We connect to xenstored and can send the INTRODUCE_DOMAIN message with
the pfn and domid. This will allow xenstored to map the domain ring and
channel to serve xenbus (domain) client requests.
xs_read()'s failure is indicated by NULL return value while @len will
report length of error string. If /tool/xenstored/domid mnode does not
exist in xenstore then we assume that Domain-0 is where xenstore is
hosted.
Try to fetch where xenstore domain leaves such that we can create the
unbound ports and seed grant table accordingly.
Finally seed xenstore with the default entries. The entries comprised at
init time (before introduce_domain) and are all domain local paths:
All of which have r/w permission alongside parent its domain path
(/local/domain/<domid>). When we gracefully shutdown or 'quit' via
HMP/QMP we properly release the domain with the release_domain()
message. Unattended shutdown is handled by xenstored itself and
hypervisor propragating the VIRQ_DOM_EXC interrupt to xenstored.
qemu allocates the SPECIALPAGE_XENSTORE page as a MAP_PRIVATE |
MAP_ANON mapping. The same mapping will be done on xenstored (via
/dev/xen/gntdev) as a MAP_SHARED mapping on the SPECIALPAGE_XENSTORE
grant-entry.
This leads to an interesting breakage of xenstored's rss counter:
BUG: Bad rss-counter state mm:00000000acd9575d idx:1 val:-1
BUG: Bad rss-counter state mm:00000000acd9575d idx:3 val:1
Since it seems nonsensical to do a MAP_PRIVATE mapping for a page
that is meant to be shared with a different process, we switch
to a MAP_SHARED | MAP_ANON mapping.
Which is used to fetch xenstore PFN and port to be used
by the guest. This is preallocated by the toolstack when
guest will just read those and use it straight away.
We support only grant table layout version 1 (for now).
We seed the initial grant table frame which contains the
preallocated reserved grant entries such as xenstore and
console which are used by the backend to map the ring
and such.
For now we limit to the most widely supported i.e. version 1.
Initializing grant table means adding support for:
* XENMAPSPACE_shared_info in XENMEM_add_to_physmap hypercall
* GNTTABOP_set_version hypercall
Also add bounds check for event channel ports and also do not set a
eventfd for a Dom0 unbound port.
When running xenstored in Domain-0 we may have a remote domid of 0
(other than DOMID_SELF) which is the case when xenstored runs on
Domain-0. In such cases remote_dom being 0 is valid.
Uses XEN_EVENTFD ioctl to assign a eventfd to an ipi port and
therefore notify the guest. This allows offloading PV IPIs
to hypervisor instead of through Qemu.
Introduce support for one shot and periodic mode of Xen PV timers,
whereby timer interrupts come through a special virq event channel
with deadlines being set through:
Joao Martins [Thu, 30 Aug 2018 18:46:38 +0000 (14:46 -0400)]
i386/xen: evtchn IPI support
IPI event channels are per cpu too and require an additional
hypercall to be implemented which is EVTCHNOP_send used when
a CPU wants to notify a target vcpu.
Joao Martins [Tue, 21 Aug 2018 16:14:20 +0000 (12:14 -0400)]
i386/xen: evtchn VIRQ support
Add VIRQ support and handle the basic ops to support event
channels: bind, unmask, close, status.
For now we only implement the 2 level ABI, which is limited
to 4k event channels on 64-bit or 1k on 32-bit. The basic
mechanism works by setting the shared_info bit array and
setting the port bit to 1 in evtchn_pending. Additionally
we have another word which depicts which groups have pending
ports.
Current implementations of test_and_set/clear_bit() are
non-atomic and because Xen HVM guests (on KVM), a word can be
accessed simultaneously by qemu and the hypervisor (and possibly
guest itself). And thus we need bitops atomic ops. The new header
is ported from what's used on Xen.
Co-developed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com> Signed-off-by: Joao Martins <joao.m.martins@oracle.com> Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Ankur Arora [Thu, 9 Aug 2018 18:24:30 +0000 (14:24 -0400)]
i386/xen: handle upcall related hypercalls
It handles HVMOP_set_param(.index = HVM_PARAM_CALLBACK_IRQ) which
sets the upcall IRQ from the guest. This is not a APIC asserted
vector and hence we need to add a new special irqchip route which
further requires special hypervisor enlightment. We gate enabling
this with KVM_CAP_XEN_HVM_GUEST which has the necessary irqchip
route type support.
Ankur Arora [Fri, 17 Aug 2018 23:52:05 +0000 (16:52 -0700)]
kvm/ioapic: mark gsi-2 used in ioapic routing init
GSI-2/IOAPIC pin-2 is treated specially while initing
IRQ routing: PIC does not use it at all while the IOAPIC
maps virq=0 to pin-2 and does not use GSI-2.
(all other GSIs are identity mapped to pins.)
This results in any later code which allocates a virq
to be assigned GSI-2. This virq is in-turn used to
remap interrupts to HYPERVISOR_CALLBACK_VECTOR (0xf3)
to deliver to the guest.
Ordinarily this would be okay, but if the event delivery is
via direct injection via KVM_REQ_EVENT (without going
through the LAPIC) we see vmentry failure.
Joao Martins [Fri, 29 Jun 2018 14:54:50 +0000 (10:54 -0400)]
i386/xen: handle register_vcpu_info
Handle the hypercall to set a per vcpu info, as opposed to using
shared_info equivalent. Also, Guests may not call
VCPUOP_register_vcpu_info and will fail in event channel operations if a
proper one isn't set in Qemu.
So derive the hva from shared_info which is where these are located when
guest doesn't seed an additional pointer.
Joao Martins [Mon, 18 Jun 2018 16:26:44 +0000 (12:26 -0400)]
i386/xen: implement HYPERVISOR_vcpu_op
This is simply when guest tries to register a vcpu_info
and since vcpu_info placement is optional in the minimum ABI
therefore we can just fail with -ENOSYS
Joao Martins [Mon, 18 Jun 2018 16:17:42 +0000 (12:17 -0400)]
i386/xen: set shared_info page
This is done by implementing HYPERVISOR_memory_op specifically
XENMEM_add_to_physmap with space XENMAPSPACE_shared_info. While
Xen removes the page with it's own, we instead use the gfn passed
by the guest.
Joao Martins [Thu, 14 Jun 2018 12:29:45 +0000 (08:29 -0400)]
i386/xen: implement HYPERCALL_xen_version
This is just meant to serve as an example on how we can implement
hypercalls. xen_version specifically since Qemu does all kind of
feature controllability. So handling that here seems appropriate.
Joao Martins [Wed, 13 Jun 2018 14:14:31 +0000 (10:14 -0400)]
i386/xen: handle guest hypercalls
This means handling the new exit reason for Xen but still
crashing on purpose. As we implement each of the hypercalls
we will then return the right return code.
Joao Martins [Tue, 12 Jun 2018 21:33:05 +0000 (22:33 +0100)]
i386/xen: handle Xen hypercall page
Firstly create the blob and set up the vmmcall. Afterwards
we update number of hypercall pages to 1, and then make the
vm ioctl to set XEN_HVM_CONFIG. This allows the guest to
make hypercalls
* remotes/dgibson/tags/ppc-for-4.0-20190219: (43 commits)
target/ppc: convert vmin* and vmax* to vector operations
target/ppc: convert vadd*s and vsub*s to vector operations
target/ppc: Split out VSCR_SAT to a vector field
target/ppc: Add set_vscr_sat
target/ppc: Use mtvscr/mfvscr for vmstate
target/ppc: Add helper_mfvscr
target/ppc: Remove vscr_nj and vscr_sat
target/ppc: Use helper_mtvscr for reset and gdb
target/ppc: Pass integer to helper_mtvscr
target/ppc: convert xxsel to vector operations
target/ppc: convert xxspltw to vector operations
target/ppc: convert xxspltib to vector operations
target/ppc: convert VSX logical operations to vector operations
target/ppc: convert vsplt[bhw] to use vector operations
target/ppc: convert vspltis[bhw] to use vector operations
target/ppc: convert vaddu[b,h,w,d] and vsubu[b,h,w,d] over to use vector operations
target/ppc: convert VMX logical instructions to use vector operations
xics: Drop the KVM ICS class
spapr/irq: Use the "simple" ICS class for KVM
xics: Handle KVM interrupt presentation from "simple" ICS code
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Peter Maydell [Mon, 18 Feb 2019 14:23:13 +0000 (14:23 +0000)]
Merge remote-tracking branch 'remotes/armbru/tags/pull-qapi-2019-02-18' into staging
QAPI patches for 2019-02-18
# gpg: Signature made Mon 18 Feb 2019 13:44:30 GMT
# gpg: using RSA key 3870B400EB918653
# gpg: Good signature from "Markus Armbruster <armbru@redhat.com>" [full]
# gpg: aka "Markus Armbruster <armbru@pond.sub.org>" [full]
# Primary key fingerprint: 354B C8B3 D7EB 2A6B 6867 4E5F 3870 B400 EB91 8653
* remotes/armbru/tags/pull-qapi-2019-02-18:
qapi: move RTC_CHANGE to the target schema
qmp: Deprecate query-events in favor of query-qmp-schema
Revert "qapi-events: add 'if' condition to implicit event enum"
qapi: remove qmp_unregister_command()
qapi: make query-cpu-definitions depend on specific targets
qapi: make query-cpu-model-expansion depend on s390 or x86
qapi: make query-gic-capabilities depend on TARGET_ARM
target.json: add a note about query-cpu* not being s390x-specific
qapi: make s390 commands depend on TARGET_S390X
qapi: make rtc-reset-reinjection and SEV depend on TARGET_I386
qapi: New module target.json
build: Deal with all of QAPI's .o in qapi/Makefile.objs
build-sys: move qmp-introspect per target
qapi: Generate QAPIEvent stuff into separate files
qapi: Prepare for system modules other than 'builtin'
qapi: Clean up modular built-in code generation a bit
qapi: Fix up documentation for recent commit a95291007b2
qapi: Belatedly document modular code generation
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Marc-André Lureau [Thu, 14 Feb 2019 15:22:51 +0000 (16:22 +0100)]
qapi: move RTC_CHANGE to the target schema
A few targets don't emit RTC_CHANGE, we could restrict the event to
the tagets that do emit it.
Note: There is a lot more of events & commands that we could restrict
to capable targets, with the cost of some additional complexity, but
the benefit of added correctness and better introspection.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20190214152251.2073-19-armbru@redhat.com>
Markus Armbruster [Thu, 14 Feb 2019 15:22:50 +0000 (16:22 +0100)]
qmp: Deprecate query-events in favor of query-qmp-schema
query-events doesn't reflect compile-time configuration. Instead of
fixing that, deprecate the command in favor of query-qmp-schema.
Libvirt prefers query-qmp-schema as of commit 22d7222ec0 "qemu: caps:
Don't call 'query-events' when we probe events from QMP schema".
It'll be in the next release.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20190214152251.2073-18-armbru@redhat.com>
The commit applied the events' conditions to the members of enum
QAPIEvent. Awkward, because it renders QAPIEvent unusable in
target-independent code as soon as we make an event target-dependent.
Reverting this has the following effects:
* ui/vnc.c can remain target independent.
* monitor_qapi_event_conf[] doesn't have to muck around with #ifdef.
* query-events again doesn't reflect conditionals. I'm going to
deprecate it in favor of query-qmp-schema.
Another option would be to split target-dependent parts off enum
QAPIEvent into a target-dependent enum. Doesn't seem worthwhile right
now.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20190214152251.2073-17-armbru@redhat.com>
Markus Armbruster [Thu, 14 Feb 2019 15:22:41 +0000 (16:22 +0100)]
qapi: New module target.json
We can't add appropriate target-specific conditionals to misc.json,
because that would make all of misc.json unusable in
target-independent code. To keep misc.json target-independent, we
need to split off target-dependent target.json.
This commit doesn't actually split off anything, it merely creates the
empty module. The next few patches will move stuff from misc.json
there.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20190214152251.2073-9-armbru@redhat.com>
Markus Armbruster [Thu, 14 Feb 2019 15:22:40 +0000 (16:22 +0100)]
build: Deal with all of QAPI's .o in qapi/Makefile.objs
Adding QAPI's .o to util-obj-y, common-obj-y and obj-y is spread over
three places: Makefile.objs takes care of target-independent generated
code, Makefile.target of target-dependent generated code, and
qapi/Makefile.objs of (target-independent) hand-written code.
Do everything in qapi/Makefile.objs.
Suggested-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20190214152251.2073-8-armbru@redhat.com>
Marc-André Lureau [Thu, 14 Feb 2019 15:22:39 +0000 (16:22 +0100)]
build-sys: move qmp-introspect per target
The following patches are going to introduce per-target #ifdef in the
schemas.
The introspection data is statically generated once, and must thus be
built per-target to reflect target-specific configuration.
Drop "do_test_visitor_in_qmp_introspect(&qmp_schema_qlit)" since the
schema is no longer in a common object. It is covered by the per-target
query-qmp-schema test instead.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20190214152251.2073-7-armbru@redhat.com>
Markus Armbruster [Thu, 14 Feb 2019 15:22:38 +0000 (16:22 +0100)]
qapi: Generate QAPIEvent stuff into separate files
Having to include qapi-events.h just for QAPIEvent is suboptimal, but
quite tolerable now. It'll become problematic when we have events
conditional on the target, because then qapi-events.h won't be usable
from target-independent code anymore. Avoid that by generating it
into separate files.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20190214152251.2073-6-armbru@redhat.com>
Markus Armbruster [Thu, 14 Feb 2019 15:22:37 +0000 (16:22 +0100)]
qapi: Prepare for system modules other than 'builtin'
The next commit wants to generate qapi-emit-events.{c.h}. To enable
that, extend QAPISchemaModularCVisitor to support additional "system
modules", i.e. modules that don't correspond to a (user-defined) QAPI
schema module.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20190214152251.2073-5-armbru@redhat.com>
Markus Armbruster [Thu, 14 Feb 2019 15:22:36 +0000 (16:22 +0100)]
qapi: Clean up modular built-in code generation a bit
We neglect to call .visit_module() for the special module we use for
built-ins. Harmless, but clean it up anyway. The
tests/qapi-schema/*.out now show the built-in module as 'module None'.
Subclasses of QAPISchemaModularCVisitor need to ._add_module() this
special module to enable code generation for built-ins. When this
hasn't been done, QAPISchemaModularCVisitor.visit_module() does
nothing for the special module. That looks like built-ins could
accidentally be generated into the wrong module when a subclass
neglects to call ._add_module(). Can't happen, because built-ins are
all visited before any other module. But that's non-obvious. Switch
off code generation explicitly.
Rename QAPISchemaModularCVisitor._begin_module() to
._begin_user_module().
New QAPISchemaModularCVisitor._is_builtin_module(), for clarity.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20190214152251.2073-4-armbru@redhat.com>
Markus Armbruster [Thu, 14 Feb 2019 15:22:34 +0000 (16:22 +0100)]
qapi: Belatedly document modular code generation
We generate code for built-ins and sub-modules into separate files
since commit cdb6610ae42 and 252dc3105fc (v2.12.0). Both commits
neglected to update documentation. Do that now.
Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20190214152251.2073-2-armbru@redhat.com>
Peter Maydell [Mon, 18 Feb 2019 11:32:00 +0000 (11:32 +0000)]
Merge remote-tracking branch 'remotes/cohuck/tags/s390x-20190218' into staging
s390x updates:
- tcg: implement STCK and friends for CONFIG_USER_ONLY
- add zpci to qemu cpu model, as pci is now always built
- add mepoch to default z14 cpu model
- add cpu model for z14 GA2
- various improvements
* remotes/cohuck/tags/s390x-20190218:
s390x: upgrade status of KVM cores to "supported"
s390x/kvm: add tracepoint to ioeventfd interface
s390x/cpumodel: add z14 GA2 model
s390x/cpumodel: default enable mepoch for z14 and later
s390x/cpumodel: mepochptff: warn when no mepoch and re-align group init
s390x: add zPCI feature to "qemu" CPU model
target/s390x: Implement STCK et al for CONFIG_USER_ONLY
target/s390x: Split out s390-tod.h
s390x: always provide pci support
s390x: Fix the confusing contributions-after-2012 license statements
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Collin Walling [Tue, 12 Feb 2019 01:16:56 +0000 (20:16 -0500)]
s390x/cpumodel: default enable mepoch for z14 and later
Latest systems and host kernels support mepoch, which is a
feature that was meant to be supported for z14 GA1 from the
get-go. Let's copy it to the z14 GA1 default CPU model.
Machines s390-ccw-virtio-3.1 and older will retain the old CPU
models and will not provide this bit nor the extended PTFF
functions in the default model.
Collin Walling [Tue, 12 Feb 2019 01:16:55 +0000 (20:16 -0500)]
s390x/cpumodel: mepochptff: warn when no mepoch and re-align group init
The extended PTFF features (qsie, qtoue, stoe, stoue) are dependent
on the multiple-epoch facility (mepoch). Let's print a warning if these
features are enabled without mepoch.
While we're at it, let's move the FEAT_GROUP_INIT for mepochptff down
the s390_feature_groups list so it can be properly indexed with its
generated S390FeatGroup enum.
Signed-off-by: Collin Walling <walling@linux.ibm.com>
Message-Id: <20190212011657.18324-1-walling@linux.ibm.com> Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com> Reviewed-by: David Hildenbrand <david@redhat.com> Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Richard Henderson [Tue, 12 Feb 2019 05:30:44 +0000 (21:30 -0800)]
target/s390x: Implement STCK et al for CONFIG_USER_ONLY
This is a non-privileged instruction that was only implemented
for system mode. However, the stck instruction is used by glibc,
so this was causing SIGILL for programs run under debian stretch.
Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: David Hildenbrand <david@redhat.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20190212053044.29015-3-richard.henderson@linaro.org> Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Cornelia Huck [Mon, 11 Feb 2019 11:32:55 +0000 (12:32 +0100)]
s390x: always provide pci support
We tried to make pci support optional on s390x in the past;
unfortunately, we still require the s390 phb to be created
unconditionally due to backwards compatibility issues.
Instead of sinking more effort into this (including compat
handling for older machines etc.) for non-obvious gains, let's
just make CONFIG_PCI something that is always set on s390x.
Note that you can still fence off pci for the _guest_ if you
provide a cpu model without the zpci feature.
Message-Id: <20190211113255.3837-1-cohuck@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: David Hildenbrand <david@redhat.com> Acked-by: Christian Borntraeger <borntraeger@de.ibm.com> Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Thomas Huth [Wed, 6 Feb 2019 12:41:33 +0000 (13:41 +0100)]
s390x: Fix the confusing contributions-after-2012 license statements
The license information in these files is rather confusing. The text
declares LGPL first, but then says that contributions after 2012 are
licensed under the GPL instead. How should the average user who just
downloaded the release tarball know which part is now GPL and which
is LGPL?
Looking at the text of the LGPL (see COPYING.LIB in the top directory),
the license clearly states how this should be done instead:
"3. You may opt to apply the terms of the ordinary GNU General Public
License instead of this License to a given copy of the Library. To do
this, you must alter all the notices that refer to this License, so
that they refer to the ordinary GNU General Public License, version 2,
instead of to this License."
Thus let's clean up the confusing statements and use the proper GPL
text only.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <1549456893-16589-1-git-send-email-thuth@redhat.com> Acked-by: Laurent Vivier <laurent@vivier.eu> Acked-by: Christian Borntraeger <borntraeger@de.ibm.com> Signed-off-by: Cornelia Huck <cohuck@redhat.com>
Richard Henderson [Fri, 15 Feb 2019 10:00:58 +0000 (10:00 +0000)]
target/ppc: convert vmin* and vmax* to vector operations
Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-18-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Richard Henderson [Fri, 15 Feb 2019 10:00:57 +0000 (10:00 +0000)]
target/ppc: convert vadd*s and vsub*s to vector operations
Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-17-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Richard Henderson [Fri, 15 Feb 2019 10:00:56 +0000 (10:00 +0000)]
target/ppc: Split out VSCR_SAT to a vector field
Change the representation of VSCR_SAT such that it is easy
to set from vector code.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-16-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Richard Henderson [Fri, 15 Feb 2019 10:00:55 +0000 (10:00 +0000)]
target/ppc: Add set_vscr_sat
This is required before changing the representation of the register.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-15-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Richard Henderson [Fri, 15 Feb 2019 10:00:54 +0000 (10:00 +0000)]
target/ppc: Use mtvscr/mfvscr for vmstate
This is required before changing the representation of the register.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-14-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Richard Henderson [Fri, 15 Feb 2019 10:00:53 +0000 (10:00 +0000)]
target/ppc: Add helper_mfvscr
This is required before changing the representation of the register.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-13-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Richard Henderson [Fri, 15 Feb 2019 10:00:52 +0000 (10:00 +0000)]
target/ppc: Remove vscr_nj and vscr_sat
These macros are no longer used.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-12-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Richard Henderson [Fri, 15 Feb 2019 10:00:51 +0000 (10:00 +0000)]
target/ppc: Use helper_mtvscr for reset and gdb
Not setting flush_to_zero from gdb_set_avr_reg was a bug.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-11-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Richard Henderson [Fri, 15 Feb 2019 10:00:50 +0000 (10:00 +0000)]
target/ppc: Pass integer to helper_mtvscr
We can re-use this helper elsewhere if we're not passing
in an entire vector register.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-10-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Richard Henderson [Fri, 15 Feb 2019 10:00:49 +0000 (10:00 +0000)]
target/ppc: convert xxsel to vector operations
Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-9-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Richard Henderson [Fri, 15 Feb 2019 10:00:48 +0000 (10:00 +0000)]
target/ppc: convert xxspltw to vector operations
Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-8-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Richard Henderson [Fri, 15 Feb 2019 10:00:47 +0000 (10:00 +0000)]
target/ppc: convert xxspltib to vector operations
Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-7-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Richard Henderson [Fri, 15 Feb 2019 10:00:46 +0000 (10:00 +0000)]
target/ppc: convert VSX logical operations to vector operations
Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-6-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Richard Henderson [Fri, 15 Feb 2019 10:00:45 +0000 (10:00 +0000)]
target/ppc: convert vsplt[bhw] to use vector operations
Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-5-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Richard Henderson [Fri, 15 Feb 2019 10:00:44 +0000 (10:00 +0000)]
target/ppc: convert vspltis[bhw] to use vector operations
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20190215100058.20015-4-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Mark Cave-Ayland [Fri, 15 Feb 2019 10:00:43 +0000 (10:00 +0000)]
target/ppc: convert vaddu[b,h,w,d] and vsubu[b,h,w,d] over to use vector operations
Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-3-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Mark Cave-Ayland [Fri, 15 Feb 2019 10:00:42 +0000 (10:00 +0000)]
target/ppc: convert VMX logical instructions to use vector operations
Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20190215100058.20015-2-mark.cave-ayland@ilande.co.uk> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Greg Kurz [Fri, 15 Feb 2019 11:40:41 +0000 (12:40 +0100)]
xics: Drop the KVM ICS class
The KVM ICS class isn't used anymore. Drop it.
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <155023084177.1011724.14693955932559990358.stgit@bahia.lan> Reviewed-by: Cédric Le Goater <clg@kaod.org> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Greg Kurz [Fri, 15 Feb 2019 11:40:35 +0000 (12:40 +0100)]
spapr/irq: Use the "simple" ICS class for KVM
The "simple" ICS class knows how to interract with KVM. Adapt sPAPR to use
it instead of the ICS KVM class.
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <155023083585.1011724.2868047424353921455.stgit@bahia.lan> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Greg Kurz [Fri, 15 Feb 2019 11:40:30 +0000 (12:40 +0100)]
xics: Handle KVM interrupt presentation from "simple" ICS code
We want to use the "simple" ICS type in both KVM and non-KVM setups.
Teach the "simple" ICS how to present interrupts to KVM and adapt
sPAPR accordingly.
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <155023082996.1011724.16237920586343905010.stgit@bahia.lan> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Greg Kurz [Fri, 15 Feb 2019 11:40:24 +0000 (12:40 +0100)]
xics: Handle KVM ICS reset from the "simple" ICS code
The KVM ICS reset handler simply writes the ICS state to KVM. This
doesn't need the overkill parent_reset logic we have today. Also
we want to use the same ICS type for the KVM and non-KVM case with
pseries.
Call icp_set_kvm_state() from the "simple" ICS reset function.
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <155023082407.1011724.1983100830860273401.stgit@bahia.lan> Reviewed-by: Cédric Le Goater <clg@kaod.org> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Greg Kurz [Fri, 15 Feb 2019 11:40:18 +0000 (12:40 +0100)]
xics: Explicitely call KVM ICS methods from the common code
The pre_save(), post_load() and synchronize_state() methods of the
ICSStateClass type are really KVM only things. Make that obvious
by dropping the indirections and directly calling the KVM functions
instead.
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <155023081817.1011724.14078777320394028836.stgit@bahia.lan> Reviewed-by: Cédric Le Goater <clg@kaod.org> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Greg Kurz [Fri, 15 Feb 2019 11:40:12 +0000 (12:40 +0100)]
xics: Drop the KVM ICP class
The KVM ICP class isn't used anymore. Drop it.
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <155023081228.1011724.12474992370439652538.stgit@bahia.lan> Reviewed-by: Cédric Le Goater <clg@kaod.org> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Greg Kurz [Fri, 15 Feb 2019 11:40:06 +0000 (12:40 +0100)]
spapr/irq: Use the base ICP class for KVM
The base ICP class knows how to interact with KVM. Adapt sPAPR to use it
instead of the ICP KVM class.
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <155023080638.1011724.792095453419098948.stgit@bahia.lan> Reviewed-by: Cédric Le Goater <clg@kaod.org> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Greg Kurz [Fri, 15 Feb 2019 11:40:00 +0000 (12:40 +0100)]
xics: Handle KVM ICP realize from the common code
The realization of KVM ICP currently follows the parent_realize logic,
which is a bit overkill here. Also we want to get rid of the KVM ICP
class. Explicitely call icp_kvm_realize() from the base ICP realize
function.
Note that ICPStateClass::parent_realize is retained because powernv
needs it.
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <155023080049.1011724.15423463482790260696.stgit@bahia.lan> Reviewed-by: Cédric Le Goater <clg@kaod.org> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Greg Kurz [Fri, 15 Feb 2019 11:39:54 +0000 (12:39 +0100)]
xics: Handle KVM ICP reset from the common code
The KVM ICP reset handler simply writes the ICP state to KVM. This
doesn't need the overkill parent_reset logic we have today. Call
icp_set_kvm_state() from the base ICP reset function instead.
Since there are no other users for ICPStateClass::parent_reset, and
it isn't currently expected to change, drop it as well.
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <155023079461.1011724.12644984391500635645.stgit@bahia.lan> Reviewed-by: Cédric Le Goater <clg@kaod.org> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Greg Kurz [Fri, 15 Feb 2019 11:39:48 +0000 (12:39 +0100)]
xics: Explicitely call KVM ICP methods from the common code
The pre_save(), post_load() and synchronize_state() methods of the
ICPStateClass type are really KVM only things. Make that obvious
by dropping the indirections and directly calling the KVM functions
instead.
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <155023078871.1011724.3083923389814185598.stgit@bahia.lan> Reviewed-by: Cédric Le Goater <clg@kaod.org> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Cédric Le Goater [Wed, 13 Feb 2019 21:07:56 +0000 (22:07 +0100)]
spapr/irq: remove the XICS offset adjustment
Now that we have changed the XICS and the XIVE interrupt backend to
have different size for their IRQ number space, we do not need to
align their source numbers anymore. Remove the offset adjustment and
wire the dual 'qirq' handler to the 'qirq' handler of the current
interrupt mode in use.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20190213210756.27032-3-clg@kaod.org> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Cédric Le Goater [Wed, 13 Feb 2019 21:07:55 +0000 (22:07 +0100)]
spapr/irq: add an 'nr_irq' parameter to initialize the backend.
When using the 'dual' interrupt mode, the source numbers of both sPAPR
IRQ backends are aligned to share a common IRQ number space and to use
a similar mapping of the machine qemu_irq array which is indexed by
the source number.
The XICS IRQ number range initially being [ 0x1000 - 0x2000 ], this
requires to change the XICS ICSState offset to 0 and to provision for
an extra 4K of source numbers and qemu_irqs which will never be used
by the machine when running under the XICS interrupt mode. This is not
an optimal solution.
Change the init() method to allocate an IRQ number space of the
expected size for the XICS sPAPR IRQ backend. It breaks the interrupt
signaling when under the 'dual' mode because source numbers have
unexpected values but next patch will fix that.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20190213210756.27032-2-clg@kaod.org> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
This causes the allocated buffer 'int_buf' to be smaller than expected
and we eventually overwrite some of glibc's control structures (see
"chunk" in https://sourceware.org/glibc/wiki/MallocInternals)
The following error is seen while trying to free int_buf:
"free(): invalid next size (fast)"
Fixes: a324d6f166 "spapr: Support ibm,dynamic-memory-v2 property" Signed-off-by: Fabiano Rosas <farosas@linux.ibm.com>
Message-Id: <20190213172926.21740-1-farosas@linux.ibm.com> Reviewed-by: Greg Kurz <groug@kaod.org> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Benjamin Herrenschmidt [Mon, 28 Jan 2019 09:46:18 +0000 (10:46 +0100)]
target/ppc: Disable ISA 2.06 PM instructions on POWER9
The ISA 2.06/2.07 Power Management instructions (doze, nap & rvwinkle)
don't exist on POWER9, don't enable them.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20190128094625.4428-13-clg@kaod.org> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
projects / users / dwmw2 / qemu.git / log