Re: [PATCH v8 11/24] erofs: introduce xattr & posixacl support

From: David Sterba
Date: Mon Sep 02 2019 - 10:20:22 EST


On Mon, Sep 02, 2019 at 09:51:59PM +0800, Chao Yu wrote:
> On 2019-9-2 21:06, David Sterba wrote:
> > On Mon, Sep 02, 2019 at 05:57:11AM -0700, Christoph Hellwig wrote:
> >>> +config EROFS_FS_XATTR
> >>> + bool "EROFS extended attributes"
> >>> + depends on EROFS_FS
> >>> + default y
> >>> + help
> >>> + Extended attributes are name:value pairs associated with inodes by
> >>> + the kernel or by users (see the attr(5) manual page, or visit
> >>> + <http://acl.bestbits.at/> for details).
> >>> +
> >>> + If unsure, say N.
> >>> +
> >>> +config EROFS_FS_POSIX_ACL
> >>> + bool "EROFS Access Control Lists"
> >>> + depends on EROFS_FS_XATTR
> >>> + select FS_POSIX_ACL
> >>> + default y
> >>
> >> Is there any good reason to make these optional these days?
> >
> > I objected against adding so many config options, not to say for the
> > standard features. The various cache strategies or other implementation
> > details have been removed but I agree that making xattr/acl configurable
> > is not necessary as well.
>
> I can see similar *_ACL option in btrfs/ext4/xfs, should we remove them as well
> due to the same reason?

Oh right, I think the reasons are historical and that we can remove the
options nowadays. From the compatibility POV this should be safe, with
ACLs compiled out, no tool would use them, and no harm done when the
code is present but not used.

There were some efforts by embedded guys to make parts of kernel more
configurable to allow removing subsystems to reduce the final image
size. In this case I don't think it would make any noticeable
difference, eg. the size of fs/btrfs/acl.o on release config is 1.6KiB,
while the whole module is over 1.3MiB.