**** BEGIN LOGGING AT Sun Jul 15 03:00:01 2018 Jul 15 08:36:26 hmmmm Jul 15 09:12:26 morning Jul 15 09:39:34 hmph, what have I done wrong? doesn't the cmake infrastructure already build outside of the source tree? Jul 15 09:39:43 Building in source directory is forbidden. Please make separated build Jul 15 09:39:44 directory. Jul 15 09:39:44 :( Jul 15 09:46:39 seems like serious issue...? Jul 15 09:47:50 well yes.... Jul 15 09:47:51 :P Jul 15 09:49:37 Create a "build" dir with user rights, iirc that's all you need to do, then try to make again from that dir. Jul 15 09:49:59 But that's a really old memory... Jul 15 09:50:19 right, except openwrt build system has infrastructure to support cmake, so it isn't that easy Jul 15 09:50:22 hh Jul 15 09:50:22 heh Jul 15 09:54:17 weird, hm Jul 15 09:57:50 morning Jul 15 10:08:36 it builds with older versions, so probably just busted cmakelists Jul 15 10:09:09 I avoid cmake like the plague though, so no idea how to fix i Jul 15 10:09:10 t Jul 15 10:11:31 Hi is flow offloading all in the firewall? Jul 15 10:11:49 Would it do any thing for a dumb AP Jul 15 10:12:12 its only for forwarding iirc Jul 15 10:12:17 K thanks Jul 15 10:14:22 bleh, fml Jul 15 10:14:27 My dumb ap has firewall and dnsmasq plus odhcp disabled Jul 15 10:16:19 why does your ap have a firewall? :D Jul 15 10:17:14 because building an image without one is an extra step? Jul 15 10:17:43 just coz its there doesn't mean you need to run it heh Jul 15 10:18:36 meh, screw this cmake nonsense Jul 15 10:18:44 come back autotools Jul 15 10:18:48 all is forgiven Jul 15 10:18:51 lol Jul 15 10:19:21 but also... why is openwrt trying to cmake . Jul 15 10:19:34 +build Jul 15 10:20:07 essentially anyway, it passes the src dir in build_dir/ as the last arg Jul 15 10:20:23 hi Jul 15 10:20:58 sup Jul 15 10:22:03 i'm stealing ldir's .55 bump for 18.06 and playing it safe Jul 15 10:27:32 raaaaaaaaaage Jul 15 10:32:45 jwh the firewall is disabled Jul 15 10:33:22 along with dnsmasq and odhcp Jul 15 10:33:26 good man Jul 15 10:33:34 I run a snapshot build Jul 15 10:34:06 If i built my bins I would build with the unneded packages removed Jul 15 10:34:12 https://gist.github.com/joeholden/590343b7506af2f96efa1daab7c54b9e#file-gistfile1-txt Jul 15 10:34:15 .. Jul 15 10:35:07 I have talked about Openwrt doing AP builds but It seems it would be to much effort Jul 15 10:35:31 heh Jul 15 10:35:58 It would be cool tho to havve a luci-app-AP that linked back in to the mane router with openwrt on it and report the states of the ap back to the mane router Jul 15 10:36:22 sounds like something to make :D Jul 15 10:36:38 Not mate I wish Jul 15 10:36:41 could just do it like the draytek vdsl modems do (except not encrypted) Jul 15 10:36:49 just broadcast a packet with info in Jul 15 10:36:49 I code as well as a lump of cool Jul 15 10:36:54 I still haven't managed to decrypt it Jul 15 10:37:10 thats some idiocy right there, encrypting useful info that isn't giving anything away Jul 15 10:37:12 Cole! Jul 15 10:37:15 * Jul 15 10:37:21 coal :P Jul 15 10:37:26 Tapper: AP builds sound indeed overkill. But a click-button solution (e.g. in Luci) to switch from router (firewall/dns/dhcp) to pure AP (transparent bridge) might be something worth looking into Jul 15 10:37:42 it's not specially hard to do by hand but it might improve the end user experience :) Jul 15 10:38:05 needs more setup wizards :D Jul 15 10:38:13 It would be cool to log on to your mane router and have the status of your APs shou up on the mane page Jul 15 10:38:37 yeah that'd be cool as well. Getting openwrt one step closer to world domination ;) Jul 15 10:38:39 Tapper: main = primary ; mane = lion neck hair Jul 15 10:38:47 also what russell-- said ;D Jul 15 10:39:12 But I like lion neck hair haha Jul 15 10:39:19 Tapper: that already exists, unifi :D Jul 15 10:39:27 I hate English Jul 15 10:39:34 lol, so do we all Jul 15 10:39:37 lol Jul 15 10:40:21 well, and what is there to love about coal? Jul 15 10:40:22 I should just talk patwa! Jul 15 10:40:36 I dont like coal! Jul 15 10:42:15 patois is french Jul 15 10:42:45 ok, I fixed it with a well named patch Jul 15 10:42:46 there are several patois too :) Jul 15 10:42:52 Tapper: there are bumper stickers here, being near coal mining areas, common ones: "Friend of Coal" I have long wanted to make the little stencil that changes that to "Friends of Cole" (Cole being a nick name here). Jul 15 10:43:42 friend of functioning lungs Jul 15 10:44:40 friend of not living on a baron sphere Jul 15 10:44:44 jwh: try setting CMAKE_BINARY_SUBDIR=build Jul 15 10:44:45 baren? Jul 15 10:44:51 jow: aaah Jul 15 10:44:58 I was scouring cmake.mk but didn't see that :( Jul 15 10:48:41 more hilarity anyway :D Jul 15 10:49:46 INSTALL FILES given unknown argument Jul 15 10:49:46 "/home/jwh/nspv2/staging_dir/toolchain-x86_64_gcc-7.3.0_musl/etc". Jul 15 10:49:47 :D Jul 15 10:51:28 possible stray \, bleh Jul 15 10:51:43 nope, not that Jul 15 10:54:47 more patches required Jul 15 11:07:35 jow: do you have some time to do the lede 17.01.5 release? Jul 15 11:09:39 which sec issue? Jul 15 11:11:28 i mean, what prompted a 1701.5? Jul 15 11:12:20 there are already a lot of patches in the branch Jul 15 11:12:36 some are also security related Jul 15 11:13:22 ntd: .5 was planned as a general maintenance release afaik Jul 15 11:13:26 hey, i'm all for a point release :) Jul 15 11:14:06 but these usually fix a remote exploit in base/included tools/kernel Jul 15 11:14:36 ntd: one of the main issues causing the lede/openwrt split was the lack of maintenance releases between major o nes Jul 15 11:14:39 * ones Jul 15 11:14:45 .3 and .4 were dnsmasq and dropbear, no? Jul 15 11:15:01 there are many bugfixes in the kernel Jul 15 11:15:07 sure Jul 15 11:15:12 but I do not know how many are relevant for normal openwrt use cases Jul 15 11:16:19 hey, current kernel is full of holes but afaik there require local access. count me in for .5 Jul 15 11:17:14 speaking of releases, why is 1806 stuck at rc1? Jul 15 11:26:26 also, huawei is quite good at providing updates for the LTE sticks (with changelogs, lordy at the number of vulnerabilities) Jul 15 11:26:46 well, they make updates, they just don't make them available through their support pages Jul 15 11:26:49 ... Jul 15 11:27:14 i've found that recent updates confuse both networkmanager/modemmanager and lede Jul 15 11:29:12 ntd: 18.06 is stuck at rc1 because I had many luci bugs to solve first, I actually planned to tag rc2 today Jul 15 11:29:20 Hauke: yes, we can do it right now Jul 15 11:30:18 nice on both counts. anyone know if modem scripts/tools have been updated for 1806? Jul 15 11:30:52 i ended up having to do testpoint downgrades of 50+ modems :) Jul 15 11:30:56 jow: ok Jul 15 11:31:18 jow: is there something documented? Jul 15 11:31:18 ah beautiful, a user of ucontext and get/setcontext... hello darkness my old fiend Jul 15 11:32:30 Hauke: let me dig out the irc logs Jul 15 11:32:56 ok Jul 15 12:22:57 hmm. I don't know how I did that but I'm getting make world to fail on missing staging_dir/host/bin/usign Jul 15 12:23:13 * f00b4r0 has a talent for breaking things in unexpected ways Jul 15 12:25:35 you probably already wiped staging_dir? Jul 15 12:25:46 i did not Jul 15 12:25:56 it's a fresh checkout/make menuconfig/make world Jul 15 12:27:17 i did strip the config down a lot since I was only interested in making sure my kernel changes build. But i'd expect the build system to prevent me from disabling some mandatory config options (in fact I'd expect them to not be configurable at all) :) Jul 15 12:29:17 there is a command to just rebuild the kernel Jul 15 12:29:31 you still need the toolchain etc Jul 15 12:29:52 anyway, it's probably some obscure bug nobody else is ever going to hit, so nevermind Jul 15 12:31:12 * f00b4r0 pings rmounce, need testing some fix that should probably be included in 18.06 Jul 15 12:32:21 hi all, I'm trying to add a new board to OpenWRT and I've a few questions, could someone help? Jul 15 12:33:27 f00b4r0: I think there are some indirect build dependencies not properly expressed, and it really is much easier to just do a simple build instead of trying to find out what make targets to invoke in which order (especially since one of the kernel build steps includes building images, which requires almost everything anyways) Jul 15 12:34:03 ack. That's exactly why I ran make world instead of make whatever builds the kernel only :) Jul 15 12:34:19 that's also why my script does that ;) Jul 15 12:35:09 the failing step is Signing package index... Jul 15 12:35:09 bash: line 3: /home/varenet/openwrt/staging_dir/host/bin/usign: No such file or directory Jul 15 12:35:09 the failure is quite simple Jul 15 12:35:09 Signing package index... Jul 15 12:35:09 bash: line 3: /home/varenet/openwrt/staging_dir/host/bin/usign: No such file or directory Jul 15 12:35:21 rah, sorry for double paste Jul 15 12:40:05 I'm very new to OpenWRT. The board I'm trying to add is the Nanopi Neo2, it's very similar to the Nanopi Neo Plus2, which is already added in the 18.06 branch Jul 15 12:40:31 So I've been looking at the commits that added support for the Plus2 https://git.openwrt.org/?p=openwrt%2Fopenwrt.git&a=search&h=HEAD&st=commit&s=NanoPi+NEO+Plus2 Jul 15 12:41:18 The files package/boot/uboot-sunxi/Makefile and target/linux/sunxi/image/cortex-a53.mk seem simple enough, I've made the changes there Jul 15 12:42:27 However, some of the changes are in the form of patches that backport the device tree and gigabit ethernet support Jul 15 12:42:44 ldir: ping Jul 15 12:43:01 For example https://git.openwrt.org/?p=openwrt/openwrt.git;a=commit;h=d6634edfc1f0177cb19e52441d71d7e3e03807c2 Jul 15 12:43:26 And https://git.openwrt.org/?p=openwrt/openwrt.git;a=commit;h=b7a1aa4df2a9834bb7339712670abbe1a05dc01c and https://git.openwrt.org/?p=openwrt/openwrt.git;a=commit;h=6247929d66f1cdc6124b5f843416c3146975f0b8 Jul 15 12:43:48 Just wondering how I go about creating these for the Neo2 :) Jul 15 12:46:52 Never worked with patching and backporting, I've read the guides on the wiki but still some pointers would be appreciated Jul 15 12:51:09 leperet: maybe this helps https://openwrt.org/docs/guide-developer/start Jul 15 12:51:54 and https://openwrt.org/docs/guide-developer/adding_new_device Jul 15 12:53:11 Hi Borromini, thanks, I've read both those pages and I think I know what files I need to add/change Jul 15 12:53:33 ok :) Jul 15 12:54:05 It's just the how I'm stuck on:) Jul 15 12:54:20 oh :( Jul 15 12:54:48 the backport patches are from upstream (ie newer kernels), but i suppose you already gathered that Jul 15 12:54:54 Yup :) Jul 15 12:55:09 For example take this commit https://git.openwrt.org/?p=openwrt/openwrt.git;a=commit;h=d6634edfc1f0177cb19e52441d71d7e3e03807c2 Jul 15 12:55:47 you know how git works i presume Jul 15 12:55:54 maybe i'm not understanding your question right Jul 15 12:56:27 eg. Antony Antony committed this for the Neo Plus2, I need to do the same for the Neo 2. It's the same hardware for GigE Jul 15 12:57:19 yeah he sent in the patch and Hauke accepted and applied it Jul 15 12:57:43 you send patches through git send-email, to the openwrt devel mailing list Jul 15 12:57:45 sounds like both hardware should share a common dtsi Jul 15 12:58:41 Yup, very similar. Some things are different (the smaller board has no emmc and no built in wireless) Jul 15 12:59:19 Just wanna know how to create a similar patch, I guess it's a patch on kernel 4.14 from something from 4.16 Jul 15 13:00:32 The link to the patch on that commit takes me to an unknown page, then when I search google for the code 27d7f9297027 I get only LEDE/OpenWRT results :) Jul 15 13:01:11 leperet: is the board already supported by mainline Linux kernel Jul 15 13:01:17 Yup Jul 15 13:01:19 and u-boot Jul 15 13:01:57 As you can see from all the commits for the Neo Plus2 https://git.openwrt.org/?p=openwrt%2Fopenwrt.git&a=search&h=HEAD&st=commit&s=NanoPi+NEO+Plus2 very few commits were required to add support Jul 15 13:02:01 Yup Jul 15 13:02:16 leperet: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=27d7f9297027 Jul 15 13:02:40 amazing, thank you Jul 15 13:03:00 you can generate a patch like this: "git format-patch -1" Jul 15 13:03:23 or click the link in the web ui ;-) Jul 15 13:05:00 haha, I'm just starting at this and want to understand properly. Am I in theory checking out the 4.14 branch of linux, applying the changes from the 4.16 commit and then generating the strange formatted patch file that Antony Antony committed here https://git.openwrt.org/?p=openwrt/openwrt.git;a=blob_plain;f=target/linux/sunxi/patches-4.14/061-arm-dts-sun50i-support-for-nanopi-neo-plus2-board.patch;hb=d6634edfc1f0177cb19e5244 Jul 15 13:06:17 it's not a strange patch file. It's a patch file committed to a git repo. Every commit being a diff itself, the patch appears prepended with '+' Jul 15 13:06:32 a patch to a patch, so to speak Jul 15 13:07:50 Yup strange just means unfamiliar to me, seems completely logical (additional lines and even new files indicated by pluses and removals by negatives, elegant) Jul 15 13:13:37 Hauke, where is the link in the webui I can use to generate a patch? Thank you all, you've given me plenty to go and investigate further on my own Jul 15 13:14:55 leperet: there is a link called patch Jul 15 15:14:08 f00b4r0: eh? Jul 15 15:14:33 'rah' as in 'argh' :) Jul 15 15:14:42 rc2 today? Really!!!??? Jul 15 15:18:28 holy shit I'm guessing from the screaming outside the game has started Jul 15 15:18:33 and something happened ;P Jul 15 16:01:25 Hello. On the live system, where are the packages' pre-rm scripts are stored? I need to manually clean up messed up makefile. Jul 15 16:12:08 stangri: its usually all in /usr/lib/opkg/ Jul 15 16:12:23 ty Jul 15 16:12:53 jow: btw, the uci del_list always returns 0 even if nothing was deleted. Is it intentional? Jul 15 16:13:13 uci del only returns 0 if stuff was actually deleted. Jul 15 16:17:32 stangri: yeah, an old bug Jul 15 16:51:14 jow: what's the preferred way of getting fixes into 18.06? Jul 15 17:08:02 hello, i've got a TPLINK WR740N v7 which seems to be identical to the v6 but there's no firmware on dl page for it, what can i do? Jul 15 17:21:28 updated openwrt/upstream, https://sdwalker.github.io/uscan/index.html Jul 15 17:42:55 sleepenthusiast: this was asked already. the wr740n v7 is a wr741 v6 Jul 15 17:48:11 so umm, how do I workaround lack of get/setcontext in musl? I don't really get what ucontext is Jul 15 17:48:20 I mean I could mark it as glibc only but that sucks Jul 15 17:59:30 or I guess.. I could go back to the last release and just cherry pick patches since ucontext stuff was only added recently (and accept some of the performance loss it intended to fix) Jul 15 18:05:24 hmm, rc2 :-) Jul 15 18:10:57 weee Jul 15 18:11:12 userland builds now Jul 15 18:39:12 @DonkeyHotei - yeah, it was asked by me before i believe, did someone else ask this too? will there be official support? Jul 15 18:39:20 now to tackle kmods Jul 15 18:40:06 f00b4r0: start / reply to a thread on the ml asking for specific commits to be cherry picked / backported Jul 15 18:40:30 f00b4r0: if the fix is not yet in master do a pr / patch to ml and ask for cherry picking / backporting in the cover letter or commit message Jul 15 18:41:03 jow: ok. I have a series of fixes for spi nor routerboards that I'm trying to get submitters to test (since I dont have the hardware), I suppose it's better to do that first before asking for cherry-picking Jul 15 18:41:23 ok makes sense. Jul 15 18:41:30 adding new devices is no longer an option I suppose? Jul 15 18:44:59 jow: uci del_list always returns 0 even if nothing was deleted. Is it intentional? uci del only returns 0 if stuff was actually deleted. Jul 15 18:45:48 stangri: you mean the cli tool? Jul 15 18:45:50 jow: reason I'm asking is that I've made my patches incremental after what I have submitted as PR#1170. If that's a no go for 18.06 I'll reorder the patches for easier picking Jul 15 18:45:55 jow: yes Jul 15 18:46:16 f00b4r0: could be too late but if there's large interest I don't see why not Jul 15 18:46:28 stangri: probably a buf Jul 15 18:46:29 *bug Jul 15 18:46:36 flyspray? Jul 15 18:46:44 dunno about large interest. the hAP mini is the cheapest device mikrotik currently offers, afaik Jul 15 18:46:51 does that count? ;) Jul 15 18:46:59 stangri: patch would be preferred ;) Jul 15 18:47:12 feel free to open a bug but I'm not sure it'll be worked on Jul 15 18:47:13 few people have been looking at hap, would be nice I guess Jul 15 18:47:45 f00b4r0: final is currently planned for next weekend sat/sun/mon Jul 15 18:47:51 until then stuff can be merged still Jul 15 18:48:06 preferably until thu/fri latest Jul 15 18:49:09 mmm, in that case, may have to send one more patch Jul 15 18:50:20 jow: ok. https://github.com/f00b4r0/openwrt/commits/rbspifixes these are the fixes. Of the last 4 commits, all but the last are 18.06 material. Jul 15 18:50:32 other bits are packages, so guess they could go in at any tiem Jul 15 18:50:34 time Jul 15 18:50:56 * f00b4r0 supposes it's impossible to have incremental PRs Jul 15 19:20:26 * f00b4r0 pulls hair. Submitter of 'support' code didn't actually try to flash device. Turns out flash chip is unsupported \o/ Jul 15 19:23:10 :D Jul 15 19:25:55 * ldir blames the French ;-) Jul 15 19:26:24 mm, using AUTOLOAD for KernelPackage, can I just specify a bunch of modules? Jul 15 19:26:41 like... AUTOLOAD:=$(call AutoProbe,blah1) $(call AutoProbe,blah2)? Jul 15 19:26:45 or... Jul 15 19:27:41 not sure its worth splitting them up since they come as a bundle Jul 15 19:28:04 and have dependencies on each other, would be a nightmare to maintain Jul 15 19:29:23 does anyone build openwrt from source? Can you try fixing a few lines in one file, build an image and then test the result with a cli command? Jul 15 19:30:21 jwh: you only need to specify the "bottom" one, and any dependencies will be automatically loaded (as long as they are direct dependencies) Jul 15 19:30:48 yeah this code is pretty terrible Jul 15 19:30:49 :D Jul 15 19:31:11 it also appears to conflict with an in-tree module, so may need to fix that too Jul 15 19:31:24 stangri: yes....? what's the real question/problem? Jul 15 19:31:25 as long as modinfo shows the dependencies, they work fine Jul 15 19:31:44 * jwh looks Jul 15 19:32:31 no dependencies declare in the modules, sigh Jul 15 19:32:43 declared Jul 15 19:32:51 karlp: thanks, PMed. Jul 15 19:33:03 nope. not happening. Jul 15 19:33:10 I *think* they will work independantly to a degree, but if you're gonna install the package, may as well load the lot Jul 15 19:33:24 otherwise its like driving a car with 3 whees Jul 15 19:33:25 wheels Jul 15 19:33:34 jeez, what a drama queen Jul 15 19:33:39 lol Jul 15 19:33:55 testing is easy, does your fix produce the required results? :D Jul 15 19:34:13 if yes; send patch, if no; try again Jul 15 19:35:36 stangri: just ask here like a rational person. Jul 15 19:36:03 you think I'm going to be the only possible person who can answer your question? Jul 15 19:36:16 asking everyone individually in PM is just wasting everyone's time. Jul 15 19:36:22 ldir: ping Jul 15 19:36:33 f00b4r0: le pong Jul 15 19:36:37 ;) Jul 15 19:45:46 karlp: need someone to test a theory in the first comment: https://bugs.openwrt.org/index.php?do=details&task_id=1654 Jul 15 19:47:03 so file a patch? what exactly were you expecting me to do via PM on that? Jul 15 19:47:17 you can build openwrt too you know? Jul 15 19:50:22 jow: hey, did you have chance to look at the patch I sent to the list for fw3? is it good or does it need tweaking? Jul 15 19:53:09 karlp: thank you, you've been a great help. Jul 15 19:59:53 np, you're welcome. Jul 15 22:24:48 newest 1806 build saw this: WARNING: Makefile 'package/feeds/luci/luci-nginx/Makefile' has a dependency on 'nginx-mod-luci', which does not exist Jul 15 22:25:19 did you update feeds? :D Jul 15 22:29:05 yes of course Jul 15 22:37:22 mmm Jul 15 22:37:23 so Jul 15 22:37:51 if I have KernelPackage definitions in a package, how do I depend on those? just the name? or do I need to prefix it with kmod or? Jul 15 22:40:02 should probably make them a submenu of the package if I can, not sure splitting them up is a worthwhile endeavour since its pretty useless without the kmods Jul 15 22:45:00 A question...is it possible for dnsmasq to route DNS queries from a specific device for a specific domain to a specific DNS server? Jul 15 22:45:11 #openwrt Jul 15 23:12:29 what is libatomic in openwrt for? does mips in general have its hardware-backed atomic instructions? Jul 15 23:12:50 since it is not selected by default i assume the answer is yes... Jul 16 00:32:16 I'm trying to patch up some files for the BR-6478AC-V2 according to some old patches from 2016, but need some guidance on interpreting and writing the changes Jul 16 00:32:19 (not a coder) Jul 16 00:32:43 specifically, the last part of this patch, dealing with mtdsplit_uimage.c Jul 16 00:32:47 http://git.panaceas.org/cgit.cgi/openwrt/em-br6478acv2/master-187ad058-pq/tree/master/jmm.patch Jul 16 00:33:13 +@@ -305,32 +313,41 @@ Jul 16 00:33:39 the patch fails because mtdsplit_uimage.c has changed since the patch was written Jul 16 00:33:57 the old .c file had a switch, while the new file has two if statements in its place Jul 16 00:34:53 many things have changed since kernel 4.4 Jul 16 00:35:20 I'm sure they have Jul 16 00:36:17 but this is the current issue I'm trying to tackle with compiling the firmware Jul 16 00:37:39 the patch is necessary due to some issues with checksums and some magic around padding and ordering Jul 16 00:40:23 https://forum.openwrt.org/viewtopic.php?id=67192&p=1 gives a run-down. there was significant work done on it by a few people, but never finished. Jul 16 00:46:20 what a mess Jul 16 00:46:45 that's not a promising comment Jul 16 00:48:51 it's an example of work that ended up deferred due to the lede-openwrt re-merge Jul 16 00:49:26 it's not likely to get serious attention during the 18.06 release cycle Jul 16 00:52:05 understandable that it fell by the wayside, not exactly a renowned router.. but I'd be okay just hacking it (or pretending I can) in the mean time Jul 16 00:53:54 the only part I'm stalled on (right now) is that switch statement Jul 16 00:59:09 the patch to that function does not look like a great approach in the long run, but its original author is not here to ask Jul 16 01:06:59 if you can still be here around 6 hours from now, many more people with ideas may be awake Jul 16 01:07:17 GREAT SUCCESS Jul 16 01:07:19 I mean hello Jul 16 01:08:09 almost done making accel-ppp work Jul 16 01:08:36 kinda horrible but what you gonna do if your libc doesn't implement commonly used functions Jul 16 01:25:59 I will check in later, thanks DonkeyHotei Jul 16 01:42:51 warrick_, do you have a link to the source for the current version? Jul 16 02:02:06 Mister_X: are you asking for the vendor gpl tarball? Jul 16 02:03:14 just the source of the file that cause an issue Jul 16 02:03:25 I want to see if I can figure out how what to change Jul 16 02:05:44 i'm not sure the author of that patch knew what he was doing. compare the image generation for the br-6475nd Jul 16 02:11:31 I just wanted to see if I could help with a quick hack Jul 16 02:11:37 to get it working Jul 16 02:17:33 looks possible **** ENDING LOGGING AT Mon Jul 16 03:00:01 2018