**** BEGIN LOGGING AT Mon Mar 30 03:00:00 2020 Mar 30 06:06:37 ynezz: piiinnnggg Mar 30 07:44:13 aparcar: meh Mar 30 07:59:38 ynezz: Did you read the debian paste thing from yesterday about profile nameing? Mar 30 08:01:09 hm, no Mar 30 08:19:56 ynezz: please read https://paste.debian.net/hidden/161e259e/ Mar 30 08:20:13 (I need to figure out how enable desktop notifiaction with my weechat client) Mar 30 11:09:21 build #276 of ramips/rt305x is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/ramips%2Frt305x/builds/276 Mar 30 12:17:52 dumb question probably: how is the nand adress on (nand read 0xXXXXXX) uboot determined? How can I obtain it? Mar 30 13:27:12 that spam mail on devel made my inbox blow up with ~450 bounces so far ;) Mar 30 13:53:36 blogic: same here, luckily its over now Mar 30 14:02:18 * ldir feels left out Mar 30 14:15:52 ldir: I can send you few spare bounces from the 2007, it was good bounce year Mar 30 14:16:17 or was it 2008? Mar 30 14:16:28 vintage bounces - tasty Mar 30 14:16:31 lol Mar 30 14:16:32 * jakllsch drizzles fabric softner in the chanel Mar 30 14:16:38 ldir: you must be feeling _very_ lonely :D Mar 30 14:17:58 I'm a horrible combination of mindlessly bored and a bit frustrated. Mar 30 14:18:35 well, I just learned how one cleans a ship's cargo hold Mar 30 14:19:09 oh that's an every day useful skill to have Mar 30 14:19:23 ldir: https://www.youtube.com/watch?v=is4cqxLM-N4 :) Mar 30 14:19:23 ldir: alias make='make 2>&1 | grep -v WARNING.*dependency' Mar 30 14:19:28 :) Mar 30 14:19:51 it's a surprisingly well done vlog Mar 30 14:20:54 ynezz: why is everything simple when you do it? Mar 30 14:22:36 * ldir reaches across to macbot - cold - hmm, folding@home is stuck for work units again Mar 30 14:22:55 helping covid crunching, i see Mar 30 14:25:44 ldir: if your macbot is bored, maybe you could give it the mvebu/cortexa53 test build? Mar 30 14:26:00 actually I preference it at 'Huntingtons' simulations as Mrs ldir's father (& his 4? brothers) died of it. Mrs ldir has a 50/50 chance of inheriting the buglet. But if it runs out of those units am more than willing for it to tackle the virus Mar 30 14:26:52 ouch, sorry to read about that Mar 30 14:27:01 ^ Mar 30 14:28:05 she may not have it. or she may have it but symptoms don't start till much later life...when we're all a bit mental anyway :-) Mar 30 14:29:03 ynezz: is that 'target build mvebu cortexa53' in gitlab ci.yaml speak? Mar 30 14:29:18 hi, is there any way to use the uboot command nand read for the whole chip? I can't seem to figure out the total size Mar 30 14:31:03 ldir: yep Mar 30 14:38:08 ldir: 'mvebe coretexa53' wont build it :) Mar 30 14:42:21 never trust a human to do a computer's job - can't even edit a file Mar 30 15:11:34 ldir: yeah, that's why I'm keeping those other targets I don't want to build test commented out, so just placing/removing the dot in front of the build task Mar 30 15:48:03 ynezz: ah that's helpful - I didn't know '.' was a comment, so I've now just copied your commit :-) Mar 30 15:50:50 its still crappy manual work, but GitLab still cant read my mind - anyway still huge timesaver Mar 30 17:09:00 trying to spin this aerohive ap170 up with something close, but I'm running into boot issues. built a wrt400n image (with a serial port tweak back down to the bootloader's 9600 baud) https://pastebin.com/ibBpue9Q anybody got some pointers on getting this to see the squashfs root? Mar 30 17:21:12 Is there a better place to ask? Mar 30 17:25:57 LoRez: this place is ok Mar 30 17:26:39 LoRez: I see you're pretty close, but are you really trying ar71xx there? Mar 30 17:28:35 LoRez: for the reference, current best practice to partition devices like that is to have just a single firmware partition holding kernel, then DTB appended, then squashfs rootfs. And let the splitters do their magic. While you have a separate kernel partition there. Mar 30 17:29:26 LoRez: if you want to boot without flashing anything then you need to be using an "initramfs" image, the way you're doing it (tftp kernel, tftp squashfs rootfs and then tell the kernel that the squashfs image is an initramfs image) just can't work at all anyhow. Mar 30 17:30:17 LoRez: so my advice: 1. do not waste time on ar71xx; 2. use initramfs image for testing everything; 3. use combined firmware partition. Mar 30 17:32:38 as I said over in #openwrt a bit back, that's the only way I seem to get the proper images. I can switch to ath79, but I don't seem to be getting useful images. my first-ish concern is backing up the existing flash, so yeah, I want to boot without flashing to start. Mar 30 17:33:04 I need to dig out where the MAC address is and probably the ART table for instance. Mar 30 17:34:32 the partitions that are showing up in the log are those that the wrt400n has built into its image. Mar 30 17:35:13 LoRez: art is always at the very end and have a fixed size. Mar 30 17:36:08 good to know Mar 30 17:36:12 LoRez: I'd guess nvram, factory and language can all go and be part of firmware. Mar 30 17:36:18 caldata is art Mar 30 17:37:01 that partition table is for the wrt400. this thing has 64M of flash and that partition table is just for 16M Mar 30 17:37:02 LoRez: ar71xx would be wasting time, I suggest to start making a suitable DTS for ath79. Mar 30 17:37:34 LoRez: can't be, you booted and m25p128 (16384 Kbytes) Mar 30 17:38:11 Probably there's additional NAND flash attached, but it doesn't change the m25p128 size. Mar 30 17:39:40 there are 4 of them. Mar 30 17:41:28 https://pastebin.com/ZD69dxuq Mar 30 17:44:11 LoRez: do you see them all on the board? Mar 30 17:44:40 yep! One's on the back side, but all 4 are there. Mar 30 17:44:43 That would be really baroque! Mar 30 17:45:23 I wonder if they're all sharing the same SPI bus. Then you need to figure out their CS pins. And hm, how are they supposed to be used?! Mar 30 17:45:31 I have no idea why they did that. I guess there's not a 64M SPI flash that was usable? It makes me wonder if they're interleaved somehow weird. Mar 30 17:45:50 yeah. One of the reasons I want to pull an image from it :) Mar 30 17:46:28 LoRez: when you boot it with vendor firmware, are you able to mount debugfs and see /sys/kernel/debug/gpio contents? Mar 30 17:47:16 It dumps me into a janky aerohive specific firmware prompt. Mar 30 17:48:03 LoRez: you can probably figure out the needed info regarding the flash from uboot sources then. Mar 30 17:51:35 the only gpio that might be useful on this one is the reset button and the CS lines for the flash. There are two leds inside, but it's an outdoor unit, so it's encased completely in metal with just the antenna and ethernet ports cut through. Mar 30 17:52:45 LoRez: flash IC's chipselect pins are controlled by SoC's GPIOs. Mar 30 17:52:53 Probably. Mar 30 22:51:54 ldir: jansson should have been converted to cmake before importing ;) Mar 30 23:00:19 5.6.0 broke iwlwifi Mar 30 23:01:08 https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net.git/commit/?id=be8c827f50a0bcd56361b31ada11dc0a3c2fd240 Mar 30 23:01:19 yeah Mar 30 23:01:49 still. Mar 30 23:01:57 people are already trying 5.6 with openwrt? Mar 30 23:01:58 paperbag time ;-) Mar 30 23:02:26 i noticed on my laptop Mar 30 23:05:12 anyone here familiar with porting python packages to OpenWrt when they require C compiling? Mar 30 23:05:22 russell--: oh I'm running boring debian with 5.4 Mar 30 23:14:53 haw iwlwifi been ported to openwrt yet? Mar 30 23:20:27 mangix: it's selectable in make menuconfig Mar 30 23:21:27 oh the firmware Mar 30 23:21:34 got it confused with iwd Mar 31 00:14:34 is there a shortcut for just rebuilding the kernel? Mar 31 00:33:34 LoRez: make package/kernel/compile I think Mar 31 00:36:27 that apparently doesn't work here. Mar 31 00:37:05 I have many feelings I'm going to bodge this dts file repeatedly. Mar 31 01:34:57 alright. I've got a ath79 image booted. Is there a way to (preferably with something already onboard) toggle the gpio pins individually, so I can find the ones hooked up to the flash? **** ENDING LOGGING AT Tue Mar 31 02:59:57 2020