| Message ID | 20170825043036.18236-2-npiggin@gmail.com |
|---|---|
| State | Accepted |
| Headers | show |
Nicholas Piggin <npiggin@gmail.com> writes: > POWER9 CPUs have independent MMU contexts per thread, so KVM does not > need to quiesce secondary threads, so the hwthread_req/hwthread_state > protocol does not have to be used. So patch it away on POWER9, and patch > away the branch from the Linux idle wakeup to kvm_start_guest that is > never used. > > Add a warning and error out of kvmppc_grab_hwthread in case it is ever > called on POWER9. > > This avoids a hwsync in the idle wakeup path on POWER9. > > Signed-off-by: Nicholas Piggin <npiggin@gmail.com> > --- > arch/powerpc/include/asm/kvm_book3s_asm.h | 4 ++++ > arch/powerpc/kernel/idle_book3s.S | 35 +++++++++++++++++++++---------- > arch/powerpc/kvm/book3s_hv.c | 14 ++++++++++++- > arch/powerpc/kvm/book3s_hv_rmhandlers.S | 8 +++++++ > 4 files changed, 49 insertions(+), 12 deletions(-) Paulus do you mind acking this and I'll put it in my ppc-kvm topic branch. cheers -- To unsubscribe from this list: send the line "unsubscribe kvm-ppc" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Fri, Aug 25, 2017 at 02:30:33PM +1000, Nicholas Piggin wrote: > POWER9 CPUs have independent MMU contexts per thread, so KVM does not > need to quiesce secondary threads, so the hwthread_req/hwthread_state > protocol does not have to be used. So patch it away on POWER9, and patch > away the branch from the Linux idle wakeup to kvm_start_guest that is > never used. > > Add a warning and error out of kvmppc_grab_hwthread in case it is ever > called on POWER9. > > This avoids a hwsync in the idle wakeup path on POWER9. > > Signed-off-by: Nicholas Piggin <npiggin@gmail.com> Some of the feature sections seem a little unnecessary except for documentation purposes (e.g. the second one added after the kvm_no_guest label), but they won't actually hurt, so: Acked-by: Paul Mackerras <paulus@ozlabs.org> -- To unsubscribe from this list: send the line "unsubscribe kvm-ppc" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Fri, 2017-08-25 at 04:30:33 UTC, Nicholas Piggin wrote: > POWER9 CPUs have independent MMU contexts per thread, so KVM does not > need to quiesce secondary threads, so the hwthread_req/hwthread_state > protocol does not have to be used. So patch it away on POWER9, and patch > away the branch from the Linux idle wakeup to kvm_start_guest that is > never used. > > Add a warning and error out of kvmppc_grab_hwthread in case it is ever > called on POWER9. > > This avoids a hwsync in the idle wakeup path on POWER9. > > Signed-off-by: Nicholas Piggin <npiggin@gmail.com> > Acked-by: Paul Mackerras <paulus@ozlabs.org> Applied to powerpc next, thanks. https://git.kernel.org/powerpc/c/94a04bc25a2c6296bd0c5e82c10e82 cheers -- To unsubscribe from this list: send the line "unsubscribe kvm-ppc" 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/include/asm/kvm_book3s_asm.h b/arch/powerpc/include/asm/kvm_book3s_asm.h index 7cea76f11c26..83596f32f50b 100644 --- a/arch/powerpc/include/asm/kvm_book3s_asm.h +++ b/arch/powerpc/include/asm/kvm_book3s_asm.h @@ -104,6 +104,10 @@ struct kvmppc_host_state { u8 napping; #ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE + /* + * hwthread_req/hwthread_state pair is used to pull sibling threads + * out of guest on pre-ISAv3.0B CPUs where threads share MMU. + */ u8 hwthread_req; u8 hwthread_state; u8 host_ipi; diff --git a/arch/powerpc/kernel/idle_book3s.S b/arch/powerpc/kernel/idle_book3s.S index bfbf0976fc09..4924647d964d 100644 --- a/arch/powerpc/kernel/idle_book3s.S +++ b/arch/powerpc/kernel/idle_book3s.S @@ -296,13 +296,20 @@ enter_winkle: /* * r3 - PSSCR value corresponding to the requested stop state. */ -power_enter_stop: #ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE - /* Tell KVM we're entering idle */ +power_enter_stop_kvm_rm: + /* + * This is currently unused because POWER9 KVM does not have to + * gather secondary threads into sibling mode, but the code is + * here in case that function is required. + * + * Tell KVM we're entering idle. + */ li r4,KVM_HWTHREAD_IN_IDLE /* DO THIS IN REAL MODE! See comment above. */ stb r4,HSTATE_HWTHREAD_STATE(r13) #endif +power_enter_stop: /* * Check if we are executing the lite variant with ESL=EC=0 */ @@ -465,6 +472,18 @@ pnv_powersave_wakeup_mce: b pnv_powersave_wakeup +#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE +kvm_start_guest_check: + li r0,KVM_HWTHREAD_IN_KERNEL + stb r0,HSTATE_HWTHREAD_STATE(r13) + /* Order setting hwthread_state vs. testing hwthread_req */ + sync + lbz r0,HSTATE_HWTHREAD_REQ(r13) + cmpwi r0,0 + beqlr + b kvm_start_guest +#endif + /* * Called from reset vector for powersave wakeups. * cr3 - set to gt if waking up with partial/complete hypervisor state loss @@ -489,15 +508,9 @@ ALT_FTR_SECTION_END_IFSET(CPU_FTR_ARCH_300) mr r3,r12 #ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE - li r0,KVM_HWTHREAD_IN_KERNEL - stb r0,HSTATE_HWTHREAD_STATE(r13) - /* Order setting hwthread_state vs. testing hwthread_req */ - sync - lbz r0,HSTATE_HWTHREAD_REQ(r13) - cmpwi r0,0 - beq 1f - b kvm_start_guest -1: +BEGIN_FTR_SECTION + bl kvm_start_guest_check +END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_300) #endif /* Return SRR1 from power7_nap() */ diff --git a/arch/powerpc/kvm/book3s_hv.c b/arch/powerpc/kvm/book3s_hv.c index 359c79cdf0cc..e34cd6fb947b 100644 --- a/arch/powerpc/kvm/book3s_hv.c +++ b/arch/powerpc/kvm/book3s_hv.c @@ -2111,6 +2111,16 @@ static int kvmppc_grab_hwthread(int cpu) struct paca_struct *tpaca; long timeout = 10000; + /* + * ISA v3.0 idle routines do not set hwthread_state or test + * hwthread_req, so they can not grab idle threads. + */ + if (cpu_has_feature(CPU_FTR_ARCH_300)) { + WARN_ON(1); + pr_err("KVM: can not control sibling threads\n"); + return -EBUSY; + } + tpaca = &paca[cpu]; /* Ensure the thread won't go into the kernel if it wakes */ @@ -2145,10 +2155,12 @@ static void kvmppc_release_hwthread(int cpu) struct paca_struct *tpaca; tpaca = &paca[cpu]; - tpaca->kvm_hstate.hwthread_req = 0; tpaca->kvm_hstate.kvm_vcpu = NULL; tpaca->kvm_hstate.kvm_vcore = NULL; tpaca->kvm_hstate.kvm_split_mode = NULL; + if (!cpu_has_feature(CPU_FTR_ARCH_300)) + tpaca->kvm_hstate.hwthread_req = 0; + } static void radix_flush_cpu(struct kvm *kvm, int cpu, struct kvm_vcpu *vcpu) diff --git a/arch/powerpc/kvm/book3s_hv_rmhandlers.S b/arch/powerpc/kvm/book3s_hv_rmhandlers.S index c52184a8efdf..3e024fd71fe8 100644 --- a/arch/powerpc/kvm/book3s_hv_rmhandlers.S +++ b/arch/powerpc/kvm/book3s_hv_rmhandlers.S @@ -149,9 +149,11 @@ END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S) subf r4, r4, r3 mtspr SPRN_DEC, r4 +BEGIN_FTR_SECTION /* hwthread_req may have got set by cede or no vcpu, so clear it */ li r0, 0 stb r0, HSTATE_HWTHREAD_REQ(r13) +END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_300) /* * For external interrupts we need to call the Linux @@ -314,6 +316,7 @@ kvm_novcpu_exit: * Relocation is off and most register values are lost. * r13 points to the PACA. * r3 contains the SRR1 wakeup value, SRR1 is trashed. + * This is not used by ISAv3.0B processors. */ .globl kvm_start_guest kvm_start_guest: @@ -432,6 +435,9 @@ kvm_secondary_got_guest: * While waiting we also need to check if we get given a vcpu to run. */ kvm_no_guest: +BEGIN_FTR_SECTION + twi 31,0,0 +END_FTR_SECTION_IFSET(CPU_FTR_ARCH_300) lbz r3, HSTATE_HWTHREAD_REQ(r13) cmpwi r3, 0 bne 53f @@ -2509,8 +2515,10 @@ kvm_do_nap: clrrdi r0, r0, 1 mtspr SPRN_CTRLT, r0 +BEGIN_FTR_SECTION li r0,1 stb r0,HSTATE_HWTHREAD_REQ(r13) +END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_300) mfspr r5,SPRN_LPCR ori r5,r5,LPCR_PECE0 | LPCR_PECE1 BEGIN_FTR_SECTION
POWER9 CPUs have independent MMU contexts per thread, so KVM does not need to quiesce secondary threads, so the hwthread_req/hwthread_state protocol does not have to be used. So patch it away on POWER9, and patch away the branch from the Linux idle wakeup to kvm_start_guest that is never used. Add a warning and error out of kvmppc_grab_hwthread in case it is ever called on POWER9. This avoids a hwsync in the idle wakeup path on POWER9. Signed-off-by: Nicholas Piggin <npiggin@gmail.com> --- arch/powerpc/include/asm/kvm_book3s_asm.h | 4 ++++ arch/powerpc/kernel/idle_book3s.S | 35 +++++++++++++++++++++---------- arch/powerpc/kvm/book3s_hv.c | 14 ++++++++++++- arch/powerpc/kvm/book3s_hv_rmhandlers.S | 8 +++++++ 4 files changed, 49 insertions(+), 12 deletions(-)