01:29 <schmars[m]> nbd: one my affected mt7915 devices recovered somewhat successfully for the first time today. will post info tomorrow

01:29 <schmars[m]> took a couple of hours though

01:33 <damo22> why does the buildbot do something different than local build?

01:33 <damo22> surely that is the cause of broken code

01:34 <damo22> because devs only test local build and code gets merged, then the buildbot builds a broken board port

01:35 <damo22> eg if the kernel gets bloated past a threshold where the decompression fails

02:02 <KGB-0> https://tests.reproducible-builds.org/openwrt/openwrt_sunxi.html has been updated. (0% images and 100.0% packages reproducible in our current test framework.)

02:08 <dwfreed> damo22: there is CONFIG_BUILDBOT to replicate what buildbot does

02:08 <dwfreed> that's all buildbot is doing anyway

04:13 <Tusker> heya guys, I was wondering if there was any example of dual boot with ubi? I currently have label="ubi", and a kernel partition for the primary partition, but if I write into the backup partition, and make uboot boot the new offset for the backup kernel, how do I redirect to the other rootfs ?

04:19 <slh> linksys wrt*, ea8300

04:34 <damo22> dwfreed: i guess my question is, why is that not set by default in local builds?

04:39 <Tusker> slh: OK, thanks, but the challenge is that the deco x20 has the kernel sitting inside a jffs2 partition, and the rootfs I have done for openwrt is on an ubifs sitting after the jffs2 partition. THe stock firmware doesn't use ubifs, so their uboot doesn't switch the rootfs for the ubifs part

04:41 <Tusker> there is a tp_boot_idx that is set to 0 or 1, which will redirect to the different jffs2 partitions, but I can't see how I can use that uboot variable to control what ubifs to mount with the kernel. The label="ubi" allows it to automount fine, but if I remove that label="ubi" the kernel won't know how to mount it

04:44 <slh> some targets use a patch to implement append-rootblock=

04:45 <slh> e.g. target/linux/ipq806x/files-6.6/arch/arm/boot/dts/qcom/qcom-ipq8064-eax500.dtsi

04:45 <Tusker> ah, interesting... any example ? or should I just grep ?

04:45 <slh> target/linux/ipq806x/patches-6.6/902-ARM-decompressor-support-for-ATAGs-rootblock-parsing.patch

04:46 <slh> grep is suggested, as it's used in a variety of cases

04:56 <Tusker> hmm... doesn't seem like it will work with ramips

05:09 <dwfreed> damo22: because there are aspects of buildbot that are harmful to local builds

08:48 <owrt-images-builds> Build [#259](https://buildbot.openwrt.org/images/#/builders/239/builds/259) of `master_starfive/generic` completed successfully.

09:02 <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.)

09:17 <owrt-images-builds> Build [#362](https://buildbot.openwrt.org/images/#/builders/124/builds/362) of `master_bcm27xx/bcm2710` completed successfully.

10:12 <owrt-images-builds> Build [#367](https://buildbot.openwrt.org/images/#/builders/187/builds/367) of `master_lantiq/xway_legacy` failed.

12:01 <owrt-images-builds> Build [#367](https://buildbot.openwrt.org/images/#/builders/205/builds/367) of `master_mpc85xx/p1020` failed.

14:19 <mrkiko> Ansuel: robimarko: did you in the end get rid of ZLOADADDR or replaced the whole u-boot in Zyxel nbg7816 ??

14:22 <mrkiko> robimarko: Ansuel: sorry, meant ZLOADER or how that thing is called :)

14:22 <mrkiko> robimarko: Ansuel: the thing that annoyes the user with the password :)

14:39 <fixitman> Hello

14:40 <fixitman> I have a GL.iNet E-750 v2. It uses atheros chipset. I also own a v1 of this device. On v1, I can use either ath10k or ath10k-ct to utilize the 5GHz radio. On v2, the ath10k-ct driver does not work -- no interface is recognized. However, the old ath10k driver does work.

14:41 <fixitman> I have some development experience and have compiled my own build of 23.05.4 to test on this device... But I am not sure where to begin with trying to debug a kernel module like this

14:41 <fixitman> If anyone has some experience with the atheros drivers and could give me some tips, I'd appreciate the help

14:42 <fixitman> My goal is to get the ath10k-ct driver working on this device

14:51 <fixitman> The manufacturer's "fix" for ath10k-ct not working in vanilla OpenWrt is to revert to the old ath10k driver... https://github.com/openwrt/openwrt/pull/15544/files

15:29 <owrt-images-builds> Build [#368](https://buildbot.openwrt.org/images/#/builders/231/builds/368) of `master_mpc85xx/p1010` failed.

16:22 <robimarko> mrkiko: The ipq806x nbg?

19:16 <f00b4r0> schmars[m]: you seem to have a more reliable reproducer than I. Mine is still alive and kicking after 10 days

19:17 <schmars[m]> ~once per day, but different client device every time

19:17 <schmars[m]> nice

19:17 <f00b4r0> fsvo "nice" :)

19:18 <schmars[m]> i have one which is hasn't crashed in a few weeks now :) but two at the busy location crash daily

19:18 <f00b4r0> *sigh*

19:18 <schmars[m]> nice that yours is alive is what i meant :)

19:18 <f00b4r0> ah lol :)

19:19 <f00b4r0> mine is fairly random. I suspect it only takes that one deadly client to kill it, and said client happens randomly

19:19 <schmars[m]> yep, think so too. busy location = higher likelyhood

19:21 <f00b4r0> yep

20:21 <owrt-images-builds> Build [#119](https://buildbot.openwrt.org/images/#/builders/10/builds/119) of `openwrt-23.05_bcm63xx/smp` failed.

22:15 <schmars[m]> nbd: nevermind what i said yesterday, it didn't recover, the people there just rebooted it themselves :-) i collected more dumps from the past few days: https://gist.github.com/pktpls/bc767a300460442c4faff14a2db53986 - the last two are interesting, they show actual errors i think. the other dumps are "as always", some device deauthenticates which has been switching between 5ghz and 2ghz before, and then the message timeouts start

23:27 <KGB-0> https://tests.reproducible-builds.org/openwrt/openwrt_omap.html has been updated. (11.1% images and 100.0% packages reproducible in our current test framework.)