gfs2: Add proper lockspace locking

JIRA: https://issues.redhat.com/browse/RHEL-116886
Upstream Status: git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2.git

Conflicts:
RHEL-9 doesn't have commit dc52cd2 ("fs: dlm: fix F_CANCELLK to cancel pending request").

GFS2 has been calling functions like dlm_lock() even after the lockspace
that these functions operate on has been released with
dlm_release_lockspace().  It has always assumed that those functions
would return -EINVAL in that case, but that was never guaranteed, and it
certainly is no longer the case since commit 4db41bf ("dlm: remove
ls_local_handle from struct dlm_ls").

To fix that, add proper lockspace locking.

Fixes: 3e11e53 ("GFS2: ignore unlock failures after withdraw")
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
(cherry picked from commit 6ab2655)

Author: Andreas Gruenbacher

fs/gfs2/file.c

@@ -1416,6 +1416,7 @@ static int gfs2_lock(struct file *file, int cmd, struct file_lock *fl)
 	struct gfs2_inode *ip = GFS2_I(file->f_mapping->host);
 	struct gfs2_sbd *sdp = GFS2_SB(file->f_mapping->host);
 	struct lm_lockstruct *ls = &sdp->sd_lockstruct;
+	int ret;
 
 	if (!(fl->fl_flags & FL_POSIX))
 		return -ENOLCK;
@@ -1429,12 +1430,18 @@ static int gfs2_lock(struct file *file, int cmd, struct file_lock *fl)
 			locks_lock_file_wait(file, fl);
 		return -EIO;
 	}
-	if (IS_GETLK(cmd))
-		return dlm_posix_get(ls->ls_dlm, ip->i_no_addr, file, fl);
-	else if (fl->fl_type == F_UNLCK)
-		return dlm_posix_unlock(ls->ls_dlm, ip->i_no_addr, file, fl);
-	else
-		return dlm_posix_lock(ls->ls_dlm, ip->i_no_addr, file, cmd, fl);
+	down_read(&ls->ls_sem);
+	ret = -ENODEV;
+	if (likely(ls->ls_dlm != NULL)) {
+		if (IS_GETLK(cmd))
+			ret = dlm_posix_get(ls->ls_dlm, ip->i_no_addr, file, fl);
+		else if (fl->fl_type == F_UNLCK)
+			ret = dlm_posix_unlock(ls->ls_dlm, ip->i_no_addr, file, fl);
+		else
+			ret = dlm_posix_lock(ls->ls_dlm, ip->i_no_addr, file, cmd, fl);
+	}
+	up_read(&ls->ls_sem);
+	return ret;
 }
 
 static void __flock_holder_uninit(struct file *file, struct gfs2_holder *fl_gh)

fs/gfs2/glock.c

@@ -796,9 +796,8 @@ __acquires(&gl->gl_lockref.lock)
 		}
 		clear_bit(GLF_PENDING_REPLY, &gl->gl_flags);
 
-		if (ret == -EINVAL && gl->gl_target == LM_ST_UNLOCKED &&
-		    target == LM_ST_UNLOCKED &&
-		    test_bit(DFL_UNMOUNT, &ls->ls_recover_flags)) {
+		if (ret == -ENODEV && gl->gl_target == LM_ST_UNLOCKED &&
+		    target == LM_ST_UNLOCKED) {
 			/*
 			 * The lockspace has been released and the lock has
 			 * been unlocked implicitly.

fs/gfs2/incore.h

@@ -657,6 +657,8 @@ struct lm_lockstruct {
 	struct completion ls_sync_wait; /* {control,mounted}_{lock,unlock} */
 	char *ls_lvb_bits;
 
+	struct rw_semaphore ls_sem;
+
 	spinlock_t ls_recover_spin; /* protects following fields */
 	unsigned long ls_recover_flags; /* DFL_ */
 	uint32_t ls_recover_mount; /* gen in first recover_done cb */

fs/gfs2/lock_dlm.c

@@ -312,8 +312,13 @@ static int gdlm_lock(struct gfs2_glock *gl, unsigned int req_state,
 	 */
 
 again:
-	error = dlm_lock(ls->ls_dlm, req, &gl->gl_lksb, lkf, strname,
-			GDLM_STRNAME_BYTES - 1, 0, gdlm_ast, gl, gdlm_bast);
+	down_read(&ls->ls_sem);
+	error = -ENODEV;
+	if (likely(ls->ls_dlm != NULL)) {
+		error = dlm_lock(ls->ls_dlm, req, &gl->gl_lksb, lkf, strname,
+				GDLM_STRNAME_BYTES - 1, 0, gdlm_ast, gl, gdlm_bast);
+	}
+	up_read(&ls->ls_sem);
 	if (error == -EBUSY) {
 		msleep(20);
 		goto again;
@@ -362,8 +367,13 @@ static void gdlm_put_lock(struct gfs2_glock *gl)
 		flags |= DLM_LKF_VALBLK;
 
 again:
-	error = dlm_unlock(ls->ls_dlm, gl->gl_lksb.sb_lkid, flags,
-			   NULL, gl);
+	down_read(&ls->ls_sem);
+	error = -ENODEV;
+	if (likely(ls->ls_dlm != NULL)) {
+		error = dlm_unlock(ls->ls_dlm, gl->gl_lksb.sb_lkid, flags,
+				   NULL, gl);
+	}
+	up_read(&ls->ls_sem);
 	if (error == -EBUSY) {
 		msleep(20);
 		goto again;
@@ -379,7 +389,12 @@ static void gdlm_put_lock(struct gfs2_glock *gl)
 static void gdlm_cancel(struct gfs2_glock *gl)
 {
 	struct lm_lockstruct *ls = &gl->gl_name.ln_sbd->sd_lockstruct;
-	dlm_unlock(ls->ls_dlm, gl->gl_lksb.sb_lkid, DLM_LKF_CANCEL, NULL, gl);
+
+	down_read(&ls->ls_sem);
+	if (likely(ls->ls_dlm != NULL)) {
+		dlm_unlock(ls->ls_dlm, gl->gl_lksb.sb_lkid, DLM_LKF_CANCEL, NULL, gl);
+	}
+	up_read(&ls->ls_sem);
 }
 
 /*
@@ -560,7 +575,11 @@ static int sync_unlock(struct gfs2_sbd *sdp, struct dlm_lksb *lksb, char *name)
 	struct lm_lockstruct *ls = &sdp->sd_lockstruct;
 	int error;
 
-	error = dlm_unlock(ls->ls_dlm, lksb->sb_lkid, 0, lksb, ls);
+	down_read(&ls->ls_sem);
+	error = -ENODEV;
+	if (likely(ls->ls_dlm != NULL))
+		error = dlm_unlock(ls->ls_dlm, lksb->sb_lkid, 0, lksb, ls);
+	up_read(&ls->ls_sem);
 	if (error) {
 		fs_err(sdp, "%s lkid %x error %d\n",
 		       name, lksb->sb_lkid, error);
@@ -587,9 +606,14 @@ static int sync_lock(struct gfs2_sbd *sdp, int mode, uint32_t flags,
 	memset(strname, 0, GDLM_STRNAME_BYTES);
 	snprintf(strname, GDLM_STRNAME_BYTES, "%8x%16x", LM_TYPE_NONDISK, num);
 
-	error = dlm_lock(ls->ls_dlm, mode, lksb, flags,
-			 strname, GDLM_STRNAME_BYTES - 1,
-			 0, sync_wait_cb, ls, NULL);
+	down_read(&ls->ls_sem);
+	error = -ENODEV;
+	if (likely(ls->ls_dlm != NULL)) {
+		error = dlm_lock(ls->ls_dlm, mode, lksb, flags,
+				 strname, GDLM_STRNAME_BYTES - 1,
+				 0, sync_wait_cb, ls, NULL);
+	}
+	up_read(&ls->ls_sem);
 	if (error) {
 		fs_err(sdp, "%s lkid %x flags %x mode %d error %d\n",
 		       name, lksb->sb_lkid, flags, mode, error);
@@ -1315,6 +1339,7 @@ static int gdlm_mount(struct gfs2_sbd *sdp, const char *table)
 	 */
 
 	INIT_DELAYED_WORK(&sdp->sd_control_work, gfs2_control_func);
+	ls->ls_dlm = NULL;
 	spin_lock_init(&ls->ls_recover_spin);
 	ls->ls_recover_flags = 0;
 	ls->ls_recover_mount = 0;
@@ -1349,6 +1374,7 @@ static int gdlm_mount(struct gfs2_sbd *sdp, const char *table)
 	 * create/join lockspace
 	 */
 
+	init_rwsem(&ls->ls_sem);
 	error = dlm_new_lockspace(fsname, cluster, flags, GDLM_LVB_SIZE,
 				  &gdlm_lockspace_ops, sdp, &ops_result,
 				  &ls->ls_dlm);
@@ -1428,10 +1454,12 @@ static void gdlm_unmount(struct gfs2_sbd *sdp)
 
 	/* mounted_lock and control_lock will be purged in dlm recovery */
 release:
+	down_write(&ls->ls_sem);
 	if (ls->ls_dlm) {
 		dlm_release_lockspace(ls->ls_dlm, 2);
 		ls->ls_dlm = NULL;
 	}
+	up_write(&ls->ls_sem);
 
 	free_recover_size(ls);
 }