04:26 <KGB-0> https://tests.reproducible-builds.org/openwrt/openwrt_sunxi.html has been updated. (0% images and 99.9% packages reproducible in our current test framework.)

05:51 <rmilecki> what is the basic SoC variant out of following: mt7620a, mt7620n, mt7621, mt7628an?

05:51 <rmilecki> i expect one to be basic one and other being variations, is that correct?

05:57 <rmilecki> (i see i missed mt7622, mt7623a and mt7629)

06:13 <stintel> Mangix: try reverting the last hostapd bump

06:14 <stintel> my phone is constantly disconnecting with that as well, reverting it made the issue go away, haven't had time to dig deeper so for now I'm just running with 2023-03-29

06:35 <slh64> rmilecki: I wouldn't phrase it that way, or would you say the same about bcm4704 vs bcm4908?

06:36 <rmilecki> slh64: bcm4704 is old MIPS while bcm4908 is 4 x ARM CPU

06:36 <rmilecki> totally different

06:36 <rmilecki> but I thought those Mediatek chipsets are more similar

06:36 <slh64> rmilecki: exactly, same about those ralink/ mediatek ones

06:37 <rmilecki> so all of them are quite different?

06:37 <rmilecki> all those I listed?

06:37 <slh64> the ones in your first list are mips, mt7623 is ARMv7, mt7622 is ARMv8

06:37 <rmilecki> oh

06:37 <slh64> different generations at least

06:38 <rmilecki> ok, thanks for info

06:38 <slh64> well, apart from mt7620a and mt7620n (the later is stripped down)

06:39 <rmilecki> so mt7620n is a stripped down version of mt7620a?

06:39 <slh64> mt7628 is inbetween mt7620 and mt7621, closer to mt7620, but usually with mt76 5 GHz wlan

06:39 <slh64> yes

06:40 <rmilecki> fun...

06:40 <slh64> mt7620 560 MHz single core, mt7621a 880 MHz dual-core (with hyperthreading on top)

06:44 <slh64> mt7620/ mt7628 are still mips24KEc, mt7621 is mips1004Kc

06:44 <slh64> https://wikidevi.wi-cat.ru/MediaTek#General_Purpose has a quite nice overview

06:45 <rmilecki> slh64: thanks!

06:45 <slh64> ...and the interesting mediatek chipsets would now be filogic 820/ 830 and the wifi7 base SOC filogic 880 (all ARMv8)

06:53 <slh64> as Mediatek's SOCs usually don't have (had) integrated wireless, in favour of PCIe, you'll often see mix-and-match, newer wireless chipsets on older SOCs (first - and for the low-cost market), only mt7622 started integrating 802.11n 2.4 GHz wireless (meaning the cheap mips1004Kc mt7621a SOCs often have mt7915 DBDC, ax in both bands - while the better ARMv8 based mt7622 SOC typically only has 11n 2.4 GHz and 11ax 5 GHz wifi)

07:37 <PaulFertser> But DBDC is 2x2 on both bands while when you use mt7622 for 802.11n on 2.4 GHz you usually get 4x4 802.11ax on 5 GHz.

07:39 <PaulFertser> Talking about DBDC on Mediatek... What's up with the mt76 driver maintenance? I know nbd isn't the only maintainer, there're two others plus two reviewers. And I understand Felix is busy with real cool useful generic stuff. But what about the others, why my trivial patch isn't getting reviewed and merged?

07:42 <PaulFertser> mt7620 wireless is handled by the rt28xx driver, not mt76.

07:49 <mrkiko> yeah, true. In some cases, when there is an mt7610E module in the mix, then mt76 is also involved. But SOC wi-fi is handled by rt28xx or something

07:50 <slh64> that's what I meant with something inbetween

07:50 <mrkiko> slh64: don't know if what you say is still accurate, for newer socs - e.g.: from mt7622 onwards, it's startring to get in SOC

07:52 <slh64> mt7622 has wifi4 4x4 2.4 GHz integrated, indeed

07:52 <mrkiko> slh64: but if I understood things right, mt7688 and so on have boyth wifi frequencies support. If I am right

07:54 <mrkiko> but - good morning to all!!

07:56 <slh64> I'm not quite sure how integrated filogic 8x0 is, there is a quite large support chip for either radio, looks a bit large for just a phy

08:00 <slh64> at least filogic 880 has wifi7 via PCIe (again)

08:03 <slh64> and filogic 820/ 830 is also using PCIe, mt7975n and mt7975p

08:16 <PaulFertser> Thank you and sorry for nagging.

10:50 <KGB-2> https://tests.reproducible-builds.org/openwrt/openwrt_x86.html has been updated. (100.0% images and 100.0% packages reproducible in our current test framework.)

11:40 <ynezz> stintel: BTW you should probably report that as it seems, that nobody is aware about the issue(s) and that hostapd update just got backported to 23.05 via 8d7d9aa4a463

11:54 <nbd> rmilecki: found a use-after-free bug in hostapd and fixed it

11:54 <nbd> rmilecki: please test the latest version from my staging tree

11:54 <rmilecki> nbd: sounds fun debugging

11:55 <rmilecki> nbd: i'm in the middle of bisecting between 4.4 and master, i'll test later today or over weekend, depending how it goes

11:55 <rmilecki> fun on my side too ;)

11:55 <nbd> between 4.4 and master ... wow, that's a huge range

11:56 <nbd> i couldn't reproduce the issue that you were seeing on reload, but i did get some quirky behavior

11:56 <nbd> and a rare crash that led me to the use-after-free bug

12:27 <mrkiko> slh64: thanks for the info, didn't know that

13:01 <owrt-2203-builds> Build [#264](https://buildbot.openwrt.org/openwrt-22.03/images/#builders/29/builds/264) of `mvebu/cortexa72` failed.

13:12 <Slimey> good day

13:56 <mrkiko> Slimey: the same to you :)

15:46 <djfe> rmilecki: MT7621ST is actually 880MHz Dual Core without Hyperthreading (the Netgear R6220 you listed)

15:46 <djfe> MT7621DAT is the hyperthreaded variant

15:46 <djfe> there are two other names as well, I don't know all specifics.

15:47 <djfe> mt7628 is low end

15:47 <djfe> mt7621 is the successor of mt7620 afaik

15:55 <mrkiko> mt7628 seems kinda an mt7620 with more mediatek and less ralink design

16:33 <owrt-2203-builds> Build [#270](https://buildbot.openwrt.org/openwrt-22.03/images/#builders/22/builds/270) of `bcm4908/generic` failed.

17:44 <KGB-0> https://tests.reproducible-builds.org/openwrt/openwrt_kirkwood.html has been updated. (100.0% images and 99.9% packages reproducible in our current test framework.)

18:01 <Mangix> stintel: meh this is a production router. Was running snapshot mostly because of the iPhone fix in mt76. I guess I'll just downgrade to 22.03 and use the 160MHz workaround

18:29 <owrt-2203-builds> Build [#265](https://buildbot.openwrt.org/openwrt-22.03/images/#builders/29/builds/265) of `mvebu/cortexa72` completed successfully.