https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=283903
--- Comment #27 from Bjoern A. Zeeb <b...@freebsd.org> --- (In reply to Guillaume Outters from comment #26) > D3 (softclock_call_cc+0x12c > rtw_tx_report_purge_timer+0x8e) seems to be a > degraded > mode under load (no time to call D1 anymore, it is stacked for later?), which > works well > but **ONLY COMPENSATES ONE OF THE TWO FREEs NORMALLY CALLED WITHIN D1**. I believe that was my initial shot in the dark and I wonder if you are still running with the review from there and if not (half of it was committed to main the other day, the LinuxKPI parts) if you want to apply the rtw88 parts. It doesn't solve things but I wonder if it would "behave better". As for the other parts, not forgotten but my urgencies are currently elsewhere for a few more days. But given I'll have to go back to rtw88 and rtw89 after that (likely next week) I had planned to have a look then. Sorry it's currently annoying behavior for you. If you have more findings that would certainly be helpful. By now I wonder if the card/firmware simply doesn't send TX reports anymore and D3 doesn't simply free them; mind you there is a problem with and ID in there as well which only takes 256 items and then wraps around. Maybe more instrumentation is needed around that... -- You are receiving this mail because: You are on the CC list for the bug.
