**** BEGIN LOGGING AT Mon Aug 14 03:00:02 2017 Aug 14 03:07:50 no luck switching the dtb files Aug 14 03:08:10 anyone else have an idea? getting "No INS backends" when running ardupilot. Aug 14 03:17:02 braincore: dtb is a compiled device tree file, it basically describes the hardware to linux Aug 14 03:17:30 u-boot loads it, makes some last-minute adjustments if needed, and passes the result to the kernel Aug 14 03:18:30 MrCurious: having to specify that manually is still really weird since it should be the default Aug 14 03:20:31 MrCurious: maybe check the eeprom? head -c 256 /sys/bus/nvmem/devices/0-00500/nvmem | hexdump -C Aug 14 03:20:31 that is what i have heard. i can only guess i messed up the os by following EDUmip instructions *shrug* Aug 14 03:21:07 its all solved now, so no sense jousting a solved issue (and it is all tucked away for the day) Aug 14 03:22:02 big accomplishment of the day was using servo output + 7002 FET to make a DC motor driver for neato-lidar, Aug 14 03:23:18 well if it really makes a difference then it suggests your board eeprom might be misprogrammed (that's the only reason I can think of that could get the wrong dtb be loaded by default), which is something jkridner might want to know about Aug 14 03:23:28 i have decided the BB blue is a god-send for hobby robotics Aug 14 03:24:01 when i set up board #2 and #3, i will keep an eye out for the issue. if they manifest it, i will send up the alarm Aug 14 03:25:10 would love if there was a hint on how to use 4 - 8 PING sensors with BB blue Aug 14 03:25:11 apparently for the blue, bytes 4-15 of the eeprom should be "A335BNLTBLA2" Aug 14 03:25:47 the A335BNLT will no doubt be okay otherwise it wouldn't boot at all, but the "BLA2" might be borked Aug 14 03:26:16 ok, now i am intrigued enough to set it up and look... Aug 14 03:27:04 the list of eeprom identifiers can be found here: https://github.com/beagleboard/image-builder#eeprom-database Aug 14 03:29:16 it contains string "A335BNLTBLA2" Aug 14 03:29:53 so most likely i botched the config up with EDUmip Aug 14 03:30:28 ok Aug 14 03:31:22 then commenting the dtb= line out should have zero effect (just make sure there isn't accidently another one) Aug 14 03:31:27 it WAS a fun 1 1/2 evening bug hunt though :) Aug 14 03:41:09 zmatt: thanks for the explanation Aug 14 04:04:58 anyone else try to use arducopter with the BB blue? Aug 14 04:19:55 is the IMU supposed to be exposed via i2c? Aug 14 04:20:27 maybe check what's in /sys/bus/i2c/devices ? Aug 14 04:20:54 yeah it looks like it's on i2c-2 Aug 14 04:21:54 probably then shows up as an iio device, check /sys/bus/iio/devices Aug 14 09:27:23 hi Aug 14 10:40:55 hello Aug 14 10:40:59 Hello Aug 14 10:42:26 i have a question to extend the beagle bone black rev c, memory beyond 3.5GB, is it possible to have application beyond this size? Aug 14 10:43:17 I want to do this with the sd card Aug 14 10:44:11 desjardin: application beyond this size? Aug 14 10:44:32 desjardin: you mean, a filesystem thats bigger than the emmc? Aug 14 10:48:48 yes, Aug 14 10:49:15 desjardin: by having it on the sd card, yes. Aug 14 10:49:23 you can just plug in a microsd with any size you want Aug 14 10:49:41 tbr: i want one with EXACTLY 4.58 bytes!!!! Aug 14 10:50:07 * tbr chops a byte into bits for LetoThe2nd Aug 14 10:50:12 thx! Aug 14 10:50:25 ok, but areyou using the fdisk tool to make it? Aug 14 10:50:27 bytes, bits, and milli-bits Aug 14 10:50:56 * ogra grabs the leftover bits to have them with some maple syrup Aug 14 10:51:33 desjardin: depends if you want to also boot from the sd card, or if its just additional memory Aug 14 10:51:36 desjardin: after flashing an image onto sd card, you can expand the filesystem by deleting the partition and recreating it _with the same start sector!_ and then using "resize2fs" to expand the filesystem Aug 14 10:52:29 or yu use parted/gparted with the SD plugged into your PC Aug 14 10:52:30 ok, with fdisk then resizefs, i know this procedure but i got some problems, i have to retry it Aug 14 10:52:31 ;) Aug 14 10:52:52 or indeed there are higher-level tools also Aug 14 10:53:08 which tool is higher? Aug 14 10:53:31 (g)parted like ogra mentioned Aug 14 10:54:14 parted does everythig in one go ... (no need to separately resize partition and fs) with the "resizepart" command Aug 14 10:54:36 (and gparted is just the gui for that) Aug 14 10:55:19 another question is: if i flash again an sd card with an image, should i replace the existing one located on the internal mmc, by pressing the F2 button Aug 14 10:56:31 S2, instead of F3 excuse me Aug 14 10:56:36 S2, instead of F2 excuse me Aug 14 10:57:09 do you see my question? Aug 14 10:57:17 the S2 button is very rarely needed Aug 14 10:57:52 the only reason is when using an sd card instead of eMMC while an old version of the bootloader is installed on eMMC Aug 14 10:58:16 (this can be solved e.g. by wiping eMMC, or at least wiping the bootloader there) Aug 14 10:58:57 S2 forces the bootloader on eMMC to be ignored Aug 14 10:59:26 ok so i suppose i just need to put the new formated sd card in the beagle bone and use the fdisk, then i should see a new dev/mmcblkx appear? Aug 14 10:59:49 ? Aug 14 10:59:55 I don't understand your question Aug 14 11:00:43 inserting an sd card always results in a new /dev/mmcblk device, regardless of whether it's formatted Aug 14 11:01:07 (if the kernel is sufficiently recent it is always /dev/mmcblk0 ) Aug 14 11:02:34 when you use df -h, you see different kinds of mmcblk appear, for example, in my case when i do not have any sd card plugged on theBB, i only see /dev/mmcblk1p1 Aug 14 11:03:15 yes, /dev/mmcblk1 is the eMMC (when using an uptodate kernel at least, on older kernels the numbering can vary) Aug 14 11:03:35 and mmcblk1p1 is the first (and usually only) partition thereof Aug 14 11:04:02 (note that df -h only shows mounted filesystems) Aug 14 11:04:42 If if insert a new sd card with the formated image, with the df -h command, i only see a mmcblk1p1, with the sd card plugged on BB Aug 14 11:06:11 there's no reason for it to get automatically mounted Aug 14 11:06:42 if you installed a bootable image onto the card, then on reboot it should automatically boot from sd instead of eMMC Aug 14 11:06:51 which command do you recommend me to see the new sd card? instead of df -h Aug 14 11:07:17 uhh, you can get a list of block devices with lsblk Aug 14 11:10:53 ok with lsblk i see my sd card, Aug 14 11:12:23 which command do you recommend to mount the sd card? Aug 14 11:16:13 ah you just want to mount the storage? earlier it sounded like you wanted more space for the whole filesystem Aug 14 11:17:11 yes i just want to mount it for the time being, sorry for confusion Aug 14 11:17:49 you can mount it with mount /dev/mmcblk0p1 /path/to/mountpoint or add a line to /etc/fstab to have it automatically mounted Aug 14 11:22:54 i i type the command mount /dev/mmcblk0p1 /home/debian/mmcblk0, i get "wrong fs type, bad option bad super block.." do you have any idea? Aug 14 11:23:16 is the partition formatted? Aug 14 11:24:35 if not, format it with mkfs.ext4 /dev/mmcblk0p1 Aug 14 11:28:59 ok it is under progress, thanks, i suppose after it is complete, i will see this sd card using df -h? if it is the case, this is good Aug 14 11:29:14 after mounting it yes Aug 14 11:33:43 sorry agin, but about mkfs.ext4 /dev/mmcblk0p1 is it long? Aug 14 11:33:55 ? Aug 14 11:35:14 wrigleys extra Aug 14 11:35:15 Sorry it this mkfs.ext4 is complete, when i mount i still have the same issue "wrong fs type, bad option bad super block." Aug 14 11:35:46 Sorry this mkfs.ext4 is now complete, when i mount i still have the same issue "wrong fs type, bad option bad super block." Aug 14 11:35:56 uhh Aug 14 11:36:07 that makes no sense Aug 14 11:36:45 is there any message in kernel log? ( dmesg | tail ) Aug 14 11:38:01 yes: "[ 675.597024] FAT-fs (mmcblk0): bogus number of reserved sectors" Aug 14 11:38:35 my kernel responds: " [ 118.627395] error 0 requesting status 0x8000900" Aug 14 11:39:02 are you trying to mount /dev/mmcblk0 instead of /dev/mmcblk0p1 ? Aug 14 11:40:01 no? "mount /dev/mmcblk0p1 /home/debian/mmcblk0" Aug 14 11:40:25 o.O Aug 14 11:41:32 then I think the logical conclusion is that your bbb is inhabited by ghosts Aug 14 11:42:12 is a ghost a technical thing? Aug 14 11:42:24 no I was just being silly Aug 14 11:42:58 ok good Aug 14 11:43:57 is it due to one reason that my sd card is 32 GB? Aug 14 11:44:10 there's no good reason why mount would fail, and the kernel log entries you showed make no sense. but I'm at work so I also don't have a lot of time to help investigating Aug 14 11:44:23 that should be fine Aug 14 11:55:09 thank you Aug 14 11:56:21 for you time anyway it was good Aug 14 12:02:39 Think part of the proble is it still either has or thinks it has a FAT filesystem: FAT-fs (mmcblk0): ... Aug 14 12:03:38 you could try nounting it with: mount -t vfat /dev/mmcblk0p1 /home/debian/mmcblk0 Aug 14 12:03:52 -t ext4 you mean Aug 14 12:04:09 assuming that you've mkdir /home/debian/mmcblk0 first Aug 14 12:04:40 zmatt: the error his kernel threw says FAT-Fs Aug 14 12:05:11 he formatted the partition with mkfs.ext4 Aug 14 12:05:13 but yes might want to try with -t ext4 Aug 14 12:05:45 Ah I came in late assumed the kernel error was from after the formating Aug 14 12:06:00 figured the format failed somehow Aug 14 12:07:34 how this command is peformed? mkfs -t ext4? Aug 14 12:07:59 -t is for the mount command Aug 14 12:08:10 it specifies the expected FS type Aug 14 12:09:49 ok i see i also did it like this: mount -t vfat ...? bu not mount -t /dev/.., is it the same than what you suggest Aug 14 12:10:26 desjardin: you need the whole thing Aug 14 12:10:54 mount -t Aug 14 12:11:16 thus: mount -t vfat /dev/mmcblk0p1 /home/debian/mmcblk0 Aug 14 12:11:27 or: mount -t ext4 /dev/mmcblk0p1 /home/debian/mmcblk0 Aug 14 12:11:50 in fact, i have reformatted with windows tool : Formater, then i restarted BB and i could mount like you suggested, and now i see the mmcblk0p1 with df -h Aug 14 12:12:05 thank you really Aug 14 12:14:24 do you know if i can install new debian paquet on this new sd card? Aug 14 12:17:34 desjardin: no, its more like a data partition. Aug 14 12:34:29 ok what do you suggest to solve this issue? Aug 14 12:35:59 use the sd card to boot from and leave the internal emmc alone Aug 14 12:36:50 or, rethink your approach anyways. its highly improbable that you actuall have more than 3gb of pure rootfs applicaiton data. Aug 14 12:38:23 do you know qt? i noticed than when i install qt paquet i reach easily the limit of 3.5 GB Aug 14 12:39:06 i know qt. but qt itself certainly doesn't bring 3gb of stuff too. Aug 14 12:39:43 mabye give your system a good check with ncdu to see where the space actually goes Aug 14 12:41:58 it is sure ! anyway do you think i can resize the sd card partition up to 30GB on BB?in order to have more applications? In other words, is the BB good for such partition size? Aug 14 12:42:52 desjardin: the problem is not the partition size. the problem is that just by mounting that partition, it is not automatically available everywhere in the rootfs. it is available in the very one directory that you mounted it to. Aug 14 12:44:30 so if you want the root filesystem to be 30GB, you have to put it on the sd card altogether Aug 14 12:46:17 ok, of course, i should boot on the sd card! Do you know how to boot on sd card automatically? is it the uenv.txt file to change? Aug 14 12:46:51 it should happen automatically Aug 14 12:47:04 AFAIK it should happen automatically, if you sd card is bootable Aug 14 12:50:19 What do You mean by "sd card bootable"? is there anything special to perform? Aug 14 12:52:08 special, like putting a proper image on it? Aug 14 12:52:12 I'll explain what i know onthis subject: for me i have to format the sd card with the linux image (debian BB jessie) and plug it into the BB then power it Aug 14 12:53:17 exactly, then, it should be bootable. Aug 14 12:53:48 on such a sd card, you can modify the partition to fill up the whole space Aug 14 12:53:56 ok this is a very simple procedure, thank a lot Aug 14 12:54:05 ok this is a very simple procedure, thanks a lot Aug 14 13:50:30 ok, thank you, i have booted on my sd card image, my sd card size is 32 GB, with the fdisk tool i tried to create a second partition, whose size is a few GB, but it lets me create a second partition with a maximum size of 3MB, i rebooted then tried using the resize2fs tool. i get this error "resize2fs: Bad magic number in super-block while trying to open /dev/mmcblk0p2 Couldn't find valid filesystem superblock." Aug 14 13:50:55 any idea? Aug 14 13:59:20 I remember resizing a partition is kinda a pain, you have to delete the old partition with fdisk then create a new one on top of it, THEN resize2fs, but maybe i'm mistaken Aug 14 14:05:02 yes this is exactly what i did Aug 14 14:07:30 unfortunately, it gives errors! however this should be possible on BB? Aug 14 14:11:21 desjardin: don't make a second partition, remove the first partition and recreate it with the same start sector (but length covering the whole card) Aug 14 14:13:10 ok but the first partition is the boot part of linux, in this case, it is not a problem for BB? Aug 14 14:16:34 the contents of the partition is never touched Aug 14 14:16:59 ok i will try right now Aug 14 14:17:07 if fdisk asks to remove the ext4 "signature" (that seems to be a new thing in later versions), say no Aug 14 14:17:17 here's an example: https://pastebin.com/raw/3hcrjACS Aug 14 14:18:15 i suppose i have to delete the second partition too? Aug 14 14:18:24 the one you created? yes Aug 14 14:20:24 be sure to use the original "Start" value when it asks for First sector Aug 14 14:20:45 (it's probably 8192, but check the original partition map to be sure) Aug 14 14:21:30 ok done! now i have only one partition of size 3.6GB but it does not let me create a second one on a total of 8GB on this sd card. Is it ok according to you? Aug 14 14:22:17 uhh, you should end with one partition that covers the whole 8GB Aug 14 14:24:09 something like https://pastebin.com/raw/mkia5Uav Aug 14 14:25:44 then run e2fsck -f /dev/mmcblk0p1 && resize2fs /dev/mmcblk0p1 Aug 14 14:27:55 ohh! if you're doing this while booted from it you'll need to reboot after changing the partition table Aug 14 14:28:01 and then refrain from using e2fsck Aug 14 14:28:59 ok thank you, i did this e2fsck command that you suggested, and i get "/dev/mmcblk0p1 is mounted" then "e2fsck: Cannot continue, aborting" Aug 14 14:29:46 correct, the e2fsck step should be skipped when the filesystem is mounted Aug 14 14:30:03 you'll also have gotten some error from fdisk that it couldn't get the kernel to reread the partition table Aug 14 14:30:09 hence the need to reboot Aug 14 14:30:18 and then you can use resize2fs Aug 14 14:31:26 ok, thank you, very useful recommandations, i am rebooting Aug 14 14:32:07 normally you should never try to mess with filesystems that are mounted! fortunately, expanding the filesystem like this is safe, ext4 supports "live resize" (but only to make it larger) :) Aug 14 14:42:00 ok i rebooted! Aug 14 14:58:23 Again, should i re-execute e2fsck -f /dev/mmcblk0p1 && resize2fs /dev/mmcblk0p1? Aug 14 14:59:24 again, e2fsck should _not_ be used on a mounted filesystem Aug 14 14:59:28 just the resize2fs Aug 14 15:30:54 again about resize2fs, after the boot, i get "resize2fs: Bad magic number in super-block while trying to open /dev/mmcblk0p2 Couldn't find valid filesystem superblock." have any ideal Aug 14 15:31:01 ? Aug 14 15:31:49 mount -o remount rw / or some such command, then it is RO, and you can operate on it. Aug 14 15:32:18 oops s/rw/ro/ Aug 14 15:36:19 desjardin: uhh, why is there still a second partition? Aug 14 15:36:35 MrCurious: ?? Aug 14 15:36:51 MrCurious: live-resize requires that it is mounted rw Aug 14 15:37:10 (offline resize requires that it is not mounted at all, neither rw nor ro) Aug 14 15:37:21 the issue is that p2 is the wrong partition and shouldn't even exist Aug 14 15:39:04 desjardin: can you show the output of fdisk -l /dev/mmcblk0 please? (use pastebin.com or similar site) Aug 14 15:41:10 this: Device Boot Start End Sectors Size Id Type /dev/mmcblk0p1 * 8192 6963199 6955008 3.3G 83 Linux /dev/mmcblk0p2 2048 8191 6144 3M 83 Linux Aug 14 15:41:54 can you see it? Aug 14 15:42:59 1. that isn't the full output 2. don't paste multiline output in irc, it just becomes a mess. that's why I told you to use pastebin.com (or similar site) Aug 14 15:45:02 i redo:sudo fdisk -l /dev/mmcblk0 Aug 14 15:45:16 Disk /dev/mmcblk0: 7.4 GiB, 7948206080 bytes, 15523840 sectors Aug 14 15:45:24 Units: sectors of 1 * 512 = 512 bytes Aug 14 15:45:35 Sector size (logical/physical): 512 bytes / 512 bytes Aug 14 15:45:42 I/O size (minimum/optimal): 512 bytes / 512 bytes Aug 14 15:45:48 Disklabel type: dos Aug 14 15:45:55 Disk identifier: 0xca52207f Aug 14 15:46:02 Device Boot Start End Sectors Size Id Type Aug 14 15:46:08 /dev/mmcblk0p1 * 8192 6963199 6955008 3.3G 83 Linux Aug 14 15:46:25 Partition table entries are not in disk order. Aug 14 15:46:39 Is it better at your side? to view it? Aug 14 15:48:01 sorry there was also this line:/dev/mmcblk0p2 2048 8191 6144 3M 83 Linux Aug 14 15:48:09 * ogra wonders if desjardin read zmatt's 2) above Aug 14 15:48:18 I'm done with you. come back when you've learned to read Aug 14 15:48:34 desjardin, use pastebin.com or paste.ubuntu.com (or whatever other paste service you like) Aug 14 15:48:53 this is just unreadable on IRC Aug 14 15:48:56 ok for pastebin.com but i need an account? Aug 14 15:50:24 * ogra knows you dont need one on paste.ubuntu.com ... just add something random in the "Poster" field ) Aug 14 15:55:09 I posted something under a name "gilles-desjardin" on http://paste.ubuntu.com/25312879/ Aug 14 15:59:15 do you see my post on paste.ubuntu.com? Aug 14 16:06:57 ok, it seems that i succeeded in getting a larger partition size on my sdcard, thanks a lot, you are great!! Aug 14 16:12:41 Hello, cant find on forums any information when the BB_X15-rev B1 was in production. To my surprise mouser sent me one today saying it came from factory now. Is this possible? ? Aug 14 17:47:14 Can anyone confirm that the bbx15 boards now rolling off the assemblyline is marked as rev c and not b1 ? Aug 14 17:48:10 that sounds weird Aug 14 17:48:37 jkridner: ^^^^ Aug 14 17:49:23 they are rev C (PG 2.x processors) Aug 14 17:49:30 I got one today after long time waiting and in etching on board it says b1. is this a misstake ? Aug 14 17:50:42 more or less, yes. the PCB didn't change, so it is accurate from a PCB perspective. Aug 14 17:51:34 So pcb still say rev b1 but new silicon ? I cant check detail now as i left it at work in the box and wrote retailer a email. Aug 14 17:51:58 Desperate_x15-b1: yup Aug 14 17:52:54 Ok then it is rev c after all. no other signs i an verify rev ? Aug 14 17:55:01 https://github.com/beagleboard/beagleboard-x15/blob/master/BeagleBoard-X15_Quick-Start-Guide.pdf Aug 14 17:56:44 Does it match the pictures in ^^^ ? Aug 14 17:57:13 I red the BOM for rev C and the identification to the processor didnt match on the last to letters. AM5728BABCXA according to BOM Aug 14 17:57:31 well, the ones that aren't A2 pictures. :( Aug 14 17:58:00 very few B1s were ever built. Aug 14 17:58:12 the EEPROM also identifies the board as rev C. Aug 14 18:00:03 Is the AM5728BABC## enough to know its the rev c? XA didnt match Aug 14 18:02:11 Neither my eyes nor my camera on my phone are good enough to read the letters. Aug 14 18:02:30 I can tell you the board is a rev C (no doubt), but can't say exactly how I need to fix the BOM. Aug 14 18:03:51 Ok, then i just relax and start using it. just wanted to make sure it isnt some return or rerfurbished old board. Thanx for your support. Aug 14 18:09:51 And now i found this thing. If anyone else have doubts http://www.ti.com/product/AM5728/datasheet/device-and-documentation-support#ext_CH09_Device_Doc_Support Aug 14 18:52:33 Hello again. Went back to the office and fetched the board. Marking on the cpu says AM5729BABCXEA AM5729? not 28 is this expected? Aug 14 21:28:36 Hi. Can I have both beaglelogic and the PRUDAQ capes installed at the same time and choose between functionality at runtime? Aug 14 22:09:13 DuncanT: theoretically, yes, but you'd have to make sure any userspace stuff was completely clean before doing the switch. Aug 14 22:09:29 because you can load DT overlays at run-time .. iirc. .. Aug 14 22:22:06 Hey there, I want to use a BME280 environmental sensor with my beaglebone black using I2C. Now I try to get the pinmux to work. I connected the sensor at P9.19 and P.9.20, but the device does not show up when running i2cdetect -y -r 0/1/2. What is the right way to debug such a problem? Do I have to create a device file or a new overlay? https://www.irccloud.com/pastebin/xAAxAZk4/ Aug 14 22:27:40 I am also not sure if I need external pull-up resistors for it to work, or if the bbb design has these internally. Aug 15 01:01:35 DuncanT: it would be a good idea to check their schematics though Aug 15 01:02:01 if both capes try to drive the same pin, sadness results Aug 15 01:05:29 Hi I am having trouble booting with circuits connected to the HDMI pins (reconfigured for gpio). I read that these pins are also boot pins. Even if the connected pins are grounded could this be an issue? Aug 15 01:05:40 yes Aug 15 01:06:05 lcd_data0-15 must be pulled in a particular direction during power-on Aug 15 01:06:54 Is there a way to do so without having to unplug and plug back in the 20 wires Im using? Aug 15 01:07:34 Or should I just get a male-female pinheader and plug them in after boot? Aug 15 01:08:03 pins 31, 41, 43, and 44 of P8 must be pulled up to 3v3 during power-on Aug 15 01:08:45 pins 32, 33, 35, 39, 40, 45, and 46 must be pulled low, the rest don't matter Aug 15 01:09:48 (there are on-board 100K resistors that normally do this) Aug 15 01:10:30 And theres no way in software to ignore them without having to change the physical circuit? Aug 15 01:10:55 also I appreciate the help. I've been troubleshooting all day and just found this irc. Aug 15 01:11:36 they're sampled by hardware and affect both initial configuration of the processor itself, and the behaviour of the boot ROM Aug 15 01:12:11 safest is to avoid driving these pins externally, i.e. use them only as outputs and not as inputs Aug 15 01:14:05 (and even then one has to be careful, since connecting them to other stuff may accidently pull against the very weak 100K resistors) Aug 15 01:20:58 Sorry I'm not great with circuits. They are connected directly to the output pins of a hardware counter. But the counter is not connected to power. Since pins 32, 33, 35, 39, 40, 45 and 46 must be pulled low can I leave those connected? Will the counter act as ground? Aug 15 01:21:19 And then I could just not use 31, 41, 43 and 44 Aug 15 01:21:55 that would work yes Aug 15 01:22:13 ... why are you connecting a hardware counter though? :) Aug 15 01:22:34 Awesome. I will try that. Aug 15 01:22:50 School project. We wanted to implement a hardware clock using a crystal oscillator Aug 15 01:23:52 heh ok Aug 15 02:48:32 I now also tried different pin combinations and enabled P9 17,18 and P9 24,26 for i2c. I still do not find anything at 0x77 or 0x76, where the BME sensor should be according to the docs Aug 15 02:49:24 I do get this error in dmesg when I issue the echo to load the driver: bbbw kernel: [ 1418.265550] bmp280: probe of 2-0077 failed with error -121 Aug 15 02:51:34 I connected my bitscope as logic analyzer, there is activity on the bus. Negotiations end with ERR 0x7f **** ENDING LOGGING AT Tue Aug 15 03:00:01 2017