It's not needed to check the return of qobject_from_jsonf()
anymore, as an assert() has been added there.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
With this commit QMP will be started in Capabilities Negotiation
mode, where the only command allowed to run is 'qmp_capabilities'.
All other commands will return CommandNotFound error. Asynchronous
messages are not delivered either.
When 'qmp_capabilities' is successfully executed QMP enters in
Command mode, where all commands (except 'qmp_capabilities') are
allowed to run and asynchronous messages are delivered.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This command will be used to enable QMP capabilities advertised
by the capabilities array.
Note that it will be mandatory to issue this command in order
to make QMP functional (although this behavior is not being
enforced by this commit).
Also, as we don't have any capabilities yet, the new command
doesn't accept any arguments. I will postpone the decision for
a format for this until we get our first capability.
Finally, this command is visible from the user Monitor too, in
the meaning that you can execute it but it won't do anything.
Making it only visible in QMP is beyond this series' goal, as
it requires changes in unrelated places.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
With capability negotiation support clients will only have a chance
to check QEMU's version (ie. issue 'query-version') after the
negotiation procedure is done.
It might be useful to clients to check QEMU's version before
negotiating features, though.
To allow that, this commit adds the QEMU's version object to the
greeting message.
Not really sure this is needed, but doesn't hurt anyway.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit adds the basic definitions for the BLOCK_IO_ERROR
event, but actual event emission will be introduced by the
next commits.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This changes the error message from "Invalid CPU index" to "Invalid
parameter index" in the human monitor.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This is a double value with optional suffixes ms, us, ns. We'll need
this to get migrate_set_downtime() QMP-ready.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This is a double value with optional suffixes G, g, M, m, K, k. We'll
need this to get migrate_set_speed() QMP-ready.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Function is_async_return() added by commit 940cc30d0d assumes
that 'data', which is returned by handlers, is always a QDict.
This is not true, as QLists can also be returned, in this case
we'll get a segfault.
Fix that by checking if 'data' is a QDict.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
When using ballooning to manage overcommitted memory on a host, a system for
guests to communicate their memory usage to the host can provide information
that will minimize the impact of ballooning on the guests. The current method
employs a daemon running in each guest that communicates memory statistics to a
host daemon at a specified time interval. The host daemon aggregates this
information and inflates and/or deflates balloons according to the level of
host memory pressure. This approach is effective but overly complex since a
daemon must be installed inside each guest and coordinated to communicate with
the host. A simpler approach is to collect memory statistics in the virtio
balloon driver and communicate them directly to the hypervisor.
Signed-off-by: Adam Litke <agl@us.ibm.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Qemu has a number of commands that can operate asynchronously (savevm, migrate,
etc) and it will be getting more. For these commands, the user monitor needs
to be suspended, but QMP monitors could continue to to accept other commands.
This patch introduces a new command API that isolates the details of handling
different monitor types from the actual command execution.
A monitor command can use this API by implementing the mhandler.cmd_async
handler (or info_async if appropriate). This function is responsible for
submitting the command and does not return any data although it may raise
errors. When the command completes, the QMPCompletion callback should be
invoked with its opaque data and the command result.
The process for submitting and completing an asynchronous command is different
for QMP and user monitors. A user monitor must be suspended at submit time and
resumed at completion time. The user_print() function must be passed to the
QMPCompletion callback so the result can be displayed properly. QMP monitors
are simpler. No submit time setup is required. When the command completes,
monitor_protocol_emitter() writes the result in JSON format.
This API can also be used to implement synchronous commands. In this case, the
cmd_async handler should immediately call the QMPCompletion callback. It is my
hope that this new interface will work for all commands, leading to a
drastically simplified monitor.c once all commands are ported.
Signed-off-by: Adam Litke <agl@us.ibm.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The returned QObject is a QList of all buses. Each bus is
represented by a QDict, which has a key with a QList of all
PCI devices attached to it. Each device is represented by
a QDict.
As has happended to other complex conversions, it's hard to
split this commit as part of it are new functions which are
called by each other.
IMPORTANT: support for printing PCI bridge attached devices
is NOT part of this commit, it's going to be added by the
next commit, as it's untested.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
CC i386-softmmu/monitor.o
cc1: warnings being treated as errors
/usr/src/RPM/BUILD/qemu-0.11.92/monitor.c: In function 'do_memory_save':
/usr/src/RPM/BUILD/qemu-0.11.92/monitor.c:1318: error: ignoring return value of 'fwrite', declared with attribute warn_unused_result
/usr/src/RPM/BUILD/qemu-0.11.92/monitor.c: In function 'do_physical_memory_save':
/usr/src/RPM/BUILD/qemu-0.11.92/monitor.c:1345: error: ignoring return value of 'fwrite', declared with attribute warn_unused_result
make[1]: *** [monitor.o] Error 1
Signed-off-by: Kirill A. Shutemov <kirill@shutemov.name>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
mon_get_cpu() can't return null pointer, because it passes its return
value to cpu_synchronize_state() first, which crashes if its argument
is null.
Remove the (pretty cheesy) handling of this non-existing error.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Commit f039a563f2 introduces
a regression as monitor_protocol_event() will return in
the first user Monitor it finds in the QLIST_FOREACH()
loop.
The right thing to do is to only delivery an asynchronous
event if the 'mon' is a QMP Monitor.
The aforementioned commit was an early version, if it was
applied to stable (it should) this one has to be applied
there too.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Instead of using the field 'readonly' of the BlockDriverState struct for passing the request,
pass the request in the flags parameter to the function.
Signed-off-by: Naphtali Sprei <nsprei@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
When using a control/QMP monitor in tandem with a regular monitor, asynchronous
messages can get lost depending on the order of the QEMU program arguments.
QEMU events issued by monitor_protocol_event() always go to cur_mon. If the
user monitor was specified on the command line first (or it has ,default), the
message will be directed to the user monitor (not the QMP monitor).
Additionally, only one QMP session is currently able to receive async messages.
To avoid this confusion, scan through the list of monitors and emit the message
on each QMP monitor.
Signed-off-by: Adam Litke <agl@us.ibm.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
It's emitted when a VNC client session is activated by QEMU,
client's information such as port, IP and auth ID (if the
session is authenticated) are provided.
Event example:
{ "event": "VNC_INITIALIZED",
"timestamp": {"seconds": 1263475302, "microseconds": 150772},
"data": {
"server": { "auth": "sasl", "family": "ipv4",
"service": "5901", "host": "0.0.0.0"},
"client": { "family": "ipv4", "service": "46089",
"host": "127.0.0.1", "sasl_username": "lcapitulino" } } }
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
It's emitted when a VNC client connects to QEMU, client's information
such as port and IP address are provided.
Note that this event is emitted right when the connection is
established. This means that it happens before authentication
procedure and session initialization.
Event example:
{ "event": "VNC_CONNECTED",
"timestamp": { "seconds": 1262976601, "microseconds": 975795 },
"data": {
"server": { "auth": "sasl", "family": "ipv4",
"service": "5901", "host": "0.0.0.0" },
"client": { "family": "ipv4", "service": "58425",
"host": "127.0.0.1" } } }
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The monitor_protocol_event() function will free the
event's data, this is wrong as 'data' management is up
to the caller.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Some commands return a QList of QDicts, which is valid,
but will trig the assert().
Just drop it.
Reported-by: Nathan Baum <nathan@parenthephobia.org.uk>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This makes do_balloon() accept megabyte values from the user
Monitor while accepting byte values for QMP.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This is a target long value in megabytes which should be
converted to bytes.
It will be used by handlers which accept a megabyte value
when in "user mode".
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Currently, when a regular command doesn't have any data to output,
QMP will emit:
{ "return": "OK" }
Returning an empty dict is better though, because dicts can support
some protocol changes in a compatible way.
So, with this commit we will return:
{ "return": {} }
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Looks like I dropped this check when addressing the 'query-'
commands request.
QMP should only handle converted commands, obviously.
Reported-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This allows to suspend command interpretation and execution
synchronously, e.g. during migration.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The monitor_vprintf() function now touches the 'mon' pointer
before calling monitor_puts(), this causes block migration
to segfault as its functions call monitor_printf() with a
NULL 'mon'.
To fix the problem this commit moves the 'mon' NULL check
from monitor_puts() to monitor_vprintf().
This can potentially hide bugs, but for some reason this has
been the behavior for a long time.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The monitor_read_command() function is readline specific
and should only be used when readline is available.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
do_balloon() should check for ballooning availability as
do_info_balloon() does.
Noted by Daniel P. Berrange <berrange@redhat.com>.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Monitor's eject command uses 'filename' for the device name
argument, but 'device' is a better name.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Return a QDict with server information. Connected clients are returned
as a QList of QDicts.
The new functions (vnc_qdict_remote_addr(), vnc_qdict_local_addr() and
put_addr_qdict()) are used to insert 'host' and 'service' information
in the returned QDict.
This patch is big, but I don't see how to split it.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Each device is represented by a QDict. The returned QObject is a QList
of all devices.
This commit should not change user output.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Each device statistic information is stored in a QDict and
the returned QObject is a QList of all devices.
This commit should not change user output.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Each block device information is stored in a QDict and the
returned QObject is a QList of all devices.
This commit should not change user output.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Return a QDict, which may contain up to more two QDicts, depending
on the type of migration we're performing.
IMPORTANT: as a QInt stores a int64_t integer, RAM values are going
to be stored as int64_t and not as uint64_t as they are today. If
this is a problem QInt will have to be changed.
This commit should not change user output.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Each mouse is represented by a QDict, the returned QObject is a QList of
all mice.
This commit should not change user output.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
snprintf() is used because the UUID_FMT is too complex for
qobject_from_jsonf().
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
All 'info' commands should use QDict, this commit also kills
monitor_print_qobject() as do_info_version() doesn't use it
anymore (and no handler will).
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
While there update the documentation as well.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Should return a QDict and should not print the user protocol bits
(eg. "c|cont").
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Monitor commands should always return values in bytes and info
commands should always return a QDict.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Also affects do_change(), because the two share eject_device().
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The do_cont() function will ask the user to enter a password if a
device is encrypted.
This is invalid under QMP, so we raise a QERR_DEVICE_ENCRYPTED
error.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
When using encrypted disk images, QEMU will prompt the user
for passwords when started.
This makes sense for the user protocol, but doesn't for QMP.
The solution is to have Monitor command which allows the user
or a Client to set passwords in advance, so that we avoid
the prompt completely.
This is what block_passwd does, for example:
(QEMU) block_passwd ide0-hd0 foobar
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
We still have handlers which will call monitor print functions
in several places. Usually to report errors.
If they do this when we are in control mode, we will be emitting
garbage to our clients.
To avoid this problem, this commit adds a way to disable those
functions. If any of them is called when in control mode, we will
emit a generic error.
Although this is far from the perfect solution, it guarantees
that only JSON is sent to Clients.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Debug, shutdown, reset, powerdown and stop are all basic events,
as they are very simple they can be added in the same commit.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Asynchronous events are generated with a call to
monitor_protocol_event().
This function builds the right data-type and emit the event
right away. The emitted data is always a JSON object and its
format is as follows:
{ "event": json-string,
"timestamp": { "seconds": json-number, "microseconds": json-number },
"data": json-value }
This design is based on ideas by Amit Shah <amit.shah@redhat.com>.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The 'info' command makes sense for the user protocol, but for QMP
it doesn't, as its return data is not well defined. That is, it
can return anything.
To fix this Avi proposes having 'query-' commands when in protocol
mode. For example, 'info balloon' would become 'query-balloon'.
The right way of supporting this would probably be to move all
info handlers to qemu-monitor.hx, add a flags field to mon_cmd_t
to identify them and then modify do_info() to do its search based
on that flag.
Unfortunately, this would require a big change in the Monitor.
To make things simpler for now, this commit takes a different
approach: a check for commands starting with "query-" is added to
toplevel QMP code, if it's true we setup things so that do_info()
is called with the appropriate arguments.
This is a hack, but is a temporary one and guarantees that query-
commands will work from the first day.
Also note that 'info' is not allowed in protocol mode.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The JSON stream parser is used to do QMP input. When there
are enough characters to be parsed it calls Monitor's
handle_qmp_command() function to handle the input.
This function's job is to check if the input is correct and
call the appropriate handler. In other words, it does for QMP
what handle_user_command() does for the user protocol.
This means that handle_qmp_command() also has to parse the
(ugly) "args_type" format to able to get the arguments names
and types expected by the handler.
The format to input commands in QMP is as follows:
{ "execute": json-string,
"id": json-value, "arguments": json-object }
Please, note that this commit also adds "id" support.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit adds specific QMP checks to do_info(), so that
it behaves as expected in QMP mode.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
In the new Monitor output is always performed by only two
functions: do_info() and monitor_call_handler().
To support QMP output, we modify those functions to check if we
are in control mode. If so, we call monitor_protocol_emitter()
to emit QMP output, otherwise we do regular output.
QMP has two types of responses to issued commands: success and
error. The outputed data is always a JSON object.
Success responses have the following format:
{ "return": json-value, "id": json-value }
Error responses have the following format:
{ "error": { "class": json-string,
"desc": json-string,
"data": json-value } "id": json-value }
Please, note that the "id" key is part of the input code, and
thus is not added in this commit.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit adds initial QMP support in QEMU. It's important
to notice that most QMP code will be part of the Monitor.
Input will be read by monitor_control_read(). Currently it
does nothing but next patches will add proper input support.
The function monitor_json_emitter(), as its name implies, is
used by the Monitor to emit JSON output. In this commit it's
used by monitor_control_event() to print our greeting message.
Finally, control mode support is also added to monitor_init(),
allowing QMP to be really enabled.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
List QMP available commands. Only valid in control mode, where
has to be used as 'query-commands.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
As this series will add a new kind of Monitor command, it's better
to rename monitor_handle_command() to what it really is:
handle_user_command().
This will avoid confusion.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit moves the loop which searches for the command
entry corresponding to a command name to its own function.
It will be used by QMP code as well.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit moves the code which calls Monitor handlers to
its own function, as it will be used by QMP code as well.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit adds a flag called 'control' to the '-monitor'
command-line option. This flag enables control mode.
The syntax is:
qemu [...] -monitor control,<device>
Where <device> is a chardev (excluding 'vc', for obvious reasons).
For example:
$ qemu [...] -monitor control,tcp:localhost:4444,server
Will run QEMU in control mode, waiting for a client TCP connection
on localhost port 4444.
NOTE: I've tried using QemuOpts for this, but turns out that it
will try to parse the device part, which should be untouched.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This flag will be set when Monitor enters "control mode", in
which the output will be defined by the QEMU Monitor Protocol.
This also introduces a macro to check if the flag is set.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit adds QError support in the Monitor.
A QError member is added to the Monitor struct. This new member
stores error information and is also used to check if an error
has occurred when the called handler returns.
Additionally, a new macro called qemu_error_new() is introduced.
It builds on top of the QemuErrorSink API and should be used in
place of qemu_error().
When all conversion to qemu_error_new() is done, qemu_error() can
be turned private.
Basically, Monitor's error flow is something like this:
1. An error occurs in the handler, it calls qemu_error_new()
2. qemu_error_new() builds a new QError object and stores it in
the Monitor struct
3. The handler returns
4. Top level Monitor code checks the Monitor struct and calls
qerror_print() to print the error
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This patch adds the option to activate non-shared storage migration from the
monitor.
The migration command is as follows:
(qemu) migrate -d tcp:0:4444 # for ordinary live migration
(qemu) migrate -d -b tcp:0:4444 # for live migration with complete storage copy
(qemu) migrate -d -i tcp:0:4444 # for live migration with incremental storage copy, storage is cow based.
Changes from v4:
- Minor coding style issues.
Signed-off-by: Liran Schour <lirans@il.ibm.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
We have code for a quite a few block formats. While I trust that all
of these formats are useful at least for some people in some
circumstances, some of them are of a kind that friends don't let
friends use in production.
This patch provides an optional block format whitelist, default off.
If a whitelist is configured with --block-drv-whitelist, QEMU proper
can use only whitelisted formats. Other programs, like qemu-img, are
not affected.
Drivers for formats off the whitelist still participate in format
probing, to ensure all programs probe exactly the same. Without that,
QEMU proper would be prone to treat images with a format off the
whitelist as raw when the image's format is probed.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Note that errors are not being converted yet.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Note that errors are not being converted yet.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Note that errors are not being converted yet.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Note that errors are not being converted yet.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Note that errors are not being converted yet.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The char event RESET is emitted when a char device is opened.
Give it a better name.
Patchworks-ID: 35287
Signed-off-by: Amit Shah <amit.shah@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Each CPU information is stored in a QDict and the returned
QObject is a QList of all CPUs.
The QDict contains the following information:
- "CPU": cpu index
- "current": "yes" or "no"
- "pc": current PC
- "halted": "yes" or "no"
The user output in the Monitor should not change and the
future monitor protocol is expected to emit something like:
[ { "CPU": 0, "current": "yes", "pc": 0x..., "halted": "no" },
{ "CPU": 1, "current": "no", "pc": 0x..., "halted": "yes" } ]
which corresponds to the following user output:
* CPU #0: pc=0x00000000fffffff0
CPU #1: pc=0x00000000fffffff0 (halted)
Patchworks-ID: 35352
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
On success return a QInt with the balloon's value.
This also introduces monitor_print_balloon() to print the
balloon information in the user protocol.
Please, note that errors are not being converted yet.
Patchworks-ID: 35351
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The returned data is always a QString.
Also introduces monitor_print_qobject(), which can be used as
a standard way to print QObjects in the user protocol format.
Patchworks-ID: 35350
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
It is important to note that it never fails, as big refactoring
of the virtio code would be needed to get the proper error code.
Patchworks-ID: 35349
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Appropriate error handling support will be needed to have
encrypted images working under the future machine protocol,
but this initial conversion will work with the current
user protocol.
Patchworks-ID: 35348
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
do_info() is special, its job is to call 'info handlers'.
This is similar to what monitor_handle_command() does,
therefore do_info() also has to distinguish among new and
old style info handlers.
This commit converts do_info() to the new QObject style and
makes the appropriate changes so that it can handle both
info handlers styles.
In the future, when all handlers are converted to QObject's
style, it will be possible to share more code with
monitor_handle_command().
This commit also introduces a new function called
monitor_user_noop(), it should be used by handlers which do
not have data to print.
This is the case of do_info().
Patchworks-ID: 35341
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit changes monitor_handle_command() to support old style
_and_ new style handlers.
New style handlers are protocol independent, they return their
data to the Monitor, which in turn decides how to print them
(ie. user protocol vs. machine protocol).
Converted handlers will use the 'user_print' member of 'mon_cmd_t'
to define its user protocol function, which will be called to print
data in the user protocol format.
Handlers which don't have 'user_print' defined are not converted
and are handled as usual.
Patchworks-ID: 35340
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This new struct member will store a pointer to a function that
should be used to output data in the user protocol format.
It will also serve as a flag to say if a given handler has already
been converted to the new QObject style.
Patchworks-ID: 35339
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commits adds a new union member to mon_cmd_t for command
handlers and convert monitor_handle_command() and qemu-monitor.hx
to use it.
This improves type safety.
Patchworks-ID: 35337
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit adds a union to mon_cmd_t for info handlers and
converts do_info() and info_cmds[] array to use it.
This improves type safety.
Next commit will convert command handlers.
Patchworks-ID: 35336
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This patch adds infrastructure to maintain memory regions which must be
restored on reset. That includes roms (vga bios and option roms on pc),
but is also used when loading linux kernels directly. Features:
- loading files is supported.
- passing blobs is supported.
- target address range is supported (for optionrom area).
- fixed target memory address is supported (linux kernel).
New in v2:
- writes to ROM are done only at initial boot.
- also handle aout and uimage loaders.
- drop unused fread_targphys() function.
The final memory layout is created once all memory regions are
registered. The option roms get addresses assigned and the
registered regions are checked against overlaps. Finally all data
is copyed to the guest memory.
Advantages:
(1) Filling memory on initial boot and on reset takes the same
code path, making reset more robust.
(2) The need to keep track of the option rom load address is gone.
(3) Due to (2) option roms can be loaded outside pc_init(). This
allows to move the pxe rom loading into the nic drivers for
example.
Additional bonus: There is a 'info roms' monitor command now.
The patch also switches over pc.c and removes the
option_rom_setup_reset() and load_option_rom() functions.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
In the very least, a change like this requires discussion on the list.
The naming convention is goofy and it causes a massive merge problem. Something
like this _must_ be presented on the list first so people can provide input
and cope with it.
This reverts commit 99a0949b72.
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The CPU state parameter is not used, remove it and adjust callers. Now we
can compile ioport.c once for all targets.
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Problem: Our file sys-queue.h is a copy of the BSD file, but there are
some additions and it's not entirely compatible. Because of that, there have
been conflicts with system headers on BSD systems. Some hacks have been
introduced in the commits 15cc923584,
f40d753718,
96555a96d7 and
3990d09adf but the fixes were fragile.
Solution: Avoid the conflict entirely by renaming the functions and the
file. Revert the previous hacks.
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
make the mux driver send mux_in and mux_out events when switching
focus while hooking up more handlers.
stop using CharDriverState->focus in monitor.c, track state using
the mux events instead. This also removes the implicit assumtion
that a muxed monitor allways has mux channel 0.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The 'i' argument type is for 32-bit only and most handlers
will use an 'int' to store its value.
It's better to fail gracefully when the user enters a value
greater than 32-bit than to get subtle casting bugs.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit drops all the code used to handle the 'args[]' array,
as now we use a dictionary to pass arguments.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
It's not used anymore, as QDict is now used to handle string
memory allocation/deallocation.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
GET_TLONG() and GET_TPHYSADDR() are not needed anymore, QInt can
handle such conversions.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
In order to help the integration with unit-tests and having a better
design, this commit splits monitor_handle_command() into two parts.
The parsing code is moved to a function called monitor_parse_command(),
while allocating memory and calling the handler is still done by
monitor_handle_command().
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit ports command handlers that receive ten arguments to use
the new monitor's dictionary.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Commit 79c4f6b080 added handler_8 and
handler_9 handling, but there isn't any command handler with those
number of arguments.
Just drop them.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit ports command handlers that receive seven arguments to
use the new monitor's dictionary.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit ports command handlers that receive six arguments to use
the new monitor's dictionary.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit ports command handlers that receive five arguments to use
the new monitor's dictionary.
Note that GET_TLONG() and GET_TPHYSADDR() macros are not used anymore.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit ports command handlers that receive four arguments to use
the new monitor's dictionary.
Note that GET_TLONG() and GET_TPHYSADDR() macros are not used anymore.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit ports command handlers that receive three arguments to use
the new monitor's dictionary.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit ports command handlers that receive two arguments to use
the new monitor's dictionary.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit ports command handlers that receive one argument to use
the new monitor's dictionary.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This commit ports command handlers that receive no arguments to use
the new monitor's dictionary.
It might seem no sense to do this, as the handlers have no arguments,
but at the end of this porting work all handlers will have the same
structure.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
With this commit monitor_handle_command() will be able to setup a
QDict with arguments to command handlers.
However, the current 'args[]' method is still being used, next
changes will port commands to get their arguments from the dictionary.
Two changes are worth noting:
1. The '/' argument type always adds the following standard keys in the
dictionary: 'count', 'format' and 'size'. This way, the argument
name used in the 'args_type' string doesn't matter
2. The optional argument type '?' doesn't need to pass the additional
'has_arg' argument, hanlders can do the same check with qdict_haskey()
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Current handlers argument types, as defined in qemu-monitor.hx file,
are a sequence of chars where each one represents one argument type
of the command handler. The number of chars is also used to know how
many arguments a given handler accepts.
This commit defines a new format, which makes mandatory the use of
a name for each argument.
For example, do_eject() command handler is currently defined as:
{ "eject", "-fB", do_eject, ... }
With the new format it becomes:
{ "eject", "force:-f,filename:B", do_eject, ... }
This way the Monitor will be capable of setting up a dictionary, using
each argument's name as the key and the argument itself as the value.
This commit also adds two new functions: key_get_info() and
next_arg_type(), both are used to parse the new format.
Currently key_get_info() consumes the 'key' part of the new format and
discards it, this way the current parsing code is not affected by this
change.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Some functions exported to be used by the Monitor as command
handlers are also called in other places as regular functions.
When those functions got ported to use the Monitor dictionary
to pass argments, the callers will have to setup a dictionary
to be able to call them.
To avoid this problem, this commit add wrappers to those functions,
so that we change the wrapper to accept the dictionary, letting
the current functions as is.
The following wrappers are being added:
- do_help_cmd()
- do_pci_device_hot_remove()
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
This patch adds some functions for error reporting to address the
problem that error messages should be routed to different destinations
depending on the context of the caller, i.e. monitor command errors
should go to the monitor, command line errors to stderr.
qemu_error() is a printf-like function to report errors.
qemu_errors_to_file() and qemu_errors_to_mon() switch the destination
for the error message to the specified file or monitor. When setting a
new destination the old one will be kept. One can switch back using
qemu_errors_to_previous(). i.e. it works like a stack.
main() calls qemu_errors_to_file(stderr), so errors go to stderr by
default. monitor callbacks are wrapped into qemu_errors_to_mon() +
qemu_errors_to_previous(), so any errors triggered by monitor commands
will go to the monitor.
Each thread has its own error message destination. qemu-kvm probably
should add a qemu_errors_to_file(stderr) call to the i/o-thread
initialization code.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
cpu_synchronize_state() is a little unreadable since the 'modified'
argument isn't self-explanatory. Simplify it by making it always
synchronize the kernel state into qemu, and automatically flush the
registers back to the kernel if they've been synchronized on this
exit.
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
kqemu introduces a number of restrictions on the i386 target. The worst is that
it prevents large memory from working in the default build.
Furthermore, kqemu is fundamentally flawed in a number of ways. It relies on
the TSC as a time source which will not be reliable on a multiple processor
system in userspace. Since most modern processors are multicore, this severely
limits the utility of kqemu.
kvm is a viable alternative for people looking to accelerate qemu and has the
benefit of being supported by the upstream Linux kernel. If someone can
implement work arounds to remove the restrictions introduced by kqemu, I'm
happy to avoid and/or revert this patch.
N.B. kqemu will still function in the 0.11 series but this patch removes it from
the 0.12 series.
Paul, please Ack or Nack this patch.
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Commit 751c6a1704 changed the monitor's
'commit' command to this behavior:
1. Any string you type as argument will cause do_commit() to
call bdrv_commit() to all devices
2. If you enter a device name, it will be the only one ignored
by do_commit() :)
The fix is to call bdrv_commit() to the specified device only and
ignore the others (when 'all' is not specified).
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Message-Id:
First step cleaning up the drives handling. This one does nothing but
removing drives_table[], still it became seriously big.
drive_get_index() is gone and is replaced by drives_get() which hands
out DriveInfo pointers instead of a table index. This needs adaption in
*tons* of places all over.
The drives are now maintained as linked list.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Add monitor commands to support passing file descriptors via
SCM_RIGHTS.
getfd assigns the passed file descriptor a name for use with other
monitor commands.
closefd allows passed file descriptors to be closed. If a monitor
command actually uses a named file descriptor, closefd will not be
required.
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Useful for testing hardware emulations or manipulating its state to
stress guest drivers.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Move registration function for the boot_set callback handler and provide
qemu_boot_set so that it can also be used outside the monitor code.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
- MCE features are initialized when VCPU is intialized according to CPUID.
- A monitor command "mce" is added to inject a MCE.
- A new interrupt mask: CPU_INTERRUPT_MCE is added to inject the MCE.
aliguori: fix build for linux-user
Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Refactor the ACL monitor interface to make full use of the monitor
command dispatcher. This also gives proper help formatting and command
completion. Note that 'acl allow' and 'acl deny' were combined to
'acl_add aclname match allow|deny [index]' for consistency reasons.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
