**** BEGIN LOGGING AT Sat Mar 27 02:59:58 2021 Mar 27 08:55:43 Hello Guys, Greeting! I have one issue using BBB with Ethernet. Sometimes the Ethernet port on a Beagle Bone Black does not work on power up. It takes either a physical reset button press or a power cycle to fix it. Mar 27 09:06:50 Can anyone please help? Mar 27 09:58:28 I thought that was fixed, but seems its unreliable still.. what kernel version/ image are you using? Mar 27 10:01:32 4.14 Mar 27 10:02:25 We still experience this issue Mar 27 10:03:49 Linux ddsdevice 4.14.71-ti-r80 #1 SMP PREEMPT Fri Oct 5 23:50:11 UTC 2018 armv7l GNU/Linux Mar 27 10:03:58 Description: Debian GNU/Linux 9.5 (stretch) Mar 27 10:06:02 Hello Guys, Greeting! I have one issue using BBB with Ethernet. Sometimes the Ethernet port on a Beagle Bone Black does not work on power up. It takes either a physical reset button press or a power cycle to fix it. Mar 27 10:06:03 Can anyone please help? Mar 27 10:06:31 echo ?! Mar 27 10:06:55 that's quite an old kernel .. is that the TI kernel repo? Mar 27 10:07:20 I lost connectivity, so thought to post it again. Sorry. Mar 27 10:08:46 It cameup with image by default. I just downloaded the image from beaglebone website and flashed it. Mar 27 10:09:30 Can you please tel me in which version this issue was handled? Mar 27 10:11:59 We are using the image bone-debian-9.5-iot-armhf-2018-10-07-4gb.img.xz. Mar 27 10:14:51 I just changed the device name, thats what you see the name ddsdevice Mar 27 11:16:09 Guys, Any update for me? Mar 27 13:05:22 2018@? whut Mar 27 13:05:26 zmatt: halp? Mar 27 13:05:46 I wouldn't use an IoT image personally .. what's your target application? Mar 27 13:22:50 I would offer a lil more assistance, but its been a while since I used my BBB's .. even if one functions as my DHCP server for my LAN (!) Mar 27 16:09:10 Hello Guys, Greeting! I have one issue using BBB with Ethernet. Sometimes the Ethernet port on a Beagle Bone Black does not work on power up. It takes either a physical reset button press or a power cycle to fix it. Mar 27 16:09:11 Can anyone please help? Mar 27 16:09:11 I am using the kernel/OS version printed below. Mar 27 16:09:12 4.14.71-ti-r80 #1 SMP PREEMPT Fri Oct 5 23:50:11 UTC 2018 armv7l GNU/Linux Mar 27 16:09:12 Debian GNU/Linux 9.5 (stretch) Mar 27 16:31:00 repeatedly asking the same questino is merely likely to get you ignored, fyi Mar 27 17:07:40 hmmm, irccloud doesn't make it easy to find mentions. Mar 27 18:13:55 kveremitz, I did not get proper reply. that's why asking that question repeatedly. If my channel has cut then who will you reply or how long you want to wait. Reply if you know otherwise it is open issue in the current version as well or say something. I asked this question many hours before, there were 166 people online but still no reply. What Mar 27 18:13:56 does it mean. Does it make sense? Mar 27 18:19:32 Thanks foe the support though. It seems it is open issue. Good bye. Mar 27 19:20:10 ¯\_(ツ)_/¯ Mar 27 19:20:14 no loss there then Mar 27 19:48:21 I think we should be more welcoming :) Mar 27 19:48:29 feel free Mar 27 20:32:52 I am super at a welcoming state? Mar 27 20:33:18 Bienvenido! Mar 27 20:33:20 WElcome! Mar 27 20:33:36 Gratzi! Mar 27 20:53:53 kveremitz: and no, the ethernet issue was never fixed, it is a hardware issue Mar 27 20:58:22 kveremitz: https://pastebin.com/z7a7Tn4C Mar 27 20:59:16 zmatt: thanks, my memory on TI matters is a tad rusty Mar 27 20:59:28 it's not a TI matter, it's the etherhet phy Mar 27 21:00:29 the phy just sucks Mar 27 21:01:47 its an external PHY yes? (I meant more broadly about BBB-related subjects) Mar 27 21:02:09 I'm marginally more au-fait with sunxi/rpi/imx hardware currently Mar 27 21:02:19 and rockchip.. all the ARMs Mar 27 21:02:57 it's a Microchip LAN8710A Mar 27 21:03:09 but I did point someone who was trying to write some bare-metal code to TFTP boot over here , since rpi is a bad platform for netbooting, and I wouldn't recommend sunxi personally either... Mar 27 21:03:14 oh M$ip Mar 27 21:03:22 yeah they ain't great Mar 27 21:03:43 I didn't like those hub chips on the older rpis either.. :D Mar 27 21:04:02 even if they were SMSC before Microchip Mar 27 21:04:14 this one was SMSC too Mar 27 21:04:24 yup sounds about right Mar 27 21:04:40 that said, their straight hub chips are OK Mar 27 21:04:58 the 250x's Mar 27 21:05:16 I have no experience with their stuff other than this phy, and this phy clearly sucks Mar 27 21:05:35 :) Mar 27 21:07:08 I've seen weird behaviour with a lan8710a on another device Mar 27 21:07:35 :/ Mar 27 21:07:38 if the smsc phy driver is enabled in the kernel, it often goes into a weird non-working state Mar 27 21:07:51 with that driver disabled, all is fine Mar 27 21:07:51 what a pita. Mar 27 21:07:58 mru: lol weird Mar 27 21:08:02 I've not seen that Mar 27 21:08:10 only with recent kernels Mar 27 21:08:27 not sure when it started Mar 27 21:08:31 I do know the kernel driver has a workaround for a phy bug related to not detecting link changes in powersave mode Mar 27 21:08:37 somewhere between 4.19 and 5.10 Mar 27 21:08:56 mru: sounds like something that should be bisected and reported? Mar 27 21:09:11 sure, send me more time Mar 27 21:09:20 hehe Mar 27 21:09:31 and it's an intermittent problem Mar 27 21:09:40 so it can take hours to determine if a particular version is good or bad Mar 27 21:09:40 intermittent problems suck Mar 27 21:11:26 jkridner: should the BBB not have an official errata doc (for the ethernet phy issue and the 3.3v regulator issues) so that people who base designs off of the BBB don't blindly copy its issues too? (like how the osd335x inherited the 3.3v issues) Mar 27 21:29:26 if they're blindly copying, are they going to read errata? Mar 27 21:29:49 no Mar 27 21:29:56 well if they don't they only have themselves to blame Mar 27 21:30:17 reading errata is definitely one of the first things I do with any chip I might have to work with Mar 27 21:30:47 not that i'm saying issues shouldn't be documented :) Mar 27 21:30:54 however if an issue is has been known for a long time yet is not clearly documented, I'd be pissed Mar 27 21:31:31 you must be pissed a lot Mar 27 21:31:43 or, well, more annoyed probably Mar 27 21:31:53 I'd don't actually get pissed that easily Mar 27 21:32:44 but I definitely respect companies more if they publish errata in an easy to find place (like TI who puts them prominently on the product page) Mar 27 21:33:25 for the BBB, having an errata doc in the same place as the schematic and other hardware design files seems appropriate **** ENDING LOGGING AT Sun Mar 28 02:59:56 2021