**** BEGIN LOGGING AT Tue Feb 22 02:59:56 2022 Feb 22 05:19:25 hello, I've been given what I believe to be a BeagleBoard rev D (if the sticker by "ASSY REV" on the silkscreen is what indicates that). I've been trying to get it to work by following the "getting started" page but haven't had much luck. I used the "Stretch for BeagleBoard via microSD card" image and burned it to an 8GB SD card with balenaEtcher. Feb 22 05:19:25 when I plug in the BeagleBoard to my computer (mac OS Catalina 10.15.5), it *seems* like it's booting. the USR0 LED flashes in the heartbeat pattern and USR1 looks like what I'd expect to see while it's loading the OS off the SD card. but I never get a network device enumerated on my computer. I tried using an FTDI USB-serial adapter to look at Feb 22 05:19:26 what's coming out of the serial port, and while I *am* getting data that looks like, well, a linux bootup, it's garbled. it looks like what you'd get with the wrong baud rate, but I have my software set exactly the way all the docs expect (115200, 8-N-1, no flow control). I've tried multiple serial terminal programs and they all give the same Feb 22 05:19:26 results. I've tried different baud rates all the way down to 9600 and they all give the same results. Feb 22 05:20:19 the one thing I haven't tried yet is just plugging something into the HDMI (well, DVI) out and see if it displays anything, but I don't know if it would? Feb 22 05:27:52 (I just tried with the Angstrom image from 2012-01-11 with the same results as well) Feb 22 08:01:36 hi Feb 22 08:03:21 I would like to ask more about BBB Feb 22 08:10:50 Add sram to bbb Feb 22 08:16:37 any recommendation? Feb 22 09:12:20 sram? why that ? Feb 22 11:48:12 JarrettB: beagleboard or beaglebone? Feb 22 11:49:56 JarrettB: also, why are you trying to flash an ancient stretch image? all current images are buster Feb 22 11:50:53 JarrettB: and angstrom is prehistoric and long dead Feb 22 11:51:34 JarrettB: is the ftdi serial adapter maybe one for 5V rather than for 3.3V ? Feb 22 12:30:15 Where can I get the cape manager patch? Feb 22 12:37:21 hi Feb 22 12:52:24 are there Debian images for bullseye or bookworm? Feb 22 12:57:03 Question about like linux I am running Ubuntu on A am335x but we were having power issues it seems that at a reboot Linux puts a certain value into the tps which sets vdd_mpu to a low voltage Feb 22 12:58:00 and at boot (after a reboot) this voltage is to low to run the RAM reliably Feb 22 12:58:18 But it is linux setting this value at reboot I am trying to figure out which process does that Feb 22 12:58:23 any ideas? Feb 22 12:58:30 or maybe it's the tps driver doing that Feb 22 13:36:15 zmatt: I'm trying to flash stretch (2018-10-07) because the buster images don't list the original beagleboard on their compatibility lists. https://beagleboard.org/beagleboard is the board I have, though slightly different from the pic I think because the pic is an earlier rev. https://beagleboard.org/latest-images is where I got the image, it's Feb 22 13:36:15 the third one in the "recommended debian images" section Feb 22 14:14:05 lucascastro: cape manager has been deprecated for quite a while, nowadays cape detection is done by u-boot, which applies the overlays to the DT before passing it to the kernel Feb 22 14:17:11 rah: there are testing images / snapshots: https://forum.beagleboard.org/t/debian-11-x-bullseye-monthly-snapshots/31280 Feb 22 14:29:54 hi, what's the usual max admissible heat for SMC resistors? think I fried a bunch when resoldering @350°C Feb 22 14:33:43 Duality: ehh, your description doesn't make quite sense... the ram isn't powered by vdd_mpu, it has its own supply, also the pmic voltages are reset by u-boot as early as possible in SPL to try to deal with erratum 1.0.24 (boot being unreliable if a warm reset happens while vdd_mpu or vdd_core is configured as OPP50 level in the pmic, since the awful tps65217 has no mechanism to ensure its voltages ... Feb 22 14:33:49 ...get reset at warm reset) Feb 22 14:35:55 Duality: if the voltage gets set too low for reliable operation once the kernel is running, that would be a major bug in either the kernel driver or the dts (or maybe the power management firmware if it's involved, but I think it only messes with the pmic on entering/exiting deep sleep) Feb 22 14:44:05 Duality: is this a BBB or a custom board? Feb 22 14:50:11 Duality: if custom, and you're using an AM335x that's not speedbinned for 1 GHz, beware when basing your dts on that of the BBB, since the BBB modifies the OPP tables (declared in the base am33xx.dtsi) to inform the kernel that 1 GHz operation is always supported (since all BBBs support 1 GHz but a very early batch had an AM335x whose efuse bits erroneously didn't indicate 1 GHz support) Feb 22 14:51:52 Duality: accidently using that modification on an am335x that doesn't support 1 GHz would cause the kernel to unintentionally overclock/overvolt the cortex-A8 Feb 22 14:52:32 zmatt: kinda wonder how 'speed' binned the parts are really... or just market segements... but OC'ing is not recommended... Feb 22 14:52:52 rcn-ee: me neither, but you probably don't want to find out in production ;) Feb 22 14:53:06 in the field, when it's 40F below, with a wind... ;) Feb 22 14:58:09 Duality: slight correction to what I said earlier: I think u-boot might actually lower the PLL frequencies first (super early) and then raise them again later once the pmic has been configured, not sure though, too lazy to dig through it to see what it actually does nowadays Feb 22 15:00:27 Duality: did you clone the Beaglebone's eeprom value? this is one place in u-boot: https://github.com/beagleboard/u-boot/blob/v2021.10-bbb.io-am335x/board/ti/am335x/board.c#L778 Feb 22 15:02:37 fixing the erratum of the pmic voltage not being reset at reset is one of various improvements the tps65218 has over the tps65217 ... (though the BBB would break that anyway with the huge capacitor on the reset line that completely absorbs the low-pulse the cpu generates on warm reset) Feb 22 15:02:47 :P Feb 22 15:04:53 prior to the new BBB revision that is Feb 22 16:43:40 anyone else having problems with the irccloud app? Feb 22 16:44:35 zmatt: fyi, beagleboard.org now has #beagle and #beagle-* as a registered project space (whatever the term for that is) Feb 22 17:27:40 the documentation seems to imply that I should be able to connect from the beagleboard to a monitor with an HDMI cable... is that the case, or is an HDMI -> DVI cable required? right now my monitor is seeing *something* but it's unhappy and says the mode is unsupported Feb 22 17:27:46 I just want to see if this damn thing is working lol Feb 22 17:28:09 am I gonna have to break out the oscilloscope's serial decoder, geez Feb 22 17:50:43 OKAY. I managed to find an HDMI -> DVI adapter and *it is booting and showing a desktop yay!* now why doesn't it work with just HDMI Feb 22 17:57:28 I guess that's just a monitor compatibility thing, oh well Feb 22 18:03:32 zmatt: what about the 0002-U-Boot-BeagleBone-Cape-Manager.patch that it's generated when git format-patch Feb 22 18:08:16 0001-am335x_evm-uEnv.txt-bootz-n-fixes.patch implement uname_boot Feb 22 18:09:04 the 0002-*.patch is not needed? Feb 22 18:21:41 alright it *does* work with HDMI, it just takes like 15 seconds before it switches into a mode that the monitor likes. so I think I'm good now. still wondering about what image is best to use with the plain old beagleboard, and I think you may be onto something wrt. the 3.3V/5V issue with my FTDI zmatt, but at least I know the thing is working so Feb 22 18:21:42 the serial is secondary Feb 22 18:24:41 zmatt: where are they? Feb 22 18:25:10 zmatt: that page only talks about them but doesn't say anything about where to get them Feb 22 18:26:02 s/page only/page/ Feb 22 18:26:47 seems like there are downloads a little further down the page rah Feb 22 18:26:56 https://forum.beagleboard.org/t/debian-11-x-bullseye-monthly-snapshots/31280#debian-11x-bullseye-minimal-snapshot-10 Feb 22 18:27:37 JarrettB: not here Feb 22 18:28:16 it's not showing up for you? Feb 22 18:28:27 hmm, apparantly that page needs javascript to.. display the page Feb 22 18:28:36 ahh lol Feb 22 18:28:51 the unfortunate reality of the modern web Feb 22 18:32:06 I found the image on that page earlier and tried it, it came up but the network interface showed the link as being down for no apparent reason so I moved on Feb 22 18:32:20 currently doing a dist-upgrade from buster Feb 22 18:32:50 I'm using an X15 Feb 22 18:33:09 presumable the image hasn't been tested on that board Feb 22 18:33:16 *presumably Feb 22 18:37:37 rah: what issue? please reply on that thread.. the v5.10.x kenrel is a little new, so still working out all the regressions. Feb 22 18:40:35 ti changed the network driver, cpsw-new now... so on am335x we were fighting issues.. on teh bbai, it seemed to work, wonder what broke on the x15 for you.. Feb 22 18:41:32 it's also the first time with systemd-networkd, wonder if it got confused with teh dual ethernet on the x15.. Feb 22 18:44:12 *oh,* apparently the serial port outputs 12Vpp cause it's compliant RS-232... I hope that didn't do anything to my FTDI Feb 22 18:44:44 JarrettB: which version do you have BeagleBoard Classic Bx or Cx revision? Feb 22 18:44:46 rcn-ee: to be honest, interacting with that mailing list in the past has drained me of all interest in doing so again Feb 22 18:45:02 https://forum.beagleboard.org/t/resend-ethernet-phy-id-u-boot-fix/2324/11 Feb 22 18:45:12 rcn-ee I think it's D. the "ASSY REV" sticker says D Feb 22 18:45:38 the one by the user/reset buttons Feb 22 18:45:58 rah: it's even worse now... ti rewrote the driver.. we have no fixup for phy... going forward.. Feb 22 18:46:52 JarrettB: ah so the last BeagleBoard Classic.. Feb 22 18:46:56 yeah Feb 22 18:47:32 rcn-ee: worse than why you just stopped communicating about it for no reason? Feb 22 18:47:50 s/than why/than when/ Feb 22 18:48:19 rah: sorry, i do have a day job... BeagleBoard is my free time.. Feb 22 18:49:32 rah the bullseye image: https://forum.beagleboard.org/t/debian-11-x-bullseye-monthly-snapshots/31280 uses u-boot v2021.10 which has that fix... which still doesn't work 100%.. Feb 22 18:50:27 JarrettB: this is the serial pinout on that old board... Look at teh Schematic of this adapter... https://www.digikey.com/en/products/detail/digi-key-electronics/bbc01/1972181 Feb 22 18:50:46 it was based on an old rs232 motherboard spec.. Feb 22 18:51:45 oh I actually have one of those pin-header-to-DB9 cables, but it's not much use to me cause I don't have a serial port on any of my computers Feb 22 18:52:21 I've only ever needed the FTDI to do serial before cause I've only used things that did low-voltage UART Feb 22 18:52:31 JarrettB: ah! i use the usb-ftdi with a real rs232 header on it... but be very very careful, the original Beagle lacks a lot of protection.. Feb 22 18:52:42 good to know! Feb 22 18:52:44 thanks! Feb 22 21:39:20 rah: note that the latest bbb revision allows the phy to be reset using a gpio, so what we really need is some code in u-boot that checks if the phy is in a weird state and resets it until it works properly Feb 22 21:59:57 rah: that forum thread does seem confusing... it does looks like mainline u-boot contains the fix, currently at least **** ENDING LOGGING AT Wed Feb 23 02:59:57 2022