Hello Florian,
ecrt_master_get_slave will receive a NULL pointer from
ec_master_find_slave_const when the query doesn't succeed (e.g. a
nonexistent slave_position) - but tries to fill the slave info response
nonetheless, resulting in kernel BUG: unable to handle kernel NULL
pointer dereference.
for your consideration, this patch addresses the issue:
hg diff master/master.c
diff -r 2eff7c993a63 master/master.c
--- a/master/master.c Tue Feb 12 17:31:08 2013 +0100
+++ b/master/master.c Thu May 01 16:32:58 2014 -0700
@@ -2624,6 +2624,7 @@
{
const ec_slave_t *slave;
unsigned int i;
+ int status = 0;
if (down_interruptible(&master->master_sem)) {
return -EINTR;
@@ -2631,6 +2632,11 @@
slave = ec_master_find_slave_const(master, 0, slave_position);
+ if (slave == NULL) {
+ status = -ENOENT;
+ goto out_get_slave;
+ }
+
slave_info->position = slave->ring_position;
slave_info->vendor_id = slave->sii.vendor_id;
slave_info->product_code = slave->sii.product_code;
@@ -2667,9 +2673,11 @@
slave_info->name[0] = 0;
}
+out_get_slave:
+
up(&master->master_sem);
- return 0;
+ return status;
}
/*****************************************************************************/
Regards,
Matthieu
--
Matthieu Bec GMTO Corp
cell : +1 626 354 9367 251 S Lake Ave, Suite 300
phone: +1 626 204 0527 Pasadena, CA 91101
_______________________________________________
etherlab-dev mailing list
etherlab-dev@etherlab.org
http://lists.etherlab.org/mailman/listinfo/etherlab-dev
