microsoft
/
WSL2-Linux-Kernel
зеркало из https://github.com/microsoft/WSL2-Linux-Kernel.git
1
0
Форкнуть 0
Код Задачи Пакеты Проекты Релизы Вики Активность

[PATCH] per task delay accounting taskstats interface: documentation fix 

Change documentation and example program to reflect the flow control issues
being addressed by the cpumask changes.

Signed-off-by: Shailabh Nagar <nagar@watson.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This commit is contained in:
Shailabh Nagar 2006-07-14 00:24:45 -07:00 коммит произвёл Linus Torvalds
Родитель ad4ecbcba7
Коммит 9e06d3f9f6
2 изменённых файлов: 361 добавлений и 301 удалений
Показать статистику Скачать Patch файл Скачать Diff файл Показать все файлы Свернуть все файлы

384
Documentation/accounting/getdelays.c
Просмотреть файл

@ -5,6 +5,7 @@
* *
* Copyright (C) Shailabh Nagar, IBM Corp. 2005 * Copyright (C) Shailabh Nagar, IBM Corp. 2005
* Copyright (C) Balbir Singh, IBM Corp. 2006 * Copyright (C) Balbir Singh, IBM Corp. 2006
* Copyright (c) Jay Lan, SGI. 2006
* *
*/ */
@ -36,13 +37,37 @@
#define err(code, fmt, arg...) do { printf(fmt, ##arg); exit(code); } while (0) #define err(code, fmt, arg...) do { printf(fmt, ##arg); exit(code); } while (0)
int done = 0; int done = 0;
int rcvbufsz=0;
char name[100];
int dbg=0, print_delays=0;
__u64 stime, utime;
#define PRINTF(fmt, arg...) { \
if (dbg) { \
printf(fmt, ##arg); \
} \
}
/* Maximum size of response requested or message sent */
#define MAX_MSG_SIZE 256
/* Maximum number of cpus expected to be specified in a cpumask */
#define MAX_CPUS 32
/* Maximum length of pathname to log file */
#define MAX_FILENAME 256
struct msgtemplate {
struct nlmsghdr n;
struct genlmsghdr g;
char buf[MAX_MSG_SIZE];
};
char cpumask[100+6*MAX_CPUS];
/* /*
* Create a raw netlink socket and bind * Create a raw netlink socket and bind
*/ */
static int create_nl_socket(int protocol, int groups) static int create_nl_socket(int protocol)
{ {
socklen_t addr_len;
int fd; int fd;
struct sockaddr_nl local; struct sockaddr_nl local;
@ -50,9 +75,16 @@ static int create_nl_socket(int protocol, int groups)
if (fd < 0) if (fd < 0)
return -1; return -1;
if (rcvbufsz)
if (setsockopt(fd, SOL_SOCKET, SO_RCVBUF,
&rcvbufsz, sizeof(rcvbufsz)) < 0) {
printf("Unable to set socket rcv buf size to %d\n",
rcvbufsz);
return -1;
}
memset(&local, 0, sizeof(local)); memset(&local, 0, sizeof(local));
local.nl_family = AF_NETLINK; local.nl_family = AF_NETLINK;
local.nl_groups = groups;
if (bind(fd, (struct sockaddr *) &local, sizeof(local)) < 0) if (bind(fd, (struct sockaddr *) &local, sizeof(local)) < 0)
goto error; goto error;
@ -63,78 +95,72 @@ static int create_nl_socket(int protocol, int groups)
return -1; return -1;
} }
int sendto_fd(int s, const char *buf, int bufLen)
{
struct sockaddr_nl nladdr;
int r;
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)); memset(&nladdr, 0, sizeof(nladdr));
nladdr.nl_family = AF_NETLINK; nladdr.nl_family = AF_NETLINK;
while ((r = sendto(sd, buf, buflen, 0, (struct sockaddr *) &nladdr,
while ((r = sendto(s, buf, bufLen, 0, (struct sockaddr *) &nladdr, sizeof(nladdr))) < buflen) {
sizeof(nladdr))) < bufLen) {
if (r > 0) { if (r > 0) {
buf += r; buf += r;
bufLen -= r; buflen -= r;
} else if (errno != EAGAIN) } else if (errno != EAGAIN)
return -1; return -1;
} }
return 0; return 0;
} }
/* /*
* Probe the controller in genetlink to find the family id * Probe the controller in genetlink to find the family id
* for the TASKSTATS family * for the TASKSTATS family
*/ */
int get_family_id(int sd) int get_family_id(int sd)
{ {
struct {
struct nlmsghdr n;
struct genlmsghdr g;
char buf[256];
} family_req;
struct { struct {
struct nlmsghdr n; struct nlmsghdr n;
struct genlmsghdr g; struct genlmsghdr g;
char buf[256]; char buf[256];
} ans; } ans;
int id; int id, rc;
struct nlattr *na; struct nlattr *na;
int rep_len; int rep_len;
/* Get family name */ strcpy(name, TASKSTATS_GENL_NAME);
family_req.n.nlmsg_type = GENL_ID_CTRL; rc = send_cmd(sd, GENL_ID_CTRL, getpid(), CTRL_CMD_GETFAMILY,
family_req.n.nlmsg_flags = NLM_F_REQUEST; CTRL_ATTR_FAMILY_NAME, (void *)name,
family_req.n.nlmsg_seq = 0; strlen(TASKSTATS_GENL_NAME)+1);
family_req.n.nlmsg_pid = getpid();
family_req.n.nlmsg_len = NLMSG_LENGTH(GENL_HDRLEN);
family_req.g.cmd = CTRL_CMD_GETFAMILY;
family_req.g.version = 0x1;
na = (struct nlattr *) GENLMSG_DATA(&family_req);
na->nla_type = CTRL_ATTR_FAMILY_NAME;
na->nla_len = strlen(TASKSTATS_GENL_NAME) + 1 + NLA_HDRLEN;
strcpy(NLA_DATA(na), TASKSTATS_GENL_NAME);
family_req.n.nlmsg_len += NLMSG_ALIGN(na->nla_len);
if (sendto_fd(sd, (char *) &family_req, family_req.n.nlmsg_len) < 0)
err(1, "error sending message via Netlink\n");
rep_len = recv(sd, &ans, sizeof(ans), 0); rep_len = recv(sd, &ans, sizeof(ans), 0);
if (ans.n.nlmsg_type == NLMSG_ERROR ||
if (rep_len < 0) (rep_len < 0) || !NLMSG_OK((&ans.n), rep_len))
err(1, "error receiving reply message via Netlink\n"); return 0;
/* Validate response message */
if (!NLMSG_OK((&ans.n), rep_len))
err(1, "invalid reply message received via Netlink\n");
if (ans.n.nlmsg_type == NLMSG_ERROR) { /* error */
printf("error received NACK - leaving\n");
exit(1);
}
na = (struct nlattr *) GENLMSG_DATA(&ans); na = (struct nlattr *) GENLMSG_DATA(&ans);
na = (struct nlattr *) ((char *) na + NLA_ALIGN(na->nla_len)); na = (struct nlattr *) ((char *) na + NLA_ALIGN(na->nla_len));
@ -144,7 +170,7 @@ int get_family_id(int sd)
return id; return id;
} }
void print_taskstats(struct taskstats *t) void print_delayacct(struct taskstats *t)
{ {
printf("\n\nCPU %15s%15s%15s%15s\n" printf("\n\nCPU %15s%15s%15s%15s\n"
" %15llu%15llu%15llu%15llu\n" " %15llu%15llu%15llu%15llu\n"
@ -160,181 +186,154 @@ void print_taskstats(struct taskstats *t)
"count", "delay total", t->swapin_count, t->swapin_delay_total); "count", "delay total", t->swapin_count, t->swapin_delay_total);
} }
void sigchld(int sig)
{
done = 1;
}
int main(int argc, char *argv[]) int main(int argc, char *argv[])
{ {
int rc; int c, rc, rep_len, aggr_len, len2, cmd_type;
int sk_nl;
struct nlmsghdr *nlh;
struct genlmsghdr *genlhdr;
char *buf;
struct taskstats_cmd_param *param;
__u16 id; __u16 id;
__u32 mypid;
struct nlattr *na; struct nlattr *na;
/* For receiving */
struct sockaddr_nl kern_nla, from_nla;
socklen_t from_nla_len;
int recv_len;
struct taskstats_reply *reply;
struct {
struct nlmsghdr n;
struct genlmsghdr g;
char buf[256];
} req;
struct {
struct nlmsghdr n;
struct genlmsghdr g;
char buf[256];
} ans;
int nl_sd = -1; int nl_sd = -1;
int rep_len;
int len = 0; int len = 0;
int aggr_len, len2;
struct sockaddr_nl nladdr;
pid_t tid = 0; pid_t tid = 0;
pid_t rtid = 0; pid_t rtid = 0;
int cmd_type = TASKSTATS_TYPE_TGID;
int c, status;
int forking = 0;
struct sigaction act = {
.sa_handler = SIG_IGN,
.sa_mask = SA_NOMASK,
};
struct sigaction tact ;
if (argc < 3) { int fd = 0;
printf("usage %s [-t tgid][-p pid][-c cmd]\n", argv[0]); int count = 0;
exit(-1); int write_file = 0;
} int maskset = 0;
char logfile[128];
int loop = 0;
tact.sa_handler = sigchld; struct msgtemplate msg;
sigemptyset(&tact.sa_mask);
if (sigaction(SIGCHLD, &tact, NULL) < 0)
err(1, "sigaction failed for SIGCHLD\n");
while (1) { while (1) {
c = getopt(argc, argv, "dw:r:m:t:p:v:l");
c = getopt(argc, argv, "t:p:c:");
if (c < 0) if (c < 0)
break; break;
switch (c) { switch (c) {
case 'd':
printf("print delayacct stats ON\n");
print_delays = 1;
break;
case 'w':
strncpy(logfile, optarg, MAX_FILENAME);
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': case 't':
tid = atoi(optarg); tid = atoi(optarg);
if (!tid) if (!tid)
err(1, "Invalid tgid\n"); err(1, "Invalid tgid\n");
cmd_type = TASKSTATS_CMD_ATTR_TGID; cmd_type = TASKSTATS_CMD_ATTR_TGID;
print_delays = 1;
break; break;
case 'p': case 'p':
tid = atoi(optarg); tid = atoi(optarg);
if (!tid) if (!tid)
err(1, "Invalid pid\n"); err(1, "Invalid pid\n");
cmd_type = TASKSTATS_CMD_ATTR_TGID; cmd_type = TASKSTATS_CMD_ATTR_PID;
print_delays = 1;
break; break;
case 'c': case 'v':
opterr = 0; printf("debug on\n");
tid = fork(); dbg = 1;
if (tid < 0) break;
err(1, "fork failed\n"); case 'l':
printf("listen forever\n");
if (tid == 0) { /* child process */ loop = 1;
if (execvp(argv[optind - 1], &argv[optind - 1]) < 0) {
exit(-1);
}
}
forking = 1;
break; break;
default: default:
printf("usage %s [-t tgid][-p pid][-c cmd]\n", argv[0]); printf("Unknown option %d\n", c);
exit(-1); exit(-1);
break;
} }
if (c == 'c')
break;
} }
/* Construct Netlink request message */ 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);
}
}
/* Send Netlink request message & get reply */ if ((nl_sd = create_nl_socket(NETLINK_GENERIC)) < 0)
if ((nl_sd =
create_nl_socket(NETLINK_GENERIC, TASKSTATS_LISTEN_GROUP)) < 0)
err(1, "error creating Netlink socket\n"); err(1, "error creating Netlink socket\n");
mypid = getpid();
id = get_family_id(nl_sd); id = get_family_id(nl_sd);
if (!id) {
printf("Error getting family id, errno %d", errno);
goto err;
}
PRINTF("family id %d\n", id);
/* Send command needed */ if (maskset) {
req.n.nlmsg_len = NLMSG_LENGTH(GENL_HDRLEN); rc = send_cmd(nl_sd, id, mypid, TASKSTATS_CMD_GET,
req.n.nlmsg_type = id; TASKSTATS_CMD_ATTR_REGISTER_CPUMASK,
req.n.nlmsg_flags = NLM_F_REQUEST; &cpumask, sizeof(cpumask));
req.n.nlmsg_seq = 0; PRINTF("Sent register cpumask, retval %d\n", rc);
req.n.nlmsg_pid = tid; if (rc < 0) {
req.g.cmd = TASKSTATS_CMD_GET; printf("error sending register cpumask\n");
na = (struct nlattr *) GENLMSG_DATA(&req); goto err;
na->nla_type = cmd_type; }
na->nla_len = sizeof(unsigned int) + NLA_HDRLEN; }
*(__u32 *) NLA_DATA(na) = tid;
req.n.nlmsg_len += NLMSG_ALIGN(na->nla_len);
if (tid) {
if (!forking && sendto_fd(nl_sd, (char *) &req, req.n.nlmsg_len) < 0) rc = send_cmd(nl_sd, id, mypid, TASKSTATS_CMD_GET,
err(1, "error sending message via Netlink\n"); cmd_type, &tid, sizeof(__u32));
PRINTF("Sent pid/tgid, retval %d\n", rc);
act.sa_handler = SIG_IGN; if (rc < 0) {
sigemptyset(&act.sa_mask); printf("error sending tid/tgid cmd\n");
if (sigaction(SIGINT, &act, NULL) < 0) goto done;
err(1, "sigaction failed for SIGINT\n"); }
}
do { do {
int i; int i;
struct pollfd pfd;
int pollres;
pfd.events = 0xffff & ~POLLOUT; rep_len = recv(nl_sd, &msg, sizeof(msg), 0);
pfd.fd = nl_sd; PRINTF("received %d bytes\n", rep_len);
pollres = poll(&pfd, 1, 5000);
if (pollres < 0 || done) {
break;
}
rep_len = recv(nl_sd, &ans, sizeof(ans), 0);
nladdr.nl_family = AF_NETLINK;
nladdr.nl_groups = TASKSTATS_LISTEN_GROUP;
if (ans.n.nlmsg_type == NLMSG_ERROR) { /* error */
printf("error received NACK - leaving\n");
exit(1);
}
if (rep_len < 0) { if (rep_len < 0) {
err(1, "error receiving reply message via Netlink\n"); printf("nonfatal reply error: errno %d\n", errno);
break; continue;
}
if (msg.n.nlmsg_type == NLMSG_ERROR ||
!NLMSG_OK((&msg.n), rep_len)) {
printf("fatal reply error, errno %d\n", errno);
goto done;
} }
/* Validate response message */ PRINTF("nlmsghdr size=%d, nlmsg_len=%d, rep_len=%d\n",
if (!NLMSG_OK((&ans.n), rep_len)) sizeof(struct nlmsghdr), msg.n.nlmsg_len, rep_len);
err(1, "invalid reply message received via Netlink\n");
rep_len = GENLMSG_PAYLOAD(&ans.n);
na = (struct nlattr *) GENLMSG_DATA(&ans); rep_len = GENLMSG_PAYLOAD(&msg.n);
na = (struct nlattr *) GENLMSG_DATA(&msg);
len = 0; len = 0;
i = 0; i = 0;
while (len < rep_len) { while (len < rep_len) {
len += NLA_ALIGN(na->nla_len); len += NLA_ALIGN(na->nla_len);
switch (na->nla_type) { switch (na->nla_type) {
case TASKSTATS_TYPE_AGGR_PID:
/* Fall through */
case TASKSTATS_TYPE_AGGR_TGID: case TASKSTATS_TYPE_AGGR_TGID:
/* Fall through */
case TASKSTATS_TYPE_AGGR_PID:
aggr_len = NLA_PAYLOAD(na->nla_len); aggr_len = NLA_PAYLOAD(na->nla_len);
len2 = 0; len2 = 0;
/* For nested attributes, na follows */ /* For nested attributes, na follows */
@ -344,33 +343,54 @@ int main(int argc, char *argv[])
switch (na->nla_type) { switch (na->nla_type) {
case TASKSTATS_TYPE_PID: case TASKSTATS_TYPE_PID:
rtid = *(int *) NLA_DATA(na); rtid = *(int *) NLA_DATA(na);
if (print_delays)
printf("PID\t%d\n", rtid);
break; break;
case TASKSTATS_TYPE_TGID: case TASKSTATS_TYPE_TGID:
rtid = *(int *) NLA_DATA(na); rtid = *(int *) NLA_DATA(na);
if (print_delays)
printf("TGID\t%d\n", rtid);
break; break;
case TASKSTATS_TYPE_STATS: case TASKSTATS_TYPE_STATS:
if (rtid == tid) { count++;
print_taskstats((struct taskstats *) if (print_delays)
NLA_DATA(na)); print_delayacct((struct taskstats *) NLA_DATA(na));
done = 1; if (fd) {
if (write(fd, NLA_DATA(na), na->nla_len) < 0) {
err(1,"write error\n");
} }
}
if (!loop)
goto done;
break;
default:
printf("Unknown nested nla_type %d\n", na->nla_type);
break; break;
} }
len2 += NLA_ALIGN(na->nla_len); len2 += NLA_ALIGN(na->nla_len);
na = (struct nlattr *) ((char *) na + len2); na = (struct nlattr *) ((char *) na + len2);
if (done) }
break; break;
}
}
na = (struct nlattr *) (GENLMSG_DATA(&ans) + len);
if (done)
break;
}
if (done)
break;
}
while (1);
default:
printf("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, sizeof(cpumask));
printf("Sent deregister mask, retval %d\n", rc);
if (rc < 0)
err(rc, "error sending deregister cpumask\n");
}
err:
close(nl_sd); close(nl_sd);
if (fd)
close(fd);
return 0; return 0;
} }

64
Documentation/accounting/taskstats.txt
Просмотреть файл

@ -26,20 +26,28 @@ leader - a process is deemed alive as long as it has any task belonging to it.
Usage Usage
----- -----
To get statistics during task's lifetime, userspace opens a unicast netlink To get statistics during a task's lifetime, userspace opens a unicast netlink
socket (NETLINK_GENERIC family) and sends commands specifying a pid or a tgid. socket (NETLINK_GENERIC family) and sends commands specifying a pid or a tgid.
The response contains statistics for a task (if pid is specified) or the sum of The response contains statistics for a task (if pid is specified) or the sum of
statistics for all tasks of the process (if tgid is specified). statistics for all tasks of the process (if tgid is specified).
To obtain statistics for tasks which are exiting, userspace opens a multicast To obtain statistics for tasks which are exiting, the userspace listener
netlink socket. Each time a task exits, its per-pid statistics is always sent sends a register command and specifies a cpumask. Whenever a task exits on
by the kernel to each listener on the multicast socket. In addition, if it is one of the cpus in the cpumask, its per-pid statistics are sent to the
the last thread exiting its thread group, an additional record containing the registered listener. Using cpumasks allows the data received by one listener
per-tgid stats are also sent. The latter contains the sum of per-pid stats for to be limited and assists in flow control over the netlink interface and is
all threads in the thread group, both past and present. explained in more detail below.
If the exiting task is the last thread exiting its thread group,
an additional record containing the per-tgid stats is also sent to userspace.
The latter contains the sum of per-pid stats for all threads in the thread
group, both past and present.
getdelays.c is a simple utility demonstrating usage of the taskstats interface getdelays.c is a simple utility demonstrating usage of the taskstats interface
for reporting delay accounting statistics. for reporting delay accounting statistics. Users can register cpumasks,
send commands and process responses, listen for per-tid/tgid exit data,
write the data received to a file and do basic flow control by increasing
receive buffer sizes.
Interface Interface
--------- ---------
@ -66,10 +74,20 @@ The messages are in the format
The taskstats payload is one of the following three kinds: The taskstats payload is one of the following three kinds:
1. Commands: Sent from user to kernel. The payload is one attribute, of type 1. Commands: Sent from user to kernel. Commands to get data on
TASKSTATS_CMD_ATTR_PID/TGID, containing a u32 pid or tgid in the attribute a pid/tgid consist of one attribute, of type TASKSTATS_CMD_ATTR_PID/TGID,
payload. The pid/tgid denotes the task/process for which userspace wants containing a u32 pid or tgid in the attribute payload. The pid/tgid denotes
statistics. the task/process for which userspace wants statistics.
Commands to register/deregister interest in exit data from a set of cpus
consist of one attribute, of type
TASKSTATS_CMD_ATTR_REGISTER/DEREGISTER_CPUMASK and contain a cpumask in the
attribute payload. The cpumask is specified as an ascii string of
comma-separated cpu ranges e.g. to listen to exit data from cpus 1,2,3,5,7,8
the cpumask would be "1-3,5,7-8". If userspace forgets to deregister interest
in cpus before closing the listening socket, the kernel cleans up its interest
set over time. However, for the sake of efficiency, an explicit deregistration
is advisable.
2. Response for a command: sent from the kernel in response to a userspace 2. Response for a command: sent from the kernel in response to a userspace
command. The payload is a series of three attributes of type: command. The payload is a series of three attributes of type:
@ -138,4 +156,26 @@ struct too much, requiring disparate userspace accounting utilities to
unnecessarily receive large structures whose fields are of no interest, then unnecessarily receive large structures whose fields are of no interest, then
extending the attributes structure would be worthwhile. extending the attributes structure would be worthwhile.
Flow control for taskstats
--------------------------
When the rate of task exits becomes large, a listener may not be able to keep
up with the kernel's rate of sending per-tid/tgid exit data leading to data
loss. This possibility gets compounded when the taskstats structure gets
extended and the number of cpus grows large.
To avoid losing statistics, userspace should do one or more of the following:
- increase the receive buffer sizes for the netlink sockets opened by
listeners to receive exit data.
- create more listeners and reduce the number of cpus being listened to by
each listener. In the extreme case, there could be one listener for each cpu.
Users may also consider setting the cpu affinity of the listener to the subset
of cpus to which it listens, especially if they are listening to just one cpu.
Despite these measures, if the userspace receives ENOBUFS error messages
indicated overflow of receive buffers, it should take measures to handle the
loss of data.
---- ----