**** BEGIN LOGGING AT Sun Apr 19 02:59:57 2020 Apr 19 03:05:32 maybe this should be backported to 4.14: https://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs-stable.git/commit/?h=linux-4.14.y&id=3cc9792fa1ea67bbfdaee28737e37f9045976b76 Apr 19 03:05:54 trying to rm a /dev/null device on f2fs gives this stack trace: https://paste.ee/p/VQziQ#9P0HO6XhwQTWxLk1r5yfUyQtaO0qW54U Apr 19 03:06:14 it's the fix per https://bugzilla.kernel.org/show_bug.cgi?id=202495 . i'd need to rebuild to confirm though. Apr 19 03:09:11 Hey all :) Anyone want to help show me how I'm missing something obvious on a kernel compile? Apr 19 03:50:30 okay, so that patch does fix the WARN_ON() when trying to e.g. 'rm /some/f2fs/filesystem/dev/null'. should i open a pr to backports-4.14? Apr 19 04:21:29 build #145 of layerscape/armv8_64b is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/openwrt-19.07/images/builders/layerscape%2Farmv8_64b/builds/145 Apr 19 04:21:30 m4t: that is probably the best approach (respectively mailing it in) Apr 19 04:22:26 i imagine nobody has hit it because all devs are normally on a tmpfs Apr 19 04:22:35 i needed a chroot for something though Apr 19 07:25:05 An very good thumbs up for whoever made the uhttpd openssl ssl defaults, checker looks awesome :) Apr 19 08:02:45 Does anyone run the nginx variant? Last time I tried it (I think almost 2 years now) it was kinda unstable, it kept being up ~day and then "not there", plenty capable x86 host so no restrictions there :) Apr 19 08:03:27 I ask because I don't have any good testing setup for that available and wouldn't like to do another live-test =D Apr 19 08:40:04 Also, more technical wireless query: When I setup multiple SSID'ds or "networks", some with sole SSID, others sharing SSID (with freq per device and other devices), do I care to set diffirent BSSID in single device between the sole SSID and then SSID sharing name? Apr 19 08:40:13 11r also used Apr 19 08:40:43 or rephrasing: would I benefit or lose something if doing such thing? Apr 19 08:42:58 jow: is the current buildbot using sqlite or mysql/postgres? Apr 19 09:38:11 olmari: still unstable from what i know Apr 19 09:46:15 mangix_: Than kyou for the input :) Apr 19 09:57:26 can someone enlighten me on the role of ucidef_set_label_macaddr()? I'm not sure what it's used for? Apr 19 09:58:57 jow: using PostgreSQL seem to speed up the builds by 10-20 minutes... Apr 19 10:36:45 Throughout anything I've done database related, I've accustomed to see sqlite never performing, and even more importantly, it fails in so many odd ways where "proper" databases works =) Apr 19 10:37:49 I once thought sqlite is okay fir fast testng and alike, but today I'm not even sure about that (: Apr 19 10:56:50 f00b4r0: There's usually one mac address printed on the router label and this function is for that. Apr 19 10:58:45 f00b4r0: It isn't used in openwrt now but the info it stored can be used if a userspace program wants an unique device identifier. Apr 19 10:59:40 gch981213: ah ok thanks. So not used in owrt. That's the bit I needed Apr 19 11:00:29 I'm finishing the plumbing for supporting mikrotik devices "properly" in a DTS world, and we can't use mtd-mac-address, so I'm looking at the next best thing Apr 19 11:04:16 what is the best resource or guide for writing a modern luci customization? Apr 19 14:15:25 only the recent code commits. Apr 19 14:15:48 there was hardly ever a "guide" to writing luci, that hasnt' really changed now that it's js instead of lua :) Apr 19 15:06:54 is my understanding that /sbin/hotplug-call is called upon kernel uevents such as firmware requests correct? Apr 19 15:49:25 Hi all, I'm thinking about having hotplug call go over ubus, but not sure how to go about it, either have a method per type (hotplug.ntp), or extending ubus acl to be able to validate arguments (service.event type: hotplug.ntp) Apr 19 15:51:41 the goal would be to finally sandbox all the daemons running as root like ntpd / udhcpc / odhcp6c that need very few privileges execpt to call hotplug Apr 19 16:09:07 adrianschmutzler: there is a target-wide fix fix commit in my staging tree for as long as the repeater ones Apr 19 16:09:57 afair the ramips situation is a bit different, as they are using OF in early bootstage to determine machine capabilities. Apr 19 16:12:16 Upstream documentation also seems outdated, so give it some time Apr 19 16:23:13 blocktrron: thanks, will have a look. no hurry, I also don't think that's terribly important, just wanted to note it. Apr 19 17:29:01 aparcar[m]: sqlite Apr 19 18:16:41 champtar: you're talking about the services that want to send a hotplug event right? not the consumers? Apr 19 20:06:04 karlp: yes, ntpd for example Apr 19 20:06:55 karlp: yes the "producers", ntpd for example Apr 19 20:09:44 which numeric prefix should i give to an f2fs patch intended for backports-4.14? Apr 19 20:20:31 going for: generic-backport/501-v4.19-f2fs-skip-verifying-block-address-non-regular-inode.patch Apr 19 20:25:53 m4t: "0xx - upstream backports", https://openwrt.org/docs/guide-developer/build-system/use-patches-with-buildsystem#naming_patches Apr 19 20:26:14 thanks Apr 19 20:26:47 funny i was already at https://openwrt.org/docs/guide-developer/build-system/use-patches-with-buildsystem#adding_or_editing_kernel_patches, i guess i should've kept scrolling :P Apr 19 21:10:20 updated openwrt/upstream, https://sdwalker.github.io/uscan/index.html Apr 19 22:32:24 can anyone else confirm what appears to be a race condition between procd/ubus/logd on x86_64? it seems to happen consistently for me when booting under qemu https://imgur.com/a/JDWEYYP Apr 19 22:32:38 that's 19.07.2 Apr 19 22:32:47 (i saw this on real hardware, too) Apr 19 22:36:41 i was just hitting in an initramfs on real hardware though iirc. Apr 19 22:58:28 seeing it at least as far back as 18.06.0 for the squashfs-combined images. not for the ext4 combined, though. another side effect is that the overlay doesn't seem to get created/mounted: https://imgur.com/a/2kezkK2 Apr 20 01:10:05 Hauke: can you backport https://github.com/openwrt/openwrt/commit/ce1798e915181e6c1f3ba735b254b37b84261303 to 19.07. Fixes compilation with uClibc-ng. **** ENDING LOGGING AT Mon Apr 20 02:59:57 2020