Hello
With a 2.6.8.1 kernel, using eql as a module, I have the problem when I
remove the interfaces under eql without using eql_emancipate ioctl (it
is ppp interfaces which are removed sometimes). When I try to add a new
slave interface after having removed several interfaces, I have the
error :
EQL_ENSLAVE failed: No space left on device
I have looked at the drivers/net/eql.c code and it seems that the
num_slaves variable is not decremented properly when a dead interface is
detected. I have modified the code to decrement num_slaves needed (after
each eql_kill_one_slave call) and it works much better. Here is the
patch.
I hope it could be merged.
Please Cc me on answer, I'm not on the list.
Regards.
--
Loïc
---
diff -Nur a/drivers/net/eql.c b/drivers/net/eql.c
--- a/drivers/net/eql.c 2004-08-14 07:37:40.000000000 +0200
+++ b/drivers/net/eql.c 2004-09-03 15:16:41.000000000 +0200
@@ -319,6 +319,7 @@
} else {
/* We found a dead slave, kill it. */
eql_kill_one_slave(slave);
+ queue->num_slaves--;
}
}
return best_slave;
@@ -392,8 +393,10 @@
slave_t *duplicate_slave = NULL;
duplicate_slave = __eql_find_slave_dev(queue, slave->dev);
- if (duplicate_slave != 0)
+ if (duplicate_slave != 0) {
eql_kill_one_slave(duplicate_slave);
+ queue->num_slaves--;
+ }
list_add(&slave->list, &queue->all_slaves);
queue->num_slaves++;
@@ -472,6 +475,7 @@
if (slave) {
eql_kill_one_slave(slave);
+ eql->queue.num_slaves--;
ret = 0;
}
}
-
: send the line "unsubscribe linux-net" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
