[lkml]   [2018]   [Aug]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: [PATCH V3 0/4] Fix kvm misconceives NVDIMM pages as reserved mmio

On 2018年08月10日 21:27, David Hildenbrand wrote:
> On 09.08.2018 12:52, Zhang Yi wrote:
>> For device specific memory space, when we move these area of pfn to
>> memory zone, we will set the page reserved flag at that time, some of
>> these reserved for device mmio, and some of these are not, such as
>> NVDIMM pmem.
>> Now, we map these dev_dax or fs_dax pages to kvm for DIMM/NVDIMM
>> backend, since these pages are reserved. the check of
>> kvm_is_reserved_pfn() misconceives those pages as MMIO. Therefor, we
>> introduce 2 page map types, MEMORY_DEVICE_FS_DAX/MEMORY_DEVICE_DEV_DAX,
>> to indentify these pages are from NVDIMM pmem. and let kvm treat these
>> as normal pages.
>> Without this patch, Many operations will be missed due to this
>> mistreatment to pmem pages. For example, a page may not have chance to
>> be unpinned for KVM guest(in kvm_release_pfn_clean); not able to be
>> marked as dirty/accessed(in kvm_set_pfn_dirty/accessed) etc.
> I am right now looking into (and trying to better document) PG_reserved
> - and having a hard time :) .
> One of the main points about reserved pages is that the struct pages are
> not to be touched. See [1] (I know that statement is fairly old, but it
> resembles what PG_reserved is actually used for nowadays - with some
> exceptions unfortunately.).
> Struct pages part of user space tables that are PG_reserved can indicate
> (as of now according to my research)
> - MMIO pages
> - Selected MMAPed pages - e.g. vDSO
> - Zero page
> - PMEM pages as you correctly state
> So I wonder, if it is really the right approach to silently go ahead and
> treat reserved pages just like they would not be reserved. Maybe the
> right approach would rather be to do something about pmem pages being
> reserved. Yes, they are never to be given to the page allocator, but I
> wonder if PG_reserved is strictly needed for that.
> [1]

Thanks David list the long history of Page reserved, By now, I think we treat nvdimm as a device not a DRAM, also has it's device driver which manager its own device memory. From this perspective, it is reasonable to set these pages as zone device memory and mark reserved flag.
@Dan @Dave, how do you think about this?

>> V1:
>> V2:
>> V3:
>> [PATCH V3 1/4] Needs Comments.
>> [PATCH V3 2/4] Update the description of MEMORY_DEVICE_DEV_DAX: Jan
>> [PATCH V3 3/4] Acked-by: Jan in V2
>> [PATCH V3 4/4] Needs Comments.
>> Zhang Yi (4):
>> kvm: remove redundant reserved page check
>> mm: introduce memory type MEMORY_DEVICE_DEV_DAX
>> mm: add a function to differentiate the pages is from DAX device
>> memory
>> kvm: add a check if pfn is from NVDIMM pmem.
>> drivers/dax/pmem.c | 1 +
>> include/linux/memremap.h | 8 ++++++++
>> include/linux/mm.h | 12 ++++++++++++
>> virt/kvm/kvm_main.c | 16 ++++++++--------
>> 4 files changed, 29 insertions(+), 8 deletions(-)

 \ /
  Last update: 2018-08-13 11:42    [W:0.159 / U:0.336 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site