summaryrefslogtreecommitdiffstats
path: root/arch/x86/kvm/x86.c
diff options
context:
space:
mode:
authorTakashi Iwai <tiwai@suse.de>2014-08-04 18:28:21 +0200
committerTakashi Iwai <tiwai@suse.de>2014-08-04 18:28:21 +0200
commite9e3bdffe38b256f920eacc455d30deba5e97655 (patch)
treecfa546f6370c117e012c12605da22ce189542127 /arch/x86/kvm/x86.c
parentf42bb22243d2ae264d721b055f836059fe35321f (diff)
parentae34a78c430c37c06404f032fb04e51315204281 (diff)
downloadop-kernel-dev-e9e3bdffe38b256f920eacc455d30deba5e97655.zip
op-kernel-dev-e9e3bdffe38b256f920eacc455d30deba5e97655.tar.gz
Merge tag 'asoc-v3.17' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus
ASoC: Updates for v3.17 This has been a pretty exciting release in terms of the framework, we've finally got support for multiple CODECs attached to a single DAI link which has been something there's been interest in as long as I've been working on ASoC. A big thanks to Benoit and Misael for their work on this. Otherwise it's been a fairly standard release for development, including more componentisation work from Lars-Peter and a good selection of both CODEC and CPU drivers. - Support for multiple CODECs attached to a single DAI, enabling systems with for example multiple DAC/speaker drivers on a single link, contributed by Benoit Cousson based on work from Misael Lopez Cruz. - Support for byte controls larger than 256 bytes based on the use of TLVs contributed by Omair Mohammed Abdullah. - More componentisation work from Lars-Peter Clausen. - The remainder of the conversions of CODEC drivers to params_width() - Drivers for Cirrus Logic CS4265, Freescale i.MX ASRC blocks, Realtek RT286 and RT5670, Rockchip RK3xxx I2S controllers and Texas Instruments TAS2552. - Lots of updates and fixes, especially to the DaVinci, Intel, Freescale, Realtek, and rcar drivers.
Diffstat (limited to 'arch/x86/kvm/x86.c')
-rw-r--r--arch/x86/kvm/x86.c12
1 files changed, 12 insertions, 0 deletions
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index f644933..ef432f8 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -5887,6 +5887,18 @@ static int inject_pending_event(struct kvm_vcpu *vcpu, bool req_int_win)
kvm_x86_ops->set_nmi(vcpu);
}
} else if (kvm_cpu_has_injectable_intr(vcpu)) {
+ /*
+ * Because interrupts can be injected asynchronously, we are
+ * calling check_nested_events again here to avoid a race condition.
+ * See https://lkml.org/lkml/2014/7/2/60 for discussion about this
+ * proposal and current concerns. Perhaps we should be setting
+ * KVM_REQ_EVENT only on certain events and not unconditionally?
+ */
+ if (is_guest_mode(vcpu) && kvm_x86_ops->check_nested_events) {
+ r = kvm_x86_ops->check_nested_events(vcpu, req_int_win);
+ if (r != 0)
+ return r;
+ }
if (kvm_x86_ops->interrupt_allowed(vcpu)) {
kvm_queue_interrupt(vcpu, kvm_cpu_get_interrupt(vcpu),
false);
OpenPOWER on IntegriCloud