HP Forums

Full Version: Linux on Prime not working
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
So I tried following Zephray Wenting's instructions (http://github.com/zephray/prinux) for how to flash Linux to the Prime. On step 2, it said that after shorting the pins and pressing the reset button I should see a new USB device called "SE Blank 6ULL" on my PC, which I didn't. Unsurprisingly, the flashing scripts gave me an error too. Did I do anything wrong or is the tutorial outdated?
I haven't tried this in a really long time, but I used to apply 3.3v onto the TP907 pad and press the reset button to trigger the SoC recovery mode.

Personally I'd try connecting the calculator to a Linux machine and look into the kernel logs for details, but that's because I don't know how to check for USB enumeration or kernel events on Windows. Maybe you've got a defective or charge-only USB cable?
(08-01-2021 02:03 PM)jfelten Wrote: [ -> ]So I tried following Zephray Wenting's instructions (http://github.com/zephray/prinux) for how to flash Linux to the Prime. On step 2, it said that after shorting the pins and pressing the reset button I should see a new USB device called "SE Blank 6ULL" on my PC, which I didn't. Unsurprisingly, the flashing scripts gave me an error too. Did I do anything wrong or is the tutorial outdated?

That project is over two years out of date. In general, I'd consider any project more than 1.5 years old to be "abandoned." YMMV...
(08-01-2021 05:06 PM)Jean-Baptiste Boric Wrote: [ -> ]I haven't tried this in a really long time, but I used to apply 3.3v onto the TP907 pad and press the reset button to trigger the SoC recovery mode.

Personally I'd try connecting the calculator to a Linux machine and look into the kernel logs for details, but that's because I don't know how to check for USB enumeration or kernel events on Windows. Maybe you've got a defective or charge-only USB cable?

So you mean connecting a 3.3v power supply to the pads next to the top-right corner of the CPU? Also, I tried using a micro-B cable instead of micro-C, connected to a Linux computer, and I noticed that after the calculator reset, it appeared as HP Prime, not "SE Blank 6ULL".
(08-02-2021 07:19 PM)jfelten Wrote: [ -> ]So you mean connecting a 3.3v power supply to the pads next to the top-right corner of the CPU? Also, I tried using a micro-B cable instead of micro-C, connected to a Linux computer, and I noticed that after the calculator reset, it appeared as HP Prime, not "SE Blank 6ULL".

My notes do say TP907, which is next to the top-right corner of the SoC, but I cannot remember if that's actually correct (it could've been TP807 if I wrote it down wrong, but I doubt it).

There is theoretically another way, which would be to short two NAND data pins together to force the SoC's ROM into recovery mode (if it can't read successfully the NAND chip, it should fallback).

If all else fails, there is the option of hooking a JTAG probe to the back of the motherboard (https://tiplanet.org/forum/viewtopic.php...27#p238167), but it's glued to the back of the case with an adhesive film. It can be separated with a plastic spatula (or carefully with a knife) and some paper sheets to stop the motherboard from gluing itself back, but unless you actually need to access the pins on the back I wouldn't recommend doing that.
I have booted linux recently using this method and can confirm it still works. What I can say is that it is sometimes finnicky to get a proper short. I have to use an object with just the right shape in order to make good contact with both pads. Also, sometimes it doesn't work if you remove the short too fast. To be safe I generally hold it for about 1 sec after releasing the reset button.
What household materials would you recommend in order to get a good short? I used a screwdriver, but as I mentioned, it didn't work.
I also tried a different screwdriver, and then aluminum foil. Nothing.
I use a pair of scissors. Nice pointy tips, fully conductive and often with an insulated handle (not safety for you, but for the device).
Well, I used a pair of scissors, and it worked. Windows gave me a notification saying "We're setting up 'SE Blank 6ull'", but the NAND backup script is still giving me an error saying that the device was not found. What do I do now?
Tried using a paperclip for shorting, and now I am getting the default firmware boot but with a twist. It boot the firmware, but it has an intense white overlay, and the screen is frozen. Back to normal after resetting though.
So apparently the error I was getting was because I was missing certain files. After resolving this error, I am encountering problems with the NAND backup script. It keeps failing at 3/33, and I am not seeing any file output in the "backup" folder. See attached screenshot for details.
[attachment=9783]
Reference URL's