**** BEGIN LOGGING AT Fri Feb 14 02:59:59 2020 Feb 14 07:23:43 karlp: no, the actual button will be ev.target in your onclick handler Feb 14 08:52:17 just tried building xback's staging tree, ran into this: Feb 14 08:52:18 ERROR: module '/home/openwrt/src/lede/build_dir/target-mips_24kc_musl/linux-ath79_generic/backports-5.4-rc8-1/drivers/net/wireless/ath/ath9k/ath9k_pci_owl_loader.ko' is missing. Feb 14 09:01:16 jow: https://pastebin.com/g6jv7wEL - writing the 'stored timestamp' parameters as a list breaks the startup. Feb 14 09:16:38 sorry if that comes across as a complaint... it's not. An observation/bugette report :-) Feb 14 09:29:47 ldir: already pushed a fix for that Feb 14 09:30:57 fantastic - thank you - rebuilding :-) Feb 14 09:45:49 argh, gmail managed to flag the whole mt7621/pause fram conversation as spam Feb 14 09:45:52 great Feb 14 09:54:35 kristrev: which pause frame conversation? Feb 14 09:55:05 [PATCH] ramips: gsw_mt7621: disable PORT 5 MAC RX/TX flow control by default Feb 14 10:01:05 found it Feb 14 10:01:10 thank you Feb 14 10:03:29 np Feb 14 10:05:04 kristrev: and the reason for flagging it as spam? those Ref: URLs? Feb 14 10:06:00 ynezz: It is a very good question, I cant find a pattern. I see some mails from netdev in spam as well, but most other mails (from netdev and openwrt-devel) are not flagged Feb 14 10:06:12 some mails have URLs, others dont Feb 14 10:06:17 I guess it is AI Feb 14 10:07:11 fwiw, not flagged spam for me, but heh, an ongoing issue. mailing lists should implement forward error correction to cope with gmail fade Feb 14 10:57:29 kristrev: ynezz: about the mt7621 pause frame issue. Currently having a easy to reproduce variant. I reduced the dma_size to 16 entries and set tx_thres half of dma_size. Now it is easy to get a timeout. Just run iperf3 on the device. Feb 14 10:58:13 It was a bit late so I did not debugged that well. Feb 14 11:00:02 In my first attempt, tx_thress was bigger then dma_size so it was only filling the 2 first entry. Feb 14 11:35:06 jow: yeah, I got ev.target, and L.bind got me to that inside the .then handler of a request, but how can you get the actyual node from another option/formvalue/section, etc? Feb 14 12:40:01 karlp: I added an getUIElement(section_id) for that Feb 14 12:42:15 karlp: otheropt.getUIElement(section_id) Feb 14 12:46:35 ok, neat. I think that's enoug from your end of it, the rest is just me dealing with raw dom manip stuff :) and this DSL of creating nodes on the fly. Feb 14 12:46:38 E() stuff. Feb 14 12:49:11 actually, wait... getUIElement() was wrong Feb 14 12:49:35 use var domNode = m.findElement('id', otheropt.cbid(section_id)); Feb 14 13:13:37 jow: I've just pushed the brcm target renames. Is there anything to do for the snapshot build bot or will it recognize the updated folder automatically? Feb 14 13:15:11 adrianschmutzler: needs a reconfig, wil ltake care of it Feb 14 13:15:51 brcm63xx->bcm63xx, brcm47xx->bcm47xx, brcm2708->bcm27xx. Thanks! Feb 14 14:23:44 hej - poszukuje pomocy kogos doswiadczonego w OpenWRT (bcm53xx) Feb 14 14:25:03 chcialbym sprobowac dodac wsparcie dla ASUS RT-AC5300 ... ale najpierw chcialbym zasiegnac opini kogos bardziej w temacie czy to w ogole wydaje sie mozliwe :) Feb 14 14:25:54 czy sterowniki brcmfmac maja szanse zadzialac? troche pomocy tez bym potrzebowal z DTS-ami :) Feb 14 14:26:08 czy moglbys pomoc? Feb 14 14:50:15 RobertP: English please Feb 14 14:51:21 my bad - thought it was a priv message :) Feb 14 14:51:26 won't happen again Feb 14 14:54:19 * Slimey hugs RobertP with a rather large squid Feb 14 14:54:40 lol Feb 14 20:32:23 symbol PACKAGE_nginx-util depends on PACKAGE_nginx-ssl-util Feb 15 01:21:43 https://i.imgur.com/v3VFsvb.png interesting panic after building a new image... Feb 15 01:48:16 error relocating /lib/libubus.so Feb 15 01:50:52 last build was from dec 11 Feb 15 02:06:50 nyt: was it a clean source tree? Feb 15 02:07:08 yes Feb 15 02:07:13 oh Feb 15 02:07:22 only difference was enabling wanguard Feb 15 02:07:31 and merging in the latest master Feb 15 02:07:36 trying again with wanguard disabled Feb 15 02:07:38 i've found that things that link to libubox break if it updates Feb 15 02:07:38 but i doubt that's it Feb 15 02:07:43 since i see some libubus changes Feb 15 02:07:53 but i ran make clean between Feb 15 02:07:58 huh Feb 15 02:08:05 i found i had to do this: Feb 15 02:08:08 make package/libubox/clean package/uclient/clean package/ustream-ssl/clean package/fstools/clean package/fwtool/clean package/mtd/clean package/opkg/clean package/procd/clean package/rpcd/clean package/ubox/clean package/ubus/clean package/ucert/clean package/uci/clean package/urandom-seed/clean package/urngd/clean package/usign/clean Feb 15 02:08:26 might be a little excessive but oh well Feb 15 02:08:29 make clean wont clean out the packages there? Feb 15 02:08:45 i'd think it would Feb 15 02:09:10 can check timestamps in build_dir maybe Feb 15 02:09:32 will making clean in any of those blow out the uuid? Feb 15 02:09:37 of the build Feb 15 02:10:00 uhm not sure. where is that kept? Feb 15 02:10:05 i wish i remembered Feb 15 02:10:10 not the kmod hash? Feb 15 02:10:16 no idea Feb 15 02:10:18 ill try it tho Feb 15 02:10:46 never ran into the issue before Feb 15 02:11:17 i think the one i ran into was an abi change without a library version bump Feb 15 02:11:34 luckily i caught it in an initramfs before reflashing it... Feb 15 02:11:57 does make dirclean blow out the uuid? Feb 15 02:12:42 how do you see the uuid? Feb 15 02:12:55 just curious, idk if this or that 'make clean' affects it Feb 15 02:18:53 make clean does not affect it afaik Feb 15 02:20:01 /overlay/etc/.extroot-uuid Feb 15 02:22:02 oh that Feb 15 02:22:14 whichever that compares against i have no idea Feb 15 02:22:24 that's on the device itself right? Feb 15 02:23:16 thats the uuid of the rootfs from the image Feb 15 02:23:22 its stored on the overlay to prevent mismatch Feb 15 02:25:03 yeah. well you can just rm it and i think it recreates it next time it's mounted as overlay Feb 15 02:37:10 yeah Feb 15 02:37:16 just trying to save the headache of having to do that Feb 15 02:37:23 just not sure where the uuid is calculated from Feb 15 02:37:33 [nyt@cc:~/archive/openwrt/x86]: grep -r 3ba130b4a864 . Feb 15 02:37:33 [nyt@cc:~/archive/openwrt/x86]: Feb 15 02:37:37 not present anywhere in the build tree even Feb 15 02:37:45 unless case insensitive Feb 15 02:37:47 so prob calc'd off something Feb 15 02:40:56 yep, not case insensitive Feb 15 02:40:56 if you run 'block detect', it matches the uuid listed for /rom Feb 15 02:41:05 i think it might be baked into the squashfs Feb 15 02:41:11 it is Feb 15 02:41:17 i said that above, its the uuid of the rootfs Feb 15 02:41:26 i just have no idea what generates that Feb 15 02:41:26 oh Feb 15 02:41:32 sry didnt see that Feb 15 02:42:17 'blkid' doesnt show it though Feb 15 02:42:37 maybe busybox doesnt parse squashfs Feb 15 02:42:42 blkid does show it Feb 15 02:42:52 for me at least Feb 15 02:45:08 shrug. i have CONFIG_BUSYBOX_CONFIG_FEATURE_VOLUMEID_SQUASHFS=y Feb 15 02:45:29 is yours busybox blkid or util-linux? Feb 15 02:50:53 package_blkid=y Feb 15 02:51:09 x86 so im not space limited Feb 15 02:53:51 rebuilt without wireguard, same loop Feb 15 02:53:59 there were some libubus changes that might be responsible Feb 15 02:54:07 i may try pulling those patches and seeing if it helps Feb 15 02:55:01 dang **** ENDING LOGGING AT Sat Feb 15 02:59:57 2020