[PATCH] kcmp: Fix ret code comparison

[Cyrill Gorcunov]

In the early draft of kcmp syscall it has been returning
[-1|0|1] values but finally [0|1|2] were merged into the
kernel, but I forgot to update the criu code. The good
thing is that because we're using rbtree the kcmp results
are still sorted and tree is balanced but sometime we may
take a wrong branch generating new ID even if the object
is present in the tree which eventually may lead to dump
faulure.

Reported-by: Deyan Doychev <deyan at 1h.com>
Signed-off-by: Cyrill Gorcunov <gorcunov at openvz.org>
---
 kcmp-ids.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)

diff --git a/kcmp-ids.c b/kcmp-ids.c
index 3f75cff33f46..f082d9730f59 100644
--- a/kcmp-ids.c
+++ b/kcmp-ids.c
@@ -149,12 +149,18 @@ static u32 kid_generate_sub(struct kid_tree *tree, struct kid_entry *e,
 				this->elem.idx, elem->idx);
 
 		parent = *new;
-		if (ret < 0)
+		if (ret == 1)
 			node = node->rb_left, new = &((*new)->rb_left);
-		else if (ret > 0)
+		else if (ret == 2)
 			node = node->rb_right, new = &((*new)->rb_right);
-		else
+		else if (ret == 0)
 			return this->subid;
+		else {
+			pr_err("kcmp failed: pid (%d %d) type %u idx (%u %u) ret %d\n",
+			       this->elem.pid, elem->pid, tree->kcmp_type,
+			       this->elem.idx, elem->idx, ret);
+			return 0;
+		}
 	}
 
 	sub = alloc_kid_entry(tree, elem);
-- 
1.8.3.1


--Rgf3q3z9SdmXC6oT--