On Thu, Jun 18, 2020 at 11:19:12AM +1000, Dave Chinner wrote:
> On Wed, Jun 17, 2020 at 07:57:29AM +0000, Satya Tangirala wrote:
> > Introduce SB_INLINECRYPT, which is set by filesystems that wish to use
> > blk-crypto for file content en/decryption. This flag maps to the
> > '-o inlinecrypt' mount option which multiple filesystems will implement,
> > and code in fs/crypto/ needs to be able to check for this mount option
> > in a filesystem-independent way.
> >
> > Signed-off-by: Satya Tangirala <satyat@xxxxxxxxxx>
> > ---
> > fs/proc_namespace.c | 1 +
> > include/linux/fs.h | 1 +
> > 2 files changed, 2 insertions(+)
> >
> > diff --git a/fs/proc_namespace.c b/fs/proc_namespace.c
> > index 3059a9394c2d..e0ff1f6ac8f1 100644
> > --- a/fs/proc_namespace.c
> > +++ b/fs/proc_namespace.c
> > @@ -49,6 +49,7 @@ static int show_sb_opts(struct seq_file *m, struct super_block *sb)
> > { SB_DIRSYNC, ",dirsync" },
> > { SB_MANDLOCK, ",mand" },
> > { SB_LAZYTIME, ",lazytime" },
> > + { SB_INLINECRYPT, ",inlinecrypt" },
> > { 0, NULL }
> > };
> > const struct proc_fs_opts *fs_infop;
>
> NACK.
>
> SB_* flgs are for functionality enabled on the superblock, not for
> indicating mount options that have been set by the user.
That's an interesting claim, given that most SB_* flags are for mount options.
E.g.:
ro => SB_RDONLY
nosuid => SB_NOSUID
nodev => SB_NODEV
noexec => SB_NOEXEC
sync => SB_SYNCHRONOUS
mand => SB_MANDLOCK
noatime => SB_NOATIME
nodiratime => SB_NODIRATIME
lazytime => SB_LAZYTIME
>
> If the mount options are directly parsed by the filesystem option
> parser (as is done later in this patchset), then the mount option
> setting should be emitted by the filesystem's ->show_options
> function, not a generic function.
>
> The option string must match what the filesystem defines, not
> require separate per-filesystem and VFS definitions of the same
> option that people could potentially get wrong (*cough* i_version vs
> iversion *cough*)....
Are you objecting to the use of a SB_* flag, or just to showing the flag in
show_sb_opts() instead of in the individual filesystems? Note that the SB_*
flag was requested by Christoph
(https://lkml.kernel.org/r/20191031183217.GF23601@xxxxxxxxxxxxx/,
https://lkml.kernel.org/r/20191031212103.GA6244@xxxxxxxxxxxxx/). We originally
used a function fscrypt_operations::inline_crypt_enabled() instead.
- Eric
