diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
index 670104628ddb..156d92b94525 100644
--- a/fs/f2fs/segment.c
+++ b/fs/f2fs/segment.c
@@ -1282,6 +1282,15 @@ static int __submit_discard_cmd(struct f2fs_sb_info *sbi,
 	}
 #endif
 
+	/*
+	 * stop issuing discard for any of below cases:
+	 * 1. device is conventional zone, but it doesn't support discard.
+	 * 2. device is regulare device, after snapshot it doesn't support
+	 * discard.
+	 */
+	if (!bdev_max_discard_sectors(bdev))
+		return -EOPNOTSUPP;
+
 	trace_f2fs_issue_discard(bdev, dc->di.start, dc->di.len);
 
 	lstart = dc->di.lstart;
diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index f05d0e43db9e..b72fa103b963 100644
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -1735,6 +1735,18 @@ static int f2fs_freeze(struct super_block *sb)
 
 static int f2fs_unfreeze(struct super_block *sb)
 {
+	struct f2fs_sb_info *sbi = F2FS_SB(sb);
+
+	/*
+	 * It will update discard_max_bytes of mounted lvm device to zero
+	 * after creating snapshot on this lvm device, let's drop all
+	 * remained discards.
+	 * We don't need to disable real-time discard because discard_max_bytes
+	 * will recover after removal of snapshot.
+	 */
+	if (test_opt(sbi, DISCARD) && !f2fs_hw_support_discard(sbi))
+		f2fs_issue_discard_timeout(sbi);
+
 	clear_sbi_flag(F2FS_SB(sb), SBI_IS_FREEZING);
 	return 0;
 }