commit fce55481360d "ppc: A couple more dummy POWER8 Book4 regs"
squashed in to rapidly a set of POWER8 Book4 regs in the wrong
routine. This patch introduces the missing gen_spr_power8_book4()
routine to fix their location.
Signed-off-by: NCédric Le Goater <clg@fr.ibm.com>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

RX buffer pools are now enabled by default for new machine types.
For older machine types, they are still disabled to avoid breaking
migration.
Signed-off-by: NThomas Huth <thuth@redhat.com>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>
Reviewed-by: NLaurent Vivier <lvivier@redhat.com>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

tl;dr:
This patch introduces an alternate way of handling the receive
buffers of the spapr-vlan device, resulting in much better
receive performance for the guest.

Full story:
One of our testers recently discovered that the performance of the
spapr-vlan device is very poor compared to other NICs, and that
a simple "ping -i 0.2 -s 65507 someip" in the guest can result
in more than 50% lost ping packets (especially with older guest
kernels < 3.17).

After doing some analysis, it was clear that there is a problem
with the way we handle the receive buffers in spapr_llan.c: The
ibmveth driver of the guest Linux kernel tries to add a lot of
buffers into several buffer pools (with 512, 2048 and 65536 byte
sizes by default, but it can be changed via the entries in the
/sys/devices/vio/1000/pool* directories of the guest). However,
the spapr-vlan device of QEMU only tries to squeeze all receive
buffer descriptors into one single page which has been supplied
by the guest during the H_REGISTER_LOGICAL_LAN call, without
taking care of different buffer sizes. This has two bad effects:
First, only a very limited number of buffer descriptors is accepted
at all. Second, we also hand 64k buffers to the guest even if
the 2k buffers would fit better - and this results in dropped packets
in the IP layer of the guest since too much skbuf memory is used.

Though it seems at a first glance like PAPR says that we should store
the receive buffer descriptors in the page that is supplied during
the H_REGISTER_LOGICAL_LAN call, chapter 16.4.1.2 in the LoPAPR spec
declares that "the contents of these descriptors are architecturally
opaque, none of these descriptors are manipulated by code above
the architected interfaces". That means we don't have to store
the RX buffer descriptors in this page, but can also manage the
receive buffers at the hypervisor level only. This is now what we
are doing here: Introducing proper RX buffer pools which are also
sorted by size of the buffers, so we can hand out a buffer with
the best fitting size when a packet has been received.

To avoid problems with migration from/to older version of QEMU,
the old behavior is also retained and enabled by default. The new
buffer management has to be enabled via a new "use-rx-buffer-pools"
property.

Now with the new buffer pool management enabled, the problem with
"ping -s 65507" is fixed for me, and the throughput of a simple
test with wget increases from creeping 3MB/s up to 20MB/s!
Signed-off-by: NThomas Huth <thuth@redhat.com>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

Refactor the code a little bit by extracting the code that reads
and writes the receive buffer list page into separate functions.
There should be no functional change in this patch, this is just
a preparation for the upcoming extensions that introduce receive
buffer pools.
Signed-off-by: NThomas Huth <thuth@redhat.com>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>
Reviewed-by: NLaurent Vivier <lvivier@redhat.com>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
[clg: squashed in patch 'ppc: Add dummy ACOP SPR' ]
Signed-off-by: NCédric Le Goater <clg@fr.ibm.com>
Reviewed-by: NThomas Huth <thuth@redhat.com>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

We should implement HW breakpoint/watchpoint, qemu supports them...
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
Reviewed-by: NThomas Huth <thuth@redhat.com>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

With appropriate AMR-like masks. Not actually used by the translation
logic at that point
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
[clg: changed spr_register_hv(SPR_IAMR) to spr_register_kvm_hv(SPR_IAMR)
      changed gen_spr_amr() prototype ]
Signed-off-by: NCédric Le Goater <clg@fr.ibm.com>
Reviewed-by: NThomas Huth <thuth@redhat.com>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

The masks weren't chosen nor applied properly. The architecture specifies
that writes to AMR are masked by UAMOR for PR=1, otherwise AMOR for HV=0.

The writes to UAMOR are masked by AMOR for HV=0
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
[clg: moved gen_spr_amr() prototype change to next patch ]
Signed-off-by: NCédric Le Goater <clg@fr.ibm.com>
Reviewed-by: NThomas Huth <thuth@redhat.com>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

Make sure we give the guest full authorization
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
Reviewed-by: NThomas Huth <thuth@redhat.com>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

It's supposed to be an instruction counter. For now make us not
crash when accessing it.
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
Reviewed-by: NThomas Huth <thuth@redhat.com>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

And move the code adjusting the MSR mask and calling kvmppc_set_papr()
to it. This allows us to add a few more things such as disabling setting
of MSR:HV and appropriate LPCR bits which will be used when fixing
the exception model.
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>
[clg: removed LPCR setting ]
Signed-off-by: NCédric Le Goater <clg@fr.ibm.com>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

We don't give them a KVM reg number to most of the registers yet as no
current KVM version supports HV mode. For DAWR and DAWRX, the KVM reg
number is needed since this register can be set by the guest via the
H_SET_MODE hypercall.
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
[clg: squashed in patch 'ppc: Add KVM numbers to some P8 SPRs'
      changed the commit log with a proposal of Thomas Huth
      removed all hunks except those related to AMOR and DAWR* ]
Signed-off-by: NCédric Le Goater <clg@fr.ibm.com>
Reviewed-by: NThomas Huth <thuth@redhat.com>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

The current set of spr_register_* macros only take the user and
supervisor function pointers. To make the transition easy, we
don't change that but we add "_hv" variants that can be used to
register all 3 sets.

To simplify the transition, users of the "old" macro will set the
hypervisor callback to be the same as the supervisor one. The new
registration function only needs to be used for registers that are
either hypervisor only or behave differently in HV mode.
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>
[clg: fixed else if condition in gen_op_mfspr() ]
Signed-off-by: NCédric Le Goater <clg@fr.ibm.com>
Reviewed-by: NThomas Huth <thuth@redhat.com>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

Add definitions for additional SPR numbers and SPR bit definitions
that will be relevant for subsequent improvements to POWER8 emulation

Also fix the definition of LPIDR which was incorrect (and is different
for server and embedded).
Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
Reviewed-by: NThomas Huth <thuth@redhat.com>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

ePAPR defines "hcall-instructions" device-tree property which contains
code to call hypercalls in ePAPR paravirtualized guests.  In general
pseries guests won't use this property, instead using the PAPR defined
hypercall interface.

However, this property has been re-used to implement a hack to allow
PR KVM to run (slightly modified) guests in some situations where it
otherwise wouldn't be able to (because the system's L0 hypervisor
doesn't forward the PAPR hypercalls to the PR KVM kernel).

Hence, this property is always present in the device tree for pseries
guests. All KVM guests use it at least to read features via the
KVM_HC_FEATURES hypercall.

The property is populated by the code returned from the KVM's
KVM_PPC_GET_PVINFO ioctl; if not implemented in the KVM, QEMU supplies
code which will fail all hypercall attempts. If QEMU does not create
the property, and the guest kernel is compiled with
CONFIG_EPAPR_PARAVIRT (which is normally the case), there is exactly
the same stub at @epapr_hypercall_start already.

Rather than maintaining this fairly useless stub implementation, it
makes more sense not to create the property in the device tree in the
first place if the host kernel does not implement it.

This changes kvmppc_get_hypercall() to return 1 if the host kernel
does not implement KVM_CAP_PPC_GET_PVINFO. The caller can use it to decide
on whether to create the property or not.

This changes the pseries machine to not create the property if KVM does
not implement KVM_PPC_GET_PVINFO. In practice this means that from now
on the property will not be created if either HV KVM or TCG is used.
Signed-off-by: NAlexey Kardashevskiy <aik@ozlabs.ru>
[reworded commit message for clarity --dwg]
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

When a qemu-system-ppc64 is started, the 64-bit mode bit
is not set in MSR.
Signed-off-by: NLaurent Vivier <lvivier@redhat.com>
Reviewed-by: NAlexander Graf <agraf@suse.de>
Signed-off-by: NDavid Gibson <david@gibson.dropbear.id.au>

ivshmem: Fixes, cleanups, device model split

# gpg: Signature made Mon 21 Mar 2016 20:33:54 GMT using RSA key ID EB918653
# gpg: Good signature from "Markus Armbruster <armbru@redhat.com>"
# gpg:                 aka "Markus Armbruster <armbru@pond.sub.org>"

* remotes/armbru/tags/pull-ivshmem-2016-03-18: (40 commits)
  contrib/ivshmem-server: Print "not for production" warning
  ivshmem: Require master to have ID zero
  ivshmem: Drop ivshmem property x-memdev
  ivshmem: Clean up after the previous commit
  ivshmem: Split ivshmem-plain, ivshmem-doorbell off ivshmem
  ivshmem: Replace int role_val by OnOffAuto master
  qdev: New DEFINE_PROP_ON_OFF_AUTO
  ivshmem: Inline check_shm_size() into its only caller
  ivshmem: Simplify memory regions for BAR 2 (shared memory)
  ivshmem: Implement shm=... with a memory backend
  ivshmem: Tighten check of property "size"
  ivshmem: Simplify how we cope with short reads from server
  ivshmem: Drop the hackish test for UNIX domain chardev
  ivshmem: Rely on server sending the ID right after the version
  ivshmem: Propagate errors through ivshmem_recv_setup()
  ivshmem: Receive shared memory synchronously in realize()
  ivshmem: Plug leaks on unplug, fix peer disconnect
  ivshmem: Disentangle ivshmem_read()
  ivshmem: Simplify rejection of invalid peer ID from server
  ivshmem: Assert interrupts are set up once
  ...
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

wxx patch queue

# gpg: Signature made Tue 22 Mar 2016 18:18:36 GMT using RSA key ID 677450AD
# gpg: Good signature from "Stefan Weil <sw@weilnetz.de>"
# gpg:                 aka "Stefan Weil <stefan.weil@weilnetz.de>"
# gpg:                 aka "Stefan Weil <stefan.weil@bib.uni-mannheim.de>"
# gpg: WARNING: This key is not certified with sufficiently trusted signatures!
# gpg:          It is not certain that the signature belongs to the owner.
# Primary key fingerprint: 4923 6FEA 75C9 5D69 8EC2  B78A E08C 21D5 6774 50AD

* remotes/weil/tags/pull-wxx-20160322:
  wxx: Add support for ncurses
  Remove unneeded include statements for setjmp.h
  Include setjmp.h in qemu/osdep.h (bug fix for w64)
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

We used to support only pdcurses for Windows, but recently Cygwin added
mingw64-i686-ncurses and mingw64-x86_64-ncurses packages which are
supported now, too.
Signed-off-by: NStefan Weil <sw@weilnetz.de>

As soon as setjmp.h is included from qemu/osdep.h, those old include
statements are no longer needed.

Add also setjmp.h to the list in scripts/clean-includes.
Signed-off-by: NStefan Weil <sw@weilnetz.de>

setjmp must be declared before sysemu/os-win32.h
because it is redefined there for 64 bit Windows.
Reviewed-by: NRichard Henderson  <rth@twiddle.net>
Tested-by: NAndrew Baumann <Andrew.Baumann@microsoft.com>
Signed-off-by: NStefan Weil <sw@weilnetz.de>

qemu-ga patch queue for 2.6

* remove unused variable

# gpg: Signature made Mon 21 Mar 2016 17:32:42 GMT using RSA key ID F108B584
# gpg: Good signature from "Michael Roth <flukshun@gmail.com>"
# gpg:                 aka "Michael Roth <mdroth@utexas.edu>"
# gpg:                 aka "Michael Roth <mdroth@linux.vnet.ibm.com>"

* remotes/mdroth/tags/qga-pull-2016-03-21-tag:
  qemu-ga: drop unused local err variable
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

usb: bugfix collection.

# gpg: Signature made Mon 21 Mar 2016 11:07:39 GMT using RSA key ID D3E87138
# gpg: Good signature from "Gerd Hoffmann (work) <kraxel@redhat.com>"
# gpg:                 aka "Gerd Hoffmann <gerd@kraxel.org>"
# gpg:                 aka "Gerd Hoffmann (private) <kraxel@gmail.com>"

* remotes/kraxel/tags/pull-usb-20160321-1:
  usb: ehci: add capability mmio write function
  hw/usb/dev-mtp: Guard inotify usage with CONFIG_INOTIFY1
  usb: fix unbound stack warning for inotify_watchfn
  usb: fix unbound stack usage for usb_mtp_add_str
  usb: fix unbounded stack warning for xhci_dma_write_u32s
  usb: Fix compilation for Windows
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

The code is okay for illustrating how things work and for testing, but
its error handling make it unfit for production use.  Print a warning
to protect the innocent.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <1458066895-20632-41-git-send-email-armbru@redhat.com>

Migration with ivshmem needs to be carefully orchestrated to work.
Exactly one peer (the "master") migrates to the destination, all other
peers need to unplug (and disconnect), migrate, plug back (and
reconnect).  This is sort of documented in qemu-doc.

If peers connect on the destination before migration completes, the
shared memory can get messed up.  This isn't documented anywhere.  Fix
that in qemu-doc.

To avoid messing up register IVPosition on migration, the server must
assign the same ID on source and destination.  ivshmem-spec.txt leaves
ID assignment unspecified, however.

Amend ivshmem-spec.txt to require the first client to receive ID zero.
The example ivshmem-server complies: it always assigns the first
unused ID.

For a bit of additional safety, enforce ID zero for the master.  This
does nothing when we're not using a server, because the ID is zero for
all peers then.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <1458066895-20632-40-git-send-email-armbru@redhat.com>

Use ivshmem-plain instead.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <1458066895-20632-39-git-send-email-armbru@redhat.com>

Move code to more sensible places.  Use the opportunity to reorder and
document IVShmemState members.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <1458066895-20632-38-git-send-email-armbru@redhat.com>

ivshmem can be configured with and without interrupt capability
(a.k.a. "doorbell").  The two configurations have largely disjoint
options, which makes for a confusing (and badly checked) user
interface.  Moreover, the device can't tell the guest whether its
doorbell is enabled.

Create two new device models ivshmem-plain and ivshmem-doorbell, and
deprecate the old one.

Changes from ivshmem:

* PCI revision is 1 instead of 0.  The new revision is fully backwards
  compatible for guests.  Guests may elect to require at least
  revision 1 to make sure they're not exposed to the funny "no shared
  memory, yet" state.

* Property "role" replaced by "master".  role=master becomes
  master=on, role=peer becomes master=off.  Default is off instead of
  auto.

* Property "use64" is gone.  The new devices always have 64 bit BARs.

Changes from ivshmem to ivshmem-plain:

* The Interrupt Pin register in PCI config space is zero (does not use
  an interrupt pin) instead of one (uses INTA).

* Property "x-memdev" is renamed to "memdev".

* Properties "shm" and "size" are gone.  Use property "memdev"
  instead.

* Property "msi" is gone.  The new device can't have MSI-X capability.
  It can't interrupt anyway.

* Properties "ioeventfd" and "vectors" are gone.  They're meaningless
  without interrupts anyway.

Changes from ivshmem to ivshmem-doorbell:

* Property "msi" is gone.  The new device always has MSI-X capability.

* Property "ioeventfd" defaults to on instead of off.

* Property "size" is gone.  The new device can only map all the shared
  memory received from the server.

Guests can easily find out whether the device is configured for
interrupts by checking for MSI-X capability.

Note: some code added in sub-optimal places to make the diff easier to
review.  The next commit will move it to more sensible places.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <1458066895-20632-37-git-send-email-armbru@redhat.com>

In preparation of making it a qdev property.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <1458066895-20632-36-git-send-email-armbru@redhat.com>

Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <1458066895-20632-35-git-send-email-armbru@redhat.com>

Improve the error messages while there.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Message-Id: <1458066895-20632-34-git-send-email-armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>

ivshmem_realize() puts the shared memory region in a container region.
Used to be necessary to permit delayed mapping of the shared memory.
However, we recently moved to synchronous mapping, in "ivshmem:
Receive shared memory synchronously in realize()" and the commit
following it.  The container is redundant since then.  Drop it.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Message-Id: <1458066895-20632-33-git-send-email-armbru@redhat.com>

ivshmem has its very own code to create and map shared memory.
Replace that with an implicitly created memory backend.  Reduces the
number of ways we create BAR 2 from three to two.

The memory-backend-file is currently available only with CONFIG_LINUX,
so this adds a second Linuxism to ivshmem (the other one is eventfd).
Should we ever need to make it portable to systems where
memory-backend-file can't be made to serve, we could create a
memory-backend-shmem that allocates memory with shm_open().

Bonus fix: shared memory files are now created with permissions 0655
instead of 0777.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Message-Id: <1458066895-20632-32-git-send-email-armbru@redhat.com>

If size_t is narrower than 64 bits, passing uint64_t ivshmem_size to
mmap() truncates.  Reject such sizes.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <1458066895-20632-31-git-send-email-armbru@redhat.com>

Short reads from a UNIX domain sockets are exceedingly unlikely when
the other side always sends eight bytes and we always read eight
bytes.  We cope with them anyway.  However, the code doing that is
rather convoluted.  Dumb it down radically.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <1458066895-20632-30-git-send-email-armbru@redhat.com>

The chardev must be capable of transmitting SCM_RIGHTS ancillary
messages.  We check it by comparing CharDriverState member filename to
"unix:".  That's almost as brittle as it is disgusting.

When the actual transmission all happened asynchronously, this check
was all we could do in realize(), and thus better than nothing.  But
now we receive at least one SCM_RIGHTS synchronously in realize(),
it's not worth its keep anymore.  Drop it.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <1458066895-20632-29-git-send-email-armbru@redhat.com>

The protocol specification (ivshmem-spec.txt, formerly
ivshmem_device_spec.txt) has always required the ID message to be sent
right at the beginning, and ivshmem-server has always complied.  The
device, however, accepts it out of order.  If an interrupt setup
arrived before it, though, it would be misinterpreted as connect
notification.  Fix the latent bug by relying on the spec and
ivshmem-server's actual behavior.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <1458066895-20632-28-git-send-email-armbru@redhat.com>

This kills off the funny state described in the previous commit.

Simplify ivshmem_io_read() accordingly, and update documentation.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Message-Id: <1458066895-20632-27-git-send-email-armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>

When configured for interrupts (property "chardev" given), we receive
the shared memory from an ivshmem server.  We do so asynchronously
after realize() completes, by setting up callbacks with
qemu_chr_add_handlers().

Keeping server I/O out of realize() that way avoids delays due to a
slow server.  This is probably relevant only for hot plug.

However, this funny "no shared memory, yet" state of the device also
causes a raft of issues that are hard or impossible to work around:

* The guest is exposed to this state: when we enter and leave it its
  shared memory contents is apruptly replaced, and device register
  IVPosition changes.

  This is a known issue.  We document that guests should not access
  the shared memory after device initialization until the IVPosition
  register becomes non-negative.

  For cold plug, the funny state is unlikely to be visible in
  practice, because we normally receive the shared memory long before
  the guest gets around to mess with the device.

  For hot plug, the timing is tighter, but the relative slowness of
  PCI device configuration has a good chance to hide the funny state.

  In either case, guests complying with the documented procedure are
  safe.

* Migration becomes racy.

  If migration completes before the shared memory setup completes on
  the source, shared memory contents is silently lost.  Fortunately,
  migration is rather unlikely to win this race.

  If the shared memory's ramblock arrives at the destination before
  shared memory setup completes, migration fails.

  There is no known way for a management application to wait for
  shared memory setup to complete.

  All you can do is retry failed migration.  You can improve your
  chances by leaving more time between running the destination QEMU
  and the migrate command.

  To mitigate silent memory loss, you need to ensure the server
  initializes shared memory exactly the same on source and
  destination.

  These issues are entirely undocumented so far.

I'd expect the server to be almost always fast enough to hide these
issues.  But then rare catastrophic races are in a way the worst kind.

This is way more trouble than I'm willing to take from any device.
Kill the funny state by receiving shared memory synchronously in
realize().  If your hot plug hangs, go kill your ivshmem server.

For easier review, this commit only makes the receive synchronous, it
doesn't add the necessary error propagation.  Without that, the funny
state persists.  The next commit will do that, and kill it off for
real.
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <1458066895-20632-26-git-send-email-armbru@redhat.com>

close_peer_eventfds() cleans up three things: ioeventfd triggers if
they exist, eventfds, and the array to store them.

Commit 98609cd8 (v1.2.0) fixed it not to clean up ioeventfd triggers
when they don't exist (property ioeventfd=off, which is the default).
Unfortunately, the fix also made it skip cleanup of the eventfds and
the array then.  This is a memory and file descriptor leak on unplug.

Additionally, the reset of nb_eventfds is skipped.  Doesn't matter on
unplug.  On peer disconnect, however, this permanently wedges the
interrupt vectors used for that peer's ID.  The eventfds stay behind,
but aren't connected to a peer anymore.  When the ID gets recycled for
a new peer, the new peer's eventfds get assigned to vectors after the
old ones.  Commonly, the device's number of vectors matches the
server's, so the new ones get dropped with a "Too many eventfd
received" message.  Interrupts either don't work (common case) or go
to the wrong vector.

Fix by narrowing the conditional to just the ioeventfd trigger
cleanup.

While there, move the "invalid" peer check to the only caller where it
can actually happen, and tighten it to reject own ID.

Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <1458066895-20632-25-git-send-email-armbru@redhat.com>