Index: arch/ppc/Kconfig =================================================================== --- 3c5e9440c6a37c3355b50608836a23c8fa4eec99/arch/ppc/Kconfig (mode:100644) +++ 672465b3b9b3e2e26a8caf74ed64aa6885c52c13/arch/ppc/Kconfig (mode:100644) @@ -1083,6 +1083,23 @@ source kernel/power/Kconfig +config SECCOMP + bool "Enable seccomp to safely compute untrusted bytecode" + depends on PROC_FS + default y + help + This kernel feature is useful for number crunching applications + that may need to compute untrusted bytecode during their + execution. By using pipes or other transports made available to + the process as file descriptors supporting the read/write + syscalls, it's possible to isolate those applications in + their own address space using seccomp. Once seccomp is + enabled via /proc//seccomp, it cannot be disabled + and the task is only allowed to execute a few safe syscalls + defined by each seccomp mode. + + If unsure, say Y. Only embedded should say N here. + endmenu config ISA_DMA_API Index: arch/ppc/kernel/entry.S =================================================================== --- 3c5e9440c6a37c3355b50608836a23c8fa4eec99/arch/ppc/kernel/entry.S (mode:100644) +++ 672465b3b9b3e2e26a8caf74ed64aa6885c52c13/arch/ppc/kernel/entry.S (mode:100644) @@ -202,7 +202,7 @@ rlwinm r11,r11,0,~_TIFL_FORCE_NOERROR stw r11,TI_LOCAL_FLAGS(r10) lwz r11,TI_FLAGS(r10) - andi. r11,r11,_TIF_SYSCALL_TRACE + andi. r11,r11,_TIF_SYSCALL_T_OR_A bne- syscall_dotrace syscall_dotrace_cont: cmplwi 0,r0,NR_syscalls @@ -237,7 +237,7 @@ SYNC MTMSRD(r10) lwz r9,TI_FLAGS(r12) - andi. r0,r9,(_TIF_SYSCALL_TRACE|_TIF_SIGPENDING|_TIF_NEED_RESCHED) + andi. r0,r9,(_TIF_SYSCALL_T_OR_A|_TIF_SIGPENDING|_TIF_NEED_RESCHED) bne- syscall_exit_work syscall_exit_cont: #if defined(CONFIG_4xx) || defined(CONFIG_BOOKE) @@ -277,7 +277,8 @@ SAVE_NVGPRS(r1) li r0,0xc00 stw r0,TRAP(r1) - bl do_syscall_trace + addi r3,r1,STACK_FRAME_OVERHEAD + bl do_syscall_trace_enter lwz r0,GPR0(r1) /* Restore original registers */ lwz r3,GPR3(r1) lwz r4,GPR4(r1) @@ -291,7 +292,7 @@ syscall_exit_work: stw r6,RESULT(r1) /* Save result */ stw r3,GPR3(r1) /* Update return value */ - andi. r0,r9,_TIF_SYSCALL_TRACE + andi. r0,r9,_TIF_SYSCALL_T_OR_A beq 5f ori r10,r10,MSR_EE SYNC @@ -303,7 +304,8 @@ li r4,0xc00 stw r4,TRAP(r1) 4: - bl do_syscall_trace + addi r3,r1,STACK_FRAME_OVERHEAD + bl do_syscall_trace_leave REST_NVGPRS(r1) 2: lwz r3,GPR3(r1) @@ -627,8 +629,8 @@ subi r1,r3,STACK_FRAME_OVERHEAD rlwinm r12,r1,0,0,18 /* current_thread_info() */ lwz r9,TI_FLAGS(r12) - andi. r0,r9,_TIF_SYSCALL_TRACE - bnel- do_syscall_trace + andi. r0,r9,_TIF_SYSCALL_T_OR_A + bnel- do_syscall_trace_leave /* fall through */ .globl ret_from_except_full Index: arch/ppc/kernel/ppc_ksyms.c =================================================================== --- 3c5e9440c6a37c3355b50608836a23c8fa4eec99/arch/ppc/kernel/ppc_ksyms.c (mode:100644) +++ 672465b3b9b3e2e26a8caf74ed64aa6885c52c13/arch/ppc/kernel/ppc_ksyms.c (mode:100644) @@ -55,7 +55,6 @@ #define EXPORT_SYMTAB_STROPS extern void transfer_to_handler(void); -extern void do_syscall_trace(void); extern void do_IRQ(struct pt_regs *regs); extern void MachineCheckException(struct pt_regs *regs); extern void AlignmentException(struct pt_regs *regs); @@ -74,7 +73,6 @@ EXPORT_SYMBOL(clear_pages); EXPORT_SYMBOL(clear_user_page); EXPORT_SYMBOL(do_signal); -EXPORT_SYMBOL(do_syscall_trace); EXPORT_SYMBOL(transfer_to_handler); EXPORT_SYMBOL(do_IRQ); EXPORT_SYMBOL(MachineCheckException); Index: arch/ppc/kernel/ptrace.c =================================================================== --- 3c5e9440c6a37c3355b50608836a23c8fa4eec99/arch/ppc/kernel/ptrace.c (mode:100644) +++ 672465b3b9b3e2e26a8caf74ed64aa6885c52c13/arch/ppc/kernel/ptrace.c (mode:100644) @@ -27,6 +27,9 @@ #include #include #include +#include +#include +#include #include #include @@ -455,11 +458,10 @@ return ret; } -void do_syscall_trace(void) +static void do_syscall_trace(void) { - if (!test_thread_flag(TIF_SYSCALL_TRACE) - || !(current->ptrace & PT_PTRACED)) - return; + /* the 0x80 provides a way for the tracing parent to distinguish + between a syscall stop and SIGTRAP delivery */ ptrace_notify(SIGTRAP | ((current->ptrace & PT_TRACESYSGOOD) ? 0x80 : 0)); @@ -473,3 +475,33 @@ current->exit_code = 0; } } + +void do_syscall_trace_enter(struct pt_regs *regs) +{ + if (test_thread_flag(TIF_SYSCALL_TRACE) + && (current->ptrace & PT_PTRACED)) + do_syscall_trace(); + + if (unlikely(current->audit_context)) + audit_syscall_entry(current, AUDIT_ARCH_PPC, + regs->gpr[0], + regs->gpr[3], regs->gpr[4], + regs->gpr[5], regs->gpr[6]); +} + +void do_syscall_trace_leave(struct pt_regs *regs) +{ + secure_computing(regs->gpr[0]); + + if (unlikely(current->audit_context)) + audit_syscall_exit(current, + (regs->ccr&0x1000)?AUDITSC_FAILURE:AUDITSC_SUCCESS, + regs->result); + + if ((test_thread_flag(TIF_SYSCALL_TRACE)) + && (current->ptrace & PT_PTRACED)) + do_syscall_trace(); +} + +EXPORT_SYMBOL(do_syscall_trace_enter); +EXPORT_SYMBOL(do_syscall_trace_leave); Index: include/asm-ppc/seccomp.h =================================================================== --- /dev/null (tree:3c5e9440c6a37c3355b50608836a23c8fa4eec99) +++ 672465b3b9b3e2e26a8caf74ed64aa6885c52c13/include/asm-ppc/seccomp.h (mode:100644) @@ -0,0 +1,10 @@ +#ifndef _ASM_SECCOMP_H + +#include + +#define __NR_seccomp_read __NR_read +#define __NR_seccomp_write __NR_write +#define __NR_seccomp_exit __NR_exit +#define __NR_seccomp_sigreturn __NR_rt_sigreturn + +#endif /* _ASM_SECCOMP_H */ Index: include/asm-ppc/thread_info.h =================================================================== --- 3c5e9440c6a37c3355b50608836a23c8fa4eec99/include/asm-ppc/thread_info.h (mode:100644) +++ 672465b3b9b3e2e26a8caf74ed64aa6885c52c13/include/asm-ppc/thread_info.h (mode:100644) @@ -77,12 +77,19 @@ #define TIF_POLLING_NRFLAG 4 /* true if poll_idle() is polling TIF_NEED_RESCHED */ #define TIF_MEMDIE 5 +#define TIF_SYSCALL_AUDIT 6 /* syscall auditing active */ +#define TIF_SECCOMP 7 /* secure computing */ + /* as above, but as bit values */ #define _TIF_SYSCALL_TRACE (1< #include -/* Request and reply types */ -#define AUDIT_GET 1000 /* Get status */ -#define AUDIT_SET 1001 /* Set status (enable/disable/auditd) */ -#define AUDIT_LIST 1002 /* List filtering rules */ -#define AUDIT_ADD 1003 /* Add filtering rule */ -#define AUDIT_DEL 1004 /* Delete filtering rule */ -#define AUDIT_USER 1005 /* Send a message from user-space */ -#define AUDIT_LOGIN 1006 /* Define the login id and informaiton */ -#define AUDIT_KERNEL 2000 /* Asynchronous audit record. NOT A REQUEST. */ +/* The netlink messages for the audit system is divided into blocks: + * 1000 - 1099 are for commanding the audit system + * 1100 - 1199 user space trusted application messages + * 1200 - 1299 messages internal to the audit daemon + * 1300 - 1399 audit event messages + * 1400 - 1499 SE Linux use + * 1500 - 1999 future use + * 2000 is for otherwise unclassified kernel audit messages + * + * Messages from 1000-1199 are bi-directional. 1200-1299 are exclusively user + * space. Anything over that is kernel --> user space communication. + */ +#define AUDIT_GET 1000 /* Get status */ +#define AUDIT_SET 1001 /* Set status (enable/disable/auditd) */ +#define AUDIT_LIST 1002 /* List syscall filtering rules */ +#define AUDIT_ADD 1003 /* Add syscall filtering rule */ +#define AUDIT_DEL 1004 /* Delete syscall filtering rule */ +#define AUDIT_USER 1005 /* Message from userspace -- deprecated */ +#define AUDIT_LOGIN 1006 /* Define the login id and information */ +#define AUDIT_WATCH_INS 1007 /* Insert file/dir watch entry */ +#define AUDIT_WATCH_REM 1008 /* Remove file/dir watch entry */ +#define AUDIT_WATCH_LIST 1009 /* List all file/dir watches */ +#define AUDIT_SIGNAL_INFO 1010 /* Get info about sender of signal to auditd */ + +#define AUDIT_FIRST_USER_MSG 1100 /* Userspace messages uninteresting to kernel */ +#define AUDIT_LAST_USER_MSG 1199 + +#define AUDIT_DAEMON_START 1200 /* Daemon startup record */ +#define AUDIT_DAEMON_END 1201 /* Daemon normal stop record */ +#define AUDIT_DAEMON_ABORT 1202 /* Daemon error stop record */ +#define AUDIT_DAEMON_CONFIG 1203 /* Daemon config change */ + +#define AUDIT_SYSCALL 1300 /* Syscall event */ +#define AUDIT_FS_WATCH 1301 /* Filesystem watch event */ +#define AUDIT_PATH 1302 /* Filename path information */ +#define AUDIT_IPC 1303 /* IPC record */ +#define AUDIT_SOCKETCALL 1304 /* sys_socketcall arguments */ +#define AUDIT_CONFIG_CHANGE 1305 /* Audit system configuration change */ +#define AUDIT_SOCKADDR 1306 /* sockaddr copied as syscall arg */ +#define AUDIT_CWD 1307 /* Current working directory */ + +#define AUDIT_AVC 1400 /* SE Linux avc denial or grant */ +#define AUDIT_SELINUX_ERR 1401 /* Internal SE Linux Errors */ +#define AUDIT_AVC_PATH 1402 /* dentry, vfsmount pair from avc */ + +#define AUDIT_KERNEL 2000 /* Asynchronous audit record. NOT A REQUEST. */ /* Rule flags */ #define AUDIT_PER_TASK 0x01 /* Apply rule at task creation (not syscall) */ @@ -132,16 +169,9 @@ #define AUDIT_ARCH_V850 (EM_V850|__AUDIT_ARCH_LE) #define AUDIT_ARCH_X86_64 (EM_X86_64|__AUDIT_ARCH_64BIT|__AUDIT_ARCH_LE) -#ifndef __KERNEL__ -struct audit_message { - struct nlmsghdr nlh; - char data[1200]; -}; -#endif - struct audit_status { __u32 mask; /* Bit mask for valid entries */ - __u32 enabled; /* 1 = enabled, 0 = disbaled */ + __u32 enabled; /* 1 = enabled, 0 = disabled */ __u32 failure; /* Failure-to-log action */ __u32 pid; /* pid of auditd process */ __u32 rate_limit; /* messages rate limit (per second) */ @@ -161,6 +191,11 @@ #ifdef __KERNEL__ +struct audit_sig_info { + uid_t uid; + pid_t pid; +}; + struct audit_buffer; struct audit_context; struct inode; @@ -185,11 +220,16 @@ /* Private API (for audit.c only) */ extern int audit_receive_filter(int type, int pid, int uid, int seq, void *data, uid_t loginuid); -extern void audit_get_stamp(struct audit_context *ctx, - struct timespec *t, unsigned int *serial); +extern unsigned int audit_serial(void); +extern void auditsc_get_stamp(struct audit_context *ctx, + struct timespec *t, unsigned int *serial); extern int audit_set_loginuid(struct task_struct *task, uid_t loginuid); extern uid_t audit_get_loginuid(struct audit_context *ctx); extern int audit_ipc_perms(unsigned long qbytes, uid_t uid, gid_t gid, mode_t mode); +extern int audit_socketcall(int nargs, unsigned long *args); +extern int audit_sockaddr(int len, void *addr); +extern int audit_avc_path(struct dentry *dentry, struct vfsmount *mnt); +extern void audit_signal_info(int sig, struct task_struct *t); #else #define audit_alloc(t) ({ 0; }) #define audit_free(t) do { ; } while (0) @@ -198,18 +238,24 @@ #define audit_getname(n) do { ; } while (0) #define audit_putname(n) do { ; } while (0) #define audit_inode(n,i) do { ; } while (0) +#define audit_receive_filter(t,p,u,s,d,l) ({ -EOPNOTSUPP; }) +#define auditsc_get_stamp(c,t,s) do { BUG(); } while (0) #define audit_get_loginuid(c) ({ -1; }) #define audit_ipc_perms(q,u,g,m) ({ 0; }) +#define audit_socketcall(n,a) ({ 0; }) +#define audit_sockaddr(len, addr) ({ 0; }) +#define audit_avc_path(dentry, mnt) ({ 0; }) +#define audit_signal_info(s,t) do { ; } while (0) #endif #ifdef CONFIG_AUDIT /* These are defined in audit.c */ /* Public API */ -extern void audit_log(struct audit_context *ctx, +extern void audit_log(struct audit_context *ctx, int type, const char *fmt, ...) - __attribute__((format(printf,2,3))); + __attribute__((format(printf,3,4))); -extern struct audit_buffer *audit_log_start(struct audit_context *ctx); +extern struct audit_buffer *audit_log_start(struct audit_context *ctx,int type); extern void audit_log_format(struct audit_buffer *ab, const char *fmt, ...) __attribute__((format(printf,2,3))); @@ -229,8 +275,8 @@ void *payload, int size); extern void audit_log_lost(const char *message); #else -#define audit_log(t,f,...) do { ; } while (0) -#define audit_log_start(t) ({ NULL; }) +#define audit_log(c,t,f,...) do { ; } while (0) +#define audit_log_start(c,t) ({ NULL; }) #define audit_log_vformat(b,f,a) do { ; } while (0) #define audit_log_format(b,f,...) do { ; } while (0) #define audit_log_end(b) do { ; } while (0) Index: init/Kconfig =================================================================== --- 3c5e9440c6a37c3355b50608836a23c8fa4eec99/init/Kconfig (mode:100644) +++ 672465b3b9b3e2e26a8caf74ed64aa6885c52c13/init/Kconfig (mode:100644) @@ -164,6 +164,7 @@ config AUDIT bool "Auditing support" + depends on NET default y if SECURITY_SELINUX help Enable auditing infrastructure that can be used with another @@ -173,7 +174,7 @@ config AUDITSYSCALL bool "Enable system-call auditing support" - depends on AUDIT && (X86 || PPC64 || ARCH_S390 || IA64 || UML) + depends on AUDIT && (X86 || PPC || PPC64 || ARCH_S390 || IA64 || UML) default y if SECURITY_SELINUX help Enable low-overhead system-call auditing infrastructure that Index: kernel/audit.c =================================================================== --- 3c5e9440c6a37c3355b50608836a23c8fa4eec99/kernel/audit.c (mode:100644) +++ 672465b3b9b3e2e26a8caf74ed64aa6885c52c13/kernel/audit.c (mode:100644) @@ -46,6 +46,8 @@ #include #include #include +#include +#include #include @@ -68,7 +70,7 @@ /* If audit records are to be written to the netlink socket, audit_pid * contains the (non-zero) pid. */ -static int audit_pid; +int audit_pid; /* If audit_limit is non-zero, limit the rate of sending audit records * to that number per second. This prevents DoS attacks, but results in @@ -77,7 +79,10 @@ /* Number of outstanding audit_buffers allowed. */ static int audit_backlog_limit = 64; -static atomic_t audit_backlog = ATOMIC_INIT(0); + +/* The identity of the user shutting down the audit system. */ +uid_t audit_sig_uid = -1; +pid_t audit_sig_pid = -1; /* Records can be lost in several ways: 0) [suppressed in audit_alloc] @@ -91,19 +96,17 @@ /* The netlink socket. */ static struct sock *audit_sock; -/* There are two lists of audit buffers. The txlist contains audit - * buffers that cannot be sent immediately to the netlink device because - * we are in an irq context (these are sent later in a tasklet). - * - * The second list is a list of pre-allocated audit buffers (if more +/* 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). */ -static DEFINE_SPINLOCK(audit_txlist_lock); static DEFINE_SPINLOCK(audit_freelist_lock); static int audit_freelist_count = 0; -static LIST_HEAD(audit_txlist); static LIST_HEAD(audit_freelist); +static struct sk_buff_head audit_skb_queue; +static struct task_struct *kauditd_task; +static DECLARE_WAIT_QUEUE_HEAD(kauditd_wait); + /* 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. */ @@ -112,7 +115,7 @@ static LIST_HEAD(audit_extlist); /* The netlink socket is only to be read by 1 CPU, which lets us assume - * that list additions and deletions never happen simultaneiously in + * that list additions and deletions never happen simultaneously in * auditsc.c */ static DECLARE_MUTEX(audit_netlink_sem); @@ -132,21 +135,14 @@ * use simultaneously. */ struct audit_buffer { struct list_head list; - struct sk_buff_head sklist; /* formatted skbs ready to send */ + struct sk_buff *skb; /* formatted skb ready to send */ struct audit_context *ctx; /* NULL or associated context */ - int len; /* used area of tmp */ - char tmp[AUDIT_BUFSIZ]; - - /* Pointer to header and contents */ - struct nlmsghdr *nlh; - int total; - int type; - int pid; }; -void audit_set_type(struct audit_buffer *ab, int type) +static void audit_set_pid(struct audit_buffer *ab, pid_t pid) { - ab->type = type; + struct nlmsghdr *nlh = (struct nlmsghdr *)ab->skb->data; + nlh->nlmsg_pid = pid; } struct audit_entry { @@ -154,9 +150,6 @@ struct audit_rule rule; }; -static void audit_log_end_irq(struct audit_buffer *ab); -static void audit_log_end_fast(struct audit_buffer *ab); - static void audit_panic(const char *message) { switch (audit_failure) @@ -227,10 +220,8 @@ if (print) { printk(KERN_WARNING - "audit: audit_lost=%d audit_backlog=%d" - " audit_rate_limit=%d audit_backlog_limit=%d\n", + "audit: audit_lost=%d audit_rate_limit=%d audit_backlog_limit=%d\n", atomic_read(&audit_lost), - atomic_read(&audit_backlog), audit_rate_limit, audit_backlog_limit); audit_panic(message); @@ -242,7 +233,8 @@ { int old = audit_rate_limit; audit_rate_limit = limit; - audit_log(NULL, "audit_rate_limit=%d old=%d by auid %u", + audit_log(NULL, AUDIT_CONFIG_CHANGE, + "audit_rate_limit=%d old=%d by auid=%u", audit_rate_limit, old, loginuid); return old; } @@ -251,7 +243,8 @@ { int old = audit_backlog_limit; audit_backlog_limit = limit; - audit_log(NULL, "audit_backlog_limit=%d old=%d by auid %u", + audit_log(NULL, AUDIT_CONFIG_CHANGE, + "audit_backlog_limit=%d old=%d by auid=%u", audit_backlog_limit, old, loginuid); return old; } @@ -262,8 +255,9 @@ if (state != 0 && state != 1) return -EINVAL; audit_enabled = state; - audit_log(NULL, "audit_enabled=%d old=%d by auid %u", - audit_enabled, old, loginuid); + audit_log(NULL, AUDIT_CONFIG_CHANGE, + "audit_enabled=%d old=%d by auid=%u", + audit_enabled, old, loginuid); return old; } @@ -275,12 +269,44 @@ && state != AUDIT_FAIL_PANIC) return -EINVAL; audit_failure = state; - audit_log(NULL, "audit_failure=%d old=%d by auid %u", - audit_failure, old, loginuid); + audit_log(NULL, AUDIT_CONFIG_CHANGE, + "audit_failure=%d old=%d by auid=%u", + audit_failure, old, loginuid); return old; } -#ifdef CONFIG_NET +int kauditd_thread(void *dummy) +{ + struct sk_buff *skb; + + while (1) { + skb = skb_dequeue(&audit_skb_queue); + if (skb) { + if (audit_pid) { + int err = netlink_unicast(audit_sock, skb, audit_pid, 0); + if (err < 0) { + BUG_ON(err != -ECONNREFUSED); /* Shoudn't happen */ + printk(KERN_ERR "audit: *NO* daemon at audit_pid=%d\n", audit_pid); + audit_pid = 0; + } + } else { + printk(KERN_ERR "%s\n", skb->data + NLMSG_SPACE(0)); + kfree_skb(skb); + } + } else { + DECLARE_WAITQUEUE(wait, current); + set_current_state(TASK_INTERRUPTIBLE); + add_wait_queue(&kauditd_wait, &wait); + + if (!skb_queue_len(&audit_skb_queue)) + schedule(); + + __set_current_state(TASK_RUNNING); + remove_wait_queue(&kauditd_wait, &wait); + } + } +} + void audit_send_reply(int pid, int seq, int type, int done, int multi, void *payload, int size) { @@ -293,13 +319,16 @@ skb = alloc_skb(len, GFP_KERNEL); if (!skb) - goto nlmsg_failure; + return; - nlh = NLMSG_PUT(skb, pid, seq, t, len - sizeof(*nlh)); + nlh = NLMSG_PUT(skb, pid, seq, t, size); nlh->nlmsg_flags = flags; data = NLMSG_DATA(nlh); memcpy(data, payload, size); - netlink_unicast(audit_sock, skb, pid, MSG_DONTWAIT); + + /* 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 */ @@ -321,10 +350,12 @@ case AUDIT_SET: case AUDIT_ADD: case AUDIT_DEL: + case AUDIT_SIGNAL_INFO: if (!cap_raised(eff_cap, CAP_AUDIT_CONTROL)) err = -EPERM; break; case AUDIT_USER: + case AUDIT_FIRST_USER_MSG...AUDIT_LAST_USER_MSG: if (!cap_raised(eff_cap, CAP_AUDIT_WRITE)) err = -EPERM; break; @@ -344,11 +375,21 @@ struct audit_buffer *ab; u16 msg_type = nlh->nlmsg_type; uid_t loginuid; /* loginuid of sender */ + struct audit_sig_info sig_data; err = audit_netlink_ok(NETLINK_CB(skb).eff_cap, msg_type); if (err) return err; + /* As soon as there's any sign of userspace auditd, start kauditd to talk to it */ + if (!kauditd_task) + kauditd_task = kthread_run(kauditd_thread, NULL, "kauditd"); + if (IS_ERR(kauditd_task)) { + err = PTR_ERR(kauditd_task); + kauditd_task = NULL; + return err; + } + pid = NETLINK_CREDS(skb)->pid; uid = NETLINK_CREDS(skb)->uid; loginuid = NETLINK_CB(skb).loginuid; @@ -363,7 +404,7 @@ status_set.rate_limit = audit_rate_limit; status_set.backlog_limit = audit_backlog_limit; status_set.lost = atomic_read(&audit_lost); - status_set.backlog = atomic_read(&audit_backlog); + status_set.backlog = skb_queue_len(&audit_skb_queue); audit_send_reply(NETLINK_CB(skb).pid, seq, AUDIT_GET, 0, 0, &status_set, sizeof(status_set)); break; @@ -382,7 +423,8 @@ if (status_get->mask & AUDIT_STATUS_PID) { int old = audit_pid; audit_pid = status_get->pid; - audit_log(NULL, "audit_pid=%d old=%d by auid %u", + audit_log(NULL, AUDIT_CONFIG_CHANGE, + "audit_pid=%d old=%d by auid=%u", audit_pid, old, loginuid); } if (status_get->mask & AUDIT_STATUS_RATE_LIMIT) @@ -392,18 +434,15 @@ loginuid); break; case AUDIT_USER: - ab = audit_log_start(NULL); + case AUDIT_FIRST_USER_MSG...AUDIT_LAST_USER_MSG: + ab = audit_log_start(NULL, msg_type); if (!ab) break; /* audit_panic has been called */ audit_log_format(ab, - "user pid=%d uid=%d length=%d loginuid=%u" + "user pid=%d uid=%u auid=%u" " msg='%.1024s'", - pid, uid, - (int)(nlh->nlmsg_len - - ((char *)data - (char *)nlh)), - loginuid, (char *)data); - ab->type = AUDIT_USER; - ab->pid = pid; + pid, uid, loginuid, (char *)data); + audit_set_pid(ab, pid); audit_log_end(ab); break; case AUDIT_ADD: @@ -412,12 +451,14 @@ return -EINVAL; /* fallthrough */ case AUDIT_LIST: -#ifdef CONFIG_AUDITSYSCALL err = audit_receive_filter(nlh->nlmsg_type, NETLINK_CB(skb).pid, uid, seq, data, loginuid); -#else - err = -EOPNOTSUPP; -#endif + break; + case AUDIT_SIGNAL_INFO: + sig_data.uid = audit_sig_uid; + sig_data.pid = audit_sig_pid; + audit_send_reply(NETLINK_CB(skb).pid, seq, AUDIT_SIGNAL_INFO, + 0, 0, &sig_data, sizeof(sig_data)); break; default: err = -EINVAL; @@ -467,87 +508,6 @@ up(&audit_netlink_sem); } -/* Move data from tmp buffer into an skb. This is an extra copy, and - * that is unfortunate. However, the copy will only occur when a record - * is being written to user space, which is already a high-overhead - * operation. (Elimination of the copy is possible, for example, by - * writing directly into a pre-allocated skb, at the cost of wasting - * memory. */ -static void audit_log_move(struct audit_buffer *ab) -{ - struct sk_buff *skb; - char *start; - int extra = ab->nlh ? 0 : NLMSG_SPACE(0); - - /* possible resubmission */ - if (ab->len == 0) - return; - - skb = skb_peek_tail(&ab->sklist); - if (!skb || skb_tailroom(skb) <= ab->len + extra) { - skb = alloc_skb(2 * ab->len + extra, GFP_ATOMIC); - if (!skb) { - ab->len = 0; /* Lose information in ab->tmp */ - audit_log_lost("out of memory in audit_log_move"); - return; - } - __skb_queue_tail(&ab->sklist, skb); - if (!ab->nlh) - ab->nlh = (struct nlmsghdr *)skb_put(skb, - NLMSG_SPACE(0)); - } - start = skb_put(skb, ab->len); - memcpy(start, ab->tmp, ab->len); - ab->len = 0; -} - -/* Iterate over the skbuff in the audit_buffer, sending their contents - * to user space. */ -static inline int audit_log_drain(struct audit_buffer *ab) -{ - struct sk_buff *skb; - - while ((skb = skb_dequeue(&ab->sklist))) { - int retval = 0; - - if (audit_pid) { - if (ab->nlh) { - ab->nlh->nlmsg_len = ab->total; - ab->nlh->nlmsg_type = ab->type; - ab->nlh->nlmsg_flags = 0; - ab->nlh->nlmsg_seq = 0; - ab->nlh->nlmsg_pid = ab->pid; - } - skb_get(skb); /* because netlink_* frees */ - retval = netlink_unicast(audit_sock, skb, audit_pid, - MSG_DONTWAIT); - } - if (retval == -EAGAIN && - (atomic_read(&audit_backlog)) < audit_backlog_limit) { - skb_queue_head(&ab->sklist, skb); - audit_log_end_irq(ab); - return 1; - } - if (retval < 0) { - if (retval == -ECONNREFUSED) { - printk(KERN_ERR - "audit: *NO* daemon at audit_pid=%d\n", - audit_pid); - audit_pid = 0; - } else - audit_log_lost("netlink socket too busy"); - } - if (!audit_pid) { /* No daemon */ - int offset = ab->nlh ? NLMSG_SPACE(0) : 0; - int len = skb->len - offset; - skb->data[offset + len] = '\0'; - printk(KERN_ERR "%s\n", skb->data + offset); - } - kfree_skb(skb); - ab->nlh = NULL; - } - return 0; -} /* Initialize audit support at boot time. */ static int __init audit_init(void) @@ -558,40 +518,13 @@ if (!audit_sock) audit_panic("cannot initialize netlink socket"); + audit_sock->sk_sndtimeo = MAX_SCHEDULE_TIMEOUT; + skb_queue_head_init(&audit_skb_queue); audit_initialized = 1; audit_enabled = audit_default; - audit_log(NULL, "initialized"); + audit_log(NULL, AUDIT_KERNEL, "initialized"); return 0; } - -#else -/* Without CONFIG_NET, we have no skbuffs. For now, print what we have - * in the buffer. */ -static void audit_log_move(struct audit_buffer *ab) -{ - printk(KERN_ERR "%*.*s\n", ab->len, ab->len, ab->tmp); - ab->len = 0; -} - -static inline int audit_log_drain(struct audit_buffer *ab) -{ - return 0; -} - -/* Initialize audit support at boot time. */ -int __init audit_init(void) -{ - printk(KERN_INFO "audit: initializing WITHOUT netlink support\n"); - audit_sock = NULL; - audit_pid = 0; - - audit_initialized = 1; - audit_enabled = audit_default; - audit_log(NULL, "initialized"); - return 0; -} -#endif - __initcall(audit_init); /* Process kernel command-line parameter at boot time. audit=0 or audit=1. */ @@ -608,6 +541,102 @@ __setup("audit=", audit_enable); +static void audit_buffer_free(struct audit_buffer *ab) +{ + unsigned long flags; + + if (!ab) + return; + + if (ab->skb) + kfree_skb(ab->skb); + + spin_lock_irqsave(&audit_freelist_lock, flags); + if (++audit_freelist_count > AUDIT_MAXFREE) + kfree(ab); + else + list_add(&ab->list, &audit_freelist); + spin_unlock_irqrestore(&audit_freelist_lock, flags); +} + +static struct audit_buffer * audit_buffer_alloc(struct audit_context *ctx, + int gfp_mask, int type) +{ + unsigned long flags; + struct audit_buffer *ab = NULL; + struct nlmsghdr *nlh; + + spin_lock_irqsave(&audit_freelist_lock, flags); + if (!list_empty(&audit_freelist)) { + ab = list_entry(audit_freelist.next, + struct audit_buffer, list); + list_del(&ab->list); + --audit_freelist_count; + } + spin_unlock_irqrestore(&audit_freelist_lock, flags); + + if (!ab) { + ab = kmalloc(sizeof(*ab), gfp_mask); + if (!ab) + goto err; + } + + ab->skb = alloc_skb(AUDIT_BUFSIZ, gfp_mask); + if (!ab->skb) + goto err; + + ab->ctx = ctx; + nlh = (struct nlmsghdr *)skb_put(ab->skb, NLMSG_SPACE(0)); + nlh->nlmsg_type = type; + nlh->nlmsg_flags = 0; + nlh->nlmsg_pid = 0; + nlh->nlmsg_seq = 0; + return ab; +err: + audit_buffer_free(ab); + return NULL; +} + +/* Compute a serial number for the audit record. Audit records are + * written to user-space as soon as they are generated, so a complete + * audit record may be written in several pieces. The timestamp of the + * record and this serial number are used by the user-space tools to + * determine which pieces belong to the same audit record. The + * (timestamp,serial) tuple is unique for each syscall and is live from + * syscall entry to syscall exit. + * + * Atomic values are only guaranteed to be 24-bit, so we count down. + * + * NOTE: Another possibility is to store the formatted records off the + * audit context (for those records that have a context), and emit them + * all at syscall exit. However, this could delay the reporting of + * significant errors until syscall exit (or never, if the system + * halts). */ +unsigned int audit_serial(void) +{ + static atomic_t serial = ATOMIC_INIT(0xffffff); + unsigned int a, b; + + do { + a = atomic_read(&serial); + if (atomic_dec_and_test(&serial)) + atomic_set(&serial, 0xffffff); + b = atomic_read(&serial); + } while (b != a - 1); + + return 0xffffff - b; +} + +static inline void audit_get_stamp(struct audit_context *ctx, + struct timespec *t, unsigned int *serial) +{ + if (ctx) + auditsc_get_stamp(ctx, t, serial); + else { + *t = CURRENT_TIME; + *serial = audit_serial(); + } +} /* Obtain an audit buffer. This routine does locking to obtain the * audit buffer, but then no locking is required for calls to @@ -615,10 +644,9 @@ * syscall, then the syscall is marked as auditable and an audit record * will be written at syscall exit. If there is no associated task, tsk * should be NULL. */ -struct audit_buffer *audit_log_start(struct audit_context *ctx) +struct audit_buffer *audit_log_start(struct audit_context *ctx, int type) { struct audit_buffer *ab = NULL; - unsigned long flags; struct timespec t; unsigned int serial; @@ -626,57 +654,48 @@ return NULL; if (audit_backlog_limit - && atomic_read(&audit_backlog) > audit_backlog_limit) { + && skb_queue_len(&audit_skb_queue) > audit_backlog_limit) { if (audit_rate_check()) printk(KERN_WARNING "audit: audit_backlog=%d > " "audit_backlog_limit=%d\n", - atomic_read(&audit_backlog), + skb_queue_len(&audit_skb_queue), audit_backlog_limit); audit_log_lost("backlog limit exceeded"); return NULL; } - spin_lock_irqsave(&audit_freelist_lock, flags); - if (!list_empty(&audit_freelist)) { - ab = list_entry(audit_freelist.next, - struct audit_buffer, list); - list_del(&ab->list); - --audit_freelist_count; - } - spin_unlock_irqrestore(&audit_freelist_lock, flags); - - if (!ab) - ab = kmalloc(sizeof(*ab), GFP_ATOMIC); + ab = audit_buffer_alloc(ctx, GFP_ATOMIC, type); if (!ab) { audit_log_lost("out of memory in audit_log_start"); return NULL; } - atomic_inc(&audit_backlog); - skb_queue_head_init(&ab->sklist); + audit_get_stamp(ab->ctx, &t, &serial); - ab->ctx = ctx; - ab->len = 0; - ab->nlh = NULL; - ab->total = 0; - ab->type = AUDIT_KERNEL; - ab->pid = 0; - -#ifdef CONFIG_AUDITSYSCALL - if (ab->ctx) - audit_get_stamp(ab->ctx, &t, &serial); - else -#endif - { - t = CURRENT_TIME; - serial = 0; - } audit_log_format(ab, "audit(%lu.%03lu:%u): ", t.tv_sec, t.tv_nsec/1000000, serial); return ab; } +/** + * audit_expand - expand skb in the audit buffer + * @ab: audit_buffer + * + * Returns 0 (no space) on failed expansion, or available space if + * successful. + */ +static inline int audit_expand(struct audit_buffer *ab, int extra) +{ + struct sk_buff *skb = ab->skb; + int ret = pskb_expand_head(skb, skb_headroom(skb), extra, + GFP_ATOMIC); + if (ret < 0) { + audit_log_lost("out of memory in audit_expand"); + return 0; + } + return skb_tailroom(skb); +} /* Format an audit message into the audit buffer. If there isn't enough * room in the audit buffer, more room will be allocated and vsnprint @@ -686,26 +705,35 @@ va_list args) { int len, avail; + struct sk_buff *skb; + va_list args2; if (!ab) return; - avail = sizeof(ab->tmp) - ab->len; - if (avail <= 0) { - audit_log_move(ab); - avail = sizeof(ab->tmp) - ab->len; + BUG_ON(!ab->skb); + skb = ab->skb; + avail = skb_tailroom(skb); + if (avail == 0) { + avail = audit_expand(ab, AUDIT_BUFSIZ); + if (!avail) + goto out; } - len = vsnprintf(ab->tmp + ab->len, avail, fmt, args); + va_copy(args2, args); + len = vsnprintf(skb->tail, avail, fmt, args); if (len >= avail) { /* The printk buffer is 1024 bytes long, so if we get * here and AUDIT_BUFSIZ is at least 1024, then we can * log everything that printk could have logged. */ - audit_log_move(ab); - avail = sizeof(ab->tmp) - ab->len; - len = vsnprintf(ab->tmp + ab->len, avail, fmt, args); - } - ab->len += (len < avail) ? len : avail; - ab->total += (len < avail) ? len : avail; + avail = audit_expand(ab, max_t(unsigned, AUDIT_BUFSIZ, 1+len-avail)); + if (!avail) + goto out; + len = vsnprintf(skb->tail, avail, fmt, args2); + } + if (len > 0) + skb_put(skb, len); +out: + return; } /* Format a message into the audit buffer. All the work is done in @@ -721,20 +749,47 @@ va_end(args); } -void audit_log_hex(struct audit_buffer *ab, const unsigned char *buf, size_t len) +/* This function will take the passed buf and convert it into a string of + * ascii hex digits. The new string is placed onto the skb. */ +void audit_log_hex(struct audit_buffer *ab, const unsigned char *buf, + size_t len) { - int i; + int i, avail, new_len; + unsigned char *ptr; + struct sk_buff *skb; + static const unsigned char *hex = "0123456789ABCDEF"; + + BUG_ON(!ab->skb); + skb = ab->skb; + avail = skb_tailroom(skb); + new_len = len<<1; + if (new_len >= avail) { + /* Round the buffer request up to the next multiple */ + new_len = AUDIT_BUFSIZ*(((new_len-avail)/AUDIT_BUFSIZ) + 1); + avail = audit_expand(ab, new_len); + if (!avail) + return; + } - for (i=0; itail; + for (i=0; i>4]; /* Upper nibble */ + *ptr++ = hex[buf[i] & 0x0F]; /* Lower nibble */ + } + *ptr = 0; + skb_put(skb, len << 1); /* new string is twice the old string */ } +/* 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). */ void audit_log_untrustedstring(struct audit_buffer *ab, const char *string) { const unsigned char *p = string; while (*p) { - if (*p == '"' || *p == ' ' || *p < 0x20 || *p > 0x7f) { + if (*p == '"' || *p < 0x21 || *p > 0x7f) { audit_log_hex(ab, string, strlen(string)); return; } @@ -743,117 +798,63 @@ audit_log_format(ab, "\"%s\"", string); } - -/* This is a helper-function to print the d_path without using a static - * buffer or allocating another buffer in addition to the one in - * audit_buffer. */ +/* This is a helper-function to print the escaped d_path */ void audit_log_d_path(struct audit_buffer *ab, const char *prefix, struct dentry *dentry, struct vfsmount *vfsmnt) { - char *p; - int len, avail; - - if (prefix) audit_log_format(ab, " %s", prefix); - - if (ab->len > 128) - audit_log_move(ab); - avail = sizeof(ab->tmp) - ab->len; - p = d_path(dentry, vfsmnt, ab->tmp + ab->len, avail); - if (IS_ERR(p)) { - /* FIXME: can we save some information here? */ - audit_log_format(ab, ""); - } else { - /* path isn't at start of buffer */ - len = (ab->tmp + sizeof(ab->tmp) - 1) - p; - memmove(ab->tmp + ab->len, p, len); - ab->len += len; - ab->total += len; - } -} - -/* Remove queued messages from the audit_txlist and send them to userspace. */ -static void audit_tasklet_handler(unsigned long arg) -{ - LIST_HEAD(list); - struct audit_buffer *ab; - unsigned long flags; + char *p, *path; - spin_lock_irqsave(&audit_txlist_lock, flags); - list_splice_init(&audit_txlist, &list); - spin_unlock_irqrestore(&audit_txlist_lock, flags); + if (prefix) + audit_log_format(ab, " %s", prefix); - while (!list_empty(&list)) { - ab = list_entry(list.next, struct audit_buffer, list); - list_del(&ab->list); - audit_log_end_fast(ab); + /* We will allow 11 spaces for ' (deleted)' to be appended */ + path = kmalloc(PATH_MAX+11, GFP_KERNEL); + if (!path) { + audit_log_format(ab, ""); + return; } + p = d_path(dentry, vfsmnt, path, PATH_MAX+11); + if (IS_ERR(p)) { /* Should never happen since we send PATH_MAX */ + /* FIXME: can we save some information here? */ + audit_log_format(ab, ""); + } else + audit_log_untrustedstring(ab, p); + kfree(path); } -static DECLARE_TASKLET(audit_tasklet, audit_tasklet_handler, 0); - /* The netlink_* functions cannot be called inside an irq context, so * the audit buffer is places on a queue and a tasklet is scheduled to * remove them from the queue outside the irq context. May be called in * any context. */ -static void audit_log_end_irq(struct audit_buffer *ab) -{ - unsigned long flags; - - if (!ab) - return; - spin_lock_irqsave(&audit_txlist_lock, flags); - list_add_tail(&ab->list, &audit_txlist); - spin_unlock_irqrestore(&audit_txlist_lock, flags); - - tasklet_schedule(&audit_tasklet); -} - -/* Send the message in the audit buffer directly to user space. May not - * be called in an irq context. */ -static void audit_log_end_fast(struct audit_buffer *ab) +void audit_log_end(struct audit_buffer *ab) { - unsigned long flags; - - BUG_ON(in_irq()); if (!ab) return; if (!audit_rate_check()) { audit_log_lost("rate limit exceeded"); } else { - audit_log_move(ab); - if (audit_log_drain(ab)) - return; + if (audit_pid) { + struct nlmsghdr *nlh = (struct nlmsghdr *)ab->skb->data; + nlh->nlmsg_len = ab->skb->len - NLMSG_SPACE(0); + skb_queue_tail(&audit_skb_queue, ab->skb); + ab->skb = NULL; + wake_up_interruptible(&kauditd_wait); + } else { + printk("%s\n", ab->skb->data + NLMSG_SPACE(0)); + } } - - atomic_dec(&audit_backlog); - spin_lock_irqsave(&audit_freelist_lock, flags); - if (++audit_freelist_count > AUDIT_MAXFREE) - kfree(ab); - else - list_add(&ab->list, &audit_freelist); - spin_unlock_irqrestore(&audit_freelist_lock, flags); -} - -/* Send or queue the message in the audit buffer, depending on the - * current context. (A convenience function that may be called in any - * context.) */ -void audit_log_end(struct audit_buffer *ab) -{ - if (in_irq()) - audit_log_end_irq(ab); - else - audit_log_end_fast(ab); + audit_buffer_free(ab); } /* Log an audit record. This is a convenience function that calls * audit_log_start, audit_log_vformat, and audit_log_end. It may be * called in any context. */ -void audit_log(struct audit_context *ctx, const char *fmt, ...) +void audit_log(struct audit_context *ctx, int type, const char *fmt, ...) { struct audit_buffer *ab; va_list args; - ab = audit_log_start(ctx); + ab = audit_log_start(ctx, type); if (ab) { va_start(args, fmt); audit_log_vformat(ab, fmt, args); Index: kernel/auditsc.c =================================================================== --- 3c5e9440c6a37c3355b50608836a23c8fa4eec99/kernel/auditsc.c (mode:100644) +++ 672465b3b9b3e2e26a8caf74ed64aa6885c52c13/kernel/auditsc.c (mode:100644) @@ -34,7 +34,8 @@ #include #include #include - +#include +#include #include #include #include @@ -112,6 +113,23 @@ mode_t mode; }; +struct audit_aux_data_socketcall { + struct audit_aux_data d; + int nargs; + unsigned long args[0]; +}; + +struct audit_aux_data_sockaddr { + struct audit_aux_data d; + int len; + char a[0]; +}; + +struct audit_aux_data_path { + struct audit_aux_data d; + struct dentry *dentry; + struct vfsmount *mnt; +}; /* The per-task audit context. */ struct audit_context { @@ -127,6 +145,8 @@ int auditable; /* 1 if record should be written */ int name_count; struct audit_names names[AUDIT_NAMES]; + struct dentry * pwd; + struct vfsmount * pwdmnt; struct audit_context *previous; /* For nested syscalls */ struct audit_aux_data *aux; @@ -157,6 +177,8 @@ struct audit_rule rule; }; +extern int audit_pid; + /* Check to see if two rules are identical. It is called from * audit_del_rule during AUDIT_DEL. */ static int audit_compare_rule(struct audit_rule *a, struct audit_rule *b) @@ -226,7 +248,6 @@ return -EFAULT; /* No matching rule */ } -#ifdef CONFIG_NET /* Copy rule from user-space to kernel-space. Called during * AUDIT_ADD. */ static int audit_copy_rule(struct audit_rule *d, struct audit_rule *s) @@ -287,7 +308,8 @@ err = audit_add_rule(entry, &audit_entlist); if (!err && (flags & AUDIT_AT_EXIT)) err = audit_add_rule(entry, &audit_extlist); - audit_log(NULL, "auid %u added an audit rule\n", loginuid); + audit_log(NULL, AUDIT_CONFIG_CHANGE, + "auid=%u added an audit rule\n", loginuid); break; case AUDIT_DEL: flags =((struct audit_rule *)data)->flags; @@ -297,7 +319,8 @@ err = audit_del_rule(data, &audit_entlist); if (!err && (flags & AUDIT_AT_EXIT)) err = audit_del_rule(data, &audit_extlist); - audit_log(NULL, "auid %u removed an audit rule\n", loginuid); + audit_log(NULL, AUDIT_CONFIG_CHANGE, + "auid=%u removed an audit rule\n", loginuid); break; default: return -EINVAL; @@ -305,7 +328,6 @@ return err; } -#endif /* Compare a task_struct with an audit_rule. Return 1 on match, 0 * otherwise. */ @@ -444,7 +466,7 @@ /* At syscall entry and exit time, this filter is called if the * audit_state is not low enough that auditing cannot take place, but is - * also not high enough that we already know we have to write and audit + * also not high enough that we already know we have to write an audit * record (i.e., the state is AUDIT_SETUP_CONTEXT or AUDIT_BUILD_CONTEXT). */ static enum audit_state audit_filter_syscall(struct task_struct *tsk, @@ -532,6 +554,12 @@ if (context->names[i].name) __putname(context->names[i].name); context->name_count = 0; + if (context->pwd) + dput(context->pwd); + if (context->pwdmnt) + mntput(context->pwdmnt); + context->pwd = NULL; + context->pwdmnt = NULL; } static inline void audit_free_aux(struct audit_context *context) @@ -539,6 +567,11 @@ struct audit_aux_data *aux; while ((aux = context->aux)) { + if (aux->type == AUDIT_AVC_PATH) { + struct audit_aux_data_path *axi = (void *)aux; + dput(axi->dentry); + mntput(axi->mnt); + } context->aux = aux->next; kfree(aux); } @@ -625,7 +658,8 @@ struct vm_area_struct *vma; get_task_comm(name, current); - audit_log_format(ab, " comm=%s", name); + audit_log_format(ab, " comm="); + audit_log_untrustedstring(ab, name); if (!mm) return; @@ -649,23 +683,24 @@ { int i; struct audit_buffer *ab; + struct audit_aux_data *aux; - ab = audit_log_start(context); + ab = audit_log_start(context, AUDIT_SYSCALL); if (!ab) return; /* audit_panic has been called */ - audit_log_format(ab, "syscall=%d", context->major); + audit_log_format(ab, "arch=%x syscall=%d", + context->arch, context->major); if (context->personality != PER_LINUX) audit_log_format(ab, " per=%lx", context->personality); - audit_log_format(ab, " arch=%x", context->arch); if (context->return_valid) audit_log_format(ab, " success=%s exit=%ld", (context->return_valid==AUDITSC_SUCCESS)?"yes":"no", context->return_code); audit_log_format(ab, " a0=%lx a1=%lx a2=%lx a3=%lx items=%d" - " pid=%d loginuid=%d uid=%d gid=%d" - " euid=%d suid=%d fsuid=%d" - " egid=%d sgid=%d fsgid=%d", + " pid=%d auid=%u uid=%u gid=%u" + " euid=%u suid=%u fsuid=%u" + " egid=%u sgid=%u fsgid=%u", context->argv[0], context->argv[1], context->argv[2], @@ -679,33 +714,57 @@ context->egid, context->sgid, context->fsgid); audit_log_task_info(ab); audit_log_end(ab); - while (context->aux) { - struct audit_aux_data *aux; - ab = audit_log_start(context); + for (aux = context->aux; aux; aux = aux->next) { + + ab = audit_log_start(context, aux->type); if (!ab) continue; /* audit_panic has been called */ - aux = context->aux; - context->aux = aux->next; - - audit_log_format(ab, "auxitem=%d", aux->type); switch (aux->type) { - case AUDIT_AUX_IPCPERM: { + case AUDIT_IPC: { struct audit_aux_data_ipcctl *axi = (void *)aux; audit_log_format(ab, - " qbytes=%lx uid=%d gid=%d mode=%x", + " qbytes=%lx iuid=%u igid=%u mode=%x", axi->qbytes, axi->uid, axi->gid, axi->mode); - } + break; } + + case AUDIT_SOCKETCALL: { + int i; + struct audit_aux_data_socketcall *axs = (void *)aux; + audit_log_format(ab, "nargs=%d", axs->nargs); + for (i=0; inargs; i++) + audit_log_format(ab, " a%d=%lx", i, axs->args[i]); + break; } + + case AUDIT_SOCKADDR: { + struct audit_aux_data_sockaddr *axs = (void *)aux; + + audit_log_format(ab, "saddr="); + audit_log_hex(ab, axs->a, axs->len); + break; } + + case AUDIT_AVC_PATH: { + struct audit_aux_data_path *axi = (void *)aux; + audit_log_d_path(ab, "path=", axi->dentry, axi->mnt); + break; } + } audit_log_end(ab); - kfree(aux); } + if (context->pwd && context->pwdmnt) { + ab = audit_log_start(context, AUDIT_CWD); + if (ab) { + audit_log_d_path(ab, "cwd=", context->pwd, context->pwdmnt); + audit_log_end(ab); + } + } for (i = 0; i < context->name_count; i++) { - ab = audit_log_start(context); + ab = audit_log_start(context, AUDIT_PATH); if (!ab) continue; /* audit_panic has been called */ + audit_log_format(ab, "item=%d", i); if (context->names[i].name) { audit_log_format(ab, " name="); @@ -713,7 +772,7 @@ } if (context->names[i].ino != (unsigned long)-1) audit_log_format(ab, " inode=%lu dev=%02x:%02x mode=%#o" - " uid=%d gid=%d rdev=%02x:%02x", + " ouid=%u ogid=%u rdev=%02x:%02x", context->names[i].ino, MAJOR(context->names[i].dev), MINOR(context->names[i].dev), @@ -741,42 +800,12 @@ /* Check for system calls that do not go through the exit * function (e.g., exit_group), then free context block. */ - if (context->in_syscall && context->auditable) + if (context->in_syscall && context->auditable && context->pid != audit_pid) audit_log_exit(context); audit_free_context(context); } -/* Compute a serial number for the audit record. Audit records are - * written to user-space as soon as they are generated, so a complete - * audit record may be written in several pieces. The timestamp of the - * record and this serial number are used by the user-space daemon to - * determine which pieces belong to the same audit record. The - * (timestamp,serial) tuple is unique for each syscall and is live from - * syscall entry to syscall exit. - * - * Atomic values are only guaranteed to be 24-bit, so we count down. - * - * NOTE: Another possibility is to store the formatted records off the - * audit context (for those records that have a context), and emit them - * all at syscall exit. However, this could delay the reporting of - * significant errors until syscall exit (or never, if the system - * halts). */ -static inline unsigned int audit_serial(void) -{ - static atomic_t serial = ATOMIC_INIT(0xffffff); - unsigned int a, b; - - do { - a = atomic_read(&serial); - if (atomic_dec_and_test(&serial)) - atomic_set(&serial, 0xffffff); - b = atomic_read(&serial); - } while (b != a - 1); - - return 0xffffff - b; -} - /* Fill in audit context at syscall entry. This only happens if the * audit context was created when the task was created and the state or * filters demand the audit context be built. If the state from the @@ -876,7 +905,7 @@ if (likely(!context)) return; - if (context->in_syscall && context->auditable) + if (context->in_syscall && context->auditable && context->pid != audit_pid) audit_log_exit(context); context->in_syscall = 0; @@ -916,6 +945,13 @@ context->names[context->name_count].name = name; context->names[context->name_count].ino = (unsigned long)-1; ++context->name_count; + if (!context->pwd) { + read_lock(¤t->fs->lock); + context->pwd = dget(current->fs->pwd); + context->pwdmnt = mntget(current->fs->pwdmnt); + read_unlock(¤t->fs->lock); + } + } /* Intercept a putname request. Called from @@ -994,34 +1030,26 @@ context->names[idx].rdev = inode->i_rdev; } -void audit_get_stamp(struct audit_context *ctx, - struct timespec *t, unsigned int *serial) +void auditsc_get_stamp(struct audit_context *ctx, + struct timespec *t, unsigned int *serial) { - if (ctx) { - t->tv_sec = ctx->ctime.tv_sec; - t->tv_nsec = ctx->ctime.tv_nsec; - *serial = ctx->serial; - ctx->auditable = 1; - } else { - *t = CURRENT_TIME; - *serial = 0; - } + t->tv_sec = ctx->ctime.tv_sec; + t->tv_nsec = ctx->ctime.tv_nsec; + *serial = ctx->serial; + ctx->auditable = 1; } -extern int audit_set_type(struct audit_buffer *ab, int type); - int audit_set_loginuid(struct task_struct *task, uid_t loginuid) { if (task->audit_context) { struct audit_buffer *ab; - ab = audit_log_start(NULL); + ab = audit_log_start(NULL, AUDIT_LOGIN); if (ab) { audit_log_format(ab, "login pid=%d uid=%u " - "old loginuid=%u new loginuid=%u", + "old auid=%u new auid=%u", task->pid, task->uid, task->audit_context->loginuid, loginuid); - audit_set_type(ab, AUDIT_LOGIN); audit_log_end(ab); } task->audit_context->loginuid = loginuid; @@ -1051,8 +1079,89 @@ ax->gid = gid; ax->mode = mode; - ax->d.type = AUDIT_AUX_IPCPERM; + ax->d.type = AUDIT_IPC; ax->d.next = context->aux; context->aux = (void *)ax; return 0; } + +int audit_socketcall(int nargs, unsigned long *args) +{ + struct audit_aux_data_socketcall *ax; + struct audit_context *context = current->audit_context; + + if (likely(!context)) + return 0; + + ax = kmalloc(sizeof(*ax) + nargs * sizeof(unsigned long), GFP_KERNEL); + if (!ax) + return -ENOMEM; + + ax->nargs = nargs; + memcpy(ax->args, args, nargs * sizeof(unsigned long)); + + ax->d.type = AUDIT_SOCKETCALL; + ax->d.next = context->aux; + context->aux = (void *)ax; + return 0; +} + +int audit_sockaddr(int len, void *a) +{ + struct audit_aux_data_sockaddr *ax; + struct audit_context *context = current->audit_context; + + if (likely(!context)) + return 0; + + ax = kmalloc(sizeof(*ax) + len, GFP_KERNEL); + if (!ax) + return -ENOMEM; + + ax->len = len; + memcpy(ax->a, a, len); + + ax->d.type = AUDIT_SOCKADDR; + ax->d.next = context->aux; + context->aux = (void *)ax; + return 0; +} + +int audit_avc_path(struct dentry *dentry, struct vfsmount *mnt) +{ + struct audit_aux_data_path *ax; + struct audit_context *context = current->audit_context; + + if (likely(!context)) + return 0; + + ax = kmalloc(sizeof(*ax), GFP_ATOMIC); + if (!ax) + return -ENOMEM; + + ax->dentry = dget(dentry); + ax->mnt = mntget(mnt); + + ax->d.type = AUDIT_AVC_PATH; + ax->d.next = context->aux; + context->aux = (void *)ax; + return 0; +} + +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->pid == 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; + } + } +} + Index: kernel/signal.c =================================================================== --- 3c5e9440c6a37c3355b50608836a23c8fa4eec99/kernel/signal.c (mode:100644) +++ 672465b3b9b3e2e26a8caf74ed64aa6885c52c13/kernel/signal.c (mode:100644) @@ -24,6 +24,7 @@ #include #include #include +#include #include #include #include @@ -667,7 +668,11 @@ && (current->uid ^ t->suid) && (current->uid ^ t->uid) && !capable(CAP_KILL)) return error; - return security_task_kill(t, info, sig); + + error = security_task_kill(t, info, sig); + if (!error) + audit_signal_info(sig, t); /* Let audit system see the signal */ + return error; } /* forward decl */ Index: net/socket.c =================================================================== --- 3c5e9440c6a37c3355b50608836a23c8fa4eec99/net/socket.c (mode:100644) +++ 672465b3b9b3e2e26a8caf74ed64aa6885c52c13/net/socket.c (mode:100644) @@ -81,6 +81,7 @@ #include #include #include +#include #ifdef CONFIG_NET_RADIO #include /* Note : will define WIRELESS_EXT */ @@ -226,7 +227,7 @@ return 0; if(copy_from_user(kaddr,uaddr,ulen)) return -EFAULT; - return 0; + return audit_sockaddr(ulen, kaddr); } /** @@ -1906,7 +1907,11 @@ /* copy_from_user should be SMP safe. */ if (copy_from_user(a, args, nargs[call])) return -EFAULT; - + + err = audit_socketcall(nargs[call]/sizeof(unsigned long), a); + if (err) + return err; + a0=a[0]; a1=a[1]; Index: security/selinux/avc.c =================================================================== --- 3c5e9440c6a37c3355b50608836a23c8fa4eec99/security/selinux/avc.c (mode:100644) +++ 672465b3b9b3e2e26a8caf74ed64aa6885c52c13/security/selinux/avc.c (mode:100644) @@ -242,7 +242,7 @@ avc_node_cachep = kmem_cache_create("avc_node", sizeof(struct avc_node), 0, SLAB_PANIC, NULL, NULL); - audit_log(current->audit_context, "AVC INITIALIZED\n"); + audit_log(current->audit_context, AUDIT_KERNEL, "AVC INITIALIZED\n"); } int avc_get_hash_stats(char *page) @@ -532,6 +532,7 @@ u16 tclass, u32 requested, struct av_decision *avd, int result, struct avc_audit_data *a) { + struct task_struct *tsk = current; struct inode *inode = NULL; u32 denied, audited; struct audit_buffer *ab; @@ -549,12 +550,18 @@ return; } - ab = audit_log_start(current->audit_context); + ab = audit_log_start(current->audit_context, AUDIT_AVC); if (!ab) return; /* audit_panic has been called */ audit_log_format(ab, "avc: %s ", denied ? "denied" : "granted"); avc_dump_av(ab, tclass,audited); audit_log_format(ab, " for "); + if (a && a->tsk) + tsk = a->tsk; + if (tsk && tsk->pid) { + audit_log_format(ab, " pid=%d comm=", tsk->pid); + audit_log_untrustedstring(ab, tsk->comm); + } if (a) { switch (a->type) { case AVC_AUDIT_DATA_IPC: @@ -566,21 +573,18 @@ case AVC_AUDIT_DATA_FS: if (a->u.fs.dentry) { struct dentry *dentry = a->u.fs.dentry; - if (a->u.fs.mnt) { - audit_log_d_path(ab, "path=", dentry, - a->u.fs.mnt); - } else { - audit_log_format(ab, " name=%s", - dentry->d_name.name); - } + if (a->u.fs.mnt) + audit_avc_path(dentry, a->u.fs.mnt); + audit_log_format(ab, " name="); + audit_log_untrustedstring(ab, dentry->d_name.name); inode = dentry->d_inode; } else if (a->u.fs.inode) { struct dentry *dentry; inode = a->u.fs.inode; dentry = d_find_alias(inode); if (dentry) { - audit_log_format(ab, " name=%s", - dentry->d_name.name); + audit_log_format(ab, " name="); + audit_log_untrustedstring(ab, dentry->d_name.name); dput(dentry); } } @@ -623,22 +627,20 @@ case AF_UNIX: u = unix_sk(sk); if (u->dentry) { - audit_log_d_path(ab, "path=", - u->dentry, u->mnt); + audit_avc_path(u->dentry, u->mnt); + audit_log_format(ab, " name="); + audit_log_untrustedstring(ab, u->dentry->d_name.name); break; } if (!u->addr) break; len = u->addr->len-sizeof(short); p = &u->addr->name->sun_path[0]; + audit_log_format(ab, " path="); if (*p) - audit_log_format(ab, - "path=%*.*s", len, - len, p); + audit_log_untrustedstring(ab, p); else - audit_log_format(ab, - "path=@%*.*s", len-1, - len-1, p+1); + audit_log_hex(ab, p, len); break; } } Index: security/selinux/hooks.c =================================================================== --- 3c5e9440c6a37c3355b50608836a23c8fa4eec99/security/selinux/hooks.c (mode:100644) +++ 672465b3b9b3e2e26a8caf74ed64aa6885c52c13/security/selinux/hooks.c (mode:100644) @@ -3419,7 +3419,7 @@ err = selinux_nlmsg_lookup(isec->sclass, nlh->nlmsg_type, &perm); if (err) { if (err == -EINVAL) { - audit_log(current->audit_context, + audit_log(current->audit_context, AUDIT_SELINUX_ERR, "SELinux: unrecognized netlink message" " type=%hu for sclass=%hu\n", nlh->nlmsg_type, isec->sclass); Index: security/selinux/nlmsgtab.c =================================================================== --- 3c5e9440c6a37c3355b50608836a23c8fa4eec99/security/selinux/nlmsgtab.c (mode:100644) +++ 672465b3b9b3e2e26a8caf74ed64aa6885c52c13/security/selinux/nlmsgtab.c (mode:100644) @@ -97,6 +97,7 @@ { AUDIT_ADD, NETLINK_AUDIT_SOCKET__NLMSG_WRITE }, { AUDIT_DEL, NETLINK_AUDIT_SOCKET__NLMSG_WRITE }, { AUDIT_USER, NETLINK_AUDIT_SOCKET__NLMSG_RELAY }, + { AUDIT_SIGNAL_INFO, NETLINK_AUDIT_SOCKET__NLMSG_READ }, }; @@ -141,8 +142,13 @@ break; case SECCLASS_NETLINK_AUDIT_SOCKET: - err = nlmsg_perm(nlmsg_type, perm, nlmsg_audit_perms, - sizeof(nlmsg_audit_perms)); + if (nlmsg_type >= AUDIT_FIRST_USER_MSG && + nlmsg_type <= AUDIT_LAST_USER_MSG) { + *perm = NETLINK_AUDIT_SOCKET__NLMSG_RELAY; + } else { + err = nlmsg_perm(nlmsg_type, perm, nlmsg_audit_perms, + sizeof(nlmsg_audit_perms)); + } break; /* No messaging from userspace, or class unknown/unhandled */ Index: security/selinux/ss/services.c =================================================================== --- 3c5e9440c6a37c3355b50608836a23c8fa4eec99/security/selinux/ss/services.c (mode:100644) +++ 672465b3b9b3e2e26a8caf74ed64aa6885c52c13/security/selinux/ss/services.c (mode:100644) @@ -365,7 +365,7 @@ goto out; if (context_struct_to_string(tcontext, &t, &tlen) < 0) goto out; - audit_log(current->audit_context, + audit_log(current->audit_context, AUDIT_SELINUX_ERR, "security_validate_transition: denied for" " oldcontext=%s newcontext=%s taskcontext=%s tclass=%s", o, n, t, policydb.p_class_val_to_name[tclass-1]); @@ -742,7 +742,7 @@ goto out; if (context_struct_to_string(newcontext, &n, &nlen) < 0) goto out; - audit_log(current->audit_context, + audit_log(current->audit_context, AUDIT_SELINUX_ERR, "security_compute_sid: invalid context %s" " for scontext=%s" " tcontext=%s"