[v3,4/5] power: reduce memory footprint of per-lcore state

Message ID 20250210213158.4165701-5-david.marchand@redhat.com (mailing list archive)
State Accepted
Delegated to: David Marchand
Headers
Series Defer lcore variables allocation |

Checks

Context Check Description
ci/checkpatch success coding style OK

Commit Message

David Marchand Feb. 10, 2025, 9:31 p.m. UTC
Now that the per-lcore state was moved into a lcore variable,
there is no reason to align a per-lcore state on a cache line to avoid
false sharing.
Remove this alignment and save a few bytes.

Fixes: 130643319579 ("power: keep per-lcore state in lcore variable")

Signed-off-by: David Marchand <david.marchand@redhat.com>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
Acked-by: Frode Nordahl <frode.nordahl@canonical.com>
---
 lib/power/rte_power_pmd_mgmt.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
  

Patch

diff --git a/lib/power/rte_power_pmd_mgmt.c b/lib/power/rte_power_pmd_mgmt.c
index 369ce3c354..8ec925bd65 100644
--- a/lib/power/rte_power_pmd_mgmt.c
+++ b/lib/power/rte_power_pmd_mgmt.c
@@ -56,7 +56,7 @@  struct queue_list_entry {
 	const struct rte_eth_rxtx_callback *cb;
 };
 
-struct __rte_cache_aligned pmd_core_cfg {
+struct pmd_core_cfg {
 	TAILQ_HEAD(queue_list_head, queue_list_entry) head;
 	/**< List of queues associated with this lcore */
 	size_t n_queues;