util-lib: rework get_process_cmdline() (#3529)

This reworks get_process_cmdline() quite substantially, fixing the following:

- Fixes:
  https://github.com/systemd/systemd/pull/3512/commits/a4e3bf4d7ac2de51191ce136ee9361ba319e106c#r66837630

- The passed max_length is also applied to the "comm" name, if comm_fallback is
  set.

- The right thing happens if max_length == 1 is specified

- when the cmdline "foobar" is abbreviated to 6 characters the result is not
  "foobar" instead of "foo...".

- trailing whitespace are removed before the ... suffix is appended. The 7
  character abbreviation of "foo barz" is hence "foo..." instead of "foo ...".

- leading whitespace are suppressed from the cmdline

- a comprehensive test case is added

2 files changed, 270 insertions, 12 deletions

diff --git a/src/basic/process-util.c b/src/basic/process-util.c
index 08fa98bb9e..b5b068ad38 100644
--- a/src/basic/process-util.c
+++ b/src/basic/process-util.c
@@ -110,6 +110,15 @@ int get_process_cmdline(pid_t pid, size_t max_length, bool comm_fallback, char *
         assert(line);
         assert(pid >= 0);
 
+        /* Retrieves a process' command line. Replaces unprintable characters while doing so by whitespace (coalescing
+         * multiple sequential ones into one). If max_length is != 0 will return a string of the specified size at most
+         * (the trailing NUL byte does count towards the length here!), abbreviated with a "..." ellipsis. If
+         * comm_fallback is true and the process has no command line set (the case for kernel threads), or has a
+         * command line that resolves to the empty string will return the "comm" name of the process instead.
+         *
+         * Returns -ESRCH if the process doesn't exist, and -ENOENT if the process has no command line (and
+         * comm_fallback is false). */
+
         p = procfs_file_alloca(pid, "cmdline");
 
         f = fopen(p, "re");
@@ -119,12 +128,22 @@ int get_process_cmdline(pid_t pid, size_t max_length, bool comm_fallback, char *
                 return -errno;
         }
 
-        if (max_length == 0) {
+        if (max_length == 1) {
+
+                /* If there's only room for one byte, return the empty string */
+                r = new0(char, 1);
+                if (!r)
+                        return -ENOMEM;
+
+                *line = r;
+                return 0;
+
+        } else if (max_length == 0) {
                 size_t len = 0, allocated = 0;
 
                 while ((c = getc(f)) != EOF) {
 
-                        if (!GREEDY_REALLOC(r, allocated, len+2)) {
+                        if (!GREEDY_REALLOC(r, allocated, len+3)) {
                                 free(r);
                                 return -ENOMEM;
                         }
@@ -136,7 +155,7 @@ int get_process_cmdline(pid_t pid, size_t max_length, bool comm_fallback, char *
                                 }
 
                                 r[len++] = c;
-                        } else
+                        } else if (len > 0)
                                 space = true;
                }
 
@@ -144,6 +163,7 @@ int get_process_cmdline(pid_t pid, size_t max_length, bool comm_fallback, char *
                         r[len] = 0;
 
         } else {
+                bool dotdotdot = false;
                 size_t left;
 
                 r = new(char, max_length);
@@ -155,28 +175,46 @@ int get_process_cmdline(pid_t pid, size_t max_length, bool comm_fallback, char *
                 while ((c = getc(f)) != EOF) {
 
                         if (isprint(c)) {
+
                                 if (space) {
-                                        if (left <= 4)
+                                        if (left <= 2) {
+                                                dotdotdot = true;
                                                 break;
+                                        }
 
                                         *(k++) = ' ';
                                         left--;
                                         space = false;
                                 }
 
-                                if (left <= 4)
+                                if (left <= 1) {
+                                        dotdotdot = true;
                                         break;
+                                }
 
                                 *(k++) = (char) c;
                                 left--;
-                        }  else
+                        }  else if (k > r)
                                 space = true;
                 }
 
-                if (left <= 4) {
-                        size_t n = MIN(left-1, 3U);
-                        memcpy(k, "...", n);
-                        k[n] = 0;
+                if (dotdotdot) {
+                        if (max_length <= 4) {
+                                k = r;
+                                left = max_length;
+                        } else {
+                                k = r + max_length - 4;
+                                left = 4;
+
+                                /* Eat up final spaces */
+                                while (k > r && isspace(k[-1])) {
+                                        k--;
+                                        left++;
+                                }
+                        }
+
+                        strncpy(k, "...", left-1);
+                        k[left] = 0;
                 } else
                         *k = 0;
         }
@@ -195,7 +233,37 @@ int get_process_cmdline(pid_t pid, size_t max_length, bool comm_fallback, char *
                 if (h < 0)
                         return h;
 
-                r = strjoin("[", t, "]", NULL);
+                if (max_length == 0)
+                        r = strjoin("[", t, "]", NULL);
+                else {
+                        size_t l;
+
+                        l = strlen(t);
+
+                        if (l + 3 <= max_length)
+                                r = strjoin("[", t, "]", NULL);
+                        else if (max_length <= 6) {
+
+                                r = new(char, max_length);
+                                if (!r)
+                                        return -ENOMEM;
+
+                                memcpy(r, "[...]", max_length-1);
+                                r[max_length-1] = 0;
+                        } else {
+                                char *e;
+
+                                t[max_length - 6] = 0;
+
+                                /* Chop off final spaces */
+                                e = strchr(t, 0);
+                                while (e > t && isspace(e[-1]))
+                                        e--;
+                                *e = 0;
+
+                                r = strjoin("[", t, "...]", NULL);
+                        }
+                }
                 if (!r)
                         return -ENOMEM;
         }
diff --git a/src/test/test-process-util.c b/src/test/test-process-util.c
index 8bb5f6e3a3..af2c9282d4 100644
--- a/src/test/test-process-util.c
+++ b/src/test/test-process-util.c
@@ -18,7 +18,10 @@
   along with systemd; If not, see <http://www.gnu.org/licenses/>.
 ***/
 
+#include <sched.h>
+#include <sys/mount.h>
 #include <sys/personality.h>
+#include <sys/prctl.h>
 #include <sys/stat.h>
 #include <sys/types.h>
 #include <sys/wait.h>
@@ -26,6 +29,7 @@
 
 #include "alloc-util.h"
 #include "architecture.h"
+#include "fd-util.h"
 #include "log.h"
 #include "macro.h"
 #include "parse-util.h"
@@ -59,7 +63,11 @@ static void test_get_process_comm(pid_t pid) {
         log_info("PID"PID_FMT" cmdline: '%s'", pid, c);
 
         assert_se(get_process_cmdline(pid, 8, false, &d) >= 0);
-        log_info("PID"PID_FMT" cmdline truncated: '%s'", pid, d);
+        log_info("PID"PID_FMT" cmdline truncated to 8: '%s'", pid, d);
+
+        free(d);
+        assert_se(get_process_cmdline(pid, 1, false, &d) >= 0);
+        log_info("PID"PID_FMT" cmdline truncated to 1: '%s'", pid, d);
 
         assert_se(get_process_ppid(pid, &e) >= 0);
         log_info("PID"PID_FMT" PPID: "PID_FMT, pid, e);
@@ -147,6 +155,187 @@ static void test_personality(void) {
 #endif
 }
 
+static void test_get_process_cmdline_harder(void) {
+        char path[] = "/tmp/test-cmdlineXXXXXX";
+        _cleanup_close_ int fd = -1;
+        _cleanup_free_ char *line = NULL;
+        pid_t pid;
+
+        if (geteuid() != 0)
+                return;
+
+        pid = fork();
+        if (pid > 0) {
+                siginfo_t si;
+
+                (void) wait_for_terminate(pid, &si);
+
+                assert_se(si.si_code == CLD_EXITED);
+                assert_se(si.si_status == 0);
+
+                return;
+        }
+
+        assert_se(pid == 0);
+        assert_se(unshare(CLONE_NEWNS) >= 0);
+
+        fd = mkostemp(path, O_CLOEXEC);
+        assert_se(fd >= 0);
+        assert_se(mount(path, "/proc/self/cmdline", "bind", MS_BIND, NULL) >= 0);
+        assert_se(unlink(path) >= 0);
+
+        assert_se(prctl(PR_SET_NAME, "testa") >= 0);
+
+        assert_se(get_process_cmdline(getpid(), 0, false, &line) == -ENOENT);
+
+        assert_se(get_process_cmdline(getpid(), 0, true, &line) >= 0);
+        assert_se(streq(line, "[testa]"));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 1, true, &line) >= 0);
+        assert_se(streq(line, ""));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 2, true, &line) >= 0);
+        assert_se(streq(line, "["));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 3, true, &line) >= 0);
+        assert_se(streq(line, "[."));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 4, true, &line) >= 0);
+        assert_se(streq(line, "[.."));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 5, true, &line) >= 0);
+        assert_se(streq(line, "[..."));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 6, true, &line) >= 0);
+        assert_se(streq(line, "[...]"));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 7, true, &line) >= 0);
+        assert_se(streq(line, "[t...]"));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 8, true, &line) >= 0);
+        assert_se(streq(line, "[testa]"));
+        line = mfree(line);
+
+        assert_se(write(fd, "\0\0\0\0\0\0\0\0\0", 10) == 10);
+
+        assert_se(get_process_cmdline(getpid(), 0, false, &line) == -ENOENT);
+
+        assert_se(get_process_cmdline(getpid(), 0, true, &line) >= 0);
+        assert_se(streq(line, "[testa]"));
+        line = mfree(line);
+
+        assert_se(write(fd, "foo\0bar\0\0\0\0\0", 10) == 10);
+
+        assert_se(get_process_cmdline(getpid(), 0, false, &line) >= 0);
+        assert_se(streq(line, "foo bar"));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 0, true, &line) >= 0);
+        assert_se(streq(line, "foo bar"));
+        line = mfree(line);
+
+        assert_se(write(fd, "quux", 4) == 4);
+        assert_se(get_process_cmdline(getpid(), 0, false, &line) >= 0);
+        assert_se(streq(line, "foo bar quux"));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 0, true, &line) >= 0);
+        assert_se(streq(line, "foo bar quux"));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 1, true, &line) >= 0);
+        assert_se(streq(line, ""));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 2, true, &line) >= 0);
+        assert_se(streq(line, "."));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 3, true, &line) >= 0);
+        assert_se(streq(line, ".."));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 4, true, &line) >= 0);
+        assert_se(streq(line, "..."));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 5, true, &line) >= 0);
+        assert_se(streq(line, "f..."));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 6, true, &line) >= 0);
+        assert_se(streq(line, "fo..."));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 7, true, &line) >= 0);
+        assert_se(streq(line, "foo..."));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 8, true, &line) >= 0);
+        assert_se(streq(line, "foo..."));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 9, true, &line) >= 0);
+        assert_se(streq(line, "foo b..."));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 10, true, &line) >= 0);
+        assert_se(streq(line, "foo ba..."));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 11, true, &line) >= 0);
+        assert_se(streq(line, "foo bar..."));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 12, true, &line) >= 0);
+        assert_se(streq(line, "foo bar..."));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 13, true, &line) >= 0);
+        assert_se(streq(line, "foo bar quux"));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 14, true, &line) >= 0);
+        assert_se(streq(line, "foo bar quux"));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 1000, true, &line) >= 0);
+        assert_se(streq(line, "foo bar quux"));
+        line = mfree(line);
+
+        assert_se(ftruncate(fd, 0) >= 0);
+        assert_se(prctl(PR_SET_NAME, "aaaa bbbb cccc") >= 0);
+
+        assert_se(get_process_cmdline(getpid(), 0, false, &line) == -ENOENT);
+
+        assert_se(get_process_cmdline(getpid(), 0, true, &line) >= 0);
+        assert_se(streq(line, "[aaaa bbbb cccc]"));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 10, true, &line) >= 0);
+        assert_se(streq(line, "[aaaa...]"));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 11, true, &line) >= 0);
+        assert_se(streq(line, "[aaaa...]"));
+        line = mfree(line);
+
+        assert_se(get_process_cmdline(getpid(), 12, true, &line) >= 0);
+        assert_se(streq(line, "[aaaa b...]"));
+        line = mfree(line);
+
+        safe_close(fd);
+        _exit(0);
+}
+
 int main(int argc, char *argv[]) {
         log_parse_environment();
         log_open();
@@ -164,6 +353,7 @@ int main(int argc, char *argv[]) {
         test_pid_is_unwaited();
         test_pid_is_alive();
         test_personality();
+        test_get_process_cmdline_harder();
 
         return 0;
 }