Re: [PATCH] hfsplus: drop ACL support

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Tue, Jul 17, 2018 at 11:29:17AM -0700, Viacheslav Dubeyko wrote:
> On Tue, 2018-07-17 at 15:18 +0200, Christoph Hellwig wrote:
> > On Mon, Jul 16, 2018 at 05:07:48PM -0700, Andrew Morton wrote:
> > > 
> > > On Sat, 14 Jul 2018 12:11:03 -0700 Viacheslav Dubeyko <slava@dubeyk
> > > o.com> wrote:
> > > 
> > > > 
> > > > On Sat, 2018-07-14 at 16:06 -0300, Ernesto A. Fernández wrote:
> > > > > 
> > > > > The HFS+ Access Control Lists have not worked at all for the
> > > > > past
> > > > > five
> > > > > years, and nobody seems to have noticed. Besides, POSIX draft
> > > > > ACLs
> > > > > are
> > > > > not compatible with MacOS. Drop the feature entirely.
> > > > > 
> > > > Bugs need to be fixed but not to drop. Otherwise, it needs to
> > > > drop the
> > > > whole HFS+ support from the kernel.
> > > Yes, we could make it depend on BROKEN and hope that someone comes
> > > along and fixes it - there's little cost in carrying such a thing.
> > > 
> > > Or maybe it can never work for some reason and should indeed be
> > > dropped, I don't know.  I was interested in understanding the
> > > reasoning
> > > behind hch's ack, but I can't find it on linux-fs-devel.  Help?
> > The issue is that hfsplus supports it's own NFSv4-like ACLs, which
> > are nothing like the Posix ACLs supported by the Linux driver.  Or
> > in other words: they were always a Linux-only incompatible extension
> > that should never have been merged.
> 
> 
> The richacls subsystem was not in the kernel at the time of ACL support
> implementation. So, the idea of Linux-like ACLs in HFS+ was not to be
> compatible with MAC OS X. The goal was to provide the opportunity to
> use the ACLs only in Linux if anybody likes this. The MAC OS X will
> ignore such xattrs.
> 
> As far as I can see, the ACL subsystem in HFS+ stops to work after
> Christoph's reworking the ACL subsystem in the Linux. So, we have
> richacls subsystem in the Linux kernel now. I believe that it makes
> sense to implement the richacls support in HFS+ at first. Then we can
> discuss to drop or not to drop the current ACL implementation in HFS+. 

Richacls are *not* in the kernel yet.

- Eric



[Index of Archives]     [Linux Ext4 Filesystem]     [Union Filesystem]     [Filesystem Testing]     [Ceph Users]     [Ecryptfs]     [AutoFS]     [Kernel Newbies]     [Share Photos]     [Security]     [Netfilter]     [Bugtraq]     [Yosemite News]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux Cachefs]     [Reiser Filesystem]     [Linux RAID]     [Samba]     [Device Mapper]     [CEPH Development]

  Powered by Linux