From: Menglong Dong <dong.menglong@xxxxxxxxxx> As Luis Chamberlain suggested, I split the patch: [init/initramfs.c: make initramfs support pivot_root] (https://lore.kernel.org/linux-fsdevel/20210520154244.20209-1-dong.menglong@xxxxxxxxxx/) into three. The goal of the series patches is to make pivot_root() support initramfs. In the first patch, I introduce the function ramdisk_exec_exist(), which is used to check the exist of 'ramdisk_execute_command' in relative path mode. In the second patch, I create a second mount, which is called 'user root', and make it become the root. Therefore, the root has a parent mount, and it can be umounted or pivot_root. Before change root, I have to check the exist of ramdisk_execute_command, because 'user root' should be umounted if ramdisk_execute_command not exist. 'user root' is mounted on '/root', and cpio is unpacked to it. So I have to use relative path to do this check, as 'user root' is not the root yet. Maybe I can do the check after change root, but it seems complex to change root back to '/'. What's weird is that I try to move 'user root' from '/root' to '/', but the absolute path lookup seems never follow the mount. That's why I introduced ramdisk_exec_exist. In the third patch, I fix rootfs_fs_type with ramfs, as it is not used directly any more, and it make no sense to switch it between ramfs and tmpfs, just fix it with ramfs to simplify the code. Menglong Dong (3): init/main.c: introduce function ramdisk_exec_exist() init/do_cmounts.c: introduce 'user_root' for initramfs init/do_mounts.c: fix rootfs_fs_type with ramfs fs/namespace.c | 2 -- include/linux/init.h | 1 - init/do_mounts.c | 82 +++++++++++++++++++++++++++++++++++++------- init/do_mounts.h | 7 +++- init/initramfs.c | 10 ++++++ init/main.c | 17 ++++++++- 6 files changed, 101 insertions(+), 18 deletions(-) -- 2.31.1
