Define two new migration parameters to be used with TLS encryption.
The 'tls-creds' parameter provides the ID of an instance of the
'tls-creds' object type, or rather a subclass such as 'tls-creds-x509'.
Providing these credentials will enable use of TLS on the migration
data stream.

If using x509 certificates, together with a migration URI that does
not include a hostname, the 'tls-hostname' parameter provides the
hostname to use when verifying the server's x509 certificate. This
allows TLS to be used in combination with fd: and exec: protocols
where a TCP connection is established by a 3rd party outside of
QEMU.

NB, this requires changing the migrate_set_parameter method in the
HMP to accept a 's' (string) value instead of 'i' (integer). This
is backwards compatible, because the parsing of strings allows the
quotes to be optional, thus any integer is also a valid string.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-26-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

The MigrateState struct uses an array for storing migration
parameters. This presumes that all future parameters will
be integers too, which is not going to be the case. There
is no functional reason why an array is used, if anything
it makes the code less clear. The QAPI schema already
defines a struct - MigrationParameters - capable of storing
all the individual parameters, so just use that instead of
an array.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-25-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

Now that the memory buffer based QEMUFile impl is gone, there
is no need for any backend to be accessing internals of the
QEMUFile struct, so it can be moved back into qemu-file.c

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-24-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

Now that the exec migration backend and savevm have converted
to use the QIOChannel based QEMUFile, there is no user remaining
for the stdio based QEMUFile impl and it can be deleted.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-23-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

Now that the tcp, unix and fd migration backends have converted
to use the QIOChannel based QEMUFile, there is no user remaining
for the sockets based QEMUFile impl and it can be deleted.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-22-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

Now that we don't have have a buffer based QemuFile
implementation, the QEMUSizedBuffer code is also
unused and can be deleted. A simpler buffer class
also exists in util/buffer.c which other code can
used as needed.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-21-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

The qemu_bufopen() method is no longer used, so the memory
buffer based QEMUFile backend can be deleted entirely.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-20-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

Convert the exec savevm code to use QIOChannel and QEMUFileChannel,
instead of the stdio APIs.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-19-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

This converts the RDMA code to provide a subclass of QIOChannel
that uses RDMA for the data transport.

This implementation of RDMA does not correctly handle non-blocking
mode. Reads might block if there was not already some pending data
and writes will block until all data is sent. This flawed behaviour
was already present in the existing impl, so appears to not be a
critical problem at this time. It should be on the list of things
to fix in the future though.

The RDMA code would be much better off it it could be split up in
a generic RDMA layer, a QIOChannel impl based on RMDA, and then
the RMDA migration glue. This is left as a future exercise for
the brave.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-18-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

Convert the exec socket migration protocol driver to use
QIOChannel and QEMUFileChannel, instead of the stdio
popen APIs. It can be unconditionally built because the
QIOChannelCommand class can report suitable error messages
on platforms which can't fork processes.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-17-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

Convert the fd socket migration protocol driver to use
QIOChannel and QEMUFileChannel, instead of plain sockets
APIs. It can be unconditionally built because the
QIOChannel APIs it uses will take care to report suitable
error messages if needed.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-16-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

Drop the current TCP socket migration driver and extend
the new generic socket driver to cope with the TCP address
format

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-15-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

The unix.c file will be nearly the same as the tcp.c file,
only differing in the initial SocketAddress creation code.
Rename unix.c to socket.c and refactor it a little to
prepare for merging the TCP code.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-14-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

Convert the unix socket migration protocol driver to use
QIOChannel and QEMUFileChannel, instead of plain sockets
APIs. It can be unconditionally built, since the socket
impl of QIOChannel will report a suitable error on platforms
where UNIX sockets are unavailable.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-13-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

The post-copy code does some I/O to/from an intermediate
in-memory buffer rather than direct to the underlying
I/O channel. Switch this code to use QIOChannelBuffer
instead of QEMUSizedBuffer.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-12-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

Currently if an application initiates an outgoing migration,
it may or may not, get an error reported back on failure. If
the error occurs synchronously to the 'migrate' command
execution, the client app will see the error message. This
is the case for DNS lookup failures. If the error occurs
asynchronously to the monitor command though, the error
will be thrown away and the client left guessing about
what went wrong. This is the case for failure to connect
to the TCP server (eg due to wrong port, or firewall
rules, or other similar errors).

In the future we'll be adding more scope for errors to
happen asynchronously with the TLS protocol handshake.
TLS errors are hard to diagnose even when they are well
reported, so discarding errors entirely will make it
impossible to debug TLS connection problems.

Management apps which do migration are already using
'query-migrate' / 'info migrate' to check up on progress
of background migration operations and to see their end
status. This is a fine place to also include the error
message when things go wrong.

This patch thus adds an 'error-desc' field to the
MigrationInfo struct, which will be populated when
the 'status' is set to 'failed':

(qemu) migrate -d tcp:localhost:9001
(qemu) info migrate
capabilities: xbzrle: off rdma-pin-all: off auto-converge: off zero-blocks: off compress: off events: off x-postcopy-ram: off
Migration status: failed (Error connecting to socket: Connection refused)
total time: 0 milliseconds

In the HMP, when doing non-detached migration, it is
also possible to display this error message directly
to the app.

(qemu) migrate tcp:localhost:9001
Error connecting to socket: Connection refused

Or with QMP

  {
    "execute": "query-migrate",
    "arguments": {}
  }
  {
    "return": {
      "status": "failed",
      "error-desc": "address resolution failed for myhost:9000: No address associated with hostname"
    }
  }

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Message-Id: <1461751518-12128-11-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

Currently creating a QEMUFile instance from a QIOChannel is
quite simple only requiring a single call to
qemu_fopen_channel_input or  qemu_fopen_channel_output
depending on the end of migration connection.

When QEMU gains TLS support, however, there will need to be
a TLS negotiation done inbetween creation of the QIOChannel
and creation of the final QEMUFile. Introduce some helper
methods that will encapsulate this logic, isolating the
migration protocol drivers from knowledge about TLS.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Acked-by: Juan Quintela <quintela@redhat.com>
Message-Id: <1461751518-12128-10-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

Introduce a new QEMUFile implementation that is based on
the QIOChannel objects. This impl is different from existing
impls in that there is no file descriptor that can be made
available, as some channels may be based on higher level
protocols such as TLS.

Although the QIOChannel based implementation can trivially
provide a bi-directional stream, initially we have separate
functions for opening input & output directions to fit with
the expectation of the current QEMUFile interface.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-Id: <1461751518-12128-9-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

Instead of relying on the default QEMUFile I/O blocking flag
state, explicitly turn on blocking I/O for outgoing migration
since it takes place in a background thread.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Message-Id: <1461751518-12128-8-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

Remove the assumption that every QEMUFile implementation has
a file descriptor available by introducing a new function
in QEMUFileOps to change the blocking state of a QEMUFile.

If not set, it will fallback to the original code using
the get_fd method.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Message-Id: <1461751518-12128-7-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

The QEMUFileOps struct contains the I/O subsystem callbacks
and the migration stage hooks. Split the hooks out into a
separate QEMUFileHooks struct to make it easier to refactor
the I/O side of QEMUFile without affecting the hooks.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Message-Id: <1461751518-12128-6-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

The QEMUFile writev_buffer / put_buffer functions are expected
to write out the full set of requested data, blocking until
complete. The qemu_fflush() caller does not expect to deal with
partial writes. Clarify the function comments and add a sanity
check to the code to catch mistaken implementations.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Message-Id: <1461751518-12128-5-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

Some of the test-vmstate.c test cases use a temporary file
while others use a memory buffer. To facilitate the future
removal of the qemu_bufopen() function, convert all the tests
to use a temporary file.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Message-Id: <1461751518-12128-4-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

The QIOChannelBuffer's close implementation will free
the internal data buffer. It failed to reset the pointer
to NULL though, so when the object is later finalized
it will free it a second time with predictable crash.

Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Message-Id: <1461751518-12128-3-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

The s390 skeys monitor command needs to write out a plain text
file. Currently it is using the QEMUFile class for this, but
work is ongoing to refactor QEMUFile and eliminate much code
related to it. The only feature qemu_fopen() gives over fopen()
is support for QEMU FD passing, but this can be achieved with
qemu_open() + fdopen() too. Switching to regular stdio FILE
APIs avoids the need to sprintf via an intermedia buffer which
slightly simplifies the code.

Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Message-Id: <1461751518-12128-2-git-send-email-berrange@redhat.com>
Signed-off-by: Amit Shah <amit.shah@redhat.com>

X86 queue, 2016-05-23

# gpg: Signature made Mon 23 May 2016 23:48:27 BST using RSA key ID 984DC5A6
# gpg: Good signature from "Eduardo Habkost <ehabkost@redhat.com>"

* remotes/ehabkost/tags/x86-pull-request:
  target-i386: kvm: Eliminate kvm_msr_entry_set()
  target-i386: kvm: Simplify MSR setting functions
  target-i386: kvm: Simplify MSR array construction
  target-i386: kvm: Increase MSR_BUF_SIZE
  target-i386: kvm: Allocate kvm_msrs struct once per VCPU
  target-i386: Call cpu_exec_init() on realize
  target-i386: Move TCG initialization to realize time
  target-i386: Move TCG initialization check to tcg_x86_init()
  cpu: Eliminate cpudef_init(), cpudef_setup()
  target-i386: Set constant model_id for qemu64/qemu32/athlon
  pc: Set CPU model-id on compat_props for pc <= 2.4
  osdep: Move default qemu_hw_version() value to a macro
  target-i386: kvm: Use X86XSaveArea struct for xsave save/load
  target-i386: Use xsave structs for ext_save_area
  target-i386: Define structs for layout of xsave area

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

migration fixes:

- ensure src block devices continue fine after a failed migration
- fail on migration blockers; helps 9p savevm/loadvm
- move autoconverge commands out of experimental state
- move the migration-specific qjson in migration/

# gpg: Signature made Mon 23 May 2016 18:15:09 BST using RSA key ID 657EF670
# gpg: Good signature from "Amit Shah <amit@amitshah.net>"
# gpg:                 aka "Amit Shah <amit@kernel.org>"
# gpg:                 aka "Amit Shah <amitshah@gmx.net>"

* remotes/amit-migration/tags/migration-2.7-1:
  migration: regain control of images when migration fails to complete
  savevm: fail if migration blockers are present
  migration: Promote improved autoconverge commands out of experimental state
  migration/qjson: Drop gratuitous use of QOM
  migration: Move qjson.[ch] to migration/

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

rng: rename RndRandom to RndRandom

# gpg: Signature made Mon 23 May 2016 16:44:58 BST using RSA key ID 657EF670
# gpg: Good signature from "Amit Shah <amit@amitshah.net>"
# gpg:                 aka "Amit Shah <amit@kernel.org>"
# gpg:                 aka "Amit Shah <amitshah@gmx.net>"

* remotes/amit-virtio-rng/tags/rng-2.7-1:
  rng-random: rename RndRandom to RngRandom

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

opencores_eth cleanups:
- use mii.h
- reduce stack usage in open_eth_start_xmit.

# gpg: Signature made Mon 23 May 2016 20:14:20 BST using RSA key ID F83FA044
# gpg: Good signature from "Max Filippov <max.filippov@cogentembedded.com>"
# gpg:                 aka "Max Filippov <jcmvbkbc@gmail.com>"

* remotes/xtensa/tags/20160523-opencores_eth:
  hw/net/opencores_eth: Allocating Large sized arrays to heap
  hw/net/opencores_eth: use mii.h

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Inline the function inside kvm_msr_entry_add().

Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

Simplify kvm_put_tscdeadline_msr() and
kvm_put_msr_feature_control() using kvm_msr_buf and the
kvm_msr_entry_add() helper.

Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

Add a helper function that appends new entries to the MSR buffer
and checks for the buffer size limit.

Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

We are dangerously close to the array limits in kvm_put_msrs()
and kvm_get_msrs(): with the default mcg_cap configuration, we
can set up to 148 MSRs in kvm_put_msrs(), and if we allow mcg_cap
to be changed, we can write up to 236 MSRs.

Use 4096 bytes for the buffer, that can hold 255 kvm_msr_entry
structs.

Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

Instead of using 2400 bytes in the stack for 150 MSR entries in
kvm_get_msrs() and kvm_put_msrs(), allocate a buffer once for
each VCPU.

Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

QOM instance_init functions are not supposed to have any side-effects,
as new objects may be created at any moment for querying property
information (see qmp_device_list_properties()).

Calling cpu_exec_init() also affects QEMU's ability to handle errors
during CPU creation, as some actions done by cpu_exec_init() can't be
reverted.

Move cpu_exec_init() call to realize so a simple object_new() won't
trigger it, and so that it is called after some basic validation of CPU
parameters.

Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

QOM instance_init functions are not supposed to have any side-effects,
as new objects may be created at any moment for querying property
information (see qmp_device_list_properties()).

Move TCG initialization to realize time so it won't be called when just
doing object_new() on a X86CPU subclass.

Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

Instead of requiring cpu.c to check if TCG was already initialized,
simply let the function be called multiple times.

Suggested-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

x86_cpudef_init() doesn't do anything anymore, cpudef_init(),
cpudef_setup(), and x86_cpudef_init() can be finally removed.

Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

Newer PC machines don't set hw_version, and older machines set
model-id on compat_props explicitly, so we don't need the
x86_cpudef_setup() code that sets model_id using
qemu_hw_version() anymore.

Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>

open_eth_start_xmit has a huge stack usage of 65536 bytes approx.
Moving large arrays to heap to reduce stack usage.

Reduce size of a buffer allocated on stack to 0x600 bytes, which is the
maximal frame length when HUGEN bit is not set in MODER, only allocate
buffer on heap when that is too small. Thus heap is not used in typical
use case.

Signed-off-by: Zhou Jie <zhoujie2011@cn.fujitsu.com>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>