]> www.infradead.org Git - users/jedix/linux-maple.git/commit
namespace: introduce SB_I_NOIDMAP flag
authorAlexander Mikhalitsyn <aleksandr.mikhalitsyn@canonical.com>
Tue, 3 Sep 2024 15:16:12 +0000 (17:16 +0200)
committerMiklos Szeredi <mszeredi@redhat.com>
Wed, 4 Sep 2024 14:47:49 +0000 (16:47 +0200)
commit2097154a10c6ee78be8796411e5d0ad81ee06ed6
tree84863cfe85a4756c0e93e5f3cf975c4d0162670c
parent506b21c945b9716a1e092189c260d9400c52fa14
namespace: introduce SB_I_NOIDMAP flag

Right now we determine if filesystem support vfs idmappings or not basing
on the FS_ALLOW_IDMAP flag presence. This "static" way works perfecly well
for local filesystems like ext4, xfs, btrfs, etc. But for network-like
filesystems like fuse, cephfs this approach is not ideal, because sometimes
proper support of vfs idmaps requires some extensions for the on-wire
protocol, which implies that changes have to be made not only in the Linux
kernel code but also in the 3rd party components like libfuse, cephfs MDS
server and so on.

We have seen that issue during our work on cephfs idmapped mounts [1] with
Christian, but right now I'm working on the idmapped mounts support for
fuse/virtiofs and I think that it is a right time for this extension.

[1] 5ccd8530dd7 ("ceph: handle idmapped mounts in create_request_message()")

Suggested-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Alexander Mikhalitsyn <aleksandr.mikhalitsyn@canonical.com>
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
fs/namespace.c
include/linux/fs.h