kvm: introduce kvm_kernel_irqchip_* functions

The KVMState struct is opaque, so provide accessors for the fields
that will be moved from current_machine to the accelerator. For now
they just forwar

kvm: introduce kvm_kernel_irqchip_* functions

The KVMState struct is opaque, so provide accessors for the fields
that will be moved from current_machine to the accelerator. For now
they just forward to the machine object, but this will change.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

show more ...

kvm: convert "-machine kvm_shadow_mem" to an accelerator property

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

tcg: add "-accel tcg,tb-size" and deprecate "-tb-size"

-tb-size fits nicely in the new framework for accelerator-specific options. It
is a very niche option, so insta-deprecate it.

Signed-off-by:

tcg: add "-accel tcg,tb-size" and deprecate "-tb-size"

-tb-size fits nicely in the new framework for accelerator-specific options. It
is a very niche option, so insta-deprecate it.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

show more ...

tcg: convert "-accel threads" to a QOM property

Replace the ad-hoc qemu_tcg_configure with generic code invoking QOM
property getters and setters. More properties (and thus more valid
-accel subopt

tcg: convert "-accel threads" to a QOM property

Replace the ad-hoc qemu_tcg_configure with generic code invoking QOM
property getters and setters. More properties (and thus more valid
-accel suboptions) will be added in the next patches, which will move
accelerator-related "-machine" options to accelerators.

Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

show more ...

accel: pass object to accel_init_machine

We will have to set QOM properties before accel_init_machine, based on the
options provided to -accel. Construct the object outside it so that it
will be po

accel: pass object to accel_init_machine

We will have to set QOM properties before accel_init_machine, based on the
options provided to -accel. Construct the object outside it so that it
will be possible to iterate on properties between object_new_with_class
and accel_init_machine.

Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

show more ...

qom: add object_new_with_class

Similar to CPU and machine classes, "-accel" class names are mangled,
so we have to first get a class via accel_find and then instantiate it.
Provide a new function to

qom: add object_new_with_class

Similar to CPU and machine classes, "-accel" class names are mangled,
so we have to first get a class via accel_find and then instantiate it.
Provide a new function to instantiate a class without going through
object_class_get_name, and use it for CPUs and machines already.

Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

show more ...

accel: compile accel/accel.c just once

Now that accel/accel.c does not use CONFIG_TCG or CONFIG_KVM anymore,
it need not be compiled once for every softmmu target.

Signed-off-by: Paolo Bonzini <pbo

accel: compile accel/accel.c just once

Now that accel/accel.c does not use CONFIG_TCG or CONFIG_KVM anymore,
it need not be compiled once for every softmmu target.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

show more ...

vl: merge -accel processing into configure_accelerators

The next step is to move the parsing of "-machine accel=..." into vl.c,
unifying it with the configure_accelerators() function that has just
b

vl: merge -accel processing into configure_accelerators

The next step is to move the parsing of "-machine accel=..." into vl.c,
unifying it with the configure_accelerators() function that has just
been introduced. This way, we will be able to desugar it into multiple
"-accel" options, without polluting accel/accel.c.

The CONFIG_TCG and CONFIG_KVM symbols are not available in vl.c, but
we can use accel_find instead to find their value at runtime. Once we
know that the binary has one of TCG or KVM, the default accelerator
can be expressed simply as "tcg:kvm", because TCG never fails to initialize.

Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

show more ...

tcg: move qemu_tcg_configure to accel/tcg/tcg-all.c

Move everything related to mttcg_enabled in accel/tcg/tcg-all.c,
which will make even more sense when "thread" becomes a QOM property.

For now, i

tcg: move qemu_tcg_configure to accel/tcg/tcg-all.c

Move everything related to mttcg_enabled in accel/tcg/tcg-all.c,
which will make even more sense when "thread" becomes a QOM property.

For now, initializing mttcg_enabled in the instance_init function
prepares for the next patch, which will only invoke qemu_tcg_configure
when the command line includes a -accel option.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

show more ...

kvm: Reallocate dirty_bmap when we change a slot

kvm_set_phys_mem can be called to reallocate a slot by something the
guest does (e.g. writing to PAM and other chipset registers).
This can happen in

kvm: Reallocate dirty_bmap when we change a slot

kvm_set_phys_mem can be called to reallocate a slot by something the
guest does (e.g. writing to PAM and other chipset registers).
This can happen in the middle of a migration, and if we're unlucky
it can now happen between the split 'sync' and 'clear'; the clear
asserts if there's no bmap to clear. Recreate the bmap whenever
we change the slot, keeping the clear path happy.

Typically this is triggered by the guest rebooting during a migrate.

Corresponds to:
https://bugzilla.redhat.com/show_bug.cgi?id=1772774
https://bugzilla.redhat.com/show_bug.cgi?id=1771032

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

show more ...

kvm: Introduce KVM irqchip change notifier

Awareness of an in kernel irqchip is usually local to the machine and its
top-level interrupt controller. However, in a few cases other things need
to kno

kvm: Introduce KVM irqchip change notifier

Awareness of an in kernel irqchip is usually local to the machine and its
top-level interrupt controller. However, in a few cases other things need
to know about it. In particular vfio devices need this in order to
accelerate interrupt delivery.

If interrupt routing is changed, such devices may need to readjust their
connection to the KVM irqchip. pci_bus_fire_intx_routing_notifier() exists
to do just this.

However, for the pseries machine type we have a situation where the routing
remains constant but the top-level irq chip itself is changed. This occurs
because of PAPR feature negotiation which allows the guest to decide
between the older XICS and newer XIVE irq chip models (both of which are
paravirtualized).

To allow devices like vfio to adjust to this change, introduce a new
notifier for the purpose kvm_irqchip_change_notify().

Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Alexey Kardashevskiy <aik@ozlabs.ru>

Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Tested-by: Alex Williamson <alex.williamson@redhat.com>
Reviewed-by: Alex Williamson <alex.williamson@redhat.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Acked-by: Alex Williamson <alex.williamson@redhat.com>

show more ...

Remove unassigned_access CPU hook

All targets have now migrated away from the old unassigned_access
hook to the new do_transaction_failed hook. This means we can remove
the core-code infrastructure

Remove unassigned_access CPU hook

All targets have now migrated away from the old unassigned_access
hook to the new do_transaction_failed hook. This means we can remove
the core-code infrastructure for that hook and the code that calls it.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Message-id: 20191108173732.11816-1-peter.maydell@linaro.org

show more ...

accel/stubs: reduce headers from tcg-stub

We don't need much for these. However I do wonder why these aren't
just null inlines in exec-all.h

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Revi

accel/stubs: reduce headers from tcg-stub

We don't need much for these. However I do wonder why these aren't
just null inlines in exec-all.h

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

show more ...

translator: inject instrumentation from plugins

Signed-off-by: Emilio G. Cota <cota@braap.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@l

translator: inject instrumentation from plugins

Signed-off-by: Emilio G. Cota <cota@braap.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

show more ...

cputlb: ensure _cmmu helper functions follow the naming standard

We document this in docs/devel/load-stores.rst so lets follow it. The
32 bit and 64 bit access functions have historically not includ

cputlb: ensure _cmmu helper functions follow the naming standard

We document this in docs/devel/load-stores.rst so lets follow it. The
32 bit and 64 bit access functions have historically not included the
sign so we leave those as is. We also introduce some signed helpers
which are used for loading immediate values in the translator.

Fixes: 282dffc8
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20191021150910.23216-1-alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

show more ...

translate-all: notify plugin code of tb_flush

Plugins might allocate per-TB data that then they get passed each
time a TB is executed (via the *userdata pointer).

Notify plugin code every time a co

translate-all: notify plugin code of tb_flush

Plugins might allocate per-TB data that then they get passed each
time a TB is executed (via the *userdata pointer).

Notify plugin code every time a code cache flush occurs, so
that plugins can then reclaim the memory of the per-TB data.

Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Emilio G. Cota <cota@braap.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>

show more ...

plugins: implement helpers for resolving hwaddr

We need to keep a local per-cpu copy of the data as other threads may
be running. Currently we can provide insight as to if the access was
IO or not a

plugins: implement helpers for resolving hwaddr

We need to keep a local per-cpu copy of the data as other threads may
be running. Currently we can provide insight as to if the access was
IO or not and give the offset into a given device (usually the main
RAMBlock). We store enough information to get details such as the
MemoryRegion which might be useful in later expansions to the API.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

show more ...

tcg: let plugins instrument virtual memory accesses

To capture all memory accesses we need hook into all the various
helper functions that are involved in memory operations as well as the
injected i

tcg: let plugins instrument virtual memory accesses

To capture all memory accesses we need hook into all the various
helper functions that are involved in memory operations as well as the
injected inline helper calls. A later commit will allow us to resolve
the actual guest HW addresses by replaying the lookup.

Signed-off-by: Emilio G. Cota <cota@braap.org>
[AJB: drop haddr handling, just deal in vaddr]
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

show more ...

atomic_template: add inline trace/plugin helpers

In preparation for plugin support.

Signed-off-by: Emilio G. Cota <cota@braap.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: R

atomic_template: add inline trace/plugin helpers

In preparation for plugin support.

Signed-off-by: Emilio G. Cota <cota@braap.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

show more ...

plugin-gen: add module for TCG-related code

We first inject empty instrumentation from translator_loop.
After translation, we go through the plugins to see what
they want to register for, filling in

plugin-gen: add module for TCG-related code

We first inject empty instrumentation from translator_loop.
After translation, we go through the plugins to see what
they want to register for, filling in the empty instrumentation.
If if turns out that some instrumentation remains unused, we
remove it.

This approach supports the following features:

- Inlining TCG code for simple operations. Note that we do not
export TCG ops to plugins. Instead, we give them a C API to
insert inlined ops. So far we only support adding an immediate
to a u64, e.g. to count events.

- "Direct" callbacks. These are callbacks that do not go via
a helper. Instead, the helper is defined at run-time, so that
the plugin code is directly called from TCG. This makes direct
callbacks as efficient as possible; they are therefore used
for very frequent events, e.g. memory callbacks.

- Passing the host address to memory callbacks. Most of this
is implemented in a later patch though.

- Instrumentation of memory accesses performed from helpers.
See the corresponding comment, as well as a later patch.

Signed-off-by: Emilio G. Cota <cota@braap.org>
[AJB: add alloc_tcg_plugin_context, use glib, rm hwaddr]
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

show more ...

cputlb: introduce get_page_addr_code_hostp

This will be used by plugins to get the host address
of instructions.

Signed-off-by: Emilio G. Cota <cota@braap.org>
Signed-off-by: Alex Bennée <alex.benn

cputlb: introduce get_page_addr_code_hostp

This will be used by plugins to get the host address
of instructions.

Signed-off-by: Emilio G. Cota <cota@braap.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

show more ...

translate-all: use cpu_in_exclusive_work_context() in tb_flush

tb_flush will be called by the plugin module from a safe
work environment. Prepare for that.

Suggested-by: Alex Bennée <alex.bennee@li

translate-all: use cpu_in_exclusive_work_context() in tb_flush

tb_flush will be called by the plugin module from a safe
work environment. Prepare for that.

Suggested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Emilio G. Cota <cota@braap.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

show more ...

cpu: introduce cpu_in_exclusive_context()

Suggested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Emilio G. Cota <cota@braap.org>
[AJB: mo

cpu: introduce cpu_in_exclusive_context()

Suggested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Emilio G. Cota <cota@braap.org>
[AJB: moved inside start/end_exclusive fns + cleanup]
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

show more ...

trace: add mmu_index to mem_info

We are going to re-use mem_info later for plugins and will need to
track the mmu_idx for softmmu code.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>

translate-all: Remove tb_alloc

Since 2ac01d6dafab, this function does only two things: assert a
lock is held, and call tcg_tb_alloc. It is used exactly once,
and its user has already done the asser

translate-all: Remove tb_alloc

Since 2ac01d6dafab, this function does only two things: assert a
lock is held, and call tcg_tb_alloc. It is used exactly once,
and its user has already done the assert.

Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Clement Deschamps <clement.deschamps@greensocs.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

show more ...