Reducing multiplexing latencies still further in wifi

For Developers
141

Any comments on performance?

142

Honestly, not a big difference in my case. It helps with the low latency I require in my network.

1 Like
143

Thank you very much, Toke <3

4 Likes
144

Yeah, seemed appropriate. To Dave! :clinking_beer_mugs:

10 Likes
145

I'm using it too. I haven't done any measurements, but by simply browsing the web and doing normal stuff. It feels like "time-to-first-byte" is faster if you catch my drift. Before there was a little "lag" from pressing enter to the page loaded and displayed.

--- a/net/mac80211/sta_info.c
+++ b/net/mac80211/sta_info.c
@@ -2889,15 +2889,7 @@ unsigned long ieee80211_sta_last_active(

 static void sta_update_codel_params(struct sta_info *sta, u32 thr)
 {
-	if (thr && thr < STA_SLOW_THRESHOLD * sta->local->num_sta) {
-		sta->cparams.target = MS2TIME(50);
-		sta->cparams.interval = MS2TIME(300);
-		sta->cparams.ecn = false;
-	} else {
-		sta->cparams.target = MS2TIME(20);
-		sta->cparams.interval = MS2TIME(100);
-		sta->cparams.ecn = true;
-	}
+	return;
 }

 void ieee80211_sta_set_expected_throughput(struct ieee80211_sta *pubsta,
5 Likes
146

This must be an unpopular patch. No one on the mailing list ever acknowledges it (Dave’s or yours).

147

Actually, it was just merged today: https://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless-next.git/commit/?id=4876376988081d636a4c4e5f03a5556386b49087

12 Likes
148

That’s amazing! I looked earlier today, then grumbled to myself for a while, then posted. Good times are ahead! Thanks Toke!

4 Likes
149

I think it'll be what it is...

It's not like Dave Taht can defend/debug it post facto...

150

You’re such a putz. It certainly is what it is.

3 Likes
151

Dave 14305 via OpenWrt Forum mail@forum.openwrt.org writes:

That’s amazing! I looked earlier today, then grumbled to myself for a
while, then posted. Good times are ahead! Thanks Toke!

You're very welcome :slight_smile:

152

Will this patch be backported to OpenWrt main/master and 24.10 branches ? Thank you.

153

"Keshav A. via OpenWrt Forum" mail@forum.openwrt.org writes:

Will this patch be backported to OpenWrt main/master and 24.10
branches ? Thank you.

I expect it will land in openwrt master whenever they do a rebase on
upstream mac80211. This patch will only land in upstream version 6.16,
though, so not sure how long that will take.

To get it into openwrt sooner, someone will have to open a pull request
with an explicit patch...

6 Likes
154

I'm just saying that dtaht had keen insight into things - and that insight was over years of time.

It's hard to find someone that could sort things out and keep it going.

Yes, I miss his contributions as many do...

1 Like
156

Any ETA when it's in, please?

157

mac80211 just got updated to 6.14.11, but the upstream revert wasn’t backported to 6.14, so the next mac80211 bump to 6.15.4 or higher will be the winner (see nbd’s comments in the PR).

https://cdn.kernel.org/pub/linux/kernel/v6.x/ChangeLog-6.15.4

2 Likes