<GenTooMan>
warning audio weak and biased to only one side.
frostsnow has quit [Ping timeout: 245 seconds]
frostsnow has joined #beagle
set_ has quit [Remote host closed the connection]
Nirav has joined #beagle
<Nirav>
Hi
<Nirav>
we need quote for below part urgently
<Nirav>
BEAGLEBONE BLK REV B
<Nirav>
Requirement Qty is 1000 nos.
<zmatt>
Nirav: this isn't a sales channel, it's a community chat
<zmatt>
Nirav: for orders contact a distributor
<zmatt>
Nirav: digikey and farnell both have plenty of beaglebone black in stock
<zmatt>
Nirav: btw, "rev B" is obsolete, it has been replaced by rev C a long long time ago
<zmatt>
(the only difference is larger eMMC)
<Nirav>
Kindly share us quote for Rev C
<Nirav>
Also send us data sheet as per below email ID
<Nirav>
nirav@aimtron.com
Nirav45 has joined #beagle
<Nirav45>
Kindly share us Quote for Rev C
<Nirav45>
Also send us Data sheet for Rev C on below Email ID
<Nirav45>
nirav@aimtron.com
Nirav has quit [Ping timeout: 250 seconds]
<Nirav45>
hi
<Nirav45>
I m waiting
CygniX- has joined #beagle
shoragan[m] has quit [Ping timeout: 268 seconds]
CygniX has quit [Read error: Connection reset by peer]
<zmatt>
Nirav45: I answered your question earlier: this is not a sales channel
<zmatt>
Nirav45: multiple distributors have plenty of BeagleBone Black (rev C) in stock, and others get more stock later this month
<zmatt>
if you want a quote, ask a distributor
thinkfat has joined #beagle
Nirav45 has quit [Quit: Client closed]
signal11 has quit [*.net *.split]
Patater has quit [*.net *.split]
robert__ has quit [*.net *.split]
ogra has quit [*.net *.split]
russell-- has quit [*.net *.split]
russell-- has joined #beagle
robert__ has joined #beagle
Patater has joined #beagle
signal11 has joined #beagle
ogra has joined #beagle
Shadyman has quit [Quit: Leaving.]
<zmatt>
hmm, the am64x sk evm is only $99 ... that's tempting, even though I have no idea what use I'd have for one
ikarso has joined #beagle
Guest34 has joined #beagle
<Guest34>
Hi, can you please provide the status and datasheet of below parts.
<Guest34>
BBONE-BLACK-4G REVISION CBEAGLEBONE BLACK REVISION C
<veremitz>
I think a ban is due ..
Guest34 has quit [Quit: Client closed]
xet7 has joined #beagle
<zmatt>
it doesn't seem to be the same person (totally different IP)
<zmatt>
though both IPs are in india...
thinkfat has quit [Quit: Client closed]
troth has joined #beagle
mvaittin has quit [Quit: node-irc says goodbye]
jkridner[m] has quit [Quit: node-irc says goodbye]
shoragan[m] has joined #beagle
troth has quit [Quit: Leaving]
mvaittin has joined #beagle
jkridner[m] has joined #beagle
troth has joined #beagle
troth_ has joined #beagle
troth_ has quit [Remote host closed the connection]
mvaittin has quit [Client Quit]
mvaittin has joined #beagle
mvaittin has quit [Quit: node-irc says goodbye]
mvaittin has joined #beagle
<veremitz>
sounds about right..
johanhenselmans has quit [Ping timeout: 272 seconds]
vagrantc has joined #beagle
johanhenselmans has joined #beagle
thinkfat has joined #beagle
thinkfat_ has joined #beagle
<thinkfat_>
there you go
thinkfat has quit [Quit: Client closed]
Stanto has joined #beagle
Stanto has quit [Quit: Client closed]
Stanto has joined #beagle
aswin has joined #beagle
<aswin>
Hi Folks, I have been working to convert the BeagleV project to Kicad6
<aswin>
Just finished the schematics and the Library part
<aswin>
If someone can go through the Kicad files and verify the schematics is complete as per the reference one, That will be really helpful for me to proceed with the project
<aswin>
Next step is to generate the netlist and create the footprint library for components
JGS_mining has joined #beagle
<JGS_mining>
Hi everyone. I have a beaglebone black v 2.5 out of a Bitmain ASIC. It has been infected with a firmware virus that prevents SD card flashing. There are no pins for UART plug and from what I have read, I can dump whats in the firmware by using that. Help?
<vagrantc>
is the broken firmware installed to eMMC ? there's a button you can press to change the boot order so it prefers microSD
<JGS_mining>
The BBB is produced for Bitmain and does not have the buttons. I am not familiar with many aspects of this. I'm a miner that recognized this as something similar to a Pi and so I've been out looking for SMEs on this thing. Comparing pictures of a standard BBB vs this one, it has spots for things that arent there. USB, the I/O pins, power and the
<JGS_mining>
button are the first things that jump out.
<JGS_mining>
This particular virus removes the libraries to be able to boot from the SD card.
<vagrantc>
can you get to a u-boot prompt on serial console?
<vagrantc>
you might be able to short the appropriate pads to get it to boot off of microSD
<vagrantc>
but since this is a custom board, you might need to ask the vendors who created it; we can only wildly speculate about it
<JGS_mining>
as far as the eMMC, I'm not sure what that is. As for a serial connection, I don't have the proper cable to connect to it.
<vagrantc>
debugging this sort of thing without a serial console is not wise
<JGS_mining>
is it possible to clean it completely of anything? A factory reset if you will?
<vagrantc>
like i said, this is a custom board, so ... ask the vendor?
<JGS_mining>
Theyre in China and aren't very forthcoming in sharing
<vagrantc>
well, if there are serial headers, get a serial console cable, plug it in, and see what you can see ...
<vagrantc>
eMMC on a beaglebone black is another storage media (somewhat like microSD, but soldered on for the BBB) ... and it is typically where the firmware is stored
<aswin>
JGS_mining: How do you know there is a virus which is blocking the device accessing the microSD?
<vagrantc>
but your custom board may not have eMMC, etc.
<JGS_mining>
Total shot in the dark here, EZSync FTDI chip USB to 3.3v TTL UART Serial Cable, Connector end, 1.5m, TTL-232R-3V3 Compatible, EZsync007 work?
<aswin>
BBB has an onboard flash. It looks like its booting from that. In that case the linux kernel running on that may have SD card disabled in the dtb.
<JGS_mining>
aswin similar, but that one looks like it has the pins soldered in. mine doesnt
<aswin>
JGS_mining: You can solder connecting wires and access that UART port.
<vagrantc>
though if you're unlucky, sounds like the virus, if you have it, installs a bootloader that disables the ability to interact
<aswin>
JGS_mining: What I recommend you is first get the serial access
<JGS_mining>
vagrantc thats exactly what it does. Everyone else is spending thousands of dollars on new control boards, used, that may or may not be infected. So I'm looking for a way to wipe it
<JGS_mining>
is there a how-to you can direct me to? I just ordered the cable and it'll be here friday.
<vagrantc>
if it has the pads for the button to switch to booting directly off of microSD, that would seem like the simplest way ... though you'll also need a serial console
<JGS_mining>
aswin root/root gets into the interface, is that going to be the same?
* vagrantc
marvels at the high security protocols :)
<aswin>
if that works
<aswin>
if they kept root/root hahaha
<JGS_mining>
Only in the past year or two has there been ANY effort at security. I bought these miners used and they came with AIDS.
<vagrantc>
or get it booting from microSD, if that's possible; then you can inspect or reinstall from a safe environment
<vagrantc>
you don't want to be booting into a compromised environment; you can't trust the output of anything in it
<vagrantc>
presuming it's possible to reinstall
<vagrantc>
e.g. bitmain has installable images, etc.
<aswin>
vagrantc: JGS_mining: has to make sure that you have the proper image to reinstall. Otherwise its just another BBB
<JGS_mining>
I have trusted images that I have flashed on other miners
<JGS_mining>
its the only thing that saved them from getting infected.
<JGS_mining>
So there isnt a pin to short like on a PC that dumps the cmos? I have to go serial into the board, jump the boot switch, and then try?
<aswin>
JGS_mining: I can see that the boot switch is not present on the board.
<aswin>
JGS_mining: You can solder two wires and short them together to start booting from the microSD card
<JGS_mining>
aswin correct. But a steady hand and a wire...
<aswin>
JGS_mining: All the 4 leds are present on the board
<JGS_mining>
aswin the switch pad is on the underside. I am unable to find which pins need to be shorted to change to the SD card. I'm guessing not too many people break the switch off and have this problem
<zmatt>
JGS_mining: they didn't "break the switch off", your board is most likely not a beaglebone but some custom derivative
<aswin>
zmatt: It is not populated. That's what I can see from the image he has shared.
<zmatt>
JGS_mining: yeah it definitely isn't
<zmatt>
this isn't a beaglebone, it's a custom board
<aswin>
zmatt: Yes! It is but it is a beaglebone derivative. Looks like it. I am not sure!
<zmatt>
that nanya part must be a combined NAND + ram chip? I'm trying to read its part number
<JGS_mining>
@zmatt absolutely its custom, but its based on a BBB. I'm starting from there and seeing what is the same and how, or if I can, to get these 9 boards back to usable instead of buying brand new control board assemblies from less than reputable sources.
<JGS_mining>
I have high res pics of this board. How can I send them?
<aswin>
JGS_mining: The switch will have 4 pads. take a continuity meter and see which of those pads are not connected together.
<zmatt>
aswin: there's no reason to assume the boot switch will work, and if it does you'd still need a bootable image to boot from (suitable for this board)
<zmatt>
since this thing boots from NAND instead of eMMC, boot works very differently from a normal beaglebone
<JGS_mining>
@zmatt I have images. When I tried to flash them is when I became aware my boards had been infected
<aswin>
zmatt: There are no definite answer for any of these questions. The best way is try and see where it leads.
<zmatt>
JGS_mining: periodically people show up who try to do something with these kinds of funky boards, generally speaking it would require a large investment of time, especially if you do not have the schematic
<zmatt>
how did you try to flash an image?
<aswin>
JGS_mining: If you have a beaglebone black you can try your image on that and attach to the interface board. See it works.
<zmatt>
aswin: an image for this thing should not be run on a beaglebone black or vice versa
<JGS_mining>
I started out by trying to use the GUI to update the firmware to 3rd party. Bitmain firmware now needs a signature to do that. I then followed the instructions provided by the 3rd party firmware vendor and copied the files onto a FAT32 SD card. Plugged it in, applied power. The red light flashes then the blue, and it did that every few seconds for
<JGS_mining>
like 15 minutes before I started doing some investigating.
<JGS_mining>
zmatt then I tried a second miner and it did the same thing
<JGS_mining>
I can only assume the rest of them are the same
<JGS_mining>
zmatt I read someone was trying just that, ordering a bunch of BBB to see if that worked a week ago. From what you said, I'm guessing that wont work?
<zmatt>
huh, that nanya chip is ddr3 memory (256 MB)... so what is it booting from?
<zmatt>
is there a chip under that big sticker?
<aswin>
zmatt: wondering flash chip is under that sticker?
<zmatt>
well that's the only place it could be hiding
<zmatt>
the chips on top are the cpu, ram, pmic, and ethernet phy
<aswin>
zmatt: If they use BeagleBoneBlack reference then the their board should have an opensource license ?
<zmatt>
lol
<zmatt>
JGS_mining: so this thing has a different RAM size and different type of flash (raw NAND instead of eMMC), which means it uses a custom-built bootloader
<zmatt>
attempting to boot a beaglebone SD card image on this thing will not work, and might even damage the hardware (if they rudely used a board id that claims it's a beaglebone black and the incompatible pin configuration happens to result in a drive conflict)
<zmatt>
you need a bootable SD card image specifically for this board
<JGS_mining>
zmatt i have it.
<zmatt>
do you?
<zmatt>
copying files to a FAT-formatted SD card does not yield a bootable SD card
<JGS_mining>
zmatt I have the original Bitmain image for this particular ASIC, as well as the 3rd party image I was TRYING to get on here
<JGS_mining>
the bitmain image is an iso
<JGS_mining>
its made for a bad firmware flash.
<JGS_mining>
zmatt if i can get that on there, then this mess is fixed.
<zmatt>
if the board were configured to prefer booting from SD over booting from NAND then merely having a bootable SD card inserted would suffice to boot from it
<zmatt>
I guess my strategy would be to hope P8.31-46 still connect to the sysboot pins, measure sysboot0-4 to confirm that the current boot order, and then use wires to override the boot order
<zmatt>
*to confirm the current boot order
<JGS_mining>
zmatt the virus removed the libraries that allow a boot from the SD card
<JGS_mining>
can you DM me please?
<zmatt>
bootrom can boot directly from SD card
<zmatt>
no, why?
<zmatt>
do you have a link to that iso ?
<JGS_mining>
To discuss paying you for your time, effort and dealing with someone who has no idea what you just said. I mine crypto currency. This is a business expense. If I cant get it fixed here, I'll have to buy new control boards from sketchy sources.
<JGS_mining>
this is a business I own.
<zmatt>
has the manufacturer gone belly-up?
<JGS_mining>
no
<JGS_mining>
let me find the link
<zmatt>
so why not ask them how to reflash it?
<zmatt>
this is a commercial product, the manufacturer is the logical first place for this kind of support
<JGS_mining>
Theyre Chinese, I don't speak it. And most of all, they sell the control boards
<JGS_mining>
Support in the mining community is nothing. Its so mass produced and mass bought, they dont care
<JGS_mining>
it wouldnt surprise me if they released this virus
<zmatt>
great product
<JGS_mining>
I agree
<aswin>
LOL
<zmatt>
oh actually the boot config can be read from the pull resistors
set_ has joined #beagle
<zmatt>
1000001000010011
<JGS_mining>
zmatt aswin I have a quarter ETH for ya'll if you get me mining again. I dont know if you're into crypto at all, if you arent, THAT is where MY expertise is and I can walk you through it.
<zmatt>
10011 is { nand, SD card, UART }
<JGS_mining>
zmatt I have ZERO idea what you just said
<zmatt>
JGS_mining: the behaviour of bootrom (which is baked into the processor) is configured using resistors
<JGS_mining>
that I understand
<zmatt>
at the topright corner of https://ibb.co/vLCJGPS you see a row of pads for 16 resistors (and one capacitor on each end, those are unimportant)
<zmatt>
and you'll see a similar row on the opposite side of the pcb
<JGS_mining>
like a raspberry pi, right?
<JGS_mining>
nevermind, I didnt say that, I see what youre talking about
<zmatt>
the 1000001000010011 is which resistors are placed on the bottom, from bottom to top
<zmatt>
the last five bits (10011) determine the boot device order
<zmatt>
which is 1. NAND 2. SD card 3. UART (xmodem)
<zmatt>
i.e. with this boot config it will not try to boot from SD card unless NAND appears to be blank
<JGS_mining>
is it possible to permanantly boot off an SD card without having to add a resistor? I don't have micromin experience or capability
<JGS_mining>
but I can scrape one off no problem
<zmatt>
on a beaglebone the pins being pulled up/down by those resistors also connect to one of the expansion headers
<zmatt>
specifically, the expansion header pins right next to those resistors
<zmatt>
I'm not sure what would be the safest config to use... I guess 10111 should be safe to use (as long as it's not plugged into the miner board)
<zmatt>
so that would just require one wire
<JGS_mining>
This just blew past my understanding.
<zmatt>
from P8.43 ("lcd_data2") to the 3.3V supply (P9.03 or P9.04)
<JGS_mining>
not plugged in...the L3_10 board? that's the only way I know how to get power to the control board. Not in the ASIC itself, thats no problem.
<zmatt>
(for safety it would be preferred to use a 1K resistor instead of a wire, if possible)
<zmatt>
power is supplied via P9: P9.01/02 is ground, P9.05/06 is 5V supply to the board
ikarso has quit [Quit: Connection closed for inactivity]
<zmatt>
the problem with having the board plugged in is that boot mode 10111 has spi boot as very first option, which means it will configure four pins to SPI mode and try to boot from that (before attempting to boot from SD card)... without knowing that's attached to those pins on the main board, there's no way to know if that's safe
<zmatt>
ehh, sorry
<zmatt>
it has spi boot as second option
<zmatt>
which means that if it turns out your SD card is, in fact, not directly bootable then it will try SPI boot
<zmatt>
ideally you'd use this override-wire only to boot an SD card image and either reflash or erase the on-board NAND
<zmatt>
and then you wouldn't need the wire anymore
<zmatt>
right, except of course you'd just want to use a jumper wire plugged in on the top side
<JGS_mining>
If I erase the nand, that means the virus is gone along with all boot instructions. Then its going to go into the hardwired nand, sd, uart. If I have the original bootable bitmain image, then it should load. Am I understanding that correctly?
<aswin>
zmatt: The board is attached to another board. So he may not have access to the connectors
<JGS_mining>
aswin THAT is a beautiful picture. I understand that! :D
<zmatt>
this override-wire/resistor would cause it to ignore NAND and unconditionally boot from SD card (more specifically, boot device list: 1. SD card 2. SPI flash 3. UART (xmodem) 4. USB netboot (rndis) )
<zmatt>
that means if your SD card is bootable, that wire will make it boot from SD card regardless of contents of NAND
<zmatt>
(or rather, bootrom wlil load u-boot from SD card... that happens next is up to the boot script of u-boot, which I have no way of knowing in your case. if the boot script prefers booting linux from NAND over booting linux from SD card then you're still kinda screwed and would probably need access to the serial console or a different custom-built u-boot)
<zmatt>
wiping NAND would also result in booting from SD card
<JGS_mining>
wiping nand is my priority because thats where the virus is.
<zmatt>
if the SD card is not bootable, then the presence of this override wire _might_ be hazardous if the control board is plugged into the L3 board
<zmatt>
right, but for that you'll need to boot from SD card
<aswin>
JGS_mining: If you are 100% sure that the ISO you have will boot the device then you can try pulling the lcd_data2 up.
<zmatt>
you were still looking up a link to the iso... I could tell you right away whether it looks bootable
<zmatt>
though right now I'm afk, back in 30-60 minutes or so
<JGS_mining>
aswin it was just a long blue light and theres a red light staying lit
<JGS_mining>
how long do I need to leave it?
<aswin>
JGS_mining: Did you check the serial port?
<JGS_mining>
the red n green lights started blinking in unison which the instructions say is a good flash. Checking it now
<JGS_mining>
i dont have the ability to check the serial port
aswin has left #beagle [#beagle]
<zmatt>
note that it won't boot from NAND as long as that resistor is there
<JGS_mining>
well. I suppose it was worth a try. when it booted it was blinking green, now its blinking red.
<JGS_mining>
I used the bitmain image with the resistor in place. it flashed red/green together.
<JGS_mining>
i took the sd and resistor off and it blinks red
<JGS_mining>
I put in the 3rd party image hoping it would load, its still blinking red and the router isnt getting asked for an ip
<zmatt>
yeah debugging something like this without serial debug cable will probably be neigh impossible
<zmatt>
there's no way for us to even guess what it's doing or why
<JGS_mining>
I have a usb/serial cable to interface the routers n switches. any way I can cut the connector off and use the wires? or is that a totally different cable
<zmatt>
no idea, I wouldn't risk it.. it might be proper rs232 instead of low-voltage ttl serial
<zmatt>
of course you have the slight problem that your beaglebone doesn't have the serial console header installed
<zmatt>
if you can find some header strip you might be able to get away with just bending the pins slightly to make them friction-fit in the beaglebone where the connector is supposed to be mounted... an advantage of the fact the header isn't mounted is that you can stick it into the backside where you can access it even while the board is plugged into the L3
<zmatt>
(you can see where the header is supposed to be in the middle-bottom of https://ibb.co/ZgCDLz5 adjacent to the P9 header. the square pad is pin 1 (ground))
<JGS_mining>
I do. this is painful. the cable isnt going to be here til monday.
<JGS_mining>
zmatt Do you have an ETH or BTC address?
<zmatt>
nope
<zmatt>
btw, to be clear: you've booted successfully from this sd card, so if this was a working recovery image, it should have recovered the board
<JGS_mining>
so hopes not gone till serial says I smoked something?
<JGS_mining>
and to get an eth or btc address, I personally use Guarda.
<zmatt>
more like, the most obvious reason for this to not work is just that the sd card image isn't a properly working recovery image
<JGS_mining>
lol which I wouldnt put past bitmain
<zmatt>
or you somehow failed to follow the undocumented proper procedure ;)
<JGS_mining>
lol I'll blame bitmain
<JGS_mining>
or you create a coinbase, kraken, binance (the list goes on of crypto exchanges, but those are the main ones I use.) From there you can cash out the eth to whatever currency is your local.
<zmatt>
here's payment: if you figure out how to fix it, document what you did for other people with this problem and get it out to the community of these boards so we don't get more people here with the same issue :P
<JGS_mining>
hahah, wouldnt you know, that was sort of my plan. I had a camera recording me soldering the resistor and it working then not working. I was going to get it out there and tell people if it un-f's your ASICs to send some ETH to yours and aswins addresses.
<JGS_mining>
looking at the forums this is out of hand
<zmatt>
and blaming the manufacturer is definitely the appropriate response... there's no excuse for reflashing to be this much hassle
JGS_mining has quit [Quit: Client closed]
<zmatt>
like, they could have included one fucking button to override the boot mode and provide a working reflasher
JGS_mining has joined #beagle
<JGS_mining>
LMAO but but what would entice people to buy the next gen machines if they can prolong the life??? *cough cough APPLE*
<zmatt>
isn't it all about the actual ASICs? like, this board is just a mere supervisor
<JGS_mining>
without this board, its an expensive door stop.
<zmatt>
also, given the prices you mentioned it's completely absurd they optimized these boards down to removing components that cost a few cents
<JGS_mining>
google image search query: large bitcoin mine
<JGS_mining>
thousands of large farms
<JGS_mining>
and thats for bitcoin. These particular miners mine the scrypt algorithm
<JGS_mining>
HUGE mines
<zmatt>
do people actually call a mining farm a "mine" ? lol
<JGS_mining>
LOL yeah. I'm a farmer miner.
<zmatt>
the whole term "mining" kinda sucks anyway, it gives people wrong ideas about what the process actually does
<JGS_mining>
Absolutely. when someone asks about it, thats the first thing I say, its a total misnomer. I'm selling my computational power to the network.
<JGS_mining>
and I get paid in that particular crypto currency
<zmatt>
and (at least with bitcoin, I assume also with other systems) the total amount of currency "mined" per week does not actually depend on how much mining is going on, which is where the analogy immediately breaks down
<JGS_mining>
nope. Its a finite amount over a given period of time. random numbers as an example...if the daily output is 100 coins, and there are 2 miners, they statistically get 50 each. when there are 10000 miners...significantly less.
<JGS_mining>
and with crypto prices climbing like a cat with its tail lit, more and more people are jumping in and the big mines dont care about paying $10k a pop when they make millions a month
<JGS_mining>
I'm small in comparison, but its totally worth my time. even this control board nonsense
* zmatt
-> zZ
<JGS_mining>
Thank you VERY much for helping me on this, and more importantly indulging my lack of knowledge. I totally understand the magnitude of patience that is needed to explain something you're an SME on to someone who just picked up the fact that a BBB exists today.
<JGS_mining>
I'm going to try to get a usable image from HiveOS, if not, there are some other places I can try that shouldnt take me back down to reinfection. I'll be back when I get a serial cable. Again, thank you! Have a great rest of your day/night