Unfortunately I am on Pi Zero 2 W, so this (modulo the changes needed to compile 32-bit Pi0 kernel, not Pi5) has been running for about 3 hours now already, and it doesn't seem to be anywhere close to completion. I'll report when it's done (hopefully within a couple of days lol).assuming you're on Pi5also refer to https://www.raspberrypi.com/documentati ... el-locallyCode:
sudo apt install git bc bison flex libssl-dev makegit clone --depth=1 --branch rpi-6.6.y https://github.com/raspberrypi/linuxcd linuxKERNEL=kernel_2712make bcm2712_defconfigmake menuconfigmake -j4 Image.gz modules dtbssudo make modules_installsudo cp arch/arm64/boot/dts/broadcom/*.dtb /boot/firmware/sudo cp arch/arm64/boot/dts/overlays/*.dtb* /boot/firmware/overlays/sudo cp arch/arm64/boot/dts/overlays/README /boot/firmware/overlays/sudo cp arch/arm64/boot/Image.gz /boot/firmware/$KERNEL.img
Even if it does produce the same symvers as official builds, it would still leave the mystery of why cross-compiling under several different environments produces different (but internally consistent) results. I don't think official builds are native builds, it would make sense to cross-compile there too.
And again, I don't have trouble building the kernel. I have trouble building a kernel with *the same* symvers as the official one. Symvers difference suggest something very serious, IIUC something as serious as pointer size difference, struct size/field order/alignment/layout difference, etc. This is what concerns me.
Statistics: Posted by provod — Thu Apr 04, 2024 9:31 pm