net/ice: fix L1 check interval

[ upstream commit ff628a22c51f9cc5f69c715005a42456a2aec4f6 ]

For edge cases where the transceiver is physically inserted first and
immediately afterwards the DPDK PF is started the LSC event may occur
outside the current setting for the maximum check interval window. This
change lengthens the check interval to account for this along with other
reported cases where the link event may be longer than 1 second.

Fixes: cf911d9 ("net/ice: support link update")

Signed-off-by: Timothy Miskell <timothy.miskell@intel.com>
Tested-by: Jonathan Tsai <jonathan1.tsai@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>

drivers/net/ice/ice_ethdev.c

@@ -3783,8 +3783,8 @@ ice_atomic_write_link_status(struct rte_eth_dev *dev,
 static int
 ice_link_update(struct rte_eth_dev *dev, int wait_to_complete)
 {
-#define CHECK_INTERVAL 100  /* 100ms */
-#define MAX_REPEAT_TIME 10  /* 1s (10 * 100ms) in total */
+#define CHECK_INTERVAL 50  /* 50ms */
+#define MAX_REPEAT_TIME 40  /* 2s (40 * 50ms) in total */
 	struct ice_hw *hw = ICE_DEV_PRIVATE_TO_HW(dev->data->dev_private);
 	struct ice_link_status link_status;
 	struct rte_eth_link link, old;