The SIGP order CPU RESET was still missing in the list of our
supported handler. This patch now adds a simple implementation,
by using the cpu_reset() function that is already available in
target-s390x/cpu.c.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>

The s390_cpu_initial_reset() function had two deficiencies: First, it
used an ioctl for the destination CPU, and this ioctl could block
nearly forever, as long as the destination CPU was running in the SIE
loop. Second, it also cleared the general purpose registers - something
it should not do according to the Principles of Operations.
Since we've already got another function for the initial CPU reset in
cpu.c, we can also use that function instead. And by using run_on_cpu()
for executing this code, we make sure that the destination CPU is
correctly kicked out of kernel mode now.
Suggested-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>

In the SCLP handler function, the condition code register must
only be set if no exception occured.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>

If the 51 most significant bits of the SCCB address are zero or equal to
the prefix, we should throw an specification exception, too.
Also moved the check for privileged mode to sclp_service_call() to have
all program checks in one place now.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>

The pointer to the SCCB should not be limited to 32 bits only.
In contrast to this, the command word parameter is only 32 bits
(the upper 32 bits should be ignored).
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>

So far, the DIAG 500 hypervisor call was only setting -EINVAL in
R2 when a guest tried to call this function with an illegal subcode.
This patch now changes the behavior so that a specification exception
is thrown instead, since this is the common behavior of other DIAG
functions (and other CPU instructions) when being called with illegal
parameters.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>

The SET-ARCHITECTURE handler in QEMU caused a program interruption.
This is wrong according to the "Principles of Operations" specification
(since SIGP should never cause a program interrupt) and was likely only
introduced for debugging purposes. Since we handle SET-ARCHITECTURE in
the kernel already and only dropped to user space in case of bad mode
parameters, we should just report INVALID PARAMETER in QEMU instead.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>

S390 can also use async page faults, to enhance guest scheduling.
In case of live migration we want to disable the feature and let
all pending request finish.
Signed-off-by: NDominik Dingel <dingel@linux.vnet.ibm.com>
Signed-off-by: NJens Freimann <jfrei@linux.vnet.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>

The partial register handling (introduced with commits
420840e5 and
3474b679 ) aimed to improve intercept
handling performance.

It made the code more complicated though. During development for life
migration/init/reset etc it turned out that this might cause several
hard to debug programming errors. With the introduction of ioeventfd
(and future irqfd patches) the qemu intercept handlers are no longer
hot-path. And therefore the partial register handling can be
removed to simplify the code.
Signed-off-by: NDominik Dingel <dingel@linux.vnet.ibm.com>
CC: Jason J. Herne <jjherne@us.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Acked-by: NAlexander Graf <agraf@suse.de>

If SIGP is called with an unknown order code, it has to return CC1
instead of CC3 and set the "invalid order" bit in the return status.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NJens Freimann <jfrei@linux.vnet.ibm.com>
Signed-off-by: NAlexander Graf <agraf@suse.de>

We've already got a helper function for calculating the
base/displacement of RS formatted instructions, so we can
get rid of the manual calculation of the SIGP order code.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NJens Freimann <jfrei@linux.vnet.ibm.com>
Signed-off-by: NAlexander Graf <agraf@suse.de>

This patch adds the missing START order to the SIGP instruction handler.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NJens Freimann <jfrei@linux.vnet.ibm.com>
Signed-off-by: NAlexander Graf <agraf@suse.de>

To make scripts/checkpatch.pl happy for the following patches,
the coding style in handle_sigp() has to be fixed first.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Acked-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NJens Freimann <jfrei@linux.vnet.ibm.com>
Signed-off-by: NAlexander Graf <agraf@suse.de>

The SIGP order STORE STATUS AT ADDRESS will be handled in
kernel space, so we do not need the stub in QEMU anymore.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NJens Freimann <jfrei@linux.vnet.ibm.com>
Signed-off-by: NAlexander Graf <agraf@suse.de>

The SIGP order defines are also available in cpu.h,
so there is no need to re-define them in kvm.c.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NJens Freimann <jfrei@linux.vnet.ibm.com>
Signed-off-by: NAlexander Graf <agraf@suse.de>

The instruction intercept handler for diagnose used only the displacement
when trying to calculate the function code. This is only correct for base
0, however; we need to perform a complete base/displacement address
calculation and use bits 48-63 as the function code.
Reviewed-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Signed-off-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NJens Freimann <jfrei@linux.vnet.ibm.com>
Signed-off-by: NAlexander Graf <agraf@suse.de>

phys_mem_alloc and its assigned values qemu_anon_ram_alloc and
legacy_s390_alloc must have identical argument lists.

legacy_s390_alloc uses the size parameter to call mmap, so size_t is
good enough for all of them.

This patch fixes compiler errors on i686 Linux hosts:

  CC    alpha-softmmu/exec.o
exec.c:752:51: error:
 initialization from incompatible pointer type [-Werror]
exec.c: In function 'qemu_ram_alloc_from_ptr':
exec.c:1139:32: error:
 comparison of distinct pointer types lacks a cast [-Werror]
exec.c: In function 'qemu_ram_remap':
exec.c:1283:21: error:
 comparison of distinct pointer types lacks a cast [-Werror]
Signed-off-by: NStefan Weil <sw@weilnetz.de>
Reviewed-by: NMarkus Armbruster <armbru@redhat.com>
Message-id: 1380481005-32399-1-git-send-email-sw@weilnetz.de
Signed-off-by: NAnthony Liguori <aliguori@amazon.com>

The IO instruction handlers now take care of setting the CC value on
their own, so that the confusing return code magic in kvm_handle_css_inst()
is not needed anymore.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>

Moved the setcc() function to cpu.h so that it can be used by other
files, too. It now also does not modify the kvm state anymore since
this gets updated during kvm_arch_put_registers() anyway.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>

The SCLP instruction is priviledged, so we should make sure that
we generate an exception when it is called from the problem state.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>

We abort() on memory allocation failure.  abort() is appropriate for
programming errors.  Maybe most memory allocation failures are
programming errors, maybe not.  But guest memory allocation failure
isn't, and aborting when the user asks for more memory than we can
provide is not nice.  exit(1) instead, and do it in just one place, so
the error message is consistent.
Tested-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Reviewed-by: NAndreas Färber <afaerber@suse.de>
Acked-by: NLaszlo Ersek <lersek@redhat.com>
Acked-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Message-id: 1375276272-15988-8-git-send-email-armbru@redhat.com
Signed-off-by: NAnthony Liguori <anthony@codemonkey.ws>

Make it a generic hook rather than a KVM hook.  Less code and
ifdeffery.

Since the only user of the hook is old S390 KVM, there's hope we can
get rid of it some day.
Acked-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: NMarkus Armbruster <armbru@redhat.com>
Acked-by: NLaszlo Ersek <lersek@redhat.com>
Acked-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Message-id: 1375276272-15988-5-git-send-email-armbru@redhat.com
Signed-off-by: NAnthony Liguori <anthony@codemonkey.ws>

There is the 'nmi' command that is used to trigger a guest dump via kdump feature on x86.
s390 uses RESTART interrupt to trigger kdump.
So, this patch provides a mean to use 'nmi' command on s390 to raise RESTART interrupt.

The CPU to receive the RESTART interrupt is the "default" one.

There is an infrastructure to select the "default" CPU using 'cpu' command.
The 'info cpus' command can be used to see which one is the "default".

In order to wire up the RESTART to 'nmi' command we had to:
1. implement the kvm_s390_cpu_restart function by exporting the existing code
2. implement s390_cpu_restart function as kvm-aware wrapper
3. modify the qmp_inject_nmi function to enable (for s390) the scan for
   "default" CPU and call s390_cpu_restart for it;
3. fix some messages.
Signed-off-by: NEugene (jno) Dvurechenski <jno@linux.vnet.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Acked-by: NAlexander Graf <agraf@suse.de>

Linux uses a check for subcode 6 to decide if other subcodes are
available. Provide a minimal implementation for subcode 6, as well
as for subcode 5.
Signed-off-by: NEugene (jno) Dvurechenski <jno@linux.vnet.ibm.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>
[Move code from kvm.c into misc_helper.c]

This alignes case statements to switch statements in the handle_diag
function as mandated by coding style.
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>

Removed the redundant return code statement from handle_instruction() - it
always returned 0 and never reports any errors to its caller, since errors
from the sub-functions are already reported via program exceptions instead.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>

Consolidated the setting of the condition code in kvm_handle_css_inst().
For the (unhandled) instructions EQBS and SQBS, we have to return
an operation exception instead of cc3. Also removed the is_ioinst()
function to avoid decoding the opcode twice.
Signed-off-by: NThomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>

'dprintf' is the name of a POSIX standard function so we should not be
stealing it for our debug macro. Rename to 'DPRINTF' (in line with
a number of other source files.)
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>
Reviewed-by: NAndreas Färber <afaerber@suse.de>
Reviewed-by: NMORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Acked-by: NRichard Henderson <rth@twiddle.net>
Acked-by: NKevin Wolf <kwolf@redhat.com>
Message-id: 1375100199-13934-3-git-send-email-peter.maydell@linaro.org
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

Commit f17ec444
exec: Change cpu_memory_rw_debug() argument to CPUState

missed to update s390x KVM code, breaking the build.

Let's fix it up.
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: NAndreas Färber <afaerber@suse.de>

This allows to get rid of the last remaining ENV_GET_CPU() in
target-s390x/ by using CPU() cast directly on the argument.

Cc: Jason J. Herne <jjherne@us.ibm.com>
Signed-off-by: NAndreas Färber <afaerber@suse.de>

Commit 3474b679 (Utilize selective
runtime reg sync for hot code paths) introduced two uses of
ENV_GET_CPU() inside target-s390x/ KVM code. In one case we can use a
direct CPU() cast instead.

Cc: Jason J. Herne <jjherne@us.ibm.com>
Signed-off-by: NAndreas Färber <afaerber@suse.de>

event_notifier_get_fd() is not available on windows hosts. Fix this by
moving the calls to event_notifier_get_fd() to the kvm code.
Reported-by: NStefan Weil <sw@weilnetz.de>
Reviewed-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NCornelia Huck <cornelia.huck@de.ibm.com>

Change Monitor::mon_cpu to CPUState as well.
Reviewed-by: Nliguang <lig.fnst@cn.fujitsu.com>
Acked-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NAndreas Färber <afaerber@suse.de>

On hosts that support ioeventfd, make use of it for host-to-guest
notifications via diagnose 500.
Signed-off-by: NCornelia Huck <cornelia.huck@de.ibm.com>

The common KVM code insists on calling kvm_arch_init_irq_routing()
as soon as it sees kernel header support for it (regardless of whether
QEMU supports it).  Provide a dummy function to satisfy this.
Signed-off-by: NAlexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: NMichael Tokarev <mjt@tls.msk.ru>

This is preparatory to the introduction of a separate freeing API.
Reported-by: NAmos Kong <akong@redhat.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>
Reviewed-by: NAmos Kong <akong@redhat.com>
Message-id: 1368454796-14989-2-git-send-email-pbonzini@redhat.com
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

On s390 the disabled wait state indicates a state of attention.
For example Linux uses that state after a panic. Lets
put the system into panicked state.

An alternative implementation would be to state
disabled-wait <address> instead of pause in the action field.
(e.g. z/OS, z/VM and other classic OSes use the address of the
disabled wait to indicate an error code).
Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Message-id: 6cf41156322e27e81a727b69f03728dbc225d5bb.1366945969.git.hutao@cn.fujitsu.com
Signed-off-by: NAnthony Liguori <aliguori@us.ibm.com>

Make use of new kvm_s390_get_registers_partial() for kvm_handle_css_inst() and
handle_hypercall() since they only need registers from the partial set and they
are called quite frequently.
Signed-off-by: NJason J. Herne <jjherne@us.ibm.com>
Signed-off-by: NAlexander Graf <agraf@suse.de>

We want to avoid expensive register synchronization IOCTL's on the hot path so
a new kvm_s390_get_registers_partial() is introduced as a compliment to
kvm_arch_get_registers().  The new function is called on the hot path, and
kvm_arch_get_registers() is called when we need the complete runtime register
state.

kvm_arch_put_registers() is updated to only sync the partial runtime set when
we've only dirtied the partial runtime set.  This is to avoid sending bad data
back to KVM if we've only partially synced the runtime register set.
Signed-off-by: NJason J. Herne <jjherne@us.ibm.com>
Reviewed-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: NAlexander Graf <agraf@suse.de>

Commit 259186a7 "cpu: Move halted and interrupt_request fields to CPUState"
seems to have missed one instance in target-s390x/kvm.c:

/home/cohuck/git/qemu/target-s390x/kvm.c: In function ‘kvm_arch_process_async_events’:
/home/cohuck/git/qemu/target-s390x/kvm.c:319: error: ‘CPUS390XState’ has no member named ‘halted’
/home/cohuck/git/qemu/target-s390x/kvm.c:320: warning: control reaches end of non-void function
make[1]: *** [target-s390x/kvm.o] Error 1

Let's just switch to cs->halted.
Signed-off-by: NCornelia Huck <cornelia.huck@de.ibm.com>
Acked-by: NAlexander Graf <agraf@suse.de>
Acked-by: NAndreas Färber <afaerber@suse.de>
Signed-off-by: NBlue Swirl <blauwirbel@gmail.com>