To be able to run a guest, we also need to implement a guest MMU.
This patch adds MMU handling for PPC64 guests.
Signed-off-by: Alexander Graf <ag...@suse.de>
---
arch/powerpc/kvm/970_mmu.c | 466 ++++++++++++++++++++++++++++++++++++++++++++
1 files changed, 466 insertions(+), 0 deletions(-)
create mode 100644 arch/powerpc/kvm/970_mmu.c
diff --git a/arch/powerpc/kvm/970_mmu.c b/arch/powerpc/kvm/970_mmu.c
new file mode 100644
index 0000000..ce693e2
--- /dev/null
+++ b/arch/powerpc/kvm/970_mmu.c
@@ -0,0 +1,466 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
+ *
+ * Copyright SUSE Linux Products GmbH 2009
+ *
+ * Authors: Alexander Graf <ag...@suse.de>
+ */
+
+#include <linux/types.h>
+#include <linux/string.h>
+#include <linux/kvm.h>
+#include <linux/kvm_host.h>
+#include <linux/highmem.h>
+
+#include <asm/tlbflush.h>
+#include <asm/kvm_ppc.h>
+#include <asm/kvm_970.h>
+
+// #define DEBUG_MMU
+
+static void kvmppc_mmu_970_reset_msr(struct kvm_vcpu *vcpu)
+{
+ kvmppc_set_msr(vcpu, MSR_SF);
+}
+
+static struct kvmppc_slb *kvmppc_mmu_970_find_slbe(struct kvmppc_vcpu_970
*vcpu_970,
+ gva_t eaddr)
+{
+ int i;
+ u64 esid = GET_ESID(eaddr);
+ u64 esid_1t = GET_ESID_1T(eaddr);
+
+ for (i = 0; i < vcpu_970->slb_nr; i++) {
+ u64 cmp_esid = esid;
+
+ if (!vcpu_970->slb[i].valid)
+ continue;
+
+ if (vcpu_970->slb[i].large)
+ cmp_esid = esid_1t;
+
+ if (vcpu_970->slb[i].esid == cmp_esid)
+ return &vcpu_970->slb[i];
+ }
+
+#ifdef DEBUG_MMU
+ printk(KERN_ERR "KVM: No SLB entry found for 0x%lx [%x | %x]\n", eaddr,
esid, esid_1t);
+ for (i = 0; i < vcpu_970->slb_nr; i++) {
+ if (vcpu_970->slb[i].vsid)
+ printk(KERN_ERR " %d: %c%c %lx %lx\n", i,
vcpu_970->slb[i].valid ? 'v' : ' ',
+ vcpu_970->slb[i].large
? 'l' : ' ',
+ vcpu_970->slb[i].esid,
+ vcpu_970->slb[i].vsid);
+ }
+#endif
+
+ return NULL;
+}
+
+static u64 kvmppc_mmu_970_ea_to_vp(struct kvm_vcpu *vcpu, gva_t eaddr, bool
data)
+{
+ struct kvmppc_slb *slb = kvmppc_mmu_970_find_slbe(to_970(vcpu), eaddr);
+
+ if (!slb)
+ return 0;
+
+ if (slb->large)
+ return (((u64)eaddr >> 12) & 0xfffffff) | (((u64)slb->vsid) <<
28);
+
+ return (((u64)eaddr >> 12) & 0xffff) | (((u64)slb->vsid) << 16);
+}
+
+static int kvmppc_mmu_970_get_pagesize(struct kvmppc_slb *slbe)
+{
+ return slbe->large ? 24 : 12;
+}
+
+static u32 kvmppc_mmu_970_get_page(struct kvmppc_slb *slbe, gva_t eaddr)
+{
+ int p = kvmppc_mmu_970_get_pagesize(slbe);
+ return ((eaddr & 0xfffffff) >> p);
+}
+
+static hva_t kvmppc_mmu_970_get_pteg(struct kvmppc_vcpu_970 *vcpu_970,
+ struct kvmppc_slb *slbe, gva_t eaddr,
+ bool second)
+{
+ u64 hash, pteg, htabsize;
+ u32 page;
+ hva_t r;
+
+ page = kvmppc_mmu_970_get_page(slbe, eaddr);
+ htabsize = ((1 << ((vcpu_970->sdr1 & 0x1f) + 11)) - 1);
+
+ hash = slbe->vsid ^ page;
+ if (second)
+ hash = ~hash;
+ hash &= ((1ULL << 39ULL) - 1ULL);
+ hash &= htabsize;
+ hash <<= 7ULL;
+
+ pteg = vcpu_970->sdr1 & 0xfffffffffffc0000ULL;
+ pteg |= hash;
+
+#ifdef DEBUG_MMU
+ printk(KERN_INFO "MMU: page=0x%x sdr1=0x%lx pteg=0x%lx vsid=0x%lx\n",
page, vcpu_970->sdr1, pteg, slbe->vsid);
+#endif
+
+ r = gfn_to_hva(vcpu_970->vcpu.kvm, pteg >> PAGE_SHIFT);
+ if (kvm_is_error_hva(r))
+ return r;
+ return r | (pteg & ~PAGE_MASK);
+}
+
+static u64 kvmppc_mmu_970_get_avpn(struct kvmppc_slb *slbe, gva_t eaddr)
+{
+ int p = kvmppc_mmu_970_get_pagesize(slbe);
+ u64 avpn;
+
+ avpn = kvmppc_mmu_970_get_page(slbe, eaddr);
+ avpn |= slbe->vsid << (28 - p);
+
+ if (p < 24)
+ avpn >>= ((80 - p) - 56) - 8;
+ else
+ avpn <<= 8;
+
+ return avpn;
+}
+
+static int kvmppc_mmu_970_xlate(struct kvm_vcpu *vcpu, gva_t eaddr,
+ struct kvmppc_pte *gpte, bool data)
+{
+ struct kvmppc_vcpu_970 *vcpu_970 = to_970(vcpu);
+ struct kvmppc_slb *slbe;
+ hva_t ptegp;
+ u64 pteg[16];
+ u64 avpn = 0;
+ int i;
+ u8 key = 0;
+ bool found = false;
+ bool perm_err = false;
+ int second = 0;
+
+ kvmppc_switch_context(vcpu, CONTEXT_HOST);
+
+ slbe = kvmppc_mmu_970_find_slbe(vcpu_970, eaddr);
+ if (!slbe)
+ goto no_seg_found;
+
+do_second:
+ ptegp = kvmppc_mmu_970_get_pteg(vcpu_970, slbe, eaddr, second);
+ if (kvm_is_error_hva(ptegp))
+ goto no_page_found;
+
+ avpn = kvmppc_mmu_970_get_avpn(slbe, eaddr);
+
+ if(copy_from_user(pteg, (void __user *)ptegp, sizeof(pteg))) {
+ printk(KERN_ERR "KVM can't copy data from 0x%lx!\n", ptegp);
+ goto no_page_found;
+ }
+
+ if ((vcpu->arch.msr & MSR_PR) && slbe->Kp)
+ key = 4;
+ else if (!(vcpu->arch.msr & MSR_PR) && slbe->Ks)
+ key = 4;
+
+ for (i=0; i<16; i+=2) {
+ u64 v = pteg[i];
+ u64 r = pteg[i+1];
+
+ // Valid check
+ if (!(v & HPTE_V_VALID))
+ continue;
+ // Hash check
+ if ((v & HPTE_V_SECONDARY) != second)
+ continue;
+
+ // AVPN compare
+ if (HPTE_V_AVPN_VAL(avpn) == HPTE_V_AVPN_VAL(v)) {
+ u8 pp = (r & HPTE_R_PP) | key;
+ int eaddr_mask = 0xFFF;
+
+ gpte->eaddr = eaddr;
+ gpte->vpage = kvmppc_mmu_970_ea_to_vp(vcpu, eaddr,
data);
+ if (slbe->large)
+ eaddr_mask = 0xFFFFFF;
+ gpte->raddr = (r & HPTE_R_RPN) | (eaddr & eaddr_mask);
+ gpte->may_execute = ((r & HPTE_R_N) ? false : true);
+ gpte->may_read = false;
+ gpte->may_write = false;
+
+ switch (pp) {
+ case 0:
+ case 1:
+ case 2:
+ case 6:
+ gpte->may_write = true;
+ /* fall through */
+ case 3:
+ case 5:
+ case 7:
+ gpte->may_read = true;
+ break;
+ }
+
+ if (!gpte->may_read) {
+ perm_err = true;
+ continue;
+ }
+#ifdef DEBUG_MMU
+ printk(KERN_ERR "KVM MMU: Translated 0x%lx [0x%lx] ->
0x%lx -> 0x%lx\n",
+ eaddr, avpn, gpte->vpage, gpte->raddr);
+#endif
+ found = true;
+ break;
+ }
+ }
+
+ // Update PTE R and C bits, so the guest's swapper knows we used the
page
+ if (found) {
+ u32 oldr = pteg[i+1];
+
+ if (gpte->may_read)
+ pteg[i+1] |= HPTE_R_R; // Accessed flag
+ if (gpte->may_write)
+ pteg[i+1] |= HPTE_R_C; // Dirty flag
+#ifdef DEBUG_MMU_PTE
+ else
+ printk(KERN_INFO "KVM: Mapping read-only page!\n");
+#endif
+
+ // Write back into the PTEG
+ if (pteg[i+1] != oldr)
+ copy_to_user((void __user *)ptegp, pteg, sizeof(pteg));
+
+ return 0;
+ } else {
+#ifdef DEBUG_MMU
+ printk(KERN_ERR "KVM MMU: No PTE found (ea=0x%llx sdr1=0x%lx
ptegp=0x%lx)\n",
+ eaddr, to_970(vcpu)->sdr1, ptegp);
+ for (i=0; i<16; i+=2)
+ printk(KERN_ERR " %02d: 0x%lx - 0x%lx (0x%lx)\n",
+ i, pteg[i], pteg[i+1], avpn);
+#endif
+
+ if (!second) {
+ second = HPTE_V_SECONDARY;
+ goto do_second;
+ }
+ }
+
+
+no_page_found:
+
+
+ if (perm_err)
+ return -EPERM;
+
+ return -ENOENT;
+
+no_seg_found:
+
+#ifdef DEBUG_MMU
+ printk(KERN_ERR "KVM MMU: Trigger segment fault\n");
+#endif
+ return -EINVAL;
+}
+
+static void kvmppc_mmu_970_slbmte(struct kvm_vcpu *vcpu, u64 rs, u64 rb)
+{
+ struct kvmppc_vcpu_970 *vcpu_970;
+ u64 esid, esid_1t;
+ int slb_nr;
+ struct kvmppc_slb *slbe;
+
+#ifdef DEBUG_MMU
+ printk(KERN_ERR "KVM MMU: slbmte(0x%lx, 0x%lx)\n", rs, rb);
+#endif
+
+ vcpu_970 = to_970(vcpu);
+
+ esid = GET_ESID(rb);
+ esid_1t = GET_ESID_1T(rb);
+ slb_nr = rb & 0xfff;
+
+ if (slb_nr > vcpu_970->slb_nr)
+ return;
+
+ slbe = &vcpu_970->slb[slb_nr];
+
+ slbe->large = (rs & SLB_VSID_L) ? 1 : 0;
+ slbe->esid = slbe->large ? esid_1t : esid;
+ slbe->vsid = rs >> 12;
+ slbe->valid = (rb & SLB_ESID_V) ? 1 : 0;
+ slbe->Ks = (rs & SLB_VSID_KS) ? 1 : 0;
+ slbe->Kp = (rs & SLB_VSID_KP) ? 1 : 0;
+ slbe->nx = (rs & SLB_VSID_N) ? 1 : 0;
+ slbe->class = (rs & SLB_VSID_C) ? 1 : 0;
+
+ slbe->orige = rb & (ESID_MASK | SLB_ESID_V);
+ slbe->origv = rs;
+
+ /* Map the new segment */
+ kvmppc_mmu_map_segment(vcpu, esid << SID_SHIFT);
+}
+
+static u64 kvmppc_mmu_970_slbmfee(struct kvm_vcpu *vcpu, u64 slb_nr)
+{
+ struct kvmppc_vcpu_970 *vcpu_970 = to_970(vcpu);
+ struct kvmppc_slb *slbe;
+
+ if (slb_nr > vcpu_970->slb_nr)
+ return 0;
+
+ slbe = &vcpu_970->slb[slb_nr];
+
+ return slbe->orige;
+}
+
+static u64 kvmppc_mmu_970_slbmfev(struct kvm_vcpu *vcpu, u64 slb_nr)
+{
+ struct kvmppc_vcpu_970 *vcpu_970 = to_970(vcpu);
+ struct kvmppc_slb *slbe;
+
+ if (slb_nr > vcpu_970->slb_nr)
+ return 0;
+
+ slbe = &vcpu_970->slb[slb_nr];
+
+ return slbe->origv;
+}
+
+static void kvmppc_mmu_970_slbie(struct kvm_vcpu *vcpu, u64 ea)
+{
+ struct kvmppc_vcpu_970 *vcpu_970 = to_970(vcpu);
+ struct kvmppc_slb *slbe;
+
+#ifdef DEBUG_MMU
+ printk(KERN_ERR "KVM MMU: slbie(0x%lx)\n", ea);
+#endif
+
+ slbe = kvmppc_mmu_970_find_slbe(vcpu_970, ea);
+
+ if (!slbe)
+ return;
+
+#ifdef DEBUG_MMU
+ printk(KERN_ERR "KVM MMU: slbie(0x%lx, 0x%lx)\n", ea, slbe->esid);
+#endif
+
+ slbe->valid = false;
+ kvmppc_mmu_pte_flush(vcpu, ea, ESID_MASK);
+}
+
+static void kvmppc_mmu_970_slbia(struct kvm_vcpu *vcpu)
+{
+ struct kvmppc_vcpu_970 *vcpu_970 = to_970(vcpu);
+ int i;
+
+#ifdef DEBUG_MMU
+ printk(KERN_ERR "KVM MMU: slbia()\n");
+#endif
+ for (i = 1; i < vcpu_970->slb_nr; i++)
+ vcpu_970->slb[i].valid = false;
+
+ kvmppc_mmu_pte_flush(vcpu, 0, 0);
+}
+
+static void kvmppc_mmu_970_mtsrin(struct kvm_vcpu *vcpu, u32 srnum, ulong
value)
+{
+ u64 rb = 0, rs = 0;
+
+ /* ESID = srnum */
+ rb |= (srnum & 0xf) << 28;
+ /* Set the valid bit */
+ rb |= 1 << 27;
+ /* Index = ESID */
+ rb |= srnum;
+
+ /* VSID = VSID */
+ rs |= (value & 0xfffffff) << 12;
+ /* flags = flags */
+ rs |= ((value >> 27) & 0xf) << 9;
+
+ kvmppc_mmu_970_slbmte(vcpu, rs, rb);
+}
+
+static void kvmppc_mmu_970_tlbie(struct kvm_vcpu *vcpu, ulong va, bool large)
+{
+ u64 mask = 0xFFFFFFFFFULL;
+#ifdef DEBUG_MMU
+ printk(KERN_ERR "KVM MMU: tlbie(0x%lx)\n", va);
+#endif
+ if (large)
+ mask = 0xFFFFFF000ULL;
+ kvmppc_mmu_pte_vflush(vcpu, va >> 12, mask);
+}
+
+static u64 kvmppc_mmu_970_esid_to_vsid(struct kvm_vcpu *vcpu, u64 esid)
+{
+ u64 r = 0;
+
+ switch (vcpu->arch.msr & (MSR_DR|MSR_IR)) {
+ case 0:
+ r = (VSID_REAL >> 16) | esid;
+ break;
+ case MSR_IR:
+ r = (VSID_REAL_IR >> 16) | esid;
+ break;
+ case MSR_DR:
+ r = (VSID_REAL_DR >> 16) | esid;
+ break;
+ case MSR_DR|MSR_IR:
+ {
+ ulong ea;
+ struct kvmppc_slb *slb;
+ ea = esid << SID_SHIFT;
+ slb = kvmppc_mmu_970_find_slbe(to_970(vcpu), ea);
+ if (slb)
+ r = slb->vsid;
+
+ break;
+ }
+ default:
+ BUG();
+ break;
+ }
+
+ return r;
+}
+
+static bool kvmppc_mmu_970_is_dcbz32(struct kvm_vcpu *vcpu)
+{
+ return (to_970(vcpu)->hid[5] & 0x80);
+}
+
+void kvmppc_mmu_970_init(struct kvm_vcpu *vcpu)
+{
+ struct kvmppc_mmu *mmu = &vcpu->arch.mmu;
+
+ mmu->mfsrin = NULL;
+ mmu->mtsrin = kvmppc_mmu_970_mtsrin;
+ mmu->slbmte = kvmppc_mmu_970_slbmte;
+ mmu->slbmfee = kvmppc_mmu_970_slbmfee;
+ mmu->slbmfev = kvmppc_mmu_970_slbmfev;
+ mmu->slbie = kvmppc_mmu_970_slbie;
+ mmu->slbia = kvmppc_mmu_970_slbia;
+ mmu->xlate = kvmppc_mmu_970_xlate;
+ mmu->reset_msr = kvmppc_mmu_970_reset_msr;
+ mmu->tlbie = kvmppc_mmu_970_tlbie;
+ mmu->esid_to_vsid = kvmppc_mmu_970_esid_to_vsid;
+ mmu->ea_to_vp = kvmppc_mmu_970_ea_to_vp;
+ mmu->is_dcbz32 = kvmppc_mmu_970_is_dcbz32;
+}
--
1.6.0.2
--
To unsubscribe from this list: send the line "unsubscribe kvm-ppc" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
