fs/fsstress: prevent orphan and zombie processes to be left

parent use signal handler to deal with SIGTERM, make sure children processes ended before parent.
---

when I run "killall fsstress" in my testcase, I can't use a simple 'wait' to wait fsstress
all over then 'umount' device. Because many orphan processes still running.

This is not a bug, but I think this patch can make fsstress be better. So I backport xfstests/fsstress here.

BTW, I modify some compile warning for opdesc_t ops[]. Hope it's not wrong ;)

Thank you,
Zorro Lang

 testcases/kernel/fs/fsstress/fsstress.c | 78
++++++++++++++++++++++++---------
 1 file changed, 58 insertions(+), 20 deletions(-)

b/testcases/kernel/fs/fsstress/fsstress.c
index 2243280..ec3f646 100644

@@ -30,7 +30,11 @@
  * http://oss.sgi.com/projects/GenInfo/SGIGPLNoticeExplan/
  */
 +#include <config.h>
 #include "global.h"
+#ifdef HAVE_SYS_PRCTL_H
+#include <sys/prctl.h>
+#endif
  #define XFS_ERRTAG_MAX		17
 @@ -160,35 +164,35 @@ opdesc_t ops[] = {
 	{OP_BULKSTAT, "bulkstat", bulkstat_f, 1, 0, 1},

{OP_BULKSTAT1, "bulkstat1", bulkstat1_f, 1, 0, 1},
#endif
- {OP_CHOWN, "chown", chown_f, 3, 1},
- {OP_CREAT, "creat", creat_f, 4, 1},
- {OP_DREAD, "dread", dread_f, 4, 0},
- {OP_DWRITE, "dwrite", dwrite_f, 4, 1},
- {OP_FDATASYNC, "fdatasync", fdatasync_f, 1, 1},
+ {OP_CHOWN, "chown", chown_f, 3, 1, 0},
+ {OP_CREAT, "creat", creat_f, 4, 1, 0},
+ {OP_DREAD, "dread", dread_f, 4, 0, 0},
+ {OP_DWRITE, "dwrite", dwrite_f, 4, 1, 0},
+ {OP_FDATASYNC, "fdatasync", fdatasync_f, 1, 1, 0},
#ifndef NO_XFS
{OP_FREESP, "freesp", freesp_f, 1, 1, 1},
#endif
- {OP_FSYNC, "fsync", fsync_f, 1, 1},
- {OP_GETDENTS, "getdents", getdents_f, 1, 0},
- {OP_LINK, "link", link_f, 1, 1},
- {OP_MKDIR, "mkdir", mkdir_f, 2, 1},
- {OP_MKNOD, "mknod", mknod_f, 2, 1},
- {OP_READ, "read", read_f, 1, 0},
- {OP_READLINK, "readlink", readlink_f, 1, 0},
- {OP_RENAME, "rename", rename_f, 2, 1},
+ {OP_FSYNC, "fsync", fsync_f, 1, 1, 0},
+ {OP_GETDENTS, "getdents", getdents_f, 1, 0, 0},
+ {OP_LINK, "link", link_f, 1, 1, 0},
+ {OP_MKDIR, "mkdir", mkdir_f, 2, 1, 0},
+ {OP_MKNOD, "mknod", mknod_f, 2, 1, 0},
+ {OP_READ, "read", read_f, 1, 0, 0},
+ {OP_READLINK, "readlink", readlink_f, 1, 0, 0},
+ {OP_RENAME, "rename", rename_f, 2, 1, 0},
#ifndef NO_XFS
{OP_RESVSP, "resvsp", resvsp_f, 1, 1, 1},
#endif
- {OP_RMDIR, "rmdir", rmdir_f, 1, 1},
- {OP_STAT, "stat", stat_f, 1, 0},
- {OP_SYMLINK, "symlink", symlink_f, 2, 1},
- {OP_SYNC, "sync", sync_f, 1, 0},
- {OP_TRUNCATE, "truncate", truncate_f, 2, 1},
- {OP_UNLINK, "unlink", unlink_f, 1, 1},
+ {OP_RMDIR, "rmdir", rmdir_f, 1, 1, 0},
+ {OP_STAT, "stat", stat_f, 1, 0, 0},
+ {OP_SYMLINK, "symlink", symlink_f, 2, 1, 0},
+ {OP_SYNC, "sync", sync_f, 1, 0, 0},
+ {OP_TRUNCATE, "truncate", truncate_f, 2, 1, 0},
+ {OP_UNLINK, "unlink", unlink_f, 1, 1, 0},
#ifndef NO_XFS
{OP_UNRESVSP, "unresvsp", unresvsp_f, 1, 1, 1},
#endif
- {OP_WRITE, "write", write_f, 4, 1},
+ {OP_WRITE, "write", write_f, 4, 1, 0},
}, *ops_end;
flist_t flist[FT_nft] = {
@@ -227,6 +231,7 @@ int no_xfs = 0;
#else
int no_xfs = 1;
#endif
+sig_atomic_t should_stop = 0;
void add_to_flist(int, int, int);
void append_pathname(pathname_t *, char *);
@@ -273,6 +278,11 @@ void usage(void);
void write_freq(void);
void zero_freq(void);
+void sg_handler(int signum)
+{
+ should_stop = 1;
+}
+
int main(int argc, char **argv)
{
char buf[10];
@@ -297,6 +307,7 @@ int main(int argc, char **argv)
#ifndef NO_XFS
xfs_error_injection_t err_inj;
#endif
+ struct sigaction action;
errrange = errtag = 0;
umask(0);
@@ -465,17 +476,44 @@ int main(int argc, char **argv)
#endif
close(fd);
unlink(buf);
+
+
if (nproc == 1) {
procid = 0;
doproc();
} else {
+ setpgid(0, 0);
+ action.sa_handler = sg_handler;
+ sigemptyset(&action.sa_mask);
+ action.sa_flags = 0;
+ if (sigaction(SIGTERM, &action, 0)) {
+ perror("sigaction failed");
+ exit(1);
+ }
+
for (i = 0; i < nproc; i++) {
if (fork() == 0) {
+
+ action.sa_handler = SIG_DFL;
+ sigemptyset(&action.sa_mask);
+ if (sigaction(SIGTERM, &action, 0))
+ return 1;
+#ifdef HAVE_SYS_PRCTL_H
+ prctl(PR_SET_PDEATHSIG, SIGKILL);
+ if (getppid() == 1) /* parent died already? */
+ return 0;
+#endif
procid = i; doproc(); return 0;
}
}
+ while (wait(&stat) > 0 && !should_stop) {
+ continue;
+ }
+ action.sa_flags = SA_RESTART;
+ sigaction(SIGTERM, &action, 0);
+ kill(-getpid(), SIGTERM);
while (wait(&stat) > 0)
continue;
}

1.9.3


------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list