> -----Original Message----- > From: kvm-ppc-owner@xxxxxxxxxxxxxxx > [mailto:kvm-ppc-owner@xxxxxxxxxxxxxxx] On Behalf Of Hollis Blanchard > Sent: Friday, December 12, 2008 5:02 AM > To: Liu Yu-B13201 > Cc: kvm-ppc@xxxxxxxxxxxxxxx > Subject: Re: [PATCH 03/11] Use arch.pid as shadow pid for E500 > > On Thu, 2008-12-11 at 17:11 +0800, Liu Yu wrote: > > Signed-off-by: Liu Yu <yu.liu@xxxxxxxxxxxxx> > > --- > > arch/powerpc/kernel/asm-offsets.c | 8 +++++++- > > 1 files changed, 7 insertions(+), 1 deletions(-) > > > > diff --git a/arch/powerpc/kernel/asm-offsets.c > b/arch/powerpc/kernel/asm-offsets.c > > index 544a27f..79fe94c 100644 > > --- a/arch/powerpc/kernel/asm-offsets.c > > +++ b/arch/powerpc/kernel/asm-offsets.c > > @@ -355,6 +355,13 @@ int main(void) > > DEFINE(PTE_SIZE, sizeof(pte_t)); > > > > #ifdef CONFIG_KVM > > +#ifdef CONFIG_44x > > + DEFINE(VCPU_SHADOW_PID, offsetof(struct kvm_vcpu, > arch.shadow_pid)); > > +#endif > > +#ifdef CONFIG_FSL_BOOKE > > + DEFINE(VCPU_SHADOW_PID, offsetof(struct kvm_vcpu, arch.pid)); > > +#endif > > + > > DEFINE(VCPU_HOST_STACK, offsetof(struct kvm_vcpu, > arch.host_stack)); > > DEFINE(VCPU_HOST_PID, offsetof(struct kvm_vcpu, arch.host_pid)); > > DEFINE(VCPU_GPRS, offsetof(struct kvm_vcpu, arch.gpr)); > > @@ -368,7 +375,6 @@ int main(void) > > DEFINE(VCPU_SPRG5, offsetof(struct kvm_vcpu, arch.sprg5)); > > DEFINE(VCPU_SPRG6, offsetof(struct kvm_vcpu, arch.sprg6)); > > DEFINE(VCPU_SPRG7, offsetof(struct kvm_vcpu, arch.sprg7)); > > - DEFINE(VCPU_SHADOW_PID, offsetof(struct kvm_vcpu, > arch.shadow_pid)); > > > > DEFINE(VCPU_LAST_INST, offsetof(struct kvm_vcpu, > arch.last_inst)); > > DEFINE(VCPU_FAULT_DEAR, offsetof(struct kvm_vcpu, > arch.fault_dear)); > > I can commit this, but we got a significant performance boost from > implementing the shadow_pid scheme. Do you think implementing > that would > be difficult? > Not difficult. But I want to compare it with some other ways before apply it. As you know, E500 has a large TLB0, so it has more room to handle multi-VMs on a single core. Current shadow pid trick indeed boost performance in the case of one VM, but for multi-VMs, your pid mapping idea seems better for E500. -- To unsubscribe from this list: send the line "unsubscribe kvm-ppc" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
