aboutsummaryrefslogtreecommitdiff
path: root/gcc/tree-eh.c
diff options
context:
space:
mode:
Diffstat (limited to 'gcc/tree-eh.c')
-rw-r--r--gcc/tree-eh.c6
1 files changed, 6 insertions, 0 deletions
diff --git a/gcc/tree-eh.c b/gcc/tree-eh.c
index 0f5931e..3ed92a5 100644
--- a/gcc/tree-eh.c
+++ b/gcc/tree-eh.c
@@ -3369,6 +3369,12 @@ unsplit_eh (eh_landing_pad lp)
return false;
}
+ /* The new destination block must not already be a destination of
+ the source block, lest we merge fallthru and eh edges and get
+ all sorts of confused. */
+ if (find_edge (e_in->src, e_out->dest))
+ return false;
+
/* ??? I can't imagine there would be PHI nodes, since by nature
of critical edge splitting this block should never have been
a dominance frontier. If cfg cleanups somehow confuse this,