The dtrace systemtap trace backend for QEMU is very powerful but it is
also somewhat unfriendly to users who aren't familiar with systemtap,
or who don't need its power right now.
stap -e "....some strange script...."
The 'log' backend for QEMU by comparison is very crude but incredibly
easy to use:
$ qemu -d trace:qio* ...some args...
23266@1547735759.137292:qio_channel_socket_new Socket new ioc=0x563a8a39d400
23266@1547735759.137305:qio_task_new Task new task=0x563a891d0570 source=0x563a8a39d400 func=0x563a86f1e6c0 opaque=0x563a89078000
23266@1547735759.137326:qio_task_thread_start Task thread start task=0x563a891d0570 worker=0x563a86f1ce50 opaque=0x563a891d9d90
23273@1547735759.137491:qio_task_thread_run Task thread run task=0x563a891d0570
23273@1547735759.137503:qio_channel_socket_connect_sync Socket connect sync ioc=0x563a8a39d400 addr=0x563a891d9d90
23273@1547735759.138108:qio_channel_socket_connect_fail Socket connect fail ioc=0x563a8a39d400
This commit introduces a way to do simple printf style logging of probe
points using systemtap. In particular it creates another set of tapsets,
one per emulator:
/usr/share/systemtap/tapset/qemu-*-log.stp
These pre-define probe functions which simply call printf() on their
arguments. The printf() format string is taken from the normal
trace-events files, with a little munging to the format specifiers
to cope with systemtap's more restrictive syntax.
With this you can now do
$ stap -e 'probe qemu.system.x86_64.log.qio*{}'
22806@1547735341399856820 qio_channel_socket_new Socket new ioc=0x56135d1d7c00
22806@1547735341399862570 qio_task_new Task new task=0x56135cd66eb0 source=0x56135d1d7c00 func=0x56135af746c0 opaque=0x56135bf06400
22806@1547735341399865943 qio_task_thread_start Task thread start task=0x56135cd66eb0 worker=0x56135af72e50 opaque=0x56135c071d70
22806@1547735341399976816 qio_task_thread_run Task thread run task=0x56135cd66eb0
We go one step further though and introduce a 'qemu-trace-stap' tool to
make this even easier
$ qemu-trace-stap run qemu-system-x86_64 'qio*'
22806@1547735341399856820 qio_channel_socket_new Socket new ioc=0x56135d1d7c00
22806@1547735341399862570 qio_task_new Task new task=0x56135cd66eb0 source=0x56135d1d7c00 func=0x56135af746c0 opaque=0x56135bf06400
22806@1547735341399865943 qio_task_thread_start Task thread start task=0x56135cd66eb0 worker=0x56135af72e50 opaque=0x56135c071d70
22806@1547735341399976816 qio_task_thread_run Task thread run task=0x56135cd66eb0
This tool is clever in that it will automatically change the
SYSTEMTAP_TAPSET env variable to point to the directory containing the
right set of probes for the QEMU binary path you give it. This is useful
if you have QEMU installed in /usr but are trying to test and trace a
binary in /home/berrange/usr/qemu-git. In that case you'd do
$ qemu-trace-stap run /home/berrange/usr/qemu-git/bin/qemu-system-x86_64 'qio*'
And it'll make sure /home/berrange/usr/qemu-git/share/systemtap/tapset
is used for the trace session
The 'qemu-trace-stap' script takes a verbose arg so you can understand
what it is running
$ qemu-trace-stap run /home/berrange/usr/qemu-git/bin/qemu-system-x86_64 'qio*'
Using tapset dir '/home/berrange/usr/qemu-git/share/systemtap/tapset' for binary '/home/berrange/usr/qemu-git/bin/qemu-system-x86_64'
Compiling script 'probe qemu.system.x86_64.log.qio* {}'
Running script, <Ctrl>-c to quit
...trace output...
It can enable multiple probes at once
$ qemu-trace-stap run qemu-system-x86_64 'qio*' 'qcrypto*' 'buffer*'
By default it monitors all existing running processes and all future
launched proceses. This can be restricted to a specific PID using the
--pid arg
$ qemu-trace-stap run --pid 2532 qemu-system-x86_64 'qio*'
Finally if you can't remember what probes are valid it can tell you
$ qemu-trace-stap list qemu-system-x86_64
ahci_check_irq
ahci_cmd_done
ahci_dma_prepare_buf
ahci_dma_prepare_buf_fail
ahci_dma_rw_buf
ahci_irq_lower
...snip...
Or list just those matching a prefix pattern
$ qemu-trace-stap list -v qemu-system-x86_64 'qio*'
Using tapset dir '/home/berrange/usr/qemu-git/share/systemtap/tapset' for binary '/home/berrange/usr/qemu-git/bin/qemu-system-x86_64'
Listing probes with name 'qemu.system.x86_64.log.qio*'
qio_channel_command_abort
qio_channel_command_new_pid
qio_channel_command_new_spawn
qio_channel_command_wait
qio_channel_file_new_fd
...snip...
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Message-id: 20190123120016.4538-5-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
In some cases it may be helpful to modify state before saving it for
migration, and then modify the state back after it has been saved. The
existing pre_save function provides half of this functionality. This
patch adds a post_save function to provide the second half.
Signed-off-by: Aaron Lindsay <aclindsa@gmail.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-id: 20181211151945.29137-2-aaron@os.amperecomputing.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The API of cpu_physical_memory_write_rom() is odd, because it
takes an AddressSpace, unlike all the other cpu_physical_memory_*
access functions. Rename it to address_space_write_rom(), and
bring its API into line with address_space_write().
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Message-id: 20181122133507.30950-3-peter.maydell@linaro.org
The generated code is for now *unconditional*. Later patches generate
the conditionals.
Note that union discriminators may not have 'if' conditionals.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20181213123724.4866-14-marcandre.lureau@redhat.com>
Message-Id: <20181213123724.4866-15-marcandre.lureau@redhat.com>
[Patches squashed, commit message tweaked]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
QAPISchemaMember gains .ifcond for enum members: inherited classes,
such as QAPISchemaObjectTypeMember, will thus have an ifcond member
after this (those different types will also use the .ifcond to store
the condition and generate conditional code in the following patches).
The generated code remains unconditional for now. Later patches
generate the conditionals.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20181213123724.4866-10-marcandre.lureau@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
RFC8259 obsoletes RFC7159. Fix a couple of URLs to point to the
newer version.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20181203175702.128701-1-eblake@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Add a new flag to mark memory region that are used as non-volatile, by
NVDIMM for example. That bit is propagated down to the flat view, and
reflected in HMP info mtree with a "nv-" prefix on the memory type.
This way, guest_phys_blocks_region_add() can skip the NV memory
regions for dumps and TCG memory clear in a following patch.
Cc: dgilbert@redhat.com
Cc: imammedo@redhat.com
Cc: pbonzini@redhat.com
Cc: guangrong.xiao@linux.intel.com
Cc: mst@redhat.com
Cc: xiaoguangrong.eric@gmail.com
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20181003114454.5662-2-marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The acceptance (aka functional, aka Avocado-based) tests are
Python files located in "tests/acceptance" that need to be run
with the Avocado libs and test runner.
Let's provide a convenient way for QEMU developers to run them,
by making use of the tests-venv with the required setup.
Also, while the Avocado test runner will take care of creating a
location to save test results to, it was understood that it's better
if the results are kept within the build tree.
Signed-off-by: Cleber Rosa <crosa@redhat.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: Caio Carrara <ccarrara@redhat.com>
Message-Id: <20181018153134.8493-3-crosa@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
The line immediate following a ".. code::" block is considered
to contains arguments to the "code directive". The lack of a
new line gives me during at parse time:
testing.rst:63: (ERROR/3) Error in "code" directive:
maximum 1 argument(s) allowed, 3 supplied.
.. code::
make check-unit V=1
testing.rst:120: (ERROR/3) Error in "code" directive:
maximum 1 argument(s) allowed, 3 supplied.
.. code::
make check-qtest V=1
Let's add the missing newlines, both for consistency and to
avoid the parsing errors.
Signed-off-by: Cleber Rosa <crosa@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-Id: <20181004161852.11673-6-crosa@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Commit 31d2dda ("build-system: remove per-test GCOV reporting", 2018-06-20)
removed users of the variables, since those uses can be replaced by a simple
overall report produced by gcovr. However, the variables were never removed.
Do it now.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
[thuth: Fixed up contextual conflicts with the patch from Eric]
Signed-off-by: Thomas Huth <thuth@redhat.com>
When we added the _with_attrs accessors we forgot to mention
them in the documentation.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20180824170422.5783-4-peter.maydell@linaro.org>
Based-on: <20180802174042.29234-1-peter.maydell@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Now that all the users of old_mmio MemoryRegion accessors
have been converted, we can remove the core code support.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20180824170422.5783-2-peter.maydell@linaro.org>
Based-on: <20180802174042.29234-1-peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
We consciously chose in commit 1a9a507b to hide QAPI type names
from the introspection output on the wire, but added a command
line option -u to unmask the type name when doing a debug build.
The unmask option still remains useful to some other forms of
automated analysis, so it will not be removed; however, when it
is not in use, the generated .c file can be hard to read. At
the time when we first introduced masking, the generated file
consisted only of a monolithic C string, so there was no clean
way to inject any comments.
Later, in commit 7d0f982b, we switched the generation to output
a QLit object, in part to make it easier for future addition of
conditional compilation. In fact, commit d626b6c1 took advantage
of this by passing a tuple instead of a bare object for encoding
the output of conditionals. By extending that tuple, we can now
interject strategic comments.
For now, type name debug aid comments are only output once per
meta-type, rather than at all uses of the number used to encode
the type within the introspection data. But this is still a lot
more convenient than having to regenerate the file with the
unmask operation temporarily turned on - merely search the
generated file for '"NNN" =' to learn the corresponding source
name and associated definition of type NNN.
The generated qapi-introspect.c changes only with the addition
of comments, such as:
| @@ -14755,6 +15240,7 @@
| { "name", QLIT_QSTR("[485]"), },
| {}
| })),
| + /* "485" = QCryptoBlockInfoLUKSSlot */
| QLIT_QDICT(((QLitDictEntry[]) {
| { "members", QLIT_QLIST(((QLitObject[]) {
| QLIT_QDICT(((QLitDictEntry[]) {
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20180827213943.33524-3-eblake@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Rebased, update to qapi-code-gen.txt corrected]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20180828120736.32323-3-armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
The generated qapi_event_send_FOO() take an Error ** argument. They
can't actually fail, because all they do with the argument is passing it
to functions that can't fail: the QObject output visitor, and the
@qmp_emit callback, which is either monitor_qapi_event_queue() or
event_test_emit().
Drop the argument, and pass &error_abort to the QObject output visitor
and @qmp_emit instead.
Suggested-by: Eric Blake <eblake@redhat.com>
Suggested-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20180815133747.25032-4-peterx@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Commit message rewritten, update to qapi-code-gen.txt corrected]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Clarify that the pre_load function in a subsection is only called if
the subsection is found; to handle a missing subsection you may
set values in the pre_load of the parent vmsd.
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Invoking 'make vm-build-freebsd' and friends with V=1 should
propagate that verbosity setting down into the build run
inside the VM. Make sure we do that. This brings it into
line with how the container tests handle V=1.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20180803085230.30574-4-peter.maydell@linaro.org>
Signed-off-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The '-trace' and '--trace' spellings are only both supported in qemu
binary, while for qemu-nbd or qemu-img only '--trace' spelling is
supported. So for the consistency of trace option invocation, we
should use double-dash spelling in our documentation.
This's also mentioned in
https://wiki.qemu.org/BiteSizedTasks#Consistent_option_usage_in_documentation
.
Signed-off-by: Yaowei Bai <baiyaowei@cmss.chinamobile.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 1530674247-31200-1-git-send-email-baiyaowei@cmss.chinamobile.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Ville Skyttä <ville.skytta@iki.fi>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 20180612065150.21110-1-ville.skytta@iki.fi
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
This will build a coverage report under the current directory in
reports/coverage. At the users option a report can be generated by
directly invoking something like:
make foo/bar/coverage-report.html
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
This can be used to remove any stale coverage data before any
particular test run. This is useful for analysing individual tests.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org>---
I'm not entirely sure who's using this information and certainly in a
CI environment it just washes over as additional noise. Later patches
will provide new reporting options so a user who wants to analyse
individual tests will be able to use that to get the information.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20180703085358.13941-33-armbru@redhat.com>
Commit cf869d5317 "qmp: support out-of-band (oob) execution" added a
general mechanism for command-independent arguments just for an
out-of-band flag:
The "control" key is introduced to store this extra flag. "control"
field is used to store arguments that are shared by all the commands,
rather than command specific arguments. Let "run-oob" be the first.
However, it failed to reject unknown members of "control". For
instance, in QMP command
{"execute": "query-name", "id": 42, "control": {"crap": true}}
"crap" gets silently ignored.
Instead of fixing this, revert the general "control" mechanism
(because YAGNI), and do it the way I initially proposed, with key
"exec-oob". Simpler code, simpler interface.
An out-of-band command
{"execute": "migrate-pause", "id": 42, "control": {"run-oob": true}}
becomes
{"exec-oob": "migrate-pause", "id": 42}
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20180703085358.13941-13-armbru@redhat.com>
[Commit message typo fixed]
Accept 'if' key in top-level elements, accepted as string or list of
string type. The following patches will modify the test visitor to
check the value is correctly saved, and generate #if/#endif code (as a
single #if/endif line or a series for a list).
Example of 'if' key:
{ 'struct': 'TestIfStruct', 'data': { 'foo': 'int' },
'if': 'defined(TEST_IF_STRUCT)' }
The generated code is for now *unconditional*. Later patches generate
the conditionals.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20180703155648.11933-2-marcandre.lureau@redhat.com>
[Commit message and Documentation improved]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Affects documentation and a few error messages.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20180703085358.13941-2-armbru@redhat.com>
Out-Of-Band handlers need to protect shared state if there is any.
Mention it in the document. Meanwhile, touch up some other places too,
either with better English, or reordering of bullets.
Suggested-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20180620073223.31964-6-peterx@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Only one existing trace event uses a floating point type. Unfortunately
float and double cannot be supported since SystemTap does not have
floating point types.
Remove float and double from the whitelist and document this limitation.
Update the migrate_transferred trace event to use uint64_t instead of
double.
Cc: Dr. David Alan Gilbert <dgilbert@redhat.com>
Cc: Daniel P. Berrangé <berrange@redhat.com>
Cc: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Message-id: 20180621150254.4922-1-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
It often happens that just a few discriminator values imply extra data in
a flat union. Existing checks did not make possible to leave other values
uncovered. Such cases had to be worked around by either stating a dummy
(empty) type or introducing another (subset) discriminator enumeration.
Both options create redundant entities in qapi files for little profit.
With this patch it is not necessary anymore to add designated union
fields for every possible value of a discriminator enumeration.
Signed-off-by: Anton Nefedov <anton.nefedov@virtuozzo.com>
Message-Id: <1529311206-76847-2-git-send-email-anton.nefedov@virtuozzo.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
This patch adds the very minimum infrastructure necessary for writing
and running functional/acceptance tests, including:
* Documentation
* The avocado_qemu.Test base test class
* One example tests (version.py)
Additional functionality is expected to be added along the tests that
require them.
Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20180530184156.15634-2-crosa@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
[ehabkost: fix typo on testing.rst]
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Use mmap_lock in user-mode to protect TCG state and the page descriptors.
In !user-mode, each vCPU has its own TCG state, so no locks needed.
Per-page locks are used to protect the page descriptors.
Per-TB locks are used in both modes to protect TB jumps.
Some notes:
- tb_lock is removed from notdirty_mem_write by passing a
locked page_collection to tb_invalidate_phys_page_fast.
- tcg_tb_lookup/remove/insert/etc have their own internal lock(s),
so there is no need to further serialize access to them.
- do_tb_flush is run in a safe async context, meaning no other
vCPU threads are running. Therefore acquiring mmap_lock there
is just to please tools such as thread sanitizer.
- Not visible in the diff, but tb_invalidate_phys_page already
has an assert_memory_lock.
- cpu_io_recompile is !user-only, so no mmap_lock there.
- Added mmap_unlock()'s before all siglongjmp's that could
be called in user-mode while mmap_lock is held.
+ Added an assert for !have_mmap_lock() after returning from
the longjmp in cpu_exec, just like we do in cpu_exec_step_atomic.
Performance numbers before/after:
Host: AMD Opteron(tm) Processor 6376
ubuntu 17.04 ppc64 bootup+shutdown time
700 +-+--+----+------+------------+-----------+------------*--+-+
| + + + + + *B |
| before ***B*** ** * |
|tb lock removal ###D### *** |
600 +-+ *** +-+
| ** # |
| *B* #D |
| *** * ## |
500 +-+ *** ### +-+
| * *** ### |
| *B* # ## |
| ** * #D# |
400 +-+ ** ## +-+
| ** ### |
| ** ## |
| ** # ## |
300 +-+ * B* #D# +-+
| B *** ### |
| * ** #### |
| * *** ### |
200 +-+ B *B #D# +-+
| #B* * ## # |
| #* ## |
| + D##D# + + + + |
100 +-+--+----+------+------------+-----------+------------+--+-+
1 8 16 Guest CPUs 48 64
png: https://imgur.com/HwmBHXe
debian jessie aarch64 bootup+shutdown time
90 +-+--+-----+-----+------------+------------+------------+--+-+
| + + + + + + |
| before ***B*** B |
80 +tb lock removal ###D### **D +-+
| **### |
| **## |
70 +-+ ** # +-+
| ** ## |
| ** # |
60 +-+ *B ## +-+
| ** ## |
| *** #D |
50 +-+ *** ## +-+
| * ** ### |
| **B* ### |
40 +-+ **** # ## +-+
| **** #D# |
| ***B** ### |
30 +-+ B***B** #### +-+
| B * * # ### |
| B ###D# |
20 +-+ D ##D## +-+
| D# |
| + + + + + + |
10 +-+--+-----+-----+------------+------------+------------+--+-+
1 8 16 Guest CPUs 48 64
png: https://imgur.com/iGpGFtv
The gains are high for 4-8 CPUs. Beyond that point, however, unrelated
lock contention significantly hurts scalability.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Emilio G. Cota <cota@braap.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
This applies to both user-mode and !user-mode emulation.
Instead of relying on a global lock, protect the list of incoming
jumps with tb->jmp_lock. This lock also protects tb->cflags,
so update all tb->cflags readers outside tb->jmp_lock to use
atomic reads via tb_cflags().
In order to find the destination TB (and therefore its jmp_lock)
from the origin TB, we introduce tb->jmp_dest[].
I considered not using a linked list of jumps, which simplifies
code and makes the struct smaller. However, it unnecessarily increases
memory usage, which results in a performance decrease. See for
instance these numbers booting+shutting down debian-arm:
Time (s) Rel. err (%) Abs. err (s) Rel. slowdown (%)
------------------------------------------------------------------------------
before 20.88 0.74 0.154512 0.
after 20.81 0.38 0.079078 -0.33524904
GTree 21.02 0.28 0.058856 0.67049808
GHashTable + xxhash 21.63 1.08 0.233604 3.5919540
Using a hash table or a binary tree to keep track of the jumps
doesn't really pay off, not only due to the increased memory usage,
but also because most TBs have only 0 or 1 jumps to them. The maximum
number of jumps when booting debian-arm that I measured is 35, but
as we can see in the histogram below a TB with that many incoming jumps
is extremely rare; the average TB has 0.80 incoming jumps.
n_jumps: 379208; avg jumps/tb: 0.801099
dist: [0.0,1.0)|▄█▁▁▁▁▁▁▁▁▁▁▁ ▁▁▁▁▁▁ ▁▁▁ ▁▁▁ ▁|[34.0,35.0]
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Emilio G. Cota <cota@braap.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Use the recently-gained QHT feature of returning the matching TB if it
already exists. This allows us to get rid of the lookup we perform
right after acquiring tb_lock.
Suggested-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Emilio G. Cota <cota@braap.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Groundwork for supporting parallel TCG generation.
We never remove entries from the radix tree, so we can use cmpxchg
to implement lockless insertions.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Emilio G. Cota <cota@braap.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
There's a common pattern in QEMU where a function needs to perform
a data load or store of an N byte integer in a particular endianness.
At the moment this is handled by doing a switch() on the size and
calling the appropriate ld*_p or st*_p function for each size.
Provide a new family of functions ldn_*_p() and stn_*_p() which
take the size as an argument and do the switch() themselves.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20180611171007.4165-2-peter.maydell@linaro.org
The function has been deprecated for 2.5 years, and there are just a handful
of users. Convert them to memory_region_init_io with NULL callbacks,
and while at it pass the right device as the owner.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
New option will be used to allow commands, which are prepared/need
to run, during preconfig state. Other commands that should be able
to run in preconfig state, should be amended to not expect machine
in initialized state or deal with it.
For compatibility reasons, commands that don't use new flag
'allow-preconfig' explicitly are not permitted to run in
preconfig state but allowed in all other states like they used
to be.
Within this patch allow following commands in preconfig state:
qmp_capabilities
query-qmp-schema
query-commands
query-command-line-options
query-status
exit-preconfig
to allow qmp connection, basic introspection and moving to the next
state.
PS:
set-numa-node and query-hotpluggable-cpus will be enabled later in
a separate patches.
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <1526057503-39287-1-git-send-email-imammedo@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[ehabkost: Changed "since 2.13" to "since 3.0"]
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Update the migration docs:
Among other changes:
* Added a general list of advice for device authors
* Reordered the section on conditional state (subsections etc)
into the order we prefer.
* Add a note about firmware
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Balamuruhan S <bala24@linux.vnet.ibm.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Message-Id: <20180503191059.19576-1-dgilbert@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Now that we can safely call QOBJECT() on QObject * as well as its
subtypes, we can have macros qobject_ref() / qobject_unref() that work
everywhere instead of having to use QINCREF() / QDECREF() for QObject
and qobject_incref() / qobject_decref() for its subtypes.
The replacement is mechanical, except I broke a long line, and added a
cast in monitor_qmp_cleanup_req_queue_locked(). Unlike
qobject_decref(), qobject_unref() doesn't accept void *.
Note that the new macros evaluate their argument exactly once, thus no
need to shout them.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20180419150145.24795-4-marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Rebased, semantic conflict resolved, commit message improved]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Alexey Perevalov <a.perevalov@samsung.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Message-Id: <1521742647-25550-5-git-send-email-a.perevalov@samsung.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Add some notes to the migration documentation for shared memory
postcopy.
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
