powerpc/kvm/book3s: Avoid using rmap to protect parallel page table update.

We now depend on kvm->mmu_lock

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200505071729.54912-17-aneesh.kumar@linux.ibm.com
This commit is contained in:
Aneesh Kumar K.V 2020-05-05 12:47:23 +05:30 committed by Michael Ellerman
parent 7769a3394b
commit 3ff8df1430
1 changed files with 9 additions and 29 deletions

View File

@ -74,8 +74,8 @@ struct kvmppc_spapr_tce_table *kvmppc_find_table(struct kvm *kvm,
EXPORT_SYMBOL_GPL(kvmppc_find_table); EXPORT_SYMBOL_GPL(kvmppc_find_table);
#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE #ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
static long kvmppc_rm_tce_to_ua(struct kvm *kvm, unsigned long tce, static long kvmppc_rm_tce_to_ua(struct kvm *kvm,
unsigned long *ua, unsigned long **prmap) unsigned long tce, unsigned long *ua)
{ {
unsigned long gfn = tce >> PAGE_SHIFT; unsigned long gfn = tce >> PAGE_SHIFT;
struct kvm_memory_slot *memslot; struct kvm_memory_slot *memslot;
@ -87,9 +87,6 @@ static long kvmppc_rm_tce_to_ua(struct kvm *kvm, unsigned long tce,
*ua = __gfn_to_hva_memslot(memslot, gfn) | *ua = __gfn_to_hva_memslot(memslot, gfn) |
(tce & ~(PAGE_MASK | TCE_PCI_READ | TCE_PCI_WRITE)); (tce & ~(PAGE_MASK | TCE_PCI_READ | TCE_PCI_WRITE));
if (prmap)
*prmap = &memslot->arch.rmap[gfn - memslot->base_gfn];
return 0; return 0;
} }
@ -116,7 +113,7 @@ static long kvmppc_rm_tce_validate(struct kvmppc_spapr_tce_table *stt,
if (iommu_tce_check_gpa(stt->page_shift, gpa)) if (iommu_tce_check_gpa(stt->page_shift, gpa))
return H_PARAMETER; return H_PARAMETER;
if (kvmppc_rm_tce_to_ua(stt->kvm, tce, &ua, NULL)) if (kvmppc_rm_tce_to_ua(stt->kvm, tce, &ua))
return H_TOO_HARD; return H_TOO_HARD;
list_for_each_entry_lockless(stit, &stt->iommu_tables, next) { list_for_each_entry_lockless(stit, &stt->iommu_tables, next) {
@ -411,7 +408,7 @@ long kvmppc_rm_h_put_tce(struct kvm_vcpu *vcpu, unsigned long liobn,
return ret; return ret;
dir = iommu_tce_direction(tce); dir = iommu_tce_direction(tce);
if ((dir != DMA_NONE) && kvmppc_rm_tce_to_ua(vcpu->kvm, tce, &ua, NULL)) if ((dir != DMA_NONE) && kvmppc_rm_tce_to_ua(vcpu->kvm, tce, &ua))
return H_PARAMETER; return H_PARAMETER;
entry = ioba >> stt->page_shift; entry = ioba >> stt->page_shift;
@ -488,7 +485,6 @@ long kvmppc_rm_h_put_tce_indirect(struct kvm_vcpu *vcpu,
struct kvmppc_spapr_tce_table *stt; struct kvmppc_spapr_tce_table *stt;
long i, ret = H_SUCCESS; long i, ret = H_SUCCESS;
unsigned long tces, entry, ua = 0; unsigned long tces, entry, ua = 0;
unsigned long *rmap = NULL;
unsigned long mmu_seq; unsigned long mmu_seq;
bool prereg = false; bool prereg = false;
struct kvmppc_spapr_tce_iommu_table *stit; struct kvmppc_spapr_tce_iommu_table *stit;
@ -530,7 +526,7 @@ long kvmppc_rm_h_put_tce_indirect(struct kvm_vcpu *vcpu,
*/ */
struct mm_iommu_table_group_mem_t *mem; struct mm_iommu_table_group_mem_t *mem;
if (kvmppc_rm_tce_to_ua(vcpu->kvm, tce_list, &ua, NULL)) if (kvmppc_rm_tce_to_ua(vcpu->kvm, tce_list, &ua))
return H_TOO_HARD; return H_TOO_HARD;
mem = mm_iommu_lookup_rm(vcpu->kvm->mm, ua, IOMMU_PAGE_SIZE_4K); mem = mm_iommu_lookup_rm(vcpu->kvm->mm, ua, IOMMU_PAGE_SIZE_4K);
@ -546,23 +542,9 @@ long kvmppc_rm_h_put_tce_indirect(struct kvm_vcpu *vcpu,
* We do not require memory to be preregistered in this case * We do not require memory to be preregistered in this case
* so lock rmap and do __find_linux_pte_or_hugepte(). * so lock rmap and do __find_linux_pte_or_hugepte().
*/ */
if (kvmppc_rm_tce_to_ua(vcpu->kvm, tce_list, &ua, &rmap)) if (kvmppc_rm_tce_to_ua(vcpu->kvm, tce_list, &ua))
return H_TOO_HARD; return H_TOO_HARD;
rmap = (void *) vmalloc_to_phys(rmap);
if (WARN_ON_ONCE_RM(!rmap))
return H_TOO_HARD;
/*
* Synchronize with the MMU notifier callbacks in
* book3s_64_mmu_hv.c (kvm_unmap_hva_range_hv etc.).
* While we have the rmap lock, code running on other CPUs
* cannot finish unmapping the host real page that backs
* this guest real page, so we are OK to access the host
* real page.
*/
lock_rmap(rmap);
arch_spin_lock(&kvm->mmu_lock.rlock.raw_lock); arch_spin_lock(&kvm->mmu_lock.rlock.raw_lock);
if (kvmppc_rm_ua_to_hpa(vcpu, mmu_seq, ua, &tces)) { if (kvmppc_rm_ua_to_hpa(vcpu, mmu_seq, ua, &tces)) {
ret = H_TOO_HARD; ret = H_TOO_HARD;
@ -582,7 +564,7 @@ long kvmppc_rm_h_put_tce_indirect(struct kvm_vcpu *vcpu,
unsigned long tce = be64_to_cpu(((u64 *)tces)[i]); unsigned long tce = be64_to_cpu(((u64 *)tces)[i]);
ua = 0; ua = 0;
if (kvmppc_rm_tce_to_ua(vcpu->kvm, tce, &ua, NULL)) { if (kvmppc_rm_tce_to_ua(vcpu->kvm, tce, &ua)) {
ret = H_PARAMETER; ret = H_PARAMETER;
goto invalidate_exit; goto invalidate_exit;
} }
@ -607,10 +589,8 @@ long kvmppc_rm_h_put_tce_indirect(struct kvm_vcpu *vcpu,
iommu_tce_kill_rm(stit->tbl, entry, npages); iommu_tce_kill_rm(stit->tbl, entry, npages);
unlock_exit: unlock_exit:
if (rmap) if (!prereg)
unlock_rmap(rmap); arch_spin_unlock(&kvm->mmu_lock.rlock.raw_lock);
arch_spin_unlock(&kvm->mmu_lock.rlock.raw_lock);
return ret; return ret;
} }