migration: Wait for semaphore before completing migration

Wait for a semaphore before completing the migration,
if the previously added capability was enabled.

Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration: Add 'pre-switchover' and 'device' statuses

Add two statuses for use when the 'pause-before-switchover'
capability is enabled.

'pre-switchover' is the state that we wait in for management
to allow us to continue.
'device' is the state we enter while serialising the devices
after management gives us the OK.

Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration: Add 'pause-before-switchover' capability

When 'pause-before-switchover' is enabled, the outgoing migration
will pause before invalidating the block devices and serializing
the device state.
At this point the management layer gets the chance to clean up any
device jobs or other device users before the migration completes.

Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration: Make cache_init() take an error parameter

Once there, take a total size instead of the size of the pages.  We
move the check that the new_size is bigger than one page from
xbzrle_cache_resize().

Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
--

Fix typo spotted by Peter Xu

migration: Move xbzrle cache resize error handling to xbzrle_cache_resize

Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>

migration: Make cache size elements use the right types

Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>

migratiom: Remove max_item_age parameter

It was not used at all since commit:

27af7d6ea5015e5ef1f7985eab94a8a218267a2b

which replaced its use by the dirty sync count.

Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>

migration: Fix migrate_test_apply for multifd parameters

They were missing when introduced on the tree

Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>

Merge remote-tracking branch 'remotes/stefanha/tags/block-pull-request' into staging

# gpg: Signature made Fri 20 Oct 2017 13:02:25 BST
# gpg:                using RSA key 0x9CA4ABB381AB73C8
# gpg: Good signature from "Stefan Hajnoczi <stefanha@redhat.com>"
# gpg:                 aka "Stefan Hajnoczi <stefanha@gmail.com>"
# Primary key fingerprint: 8695 A8BF D3F9 7CDA AC35  775A 9CA4 ABB3 81AB 73C8

* remotes/stefanha/tags/block-pull-request:
  oslib-posix: Fix compiler warning and some data types

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

Merge remote-tracking branch 'remotes/cohuck/tags/s390x-20171020' into staging

The last big chunk of s390x changes:
- (experimental) smp support under tcg
- provide the virtio-input devices for virtio-ccw
- improve error handling in the css code
- enable some simple virtio tests for s390x
- low-address protection in tcg
- some more cleanups and fixes

# gpg: Signature made Fri 20 Oct 2017 12:49:22 BST
# gpg:                using RSA key 0xDECF6B93C6F02FAF
# gpg: Good signature from "Cornelia Huck <conny@cornelia-huck.de>"
# gpg:                 aka "Cornelia Huck <huckc@linux.vnet.ibm.com>"
# gpg:                 aka "Cornelia Huck <cornelia.huck@de.ibm.com>"
# gpg:                 aka "Cornelia Huck <cohuck@kernel.org>"
# gpg:                 aka "Cornelia Huck <cohuck@redhat.com>"
# Primary key fingerprint: C3D0 D66D C362 4FF6 A8C0  18CE DECF 6B93 C6F0 2FAF

* remotes/cohuck/tags/s390x-20171020: (46 commits)
  s390x/tcg: low-address protection support
  accel/tcg: allow to invalidate a write TLB entry immediately
  tests: Enable the very simple virtio tests on s390x, too
  libqtest: Add qtest_[v]startf()
  s390x: refactor error handling for MSCH handler
  s390x: refactor error handling for HSCH handler
  s390x: refactor error handling for CSCH handler
  s390x: refactor error handling for XSCH handler
  s390x: improve error handling for SSCH and RSCH
  s390x/css: IO instr handler ending control
  s390x: move s390x_new_cpu() into board code
  s390x: fix cpu object referrence leak in s390x_new_cpu()
  s390x/event-facility: variable-length event masks
  s390x/MAINTAINERS: add mailing list
  virtio-ccw: Add the virtio-input devices for CCW bus
  target/s390x: special handling when starting a CPU with WAIT PSW
  s390x/tcg: refactor stfl(e) to use s390_get_feat_block()
  s390x/tcg: unlock NMI
  s390x/cpumodel: allow to enable SENSE RUNNING STATUS for qemu
  s390x/tcg: switch to new SIGP handling code
  ...

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

Merge remote-tracking branch 'remotes/famz/tags/docker-pull-request' into staging

# gpg: Signature made Fri 20 Oct 2017 07:30:45 BST
# gpg:                using RSA key 0xCA35624C6A9171C6
# gpg: Good signature from "Fam Zheng <famz@redhat.com>"
# 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: 5003 7CB7 9706 0F76 F021  AD56 CA35 624C 6A91 71C6

* remotes/famz/tags/docker-pull-request:
  docker: Fix PATH for ccache
  docker: fix out-of-tree 'make docker-test-build@debian-powerpc-cross'
  docker: allow running from srcdir != builddir build
  docker: cleanup temp directory after test
  docker: Don't allocate tty unless DEBUG=1

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

s390x/tcg: low-address protection support

This is a neat way to implement low address protection, whereby
only the first 512 bytes of the first two pages (each 4096 bytes) of
every address space are protected.

Store a tec of 0 for the access exception, this is what is defined by
Enhanced Suppression on Protection in case of a low address protection
(Bit 61 set to 0, rest undefined).

We have to make sure to to pass the access address, not the masked page
address into mmu_translate*().

Drop the check from testblock. So we can properly test this via
kvm-unit-tests.

This will check every access going through one of the MMUs.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20171016202358.3633-3-david@redhat.com>
[CH: restored error message for access register mode]
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

accel/tcg: allow to invalidate a write TLB entry immediately

Background: s390x implements Low-Address Protection (LAP). If LAP is
enabled, writing to effective addresses (before any translation)
0-511 and 4096-4607 triggers a protection exception.

So we have subpage protection on the first two pages of every address
space (where the lowcore - the CPU private data resides).

By immediately invalidating the write entry but allowing the caller to
continue, we force every write access onto these first two pages into
the slow path. we will get a tlb fault with the specific accessed
addresses and can then evaluate if protection applies or not.

We have to make sure to ignore the invalid bit if tlb_fill() succeeds.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20171016202358.3633-2-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

tests: Enable the very simple virtio tests on s390x, too

These tests can easily be used on s390x, too. We just have to make
sure to use the virtio-xxx-ccw devices instead of virtio-xxx-pci.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <1508336428-20511-3-git-send-email-thuth@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

libqtest: Add qtest_[v]startf()

We have several callers that were formatting the argument strings
themselves; consolidate this effort by adding new convenience
functions directly in libqtest, and update some call-sites that
can benefit from it.

Note that the new functions qtest_startf() and qtest_vstartf()
behave more like qtest_init() (the caller must assign global_qtest
after the fact, rather than getting it implicitly set).  This helps
us prepare for future patches that get rid of the global variable,
by explicitly highlighting which tests still depend on it now.

Signed-off-by: Eric Blake <eblake@redhat.com>
[thuth: Dropped the hunks that do not apply cleanly to qemu master
 yet and added the missing g_free(args) in qtest_vstartf()]
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <1508336428-20511-2-git-send-email-thuth@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x: refactor error handling for MSCH handler

Simplify the error handling of the MSCH.  Let the code detecting the
condition tell (in a less ambiguous way) how it's to be handled. No
changes in behavior.

Signed-off-by: Halil Pasic <pasic@linux.vnet.ibm.com>
Message-Id: <20171017140453.51099-8-pasic@linux.vnet.ibm.com>
[CH: fix return code for fctl != 0]
Reviewed-by: Dong Jia Shi <bjsdjshi@linux.vnet.ibm.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x: refactor error handling for HSCH handler

Simplify the error handling of the HSCH.  Let the code detecting the
condition tell (in a less ambiguous way) how it's to be handled. No
changes in behavior.

Signed-off-by: Halil Pasic <pasic@linux.vnet.ibm.com>
Message-Id: <20171017140453.51099-7-pasic@linux.vnet.ibm.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Dong Jia Shi <bjsdjshi@linux.vnet.ibm.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x: refactor error handling for CSCH handler

Simplify the error handling of the CSCH.  Let the code detecting the
condition tell (in a less ambiguous way) how it's to be handled. No
changes in behavior.

Signed-off-by: Halil Pasic <pasic@linux.vnet.ibm.com>
Message-Id: <20171017140453.51099-6-pasic@linux.vnet.ibm.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Dong Jia Shi <bjsdjshi@linux.vnet.ibm.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x: refactor error handling for XSCH handler

Simplify the error handling of the XSCH.  Let the code detecting the
condition tell (in a less ambiguous way) how it's to be handled. No
changes in behavior.

Signed-off-by: Halil Pasic <pasic@linux.vnet.ibm.com>
Message-Id: <20171017140453.51099-5-pasic@linux.vnet.ibm.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Dong Jia Shi <bjsdjshi@linux.vnet.ibm.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x: improve error handling for SSCH and RSCH

Simplify the error handling of the SSCH and RSCH handler avoiding
arbitrary and cryptic error codes being used to tell how the instruction
is supposed to end.  Let the code detecting the condition tell how it's
to be handled in a less ambiguous way.  It's best to handle SSCH and RSCH
in one go as the emulation of the two shares a lot of code.

For passthrough this change isn't pure refactoring, but changes the way
kernel reported EFAULT is handled. After clarifying the kernel interface
we decided that EFAULT shall be mapped to unit exception.  Same goes for
unexpected error codes and absence of required ORB flags.

Signed-off-by: Halil Pasic <pasic@linux.vnet.ibm.com>
Message-Id: <20171017140453.51099-4-pasic@linux.vnet.ibm.com>
Tested-by: Dong Jia Shi <bjsdjshi@linux.vnet.ibm.com>
[CH: cosmetic changes]
Reviewed-by: Dong Jia Shi <bjsdjshi@linux.vnet.ibm.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/css: IO instr handler ending control

CSS code needs to tell the IO instruction handlers located in ioinst.c
how the emulated instruction should be ended. Currently this is done by
returning generic (POSIX) error codes, and mapping them to outcomes like
condition codes. This makes bugs easy to create and hard to recognize.

As a preparation for moving away from (mis)using generic error codes for
flow control let us introduce a type which tells the instruction
handler function how to end the instruction, in a more straight-forward
and less ambiguous way.

Signed-off-by: Halil Pasic <pasic@linux.vnet.ibm.com>
Message-Id: <20171017140453.51099-3-pasic@linux.vnet.ibm.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Dong Jia Shi <bjsdjshi@linux.vnet.ibm.com>
[CH: cosmetic changes]
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x: move s390x_new_cpu() into board code

s390-virtio-ccw.c is the sole user of s390x_new_cpu(),
so move this helper there.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <1508253203-119237-1-git-send-email-imammedo@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x: fix cpu object referrence leak in s390x_new_cpu()

object_new() returns cpu with refcnt == 1 and after realize
refcnt == 2*. s390x_new_cpu() as an owner of the first refcnt
should have released it on exit in both cases (on error and
success) to avoid it leaking. Do so for both cases.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <1508247680-98800-2-git-send-email-imammedo@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/event-facility: variable-length event masks

The architecture supports masks of variable length for sclp write
event mask. We currently only support 4 byte event masks, as that
is what Linux uses.

Let's extend this to the maximum mask length supported by the
architecture and return 0 to the guest for the mask bits we don't
support in core.

Initial patch by: Cornelia Huck <cornelia.huck@de.ibm.com>

Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Jason J. Herne <jjherne@linux.vnet.ibm.com>
Message-Id: <1507729193-9747-1-git-send-email-jjherne@linux.vnet.ibm.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/MAINTAINERS: add mailing list

There's a new qemu-s390x@nongnu.org mailing list: add it to the
s390 sections.

Reviewed-by: Thomas Huth <thuth@redhat.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

virtio-ccw: Add the virtio-input devices for CCW bus

Wire up the virtio-input HID devices (keyboard, mouse, tablet)
for the CCW bus. The virtio-input is a virtio-1 device,
so disable legacy revision 0.

Signed-off-by: Farhan Ali <alifm@linux.vnet.ibm.com>
Reviewed-by: Halil Pasic <pasic@linux.vnet.ibm.com>
Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com>
Message-Id: <6a8ea4c503ee32c2ca7fa608b5f2f547009be8ee.1507557166.git.alifm@linux.vnet.ibm.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

target/s390x: special handling when starting a CPU with WAIT PSW

When we try to start a CPU with a WAIT PSW, we have to take care that
TCG will actually try to continue executing instructions.

We must therefore really only unhalt the CPU if we don't have a WAIT
PSW. Also document the special order for restart interrupts, which
load a new PSW and change the state to operating.

To keep KVM working, simply don't have a look at the WAIT bit when
loading the PSW. Otherwise the behavior of a restart interrupt when
a CPU stopped would be changed.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-31-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: refactor stfl(e) to use s390_get_feat_block()

Refactor it to use s390_get_feat_block(). Directly write into the mapped
lowcore with stfl and make sure it is really only compiled if needed.

While at it, add an alignment check for STFLE and avoid
potential_page_fault() by properly restoring the CPU state.

Due to s390_get_feat_block(), we will now also indicate the
"Configuration-z-architectural-mode", which is with new SIGP code the
right thing to do.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-30-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: unlock NMI

Nothing hindering us anymore from unlocking the restart code (used for
NMI).

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-29-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/cpumodel: allow to enable SENSE RUNNING STATUS for qemu

As we properly implement it, allow to enable it.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-28-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: switch to new SIGP handling code

This effectively enables experimental SMP support. Floating interrupts are
still a mess, so allow it but print a big warning. There also seems
to be a problem with CPU hotplug (after the main loop started).

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-27-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
[CH: changed insn-data.def as pointed out by Richard]
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: flush the tlb on SIGP SET PREFIX

Thanks to Aurelien Jarno for doing this in his prototype.

We can flush the whole TLB as this should happen really rarely.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-26-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: implement STOP and RESET interrupts for TCG

Implement them like KVM implements/handles them. Both can only be
triggered via SIGP instructions. RESET has (almost) the lowest priority if
the CPU is running, and the highest if the CPU is STOPPED. This is handled
in SIGP code already. On delivery, we only have to care about the
"CPU running" scenario.

STOP is defined to be delivered after all other interrupts have been
delivered. Therefore it has the actual lowest priority.

As both can wake up a CPU if sleeping, indicate them correctly to
external code (e.g. cpu_has_work()).

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-25-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: implement SIGP CONDITIONAL EMERGENCY SIGNAL

Mostly analogous to the kernel/KVM version (so I assume the checks are
correct :) ). As a preparation for TCG.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-24-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: implement SIGP EMERGENCY SIGNAL

As preparation for TCG.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-23-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: implement SIGP EXTERNAL CALL

As preparation for TCG.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-22-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: implement SIGP SENSE

Add it as preparation for TCG. Sensing could later be done completely
lockless.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-21-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: implement SIGP SENSE RUNNING STATUS

Preparation for TCG, for KVM is this is completely handled in the
kernel.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-20-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/kvm: factor out actual handling of STOP interrupts

For KVM, the KVM module decides when a STOP can be performed (when the
STOP interrupt can be processed). Factor it out so we can use it
later for TCG.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-19-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/kvm: factor out SIGP code into sigp.c

We want to use the same code base for TCG, so let's cleanly factor it
out.

The sigp mutex is currently not really needed, as everything is
protected by the iothread mutex. But this could change later, so leave
it in place and initialize it properly from common code.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-17-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/kvm: drop two debug prints

Preparation for moving it out of kvm.c.

Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-16-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/kvm: factor out storing of adtl CPU status

Called from SIGP code to be factored out, so let's move it. Add a
FIXME for TCG code in the future.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-15-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/kvm: factor out storing of CPU status

Factor it out into s390_store_status(), to be used also by TCG later on.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-14-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/kvm: generalize SIGP stop and restart interrupt injection

Preparation for factoring it out into !kvm code.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-13-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/kvm: pass ipb directly into handle_sigp()

No need to pass kvm_run. Pass parameters alphabetically ordered.

Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-12-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

target/s390x: interpret PSW_MASK_WAIT only for TCG

KVM handles the wait PSW itself and triggers a WAIT ICPT in case it
really wants to sleep (disabled wait).

This will later allow us to change the order of loading a restart
interrupt and setting a CPU to OPERATING on SIGP RESTART without
changing KVM behavior.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-11-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: handle WAIT PSWs during interrupt injection

If we encounter a WAIT PSW, we have to halt immediately. Using
cpu_loop_exit() at this point feels wrong. Simply leaving
cs->exception_index set doesn't result in an immediate stop.

This is also necessary to properly handle SIGP STOP interrupts later.

The CPU_INTERRUPT_HALT will be processed immediately and properly set
the CPU to halted (also resetting cs->exception_index to EXCP_HLT)

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-10-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

target/s390x: factor out handling of WAIT PSW into s390_handle_wait()

This will now also detect crashes under TCG. We can directly use
cpu->env.psw.addr instead of kvm_run, as we do a cpu_synchronize_state().

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-9-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: a CPU cannot switch state due to an interrupt

Going to OPERATING here looks wrong. A CPU should even never be
!OPERATING at this point. Unhalting will already be done in
cpu_handle_halt() if there is work, so we can drop this statement
completely.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-8-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: STOPPED cpus can never wake up

Interrupts can't wake such CPUs up. SIGP from other CPUs has to be used
to toggle the state.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-7-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: take care of external interrupt subclasses

We can now let go of INTERRUPT_EXT. When cr0 changes, we have to
revalidate if we now have a pending external interrupt, just like
when the PSW (or SYSTEM MASK only) changes.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-6-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: rework checking for deliverable interrupts

Currently, enabling/disabling of interrupts is not really supported.

Let's improve interrupt handling code by explicitly checking for
deliverable interrupts only. This is the first step. Checking for
external interrupt subclasses will be done next.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-5-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: injection of emergency signals and external calls

Preparation for new TCG SIGP code. Especially also prepare for
indicating that another external call is already pending.

Take care of interrupt priority.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-4-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: cleanup service interrupt injection

There are still some leftovers from old virtio interrupts in there.
Most importantly, we don't have to queue service interrupts anymore.
Just like KVM, we can simply multiplex the SCLP service interrupts and
avoid the queue.

Also, now only valid parameters/cpu_addr will be stored on service
interrupts.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-3-david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/tcg: turn INTERRUPT_EXT into a mask

External interrupts are currently all handled like floating external
interrupts, they are queued. Let's prepare for a split of floating
and local interrupts by turning INTERRUPT_EXT into a mask.

While we can have various floating external interrupts of one kind, there
is usually only one (or a fixed number) of the local external interrupts.

So turn INTERRUPT_EXT into a mask and properly indicate the kind of
external interrupt. Floating interrupts will have to moved out of
one CPU instance later once we have SMP support.

The only floating external interrupts used right now are SERVICE
interrupts, so let's use that name. Following patches will clean up
SERVICE interrupt injection.

This get's rid of the ugly special handling for cpu timer and clock
comparator interrupts. And we really only store the parameters as
defined by the PoP.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20170928203708.9376-2-david@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

s390x/css: be more consistent if broken beyond repair

Calling do_subchannel_work with no function control flags set in SCSW is
a programming error. Currently we handle this differently in
do_subchannel_work_virtual and do_subchannel_work_passthrough. Let's be
consistent and guard with a common assert against this programming error.

Signed-off-by: Halil Pasic <pasic@linux.vnet.ibm.com>
Message-Id: <20171004154144.88995-2-pasic@linux.vnet.ibm.com>
Reviewed-by: Dong Jia Shi <bjsdjshi@linux.vnet.ibm.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

S390: use g_new() family of functions

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
[PMD: more changes in hw/s390x/css.c, added target/s390x/cpu_models.c]
Message-Id: <20171006235023.11952-27-f4bug@amsat.org>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>

Merge remote-tracking branch 'remotes/stefanberger/tags/pull-tpm-2017-10-19-1' into staging

Merge tpm 2017/10/19 v1

# gpg: Signature made Thu 19 Oct 2017 16:42:39 BST
# gpg:                using RSA key 0x75AD65802A0B4211
# gpg: Good signature from "Stefan Berger <stefanb@linux.vnet.ibm.com>"
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: B818 B9CA DF90 89C2 D5CE  C66B 75AD 6580 2A0B 4211

* remotes/stefanberger/tags/pull-tpm-2017-10-19-1: (21 commits)
  tpm: move recv_data_callback to TPM interface
  tpm: add a QOM TPM interface
  tpm-tis: fold TPMTISEmuState in TPMState
  tpm-tis: remove tpm_tis.h header
  tpm-tis: move TPMState to TIS header
  tpm: remove locty_data from TPMState
  tpm-emulator: fix error handling
  tpm: add TPMBackendCmd to hold the request state
  tpm: remove locty argument from receive_cb
  tpm: remove needless cast
  tpm: remove unused TPMBackendCmd
  tpm: remove configure_tpm() hop
  tpm: remove init() class method
  tpm: remove TPMDriverOps
  tpm: move TPMSizedBuffer to tpm_tis.h
  tpm: remove tpm_register_driver()
  tpm: replace tpm_get_backend_driver() to drop be_drivers
  tpm: lookup tpm backend class in tpm_driver_find_by_type()
  tpm: make tpm_get_backend_driver() static
  tpm-tis: remove RAISE_STS_IRQ
  ...

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

oslib-posix: Fix compiler warning and some data types

gcc warning:

/qemu/util/oslib-posix.c:304:11: error:
 variable ‘addr’ might be clobbered by ‘longjmp’ or ‘vfork’
 [-Werror=clobbered]

Fix also some related data types:

numpages, hpagesize are used as pointer offset.
Always use size_t for them and also for the derived
numpages_per_thread and size_per_thread.

Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Message-id: 20171016202912.1117-1-sw@weilnetz.de
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

Merge remote-tracking branch 'remotes/mcayland/tags/qemu-openbios-signed' into staging

Update OpenBIOS images

# gpg: Signature made Thu 19 Oct 2017 20:08:44 BST
# gpg:                using RSA key 0x5BC2C56FAE0F321F
# gpg: Good signature from "Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>"
# Primary key fingerprint: CC62 1AB9 8E82 200D 915C  C9C4 5BC2 C56F AE0F 321F

* remotes/mcayland/tags/qemu-openbios-signed:
  Update OpenBIOS images to 83818bd built from submodule.

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

docker: Fix PATH for ccache

Before bcd7f06f57fb6f780a3e2f7a46c22b6f6c8238aa we source /etc/profile
so the PATH included the right paths to ccache binaries. Now we need to
update $PATH explicitly from run script.

Keep the old /usr/lib around just so that in the future, ccache from 32
bit images will just work.

Signed-off-by: Fam Zheng <famz@redhat.com>
Message-Id: <20171018073841.30062-1-famz@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>

docker: fix out-of-tree 'make docker-test-build@debian-powerpc-cross'

Without this patch, it fails with:

IOError: [Errno 2] No such file or directory:
 'tests/docker/dockerfiles/debian-apt-fake.sh'

Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <150841324224.31292.4012464539983692364.stgit@bahia.lan>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Fam Zheng <famz@redhat.com>

docker: allow running from srcdir != builddir build

The new script uses "git submodule", which is picky about being invoked
from the top of the git checkout.  Invoke the script from $(SRC_PATH)
to avoid git's wrath.

Fixes: b7f404201e45e99da23b9764dec27ce5f965cdcd
Cc: Fam Zheng <famz@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <1508331989-142364-1-git-send-email-pbonzini@redhat.com>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>

docker: cleanup temp directory after test

There are temp directories named "docker-src.*" after doing docker
tests.  I don't see much point in keeping that (it only contains the
qemu.tar which is exactly current tree, and the copied "run" file).
Let's remove it after test finished.

CC: "Alex Bennée" <alex.bennee@linaro.org>
CC: Fam Zheng <famz@redhat.com>
CC: "Philippe Mathieu-Daudé" <f4bug@amsat.org>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20171017071247.32355-1-peterx@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>

docker: Don't allocate tty unless DEBUG=1

The existence of tty in the container seems to urge gcc into colorizing
the errors, but the escape chars will clutter the report once turned
into email replies on patchew. Move -t to debug mode.

Reported-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
Message-Id: <20171013011954.9975-1-famz@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Fam Zheng <famz@redhat.com>

Update OpenBIOS images to 83818bd built from submodule.

Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>

Merge remote-tracking branch 'remotes/mcayland/tags/qemu-sparc-signed' into staging

qemu-sparc update

# gpg: Signature made Thu 19 Oct 2017 07:50:16 BST
# gpg:                using RSA key 0x5BC2C56FAE0F321F
# gpg: Good signature from "Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>"
# Primary key fingerprint: CC62 1AB9 8E82 200D 915C  C9C4 5BC2 C56F AE0F 321F

* remotes/mcayland/tags/qemu-sparc-signed:
  sun4u: fix assert when adding NICs which aren't the in-built model
  sun4u: update PCI topology to include simba PCI bridges

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

Merge remote-tracking branch 'remotes/kraxel/tags/seabios-1.11-prerelease-20171019-pull-request' into staging

seabios: update to 1.11 prerelease

# gpg: Signature made Thu 19 Oct 2017 09:19:33 BST
# gpg:                using RSA key 0x4CB6D8EED3E87138
# 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>"
# Primary key fingerprint: A032 8CFF B93A 17A7 9901  FE7D 4CB6 D8EE D3E8 7138

* remotes/kraxel/tags/seabios-1.11-prerelease-20171019-pull-request:
  seabios: update to 1.11 prerelease

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

build: Fix dtc-checkout race condition in Makefile

This was introduced by:
    commit aef45d51d1204f3335fb99de6658e0c5612c2b67
    Author: Daniel P. Berrange <berrange@redhat.com>
    Date:   Fri Sep 29 11:11:56 2017 +0100

build: automatically handle GIT submodule checkout for dtc

On my system, I see the following with a fresh clone:

% ./configure --disable-gtk --target-list=aarch64-softmmu
% make -j8
  GEN     aarch64-softmmu/config-devices.mak.tmp
  GEN     config-host.h
mkdir -p dtc/libfdt
  GIT     ui/keycodemapdb dtc
mkdir -p dtc/tests
  GEN     qemu-options.def
[snip]
  GEN     migration/trace.h
make: *** [git-submodule-update] Error 1
make: *** Waiting for unfinished jobs....

Upon closer inspection, the root cause of the error is:

% git submodule update --init ui/keycodemapdb dtc
fatal: destination path 'dtc' already exists and is not an empty directory.
Clone of 'git://git.qemu-project.org/dtc.git' into submodule path 'dtc' failed

This patch fixes this race condition by forcing the 'dtc/%' rule which caused
'dtc' to be non-empty to wait on '.git-submodule-status'.

Signed-off-by: Aaron Lindsay <alindsay@codeaurora.org>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Acked-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-id: 1508352023-28591-1-git-send-email-alindsay@codeaurora.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

tpm: move recv_data_callback to TPM interface

Simplify the TPM backend setup, move callback to TPM interface.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm: add a QOM TPM interface

This will simplify backend / interface objects relationship, so the
frontend interface will simply have to implement the TPM QOM interface.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm-tis: fold TPMTISEmuState in TPMState

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm-tis: remove tpm_tis.h header

The definitions are now private to TIS implementation.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm-tis: move TPMState to TIS header

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm: remove locty_data from TPMState

Keep it internal to tpm-tis instead.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm-emulator: fix error handling

The previous patch cleaned up a bit error handling, and exposed an
existing bug: error_report_err() could be called with a NULL error.
Instead, make tpm_emulator_set_locality() set the error.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm: add TPMBackendCmd to hold the request state

This simplifies a bit locality handling, and argument passing, and
could pave the way to queuing requests (if that makes sense).

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm: remove locty argument from receive_cb

The tpm_state is passed as argument, the assert() is pointless since
we give it the value of tpm_state->locty_number already.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm: remove needless cast

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm: remove unused TPMBackendCmd

There is only handling of request so far in both backends.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm: remove configure_tpm() hop

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm: remove init() class method

No backend use it.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm: remove TPMDriverOps

Use TPMBackendClass to hold class methods/fields.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm: move TPMSizedBuffer to tpm_tis.h

Close to where it's being used.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm: remove tpm_register_driver()

No more users of be_drivers[], drop that too.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm: replace tpm_get_backend_driver() to drop be_drivers

Use tpm_driver_find_by_type() instead.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm: lookup tpm backend class in tpm_driver_find_by_type()

One step towards removing TPMDriverOps and driver registration.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm: make tpm_get_backend_driver() static

No need to export the function.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm-tis: remove RAISE_STS_IRQ

This look like temporary hacking code. It shouldn't be necessary in
release code, or there should be a runtime option for it.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

tpm-tis: remove unused hw_access argument

This argument is always false, simplify the code.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>

Merge remote-tracking branch 'remotes/bonzini/tags/for-upstream' into staging

* TCG 8-byte atomic accesses bugfix (Andrew)
* Report disk rotation rate (Daniel)
* Report invalid scsi-disk block size configuration (Mark)
* KVM and memory API MemoryListener fixes (David, Maxime, Peter Xu)
* x86 CPU hotplug crash fix (Igor)
* Load/store API documentation (Peter Maydell)
* Small fixes by myself and Thomas
* qdev DEVICE_DELETED deferral (Michael)

# gpg: Signature made Wed 18 Oct 2017 10:56:24 BST
# gpg:                using RSA key 0xBFFBD25F78C7AE83
# gpg: Good signature from "Paolo Bonzini <bonzini@gnu.org>"
# gpg:                 aka "Paolo Bonzini <pbonzini@redhat.com>"
# Primary key fingerprint: 46F5 9FBD 57D6 12E7 BFD4  E2F7 7E15 100C CD36 69B1
#      Subkey fingerprint: F133 3857 4B66 2389 866C  7682 BFFB D25F 78C7 AE83

* remotes/bonzini/tags/for-upstream: (29 commits)
  scsi: reject configurations with logical block size > physical block size
  qdev: defer DEVICE_DEL event until instance_finalize()
  Revert "qdev: Free QemuOpts when the QOM path goes away"
  qdev: store DeviceState's canonical path to use when unparenting
  qemu-pr-helper: use new libmultipath API
  watch_mem_write: implement 8-byte accesses
  notdirty_mem_write: implement 8-byte accesses
  memory: reuse section_from_flat_range()
  kvm: simplify kvm_align_section()
  kvm: region_add and region_del is not called on updates
  kvm: fix error message when failing to unregister slot
  kvm: tolerate non-existing slot for log_start/log_stop/log_sync
  kvm: fix alignment of ram address
  memory: call log_start after region_add
  target/i386: trap on instructions longer than >15 bytes
  target/i386: introduce x86_ld*_code
  tco: add trace events
  docs/devel/loads-stores.rst: Document our various load and store APIs
  nios2: define tcg_env
  build: remove CONFIG_LIBDECNUMBER
  ...

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

Merge remote-tracking branch 'remotes/riku/tags/pull-linux-user-20171018' into staging

Linux-user updates for Qemu 2.11

# gpg: Signature made Wed 18 Oct 2017 13:20:14 BST
# gpg:                using RSA key 0xB44890DEDE3C9BC0
# gpg: Good signature from "Riku Voipio <riku.voipio@iki.fi>"
# gpg:                 aka "Riku Voipio <riku.voipio@linaro.org>"
# Primary key fingerprint: FF82 03C8 C391 98AE 0581  41EF B448 90DE DE3C 9BC0

* remotes/riku/tags/pull-linux-user-20171018:
  linux-user: Fix TARGET_MTIOCTOP/MTIOCGET/MTIOCPOS values
  linux-user/main: support dfilter
  linux-user: Fix target FS_IOC_GETFLAGS and FS_IOC_SETFLAGS numbers
  linux-user/sh4: Reduce TARGET_VIRT_ADDR_SPACE_BITS to 31
  linux-user: Tidy and enforce reserved_va initialization
  tcg: Fix off-by-one in assert in page_set_flags
  linux-user: Allow -R values up to 0xffff0000 for 32-bit ARM guests
  linux-user: remove duplicate break in syscall
  target/m68k,linux-user: manage FP registers in ucontext
  linux-user: fix O_TMPFILE handling

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

Merge remote-tracking branch 'remotes/kraxel/tags/opengl-20171017-pull-request' into staging

ui: opengl updates for dma-buf support.

# gpg: Signature made Tue 17 Oct 2017 12:13:36 BST
# gpg:                using RSA key 0x4CB6D8EED3E87138
# 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>"
# Primary key fingerprint: A032 8CFF B93A 17A7 9901  FE7D 4CB6 D8EE D3E8 7138

* remotes/kraxel/tags/opengl-20171017-pull-request:
  egl-headless: add dmabuf support
  egl-helpers: add egl_texture_blit and egl_texture_blend
  egl-helpers: add dmabuf import support
  opengl: add flipping vertex shader
  opengl: move shader init from console-gl.c to shader.c
  console: add support for dmabufs

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

Merge remote-tracking branch 'remotes/kraxel/tags/vga-20171017-pull-request' into staging

cirrus: bugfixes, with some vga cleanups.

# gpg: Signature made Tue 17 Oct 2017 09:24:37 BST
# gpg:                using RSA key 0x4CB6D8EED3E87138
# 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>"
# Primary key fingerprint: A032 8CFF B93A 17A7 9901  FE7D 4CB6 D8EE D3E8 7138

* remotes/kraxel/tags/vga-20171017-pull-request:
  cirrus: fix oob access in mode4and5 write functions
  vga: add ram_addr_t cast
  vga: handle cirrus vbe mode wraparounds.
  vga: drop line_offset variable

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

seabios: update to 1.11 prerelease

This is the seabios update for qemu 2.11.  Well, almost, seabios is in
freeze for the upcoming 1.11 release.  This updates seabios to current
git master snapshot, and it will be updated again to 1.11 final before
the 2.11 release.

With this two-step seabios gets some more wide testing before the actual
release and the update to 1.11 final (which will most likely happen
after qemu freeze) should have bugfix patches only.

git shortlog
============

Aleksandr Bezzubikov (3):
      pci: refactor pci_find_capapibilty to get bdf as the first argument instead of the whole pci_device
      pci: add QEMU-specific PCI capability structure
      pci: enable RedHat PCI bridges to reserve additional resources on PCI init

Ben Warren (5):
      QEMU DMA: Add DMA write capability
      romfile-loader: Switch to using named structs
      QEMU fw_cfg: Add command to write back address of file
      QEMU fw_cfg: Add functions for accessing files by key
      QEMU fw_cfg: Write fw_cfg back on S3 resume

Daniel Verkamp (5):
      nvme: support NVMe 1.0 controllers
      nvme: extend command timeout to 5 seconds
      nvme: fix reversed loop condition in cmd_readwrite
      nvme: fix extraction of status code bits
      nvme: fix copy-paste mistake in comment

Filippo Sironi (1):
      nvme: Use the Maximum Queue Entries Supported (MQES) to initialize I/O queues

Gerd Hoffmann (7):
      usb: add hub portmap
      usb-xhci: use hub portmap
      std: add cp437 to unicode map
      kbd: make enqueue_key public, add ascii_to_keycode
      romfile: add support for constant files.
      paravirt: serial console configuration.
      add serial console support

Igor Mammedov (1):
      drop "etc/boot-cpus" fw_cfg file and reuse legacy QEMU_CFG_NB_CPUS

Jason Wang (1):
      virtio: IOMMU support

Julian Stecklina (2):
      block: add NVMe boot support
      nvme: fix out of memory behavior

Julius Werner (1):
      coreboot: Adapt to upstream CBMEM console changes

Kevin O'Connor (26):
      usb: Make usb_time_sigatt variable static
      tpm: Add comment banners to tcg.c separating major parts of spec
      tpm: Don't call tpm_set_failure() from tpm12_get_capability()
      tpm: Move code around in tcgbios.c to keep like code together
      acpi: Generalize find_fadt() and find_tcpa_by_rsdp() into find_acpi_table()
      tpm: Don't call tpm_build_and_send_cmd() from tpm20_stirrandom()
      tpm: Rework tpm_build_and_send_cmd() into tpm_simple_cmd()
      ps2port: Disable keyboard/mouse prior to resetting ps2 controller
      docs: Note release dates for 1.10.1 and 1.10.2
      resume: Don't attempt to use generic reboot mechanisms on QEMU
      boot: Increase description size in boot menu
      src: Minor - remove tab characters that slipped into SeaBIOS C code
      NVMe: Allow NVMe to be enabled on real hardware
      smm: Backup and restore A20 on an SMI based mode switch
      stacks: Make sure to initialize Call16Data
      stacks: Don't update the A20 settings if they haven't changed
      stacks: There is no need to disable NMI if it is already disabled
      vga: Fix bug in stdvga_get_linesize()
      docs: Fix typos in Memory_Model.md
      tcgbios: Fix use of unitialized variable
      boot: Rename drive_g to drive
      disk: Don't require the 'struct drive_s' to be in the f-segment
      block: Rename disk_op_s->drive_gf to drive_fl
      virtio: Allocate drive_s storage in low memory
      xhci: Build TRBs directly in xhci_trb_queue()
      xhci: Verify the device is still present in xhci_cmd_submit()

Ladi Prosek (1):
      ahci: Set upper 32-bit registers to zero

Patrick Rudolph (4):
      SeaVGABios/cbvga: Advertise correct pixel format
      SeaVGABIOS/vbe: Query driver for scanline pitch v2
      SeaVGABios/cbvga: Use active mode to clear screen
      SeaVGABios/cbvga: Advertise compatible VESA modes

Paul Menzel (1):
      vgasrc: Increase debug level

Petr Berky (1):
      config: Add function to check if fw_cfg exists

Ricardo Ribalda Delgado (1):
      serialio: Support for mmap serial ports

Roman Kagan (11):
      blockcmd: accept only disks and CD-ROMs
      blockcmd: generic SCSI luns enumeration
      virtio-scsi: enumerate luns with REPORT LUNS
      esp-scsi: enumerate luns with REPORT LUNS
      usb-uas: enumerate luns with REPORT LUNS
      pvscsi: fix the comment about lun enumeration
      mpt-scsi: try to enumerate luns with REPORT LUNS
      lsi-scsi: reset in case of a serious problem
      lsi-scsi: try to enumerate luns with REPORT LUNS
      blockcmd: start REPORT_LUNS with the smallest buffer
      Revert "lsi-scsi: reset in case of a serious problem"

Stefan Berger (1):
      tpm: Log TPM 2 digest structure in little endian format

Youness Alaoui (1):
      nvme: Enable NVMe support for non-qemu hardware

Zeh, Werner (1):
      ahci: Disable Native Command Queueing

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

sun4u: fix assert when adding NICs which aren't the in-built model

Commit 8d93297 introduced a bug whereby non-inbuilt NICs are realized before
setting the default MAC address causing an assert. Switch NIC creation
over from pci_create_simple() to pci_create() which works exactly the
same except omitting the realize as originally intended.

Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Artyom Tarasenko <atar4qemu@gmail.com>

sun4u: update PCI topology to include simba PCI bridges

This patch updates the sun4u model to being much closer to a real Ultra 5
by moving devices behind the 2 simba PCI bridges (A and B) as found on real
hardware.

The most noticeable change introduced by this patchset is that in-built devices
are no longer attached to the PCI root bus, but instead behind PCI bridge A.
Along with this the interrupt routing is updated accordingly to match the
official documentation.

Since the existing code currently bypasses the PCI bridge interrupt
swizzling, the interrupt mapping functions are reorganised so that
pci_pbm_map_irq() is used by the PCI bridges and pci_apb_map_irq() is
used by the PCI host bridge.

Behind the sabre PCI host bridge, the PCI IO space now needs to be
split into two separate halves at 0x8000000. Therefore we also setup a new
PCI IO space region of increased size on the PCI host bridge and enable
32-bit PCI IO accesses to allow IO accesses to reach devices behind PCI
bridge B correctly.

As part of this change we also combine the onboard sunhme NIC and the ebus
into a single multi-function device as done on a real Ultra 5. For other
NICs the existing behaviour is preserved, i.e. we initialise them and
place them into the next free slot on PCI bus B.

Finally we mark the physically unavailable slots (plus slot 0 in busA) as
reserved to ensure that users can't plug devices into non-existent slots
which will break interrupt routing.

Note: since this commit changes PCI topology and interrupt routing, an
updated openbios-sparc64 binary is included with this commit containing the
associated changes to maintain bisectability.

Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Artyom Tarasenko <atar4qemu@gmail.com>

scsi: reject configurations with logical block size > physical block size

Logical block size of a SCSI disk should never be larger than
physical block size. From an ATA/SCSI perspective, it makes no sense
to have the logical block size greater than the physical block size,
and it cannot even be effectively expressed in the command set. The
whole point of adding the physical block size to the ATA/SCSI command
set was to communicate a desire for a larger block size (than logical),
while maintaining backwards compatibility with legacy 512 byte block
size.

When setting logical_block_size > physical_block_size, QEMU cannot express
it in READ CAPACITY(16) output, and all it can do is set the physical
block exponent to 0 (i.e. logical_block_size == physical_block_size).
Reporting the error properly, however, is better.

Signed-off-by: Mark Kanda <mark.kanda@oracle.com>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Message-Id: <1508185024-5840-1-git-send-email-mark.kanda@oracle.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

qdev: defer DEVICE_DEL event until instance_finalize()

DEVICE_DEL is currently emitted when a Device is unparented, as
opposed to when it is finalized. The main design motivation for this
seems to be that after unparent()/unrealize(), the Device is no
longer visible to the guest, and thus the operation is complete
from the perspective of management.

However, there are cases where remaining host-side cleanup is also
pertinent to management. The is generally handled by treating these
resources as aspects of the "backend", which can be managed via
separate interfaces/events, such as blockdev_add/del, netdev_add/del,
object_add/del, etc, but some devices do not have this level of
compartmentalization, namely vfio-pci, and possibly to lend themselves
well to it.

In the case of vfio-pci, the "backend" cleanup happens as part of
the finalization of the vfio-pci device itself, in particular the
cleanup of the VFIO group FD. Failing to wait for this cleanup can
result in tools like libvirt attempting to rebind the device to
the host while it's still being used by VFIO, which can result in
host crashes or other misbehavior depending on the host driver.

Deferring DEVICE_DEL still affords us the ability to manage backends
explicitly, while also addressing cases like vfio-pci's, so we
implement that approach here.

An alternative proposal involving having VFIO emit a separate event
to denote completion of host-side cleanup was discussed, but the
prevailing opinion seems to be that it is not worth the added
complexity, and leaves the issue open for other Device implementations
to solve in the future.

Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Tested-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20171016222315.407-4-mdroth@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Revert "qdev: Free QemuOpts when the QOM path goes away"

This reverts commit abed886ec60cf239a03515cf0b30fb11fa964c44.

This patch originally addressed an issue where a DEVICE_DELETED
event could be emitted (in device_unparent()) before a Device's
QemuOpts were cleaned up (in device_finalize()), leading to a
"duplicate ID" error if management attempted to immediately add
a device with the same ID in response to the DEVICE_DELETED event.

An alternative will be implemented in a subsequent patch where we
defer the DEVICE_DELETED event until device_finalize(), which would
also prevent the race, so we revert the original fix in preparation.

Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Tested-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20171016222315.407-3-mdroth@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>