[v3,0/5] Add ignore-external migration capability

Message

Yury Kotov Feb. 15, 2019, 5:45 p.m. UTC

Hi,

The series adds a migration capability, which allows to skip shared RAM blocks
during the migration. It's useful for fast local migration. E.g. to update QEMU
for the running guests.

Usage example:
1. Start source VM:
   qemu-system-x86 \
     -m 4G \
     -object memory-backend-file,id=mem0,size=4G,share=on,mem-path=/dev/shm/mem0 \
     -numa node,memdev=mem0 \
     -qmp unix:/tmp/qemu-qmp-1.sock,server,nowait \

2. Start target VM:
   qemu-system-x86 \
     -m 4G \
     -object memory-backend-file,id=mem0,size=4G,share=on,mem-path=/dev/shm/mem0 \
     -numa node,memdev=mem0 \
     -qmp unix:/tmp/qemu-qmp-2.sock,server,nowait \
     -incoming defer

3. Enable ignore-shared capability on both VMs:
   { "execute": "migrate-set-capabilities" , "arguments":
     { "capabilities": [ { "capability": "x-ignore-shared", "state": true } ] } }

4. Start migration.

Another use case I keep in mind is to migrate to file. Usage is very similar.

V2 to V3:
* Split "migration: Introduce ignore-shared capability"
* Serialize the capabilities as strings rather than as indexes
* Don't allow to enable postcopy and ignore-shared together
* Skip the test for OSs which don't have /dev/shm
* Add a check whether shared RAM has been really skipped

V1 to V2:
* Keep migration stream compatibility
* Reuse the existing code to ignore unwanted RAMBlocks
* Add capability validation feature
* ignore-external -> ignore-shared

Regards,
Yury

Yury Kotov (5):
  exec: Change RAMBlockIterFunc definition
  migration: Introduce ignore-shared capability
  migration: Add an ability to ignore shared RAM blocks
  tests/migration-test: Add a test for ignore-shared capability
  migration: Add capabilities validation

 exec.c                    |  38 +++++------
 include/exec/cpu-common.h |   7 +-
 migration/migration.c     |  14 ++++
 migration/migration.h     |   5 +-
 migration/postcopy-ram.c  |  48 +++++++------
 migration/ram.c           | 110 ++++++++++++++++++++++--------
 migration/rdma.c          |   9 ++-
 migration/savevm.c        | 137 ++++++++++++++++++++++++++++++++++++++
 qapi/migration.json       |   5 +-
 stubs/ram-block.c         |  15 +++++
 tests/migration-test.c    | 131 +++++++++++++++++++++++++++++-------
 util/vfio-helpers.c       |   6 +-
 12 files changed, 420 insertions(+), 105 deletions(-)

Comments

[e-mail] Feb. 15, 2019, 6:17 p.m. UTC | #1

Patchew URL: https://patchew.org/QEMU/[e-mail]/



Hi,

This series seems to have some coding style problems. See output below for
more information:

Message-id: [e-mail]
Subject: [Qemu-devel] [PATCH v3 0/5] Add ignore-external migration capability
Type: series

=== TEST SCRIPT BEGIN ===
#!/bin/bash
git config --local diff.renamelimit 0
git config --local diff.renames True
git config --local diff.algorithm histogram
./scripts/checkpatch.pl --mailback base..
=== TEST SCRIPT END ===

Updating 3c8cf5a9c21ff8782164d1def7f44bd888713384
Switched to a new branch 'test'
5f9e90967e migration: Add capabilities validation
fad48088a8 tests/migration-test: Add a test for ignore-shared capability
68347ad2eb migration: Add an ability to ignore shared RAM blocks
e3d2fd0298 migration: Introduce ignore-shared capability
73d9e23937 exec: Change RAMBlockIterFunc definition

=== OUTPUT BEGIN ===
1/5 Checking commit 73d9e23937e9 (exec: Change RAMBlockIterFunc definition)
2/5 Checking commit e3d2fd0298a4 (migration: Introduce ignore-shared capability)
3/5 Checking commit 68347ad2eb24 (migration: Add an ability to ignore shared RAM blocks)
ERROR: Macros with multiple statements should be enclosed in a do - while loop
#146: FILE: migration/ram.c:169:
+#define RAMBLOCK_FOREACH_NOT_IGNORED(block)            \
+    INTERNAL_RAMBLOCK_FOREACH(block)                   \
+        if (ramblock_is_ignored(block)) {} else

ERROR: trailing statements should be on next line
#148: FILE: migration/ram.c:171:
+        if (ramblock_is_ignored(block)) {} else

total: 2 errors, 0 warnings, 386 lines checked

Patch 3/5 has style problems, please review.  If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.

4/5 Checking commit fad48088a805 (tests/migration-test: Add a test for ignore-shared capability)
5/5 Checking commit 5f9e90967e69 (migration: Add capabilities validation)
=== OUTPUT END ===

Test command exited with code: 1


The full log is available at
http://patchew.org/logs/[e-mail]/testing.checkpatch/?type=message.
---
Email generated automatically by Patchew [http://patchew.org/].
Please send your feedback to [e-mail]

[e-mail] Feb. 15, 2019, 6:21 p.m. UTC | #2

Patchew URL: https://patchew.org/QEMU/[e-mail]/



Hi,

This series seems to have some coding style problems. See output below for
more information:

Message-id: [e-mail]
Subject: [Qemu-devel] [PATCH v3 0/5] Add ignore-external migration capability
Type: series

=== TEST SCRIPT BEGIN ===
#!/bin/bash
git config --local diff.renamelimit 0
git config --local diff.renames True
git config --local diff.algorithm histogram
./scripts/checkpatch.pl --mailback base..
=== TEST SCRIPT END ===

Updating 3c8cf5a9c21ff8782164d1def7f44bd888713384
From https://github.com/patchew-project/qemu
 * [new tag]         patchew/[e-mail] -> patchew/[e-mail]
 * [new tag]         patchew/[e-mail] -> patchew/[e-mail]
Submodule 'capstone' (https://git.qemu.org/git/capstone.git) registered for path 'capstone'
Submodule 'dtc' (https://git.qemu.org/git/dtc.git) registered for path 'dtc'
Submodule 'roms/QemuMacDrivers' (https://git.qemu.org/git/QemuMacDrivers.git) registered for path 'roms/QemuMacDrivers'
Submodule 'roms/SLOF' (https://git.qemu.org/git/SLOF.git) registered for path 'roms/SLOF'
Submodule 'roms/ipxe' (https://git.qemu.org/git/ipxe.git) registered for path 'roms/ipxe'
Submodule 'roms/openbios' (https://git.qemu.org/git/openbios.git) registered for path 'roms/openbios'
Submodule 'roms/openhackware' (https://git.qemu.org/git/openhackware.git) registered for path 'roms/openhackware'
Submodule 'roms/qemu-palcode' (https://git.qemu.org/git/qemu-palcode.git) registered for path 'roms/qemu-palcode'
Submodule 'roms/seabios' (https://git.qemu.org/git/seabios.git/) registered for path 'roms/seabios'
Submodule 'roms/seabios-hppa' (https://github.com/hdeller/seabios-hppa.git) registered for path 'roms/seabios-hppa'
Submodule 'roms/sgabios' (https://git.qemu.org/git/sgabios.git) registered for path 'roms/sgabios'
Submodule 'roms/skiboot' (https://git.qemu.org/git/skiboot.git) registered for path 'roms/skiboot'
Submodule 'roms/u-boot' (https://git.qemu.org/git/u-boot.git) registered for path 'roms/u-boot'
Submodule 'roms/u-boot-sam460ex' (https://git.qemu.org/git/u-boot-sam460ex.git) registered for path 'roms/u-boot-sam460ex'
Submodule 'tests/fp/berkeley-softfloat-3' (https://github.com/cota/berkeley-softfloat-3) registered for path 'tests/fp/berkeley-softfloat-3'
Submodule 'tests/fp/berkeley-testfloat-3' (https://github.com/cota/berkeley-testfloat-3) registered for path 'tests/fp/berkeley-testfloat-3'
Submodule 'ui/keycodemapdb' (https://git.qemu.org/git/keycodemapdb.git) registered for path 'ui/keycodemapdb'
Cloning into 'capstone'...
Submodule path 'capstone': checked out '22ead3e0bfdb87516656453336160e0a37b066bf'
Cloning into 'dtc'...
Submodule path 'dtc': checked out '88f18909db731a627456f26d779445f84e449536'
Cloning into 'roms/QemuMacDrivers'...
Submodule path 'roms/QemuMacDrivers': checked out '90c488d5f4a407342247b9ea869df1c2d9c8e266'
Cloning into 'roms/SLOF'...
Submodule path 'roms/SLOF': checked out 'a5b428e1c1eae703bdd62a3f527223c291ee3fdc'
Cloning into 'roms/ipxe'...
Submodule path 'roms/ipxe': checked out 'de4565cbe76ea9f7913a01f331be3ee901bb6e17'
Cloning into 'roms/openbios'...
Submodule path 'roms/openbios': checked out '3464681b2b5983df80086a40179d324102347da3'
Cloning into 'roms/openhackware'...
Submodule path 'roms/openhackware': checked out 'c559da7c8eec5e45ef1f67978827af6f0b9546f5'
Cloning into 'roms/qemu-palcode'...
Submodule path 'roms/qemu-palcode': checked out '51c237d7e20d05100eacadee2f61abc17e6bc097'
Cloning into 'roms/seabios'...
Submodule path 'roms/seabios': checked out 'a698c8995ffb2838296ec284fe3c4ad33dfca307'
Cloning into 'roms/seabios-hppa'...
Submodule path 'roms/seabios-hppa': checked out '1ef99a01572c2581c30e16e6fe69e9ea2ef92ce0'
Cloning into 'roms/sgabios'...
Submodule path 'roms/sgabios': checked out 'cbaee52287e5f32373181cff50a00b6c4ac9015a'
Cloning into 'roms/skiboot'...
Submodule path 'roms/skiboot': checked out 'e0ee24c27a172bcf482f6f2bc905e6211c134bcc'
Cloning into 'roms/u-boot'...
Submodule path 'roms/u-boot': checked out 'd85ca029f257b53a96da6c2fb421e78a003a9943'
Cloning into 'roms/u-boot-sam460ex'...
Submodule path 'roms/u-boot-sam460ex': checked out '60b3916f33e617a815973c5a6df77055b2e3a588'
Cloning into 'tests/fp/berkeley-softfloat-3'...
Submodule path 'tests/fp/berkeley-softfloat-3': checked out 'b64af41c3276f97f0e181920400ee056b9c88037'
Cloning into 'tests/fp/berkeley-testfloat-3'...
Submodule path 'tests/fp/berkeley-testfloat-3': checked out '5a59dcec19327396a011a17fd924aed4fec416b3'
Cloning into 'ui/keycodemapdb'...
Submodule path 'ui/keycodemapdb': checked out '6b3d716e2b6472eb7189d3220552280ef3d832ce'
Switched to a new branch 'test'
5f9e909 migration: Add capabilities validation
fad4808 tests/migration-test: Add a test for ignore-shared capability
68347ad migration: Add an ability to ignore shared RAM blocks
e3d2fd0 migration: Introduce ignore-shared capability
73d9e23 exec: Change RAMBlockIterFunc definition

=== OUTPUT BEGIN ===
1/5 Checking commit 73d9e23937e9 (exec: Change RAMBlockIterFunc definition)
2/5 Checking commit e3d2fd0298a4 (migration: Introduce ignore-shared capability)
3/5 Checking commit 68347ad2eb24 (migration: Add an ability to ignore shared RAM blocks)
ERROR: Macros with multiple statements should be enclosed in a do - while loop
#146: FILE: migration/ram.c:169:
+#define RAMBLOCK_FOREACH_NOT_IGNORED(block)            \
+    INTERNAL_RAMBLOCK_FOREACH(block)                   \
+        if (ramblock_is_ignored(block)) {} else

ERROR: trailing statements should be on next line
#148: FILE: migration/ram.c:171:
+        if (ramblock_is_ignored(block)) {} else

total: 2 errors, 0 warnings, 386 lines checked

Patch 3/5 has style problems, please review.  If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.

4/5 Checking commit fad48088a805 (tests/migration-test: Add a test for ignore-shared capability)
5/5 Checking commit 5f9e90967e69 (migration: Add capabilities validation)
=== OUTPUT END ===

Test command exited with code: 1


The full log is available at
http://patchew.org/logs/[e-mail]/testing.checkpatch/?type=message.
---
Email generated automatically by Patchew [http://patchew.org/].
Please send your feedback to [e-mail]

Dr. David Alan Gilbert Feb. 20, 2019, 5:42 p.m. UTC | #3

* Yury Kotov ([e-mail]) wrote:
> Hi,
> 
> The series adds a migration capability, which allows to skip shared RAM blocks
> during the migration. It's useful for fast local migration. E.g. to update QEMU
> for the running guests.

Other than the title of this cover letter being out of date, I think
we're about OK!

Dave

> Usage example:
> 1. Start source VM:
>    qemu-system-x86 \
>      -m 4G \
>      -object memory-backend-file,id=mem0,size=4G,share=on,mem-path=/dev/shm/mem0 \
>      -numa node,memdev=mem0 \
>      -qmp unix:/tmp/qemu-qmp-1.sock,server,nowait \
> 
> 2. Start target VM:
>    qemu-system-x86 \
>      -m 4G \
>      -object memory-backend-file,id=mem0,size=4G,share=on,mem-path=/dev/shm/mem0 \
>      -numa node,memdev=mem0 \
>      -qmp unix:/tmp/qemu-qmp-2.sock,server,nowait \
>      -incoming defer
> 
> 3. Enable ignore-shared capability on both VMs:
>    { "execute": "migrate-set-capabilities" , "arguments":
>      { "capabilities": [ { "capability": "x-ignore-shared", "state": true } ] } }
> 
> 4. Start migration.
> 
> Another use case I keep in mind is to migrate to file. Usage is very similar.
> 
> V2 to V3:
> * Split "migration: Introduce ignore-shared capability"
> * Serialize the capabilities as strings rather than as indexes
> * Don't allow to enable postcopy and ignore-shared together
> * Skip the test for OSs which don't have /dev/shm
> * Add a check whether shared RAM has been really skipped
> 
> V1 to V2:
> * Keep migration stream compatibility
> * Reuse the existing code to ignore unwanted RAMBlocks
> * Add capability validation feature
> * ignore-external -> ignore-shared
> 
> Regards,
> Yury
> 
> Yury Kotov (5):
>   exec: Change RAMBlockIterFunc definition
>   migration: Introduce ignore-shared capability
>   migration: Add an ability to ignore shared RAM blocks
>   tests/migration-test: Add a test for ignore-shared capability
>   migration: Add capabilities validation
> 
>  exec.c                    |  38 +++++------
>  include/exec/cpu-common.h |   7 +-
>  migration/migration.c     |  14 ++++
>  migration/migration.h     |   5 +-
>  migration/postcopy-ram.c  |  48 +++++++------
>  migration/ram.c           | 110 ++++++++++++++++++++++--------
>  migration/rdma.c          |   9 ++-
>  migration/savevm.c        | 137 ++++++++++++++++++++++++++++++++++++++
>  qapi/migration.json       |   5 +-
>  stubs/ram-block.c         |  15 +++++
>  tests/migration-test.c    | 131 +++++++++++++++++++++++++++++-------
>  util/vfio-helpers.c       |   6 +-
>  12 files changed, 420 insertions(+), 105 deletions(-)
> 
> -- 
> 2.20.1
> 
--
Dr. David Alan Gilbert / [e-mail] / Manchester, UK

Dr. David Alan Gilbert Feb. 20, 2019, 5:43 p.m. UTC | #4

* [e-mail] ([e-mail]) wrote:
> Patchew URL: https://patchew.org/QEMU/[e-mail]/
> 
> 
> 
> Hi,
> 
> This series seems to have some coding style problems. See output below for
> more information:
> 
> Message-id: [e-mail]
> Subject: [Qemu-devel] [PATCH v3 0/5] Add ignore-external migration capability
> Type: series
> 
> === TEST SCRIPT BEGIN ===
> #!/bin/bash
> git config --local diff.renamelimit 0
> git config --local diff.renames True
> git config --local diff.algorithm histogram
> ./scripts/checkpatch.pl --mailback base..
> === TEST SCRIPT END ===
> 
> Updating 3c8cf5a9c21ff8782164d1def7f44bd888713384
> Switched to a new branch 'test'
> 5f9e90967e migration: Add capabilities validation
> fad48088a8 tests/migration-test: Add a test for ignore-shared capability
> 68347ad2eb migration: Add an ability to ignore shared RAM blocks
> e3d2fd0298 migration: Introduce ignore-shared capability
> 73d9e23937 exec: Change RAMBlockIterFunc definition
> 
> === OUTPUT BEGIN ===
> 1/5 Checking commit 73d9e23937e9 (exec: Change RAMBlockIterFunc definition)
> 2/5 Checking commit e3d2fd0298a4 (migration: Introduce ignore-shared capability)
> 3/5 Checking commit 68347ad2eb24 (migration: Add an ability to ignore shared RAM blocks)
> ERROR: Macros with multiple statements should be enclosed in a do - while loop
> #146: FILE: migration/ram.c:169:
> +#define RAMBLOCK_FOREACH_NOT_IGNORED(block)            \
> +    INTERNAL_RAMBLOCK_FOREACH(block)                   \
> +        if (ramblock_is_ignored(block)) {} else

This is OK, it's just a copy of the existing macro we have for the loop

> ERROR: trailing statements should be on next line
> #148: FILE: migration/ram.c:171:
> +        if (ramblock_is_ignored(block)) {} else
> 
> total: 2 errors, 0 warnings, 386 lines checked
> 
> Patch 3/5 has style problems, please review.  If any of these errors
> are false positives report them to the maintainer, see
> CHECKPATCH in MAINTAINERS.
> 
> 4/5 Checking commit fad48088a805 (tests/migration-test: Add a test for ignore-shared capability)
> 5/5 Checking commit 5f9e90967e69 (migration: Add capabilities validation)
> === OUTPUT END ===
> 
> Test command exited with code: 1
> 
> 
> The full log is available at
> http://patchew.org/logs/[e-mail]/testing.checkpatch/?type=message.
> ---
> Email generated automatically by Patchew [http://patchew.org/].
> Please send your feedback to [e-mail]
--
Dr. David Alan Gilbert / [e-mail] / Manchester, UK

[e-mail] Feb. 27, 2019, 3:25 p.m. UTC | #5

Patchew URL: https://patchew.org/QEMU/[e-mail]/



Hi,

This series seems to have some coding style problems. See output below for
more information:

Message-id: [e-mail]
Subject: [Qemu-devel] [PATCH v3 0/5] Add ignore-external migration capability
Type: series

=== TEST SCRIPT BEGIN ===
#!/bin/bash
git rev-parse base > /dev/null || exit 0
git config --local diff.renamelimit 0
git config --local diff.renames True
git config --local diff.algorithm histogram
./scripts/checkpatch.pl --mailback base..
=== TEST SCRIPT END ===

Updating 3c8cf5a9c21ff8782164d1def7f44bd888713384
Switched to a new branch 'test'
74b71b6821 migration: Add capabilities validation
6d93a8d489 tests/migration-test: Add a test for ignore-shared capability
49c2e3db2c migration: Add an ability to ignore shared RAM blocks
96e230b463 migration: Introduce ignore-shared capability
c56095df53 exec: Change RAMBlockIterFunc definition

=== OUTPUT BEGIN ===
1/5 Checking commit c56095df53b1 (exec: Change RAMBlockIterFunc definition)
2/5 Checking commit 96e230b46352 (migration: Introduce ignore-shared capability)
3/5 Checking commit 49c2e3db2cab (migration: Add an ability to ignore shared RAM blocks)
ERROR: Macros with multiple statements should be enclosed in a do - while loop
#147: FILE: migration/ram.c:169:
+#define RAMBLOCK_FOREACH_NOT_IGNORED(block)            \
+    INTERNAL_RAMBLOCK_FOREACH(block)                   \
+        if (ramblock_is_ignored(block)) {} else

ERROR: trailing statements should be on next line
#149: FILE: migration/ram.c:171:
+        if (ramblock_is_ignored(block)) {} else

total: 2 errors, 0 warnings, 386 lines checked

Patch 3/5 has style problems, please review.  If any of these errors
are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.

4/5 Checking commit 6d93a8d48994 (tests/migration-test: Add a test for ignore-shared capability)
5/5 Checking commit 74b71b6821b2 (migration: Add capabilities validation)
=== OUTPUT END ===

Test command exited with code: 1


The full log is available at
http://patchew.org/logs/[e-mail]/testing.checkpatch/?type=message.
---
Email generated automatically by Patchew [http://patchew.org/].
Please send your feedback to [e-mail]

Dr. David Alan Gilbert March 5, 2019, 2:05 p.m. UTC | #6

* Yury Kotov ([e-mail]) wrote:
> Hi,
> 
> The series adds a migration capability, which allows to skip shared RAM blocks
> during the migration. It's useful for fast local migration. E.g. to update QEMU
> for the running guests.

Queued

> 
> Usage example:
> 1. Start source VM:
>    qemu-system-x86 \
>      -m 4G \
>      -object memory-backend-file,id=mem0,size=4G,share=on,mem-path=/dev/shm/mem0 \
>      -numa node,memdev=mem0 \
>      -qmp unix:/tmp/qemu-qmp-1.sock,server,nowait \
> 
> 2. Start target VM:
>    qemu-system-x86 \
>      -m 4G \
>      -object memory-backend-file,id=mem0,size=4G,share=on,mem-path=/dev/shm/mem0 \
>      -numa node,memdev=mem0 \
>      -qmp unix:/tmp/qemu-qmp-2.sock,server,nowait \
>      -incoming defer
> 
> 3. Enable ignore-shared capability on both VMs:
>    { "execute": "migrate-set-capabilities" , "arguments":
>      { "capabilities": [ { "capability": "x-ignore-shared", "state": true } ] } }
> 
> 4. Start migration.
> 
> Another use case I keep in mind is to migrate to file. Usage is very similar.
> 
> V2 to V3:
> * Split "migration: Introduce ignore-shared capability"
> * Serialize the capabilities as strings rather than as indexes
> * Don't allow to enable postcopy and ignore-shared together
> * Skip the test for OSs which don't have /dev/shm
> * Add a check whether shared RAM has been really skipped
> 
> V1 to V2:
> * Keep migration stream compatibility
> * Reuse the existing code to ignore unwanted RAMBlocks
> * Add capability validation feature
> * ignore-external -> ignore-shared
> 
> Regards,
> Yury
> 
> Yury Kotov (5):
>   exec: Change RAMBlockIterFunc definition
>   migration: Introduce ignore-shared capability
>   migration: Add an ability to ignore shared RAM blocks
>   tests/migration-test: Add a test for ignore-shared capability
>   migration: Add capabilities validation
> 
>  exec.c                    |  38 +++++------
>  include/exec/cpu-common.h |   7 +-
>  migration/migration.c     |  14 ++++
>  migration/migration.h     |   5 +-
>  migration/postcopy-ram.c  |  48 +++++++------
>  migration/ram.c           | 110 ++++++++++++++++++++++--------
>  migration/rdma.c          |   9 ++-
>  migration/savevm.c        | 137 ++++++++++++++++++++++++++++++++++++++
>  qapi/migration.json       |   5 +-
>  stubs/ram-block.c         |  15 +++++
>  tests/migration-test.c    | 131 +++++++++++++++++++++++++++++-------
>  util/vfio-helpers.c       |   6 +-
>  12 files changed, 420 insertions(+), 105 deletions(-)
> 
> -- 
> 2.20.1
> 
> 
--
Dr. David Alan Gilbert / [e-mail] / Manchester, UK

Dr. David Alan Gilbert March 5, 2019, 6:06 p.m. UTC | #7

* Dr. David Alan Gilbert ([e-mail]) wrote:
> * Yury Kotov ([e-mail]) wrote:
> > Hi,
> > 
> > The series adds a migration capability, which allows to skip shared RAM blocks
> > during the migration. It's useful for fast local migration. E.g. to update QEMU
> > for the running guests.
> 
> Queued

Hi Yury,
  The test is failing badly on aarch64 tcg for me (i.e. aarch64 tcg on
x86);  I'm going to keep it merged but keep the test disabled for now.
Can you have a look to see if you can see what's going wrong?

Dave

> > 
> > Usage example:
> > 1. Start source VM:
> >    qemu-system-x86 \
> >      -m 4G \
> >      -object memory-backend-file,id=mem0,size=4G,share=on,mem-path=/dev/shm/mem0 \
> >      -numa node,memdev=mem0 \
> >      -qmp unix:/tmp/qemu-qmp-1.sock,server,nowait \
> > 
> > 2. Start target VM:
> >    qemu-system-x86 \
> >      -m 4G \
> >      -object memory-backend-file,id=mem0,size=4G,share=on,mem-path=/dev/shm/mem0 \
> >      -numa node,memdev=mem0 \
> >      -qmp unix:/tmp/qemu-qmp-2.sock,server,nowait \
> >      -incoming defer
> > 
> > 3. Enable ignore-shared capability on both VMs:
> >    { "execute": "migrate-set-capabilities" , "arguments":
> >      { "capabilities": [ { "capability": "x-ignore-shared", "state": true } ] } }
> > 
> > 4. Start migration.
> > 
> > Another use case I keep in mind is to migrate to file. Usage is very similar.
> > 
> > V2 to V3:
> > * Split "migration: Introduce ignore-shared capability"
> > * Serialize the capabilities as strings rather than as indexes
> > * Don't allow to enable postcopy and ignore-shared together
> > * Skip the test for OSs which don't have /dev/shm
> > * Add a check whether shared RAM has been really skipped
> > 
> > V1 to V2:
> > * Keep migration stream compatibility
> > * Reuse the existing code to ignore unwanted RAMBlocks
> > * Add capability validation feature
> > * ignore-external -> ignore-shared
> > 
> > Regards,
> > Yury
> > 
> > Yury Kotov (5):
> >   exec: Change RAMBlockIterFunc definition
> >   migration: Introduce ignore-shared capability
> >   migration: Add an ability to ignore shared RAM blocks
> >   tests/migration-test: Add a test for ignore-shared capability
> >   migration: Add capabilities validation
> > 
> >  exec.c                    |  38 +++++------
> >  include/exec/cpu-common.h |   7 +-
> >  migration/migration.c     |  14 ++++
> >  migration/migration.h     |   5 +-
> >  migration/postcopy-ram.c  |  48 +++++++------
> >  migration/ram.c           | 110 ++++++++++++++++++++++--------
> >  migration/rdma.c          |   9 ++-
> >  migration/savevm.c        | 137 ++++++++++++++++++++++++++++++++++++++
> >  qapi/migration.json       |   5 +-
> >  stubs/ram-block.c         |  15 +++++
> >  tests/migration-test.c    | 131 +++++++++++++++++++++++++++++-------
> >  util/vfio-helpers.c       |   6 +-
> >  12 files changed, 420 insertions(+), 105 deletions(-)
> > 
> > -- 
> > 2.20.1
> > 
> > 
> --
> Dr. David Alan Gilbert / [e-mail] / Manchester, UK
--
Dr. David Alan Gilbert / [e-mail] / Manchester, UK

Yury Kotov March 6, 2019, 10:47 a.m. UTC | #8

05.03.2019, 21:06, "Dr. David Alan Gilbert" <[e-mail]>:
> * Dr. David Alan Gilbert ([e-mail]) wrote:
>>  * Yury Kotov ([e-mail]) wrote:
>>  > Hi,
>>  >
>>  > The series adds a migration capability, which allows to skip shared RAM blocks
>>  > during the migration. It's useful for fast local migration. E.g. to update QEMU
>>  > for the running guests.
>>
>>  Queued
>
> Hi Yury,
>   The test is failing badly on aarch64 tcg for me (i.e. aarch64 tcg on
> x86); I'm going to keep it merged but keep the test disabled for now.
> Can you have a look to see if you can see what's going wrong?
>
> Dave
>

Hi,

Yes, of course. I'll look at this.

Regards,
Yury

>>  >
>>  > Usage example:
>>  > 1. Start source VM:
>>  > qemu-system-x86 \
>>  > -m 4G \
>>  > -object memory-backend-file,id=mem0,size=4G,share=on,mem-path=/dev/shm/mem0 \
>>  > -numa node,memdev=mem0 \
>>  > -qmp unix:/tmp/qemu-qmp-1.sock,server,nowait \
>>  >
>>  > 2. Start target VM:
>>  > qemu-system-x86 \
>>  > -m 4G \
>>  > -object memory-backend-file,id=mem0,size=4G,share=on,mem-path=/dev/shm/mem0 \
>>  > -numa node,memdev=mem0 \
>>  > -qmp unix:/tmp/qemu-qmp-2.sock,server,nowait \
>>  > -incoming defer
>>  >
>>  > 3. Enable ignore-shared capability on both VMs:
>>  > { "execute": "migrate-set-capabilities" , "arguments":
>>  > { "capabilities": [ { "capability": "x-ignore-shared", "state": true } ] } }
>>  >
>>  > 4. Start migration.
>>  >
>>  > Another use case I keep in mind is to migrate to file. Usage is very similar.
>>  >
>>  > V2 to V3:
>>  > * Split "migration: Introduce ignore-shared capability"
>>  > * Serialize the capabilities as strings rather than as indexes
>>  > * Don't allow to enable postcopy and ignore-shared together
>>  > * Skip the test for OSs which don't have /dev/shm
>>  > * Add a check whether shared RAM has been really skipped
>>  >
>>  > V1 to V2:
>>  > * Keep migration stream compatibility
>>  > * Reuse the existing code to ignore unwanted RAMBlocks
>>  > * Add capability validation feature
>>  > * ignore-external -> ignore-shared
>>  >
>>  > Regards,
>>  > Yury
>>  >
>>  > Yury Kotov (5):
>>  > exec: Change RAMBlockIterFunc definition
>>  > migration: Introduce ignore-shared capability
>>  > migration: Add an ability to ignore shared RAM blocks
>>  > tests/migration-test: Add a test for ignore-shared capability
>>  > migration: Add capabilities validation
>>  >
>>  > exec.c | 38 +++++------
>>  > include/exec/cpu-common.h | 7 +-
>>  > migration/migration.c | 14 ++++
>>  > migration/migration.h | 5 +-
>>  > migration/postcopy-ram.c | 48 +++++++------
>>  > migration/ram.c | 110 ++++++++++++++++++++++--------
>>  > migration/rdma.c | 9 ++-
>>  > migration/savevm.c | 137 ++++++++++++++++++++++++++++++++++++++
>>  > qapi/migration.json | 5 +-
>>  > stubs/ram-block.c | 15 +++++
>>  > tests/migration-test.c | 131 +++++++++++++++++++++++++++++-------
>>  > util/vfio-helpers.c | 6 +-
>>  > 12 files changed, 420 insertions(+), 105 deletions(-)
>>  >
>>  > --
>>  > 2.20.1
>>  >
>>  >
>>  --
>>  Dr. David Alan Gilbert / [e-mail] / Manchester, UK
> --
> Dr. David Alan Gilbert / [e-mail] / Manchester, UK

Yury Kotov March 14, 2019, 11:05 a.m. UTC | #9

Hi,

I've sent a mail about this problem:
[RFC PATCH] QEMU may write to system_memory before guest starts

Regards,
Yury

06.03.2019, 13:47, "Yury Kotov" <[e-mail]>:
> 05.03.2019, 21:06, "Dr. David Alan Gilbert" <[e-mail]>:
>>  * Dr. David Alan Gilbert ([e-mail]) wrote:
>>>   * Yury Kotov ([e-mail]) wrote:
>>>   > Hi,
>>>   >
>>>   > The series adds a migration capability, which allows to skip shared RAM blocks
>>>   > during the migration. It's useful for fast local migration. E.g. to update QEMU
>>>   > for the running guests.
>>>
>>>   Queued
>>
>>  Hi Yury,
>>    The test is failing badly on aarch64 tcg for me (i.e. aarch64 tcg on
>>  x86); I'm going to keep it merged but keep the test disabled for now.
>>  Can you have a look to see if you can see what's going wrong?
>>
>>  Dave
>
> Hi,
>
> Yes, of course. I'll look at this.
>
> Regards,
> Yury
>
>>>   >
>>>   > Usage example:
>>>   > 1. Start source VM:
>>>   > qemu-system-x86 \
>>>   > -m 4G \
>>>   > -object memory-backend-file,id=mem0,size=4G,share=on,mem-path=/dev/shm/mem0 \
>>>   > -numa node,memdev=mem0 \
>>>   > -qmp unix:/tmp/qemu-qmp-1.sock,server,nowait \
>>>   >
>>>   > 2. Start target VM:
>>>   > qemu-system-x86 \
>>>   > -m 4G \
>>>   > -object memory-backend-file,id=mem0,size=4G,share=on,mem-path=/dev/shm/mem0 \
>>>   > -numa node,memdev=mem0 \
>>>   > -qmp unix:/tmp/qemu-qmp-2.sock,server,nowait \
>>>   > -incoming defer
>>>   >
>>>   > 3. Enable ignore-shared capability on both VMs:
>>>   > { "execute": "migrate-set-capabilities" , "arguments":
>>>   > { "capabilities": [ { "capability": "x-ignore-shared", "state": true } ] } }
>>>   >
>>>   > 4. Start migration.
>>>   >
>>>   > Another use case I keep in mind is to migrate to file. Usage is very similar.
>>>   >
>>>   > V2 to V3:
>>>   > * Split "migration: Introduce ignore-shared capability"
>>>   > * Serialize the capabilities as strings rather than as indexes
>>>   > * Don't allow to enable postcopy and ignore-shared together
>>>   > * Skip the test for OSs which don't have /dev/shm
>>>   > * Add a check whether shared RAM has been really skipped
>>>   >
>>>   > V1 to V2:
>>>   > * Keep migration stream compatibility
>>>   > * Reuse the existing code to ignore unwanted RAMBlocks
>>>   > * Add capability validation feature
>>>   > * ignore-external -> ignore-shared
>>>   >
>>>   > Regards,
>>>   > Yury
>>>   >
>>>   > Yury Kotov (5):
>>>   > exec: Change RAMBlockIterFunc definition
>>>   > migration: Introduce ignore-shared capability
>>>   > migration: Add an ability to ignore shared RAM blocks
>>>   > tests/migration-test: Add a test for ignore-shared capability
>>>   > migration: Add capabilities validation
>>>   >
>>>   > exec.c | 38 +++++------
>>>   > include/exec/cpu-common.h | 7 +-
>>>   > migration/migration.c | 14 ++++
>>>   > migration/migration.h | 5 +-
>>>   > migration/postcopy-ram.c | 48 +++++++------
>>>   > migration/ram.c | 110 ++++++++++++++++++++++--------
>>>   > migration/rdma.c | 9 ++-
>>>   > migration/savevm.c | 137 ++++++++++++++++++++++++++++++++++++++
>>>   > qapi/migration.json | 5 +-
>>>   > stubs/ram-block.c | 15 +++++
>>>   > tests/migration-test.c | 131 +++++++++++++++++++++++++++++-------
>>>   > util/vfio-helpers.c | 6 +-
>>>   > 12 files changed, 420 insertions(+), 105 deletions(-)
>>>   >
>>>   > --
>>>   > 2.20.1
>>>   >
>>>   >
>>>   --
>>>   Dr. David Alan Gilbert / [e-mail] / Manchester, UK
>>  --
>>  Dr. David Alan Gilbert / [e-mail] / Manchester, UK