<Armbian-Discord> <T​onymac32> "Synchronous Abort" handler, esr 0x96000004 elr: 0000000001065468 lr : 0000000001054df4 (reloc) elr: 000000007dfb4468 lr : 000000007dfa3df4 x0 : 907c3837ab9509de x1 : 000000007dfbc140 x2 : 0000000000000010 x3 : 0000000000000000 x4 : 0000000000000000 x5 : 907c3837ab9509de x6 : 000000007bf54e10 x7 : 0000000000000000 x8 : 0000000000000007 x9 : 0000000000000000 x10: 0000000000000178 x11: 000000007bf40e2c x12:
<Armbian-Discord> 00000000000000a4 x13: 000000007bf40de8 x14: 000000007a81e000 x15: 0000000000000020 x16: 000000007bf3c2a7 x17: 0000000000000000 x18: 000000007bf4cdb0 x19: 000000007af41040 x20: 000000007df4fb20 x21: 000000007dfbc140 x22: 0000000000001000 x23: 000000007bf54d60 x24: 000000007dfdac00 x25: 0000000001000000 x26: 0000000001000000 x27: 0000000001000000 x28: 0000000000001000 x29: 000000007bf40de0 Code: eb02009f 54000061 52800000
<Armbian-Discord> 14000006 (386468a3) Resetting CPU ... resetting ...
<Armbian-Discord> <T​onymac32> U-boot fail, I'll have to take a look
<Armbian-Discord> <T​onymac32> shoot first ask questions later image building, unfortunately
<Armbian-Discord> <T​onymac32> My apologies for not getting to it sooner, I was given a heads up to test new images, but never imagined they were being this reckless: https://github.com/armbian/build/commit/15cab819fe5f4b0496f65ed39fccc426e635ce82
<Armbian-Discord> <T​onymac32> This commit bumps u-boot to a new version for all Amlogic boards with no testing at all. 22.07 breaks the Le Potato just as it broke the VIM3l @rpardini fixed.
<Armbian-Discord> <T​onymac32> Let me fix it.
<Armbian-Discord> <T​onymac32> PR submitted
<Armbian-Discord> <I​gorPec> good morning
<Armbian-Discord> <I​gorPec> i tested 3 amlogic boards and all worked ... rebuilding images
<Armbian-Discord> <a​deepv> It seems that somewhere there is either a buffer overflow or something like that. and it does not show up in all cases.
<Armbian-Discord> <I​gorPec> yes. anyway i have rebuilding images one by one
<Armbian-Discord> <I​gorPec> yours should be fine
<Armbian-Discord> <I​gorPec> it seems with uboot we always need to test all boards
<Armbian-Discord> <r​pardini> This, my guess, is RAM-size related. All my 4g boards work with 07 (and 10)
<Armbian-Discord> <I​gorPec> you mean that's why lepotato failed?
<Armbian-Discord> <r​pardini> Could be.
<Armbian-Discord> <a​deepv> i fully disable EFI config for u-boot
<Armbian-Discord> <M​anoftheSea> Do u-boot devs test on boards?
<Armbian-Discord> <n​armstrong> Not enough...
<Armbian-Discord> <d​sx724> Every u-boot release is basically a full re-test due to changes. The target is moving too fast with all the formalization changes going on.
<Armbian-Discord> <d​sx724> The whole device-tree between Linux and u-boot is completely non-scalable.
<Armbian-Discord> <c​0rnelius> Thats an intense stance. I don't think the problem with booting the Le Potato had anything to do with moving from v2022.01 to 07. I downloaded the img and reflashed uboot on it. The same error that was reported is happening regardless of the u-boot release. To clarify, I have a two Le Potato's running on Linux 6.x.y and u-boot v2022.10. They boot fine.
<Armbian-Discord> <I​gorPec> Did you check uboot version on those images? I did it very fast
<Armbian-Discord> <r​pardini> Ok so what is going on, then?
<Armbian-Discord> <c​0rnelius> Thats a good question and I'm not sure yet. I haven't had time to really inspect it. To get Armbian to boot, I flashed the latest u-boot to the SD and dropped in my own kernel. I also added an extlinux.conf file. https://paste.debian.net/1259293/
_whitelogger has joined #armbian-amlogic