@@ -244,6 +244,7 @@ struct kvm_arch_memory_slot {
struct kvm_arch {
unsigned int lpid;
#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
+ unsigned int tlb_sets;
unsigned long hpt_virt;
struct revmap_entry *revmap;
atomic64_t mmio_update;
@@ -487,6 +487,7 @@ int main(void)
/* book3s */
#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
+ DEFINE(KVM_TLB_SETS, offsetof(struct kvm, arch.tlb_sets));
DEFINE(KVM_SDR1, offsetof(struct kvm, arch.sdr1));
DEFINE(KVM_HOST_LPID, offsetof(struct kvm, arch.host_lpid));
DEFINE(KVM_HOST_LPCR, offsetof(struct kvm, arch.host_lpcr));
@@ -3265,8 +3265,11 @@ static int kvmppc_core_init_vm_hv(struct kvm *kvm)
* Since we don't flush the TLB when tearing down a VM,
* and this lpid might have previously been used,
* make sure we flush on each core before running the new VM.
+ * On POWER9, the tlbie in mmu_partition_table_set_entry()
+ * does this flush for us.
*/
- cpumask_setall(&kvm->arch.need_tlb_flush);
+ if (!cpu_has_feature(CPU_FTR_ARCH_300))
+ cpumask_setall(&kvm->arch.need_tlb_flush);
/* Start out with the default set of hcalls enabled */
memcpy(kvm->arch.enabled_hcalls, default_enabled_hcalls,
@@ -3292,6 +3295,17 @@ static int kvmppc_core_init_vm_hv(struct kvm *kvm)
kvm->arch.lpcr = lpcr;
/*
+ * Work out how many sets the TLB has, for the use of
+ * the TLB invalidation loop in book3s_hv_rmhandlers.S.
+ */
+ if (cpu_has_feature(CPU_FTR_ARCH_300))
+ kvm->arch.tlb_sets = POWER9_TLB_SETS_HASH; /* 256 */
+ else if (cpu_has_feature(CPU_FTR_ARCH_207S))
+ kvm->arch.tlb_sets = POWER8_TLB_SETS; /* 512 */
+ else
+ kvm->arch.tlb_sets = POWER7_TLB_SETS; /* 128 */
+
+ /*
* Track that we now have a HV mode VM active. This blocks secondary
* CPU threads from coming online.
*/
@@ -3733,3 +3747,4 @@ module_exit(kvmppc_book3s_exit_hv);
MODULE_LICENSE("GPL");
MODULE_ALIAS_MISCDEV(KVM_MINOR);
MODULE_ALIAS("devname:kvm");
+
@@ -425,13 +425,18 @@ static void do_tlbies(struct kvm *kvm, unsigned long *rbvalues,
{
long i;
+ /*
+ * We use the POWER9 5-operand versions of tlbie and tlbiel here.
+ * Since we are using RIC=0 PRS=0 R=0, and P7/P8 tlbiel ignores
+ * the RS field, this is backwards-compatible with P7 and P8.
+ */
if (global) {
while (!try_lock_tlbie(&kvm->arch.tlbie_lock))
cpu_relax();
if (need_sync)
asm volatile("ptesync" : : : "memory");
for (i = 0; i < npages; ++i)
- asm volatile(PPC_TLBIE(%1,%0) : :
+ asm volatile(PPC_TLBIE_5(%0,%1,0,0,0) : :
"r" (rbvalues[i]), "r" (kvm->arch.lpid));
asm volatile("eieio; tlbsync; ptesync" : : : "memory");
kvm->arch.tlbie_lock = 0;
@@ -439,7 +444,8 @@ static void do_tlbies(struct kvm *kvm, unsigned long *rbvalues,
if (need_sync)
asm volatile("ptesync" : : : "memory");
for (i = 0; i < npages; ++i)
- asm volatile("tlbiel %0" : : "r" (rbvalues[i]));
+ asm volatile(PPC_TLBIEL(%0,%1,0,0,0) : :
+ "r" (rbvalues[i]), "r" (0));
asm volatile("ptesync" : : : "memory");
}
}
@@ -613,12 +613,8 @@ END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_300)
stdcx. r7,0,r6
bne 23b
/* Flush the TLB of any entries for this LPID */
- /* use arch 2.07S as a proxy for POWER8 */
-BEGIN_FTR_SECTION
- li r6,512 /* POWER8 has 512 sets */
-FTR_SECTION_ELSE
- li r6,128 /* POWER7 has 128 sets */
-ALT_FTR_SECTION_END_IFSET(CPU_FTR_ARCH_207S)
+ lwz r6,KVM_TLB_SETS(r9)
+ li r0,0 /* RS for P9 version of tlbiel */
mtctr r6
li r7,0x800 /* IS field = 0b10 */
ptesync
POWER9 adds new capabilities to the tlbie (TLB invalidate entry) and tlbiel (local tlbie) instructions. Both instructions get a set of new parameters (RIC, PRS and R) which appear as bits in the instruction word. The tlbiel instruction now has a second register operand, which contains a PID and/or LPID value if needed, and should otherwise contain 0. This adapts KVM-HV's usage of tlbie and tlbiel to work on POWER9 as well as older processors. Since we only handle HPT guests so far, we need RIC=0 PRS=0 R=0, which ends up with the same instruction word as on previous processors, so we don't need to conditionally execute different instructions depending on the processor. The local flush on first entry to a guest in book3s_hv_rmhandlers.S is a loop which depends on the number of TLB sets. Rather than using feature sections to set the number of iterations based on which CPU we're on, we now work out this number at VM creation time and store it in the kvm_arch struct. That will make it possible to get the number from the device tree in future, which will help with compatibility with future processors. Since mmu_partition_table_set_entry() does a global flush of the whole LPID, we don't need to do the TLB flush on first entry to the guest on each processor. Therefore we don't set all bits in the tlb_need_flush bitmap on VM startup on POWER9. Signed-off-by: Paul Mackerras <paulus@ozlabs.org> --- arch/powerpc/include/asm/kvm_host.h | 1 + arch/powerpc/kernel/asm-offsets.c | 1 + arch/powerpc/kvm/book3s_hv.c | 17 ++++++++++++++++- arch/powerpc/kvm/book3s_hv_rm_mmu.c | 10 ++++++++-- arch/powerpc/kvm/book3s_hv_rmhandlers.S | 8 ++------ 5 files changed, 28 insertions(+), 9 deletions(-)