Re: [PATCH RESEND v4] powerpc: kvm: fix rare but potential deadlock scene

2013-11-20 Thread Alexander Graf
> Am 20.11.2013 um 03:42 schrieb Liu ping fan : > >> On Tue, Nov 19, 2013 at 6:39 PM, Alexander Graf wrote: >> >>> On 19.11.2013, at 07:12, Liu Ping Fan wrote: >>> >>> Since kvmppc_hv_find_lock_hpte() is called from both virtmode and >>> realmode, so it can trigger the deadlock. >>> >>> Sup

Re: [PATCH RESEND v4] powerpc: kvm: fix rare but potential deadlock scene

2013-11-19 Thread Liu ping fan
On Tue, Nov 19, 2013 at 6:39 PM, Alexander Graf wrote: > > On 19.11.2013, at 07:12, Liu Ping Fan wrote: > >> Since kvmppc_hv_find_lock_hpte() is called from both virtmode and >> realmode, so it can trigger the deadlock. >> >> Suppose the following scene: >> >> Two physical cpuM, cpuN, two VM inst

Re: [PATCH RESEND v4] powerpc: kvm: fix rare but potential deadlock scene

2013-11-19 Thread Alexander Graf
On 19.11.2013, at 07:12, Liu Ping Fan wrote: > Since kvmppc_hv_find_lock_hpte() is called from both virtmode and > realmode, so it can trigger the deadlock. > > Suppose the following scene: > > Two physical cpuM, cpuN, two VM instances A, B, each VM has a group of > vcpus. > > If on cpuM, vcp

[PATCH RESEND v4] powerpc: kvm: fix rare but potential deadlock scene

2013-11-18 Thread Liu Ping Fan
Since kvmppc_hv_find_lock_hpte() is called from both virtmode and realmode, so it can trigger the deadlock. Suppose the following scene: Two physical cpuM, cpuN, two VM instances A, B, each VM has a group of vcpus. If on cpuM, vcpu_A_1 holds bitlock X (HPTE_V_HVLOCK), then is switched out, and o