mdmon: wait after trying to kill

Now that mdmon handles sigterm if another monitor wants to take over it
should wait until all managed arrays are clean.  So make WaitClean()
available to mdmon and teach try_kill_monitor() to wait on each subarray
in the container.

...since we may be communicating with a dieing process, we need to
block SIGPIPE earlier.

Signed-off-by: Dan Williams <dan.j.williams@intel.com>

diff --git a/Makefile b/Makefile
index f26eeaceb0946d62802999b9505c8fb4f8042d1b..89427774bda00558d5f1f4121799befc6c535afb 100644 (file)
--- a/Makefile
+++ b/Makefile
@@ -85,7 +85,7 @@ SRCS =  mdadm.c config.c mdstat.c  ReadMe.c util.c Manage.c Assemble.c Build.c \
 
 MON_OBJS = mdmon.o monitor.o managemon.o util.o mdstat.o sysfs.o config.o \
        Kill.o sg_io.o dlink.o ReadMe.o super0.o super1.o super-intel.o \
-       super-ddf.o sha1.o crc32.o msg.o
+       super-ddf.o sha1.o crc32.o msg.o Monitor.o
 
 
 STATICSRC = pwgr.c

diff --git a/managemon.c b/managemon.c
index a8af614af0be5304b9349b95eaa80226a59aeb7d..023431276db99ee5b859eda1877cfc52961c7a9f 100644 (file)
--- a/managemon.c
+++ b/managemon.c
@@ -504,13 +504,7 @@ void manage(struct mdstat_ent *mdstat, struct supertype *container)
                        manage_container(mdstat, container);
                        continue;
                }
-               if (mdstat->metadata_version == NULL ||
-                   strncmp(mdstat->metadata_version, "external:", 9) != 0 ||
-                   !is_subarray(mdstat->metadata_version+9) ||
-                   strncmp(mdstat->metadata_version+10, container->devname,
-                           strlen(container->devname)) != 0 ||
-                   mdstat->metadata_version[10+strlen(container->devname)]
-                     != '/')
+               if (!is_container_member(mdstat, container->devname))
                        /* Not for this array */
                        continue;
                /* Looks like a member of this container */

diff --git a/mdmon.c b/mdmon.c
index 7ef0c80f5b38568c7f1ecfda04fe0bcd79bb4ed5..d40adb26706ae732075f591eacd540fc1c99bc3c 100644 (file)
--- a/mdmon.c
+++ b/mdmon.c
@@ -105,11 +105,25 @@ int make_pidfile(char *devname, int o_excl)
        return 0;
 }
 
+int is_container_member(struct mdstat_ent *mdstat, char *container)
+{
+       if (mdstat->metadata_version == NULL ||
+           strncmp(mdstat->metadata_version, "external:", 9) != 0 ||
+           !is_subarray(mdstat->metadata_version+9) ||
+           strncmp(mdstat->metadata_version+10, container, strlen(container)) != 0 ||
+           mdstat->metadata_version[10+strlen(container)] != '/')
+               return 0;
+
+       return 1;
+}
+
+void remove_pidfile(char *devname);
 static void try_kill_monitor(char *devname)
 {
        char buf[100];
        int fd;
        pid_t pid;
+       struct mdstat_ent *mdstat;
 
        sprintf(buf, "/var/run/mdadm/%s.pid", devname);
        fd = open(buf, O_RDONLY);
@@ -135,8 +149,19 @@ static void try_kill_monitor(char *devname)
                return;
        }
 
-       if (strstr(buf, "mdmon") != NULL)
-               kill(pid, SIGTERM);
+       if (!strstr(buf, "mdmon"))
+               return;
+
+       kill(pid, SIGTERM);
+
+       mdstat = mdstat_read(0, 0);
+       for ( ; mdstat; mdstat = mdstat->next)
+               if (is_container_member(mdstat, devname)) {
+                       sprintf(buf, "/dev/%s", mdstat->dev);
+                       WaitClean(buf);
+               }
+       free_mdstat(mdstat);
+       remove_pidfile(devname);
 }
 
 void remove_pidfile(char *devname)
@@ -268,6 +293,26 @@ int main(int argc, char *argv[])
        container->devname = devnum2devname(container->devnum);
        container->device_name = argv[1];
 
+       /* SIGUSR is sent between parent and child.  So both block it
+        * and enable it only with pselect.
+        */
+       sigemptyset(&set);
+       sigaddset(&set, SIGUSR1);
+       sigaddset(&set, SIGHUP);
+       sigaddset(&set, SIGALRM);
+       sigaddset(&set, SIGTERM);
+       sigprocmask(SIG_BLOCK, &set, NULL);
+       act.sa_handler = wake_me;
+       act.sa_flags = 0;
+       sigaction(SIGUSR1, &act, NULL);
+       sigaction(SIGALRM, &act, NULL);
+       act.sa_handler = hup;
+       sigaction(SIGHUP, &act, NULL);
+       act.sa_handler = term;
+       sigaction(SIGTERM, &act, NULL);
+       act.sa_handler = SIG_IGN;
+       sigaction(SIGPIPE, &act, NULL);
+
        /* If this fails, we hope it already exists */
        mkdir("/var/run/mdadm", 0600);
        /* pid file lives in /var/run/mdadm/mdXX.pid */
@@ -364,26 +409,6 @@ int main(int argc, char *argv[])
 
        mlockall(MCL_FUTURE);
 
-       /* SIGUSR is sent between parent and child.  So both block it
-        * and enable it only with pselect.
-        */
-       sigemptyset(&set);
-       sigaddset(&set, SIGUSR1);
-       sigaddset(&set, SIGHUP);
-       sigaddset(&set, SIGALRM);
-       sigaddset(&set, SIGTERM);
-       sigprocmask(SIG_BLOCK, &set, NULL);
-       act.sa_handler = wake_me;
-       act.sa_flags = 0;
-       sigaction(SIGUSR1, &act, NULL);
-       sigaction(SIGALRM, &act, NULL);
-       act.sa_handler = hup;
-       sigaction(SIGHUP, &act, NULL);
-       act.sa_handler = term;
-       sigaction(SIGTERM, &act, NULL);
-       act.sa_handler = SIG_IGN;
-       sigaction(SIGPIPE, &act, NULL);
-
        if (clone_monitor(container) < 0) {
                fprintf(stderr, "mdmon: failed to start monitor process: %s\n",
                        strerror(errno));

diff --git a/mdmon.h b/mdmon.h
index 2ce1fe6924790e53267d6a4be5225d2048973693..8a5cd4a8af6c8948a2605f89d4d16e7b4e8d260e 100644 (file)
--- a/mdmon.h
+++ b/mdmon.h
@@ -54,6 +54,7 @@ extern int sigterm;
 
 int read_dev_state(int fd);
 int get_resync_start(struct active_array *a);
+int is_container_member(struct mdstat_ent *mdstat, char *container);
 
 struct mdstat_ent *mdstat_read(int hold, int start);