For example, allow:
ndctl list -b ndbus0
...in addition to the existing provider based method:
ndctl list -b ACPI.NFIT
...or the id based method:
ndctl list -b 0
Cc: Vishal Verma
Reported-by: Dave Jiang
Signed-off-by: Dan Williams
Hi:
I am trying to debug an issue related to my private built kernel's (4.11
from upstream with ACPI/NFIT/NVDIMM enabled) interaction with a persistent
memory device's type.
We have a way to emulate a software based NVDIMM device into a guest, which is
running my private built kernel. When
On Wed, Jul 19, 2017 at 11:51:12AM -0600, Ross Zwisler wrote:
> On Wed, Jul 19, 2017 at 04:16:59PM +0200, Jan Kara wrote:
> > On Wed 28-06-17 16:01:48, Ross Zwisler wrote:
> > > To be able to use the common 4k zero page in DAX we need to have our PTE
> > > fault path look more like our PMD fault
On 07/19/2017 02:33 PM, Dan Williams wrote:
> Commit 10663a60d723 "ndctl, list: add '--device-dax' option" tried to
> cleanup the default namespace output by putting the device-dax details
> behind an extra option. This was done in anticipation of device-dax
> sub-division support, but the kernel
If libnvdimm registers namespace0.0 and namespace0.1 as
device-dax-regions there will be 2 entries in /sys/class/dax, dax0.0 and
dax0.1. However, libdaxctl does not account for the fact that these two
devices are hosted by 2 separate parent devices. Fix this assumption and
create separate regions
Add a simple sanity check for multiple dax instances per
libnvdimm-region support. This is known to fail with a sysfs duplicate
filename warning on kernels prior to 4.13-final.
Signed-off-by: Dan Williams
---
test/Makefile.am |1 +
test/multi-dax.sh | 61
Commit 10663a60d723 "ndctl, list: add '--device-dax' option" tried to
cleanup the default namespace output by putting the device-dax details
behind an extra option. This was done in anticipation of device-dax
sub-division support, but the kernel has since added support for
multiple-pmem namespaces
The region id is meant to uniquely identify a host memory range, but in
the case of the libnvdimm namespace the host memory range may be
sub-divided into multiple namespaces leading to ambiguous output like
the following
# daxctl list -R
[
{
"id":1,
If someone turns on libndctl debug they automatically get libdaxctl
debug output from the internal library instance.
Signed-off-by: Dan Williams
---
ndctl/lib/libndctl.c |2 ++
1 file changed, 2 insertions(+)
diff --git a/ndctl/lib/libndctl.c
Clean up region_path to carry the region device path directly. This
aligns libdaxctl with libndctl internal path names, and is a preparation
for carrying absolute path names that are needed for multi-device-dax
per region support.
Signed-off-by: Dan Williams
---
A test report highlighted the fact that the unit tests never verified
that multiple-namespaces in a region could be in device-dax mode. This
was fixed in the kernel [1], but there is some fallout to handle in
ndctl and daxctl:
1/ teach libdaxctl that there may be multiple dax-region instances
Fix warnings of the form...
WARNING: CPU: 10 PID: 4983 at fs/sysfs/dir.c:31 sysfs_warn_dup+0x62/0x80
sysfs: cannot create duplicate filename '/class/dax/dax12.0'
Call Trace:
dump_stack+0x63/0x86
__warn+0xcb/0xf0
warn_slowpath_fmt+0x5a/0x80
?
On Wed, Jul 19, 2017 at 04:16:59PM +0200, Jan Kara wrote:
> On Wed 28-06-17 16:01:48, Ross Zwisler wrote:
> > To be able to use the common 4k zero page in DAX we need to have our PTE
> > fault path look more like our PMD fault path where a PTE entry can be
> > marked as dirty and writeable as it
On Wed, Jul 19, 2017 at 05:33:14PM +0200, Jan Kara wrote:
> On Wed 28-06-17 16:01:50, Ross Zwisler wrote:
> > Another major change is that we remove dax_pfn_mkwrite() from our fault
> > flow, and instead rely on the page fault itself to make the PTE dirty and
> > writeable. The following
On Wed 28-06-17 16:01:50, Ross Zwisler wrote:
> Another major change is that we remove dax_pfn_mkwrite() from our fault
> flow, and instead rely on the page fault itself to make the PTE dirty and
> writeable. The following description from the patch adding the
> vm_insert_mixed_mkwrite() call
On 07/19/2017 02:48 AM, Bob Liu wrote:
>> Option 2: Provide the user with HMAT performance data directly in
>> sysfs, allowing applications to directly access it without the need
>> for the library and daemon.
>>
> Is it possible to do the memory allocation automatically by the
> kernel and
On Wed 28-06-17 16:01:48, Ross Zwisler wrote:
> To be able to use the common 4k zero page in DAX we need to have our PTE
> fault path look more like our PMD fault path where a PTE entry can be
> marked as dirty and writeable as it is first inserted, rather than waiting
> for a follow-up
On 2017/7/7 5:52, Ross Zwisler wrote:
> Quick Summary
>
> Platforms in the very near future will have multiple types of memory
> attached to a single CPU. These disparate memory ranges will have some
> characteristics in common, such as CPU cache coherence, but they can have
> wide
This Message was undeliverable due to the following reason:
Your message was not delivered because the destination computer was
not reachable within the allowed queue period. The amount of time
a message is queued before it is returned depends on local configura-
tion parameters.
Most likely
19 matches
Mail list logo