Re: [PATCH v2 3/5] libnvdimm/nvdimm/flush: Allow architecture to override the flush barrier

On Thu, May 21, 2020 at 7:39 AM Jeff Moyer [off-list ref] wrote:

Dan Williams [off-list ref] writes:

quoted

quoted

But I agree with your concern that if we have older kernel/applications
that continue to use `dcbf` on future hardware we will end up
having issues w.r.t powerfail consistency. The plan is what you outlined
above as tighter ecosystem control. Considering we don't have a pmem
device generally available, we get both kernel and userspace upgraded
to use these new instructions before such a device is made available.

I thought power already supported NVDIMM-N, no?  So are you saying that
those devices will continue to work with the existing flushing and
fencing mechanisms?

quoted

Ok, I think a compile time kernel option with a runtime override
satisfies my concern. Does that work for you?

The compile time option only helps when running newer kernels.  I'm not
sure how you would even begin to audit userspace applications (keep in
mind, not every application is open source, and not every application
uses pmdk).  I also question the merits of forcing the administrator to
make the determination of whether all applications on the system will
work properly.  Really, you have to rely on the vendor to tell you the
platform is supported, and at that point, why put further hurdles in the
way?

I'm thoroughly confused by this. I thought this was exactly the role
of a Linux distribution vendor. ISVs qualify their application on a
hardware-platform + distribution combination and the distribution owns
picking ABI defaults like CONFIG_SYSFS_DEPRECATED regardless of
whether they can guarantee that all apps are updated to the new
semantics.

The administrator is not forced, the administrator if afforded an
override in the extreme case that they find an exception to what was
qualified and need to override the distribution's compile-time choice.

The decision to require different instructions on ppc is unfortunate,
but one I'm sure we have no control over.  I don't see any merit in the
kernel disallowing MAP_SYNC access on these platforms.  Ideally, we'd
have some way of ensuring older kernels don't work with these new
platforms, but I don't think that's possible.

I see disabling MAP_SYNC as the more targeted form of "ensursing older
kernels don't work.

So I guess we agree that something should break when baseline
assumptions change, we just don't yet agree on where that break should
happen?