diff options
| author | Peter Maydell <peter.maydell@linaro.org> | 2015-11-05 14:31:24 +0000 |
|---|---|---|
| committer | Peter Maydell <peter.maydell@linaro.org> | 2015-11-05 14:31:24 +0000 |
| commit | 496c1b19facc7b850fa0c09899fcc07a0702fbfd (patch) | |
| tree | 123bc6e3e5fb4ab3c8c0d5ba900ff3b2da4a0e57 /hw/i386/kvm/clock.c | |
| parent | 8835b9df3bddf332c883c861d6a1defc12c4ebe9 (diff) | |
| parent | e01dd3da5cf9aa90ae844d3b86c2c2762066edac (diff) | |
Merge remote-tracking branch 'remotes/bonzini/tags/for-upstream' into staging
* Guest ABI fixes for PC machines (hw_version)
* Fixes for recent Perl
* John Snow's configure fixes
* file-backed RAM improvements (Igor, Pavel)
* -Werror=clobbered fixes (Stefan)
* Kill -d ioport
* Fix qemu-system-s390x
* Performance improvement for kvmclock migration
# gpg: Signature made Thu 05 Nov 2015 13:42:27 GMT using RSA key ID 78C7AE83
# gpg: Good signature from "Paolo Bonzini <bonzini@gnu.org>"
# gpg: aka "Paolo Bonzini <pbonzini@redhat.com>"
* remotes/bonzini/tags/for-upstream:
iscsi: Translate scsi sense into error code
Revert "Introduce cpu_clean_all_dirty"
kvmclock: add a new function to update env->tsc.
configure: disable FORTIFY_SOURCE under clang
backends/hostmem-file: Allow to specify full pathname for backing file
configure: disallow ccache during compile tests
cpu-exec: Fix compiler warning (-Werror=clobbered)
memory: call begin, log_start and commit when registering a new listener
megasas: Use qemu_hw_version() instead of QEMU_VERSION
osdep: Rename qemu_{get, set}_version() to qemu_{, set_}hw_version()
pc: Set hw_version on all machine classes
qemu-log: remove -d ioport
ioport: do not use CPU_LOG_IOPORT
target-i386: fix pcmpxstrx equal-ordered (strstr) mode
scripts/text2pod.pl: Escape left brace
file_ram_alloc: propagate error to caller instead of terminating QEMU
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Diffstat (limited to 'hw/i386/kvm/clock.c')
| -rw-r--r-- | hw/i386/kvm/clock.c | 18 |
1 files changed, 2 insertions, 16 deletions
diff --git a/hw/i386/kvm/clock.c b/hw/i386/kvm/clock.c index efdf165848..0593a3f1f5 100644 --- a/hw/i386/kvm/clock.c +++ b/hw/i386/kvm/clock.c @@ -17,7 +17,7 @@ #include "qemu/host-utils.h" #include "sysemu/sysemu.h" #include "sysemu/kvm.h" -#include "sysemu/cpus.h" +#include "kvm_i386.h" #include "hw/sysbus.h" #include "hw/kvm/clock.h" @@ -125,21 +125,7 @@ static void kvmclock_vm_state_change(void *opaque, int running, return; } - cpu_synchronize_all_states(); - /* In theory, the cpu_synchronize_all_states() call above wouldn't - * affect the rest of the code, as the VCPU state inside CPUState - * is supposed to always match the VCPU state on the kernel side. - * - * In practice, calling cpu_synchronize_state() too soon will load the - * kernel-side APIC state into X86CPU.apic_state too early, APIC state - * won't be reloaded later because CPUState.vcpu_dirty==true, and - * outdated APIC state may be migrated to another host. - * - * The real fix would be to make sure outdated APIC state is read - * from the kernel again when necessary. While this is not fixed, we - * need the cpu_clean_all_dirty() call below. - */ - cpu_clean_all_dirty(); + kvm_synchronize_all_tsc(); ret = kvm_vm_ioctl(kvm_state, KVM_GET_CLOCK, &data); if (ret < 0) { |