**** BEGIN LOGGING AT Wed Mar 10 02:59:56 2021 Mar 10 03:04:24 Can anyone point me in the direction to where I can find the stock compat string for the the ER10x on the stock system so I can put it into the SUPPORTED_DEVICES string in the .mk? Mar 10 03:20:45 /tmp/device_model? Mar 10 03:22:57 rsalvaterra: both gdma and hsdma are marked "disabled" in mt7621.dtsi, does any device actually enable them? Mar 10 03:36:16 >KGB-1< https://tests.reproducible-builds.org/openwrt/openwrt_ar71xx.html has been updated. (99.1% images and 98.4% packages reproducible in our current test framework.) Mar 10 05:40:47 Well.. I've got Openwrt installed, the lower ports set, running 5.10.20 on this ER10x.. Now the fun can begin Mar 10 05:41:08 although I'm not impressed with the throughput, but I think that's a known limitation Mar 10 06:19:42 Grommish: there is two 1gbe links and to forward something close to it - it needs hw acceleration. mt7621 has it Mar 10 06:20:00 device is supposed to be Mar 10 06:21:15 to be/ to have Mar 10 06:22:22 Yeah, I'v got ports 0-4 working, 5-9 not yet, but the port 2 to 0 speedtest maxes out at under 600 down Mar 10 06:22:28 Grommish: stock have nothing to do about it. you need a working device tree where you define your device name that later be used across the tree. mt7621. Mar 10 06:23:19 damex: Well, I was being particularly dense about the device. Ubiquity is hash checking under v2 Mar 10 06:23:26 second switch is RTL8367RB Mar 10 06:23:37 damex: So I can't just load the bin via the online internet or add system image command Mar 10 06:24:44 it might work like on er-x but i have no clue what exactly is needed to make it work Mar 10 06:25:26 check er-x implementation - it works upgrading from some edgeos firmwares. not all of them Mar 10 06:27:11 Yeah, v1 I think you could, v2 you can't Mar 10 06:31:58 damex: I just tftp load the initial image and be done with it.. Now, I need to figure out hte rest of it Mar 10 06:34:42 damex: mt7530 mdio-bus:1f: configuring for fixed/rgmii link mode rgmii seems to be up, but nothing on the ports Mar 10 07:27:51 Grommish: because the switch is on port 5 of the MT7621 switch, you have to the RGMII2 group to function RGMII. Mar 10 07:30:49 brb, reboot the system Mar 10 07:31:07 Rene__: I'm showing the mac@1 under the ethernet as disabled.. I'm going to try to set it to Okay and see what happens Mar 10 07:37:25 Grommish: when i was testing the 5-port ERX a few years ago i would see either 500Mbps, 750Mbps or about a gig on different runs of iperf through a NAT, but it was consistent within the run Mar 10 07:37:56 the er10x seems to be unavailable at the moment, i was thinking of buying one, but i guess not Mar 10 07:39:45 I've got the stock dts I pulled and I'm trying to see what is different from the dts I pulled from the openwrt run **** BEGIN LOGGING AT Wed Mar 10 07:45:51 2021 Mar 10 07:59:22 Rene__: I'm showing the mac@1 under the ethernet as disabled.. I'm going to try to set it to Okay and see what happens <-- Grommish Mar 10 08:02:04 https://github.com/openwrt/openwrt/blob/master/target/linux/ramips/dts/mt7621_ubnt_edgerouter-x.dtsi#L108 remove the "RGMII2" from the list so that the RGMII2 pin get the funtion RGMII Mar 10 08:03:38 mangix: AFAIK disabling mac1 doesn't change RGMII2 mode. Mar 10 08:05:34 Rene__: Will try that here shortly :) Mar 10 08:07:11 Grommish: mangix: Hmm I am wrong. RGMII2 function only has to change to RGMII function when mac@1 wants to talk to port 5. Mar 10 08:07:57 Rene__: This is the stock dts I pulled from EdgeOS, if it helps.. but it's based on the 4.14 kernel.. https://gist.github.com/Grommish/7886f7be0d07b59978226844f2f10d9e Mar 10 08:08:40 I'm trying to see what matches and what doesn't from the 5.10 dts for mt7621.dtsi Mar 10 08:10:22 I've had no experience with the device trees.. my primary device has it built in and I never had to deal with it Mar 10 08:13:48 Oh FFS.. Well..Regardless of anything else.. it doesn't work if the driver isn't turned on in the kernel Mar 10 08:13:57 RTL8367B_PHY [=n] Mar 10 08:14:03 Grommish: I think this is the setup code for er-10x https://github.com/Lochnair/kernel_e50/blob/v2.0.9/master/drivers/net/ethernet/mediatek/gsw_mt7621.c#L283 Mar 10 08:14:31 lipnitsk: Yes, some do, so they're probably disabled for some reason, I don't know which. Mar 10 08:18:53 So, I'll turn it on and see if it actually does anything :D Mar 10 08:20:33 Grommish: do you have the dmesg of ubnt firmware? look for "change HW-TRAP to 0x" to see how the mt7530 is configured. Mar 10 08:21:07 Rene__: Sure.. one minute.. I'm building and I'll flash and check Mar 10 08:30:53 mangix: I now have a 100 % working configuration with Linux 5.10 on ramips/mt7621. I'm going to apply your musl series, make dirclean an build a new image. Hopefully the problems I've been seeing were caused by a dirty tree. Mar 10 08:33:33 Rene__: Bootlogs.. OpenWrt on top, Stock EdgeOS at the bottom Mar 10 08:33:35 Rene__: https://gist.github.com/Grommish/00aeb17c88b6ce92d9700435c0a4f290 Mar 10 08:39:26 Grommish: thanks Mar 10 08:44:34 Hmm… I just noticed this, probably not fatal, but ominous, nonetheless… Mar 10 08:44:36 [ 1.044918] BUG: Bad rss-counter state mm:(ptrval) type:MM_ANONPAGES val:1 Mar 10 08:45:43 rsalvaterra: i'll join in the mt7621 fun in a month Mar 10 08:46:57 mangix: Nice! You're gonna love/hate it. :P Mar 10 08:51:53 I've had mt7621 devices before Mar 10 08:52:11 it's a mixed bag Mar 10 08:55:54 lipnitsk: On what device have you seen the bad rss-counter bug? Grommish, do you see it on the ER10x? I can't find it in your boot log. Mar 10 08:56:58 rsalvaterra: I'm not using the 1.2.2 if that makes a difference.. I'm using the main branch and the testing kernel Mar 10 08:58:19 Grommish: I'm not using musl 1.2.2 either (not yet, at least). Mar 10 08:58:45 And this is strictly a kernel issue, no userspace involved. Mar 10 09:05:45 russell--: re: the teql hotplug discussion in the tc-tiny PR, are you OK with simply dropping the '+tc' dependency as I suggested? Mar 10 09:07:42 guidosarducci: Yes, please. It's weird, to say the least. :P Mar 10 09:07:46 Rene__: Getting closer: https://gist.github.com/Grommish/39f61eb7e2a3141da54d790e9fbcff0a Mar 10 09:09:22 Grommish: Based on the HW-TRAP settings from stock firmware. P5 is used and set in RGMII mode. Based on the gsw_mt7621 code no extra delay is needed so phy-mode = "rgmii"; Mar 10 09:17:44 Rene__: in the mt7621.dtsi, pcie: pcie@1e140000 is status = "disabled"; Would that make a difference? Mar 10 09:18:15 Sorry, again, my tree knowledge is almost non-existant Mar 10 09:19:10 Grommish: no, I am also not a dt expert, but pcie has no part in the ethernet part. Mar 10 09:19:32 rsalvaterra: I knew you'd say that, but really it's more about what's better/correct rather than less weird. Things aren't ideal now with kitchen-sink packages like kmod-sched, but the proposed change is still an improvement. jow was pinged as part of the prior discussion so I hope to hear from him too. Mar 10 09:20:55 Rene__: This is the current dts I'm using.. https://gist.github.com/Grommish/21397fee2a2b4c6e4ddf1d6d7ad7ee47 Mar 10 09:23:31 Grommish: Is the realtek rtl8367b a mainline kernel driver? Mar 10 09:24:29 guidosarducci: Sure, it's just my €0,02. ;) The kmod-sched package really needs to be split up (and probably sqm-scripts, since most people just want to "eat" cake)… Mar 10 09:25:18 guidosarducci: it's up to the taste police ... i still kind of like my add a package for the script, it's no different size-wise than your patch, you need all the parts one way or the other. secondly, if you really want to save space by splitting the kmod, it doesn't look that difficult. Mar 10 09:27:06 russell--: The problem is, if you split kmod-sched, sqm-scripts will still depend on all the split packages, because I contains scripts which use all the schedulers, even if you're not using them at all. Mar 10 09:27:34 no different than now Mar 10 09:27:45 rsalvaterra: not a problem: compat virtual packages, etc. Mar 10 09:28:42 and consumers of those packages would *want* to adjust their own dependencies. Mar 10 09:28:51 Rene__: in 5.10, yes Mar 10 09:30:27 Rene__: https://gist.github.com/Grommish/a645dcbbf5d1457250d5768f85b891e4 I've enabled the drivers anyway Mar 10 09:31:27 so, mt7621 owners Mar 10 09:31:36 is HWNAT functional? Mar 10 09:31:46 I've seen mixed reports Mar 10 09:32:39 rsalvaterra: I think russell-- is only talking about splitting off kmod-sched-teql, and not the larger refactoring I've wanted for a long time. That's doable, I just worry about having time. Do we know when 21.02 is expected, 'cause I'd like to get the latest iproute2 in there? Mar 10 09:33:25 mangix: Not yet, on 5.10. I think nbd is working on it. Mar 10 09:33:47 mangix: https://git.openwrt.org/?p=openwrt/staging/nbd.git;a=summary Mar 10 09:34:27 oh it's not working in general? what about 5.4? Mar 10 09:34:58 mangix: please try it with 5.10 (using the commit from my staging tree) and let me know if it works for you Mar 10 09:35:00 guidosarducci: I believe the timeframe is the usual one, "when it's done", but since it was branched already, it should be out in the next few months, I wager… :/ Mar 10 09:35:03 mangix: it works for me Mar 10 09:35:33 rsalvaterra: BCM6838 is documented here: http://www.datashed.science/misc/bcm/gpl/broadcom-sdk-416L05/shared/opensource/include/bcm963xx/6838_map_part.h Mar 10 09:35:51 mangix: I can't really give you a definitive answer, as I never used hardware NAT (conflicts with SQM). Mar 10 09:35:52 and it's supported on u-boot too: https://github.com/u-boot/u-boot/blob/master/arch/mips/dts/brcm%2Cbcm6838.dtsi Mar 10 09:36:09 russell--: I'll look into splitting off a kmod-sched-teql. Suggestions for the top-level teql pkg name? Mar 10 09:36:24 nbd_: great Mar 10 09:36:55 noltari: I see GPON definitions on that header…! :O Mar 10 09:39:24 Grommish: I see mixed results about the MDIO bus of realtek device. I may work on the MDIO bus of the internal switch. If that is that case that you have to say to the realtek driver which mdio bus to use. Mar 10 09:40:49 russell--: but how likely is that split teql package to be backported? There's a good case to be made for iproute2... Mar 10 09:41:08 noltari: Is there an easy way to find the CFE password? :) Mar 10 09:41:13 Other option is that it uses gpio-bit-banging to transfer mdio data to the realtek device. Then you have to specify the two gpio pins that are used. Mar 10 09:41:33 rsalvaterra: first time I know about such thing as a CFE password xD Mar 10 09:41:53 noltari: Oh, my… :P Mar 10 09:42:08 none of my devices have a CFE password Mar 10 09:42:12 guidosarducci: teql-script Mar 10 09:42:14 or something Mar 10 09:42:36 Yeah, when you try to break the boot process, CFE asks for a password. :) Mar 10 09:42:46 BCM6838 seems to be newer, maybe that's why it has a password... Mar 10 09:42:51 russell--: do we have "hotplug" packages? Mar 10 09:43:50 rsalvaterra: maybe you could find the password in the official firmware or GPL sources (if available) Mar 10 09:44:09 Grommish: An other thing is that currently MT7530 doesn't support fixed link on PORT 5 when used as switch port. It always expect an PHY on the other side. Mar 10 09:44:27 noltari: I sent an email to Altice asking for the sources. The answer was crickets, of course. :P Mar 10 09:44:44 WTH, visting family and seeing 10% packet loss and up to 600ms latencies. Congratulations on your Puma6 modem purchase! Mar 10 09:44:53 rsalvaterra: yeah I sent a request to ZyXEL a month ago and I got the same answer Mar 10 09:45:30 git grep PKG_NAME Mar 10 09:45:42 "This firmware was released more than 3 years ago, so we aren't forced to keep the GPL sources that long" Mar 10 09:45:43 WTF? Mar 10 09:46:38 3 years is a relevant duration if they provided a written offer in lieu of distributing source with the binary, iirc Mar 10 09:46:52 russell--: already did: we have things like "button-hotplug", so "teql-hotplug" sounds good. Mar 10 09:47:15 there is qos-scripts too Mar 10 09:47:50 russell--: true, though that class of packages has pretty broad functionality... Mar 10 09:48:06 and a bunch more in the packages feed Mar 10 09:48:34 no $foo-hotplug in packages feed Mar 10 09:50:36 russell--: I was thinking under package/kernel rather than package feed. Problem? Mar 10 09:51:23 no, just surveying existing PKG_NAMES Mar 10 09:51:53 who actually uses teql? Mar 10 09:52:23 noltari: If they actually delete the sources (I doubt it), it should be forbidden… :/ Mar 10 09:52:57 russell--: no one that I know, but someone will surely scream if it breaks... Mar 10 09:55:00 russell--: dwmw2 uses (or used) teql. He actually wrote that script. Mar 10 09:55:22 guidosarducci: can you look into https://github.com/openwrt/openwrt/pull/3004/commits/797b478f13a6abd129049d2e38362f2bb5272c40 ? Mar 10 09:55:51 russell--: https://dev.archive.openwrt.org/ticket/11192 Mar 10 09:56:34 Ok, for people who know device trees.. How do I enabled ethernet@1e100000 / mac@1? Mar 10 09:56:40 russell--: what's up? Mar 10 09:57:04 I know it has to have status = "okay" but I don't know how to call it by the & Mar 10 10:00:18 funny Mar 10 10:00:40 when going to the totolink page on aliexpress, all the links direct to the portuguese version of aliexpress Mar 10 10:03:07 mangix: I'd asked for more details on root cause: https://github.com/openwrt/openwrt/pull/3004#issuecomment-764197965. Can you post build/error logs? Mar 10 10:04:15 mangix: hadn't been able to reproduce and don't know the right place any fix would belong i.e. musl, kernel, etc. Mar 10 10:04:18 mangix: Brazilian, you mean. :P Mar 10 10:05:05 rsalvaterra: it's pt.aliexpress Mar 10 10:06:37 dwmw2: just curious about the teql use-case Mar 10 10:06:37 mangix: Yeah, but it's an automated translation… It sucks so bad I have the AliExpress app in English. Mar 10 10:07:37 rsalvaterra: ah ok Mar 10 10:08:07 mangix: Is the totolink the mt7621 device you ordered? Mar 10 10:08:22 yeah Mar 10 10:08:45 Which model is it? :) Mar 10 10:09:05 https://www.aliexpress.com/item/1005001942450125.html?spm=a2g0s.9042311.0.0.658c4c4dsFKrOu Mar 10 10:09:53 mangix: Expensive… I got this one: https://www.aliexpress.com/item/4001037617701.html Mar 10 10:10:14 pretty sure anything ax is expensive Mar 10 10:10:38 although the intel m.2 ax cards are selling for pretty cheap Mar 10 10:10:43 Oh. Missed the ax part, you're right. :P Mar 10 10:11:07 ax200 that is. probably because there's an ax210 Mar 10 10:11:12 jow: have you seen this? https://github.com/openwrt/packages/issues/14897 Mar 10 10:11:21 mangix: Stay away from Intel ax cards… especially for AP mode. Mar 10 10:12:36 i seem to remember ath10k gaining support for USB devices. wonder what happened to those Mar 10 10:14:44 rsalvaterra: they work fine in a laptop. except i dealt with a hillarious issue recently Mar 10 10:15:16 the wifi card stopped working Mar 10 10:15:30 and the laptop had no way of killing power to it Mar 10 10:16:36 i ended up disassembling it to disconnect the battery Mar 10 10:17:36 mangix: Happens to me when I disconnect my dad's laptop (iwlwifi AC 3165) from the AP. I have to reboot for it to come to life again. Mar 10 10:18:17 guidosarducci: https://gist.github.com/neheb/e432053147d4f2d23074d4292ad6a744 Mar 10 10:18:35 rsalvaterra: reboot and holding power button were not enough on this laptop Mar 10 10:20:11 mangix: To me it feels like Intel does the bare minimum to get their Wi-Fi devices working. I don't know why, they should take a long look at their graphics team, to learn how it's done. Mar 10 10:22:59 rsalvaterra: we talking linux or windows? Mar 10 10:23:18 rsalvaterra, well even graphics team takes time to fix bugs - like ~4 stable kernel releases to fix non usable graphics (5.7(broken)->5.10) Mar 10 10:24:06 mangix: I only tal Linux, sorry if it wasn't clear. :) Mar 10 10:24:09 *talk Mar 10 10:28:04 rsalvaterra: that card might be one with dead support unfortunately Mar 10 10:28:22 the 8265 is still well supported oddly enough Mar 10 10:28:36 https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/commit/?id=346057dbe7c7595748e61b8a5b962e5c7316924b Mar 10 10:29:18 mangix: Can't replace it. Lenovo signed BIOS, card whitelist. It's atrocious. Mar 10 10:29:46 mangix: FYI, your commit message said problem with "__maybe_inline" definition, about which I could find *nothing* when I last looked, but your logs instead show a problem with "__always_inline". Important distinction :) I'm more worried about the 2700 "invalid symbol" error messages in that log, which I've never seen. I gather the patch makes it compile finally, but does it actually work? Mar 10 10:30:02 arent those whitelists/lenovo bios entries hackable ? Mar 10 10:30:30 plntyk: Signed bios… change the signature and it won't flash. Mar 10 10:30:34 ah k Mar 10 10:30:38 guidosarducci: why wouldn't it? Mar 10 10:30:46 so there previous hacks where on unsigned hw Mar 10 10:30:52 bios Mar 10 10:31:01 rsalvaterra: ahem Mar 10 10:31:11 spi flash Mar 10 10:31:23 mangix: Don't tell me to desolder the flash chip, please. Mar 10 10:31:30 lol no Mar 10 10:31:38 use a SOIC8 clip Mar 10 10:32:09 mangix: Have you tried to disassemble a cheap laptop? It's not pretty. :P Mar 10 10:32:20 i have unfortunately Mar 10 10:33:04 I've read somewhere someone did it, and actually hacked the BIOS (some Portuguese guy, actually; even crazier than me). Mar 10 10:33:47 * russell-- recalls a linuxconf talk, i think it was about keyboards though Mar 10 10:34:18 anyway, these signed BIOSes are annoyingh Mar 10 10:35:27 russell--: The ticket made the use case fairly clear, didn't it? I had bonded ADSL lines, needed to use them both equally for uplink Mar 10 10:35:40 rsalvaterra: my interpretation of the commit i linked to is that those are the currently supported models. those and newer. Mar 10 10:36:25 dwmw2: yes, that is clear Mar 10 10:37:16 https://www.youtube.com/watch?v=Fzmm87oVQ6c Mar 10 10:37:19 mangix: I think I've seen firmware updates for the 3165 (it's the same firmware as the 7265D) last year, or the year before. Mar 10 10:37:36 "My personal fight against the modern laptop" Mar 10 10:38:12 -rw-r--r-- 1 root root 1036444 dez 6 19:56 iwlwifi-7265D-29.ucode Mar 10 10:38:26 Hmm… last December…? Mar 10 10:38:59 Yep, checks out: https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/log/iwlwifi-7265D-29.ucode Mar 10 10:39:48 (I manually manage my firmware, the size of the linux-firmware package is getting out of hand.) Mar 10 11:08:29 rsalvaterra: Are your mt7621 ports called lanX or ethX? Mar 10 11:11:05 Grommish: lanx and wan. Mar 10 11:11:56 rsalvaterra: See.. my image Tries to do that, but doesn't work unless I edit the /etc/config/network to read eth1-eth4 and eth0 for wan Mar 10 11:12:05 Hmmm Mar 10 11:12:56 only then will the interfaces come up or even show.. I wonder why Mar 10 11:15:57 mangix: Image built. I hope it works, fingers crossed… Mar 10 11:20:31 How is this even possible…?! Mar 10 11:20:58 mangix: [ 14.256639] mt7603e 0000:02:00.0: Invalid MAC address, using random address 96:ea:e3:4d:72:02 Mar 10 11:20:59 mangix: [ 16.665500] mt7615e 0000:01:00.0: Invalid MAC address, using random address 76:10:cd:73:4c:95 Mar 10 11:21:03 With musl 1.2.2. Mar 10 11:26:34 mangix: I am 100 % sure this is caused by the musl update. I reverted to the previous image and the problem is gone. Mar 10 11:26:53 Weird as hell…! Mar 10 11:29:28 Anyone have an idea why the ER-x clone I made wants to use ethX instead of lanX/wan? https://gist.github.com/Grommish/f34681b287f4a53a864cdbf2548e02d1 Mar 10 11:30:06 the labels are all lan1/lan2/lan3/lan4 etc in the dtsi Mar 10 11:30:56 lipnitsk: You have experience with mt7621 device trees, don't you? :) Mar 10 11:31:56 rsalvaterra: what kernel modules are you running? Mar 10 11:32:34 mangix: Modules? Only the compat ones. I do monolithic builds. Let me show you… Mar 10 11:33:07 mangix: https://paste.debian.net/1188669/ Mar 10 11:39:02 Ah yes. mt76 needs to be fixed :) Mar 10 11:40:03 Now he tells me… *facepalm* Mar 10 11:40:15 :P Mar 10 11:40:58 Do you know where I should look, in order to fix this, even if in a hackish way, in my tree? :) Mar 10 11:42:06 https://git.kernel.org/pub/scm/linux/kernel/git/backports/backports.git/commit/?id=473fbe303f36b84c6653dd84ee302d1b70078479 Mar 10 11:42:20 Some fix like that is needed Mar 10 11:43:27 Gah…! The stddef problem! Mar 10 11:43:41 Thanks for the pointer! :) Mar 10 11:43:46 * rsalvaterra starts digging… Mar 10 11:50:27 rsalvaterra: this was actually always a problem, but musl’s headers never caused any issues until 1.2 where __BIG and __LITTLE_ENDIAN are always defined by every header. Mar 10 11:50:35 mangix: https://git.openwrt.org/?p=openwrt/openwrt.git;a=blob;f=package/kernel/mt76/Makefile;h=810ee4c3d43314b8bcc3a4f9f4ce43ed28c08150;hb=HEAD#l238 Mar 10 11:50:50 Changing = to += should suffice…? Mar 10 11:51:34 No Mar 10 11:51:46 Grommish: target/linux/ramips/dts/mt7621_ubnt_edgerouter-x.dtsi Mar 10 11:52:14 mac80211 has some -noincludelibcheaders option Mar 10 11:52:40 that needs to be added to mt76 Mar 10 11:53:26 erm sorry, backporta Mar 10 11:53:35 *backports Mar 10 11:54:00 noltari: all my RPi0W on recent master seem to keep their LED on after boot, this wasn't the case before. any clue? Mar 10 11:58:06 hmmm and now I'm not even able to sysupgrade anymore Mar 10 11:58:14 Wed Mar 10 11:54:11 UTC 2021 upgrade: Image metadata not present Mar 10 11:58:15 Wed Mar 10 11:54:11 UTC 2021 upgrade: Use sysupgrade -F to override this check when downgrading or flashing to vendor firmware Mar 10 11:58:40 https://gist.github.com/stintel/d34065ce76aac6483cd50997aab7b810 Mar 10 11:59:14 so the image is signed somehow, breaking the metadata check ? Mar 10 11:59:38 stintel: there were some changes from the RPi Foundation on latest patches update which added support for LEDs on newer devices, but it could affect the old ones... Mar 10 12:00:14 noltari: the LED was already supported and disabled when boot finished (after wifi) Mar 10 12:01:02 stintel: yes I know, but as I already said "it could also affect the old models" Mar 10 12:02:22 oh Mar 10 12:02:31 done) Mar 10 12:02:32 status_led_on Mar 10 12:02:42 that's basically expected behaviour then Mar 10 12:02:47 and it was broken before Mar 10 12:03:29 and this is not configurable even Mar 10 12:03:39 that's slightly disappointing Mar 10 12:04:24 noltari: do you have any idea about sysupgrade ? Mar 10 12:05:09 stintel: after downloading latest sysupgrade image, it looks like I can't extract the squashfs one, but the ext4 image can be extracted... Mar 10 12:05:11 strange Mar 10 12:05:27 it's like the squashfs image is corrupted or something like that... Mar 10 12:06:27 >KGB-1< https://tests.reproducible-builds.org/openwrt/openwrt_ath79.html has been updated. (99.1% images and 98.3% packages reproducible in our current test framework.) Mar 10 12:09:47 noltari: I can flash the image I built just fine on a zero w running an older version of openwrt Mar 10 12:10:11 and all 4 images that are built gunzip fine also Mar 10 12:10:24 ok, I will try to look into it when I have time Mar 10 12:10:59 for the zero w in my bedroom I'll manually disable the LED until I can come up with something nice to make this configurable Mar 10 12:12:02 if I find more info about sysupgrade I'll let you know Mar 10 12:31:29 noltari: hmm, now cannot reproduce on other Zero W Mar 10 12:31:53 nor on the same. I guess something went wrong when uploading the image Mar 10 12:38:29 jow: (repost) do you remember why you added loopback to the interfaces in https://git.openwrt.org/c8a8f8fd ? Mar 10 12:39:33 stintel: I think I did it to make discovery tools enumerate the local instance as well Mar 10 13:06:32 mangix: I find it odd having this header problem with mt76 but not with ath10k-ct… Mar 10 14:59:14 jow: you mean discovery tools other than lldp itself ? Mar 10 15:50:49 rsalvaterra: not really a device tree expert. what's the question? Mar 10 15:55:05 lipnitsk: It's not me, Grommish is having some issues configuring the switch(es?) on the device tree of an ER10x. :) Mar 10 15:55:46 the rss counter bug is more or less a red herring btw Mar 10 15:55:51 lipnitsk: In other news, thanks to mangix, I learned mt76 is broken with musl 1.2.2. Mar 10 15:56:21 see https://lore.kernel.org/lkml/CALCv0x1NauG_13DmmzwYaRDaq3qjmvEdyi7=XzF04KR06Q=WHA@mail.gmail.com/ I dug into it Mar 10 15:56:44 lipnitsk: Ah, yes, I already read your email exchange with Eric W. Biederman. :) Mar 10 15:57:10 i think we need to investigate what's calling stuff so early but it's not a showstopper Mar 10 15:57:27 Oh, and Linus chimed in too. Mar 10 15:57:34 yeah... Mar 10 15:58:32 "MOST of the later MIPS architectures walked away from the pure virtual cache setups." Mar 10 15:58:46 Yeah, this is one of them. Caches are either PIPT or VIPT. Mar 10 15:58:56 No issue there. Mar 10 16:04:34 some hotplug call or something is just happening way too early, before MM is even fully initialized Mar 10 16:06:50 stintel: at that time I might have thought that lldpcli needs it Mar 10 16:07:26 stintel: or it might have been the case / I might have thought that lldpd needs at least one existing interface to start, hence loopback to give it something to work with Mar 10 16:08:03 stintel: in any case feel free to drop it if you cannot figure out a sane usecase for it to be there Mar 10 16:08:22 jow: thanks, I'll do some more experimenting, and if it doesn't crash without it and when "lan" doesn't exist, I can accept https://patchwork.ozlabs.org/project/openwrt/patch/20201211070511.3515348-1-john@phrozen.org/ Mar 10 16:08:39 could also have been that I simply added loopback and lan to demonstrate that multiple interfaces can be passed via uci list notation Mar 10 16:40:37 Grommish: you might have already seen this, but I downloaded ubnt GPL archive and it has some DTS stuff for er-10x Mar 10 16:40:56 https://github.com/lipnitsk/linux/commit/141209082c39970fca5b16996c95888c61a1b6d9#diff-27e0fa91df84b591e242d7a818d849148daf3fec5179e037cac29deda48b398f Mar 10 16:41:30 that's based on linux 4.14 so no dsa etc though. Mar 10 20:05:11 rsalvaterra: the issue is that both codebases match differently against big and little endian. Mar 10 20:07:20 mangix: I see… I mean, I understand the problem, but I don't grok the code enough to understand how to fix it. :) Mar 10 20:13:25 rsalvaterra: did you try the += trick? Maybe it actually works... Mar 10 20:13:56 mangix: I haven't. I trusted your judgement… :) Mar 10 20:15:39 And sure enough, ath10k-ct also does =, and works fine… Mar 10 21:03:47 Who is this KGB guy? Mar 10 21:04:23 I don't know, but seems to be enjoying himself. Mar 10 21:11:11 what's happening? Mar 10 21:19:15 mangix: I'm doing a build with the mt76 hack, let's see how it goes… Mar 10 21:21:53 oh, great -ChanServ- You are not authorized to (de)op rmilecki on #openwrt-devel Mar 10 21:22:39 rsalvaterra: any progress on the mt76 issue? Mar 10 21:40:43 Borromini: Which one? :) Mar 10 21:41:43 Borromini: Today I found out (the hard way) mt76 doesn't work correctly with musl 1.2.2. Mar 10 21:45:06 oh :( Mar 10 21:45:20 thought you were still chipping away at the mac problem Mar 10 21:46:35 Borromini: The MAC problem is the symptom. Mar 10 21:47:06 ok Mar 10 21:47:55 I'm trying now what I believe is a big hammer, but let's see how the build goes… Mar 10 21:51:32 :) Mar 10 22:20:32 … south. I'll have to TFTP my out of this one. Mar 10 22:20:57 Oh, well… Mar 10 22:23:17 jow: how does nlbwmon check for ipv6 connectivity? I have native ipv6 and yet nlbwmon on my router (with multiple openwrt devices downstream) shows my network as 0 ipv6 capable devices >_> Mar 10 22:23:45 it used to be better, with a few reboots of switches/APs downstream sometimes, but now it stubbornly keeps reporting zero Mar 10 22:24:33 this is on 21.02 but i suppose master codebase doesn't diverge much at this point Mar 10 22:33:55 jow: thanks for the feedback Mar 10 22:46:26 jow: could you take a look at my last fw3 updates for DSCP/MARK please? I've updated as discussed: https://github.com/guidosarducci/firewall3/commits/master-fix-dscp-mark. Thanks for the help! Mar 10 22:50:41 Borromini: really freenode is to blame here, if a client keeps being killed with excess flood they should just temp gline it Mar 10 23:11:49 rsalvaterra: I don't have any hardware yet so I can't really help Mar 10 23:15:34 mangix: No worries, there's no rush. I'm using the Omnia now. :) Mar 10 23:16:17 (And I have the Archer C6 with the exact same configuration, in case I brick the Omnia. :P) Mar 11 00:40:17 is it expected that tcpdump doesn't really work on dsa interfaces? Mar 11 00:40:34 I'm only seeing broadcast/multicast frames Mar 11 01:05:58 stintel: what device? Mar 11 01:06:34 Netgear GS108T v3 Mar 11 01:07:49 hmm no idea Mar 11 02:37:57 >KGB-0< https://tests.reproducible-builds.org/openwrt/openwrt_omap.html has been updated. (0% images and 96.1% packages reproducible in our current test framework.) **** ENDING LOGGING AT Thu Mar 11 02:59:57 2021