01:04 <owrt-images-builds> Build [#132](https://buildbot.staging.openwrt.org/images/#/builders/25/builds/132) of `master_gemini/generic` completed successfully.

03:02 <russell--> f00b4r0: i followed the u-boot environment changes suggested on the wiki page: https://openwrt.org/toh/watchguard/firebox-m300

03:02 <russell--> printenv seems to suggest they are set as suggested

03:04 <dwfreed> pastebin your printenv ?

03:04 <russell--> ext2load seems to work, but bootm is not happy

03:06 <russell--> dwfreed: http://sprunge.us/zZ6pVs

03:14 <russell--> MB-M300 => version

03:15 <russell--> U-Boot 2014.07 (Aug 12 2015 - 13:55:34)

03:15 <russell--> GNU ld (crosstool-NG 1.20.0) 2.22

03:15 <russell--> ppc-gcc (crosstool-NG 1.20.0) 4.8.3

03:16 <dwfreed> your bootcmd seems wrong

03:17 <dwfreed> I'd have to boot my m300 to be sure

03:30 <dwfreed> yeah, it is

03:30 <dwfreed> russell--: setenv bootcmd 'run wgBootSysA'

03:30 <dwfreed> saveenv and reset

03:35 <russell--> no difference

03:36 <dwfreed> what's the new printenv?

03:37 <russell--> no matter what bootcmd is set to, i get the menu, SysA is selected, and i get the bad crc

03:37 <russell--> Verifying Hash Integrity ... crc32 error!

03:37 <russell--> Bad hash value for 'hash@1' hash node in 'kernel-1' image node

03:37 <russell--> Bad Data Hash

03:37 <dwfreed> sure your sd card isn't bad?

03:38 <dwfreed> try validating that

03:39 <russell--> i'm using the same one that came in it, read it with dd got save a backup, wrote zeros at it and then the sdcard image, no hint of a problem with the media

03:39 <dwfreed> that output suggests otherwise :P

03:39 <russell--> to* save a backp

03:40 <russell--> i can read the image with ext2load, no errors

03:40 <dwfreed> just because it loads doesn't mean it isn't still corrupted

03:41 <dwfreed> humor me and take an md5sum of the image file and the same size at the start of the sd card

03:41 <russell--> MB-M300 => ext2load mmc 0:1 $loadaddr $OpenWrt_kernel

03:41 <russell--> 6040740 bytes read in 322 ms (17.9 MiB/s)

03:41 <russell--> can't hurt

04:08 <KGB-0> https://tests.reproducible-builds.org/openwrt/openwrt_kirkwood.html has been updated. (100.0% images and 100.0% packages reproducible in our current test framework.)

04:13 <russell--> dwfreed: bingo. i think what might have happened is my sdcard got remounted and was mounted as a file system when i dd'd to it and it got scribbled on. reflashing the sdcard, and it boots.

04:14 <dwfreed> yeah, you probably had an automount kick in mid-flash or something

04:14 <dwfreed> automounters suck

04:15 <russell--> on the bright side, the sdcard itself is fine

04:15 <russell--> and even brighter side, box works

04:15 <dwfreed> woo

04:16 <dwfreed> one of these days I'll get mine up and running

04:18 <owrt-images-builds> Build [#134](https://buildbot.staging.openwrt.org/images/#/builders/76/builds/134) of `master_ipq40xx/generic` completed successfully.

05:05 <owrt-images-builds> Build [#134](https://buildbot.staging.openwrt.org/images/#/builders/59/builds/134) of `master_apm821xx/sata` completed successfully.

05:37 <owrt-images-builds> Build [#134](https://buildbot.staging.openwrt.org/images/#/builders/65/builds/134) of `master_at91/sam9x` completed successfully.

06:14 <dwfreed> I believe it is mimicing the default oem config

06:26 <owrt-images-builds> Build [#136](https://buildbot.staging.openwrt.org/images/#/builders/61/builds/136) of `master_bcm47xx/legacy` completed successfully.

07:06 <owrt-images-builds> Build [#136](https://buildbot.staging.openwrt.org/images/#/builders/51/builds/136) of `master_ath79/tiny` completed successfully.

07:30 <russell--> uci-defaults scripts run too early to nuke the network interface stanzas

07:36 <russell--> i guess i can just override /etc/board.d/02_network

07:44 <dwfreed> they also only run once

07:44 <dwfreed> so you can just override the config after

07:53 <russell--> it isn't unprecedented, but imho, it is weird: git grep ucidef_set_interface | grep ipaddr

07:54 <russell--> the board.d override works, but it is a bit fragile

10:04 <torv> Mangix: You probably accidentally opened a "container tab" in Firefox. (Re: "[...] Firefox thinks OpenWrt forums are Banking")

10:35 <hitech95> Any idea on why I have SCLK and CS in a spi bus but no MISO and MOSI? I'm using spidev_test utility and the pinmuxing seems correct.

14:52 <johnf> what causes devices to appear in the table of hardware? I see the GL-MT3000 is supported, but it's not in the ToH

14:52 <johnf> https://firmware-selector.openwrt.org/?version=23.05.0-rc4&target=mediatek%2Ffilogic&id=glinet_gl-mt3000

14:52 <johnf> it is in the firmware selector though

14:53 <johnf> I thought it was automated to some degree

16:32 <robimarko> Its a manual process, somebody has to add it

17:33 <Mangix> torv: makes sense

17:38 <hauke> My IPv6 connection does not work after my ISP disconnected me on 23.05.0-rc3. Every 24h my ISP disconnects the PPPOE connection on top of my fiber line. IPv4 is configured correctly, but IPv6 is not configfured correctly, I am only geting a link local IPv6 address. it looks like this: "local LL address fe80::ada4:9af5:db1c:6552"

17:39 <hauke> I assume that the ISP expects that I do a neighbor discovery to find the router

17:39 <hauke> can I trigegr this manually for experiments?

17:39 <hauke> when I first connect to the line IPv6 is working fine, it only breaks after time

17:48 <lemmi> did you try dhcpv6-pd? usually that's what's being used.

17:59 <lemmi> ra with ppp is rather useless anyway, since it can't be easily bridged or forwared for the rest of your network

18:00 <hauke> lemmi: I configured it like this: https://paste.debian.net/1294347/

18:01 <hauke> don't you use RA for the interface itself and then PD to get a prefix for the network?

18:03 <lemmi> you can do that, but that's not really necessary

18:04 <lemmi> my default route is via a link local address, so the ras are completely useless here, which is why i disabled them

18:05 <hauke> ok

18:08 <lemmi> so as long PD works, i wouldn't worry

18:11 <lemmi> hauke: oh.. and and i'm not sure `pppoe-wan` is the correct device. it might need an @ in front, or be `@wan`, or if in doubt use `wan.7` as well

18:12 <lemmi> but i always get that wrong and can't remember what the correct thing is... so i tend to play around until it works.

18:13 <hauke> the strange thing is that it works with the first connection and only fails after a disconenct

18:16 <lemmi> maybe the dhcp client needs a push somehow

18:20 <hauke> now I reloaded the network configuration and ppp was restarted, then IPv6 started working fine

18:20 <hauke> I will monitor this

18:25 <lemmi> i'd try and use `@wan` as the device. it might be that otherwise the interface block isn't properly triggered on a reconnect. (thinking about it, `wan.7` can't work btw, since dhcp needs to happen over pppoe)

18:29 <iocampomx> Hi, I'm trying to define the mac address in 02_network, but somehow isn't working. I first checked that the right board_name in /tmp/sysinfo/board_name matches the case within the ramips_setup_macs() function. I've defined label_mac, lan_mac and wan_mac. label_mac is extracted from factory. lan_mac = label_mac, wan_mac + lan_mac + 1. I tested those three lines (loading mac, adding 1 number to mac) within OpenWrt and

18:29 <iocampomx> works, what I haven't tested (and I don't know if those works) are the functions: ucidef_set_interface_macaddr for "lan" & "wan".

18:31 <iocampomx> I would like to understand if there is anything special in /etc/config/network I should do to match those "lan" & "wan" labels. And/or, if there is any other file I need to update to make it work.

18:31 <iocampomx> For instance, I see there is a function ramips_setup_interfaces() within 02_network file, but I haven't touched it. Should I?

18:46 <iocampomx> A follow-up question, Is 02_network only once in the first boot after performing an upgrade?

19:36 <hauke> lemmi: I removed the ipv6 option from the pppoe interface, I seletced ipv6 = 1, then I have to manually configure it, it looks like this type does not notice when the link broke down. I selected auto now and it works so far, see https://openwrt.org/docs/guide-user/network/ipv6/configuration#ppp-based_protocols_and_option_ipv6

21:40 <iocampomx> So, which is the best way to define the mac address of the LAN & WAN address? (I already know how to get and calculate it). I'm trying to understand if I should use 02_network, or init.d scripts, or rc?

21:42 <stintel> the "best" way is to do it in the DTS, reading it from wherever mfg stored it in flash

21:44 <iocampomx> So, it's a mediatek SoC, I'm reading it already in DTS using $wmac... It works, but that is only the mac address for thw WLAN (WiFi).

21:44 <iocampomx> Now, I'm trying to define the mac address for WAN & LAN, which will be WLAN + 1, WLAN + 2

21:44 <iocampomx> Is that something that I can also define in DTS?

21:44 <stintel> I believe so

21:46 <stintel> target/linux/generic/pending-6.1/804-nvmem-core-support-mac-base-fixed-layout-cells.patch

21:46 <linusw> iocampomx: the DTS just use compatible = "nvmem-cells" to create an arbitrary readout at some offset from a flash partition if it is stored somewhere fixed in flash.

21:47 <iocampomx> So, asumming that only the WLAN mac addre is stored on the flash partition, and the LAN & WAN mac address should derive from there (adding 1 and 2 hex values), what I can do?

21:48 <iocampomx> (To be more clear, besides the WLAN mac address, I don't see the mac address for LAN & WAN. I also don't want to modify the Factory partition)

21:49 <stintel> yeah you should never modify mac addresses in flash

21:49 <stintel> git grep mac-base

21:49 <stintel> plenty of examples

21:50 <karlp> (assuming they were actualyl uniquely programmed at a competent factory)

21:50 <stintel> you should look at what stock firmware does probably

21:50 <iocampomx> Yes, it's a unique mac already flashed there, but just for the WLAN, I know how to get it and calculate the rest.

21:50 <iocampomx> For instance. wlan_mac=$(hexdump -s 4 -n 6 -e '6/1 "%02x:"' /dev/$(cat /proc/mtd|grep -i "Factory"|cut -d':' -f1)|cut -c-17)

21:50 <iocampomx> From there I can do: macaddr_add "$wlan_mac" 1, macaddr_add "$wlan_mac" 2, etc.

21:51 <stintel> it happens that there is 1 unique mac in the flash for e.g wlan, then wan is + 1 lan + 2

21:51 <stintel> you can do all that with mac-base in DTS

21:51 <stintel> https://www.kernel.org/doc/Documentation/devicetree/bindings/nvmem/layouts/fixed-cell.yaml

21:53 <iocampomx> Like in this example: https://github.com/openwrt/openwrt/blob/9536446965e853231e34c4e5dc4cf13f838b9e90/target/linux/ath79/dts/ar7161_dlink_dir-825-b1.dts#L207-L217 ?

21:55 <stintel> iiuc that reads 2 mac addresses from flash, in ascii format (length 17 as documented in fixed-cell.yaml)

21:56 <stintel> then nvmem-cells = <&macaddr_wan 0> uses exactly that mac, and nvmem-cells = <&macaddr_wan 1> would be offset 1 (so it's mac from flash +1)

21:56 <stintel> (iiuc)

21:57 <stintel> I haven't used mac-base myself as it's fairly new, but it's what's accepted upstream and everything new should use that

21:58 <iocampomx> Ok! I will play with it.

23:37 <KGB-2> https://tests.reproducible-builds.org/openwrt/openwrt_lantiq.html has been updated. (95.9% images and 100.0% packages reproducible in our current test framework.)