[linux-2.6-block.git] / Documentation / accounting / getdelays.c

/* getdelays.c
 *
 * Utility to get per-pid and per-tgid delay accounting statistics
 * Also illustrates usage of the taskstats interface
 *
 * Copyright (C) Shailabh Nagar, IBM Corp. 2005
 * Copyright (C) Balbir Singh, IBM Corp. 2006
 * Copyright (c) Jay Lan, SGI. 2006
 *
 * Compile with
 *	gcc -I/usr/src/linux/include getdelays.c -o getdelays
 */

#include <stdio.h>
#include <stdlib.h>
#include <errno.h>
#include <unistd.h>
#include <poll.h>
#include <string.h>
#include <fcntl.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/socket.h>
#include <signal.h>

#include <linux/genetlink.h>
#include <linux/taskstats.h>
#include <linux/cgroupstats.h>

/*
 * Generic macros for dealing with netlink sockets. Might be duplicated
 * elsewhere. It is recommended that commercial grade applications use
 * libnl or libnetlink and use the interfaces provided by the library
 */
#define GENLMSG_DATA(glh)	((void *)(NLMSG_DATA(glh) + GENL_HDRLEN))
#define GENLMSG_PAYLOAD(glh)	(NLMSG_PAYLOAD(glh, 0) - GENL_HDRLEN)
#define NLA_DATA(na)		((void *)((char*)(na) + NLA_HDRLEN))
#define NLA_PAYLOAD(len)	(len - NLA_HDRLEN)

#define err(code, fmt, arg...)			\
	do {					\
		fprintf(stderr, fmt, ##arg);	\
		exit(code);			\
	} while (0)

int done;
int rcvbufsz;
char name[100];
int dbg;
int print_delays;
int print_io_accounting;
int print_task_context_switch_counts;
__u64 stime, utime;

#define PRINTF(fmt, arg...) {			\
	    if (dbg) {				\
		printf(fmt, ##arg);		\
	    }					\
	}

/* Maximum size of response requested or message sent */
#define MAX_MSG_SIZE	1024
/* Maximum number of cpus expected to be specified in a cpumask */
#define MAX_CPUS	32

struct msgtemplate {
	struct nlmsghdr n;
	struct genlmsghdr g;
	char buf[MAX_MSG_SIZE];
};

char cpumask[100+6*MAX_CPUS];

static void usage(void)
{
	fprintf(stderr, "getdelays [-dilv] [-w logfile] [-r bufsize] "
			"[-m cpumask] [-t tgid] [-p pid]\n");
	fprintf(stderr, "  -d: print delayacct stats\n");
	fprintf(stderr, "  -i: print IO accounting (works only with -p)\n");
	fprintf(stderr, "  -l: listen forever\n");
	fprintf(stderr, "  -v: debug on\n");
	fprintf(stderr, "  -C: container path\n");
}

/*
 * Create a raw netlink socket and bind
 */
static int create_nl_socket(int protocol)
{
	int fd;
	struct sockaddr_nl local;

	fd = socket(AF_NETLINK, SOCK_RAW, protocol);
	if (fd < 0)
		return -1;

	if (rcvbufsz)
		if (setsockopt(fd, SOL_SOCKET, SO_RCVBUF,
				&rcvbufsz, sizeof(rcvbufsz)) < 0) {
			fprintf(stderr, "Unable to set socket rcv buf size "
					"to %d\n",
				rcvbufsz);
			return -1;
		}

	memset(&local, 0, sizeof(local));
	local.nl_family = AF_NETLINK;

	if (bind(fd, (struct sockaddr *) &local, sizeof(local)) < 0)
		goto error;

	return fd;
error:
	close(fd);
	return -1;
}


int send_cmd(int sd, __u16 nlmsg_type, __u32 nlmsg_pid,
	     __u8 genl_cmd, __u16 nla_type,
	     void *nla_data, int nla_len)
{
	struct nlattr *na;
	struct sockaddr_nl nladdr;
	int r, buflen;
	char *buf;

	struct msgtemplate msg;

	msg.n.nlmsg_len = NLMSG_LENGTH(GENL_HDRLEN);
	msg.n.nlmsg_type = nlmsg_type;
	msg.n.nlmsg_flags = NLM_F_REQUEST;
	msg.n.nlmsg_seq = 0;
	msg.n.nlmsg_pid = nlmsg_pid;
	msg.g.cmd = genl_cmd;
	msg.g.version = 0x1;
	na = (struct nlattr *) GENLMSG_DATA(&msg);
	na->nla_type = nla_type;
	na->nla_len = nla_len + 1 + NLA_HDRLEN;
	memcpy(NLA_DATA(na), nla_data, nla_len);
	msg.n.nlmsg_len += NLMSG_ALIGN(na->nla_len);

	buf = (char *) &msg;
	buflen = msg.n.nlmsg_len ;
	memset(&nladdr, 0, sizeof(nladdr));
	nladdr.nl_family = AF_NETLINK;
	while ((r = sendto(sd, buf, buflen, 0, (struct sockaddr *) &nladdr,
			   sizeof(nladdr))) < buflen) {
		if (r > 0) {
			buf += r;
			buflen -= r;
		} else if (errno != EAGAIN)
			return -1;
	}
	return 0;
}


/*
 * Probe the controller in genetlink to find the family id
 * for the TASKSTATS family
 */
int get_family_id(int sd)
{
	struct {
		struct nlmsghdr n;
		struct genlmsghdr g;
		char buf[256];
	} ans;

	int id = 0, rc;
	struct nlattr *na;
	int rep_len;

	strcpy(name, TASKSTATS_GENL_NAME);
	rc = send_cmd(sd, GENL_ID_CTRL, getpid(), CTRL_CMD_GETFAMILY,
			CTRL_ATTR_FAMILY_NAME, (void *)name,
			strlen(TASKSTATS_GENL_NAME)+1);

	rep_len = recv(sd, &ans, sizeof(ans), 0);
	if (ans.n.nlmsg_type == NLMSG_ERROR ||
	    (rep_len < 0) || !NLMSG_OK((&ans.n), rep_len))
		return 0;

	na = (struct nlattr *) GENLMSG_DATA(&ans);
	na = (struct nlattr *) ((char *) na + NLA_ALIGN(na->nla_len));
	if (na->nla_type == CTRL_ATTR_FAMILY_ID) {
		id = *(__u16 *) NLA_DATA(na);
	}
	return id;
}

void print_delayacct(struct taskstats *t)
{
	printf("\n\nCPU   %15s%15s%15s%15s\n"
	       "      %15llu%15llu%15llu%15llu\n"
	       "IO    %15s%15s\n"
	       "      %15llu%15llu\n"
	       "SWAP  %15s%15s\n"
	       "      %15llu%15llu\n"
	       "RECLAIM  %12s%15s\n"
	       "      %15llu%15llu\n",
	       "count", "real total", "virtual total", "delay total",
	       (unsigned long long)t->cpu_count,
	       (unsigned long long)t->cpu_run_real_total,
	       (unsigned long long)t->cpu_run_virtual_total,
	       (unsigned long long)t->cpu_delay_total,
	       "count", "delay total",
	       (unsigned long long)t->blkio_count,
	       (unsigned long long)t->blkio_delay_total,
	       "count", "delay total",
	       (unsigned long long)t->swapin_count,
	       (unsigned long long)t->swapin_delay_total,
	       "count", "delay total",
	       (unsigned long long)t->freepages_count,
	       (unsigned long long)t->freepages_delay_total);
}

void task_context_switch_counts(struct taskstats *t)
{
	printf("\n\nTask   %15s%15s\n"
	       "       %15llu%15llu\n",
	       "voluntary", "nonvoluntary",
	       (unsigned long long)t->nvcsw, (unsigned long long)t->nivcsw);
}

void print_cgroupstats(struct cgroupstats *c)
{
	printf("sleeping %llu, blocked %llu, running %llu, stopped %llu, "
		"uninterruptible %llu\n", (unsigned long long)c->nr_sleeping,
		(unsigned long long)c->nr_io_wait,
		(unsigned long long)c->nr_running,
		(unsigned long long)c->nr_stopped,
		(unsigned long long)c->nr_uninterruptible);
}


void print_ioacct(struct taskstats *t)
{
	printf("%s: read=%llu, write=%llu, cancelled_write=%llu\n",
		t->ac_comm,
		(unsigned long long)t->read_bytes,
		(unsigned long long)t->write_bytes,
		(unsigned long long)t->cancelled_write_bytes);
}

int main(int argc, char *argv[])
{
	int c, rc, rep_len, aggr_len, len2, cmd_type;
	__u16 id;
	__u32 mypid;

	struct nlattr *na;
	int nl_sd = -1;
	int len = 0;
	pid_t tid = 0;
	pid_t rtid = 0;

	int fd = 0;
	int count = 0;
	int write_file = 0;
	int maskset = 0;
	char *logfile = NULL;
	int loop = 0;
	int containerset = 0;
	char containerpath[1024];
	int cfd = 0;

	struct msgtemplate msg;

	while (1) {
		c = getopt(argc, argv, "qdiw:r:m:t:p:vlC:");
		if (c < 0)
			break;

		switch (c) {
		case 'd':
			printf("print delayacct stats ON\n");
			print_delays = 1;
			break;
		case 'i':
			printf("printing IO accounting\n");
			print_io_accounting = 1;
			break;
		case 'q':
			printf("printing task/process context switch rates\n");
			print_task_context_switch_counts = 1;
			break;
		case 'C':
			containerset = 1;
			strncpy(containerpath, optarg, strlen(optarg) + 1);
			break;
		case 'w':
			logfile = strdup(optarg);
			printf("write to file %s\n", logfile);
			write_file = 1;
			break;
		case 'r':
			rcvbufsz = atoi(optarg);
			printf("receive buf size %d\n", rcvbufsz);
			if (rcvbufsz < 0)
				err(1, "Invalid rcv buf size\n");
			break;
		case 'm':
			strncpy(cpumask, optarg, sizeof(cpumask));
			maskset = 1;
			printf("cpumask %s maskset %d\n", cpumask, maskset);
			break;
		case 't':
			tid = atoi(optarg);
			if (!tid)
				err(1, "Invalid tgid\n");
			cmd_type = TASKSTATS_CMD_ATTR_TGID;
			break;
		case 'p':
			tid = atoi(optarg);
			if (!tid)
				err(1, "Invalid pid\n");
			cmd_type = TASKSTATS_CMD_ATTR_PID;
			break;
		case 'v':
			printf("debug on\n");
			dbg = 1;
			break;
		case 'l':
			printf("listen forever\n");
			loop = 1;
			break;
		default:
			usage();
			exit(-1);
		}
	}

	if (write_file) {
		fd = open(logfile, O_WRONLY | O_CREAT | O_TRUNC,
			  S_IRUSR | S_IWUSR | S_IRGRP | S_IROTH);
		if (fd == -1) {
			perror("Cannot open output file\n");
			exit(1);
		}
	}

	if ((nl_sd = create_nl_socket(NETLINK_GENERIC)) < 0)
		err(1, "error creating Netlink socket\n");


	mypid = getpid();
	id = get_family_id(nl_sd);
	if (!id) {
		fprintf(stderr, "Error getting family id, errno %d\n", errno);
		goto err;
	}
	PRINTF("family id %d\n", id);

	if (maskset) {
		rc = send_cmd(nl_sd, id, mypid, TASKSTATS_CMD_GET,
			      TASKSTATS_CMD_ATTR_REGISTER_CPUMASK,
			      &cpumask, strlen(cpumask) + 1);
		PRINTF("Sent register cpumask, retval %d\n", rc);
		if (rc < 0) {
			fprintf(stderr, "error sending register cpumask\n");
			goto err;
		}
	}

	if (tid && containerset) {
		fprintf(stderr, "Select either -t or -C, not both\n");
		goto err;
	}

	if (tid) {
		rc = send_cmd(nl_sd, id, mypid, TASKSTATS_CMD_GET,
			      cmd_type, &tid, sizeof(__u32));
		PRINTF("Sent pid/tgid, retval %d\n", rc);
		if (rc < 0) {
			fprintf(stderr, "error sending tid/tgid cmd\n");
			goto done;
		}
	}

	if (containerset) {
		cfd = open(containerpath, O_RDONLY);
		if (cfd < 0) {
			perror("error opening container file");
			goto err;
		}
		rc = send_cmd(nl_sd, id, mypid, CGROUPSTATS_CMD_GET,
			      CGROUPSTATS_CMD_ATTR_FD, &cfd, sizeof(__u32));
		if (rc < 0) {
			perror("error sending cgroupstats command");
			goto err;
		}
	}

	do {
		int i;

		rep_len = recv(nl_sd, &msg, sizeof(msg), 0);
		PRINTF("received %d bytes\n", rep_len);

		if (rep_len < 0) {
			fprintf(stderr, "nonfatal reply error: errno %d\n",
				errno);
			continue;
		}
		if (msg.n.nlmsg_type == NLMSG_ERROR ||
		    !NLMSG_OK((&msg.n), rep_len)) {
			struct nlmsgerr *err = NLMSG_DATA(&msg);
			fprintf(stderr, "fatal reply error,  errno %d\n",
				err->error);
			goto done;
		}

		PRINTF("nlmsghdr size=%zu, nlmsg_len=%d, rep_len=%d\n",
		       sizeof(struct nlmsghdr), msg.n.nlmsg_len, rep_len);


		rep_len = GENLMSG_PAYLOAD(&msg.n);

		na = (struct nlattr *) GENLMSG_DATA(&msg);
		len = 0;
		i = 0;
		while (len < rep_len) {
			len += NLA_ALIGN(na->nla_len);
			switch (na->nla_type) {
			case TASKSTATS_TYPE_AGGR_TGID:
				/* Fall through */
			case TASKSTATS_TYPE_AGGR_PID:
				aggr_len = NLA_PAYLOAD(na->nla_len);
				len2 = 0;
				/* For nested attributes, na follows */
				na = (struct nlattr *) NLA_DATA(na);
				done = 0;
				while (len2 < aggr_len) {
					switch (na->nla_type) {
					case TASKSTATS_TYPE_PID:
						rtid = *(int *) NLA_DATA(na);
						if (print_delays)
							printf("PID\t%d\n", rtid);
						break;
					case TASKSTATS_TYPE_TGID:
						rtid = *(int *) NLA_DATA(na);
						if (print_delays)
							printf("TGID\t%d\n", rtid);
						break;
					case TASKSTATS_TYPE_STATS:
						count++;
						if (print_delays)
							print_delayacct((struct taskstats *) NLA_DATA(na));
						if (print_io_accounting)
							print_ioacct((struct taskstats *) NLA_DATA(na));
						if (print_task_context_switch_counts)
							task_context_switch_counts((struct taskstats *) NLA_DATA(na));
						if (fd) {
							if (write(fd, NLA_DATA(na), na->nla_len) < 0) {
								err(1,"write error\n");
							}
						}
						if (!loop)
							goto done;
						break;
					default:
						fprintf(stderr, "Unknown nested"
							" nla_type %d\n",
							na->nla_type);
						break;
					}
					len2 += NLA_ALIGN(na->nla_len);
					na = (struct nlattr *) ((char *) na + len2);
				}
				break;

			case CGROUPSTATS_TYPE_CGROUP_STATS:
				print_cgroupstats(NLA_DATA(na));
				break;
			default:
				fprintf(stderr, "Unknown nla_type %d\n",
					na->nla_type);
				break;
			}
			na = (struct nlattr *) (GENLMSG_DATA(&msg) + len);
		}
	} while (loop);
done:
	if (maskset) {
		rc = send_cmd(nl_sd, id, mypid, TASKSTATS_CMD_GET,
			      TASKSTATS_CMD_ATTR_DEREGISTER_CPUMASK,
			      &cpumask, strlen(cpumask) + 1);
		printf("Sent deregister mask, retval %d\n", rc);
		if (rc < 0)
			err(rc, "error sending deregister cpumask\n");
	}
err:
	close(nl_sd);
	if (fd)
		close(fd);
	if (cfd)
		close(cfd);
	return 0;
}
Commit	Line	Data
a3baf649 SN	1	/* getdelays.c
	2	*
	3	* Utility to get per-pid and per-tgid delay accounting statistics
	4	* Also illustrates usage of the taskstats interface
	5	*
	6	* Copyright (C) Shailabh Nagar, IBM Corp. 2005
	7	* Copyright (C) Balbir Singh, IBM Corp. 2006
9e06d3f9	8	* Copyright (c) Jay Lan, SGI. 2006
a3baf649	9	*
d2f7bf13 AM	10	* Compile with
d2f7bf13 AM	11	* gcc -I/usr/src/linux/include getdelays.c -o getdelays
a3baf649 SN	12	*/
	13
	14	#include <stdio.h>
	15	#include <stdlib.h>
	16	#include <errno.h>
	17	#include <unistd.h>
	18	#include <poll.h>
	19	#include <string.h>
	20	#include <fcntl.h>
	21	#include <sys/types.h>
	22	#include <sys/stat.h>
	23	#include <sys/socket.h>
a3baf649 SN	24	#include <signal.h>
	25
	26	#include <linux/genetlink.h>
	27	#include <linux/taskstats.h>
546040dc	28	#include <linux/cgroupstats.h>
a3baf649 SN	29
	30	/*
	31	* Generic macros for dealing with netlink sockets. Might be duplicated
	32	* elsewhere. It is recommended that commercial grade applications use
	33	* libnl or libnetlink and use the interfaces provided by the library
	34	*/
	35	#define GENLMSG_DATA(glh) ((void *)(NLMSG_DATA(glh) + GENL_HDRLEN))
	36	#define GENLMSG_PAYLOAD(glh) (NLMSG_PAYLOAD(glh, 0) - GENL_HDRLEN)
	37	#define NLA_DATA(na) ((void )((char)(na) + NLA_HDRLEN))
	38	#define NLA_PAYLOAD(len) (len - NLA_HDRLEN)
	39
d2f7bf13 AM	40	#define err(code, fmt, arg...) \
	41	do { \
	42	fprintf(stderr, fmt, ##arg); \
	43	exit(code); \
	44	} while (0)
	45
	46	int done;
	47	int rcvbufsz;
	48	char name[100];
	49	int dbg;
	50	int print_delays;
cf709844	51	int print_io_accounting;
b663a79c	52	int print_task_context_switch_counts;
9e06d3f9	53	__u64 stime, utime;
d2f7bf13	54
9e06d3f9 SN	55	#define PRINTF(fmt, arg...) { \
	56	if (dbg) { \
	57	printf(fmt, ##arg); \
	58	} \
	59	}
	60
	61	/* Maximum size of response requested or message sent */
88040230	62	#define MAX_MSG_SIZE 1024
9e06d3f9 SN	63	/* Maximum number of cpus expected to be specified in a cpumask */
9e06d3f9 SN	64	#define MAX_CPUS 32
9e06d3f9 SN	65
	66	struct msgtemplate {
	67	struct nlmsghdr n;
	68	struct genlmsghdr g;
	69	char buf[MAX_MSG_SIZE];
	70	};
	71
	72	char cpumask[100+6*MAX_CPUS];
a3baf649	73
f16825bb RD	74	static void usage(void)
	75	{
	76	fprintf(stderr, "getdelays [-dilv] [-w logfile] [-r bufsize] "
	77	"[-m cpumask] [-t tgid] [-p pid]\n");
	78	fprintf(stderr, " -d: print delayacct stats\n");
	79	fprintf(stderr, " -i: print IO accounting (works only with -p)\n");
	80	fprintf(stderr, " -l: listen forever\n");
	81	fprintf(stderr, " -v: debug on\n");
546040dc	82	fprintf(stderr, " -C: container path\n");
f16825bb RD	83	}
f16825bb RD	84
a3baf649 SN	85	/*
	86	* Create a raw netlink socket and bind
	87	*/
9e06d3f9	88	static int create_nl_socket(int protocol)
a3baf649	89	{
9e06d3f9 SN	90	int fd;
	91	struct sockaddr_nl local;
	92
	93	fd = socket(AF_NETLINK, SOCK_RAW, protocol);
	94	if (fd < 0)
	95	return -1;
	96
	97	if (rcvbufsz)
	98	if (setsockopt(fd, SOL_SOCKET, SO_RCVBUF,
	99	&rcvbufsz, sizeof(rcvbufsz)) < 0) {
d2f7bf13 AM	100	fprintf(stderr, "Unable to set socket rcv buf size "
	101	"to %d\n",
	102	rcvbufsz);
9e06d3f9 SN	103	return -1;
9e06d3f9 SN	104	}
a3baf649	105
9e06d3f9 SN	106	memset(&local, 0, sizeof(local));
9e06d3f9 SN	107	local.nl_family = AF_NETLINK;
a3baf649	108
9e06d3f9 SN	109	if (bind(fd, (struct sockaddr *) &local, sizeof(local)) < 0)
9e06d3f9 SN	110	goto error;
a3baf649	111
9e06d3f9 SN	112	return fd;
	113	error:
	114	close(fd);
	115	return -1;
a3baf649 SN	116	}
a3baf649 SN	117
9e06d3f9 SN	118
	119	int send_cmd(int sd, __u16 nlmsg_type, __u32 nlmsg_pid,
	120	__u8 genl_cmd, __u16 nla_type,
	121	void *nla_data, int nla_len)
a3baf649	122	{
9e06d3f9 SN	123	struct nlattr *na;
	124	struct sockaddr_nl nladdr;
	125	int r, buflen;
	126	char *buf;
	127
	128	struct msgtemplate msg;
	129
	130	msg.n.nlmsg_len = NLMSG_LENGTH(GENL_HDRLEN);
	131	msg.n.nlmsg_type = nlmsg_type;
	132	msg.n.nlmsg_flags = NLM_F_REQUEST;
	133	msg.n.nlmsg_seq = 0;
	134	msg.n.nlmsg_pid = nlmsg_pid;
	135	msg.g.cmd = genl_cmd;
	136	msg.g.version = 0x1;
	137	na = (struct nlattr *) GENLMSG_DATA(&msg);
	138	na->nla_type = nla_type;
	139	na->nla_len = nla_len + 1 + NLA_HDRLEN;
	140	memcpy(NLA_DATA(na), nla_data, nla_len);
	141	msg.n.nlmsg_len += NLMSG_ALIGN(na->nla_len);
	142
	143	buf = (char *) &msg;
	144	buflen = msg.n.nlmsg_len ;
	145	memset(&nladdr, 0, sizeof(nladdr));
	146	nladdr.nl_family = AF_NETLINK;
	147	while ((r = sendto(sd, buf, buflen, 0, (struct sockaddr *) &nladdr,
	148	sizeof(nladdr))) < buflen) {
	149	if (r > 0) {
	150	buf += r;
	151	buflen -= r;
	152	} else if (errno != EAGAIN)
	153	return -1;
	154	}
	155	return 0;
a3baf649 SN	156	}
a3baf649 SN	157
9e06d3f9	158
a3baf649 SN	159	/*
	160	* Probe the controller in genetlink to find the family id
	161	* for the TASKSTATS family
	162	*/
	163	int get_family_id(int sd)
	164	{
9e06d3f9 SN	165	struct {
	166	struct nlmsghdr n;
	167	struct genlmsghdr g;
	168	char buf[256];
	169	} ans;
	170
10e6f32b	171	int id = 0, rc;
9e06d3f9 SN	172	struct nlattr *na;
	173	int rep_len;
	174
	175	strcpy(name, TASKSTATS_GENL_NAME);
	176	rc = send_cmd(sd, GENL_ID_CTRL, getpid(), CTRL_CMD_GETFAMILY,
	177	CTRL_ATTR_FAMILY_NAME, (void *)name,
	178	strlen(TASKSTATS_GENL_NAME)+1);
	179
	180	rep_len = recv(sd, &ans, sizeof(ans), 0);
	181	if (ans.n.nlmsg_type == NLMSG_ERROR \|\|
	182	(rep_len < 0) \|\| !NLMSG_OK((&ans.n), rep_len))
	183	return 0;
a3baf649	184
9e06d3f9 SN	185	na = (struct nlattr *) GENLMSG_DATA(&ans);
	186	na = (struct nlattr ) ((char ) na + NLA_ALIGN(na->nla_len));
	187	if (na->nla_type == CTRL_ATTR_FAMILY_ID) {
	188	id = (__u16 ) NLA_DATA(na);
	189	}
	190	return id;
a3baf649 SN	191	}
a3baf649 SN	192
9e06d3f9	193	void print_delayacct(struct taskstats *t)
a3baf649	194	{
9e06d3f9 SN	195	printf("\n\nCPU %15s%15s%15s%15s\n"
	196	" %15llu%15llu%15llu%15llu\n"
	197	"IO %15s%15s\n"
	198	" %15llu%15llu\n"
9b0975a2 KK	199	"SWAP %15s%15s\n"
	200	" %15llu%15llu\n"
	201	"RECLAIM %12s%15s\n"
efbee7f1	202	" %15llu%15llu\n",
9e06d3f9	203	"count", "real total", "virtual total", "delay total",
66659313 RD	204	(unsigned long long)t->cpu_count,
	205	(unsigned long long)t->cpu_run_real_total,
	206	(unsigned long long)t->cpu_run_virtual_total,
	207	(unsigned long long)t->cpu_delay_total,
9e06d3f9	208	"count", "delay total",
66659313 RD	209	(unsigned long long)t->blkio_count,
66659313 RD	210	(unsigned long long)t->blkio_delay_total,
9b0975a2	211	"count", "delay total",
66659313 RD	212	(unsigned long long)t->swapin_count,
	213	(unsigned long long)t->swapin_delay_total,
	214	"count", "delay total",
	215	(unsigned long long)t->freepages_count,
	216	(unsigned long long)t->freepages_delay_total);
a3baf649 SN	217	}
a3baf649 SN	218
b663a79c MU	219	void task_context_switch_counts(struct taskstats *t)
	220	{
	221	printf("\n\nTask %15s%15s\n"
10e6f32b	222	" %15llu%15llu\n",
b663a79c	223	"voluntary", "nonvoluntary",
66659313	224	(unsigned long long)t->nvcsw, (unsigned long long)t->nivcsw);
b663a79c MU	225	}
b663a79c MU	226
546040dc BS	227	void print_cgroupstats(struct cgroupstats *c)
	228	{
	229	printf("sleeping %llu, blocked %llu, running %llu, stopped %llu, "
66659313 RD	230	"uninterruptible %llu\n", (unsigned long long)c->nr_sleeping,
	231	(unsigned long long)c->nr_io_wait,
	232	(unsigned long long)c->nr_running,
	233	(unsigned long long)c->nr_stopped,
	234	(unsigned long long)c->nr_uninterruptible);
546040dc BS	235	}
	236
	237
cf709844 AM	238	void print_ioacct(struct taskstats *t)
	239	{
	240	printf("%s: read=%llu, write=%llu, cancelled_write=%llu\n",
	241	t->ac_comm,
	242	(unsigned long long)t->read_bytes,
	243	(unsigned long long)t->write_bytes,
	244	(unsigned long long)t->cancelled_write_bytes);
	245	}
	246
a3baf649 SN	247	int main(int argc, char *argv[])
a3baf649 SN	248	{
9e06d3f9 SN	249	int c, rc, rep_len, aggr_len, len2, cmd_type;
	250	__u16 id;
	251	__u32 mypid;
	252
	253	struct nlattr *na;
	254	int nl_sd = -1;
	255	int len = 0;
	256	pid_t tid = 0;
	257	pid_t rtid = 0;
	258
	259	int fd = 0;
	260	int count = 0;
	261	int write_file = 0;
	262	int maskset = 0;
7f76c403	263	char *logfile = NULL;
9e06d3f9	264	int loop = 0;
546040dc BS	265	int containerset = 0;
	266	char containerpath[1024];
	267	int cfd = 0;
9e06d3f9 SN	268
	269	struct msgtemplate msg;
	270
	271	while (1) {
546040dc	272	c = getopt(argc, argv, "qdiw:r:m:t:p:vlC:");
9e06d3f9 SN	273	if (c < 0)
9e06d3f9 SN	274	break;
a3baf649	275
9e06d3f9 SN	276	switch (c) {
	277	case 'd':
	278	printf("print delayacct stats ON\n");
	279	print_delays = 1;
	280	break;
cf709844 AM	281	case 'i':
	282	printf("printing IO accounting\n");
	283	print_io_accounting = 1;
	284	break;
b663a79c MU	285	case 'q':
	286	printf("printing task/process context switch rates\n");
	287	print_task_context_switch_counts = 1;
	288	break;
546040dc BS	289	case 'C':
	290	containerset = 1;
	291	strncpy(containerpath, optarg, strlen(optarg) + 1);
	292	break;
9e06d3f9	293	case 'w':
7f76c403	294	logfile = strdup(optarg);
9e06d3f9 SN	295	printf("write to file %s\n", logfile);
	296	write_file = 1;
	297	break;
	298	case 'r':
	299	rcvbufsz = atoi(optarg);
	300	printf("receive buf size %d\n", rcvbufsz);
	301	if (rcvbufsz < 0)
	302	err(1, "Invalid rcv buf size\n");
	303	break;
	304	case 'm':
	305	strncpy(cpumask, optarg, sizeof(cpumask));
	306	maskset = 1;
	307	printf("cpumask %s maskset %d\n", cpumask, maskset);
	308	break;
	309	case 't':
	310	tid = atoi(optarg);
	311	if (!tid)
	312	err(1, "Invalid tgid\n");
	313	cmd_type = TASKSTATS_CMD_ATTR_TGID;
9e06d3f9 SN	314	break;
	315	case 'p':
	316	tid = atoi(optarg);
	317	if (!tid)
	318	err(1, "Invalid pid\n");
	319	cmd_type = TASKSTATS_CMD_ATTR_PID;
9e06d3f9 SN	320	break;
	321	case 'v':
	322	printf("debug on\n");
	323	dbg = 1;
	324	break;
	325	case 'l':
	326	printf("listen forever\n");
	327	loop = 1;
	328	break;
	329	default:
f16825bb	330	usage();
9e06d3f9	331	exit(-1);
a3baf649	332	}
a3baf649	333	}
a3baf649	334
9e06d3f9 SN	335	if (write_file) {
	336	fd = open(logfile, O_WRONLY \| O_CREAT \| O_TRUNC,
	337	S_IRUSR \| S_IWUSR \| S_IRGRP \| S_IROTH);
	338	if (fd == -1) {
	339	perror("Cannot open output file\n");
	340	exit(1);
	341	}
	342	}
a3baf649	343
9e06d3f9 SN	344	if ((nl_sd = create_nl_socket(NETLINK_GENERIC)) < 0)
9e06d3f9 SN	345	err(1, "error creating Netlink socket\n");
a3baf649	346
a3baf649	347
9e06d3f9 SN	348	mypid = getpid();
	349	id = get_family_id(nl_sd);
	350	if (!id) {
d2f7bf13	351	fprintf(stderr, "Error getting family id, errno %d\n", errno);
9e06d3f9	352	goto err;
a3baf649	353	}
9e06d3f9 SN	354	PRINTF("family id %d\n", id);
	355
	356	if (maskset) {
	357	rc = send_cmd(nl_sd, id, mypid, TASKSTATS_CMD_GET,
	358	TASKSTATS_CMD_ATTR_REGISTER_CPUMASK,
7d1bdca9	359	&cpumask, strlen(cpumask) + 1);
9e06d3f9 SN	360	PRINTF("Sent register cpumask, retval %d\n", rc);
9e06d3f9 SN	361	if (rc < 0) {
d2f7bf13	362	fprintf(stderr, "error sending register cpumask\n");
9e06d3f9 SN	363	goto err;
9e06d3f9 SN	364	}
a3baf649 SN	365	}
a3baf649 SN	366
546040dc BS	367	if (tid && containerset) {
	368	fprintf(stderr, "Select either -t or -C, not both\n");
	369	goto err;
	370	}
	371
9e06d3f9 SN	372	if (tid) {
	373	rc = send_cmd(nl_sd, id, mypid, TASKSTATS_CMD_GET,
	374	cmd_type, &tid, sizeof(__u32));
	375	PRINTF("Sent pid/tgid, retval %d\n", rc);
	376	if (rc < 0) {
d2f7bf13	377	fprintf(stderr, "error sending tid/tgid cmd\n");
9e06d3f9 SN	378	goto done;
9e06d3f9 SN	379	}
a3baf649 SN	380	}
a3baf649 SN	381
546040dc BS	382	if (containerset) {
	383	cfd = open(containerpath, O_RDONLY);
	384	if (cfd < 0) {
	385	perror("error opening container file");
	386	goto err;
	387	}
	388	rc = send_cmd(nl_sd, id, mypid, CGROUPSTATS_CMD_GET,
	389	CGROUPSTATS_CMD_ATTR_FD, &cfd, sizeof(__u32));
	390	if (rc < 0) {
	391	perror("error sending cgroupstats command");
	392	goto err;
	393	}
	394	}
	395
9e06d3f9 SN	396	do {
9e06d3f9 SN	397	int i;
a3baf649	398
9e06d3f9 SN	399	rep_len = recv(nl_sd, &msg, sizeof(msg), 0);
9e06d3f9 SN	400	PRINTF("received %d bytes\n", rep_len);
a3baf649	401
9e06d3f9	402	if (rep_len < 0) {
d2f7bf13 AM	403	fprintf(stderr, "nonfatal reply error: errno %d\n",
d2f7bf13 AM	404	errno);
9e06d3f9 SN	405	continue;
	406	}
	407	if (msg.n.nlmsg_type == NLMSG_ERROR \|\|
	408	!NLMSG_OK((&msg.n), rep_len)) {
7d1bdca9	409	struct nlmsgerr *err = NLMSG_DATA(&msg);
d2f7bf13 AM	410	fprintf(stderr, "fatal reply error, errno %d\n",
d2f7bf13 AM	411	err->error);
9e06d3f9 SN	412	goto done;
	413	}
	414
10e6f32b	415	PRINTF("nlmsghdr size=%zu, nlmsg_len=%d, rep_len=%d\n",
9e06d3f9 SN	416	sizeof(struct nlmsghdr), msg.n.nlmsg_len, rep_len);
	417
	418
	419	rep_len = GENLMSG_PAYLOAD(&msg.n);
	420
	421	na = (struct nlattr *) GENLMSG_DATA(&msg);
	422	len = 0;
	423	i = 0;
	424	while (len < rep_len) {
	425	len += NLA_ALIGN(na->nla_len);
	426	switch (na->nla_type) {
	427	case TASKSTATS_TYPE_AGGR_TGID:
	428	/* Fall through */
	429	case TASKSTATS_TYPE_AGGR_PID:
	430	aggr_len = NLA_PAYLOAD(na->nla_len);
	431	len2 = 0;
	432	/* For nested attributes, na follows */
	433	na = (struct nlattr *) NLA_DATA(na);
	434	done = 0;
	435	while (len2 < aggr_len) {
	436	switch (na->nla_type) {
	437	case TASKSTATS_TYPE_PID:
	438	rtid = (int ) NLA_DATA(na);
	439	if (print_delays)
	440	printf("PID\t%d\n", rtid);
	441	break;
	442	case TASKSTATS_TYPE_TGID:
	443	rtid = (int ) NLA_DATA(na);
	444	if (print_delays)
	445	printf("TGID\t%d\n", rtid);
	446	break;
	447	case TASKSTATS_TYPE_STATS:
	448	count++;
	449	if (print_delays)
	450	print_delayacct((struct taskstats *) NLA_DATA(na));
cf709844 AM	451	if (print_io_accounting)
cf709844 AM	452	print_ioacct((struct taskstats *) NLA_DATA(na));
b663a79c MU	453	if (print_task_context_switch_counts)
b663a79c MU	454	task_context_switch_counts((struct taskstats *) NLA_DATA(na));
9e06d3f9 SN	455	if (fd) {
	456	if (write(fd, NLA_DATA(na), na->nla_len) < 0) {
	457	err(1,"write error\n");
	458	}
	459	}
	460	if (!loop)
	461	goto done;
	462	break;
	463	default:
d2f7bf13 AM	464	fprintf(stderr, "Unknown nested"
	465	" nla_type %d\n",
	466	na->nla_type);
9e06d3f9 SN	467	break;
	468	}
	469	len2 += NLA_ALIGN(na->nla_len);
	470	na = (struct nlattr ) ((char ) na + len2);
	471	}
	472	break;
	473
546040dc BS	474	case CGROUPSTATS_TYPE_CGROUP_STATS:
	475	print_cgroupstats(NLA_DATA(na));
	476	break;
9e06d3f9	477	default:
d2f7bf13 AM	478	fprintf(stderr, "Unknown nla_type %d\n",
d2f7bf13 AM	479	na->nla_type);
9e06d3f9	480	break;
a3baf649	481	}
9e06d3f9	482	na = (struct nlattr *) (GENLMSG_DATA(&msg) + len);
a3baf649	483	}
9e06d3f9 SN	484	} while (loop);
	485	done:
	486	if (maskset) {
	487	rc = send_cmd(nl_sd, id, mypid, TASKSTATS_CMD_GET,
	488	TASKSTATS_CMD_ATTR_DEREGISTER_CPUMASK,
7d1bdca9	489	&cpumask, strlen(cpumask) + 1);
9e06d3f9 SN	490	printf("Sent deregister mask, retval %d\n", rc);
	491	if (rc < 0)
	492	err(rc, "error sending deregister cpumask\n");
a3baf649	493	}
9e06d3f9 SN	494	err:
	495	close(nl_sd);
	496	if (fd)
	497	close(fd);
546040dc BS	498	if (cfd)
546040dc BS	499	close(cfd);
9e06d3f9	500	return 0;
a3baf649	501	}