Patchwork [2/6] KVM: PPC: Book3S HV: Make HTAB code LE host aware

login
register
mail settings
Submitter Alexander Graf
Date June 11, 2014, 10:33 a.m.
Message ID <1402482831-8941-3-git-send-email-agraf@suse.de>
Download mbox | patch
Permalink /patch/358603/
State New
Headers show

Comments

Alexander Graf - June 11, 2014, 10:33 a.m.
When running on an LE host all data structures are kept in little endian
byte order. However, the HTAB still needs to be maintained in big endian.

So every time we access any HTAB we need to make sure we do so in the right
byte order. Fix up all accesses to manually byte swap.

Signed-off-by: Alexander Graf <agraf@suse.de>
---
 arch/powerpc/include/asm/kvm_book3s_64.h |  14 +++-
 arch/powerpc/kvm/book3s_64_mmu_hv.c      |  95 +++++++++++++-------------
 arch/powerpc/kvm/book3s_hv_rm_mmu.c      | 111 ++++++++++++++++++-------------
 3 files changed, 126 insertions(+), 94 deletions(-)

Patch

diff --git a/arch/powerpc/include/asm/kvm_book3s_64.h b/arch/powerpc/include/asm/kvm_book3s_64.h
index fddb72b..2e99a3e 100644
--- a/arch/powerpc/include/asm/kvm_book3s_64.h
+++ b/arch/powerpc/include/asm/kvm_book3s_64.h
@@ -61,18 +61,26 @@  extern unsigned long kvm_rma_pages;
 
 static inline long try_lock_hpte(unsigned long *hpte, unsigned long bits)
 {
-	unsigned long tmp, old;
+	unsigned long tmp, old, lockbit;
+
+	/*
+	 * We load/store in native endian, but the HTAB is in big endian. If
+	 * we byte swap all data we apply on the PTE we're implicitly correct
+	 * again.
+	 */
+	lockbit = cpu_to_be64(HPTE_V_HVLOCK);
+	bits = cpu_to_be64(bits);
 
 	asm volatile("	ldarx	%0,0,%2\n"
 		     "	and.	%1,%0,%3\n"
 		     "	bne	2f\n"
-		     "	ori	%0,%0,%4\n"
+		     "	or	%0,%0,%4\n"
 		     "  stdcx.	%0,0,%2\n"
 		     "	beq+	2f\n"
 		     "	mr	%1,%3\n"
 		     "2:	isync"
 		     : "=&r" (tmp), "=&r" (old)
-		     : "r" (hpte), "r" (bits), "i" (HPTE_V_HVLOCK)
+		     : "r" (hpte), "r" (bits), "r" (lockbit)
 		     : "cc", "memory");
 	return old == 0;
 }
diff --git a/arch/powerpc/kvm/book3s_64_mmu_hv.c b/arch/powerpc/kvm/book3s_64_mmu_hv.c
index 8056107..f53cf2e 100644
--- a/arch/powerpc/kvm/book3s_64_mmu_hv.c
+++ b/arch/powerpc/kvm/book3s_64_mmu_hv.c
@@ -474,12 +474,12 @@  static int kvmppc_mmu_book3s_64_hv_xlate(struct kvm_vcpu *vcpu, gva_t eaddr,
 		return -ENOENT;
 	}
 	hptep = (unsigned long *)(kvm->arch.hpt_virt + (index << 4));
-	v = hptep[0] & ~HPTE_V_HVLOCK;
+	v = be64_to_cpu(hptep[0]) & ~HPTE_V_HVLOCK;
 	gr = kvm->arch.revmap[index].guest_rpte;
 
 	/* Unlock the HPTE */
 	asm volatile("lwsync" : : : "memory");
-	hptep[0] = v;
+	hptep[0] = cpu_to_be64(v);
 	preempt_enable();
 
 	gpte->eaddr = eaddr;
@@ -611,11 +611,11 @@  int kvmppc_book3s_hv_page_fault(struct kvm_run *run, struct kvm_vcpu *vcpu,
 	preempt_disable();
 	while (!try_lock_hpte(hptep, HPTE_V_HVLOCK))
 		cpu_relax();
-	hpte[0] = hptep[0] & ~HPTE_V_HVLOCK;
-	hpte[1] = hptep[1];
+	hpte[0] = be64_to_cpu(hptep[0]) & ~HPTE_V_HVLOCK;
+	hpte[1] = be64_to_cpu(hptep[1]);
 	hpte[2] = r = rev->guest_rpte;
 	asm volatile("lwsync" : : : "memory");
-	hptep[0] = hpte[0];
+	hptep[0] = cpu_to_be64(hpte[0]);
 	preempt_enable();
 
 	if (hpte[0] != vcpu->arch.pgfault_hpte[0] ||
@@ -731,8 +731,9 @@  int kvmppc_book3s_hv_page_fault(struct kvm_run *run, struct kvm_vcpu *vcpu,
 	preempt_disable();
 	while (!try_lock_hpte(hptep, HPTE_V_HVLOCK))
 		cpu_relax();
-	if ((hptep[0] & ~HPTE_V_HVLOCK) != hpte[0] || hptep[1] != hpte[1] ||
-	    rev->guest_rpte != hpte[2])
+	if ((be64_to_cpu(hptep[0]) & ~HPTE_V_HVLOCK) != hpte[0] ||
+		be64_to_cpu(hptep[1]) != hpte[1] ||
+		rev->guest_rpte != hpte[2])
 		/* HPTE has been changed under us; let the guest retry */
 		goto out_unlock;
 	hpte[0] = (hpte[0] & ~HPTE_V_ABSENT) | HPTE_V_VALID;
@@ -752,20 +753,20 @@  int kvmppc_book3s_hv_page_fault(struct kvm_run *run, struct kvm_vcpu *vcpu,
 	rcbits = *rmap >> KVMPPC_RMAP_RC_SHIFT;
 	r &= rcbits | ~(HPTE_R_R | HPTE_R_C);
 
-	if (hptep[0] & HPTE_V_VALID) {
+	if (be64_to_cpu(hptep[0]) & HPTE_V_VALID) {
 		/* HPTE was previously valid, so we need to invalidate it */
 		unlock_rmap(rmap);
-		hptep[0] |= HPTE_V_ABSENT;
+		hptep[0] |= cpu_to_be64(HPTE_V_ABSENT);
 		kvmppc_invalidate_hpte(kvm, hptep, index);
 		/* don't lose previous R and C bits */
-		r |= hptep[1] & (HPTE_R_R | HPTE_R_C);
+		r |= be64_to_cpu(hptep[1]) & (HPTE_R_R | HPTE_R_C);
 	} else {
 		kvmppc_add_revmap_chain(kvm, rev, rmap, index, 0);
 	}
 
-	hptep[1] = r;
+	hptep[1] = cpu_to_be64(r);
 	eieio();
-	hptep[0] = hpte[0];
+	hptep[0] = cpu_to_be64(hpte[0]);
 	asm volatile("ptesync" : : : "memory");
 	preempt_enable();
 	if (page && hpte_is_writable(r))
@@ -784,7 +785,7 @@  int kvmppc_book3s_hv_page_fault(struct kvm_run *run, struct kvm_vcpu *vcpu,
 	return ret;
 
  out_unlock:
-	hptep[0] &= ~HPTE_V_HVLOCK;
+	hptep[0] &= ~cpu_to_be64(HPTE_V_HVLOCK);
 	preempt_enable();
 	goto out_put;
 }
@@ -880,7 +881,7 @@  static int kvm_unmap_rmapp(struct kvm *kvm, unsigned long *rmapp,
 		if (!try_lock_hpte(hptep, HPTE_V_HVLOCK)) {
 			/* unlock rmap before spinning on the HPTE lock */
 			unlock_rmap(rmapp);
-			while (hptep[0] & HPTE_V_HVLOCK)
+			while (be64_to_cpu(hptep[0]) & HPTE_V_HVLOCK)
 				cpu_relax();
 			continue;
 		}
@@ -899,14 +900,14 @@  static int kvm_unmap_rmapp(struct kvm *kvm, unsigned long *rmapp,
 
 		/* Now check and modify the HPTE */
 		ptel = rev[i].guest_rpte;
-		psize = hpte_page_size(hptep[0], ptel);
-		if ((hptep[0] & HPTE_V_VALID) &&
+		psize = hpte_page_size(be64_to_cpu(hptep[0]), ptel);
+		if ((be64_to_cpu(hptep[0]) & HPTE_V_VALID) &&
 		    hpte_rpn(ptel, psize) == gfn) {
 			if (kvm->arch.using_mmu_notifiers)
-				hptep[0] |= HPTE_V_ABSENT;
+				hptep[0] |= cpu_to_be64(HPTE_V_ABSENT);
 			kvmppc_invalidate_hpte(kvm, hptep, i);
 			/* Harvest R and C */
-			rcbits = hptep[1] & (HPTE_R_R | HPTE_R_C);
+			rcbits = be64_to_cpu(hptep[1]) & (HPTE_R_R | HPTE_R_C);
 			*rmapp |= rcbits << KVMPPC_RMAP_RC_SHIFT;
 			if (rcbits & ~rev[i].guest_rpte) {
 				rev[i].guest_rpte = ptel | rcbits;
@@ -914,7 +915,7 @@  static int kvm_unmap_rmapp(struct kvm *kvm, unsigned long *rmapp,
 			}
 		}
 		unlock_rmap(rmapp);
-		hptep[0] &= ~HPTE_V_HVLOCK;
+		hptep[0] &= ~cpu_to_be64(HPTE_V_HVLOCK);
 	}
 	return 0;
 }
@@ -981,19 +982,20 @@  static int kvm_age_rmapp(struct kvm *kvm, unsigned long *rmapp,
 		j = rev[i].forw;
 
 		/* If this HPTE isn't referenced, ignore it */
-		if (!(hptep[1] & HPTE_R_R))
+		if (!(be64_to_cpu(hptep[1]) & HPTE_R_R))
 			continue;
 
 		if (!try_lock_hpte(hptep, HPTE_V_HVLOCK)) {
 			/* unlock rmap before spinning on the HPTE lock */
 			unlock_rmap(rmapp);
-			while (hptep[0] & HPTE_V_HVLOCK)
+			while (be64_to_cpu(hptep[0]) & HPTE_V_HVLOCK)
 				cpu_relax();
 			goto retry;
 		}
 
 		/* Now check and modify the HPTE */
-		if ((hptep[0] & HPTE_V_VALID) && (hptep[1] & HPTE_R_R)) {
+		if ((be64_to_cpu(hptep[0]) & HPTE_V_VALID) &&
+		    (be64_to_cpu(hptep[1]) & HPTE_R_R)) {
 			kvmppc_clear_ref_hpte(kvm, hptep, i);
 			if (!(rev[i].guest_rpte & HPTE_R_R)) {
 				rev[i].guest_rpte |= HPTE_R_R;
@@ -1001,7 +1003,7 @@  static int kvm_age_rmapp(struct kvm *kvm, unsigned long *rmapp,
 			}
 			ret = 1;
 		}
-		hptep[0] &= ~HPTE_V_HVLOCK;
+		hptep[0] &= ~cpu_to_be64(HPTE_V_HVLOCK);
 	} while ((i = j) != head);
 
 	unlock_rmap(rmapp);
@@ -1035,7 +1037,7 @@  static int kvm_test_age_rmapp(struct kvm *kvm, unsigned long *rmapp,
 		do {
 			hp = (unsigned long *)(kvm->arch.hpt_virt + (i << 4));
 			j = rev[i].forw;
-			if (hp[1] & HPTE_R_R)
+			if (be64_to_cpu(hp[1]) & HPTE_R_R)
 				goto out;
 		} while ((i = j) != head);
 	}
@@ -1091,6 +1093,7 @@  static int kvm_test_clear_dirty_npages(struct kvm *kvm, unsigned long *rmapp)
 
 	i = head = *rmapp & KVMPPC_RMAP_INDEX;
 	do {
+		unsigned long hptep1;
 		hptep = (unsigned long *) (kvm->arch.hpt_virt + (i << 4));
 		j = rev[i].forw;
 
@@ -1108,29 +1111,30 @@  static int kvm_test_clear_dirty_npages(struct kvm *kvm, unsigned long *rmapp)
 		 * Otherwise we need to do the tlbie even if C==0 in
 		 * order to pick up any delayed writeback of C.
 		 */
-		if (!(hptep[1] & HPTE_R_C) &&
-		    (!hpte_is_writable(hptep[1]) || vcpus_running(kvm)))
+		hptep1 = be64_to_cpu(hptep[1]);
+		if (!(hptep1 & HPTE_R_C) &&
+		    (!hpte_is_writable(hptep1) || vcpus_running(kvm)))
 			continue;
 
 		if (!try_lock_hpte(hptep, HPTE_V_HVLOCK)) {
 			/* unlock rmap before spinning on the HPTE lock */
 			unlock_rmap(rmapp);
-			while (hptep[0] & HPTE_V_HVLOCK)
+			while (hptep[0] & cpu_to_be64(HPTE_V_HVLOCK))
 				cpu_relax();
 			goto retry;
 		}
 
 		/* Now check and modify the HPTE */
-		if (!(hptep[0] & HPTE_V_VALID))
+		if (!(hptep[0] & cpu_to_be64(HPTE_V_VALID)))
 			continue;
 
 		/* need to make it temporarily absent so C is stable */
-		hptep[0] |= HPTE_V_ABSENT;
+		hptep[0] |= cpu_to_be64(HPTE_V_ABSENT);
 		kvmppc_invalidate_hpte(kvm, hptep, i);
-		v = hptep[0];
-		r = hptep[1];
+		v = be64_to_cpu(hptep[0]);
+		r = be64_to_cpu(hptep[1]);
 		if (r & HPTE_R_C) {
-			hptep[1] = r & ~HPTE_R_C;
+			hptep[1] = cpu_to_be64(r & ~HPTE_R_C);
 			if (!(rev[i].guest_rpte & HPTE_R_C)) {
 				rev[i].guest_rpte |= HPTE_R_C;
 				note_hpte_modification(kvm, &rev[i]);
@@ -1143,7 +1147,7 @@  static int kvm_test_clear_dirty_npages(struct kvm *kvm, unsigned long *rmapp)
 		}
 		v &= ~(HPTE_V_ABSENT | HPTE_V_HVLOCK);
 		v |= HPTE_V_VALID;
-		hptep[0] = v;
+		hptep[0] = cpu_to_be64(v);
 	} while ((i = j) != head);
 
 	unlock_rmap(rmapp);
@@ -1316,7 +1320,8 @@  static int hpte_dirty(struct revmap_entry *revp, unsigned long *hptp)
 
 	/* Also need to consider changes in reference and changed bits */
 	rcbits_unset = ~revp->guest_rpte & (HPTE_R_R | HPTE_R_C);
-	if ((hptp[0] & HPTE_V_VALID) && (hptp[1] & rcbits_unset))
+	if ((be64_to_cpu(hptp[0]) & HPTE_V_VALID) &&
+	    (be64_to_cpu(hptp[1]) & rcbits_unset))
 		return 1;
 
 	return 0;
@@ -1337,10 +1342,10 @@  static long record_hpte(unsigned long flags, unsigned long *hptp,
 		return 0;
 
 	valid = 0;
-	if (hptp[0] & (HPTE_V_VALID | HPTE_V_ABSENT)) {
+	if (be64_to_cpu(hptp[0]) & (HPTE_V_VALID | HPTE_V_ABSENT)) {
 		valid = 1;
 		if ((flags & KVM_GET_HTAB_BOLTED_ONLY) &&
-		    !(hptp[0] & HPTE_V_BOLTED))
+		    !(be64_to_cpu(hptp[0]) & HPTE_V_BOLTED))
 			valid = 0;
 	}
 	if (valid != want_valid)
@@ -1352,7 +1357,7 @@  static long record_hpte(unsigned long flags, unsigned long *hptp,
 		preempt_disable();
 		while (!try_lock_hpte(hptp, HPTE_V_HVLOCK))
 			cpu_relax();
-		v = hptp[0];
+		v = be64_to_cpu(hptp[0]);
 
 		/* re-evaluate valid and dirty from synchronized HPTE value */
 		valid = !!(v & HPTE_V_VALID);
@@ -1360,9 +1365,9 @@  static long record_hpte(unsigned long flags, unsigned long *hptp,
 
 		/* Harvest R and C into guest view if necessary */
 		rcbits_unset = ~revp->guest_rpte & (HPTE_R_R | HPTE_R_C);
-		if (valid && (rcbits_unset & hptp[1])) {
-			revp->guest_rpte |= (hptp[1] & (HPTE_R_R | HPTE_R_C)) |
-				HPTE_GR_MODIFIED;
+		if (valid && (rcbits_unset & be64_to_cpu(hptp[1]))) {
+			revp->guest_rpte |= (be64_to_cpu(hptp[1]) &
+				(HPTE_R_R | HPTE_R_C)) | HPTE_GR_MODIFIED;
 			dirty = 1;
 		}
 
@@ -1381,13 +1386,13 @@  static long record_hpte(unsigned long flags, unsigned long *hptp,
 			revp->guest_rpte = r;
 		}
 		asm volatile(PPC_RELEASE_BARRIER "" : : : "memory");
-		hptp[0] &= ~HPTE_V_HVLOCK;
+		hptp[0] &= ~cpu_to_be64(HPTE_V_HVLOCK);
 		preempt_enable();
 		if (!(valid == want_valid && (first_pass || dirty)))
 			ok = 0;
 	}
-	hpte[0] = v;
-	hpte[1] = r;
+	hpte[0] = cpu_to_be64(v);
+	hpte[1] = cpu_to_be64(r);
 	return ok;
 }
 
@@ -1551,7 +1556,7 @@  static ssize_t kvm_htab_write(struct file *file, const char __user *buf,
 			lbuf += 2;
 			nb += HPTE_SIZE;
 
-			if (hptp[0] & (HPTE_V_VALID | HPTE_V_ABSENT))
+			if (be64_to_cpu(hptp[0]) & (HPTE_V_VALID | HPTE_V_ABSENT))
 				kvmppc_do_h_remove(kvm, 0, i, 0, tmp);
 			err = -EIO;
 			ret = kvmppc_virtmode_do_h_enter(kvm, H_EXACT, i, v, r,
@@ -1577,7 +1582,7 @@  static ssize_t kvm_htab_write(struct file *file, const char __user *buf,
 		}
 
 		for (j = 0; j < hdr.n_invalid; ++j) {
-			if (hptp[0] & (HPTE_V_VALID | HPTE_V_ABSENT))
+			if (be64_to_cpu(hptp[0]) & (HPTE_V_VALID | HPTE_V_ABSENT))
 				kvmppc_do_h_remove(kvm, 0, i, 0, tmp);
 			++i;
 			hptp += 2;
diff --git a/arch/powerpc/kvm/book3s_hv_rm_mmu.c b/arch/powerpc/kvm/book3s_hv_rm_mmu.c
index ac840c6..87624ab 100644
--- a/arch/powerpc/kvm/book3s_hv_rm_mmu.c
+++ b/arch/powerpc/kvm/book3s_hv_rm_mmu.c
@@ -157,7 +157,7 @@  static pte_t lookup_linux_pte_and_update(pgd_t *pgdir, unsigned long hva,
 static inline void unlock_hpte(unsigned long *hpte, unsigned long hpte_v)
 {
 	asm volatile(PPC_RELEASE_BARRIER "" : : : "memory");
-	hpte[0] = hpte_v;
+	hpte[0] = cpu_to_be64(hpte_v);
 }
 
 long kvmppc_do_h_enter(struct kvm *kvm, unsigned long flags,
@@ -277,7 +277,7 @@  long kvmppc_do_h_enter(struct kvm *kvm, unsigned long flags,
 		pte_index &= ~7UL;
 		hpte = (unsigned long *)(kvm->arch.hpt_virt + (pte_index << 4));
 		for (i = 0; i < 8; ++i) {
-			if ((*hpte & HPTE_V_VALID) == 0 &&
+			if ((be64_to_cpu(*hpte) & HPTE_V_VALID) == 0 &&
 			    try_lock_hpte(hpte, HPTE_V_HVLOCK | HPTE_V_VALID |
 					  HPTE_V_ABSENT))
 				break;
@@ -292,11 +292,13 @@  long kvmppc_do_h_enter(struct kvm *kvm, unsigned long flags,
 			 */
 			hpte -= 16;
 			for (i = 0; i < 8; ++i) {
+				u64 pte;
 				while (!try_lock_hpte(hpte, HPTE_V_HVLOCK))
 					cpu_relax();
-				if (!(*hpte & (HPTE_V_VALID | HPTE_V_ABSENT)))
+				pte = be64_to_cpu(*hpte);
+				if (!(pte & (HPTE_V_VALID | HPTE_V_ABSENT)))
 					break;
-				*hpte &= ~HPTE_V_HVLOCK;
+				*hpte &= ~cpu_to_be64(HPTE_V_HVLOCK);
 				hpte += 2;
 			}
 			if (i == 8)
@@ -308,10 +310,13 @@  long kvmppc_do_h_enter(struct kvm *kvm, unsigned long flags,
 		if (!try_lock_hpte(hpte, HPTE_V_HVLOCK | HPTE_V_VALID |
 				   HPTE_V_ABSENT)) {
 			/* Lock the slot and check again */
+			u64 pte;
+
 			while (!try_lock_hpte(hpte, HPTE_V_HVLOCK))
 				cpu_relax();
-			if (*hpte & (HPTE_V_VALID | HPTE_V_ABSENT)) {
-				*hpte &= ~HPTE_V_HVLOCK;
+			pte = be64_to_cpu(*hpte);
+			if (pte & (HPTE_V_VALID | HPTE_V_ABSENT)) {
+				*hpte &= ~cpu_to_be64(HPTE_V_HVLOCK);
 				return H_PTEG_FULL;
 			}
 		}
@@ -347,11 +352,11 @@  long kvmppc_do_h_enter(struct kvm *kvm, unsigned long flags,
 		}
 	}
 
-	hpte[1] = ptel;
+	hpte[1] = cpu_to_be64(ptel);
 
 	/* Write the first HPTE dword, unlocking the HPTE and making it valid */
 	eieio();
-	hpte[0] = pteh;
+	hpte[0] = cpu_to_be64(pteh);
 	asm volatile("ptesync" : : : "memory");
 
 	*pte_idx_ret = pte_index;
@@ -471,34 +476,39 @@  long kvmppc_do_h_remove(struct kvm *kvm, unsigned long flags,
 	unsigned long *hpte;
 	unsigned long v, r, rb;
 	struct revmap_entry *rev;
+	u64 pte;
 
 	if (pte_index >= kvm->arch.hpt_npte)
 		return H_PARAMETER;
 	hpte = (unsigned long *)(kvm->arch.hpt_virt + (pte_index << 4));
 	while (!try_lock_hpte(hpte, HPTE_V_HVLOCK))
 		cpu_relax();
-	if ((hpte[0] & (HPTE_V_ABSENT | HPTE_V_VALID)) == 0 ||
-	    ((flags & H_AVPN) && (hpte[0] & ~0x7fUL) != avpn) ||
-	    ((flags & H_ANDCOND) && (hpte[0] & avpn) != 0)) {
-		hpte[0] &= ~HPTE_V_HVLOCK;
+	pte = be64_to_cpu(hpte[0]);
+	if ((pte & (HPTE_V_ABSENT | HPTE_V_VALID)) == 0 ||
+	    ((flags & H_AVPN) && (pte & ~0x7fUL) != avpn) ||
+	    ((flags & H_ANDCOND) && (pte & avpn) != 0)) {
+		hpte[0] &= ~cpu_to_be64(HPTE_V_HVLOCK);
 		return H_NOT_FOUND;
 	}
 
 	rev = real_vmalloc_addr(&kvm->arch.revmap[pte_index]);
-	v = hpte[0] & ~HPTE_V_HVLOCK;
+	v = pte & ~HPTE_V_HVLOCK;
 	if (v & HPTE_V_VALID) {
-		hpte[0] &= ~HPTE_V_VALID;
-		rb = compute_tlbie_rb(v, hpte[1], pte_index);
+		u64 pte1;
+
+		pte1 = be64_to_cpu(hpte[1]);
+		hpte[0] &= ~cpu_to_be64(HPTE_V_VALID);
+		rb = compute_tlbie_rb(v, pte1, pte_index);
 		do_tlbies(kvm, &rb, 1, global_invalidates(kvm, flags), true);
 		/* Read PTE low word after tlbie to get final R/C values */
-		remove_revmap_chain(kvm, pte_index, rev, v, hpte[1]);
+		remove_revmap_chain(kvm, pte_index, rev, v, pte1);
 	}
 	r = rev->guest_rpte & ~HPTE_GR_RESERVED;
 	note_hpte_modification(kvm, rev);
 	unlock_hpte(hpte, 0);
 
-	hpret[0] = v;
-	hpret[1] = r;
+	hpret[0] = cpu_to_be64(v);
+	hpret[1] = cpu_to_be64(r);
 	return H_SUCCESS;
 }
 EXPORT_SYMBOL_GPL(kvmppc_do_h_remove);
@@ -520,6 +530,7 @@  long kvmppc_h_bulk_remove(struct kvm_vcpu *vcpu)
 	int global;
 	long int ret = H_SUCCESS;
 	struct revmap_entry *rev, *revs[4];
+	u64 hp0;
 
 	global = global_invalidates(kvm, 0);
 	for (i = 0; i < 4 && ret == H_SUCCESS; ) {
@@ -552,23 +563,24 @@  long kvmppc_h_bulk_remove(struct kvm_vcpu *vcpu)
 					cpu_relax();
 			}
 			found = 0;
-			if (hp[0] & (HPTE_V_ABSENT | HPTE_V_VALID)) {
+			hp0 = be64_to_cpu(hp[0]);
+			if (hp0 & (HPTE_V_ABSENT | HPTE_V_VALID)) {
 				switch (flags & 3) {
 				case 0:		/* absolute */
 					found = 1;
 					break;
 				case 1:		/* andcond */
-					if (!(hp[0] & args[j + 1]))
+					if (!(hp0 & args[j + 1]))
 						found = 1;
 					break;
 				case 2:		/* AVPN */
-					if ((hp[0] & ~0x7fUL) == args[j + 1])
+					if ((hp0 & ~0x7fUL) == args[j + 1])
 						found = 1;
 					break;
 				}
 			}
 			if (!found) {
-				hp[0] &= ~HPTE_V_HVLOCK;
+				hp[0] &= ~cpu_to_be64(HPTE_V_HVLOCK);
 				args[j] = ((0x90 | flags) << 56) + pte_index;
 				continue;
 			}
@@ -577,7 +589,7 @@  long kvmppc_h_bulk_remove(struct kvm_vcpu *vcpu)
 			rev = real_vmalloc_addr(&kvm->arch.revmap[pte_index]);
 			note_hpte_modification(kvm, rev);
 
-			if (!(hp[0] & HPTE_V_VALID)) {
+			if (!(hp0 & HPTE_V_VALID)) {
 				/* insert R and C bits from PTE */
 				rcbits = rev->guest_rpte & (HPTE_R_R|HPTE_R_C);
 				args[j] |= rcbits << (56 - 5);
@@ -585,8 +597,10 @@  long kvmppc_h_bulk_remove(struct kvm_vcpu *vcpu)
 				continue;
 			}
 
-			hp[0] &= ~HPTE_V_VALID;		/* leave it locked */
-			tlbrb[n] = compute_tlbie_rb(hp[0], hp[1], pte_index);
+			/* leave it locked */
+			hp[0] &= ~cpu_to_be64(HPTE_V_VALID);
+			tlbrb[n] = compute_tlbie_rb(be64_to_cpu(hp[0]),
+				be64_to_cpu(hp[1]), pte_index);
 			indexes[n] = j;
 			hptes[n] = hp;
 			revs[n] = rev;
@@ -605,7 +619,8 @@  long kvmppc_h_bulk_remove(struct kvm_vcpu *vcpu)
 			pte_index = args[j] & ((1ul << 56) - 1);
 			hp = hptes[k];
 			rev = revs[k];
-			remove_revmap_chain(kvm, pte_index, rev, hp[0], hp[1]);
+			remove_revmap_chain(kvm, pte_index, rev,
+				be64_to_cpu(hp[0]), be64_to_cpu(hp[1]));
 			rcbits = rev->guest_rpte & (HPTE_R_R|HPTE_R_C);
 			args[j] |= rcbits << (56 - 5);
 			hp[0] = 0;
@@ -623,6 +638,7 @@  long kvmppc_h_protect(struct kvm_vcpu *vcpu, unsigned long flags,
 	unsigned long *hpte;
 	struct revmap_entry *rev;
 	unsigned long v, r, rb, mask, bits;
+	u64 pte;
 
 	if (pte_index >= kvm->arch.hpt_npte)
 		return H_PARAMETER;
@@ -630,13 +646,14 @@  long kvmppc_h_protect(struct kvm_vcpu *vcpu, unsigned long flags,
 	hpte = (unsigned long *)(kvm->arch.hpt_virt + (pte_index << 4));
 	while (!try_lock_hpte(hpte, HPTE_V_HVLOCK))
 		cpu_relax();
-	if ((hpte[0] & (HPTE_V_ABSENT | HPTE_V_VALID)) == 0 ||
-	    ((flags & H_AVPN) && (hpte[0] & ~0x7fUL) != avpn)) {
-		hpte[0] &= ~HPTE_V_HVLOCK;
+	pte = be64_to_cpu(hpte[0]);
+	if ((pte & (HPTE_V_ABSENT | HPTE_V_VALID)) == 0 ||
+	    ((flags & H_AVPN) && (pte & ~0x7fUL) != avpn)) {
+		hpte[0] &= ~cpu_to_be64(HPTE_V_HVLOCK);
 		return H_NOT_FOUND;
 	}
 
-	v = hpte[0];
+	v = pte;
 	bits = (flags << 55) & HPTE_R_PP0;
 	bits |= (flags << 48) & HPTE_R_KEY_HI;
 	bits |= flags & (HPTE_R_PP | HPTE_R_N | HPTE_R_KEY_LO);
@@ -650,12 +667,12 @@  long kvmppc_h_protect(struct kvm_vcpu *vcpu, unsigned long flags,
 		rev->guest_rpte = r;
 		note_hpte_modification(kvm, rev);
 	}
-	r = (hpte[1] & ~mask) | bits;
+	r = (be64_to_cpu(hpte[1]) & ~mask) | bits;
 
 	/* Update HPTE */
 	if (v & HPTE_V_VALID) {
 		rb = compute_tlbie_rb(v, r, pte_index);
-		hpte[0] = v & ~HPTE_V_VALID;
+		hpte[0] = cpu_to_be64(v & ~HPTE_V_VALID);
 		do_tlbies(kvm, &rb, 1, global_invalidates(kvm, flags), true);
 		/*
 		 * If the host has this page as readonly but the guest
@@ -681,9 +698,9 @@  long kvmppc_h_protect(struct kvm_vcpu *vcpu, unsigned long flags,
 			}
 		}
 	}
-	hpte[1] = r;
+	hpte[1] = cpu_to_be64(r);
 	eieio();
-	hpte[0] = v & ~HPTE_V_HVLOCK;
+	hpte[0] = cpu_to_be64(v & ~HPTE_V_HVLOCK);
 	asm volatile("ptesync" : : : "memory");
 	return H_SUCCESS;
 }
@@ -705,8 +722,8 @@  long kvmppc_h_read(struct kvm_vcpu *vcpu, unsigned long flags,
 	rev = real_vmalloc_addr(&kvm->arch.revmap[pte_index]);
 	for (i = 0; i < n; ++i, ++pte_index) {
 		hpte = (unsigned long *)(kvm->arch.hpt_virt + (pte_index << 4));
-		v = hpte[0] & ~HPTE_V_HVLOCK;
-		r = hpte[1];
+		v = be64_to_cpu(hpte[0]) & ~HPTE_V_HVLOCK;
+		r = be64_to_cpu(hpte[1]);
 		if (v & HPTE_V_ABSENT) {
 			v &= ~HPTE_V_ABSENT;
 			v |= HPTE_V_VALID;
@@ -726,8 +743,9 @@  void kvmppc_invalidate_hpte(struct kvm *kvm, unsigned long *hptep,
 {
 	unsigned long rb;
 
-	hptep[0] &= ~HPTE_V_VALID;
-	rb = compute_tlbie_rb(hptep[0], hptep[1], pte_index);
+	hptep[0] &= ~cpu_to_be64(HPTE_V_VALID);
+	rb = compute_tlbie_rb(be64_to_cpu(hptep[0]), be64_to_cpu(hptep[1]),
+			      pte_index);
 	do_tlbies(kvm, &rb, 1, 1, true);
 }
 EXPORT_SYMBOL_GPL(kvmppc_invalidate_hpte);
@@ -738,8 +756,9 @@  void kvmppc_clear_ref_hpte(struct kvm *kvm, unsigned long *hptep,
 	unsigned long rb;
 	unsigned char rbyte;
 
-	rb = compute_tlbie_rb(hptep[0], hptep[1], pte_index);
-	rbyte = (hptep[1] & ~HPTE_R_R) >> 8;
+	rb = compute_tlbie_rb(be64_to_cpu(hptep[0]), be64_to_cpu(hptep[1]),
+			      pte_index);
+	rbyte = (be64_to_cpu(hptep[1]) & ~HPTE_R_R) >> 8;
 	/* modify only the second-last byte, which contains the ref bit */
 	*((char *)hptep + 14) = rbyte;
 	do_tlbies(kvm, &rb, 1, 1, false);
@@ -801,7 +820,7 @@  long kvmppc_hv_find_lock_hpte(struct kvm *kvm, gva_t eaddr, unsigned long slb_v,
 
 		for (i = 0; i < 16; i += 2) {
 			/* Read the PTE racily */
-			v = hpte[i] & ~HPTE_V_HVLOCK;
+			v = be64_to_cpu(hpte[i]) & ~HPTE_V_HVLOCK;
 
 			/* Check valid/absent, hash, segment size and AVPN */
 			if (!(v & valid) || (v & mask) != val)
@@ -810,8 +829,8 @@  long kvmppc_hv_find_lock_hpte(struct kvm *kvm, gva_t eaddr, unsigned long slb_v,
 			/* Lock the PTE and read it under the lock */
 			while (!try_lock_hpte(&hpte[i], HPTE_V_HVLOCK))
 				cpu_relax();
-			v = hpte[i] & ~HPTE_V_HVLOCK;
-			r = hpte[i+1];
+			v = be64_to_cpu(hpte[i]) & ~HPTE_V_HVLOCK;
+			r = be64_to_cpu(hpte[i+1]);
 
 			/*
 			 * Check the HPTE again, including large page size
@@ -825,7 +844,7 @@  long kvmppc_hv_find_lock_hpte(struct kvm *kvm, gva_t eaddr, unsigned long slb_v,
 				return (hash << 3) + (i >> 1);
 
 			/* Unlock and move on */
-			hpte[i] = v;
+			hpte[i] = cpu_to_be64(v);
 		}
 
 		if (val & HPTE_V_SECONDARY)
@@ -871,8 +890,8 @@  long kvmppc_hpte_hv_fault(struct kvm_vcpu *vcpu, unsigned long addr,
 		return 0;		/* for prot fault, HPTE disappeared */
 	}
 	hpte = (unsigned long *)(kvm->arch.hpt_virt + (index << 4));
-	v = hpte[0] & ~HPTE_V_HVLOCK;
-	r = hpte[1];
+	v = be64_to_cpu(hpte[0]) & ~HPTE_V_HVLOCK;
+	r = be64_to_cpu(hpte[1]);
 	rev = real_vmalloc_addr(&kvm->arch.revmap[index]);
 	gr = rev->guest_rpte;