From: Nick Peng <pymumu@gmail.com>
To: musl@lists.openwall.com
Subject: [musl] BUG: Calling readdir/dirfd after vfork will cause deadlock.
Date: Sat, 25 Jun 2022 11:40:17 +0800 [thread overview]
Message-ID: <CALkM5nvbb27vzH--43Oo1k0EjH68d9fUpEgY8RYAwvTPNkUc4g@mail.gmail.com> (raw)
[-- Attachment #1: Type: text/plain, Size: 2603 bytes --]
Description: After vfork, calling functions such as readdir/dirfd may
cause deadlock. GNU C is OK.
Also tested on x86-64 with musl, no deadlock, but
seems never exit, slower than GNU C.
Version: latest, musl-1.2.3
OS: debian bullseye 64bit OS. and asus router
CPU: raspberrypi aarch64, mips32
Reproduce Code:
#include <dirent.h>
#include <pthread.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <sys/types.h>
#include <sys/wait.h>
#include <unistd.h>
#include <string.h>
pthread_mutex_t lock = PTHREAD_MUTEX_INITIALIZER;
pthread_cond_t cond = PTHREAD_COND_INITIALIZER;
struct tlog_log *logs = NULL;
int do_exit = 0;
void *test(void *arg)
{
int i = 0;
for (i = 0; i < 10000000; i++) {
char *b = malloc(4096);
memset(b, 0, 4096);
free(b);
}
do_exit = 1;
return NULL;
}
void lockfunc()
{
char path_name[4096];
DIR *dir = NULL;
struct dirent *ent;
snprintf(path_name, sizeof(path_name), "/proc/self/fd/");
dir = opendir(path_name);
if (dir == NULL) {
goto errout;
}
while ((ent = readdir(dir)) != NULL) {
}
closedir(dir);
return;
errout:
if (dir) {
closedir(dir);
}
return;
}
void *test_fork(void *arg)
{
int count = 0;
while (do_exit == 0) {
printf("test fork count %d\n", count++);
int pid = vfork();
if (pid < 0) {
return NULL;
} else if (pid == 0) {
lockfunc();
_exit(0);
}
int status;
waitpid(pid, &status, 0);
}
return NULL;
}
int main(int argc, char *argv[])
{
pthread_attr_t attr;
pthread_t threads[10];
pthread_t fork_test;
int i;
int ret;
pthread_attr_init(&attr);
ret = pthread_create(&fork_test, &attr, test_fork, NULL);
for (i = 0; i < 10; i++) {
ret = pthread_create(&threads[i], &attr, test, NULL);
if (ret != 0) {
return 1;
}
}
for (i = 0; i < 10; i++) {
void *retval = NULL;
pthread_join(threads[i], &retval);
}
void *retval = NULL;
pthread_join(fork_test, &retval);
printf("exit\n");
getchar();
return 0;
}
Log:
pi@raspberrypi:~/code/tinylog/test $ ./test
test fork count 0
test fork count 1 <-- lock here
^C
gdb backtrace:
x0000000000409524 in __lock ()
(gdb) bt
#0 0x0000000000409524 in __lock ()
#1 0x0000000000406278 in __libc_malloc_impl ()
[-- Attachment #2: Type: text/html, Size: 3318 bytes --]
next reply other threads:[~2022-06-25 11:09 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-06-25 3:40 Nick Peng [this message]
2022-06-25 12:51 ` Szabolcs Nagy
2022-06-27 4:05 ` Nick Peng
2022-06-27 10:38 ` Laurent Bercot
2022-06-27 14:21 ` Rich Felker
2022-06-27 14:29 ` Markus Wichmann
2022-06-27 7:42 ` Florian Weimer
2022-06-27 8:37 ` Szabolcs Nagy
2022-06-27 9:23 ` Florian Weimer
2022-06-27 17:33 ` Szabolcs Nagy
2022-06-25 12:56 ` Alex Xu (Hello71)
2022-06-27 7:44 ` Florian Weimer
2022-06-27 12:28 ` Alex Xu (Hello71)
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CALkM5nvbb27vzH--43Oo1k0EjH68d9fUpEgY8RYAwvTPNkUc4g@mail.gmail.com \
--to=pymumu@gmail.com \
--cc=musl@lists.openwall.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
Code repositories for project(s) associated with this public inbox
https://git.vuxu.org/mirror/musl/
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).