #ifdef CONFIG_SMP +/* + * This function tries to deactivate a timer. Upon successful (ret >= 0) + * exit the timer is not queued and the handler is not running on any CPU. + * + * It must not be called from interrupt contexts. + */ +int try_to_del_timer_sync(struct timer_list *timer) +{ + timer_base_t *base; + unsigned long flags; + int ret = -1; + + base = lock_timer_base(timer, &flags); + + if (base->running_timer == timer) + goto out; + + ret = 0; + if (timer_pending(timer)) { + detach_timer(timer, 1); + ret = 1; + } +out: + spin_unlock_irqrestore(&base->lock, flags); + + return ret; +} + /*** * del_timer_sync - deactivate a timer and wait for the handler to finish. * @timer: the timer to be deactivated @@ -384,28 +412,13 @@ EXPORT_SYMBOL(del_timer); */ int del_timer_sync(struct timer_list *timer) { - timer_base_t *base; - unsigned long flags; - int ret = -1; - check_timer(timer); - do { - base = lock_timer_base(timer, &flags); - - if (base->running_timer == timer) - goto unlock; - - ret = 0; - if (timer_pending(timer)) { - detach_timer(timer, 1); - ret = 1; - } -unlock: - spin_unlock_irqrestore(&base->lock, flags); - } while (ret < 0); - - return ret; + for (;;) { + int ret = try_to_del_timer_sync(timer); + if (ret >= 0) + return ret; + } } EXPORT_SYMBOL(del_timer_sync); -- cgit v1.2.3-70-g09d2 From f972be33ce6a08b5f096ba013c7459a3a82f5f39 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Thu, 23 Jun 2005 00:09:00 -0700 Subject: [PATCH] posix-timers: use try_to_del_timer_sync() sys_timer_settime/sys_timer_delete needs to delete k_itimer->real.timer synchronously while holding ->it_lock, which is also locked in posix_timer_fn. This patch removes timer_active/set_timer_inactive which plays with timer_list's internals in favour of using try_to_del_timer_sync(), which was introduced in the previous patch. Signed-off-by: Oleg Nesterov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/posix-timers.c | 34 +++++++--------------------------- 1 file changed, 7 insertions(+), 27 deletions(-) (limited to 'kernel') diff --git a/kernel/posix-timers.c b/kernel/posix-timers.c index cabb63fc9e1..5b7b4736d82 100644 --- a/kernel/posix-timers.c +++ b/kernel/posix-timers.c @@ -88,23 +88,6 @@ static kmem_cache_t *posix_timers_cache; static struct idr posix_timers_id; static DEFINE_SPINLOCK(idr_lock); -/* - * Just because the timer is not in the timer list does NOT mean it is - * inactive. It could be in the "fire" routine getting a new expire time. - */ -#define TIMER_INACTIVE 1 - -#ifdef CONFIG_SMP -# define timer_active(tmr) \ - ((tmr)->it.real.timer.entry.prev != (void *)TIMER_INACTIVE) -# define set_timer_inactive(tmr) \ - do { \ - (tmr)->it.real.timer.entry.prev = (void *)TIMER_INACTIVE; \ - } while (0) -#else -# define timer_active(tmr) BARFY // error to use outside of SMP -# define set_timer_inactive(tmr) do { } while (0) -#endif /* * we assume that the new SIGEV_THREAD_ID shares no bits with the other * SIGEV values. Here we put out an error if this assumption fails. @@ -226,7 +209,6 @@ static inline int common_timer_create(struct k_itimer *new_timer) init_timer(&new_timer->it.real.timer); new_timer->it.real.timer.data = (unsigned long) new_timer; new_timer->it.real.timer.function = posix_timer_fn; - set_timer_inactive(new_timer); return 0; } @@ -480,7 +462,6 @@ static void posix_timer_fn(unsigned long __data) int do_notify = 1; spin_lock_irqsave(&timr->it_lock, flags); - set_timer_inactive(timr); if (!list_empty(&timr->it.real.abs_timer_entry)) { spin_lock(&abs_list.lock); do { @@ -983,8 +964,8 @@ common_timer_set(struct k_itimer *timr, int flags, * careful here. If smp we could be in the "fire" routine which will * be spinning as we hold the lock. But this is ONLY an SMP issue. */ + if (try_to_del_timer_sync(&timr->it.real.timer) < 0) { #ifdef CONFIG_SMP - if (timer_active(timr) && !del_timer(&timr->it.real.timer)) /* * It can only be active if on an other cpu. Since * we have cleared the interval stuff above, it should @@ -994,11 +975,9 @@ common_timer_set(struct k_itimer *timr, int flags, * a "retry" exit status. */ return TIMER_RETRY; - - set_timer_inactive(timr); -#else - del_timer(&timr->it.real.timer); #endif + } + remove_from_abslist(timr); timr->it_requeue_pending = (timr->it_requeue_pending + 2) & @@ -1083,8 +1062,9 @@ retry: static inline int common_timer_del(struct k_itimer *timer) { timer->it.real.incr = 0; + + if (try_to_del_timer_sync(&timer->it.real.timer) < 0) { #ifdef CONFIG_SMP - if (timer_active(timer) && !del_timer(&timer->it.real.timer)) /* * It can only be active if on an other cpu. Since * we have cleared the interval stuff above, it should @@ -1094,9 +1074,9 @@ static inline int common_timer_del(struct k_itimer *timer) * a "retry" exit status. */ return TIMER_RETRY; -#else - del_timer(&timer->it.real.timer); #endif + } + remove_from_abslist(timer); return 0; -- cgit v1.2.3-70-g09d2 From ab4af03a4054bd78bcabfb2214c9597201beae35 Mon Sep 17 00:00:00 2001 From: Greg Edwards Date: Thu, 23 Jun 2005 00:09:05 -0700 Subject: [PATCH] CON_CONSDEV bit not set correctly on last console According to include/linux/console.h, CON_CONSDEV flag should be set on the last console specified on the boot command line: 86 #define CON_PRINTBUFFER (1) 87 #define CON_CONSDEV (2) /* Last on the command line */ 88 #define CON_ENABLED (4) 89 #define CON_BOOT (8) This does not currently happen if there is more than one console specified on the boot commandline. Instead, it gets set on the first console on the command line. This can cause problems for things like kdb that look for the CON_CONSDEV flag to see if the console is valid. Additionaly, it doesn't look like CON_CONSDEV is reassigned to the next preferred console at unregister time if the console being unregistered currently has that bit set. Example (from sn2 ia64): elilo vmlinuz root= console=ttyS0 console=ttySG0 in this case, the flags on ttySG console struct will be 0x4 (should be 0x6). Attached patch against bk fixes both issues for the cases I looked at. It uses selected_console (which gets incremented for each console specified on the command line) as the indicator of which console to set CON_CONSDEV on. When adding the console to the list, if the previous one had CON_CONSDEV set, it masks it out. Tested on ia64 and x86. The problem with the current behavior is it breaks overriding the default from the boot line. In the ia64 case, there may be a global append line defining console=a in elilo.conf. Then you want to boot your kernel, and want to override the default by passing console=b on the boot line. elilo constructs the kernel cmdline by starting with the value of the global append line, then tacks on whatever else you specify, which puts console=b last. Signed-off-by: Greg Edwards Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/printk.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/printk.c b/kernel/printk.c index 01b58d7d17f..3a442bfb8be 100644 --- a/kernel/printk.c +++ b/kernel/printk.c @@ -876,8 +876,10 @@ void register_console(struct console * console) break; console->flags |= CON_ENABLED; console->index = console_cmdline[i].index; - if (i == preferred_console) + if (i == selected_console) { console->flags |= CON_CONSDEV; + preferred_console = selected_console; + } break; } @@ -897,6 +899,8 @@ void register_console(struct console * console) if ((console->flags & CON_CONSDEV) || console_drivers == NULL) { console->next = console_drivers; console_drivers = console; + if (console->next) + console->next->flags &= ~CON_CONSDEV; } else { console->next = console_drivers->next; console_drivers->next = console; @@ -937,10 +941,14 @@ int unregister_console(struct console * console) /* If last console is removed, we re-enable picking the first * one that gets registered. Without that, pmac early boot console * would prevent fbcon from taking over. + * + * If this isn't the last console and it has CON_CONSDEV set, we + * need to set it on the next preferred console. */ if (console_drivers == NULL) preferred_console = selected_console; - + else if (console->flags & CON_CONSDEV) + console_drivers->flags |= CON_CONSDEV; release_console_sem(); return res; -- cgit v1.2.3-70-g09d2 From be5b4fbd017d12e0d09ea0528a5839ce2ed2c8c8 Mon Sep 17 00:00:00 2001 From: Jesper Juhl Date: Thu, 23 Jun 2005 00:09:09 -0700 Subject: [PATCH] preempt_count is int - remove cast and don't assign to unsigned type In kernel/sched.c the return value from preempt_count() is cast to an int. That made sense when preempt_count was defined as different types on is not needed and should go away. The patch removes the cast. In kernel/timer.c the return value from preempt_count() is assigned to a variable of type u32 and then that unsigned value is later compared to preempt_count(). Since preempt_count() returns an int, an int is what should be used to store its return value. Storing the result in an unsigned 32bit integer made a tiny bit of sense back when preempt_count was different types on different archs, but no more - let's not play signed vs unsigned comparison games when we don't have to. The patch modifies the code to use an int to hold the value. While I was around that bit of code I also made two changes to a nearby (related) printk() - I modified it to specify the loglevel explicitly and also broke the line into a few pieces to avoid it being longer than 80 chars and clarified the text a bit. Signed-off-by: Jesper Juhl Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/sched.c | 2 +- kernel/timer.c | 8 ++++++-- 2 files changed, 7 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/sched.c b/kernel/sched.c index deca041fc36..6ee4515d5a2 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -2576,7 +2576,7 @@ void fastcall add_preempt_count(int val) /* * Underflow? */ - BUG_ON(((int)preempt_count() < 0)); + BUG_ON((preempt_count() < 0)); preempt_count() += val; /* * Spinlock count overflowing soon? diff --git a/kernel/timer.c b/kernel/timer.c index 1f986c16d89..51ff917c959 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -489,10 +489,14 @@ static inline void __run_timers(tvec_base_t *base) detach_timer(timer, 1); spin_unlock_irq(&base->t_base.lock); { - u32 preempt_count = preempt_count(); + int preempt_count = preempt_count(); fn(data); if (preempt_count != preempt_count()) { - printk("huh, entered %p with %08x, exited with %08x?\n", fn, preempt_count, preempt_count()); + printk(KERN_WARNING "huh, entered %p " + "with preempt_count %08x, exited" + " with %08x?\n", + fn, preempt_count, + preempt_count()); BUG(); } } -- cgit v1.2.3-70-g09d2 From 5f45f1a78fbac3cc859ec10c5366e97d20d40fa2 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 23 Jun 2005 00:09:12 -0700 Subject: [PATCH] remove duplicate get_dentry functions in various places Various filesystem drivers have grown a get_dentry() function that's a duplicate of lookup_one_len, except that it doesn't take a maximum length argument and doesn't check for \0 or / in the passed in filename. Switch all these places to use lookup_one_len. Signed-off-by: Christoph Hellwig Cc: Greg KH Cc: Paul Jackson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/usb/core/inode.c | 13 +------------ fs/debugfs/inode.c | 12 +----------- fs/sysfs/dir.c | 5 +++-- fs/sysfs/file.c | 5 +++-- fs/sysfs/group.c | 4 +++- fs/sysfs/inode.c | 10 ---------- fs/sysfs/sysfs.h | 1 - kernel/cpuset.c | 8 +------- 8 files changed, 12 insertions(+), 46 deletions(-) (limited to 'kernel') diff --git a/drivers/usb/core/inode.c b/drivers/usb/core/inode.c index f9f9561c6ba..c3e3a95d380 100644 --- a/drivers/usb/core/inode.c +++ b/drivers/usb/core/inode.c @@ -453,17 +453,6 @@ static int usbfs_fill_super(struct super_block *sb, void *data, int silent) return 0; } -static struct dentry * get_dentry(struct dentry *parent, const char *name) -{ - struct qstr qstr; - - qstr.name = name; - qstr.len = strlen(name); - qstr.hash = full_name_hash(name,qstr.len); - return lookup_hash(&qstr,parent); -} - - /* * fs_create_by_name - create a file, given a name * @name: name of file @@ -496,7 +485,7 @@ static int fs_create_by_name (const char *name, mode_t mode, *dentry = NULL; down(&parent->d_inode->i_sem); - *dentry = get_dentry (parent, name); + *dentry = lookup_one_len(name, parent, strlen(name)); if (!IS_ERR(dentry)) { if ((mode & S_IFMT) == S_IFDIR) error = usbfs_mkdir (parent->d_inode, *dentry, mode); diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index b529786699e..a86ac4aeaed 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -110,16 +110,6 @@ static int debug_fill_super(struct super_block *sb, void *data, int silent) return simple_fill_super(sb, DEBUGFS_MAGIC, debug_files); } -static struct dentry * get_dentry(struct dentry *parent, const char *name) -{ - struct qstr qstr; - - qstr.name = name; - qstr.len = strlen(name); - qstr.hash = full_name_hash(name,qstr.len); - return lookup_hash(&qstr,parent); -} - static struct super_block *debug_get_sb(struct file_system_type *fs_type, int flags, const char *dev_name, void *data) @@ -157,7 +147,7 @@ static int debugfs_create_by_name(const char *name, mode_t mode, *dentry = NULL; down(&parent->d_inode->i_sem); - *dentry = get_dentry (parent, name); + *dentry = lookup_one_len(name, parent, strlen(name)); if (!IS_ERR(dentry)) { if ((mode & S_IFMT) == S_IFDIR) error = debugfs_mkdir(parent->d_inode, *dentry, mode); diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index 37d7a6875d8..59734ba1ee6 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -8,6 +8,7 @@ #include #include #include +#include #include "sysfs.h" DECLARE_RWSEM(sysfs_rename_sem); @@ -99,7 +100,7 @@ static int create_dir(struct kobject * k, struct dentry * p, umode_t mode = S_IFDIR| S_IRWXU | S_IRUGO | S_IXUGO; down(&p->d_inode->i_sem); - *d = sysfs_get_dentry(p,n); + *d = lookup_one_len(n, p, strlen(n)); if (!IS_ERR(*d)) { error = sysfs_make_dirent(p->d_fsdata, *d, k, mode, SYSFS_DIR); if (!error) { @@ -315,7 +316,7 @@ int sysfs_rename_dir(struct kobject * kobj, const char *new_name) down(&parent->d_inode->i_sem); - new_dentry = sysfs_get_dentry(parent, new_name); + new_dentry = lookup_one_len(new_name, parent, strlen(new_name)); if (!IS_ERR(new_dentry)) { if (!new_dentry->d_inode) { error = kobject_set_name(kobj, "%s", new_name); diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c index 849aac11546..e9cfa39f409 100644 --- a/fs/sysfs/file.c +++ b/fs/sysfs/file.c @@ -5,6 +5,7 @@ #include #include #include +#include #include #include @@ -400,7 +401,7 @@ int sysfs_update_file(struct kobject * kobj, const struct attribute * attr) int res = -ENOENT; down(&dir->d_inode->i_sem); - victim = sysfs_get_dentry(dir, attr->name); + victim = lookup_one_len(attr->name, dir, strlen(attr->name)); if (!IS_ERR(victim)) { /* make sure dentry is really there */ if (victim->d_inode && @@ -443,7 +444,7 @@ int sysfs_chmod_file(struct kobject *kobj, struct attribute *attr, mode_t mode) int res = -ENOENT; down(&dir->d_inode->i_sem); - victim = sysfs_get_dentry(dir, attr->name); + victim = lookup_one_len(attr->name, dir, strlen(attr->name)); if (!IS_ERR(victim)) { if (victim->d_inode && (victim->d_parent->d_inode == dir->d_inode)) { diff --git a/fs/sysfs/group.c b/fs/sysfs/group.c index f11ac5ea702..122145b0895 100644 --- a/fs/sysfs/group.c +++ b/fs/sysfs/group.c @@ -11,6 +11,7 @@ #include #include #include +#include #include #include "sysfs.h" @@ -68,7 +69,8 @@ void sysfs_remove_group(struct kobject * kobj, struct dentry * dir; if (grp->name) - dir = sysfs_get_dentry(kobj->dentry,grp->name); + dir = lookup_one_len(grp->name, kobj->dentry, + strlen(grp->name)); else dir = dget(kobj->dentry); diff --git a/fs/sysfs/inode.c b/fs/sysfs/inode.c index 565cac1d420..8de13bafaa7 100644 --- a/fs/sysfs/inode.c +++ b/fs/sysfs/inode.c @@ -166,16 +166,6 @@ int sysfs_create(struct dentry * dentry, int mode, int (*init)(struct inode *)) return error; } -struct dentry * sysfs_get_dentry(struct dentry * parent, const char * name) -{ - struct qstr qstr; - - qstr.name = name; - qstr.len = strlen(name); - qstr.hash = full_name_hash(name,qstr.len); - return lookup_hash(&qstr,parent); -} - /* * Get the name for corresponding element represented by the given sysfs_dirent */ diff --git a/fs/sysfs/sysfs.h b/fs/sysfs/sysfs.h index 29da6f5f07c..3f8953e0e5d 100644 --- a/fs/sysfs/sysfs.h +++ b/fs/sysfs/sysfs.h @@ -7,7 +7,6 @@ extern int sysfs_create(struct dentry *, int mode, int (*init)(struct inode *)); extern int sysfs_make_dirent(struct sysfs_dirent *, struct dentry *, void *, umode_t, int); -extern struct dentry * sysfs_get_dentry(struct dentry *, const char *); extern int sysfs_add_file(struct dentry *, const struct attribute *, int); extern void sysfs_hash_and_remove(struct dentry * dir, const char * name); diff --git a/kernel/cpuset.c b/kernel/cpuset.c index 00e8f257551..79dd929f408 100644 --- a/kernel/cpuset.c +++ b/kernel/cpuset.c @@ -228,13 +228,7 @@ static struct dentry_operations cpuset_dops = { static struct dentry *cpuset_get_dentry(struct dentry *parent, const char *name) { - struct qstr qstr; - struct dentry *d; - - qstr.name = name; - qstr.len = strlen(name); - qstr.hash = full_name_hash(name, qstr.len); - d = lookup_hash(&qstr, parent); + struct dentry *d = lookup_one_len(name, parent, strlen(name)); if (!IS_ERR(d)) d->d_op = &cpuset_dops; return d; -- cgit v1.2.3-70-g09d2 From df164db5fd16888ddbe2a63a47b2f6dda9a428b5 Mon Sep 17 00:00:00 2001 From: Alexander Nyberg Date: Thu, 23 Jun 2005 00:09:13 -0700 Subject: [PATCH] avoid resursive oopses Prevent recursive faults in do_exit() by leaving the task alone and wait for reboot. This may allow a more graceful shutdown and possibly save the original oops. Signed-off-by: Alexander Nyberg Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/exit.c | 11 +++++++++++ 1 file changed, 11 insertions(+) (limited to 'kernel') diff --git a/kernel/exit.c b/kernel/exit.c index 2ef2ad54020..c2bdf6fb61a 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -793,6 +793,17 @@ fastcall NORET_TYPE void do_exit(long code) ptrace_notify((PTRACE_EVENT_EXIT << 8) | SIGTRAP); } + /* + * We're taking recursive faults here in do_exit. Safest is to just + * leave this task alone and wait for reboot. + */ + if (unlikely(tsk->flags & PF_EXITING)) { + printk(KERN_ALERT + "Fixing recursive fault but reboot is needed!\n"); + set_current_state(TASK_UNINTERRUPTIBLE); + schedule(); + } + tsk->flags |= PF_EXITING; /* -- cgit v1.2.3-70-g09d2 From b94cce926b2b902b79380ccba370d6f9f2980de0 Mon Sep 17 00:00:00 2001 From: Hien Nguyen Date: Thu, 23 Jun 2005 00:09:19 -0700 Subject: [PATCH] kprobes: function-return probes This patch adds function-return probes to kprobes for the i386 architecture. This enables you to establish a handler to be run when a function returns. 1. API Two new functions are added to kprobes: int register_kretprobe(struct kretprobe *rp); void unregister_kretprobe(struct kretprobe *rp); 2. Registration and unregistration 2.1 Register To register a function-return probe, the user populates the following fields in a kretprobe object and calls register_kretprobe() with the kretprobe address as an argument: kp.addr - the function's address handler - this function is run after the ret instruction executes, but before control returns to the return address in the caller. maxactive - The maximum number of instances of the probed function that can be active concurrently. For example, if the function is non- recursive and is called with a spinlock or mutex held, maxactive = 1 should be enough. If the function is non-recursive and can never relinquish the CPU (e.g., via a semaphore or preemption), NR_CPUS should be enough. maxactive is used to determine how many kretprobe_instance objects to allocate for this particular probed function. If maxactive <= 0, it is set to a default value (if CONFIG_PREEMPT maxactive=max(10, 2 * NR_CPUS) else maxactive=NR_CPUS) For example: struct kretprobe rp; rp.kp.addr = /* entrypoint address */ rp.handler = /*return probe handler */ rp.maxactive = /* e.g., 1 or NR_CPUS or 0, see the above explanation */ register_kretprobe(&rp); The following field may also be of interest: nmissed - Initialized to zero when the function-return probe is registered, and incremented every time the probed function is entered but there is no kretprobe_instance object available for establishing the function-return probe (i.e., because maxactive was set too low). 2.2 Unregister To unregiter a function-return probe, the user calls unregister_kretprobe() with the same kretprobe object as registered previously. If a probed function is running when the return probe is unregistered, the function will return as expected, but the handler won't be run. 3. Limitations 3.1 This patch supports only the i386 architecture, but patches for x86_64 and ppc64 are anticipated soon. 3.2 Return probes operates by replacing the return address in the stack (or in a known register, such as the lr register for ppc). This may cause __builtin_return_address(0), when invoked from the return-probed function, to return the address of the return-probes trampoline. 3.3 This implementation uses the "Multiprobes at an address" feature in 2.6.12-rc3-mm3. 3.4 Due to a limitation in multi-probes, you cannot currently establish a return probe and a jprobe on the same function. A patch to remove this limitation is being tested. This feature is required by SystemTap (http://sourceware.org/systemtap), and reflects ideas contributed by several SystemTap developers, including Will Cohen and Ananth Mavinakayanahalli. Signed-off-by: Hien Nguyen Signed-off-by: Prasanna S Panchamukhi Signed-off-by: Frederik Deweerdt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/kprobes.c | 102 +++++++++++++++++++++- arch/i386/kernel/process.c | 15 ++++ include/asm-i386/kprobes.h | 3 + include/linux/kprobes.h | 90 ++++++++++++++++++- kernel/kprobes.c | 213 +++++++++++++++++++++++++++++++++++++++++++-- 5 files changed, 415 insertions(+), 8 deletions(-) (limited to 'kernel') diff --git a/arch/i386/kernel/kprobes.c b/arch/i386/kernel/kprobes.c index 59ff9b45506..048f754bbe2 100644 --- a/arch/i386/kernel/kprobes.c +++ b/arch/i386/kernel/kprobes.c @@ -23,6 +23,9 @@ * Rusty Russell). * 2004-July Suparna Bhattacharya added jumper probes * interface to access function arguments. + * 2005-May Hien Nguyen , Jim Keniston + * and Prasanna S Panchamukhi + * added function-return probes. */ #include @@ -91,6 +94,53 @@ static inline void prepare_singlestep(struct kprobe *p, struct pt_regs *regs) regs->eip = (unsigned long)&p->ainsn.insn; } +struct task_struct *arch_get_kprobe_task(void *ptr) +{ + return ((struct thread_info *) (((unsigned long) ptr) & + (~(THREAD_SIZE -1))))->task; +} + +void arch_prepare_kretprobe(struct kretprobe *rp, struct pt_regs *regs) +{ + unsigned long *sara = (unsigned long *)®s->esp; + struct kretprobe_instance *ri; + static void *orig_ret_addr; + + /* + * Save the return address when the return probe hits + * the first time, and use it to populate the (krprobe + * instance)->ret_addr for subsequent return probes at + * the same addrress since stack address would have + * the kretprobe_trampoline by then. + */ + if (((void*) *sara) != kretprobe_trampoline) + orig_ret_addr = (void*) *sara; + + if ((ri = get_free_rp_inst(rp)) != NULL) { + ri->rp = rp; + ri->stack_addr = sara; + ri->ret_addr = orig_ret_addr; + add_rp_inst(ri); + /* Replace the return addr with trampoline addr */ + *sara = (unsigned long) &kretprobe_trampoline; + } else { + rp->nmissed++; + } +} + +void arch_kprobe_flush_task(struct task_struct *tk, spinlock_t *kp_lock) +{ + unsigned long flags = 0; + struct kretprobe_instance *ri; + spin_lock_irqsave(kp_lock, flags); + while ((ri = get_rp_inst_tsk(tk)) != NULL) { + *((unsigned long *)(ri->stack_addr)) = + (unsigned long) ri->ret_addr; + recycle_rp_inst(ri); + } + spin_unlock_irqrestore(kp_lock, flags); +} + /* * Interrupts are disabled on entry as trap3 is an interrupt gate and they * remain disabled thorough out this function. @@ -183,6 +233,55 @@ no_kprobe: return ret; } +/* + * For function-return probes, init_kprobes() establishes a probepoint + * here. When a retprobed function returns, this probe is hit and + * trampoline_probe_handler() runs, calling the kretprobe's handler. + */ + void kretprobe_trampoline_holder(void) + { + asm volatile ( ".global kretprobe_trampoline\n" + "kretprobe_trampoline: \n" + "nop\n"); + } + +/* + * Called when we hit the probe point at kretprobe_trampoline + */ +int trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs) +{ + struct task_struct *tsk; + struct kretprobe_instance *ri; + struct hlist_head *head; + struct hlist_node *node; + unsigned long *sara = ((unsigned long *) ®s->esp) - 1; + + tsk = arch_get_kprobe_task(sara); + head = kretprobe_inst_table_head(tsk); + + hlist_for_each_entry(ri, node, head, hlist) { + if (ri->stack_addr == sara && ri->rp) { + if (ri->rp->handler) + ri->rp->handler(ri, regs); + } + } + return 0; +} + +void trampoline_post_handler(struct kprobe *p, struct pt_regs *regs, + unsigned long flags) +{ + struct kretprobe_instance *ri; + /* RA already popped */ + unsigned long *sara = ((unsigned long *)®s->esp) - 1; + + while ((ri = get_rp_inst(sara))) { + regs->eip = (unsigned long)ri->ret_addr; + recycle_rp_inst(ri); + } + regs->eflags &= ~TF_MASK; +} + /* * Called after single-stepping. p->addr is the address of the * instruction whose first byte has been replaced by the "int 3" @@ -266,7 +365,8 @@ static inline int post_kprobe_handler(struct pt_regs *regs) if (current_kprobe->post_handler) current_kprobe->post_handler(current_kprobe, regs, 0); - resume_execution(current_kprobe, regs); + if (current_kprobe->post_handler != trampoline_post_handler) + resume_execution(current_kprobe, regs); regs->eflags |= kprobe_saved_eflags; unlock_kprobes(); diff --git a/arch/i386/kernel/process.c b/arch/i386/kernel/process.c index be3efba7caf..aea2ce1145d 100644 --- a/arch/i386/kernel/process.c +++ b/arch/i386/kernel/process.c @@ -37,6 +37,7 @@ #include #include #include +#include #include #include @@ -339,6 +340,13 @@ void exit_thread(void) struct task_struct *tsk = current; struct thread_struct *t = &tsk->thread; + /* + * Remove function-return probe instances associated with this task + * and put them back on the free list. Do not insert an exit probe for + * this function, it will be disabled by kprobe_flush_task if you do. + */ + kprobe_flush_task(tsk); + /* The process may have allocated an io port bitmap... nuke it. */ if (unlikely(NULL != t->io_bitmap_ptr)) { int cpu = get_cpu(); @@ -362,6 +370,13 @@ void flush_thread(void) { struct task_struct *tsk = current; + /* + * Remove function-return probe instances associated with this task + * and put them back on the free list. Do not insert an exit probe for + * this function, it will be disabled by kprobe_flush_task if you do. + */ + kprobe_flush_task(tsk); + memset(tsk->thread.debugreg, 0, sizeof(unsigned long)*8); memset(tsk->thread.tls_array, 0, sizeof(tsk->thread.tls_array)); /* diff --git a/include/asm-i386/kprobes.h b/include/asm-i386/kprobes.h index 4092f68d123..8b6d3a90cd7 100644 --- a/include/asm-i386/kprobes.h +++ b/include/asm-i386/kprobes.h @@ -39,6 +39,9 @@ typedef u8 kprobe_opcode_t; : (((unsigned long)current_thread_info()) + THREAD_SIZE - (ADDR))) #define JPROBE_ENTRY(pentry) (kprobe_opcode_t *)pentry +#define ARCH_SUPPORTS_KRETPROBES + +void kretprobe_trampoline(void); /* Architecture specific copy of original instruction*/ struct arch_specific_insn { diff --git a/include/linux/kprobes.h b/include/linux/kprobes.h index 99ddba5a4e0..fba39f87efe 100644 --- a/include/linux/kprobes.h +++ b/include/linux/kprobes.h @@ -25,21 +25,31 @@ * Rusty Russell). * 2004-July Suparna Bhattacharya added jumper probes * interface to access function arguments. + * 2005-May Hien Nguyen and Jim Keniston + * and Prasanna S Panchamukhi + * added function-return probes. */ #include #include #include #include +#include + #include struct kprobe; struct pt_regs; +struct kretprobe; +struct kretprobe_instance; typedef int (*kprobe_pre_handler_t) (struct kprobe *, struct pt_regs *); typedef int (*kprobe_break_handler_t) (struct kprobe *, struct pt_regs *); typedef void (*kprobe_post_handler_t) (struct kprobe *, struct pt_regs *, unsigned long flags); typedef int (*kprobe_fault_handler_t) (struct kprobe *, struct pt_regs *, int trapnr); +typedef int (*kretprobe_handler_t) (struct kretprobe_instance *, + struct pt_regs *); + struct kprobe { struct hlist_node hlist; @@ -85,6 +95,62 @@ struct jprobe { kprobe_opcode_t *entry; /* probe handling code to jump to */ }; +#ifdef ARCH_SUPPORTS_KRETPROBES +extern int trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs); +extern void trampoline_post_handler(struct kprobe *p, struct pt_regs *regs, + unsigned long flags); +extern struct task_struct *arch_get_kprobe_task(void *ptr); +extern void arch_prepare_kretprobe(struct kretprobe *rp, struct pt_regs *regs); +extern void arch_kprobe_flush_task(struct task_struct *tk, spinlock_t *kp_lock); +#else /* ARCH_SUPPORTS_KRETPROBES */ +static inline void kretprobe_trampoline(void) +{ +} +static inline int trampoline_probe_handler(struct kprobe *p, + struct pt_regs *regs) +{ + return 0; +} +static inline void trampoline_post_handler(struct kprobe *p, + struct pt_regs *regs, unsigned long flags) +{ +} +static inline void arch_prepare_kretprobe(struct kretprobe *rp, + struct pt_regs *regs) +{ +} +static inline void arch_kprobe_flush_task(struct task_struct *tk) +{ +} +#define arch_get_kprobe_task(ptr) ((struct task_struct *)NULL) +#endif /* ARCH_SUPPORTS_KRETPROBES */ +/* + * Function-return probe - + * Note: + * User needs to provide a handler function, and initialize maxactive. + * maxactive - The maximum number of instances of the probed function that + * can be active concurrently. + * nmissed - tracks the number of times the probed function's return was + * ignored, due to maxactive being too low. + * + */ +struct kretprobe { + struct kprobe kp; + kretprobe_handler_t handler; + int maxactive; + int nmissed; + struct hlist_head free_instances; + struct hlist_head used_instances; +}; + +struct kretprobe_instance { + struct hlist_node uflist; /* either on free list or used list */ + struct hlist_node hlist; + struct kretprobe *rp; + void *ret_addr; + void *stack_addr; +}; + #ifdef CONFIG_KPROBES /* Locks kprobe: irq must be disabled */ void lock_kprobes(void); @@ -104,6 +170,7 @@ extern void show_registers(struct pt_regs *regs); /* Get the kprobe at this addr (if any). Must have called lock_kprobes */ struct kprobe *get_kprobe(void *addr); +struct hlist_head * kretprobe_inst_table_head(struct task_struct *tsk); int register_kprobe(struct kprobe *p); void unregister_kprobe(struct kprobe *p); @@ -113,7 +180,16 @@ int register_jprobe(struct jprobe *p); void unregister_jprobe(struct jprobe *p); void jprobe_return(void); -#else +int register_kretprobe(struct kretprobe *rp); +void unregister_kretprobe(struct kretprobe *rp); + +struct kretprobe_instance *get_free_rp_inst(struct kretprobe *rp); +struct kretprobe_instance *get_rp_inst(void *sara); +struct kretprobe_instance *get_rp_inst_tsk(struct task_struct *tk); +void add_rp_inst(struct kretprobe_instance *ri); +void kprobe_flush_task(struct task_struct *tk); +void recycle_rp_inst(struct kretprobe_instance *ri); +#else /* CONFIG_KPROBES */ static inline int kprobe_running(void) { return 0; @@ -135,5 +211,15 @@ static inline void unregister_jprobe(struct jprobe *p) static inline void jprobe_return(void) { } -#endif +static inline int register_kretprobe(struct kretprobe *rp) +{ + return -ENOSYS; +} +static inline void unregister_kretprobe(struct kretprobe *rp) +{ +} +static inline void kprobe_flush_task(struct task_struct *tk) +{ +} +#endif /* CONFIG_KPROBES */ #endif /* _LINUX_KPROBES_H */ diff --git a/kernel/kprobes.c b/kernel/kprobes.c index 037142b72a4..692fbf75ab4 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -27,6 +27,9 @@ * interface to access function arguments. * 2004-Sep Prasanna S Panchamukhi Changed Kprobes * exceptions notifier to be first on the priority list. + * 2005-May Hien Nguyen , Jim Keniston + * and Prasanna S Panchamukhi + * added function-return probes. */ #include #include @@ -41,6 +44,7 @@ #define KPROBE_TABLE_SIZE (1 << KPROBE_HASH_BITS) static struct hlist_head kprobe_table[KPROBE_TABLE_SIZE]; +static struct hlist_head kretprobe_inst_table[KPROBE_TABLE_SIZE]; unsigned int kprobe_cpu = NR_CPUS; static DEFINE_SPINLOCK(kprobe_lock); @@ -78,7 +82,7 @@ struct kprobe *get_kprobe(void *addr) * Aggregate handlers for multiple kprobes support - these handlers * take care of invoking the individual kprobe handlers on p->list */ -int aggr_pre_handler(struct kprobe *p, struct pt_regs *regs) +static int aggr_pre_handler(struct kprobe *p, struct pt_regs *regs) { struct kprobe *kp; @@ -92,8 +96,8 @@ int aggr_pre_handler(struct kprobe *p, struct pt_regs *regs) return 0; } -void aggr_post_handler(struct kprobe *p, struct pt_regs *regs, - unsigned long flags) +static void aggr_post_handler(struct kprobe *p, struct pt_regs *regs, + unsigned long flags) { struct kprobe *kp; @@ -107,7 +111,8 @@ void aggr_post_handler(struct kprobe *p, struct pt_regs *regs, return; } -int aggr_fault_handler(struct kprobe *p, struct pt_regs *regs, int trapnr) +static int aggr_fault_handler(struct kprobe *p, struct pt_regs *regs, + int trapnr) { /* * if we faulted "during" the execution of a user specified @@ -120,6 +125,135 @@ int aggr_fault_handler(struct kprobe *p, struct pt_regs *regs, int trapnr) return 0; } +struct kprobe trampoline_p = { + .addr = (kprobe_opcode_t *) &kretprobe_trampoline, + .pre_handler = trampoline_probe_handler, + .post_handler = trampoline_post_handler +}; + +struct kretprobe_instance *get_free_rp_inst(struct kretprobe *rp) +{ + struct hlist_node *node; + struct kretprobe_instance *ri; + hlist_for_each_entry(ri, node, &rp->free_instances, uflist) + return ri; + return NULL; +} + +static struct kretprobe_instance *get_used_rp_inst(struct kretprobe *rp) +{ + struct hlist_node *node; + struct kretprobe_instance *ri; + hlist_for_each_entry(ri, node, &rp->used_instances, uflist) + return ri; + return NULL; +} + +struct kretprobe_instance *get_rp_inst(void *sara) +{ + struct hlist_head *head; + struct hlist_node *node; + struct task_struct *tsk; + struct kretprobe_instance *ri; + + tsk = arch_get_kprobe_task(sara); + head = &kretprobe_inst_table[hash_ptr(tsk, KPROBE_HASH_BITS)]; + hlist_for_each_entry(ri, node, head, hlist) { + if (ri->stack_addr == sara) + return ri; + } + return NULL; +} + +void add_rp_inst(struct kretprobe_instance *ri) +{ + struct task_struct *tsk; + /* + * Remove rp inst off the free list - + * Add it back when probed function returns + */ + hlist_del(&ri->uflist); + tsk = arch_get_kprobe_task(ri->stack_addr); + /* Add rp inst onto table */ + INIT_HLIST_NODE(&ri->hlist); + hlist_add_head(&ri->hlist, + &kretprobe_inst_table[hash_ptr(tsk, KPROBE_HASH_BITS)]); + + /* Also add this rp inst to the used list. */ + INIT_HLIST_NODE(&ri->uflist); + hlist_add_head(&ri->uflist, &ri->rp->used_instances); +} + +void recycle_rp_inst(struct kretprobe_instance *ri) +{ + /* remove rp inst off the rprobe_inst_table */ + hlist_del(&ri->hlist); + if (ri->rp) { + /* remove rp inst off the used list */ + hlist_del(&ri->uflist); + /* put rp inst back onto the free list */ + INIT_HLIST_NODE(&ri->uflist); + hlist_add_head(&ri->uflist, &ri->rp->free_instances); + } else + /* Unregistering */ + kfree(ri); +} + +struct hlist_head * kretprobe_inst_table_head(struct task_struct *tsk) +{ + return &kretprobe_inst_table[hash_ptr(tsk, KPROBE_HASH_BITS)]; +} + +struct kretprobe_instance *get_rp_inst_tsk(struct task_struct *tk) +{ + struct task_struct *tsk; + struct hlist_head *head; + struct hlist_node *node; + struct kretprobe_instance *ri; + + head = &kretprobe_inst_table[hash_ptr(tk, KPROBE_HASH_BITS)]; + + hlist_for_each_entry(ri, node, head, hlist) { + tsk = arch_get_kprobe_task(ri->stack_addr); + if (tsk == tk) + return ri; + } + return NULL; +} + +/* + * This function is called from do_exit or do_execv when task tk's stack is + * about to be recycled. Recycle any function-return probe instances + * associated with this task. These represent probed functions that have + * been called but may never return. + */ +void kprobe_flush_task(struct task_struct *tk) +{ + arch_kprobe_flush_task(tk, &kprobe_lock); +} + +/* + * This kprobe pre_handler is registered with every kretprobe. When probe + * hits it will set up the return probe. + */ +static int pre_handler_kretprobe(struct kprobe *p, struct pt_regs *regs) +{ + struct kretprobe *rp = container_of(p, struct kretprobe, kp); + + /*TODO: consider to only swap the RA after the last pre_handler fired */ + arch_prepare_kretprobe(rp, regs); + return 0; +} + +static inline void free_rp_inst(struct kretprobe *rp) +{ + struct kretprobe_instance *ri; + while ((ri = get_free_rp_inst(rp)) != NULL) { + hlist_del(&ri->uflist); + kfree(ri); + } +} + /* * Fill in the required fields of the "manager kprobe". Replace the * earlier kprobe in the hlist with the manager kprobe @@ -257,16 +391,82 @@ void unregister_jprobe(struct jprobe *jp) unregister_kprobe(&jp->kp); } +#ifdef ARCH_SUPPORTS_KRETPROBES + +int register_kretprobe(struct kretprobe *rp) +{ + int ret = 0; + struct kretprobe_instance *inst; + int i; + + rp->kp.pre_handler = pre_handler_kretprobe; + + /* Pre-allocate memory for max kretprobe instances */ + if (rp->maxactive <= 0) { +#ifdef CONFIG_PREEMPT + rp->maxactive = max(10, 2 * NR_CPUS); +#else + rp->maxactive = NR_CPUS; +#endif + } + INIT_HLIST_HEAD(&rp->used_instances); + INIT_HLIST_HEAD(&rp->free_instances); + for (i = 0; i < rp->maxactive; i++) { + inst = kmalloc(sizeof(struct kretprobe_instance), GFP_KERNEL); + if (inst == NULL) { + free_rp_inst(rp); + return -ENOMEM; + } + INIT_HLIST_NODE(&inst->uflist); + hlist_add_head(&inst->uflist, &rp->free_instances); + } + + rp->nmissed = 0; + /* Establish function entry probe point */ + if ((ret = register_kprobe(&rp->kp)) != 0) + free_rp_inst(rp); + return ret; +} + +#else /* ARCH_SUPPORTS_KRETPROBES */ + +int register_kretprobe(struct kretprobe *rp) +{ + return -ENOSYS; +} + +#endif /* ARCH_SUPPORTS_KRETPROBES */ + +void unregister_kretprobe(struct kretprobe *rp) +{ + unsigned long flags; + struct kretprobe_instance *ri; + + unregister_kprobe(&rp->kp); + /* No race here */ + spin_lock_irqsave(&kprobe_lock, flags); + free_rp_inst(rp); + while ((ri = get_used_rp_inst(rp)) != NULL) { + ri->rp = NULL; + hlist_del(&ri->uflist); + } + spin_unlock_irqrestore(&kprobe_lock, flags); +} + static int __init init_kprobes(void) { int i, err = 0; /* FIXME allocate the probe table, currently defined statically */ /* initialize all list heads */ - for (i = 0; i < KPROBE_TABLE_SIZE; i++) + for (i = 0; i < KPROBE_TABLE_SIZE; i++) { INIT_HLIST_HEAD(&kprobe_table[i]); + INIT_HLIST_HEAD(&kretprobe_inst_table[i]); + } err = register_die_notifier(&kprobe_exceptions_nb); + /* Register the trampoline probe for return probe */ + register_kprobe(&trampoline_p); return err; } @@ -277,3 +477,6 @@ EXPORT_SYMBOL_GPL(unregister_kprobe); EXPORT_SYMBOL_GPL(register_jprobe); EXPORT_SYMBOL_GPL(unregister_jprobe); EXPORT_SYMBOL_GPL(jprobe_return); +EXPORT_SYMBOL_GPL(register_kretprobe); +EXPORT_SYMBOL_GPL(unregister_kretprobe); + -- cgit v1.2.3-70-g09d2 From 7e1048b11c5afe79aac46a42e3ccec86b8365c6d Mon Sep 17 00:00:00 2001 From: Rusty Lynch Date: Thu, 23 Jun 2005 00:09:25 -0700 Subject: [PATCH] Move kprobe [dis]arming into arch specific code The architecture independent code of the current kprobes implementation is arming and disarming kprobes at registration time. The problem is that the code is assuming that arming and disarming is a just done by a simple write of some magic value to an address. This is problematic for ia64 where our instructions look more like structures, and we can not insert break points by just doing something like: *p->addr = BREAKPOINT_INSTRUCTION; The following patch to 2.6.12-rc4-mm2 adds two new architecture dependent functions: * void arch_arm_kprobe(struct kprobe *p) * void arch_disarm_kprobe(struct kprobe *p) and then adds the new functions for each of the architectures that already implement kprobes (spar64/ppc64/i386/x86_64). I thought arch_[dis]arm_kprobe was the most descriptive of what was really happening, but each of the architectures already had a disarm_kprobe() function that was really a "disarm and do some other clean-up items as needed when you stumble across a recursive kprobe." So... I took the liberty of changing the code that was calling disarm_kprobe() to call arch_disarm_kprobe(), and then do the cleanup in the block of code dealing with the recursive kprobe case. So far this patch as been tested on i386, x86_64, and ppc64, but still needs to be tested in sparc64. Signed-off-by: Rusty Lynch Signed-off-by: Anil S Keshavamurthy Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/kprobes.c | 19 +++++++++++++++---- arch/ppc64/kernel/kprobes.c | 19 +++++++++++++++---- arch/sparc64/kernel/kprobes.c | 31 ++++++++++++++++++------------- arch/x86_64/kernel/kprobes.c | 26 ++++++++++++++++++-------- include/linux/kprobes.h | 2 ++ kernel/kprobes.c | 12 ++++-------- 6 files changed, 72 insertions(+), 37 deletions(-) (limited to 'kernel') diff --git a/arch/i386/kernel/kprobes.c b/arch/i386/kernel/kprobes.c index 048f754bbe2..2314d8d306f 100644 --- a/arch/i386/kernel/kprobes.c +++ b/arch/i386/kernel/kprobes.c @@ -33,6 +33,7 @@ #include #include #include +#include #include #include @@ -71,16 +72,25 @@ int arch_prepare_kprobe(struct kprobe *p) void arch_copy_kprobe(struct kprobe *p) { memcpy(p->ainsn.insn, p->addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t)); + p->opcode = *p->addr; } -void arch_remove_kprobe(struct kprobe *p) +void arch_arm_kprobe(struct kprobe *p) { + *p->addr = BREAKPOINT_INSTRUCTION; + flush_icache_range((unsigned long) p->addr, + (unsigned long) p->addr + sizeof(kprobe_opcode_t)); } -static inline void disarm_kprobe(struct kprobe *p, struct pt_regs *regs) +void arch_disarm_kprobe(struct kprobe *p) { *p->addr = p->opcode; - regs->eip = (unsigned long)p->addr; + flush_icache_range((unsigned long) p->addr, + (unsigned long) p->addr + sizeof(kprobe_opcode_t)); +} + +void arch_remove_kprobe(struct kprobe *p) +{ } static inline void prepare_singlestep(struct kprobe *p, struct pt_regs *regs) @@ -177,7 +187,8 @@ static int kprobe_handler(struct pt_regs *regs) unlock_kprobes(); goto no_kprobe; } - disarm_kprobe(p, regs); + arch_disarm_kprobe(p); + regs->eip = (unsigned long)p->addr; ret = 1; } else { p = current_kprobe; diff --git a/arch/ppc64/kernel/kprobes.c b/arch/ppc64/kernel/kprobes.c index e950a2058a1..8c0920a6d03 100644 --- a/arch/ppc64/kernel/kprobes.c +++ b/arch/ppc64/kernel/kprobes.c @@ -32,6 +32,7 @@ #include #include #include +#include #include #include @@ -61,16 +62,25 @@ int arch_prepare_kprobe(struct kprobe *p) void arch_copy_kprobe(struct kprobe *p) { memcpy(p->ainsn.insn, p->addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t)); + p->opcode = *p->addr; } -void arch_remove_kprobe(struct kprobe *p) +void arch_arm_kprobe(struct kprobe *p) { + *p->addr = BREAKPOINT_INSTRUCTION; + flush_icache_range((unsigned long) p->addr, + (unsigned long) p->addr + sizeof(kprobe_opcode_t)); } -static inline void disarm_kprobe(struct kprobe *p, struct pt_regs *regs) +void arch_disarm_kprobe(struct kprobe *p) { *p->addr = p->opcode; - regs->nip = (unsigned long)p->addr; + flush_icache_range((unsigned long) p->addr, + (unsigned long) p->addr + sizeof(kprobe_opcode_t)); +} + +void arch_remove_kprobe(struct kprobe *p) +{ } static inline void prepare_singlestep(struct kprobe *p, struct pt_regs *regs) @@ -101,7 +111,8 @@ static inline int kprobe_handler(struct pt_regs *regs) unlock_kprobes(); goto no_kprobe; } - disarm_kprobe(p, regs); + arch_disarm_kprobe(p); + regs->nip = (unsigned long)p->addr; ret = 1; } else { p = current_kprobe; diff --git a/arch/sparc64/kernel/kprobes.c b/arch/sparc64/kernel/kprobes.c index 7066d7ba667..d67195ba3fa 100644 --- a/arch/sparc64/kernel/kprobes.c +++ b/arch/sparc64/kernel/kprobes.c @@ -6,7 +6,6 @@ #include #include #include - #include #include @@ -47,6 +46,19 @@ void arch_copy_kprobe(struct kprobe *p) { p->ainsn.insn[0] = *p->addr; p->ainsn.insn[1] = BREAKPOINT_INSTRUCTION_2; + p->opcode = *p->addr; +} + +void arch_arm_kprobe(struct kprobe *p) +{ + *p->addr = BREAKPOINT_INSTRUCTION; + flushi(p->addr); +} + +void arch_disarm_kprobe(struct kprobe *p) +{ + *p->addr = p->opcode; + flushi(p->addr); } void arch_remove_kprobe(struct kprobe *p) @@ -78,17 +90,6 @@ static inline void prepare_singlestep(struct kprobe *p, struct pt_regs *regs) } } -static inline void disarm_kprobe(struct kprobe *p, struct pt_regs *regs) -{ - *p->addr = p->opcode; - flushi(p->addr); - - regs->tpc = (unsigned long) p->addr; - regs->tnpc = current_kprobe_orig_tnpc; - regs->tstate = ((regs->tstate & ~TSTATE_PIL) | - current_kprobe_orig_tstate_pil); -} - static int kprobe_handler(struct pt_regs *regs) { struct kprobe *p; @@ -109,7 +110,11 @@ static int kprobe_handler(struct pt_regs *regs) unlock_kprobes(); goto no_kprobe; } - disarm_kprobe(p, regs); + arch_disarm_kprobe(p); + regs->tpc = (unsigned long) p->addr; + regs->tnpc = current_kprobe_orig_tnpc; + regs->tstate = ((regs->tstate & ~TSTATE_PIL) | + current_kprobe_orig_tstate_pil); ret = 1; } else { p = current_kprobe; diff --git a/arch/x86_64/kernel/kprobes.c b/arch/x86_64/kernel/kprobes.c index 203672ca740..324bf57925a 100644 --- a/arch/x86_64/kernel/kprobes.c +++ b/arch/x86_64/kernel/kprobes.c @@ -39,7 +39,7 @@ #include #include #include - +#include #include #include @@ -216,19 +216,28 @@ void arch_copy_kprobe(struct kprobe *p) BUG_ON((s64) (s32) disp != disp); /* Sanity check. */ *ripdisp = disp; } + p->opcode = *p->addr; } -void arch_remove_kprobe(struct kprobe *p) +void arch_arm_kprobe(struct kprobe *p) { - up(&kprobe_mutex); - free_insn_slot(p->ainsn.insn); - down(&kprobe_mutex); + *p->addr = BREAKPOINT_INSTRUCTION; + flush_icache_range((unsigned long) p->addr, + (unsigned long) p->addr + sizeof(kprobe_opcode_t)); } -static inline void disarm_kprobe(struct kprobe *p, struct pt_regs *regs) +void arch_disarm_kprobe(struct kprobe *p) { *p->addr = p->opcode; - regs->rip = (unsigned long)p->addr; + flush_icache_range((unsigned long) p->addr, + (unsigned long) p->addr + sizeof(kprobe_opcode_t)); +} + +void arch_remove_kprobe(struct kprobe *p) +{ + up(&kprobe_mutex); + free_insn_slot(p->ainsn.insn); + down(&kprobe_mutex); } static void prepare_singlestep(struct kprobe *p, struct pt_regs *regs) @@ -311,7 +320,8 @@ int kprobe_handler(struct pt_regs *regs) unlock_kprobes(); goto no_kprobe; } - disarm_kprobe(p, regs); + arch_disarm_kprobe(p); + regs->rip = (unsigned long)p->addr; ret = 1; } else { p = current_kprobe; diff --git a/include/linux/kprobes.h b/include/linux/kprobes.h index fba39f87efe..0f90466fb8b 100644 --- a/include/linux/kprobes.h +++ b/include/linux/kprobes.h @@ -165,6 +165,8 @@ static inline int kprobe_running(void) extern int arch_prepare_kprobe(struct kprobe *p); extern void arch_copy_kprobe(struct kprobe *p); +extern void arch_arm_kprobe(struct kprobe *p); +extern void arch_disarm_kprobe(struct kprobe *p); extern void arch_remove_kprobe(struct kprobe *p); extern void show_registers(struct pt_regs *regs); diff --git a/kernel/kprobes.c b/kernel/kprobes.c index 692fbf75ab4..e8e0ae8a6e1 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -261,7 +261,7 @@ static inline void free_rp_inst(struct kretprobe *rp) static inline void add_aggr_kprobe(struct kprobe *ap, struct kprobe *p) { ap->addr = p->addr; - ap->opcode = p->opcode; + memcpy(&ap->opcode, &p->opcode, sizeof(kprobe_opcode_t)); memcpy(&ap->ainsn, &p->ainsn, sizeof(struct arch_specific_insn)); ap->pre_handler = aggr_pre_handler; @@ -304,10 +304,8 @@ static int register_aggr_kprobe(struct kprobe *old_p, struct kprobe *p) /* kprobe removal house-keeping routines */ static inline void cleanup_kprobe(struct kprobe *p, unsigned long flags) { - *p->addr = p->opcode; + arch_disarm_kprobe(p); hlist_del(&p->hlist); - flush_icache_range((unsigned long) p->addr, - (unsigned long) p->addr + sizeof(kprobe_opcode_t)); spin_unlock_irqrestore(&kprobe_lock, flags); arch_remove_kprobe(p); } @@ -344,10 +342,8 @@ int register_kprobe(struct kprobe *p) hlist_add_head(&p->hlist, &kprobe_table[hash_ptr(p->addr, KPROBE_HASH_BITS)]); - p->opcode = *p->addr; - *p->addr = BREAKPOINT_INSTRUCTION; - flush_icache_range((unsigned long) p->addr, - (unsigned long) p->addr + sizeof(kprobe_opcode_t)); + arch_arm_kprobe(p); + out: spin_unlock_irqrestore(&kprobe_lock, flags); rm_kprobe: -- cgit v1.2.3-70-g09d2 From 0aa55e4d7db822059fe8132fe9f2b7773c48216c Mon Sep 17 00:00:00 2001 From: Hien Nguyen Date: Thu, 23 Jun 2005 00:09:26 -0700 Subject: [PATCH] kprobes: moves lock-unlock to non-arch kprobe_flush_task This patch moves the lock/unlock of the arch specific kprobe_flush_task() to the non-arch specific kprobe_flusk_task(). Signed-off-by: Hien Nguyen Acked-by: Prasanna S Panchamukhi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/kprobes.c | 5 +---- include/linux/kprobes.h | 3 +-- kernel/kprobes.c | 5 ++++- 3 files changed, 6 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/arch/i386/kernel/kprobes.c b/arch/i386/kernel/kprobes.c index 2314d8d306f..b8e2bae0ab4 100644 --- a/arch/i386/kernel/kprobes.c +++ b/arch/i386/kernel/kprobes.c @@ -138,17 +138,14 @@ void arch_prepare_kretprobe(struct kretprobe *rp, struct pt_regs *regs) } } -void arch_kprobe_flush_task(struct task_struct *tk, spinlock_t *kp_lock) +void arch_kprobe_flush_task(struct task_struct *tk) { - unsigned long flags = 0; struct kretprobe_instance *ri; - spin_lock_irqsave(kp_lock, flags); while ((ri = get_rp_inst_tsk(tk)) != NULL) { *((unsigned long *)(ri->stack_addr)) = (unsigned long) ri->ret_addr; recycle_rp_inst(ri); } - spin_unlock_irqrestore(kp_lock, flags); } /* diff --git a/include/linux/kprobes.h b/include/linux/kprobes.h index 0f90466fb8b..461391decc4 100644 --- a/include/linux/kprobes.h +++ b/include/linux/kprobes.h @@ -33,7 +33,6 @@ #include #include #include -#include #include @@ -101,7 +100,7 @@ extern void trampoline_post_handler(struct kprobe *p, struct pt_regs *regs, unsigned long flags); extern struct task_struct *arch_get_kprobe_task(void *ptr); extern void arch_prepare_kretprobe(struct kretprobe *rp, struct pt_regs *regs); -extern void arch_kprobe_flush_task(struct task_struct *tk, spinlock_t *kp_lock); +extern void arch_kprobe_flush_task(struct task_struct *tk); #else /* ARCH_SUPPORTS_KRETPROBES */ static inline void kretprobe_trampoline(void) { diff --git a/kernel/kprobes.c b/kernel/kprobes.c index e8e0ae8a6e1..dd42e717dd3 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -229,7 +229,10 @@ struct kretprobe_instance *get_rp_inst_tsk(struct task_struct *tk) */ void kprobe_flush_task(struct task_struct *tk) { - arch_kprobe_flush_task(tk, &kprobe_lock); + unsigned long flags = 0; + spin_lock_irqsave(&kprobe_lock, flags); + arch_kprobe_flush_task(tk); + spin_unlock_irqrestore(&kprobe_lock, flags); } /* -- cgit v1.2.3-70-g09d2 From ea32c65cc2d2294c04e9f81d0578a6f51febfdbf Mon Sep 17 00:00:00 2001 From: Prasanna S Panchamukhi Date: Thu, 23 Jun 2005 00:09:36 -0700 Subject: [PATCH] kprobes: Temporary disarming of reentrant probe In situations where a kprobes handler calls a routine which has a probe on it, then kprobes_handler() disarms the new probe forever. This patch removes the above limitation by temporarily disarming the new probe. When the another probe hits while handling the old probe, the kprobes_handler() saves previous kprobes state and handles the new probe without calling the new kprobes registered handlers. kprobe_post_handler() restores back the previous kprobes state and the normal execution continues. However on x86_64 architecture, re-rentrancy is provided only through pre_handler(). If a routine having probe is referenced through post_handler(), then the probes on that routine are disarmed forever, since the exception stack is gets changed after the processor single steps the instruction of the new probe. This patch includes generic changes to support temporary disarming on reentrancy of probes. Signed-of-by: Prasanna S Panchamukhi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/kprobes.h | 9 +++++++++ kernel/kprobes.c | 1 + 2 files changed, 10 insertions(+) (limited to 'kernel') diff --git a/include/linux/kprobes.h b/include/linux/kprobes.h index 461391decc4..5e1a7b0d7b3 100644 --- a/include/linux/kprobes.h +++ b/include/linux/kprobes.h @@ -36,6 +36,12 @@ #include +/* kprobe_status settings */ +#define KPROBE_HIT_ACTIVE 0x00000001 +#define KPROBE_HIT_SS 0x00000002 +#define KPROBE_REENTER 0x00000004 +#define KPROBE_HIT_SSDONE 0x00000008 + struct kprobe; struct pt_regs; struct kretprobe; @@ -55,6 +61,9 @@ struct kprobe { /* list of kprobes for multi-handler support */ struct list_head list; + /*count the number of times this probe was temporarily disarmed */ + unsigned long nmissed; + /* location of the probe point */ kprobe_opcode_t *addr; diff --git a/kernel/kprobes.c b/kernel/kprobes.c index dd42e717dd3..456ecedff2d 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -335,6 +335,7 @@ int register_kprobe(struct kprobe *p) } spin_lock_irqsave(&kprobe_lock, flags); old_p = get_kprobe(p->addr); + p->nmissed = 0; if (old_p) { ret = register_aggr_kprobe(old_p, p); goto out; -- cgit v1.2.3-70-g09d2 From 8b0914ea7475615c7c8965c1ac8fe4069270f25c Mon Sep 17 00:00:00 2001 From: Prasanna S Panchamukhi Date: Thu, 23 Jun 2005 00:09:41 -0700 Subject: [PATCH] jprobes: allow a jprobe to coexist with muliple kprobes Presently either multiple kprobes or only one jprobe could be inserted. This patch removes the above limitation and allows one jprobe and multiple kprobes to coexist at the same address. However multiple jprobes cannot coexist with multiple kprobes. Currently I am working on the prototype to allow multiple jprobes coexist with multiple kprobes. Signed-off-by: Ananth N Mavinakayanhalli Signed-off-by: Prasanna S Panchamukhi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/kprobes.c | 61 ++++++++++++++++++++++++++++++++++++++++++++++---------- 1 file changed, 51 insertions(+), 10 deletions(-) (limited to 'kernel') diff --git a/kernel/kprobes.c b/kernel/kprobes.c index 456ecedff2d..334f37472c5 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -89,9 +89,10 @@ static int aggr_pre_handler(struct kprobe *p, struct pt_regs *regs) list_for_each_entry(kp, &p->list, list) { if (kp->pre_handler) { curr_kprobe = kp; - kp->pre_handler(kp, regs); - curr_kprobe = NULL; + if (kp->pre_handler(kp, regs)) + return 1; } + curr_kprobe = NULL; } return 0; } @@ -125,6 +126,19 @@ static int aggr_fault_handler(struct kprobe *p, struct pt_regs *regs, return 0; } +static int aggr_break_handler(struct kprobe *p, struct pt_regs *regs) +{ + struct kprobe *kp = curr_kprobe; + if (curr_kprobe && kp->break_handler) { + if (kp->break_handler(kp, regs)) { + curr_kprobe = NULL; + return 1; + } + } + curr_kprobe = NULL; + return 0; +} + struct kprobe trampoline_p = { .addr = (kprobe_opcode_t *) &kretprobe_trampoline, .pre_handler = trampoline_probe_handler, @@ -257,19 +271,46 @@ static inline void free_rp_inst(struct kretprobe *rp) } } +/* + * Keep all fields in the kprobe consistent + */ +static inline void copy_kprobe(struct kprobe *old_p, struct kprobe *p) +{ + memcpy(&p->opcode, &old_p->opcode, sizeof(kprobe_opcode_t)); + memcpy(&p->ainsn, &old_p->ainsn, sizeof(struct arch_specific_insn)); +} + +/* +* Add the new probe to old_p->list. Fail if this is the +* second jprobe at the address - two jprobes can't coexist +*/ +static int add_new_kprobe(struct kprobe *old_p, struct kprobe *p) +{ + struct kprobe *kp; + + if (p->break_handler) { + list_for_each_entry(kp, &old_p->list, list) { + if (kp->break_handler) + return -EEXIST; + } + list_add_tail(&p->list, &old_p->list); + } else + list_add(&p->list, &old_p->list); + return 0; +} + /* * Fill in the required fields of the "manager kprobe". Replace the * earlier kprobe in the hlist with the manager kprobe */ static inline void add_aggr_kprobe(struct kprobe *ap, struct kprobe *p) { + copy_kprobe(p, ap); ap->addr = p->addr; - memcpy(&ap->opcode, &p->opcode, sizeof(kprobe_opcode_t)); - memcpy(&ap->ainsn, &p->ainsn, sizeof(struct arch_specific_insn)); - ap->pre_handler = aggr_pre_handler; ap->post_handler = aggr_post_handler; ap->fault_handler = aggr_fault_handler; + ap->break_handler = aggr_break_handler; INIT_LIST_HEAD(&ap->list); list_add(&p->list, &ap->list); @@ -290,16 +331,16 @@ static int register_aggr_kprobe(struct kprobe *old_p, struct kprobe *p) int ret = 0; struct kprobe *ap; - if (old_p->break_handler || p->break_handler) { - ret = -EEXIST; /* kprobe and jprobe can't (yet) coexist */ - } else if (old_p->pre_handler == aggr_pre_handler) { - list_add(&p->list, &old_p->list); + if (old_p->pre_handler == aggr_pre_handler) { + copy_kprobe(old_p, p); + ret = add_new_kprobe(old_p, p); } else { ap = kcalloc(1, sizeof(struct kprobe), GFP_ATOMIC); if (!ap) return -ENOMEM; add_aggr_kprobe(ap, old_p); - list_add(&p->list, &ap->list); + copy_kprobe(ap, p); + ret = add_new_kprobe(ap, p); } return ret; } -- cgit v1.2.3-70-g09d2 From d6e711448137ca3301512cec41a2c2ce852b3d0a Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Thu, 23 Jun 2005 00:09:43 -0700 Subject: [PATCH] setuid core dump Add a new `suid_dumpable' sysctl: This value can be used to query and set the core dump mode for setuid or otherwise protected/tainted binaries. The modes are 0 - (default) - traditional behaviour. Any process which has changed privilege levels or is execute only will not be dumped 1 - (debug) - all processes dump core when possible. The core dump is owned by the current user and no security is applied. This is intended for system debugging situations only. Ptrace is unchecked. 2 - (suidsafe) - any binary which normally would not be dumped is dumped readable by root only. This allows the end user to remove such a dump but not access it directly. For security reasons core dumps in this mode will not overwrite one another or other files. This mode is appropriate when adminstrators are attempting to debug problems in a normal environment. (akpm: > > +EXPORT_SYMBOL(suid_dumpable); > > EXPORT_SYMBOL_GPL? No problem to me. > > if (current->euid == current->uid && current->egid == current->gid) > > current->mm->dumpable = 1; > > Should this be SUID_DUMP_USER? Actually the feedback I had from last time was that the SUID_ defines should go because its clearer to follow the numbers. They can go everywhere (and there are lots of places where dumpable is tested/used as a bool in untouched code) > Maybe this should be renamed to `dump_policy' or something. Doing that > would help us catch any code which isn't using the #defines, too. Fair comment. The patch was designed to be easy to maintain for Red Hat rather than for merging. Changing that field would create a gigantic diff because it is used all over the place. ) Signed-off-by: Alan Cox Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/sysctl/kernel.txt | 20 ++++++++++++++++++++ fs/exec.c | 23 +++++++++++++++++++++-- fs/proc/base.c | 6 ++++-- include/linux/binfmts.h | 5 +++++ include/linux/sched.h | 2 +- include/linux/sysctl.h | 1 + kernel/sys.c | 22 +++++++++++----------- kernel/sysctl.c | 9 +++++++++ security/commoncap.c | 2 +- security/dummy.c | 2 +- 10 files changed, 74 insertions(+), 18 deletions(-) (limited to 'kernel') diff --git a/Documentation/sysctl/kernel.txt b/Documentation/sysctl/kernel.txt index 35159176997..9f11d36a8c1 100644 --- a/Documentation/sysctl/kernel.txt +++ b/Documentation/sysctl/kernel.txt @@ -49,6 +49,7 @@ show up in /proc/sys/kernel: - shmmax [ sysv ipc ] - shmmni - stop-a [ SPARC only ] +- suid_dumpable - sysrq ==> Documentation/sysrq.txt - tainted - threads-max @@ -300,6 +301,25 @@ kernel. This value defaults to SHMMAX. ============================================================== +suid_dumpable: + +This value can be used to query and set the core dump mode for setuid +or otherwise protected/tainted binaries. The modes are + +0 - (default) - traditional behaviour. Any process which has changed + privilege levels or is execute only will not be dumped +1 - (debug) - all processes dump core when possible. The core dump is + owned by the current user and no security is applied. This is + intended for system debugging situations only. Ptrace is unchecked. +2 - (suidsafe) - any binary which normally would not be dumped is dumped + readable by root only. This allows the end user to remove + such a dump but not access it directly. For security reasons + core dumps in this mode will not overwrite one another or + other files. This mode is appropriate when adminstrators are + attempting to debug problems in a normal environment. + +============================================================== + tainted: Non-zero if the kernel has been tainted. Numeric values, which diff --git a/fs/exec.c b/fs/exec.c index 3a4b35a14c0..48871917d36 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -58,6 +58,9 @@ int core_uses_pid; char core_pattern[65] = "core"; +int suid_dumpable = 0; + +EXPORT_SYMBOL(suid_dumpable); /* The maximal length of core_pattern is also specified in sysctl.c */ static struct linux_binfmt *formats; @@ -864,6 +867,9 @@ int flush_old_exec(struct linux_binprm * bprm) if (current->euid == current->uid && current->egid == current->gid) current->mm->dumpable = 1; + else + current->mm->dumpable = suid_dumpable; + name = bprm->filename; /* Copies the binary name from after last slash */ @@ -884,7 +890,7 @@ int flush_old_exec(struct linux_binprm * bprm) permission(bprm->file->f_dentry->d_inode,MAY_READ, NULL) || (bprm->interp_flags & BINPRM_FLAGS_ENFORCE_NONDUMP)) { suid_keys(current); - current->mm->dumpable = 0; + current->mm->dumpable = suid_dumpable; } /* An exec changes our domain. We are no longer part of the thread @@ -1432,6 +1438,8 @@ int do_coredump(long signr, int exit_code, struct pt_regs * regs) struct inode * inode; struct file * file; int retval = 0; + int fsuid = current->fsuid; + int flag = 0; binfmt = current->binfmt; if (!binfmt || !binfmt->core_dump) @@ -1441,6 +1449,16 @@ int do_coredump(long signr, int exit_code, struct pt_regs * regs) up_write(&mm->mmap_sem); goto fail; } + + /* + * We cannot trust fsuid as being the "true" uid of the + * process nor do we know its entire history. We only know it + * was tainted so we dump it as root in mode 2. + */ + if (mm->dumpable == 2) { /* Setuid core dump mode */ + flag = O_EXCL; /* Stop rewrite attacks */ + current->fsuid = 0; /* Dump root private */ + } mm->dumpable = 0; init_completion(&mm->core_done); spin_lock_irq(¤t->sighand->siglock); @@ -1466,7 +1484,7 @@ int do_coredump(long signr, int exit_code, struct pt_regs * regs) lock_kernel(); format_corename(corename, core_pattern, signr); unlock_kernel(); - file = filp_open(corename, O_CREAT | 2 | O_NOFOLLOW | O_LARGEFILE, 0600); + file = filp_open(corename, O_CREAT | 2 | O_NOFOLLOW | O_LARGEFILE | flag, 0600); if (IS_ERR(file)) goto fail_unlock; inode = file->f_dentry->d_inode; @@ -1491,6 +1509,7 @@ int do_coredump(long signr, int exit_code, struct pt_regs * regs) close_fail: filp_close(file, NULL); fail_unlock: + current->fsuid = fsuid; complete_all(&mm->core_done); fail: return retval; diff --git a/fs/proc/base.c b/fs/proc/base.c index e31903aadd9..ace151fa487 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -314,7 +314,7 @@ static int may_ptrace_attach(struct task_struct *task) (current->gid != task->gid)) && !capable(CAP_SYS_PTRACE)) goto out; rmb(); - if (!task->mm->dumpable && !capable(CAP_SYS_PTRACE)) + if (task->mm->dumpable != 1 && !capable(CAP_SYS_PTRACE)) goto out; if (security_ptrace(current, task)) goto out; @@ -1113,7 +1113,9 @@ static int task_dumpable(struct task_struct *task) if (mm) dumpable = mm->dumpable; task_unlock(task); - return dumpable; + if(dumpable == 1) + return 1; + return 0; } diff --git a/include/linux/binfmts.h b/include/linux/binfmts.h index 7e736e201c4..c1e82c51444 100644 --- a/include/linux/binfmts.h +++ b/include/linux/binfmts.h @@ -69,6 +69,11 @@ extern void remove_arg_zero(struct linux_binprm *); extern int search_binary_handler(struct linux_binprm *,struct pt_regs *); extern int flush_old_exec(struct linux_binprm * bprm); +extern int suid_dumpable; +#define SUID_DUMP_DISABLE 0 /* No setuid dumping */ +#define SUID_DUMP_USER 1 /* Dump as user of process */ +#define SUID_DUMP_ROOT 2 /* Dump as root */ + /* Stack area protections */ #define EXSTACK_DEFAULT 0 /* Whatever the arch defaults to */ #define EXSTACK_DISABLE_X 1 /* Disable executable stacks */ diff --git a/include/linux/sched.h b/include/linux/sched.h index b58afd97a18..901742f9238 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -246,7 +246,7 @@ struct mm_struct { unsigned long saved_auxv[42]; /* for /proc/PID/auxv */ - unsigned dumpable:1; + unsigned dumpable:2; cpumask_t cpu_vm_mask; /* Architecture-specific MM context */ diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h index a17745c80a9..614e939c78a 100644 --- a/include/linux/sysctl.h +++ b/include/linux/sysctl.h @@ -136,6 +136,7 @@ enum KERN_UNKNOWN_NMI_PANIC=66, /* int: unknown nmi panic flag */ KERN_BOOTLOADER_TYPE=67, /* int: boot loader type */ KERN_RANDOMIZE=68, /* int: randomize virtual address space */ + KERN_SETUID_DUMPABLE=69, /* int: behaviour of dumps for setuid core */ }; diff --git a/kernel/sys.c b/kernel/sys.c index f006632c2ba..0a2c8cda963 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -525,7 +525,7 @@ asmlinkage long sys_setregid(gid_t rgid, gid_t egid) } if (new_egid != old_egid) { - current->mm->dumpable = 0; + current->mm->dumpable = suid_dumpable; smp_wmb(); } if (rgid != (gid_t) -1 || @@ -556,7 +556,7 @@ asmlinkage long sys_setgid(gid_t gid) { if(old_egid != gid) { - current->mm->dumpable=0; + current->mm->dumpable = suid_dumpable; smp_wmb(); } current->gid = current->egid = current->sgid = current->fsgid = gid; @@ -565,7 +565,7 @@ asmlinkage long sys_setgid(gid_t gid) { if(old_egid != gid) { - current->mm->dumpable=0; + current->mm->dumpable = suid_dumpable; smp_wmb(); } current->egid = current->fsgid = gid; @@ -596,7 +596,7 @@ static int set_user(uid_t new_ruid, int dumpclear) if(dumpclear) { - current->mm->dumpable = 0; + current->mm->dumpable = suid_dumpable; smp_wmb(); } current->uid = new_ruid; @@ -653,7 +653,7 @@ asmlinkage long sys_setreuid(uid_t ruid, uid_t euid) if (new_euid != old_euid) { - current->mm->dumpable=0; + current->mm->dumpable = suid_dumpable; smp_wmb(); } current->fsuid = current->euid = new_euid; @@ -703,7 +703,7 @@ asmlinkage long sys_setuid(uid_t uid) if (old_euid != uid) { - current->mm->dumpable = 0; + current->mm->dumpable = suid_dumpable; smp_wmb(); } current->fsuid = current->euid = uid; @@ -748,7 +748,7 @@ asmlinkage long sys_setresuid(uid_t ruid, uid_t euid, uid_t suid) if (euid != (uid_t) -1) { if (euid != current->euid) { - current->mm->dumpable = 0; + current->mm->dumpable = suid_dumpable; smp_wmb(); } current->euid = euid; @@ -798,7 +798,7 @@ asmlinkage long sys_setresgid(gid_t rgid, gid_t egid, gid_t sgid) if (egid != (gid_t) -1) { if (egid != current->egid) { - current->mm->dumpable = 0; + current->mm->dumpable = suid_dumpable; smp_wmb(); } current->egid = egid; @@ -845,7 +845,7 @@ asmlinkage long sys_setfsuid(uid_t uid) { if (uid != old_fsuid) { - current->mm->dumpable = 0; + current->mm->dumpable = suid_dumpable; smp_wmb(); } current->fsuid = uid; @@ -875,7 +875,7 @@ asmlinkage long sys_setfsgid(gid_t gid) { if (gid != old_fsgid) { - current->mm->dumpable = 0; + current->mm->dumpable = suid_dumpable; smp_wmb(); } current->fsgid = gid; @@ -1652,7 +1652,7 @@ asmlinkage long sys_prctl(int option, unsigned long arg2, unsigned long arg3, error = 1; break; case PR_SET_DUMPABLE: - if (arg2 != 0 && arg2 != 1) { + if (arg2 < 0 || arg2 > 2) { error = -EINVAL; break; } diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 701d12c6306..24a4d12d5aa 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -58,6 +58,7 @@ extern int sysctl_overcommit_ratio; extern int max_threads; extern int sysrq_enabled; extern int core_uses_pid; +extern int suid_dumpable; extern char core_pattern[]; extern int cad_pid; extern int pid_max; @@ -950,6 +951,14 @@ static ctl_table fs_table[] = { .proc_handler = &proc_dointvec, }, #endif + { + .ctl_name = KERN_SETUID_DUMPABLE, + .procname = "suid_dumpable", + .data = &suid_dumpable, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = &proc_dointvec, + }, { .ctl_name = 0 } }; diff --git a/security/commoncap.c b/security/commoncap.c index 849b8c338ee..04c12f58d65 100644 --- a/security/commoncap.c +++ b/security/commoncap.c @@ -149,7 +149,7 @@ void cap_bprm_apply_creds (struct linux_binprm *bprm, int unsafe) if (bprm->e_uid != current->uid || bprm->e_gid != current->gid || !cap_issubset (new_permitted, current->cap_permitted)) { - current->mm->dumpable = 0; + current->mm->dumpable = suid_dumpable; if (unsafe & ~LSM_UNSAFE_PTRACE_CAP) { if (!capable(CAP_SETUID)) { diff --git a/security/dummy.c b/security/dummy.c index b32eff14654..6ff88758647 100644 --- a/security/dummy.c +++ b/security/dummy.c @@ -130,7 +130,7 @@ static void dummy_bprm_free_security (struct linux_binprm *bprm) static void dummy_bprm_apply_creds (struct linux_binprm *bprm, int unsafe) { if (bprm->e_uid != current->uid || bprm->e_gid != current->gid) { - current->mm->dumpable = 0; + current->mm->dumpable = suid_dumpable; if ((unsafe & ~LSM_UNSAFE_PTRACE_CAP) && !capable(CAP_SETUID)) { bprm->e_uid = current->uid; -- cgit v1.2.3-70-g09d2 From 4fea2838aa00b9e59efde974dcdb455608192811 Mon Sep 17 00:00:00 2001 From: Kirill Korotaev Date: Thu, 23 Jun 2005 00:09:51 -0700 Subject: [PATCH] Software suspend and recalc sigpending bug fix This patch fixes recalc_sigpending() to work correctly with tasks which are being freezed. The problem is that freeze_processes() sets PF_FREEZE and TIF_SIGPENDING flags on tasks, but recalc_sigpending() called from e.g. sys_rt_sigtimedwait or any other kernel place will clear TIF_SIGPENDING due to no pending signals queued and the tasks won't be freezed until it recieves a real signal or freezed_processes() fail due to timeout. Signed-Off-By: Kirill Korotaev Signed-Off-By: Alexey Kuznetsov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/signal.c | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel') diff --git a/kernel/signal.c b/kernel/signal.c index c89821b69ae..d1258729a5f 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -213,6 +213,7 @@ static inline int has_pending_signals(sigset_t *signal, sigset_t *blocked) fastcall void recalc_sigpending_tsk(struct task_struct *t) { if (t->signal->group_stop_count > 0 || + (t->flags & PF_FREEZE) || PENDING(&t->pending, &t->blocked) || PENDING(&t->signal->shared_pending, &t->blocked)) set_tsk_thread_flag(t, TIF_SIGPENDING); -- cgit v1.2.3-70-g09d2 From 71a2224d7d1cefc23a1ac80bba421cc069cc3257 Mon Sep 17 00:00:00 2001 From: Christoph Lameter Date: Thu, 23 Jun 2005 00:10:05 -0700 Subject: [PATCH] Optimize sys_times for a single thread process Avoid taking the tasklist_lock in sys_times if the process is single threaded. In a NUMA system taking the tasklist_lock may cause a bouncing cacheline if multiple independent processes continually call sys_times to measure their performance. Signed-off-by: Christoph Lameter Signed-off-by: Shai Fultheim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/exit.c | 5 ++++ kernel/sys.c | 86 +++++++++++++++++++++++++++++++++++++++++------------------ 2 files changed, 65 insertions(+), 26 deletions(-) (limited to 'kernel') diff --git a/kernel/exit.c b/kernel/exit.c index c2bdf6fb61a..3ebcd60a19c 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -72,6 +72,11 @@ repeat: BUG_ON(!list_empty(&p->ptrace_list) || !list_empty(&p->ptrace_children)); __exit_signal(p); __exit_sighand(p); + /* + * Note that the fastpath in sys_times depends on __exit_signal having + * updated the counters before a task is removed from the tasklist of + * the process by __unhash_process. + */ __unhash_process(p); /* diff --git a/kernel/sys.c b/kernel/sys.c index 0a2c8cda963..5a9d6b07501 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -894,35 +894,69 @@ asmlinkage long sys_times(struct tms __user * tbuf) */ if (tbuf) { struct tms tmp; - struct task_struct *tsk = current; - struct task_struct *t; cputime_t utime, stime, cutime, cstime; - read_lock(&tasklist_lock); - utime = tsk->signal->utime; - stime = tsk->signal->stime; - t = tsk; - do { - utime = cputime_add(utime, t->utime); - stime = cputime_add(stime, t->stime); - t = next_thread(t); - } while (t != tsk); - - /* - * While we have tasklist_lock read-locked, no dying thread - * can be updating current->signal->[us]time. Instead, - * we got their counts included in the live thread loop. - * However, another thread can come in right now and - * do a wait call that updates current->signal->c[us]time. - * To make sure we always see that pair updated atomically, - * we take the siglock around fetching them. - */ - spin_lock_irq(&tsk->sighand->siglock); - cutime = tsk->signal->cutime; - cstime = tsk->signal->cstime; - spin_unlock_irq(&tsk->sighand->siglock); - read_unlock(&tasklist_lock); +#ifdef CONFIG_SMP + if (thread_group_empty(current)) { + /* + * Single thread case without the use of any locks. + * + * We may race with release_task if two threads are + * executing. However, release task first adds up the + * counters (__exit_signal) before removing the task + * from the process tasklist (__unhash_process). + * __exit_signal also acquires and releases the + * siglock which results in the proper memory ordering + * so that the list modifications are always visible + * after the counters have been updated. + * + * If the counters have been updated by the second thread + * but the thread has not yet been removed from the list + * then the other branch will be executing which will + * block on tasklist_lock until the exit handling of the + * other task is finished. + * + * This also implies that the sighand->siglock cannot + * be held by another processor. So we can also + * skip acquiring that lock. + */ + utime = cputime_add(current->signal->utime, current->utime); + stime = cputime_add(current->signal->utime, current->stime); + cutime = current->signal->cutime; + cstime = current->signal->cstime; + } else +#endif + { + + /* Process with multiple threads */ + struct task_struct *tsk = current; + struct task_struct *t; + read_lock(&tasklist_lock); + utime = tsk->signal->utime; + stime = tsk->signal->stime; + t = tsk; + do { + utime = cputime_add(utime, t->utime); + stime = cputime_add(stime, t->stime); + t = next_thread(t); + } while (t != tsk); + + /* + * While we have tasklist_lock read-locked, no dying thread + * can be updating current->signal->[us]time. Instead, + * we got their counts included in the live thread loop. + * However, another thread can come in right now and + * do a wait call that updates current->signal->c[us]time. + * To make sure we always see that pair updated atomically, + * we take the siglock around fetching them. + */ + spin_lock_irq(&tsk->sighand->siglock); + cutime = tsk->signal->cutime; + cstime = tsk->signal->cstime; + spin_unlock_irq(&tsk->sighand->siglock); + read_unlock(&tasklist_lock); + } tmp.tms_utime = cputime_to_clock_t(utime); tmp.tms_stime = cputime_to_clock_t(stime); tmp.tms_cutime = cputime_to_clock_t(cutime); -- cgit v1.2.3-70-g09d2 From c43dc2fd885b5658cfd7cedb7bcca20910c517a4 Mon Sep 17 00:00:00 2001 From: Benjamin LaHaise Date: Thu, 23 Jun 2005 00:10:27 -0700 Subject: [PATCH] aio: make wait_queue ->task ->private In the upcoming aio_down patch, it is useful to store a private data pointer in the kiocb's wait_queue. Since we provide our own wake up function and do not require the task_struct pointer, it makes sense to convert the task pointer into a generic private pointer. Signed-off-by: Benjamin LaHaise Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/wait.h | 16 ++++++++-------- kernel/sched.c | 2 +- 2 files changed, 9 insertions(+), 9 deletions(-) (limited to 'kernel') diff --git a/include/linux/wait.h b/include/linux/wait.h index c9486c3efb4..d38c9fecdc3 100644 --- a/include/linux/wait.h +++ b/include/linux/wait.h @@ -33,7 +33,7 @@ int default_wake_function(wait_queue_t *wait, unsigned mode, int sync, void *key struct __wait_queue { unsigned int flags; #define WQ_FLAG_EXCLUSIVE 0x01 - struct task_struct * task; + void *private; wait_queue_func_t func; struct list_head task_list; }; @@ -60,7 +60,7 @@ typedef struct __wait_queue_head wait_queue_head_t; */ #define __WAITQUEUE_INITIALIZER(name, tsk) { \ - .task = tsk, \ + .private = tsk, \ .func = default_wake_function, \ .task_list = { NULL, NULL } } @@ -86,7 +86,7 @@ static inline void init_waitqueue_head(wait_queue_head_t *q) static inline void init_waitqueue_entry(wait_queue_t *q, struct task_struct *p) { q->flags = 0; - q->task = p; + q->private = p; q->func = default_wake_function; } @@ -94,7 +94,7 @@ static inline void init_waitqueue_func_entry(wait_queue_t *q, wait_queue_func_t func) { q->flags = 0; - q->task = NULL; + q->private = NULL; q->func = func; } @@ -110,7 +110,7 @@ static inline int waitqueue_active(wait_queue_head_t *q) * aio specifies a wait queue entry with an async notification * callback routine, not associated with any task. */ -#define is_sync_wait(wait) (!(wait) || ((wait)->task)) +#define is_sync_wait(wait) (!(wait) || ((wait)->private)) extern void FASTCALL(add_wait_queue(wait_queue_head_t *q, wait_queue_t * wait)); extern void FASTCALL(add_wait_queue_exclusive(wait_queue_head_t *q, wait_queue_t * wait)); @@ -384,7 +384,7 @@ int wake_bit_function(wait_queue_t *wait, unsigned mode, int sync, void *key); #define DEFINE_WAIT(name) \ wait_queue_t name = { \ - .task = current, \ + .private = current, \ .func = autoremove_wake_function, \ .task_list = LIST_HEAD_INIT((name).task_list), \ } @@ -393,7 +393,7 @@ int wake_bit_function(wait_queue_t *wait, unsigned mode, int sync, void *key); struct wait_bit_queue name = { \ .key = __WAIT_BIT_KEY_INITIALIZER(word, bit), \ .wait = { \ - .task = current, \ + .private = current, \ .func = wake_bit_function, \ .task_list = \ LIST_HEAD_INIT((name).wait.task_list), \ @@ -402,7 +402,7 @@ int wake_bit_function(wait_queue_t *wait, unsigned mode, int sync, void *key); #define init_wait(wait) \ do { \ - (wait)->task = current; \ + (wait)->private = current; \ (wait)->func = autoremove_wake_function; \ INIT_LIST_HEAD(&(wait)->task_list); \ } while (0) diff --git a/kernel/sched.c b/kernel/sched.c index 6ee4515d5a2..76080d142e3 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -2869,7 +2869,7 @@ need_resched: int default_wake_function(wait_queue_t *curr, unsigned mode, int sync, void *key) { - task_t *p = curr->task; + task_t *p = curr->private; return try_to_wake_up(p, mode, sync); } -- cgit v1.2.3-70-g09d2 From 7888e7ff4ee579442128d7d12a9c9dbf2cf7de6a Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 23 Jun 2005 22:00:51 -0700 Subject: [PATCH] Keys: Pass session keyring to call_usermodehelper() The attached patch makes it possible to pass a session keyring through to the process spawned by call_usermodehelper(). This allows patch 3/3 to pass an authorisation key through to /sbin/request-key, thus permitting better access controls when doing just-in-time key creation. Signed-Off-By: David Howells Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/key.h | 10 +++++++++- include/linux/kmod.h | 13 ++++++++++++- kernel/kmod.c | 17 +++++++++++++---- security/keys/request_key.c | 2 +- 4 files changed, 35 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/include/linux/key.h b/include/linux/key.h index 2c24ffaca86..2bfbf88d274 100644 --- a/include/linux/key.h +++ b/include/linux/key.h @@ -273,14 +273,22 @@ extern void key_fsuid_changed(struct task_struct *tsk); extern void key_fsgid_changed(struct task_struct *tsk); extern void key_init(void); +#define __install_session_keyring(tsk, keyring) \ +({ \ + struct key *old_session = tsk->signal->session_keyring; \ + tsk->signal->session_keyring = keyring; \ + old_session; \ +}) + #else /* CONFIG_KEYS */ #define key_validate(k) 0 #define key_serial(k) 0 -#define key_get(k) NULL +#define key_get(k) ({ NULL; }) #define key_put(k) do { } while(0) #define alloc_uid_keyring(u) 0 #define switch_uid_keyring(u) do { } while(0) +#define __install_session_keyring(t, k) ({ NULL; }) #define copy_keys(f,t) 0 #define copy_thread_group_keys(t) 0 #define exit_keys(t) do { } while(0) diff --git a/include/linux/kmod.h b/include/linux/kmod.h index 95d0e4b0814..e4a23154940 100644 --- a/include/linux/kmod.h +++ b/include/linux/kmod.h @@ -19,6 +19,7 @@ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ +#include #include #include #include @@ -34,7 +35,17 @@ static inline int request_module(const char * name, ...) { return -ENOSYS; } #endif #define try_then_request_module(x, mod...) ((x) ?: (request_module(mod), (x))) -extern int call_usermodehelper(char *path, char *argv[], char *envp[], int wait); + +struct key; +extern int call_usermodehelper_keys(char *path, char *argv[], char *envp[], + struct key *session_keyring, int wait); + +static inline int +call_usermodehelper(char *path, char **argv, char **envp, int wait) +{ + return call_usermodehelper_keys(path, argv, envp, NULL, wait); +} + extern void usermodehelper_init(void); #endif /* __LINUX_KMOD_H__ */ diff --git a/kernel/kmod.c b/kernel/kmod.c index eed53d4f523..44166e3bb8a 100644 --- a/kernel/kmod.c +++ b/kernel/kmod.c @@ -120,6 +120,7 @@ struct subprocess_info { char *path; char **argv; char **envp; + struct key *ring; int wait; int retval; }; @@ -130,16 +131,21 @@ struct subprocess_info { static int ____call_usermodehelper(void *data) { struct subprocess_info *sub_info = data; + struct key *old_session; int retval; - /* Unblock all signals. */ + /* Unblock all signals and set the session keyring. */ + key_get(sub_info->ring); flush_signals(current); spin_lock_irq(¤t->sighand->siglock); + old_session = __install_session_keyring(current, sub_info->ring); flush_signal_handlers(current, 1); sigemptyset(¤t->blocked); recalc_sigpending(); spin_unlock_irq(¤t->sighand->siglock); + key_put(old_session); + /* We can run anywhere, unlike our parent keventd(). */ set_cpus_allowed(current, CPU_MASK_ALL); @@ -211,10 +217,11 @@ static void __call_usermodehelper(void *data) } /** - * call_usermodehelper - start a usermode application + * call_usermodehelper_keys - start a usermode application * @path: pathname for the application * @argv: null-terminated argument list * @envp: null-terminated environment list + * @session_keyring: session keyring for process (NULL for an empty keyring) * @wait: wait for the application to finish and return status. * * Runs a user-space application. The application is started @@ -224,7 +231,8 @@ static void __call_usermodehelper(void *data) * Must be called from process context. Returns a negative error code * if program was not execed successfully, or 0. */ -int call_usermodehelper(char *path, char **argv, char **envp, int wait) +int call_usermodehelper_keys(char *path, char **argv, char **envp, + struct key *session_keyring, int wait) { DECLARE_COMPLETION(done); struct subprocess_info sub_info = { @@ -232,6 +240,7 @@ int call_usermodehelper(char *path, char **argv, char **envp, int wait) .path = path, .argv = argv, .envp = envp, + .ring = session_keyring, .wait = wait, .retval = 0, }; @@ -247,7 +256,7 @@ int call_usermodehelper(char *path, char **argv, char **envp, int wait) wait_for_completion(&done); return sub_info.retval; } -EXPORT_SYMBOL(call_usermodehelper); +EXPORT_SYMBOL(call_usermodehelper_keys); void __init usermodehelper_init(void) { diff --git a/security/keys/request_key.c b/security/keys/request_key.c index 1f6c0940297..1919540f047 100644 --- a/security/keys/request_key.c +++ b/security/keys/request_key.c @@ -88,7 +88,7 @@ static int call_request_key(struct key *key, argv[i] = NULL; /* do it */ - return call_usermodehelper(argv[0], argv, envp, 1); + return call_usermodehelper_keys(argv[0], argv, envp, NULL, 1); } /* end call_request_key() */ -- cgit v1.2.3-70-g09d2 From 3e30148c3d524a9c1c63ca28261bc24c457eb07a Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 23 Jun 2005 22:00:56 -0700 Subject: [PATCH] Keys: Make request-key create an authorisation key The attached patch makes the following changes: (1) There's a new special key type called ".request_key_auth". This is an authorisation key for when one process requests a key and another process is started to construct it. This type of key cannot be created by the user; nor can it be requested by kernel services. Authorisation keys hold two references: (a) Each refers to a key being constructed. When the key being constructed is instantiated the authorisation key is revoked, rendering it of no further use. (b) The "authorising process". This is either: (i) the process that called request_key(), or: (ii) if the process that called request_key() itself had an authorisation key in its session keyring, then the authorising process referred to by that authorisation key will also be referred to by the new authorisation key. This means that the process that initiated a chain of key requests will authorise the lot of them, and will, by default, wind up with the keys obtained from them in its keyrings. (2) request_key() creates an authorisation key which is then passed to /sbin/request-key in as part of a new session keyring. (3) When request_key() is searching for a key to hand back to the caller, if it comes across an authorisation key in the session keyring of the calling process, it will also search the keyrings of the process specified therein and it will use the specified process's credentials (fsuid, fsgid, groups) to do that rather than the calling process's credentials. This allows a process started by /sbin/request-key to find keys belonging to the authorising process. (4) A key can be read, even if the process executing KEYCTL_READ doesn't have direct read or search permission if that key is contained within the keyrings of a process specified by an authorisation key found within the calling process's session keyring, and is searchable using the credentials of the authorising process. This allows a process started by /sbin/request-key to read keys belonging to the authorising process. (5) The magic KEY_SPEC_*_KEYRING key IDs when passed to KEYCTL_INSTANTIATE or KEYCTL_NEGATE will specify a keyring of the authorising process, rather than the process doing the instantiation. (6) One of the process keyrings can be nominated as the default to which request_key() should attach new keys if not otherwise specified. This is done with KEYCTL_SET_REQKEY_KEYRING and one of the KEY_REQKEY_DEFL_* constants. The current setting can also be read using this call. (7) request_key() is partially interruptible. If it is waiting for another process to finish constructing a key, it can be interrupted. This permits a request-key cycle to be broken without recourse to rebooting. Signed-Off-By: David Howells Signed-Off-By: Benoit Boissinot Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/keys.txt | 34 ++++++++ include/linux/key-ui.h | 41 ++++++++- include/linux/key.h | 9 +- include/linux/keyctl.h | 11 +++ include/linux/sched.h | 8 +- kernel/sys.c | 2 +- security/keys/Makefile | 5 +- security/keys/compat.c | 7 +- security/keys/internal.h | 45 +++++++++- security/keys/key.c | 24 ++++-- security/keys/keyctl.c | 176 ++++++++++++++++++++++++------------- security/keys/keyring.c | 67 ++++++++++++-- security/keys/process_keys.c | 179 +++++++++++++++++++++++--------------- security/keys/request_key.c | 182 ++++++++++++++++++++++++++++++++------- security/keys/request_key_auth.c | 180 ++++++++++++++++++++++++++++++++++++++ 15 files changed, 779 insertions(+), 191 deletions(-) create mode 100644 security/keys/request_key_auth.c (limited to 'kernel') diff --git a/Documentation/keys.txt b/Documentation/keys.txt index 3df40c1fe15..0321ded4b9a 100644 --- a/Documentation/keys.txt +++ b/Documentation/keys.txt @@ -591,6 +591,37 @@ The keyctl syscall functions are: this case too. + (*) Set the default request-key destination keyring. + + long keyctl(KEYCTL_SET_REQKEY_KEYRING, int reqkey_defl); + + This sets the default keyring to which implicitly requested keys will be + attached for this thread. reqkey_defl should be one of these constants: + + CONSTANT VALUE NEW DEFAULT KEYRING + ====================================== ====== ======================= + KEY_REQKEY_DEFL_NO_CHANGE -1 No change + KEY_REQKEY_DEFL_DEFAULT 0 Default[1] + KEY_REQKEY_DEFL_THREAD_KEYRING 1 Thread keyring + KEY_REQKEY_DEFL_PROCESS_KEYRING 2 Process keyring + KEY_REQKEY_DEFL_SESSION_KEYRING 3 Session keyring + KEY_REQKEY_DEFL_USER_KEYRING 4 User keyring + KEY_REQKEY_DEFL_USER_SESSION_KEYRING 5 User session keyring + KEY_REQKEY_DEFL_GROUP_KEYRING 6 Group keyring + + The old default will be returned if successful and error EINVAL will be + returned if reqkey_defl is not one of the above values. + + The default keyring can be overridden by the keyring indicated to the + request_key() system call. + + Note that this setting is inherited across fork/exec. + + [1] The default default is: the thread keyring if there is one, otherwise + the process keyring if there is one, otherwise the session keyring if + there is one, otherwise the user default session keyring. + + =============== KERNEL SERVICES =============== @@ -626,6 +657,9 @@ payload contents" for more information. Should the function fail error ENOKEY, EKEYEXPIRED or EKEYREVOKED will be returned. + If successful, the key will have been attached to the default keyring for + implicitly obtained request-key keys, as set by KEYCTL_SET_REQKEY_KEYRING. + (*) When it is no longer required, the key should be released using: diff --git a/include/linux/key-ui.h b/include/linux/key-ui.h index 159ca8d54e9..cc326174a80 100644 --- a/include/linux/key-ui.h +++ b/include/linux/key-ui.h @@ -1,4 +1,4 @@ -/* key-ui.h: key userspace interface stuff for use by keyfs +/* key-ui.h: key userspace interface stuff * * Copyright (C) 2004 Red Hat, Inc. All Rights Reserved. * Written by David Howells (dhowells@redhat.com) @@ -84,8 +84,45 @@ static inline int key_any_permission(const struct key *key, key_perm_t perm) return kperm != 0; } +static inline int key_task_groups_search(struct task_struct *tsk, gid_t gid) +{ + int ret; + + task_lock(tsk); + ret = groups_search(tsk->group_info, gid); + task_unlock(tsk); + return ret; +} + +static inline int key_task_permission(const struct key *key, + struct task_struct *context, + key_perm_t perm) +{ + key_perm_t kperm; + + if (key->uid == context->fsuid) { + kperm = key->perm >> 16; + } + else if (key->gid != -1 && + key->perm & KEY_GRP_ALL && ( + key->gid == context->fsgid || + key_task_groups_search(context, key->gid) + ) + ) { + kperm = key->perm >> 8; + } + else { + kperm = key->perm; + } + + kperm = kperm & perm & KEY_ALL; + + return kperm == perm; + +} -extern struct key *lookup_user_key(key_serial_t id, int create, int part, +extern struct key *lookup_user_key(struct task_struct *context, + key_serial_t id, int create, int partial, key_perm_t perm); extern long join_session_keyring(const char *name); diff --git a/include/linux/key.h b/include/linux/key.h index 2bfbf88d274..970bbd916cf 100644 --- a/include/linux/key.h +++ b/include/linux/key.h @@ -199,10 +199,12 @@ extern int key_payload_reserve(struct key *key, size_t datalen); extern int key_instantiate_and_link(struct key *key, const void *data, size_t datalen, - struct key *keyring); + struct key *keyring, + struct key *instkey); extern int key_negate_and_link(struct key *key, unsigned timeout, - struct key *keyring); + struct key *keyring, + struct key *instkey); extern void key_revoke(struct key *key); extern void key_put(struct key *key); @@ -245,9 +247,6 @@ extern struct key *keyring_search(struct key *keyring, struct key_type *type, const char *description); -extern struct key *search_process_keyrings(struct key_type *type, - const char *description); - extern int keyring_add_key(struct key *keyring, struct key *key); diff --git a/include/linux/keyctl.h b/include/linux/keyctl.h index 381dedc370a..8d7c59a29e0 100644 --- a/include/linux/keyctl.h +++ b/include/linux/keyctl.h @@ -20,6 +20,16 @@ #define KEY_SPEC_USER_SESSION_KEYRING -5 /* - key ID for UID-session keyring */ #define KEY_SPEC_GROUP_KEYRING -6 /* - key ID for GID-specific keyring */ +/* request-key default keyrings */ +#define KEY_REQKEY_DEFL_NO_CHANGE -1 +#define KEY_REQKEY_DEFL_DEFAULT 0 +#define KEY_REQKEY_DEFL_THREAD_KEYRING 1 +#define KEY_REQKEY_DEFL_PROCESS_KEYRING 2 +#define KEY_REQKEY_DEFL_SESSION_KEYRING 3 +#define KEY_REQKEY_DEFL_USER_KEYRING 4 +#define KEY_REQKEY_DEFL_USER_SESSION_KEYRING 5 +#define KEY_REQKEY_DEFL_GROUP_KEYRING 6 + /* keyctl commands */ #define KEYCTL_GET_KEYRING_ID 0 /* ask for a keyring's ID */ #define KEYCTL_JOIN_SESSION_KEYRING 1 /* join or start named session keyring */ @@ -35,5 +45,6 @@ #define KEYCTL_READ 11 /* read a key or keyring's contents */ #define KEYCTL_INSTANTIATE 12 /* instantiate a partially constructed key */ #define KEYCTL_NEGATE 13 /* negate a partially constructed key */ +#define KEYCTL_SET_REQKEY_KEYRING 14 /* set default request-key keyring */ #endif /* _LINUX_KEYCTL_H */ diff --git a/include/linux/sched.h b/include/linux/sched.h index 901742f9238..2c69682b044 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -561,9 +561,10 @@ struct group_info { groups_free(group_info); \ } while (0) -struct group_info *groups_alloc(int gidsetsize); -void groups_free(struct group_info *group_info); -int set_current_groups(struct group_info *group_info); +extern struct group_info *groups_alloc(int gidsetsize); +extern void groups_free(struct group_info *group_info); +extern int set_current_groups(struct group_info *group_info); +extern int groups_search(struct group_info *group_info, gid_t grp); /* access the groups "array" with this macro */ #define GROUP_AT(gi, i) \ ((gi)->blocks[(i)/NGROUPS_PER_BLOCK][(i)%NGROUPS_PER_BLOCK]) @@ -660,6 +661,7 @@ struct task_struct { struct user_struct *user; #ifdef CONFIG_KEYS struct key *thread_keyring; /* keyring private to this thread */ + unsigned char jit_keyring; /* default keyring to attach requested keys to */ #endif int oomkilladj; /* OOM kill score adjustment (bit shift). */ char comm[TASK_COMM_LEN]; /* executable name excluding path diff --git a/kernel/sys.c b/kernel/sys.c index 5a9d6b07501..da24bc1292d 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -1259,7 +1259,7 @@ static void groups_sort(struct group_info *group_info) } /* a simple bsearch */ -static int groups_search(struct group_info *group_info, gid_t grp) +int groups_search(struct group_info *group_info, gid_t grp) { int left, right; diff --git a/security/keys/Makefile b/security/keys/Makefile index ddb495d6506..c392d750b20 100644 --- a/security/keys/Makefile +++ b/security/keys/Makefile @@ -7,8 +7,9 @@ obj-y := \ keyring.o \ keyctl.o \ process_keys.o \ - user_defined.o \ - request_key.o + request_key.o \ + request_key_auth.o \ + user_defined.o obj-$(CONFIG_KEYS_COMPAT) += compat.o obj-$(CONFIG_PROC_FS) += proc.o diff --git a/security/keys/compat.c b/security/keys/compat.c index aff8b22dcb5..3303673c636 100644 --- a/security/keys/compat.c +++ b/security/keys/compat.c @@ -1,6 +1,6 @@ /* compat.c: 32-bit compatibility syscall for 64-bit systems * - * Copyright (C) 2004 Red Hat, Inc. All Rights Reserved. + * Copyright (C) 2004-5 Red Hat, Inc. All Rights Reserved. * Written by David Howells (dhowells@redhat.com) * * This program is free software; you can redistribute it and/or @@ -24,7 +24,7 @@ * - if you can, you should call sys_keyctl directly */ asmlinkage long compat_sys_keyctl(u32 option, - u32 arg2, u32 arg3, u32 arg4, u32 arg5) + u32 arg2, u32 arg3, u32 arg4, u32 arg5) { switch (option) { case KEYCTL_GET_KEYRING_ID: @@ -71,6 +71,9 @@ asmlinkage long compat_sys_keyctl(u32 option, case KEYCTL_NEGATE: return keyctl_negate_key(arg2, arg3, arg4); + case KEYCTL_SET_REQKEY_KEYRING: + return keyctl_set_reqkey_keyring(arg2); + default: return -EOPNOTSUPP; } diff --git a/security/keys/internal.h b/security/keys/internal.h index 67b2b93a748..46c8602661c 100644 --- a/security/keys/internal.h +++ b/security/keys/internal.h @@ -1,6 +1,6 @@ /* internal.h: authentication token and access key management internal defs * - * Copyright (C) 2003 Red Hat, Inc. All Rights Reserved. + * Copyright (C) 2003-5 Red Hat, Inc. All Rights Reserved. * Written by David Howells (dhowells@redhat.com) * * This program is free software; you can redistribute it and/or @@ -15,6 +15,16 @@ #include #include +#if 0 +#define kenter(FMT, a...) printk("==> %s("FMT")\n",__FUNCTION__ , ## a) +#define kleave(FMT, a...) printk("<== %s()"FMT"\n",__FUNCTION__ , ## a) +#define kdebug(FMT, a...) printk(FMT"\n" , ## a) +#else +#define kenter(FMT, a...) do {} while(0) +#define kleave(FMT, a...) do {} while(0) +#define kdebug(FMT, a...) do {} while(0) +#endif + extern struct key_type key_type_dead; extern struct key_type key_type_user; @@ -66,20 +76,46 @@ extern struct key *__keyring_search_one(struct key *keyring, const char *description, key_perm_t perm); +extern struct key *keyring_search_instkey(struct key *keyring, + key_serial_t target_id); + typedef int (*key_match_func_t)(const struct key *, const void *); extern struct key *keyring_search_aux(struct key *keyring, + struct task_struct *tsk, struct key_type *type, const void *description, key_match_func_t match); -extern struct key *search_process_keyrings_aux(struct key_type *type, - const void *description, - key_match_func_t match); +extern struct key *search_process_keyrings(struct key_type *type, + const void *description, + key_match_func_t match, + struct task_struct *tsk); extern struct key *find_keyring_by_name(const char *name, key_serial_t bound); extern int install_thread_keyring(struct task_struct *tsk); +extern int install_process_keyring(struct task_struct *tsk); + +extern struct key *request_key_and_link(struct key_type *type, + const char *description, + const char *callout_info, + struct key *dest_keyring); + +/* + * request_key authorisation + */ +struct request_key_auth { + struct key *target_key; + struct task_struct *context; + pid_t pid; +}; + +extern struct key_type key_type_request_key_auth; +extern struct key *request_key_auth_new(struct key *target, + struct key **_rkakey); + +extern struct key *key_get_instantiation_authkey(key_serial_t target_id); /* * keyctl functions @@ -100,6 +136,7 @@ extern long keyctl_setperm_key(key_serial_t, key_perm_t); extern long keyctl_instantiate_key(key_serial_t, const void __user *, size_t, key_serial_t); extern long keyctl_negate_key(key_serial_t, unsigned, key_serial_t); +extern long keyctl_set_reqkey_keyring(int); /* diff --git a/security/keys/key.c b/security/keys/key.c index 1fdfccb3fe4..3304d37bb37 100644 --- a/security/keys/key.c +++ b/security/keys/key.c @@ -1,6 +1,6 @@ /* key.c: basic authentication token and access key management * - * Copyright (C) 2004 Red Hat, Inc. All Rights Reserved. + * Copyright (C) 2004-5 Red Hat, Inc. All Rights Reserved. * Written by David Howells (dhowells@redhat.com) * * This program is free software; you can redistribute it and/or @@ -391,7 +391,8 @@ EXPORT_SYMBOL(key_payload_reserve); static int __key_instantiate_and_link(struct key *key, const void *data, size_t datalen, - struct key *keyring) + struct key *keyring, + struct key *instkey) { int ret, awaken; @@ -419,6 +420,10 @@ static int __key_instantiate_and_link(struct key *key, /* and link it into the destination keyring */ if (keyring) ret = __key_link(keyring, key); + + /* disable the authorisation key */ + if (instkey) + key_revoke(instkey); } } @@ -439,19 +444,21 @@ static int __key_instantiate_and_link(struct key *key, int key_instantiate_and_link(struct key *key, const void *data, size_t datalen, - struct key *keyring) + struct key *keyring, + struct key *instkey) { int ret; if (keyring) down_write(&keyring->sem); - ret = __key_instantiate_and_link(key, data, datalen, keyring); + ret = __key_instantiate_and_link(key, data, datalen, keyring, instkey); if (keyring) up_write(&keyring->sem); return ret; + } /* end key_instantiate_and_link() */ EXPORT_SYMBOL(key_instantiate_and_link); @@ -462,7 +469,8 @@ EXPORT_SYMBOL(key_instantiate_and_link); */ int key_negate_and_link(struct key *key, unsigned timeout, - struct key *keyring) + struct key *keyring, + struct key *instkey) { struct timespec now; int ret, awaken; @@ -495,6 +503,10 @@ int key_negate_and_link(struct key *key, /* and link it into the destination keyring */ if (keyring) ret = __key_link(keyring, key); + + /* disable the authorisation key */ + if (instkey) + key_revoke(instkey); } up_write(&key_construction_sem); @@ -781,7 +793,7 @@ struct key *key_create_or_update(struct key *keyring, } /* instantiate it and link it into the target keyring */ - ret = __key_instantiate_and_link(key, payload, plen, keyring); + ret = __key_instantiate_and_link(key, payload, plen, keyring, NULL); if (ret < 0) { key_put(key); key = ERR_PTR(ret); diff --git a/security/keys/keyctl.c b/security/keys/keyctl.c index cedb7326de2..fea262860ea 100644 --- a/security/keys/keyctl.c +++ b/security/keys/keyctl.c @@ -1,6 +1,6 @@ /* keyctl.c: userspace keyctl operations * - * Copyright (C) 2004 Red Hat, Inc. All Rights Reserved. + * Copyright (C) 2004-5 Red Hat, Inc. All Rights Reserved. * Written by David Howells (dhowells@redhat.com) * * This program is free software; you can redistribute it and/or @@ -49,6 +49,13 @@ asmlinkage long sys_add_key(const char __user *_type, goto error; type[31] = '\0'; + if (!type[0]) + goto error; + + ret = -EPERM; + if (type[0] == '.') + goto error; + ret = -EFAULT; dlen = strnlen_user(_description, PAGE_SIZE - 1); if (dlen <= 0) @@ -82,7 +89,7 @@ asmlinkage long sys_add_key(const char __user *_type, } /* find the target keyring (which must be writable) */ - keyring = lookup_user_key(ringid, 1, 0, KEY_WRITE); + keyring = lookup_user_key(NULL, ringid, 1, 0, KEY_WRITE); if (IS_ERR(keyring)) { ret = PTR_ERR(keyring); goto error3; @@ -181,7 +188,7 @@ asmlinkage long sys_request_key(const char __user *_type, /* get the destination keyring if specified */ dest = NULL; if (destringid) { - dest = lookup_user_key(destringid, 1, 0, KEY_WRITE); + dest = lookup_user_key(NULL, destringid, 1, 0, KEY_WRITE); if (IS_ERR(dest)) { ret = PTR_ERR(dest); goto error3; @@ -196,23 +203,15 @@ asmlinkage long sys_request_key(const char __user *_type, } /* do the search */ - key = request_key(ktype, description, callout_info); + key = request_key_and_link(ktype, description, callout_info, dest); if (IS_ERR(key)) { ret = PTR_ERR(key); goto error5; } - /* link the resulting key to the destination keyring */ - if (dest) { - ret = key_link(dest, key); - if (ret < 0) - goto error6; - } - ret = key->serial; - error6: - key_put(key); + key_put(key); error5: key_type_put(ktype); error4: @@ -237,7 +236,7 @@ long keyctl_get_keyring_ID(key_serial_t id, int create) struct key *key; long ret; - key = lookup_user_key(id, create, 0, KEY_SEARCH); + key = lookup_user_key(NULL, id, create, 0, KEY_SEARCH); if (IS_ERR(key)) { ret = PTR_ERR(key); goto error; @@ -324,7 +323,7 @@ long keyctl_update_key(key_serial_t id, } /* find the target key (which must be writable) */ - key = lookup_user_key(id, 0, 0, KEY_WRITE); + key = lookup_user_key(NULL, id, 0, 0, KEY_WRITE); if (IS_ERR(key)) { ret = PTR_ERR(key); goto error2; @@ -352,7 +351,7 @@ long keyctl_revoke_key(key_serial_t id) struct key *key; long ret; - key = lookup_user_key(id, 0, 0, KEY_WRITE); + key = lookup_user_key(NULL, id, 0, 0, KEY_WRITE); if (IS_ERR(key)) { ret = PTR_ERR(key); goto error; @@ -378,7 +377,7 @@ long keyctl_keyring_clear(key_serial_t ringid) struct key *keyring; long ret; - keyring = lookup_user_key(ringid, 1, 0, KEY_WRITE); + keyring = lookup_user_key(NULL, ringid, 1, 0, KEY_WRITE); if (IS_ERR(keyring)) { ret = PTR_ERR(keyring); goto error; @@ -404,13 +403,13 @@ long keyctl_keyring_link(key_serial_t id, key_serial_t ringid) struct key *keyring, *key; long ret; - keyring = lookup_user_key(ringid, 1, 0, KEY_WRITE); + keyring = lookup_user_key(NULL, ringid, 1, 0, KEY_WRITE); if (IS_ERR(keyring)) { ret = PTR_ERR(keyring); goto error; } - key = lookup_user_key(id, 1, 0, KEY_LINK); + key = lookup_user_key(NULL, id, 1, 0, KEY_LINK); if (IS_ERR(key)) { ret = PTR_ERR(key); goto error2; @@ -438,13 +437,13 @@ long keyctl_keyring_unlink(key_serial_t id, key_serial_t ringid) struct key *keyring, *key; long ret; - keyring = lookup_user_key(ringid, 0, 0, KEY_WRITE); + keyring = lookup_user_key(NULL, ringid, 0, 0, KEY_WRITE); if (IS_ERR(keyring)) { ret = PTR_ERR(keyring); goto error; } - key = lookup_user_key(id, 0, 0, 0); + key = lookup_user_key(NULL, id, 0, 0, 0); if (IS_ERR(key)) { ret = PTR_ERR(key); goto error2; @@ -475,16 +474,29 @@ long keyctl_describe_key(key_serial_t keyid, char __user *buffer, size_t buflen) { - struct key *key; + struct key *key, *instkey; char *tmpbuf; long ret; - key = lookup_user_key(keyid, 0, 1, KEY_VIEW); + key = lookup_user_key(NULL, keyid, 0, 1, KEY_VIEW); if (IS_ERR(key)) { + /* viewing a key under construction is permitted if we have the + * authorisation token handy */ + if (PTR_ERR(key) == -EACCES) { + instkey = key_get_instantiation_authkey(keyid); + if (!IS_ERR(instkey)) { + key_put(instkey); + key = lookup_user_key(NULL, keyid, 0, 1, 0); + if (!IS_ERR(key)) + goto okay; + } + } + ret = PTR_ERR(key); goto error; } +okay: /* calculate how much description we're going to return */ ret = -ENOMEM; tmpbuf = kmalloc(PAGE_SIZE, GFP_KERNEL); @@ -568,7 +580,7 @@ long keyctl_keyring_search(key_serial_t ringid, goto error2; /* get the keyring at which to begin the search */ - keyring = lookup_user_key(ringid, 0, 0, KEY_SEARCH); + keyring = lookup_user_key(NULL, ringid, 0, 0, KEY_SEARCH); if (IS_ERR(keyring)) { ret = PTR_ERR(keyring); goto error2; @@ -577,7 +589,7 @@ long keyctl_keyring_search(key_serial_t ringid, /* get the destination keyring if specified */ dest = NULL; if (destringid) { - dest = lookup_user_key(destringid, 1, 0, KEY_WRITE); + dest = lookup_user_key(NULL, destringid, 1, 0, KEY_WRITE); if (IS_ERR(dest)) { ret = PTR_ERR(dest); goto error3; @@ -656,24 +668,23 @@ long keyctl_read_key(key_serial_t keyid, char __user *buffer, size_t buflen) long ret; /* find the key first */ - key = lookup_user_key(keyid, 0, 0, 0); + key = lookup_user_key(NULL, keyid, 0, 0, 0); if (!IS_ERR(key)) { /* see if we can read it directly */ if (key_permission(key, KEY_READ)) goto can_read_key; - /* can't; see if it's searchable from this process's - * keyrings */ - ret = -ENOKEY; - if (key_permission(key, KEY_SEARCH)) { - /* okay - we do have search permission on the key - * itself, but do we have the key? */ - skey = search_process_keyrings_aux(key->type, key, - keyctl_read_key_same); - if (!IS_ERR(skey)) - goto can_read_key2; - } - + /* we can't; see if it's searchable from this process's + * keyrings + * - we automatically take account of the fact that it may be + * dangling off an instantiation key + */ + skey = search_process_keyrings(key->type, key, + keyctl_read_key_same, current); + if (!IS_ERR(skey)) + goto can_read_key2; + + ret = PTR_ERR(skey); goto error2; } @@ -719,7 +730,7 @@ long keyctl_chown_key(key_serial_t id, uid_t uid, gid_t gid) if (uid == (uid_t) -1 && gid == (gid_t) -1) goto error; - key = lookup_user_key(id, 1, 1, 0); + key = lookup_user_key(NULL, id, 1, 1, 0); if (IS_ERR(key)) { ret = PTR_ERR(key); goto error; @@ -776,7 +787,7 @@ long keyctl_setperm_key(key_serial_t id, key_perm_t perm) if (perm & ~(KEY_USR_ALL | KEY_GRP_ALL | KEY_OTH_ALL)) goto error; - key = lookup_user_key(id, 1, 1, 0); + key = lookup_user_key(NULL, id, 1, 1, 0); if (IS_ERR(key)) { ret = PTR_ERR(key); goto error; @@ -809,7 +820,8 @@ long keyctl_instantiate_key(key_serial_t id, size_t plen, key_serial_t ringid) { - struct key *key, *keyring; + struct request_key_auth *rka; + struct key *instkey, *keyring; void *payload; long ret; @@ -831,18 +843,21 @@ long keyctl_instantiate_key(key_serial_t id, goto error2; } - /* find the target key (which must be writable) */ - key = lookup_user_key(id, 0, 1, KEY_WRITE); - if (IS_ERR(key)) { - ret = PTR_ERR(key); + /* find the instantiation authorisation key */ + instkey = key_get_instantiation_authkey(id); + if (IS_ERR(instkey)) { + ret = PTR_ERR(instkey); goto error2; } - /* find the destination keyring if present (which must also be - * writable) */ + rka = instkey->payload.data; + + /* find the destination keyring amongst those belonging to the + * requesting task */ keyring = NULL; if (ringid) { - keyring = lookup_user_key(ringid, 1, 0, KEY_WRITE); + keyring = lookup_user_key(rka->context, ringid, 1, 0, + KEY_WRITE); if (IS_ERR(keyring)) { ret = PTR_ERR(keyring); goto error3; @@ -850,11 +865,12 @@ long keyctl_instantiate_key(key_serial_t id, } /* instantiate the key and link it into a keyring */ - ret = key_instantiate_and_link(key, payload, plen, keyring); + ret = key_instantiate_and_link(rka->target_key, payload, plen, + keyring, instkey); key_put(keyring); error3: - key_put(key); + key_put(instkey); error2: kfree(payload); error: @@ -869,21 +885,24 @@ long keyctl_instantiate_key(key_serial_t id, */ long keyctl_negate_key(key_serial_t id, unsigned timeout, key_serial_t ringid) { - struct key *key, *keyring; + struct request_key_auth *rka; + struct key *instkey, *keyring; long ret; - /* find the target key (which must be writable) */ - key = lookup_user_key(id, 0, 1, KEY_WRITE); - if (IS_ERR(key)) { - ret = PTR_ERR(key); + /* find the instantiation authorisation key */ + instkey = key_get_instantiation_authkey(id); + if (IS_ERR(instkey)) { + ret = PTR_ERR(instkey); goto error; } + rka = instkey->payload.data; + /* find the destination keyring if present (which must also be * writable) */ keyring = NULL; if (ringid) { - keyring = lookup_user_key(ringid, 1, 0, KEY_WRITE); + keyring = lookup_user_key(NULL, ringid, 1, 0, KEY_WRITE); if (IS_ERR(keyring)) { ret = PTR_ERR(keyring); goto error2; @@ -891,16 +910,54 @@ long keyctl_negate_key(key_serial_t id, unsigned timeout, key_serial_t ringid) } /* instantiate the key and link it into a keyring */ - ret = key_negate_and_link(key, timeout, keyring); + ret = key_negate_and_link(rka->target_key, timeout, keyring, instkey); key_put(keyring); error2: - key_put(key); + key_put(instkey); error: return ret; } /* end keyctl_negate_key() */ +/*****************************************************************************/ +/* + * set the default keyring in which request_key() will cache keys + * - return the old setting + */ +long keyctl_set_reqkey_keyring(int reqkey_defl) +{ + int ret; + + switch (reqkey_defl) { + case KEY_REQKEY_DEFL_THREAD_KEYRING: + ret = install_thread_keyring(current); + if (ret < 0) + return ret; + goto set; + + case KEY_REQKEY_DEFL_PROCESS_KEYRING: + ret = install_process_keyring(current); + if (ret < 0) + return ret; + + case KEY_REQKEY_DEFL_DEFAULT: + case KEY_REQKEY_DEFL_SESSION_KEYRING: + case KEY_REQKEY_DEFL_USER_KEYRING: + case KEY_REQKEY_DEFL_USER_SESSION_KEYRING: + set: + current->jit_keyring = reqkey_defl; + + case KEY_REQKEY_DEFL_NO_CHANGE: + return current->jit_keyring; + + case KEY_REQKEY_DEFL_GROUP_KEYRING: + default: + return -EINVAL; + } + +} /* end keyctl_set_reqkey_keyring() */ + /*****************************************************************************/ /* * the key control system call @@ -971,6 +1028,9 @@ asmlinkage long sys_keyctl(int option, unsigned long arg2, unsigned long arg3, (unsigned) arg3, (key_serial_t) arg4); + case KEYCTL_SET_REQKEY_KEYRING: + return keyctl_set_reqkey_keyring(arg2); + default: return -EOPNOTSUPP; } diff --git a/security/keys/keyring.c b/security/keys/keyring.c index c9a5de19748..90a551e4da6 100644 --- a/security/keys/keyring.c +++ b/security/keys/keyring.c @@ -1,6 +1,6 @@ /* keyring.c: keyring handling * - * Copyright (C) 2004 Red Hat, Inc. All Rights Reserved. + * Copyright (C) 2004-5 Red Hat, Inc. All Rights Reserved. * Written by David Howells (dhowells@redhat.com) * * This program is free software; you can redistribute it and/or @@ -308,7 +308,7 @@ struct key *keyring_alloc(const char *description, uid_t uid, gid_t gid, uid, gid, KEY_USR_ALL, not_in_quota); if (!IS_ERR(keyring)) { - ret = key_instantiate_and_link(keyring, NULL, 0, dest); + ret = key_instantiate_and_link(keyring, NULL, 0, dest, NULL); if (ret < 0) { key_put(keyring); keyring = ERR_PTR(ret); @@ -326,11 +326,12 @@ struct key *keyring_alloc(const char *description, uid_t uid, gid_t gid, * - we only find keys on which we have search permission * - we use the supplied match function to see if the description (or other * feature of interest) matches - * - we readlock the keyrings as we search down the tree + * - we rely on RCU to prevent the keyring lists from disappearing on us * - we return -EAGAIN if we didn't find any matching key * - we return -ENOKEY if we only found negative matching keys */ struct key *keyring_search_aux(struct key *keyring, + struct task_struct *context, struct key_type *type, const void *description, key_match_func_t match) @@ -352,7 +353,7 @@ struct key *keyring_search_aux(struct key *keyring, /* top keyring must have search permission to begin the search */ key = ERR_PTR(-EACCES); - if (!key_permission(keyring, KEY_SEARCH)) + if (!key_task_permission(keyring, context, KEY_SEARCH)) goto error; key = ERR_PTR(-ENOTDIR); @@ -392,7 +393,7 @@ struct key *keyring_search_aux(struct key *keyring, continue; /* key must have search permissions */ - if (!key_permission(key, KEY_SEARCH)) + if (!key_task_permission(key, context, KEY_SEARCH)) continue; /* we set a different error code if we find a negative key */ @@ -418,7 +419,7 @@ struct key *keyring_search_aux(struct key *keyring, if (sp >= KEYRING_SEARCH_MAX_DEPTH) continue; - if (!key_permission(key, KEY_SEARCH)) + if (!key_task_permission(key, context, KEY_SEARCH)) continue; /* stack the current position */ @@ -468,7 +469,11 @@ struct key *keyring_search(struct key *keyring, struct key_type *type, const char *description) { - return keyring_search_aux(keyring, type, description, type->match); + if (!type->match) + return ERR_PTR(-ENOKEY); + + return keyring_search_aux(keyring, current, + type, description, type->match); } /* end keyring_search() */ @@ -496,7 +501,8 @@ struct key *__keyring_search_one(struct key *keyring, key = klist->keys[loop]; if (key->type == ktype && - key->type->match(key, description) && + (!key->type->match || + key->type->match(key, description)) && key_permission(key, perm) && !test_bit(KEY_FLAG_REVOKED, &key->flags) ) @@ -515,6 +521,51 @@ struct key *__keyring_search_one(struct key *keyring, } /* end __keyring_search_one() */ +/*****************************************************************************/ +/* + * search for an instantiation authorisation key matching a target key + * - the RCU read lock must be held by the caller + * - a target_id of zero specifies any valid token + */ +struct key *keyring_search_instkey(struct key *keyring, + key_serial_t target_id) +{ + struct request_key_auth *rka; + struct keyring_list *klist; + struct key *instkey; + int loop; + + klist = rcu_dereference(keyring->payload.subscriptions); + if (klist) { + for (loop = 0; loop < klist->nkeys; loop++) { + instkey = klist->keys[loop]; + + if (instkey->type != &key_type_request_key_auth) + continue; + + rka = instkey->payload.data; + if (target_id && rka->target_key->serial != target_id) + continue; + + /* the auth key is revoked during instantiation */ + if (!test_bit(KEY_FLAG_REVOKED, &instkey->flags)) + goto found; + + instkey = ERR_PTR(-EKEYREVOKED); + goto error; + } + } + + instkey = ERR_PTR(-EACCES); + goto error; + +found: + atomic_inc(&instkey->usage); +error: + return instkey; + +} /* end keyring_search_instkey() */ + /*****************************************************************************/ /* * find a keyring with the specified name diff --git a/security/keys/process_keys.c b/security/keys/process_keys.c index 972e3017268..34db087bbcc 100644 --- a/security/keys/process_keys.c +++ b/security/keys/process_keys.c @@ -165,7 +165,7 @@ int install_thread_keyring(struct task_struct *tsk) /* * make sure a process keyring is installed */ -static int install_process_keyring(struct task_struct *tsk) +int install_process_keyring(struct task_struct *tsk) { unsigned long flags; struct key *keyring; @@ -376,12 +376,13 @@ void key_fsgid_changed(struct task_struct *tsk) * - we return -EAGAIN if we didn't find any matching key * - we return -ENOKEY if we found only negative matching keys */ -struct key *search_process_keyrings_aux(struct key_type *type, - const void *description, - key_match_func_t match) +struct key *search_process_keyrings(struct key_type *type, + const void *description, + key_match_func_t match, + struct task_struct *context) { - struct task_struct *tsk = current; - struct key *key, *ret, *err; + struct request_key_auth *rka; + struct key *key, *ret, *err, *instkey; /* we want to return -EAGAIN or -ENOKEY if any of the keyrings were * searchable, but we failed to find a key or we found a negative key; @@ -395,9 +396,9 @@ struct key *search_process_keyrings_aux(struct key_type *type, err = ERR_PTR(-EAGAIN); /* search the thread keyring first */ - if (tsk->thread_keyring) { - key = keyring_search_aux(tsk->thread_keyring, type, - description, match); + if (context->thread_keyring) { + key = keyring_search_aux(context->thread_keyring, + context, type, description, match); if (!IS_ERR(key)) goto found; @@ -415,9 +416,9 @@ struct key *search_process_keyrings_aux(struct key_type *type, } /* search the process keyring second */ - if (tsk->signal->process_keyring) { - key = keyring_search_aux(tsk->signal->process_keyring, - type, description, match); + if (context->signal->process_keyring) { + key = keyring_search_aux(context->signal->process_keyring, + context, type, description, match); if (!IS_ERR(key)) goto found; @@ -434,53 +435,93 @@ struct key *search_process_keyrings_aux(struct key_type *type, } } - /* search the session keyring last */ - if (tsk->signal->session_keyring) { + /* search the session keyring */ + if (context->signal->session_keyring) { rcu_read_lock(); key = keyring_search_aux( - rcu_dereference(tsk->signal->session_keyring), - type, description, match); + rcu_dereference(context->signal->session_keyring), + context, type, description, match); rcu_read_unlock(); + + if (!IS_ERR(key)) + goto found; + + switch (PTR_ERR(key)) { + case -EAGAIN: /* no key */ + if (ret) + break; + case -ENOKEY: /* negative key */ + ret = key; + break; + default: + err = key; + break; + } + + /* if this process has a session keyring and that has an + * instantiation authorisation key in the bottom level, then we + * also search the keyrings of the process mentioned there */ + if (context != current) + goto no_key; + + rcu_read_lock(); + instkey = __keyring_search_one( + rcu_dereference(context->signal->session_keyring), + &key_type_request_key_auth, NULL, 0); + rcu_read_unlock(); + + if (IS_ERR(instkey)) + goto no_key; + + rka = instkey->payload.data; + + key = search_process_keyrings(type, description, match, + rka->context); + key_put(instkey); + + if (!IS_ERR(key)) + goto found; + + switch (PTR_ERR(key)) { + case -EAGAIN: /* no key */ + if (ret) + break; + case -ENOKEY: /* negative key */ + ret = key; + break; + default: + err = key; + break; + } } + /* or search the user-session keyring */ else { - key = keyring_search_aux(tsk->user->session_keyring, - type, description, match); - } - - if (!IS_ERR(key)) - goto found; + key = keyring_search_aux(context->user->session_keyring, + context, type, description, match); + if (!IS_ERR(key)) + goto found; - switch (PTR_ERR(key)) { - case -EAGAIN: /* no key */ - if (ret) + switch (PTR_ERR(key)) { + case -EAGAIN: /* no key */ + if (ret) + break; + case -ENOKEY: /* negative key */ + ret = key; break; - case -ENOKEY: /* negative key */ - ret = key; - break; - default: - err = key; - break; + default: + err = key; + break; + } } + +no_key: /* no key - decide on the error we're going to go for */ key = ret ? ret : err; - found: +found: return key; -} /* end search_process_keyrings_aux() */ - -/*****************************************************************************/ -/* - * search the process keyrings for the first matching key - * - we return -EAGAIN if we didn't find any matching key - * - we return -ENOKEY if we found only negative matching keys - */ -struct key *search_process_keyrings(struct key_type *type, - const char *description) -{ - return search_process_keyrings_aux(type, description, type->match); - } /* end search_process_keyrings() */ /*****************************************************************************/ @@ -489,72 +530,73 @@ struct key *search_process_keyrings(struct key_type *type, * - don't create special keyrings unless so requested * - partially constructed keys aren't found unless requested */ -struct key *lookup_user_key(key_serial_t id, int create, int partial, - key_perm_t perm) +struct key *lookup_user_key(struct task_struct *context, key_serial_t id, + int create, int partial, key_perm_t perm) { - struct task_struct *tsk = current; - unsigned long flags; struct key *key; int ret; + if (!context) + context = current; + key = ERR_PTR(-ENOKEY); switch (id) { case KEY_SPEC_THREAD_KEYRING: - if (!tsk->thread_keyring) { + if (!context->thread_keyring) { if (!create) goto error; - ret = install_thread_keyring(tsk); + ret = install_thread_keyring(context); if (ret < 0) { key = ERR_PTR(ret); goto error; } } - key = tsk->thread_keyring; + key = context->thread_keyring; atomic_inc(&key->usage); break; case KEY_SPEC_PROCESS_KEYRING: - if (!tsk->signal->process_keyring) { + if (!context->signal->process_keyring) { if (!create) goto error; - ret = install_process_keyring(tsk); + ret = install_process_keyring(context); if (ret < 0) { key = ERR_PTR(ret); goto error; } } - key = tsk->signal->process_keyring; + key = context->signal->process_keyring; atomic_inc(&key->usage); break; case KEY_SPEC_SESSION_KEYRING: - if (!tsk->signal->session_keyring) { + if (!context->signal->session_keyring) { /* always install a session keyring upon access if one * doesn't exist yet */ ret = install_session_keyring( - tsk, tsk->user->session_keyring); + context, context->user->session_keyring); if (ret < 0) goto error; } - spin_lock_irqsave(&tsk->sighand->siglock, flags); - key = tsk->signal->session_keyring; + rcu_read_lock(); + key = rcu_dereference(context->signal->session_keyring); atomic_inc(&key->usage); - spin_unlock_irqrestore(&tsk->sighand->siglock, flags); + rcu_read_unlock(); break; case KEY_SPEC_USER_KEYRING: - key = tsk->user->uid_keyring; + key = context->user->uid_keyring; atomic_inc(&key->usage); break; case KEY_SPEC_USER_SESSION_KEYRING: - key = tsk->user->session_keyring; + key = context->user->session_keyring; atomic_inc(&key->usage); break; @@ -574,7 +616,7 @@ struct key *lookup_user_key(key_serial_t id, int create, int partial, break; } - /* check the status and permissions */ + /* check the status */ if (perm) { ret = key_validate(key); if (ret < 0) @@ -585,8 +627,10 @@ struct key *lookup_user_key(key_serial_t id, int create, int partial, if (!partial && !test_bit(KEY_FLAG_INSTANTIATED, &key->flags)) goto invalid_key; + /* check the permissions */ ret = -EACCES; - if (!key_permission(key, perm)) + + if (!key_task_permission(key, context, perm)) goto invalid_key; error: @@ -609,7 +653,6 @@ struct key *lookup_user_key(key_serial_t id, int create, int partial, long join_session_keyring(const char *name) { struct task_struct *tsk = current; - unsigned long flags; struct key *keyring; long ret; @@ -619,9 +662,9 @@ long join_session_keyring(const char *name) if (ret < 0) goto error; - spin_lock_irqsave(&tsk->sighand->siglock, flags); - ret = tsk->signal->session_keyring->serial; - spin_unlock_irqrestore(&tsk->sighand->siglock, flags); + rcu_read_lock(); + ret = rcu_dereference(tsk->signal->session_keyring)->serial; + rcu_read_unlock(); goto error; } diff --git a/security/keys/request_key.c b/security/keys/request_key.c index 54aa7b70e63..dfcd983af1f 100644 --- a/security/keys/request_key.c +++ b/security/keys/request_key.c @@ -1,6 +1,6 @@ /* request_key.c: request a key from userspace * - * Copyright (C) 2004 Red Hat, Inc. All Rights Reserved. + * Copyright (C) 2004-5 Red Hat, Inc. All Rights Reserved. * Written by David Howells (dhowells@redhat.com) * * This program is free software; you can redistribute it and/or @@ -13,6 +13,7 @@ #include #include #include +#include #include "internal.h" struct key_construction { @@ -27,18 +28,26 @@ DECLARE_WAIT_QUEUE_HEAD(request_key_conswq); /* * request userspace finish the construction of a key * - execute "/sbin/request-key " - * - if callout_info is an empty string, it'll be rendered as a "-" instead */ static int call_request_key(struct key *key, const char *op, const char *callout_info) { struct task_struct *tsk = current; - unsigned long flags; key_serial_t prkey, sskey; + struct key *session_keyring, *rkakey; char *argv[10], *envp[3], uid_str[12], gid_str[12]; char key_str[12], keyring_str[3][12]; - int i; + int ret, i; + + kenter("{%d},%s,%s", key->serial, op, callout_info); + + /* generate a new session keyring with an auth key in it */ + session_keyring = request_key_auth_new(key, &rkakey); + if (IS_ERR(session_keyring)) { + ret = PTR_ERR(session_keyring); + goto error; + } /* record the UID and GID */ sprintf(uid_str, "%d", current->fsuid); @@ -55,17 +64,17 @@ static int call_request_key(struct key *key, if (tsk->signal->process_keyring) prkey = tsk->signal->process_keyring->serial; - sskey = 0; - spin_lock_irqsave(&tsk->sighand->siglock, flags); - if (tsk->signal->session_keyring) - sskey = tsk->signal->session_keyring->serial; - spin_unlock_irqrestore(&tsk->sighand->siglock, flags); - + sprintf(keyring_str[1], "%d", prkey); - if (!sskey) + if (tsk->signal->session_keyring) { + rcu_read_lock(); + sskey = rcu_dereference(tsk->signal->session_keyring)->serial; + rcu_read_unlock(); + } + else { sskey = tsk->user->session_keyring->serial; + } - sprintf(keyring_str[1], "%d", prkey); sprintf(keyring_str[2], "%d", sskey); /* set up a minimal environment */ @@ -84,11 +93,20 @@ static int call_request_key(struct key *key, argv[i++] = keyring_str[0]; argv[i++] = keyring_str[1]; argv[i++] = keyring_str[2]; - argv[i++] = callout_info[0] ? (char *) callout_info : "-"; + argv[i++] = (char *) callout_info; argv[i] = NULL; /* do it */ - return call_usermodehelper_keys(argv[0], argv, envp, NULL, 1); + ret = call_usermodehelper_keys(argv[0], argv, envp, session_keyring, 1); + + /* dispose of the special keys */ + key_revoke(rkakey); + key_put(rkakey); + key_put(session_keyring); + + error: + kleave(" = %d", ret); + return ret; } /* end call_request_key() */ @@ -107,6 +125,8 @@ static struct key *__request_key_construction(struct key_type *type, struct key *key; int ret, negated; + kenter("%s,%s,%s", type->name, description, callout_info); + /* create a key and add it to the queue */ key = key_alloc(type, description, current->fsuid, current->fsgid, KEY_USR_ALL, 0); @@ -143,6 +163,7 @@ static struct key *__request_key_construction(struct key_type *type, } out: + kleave(" = %p", key); return key; request_failed: @@ -216,6 +237,9 @@ static struct key *request_key_construction(struct key_type *type, DECLARE_WAITQUEUE(myself, current); + kenter("%s,%s,{%d},%s", + type->name, description, user->uid, callout_info); + /* see if there's such a key under construction already */ down_write(&key_construction_sem); @@ -232,6 +256,7 @@ static struct key *request_key_construction(struct key_type *type, /* see about getting userspace to construct the key */ key = __request_key_construction(type, description, callout_info); error: + kleave(" = %p", key); return key; /* someone else has the same key under construction @@ -245,9 +270,11 @@ static struct key *request_key_construction(struct key_type *type, add_wait_queue(&request_key_conswq, &myself); for (;;) { - set_current_state(TASK_UNINTERRUPTIBLE); + set_current_state(TASK_INTERRUPTIBLE); if (!test_bit(KEY_FLAG_USER_CONSTRUCT, &ckey->flags)) break; + if (signal_pending(current)) + break; schedule(); } @@ -265,23 +292,85 @@ static struct key *request_key_construction(struct key_type *type, } /* end request_key_construction() */ +/*****************************************************************************/ +/* + * link a freshly minted key to an appropriate destination keyring + */ +static void request_key_link(struct key *key, struct key *dest_keyring) +{ + struct task_struct *tsk = current; + struct key *drop = NULL; + + kenter("{%d},%p", key->serial, dest_keyring); + + /* find the appropriate keyring */ + if (!dest_keyring) { + switch (tsk->jit_keyring) { + case KEY_REQKEY_DEFL_DEFAULT: + case KEY_REQKEY_DEFL_THREAD_KEYRING: + dest_keyring = tsk->thread_keyring; + if (dest_keyring) + break; + + case KEY_REQKEY_DEFL_PROCESS_KEYRING: + dest_keyring = tsk->signal->process_keyring; + if (dest_keyring) + break; + + case KEY_REQKEY_DEFL_SESSION_KEYRING: + rcu_read_lock(); + dest_keyring = key_get( + rcu_dereference(tsk->signal->session_keyring)); + rcu_read_unlock(); + drop = dest_keyring; + + if (dest_keyring) + break; + + case KEY_REQKEY_DEFL_USER_SESSION_KEYRING: + dest_keyring = current->user->session_keyring; + break; + + case KEY_REQKEY_DEFL_USER_KEYRING: + dest_keyring = current->user->uid_keyring; + break; + + case KEY_REQKEY_DEFL_GROUP_KEYRING: + default: + BUG(); + } + } + + /* and attach the key to it */ + key_link(dest_keyring, key); + + key_put(drop); + + kleave(""); + +} /* end request_key_link() */ + /*****************************************************************************/ /* * request a key * - search the process's keyrings * - check the list of keys being created or updated - * - call out to userspace for a key if requested (supplementary info can be - * passed) + * - call out to userspace for a key if supplementary info was provided + * - cache the key in an appropriate keyring */ -struct key *request_key(struct key_type *type, - const char *description, - const char *callout_info) +struct key *request_key_and_link(struct key_type *type, + const char *description, + const char *callout_info, + struct key *dest_keyring) { struct key_user *user; struct key *key; + kenter("%s,%s,%s,%p", + type->name, description, callout_info, dest_keyring); + /* search all the process keyrings for a key */ - key = search_process_keyrings_aux(type, description, type->match); + key = search_process_keyrings(type, description, type->match, current); if (PTR_ERR(key) == -EAGAIN) { /* the search failed, but the keyrings were searchable, so we @@ -292,12 +381,13 @@ struct key *request_key(struct key_type *type, /* - get hold of the user's construction queue */ user = key_user_lookup(current->fsuid); - if (!user) { - key = ERR_PTR(-ENOMEM); - goto error; - } + if (!user) + goto nomem; + + do { + if (signal_pending(current)) + goto interrupted; - for (;;) { /* ask userspace (returns NULL if it waited on a key * being constructed) */ key = request_key_construction(type, description, @@ -307,18 +397,46 @@ struct key *request_key(struct key_type *type, /* someone else made the key we want, so we need to * search again as it might now be available to us */ - key = search_process_keyrings_aux(type, description, - type->match); - if (PTR_ERR(key) != -EAGAIN) - break; - } + key = search_process_keyrings(type, description, + type->match, current); + + } while (PTR_ERR(key) == -EAGAIN); key_user_put(user); + + /* link the new key into the appropriate keyring */ + if (!PTR_ERR(key)) + request_key_link(key, dest_keyring); } - error: +error: + kleave(" = %p", key); return key; +nomem: + key = ERR_PTR(-ENOMEM); + goto error; + +interrupted: + key_user_put(user); + key = ERR_PTR(-EINTR); + goto error; + +} /* end request_key_and_link() */ + +/*****************************************************************************/ +/* + * request a key + * - search the process's keyrings + * - check the list of keys being created or updated + * - call out to userspace for a key if supplementary info was provided + */ +struct key *request_key(struct key_type *type, + const char *description, + const char *callout_info) +{ + return request_key_and_link(type, description, callout_info, NULL); + } /* end request_key() */ EXPORT_SYMBOL(request_key); diff --git a/security/keys/request_key_auth.c b/security/keys/request_key_auth.c new file mode 100644 index 00000000000..f2226463222 --- /dev/null +++ b/security/keys/request_key_auth.c @@ -0,0 +1,180 @@ +/* request_key_auth.c: request key authorisation controlling key def + * + * Copyright (C) 2005 Red Hat, Inc. All Rights Reserved. + * Written by David Howells (dhowells@redhat.com) + * + * 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. + */ + +#include +#include +#include +#include +#include "internal.h" + +static int request_key_auth_instantiate(struct key *, const void *, size_t); +static void request_key_auth_describe(const struct key *, struct seq_file *); +static void request_key_auth_destroy(struct key *); + +/* + * the request-key authorisation key type definition + */ +struct key_type key_type_request_key_auth = { + .name = ".request_key_auth", + .def_datalen = sizeof(struct request_key_auth), + .instantiate = request_key_auth_instantiate, + .describe = request_key_auth_describe, + .destroy = request_key_auth_destroy, +}; + +/*****************************************************************************/ +/* + * instantiate a request-key authorisation record + */ +static int request_key_auth_instantiate(struct key *key, + const void *data, + size_t datalen) +{ + struct request_key_auth *rka, *irka; + struct key *instkey; + int ret; + + ret = -ENOMEM; + rka = kmalloc(sizeof(*rka), GFP_KERNEL); + if (rka) { + /* see if the calling process is already servicing the key + * request of another process */ + instkey = key_get_instantiation_authkey(0); + if (!IS_ERR(instkey)) { + /* it is - use that instantiation context here too */ + irka = instkey->payload.data; + rka->context = irka->context; + rka->pid = irka->pid; + key_put(instkey); + } + else { + /* it isn't - use this process as the context */ + rka->context = current; + rka->pid = current->pid; + } + + rka->target_key = key_get((struct key *) data); + key->payload.data = rka; + ret = 0; + } + + return ret; + +} /* end request_key_auth_instantiate() */ + +/*****************************************************************************/ +/* + * + */ +static void request_key_auth_describe(const struct key *key, + struct seq_file *m) +{ + struct request_key_auth *rka = key->payload.data; + + seq_puts(m, "key:"); + seq_puts(m, key->description); + seq_printf(m, " pid:%d", rka->pid); + +} /* end request_key_auth_describe() */ + +/*****************************************************************************/ +/* + * destroy an instantiation authorisation token key + */ +static void request_key_auth_destroy(struct key *key) +{ + struct request_key_auth *rka = key->payload.data; + + kenter("{%d}", key->serial); + + key_put(rka->target_key); + +} /* end request_key_auth_destroy() */ + +/*****************************************************************************/ +/* + * create a session keyring to be for the invokation of /sbin/request-key and + * stick an authorisation token in it + */ +struct key *request_key_auth_new(struct key *target, struct key **_rkakey) +{ + struct key *keyring, *rkakey = NULL; + char desc[20]; + int ret; + + kenter("%d,", target->serial); + + /* allocate a new session keyring */ + sprintf(desc, "_req.%u", target->serial); + + keyring = keyring_alloc(desc, current->fsuid, current->fsgid, 1, NULL); + if (IS_ERR(keyring)) { + kleave("= %ld", PTR_ERR(keyring)); + return keyring; + } + + /* allocate the auth key */ + sprintf(desc, "%x", target->serial); + + rkakey = key_alloc(&key_type_request_key_auth, desc, + current->fsuid, current->fsgid, + KEY_USR_VIEW, 1); + if (IS_ERR(rkakey)) { + key_put(keyring); + kleave("= %ld", PTR_ERR(rkakey)); + return rkakey; + } + + /* construct and attach to the keyring */ + ret = key_instantiate_and_link(rkakey, target, 0, keyring, NULL); + if (ret < 0) { + key_revoke(rkakey); + key_put(rkakey); + key_put(keyring); + kleave("= %d", ret); + return ERR_PTR(ret); + } + + *_rkakey = rkakey; + kleave(" = {%d} ({%d})", keyring->serial, rkakey->serial); + return keyring; + +} /* end request_key_auth_new() */ + +/*****************************************************************************/ +/* + * get the authorisation key for instantiation of a specific key if attached to + * the current process's keyrings + * - this key is inserted into a keyring and that is set as /sbin/request-key's + * session keyring + * - a target_id of zero specifies any valid token + */ +struct key *key_get_instantiation_authkey(key_serial_t target_id) +{ + struct task_struct *tsk = current; + struct key *instkey; + + /* we must have our own personal session keyring */ + if (!tsk->signal->session_keyring) + return ERR_PTR(-EACCES); + + /* and it must contain a suitable request authorisation key + * - lock RCU against session keyring changing + */ + rcu_read_lock(); + + instkey = keyring_search_instkey( + rcu_dereference(tsk->signal->session_keyring), target_id); + + rcu_read_unlock(); + return instkey; + +} /* end key_get_instantiation_authkey() */ -- cgit v1.2.3-70-g09d2 From c988d2b2845495373f666a381d354a7f80981d62 Mon Sep 17 00:00:00 2001 From: Matt Domsch Date: Thu, 23 Jun 2005 22:05:15 -0700 Subject: [PATCH] modules: add version and srcversion to sysfs This patch adds version and srcversion files to /sys/module/${modulename} containing the version and srcversion fields of the module's modinfo section (if present). /sys/module/e1000 |-- srcversion `-- version This patch differs slightly from the version posted in January, as it now uses the new kstrdup() call in -mm. Why put this in sysfs? a) Tools like DKMS, which deal with changing out individual kernel modules without replacing the whole kernel, can behave smarter if they can tell the version of a given module. The autoinstaller feature, for example, which determines if your system has a "good" version of a driver (i.e. if the one provided by DKMS has a newer verson than that provided by the kernel package installed), and to automatically compile and install a newer version if DKMS has it but your kernel doesn't yet have that version. b) Because sysadmins manually, or with tools like DKMS, can switch out modules on the file system, you can't count on 'modinfo foo.ko', which looks at /lib/modules/${kernelver}/... actually matching what is loaded into the kernel already. Hence asking sysfs for this. c) as the unbind-driver-from-device work takes shape, it will be possible to rebind a driver that's built-in (no .ko to modinfo for the version) to a newly loaded module. sysfs will have the currently-built-in version info, for comparison. d) tech support scripts can then easily grab the version info for what's running presently - a question I get often. There has been renewed interest in this patch on linux-scsi by driver authors. As the idea originated from GregKH, I leave his Signed-off-by: intact, though the implementation is nearly completely new. Compiled and run on x86 and x86_64. From: Matthew Dobson build fix From: Thierry Vignaud build fix From: Matthew Dobson warning fix Signed-off-by: Greg Kroah-Hartman Signed-off-by: Matt Domsch Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/module.h | 5 +++ kernel/module.c | 95 ++++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 100 insertions(+) (limited to 'kernel') diff --git a/include/linux/module.h b/include/linux/module.h index 0e432a0f4ae..f05372b7fe7 100644 --- a/include/linux/module.h +++ b/include/linux/module.h @@ -51,6 +51,9 @@ struct module_attribute { ssize_t (*show)(struct module_attribute *, struct module *, char *); ssize_t (*store)(struct module_attribute *, struct module *, const char *, size_t count); + void (*setup)(struct module *, const char *); + int (*test)(struct module *); + void (*free)(struct module *); }; struct module_kobject @@ -239,6 +242,8 @@ struct module /* Sysfs stuff. */ struct module_kobject mkobj; struct module_param_attrs *param_attrs; + const char *version; + const char *srcversion; /* Exported symbols */ const struct kernel_symbol *syms; diff --git a/kernel/module.c b/kernel/module.c index a566745dde6..0494c89a0d2 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -35,6 +35,7 @@ #include #include #include +#include #include #include #include @@ -370,6 +371,43 @@ static inline void percpu_modcopy(void *pcpudst, const void *src, #endif /* CONFIG_SMP */ #ifdef CONFIG_MODULE_UNLOAD +#define MODINFO_ATTR(field) \ +static void setup_modinfo_##field(struct module *mod, const char *s) \ +{ \ + mod->field = kstrdup(s, GFP_KERNEL); \ +} \ +static ssize_t show_modinfo_##field(struct module_attribute *mattr, \ + struct module *mod, char *buffer) \ +{ \ + return sprintf(buffer, "%s\n", mod->field); \ +} \ +static int modinfo_##field##_exists(struct module *mod) \ +{ \ + return mod->field != NULL; \ +} \ +static void free_modinfo_##field(struct module *mod) \ +{ \ + kfree(mod->field); \ + mod->field = NULL; \ +} \ +static struct module_attribute modinfo_##field = { \ + .attr = { .name = __stringify(field), .mode = 0444, \ + .owner = THIS_MODULE }, \ + .show = show_modinfo_##field, \ + .setup = setup_modinfo_##field, \ + .test = modinfo_##field##_exists, \ + .free = free_modinfo_##field, \ +}; + +MODINFO_ATTR(version); +MODINFO_ATTR(srcversion); + +static struct module_attribute *modinfo_attrs[] = { + &modinfo_version, + &modinfo_srcversion, + NULL, +}; + /* Init the unload section of the module. */ static void module_unload_init(struct module *mod) { @@ -1031,6 +1069,32 @@ static void module_remove_refcnt_attr(struct module *mod) } #endif +#ifdef CONFIG_MODULE_UNLOAD +static int module_add_modinfo_attrs(struct module *mod) +{ + struct module_attribute *attr; + int error = 0; + int i; + + for (i = 0; (attr = modinfo_attrs[i]) && !error; i++) { + if (!attr->test || + (attr->test && attr->test(mod))) + error = sysfs_create_file(&mod->mkobj.kobj,&attr->attr); + } + return error; +} + +static void module_remove_modinfo_attrs(struct module *mod) +{ + struct module_attribute *attr; + int i; + + for (i = 0; (attr = modinfo_attrs[i]); i++) { + sysfs_remove_file(&mod->mkobj.kobj,&attr->attr); + attr->free(mod); + } +} +#endif static int mod_sysfs_setup(struct module *mod, struct kernel_param *kparam, @@ -1056,6 +1120,12 @@ static int mod_sysfs_setup(struct module *mod, if (err) goto out_unreg; +#ifdef CONFIG_MODULE_UNLOAD + err = module_add_modinfo_attrs(mod); + if (err) + goto out_unreg; +#endif + return 0; out_unreg: @@ -1066,6 +1136,9 @@ out: static void mod_kobject_remove(struct module *mod) { +#ifdef CONFIG_MODULE_UNLOAD + module_remove_modinfo_attrs(mod); +#endif module_remove_refcnt_attr(mod); module_param_sysfs_remove(mod); @@ -1311,6 +1384,23 @@ static char *get_modinfo(Elf_Shdr *sechdrs, return NULL; } +#ifdef CONFIG_MODULE_UNLOAD +static void setup_modinfo(struct module *mod, Elf_Shdr *sechdrs, + unsigned int infoindex) +{ + struct module_attribute *attr; + int i; + + for (i = 0; (attr = modinfo_attrs[i]); i++) { + if (attr->setup) + attr->setup(mod, + get_modinfo(sechdrs, + infoindex, + attr->attr.name)); + } +} +#endif + #ifdef CONFIG_KALLSYMS int is_exported(const char *name, const struct module *mod) { @@ -1615,6 +1705,11 @@ static struct module *load_module(void __user *umod, /* Set up license info based on the info section */ set_license(mod, get_modinfo(sechdrs, infoindex, "license")); +#ifdef CONFIG_MODULE_UNLOAD + /* Set up MODINFO_ATTR fields */ + setup_modinfo(mod, sechdrs, infoindex); +#endif + /* Fix up syms, so that st_value is a pointer to location. */ err = simplify_symbols(sechdrs, symindex, strtab, versindex, pcpuindex, mod); -- cgit v1.2.3-70-g09d2 From 52c1da39534fb382c061de58b65f678ad74b59f5 Mon Sep 17 00:00:00 2001 From: Adrian Bunk Date: Thu, 23 Jun 2005 22:05:33 -0700 Subject: [PATCH] make various thing static Another rollup of patches which give various symbols static scope Signed-off-by: Adrian Bunk Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/media/common/saa7146_fops.c | 2 +- drivers/media/video/tvaudio.c | 22 +++++++++++----------- drivers/scsi/hosts.c | 2 +- drivers/scsi/scsi.c | 6 ++++-- drivers/scsi/scsi_debug.c | 2 +- drivers/scsi/scsi_lib.c | 2 +- drivers/scsi/scsi_priv.h | 4 ---- drivers/scsi/scsi_sysfs.c | 4 ++-- fs/namespace.c | 2 +- fs/reiserfs/stree.c | 2 +- include/linux/irq.h | 1 - include/linux/namespace.h | 1 - include/net/sctp/sm.h | 6 ------ kernel/irq/spurious.c | 2 +- kernel/module.c | 2 +- kernel/power/swsusp.c | 2 +- net/sctp/sm_statefuns.c | 16 ++++++++++++++-- 17 files changed, 40 insertions(+), 38 deletions(-) (limited to 'kernel') diff --git a/drivers/media/common/saa7146_fops.c b/drivers/media/common/saa7146_fops.c index cb826c9adfe..c04fd11526e 100644 --- a/drivers/media/common/saa7146_fops.c +++ b/drivers/media/common/saa7146_fops.c @@ -403,7 +403,7 @@ static struct file_operations video_fops = .llseek = no_llseek, }; -void vv_callback(struct saa7146_dev *dev, unsigned long status) +static void vv_callback(struct saa7146_dev *dev, unsigned long status) { u32 isr = status; diff --git a/drivers/media/video/tvaudio.c b/drivers/media/video/tvaudio.c index 5430b25b910..9a493bea76d 100644 --- a/drivers/media/video/tvaudio.c +++ b/drivers/media/video/tvaudio.c @@ -1236,17 +1236,17 @@ static int ta8874z_checkit(struct CHIPSTATE *chip) /* audio chip descriptions - struct CHIPDESC */ /* insmod options to enable/disable individual audio chips */ -int tda8425 = 1; -int tda9840 = 1; -int tda9850 = 1; -int tda9855 = 1; -int tda9873 = 1; -int tda9874a = 1; -int tea6300 = 0; // address clash with msp34xx -int tea6320 = 0; // address clash with msp34xx -int tea6420 = 1; -int pic16c54 = 1; -int ta8874z = 0; // address clash with tda9840 +static int tda8425 = 1; +static int tda9840 = 1; +static int tda9850 = 1; +static int tda9855 = 1; +static int tda9873 = 1; +static int tda9874a = 1; +static int tea6300 = 0; // address clash with msp34xx +static int tea6320 = 0; // address clash with msp34xx +static int tea6420 = 1; +static int pic16c54 = 1; +static int ta8874z = 0; // address clash with tda9840 module_param(tda8425, int, 0444); module_param(tda9840, int, 0444); diff --git a/drivers/scsi/hosts.c b/drivers/scsi/hosts.c index ba347576d99..d7a38b6713f 100644 --- a/drivers/scsi/hosts.c +++ b/drivers/scsi/hosts.c @@ -56,7 +56,7 @@ static struct class shost_class = { * @shost: pointer to struct Scsi_Host * recovery: recovery requested to run. **/ -void scsi_host_cancel(struct Scsi_Host *shost, int recovery) +static void scsi_host_cancel(struct Scsi_Host *shost, int recovery) { struct scsi_device *sdev; diff --git a/drivers/scsi/scsi.c b/drivers/scsi/scsi.c index 5578ae9a9e4..1cb5f7d4f27 100644 --- a/drivers/scsi/scsi.c +++ b/drivers/scsi/scsi.c @@ -68,6 +68,8 @@ #include "scsi_priv.h" #include "scsi_logging.h" +static void scsi_done(struct scsi_cmnd *cmd); +static int scsi_retry_command(struct scsi_cmnd *cmd); /* * Definitions and constants. @@ -741,7 +743,7 @@ static DEFINE_PER_CPU(struct list_head, scsi_done_q); * * This function is interrupt context safe. */ -void scsi_done(struct scsi_cmnd *cmd) +static void scsi_done(struct scsi_cmnd *cmd) { /* * We don't have to worry about this one timing out any more. @@ -836,7 +838,7 @@ static void scsi_softirq(struct softirq_action *h) * level drivers should not become re-entrant as a result of * this. */ -int scsi_retry_command(struct scsi_cmnd *cmd) +static int scsi_retry_command(struct scsi_cmnd *cmd) { /* * Restore the SCSI command state. diff --git a/drivers/scsi/scsi_debug.c b/drivers/scsi/scsi_debug.c index e0208886b45..322b5a41a36 100644 --- a/drivers/scsi/scsi_debug.c +++ b/drivers/scsi/scsi_debug.c @@ -1783,7 +1783,7 @@ static void __exit scsi_debug_exit(void) device_initcall(scsi_debug_init); module_exit(scsi_debug_exit); -void pseudo_0_release(struct device * dev) +static void pseudo_0_release(struct device * dev) { if (SCSI_DEBUG_OPT_NOISE & scsi_debug_opts) printk(KERN_INFO "scsi_debug: pseudo_0_release() called\n"); diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index 9f996499fa9..621dee8b8cb 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -44,7 +44,7 @@ struct scsi_host_sg_pool { #endif #define SP(x) { x, "sgpool-" #x } -struct scsi_host_sg_pool scsi_sg_pools[] = { +static struct scsi_host_sg_pool scsi_sg_pools[] = { SP(8), SP(16), SP(32), diff --git a/drivers/scsi/scsi_priv.h b/drivers/scsi/scsi_priv.h index c01580df447..96d4f745975 100644 --- a/drivers/scsi/scsi_priv.h +++ b/drivers/scsi/scsi_priv.h @@ -61,8 +61,6 @@ extern void scsi_exit_hosts(void); extern int scsi_dispatch_cmd(struct scsi_cmnd *cmd); extern int scsi_setup_command_freelist(struct Scsi_Host *shost); extern void scsi_destroy_command_freelist(struct Scsi_Host *shost); -extern void scsi_done(struct scsi_cmnd *cmd); -extern int scsi_retry_command(struct scsi_cmnd *cmd); extern int scsi_insert_special_req(struct scsi_request *sreq, int); extern void scsi_init_cmd_from_req(struct scsi_cmnd *cmd, struct scsi_request *sreq); @@ -136,7 +134,6 @@ extern void scsi_exit_sysctl(void); #endif /* CONFIG_SYSCTL */ /* scsi_sysfs.c */ -extern void scsi_device_dev_release(struct device *); extern int scsi_sysfs_add_sdev(struct scsi_device *); extern int scsi_sysfs_add_host(struct Scsi_Host *); extern int scsi_sysfs_register(void); @@ -145,7 +142,6 @@ extern void scsi_sysfs_device_initialize(struct scsi_device *); extern int scsi_sysfs_target_initialize(struct scsi_device *); extern struct scsi_transport_template blank_transport_template; -extern struct class sdev_class; extern struct bus_type scsi_bus_type; /* diff --git a/drivers/scsi/scsi_sysfs.c b/drivers/scsi/scsi_sysfs.c index 93b41100a6d..beed7fbe1cb 100644 --- a/drivers/scsi/scsi_sysfs.c +++ b/drivers/scsi/scsi_sysfs.c @@ -150,7 +150,7 @@ static void scsi_device_cls_release(struct class_device *class_dev) put_device(&sdev->sdev_gendev); } -void scsi_device_dev_release(struct device *dev) +static void scsi_device_dev_release(struct device *dev) { struct scsi_device *sdev; struct device *parent; @@ -185,7 +185,7 @@ void scsi_device_dev_release(struct device *dev) put_device(parent); } -struct class sdev_class = { +static struct class sdev_class = { .name = "scsi_device", .release = scsi_device_cls_release, }; diff --git a/fs/namespace.c b/fs/namespace.c index 3b93e5d750e..208c079e9fd 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -337,7 +337,7 @@ int may_umount(struct vfsmount *mnt) EXPORT_SYMBOL(may_umount); -void umount_tree(struct vfsmount *mnt) +static void umount_tree(struct vfsmount *mnt) { struct vfsmount *p; LIST_HEAD(kill); diff --git a/fs/reiserfs/stree.c b/fs/reiserfs/stree.c index c47f8fd31a2..63158491e15 100644 --- a/fs/reiserfs/stree.c +++ b/fs/reiserfs/stree.c @@ -223,7 +223,7 @@ extern struct tree_balance * cur_tb; const struct reiserfs_key MIN_KEY = {0, 0, {{0, 0},}}; /* Maximal possible key. It is never in the tree. */ -const struct reiserfs_key MAX_KEY = { +static const struct reiserfs_key MAX_KEY = { __constant_cpu_to_le32(0xffffffff), __constant_cpu_to_le32(0xffffffff), {{__constant_cpu_to_le32(0xffffffff), diff --git a/include/linux/irq.h b/include/linux/irq.h index 7fc1022be9e..12277799c00 100644 --- a/include/linux/irq.h +++ b/include/linux/irq.h @@ -88,7 +88,6 @@ extern fastcall int handle_IRQ_event(unsigned int irq, struct pt_regs *regs, struct irqaction *action); extern fastcall unsigned int __do_IRQ(unsigned int irq, struct pt_regs *regs); extern void note_interrupt(unsigned int irq, irq_desc_t *desc, int action_ret); -extern void report_bad_irq(unsigned int irq, irq_desc_t *desc, int action_ret); extern int can_request_irq(unsigned int irq, unsigned long irqflags); extern void init_irq_proc(void); diff --git a/include/linux/namespace.h b/include/linux/namespace.h index 9eca1558d72..697991b69f9 100644 --- a/include/linux/namespace.h +++ b/include/linux/namespace.h @@ -12,7 +12,6 @@ struct namespace { struct rw_semaphore sem; }; -extern void umount_tree(struct vfsmount *); extern int copy_namespace(int, struct task_struct *); extern void __put_namespace(struct namespace *namespace); diff --git a/include/net/sctp/sm.h b/include/net/sctp/sm.h index a53e08a45e3..88d9fe5975d 100644 --- a/include/net/sctp/sm.h +++ b/include/net/sctp/sm.h @@ -131,7 +131,6 @@ sctp_state_fn_t sctp_sf_do_ecne; sctp_state_fn_t sctp_sf_ootb; sctp_state_fn_t sctp_sf_pdiscard; sctp_state_fn_t sctp_sf_violation; -sctp_state_fn_t sctp_sf_violation_chunklen; sctp_state_fn_t sctp_sf_discard_chunk; sctp_state_fn_t sctp_sf_do_5_2_1_siminit; sctp_state_fn_t sctp_sf_do_5_2_2_dupinit; @@ -259,11 +258,6 @@ struct sctp_chunk *sctp_make_fwdtsn(const struct sctp_association *asoc, void sctp_chunk_assign_tsn(struct sctp_chunk *); void sctp_chunk_assign_ssn(struct sctp_chunk *); -sctp_disposition_t sctp_stop_t1_and_abort(sctp_cmd_seq_t *commands, - __u16 error, - const struct sctp_association *asoc, - struct sctp_transport *transport); - /* Prototypes for statetable processing. */ int sctp_do_sm(sctp_event_t event_type, sctp_subtype_t subtype, diff --git a/kernel/irq/spurious.c b/kernel/irq/spurious.c index f6297c30690..ba039e827d5 100644 --- a/kernel/irq/spurious.c +++ b/kernel/irq/spurious.c @@ -45,7 +45,7 @@ __report_bad_irq(unsigned int irq, irq_desc_t *desc, irqreturn_t action_ret) } } -void report_bad_irq(unsigned int irq, irq_desc_t *desc, irqreturn_t action_ret) +static void report_bad_irq(unsigned int irq, irq_desc_t *desc, irqreturn_t action_ret) { static int count = 100; diff --git a/kernel/module.c b/kernel/module.c index 0494c89a0d2..068e271ab3a 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -730,7 +730,7 @@ static int obsparm_copy_string(const char *val, struct kernel_param *kp) return 0; } -int set_obsolete(const char *val, struct kernel_param *kp) +static int set_obsolete(const char *val, struct kernel_param *kp) { unsigned int min, max; unsigned int size, maxsize; diff --git a/kernel/power/swsusp.c b/kernel/power/swsusp.c index 90b3b68dee3..53f9f8720ee 100644 --- a/kernel/power/swsusp.c +++ b/kernel/power/swsusp.c @@ -81,7 +81,7 @@ static int nr_copy_pages_check; extern char resume_file[]; /* Local variables that should not be affected by save */ -unsigned int nr_copy_pages __nosavedata = 0; +static unsigned int nr_copy_pages __nosavedata = 0; /* Suspend pagedir is allocated before final copy, therefore it must be freed after resume diff --git a/net/sctp/sm_statefuns.c b/net/sctp/sm_statefuns.c index 058189684c7..86073df418f 100644 --- a/net/sctp/sm_statefuns.c +++ b/net/sctp/sm_statefuns.c @@ -92,6 +92,17 @@ static sctp_disposition_t sctp_sf_shut_8_4_5(const struct sctp_endpoint *ep, sctp_cmd_seq_t *commands); static struct sctp_sackhdr *sctp_sm_pull_sack(struct sctp_chunk *chunk); +static sctp_disposition_t sctp_stop_t1_and_abort(sctp_cmd_seq_t *commands, + __u16 error, + const struct sctp_association *asoc, + struct sctp_transport *transport); + +static sctp_disposition_t sctp_sf_violation_chunklen( + const struct sctp_endpoint *ep, + const struct sctp_association *asoc, + const sctp_subtype_t type, + void *arg, + sctp_cmd_seq_t *commands); /* Small helper function that checks if the chunk length * is of the appropriate length. The 'required_length' argument @@ -2328,7 +2339,7 @@ sctp_disposition_t sctp_sf_cookie_echoed_abort(const struct sctp_endpoint *ep, * * This is common code called by several sctp_sf_*_abort() functions above. */ -sctp_disposition_t sctp_stop_t1_and_abort(sctp_cmd_seq_t *commands, +static sctp_disposition_t sctp_stop_t1_and_abort(sctp_cmd_seq_t *commands, __u16 error, const struct sctp_association *asoc, struct sctp_transport *transport) @@ -3687,7 +3698,8 @@ sctp_disposition_t sctp_sf_violation(const struct sctp_endpoint *ep, * * Generate an ABORT chunk and terminate the association. */ -sctp_disposition_t sctp_sf_violation_chunklen(const struct sctp_endpoint *ep, +static sctp_disposition_t sctp_sf_violation_chunklen( + const struct sctp_endpoint *ep, const struct sctp_association *asoc, const sctp_subtype_t type, void *arg, -- cgit v1.2.3-70-g09d2