wooosaiiii has quit [Remote host closed the connection]
wooosaiiii has joined #u-boot
fdanis_away is now known as fdanis
agust has joined #u-boot
smartin has joined #u-boot
mckoan|away is now known as mckoan
frieder has joined #u-boot
<j`ey>
marex: sorry, I was just grumpy and it was nearly midnight, so it felt like a good time to stop :p
tnovotny has joined #u-boot
milkylainen has joined #u-boot
monstr has joined #u-boot
guillaume_g has joined #u-boot
qschulz has joined #u-boot
sszy has joined #u-boot
mmu_man has joined #u-boot
mmu_man has quit [Ping timeout: 250 seconds]
alan_o has quit [Ping timeout: 245 seconds]
matthias_bgg has joined #u-boot
PatDelaunay has joined #u-boot
alan_o has joined #u-boot
eduardas has joined #u-boot
Xavier7 has quit [Quit: IRcap 8.72 ]
mmu_man has joined #u-boot
PatDelaunay has quit [Quit: Client closed]
mmu_man has quit [Ping timeout: 256 seconds]
mmu_man has joined #u-boot
mmu_man has quit [Ping timeout: 256 seconds]
jwillikers has joined #u-boot
alpernebbi has joined #u-boot
PatDelaunay has joined #u-boot
monstr has quit [Ping timeout: 252 seconds]
monstr has joined #u-boot
mmu_man has joined #u-boot
<marex>
j`ey: you could still be a bit more specific -- which platform, how are you loading the blobs etc.
<j`ey>
marex: it's on the Apple M1, there's a program (called m1n1) that copies u-boot to memory somewhere and jumps to it
<marex>
oh that
<marex>
j`ey: look at U-Boot load address, make sure that u-boot.bin or whatever you load ends up at that address
<marex>
then, take u-boot.bin , pad it so the next (kernel arm64 Image) blob is at 2 MiB offset
<marex>
then concatenate them
<marex>
then boot the result , and then booti from U-Boot the kernel Image at wherever you originally loaded the combo blob + 2 MiB
<marex>
that _should_ work assuming u-boot relocation doesn't corrupt the concatenated Image (it shouldn't unless you load it very close to the u-boot relocation address , or early malloc area)
<j`ey>
I will give that a try
<j`ey>
I've been trying something where I include the kernel image in the u-boot build with '.incbin'
<marex>
j`ey: probably dont bother, cat u-boot.bin padding Image > output.blob should do fine
<j`ey>
yeah this is just what I last tried last night, I will try again the other way
mmu_man has quit [Ping timeout: 240 seconds]
<j`ey>
marex: that has gotten me further! getting "FDT and ATAGS support not compiled in" now, looking at the code to try find out which bit exactly is failing
<marex>
the FDT bit
<marex>
did you do booti on the kernel image ?
<marex>
you might need DT too, I'm not sure how that works on the m1
<marex>
either you can build the DT into the kernel image or bundle it again at the end
<j`ey>
yeah I booti the kernel image
<marex>
but then, it might be easier to assemble fitImage of kernel Image and DT and paste it at the end of your bootloader blob instead of the arm64 Image
<marex>
as a bonus, you won't need such a huge padding, since the bootloader would relocate the components to the correct location
<marex>
j`ey: dont you need to booti <kernel image> - <fdt address> ?
<j`ey>
Im not sure, this is my first time using uboot!
<marex>
j`ey: isnt there some howto for the M1 ?
<marex>
if it behaves like a normal arm64 machine, then booti <kernel> - <fdt>
<j`ey>
I'll try that, just have to figure out the fdt address!
<marex>
do you have a .dtb blob for the board ?
<j`ey>
yeah
<marex>
j`ey: well, then just stick at at another 2 MiB pad past the Image and start with that
monstr has quit [Remote host closed the connection]
mmu_man has joined #u-boot
eduardas has quit [Quit: Konversation terminated!]
sys64738 has joined #u-boot
jwillikers has quit [Read error: Connection reset by peer]
jwillikers has joined #u-boot
<sys64738>
hi, im trying to limit the amount of physical memory linux can access on a zynq (because the second core needs to run a baremetal thing, which needs a separate dram range to live in), and, while i can select eg. the uart baudrate in the dtb passed to the kernel, the memory settings seem to persist from uboots builtin device tree, instead of those passed to the kernel from the dtb file loaded from
<sys64738>
the boot medium. what could be the cause of this and is there any way to make linux use the memory configuration in the dtb file instead? (or should i ask this in a different channel?) thanks
<j`ey>
marex: I got further now, "Starting kernel ...", but then nothing :'(