emo/busybox
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

uevent: make it survive ENOBUFS

Browse Source 
function                                             old     new   delta
uevent_main                                          285     337     +52

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
This commit is contained in:
Denys Vlasenko 2020-11-22 16:22:16 +01:00
parent 1c9cd9b71b
commit cb143b9793
1 changed files with 21 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

33
util-linux/uevent.c
View File

@ -30,19 +30,19 @@
#define INIT_G() do { setup_common_bufsiz(); } while (0) #define INIT_G() do { setup_common_bufsiz(); } while (0)
enum { enum {
MAX_ENV = COMMON_BUFSIZE / sizeof(char*) - 1, MAX_ENV = COMMON_BUFSIZE / sizeof(char*) - 1,
/* sizeof(env[0]) instead of sizeof(char*) // ^^^sizeof(env[0]) instead of sizeof(char*)
* makes gcc-6.3.0 emit "strict-aliasing" warning. // makes gcc-6.3.0 emit "strict-aliasing" warning.
*/
};
enum { // socket receive buffer of 2MiB proved to be too small:
/* socket receive buffer of 2MiB proved to be too small: // http://lists.busybox.net/pipermail/busybox/2019-December/087665.html
* http://lists.busybox.net/pipermail/busybox/2019-December/087665.html // udevd seems to use a whooping 128MiB.
* Udevd seems to use a whooping 128MiB. // The socket receive buffer size is just a resource limit.
* The socket receive buffer size is just a resource limit. // The buffers are allocated lazily so the memory is not wasted.
* The buffers are allocated lazily so the memory is not wasted.
*/
KERN_RCVBUF = 128 * 1024 * 1024, KERN_RCVBUF = 128 * 1024 * 1024,
// Might be made smaller: the kernel v5.4 passes up to 32 environment
// variables with a total of 2kb on each event.
// On top of that the action string and device path are added.
USER_RCVBUF = 16 * 1024, USER_RCVBUF = 16 * 1024,
}; };
@ -61,6 +61,7 @@ int uevent_main(int argc UNUSED_PARAM, char **argv)
// Reproducer: // Reproducer:
// uevent mdev & // uevent mdev &
// find /sys -name uevent -exec sh -c 'echo add >"{}"' ';' // find /sys -name uevent -exec sh -c 'echo add >"{}"' ';'
reopen:
fd = create_and_bind_to_netlink(NETLINK_KOBJECT_UEVENT, /*groups:*/ 1 << 0, KERN_RCVBUF); fd = create_and_bind_to_netlink(NETLINK_KOBJECT_UEVENT, /*groups:*/ 1 << 0, KERN_RCVBUF);
for (;;) { for (;;) {
@ -82,8 +83,16 @@ int uevent_main(int argc UNUSED_PARAM, char **argv)
// Here we block, possibly for a very long time // Here we block, possibly for a very long time
len = safe_read(fd, netbuf, USER_RCVBUF - 1); len = safe_read(fd, netbuf, USER_RCVBUF - 1);
if (len < 0) if (len < 0) {
if (errno == ENOBUFS) {
// Ran out of socket receive buffer
bb_simple_error_msg("uevent overrun");
close(fd);
munmap(netbuf, USER_RCVBUF);
goto reopen;
}
bb_simple_perror_msg_and_die("read"); bb_simple_perror_msg_and_die("read");
}
end = netbuf + len; end = netbuf + len;
*end = '\0'; *end = '\0';