qemu-devel
[Top][All Lists]
Advanced

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

Re: [Qemu-devel] [PATCH 00/15] NVDIMM: introduce nvdimm label support


From: Xiao Guangrong
Subject: Re: [Qemu-devel] [PATCH 00/15] NVDIMM: introduce nvdimm label support
Date: Wed, 23 Mar 2016 10:46:56 +0800
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.5.1


Sorry for the delay and i just come back from my vacation.

On 03/23/2016 04:30 AM, Stefan Hajnoczi wrote:
On Tue, Mar 22, 2016 at 08:37:40AM -0700, Dan Williams wrote:
On Thu, Mar 17, 2016 at 3:04 AM, Stefan Hajnoczi <address@hidden> wrote:
On Thu, Mar 17, 2016 at 04:32:46PM +0800, Xiao Guangrong wrote:
This patchset is against commit d4207b223eef3 (fw-cfg: support writeable
blobs) on pci branch of Michael's git tree and can be found at:
       https://github.com/xiaogr/qemu.git nvdimm-label-v1

This is the last part of vNVDIMM implementation which introduces nvdimm
label support

Currently Linux NVDIMM driver does not support namespace operation on this
kind of PMEM, apply below changes to support dynamical namespace:

@@ -798,7 +823,8 @@ static int acpi_nfit_register_dimms(struct acpi_nfit_desc *a
                         continue;
                 }

-               if (nfit_mem->bdw && nfit_mem->memdev_pmem)
+               //if (nfit_mem->bdw && nfit_mem->memdev_pmem)
+               if (nfit_mem->memdev_pmem)
                         flags |= NDD_ALIASING;

Not a blocker for this patch series, but why does Linux require Block
Device Window to enable namespace support?

A namespace label delineates aliased capacity between the pmem and
block-window access mechanisms.  If there is no aliased capacity then
the size of the namespace can be directly derived from the nfit range
and a label need not be considered.  Contiguous (dax-capable)
sub-divisions of pmem can be had via partitioning of the resulting
gendisk.

Xiao Guangrong: Given this new information, what is the purpose of the
QEMU patches for ACPI DSM and especially the namespace label support?


The initiation goal we introduced vlabel support is:
| Yes, I see Linux driver supports label-less vNVDIMM that is exact current QEMU
| doing. However, label-less is only Linux specific implementation (as it
| completely bypasses namespace), other OS vendors (e.g Microsoft) will use 
label
| storage to address their own requirements,or they do not follow namespace spec
| at all. Another reason is that label is essential for PBLK support.

I did not get the windows driver so far, so this is not tested on windows yet,
just double checked with windows driver team, they told me that:
"It always assumes there is one there so we will need to change how this works"
So the update of windows side will be made.

The QEMU NVDIMM device only supports pmem so now I'm not sure we need
namespace labels or the ACPI DSM at all.


We are planing to support PBLK which "simulated BLK it can simply make one big
aperture that happens to overlap the same address range as PMEM", it will "allow
enabling BTT on BLK namespaces to support power-fail-safe sector updates for a
filesystem on PMEM". The idea was contributed by Dan in our internal 
discussion...






reply via email to

[Prev in Thread] Current Thread [Next in Thread]