| /* |
| * This file is part of ltrace. |
| * Copyright (C) 2011,2012,2013 Petr Machata, Red Hat Inc. |
| * Copyright (C) 2010 Arnaud Patard, Mandriva SA |
| * Copyright (C) 1998,2001,2002,2003,2004,2007,2008,2009 Juan Cespedes |
| * Copyright (C) 2008 Luis Machado, IBM Corporation |
| * Copyright (C) 2006 Ian Wienand |
| * Copyright (C) 2006 Paul Gilliam, IBM Corporation |
| * |
| * This program is free software; you can redistribute it and/or |
| * modify it under the terms of the GNU General Public License as |
| * published by the Free Software Foundation; either version 2 of the |
| * License, or (at your option) any later version. |
| * |
| * This program is distributed in the hope that it will be useful, but |
| * WITHOUT ANY WARRANTY; without even the implied warranty of |
| * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| * General Public License for more details. |
| * |
| * You should have received a copy of the GNU General Public License |
| * along with this program; if not, write to the Free Software |
| * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA |
| * 02110-1301 USA |
| */ |
| |
| #include "config.h" |
| |
| #define _GNU_SOURCE |
| #include <assert.h> |
| #include <errno.h> |
| #include <signal.h> |
| #include <stdio.h> |
| #include <stdlib.h> |
| #include <string.h> |
| #include <stdbool.h> |
| |
| #include "backend.h" |
| #include "breakpoint.h" |
| #include "common.h" |
| #include "fetch.h" |
| #include "library.h" |
| #include "proc.h" |
| #include "prototype.h" |
| #include "summary.h" |
| #include "value_dict.h" |
| |
| static void handle_signal(Event *event); |
| static void handle_exit(Event *event); |
| static void handle_exit_signal(Event *event); |
| static void handle_syscall(Event *event); |
| static void handle_arch_syscall(Event *event); |
| static void handle_sysret(Event *event); |
| static void handle_arch_sysret(Event *event); |
| static void handle_clone(Event *event); |
| static void handle_exec(Event *event); |
| static void handle_breakpoint(Event *event); |
| static void handle_new(Event *event); |
| |
| static void callstack_push_syscall(struct process *proc, int sysnum); |
| static void callstack_push_symfunc(struct process *proc, struct breakpoint *bp); |
| /* XXX Stack maintenance should be moved to a dedicated module, or to |
| * proc.c, and push/pop should be visible outside this module. For |
| * now, because we need this in proc.c, this is non-static. */ |
| void callstack_pop(struct process *proc); |
| |
| static char *shortsignal(struct process *proc, int signum); |
| static char *sysname(struct process *proc, int sysnum); |
| static char *arch_sysname(struct process *proc, int sysnum); |
| |
| static Event * |
| call_handler(struct process *proc, Event *event) |
| { |
| assert(proc != NULL); |
| |
| struct event_handler *handler = proc->event_handler; |
| if (handler == NULL) |
| return event; |
| |
| return (*handler->on_event) (handler, event); |
| } |
| |
| void |
| handle_event(Event *event) |
| { |
| if (exiting == 1) { |
| debug(1, "ltrace about to exit"); |
| os_ltrace_exiting(); |
| exiting = 2; |
| } |
| debug(DEBUG_FUNCTION, "handle_event(pid=%d, type=%d)", |
| event->proc ? event->proc->pid : -1, event->type); |
| |
| /* If the thread group or an individual task define an |
| overriding event handler, give them a chance to kick in. |
| We will end up calling both handlers, if the first one |
| doesn't sink the event. */ |
| if (event->proc != NULL) { |
| event = call_handler(event->proc, event); |
| if (event == NULL) |
| /* It was handled. */ |
| return; |
| |
| /* Note: the previous handler has a chance to alter |
| * the event. */ |
| if (event->proc != NULL |
| && event->proc->leader != NULL |
| && event->proc != event->proc->leader) { |
| event = call_handler(event->proc->leader, event); |
| if (event == NULL) |
| return; |
| } |
| } |
| |
| switch (event->type) { |
| case EVENT_NONE: |
| debug(1, "event: none"); |
| return; |
| |
| case EVENT_SIGNAL: |
| assert(event->proc != NULL); |
| debug(1, "[%d] event: signal (%s [%d])", |
| event->proc->pid, |
| shortsignal(event->proc, event->e_un.signum), |
| event->e_un.signum); |
| handle_signal(event); |
| return; |
| |
| case EVENT_EXIT: |
| assert(event->proc != NULL); |
| debug(1, "[%d] event: exit (%d)", |
| event->proc->pid, |
| event->e_un.ret_val); |
| handle_exit(event); |
| return; |
| |
| case EVENT_EXIT_SIGNAL: |
| assert(event->proc != NULL); |
| debug(1, "[%d] event: exit signal (%s [%d])", |
| event->proc->pid, |
| shortsignal(event->proc, event->e_un.signum), |
| event->e_un.signum); |
| handle_exit_signal(event); |
| return; |
| |
| case EVENT_SYSCALL: |
| assert(event->proc != NULL); |
| debug(1, "[%d] event: syscall (%s [%d])", |
| event->proc->pid, |
| sysname(event->proc, event->e_un.sysnum), |
| event->e_un.sysnum); |
| handle_syscall(event); |
| return; |
| |
| case EVENT_SYSRET: |
| assert(event->proc != NULL); |
| debug(1, "[%d] event: sysret (%s [%d])", |
| event->proc->pid, |
| sysname(event->proc, event->e_un.sysnum), |
| event->e_un.sysnum); |
| handle_sysret(event); |
| return; |
| |
| case EVENT_ARCH_SYSCALL: |
| assert(event->proc != NULL); |
| debug(1, "[%d] event: arch_syscall (%s [%d])", |
| event->proc->pid, |
| arch_sysname(event->proc, event->e_un.sysnum), |
| event->e_un.sysnum); |
| handle_arch_syscall(event); |
| return; |
| |
| case EVENT_ARCH_SYSRET: |
| assert(event->proc != NULL); |
| debug(1, "[%d] event: arch_sysret (%s [%d])", |
| event->proc->pid, |
| arch_sysname(event->proc, event->e_un.sysnum), |
| event->e_un.sysnum); |
| handle_arch_sysret(event); |
| return; |
| |
| case EVENT_CLONE: |
| case EVENT_VFORK: |
| assert(event->proc != NULL); |
| debug(1, "[%d] event: clone (%u)", |
| event->proc->pid, event->e_un.newpid); |
| handle_clone(event); |
| return; |
| |
| case EVENT_EXEC: |
| assert(event->proc != NULL); |
| debug(1, "[%d] event: exec()", |
| event->proc->pid); |
| handle_exec(event); |
| return; |
| |
| case EVENT_BREAKPOINT: |
| assert(event->proc != NULL); |
| debug(1, "[%d] event: breakpoint %p", |
| event->proc->pid, event->e_un.brk_addr); |
| handle_breakpoint(event); |
| return; |
| |
| case EVENT_NEW: |
| debug(1, "[%d] event: new process", |
| event->e_un.newpid); |
| handle_new(event); |
| return; |
| default: |
| fprintf(stderr, "Error! unknown event?\n"); |
| exit(1); |
| } |
| } |
| |
| typedef struct Pending_New Pending_New; |
| struct Pending_New { |
| pid_t pid; |
| Pending_New * next; |
| }; |
| static Pending_New * pending_news = NULL; |
| |
| static int |
| pending_new(pid_t pid) { |
| Pending_New * p; |
| |
| debug(DEBUG_FUNCTION, "pending_new(%d)", pid); |
| |
| p = pending_news; |
| while (p) { |
| if (p->pid == pid) { |
| return 1; |
| } |
| p = p->next; |
| } |
| return 0; |
| } |
| |
| static void |
| pending_new_insert(pid_t pid) { |
| Pending_New * p; |
| |
| debug(DEBUG_FUNCTION, "pending_new_insert(%d)", pid); |
| |
| p = malloc(sizeof(Pending_New)); |
| if (!p) { |
| perror("malloc()"); |
| exit(1); |
| } |
| p->pid = pid; |
| p->next = pending_news; |
| pending_news = p; |
| } |
| |
| static void |
| pending_new_remove(pid_t pid) |
| { |
| debug(DEBUG_FUNCTION, "pending_new_remove(%d)", pid); |
| |
| Pending_New **pp; |
| for (pp = &pending_news; *pp != NULL; pp = &(*pp)->next) |
| if ((*pp)->pid == pid) { |
| Pending_New *p = *pp; |
| *pp = p->next; |
| free(p); |
| return; |
| } |
| } |
| |
| static void |
| handle_clone(Event *event) |
| { |
| debug(DEBUG_FUNCTION, "handle_clone(pid=%d)", event->proc->pid); |
| |
| struct process *proc = malloc(sizeof(*proc)); |
| pid_t newpid = event->e_un.newpid; |
| if (proc == NULL |
| || process_clone(proc, event->proc, newpid) < 0) { |
| free(proc); |
| proc = NULL; |
| fprintf(stderr, |
| "Couldn't initialize tracing of process %d.\n", |
| newpid); |
| |
| } else { |
| proc->parent = event->proc; |
| /* We save register values to the arch pointer, and |
| * these need to be per-thread. XXX arch_ptr should |
| * be retired in favor of fetch interface anyway. */ |
| proc->arch_ptr = NULL; |
| } |
| |
| if (pending_new(newpid)) { |
| pending_new_remove(newpid); |
| |
| if (proc != NULL) { |
| proc->event_handler = NULL; |
| if (event->proc->state == STATE_ATTACHED |
| && options.follow) |
| proc->state = STATE_ATTACHED; |
| else |
| proc->state = STATE_IGNORED; |
| } |
| |
| continue_process(newpid); |
| |
| } else if (proc != NULL) { |
| proc->state = STATE_BEING_CREATED; |
| } |
| |
| if (event->type != EVENT_VFORK) |
| continue_process(event->proc->pid); |
| else if (proc != NULL) |
| continue_after_vfork(proc); |
| else |
| continue_process(newpid); |
| } |
| |
| static void |
| handle_new(Event *event) |
| { |
| debug(DEBUG_FUNCTION, "handle_new(pid=%d)", event->e_un.newpid); |
| |
| struct process *proc = pid2proc(event->e_un.newpid); |
| if (!proc) { |
| pending_new_insert(event->e_un.newpid); |
| } else { |
| assert(proc->state == STATE_BEING_CREATED); |
| if (options.follow) { |
| proc->state = STATE_ATTACHED; |
| } else { |
| proc->state = STATE_IGNORED; |
| } |
| continue_process(proc->pid); |
| } |
| } |
| |
| static char * |
| shortsignal(struct process *proc, int signum) |
| { |
| static char *signalent0[] = { |
| #include "signalent.h" |
| }; |
| static char *signalent1[] = { |
| #include "signalent1.h" |
| }; |
| static char **signalents[] = { signalent0, signalent1 }; |
| int nsignals[] = { sizeof signalent0 / sizeof signalent0[0], |
| sizeof signalent1 / sizeof signalent1[0] |
| }; |
| |
| debug(DEBUG_FUNCTION, "shortsignal(pid=%d, signum=%d)", proc->pid, signum); |
| |
| assert(proc->personality < sizeof signalents / sizeof signalents[0]); |
| if (signum < 0 || signum >= nsignals[proc->personality]) { |
| return "UNKNOWN_SIGNAL"; |
| } else { |
| return signalents[proc->personality][signum]; |
| } |
| } |
| |
| static char * |
| sysname(struct process *proc, int sysnum) |
| { |
| static char result[128]; |
| static char *syscallent0[] = { |
| #include "syscallent.h" |
| }; |
| static char *syscallent1[] = { |
| #include "syscallent1.h" |
| }; |
| static char **syscallents[] = { syscallent0, syscallent1 }; |
| int nsyscalls[] = { |
| sizeof syscallent0 / sizeof syscallent0[0], |
| sizeof syscallent1 / sizeof syscallent1[0], |
| }; |
| |
| debug(DEBUG_FUNCTION, "sysname(pid=%d, sysnum=%d)", proc->pid, sysnum); |
| |
| assert(proc->personality < sizeof syscallents / sizeof syscallents[0]); |
| if (sysnum < 0 || sysnum >= nsyscalls[proc->personality]) { |
| sprintf(result, "SYS_%d", sysnum); |
| return result; |
| } else { |
| return syscallents[proc->personality][sysnum]; |
| } |
| } |
| |
| static char * |
| arch_sysname(struct process *proc, int sysnum) |
| { |
| static char result[128]; |
| static char *arch_syscallent[] = { |
| #include "arch_syscallent.h" |
| }; |
| int nsyscalls = sizeof arch_syscallent / sizeof arch_syscallent[0]; |
| |
| debug(DEBUG_FUNCTION, "arch_sysname(pid=%d, sysnum=%d)", proc->pid, sysnum); |
| |
| if (sysnum < 0 || sysnum >= nsyscalls) { |
| sprintf(result, "ARCH_%d", sysnum); |
| return result; |
| } else { |
| sprintf(result, "ARCH_%s", arch_syscallent[sysnum]); |
| return result; |
| } |
| } |
| |
| #ifndef HAVE_STRSIGNAL |
| # define strsignal(SIGNUM) "???" |
| #endif |
| |
| static void |
| handle_signal(Event *event) { |
| debug(DEBUG_FUNCTION, "handle_signal(pid=%d, signum=%d)", event->proc->pid, event->e_un.signum); |
| if (event->proc->state != STATE_IGNORED && !options.no_signals) { |
| output_line(event->proc, "--- %s (%s) ---", |
| shortsignal(event->proc, event->e_un.signum), |
| strsignal(event->e_un.signum)); |
| } |
| continue_after_signal(event->proc->pid, event->e_un.signum); |
| } |
| |
| static int |
| init_syscall_symbol(struct library_symbol *libsym, const char *name) |
| { |
| static struct library syscall_lib; |
| |
| if (syscall_lib.protolib == NULL) { |
| struct protolib *protolib |
| = protolib_cache_load(&g_protocache, "syscalls", 0, 1); |
| if (protolib == NULL) { |
| fprintf(stderr, "Couldn't load system call prototypes:" |
| " %s.\n", strerror(errno)); |
| |
| /* Instead, get a fake one just so we can |
| * carry on, limping. */ |
| protolib = malloc(sizeof *protolib); |
| if (protolib == NULL) { |
| fprintf(stderr, "Couldn't even allocate a fake " |
| "prototype library: %s.\n", |
| strerror(errno)); |
| abort(); |
| } |
| protolib_init(protolib); |
| } |
| |
| assert(protolib != NULL); |
| if (library_init(&syscall_lib, LT_LIBTYPE_SYSCALL) < 0) { |
| fprintf(stderr, "Couldn't initialize system call " |
| "library: %s.\n", strerror(errno)); |
| abort(); |
| } |
| |
| library_set_soname(&syscall_lib, "SYS", 0); |
| syscall_lib.protolib = protolib; |
| } |
| |
| if (library_symbol_init(libsym, 0, name, 0, LS_TOPLT_NONE) < 0) |
| return -1; |
| |
| libsym->lib = &syscall_lib; |
| return 0; |
| } |
| |
| /* Account the unfinished functions on the call stack. */ |
| static void |
| account_current_callstack(struct process *proc) |
| { |
| if (! options.summary) |
| return; |
| |
| struct timedelta spent[proc->callstack_depth]; |
| |
| size_t i; |
| for (i = 0; i < proc->callstack_depth; ++i) { |
| struct callstack_element *elem = &proc->callstack[i]; |
| spent[i] = calc_time_spent(elem->enter_time); |
| } |
| |
| for (i = 0; i < proc->callstack_depth; ++i) { |
| struct callstack_element *elem = &proc->callstack[i]; |
| struct library_symbol syscall, *libsym = NULL; |
| if (elem->is_syscall) { |
| const char *name = sysname(proc, elem->c_un.syscall); |
| if (init_syscall_symbol(&syscall, name) >= 0) |
| libsym = &syscall; |
| |
| } else { |
| libsym = elem->c_un.libfunc; |
| } |
| |
| if (libsym != NULL) { |
| summary_account_call(libsym, spent[i]); |
| |
| if (elem->is_syscall) |
| library_symbol_destroy(&syscall); |
| } |
| } |
| } |
| |
| static void |
| handle_exit(Event *event) { |
| debug(DEBUG_FUNCTION, "handle_exit(pid=%d, status=%d)", event->proc->pid, event->e_un.ret_val); |
| if (event->proc->state != STATE_IGNORED) { |
| output_line(event->proc, "+++ exited (status %d) +++", |
| event->e_un.ret_val); |
| } |
| |
| account_current_callstack(event->proc); |
| remove_process(event->proc); |
| } |
| |
| static void |
| handle_exit_signal(Event *event) { |
| debug(DEBUG_FUNCTION, "handle_exit_signal(pid=%d, signum=%d)", event->proc->pid, event->e_un.signum); |
| if (event->proc->state != STATE_IGNORED) { |
| output_line(event->proc, "+++ killed by %s +++", |
| shortsignal(event->proc, event->e_un.signum)); |
| } |
| |
| account_current_callstack(event->proc); |
| remove_process(event->proc); |
| } |
| |
| static void |
| output_syscall(struct process *proc, const char *name, enum tof tof, |
| bool left, struct timedelta *spent) |
| { |
| if (left) |
| assert(spent == NULL); |
| |
| struct library_symbol syscall; |
| if (init_syscall_symbol(&syscall, name) >= 0) { |
| if (left) { |
| if (! options.summary) |
| output_left(tof, proc, &syscall); |
| } else if (options.summary) { |
| summary_account_call(&syscall, *spent); |
| } else { |
| output_right(tof, proc, &syscall, spent); |
| } |
| |
| library_symbol_destroy(&syscall); |
| } |
| } |
| |
| static void |
| output_syscall_left(struct process *proc, const char *name) |
| { |
| output_syscall(proc, name, LT_TOF_SYSCALL, true, NULL); |
| } |
| |
| static void |
| output_syscall_right(struct process *proc, const char *name, |
| struct timedelta *spent) |
| { |
| output_syscall(proc, name, LT_TOF_SYSCALLR, false, spent); |
| } |
| |
| static void |
| handle_syscall(Event *event) |
| { |
| debug(DEBUG_FUNCTION, "handle_syscall(pid=%d, sysnum=%d)", event->proc->pid, event->e_un.sysnum); |
| if (event->proc->state != STATE_IGNORED) { |
| callstack_push_syscall(event->proc, event->e_un.sysnum); |
| if (options.syscalls) |
| output_syscall_left(event->proc, |
| sysname(event->proc, |
| event->e_un.sysnum)); |
| } |
| continue_after_syscall(event->proc, event->e_un.sysnum, 0); |
| } |
| |
| static void |
| handle_exec(Event *event) |
| { |
| struct process *proc = event->proc; |
| |
| /* Save the PID so that we can use it after unsuccessful |
| * process_exec. */ |
| pid_t pid = proc->pid; |
| |
| debug(DEBUG_FUNCTION, "handle_exec(pid=%d)", proc->pid); |
| if (proc->state == STATE_IGNORED) { |
| untrace: |
| untrace_pid(pid); |
| remove_process(proc); |
| return; |
| } |
| output_line(proc, "--- Called exec() ---"); |
| |
| account_current_callstack(proc); |
| |
| if (process_exec(proc) < 0) { |
| fprintf(stderr, |
| "couldn't reinitialize process %d after exec\n", pid); |
| goto untrace; |
| } |
| |
| continue_after_exec(proc); |
| } |
| |
| static void |
| handle_arch_syscall(Event *event) { |
| debug(DEBUG_FUNCTION, "handle_arch_syscall(pid=%d, sysnum=%d)", event->proc->pid, event->e_un.sysnum); |
| if (event->proc->state != STATE_IGNORED) { |
| callstack_push_syscall(event->proc, 0xf0000 + event->e_un.sysnum); |
| if (options.syscalls) { |
| output_syscall_left(event->proc, |
| arch_sysname(event->proc, |
| event->e_un.sysnum)); |
| } |
| } |
| continue_process(event->proc->pid); |
| } |
| |
| static void |
| handle_x_sysret(Event *event, char *(*name_cb)(struct process *, int)) |
| { |
| debug(DEBUG_FUNCTION, "handle_x_sysret(pid=%d, sysnum=%d)", |
| event->proc->pid, event->e_un.sysnum); |
| |
| unsigned d = event->proc->callstack_depth; |
| assert(d > 0); |
| struct callstack_element *elem = &event->proc->callstack[d - 1]; |
| assert(elem->is_syscall); |
| |
| if (event->proc->state != STATE_IGNORED) { |
| struct timedelta spent = calc_time_spent(elem->enter_time); |
| if (options.syscalls) |
| output_syscall_right(event->proc, |
| name_cb(event->proc, |
| event->e_un.sysnum), |
| &spent); |
| |
| callstack_pop(event->proc); |
| } |
| continue_after_syscall(event->proc, event->e_un.sysnum, 1); |
| } |
| |
| static void |
| handle_sysret(Event *event) |
| { |
| handle_x_sysret(event, &sysname); |
| } |
| |
| static void |
| handle_arch_sysret(Event *event) |
| { |
| handle_x_sysret(event, &arch_sysname); |
| } |
| |
| static void |
| output_right_tos(struct process *proc) |
| { |
| size_t d = proc->callstack_depth; |
| assert(d > 0); |
| struct callstack_element *elem = &proc->callstack[d - 1]; |
| assert(! elem->is_syscall); |
| |
| if (proc->state != STATE_IGNORED) { |
| struct timedelta spent = calc_time_spent(elem->enter_time); |
| if (options.summary) |
| summary_account_call(elem->c_un.libfunc, spent); |
| else |
| output_right(LT_TOF_FUNCTIONR, proc, elem->c_un.libfunc, |
| &spent); |
| } |
| } |
| |
| #ifndef ARCH_HAVE_SYMBOL_RET |
| void arch_symbol_ret(struct process *proc, struct library_symbol *libsym) |
| { |
| } |
| #endif |
| |
| static void |
| handle_breakpoint(Event *event) |
| { |
| int i, j; |
| struct breakpoint *sbp; |
| struct process *leader = event->proc->leader; |
| void *brk_addr = event->e_un.brk_addr; |
| |
| /* The leader has terminated. */ |
| if (leader == NULL) { |
| continue_process(event->proc->pid); |
| return; |
| } |
| |
| debug(DEBUG_FUNCTION, "handle_breakpoint(pid=%d, addr=%p)", |
| event->proc->pid, brk_addr); |
| debug(2, "event: breakpoint (%p)", brk_addr); |
| |
| for (i = event->proc->callstack_depth - 1; i >= 0; i--) { |
| if (brk_addr == event->proc->callstack[i].return_addr) { |
| for (j = event->proc->callstack_depth - 1; j > i; j--) |
| callstack_pop(event->proc); |
| |
| struct library_symbol *libsym = |
| event->proc->callstack[i].c_un.libfunc; |
| |
| arch_symbol_ret(event->proc, libsym); |
| output_right_tos(event->proc); |
| callstack_pop(event->proc); |
| |
| /* Pop also any other entries that seem like |
| * they are linked to the current one: they |
| * have the same return address, but were made |
| * for different symbols. This should only |
| * happen for entry point tracing, i.e. for -x |
| * everywhere, or -x and -e on MIPS. */ |
| while (event->proc->callstack_depth > 0) { |
| struct callstack_element *prev; |
| size_t d = event->proc->callstack_depth; |
| prev = &event->proc->callstack[d - 1]; |
| |
| if (prev->c_un.libfunc == libsym |
| || prev->return_addr != brk_addr) |
| break; |
| |
| arch_symbol_ret(event->proc, |
| prev->c_un.libfunc); |
| output_right_tos(event->proc); |
| callstack_pop(event->proc); |
| } |
| |
| /* Maybe the previous callstack_pop's got rid |
| * of the breakpoint, but if we are in a |
| * recursive call, it's still enabled. In |
| * that case we need to skip it properly. */ |
| if ((sbp = address2bpstruct(leader, brk_addr)) != NULL) { |
| continue_after_breakpoint(event->proc, sbp); |
| } else { |
| set_instruction_pointer(event->proc, brk_addr); |
| continue_process(event->proc->pid); |
| } |
| return; |
| } |
| } |
| |
| if ((sbp = address2bpstruct(leader, brk_addr)) != NULL) |
| breakpoint_on_hit(sbp, event->proc); |
| else if (event->proc->state != STATE_IGNORED) |
| output_line(event->proc, |
| "unexpected breakpoint at %p", brk_addr); |
| |
| /* breakpoint_on_hit may delete its own breakpoint, so we have |
| * to look it up again. */ |
| if ((sbp = address2bpstruct(leader, brk_addr)) != NULL) { |
| |
| if (event->proc->state != STATE_IGNORED |
| && sbp->libsym != NULL) { |
| event->proc->stack_pointer = get_stack_pointer(event->proc); |
| callstack_push_symfunc(event->proc, sbp); |
| if (! options.summary) |
| output_left(LT_TOF_FUNCTION, event->proc, |
| sbp->libsym); |
| } |
| |
| breakpoint_on_continue(sbp, event->proc); |
| return; |
| } else { |
| set_instruction_pointer(event->proc, brk_addr); |
| } |
| |
| continue_process(event->proc->pid); |
| } |
| |
| static void |
| callstack_push_syscall(struct process *proc, int sysnum) |
| { |
| struct callstack_element *elem; |
| |
| debug(DEBUG_FUNCTION, "callstack_push_syscall(pid=%d, sysnum=%d)", proc->pid, sysnum); |
| /* FIXME: not good -- should use dynamic allocation. 19990703 mortene. */ |
| if (proc->callstack_depth == MAX_CALLDEPTH - 1) { |
| fprintf(stderr, "%s: Error: call nesting too deep!\n", __func__); |
| abort(); |
| return; |
| } |
| |
| elem = &proc->callstack[proc->callstack_depth]; |
| *elem = (struct callstack_element){}; |
| elem->is_syscall = 1; |
| elem->c_un.syscall = sysnum; |
| elem->return_addr = NULL; |
| |
| proc->callstack_depth++; |
| if (opt_T || options.summary) { |
| struct timezone tz; |
| gettimeofday(&elem->enter_time, &tz); |
| } |
| } |
| |
| static void |
| callstack_push_symfunc(struct process *proc, struct breakpoint *bp) |
| { |
| struct callstack_element *elem; |
| |
| debug(DEBUG_FUNCTION, "callstack_push_symfunc(pid=%d, symbol=%s)", |
| proc->pid, bp->libsym->name); |
| /* FIXME: not good -- should use dynamic allocation. 19990703 mortene. */ |
| if (proc->callstack_depth == MAX_CALLDEPTH - 1) { |
| fprintf(stderr, "%s: Error: call nesting too deep!\n", __func__); |
| abort(); |
| return; |
| } |
| |
| elem = &proc->callstack[proc->callstack_depth++]; |
| *elem = (struct callstack_element){}; |
| elem->is_syscall = 0; |
| elem->c_un.libfunc = bp->libsym; |
| |
| struct breakpoint *rbp = NULL; |
| if (breakpoint_get_return_bp(&rbp, bp, proc) == 0 |
| && rbp != NULL) { |
| struct breakpoint *ext_rbp = insert_breakpoint(proc, rbp); |
| if (ext_rbp != rbp) { |
| breakpoint_destroy(rbp); |
| free(rbp); |
| rbp = ext_rbp; |
| } |
| } |
| |
| elem->return_addr = rbp != NULL ? rbp->addr : 0; |
| |
| if (opt_T || options.summary) { |
| struct timezone tz; |
| gettimeofday(&elem->enter_time, &tz); |
| } |
| } |
| |
| void |
| callstack_pop(struct process *proc) |
| { |
| struct callstack_element *elem; |
| assert(proc->callstack_depth > 0); |
| |
| debug(DEBUG_FUNCTION, "callstack_pop(pid=%d)", proc->pid); |
| elem = &proc->callstack[proc->callstack_depth - 1]; |
| if (!elem->is_syscall && elem->return_addr) { |
| struct breakpoint *bp |
| = address2bpstruct(proc->leader, elem->return_addr); |
| if (bp != NULL) { |
| breakpoint_on_hit(bp, proc); |
| delete_breakpoint(proc, bp); |
| } |
| } |
| |
| if (elem->fetch_context != NULL) |
| fetch_arg_done(elem->fetch_context); |
| |
| if (elem->arguments != NULL) { |
| val_dict_destroy(elem->arguments); |
| free(elem->arguments); |
| } |
| |
| proc->callstack_depth--; |
| } |