On Tue, Sep 16, 2025 at 03:22:54PM +0200, Paolo Abeni wrote:

On 9/15/25 5:58 AM, Erni Sri Satya Vennela wrote:

quoted

Report standard counter stats->rx_missed_errors
using hc_rx_discards_no_wqe from the hardware.

Add a dedicated workqueue to periodically run
mana_query_gf_stats every 2 seconds to get the latest
info in eth_stats and define a driver capability flag
to notify hardware of the periodic queries.

To avoid repeated failures and log flooding, the workqueue
is not rescheduled if mana_query_gf_stats fails.

Can the failure root cause be a "transient" one? If so, this looks like
a dangerous strategy; is such scenario, AFAICS, stats will be broken
until the device is removed and re-probed.

/P

After internal discussion, We are planning to fix this issue following
the below approach:

Stop rescheduling the work queue only upon detecting HWC timeout.
In this case:
1. Reset all stats to zero to avoid stale reporting.
2. Introduce a driver flag to detect the first occurrence of HWC timeout.
3. Log a warn_once during subsequent calls to mana_get_stats64 to signal
   the issue.

Thanks,
Vennela