aboutsummaryrefslogtreecommitdiffstats
path: root/kernel
diff options
context:
space:
mode:
Diffstat (limited to 'kernel')
-rw-r--r--kernel/Makefile8
-rw-r--r--kernel/audit.c2
-rw-r--r--kernel/auditsc.c2
-rw-r--r--kernel/cgroup.c15
-rw-r--r--kernel/compat.c7
-rw-r--r--kernel/configs.c4
-rw-r--r--kernel/cpuset.c10
-rw-r--r--kernel/cred.c6
-rw-r--r--kernel/debug/debug_core.c2
-rw-r--r--kernel/debug/gdbstub.c22
-rw-r--r--kernel/debug/kdb/kdb_bt.c5
-rw-r--r--kernel/debug/kdb/kdb_cmds4
-rw-r--r--kernel/debug/kdb/kdb_debugger.c21
-rw-r--r--kernel/debug/kdb/kdb_io.c36
-rw-r--r--kernel/debug/kdb/kdb_main.c4
-rw-r--r--kernel/debug/kdb/kdb_private.h3
-rw-r--r--kernel/delayacct.c2
-rw-r--r--kernel/events/core.c67
-rw-r--r--kernel/exit.c3
-rw-r--r--kernel/fork.c89
-rw-r--r--kernel/futex.c58
-rw-r--r--kernel/gcov/Kconfig2
-rw-r--r--kernel/irq/Kconfig4
-rw-r--r--kernel/irq/Makefile1
-rw-r--r--kernel/irq/chip.c2
-rw-r--r--kernel/irq/devres.c2
-rw-r--r--kernel/irq/generic-chip.c4
-rw-r--r--kernel/irq/irqdesc.c37
-rw-r--r--kernel/irq/irqdomain.c184
-rw-r--r--kernel/irq/manage.c17
-rw-r--r--kernel/kexec.c2
-rw-r--r--kernel/kmod.c2
-rw-r--r--kernel/lockdep.c45
-rw-r--r--kernel/module.c80
-rw-r--r--kernel/notifier.c31
-rw-r--r--kernel/panic.c2
-rw-r--r--kernel/params.c18
-rw-r--r--kernel/pid.c1
-rw-r--r--kernel/posix-cpu-timers.c5
-rw-r--r--kernel/power/Kconfig8
-rw-r--r--kernel/printk.c8
-rw-r--r--kernel/ptrace.c23
-rw-r--r--kernel/rcupdate.c2
-rw-r--r--kernel/rcutorture.c4
-rw-r--r--kernel/rcutree_trace.c2
-rw-r--r--kernel/resource.c28
-rw-r--r--kernel/rwsem.c2
-rw-r--r--kernel/sched.c158
-rw-r--r--kernel/sched_features.h4
-rw-r--r--kernel/sched_rt.c4
-rw-r--r--kernel/signal.c17
-rw-r--r--kernel/stop_machine.c2
-rw-r--r--kernel/sys.c85
-rw-r--r--kernel/sys_ni.c1
-rw-r--r--kernel/sysctl_binary.c2
-rw-r--r--kernel/sysctl_check.c2
-rw-r--r--kernel/taskstats.c21
-rw-r--r--kernel/time/alarmtimer.c18
-rw-r--r--kernel/trace/Kconfig2
-rw-r--r--kernel/trace/blktrace.c21
-rw-r--r--kernel/trace/trace.h2
-rw-r--r--kernel/trace/trace_mmiotrace.c2
-rw-r--r--kernel/tsacct.c15
-rw-r--r--kernel/workqueue.c7
64 files changed, 894 insertions, 355 deletions
diff --git a/kernel/Makefile b/kernel/Makefile
index 2d64cfcc8b42..eca595e2fd52 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -10,7 +10,7 @@ obj-y = sched.o fork.o exec_domain.o panic.o printk.o \
10 kthread.o wait.o kfifo.o sys_ni.o posix-cpu-timers.o mutex.o \ 10 kthread.o wait.o kfifo.o sys_ni.o posix-cpu-timers.o mutex.o \
11 hrtimer.o rwsem.o nsproxy.o srcu.o semaphore.o \ 11 hrtimer.o rwsem.o nsproxy.o srcu.o semaphore.o \
12 notifier.o ksysfs.o pm_qos_params.o sched_clock.o cred.o \ 12 notifier.o ksysfs.o pm_qos_params.o sched_clock.o cred.o \
13 async.o range.o jump_label.o 13 async.o range.o
14obj-y += groups.o 14obj-y += groups.o
15 15
16ifdef CONFIG_FUNCTION_TRACER 16ifdef CONFIG_FUNCTION_TRACER
@@ -107,6 +107,7 @@ obj-$(CONFIG_PERF_EVENTS) += events/
107obj-$(CONFIG_USER_RETURN_NOTIFIER) += user-return-notifier.o 107obj-$(CONFIG_USER_RETURN_NOTIFIER) += user-return-notifier.o
108obj-$(CONFIG_PADATA) += padata.o 108obj-$(CONFIG_PADATA) += padata.o
109obj-$(CONFIG_CRASH_DUMP) += crash_dump.o 109obj-$(CONFIG_CRASH_DUMP) += crash_dump.o
110obj-$(CONFIG_JUMP_LABEL) += jump_label.o
110 111
111ifneq ($(CONFIG_SCHED_OMIT_FRAME_POINTER),y) 112ifneq ($(CONFIG_SCHED_OMIT_FRAME_POINTER),y)
112# According to Alan Modra <alan@linuxcare.com.au>, the -fno-omit-frame-pointer is 113# According to Alan Modra <alan@linuxcare.com.au>, the -fno-omit-frame-pointer is
@@ -125,11 +126,10 @@ targets += config_data.gz
125$(obj)/config_data.gz: $(KCONFIG_CONFIG) FORCE 126$(obj)/config_data.gz: $(KCONFIG_CONFIG) FORCE
126 $(call if_changed,gzip) 127 $(call if_changed,gzip)
127 128
128quiet_cmd_ikconfiggz = IKCFG $@ 129 filechk_ikconfiggz = (echo "static const char kernel_config_data[] __used = MAGIC_START"; cat $< | scripts/bin2c; echo "MAGIC_END;")
129 cmd_ikconfiggz = (echo "static const char kernel_config_data[] __used = MAGIC_START"; cat $< | scripts/bin2c; echo "MAGIC_END;") > $@
130targets += config_data.h 130targets += config_data.h
131$(obj)/config_data.h: $(obj)/config_data.gz FORCE 131$(obj)/config_data.h: $(obj)/config_data.gz FORCE
132 $(call if_changed,ikconfiggz) 132 $(call filechk,ikconfiggz)
133 133
134$(obj)/time.o: $(obj)/timeconst.h 134$(obj)/time.o: $(obj)/timeconst.h
135 135
diff --git a/kernel/audit.c b/kernel/audit.c
index 52501b5d4902..0a1355ca3d79 100644
--- a/kernel/audit.c
+++ b/kernel/audit.c
@@ -43,7 +43,7 @@
43 43
44#include <linux/init.h> 44#include <linux/init.h>
45#include <asm/types.h> 45#include <asm/types.h>
46#include <asm/atomic.h> 46#include <linux/atomic.h>
47#include <linux/mm.h> 47#include <linux/mm.h>
48#include <linux/module.h> 48#include <linux/module.h>
49#include <linux/slab.h> 49#include <linux/slab.h>
diff --git a/kernel/auditsc.c b/kernel/auditsc.c
index 00d79df03e76..ce4b054acee5 100644
--- a/kernel/auditsc.c
+++ b/kernel/auditsc.c
@@ -44,7 +44,7 @@
44 44
45#include <linux/init.h> 45#include <linux/init.h>
46#include <asm/types.h> 46#include <asm/types.h>
47#include <asm/atomic.h> 47#include <linux/atomic.h>
48#include <linux/fs.h> 48#include <linux/fs.h>
49#include <linux/namei.h> 49#include <linux/namei.h>
50#include <linux/mm.h> 50#include <linux/mm.h>
diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index e1c72c0f512b..1d2b6ceea95d 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -27,9 +27,11 @@
27 */ 27 */
28 28
29#include <linux/cgroup.h> 29#include <linux/cgroup.h>
30#include <linux/cred.h>
30#include <linux/ctype.h> 31#include <linux/ctype.h>
31#include <linux/errno.h> 32#include <linux/errno.h>
32#include <linux/fs.h> 33#include <linux/fs.h>
34#include <linux/init_task.h>
33#include <linux/kernel.h> 35#include <linux/kernel.h>
34#include <linux/list.h> 36#include <linux/list.h>
35#include <linux/mm.h> 37#include <linux/mm.h>
@@ -59,7 +61,7 @@
59#include <linux/poll.h> 61#include <linux/poll.h>
60#include <linux/flex_array.h> /* used in cgroup_attach_proc */ 62#include <linux/flex_array.h> /* used in cgroup_attach_proc */
61 63
62#include <asm/atomic.h> 64#include <linux/atomic.h>
63 65
64static DEFINE_MUTEX(cgroup_mutex); 66static DEFINE_MUTEX(cgroup_mutex);
65 67
@@ -1514,6 +1516,7 @@ static struct dentry *cgroup_mount(struct file_system_type *fs_type,
1514 struct cgroup *root_cgrp = &root->top_cgroup; 1516 struct cgroup *root_cgrp = &root->top_cgroup;
1515 struct inode *inode; 1517 struct inode *inode;
1516 struct cgroupfs_root *existing_root; 1518 struct cgroupfs_root *existing_root;
1519 const struct cred *cred;
1517 int i; 1520 int i;
1518 1521
1519 BUG_ON(sb->s_root != NULL); 1522 BUG_ON(sb->s_root != NULL);
@@ -1593,7 +1596,9 @@ static struct dentry *cgroup_mount(struct file_system_type *fs_type,
1593 BUG_ON(!list_empty(&root_cgrp->children)); 1596 BUG_ON(!list_empty(&root_cgrp->children));
1594 BUG_ON(root->number_of_cgroups != 1); 1597 BUG_ON(root->number_of_cgroups != 1);
1595 1598
1599 cred = override_creds(&init_cred);
1596 cgroup_populate_dir(root_cgrp); 1600 cgroup_populate_dir(root_cgrp);
1601 revert_creds(cred);
1597 mutex_unlock(&cgroup_mutex); 1602 mutex_unlock(&cgroup_mutex);
1598 mutex_unlock(&inode->i_mutex); 1603 mutex_unlock(&inode->i_mutex);
1599 } else { 1604 } else {
@@ -1697,7 +1702,6 @@ int cgroup_path(const struct cgroup *cgrp, char *buf, int buflen)
1697{ 1702{
1698 char *start; 1703 char *start;
1699 struct dentry *dentry = rcu_dereference_check(cgrp->dentry, 1704 struct dentry *dentry = rcu_dereference_check(cgrp->dentry,
1700 rcu_read_lock_held() ||
1701 cgroup_lock_is_held()); 1705 cgroup_lock_is_held());
1702 1706
1703 if (!dentry || cgrp == dummytop) { 1707 if (!dentry || cgrp == dummytop) {
@@ -1723,7 +1727,6 @@ int cgroup_path(const struct cgroup *cgrp, char *buf, int buflen)
1723 break; 1727 break;
1724 1728
1725 dentry = rcu_dereference_check(cgrp->dentry, 1729 dentry = rcu_dereference_check(cgrp->dentry,
1726 rcu_read_lock_held() ||
1727 cgroup_lock_is_held()); 1730 cgroup_lock_is_held());
1728 if (!cgrp->parent) 1731 if (!cgrp->parent)
1729 continue; 1732 continue;
@@ -4814,8 +4817,7 @@ unsigned short css_id(struct cgroup_subsys_state *css)
4814 * on this or this is under rcu_read_lock(). Once css->id is allocated, 4817 * on this or this is under rcu_read_lock(). Once css->id is allocated,
4815 * it's unchanged until freed. 4818 * it's unchanged until freed.
4816 */ 4819 */
4817 cssid = rcu_dereference_check(css->id, 4820 cssid = rcu_dereference_check(css->id, atomic_read(&css->refcnt));
4818 rcu_read_lock_held() || atomic_read(&css->refcnt));
4819 4821
4820 if (cssid) 4822 if (cssid)
4821 return cssid->id; 4823 return cssid->id;
@@ -4827,8 +4829,7 @@ unsigned short css_depth(struct cgroup_subsys_state *css)
4827{ 4829{
4828 struct css_id *cssid; 4830 struct css_id *cssid;
4829 4831
4830 cssid = rcu_dereference_check(css->id, 4832 cssid = rcu_dereference_check(css->id, atomic_read(&css->refcnt));
4831 rcu_read_lock_held() || atomic_read(&css->refcnt));
4832 4833
4833 if (cssid) 4834 if (cssid)
4834 return cssid->depth; 4835 return cssid->depth;
diff --git a/kernel/compat.c b/kernel/compat.c
index fc9eb093acd5..e2435ee9993a 100644
--- a/kernel/compat.c
+++ b/kernel/compat.c
@@ -158,6 +158,7 @@ int put_compat_timespec(const struct timespec *ts, struct compat_timespec __user
158 __put_user(ts->tv_sec, &cts->tv_sec) || 158 __put_user(ts->tv_sec, &cts->tv_sec) ||
159 __put_user(ts->tv_nsec, &cts->tv_nsec)) ? -EFAULT : 0; 159 __put_user(ts->tv_nsec, &cts->tv_nsec)) ? -EFAULT : 0;
160} 160}
161EXPORT_SYMBOL_GPL(put_compat_timespec);
161 162
162static long compat_nanosleep_restart(struct restart_block *restart) 163static long compat_nanosleep_restart(struct restart_block *restart)
163{ 164{
@@ -890,6 +891,7 @@ sigset_from_compat (sigset_t *set, compat_sigset_t *compat)
890 case 1: set->sig[0] = compat->sig[0] | (((long)compat->sig[1]) << 32 ); 891 case 1: set->sig[0] = compat->sig[0] | (((long)compat->sig[1]) << 32 );
891 } 892 }
892} 893}
894EXPORT_SYMBOL_GPL(sigset_from_compat);
893 895
894asmlinkage long 896asmlinkage long
895compat_sys_rt_sigtimedwait (compat_sigset_t __user *uthese, 897compat_sys_rt_sigtimedwait (compat_sigset_t __user *uthese,
@@ -991,11 +993,8 @@ asmlinkage long compat_sys_rt_sigsuspend(compat_sigset_t __user *unewset, compat
991 sigset_from_compat(&newset, &newset32); 993 sigset_from_compat(&newset, &newset32);
992 sigdelsetmask(&newset, sigmask(SIGKILL)|sigmask(SIGSTOP)); 994 sigdelsetmask(&newset, sigmask(SIGKILL)|sigmask(SIGSTOP));
993 995
994 spin_lock_irq(&current->sighand->siglock);
995 current->saved_sigmask = current->blocked; 996 current->saved_sigmask = current->blocked;
996 current->blocked = newset; 997 set_current_blocked(&newset);
997 recalc_sigpending();
998 spin_unlock_irq(&current->sighand->siglock);
999 998
1000 current->state = TASK_INTERRUPTIBLE; 999 current->state = TASK_INTERRUPTIBLE;
1001 schedule(); 1000 schedule();
diff --git a/kernel/configs.c b/kernel/configs.c
index b4066b44a99d..42e8fa075eed 100644
--- a/kernel/configs.c
+++ b/kernel/configs.c
@@ -92,8 +92,8 @@ static void __exit ikconfig_cleanup(void)
92module_init(ikconfig_init); 92module_init(ikconfig_init);
93module_exit(ikconfig_cleanup); 93module_exit(ikconfig_cleanup);
94 94
95#endif /* CONFIG_IKCONFIG_PROC */
96
95MODULE_LICENSE("GPL"); 97MODULE_LICENSE("GPL");
96MODULE_AUTHOR("Randy Dunlap"); 98MODULE_AUTHOR("Randy Dunlap");
97MODULE_DESCRIPTION("Echo the kernel .config file used to build the kernel"); 99MODULE_DESCRIPTION("Echo the kernel .config file used to build the kernel");
98
99#endif /* CONFIG_IKCONFIG_PROC */
diff --git a/kernel/cpuset.c b/kernel/cpuset.c
index 9c9b7545c810..10131fdaff70 100644
--- a/kernel/cpuset.c
+++ b/kernel/cpuset.c
@@ -55,7 +55,7 @@
55#include <linux/sort.h> 55#include <linux/sort.h>
56 56
57#include <asm/uaccess.h> 57#include <asm/uaccess.h>
58#include <asm/atomic.h> 58#include <linux/atomic.h>
59#include <linux/mutex.h> 59#include <linux/mutex.h>
60#include <linux/workqueue.h> 60#include <linux/workqueue.h>
61#include <linux/cgroup.h> 61#include <linux/cgroup.h>
@@ -2460,11 +2460,19 @@ static int cpuset_spread_node(int *rotor)
2460 2460
2461int cpuset_mem_spread_node(void) 2461int cpuset_mem_spread_node(void)
2462{ 2462{
2463 if (current->cpuset_mem_spread_rotor == NUMA_NO_NODE)
2464 current->cpuset_mem_spread_rotor =
2465 node_random(&current->mems_allowed);
2466
2463 return cpuset_spread_node(&current->cpuset_mem_spread_rotor); 2467 return cpuset_spread_node(&current->cpuset_mem_spread_rotor);
2464} 2468}
2465 2469
2466int cpuset_slab_spread_node(void) 2470int cpuset_slab_spread_node(void)
2467{ 2471{
2472 if (current->cpuset_slab_spread_rotor == NUMA_NO_NODE)
2473 current->cpuset_slab_spread_rotor =
2474 node_random(&current->mems_allowed);
2475
2468 return cpuset_spread_node(&current->cpuset_slab_spread_rotor); 2476 return cpuset_spread_node(&current->cpuset_slab_spread_rotor);
2469} 2477}
2470 2478
diff --git a/kernel/cred.c b/kernel/cred.c
index 174fa84eca30..8ef31f53c44c 100644
--- a/kernel/cred.c
+++ b/kernel/cred.c
@@ -508,10 +508,8 @@ int commit_creds(struct cred *new)
508 key_fsgid_changed(task); 508 key_fsgid_changed(task);
509 509
510 /* do it 510 /* do it
511 * - What if a process setreuid()'s and this brings the 511 * RLIMIT_NPROC limits on user->processes have already been checked
512 * new uid over his NPROC rlimit? We can check this now 512 * in set_user().
513 * cheaply with the new uid cache, so if it matters
514 * we should be checking for it. -DaveM
515 */ 513 */
516 alter_cred_subscribers(new, 2); 514 alter_cred_subscribers(new, 2);
517 if (new->user != old->user) 515 if (new->user != old->user)
diff --git a/kernel/debug/debug_core.c b/kernel/debug/debug_core.c
index bad6786dee88..0d7c08784efb 100644
--- a/kernel/debug/debug_core.c
+++ b/kernel/debug/debug_core.c
@@ -51,7 +51,7 @@
51 51
52#include <asm/cacheflush.h> 52#include <asm/cacheflush.h>
53#include <asm/byteorder.h> 53#include <asm/byteorder.h>
54#include <asm/atomic.h> 54#include <linux/atomic.h>
55#include <asm/system.h> 55#include <asm/system.h>
56 56
57#include "debug_core.h" 57#include "debug_core.h"
diff --git a/kernel/debug/gdbstub.c b/kernel/debug/gdbstub.c
index a11db956dd62..34872482315e 100644
--- a/kernel/debug/gdbstub.c
+++ b/kernel/debug/gdbstub.c
@@ -42,6 +42,8 @@
42/* Our I/O buffers. */ 42/* Our I/O buffers. */
43static char remcom_in_buffer[BUFMAX]; 43static char remcom_in_buffer[BUFMAX];
44static char remcom_out_buffer[BUFMAX]; 44static char remcom_out_buffer[BUFMAX];
45static int gdbstub_use_prev_in_buf;
46static int gdbstub_prev_in_buf_pos;
45 47
46/* Storage for the registers, in GDB format. */ 48/* Storage for the registers, in GDB format. */
47static unsigned long gdb_regs[(NUMREGBYTES + 49static unsigned long gdb_regs[(NUMREGBYTES +
@@ -58,6 +60,13 @@ static int gdbstub_read_wait(void)
58 int ret = -1; 60 int ret = -1;
59 int i; 61 int i;
60 62
63 if (unlikely(gdbstub_use_prev_in_buf)) {
64 if (gdbstub_prev_in_buf_pos < gdbstub_use_prev_in_buf)
65 return remcom_in_buffer[gdbstub_prev_in_buf_pos++];
66 else
67 gdbstub_use_prev_in_buf = 0;
68 }
69
61 /* poll any additional I/O interfaces that are defined */ 70 /* poll any additional I/O interfaces that are defined */
62 while (ret < 0) 71 while (ret < 0)
63 for (i = 0; kdb_poll_funcs[i] != NULL; i++) { 72 for (i = 0; kdb_poll_funcs[i] != NULL; i++) {
@@ -109,7 +118,6 @@ static void get_packet(char *buffer)
109 buffer[count] = ch; 118 buffer[count] = ch;
110 count = count + 1; 119 count = count + 1;
111 } 120 }
112 buffer[count] = 0;
113 121
114 if (ch == '#') { 122 if (ch == '#') {
115 xmitcsum = hex_to_bin(gdbstub_read_wait()) << 4; 123 xmitcsum = hex_to_bin(gdbstub_read_wait()) << 4;
@@ -124,6 +132,7 @@ static void get_packet(char *buffer)
124 if (dbg_io_ops->flush) 132 if (dbg_io_ops->flush)
125 dbg_io_ops->flush(); 133 dbg_io_ops->flush();
126 } 134 }
135 buffer[count] = 0;
127 } while (checksum != xmitcsum); 136 } while (checksum != xmitcsum);
128} 137}
129 138
@@ -1082,12 +1091,11 @@ int gdbstub_state(struct kgdb_state *ks, char *cmd)
1082 case 'c': 1091 case 'c':
1083 strcpy(remcom_in_buffer, cmd); 1092 strcpy(remcom_in_buffer, cmd);
1084 return 0; 1093 return 0;
1085 case '?': 1094 case '$':
1086 gdb_cmd_status(ks); 1095 strcpy(remcom_in_buffer, cmd);
1087 break; 1096 gdbstub_use_prev_in_buf = strlen(remcom_in_buffer);
1088 case '\0': 1097 gdbstub_prev_in_buf_pos = 0;
1089 strcpy(remcom_out_buffer, ""); 1098 return 0;
1090 break;
1091 } 1099 }
1092 dbg_io_ops->write_char('+'); 1100 dbg_io_ops->write_char('+');
1093 put_packet(remcom_out_buffer); 1101 put_packet(remcom_out_buffer);
diff --git a/kernel/debug/kdb/kdb_bt.c b/kernel/debug/kdb/kdb_bt.c
index 2f62fe85f16a..7179eac7b41c 100644
--- a/kernel/debug/kdb/kdb_bt.c
+++ b/kernel/debug/kdb/kdb_bt.c
@@ -112,9 +112,8 @@ kdb_bt(int argc, const char **argv)
112 unsigned long addr; 112 unsigned long addr;
113 long offset; 113 long offset;
114 114
115 kdbgetintenv("BTARGS", &argcount); /* Arguments to print */ 115 /* Prompt after each proc in bta */
116 kdbgetintenv("BTAPROMPT", &btaprompt); /* Prompt after each 116 kdbgetintenv("BTAPROMPT", &btaprompt);
117 * proc in bta */
118 117
119 if (strcmp(argv[0], "bta") == 0) { 118 if (strcmp(argv[0], "bta") == 0) {
120 struct task_struct *g, *p; 119 struct task_struct *g, *p;
diff --git a/kernel/debug/kdb/kdb_cmds b/kernel/debug/kdb/kdb_cmds
index 56c88e4db309..9834ad303ab6 100644
--- a/kernel/debug/kdb/kdb_cmds
+++ b/kernel/debug/kdb/kdb_cmds
@@ -18,16 +18,12 @@ defcmd dumpcommon "" "Common kdb debugging"
18endefcmd 18endefcmd
19 19
20defcmd dumpall "" "First line debugging" 20defcmd dumpall "" "First line debugging"
21 set BTSYMARG 1
22 set BTARGS 9
23 pid R 21 pid R
24 -dumpcommon 22 -dumpcommon
25 -bta 23 -bta
26endefcmd 24endefcmd
27 25
28defcmd dumpcpu "" "Same as dumpall but only tasks on cpus" 26defcmd dumpcpu "" "Same as dumpall but only tasks on cpus"
29 set BTSYMARG 1
30 set BTARGS 9
31 pid R 27 pid R
32 -dumpcommon 28 -dumpcommon
33 -btc 29 -btc
diff --git a/kernel/debug/kdb/kdb_debugger.c b/kernel/debug/kdb/kdb_debugger.c
index dd0b1b7dd02c..d9ca9aa481ec 100644
--- a/kernel/debug/kdb/kdb_debugger.c
+++ b/kernel/debug/kdb/kdb_debugger.c
@@ -30,6 +30,8 @@ EXPORT_SYMBOL_GPL(kdb_poll_funcs);
30int kdb_poll_idx = 1; 30int kdb_poll_idx = 1;
31EXPORT_SYMBOL_GPL(kdb_poll_idx); 31EXPORT_SYMBOL_GPL(kdb_poll_idx);
32 32
33static struct kgdb_state *kdb_ks;
34
33int kdb_stub(struct kgdb_state *ks) 35int kdb_stub(struct kgdb_state *ks)
34{ 36{
35 int error = 0; 37 int error = 0;
@@ -39,6 +41,7 @@ int kdb_stub(struct kgdb_state *ks)
39 kdb_dbtrap_t db_result = KDB_DB_NOBPT; 41 kdb_dbtrap_t db_result = KDB_DB_NOBPT;
40 int i; 42 int i;
41 43
44 kdb_ks = ks;
42 if (KDB_STATE(REENTRY)) { 45 if (KDB_STATE(REENTRY)) {
43 reason = KDB_REASON_SWITCH; 46 reason = KDB_REASON_SWITCH;
44 KDB_STATE_CLEAR(REENTRY); 47 KDB_STATE_CLEAR(REENTRY);
@@ -123,20 +126,8 @@ int kdb_stub(struct kgdb_state *ks)
123 KDB_STATE_CLEAR(PAGER); 126 KDB_STATE_CLEAR(PAGER);
124 kdbnearsym_cleanup(); 127 kdbnearsym_cleanup();
125 if (error == KDB_CMD_KGDB) { 128 if (error == KDB_CMD_KGDB) {
126 if (KDB_STATE(DOING_KGDB) || KDB_STATE(DOING_KGDB2)) { 129 if (KDB_STATE(DOING_KGDB))
127 /*
128 * This inteface glue which allows kdb to transition in into
129 * the gdb stub. In order to do this the '?' or '' gdb serial
130 * packet response is processed here. And then control is
131 * passed to the gdbstub.
132 */
133 if (KDB_STATE(DOING_KGDB))
134 gdbstub_state(ks, "?");
135 else
136 gdbstub_state(ks, "");
137 KDB_STATE_CLEAR(DOING_KGDB); 130 KDB_STATE_CLEAR(DOING_KGDB);
138 KDB_STATE_CLEAR(DOING_KGDB2);
139 }
140 return DBG_PASS_EVENT; 131 return DBG_PASS_EVENT;
141 } 132 }
142 kdb_bp_install(ks->linux_regs); 133 kdb_bp_install(ks->linux_regs);
@@ -166,3 +157,7 @@ int kdb_stub(struct kgdb_state *ks)
166 return kgdb_info[ks->cpu].ret_state; 157 return kgdb_info[ks->cpu].ret_state;
167} 158}
168 159
160void kdb_gdb_state_pass(char *buf)
161{
162 gdbstub_state(kdb_ks, buf);
163}
diff --git a/kernel/debug/kdb/kdb_io.c b/kernel/debug/kdb/kdb_io.c
index 96fdaac46a80..4802eb5840e1 100644
--- a/kernel/debug/kdb/kdb_io.c
+++ b/kernel/debug/kdb/kdb_io.c
@@ -31,15 +31,21 @@ char kdb_prompt_str[CMD_BUFLEN];
31 31
32int kdb_trap_printk; 32int kdb_trap_printk;
33 33
34static void kgdb_transition_check(char *buffer) 34static int kgdb_transition_check(char *buffer)
35{ 35{
36 int slen = strlen(buffer); 36 if (buffer[0] != '+' && buffer[0] != '$') {
37 if (strncmp(buffer, "$?#3f", slen) != 0 &&
38 strncmp(buffer, "$qSupported#37", slen) != 0 &&
39 strncmp(buffer, "+$qSupported#37", slen) != 0) {
40 KDB_STATE_SET(KGDB_TRANS); 37 KDB_STATE_SET(KGDB_TRANS);
41 kdb_printf("%s", buffer); 38 kdb_printf("%s", buffer);
39 } else {
40 int slen = strlen(buffer);
41 if (slen > 3 && buffer[slen - 3] == '#') {
42 kdb_gdb_state_pass(buffer);
43 strcpy(buffer, "kgdb");
44 KDB_STATE_SET(DOING_KGDB);
45 return 1;
46 }
42 } 47 }
48 return 0;
43} 49}
44 50
45static int kdb_read_get_key(char *buffer, size_t bufsize) 51static int kdb_read_get_key(char *buffer, size_t bufsize)
@@ -251,6 +257,10 @@ poll_again:
251 case 13: /* enter */ 257 case 13: /* enter */
252 *lastchar++ = '\n'; 258 *lastchar++ = '\n';
253 *lastchar++ = '\0'; 259 *lastchar++ = '\0';
260 if (!KDB_STATE(KGDB_TRANS)) {
261 KDB_STATE_SET(KGDB_TRANS);
262 kdb_printf("%s", buffer);
263 }
254 kdb_printf("\n"); 264 kdb_printf("\n");
255 return buffer; 265 return buffer;
256 case 4: /* Del */ 266 case 4: /* Del */
@@ -382,22 +392,26 @@ poll_again:
382 * printed characters if we think that 392 * printed characters if we think that
383 * kgdb is connecting, until the check 393 * kgdb is connecting, until the check
384 * fails */ 394 * fails */
385 if (!KDB_STATE(KGDB_TRANS)) 395 if (!KDB_STATE(KGDB_TRANS)) {
386 kgdb_transition_check(buffer); 396 if (kgdb_transition_check(buffer))
387 else 397 return buffer;
398 } else {
388 kdb_printf("%c", key); 399 kdb_printf("%c", key);
400 }
389 } 401 }
390 /* Special escape to kgdb */ 402 /* Special escape to kgdb */
391 if (lastchar - buffer >= 5 && 403 if (lastchar - buffer >= 5 &&
392 strcmp(lastchar - 5, "$?#3f") == 0) { 404 strcmp(lastchar - 5, "$?#3f") == 0) {
405 kdb_gdb_state_pass(lastchar - 5);
393 strcpy(buffer, "kgdb"); 406 strcpy(buffer, "kgdb");
394 KDB_STATE_SET(DOING_KGDB); 407 KDB_STATE_SET(DOING_KGDB);
395 return buffer; 408 return buffer;
396 } 409 }
397 if (lastchar - buffer >= 14 && 410 if (lastchar - buffer >= 11 &&
398 strcmp(lastchar - 14, "$qSupported#37") == 0) { 411 strcmp(lastchar - 11, "$qSupported") == 0) {
412 kdb_gdb_state_pass(lastchar - 11);
399 strcpy(buffer, "kgdb"); 413 strcpy(buffer, "kgdb");
400 KDB_STATE_SET(DOING_KGDB2); 414 KDB_STATE_SET(DOING_KGDB);
401 return buffer; 415 return buffer;
402 } 416 }
403 } 417 }
diff --git a/kernel/debug/kdb/kdb_main.c b/kernel/debug/kdb/kdb_main.c
index be14779bcef6..63786e71a3cd 100644
--- a/kernel/debug/kdb/kdb_main.c
+++ b/kernel/debug/kdb/kdb_main.c
@@ -145,7 +145,6 @@ static char *__env[] = {
145#endif 145#endif
146 "RADIX=16", 146 "RADIX=16",
147 "MDCOUNT=8", /* lines of md output */ 147 "MDCOUNT=8", /* lines of md output */
148 "BTARGS=9", /* 9 possible args in bt */
149 KDB_PLATFORM_ENV, 148 KDB_PLATFORM_ENV,
150 "DTABCOUNT=30", 149 "DTABCOUNT=30",
151 "NOSECT=1", 150 "NOSECT=1",
@@ -172,6 +171,7 @@ static char *__env[] = {
172 (char *)0, 171 (char *)0,
173 (char *)0, 172 (char *)0,
174 (char *)0, 173 (char *)0,
174 (char *)0,
175}; 175};
176 176
177static const int __nenv = (sizeof(__env) / sizeof(char *)); 177static const int __nenv = (sizeof(__env) / sizeof(char *));
@@ -1386,7 +1386,7 @@ int kdb_main_loop(kdb_reason_t reason, kdb_reason_t reason2, int error,
1386 } 1386 }
1387 1387
1388 if (result == KDB_CMD_KGDB) { 1388 if (result == KDB_CMD_KGDB) {
1389 if (!(KDB_STATE(DOING_KGDB) || KDB_STATE(DOING_KGDB2))) 1389 if (!KDB_STATE(DOING_KGDB))
1390 kdb_printf("Entering please attach debugger " 1390 kdb_printf("Entering please attach debugger "
1391 "or use $D#44+ or $3#33\n"); 1391 "or use $D#44+ or $3#33\n");
1392 break; 1392 break;
diff --git a/kernel/debug/kdb/kdb_private.h b/kernel/debug/kdb/kdb_private.h
index 35d69ed1dfb5..e381d105b40b 100644
--- a/kernel/debug/kdb/kdb_private.h
+++ b/kernel/debug/kdb/kdb_private.h
@@ -21,7 +21,6 @@
21#define KDB_CMD_SS (-1003) 21#define KDB_CMD_SS (-1003)
22#define KDB_CMD_SSB (-1004) 22#define KDB_CMD_SSB (-1004)
23#define KDB_CMD_KGDB (-1005) 23#define KDB_CMD_KGDB (-1005)
24#define KDB_CMD_KGDB2 (-1006)
25 24
26/* Internal debug flags */ 25/* Internal debug flags */
27#define KDB_DEBUG_FLAG_BP 0x0002 /* Breakpoint subsystem debug */ 26#define KDB_DEBUG_FLAG_BP 0x0002 /* Breakpoint subsystem debug */
@@ -146,7 +145,6 @@ extern int kdb_state;
146 * keyboard on this cpu */ 145 * keyboard on this cpu */
147#define KDB_STATE_KEXEC 0x00040000 /* kexec issued */ 146#define KDB_STATE_KEXEC 0x00040000 /* kexec issued */
148#define KDB_STATE_DOING_KGDB 0x00080000 /* kgdb enter now issued */ 147#define KDB_STATE_DOING_KGDB 0x00080000 /* kgdb enter now issued */
149#define KDB_STATE_DOING_KGDB2 0x00100000 /* kgdb enter now issued */
150#define KDB_STATE_KGDB_TRANS 0x00200000 /* Transition to kgdb */ 148#define KDB_STATE_KGDB_TRANS 0x00200000 /* Transition to kgdb */
151#define KDB_STATE_ARCH 0xff000000 /* Reserved for arch 149#define KDB_STATE_ARCH 0xff000000 /* Reserved for arch
152 * specific use */ 150 * specific use */
@@ -218,6 +216,7 @@ extern void kdb_print_nameval(const char *name, unsigned long val);
218extern void kdb_send_sig_info(struct task_struct *p, struct siginfo *info); 216extern void kdb_send_sig_info(struct task_struct *p, struct siginfo *info);
219extern void kdb_meminfo_proc_show(void); 217extern void kdb_meminfo_proc_show(void);
220extern char *kdb_getstr(char *, size_t, char *); 218extern char *kdb_getstr(char *, size_t, char *);
219extern void kdb_gdb_state_pass(char *buf);
221 220
222/* Defines for kdb_symbol_print */ 221/* Defines for kdb_symbol_print */
223#define KDB_SP_SPACEB 0x0001 /* Space before string */ 222#define KDB_SP_SPACEB 0x0001 /* Space before string */
diff --git a/kernel/delayacct.c b/kernel/delayacct.c
index ead9b610aa71..418b3f7053aa 100644
--- a/kernel/delayacct.c
+++ b/kernel/delayacct.c
@@ -19,8 +19,10 @@
19#include <linux/time.h> 19#include <linux/time.h>
20#include <linux/sysctl.h> 20#include <linux/sysctl.h>
21#include <linux/delayacct.h> 21#include <linux/delayacct.h>
22#include <linux/module.h>
22 23
23int delayacct_on __read_mostly = 1; /* Delay accounting turned on/off */ 24int delayacct_on __read_mostly = 1; /* Delay accounting turned on/off */
25EXPORT_SYMBOL_GPL(delayacct_on);
24struct kmem_cache *delayacct_cache; 26struct kmem_cache *delayacct_cache;
25 27
26static int __init delayacct_setup_disable(char *str) 28static int __init delayacct_setup_disable(char *str)
diff --git a/kernel/events/core.c b/kernel/events/core.c
index b8785e26ee1c..0f857782d06f 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -399,14 +399,54 @@ void perf_cgroup_switch(struct task_struct *task, int mode)
399 local_irq_restore(flags); 399 local_irq_restore(flags);
400} 400}
401 401
402static inline void perf_cgroup_sched_out(struct task_struct *task) 402static inline void perf_cgroup_sched_out(struct task_struct *task,
403 struct task_struct *next)
403{ 404{
404 perf_cgroup_switch(task, PERF_CGROUP_SWOUT); 405 struct perf_cgroup *cgrp1;
406 struct perf_cgroup *cgrp2 = NULL;
407
408 /*
409 * we come here when we know perf_cgroup_events > 0
410 */
411 cgrp1 = perf_cgroup_from_task(task);
412
413 /*
414 * next is NULL when called from perf_event_enable_on_exec()
415 * that will systematically cause a cgroup_switch()
416 */
417 if (next)
418 cgrp2 = perf_cgroup_from_task(next);
419
420 /*
421 * only schedule out current cgroup events if we know
422 * that we are switching to a different cgroup. Otherwise,
423 * do no touch the cgroup events.
424 */
425 if (cgrp1 != cgrp2)
426 perf_cgroup_switch(task, PERF_CGROUP_SWOUT);
405} 427}
406 428
407static inline void perf_cgroup_sched_in(struct task_struct *task) 429static inline void perf_cgroup_sched_in(struct task_struct *prev,
430 struct task_struct *task)
408{ 431{
409 perf_cgroup_switch(task, PERF_CGROUP_SWIN); 432 struct perf_cgroup *cgrp1;
433 struct perf_cgroup *cgrp2 = NULL;
434
435 /*
436 * we come here when we know perf_cgroup_events > 0
437 */
438 cgrp1 = perf_cgroup_from_task(task);
439
440 /* prev can never be NULL */
441 cgrp2 = perf_cgroup_from_task(prev);
442
443 /*
444 * only need to schedule in cgroup events if we are changing
445 * cgroup during ctxsw. Cgroup events were not scheduled
446 * out of ctxsw out if that was not the case.
447 */
448 if (cgrp1 != cgrp2)
449 perf_cgroup_switch(task, PERF_CGROUP_SWIN);
410} 450}
411 451
412static inline int perf_cgroup_connect(int fd, struct perf_event *event, 452static inline int perf_cgroup_connect(int fd, struct perf_event *event,
@@ -518,11 +558,13 @@ static inline void update_cgrp_time_from_cpuctx(struct perf_cpu_context *cpuctx)
518{ 558{
519} 559}
520 560
521static inline void perf_cgroup_sched_out(struct task_struct *task) 561static inline void perf_cgroup_sched_out(struct task_struct *task,
562 struct task_struct *next)
522{ 563{
523} 564}
524 565
525static inline void perf_cgroup_sched_in(struct task_struct *task) 566static inline void perf_cgroup_sched_in(struct task_struct *prev,
567 struct task_struct *task)
526{ 568{
527} 569}
528 570
@@ -1988,7 +2030,7 @@ void __perf_event_task_sched_out(struct task_struct *task,
1988 * cgroup event are system-wide mode only 2030 * cgroup event are system-wide mode only
1989 */ 2031 */
1990 if (atomic_read(&__get_cpu_var(perf_cgroup_events))) 2032 if (atomic_read(&__get_cpu_var(perf_cgroup_events)))
1991 perf_cgroup_sched_out(task); 2033 perf_cgroup_sched_out(task, next);
1992} 2034}
1993 2035
1994static void task_ctx_sched_out(struct perf_event_context *ctx) 2036static void task_ctx_sched_out(struct perf_event_context *ctx)
@@ -2153,7 +2195,8 @@ static void perf_event_context_sched_in(struct perf_event_context *ctx,
2153 * accessing the event control register. If a NMI hits, then it will 2195 * accessing the event control register. If a NMI hits, then it will
2154 * keep the event running. 2196 * keep the event running.
2155 */ 2197 */
2156void __perf_event_task_sched_in(struct task_struct *task) 2198void __perf_event_task_sched_in(struct task_struct *prev,
2199 struct task_struct *task)
2157{ 2200{
2158 struct perf_event_context *ctx; 2201 struct perf_event_context *ctx;
2159 int ctxn; 2202 int ctxn;
@@ -2171,7 +2214,7 @@ void __perf_event_task_sched_in(struct task_struct *task)
2171 * cgroup event are system-wide mode only 2214 * cgroup event are system-wide mode only
2172 */ 2215 */
2173 if (atomic_read(&__get_cpu_var(perf_cgroup_events))) 2216 if (atomic_read(&__get_cpu_var(perf_cgroup_events)))
2174 perf_cgroup_sched_in(task); 2217 perf_cgroup_sched_in(prev, task);
2175} 2218}
2176 2219
2177static u64 perf_calculate_period(struct perf_event *event, u64 nsec, u64 count) 2220static u64 perf_calculate_period(struct perf_event *event, u64 nsec, u64 count)
@@ -2427,7 +2470,7 @@ static void perf_event_enable_on_exec(struct perf_event_context *ctx)
2427 * ctxswin cgroup events which are already scheduled 2470 * ctxswin cgroup events which are already scheduled
2428 * in. 2471 * in.
2429 */ 2472 */
2430 perf_cgroup_sched_out(current); 2473 perf_cgroup_sched_out(current, NULL);
2431 2474
2432 raw_spin_lock(&ctx->lock); 2475 raw_spin_lock(&ctx->lock);
2433 task_ctx_sched_out(ctx); 2476 task_ctx_sched_out(ctx);
@@ -3353,8 +3396,8 @@ static int perf_event_index(struct perf_event *event)
3353} 3396}
3354 3397
3355static void calc_timer_values(struct perf_event *event, 3398static void calc_timer_values(struct perf_event *event,
3356 u64 *running, 3399 u64 *enabled,
3357 u64 *enabled) 3400 u64 *running)
3358{ 3401{
3359 u64 now, ctx_time; 3402 u64 now, ctx_time;
3360 3403
diff --git a/kernel/exit.c b/kernel/exit.c
index 73bb192a3d32..2913b3509d42 100644
--- a/kernel/exit.c
+++ b/kernel/exit.c
@@ -85,7 +85,6 @@ static void __exit_signal(struct task_struct *tsk)
85 struct tty_struct *uninitialized_var(tty); 85 struct tty_struct *uninitialized_var(tty);
86 86
87 sighand = rcu_dereference_check(tsk->sighand, 87 sighand = rcu_dereference_check(tsk->sighand,
88 rcu_read_lock_held() ||
89 lockdep_tasklist_lock_is_held()); 88 lockdep_tasklist_lock_is_held());
90 spin_lock(&sighand->siglock); 89 spin_lock(&sighand->siglock);
91 90
@@ -898,7 +897,6 @@ NORET_TYPE void do_exit(long code)
898 897
899 profile_task_exit(tsk); 898 profile_task_exit(tsk);
900 899
901 WARN_ON(atomic_read(&tsk->fs_excl));
902 WARN_ON(blk_needs_flush_plug(tsk)); 900 WARN_ON(blk_needs_flush_plug(tsk));
903 901
904 if (unlikely(in_interrupt())) 902 if (unlikely(in_interrupt()))
@@ -982,6 +980,7 @@ NORET_TYPE void do_exit(long code)
982 trace_sched_process_exit(tsk); 980 trace_sched_process_exit(tsk);
983 981
984 exit_sem(tsk); 982 exit_sem(tsk);
983 exit_shm(tsk);
985 exit_files(tsk); 984 exit_files(tsk);
986 exit_fs(tsk); 985 exit_fs(tsk);
987 check_stack_usage(); 986 check_stack_usage();
diff --git a/kernel/fork.c b/kernel/fork.c
index aeae5b11b62e..8e6b6f4fb272 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -80,7 +80,7 @@
80 * Protected counters by write_lock_irq(&tasklist_lock) 80 * Protected counters by write_lock_irq(&tasklist_lock)
81 */ 81 */
82unsigned long total_forks; /* Handle normal Linux uptimes. */ 82unsigned long total_forks; /* Handle normal Linux uptimes. */
83int nr_threads; /* The idle threads do not count.. */ 83int nr_threads; /* The idle threads do not count.. */
84 84
85int max_threads; /* tunable limit on nr_threads */ 85int max_threads; /* tunable limit on nr_threads */
86 86
@@ -232,7 +232,7 @@ void __init fork_init(unsigned long mempages)
232 /* 232 /*
233 * we need to allow at least 20 threads to boot a system 233 * we need to allow at least 20 threads to boot a system
234 */ 234 */
235 if(max_threads < 20) 235 if (max_threads < 20)
236 max_threads = 20; 236 max_threads = 20;
237 237
238 init_task.signal->rlim[RLIMIT_NPROC].rlim_cur = max_threads/2; 238 init_task.signal->rlim[RLIMIT_NPROC].rlim_cur = max_threads/2;
@@ -268,7 +268,7 @@ static struct task_struct *dup_task_struct(struct task_struct *orig)
268 return NULL; 268 return NULL;
269 } 269 }
270 270
271 err = arch_dup_task_struct(tsk, orig); 271 err = arch_dup_task_struct(tsk, orig);
272 if (err) 272 if (err)
273 goto out; 273 goto out;
274 274
@@ -288,9 +288,11 @@ static struct task_struct *dup_task_struct(struct task_struct *orig)
288 tsk->stack_canary = get_random_int(); 288 tsk->stack_canary = get_random_int();
289#endif 289#endif
290 290
291 /* One for us, one for whoever does the "release_task()" (usually parent) */ 291 /*
292 atomic_set(&tsk->usage,2); 292 * One for us, one for whoever does the "release_task()" (usually
293 atomic_set(&tsk->fs_excl, 0); 293 * parent)
294 */
295 atomic_set(&tsk->usage, 2);
294#ifdef CONFIG_BLK_DEV_IO_TRACE 296#ifdef CONFIG_BLK_DEV_IO_TRACE
295 tsk->btrace_seq = 0; 297 tsk->btrace_seq = 0;
296#endif 298#endif
@@ -438,7 +440,7 @@ fail_nomem:
438 goto out; 440 goto out;
439} 441}
440 442
441static inline int mm_alloc_pgd(struct mm_struct * mm) 443static inline int mm_alloc_pgd(struct mm_struct *mm)
442{ 444{
443 mm->pgd = pgd_alloc(mm); 445 mm->pgd = pgd_alloc(mm);
444 if (unlikely(!mm->pgd)) 446 if (unlikely(!mm->pgd))
@@ -446,7 +448,7 @@ static inline int mm_alloc_pgd(struct mm_struct * mm)
446 return 0; 448 return 0;
447} 449}
448 450
449static inline void mm_free_pgd(struct mm_struct * mm) 451static inline void mm_free_pgd(struct mm_struct *mm)
450{ 452{
451 pgd_free(mm, mm->pgd); 453 pgd_free(mm, mm->pgd);
452} 454}
@@ -483,7 +485,7 @@ static void mm_init_aio(struct mm_struct *mm)
483#endif 485#endif
484} 486}
485 487
486static struct mm_struct * mm_init(struct mm_struct * mm, struct task_struct *p) 488static struct mm_struct *mm_init(struct mm_struct *mm, struct task_struct *p)
487{ 489{
488 atomic_set(&mm->mm_users, 1); 490 atomic_set(&mm->mm_users, 1);
489 atomic_set(&mm->mm_count, 1); 491 atomic_set(&mm->mm_count, 1);
@@ -514,9 +516,9 @@ static struct mm_struct * mm_init(struct mm_struct * mm, struct task_struct *p)
514/* 516/*
515 * Allocate and initialize an mm_struct. 517 * Allocate and initialize an mm_struct.
516 */ 518 */
517struct mm_struct * mm_alloc(void) 519struct mm_struct *mm_alloc(void)
518{ 520{
519 struct mm_struct * mm; 521 struct mm_struct *mm;
520 522
521 mm = allocate_mm(); 523 mm = allocate_mm();
522 if (!mm) 524 if (!mm)
@@ -584,7 +586,7 @@ void added_exe_file_vma(struct mm_struct *mm)
584void removed_exe_file_vma(struct mm_struct *mm) 586void removed_exe_file_vma(struct mm_struct *mm)
585{ 587{
586 mm->num_exe_file_vmas--; 588 mm->num_exe_file_vmas--;
587 if ((mm->num_exe_file_vmas == 0) && mm->exe_file){ 589 if ((mm->num_exe_file_vmas == 0) && mm->exe_file) {
588 fput(mm->exe_file); 590 fput(mm->exe_file);
589 mm->exe_file = NULL; 591 mm->exe_file = NULL;
590 } 592 }
@@ -776,9 +778,9 @@ fail_nocontext:
776 return NULL; 778 return NULL;
777} 779}
778 780
779static int copy_mm(unsigned long clone_flags, struct task_struct * tsk) 781static int copy_mm(unsigned long clone_flags, struct task_struct *tsk)
780{ 782{
781 struct mm_struct * mm, *oldmm; 783 struct mm_struct *mm, *oldmm;
782 int retval; 784 int retval;
783 785
784 tsk->min_flt = tsk->maj_flt = 0; 786 tsk->min_flt = tsk->maj_flt = 0;
@@ -845,7 +847,7 @@ static int copy_fs(unsigned long clone_flags, struct task_struct *tsk)
845 return 0; 847 return 0;
846} 848}
847 849
848static int copy_files(unsigned long clone_flags, struct task_struct * tsk) 850static int copy_files(unsigned long clone_flags, struct task_struct *tsk)
849{ 851{
850 struct files_struct *oldf, *newf; 852 struct files_struct *oldf, *newf;
851 int error = 0; 853 int error = 0;
@@ -1109,6 +1111,7 @@ static struct task_struct *copy_process(unsigned long clone_flags,
1109 p->real_cred->user != INIT_USER) 1111 p->real_cred->user != INIT_USER)
1110 goto bad_fork_free; 1112 goto bad_fork_free;
1111 } 1113 }
1114 current->flags &= ~PF_NPROC_EXCEEDED;
1112 1115
1113 retval = copy_creds(p, clone_flags); 1116 retval = copy_creds(p, clone_flags);
1114 if (retval < 0) 1117 if (retval < 0)
@@ -1167,13 +1170,17 @@ static struct task_struct *copy_process(unsigned long clone_flags,
1167 cgroup_fork(p); 1170 cgroup_fork(p);
1168#ifdef CONFIG_NUMA 1171#ifdef CONFIG_NUMA
1169 p->mempolicy = mpol_dup(p->mempolicy); 1172 p->mempolicy = mpol_dup(p->mempolicy);
1170 if (IS_ERR(p->mempolicy)) { 1173 if (IS_ERR(p->mempolicy)) {
1171 retval = PTR_ERR(p->mempolicy); 1174 retval = PTR_ERR(p->mempolicy);
1172 p->mempolicy = NULL; 1175 p->mempolicy = NULL;
1173 goto bad_fork_cleanup_cgroup; 1176 goto bad_fork_cleanup_cgroup;
1174 } 1177 }
1175 mpol_fix_fork_child_flag(p); 1178 mpol_fix_fork_child_flag(p);
1176#endif 1179#endif
1180#ifdef CONFIG_CPUSETS
1181 p->cpuset_mem_spread_rotor = NUMA_NO_NODE;
1182 p->cpuset_slab_spread_rotor = NUMA_NO_NODE;
1183#endif
1177#ifdef CONFIG_TRACE_IRQFLAGS 1184#ifdef CONFIG_TRACE_IRQFLAGS
1178 p->irq_events = 0; 1185 p->irq_events = 0;
1179#ifdef __ARCH_WANT_INTERRUPTS_ON_CTXSW 1186#ifdef __ARCH_WANT_INTERRUPTS_ON_CTXSW
@@ -1213,25 +1220,33 @@ static struct task_struct *copy_process(unsigned long clone_flags,
1213 retval = perf_event_init_task(p); 1220 retval = perf_event_init_task(p);
1214 if (retval) 1221 if (retval)
1215 goto bad_fork_cleanup_policy; 1222 goto bad_fork_cleanup_policy;
1216 1223 retval = audit_alloc(p);
1217 if ((retval = audit_alloc(p))) 1224 if (retval)
1218 goto bad_fork_cleanup_policy; 1225 goto bad_fork_cleanup_policy;
1219 /* copy all the process information */ 1226 /* copy all the process information */
1220 if ((retval = copy_semundo(clone_flags, p))) 1227 retval = copy_semundo(clone_flags, p);
1228 if (retval)
1221 goto bad_fork_cleanup_audit; 1229 goto bad_fork_cleanup_audit;
1222 if ((retval = copy_files(clone_flags, p))) 1230 retval = copy_files(clone_flags, p);
1231 if (retval)
1223 goto bad_fork_cleanup_semundo; 1232 goto bad_fork_cleanup_semundo;
1224 if ((retval = copy_fs(clone_flags, p))) 1233 retval = copy_fs(clone_flags, p);
1234 if (retval)
1225 goto bad_fork_cleanup_files; 1235 goto bad_fork_cleanup_files;
1226 if ((retval = copy_sighand(clone_flags, p))) 1236 retval = copy_sighand(clone_flags, p);
1237 if (retval)
1227 goto bad_fork_cleanup_fs; 1238 goto bad_fork_cleanup_fs;
1228 if ((retval = copy_signal(clone_flags, p))) 1239 retval = copy_signal(clone_flags, p);
1240 if (retval)
1229 goto bad_fork_cleanup_sighand; 1241 goto bad_fork_cleanup_sighand;
1230 if ((retval = copy_mm(clone_flags, p))) 1242 retval = copy_mm(clone_flags, p);
1243 if (retval)
1231 goto bad_fork_cleanup_signal; 1244 goto bad_fork_cleanup_signal;
1232 if ((retval = copy_namespaces(clone_flags, p))) 1245 retval = copy_namespaces(clone_flags, p);
1246 if (retval)
1233 goto bad_fork_cleanup_mm; 1247 goto bad_fork_cleanup_mm;
1234 if ((retval = copy_io(clone_flags, p))) 1248 retval = copy_io(clone_flags, p);
1249 if (retval)
1235 goto bad_fork_cleanup_namespaces; 1250 goto bad_fork_cleanup_namespaces;
1236 retval = copy_thread(clone_flags, stack_start, stack_size, p, regs); 1251 retval = copy_thread(clone_flags, stack_start, stack_size, p, regs);
1237 if (retval) 1252 if (retval)
@@ -1253,7 +1268,7 @@ static struct task_struct *copy_process(unsigned long clone_flags,
1253 /* 1268 /*
1254 * Clear TID on mm_release()? 1269 * Clear TID on mm_release()?
1255 */ 1270 */
1256 p->clear_child_tid = (clone_flags & CLONE_CHILD_CLEARTID) ? child_tidptr: NULL; 1271 p->clear_child_tid = (clone_flags & CLONE_CHILD_CLEARTID) ? child_tidptr : NULL;
1257#ifdef CONFIG_BLOCK 1272#ifdef CONFIG_BLOCK
1258 p->plug = NULL; 1273 p->plug = NULL;
1259#endif 1274#endif
@@ -1321,7 +1336,7 @@ static struct task_struct *copy_process(unsigned long clone_flags,
1321 * it's process group. 1336 * it's process group.
1322 * A fatal signal pending means that current will exit, so the new 1337 * A fatal signal pending means that current will exit, so the new
1323 * thread can't slip out of an OOM kill (or normal SIGKILL). 1338 * thread can't slip out of an OOM kill (or normal SIGKILL).
1324 */ 1339 */
1325 recalc_sigpending(); 1340 recalc_sigpending();
1326 if (signal_pending(current)) { 1341 if (signal_pending(current)) {
1327 spin_unlock(&current->sighand->siglock); 1342 spin_unlock(&current->sighand->siglock);
@@ -1682,12 +1697,14 @@ SYSCALL_DEFINE1(unshare, unsigned long, unshare_flags)
1682 */ 1697 */
1683 if (unshare_flags & (CLONE_NEWIPC|CLONE_SYSVSEM)) 1698 if (unshare_flags & (CLONE_NEWIPC|CLONE_SYSVSEM))
1684 do_sysvsem = 1; 1699 do_sysvsem = 1;
1685 if ((err = unshare_fs(unshare_flags, &new_fs))) 1700 err = unshare_fs(unshare_flags, &new_fs);
1701 if (err)
1686 goto bad_unshare_out; 1702 goto bad_unshare_out;
1687 if ((err = unshare_fd(unshare_flags, &new_fd))) 1703 err = unshare_fd(unshare_flags, &new_fd);
1704 if (err)
1688 goto bad_unshare_cleanup_fs; 1705 goto bad_unshare_cleanup_fs;
1689 if ((err = unshare_nsproxy_namespaces(unshare_flags, &new_nsproxy, 1706 err = unshare_nsproxy_namespaces(unshare_flags, &new_nsproxy, new_fs);
1690 new_fs))) 1707 if (err)
1691 goto bad_unshare_cleanup_fd; 1708 goto bad_unshare_cleanup_fd;
1692 1709
1693 if (new_fs || new_fd || do_sysvsem || new_nsproxy) { 1710 if (new_fs || new_fd || do_sysvsem || new_nsproxy) {
diff --git a/kernel/futex.c b/kernel/futex.c
index 3fbc76cbb9aa..11cbe052b2e8 100644
--- a/kernel/futex.c
+++ b/kernel/futex.c
@@ -218,6 +218,8 @@ static void drop_futex_key_refs(union futex_key *key)
218 * @uaddr: virtual address of the futex 218 * @uaddr: virtual address of the futex
219 * @fshared: 0 for a PROCESS_PRIVATE futex, 1 for PROCESS_SHARED 219 * @fshared: 0 for a PROCESS_PRIVATE futex, 1 for PROCESS_SHARED
220 * @key: address where result is stored. 220 * @key: address where result is stored.
221 * @rw: mapping needs to be read/write (values: VERIFY_READ,
222 * VERIFY_WRITE)
221 * 223 *
222 * Returns a negative error code or 0 224 * Returns a negative error code or 0
223 * The key words are stored in *key on success. 225 * The key words are stored in *key on success.
@@ -229,12 +231,12 @@ static void drop_futex_key_refs(union futex_key *key)
229 * lock_page() might sleep, the caller should not hold a spinlock. 231 * lock_page() might sleep, the caller should not hold a spinlock.
230 */ 232 */
231static int 233static int
232get_futex_key(u32 __user *uaddr, int fshared, union futex_key *key) 234get_futex_key(u32 __user *uaddr, int fshared, union futex_key *key, int rw)
233{ 235{
234 unsigned long address = (unsigned long)uaddr; 236 unsigned long address = (unsigned long)uaddr;
235 struct mm_struct *mm = current->mm; 237 struct mm_struct *mm = current->mm;
236 struct page *page, *page_head; 238 struct page *page, *page_head;
237 int err; 239 int err, ro = 0;
238 240
239 /* 241 /*
240 * The futex address must be "naturally" aligned. 242 * The futex address must be "naturally" aligned.
@@ -262,8 +264,18 @@ get_futex_key(u32 __user *uaddr, int fshared, union futex_key *key)
262 264
263again: 265again:
264 err = get_user_pages_fast(address, 1, 1, &page); 266 err = get_user_pages_fast(address, 1, 1, &page);
267 /*
268 * If write access is not required (eg. FUTEX_WAIT), try
269 * and get read-only access.
270 */
271 if (err == -EFAULT && rw == VERIFY_READ) {
272 err = get_user_pages_fast(address, 1, 0, &page);
273 ro = 1;
274 }
265 if (err < 0) 275 if (err < 0)
266 return err; 276 return err;
277 else
278 err = 0;
267 279
268#ifdef CONFIG_TRANSPARENT_HUGEPAGE 280#ifdef CONFIG_TRANSPARENT_HUGEPAGE
269 page_head = page; 281 page_head = page;
@@ -305,6 +317,13 @@ again:
305 if (!page_head->mapping) { 317 if (!page_head->mapping) {
306 unlock_page(page_head); 318 unlock_page(page_head);
307 put_page(page_head); 319 put_page(page_head);
320 /*
321 * ZERO_PAGE pages don't have a mapping. Avoid a busy loop
322 * trying to find one. RW mapping would have COW'd (and thus
323 * have a mapping) so this page is RO and won't ever change.
324 */
325 if ((page_head == ZERO_PAGE(address)))
326 return -EFAULT;
308 goto again; 327 goto again;
309 } 328 }
310 329
@@ -316,6 +335,15 @@ again:
316 * the object not the particular process. 335 * the object not the particular process.
317 */ 336 */
318 if (PageAnon(page_head)) { 337 if (PageAnon(page_head)) {
338 /*
339 * A RO anonymous page will never change and thus doesn't make
340 * sense for futex operations.
341 */
342 if (ro) {
343 err = -EFAULT;
344 goto out;
345 }
346
319 key->both.offset |= FUT_OFF_MMSHARED; /* ref taken on mm */ 347 key->both.offset |= FUT_OFF_MMSHARED; /* ref taken on mm */
320 key->private.mm = mm; 348 key->private.mm = mm;
321 key->private.address = address; 349 key->private.address = address;
@@ -327,9 +355,10 @@ again:
327 355
328 get_futex_key_refs(key); 356 get_futex_key_refs(key);
329 357
358out:
330 unlock_page(page_head); 359 unlock_page(page_head);
331 put_page(page_head); 360 put_page(page_head);
332 return 0; 361 return err;
333} 362}
334 363
335static inline void put_futex_key(union futex_key *key) 364static inline void put_futex_key(union futex_key *key)
@@ -355,8 +384,8 @@ static int fault_in_user_writeable(u32 __user *uaddr)
355 int ret; 384 int ret;
356 385
357 down_read(&mm->mmap_sem); 386 down_read(&mm->mmap_sem);
358 ret = get_user_pages(current, mm, (unsigned long)uaddr, 387 ret = fixup_user_fault(current, mm, (unsigned long)uaddr,
359 1, 1, 0, NULL, NULL); 388 FAULT_FLAG_WRITE);
360 up_read(&mm->mmap_sem); 389 up_read(&mm->mmap_sem);
361 390
362 return ret < 0 ? ret : 0; 391 return ret < 0 ? ret : 0;
@@ -940,7 +969,7 @@ futex_wake(u32 __user *uaddr, unsigned int flags, int nr_wake, u32 bitset)
940 if (!bitset) 969 if (!bitset)
941 return -EINVAL; 970 return -EINVAL;
942 971
943 ret = get_futex_key(uaddr, flags & FLAGS_SHARED, &key); 972 ret = get_futex_key(uaddr, flags & FLAGS_SHARED, &key, VERIFY_READ);
944 if (unlikely(ret != 0)) 973 if (unlikely(ret != 0))
945 goto out; 974 goto out;
946 975
@@ -986,10 +1015,10 @@ futex_wake_op(u32 __user *uaddr1, unsigned int flags, u32 __user *uaddr2,
986 int ret, op_ret; 1015 int ret, op_ret;
987 1016
988retry: 1017retry:
989 ret = get_futex_key(uaddr1, flags & FLAGS_SHARED, &key1); 1018 ret = get_futex_key(uaddr1, flags & FLAGS_SHARED, &key1, VERIFY_READ);
990 if (unlikely(ret != 0)) 1019 if (unlikely(ret != 0))
991 goto out; 1020 goto out;
992 ret = get_futex_key(uaddr2, flags & FLAGS_SHARED, &key2); 1021 ret = get_futex_key(uaddr2, flags & FLAGS_SHARED, &key2, VERIFY_WRITE);
993 if (unlikely(ret != 0)) 1022 if (unlikely(ret != 0))
994 goto out_put_key1; 1023 goto out_put_key1;
995 1024
@@ -1243,10 +1272,11 @@ retry:
1243 pi_state = NULL; 1272 pi_state = NULL;
1244 } 1273 }
1245 1274
1246 ret = get_futex_key(uaddr1, flags & FLAGS_SHARED, &key1); 1275 ret = get_futex_key(uaddr1, flags & FLAGS_SHARED, &key1, VERIFY_READ);
1247 if (unlikely(ret != 0)) 1276 if (unlikely(ret != 0))
1248 goto out; 1277 goto out;
1249 ret = get_futex_key(uaddr2, flags & FLAGS_SHARED, &key2); 1278 ret = get_futex_key(uaddr2, flags & FLAGS_SHARED, &key2,
1279 requeue_pi ? VERIFY_WRITE : VERIFY_READ);
1250 if (unlikely(ret != 0)) 1280 if (unlikely(ret != 0))
1251 goto out_put_key1; 1281 goto out_put_key1;
1252 1282
@@ -1790,7 +1820,7 @@ static int futex_wait_setup(u32 __user *uaddr, u32 val, unsigned int flags,
1790 * while the syscall executes. 1820 * while the syscall executes.
1791 */ 1821 */
1792retry: 1822retry:
1793 ret = get_futex_key(uaddr, flags & FLAGS_SHARED, &q->key); 1823 ret = get_futex_key(uaddr, flags & FLAGS_SHARED, &q->key, VERIFY_READ);
1794 if (unlikely(ret != 0)) 1824 if (unlikely(ret != 0))
1795 return ret; 1825 return ret;
1796 1826
@@ -1941,7 +1971,7 @@ static int futex_lock_pi(u32 __user *uaddr, unsigned int flags, int detect,
1941 } 1971 }
1942 1972
1943retry: 1973retry:
1944 ret = get_futex_key(uaddr, flags & FLAGS_SHARED, &q.key); 1974 ret = get_futex_key(uaddr, flags & FLAGS_SHARED, &q.key, VERIFY_WRITE);
1945 if (unlikely(ret != 0)) 1975 if (unlikely(ret != 0))
1946 goto out; 1976 goto out;
1947 1977
@@ -2060,7 +2090,7 @@ retry:
2060 if ((uval & FUTEX_TID_MASK) != vpid) 2090 if ((uval & FUTEX_TID_MASK) != vpid)
2061 return -EPERM; 2091 return -EPERM;
2062 2092
2063 ret = get_futex_key(uaddr, flags & FLAGS_SHARED, &key); 2093 ret = get_futex_key(uaddr, flags & FLAGS_SHARED, &key, VERIFY_WRITE);
2064 if (unlikely(ret != 0)) 2094 if (unlikely(ret != 0))
2065 goto out; 2095 goto out;
2066 2096
@@ -2249,7 +2279,7 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags,
2249 debug_rt_mutex_init_waiter(&rt_waiter); 2279 debug_rt_mutex_init_waiter(&rt_waiter);
2250 rt_waiter.task = NULL; 2280 rt_waiter.task = NULL;
2251 2281
2252 ret = get_futex_key(uaddr2, flags & FLAGS_SHARED, &key2); 2282 ret = get_futex_key(uaddr2, flags & FLAGS_SHARED, &key2, VERIFY_WRITE);
2253 if (unlikely(ret != 0)) 2283 if (unlikely(ret != 0))
2254 goto out; 2284 goto out;
2255 2285
diff --git a/kernel/gcov/Kconfig b/kernel/gcov/Kconfig
index 5bf924d80b5c..a92028196cc1 100644
--- a/kernel/gcov/Kconfig
+++ b/kernel/gcov/Kconfig
@@ -3,7 +3,7 @@ menu "GCOV-based kernel profiling"
3config GCOV_KERNEL 3config GCOV_KERNEL
4 bool "Enable gcov-based kernel profiling" 4 bool "Enable gcov-based kernel profiling"
5 depends on DEBUG_FS 5 depends on DEBUG_FS
6 select CONSTRUCTORS 6 select CONSTRUCTORS if !UML
7 default n 7 default n
8 ---help--- 8 ---help---
9 This option enables gcov-based code profiling (e.g. for code coverage 9 This option enables gcov-based code profiling (e.g. for code coverage
diff --git a/kernel/irq/Kconfig b/kernel/irq/Kconfig
index d1d051b38e0b..5a38bf4de641 100644
--- a/kernel/irq/Kconfig
+++ b/kernel/irq/Kconfig
@@ -52,6 +52,10 @@ config IRQ_EDGE_EOI_HANDLER
52config GENERIC_IRQ_CHIP 52config GENERIC_IRQ_CHIP
53 bool 53 bool
54 54
55# Generic irq_domain hw <--> linux irq number translation
56config IRQ_DOMAIN
57 bool
58
55# Support forced irq threading 59# Support forced irq threading
56config IRQ_FORCED_THREADING 60config IRQ_FORCED_THREADING
57 bool 61 bool
diff --git a/kernel/irq/Makefile b/kernel/irq/Makefile
index 73290056cfb6..fff17381f0af 100644
--- a/kernel/irq/Makefile
+++ b/kernel/irq/Makefile
@@ -2,6 +2,7 @@
2obj-y := irqdesc.o handle.o manage.o spurious.o resend.o chip.o dummychip.o devres.o 2obj-y := irqdesc.o handle.o manage.o spurious.o resend.o chip.o dummychip.o devres.o
3obj-$(CONFIG_GENERIC_IRQ_CHIP) += generic-chip.o 3obj-$(CONFIG_GENERIC_IRQ_CHIP) += generic-chip.o
4obj-$(CONFIG_GENERIC_IRQ_PROBE) += autoprobe.o 4obj-$(CONFIG_GENERIC_IRQ_PROBE) += autoprobe.o
5obj-$(CONFIG_IRQ_DOMAIN) += irqdomain.o
5obj-$(CONFIG_PROC_FS) += proc.o 6obj-$(CONFIG_PROC_FS) += proc.o
6obj-$(CONFIG_GENERIC_PENDING_IRQ) += migration.o 7obj-$(CONFIG_GENERIC_PENDING_IRQ) += migration.o
7obj-$(CONFIG_PM_SLEEP) += pm.o 8obj-$(CONFIG_PM_SLEEP) += pm.o
diff --git a/kernel/irq/chip.c b/kernel/irq/chip.c
index d5a3009da71a..dc5114b4c16c 100644
--- a/kernel/irq/chip.c
+++ b/kernel/irq/chip.c
@@ -178,7 +178,7 @@ void irq_shutdown(struct irq_desc *desc)
178 desc->depth = 1; 178 desc->depth = 1;
179 if (desc->irq_data.chip->irq_shutdown) 179 if (desc->irq_data.chip->irq_shutdown)
180 desc->irq_data.chip->irq_shutdown(&desc->irq_data); 180 desc->irq_data.chip->irq_shutdown(&desc->irq_data);
181 if (desc->irq_data.chip->irq_disable) 181 else if (desc->irq_data.chip->irq_disable)
182 desc->irq_data.chip->irq_disable(&desc->irq_data); 182 desc->irq_data.chip->irq_disable(&desc->irq_data);
183 else 183 else
184 desc->irq_data.chip->irq_mask(&desc->irq_data); 184 desc->irq_data.chip->irq_mask(&desc->irq_data);
diff --git a/kernel/irq/devres.c b/kernel/irq/devres.c
index 1ef4ffcdfa55..bd8e788d71e0 100644
--- a/kernel/irq/devres.c
+++ b/kernel/irq/devres.c
@@ -87,8 +87,8 @@ void devm_free_irq(struct device *dev, unsigned int irq, void *dev_id)
87{ 87{
88 struct irq_devres match_data = { irq, dev_id }; 88 struct irq_devres match_data = { irq, dev_id };
89 89
90 free_irq(irq, dev_id);
91 WARN_ON(devres_destroy(dev, devm_irq_release, devm_irq_match, 90 WARN_ON(devres_destroy(dev, devm_irq_release, devm_irq_match,
92 &match_data)); 91 &match_data));
92 free_irq(irq, dev_id);
93} 93}
94EXPORT_SYMBOL(devm_free_irq); 94EXPORT_SYMBOL(devm_free_irq);
diff --git a/kernel/irq/generic-chip.c b/kernel/irq/generic-chip.c
index 3a2cab407b93..e38544dddb18 100644
--- a/kernel/irq/generic-chip.c
+++ b/kernel/irq/generic-chip.c
@@ -246,7 +246,7 @@ void irq_setup_generic_chip(struct irq_chip_generic *gc, u32 msk,
246 gc->mask_cache = irq_reg_readl(gc->reg_base + ct->regs.mask); 246 gc->mask_cache = irq_reg_readl(gc->reg_base + ct->regs.mask);
247 247
248 for (i = gc->irq_base; msk; msk >>= 1, i++) { 248 for (i = gc->irq_base; msk; msk >>= 1, i++) {
249 if (!msk & 0x01) 249 if (!(msk & 0x01))
250 continue; 250 continue;
251 251
252 if (flags & IRQ_GC_INIT_NESTED_LOCK) 252 if (flags & IRQ_GC_INIT_NESTED_LOCK)
@@ -301,7 +301,7 @@ void irq_remove_generic_chip(struct irq_chip_generic *gc, u32 msk,
301 raw_spin_unlock(&gc_lock); 301 raw_spin_unlock(&gc_lock);
302 302
303 for (; msk; msk >>= 1, i++) { 303 for (; msk; msk >>= 1, i++) {
304 if (!msk & 0x01) 304 if (!(msk & 0x01))
305 continue; 305 continue;
306 306
307 /* Remove handler first. That will mask the irq line */ 307 /* Remove handler first. That will mask the irq line */
diff --git a/kernel/irq/irqdesc.c b/kernel/irq/irqdesc.c
index 4c60a50e66b2..039b889ea053 100644
--- a/kernel/irq/irqdesc.c
+++ b/kernel/irq/irqdesc.c
@@ -70,7 +70,8 @@ static inline void desc_smp_init(struct irq_desc *desc, int node) { }
70static inline int desc_node(struct irq_desc *desc) { return 0; } 70static inline int desc_node(struct irq_desc *desc) { return 0; }
71#endif 71#endif
72 72
73static void desc_set_defaults(unsigned int irq, struct irq_desc *desc, int node) 73static void desc_set_defaults(unsigned int irq, struct irq_desc *desc, int node,
74 struct module *owner)
74{ 75{
75 int cpu; 76 int cpu;
76 77
@@ -86,6 +87,7 @@ static void desc_set_defaults(unsigned int irq, struct irq_desc *desc, int node)
86 desc->irq_count = 0; 87 desc->irq_count = 0;
87 desc->irqs_unhandled = 0; 88 desc->irqs_unhandled = 0;
88 desc->name = NULL; 89 desc->name = NULL;
90 desc->owner = owner;
89 for_each_possible_cpu(cpu) 91 for_each_possible_cpu(cpu)
90 *per_cpu_ptr(desc->kstat_irqs, cpu) = 0; 92 *per_cpu_ptr(desc->kstat_irqs, cpu) = 0;
91 desc_smp_init(desc, node); 93 desc_smp_init(desc, node);
@@ -128,7 +130,7 @@ static void free_masks(struct irq_desc *desc)
128static inline void free_masks(struct irq_desc *desc) { } 130static inline void free_masks(struct irq_desc *desc) { }
129#endif 131#endif
130 132
131static struct irq_desc *alloc_desc(int irq, int node) 133static struct irq_desc *alloc_desc(int irq, int node, struct module *owner)
132{ 134{
133 struct irq_desc *desc; 135 struct irq_desc *desc;
134 gfp_t gfp = GFP_KERNEL; 136 gfp_t gfp = GFP_KERNEL;
@@ -147,7 +149,7 @@ static struct irq_desc *alloc_desc(int irq, int node)
147 raw_spin_lock_init(&desc->lock); 149 raw_spin_lock_init(&desc->lock);
148 lockdep_set_class(&desc->lock, &irq_desc_lock_class); 150 lockdep_set_class(&desc->lock, &irq_desc_lock_class);
149 151
150 desc_set_defaults(irq, desc, node); 152 desc_set_defaults(irq, desc, node, owner);
151 153
152 return desc; 154 return desc;
153 155
@@ -173,13 +175,14 @@ static void free_desc(unsigned int irq)
173 kfree(desc); 175 kfree(desc);
174} 176}
175 177
176static int alloc_descs(unsigned int start, unsigned int cnt, int node) 178static int alloc_descs(unsigned int start, unsigned int cnt, int node,
179 struct module *owner)
177{ 180{
178 struct irq_desc *desc; 181 struct irq_desc *desc;
179 int i; 182 int i;
180 183
181 for (i = 0; i < cnt; i++) { 184 for (i = 0; i < cnt; i++) {
182 desc = alloc_desc(start + i, node); 185 desc = alloc_desc(start + i, node, owner);
183 if (!desc) 186 if (!desc)
184 goto err; 187 goto err;
185 mutex_lock(&sparse_irq_lock); 188 mutex_lock(&sparse_irq_lock);
@@ -227,7 +230,7 @@ int __init early_irq_init(void)
227 nr_irqs = initcnt; 230 nr_irqs = initcnt;
228 231
229 for (i = 0; i < initcnt; i++) { 232 for (i = 0; i < initcnt; i++) {
230 desc = alloc_desc(i, node); 233 desc = alloc_desc(i, node, NULL);
231 set_bit(i, allocated_irqs); 234 set_bit(i, allocated_irqs);
232 irq_insert_desc(i, desc); 235 irq_insert_desc(i, desc);
233 } 236 }
@@ -261,7 +264,7 @@ int __init early_irq_init(void)
261 alloc_masks(&desc[i], GFP_KERNEL, node); 264 alloc_masks(&desc[i], GFP_KERNEL, node);
262 raw_spin_lock_init(&desc[i].lock); 265 raw_spin_lock_init(&desc[i].lock);
263 lockdep_set_class(&desc[i].lock, &irq_desc_lock_class); 266 lockdep_set_class(&desc[i].lock, &irq_desc_lock_class);
264 desc_set_defaults(i, &desc[i], node); 267 desc_set_defaults(i, &desc[i], node, NULL);
265 } 268 }
266 return arch_early_irq_init(); 269 return arch_early_irq_init();
267} 270}
@@ -276,8 +279,16 @@ static void free_desc(unsigned int irq)
276 dynamic_irq_cleanup(irq); 279 dynamic_irq_cleanup(irq);
277} 280}
278 281
279static inline int alloc_descs(unsigned int start, unsigned int cnt, int node) 282static inline int alloc_descs(unsigned int start, unsigned int cnt, int node,
283 struct module *owner)
280{ 284{
285 u32 i;
286
287 for (i = 0; i < cnt; i++) {
288 struct irq_desc *desc = irq_to_desc(start + i);
289
290 desc->owner = owner;
291 }
281 return start; 292 return start;
282} 293}
283 294
@@ -333,11 +344,13 @@ EXPORT_SYMBOL_GPL(irq_free_descs);
333 * @from: Start the search from this irq number 344 * @from: Start the search from this irq number
334 * @cnt: Number of consecutive irqs to allocate. 345 * @cnt: Number of consecutive irqs to allocate.
335 * @node: Preferred node on which the irq descriptor should be allocated 346 * @node: Preferred node on which the irq descriptor should be allocated
347 * @owner: Owning module (can be NULL)
336 * 348 *
337 * Returns the first irq number or error code 349 * Returns the first irq number or error code
338 */ 350 */
339int __ref 351int __ref
340irq_alloc_descs(int irq, unsigned int from, unsigned int cnt, int node) 352__irq_alloc_descs(int irq, unsigned int from, unsigned int cnt, int node,
353 struct module *owner)
341{ 354{
342 int start, ret; 355 int start, ret;
343 356
@@ -366,13 +379,13 @@ irq_alloc_descs(int irq, unsigned int from, unsigned int cnt, int node)
366 379
367 bitmap_set(allocated_irqs, start, cnt); 380 bitmap_set(allocated_irqs, start, cnt);
368 mutex_unlock(&sparse_irq_lock); 381 mutex_unlock(&sparse_irq_lock);
369 return alloc_descs(start, cnt, node); 382 return alloc_descs(start, cnt, node, owner);
370 383
371err: 384err:
372 mutex_unlock(&sparse_irq_lock); 385 mutex_unlock(&sparse_irq_lock);
373 return ret; 386 return ret;
374} 387}
375EXPORT_SYMBOL_GPL(irq_alloc_descs); 388EXPORT_SYMBOL_GPL(__irq_alloc_descs);
376 389
377/** 390/**
378 * irq_reserve_irqs - mark irqs allocated 391 * irq_reserve_irqs - mark irqs allocated
@@ -440,7 +453,7 @@ void dynamic_irq_cleanup(unsigned int irq)
440 unsigned long flags; 453 unsigned long flags;
441 454
442 raw_spin_lock_irqsave(&desc->lock, flags); 455 raw_spin_lock_irqsave(&desc->lock, flags);
443 desc_set_defaults(irq, desc, desc_node(desc)); 456 desc_set_defaults(irq, desc, desc_node(desc), NULL);
444 raw_spin_unlock_irqrestore(&desc->lock, flags); 457 raw_spin_unlock_irqrestore(&desc->lock, flags);
445} 458}
446 459
diff --git a/kernel/irq/irqdomain.c b/kernel/irq/irqdomain.c
new file mode 100644
index 000000000000..b57a3776de44
--- /dev/null
+++ b/kernel/irq/irqdomain.c
@@ -0,0 +1,184 @@
1#include <linux/irq.h>
2#include <linux/irqdomain.h>
3#include <linux/module.h>
4#include <linux/mutex.h>
5#include <linux/of.h>
6#include <linux/of_address.h>
7#include <linux/slab.h>
8
9static LIST_HEAD(irq_domain_list);
10static DEFINE_MUTEX(irq_domain_mutex);
11
12/**
13 * irq_domain_add() - Register an irq_domain
14 * @domain: ptr to initialized irq_domain structure
15 *
16 * Registers an irq_domain structure. The irq_domain must at a minimum be
17 * initialized with an ops structure pointer, and either a ->to_irq hook or
18 * a valid irq_base value. Everything else is optional.
19 */
20void irq_domain_add(struct irq_domain *domain)
21{
22 struct irq_data *d;
23 int hwirq;
24
25 /*
26 * This assumes that the irq_domain owner has already allocated
27 * the irq_descs. This block will be removed when support for dynamic
28 * allocation of irq_descs is added to irq_domain.
29 */
30 for (hwirq = 0; hwirq < domain->nr_irq; hwirq++) {
31 d = irq_get_irq_data(irq_domain_to_irq(domain, hwirq));
32 if (!d) {
33 WARN(1, "error: assigning domain to non existant irq_desc");
34 return;
35 }
36 if (d->domain) {
37 /* things are broken; just report, don't clean up */
38 WARN(1, "error: irq_desc already assigned to a domain");
39 return;
40 }
41 d->domain = domain;
42 d->hwirq = hwirq;
43 }
44
45 mutex_lock(&irq_domain_mutex);
46 list_add(&domain->list, &irq_domain_list);
47 mutex_unlock(&irq_domain_mutex);
48}
49
50/**
51 * irq_domain_del() - Unregister an irq_domain
52 * @domain: ptr to registered irq_domain.
53 */
54void irq_domain_del(struct irq_domain *domain)
55{
56 struct irq_data *d;
57 int hwirq;
58
59 mutex_lock(&irq_domain_mutex);
60 list_del(&domain->list);
61 mutex_unlock(&irq_domain_mutex);
62
63 /* Clear the irq_domain assignments */
64 for (hwirq = 0; hwirq < domain->nr_irq; hwirq++) {
65 d = irq_get_irq_data(irq_domain_to_irq(domain, hwirq));
66 d->domain = NULL;
67 }
68}
69
70#if defined(CONFIG_OF_IRQ)
71/**
72 * irq_create_of_mapping() - Map a linux irq number from a DT interrupt spec
73 *
74 * Used by the device tree interrupt mapping code to translate a device tree
75 * interrupt specifier to a valid linux irq number. Returns either a valid
76 * linux IRQ number or 0.
77 *
78 * When the caller no longer need the irq number returned by this function it
79 * should arrange to call irq_dispose_mapping().
80 */
81unsigned int irq_create_of_mapping(struct device_node *controller,
82 const u32 *intspec, unsigned int intsize)
83{
84 struct irq_domain *domain;
85 unsigned long hwirq;
86 unsigned int irq, type;
87 int rc = -EINVAL;
88
89 /* Find a domain which can translate the irq spec */
90 mutex_lock(&irq_domain_mutex);
91 list_for_each_entry(domain, &irq_domain_list, list) {
92 if (!domain->ops->dt_translate)
93 continue;
94 rc = domain->ops->dt_translate(domain, controller,
95 intspec, intsize, &hwirq, &type);
96 if (rc == 0)
97 break;
98 }
99 mutex_unlock(&irq_domain_mutex);
100
101 if (rc != 0)
102 return 0;
103
104 irq = irq_domain_to_irq(domain, hwirq);
105 if (type != IRQ_TYPE_NONE)
106 irq_set_irq_type(irq, type);
107 pr_debug("%s: mapped hwirq=%i to irq=%i, flags=%x\n",
108 controller->full_name, (int)hwirq, irq, type);
109 return irq;
110}
111EXPORT_SYMBOL_GPL(irq_create_of_mapping);
112
113/**
114 * irq_dispose_mapping() - Discard a mapping created by irq_create_of_mapping()
115 * @irq: linux irq number to be discarded
116 *
117 * Calling this function indicates the caller no longer needs a reference to
118 * the linux irq number returned by a prior call to irq_create_of_mapping().
119 */
120void irq_dispose_mapping(unsigned int irq)
121{
122 /*
123 * nothing yet; will be filled when support for dynamic allocation of
124 * irq_descs is added to irq_domain
125 */
126}
127EXPORT_SYMBOL_GPL(irq_dispose_mapping);
128
129int irq_domain_simple_dt_translate(struct irq_domain *d,
130 struct device_node *controller,
131 const u32 *intspec, unsigned int intsize,
132 unsigned long *out_hwirq, unsigned int *out_type)
133{
134 if (d->of_node != controller)
135 return -EINVAL;
136 if (intsize < 1)
137 return -EINVAL;
138
139 *out_hwirq = intspec[0];
140 *out_type = IRQ_TYPE_NONE;
141 if (intsize > 1)
142 *out_type = intspec[1] & IRQ_TYPE_SENSE_MASK;
143 return 0;
144}
145
146struct irq_domain_ops irq_domain_simple_ops = {
147 .dt_translate = irq_domain_simple_dt_translate,
148};
149EXPORT_SYMBOL_GPL(irq_domain_simple_ops);
150
151/**
152 * irq_domain_create_simple() - Set up a 'simple' translation range
153 */
154void irq_domain_add_simple(struct device_node *controller, int irq_base)
155{
156 struct irq_domain *domain;
157
158 domain = kzalloc(sizeof(*domain), GFP_KERNEL);
159 if (!domain) {
160 WARN_ON(1);
161 return;
162 }
163
164 domain->irq_base = irq_base;
165 domain->of_node = of_node_get(controller);
166 domain->ops = &irq_domain_simple_ops;
167 irq_domain_add(domain);
168}
169EXPORT_SYMBOL_GPL(irq_domain_add_simple);
170
171void irq_domain_generate_simple(const struct of_device_id *match,
172 u64 phys_base, unsigned int irq_start)
173{
174 struct device_node *node;
175 pr_info("looking for phys_base=%llx, irq_start=%i\n",
176 (unsigned long long) phys_base, (int) irq_start);
177 node = of_find_matching_node_by_address(NULL, match, phys_base);
178 if (node)
179 irq_domain_add_simple(node, irq_start);
180 else
181 pr_info("no node found\n");
182}
183EXPORT_SYMBOL_GPL(irq_domain_generate_simple);
184#endif /* CONFIG_OF_IRQ */
diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c
index 0a7840aeb0fb..9b956fa20308 100644
--- a/kernel/irq/manage.c
+++ b/kernel/irq/manage.c
@@ -883,6 +883,8 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
883 883
884 if (desc->irq_data.chip == &no_irq_chip) 884 if (desc->irq_data.chip == &no_irq_chip)
885 return -ENOSYS; 885 return -ENOSYS;
886 if (!try_module_get(desc->owner))
887 return -ENODEV;
886 /* 888 /*
887 * Some drivers like serial.c use request_irq() heavily, 889 * Some drivers like serial.c use request_irq() heavily,
888 * so we have to be careful not to interfere with a 890 * so we have to be careful not to interfere with a
@@ -906,8 +908,10 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
906 */ 908 */
907 nested = irq_settings_is_nested_thread(desc); 909 nested = irq_settings_is_nested_thread(desc);
908 if (nested) { 910 if (nested) {
909 if (!new->thread_fn) 911 if (!new->thread_fn) {
910 return -EINVAL; 912 ret = -EINVAL;
913 goto out_mput;
914 }
911 /* 915 /*
912 * Replace the primary handler which was provided from 916 * Replace the primary handler which was provided from
913 * the driver for non nested interrupt handling by the 917 * the driver for non nested interrupt handling by the
@@ -929,8 +933,10 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
929 933
930 t = kthread_create(irq_thread, new, "irq/%d-%s", irq, 934 t = kthread_create(irq_thread, new, "irq/%d-%s", irq,
931 new->name); 935 new->name);
932 if (IS_ERR(t)) 936 if (IS_ERR(t)) {
933 return PTR_ERR(t); 937 ret = PTR_ERR(t);
938 goto out_mput;
939 }
934 /* 940 /*
935 * We keep the reference to the task struct even if 941 * We keep the reference to the task struct even if
936 * the thread dies to avoid that the interrupt code 942 * the thread dies to avoid that the interrupt code
@@ -1095,6 +1101,8 @@ out_thread:
1095 kthread_stop(t); 1101 kthread_stop(t);
1096 put_task_struct(t); 1102 put_task_struct(t);
1097 } 1103 }
1104out_mput:
1105 module_put(desc->owner);
1098 return ret; 1106 return ret;
1099} 1107}
1100 1108
@@ -1203,6 +1211,7 @@ static struct irqaction *__free_irq(unsigned int irq, void *dev_id)
1203 put_task_struct(action->thread); 1211 put_task_struct(action->thread);
1204 } 1212 }
1205 1213
1214 module_put(desc->owner);
1206 return action; 1215 return action;
1207} 1216}
1208 1217
diff --git a/kernel/kexec.c b/kernel/kexec.c
index 8d814cbc8109..296fbc84d659 100644
--- a/kernel/kexec.c
+++ b/kernel/kexec.c
@@ -1095,7 +1095,7 @@ size_t crash_get_memory_size(void)
1095 size_t size = 0; 1095 size_t size = 0;
1096 mutex_lock(&kexec_mutex); 1096 mutex_lock(&kexec_mutex);
1097 if (crashk_res.end != crashk_res.start) 1097 if (crashk_res.end != crashk_res.start)
1098 size = crashk_res.end - crashk_res.start + 1; 1098 size = resource_size(&crashk_res);
1099 mutex_unlock(&kexec_mutex); 1099 mutex_unlock(&kexec_mutex);
1100 return size; 1100 return size;
1101} 1101}
diff --git a/kernel/kmod.c b/kernel/kmod.c
index 47613dfb7b28..ddc7644c1305 100644
--- a/kernel/kmod.c
+++ b/kernel/kmod.c
@@ -274,7 +274,7 @@ static void __call_usermodehelper(struct work_struct *work)
274 * (used for preventing user land processes from being created after the user 274 * (used for preventing user land processes from being created after the user
275 * land has been frozen during a system-wide hibernation or suspend operation). 275 * land has been frozen during a system-wide hibernation or suspend operation).
276 */ 276 */
277static int usermodehelper_disabled; 277static int usermodehelper_disabled = 1;
278 278
279/* Number of helpers running */ 279/* Number of helpers running */
280static atomic_t running_helpers = ATOMIC_INIT(0); 280static atomic_t running_helpers = ATOMIC_INIT(0);
diff --git a/kernel/lockdep.c b/kernel/lockdep.c
index 3956f5149e25..91d67ce3a8d5 100644
--- a/kernel/lockdep.c
+++ b/kernel/lockdep.c
@@ -2468,7 +2468,7 @@ mark_held_locks(struct task_struct *curr, enum mark_type mark)
2468 2468
2469 BUG_ON(usage_bit >= LOCK_USAGE_STATES); 2469 BUG_ON(usage_bit >= LOCK_USAGE_STATES);
2470 2470
2471 if (hlock_class(hlock)->key == &__lockdep_no_validate__) 2471 if (hlock_class(hlock)->key == __lockdep_no_validate__.subkeys)
2472 continue; 2472 continue;
2473 2473
2474 if (!mark_lock(curr, hlock, usage_bit)) 2474 if (!mark_lock(curr, hlock, usage_bit))
@@ -2485,23 +2485,9 @@ static void __trace_hardirqs_on_caller(unsigned long ip)
2485{ 2485{
2486 struct task_struct *curr = current; 2486 struct task_struct *curr = current;
2487 2487
2488 if (DEBUG_LOCKS_WARN_ON(unlikely(early_boot_irqs_disabled)))
2489 return;
2490
2491 if (unlikely(curr->hardirqs_enabled)) {
2492 /*
2493 * Neither irq nor preemption are disabled here
2494 * so this is racy by nature but losing one hit
2495 * in a stat is not a big deal.
2496 */
2497 __debug_atomic_inc(redundant_hardirqs_on);
2498 return;
2499 }
2500 /* we'll do an OFF -> ON transition: */ 2488 /* we'll do an OFF -> ON transition: */
2501 curr->hardirqs_enabled = 1; 2489 curr->hardirqs_enabled = 1;
2502 2490
2503 if (DEBUG_LOCKS_WARN_ON(current->hardirq_context))
2504 return;
2505 /* 2491 /*
2506 * We are going to turn hardirqs on, so set the 2492 * We are going to turn hardirqs on, so set the
2507 * usage bit for all held locks: 2493 * usage bit for all held locks:
@@ -2529,9 +2515,25 @@ void trace_hardirqs_on_caller(unsigned long ip)
2529 if (unlikely(!debug_locks || current->lockdep_recursion)) 2515 if (unlikely(!debug_locks || current->lockdep_recursion))
2530 return; 2516 return;
2531 2517
2518 if (unlikely(current->hardirqs_enabled)) {
2519 /*
2520 * Neither irq nor preemption are disabled here
2521 * so this is racy by nature but losing one hit
2522 * in a stat is not a big deal.
2523 */
2524 __debug_atomic_inc(redundant_hardirqs_on);
2525 return;
2526 }
2527
2532 if (DEBUG_LOCKS_WARN_ON(!irqs_disabled())) 2528 if (DEBUG_LOCKS_WARN_ON(!irqs_disabled()))
2533 return; 2529 return;
2534 2530
2531 if (DEBUG_LOCKS_WARN_ON(unlikely(early_boot_irqs_disabled)))
2532 return;
2533
2534 if (DEBUG_LOCKS_WARN_ON(current->hardirq_context))
2535 return;
2536
2535 current->lockdep_recursion = 1; 2537 current->lockdep_recursion = 1;
2536 __trace_hardirqs_on_caller(ip); 2538 __trace_hardirqs_on_caller(ip);
2537 current->lockdep_recursion = 0; 2539 current->lockdep_recursion = 0;
@@ -2872,10 +2874,7 @@ static int mark_lock(struct task_struct *curr, struct held_lock *this,
2872void lockdep_init_map(struct lockdep_map *lock, const char *name, 2874void lockdep_init_map(struct lockdep_map *lock, const char *name,
2873 struct lock_class_key *key, int subclass) 2875 struct lock_class_key *key, int subclass)
2874{ 2876{
2875 int i; 2877 memset(lock, 0, sizeof(*lock));
2876
2877 for (i = 0; i < NR_LOCKDEP_CACHING_CLASSES; i++)
2878 lock->class_cache[i] = NULL;
2879 2878
2880#ifdef CONFIG_LOCK_STAT 2879#ifdef CONFIG_LOCK_STAT
2881 lock->cpu = raw_smp_processor_id(); 2880 lock->cpu = raw_smp_processor_id();
@@ -3112,7 +3111,13 @@ static int match_held_lock(struct held_lock *hlock, struct lockdep_map *lock)
3112 if (!class) 3111 if (!class)
3113 class = look_up_lock_class(lock, 0); 3112 class = look_up_lock_class(lock, 0);
3114 3113
3115 if (DEBUG_LOCKS_WARN_ON(!class)) 3114 /*
3115 * If look_up_lock_class() failed to find a class, we're trying
3116 * to test if we hold a lock that has never yet been acquired.
3117 * Clearly if the lock hasn't been acquired _ever_, we're not
3118 * holding it either, so report failure.
3119 */
3120 if (!class)
3116 return 0; 3121 return 0;
3117 3122
3118 if (DEBUG_LOCKS_WARN_ON(!hlock->nest_lock)) 3123 if (DEBUG_LOCKS_WARN_ON(!hlock->nest_lock))
diff --git a/kernel/module.c b/kernel/module.c
index 795bdc7f5c3f..04379f92f843 100644
--- a/kernel/module.c
+++ b/kernel/module.c
@@ -545,9 +545,9 @@ static void setup_modinfo_##field(struct module *mod, const char *s) \
545 mod->field = kstrdup(s, GFP_KERNEL); \ 545 mod->field = kstrdup(s, GFP_KERNEL); \
546} \ 546} \
547static ssize_t show_modinfo_##field(struct module_attribute *mattr, \ 547static ssize_t show_modinfo_##field(struct module_attribute *mattr, \
548 struct module *mod, char *buffer) \ 548 struct module_kobject *mk, char *buffer) \
549{ \ 549{ \
550 return sprintf(buffer, "%s\n", mod->field); \ 550 return sprintf(buffer, "%s\n", mk->mod->field); \
551} \ 551} \
552static int modinfo_##field##_exists(struct module *mod) \ 552static int modinfo_##field##_exists(struct module *mod) \
553{ \ 553{ \
@@ -902,9 +902,9 @@ void symbol_put_addr(void *addr)
902EXPORT_SYMBOL_GPL(symbol_put_addr); 902EXPORT_SYMBOL_GPL(symbol_put_addr);
903 903
904static ssize_t show_refcnt(struct module_attribute *mattr, 904static ssize_t show_refcnt(struct module_attribute *mattr,
905 struct module *mod, char *buffer) 905 struct module_kobject *mk, char *buffer)
906{ 906{
907 return sprintf(buffer, "%u\n", module_refcount(mod)); 907 return sprintf(buffer, "%u\n", module_refcount(mk->mod));
908} 908}
909 909
910static struct module_attribute refcnt = { 910static struct module_attribute refcnt = {
@@ -952,11 +952,11 @@ static inline int module_unload_init(struct module *mod)
952#endif /* CONFIG_MODULE_UNLOAD */ 952#endif /* CONFIG_MODULE_UNLOAD */
953 953
954static ssize_t show_initstate(struct module_attribute *mattr, 954static ssize_t show_initstate(struct module_attribute *mattr,
955 struct module *mod, char *buffer) 955 struct module_kobject *mk, char *buffer)
956{ 956{
957 const char *state = "unknown"; 957 const char *state = "unknown";
958 958
959 switch (mod->state) { 959 switch (mk->mod->state) {
960 case MODULE_STATE_LIVE: 960 case MODULE_STATE_LIVE:
961 state = "live"; 961 state = "live";
962 break; 962 break;
@@ -975,10 +975,27 @@ static struct module_attribute initstate = {
975 .show = show_initstate, 975 .show = show_initstate,
976}; 976};
977 977
978static ssize_t store_uevent(struct module_attribute *mattr,
979 struct module_kobject *mk,
980 const char *buffer, size_t count)
981{
982 enum kobject_action action;
983
984 if (kobject_action_type(buffer, count, &action) == 0)
985 kobject_uevent(&mk->kobj, action);
986 return count;
987}
988
989struct module_attribute module_uevent = {
990 .attr = { .name = "uevent", .mode = 0200 },
991 .store = store_uevent,
992};
993
978static struct module_attribute *modinfo_attrs[] = { 994static struct module_attribute *modinfo_attrs[] = {
979 &modinfo_version, 995 &modinfo_version,
980 &modinfo_srcversion, 996 &modinfo_srcversion,
981 &initstate, 997 &initstate,
998 &module_uevent,
982#ifdef CONFIG_MODULE_UNLOAD 999#ifdef CONFIG_MODULE_UNLOAD
983 &refcnt, 1000 &refcnt,
984#endif 1001#endif
@@ -1187,7 +1204,7 @@ struct module_sect_attrs
1187}; 1204};
1188 1205
1189static ssize_t module_sect_show(struct module_attribute *mattr, 1206static ssize_t module_sect_show(struct module_attribute *mattr,
1190 struct module *mod, char *buf) 1207 struct module_kobject *mk, char *buf)
1191{ 1208{
1192 struct module_sect_attr *sattr = 1209 struct module_sect_attr *sattr =
1193 container_of(mattr, struct module_sect_attr, mattr); 1210 container_of(mattr, struct module_sect_attr, mattr);
@@ -1697,6 +1714,15 @@ static void unset_module_core_ro_nx(struct module *mod) { }
1697static void unset_module_init_ro_nx(struct module *mod) { } 1714static void unset_module_init_ro_nx(struct module *mod) { }
1698#endif 1715#endif
1699 1716
1717void __weak module_free(struct module *mod, void *module_region)
1718{
1719 vfree(module_region);
1720}
1721
1722void __weak module_arch_cleanup(struct module *mod)
1723{
1724}
1725
1700/* Free a module, remove from lists, etc. */ 1726/* Free a module, remove from lists, etc. */
1701static void free_module(struct module *mod) 1727static void free_module(struct module *mod)
1702{ 1728{
@@ -1851,6 +1877,26 @@ static int simplify_symbols(struct module *mod, const struct load_info *info)
1851 return ret; 1877 return ret;
1852} 1878}
1853 1879
1880int __weak apply_relocate(Elf_Shdr *sechdrs,
1881 const char *strtab,
1882 unsigned int symindex,
1883 unsigned int relsec,
1884 struct module *me)
1885{
1886 pr_err("module %s: REL relocation unsupported\n", me->name);
1887 return -ENOEXEC;
1888}
1889
1890int __weak apply_relocate_add(Elf_Shdr *sechdrs,
1891 const char *strtab,
1892 unsigned int symindex,
1893 unsigned int relsec,
1894 struct module *me)
1895{
1896 pr_err("module %s: RELA relocation unsupported\n", me->name);
1897 return -ENOEXEC;
1898}
1899
1854static int apply_relocations(struct module *mod, const struct load_info *info) 1900static int apply_relocations(struct module *mod, const struct load_info *info)
1855{ 1901{
1856 unsigned int i; 1902 unsigned int i;
@@ -2235,6 +2281,11 @@ static void dynamic_debug_remove(struct _ddebug *debug)
2235 ddebug_remove_module(debug->modname); 2281 ddebug_remove_module(debug->modname);
2236} 2282}
2237 2283
2284void * __weak module_alloc(unsigned long size)
2285{
2286 return size == 0 ? NULL : vmalloc_exec(size);
2287}
2288
2238static void *module_alloc_update_bounds(unsigned long size) 2289static void *module_alloc_update_bounds(unsigned long size)
2239{ 2290{
2240 void *ret = module_alloc(size); 2291 void *ret = module_alloc(size);
@@ -2645,6 +2696,14 @@ static void flush_module_icache(const struct module *mod)
2645 set_fs(old_fs); 2696 set_fs(old_fs);
2646} 2697}
2647 2698
2699int __weak module_frob_arch_sections(Elf_Ehdr *hdr,
2700 Elf_Shdr *sechdrs,
2701 char *secstrings,
2702 struct module *mod)
2703{
2704 return 0;
2705}
2706
2648static struct module *layout_and_allocate(struct load_info *info) 2707static struct module *layout_and_allocate(struct load_info *info)
2649{ 2708{
2650 /* Module within temporary copy. */ 2709 /* Module within temporary copy. */
@@ -2716,6 +2775,13 @@ static void module_deallocate(struct module *mod, struct load_info *info)
2716 module_free(mod, mod->module_core); 2775 module_free(mod, mod->module_core);
2717} 2776}
2718 2777
2778int __weak module_finalize(const Elf_Ehdr *hdr,
2779 const Elf_Shdr *sechdrs,
2780 struct module *me)
2781{
2782 return 0;
2783}
2784
2719static int post_relocation(struct module *mod, const struct load_info *info) 2785static int post_relocation(struct module *mod, const struct load_info *info)
2720{ 2786{
2721 /* Sort exception table now relocations are done. */ 2787 /* Sort exception table now relocations are done. */
diff --git a/kernel/notifier.c b/kernel/notifier.c
index 2488ba7eb568..8d7b435806c9 100644
--- a/kernel/notifier.c
+++ b/kernel/notifier.c
@@ -525,37 +525,6 @@ void srcu_init_notifier_head(struct srcu_notifier_head *nh)
525} 525}
526EXPORT_SYMBOL_GPL(srcu_init_notifier_head); 526EXPORT_SYMBOL_GPL(srcu_init_notifier_head);
527 527
528/**
529 * register_reboot_notifier - Register function to be called at reboot time
530 * @nb: Info about notifier function to be called
531 *
532 * Registers a function with the list of functions
533 * to be called at reboot time.
534 *
535 * Currently always returns zero, as blocking_notifier_chain_register()
536 * always returns zero.
537 */
538int register_reboot_notifier(struct notifier_block *nb)
539{
540 return blocking_notifier_chain_register(&reboot_notifier_list, nb);
541}
542EXPORT_SYMBOL(register_reboot_notifier);
543
544/**
545 * unregister_reboot_notifier - Unregister previously registered reboot notifier
546 * @nb: Hook to be unregistered
547 *
548 * Unregisters a previously registered reboot
549 * notifier function.
550 *
551 * Returns zero on success, or %-ENOENT on failure.
552 */
553int unregister_reboot_notifier(struct notifier_block *nb)
554{
555 return blocking_notifier_chain_unregister(&reboot_notifier_list, nb);
556}
557EXPORT_SYMBOL(unregister_reboot_notifier);
558
559static ATOMIC_NOTIFIER_HEAD(die_chain); 528static ATOMIC_NOTIFIER_HEAD(die_chain);
560 529
561int notrace __kprobes notify_die(enum die_val val, const char *str, 530int notrace __kprobes notify_die(enum die_val val, const char *str,
diff --git a/kernel/panic.c b/kernel/panic.c
index 69231670eb95..d7bb6974efb5 100644
--- a/kernel/panic.c
+++ b/kernel/panic.c
@@ -119,6 +119,8 @@ NORET_TYPE void panic(const char * fmt, ...)
119 } 119 }
120 mdelay(PANIC_TIMER_STEP); 120 mdelay(PANIC_TIMER_STEP);
121 } 121 }
122 }
123 if (panic_timeout != 0) {
122 /* 124 /*
123 * This will not be a clean reboot, with everything 125 * This will not be a clean reboot, with everything
124 * shutting down. But if there is a chance of 126 * shutting down. But if there is a chance of
diff --git a/kernel/params.c b/kernel/params.c
index ed72e1330862..22df3e0d142a 100644
--- a/kernel/params.c
+++ b/kernel/params.c
@@ -225,8 +225,8 @@ int parse_args(const char *name,
225 int ret; \ 225 int ret; \
226 \ 226 \
227 ret = strtolfn(val, 0, &l); \ 227 ret = strtolfn(val, 0, &l); \
228 if (ret == -EINVAL || ((type)l != l)) \ 228 if (ret < 0 || ((type)l != l)) \
229 return -EINVAL; \ 229 return ret < 0 ? ret : -EINVAL; \
230 *((type *)kp->arg) = l; \ 230 *((type *)kp->arg) = l; \
231 return 0; \ 231 return 0; \
232 } \ 232 } \
@@ -511,7 +511,7 @@ struct module_param_attrs
511#define to_param_attr(n) container_of(n, struct param_attribute, mattr) 511#define to_param_attr(n) container_of(n, struct param_attribute, mattr)
512 512
513static ssize_t param_attr_show(struct module_attribute *mattr, 513static ssize_t param_attr_show(struct module_attribute *mattr,
514 struct module *mod, char *buf) 514 struct module_kobject *mk, char *buf)
515{ 515{
516 int count; 516 int count;
517 struct param_attribute *attribute = to_param_attr(mattr); 517 struct param_attribute *attribute = to_param_attr(mattr);
@@ -531,7 +531,7 @@ static ssize_t param_attr_show(struct module_attribute *mattr,
531 531
532/* sysfs always hands a nul-terminated string in buf. We rely on that. */ 532/* sysfs always hands a nul-terminated string in buf. We rely on that. */
533static ssize_t param_attr_store(struct module_attribute *mattr, 533static ssize_t param_attr_store(struct module_attribute *mattr,
534 struct module *owner, 534 struct module_kobject *km,
535 const char *buf, size_t len) 535 const char *buf, size_t len)
536{ 536{
537 int err; 537 int err;
@@ -730,6 +730,10 @@ static struct module_kobject * __init locate_module_kobject(const char *name)
730 mk->kobj.kset = module_kset; 730 mk->kobj.kset = module_kset;
731 err = kobject_init_and_add(&mk->kobj, &module_ktype, NULL, 731 err = kobject_init_and_add(&mk->kobj, &module_ktype, NULL,
732 "%s", name); 732 "%s", name);
733#ifdef CONFIG_MODULES
734 if (!err)
735 err = sysfs_create_file(&mk->kobj, &module_uevent.attr);
736#endif
733 if (err) { 737 if (err) {
734 kobject_put(&mk->kobj); 738 kobject_put(&mk->kobj);
735 printk(KERN_ERR 739 printk(KERN_ERR
@@ -807,7 +811,7 @@ static void __init param_sysfs_builtin(void)
807} 811}
808 812
809ssize_t __modver_version_show(struct module_attribute *mattr, 813ssize_t __modver_version_show(struct module_attribute *mattr,
810 struct module *mod, char *buf) 814 struct module_kobject *mk, char *buf)
811{ 815{
812 struct module_version_attribute *vattr = 816 struct module_version_attribute *vattr =
813 container_of(mattr, struct module_version_attribute, mattr); 817 container_of(mattr, struct module_version_attribute, mattr);
@@ -852,7 +856,7 @@ static ssize_t module_attr_show(struct kobject *kobj,
852 if (!attribute->show) 856 if (!attribute->show)
853 return -EIO; 857 return -EIO;
854 858
855 ret = attribute->show(attribute, mk->mod, buf); 859 ret = attribute->show(attribute, mk, buf);
856 860
857 return ret; 861 return ret;
858} 862}
@@ -871,7 +875,7 @@ static ssize_t module_attr_store(struct kobject *kobj,
871 if (!attribute->store) 875 if (!attribute->store)
872 return -EIO; 876 return -EIO;
873 877
874 ret = attribute->store(attribute, mk->mod, buf, len); 878 ret = attribute->store(attribute, mk, buf, len);
875 879
876 return ret; 880 return ret;
877} 881}
diff --git a/kernel/pid.c b/kernel/pid.c
index 57a8346a270e..e432057f3b21 100644
--- a/kernel/pid.c
+++ b/kernel/pid.c
@@ -405,7 +405,6 @@ struct task_struct *pid_task(struct pid *pid, enum pid_type type)
405 if (pid) { 405 if (pid) {
406 struct hlist_node *first; 406 struct hlist_node *first;
407 first = rcu_dereference_check(hlist_first_rcu(&pid->tasks[type]), 407 first = rcu_dereference_check(hlist_first_rcu(&pid->tasks[type]),
408 rcu_read_lock_held() ||
409 lockdep_tasklist_lock_is_held()); 408 lockdep_tasklist_lock_is_held());
410 if (first) 409 if (first)
411 result = hlist_entry(first, struct task_struct, pids[(type)].node); 410 result = hlist_entry(first, struct task_struct, pids[(type)].node);
diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c
index 58f405b581e7..c8008dd58ef2 100644
--- a/kernel/posix-cpu-timers.c
+++ b/kernel/posix-cpu-timers.c
@@ -250,7 +250,7 @@ void thread_group_cputime(struct task_struct *tsk, struct task_cputime *times)
250 do { 250 do {
251 times->utime = cputime_add(times->utime, t->utime); 251 times->utime = cputime_add(times->utime, t->utime);
252 times->stime = cputime_add(times->stime, t->stime); 252 times->stime = cputime_add(times->stime, t->stime);
253 times->sum_exec_runtime += t->se.sum_exec_runtime; 253 times->sum_exec_runtime += task_sched_runtime(t);
254 } while_each_thread(tsk, t); 254 } while_each_thread(tsk, t);
255out: 255out:
256 rcu_read_unlock(); 256 rcu_read_unlock();
@@ -312,7 +312,8 @@ static int cpu_clock_sample_group(const clockid_t which_clock,
312 cpu->cpu = cputime.utime; 312 cpu->cpu = cputime.utime;
313 break; 313 break;
314 case CPUCLOCK_SCHED: 314 case CPUCLOCK_SCHED:
315 cpu->sched = thread_group_sched_runtime(p); 315 thread_group_cputime(p, &cputime);
316 cpu->sched = cputime.sum_exec_runtime;
316 break; 317 break;
317 } 318 }
318 return 0; 319 return 0;
diff --git a/kernel/power/Kconfig b/kernel/power/Kconfig
index 7b856b3458d2..3744c594b19b 100644
--- a/kernel/power/Kconfig
+++ b/kernel/power/Kconfig
@@ -193,8 +193,8 @@ config APM_EMULATION
193 notification of APM "events" (e.g. battery status change). 193 notification of APM "events" (e.g. battery status change).
194 194
195 In order to use APM, you will need supporting software. For location 195 In order to use APM, you will need supporting software. For location
196 and more information, read <file:Documentation/power/pm.txt> and the 196 and more information, read <file:Documentation/power/apm-acpi.txt>
197 Battery Powered Linux mini-HOWTO, available from 197 and the Battery Powered Linux mini-HOWTO, available from
198 <http://www.tldp.org/docs.html#howto>. 198 <http://www.tldp.org/docs.html#howto>.
199 199
200 This driver does not spin down disk drives (see the hdparm(8) 200 This driver does not spin down disk drives (see the hdparm(8)
@@ -231,3 +231,7 @@ config PM_CLK
231config PM_GENERIC_DOMAINS 231config PM_GENERIC_DOMAINS
232 bool 232 bool
233 depends on PM 233 depends on PM
234
235config PM_GENERIC_DOMAINS_RUNTIME
236 def_bool y
237 depends on PM_RUNTIME && PM_GENERIC_DOMAINS
diff --git a/kernel/printk.c b/kernel/printk.c
index 37dff3429adb..28a40d8171b8 100644
--- a/kernel/printk.c
+++ b/kernel/printk.c
@@ -318,8 +318,10 @@ static int check_syslog_permissions(int type, bool from_file)
318 return 0; 318 return 0;
319 /* For historical reasons, accept CAP_SYS_ADMIN too, with a warning */ 319 /* For historical reasons, accept CAP_SYS_ADMIN too, with a warning */
320 if (capable(CAP_SYS_ADMIN)) { 320 if (capable(CAP_SYS_ADMIN)) {
321 WARN_ONCE(1, "Attempt to access syslog with CAP_SYS_ADMIN " 321 printk_once(KERN_WARNING "%s (%d): "
322 "but no CAP_SYSLOG (deprecated).\n"); 322 "Attempt to access syslog with CAP_SYS_ADMIN "
323 "but no CAP_SYSLOG (deprecated).\n",
324 current->comm, task_pid_nr(current));
323 return 0; 325 return 0;
324 } 326 }
325 return -EPERM; 327 return -EPERM;
@@ -1602,7 +1604,7 @@ static int __init printk_late_init(void)
1602 struct console *con; 1604 struct console *con;
1603 1605
1604 for_each_console(con) { 1606 for_each_console(con) {
1605 if (con->flags & CON_BOOT) { 1607 if (!keep_bootcon && con->flags & CON_BOOT) {
1606 printk(KERN_INFO "turn off boot console %s%d\n", 1608 printk(KERN_INFO "turn off boot console %s%d\n",
1607 con->name, con->index); 1609 con->name, con->index);
1608 unregister_console(con); 1610 unregister_console(con);
diff --git a/kernel/ptrace.c b/kernel/ptrace.c
index 9de3ecfd20f9..a70d2a5d8c7b 100644
--- a/kernel/ptrace.c
+++ b/kernel/ptrace.c
@@ -744,20 +744,17 @@ int ptrace_request(struct task_struct *child, long request,
744 break; 744 break;
745 745
746 si = child->last_siginfo; 746 si = child->last_siginfo;
747 if (unlikely(!si || si->si_code >> 8 != PTRACE_EVENT_STOP)) 747 if (likely(si && (si->si_code >> 8) == PTRACE_EVENT_STOP)) {
748 break; 748 child->jobctl |= JOBCTL_LISTENING;
749 749 /*
750 child->jobctl |= JOBCTL_LISTENING; 750 * If NOTIFY is set, it means event happened between
751 751 * start of this trap and now. Trigger re-trap.
752 /* 752 */
753 * If NOTIFY is set, it means event happened between start 753 if (child->jobctl & JOBCTL_TRAP_NOTIFY)
754 * of this trap and now. Trigger re-trap immediately. 754 signal_wake_up(child, true);
755 */ 755 ret = 0;
756 if (child->jobctl & JOBCTL_TRAP_NOTIFY) 756 }
757 signal_wake_up(child, true);
758
759 unlock_task_sighand(child, &flags); 757 unlock_task_sighand(child, &flags);
760 ret = 0;
761 break; 758 break;
762 759
763 case PTRACE_DETACH: /* detach a process that was attached. */ 760 case PTRACE_DETACH: /* detach a process that was attached. */
diff --git a/kernel/rcupdate.c b/kernel/rcupdate.c
index 7784bd216b6a..ddddb320be61 100644
--- a/kernel/rcupdate.c
+++ b/kernel/rcupdate.c
@@ -37,7 +37,7 @@
37#include <linux/smp.h> 37#include <linux/smp.h>
38#include <linux/interrupt.h> 38#include <linux/interrupt.h>
39#include <linux/sched.h> 39#include <linux/sched.h>
40#include <asm/atomic.h> 40#include <linux/atomic.h>
41#include <linux/bitops.h> 41#include <linux/bitops.h>
42#include <linux/percpu.h> 42#include <linux/percpu.h>
43#include <linux/notifier.h> 43#include <linux/notifier.h>
diff --git a/kernel/rcutorture.c b/kernel/rcutorture.c
index 2e138db03382..98f51b13bb7e 100644
--- a/kernel/rcutorture.c
+++ b/kernel/rcutorture.c
@@ -33,7 +33,7 @@
33#include <linux/rcupdate.h> 33#include <linux/rcupdate.h>
34#include <linux/interrupt.h> 34#include <linux/interrupt.h>
35#include <linux/sched.h> 35#include <linux/sched.h>
36#include <asm/atomic.h> 36#include <linux/atomic.h>
37#include <linux/bitops.h> 37#include <linux/bitops.h>
38#include <linux/completion.h> 38#include <linux/completion.h>
39#include <linux/moduleparam.h> 39#include <linux/moduleparam.h>
@@ -941,7 +941,6 @@ static void rcu_torture_timer(unsigned long unused)
941 idx = cur_ops->readlock(); 941 idx = cur_ops->readlock();
942 completed = cur_ops->completed(); 942 completed = cur_ops->completed();
943 p = rcu_dereference_check(rcu_torture_current, 943 p = rcu_dereference_check(rcu_torture_current,
944 rcu_read_lock_held() ||
945 rcu_read_lock_bh_held() || 944 rcu_read_lock_bh_held() ||
946 rcu_read_lock_sched_held() || 945 rcu_read_lock_sched_held() ||
947 srcu_read_lock_held(&srcu_ctl)); 946 srcu_read_lock_held(&srcu_ctl));
@@ -1002,7 +1001,6 @@ rcu_torture_reader(void *arg)
1002 idx = cur_ops->readlock(); 1001 idx = cur_ops->readlock();
1003 completed = cur_ops->completed(); 1002 completed = cur_ops->completed();
1004 p = rcu_dereference_check(rcu_torture_current, 1003 p = rcu_dereference_check(rcu_torture_current,
1005 rcu_read_lock_held() ||
1006 rcu_read_lock_bh_held() || 1004 rcu_read_lock_bh_held() ||
1007 rcu_read_lock_sched_held() || 1005 rcu_read_lock_sched_held() ||
1008 srcu_read_lock_held(&srcu_ctl)); 1006 srcu_read_lock_held(&srcu_ctl));
diff --git a/kernel/rcutree_trace.c b/kernel/rcutree_trace.c
index 4e144876dc68..3b0c0986afc0 100644
--- a/kernel/rcutree_trace.c
+++ b/kernel/rcutree_trace.c
@@ -31,7 +31,7 @@
31#include <linux/rcupdate.h> 31#include <linux/rcupdate.h>
32#include <linux/interrupt.h> 32#include <linux/interrupt.h>
33#include <linux/sched.h> 33#include <linux/sched.h>
34#include <asm/atomic.h> 34#include <linux/atomic.h>
35#include <linux/bitops.h> 35#include <linux/bitops.h>
36#include <linux/module.h> 36#include <linux/module.h>
37#include <linux/completion.h> 37#include <linux/completion.h>
diff --git a/kernel/resource.c b/kernel/resource.c
index 3ff40178dce7..c8dc249da5ce 100644
--- a/kernel/resource.c
+++ b/kernel/resource.c
@@ -419,6 +419,9 @@ static int __find_resource(struct resource *root, struct resource *old,
419 else 419 else
420 tmp.end = root->end; 420 tmp.end = root->end;
421 421
422 if (tmp.end < tmp.start)
423 goto next;
424
422 resource_clip(&tmp, constraint->min, constraint->max); 425 resource_clip(&tmp, constraint->min, constraint->max);
423 arch_remove_reservations(&tmp); 426 arch_remove_reservations(&tmp);
424 427
@@ -436,8 +439,10 @@ static int __find_resource(struct resource *root, struct resource *old,
436 return 0; 439 return 0;
437 } 440 }
438 } 441 }
439 if (!this) 442
443next: if (!this || this->end == root->end)
440 break; 444 break;
445
441 if (this != old) 446 if (this != old)
442 tmp.start = this->end + 1; 447 tmp.start = this->end + 1;
443 this = this->sibling; 448 this = this->sibling;
@@ -553,6 +558,27 @@ int allocate_resource(struct resource *root, struct resource *new,
553 558
554EXPORT_SYMBOL(allocate_resource); 559EXPORT_SYMBOL(allocate_resource);
555 560
561/**
562 * lookup_resource - find an existing resource by a resource start address
563 * @root: root resource descriptor
564 * @start: resource start address
565 *
566 * Returns a pointer to the resource if found, NULL otherwise
567 */
568struct resource *lookup_resource(struct resource *root, resource_size_t start)
569{
570 struct resource *res;
571
572 read_lock(&resource_lock);
573 for (res = root->child; res; res = res->sibling) {
574 if (res->start == start)
575 break;
576 }
577 read_unlock(&resource_lock);
578
579 return res;
580}
581
556/* 582/*
557 * Insert a resource into the resource tree. If successful, return NULL, 583 * Insert a resource into the resource tree. If successful, return NULL,
558 * otherwise return the conflicting resource (compare to __request_resource()) 584 * otherwise return the conflicting resource (compare to __request_resource())
diff --git a/kernel/rwsem.c b/kernel/rwsem.c
index 176e5e56ffab..9f48f3d82e9b 100644
--- a/kernel/rwsem.c
+++ b/kernel/rwsem.c
@@ -11,7 +11,7 @@
11#include <linux/rwsem.h> 11#include <linux/rwsem.h>
12 12
13#include <asm/system.h> 13#include <asm/system.h>
14#include <asm/atomic.h> 14#include <linux/atomic.h>
15 15
16/* 16/*
17 * lock for reading 17 * lock for reading
diff --git a/kernel/sched.c b/kernel/sched.c
index 9aaf567c5da5..b50b0f0c9aa9 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -75,6 +75,9 @@
75#include <asm/tlb.h> 75#include <asm/tlb.h>
76#include <asm/irq_regs.h> 76#include <asm/irq_regs.h>
77#include <asm/mutex.h> 77#include <asm/mutex.h>
78#ifdef CONFIG_PARAVIRT
79#include <asm/paravirt.h>
80#endif
78 81
79#include "sched_cpupri.h" 82#include "sched_cpupri.h"
80#include "workqueue_sched.h" 83#include "workqueue_sched.h"
@@ -528,6 +531,12 @@ struct rq {
528#ifdef CONFIG_IRQ_TIME_ACCOUNTING 531#ifdef CONFIG_IRQ_TIME_ACCOUNTING
529 u64 prev_irq_time; 532 u64 prev_irq_time;
530#endif 533#endif
534#ifdef CONFIG_PARAVIRT
535 u64 prev_steal_time;
536#endif
537#ifdef CONFIG_PARAVIRT_TIME_ACCOUNTING
538 u64 prev_steal_time_rq;
539#endif
531 540
532 /* calc_load related fields */ 541 /* calc_load related fields */
533 unsigned long calc_load_update; 542 unsigned long calc_load_update;
@@ -581,7 +590,6 @@ static inline int cpu_of(struct rq *rq)
581 590
582#define rcu_dereference_check_sched_domain(p) \ 591#define rcu_dereference_check_sched_domain(p) \
583 rcu_dereference_check((p), \ 592 rcu_dereference_check((p), \
584 rcu_read_lock_held() || \
585 lockdep_is_held(&sched_domains_mutex)) 593 lockdep_is_held(&sched_domains_mutex))
586 594
587/* 595/*
@@ -1921,10 +1929,28 @@ void account_system_vtime(struct task_struct *curr)
1921} 1929}
1922EXPORT_SYMBOL_GPL(account_system_vtime); 1930EXPORT_SYMBOL_GPL(account_system_vtime);
1923 1931
1924static void update_rq_clock_task(struct rq *rq, s64 delta) 1932#endif /* CONFIG_IRQ_TIME_ACCOUNTING */
1933
1934#ifdef CONFIG_PARAVIRT
1935static inline u64 steal_ticks(u64 steal)
1925{ 1936{
1926 s64 irq_delta; 1937 if (unlikely(steal > NSEC_PER_SEC))
1938 return div_u64(steal, TICK_NSEC);
1927 1939
1940 return __iter_div_u64_rem(steal, TICK_NSEC, &steal);
1941}
1942#endif
1943
1944static void update_rq_clock_task(struct rq *rq, s64 delta)
1945{
1946/*
1947 * In theory, the compile should just see 0 here, and optimize out the call
1948 * to sched_rt_avg_update. But I don't trust it...
1949 */
1950#if defined(CONFIG_IRQ_TIME_ACCOUNTING) || defined(CONFIG_PARAVIRT_TIME_ACCOUNTING)
1951 s64 steal = 0, irq_delta = 0;
1952#endif
1953#ifdef CONFIG_IRQ_TIME_ACCOUNTING
1928 irq_delta = irq_time_read(cpu_of(rq)) - rq->prev_irq_time; 1954 irq_delta = irq_time_read(cpu_of(rq)) - rq->prev_irq_time;
1929 1955
1930 /* 1956 /*
@@ -1947,12 +1973,35 @@ static void update_rq_clock_task(struct rq *rq, s64 delta)
1947 1973
1948 rq->prev_irq_time += irq_delta; 1974 rq->prev_irq_time += irq_delta;
1949 delta -= irq_delta; 1975 delta -= irq_delta;
1976#endif
1977#ifdef CONFIG_PARAVIRT_TIME_ACCOUNTING
1978 if (static_branch((&paravirt_steal_rq_enabled))) {
1979 u64 st;
1980
1981 steal = paravirt_steal_clock(cpu_of(rq));
1982 steal -= rq->prev_steal_time_rq;
1983
1984 if (unlikely(steal > delta))
1985 steal = delta;
1986
1987 st = steal_ticks(steal);
1988 steal = st * TICK_NSEC;
1989
1990 rq->prev_steal_time_rq += steal;
1991
1992 delta -= steal;
1993 }
1994#endif
1995
1950 rq->clock_task += delta; 1996 rq->clock_task += delta;
1951 1997
1952 if (irq_delta && sched_feat(NONIRQ_POWER)) 1998#if defined(CONFIG_IRQ_TIME_ACCOUNTING) || defined(CONFIG_PARAVIRT_TIME_ACCOUNTING)
1953 sched_rt_avg_update(rq, irq_delta); 1999 if ((irq_delta + steal) && sched_feat(NONTASK_POWER))
2000 sched_rt_avg_update(rq, irq_delta + steal);
2001#endif
1954} 2002}
1955 2003
2004#ifdef CONFIG_IRQ_TIME_ACCOUNTING
1956static int irqtime_account_hi_update(void) 2005static int irqtime_account_hi_update(void)
1957{ 2006{
1958 struct cpu_usage_stat *cpustat = &kstat_this_cpu.cpustat; 2007 struct cpu_usage_stat *cpustat = &kstat_this_cpu.cpustat;
@@ -1987,12 +2036,7 @@ static int irqtime_account_si_update(void)
1987 2036
1988#define sched_clock_irqtime (0) 2037#define sched_clock_irqtime (0)
1989 2038
1990static void update_rq_clock_task(struct rq *rq, s64 delta) 2039#endif
1991{
1992 rq->clock_task += delta;
1993}
1994
1995#endif /* CONFIG_IRQ_TIME_ACCOUNTING */
1996 2040
1997#include "sched_idletask.c" 2041#include "sched_idletask.c"
1998#include "sched_fair.c" 2042#include "sched_fair.c"
@@ -3021,7 +3065,7 @@ static void finish_task_switch(struct rq *rq, struct task_struct *prev)
3021#ifdef __ARCH_WANT_INTERRUPTS_ON_CTXSW 3065#ifdef __ARCH_WANT_INTERRUPTS_ON_CTXSW
3022 local_irq_disable(); 3066 local_irq_disable();
3023#endif /* __ARCH_WANT_INTERRUPTS_ON_CTXSW */ 3067#endif /* __ARCH_WANT_INTERRUPTS_ON_CTXSW */
3024 perf_event_task_sched_in(current); 3068 perf_event_task_sched_in(prev, current);
3025#ifdef __ARCH_WANT_INTERRUPTS_ON_CTXSW 3069#ifdef __ARCH_WANT_INTERRUPTS_ON_CTXSW
3026 local_irq_enable(); 3070 local_irq_enable();
3027#endif /* __ARCH_WANT_INTERRUPTS_ON_CTXSW */ 3071#endif /* __ARCH_WANT_INTERRUPTS_ON_CTXSW */
@@ -3681,30 +3725,6 @@ unsigned long long task_sched_runtime(struct task_struct *p)
3681} 3725}
3682 3726
3683/* 3727/*
3684 * Return sum_exec_runtime for the thread group.
3685 * In case the task is currently running, return the sum plus current's
3686 * pending runtime that have not been accounted yet.
3687 *
3688 * Note that the thread group might have other running tasks as well,
3689 * so the return value not includes other pending runtime that other
3690 * running tasks might have.
3691 */
3692unsigned long long thread_group_sched_runtime(struct task_struct *p)
3693{
3694 struct task_cputime totals;
3695 unsigned long flags;
3696 struct rq *rq;
3697 u64 ns;
3698
3699 rq = task_rq_lock(p, &flags);
3700 thread_group_cputime(p, &totals);
3701 ns = totals.sum_exec_runtime + do_task_delta_exec(p, rq);
3702 task_rq_unlock(rq, p, &flags);
3703
3704 return ns;
3705}
3706
3707/*
3708 * Account user cpu time to a process. 3728 * Account user cpu time to a process.
3709 * @p: the process that the cpu time gets accounted to 3729 * @p: the process that the cpu time gets accounted to
3710 * @cputime: the cpu time spent in user space since the last update 3730 * @cputime: the cpu time spent in user space since the last update
@@ -3845,6 +3865,25 @@ void account_idle_time(cputime_t cputime)
3845 cpustat->idle = cputime64_add(cpustat->idle, cputime64); 3865 cpustat->idle = cputime64_add(cpustat->idle, cputime64);
3846} 3866}
3847 3867
3868static __always_inline bool steal_account_process_tick(void)
3869{
3870#ifdef CONFIG_PARAVIRT
3871 if (static_branch(&paravirt_steal_enabled)) {
3872 u64 steal, st = 0;
3873
3874 steal = paravirt_steal_clock(smp_processor_id());
3875 steal -= this_rq()->prev_steal_time;
3876
3877 st = steal_ticks(steal);
3878 this_rq()->prev_steal_time += st * TICK_NSEC;
3879
3880 account_steal_time(st);
3881 return st;
3882 }
3883#endif
3884 return false;
3885}
3886
3848#ifndef CONFIG_VIRT_CPU_ACCOUNTING 3887#ifndef CONFIG_VIRT_CPU_ACCOUNTING
3849 3888
3850#ifdef CONFIG_IRQ_TIME_ACCOUNTING 3889#ifdef CONFIG_IRQ_TIME_ACCOUNTING
@@ -3876,6 +3915,9 @@ static void irqtime_account_process_tick(struct task_struct *p, int user_tick,
3876 cputime64_t tmp = cputime_to_cputime64(cputime_one_jiffy); 3915 cputime64_t tmp = cputime_to_cputime64(cputime_one_jiffy);
3877 struct cpu_usage_stat *cpustat = &kstat_this_cpu.cpustat; 3916 struct cpu_usage_stat *cpustat = &kstat_this_cpu.cpustat;
3878 3917
3918 if (steal_account_process_tick())
3919 return;
3920
3879 if (irqtime_account_hi_update()) { 3921 if (irqtime_account_hi_update()) {
3880 cpustat->irq = cputime64_add(cpustat->irq, tmp); 3922 cpustat->irq = cputime64_add(cpustat->irq, tmp);
3881 } else if (irqtime_account_si_update()) { 3923 } else if (irqtime_account_si_update()) {
@@ -3929,6 +3971,9 @@ void account_process_tick(struct task_struct *p, int user_tick)
3929 return; 3971 return;
3930 } 3972 }
3931 3973
3974 if (steal_account_process_tick())
3975 return;
3976
3932 if (user_tick) 3977 if (user_tick)
3933 account_user_time(p, cputime_one_jiffy, one_jiffy_scaled); 3978 account_user_time(p, cputime_one_jiffy, one_jiffy_scaled);
3934 else if ((p != rq->idle) || (irq_count() != HARDIRQ_OFFSET)) 3979 else if ((p != rq->idle) || (irq_count() != HARDIRQ_OFFSET))
@@ -4210,9 +4255,9 @@ pick_next_task(struct rq *rq)
4210} 4255}
4211 4256
4212/* 4257/*
4213 * schedule() is the main scheduler function. 4258 * __schedule() is the main scheduler function.
4214 */ 4259 */
4215asmlinkage void __sched schedule(void) 4260static void __sched __schedule(void)
4216{ 4261{
4217 struct task_struct *prev, *next; 4262 struct task_struct *prev, *next;
4218 unsigned long *switch_count; 4263 unsigned long *switch_count;
@@ -4253,16 +4298,6 @@ need_resched:
4253 if (to_wakeup) 4298 if (to_wakeup)
4254 try_to_wake_up_local(to_wakeup); 4299 try_to_wake_up_local(to_wakeup);
4255 } 4300 }
4256
4257 /*
4258 * If we are going to sleep and we have plugged IO
4259 * queued, make sure to submit it to avoid deadlocks.
4260 */
4261 if (blk_needs_flush_plug(prev)) {
4262 raw_spin_unlock(&rq->lock);
4263 blk_schedule_flush_plug(prev);
4264 raw_spin_lock(&rq->lock);
4265 }
4266 } 4301 }
4267 switch_count = &prev->nvcsw; 4302 switch_count = &prev->nvcsw;
4268 } 4303 }
@@ -4300,6 +4335,26 @@ need_resched:
4300 if (need_resched()) 4335 if (need_resched())
4301 goto need_resched; 4336 goto need_resched;
4302} 4337}
4338
4339static inline void sched_submit_work(struct task_struct *tsk)
4340{
4341 if (!tsk->state)
4342 return;
4343 /*
4344 * If we are going to sleep and we have plugged IO queued,
4345 * make sure to submit it to avoid deadlocks.
4346 */
4347 if (blk_needs_flush_plug(tsk))
4348 blk_schedule_flush_plug(tsk);
4349}
4350
4351asmlinkage void __sched schedule(void)
4352{
4353 struct task_struct *tsk = current;
4354
4355 sched_submit_work(tsk);
4356 __schedule();
4357}
4303EXPORT_SYMBOL(schedule); 4358EXPORT_SYMBOL(schedule);
4304 4359
4305#ifdef CONFIG_MUTEX_SPIN_ON_OWNER 4360#ifdef CONFIG_MUTEX_SPIN_ON_OWNER
@@ -4366,7 +4421,7 @@ asmlinkage void __sched notrace preempt_schedule(void)
4366 4421
4367 do { 4422 do {
4368 add_preempt_count_notrace(PREEMPT_ACTIVE); 4423 add_preempt_count_notrace(PREEMPT_ACTIVE);
4369 schedule(); 4424 __schedule();
4370 sub_preempt_count_notrace(PREEMPT_ACTIVE); 4425 sub_preempt_count_notrace(PREEMPT_ACTIVE);
4371 4426
4372 /* 4427 /*
@@ -4394,7 +4449,7 @@ asmlinkage void __sched preempt_schedule_irq(void)
4394 do { 4449 do {
4395 add_preempt_count(PREEMPT_ACTIVE); 4450 add_preempt_count(PREEMPT_ACTIVE);
4396 local_irq_enable(); 4451 local_irq_enable();
4397 schedule(); 4452 __schedule();
4398 local_irq_disable(); 4453 local_irq_disable();
4399 sub_preempt_count(PREEMPT_ACTIVE); 4454 sub_preempt_count(PREEMPT_ACTIVE);
4400 4455
@@ -5519,7 +5574,7 @@ static inline int should_resched(void)
5519static void __cond_resched(void) 5574static void __cond_resched(void)
5520{ 5575{
5521 add_preempt_count(PREEMPT_ACTIVE); 5576 add_preempt_count(PREEMPT_ACTIVE);
5522 schedule(); 5577 __schedule();
5523 sub_preempt_count(PREEMPT_ACTIVE); 5578 sub_preempt_count(PREEMPT_ACTIVE);
5524} 5579}
5525 5580
@@ -7374,6 +7429,7 @@ static void __sdt_free(const struct cpumask *cpu_map)
7374 struct sched_domain *sd = *per_cpu_ptr(sdd->sd, j); 7429 struct sched_domain *sd = *per_cpu_ptr(sdd->sd, j);
7375 if (sd && (sd->flags & SD_OVERLAP)) 7430 if (sd && (sd->flags & SD_OVERLAP))
7376 free_sched_groups(sd->groups, 0); 7431 free_sched_groups(sd->groups, 0);
7432 kfree(*per_cpu_ptr(sdd->sd, j));
7377 kfree(*per_cpu_ptr(sdd->sg, j)); 7433 kfree(*per_cpu_ptr(sdd->sg, j));
7378 kfree(*per_cpu_ptr(sdd->sgp, j)); 7434 kfree(*per_cpu_ptr(sdd->sgp, j));
7379 } 7435 }
diff --git a/kernel/sched_features.h b/kernel/sched_features.h
index 1e7066d76c26..2e74677cb040 100644
--- a/kernel/sched_features.h
+++ b/kernel/sched_features.h
@@ -61,9 +61,9 @@ SCHED_FEAT(LB_BIAS, 1)
61SCHED_FEAT(OWNER_SPIN, 1) 61SCHED_FEAT(OWNER_SPIN, 1)
62 62
63/* 63/*
64 * Decrement CPU power based on irq activity 64 * Decrement CPU power based on time not spent running tasks
65 */ 65 */
66SCHED_FEAT(NONIRQ_POWER, 1) 66SCHED_FEAT(NONTASK_POWER, 1)
67 67
68/* 68/*
69 * Queue remote wakeups on the target CPU and process them 69 * Queue remote wakeups on the target CPU and process them
diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c
index 97540f0c9e47..af1177858be3 100644
--- a/kernel/sched_rt.c
+++ b/kernel/sched_rt.c
@@ -1050,7 +1050,7 @@ select_task_rq_rt(struct task_struct *p, int sd_flag, int flags)
1050 */ 1050 */
1051 if (curr && unlikely(rt_task(curr)) && 1051 if (curr && unlikely(rt_task(curr)) &&
1052 (curr->rt.nr_cpus_allowed < 2 || 1052 (curr->rt.nr_cpus_allowed < 2 ||
1053 curr->prio < p->prio) && 1053 curr->prio <= p->prio) &&
1054 (p->rt.nr_cpus_allowed > 1)) { 1054 (p->rt.nr_cpus_allowed > 1)) {
1055 int target = find_lowest_rq(p); 1055 int target = find_lowest_rq(p);
1056 1056
@@ -1581,7 +1581,7 @@ static void task_woken_rt(struct rq *rq, struct task_struct *p)
1581 p->rt.nr_cpus_allowed > 1 && 1581 p->rt.nr_cpus_allowed > 1 &&
1582 rt_task(rq->curr) && 1582 rt_task(rq->curr) &&
1583 (rq->curr->rt.nr_cpus_allowed < 2 || 1583 (rq->curr->rt.nr_cpus_allowed < 2 ||
1584 rq->curr->prio < p->prio)) 1584 rq->curr->prio <= p->prio))
1585 push_rt_tasks(rq); 1585 push_rt_tasks(rq);
1586} 1586}
1587 1587
diff --git a/kernel/signal.c b/kernel/signal.c
index d7f70aed1cc0..291c9700be75 100644
--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -3102,15 +3102,11 @@ SYSCALL_DEFINE0(sgetmask)
3102 3102
3103SYSCALL_DEFINE1(ssetmask, int, newmask) 3103SYSCALL_DEFINE1(ssetmask, int, newmask)
3104{ 3104{
3105 int old; 3105 int old = current->blocked.sig[0];
3106 3106 sigset_t newset;
3107 spin_lock_irq(&current->sighand->siglock);
3108 old = current->blocked.sig[0];
3109 3107
3110 siginitset(&current->blocked, newmask & ~(sigmask(SIGKILL)| 3108 siginitset(&newset, newmask & ~(sigmask(SIGKILL) | sigmask(SIGSTOP)));
3111 sigmask(SIGSTOP))); 3109 set_current_blocked(&newset);
3112 recalc_sigpending();
3113 spin_unlock_irq(&current->sighand->siglock);
3114 3110
3115 return old; 3111 return old;
3116} 3112}
@@ -3167,11 +3163,8 @@ SYSCALL_DEFINE2(rt_sigsuspend, sigset_t __user *, unewset, size_t, sigsetsize)
3167 return -EFAULT; 3163 return -EFAULT;
3168 sigdelsetmask(&newset, sigmask(SIGKILL)|sigmask(SIGSTOP)); 3164 sigdelsetmask(&newset, sigmask(SIGKILL)|sigmask(SIGSTOP));
3169 3165
3170 spin_lock_irq(&current->sighand->siglock);
3171 current->saved_sigmask = current->blocked; 3166 current->saved_sigmask = current->blocked;
3172 current->blocked = newset; 3167 set_current_blocked(&newset);
3173 recalc_sigpending();
3174 spin_unlock_irq(&current->sighand->siglock);
3175 3168
3176 current->state = TASK_INTERRUPTIBLE; 3169 current->state = TASK_INTERRUPTIBLE;
3177 schedule(); 3170 schedule();
diff --git a/kernel/stop_machine.c b/kernel/stop_machine.c
index c1124752e1d3..ba5070ce5765 100644
--- a/kernel/stop_machine.c
+++ b/kernel/stop_machine.c
@@ -19,7 +19,7 @@
19#include <linux/interrupt.h> 19#include <linux/interrupt.h>
20#include <linux/kallsyms.h> 20#include <linux/kallsyms.h>
21 21
22#include <asm/atomic.h> 22#include <linux/atomic.h>
23 23
24/* 24/*
25 * Structure to determine completion condition and record errors. May 25 * Structure to determine completion condition and record errors. May
diff --git a/kernel/sys.c b/kernel/sys.c
index e4128b278f23..18ee1d2f6474 100644
--- a/kernel/sys.c
+++ b/kernel/sys.c
@@ -8,7 +8,6 @@
8#include <linux/mm.h> 8#include <linux/mm.h>
9#include <linux/utsname.h> 9#include <linux/utsname.h>
10#include <linux/mman.h> 10#include <linux/mman.h>
11#include <linux/notifier.h>
12#include <linux/reboot.h> 11#include <linux/reboot.h>
13#include <linux/prctl.h> 12#include <linux/prctl.h>
14#include <linux/highuid.h> 13#include <linux/highuid.h>
@@ -38,6 +37,8 @@
38#include <linux/fs_struct.h> 37#include <linux/fs_struct.h>
39#include <linux/gfp.h> 38#include <linux/gfp.h>
40#include <linux/syscore_ops.h> 39#include <linux/syscore_ops.h>
40#include <linux/version.h>
41#include <linux/ctype.h>
41 42
42#include <linux/compat.h> 43#include <linux/compat.h>
43#include <linux/syscalls.h> 44#include <linux/syscalls.h>
@@ -45,6 +46,8 @@
45#include <linux/user_namespace.h> 46#include <linux/user_namespace.h>
46 47
47#include <linux/kmsg_dump.h> 48#include <linux/kmsg_dump.h>
49/* Move somewhere else to avoid recompiling? */
50#include <generated/utsrelease.h>
48 51
49#include <asm/uaccess.h> 52#include <asm/uaccess.h>
50#include <asm/io.h> 53#include <asm/io.h>
@@ -320,6 +323,37 @@ void kernel_restart_prepare(char *cmd)
320} 323}
321 324
322/** 325/**
326 * register_reboot_notifier - Register function to be called at reboot time
327 * @nb: Info about notifier function to be called
328 *
329 * Registers a function with the list of functions
330 * to be called at reboot time.
331 *
332 * Currently always returns zero, as blocking_notifier_chain_register()
333 * always returns zero.
334 */
335int register_reboot_notifier(struct notifier_block *nb)
336{
337 return blocking_notifier_chain_register(&reboot_notifier_list, nb);
338}
339EXPORT_SYMBOL(register_reboot_notifier);
340
341/**
342 * unregister_reboot_notifier - Unregister previously registered reboot notifier
343 * @nb: Hook to be unregistered
344 *
345 * Unregisters a previously registered reboot
346 * notifier function.
347 *
348 * Returns zero on success, or %-ENOENT on failure.
349 */
350int unregister_reboot_notifier(struct notifier_block *nb)
351{
352 return blocking_notifier_chain_unregister(&reboot_notifier_list, nb);
353}
354EXPORT_SYMBOL(unregister_reboot_notifier);
355
356/**
323 * kernel_restart - reboot the system 357 * kernel_restart - reboot the system
324 * @cmd: pointer to buffer containing command to execute for restart 358 * @cmd: pointer to buffer containing command to execute for restart
325 * or %NULL 359 * or %NULL
@@ -591,11 +625,18 @@ static int set_user(struct cred *new)
591 if (!new_user) 625 if (!new_user)
592 return -EAGAIN; 626 return -EAGAIN;
593 627
628 /*
629 * We don't fail in case of NPROC limit excess here because too many
630 * poorly written programs don't check set*uid() return code, assuming
631 * it never fails if called by root. We may still enforce NPROC limit
632 * for programs doing set*uid()+execve() by harmlessly deferring the
633 * failure to the execve() stage.
634 */
594 if (atomic_read(&new_user->processes) >= rlimit(RLIMIT_NPROC) && 635 if (atomic_read(&new_user->processes) >= rlimit(RLIMIT_NPROC) &&
595 new_user != INIT_USER) { 636 new_user != INIT_USER)
596 free_uid(new_user); 637 current->flags |= PF_NPROC_EXCEEDED;
597 return -EAGAIN; 638 else
598 } 639 current->flags &= ~PF_NPROC_EXCEEDED;
599 640
600 free_uid(new->user); 641 free_uid(new->user);
601 new->user = new_user; 642 new->user = new_user;
@@ -1124,6 +1165,34 @@ DECLARE_RWSEM(uts_sem);
1124#define override_architecture(name) 0 1165#define override_architecture(name) 0
1125#endif 1166#endif
1126 1167
1168/*
1169 * Work around broken programs that cannot handle "Linux 3.0".
1170 * Instead we map 3.x to 2.6.40+x, so e.g. 3.0 would be 2.6.40
1171 */
1172static int override_release(char __user *release, int len)
1173{
1174 int ret = 0;
1175 char buf[len];
1176
1177 if (current->personality & UNAME26) {
1178 char *rest = UTS_RELEASE;
1179 int ndots = 0;
1180 unsigned v;
1181
1182 while (*rest) {
1183 if (*rest == '.' && ++ndots >= 3)
1184 break;
1185 if (!isdigit(*rest) && *rest != '.')
1186 break;
1187 rest++;
1188 }
1189 v = ((LINUX_VERSION_CODE >> 8) & 0xff) + 40;
1190 snprintf(buf, len, "2.6.%u%s", v, rest);
1191 ret = copy_to_user(release, buf, len);
1192 }
1193 return ret;
1194}
1195
1127SYSCALL_DEFINE1(newuname, struct new_utsname __user *, name) 1196SYSCALL_DEFINE1(newuname, struct new_utsname __user *, name)
1128{ 1197{
1129 int errno = 0; 1198 int errno = 0;
@@ -1133,6 +1202,8 @@ SYSCALL_DEFINE1(newuname, struct new_utsname __user *, name)
1133 errno = -EFAULT; 1202 errno = -EFAULT;
1134 up_read(&uts_sem); 1203 up_read(&uts_sem);
1135 1204
1205 if (!errno && override_release(name->release, sizeof(name->release)))
1206 errno = -EFAULT;
1136 if (!errno && override_architecture(name)) 1207 if (!errno && override_architecture(name))
1137 errno = -EFAULT; 1208 errno = -EFAULT;
1138 return errno; 1209 return errno;
@@ -1154,6 +1225,8 @@ SYSCALL_DEFINE1(uname, struct old_utsname __user *, name)
1154 error = -EFAULT; 1225 error = -EFAULT;
1155 up_read(&uts_sem); 1226 up_read(&uts_sem);
1156 1227
1228 if (!error && override_release(name->release, sizeof(name->release)))
1229 error = -EFAULT;
1157 if (!error && override_architecture(name)) 1230 if (!error && override_architecture(name))
1158 error = -EFAULT; 1231 error = -EFAULT;
1159 return error; 1232 return error;
@@ -1188,6 +1261,8 @@ SYSCALL_DEFINE1(olduname, struct oldold_utsname __user *, name)
1188 1261
1189 if (!error && override_architecture(name)) 1262 if (!error && override_architecture(name))
1190 error = -EFAULT; 1263 error = -EFAULT;
1264 if (!error && override_release(name->release, sizeof(name->release)))
1265 error = -EFAULT;
1191 return error ? -EFAULT : 0; 1266 return error ? -EFAULT : 0;
1192} 1267}
1193#endif 1268#endif
diff --git a/kernel/sys_ni.c b/kernel/sys_ni.c
index 62cbc8877fef..a9a5de07c4f1 100644
--- a/kernel/sys_ni.c
+++ b/kernel/sys_ni.c
@@ -16,7 +16,6 @@ asmlinkage long sys_ni_syscall(void)
16 return -ENOSYS; 16 return -ENOSYS;
17} 17}
18 18
19cond_syscall(sys_nfsservctl);
20cond_syscall(sys_quotactl); 19cond_syscall(sys_quotactl);
21cond_syscall(sys32_quotactl); 20cond_syscall(sys32_quotactl);
22cond_syscall(sys_acct); 21cond_syscall(sys_acct);
diff --git a/kernel/sysctl_binary.c b/kernel/sysctl_binary.c
index 3b8e028b9601..e8bffbe2ba4b 100644
--- a/kernel/sysctl_binary.c
+++ b/kernel/sysctl_binary.c
@@ -1,6 +1,6 @@
1#include <linux/stat.h> 1#include <linux/stat.h>
2#include <linux/sysctl.h> 2#include <linux/sysctl.h>
3#include "../fs/xfs/linux-2.6/xfs_sysctl.h" 3#include "../fs/xfs/xfs_sysctl.h"
4#include <linux/sunrpc/debug.h> 4#include <linux/sunrpc/debug.h>
5#include <linux/string.h> 5#include <linux/string.h>
6#include <net/ip_vs.h> 6#include <net/ip_vs.h>
diff --git a/kernel/sysctl_check.c b/kernel/sysctl_check.c
index 4e4932a7b360..362da653813d 100644
--- a/kernel/sysctl_check.c
+++ b/kernel/sysctl_check.c
@@ -1,6 +1,6 @@
1#include <linux/stat.h> 1#include <linux/stat.h>
2#include <linux/sysctl.h> 2#include <linux/sysctl.h>
3#include "../fs/xfs/linux-2.6/xfs_sysctl.h" 3#include "../fs/xfs/xfs_sysctl.h"
4#include <linux/sunrpc/debug.h> 4#include <linux/sunrpc/debug.h>
5#include <linux/string.h> 5#include <linux/string.h>
6#include <net/ip_vs.h> 6#include <net/ip_vs.h>
diff --git a/kernel/taskstats.c b/kernel/taskstats.c
index fc0f22005417..e66046456f4f 100644
--- a/kernel/taskstats.c
+++ b/kernel/taskstats.c
@@ -28,7 +28,7 @@
28#include <linux/fs.h> 28#include <linux/fs.h>
29#include <linux/file.h> 29#include <linux/file.h>
30#include <net/genetlink.h> 30#include <net/genetlink.h>
31#include <asm/atomic.h> 31#include <linux/atomic.h>
32 32
33/* 33/*
34 * Maximum length of a cpumask that can be specified in 34 * Maximum length of a cpumask that can be specified in
@@ -291,30 +291,28 @@ static int add_del_listener(pid_t pid, const struct cpumask *mask, int isadd)
291 if (!cpumask_subset(mask, cpu_possible_mask)) 291 if (!cpumask_subset(mask, cpu_possible_mask))
292 return -EINVAL; 292 return -EINVAL;
293 293
294 s = NULL;
295 if (isadd == REGISTER) { 294 if (isadd == REGISTER) {
296 for_each_cpu(cpu, mask) { 295 for_each_cpu(cpu, mask) {
297 if (!s) 296 s = kmalloc_node(sizeof(struct listener),
298 s = kmalloc_node(sizeof(struct listener), 297 GFP_KERNEL, cpu_to_node(cpu));
299 GFP_KERNEL, cpu_to_node(cpu));
300 if (!s) 298 if (!s)
301 goto cleanup; 299 goto cleanup;
300
302 s->pid = pid; 301 s->pid = pid;
303 INIT_LIST_HEAD(&s->list);
304 s->valid = 1; 302 s->valid = 1;
305 303
306 listeners = &per_cpu(listener_array, cpu); 304 listeners = &per_cpu(listener_array, cpu);
307 down_write(&listeners->sem); 305 down_write(&listeners->sem);
308 list_for_each_entry_safe(s2, tmp, &listeners->list, list) { 306 list_for_each_entry(s2, &listeners->list, list) {
309 if (s2->pid == pid) 307 if (s2->pid == pid && s2->valid)
310 goto next_cpu; 308 goto exists;
311 } 309 }
312 list_add(&s->list, &listeners->list); 310 list_add(&s->list, &listeners->list);
313 s = NULL; 311 s = NULL;
314next_cpu: 312exists:
315 up_write(&listeners->sem); 313 up_write(&listeners->sem);
314 kfree(s); /* nop if NULL */
316 } 315 }
317 kfree(s);
318 return 0; 316 return 0;
319 } 317 }
320 318
@@ -657,6 +655,7 @@ static struct genl_ops taskstats_ops = {
657 .cmd = TASKSTATS_CMD_GET, 655 .cmd = TASKSTATS_CMD_GET,
658 .doit = taskstats_user_cmd, 656 .doit = taskstats_user_cmd,
659 .policy = taskstats_cmd_get_policy, 657 .policy = taskstats_cmd_get_policy,
658 .flags = GENL_ADMIN_PERM,
660}; 659};
661 660
662static struct genl_ops cgroupstats_ops = { 661static struct genl_ops cgroupstats_ops = {
diff --git a/kernel/time/alarmtimer.c b/kernel/time/alarmtimer.c
index 59f369f98a04..ea5e1a928d5b 100644
--- a/kernel/time/alarmtimer.c
+++ b/kernel/time/alarmtimer.c
@@ -441,6 +441,8 @@ static int alarm_timer_create(struct k_itimer *new_timer)
441static void alarm_timer_get(struct k_itimer *timr, 441static void alarm_timer_get(struct k_itimer *timr,
442 struct itimerspec *cur_setting) 442 struct itimerspec *cur_setting)
443{ 443{
444 memset(cur_setting, 0, sizeof(struct itimerspec));
445
444 cur_setting->it_interval = 446 cur_setting->it_interval =
445 ktime_to_timespec(timr->it.alarmtimer.period); 447 ktime_to_timespec(timr->it.alarmtimer.period);
446 cur_setting->it_value = 448 cur_setting->it_value =
@@ -479,11 +481,17 @@ static int alarm_timer_set(struct k_itimer *timr, int flags,
479 if (!rtcdev) 481 if (!rtcdev)
480 return -ENOTSUPP; 482 return -ENOTSUPP;
481 483
482 /* Save old values */ 484 /*
483 old_setting->it_interval = 485 * XXX HACK! Currently we can DOS a system if the interval
484 ktime_to_timespec(timr->it.alarmtimer.period); 486 * period on alarmtimers is too small. Cap the interval here
485 old_setting->it_value = 487 * to 100us and solve this properly in a future patch! -jstultz
486 ktime_to_timespec(timr->it.alarmtimer.node.expires); 488 */
489 if ((new_setting->it_interval.tv_sec == 0) &&
490 (new_setting->it_interval.tv_nsec < 100000))
491 new_setting->it_interval.tv_nsec = 100000;
492
493 if (old_setting)
494 alarm_timer_get(timr, old_setting);
487 495
488 /* If the timer was already set, cancel it */ 496 /* If the timer was already set, cancel it */
489 alarm_cancel(&timr->it.alarmtimer); 497 alarm_cancel(&timr->it.alarmtimer);
diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig
index 2ad39e556cb4..cd3134510f3d 100644
--- a/kernel/trace/Kconfig
+++ b/kernel/trace/Kconfig
@@ -82,7 +82,7 @@ config EVENT_POWER_TRACING_DEPRECATED
82 power:power_frequency 82 power:power_frequency
83 This is for userspace compatibility 83 This is for userspace compatibility
84 and will vanish after 5 kernel iterations, 84 and will vanish after 5 kernel iterations,
85 namely 2.6.41. 85 namely 3.1.
86 86
87config CONTEXT_SWITCH_TRACER 87config CONTEXT_SWITCH_TRACER
88 bool 88 bool
diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c
index 6957aa298dfa..7c910a5593a6 100644
--- a/kernel/trace/blktrace.c
+++ b/kernel/trace/blktrace.c
@@ -206,6 +206,8 @@ static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes,
206 what |= MASK_TC_BIT(rw, RAHEAD); 206 what |= MASK_TC_BIT(rw, RAHEAD);
207 what |= MASK_TC_BIT(rw, META); 207 what |= MASK_TC_BIT(rw, META);
208 what |= MASK_TC_BIT(rw, DISCARD); 208 what |= MASK_TC_BIT(rw, DISCARD);
209 what |= MASK_TC_BIT(rw, FLUSH);
210 what |= MASK_TC_BIT(rw, FUA);
209 211
210 pid = tsk->pid; 212 pid = tsk->pid;
211 if (act_log_check(bt, what, sector, pid)) 213 if (act_log_check(bt, what, sector, pid))
@@ -1054,6 +1056,9 @@ static void fill_rwbs(char *rwbs, const struct blk_io_trace *t)
1054 goto out; 1056 goto out;
1055 } 1057 }
1056 1058
1059 if (tc & BLK_TC_FLUSH)
1060 rwbs[i++] = 'F';
1061
1057 if (tc & BLK_TC_DISCARD) 1062 if (tc & BLK_TC_DISCARD)
1058 rwbs[i++] = 'D'; 1063 rwbs[i++] = 'D';
1059 else if (tc & BLK_TC_WRITE) 1064 else if (tc & BLK_TC_WRITE)
@@ -1063,10 +1068,10 @@ static void fill_rwbs(char *rwbs, const struct blk_io_trace *t)
1063 else 1068 else
1064 rwbs[i++] = 'N'; 1069 rwbs[i++] = 'N';
1065 1070
1071 if (tc & BLK_TC_FUA)
1072 rwbs[i++] = 'F';
1066 if (tc & BLK_TC_AHEAD) 1073 if (tc & BLK_TC_AHEAD)
1067 rwbs[i++] = 'A'; 1074 rwbs[i++] = 'A';
1068 if (tc & BLK_TC_BARRIER)
1069 rwbs[i++] = 'B';
1070 if (tc & BLK_TC_SYNC) 1075 if (tc & BLK_TC_SYNC)
1071 rwbs[i++] = 'S'; 1076 rwbs[i++] = 'S';
1072 if (tc & BLK_TC_META) 1077 if (tc & BLK_TC_META)
@@ -1132,7 +1137,7 @@ typedef int (blk_log_action_t) (struct trace_iterator *iter, const char *act);
1132 1137
1133static int blk_log_action_classic(struct trace_iterator *iter, const char *act) 1138static int blk_log_action_classic(struct trace_iterator *iter, const char *act)
1134{ 1139{
1135 char rwbs[6]; 1140 char rwbs[RWBS_LEN];
1136 unsigned long long ts = iter->ts; 1141 unsigned long long ts = iter->ts;
1137 unsigned long nsec_rem = do_div(ts, NSEC_PER_SEC); 1142 unsigned long nsec_rem = do_div(ts, NSEC_PER_SEC);
1138 unsigned secs = (unsigned long)ts; 1143 unsigned secs = (unsigned long)ts;
@@ -1148,7 +1153,7 @@ static int blk_log_action_classic(struct trace_iterator *iter, const char *act)
1148 1153
1149static int blk_log_action(struct trace_iterator *iter, const char *act) 1154static int blk_log_action(struct trace_iterator *iter, const char *act)
1150{ 1155{
1151 char rwbs[6]; 1156 char rwbs[RWBS_LEN];
1152 const struct blk_io_trace *t = te_blk_io_trace(iter->ent); 1157 const struct blk_io_trace *t = te_blk_io_trace(iter->ent);
1153 1158
1154 fill_rwbs(rwbs, t); 1159 fill_rwbs(rwbs, t);
@@ -1561,7 +1566,7 @@ static const struct {
1561} mask_maps[] = { 1566} mask_maps[] = {
1562 { BLK_TC_READ, "read" }, 1567 { BLK_TC_READ, "read" },
1563 { BLK_TC_WRITE, "write" }, 1568 { BLK_TC_WRITE, "write" },
1564 { BLK_TC_BARRIER, "barrier" }, 1569 { BLK_TC_FLUSH, "flush" },
1565 { BLK_TC_SYNC, "sync" }, 1570 { BLK_TC_SYNC, "sync" },
1566 { BLK_TC_QUEUE, "queue" }, 1571 { BLK_TC_QUEUE, "queue" },
1567 { BLK_TC_REQUEUE, "requeue" }, 1572 { BLK_TC_REQUEUE, "requeue" },
@@ -1573,6 +1578,7 @@ static const struct {
1573 { BLK_TC_META, "meta" }, 1578 { BLK_TC_META, "meta" },
1574 { BLK_TC_DISCARD, "discard" }, 1579 { BLK_TC_DISCARD, "discard" },
1575 { BLK_TC_DRV_DATA, "drv_data" }, 1580 { BLK_TC_DRV_DATA, "drv_data" },
1581 { BLK_TC_FUA, "fua" },
1576}; 1582};
1577 1583
1578static int blk_trace_str2mask(const char *str) 1584static int blk_trace_str2mask(const char *str)
@@ -1788,6 +1794,9 @@ void blk_fill_rwbs(char *rwbs, u32 rw, int bytes)
1788{ 1794{
1789 int i = 0; 1795 int i = 0;
1790 1796
1797 if (rw & REQ_FLUSH)
1798 rwbs[i++] = 'F';
1799
1791 if (rw & WRITE) 1800 if (rw & WRITE)
1792 rwbs[i++] = 'W'; 1801 rwbs[i++] = 'W';
1793 else if (rw & REQ_DISCARD) 1802 else if (rw & REQ_DISCARD)
@@ -1797,6 +1806,8 @@ void blk_fill_rwbs(char *rwbs, u32 rw, int bytes)
1797 else 1806 else
1798 rwbs[i++] = 'N'; 1807 rwbs[i++] = 'N';
1799 1808
1809 if (rw & REQ_FUA)
1810 rwbs[i++] = 'F';
1800 if (rw & REQ_RAHEAD) 1811 if (rw & REQ_RAHEAD)
1801 rwbs[i++] = 'A'; 1812 rwbs[i++] = 'A';
1802 if (rw & REQ_SYNC) 1813 if (rw & REQ_SYNC)
diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h
index 3f381d0b20a8..616846bcfee5 100644
--- a/kernel/trace/trace.h
+++ b/kernel/trace/trace.h
@@ -2,7 +2,7 @@
2#define _LINUX_KERNEL_TRACE_H 2#define _LINUX_KERNEL_TRACE_H
3 3
4#include <linux/fs.h> 4#include <linux/fs.h>
5#include <asm/atomic.h> 5#include <linux/atomic.h>
6#include <linux/sched.h> 6#include <linux/sched.h>
7#include <linux/clocksource.h> 7#include <linux/clocksource.h>
8#include <linux/ring_buffer.h> 8#include <linux/ring_buffer.h>
diff --git a/kernel/trace/trace_mmiotrace.c b/kernel/trace/trace_mmiotrace.c
index 017fa376505d..fd3c8aae55e5 100644
--- a/kernel/trace/trace_mmiotrace.c
+++ b/kernel/trace/trace_mmiotrace.c
@@ -12,7 +12,7 @@
12#include <linux/slab.h> 12#include <linux/slab.h>
13#include <linux/time.h> 13#include <linux/time.h>
14 14
15#include <asm/atomic.h> 15#include <linux/atomic.h>
16 16
17#include "trace.h" 17#include "trace.h"
18#include "trace_output.h" 18#include "trace_output.h"
diff --git a/kernel/tsacct.c b/kernel/tsacct.c
index 24dc60d9fa1f..5bbfac85866e 100644
--- a/kernel/tsacct.c
+++ b/kernel/tsacct.c
@@ -78,6 +78,7 @@ void bacct_add_tsk(struct taskstats *stats, struct task_struct *tsk)
78 78
79#define KB 1024 79#define KB 1024
80#define MB (1024*KB) 80#define MB (1024*KB)
81#define KB_MASK (~(KB-1))
81/* 82/*
82 * fill in extended accounting fields 83 * fill in extended accounting fields
83 */ 84 */
@@ -95,14 +96,14 @@ void xacct_add_tsk(struct taskstats *stats, struct task_struct *p)
95 stats->hiwater_vm = get_mm_hiwater_vm(mm) * PAGE_SIZE / KB; 96 stats->hiwater_vm = get_mm_hiwater_vm(mm) * PAGE_SIZE / KB;
96 mmput(mm); 97 mmput(mm);
97 } 98 }
98 stats->read_char = p->ioac.rchar; 99 stats->read_char = p->ioac.rchar & KB_MASK;
99 stats->write_char = p->ioac.wchar; 100 stats->write_char = p->ioac.wchar & KB_MASK;
100 stats->read_syscalls = p->ioac.syscr; 101 stats->read_syscalls = p->ioac.syscr & KB_MASK;
101 stats->write_syscalls = p->ioac.syscw; 102 stats->write_syscalls = p->ioac.syscw & KB_MASK;
102#ifdef CONFIG_TASK_IO_ACCOUNTING 103#ifdef CONFIG_TASK_IO_ACCOUNTING
103 stats->read_bytes = p->ioac.read_bytes; 104 stats->read_bytes = p->ioac.read_bytes & KB_MASK;
104 stats->write_bytes = p->ioac.write_bytes; 105 stats->write_bytes = p->ioac.write_bytes & KB_MASK;
105 stats->cancelled_write_bytes = p->ioac.cancelled_write_bytes; 106 stats->cancelled_write_bytes = p->ioac.cancelled_write_bytes & KB_MASK;
106#else 107#else
107 stats->read_bytes = 0; 108 stats->read_bytes = 0;
108 stats->write_bytes = 0; 109 stats->write_bytes = 0;
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 25fb1b0e53fa..1783aabc6128 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -2412,8 +2412,13 @@ reflush:
2412 2412
2413 for_each_cwq_cpu(cpu, wq) { 2413 for_each_cwq_cpu(cpu, wq) {
2414 struct cpu_workqueue_struct *cwq = get_cwq(cpu, wq); 2414 struct cpu_workqueue_struct *cwq = get_cwq(cpu, wq);
2415 bool drained;
2415 2416
2416 if (!cwq->nr_active && list_empty(&cwq->delayed_works)) 2417 spin_lock_irq(&cwq->gcwq->lock);
2418 drained = !cwq->nr_active && list_empty(&cwq->delayed_works);
2419 spin_unlock_irq(&cwq->gcwq->lock);
2420
2421 if (drained)
2417 continue; 2422 continue;
2418 2423
2419 if (++flush_cnt == 10 || 2424 if (++flush_cnt == 10 ||
generated by cgit v1.2.2 (git 2.25.0) at 2025-12-28 03:37:24 -0500