1
0
mirror of https://github.com/mpv-player/mpv synced 2024-12-28 10:02:17 +00:00
mpv/common/msg.c
Kacper Michajłow 9030e6a7ad msg: use fwrite
No need to recalc string length.
2024-04-13 18:23:16 +02:00

1118 lines
37 KiB
C

/*
* This file is part of mpv.
*
* mpv is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* mpv 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 Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with mpv. If not, see <http://www.gnu.org/licenses/>.
*/
#include <assert.h>
#include <stdarg.h>
#include <stdatomic.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include "mpv_talloc.h"
#include "misc/bstr.h"
#include "common/common.h"
#include "common/global.h"
#include "misc/bstr.h"
#include "options/options.h"
#include "options/path.h"
#include "osdep/terminal.h"
#include "osdep/io.h"
#include "osdep/threads.h"
#include "osdep/timer.h"
#include "libmpv/client.h"
#include "msg.h"
#include "msg_control.h"
// log buffer size (lines) logfile level
#define FILE_BUF 100
// lines to accumulate before any client requests the terminal loglevel
#define EARLY_TERM_BUF 100
// logfile lines to accumulate during init before we know the log file name.
// thousands of logfile lines during init can happen (especially with many
// scripts, big config, etc), so we set 5000. If it cycles and messages are
// overwritten, then the first (virtual) log line indicates how many were lost.
#define EARLY_FILE_BUF 5000
struct mp_log_root {
struct mpv_global *global;
mp_mutex lock;
mp_mutex log_file_lock;
mp_cond log_file_wakeup;
// --- protected by lock
char **msg_levels;
bool use_terminal; // make accesses to stderr/stdout
bool module;
bool show_time;
int blank_lines; // number of lines usable by status
int status_lines; // number of current status lines
bool color[STDERR_FILENO + 1];
bool isatty[STDERR_FILENO + 1];
int verbose;
bool really_quiet;
bool force_stderr;
struct mp_log_buffer **buffers;
int num_buffers;
struct mp_log_buffer *early_buffer;
struct mp_log_buffer *early_filebuffer;
FILE *stats_file;
bstr buffer;
bstr term_msg;
bstr term_msg_tmp;
bstr status_line;
struct mp_log *status_log;
bstr term_status_msg;
// --- must be accessed atomically
/* This is incremented every time the msglevels must be reloaded.
* (This is perhaps better than maintaining a globally accessible and
* synchronized mp_log tree.) */
atomic_ulong reload_counter;
// --- owner thread only (caller of mp_msg_init() etc.)
char *log_path;
char *stats_path;
mp_thread log_file_thread;
// --- owner thread only, but frozen while log_file_thread is running
FILE *log_file;
struct mp_log_buffer *log_file_buffer;
// --- protected by log_file_lock
bool log_file_thread_active; // also termination signal for the thread
int module_indent;
};
struct mp_log {
struct mp_log_root *root;
const char *prefix;
const char *verbose_prefix;
int max_level; // minimum log level for this instance
int level; // minimum log level for any outputs
int terminal_level; // minimum log level for terminal output
atomic_ulong reload_counter;
bstr partial[MSGL_MAX + 1];
};
struct mp_log_buffer {
struct mp_log_root *root;
mp_mutex lock;
// --- protected by lock
struct mp_log_buffer_entry **entries; // ringbuffer
int capacity; // total space in entries[]
int entry0; // first (oldest) entry index
int num_entries; // number of valid entries after entry0
uint64_t dropped; // number of skipped entries
bool silent;
// --- immutable
void (*wakeup_cb)(void *ctx);
void *wakeup_cb_ctx;
int level;
};
static const struct mp_log null_log = {0};
struct mp_log *const mp_null_log = (struct mp_log *)&null_log;
static bool match_mod(const char *name, const char *mod)
{
if (!strcmp(mod, "all"))
return true;
// Path prefix matches
bstr b = bstr0(name);
return bstr_eatstart0(&b, mod) && (bstr_eatstart0(&b, "/") || !b.len);
}
static void update_loglevel(struct mp_log *log)
{
struct mp_log_root *root = log->root;
mp_mutex_lock(&root->lock);
log->level = MSGL_STATUS + root->verbose; // default log level
if (root->really_quiet)
log->level = -1;
for (int n = 0; root->msg_levels && root->msg_levels[n * 2 + 0]; n++) {
if (match_mod(log->verbose_prefix, root->msg_levels[n * 2 + 0]))
log->level = mp_msg_find_level(root->msg_levels[n * 2 + 1]);
}
log->terminal_level = log->level;
for (int n = 0; n < log->root->num_buffers; n++) {
int buffer_level = log->root->buffers[n]->level;
if (buffer_level == MP_LOG_BUFFER_MSGL_LOGFILE)
buffer_level = MSGL_DEBUG;
if (buffer_level != MP_LOG_BUFFER_MSGL_TERM)
log->level = MPMAX(log->level, buffer_level);
}
if (log->root->log_file)
log->level = MPMAX(log->level, MSGL_DEBUG);
if (log->root->stats_file)
log->level = MPMAX(log->level, MSGL_STATS);
log->level = MPMIN(log->level, log->max_level);
atomic_store(&log->reload_counter, atomic_load(&log->root->reload_counter));
mp_mutex_unlock(&root->lock);
}
// Set (numerically) the maximum level that should still be output for this log
// instances. E.g. lev=MSGL_WARN => show only warnings and errors.
void mp_msg_set_max_level(struct mp_log *log, int lev)
{
if (!log->root)
return;
mp_mutex_lock(&log->root->lock);
log->max_level = MPCLAMP(lev, -1, MSGL_MAX);
mp_mutex_unlock(&log->root->lock);
update_loglevel(log);
}
// Get the current effective msg level.
// Thread-safety: see mp_msg().
int mp_msg_level(struct mp_log *log)
{
struct mp_log_root *root = log->root;
if (!root)
return -1;
if (atomic_load_explicit(&log->reload_counter, memory_order_relaxed) !=
atomic_load_explicit(&root->reload_counter, memory_order_relaxed))
{
update_loglevel(log);
}
return log->level;
}
static inline int term_msg_fileno(struct mp_log_root *root, int lev)
{
return (root->force_stderr || lev == MSGL_STATUS || lev == MSGL_FATAL ||
lev == MSGL_ERR || lev == MSGL_WARN) ? STDERR_FILENO : STDOUT_FILENO;
}
// Reposition cursor and clear lines for outputting the status line. In certain
// cases, like term OSD and subtitle display, the status can consist of
// multiple lines.
static void prepare_prefix(struct mp_log_root *root, bstr *out, int lev, int term_lines)
{
int new_lines = lev == MSGL_STATUS ? term_lines : 0;
out->len = 0;
if (!root->isatty[term_msg_fileno(root, lev)]) {
if (root->status_lines)
bstr_xappend(root, out, bstr0("\n"));
root->status_lines = new_lines;
return;
}
// Set cursor state
if (new_lines && !root->status_lines) {
bstr_xappend(root, out, bstr0(TERM_ESC_HIDE_CURSOR));
} else if (!new_lines && root->status_lines) {
bstr_xappend(root, out, bstr0(TERM_ESC_RESTORE_CURSOR));
}
int line_skip = 0;
if (root->status_lines) {
// Clear previous status line
bstr_xappend(root, out, bstr0("\033[1K\r"));
bstr up_clear = bstr0("\033[A\033[K");
for (int i = 1; i < root->status_lines; ++i)
bstr_xappend(root, out, up_clear);
assert(root->status_lines > 0 && root->blank_lines >= root->status_lines);
line_skip = root->blank_lines - root->status_lines;
}
if (new_lines)
line_skip -= MPMAX(0, root->blank_lines - new_lines);
if (line_skip)
bstr_xappend_asprintf(root, out, line_skip > 0 ? "\033[%dA" : "\033[%dB", abs(line_skip));
root->blank_lines = MPMAX(0, root->blank_lines - term_lines);
root->status_lines = new_lines;
root->blank_lines += root->status_lines;
}
void mp_msg_flush_status_line(struct mp_log *log, bool clear)
{
if (!log->root)
return;
mp_mutex_lock(&log->root->lock);
if (!log->root->status_lines)
goto done;
if (!clear) {
if (log->root->isatty[STDERR_FILENO])
fprintf(stderr, TERM_ESC_RESTORE_CURSOR);
fprintf(stderr, "\n");
log->root->blank_lines = 0;
log->root->status_lines = 0;
goto done;
}
bstr term_msg = {0};
prepare_prefix(log->root, &term_msg, MSGL_STATUS, 0);
if (term_msg.len) {
fprintf(stderr, "%.*s", BSTR_P(term_msg));
talloc_free(term_msg.start);
}
done:
log->root->status_line.len = 0;
mp_mutex_unlock(&log->root->lock);
}
void mp_msg_set_term_title(struct mp_log *log, const char *title)
{
if (log->root && title) {
// Lock because printf to terminal is not necessarily atomic.
mp_mutex_lock(&log->root->lock);
fprintf(stderr, "\033]0;%s\007", title);
mp_mutex_unlock(&log->root->lock);
}
}
bool mp_msg_has_status_line(struct mpv_global *global)
{
struct mp_log_root *root = global->log->root;
mp_mutex_lock(&root->lock);
bool r = root->status_lines > 0;
mp_mutex_unlock(&root->lock);
return r;
}
static void set_term_color(void *talloc_ctx, bstr *text, int c)
{
if (c == -1) {
bstr_xappend(talloc_ctx, text, bstr0("\033[0m"));
return;
}
bstr_xappend_asprintf(talloc_ctx, text, "\033[%d;3%dm", c >> 3, c & 7);
}
static void set_msg_color(void *talloc_ctx, bstr *text, int lev)
{
static const int v_colors[] = {9, 1, 3, -1, -1, 2, 8, 8, 8, -1};
set_term_color(talloc_ctx, text, v_colors[lev]);
}
static void pretty_print_module(struct mp_log_root *root, bstr *text,
const char *prefix, int lev)
{
size_t prefix_len = strlen(prefix);
root->module_indent = MPMAX(10, MPMAX(root->module_indent, prefix_len));
bool color = root->color[term_msg_fileno(root, lev)];
// Use random color based on the name of the module
if (color) {
unsigned int mod = 0;
for (int i = 0; i < prefix_len; ++i)
mod = mod * 33 + prefix[i];
set_term_color(root, text, (mod + 1) % 15 + 1);
}
bstr_xappend_asprintf(root, text, "%*s", root->module_indent, prefix);
if (color)
set_term_color(root, text, -1);
bstr_xappend(root, text, bstr0(": "));
if (color)
set_msg_color(root, text, lev);
}
static bool test_terminal_level(struct mp_log *log, int lev)
{
return lev <= log->terminal_level && log->root->use_terminal &&
!(lev == MSGL_STATUS && terminal_in_background());
}
// This is very basic way to infer needed width for a string.
static int term_disp_width(bstr str)
{
int width = 0;
while (str.len) {
if (bstr_eatstart0(&str, "\033[")) {
while (str.len && !((*str.start >= '@' && *str.start <= '~') || *str.start == 'm'))
str = bstr_cut(str, 1);
str = bstr_cut(str, 1);
continue;
}
bstr code = bstr_split_utf8(str, &str);
if (code.len == 0)
return 0;
if (code.len == 1 && *code.start == '\n')
continue;
// Only single-width characters are supported
width++;
// Assume that everything before \r should be discarded for simplicity
if (code.len == 1 && *code.start == '\r')
width = 0;
}
return width;
}
static void append_terminal_line(struct mp_log *log, int lev,
bstr text, bstr *term_msg, int *line_w)
{
struct mp_log_root *root = log->root;
size_t start = term_msg->len;
if (root->show_time)
bstr_xappend_asprintf(root, term_msg, "[%10.6f] ", mp_time_sec());
const char *log_prefix = (lev >= MSGL_V) || root->verbose || root->module
? log->verbose_prefix : log->prefix;
if (log_prefix) {
if (root->module) {
pretty_print_module(root, term_msg, log_prefix, lev);
} else {
bstr_xappend_asprintf(root, term_msg, "[%s] ", log_prefix);
}
}
bstr_xappend(root, term_msg, text);
*line_w = root->isatty[term_msg_fileno(root, lev)]
? term_disp_width(bstr_splice(*term_msg, start, term_msg->len)) : 0;
}
static struct mp_log_buffer_entry *log_buffer_read(struct mp_log_buffer *buffer)
{
assert(buffer->num_entries);
struct mp_log_buffer_entry *res = buffer->entries[buffer->entry0];
buffer->entry0 = (buffer->entry0 + 1) % buffer->capacity;
buffer->num_entries -= 1;
return res;
}
static void write_msg_to_buffers(struct mp_log *log, int lev, bstr text)
{
struct mp_log_root *root = log->root;
for (int n = 0; n < root->num_buffers; n++) {
struct mp_log_buffer *buffer = root->buffers[n];
bool wakeup = false;
mp_mutex_lock(&buffer->lock);
int buffer_level = buffer->level;
if (buffer_level == MP_LOG_BUFFER_MSGL_TERM)
buffer_level = log->terminal_level;
if (buffer_level == MP_LOG_BUFFER_MSGL_LOGFILE)
buffer_level = MPMAX(log->terminal_level, MSGL_DEBUG);
if (lev <= buffer_level && lev != MSGL_STATUS) {
if (buffer->level == MP_LOG_BUFFER_MSGL_LOGFILE) {
// If the buffer is full, block until we can write again,
// unless there's no write thread (died, or early filebuffer)
bool dead = false;
while (buffer->num_entries == buffer->capacity && !dead) {
// Temporary unlock is OK; buffer->level is immutable, and
// buffer can't go away because the global log lock is held.
mp_mutex_unlock(&buffer->lock);
mp_mutex_lock(&root->log_file_lock);
if (root->log_file_thread_active) {
mp_cond_wait(&root->log_file_wakeup,
&root->log_file_lock);
} else {
dead = true;
}
mp_mutex_unlock(&root->log_file_lock);
mp_mutex_lock(&buffer->lock);
}
}
if (buffer->num_entries == buffer->capacity) {
struct mp_log_buffer_entry *skip = log_buffer_read(buffer);
talloc_free(skip);
buffer->dropped += 1;
}
struct mp_log_buffer_entry *entry = talloc_ptrtype(NULL, entry);
*entry = (struct mp_log_buffer_entry) {
.prefix = talloc_strdup(entry, log->verbose_prefix),
.level = lev,
.text = bstrdup0(entry, text),
};
int pos = (buffer->entry0 + buffer->num_entries) % buffer->capacity;
buffer->entries[pos] = entry;
buffer->num_entries += 1;
if (buffer->wakeup_cb && !buffer->silent)
wakeup = true;
}
mp_mutex_unlock(&buffer->lock);
if (wakeup)
buffer->wakeup_cb(buffer->wakeup_cb_ctx);
}
}
static void dump_stats(struct mp_log *log, int lev, bstr text)
{
struct mp_log_root *root = log->root;
if (lev == MSGL_STATS && root->stats_file)
fprintf(root->stats_file, "%"PRId64" %.*s\n", mp_time_ns(), BSTR_P(text));
}
static void write_term_msg(struct mp_log *log, int lev, bstr text, bstr *out)
{
struct mp_log_root *root = log->root;
bool print_term = test_terminal_level(log, lev);
int fileno = term_msg_fileno(root, lev);
int term_w = 0, term_h = 0;
if (print_term && root->isatty[fileno])
terminal_get_size(&term_w, &term_h);
out->len = 0;
// Split away each line. Normally we require full lines; buffer partial
// lines if they happen.
root->term_msg_tmp.len = 0;
int term_msg_lines = 0;
bstr str = text;
while (str.len) {
bstr line = bstr_getline(str, &str);
if (line.start[line.len - 1] != '\n') {
assert(str.len == 0);
str = line;
break;
}
if (print_term) {
int line_w;
append_terminal_line(log, lev, line, &root->term_msg_tmp, &line_w);
term_msg_lines += (!line_w || !term_w)
? 1 : (line_w + term_w - 1) / term_w;
}
write_msg_to_buffers(log, lev, line);
}
if (lev == MSGL_STATUS) {
int line_w = 0;
if (str.len && print_term)
append_terminal_line(log, lev, str, &root->term_msg_tmp, &line_w);
term_msg_lines += !term_w ? (str.len ? 1 : 0)
: (line_w + term_w - 1) / term_w;
} else if (str.len) {
bstr_xappend(NULL, &log->partial[lev], str);
}
if (print_term && (root->term_msg_tmp.len || lev == MSGL_STATUS)) {
prepare_prefix(root, out, lev, term_msg_lines);
if (root->color[fileno] && root->term_msg_tmp.len) {
set_msg_color(root, out, lev);
set_term_color(root, &root->term_msg_tmp, -1);
}
bstr_xappend(root, out, root->term_msg_tmp);
}
}
void mp_msg_va(struct mp_log *log, int lev, const char *format, va_list va)
{
if (!mp_msg_test(log, lev))
return; // do not display
struct mp_log_root *root = log->root;
mp_mutex_lock(&root->lock);
root->buffer.len = 0;
if (log->partial[lev].len)
bstr_xappend(root, &root->buffer, log->partial[lev]);
log->partial[lev].len = 0;
bstr_xappend_vasprintf(root, &root->buffer, format, va);
// Remember last status message and restore it to ensure that it is
// always displayed
if (lev == MSGL_STATUS) {
root->status_log = log;
root->status_line.len = 0;
// Use bstr_xappend instead bstrdup to reuse allocated memory
if (root->buffer.len)
bstr_xappend(root, &root->status_line, root->buffer);
}
if (lev == MSGL_STATS) {
dump_stats(log, lev, root->buffer);
} else if (lev == MSGL_STATUS && !test_terminal_level(log, lev)) {
/* discard */
} else {
write_term_msg(log, lev, root->buffer, &root->term_msg);
root->term_status_msg.len = 0;
if (lev != MSGL_STATUS && root->status_line.len && root->status_log &&
test_terminal_level(root->status_log, MSGL_STATUS))
{
write_term_msg(root->status_log, MSGL_STATUS, root->status_line,
&root->term_status_msg);
}
int fileno = term_msg_fileno(root, lev);
FILE *stream = fileno == STDERR_FILENO ? stderr : stdout;
if (root->term_msg.len) {
fwrite(root->term_msg.start, root->term_msg.len, 1, stream);
if (root->term_status_msg.len)
fwrite(root->term_status_msg.start, root->term_status_msg.len, 1, stream);
fflush(stream);
}
}
mp_mutex_unlock(&root->lock);
}
static void destroy_log(void *ptr)
{
struct mp_log *log = ptr;
// This is not managed via talloc itself, because mp_msg calls must be
// thread-safe, while talloc is not thread-safe.
for (int lvl = 0; lvl <= MSGL_MAX; ++lvl)
talloc_free(log->partial[lvl].start);
}
// Create a new log context, which uses talloc_ctx as talloc parent, and parent
// as logical parent.
// The name is the prefix put before the output. It's usually prefixed by the
// parent's name. If the name starts with "/", the parent's name is not
// prefixed (except in verbose mode), and if it starts with "!", the name is
// not printed at all (except in verbose mode).
// If name is NULL, the parent's name/prefix is used.
// Thread-safety: fully thread-safe, but keep in mind that talloc is not (so
// talloc_ctx must be owned by the current thread).
struct mp_log *mp_log_new(void *talloc_ctx, struct mp_log *parent,
const char *name)
{
assert(parent);
struct mp_log *log = talloc_zero(talloc_ctx, struct mp_log);
if (!parent->root)
return log; // same as null_log
talloc_set_destructor(log, destroy_log);
log->root = parent->root;
log->max_level = MSGL_MAX;
if (name) {
if (name[0] == '!') {
name = &name[1];
} else if (name[0] == '/') {
name = &name[1];
log->prefix = talloc_strdup(log, name);
} else {
log->prefix = parent->prefix
? talloc_asprintf(log, "%s/%s", parent->prefix, name)
: talloc_strdup(log, name);
}
log->verbose_prefix = parent->prefix
? talloc_asprintf(log, "%s/%s", parent->prefix, name)
: talloc_strdup(log, name);
if (log->prefix && !log->prefix[0])
log->prefix = NULL;
if (!log->verbose_prefix[0])
log->verbose_prefix = "global";
} else {
log->prefix = talloc_strdup(log, parent->prefix);
log->verbose_prefix = talloc_strdup(log, parent->verbose_prefix);
}
return log;
}
void mp_msg_init(struct mpv_global *global)
{
assert(!global->log);
struct mp_log_root *root = talloc_zero(NULL, struct mp_log_root);
*root = (struct mp_log_root){
.global = global,
.reload_counter = 1,
};
mp_mutex_init(&root->lock);
mp_mutex_init(&root->log_file_lock);
mp_cond_init(&root->log_file_wakeup);
struct mp_log dummy = { .root = root };
struct mp_log *log = mp_log_new(root, &dummy, "");
global->log = log;
}
static MP_THREAD_VOID log_file_thread(void *p)
{
struct mp_log_root *root = p;
mp_thread_set_name("log");
mp_mutex_lock(&root->log_file_lock);
while (root->log_file_thread_active) {
struct mp_log_buffer_entry *e =
mp_msg_log_buffer_read(root->log_file_buffer);
if (e) {
mp_mutex_unlock(&root->log_file_lock);
fprintf(root->log_file, "[%8.3f][%c][%s] %s",
mp_time_sec(),
mp_log_levels[e->level][0], e->prefix, e->text);
fflush(root->log_file);
mp_mutex_lock(&root->log_file_lock);
talloc_free(e);
// Multiple threads might be blocked if the log buffer was full.
mp_cond_broadcast(&root->log_file_wakeup);
} else {
mp_cond_wait(&root->log_file_wakeup, &root->log_file_lock);
}
}
mp_mutex_unlock(&root->log_file_lock);
MP_THREAD_RETURN();
}
static void wakeup_log_file(void *p)
{
struct mp_log_root *root = p;
mp_mutex_lock(&root->log_file_lock);
mp_cond_broadcast(&root->log_file_wakeup);
mp_mutex_unlock(&root->log_file_lock);
}
// Only to be called from the main thread.
static void terminate_log_file_thread(struct mp_log_root *root)
{
bool wait_terminate = false;
mp_mutex_lock(&root->log_file_lock);
if (root->log_file_thread_active) {
root->log_file_thread_active = false;
mp_cond_broadcast(&root->log_file_wakeup);
wait_terminate = true;
}
mp_mutex_unlock(&root->log_file_lock);
if (wait_terminate)
mp_thread_join(root->log_file_thread);
mp_msg_log_buffer_destroy(root->log_file_buffer);
root->log_file_buffer = NULL;
if (root->log_file)
fclose(root->log_file);
root->log_file = NULL;
}
// If opt is different from *current_path, update *current_path and return true.
// No lock must be held; passed values must be accessible without.
static bool check_new_path(struct mpv_global *global, char *opt,
char **current_path)
{
void *tmp = talloc_new(NULL);
bool res = false;
char *new_path = mp_get_user_path(tmp, global, opt);
if (!new_path)
new_path = "";
char *old_path = *current_path ? *current_path : "";
if (strcmp(old_path, new_path) != 0) {
talloc_free(*current_path);
*current_path = NULL;
if (new_path[0])
*current_path = talloc_strdup(NULL, new_path);
res = true;
}
talloc_free(tmp);
return res;
}
void mp_msg_update_msglevels(struct mpv_global *global, struct MPOpts *opts)
{
struct mp_log_root *root = global->log->root;
mp_mutex_lock(&root->lock);
root->verbose = opts->verbose;
root->really_quiet = opts->msg_really_quiet;
root->module = opts->msg_module;
root->use_terminal = opts->use_terminal;
root->show_time = opts->msg_time;
if (root->really_quiet)
root->status_lines = 0;
for (int i = STDOUT_FILENO; i <= STDERR_FILENO && root->use_terminal; ++i) {
root->isatty[i] = isatty(i);
root->color[i] = opts->msg_color && root->isatty[i];
}
m_option_type_msglevels.free(&root->msg_levels);
m_option_type_msglevels.copy(NULL, &root->msg_levels, &opts->msg_levels);
atomic_fetch_add(&root->reload_counter, 1);
mp_mutex_unlock(&root->lock);
if (check_new_path(global, opts->log_file, &root->log_path)) {
terminate_log_file_thread(root);
if (root->log_path) {
root->log_file = fopen(root->log_path, "wb");
if (root->log_file) {
// if a logfile is created and the early filebuf still exists,
// flush and destroy the early buffer
mp_mutex_lock(&root->lock);
struct mp_log_buffer *earlybuf = root->early_filebuffer;
if (earlybuf)
root->early_filebuffer = NULL; // but it still logs msgs
mp_mutex_unlock(&root->lock);
if (earlybuf) {
// flush, destroy before creating the normal logfile buf,
// as once the new one is created (specifically, its write
// thread), then MSGL_LOGFILE messages become blocking, but
// the early logfile buf is without dequeue - can deadlock.
// note: timestamp is unknown, we use 0.000 as indication.
// note: new messages while iterating are still flushed.
struct mp_log_buffer_entry *e;
while ((e = mp_msg_log_buffer_read(earlybuf))) {
fprintf(root->log_file, "[%8.3f][%c][%s] %s", 0.0,
mp_log_levels[e->level][0], e->prefix, e->text);
talloc_free(e);
}
mp_msg_log_buffer_destroy(earlybuf); // + remove from root
}
root->log_file_buffer =
mp_msg_log_buffer_new(global, FILE_BUF, MP_LOG_BUFFER_MSGL_LOGFILE,
wakeup_log_file, root);
root->log_file_thread_active = true;
if (mp_thread_create(&root->log_file_thread, log_file_thread,
root))
{
root->log_file_thread_active = false;
terminate_log_file_thread(root);
}
} else {
mp_err(global->log, "Failed to open log file '%s'\n",
root->log_path);
}
}
}
if (check_new_path(global, opts->dump_stats, &root->stats_path)) {
bool open_error = false;
mp_mutex_lock(&root->lock);
if (root->stats_file)
fclose(root->stats_file);
root->stats_file = NULL;
if (root->stats_path) {
root->stats_file = fopen(root->stats_path, "wb");
open_error = !root->stats_file;
}
mp_mutex_unlock(&root->lock);
if (open_error) {
mp_err(global->log, "Failed to open stats file '%s'\n",
root->stats_path);
}
}
}
void mp_msg_force_stderr(struct mpv_global *global, bool force_stderr)
{
struct mp_log_root *root = global->log->root;
mp_mutex_lock(&root->lock);
root->force_stderr = force_stderr;
mp_mutex_unlock(&root->lock);
}
// Only to be called from the main thread.
bool mp_msg_has_log_file(struct mpv_global *global)
{
struct mp_log_root *root = global->log->root;
return !!root->log_file;
}
void mp_msg_uninit(struct mpv_global *global)
{
struct mp_log_root *root = global->log->root;
mp_msg_flush_status_line(global->log, true);
if (root->really_quiet && root->isatty[STDERR_FILENO])
fprintf(stderr, TERM_ESC_RESTORE_CURSOR);
terminate_log_file_thread(root);
mp_msg_log_buffer_destroy(root->early_buffer);
mp_msg_log_buffer_destroy(root->early_filebuffer);
assert(root->num_buffers == 0);
if (root->stats_file)
fclose(root->stats_file);
talloc_free(root->stats_path);
talloc_free(root->log_path);
m_option_type_msglevels.free(&root->msg_levels);
mp_mutex_destroy(&root->lock);
mp_mutex_destroy(&root->log_file_lock);
mp_cond_destroy(&root->log_file_wakeup);
talloc_free(root);
global->log = NULL;
}
// early logging store log messages before they have a known destination.
// there are two early log buffers which are similar logically, and both cease
// function (if still exist, independently) once the log destination is known,
// or mpv init is complete (typically, after all clients/scripts init is done).
//
// - "normal" early_buffer, holds early terminal-level logs, and is handed over
// to the first client which requests such log buffer, so that it sees older
// messages too. further clients which request a log buffer get a new one
// which accumulates messages starting at this point in time.
//
// - early_filebuffer - early log-file messages until a log file name is known.
// main cases where meaningful messages are accumulated before the filename
// is known are when log-file is set at mpv.conf, or from script/client init.
// once a file name is known, the early buffer is flushed and destroyed.
// unlike the "proper" log-file buffer, the early filebuffer is not backed by
// a write thread, and hence non-blocking (can overwrite old messages).
// it's also bigger than the actual file buffer (early: 5000, actual: 100).
static void mp_msg_set_early_logging_raw(struct mpv_global *global, bool enable,
struct mp_log_buffer **root_logbuf,
int size, int level)
{
struct mp_log_root *root = global->log->root;
mp_mutex_lock(&root->lock);
if (enable != !!*root_logbuf) {
if (enable) {
mp_mutex_unlock(&root->lock);
struct mp_log_buffer *buf =
mp_msg_log_buffer_new(global, size, level, NULL, NULL);
mp_mutex_lock(&root->lock);
assert(!*root_logbuf); // no concurrent calls to this function
*root_logbuf = buf;
} else {
struct mp_log_buffer *buf = *root_logbuf;
*root_logbuf = NULL;
mp_mutex_unlock(&root->lock);
mp_msg_log_buffer_destroy(buf);
return;
}
}
mp_mutex_unlock(&root->lock);
}
void mp_msg_set_early_logging(struct mpv_global *global, bool enable)
{
struct mp_log_root *root = global->log->root;
mp_msg_set_early_logging_raw(global, enable, &root->early_buffer,
EARLY_TERM_BUF, MP_LOG_BUFFER_MSGL_TERM);
// normally MSGL_LOGFILE buffer gets a write thread, but not the early buf
mp_msg_set_early_logging_raw(global, enable, &root->early_filebuffer,
EARLY_FILE_BUF, MP_LOG_BUFFER_MSGL_LOGFILE);
}
struct mp_log_buffer *mp_msg_log_buffer_new(struct mpv_global *global,
int size, int level,
void (*wakeup_cb)(void *ctx),
void *wakeup_cb_ctx)
{
struct mp_log_root *root = global->log->root;
mp_mutex_lock(&root->lock);
if (level == MP_LOG_BUFFER_MSGL_TERM) {
// The first thing which creates a terminal-level log buffer gets the
// early log buffer, if it exists. This is supposed to enable a script
// to grab log messages from before it was initialized. It's OK that
// this works only for 1 script and only once.
if (root->early_buffer) {
struct mp_log_buffer *buffer = root->early_buffer;
root->early_buffer = NULL;
mp_msg_log_buffer_resize(buffer, size);
buffer->wakeup_cb = wakeup_cb;
buffer->wakeup_cb_ctx = wakeup_cb_ctx;
mp_mutex_unlock(&root->lock);
return buffer;
}
}
assert(size > 0);
struct mp_log_buffer *buffer = talloc_ptrtype(NULL, buffer);
*buffer = (struct mp_log_buffer) {
.root = root,
.level = level,
.entries = talloc_array(buffer, struct mp_log_buffer_entry *, size),
.capacity = size,
.wakeup_cb = wakeup_cb,
.wakeup_cb_ctx = wakeup_cb_ctx,
};
mp_mutex_init(&buffer->lock);
MP_TARRAY_APPEND(root, root->buffers, root->num_buffers, buffer);
atomic_fetch_add(&root->reload_counter, 1);
mp_mutex_unlock(&root->lock);
return buffer;
}
void mp_msg_log_buffer_resize(struct mp_log_buffer *buffer, int size)
{
mp_mutex_lock(&buffer->lock);
assert(size > 0);
if (buffer->capacity < size &&
buffer->entry0 + buffer->num_entries <= buffer->capacity) {
// shortcut if buffer doesn't wrap
buffer->entries = talloc_realloc(buffer, buffer->entries,
struct mp_log_buffer_entry *, size);
} else if (buffer->capacity != size) {
struct mp_log_buffer_entry **entries =
talloc_array(buffer, struct mp_log_buffer_entry *, size);
int num_entries = 0;
for (int i = buffer->num_entries - 1; i >= 0; i--) {
int entry = (buffer->entry0 + i) % buffer->num_entries;
struct mp_log_buffer_entry *res = buffer->entries[entry];
if (num_entries < size) {
entries[num_entries++] = res;
} else {
talloc_free(res);
buffer->dropped += 1;
}
}
talloc_free(buffer->entries);
buffer->entries = entries;
buffer->entry0 = 0;
buffer->num_entries = num_entries;
}
buffer->capacity = size;
mp_mutex_unlock(&buffer->lock);
}
void mp_msg_log_buffer_set_silent(struct mp_log_buffer *buffer, bool silent)
{
mp_mutex_lock(&buffer->lock);
buffer->silent = silent;
mp_mutex_unlock(&buffer->lock);
}
void mp_msg_log_buffer_destroy(struct mp_log_buffer *buffer)
{
if (!buffer)
return;
struct mp_log_root *root = buffer->root;
mp_mutex_lock(&root->lock);
for (int n = 0; n < root->num_buffers; n++) {
if (root->buffers[n] == buffer) {
MP_TARRAY_REMOVE_AT(root->buffers, root->num_buffers, n);
goto found;
}
}
MP_ASSERT_UNREACHABLE();
found:
while (buffer->num_entries)
talloc_free(log_buffer_read(buffer));
mp_mutex_destroy(&buffer->lock);
talloc_free(buffer);
atomic_fetch_add(&root->reload_counter, 1);
mp_mutex_unlock(&root->lock);
}
// Return a queued message, or if the buffer is empty, NULL.
// Thread-safety: one buffer can be read by a single thread only.
struct mp_log_buffer_entry *mp_msg_log_buffer_read(struct mp_log_buffer *buffer)
{
struct mp_log_buffer_entry *res = NULL;
mp_mutex_lock(&buffer->lock);
if (!buffer->silent && buffer->num_entries) {
if (buffer->dropped) {
res = talloc_ptrtype(NULL, res);
*res = (struct mp_log_buffer_entry) {
.prefix = "overflow",
.level = MSGL_FATAL,
.text = talloc_asprintf(res,
"log message buffer overflow: %"PRId64" messages skipped\n",
buffer->dropped),
};
buffer->dropped = 0;
} else {
res = log_buffer_read(buffer);
}
}
mp_mutex_unlock(&buffer->lock);
return res;
}
// Thread-safety: fully thread-safe, but keep in mind that the lifetime of
// log must be guaranteed during the call.
// Never call this from signal handlers.
void mp_msg(struct mp_log *log, int lev, const char *format, ...)
{
va_list va;
va_start(va, format);
mp_msg_va(log, lev, format, va);
va_end(va);
}
const char *const mp_log_levels[MSGL_MAX + 1] = {
[MSGL_FATAL] = "fatal",
[MSGL_ERR] = "error",
[MSGL_WARN] = "warn",
[MSGL_INFO] = "info",
[MSGL_STATUS] = "status",
[MSGL_V] = "v",
[MSGL_DEBUG] = "debug",
[MSGL_TRACE] = "trace",
[MSGL_STATS] = "stats",
};
const int mp_mpv_log_levels[MSGL_MAX + 1] = {
[MSGL_FATAL] = MPV_LOG_LEVEL_FATAL,
[MSGL_ERR] = MPV_LOG_LEVEL_ERROR,
[MSGL_WARN] = MPV_LOG_LEVEL_WARN,
[MSGL_INFO] = MPV_LOG_LEVEL_INFO,
[MSGL_STATUS] = 0, // never used
[MSGL_V] = MPV_LOG_LEVEL_V,
[MSGL_DEBUG] = MPV_LOG_LEVEL_DEBUG,
[MSGL_TRACE] = MPV_LOG_LEVEL_TRACE,
[MSGL_STATS] = 0, // never used
};
int mp_msg_find_level(const char *s)
{
for (int n = 0; n < MP_ARRAY_SIZE(mp_log_levels); n++) {
if (mp_log_levels[n] && !strcasecmp(s, mp_log_levels[n]))
return n;
}
return -1;
}