Commit 37cbd41d authored by Kirill Korotaev's avatar Kirill Korotaev Committed by Linus Torvalds

[PATCH] fixed pidhashing patch

This patch fixes strange and obscure pid implementation in current kernels:
- it removes calling of put_task_struct() from detach_pid()
  under tasklist_lock. This allows to use blocking calls
  in security_task_free() hooks (in __put_task_struct()).
- it saves some space = 5*5 ints = 100 bytes in task_struct
- it's smaller and tidy, more straigthforward and doesn't use
  any knowledge about pids using and assignment.
- it removes pid_links and pid_struct doesn't hold reference counters
  on task_struct. instead, new pid_structs and linked altogether and
  only one of them is inserted in hash_list.

Signed-off-by: Kirill Korotaev (kksx@mail.ru)
Signed-off-by: default avatarWilliam Irwin <wli@holomorphy.com>
Signed-off-by: default avatarLinus Torvalds <torvalds@osdl.org>
parent 3345ee71
......@@ -424,7 +424,6 @@ void do_tty_hangup(void *data)
struct file * cons_filp = NULL;
struct file *filp, *f = NULL;
struct task_struct *p;
struct pid *pid;
int closecount = 0, n;
if (!tty)
......@@ -495,8 +494,7 @@ void do_tty_hangup(void *data)
read_lock(&tasklist_lock);
if (tty->session > 0) {
struct list_head *l;
for_each_task_pid(tty->session, PIDTYPE_SID, p, l, pid) {
do_each_task_pid(tty->session, PIDTYPE_SID, p) {
if (p->signal->tty == tty)
p->signal->tty = NULL;
if (!p->signal->leader)
......@@ -505,7 +503,7 @@ void do_tty_hangup(void *data)
send_group_sig_info(SIGCONT, SEND_SIG_PRIV, p);
if (tty->pgrp > 0)
p->signal->tty_old_pgrp = tty->pgrp;
}
} while_each_task_pid(tty->session, PIDTYPE_SID, p);
}
read_unlock(&tasklist_lock);
......@@ -577,8 +575,6 @@ void disassociate_ctty(int on_exit)
{
struct tty_struct *tty;
struct task_struct *p;
struct list_head *l;
struct pid *pid;
int tty_pgrp = -1;
lock_kernel();
......@@ -607,8 +603,9 @@ void disassociate_ctty(int on_exit)
tty->pgrp = -1;
read_lock(&tasklist_lock);
for_each_task_pid(current->signal->session, PIDTYPE_SID, p, l, pid)
do_each_task_pid(current->signal->session, PIDTYPE_SID, p) {
p->signal->tty = NULL;
} while_each_task_pid(current->signal->session, PIDTYPE_SID, p);
read_unlock(&tasklist_lock);
unlock_kernel();
}
......@@ -1260,15 +1257,15 @@ static void release_dev(struct file * filp)
*/
if (tty_closing || o_tty_closing) {
struct task_struct *p;
struct list_head *l;
struct pid *pid;
read_lock(&tasklist_lock);
for_each_task_pid(tty->session, PIDTYPE_SID, p, l, pid)
do_each_task_pid(tty->session, PIDTYPE_SID, p) {
p->signal->tty = NULL;
} while_each_task_pid(tty->session, PIDTYPE_SID, p);
if (o_tty)
for_each_task_pid(o_tty->session, PIDTYPE_SID, p,l, pid)
do_each_task_pid(o_tty->session, PIDTYPE_SID, p) {
p->signal->tty = NULL;
} while_each_task_pid(o_tty->session, PIDTYPE_SID, p);
read_unlock(&tasklist_lock);
}
......@@ -1638,8 +1635,6 @@ static int fionbio(struct file *file, int __user *p)
static int tiocsctty(struct tty_struct *tty, int arg)
{
struct list_head *l;
struct pid *pid;
task_t *p;
if (current->signal->leader &&
......@@ -1662,8 +1657,9 @@ static int tiocsctty(struct tty_struct *tty, int arg)
*/
read_lock(&tasklist_lock);
for_each_task_pid(tty->session, PIDTYPE_SID, p, l, pid)
do_each_task_pid(tty->session, PIDTYPE_SID, p) {
p->signal->tty = NULL;
} while_each_task_pid(tty->session, PIDTYPE_SID, p);
read_unlock(&tasklist_lock);
} else
return -EPERM;
......@@ -1970,8 +1966,6 @@ static void __do_SAK(void *arg)
#else
struct tty_struct *tty = arg;
struct task_struct *p;
struct list_head *l;
struct pid *pid;
int session;
int i;
struct file *filp;
......@@ -1984,7 +1978,7 @@ static void __do_SAK(void *arg)
if (tty->driver->flush_buffer)
tty->driver->flush_buffer(tty);
read_lock(&tasklist_lock);
for_each_task_pid(session, PIDTYPE_SID, p, l, pid) {
do_each_task_pid(session, PIDTYPE_SID, p) {
if (p->signal->tty == tty || session > 0) {
printk(KERN_NOTICE "SAK: killed process %d"
" (%s): p->signal->session==tty->session\n",
......@@ -2011,7 +2005,7 @@ static void __do_SAK(void *arg)
spin_unlock(&p->files->file_lock);
}
task_unlock(p);
}
} while_each_task_pid(session, PIDTYPE_SID, p);
read_unlock(&tasklist_lock);
#endif
}
......
......@@ -497,11 +497,9 @@ void send_sigio(struct fown_struct *fown, int fd, int band)
send_sigio_to_task(p, fown, fd, band);
}
} else {
struct list_head *l;
struct pid *pidptr;
for_each_task_pid(-pid, PIDTYPE_PGID, p, l, pidptr) {
do_each_task_pid(-pid, PIDTYPE_PGID, p) {
send_sigio_to_task(p, fown, fd, band);
}
} while_each_task_pid(-pid, PIDTYPE_PGID, p);
}
read_unlock(&tasklist_lock);
out_unlock_fown:
......@@ -534,11 +532,9 @@ int send_sigurg(struct fown_struct *fown)
send_sigurg_to_task(p, fown);
}
} else {
struct list_head *l;
struct pid *pidptr;
for_each_task_pid(-pid, PIDTYPE_PGID, p, l, pidptr) {
do_each_task_pid(-pid, PIDTYPE_PGID, p) {
send_sigurg_to_task(p, fown);
}
} while_each_task_pid(-pid, PIDTYPE_PGID, p);
}
read_unlock(&tasklist_lock);
out_unlock_fown:
......
......@@ -778,10 +778,9 @@ static struct inode_operations proc_pid_link_inode_operations = {
.follow_link = proc_pid_follow_link
};
static int pid_alive(struct task_struct *p)
static inline int pid_alive(struct task_struct *p)
{
BUG_ON(p->pids[PIDTYPE_PID].pidptr != &p->pids[PIDTYPE_PID].pid);
return atomic_read(&p->pids[PIDTYPE_PID].pid.count);
return p->pids[PIDTYPE_PID].nr != 0;
}
#define NUMBUF 10
......
......@@ -12,35 +12,22 @@ enum pid_type
struct pid
{
/* Try to keep hash_chain in the same cacheline as nr for find_pid */
struct hlist_node hash_chain;
/* Try to keep pid_chain in the same cacheline as nr for find_pid */
int nr;
atomic_t count;
struct task_struct *task;
struct list_head task_list;
};
struct pid_link
{
struct list_head pid_chain;
struct pid *pidptr;
struct pid pid;
struct hlist_node pid_chain;
/* list of pids with the same nr, only one of them is in the hash */
struct list_head pid_list;
};
#define pid_task(elem, type) \
list_entry(elem, struct task_struct, pids[type].pid_chain)
list_entry(elem, struct task_struct, pids[type].pid_list)
/*
* attach_pid() and link_pid() must be called with the tasklist_lock
* attach_pid() and detach_pid() must be called with the tasklist_lock
* write-held.
*/
extern int FASTCALL(attach_pid(struct task_struct *task, enum pid_type type, int nr));
extern void FASTCALL(link_pid(struct task_struct *task, struct pid_link *link, struct pid *pid));
/*
* detach_pid() must be called with the tasklist_lock write-held.
*/
extern void FASTCALL(detach_pid(struct task_struct *task, enum pid_type));
/*
......@@ -53,13 +40,16 @@ extern int alloc_pidmap(void);
extern void FASTCALL(free_pidmap(int));
extern void switch_exec_pids(struct task_struct *leader, struct task_struct *thread);
#define for_each_task_pid(who, type, task, elem, pid) \
if ((pid = find_pid(type, who))) \
for (elem = pid->task_list.next, \
prefetch(elem->next), \
task = pid_task(elem, type); \
elem != &pid->task_list; \
elem = elem->next, prefetch(elem->next), \
task = pid_task(elem, type))
#define do_each_task_pid(who, type, task) \
if ((task = find_task_by_pid_type(type, who))) { \
prefetch((task)->pids[type].pid_list.next); \
do {
#define while_each_task_pid(who, type, task) \
task = pid_task((task)->pids[type].pid_list.next,\
type); \
prefetch((task)->pids[type].pid_list.next); \
} while (hlist_unhashed(&(task)->pids[type].pid_chain));\
} \
#endif /* _LINUX_PID_H */
......@@ -494,7 +494,7 @@ struct task_struct {
struct task_struct *group_leader; /* threadgroup leader */
/* PID/PID hash table linkage. */
struct pid_link pids[PIDTYPE_MAX];
struct pid pids[PIDTYPE_MAX];
wait_queue_head_t wait_chldexit; /* for wait4() */
struct completion *vfork_done; /* for vfork() */
......@@ -673,7 +673,8 @@ extern struct task_struct init_task;
extern struct mm_struct init_mm;
extern struct task_struct *find_task_by_pid(int pid);
#define find_task_by_pid(nr) find_task_by_pid_type(PIDTYPE_PID, nr)
extern struct task_struct *find_task_by_pid_type(int type, int pid);
extern void set_special_pids(pid_t session, pid_t pgrp);
extern void __set_special_pids(pid_t session, pid_t pgrp);
......@@ -876,9 +877,7 @@ extern task_t * FASTCALL(next_thread(const task_t *p));
static inline int thread_group_empty(task_t *p)
{
struct pid *pid = p->pids[PIDTYPE_TGID].pidptr;
return pid->task_list.next->next == &pid->task_list;
return list_empty(&p->pids[PIDTYPE_TGID].pid_list);
}
#define delay_group_leader(p) \
......
......@@ -89,14 +89,12 @@ static inline void cap_set_pg(int pgrp, kernel_cap_t *effective,
kernel_cap_t *permitted)
{
task_t *g, *target;
struct list_head *l;
struct pid *pid;
for_each_task_pid(pgrp, PIDTYPE_PGID, g, l, pid) {
do_each_task_pid(pgrp, PIDTYPE_PGID, g) {
target = g;
while_each_thread(g, target)
security_capset_set(target, effective, inheritable, permitted);
}
} while_each_task_pid(pgrp, PIDTYPE_PGID, g);
}
/*
......
......@@ -124,16 +124,15 @@ void unhash_process(struct task_struct *p)
int session_of_pgrp(int pgrp)
{
struct task_struct *p;
struct list_head *l;
struct pid *pid;
int sid = -1;
read_lock(&tasklist_lock);
for_each_task_pid(pgrp, PIDTYPE_PGID, p, l, pid)
do_each_task_pid(pgrp, PIDTYPE_PGID, p) {
if (p->signal->session > 0) {
sid = p->signal->session;
goto out;
}
} while_each_task_pid(pgrp, PIDTYPE_PGID, p);
p = find_task_by_pid(pgrp);
if (p)
sid = p->signal->session;
......@@ -154,11 +153,9 @@ int session_of_pgrp(int pgrp)
static int will_become_orphaned_pgrp(int pgrp, task_t *ignored_task)
{
struct task_struct *p;
struct list_head *l;
struct pid *pid;
int ret = 1;
for_each_task_pid(pgrp, PIDTYPE_PGID, p, l, pid) {
do_each_task_pid(pgrp, PIDTYPE_PGID, p) {
if (p == ignored_task
|| p->state >= TASK_ZOMBIE
|| p->real_parent->pid == 1)
......@@ -168,7 +165,7 @@ static int will_become_orphaned_pgrp(int pgrp, task_t *ignored_task)
ret = 0;
break;
}
}
} while_each_task_pid(pgrp, PIDTYPE_PGID, p);
return ret; /* (sighing) "Often!" */
}
......@@ -187,10 +184,8 @@ static inline int has_stopped_jobs(int pgrp)
{
int retval = 0;
struct task_struct *p;
struct list_head *l;
struct pid *pid;
for_each_task_pid(pgrp, PIDTYPE_PGID, p, l, pid) {
do_each_task_pid(pgrp, PIDTYPE_PGID, p) {
if (p->state != TASK_STOPPED)
continue;
......@@ -206,7 +201,7 @@ static inline int has_stopped_jobs(int pgrp)
retval = 1;
break;
}
} while_each_task_pid(pgrp, PIDTYPE_PGID, p);
return retval;
}
......@@ -849,9 +844,6 @@ asmlinkage long sys_exit(int error_code)
task_t fastcall *next_thread(const task_t *p)
{
const struct pid_link *link = p->pids + PIDTYPE_TGID;
const struct list_head *tmp, *head = &link->pidptr->task_list;
#ifdef CONFIG_SMP
if (!p->sighand)
BUG();
......@@ -859,11 +851,7 @@ task_t fastcall *next_thread(const task_t *p)
!rwlock_is_locked(&tasklist_lock))
BUG();
#endif
tmp = link->pid_chain.next;
if (tmp == head)
tmp = head->next;
return pid_task(tmp, PIDTYPE_TGID);
return pid_task(p->pids[PIDTYPE_TGID].pid_list.next, PIDTYPE_TGID);
}
EXPORT_SYMBOL(next_thread);
......
......@@ -1124,14 +1124,13 @@ static task_t *copy_process(unsigned long clone_flags,
__ptrace_link(p, current->parent);
attach_pid(p, PIDTYPE_PID, p->pid);
attach_pid(p, PIDTYPE_TGID, p->tgid);
if (thread_group_leader(p)) {
attach_pid(p, PIDTYPE_TGID, p->tgid);
attach_pid(p, PIDTYPE_PGID, process_group(p));
attach_pid(p, PIDTYPE_SID, p->signal->session);
if (p->pid)
__get_cpu_var(process_counts)++;
} else
link_pid(p, p->pids + PIDTYPE_TGID, &p->group_leader->pids[PIDTYPE_TGID].pid);
}
nr_threads++;
write_unlock_irq(&tasklist_lock);
......
......@@ -146,74 +146,66 @@ int alloc_pidmap(void)
return -1;
}
fastcall struct pid *find_pid(enum pid_type type, int nr)
struct pid * fastcall find_pid(enum pid_type type, int nr)
{
struct hlist_node *elem;
struct pid *pid;
hlist_for_each_entry(pid, elem,
&pid_hash[type][pid_hashfn(nr)], hash_chain) {
&pid_hash[type][pid_hashfn(nr)], pid_chain) {
if (pid->nr == nr)
return pid;
}
return NULL;
}
void fastcall link_pid(task_t *task, struct pid_link *link, struct pid *pid)
{
atomic_inc(&pid->count);
list_add_tail(&link->pid_chain, &pid->task_list);
link->pidptr = pid;
}
int fastcall attach_pid(task_t *task, enum pid_type type, int nr)
{
struct pid *pid = find_pid(type, nr);
if (pid)
atomic_inc(&pid->count);
else {
pid = &task->pids[type].pid;
pid->nr = nr;
atomic_set(&pid->count, 1);
INIT_LIST_HEAD(&pid->task_list);
pid->task = task;
get_task_struct(task);
hlist_add_head(&pid->hash_chain,
struct pid *pid, *task_pid;
task_pid = &task->pids[type];
pid = find_pid(type, nr);
if (pid == NULL) {
hlist_add_head(&task_pid->pid_chain,
&pid_hash[type][pid_hashfn(nr)]);
INIT_LIST_HEAD(&task_pid->pid_list);
} else {
INIT_HLIST_NODE(&task_pid->pid_chain);
list_add_tail(&task_pid->pid_list, &pid->pid_list);
}
list_add_tail(&task->pids[type].pid_chain, &pid->task_list);
task->pids[type].pidptr = pid;
task_pid->nr = nr;
return 0;
}
static inline int __detach_pid(task_t *task, enum pid_type type)
{
struct pid_link *link = task->pids + type;
struct pid *pid = link->pidptr;
struct pid *pid, *pid_next;
int nr;
list_del(&link->pid_chain);
if (!atomic_dec_and_test(&pid->count))
return 0;
pid = &task->pids[type];
if (!hlist_unhashed(&pid->pid_chain)) {
hlist_del(&pid->pid_chain);
if (!list_empty(&pid->pid_list)) {
pid_next = list_entry(pid->pid_list.next,
struct pid, pid_list);
/* insert next pid from pid_list to hash */
hlist_add_head(&pid_next->pid_chain,
&pid_hash[type][pid_hashfn(pid_next->nr)]);
}
}
list_del(&pid->pid_list);
nr = pid->nr;
hlist_del(&pid->hash_chain);
put_task_struct(pid->task);
pid->nr = 0;
return nr;
}
static void _detach_pid(task_t *task, enum pid_type type)
{
__detach_pid(task, type);
}
void fastcall detach_pid(task_t *task, enum pid_type type)
{
int nr = __detach_pid(task, type);
int nr;
nr = __detach_pid(task, type);
if (!nr)
return;
......@@ -223,16 +215,18 @@ void fastcall detach_pid(task_t *task, enum pid_type type)
free_pidmap(nr);
}
task_t *find_task_by_pid(int nr)
task_t *find_task_by_pid_type(int type, int nr)
{
struct pid *pid = find_pid(PIDTYPE_PID, nr);
struct pid *pid;
pid = find_pid(type, nr);
if (!pid)
return NULL;
return pid_task(pid->task_list.next, PIDTYPE_PID);
return pid_task(&pid->pid_list, type);
}
EXPORT_SYMBOL(find_task_by_pid);
EXPORT_SYMBOL(find_task_by_pid_type);
/*
* This function switches the PIDs if a non-leader thread calls
......@@ -241,13 +235,13 @@ EXPORT_SYMBOL(find_task_by_pid);
*/
void switch_exec_pids(task_t *leader, task_t *thread)
{
_detach_pid(leader, PIDTYPE_PID);
_detach_pid(leader, PIDTYPE_TGID);
_detach_pid(leader, PIDTYPE_PGID);
_detach_pid(leader, PIDTYPE_SID);
__detach_pid(leader, PIDTYPE_PID);
__detach_pid(leader, PIDTYPE_TGID);
__detach_pid(leader, PIDTYPE_PGID);
__detach_pid(leader, PIDTYPE_SID);
_detach_pid(thread, PIDTYPE_PID);
_detach_pid(thread, PIDTYPE_TGID);
__detach_pid(thread, PIDTYPE_PID);
__detach_pid(thread, PIDTYPE_TGID);
leader->pid = leader->tgid = thread->pid;
thread->pid = thread->tgid;
......
......@@ -1110,8 +1110,6 @@ int group_send_sig_info(int sig, struct siginfo *info, struct task_struct *p)
int __kill_pg_info(int sig, struct siginfo *info, pid_t pgrp)
{
struct task_struct *p;
struct list_head *l;
struct pid *pid;
int retval, success;
if (pgrp <= 0)
......@@ -1119,11 +1117,11 @@ int __kill_pg_info(int sig, struct siginfo *info, pid_t pgrp)
success = 0;
retval = -ESRCH;
for_each_task_pid(pgrp, PIDTYPE_PGID, p, l, pid) {
do_each_task_pid(pgrp, PIDTYPE_PGID, p) {
int err = group_send_sig_info(sig, info, p);
success |= !err;
retval = err;
}
} while_each_task_pid(pgrp, PIDTYPE_PGID, p);
return success ? 0 : retval;
}
......@@ -1150,8 +1148,6 @@ int
kill_sl_info(int sig, struct siginfo *info, pid_t sid)
{
int err, retval = -EINVAL;
struct pid *pid;
struct list_head *l;
struct task_struct *p;
if (sid <= 0)
......@@ -1159,13 +1155,13 @@ kill_sl_info(int sig, struct siginfo *info, pid_t sid)
retval = -ESRCH;
read_lock(&tasklist_lock);
for_each_task_pid(sid, PIDTYPE_SID, p, l, pid) {
do_each_task_pid(sid, PIDTYPE_SID, p) {
if (!p->signal->leader)
continue;
err = group_send_sig_info(sig, info, p);
if (retval)
retval = err;
}
} while_each_task_pid(sid, PIDTYPE_SID, p);
read_unlock(&tasklist_lock);
out:
return retval;
......
......@@ -310,8 +310,6 @@ asmlinkage long sys_setpriority(int which, int who, int niceval)
{
struct task_struct *g, *p;
struct user_struct *user;
struct pid *pid;
struct list_head *l;
int error = -EINVAL;
if (which > 2 || which < 0)
......@@ -336,8 +334,9 @@ asmlinkage long sys_setpriority(int which, int who, int niceval)
case PRIO_PGRP:
if (!who)
who = process_group(current);
for_each_task_pid(who, PIDTYPE_PGID, p, l, pid)
do_each_task_pid(who, PIDTYPE_PGID, p) {
error = set_one_prio(p, niceval, error);
} while_each_task_pid(who, PIDTYPE_PGID, p);
break;
case PRIO_USER:
if (!who)
......@@ -371,8 +370,6 @@ asmlinkage long sys_setpriority(int which, int who, int niceval)
asmlinkage long sys_getpriority(int which, int who)
{
struct task_struct *g, *p;
struct list_head *l;
struct pid *pid;
struct user_struct *user;
long niceval, retval = -ESRCH;
......@@ -394,11 +391,11 @@ asmlinkage long sys_getpriority(int which, int who)
case PRIO_PGRP:
if (!who)
who = process_group(current);
for_each_task_pid(who, PIDTYPE_PGID, p, l, pid) {
do_each_task_pid(who, PIDTYPE_PGID, p) {
niceval = 20 - task_nice(p);
if (niceval > retval)
retval = niceval;
}
} while_each_task_pid(who, PIDTYPE_PGID, p);
break;
case PRIO_USER:
if (!who)
......@@ -1044,12 +1041,11 @@ asmlinkage long sys_setpgid(pid_t pid, pid_t pgid)
if (pgid != pid) {
struct task_struct *p;
struct pid *pid;
struct list_head *l;
for_each_task_pid(pgid, PIDTYPE_PGID, p, l, pid)
do_each_task_pid(pgid, PIDTYPE_PGID, p) {
if (p->signal->session == current->signal->session)
goto ok_pgid;
} while_each_task_pid(pgid, PIDTYPE_PGID, p);
goto out;
}
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment