On Sat, 27 Feb 2010 16:07:51 +0100
Luca Berra <bluca@xxxxxxxxxx> wrote:
> there are two cases that would result in two mdmon processes for the
> same array
> takeover not detecting pid file in ALT_RUN
> if running mdmon a second time without "--takeover"
>
> Signed-off-by: Luca Berra <bluca@xxxxxxxxxx>
> ---
> mdmon.c | 10 +++++++++-
> msg.c | 2 +-
> util.c | 4 ++--
> 3 files changed, 12 insertions(+), 4 deletions(-)
>
> diff --git a/mdmon.c b/mdmon.c
> index d593ebc..a2b7946 100644
> --- a/mdmon.c
> +++ b/mdmon.c
> @@ -441,7 +441,6 @@ static int mdmon(char *devname, int devnum, int must_fork, int takeover)
> sigaction(SIGPIPE, &act, NULL);
>
> if (takeover) {
> - pid_dir = VAR_RUN;
> victim = mdmon_pid(container->devnum);
> if (victim < 0) {
> pid_dir = ALT_RUN;
Is this just a micro-optimisation? or does it have some real functional
significance?
> @@ -453,6 +452,15 @@ static int mdmon(char *devname, int devnum, int must_fork, int takeover)
>
> ignore = chdir("/");
> if (victim < 0) {
> + pid_dir = ALT_RUN;
> + if (ping_monitor(container->devname) == 0) {
> + fprintf(stderr, "mdmon: %s already managed\n",
> + container->devname);
> + if (!takeover)
> + fprintf(stderr, "\trun mdmon --takeover instead\n");
> + exit(3);
> + }
> + pid_dir = VAR_RUN;
> if (ping_monitor(container->devname) == 0) {
> fprintf(stderr, "mdmon: %s already managed\n",
> container->devname);
I don't understand this. We always know where the pid file is, why do we
have to check again?
> diff --git a/msg.c b/msg.c
> index cc07b96..d2d8445 100644
> --- a/msg.c
> +++ b/msg.c
> @@ -147,7 +147,7 @@ int connect_monitor(char *devname)
> int pos;
> char *c;
>
> - pos = sprintf(path, "%s/", VAR_RUN);
> + pos = sprintf(path, "%s/", pid_dir);
> if (is_subarray(devname)) {
> devname++;
> c = strchr(devname, '/');
This is correct, thanks.
> diff --git a/util.c b/util.c
> index 1def2a0..2fe566d 100644
> --- a/util.c
> +++ b/util.c
> @@ -1475,11 +1475,11 @@ int mdmon_pid(int devnum)
> fd = open(path, O_RDONLY | O_NOATIME, 0);
>
> if (fd < 0)
> - return 0;
> + return -1;
> n = read(fd, pid, 9);
> close(fd);
> if (n <= 0)
> - return 0;
> + return -1;
> return atoi(pid);
> }
>
This is needed too.
For now I'll keep just the last two chunks.
Thanks.
NeilBrown
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
