Re: [RFC PATCH] xfs: support for non-mmu architectures

[Dave Chinner]

On Fri, Nov 20, 2015 at 03:43:20PM +0200, Octavian Purdila wrote:
> On Fri, Nov 20, 2015 at 1:24 AM, Dave Chinner <david@xxxxxxxxxxxxx> wrote:
> > On Wed, Nov 18, 2015 at 12:46:21AM +0200, Octavian Purdila wrote:
> >> Naive implementation for non-mmu architectures: allocate physically
> >> contiguous xfs buffers with alloc_pages. Terribly inefficient with
> >> memory and fragmentation on high I/O loads but it may be good enough
> >> for basic usage (which most non-mmu architectures will need).
> >
> > Can you please explain why you want to use XFS on low end, basic
> > non-MMU devices? XFS is a high performance, enterprise/HPC level
> > filesystem - it's not a filesystem designed for small IoT level
> > devices - so I'm struggling to see why we'd want to expend any
> > effort to make XFS work on such devices....
> >
> 
> Hi David,
> 
> Yes XFS as the main fs on this type of devices does not make sense,
> but does it hurt to be able to perform basic operation on XFS from
> these devices? Perhaps accessing an external medium formatted with
> XFS?
> 
> Another example is accessing VM images that are formatted with XFS.
> Currently we can do that with tools like libguestfs that use a VM in
> the background. I am working on a lighter solution for that where we
> compile the Linux kernel as a library [1]. This allows access to the
> filesystem without the need to use a full VM.

That's hardly a "lighter solution".

I'm kinda tired of the ongoing "hack random shit" approach to
container development. If you need a XFS-FUSE module to allow safe
userspace access to XFS fielsystems then maybe, just maybe, it makes
sense to ask the XFS developers how to best go about providing a
reliable, up-to-date, tested, maintained and supported XFS-FUSE
module?

IOWs, a "lighter solution" is to use the libxfs code base that we
already maintain across kernel and userspace in the xfsprogs package
and write a FUSE wrapper around that. That, immediately, will give
you full read-only access to XFS filesystem images via FUSE. Then we
(the XFS developers) can test the XFS-FUSE module under normal
development conditions as we modify the xfsprogs code base (e.g. via
xfstests) and ensure we always release a working, up-to-date FUSE
wrapper with each xfsprogs release.

And then once a proper read-only FUSE wrapper has been written, then
we can discuss what is necessary to enable write access via porting
the necessary parts of the kernel code across to the userspace
libxfs codebase and hooking them up to the FUSE API...

Hmmm?

> And a final example is linking the bootloader code with LKL to access
> the filesystem. This has a hard requirement on non-mmu.

No way. We *can't* support filesystems that have had bootloaders
make arbitrary changes to the filesystem without the knowlege of the
OS that *owns the filesystem*. Similarly, we cannot support random
applications that internally mount and modify filesystem images in
ways we can't see, control, test or simulate.  Sure, they use the
kernel code, but that doesn't stop them from doing stupid shit that
could corrupt the filesystem image. So, no, we are not going to
support giving random applications direct access to XFS filesystem
images, even via LKL.

> So, IMHO there are usecases for using XFS on non-mmu architectures. I
> think it all boils down to: is the patch simple enough to not put an
> unreasonable maintenance burden on developers?

Look at it this way: a user tries to mount their
XFS-on-LKL-on-FUSE, XFS throughs a memory allocation deadlock
warning because of a high order allocation during mount failing
(i.e. due to the vmalloc no-mmu hacks).

Who has to spend effort to find out why the error is being thrown?
It's not the FUSE developers. It's not the LKL developers. XFS is
going to be blamed, because it's an XFS error message, and it will
be up to the XFS developers to /prove/ that it's not an XFS problem.
And then google will find these complaints about XFS causing all
sorts of problems, and we're back to the bad old days of everyone
blaming XFS for shortcomings and problems in other code we have no
control over..

I really don't see how using LKL to give userspace access to XFS
filesystems is a better solution than actually writing a proper,
supported XFS-FUSE module. LKL is so full of compromises that it's
going to be unworkable and unsupportable in practice...

Cheers,

Dave.
-- 
Dave Chinner
david@xxxxxxxxxxxxx
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/