**** BEGIN LOGGING AT Fri Mar 01 02:59:57 2019 Mar 01 08:11:13 * russell-- has been having trouble with ubnt m xw devices (tried ar71xx on nano-m5-xw, no ethernet traffic; ath79 with ynezz pull request patches on nano-m2-xw, also no ethernet traffic). reverted to r4983-g7d905f1e95, works again. Mar 01 08:12:01 this PR: https://github.com/openwrt/openwrt/pull/1737 Mar 01 09:30:03 russell--: so it works fine for you with #1735 ? Mar 01 09:31:24 BTW is anyone using git-pw with openwrt's patchwork? I'm trying to get it working, but getting this strange error `requests.exceptions.HTTPError: 404 Client Error: Not Found for url: https://patchwork.ozlabs.org/api/1.1/patches/1046790/` Mar 01 09:31:37 ynezz: yes Mar 01 09:32:36 these are device 10km away. luckily i have another ap nearby and can connect over wifi to recover ... except for the nano-m5-xw, which i'm about to try ;-) Mar 01 09:34:02 good luck :) Mar 01 09:34:30 also, thank you ipv6 link local addresses Mar 01 09:51:00 the nano-m5-xw also worked, yay! Mar 01 10:00:20 ynezz: works fine here Mar 01 10:00:48 URL must be https://patchwork.ozlabs.org/api/1.0 Mar 01 10:31:44 mangix: ah, great, thanks! Mar 01 12:14:40 ldir: you can probably fix your OSX build breakage by fixing this function https://github.com/squashfskit/squashfskit/blob/master/squashfs-tools/version.sh#L56 Mar 01 12:15:46 probably some incompatible date switches? Mar 01 12:58:25 lynxis: FYI ---^ and https://bugs.openwrt.org/index.php?do=details&task_id=2153 Mar 01 13:07:59 hi all, i need update dts only to reflect console speed update in uboot for linkit smart Mar 01 13:17:40 how to just recompile kernel with dts Mar 01 14:54:02 dedeckeh: this build breaks relay mode as far as I can tell: https://git.openwrt.org/?p=openwrt/openwrt.git;a=commitdiff;h=331963717bf4781ead1ff03b56de2802b9e05164 Mar 01 14:54:17 dedeckeh: when I build without that change relay mode works fine Mar 01 14:56:00 dedeckeh: my config: https://pastebin.com/VNY1ZG6R Mar 01 14:58:30 netprince:did you test the latest version (https://git.openwrt.org/?p=openwrt/openwrt.git;a=commit;h=c8153722a2eddceb6a8df5fd1a676d6d1f688640) ? Mar 01 14:58:58 dedeckeh: yes Mar 01 15:03:01 netprince:can you set loglevel to 7 and send the output via mail ? Mar 01 15:04:37 dedeckeh: yes, I have a few things keeping me busy this morning, might take me a little time. Mar 01 15:05:49 ynezz: pong Mar 01 15:06:34 netprince:np Mar 01 16:40:10 dedeckeh: is there a uci option for odhcpd to log to a file? Or should I just grep from logread after a bit? Mar 01 16:40:57 netprince:use logread | grep odhcpd Mar 01 16:41:08 ok Mar 01 16:43:32 dedeckeh: what is your email address? Mar 01 16:43:44 dedeckeh@gmail.com Mar 01 16:53:44 dedeckeh: just sent, also trying to see what I can see from this log file Mar 01 16:54:01 netprince:looked into your config; can you remove the config dhcp 'wan' section and check if you still have the issue ? Mar 01 16:54:44 sure, I will try that Mar 01 16:57:32 dedeckeh: no, still cant ping the gateway after removing config dhcp 'wan' section **** BEGIN LOGGING AT Fri Mar 01 17:26:40 2019 Mar 01 20:30:41 Any "tricks" on how to get "patchwork" not to mangle/lose patches send to the list by email? Mar 01 20:31:24 `git format-patch HEAD^` and then emailing it, text-only delivery, hasn't been very robust for me Mar 01 20:31:39 (Thunderbird, macOS) Mar 01 20:32:12 I believe git-send-email is the preferred method, though getting the arguments right can be a bit of an exercise. Mar 01 20:32:40 Ugh, OK, I looked at that once and yes, quite the list of arguments! Mar 01 20:32:51 I'll give it a shot again Mar 01 20:33:19 When I submitted a patch with it a long time ago, I just emailed them to myself initially until I was sure I had everything right. Mar 01 20:33:49 Good idea, thanks -- trying --dry-run wasn't terribly effective for me Mar 01 20:52:17 Is there a staging tree with Linux 4.19 for the ath79 target up? Mar 01 20:53:45 there is this PR: https://github.com/openwrt/openwrt/pull/1635 Mar 01 20:55:16 Thanks tmn505! :thumbs-up: Mar 01 20:55:37 Now I can beat on NAND support Mar 01 22:37:20 https://github.com/kaloz/mwlwifi/issues/353 Mar 01 22:43:54 Well, patchwork apparently just hates me -- even with referring to the thread, it created a new patch series. Mar 01 22:44:30 ldir: ping Mar 02 00:38:51 Is there a way, from the command line, that the underlying MTD partition can be determined for a UBI device? Mar 02 00:39:45 ubinfo and `find /proc -name '*ubi*'` didn't seem to ehlp Mar 02 00:44:12 Found it -- /sys/devices/virtual/ubi/ubi0/mtd_num Mar 02 00:47:04 er, how would one configure a monitor port to sniff VHT-80 traffic with channel set to 40? Mar 02 00:47:31 this does not work it seems: iw dev moni300 set freq 5200 80 5190 Mar 02 00:51:30 jeffsf: claws-mail seems to be okay'ish for mailing patches (needs for attention, as in disabling signing), but git-send-email is the preferred option Mar 02 00:52:12 ahh, tis: iw dev moni300 set freq 5200 80 5210 Mar 02 00:52:37 pkgadd_: Thanks, got most of the command-line arguments figured out, but somehow, even setting --in-reply-to= Patchwork decided it was a new series :( Mar 02 00:53:03 At least I'm over the worst of it! Mar 02 00:53:41 Back to EA8300 booting... Mar 02 00:54:32 how are you getting along? looks like a cool device, just quite a bit more involved to get it ported than I hoped Mar 02 00:55:08 Other than the QCA9888 problems with crashing, I've got most of it worked, I think. Mar 02 00:55:36 Having challenges with overriding the boot loader args, even with the patch used for other IPQ devices Mar 02 00:56:02 So tweaking my local bootargs in U-Boot so that I can debug some strangeness with sysupgrade Mar 02 00:56:32 Then I can go back to getting it bootable for Joe Average without needing serial Mar 02 00:57:27 yep, that might make it easier for a start Mar 02 00:58:12 Want to wrap up the EA8300 so I can get on NAND for ath79 on Linux 4.19 (GL.iNet AR750S) Mar 02 00:58:32 As well as not feeling like I've got a $130 paperweight ;) Mar 02 01:00:41 Not used to how fast NAND flashes compared to the NOR! Mar 02 01:00:57 regarding QCA9888, ath10k{,-ct} has quite some problems without the correct board files and pre-cal, it's basically the most likely cause for issues Mar 02 01:01:10 I had quite serious issues on my nbg6817 before that got sorted Mar 02 01:01:56 (didn't work at all, without reducing txpower by 3 dBi; now it's all fine with defaults) Mar 02 01:02:10 Yeah, I know what OEM is using and have that in place. greearb_ looked at some of my output, but nothing super obvious there Mar 02 01:02:50 I need to figure out how to use the Swiss army knife to wrap the OEM firmware and see if that is at least functional Mar 02 01:03:09 I just have a haunting suspicion that I'm missing or have a GPIO wrong somewhere Mar 02 01:04:52 I did see a commit on `master` around "mac80211: add a fix to prevent unsafe queue wake calls during restart" Mar 02 01:05:28 and I saw that there were new CT driver/firmware releases as well Mar 02 01:20:48 (sysupgrade working now, little yay of the day) Mar 02 01:21:16 :) Mar 02 01:21:52 can you get telnet or ssh access on the OEM firmware (easily)? Mar 02 01:22:27 Yes, it's been on serial since the day I tore off the shrink wrap ;) Mar 02 01:22:40 Have full root access Mar 02 01:22:41 if yes, a simple start might be to hardcode OpenWrt to one partition set and setting the bootorder accordingly Mar 02 01:23:27 well, you have serial, Joes Sixpack likely won't - that's why I was asking for enduser available telnet/ ssh Mar 02 01:24:37 Oh, access _without_ serial? No, haven't looked for weaknesses in the OEM firmware yet Mar 02 01:25:09 Just confirmed how the "sign" their firmware that comes in through the web interface -- that I've got working too Mar 02 01:25:43 s/how the/how they/ Mar 02 01:26:04 e.g. ZyXEL allows enabling telnet and ssh from the normal (expert) GUI, Netgear from a 'hidden' /debug.htm page Mar 02 01:26:36 I've got full image of the running device's file system, as well as Mar 02 01:27:02 unpacking the UBI filesystem from the OEM binary upgrade firmware Mar 02 01:27:10 so might have a look through that Mar 02 01:29:19 I didn't see sshd or anything "exciting" in `find ubifs/ -executable` Mar 02 01:30:57 ubifs/etc/registration.d/15_ssh_server is tantalizing, but /etc/init.d/service_sshd.sh doesn't exist Mar 02 01:48:24 LOL at OEM code -- echo -ne "\tProgramming NOR flash (takes a long, long while, please wait)...\n" >> /dev/console Mar 02 01:51:42 hmm, doesn't the ea8300 have NAND? Mar 02 01:52:04 but yes, >=16 MB NOR and/ or 4K sectors can take quite a while Mar 02 01:56:53 Generic code for Linksys -- just 256 MB NAND, direct boots from it Mar 02 01:57:43 It's a nice box, looking forward to getting it up and solid with three radios Mar 02 01:58:39 yep, similar situation with the OEM firmware code on zyxel - interestingly looking for "daulflag" (yes, that typo isn't mine) throughout the code ;) Mar 02 02:13:08 OK, dinner time here in California -- have a great night/weekend all **** ENDING LOGGING AT Sat Mar 02 02:59:57 2019