On Fri, Jul 20, 2018 at 01:46:06PM -0700, Verma, Vishal L wrote:
>
> On Thu, 2018-07-12 at 09:47 -0600, Keith Busch wrote:
> > This patch will find the max contiguous area to determine the largest
> > pmem namespace size that can be created. If the requested size exceeds
> > the largest available, ENOSPC error will be returned.
> >
> > This fixes the allocation underrun error and wrong error return code
> > that have otherwise been observed as the following kernel warning:
> >
> > WARNING: CPU: <CPU> PID: <PID> at drivers/nvdimm/namespace_devs.c:913 size_store
> >
> > Fixes: a1f3e4d6a0c3 ("libnvdimm, region: update nd_region_available_dpa() for multi-pmem support")
> > Cc: <stable@xxxxxxxxxxxxxxx>
> > Signed-off-by: Keith Busch <keith.busch@xxxxxxxxx>
>
> Hi Keith,
>
> I was testing these patches and I found:
>
> When booting a VM which has both, a qemu ACPI.NFIT bus, and nfit_test
> buses, initially the nfit_test buses show correct max_available_extent.
> But the qemu ACPI.NFIT bus regions (which have an automatic full-
> capacity namespace created on them when they come up) show
> max_available_extent of the full region size, even as the
> available_size attr is zero.
The max extents only counts the free pmem that it can reserve.
We shouldn't have been able to reserve non-free pmem, so it sounds like
something must be wrong with how the resources were set up.
I'll make a similar qemu config and see why/if the resource was
considered free.
