On 1/16/26 14:02, Bruce Richardson wrote:
On Mon, Jan 12, 2026 at 02:53:11PM +0100, Thierry Herbelot wrote:
When adding the i40e PMD, the rss_hash_conf_get function was changed to
update the RSS key len.
The corresponding functions for ixgbe & e1000 were not updated to
follow the new convention.
Fixes: 8a387fa85f02 ("ethdev: more RSS flags")
Signed-off-by: Thierry Herbelot <[email protected]>
---
drivers/net/intel/e1000/igb_rxtx.c | 1 +
drivers/net/intel/e1000/igc_ethdev.c | 1 +
drivers/net/intel/ixgbe/ixgbe_rxtx.c | 1 +
3 files changed, 3 insertions(+)
diff --git a/drivers/net/intel/e1000/igb_rxtx.c
b/drivers/net/intel/e1000/igb_rxtx.c
index cdd7a3338f61..d5c2bc013702 100644
--- a/drivers/net/intel/e1000/igb_rxtx.c
+++ b/drivers/net/intel/e1000/igb_rxtx.c
@@ -2042,6 +2042,7 @@ int eth_igb_rss_hash_conf_get(struct rte_eth_dev *dev,
hash_key[(i * 4) + 2] = (rss_key >> 16) & 0x000000FF;
hash_key[(i * 4) + 3] = (rss_key >> 24) & 0x000000FF;
}
+ rss_conf->rss_key_len = 40;
}
/* Get RSS functions configured in MRQC register */
diff --git a/drivers/net/intel/e1000/igc_ethdev.c
b/drivers/net/intel/e1000/igc_ethdev.c
index b9c91d2446c0..9f6ada3bc6cd 100644
--- a/drivers/net/intel/e1000/igc_ethdev.c
+++ b/drivers/net/intel/e1000/igc_ethdev.c
@@ -2559,6 +2559,7 @@ eth_igc_rss_hash_conf_get(struct rte_eth_dev *dev,
/* read RSS key from register */
for (i = 0; i < IGC_HKEY_MAX_INDEX; i++)
hash_key[i] = E1000_READ_REG_LE_VALUE(hw,
E1000_RSSRK(i));
+ rss_conf->rss_key_len = IGC_HKEY_MAX_INDEX * sizeof(uint32_t);
}
/* get RSS functions configured in MRQC register */
diff --git a/drivers/net/intel/ixgbe/ixgbe_rxtx.c
b/drivers/net/intel/ixgbe/ixgbe_rxtx.c
index a7583c178a14..a2c5ffe0b3b0 100644
--- a/drivers/net/intel/ixgbe/ixgbe_rxtx.c
+++ b/drivers/net/intel/ixgbe/ixgbe_rxtx.c
@@ -3756,6 +3756,7 @@ ixgbe_dev_rss_hash_conf_get(struct rte_eth_dev *dev,
hash_key[(i * 4) + 2] = (rss_key >> 16) & 0x000000FF;
hash_key[(i * 4) + 3] = (rss_key >> 24) & 0x000000FF;
}
+ rss_conf->rss_key_len = 40;
}
Acked-by: Bruce Richardson <[email protected]>
Code is correct. However, do you think it would be better to use
IGB_HKEY_MAX_INDEX and IXGBE_HKEY_MAX_INDEX in those driver assignments? I
realise the loops don't use them, but they probably should.
Hello Bruce,
I will resend the patch, using the symbolic constants, as suggested.
Thanks
Thierry
/Bruce
--
Thierry Herbelot
Senior Software Engineer
http://www.6wind.com/
Follow us:
https://www.linkedin.com/company/6wind/
https://twitter.com/6WINDsoftware
https://www.youtube.com/user/6windsoftware
