Cannot POST/BOOT after an initial 'update-all.sh' execution completes successfully [Solved]

Just got a MiSTer and need some help? Join the forum and post your questions here!
User avatar
fignutz
Posts: 6
Joined: Tue May 17, 2022 6:45 pm
Has thanked: 3 times

Cannot POST/BOOT after an initial 'update-all.sh' execution completes successfully [Solved]

Unread post by fignutz »

This is my first setup/build of the MiSTer FPGA project. Thanks to excellent guidance from this community's
forums and social blades (e.g. YouTube, Discord, etc.), this build was created using the following components:

* MiSTer Addons USB Expansion Board
* MiSTer Addons I/O Digital Board
* MiSTer Addons USB Bracket adapter
* MiSTer Addons DC Jumper
* SDRAM XS-DS v2.9 128MB module (2 qty; both DE-10 slots occupied)
* MiSTer RTC Real Time Clock Board v1.3
* Noctua NF-A4x10 Fan replacement
* SanDisk Ultra A1 1TB MicroSD Card
* MeanWell GST25A05-P1J 5V PowerSupply

As the title of this thread's subject implies, running the update-all.sh script instead of the default
update.sh script provided by default by the MiSTer FUSION Image will complete
successfully. However, once the script attempts to reboot the system, the system will attempt to power on, but
none of the I/O Board LEDS will light up. Additionally, there is no VIDEO Out across the HDMI port, suggesting that
the system cannot POST.

To get back to a working state, I had to:
1) power off the MiSTer
2) remove the MicrosSD Card
3) reset the card's partitions (Windows Disk Management Tool)
4) reimage the card with the FUSION image
5) reinsert into the MiSTer and power back on.

Before I try to run the update-all.sh script again, has this kind of thing happened before, and if so are there workarounds
to using it that won't soft-brick my MiSTer?

[[ UPDATE ]]
I went ahead and tried running the default/bundled 'update.sh' script instead this time. The script ran to completion, and then proceeded
to reboot the MiSTer. However, just like the 'update-all' script, the system does not POST: no screen, no LEDS, but fan on I/O board is running.

I'm now doing the workaround mentioned here in this thread => https://misterfpga.org/viewtopic.php?p=36450#p36450
I have no idea if this will work for me as it did for the OP of that thread, but if it does, is the common practice to report this as a bug
here in the forums, or on Theypsilion's GitHub project Issues tracker?

[[ LAST UPDATE ]]
Special Thanks to bbond007 for helping me come to the answer here.

TL;DR: the SanDisk Ultra A1 1TB card is the problem.

Redoing the initial setup process with a smaller non-Ultra SanDisk brand Card (16GB), allowed the process to complete and reboot without issues for both update scripts. Sort of a shame, as this card was pretty pricey.
bbond007
Top Contributor
Posts: 519
Joined: Tue May 26, 2020 5:06 am
Has thanked: 85 times
Been thanked: 198 times

Re: Cannot POST/BOOT after an initial 'update-all.sh' execution completes successfully [Solved]

Unread post by bbond007 »

fignutz wrote: Tue May 17, 2022 7:51 pm

To get back to a working state, I had to:
1) power off the MiSTer
2) remove the MicrosSD Card
3) reset the card's partitions (Windows Disk Management Tool)
4) reimage the card with the FUSION image
5) reinsert into the MiSTer and power back on.
That is kind of the thermonuclear approach to fixing the problem...

Next time try and just replace the /media/fat/MiSTer file and it will probably work again.

It can be usually fixed using WinSCP and you don't even need to remove the SD card, and I've never resorted to reimaging.

If that does not work, then also replace the following:

Code: Select all

/media/fat/menu.rbf
/media/fat/linux/linux.img
/media/fat/linux/zImage_dtb
User avatar
fignutz
Posts: 6
Joined: Tue May 17, 2022 6:45 pm
Has thanked: 3 times

Re: Cannot POST/BOOT after an initial 'update-all.sh' execution completes successfully [Solved]

Unread post by fignutz »

Hi there @bbond007 ,

I should've made it more clear that, in this state, the MiSTer (for me) does not show up at all via SCP/SSH/SFTP. So, doing a patch in of the MiSTer file as you've suggested would not be possible (for me, in this circumstance).
User avatar
Wave
Posts: 129
Joined: Tue Jan 04, 2022 6:57 pm
Has thanked: 86 times
Been thanked: 40 times

Re: Cannot POST/BOOT after an initial 'update-all.sh' execution completes successfully [Solved]

Unread post by Wave »

Sounds frustrating. :/

Just as a note, I successfully use this SanDisk 1TB A1 with the MiSTer. Did the initial setup about six months ago; wonder if I'd run into trouble like you did using the current setup software.

Another thing is, I do often hear MicroSD cards bought online can be counterfeit.
she/her
User avatar
fignutz
Posts: 6
Joined: Tue May 17, 2022 6:45 pm
Has thanked: 3 times

Re: Cannot POST/BOOT after an initial 'update-all.sh' execution completes successfully [Solved]

Unread post by fignutz »

Wave wrote: Tue May 17, 2022 11:40 pm Sounds frustrating. :/

Just as a note, I successfully use this SanDisk 1TB A1 with the MiSTer. Did the initial setup about six months ago; wonder if I'd run into trouble like you did using the current setup software.

Another thing is, I do often hear MicroSD cards bought online can be counterfeit.
Mine was acquired from a 'Trusted' 3rd Party vendor on NewEgg.com (shipped from Shenzhen). If it was counterfeit, its a very convincing one (eg. SanDisk machine sealed packaging and paperwork) :? .
User avatar
Wave
Posts: 129
Joined: Tue Jan 04, 2022 6:57 pm
Has thanked: 86 times
Been thanked: 40 times

Re: Cannot POST/BOOT after an initial 'update-all.sh' execution completes successfully [Solved]

Unread post by Wave »

Apparently they can be pretty convincing, even in packaging. It's easy enough to test on your PC (by trying to fill it to capacity, for ex) if suspected.
she/her
chromium
Posts: 31
Joined: Sun May 15, 2022 10:25 am
Has thanked: 4 times
Been thanked: 7 times

Re: Cannot POST/BOOT after an initial 'update-all.sh' execution completes successfully [Solved]

Unread post by chromium »

I just ran update_all while connected to my 1084S via Direct Video, once the update completed it rebooted and now the image won't sync :(
bbond007
Top Contributor
Posts: 519
Joined: Tue May 26, 2020 5:06 am
Has thanked: 85 times
Been thanked: 198 times

Re: Cannot POST/BOOT after an initial 'update-all.sh' execution completes successfully [Solved]

Unread post by bbond007 »

fignutz wrote: Tue May 17, 2022 7:51 pm Redoing the initial setup process with a smaller non-Ultra SanDisk brand Card (16GB), allowed the process to complete and reboot without issues for both update scripts. Sort of a shame, as this card was pretty pricey.
I initially had 64GB card that turned read-only. Worked for a long time though (maybe a year). I kept modifying an INI file and and the changes would not persist after a reboot.

Sorry about your card :(
chromium
Posts: 31
Joined: Sun May 15, 2022 10:25 am
Has thanked: 4 times
Been thanked: 7 times

Re: Cannot POST/BOOT after an initial 'update-all.sh' execution completes successfully [Solved]

Unread post by chromium »

https://github.com/MiSTer-devel/Main_Mi ... r_20220511

I reverted back to the version there and now it works again.

So the latest version has broken direct video? At least for me it has
bbond007
Top Contributor
Posts: 519
Joined: Tue May 26, 2020 5:06 am
Has thanked: 85 times
Been thanked: 198 times

Re: Cannot POST/BOOT after an initial 'update-all.sh' execution completes successfully [Solved]

Unread post by bbond007 »

Wave wrote: Tue May 17, 2022 11:40 pm Just as a note, I successfully use this SanDisk 1TB A1 with the MiSTer. Did the initial setup about six months ago; wonder if I'd run into trouble like you did using the current setup software.
I think B&H is probably one of the most reliable vendors out there. I definitely would trust them over Newegg at this point.
User avatar
fignutz
Posts: 6
Joined: Tue May 17, 2022 6:45 pm
Has thanked: 3 times

Re: Cannot POST/BOOT after an initial 'update-all.sh' execution completes successfully [Solved]

Unread post by fignutz »

Indeed (B&H). I completely forgot about using that outlet. I'm currently running a comprehensive fill test on my possible faulty/illegitimate MicroSD card now with H2testw
Post Reply