At the moment the Arm implementations of kvm_arch_{get,put}_registers()
don't support having QEMU change the values of system registers
(aka coprocessor registers for AArch32). This is because although
kvm_arch_get_registers() calls write_list_to_cpustate() to
update the CPU state struct fields (so QEMU code can read the
values in the usual way), kvm_arch_put_registers() does not
call write_cpustate_to_list(), meaning that any changes to
the CPU state struct fields will not be passed back to KVM.
The rationale for this design is documented in a comment in the
AArch32 kvm_arch_put_registers() -- writing the values in the
cpregs list into the CPU state struct is "lossy" because the
write of a register might not succeed, and so if we blindly
copy the CPU state values back again we will incorrectly
change register values for the guest. The assumption was that
no QEMU code would need to write to the registers.
However, when we implemented debug support for KVM guests, we
broke that assumption: the code to handle "set the guest up
to take a breakpoint exception" does so by updating various
guest registers including ESR_EL1.
Support this by making kvm_arch_put_registers() synchronize
CPU state back into the list. We sync only those registers
where the initial write succeeds, which should be sufficient.
This commit is the same as commit 823e1b3818 which we
had to revert in commit 942f99c825, except that the bug
which was preventing EDK2 guest firmware running has been fixed:
kvm_arm_reset_vcpu() now calls write_list_to_cpustate().
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Tested-by: Eric Auger <eric.auger@redhat.com>
This reverts commit 823e1b3818,
which introduces a regression running EDK2 guest firmware
under KVM:
error: kvm run failed Function not implemented
PC=000000013f5a6208 X00=00000000404003c4 X01=000000000000003a
X02=0000000000000000 X03=00000000404003c4 X04=0000000000000000
X05=0000000096000046 X06=000000013d2ef270 X07=000000013e3d1710
X08=09010755ffaf8ba8 X09=ffaf8b9cfeeb5468 X10=feeb546409010756
X11=09010757ffaf8b90 X12=feeb50680903068b X13=090306a1ffaf8bc0
X14=0000000000000000 X15=0000000000000000 X16=000000013f872da0
X17=00000000ffffa6ab X18=0000000000000000 X19=000000013f5a92d0
X20=000000013f5a7a78 X21=000000000000003a X22=000000013f5a7ab2
X23=000000013f5a92e8 X24=000000013f631090 X25=0000000000000010
X26=0000000000000100 X27=000000013f89501b X28=000000013e3d14e0
X29=000000013e3d12a0 X30=000000013f5a2518 SP=000000013b7be0b0
PSTATE=404003c4 -Z-- EL1t
with
[ 3507.926571] kvm [35042]: load/store instruction decoding not implemented
in the host dmesg.
Revert the change for the moment until we can investigate the
cause of the regression.
Reported-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Instead of gating the A32/T32 FP16 conversion instructions on
the ARM_FEATURE_VFP_FP16 flag, switch to our new approach of
looking at ID register bits. In this case MVFR1 fields FPHP
and SIMDHP indicate the presence of these insns.
This change doesn't alter behaviour for any of our CPUs.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20190222170936.13268-2-peter.maydell@linaro.org
At the moment the Arm implementations of kvm_arch_{get,put}_registers()
don't support having QEMU change the values of system registers
(aka coprocessor registers for AArch32). This is because although
kvm_arch_get_registers() calls write_list_to_cpustate() to
update the CPU state struct fields (so QEMU code can read the
values in the usual way), kvm_arch_put_registers() does not
call write_cpustate_to_list(), meaning that any changes to
the CPU state struct fields will not be passed back to KVM.
The rationale for this design is documented in a comment in the
AArch32 kvm_arch_put_registers() -- writing the values in the
cpregs list into the CPU state struct is "lossy" because the
write of a register might not succeed, and so if we blindly
copy the CPU state values back again we will incorrectly
change register values for the guest. The assumption was that
no QEMU code would need to write to the registers.
However, when we implemented debug support for KVM guests, we
broke that assumption: the code to handle "set the guest up
to take a breakpoint exception" does so by updating various
guest registers including ESR_EL1.
Support this by making kvm_arch_put_registers() synchronize
CPU state back into the list. We sync only those registers
where the initial write succeeds, which should be sufficient.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Tested-by: Alex Bennée <alex.bennee@linaro.org>
Tested-by: Dongjiu Geng <gengdongjiu@huawei.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20181113180154.17903-5-richard.henderson@linaro.org
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Assert that the value to be written is the correct size.
No change in functionality here, just mirroring the same
function from kvm64.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20181113180154.17903-4-richard.henderson@linaro.org
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Hyp mode is an exception to the general rule that each AArch32
mode has its own r13, r14 and SPSR -- it has a banked r13 and
SPSR but shares its r14 with User and System mode. We were
incorrectly implementing it as banked, which meant that on
entry to Hyp mode r14 was 0 rather than the USR/SYS r14.
We provide a new function r14_bank_number() which is like
the existing bank_number() but provides the index into
env->banked_r14[]; bank_number() provides the index to use
for env->banked_r13[] and env->banked_cpsr[].
All the points in the code that were using bank_number()
to index into env->banked_r14[] are updated for consintency:
* switch_mode() -- this is the only place where we fix
an actual bug
* aarch64_sync_32_to_64() and aarch64_sync_64_to_32():
no behavioural change as we already special-cased Hyp R14
* kvm32.c: no behavioural change since the guest can't ever
be in Hyp mode, but conceptually the right thing to do
* msr_banked()/mrs_banked(): we can never get to the case
that accesses banked_r14[] with tgtmode == ARM_CPU_MODE_HYP,
so no behavioural change
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20181109173553.22341-2-peter.maydell@linaro.org
This patch extends the qemu-kvm state sync logic with support for
KVM_GET/SET_VCPU_EVENTS, giving access to yet missing SError exception.
And also it can support the exception state migration.
The SError exception states include SError pending state and ESR value,
the kvm_put/get_vcpu_events() will be called when set or get system
registers. When do migration, if source machine has SError pending,
QEMU will do this migration regardless whether the target machine supports
to specify guest ESR value, because if target machine does not support that,
it can also inject the SError with zero ESR value.
Signed-off-by: Dongjiu Geng <gengdongjiu@huawei.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 1538067351-23931-3-git-send-email-gengdongjiu@huawei.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
KVM implies V7VE, which implies ARM_DIV and THUMB_DIV. The conditional
detection here is therefore unnecessary. Because V7VE is already
unconditionally specified for all KVM hosts, ARM_DIV and THUMB_DIV are
already indirectly specified and do not need to be included here at all.
Signed-off-by: Aaron Lindsay <alindsay@codeaurora.org>
Message-id: 1529699547-17044-6-git-send-email-alindsay@codeaurora.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Currently we query the host CPU features in the class init function
for the TYPE_ARM_HOST_CPU class, so that we can later copy them
from the class object into the instance object in the object
instance init function. This is awkward for implementing "-cpu max",
which should work like "-cpu host" for KVM but like "cpu with all
implemented features" for TCG.
Move the place where we store the information about the host CPU from
a class object to static variables in kvm.c, and then in the instance
init function call a new kvm_arm_set_cpu_features_from_host()
function which will query the host kernel if necessary and then
fill in the CPU instance fields.
This allows us to drop the special class struct and class init
function for TYPE_ARM_HOST_CPU entirely.
We can't delay the probe until realize, because the ARM
instance_post_init hook needs to look at the feature bits we
set, so we need to do it in the initfn. This is safe because
the probing doesn't affect the actual VM state (it creates a
separate scratch VM to do its testing), but the probe might fail.
Because we can't report errors in retrieving the host features
in the initfn, we check this belatedly in the realize function
(the intervening code will be able to cope with the relevant
fields in the CPU structure being zero).
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20180308130626.12393-2-peter.maydell@linaro.org
Helpers that return a pointer into env->vfp.regs so that we isolate
the logic of how to index the regs array for different cpu modes.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20180119045438.28582-7-richard.henderson@linaro.org
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
If a KVM PMU init or set-irq attr call fails we just silently stop
the PMU DT node generation. The only way they could fail, though,
is if the attr's respective KVM has-attr call fails. But that should
never happen if KVM advertises the PMU capability, because both
attrs have been available since the capability was introduced. Let's
just abort if this should-never-happen stuff does happen, because,
if it does, then something is obviously horribly wrong.
Signed-off-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: Christoffer Dall <cdall@linaro.org>
Message-id: 1500471597-2517-5-git-send-email-drjones@redhat.com
[PMM: change kvm32.c kvm_arm_pmu_init() to the new API too]
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
When adding a PMU with a userspace irqchip we skip the set-irq
stage of device creation. Split the 'create' function into two
functions 'init' and 'set-irq' so they may be called separately.
Signed-off-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: Christoffer Dall <cdall@linaro.org>
Message-id: 1500471597-2517-3-git-send-email-drjones@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
We've currently got 18 architectures in QEMU, and thus 18 target-xxx
folders in the root folder of the QEMU source tree. More architectures
(e.g. RISC-V, AVR) are likely to be included soon, too, so the main
folder of the QEMU sources slowly gets quite overcrowded with the
target-xxx folders.
To disburden the main folder a little bit, let's move the target-xxx
folders into a dedicated target/ folder, so that target-xxx/ simply
becomes target/xxx/ instead.
Acked-by: Laurent Vivier <laurent@vivier.eu> [m68k part]
Acked-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de> [tricore part]
Acked-by: Michael Walle <michael@walle.cc> [lm32 part]
Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com> [s390x part]
Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com> [s390x part]
Acked-by: Eduardo Habkost <ehabkost@redhat.com> [i386 part]
Acked-by: Artyom Tarasenko <atar4qemu@gmail.com> [sparc part]
Acked-by: Richard Henderson <rth@twiddle.net> [alpha part]
Acked-by: Max Filippov <jcmvbkbc@gmail.com> [xtensa part]
Reviewed-by: David Gibson <david@gibson.dropbear.id.au> [ppc part]
Acked-by: Edgar E. Iglesias <edgar.iglesias@xilinx.com> [crisµblaze part]
Acked-by: Guan Xuetao <gxt@mprc.pku.edu.cn> [unicore32 part]
Signed-off-by: Thomas Huth <thuth@redhat.com>
