On Thu, Jan 18, 2024 at 12:04 AM Christian Marangi <ansuelsmth@xxxxxxxxx> wrote: > Some followup on this... I manage to enable DEBUG_LL and can have debug > output from the decompressor... Yeah that is helpful! > From what I can see fdt_check_mem_start is not called at all... > > What I'm using with kernel config are: > CONFIG_ARM_APPENDED_DTB=y > CONFIG_ARM_ATAG_DTB_COMPAT=y > And a downstream patch that mangle all the atags and takes only the > cmdline one. > > The load and entry point is: > 0x42208000 > > With the current setup I have this (I also added some debug log that > print what is actually passed to do decompress > > DTB:0x42AED270 (0x00008BA7) > Uncompressing Linux... > 40208000 > 4220F10C done, booting the kernel. > > Where 40208000 is the value of output_start and 4220F10C is input_data. > > And I think this confirm that it's getting loaded in the wrong position > actually in reserved memory... But how this is possible??? Hope can > someone help me in this since I wasted the entire day with this and > didn't manage to make any progress... aside from having fun with the > head.S assembly code. I have no idea how this happens, but when I boot images I do it using fastboot like this: fastboot --base 40200000 --cmdline "console=ttyMSM0,115200,n8" boot zImage So I definitely hammer it to boot from 0x40200000 (+0x8000). Yours, Linus Walleij
