05:52 <Armbian-Discord> <T​onymac32> I tested Potato, C2, and K2 with the images that wre complaint. None booted. I reverted U-boot and all boot

05:53 <Armbian-Discord> <T​onymac32> so 100% of my test boards failed 22.07 😉

05:53 <Armbian-Discord> <T​onymac32> I haven't tried VIM, but it's also a GXL so...

05:56 <Armbian-Discord> <T​onymac32> What 3, out of curiosity?

11:00 <Armbian-Discord> <r​pardini> Hmmm -- using extlinux might avoid hitting loadaddr and such in bootscripts, which might contribute to the problem. Specially on lower RAM boards? I'm speculating.

13:05 <Armbian-Discord> <T​onymac32> Possibly

14:48 <Armbian-Discord> <c​0rnelius> @rpardini maybe. that boot.cmd file is very busy. extlinux I find to be way easier to use in general.

17:11 <Armbian-Discord> <r​pardini> Seems to me, Neil has all the correct addresses in the _defconfig in u-boot, and our bootscript just ignores those and uses the same addresses across the family. extlinux should work better indeed...

17:48 <Armbian-Discord> <T​onymac32> facepalm

17:49 <Armbian-Discord> <T​onymac32> Ok, so I'm hearing "meson64 needs split up if we keep doing it our way"

17:50 <Armbian-Discord> <T​onymac32> Because you can't reconcile G12 and later with GXL/L/M

17:51 <Armbian-Discord> <T​onymac32> I don't remember how we package the boot script. A GX/L/M one and a G12+ one might be worthwhile

17:52 <Armbian-Discord> <T​onymac32> Or modify to use the defconfig defaults