From mboxrd@z Thu Jan 1 00:00:00 1970 From: Peter Korsgaard Date: Sun, 27 Mar 2011 23:12:51 +0200 Subject: [Buildroot] Booting with ramdisk In-Reply-To: <201103211546.22271.dargaud@lpsc.in2p3.fr> (Guillaume Dargaud's message of "Mon, 21 Mar 2011 15:46:22 +0100") References: <201103211546.22271.dargaud@lpsc.in2p3.fr> Message-ID: <8739m8mfn0.fsf@macbook.be.48ers.dk> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net >>>>> "Guillaume" == Guillaume Dargaud writes: Hi, Guillaume> Now I want to embed that OS in the kernel (no need for a Guillaume> network anymore as the kernel is on a flash mem). So I did Guillaume> the following: - add ROOTFS CPIO to the buildroot options Guillaume> and make a new version - point CONFIG_INITRAMFS_SOURCE to Guillaume> the resulting rootfs.cpio file That sounds correct. If you build your kernel in buildroot we even have initramfs option that handles it all, but the above should afaik work as well. Guillaume> - change the kernel parameters to "console=ttyUL0,115200 root=/dev/ram" (that's in a dts file) That shouldn't be needed. Guillaume> The kernel stops instantly without even a blip on the Guillaume> console. A debugger reveals that it stops at 0x401718... I Guillaume> have no idea what's going on at this stage. Strange. What architecture/system is this? How much did the (compressed) kernel grow? Perhaps it's too big / you'll need to load it higher. For most archs, the kernel want to decompress into address 0, so if you load it too low, you might end up overwriting the compressed kernel as you're decompressing it. Perhaps you could add some debugging prints to the bootloader handover / kernel decompression? -- Bye, Peter Korsgaard