| Message ID | 1404142497-6430-5-git-send-email-mihai.caraman@freescale.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On 30.06.14 17:34, Mihai Caraman wrote: > Add KVM Book3E AltiVec support. KVM Book3E FPU support gracefully reuse host > infrastructure so follow the same approach for AltiVec. > > Signed-off-by: Mihai Caraman <mihai.caraman@freescale.com> Same comment here - I fail to see how we refetch Altivec state after a context switch. Alex -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
> -----Original Message----- > From: Alexander Graf [mailto:agraf@suse.de] > Sent: Thursday, July 03, 2014 3:32 PM > To: Caraman Mihai Claudiu-B02008; kvm-ppc@vger.kernel.org > Cc: kvm@vger.kernel.org; linuxppc-dev@lists.ozlabs.org > Subject: Re: [PATCH 4/6 v2] KVM: PPC: Book3E: Add AltiVec support > > > On 30.06.14 17:34, Mihai Caraman wrote: > > Add KVM Book3E AltiVec support. KVM Book3E FPU support gracefully reuse > host > > infrastructure so follow the same approach for AltiVec. > > > > Signed-off-by: Mihai Caraman <mihai.caraman@freescale.com> > > Same comment here - I fail to see how we refetch Altivec state after a > context switch. See previous comment. I also run my usual Altivec stress test consisting in a guest and host process running affine to a physical core an competing for the same unit's resources using different data sets. -Mike -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, 2014-06-30 at 18:34 +0300, Mihai Caraman wrote: > Add KVM Book3E AltiVec support. KVM Book3E FPU support gracefully reuse host > infrastructure so follow the same approach for AltiVec. > > Signed-off-by: Mihai Caraman <mihai.caraman@freescale.com> > --- > v2: > - integrate Paul's FP/VMX/VSX changes > > arch/powerpc/kvm/booke.c | 67 ++++++++++++++++++++++++++++++++++++++++++++++-- > 1 file changed, 65 insertions(+), 2 deletions(-) I had to apply the whole patchset to get proper context for reviewing this, and found some nits: > case BOOKE_INTERRUPT_SPE_FP_DATA_ALTIVEC_ASSIST: > if (kvmppc_supports_spe() || kvmppc_supports_altivec()) { > kvmppc_booke_queue_irqprio(vcpu, > BOOKE_IRQPRIO_SPE_FP_DATA_ALTIVEC_ASSIST); > r = RESUME_GUEST; > } else { > /* > * These really should never happen without CONFIG_SPE, > * as we should never enable the real MSR[SPE] in the > * guest. > */ Besides the comment not being updated for Altivec, it's not true on HV, where the guest can enable MSR[VEC] all by itself. For HV, the reason we shouldn't be able to get here is that we disable KVM on e6500 if CONFIG_ALTIVEC is not enabled, and no other HV core supports either SPE or Altivec. > pr_crit("%s: unexpected SPE interrupt %u at %08lx\n", > __func__, exit_nr, vcpu->arch.pc); Error string will say SPE regardless of what sort of chip you're on. Given that this is explicitly on the "no support for Altivec or SPE" path, "SPE/Altivec" phrasing seems appropriate. Of course we have bigger problems than that if we ever reach this code. :-) -Scott -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c index 4cc9b26..4ba75f6 100644 --- a/arch/powerpc/kvm/booke.c +++ b/arch/powerpc/kvm/booke.c @@ -100,6 +100,19 @@ static inline bool kvmppc_supports_spe(void) return false; } +/* + * Always returns true if AltiVec unit is present, + * see kvmppc_core_check_processor_compat(). + */ +static inline bool kvmppc_supports_altivec(void) +{ +#ifdef CONFIG_ALTIVEC + if (cpu_has_feature(CPU_FTR_ALTIVEC)) + return true; +#endif + return false; +} + #ifdef CONFIG_SPE void kvmppc_vcpu_disable_spe(struct kvm_vcpu *vcpu) { @@ -178,6 +191,40 @@ static void kvmppc_vcpu_sync_fpu(struct kvm_vcpu *vcpu) #endif } +/* + * Simulate AltiVec unavailable fault to load guest state + * from thread to AltiVec unit. + * It requires to be called with preemption disabled. + */ +static inline void kvmppc_load_guest_altivec(struct kvm_vcpu *vcpu) +{ +#ifdef CONFIG_ALTIVEC + if (kvmppc_supports_altivec()) { + if (!(current->thread.regs->msr & MSR_VEC)) { + enable_kernel_altivec(); + load_vr_state(&vcpu->arch.vr); + current->thread.vr_save_area = &vcpu->arch.vr; + current->thread.regs->msr |= MSR_VEC; + } + } +#endif +} + +/* + * Save guest vcpu AltiVec state into thread. + * It requires to be called with preemption disabled. + */ +static inline void kvmppc_save_guest_altivec(struct kvm_vcpu *vcpu) +{ +#ifdef CONFIG_ALTIVEC + if (kvmppc_supports_altivec()) { + if (current->thread.regs->msr & MSR_VEC) + giveup_altivec(current); + current->thread.vr_save_area = NULL; + } +#endif +} + static void kvmppc_vcpu_sync_debug(struct kvm_vcpu *vcpu) { /* Synchronize guest's desire to get debug interrupts into shadow MSR */ @@ -749,6 +796,17 @@ int kvmppc_vcpu_run(struct kvm_run *kvm_run, struct kvm_vcpu *vcpu) kvmppc_load_guest_fp(vcpu); #endif +#ifdef CONFIG_ALTIVEC + /* Save userspace AltiVec state in stack */ + if (kvmppc_supports_altivec()) + enable_kernel_altivec(); + /* + * Since we can't trap on MSR_VEC in GS-mode, we consider the guest + * as always using the AltiVec. + */ + kvmppc_load_guest_altivec(vcpu); +#endif + /* Switch to guest debug context */ debug = vcpu->arch.shadow_dbg_reg; switch_booke_debug_regs(&debug); @@ -771,6 +829,10 @@ int kvmppc_vcpu_run(struct kvm_run *kvm_run, struct kvm_vcpu *vcpu) kvmppc_save_guest_fp(vcpu); #endif +#ifdef CONFIG_ALTIVEC + kvmppc_save_guest_altivec(vcpu); +#endif + out: vcpu->mode = OUTSIDE_GUEST_MODE; return ret; @@ -1014,7 +1076,7 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu, break; case BOOKE_INTERRUPT_SPE_ALTIVEC_UNAVAIL: { - if (kvmppc_supports_spe()) { + if (kvmppc_supports_spe() || kvmppc_supports_altivec()) { bool enabled = false; #if !defined(CONFIG_KVM_BOOKE_HV) && defined(CONFIG_SPE) @@ -1040,7 +1102,7 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu, } case BOOKE_INTERRUPT_SPE_FP_DATA_ALTIVEC_ASSIST: - if (kvmppc_supports_spe()) { + if (kvmppc_supports_spe() || kvmppc_supports_altivec()) { kvmppc_booke_queue_irqprio(vcpu, BOOKE_IRQPRIO_SPE_FP_DATA_ALTIVEC_ASSIST); r = RESUME_GUEST; @@ -1249,6 +1311,7 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu, /* interrupts now hard-disabled */ kvmppc_fix_ee_before_entry(); kvmppc_load_guest_fp(vcpu); + kvmppc_load_guest_altivec(vcpu); } }
Add KVM Book3E AltiVec support. KVM Book3E FPU support gracefully reuse host infrastructure so follow the same approach for AltiVec. Signed-off-by: Mihai Caraman <mihai.caraman@freescale.com> --- v2: - integrate Paul's FP/VMX/VSX changes arch/powerpc/kvm/booke.c | 67 ++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 65 insertions(+), 2 deletions(-)