Highlights:
* New "-numa cpu" option
* NUMA distance configuration
* migration/i386 vmstatification
-----BEGIN PGP SIGNATURE-----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=48wH
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'ehabkost/tags/x86-and-machine-pull-request' into staging
x86 and machine queue, 2017-05-11
Highlights:
* New "-numa cpu" option
* NUMA distance configuration
* migration/i386 vmstatification
# gpg: Signature made Thu 11 May 2017 08:16:07 PM BST
# gpg: using RSA key 0x2807936F984DC5A6
# gpg: Good signature from "Eduardo Habkost <ehabkost@redhat.com>"
# gpg: Note: This key has expired!
# Primary key fingerprint: 5A32 2FD5 ABC4 D3DB ACCF D1AA 2807 936F 984D C5A6
* ehabkost/tags/x86-and-machine-pull-request: (29 commits)
migration/i386: Remove support for pre-0.12 formats
vmstatification: i386 FPReg
migration/i386: Remove old non-softfloat 64bit FP support
tests: check -numa node,cpu=props_list usecase
numa: add '-numa cpu,...' option for property based node mapping
numa: remove node_cpu bitmaps as they are no longer used
numa: use possible_cpus for not mapped CPUs check
machine: call machine init from wrapper
numa: remove no longer need numa_post_machine_init()
tests: numa: add case for QMP command query-cpus
QMP: include CpuInstanceProperties into query_cpus output output
virt-arm: get numa node mapping from possible_cpus instead of numa_get_node_for_cpu()
spapr: get numa node mapping from possible_cpus instead of numa_get_node_for_cpu()
pc: get numa node mapping from possible_cpus instead of numa_get_node_for_cpu()
numa: do default mapping based on possible_cpus instead of node_cpu bitmaps
numa: mirror cpu to node mapping in MachineState::possible_cpus
numa: add check that board supports cpu_index to node mapping
virt-arm: add node-id property to CPU
pc: add node-id property to CPU
spapr: add node-id property to sPAPR core
...
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This pull request supersedes the one from yesterday (20170510), fixing
an important style bug in one patch, and adding an extra couple of
simple patches.
Highlights of this set:
* Some fixes for POWER9
* TCG support for POWER9 radix MMU
* VGA rom for Mac machine types
* Fixes for the XICS interrupt controller
* MTTCG support for ppc targets
As suggested by Paolo, I've tried to add the Docker tests to my
standard pre-pull-request tests. I haven't wholly suceeded; this has
been tested with some of the Docker images, but others I haven't
managed due to problems that as best I can tell are not due to
problems in this patch series. I'll continue working on this for
future pull requests. Specifically, 'travis', 'fedora', and 'centos6'
seem to work. 'min-glib' jammed while gtesting moxie, which seems
very unlikely to be caused by this series. 'ubuntu', 'debian' and
'debian-bootstrap' hit build errors almost immediately that look like
problems with the container configuration, and 'debian-*-cross' hit
build errors later on which also look like missing dependencies from
the container.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2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=D0Hk
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'dgibson/tags/ppc-for-2.10-20170511' into staging
ppc patch queue for 2017-05-11
This pull request supersedes the one from yesterday (20170510), fixing
an important style bug in one patch, and adding an extra couple of
simple patches.
Highlights of this set:
* Some fixes for POWER9
* TCG support for POWER9 radix MMU
* VGA rom for Mac machine types
* Fixes for the XICS interrupt controller
* MTTCG support for ppc targets
As suggested by Paolo, I've tried to add the Docker tests to my
standard pre-pull-request tests. I haven't wholly suceeded; this has
been tested with some of the Docker images, but others I haven't
managed due to problems that as best I can tell are not due to
problems in this patch series. I'll continue working on this for
future pull requests. Specifically, 'travis', 'fedora', and 'centos6'
seem to work. 'min-glib' jammed while gtesting moxie, which seems
very unlikely to be caused by this series. 'ubuntu', 'debian' and
'debian-bootstrap' hit build errors almost immediately that look like
problems with the container configuration, and 'debian-*-cross' hit
build errors later on which also look like missing dependencies from
the container.
# gpg: Signature made Thu 11 May 2017 05:13:46 AM BST
# gpg: using RSA key 0x6C38CACA20D9B392
# gpg: Good signature from "David Gibson <david@gibson.dropbear.id.au>"
# gpg: aka "David Gibson (kernel.org) <dwg@kernel.org>"
# gpg: aka "David Gibson (Red Hat) <dgibson@redhat.com>"
# gpg: aka "David Gibson (ozlabs.org) <dgibson@ozlabs.org>"
# Primary key fingerprint: 75F4 6586 AE61 A66C C44E 87DC 6C38 CACA 20D9 B392
* dgibson/tags/ppc-for-2.10-20170511: (23 commits)
target/ppc: Avoid printing wrong aliases in CPU help text
pnv: Fix build failures on some host platforms
target/ppc: Allow workarounds for POWER9 DD1
spapr: Don't accidentally advertise HTM support on POWER9
ppc: xics: fix compilation with CentOS 6
target/ppc: Enable RADIX mmu mode for pseries TCG guest
target/ppc: Implement ISA V3.00 radix page fault handler
target/ppc: Change tlbie invalid fields for POWER9 support
target/ppc: Update tlbie to check privilege level based on GTSE
target/ppc: Set UPRT and GTSE on all cpus in H_REGISTER_PROCESS_TABLE
ppc: add qemu_vga.ndrv ROM to fw_cfg interface for NewWorld Macs
ppc: add qemu_vga.ndrv ROM to fw_cfg interface for OldWorld Macs
Add QemuMacDrivers qemu_vga.ndrv revision d4e7d7a built as submodule
Add QemuMacDrivers as submodule
ppc/xics: preserve P and Q bits for KVM IRQs
ppc/xics: Fix stale irq->status bits after get
target/ppc: do not reset reserve_addr in exec_enter
tcg: enable MTTCG by default for PPC64 on x86
cpus: Fix CPU unplug for MTTCG
target/ppc: Generate fence operations
...
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
If trace backend is set to TRACE_NOP, trace_get_vcpu_event_count
returns 0, cause bitmap_new call abort.
The abort can be triggered as follows:
$ ./configure --enable-trace-backend=nop --target-list=x86_64-softmmu
$ gdb ./x86_64-softmmu/qemu-system-x86_64 -M q35,accel=kvm -m 1G
(gdb) bt
#0 0x00007ffff04e25f7 in raise () from /lib64/libc.so.6
#1 0x00007ffff04e3ce8 in abort () from /lib64/libc.so.6
#2 0x00005555559de905 in bitmap_new (nbits=<optimized out>)
at /home/root/git/qemu2.git/include/qemu/bitmap.h:96
#3 cpu_common_initfn (obj=0x555556621d30) at qom/cpu.c:399
#4 0x0000555555a11869 in object_init_with_type (obj=0x555556621d30, ti=0x55555656bbb0) at qom/object.c:341
#5 0x0000555555a11869 in object_init_with_type (obj=0x555556621d30, ti=0x55555656bd30) at qom/object.c:341
#6 0x0000555555a11efc in object_initialize_with_type (data=data@entry=0x555556621d30, size=76560,
type=type@entry=0x55555656bd30) at qom/object.c:376
#7 0x0000555555a12061 in object_new_with_type (type=0x55555656bd30) at qom/object.c:484
#8 0x0000555555a121c5 in object_new (typename=typename@entry=0x555556550340 "qemu64-x86_64-cpu")
at qom/object.c:494
#9 0x00005555557f6e3d in pc_new_cpu (typename=typename@entry=0x555556550340 "qemu64-x86_64-cpu", apic_id=0,
errp=errp@entry=0x5555565391b0 <error_fatal>) at /home/root/git/qemu2.git/hw/i386/pc.c:1101
#10 0x00005555557fa33e in pc_cpus_init (pcms=pcms@entry=0x5555565f9690)
at /home/root/git/qemu2.git/hw/i386/pc.c:1184
#11 0x00005555557fe0f6 in pc_q35_init (machine=0x5555565f9690) at /home/root/git/qemu2.git/hw/i386/pc_q35.c:121
#12 0x000055555574fbad in main (argc=<optimized out>, argv=<optimized out>, envp=<optimized out>) at vl.c:4562
Signed-off-by: Anthony Xu <anthony.xu@intel.com>
Message-id: 1494369432-15418-1-git-send-email-anthony.xu@intel.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
The main loop uses aio_disable_external()/aio_enable_external() to
temporarily disable processing of external AioContext clients like
device emulation.
This allows monitor commands to quiesce I/O and prevent the guest from
submitting new requests while a monitor command is in progress.
The aio_enable_external() API is currently broken when an IOThread is in
aio_poll() waiting for fd activity when the main loop re-enables
external clients. Incrementing ctx->external_disable_cnt does not wake
the IOThread from ppoll(2) so fd processing remains suspended and leads
to unresponsive emulated devices.
This patch adds an aio_notify() call to aio_enable_external() so the
IOThread is kicked out of ppoll(2) and will re-arm the file descriptors.
The bug can be reproduced as follows:
$ qemu -M accel=kvm -m 1024 \
-object iothread,id=iothread0 \
-device virtio-scsi-pci,iothread=iothread0,id=virtio-scsi-pci0 \
-drive if=none,id=drive0,aio=native,cache=none,format=raw,file=test.img \
-device scsi-hd,id=scsi-hd0,drive=drive0 \
-qmp tcp::5555,server,nowait
$ scripts/qmp/qmp-shell localhost:5555
(qemu) blockdev-snapshot-sync device=drive0 snapshot-file=sn1.qcow2
mode=absolute-paths format=qcow2
After blockdev-snapshot-sync completes the SCSI disk will be
unresponsive. This leads to request timeouts inside the guest.
Reported-by: Qianqian Zhu <qizhu@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20170508180705.20609-1-stefanha@redhat.com
Suggested-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Since we are already in coroutine context during the body of
bdrv_co_get_block_status(), we can shave off a few layers of
wrappers when recursing to query the protocol when a format driver
returned BDRV_BLOCK_RAW.
Note that we are already using the correct recursion later on in
the same function, when probing whether the protocol layer is sparse
in order to find out if we can add BDRV_BLOCK_ZERO to an existing
BDRV_BLOCK_DATA|BDRV_BLOCK_OFFSET_VALID.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 20170504173745.27414-1-eblake@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
The GThread implementation is not functional enough to actually
run QEMU reliably. While it was potentially useful for debugging,
we have a scripts/qemugdb/coroutine.py to enable tracing of
ucontext coroutines in GDB, so that removes the only reason for
GThread to exist.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Acked-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Remove support for versions of the CPU state prior to 11
which is the version used in qemu 0.12 - you'd be pretty
lucky if you got a migration stream to work from anything
that old anyway. This doesn't affect the machine type
definition in any way.
My main reason for doing this is the hack for sysenter_esp/eip
that uses .get/.put's in state versions less than 7 (that's
prior to somewhere before 0.10).
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-Id: <20170405190024.27581-4-dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Convert the fpreg save/restore to use VMSTATE_ macros rather than
.get/.put.
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-Id: <20170405190024.27581-3-dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Long long ago, we used to support storing the x86 FP registers in
a 64bit format.
Then c31da136a0 in v0.14-rc0 removed
the last support for writing that in the migration format.
Even before that, it was only used if you had softfloat disabled
(i.e. !USE_X86LDOUBLE) so in practice use of it in even earlier
qemu is unlikely for most users.
Kill it off, it's complicated, and possibly broken.
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-Id: <20170405190024.27581-2-dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
legacy cpu to node mapping is using cpu index values to map
VCPU to node with help of '-numa node,nodeid=node,cpus=x[-y]'
option. However cpu index is internal concept and QEMU users
have to guess /reimplement qemu's logic/ to map it to
a concrete cpu socket/core/thread to make sane CPUs
placement across numa nodes.
This patch allows to map cpu objects to numa nodes using
the same properties as used for cpus with -device/device_add
(socket-id/core-id/thread-id/node-id).
At present valid properties/values to address CPUs could be
fetched using hotpluggable-cpus monitor/qmp command, it will
require user to start qemu twice when creating domain to fetch
possible CPUs for a machine type/-smp layout first and
then the second time with numa explicit mapping for actual
usage. The first step results could be saved and reused to
set/change mapping later as far as machine type/-smp stays
the same.
Proposed impl. supports exact and wildcard matching to
simplify CLI and allow to set mapping for a specific cpu
or group of cpu objects specified by matched properties.
For example:
# exact mapping x86
-numa cpu,node-id=x,socket-id=y,core-id=z,thread-id=n
# exact mapping SPAPR
-numa cpu,node-id=x,core-id=y
# wildcard mapping, all cpu objects that match socket-id=y
# are mapped to node-id=x
-numa cpu,node-id=x,socket-id=y
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <1494415802-227633-18-git-send-email-imammedo@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Postfactum "CPU(s) present in multiple NUMA nodes" check
was the last user of node_cpu bitmaps, but it's not need
as machine_set_cpu_numa_node() does the similar check at
the time mapping is set for cpus (i.e. when -numa cpus=
is parsed) and ensures that cpu can be mapped only to
one node.
Remove duplicate check based on node_cpu bitmaps and
since the last user is gone remove node_cpu as well,
which completes internal transition from legacy bitmap
based mapping storage to possible_cpus storage.
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Message-Id: <1494415802-227633-17-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
and remove corresponding part in numa.c that uses
node_cpu bitmaps.
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Message-Id: <1494415802-227633-16-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
add machine_run_board_init() wrapper that calls machine
init for now but in follow up patches it will be used
to run generic machine code that should run before
machine init.
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Message-Id: <1494415802-227633-15-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
CPUState::numa_node is still in use but now it's set by
board when it creates CPU objects. So there isn't any
need to set it again after all CPU's are created,
since it's been already set.
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Message-Id: <1494415802-227633-14-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
if board supports CpuInstanceProperties, report them for
each CPU thread listed. Main motivation for this is to
provide these properties introspection via QMP interface
for using in test cases to verify numa node to cpu mapping,
which includes not only boards that support cpu hotplug
and have this info in query-hotpluggable-cpus (pc/spapr)
but also for boards that don't not support hotpluggable-cpus
but support numa mapping (virt-arm).
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <1494415802-227633-12-git-send-email-imammedo@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Message-Id: <1494415802-227633-11-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
it's safe to remove thread node_id != core node_id error
branch as machine_set_cpu_numa_node() also does mismatch
check and is called even before any CPU is created.
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <1494415802-227633-10-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Message-Id: <1494415802-227633-9-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Message-Id: <1494415802-227633-8-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Introduce machine_set_cpu_numa_node() helper that stores
node mapping for CPU in MachineState::possible_cpus.
CPU and node it belongs to is specified by 'props' argument.
Patch doesn't remove old way of storing mapping in
numa_info[X].node_cpu as removing it at the same time
makes patch rather big. Instead it just mirrors mapping
in possible_cpus and follow up per target patches will
switch to possible_cpus and numa_info[X].node_cpu will
be removed once there isn't any users left.
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Message-Id: <1494415802-227633-7-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Default node mapping initialization already checks that board
supports cpu_index to node mapping and refuses to start if
it's not supported. Do the same for explicitly provided
mapping "-numa node,cpus=..."
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <1494415802-227633-6-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
it will allow switching from cpu_index to property based
numa mapping in follow up patches.
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Message-Id: <1494415802-227633-5-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
it will allow switching from cpu_index to property based
numa mapping in follow up patches.
PS:
patch changes default value of CPUState::numa_node from 0
to CPU_UNSET_NUMA_NODE_ID. The only place for x86 that
would affected is monitor's 'infor numa' command which
uses that field. However legacy 0 value is still preserved
by pc_cpu_pre_plug() in this patch if user/numa.c hasn't
set it explicitly, so there is no change in behavior.
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <1494415802-227633-4-git-send-email-imammedo@redhat.com>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
it will allow switching from cpu_index to core based numa
mapping in follow up patches.
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <1494415802-227633-3-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Originally CPU threads were by default assigned in
round-robin fashion. However it was causing issues in
guest since CPU threads from the same socket/core could
be placed on different NUMA nodes.
Commit fb43b73b (pc: fix default VCPU to NUMA node mapping)
fixed it by grouping threads within a socket on the same node
introducing cpu_index_to_socket_id() callback and commit
20bb648d (spapr: Fix default NUMA node allocation for threads)
reused callback to fix similar issues for SPAPR machine
even though socket doesn't make much sense there.
As result QEMU ended up having 3 default distribution rules
used by 3 targets /virt-arm, spapr, pc/.
In effort of moving NUMA mapping for CPUs into possible_cpus,
generalize default mapping in numa.c by making boards decide
on default mapping and let them explicitly tell generic
numa code to which node a CPU thread belongs to by replacing
cpu_index_to_socket_id() with @cpu_index_to_instance_props()
which provides default node_id assigned by board to specified
cpu_index.
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Message-Id: <1494415802-227633-2-git-send-email-imammedo@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Currently cpu_index is implicitly auto assigned during
cpu.realize() time cpu_exec_realizefn()->cpu_list_add().
It happens to match index in possible_cpus so take
control over it and make board initialize cpu_index
to possible_cpus index explicitly. It will at least
document that board is in control of it and when
'-device cpu' support comes it will keep cpu_index
stable regardless of order cpus are created so it won't
break migration.
Within this series it will be used for internal
conversion from storing cpu_index based NUMA node
bitmaps to property based mapping with possible_cpus,
And will allow map cpu_index to a CPU entry in
possible_cpus array.
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Message-Id: <1493816238-33120-5-git-send-email-imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
for now precalculate and store mp_afinity in possible_cpus
as ARM cpus don't have socket/core/thread-id properties yet.
In follow patches possible_cpus will be used for storing
and setting NUMA node mapping and replace legacy bitmap
based numa_info[node_id].node_cpu/numa_get_node_for_cpu()
For the lack of better idea, this patch cannibalizes
possible_cpus.cpus[x].props.thread_id so that
*_cpu_index_to_props() callback could return addressable
by props CPU which will be used by machine_set_cpu_numa_node()
in follow up patches to assign a CPU to node. But
cannibalizing is fine for now as that thread_id isn't exposed
to users (no hotpluggable_cpus callback support for ARM yet)
and it will be used only internally until 'device_add cpu'
is supported where we can decide on which properties to use.
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <1493816238-33120-4-git-send-email-imammedo@redhat.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <1493816238-33120-3-git-send-email-imammedo@redhat.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
When there are more nodes than available memory to put the minimum
allowed memory by node, all the memory is put on the last node.
This is because we put (ram_size / nb_numa_nodes) &
~((1 << mc->numa_mem_align_shift) - 1); on each node, and in this
case the value is 0. This is particularly true with pseries,
as the memory must be aligned to 256MB.
To avoid this problem, this patch uses an error diffusion algorithm [1]
to distribute equally the memory on nodes.
We introduce numa_auto_assign_ram() function in MachineClass
to keep compatibility between machine type versions.
The legacy function is used with pseries-2.9, pc-q35-2.9 and
pc-i440fx-2.9 (and previous), the new one with all others.
Example:
qemu-system-ppc64 -S -nographic -nodefaults -monitor stdio -m 1G -smp 8 \
-numa node -numa node -numa node \
-numa node -numa node -numa node
Before:
(qemu) info numa
6 nodes
node 0 cpus: 0 6
node 0 size: 0 MB
node 1 cpus: 1 7
node 1 size: 0 MB
node 2 cpus: 2
node 2 size: 0 MB
node 3 cpus: 3
node 3 size: 0 MB
node 4 cpus: 4
node 4 size: 0 MB
node 5 cpus: 5
node 5 size: 1024 MB
After:
(qemu) info numa
6 nodes
node 0 cpus: 0 6
node 0 size: 0 MB
node 1 cpus: 1 7
node 1 size: 256 MB
node 2 cpus: 2
node 2 size: 0 MB
node 3 cpus: 3
node 3 size: 256 MB
node 4 cpus: 4
node 4 size: 256 MB
node 5 cpus: 5
node 5 size: 256 MB
[1] https://en.wikipedia.org/wiki/Error_diffusion
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Message-Id: <20170502162955.1610-2-lvivier@redhat.com>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
[ehabkost: s/ram_size/size/ at numa_default_auto_assign_ram()]
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
This patch is going to add SLIT table support in QEMU, and provides
additional option `dist` for command `-numa` to allow user set vNUMA
distance by QEMU command.
With this patch, when a user wants to create a guest that contains
several vNUMA nodes and also wants to set distance among those nodes,
the QEMU command would like:
```
-numa node,nodeid=0,cpus=0 \
-numa node,nodeid=1,cpus=1 \
-numa node,nodeid=2,cpus=2 \
-numa node,nodeid=3,cpus=3 \
-numa dist,src=0,dst=1,val=21 \
-numa dist,src=0,dst=2,val=31 \
-numa dist,src=0,dst=3,val=41 \
-numa dist,src=1,dst=2,val=21 \
-numa dist,src=1,dst=3,val=31 \
-numa dist,src=2,dst=3,val=21 \
```
Signed-off-by: He Chen <he.chen@linux.intel.com>
Message-Id: <1493260558-20728-1-git-send-email-he.chen@linux.intel.com>
Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
I volunteer to review linux-user patches.
Adding myself will help to not miss some of them.
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Acked-by: Riku Voipio <riku.voipio@linaro.org>
Message-id: 20170510153950.29343-1-laurent@vivier.eu
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Change the nested if statements into a flat format, to make
it clearer what validation / capping is being performed on
different CPUID index values.
NB this changes behaviour when "index > env->cpuid_xlevel2".
This won't have any guest-visible effect because no there is
no CPUID[0xC0000001] feature supported by TCG, and KVM code
will never call cpu_x86_cpuid() with such an index value.
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <20170509132736.10071-2-berrange@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
util/qemu-progress.c is currently unmaintained. The only user of its
functionality is qemu-img, so it effectively is part of the block layer.
Suggested-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170428165517.30341-1-mreitz@redhat.com
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Passing a byte offset, but sector count, when we ultimately
want to operate on cluster granularity, is madness. Clean up
the external interfaces to take both offset and count as bytes,
while still keeping the assertion added previously that the
caller must align the values to a cluster. Then rename things
to make sure backports don't get confused by changed units:
instead of qcow2_discard_clusters() and qcow2_zero_clusters(),
we now have qcow2_cluster_discard() and qcow2_cluster_zeroize().
The internal functions still operate on clusters at a time, and
return an int for number of cleared clusters; but on an image
with 2M clusters, a single L2 table holds 256k entries that each
represent a 2M cluster, totalling well over INT_MAX bytes if we
ever had a request for that many bytes at once. All our callers
currently limit themselves to 32-bit bytes (and therefore fewer
clusters), but by making this function 64-bit clean, we have one
less place to clean up if we later improve the block layer to
support 64-bit bytes through all operations (with the block layer
auto-fragmenting on behalf of more-limited drivers), rather than
the current state where some interfaces are artificially limited
to INT_MAX at a time.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-13-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
We already audited (in commit 0c1bd469) that qcow2_discard_clusters()
is only passed cluster-aligned start values; but we can further
tighten the assertion that the only unaligned end value is at EOF.
Recent commits have taken advantage of an unaligned tail cluster,
for both discard and write zeroes.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-12-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
We've already improved discards to operate efficiently on the tail
of an unaligned qcow2 image; it's time to make a similar improvement
to write zeroes. The special case is only valid at the tail
cluster of a file, where we must recognize that any sectors beyond
the image end would implicitly read as zero, and therefore should
not penalize our logic for widening a partial cluster into writing
the whole cluster as zero.
However, note that for now, the special case of end-of-file is only
recognized if there is no backing file, or if the backing file has
the same length; that's because when the backing file is shorter
than the active layer, we don't have code in place to recognize
that reads of a sector unallocated at the top and beyond the backing
end-of-file are implicitly zero. It's not much of a real loss,
because most people don't use images that aren't cluster-aligned,
or where the active layer is a different size than the backing
layer (especially where the difference falls within a single cluster).
Update test 154 to cover the new scenarios, using two images of
intentionally differing length.
While at it, fix the test to gracefully skip when run as
./check -qcow2 -o compat=0.10 154
since the older format lacks zero clusters already required earlier
in the test.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-11-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
No tests were covering write zeroes with unmap. Additionally,
I needed to prove that my previous patches for correct status
reporting and write zeroes optimizations actually had an impact.
The test works for cluster_size between 8k and 2M (for smaller
sizes, it fails because our allocation patterns are not contiguous
with small clusters - in part, the largest consecutive allocation
we tend to get is often bounded by the size covered by one L2
table).
Note that testing for zero clusters is tricky: 'qemu-io map'
reports whether data comes from the current layer of the image
(useful for sniffing out which regions of the file have
QCOW_OFLAG_ZERO) - but doesn't show which clusters have mappings;
while 'qemu-img map' sees "zero":true for both unallocated and
zero clusters for any qcow2 with no backing layer (so less useful
at detecting true zero clusters), but reliably shows mappings.
So we have to rely on both queries side-by-side at each point of
the test.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-10-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Although _filter_qemu_img_map documents that it scrubs offsets, it
was only doing so for human mode. Of the existing tests using the
filter (97, 122, 150, 154, 176), two of them are affected, but it
does not hurt the validity of the tests to not require particular
mappings (another test, 66, uses offsets but intentionally does not
pass through _filter_qemu_img_map, because it checks that offsets
are unchanged before and after an operation).
Another justification for this patch is that it will allow a future
patch to utilize 'qemu-img map --output=json' to check the status of
preallocated zero clusters without regards to the mapping (since
the qcow2 mapping can be very sensitive to the chosen cluster size,
when preallocation is not in use).
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-9-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Similar to discard_single_l2(), we should try to avoid dirtying
the L2 cache when the cluster we are changing already has the
right characteristics.
Note that by the time we get to zero_single_l2(), BDRV_REQ_MAY_UNMAP
is a requirement to unallocate a cluster (this is because the block
layer clears that flag if discard.* flags during open requested that
we never punch holes - see the conversation around commit 170f4b2e,
https://lists.gnu.org/archive/html/qemu-devel/2016-09/msg07306.html).
Therefore, this patch can only reuse a zero cluster as-is if either
unmapping is not requested, or if the zero cluster was not associated
with an allocation.
Technically, there are some cases where an unallocated cluster
already reads as all zeroes (namely, when there is no backing file
[easy: check bs->backing], or when the backing file also reads as
zeroes [harder: we can't check bdrv_get_block_status since we are
already holding the lock]), where the guest would not immediately see
a difference if we left that cluster unallocated. But if the user
did not request unmapping, leaving an unallocated cluster is wrong;
and even if the user DID request unmapping, keeping a cluster
unallocated risks a subtle semantic change of guest-visible contents
if a backing file is later added, and it is not worth auditing
whether all internal uses such as mirror properly avoid an unmap
request. Thus, this patch is intentionally limited to just clusters
that are already marked as zero.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20170507000552.20847-8-eblake@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Treat plain zero clusters differently from allocated ones, so that
we can simplify the logic of checking whether an offset is present.
Do this by splitting QCOW2_CLUSTER_ZERO into two new enums,
QCOW2_CLUSTER_ZERO_PLAIN and QCOW2_CLUSTER_ZERO_ALLOC.
I tried to arrange the enum so that we could use
'ret <= QCOW2_CLUSTER_ZERO_PLAIN' for all unallocated types, and
'ret >= QCOW2_CLUSTER_ZERO_ALLOC' for allocated types, although
I didn't actually end up taking advantage of the layout.
In many cases, this leads to simpler code, by properly combining
cases (sometimes, both zero types pair together, other times,
plain zero is more like unallocated while allocated zero is more
like normal).
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-id: 20170507000552.20847-7-eblake@redhat.com
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Although it doesn't add all that much type safety (this is C, after
all), it does add a bit of legibility to use the name QCow2ClusterType
instead of a plain int.
In particular, qcow2_get_cluster_offset() has an overloaded return
type; a QCow2ClusterType on success, and -errno on failure; keeping
the cluster type in a separate variable makes it slightly easier for
the next patch to make further computations based on the type.
Suggested-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-id: 20170507000552.20847-6-eblake@redhat.com
[mreitz: Use the new type in two more places (one of them pulled from
the next patch)]
Signed-off-by: Max Reitz <mreitz@redhat.com>
Stefan Hajnoczi