On 6/1/22 09:39, Min Hu (Connor) wrote:
From: Huisong Li <lihuis...@huawei.com>
The dev->data->mac_addrs[0] will be changed to a new MAC address when
applications modify the default MAC address by
rte_eth_dev_default_mac_addr_set(). However, if the new default one has
been added as a non-default MAC address by rte_eth_dev_mac_addr_add(), the
the rte_eth_dev_default_mac_addr_set() doesn't remove it from the mac_addrs
list. As a result, one MAC address occupies two indexes in the list. Like:
add(MAC1)
add(MAC2)
add(MAC3)
add(MAC4)
set_default(MAC3)
default=MAC3, filters=MAC1, MAC2, MAC3, MAC4
In addition, some PMDs, such as i40e, ice, hns3 and so on, do remove the
old default MAC when set default MAC. If user continues to do
set_default(MAC5), and the mac_addrs list is default=MAC5, filters=(MAC1,
MAC2, MAC3, MAC4). At this moment, user can still view MAC3 from the list,
but packets with MAC3 aren't actually received by the PMD.
IMHO, the main problem is inconsistency which exists right now.
rte_eth_dev_mac_addr_add() checks for duplicate MAC addition
including the default one (index zero) and extends the entry
pool mask (including zero entry case).
However, the patch above does not extend zero entry pool mask.
So, the result will depend on order which is bad:
A. Set default to A, add MAC A with pool 2 => pool mask has 2
B. Add MAC A with pool 2, set default to A => pool mask is empty
Am I missing something in the code?
What is the right/intended behaviour?
Fixes: 854d8ad4ef68 ("ethdev: add default mac address modifier")
Cc: sta...@dpdk.org
Signed-off-by: Huisong Li <lihuis...@huawei.com>
Signed-off-by: Min Hu <humi...@huawei.com>
---
lib/ethdev/rte_ethdev.c | 39 +++++++++++++++++++++++++++++++++++++--
1 file changed, 37 insertions(+), 2 deletions(-)
diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c
index 46c088dc88..fc9ca8d6fd 100644
--- a/lib/ethdev/rte_ethdev.c
+++ b/lib/ethdev/rte_ethdev.c
@@ -4260,7 +4260,10 @@ rte_eth_dev_mac_addr_remove(uint16_t port_id, struct
rte_ether_addr *addr)
int
rte_eth_dev_default_mac_addr_set(uint16_t port_id, struct rte_ether_addr
*addr)
{
+ uint64_t mac_pool_sel_bk = 0;
struct rte_eth_dev *dev;
+ uint32_t pool;
+ int index;
int ret;
RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
@@ -4278,16 +4281,48 @@ rte_eth_dev_default_mac_addr_set(uint16_t port_id,
struct rte_ether_addr *addr)
RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->mac_addr_set, -ENOTSUP);
+ /*
+ * If the address has been added as a non-default MAC address by
+ * rte_eth_dev_mac_addr_add API, it should be removed from
+ * dev->data->mac_addrs[].
+ */
+ index = eth_dev_get_mac_addr_index(port_id, addr);
+ if (index > 0) {
+ /* remove address in NIC data structure */
+ mac_pool_sel_bk = dev->data->mac_pool_sel[index];
+ ret = rte_eth_dev_mac_addr_remove(port_id, addr);
+ if (ret < 0) {
+ RTE_ETHDEV_LOG(ERR,
+ "Delete MAC address from the MAC list of ethdev port
%u.\n",
+ port_id);
+ return ret;
+ }
+ /* reset pool bitmap */
+ dev->data->mac_pool_sel[index] = 0;
+ }
+
ret = (*dev->dev_ops->mac_addr_set)(dev, addr);
if (ret < 0)
- return ret;
+ goto back;
/* Update default address in NIC data structure */
rte_ether_addr_copy(addr, &dev->data->mac_addrs[0]);
return 0;
-}
+back:
+ if (index > 0) {
+ pool = 0;
+ do {
+ if (mac_pool_sel_bk & UINT64_C(1))
+ rte_eth_dev_mac_addr_add(port_id, addr, pool);
Don't we want to have at least error logs in the case of rollback
failure here?
+ mac_pool_sel_bk >>= 1;
+ pool++;
+ } while (mac_pool_sel_bk);
Please, compare vs 0 explicitly.
+ }
+
+ return ret;
+}
/*
* Returns index into MAC address array of addr. Use 00:00:00:00:00:00 to find
