In order to align with QEMU ,four more parameters about NVDIMM will
be introduced into Libvirt xml.
1.alignsize
The 'alignsize' option allows users to specify the proper alignment. When
mmap(2) the backend files, QEMU uses the host page size by default as
the alignment of mapping address.
Four new parameters were introduced into libvirt xml, including
'align', 'pmem', 'persistence' and 'unarmed', which are related to
NVDIMM memory device. So we need parse and format the xml to save
these configurations.Besides, more testcases related to these
parameters were added to verify the
According to the result parsing from xml, add corresponding properties
into QEMU command line, including 'align', 'pmem', 'persistence' and
'nvdimm-persistence'. And add testcases related to these properties.
Signed-off-by: Zhong,Luyao
---
src/qemu/qemu_command.c| 25
' properties in QEMU, and update xml parsing, formating and
qemu command-line generating process for NVDIMM.
Thanks,
Zhong, Luyao
Luyao Zhong (3):
xml: introduce more config elements for NVDIMM memory
xml: update xml parsing and formating about NVDIMM memory
qemu: update qemu command-line generating
Hi Han,
I'm not sure which release my patches will merge into. How about adding
the patch to update the release news after my last version of these
patches. Waiting for more reviews and comments.
Regards,
Luyao Zhong
On 2018/10/18 上午9:10, Han Han wrote:
On Wed, Oct 17, 2018 at 10:25 AM
polite ping
On 2018/10/17 上午10:21, Luyao Zhong wrote:
Hi libvirt experts,
This is the RFC for updating NVDIMM support in libvirt.
QEMU has supported four more properties which libvirt has not introduced
yet, including 'align', 'pmem', 'nvdimm-persistences' and 'unarmed'.
The 'align' property
original:
...
...
merge two examples above into one:
...
Signed-off-by: Luyao Zhong
---
docs/formatdomain.html.in | 14 +-
1 file changed, 1 insertion(+), 13 deletions(-)
diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in
index 7f07bb7
On 2019/1/16 下午4:16, Peter Krempa wrote:
On Wed, Jan 16, 2019 at 14:58:21 +0800, Luyao Zhong wrote:
original:
...
...
merge two examples above into one:
...
Signed-off-by: Luyao Zhong
What's the benefit of the above change? We have multiple instances
On 2018/12/17 下午7:05, Daniel P. Berrangé wrote:
On Mon, Dec 17, 2018 at 05:26:12PM +0800, Luyao Zhong wrote:
The 'pmem' option allows users to specify whether the backend
storage of memory-backend-file is a real persistent memory:
What does this actually do in practice ? Why does it matter
On 2018/12/17 下午7:07, Daniel P. Berrangé wrote:
On Mon, Dec 17, 2018 at 05:26:13PM +0800, Luyao Zhong wrote:
The 'unarmed' option allows users to mark vNVDIMM read-only:
Why don't we call this element then ?
Just keep the same with QEMU, hah~ :)
...
/dev
On 2018/12/21 上午1:10, John Ferlan wrote:
$SUBJ:
docs: Add news for new nvdimm options
On 12/20/18 4:14 AM, Luyao Zhong wrote:
add more configure options for NVDIMM
Signed-off-by: Luyao Zhong
---
docs/news.xml | 9 +
1 file changed, 9 insertions(+)
diff --git a/docs/news.xml b
According to the result parsing from xml, add align property
into QEMU command line.
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_command.c| 14 +
.../memory-hotplug-nvdimm-align.x86_64-latest.args | 36 ++
tests/qemuxml2argvtest.c
According to the result parsing from xml, add pmem property
into QEMU command line.
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_command.c| 14 +
.../memory-hotplug-nvdimm-pmem.x86_64-latest.args | 36 ++
tests/qemuxml2argvtest.c
According to the result parsing from xml, add unarmed property
into QEMU command line.
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_command.c| 16 --
src/qemu/qemu_command.h| 3 +-
src/qemu/qemu_hotplug.c
This capability tracks if nvdimm has unarmed attribute or not.
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_capabilities.c | 8 ++
src/qemu/qemu_capabilities.h | 1 +
.../caps_2.10.0.x86_64.replies | 62 -
tests
add more configure options for NVDIMM
Signed-off-by: Luyao Zhong
---
docs/news.xml | 9 +
1 file changed, 9 insertions(+)
diff --git a/docs/news.xml b/docs/news.xml
index 5bdbd34..dde98a1 100644
--- a/docs/news.xml
+++ b/docs/news.xml
@@ -65,6 +65,15 @@
qemu: Add support
:
...
/dev/dax0.0
2
4094
0
2
...
Signed-off-by: Luyao Zhong
---
docs/formatdomain.html.in | 23 +++--
docs/schemas/domaincommon.rng | 13 +++--
src/conf
The 'pmem' option allows users to specify whether the backend
storage of memory-backend-file is a real persistent memory:
...
/dev/dax0.0
4094
0
2
...
Signed-off-by: Luyao Zhong
This capability tracks if memory-backend-file has align
attribute or not.
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_capabilities.c | 4 +++-
src/qemu/qemu_capabilities.h | 1 +
tests/qemucapabilitiesdata/caps_2.12.0.aarch64.xml | 1 +
tests
Depracate DO_TEST to do nvdimm qemuxml2argvdata tests, because
DO_TEST_CAPS_LATEST is a better choice. the DO_TEST need
to specify all qemu capabilities and is not easy for scaling.
Signed-off-by: Luyao Zhong
---
.../memory-hotplug-nvdimm-access.args | 31
Hi libvirt experts,
There are some gaps between qemu and libvirt, libvirt has not
supported several config options about NVDIMM memory while
qemu has now, including 'align', 'pmem', 'unarmed'.
So these three options is introduced in this patch set.
Regards,
Luyao Zhong
Luyao Zhong (11
The 'unarmed' option allows users to mark vNVDIMM read-only:
...
/dev/dax0.0
4094
0
2
...
Signed-off-by: Luyao Zhong
---
docs/formatdomain.html.in | 45
This capability tracks if memory-backend-file has pmem
attribute or not.
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_capabilities.c | 4
src/qemu/qemu_capabilities.h | 3 +++
tests/qemucapabilitiesdata/caps_3.1.0.ppc64.xml | 1 +
tests
On 2018/12/14 上午9:06, John Ferlan wrote:
No need to CC developers for libvirt, we're all subscribed to the list
anyways and generally are faithful in reading. Reviews are a different
story.
On 12/12/18 7:52 AM, Luyao Zhong wrote:
Hi libvirt experts,
This is the RFC v3 for updating NVDIMM
On 2018/12/14 上午9:07, John Ferlan wrote:
On 12/12/18 7:52 AM, Luyao Zhong wrote:
1.alignsize
The 'alignsize' option allows users to specify the proper alignment.
2.pmem
The 'pmem' option allows users to specify whether the backend storage of
memory-backend-file is a real persistent memory
On 2018/12/14 上午9:10, John Ferlan wrote:
On 12/12/18 7:53 AM, Luyao Zhong wrote:
According to the result parsing from xml, add corresponding properties
into QEMU command line, including 'align', 'pmem' and 'unarmed'.
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_command.c
On 2018/12/14 上午9:09, John Ferlan wrote:
On 12/12/18 7:52 AM, Luyao Zhong wrote:
-object memory-backend-file[,align=][,pmem=]
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_capabilities.c | 8 +++-
src/qemu/qemu_capabilities.h | 4
On 2018/12/14 上午10:17, John Ferlan wrote:
On 12/12/18 7:53 AM, Luyao Zhong wrote:
add more configure options for NVDIMM
Signed-off-by: Luyao Zhong
---
docs/news.xml | 9 +
1 file changed, 9 insertions(+)
It will be OK to just have one of these assuming each of the options
On 2018/12/20 下午5:14, Luyao Zhong wrote:
According to the result parsing from xml, add unarmed property
into QEMU command line:
-device nvdimm,...[,unarmed=on/off]
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_command.c| 16 --
src/qemu/qemu_command.h
This capability tracks if nvdimm has unarmed attribute or not.
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_capabilities.c | 8 ++
src/qemu/qemu_capabilities.h | 1 +
.../caps_2.10.0.x86_64.replies | 62 -
tests
Depracate DO_TEST to do nvdimm qemuxml2argvdata tests, because
DO_TEST_CAPS_LATEST is a better choice. the DO_TEST need
to specify all qemu capabilities and is not easy for scaling.
Signed-off-by: Luyao Zhong
---
.../memory-hotplug-nvdimm-access.args | 31
:
...
/dev/dax0.0
2
4094
0
2
...
Signed-off-by: Luyao Zhong
---
docs/formatdomain.html.in | 23 +++--
docs/schemas/domaincommon.rng | 13 +++--
src/conf
According to the result parsing from xml, add align property
into QEMU command line:
-object memory-backend-file,...[,align=xxx]
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_command.c| 14 +
.../memory-hotplug-nvdimm-align.x86_64-latest.args | 36
The 'pmem' option allows users to specify whether the backend
storage of memory-backend-file is a real persistent memory:
...
/dev/dax0.0
4094
0
2
...
Signed-off-by: Luyao Zhong
This capability tracks if memory-backend-file has align
attribute or not.
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_capabilities.c | 4 +++-
src/qemu/qemu_capabilities.h | 1 +
tests/qemucapabilitiesdata/caps_2.12.0.aarch64.xml | 1 +
tests
This capability tracks if memory-backend-file has pmem
attribute or not.
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_capabilities.c | 4
src/qemu/qemu_capabilities.h | 3 +++
tests/qemucapabilitiesdata/caps_3.1.0.ppc64.xml | 1 +
tests
to 'alignsize' and 'readonly'
in domain xml file for user-friendly definition.
Regards,
Luyao Zhong
Luyao Zhong (11):
nvdimm: introduce 'alignsize' element into xml for NVDIMM memory
nvdimm: introduce 'pmem' element into xml for NVDIMM memory
nvdimm: introduce 'readonly' element into xml for NVDIMM
The 'readonly' option allows users to mark vNVDIMM read-only:
...
/dev/dax0.0
4094
0
2
...
Signed-off-by: Luyao Zhong
---
docs/formatdomain.html.in | 45
add more configure options for NVDIMM
Signed-off-by: Luyao Zhong
---
docs/news.xml | 9 +
1 file changed, 9 insertions(+)
diff --git a/docs/news.xml b/docs/news.xml
index e9b6bb7..2076cd5 100644
--- a/docs/news.xml
+++ b/docs/news.xml
@@ -79,6 +79,15 @@
qemu: Add support
According to the result parsing from xml, add unarmed property
into QEMU command line:
-device nvdimm,...[,unarmed=on/off]
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_command.c| 16 --
src/qemu/qemu_command.h| 3 +-
src/qemu
According to the result parsing from xml, add pmem property
into QEMU command line:
-object memory-backend-file,...[,pmem=on/off]
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_command.c| 14 +
.../memory-hotplug-nvdimm-pmem.x86_64-latest.args | 36
add more configure options for NVDIMM
Signed-off-by: Luyao Zhong
---
docs/news.xml | 9 +
1 file changed, 9 insertions(+)
diff --git a/docs/news.xml b/docs/news.xml
index 5bdbd34..0112c91 100644
--- a/docs/news.xml
+++ b/docs/news.xml
@@ -38,6 +38,15 @@
would normally
-object memory-backend-file[,align=][,pmem=]
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_capabilities.c | 8 +++-
src/qemu/qemu_capabilities.h | 4
tests/qemucapabilitiesdata/caps_2.12.0.aarch64.xml | 1 +
tests/qemucapabilitiesdata
can be configured respectively or simultaneously in domain
xml file, here is an example:
...
/dev/dax0.0
2
4094
0
2
...
Signed-off-by: Luyao Zhong
---
docs
ards,
Luyao Zhong
Luyao Zhong (4):
nvdimm: introduce more config elements into xml for NVDIMM memory
nvdimm: add nvdimm-related qemucapabilities check
nvdimm: update qemu command-line generating for NVDIMM memory
nvdimm: update news.xml
docs/formatdomain.html.in
According to the result parsing from xml, add corresponding properties
into QEMU command line, including 'align', 'pmem' and 'unarmed'.
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_command.c| 32 ++
.../memory-hotplug-nvdimm-align.args
On 2018/11/29 下午4:52, Peter Krempa wrote:
On Thu, Nov 29, 2018 at 12:08:58 +0800, Luyao Zhong wrote:
On 2018/11/28 下午10:32, Peter Krempa wrote:
On Wed, Nov 28, 2018 at 22:09:01 +0800, Luyao Zhong wrote:
According to the result parsing from xml, add corresponding properties
into QEMU command
On 2018/11/28 下午10:32, Peter Krempa wrote:
On Wed, Nov 28, 2018 at 22:09:01 +0800, Luyao Zhong wrote:
According to the result parsing from xml, add corresponding properties
into QEMU command line, including 'align', 'pmem', 'unarmed' and
'nvdimm-persistence'.
Signed-off-by: Luyao Zhong
On 2018/11/28 下午10:33, Peter Krempa wrote:
On Wed, Nov 28, 2018 at 22:09:19 +0800, Luyao Zhong wrote:
Update news.xml and add description about new supported configurations
about NVDIMM in html, including 'alignsize', 'pmem', 'unarmed' and
'persistence'.
Signed-off-by: Luyao Zhong
options here. I'd like
you give some suggestions.
Thank you in advance.
Regards,
Luyao Zhong
Luyao Zhong (3):
nvdimm: introduce more config elements into xml for NVDIMM memory
nvdimm: update qemu command-line generating for NVDIMM memory
nvdimm: update docs for introducing more NVDIMM
According to the result parsing from xml, add corresponding properties
into QEMU command line, including 'align', 'pmem', 'unarmed' and
'nvdimm-persistence'.
Signed-off-by: Luyao Zhong
---
src/qemu/qemu_capabilities.c | 17 +++
src/qemu/qemu_capabilities.h
4094
0
2
cpu
...
Signed-off-by: Luyao Zhong
---
docs/schemas/domaincommon.rng | 31 +-
src/conf/domain_conf.c| 94 +--
src/conf/domain_conf.h
Update news.xml and add description about new supported configurations
about NVDIMM in html, including 'alignsize', 'pmem', 'unarmed' and
'persistence'.
Signed-off-by: Luyao Zhong
---
docs/formatdomain.html.in | 94 ---
docs/news.xml | 9
2
The NVDIMM backend file can be a normal file or a real device file,
Current xml example and explainations may mislead users. So add more
info about the NVDIMM related elements and update the xml examples.
Signed-off-by: Luyao Zhong
---
docs/formatdomain.html.in | 9 ++---
1 file changed, 6
Hi libvirt experts,
I have some questions about live migration.
* If a live migration failed during migrating, will the domain exist on the
destination host?
* Is the flag VIR_MIGRATE_PAUSED make sense to live migration? It's a little
confusing for me. Does that indicate if I set this flag,
Attribute ``migratable`` will be 'no' by default, and 'yes' indicates
that it allows operating system or hypervisor migrating the memory
pages between different memory nodes, that also means we will not
rely on hypervisor to set the memory policy or memory affinity, we only
RFC discussion record link:
https://www.redhat.com/archives/libvir-list/2020-August/msg00960.html
Luyao Zhong (1):
introduce an attribute "migratable" to numatune memory element
docs/formatdomain.rst | 8 +++-
docs/schemas/domaincommon.rng
Hi Libvirt experts,
I would like enhence the numatune snippet configuration. Given a example
snippet:
...
...
Currently, attribute mode is either 'interleave', 'strict', or 'preferred',
I propose to add a new 'default' option. I give the reason as following.
---
include/libvirt/libvirt-domain.h | 1 +
src/conf/numa_conf.c | 9 +
src/qemu/qemu_command.c | 6 ++-
src/qemu/qemu_process.c | 27 +
src/util/virnuma.c| 3 ++
and not setting any specific memory policies explicitly.
RFC discussion:
https://www.redhat.com/archives/libvir-list/2020-November/msg01256.html
Regards,
Luyao
Luyao Zhong (3):
docs: add docs for 'restrictive' option for mode in numatune
schema: add 'restrictive' config option for mode
When user would like use cgroups to restrict the allowed memory
nodes, and require not setting any specific memory policy, then
'restrictive' mode is useful.
---
docs/formatdomain.rst | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/docs/formatdomain.rst
support 'restrictive' mode in memory element and memnode
element in numatune:
...
...
---
docs/schemas/domaincommon.rng | 2 ++
1 file changed, 2 insertions(+)
diff --git a/docs/schemas/domaincommon.rng b/docs/schemas/domaincommon.rng
index
support 'restrictive' mode in memory element and memnode
element in numatune:
...
...
---
docs/schemas/domaincommon.rng | 2 ++
1 file changed, 2 insertions(+)
diff --git a/docs/schemas/domaincommon.rng b/docs/schemas/domaincommon.rng
index
and not setting any specific memory policies explicitly.
RFC discussion:
https://www.redhat.com/archives/libvir-list/2020-November/msg01256.html
Regards,
Luyao
Luyao Zhong (3):
docs: add docs for 'restrictive' option for mode in numatune
schema: add 'restrictive' config option for mode
When user would like use cgroups to restrict the allowed memory
nodes, and require not setting any specific memory policy, then
'restrictive' mode is useful.
---
docs/formatdomain.rst | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/docs/formatdomain.rst
---
include/libvirt/libvirt-domain.h | 1 +
src/conf/numa_conf.c | 9 +
src/qemu/qemu_command.c | 6 ++-
src/qemu/qemu_process.c | 27 +
src/util/virnuma.c| 3 ++
nks for your patience to read this. Appreciate if you could give comments.
Regards,
Luyao
Luyao Zhong (1):
support system default memory policy with numatune
docs/formatdomain.rst | 12 -
docs/schemas/domaincommon.rng | 7 +++
include/libvir
This patch seeks the support of system default memory policy
when numatune is configured. Before this patch, numatune only
has three memory modes: static, interleave and prefered.
These memory policies are ultimately set by mbind() system call.
Memory policy could be 'hard coded' into the kernel,
When user would like use cgroups to restrict the allowed memory
nodes, and require not setting any specific memory policy, then
'restrictive' mode is useful.
Reviewed-by: Daniel Henrique Barboza
---
docs/formatdomain.rst | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git
Reviewed-by: Daniel Henrique Barboza
---
include/libvirt/libvirt-domain.h | 1 +
src/conf/numa_conf.c | 9 +
src/qemu/qemu_command.c | 6 ++-
src/qemu/qemu_process.c | 27 +
src/util/virnuma.c
and not setting any specific memory policies explicitly.
RFC discussion:
https://www.redhat.com/archives/libvir-list/2020-November/msg01256.html
Regards,
Luyao
Luyao Zhong (3):
docs: add docs for 'restrictive' option for mode in numatune
schema: add 'restrictive' config option for mode
support 'restrictive' mode in memory element and memnode
element in numatune:
...
...
Reviewed-by: Daniel Henrique Barboza
---
docs/schemas/domaincommon.rng | 2 ++
1 file changed, 2 insertions(+)
diff --git a/docs/schemas/domaincommon.rng
Reviewed-by: Daniel Henrique Barboza
Signed-off-by: Luyao Zhong
---
include/libvirt/libvirt-domain.h | 1 +
src/conf/numa_conf.c | 9 +
src/qemu/qemu_command.c | 6 ++-
src/qemu/qemu_process.c | 27
When user would like use cgroups to restrict the allowed memory
nodes, and require not setting any specific memory policy, then
'restrictive' mode is useful.
Reviewed-by: Daniel Henrique Barboza
Signed-off-by: Luyao Zhong
---
docs/formatdomain.rst | 7 +--
1 file changed, 5 insertions
and not setting any specific memory policies explicitly.
RFC discussion:
https://www.redhat.com/archives/libvir-list/2020-November/msg01256.html
Regards,
Luyao
Luyao Zhong (3):
docs: add docs for 'restrictive' option for mode in numatune
schema: add 'restrictive' config option for mode
support 'restrictive' mode in memory element and memnode
element in numatune:
...
...
Reviewed-by: Daniel Henrique Barboza
Signed-off-by: Luyao Zhong
---
docs/schemas/domaincommon.rng | 2 ++
1 file changed, 2 insertions(+)
diff --git a/docs
Reviewed-by: Daniel Henrique Barboza
Signed-off-by: Luyao Zhong
---
include/libvirt/libvirt-domain.h | 1 +
src/conf/numa_conf.c | 9
src/qemu/qemu_command.c | 6 ++-
src/qemu/qemu_process.c | 27
and not setting any specific memory policies explicitly.
RFC discussion:
https://www.redhat.com/archives/libvir-list/2020-November/msg01256.html
Regards,
Luyao
Luyao Zhong (3):
docs: add docs for 'restrictive' option for mode in numatune
schema: add 'restrictive' config option for mode
support 'restrictive' mode in memory element and memnode
element in numatune:
...
...
Reviewed-by: Daniel Henrique Barboza
Signed-off-by: Luyao Zhong
---
docs/schemas/domaincommon.rng | 2 ++
1 file changed, 2 insertions(+)
diff --git a/docs
When user would like use cgroups to restrict the allowed memory
nodes, and require not setting any specific memory policy, then
'restrictive' mode is useful.
Reviewed-by: Daniel Henrique Barboza
Signed-off-by: Luyao Zhong
---
docs/formatdomain.rst | 7 +--
1 file changed, 5 insertions
support 'restrictive' mode in memory element and memnode
element in numatune:
...
...
Reviewed-by: Daniel Henrique Barboza
Signed-off-by: Luyao Zhong
---
docs/schemas/domaincommon.rng | 2 ++
1 file changed, 2 insertions(+)
diff --git a/docs
When user would like use cgroups to restrict the allowed memory
nodes, and require not setting any specific memory policy, then
'restrictive' mode is useful.
Reviewed-by: Daniel Henrique Barboza
Signed-off-by: Luyao Zhong
---
docs/formatdomain.rst | 7 +--
1 file changed, 5 insertions
Reviewed-by: Daniel Henrique Barboza
Signed-off-by: Luyao Zhong
---
include/libvirt/libvirt-domain.h | 1 +
src/conf/numa_conf.c | 9
src/qemu/qemu_command.c | 6 ++-
src/qemu/qemu_process.c | 27
it is full,
then on the next one and so on).
In a word, if a user requires default mode(MPOL_DEFAULT), that means they want
kernel decide the memory allocation and also want the cgroups to restrict memory
nodes, "restrictive" mode will be useful.
BR,
Luyao
Luyao Zhong (3):
docs: add docs for
84 matches
Mail list logo
