From c16eb6cf753d5df0779c284126d7356d2377fdb3 Mon Sep 17 00:00:00 2001
From: Sebastien GODARD <sysstat@users.noreply.github.com>
Date: Mon, 22 Oct 2018 18:06:48 +0200
Subject: [PATCH] pidstat: Now handle processes with spaces in their name
 properly

pidstat had a bug which made it unable to handle processes with spaces
in their name, eg:

$ cat /proc/5768/stat
5768 (Plex New Transc) S 1264 1252 1252 0 -1 1077960704 17317 0 18 0
9260 137 0 0 15 -5 8 0 1430749 186589184 16704 18446744073709551615 1 1
0 0 0 0 0 0 8404998 18446744073709551615 0 0 17 1 0 0 22 0 0 0 0 0 0 0 0
0 0

This patch fixes the problem.

Reported by Chris Grindstaff (@cgrinds).

Signed-off-by: Sebastien GODARD <sysstat@users.noreply.github.com>
---
 pidstat.c | 59 +++++++++++++++++++++++++++++++++++------------------------
 1 file changed, 35 insertions(+), 24 deletions(-)

diff --git a/pidstat.c b/pidstat.c
index bc22aa8..af48f29 100644
--- a/pidstat.c
+++ b/pidstat.c
@@ -23,6 +23,7 @@
 #include <string.h>
 #include <stdlib.h>
 #include <unistd.h>
+#include <fcntl.h>
 #include <signal.h>
 #include <dirent.h>
 #include <ctype.h>
@@ -307,9 +308,10 @@ void read_proc_meminfo(void)
 int read_proc_pid_stat(unsigned int pid, struct pid_stats *pst,
 		       unsigned int *thread_nr, unsigned int tgid)
 {
-	FILE *fp;
-	char filename[128], format[256], comm[MAX_COMM_LEN + 1];
-	size_t len;
+	int fd, sz, rc, commsz;
+	char filename[128];
+	static char buffer[1024 + 1];
+	char *start, *end;
 
 	if (tgid) {
 		sprintf(filename, TASK_STAT, tgid, pid);
@@ -318,36 +320,45 @@ int read_proc_pid_stat(unsigned int pid, struct pid_stats *pst,
 		sprintf(filename, PID_STAT, pid);
 	}
 
-	if ((fp = fopen(filename, "r")) == NULL)
+	if ((fd = open(filename, O_RDONLY)) < 0)
 		/* No such process */
 		return 1;
 
-	sprintf(format, "%%*d (%%%ds %%*s %%*d %%*d %%*d %%*d %%*d %%*u %%lu %%lu"
-		" %%lu %%lu %%lu %%lu %%lu %%lu %%*d %%*d %%u %%*u %%*d %%lu %%lu"
-		" %%*u %%*u %%*u %%*u %%*u %%*u %%*u %%*u %%*u %%*u %%*u %%*u %%*u"
-		" %%*u %%u %%*u %%*u %%*u %%lu %%lu\\n", MAX_COMM_LEN);
-
-	fscanf(fp, format, comm,
-	       &pst->minflt, &pst->cminflt, &pst->majflt, &pst->cmajflt,
-	       &pst->utime,  &pst->stime, &pst->cutime, &pst->cstime,
-	       thread_nr, &pst->vsz, &pst->rss, &pst->processor,
-	       &pst->gtime, &pst->cgtime);
+	sz = read(fd, buffer, 1024);
+	close(fd);
+	if (sz <= 0)
+		return 1;
+	buffer[sz] = '\0';
 
-	fclose(fp);
+	if ((start = strchr(buffer, '(')) == NULL)
+		return 1;
+	start += 1;
+	if ((end = strrchr(start, ')')) == NULL)
+		return 1;
+	commsz = end - start;
+	if (commsz >= MAX_COMM_LEN)
+		return 1;
+	memcpy(pst->comm, start, commsz);
+	pst->comm[commsz] = '\0';
+	start = end + 2;
+
+	rc = sscanf(start,
+		    "%*s %*d %*d %*d %*d %*d %*u %lu %lu"
+		    " %lu %lu %lu %lu %ld %ld %*d %*d %u %*u %*d %lu %lu"
+		    " %*u %*u %*u %*u %*u %*u %*u %*u %*u %*u %*u %*u %*u"
+		    " %*u %u %*u %*u %*lu %lu %ld\n",
+		    &pst->minflt, &pst->cminflt, &pst->majflt, &pst->cmajflt,
+		    &pst->utime,  &pst->stime, &pst->cutime, &pst->cstime,
+		    thread_nr, &pst->vsz, &pst->rss, &pst->processor,
+		    &pst->gtime, &pst->cgtime);
+
+	if (rc < 14)
+		return 1;
 
 	/* Convert to kB */
 	pst->vsz >>= 10;
 	pst->rss = PG_TO_KB(pst->rss);
 
-	strncpy(pst->comm, comm, MAX_COMM_LEN);
-	pst->comm[MAX_COMM_LEN - 1] = '\0';
-
-	/* Remove trailing ')' */
-	len = strlen(pst->comm);
-	if (len && (pst->comm[len - 1] == ')')) {
-		pst->comm[len - 1] = '\0';
-	}
-
 	pst->pid = pid;
 	pst->tgid = tgid;
 	return 0;
-- 
2.14.4