diff options
Diffstat (limited to 'kernel')
| -rw-r--r-- | kernel/cgroup/cgroup-internal.h | 26 | ||||
| -rw-r--r-- | kernel/cgroup/cgroup-v1.c | 4 | ||||
| -rw-r--r-- | kernel/cgroup/cgroup.c | 12 | ||||
| -rw-r--r-- | kernel/events/callchain.c | 2 | ||||
| -rw-r--r-- | kernel/memremap.c | 2 | ||||
| -rw-r--r-- | kernel/printk/printk.c | 3 | ||||
| -rw-r--r-- | kernel/sys.c | 95 | ||||
| -rw-r--r-- | kernel/sysctl.c | 18 | ||||
| -rw-r--r-- | kernel/trace/Kconfig | 12 | ||||
| -rw-r--r-- | kernel/trace/Makefile | 5 | ||||
| -rw-r--r-- | kernel/trace/trace_kprobe.c | 9 | ||||
| -rw-r--r-- | kernel/user_namespace.c | 24 | ||||
| -rw-r--r-- | kernel/utsname_sysctl.c | 41 | ||||
| -rw-r--r-- | kernel/workqueue.c | 45 |
14 files changed, 193 insertions, 105 deletions
diff --git a/kernel/cgroup/cgroup-internal.h b/kernel/cgroup/cgroup-internal.h index 77ff1cd6a252..75568fcf2180 100644 --- a/kernel/cgroup/cgroup-internal.h +++ b/kernel/cgroup/cgroup-internal.h @@ -8,6 +8,32 @@ #include <linux/list.h> #include <linux/refcount.h> +#define TRACE_CGROUP_PATH_LEN 1024 +extern spinlock_t trace_cgroup_path_lock; +extern char trace_cgroup_path[TRACE_CGROUP_PATH_LEN]; + +/* + * cgroup_path() takes a spin lock. It is good practice not to take + * spin locks within trace point handlers, as they are mostly hidden + * from normal view. As cgroup_path() can take the kernfs_rename_lock + * spin lock, it is best to not call that function from the trace event + * handler. + * + * Note: trace_cgroup_##type##_enabled() is a static branch that will only + * be set when the trace event is enabled. + */ +#define TRACE_CGROUP_PATH(type, cgrp, ...) \ + do { \ + if (trace_cgroup_##type##_enabled()) { \ + spin_lock(&trace_cgroup_path_lock); \ + cgroup_path(cgrp, trace_cgroup_path, \ + TRACE_CGROUP_PATH_LEN); \ + trace_cgroup_##type(cgrp, trace_cgroup_path, \ + ##__VA_ARGS__); \ + spin_unlock(&trace_cgroup_path_lock); \ + } \ + } while (0) + /* * A cgroup can be associated with multiple css_sets as different tasks may * belong to different cgroups on different hierarchies. In the other diff --git a/kernel/cgroup/cgroup-v1.c b/kernel/cgroup/cgroup-v1.c index 8b4f0768efd6..51063e7a93c2 100644 --- a/kernel/cgroup/cgroup-v1.c +++ b/kernel/cgroup/cgroup-v1.c @@ -135,7 +135,7 @@ int cgroup_transfer_tasks(struct cgroup *to, struct cgroup *from) if (task) { ret = cgroup_migrate(task, false, &mgctx); if (!ret) - trace_cgroup_transfer_tasks(to, task, false); + TRACE_CGROUP_PATH(transfer_tasks, to, task, false); put_task_struct(task); } } while (task && !ret); @@ -865,7 +865,7 @@ static int cgroup1_rename(struct kernfs_node *kn, struct kernfs_node *new_parent ret = kernfs_rename(kn, new_parent, new_name_str); if (!ret) - trace_cgroup_rename(cgrp); + TRACE_CGROUP_PATH(rename, cgrp); mutex_unlock(&cgroup_mutex); diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c index 35cf3d71f8aa..aae10baf1902 100644 --- a/kernel/cgroup/cgroup.c +++ b/kernel/cgroup/cgroup.c @@ -83,6 +83,9 @@ EXPORT_SYMBOL_GPL(cgroup_mutex); EXPORT_SYMBOL_GPL(css_set_lock); #endif +DEFINE_SPINLOCK(trace_cgroup_path_lock); +char trace_cgroup_path[TRACE_CGROUP_PATH_LEN]; + /* * Protects cgroup_idr and css_idr so that IDs can be released without * grabbing cgroup_mutex. @@ -2638,7 +2641,7 @@ int cgroup_attach_task(struct cgroup *dst_cgrp, struct task_struct *leader, cgroup_migrate_finish(&mgctx); if (!ret) - trace_cgroup_attach_task(dst_cgrp, leader, threadgroup); + TRACE_CGROUP_PATH(attach_task, dst_cgrp, leader, threadgroup); return ret; } @@ -4636,7 +4639,7 @@ static void css_release_work_fn(struct work_struct *work) struct cgroup *tcgrp; /* cgroup release path */ - trace_cgroup_release(cgrp); + TRACE_CGROUP_PATH(release, cgrp); if (cgroup_on_dfl(cgrp)) cgroup_rstat_flush(cgrp); @@ -4979,7 +4982,7 @@ int cgroup_mkdir(struct kernfs_node *parent_kn, const char *name, umode_t mode) if (ret) goto out_destroy; - trace_cgroup_mkdir(cgrp); + TRACE_CGROUP_PATH(mkdir, cgrp); /* let's create and online css's */ kernfs_activate(kn); @@ -5167,9 +5170,8 @@ int cgroup_rmdir(struct kernfs_node *kn) return 0; ret = cgroup_destroy_locked(cgrp); - if (!ret) - trace_cgroup_rmdir(cgrp); + TRACE_CGROUP_PATH(rmdir, cgrp); cgroup_kn_unlock(kn); return ret; diff --git a/kernel/events/callchain.c b/kernel/events/callchain.c index c187aa3df3c8..24a77c34e9ad 100644 --- a/kernel/events/callchain.c +++ b/kernel/events/callchain.c @@ -4,7 +4,7 @@ * Copyright (C) 2008 Thomas Gleixner <tglx@linutronix.de> * Copyright (C) 2008-2011 Red Hat, Inc., Ingo Molnar * Copyright (C) 2008-2011 Red Hat, Inc., Peter Zijlstra - * Copyright © 2009 Paul Mackerras, IBM Corp. <paulus@au1.ibm.com> + * Copyright © 2009 Paul Mackerras, IBM Corp. <paulus@au1.ibm.com> * * For licensing details see kernel-base/COPYING */ diff --git a/kernel/memremap.c b/kernel/memremap.c index 1f87ea6b6545..d57d58f77409 100644 --- a/kernel/memremap.c +++ b/kernel/memremap.c @@ -43,7 +43,7 @@ static unsigned long order_at(struct resource *res, unsigned long pgoff) pgoff += 1UL << order, order = order_at((res), pgoff)) #if IS_ENABLED(CONFIG_DEVICE_PRIVATE) -int device_private_entry_fault(struct vm_area_struct *vma, +vm_fault_t device_private_entry_fault(struct vm_area_struct *vma, unsigned long addr, swp_entry_t entry, unsigned int flags, diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index 918f386b2f6e..924e37fb1620 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -66,6 +66,9 @@ int console_printk[4] = { CONSOLE_LOGLEVEL_DEFAULT, /* default_console_loglevel */ }; +atomic_t ignore_console_lock_warning __read_mostly = ATOMIC_INIT(0); +EXPORT_SYMBOL(ignore_console_lock_warning); + /* * Low level drivers may need that to know if they can schedule in * their unblank() callback or not. So let's export it. diff --git a/kernel/sys.c b/kernel/sys.c index e27b51d3facd..cf5c67533ff1 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -1237,18 +1237,19 @@ static int override_release(char __user *release, size_t len) SYSCALL_DEFINE1(newuname, struct new_utsname __user *, name) { - int errno = 0; + struct new_utsname tmp; down_read(&uts_sem); - if (copy_to_user(name, utsname(), sizeof *name)) - errno = -EFAULT; + memcpy(&tmp, utsname(), sizeof(tmp)); up_read(&uts_sem); + if (copy_to_user(name, &tmp, sizeof(tmp))) + return -EFAULT; - if (!errno && override_release(name->release, sizeof(name->release))) - errno = -EFAULT; - if (!errno && override_architecture(name)) - errno = -EFAULT; - return errno; + if (override_release(name->release, sizeof(name->release))) + return -EFAULT; + if (override_architecture(name)) + return -EFAULT; + return 0; } #ifdef __ARCH_WANT_SYS_OLD_UNAME @@ -1257,55 +1258,46 @@ SYSCALL_DEFINE1(newuname, struct new_utsname __user *, name) */ SYSCALL_DEFINE1(uname, struct old_utsname __user *, name) { - int error = 0; + struct old_utsname tmp; if (!name) return -EFAULT; down_read(&uts_sem); - if (copy_to_user(name, utsname(), sizeof(*name))) - error = -EFAULT; + memcpy(&tmp, utsname(), sizeof(tmp)); up_read(&uts_sem); + if (copy_to_user(name, &tmp, sizeof(tmp))) + return -EFAULT; - if (!error && override_release(name->release, sizeof(name->release))) - error = -EFAULT; - if (!error && override_architecture(name)) - error = -EFAULT; - return error; + if (override_release(name->release, sizeof(name->release))) + return -EFAULT; + if (override_architecture(name)) + return -EFAULT; + return 0; } SYSCALL_DEFINE1(olduname, struct oldold_utsname __user *, name) { - int error; + struct oldold_utsname tmp = {}; if (!name) return -EFAULT; - if (!access_ok(VERIFY_WRITE, name, sizeof(struct oldold_utsname))) - return -EFAULT; down_read(&uts_sem); - error = __copy_to_user(&name->sysname, &utsname()->sysname, - __OLD_UTS_LEN); - error |= __put_user(0, name->sysname + __OLD_UTS_LEN); - error |= __copy_to_user(&name->nodename, &utsname()->nodename, - __OLD_UTS_LEN); - error |= __put_user(0, name->nodename + __OLD_UTS_LEN); - error |= __copy_to_user(&name->release, &utsname()->release, - __OLD_UTS_LEN); - error |= __put_user(0, name->release + __OLD_UTS_LEN); - error |= __copy_to_user(&name->version, &utsname()->version, - __OLD_UTS_LEN); - error |= __put_user(0, name->version + __OLD_UTS_LEN); - error |= __copy_to_user(&name->machine, &utsname()->machine, - __OLD_UTS_LEN); - error |= __put_user(0, name->machine + __OLD_UTS_LEN); + memcpy(&tmp.sysname, &utsname()->sysname, __OLD_UTS_LEN); + memcpy(&tmp.nodename, &utsname()->nodename, __OLD_UTS_LEN); + memcpy(&tmp.release, &utsname()->release, __OLD_UTS_LEN); + memcpy(&tmp.version, &utsname()->version, __OLD_UTS_LEN); + memcpy(&tmp.machine, &utsname()->machine, __OLD_UTS_LEN); up_read(&uts_sem); + if (copy_to_user(name, &tmp, sizeof(tmp))) + return -EFAULT; - if (!error && override_architecture(name)) - error = -EFAULT; - if (!error && override_release(name->release, sizeof(name->release))) - error = -EFAULT; - return error ? -EFAULT : 0; + if (override_architecture(name)) + return -EFAULT; + if (override_release(name->release, sizeof(name->release))) + return -EFAULT; + return 0; } #endif @@ -1319,17 +1311,18 @@ SYSCALL_DEFINE2(sethostname, char __user *, name, int, len) if (len < 0 || len > __NEW_UTS_LEN) return -EINVAL; - down_write(&uts_sem); errno = -EFAULT; if (!copy_from_user(tmp, name, len)) { - struct new_utsname *u = utsname(); + struct new_utsname *u; + down_write(&uts_sem); + u = utsname(); memcpy(u->nodename, tmp, len); memset(u->nodename + len, 0, sizeof(u->nodename) - len); errno = 0; uts_proc_notify(UTS_PROC_HOSTNAME); + up_write(&uts_sem); } - up_write(&uts_sem); return errno; } @@ -1337,8 +1330,9 @@ SYSCALL_DEFINE2(sethostname, char __user *, name, int, len) SYSCALL_DEFINE2(gethostname, char __user *, name, int, len) { - int i, errno; + int i; struct new_utsname *u; + char tmp[__NEW_UTS_LEN + 1]; if (len < 0) return -EINVAL; @@ -1347,11 +1341,11 @@ SYSCALL_DEFINE2(gethostname, char __user *, name, int, len) i = 1 + strlen(u->nodename); if (i > len) i = len; - errno = 0; - if (copy_to_user(name, u->nodename, i)) - errno = -EFAULT; + memcpy(tmp, u->nodename, i); up_read(&uts_sem); - return errno; + if (copy_to_user(name, tmp, i)) + return -EFAULT; + return 0; } #endif @@ -1370,17 +1364,18 @@ SYSCALL_DEFINE2(setdomainname, char __user *, name, int, len) if (len < 0 || len > __NEW_UTS_LEN) return -EINVAL; - down_write(&uts_sem); errno = -EFAULT; if (!copy_from_user(tmp, name, len)) { - struct new_utsname *u = utsname(); + struct new_utsname *u; + down_write(&uts_sem); + u = utsname(); memcpy(u->domainname, tmp, len); memset(u->domainname + len, 0, sizeof(u->domainname) - len); errno = 0; uts_proc_notify(UTS_PROC_DOMAINNAME); + up_write(&uts_sem); } - up_write(&uts_sem); return errno; } diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 71ceb6c13c1a..cc02050fd0c4 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -1808,6 +1808,24 @@ static struct ctl_table fs_table[] = { .extra2 = &one, }, { + .procname = "protected_fifos", + .data = &sysctl_protected_fifos, + .maxlen = sizeof(int), + .mode = 0600, + .proc_handler = proc_dointvec_minmax, + .extra1 = &zero, + .extra2 = &two, + }, + { + .procname = "protected_regular", + .data = &sysctl_protected_regular, + .maxlen = sizeof(int), + .mode = 0600, + .proc_handler = proc_dointvec_minmax, + .extra1 = &zero, + .extra2 = &two, + }, + { .procname = "suid_dumpable", .data = &suid_dumpable, .maxlen = sizeof(int), diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index c042a455afc6..5e3de28c7677 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -774,6 +774,18 @@ config TRACING_EVENTS_GPIO help Enable tracing events for gpio subsystem +config GCOV_PROFILE_FTRACE + bool "Enable GCOV profiling on ftrace subsystem" + depends on GCOV_KERNEL + help + Enable GCOV profiling on ftrace subsystem for checking + which functions/lines are tested. + + If unsure, say N. + + Note that on a kernel compiled with this config, ftrace will + run significantly slower. + endif # FTRACE endif # TRACING_SUPPORT diff --git a/kernel/trace/Makefile b/kernel/trace/Makefile index 98d53b39a8ee..f81dadbc7c4a 100644 --- a/kernel/trace/Makefile +++ b/kernel/trace/Makefile @@ -23,6 +23,11 @@ ifdef CONFIG_TRACING_BRANCHES KBUILD_CFLAGS += -DDISABLE_BRANCH_PROFILING endif +# for GCOV coverage profiling +ifdef CONFIG_GCOV_PROFILE_FTRACE +GCOV_PROFILE := y +endif + CFLAGS_trace_benchmark.o := -I$(src) CFLAGS_trace_events_filter.o := -I$(src) diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 2bfb2bbeb3d2..c30032367aab 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -513,7 +513,14 @@ static bool within_notrace_func(struct trace_kprobe *tk) if (!addr || !kallsyms_lookup_size_offset(addr, &size, &offset)) return false; - return !ftrace_location_range(addr - offset, addr - offset + size); + /* Get the entry address of the target function */ + addr -= offset; + + /* + * Since ftrace_location_range() does inclusive range check, we need + * to subtract 1 byte from the end address. + */ + return !ftrace_location_range(addr, addr + size - 1); } #else #define within_notrace_func(tk) (false) diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c index c3d7583fcd21..e5222b5fb4fe 100644 --- a/kernel/user_namespace.c +++ b/kernel/user_namespace.c @@ -859,7 +859,16 @@ static ssize_t map_write(struct file *file, const char __user *buf, unsigned idx; struct uid_gid_extent extent; char *kbuf = NULL, *pos, *next_line; - ssize_t ret = -EINVAL; + ssize_t ret; + + /* Only allow < page size writes at the beginning of the file */ + if ((*ppos != 0) || (count >= PAGE_SIZE)) + return -EINVAL; + + /* Slurp in the user data */ + kbuf = memdup_user_nul(buf, count); + if (IS_ERR(kbuf)) + return PTR_ERR(kbuf); /* * The userns_state_mutex serializes all writes to any given map. @@ -895,19 +904,6 @@ static ssize_t map_write(struct file *file, const char __user *buf, if (cap_valid(cap_setid) && !file_ns_capable(file, ns, CAP_SYS_ADMIN)) goto out; - /* Only allow < page size writes at the beginning of the file */ - ret = -EINVAL; - if ((*ppos != 0) || (count >= PAGE_SIZE)) - goto out; - - /* Slurp in the user data */ - kbuf = memdup_user_nul(buf, count); - if (IS_ERR(kbuf)) { - ret = PTR_ERR(kbuf); - kbuf = NULL; - goto out; - } - /* Parse the user data */ ret = -EINVAL; pos = kbuf; diff --git a/kernel/utsname_sysctl.c b/kernel/utsname_sysctl.c index 233cd8fc6910..258033d62cb3 100644 --- a/kernel/utsname_sysctl.c +++ b/kernel/utsname_sysctl.c @@ -18,7 +18,7 @@ #ifdef CONFIG_PROC_SYSCTL -static void *get_uts(struct ctl_table *table, int write) +static void *get_uts(struct ctl_table *table) { char *which = table->data; struct uts_namespace *uts_ns; @@ -26,21 +26,9 @@ static void *get_uts(struct ctl_table *table, int write) uts_ns = current->nsproxy->uts_ns; which = (which - (char *)&init_uts_ns) + (char *)uts_ns; - if (!write) - down_read(&uts_sem); - else - down_write(&uts_sem); return which; } -static void put_uts(struct ctl_table *table, int write, void *which) -{ - if (!write) - up_read(&uts_sem); - else - up_write(&uts_sem); -} - /* * Special case of dostring for the UTS structure. This has locks * to observe. Should this be in kernel/sys.c ???? @@ -50,13 +38,34 @@ static int proc_do_uts_string(struct ctl_table *table, int write, { struct ctl_table uts_table; int r; + char tmp_data[__NEW_UTS_LEN + 1]; + memcpy(&uts_table, table, sizeof(uts_table)); - uts_table.data = get_uts(table, write); + uts_table.data = tmp_data; + + /* + * Buffer the value in tmp_data so that proc_dostring() can be called + * without holding any locks. + * We also need to read the original value in the write==1 case to + * support partial writes. + */ + down_read(&uts_sem); + memcpy(tmp_data, get_uts(table), sizeof(tmp_data)); + up_read(&uts_sem); r = proc_dostring(&uts_table, write, buffer, lenp, ppos); - put_uts(table, write, uts_table.data); - if (write) + if (write) { + /* + * Write back the new value. + * Note that, since we dropped uts_sem, the result can + * theoretically be incorrect if there are two parallel writes + * at non-zero offsets to the same sysctl. + */ + down_write(&uts_sem); + memcpy(get_uts(table), tmp_data, sizeof(tmp_data)); + up_write(&uts_sem); proc_sys_poll_notify(table->poll); + } return r; } diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 78b192071ef7..60e80198c3df 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -2652,6 +2652,9 @@ void flush_workqueue(struct workqueue_struct *wq) if (WARN_ON(!wq_online)) return; + lock_map_acquire(&wq->lockdep_map); + lock_map_release(&wq->lockdep_map); + mutex_lock(&wq->mutex); /* @@ -2843,7 +2846,8 @@ reflush: } EXPORT_SYMBOL_GPL(drain_workqueue); -static bool start_flush_work(struct work_struct *work, struct wq_barrier *barr) +static bool start_flush_work(struct work_struct *work, struct wq_barrier *barr, + bool from_cancel) { struct worker *worker = NULL; struct worker_pool *pool; @@ -2885,7 +2889,8 @@ static bool start_flush_work(struct work_struct *work, struct wq_barrier *barr) * workqueues the deadlock happens when the rescuer stalls, blocking * forward progress. */ - if (pwq->wq->saved_max_active == 1 || pwq->wq->rescuer) { + if (!from_cancel && + (pwq->wq->saved_max_active == 1 || pwq->wq->rescuer)) { lock_map_acquire(&pwq->wq->lockdep_map); lock_map_release(&pwq->wq->lockdep_map); } @@ -2896,6 +2901,27 @@ already_gone: return false; } +static bool __flush_work(struct work_struct *work, bool from_cancel) +{ + struct wq_barrier barr; + + if (WARN_ON(!wq_online)) + return false; + + if (!from_cancel) { + lock_map_acquire(&work->lockdep_map); + lock_map_release(&work->lockdep_map); + } + + if (start_flush_work(work, &barr, from_cancel)) { + wait_for_completion(&barr.done); + destroy_work_on_stack(&barr.work); + return true; + } else { + return false; + } +} + /** * flush_work - wait for a work to finish executing the last queueing instance * @work: the work to flush @@ -2909,18 +2935,7 @@ already_gone: */ bool flush_work(struct work_struct *work) { - struct wq_barrier barr; - - if (WARN_ON(!wq_online)) - return false; - - if (start_flush_work(work, &barr)) { - wait_for_completion(&barr.done); - destroy_work_on_stack(&barr.work); - return true; - } else { - return false; - } + return __flush_work(work, false); } EXPORT_SYMBOL_GPL(flush_work); @@ -2986,7 +3001,7 @@ static bool __cancel_work_timer(struct work_struct *work, bool is_dwork) * isn't executing. */ if (wq_online) - flush_work(work); + __flush_work(work, true); clear_work_data(work); |
