关键词:coredump、core_pattern、coredump_filter等等。
应用程序在运行过程中由于各种异常或者bug导致退出,在满足一定条件下产生一个core文件。
通常core文件包含了程序运行时内存、寄存器状态、堆栈指针、内存管理信息以及函数调用堆栈信息。
core就是程序当前工作转改存储生成的一个文件,通过工具分析这个文件,可以定位到程序异常退出的时候对应的堆栈调用等信息,找出问题点并解决。
1. 配置coredump
如果需要使用需要通过ulimit进行设置,可以通过ulimit -c查看当前系统是否支持coredump。如果为0,则表示coredump被关闭。
通过ulimit -c unlimited可以打开coredump。
coredump文件默认存储位置与可执行文件在同一目录下,文件名为core。
可以通过/proc/sys/kernel/core_pattern进行设置。
%p 出Core进程的PID %u 出Core进程的UID %s 造成Core的signal号 %t 出Core的时间,从1970-01-0100:00:00开始的秒数 %e 出Core进程对应的可执行文件名
通过echo "core-%e-%p-%s-%t" > /proc/sys/kernel/core_pattern。
在每个进程下都有coredump_filter节点/proc/<pid>/coredump_filter。
通过配置coredump_filter可以选择需在coredump的时候,将哪些内容dump到core文件中。
- (bit 0) anonymous private memory - (bit 1) anonymous shared memory - (bit 2) file-backed private memory - (bit 3) file-backed shared memory - (bit 4) ELF header pages in file-backed private memory areas (it is effective only if the bit 2 is cleared) - (bit 5) hugetlb private memory - (bit 6) hugetlb shared memory - (bit 7) DAX private memory - (bit 8) DAX shared memory
coredump_filter的默认值是0x33,也即发生coredump时会将所有anonymous内存、ELF头页面、hugetlb private memory内容保存。
coredump_filter可以被子进程继承,可以echo 0xXX > /proc/self/coredump_filter设置当前进程的coredump_filter。
static ssize_t proc_coredump_filter_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos) { ... ret = kstrtouint_from_user(buf, count, 0, &val);-------------------------将buf转换成val值。 if (ret < 0) return ret; ... for (i = 0, mask = 1; i < MMF_DUMP_FILTER_BITS; i++, mask <<= 1) { if (val & mask) set_bit(i + MMF_DUMP_FILTER_SHIFT, &mm->flags);------------------将coredump_filter的值映射到mm->flags上,后续coredump时使用。 else clear_bit(i + MMF_DUMP_FILTER_SHIFT, &mm->flags); } ... }
其中MMF_DUMP_FILTER_SHIFT为2,所以flags和coredump_filter存在如下对应关系。
#define MMF_DUMP_ANON_PRIVATE 2 #define MMF_DUMP_ANON_SHARED 3 #define MMF_DUMP_MAPPED_PRIVATE 4 #define MMF_DUMP_MAPPED_SHARED 5 #define MMF_DUMP_ELF_HEADERS 6 #define MMF_DUMP_HUGETLB_PRIVATE 7 #define MMF_DUMP_HUGETLB_SHARED 8 #define MMF_DUMP_DAX_PRIVATE 9 #define MMF_DUMP_DAX_SHARED 10
2. coredump原理
在do_signal()中根据信号判断是否触发coredump,当然还跟coredump limit、mm->flags等等相关。
满足coredump条件后,由do_coredump()进行coredump文件生成,核心是由binfmt->core_dump()进行的。
2.1 触发coredump的条件?
在内核返回用户空间的时候,会调用do_signal()处理信号。
static void do_signal(struct pt_regs *regs, int syscall) { unsigned int retval = 0, continue_addr = 0, restart_addr = 0; struct ksignal ksig; ... if (get_signal(&ksig)) { ... } ... } int get_signal(struct ksignal *ksig) { ... for (;;) { struct k_sigaction *ka; ... signr = dequeue_signal(current, ¤t->blocked, &ksig->info); ... /* Trace actually delivered signals. */ trace_signal_deliver(signr, &ksig->info, ka); ... if (sig_kernel_coredump(signr)) { if (print_fatal_signals)------------------------------可以通过kernel.print-fatal-signals = 1进行设置,对应的节点是/proc/sys/kernel/print-fatal-signals。 print_fatal_signal(ksig->info.si_signo);----------打印当前信号及当前场景的栈信息。 proc_coredump_connector(current); do_coredump(&ksig->info); } ... } spin_unlock_irq(&sighand->siglock); ksig->sig = signr; return ksig->sig > 0; } #define sig_kernel_coredump(sig) siginmask(sig, SIG_KERNEL_COREDUMP_MASK)
#define SIG_KERNEL_COREDUMP_MASK (
rt_sigmask(SIGQUIT) | rt_sigmask(SIGILL) |
rt_sigmask(SIGTRAP) | rt_sigmask(SIGABRT) |
rt_sigmask(SIGFPE) | rt_sigmask(SIGSEGV) |
rt_sigmask(SIGBUS) | rt_sigmask(SIGSYS) |
rt_sigmask(SIGXCPU) | rt_sigmask(SIGXFSZ) |
SIGEMT_MASK )
在get_signal()中,判断信号是否会导致coredump。这些信号包括SIGQUIT、SIGILL、SIGTRAP、SIGABRT、SIGFPE、SIGSEGV、SIGBUS、SIGSYS、SIGXCPU、SIGXFSZ。
“终止w/core”表示在进程当前工作目录的core文件中复制了该进程的存储图像(该文件名为core,由此可以看出这种功能很久之前就是UNIX功能的一部分)。
void proc_coredump_connector(struct task_struct *task) { struct cn_msg *msg; struct proc_event *ev; __u8 buffer[CN_PROC_MSG_SIZE] __aligned(8); if (atomic_read(&proc_event_num_listeners) < 1) return; msg = buffer_to_cn_msg(buffer); ev = (struct proc_event *)msg->data; memset(&ev->event_data, 0, sizeof(ev->event_data)); ev->timestamp_ns = ktime_get_ns(); ev->what = PROC_EVENT_COREDUMP; ev->event_data.coredump.process_pid = task->pid; ev->event_data.coredump.process_tgid = task->tgid; memcpy(&msg->id, &cn_proc_event_id, sizeof(msg->id)); msg->ack = 0; /* not used */ msg->len = sizeof(*ev); msg->flags = 0; /* not used */ send_msg(msg); }
2.2 coredump如何生成?
void do_coredump(const siginfo_t *siginfo) { struct core_state core_state; struct core_name cn; struct mm_struct *mm = current->mm; struct linux_binfmt * binfmt; const struct cred *old_cred; struct cred *cred; int retval = 0; int ispipe; struct files_struct *displaced; /* require nonrelative corefile path and be extra careful */ bool need_suid_safe = false; bool core_dumped = false; static atomic_t core_dump_count = ATOMIC_INIT(0); struct coredump_params cprm = { .siginfo = siginfo, .regs = signal_pt_regs(), .limit = rlimit(RLIMIT_CORE),-----------------------------------获取系统对于coredump的限制。 /* * We must use the same mm->flags while dumping core to avoid * inconsistency of bit flags, since this flag is not protected * by any locks. */ .mm_flags = mm->flags, }; audit_core_dumps(siginfo->si_signo); binfmt = mm->binfmt;------------------------------------------------获取当前进程所使用的程序加载器。 if (!binfmt || !binfmt->core_dump) goto fail; if (!__get_dumpable(cprm.mm_flags))---------------------------------从当前进程的mm->flags中取低两位判断是否可以coredump,SUID_DUMP_DISABLE(0)不可以,其他情况都可以。 goto fail; cred = prepare_creds(); if (!cred) goto fail; /* * We cannot trust fsuid as being the "true" uid of the process * nor do we know its entire history. We only know it was tainted * so we dump it as root in mode 2, and only into a controlled * environment (pipe handler or fully qualified path). */ if (__get_dumpable(cprm.mm_flags) == SUID_DUMP_ROOT) {--------------区分SUID_DUMP_USER和SUID_DUMP_ROOT。 /* Setuid core dump mode */ cred->fsuid = GLOBAL_ROOT_UID; /* Dump root private */ need_suid_safe = true; } retval = coredump_wait(siginfo->si_signo, &core_state); if (retval < 0) goto fail_creds; old_cred = override_creds(cred); ispipe = format_corename(&cn, &cprm);-------------------------------根据core_pattern判断是否是ispipe,然后根据core_pattern的设置生成coredump文件名称。 if (ispipe) {-------------------------------------------------------通过管道处理coredump信息。 int dump_count; char **helper_argv; struct subprocess_info *sub_info; if (ispipe < 0) { printk(KERN_WARNING "format_corename failed "); printk(KERN_WARNING "Aborting core "); goto fail_unlock; } if (cprm.limit == 1) { printk(KERN_WARNING "Process %d(%s) has RLIMIT_CORE set to 1 ", task_tgid_vnr(current), current->comm); printk(KERN_WARNING "Aborting core "); goto fail_unlock; } cprm.limit = RLIM_INFINITY; dump_count = atomic_inc_return(&core_dump_count); if (core_pipe_limit && (core_pipe_limit < dump_count)) { printk(KERN_WARNING "Pid %d(%s) over core_pipe_limit ", task_tgid_vnr(current), current->comm); printk(KERN_WARNING "Skipping core dump "); goto fail_dropcount; } helper_argv = argv_split(GFP_KERNEL, cn.corename, NULL);----------将cn.corename参数进行拆分。 if (!helper_argv) { printk(KERN_WARNING "%s failed to allocate memory ", __func__); goto fail_dropcount; } retval = -ENOMEM; sub_info = call_usermodehelper_setup(helper_argv[0], helper_argv, NULL, GFP_KERNEL, umh_pipe_setup, NULL, &cprm);---------------------通过usermodehelper调用用户空间的helper_argv[0]程序进行core_pattern。 if (sub_info) retval = call_usermodehelper_exec(sub_info, UMH_WAIT_EXEC);-----------------------------UMH_WAIT_EXEC表示在内核exec用户空间程序之后就退出,此时用户空间程序就通过pipe等待接收数据。 argv_free(helper_argv); if (retval) { printk(KERN_INFO "Core dump to |%s pipe failed ", cn.corename); goto close_fail; } } else { struct inode *inode; int open_flags = O_CREAT | O_RDWR | O_NOFOLLOW | O_LARGEFILE | O_EXCL; if (cprm.limit < binfmt->min_coredump) goto fail_unlock; if (need_suid_safe && cn.corename[0] != '/') { printk(KERN_WARNING "Pid %d(%s) can only dump core " "to fully qualified path! ", task_tgid_vnr(current), current->comm); printk(KERN_WARNING "Skipping core dump "); goto fail_unlock; } if (!need_suid_safe) { mm_segment_t old_fs; old_fs = get_fs(); set_fs(KERNEL_DS); /* * If it doesn't exist, that's fine. If there's some * other problem, we'll catch it at the filp_open(). */ (void) sys_unlink((const char __user *)cn.corename); set_fs(old_fs); } if (need_suid_safe) {---------------------------------------------创建coredump文件。 struct path root; task_lock(&init_task); get_fs_root(init_task.fs, &root); task_unlock(&init_task); cprm.file = file_open_root(root.dentry, root.mnt, cn.corename, open_flags, 0600); path_put(&root); } else { cprm.file = filp_open(cn.corename, open_flags, 0600); } if (IS_ERR(cprm.file)) goto fail_unlock; inode = file_inode(cprm.file); if (inode->i_nlink > 1)------------------------------------------coredummp文件不能有多个硬链接。 goto close_fail; if (d_unhashed(cprm.file->f_path.dentry)) goto close_fail; if (!S_ISREG(inode->i_mode))--------------------------------------coredump文件必须为普通文件。 goto close_fail; if (!uid_eq(inode->i_uid, current_fsuid())) goto close_fail; if ((inode->i_mode & 0677) != 0600) goto close_fail; if (!(cprm.file->f_mode & FMODE_CAN_WRITE))-----------------------coredump文件必须可写。 goto close_fail; if (do_truncate(cprm.file->f_path.dentry, 0, 0, cprm.file)) goto close_fail; } /* get us an unshared descriptor table; almost always a no-op */ retval = unshare_files(&displaced); if (retval) goto close_fail; if (displaced) put_files_struct(displaced); if (!dump_interrupted()) { file_start_write(cprm.file); core_dumped = binfmt->core_dump(&cprm);---------------------------调用对应程序加载器的core_dump进行处理,将数据写入到cprm.file中。 file_end_write(cprm.file); } if (ispipe && core_pipe_limit) wait_for_dump_helpers(cprm.file); close_fail: if (cprm.file) filp_close(cprm.file, NULL); fail_dropcount: if (ispipe) atomic_dec(&core_dump_count); fail_unlock: kfree(cn.corename); coredump_finish(mm, core_dumped); revert_creds(old_cred); fail_creds: put_cred(cred); fail: return; }
format_corename()根据core_pattern中的设置,生成coredump文件名。并且判断coredump文件生成方式,ispipe为真则通过管道传输给其他应用处理;否则直接保存成文件。
static int format_corename(struct core_name *cn, struct coredump_params *cprm) { const struct cred *cred = current_cred(); const char *pat_ptr = core_pattern; int ispipe = (*pat_ptr == '|');------------------------------------------|表示通过pipe处理coredump文件。 int pid_in_pattern = 0; int err = 0; cn->used = 0; cn->corename = NULL; if (expand_corename(cn, core_name_size)) return -ENOMEM; cn->corename[0] = '