**** BEGIN LOGGING AT Mon Apr 27 02:59:57 2020 Apr 27 03:06:45 build #295 of octeon/generic is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/octeon%2Fgeneric/builds/295 Apr 27 03:13:47 build #290 of malta/be is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/malta%2Fbe/builds/290 Apr 27 03:13:56 swalker: not sure I mentioned this. perl-inline is watching the wrong URL Apr 27 03:17:19 love upstream changes Apr 27 03:17:24 updated openwrt/upstream, https://sdwalker.github.io/uscan/index.html Apr 27 03:33:43 swalker: flashrom also has some...weird version handling Apr 27 03:37:07 build #288 of ath79/generic is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/ath79%2Fgeneric/builds/288 Apr 27 03:59:33 build #285 of samsung/s5pv210 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/samsung%2Fs5pv210/builds/285 Apr 27 04:36:33 build #280 of ramips/mt7620 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/ramips%2Fmt7620/builds/280 Apr 27 04:38:08 build #290 of sunxi/cortexa7 is complete: Failure [failed defconfig dltar] Build details are at http://buildbot.openwrt.org/master/images/builders/sunxi%2Fcortexa7/builds/290 blamelist: Lucian Cristian , Xiaobo Tian , Li Zhang , Petr ?tetiar , David Bauer @david-bauer.net>, Luiz Angelo Daros de Luca , Linus L?ssing , Adrian Schmutzler , Rosen Penev , Kevin Darbyshire-Bryant , Matt Merhar , DENG Qingfang , Sungbo Eo , Apr 27 04:38:08 Yangbo Lu , weidong jia Apr 27 04:49:48 build #274 of lantiq/falcon is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/lantiq%2Ffalcon/builds/274 Apr 27 05:29:46 build #290 of rb532/generic is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/rb532%2Fgeneric/builds/290 Apr 27 05:43:40 build #305 of ramips/rt305x is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/ramips%2Frt305x/builds/305 Apr 27 05:48:10 build #306 of mpc85xx/p1020 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/mpc85xx%2Fp1020/builds/306 Apr 27 06:47:02 build #301 of ramips/mt76x8 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/ramips%2Fmt76x8/builds/301 Apr 27 07:10:07 build #299 of lantiq/xway is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/lantiq%2Fxway/builds/299 Apr 27 07:12:44 russell--: https://bugs.openwrt.org/index.php?do=details&task_id=3032 Apr 27 07:24:38 ynezz: reverting to 0.177 seems to fix it Apr 27 07:27:13 build #298 of armvirt/64 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/armvirt%2F64/builds/298 Apr 27 07:34:32 russell--: I'm afraid, that reporting upstream issue in OpenWrt is not going to have desired effect (fix) :) Apr 27 07:39:56 it's okay, i have a local fork that works now. Apr 27 07:46:05 build #296 of zynq/generic is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/zynq%2Fgeneric/builds/296 Apr 27 07:50:57 philipp64: FYI, connection timeout from my MTA trying to reach yours (after reply to m-l message) Apr 27 07:53:22 build #296 of mxs/generic is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/mxs%2Fgeneric/builds/296 Apr 27 08:14:07 build #280 of layerscape/armv7 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/layerscape%2Farmv7/builds/280 Apr 27 08:14:09 blocktrron: it looks like phylink and sfp are built-in for every target Apr 27 08:41:12 Hi. https://bugs.openwrt.org/index.php?do=details&task_id=3045 is annoying for me. Other than a bisect, any idea of how to debug it? Apr 27 08:42:03 RedDwarf: does not sound like the kind of bug that will ever gain any traction Apr 27 08:42:58 so bisecting seems to be like the best next step Apr 27 08:43:49 also figure out whether it happens with vanilla Linux too (does it work as expected on a laptop?) Apr 27 08:44:13 I don't currently have a serial cable connected to the device (and the COVID lockdown...). How safe in general is to flash random git commits (i.e. will I keep having SSH access?) Apr 27 08:44:30 yes, on my laptop it works fine Apr 27 08:45:27 depends on the device and its recovery options Apr 27 08:45:40 if the bootloader can do tftp or has some kind of emergency flash then it should be pretty safe Apr 27 08:45:51 ssh access should always be there as long as the firmware boots Apr 27 09:02:55 jow: "DATA CHANNEL IMPENDING FAILURE DATA ERROR RATE TOO HIGH" <- slashdirt 02 root disk (raid1) ;P Apr 27 09:05:55 f00b4r0: hm, feeld free to shut down that builder Apr 27 09:06:29 I'll let it run as long as it can. Although right now the VM network doesn't seem to want to come back :P Apr 27 09:07:09 I need to find a suitable replacement for the disk; but I don't know when I'll be able to physically replace. Apr 27 09:23:21 morning. Hopefully I'm less stressed/grumpy today Apr 27 09:25:44 get off my lawn! Apr 27 09:25:50 :P Apr 27 09:26:08 lol Apr 27 09:29:10 Well I've fixed my rrdtool always displaying 'e' notation labels in logarithmic mode, then I came up with a better implementation...that was driving me to and the definition of insanity yesterday. I fixed an indentation snafu and probably a bug in rrdtools' libpng Apr 27 09:31:40 The overnight builds on umdns & relayd on snapshot appear to have worked, so next jobs are to bump those in 19.07, push the rrdtool improv/fixes and do a bit more finessing of sqm-qdisc-graphs. Apr 27 09:32:35 and I've finessed my https-dns-proxy PR Apr 27 09:34:04 ldir: isn't it enough to override the number format to get rid of the exponential notation? Apr 27 09:35:40 The number format appears to apply to the labels, not the y-axis labels... but I could have been immensely stupid Apr 27 09:36:25 dengqf6: the symbols are not present in the generic target configuration Apr 27 09:40:24 also bear in mind I'm the only idiot using the logarithmic y scaling option. Apr 27 09:40:26 ldir: ah yes, that might be (the y-axis labels) Apr 27 09:43:14 ldir: seems at least rrdtool 1.4.9 is needed to override the y axis lables Apr 27 09:43:43 maybe we should simply bumpt to a newer one (will also look better), but last time I checked it was extremely huge compared to 1.0.x Apr 27 09:44:02 it is. Apr 27 09:44:11 lols @ "extremely huge" Apr 27 09:44:14 I tried it a few months ago Apr 27 09:44:34 depends on stuff like libcairo and libfreetype Apr 27 09:44:40 jow: was 02 updated recently? Network doesn't seem to start at all Apr 27 09:44:43 there was something removed in current versions too iirc Apr 27 09:44:44 and likely half of freedesktop behind that Apr 27 09:44:47 no matter how many times I reboot it Apr 27 09:44:54 f00b4r0: not that I am aware of Apr 27 09:44:57 I think I tried some intermediate version. Apr 27 09:45:09 f00b4r0: you could simply nuke the vm and we'll reprovision it Apr 27 09:45:23 i'd rather not. thats work for me ;) Apr 27 09:45:32 not having any kind of access doesn't help Apr 27 09:45:41 the VM is alive. It's just not connected Apr 27 09:46:08 lemme see if I can somehow give you remote access to virsh console Apr 27 09:46:10 maybe try to access it by its ipv6 ll in case dhclient did weird things again? Apr 27 09:46:11 no serial? Apr 27 09:47:25 Namidairo: more like no account no password Apr 27 09:47:27 :P Apr 27 11:08:14 jow: you've got an rrdtool1 pull request ;-) Apr 27 14:11:40 philipp64: pong Apr 27 14:29:35 a no-feeds build just gave me these warnings: http://sprunge.us/vyTh9c Apr 27 14:33:01 (on a fresh clone) Apr 27 14:41:39 Hauke: do you happen to know if the mkimage bump was tested on macos or am I about to find out? :-) Apr 27 14:45:00 kaboooom! Apr 27 16:16:50 I’m trying to test x86_64 builds on KVM, ran into the following… I’ve provisioned an e1000 interface on the VM (the underlying hardware is a Xeon D with i210 and i350 interfaces, which use the igb driver). however, the VM is detecting an i350… but the driver isn’t binding to it, even though it’s loaded. how to troubleshoot this? Apr 27 16:46:19 philipp64: do you checked the pci ids? Apr 27 16:47:09 00:03.0 0200: 8086:1520 (rev 01) Apr 27 16:52:15 philipp64: choosed/installed proper drivers in openwrt? Apr 27 16:58:59 philipp64: do you have kmod-igbvf installed? Apr 27 16:59:44 no… had configured the interface as an “e1000” so figured that driver would bind to it. Apr 27 16:59:44 I used this: https://duckduckgo.com/html?q=%228086+1520%22+site%3Acateee.net Apr 27 17:02:33 okay, I’ll try it. I thought the “e1000” configured in the VM is what the interface would be exposed as, but it’s showing the raw PCI-IDs of the underlying device. Apr 27 17:29:43 Hi I am getting this in my logs Apr 27 17:29:45 Mon Apr 27 18:18:28 2020 daemon.warn dnsmasq[9992]: Maximum number of concurrent DNS queries reached (max: 200) Apr 27 17:29:48 PaulFertser: okay, kmod-igbvf did it… Apr 27 17:29:59 Then all internet stops. Apr 27 17:46:39 philipp64: cateee.net rules :) Apr 27 17:53:43 ldir: I build tested the mkimage bump on freebsd and added some fixes Apr 27 17:53:59 ldir: could you try it on MacOS, I hope it is similar Apr 27 17:54:24 someone's testing for FreeBSD? very cool! Apr 27 17:55:09 wanted to create a builder on FreeBSD but didn't expect that this would work easily Apr 27 17:55:11 decke[m]: only if mkimage is building becasue I do not have MacOS Apr 27 17:55:48 don't have a MacOS either but could help with FreeBSD Apr 27 18:07:03 Hauke: https://paste.ubuntu.com/p/WJW7Dm95d5/ which is all a bit dd Apr 27 18:07:07 odd Apr 27 18:10:41 ldir: what does "uname -m" return? Apr 27 18:11:02 x86_64 Apr 27 18:12:21 ldir: could you try to remove the "undefine MK_ARCH" in line 37? Apr 27 18:14:20 seems to solve it. Apr 27 18:16:53 ldir: I think it is fine it remove it Apr 27 18:17:02 does macOS use a special make version? Apr 27 18:17:38 GNU Make 3.81 This program built for i386-apple-darwin11.3.0 Apr 27 18:18:04 But I guess Apple could have done anything to it :-) Apr 27 18:19:37 ldir: undefine was added in make 3.82 Apr 27 18:20:03 ldir: https://github.com/ESMCI/cime/issues/1033 Apr 27 18:20:08 ha! ahhh, that'll be it then :-) Apr 27 18:20:49 make 3.82 is almost 10 years old: https://lists.gnu.org/archive/html/info-gnu/2010-07/msg00023.html Apr 27 18:22:53 ldir: could you please add a patch removing this line 37 from u-boot makefile, I think this is save, I do not see that this variable name is used anywhere else Apr 27 18:23:29 sure, I'm slightly confused by 'mkimage' vs uboot. though Apr 27 18:23:51 ldir: I think you will have the same problem everywhere where u-boot is build also for targets Apr 27 18:24:18 mkimage is a tool development in the u-bot project Apr 27 18:24:42 ok, will look into it :-) Apr 27 18:33:08 ldir: nbd: Is it possible to update make on MacOS? Apr 27 18:33:51 Hauke: once we've built the toolchain won't we have our own make? Apr 27 18:35:13 ldir: I do not think so Apr 27 18:36:13 am slightly surprised the make provided by apple with xtools is so old. Apr 27 18:40:04 ldir: don't they ship updates to xtools? Apr 27 18:40:13 brew offers make 4.3 (as gmake) Apr 27 18:54:39 Hauke: travis allows you to run builds on macos. we could try compiling the tools at least in a travis run. Apr 27 19:38:05 I'd sort of point out that the build-prereq make script says "Please install GNU make v3.81 or later" Apr 27 19:54:02 I actually think a realistic way forward is to bump the pre-req to 3.82 and macos users will have to install a suitable make from homebrew...something we have to do for other gnu tools anyway. Apr 27 20:13:22 Hello! I have a system where I'm installing some packages to an SD card mounted at /opt (because I'm out of room on the spi-flash, but when I opkg --dest opt install packages there, I'm not getting my init scripts and such in /etc. How do I solve that? I have my path and library path set so it picks up what's installed there (under bin, usr/bin, etc.) but my my /etc/ files aren't getting where they need to go. Apr 27 20:13:46 Should I have some type of post-install script that symlinks them perhaps? Apr 27 20:14:11 lol, I suppose I could do some ugly unionfs hack Apr 27 20:19:33 I suppose opkg --dest wasn't designed for this :( Apr 27 20:24:46 hi, where do I send a patch I have for netifd? Apr 27 20:28:46 pespin__: the mailing list should be ok. [PATCH netifd] Apr 27 20:32:11 PaulFertser, openwrt-devel@lists.openwrt.org ? Apr 27 20:33:45 pespin__: yes Apr 27 20:35:43 I see this is not supported https://openwrt.org/docs/guide-user/additional-software/opkg Apr 27 20:36:54 ldir: I am also fine with increasing the minimum required make version Apr 27 21:08:17 strange, I don't receive a subscription email when registering https://lists.openwrt.org/mailman/listinfo/openwrt-devel with my email server Apr 27 21:08:27 but using another email server i do receive the email Apr 27 21:08:39 on my email server I see no access from postfix Apr 27 21:09:06 and I can receive email from other sides and ml there Apr 27 21:14:34 Hauke: did you see the config related to the nft_reject_ipv4.ko stuff? Apr 27 21:18:24 anyone around who’s using the x86_64 EFI stuff? bonus points if you have it working with KVM? Apr 27 21:54:06 philipp64: did not look into it again Apr 27 21:57:05 I’m trying to validate the EFI image manually… first step was trying to mount the image manually using “losetup -f …” but that complained about the file size not being a multiple of 512… Apr 27 23:17:58 aparcar[m]: ping Apr 27 23:21:19 anyone know how long the metadata is? Apr 27 23:28:30 philipp64: I don't know anything about the EFI stuff. What metadata do you mean? The one attached to images or the JSON files created separately? Apr 27 23:52:00 uhhhh Apr 27 23:52:13 do these ag71xx changes demand usage of kernel 5.4? Apr 28 00:21:28 build #282 of mpc85xx/p2020 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/mpc85xx%2Fp2020/builds/282 Apr 28 00:46:16 hi. is there any openwrt CI where the snapshots / images are generated and in which I can find the build artifacts (like jenkens or alike)? Apr 28 00:47:39 more specifically I'm looking for build logs and .config file use to generate the images: Apr 28 00:47:40 https://openwrt.org/toh/tp-link/archer-c50 Apr 28 00:47:49 *used Apr 28 00:48:35 the buildbot source is in the openwrt git Apr 28 00:50:09 build #291 of sunxi/cortexa7 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/sunxi%2Fcortexa7/builds/291 Apr 28 00:57:45 DonkeyHotei: I was actually looking for http://buildbot.openwrt.org/master/images/builders/ramips%2Fmt76x8/builds/301 Apr 28 00:58:05 nonetheless, I can't find a full build log there : ( Apr 28 00:59:17 http://buildbot.openwrt.org/master/images/ Apr 28 01:08:01 pkgadd: thanks, I'm aware of this URL, that's how I arrived to the build #301 I've posted above. What I'm really looking for is a full log.txt that was run inside the buildbot, if it exists at least ... Apr 28 01:08:46 *was ran Apr 28 01:14:07 on Ubuntu, any clue what provides that dep (apparently no distro package can provide it): Apr 28 01:14:09 /mnt/git/openwrt/build_dir/hostpkg/libubox-2020-02-27-7da66430/blobmsg_json.c:23:11: fatal error: json/json.h: No such file or directory Apr 28 01:14:09 #include Apr 28 01:14:09 ^~~~~~~~~~~~~ Apr 28 01:24:52 the logs are there, the stdio logs Apr 28 01:58:31 build #292 of x86/64 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/x86%2F64/builds/292 Apr 28 02:24:49 gromero: libjson-c-dev **** ENDING LOGGING AT Tue Apr 28 02:59:59 2020