[PATCH] lapic3: kvm: qemu: fix a possible deadlock situation in save/restore

[PATCH] lapic3: kvm: qemu: fix a possible deadlock situation in save/restore

[He, Qing]

[-- Attachment #1: Type: text/plain, Size: 1369 bytes --]

    kvm: qemu: fix a possible deadlock situation in save/restore
    
    If a VM using userspace irqchip is saved when it's in HLT state,
    and is restored to use the in-kernel irqchip, the HLT state flag
    is also restored. This HLT flag in userspace HLT handling prevents
    kvm kernel driver from being run. And at the same time, the original
    userspace irqchip is now unused thus will not generate any interrupt
    to wake up the HLT state. This patch fixes this deadlock situation
    by dropping the HLT state at restore time
    
    Signed-off-by: Qing He <qing.he-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
    Signed-off-by: Yaozu (Eddie) Dong <eddie.dong-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 qemu/vl.c |    3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)

diff --git a/qemu/vl.c b/qemu/vl.c
index b14233c..7150d48 100644
--- a/qemu/vl.c
+++ b/qemu/vl.c
@@ -5428,6 +5428,9 @@ int cpu_load(QEMUFile *f, void *opaque, int
version_id)
     tlb_flush(env, 1);
 #ifdef USE_KVM
     if (kvm_allowed) {
+        /* when in-kernel irqchip is used, HF_HALTED_MASK causes
deadlock
+           because no userspace IRQs will ever clear this flag */
+        env->hflags &= ~HF_HALTED_MASK;
         for (i = 0; i < NR_IRQ_WORDS ; i++) {
             qemu_get_be32s(f, &env->kvm_interrupt_bitmap[i]);
         }


[-- Attachment #2: kvm-no-hlt-on-restore.patch --]
[-- Type: application/octet-stream, Size: 1276 bytes --]

    kvm: qemu: fix a possible deadlock situation in save/restore
    
    If a VM using userspace irqchip is saved when it's in HLT state,
    and is restored to use the in-kernel irqchip, the HLT state flag
    is also restored. This HLT flag in userspace HLT handling prevents
    kvm kernel driver from being run. And at the same time, the original
    userspace irqchip is now unused thus will not generate any interrupt
    to wake up the HLT state. This patch fixes this deadlock situation
    by dropping the HLT state at restore time
    
    Signed-off-by: Qing He <qing.he@intel.com>
    Signed-off-by: Yaozu (Eddie) Dong <eddie.dong@intel.com>
---
 qemu/vl.c |    3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)

diff --git a/qemu/vl.c b/qemu/vl.c
index b14233c..7150d48 100644
--- a/qemu/vl.c
+++ b/qemu/vl.c
@@ -5428,6 +5428,9 @@ int cpu_load(QEMUFile *f, void *opaque, int version_id)
     tlb_flush(env, 1);
 #ifdef USE_KVM
     if (kvm_allowed) {
+        /* when in-kernel irqchip is used, HF_HALTED_MASK causes deadlock
+           because no userspace IRQs will ever clear this flag */
+        env->hflags &= ~HF_HALTED_MASK;
         for (i = 0; i < NR_IRQ_WORDS ; i++) {
             qemu_get_be32s(f, &env->kvm_interrupt_bitmap[i]);
         }


[-- Attachment #3: Type: text/plain, Size: 315 bytes --]

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/

[-- Attachment #4: Type: text/plain, Size: 186 bytes --]

_______________________________________________
kvm-devel mailing list
kvm-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
https://lists.sourceforge.net/lists/listinfo/kvm-devel

Re: [PATCH] lapic3: kvm: qemu: fix a possible deadlock situation in save/restore

[Avi Kivity]

He, Qing wrote:
>     kvm: qemu: fix a possible deadlock situation in save/restore
>     
>     If a VM using userspace irqchip is saved when it's in HLT state,
>     and is restored to use the in-kernel irqchip, the HLT state flag
>     is also restored. This HLT flag in userspace HLT handling prevents
>     kvm kernel driver from being run. And at the same time, the original
>     userspace irqchip is now unused thus will not generate any interrupt
>     to wake up the HLT state. This patch fixes this deadlock situation
>     by dropping the HLT state at restore time
>     
>   

Applied, thanks.

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/