Merge tag 'pull-ppc-20230626' of https://github.com/legoater/qemu into staging

ppc queue:

* New maintainers
* Nested implementation cleanups
* Various cleanups of the CPU implementation
* SMT support for pseries
* Improvements of the XIVE2 TIMA modeling
* Extra avocado tests for pseries

# -----BEGIN PGP SIGNATURE-----
#
# iQIzBAABCAAdFiEEoPZlSPBIlev+awtgUaNDx8/77KEFAmSZKF8ACgkQUaNDx8/7
# 7KGSiBAAlHC4S9J5ujzTIojaWY72d2ZinkC+WpBus9Wr91DqaUSUd/JbzDxQCvXh
# dBWEbcyQ+abb8M3OQ3fMq9TfD2/LhxxXb+uwHIJ+ylITBnsRVCQv/4/gi3EkpRid
# h4q3wYH8OYNfCQd/cWYXNgCSNj1nS9sRrEKFXaB0JeQWHzHxriJS/SoIhilqvUru
# LFEytWNb3bxRkEkt8oAetOa9+DNLowUQ9IdzswqGcib09po3b1k4+ThfcvzU9nAc
# ek31/h1W6cJbOJcgRO2dhWUZYp7cfmcnOa02E84tGFvvY/kYbjzPZZnoniSXD4uf
# YWFCoB3VxUoZ/YKCT/pDKHVdXmLLrfckNbo9vQNEcwmjr8m0Q3d1ewD5O9oNRpgN
# H0QMENfsdojztosOm3KPQ20aqNf1R7rQegYTiWf3B2fKZ6PIqnn3tBPxaEDkH7NC
# GTAKnBhF48lcHSF/4XOfGdmqhGgPRWX/Tv0wia7RY/A4NEfiIImIu+nYSGNBbu3y
# 7xlmtcumTlsRityOZnYI3bN5ubv++XPwU5NIJPACqvAbhif2rf1vQ9rMkkK785GL
# ciJ/5f6zXsLU7DfWP+qbTBizchQgigXnRZEEc7Seo6Bwtru22oxug0qQZ5QCgyXl
# Fg5Xuoq/6T4JC75pvxh1BjVlZc3Okzbfmsj+aZNrXO581HVJ2JI=
# =XLtJ
# -----END PGP SIGNATURE-----
# gpg: Signature made Mon 26 Jun 2023 07:55:43 AM CEST
# gpg:                using RSA key A0F66548F04895EBFE6B0B6051A343C7CFFBECA1
# gpg: Good signature from "Cédric Le Goater <clg@kaod.org>" [undefined]
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: A0F6 6548 F048 95EB FE6B  0B60 51A3 43C7 CFFB ECA1

* tag 'pull-ppc-20230626' of https://github.com/legoater/qemu: (30 commits)
  tests/avocado: ppc test VOF bios Linux boot
  pnv/xive2: Check TIMA special ops against a dedicated array for P10
  pnv/xive2: Add a get_config() method on the presenter class
  tests/avocado: Add ppc64 pseries multiprocessor boot tests
  tests/avocado: boot ppc64 pseries to Linux VFS mount
  spapr: TCG allow up to 8-thread SMT on POWER8 and newer CPUs
  hw/ppc/spapr: Test whether TCG is enabled with tcg_enabled()
  target/ppc: Add msgsnd/p and DPDES SMT support
  target/ppc: Add support for SMT CTRL register
  target/ppc: Add initial flags and helpers for SMT support
  target/ppc: Fix sc instruction handling of LEV field
  target/ppc: Better CTRL SPR implementation
  target/ppc: Add ISA v3.1 LEV indication in SRR1 for system call interrupts
  target/ppc: Implement HEIR SPR
  target/ppc: Add SRR1 prefix indication to interrupt handlers
  target/ppc: Change partition-scope translate interface
  target/ppc: Fix instruction loading endianness in alignment interrupt
  ppc/spapr: Move spapr nested HV to a new file
  ppc/spapr: load and store l2 state with helper functions
  ppc/spapr: Add a nested state struct
  ...

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

15 files changed, 547 insertions, 411 deletions

diff --git a/hw/intc/pnv_xive.c b/hw/intc/pnv_xive.c
index 622f9d2..e536b3e 100644
--- a/hw/intc/pnv_xive.c
+++ b/hw/intc/pnv_xive.c
@@ -479,6 +479,16 @@ static int pnv_xive_match_nvt(XivePresenter *xptr, uint8_t format,
     return count;
 }
 
+static uint32_t pnv_xive_presenter_get_config(XivePresenter *xptr)
+{
+    uint32_t cfg = 0;
+
+    /* TIMA GEN1 is all P9 knows */
+    cfg |= XIVE_PRESENTER_GEN1_TIMA_OS;
+
+    return cfg;
+}
+
 static uint8_t pnv_xive_get_block_id(XiveRouter *xrtr)
 {
     return pnv_xive_block_id(PNV_XIVE(xrtr));
@@ -1991,6 +2001,7 @@ static void pnv_xive_class_init(ObjectClass *klass, void *data)
 
     xnc->notify = pnv_xive_notify;
     xpc->match_nvt  = pnv_xive_match_nvt;
+    xpc->get_config = pnv_xive_presenter_get_config;
 };
 
 static const TypeInfo pnv_xive_info = {
diff --git a/hw/intc/pnv_xive2.c b/hw/intc/pnv_xive2.c
index ec1edeb..ed438a2 100644
--- a/hw/intc/pnv_xive2.c
+++ b/hw/intc/pnv_xive2.c
@@ -501,6 +501,17 @@ static int pnv_xive2_match_nvt(XivePresenter *xptr, uint8_t format,
     return count;
 }
 
+static uint32_t pnv_xive2_presenter_get_config(XivePresenter *xptr)
+{
+    PnvXive2 *xive = PNV_XIVE2(xptr);
+    uint32_t cfg = 0;
+
+    if (xive->cq_regs[CQ_XIVE_CFG >> 3] & CQ_XIVE_CFG_GEN1_TIMA_OS) {
+        cfg |= XIVE_PRESENTER_GEN1_TIMA_OS;
+    }
+    return cfg;
+}
+
 static uint8_t pnv_xive2_get_block_id(Xive2Router *xrtr)
 {
     return pnv_xive2_block_id(PNV_XIVE2(xrtr));
@@ -1645,17 +1656,6 @@ static const MemoryRegionOps pnv_xive2_ic_tm_indirect_ops = {
 /*
  * TIMA ops
  */
-
-/*
- * Special TIMA offsets to handle accesses in a POWER10 way.
- *
- * Only the CAM line updates done by the hypervisor should be handled
- * specifically.
- */
-#define HV_PAGE_OFFSET         (XIVE_TM_HV_PAGE << TM_SHIFT)
-#define HV_PUSH_OS_CTX_OFFSET  (HV_PAGE_OFFSET | (TM_QW1_OS + TM_WORD2))
-#define HV_PULL_OS_CTX_OFFSET  (HV_PAGE_OFFSET | TM_SPC_PULL_OS_CTX)
-
 static void pnv_xive2_tm_write(void *opaque, hwaddr offset,
                                uint64_t value, unsigned size)
 {
@@ -1663,18 +1663,7 @@ static void pnv_xive2_tm_write(void *opaque, hwaddr offset,
     PnvXive2 *xive = pnv_xive2_tm_get_xive(cpu);
     XiveTCTX *tctx = XIVE_TCTX(pnv_cpu_state(cpu)->intc);
     XivePresenter *xptr = XIVE_PRESENTER(xive);
-    bool gen1_tima_os =
-        xive->cq_regs[CQ_XIVE_CFG >> 3] & CQ_XIVE_CFG_GEN1_TIMA_OS;
-
-    offset &= TM_ADDRESS_MASK;
-
-    /* TODO: should we switch the TM ops table instead ? */
-    if (!gen1_tima_os && offset == HV_PUSH_OS_CTX_OFFSET) {
-        xive2_tm_push_os_ctx(xptr, tctx, offset, value, size);
-        return;
-    }
 
-    /* Other TM ops are the same as XIVE1 */
     xive_tctx_tm_write(xptr, tctx, offset, value, size);
 }
 
@@ -1684,17 +1673,7 @@ static uint64_t pnv_xive2_tm_read(void *opaque, hwaddr offset, unsigned size)
     PnvXive2 *xive = pnv_xive2_tm_get_xive(cpu);
     XiveTCTX *tctx = XIVE_TCTX(pnv_cpu_state(cpu)->intc);
     XivePresenter *xptr = XIVE_PRESENTER(xive);
-    bool gen1_tima_os =
-        xive->cq_regs[CQ_XIVE_CFG >> 3] & CQ_XIVE_CFG_GEN1_TIMA_OS;
-
-    offset &= TM_ADDRESS_MASK;
-
-    /* TODO: should we switch the TM ops table instead ? */
-    if (!gen1_tima_os && offset == HV_PULL_OS_CTX_OFFSET) {
-        return xive2_tm_pull_os_ctx(xptr, tctx, offset, size);
-    }
 
-    /* Other TM ops are the same as XIVE1 */
     return xive_tctx_tm_read(xptr, tctx, offset, size);
 }
 
@@ -1987,6 +1966,7 @@ static void pnv_xive2_class_init(ObjectClass *klass, void *data)
     xnc->notify    = pnv_xive2_notify;
 
     xpc->match_nvt  = pnv_xive2_match_nvt;
+    xpc->get_config = pnv_xive2_presenter_get_config;
 };
 
 static const TypeInfo pnv_xive2_info = {
diff --git a/hw/intc/spapr_xive.c b/hw/intc/spapr_xive.c
index dc641cc..8bcab28 100644
--- a/hw/intc/spapr_xive.c
+++ b/hw/intc/spapr_xive.c
@@ -475,6 +475,21 @@ static int spapr_xive_match_nvt(XivePresenter *xptr, uint8_t format,
     return count;
 }
 
+static uint32_t spapr_xive_presenter_get_config(XivePresenter *xptr)
+{
+    uint32_t cfg = 0;
+
+    /*
+     * Let's claim GEN1 TIMA format. If running with KVM on P10, the
+     * correct answer is deep in the hardware and not accessible to
+     * us.  But it shouldn't matter as it only affects the presenter
+     * as seen by a guest OS.
+     */
+    cfg |= XIVE_PRESENTER_GEN1_TIMA_OS;
+
+    return cfg;
+}
+
 static uint8_t spapr_xive_get_block_id(XiveRouter *xrtr)
 {
     return SPAPR_XIVE_BLOCK_ID;
@@ -832,6 +847,7 @@ static void spapr_xive_class_init(ObjectClass *klass, void *data)
     sicc->post_load = spapr_xive_post_load;
 
     xpc->match_nvt  = spapr_xive_match_nvt;
+    xpc->get_config = spapr_xive_presenter_get_config;
     xpc->in_kernel  = spapr_xive_in_kernel_xptr;
 }
 
diff --git a/hw/intc/xive.c b/hw/intc/xive.c
index 5204c14..84c079b 100644
--- a/hw/intc/xive.c
+++ b/hw/intc/xive.c
@@ -20,6 +20,7 @@
 #include "monitor/monitor.h"
 #include "hw/irq.h"
 #include "hw/ppc/xive.h"
+#include "hw/ppc/xive2.h"
 #include "hw/ppc/xive_regs.h"
 #include "trace.h"
 
@@ -461,6 +462,13 @@ static void xive_tm_push_os_ctx(XivePresenter *xptr, XiveTCTX *tctx,
     }
 }
 
+static uint32_t xive_presenter_get_config(XivePresenter *xptr)
+{
+    XivePresenterClass *xpc = XIVE_PRESENTER_GET_CLASS(xptr);
+
+    return xpc->get_config(xptr);
+}
+
 /*
  * Define a mapping of "special" operations depending on the TIMA page
  * offset and the size of the operation.
@@ -497,14 +505,47 @@ static const XiveTmOp xive_tm_operations[] = {
     { XIVE_TM_HV_PAGE, TM_SPC_PULL_POOL_CTX,  8, NULL, xive_tm_pull_pool_ctx },
 };
 
-static const XiveTmOp *xive_tm_find_op(hwaddr offset, unsigned size, bool write)
+static const XiveTmOp xive2_tm_operations[] = {
+    /*
+     * MMIOs below 2K : raw values and special operations without side
+     * effects
+     */
+    { XIVE_TM_OS_PAGE, TM_QW1_OS + TM_CPPR,   1, xive_tm_set_os_cppr, NULL },
+    { XIVE_TM_HV_PAGE, TM_QW1_OS + TM_WORD2,  4, xive2_tm_push_os_ctx, NULL },
+    { XIVE_TM_HV_PAGE, TM_QW3_HV_PHYS + TM_CPPR, 1, xive_tm_set_hv_cppr, NULL },
+    { XIVE_TM_HV_PAGE, TM_QW3_HV_PHYS + TM_WORD2, 1, xive_tm_vt_push, NULL },
+    { XIVE_TM_HV_PAGE, TM_QW3_HV_PHYS + TM_WORD2, 1, NULL, xive_tm_vt_poll },
+
+    /* MMIOs above 2K : special operations with side effects */
+    { XIVE_TM_OS_PAGE, TM_SPC_ACK_OS_REG,     2, NULL, xive_tm_ack_os_reg },
+    { XIVE_TM_OS_PAGE, TM_SPC_SET_OS_PENDING, 1, xive_tm_set_os_pending, NULL },
+    { XIVE_TM_HV_PAGE, TM_SPC_PULL_OS_CTX,    4, NULL, xive2_tm_pull_os_ctx },
+    { XIVE_TM_HV_PAGE, TM_SPC_PULL_OS_CTX,    8, NULL, xive2_tm_pull_os_ctx },
+    { XIVE_TM_HV_PAGE, TM_SPC_ACK_HV_REG,     2, NULL, xive_tm_ack_hv_reg },
+    { XIVE_TM_HV_PAGE, TM_SPC_PULL_POOL_CTX,  4, NULL, xive_tm_pull_pool_ctx },
+    { XIVE_TM_HV_PAGE, TM_SPC_PULL_POOL_CTX,  8, NULL, xive_tm_pull_pool_ctx },
+};
+
+static const XiveTmOp *xive_tm_find_op(XivePresenter *xptr, hwaddr offset,
+                                       unsigned size, bool write)
 {
     uint8_t page_offset = (offset >> TM_SHIFT) & 0x3;
     uint32_t op_offset = offset & TM_ADDRESS_MASK;
-    int i;
+    const XiveTmOp *tm_ops;
+    int i, tm_ops_count;
+    uint32_t cfg;
+
+    cfg = xive_presenter_get_config(xptr);
+    if (cfg & XIVE_PRESENTER_GEN1_TIMA_OS) {
+        tm_ops = xive_tm_operations;
+        tm_ops_count = ARRAY_SIZE(xive_tm_operations);
+    } else {
+        tm_ops = xive2_tm_operations;
+        tm_ops_count = ARRAY_SIZE(xive2_tm_operations);
+    }
 
-    for (i = 0; i < ARRAY_SIZE(xive_tm_operations); i++) {
-        const XiveTmOp *xto = &xive_tm_operations[i];
+    for (i = 0; i < tm_ops_count; i++) {
+        const XiveTmOp *xto = &tm_ops[i];
 
         /* Accesses done from a more privileged TIMA page is allowed */
         if (xto->page_offset >= page_offset &&
@@ -535,7 +576,7 @@ void xive_tctx_tm_write(XivePresenter *xptr, XiveTCTX *tctx, hwaddr offset,
      * First, check for special operations in the 2K region
      */
     if (offset & TM_SPECIAL_OP) {
-        xto = xive_tm_find_op(offset, size, true);
+        xto = xive_tm_find_op(tctx->xptr, offset, size, true);
         if (!xto) {
             qemu_log_mask(LOG_GUEST_ERROR, "XIVE: invalid write access at TIMA "
                           "@%"HWADDR_PRIx"\n", offset);
@@ -548,7 +589,7 @@ void xive_tctx_tm_write(XivePresenter *xptr, XiveTCTX *tctx, hwaddr offset,
     /*
      * Then, for special operations in the region below 2K.
      */
-    xto = xive_tm_find_op(offset, size, true);
+    xto = xive_tm_find_op(tctx->xptr, offset, size, true);
     if (xto) {
         xto->write_handler(xptr, tctx, offset, value, size);
         return;
@@ -574,7 +615,7 @@ uint64_t xive_tctx_tm_read(XivePresenter *xptr, XiveTCTX *tctx, hwaddr offset,
      * First, check for special operations in the 2K region
      */
     if (offset & TM_SPECIAL_OP) {
-        xto = xive_tm_find_op(offset, size, false);
+        xto = xive_tm_find_op(tctx->xptr, offset, size, false);
         if (!xto) {
             qemu_log_mask(LOG_GUEST_ERROR, "XIVE: invalid read access to TIMA"
                           "@%"HWADDR_PRIx"\n", offset);
@@ -587,7 +628,7 @@ uint64_t xive_tctx_tm_read(XivePresenter *xptr, XiveTCTX *tctx, hwaddr offset,
     /*
      * Then, for special operations in the region below 2K.
      */
-    xto = xive_tm_find_op(offset, size, false);
+    xto = xive_tm_find_op(tctx->xptr, offset, size, false);
     if (xto) {
         ret = xto->read_handler(xptr, tctx, offset, size);
         goto out;
diff --git a/hw/pci-host/pnv_phb4.c b/hw/pci-host/pnv_phb4.c
index 542f9e2..6232cbe 100644
--- a/hw/pci-host/pnv_phb4.c
+++ b/hw/pci-host/pnv_phb4.c
@@ -133,13 +133,13 @@ static void pnv_phb4_rc_config_write(PnvPHB4 *phb, unsigned off,
     PCIDevice *pdev;
 
     if (size != 4) {
-        phb_error(phb, "rc_config_write invalid size %d\n", size);
+        phb_error(phb, "rc_config_write invalid size %d", size);
         return;
     }
 
     pdev = pci_find_device(pci->bus, 0, 0);
     if (!pdev) {
-        phb_error(phb, "rc_config_write device not found\n");
+        phb_error(phb, "rc_config_write device not found");
         return;
     }
 
@@ -155,13 +155,13 @@ static uint64_t pnv_phb4_rc_config_read(PnvPHB4 *phb, unsigned off,
     uint64_t val;
 
     if (size != 4) {
-        phb_error(phb, "rc_config_read invalid size %d\n", size);
+        phb_error(phb, "rc_config_read invalid size %d", size);
         return ~0ull;
     }
 
     pdev = pci_find_device(pci->bus, 0, 0);
     if (!pdev) {
-        phb_error(phb, "rc_config_read device not found\n");
+        phb_error(phb, "rc_config_read device not found");
         return ~0ull;
     }
 
@@ -1039,19 +1039,19 @@ static void pnv_pec_stk_nest_xscom_write(void *opaque, hwaddr addr,
         if (phb->nest_regs[PEC_NEST_STK_BAR_EN] &
             (PEC_NEST_STK_BAR_EN_MMIO0 |
              PEC_NEST_STK_BAR_EN_MMIO1)) {
-            phb_pec_error(pec, "Changing enabled BAR unsupported\n");
+            phb_pec_error(pec, "Changing enabled BAR unsupported");
         }
         phb->nest_regs[reg] = val & 0xffffffffff000000ull;
         break;
     case PEC_NEST_STK_PHB_REGS_BAR:
         if (phb->nest_regs[PEC_NEST_STK_BAR_EN] & PEC_NEST_STK_BAR_EN_PHB) {
-            phb_pec_error(pec, "Changing enabled BAR unsupported\n");
+            phb_pec_error(pec, "Changing enabled BAR unsupported");
         }
         phb->nest_regs[reg] = val & 0xffffffffffc00000ull;
         break;
     case PEC_NEST_STK_INT_BAR:
         if (phb->nest_regs[PEC_NEST_STK_BAR_EN] & PEC_NEST_STK_BAR_EN_INT) {
-            phb_pec_error(pec, "Changing enabled BAR unsupported\n");
+            phb_pec_error(pec, "Changing enabled BAR unsupported");
         }
         phb->nest_regs[reg] = val & 0xfffffff000000000ull;
         break;
diff --git a/hw/ppc/meson.build b/hw/ppc/meson.build
index c927337..a313d4b 100644
--- a/hw/ppc/meson.build
+++ b/hw/ppc/meson.build
@@ -15,6 +15,7 @@ ppc_ss.add(when: 'CONFIG_PSERIES', if_true: files(
   'spapr_vio.c',
   'spapr_events.c',
   'spapr_hcall.c',
+  'spapr_nested.c',
   'spapr_iommu.c',
   'spapr_rtas.c',
   'spapr_pci.c',
diff --git a/hw/ppc/pnv.c b/hw/ppc/pnv.c
index 590fc64..fc08317 100644
--- a/hw/ppc/pnv.c
+++ b/hw/ppc/pnv.c
@@ -799,7 +799,8 @@ static void pnv_init(MachineState *machine)
     DeviceState *dev;
 
     if (kvm_enabled()) {
-        error_report("The powernv machine does not work with KVM acceleration");
+        error_report("machine %s does not support the KVM accelerator",
+                     mc->name);
         exit(EXIT_FAILURE);
     }
 
diff --git a/hw/ppc/ppc.c b/hw/ppc/ppc.c
index 1b1220c..82e4408 100644
--- a/hw/ppc/ppc.c
+++ b/hw/ppc/ppc.c
@@ -1436,6 +1436,12 @@ int ppc_cpu_pir(PowerPCCPU *cpu)
     return env->spr_cb[SPR_PIR].default_value;
 }
 
+int ppc_cpu_tir(PowerPCCPU *cpu)
+{
+    CPUPPCState *env = &cpu->env;
+    return env->spr_cb[SPR_TIR].default_value;
+}
+
 PowerPCCPU *ppc_get_vcpu_by_pir(int pir)
 {
     CPUState *cs;
diff --git a/hw/ppc/ppc440_bamboo.c b/hw/ppc/ppc440_bamboo.c
index f969fa3..f061b8c 100644
--- a/hw/ppc/ppc440_bamboo.c
+++ b/hw/ppc/ppc440_bamboo.c
@@ -19,7 +19,6 @@
 #include "hw/pci/pci.h"
 #include "hw/boards.h"
 #include "sysemu/kvm.h"
-#include "kvm_ppc.h"
 #include "sysemu/device_tree.h"
 #include "hw/loader.h"
 #include "elf.h"
@@ -97,16 +96,6 @@ static int bamboo_load_device_tree(MachineState *machine,
         fprintf(stderr, "couldn't set /chosen/bootargs\n");
     }
 
-    /*
-     * Copy data from the host device tree into the guest. Since the guest can
-     * directly access the timebase without host involvement, we must expose
-     * the correct frequencies.
-     */
-    if (kvm_enabled()) {
-        tb_freq = kvmppc_get_tbfreq();
-        clock_freq = kvmppc_get_clockfreq();
-    }
-
     qemu_fdt_setprop_cell(fdt, "/cpus/cpu@0", "clock-frequency",
                           clock_freq);
     qemu_fdt_setprop_cell(fdt, "/cpus/cpu@0", "timebase-frequency",
@@ -175,6 +164,12 @@ static void bamboo_init(MachineState *machine)
     int success;
     int i;
 
+    if (kvm_enabled()) {
+        error_report("machine %s does not support the KVM accelerator",
+                     MACHINE_GET_CLASS(machine)->name);
+        exit(EXIT_FAILURE);
+    }
+
     cpu = POWERPC_CPU(cpu_create(machine->cpu_type));
     env = &cpu->env;
 
diff --git a/hw/ppc/prep.c b/hw/ppc/prep.c
index 33bf232..d9231c7 100644
--- a/hw/ppc/prep.c
+++ b/hw/ppc/prep.c
@@ -45,7 +45,6 @@
 #include "trace.h"
 #include "elf.h"
 #include "qemu/units.h"
-#include "kvm_ppc.h"
 
 /* SMP is not enabled, for now */
 #define MAX_CPUS 1
@@ -245,6 +244,12 @@ static void ibm_40p_init(MachineState *machine)
     long kernel_size = 0, initrd_size = 0;
     char boot_device;
 
+    if (kvm_enabled()) {
+        error_report("machine %s does not support the KVM accelerator",
+                     MACHINE_GET_CLASS(machine)->name);
+        exit(EXIT_FAILURE);
+    }
+
     /* init CPU */
     cpu = POWERPC_CPU(cpu_create(machine->cpu_type));
     env = &cpu->env;
@@ -392,18 +397,7 @@ static void ibm_40p_init(MachineState *machine)
     fw_cfg_add_i16(fw_cfg, FW_CFG_PPC_HEIGHT, graphic_height);
     fw_cfg_add_i16(fw_cfg, FW_CFG_PPC_DEPTH, graphic_depth);
 
-    fw_cfg_add_i32(fw_cfg, FW_CFG_PPC_IS_KVM, kvm_enabled());
-    if (kvm_enabled()) {
-        uint8_t *hypercall;
-
-        fw_cfg_add_i32(fw_cfg, FW_CFG_PPC_TBFREQ, kvmppc_get_tbfreq());
-        hypercall = g_malloc(16);
-        kvmppc_get_hypercall(env, hypercall, 16);
-        fw_cfg_add_bytes(fw_cfg, FW_CFG_PPC_KVM_HC, hypercall, 16);
-        fw_cfg_add_i32(fw_cfg, FW_CFG_PPC_KVM_PID, getpid());
-    } else {
-        fw_cfg_add_i32(fw_cfg, FW_CFG_PPC_TBFREQ, NANOSECONDS_PER_SECOND);
-    }
+    fw_cfg_add_i32(fw_cfg, FW_CFG_PPC_TBFREQ, NANOSECONDS_PER_SECOND);
     fw_cfg_add_i16(fw_cfg, FW_CFG_BOOT_DEVICE, boot_device);
     qemu_register_boot_set(fw_cfg_boot_set, fw_cfg);
 
diff --git a/hw/ppc/spapr.c b/hw/ppc/spapr.c
index dcb7f1c..54dbfd7 100644
--- a/hw/ppc/spapr.c
+++ b/hw/ppc/spapr.c
@@ -61,6 +61,7 @@
 
 #include "hw/ppc/fdt.h"
 #include "hw/ppc/spapr.h"
+#include "hw/ppc/spapr_nested.h"
 #include "hw/ppc/spapr_vio.h"
 #include "hw/ppc/vof.h"
 #include "hw/qdev-properties.h"
@@ -2524,10 +2525,19 @@ static void spapr_set_vsmt_mode(SpaprMachineState *spapr, Error **errp)
     int ret;
     unsigned int smp_threads = ms->smp.threads;
 
-    if (!kvm_enabled() && (smp_threads > 1)) {
-        error_setg(errp, "TCG cannot support more than 1 thread/core "
-                   "on a pseries machine");
-        return;
+    if (tcg_enabled()) {
+        if (smp_threads > 1 &&
+            !ppc_type_check_compat(ms->cpu_type, CPU_POWERPC_LOGICAL_2_07, 0,
+                                   spapr->max_compat_pvr)) {
+            error_setg(errp, "TCG only supports SMT on POWER8 or newer CPUs");
+            return;
+        }
+
+        if (smp_threads > 8) {
+            error_setg(errp, "TCG cannot support more than 8 threads/core "
+                       "on a pseries machine");
+            return;
+        }
     }
     if (!is_power_of_2(smp_threads)) {
         error_setg(errp, "Cannot support %d threads/core on a pseries "
diff --git a/hw/ppc/spapr_caps.c b/hw/ppc/spapr_caps.c
index 3fd45a6d..5a0755d 100644
--- a/hw/ppc/spapr_caps.c
+++ b/hw/ppc/spapr_caps.c
@@ -473,6 +473,20 @@ static void cap_nested_kvm_hv_apply(SpaprMachineState *spapr,
                 error_append_hint(errp,
                                   "Try appending -machine cap-nested-hv=off\n");
         }
+    } else if (tcg_enabled()) {
+        MachineState *ms = MACHINE(spapr);
+        unsigned int smp_threads = ms->smp.threads;
+
+        /*
+         * Nested-HV vCPU env state to L2, so SMT-shared SPR updates, for
+         * example, do not necessarily update the correct SPR value on sibling
+         * threads that are in a different guest/host context.
+         */
+        if (smp_threads > 1) {
+            error_setg(errp, "TCG does not support nested-HV with SMT");
+            error_append_hint(errp, "Try appending -machine cap-nested-hv=off "
+                                    "or use threads=1 with -smp\n");
+        }
     }
 }
 
diff --git a/hw/ppc/spapr_cpu_core.c b/hw/ppc/spapr_cpu_core.c
index 9b88dd5..a4e3c2f 100644
--- a/hw/ppc/spapr_cpu_core.c
+++ b/hw/ppc/spapr_cpu_core.c
@@ -255,7 +255,7 @@ static void spapr_cpu_core_unrealize(DeviceState *dev)
 }
 
 static bool spapr_realize_vcpu(PowerPCCPU *cpu, SpaprMachineState *spapr,
-                               SpaprCpuCore *sc, Error **errp)
+                               SpaprCpuCore *sc, int thread_index, Error **errp)
 {
     CPUPPCState *env = &cpu->env;
     CPUState *cs = CPU(cpu);
@@ -267,6 +267,9 @@ static bool spapr_realize_vcpu(PowerPCCPU *cpu, SpaprMachineState *spapr,
     cpu_ppc_set_vhyp(cpu, PPC_VIRTUAL_HYPERVISOR(spapr));
     kvmppc_set_papr(cpu);
 
+    env->spr_cb[SPR_PIR].default_value = cs->cpu_index;
+    env->spr_cb[SPR_TIR].default_value = thread_index;
+
     /* Set time-base frequency to 512 MHz. vhyp must be set first. */
     cpu_ppc_tb_init(env, SPAPR_TIMEBASE_FREQ);
 
@@ -337,7 +340,7 @@ static void spapr_cpu_core_realize(DeviceState *dev, Error **errp)
     for (i = 0; i < cc->nr_threads; i++) {
         sc->threads[i] = spapr_create_vcpu(sc, i, errp);
         if (!sc->threads[i] ||
-            !spapr_realize_vcpu(sc->threads[i], spapr, sc, errp)) {
+            !spapr_realize_vcpu(sc->threads[i], spapr, sc, i, errp)) {
             spapr_cpu_core_unrealize(dev);
             return;
         }
diff --git a/hw/ppc/spapr_hcall.c b/hw/ppc/spapr_hcall.c
index b904755..002ea0b 100644
--- a/hw/ppc/spapr_hcall.c
+++ b/hw/ppc/spapr_hcall.c
@@ -13,6 +13,7 @@
 #include "hw/ppc/ppc.h"
 #include "hw/ppc/spapr.h"
 #include "hw/ppc/spapr_cpu_core.h"
+#include "hw/ppc/spapr_nested.h"
 #include "mmu-hash64.h"
 #include "cpu-models.h"
 #include "trace.h"
@@ -1498,349 +1499,17 @@ target_ulong spapr_hypercall(PowerPCCPU *cpu, target_ulong opcode,
 }
 
 #ifdef CONFIG_TCG
-#define PRTS_MASK      0x1f
-
-static target_ulong h_set_ptbl(PowerPCCPU *cpu,
-                               SpaprMachineState *spapr,
-                               target_ulong opcode,
-                               target_ulong *args)
-{
-    target_ulong ptcr = args[0];
-
-    if (!spapr_get_cap(spapr, SPAPR_CAP_NESTED_KVM_HV)) {
-        return H_FUNCTION;
-    }
-
-    if ((ptcr & PRTS_MASK) + 12 - 4 > 12) {
-        return H_PARAMETER;
-    }
-
-    spapr->nested_ptcr = ptcr; /* Save new partition table */
-
-    return H_SUCCESS;
-}
-
-static target_ulong h_tlb_invalidate(PowerPCCPU *cpu,
-                                     SpaprMachineState *spapr,
-                                     target_ulong opcode,
-                                     target_ulong *args)
-{
-    /*
-     * The spapr virtual hypervisor nested HV implementation retains no L2
-     * translation state except for TLB. And the TLB is always invalidated
-     * across L1<->L2 transitions, so nothing is required here.
-     */
-
-    return H_SUCCESS;
-}
-
-static target_ulong h_copy_tofrom_guest(PowerPCCPU *cpu,
-                                        SpaprMachineState *spapr,
-                                        target_ulong opcode,
-                                        target_ulong *args)
-{
-    /*
-     * This HCALL is not required, L1 KVM will take a slow path and walk the
-     * page tables manually to do the data copy.
-     */
-    return H_FUNCTION;
-}
-
-/*
- * When this handler returns, the environment is switched to the L2 guest
- * and TCG begins running that. spapr_exit_nested() performs the switch from
- * L2 back to L1 and returns from the H_ENTER_NESTED hcall.
- */
-static target_ulong h_enter_nested(PowerPCCPU *cpu,
-                                   SpaprMachineState *spapr,
-                                   target_ulong opcode,
-                                   target_ulong *args)
-{
-    PowerPCCPUClass *pcc = POWERPC_CPU_GET_CLASS(cpu);
-    CPUState *cs = CPU(cpu);
-    CPUPPCState *env = &cpu->env;
-    SpaprCpuState *spapr_cpu = spapr_cpu_state(cpu);
-    target_ulong hv_ptr = args[0];
-    target_ulong regs_ptr = args[1];
-    target_ulong hdec, now = cpu_ppc_load_tbl(env);
-    target_ulong lpcr, lpcr_mask;
-    struct kvmppc_hv_guest_state *hvstate;
-    struct kvmppc_hv_guest_state hv_state;
-    struct kvmppc_pt_regs *regs;
-    hwaddr len;
-
-    if (spapr->nested_ptcr == 0) {
-        return H_NOT_AVAILABLE;
-    }
-
-    len = sizeof(*hvstate);
-    hvstate = address_space_map(CPU(cpu)->as, hv_ptr, &len, false,
-                                MEMTXATTRS_UNSPECIFIED);
-    if (len != sizeof(*hvstate)) {
-        address_space_unmap(CPU(cpu)->as, hvstate, len, 0, false);
-        return H_PARAMETER;
-    }
-
-    memcpy(&hv_state, hvstate, len);
-
-    address_space_unmap(CPU(cpu)->as, hvstate, len, len, false);
-
-    /*
-     * We accept versions 1 and 2. Version 2 fields are unused because TCG
-     * does not implement DAWR*.
-     */
-    if (hv_state.version > HV_GUEST_STATE_VERSION) {
-        return H_PARAMETER;
-    }
-
-    spapr_cpu->nested_host_state = g_try_new(CPUPPCState, 1);
-    if (!spapr_cpu->nested_host_state) {
-        return H_NO_MEM;
-    }
-
-    memcpy(spapr_cpu->nested_host_state, env, sizeof(CPUPPCState));
-
-    len = sizeof(*regs);
-    regs = address_space_map(CPU(cpu)->as, regs_ptr, &len, false,
-                                MEMTXATTRS_UNSPECIFIED);
-    if (!regs || len != sizeof(*regs)) {
-        address_space_unmap(CPU(cpu)->as, regs, len, 0, false);
-        g_free(spapr_cpu->nested_host_state);
-        return H_P2;
-    }
-
-    len = sizeof(env->gpr);
-    assert(len == sizeof(regs->gpr));
-    memcpy(env->gpr, regs->gpr, len);
-
-    env->lr = regs->link;
-    env->ctr = regs->ctr;
-    cpu_write_xer(env, regs->xer);
-    ppc_set_cr(env, regs->ccr);
-
-    env->msr = regs->msr;
-    env->nip = regs->nip;
-
-    address_space_unmap(CPU(cpu)->as, regs, len, len, false);
-
-    env->cfar = hv_state.cfar;
-
-    assert(env->spr[SPR_LPIDR] == 0);
-    env->spr[SPR_LPIDR] = hv_state.lpid;
-
-    lpcr_mask = LPCR_DPFD | LPCR_ILE | LPCR_AIL | LPCR_LD | LPCR_MER;
-    lpcr = (env->spr[SPR_LPCR] & ~lpcr_mask) | (hv_state.lpcr & lpcr_mask);
-    lpcr |= LPCR_HR | LPCR_UPRT | LPCR_GTSE | LPCR_HVICE | LPCR_HDICE;
-    lpcr &= ~LPCR_LPES0;
-    env->spr[SPR_LPCR] = lpcr & pcc->lpcr_mask;
-
-    env->spr[SPR_PCR] = hv_state.pcr;
-    /* hv_state.amor is not used */
-    env->spr[SPR_DPDES] = hv_state.dpdes;
-    env->spr[SPR_HFSCR] = hv_state.hfscr;
-    hdec = hv_state.hdec_expiry - now;
-    spapr_cpu->nested_tb_offset = hv_state.tb_offset;
-    /* TCG does not implement DAWR*, CIABR, PURR, SPURR, IC, VTB, HEIR SPRs*/
-    env->spr[SPR_SRR0] = hv_state.srr0;
-    env->spr[SPR_SRR1] = hv_state.srr1;
-    env->spr[SPR_SPRG0] = hv_state.sprg[0];
-    env->spr[SPR_SPRG1] = hv_state.sprg[1];
-    env->spr[SPR_SPRG2] = hv_state.sprg[2];
-    env->spr[SPR_SPRG3] = hv_state.sprg[3];
-    env->spr[SPR_BOOKS_PID] = hv_state.pidr;
-    env->spr[SPR_PPR] = hv_state.ppr;
-
-    cpu_ppc_hdecr_init(env);
-    cpu_ppc_store_hdecr(env, hdec);
-
-    /*
-     * The hv_state.vcpu_token is not needed. It is used by the KVM
-     * implementation to remember which L2 vCPU last ran on which physical
-     * CPU so as to invalidate process scope translations if it is moved
-     * between physical CPUs. For now TLBs are always flushed on L1<->L2
-     * transitions so this is not a problem.
-     *
-     * Could validate that the same vcpu_token does not attempt to run on
-     * different L1 vCPUs at the same time, but that would be a L1 KVM bug
-     * and it's not obviously worth a new data structure to do it.
-     */
-
-    env->tb_env->tb_offset += spapr_cpu->nested_tb_offset;
-    spapr_cpu->in_nested = true;
-
-    hreg_compute_hflags(env);
-    ppc_maybe_interrupt(env);
-    tlb_flush(cs);
-    env->reserve_addr = -1; /* Reset the reservation */
-
-    /*
-     * The spapr hcall helper sets env->gpr[3] to the return value, but at
-     * this point the L1 is not returning from the hcall but rather we
-     * start running the L2, so r3 must not be clobbered, so return env->gpr[3]
-     * to leave it unchanged.
-     */
-    return env->gpr[3];
-}
-
-void spapr_exit_nested(PowerPCCPU *cpu, int excp)
-{
-    CPUState *cs = CPU(cpu);
-    CPUPPCState *env = &cpu->env;
-    SpaprCpuState *spapr_cpu = spapr_cpu_state(cpu);
-    target_ulong r3_return = env->excp_vectors[excp]; /* hcall return value */
-    target_ulong hv_ptr = spapr_cpu->nested_host_state->gpr[4];
-    target_ulong regs_ptr = spapr_cpu->nested_host_state->gpr[5];
-    struct kvmppc_hv_guest_state *hvstate;
-    struct kvmppc_pt_regs *regs;
-    hwaddr len;
-
-    assert(spapr_cpu->in_nested);
-
-    cpu_ppc_hdecr_exit(env);
-
-    len = sizeof(*hvstate);
-    hvstate = address_space_map(CPU(cpu)->as, hv_ptr, &len, true,
-                                MEMTXATTRS_UNSPECIFIED);
-    if (len != sizeof(*hvstate)) {
-        address_space_unmap(CPU(cpu)->as, hvstate, len, 0, true);
-        r3_return = H_PARAMETER;
-        goto out_restore_l1;
-    }
-
-    hvstate->cfar = env->cfar;
-    hvstate->lpcr = env->spr[SPR_LPCR];
-    hvstate->pcr = env->spr[SPR_PCR];
-    hvstate->dpdes = env->spr[SPR_DPDES];
-    hvstate->hfscr = env->spr[SPR_HFSCR];
-
-    if (excp == POWERPC_EXCP_HDSI) {
-        hvstate->hdar = env->spr[SPR_HDAR];
-        hvstate->hdsisr = env->spr[SPR_HDSISR];
-        hvstate->asdr = env->spr[SPR_ASDR];
-    } else if (excp == POWERPC_EXCP_HISI) {
-        hvstate->asdr = env->spr[SPR_ASDR];
-    }
-
-    /* HEIR should be implemented for HV mode and saved here. */
-    hvstate->srr0 = env->spr[SPR_SRR0];
-    hvstate->srr1 = env->spr[SPR_SRR1];
-    hvstate->sprg[0] = env->spr[SPR_SPRG0];
-    hvstate->sprg[1] = env->spr[SPR_SPRG1];
-    hvstate->sprg[2] = env->spr[SPR_SPRG2];
-    hvstate->sprg[3] = env->spr[SPR_SPRG3];
-    hvstate->pidr = env->spr[SPR_BOOKS_PID];
-    hvstate->ppr = env->spr[SPR_PPR];
-
-    /* Is it okay to specify write length larger than actual data written? */
-    address_space_unmap(CPU(cpu)->as, hvstate, len, len, true);
-
-    len = sizeof(*regs);
-    regs = address_space_map(CPU(cpu)->as, regs_ptr, &len, true,
-                                MEMTXATTRS_UNSPECIFIED);
-    if (!regs || len != sizeof(*regs)) {
-        address_space_unmap(CPU(cpu)->as, regs, len, 0, true);
-        r3_return = H_P2;
-        goto out_restore_l1;
-    }
-
-    len = sizeof(env->gpr);
-    assert(len == sizeof(regs->gpr));
-    memcpy(regs->gpr, env->gpr, len);
-
-    regs->link = env->lr;
-    regs->ctr = env->ctr;
-    regs->xer = cpu_read_xer(env);
-    regs->ccr = ppc_get_cr(env);
-
-    if (excp == POWERPC_EXCP_MCHECK ||
-        excp == POWERPC_EXCP_RESET ||
-        excp == POWERPC_EXCP_SYSCALL) {
-        regs->nip = env->spr[SPR_SRR0];
-        regs->msr = env->spr[SPR_SRR1] & env->msr_mask;
-    } else {
-        regs->nip = env->spr[SPR_HSRR0];
-        regs->msr = env->spr[SPR_HSRR1] & env->msr_mask;
-    }
-
-    /* Is it okay to specify write length larger than actual data written? */
-    address_space_unmap(CPU(cpu)->as, regs, len, len, true);
-
-out_restore_l1:
-    memcpy(env->gpr, spapr_cpu->nested_host_state->gpr, sizeof(env->gpr));
-    env->lr = spapr_cpu->nested_host_state->lr;
-    env->ctr = spapr_cpu->nested_host_state->ctr;
-    memcpy(env->crf, spapr_cpu->nested_host_state->crf, sizeof(env->crf));
-    env->cfar = spapr_cpu->nested_host_state->cfar;
-    env->xer = spapr_cpu->nested_host_state->xer;
-    env->so = spapr_cpu->nested_host_state->so;
-    env->ov = spapr_cpu->nested_host_state->ov;
-    env->ov32 = spapr_cpu->nested_host_state->ov32;
-    env->ca32 = spapr_cpu->nested_host_state->ca32;
-    env->msr = spapr_cpu->nested_host_state->msr;
-    env->nip = spapr_cpu->nested_host_state->nip;
-
-    assert(env->spr[SPR_LPIDR] != 0);
-    env->spr[SPR_LPCR] = spapr_cpu->nested_host_state->spr[SPR_LPCR];
-    env->spr[SPR_LPIDR] = spapr_cpu->nested_host_state->spr[SPR_LPIDR];
-    env->spr[SPR_PCR] = spapr_cpu->nested_host_state->spr[SPR_PCR];
-    env->spr[SPR_DPDES] = 0;
-    env->spr[SPR_HFSCR] = spapr_cpu->nested_host_state->spr[SPR_HFSCR];
-    env->spr[SPR_SRR0] = spapr_cpu->nested_host_state->spr[SPR_SRR0];
-    env->spr[SPR_SRR1] = spapr_cpu->nested_host_state->spr[SPR_SRR1];
-    env->spr[SPR_SPRG0] = spapr_cpu->nested_host_state->spr[SPR_SPRG0];
-    env->spr[SPR_SPRG1] = spapr_cpu->nested_host_state->spr[SPR_SPRG1];
-    env->spr[SPR_SPRG2] = spapr_cpu->nested_host_state->spr[SPR_SPRG2];
-    env->spr[SPR_SPRG3] = spapr_cpu->nested_host_state->spr[SPR_SPRG3];
-    env->spr[SPR_BOOKS_PID] = spapr_cpu->nested_host_state->spr[SPR_BOOKS_PID];
-    env->spr[SPR_PPR] = spapr_cpu->nested_host_state->spr[SPR_PPR];
-
-    /*
-     * Return the interrupt vector address from H_ENTER_NESTED to the L1
-     * (or error code).
-     */
-    env->gpr[3] = r3_return;
-
-    env->tb_env->tb_offset -= spapr_cpu->nested_tb_offset;
-    spapr_cpu->in_nested = false;
-
-    hreg_compute_hflags(env);
-    ppc_maybe_interrupt(env);
-    tlb_flush(cs);
-    env->reserve_addr = -1; /* Reset the reservation */
-
-    g_free(spapr_cpu->nested_host_state);
-    spapr_cpu->nested_host_state = NULL;
-}
-
-static void hypercall_register_nested(void)
-{
-    spapr_register_hypercall(KVMPPC_H_SET_PARTITION_TABLE, h_set_ptbl);
-    spapr_register_hypercall(KVMPPC_H_ENTER_NESTED, h_enter_nested);
-    spapr_register_hypercall(KVMPPC_H_TLB_INVALIDATE, h_tlb_invalidate);
-    spapr_register_hypercall(KVMPPC_H_COPY_TOFROM_GUEST, h_copy_tofrom_guest);
-}
-
 static void hypercall_register_softmmu(void)
 {
     /* DO NOTHING */
 }
 #else
-void spapr_exit_nested(PowerPCCPU *cpu, int excp)
-{
-    g_assert_not_reached();
-}
-
 static target_ulong h_softmmu(PowerPCCPU *cpu, SpaprMachineState *spapr,
                             target_ulong opcode, target_ulong *args)
 {
     g_assert_not_reached();
 }
 
-static void hypercall_register_nested(void)
-{
-    /* DO NOTHING */
-}
-
 static void hypercall_register_softmmu(void)
 {
     /* hcall-pft */
@@ -1910,7 +1579,7 @@ static void hypercall_register_types(void)
 
     spapr_register_hypercall(KVMPPC_H_UPDATE_DT, h_update_dt);
 
-    hypercall_register_nested();
+    spapr_register_nested();
 }
 
 type_init(hypercall_register_types)
diff --git a/hw/ppc/spapr_nested.c b/hw/ppc/spapr_nested.c
new file mode 100644
index 0000000..121aa96
--- /dev/null
+++ b/hw/ppc/spapr_nested.c
@@ -0,0 +1,395 @@
+#include "qemu/osdep.h"
+#include "qemu/cutils.h"
+#include "exec/exec-all.h"
+#include "helper_regs.h"
+#include "hw/ppc/ppc.h"
+#include "hw/ppc/spapr.h"
+#include "hw/ppc/spapr_cpu_core.h"
+#include "hw/ppc/spapr_nested.h"
+
+#ifdef CONFIG_TCG
+#define PRTS_MASK      0x1f
+
+static target_ulong h_set_ptbl(PowerPCCPU *cpu,
+                               SpaprMachineState *spapr,
+                               target_ulong opcode,
+                               target_ulong *args)
+{
+    target_ulong ptcr = args[0];
+
+    if (!spapr_get_cap(spapr, SPAPR_CAP_NESTED_KVM_HV)) {
+        return H_FUNCTION;
+    }
+
+    if ((ptcr & PRTS_MASK) + 12 - 4 > 12) {
+        return H_PARAMETER;
+    }
+
+    spapr->nested_ptcr = ptcr; /* Save new partition table */
+
+    return H_SUCCESS;
+}
+
+static target_ulong h_tlb_invalidate(PowerPCCPU *cpu,
+                                     SpaprMachineState *spapr,
+                                     target_ulong opcode,
+                                     target_ulong *args)
+{
+    /*
+     * The spapr virtual hypervisor nested HV implementation retains no L2
+     * translation state except for TLB. And the TLB is always invalidated
+     * across L1<->L2 transitions, so nothing is required here.
+     */
+
+    return H_SUCCESS;
+}
+
+static target_ulong h_copy_tofrom_guest(PowerPCCPU *cpu,
+                                        SpaprMachineState *spapr,
+                                        target_ulong opcode,
+                                        target_ulong *args)
+{
+    /*
+     * This HCALL is not required, L1 KVM will take a slow path and walk the
+     * page tables manually to do the data copy.
+     */
+    return H_FUNCTION;
+}
+
+static void nested_save_state(struct nested_ppc_state *save, PowerPCCPU *cpu)
+{
+    CPUPPCState *env = &cpu->env;
+
+    memcpy(save->gpr, env->gpr, sizeof(save->gpr));
+
+    save->lr = env->lr;
+    save->ctr = env->ctr;
+    save->cfar = env->cfar;
+    save->msr = env->msr;
+    save->nip = env->nip;
+
+    save->cr = ppc_get_cr(env);
+    save->xer = cpu_read_xer(env);
+
+    save->lpcr = env->spr[SPR_LPCR];
+    save->lpidr = env->spr[SPR_LPIDR];
+    save->pcr = env->spr[SPR_PCR];
+    save->dpdes = env->spr[SPR_DPDES];
+    save->hfscr = env->spr[SPR_HFSCR];
+    save->srr0 = env->spr[SPR_SRR0];
+    save->srr1 = env->spr[SPR_SRR1];
+    save->sprg0 = env->spr[SPR_SPRG0];
+    save->sprg1 = env->spr[SPR_SPRG1];
+    save->sprg2 = env->spr[SPR_SPRG2];
+    save->sprg3 = env->spr[SPR_SPRG3];
+    save->pidr = env->spr[SPR_BOOKS_PID];
+    save->ppr = env->spr[SPR_PPR];
+
+    save->tb_offset = env->tb_env->tb_offset;
+}
+
+static void nested_load_state(PowerPCCPU *cpu, struct nested_ppc_state *load)
+{
+    CPUState *cs = CPU(cpu);
+    CPUPPCState *env = &cpu->env;
+
+    memcpy(env->gpr, load->gpr, sizeof(env->gpr));
+
+    env->lr = load->lr;
+    env->ctr = load->ctr;
+    env->cfar = load->cfar;
+    env->msr = load->msr;
+    env->nip = load->nip;
+
+    ppc_set_cr(env, load->cr);
+    cpu_write_xer(env, load->xer);
+
+    env->spr[SPR_LPCR] = load->lpcr;
+    env->spr[SPR_LPIDR] = load->lpidr;
+    env->spr[SPR_PCR] = load->pcr;
+    env->spr[SPR_DPDES] = load->dpdes;
+    env->spr[SPR_HFSCR] = load->hfscr;
+    env->spr[SPR_SRR0] = load->srr0;
+    env->spr[SPR_SRR1] = load->srr1;
+    env->spr[SPR_SPRG0] = load->sprg0;
+    env->spr[SPR_SPRG1] = load->sprg1;
+    env->spr[SPR_SPRG2] = load->sprg2;
+    env->spr[SPR_SPRG3] = load->sprg3;
+    env->spr[SPR_BOOKS_PID] = load->pidr;
+    env->spr[SPR_PPR] = load->ppr;
+
+    env->tb_env->tb_offset = load->tb_offset;
+
+    /*
+     * MSR updated, compute hflags and possible interrupts.
+     */
+    hreg_compute_hflags(env);
+    ppc_maybe_interrupt(env);
+
+    /*
+     * Nested HV does not tag TLB entries between L1 and L2, so must
+     * flush on transition.
+     */
+    tlb_flush(cs);
+    env->reserve_addr = -1; /* Reset the reservation */
+}
+
+/*
+ * When this handler returns, the environment is switched to the L2 guest
+ * and TCG begins running that. spapr_exit_nested() performs the switch from
+ * L2 back to L1 and returns from the H_ENTER_NESTED hcall.
+ */
+static target_ulong h_enter_nested(PowerPCCPU *cpu,
+                                   SpaprMachineState *spapr,
+                                   target_ulong opcode,
+                                   target_ulong *args)
+{
+    PowerPCCPUClass *pcc = POWERPC_CPU_GET_CLASS(cpu);
+    CPUPPCState *env = &cpu->env;
+    SpaprCpuState *spapr_cpu = spapr_cpu_state(cpu);
+    struct nested_ppc_state l2_state;
+    target_ulong hv_ptr = args[0];
+    target_ulong regs_ptr = args[1];
+    target_ulong hdec, now = cpu_ppc_load_tbl(env);
+    target_ulong lpcr, lpcr_mask;
+    struct kvmppc_hv_guest_state *hvstate;
+    struct kvmppc_hv_guest_state hv_state;
+    struct kvmppc_pt_regs *regs;
+    hwaddr len;
+
+    if (spapr->nested_ptcr == 0) {
+        return H_NOT_AVAILABLE;
+    }
+
+    len = sizeof(*hvstate);
+    hvstate = address_space_map(CPU(cpu)->as, hv_ptr, &len, false,
+                                MEMTXATTRS_UNSPECIFIED);
+    if (len != sizeof(*hvstate)) {
+        address_space_unmap(CPU(cpu)->as, hvstate, len, 0, false);
+        return H_PARAMETER;
+    }
+
+    memcpy(&hv_state, hvstate, len);
+
+    address_space_unmap(CPU(cpu)->as, hvstate, len, len, false);
+
+    /*
+     * We accept versions 1 and 2. Version 2 fields are unused because TCG
+     * does not implement DAWR*.
+     */
+    if (hv_state.version > HV_GUEST_STATE_VERSION) {
+        return H_PARAMETER;
+    }
+
+    if (hv_state.lpid == 0) {
+        return H_PARAMETER;
+    }
+
+    spapr_cpu->nested_host_state = g_try_new(struct nested_ppc_state, 1);
+    if (!spapr_cpu->nested_host_state) {
+        return H_NO_MEM;
+    }
+
+    assert(env->spr[SPR_LPIDR] == 0);
+    assert(env->spr[SPR_DPDES] == 0);
+    nested_save_state(spapr_cpu->nested_host_state, cpu);
+
+    len = sizeof(*regs);
+    regs = address_space_map(CPU(cpu)->as, regs_ptr, &len, false,
+                                MEMTXATTRS_UNSPECIFIED);
+    if (!regs || len != sizeof(*regs)) {
+        address_space_unmap(CPU(cpu)->as, regs, len, 0, false);
+        g_free(spapr_cpu->nested_host_state);
+        return H_P2;
+    }
+
+    len = sizeof(l2_state.gpr);
+    assert(len == sizeof(regs->gpr));
+    memcpy(l2_state.gpr, regs->gpr, len);
+
+    l2_state.lr = regs->link;
+    l2_state.ctr = regs->ctr;
+    l2_state.xer = regs->xer;
+    l2_state.cr = regs->ccr;
+    l2_state.msr = regs->msr;
+    l2_state.nip = regs->nip;
+
+    address_space_unmap(CPU(cpu)->as, regs, len, len, false);
+
+    l2_state.cfar = hv_state.cfar;
+    l2_state.lpidr = hv_state.lpid;
+
+    lpcr_mask = LPCR_DPFD | LPCR_ILE | LPCR_AIL | LPCR_LD | LPCR_MER;
+    lpcr = (env->spr[SPR_LPCR] & ~lpcr_mask) | (hv_state.lpcr & lpcr_mask);
+    lpcr |= LPCR_HR | LPCR_UPRT | LPCR_GTSE | LPCR_HVICE | LPCR_HDICE;
+    lpcr &= ~LPCR_LPES0;
+    l2_state.lpcr = lpcr & pcc->lpcr_mask;
+
+    l2_state.pcr = hv_state.pcr;
+    /* hv_state.amor is not used */
+    l2_state.dpdes = hv_state.dpdes;
+    l2_state.hfscr = hv_state.hfscr;
+    /* TCG does not implement DAWR*, CIABR, PURR, SPURR, IC, VTB, HEIR SPRs*/
+    l2_state.srr0 = hv_state.srr0;
+    l2_state.srr1 = hv_state.srr1;
+    l2_state.sprg0 = hv_state.sprg[0];
+    l2_state.sprg1 = hv_state.sprg[1];
+    l2_state.sprg2 = hv_state.sprg[2];
+    l2_state.sprg3 = hv_state.sprg[3];
+    l2_state.pidr = hv_state.pidr;
+    l2_state.ppr = hv_state.ppr;
+    l2_state.tb_offset = env->tb_env->tb_offset + hv_state.tb_offset;
+
+    /*
+     * Switch to the nested guest environment and start the "hdec" timer.
+     */
+    nested_load_state(cpu, &l2_state);
+
+    hdec = hv_state.hdec_expiry - now;
+    cpu_ppc_hdecr_init(env);
+    cpu_ppc_store_hdecr(env, hdec);
+
+    /*
+     * The hv_state.vcpu_token is not needed. It is used by the KVM
+     * implementation to remember which L2 vCPU last ran on which physical
+     * CPU so as to invalidate process scope translations if it is moved
+     * between physical CPUs. For now TLBs are always flushed on L1<->L2
+     * transitions so this is not a problem.
+     *
+     * Could validate that the same vcpu_token does not attempt to run on
+     * different L1 vCPUs at the same time, but that would be a L1 KVM bug
+     * and it's not obviously worth a new data structure to do it.
+     */
+
+    spapr_cpu->in_nested = true;
+
+    /*
+     * The spapr hcall helper sets env->gpr[3] to the return value, but at
+     * this point the L1 is not returning from the hcall but rather we
+     * start running the L2, so r3 must not be clobbered, so return env->gpr[3]
+     * to leave it unchanged.
+     */
+    return env->gpr[3];
+}
+
+void spapr_exit_nested(PowerPCCPU *cpu, int excp)
+{
+    CPUPPCState *env = &cpu->env;
+    SpaprCpuState *spapr_cpu = spapr_cpu_state(cpu);
+    struct nested_ppc_state l2_state;
+    target_ulong hv_ptr = spapr_cpu->nested_host_state->gpr[4];
+    target_ulong regs_ptr = spapr_cpu->nested_host_state->gpr[5];
+    target_ulong hsrr0, hsrr1, hdar, asdr, hdsisr;
+    struct kvmppc_hv_guest_state *hvstate;
+    struct kvmppc_pt_regs *regs;
+    hwaddr len;
+
+    assert(spapr_cpu->in_nested);
+
+    nested_save_state(&l2_state, cpu);
+    hsrr0 = env->spr[SPR_HSRR0];
+    hsrr1 = env->spr[SPR_HSRR1];
+    hdar = env->spr[SPR_HDAR];
+    hdsisr = env->spr[SPR_HDSISR];
+    asdr = env->spr[SPR_ASDR];
+
+    /*
+     * Switch back to the host environment (including for any error).
+     */
+    assert(env->spr[SPR_LPIDR] != 0);
+    nested_load_state(cpu, spapr_cpu->nested_host_state);
+    env->gpr[3] = env->excp_vectors[excp]; /* hcall return value */
+
+    cpu_ppc_hdecr_exit(env);
+
+    spapr_cpu->in_nested = false;
+
+    g_free(spapr_cpu->nested_host_state);
+    spapr_cpu->nested_host_state = NULL;
+
+    len = sizeof(*hvstate);
+    hvstate = address_space_map(CPU(cpu)->as, hv_ptr, &len, true,
+                                MEMTXATTRS_UNSPECIFIED);
+    if (len != sizeof(*hvstate)) {
+        address_space_unmap(CPU(cpu)->as, hvstate, len, 0, true);
+        env->gpr[3] = H_PARAMETER;
+        return;
+    }
+
+    hvstate->cfar = l2_state.cfar;
+    hvstate->lpcr = l2_state.lpcr;
+    hvstate->pcr = l2_state.pcr;
+    hvstate->dpdes = l2_state.dpdes;
+    hvstate->hfscr = l2_state.hfscr;
+
+    if (excp == POWERPC_EXCP_HDSI) {
+        hvstate->hdar = hdar;
+        hvstate->hdsisr = hdsisr;
+        hvstate->asdr = asdr;
+    } else if (excp == POWERPC_EXCP_HISI) {
+        hvstate->asdr = asdr;
+    }
+
+    /* HEIR should be implemented for HV mode and saved here. */
+    hvstate->srr0 = l2_state.srr0;
+    hvstate->srr1 = l2_state.srr1;
+    hvstate->sprg[0] = l2_state.sprg0;
+    hvstate->sprg[1] = l2_state.sprg1;
+    hvstate->sprg[2] = l2_state.sprg2;
+    hvstate->sprg[3] = l2_state.sprg3;
+    hvstate->pidr = l2_state.pidr;
+    hvstate->ppr = l2_state.ppr;
+
+    /* Is it okay to specify write length larger than actual data written? */
+    address_space_unmap(CPU(cpu)->as, hvstate, len, len, true);
+
+    len = sizeof(*regs);
+    regs = address_space_map(CPU(cpu)->as, regs_ptr, &len, true,
+                                MEMTXATTRS_UNSPECIFIED);
+    if (!regs || len != sizeof(*regs)) {
+        address_space_unmap(CPU(cpu)->as, regs, len, 0, true);
+        env->gpr[3] = H_P2;
+        return;
+    }
+
+    len = sizeof(env->gpr);
+    assert(len == sizeof(regs->gpr));
+    memcpy(regs->gpr, l2_state.gpr, len);
+
+    regs->link = l2_state.lr;
+    regs->ctr = l2_state.ctr;
+    regs->xer = l2_state.xer;
+    regs->ccr = l2_state.cr;
+
+    if (excp == POWERPC_EXCP_MCHECK ||
+        excp == POWERPC_EXCP_RESET ||
+        excp == POWERPC_EXCP_SYSCALL) {
+        regs->nip = l2_state.srr0;
+        regs->msr = l2_state.srr1 & env->msr_mask;
+    } else {
+        regs->nip = hsrr0;
+        regs->msr = hsrr1 & env->msr_mask;
+    }
+
+    /* Is it okay to specify write length larger than actual data written? */
+    address_space_unmap(CPU(cpu)->as, regs, len, len, true);
+}
+
+void spapr_register_nested(void)
+{
+    spapr_register_hypercall(KVMPPC_H_SET_PARTITION_TABLE, h_set_ptbl);
+    spapr_register_hypercall(KVMPPC_H_ENTER_NESTED, h_enter_nested);
+    spapr_register_hypercall(KVMPPC_H_TLB_INVALIDATE, h_tlb_invalidate);
+    spapr_register_hypercall(KVMPPC_H_COPY_TOFROM_GUEST, h_copy_tofrom_guest);
+}
+#else
+void spapr_exit_nested(PowerPCCPU *cpu, int excp)
+{
+    g_assert_not_reached();
+}
+
+void spapr_register_nested(void)
+{
+    /* DO NOTHING */
+}
+#endif