pid namespaces: define is_global_init() and is_container_init()

is_init() is an ambiguous name for the pid==1 check.  Split it into
is_global_init() and is_container_init().

A cgroup init has it's tsk->pid == 1.

A global init also has it's tsk->pid == 1 and it's active pid namespace
is the init_pid_ns.  But rather than check the active pid namespace,
compare the task structure with 'init_pid_ns.child_reaper', which is
initialized during boot to the /sbin/init process and never changes.

Changelog:

	2.6.22-rc4-mm2-pidns1:
	- Use 'init_pid_ns.child_reaper' to determine if a given task is the
	  global init (/sbin/init) process. This would improve performance
	  and remove dependence on the task_pid().

	2.6.21-mm2-pidns2:

	- [Sukadev Bhattiprolu] Changed is_container_init() calls in {powerpc,
	  ppc,avr32}/traps.c for the _exception() call to is_global_init().
	  This way, we kill only the cgroup if the cgroup's init has a
	  bug rather than force a kernel panic.

[akpm@linux-foundation.org: fix comment]
[sukadev@us.ibm.com: Use is_global_init() in arch/m32r/mm/fault.c]
[bunk@stusta.de: kernel/pid.c: remove unused exports]
[sukadev@us.ibm.com: Fix capability.c to work with threaded init]
Signed-off-by: Serge E. Hallyn <serue@us.ibm.com>
Signed-off-by: Sukadev Bhattiprolu <sukadev@us.ibm.com>
Acked-by: Pavel Emelianov <xemul@openvz.org>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Herbert Poetzel <herbert@13thfloor.at>
Cc: Kirill Korotaev <dev@sw.ru>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

arch/alpha/mm/fault.c

@@ -188,7 +188,7 @@ do_page_fault(unsigned long address, unsigned long mmcsr,
 	/* We ran out of memory, or some other thing happened to us that
 	   made us unable to handle the page fault gracefully.  */
  out_of_memory:
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		yield();
 		down_read(&mm->mmap_sem);
 		goto survive;

arch/arm/mm/fault.c

@@ -197,7 +197,7 @@ survive:
 	return fault;
 
 out_of_memory:
-	if (!is_init(tsk))
+	if (!is_global_init(tsk))
 		goto out;
 
 	/*

arch/avr32/kernel/traps.c

@@ -89,7 +89,7 @@ void _exception(long signr, struct pt_regs *regs, int code,
 	 * generate the same exception over and over again and we get
 	 * nowhere.  Better to kill it and let the kernel panic.
 	 */
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		__sighandler_t handler;
 
 		spin_lock_irq(&current->sighand->siglock);

arch/avr32/mm/fault.c

@@ -160,7 +160,7 @@ bad_area:
 		if (exception_trace && printk_ratelimit())
 			printk("%s%s[%d]: segfault at %08lx pc %08lx "
 			       "sp %08lx ecr %lu\n",
-			       is_init(tsk) ? KERN_EMERG : KERN_INFO,
+			       is_global_init(tsk) ? KERN_EMERG : KERN_INFO,
 			       tsk->comm, tsk->pid, address, regs->pc,
 			       regs->sp, ecr);
 		_exception(SIGSEGV, regs, code, address);
@@ -209,7 +209,7 @@ no_context:
 	 */
 out_of_memory:
 	up_read(&mm->mmap_sem);
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		yield();
 		down_read(&mm->mmap_sem);
 		goto survive;
@@ -231,7 +231,7 @@ do_sigbus:
 	if (exception_trace)
 		printk("%s%s[%d]: bus error at %08lx pc %08lx "
 		       "sp %08lx ecr %lu\n",
-		       is_init(tsk) ? KERN_EMERG : KERN_INFO,
+		       is_global_init(tsk) ? KERN_EMERG : KERN_INFO,
 		       tsk->comm, tsk->pid, address, regs->pc,
 		       regs->sp, ecr);
 

arch/ia64/mm/fault.c

@@ -274,7 +274,7 @@ ia64_do_page_fault (unsigned long address, unsigned long isr, struct pt_regs *re
 
   out_of_memory:
 	up_read(&mm->mmap_sem);
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		yield();
 		down_read(&mm->mmap_sem);
 		goto survive;

arch/m32r/mm/fault.c

@@ -271,7 +271,7 @@ no_context:
  */
 out_of_memory:
 	up_read(&mm->mmap_sem);
-	if (is_init(tsk)) {
+	if (is_global_init(tsk)) {
 		yield();
 		down_read(&mm->mmap_sem);
 		goto survive;

arch/m68k/mm/fault.c

@@ -180,7 +180,7 @@ good_area:
  */
 out_of_memory:
 	up_read(&mm->mmap_sem);
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		yield();
 		down_read(&mm->mmap_sem);
 		goto survive;

arch/mips/mm/fault.c

@@ -173,7 +173,7 @@ no_context:
  */
 out_of_memory:
 	up_read(&mm->mmap_sem);
-	if (is_init(tsk)) {
+	if (is_global_init(tsk)) {
 		yield();
 		down_read(&mm->mmap_sem);
 		goto survive;

arch/powerpc/kernel/traps.c

@@ -201,7 +201,7 @@ void _exception(int signr, struct pt_regs *regs, int code, unsigned long addr)
 	 * generate the same exception over and over again and we get
 	 * nowhere.  Better to kill it and let the kernel panic.
 	 */
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		__sighandler_t handler;
 
 		spin_lock_irq(&current->sighand->siglock);

arch/powerpc/mm/fault.c

@@ -375,7 +375,7 @@ bad_area_nosemaphore:
  */
 out_of_memory:
 	up_read(&mm->mmap_sem);
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		yield();
 		down_read(&mm->mmap_sem);
 		goto survive;

arch/powerpc/platforms/pseries/ras.c

@@ -332,7 +332,7 @@ static int recover_mce(struct pt_regs *regs, struct rtas_error_log * err)
 		   err->disposition == RTAS_DISP_NOT_RECOVERED &&
 		   err->target == RTAS_TARGET_MEMORY &&
 		   err->type == RTAS_TYPE_ECC_UNCORR &&
-		   !(current->pid == 0 || is_init(current))) {
+		   !(current->pid == 0 || is_global_init(current))) {
 		/* Kill off a user process with an ECC error */
 		printk(KERN_ERR "MCE: uncorrectable ecc error for pid %d\n",
 		       current->pid);

arch/ppc/kernel/traps.c

@@ -121,7 +121,7 @@ void _exception(int signr, struct pt_regs *regs, int code, unsigned long addr)
 	 * generate the same exception over and over again and we get
 	 * nowhere.  Better to kill it and let the kernel panic.
 	 */
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		__sighandler_t handler;
 
 		spin_lock_irq(&current->sighand->siglock);

arch/ppc/mm/fault.c

@@ -290,7 +290,7 @@ bad_area:
  */
 out_of_memory:
 	up_read(&mm->mmap_sem);
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		yield();
 		down_read(&mm->mmap_sem);
 		goto survive;

arch/s390/lib/uaccess_pt.c

@@ -64,7 +64,7 @@ out:
 
 out_of_memory:
 	up_read(&mm->mmap_sem);
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		yield();
 		down_read(&mm->mmap_sem);
 		goto survive;

arch/s390/mm/fault.c

@@ -211,7 +211,7 @@ static int do_out_of_memory(struct pt_regs *regs, unsigned long error_code,
 	struct mm_struct *mm = tsk->mm;
 
 	up_read(&mm->mmap_sem);
-	if (is_init(tsk)) {
+	if (is_global_init(tsk)) {
 		yield();
 		down_read(&mm->mmap_sem);
 		return 1;

arch/sh/mm/fault.c

@@ -207,7 +207,7 @@ no_context:
  */
 out_of_memory:
 	up_read(&mm->mmap_sem);
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		yield();
 		down_read(&mm->mmap_sem);
 		goto survive;

arch/sh64/mm/fault.c

@@ -278,7 +278,7 @@ bad_area:
 			show_regs(regs);
 #endif
 		}
-		if (is_init(tsk)) {
+		if (is_global_init(tsk)) {
 			panic("INIT had user mode bad_area\n");
 		}
 		tsk->thread.address = address;
@@ -320,14 +320,14 @@ no_context:
  * us unable to handle the page fault gracefully.
  */
 out_of_memory:
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		panic("INIT out of memory\n");
 		yield();
 		goto survive;
 	}
 	printk("fault:Out of memory\n");
 	up_read(&mm->mmap_sem);
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		yield();
 		down_read(&mm->mmap_sem);
 		goto survive;

arch/um/kernel/trap.c

@@ -108,7 +108,7 @@ out_nosemaphore:
  * us unable to handle the page fault gracefully.
  */
 out_of_memory:
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		up_read(&mm->mmap_sem);
 		yield();
 		down_read(&mm->mmap_sem);

arch/x86/lib/usercopy_32.c

@@ -748,7 +748,7 @@ survive:
 			retval = get_user_pages(current, current->mm,
 					(unsigned long )to, 1, 1, 0, &pg, NULL);
 
-			if (retval == -ENOMEM && is_init(current)) {
+			if (retval == -ENOMEM && is_global_init(current)) {
 				up_read(&current->mm->mmap_sem);
 				congestion_wait(WRITE, HZ/50);
 				goto survive;

arch/x86/mm/fault_32.c

@@ -587,7 +587,7 @@ no_context:
  */
 out_of_memory:
 	up_read(&mm->mmap_sem);
-	if (is_init(tsk)) {
+	if (is_global_init(tsk)) {
 		yield();
 		down_read(&mm->mmap_sem);
 		goto survive;

arch/x86/mm/fault_64.c

@@ -554,7 +554,7 @@ no_context:
  */
 out_of_memory:
 	up_read(&mm->mmap_sem);
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		yield();
 		goto again;
 	}

arch/xtensa/mm/fault.c

@@ -145,7 +145,7 @@ bad_area:
 	 */
 out_of_memory:
 	up_read(&mm->mmap_sem);
-	if (is_init(current)) {
+	if (is_global_init(current)) {
 		yield();
 		down_read(&mm->mmap_sem);
 		goto survive;

drivers/char/sysrq.c

@@ -251,7 +251,7 @@ static void send_sig_all(int sig)
 	struct task_struct *p;
 
 	for_each_process(p) {
-		if (p->mm && !is_init(p))
+		if (p->mm && !is_global_init(p))
 			/* Not swapper, init nor kernel thread */
 			force_sig(sig, p);
 	}

include/linux/sched.h

@@ -1237,12 +1237,20 @@ static inline int pid_alive(struct task_struct *p)
 }
 
 /**
- * is_init - check if a task structure is init
+ * is_global_init - check if a task structure is init
  * @tsk: Task structure to be checked.
  *
  * Check if a task structure is the first user space task the kernel created.
+ *
+ * TODO: We should inline this function after some cleanups in pid_namespace.h
+ */
+extern int is_global_init(struct task_struct *tsk);
+
+/*
+ * is_container_init:
+ * check whether in the task is init in its own pid namespace.
  */
-static inline int is_init(struct task_struct *tsk)
+static inline int is_container_init(struct task_struct *tsk)
 {
 	return tsk->pid == 1;
 }

kernel/capability.c

@@ -12,6 +12,7 @@
 #include <linux/module.h>
 #include <linux/security.h>
 #include <linux/syscalls.h>
+#include <linux/pid_namespace.h>
 #include <asm/uaccess.h>
 
 /*
@@ -129,7 +130,7 @@ static inline int cap_set_all(kernel_cap_t *effective,
      int found = 0;
 
      do_each_thread(g, target) {
-             if (target == current || is_init(target))
+             if (target == current || is_container_init(target->group_leader))
                      continue;
              found = 1;
 	     if (security_capset_check(target, effective, inheritable,

kernel/exit.c

@@ -221,7 +221,7 @@ static int will_become_orphaned_pgrp(struct pid *pgrp, struct task_struct *ignor
 	do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
 		if (p == ignored_task
 				|| p->exit_state
-				|| is_init(p->real_parent))
+				|| is_global_init(p->real_parent))
 			continue;
 		if (task_pgrp(p->real_parent) != pgrp &&
 		    task_session(p->real_parent) == task_session(p)) {

kernel/kexec.c

@@ -51,7 +51,7 @@ struct resource crashk_res = {
 
 int kexec_should_crash(struct task_struct *p)
 {
-	if (in_interrupt() || !p->pid || is_init(p) || panic_on_oops)
+	if (in_interrupt() || !p->pid || is_global_init(p) || panic_on_oops)
 		return 1;
 	return 0;
 }

kernel/pid.c

@@ -70,6 +70,11 @@ struct pid_namespace init_pid_ns = {
 	.child_reaper = &init_task
 };
 
+int is_global_init(struct task_struct *tsk)
+{
+	return tsk == init_pid_ns.child_reaper;
+}
+
 /*
  * Note: disable interrupts while the pidmap_lock is held as an
  * interrupt might come in and do read_lock(&tasklist_lock).

kernel/signal.c

@@ -256,7 +256,7 @@ flush_signal_handlers(struct task_struct *t, int force_default)
 
 int unhandled_signal(struct task_struct *tsk, int sig)
 {
-	if (is_init(tsk))
+	if (is_global_init(tsk))
 		return 1;
 	if (tsk->ptrace & PT_PTRACED)
 		return 0;

kernel/sysctl.c

@@ -1888,7 +1888,7 @@ int proc_dointvec_bset(struct ctl_table *table, int write, struct file *filp,
 		return -EPERM;
 	}
 
-	op = is_init(current) ? OP_SET : OP_AND;
+	op = is_global_init(current) ? OP_SET : OP_AND;
 	return do_proc_dointvec(table,write,filp,buffer,lenp,ppos,
 				do_proc_dointvec_bset_conv,&op);
 }

mm/oom_kill.c

@@ -212,7 +212,7 @@ static struct task_struct *select_bad_process(unsigned long *ppoints)
 		if (!p->mm)
 			continue;
 		/* skip the init task */
-		if (is_init(p))
+		if (is_global_init(p))
 			continue;
 
 		/*
@@ -265,7 +265,7 @@ static struct task_struct *select_bad_process(unsigned long *ppoints)
  */
 static void __oom_kill_task(struct task_struct *p, int verbose)
 {
-	if (is_init(p)) {
+	if (is_global_init(p)) {
 		WARN_ON(1);
 		printk(KERN_WARNING "tried to kill init!\n");
 		return;

security/commoncap.c

@@ -23,6 +23,7 @@
 #include <linux/xattr.h>
 #include <linux/hugetlb.h>
 #include <linux/mount.h>
+#include <linux/sched.h>
 
 #ifdef CONFIG_SECURITY_FILE_CAPABILITIES
 /*
@@ -334,7 +335,7 @@ void cap_bprm_apply_creds (struct linux_binprm *bprm, int unsafe)
 	/* For init, we want to retain the capabilities set
 	 * in the init_task struct. Thus we skip the usual
 	 * capability rules */
-	if (!is_init(current)) {
+	if (!is_global_init(current)) {
 		current->cap_permitted = new_permitted;
 		current->cap_effective = bprm->cap_effective ?
 				new_permitted : 0;