**** BEGIN LOGGING AT Tue Mar 01 02:59:56 2022 Mar 01 03:58:45 on BeagleBone Black, with the new monthly Debian 10 console snapshot https://forum.beagleboard.org/t/debian-10-x-buster-monthly-snapshots/31203, /sys/devices/platform/ocp/56000000.sgx is present by default, but why does that disappear if its default 4.19 kernel is replaced with any kernel listed here https://forum.beagleboard.org/t/debian-10-x-11-x-kernel-updates/30928 that is not the 4.19 kernel? is it a device tree problem? I Mar 01 03:58:45 tried to find what source code is producing the "56000000" to follow it through step by step to investigate, but while I see it in some sources for the other kernel versions, I do not see it in the sources for 4.19, either here https://github.com/RobertCNelson/bb-kernel/tree/am33x-v4.19/patches/drivers/ti or here https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/arch/arm/boot/dts/am33xx.dtsi?h=v4.19.71. Is the Mar 01 03:58:46 presence of /sys/devices/platform/ocp/56000000.sgx a hard requirement for SGX530 to function, and if so, what causes it to appear? Mar 01 03:59:52 tacokoneko: the sgx driver is packaged separately, so if you install a different kernel be sure to also install the package for the kernel module Mar 01 04:01:07 also, beware that the sgx userspace libs needs to match the version of the sgx kernel driver, so your ability to mix and match things will be limited Mar 01 04:02:46 the sgx kernel driver package is named ti-sgx-ti335x-modules-VERSION where VERSION is the kernel version Mar 01 04:04:36 though actually, sgx should appear in sysfs regardless of whether the driver is present Mar 01 04:04:53 e.g. I have /sys/bus/platform/devices/56000000.sgx even though I don't have the driver Mar 01 04:05:31 the path may however not be /sys/devices/platform/ocp/56000000.sgx though, that may vary depending on kernel version Mar 01 04:05:44 (but that has no relevance to whether sgx works or not) Mar 01 04:13:47 tacokoneko: e.g. in 5.10-ti its symlink in /sys/bus/platform/devices/ is named 56000000.gpu instead of 56000000.sgx, and its canonical path (to which the symlink resolves) is /sys/devices/platform/ocp/5600fe00.target-module/56000000.gpu Mar 01 04:14:00 but that's just an implementation detail, it's not relevant Mar 01 04:14:15 in general, sysfs paths are not stable Mar 01 04:54:58 thanks that's a good explanation, I was confused because I didn't see "56000000.sgx on minor 1" from pvrsrvkm in kernel log with newer kernels but now I see you are right and it has changed to "56000000.gpu" Mar 01 06:19:01 with that confusion over the platform device resolved, I now have SGX530 fully working (kmscube) with the 1.17 driver and kernel 5.10 on RobertCNelson's new Debian 11 image :) this will allow me to fully update my workflow Mar 01 07:20:29 Hello Mar 01 07:21:54 i brought BeagleBone black AI Mar 01 07:23:47 i need technical specification how to interact with 2 core Mar 01 07:26:07 how can i dedicate one core for specific task Mar 01 08:00:18 Hi, Hours ago I was discussing about wrong IP assigned to one of my device sometimes after reboot. Connman log reads as follows: Mar 01 08:00:18 Nov 03 17:16:43 ddsdevice kernel: cpsw 4a100000.ethernet: Detected MACID = c8:df:84:d5:d4:f5 Mar 01 08:00:19 Nov 03 17:16:43 ddsdevice kernel: cpsw 4a100000.ethernet: initialized cpsw ale version 1.4 Mar 01 08:00:19 Nov 03 17:16:43 ddsdevice kernel: cpsw 4a100000.ethernet: ALE Table size 1024 Mar 01 08:00:20 Nov 03 17:16:43 ddsdevice kernel: cpsw 4a100000.ethernet: cpts: overflow check period 1250 (jiffies) Mar 01 08:00:20 Nov 03 17:16:43 ddsdevice kernel: cpuidle: enable-method property 'ti,am3352' found operations Mar 01 08:00:21 Nov 03 17:16:51 ddsdevice sh[968]: am335x_evm: cpsw_0_mac: [C8:DF:84:D5:D4:F5] Mar 01 08:00:21 Nov 03 17:16:51 ddsdevice sh[968]: am335x_evm: cpsw_1_mac: [C8:DF:84:D5:D4:F7] Mar 01 08:00:22 Nov 03 17:16:51 ddsdevice sh[968]: am335x_evm: cpsw_2_mac: [C8:DF:84:D5:D4:F6] Mar 01 08:00:22 Nov 03 17:16:51 ddsdevice sh[968]: am335x_evm: cpsw_3_mac: [C8:DF:84:D5:D4:F8] Mar 01 08:00:23 Nov 03 17:16:51 ddsdevice sh[968]: am335x_evm: cpsw_4_mac: [C8:DF:84:D5:D4:F9] Mar 01 08:00:23 Nov 03 17:16:51 ddsdevice sh[968]: am335x_evm: cpsw_5_mac: [C8:DF:84:D5:D4:FA] Mar 01 08:00:24 Nov 03 17:16:52 ddsdevice unknown: chkphy:eth0 good Mar 01 08:00:24 Nov 03 17:16:53 ddsdevice connmand[1162]: Connection Manager version 1.35 Mar 01 08:00:25 Nov 03 17:16:54 ddsdevice kernel: using random self ethernet address Mar 01 08:00:25 Nov 03 17:16:54 ddsdevice kernel: using random host ethernet address Mar 01 08:00:26 Nov 03 17:16:54 ddsdevice kernel: using random self ethernet address Mar 01 08:00:26 Nov 03 17:16:54 ddsdevice kernel: using random host ethernet address Mar 01 08:00:38 Nov 03 17:16:56 ddsdevice connmand[1162]: Ignoring interface usb0 (filtered) Mar 01 08:00:38 Nov 03 17:16:56 ddsdevice connmand[1162]: Adding interface usb0 [ gadget ] Mar 01 08:00:39 Nov 03 17:16:56 ddsdevice connmand[1162]: eth0 {update} flags 36867 Mar 01 08:00:39 Nov 03 17:16:56 ddsdevice connmand[1162]: eth0 {newlink} index 4 address C8:DF:84:D5:D4:F5 mtu 1500 Mar 01 08:00:40 Nov 03 17:16:56 ddsdevice connmand[1162]: eth0 {newlink} index 4 operstate 2 Mar 01 08:00:40 Nov 03 17:16:56 ddsdevice connmand[1162]: usb0 {newlink} index 5 address C8:DF:84:D5:D4:F7 mtu 1500 Mar 01 08:00:41 Nov 03 17:16:56 ddsdevice connmand[1162]: usb0 {newlink} index 5 operstate 2 Mar 01 08:00:41 Nov 03 17:16:56 ddsdevice connmand[1162]: Ignoring interface usb0 (filtered) Mar 01 08:00:42 Nov 03 17:16:56 ddsdevice connmand[1162]: usb1 {create} index 6 type 1 Mar 01 08:00:42 Nov 03 17:16:56 ddsdevice connmand[1162]: usb1 {update} flags 4098 Mar 01 08:00:43 Nov 03 17:16:56 ddsdevice connmand[1162]: usb1 {newlink} index 6 address 00:00:00:00:00:00 mtu 1500 Mar 01 08:00:43 Nov 03 17:16:56 ddsdevice connmand[1162]: usb1 {newlink} index 6 operstate 2 Mar 01 08:00:44 Nov 03 17:16:56 ddsdevice connmand[1162]: Ignoring interface usb1 (filtered) Mar 01 08:00:44 Nov 03 17:16:56 ddsdevice connmand[1162]: Adding interface usb1 [ gadget ] Mar 01 08:00:45 Nov 03 17:16:56 ddsdevice connmand[1162]: usb1 {newlink} index 6 address C8:DF:84:D5:D4:FA mtu 1500 Mar 01 08:00:45 Nov 03 17:16:56 ddsdevice connmand[1162]: usb1 {newlink} index 6 operstate 2 Mar 01 08:00:46 Nov 03 17:16:56 ddsdevice connmand[1162]: Ignoring interface usb1 (filtered) Mar 01 08:00:46 Nov 03 17:16:56 ddsdevice connmand[1162]: The name net.connman.vpn was not provided by any .service files Mar 01 08:00:58 Nov 03 17:16:58 ddsdevice connmand[1162]: usb1 {update} flags 69699 Mar 01 08:00:58 Nov 03 17:16:58 ddsdevice connmand[1162]: usb1 {newlink} index 6 address C8:DF:84:D5:D4:FA mtu 1500 Mar 01 08:00:59 Nov 03 17:16:58 ddsdevice connmand[1162]: usb1 {newlink} index 6 operstate 6 Mar 01 08:00:59 Nov 03 17:16:58 ddsdevice connmand[1162]: Ignoring interface usb1 (filtered) Mar 01 08:01:00 Nov 03 17:16:58 ddsdevice connmand[1162]: usb1 {add} route 192.168.6.0 gw 0.0.0.0 scope 253 Mar 01 08:01:00 Nov 03 17:16:58 ddsdevice connmand[1162]: usb1 {add} route fe80:: gw :: scope 0 Mar 01 08:01:01 Nov 03 17:16:58 ddsdevice connmand[1162]: usb1 {del} address 192.168.6.2/24 label usb1 Mar 01 08:01:01 Nov 03 17:16:58 ddsdevice connmand[1162]: usb1 {del} route 192.168.6.0 gw 0.0.0.0 scope 253 Mar 01 08:01:02 Nov 03 17:16:58 ddsdevice connmand[1162]: usb1 {add} address 192.168.6.2/30 label usb1 family 2 Mar 01 08:01:02 Nov 03 17:16:58 ddsdevice connmand[1162]: usb1 {add} route 192.168.6.0 gw 0.0.0.0 scope 253 Mar 01 08:01:03 Nov 03 17:17:44 ddsdevice connmand[1162]: eth0 {add} address 169.254.242.141/16 label eth0 family 2 Mar 01 08:01:03 Nov 03 17:17:44 ddsdevice connmand[1162]: eth0 {add} route 169.254.0.0 gw 0.0.0.0 scope 253 Mar 01 08:01:04 Nov 03 17:17:44 ddsdevice connmand[1162]: eth0 {add} route 0.0.0.0 gw 0.0.0.0 scope 253 Mar 01 09:32:17 Raajeshwar: do NOT EVER paste a giant blob of text like that into chat Mar 01 09:32:35 use a paste service like pastebin.com Mar 01 09:32:48 Sorry Mar 01 09:33:06 Is there a way to remove it? Mar 01 09:33:11 no Mar 01 09:34:05 I won't do it again. Sorry about that. Mar 01 09:38:23 I'm not very familiar with connman, I don't use it myself, but the 45 second delay before connman adds the link-local address (169.254.242.141) suggests to me it timed out trying to obtain an IP address via DHCP Mar 01 09:39:47 "chkphy:eth0 good" is that from that article you linked earlier? (about the phy issue) because I would not recommend keeping that installed, as that workaround is potentially hazardous for the hardware Mar 01 09:40:06 (and clearly not relevant for your problem anyway) Mar 01 09:40:19 Yes it is from that installed from that link. Mar 01 09:41:44 I get link-local address 1/7 times on rebooting. Don't know why DHCP couldn't obtain IP. Mar 01 09:48:24 what about when you disconnect ethernet and plug it back in? Mar 01 09:50:29 if you can reproduce the problem that way, you can try logging in via usb and using a packet sniffer to see what's happening when you plug in ethernet Mar 01 09:51:11 e.g. (after installing tcpdump with apt): sudo tcpdump -e -i eth0 -p -n -U '(outbound && udp dst port bootps) || (inbound && udp src port bootps)' Mar 01 09:52:11 optionally add the -v option to get a lot of details (spammy!) Mar 01 09:52:35 Ok Mar 01 09:53:46 actually, adding -v is highly recommended, without it doesn't really show much info Mar 01 09:54:24 Ok. I will try. Mar 01 10:11:44 the sequence you normally expect to see is: beaglebone broadcasts "Discover", dhcp server replies with "Offer", beaglebone broadcasts "Request", dhcp server responds with "ACK": https://pastebin.com/raw/xVY8tEp5 (lots of verbiage trimmed) Mar 01 10:12:17 Disconnecting and Reconnecting ethernet didn't help. systemctl restart networking didn't help. Now I am installing tcpdump to test what you have suggested. Mar 01 10:12:32 what do you mean by "didn't help" ? Mar 01 10:12:47 Means, didn't get proper IP Mar 01 10:12:48 the question was whether it reproduces the problem (i.e. if it behaves similar to reboot) Mar 01 10:13:32 if it reproduces the problem, that's the desirable outcome since it allows you to debug the issue Mar 01 10:13:54 (since you can't really run a packet sniffer during boot obviously) Mar 01 10:14:38 Yes. It reproduces the problem. Mar 01 10:15:26 also the "networking" service (ifupdown) is irrelevant (and honestly shouldn't even be installed) since you're using connman Mar 01 10:16:07 Oh. Ok. Mar 01 10:16:22 and good, then try to see if you can use tcpdump to capture what's going on after you plug in ethernet Mar 01 10:16:52 Yes. I will get back after checking that. Mar 01 10:17:17 (i.e. unplug ethernet, start tcpdump (via usb), plug ethernet in, see what output you get) Mar 01 15:09:55 Anybody experience with hooking 5GHz USB-Wifi dongles to a beaglebone? I needed the extra channels as lots of Beaglebones are at the same location. Mar 01 15:10:38 And they all need to function as separate accesspoints. Mar 01 16:45:19 what? why all as access points/ Mar 02 00:17:35 hey does anyone here know how to get rc_test_mpu to run without an i2c error on the beaglebone blue? Mar 02 00:18:54 what sort of i2c error? also, what kernel version are you running? Mar 02 00:20:09 ERROR: in rc_i2c_write_byte, system write returned -1, expected 2 Mar 02 00:20:09 ERROR: resetting MPU, I@C write to reset bit failed Mar 02 00:20:10 ERROR: failed to reset mpu9250 Mar 02 00:20:17 as for kernel version, give me one sec Mar 02 00:21:04 its Linux 4.14.108-ti-r131 armv71 Mar 02 00:21:48 ah, you're using a fairly old image? have you tried reflashing to the latest official image? ( https://beagleboard.org/latest-images ) Mar 02 00:22:29 ... did i flash the nand storage on the beaglebone wrong? to my knowledge i was running latest image Mar 02 00:22:56 what does "cat /etc/dogtag" say? Mar 02 00:23:21 Debian Stretch imgtec Image 2020-04-06 Mar 02 00:23:38 oh that's a really weird image Mar 02 00:24:28 that's an unusual custom image built for the imgtec sdk Mar 02 00:24:55 ok yeah thats the image i flashed then Mar 02 00:25:15 yeah don't do that :D get the AM3358 Debian 10.3 2020-04-06 IoT image Mar 02 00:25:44 either the SD image if you want to boot from SD, or the eMMC Flasher image (under the "Flasher" Debian images) section if you want to reflash eMMC Mar 02 00:26:05 (reflashing eMMC is highly recommended over booting from sd, unless you explicitly have stuff on eMMC you need to preserve) Mar 02 00:26:55 ok, im grabbing the image now Mar 02 00:27:15 does that image come with librobotcontrol or will i have to reinstall it? Mar 02 00:27:42 it comes with it Mar 02 00:27:53 oh thats a relief Mar 02 00:28:43 ok its done downloading, lemme flash my eMMC and ill come back with results on if its still being a jerk or not Mar 02 00:29:05 that downloads page could really use some kind of redesign... I think it's really not clear enough what you're supposed to download, especially for new users Mar 02 00:30:37 case in point: im a new user Mar 02 00:31:09 but it does help that they have an irc like this Mar 02 00:35:18 yeah, it can sometimes take a bit of time to get a response (hence the "please be patient" in topic), but often enough there are knowledgeable people here Mar 02 00:40:05 ok, currently flashing the eMMC after getting it onto the sd card Mar 02 01:00:06 was there anything i was supposed to change before flashing the eMMC? cause its no longer letting me remote into the board at 192.168.7.2 Mar 02 01:00:25 im gonna try reflashing jic, but let me know if i messed something up Mar 02 01:11:51 ok it was something i did apparently, it is working now zmatt, however the i2c bug is still there Mar 02 02:33:49 Hello, has anyone noticed all the new "advertisements" under git and its controls? Mar 02 02:34:24 The other day, I wanted to pull from git, git decided to just blurt out tons of extra info. not related to my pull. Mar 02 02:35:01 On the BBB, I keep getting awkward colored git output. Mar 02 02:35:09 Has anyone else noticed this on their boards? **** ENDING LOGGING AT Wed Mar 02 02:59:56 2022