On 05/03/2025 22:20, Sebastian Andrzej Siewior wrote:
On 2025-03-05 21:44:23 [+0200], Tariq Toukan wrote:
Hi,
Hi,
Thanks for your patch.
IIUC you remove here the per-ring page_pool stats, and keep only the summed
stats.
I guess the reason for this is that the page_pool strings have no per-ring
variants.
59 static const char pp_stats[][ETH_GSTRING_LEN] = {
60 "rx_pp_alloc_fast",
61 "rx_pp_alloc_slow",
62 "rx_pp_alloc_slow_ho",
63 "rx_pp_alloc_empty",
64 "rx_pp_alloc_refill",
65 "rx_pp_alloc_waive",
66 "rx_pp_recycle_cached",
67 "rx_pp_recycle_cache_full",
68 "rx_pp_recycle_ring",
69 "rx_pp_recycle_ring_full",
70 "rx_pp_recycle_released_ref",
71 };
Is this the only reason?
Yes. I haven't seen any reason to keep it. It is only copied around.
I like the direction of this patch, but we won't give up the per-ring
counters. Please keep them.
Hmm. Okay. I guess I could stuff a struct there. But it really looks
like waste since it is not used.
Of course they are used.
Per-ring (per-pool) counters are exposed via ethtool -S.
