**** BEGIN LOGGING AT Wed Oct 30 02:59:57 2019 Oct 30 07:33:46 make kernel_menuconfig in latest master fails on openwrt/build_dir/target-mips_24kc_musl/linux-ath79_generic/linux-4.19.79/scripts/kconfig/zconf.tab.c Oct 30 07:42:28 on which host platform and whats complete error message, provide output of `make kernel_menuconfig V=sc -j1 JOBS=1` Oct 30 07:43:46 as you can see, it's all green http://buildbot.openwrt.org/master/images/grid ;) Oct 30 08:45:51 ynezz, I have tried that, but it did work this time... and regular make menuconfig did work too Oct 30 08:49:47 hm, interesting Oct 30 11:02:42 rmilecki: KanjiMonster: jow: ping http://lists.infradead.org/pipermail/openwrt-adm/2019-October/thread.html Oct 30 11:11:50 ynezz: ? see that link you posted Oct 30 11:13:02 rmilecki: sorry :( http://lists.infradead.org/pipermail/openwrt-adm/2019-October/001146.html proper one Oct 30 11:44:06 make[3] -C package/firmware/b43legacy-firmware compile Oct 30 11:44:18 It still does not want to go. And I do not have broadcom fw Oct 30 11:44:40 I mean, I am compiling a fw for Atheros router, still get this line in make log Oct 30 11:49:07 Necrosporus: you still need to share the complete error message, as ynezz asked. Oct 30 11:57:21 karlp, ynezz, https://pastebin.com/raw/YrVbEx4Y here is output with error. However I have since built firmware image two times and run git pull once, so it doesn't show this error anymore Oct 30 12:17:58 Necrosporus: how to reproduce it? Oct 30 12:18:15 No idea Oct 30 12:18:50 is this clean checkout of some revision, you updated some older repo to the recent master etc. Oct 30 12:23:15 it was checkout of master f526e85426aa8d232b167e871a0b329f0205f827 with a patch (probably unrelated), after building he image and running make clean, then make menuconfig several times Oct 30 12:23:58 echo CONFIG_FAT_DEFAULT_UTF8=y >> target/linux/generic/config-4.19 Oct 30 12:24:21 I have also applied this command to make kernel support non-latin filenames on USB flash with vFAT Oct 30 12:25:17 ynezz, do you think it worth investigating? Oct 30 12:26:47 nope, it looks like some broken dependency tracking or something like that Oct 30 12:27:43 After that I have tried to reset kernel config, add option I wanted manually with the command above (it was only thing I wanted to change anyway) and run make to make the image Oct 30 12:28:08 Then after image was built I run make kernel_menuconfig again but it worked Oct 30 12:36:38 I can reproduce it here as well with `make target/linux/clean kernel_menuconfig` Oct 30 13:41:53 It seems that sysupgrade preserves files etc/group etc/passwd etc/shadow despite them not being present in sysupgrade.conf and despite option -u, which should skip files identical to /rom. My files were not modified. Is it a bug? Oct 30 13:50:04 Yep, still does it (run sysupgrade again) Oct 30 14:04:07 so.. they weren't modified, and they were the same afterwards, so... everything is as it shoudl be? Oct 30 14:41:04 karlp, except they take place in JFFS2 partition. Oct 30 14:41:09 just binwalked tplink firmware, a sugar-coated luci that is Oct 30 14:41:26 i hope, i hope, tp-link donated some fund to openwrt project Oct 30 14:41:39 otherwise, should they open source their ui Oct 30 14:42:42 They should do it anyway Oct 30 14:57:20 rr123 ever mess with firmware slap? https://github.com/ChrisTheCoolHut/Firmware_Slap Oct 30 15:08:58 Slimey: thanks, unaware, trying, pip install did not work, install from source now Oct 30 15:10:37 pretty complicated stuff though , will use it to check my own work for defense Oct 30 16:49:12 Any ideas as to whether this is worth implementing long term? https://www.phoronix.com/scan.php?page=news_item&px=Intel-IWD-1.0-Released Oct 30 16:51:49 thagabe: adoes it support AP mode? Oct 30 16:52:46 Hauke: iirc its a client oriented thing, so client / mesh mode only Oct 30 16:52:56 or wifi direct Oct 30 17:11:52 and glibc only Oct 30 17:27:10 have people used it extensively on other usecases? Oct 30 17:30:28 swalker: boost, gkermit, and emailrelay seem to be bugged on uscan. can you take a look? Oct 30 20:26:24 greearb_: PACKAGE_ATH_DEBUG [=n] Oct 30 20:26:40 do you need this to be y? Oct 30 20:27:27 yes Oct 30 20:27:33 k Oct 30 20:28:28 anything else? like set COLLECT_KERNEL_DEBUG to y as well? Oct 30 21:05:35 Hmm, can someone direct me to additional documentation (beyond that in menuconfig) about COLLECT_KERNEL_DEBUG? Oct 30 21:05:57 I thought it might generate System.map but I have that with this option set to n Oct 30 21:06:19 google is not helping so far... Oct 30 21:06:41 I don't think you need COLLECT_KERNEL_DEBUG Oct 30 21:07:37 greearb_: I did a quick build (no make clean) after setting ath10k_debug - anyway to check at runtime if it took Oct 30 21:08:43 Look for things that look at least something like this: [601792.384249] ath10k_pci 0000:03:00.0: set bitrate mask, vid: 62 band: 0 Oct 30 21:08:43 [601792.384252] ath10k_pci 0000:03:00.0: mask: band: 0 legacy: 0xfff gi: 0 Oct 30 21:08:43 [601792.384254] ath10k_pci 0000:03:00.0: ht_mcs: ff ff ff 00 00 00 00 00 00 00 Oct 30 21:09:01 k Oct 30 21:09:54 i'm also interested in collect kernel debug for reasons beyond ath10k (but thankyou for speaking up) Oct 30 21:10:25 I don't actually know what it does, but that tells me it probably isn't needed for debugging ath10k at least :) Oct 30 21:19:25 how does one run 2 wireguard clients on the same openwrt router? Trying to map 1 wireguard client to WiFi1 and 2nd client to WiFi2 Oct 30 21:19:39 But it seems to always default on one or the other, no matter which WiFi I connect to Oct 30 21:41:29 * nmrh sighs... i'll have to build it the slow way Oct 30 23:40:28 https://support.mozilla.org/en-US/kb/canary-domain-use-application-dnsnet Oct 30 23:41:46 * russell-- wonders how to configure dnsmasq to turn off use-application-dns on local networks Oct 31 00:11:26 russell--: I just added it to adblock's blacklist, ** server can't find use-application-dns.net: NXDOMAIN Oct 31 00:23:30 dnsmasq address=/use-application-dns.net/ allegedly works too Oct 31 00:29:24 probably, yes. but as I've been using adblock anyways, that domain was just one more to add to the list Oct 31 00:32:58 Hi can this be used to crack open closed wifi driver blobs? https://github.com/travisgoodspeed/md380tools/wiki/GHIDRA Oct 31 00:36:39 Tapper: using disassembler practices like that is problematic in terms of mainline compatibility. it might be easier to sort out some cornercases, but it taints your source to the extent of becoming poisonous and unmergeable (e.g. acx100/ tiacx or the pwc webcam driver) Oct 31 00:37:19 O make sence. so good for a hack job but not for mane line. Oct 31 00:38:21 I don't really think it would help a hack job that much either Oct 31 00:38:56 think about it, as the big picture is basically unattainable that way Oct 31 00:39:20 the result of any disassembler is pretty bad Oct 31 00:39:45 think of unrolled loops, referred procedures, etc. Oct 31 00:39:59 --> you don't see the wood among the trees Oct 31 00:40:17 it only helps if you already know pretty well how the device works in general Oct 31 00:40:24 I wish work had not stoped on the mwlwifi" driver. I fear getting left behind on the next big kernel bump Oct 31 00:41:13 pkgadd I understand I think. Oct 31 00:41:47 s/referred/dereferred/ Oct 31 00:42:03 ? Oct 31 00:42:14 00:39:45 < pkgadd> think of unrolled loops, dereferred procedures, etc. Oct 31 00:42:19 fixing a typo Oct 31 00:43:00 O lol I feel your pane! :D Oct 31 00:44:48 try to disassemble a very simple binary (for which you have the source), something like /bin/true (or slightly more complex). the resulting disassembled source is not pretty Oct 31 00:45:00 and for anything more complex than that just confusing Oct 31 00:52:04 pkgadd This fork of the driver says that it is 69 commits ahead of kaloz:master. https://github.com/eduperez/mwlwifi_LEDE Oct 31 00:52:24 How do i get a list of what's difrent? Oct 31 00:53:13 It does not mater I found it it's all Merge remote-tracking branch 'upstream/master' Oct 31 00:53:25 sorry for the noise Oct 31 00:55:50 git clone https://github.com/kaloz/mwlwifi && git remote add https://github.com/eduperez/mwlwifi_LEDE eduperez && cd mwlwifi/ && git remote add eduperez https://github.com/eduperez/mwlwifi_LEDE && git fetch eduperez && git diff master..eduperez/master Oct 31 00:59:11 err Oct 31 00:59:35 git clone https://github.com/kaloz/mwlwifi && cd mwlwifi/ && git remote add eduperez https://github.com/eduperez/mwlwifi_LEDE && git fetch eduperez && git diff master..eduperez/master Oct 31 01:50:38 mangix: switched boost's watch file from sf to bintray, gkermit's been wrong for years, what's wrong with emailrelay? Oct 31 01:59:15 btw new cve additions are busted until cvechecker is updated to handle nvd's xml->json change **** ENDING LOGGING AT Thu Oct 31 02:59:58 2019