From: Nick Clark <nicholas.clark@xxxxxxxxx> Hi all, I wrote a couple of patches for fuse2fs (from the e2fsprogs collection). One is a minor bugfix to the displayed name in the mount command's output, and the other is a new 'fakeroot' mount option, as described below. When building cross-compiled Linux systems (for use on routers, consumer devices, etc), it's often necessary to build rootfs filesystem images. This commit adds a new 'fakeroot' mount option that gives a user full control over the mounted filesystem, within the confines of FUSE. When used, the new option allows treats allowed users as uid 0 for the purposes of permission checks (within the mounted filesystem). This allows the owner of a .img file to mount the file and make whatever changes they want to any files inside of the image. If the option is left disabled at mount-time, fuse2fs acts the same as before - only root can edit "root"-owned files inside of the image. But when enabled, fuse2fs becomes a powerful tool to build/modify filesystem images. What do you all think? -Nick Nicholas Clark (2): Fuse2fs: fix 'mount' entry in some cases. Fuse2fs: add fakeroot option. acinclude.m4 | 47 ++++++++++++++++ configure | 134 ++++++++++++++++++++++++++++++++++++++++++++++ configure.ac | 4 ++ lib/config.h.in | 6 +++ misc/fuse2fs.1.in | 3 ++ misc/fuse2fs.c | 31 ++++++++--- 6 files changed, 217 insertions(+), 8 deletions(-) -- 2.17.1
