diff options
Diffstat (limited to 'kernel')
41 files changed, 2113 insertions, 773 deletions
diff --git a/kernel/Makefile b/kernel/Makefile index 58908f9d156a..f6ef00f4f90f 100644 --- a/kernel/Makefile +++ b/kernel/Makefile @@ -20,7 +20,6 @@ obj-$(CONFIG_SMP) += cpu.o spinlock.o obj-$(CONFIG_DEBUG_SPINLOCK) += spinlock.o obj-$(CONFIG_UID16) += uid16.o obj-$(CONFIG_MODULES) += module.o -obj-$(CONFIG_OBSOLETE_INTERMODULE) += intermodule.o obj-$(CONFIG_KALLSYMS) += kallsyms.o obj-$(CONFIG_PM) += power/ obj-$(CONFIG_BSD_PROCESS_ACCT) += acct.o diff --git a/kernel/acct.c b/kernel/acct.c index b327f4d20104..368c4f03fe0e 100644 --- a/kernel/acct.c +++ b/kernel/acct.c @@ -75,7 +75,7 @@ int acct_parm[3] = {4, 2, 30}; /* * External references and all of the globals. */ -static void do_acct_process(long, struct file *); +static void do_acct_process(struct file *); /* * This structure is used so that all the data protected by lock @@ -118,7 +118,7 @@ static int check_free_space(struct file *file) spin_unlock(&acct_globals.lock); /* May block */ - if (vfs_statfs(file->f_dentry->d_inode->i_sb, &sbuf)) + if (vfs_statfs(file->f_dentry, &sbuf)) return res; suspend = sbuf.f_blocks * SUSPEND; resume = sbuf.f_blocks * RESUME; @@ -196,7 +196,7 @@ static void acct_file_reopen(struct file *file) if (old_acct) { mnt_unpin(old_acct->f_vfsmnt); spin_unlock(&acct_globals.lock); - do_acct_process(0, old_acct); + do_acct_process(old_acct); filp_close(old_acct, NULL); spin_lock(&acct_globals.lock); } @@ -419,16 +419,15 @@ static u32 encode_float(u64 value) /* * do_acct_process does all actual work. Caller holds the reference to file. */ -static void do_acct_process(long exitcode, struct file *file) +static void do_acct_process(struct file *file) { + struct pacct_struct *pacct = ¤t->signal->pacct; acct_t ac; mm_segment_t fs; - unsigned long vsize; unsigned long flim; u64 elapsed; u64 run_time; struct timespec uptime; - unsigned long jiffies; /* * First check to see if there is enough free_space to continue @@ -469,12 +468,6 @@ static void do_acct_process(long exitcode, struct file *file) #endif do_div(elapsed, AHZ); ac.ac_btime = xtime.tv_sec - elapsed; - jiffies = cputime_to_jiffies(cputime_add(current->utime, - current->signal->utime)); - ac.ac_utime = encode_comp_t(jiffies_to_AHZ(jiffies)); - jiffies = cputime_to_jiffies(cputime_add(current->stime, - current->signal->stime)); - ac.ac_stime = encode_comp_t(jiffies_to_AHZ(jiffies)); /* we really need to bite the bullet and change layout */ ac.ac_uid = current->uid; ac.ac_gid = current->gid; @@ -496,37 +489,18 @@ static void do_acct_process(long exitcode, struct file *file) old_encode_dev(tty_devnum(current->signal->tty)) : 0; read_unlock(&tasklist_lock); - ac.ac_flag = 0; - if (current->flags & PF_FORKNOEXEC) - ac.ac_flag |= AFORK; - if (current->flags & PF_SUPERPRIV) - ac.ac_flag |= ASU; - if (current->flags & PF_DUMPCORE) - ac.ac_flag |= ACORE; - if (current->flags & PF_SIGNALED) - ac.ac_flag |= AXSIG; - - vsize = 0; - if (current->mm) { - struct vm_area_struct *vma; - down_read(¤t->mm->mmap_sem); - vma = current->mm->mmap; - while (vma) { - vsize += vma->vm_end - vma->vm_start; - vma = vma->vm_next; - } - up_read(¤t->mm->mmap_sem); - } - vsize = vsize / 1024; - ac.ac_mem = encode_comp_t(vsize); + spin_lock(¤t->sighand->siglock); + ac.ac_utime = encode_comp_t(jiffies_to_AHZ(cputime_to_jiffies(pacct->ac_utime))); + ac.ac_stime = encode_comp_t(jiffies_to_AHZ(cputime_to_jiffies(pacct->ac_stime))); + ac.ac_flag = pacct->ac_flag; + ac.ac_mem = encode_comp_t(pacct->ac_mem); + ac.ac_minflt = encode_comp_t(pacct->ac_minflt); + ac.ac_majflt = encode_comp_t(pacct->ac_majflt); + ac.ac_exitcode = pacct->ac_exitcode; + spin_unlock(¤t->sighand->siglock); ac.ac_io = encode_comp_t(0 /* current->io_usage */); /* %% */ ac.ac_rw = encode_comp_t(ac.ac_io / 1024); - ac.ac_minflt = encode_comp_t(current->signal->min_flt + - current->min_flt); - ac.ac_majflt = encode_comp_t(current->signal->maj_flt + - current->maj_flt); ac.ac_swaps = encode_comp_t(0); - ac.ac_exitcode = exitcode; /* * Kernel segment override to datasegment and write it @@ -546,12 +520,63 @@ static void do_acct_process(long exitcode, struct file *file) } /** + * acct_init_pacct - initialize a new pacct_struct + */ +void acct_init_pacct(struct pacct_struct *pacct) +{ + memset(pacct, 0, sizeof(struct pacct_struct)); + pacct->ac_utime = pacct->ac_stime = cputime_zero; +} + +/** + * acct_collect - collect accounting information into pacct_struct + * @exitcode: task exit code + * @group_dead: not 0, if this thread is the last one in the process. + */ +void acct_collect(long exitcode, int group_dead) +{ + struct pacct_struct *pacct = ¤t->signal->pacct; + unsigned long vsize = 0; + + if (group_dead && current->mm) { + struct vm_area_struct *vma; + down_read(¤t->mm->mmap_sem); + vma = current->mm->mmap; + while (vma) { + vsize += vma->vm_end - vma->vm_start; + vma = vma->vm_next; + } + up_read(¤t->mm->mmap_sem); + } + + spin_lock_irq(¤t->sighand->siglock); + if (group_dead) + pacct->ac_mem = vsize / 1024; + if (thread_group_leader(current)) { + pacct->ac_exitcode = exitcode; + if (current->flags & PF_FORKNOEXEC) + pacct->ac_flag |= AFORK; + } + if (current->flags & PF_SUPERPRIV) + pacct->ac_flag |= ASU; + if (current->flags & PF_DUMPCORE) + pacct->ac_flag |= ACORE; + if (current->flags & PF_SIGNALED) + pacct->ac_flag |= AXSIG; + pacct->ac_utime = cputime_add(pacct->ac_utime, current->utime); + pacct->ac_stime = cputime_add(pacct->ac_stime, current->stime); + pacct->ac_minflt += current->min_flt; + pacct->ac_majflt += current->maj_flt; + spin_unlock_irq(¤t->sighand->siglock); +} + +/** * acct_process - now just a wrapper around do_acct_process * @exitcode: task exit code * * handles process accounting for an exiting task */ -void acct_process(long exitcode) +void acct_process() { struct file *file = NULL; @@ -570,7 +595,7 @@ void acct_process(long exitcode) get_file(file); spin_unlock(&acct_globals.lock); - do_acct_process(exitcode, file); + do_acct_process(file); fput(file); } @@ -599,9 +624,7 @@ void acct_update_integrals(struct task_struct *tsk) */ void acct_clear_integrals(struct task_struct *tsk) { - if (tsk) { - tsk->acct_stimexpd = 0; - tsk->acct_rss_mem1 = 0; - tsk->acct_vm_mem1 = 0; - } + tsk->acct_stimexpd = 0; + tsk->acct_rss_mem1 = 0; + tsk->acct_vm_mem1 = 0; } diff --git a/kernel/audit.c b/kernel/audit.c index df57b493e1cb..7dfac7031bd7 100644 --- a/kernel/audit.c +++ b/kernel/audit.c @@ -56,6 +56,7 @@ #include <linux/skbuff.h> #include <linux/netlink.h> #include <linux/selinux.h> +#include <linux/inotify.h> #include "audit.h" @@ -89,6 +90,7 @@ static int audit_backlog_wait_overflow = 0; /* The identity of the user shutting down the audit system. */ uid_t audit_sig_uid = -1; pid_t audit_sig_pid = -1; +u32 audit_sig_sid = 0; /* Records can be lost in several ways: 0) [suppressed in audit_alloc] @@ -102,6 +104,12 @@ static atomic_t audit_lost = ATOMIC_INIT(0); /* The netlink socket. */ static struct sock *audit_sock; +/* Inotify handle. */ +struct inotify_handle *audit_ih; + +/* Hash for inode-based rules */ +struct list_head audit_inode_hash[AUDIT_INODE_BUCKETS]; + /* The audit_freelist is a list of pre-allocated audit buffers (if more * than AUDIT_MAXFREE are in use, the audit buffer is freed instead of * being placed on the freelist). */ @@ -114,10 +122,8 @@ static struct task_struct *kauditd_task; static DECLARE_WAIT_QUEUE_HEAD(kauditd_wait); static DECLARE_WAIT_QUEUE_HEAD(audit_backlog_wait); -/* The netlink socket is only to be read by 1 CPU, which lets us assume - * that list additions and deletions never happen simultaneously in - * auditsc.c */ -DEFINE_MUTEX(audit_netlink_mutex); +/* Serialize requests from userspace. */ +static DEFINE_MUTEX(audit_cmd_mutex); /* AUDIT_BUFSIZ is the size of the temporary buffer used for formatting * audit records. Since printk uses a 1024 byte buffer, this buffer @@ -250,7 +256,7 @@ static int audit_set_rate_limit(int limit, uid_t loginuid, u32 sid) "audit_rate_limit=%d old=%d by auid=%u", limit, old, loginuid); audit_rate_limit = limit; - return old; + return 0; } static int audit_set_backlog_limit(int limit, uid_t loginuid, u32 sid) @@ -273,7 +279,7 @@ static int audit_set_backlog_limit(int limit, uid_t loginuid, u32 sid) "audit_backlog_limit=%d old=%d by auid=%u", limit, old, loginuid); audit_backlog_limit = limit; - return old; + return 0; } static int audit_set_enabled(int state, uid_t loginuid, u32 sid) @@ -299,7 +305,7 @@ static int audit_set_enabled(int state, uid_t loginuid, u32 sid) "audit_enabled=%d old=%d by auid=%u", state, old, loginuid); audit_enabled = state; - return old; + return 0; } static int audit_set_failure(int state, uid_t loginuid, u32 sid) @@ -327,7 +333,7 @@ static int audit_set_failure(int state, uid_t loginuid, u32 sid) "audit_failure=%d old=%d by auid=%u", state, old, loginuid); audit_failure = state; - return old; + return 0; } static int kauditd_thread(void *dummy) @@ -363,9 +369,52 @@ static int kauditd_thread(void *dummy) remove_wait_queue(&kauditd_wait, &wait); } } +} + +int audit_send_list(void *_dest) +{ + struct audit_netlink_list *dest = _dest; + int pid = dest->pid; + struct sk_buff *skb; + + /* wait for parent to finish and send an ACK */ + mutex_lock(&audit_cmd_mutex); + mutex_unlock(&audit_cmd_mutex); + + while ((skb = __skb_dequeue(&dest->q)) != NULL) + netlink_unicast(audit_sock, skb, pid, 0); + + kfree(dest); + return 0; } +struct sk_buff *audit_make_reply(int pid, int seq, int type, int done, + int multi, void *payload, int size) +{ + struct sk_buff *skb; + struct nlmsghdr *nlh; + int len = NLMSG_SPACE(size); + void *data; + int flags = multi ? NLM_F_MULTI : 0; + int t = done ? NLMSG_DONE : type; + + skb = alloc_skb(len, GFP_KERNEL); + if (!skb) + return NULL; + + nlh = NLMSG_PUT(skb, pid, seq, t, size); + nlh->nlmsg_flags = flags; + data = NLMSG_DATA(nlh); + memcpy(data, payload, size); + return skb; + +nlmsg_failure: /* Used by NLMSG_PUT */ + if (skb) + kfree_skb(skb); + return NULL; +} + /** * audit_send_reply - send an audit reply message via netlink * @pid: process id to send reply to @@ -383,29 +432,13 @@ void audit_send_reply(int pid, int seq, int type, int done, int multi, void *payload, int size) { struct sk_buff *skb; - struct nlmsghdr *nlh; - int len = NLMSG_SPACE(size); - void *data; - int flags = multi ? NLM_F_MULTI : 0; - int t = done ? NLMSG_DONE : type; - - skb = alloc_skb(len, GFP_KERNEL); + skb = audit_make_reply(pid, seq, type, done, multi, payload, size); if (!skb) return; - - nlh = NLMSG_PUT(skb, pid, seq, t, size); - nlh->nlmsg_flags = flags; - data = NLMSG_DATA(nlh); - memcpy(data, payload, size); - /* Ignore failure. It'll only happen if the sender goes away, because our timeout is set to infinite. */ netlink_unicast(audit_sock, skb, pid, 0); return; - -nlmsg_failure: /* Used by NLMSG_PUT */ - if (skb) - kfree_skb(skb); } /* @@ -451,7 +484,9 @@ static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh) struct audit_buffer *ab; u16 msg_type = nlh->nlmsg_type; uid_t loginuid; /* loginuid of sender */ - struct audit_sig_info sig_data; + struct audit_sig_info *sig_data; + char *ctx; + u32 len; err = audit_netlink_ok(NETLINK_CB(skb).eff_cap, msg_type); if (err) @@ -503,12 +538,9 @@ static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh) if (status_get->mask & AUDIT_STATUS_PID) { int old = audit_pid; if (sid) { - char *ctx = NULL; - u32 len; - int rc; - if ((rc = selinux_ctxid_to_string( + if ((err = selinux_ctxid_to_string( sid, &ctx, &len))) - return rc; + return err; else audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, @@ -523,10 +555,10 @@ static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh) audit_pid = status_get->pid; } if (status_get->mask & AUDIT_STATUS_RATE_LIMIT) - audit_set_rate_limit(status_get->rate_limit, + err = audit_set_rate_limit(status_get->rate_limit, loginuid, sid); if (status_get->mask & AUDIT_STATUS_BACKLOG_LIMIT) - audit_set_backlog_limit(status_get->backlog_limit, + err = audit_set_backlog_limit(status_get->backlog_limit, loginuid, sid); break; case AUDIT_USER: @@ -544,8 +576,6 @@ static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh) "user pid=%d uid=%u auid=%u", pid, uid, loginuid); if (sid) { - char *ctx = NULL; - u32 len; if (selinux_ctxid_to_string( sid, &ctx, &len)) { audit_log_format(ab, @@ -584,10 +614,21 @@ static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh) loginuid, sid); break; case AUDIT_SIGNAL_INFO: - sig_data.uid = audit_sig_uid; - sig_data.pid = audit_sig_pid; + err = selinux_ctxid_to_string(audit_sig_sid, &ctx, &len); + if (err) + return err; + sig_data = kmalloc(sizeof(*sig_data) + len, GFP_KERNEL); + if (!sig_data) { + kfree(ctx); + return -ENOMEM; + } + sig_data->uid = audit_sig_uid; + sig_data->pid = audit_sig_pid; + memcpy(sig_data->ctx, ctx, len); + kfree(ctx); audit_send_reply(NETLINK_CB(skb).pid, seq, AUDIT_SIGNAL_INFO, - 0, 0, &sig_data, sizeof(sig_data)); + 0, 0, sig_data, sizeof(*sig_data) + len); + kfree(sig_data); break; default: err = -EINVAL; @@ -629,20 +670,30 @@ static void audit_receive(struct sock *sk, int length) struct sk_buff *skb; unsigned int qlen; - mutex_lock(&audit_netlink_mutex); + mutex_lock(&audit_cmd_mutex); for (qlen = skb_queue_len(&sk->sk_receive_queue); qlen; qlen--) { skb = skb_dequeue(&sk->sk_receive_queue); audit_receive_skb(skb); kfree_skb(skb); } - mutex_unlock(&audit_netlink_mutex); + mutex_unlock(&audit_cmd_mutex); } +#ifdef CONFIG_AUDITSYSCALL +static const struct inotify_operations audit_inotify_ops = { + .handle_event = audit_handle_ievent, + .destroy_watch = audit_free_parent, +}; +#endif /* Initialize audit support at boot time. */ static int __init audit_init(void) { +#ifdef CONFIG_AUDITSYSCALL + int i; +#endif + printk(KERN_INFO "audit: initializing netlink socket (%s)\n", audit_default ? "enabled" : "disabled"); audit_sock = netlink_kernel_create(NETLINK_AUDIT, 0, audit_receive, @@ -661,6 +712,16 @@ static int __init audit_init(void) selinux_audit_set_callback(&selinux_audit_rule_update); audit_log(NULL, GFP_KERNEL, AUDIT_KERNEL, "initialized"); + +#ifdef CONFIG_AUDITSYSCALL + audit_ih = inotify_init(&audit_inotify_ops); + if (IS_ERR(audit_ih)) + audit_panic("cannot initialize inotify handle"); + + for (i = 0; i < AUDIT_INODE_BUCKETS; i++) + INIT_LIST_HEAD(&audit_inode_hash[i]); +#endif + return 0; } __initcall(audit_init); @@ -690,10 +751,12 @@ static void audit_buffer_free(struct audit_buffer *ab) kfree_skb(ab->skb); spin_lock_irqsave(&audit_freelist_lock, flags); - if (++audit_freelist_count > AUDIT_MAXFREE) + if (audit_freelist_count > AUDIT_MAXFREE) kfree(ab); - else + else { + audit_freelist_count++; list_add(&ab->list, &audit_freelist); + } spin_unlock_irqrestore(&audit_freelist_lock, flags); } @@ -988,28 +1051,76 @@ void audit_log_hex(struct audit_buffer *ab, const unsigned char *buf, skb_put(skb, len << 1); /* new string is twice the old string */ } +/* + * Format a string of no more than slen characters into the audit buffer, + * enclosed in quote marks. + */ +static void audit_log_n_string(struct audit_buffer *ab, size_t slen, + const char *string) +{ + int avail, new_len; + unsigned char *ptr; + struct sk_buff *skb; + + BUG_ON(!ab->skb); + skb = ab->skb; + avail = skb_tailroom(skb); + new_len = slen + 3; /* enclosing quotes + null terminator */ + if (new_len > avail) { + avail = audit_expand(ab, new_len); + if (!avail) + return; + } + ptr = skb->tail; + *ptr++ = '"'; + memcpy(ptr, string, slen); + ptr += slen; + *ptr++ = '"'; + *ptr = 0; + skb_put(skb, slen + 2); /* don't include null terminator */ +} + /** - * audit_log_unstrustedstring - log a string that may contain random characters + * audit_log_n_unstrustedstring - log a string that may contain random characters * @ab: audit_buffer + * @len: lenth of string (not including trailing null) * @string: string to be logged * * This code will escape a string that is passed to it if the string * contains a control character, unprintable character, double quote mark, * or a space. Unescaped strings will start and end with a double quote mark. * Strings that are escaped are printed in hex (2 digits per char). + * + * The caller specifies the number of characters in the string to log, which may + * or may not be the entire string. */ -void audit_log_untrustedstring(struct audit_buffer *ab, const char *string) +const char *audit_log_n_untrustedstring(struct audit_buffer *ab, size_t len, + const char *string) { const unsigned char *p = string; while (*p) { if (*p == '"' || *p < 0x21 || *p > 0x7f) { - audit_log_hex(ab, string, strlen(string)); - return; + audit_log_hex(ab, string, len); + return string + len + 1; } p++; } - audit_log_format(ab, "\"%s\"", string); + audit_log_n_string(ab, len, string); + return p + 1; +} + +/** + * audit_log_unstrustedstring - log a string that may contain random characters + * @ab: audit_buffer + * @string: string to be logged + * + * Same as audit_log_n_unstrustedstring(), except that strlen is used to + * determine string length. + */ +const char *audit_log_untrustedstring(struct audit_buffer *ab, const char *string) +{ + return audit_log_n_untrustedstring(ab, strlen(string), string); } /* This is a helper-function to print the escaped d_path */ diff --git a/kernel/audit.h b/kernel/audit.h index 6f733920fd32..8323e4132a33 100644 --- a/kernel/audit.h +++ b/kernel/audit.h @@ -19,9 +19,9 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ -#include <linux/mutex.h> #include <linux/fs.h> #include <linux/audit.h> +#include <linux/skbuff.h> /* 0 = no checking 1 = put_count checking @@ -53,6 +53,18 @@ enum audit_state { }; /* Rule lists */ +struct audit_parent; + +struct audit_watch { + atomic_t count; /* reference count */ + char *path; /* insertion path */ + dev_t dev; /* associated superblock device */ + unsigned long ino; /* associated inode number */ + struct audit_parent *parent; /* associated parent */ + struct list_head wlist; /* entry in parent->watches list */ + struct list_head rules; /* associated rules */ +}; + struct audit_field { u32 type; u32 val; @@ -70,6 +82,9 @@ struct audit_krule { u32 buflen; /* for data alloc on list rules */ u32 field_count; struct audit_field *fields; + struct audit_field *inode_f; /* quick access to an inode field */ + struct audit_watch *watch; /* associated watch */ + struct list_head rlist; /* entry in audit_watch.rules list */ }; struct audit_entry { @@ -78,15 +93,53 @@ struct audit_entry { struct audit_krule rule; }; - extern int audit_pid; -extern int audit_comparator(const u32 left, const u32 op, const u32 right); +#define AUDIT_INODE_BUCKETS 32 +extern struct list_head audit_inode_hash[AUDIT_INODE_BUCKETS]; + +static inline int audit_hash_ino(u32 ino) +{ + return (ino & (AUDIT_INODE_BUCKETS-1)); +} + +extern int audit_comparator(const u32 left, const u32 op, const u32 right); +extern int audit_compare_dname_path(const char *dname, const char *path, + int *dirlen); +extern struct sk_buff * audit_make_reply(int pid, int seq, int type, + int done, int multi, + void *payload, int size); extern void audit_send_reply(int pid, int seq, int type, int done, int multi, void *payload, int size); extern void audit_log_lost(const char *message); extern void audit_panic(const char *message); -extern struct mutex audit_netlink_mutex; +struct audit_netlink_list { + int pid; + struct sk_buff_head q; +}; + +int audit_send_list(void *); + +struct inotify_watch; +extern void audit_free_parent(struct inotify_watch *); +extern void audit_handle_ievent(struct inotify_watch *, u32, u32, u32, + const char *, struct inode *); extern int selinux_audit_rule_update(void); + +#ifdef CONFIG_AUDITSYSCALL +extern void __audit_signal_info(int sig, struct task_struct *t); +static inline void audit_signal_info(int sig, struct task_struct *t) +{ + if (unlikely(audit_pid && t->tgid == audit_pid)) + __audit_signal_info(sig, t); +} +extern enum audit_state audit_filter_inodes(struct task_struct *, + struct audit_context *); +extern void audit_set_auditable(struct audit_context *); +#else +#define audit_signal_info(s,t) +#define audit_filter_inodes(t,c) AUDIT_DISABLED +#define audit_set_auditable(c) +#endif diff --git a/kernel/auditfilter.c b/kernel/auditfilter.c index 7c134906d689..4c99d2c586ed 100644 --- a/kernel/auditfilter.c +++ b/kernel/auditfilter.c @@ -22,13 +22,59 @@ #include <linux/kernel.h> #include <linux/audit.h> #include <linux/kthread.h> +#include <linux/mutex.h> +#include <linux/fs.h> +#include <linux/namei.h> #include <linux/netlink.h> +#include <linux/sched.h> +#include <linux/inotify.h> #include <linux/selinux.h> #include "audit.h" -/* There are three lists of rules -- one to search at task creation - * time, one to search at syscall entry time, and another to search at - * syscall exit time. */ +/* + * Locking model: + * + * audit_filter_mutex: + * Synchronizes writes and blocking reads of audit's filterlist + * data. Rcu is used to traverse the filterlist and access + * contents of structs audit_entry, audit_watch and opaque + * selinux rules during filtering. If modified, these structures + * must be copied and replace their counterparts in the filterlist. + * An audit_parent struct is not accessed during filtering, so may + * be written directly provided audit_filter_mutex is held. + */ + +/* + * Reference counting: + * + * audit_parent: lifetime is from audit_init_parent() to receipt of an IN_IGNORED + * event. Each audit_watch holds a reference to its associated parent. + * + * audit_watch: if added to lists, lifetime is from audit_init_watch() to + * audit_remove_watch(). Additionally, an audit_watch may exist + * temporarily to assist in searching existing filter data. Each + * audit_krule holds a reference to its associated watch. + */ + +struct audit_parent { + struct list_head ilist; /* entry in inotify registration list */ + struct list_head watches; /* associated watches */ + struct inotify_watch wdata; /* inotify watch data */ + unsigned flags; /* status flags */ +}; + +/* + * audit_parent status flags: + * + * AUDIT_PARENT_INVALID - set anytime rules/watches are auto-removed due to + * a filesystem event to ensure we're adding audit watches to a valid parent. + * Technically not needed for IN_DELETE_SELF or IN_UNMOUNT events, as we cannot + * receive them while we have nameidata, but must be used for IN_MOVE_SELF which + * we can receive while holding nameidata. + */ +#define AUDIT_PARENT_INVALID 0x001 + +/* Audit filter lists, defined in <linux/audit.h> */ struct list_head audit_filter_list[AUDIT_NR_FILTERS] = { LIST_HEAD_INIT(audit_filter_list[0]), LIST_HEAD_INIT(audit_filter_list[1]), @@ -41,9 +87,53 @@ struct list_head audit_filter_list[AUDIT_NR_FILTERS] = { #endif }; +static DEFINE_MUTEX(audit_filter_mutex); + +/* Inotify handle */ +extern struct inotify_handle *audit_ih; + +/* Inotify events we care about. */ +#define AUDIT_IN_WATCH IN_MOVE|IN_CREATE|IN_DELETE|IN_DELETE_SELF|IN_MOVE_SELF + +void audit_free_parent(struct inotify_watch *i_watch) +{ + struct audit_parent *parent; + + parent = container_of(i_watch, struct audit_parent, wdata); + WARN_ON(!list_empty(&parent->watches)); + kfree(parent); +} + +static inline void audit_get_watch(struct audit_watch *watch) +{ + atomic_inc(&watch->count); +} + +static void audit_put_watch(struct audit_watch *watch) +{ + if (atomic_dec_and_test(&watch->count)) { + WARN_ON(watch->parent); + WARN_ON(!list_empty(&watch->rules)); + kfree(watch->path); + kfree(watch); + } +} + +static void audit_remove_watch(struct audit_watch *watch) +{ + list_del(&watch->wlist); + put_inotify_watch(&watch->parent->wdata); + watch->parent = NULL; + audit_put_watch(watch); /* match initial get */ +} + static inline void audit_free_rule(struct audit_entry *e) { int i; + + /* some rules don't have associated watches */ + if (e->rule.watch) + audit_put_watch(e->rule.watch); if (e->rule.fields) for (i = 0; i < e->rule.field_count; i++) { struct audit_field *f = &e->rule.fields[i]; @@ -60,6 +150,50 @@ static inline void audit_free_rule_rcu(struct rcu_head *head) audit_free_rule(e); } +/* Initialize a parent watch entry. */ +static struct audit_parent *audit_init_parent(struct nameidata *ndp) +{ + struct audit_parent *parent; + s32 wd; + + parent = kzalloc(sizeof(*parent), GFP_KERNEL); + if (unlikely(!parent)) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&parent->watches); + parent->flags = 0; + + inotify_init_watch(&parent->wdata); + /* grab a ref so inotify watch hangs around until we take audit_filter_mutex */ + get_inotify_watch(&parent->wdata); + wd = inotify_add_watch(audit_ih, &parent->wdata, ndp->dentry->d_inode, + AUDIT_IN_WATCH); + if (wd < 0) { + audit_free_parent(&parent->wdata); + return ERR_PTR(wd); + } + + return parent; +} + +/* Initialize a watch entry. */ +static struct audit_watch *audit_init_watch(char *path) +{ + struct audit_watch *watch; + + watch = kzalloc(sizeof(*watch), GFP_KERNEL); + if (unlikely(!watch)) + return ERR_PTR(-ENOMEM); + + INIT_LIST_HEAD(&watch->rules); + atomic_set(&watch->count, 1); + watch->path = path; + watch->dev = (dev_t)-1; + watch->ino = (unsigned long)-1; + + return watch; +} + /* Initialize an audit filterlist entry. */ static inline struct audit_entry *audit_init_entry(u32 field_count) { @@ -107,6 +241,43 @@ static char *audit_unpack_string(void **bufp, size_t *remain, size_t len) return str; } +/* Translate an inode field to kernel respresentation. */ +static inline int audit_to_inode(struct audit_krule *krule, + struct audit_field *f) +{ + if (krule->listnr != AUDIT_FILTER_EXIT || + krule->watch || krule->inode_f) + return -EINVAL; + + krule->inode_f = f; + return 0; +} + +/* Translate a watch string to kernel respresentation. */ +static int audit_to_watch(struct audit_krule *krule, char *path, int len, + u32 op) +{ + struct audit_watch *watch; + + if (!audit_ih) + return -EOPNOTSUPP; + + if (path[0] != '/' || path[len-1] == '/' || + krule->listnr != AUDIT_FILTER_EXIT || + op & ~AUDIT_EQUAL || + krule->inode_f || krule->watch) /* 1 inode # per rule, for hash */ + return -EINVAL; + + watch = audit_init_watch(path); + if (unlikely(IS_ERR(watch))) + return PTR_ERR(watch); + + audit_get_watch(watch); + krule->watch = watch; + + return 0; +} + /* Common user-space to kernel rule translation. */ static inline struct audit_entry *audit_to_entry_common(struct audit_rule *rule) { @@ -128,8 +299,11 @@ static inline struct audit_entry *audit_to_entry_common(struct audit_rule *rule) #endif ; } - if (rule->action != AUDIT_NEVER && rule->action != AUDIT_POSSIBLE && - rule->action != AUDIT_ALWAYS) + if (unlikely(rule->action == AUDIT_POSSIBLE)) { + printk(KERN_ERR "AUDIT_POSSIBLE is deprecated\n"); + goto exit_err; + } + if (rule->action != AUDIT_NEVER && rule->action != AUDIT_ALWAYS) goto exit_err; if (rule->field_count > AUDIT_MAX_FIELDS) goto exit_err; @@ -158,6 +332,7 @@ exit_err: static struct audit_entry *audit_rule_to_entry(struct audit_rule *rule) { struct audit_entry *entry; + struct audit_field *f; int err = 0; int i; @@ -172,14 +347,37 @@ static struct audit_entry *audit_rule_to_entry(struct audit_rule *rule) f->type = rule->fields[i] & ~(AUDIT_NEGATE|AUDIT_OPERATORS); f->val = rule->values[i]; - if (f->type & AUDIT_UNUSED_BITS || - f->type == AUDIT_SE_USER || - f->type == AUDIT_SE_ROLE || - f->type == AUDIT_SE_TYPE || - f->type == AUDIT_SE_SEN || - f->type == AUDIT_SE_CLR) { - err = -EINVAL; + err = -EINVAL; + switch(f->type) { + default: goto exit_free; + case AUDIT_PID: + case AUDIT_UID: + case AUDIT_EUID: + case AUDIT_SUID: + case AUDIT_FSUID: + case AUDIT_GID: + case AUDIT_EGID: + case AUDIT_SGID: + case AUDIT_FSGID: + case AUDIT_LOGINUID: + case AUDIT_PERS: + case AUDIT_ARCH: + case AUDIT_MSGTYPE: + case AUDIT_DEVMAJOR: + case AUDIT_DEVMINOR: + case AUDIT_EXIT: + case AUDIT_SUCCESS: + case AUDIT_ARG0: + case AUDIT_ARG1: + case AUDIT_ARG2: + case AUDIT_ARG3: + break; + case AUDIT_INODE: + err = audit_to_inode(&entry->rule, f); + if (err) + goto exit_free; + break; } entry->rule.vers_ops = (f->op & AUDIT_OPERATORS) ? 2 : 1; @@ -196,6 +394,18 @@ static struct audit_entry *audit_rule_to_entry(struct audit_rule *rule) } } + f = entry->rule.inode_f; + if (f) { + switch(f->op) { + case AUDIT_NOT_EQUAL: + entry->rule.inode_f = NULL; + case AUDIT_EQUAL: + break; + default: + goto exit_free; + } + } + exit_nofree: return entry; @@ -210,6 +420,7 @@ static struct audit_entry *audit_data_to_entry(struct audit_rule_data *data, { int err = 0; struct audit_entry *entry; + struct audit_field *f; void *bufp; size_t remain = datasz - sizeof(struct audit_rule_data); int i; @@ -235,6 +446,29 @@ static struct audit_entry *audit_data_to_entry(struct audit_rule_data *data, f->se_str = NULL; f->se_rule = NULL; switch(f->type) { + case AUDIT_PID: + case AUDIT_UID: + case AUDIT_EUID: + case AUDIT_SUID: + case AUDIT_FSUID: + case AUDIT_GID: + case AUDIT_EGID: + case AUDIT_SGID: + case AUDIT_FSGID: + case AUDIT_LOGINUID: + case AUDIT_PERS: + case AUDIT_ARCH: + case AUDIT_MSGTYPE: + case AUDIT_PPID: + case AUDIT_DEVMAJOR: + case AUDIT_DEVMINOR: + case AUDIT_EXIT: + case AUDIT_SUCCESS: + case AUDIT_ARG0: + case AUDIT_ARG1: + case AUDIT_ARG2: + case AUDIT_ARG3: + break; case AUDIT_SE_USER: case AUDIT_SE_ROLE: case AUDIT_SE_TYPE: @@ -260,6 +494,37 @@ static struct audit_entry *audit_data_to_entry(struct audit_rule_data *data, } else f->se_str = str; break; + case AUDIT_WATCH: + str = audit_unpack_string(&bufp, &remain, f->val); + if (IS_ERR(str)) + goto exit_free; + entry->rule.buflen += f->val; + + err = audit_to_watch(&entry->rule, str, f->val, f->op); + if (err) { + kfree(str); + goto exit_free; + } + break; + case AUDIT_INODE: + err = audit_to_inode(&entry->rule, f); + if (err) + goto exit_free; + break; + default: + goto exit_free; + } + } + + f = entry->rule.inode_f; + if (f) { + switch(f->op) { + case AUDIT_NOT_EQUAL: + entry->rule.inode_f = NULL; + case AUDIT_EQUAL: + break; + default: + goto exit_free; } } @@ -291,7 +556,7 @@ static struct audit_rule *audit_krule_to_rule(struct audit_krule *krule) rule = kmalloc(sizeof(*rule), GFP_KERNEL); if (unlikely(!rule)) - return ERR_PTR(-ENOMEM); + return NULL; memset(rule, 0, sizeof(*rule)); rule->flags = krule->flags | krule->listnr; @@ -322,7 +587,7 @@ static struct audit_rule_data *audit_krule_to_data(struct audit_krule *krule) data = kmalloc(sizeof(*data) + krule->buflen, GFP_KERNEL); if (unlikely(!data)) - return ERR_PTR(-ENOMEM); + return NULL; memset(data, 0, sizeof(*data)); data->flags = krule->flags | krule->listnr; @@ -343,6 +608,10 @@ static struct audit_rule_data *audit_krule_to_data(struct audit_krule *krule) data->buflen += data->values[i] = audit_pack_string(&bufp, f->se_str); break; + case AUDIT_WATCH: + data->buflen += data->values[i] = + audit_pack_string(&bufp, krule->watch->path); + break; default: data->values[i] = f->val; } @@ -378,6 +647,10 @@ static int audit_compare_rule(struct audit_krule *a, struct audit_krule *b) if (strcmp(a->fields[i].se_str, b->fields[i].se_str)) return 1; break; + case AUDIT_WATCH: + if (strcmp(a->watch->path, b->watch->path)) + return 1; + break; default: if (a->fields[i].val != b->fields[i].val) return 1; @@ -391,6 +664,32 @@ static int audit_compare_rule(struct audit_krule *a, struct audit_krule *b) return 0; } +/* Duplicate the given audit watch. The new watch's rules list is initialized + * to an empty list and wlist is undefined. */ +static struct audit_watch *audit_dupe_watch(struct audit_watch *old) +{ + char *path; + struct audit_watch *new; + + path = kstrdup(old->path, GFP_KERNEL); + if (unlikely(!path)) + return ERR_PTR(-ENOMEM); + + new = audit_init_watch(path); + if (unlikely(IS_ERR(new))) { + kfree(path); + goto out; + } + + new->dev = old->dev; + new->ino = old->ino; + get_inotify_watch(&old->parent->wdata); + new->parent = old->parent; + +out: + return new; +} + /* Duplicate selinux field information. The se_rule is opaque, so must be * re-initialized. */ static inline int audit_dupe_selinux_field(struct audit_field *df, @@ -422,8 +721,11 @@ static inline int audit_dupe_selinux_field(struct audit_field *df, /* Duplicate an audit rule. This will be a deep copy with the exception * of the watch - that pointer is carried over. The selinux specific fields * will be updated in the copy. The point is to be able to replace the old - * rule with the new rule in the filterlist, then free the old rule. */ -static struct audit_entry *audit_dupe_rule(struct audit_krule *old) + * rule with the new rule in the filterlist, then free the old rule. + * The rlist element is undefined; list manipulations are handled apart from + * the initial copy. */ +static struct audit_entry *audit_dupe_rule(struct audit_krule *old, + struct audit_watch *watch) { u32 fcount = old->field_count; struct audit_entry *entry; @@ -442,6 +744,8 @@ static struct audit_entry *audit_dupe_rule(struct audit_krule *old) for (i = 0; i < AUDIT_BITMASK_SIZE; i++) new->mask[i] = old->mask[i]; new->buflen = old->buflen; + new->inode_f = old->inode_f; + new->watch = NULL; new->field_count = old->field_count; memcpy(new->fields, old->fields, sizeof(struct audit_field) * fcount); @@ -463,68 +767,409 @@ static struct audit_entry *audit_dupe_rule(struct audit_krule *old) } } + if (watch) { + audit_get_watch(watch); + new->watch = watch; + } + return entry; } -/* Add rule to given filterlist if not a duplicate. Protected by - * audit_netlink_mutex. */ +/* Update inode info in audit rules based on filesystem event. */ +static void audit_update_watch(struct audit_parent *parent, + const char *dname, dev_t dev, + unsigned long ino, unsigned invalidating) +{ + struct audit_watch *owatch, *nwatch, *nextw; + struct audit_krule *r, *nextr; + struct audit_entry *oentry, *nentry; + struct audit_buffer *ab; + + mutex_lock(&audit_filter_mutex); + list_for_each_entry_safe(owatch, nextw, &parent->watches, wlist) { + if (audit_compare_dname_path(dname, owatch->path, NULL)) + continue; + + /* If the update involves invalidating rules, do the inode-based + * filtering now, so we don't omit records. */ + if (invalidating && + audit_filter_inodes(current, current->audit_context) == AUDIT_RECORD_CONTEXT) + audit_set_auditable(current->audit_context); + + nwatch = audit_dupe_watch(owatch); + if (unlikely(IS_ERR(nwatch))) { + mutex_unlock(&audit_filter_mutex); + audit_panic("error updating watch, skipping"); + return; + } + nwatch->dev = dev; + nwatch->ino = ino; + + list_for_each_entry_safe(r, nextr, &owatch->rules, rlist) { + + oentry = container_of(r, struct audit_entry, rule); + list_del(&oentry->rule.rlist); + list_del_rcu(&oentry->list); + + nentry = audit_dupe_rule(&oentry->rule, nwatch); + if (unlikely(IS_ERR(nentry))) + audit_panic("error updating watch, removing"); + else { + int h = audit_hash_ino((u32)ino); + list_add(&nentry->rule.rlist, &nwatch->rules); + list_add_rcu(&nentry->list, &audit_inode_hash[h]); + } + + call_rcu(&oentry->rcu, audit_free_rule_rcu); + } + + ab = audit_log_start(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE); + audit_log_format(ab, "audit updated rules specifying watch="); + audit_log_untrustedstring(ab, owatch->path); + audit_log_format(ab, " with dev=%u ino=%lu\n", dev, ino); + audit_log_end(ab); + + audit_remove_watch(owatch); + goto add_watch_to_parent; /* event applies to a single watch */ + } + mutex_unlock(&audit_filter_mutex); + return; + +add_watch_to_parent: + list_add(&nwatch->wlist, &parent->watches); + mutex_unlock(&audit_filter_mutex); + return; +} + +/* Remove all watches & rules associated with a parent that is going away. */ +static void audit_remove_parent_watches(struct audit_parent *parent) +{ + struct audit_watch *w, *nextw; + struct audit_krule *r, *nextr; + struct audit_entry *e; + + mutex_lock(&audit_filter_mutex); + parent->flags |= AUDIT_PARENT_INVALID; + list_for_each_entry_safe(w, nextw, &parent->watches, wlist) { + list_for_each_entry_safe(r, nextr, &w->rules, rlist) { + e = container_of(r, struct audit_entry, rule); + list_del(&r->rlist); + list_del_rcu(&e->list); + call_rcu(&e->rcu, audit_free_rule_rcu); + + audit_log(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE, + "audit implicitly removed rule from list=%d\n", + AUDIT_FILTER_EXIT); + } + audit_remove_watch(w); + } + mutex_unlock(&audit_filter_mutex); +} + +/* Unregister inotify watches for parents on in_list. + * Generates an IN_IGNORED event. */ +static void audit_inotify_unregister(struct list_head *in_list) +{ + struct audit_parent *p, *n; + + list_for_each_entry_safe(p, n, in_list, ilist) { + list_del(&p->ilist); + inotify_rm_watch(audit_ih, &p->wdata); + /* the put matching the get in audit_do_del_rule() */ + put_inotify_watch(&p->wdata); + } +} + +/* Find an existing audit rule. + * Caller must hold audit_filter_mutex to prevent stale rule data. */ +static struct audit_entry *audit_find_rule(struct audit_entry *entry, + struct list_head *list) +{ + struct audit_entry *e, *found = NULL; + int h; + + if (entry->rule.watch) { + /* we don't know the inode number, so must walk entire hash */ + for (h = 0; h < AUDIT_INODE_BUCKETS; h++) { + list = &audit_inode_hash[h]; + list_for_each_entry(e, list, list) + if (!audit_compare_rule(&entry->rule, &e->rule)) { + found = e; + goto out; + } + } + goto out; + } + + list_for_each_entry(e, list, list) + if (!audit_compare_rule(&entry->rule, &e->rule)) { + found = e; + goto out; + } + +out: + return found; +} + +/* Get path information necessary for adding watches. */ +static int audit_get_nd(char *path, struct nameidata **ndp, + struct nameidata **ndw) +{ + struct nameidata *ndparent, *ndwatch; + int err; + + ndparent = kmalloc(sizeof(*ndparent), GFP_KERNEL); + if (unlikely(!ndparent)) + return -ENOMEM; + + ndwatch = kmalloc(sizeof(*ndwatch), GFP_KERNEL); + if (unlikely(!ndwatch)) { + kfree(ndparent); + return -ENOMEM; + } + + err = path_lookup(path, LOOKUP_PARENT, ndparent); + if (err) { + kfree(ndparent); + kfree(ndwatch); + return err; + } + + err = path_lookup(path, 0, ndwatch); + if (err) { + kfree(ndwatch); + ndwatch = NULL; + } + + *ndp = ndparent; + *ndw = ndwatch; + + return 0; +} + +/* Release resources used for watch path information. */ +static void audit_put_nd(struct nameidata *ndp, struct nameidata *ndw) +{ + if (ndp) { + path_release(ndp); + kfree(ndp); + } + if (ndw) { + path_release(ndw); + kfree(ndw); + } +} + +/* Associate the given rule with an existing parent inotify_watch. + * Caller must hold audit_filter_mutex. */ +static void audit_add_to_parent(struct audit_krule *krule, + struct audit_parent *parent) +{ + struct audit_watch *w, *watch = krule->watch; + int watch_found = 0; + + list_for_each_entry(w, &parent->watches, wlist) { + if (strcmp(watch->path, w->path)) + continue; + + watch_found = 1; + + /* put krule's and initial refs to temporary watch */ + audit_put_watch(watch); + audit_put_watch(watch); + + audit_get_watch(w); + krule->watch = watch = w; + break; + } + + if (!watch_found) { + get_inotify_watch(&parent->wdata); + watch->parent = parent; + + list_add(&watch->wlist, &parent->watches); + } + list_add(&krule->rlist, &watch->rules); +} + +/* Find a matching watch entry, or add this one. + * Caller must hold audit_filter_mutex. */ +static int audit_add_watch(struct audit_krule *krule, struct nameidata *ndp, + struct nameidata *ndw) +{ + struct audit_watch *watch = krule->watch; + struct inotify_watch *i_watch; + struct audit_parent *parent; + int ret = 0; + + /* update watch filter fields */ + if (ndw) { + watch->dev = ndw->dentry->d_inode->i_sb->s_dev; + watch->ino = ndw->dentry->d_inode->i_ino; + } + + /* The audit_filter_mutex must not be held during inotify calls because + * we hold it during inotify event callback processing. If an existing + * inotify watch is found, inotify_find_watch() grabs a reference before + * returning. + */ + mutex_unlock(&audit_filter_mutex); + + if (inotify_find_watch(audit_ih, ndp->dentry->d_inode, &i_watch) < 0) { + parent = audit_init_parent(ndp); + if (IS_ERR(parent)) { + /* caller expects mutex locked */ + mutex_lock(&audit_filter_mutex); + return PTR_ERR(parent); + } + } else + parent = container_of(i_watch, struct audit_parent, wdata); + + mutex_lock(&audit_filter_mutex); + + /* parent was moved before we took audit_filter_mutex */ + if (parent->flags & AUDIT_PARENT_INVALID) + ret = -ENOENT; + else + audit_add_to_parent(krule, parent); + + /* match get in audit_init_parent or inotify_find_watch */ + put_inotify_watch(&parent->wdata); + return ret; +} + +/* Add rule to given filterlist if not a duplicate. */ static inline int audit_add_rule(struct audit_entry *entry, - struct list_head *list) + struct list_head *list) { struct audit_entry *e; + struct audit_field *inode_f = entry->rule.inode_f; + struct audit_watch *watch = entry->rule.watch; + struct nameidata *ndp, *ndw; + int h, err, putnd_needed = 0; + + if (inode_f) { + h = audit_hash_ino(inode_f->val); + list = &audit_inode_hash[h]; + } - /* Do not use the _rcu iterator here, since this is the only - * addition routine. */ - list_for_each_entry(e, list, list) { - if (!audit_compare_rule(&entry->rule, &e->rule)) - return -EEXIST; + mutex_lock(&audit_filter_mutex); + e = audit_find_rule(entry, list); + mutex_unlock(&audit_filter_mutex); + if (e) { + err = -EEXIST; + goto error; + } + + /* Avoid calling path_lookup under audit_filter_mutex. */ + if (watch) { + err = audit_get_nd(watch->path, &ndp, &ndw); + if (err) + goto error; + putnd_needed = 1; + } + + mutex_lock(&audit_filter_mutex); + if (watch) { + /* audit_filter_mutex is dropped and re-taken during this call */ + err = audit_add_watch(&entry->rule, ndp, ndw); + if (err) { + mutex_unlock(&audit_filter_mutex); + goto error; + } + h = audit_hash_ino((u32)watch->ino); + list = &audit_inode_hash[h]; } if (entry->rule.flags & AUDIT_FILTER_PREPEND) { list_add_rcu(&entry->list, list); + entry->rule.flags &= ~AUDIT_FILTER_PREPEND; } else { list_add_tail_rcu(&entry->list, list); } + mutex_unlock(&audit_filter_mutex); - return 0; + if (putnd_needed) + audit_put_nd(ndp, ndw); + + return 0; + +error: + if (putnd_needed) + audit_put_nd(ndp, ndw); + if (watch) + audit_put_watch(watch); /* tmp watch, matches initial get */ + return err; } -/* Remove an existing rule from filterlist. Protected by - * audit_netlink_mutex. */ +/* Remove an existing rule from filterlist. */ static inline int audit_del_rule(struct audit_entry *entry, struct list_head *list) { struct audit_entry *e; + struct audit_field *inode_f = entry->rule.inode_f; + struct audit_watch *watch, *tmp_watch = entry->rule.watch; + LIST_HEAD(inotify_list); + int h, ret = 0; + + if (inode_f) { + h = audit_hash_ino(inode_f->val); + list = &audit_inode_hash[h]; + } - /* Do not use the _rcu iterator here, since this is the only - * deletion routine. */ - list_for_each_entry(e, list, list) { - if (!audit_compare_rule(&entry->rule, &e->rule)) { - list_del_rcu(&e->list); - call_rcu(&e->rcu, audit_free_rule_rcu); - return 0; + mutex_lock(&audit_filter_mutex); + e = audit_find_rule(entry, list); + if (!e) { + mutex_unlock(&audit_filter_mutex); + ret = -ENOENT; + goto out; + } + + watch = e->rule.watch; + if (watch) { + struct audit_parent *parent = watch->parent; + + list_del(&e->rule.rlist); + + if (list_empty(&watch->rules)) { + audit_remove_watch(watch); + + if (list_empty(&parent->watches)) { + /* Put parent on the inotify un-registration + * list. Grab a reference before releasing + * audit_filter_mutex, to be released in + * audit_inotify_unregister(). */ + list_add(&parent->ilist, &inotify_list); + get_inotify_watch(&parent->wdata); + } } } - return -ENOENT; /* No matching rule */ + + list_del_rcu(&e->list); + call_rcu(&e->rcu, audit_free_rule_rcu); + + mutex_unlock(&audit_filter_mutex); + + if (!list_empty(&inotify_list)) + audit_inotify_unregister(&inotify_list); + +out: + if (tmp_watch) + audit_put_watch(tmp_watch); /* match initial get */ + + return ret; } /* List rules using struct audit_rule. Exists for backward * compatibility with userspace. */ -static int audit_list(void *_dest) +static void audit_list(int pid, int seq, struct sk_buff_head *q) { - int pid, seq; - int *dest = _dest; + struct sk_buff *skb; struct audit_entry *entry; int i; - pid = dest[0]; - seq = dest[1]; - kfree(dest); - - mutex_lock(&audit_netlink_mutex); - - /* The *_rcu iterators not needed here because we are - always called with audit_netlink_mutex held. */ + /* This is a blocking read, so use audit_filter_mutex instead of rcu + * iterator to sync with list writers. */ for (i=0; i<AUDIT_NR_FILTERS; i++) { list_for_each_entry(entry, &audit_filter_list[i], list) { struct audit_rule *rule; @@ -532,33 +1177,41 @@ static int audit_list(void *_dest) rule = audit_krule_to_rule(&entry->rule); if (unlikely(!rule)) break; - audit_send_reply(pid, seq, AUDIT_LIST, 0, 1, + skb = audit_make_reply(pid, seq, AUDIT_LIST, 0, 1, rule, sizeof(*rule)); + if (skb) + skb_queue_tail(q, skb); kfree(rule); } } - audit_send_reply(pid, seq, AUDIT_LIST, 1, 1, NULL, 0); - - mutex_unlock(&audit_netlink_mutex); - return 0; + for (i = 0; i < AUDIT_INODE_BUCKETS; i++) { + list_for_each_entry(entry, &audit_inode_hash[i], list) { + struct audit_rule *rule; + + rule = audit_krule_to_rule(&entry->rule); + if (unlikely(!rule)) + break; + skb = audit_make_reply(pid, seq, AUDIT_LIST, 0, 1, + rule, sizeof(*rule)); + if (skb) + skb_queue_tail(q, skb); + kfree(rule); + } + } + skb = audit_make_reply(pid, seq, AUDIT_LIST, 1, 1, NULL, 0); + if (skb) + skb_queue_tail(q, skb); } /* List rules using struct audit_rule_data. */ -static int audit_list_rules(void *_dest) +static void audit_list_rules(int pid, int seq, struct sk_buff_head *q) { - int pid, seq; - int *dest = _dest; + struct sk_buff *skb; struct audit_entry *e; int i; - pid = dest[0]; - seq = dest[1]; - kfree(dest); - - mutex_lock(&audit_netlink_mutex); - - /* The *_rcu iterators not needed here because we are - always called with audit_netlink_mutex held. */ + /* This is a blocking read, so use audit_filter_mutex instead of rcu + * iterator to sync with list writers. */ for (i=0; i<AUDIT_NR_FILTERS; i++) { list_for_each_entry(e, &audit_filter_list[i], list) { struct audit_rule_data *data; @@ -566,15 +1219,30 @@ static int audit_list_rules(void *_dest) data = audit_krule_to_data(&e->rule); if (unlikely(!data)) break; - audit_send_reply(pid, seq, AUDIT_LIST_RULES, 0, 1, - data, sizeof(*data)); + skb = audit_make_reply(pid, seq, AUDIT_LIST_RULES, 0, 1, + data, sizeof(*data) + data->buflen); + if (skb) + skb_queue_tail(q, skb); kfree(data); } } - audit_send_reply(pid, seq, AUDIT_LIST_RULES, 1, 1, NULL, 0); + for (i=0; i< AUDIT_INODE_BUCKETS; i++) { + list_for_each_entry(e, &audit_inode_hash[i], list) { + struct audit_rule_data *data; - mutex_unlock(&audit_netlink_mutex); - return 0; + data = audit_krule_to_data(&e->rule); + if (unlikely(!data)) + break; + skb = audit_make_reply(pid, seq, AUDIT_LIST_RULES, 0, 1, + data, sizeof(*data) + data->buflen); + if (skb) + skb_queue_tail(q, skb); + kfree(data); + } + } + skb = audit_make_reply(pid, seq, AUDIT_LIST_RULES, 1, 1, NULL, 0); + if (skb) + skb_queue_tail(q, skb); } /** @@ -592,7 +1260,7 @@ int audit_receive_filter(int type, int pid, int uid, int seq, void *data, size_t datasz, uid_t loginuid, u32 sid) { struct task_struct *tsk; - int *dest; + struct audit_netlink_list *dest; int err = 0; struct audit_entry *entry; @@ -605,18 +1273,22 @@ int audit_receive_filter(int type, int pid, int uid, int seq, void *data, * happen if we're actually running in the context of auditctl * trying to _send_ the stuff */ - dest = kmalloc(2 * sizeof(int), GFP_KERNEL); + dest = kmalloc(sizeof(struct audit_netlink_list), GFP_KERNEL); if (!dest) return -ENOMEM; - dest[0] = pid; - dest[1] = seq; + dest->pid = pid; + skb_queue_head_init(&dest->q); + mutex_lock(&audit_filter_mutex); if (type == AUDIT_LIST) - tsk = kthread_run(audit_list, dest, "audit_list"); + audit_list(pid, seq, &dest->q); else - tsk = kthread_run(audit_list_rules, dest, - "audit_list_rules"); + audit_list_rules(pid, seq, &dest->q); + mutex_unlock(&audit_filter_mutex); + + tsk = kthread_run(audit_send_list, dest, "audit_send_list"); if (IS_ERR(tsk)) { + skb_queue_purge(&dest->q); kfree(dest); err = PTR_ERR(tsk); } @@ -632,6 +1304,7 @@ int audit_receive_filter(int type, int pid, int uid, int seq, void *data, err = audit_add_rule(entry, &audit_filter_list[entry->rule.listnr]); + if (sid) { char *ctx = NULL; u32 len; @@ -712,7 +1385,43 @@ int audit_comparator(const u32 left, const u32 op, const u32 right) return 0; } +/* Compare given dentry name with last component in given path, + * return of 0 indicates a match. */ +int audit_compare_dname_path(const char *dname, const char *path, + int *dirlen) +{ + int dlen, plen; + const char *p; + if (!dname || !path) + return 1; + + dlen = strlen(dname); + plen = strlen(path); + if (plen < dlen) + return 1; + + /* disregard trailing slashes */ + p = path + plen - 1; + while ((*p == '/') && (p > path)) + p--; + + /* find last path component */ + p = p - dlen + 1; + if (p < path) + return 1; + else if (p > path) { + if (*--p != '/') + return 1; + else + p++; + } + + /* return length of path's directory component */ + if (dirlen) + *dirlen = p - path; + return strncmp(p, dname, dlen); +} static int audit_filter_user_rules(struct netlink_skb_parms *cb, struct audit_krule *rule, @@ -744,7 +1453,6 @@ static int audit_filter_user_rules(struct netlink_skb_parms *cb, } switch (rule->action) { case AUDIT_NEVER: *state = AUDIT_DISABLED; break; - case AUDIT_POSSIBLE: *state = AUDIT_BUILD_CONTEXT; break; case AUDIT_ALWAYS: *state = AUDIT_RECORD_CONTEXT; break; } return 1; @@ -826,32 +1534,65 @@ static inline int audit_rule_has_selinux(struct audit_krule *rule) int selinux_audit_rule_update(void) { struct audit_entry *entry, *n, *nentry; + struct audit_watch *watch; int i, err = 0; - /* audit_netlink_mutex synchronizes the writers */ - mutex_lock(&audit_netlink_mutex); + /* audit_filter_mutex synchronizes the writers */ + mutex_lock(&audit_filter_mutex); for (i = 0; i < AUDIT_NR_FILTERS; i++) { list_for_each_entry_safe(entry, n, &audit_filter_list[i], list) { if (!audit_rule_has_selinux(&entry->rule)) continue; - nentry = audit_dupe_rule(&entry->rule); + watch = entry->rule.watch; + nentry = audit_dupe_rule(&entry->rule, watch); if (unlikely(IS_ERR(nentry))) { /* save the first error encountered for the * return value */ if (!err) err = PTR_ERR(nentry); audit_panic("error updating selinux filters"); + if (watch) + list_del(&entry->rule.rlist); list_del_rcu(&entry->list); } else { + if (watch) { + list_add(&nentry->rule.rlist, + &watch->rules); + list_del(&entry->rule.rlist); + } list_replace_rcu(&entry->list, &nentry->list); } call_rcu(&entry->rcu, audit_free_rule_rcu); } } - mutex_unlock(&audit_netlink_mutex); + mutex_unlock(&audit_filter_mutex); return err; } + +/* Update watch data in audit rules based on inotify events. */ +void audit_handle_ievent(struct inotify_watch *i_watch, u32 wd, u32 mask, + u32 cookie, const char *dname, struct inode *inode) +{ + struct audit_parent *parent; + + parent = container_of(i_watch, struct audit_parent, wdata); + + if (mask & (IN_CREATE|IN_MOVED_TO) && inode) + audit_update_watch(parent, dname, inode->i_sb->s_dev, + inode->i_ino, 0); + else if (mask & (IN_DELETE|IN_MOVED_FROM)) + audit_update_watch(parent, dname, (dev_t)-1, (unsigned long)-1, 1); + /* inotify automatically removes the watch and sends IN_IGNORED */ + else if (mask & (IN_DELETE_SELF|IN_UNMOUNT)) + audit_remove_parent_watches(parent); + /* inotify does not remove the watch, so remove it manually */ + else if(mask & IN_MOVE_SELF) { + audit_remove_parent_watches(parent); + inotify_remove_watch_locked(audit_ih, i_watch); + } else if (mask & IN_IGNORED) + put_inotify_watch(i_watch); +} diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 1c03a4ed1b27..9ebd96fda295 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -3,7 +3,7 @@ * * Copyright 2003-2004 Red Hat Inc., Durham, North Carolina. * Copyright 2005 Hewlett-Packard Development Company, L.P. - * Copyright (C) 2005 IBM Corporation + * Copyright (C) 2005, 2006 IBM Corporation * All Rights Reserved. * * This program is free software; you can redistribute it and/or modify @@ -29,6 +29,9 @@ * this file -- see entry.S) is based on a GPL'd patch written by * okir@suse.de and Copyright 2003 SuSE Linux AG. * + * POSIX message queue support added by George Wilson <ltcgcw@us.ibm.com>, + * 2006. + * * The support of additional filter rules compares (>, <, >=, <=) was * added by Dustin Kirkland <dustin.kirkland@us.ibm.com>, 2005. * @@ -49,6 +52,7 @@ #include <linux/module.h> #include <linux/mount.h> #include <linux/socket.h> +#include <linux/mqueue.h> #include <linux/audit.h> #include <linux/personality.h> #include <linux/time.h> @@ -59,6 +63,8 @@ #include <linux/list.h> #include <linux/tty.h> #include <linux/selinux.h> +#include <linux/binfmts.h> +#include <linux/syscalls.h> #include "audit.h" @@ -76,6 +82,9 @@ extern int audit_enabled; * path_lookup. */ #define AUDIT_NAMES_RESERVED 7 +/* Indicates that audit should log the full pathname. */ +#define AUDIT_NAME_FULL -1 + /* When fs/namei.c:getname() is called, we store the pointer in name and * we don't let putname() free it (instead we free all of the saved * pointers at syscall exit time). @@ -83,8 +92,9 @@ extern int audit_enabled; * Further, in fs/namei.c:path_lookup() we store the inode and device. */ struct audit_names { const char *name; + int name_len; /* number of name's characters to log */ + unsigned name_put; /* call __putname() for this name */ unsigned long ino; - unsigned long pino; dev_t dev; umode_t mode; uid_t uid; @@ -100,6 +110,33 @@ struct audit_aux_data { #define AUDIT_AUX_IPCPERM 0 +struct audit_aux_data_mq_open { + struct audit_aux_data d; + int oflag; + mode_t mode; + struct mq_attr attr; +}; + +struct audit_aux_data_mq_sendrecv { + struct audit_aux_data d; + mqd_t mqdes; + size_t msg_len; + unsigned int msg_prio; + struct timespec abs_timeout; +}; + +struct audit_aux_data_mq_notify { + struct audit_aux_data d; + mqd_t mqdes; + struct sigevent notification; +}; + +struct audit_aux_data_mq_getsetattr { + struct audit_aux_data d; + mqd_t mqdes; + struct mq_attr mqstat; +}; + struct audit_aux_data_ipcctl { struct audit_aux_data d; struct ipc_perm p; @@ -110,6 +147,13 @@ struct audit_aux_data_ipcctl { u32 osid; }; +struct audit_aux_data_execve { + struct audit_aux_data d; + int argc; + int envc; + char mem[0]; +}; + struct audit_aux_data_socketcall { struct audit_aux_data d; int nargs; @@ -148,7 +192,7 @@ struct audit_context { struct audit_aux_data *aux; /* Save things to print about task_struct */ - pid_t pid; + pid_t pid, ppid; uid_t uid, euid, suid, fsuid; gid_t gid, egid, sgid, fsgid; unsigned long personality; @@ -160,12 +204,13 @@ struct audit_context { #endif }; - +/* Determine if any context name data matches a rule's watch data */ /* Compare a task_struct with an audit_rule. Return 1 on match, 0 * otherwise. */ static int audit_filter_rules(struct task_struct *tsk, struct audit_krule *rule, struct audit_context *ctx, + struct audit_names *name, enum audit_state *state) { int i, j, need_sid = 1; @@ -179,6 +224,10 @@ static int audit_filter_rules(struct task_struct *tsk, case AUDIT_PID: result = audit_comparator(tsk->pid, f->op, f->val); break; + case AUDIT_PPID: + if (ctx) + result = audit_comparator(ctx->ppid, f->op, f->val); + break; case AUDIT_UID: result = audit_comparator(tsk->uid, f->op, f->val); break; @@ -224,7 +273,10 @@ static int audit_filter_rules(struct task_struct *tsk, } break; case AUDIT_DEVMAJOR: - if (ctx) { + if (name) + result = audit_comparator(MAJOR(name->dev), + f->op, f->val); + else if (ctx) { for (j = 0; j < ctx->name_count; j++) { if (audit_comparator(MAJOR(ctx->names[j].dev), f->op, f->val)) { ++result; @@ -234,7 +286,10 @@ static int audit_filter_rules(struct task_struct *tsk, } break; case AUDIT_DEVMINOR: - if (ctx) { + if (name) + result = audit_comparator(MINOR(name->dev), + f->op, f->val); + else if (ctx) { for (j = 0; j < ctx->name_count; j++) { if (audit_comparator(MINOR(ctx->names[j].dev), f->op, f->val)) { ++result; @@ -244,16 +299,22 @@ static int audit_filter_rules(struct task_struct *tsk, } break; case AUDIT_INODE: - if (ctx) { + if (name) + result = (name->ino == f->val); + else if (ctx) { for (j = 0; j < ctx->name_count; j++) { - if (audit_comparator(ctx->names[j].ino, f->op, f->val) || - audit_comparator(ctx->names[j].pino, f->op, f->val)) { + if (audit_comparator(ctx->names[j].ino, f->op, f->val)) { ++result; break; } } } break; + case AUDIT_WATCH: + if (name && rule->watch->ino != (unsigned long)-1) + result = (name->dev == rule->watch->dev && + name->ino == rule->watch->ino); + break; case AUDIT_LOGINUID: result = 0; if (ctx) @@ -294,7 +355,6 @@ static int audit_filter_rules(struct task_struct *tsk, } switch (rule->action) { case AUDIT_NEVER: *state = AUDIT_DISABLED; break; - case AUDIT_POSSIBLE: *state = AUDIT_BUILD_CONTEXT; break; case AUDIT_ALWAYS: *state = AUDIT_RECORD_CONTEXT; break; } return 1; @@ -311,7 +371,7 @@ static enum audit_state audit_filter_task(struct task_struct *tsk) rcu_read_lock(); list_for_each_entry_rcu(e, &audit_filter_list[AUDIT_FILTER_TASK], list) { - if (audit_filter_rules(tsk, &e->rule, NULL, &state)) { + if (audit_filter_rules(tsk, &e->rule, NULL, NULL, &state)) { rcu_read_unlock(); return state; } @@ -341,8 +401,47 @@ static enum audit_state audit_filter_syscall(struct task_struct *tsk, int bit = AUDIT_BIT(ctx->major); list_for_each_entry_rcu(e, list, list) { - if ((e->rule.mask[word] & bit) == bit - && audit_filter_rules(tsk, &e->rule, ctx, &state)) { + if ((e->rule.mask[word] & bit) == bit && + audit_filter_rules(tsk, &e->rule, ctx, NULL, + &state)) { + rcu_read_unlock(); + return state; + } + } + } + rcu_read_unlock(); + return AUDIT_BUILD_CONTEXT; +} + +/* At syscall exit time, this filter is called if any audit_names[] have been + * collected during syscall processing. We only check rules in sublists at hash + * buckets applicable to the inode numbers in audit_names[]. + * Regarding audit_state, same rules apply as for audit_filter_syscall(). + */ +enum audit_state audit_filter_inodes(struct task_struct *tsk, + struct audit_context *ctx) +{ + int i; + struct audit_entry *e; + enum audit_state state; + + if (audit_pid && tsk->tgid == audit_pid) + return AUDIT_DISABLED; + + rcu_read_lock(); + for (i = 0; i < ctx->name_count; i++) { + int word = AUDIT_WORD(ctx->major); + int bit = AUDIT_BIT(ctx->major); + struct audit_names *n = &ctx->names[i]; + int h = audit_hash_ino((u32)n->ino); + struct list_head *list = &audit_inode_hash[h]; + + if (list_empty(list)) + continue; + + list_for_each_entry_rcu(e, list, list) { + if ((e->rule.mask[word] & bit) == bit && + audit_filter_rules(tsk, &e->rule, ctx, n, &state)) { rcu_read_unlock(); return state; } @@ -352,6 +451,11 @@ static enum audit_state audit_filter_syscall(struct task_struct *tsk, return AUDIT_BUILD_CONTEXT; } +void audit_set_auditable(struct audit_context *ctx) +{ + ctx->auditable = 1; +} + static inline struct audit_context *audit_get_context(struct task_struct *tsk, int return_valid, int return_code) @@ -365,12 +469,22 @@ static inline struct audit_context *audit_get_context(struct task_struct *tsk, if (context->in_syscall && !context->auditable) { enum audit_state state; + state = audit_filter_syscall(tsk, context, &audit_filter_list[AUDIT_FILTER_EXIT]); + if (state == AUDIT_RECORD_CONTEXT) { + context->auditable = 1; + goto get_context; + } + + state = audit_filter_inodes(tsk, context); if (state == AUDIT_RECORD_CONTEXT) context->auditable = 1; + } +get_context: context->pid = tsk->pid; + context->ppid = sys_getppid(); /* sic. tsk == current in all cases */ context->uid = tsk->uid; context->gid = tsk->gid; context->euid = tsk->euid; @@ -413,7 +527,7 @@ static inline void audit_free_names(struct audit_context *context) #endif for (i = 0; i < context->name_count; i++) { - if (context->names[i].name) + if (context->names[i].name && context->names[i].name_put) __putname(context->names[i].name); } context->name_count = 0; @@ -606,7 +720,7 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts tty = "(none)"; audit_log_format(ab, " a0=%lx a1=%lx a2=%lx a3=%lx items=%d" - " pid=%d auid=%u uid=%u gid=%u" + " ppid=%d pid=%d auid=%u uid=%u gid=%u" " euid=%u suid=%u fsuid=%u" " egid=%u sgid=%u fsgid=%u tty=%s", context->argv[0], @@ -614,6 +728,7 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts context->argv[2], context->argv[3], context->name_count, + context->ppid, context->pid, context->loginuid, context->uid, @@ -630,11 +745,48 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts continue; /* audit_panic has been called */ switch (aux->type) { + case AUDIT_MQ_OPEN: { + struct audit_aux_data_mq_open *axi = (void *)aux; + audit_log_format(ab, + "oflag=0x%x mode=%#o mq_flags=0x%lx mq_maxmsg=%ld " + "mq_msgsize=%ld mq_curmsgs=%ld", + axi->oflag, axi->mode, axi->attr.mq_flags, + axi->attr.mq_maxmsg, axi->attr.mq_msgsize, + axi->attr.mq_curmsgs); + break; } + + case AUDIT_MQ_SENDRECV: { + struct audit_aux_data_mq_sendrecv *axi = (void *)aux; + audit_log_format(ab, + "mqdes=%d msg_len=%zd msg_prio=%u " + "abs_timeout_sec=%ld abs_timeout_nsec=%ld", + axi->mqdes, axi->msg_len, axi->msg_prio, + axi->abs_timeout.tv_sec, axi->abs_timeout.tv_nsec); + break; } + + case AUDIT_MQ_NOTIFY: { + struct audit_aux_data_mq_notify *axi = (void *)aux; + audit_log_format(ab, + "mqdes=%d sigev_signo=%d", + axi->mqdes, + axi->notification.sigev_signo); + break; } + + case AUDIT_MQ_GETSETATTR: { + struct audit_aux_data_mq_getsetattr *axi = (void *)aux; + audit_log_format(ab, + "mqdes=%d mq_flags=0x%lx mq_maxmsg=%ld mq_msgsize=%ld " + "mq_curmsgs=%ld ", + axi->mqdes, + axi->mqstat.mq_flags, axi->mqstat.mq_maxmsg, + axi->mqstat.mq_msgsize, axi->mqstat.mq_curmsgs); + break; } + case AUDIT_IPC: { struct audit_aux_data_ipcctl *axi = (void *)aux; audit_log_format(ab, - " qbytes=%lx iuid=%u igid=%u mode=%x", - axi->qbytes, axi->uid, axi->gid, axi->mode); + "ouid=%u ogid=%u mode=%x", + axi->uid, axi->gid, axi->mode); if (axi->osid != 0) { char *ctx = NULL; u32 len; @@ -652,19 +804,18 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts case AUDIT_IPC_SET_PERM: { struct audit_aux_data_ipcctl *axi = (void *)aux; audit_log_format(ab, - " new qbytes=%lx new iuid=%u new igid=%u new mode=%x", + "qbytes=%lx ouid=%u ogid=%u mode=%x", axi->qbytes, axi->uid, axi->gid, axi->mode); - if (axi->osid != 0) { - char *ctx = NULL; - u32 len; - if (selinux_ctxid_to_string( - axi->osid, &ctx, &len)) { - audit_log_format(ab, " osid=%u", - axi->osid); - call_panic = 1; - } else - audit_log_format(ab, " obj=%s", ctx); - kfree(ctx); + break; } + + case AUDIT_EXECVE: { + struct audit_aux_data_execve *axi = (void *)aux; + int i; + const char *p; + for (i = 0, p = axi->mem; i < axi->argc; i++) { + audit_log_format(ab, "a%d=", i); + p = audit_log_untrustedstring(ab, p); + audit_log_format(ab, "\n"); } break; } @@ -700,8 +851,7 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts } } for (i = 0; i < context->name_count; i++) { - unsigned long ino = context->names[i].ino; - unsigned long pino = context->names[i].pino; + struct audit_names *n = &context->names[i]; ab = audit_log_start(context, GFP_KERNEL, AUDIT_PATH); if (!ab) @@ -709,33 +859,47 @@ static void audit_log_exit(struct audit_context *context, struct task_struct *ts audit_log_format(ab, "item=%d", i); - audit_log_format(ab, " name="); - if (context->names[i].name) - audit_log_untrustedstring(ab, context->names[i].name); - else - audit_log_format(ab, "(null)"); - - if (pino != (unsigned long)-1) - audit_log_format(ab, " parent=%lu", pino); - if (ino != (unsigned long)-1) - audit_log_format(ab, " inode=%lu", ino); - if ((pino != (unsigned long)-1) || (ino != (unsigned long)-1)) - audit_log_format(ab, " dev=%02x:%02x mode=%#o" - " ouid=%u ogid=%u rdev=%02x:%02x", - MAJOR(context->names[i].dev), - MINOR(context->names[i].dev), - context->names[i].mode, - context->names[i].uid, - context->names[i].gid, - MAJOR(context->names[i].rdev), - MINOR(context->names[i].rdev)); - if (context->names[i].osid != 0) { + if (n->name) { + switch(n->name_len) { + case AUDIT_NAME_FULL: + /* log the full path */ + audit_log_format(ab, " name="); + audit_log_untrustedstring(ab, n->name); + break; + case 0: + /* name was specified as a relative path and the + * directory component is the cwd */ + audit_log_d_path(ab, " name=", context->pwd, + context->pwdmnt); + break; + default: + /* log the name's directory component */ + audit_log_format(ab, " name="); + audit_log_n_untrustedstring(ab, n->name_len, + n->name); + } + } else + audit_log_format(ab, " name=(null)"); + + if (n->ino != (unsigned long)-1) { + audit_log_format(ab, " inode=%lu" + " dev=%02x:%02x mode=%#o" + " ouid=%u ogid=%u rdev=%02x:%02x", + n->ino, + MAJOR(n->dev), + MINOR(n->dev), + n->mode, + n->uid, + n->gid, + MAJOR(n->rdev), + MINOR(n->rdev)); + } + if (n->osid != 0) { char *ctx = NULL; u32 len; if (selinux_ctxid_to_string( - context->names[i].osid, &ctx, &len)) { - audit_log_format(ab, " osid=%u", - context->names[i].osid); + n->osid, &ctx, &len)) { + audit_log_format(ab, " osid=%u", n->osid); call_panic = 2; } else audit_log_format(ab, " obj=%s", ctx); @@ -908,11 +1072,11 @@ void audit_syscall_exit(int valid, long return_code) * Add a name to the list of audit names for this context. * Called from fs/namei.c:getname(). */ -void audit_getname(const char *name) +void __audit_getname(const char *name) { struct audit_context *context = current->audit_context; - if (!context || IS_ERR(name) || !name) + if (IS_ERR(name) || !name) return; if (!context->in_syscall) { @@ -925,6 +1089,8 @@ void audit_getname(const char *name) } BUG_ON(context->name_count >= AUDIT_NAMES); context->names[context->name_count].name = name; + context->names[context->name_count].name_len = AUDIT_NAME_FULL; + context->names[context->name_count].name_put = 1; context->names[context->name_count].ino = (unsigned long)-1; ++context->name_count; if (!context->pwd) { @@ -991,11 +1157,10 @@ static void audit_inode_context(int idx, const struct inode *inode) * audit_inode - store the inode and device from a lookup * @name: name being audited * @inode: inode being audited - * @flags: lookup flags (as used in path_lookup()) * * Called from fs/namei.c:path_lookup(). */ -void __audit_inode(const char *name, const struct inode *inode, unsigned flags) +void __audit_inode(const char *name, const struct inode *inode) { int idx; struct audit_context *context = current->audit_context; @@ -1021,20 +1186,13 @@ void __audit_inode(const char *name, const struct inode *inode, unsigned flags) ++context->ino_count; #endif } + context->names[idx].ino = inode->i_ino; context->names[idx].dev = inode->i_sb->s_dev; context->names[idx].mode = inode->i_mode; context->names[idx].uid = inode->i_uid; context->names[idx].gid = inode->i_gid; context->names[idx].rdev = inode->i_rdev; audit_inode_context(idx, inode); - if ((flags & LOOKUP_PARENT) && (strcmp(name, "/") != 0) && - (strcmp(name, ".") != 0)) { - context->names[idx].ino = (unsigned long)-1; - context->names[idx].pino = inode->i_ino; - } else { - context->names[idx].ino = inode->i_ino; - context->names[idx].pino = (unsigned long)-1; - } } /** @@ -1056,51 +1214,40 @@ void __audit_inode_child(const char *dname, const struct inode *inode, { int idx; struct audit_context *context = current->audit_context; + const char *found_name = NULL; + int dirlen = 0; if (!context->in_syscall) return; /* determine matching parent */ - if (dname) - for (idx = 0; idx < context->name_count; idx++) - if (context->names[idx].pino == pino) { - const char *n; - const char *name = context->names[idx].name; - int dlen = strlen(dname); - int nlen = name ? strlen(name) : 0; - - if (nlen < dlen) - continue; - - /* disregard trailing slashes */ - n = name + nlen - 1; - while ((*n == '/') && (n > name)) - n--; - - /* find last path component */ - n = n - dlen + 1; - if (n < name) - continue; - else if (n > name) { - if (*--n != '/') - continue; - else - n++; - } - - if (strncmp(n, dname, dlen) == 0) - goto update_context; + if (!dname) + goto update_context; + for (idx = 0; idx < context->name_count; idx++) + if (context->names[idx].ino == pino) { + const char *name = context->names[idx].name; + + if (!name) + continue; + + if (audit_compare_dname_path(dname, name, &dirlen) == 0) { + context->names[idx].name_len = dirlen; + found_name = name; + break; } + } - /* catch-all in case match not found */ +update_context: idx = context->name_count++; - context->names[idx].name = NULL; - context->names[idx].pino = pino; #if AUDIT_DEBUG context->ino_count++; #endif + /* Re-use the name belonging to the slot for a matching parent directory. + * All names for this context are relinquished in audit_free_names() */ + context->names[idx].name = found_name; + context->names[idx].name_len = AUDIT_NAME_FULL; + context->names[idx].name_put = 0; /* don't call __putname() */ -update_context: if (inode) { context->names[idx].ino = inode->i_ino; context->names[idx].dev = inode->i_sb->s_dev; @@ -1109,7 +1256,8 @@ update_context: context->names[idx].gid = inode->i_gid; context->names[idx].rdev = inode->i_rdev; audit_inode_context(idx, inode); - } + } else + context->names[idx].ino = (unsigned long)-1; } /** @@ -1142,18 +1290,23 @@ void auditsc_get_stamp(struct audit_context *ctx, */ int audit_set_loginuid(struct task_struct *task, uid_t loginuid) { - if (task->audit_context) { - struct audit_buffer *ab; - - ab = audit_log_start(NULL, GFP_KERNEL, AUDIT_LOGIN); - if (ab) { - audit_log_format(ab, "login pid=%d uid=%u " - "old auid=%u new auid=%u", - task->pid, task->uid, - task->audit_context->loginuid, loginuid); - audit_log_end(ab); + struct audit_context *context = task->audit_context; + + if (context) { + /* Only log if audit is enabled */ + if (context->in_syscall) { + struct audit_buffer *ab; + + ab = audit_log_start(NULL, GFP_KERNEL, AUDIT_LOGIN); + if (ab) { + audit_log_format(ab, "login pid=%d uid=%u " + "old auid=%u new auid=%u", + task->pid, task->uid, + context->loginuid, loginuid); + audit_log_end(ab); + } } - task->audit_context->loginuid = loginuid; + context->loginuid = loginuid; } return 0; } @@ -1170,16 +1323,193 @@ uid_t audit_get_loginuid(struct audit_context *ctx) } /** - * audit_ipc_obj - record audit data for ipc object - * @ipcp: ipc permissions + * __audit_mq_open - record audit data for a POSIX MQ open + * @oflag: open flag + * @mode: mode bits + * @u_attr: queue attributes * * Returns 0 for success or NULL context or < 0 on error. */ -int audit_ipc_obj(struct kern_ipc_perm *ipcp) +int __audit_mq_open(int oflag, mode_t mode, struct mq_attr __user *u_attr) { - struct audit_aux_data_ipcctl *ax; + struct audit_aux_data_mq_open *ax; + struct audit_context *context = current->audit_context; + + if (!audit_enabled) + return 0; + + if (likely(!context)) + return 0; + + ax = kmalloc(sizeof(*ax), GFP_ATOMIC); + if (!ax) + return -ENOMEM; + + if (u_attr != NULL) { + if (copy_from_user(&ax->attr, u_attr, sizeof(ax->attr))) { + kfree(ax); + return -EFAULT; + } + } else + memset(&ax->attr, 0, sizeof(ax->attr)); + + ax->oflag = oflag; + ax->mode = mode; + + ax->d.type = AUDIT_MQ_OPEN; + ax->d.next = context->aux; + context->aux = (void *)ax; + return 0; +} + +/** + * __audit_mq_timedsend - record audit data for a POSIX MQ timed send + * @mqdes: MQ descriptor + * @msg_len: Message length + * @msg_prio: Message priority + * @abs_timeout: Message timeout in absolute time + * + * Returns 0 for success or NULL context or < 0 on error. + */ +int __audit_mq_timedsend(mqd_t mqdes, size_t msg_len, unsigned int msg_prio, + const struct timespec __user *u_abs_timeout) +{ + struct audit_aux_data_mq_sendrecv *ax; + struct audit_context *context = current->audit_context; + + if (!audit_enabled) + return 0; + + if (likely(!context)) + return 0; + + ax = kmalloc(sizeof(*ax), GFP_ATOMIC); + if (!ax) + return -ENOMEM; + + if (u_abs_timeout != NULL) { + if (copy_from_user(&ax->abs_timeout, u_abs_timeout, sizeof(ax->abs_timeout))) { + kfree(ax); + return -EFAULT; + } + } else + memset(&ax->abs_timeout, 0, sizeof(ax->abs_timeout)); + + ax->mqdes = mqdes; + ax->msg_len = msg_len; + ax->msg_prio = msg_prio; + + ax->d.type = AUDIT_MQ_SENDRECV; + ax->d.next = context->aux; + context->aux = (void *)ax; + return 0; +} + +/** + * __audit_mq_timedreceive - record audit data for a POSIX MQ timed receive + * @mqdes: MQ descriptor + * @msg_len: Message length + * @msg_prio: Message priority + * @abs_timeout: Message timeout in absolute time + * + * Returns 0 for success or NULL context or < 0 on error. + */ +int __audit_mq_timedreceive(mqd_t mqdes, size_t msg_len, + unsigned int __user *u_msg_prio, + const struct timespec __user *u_abs_timeout) +{ + struct audit_aux_data_mq_sendrecv *ax; + struct audit_context *context = current->audit_context; + + if (!audit_enabled) + return 0; + + if (likely(!context)) + return 0; + + ax = kmalloc(sizeof(*ax), GFP_ATOMIC); + if (!ax) + return -ENOMEM; + + if (u_msg_prio != NULL) { + if (get_user(ax->msg_prio, u_msg_prio)) { + kfree(ax); + return -EFAULT; + } + } else + ax->msg_prio = 0; + + if (u_abs_timeout != NULL) { + if (copy_from_user(&ax->abs_timeout, u_abs_timeout, sizeof(ax->abs_timeout))) { + kfree(ax); + return -EFAULT; + } + } else + memset(&ax->abs_timeout, 0, sizeof(ax->abs_timeout)); + + ax->mqdes = mqdes; + ax->msg_len = msg_len; + + ax->d.type = AUDIT_MQ_SENDRECV; + ax->d.next = context->aux; + context->aux = (void *)ax; + return 0; +} + +/** + * __audit_mq_notify - record audit data for a POSIX MQ notify + * @mqdes: MQ descriptor + * @u_notification: Notification event + * + * Returns 0 for success or NULL context or < 0 on error. + */ + +int __audit_mq_notify(mqd_t mqdes, const struct sigevent __user *u_notification) +{ + struct audit_aux_data_mq_notify *ax; + struct audit_context *context = current->audit_context; + + if (!audit_enabled) + return 0; + + if (likely(!context)) + return 0; + + ax = kmalloc(sizeof(*ax), GFP_ATOMIC); + if (!ax) + return -ENOMEM; + + if (u_notification != NULL) { + if (copy_from_user(&ax->notification, u_notification, sizeof(ax->notification))) { + kfree(ax); + return -EFAULT; + } + } else + memset(&ax->notification, 0, sizeof(ax->notification)); + + ax->mqdes = mqdes; + + ax->d.type = AUDIT_MQ_NOTIFY; + ax->d.next = context->aux; + context->aux = (void *)ax; + return 0; +} + +/** + * __audit_mq_getsetattr - record audit data for a POSIX MQ get/set attribute + * @mqdes: MQ descriptor + * @mqstat: MQ flags + * + * Returns 0 for success or NULL context or < 0 on error. + */ +int __audit_mq_getsetattr(mqd_t mqdes, struct mq_attr *mqstat) +{ + struct audit_aux_data_mq_getsetattr *ax; struct audit_context *context = current->audit_context; + if (!audit_enabled) + return 0; + if (likely(!context)) return 0; @@ -1187,6 +1517,30 @@ int audit_ipc_obj(struct kern_ipc_perm *ipcp) if (!ax) return -ENOMEM; + ax->mqdes = mqdes; + ax->mqstat = *mqstat; + + ax->d.type = AUDIT_MQ_GETSETATTR; + ax->d.next = context->aux; + context->aux = (void *)ax; + return 0; +} + +/** + * audit_ipc_obj - record audit data for ipc object + * @ipcp: ipc permissions + * + * Returns 0 for success or NULL context or < 0 on error. + */ +int __audit_ipc_obj(struct kern_ipc_perm *ipcp) +{ + struct audit_aux_data_ipcctl *ax; + struct audit_context *context = current->audit_context; + + ax = kmalloc(sizeof(*ax), GFP_ATOMIC); + if (!ax) + return -ENOMEM; + ax->uid = ipcp->uid; ax->gid = ipcp->gid; ax->mode = ipcp->mode; @@ -1204,17 +1558,15 @@ int audit_ipc_obj(struct kern_ipc_perm *ipcp) * @uid: msgq user id * @gid: msgq group id * @mode: msgq mode (permissions) + * @ipcp: in-kernel IPC permissions * * Returns 0 for success or NULL context or < 0 on error. */ -int audit_ipc_set_perm(unsigned long qbytes, uid_t uid, gid_t gid, mode_t mode, struct kern_ipc_perm *ipcp) +int __audit_ipc_set_perm(unsigned long qbytes, uid_t uid, gid_t gid, mode_t mode) { struct audit_aux_data_ipcctl *ax; struct audit_context *context = current->audit_context; - if (likely(!context)) - return 0; - ax = kmalloc(sizeof(*ax), GFP_ATOMIC); if (!ax) return -ENOMEM; @@ -1223,7 +1575,6 @@ int audit_ipc_set_perm(unsigned long qbytes, uid_t uid, gid_t gid, mode_t mode, ax->uid = uid; ax->gid = gid; ax->mode = mode; - selinux_get_ipc_sid(ipcp, &ax->osid); ax->d.type = AUDIT_IPC_SET_PERM; ax->d.next = context->aux; @@ -1231,6 +1582,39 @@ int audit_ipc_set_perm(unsigned long qbytes, uid_t uid, gid_t gid, mode_t mode, return 0; } +int audit_bprm(struct linux_binprm *bprm) +{ + struct audit_aux_data_execve *ax; + struct audit_context *context = current->audit_context; + unsigned long p, next; + void *to; + + if (likely(!audit_enabled || !context)) + return 0; + + ax = kmalloc(sizeof(*ax) + PAGE_SIZE * MAX_ARG_PAGES - bprm->p, + GFP_KERNEL); + if (!ax) + return -ENOMEM; + + ax->argc = bprm->argc; + ax->envc = bprm->envc; + for (p = bprm->p, to = ax->mem; p < MAX_ARG_PAGES*PAGE_SIZE; p = next) { + struct page *page = bprm->page[p / PAGE_SIZE]; + void *kaddr = kmap(page); + next = (p + PAGE_SIZE) & ~(PAGE_SIZE - 1); + memcpy(to, kaddr + (p & (PAGE_SIZE - 1)), next - p); + to += next - p; + kunmap(page); + } + + ax->d.type = AUDIT_EXECVE; + ax->d.next = context->aux; + context->aux = (void *)ax; + return 0; +} + + /** * audit_socketcall - record audit data for sys_socketcall * @nargs: number of args @@ -1325,19 +1709,20 @@ int audit_avc_path(struct dentry *dentry, struct vfsmount *mnt) * If the audit subsystem is being terminated, record the task (pid) * and uid that is doing that. */ -void audit_signal_info(int sig, struct task_struct *t) +void __audit_signal_info(int sig, struct task_struct *t) { extern pid_t audit_sig_pid; extern uid_t audit_sig_uid; - - if (unlikely(audit_pid && t->tgid == audit_pid)) { - if (sig == SIGTERM || sig == SIGHUP) { - struct audit_context *ctx = current->audit_context; - audit_sig_pid = current->pid; - if (ctx) - audit_sig_uid = ctx->loginuid; - else - audit_sig_uid = current->uid; - } + extern u32 audit_sig_sid; + + if (sig == SIGTERM || sig == SIGHUP || sig == SIGUSR1) { + struct task_struct *tsk = current; + struct audit_context *ctx = tsk->audit_context; + audit_sig_pid = tsk->pid; + if (ctx) + audit_sig_uid = ctx->loginuid; + else + audit_sig_uid = tsk->uid; + selinux_get_task_sid(tsk, &audit_sig_sid); } } diff --git a/kernel/compat.c b/kernel/compat.c index c1601a84f8d8..126dee9530aa 100644 --- a/kernel/compat.c +++ b/kernel/compat.c @@ -21,6 +21,7 @@ #include <linux/unistd.h> #include <linux/security.h> #include <linux/timex.h> +#include <linux/migrate.h> #include <asm/uaccess.h> @@ -729,17 +730,10 @@ void sigset_from_compat (sigset_t *set, compat_sigset_t *compat) { switch (_NSIG_WORDS) { -#if defined (__COMPAT_ENDIAN_SWAP__) - case 4: set->sig[3] = compat->sig[7] | (((long)compat->sig[6]) << 32 ); - case 3: set->sig[2] = compat->sig[5] | (((long)compat->sig[4]) << 32 ); - case 2: set->sig[1] = compat->sig[3] | (((long)compat->sig[2]) << 32 ); - case 1: set->sig[0] = compat->sig[1] | (((long)compat->sig[0]) << 32 ); -#else case 4: set->sig[3] = compat->sig[6] | (((long)compat->sig[7]) << 32 ); case 3: set->sig[2] = compat->sig[4] | (((long)compat->sig[5]) << 32 ); case 2: set->sig[1] = compat->sig[2] | (((long)compat->sig[3]) << 32 ); case 1: set->sig[0] = compat->sig[0] | (((long)compat->sig[1]) << 32 ); -#endif } } @@ -934,3 +928,25 @@ asmlinkage long compat_sys_adjtimex(struct compat_timex __user *utp) return ret; } + +#ifdef CONFIG_NUMA +asmlinkage long compat_sys_move_pages(pid_t pid, unsigned long nr_pages, + compat_uptr_t __user *pages32, + const int __user *nodes, + int __user *status, + int flags) +{ + const void __user * __user *pages; + int i; + + pages = compat_alloc_user_space(nr_pages * sizeof(void *)); + for (i = 0; i < nr_pages; i++) { + compat_uptr_t p; + + if (get_user(p, pages32 + i) || + put_user(compat_ptr(p), pages + i)) + return -EFAULT; + } + return sys_move_pages(pid, nr_pages, pages, nodes, status, flags); +} +#endif diff --git a/kernel/cpuset.c b/kernel/cpuset.c index ab81fdd4572b..b602f73fb38d 100644 --- a/kernel/cpuset.c +++ b/kernel/cpuset.c @@ -41,6 +41,7 @@ #include <linux/rcupdate.h> #include <linux/sched.h> #include <linux/seq_file.h> +#include <linux/security.h> #include <linux/slab.h> #include <linux/smp_lock.h> #include <linux/spinlock.h> @@ -392,11 +393,11 @@ static int cpuset_fill_super(struct super_block *sb, void *unused_data, return 0; } -static struct super_block *cpuset_get_sb(struct file_system_type *fs_type, - int flags, const char *unused_dev_name, - void *data) +static int cpuset_get_sb(struct file_system_type *fs_type, + int flags, const char *unused_dev_name, + void *data, struct vfsmount *mnt) { - return get_sb_single(fs_type, flags, data, cpuset_fill_super); + return get_sb_single(fs_type, flags, data, cpuset_fill_super, mnt); } static struct file_system_type cpuset_fs_type = { @@ -1177,6 +1178,7 @@ static int attach_task(struct cpuset *cs, char *pidbuf, char **ppathbuf) cpumask_t cpus; nodemask_t from, to; struct mm_struct *mm; + int retval; if (sscanf(pidbuf, "%d", &pid) != 1) return -EIO; @@ -1205,6 +1207,12 @@ static int attach_task(struct cpuset *cs, char *pidbuf, char **ppathbuf) get_task_struct(tsk); } + retval = security_task_setscheduler(tsk, 0, NULL); + if (retval) { + put_task_struct(tsk); + return retval; + } + mutex_lock(&callback_mutex); task_lock(tsk); diff --git a/kernel/exit.c b/kernel/exit.c index e95b93282210..e76bd02e930e 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -36,6 +36,7 @@ #include <linux/compat.h> #include <linux/pipe_fs_i.h> #include <linux/audit.h> /* for audit_free() */ +#include <linux/resource.h> #include <asm/uaccess.h> #include <asm/unistd.h> @@ -45,8 +46,6 @@ extern void sem_exit (void); extern struct task_struct *child_reaper; -int getrusage(struct task_struct *, int, struct rusage __user *); - static void exit_mm(struct task_struct * tsk); static void __unhash_process(struct task_struct *p) @@ -579,7 +578,7 @@ static void exit_mm(struct task_struct * tsk) down_read(&mm->mmap_sem); } atomic_inc(&mm->mm_count); - if (mm != tsk->active_mm) BUG(); + BUG_ON(mm != tsk->active_mm); /* more a memory barrier than a real lock */ task_lock(tsk); tsk->mm = NULL; @@ -881,14 +880,6 @@ fastcall NORET_TYPE void do_exit(long code) tsk->flags |= PF_EXITING; - /* - * Make sure we don't try to process any timer firings - * while we are already exiting. - */ - tsk->it_virt_expires = cputime_zero; - tsk->it_prof_expires = cputime_zero; - tsk->it_sched_expires = 0; - if (unlikely(in_atomic())) printk(KERN_INFO "note: %s[%d] exited with preempt_count %d\n", current->comm, current->pid, @@ -903,11 +894,11 @@ fastcall NORET_TYPE void do_exit(long code) if (group_dead) { hrtimer_cancel(&tsk->signal->real_timer); exit_itimers(tsk->signal); - acct_process(code); } + acct_collect(code, group_dead); if (unlikely(tsk->robust_list)) exit_robust_list(tsk); -#ifdef CONFIG_COMPAT +#if defined(CONFIG_FUTEX) && defined(CONFIG_COMPAT) if (unlikely(tsk->compat_robust_list)) compat_exit_robust_list(tsk); #endif @@ -915,6 +906,8 @@ fastcall NORET_TYPE void do_exit(long code) audit_free(tsk); exit_mm(tsk); + if (group_dead) + acct_process(); exit_sem(tsk); __exit_files(tsk); __exit_fs(tsk); @@ -1538,8 +1531,7 @@ check_continued: if (options & __WNOTHREAD) break; tsk = next_thread(tsk); - if (tsk->signal != current->signal) - BUG(); + BUG_ON(tsk->signal != current->signal); } while (tsk != current); read_unlock(&tasklist_lock); diff --git a/kernel/fork.c b/kernel/fork.c index ac8100e3088a..dfd10cb370c3 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -368,6 +368,8 @@ void fastcall __mmdrop(struct mm_struct *mm) */ void mmput(struct mm_struct *mm) { + might_sleep(); + if (atomic_dec_and_test(&mm->mm_users)) { exit_aio(mm); exit_mmap(mm); @@ -623,6 +625,7 @@ out: /* * Allocate a new files structure and copy contents from the * passed in files structure. + * errorp will be valid only when the returned files_struct is NULL. */ static struct files_struct *dup_fd(struct files_struct *oldf, int *errorp) { @@ -631,6 +634,7 @@ static struct files_struct *dup_fd(struct files_struct *oldf, int *errorp) int open_files, size, i, expand; struct fdtable *old_fdt, *new_fdt; + *errorp = -ENOMEM; newf = alloc_files(); if (!newf) goto out; @@ -744,7 +748,6 @@ static int copy_files(unsigned long clone_flags, struct task_struct * tsk) * break this. */ tsk->files = NULL; - error = -ENOMEM; newf = dup_fd(oldf, &error); if (!newf) goto out; @@ -871,6 +874,7 @@ static inline int copy_signal(unsigned long clone_flags, struct task_struct * ts tsk->it_prof_expires = secs_to_cputime(sig->rlim[RLIMIT_CPU].rlim_cur); } + acct_init_pacct(&sig->pacct); return 0; } diff --git a/kernel/futex.c b/kernel/futex.c index 5699c512057b..e1a380c77a5a 100644 --- a/kernel/futex.c +++ b/kernel/futex.c @@ -1056,11 +1056,11 @@ asmlinkage long sys_futex(u32 __user *uaddr, int op, int val, (unsigned long)uaddr2, val2, val3); } -static struct super_block * -futexfs_get_sb(struct file_system_type *fs_type, - int flags, const char *dev_name, void *data) +static int futexfs_get_sb(struct file_system_type *fs_type, + int flags, const char *dev_name, void *data, + struct vfsmount *mnt) { - return get_sb_pseudo(fs_type, "futex", NULL, 0xBAD1DEA); + return get_sb_pseudo(fs_type, "futex", NULL, 0xBAD1DEA, mnt); } static struct file_system_type futex_fs_type = { diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c index b7f0388bd71c..55601b3ce60e 100644 --- a/kernel/hrtimer.c +++ b/kernel/hrtimer.c @@ -98,7 +98,6 @@ static DEFINE_PER_CPU(struct hrtimer_base, hrtimer_bases[MAX_HRTIMER_BASES]) = /** * ktime_get_ts - get the monotonic clock in timespec format - * * @ts: pointer to timespec variable * * The function calculates the monotonic clock from the realtime @@ -238,7 +237,6 @@ lock_hrtimer_base(const struct hrtimer *timer, unsigned long *flags) # ifndef CONFIG_KTIME_SCALAR /** * ktime_add_ns - Add a scalar nanoseconds value to a ktime_t variable - * * @kt: addend * @nsec: the scalar nsec value to add * @@ -299,7 +297,6 @@ void unlock_hrtimer_base(const struct hrtimer *timer, unsigned long *flags) /** * hrtimer_forward - forward the timer expiry - * * @timer: hrtimer to forward * @now: forward past this time * @interval: the interval to forward @@ -393,7 +390,7 @@ static void __remove_hrtimer(struct hrtimer *timer, struct hrtimer_base *base) if (base->first == &timer->node) base->first = rb_next(&timer->node); rb_erase(&timer->node, &base->active); - timer->node.rb_parent = HRTIMER_INACTIVE; + rb_set_parent(&timer->node, &timer->node); } /* @@ -411,7 +408,6 @@ remove_hrtimer(struct hrtimer *timer, struct hrtimer_base *base) /** * hrtimer_start - (re)start an relative timer on the current CPU - * * @timer: the timer to be added * @tim: expiry time * @mode: expiry mode: absolute (HRTIMER_ABS) or relative (HRTIMER_REL) @@ -456,17 +452,17 @@ hrtimer_start(struct hrtimer *timer, ktime_t tim, const enum hrtimer_mode mode) return ret; } +EXPORT_SYMBOL_GPL(hrtimer_start); /** * hrtimer_try_to_cancel - try to deactivate a timer - * * @timer: hrtimer to stop * * Returns: * 0 when the timer was not active * 1 when the timer was active * -1 when the timer is currently excuting the callback function and - * can not be stopped + * cannot be stopped */ int hrtimer_try_to_cancel(struct hrtimer *timer) { @@ -484,10 +480,10 @@ int hrtimer_try_to_cancel(struct hrtimer *timer) return ret; } +EXPORT_SYMBOL_GPL(hrtimer_try_to_cancel); /** * hrtimer_cancel - cancel a timer and wait for the handler to finish. - * * @timer: the timer to be cancelled * * Returns: @@ -504,10 +500,10 @@ int hrtimer_cancel(struct hrtimer *timer) cpu_relax(); } } +EXPORT_SYMBOL_GPL(hrtimer_cancel); /** * hrtimer_get_remaining - get remaining time for the timer - * * @timer: the timer to read */ ktime_t hrtimer_get_remaining(const struct hrtimer *timer) @@ -522,6 +518,7 @@ ktime_t hrtimer_get_remaining(const struct hrtimer *timer) return rem; } +EXPORT_SYMBOL_GPL(hrtimer_get_remaining); #ifdef CONFIG_NO_IDLE_HZ /** @@ -560,7 +557,6 @@ ktime_t hrtimer_get_next_event(void) /** * hrtimer_init - initialize a timer to the given clock - * * @timer: the timer to be initialized * @clock_id: the clock to be used * @mode: timer mode abs/rel @@ -572,18 +568,18 @@ void hrtimer_init(struct hrtimer *timer, clockid_t clock_id, memset(timer, 0, sizeof(struct hrtimer)); - bases = per_cpu(hrtimer_bases, raw_smp_processor_id()); + bases = __raw_get_cpu_var(hrtimer_bases); if (clock_id == CLOCK_REALTIME && mode != HRTIMER_ABS) clock_id = CLOCK_MONOTONIC; timer->base = &bases[clock_id]; - timer->node.rb_parent = HRTIMER_INACTIVE; + rb_set_parent(&timer->node, &timer->node); } +EXPORT_SYMBOL_GPL(hrtimer_init); /** * hrtimer_get_res - get the timer resolution for a clock - * * @which_clock: which clock to query * @tp: pointer to timespec variable to store the resolution * @@ -594,11 +590,12 @@ int hrtimer_get_res(const clockid_t which_clock, struct timespec *tp) { struct hrtimer_base *bases; - bases = per_cpu(hrtimer_bases, raw_smp_processor_id()); + bases = __raw_get_cpu_var(hrtimer_bases); *tp = ktime_to_timespec(bases[which_clock].resolution); return 0; } +EXPORT_SYMBOL_GPL(hrtimer_get_res); /* * Expire the per base hrtimer-queue: diff --git a/kernel/intermodule.c b/kernel/intermodule.c deleted file mode 100644 index 55b1e5b85db9..000000000000 --- a/kernel/intermodule.c +++ /dev/null @@ -1,184 +0,0 @@ -/* Deprecated, do not use. Moved from module.c to here. --RR */ - -/* Written by Keith Owens <kaos@ocs.com.au> Oct 2000 */ -#include <linux/module.h> -#include <linux/kmod.h> -#include <linux/spinlock.h> -#include <linux/list.h> -#include <linux/slab.h> - -/* inter_module functions are always available, even when the kernel is - * compiled without modules. Consumers of inter_module_xxx routines - * will always work, even when both are built into the kernel, this - * approach removes lots of #ifdefs in mainline code. - */ - -static struct list_head ime_list = LIST_HEAD_INIT(ime_list); -static DEFINE_SPINLOCK(ime_lock); -static int kmalloc_failed; - -struct inter_module_entry { - struct list_head list; - const char *im_name; - struct module *owner; - const void *userdata; -}; - -/** - * inter_module_register - register a new set of inter module data. - * @im_name: an arbitrary string to identify the data, must be unique - * @owner: module that is registering the data, always use THIS_MODULE - * @userdata: pointer to arbitrary userdata to be registered - * - * Description: Check that the im_name has not already been registered, - * complain if it has. For new data, add it to the inter_module_entry - * list. - */ -void inter_module_register(const char *im_name, struct module *owner, const void *userdata) -{ - struct list_head *tmp; - struct inter_module_entry *ime, *ime_new; - - if (!(ime_new = kzalloc(sizeof(*ime), GFP_KERNEL))) { - /* Overloaded kernel, not fatal */ - printk(KERN_ERR - "Aiee, inter_module_register: cannot kmalloc entry for '%s'\n", - im_name); - kmalloc_failed = 1; - return; - } - ime_new->im_name = im_name; - ime_new->owner = owner; - ime_new->userdata = userdata; - - spin_lock(&ime_lock); - list_for_each(tmp, &ime_list) { - ime = list_entry(tmp, struct inter_module_entry, list); - if (strcmp(ime->im_name, im_name) == 0) { - spin_unlock(&ime_lock); - kfree(ime_new); - /* Program logic error, fatal */ - printk(KERN_ERR "inter_module_register: duplicate im_name '%s'", im_name); - BUG(); - } - } - list_add(&(ime_new->list), &ime_list); - spin_unlock(&ime_lock); -} - -/** - * inter_module_unregister - unregister a set of inter module data. - * @im_name: an arbitrary string to identify the data, must be unique - * - * Description: Check that the im_name has been registered, complain if - * it has not. For existing data, remove it from the - * inter_module_entry list. - */ -void inter_module_unregister(const char *im_name) -{ - struct list_head *tmp; - struct inter_module_entry *ime; - - spin_lock(&ime_lock); - list_for_each(tmp, &ime_list) { - ime = list_entry(tmp, struct inter_module_entry, list); - if (strcmp(ime->im_name, im_name) == 0) { - list_del(&(ime->list)); - spin_unlock(&ime_lock); - kfree(ime); - return; - } - } - spin_unlock(&ime_lock); - if (kmalloc_failed) { - printk(KERN_ERR - "inter_module_unregister: no entry for '%s', " - "probably caused by previous kmalloc failure\n", - im_name); - return; - } - else { - /* Program logic error, fatal */ - printk(KERN_ERR "inter_module_unregister: no entry for '%s'", im_name); - BUG(); - } -} - -/** - * inter_module_get - return arbitrary userdata from another module. - * @im_name: an arbitrary string to identify the data, must be unique - * - * Description: If the im_name has not been registered, return NULL. - * Try to increment the use count on the owning module, if that fails - * then return NULL. Otherwise return the userdata. - */ -static const void *inter_module_get(const char *im_name) -{ - struct list_head *tmp; - struct inter_module_entry *ime; - const void *result = NULL; - - spin_lock(&ime_lock); - list_for_each(tmp, &ime_list) { - ime = list_entry(tmp, struct inter_module_entry, list); - if (strcmp(ime->im_name, im_name) == 0) { - if (try_module_get(ime->owner)) - result = ime->userdata; - break; - } - } - spin_unlock(&ime_lock); - return(result); -} - -/** - * inter_module_get_request - im get with automatic request_module. - * @im_name: an arbitrary string to identify the data, must be unique - * @modname: module that is expected to register im_name - * - * Description: If inter_module_get fails, do request_module then retry. - */ -const void *inter_module_get_request(const char *im_name, const char *modname) -{ - const void *result = inter_module_get(im_name); - if (!result) { - request_module("%s", modname); - result = inter_module_get(im_name); - } - return(result); -} - -/** - * inter_module_put - release use of data from another module. - * @im_name: an arbitrary string to identify the data, must be unique - * - * Description: If the im_name has not been registered, complain, - * otherwise decrement the use count on the owning module. - */ -void inter_module_put(const char *im_name) -{ - struct list_head *tmp; - struct inter_module_entry *ime; - - spin_lock(&ime_lock); - list_for_each(tmp, &ime_list) { - ime = list_entry(tmp, struct inter_module_entry, list); - if (strcmp(ime->im_name, im_name) == 0) { - if (ime->owner) - module_put(ime->owner); - spin_unlock(&ime_lock); - return; - } - } - spin_unlock(&ime_lock); - printk(KERN_ERR "inter_module_put: no entry for '%s'", im_name); - BUG(); -} - -EXPORT_SYMBOL(inter_module_register); -EXPORT_SYMBOL(inter_module_unregister); -EXPORT_SYMBOL(inter_module_get_request); -EXPORT_SYMBOL(inter_module_put); - -MODULE_LICENSE("GPL"); - diff --git a/kernel/irq/handle.c b/kernel/irq/handle.c index 51df337b37db..0f6530117105 100644 --- a/kernel/irq/handle.c +++ b/kernel/irq/handle.c @@ -76,10 +76,11 @@ irqreturn_t no_action(int cpl, void *dev_id, struct pt_regs *regs) /* * Have got an event to handle: */ -fastcall int handle_IRQ_event(unsigned int irq, struct pt_regs *regs, +fastcall irqreturn_t handle_IRQ_event(unsigned int irq, struct pt_regs *regs, struct irqaction *action) { - int ret, retval = 0, status = 0; + irqreturn_t ret, retval = IRQ_NONE; + unsigned int status = 0; if (!(action->flags & SA_INTERRUPT)) local_irq_enable(); diff --git a/kernel/irq/migration.c b/kernel/irq/migration.c index 134f9f2e0e39..a12d00eb5e7c 100644 --- a/kernel/irq/migration.c +++ b/kernel/irq/migration.c @@ -30,7 +30,7 @@ void move_native_irq(int irq) desc->move_irq = 0; - if (likely(cpus_empty(pending_irq_cpumask[irq]))) + if (unlikely(cpus_empty(pending_irq_cpumask[irq]))) return; if (!desc->handler->set_affinity) @@ -49,7 +49,7 @@ void move_native_irq(int irq) * cause some ioapics to mal-function. * Being paranoid i guess! */ - if (unlikely(!cpus_empty(tmp))) { + if (likely(!cpus_empty(tmp))) { if (likely(!(desc->status & IRQ_DISABLED))) desc->handler->disable(irq); diff --git a/kernel/irq/proc.c b/kernel/irq/proc.c index d03b5eef8ce0..afacd6f585fa 100644 --- a/kernel/irq/proc.c +++ b/kernel/irq/proc.c @@ -24,6 +24,8 @@ static struct proc_dir_entry *smp_affinity_entry[NR_IRQS]; #ifdef CONFIG_GENERIC_PENDING_IRQ void proc_set_irq_affinity(unsigned int irq, cpumask_t mask_val) { + set_balance_irq_affinity(irq, mask_val); + /* * Save these away for later use. Re-progam when the * interrupt is pending @@ -33,6 +35,7 @@ void proc_set_irq_affinity(unsigned int irq, cpumask_t mask_val) #else void proc_set_irq_affinity(unsigned int irq, cpumask_t mask_val) { + set_balance_irq_affinity(irq, mask_val); irq_affinity[irq] = mask_val; irq_desc[irq].handler->set_affinity(irq, mask_val); } diff --git a/kernel/irq/spurious.c b/kernel/irq/spurious.c index 7df9abd5ec86..b2fb3c18d06b 100644 --- a/kernel/irq/spurious.c +++ b/kernel/irq/spurious.c @@ -11,7 +11,7 @@ #include <linux/kallsyms.h> #include <linux/interrupt.h> -static int irqfixup; +static int irqfixup __read_mostly; /* * Recovery handler for misrouted interrupts. @@ -136,9 +136,9 @@ static void report_bad_irq(unsigned int irq, irq_desc_t *desc, irqreturn_t actio void note_interrupt(unsigned int irq, irq_desc_t *desc, irqreturn_t action_ret, struct pt_regs *regs) { - if (action_ret != IRQ_HANDLED) { + if (unlikely(action_ret != IRQ_HANDLED)) { desc->irqs_unhandled++; - if (action_ret != IRQ_NONE) + if (unlikely(action_ret != IRQ_NONE)) report_bad_irq(irq, desc, action_ret); } @@ -152,11 +152,11 @@ void note_interrupt(unsigned int irq, irq_desc_t *desc, irqreturn_t action_ret, } desc->irq_count++; - if (desc->irq_count < 100000) + if (likely(desc->irq_count < 100000)) return; desc->irq_count = 0; - if (desc->irqs_unhandled > 99900) { + if (unlikely(desc->irqs_unhandled > 99900)) { /* * The interrupt is stuck */ @@ -171,7 +171,7 @@ void note_interrupt(unsigned int irq, irq_desc_t *desc, irqreturn_t action_ret, desc->irqs_unhandled = 0; } -int noirqdebug; +int noirqdebug __read_mostly; int __init noirqdebug_setup(char *str) { diff --git a/kernel/kexec.c b/kernel/kexec.c index bf39d28e4c0e..58f0f382597c 100644 --- a/kernel/kexec.c +++ b/kernel/kexec.c @@ -902,14 +902,14 @@ static int kimage_load_segment(struct kimage *image, * kexec does not sync, or unmount filesystems so if you need * that to happen you need to do that yourself. */ -struct kimage *kexec_image = NULL; -static struct kimage *kexec_crash_image = NULL; +struct kimage *kexec_image; +struct kimage *kexec_crash_image; /* * A home grown binary mutex. * Nothing can wait so this mutex is safe to use * in interrupt context :) */ -static int kexec_lock = 0; +static int kexec_lock; asmlinkage long sys_kexec_load(unsigned long entry, unsigned long nr_segments, struct kexec_segment __user *segments, diff --git a/kernel/ksysfs.c b/kernel/ksysfs.c index f119e098e67b..9e28478a17a5 100644 --- a/kernel/ksysfs.c +++ b/kernel/ksysfs.c @@ -14,6 +14,7 @@ #include <linux/sysfs.h> #include <linux/module.h> #include <linux/init.h> +#include <linux/kexec.h> #define KERNEL_ATTR_RO(_name) \ static struct subsys_attribute _name##_attr = __ATTR_RO(_name) @@ -48,6 +49,20 @@ static ssize_t uevent_helper_store(struct subsystem *subsys, const char *page, s KERNEL_ATTR_RW(uevent_helper); #endif +#ifdef CONFIG_KEXEC +static ssize_t kexec_loaded_show(struct subsystem *subsys, char *page) +{ + return sprintf(page, "%d\n", !!kexec_image); +} +KERNEL_ATTR_RO(kexec_loaded); + +static ssize_t kexec_crash_loaded_show(struct subsystem *subsys, char *page) +{ + return sprintf(page, "%d\n", !!kexec_crash_image); +} +KERNEL_ATTR_RO(kexec_crash_loaded); +#endif /* CONFIG_KEXEC */ + decl_subsys(kernel, NULL, NULL); EXPORT_SYMBOL_GPL(kernel_subsys); @@ -56,6 +71,10 @@ static struct attribute * kernel_attrs[] = { &uevent_seqnum_attr.attr, &uevent_helper_attr.attr, #endif +#ifdef CONFIG_KEXEC + &kexec_loaded_attr.attr, + &kexec_crash_loaded_attr.attr, +#endif NULL }; diff --git a/kernel/kthread.c b/kernel/kthread.c index c5f3c6613b6d..24be714b04c7 100644 --- a/kernel/kthread.c +++ b/kernel/kthread.c @@ -45,6 +45,13 @@ struct kthread_stop_info static DEFINE_MUTEX(kthread_stop_lock); static struct kthread_stop_info kthread_stop_info; +/** + * kthread_should_stop - should this kthread return now? + * + * When someone calls kthread_stop on your kthread, it will be woken + * and this will return true. You should then return, and your return + * value will be passed through to kthread_stop(). + */ int kthread_should_stop(void) { return (kthread_stop_info.k == current); @@ -122,6 +129,25 @@ static void keventd_create_kthread(void *_create) complete(&create->done); } +/** + * kthread_create - create a kthread. + * @threadfn: the function to run until signal_pending(current). + * @data: data ptr for @threadfn. + * @namefmt: printf-style name for the thread. + * + * Description: This helper function creates and names a kernel + * thread. The thread will be stopped: use wake_up_process() to start + * it. See also kthread_run(), kthread_create_on_cpu(). + * + * When woken, the thread will run @threadfn() with @data as its + * argument. @threadfn can either call do_exit() directly if it is a + * standalone thread for which noone will call kthread_stop(), or + * return when 'kthread_should_stop()' is true (which means + * kthread_stop() has been called). The return value should be zero + * or a negative error number; it will be passed to kthread_stop(). + * + * Returns a task_struct or ERR_PTR(-ENOMEM). + */ struct task_struct *kthread_create(int (*threadfn)(void *data), void *data, const char namefmt[], @@ -156,6 +182,15 @@ struct task_struct *kthread_create(int (*threadfn)(void *data), } EXPORT_SYMBOL(kthread_create); +/** + * kthread_bind - bind a just-created kthread to a cpu. + * @k: thread created by kthread_create(). + * @cpu: cpu (might not be online, must be possible) for @k to run on. + * + * Description: This function is equivalent to set_cpus_allowed(), + * except that @cpu doesn't need to be online, and the thread must be + * stopped (i.e., just returned from kthread_create(). + */ void kthread_bind(struct task_struct *k, unsigned int cpu) { BUG_ON(k->state != TASK_INTERRUPTIBLE); @@ -166,12 +201,36 @@ void kthread_bind(struct task_struct *k, unsigned int cpu) } EXPORT_SYMBOL(kthread_bind); +/** + * kthread_stop - stop a thread created by kthread_create(). + * @k: thread created by kthread_create(). + * + * Sets kthread_should_stop() for @k to return true, wakes it, and + * waits for it to exit. Your threadfn() must not call do_exit() + * itself if you use this function! This can also be called after + * kthread_create() instead of calling wake_up_process(): the thread + * will exit without calling threadfn(). + * + * Returns the result of threadfn(), or %-EINTR if wake_up_process() + * was never called. + */ int kthread_stop(struct task_struct *k) { return kthread_stop_sem(k, NULL); } EXPORT_SYMBOL(kthread_stop); +/** + * kthread_stop_sem - stop a thread created by kthread_create(). + * @k: thread created by kthread_create(). + * @s: semaphore that @k waits on while idle. + * + * Does essentially the same thing as kthread_stop() above, but wakes + * @k by calling up(@s). + * + * Returns the result of threadfn(), or %-EINTR if wake_up_process() + * was never called. + */ int kthread_stop_sem(struct task_struct *k, struct semaphore *s) { int ret; @@ -210,5 +269,5 @@ static __init int helper_init(void) return 0; } -core_initcall(helper_init); +core_initcall(helper_init); diff --git a/kernel/module.c b/kernel/module.c index bbe04862e1b0..d75275de1c28 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -1326,7 +1326,7 @@ int is_exported(const char *name, const struct module *mod) if (!mod && lookup_symbol(name, __start___ksymtab, __stop___ksymtab)) return 1; else - if (lookup_symbol(name, mod->syms, mod->syms + mod->num_syms)) + if (mod && lookup_symbol(name, mod->syms, mod->syms + mod->num_syms)) return 1; else return 0; diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c index 520f6c59948d..d38d9ec3276c 100644 --- a/kernel/posix-cpu-timers.c +++ b/kernel/posix-cpu-timers.c @@ -555,9 +555,6 @@ static void arm_timer(struct k_itimer *timer, union cpu_time_count now) struct cpu_timer_list *next; unsigned long i; - if (CPUCLOCK_PERTHREAD(timer->it_clock) && (p->flags & PF_EXITING)) - return; - head = (CPUCLOCK_PERTHREAD(timer->it_clock) ? p->cpu_timers : p->signal->cpu_timers); head += CPUCLOCK_WHICH(timer->it_clock); @@ -1173,6 +1170,9 @@ static void check_process_timers(struct task_struct *tsk, } t = tsk; do { + if (unlikely(t->flags & PF_EXITING)) + continue; + ticks = cputime_add(cputime_add(t->utime, t->stime), prof_left); if (!cputime_eq(prof_expires, cputime_zero) && @@ -1193,11 +1193,7 @@ static void check_process_timers(struct task_struct *tsk, t->it_sched_expires > sched)) { t->it_sched_expires = sched; } - - do { - t = next_thread(t); - } while (unlikely(t->flags & PF_EXITING)); - } while (t != tsk); + } while ((t = next_thread(t)) != tsk); } } @@ -1289,30 +1285,30 @@ void run_posix_cpu_timers(struct task_struct *tsk) #undef UNEXPIRED - BUG_ON(tsk->exit_state); - /* * Double-check with locks held. */ read_lock(&tasklist_lock); - spin_lock(&tsk->sighand->siglock); + if (likely(tsk->signal != NULL)) { + spin_lock(&tsk->sighand->siglock); - /* - * Here we take off tsk->cpu_timers[N] and tsk->signal->cpu_timers[N] - * all the timers that are firing, and put them on the firing list. - */ - check_thread_timers(tsk, &firing); - check_process_timers(tsk, &firing); + /* + * Here we take off tsk->cpu_timers[N] and tsk->signal->cpu_timers[N] + * all the timers that are firing, and put them on the firing list. + */ + check_thread_timers(tsk, &firing); + check_process_timers(tsk, &firing); - /* - * We must release these locks before taking any timer's lock. - * There is a potential race with timer deletion here, as the - * siglock now protects our private firing list. We have set - * the firing flag in each timer, so that a deletion attempt - * that gets the timer lock before we do will give it up and - * spin until we've taken care of that timer below. - */ - spin_unlock(&tsk->sighand->siglock); + /* + * We must release these locks before taking any timer's lock. + * There is a potential race with timer deletion here, as the + * siglock now protects our private firing list. We have set + * the firing flag in each timer, so that a deletion attempt + * that gets the timer lock before we do will give it up and + * spin until we've taken care of that timer below. + */ + spin_unlock(&tsk->sighand->siglock); + } read_unlock(&tasklist_lock); /* diff --git a/kernel/power/Kconfig b/kernel/power/Kconfig index ce0dfb8f4a4e..fc311a4673a2 100644 --- a/kernel/power/Kconfig +++ b/kernel/power/Kconfig @@ -36,6 +36,15 @@ config PM_DEBUG code. This is helpful when debugging and reporting various PM bugs, like suspend support. +config PM_TRACE + bool "Suspend/resume event tracing" + depends on PM && PM_DEBUG && X86_32 + default y + ---help--- + This enables some cheesy code to save the last PM event point in the + RTC across reboots, so that you can debug a machine that just hangs + during suspend (or more commonly, during resume). + config SOFTWARE_SUSPEND bool "Software Suspend" depends on PM && SWAP && (X86 && (!SMP || SUSPEND_SMP)) || ((FRV || PPC32) && !SMP) diff --git a/kernel/power/disk.c b/kernel/power/disk.c index 81d4d982f3f0..e13e74067845 100644 --- a/kernel/power/disk.c +++ b/kernel/power/disk.c @@ -231,7 +231,7 @@ static int software_resume(void) late_initcall(software_resume); -static char * pm_disk_modes[] = { +static const char * const pm_disk_modes[] = { [PM_DISK_FIRMWARE] = "firmware", [PM_DISK_PLATFORM] = "platform", [PM_DISK_SHUTDOWN] = "shutdown", diff --git a/kernel/power/main.c b/kernel/power/main.c index a6d9ef46009e..6d295c776794 100644 --- a/kernel/power/main.c +++ b/kernel/power/main.c @@ -15,7 +15,7 @@ #include <linux/errno.h> #include <linux/init.h> #include <linux/pm.h> - +#include <linux/console.h> #include "power.h" @@ -86,6 +86,7 @@ static int suspend_prepare(suspend_state_t state) goto Thaw; } + suspend_console(); if ((error = device_suspend(PMSG_SUSPEND))) { printk(KERN_ERR "Some devices failed to suspend\n"); goto Finish; @@ -133,6 +134,7 @@ int suspend_enter(suspend_state_t state) static void suspend_finish(suspend_state_t state) { device_resume(); + resume_console(); thaw_processes(); enable_nonboot_cpus(); if (pm_ops && pm_ops->finish) @@ -143,7 +145,7 @@ static void suspend_finish(suspend_state_t state) -static char *pm_states[PM_SUSPEND_MAX] = { +static const char * const pm_states[PM_SUSPEND_MAX] = { [PM_SUSPEND_STANDBY] = "standby", [PM_SUSPEND_MEM] = "mem", #ifdef CONFIG_SOFTWARE_SUSPEND @@ -260,7 +262,7 @@ static ssize_t state_show(struct subsystem * subsys, char * buf) static ssize_t state_store(struct subsystem * subsys, const char * buf, size_t n) { suspend_state_t state = PM_SUSPEND_STANDBY; - char ** s; + const char * const *s; char *p; int error; int len; diff --git a/kernel/power/power.h b/kernel/power/power.h index f06f12f21767..57a792982fb9 100644 --- a/kernel/power/power.h +++ b/kernel/power/power.h @@ -55,7 +55,7 @@ struct snapshot_handle { unsigned int page; unsigned int page_offset; unsigned int prev; - struct pbe *pbe; + struct pbe *pbe, *last_pbe; void *buffer; unsigned int buf_offset; }; diff --git a/kernel/power/snapshot.c b/kernel/power/snapshot.c index 3eeedbb13b78..24c96f354231 100644 --- a/kernel/power/snapshot.c +++ b/kernel/power/snapshot.c @@ -150,6 +150,10 @@ int restore_highmem(void) } return 0; } +#else +static inline unsigned int count_highmem_pages(void) {return 0;} +static inline int save_highmem(void) {return 0;} +static inline int restore_highmem(void) {return 0;} #endif static int pfn_is_nosave(unsigned long pfn) @@ -293,62 +297,29 @@ static inline void create_pbe_list(struct pbe *pblist, unsigned int nr_pages) } } -/** - * On resume it is necessary to trace and eventually free the unsafe - * pages that have been allocated, because they are needed for I/O - * (on x86-64 we likely will "eat" these pages once again while - * creating the temporary page translation tables) - */ - -struct eaten_page { - struct eaten_page *next; - char padding[PAGE_SIZE - sizeof(void *)]; -}; - -static struct eaten_page *eaten_pages = NULL; - -static void release_eaten_pages(void) -{ - struct eaten_page *p, *q; - - p = eaten_pages; - while (p) { - q = p->next; - /* We don't want swsusp_free() to free this page again */ - ClearPageNosave(virt_to_page(p)); - free_page((unsigned long)p); - p = q; - } - eaten_pages = NULL; -} +static unsigned int unsafe_pages; /** * @safe_needed - on resume, for storing the PBE list and the image, * we can only use memory pages that do not conflict with the pages - * which had been used before suspend. + * used before suspend. * * The unsafe pages are marked with the PG_nosave_free flag - * - * Allocated but unusable (ie eaten) memory pages should be marked - * so that swsusp_free() can release them + * and we count them using unsafe_pages */ static inline void *alloc_image_page(gfp_t gfp_mask, int safe_needed) { void *res; + res = (void *)get_zeroed_page(gfp_mask); if (safe_needed) - do { + while (res && PageNosaveFree(virt_to_page(res))) { + /* The page is unsafe, mark it for swsusp_free() */ + SetPageNosave(virt_to_page(res)); + unsafe_pages++; res = (void *)get_zeroed_page(gfp_mask); - if (res && PageNosaveFree(virt_to_page(res))) { - /* This is for swsusp_free() */ - SetPageNosave(virt_to_page(res)); - ((struct eaten_page *)res)->next = eaten_pages; - eaten_pages = res; - } - } while (res && PageNosaveFree(virt_to_page(res))); - else - res = (void *)get_zeroed_page(gfp_mask); + } if (res) { SetPageNosave(virt_to_page(res)); SetPageNosaveFree(virt_to_page(res)); @@ -374,7 +345,8 @@ unsigned long get_safe_page(gfp_t gfp_mask) * On each page we set up a list of struct_pbe elements. */ -struct pbe *alloc_pagedir(unsigned int nr_pages, gfp_t gfp_mask, int safe_needed) +static struct pbe *alloc_pagedir(unsigned int nr_pages, gfp_t gfp_mask, + int safe_needed) { unsigned int num; struct pbe *pblist, *pbe; @@ -642,6 +614,8 @@ static int mark_unsafe_pages(struct pbe *pblist) return -EFAULT; } + unsafe_pages = 0; + return 0; } @@ -719,42 +693,99 @@ static inline struct pbe *unpack_orig_addresses(unsigned long *buf, } /** - * create_image - use metadata contained in the PBE list + * prepare_image - use metadata contained in the PBE list * pointed to by pagedir_nosave to mark the pages that will * be overwritten in the process of restoring the system - * memory state from the image and allocate memory for - * the image avoiding these pages + * memory state from the image ("unsafe" pages) and allocate + * memory for the image + * + * The idea is to allocate the PBE list first and then + * allocate as many pages as it's needed for the image data, + * but not to assign these pages to the PBEs initially. + * Instead, we just mark them as allocated and create a list + * of "safe" which will be used later */ -static int create_image(struct snapshot_handle *handle) +struct safe_page { + struct safe_page *next; + char padding[PAGE_SIZE - sizeof(void *)]; +}; + +static struct safe_page *safe_pages; + +static int prepare_image(struct snapshot_handle *handle) { int error = 0; - struct pbe *p, *pblist; + unsigned int nr_pages = nr_copy_pages; + struct pbe *p, *pblist = NULL; p = pagedir_nosave; error = mark_unsafe_pages(p); if (!error) { - pblist = alloc_pagedir(nr_copy_pages, GFP_ATOMIC, 1); + pblist = alloc_pagedir(nr_pages, GFP_ATOMIC, 1); if (pblist) copy_page_backup_list(pblist, p); free_pagedir(p, 0); if (!pblist) error = -ENOMEM; } - if (!error) - error = alloc_data_pages(pblist, GFP_ATOMIC, 1); + safe_pages = NULL; + if (!error && nr_pages > unsafe_pages) { + nr_pages -= unsafe_pages; + while (nr_pages--) { + struct safe_page *ptr; + + ptr = (struct safe_page *)get_zeroed_page(GFP_ATOMIC); + if (!ptr) { + error = -ENOMEM; + break; + } + if (!PageNosaveFree(virt_to_page(ptr))) { + /* The page is "safe", add it to the list */ + ptr->next = safe_pages; + safe_pages = ptr; + } + /* Mark the page as allocated */ + SetPageNosave(virt_to_page(ptr)); + SetPageNosaveFree(virt_to_page(ptr)); + } + } if (!error) { - release_eaten_pages(); pagedir_nosave = pblist; } else { - pagedir_nosave = NULL; handle->pbe = NULL; - nr_copy_pages = 0; - nr_meta_pages = 0; + swsusp_free(); } return error; } +static void *get_buffer(struct snapshot_handle *handle) +{ + struct pbe *pbe = handle->pbe, *last = handle->last_pbe; + struct page *page = virt_to_page(pbe->orig_address); + + if (PageNosave(page) && PageNosaveFree(page)) { + /* + * We have allocated the "original" page frame and we can + * use it directly to store the read page + */ + pbe->address = 0; + if (last && last->next) + last->next = NULL; + return (void *)pbe->orig_address; + } + /* + * The "original" page frame has not been allocated and we have to + * use a "safe" page frame to store the read page + */ + pbe->address = (unsigned long)safe_pages; + safe_pages = safe_pages->next; + if (last) + last->next = pbe; + handle->last_pbe = pbe; + return (void *)pbe->address; +} + /** * snapshot_write_next - used for writing the system memory snapshot. * @@ -799,15 +830,16 @@ int snapshot_write_next(struct snapshot_handle *handle, size_t count) } else if (handle->prev <= nr_meta_pages) { handle->pbe = unpack_orig_addresses(buffer, handle->pbe); if (!handle->pbe) { - error = create_image(handle); + error = prepare_image(handle); if (error) return error; handle->pbe = pagedir_nosave; - handle->buffer = (void *)handle->pbe->address; + handle->last_pbe = NULL; + handle->buffer = get_buffer(handle); } } else { handle->pbe = handle->pbe->next; - handle->buffer = (void *)handle->pbe->address; + handle->buffer = get_buffer(handle); } handle->prev = handle->page; } diff --git a/kernel/power/swsusp.c b/kernel/power/swsusp.c index c4016cbbd3e0..17f669c83012 100644 --- a/kernel/power/swsusp.c +++ b/kernel/power/swsusp.c @@ -67,9 +67,9 @@ unsigned int count_highmem_pages(void); int save_highmem(void); int restore_highmem(void); #else -static int save_highmem(void) { return 0; } -static int restore_highmem(void) { return 0; } -static unsigned int count_highmem_pages(void) { return 0; } +static inline int save_highmem(void) { return 0; } +static inline int restore_highmem(void) { return 0; } +static inline unsigned int count_highmem_pages(void) { return 0; } #endif /** @@ -175,6 +175,12 @@ void free_all_swap_pages(int swap, struct bitmap_page *bitmap) */ #define SHRINK_BITE 10000 +static inline unsigned long __shrink_memory(long tmp) +{ + if (tmp > SHRINK_BITE) + tmp = SHRINK_BITE; + return shrink_all_memory(tmp); +} int swsusp_shrink_memory(void) { @@ -192,15 +198,17 @@ int swsusp_shrink_memory(void) PAGES_FOR_IO; tmp = size; for_each_zone (zone) - if (!is_highmem(zone)) + if (!is_highmem(zone) && populated_zone(zone)) { tmp -= zone->free_pages; + tmp += zone->lowmem_reserve[ZONE_NORMAL]; + } if (tmp > 0) { - tmp = shrink_all_memory(SHRINK_BITE); + tmp = __shrink_memory(tmp); if (!tmp) return -ENOMEM; pages += tmp; } else if (size > image_size / PAGE_SIZE) { - tmp = shrink_all_memory(SHRINK_BITE); + tmp = __shrink_memory(size - (image_size / PAGE_SIZE)); pages += tmp; } printk("\b%c", p[i++%4]); diff --git a/kernel/printk.c b/kernel/printk.c index c056f3324432..95b7fe17f124 100644 --- a/kernel/printk.c +++ b/kernel/printk.c @@ -24,6 +24,7 @@ #include <linux/console.h> #include <linux/init.h> #include <linux/module.h> +#include <linux/moduleparam.h> #include <linux/interrupt.h> /* For in_interrupt() */ #include <linux/config.h> #include <linux/delay.h> @@ -67,6 +68,7 @@ EXPORT_SYMBOL(oops_in_progress); * driver system. */ static DECLARE_MUTEX(console_sem); +static DECLARE_MUTEX(secondary_console_sem); struct console *console_drivers; /* * This is used for debugging the mess that is the VT code by @@ -76,7 +78,7 @@ struct console *console_drivers; * path in the console code where we end up in places I want * locked without the console sempahore held */ -static int console_locked; +static int console_locked, console_suspended; /* * logbuf_lock protects log_buf, log_start, log_end, con_start and logged_chars @@ -326,7 +328,9 @@ static void __call_console_drivers(unsigned long start, unsigned long end) struct console *con; for (con = console_drivers; con; con = con->next) { - if ((con->flags & CON_ENABLED) && con->write) + if ((con->flags & CON_ENABLED) && con->write && + (cpu_online(smp_processor_id()) || + (con->flags & CON_ANYTIME))) con->write(con, &LOG_BUF(start), end - start); } } @@ -436,6 +440,7 @@ static int printk_time = 1; #else static int printk_time = 0; #endif +module_param(printk_time, int, S_IRUGO | S_IWUSR); static int __init printk_time_setup(char *str) { @@ -452,6 +457,18 @@ __attribute__((weak)) unsigned long long printk_clock(void) return sched_clock(); } +/* Check if we have any console registered that can be called early in boot. */ +static int have_callable_console(void) +{ + struct console *con; + + for (con = console_drivers; con; con = con->next) + if (con->flags & CON_ANYTIME) + return 1; + + return 0; +} + /** * printk - print a kernel message * @fmt: format string @@ -565,27 +582,29 @@ asmlinkage int vprintk(const char *fmt, va_list args) log_level_unknown = 1; } - if (!cpu_online(smp_processor_id())) { + if (!down_trylock(&console_sem)) { /* - * Some console drivers may assume that per-cpu resources have - * been allocated. So don't allow them to be called by this - * CPU until it is officially up. We shouldn't be calling into - * random console drivers on a CPU which doesn't exist yet.. + * We own the drivers. We can drop the spinlock and + * let release_console_sem() print the text, maybe ... */ + console_locked = 1; printk_cpu = UINT_MAX; spin_unlock_irqrestore(&logbuf_lock, flags); - goto out; - } - if (!down_trylock(&console_sem)) { - console_locked = 1; + /* - * We own the drivers. We can drop the spinlock and let - * release_console_sem() print the text + * Console drivers may assume that per-cpu resources have + * been allocated. So unless they're explicitly marked as + * being able to cope (CON_ANYTIME) don't call them until + * this CPU is officially up. */ - printk_cpu = UINT_MAX; - spin_unlock_irqrestore(&logbuf_lock, flags); - console_may_schedule = 0; - release_console_sem(); + if (cpu_online(smp_processor_id()) || have_callable_console()) { + console_may_schedule = 0; + release_console_sem(); + } else { + /* Release by hand to avoid flushing the buffer. */ + console_locked = 0; + up(&console_sem); + } } else { /* * Someone else owns the drivers. We drop the spinlock, which @@ -595,7 +614,7 @@ asmlinkage int vprintk(const char *fmt, va_list args) printk_cpu = UINT_MAX; spin_unlock_irqrestore(&logbuf_lock, flags); } -out: + preempt_enable(); return printed_len; } @@ -698,6 +717,23 @@ int __init add_preferred_console(char *name, int idx, char *options) } /** + * suspend_console - suspend the console subsystem + * + * This disables printk() while we go into suspend states + */ +void suspend_console(void) +{ + acquire_console_sem(); + console_suspended = 1; +} + +void resume_console(void) +{ + console_suspended = 0; + release_console_sem(); +} + +/** * acquire_console_sem - lock the console system for exclusive use. * * Acquires a semaphore which guarantees that the caller has @@ -708,6 +744,10 @@ int __init add_preferred_console(char *name, int idx, char *options) void acquire_console_sem(void) { BUG_ON(in_interrupt()); + if (console_suspended) { + down(&secondary_console_sem); + return; + } down(&console_sem); console_locked = 1; console_may_schedule = 1; @@ -750,6 +790,10 @@ void release_console_sem(void) unsigned long _con_start, _log_end; unsigned long wake_klogd = 0; + if (console_suspended) { + up(&secondary_console_sem); + return; + } for ( ; ; ) { spin_lock_irqsave(&logbuf_lock, flags); wake_klogd |= log_start - log_end; diff --git a/kernel/rcupdate.c b/kernel/rcupdate.c index 2058f88c7bbb..20e9710fc21c 100644 --- a/kernel/rcupdate.c +++ b/kernel/rcupdate.c @@ -612,14 +612,6 @@ void synchronize_rcu(void) wait_for_completion(&rcu.completion); } -/* - * Deprecated, use synchronize_rcu() or synchronize_sched() instead. - */ -void synchronize_kernel(void) -{ - synchronize_rcu(); -} - module_param(blimit, int, 0); module_param(qhimark, int, 0); module_param(qlowmark, int, 0); @@ -627,7 +619,6 @@ module_param(qlowmark, int, 0); module_param(rsinterval, int, 0); #endif EXPORT_SYMBOL_GPL(rcu_batches_completed); -EXPORT_SYMBOL_GPL_FUTURE(call_rcu); /* WARNING: GPL-only in April 2006. */ -EXPORT_SYMBOL_GPL_FUTURE(call_rcu_bh); /* WARNING: GPL-only in April 2006. */ +EXPORT_SYMBOL_GPL(call_rcu); +EXPORT_SYMBOL_GPL(call_rcu_bh); EXPORT_SYMBOL_GPL(synchronize_rcu); -EXPORT_SYMBOL_GPL_FUTURE(synchronize_kernel); /* WARNING: GPL-only in April 2006. */ diff --git a/kernel/sched.c b/kernel/sched.c index c13f1bd2df7d..f06d059edef5 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -3886,6 +3886,10 @@ long sched_setaffinity(pid_t pid, cpumask_t new_mask) !capable(CAP_SYS_NICE)) goto out_unlock; + retval = security_task_setscheduler(p, 0, NULL); + if (retval) + goto out_unlock; + cpus_allowed = cpuset_cpus_allowed(p); cpus_and(new_mask, new_mask, cpus_allowed); retval = set_cpus_allowed(p, new_mask); @@ -3954,7 +3958,10 @@ long sched_getaffinity(pid_t pid, cpumask_t *mask) if (!p) goto out_unlock; - retval = 0; + retval = security_task_getscheduler(p); + if (retval) + goto out_unlock; + cpus_and(*mask, p->cpus_allowed, cpu_online_map); out_unlock: @@ -4046,6 +4053,9 @@ asmlinkage long sys_sched_yield(void) static inline void __cond_resched(void) { +#ifdef CONFIG_DEBUG_SPINLOCK_SLEEP + __might_sleep(__FILE__, __LINE__); +#endif /* * The BKS might be reacquired before we have dropped * PREEMPT_ACTIVE, which could trigger a second @@ -4142,7 +4152,7 @@ EXPORT_SYMBOL(yield); */ void __sched io_schedule(void) { - struct runqueue *rq = &per_cpu(runqueues, raw_smp_processor_id()); + struct runqueue *rq = &__raw_get_cpu_var(runqueues); atomic_inc(&rq->nr_iowait); schedule(); @@ -4153,7 +4163,7 @@ EXPORT_SYMBOL(io_schedule); long __sched io_schedule_timeout(long timeout) { - struct runqueue *rq = &per_cpu(runqueues, raw_smp_processor_id()); + struct runqueue *rq = &__raw_get_cpu_var(runqueues); long ret; atomic_inc(&rq->nr_iowait); @@ -4746,6 +4756,8 @@ static int migration_call(struct notifier_block *nfb, unsigned long action, break; #ifdef CONFIG_HOTPLUG_CPU case CPU_UP_CANCELED: + if (!cpu_rq(cpu)->migration_thread) + break; /* Unbind it from offline cpu so it can run. Fall thru. */ kthread_bind(cpu_rq(cpu)->migration_thread, any_online_cpu(cpu_online_map)); diff --git a/kernel/signal.c b/kernel/signal.c index e5f8aea78ffe..1b3c921737e2 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -23,12 +23,12 @@ #include <linux/syscalls.h> #include <linux/ptrace.h> #include <linux/signal.h> -#include <linux/audit.h> #include <linux/capability.h> #include <asm/param.h> #include <asm/uaccess.h> #include <asm/unistd.h> #include <asm/siginfo.h> +#include "audit.h" /* audit_signal_info() */ /* * SLAB caches for signal bits. diff --git a/kernel/softirq.c b/kernel/softirq.c index 336f92d64e2e..9e2f1c6e73d7 100644 --- a/kernel/softirq.c +++ b/kernel/softirq.c @@ -470,6 +470,8 @@ static int cpu_callback(struct notifier_block *nfb, break; #ifdef CONFIG_HOTPLUG_CPU case CPU_UP_CANCELED: + if (!per_cpu(ksoftirqd, hotcpu)) + break; /* Unbind so it can run. Fall thru. */ kthread_bind(per_cpu(ksoftirqd, hotcpu), any_online_cpu(cpu_online_map)); diff --git a/kernel/softlockup.c b/kernel/softlockup.c index 14c7faf02909..b5c3b94e01ce 100644 --- a/kernel/softlockup.c +++ b/kernel/softlockup.c @@ -36,7 +36,7 @@ static struct notifier_block panic_block = { void touch_softlockup_watchdog(void) { - per_cpu(touch_timestamp, raw_smp_processor_id()) = jiffies; + __raw_get_cpu_var(touch_timestamp) = jiffies; } EXPORT_SYMBOL(touch_softlockup_watchdog); @@ -127,6 +127,8 @@ cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu) break; #ifdef CONFIG_HOTPLUG_CPU case CPU_UP_CANCELED: + if (!per_cpu(watchdog_task, hotcpu)) + break; /* Unbind so it can run. Fall thru. */ kthread_bind(per_cpu(watchdog_task, hotcpu), any_online_cpu(cpu_online_map)); diff --git a/kernel/stop_machine.c b/kernel/stop_machine.c index dcfb5d731466..2c0aacc37c55 100644 --- a/kernel/stop_machine.c +++ b/kernel/stop_machine.c @@ -4,6 +4,7 @@ #include <linux/cpu.h> #include <linux/err.h> #include <linux/syscalls.h> +#include <linux/kthread.h> #include <asm/atomic.h> #include <asm/semaphore.h> #include <asm/uaccess.h> @@ -25,13 +26,11 @@ static unsigned int stopmachine_num_threads; static atomic_t stopmachine_thread_ack; static DECLARE_MUTEX(stopmachine_mutex); -static int stopmachine(void *cpu) +static int stopmachine(void *unused) { int irqs_disabled = 0; int prepared = 0; - set_cpus_allowed(current, cpumask_of_cpu((int)(long)cpu)); - /* Ack: we are alive */ smp_mb(); /* Theoretically the ack = 0 might not be on this CPU yet. */ atomic_inc(&stopmachine_thread_ack); @@ -85,7 +84,8 @@ static void stopmachine_set_state(enum stopmachine_state state) static int stop_machine(void) { - int i, ret = 0; + int ret = 0; + unsigned int i; struct sched_param param = { .sched_priority = MAX_RT_PRIO-1 }; /* One high-prio thread per cpu. We'll do this one. */ @@ -96,11 +96,16 @@ static int stop_machine(void) stopmachine_state = STOPMACHINE_WAIT; for_each_online_cpu(i) { + struct task_struct *tsk; if (i == raw_smp_processor_id()) continue; - ret = kernel_thread(stopmachine, (void *)(long)i,CLONE_KERNEL); - if (ret < 0) + tsk = kthread_create(stopmachine, NULL, "stopmachine"); + if (IS_ERR(tsk)) { + ret = PTR_ERR(tsk); break; + } + kthread_bind(tsk, i); + wake_up_process(tsk); stopmachine_num_threads++; } diff --git a/kernel/sys.c b/kernel/sys.c index 0b6ec0e7936f..2d5179c67cec 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -13,7 +13,6 @@ #include <linux/notifier.h> #include <linux/reboot.h> #include <linux/prctl.h> -#include <linux/init.h> #include <linux/highuid.h> #include <linux/fs.h> #include <linux/kernel.h> @@ -57,6 +56,12 @@ #ifndef GET_FPEXC_CTL # define GET_FPEXC_CTL(a,b) (-EINVAL) #endif +#ifndef GET_ENDIAN +# define GET_ENDIAN(a,b) (-EINVAL) +#endif +#ifndef SET_ENDIAN +# define SET_ENDIAN(a,b) (-EINVAL) +#endif /* * this is where the system-wide overflow UID and GID are defined, for @@ -132,14 +137,15 @@ static int __kprobes notifier_call_chain(struct notifier_block **nl, unsigned long val, void *v) { int ret = NOTIFY_DONE; - struct notifier_block *nb; + struct notifier_block *nb, *next_nb; nb = rcu_dereference(*nl); while (nb) { + next_nb = rcu_dereference(nb->next); ret = nb->notifier_call(nb, val, v); if ((ret & NOTIFY_STOP_MASK) == NOTIFY_STOP_MASK) break; - nb = rcu_dereference(nb->next); + nb = next_nb; } return ret; } @@ -583,7 +589,7 @@ void emergency_restart(void) } EXPORT_SYMBOL_GPL(emergency_restart); -void kernel_restart_prepare(char *cmd) +static void kernel_restart_prepare(char *cmd) { blocking_notifier_call_chain(&reboot_notifier_list, SYS_RESTART, cmd); system_state = SYSTEM_RESTART; @@ -617,7 +623,7 @@ EXPORT_SYMBOL_GPL(kernel_restart); * Move into place and start executing a preloaded standalone * executable. If nothing was preloaded return an error. */ -void kernel_kexec(void) +static void kernel_kexec(void) { #ifdef CONFIG_KEXEC struct kimage *image; @@ -631,7 +637,6 @@ void kernel_kexec(void) machine_kexec(image); #endif } -EXPORT_SYMBOL_GPL(kernel_kexec); void kernel_shutdown_prepare(enum system_states state) { @@ -1860,23 +1865,20 @@ out: * fields when reaping, so a sample either gets all the additions of a * given child after it's reaped, or none so this sample is before reaping. * - * tasklist_lock locking optimisation: - * If we are current and single threaded, we do not need to take the tasklist - * lock or the siglock. No one else can take our signal_struct away, - * no one else can reap the children to update signal->c* counters, and - * no one else can race with the signal-> fields. - * If we do not take the tasklist_lock, the signal-> fields could be read - * out of order while another thread was just exiting. So we place a - * read memory barrier when we avoid the lock. On the writer side, - * write memory barrier is implied in __exit_signal as __exit_signal releases - * the siglock spinlock after updating the signal-> fields. - * - * We don't really need the siglock when we access the non c* fields - * of the signal_struct (for RUSAGE_SELF) even in multithreaded - * case, since we take the tasklist lock for read and the non c* signal-> - * fields are updated only in __exit_signal, which is called with - * tasklist_lock taken for write, hence these two threads cannot execute - * concurrently. + * Locking: + * We need to take the siglock for CHILDEREN, SELF and BOTH + * for the cases current multithreaded, non-current single threaded + * non-current multithreaded. Thread traversal is now safe with + * the siglock held. + * Strictly speaking, we donot need to take the siglock if we are current and + * single threaded, as no one else can take our signal_struct away, no one + * else can reap the children to update signal->c* counters, and no one else + * can race with the signal-> fields. If we do not take any lock, the + * signal-> fields could be read out of order while another thread was just + * exiting. So we should place a read memory barrier when we avoid the lock. + * On the writer side, write memory barrier is implied in __exit_signal + * as __exit_signal releases the siglock spinlock after updating the signal-> + * fields. But we don't do this yet to keep things simple. * */ @@ -1885,35 +1887,25 @@ static void k_getrusage(struct task_struct *p, int who, struct rusage *r) struct task_struct *t; unsigned long flags; cputime_t utime, stime; - int need_lock = 0; memset((char *) r, 0, sizeof *r); utime = stime = cputime_zero; - if (p != current || !thread_group_empty(p)) - need_lock = 1; - - if (need_lock) { - read_lock(&tasklist_lock); - if (unlikely(!p->signal)) { - read_unlock(&tasklist_lock); - return; - } - } else - /* See locking comments above */ - smp_rmb(); + rcu_read_lock(); + if (!lock_task_sighand(p, &flags)) { + rcu_read_unlock(); + return; + } switch (who) { case RUSAGE_BOTH: case RUSAGE_CHILDREN: - spin_lock_irqsave(&p->sighand->siglock, flags); utime = p->signal->cutime; stime = p->signal->cstime; r->ru_nvcsw = p->signal->cnvcsw; r->ru_nivcsw = p->signal->cnivcsw; r->ru_minflt = p->signal->cmin_flt; r->ru_majflt = p->signal->cmaj_flt; - spin_unlock_irqrestore(&p->sighand->siglock, flags); if (who == RUSAGE_CHILDREN) break; @@ -1941,8 +1933,9 @@ static void k_getrusage(struct task_struct *p, int who, struct rusage *r) BUG(); } - if (need_lock) - read_unlock(&tasklist_lock); + unlock_task_sighand(p, &flags); + rcu_read_unlock(); + cputime_to_timeval(utime, &r->ru_utime); cputime_to_timeval(stime, &r->ru_stime); } @@ -2057,6 +2050,13 @@ asmlinkage long sys_prctl(int option, unsigned long arg2, unsigned long arg3, return -EFAULT; return 0; } + case PR_GET_ENDIAN: + error = GET_ENDIAN(current, arg2); + break; + case PR_SET_ENDIAN: + error = SET_ENDIAN(current, arg2); + break; + default: error = -EINVAL; break; diff --git a/kernel/sys_ni.c b/kernel/sys_ni.c index 5433195040f1..6991bece67e8 100644 --- a/kernel/sys_ni.c +++ b/kernel/sys_ni.c @@ -87,6 +87,7 @@ cond_syscall(sys_inotify_init); cond_syscall(sys_inotify_add_watch); cond_syscall(sys_inotify_rm_watch); cond_syscall(sys_migrate_pages); +cond_syscall(sys_move_pages); cond_syscall(sys_chown16); cond_syscall(sys_fchown16); cond_syscall(sys_getegid16); @@ -132,3 +133,4 @@ cond_syscall(sys_mincore); cond_syscall(sys_madvise); cond_syscall(sys_mremap); cond_syscall(sys_remap_file_pages); +cond_syscall(compat_sys_move_pages); diff --git a/kernel/sysctl.c b/kernel/sysctl.c index e82726faeeff..2c0e65819448 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -59,6 +59,7 @@ extern int proc_nr_files(ctl_table *table, int write, struct file *filp, extern int C_A_D; extern int sysctl_overcommit_memory; extern int sysctl_overcommit_ratio; +extern int sysctl_panic_on_oom; extern int max_threads; extern int sysrq_enabled; extern int core_uses_pid; @@ -142,7 +143,6 @@ static struct ctl_table_header root_table_header = static ctl_table kern_table[]; static ctl_table vm_table[]; -static ctl_table proc_table[]; static ctl_table fs_table[]; static ctl_table debug_table[]; static ctl_table dev_table[]; @@ -150,7 +150,7 @@ extern ctl_table random_table[]; #ifdef CONFIG_UNIX98_PTYS extern ctl_table pty_table[]; #endif -#ifdef CONFIG_INOTIFY +#ifdef CONFIG_INOTIFY_USER extern ctl_table inotify_table[]; #endif @@ -202,12 +202,6 @@ static ctl_table root_table[] = { }, #endif { - .ctl_name = CTL_PROC, - .procname = "proc", - .mode = 0555, - .child = proc_table, - }, - { .ctl_name = CTL_FS, .procname = "fs", .mode = 0555, @@ -398,7 +392,7 @@ static ctl_table kern_table[] = { .strategy = &sysctl_string, }, #endif -#ifdef CONFIG_HOTPLUG +#if defined(CONFIG_HOTPLUG) && defined(CONFIG_NET) { .ctl_name = KERN_HOTPLUG, .procname = "hotplug", @@ -702,6 +696,14 @@ static ctl_table vm_table[] = { .proc_handler = &proc_dointvec, }, { + .ctl_name = VM_PANIC_ON_OOM, + .procname = "panic_on_oom", + .data = &sysctl_panic_on_oom, + .maxlen = sizeof(sysctl_panic_on_oom), + .mode = 0644, + .proc_handler = &proc_dointvec, + }, + { .ctl_name = VM_OVERCOMMIT_RATIO, .procname = "overcommit_ratio", .data = &sysctl_overcommit_ratio, @@ -918,10 +920,6 @@ static ctl_table vm_table[] = { { .ctl_name = 0 } }; -static ctl_table proc_table[] = { - { .ctl_name = 0 } -}; - static ctl_table fs_table[] = { { .ctl_name = FS_NRINODE, @@ -1028,7 +1026,7 @@ static ctl_table fs_table[] = { .mode = 0644, .proc_handler = &proc_doulongvec_minmax, }, -#ifdef CONFIG_INOTIFY +#ifdef CONFIG_INOTIFY_USER { .ctl_name = FS_INOTIFY, .procname = "inotify", diff --git a/kernel/timer.c b/kernel/timer.c index 9e49deed468c..eb97371b87d8 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -146,7 +146,7 @@ static void internal_add_timer(tvec_base_t *base, struct timer_list *timer) void fastcall init_timer(struct timer_list *timer) { timer->entry.next = NULL; - timer->base = per_cpu(tvec_bases, raw_smp_processor_id()); + timer->base = __raw_get_cpu_var(tvec_bases); } EXPORT_SYMBOL(init_timer); @@ -383,23 +383,19 @@ EXPORT_SYMBOL(del_timer_sync); static int cascade(tvec_base_t *base, tvec_t *tv, int index) { /* cascade all the timers from tv up one level */ - struct list_head *head, *curr; + struct timer_list *timer, *tmp; + struct list_head tv_list; + + list_replace_init(tv->vec + index, &tv_list); - head = tv->vec + index; - curr = head->next; /* - * We are removing _all_ timers from the list, so we don't have to - * detach them individually, just clear the list afterwards. + * We are removing _all_ timers from the list, so we + * don't have to detach them individually. */ - while (curr != head) { - struct timer_list *tmp; - - tmp = list_entry(curr, struct timer_list, entry); - BUG_ON(tmp->base != base); - curr = curr->next; - internal_add_timer(base, tmp); + list_for_each_entry_safe(timer, tmp, &tv_list, entry) { + BUG_ON(timer->base != base); + internal_add_timer(base, timer); } - INIT_LIST_HEAD(head); return index; } @@ -419,10 +415,10 @@ static inline void __run_timers(tvec_base_t *base) spin_lock_irq(&base->lock); while (time_after_eq(jiffies, base->timer_jiffies)) { - struct list_head work_list = LIST_HEAD_INIT(work_list); + struct list_head work_list; struct list_head *head = &work_list; int index = base->timer_jiffies & TVR_MASK; - + /* * Cascade timers: */ @@ -431,8 +427,8 @@ static inline void __run_timers(tvec_base_t *base) (!cascade(base, &base->tv3, INDEX(1))) && !cascade(base, &base->tv4, INDEX(2))) cascade(base, &base->tv5, INDEX(3)); - ++base->timer_jiffies; - list_splice_init(base->tv1.vec + index, &work_list); + ++base->timer_jiffies; + list_replace_init(base->tv1.vec + index, &work_list); while (!list_empty(head)) { void (*fn)(unsigned long); unsigned long data; diff --git a/kernel/user.c b/kernel/user.c index 2116642f42c6..6408c0424291 100644 --- a/kernel/user.c +++ b/kernel/user.c @@ -140,7 +140,7 @@ struct user_struct * alloc_uid(uid_t uid) atomic_set(&new->processes, 0); atomic_set(&new->files, 0); atomic_set(&new->sigpending, 0); -#ifdef CONFIG_INOTIFY +#ifdef CONFIG_INOTIFY_USER atomic_set(&new->inotify_watches, 0); atomic_set(&new->inotify_devs, 0); #endif @@ -148,7 +148,7 @@ struct user_struct * alloc_uid(uid_t uid) new->mq_bytes = 0; new->locked_shm = 0; - if (alloc_uid_keyring(new) < 0) { + if (alloc_uid_keyring(new, current) < 0) { kmem_cache_free(uid_cachep, new); return NULL; } diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 880fb415a8f6..565cf7a1febd 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -428,22 +428,34 @@ int schedule_delayed_work_on(int cpu, return ret; } -int schedule_on_each_cpu(void (*func) (void *info), void *info) +/** + * schedule_on_each_cpu - call a function on each online CPU from keventd + * @func: the function to call + * @info: a pointer to pass to func() + * + * Returns zero on success. + * Returns -ve errno on failure. + * + * Appears to be racy against CPU hotplug. + * + * schedule_on_each_cpu() is very slow. + */ +int schedule_on_each_cpu(void (*func)(void *info), void *info) { int cpu; - struct work_struct *work; + struct work_struct *works; - work = kmalloc(NR_CPUS * sizeof(struct work_struct), GFP_KERNEL); - - if (!work) + works = alloc_percpu(struct work_struct); + if (!works) return -ENOMEM; + for_each_online_cpu(cpu) { - INIT_WORK(work + cpu, func, info); + INIT_WORK(per_cpu_ptr(works, cpu), func, info); __queue_work(per_cpu_ptr(keventd_wq->cpu_wq, cpu), - work + cpu); + per_cpu_ptr(works, cpu)); } flush_workqueue(keventd_wq); - kfree(work); + free_percpu(works); return 0; } @@ -531,11 +543,11 @@ int current_is_keventd(void) static void take_over_work(struct workqueue_struct *wq, unsigned int cpu) { struct cpu_workqueue_struct *cwq = per_cpu_ptr(wq->cpu_wq, cpu); - LIST_HEAD(list); + struct list_head list; struct work_struct *work; spin_lock_irq(&cwq->lock); - list_splice_init(&cwq->worklist, &list); + list_replace_init(&cwq->worklist, &list); while (!list_empty(&list)) { printk("Taking work for %s\n", wq->name); @@ -578,6 +590,8 @@ static int workqueue_cpu_callback(struct notifier_block *nfb, case CPU_UP_CANCELED: list_for_each_entry(wq, &workqueues, list) { + if (!per_cpu_ptr(wq->cpu_wq, hotcpu)->thread) + continue; /* Unbind so it can run. */ kthread_bind(per_cpu_ptr(wq->cpu_wq, hotcpu)->thread, any_online_cpu(cpu_online_map)); |