Merge remote-tracking branch 'remotes/aurel/tags/pull-sh4-next-20150612' into staging

sh4 linux-user cpu and hwcap
misc optimizations and cleanup
convert r2d to new MMIO accessor style

# gpg: Signature made Fri Jun 12 11:28:43 2015 BST using RSA key ID 1DDD8C9B
# gpg: Good signature from "Aurelien Jarno <aurelien@aurel32.net>"
# gpg:                 aka "Aurelien Jarno <aurelien@jarno.fr>"
# gpg:                 aka "Aurelien Jarno <aurel32@debian.org>"
# 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: 7746 2642 A9EF 94FD 0F77  196D BA9C 7806 1DDD 8C9B

* remotes/aurel/tags/pull-sh4-next-20150612:
  target-sh4: remove dead code
  target-sh4: factorize fmov implementation
  target-sh4: split out Q and M from of SR and optimize div1
  target-sh4: optimize negc using add2 and sub2
  target-sh4: optimize subc using sub2
  target-sh4: optimize addc using add2
  target-sh4: Split out T from SR
  target-sh4: use bit number for SR constants
  sh4/r2d: convert to new MMIO accessor style
  linux-user: Add HWCAP for SH4
  linux-user: Default sh4 to sh7785

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
diff --git a/MAINTAINERS b/MAINTAINERS
index 4ed8215..e728d3a 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -1014,8 +1014,6 @@
 S: Maintained
 F: include/migration/
 F: migration/
-F: savevm.c
-F: arch_init.c
 F: scripts/vmstate-static-checker.py
 F: tests/vmstate-static-checker-data/
 
diff --git a/Makefile.target b/Makefile.target
index ec5b92c..3e7aafd 100644
--- a/Makefile.target
+++ b/Makefile.target
@@ -132,9 +132,10 @@
 obj-y += qtest.o bootdevice.o
 obj-y += hw/
 obj-$(CONFIG_KVM) += kvm-all.o
-obj-y += memory.o savevm.o cputlb.o
+obj-y += memory.o cputlb.o
 obj-y += memory_mapping.o
 obj-y += dump.o
+obj-y += migration/ram.o migration/savevm.o
 LIBS := $(libs_softmmu) $(LIBS)
 
 # xen support
diff --git a/arch_init.c b/arch_init.c
index d294474..725c638 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -22,46 +22,15 @@
  * THE SOFTWARE.
  */
 #include <stdint.h>
-#include <stdarg.h>
-#include <stdlib.h>
-#include <zlib.h>
-#ifndef _WIN32
-#include <sys/types.h>
-#include <sys/mman.h>
-#endif
-#include "config.h"
-#include "monitor/monitor.h"
 #include "sysemu/sysemu.h"
-#include "qemu/bitops.h"
-#include "qemu/bitmap.h"
 #include "sysemu/arch_init.h"
-#include "audio/audio.h"
-#include "hw/i386/pc.h"
 #include "hw/pci/pci.h"
 #include "hw/audio/audio.h"
-#include "sysemu/kvm.h"
-#include "migration/migration.h"
 #include "hw/i386/smbios.h"
-#include "exec/address-spaces.h"
-#include "hw/audio/pcspk.h"
-#include "migration/page_cache.h"
 #include "qemu/config-file.h"
 #include "qemu/error-report.h"
 #include "qmp-commands.h"
-#include "trace.h"
-#include "exec/cpu-all.h"
-#include "exec/ram_addr.h"
 #include "hw/acpi/acpi.h"
-#include "qemu/host-utils.h"
-#include "qemu/rcu_queue.h"
-
-#ifdef DEBUG_ARCH_INIT
-#define DPRINTF(fmt, ...) \
-    do { fprintf(stdout, "arch_init: " fmt, ## __VA_ARGS__); } while (0)
-#else
-#define DPRINTF(fmt, ...) \
-    do { } while (0)
-#endif
 
 #ifdef TARGET_SPARC
 int graphic_width = 1024;
@@ -111,24 +80,6 @@
 #endif
 
 const uint32_t arch_type = QEMU_ARCH;
-static bool mig_throttle_on;
-static int dirty_rate_high_cnt;
-static void check_guest_throttling(void);
-
-static uint64_t bitmap_sync_count;
-
-/***********************************************************/
-/* ram save/restore */
-
-#define RAM_SAVE_FLAG_FULL     0x01 /* Obsolete, not used anymore */
-#define RAM_SAVE_FLAG_COMPRESS 0x02
-#define RAM_SAVE_FLAG_MEM_SIZE 0x04
-#define RAM_SAVE_FLAG_PAGE     0x08
-#define RAM_SAVE_FLAG_EOS      0x10
-#define RAM_SAVE_FLAG_CONTINUE 0x20
-#define RAM_SAVE_FLAG_XBZRLE   0x40
-/* 0x80 is reserved in migration.h start with 0x100 next */
-#define RAM_SAVE_FLAG_COMPRESS_PAGE    0x100
 
 static struct defconfig_file {
     const char *filename;
@@ -139,8 +90,6 @@
     { NULL }, /* end of list */
 };
 
-static const uint8_t ZERO_TARGET_PAGE[TARGET_PAGE_SIZE];
-
 int qemu_read_default_config_files(bool userconfig)
 {
     int ret;
@@ -159,1517 +108,6 @@
     return 0;
 }
 
-static inline bool is_zero_range(uint8_t *p, uint64_t size)
-{
-    return buffer_find_nonzero_offset(p, size) == size;
-}
-
-/* struct contains XBZRLE cache and a static page
-   used by the compression */
-static struct {
-    /* buffer used for XBZRLE encoding */
-    uint8_t *encoded_buf;
-    /* buffer for storing page content */
-    uint8_t *current_buf;
-    /* Cache for XBZRLE, Protected by lock. */
-    PageCache *cache;
-    QemuMutex lock;
-} XBZRLE;
-
-/* buffer used for XBZRLE decoding */
-static uint8_t *xbzrle_decoded_buf;
-
-static void XBZRLE_cache_lock(void)
-{
-    if (migrate_use_xbzrle())
-        qemu_mutex_lock(&XBZRLE.lock);
-}
-
-static void XBZRLE_cache_unlock(void)
-{
-    if (migrate_use_xbzrle())
-        qemu_mutex_unlock(&XBZRLE.lock);
-}
-
-/*
- * called from qmp_migrate_set_cache_size in main thread, possibly while
- * a migration is in progress.
- * A running migration maybe using the cache and might finish during this
- * call, hence changes to the cache are protected by XBZRLE.lock().
- */
-int64_t xbzrle_cache_resize(int64_t new_size)
-{
-    PageCache *new_cache;
-    int64_t ret;
-
-    if (new_size < TARGET_PAGE_SIZE) {
-        return -1;
-    }
-
-    XBZRLE_cache_lock();
-
-    if (XBZRLE.cache != NULL) {
-        if (pow2floor(new_size) == migrate_xbzrle_cache_size()) {
-            goto out_new_size;
-        }
-        new_cache = cache_init(new_size / TARGET_PAGE_SIZE,
-                                        TARGET_PAGE_SIZE);
-        if (!new_cache) {
-            error_report("Error creating cache");
-            ret = -1;
-            goto out;
-        }
-
-        cache_fini(XBZRLE.cache);
-        XBZRLE.cache = new_cache;
-    }
-
-out_new_size:
-    ret = pow2floor(new_size);
-out:
-    XBZRLE_cache_unlock();
-    return ret;
-}
-
-/* accounting for migration statistics */
-typedef struct AccountingInfo {
-    uint64_t dup_pages;
-    uint64_t skipped_pages;
-    uint64_t norm_pages;
-    uint64_t iterations;
-    uint64_t xbzrle_bytes;
-    uint64_t xbzrle_pages;
-    uint64_t xbzrle_cache_miss;
-    double xbzrle_cache_miss_rate;
-    uint64_t xbzrle_overflows;
-} AccountingInfo;
-
-static AccountingInfo acct_info;
-
-static void acct_clear(void)
-{
-    memset(&acct_info, 0, sizeof(acct_info));
-}
-
-uint64_t dup_mig_bytes_transferred(void)
-{
-    return acct_info.dup_pages * TARGET_PAGE_SIZE;
-}
-
-uint64_t dup_mig_pages_transferred(void)
-{
-    return acct_info.dup_pages;
-}
-
-uint64_t skipped_mig_bytes_transferred(void)
-{
-    return acct_info.skipped_pages * TARGET_PAGE_SIZE;
-}
-
-uint64_t skipped_mig_pages_transferred(void)
-{
-    return acct_info.skipped_pages;
-}
-
-uint64_t norm_mig_bytes_transferred(void)
-{
-    return acct_info.norm_pages * TARGET_PAGE_SIZE;
-}
-
-uint64_t norm_mig_pages_transferred(void)
-{
-    return acct_info.norm_pages;
-}
-
-uint64_t xbzrle_mig_bytes_transferred(void)
-{
-    return acct_info.xbzrle_bytes;
-}
-
-uint64_t xbzrle_mig_pages_transferred(void)
-{
-    return acct_info.xbzrle_pages;
-}
-
-uint64_t xbzrle_mig_pages_cache_miss(void)
-{
-    return acct_info.xbzrle_cache_miss;
-}
-
-double xbzrle_mig_cache_miss_rate(void)
-{
-    return acct_info.xbzrle_cache_miss_rate;
-}
-
-uint64_t xbzrle_mig_pages_overflow(void)
-{
-    return acct_info.xbzrle_overflows;
-}
-
-/* This is the last block that we have visited serching for dirty pages
- */
-static RAMBlock *last_seen_block;
-/* This is the last block from where we have sent data */
-static RAMBlock *last_sent_block;
-static ram_addr_t last_offset;
-static unsigned long *migration_bitmap;
-static uint64_t migration_dirty_pages;
-static uint32_t last_version;
-static bool ram_bulk_stage;
-
-struct CompressParam {
-    bool start;
-    bool done;
-    QEMUFile *file;
-    QemuMutex mutex;
-    QemuCond cond;
-    RAMBlock *block;
-    ram_addr_t offset;
-};
-typedef struct CompressParam CompressParam;
-
-struct DecompressParam {
-    bool start;
-    QemuMutex mutex;
-    QemuCond cond;
-    void *des;
-    uint8 *compbuf;
-    int len;
-};
-typedef struct DecompressParam DecompressParam;
-
-static CompressParam *comp_param;
-static QemuThread *compress_threads;
-/* comp_done_cond is used to wake up the migration thread when
- * one of the compression threads has finished the compression.
- * comp_done_lock is used to co-work with comp_done_cond.
- */
-static QemuMutex *comp_done_lock;
-static QemuCond *comp_done_cond;
-/* The empty QEMUFileOps will be used by file in CompressParam */
-static const QEMUFileOps empty_ops = { };
-
-static bool compression_switch;
-static bool quit_comp_thread;
-static bool quit_decomp_thread;
-static DecompressParam *decomp_param;
-static QemuThread *decompress_threads;
-static uint8_t *compressed_data_buf;
-
-static int do_compress_ram_page(CompressParam *param);
-
-static void *do_data_compress(void *opaque)
-{
-    CompressParam *param = opaque;
-
-    while (!quit_comp_thread) {
-        qemu_mutex_lock(&param->mutex);
-        /* Re-check the quit_comp_thread in case of
-         * terminate_compression_threads is called just before
-         * qemu_mutex_lock(&param->mutex) and after
-         * while(!quit_comp_thread), re-check it here can make
-         * sure the compression thread terminate as expected.
-         */
-        while (!param->start && !quit_comp_thread) {
-            qemu_cond_wait(&param->cond, &param->mutex);
-        }
-        if (!quit_comp_thread) {
-            do_compress_ram_page(param);
-        }
-        param->start = false;
-        qemu_mutex_unlock(&param->mutex);
-
-        qemu_mutex_lock(comp_done_lock);
-        param->done = true;
-        qemu_cond_signal(comp_done_cond);
-        qemu_mutex_unlock(comp_done_lock);
-    }
-
-    return NULL;
-}
-
-static inline void terminate_compression_threads(void)
-{
-    int idx, thread_count;
-
-    thread_count = migrate_compress_threads();
-    quit_comp_thread = true;
-    for (idx = 0; idx < thread_count; idx++) {
-        qemu_mutex_lock(&comp_param[idx].mutex);
-        qemu_cond_signal(&comp_param[idx].cond);
-        qemu_mutex_unlock(&comp_param[idx].mutex);
-    }
-}
-
-void migrate_compress_threads_join(void)
-{
-    int i, thread_count;
-
-    if (!migrate_use_compression()) {
-        return;
-    }
-    terminate_compression_threads();
-    thread_count = migrate_compress_threads();
-    for (i = 0; i < thread_count; i++) {
-        qemu_thread_join(compress_threads + i);
-        qemu_fclose(comp_param[i].file);
-        qemu_mutex_destroy(&comp_param[i].mutex);
-        qemu_cond_destroy(&comp_param[i].cond);
-    }
-    qemu_mutex_destroy(comp_done_lock);
-    qemu_cond_destroy(comp_done_cond);
-    g_free(compress_threads);
-    g_free(comp_param);
-    g_free(comp_done_cond);
-    g_free(comp_done_lock);
-    compress_threads = NULL;
-    comp_param = NULL;
-    comp_done_cond = NULL;
-    comp_done_lock = NULL;
-}
-
-void migrate_compress_threads_create(void)
-{
-    int i, thread_count;
-
-    if (!migrate_use_compression()) {
-        return;
-    }
-    quit_comp_thread = false;
-    compression_switch = true;
-    thread_count = migrate_compress_threads();
-    compress_threads = g_new0(QemuThread, thread_count);
-    comp_param = g_new0(CompressParam, thread_count);
-    comp_done_cond = g_new0(QemuCond, 1);
-    comp_done_lock = g_new0(QemuMutex, 1);
-    qemu_cond_init(comp_done_cond);
-    qemu_mutex_init(comp_done_lock);
-    for (i = 0; i < thread_count; i++) {
-        /* com_param[i].file is just used as a dummy buffer to save data, set
-         * it's ops to empty.
-         */
-        comp_param[i].file = qemu_fopen_ops(NULL, &empty_ops);
-        comp_param[i].done = true;
-        qemu_mutex_init(&comp_param[i].mutex);
-        qemu_cond_init(&comp_param[i].cond);
-        qemu_thread_create(compress_threads + i, "compress",
-                           do_data_compress, comp_param + i,
-                           QEMU_THREAD_JOINABLE);
-    }
-}
-
-/**
- * save_page_header: Write page header to wire
- *
- * If this is the 1st block, it also writes the block identification
- *
- * Returns: Number of bytes written
- *
- * @f: QEMUFile where to send the data
- * @block: block that contains the page we want to send
- * @offset: offset inside the block for the page
- *          in the lower bits, it contains flags
- */
-static size_t save_page_header(QEMUFile *f, RAMBlock *block, ram_addr_t offset)
-{
-    size_t size;
-
-    qemu_put_be64(f, offset);
-    size = 8;
-
-    if (!(offset & RAM_SAVE_FLAG_CONTINUE)) {
-        qemu_put_byte(f, strlen(block->idstr));
-        qemu_put_buffer(f, (uint8_t *)block->idstr,
-                        strlen(block->idstr));
-        size += 1 + strlen(block->idstr);
-    }
-    return size;
-}
-
-/* Update the xbzrle cache to reflect a page that's been sent as all 0.
- * The important thing is that a stale (not-yet-0'd) page be replaced
- * by the new data.
- * As a bonus, if the page wasn't in the cache it gets added so that
- * when a small write is made into the 0'd page it gets XBZRLE sent
- */
-static void xbzrle_cache_zero_page(ram_addr_t current_addr)
-{
-    if (ram_bulk_stage || !migrate_use_xbzrle()) {
-        return;
-    }
-
-    /* We don't care if this fails to allocate a new cache page
-     * as long as it updated an old one */
-    cache_insert(XBZRLE.cache, current_addr, ZERO_TARGET_PAGE,
-                 bitmap_sync_count);
-}
-
-#define ENCODING_FLAG_XBZRLE 0x1
-
-/**
- * save_xbzrle_page: compress and send current page
- *
- * Returns: 1 means that we wrote the page
- *          0 means that page is identical to the one already sent
- *          -1 means that xbzrle would be longer than normal
- *
- * @f: QEMUFile where to send the data
- * @current_data:
- * @current_addr:
- * @block: block that contains the page we want to send
- * @offset: offset inside the block for the page
- * @last_stage: if we are at the completion stage
- * @bytes_transferred: increase it with the number of transferred bytes
- */
-static int save_xbzrle_page(QEMUFile *f, uint8_t **current_data,
-                            ram_addr_t current_addr, RAMBlock *block,
-                            ram_addr_t offset, bool last_stage,
-                            uint64_t *bytes_transferred)
-{
-    int encoded_len = 0, bytes_xbzrle;
-    uint8_t *prev_cached_page;
-
-    if (!cache_is_cached(XBZRLE.cache, current_addr, bitmap_sync_count)) {
-        acct_info.xbzrle_cache_miss++;
-        if (!last_stage) {
-            if (cache_insert(XBZRLE.cache, current_addr, *current_data,
-                             bitmap_sync_count) == -1) {
-                return -1;
-            } else {
-                /* update *current_data when the page has been
-                   inserted into cache */
-                *current_data = get_cached_data(XBZRLE.cache, current_addr);
-            }
-        }
-        return -1;
-    }
-
-    prev_cached_page = get_cached_data(XBZRLE.cache, current_addr);
-
-    /* save current buffer into memory */
-    memcpy(XBZRLE.current_buf, *current_data, TARGET_PAGE_SIZE);
-
-    /* XBZRLE encoding (if there is no overflow) */
-    encoded_len = xbzrle_encode_buffer(prev_cached_page, XBZRLE.current_buf,
-                                       TARGET_PAGE_SIZE, XBZRLE.encoded_buf,
-                                       TARGET_PAGE_SIZE);
-    if (encoded_len == 0) {
-        DPRINTF("Skipping unmodified page\n");
-        return 0;
-    } else if (encoded_len == -1) {
-        DPRINTF("Overflow\n");
-        acct_info.xbzrle_overflows++;
-        /* update data in the cache */
-        if (!last_stage) {
-            memcpy(prev_cached_page, *current_data, TARGET_PAGE_SIZE);
-            *current_data = prev_cached_page;
-        }
-        return -1;
-    }
-
-    /* we need to update the data in the cache, in order to get the same data */
-    if (!last_stage) {
-        memcpy(prev_cached_page, XBZRLE.current_buf, TARGET_PAGE_SIZE);
-    }
-
-    /* Send XBZRLE based compressed page */
-    bytes_xbzrle = save_page_header(f, block, offset | RAM_SAVE_FLAG_XBZRLE);
-    qemu_put_byte(f, ENCODING_FLAG_XBZRLE);
-    qemu_put_be16(f, encoded_len);
-    qemu_put_buffer(f, XBZRLE.encoded_buf, encoded_len);
-    bytes_xbzrle += encoded_len + 1 + 2;
-    acct_info.xbzrle_pages++;
-    acct_info.xbzrle_bytes += bytes_xbzrle;
-    *bytes_transferred += bytes_xbzrle;
-
-    return 1;
-}
-
-static inline
-ram_addr_t migration_bitmap_find_and_reset_dirty(MemoryRegion *mr,
-                                                 ram_addr_t start)
-{
-    unsigned long base = mr->ram_addr >> TARGET_PAGE_BITS;
-    unsigned long nr = base + (start >> TARGET_PAGE_BITS);
-    uint64_t mr_size = TARGET_PAGE_ALIGN(memory_region_size(mr));
-    unsigned long size = base + (mr_size >> TARGET_PAGE_BITS);
-
-    unsigned long next;
-
-    if (ram_bulk_stage && nr > base) {
-        next = nr + 1;
-    } else {
-        next = find_next_bit(migration_bitmap, size, nr);
-    }
-
-    if (next < size) {
-        clear_bit(next, migration_bitmap);
-        migration_dirty_pages--;
-    }
-    return (next - base) << TARGET_PAGE_BITS;
-}
-
-static void migration_bitmap_sync_range(ram_addr_t start, ram_addr_t length)
-{
-    migration_dirty_pages +=
-        cpu_physical_memory_sync_dirty_bitmap(migration_bitmap, start, length);
-}
-
-
-/* Fix me: there are too many global variables used in migration process. */
-static int64_t start_time;
-static int64_t bytes_xfer_prev;
-static int64_t num_dirty_pages_period;
-static uint64_t xbzrle_cache_miss_prev;
-static uint64_t iterations_prev;
-
-static void migration_bitmap_sync_init(void)
-{
-    start_time = 0;
-    bytes_xfer_prev = 0;
-    num_dirty_pages_period = 0;
-    xbzrle_cache_miss_prev = 0;
-    iterations_prev = 0;
-}
-
-/* Called with iothread lock held, to protect ram_list.dirty_memory[] */
-static void migration_bitmap_sync(void)
-{
-    RAMBlock *block;
-    uint64_t num_dirty_pages_init = migration_dirty_pages;
-    MigrationState *s = migrate_get_current();
-    int64_t end_time;
-    int64_t bytes_xfer_now;
-
-    bitmap_sync_count++;
-
-    if (!bytes_xfer_prev) {
-        bytes_xfer_prev = ram_bytes_transferred();
-    }
-
-    if (!start_time) {
-        start_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME);
-    }
-
-    trace_migration_bitmap_sync_start();
-    address_space_sync_dirty_bitmap(&address_space_memory);
-
-    rcu_read_lock();
-    QLIST_FOREACH_RCU(block, &ram_list.blocks, next) {
-        migration_bitmap_sync_range(block->mr->ram_addr, block->used_length);
-    }
-    rcu_read_unlock();
-
-    trace_migration_bitmap_sync_end(migration_dirty_pages
-                                    - num_dirty_pages_init);
-    num_dirty_pages_period += migration_dirty_pages - num_dirty_pages_init;
-    end_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME);
-
-    /* more than 1 second = 1000 millisecons */
-    if (end_time > start_time + 1000) {
-        if (migrate_auto_converge()) {
-            /* The following detection logic can be refined later. For now:
-               Check to see if the dirtied bytes is 50% more than the approx.
-               amount of bytes that just got transferred since the last time we
-               were in this routine. If that happens >N times (for now N==4)
-               we turn on the throttle down logic */
-            bytes_xfer_now = ram_bytes_transferred();
-            if (s->dirty_pages_rate &&
-               (num_dirty_pages_period * TARGET_PAGE_SIZE >
-                   (bytes_xfer_now - bytes_xfer_prev)/2) &&
-               (dirty_rate_high_cnt++ > 4)) {
-                    trace_migration_throttle();
-                    mig_throttle_on = true;
-                    dirty_rate_high_cnt = 0;
-             }
-             bytes_xfer_prev = bytes_xfer_now;
-        } else {
-             mig_throttle_on = false;
-        }
-        if (migrate_use_xbzrle()) {
-            if (iterations_prev != acct_info.iterations) {
-                acct_info.xbzrle_cache_miss_rate =
-                   (double)(acct_info.xbzrle_cache_miss -
-                            xbzrle_cache_miss_prev) /
-                   (acct_info.iterations - iterations_prev);
-            }
-            iterations_prev = acct_info.iterations;
-            xbzrle_cache_miss_prev = acct_info.xbzrle_cache_miss;
-        }
-        s->dirty_pages_rate = num_dirty_pages_period * 1000
-            / (end_time - start_time);
-        s->dirty_bytes_rate = s->dirty_pages_rate * TARGET_PAGE_SIZE;
-        start_time = end_time;
-        num_dirty_pages_period = 0;
-    }
-    s->dirty_sync_count = bitmap_sync_count;
-}
-
-/**
- * save_zero_page: Send the zero page to the stream
- *
- * Returns: Number of pages written.
- *
- * @f: QEMUFile where to send the data
- * @block: block that contains the page we want to send
- * @offset: offset inside the block for the page
- * @p: pointer to the page
- * @bytes_transferred: increase it with the number of transferred bytes
- */
-static int save_zero_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset,
-                          uint8_t *p, uint64_t *bytes_transferred)
-{
-    int pages = -1;
-
-    if (is_zero_range(p, TARGET_PAGE_SIZE)) {
-        acct_info.dup_pages++;
-        *bytes_transferred += save_page_header(f, block,
-                                               offset | RAM_SAVE_FLAG_COMPRESS);
-        qemu_put_byte(f, 0);
-        *bytes_transferred += 1;
-        pages = 1;
-    }
-
-    return pages;
-}
-
-/**
- * ram_save_page: Send the given page to the stream
- *
- * Returns: Number of pages written.
- *
- * @f: QEMUFile where to send the data
- * @block: block that contains the page we want to send
- * @offset: offset inside the block for the page
- * @last_stage: if we are at the completion stage
- * @bytes_transferred: increase it with the number of transferred bytes
- */
-static int ram_save_page(QEMUFile *f, RAMBlock* block, ram_addr_t offset,
-                         bool last_stage, uint64_t *bytes_transferred)
-{
-    int pages = -1;
-    uint64_t bytes_xmit;
-    ram_addr_t current_addr;
-    MemoryRegion *mr = block->mr;
-    uint8_t *p;
-    int ret;
-    bool send_async = true;
-
-    p = memory_region_get_ram_ptr(mr) + offset;
-
-    /* In doubt sent page as normal */
-    bytes_xmit = 0;
-    ret = ram_control_save_page(f, block->offset,
-                           offset, TARGET_PAGE_SIZE, &bytes_xmit);
-    if (bytes_xmit) {
-        *bytes_transferred += bytes_xmit;
-        pages = 1;
-    }
-
-    XBZRLE_cache_lock();
-
-    current_addr = block->offset + offset;
-
-    if (block == last_sent_block) {
-        offset |= RAM_SAVE_FLAG_CONTINUE;
-    }
-    if (ret != RAM_SAVE_CONTROL_NOT_SUPP) {
-        if (ret != RAM_SAVE_CONTROL_DELAYED) {
-            if (bytes_xmit > 0) {
-                acct_info.norm_pages++;
-            } else if (bytes_xmit == 0) {
-                acct_info.dup_pages++;
-            }
-        }
-    } else {
-        pages = save_zero_page(f, block, offset, p, bytes_transferred);
-        if (pages > 0) {
-            /* Must let xbzrle know, otherwise a previous (now 0'd) cached
-             * page would be stale
-             */
-            xbzrle_cache_zero_page(current_addr);
-        } else if (!ram_bulk_stage && migrate_use_xbzrle()) {
-            pages = save_xbzrle_page(f, &p, current_addr, block,
-                                     offset, last_stage, bytes_transferred);
-            if (!last_stage) {
-                /* Can't send this cached data async, since the cache page
-                 * might get updated before it gets to the wire
-                 */
-                send_async = false;
-            }
-        }
-    }
-
-    /* XBZRLE overflow or normal page */
-    if (pages == -1) {
-        *bytes_transferred += save_page_header(f, block,
-                                               offset | RAM_SAVE_FLAG_PAGE);
-        if (send_async) {
-            qemu_put_buffer_async(f, p, TARGET_PAGE_SIZE);
-        } else {
-            qemu_put_buffer(f, p, TARGET_PAGE_SIZE);
-        }
-        *bytes_transferred += TARGET_PAGE_SIZE;
-        pages = 1;
-        acct_info.norm_pages++;
-    }
-
-    XBZRLE_cache_unlock();
-
-    return pages;
-}
-
-static int do_compress_ram_page(CompressParam *param)
-{
-    int bytes_sent, blen;
-    uint8_t *p;
-    RAMBlock *block = param->block;
-    ram_addr_t offset = param->offset;
-
-    p = memory_region_get_ram_ptr(block->mr) + (offset & TARGET_PAGE_MASK);
-
-    bytes_sent = save_page_header(param->file, block, offset |
-                                  RAM_SAVE_FLAG_COMPRESS_PAGE);
-    blen = qemu_put_compression_data(param->file, p, TARGET_PAGE_SIZE,
-                                     migrate_compress_level());
-    bytes_sent += blen;
-
-    return bytes_sent;
-}
-
-static inline void start_compression(CompressParam *param)
-{
-    param->done = false;
-    qemu_mutex_lock(&param->mutex);
-    param->start = true;
-    qemu_cond_signal(&param->cond);
-    qemu_mutex_unlock(&param->mutex);
-}
-
-static inline void start_decompression(DecompressParam *param)
-{
-    qemu_mutex_lock(&param->mutex);
-    param->start = true;
-    qemu_cond_signal(&param->cond);
-    qemu_mutex_unlock(&param->mutex);
-}
-
-static uint64_t bytes_transferred;
-
-static void flush_compressed_data(QEMUFile *f)
-{
-    int idx, len, thread_count;
-
-    if (!migrate_use_compression()) {
-        return;
-    }
-    thread_count = migrate_compress_threads();
-    for (idx = 0; idx < thread_count; idx++) {
-        if (!comp_param[idx].done) {
-            qemu_mutex_lock(comp_done_lock);
-            while (!comp_param[idx].done && !quit_comp_thread) {
-                qemu_cond_wait(comp_done_cond, comp_done_lock);
-            }
-            qemu_mutex_unlock(comp_done_lock);
-        }
-        if (!quit_comp_thread) {
-            len = qemu_put_qemu_file(f, comp_param[idx].file);
-            bytes_transferred += len;
-        }
-    }
-}
-
-static inline void set_compress_params(CompressParam *param, RAMBlock *block,
-                                       ram_addr_t offset)
-{
-    param->block = block;
-    param->offset = offset;
-}
-
-static int compress_page_with_multi_thread(QEMUFile *f, RAMBlock *block,
-                                           ram_addr_t offset,
-                                           uint64_t *bytes_transferred)
-{
-    int idx, thread_count, bytes_xmit = -1, pages = -1;
-
-    thread_count = migrate_compress_threads();
-    qemu_mutex_lock(comp_done_lock);
-    while (true) {
-        for (idx = 0; idx < thread_count; idx++) {
-            if (comp_param[idx].done) {
-                bytes_xmit = qemu_put_qemu_file(f, comp_param[idx].file);
-                set_compress_params(&comp_param[idx], block, offset);
-                start_compression(&comp_param[idx]);
-                pages = 1;
-                acct_info.norm_pages++;
-                *bytes_transferred += bytes_xmit;
-                break;
-            }
-        }
-        if (pages > 0) {
-            break;
-        } else {
-            qemu_cond_wait(comp_done_cond, comp_done_lock);
-        }
-    }
-    qemu_mutex_unlock(comp_done_lock);
-
-    return pages;
-}
-
-/**
- * ram_save_compressed_page: compress the given page and send it to the stream
- *
- * Returns: Number of pages written.
- *
- * @f: QEMUFile where to send the data
- * @block: block that contains the page we want to send
- * @offset: offset inside the block for the page
- * @last_stage: if we are at the completion stage
- * @bytes_transferred: increase it with the number of transferred bytes
- */
-static int ram_save_compressed_page(QEMUFile *f, RAMBlock *block,
-                                    ram_addr_t offset, bool last_stage,
-                                    uint64_t *bytes_transferred)
-{
-    int pages = -1;
-    uint64_t bytes_xmit;
-    MemoryRegion *mr = block->mr;
-    uint8_t *p;
-    int ret;
-
-    p = memory_region_get_ram_ptr(mr) + offset;
-
-    bytes_xmit = 0;
-    ret = ram_control_save_page(f, block->offset,
-                                offset, TARGET_PAGE_SIZE, &bytes_xmit);
-    if (bytes_xmit) {
-        *bytes_transferred += bytes_xmit;
-        pages = 1;
-    }
-    if (block == last_sent_block) {
-        offset |= RAM_SAVE_FLAG_CONTINUE;
-    }
-    if (ret != RAM_SAVE_CONTROL_NOT_SUPP) {
-        if (ret != RAM_SAVE_CONTROL_DELAYED) {
-            if (bytes_xmit > 0) {
-                acct_info.norm_pages++;
-            } else if (bytes_xmit == 0) {
-                acct_info.dup_pages++;
-            }
-        }
-    } else {
-        /* When starting the process of a new block, the first page of
-         * the block should be sent out before other pages in the same
-         * block, and all the pages in last block should have been sent
-         * out, keeping this order is important, because the 'cont' flag
-         * is used to avoid resending the block name.
-         */
-        if (block != last_sent_block) {
-            flush_compressed_data(f);
-            pages = save_zero_page(f, block, offset, p, bytes_transferred);
-            if (pages == -1) {
-                set_compress_params(&comp_param[0], block, offset);
-                /* Use the qemu thread to compress the data to make sure the
-                 * first page is sent out before other pages
-                 */
-                bytes_xmit = do_compress_ram_page(&comp_param[0]);
-                acct_info.norm_pages++;
-                qemu_put_qemu_file(f, comp_param[0].file);
-                *bytes_transferred += bytes_xmit;
-                pages = 1;
-            }
-        } else {
-            pages = save_zero_page(f, block, offset, p, bytes_transferred);
-            if (pages == -1) {
-                pages = compress_page_with_multi_thread(f, block, offset,
-                                                        bytes_transferred);
-            }
-        }
-    }
-
-    return pages;
-}
-
-/**
- * ram_find_and_save_block: Finds a dirty page and sends it to f
- *
- * Called within an RCU critical section.
- *
- * Returns:  The number of pages written
- *           0 means no dirty pages
- *
- * @f: QEMUFile where to send the data
- * @last_stage: if we are at the completion stage
- * @bytes_transferred: increase it with the number of transferred bytes
- */
-
-static int ram_find_and_save_block(QEMUFile *f, bool last_stage,
-                                   uint64_t *bytes_transferred)
-{
-    RAMBlock *block = last_seen_block;
-    ram_addr_t offset = last_offset;
-    bool complete_round = false;
-    int pages = 0;
-    MemoryRegion *mr;
-
-    if (!block)
-        block = QLIST_FIRST_RCU(&ram_list.blocks);
-
-    while (true) {
-        mr = block->mr;
-        offset = migration_bitmap_find_and_reset_dirty(mr, offset);
-        if (complete_round && block == last_seen_block &&
-            offset >= last_offset) {
-            break;
-        }
-        if (offset >= block->used_length) {
-            offset = 0;
-            block = QLIST_NEXT_RCU(block, next);
-            if (!block) {
-                block = QLIST_FIRST_RCU(&ram_list.blocks);
-                complete_round = true;
-                ram_bulk_stage = false;
-                if (migrate_use_xbzrle()) {
-                    /* If xbzrle is on, stop using the data compression at this
-                     * point. In theory, xbzrle can do better than compression.
-                     */
-                    flush_compressed_data(f);
-                    compression_switch = false;
-                }
-            }
-        } else {
-            if (compression_switch && migrate_use_compression()) {
-                pages = ram_save_compressed_page(f, block, offset, last_stage,
-                                                 bytes_transferred);
-            } else {
-                pages = ram_save_page(f, block, offset, last_stage,
-                                      bytes_transferred);
-            }
-
-            /* if page is unmodified, continue to the next */
-            if (pages > 0) {
-                last_sent_block = block;
-                break;
-            }
-        }
-    }
-
-    last_seen_block = block;
-    last_offset = offset;
-
-    return pages;
-}
-
-void acct_update_position(QEMUFile *f, size_t size, bool zero)
-{
-    uint64_t pages = size / TARGET_PAGE_SIZE;
-    if (zero) {
-        acct_info.dup_pages += pages;
-    } else {
-        acct_info.norm_pages += pages;
-        bytes_transferred += size;
-        qemu_update_position(f, size);
-    }
-}
-
-static ram_addr_t ram_save_remaining(void)
-{
-    return migration_dirty_pages;
-}
-
-uint64_t ram_bytes_remaining(void)
-{
-    return ram_save_remaining() * TARGET_PAGE_SIZE;
-}
-
-uint64_t ram_bytes_transferred(void)
-{
-    return bytes_transferred;
-}
-
-uint64_t ram_bytes_total(void)
-{
-    RAMBlock *block;
-    uint64_t total = 0;
-
-    rcu_read_lock();
-    QLIST_FOREACH_RCU(block, &ram_list.blocks, next)
-        total += block->used_length;
-    rcu_read_unlock();
-    return total;
-}
-
-void free_xbzrle_decoded_buf(void)
-{
-    g_free(xbzrle_decoded_buf);
-    xbzrle_decoded_buf = NULL;
-}
-
-static void migration_end(void)
-{
-    if (migration_bitmap) {
-        memory_global_dirty_log_stop();
-        g_free(migration_bitmap);
-        migration_bitmap = NULL;
-    }
-
-    XBZRLE_cache_lock();
-    if (XBZRLE.cache) {
-        cache_fini(XBZRLE.cache);
-        g_free(XBZRLE.encoded_buf);
-        g_free(XBZRLE.current_buf);
-        XBZRLE.cache = NULL;
-        XBZRLE.encoded_buf = NULL;
-        XBZRLE.current_buf = NULL;
-    }
-    XBZRLE_cache_unlock();
-}
-
-static void ram_migration_cancel(void *opaque)
-{
-    migration_end();
-}
-
-static void reset_ram_globals(void)
-{
-    last_seen_block = NULL;
-    last_sent_block = NULL;
-    last_offset = 0;
-    last_version = ram_list.version;
-    ram_bulk_stage = true;
-}
-
-#define MAX_WAIT 50 /* ms, half buffered_file limit */
-
-
-/* Each of ram_save_setup, ram_save_iterate and ram_save_complete has
- * long-running RCU critical section.  When rcu-reclaims in the code
- * start to become numerous it will be necessary to reduce the
- * granularity of these critical sections.
- */
-
-static int ram_save_setup(QEMUFile *f, void *opaque)
-{
-    RAMBlock *block;
-    int64_t ram_bitmap_pages; /* Size of bitmap in pages, including gaps */
-
-    mig_throttle_on = false;
-    dirty_rate_high_cnt = 0;
-    bitmap_sync_count = 0;
-    migration_bitmap_sync_init();
-
-    if (migrate_use_xbzrle()) {
-        XBZRLE_cache_lock();
-        XBZRLE.cache = cache_init(migrate_xbzrle_cache_size() /
-                                  TARGET_PAGE_SIZE,
-                                  TARGET_PAGE_SIZE);
-        if (!XBZRLE.cache) {
-            XBZRLE_cache_unlock();
-            error_report("Error creating cache");
-            return -1;
-        }
-        XBZRLE_cache_unlock();
-
-        /* We prefer not to abort if there is no memory */
-        XBZRLE.encoded_buf = g_try_malloc0(TARGET_PAGE_SIZE);
-        if (!XBZRLE.encoded_buf) {
-            error_report("Error allocating encoded_buf");
-            return -1;
-        }
-
-        XBZRLE.current_buf = g_try_malloc(TARGET_PAGE_SIZE);
-        if (!XBZRLE.current_buf) {
-            error_report("Error allocating current_buf");
-            g_free(XBZRLE.encoded_buf);
-            XBZRLE.encoded_buf = NULL;
-            return -1;
-        }
-
-        acct_clear();
-    }
-
-    /* iothread lock needed for ram_list.dirty_memory[] */
-    qemu_mutex_lock_iothread();
-    qemu_mutex_lock_ramlist();
-    rcu_read_lock();
-    bytes_transferred = 0;
-    reset_ram_globals();
-
-    ram_bitmap_pages = last_ram_offset() >> TARGET_PAGE_BITS;
-    migration_bitmap = bitmap_new(ram_bitmap_pages);
-    bitmap_set(migration_bitmap, 0, ram_bitmap_pages);
-
-    /*
-     * Count the total number of pages used by ram blocks not including any
-     * gaps due to alignment or unplugs.
-     */
-    migration_dirty_pages = ram_bytes_total() >> TARGET_PAGE_BITS;
-
-    memory_global_dirty_log_start();
-    migration_bitmap_sync();
-    qemu_mutex_unlock_ramlist();
-    qemu_mutex_unlock_iothread();
-
-    qemu_put_be64(f, ram_bytes_total() | RAM_SAVE_FLAG_MEM_SIZE);
-
-    QLIST_FOREACH_RCU(block, &ram_list.blocks, next) {
-        qemu_put_byte(f, strlen(block->idstr));
-        qemu_put_buffer(f, (uint8_t *)block->idstr, strlen(block->idstr));
-        qemu_put_be64(f, block->used_length);
-    }
-
-    rcu_read_unlock();
-
-    ram_control_before_iterate(f, RAM_CONTROL_SETUP);
-    ram_control_after_iterate(f, RAM_CONTROL_SETUP);
-
-    qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
-
-    return 0;
-}
-
-static int ram_save_iterate(QEMUFile *f, void *opaque)
-{
-    int ret;
-    int i;
-    int64_t t0;
-    int pages_sent = 0;
-
-    rcu_read_lock();
-    if (ram_list.version != last_version) {
-        reset_ram_globals();
-    }
-
-    /* Read version before ram_list.blocks */
-    smp_rmb();
-
-    ram_control_before_iterate(f, RAM_CONTROL_ROUND);
-
-    t0 = qemu_clock_get_ns(QEMU_CLOCK_REALTIME);
-    i = 0;
-    while ((ret = qemu_file_rate_limit(f)) == 0) {
-        int pages;
-
-        pages = ram_find_and_save_block(f, false, &bytes_transferred);
-        /* no more pages to sent */
-        if (pages == 0) {
-            break;
-        }
-        pages_sent += pages;
-        acct_info.iterations++;
-        check_guest_throttling();
-        /* we want to check in the 1st loop, just in case it was the 1st time
-           and we had to sync the dirty bitmap.
-           qemu_get_clock_ns() is a bit expensive, so we only check each some
-           iterations
-        */
-        if ((i & 63) == 0) {
-            uint64_t t1 = (qemu_clock_get_ns(QEMU_CLOCK_REALTIME) - t0) / 1000000;
-            if (t1 > MAX_WAIT) {
-                DPRINTF("big wait: %" PRIu64 " milliseconds, %d iterations\n",
-                        t1, i);
-                break;
-            }
-        }
-        i++;
-    }
-    flush_compressed_data(f);
-    rcu_read_unlock();
-
-    /*
-     * Must occur before EOS (or any QEMUFile operation)
-     * because of RDMA protocol.
-     */
-    ram_control_after_iterate(f, RAM_CONTROL_ROUND);
-
-    qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
-    bytes_transferred += 8;
-
-    ret = qemu_file_get_error(f);
-    if (ret < 0) {
-        return ret;
-    }
-
-    return pages_sent;
-}
-
-/* Called with iothread lock */
-static int ram_save_complete(QEMUFile *f, void *opaque)
-{
-    rcu_read_lock();
-
-    migration_bitmap_sync();
-
-    ram_control_before_iterate(f, RAM_CONTROL_FINISH);
-
-    /* try transferring iterative blocks of memory */
-
-    /* flush all remaining blocks regardless of rate limiting */
-    while (true) {
-        int pages;
-
-        pages = ram_find_and_save_block(f, true, &bytes_transferred);
-        /* no more blocks to sent */
-        if (pages == 0) {
-            break;
-        }
-    }
-
-    flush_compressed_data(f);
-    ram_control_after_iterate(f, RAM_CONTROL_FINISH);
-    migration_end();
-
-    rcu_read_unlock();
-    qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
-
-    return 0;
-}
-
-static uint64_t ram_save_pending(QEMUFile *f, void *opaque, uint64_t max_size)
-{
-    uint64_t remaining_size;
-
-    remaining_size = ram_save_remaining() * TARGET_PAGE_SIZE;
-
-    if (remaining_size < max_size) {
-        qemu_mutex_lock_iothread();
-        rcu_read_lock();
-        migration_bitmap_sync();
-        rcu_read_unlock();
-        qemu_mutex_unlock_iothread();
-        remaining_size = ram_save_remaining() * TARGET_PAGE_SIZE;
-    }
-    return remaining_size;
-}
-
-static int load_xbzrle(QEMUFile *f, ram_addr_t addr, void *host)
-{
-    unsigned int xh_len;
-    int xh_flags;
-
-    if (!xbzrle_decoded_buf) {
-        xbzrle_decoded_buf = g_malloc(TARGET_PAGE_SIZE);
-    }
-
-    /* extract RLE header */
-    xh_flags = qemu_get_byte(f);
-    xh_len = qemu_get_be16(f);
-
-    if (xh_flags != ENCODING_FLAG_XBZRLE) {
-        error_report("Failed to load XBZRLE page - wrong compression!");
-        return -1;
-    }
-
-    if (xh_len > TARGET_PAGE_SIZE) {
-        error_report("Failed to load XBZRLE page - len overflow!");
-        return -1;
-    }
-    /* load data and decode */
-    qemu_get_buffer(f, xbzrle_decoded_buf, xh_len);
-
-    /* decode RLE */
-    if (xbzrle_decode_buffer(xbzrle_decoded_buf, xh_len, host,
-                             TARGET_PAGE_SIZE) == -1) {
-        error_report("Failed to load XBZRLE page - decode error!");
-        return -1;
-    }
-
-    return 0;
-}
-
-/* Must be called from within a rcu critical section.
- * Returns a pointer from within the RCU-protected ram_list.
- */
-static inline void *host_from_stream_offset(QEMUFile *f,
-                                            ram_addr_t offset,
-                                            int flags)
-{
-    static RAMBlock *block = NULL;
-    char id[256];
-    uint8_t len;
-
-    if (flags & RAM_SAVE_FLAG_CONTINUE) {
-        if (!block || block->max_length <= offset) {
-            error_report("Ack, bad migration stream!");
-            return NULL;
-        }
-
-        return memory_region_get_ram_ptr(block->mr) + offset;
-    }
-
-    len = qemu_get_byte(f);
-    qemu_get_buffer(f, (uint8_t *)id, len);
-    id[len] = 0;
-
-    QLIST_FOREACH_RCU(block, &ram_list.blocks, next) {
-        if (!strncmp(id, block->idstr, sizeof(id)) &&
-            block->max_length > offset) {
-            return memory_region_get_ram_ptr(block->mr) + offset;
-        }
-    }
-
-    error_report("Can't find block %s!", id);
-    return NULL;
-}
-
-/*
- * If a page (or a whole RDMA chunk) has been
- * determined to be zero, then zap it.
- */
-void ram_handle_compressed(void *host, uint8_t ch, uint64_t size)
-{
-    if (ch != 0 || !is_zero_range(host, size)) {
-        memset(host, ch, size);
-    }
-}
-
-static void *do_data_decompress(void *opaque)
-{
-    DecompressParam *param = opaque;
-    unsigned long pagesize;
-
-    while (!quit_decomp_thread) {
-        qemu_mutex_lock(&param->mutex);
-        while (!param->start && !quit_decomp_thread) {
-            qemu_cond_wait(&param->cond, &param->mutex);
-            pagesize = TARGET_PAGE_SIZE;
-            if (!quit_decomp_thread) {
-                /* uncompress() will return failed in some case, especially
-                 * when the page is dirted when doing the compression, it's
-                 * not a problem because the dirty page will be retransferred
-                 * and uncompress() won't break the data in other pages.
-                 */
-                uncompress((Bytef *)param->des, &pagesize,
-                           (const Bytef *)param->compbuf, param->len);
-            }
-            param->start = false;
-        }
-        qemu_mutex_unlock(&param->mutex);
-    }
-
-    return NULL;
-}
-
-void migrate_decompress_threads_create(void)
-{
-    int i, thread_count;
-
-    thread_count = migrate_decompress_threads();
-    decompress_threads = g_new0(QemuThread, thread_count);
-    decomp_param = g_new0(DecompressParam, thread_count);
-    compressed_data_buf = g_malloc0(compressBound(TARGET_PAGE_SIZE));
-    quit_decomp_thread = false;
-    for (i = 0; i < thread_count; i++) {
-        qemu_mutex_init(&decomp_param[i].mutex);
-        qemu_cond_init(&decomp_param[i].cond);
-        decomp_param[i].compbuf = g_malloc0(compressBound(TARGET_PAGE_SIZE));
-        qemu_thread_create(decompress_threads + i, "decompress",
-                           do_data_decompress, decomp_param + i,
-                           QEMU_THREAD_JOINABLE);
-    }
-}
-
-void migrate_decompress_threads_join(void)
-{
-    int i, thread_count;
-
-    quit_decomp_thread = true;
-    thread_count = migrate_decompress_threads();
-    for (i = 0; i < thread_count; i++) {
-        qemu_mutex_lock(&decomp_param[i].mutex);
-        qemu_cond_signal(&decomp_param[i].cond);
-        qemu_mutex_unlock(&decomp_param[i].mutex);
-    }
-    for (i = 0; i < thread_count; i++) {
-        qemu_thread_join(decompress_threads + i);
-        qemu_mutex_destroy(&decomp_param[i].mutex);
-        qemu_cond_destroy(&decomp_param[i].cond);
-        g_free(decomp_param[i].compbuf);
-    }
-    g_free(decompress_threads);
-    g_free(decomp_param);
-    g_free(compressed_data_buf);
-    decompress_threads = NULL;
-    decomp_param = NULL;
-    compressed_data_buf = NULL;
-}
-
-static void decompress_data_with_multi_threads(uint8_t *compbuf,
-                                               void *host, int len)
-{
-    int idx, thread_count;
-
-    thread_count = migrate_decompress_threads();
-    while (true) {
-        for (idx = 0; idx < thread_count; idx++) {
-            if (!decomp_param[idx].start) {
-                memcpy(decomp_param[idx].compbuf, compbuf, len);
-                decomp_param[idx].des = host;
-                decomp_param[idx].len = len;
-                start_decompression(&decomp_param[idx]);
-                break;
-            }
-        }
-        if (idx < thread_count) {
-            break;
-        }
-    }
-}
-
-static int ram_load(QEMUFile *f, void *opaque, int version_id)
-{
-    int flags = 0, ret = 0;
-    static uint64_t seq_iter;
-    int len = 0;
-
-    seq_iter++;
-
-    if (version_id != 4) {
-        ret = -EINVAL;
-    }
-
-    /* This RCU critical section can be very long running.
-     * When RCU reclaims in the code start to become numerous,
-     * it will be necessary to reduce the granularity of this
-     * critical section.
-     */
-    rcu_read_lock();
-    while (!ret && !(flags & RAM_SAVE_FLAG_EOS)) {
-        ram_addr_t addr, total_ram_bytes;
-        void *host;
-        uint8_t ch;
-
-        addr = qemu_get_be64(f);
-        flags = addr & ~TARGET_PAGE_MASK;
-        addr &= TARGET_PAGE_MASK;
-
-        switch (flags & ~RAM_SAVE_FLAG_CONTINUE) {
-        case RAM_SAVE_FLAG_MEM_SIZE:
-            /* Synchronize RAM block list */
-            total_ram_bytes = addr;
-            while (!ret && total_ram_bytes) {
-                RAMBlock *block;
-                uint8_t len;
-                char id[256];
-                ram_addr_t length;
-
-                len = qemu_get_byte(f);
-                qemu_get_buffer(f, (uint8_t *)id, len);
-                id[len] = 0;
-                length = qemu_get_be64(f);
-
-                QLIST_FOREACH_RCU(block, &ram_list.blocks, next) {
-                    if (!strncmp(id, block->idstr, sizeof(id))) {
-                        if (length != block->used_length) {
-                            Error *local_err = NULL;
-
-                            ret = qemu_ram_resize(block->offset, length, &local_err);
-                            if (local_err) {
-                                error_report_err(local_err);
-                            }
-                        }
-                        break;
-                    }
-                }
-
-                if (!block) {
-                    error_report("Unknown ramblock \"%s\", cannot "
-                                 "accept migration", id);
-                    ret = -EINVAL;
-                }
-
-                total_ram_bytes -= length;
-            }
-            break;
-        case RAM_SAVE_FLAG_COMPRESS:
-            host = host_from_stream_offset(f, addr, flags);
-            if (!host) {
-                error_report("Illegal RAM offset " RAM_ADDR_FMT, addr);
-                ret = -EINVAL;
-                break;
-            }
-            ch = qemu_get_byte(f);
-            ram_handle_compressed(host, ch, TARGET_PAGE_SIZE);
-            break;
-        case RAM_SAVE_FLAG_PAGE:
-            host = host_from_stream_offset(f, addr, flags);
-            if (!host) {
-                error_report("Illegal RAM offset " RAM_ADDR_FMT, addr);
-                ret = -EINVAL;
-                break;
-            }
-            qemu_get_buffer(f, host, TARGET_PAGE_SIZE);
-            break;
-        case RAM_SAVE_FLAG_COMPRESS_PAGE:
-            host = host_from_stream_offset(f, addr, flags);
-            if (!host) {
-                error_report("Invalid RAM offset " RAM_ADDR_FMT, addr);
-                ret = -EINVAL;
-                break;
-            }
-
-            len = qemu_get_be32(f);
-            if (len < 0 || len > compressBound(TARGET_PAGE_SIZE)) {
-                error_report("Invalid compressed data length: %d", len);
-                ret = -EINVAL;
-                break;
-            }
-            qemu_get_buffer(f, compressed_data_buf, len);
-            decompress_data_with_multi_threads(compressed_data_buf, host, len);
-            break;
-        case RAM_SAVE_FLAG_XBZRLE:
-            host = host_from_stream_offset(f, addr, flags);
-            if (!host) {
-                error_report("Illegal RAM offset " RAM_ADDR_FMT, addr);
-                ret = -EINVAL;
-                break;
-            }
-            if (load_xbzrle(f, addr, host) < 0) {
-                error_report("Failed to decompress XBZRLE page at "
-                             RAM_ADDR_FMT, addr);
-                ret = -EINVAL;
-                break;
-            }
-            break;
-        case RAM_SAVE_FLAG_EOS:
-            /* normal exit */
-            break;
-        default:
-            if (flags & RAM_SAVE_FLAG_HOOK) {
-                ram_control_load_hook(f, flags);
-            } else {
-                error_report("Unknown combination of migration flags: %#x",
-                             flags);
-                ret = -EINVAL;
-            }
-        }
-        if (!ret) {
-            ret = qemu_file_get_error(f);
-        }
-    }
-
-    rcu_read_unlock();
-    DPRINTF("Completed load of VM with exit code %d seq iteration "
-            "%" PRIu64 "\n", ret, seq_iter);
-    return ret;
-}
-
-static SaveVMHandlers savevm_ram_handlers = {
-    .save_live_setup = ram_save_setup,
-    .save_live_iterate = ram_save_iterate,
-    .save_live_complete = ram_save_complete,
-    .save_live_pending = ram_save_pending,
-    .load_state = ram_load,
-    .cancel = ram_migration_cancel,
-};
-
-void ram_mig_init(void)
-{
-    qemu_mutex_init(&XBZRLE.lock);
-    register_savevm_live(NULL, "ram", 0, 4, &savevm_ram_handlers, NULL);
-}
-
 struct soundhw {
     const char *name;
     const char *descr;
@@ -1869,52 +307,3 @@
 
     return info;
 }
-
-/* Stub function that's gets run on the vcpu when its brought out of the
-   VM to run inside qemu via async_run_on_cpu()*/
-static void mig_sleep_cpu(void *opq)
-{
-    qemu_mutex_unlock_iothread();
-    g_usleep(30*1000);
-    qemu_mutex_lock_iothread();
-}
-
-/* To reduce the dirty rate explicitly disallow the VCPUs from spending
-   much time in the VM. The migration thread will try to catchup.
-   Workload will experience a performance drop.
-*/
-static void mig_throttle_guest_down(void)
-{
-    CPUState *cpu;
-
-    qemu_mutex_lock_iothread();
-    CPU_FOREACH(cpu) {
-        async_run_on_cpu(cpu, mig_sleep_cpu, NULL);
-    }
-    qemu_mutex_unlock_iothread();
-}
-
-static void check_guest_throttling(void)
-{
-    static int64_t t0;
-    int64_t        t1;
-
-    if (!mig_throttle_on) {
-        return;
-    }
-
-    if (!t0)  {
-        t0 = qemu_clock_get_ns(QEMU_CLOCK_REALTIME);
-        return;
-    }
-
-    t1 = qemu_clock_get_ns(QEMU_CLOCK_REALTIME);
-
-    /* If it has been more than 40 ms since the last time the guest
-     * was throttled then do it again.
-     */
-    if (40 < (t1-t0)/1000000) {
-        mig_throttle_guest_down();
-        t0 = t1;
-    }
-}
diff --git a/cpus.c b/cpus.c
index f38b858..b85fb5f 100644
--- a/cpus.c
+++ b/cpus.c
@@ -480,6 +480,7 @@
     .name = "timer/icount",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = icount_state_needed,
     .fields = (VMStateField[]) {
         VMSTATE_INT64(qemu_icount_bias, TimersState),
         VMSTATE_INT64(qemu_icount, TimersState),
@@ -497,13 +498,9 @@
         VMSTATE_INT64_V(cpu_clock_offset, TimersState, 2),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &icount_vmstate_timers,
-            .needed = icount_state_needed,
-        }, {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &icount_vmstate_timers,
+        NULL
     }
 };
 
diff --git a/default-configs/mips-softmmu.mak b/default-configs/mips-softmmu.mak
index fd0607d..44467c3 100644
--- a/default-configs/mips-softmmu.mak
+++ b/default-configs/mips-softmmu.mak
@@ -24,14 +24,9 @@
 CONFIG_IDE_ISA=y
 CONFIG_IDE_PIIX=y
 CONFIG_NE2000_ISA=y
-CONFIG_RC4030=y
-CONFIG_DP8393X=y
-CONFIG_DS1225Y=y
 CONFIG_MIPSNET=y
 CONFIG_PFLASH_CFI01=y
-CONFIG_G364FB=y
 CONFIG_I8259=y
-CONFIG_JAZZ_LED=y
 CONFIG_MC146818RTC=y
 CONFIG_ISA_TESTDEV=y
 CONFIG_EMPTY_SLOT=y
diff --git a/default-configs/mips64-softmmu.mak b/default-configs/mips64-softmmu.mak
index b8c7910..66ed5f9 100644
--- a/default-configs/mips64-softmmu.mak
+++ b/default-configs/mips64-softmmu.mak
@@ -29,6 +29,7 @@
 CONFIG_DS1225Y=y
 CONFIG_MIPSNET=y
 CONFIG_PFLASH_CFI01=y
+CONFIG_JAZZ=y
 CONFIG_G364FB=y
 CONFIG_I8259=y
 CONFIG_JAZZ_LED=y
diff --git a/default-configs/mips64el-softmmu.mak b/default-configs/mips64el-softmmu.mak
index ae4274b..bfca2b2 100644
--- a/default-configs/mips64el-softmmu.mak
+++ b/default-configs/mips64el-softmmu.mak
@@ -31,6 +31,7 @@
 CONFIG_MIPSNET=y
 CONFIG_PFLASH_CFI01=y
 CONFIG_FULONG=y
+CONFIG_JAZZ=y
 CONFIG_G364FB=y
 CONFIG_I8259=y
 CONFIG_JAZZ_LED=y
diff --git a/default-configs/mipsel-softmmu.mak b/default-configs/mipsel-softmmu.mak
index 1e2374b..0162ef0 100644
--- a/default-configs/mipsel-softmmu.mak
+++ b/default-configs/mipsel-softmmu.mak
@@ -24,14 +24,9 @@
 CONFIG_IDE_ISA=y
 CONFIG_IDE_PIIX=y
 CONFIG_NE2000_ISA=y
-CONFIG_RC4030=y
-CONFIG_DP8393X=y
-CONFIG_DS1225Y=y
 CONFIG_MIPSNET=y
 CONFIG_PFLASH_CFI01=y
-CONFIG_G364FB=y
 CONFIG_I8259=y
-CONFIG_JAZZ_LED=y
 CONFIG_MC146818RTC=y
 CONFIG_ISA_TESTDEV=y
 CONFIG_EMPTY_SLOT=y
diff --git a/disas/mips.c b/disas/mips.c
index 1afe0c5..32940fe 100644
--- a/disas/mips.c
+++ b/disas/mips.c
@@ -2238,6 +2238,8 @@
 {"ceil.l.s", "D,S",	0x4600000a, 0xffff003f, WR_D|RD_S|FP_S|FP_D,	0,		I3|I33	},
 {"ceil.w.d", "D,S",	0x4620000e, 0xffff003f, WR_D|RD_S|FP_S|FP_D,	0,		I2	},
 {"ceil.w.s", "D,S",	0x4600000e, 0xffff003f, WR_D|RD_S|FP_S,		0,		I2	},
+{"mfhc0",   "t,G,H",    0x40400000, 0xffe007f8, LCD|WR_t|RD_C0,       0, I33},
+{"mthc0",   "t,G,H",    0x40c00000, 0xffe007f8, COD|RD_t|WR_C0|WR_CC, 0, I33},
 {"cfc0",    "t,G",	0x40400000, 0xffe007ff,	LCD|WR_t|RD_C0,		0,		I1	},
 {"cfc1",    "t,G",	0x44400000, 0xffe007ff,	LCD|WR_t|RD_C1|FP_S,	0,		I1	},
 {"cfc1",    "t,S",	0x44400000, 0xffe007ff,	LCD|WR_t|RD_C1|FP_S,	0,		I1	},
@@ -2407,6 +2409,7 @@
 {"emt",     "",		0x41600be1, 0xffffffff, TRAP,			0,		MT32	},
 {"emt",     "t",	0x41600be1, 0xffe0ffff, TRAP|WR_t,		0,		MT32	},
 {"eret",    "",         0x42000018, 0xffffffff, 0,      		0,		I3|I32	},
+{"eretnc",  "",         0x42000058, 0xffffffff, 0,                    0, I33},
 {"evpe",    "",		0x41600021, 0xffffffff, TRAP,			0,		MT32	},
 {"evpe",    "t",	0x41600021, 0xffe0ffff, TRAP|WR_t,		0,		MT32	},
 {"ext",     "t,r,+A,+C", 0x7c000000, 0xfc00003f, WR_t|RD_s,    		0,		I33	},
diff --git a/docs/migration.txt b/docs/migration.txt
index 0492a45..f6df4be 100644
--- a/docs/migration.txt
+++ b/docs/migration.txt
@@ -257,6 +257,7 @@
     .minimum_version_id = 1,
     .pre_save = ide_drive_pio_pre_save,
     .post_load = ide_drive_pio_post_load,
+    .needed = ide_drive_pio_state_needed,
     .fields = (VMStateField[]) {
         VMSTATE_INT32(req_nb_sectors, IDEState),
         VMSTATE_VARRAY_INT32(io_buffer, IDEState, io_buffer_total_len, 1,
@@ -279,13 +280,9 @@
         .... several fields ....
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection []) {
-        {
-            .vmsd = &vmstate_ide_drive_pio_state,
-            .needed = ide_drive_pio_state_needed,
-        }, {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_ide_drive_pio_state,
+        NULL
     }
 };
 
diff --git a/docs/specs/fw_cfg.txt b/docs/specs/fw_cfg.txt
index 6accd92..74351dd 100644
--- a/docs/specs/fw_cfg.txt
+++ b/docs/specs/fw_cfg.txt
@@ -203,3 +203,24 @@
 
 NOTE: This function is deprecated, and will be completely removed
 starting with QEMU v2.4.
+
+== Externally Provided Items ==
+
+As of v2.4, "file" fw_cfg items (i.e., items with selector keys above
+FW_CFG_FILE_FIRST, and with a corresponding entry in the fw_cfg file
+directory structure) may be inserted via the QEMU command line, using
+the following syntax:
+
+    -fw_cfg [name=]<item_name>,file=<path>
+
+where <item_name> is the fw_cfg item name, and <path> is the location
+on the host file system of a file containing the data to be inserted.
+
+NOTE: Users *SHOULD* choose item names beginning with the prefix "opt/"
+when using the "-fw_cfg" command line option, to avoid conflicting with
+item names used internally by QEMU. For instance:
+
+    -fw_cfg name=opt/my_item_name,file=./my_blob.bin
+
+Similarly, QEMU developers *SHOULD NOT* use item names prefixed with
+"opt/" when inserting items programmatically, e.g. via fw_cfg_add_file().
diff --git a/exec.c b/exec.c
index 487583b..76bfc4a 100644
--- a/exec.c
+++ b/exec.c
@@ -454,6 +454,7 @@
     .name = "cpu_common/exception_index",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = cpu_common_exception_index_needed,
     .fields = (VMStateField[]) {
         VMSTATE_INT32(exception_index, CPUState),
         VMSTATE_END_OF_LIST()
@@ -471,13 +472,9 @@
         VMSTATE_UINT32(interrupt_request, CPUState),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_cpu_common_exception_index,
-            .needed = cpu_common_exception_index_needed,
-        } , {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_cpu_common_exception_index,
+        NULL
     }
 };
 
@@ -3348,14 +3345,20 @@
     return res;
 }
 
-void qemu_ram_foreach_block(RAMBlockIterFunc func, void *opaque)
+int qemu_ram_foreach_block(RAMBlockIterFunc func, void *opaque)
 {
     RAMBlock *block;
+    int ret = 0;
 
     rcu_read_lock();
     QLIST_FOREACH_RCU(block, &ram_list.blocks, next) {
-        func(block->host, block->offset, block->used_length, opaque);
+        ret = func(block->idstr, block->host, block->offset,
+                   block->used_length, opaque);
+        if (ret) {
+            break;
+        }
     }
     rcu_read_unlock();
+    return ret;
 }
 #endif
diff --git a/hw/9pfs/virtio-9p-device.c b/hw/9pfs/virtio-9p-device.c
index 60f9ff9..3f4c9e7 100644
--- a/hw/9pfs/virtio-9p-device.c
+++ b/hw/9pfs/virtio-9p-device.c
@@ -140,7 +140,8 @@
 /* virtio-9p device */
 
 static Property virtio_9p_properties[] = {
-    DEFINE_VIRTIO_9P_PROPERTIES(V9fsState, fsconf),
+    DEFINE_PROP_STRING("mount_tag", V9fsState, fsconf.tag),
+    DEFINE_PROP_STRING("fsdev", V9fsState, fsconf.fsdev_id),
     DEFINE_PROP_END_OF_LIST(),
 };
 
diff --git a/hw/9pfs/virtio-9p.h b/hw/9pfs/virtio-9p.h
index 58dafa9..2e7d488 100644
--- a/hw/9pfs/virtio-9p.h
+++ b/hw/9pfs/virtio-9p.h
@@ -391,8 +391,4 @@
 #define VIRTIO_9P(obj) \
         OBJECT_CHECK(V9fsState, (obj), TYPE_VIRTIO_9P)
 
-#define DEFINE_VIRTIO_9P_PROPERTIES(_state, _field)             \
-        DEFINE_PROP_STRING("mount_tag", _state, _field.tag),    \
-        DEFINE_PROP_STRING("fsdev", _state, _field.fsdev_id)
-
 #endif
diff --git a/hw/acpi/ich9.c b/hw/acpi/ich9.c
index 25bc023..8a64ffb 100644
--- a/hw/acpi/ich9.c
+++ b/hw/acpi/ich9.c
@@ -152,6 +152,7 @@
     .version_id = 1,
     .minimum_version_id = 1,
     .minimum_version_id_old = 1,
+    .needed = vmstate_test_use_memhp,
     .fields      = (VMStateField[]) {
         VMSTATE_MEMORY_HOTPLUG(acpi_memory_hotplug, ICH9LPCPMRegs),
         VMSTATE_END_OF_LIST()
@@ -175,12 +176,9 @@
         VMSTATE_UINT32(smi_sts, ICH9LPCPMRegs),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_memhp_state,
-            .needed = vmstate_test_use_memhp,
-        },
-        VMSTATE_END_OF_LIST()
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_memhp_state,
+        NULL
     }
 };
 
diff --git a/hw/acpi/piix4.c b/hw/acpi/piix4.c
index b730ca6..3bd1d5a 100644
--- a/hw/acpi/piix4.c
+++ b/hw/acpi/piix4.c
@@ -260,6 +260,7 @@
     .version_id = 1,
     .minimum_version_id = 1,
     .minimum_version_id_old = 1,
+    .needed = vmstate_test_use_memhp,
     .fields      = (VMStateField[]) {
         VMSTATE_MEMORY_HOTPLUG(acpi_memory_hotplug, PIIX4PMState),
         VMSTATE_END_OF_LIST()
@@ -298,12 +299,9 @@
                             vmstate_test_use_acpi_pci_hotplug),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_memhp_state,
-            .needed = vmstate_test_use_memhp,
-        },
-        VMSTATE_END_OF_LIST()
+    .subsections = (const VMStateDescription*[]) {
+         &vmstate_memhp_state,
+         NULL
     }
 };
 
diff --git a/hw/block/fdc.c b/hw/block/fdc.c
index 6e79459..5e1b67e 100644
--- a/hw/block/fdc.c
+++ b/hw/block/fdc.c
@@ -699,6 +699,7 @@
     .name = "fdrive/media_changed",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = fdrive_media_changed_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT8(media_changed, FDrive),
         VMSTATE_END_OF_LIST()
@@ -716,6 +717,7 @@
     .name = "fdrive/media_rate",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = fdrive_media_rate_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT8(media_rate, FDrive),
         VMSTATE_END_OF_LIST()
@@ -733,6 +735,7 @@
     .name = "fdrive/perpendicular",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = fdrive_perpendicular_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT8(perpendicular, FDrive),
         VMSTATE_END_OF_LIST()
@@ -756,19 +759,11 @@
         VMSTATE_UINT8(sect, FDrive),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_fdrive_media_changed,
-            .needed = &fdrive_media_changed_needed,
-        } , {
-            .vmsd = &vmstate_fdrive_media_rate,
-            .needed = &fdrive_media_rate_needed,
-        } , {
-            .vmsd = &vmstate_fdrive_perpendicular,
-            .needed = &fdrive_perpendicular_needed,
-        } , {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_fdrive_media_changed,
+        &vmstate_fdrive_media_rate,
+        &vmstate_fdrive_perpendicular,
+        NULL
     }
 };
 
@@ -833,6 +828,7 @@
     .name = "fdc/reset_sensei",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = fdc_reset_sensei_needed,
     .fields = (VMStateField[]) {
         VMSTATE_INT32(reset_sensei, FDCtrl),
         VMSTATE_END_OF_LIST()
@@ -850,6 +846,7 @@
     .name = "fdc/result_timer",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = fdc_result_timer_needed,
     .fields = (VMStateField[]) {
         VMSTATE_TIMER_PTR(result_timer, FDCtrl),
         VMSTATE_END_OF_LIST()
@@ -867,6 +864,7 @@
     .name = "fdc/phase",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = fdc_phase_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT8(phase, FDCtrl),
         VMSTATE_END_OF_LIST()
@@ -911,19 +909,11 @@
                              vmstate_fdrive, FDrive),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_fdc_reset_sensei,
-            .needed = fdc_reset_sensei_needed,
-        } , {
-            .vmsd = &vmstate_fdc_result_timer,
-            .needed = fdc_result_timer_needed,
-        } , {
-            .vmsd = &vmstate_fdc_phase,
-            .needed = fdc_phase_needed,
-        } , {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_fdc_reset_sensei,
+        &vmstate_fdc_result_timer,
+        &vmstate_fdc_phase,
+        NULL
     }
 };
 
diff --git a/hw/char/serial.c b/hw/char/serial.c
index 55011cf..513d73c 100644
--- a/hw/char/serial.c
+++ b/hw/char/serial.c
@@ -662,6 +662,7 @@
     .name = "serial/thr_ipending",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = serial_thr_ipending_needed,
     .fields = (VMStateField[]) {
         VMSTATE_INT32(thr_ipending, SerialState),
         VMSTATE_END_OF_LIST()
@@ -678,6 +679,7 @@
     .name = "serial/tsr",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = serial_tsr_needed,
     .fields = (VMStateField[]) {
         VMSTATE_INT32(tsr_retry, SerialState),
         VMSTATE_UINT8(thr, SerialState),
@@ -697,6 +699,7 @@
     .name = "serial/recv_fifo",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = serial_recv_fifo_needed,
     .fields = (VMStateField[]) {
         VMSTATE_STRUCT(recv_fifo, SerialState, 1, vmstate_fifo8, Fifo8),
         VMSTATE_END_OF_LIST()
@@ -713,6 +716,7 @@
     .name = "serial/xmit_fifo",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = serial_xmit_fifo_needed,
     .fields = (VMStateField[]) {
         VMSTATE_STRUCT(xmit_fifo, SerialState, 1, vmstate_fifo8, Fifo8),
         VMSTATE_END_OF_LIST()
@@ -729,6 +733,7 @@
     .name = "serial/fifo_timeout_timer",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = serial_fifo_timeout_timer_needed,
     .fields = (VMStateField[]) {
         VMSTATE_TIMER_PTR(fifo_timeout_timer, SerialState),
         VMSTATE_END_OF_LIST()
@@ -745,6 +750,7 @@
     .name = "serial/timeout_ipending",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = serial_timeout_ipending_needed,
     .fields = (VMStateField[]) {
         VMSTATE_INT32(timeout_ipending, SerialState),
         VMSTATE_END_OF_LIST()
@@ -760,6 +766,7 @@
 static const VMStateDescription vmstate_serial_poll = {
     .name = "serial/poll",
     .version_id = 1,
+    .needed = serial_poll_needed,
     .minimum_version_id = 1,
     .fields = (VMStateField[]) {
         VMSTATE_INT32(poll_msl, SerialState),
@@ -788,31 +795,15 @@
         VMSTATE_UINT8_V(fcr_vmstate, SerialState, 3),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_serial_thr_ipending,
-            .needed = &serial_thr_ipending_needed,
-        } , {
-            .vmsd = &vmstate_serial_tsr,
-            .needed = &serial_tsr_needed,
-        } , {
-            .vmsd = &vmstate_serial_recv_fifo,
-            .needed = &serial_recv_fifo_needed,
-        } , {
-            .vmsd = &vmstate_serial_xmit_fifo,
-            .needed = &serial_xmit_fifo_needed,
-        } , {
-            .vmsd = &vmstate_serial_fifo_timeout_timer,
-            .needed = &serial_fifo_timeout_timer_needed,
-        } , {
-            .vmsd = &vmstate_serial_timeout_ipending,
-            .needed = &serial_timeout_ipending_needed,
-        } , {
-            .vmsd = &vmstate_serial_poll,
-            .needed = &serial_poll_needed,
-        } , {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_serial_thr_ipending,
+        &vmstate_serial_tsr,
+        &vmstate_serial_recv_fifo,
+        &vmstate_serial_xmit_fifo,
+        &vmstate_serial_fifo_timeout_timer,
+        &vmstate_serial_timeout_ipending,
+        &vmstate_serial_poll,
+        NULL
     }
 };
 
diff --git a/hw/char/virtio-serial-bus.c b/hw/char/virtio-serial-bus.c
index f893523..d451b22 100644
--- a/hw/char/virtio-serial-bus.c
+++ b/hw/char/virtio-serial-bus.c
@@ -1083,7 +1083,8 @@
 }
 
 static Property virtio_serial_properties[] = {
-    DEFINE_VIRTIO_SERIAL_PROPERTIES(VirtIOSerial, serial),
+    DEFINE_PROP_UINT32("max_ports", VirtIOSerial, serial.max_virtserial_ports,
+                                                  31),
     DEFINE_PROP_END_OF_LIST(),
 };
 
diff --git a/hw/display/Makefile.objs b/hw/display/Makefile.objs
index 3ea106d..61c80f3 100644
--- a/hw/display/Makefile.objs
+++ b/hw/display/Makefile.objs
@@ -34,3 +34,5 @@
 obj-$(CONFIG_VGA) += vga.o
 
 common-obj-$(CONFIG_QXL) += qxl.o qxl-logger.o qxl-render.o
+
+obj-$(CONFIG_VIRTIO) += virtio-gpu.o
diff --git a/hw/display/qxl.c b/hw/display/qxl.c
index b220e2d..722146e 100644
--- a/hw/display/qxl.c
+++ b/hw/display/qxl.c
@@ -2220,6 +2220,7 @@
     .name               = "qxl/monitors-config",
     .version_id         = 1,
     .minimum_version_id = 1,
+    .needed = qxl_monitors_config_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64(guest_monitors_config, PCIQXLDevice),
         VMSTATE_END_OF_LIST()
@@ -2253,13 +2254,9 @@
         VMSTATE_UINT64(guest_cursor, PCIQXLDevice),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &qxl_vmstate_monitors_config,
-            .needed = qxl_monitors_config_needed,
-        }, {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &qxl_vmstate_monitors_config,
+        NULL
     }
 };
 
diff --git a/hw/display/vga-pci.c b/hw/display/vga-pci.c
index ff5dfb2..0ed44c7 100644
--- a/hw/display/vga-pci.c
+++ b/hw/display/vga-pci.c
@@ -54,9 +54,7 @@
     VGACommonState vga;
     uint32_t flags;
     MemoryRegion mmio;
-    MemoryRegion ioport;
-    MemoryRegion bochs;
-    MemoryRegion qext;
+    MemoryRegion mrs[3];
 } PCIVGAState;
 
 #define TYPE_PCI_VGA "pci-vga"
@@ -76,16 +74,16 @@
 static uint64_t pci_vga_ioport_read(void *ptr, hwaddr addr,
                                     unsigned size)
 {
-    PCIVGAState *d = ptr;
+    VGACommonState *s = ptr;
     uint64_t ret = 0;
 
     switch (size) {
     case 1:
-        ret = vga_ioport_read(&d->vga, addr);
+        ret = vga_ioport_read(s, addr + 0x3c0);
         break;
     case 2:
-        ret  = vga_ioport_read(&d->vga, addr);
-        ret |= vga_ioport_read(&d->vga, addr+1) << 8;
+        ret  = vga_ioport_read(s, addr + 0x3c0);
+        ret |= vga_ioport_read(s, addr + 0x3c1) << 8;
         break;
     }
     return ret;
@@ -94,11 +92,11 @@
 static void pci_vga_ioport_write(void *ptr, hwaddr addr,
                                  uint64_t val, unsigned size)
 {
-    PCIVGAState *d = ptr;
+    VGACommonState *s = ptr;
 
     switch (size) {
     case 1:
-        vga_ioport_write(&d->vga, addr + 0x3c0, val);
+        vga_ioport_write(s, addr + 0x3c0, val);
         break;
     case 2:
         /*
@@ -106,8 +104,8 @@
          * indexed registers with a single word write because the
          * index byte is updated first.
          */
-        vga_ioport_write(&d->vga, addr + 0x3c0, val & 0xff);
-        vga_ioport_write(&d->vga, addr + 0x3c1, (val >> 8) & 0xff);
+        vga_ioport_write(s, addr + 0x3c0, val & 0xff);
+        vga_ioport_write(s, addr + 0x3c1, (val >> 8) & 0xff);
         break;
     }
 }
@@ -125,21 +123,21 @@
 static uint64_t pci_vga_bochs_read(void *ptr, hwaddr addr,
                                    unsigned size)
 {
-    PCIVGAState *d = ptr;
+    VGACommonState *s = ptr;
     int index = addr >> 1;
 
-    vbe_ioport_write_index(&d->vga, 0, index);
-    return vbe_ioport_read_data(&d->vga, 0);
+    vbe_ioport_write_index(s, 0, index);
+    return vbe_ioport_read_data(s, 0);
 }
 
 static void pci_vga_bochs_write(void *ptr, hwaddr addr,
                                 uint64_t val, unsigned size)
 {
-    PCIVGAState *d = ptr;
+    VGACommonState *s = ptr;
     int index = addr >> 1;
 
-    vbe_ioport_write_index(&d->vga, 0, index);
-    vbe_ioport_write_data(&d->vga, 0, val);
+    vbe_ioport_write_index(s, 0, index);
+    vbe_ioport_write_data(s, 0, val);
 }
 
 static const MemoryRegionOps pci_vga_bochs_ops = {
@@ -154,13 +152,13 @@
 
 static uint64_t pci_vga_qext_read(void *ptr, hwaddr addr, unsigned size)
 {
-    PCIVGAState *d = ptr;
+    VGACommonState *s = ptr;
 
     switch (addr) {
     case PCI_VGA_QEXT_REG_SIZE:
         return PCI_VGA_QEXT_SIZE;
     case PCI_VGA_QEXT_REG_BYTEORDER:
-        return d->vga.big_endian_fb ?
+        return s->big_endian_fb ?
             PCI_VGA_QEXT_BIG_ENDIAN : PCI_VGA_QEXT_LITTLE_ENDIAN;
     default:
         return 0;
@@ -170,15 +168,15 @@
 static void pci_vga_qext_write(void *ptr, hwaddr addr,
                                uint64_t val, unsigned size)
 {
-    PCIVGAState *d = ptr;
+    VGACommonState *s = ptr;
 
     switch (addr) {
     case PCI_VGA_QEXT_REG_BYTEORDER:
         if (val == PCI_VGA_QEXT_BIG_ENDIAN) {
-            d->vga.big_endian_fb = true;
+            s->big_endian_fb = true;
         }
         if (val == PCI_VGA_QEXT_LITTLE_ENDIAN) {
-            d->vga.big_endian_fb = false;
+            s->big_endian_fb = false;
         }
         break;
     }
@@ -206,10 +204,34 @@
     .endianness = DEVICE_LITTLE_ENDIAN,
 };
 
+static void pci_std_vga_mmio_region_init(VGACommonState *s,
+                                         MemoryRegion *parent,
+                                         MemoryRegion *subs,
+                                         bool qext)
+{
+    memory_region_init_io(&subs[0], NULL, &pci_vga_ioport_ops, s,
+                          "vga ioports remapped", PCI_VGA_IOPORT_SIZE);
+    memory_region_add_subregion(parent, PCI_VGA_IOPORT_OFFSET,
+                                &subs[0]);
+
+    memory_region_init_io(&subs[1], NULL, &pci_vga_bochs_ops, s,
+                          "bochs dispi interface", PCI_VGA_BOCHS_SIZE);
+    memory_region_add_subregion(parent, PCI_VGA_BOCHS_OFFSET,
+                                &subs[1]);
+
+    if (qext) {
+        memory_region_init_io(&subs[2], NULL, &pci_vga_qext_ops, s,
+                              "qemu extended regs", PCI_VGA_QEXT_SIZE);
+        memory_region_add_subregion(parent, PCI_VGA_QEXT_OFFSET,
+                                    &subs[2]);
+    }
+}
+
 static void pci_std_vga_realize(PCIDevice *dev, Error **errp)
 {
     PCIVGAState *d = PCI_VGA(dev);
     VGACommonState *s = &d->vga;
+    bool qext = false;
 
     /* vga + console init */
     vga_common_init(s, OBJECT(dev), true);
@@ -224,23 +246,12 @@
     /* mmio bar for vga register access */
     if (d->flags & (1 << PCI_VGA_FLAG_ENABLE_MMIO)) {
         memory_region_init(&d->mmio, NULL, "vga.mmio", 4096);
-        memory_region_init_io(&d->ioport, NULL, &pci_vga_ioport_ops, d,
-                              "vga ioports remapped", PCI_VGA_IOPORT_SIZE);
-        memory_region_init_io(&d->bochs, NULL, &pci_vga_bochs_ops, d,
-                              "bochs dispi interface", PCI_VGA_BOCHS_SIZE);
-
-        memory_region_add_subregion(&d->mmio, PCI_VGA_IOPORT_OFFSET,
-                                    &d->ioport);
-        memory_region_add_subregion(&d->mmio, PCI_VGA_BOCHS_OFFSET,
-                                    &d->bochs);
 
         if (d->flags & (1 << PCI_VGA_FLAG_ENABLE_QEXT)) {
-            memory_region_init_io(&d->qext, NULL, &pci_vga_qext_ops, d,
-                                  "qemu extended regs", PCI_VGA_QEXT_SIZE);
-            memory_region_add_subregion(&d->mmio, PCI_VGA_QEXT_OFFSET,
-                                        &d->qext);
+            qext = true;
             pci_set_byte(&d->dev.config[PCI_REVISION_ID], 2);
         }
+        pci_std_vga_mmio_region_init(s, &d->mmio, d->mrs, qext);
 
         pci_register_bar(&d->dev, 2, PCI_BASE_ADDRESS_SPACE_MEMORY, &d->mmio);
     }
@@ -262,6 +273,7 @@
 {
     PCIVGAState *d = PCI_VGA(dev);
     VGACommonState *s = &d->vga;
+    bool qext = false;
 
     /* vga + console init */
     vga_common_init(s, OBJECT(dev), false);
@@ -269,23 +281,12 @@
 
     /* mmio bar */
     memory_region_init(&d->mmio, OBJECT(dev), "vga.mmio", 4096);
-    memory_region_init_io(&d->ioport, OBJECT(dev), &pci_vga_ioport_ops, d,
-                          "vga ioports remapped", PCI_VGA_IOPORT_SIZE);
-    memory_region_init_io(&d->bochs, OBJECT(dev), &pci_vga_bochs_ops, d,
-                          "bochs dispi interface", PCI_VGA_BOCHS_SIZE);
-
-    memory_region_add_subregion(&d->mmio, PCI_VGA_IOPORT_OFFSET,
-                                &d->ioport);
-    memory_region_add_subregion(&d->mmio, PCI_VGA_BOCHS_OFFSET,
-                                &d->bochs);
 
     if (d->flags & (1 << PCI_VGA_FLAG_ENABLE_QEXT)) {
-        memory_region_init_io(&d->qext, NULL, &pci_vga_qext_ops, d,
-                              "qemu extended regs", PCI_VGA_QEXT_SIZE);
-        memory_region_add_subregion(&d->mmio, PCI_VGA_QEXT_OFFSET,
-                                    &d->qext);
+        qext = true;
         pci_set_byte(&d->dev.config[PCI_REVISION_ID], 2);
     }
+    pci_std_vga_mmio_region_init(s, &d->mmio, d->mrs, qext);
 
     pci_register_bar(&d->dev, 0, PCI_BASE_ADDRESS_MEM_PREFETCH, &s->vram);
     pci_register_bar(&d->dev, 2, PCI_BASE_ADDRESS_SPACE_MEMORY, &d->mmio);
diff --git a/hw/display/vga.c b/hw/display/vga.c
index d1d296c..b35d523 100644
--- a/hw/display/vga.c
+++ b/hw/display/vga.c
@@ -2035,6 +2035,7 @@
     .name = "vga.endian",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = vga_endian_state_needed,
     .fields = (VMStateField[]) {
         VMSTATE_BOOL(big_endian_fb, VGACommonState),
         VMSTATE_END_OF_LIST()
@@ -2078,13 +2079,9 @@
         VMSTATE_UINT32(vbe_bank_mask, VGACommonState),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection []) {
-        {
-            .vmsd = &vmstate_vga_endian,
-            .needed = vga_endian_state_needed,
-        }, {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_vga_endian,
+        NULL
     }
 };
 
diff --git a/hw/display/virtio-gpu.c b/hw/display/virtio-gpu.c
new file mode 100644
index 0000000..4b10ca1
--- /dev/null
+++ b/hw/display/virtio-gpu.c
@@ -0,0 +1,918 @@
+/*
+ * Virtio GPU Device
+ *
+ * Copyright Red Hat, Inc. 2013-2014
+ *
+ * Authors:
+ *     Dave Airlie <airlied@redhat.com>
+ *     Gerd Hoffmann <kraxel@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.
+ * See the COPYING file in the top-level directory.
+ */
+
+#include "qemu-common.h"
+#include "qemu/iov.h"
+#include "ui/console.h"
+#include "trace.h"
+#include "hw/virtio/virtio.h"
+#include "hw/virtio/virtio-gpu.h"
+#include "hw/virtio/virtio-bus.h"
+
+static struct virtio_gpu_simple_resource*
+virtio_gpu_find_resource(VirtIOGPU *g, uint32_t resource_id);
+
+static void update_cursor_data_simple(VirtIOGPU *g,
+                                      struct virtio_gpu_scanout *s,
+                                      uint32_t resource_id)
+{
+    struct virtio_gpu_simple_resource *res;
+    uint32_t pixels;
+
+    res = virtio_gpu_find_resource(g, resource_id);
+    if (!res) {
+        return;
+    }
+
+    if (pixman_image_get_width(res->image)  != s->current_cursor->width ||
+        pixman_image_get_height(res->image) != s->current_cursor->height) {
+        return;
+    }
+
+    pixels = s->current_cursor->width * s->current_cursor->height;
+    memcpy(s->current_cursor->data,
+           pixman_image_get_data(res->image),
+           pixels * sizeof(uint32_t));
+}
+
+static void update_cursor(VirtIOGPU *g, struct virtio_gpu_update_cursor *cursor)
+{
+    struct virtio_gpu_scanout *s;
+
+    if (cursor->pos.scanout_id >= g->conf.max_outputs) {
+        return;
+    }
+    s = &g->scanout[cursor->pos.scanout_id];
+
+    if (cursor->hdr.type != VIRTIO_GPU_CMD_MOVE_CURSOR) {
+        if (!s->current_cursor) {
+            s->current_cursor = cursor_alloc(64, 64);
+        }
+
+        s->current_cursor->hot_x = cursor->hot_x;
+        s->current_cursor->hot_y = cursor->hot_y;
+
+        if (cursor->resource_id > 0) {
+            update_cursor_data_simple(g, s, cursor->resource_id);
+        }
+        dpy_cursor_define(s->con, s->current_cursor);
+    }
+    dpy_mouse_set(s->con, cursor->pos.x, cursor->pos.y,
+                  cursor->resource_id ? 1 : 0);
+}
+
+static void virtio_gpu_get_config(VirtIODevice *vdev, uint8_t *config)
+{
+    VirtIOGPU *g = VIRTIO_GPU(vdev);
+    memcpy(config, &g->virtio_config, sizeof(g->virtio_config));
+}
+
+static void virtio_gpu_set_config(VirtIODevice *vdev, const uint8_t *config)
+{
+    VirtIOGPU *g = VIRTIO_GPU(vdev);
+    struct virtio_gpu_config vgconfig;
+
+    memcpy(&vgconfig, config, sizeof(g->virtio_config));
+
+    if (vgconfig.events_clear) {
+        g->virtio_config.events_read &= ~vgconfig.events_clear;
+    }
+}
+
+static uint64_t virtio_gpu_get_features(VirtIODevice *vdev, uint64_t features)
+{
+    return features;
+}
+
+static void virtio_gpu_notify_event(VirtIOGPU *g, uint32_t event_type)
+{
+    g->virtio_config.events_read |= event_type;
+    virtio_notify_config(&g->parent_obj);
+}
+
+static struct virtio_gpu_simple_resource *
+virtio_gpu_find_resource(VirtIOGPU *g, uint32_t resource_id)
+{
+    struct virtio_gpu_simple_resource *res;
+
+    QTAILQ_FOREACH(res, &g->reslist, next) {
+        if (res->resource_id == resource_id) {
+            return res;
+        }
+    }
+    return NULL;
+}
+
+void virtio_gpu_ctrl_response(VirtIOGPU *g,
+                              struct virtio_gpu_ctrl_command *cmd,
+                              struct virtio_gpu_ctrl_hdr *resp,
+                              size_t resp_len)
+{
+    size_t s;
+
+    if (cmd->cmd_hdr.flags & VIRTIO_GPU_FLAG_FENCE) {
+        resp->flags |= VIRTIO_GPU_FLAG_FENCE;
+        resp->fence_id = cmd->cmd_hdr.fence_id;
+        resp->ctx_id = cmd->cmd_hdr.ctx_id;
+    }
+    s = iov_from_buf(cmd->elem.in_sg, cmd->elem.in_num, 0, resp, resp_len);
+    if (s != resp_len) {
+        qemu_log_mask(LOG_GUEST_ERROR,
+                      "%s: response size incorrect %zu vs %zu\n",
+                      __func__, s, resp_len);
+    }
+    virtqueue_push(cmd->vq, &cmd->elem, s);
+    virtio_notify(VIRTIO_DEVICE(g), cmd->vq);
+    cmd->finished = true;
+}
+
+void virtio_gpu_ctrl_response_nodata(VirtIOGPU *g,
+                                     struct virtio_gpu_ctrl_command *cmd,
+                                     enum virtio_gpu_ctrl_type type)
+{
+    struct virtio_gpu_ctrl_hdr resp;
+
+    memset(&resp, 0, sizeof(resp));
+    resp.type = type;
+    virtio_gpu_ctrl_response(g, cmd, &resp, sizeof(resp));
+}
+
+static void
+virtio_gpu_fill_display_info(VirtIOGPU *g,
+                             struct virtio_gpu_resp_display_info *dpy_info)
+{
+    int i;
+
+    for (i = 0; i < g->conf.max_outputs; i++) {
+        if (g->enabled_output_bitmask & (1 << i)) {
+            dpy_info->pmodes[i].enabled = 1;
+            dpy_info->pmodes[i].r.width = g->req_state[i].width;
+            dpy_info->pmodes[i].r.height = g->req_state[i].height;
+        }
+    }
+}
+
+void virtio_gpu_get_display_info(VirtIOGPU *g,
+                                 struct virtio_gpu_ctrl_command *cmd)
+{
+    struct virtio_gpu_resp_display_info display_info;
+
+    trace_virtio_gpu_cmd_get_display_info();
+    memset(&display_info, 0, sizeof(display_info));
+    display_info.hdr.type = VIRTIO_GPU_RESP_OK_DISPLAY_INFO;
+    virtio_gpu_fill_display_info(g, &display_info);
+    virtio_gpu_ctrl_response(g, cmd, &display_info.hdr,
+                             sizeof(display_info));
+}
+
+static pixman_format_code_t get_pixman_format(uint32_t virtio_gpu_format)
+{
+    switch (virtio_gpu_format) {
+#ifdef HOST_WORDS_BIGENDIAN
+    case VIRTIO_GPU_FORMAT_B8G8R8X8_UNORM:
+        return PIXMAN_b8g8r8x8;
+    case VIRTIO_GPU_FORMAT_B8G8R8A8_UNORM:
+        return PIXMAN_b8g8r8a8;
+    case VIRTIO_GPU_FORMAT_X8R8G8B8_UNORM:
+        return PIXMAN_x8r8g8b8;
+    case VIRTIO_GPU_FORMAT_A8R8G8B8_UNORM:
+        return PIXMAN_a8r8g8b8;
+    case VIRTIO_GPU_FORMAT_R8G8B8X8_UNORM:
+        return PIXMAN_r8g8b8x8;
+    case VIRTIO_GPU_FORMAT_R8G8B8A8_UNORM:
+        return PIXMAN_r8g8b8a8;
+    case VIRTIO_GPU_FORMAT_X8B8G8R8_UNORM:
+        return PIXMAN_x8b8g8r8;
+    case VIRTIO_GPU_FORMAT_A8B8G8R8_UNORM:
+        return PIXMAN_a8b8g8r8;
+#else
+    case VIRTIO_GPU_FORMAT_B8G8R8X8_UNORM:
+        return PIXMAN_x8r8g8b8;
+    case VIRTIO_GPU_FORMAT_B8G8R8A8_UNORM:
+        return PIXMAN_a8r8g8b8;
+    case VIRTIO_GPU_FORMAT_X8R8G8B8_UNORM:
+        return PIXMAN_b8g8r8x8;
+    case VIRTIO_GPU_FORMAT_A8R8G8B8_UNORM:
+        return PIXMAN_b8g8r8a8;
+    case VIRTIO_GPU_FORMAT_R8G8B8X8_UNORM:
+        return PIXMAN_x8b8g8r8;
+    case VIRTIO_GPU_FORMAT_R8G8B8A8_UNORM:
+        return PIXMAN_a8b8g8r8;
+    case VIRTIO_GPU_FORMAT_X8B8G8R8_UNORM:
+        return PIXMAN_r8g8b8x8;
+    case VIRTIO_GPU_FORMAT_A8B8G8R8_UNORM:
+        return PIXMAN_r8g8b8a8;
+#endif
+    default:
+        return 0;
+    }
+}
+
+static void virtio_gpu_resource_create_2d(VirtIOGPU *g,
+                                          struct virtio_gpu_ctrl_command *cmd)
+{
+    pixman_format_code_t pformat;
+    struct virtio_gpu_simple_resource *res;
+    struct virtio_gpu_resource_create_2d c2d;
+
+    VIRTIO_GPU_FILL_CMD(c2d);
+    trace_virtio_gpu_cmd_res_create_2d(c2d.resource_id, c2d.format,
+                                       c2d.width, c2d.height);
+
+    if (c2d.resource_id == 0) {
+        qemu_log_mask(LOG_GUEST_ERROR, "%s: resource id 0 is not allowed\n",
+                      __func__);
+        cmd->error = VIRTIO_GPU_RESP_ERR_INVALID_RESOURCE_ID;
+        return;
+    }
+
+    res = virtio_gpu_find_resource(g, c2d.resource_id);
+    if (res) {
+        qemu_log_mask(LOG_GUEST_ERROR, "%s: resource already exists %d\n",
+                      __func__, c2d.resource_id);
+        cmd->error = VIRTIO_GPU_RESP_ERR_INVALID_RESOURCE_ID;
+        return;
+    }
+
+    res = g_new0(struct virtio_gpu_simple_resource, 1);
+
+    res->width = c2d.width;
+    res->height = c2d.height;
+    res->format = c2d.format;
+    res->resource_id = c2d.resource_id;
+
+    pformat = get_pixman_format(c2d.format);
+    if (!pformat) {
+        qemu_log_mask(LOG_GUEST_ERROR,
+                      "%s: host couldn't handle guest format %d\n",
+                      __func__, c2d.format);
+        cmd->error = VIRTIO_GPU_RESP_ERR_INVALID_PARAMETER;
+        return;
+    }
+    res->image = pixman_image_create_bits(pformat,
+                                          c2d.width,
+                                          c2d.height,
+                                          NULL, 0);
+
+    if (!res->image) {
+        qemu_log_mask(LOG_GUEST_ERROR,
+                      "%s: resource creation failed %d %d %d\n",
+                      __func__, c2d.resource_id, c2d.width, c2d.height);
+        g_free(res);
+        cmd->error = VIRTIO_GPU_RESP_ERR_OUT_OF_MEMORY;
+        return;
+    }
+
+    QTAILQ_INSERT_HEAD(&g->reslist, res, next);
+}
+
+static void virtio_gpu_resource_destroy(VirtIOGPU *g,
+                                        struct virtio_gpu_simple_resource *res)
+{
+    pixman_image_unref(res->image);
+    QTAILQ_REMOVE(&g->reslist, res, next);
+    g_free(res);
+}
+
+static void virtio_gpu_resource_unref(VirtIOGPU *g,
+                                      struct virtio_gpu_ctrl_command *cmd)
+{
+    struct virtio_gpu_simple_resource *res;
+    struct virtio_gpu_resource_unref unref;
+
+    VIRTIO_GPU_FILL_CMD(unref);
+    trace_virtio_gpu_cmd_res_unref(unref.resource_id);
+
+    res = virtio_gpu_find_resource(g, unref.resource_id);
+    if (!res) {
+        qemu_log_mask(LOG_GUEST_ERROR, "%s: illegal resource specified %d\n",
+                      __func__, unref.resource_id);
+        cmd->error = VIRTIO_GPU_RESP_ERR_INVALID_RESOURCE_ID;
+        return;
+    }
+    virtio_gpu_resource_destroy(g, res);
+}
+
+static void virtio_gpu_transfer_to_host_2d(VirtIOGPU *g,
+                                           struct virtio_gpu_ctrl_command *cmd)
+{
+    struct virtio_gpu_simple_resource *res;
+    int h;
+    uint32_t src_offset, dst_offset, stride;
+    int bpp;
+    pixman_format_code_t format;
+    struct virtio_gpu_transfer_to_host_2d t2d;
+
+    VIRTIO_GPU_FILL_CMD(t2d);
+    trace_virtio_gpu_cmd_res_xfer_toh_2d(t2d.resource_id);
+
+    res = virtio_gpu_find_resource(g, t2d.resource_id);
+    if (!res || !res->iov) {
+        qemu_log_mask(LOG_GUEST_ERROR, "%s: illegal resource specified %d\n",
+                      __func__, t2d.resource_id);
+        cmd->error = VIRTIO_GPU_RESP_ERR_INVALID_RESOURCE_ID;
+        return;
+    }
+
+    if (t2d.r.x > res->width ||
+        t2d.r.y > res->height ||
+        t2d.r.width > res->width ||
+        t2d.r.height > res->height ||
+        t2d.r.x + t2d.r.width > res->width ||
+        t2d.r.y + t2d.r.height > res->height) {
+        qemu_log_mask(LOG_GUEST_ERROR, "%s: transfer bounds outside resource"
+                      " bounds for resource %d: %d %d %d %d vs %d %d\n",
+                      __func__, t2d.resource_id, t2d.r.x, t2d.r.y,
+                      t2d.r.width, t2d.r.height, res->width, res->height);
+        cmd->error = VIRTIO_GPU_RESP_ERR_INVALID_PARAMETER;
+        return;
+    }
+
+    format = pixman_image_get_format(res->image);
+    bpp = (PIXMAN_FORMAT_BPP(format) + 7) / 8;
+    stride = pixman_image_get_stride(res->image);
+
+    if (t2d.offset || t2d.r.x || t2d.r.y ||
+        t2d.r.width != pixman_image_get_width(res->image)) {
+        void *img_data = pixman_image_get_data(res->image);
+        for (h = 0; h < t2d.r.height; h++) {
+            src_offset = t2d.offset + stride * h;
+            dst_offset = (t2d.r.y + h) * stride + (t2d.r.x * bpp);
+
+            iov_to_buf(res->iov, res->iov_cnt, src_offset,
+                       (uint8_t *)img_data
+                       + dst_offset, t2d.r.width * bpp);
+        }
+    } else {
+        iov_to_buf(res->iov, res->iov_cnt, 0,
+                   pixman_image_get_data(res->image),
+                   pixman_image_get_stride(res->image)
+                   * pixman_image_get_height(res->image));
+    }
+}
+
+static void virtio_gpu_resource_flush(VirtIOGPU *g,
+                                      struct virtio_gpu_ctrl_command *cmd)
+{
+    struct virtio_gpu_simple_resource *res;
+    struct virtio_gpu_resource_flush rf;
+    pixman_region16_t flush_region;
+    int i;
+
+    VIRTIO_GPU_FILL_CMD(rf);
+    trace_virtio_gpu_cmd_res_flush(rf.resource_id,
+                                   rf.r.width, rf.r.height, rf.r.x, rf.r.y);
+
+    res = virtio_gpu_find_resource(g, rf.resource_id);
+    if (!res) {
+        qemu_log_mask(LOG_GUEST_ERROR, "%s: illegal resource specified %d\n",
+                      __func__, rf.resource_id);
+        cmd->error = VIRTIO_GPU_RESP_ERR_INVALID_RESOURCE_ID;
+        return;
+    }
+
+    if (rf.r.x > res->width ||
+        rf.r.y > res->height ||
+        rf.r.width > res->width ||
+        rf.r.height > res->height ||
+        rf.r.x + rf.r.width > res->width ||
+        rf.r.y + rf.r.height > res->height) {
+        qemu_log_mask(LOG_GUEST_ERROR, "%s: flush bounds outside resource"
+                      " bounds for resource %d: %d %d %d %d vs %d %d\n",
+                      __func__, rf.resource_id, rf.r.x, rf.r.y,
+                      rf.r.width, rf.r.height, res->width, res->height);
+        cmd->error = VIRTIO_GPU_RESP_ERR_INVALID_PARAMETER;
+        return;
+    }
+
+    pixman_region_init_rect(&flush_region,
+                            rf.r.x, rf.r.y, rf.r.width, rf.r.height);
+    for (i = 0; i < VIRTIO_GPU_MAX_SCANOUT; i++) {
+        struct virtio_gpu_scanout *scanout;
+        pixman_region16_t region, finalregion;
+        pixman_box16_t *extents;
+
+        if (!(res->scanout_bitmask & (1 << i))) {
+            continue;
+        }
+        scanout = &g->scanout[i];
+
+        pixman_region_init(&finalregion);
+        pixman_region_init_rect(&region, scanout->x, scanout->y,
+                                scanout->width, scanout->height);
+
+        pixman_region_intersect(&finalregion, &flush_region, &region);
+        pixman_region_translate(&finalregion, -scanout->x, -scanout->y);
+        extents = pixman_region_extents(&finalregion);
+        /* work out the area we need to update for each console */
+        dpy_gfx_update(g->scanout[i].con,
+                       extents->x1, extents->y1,
+                       extents->x2 - extents->x1,
+                       extents->y2 - extents->y1);
+
+        pixman_region_fini(&region);
+        pixman_region_fini(&finalregion);
+    }
+    pixman_region_fini(&flush_region);
+}
+
+static void virtio_gpu_set_scanout(VirtIOGPU *g,
+                                   struct virtio_gpu_ctrl_command *cmd)
+{
+    struct virtio_gpu_simple_resource *res;
+    struct virtio_gpu_scanout *scanout;
+    pixman_format_code_t format;
+    uint32_t offset;
+    int bpp;
+    struct virtio_gpu_set_scanout ss;
+
+    VIRTIO_GPU_FILL_CMD(ss);
+    trace_virtio_gpu_cmd_set_scanout(ss.scanout_id, ss.resource_id,
+                                     ss.r.width, ss.r.height, ss.r.x, ss.r.y);
+
+    g->enable = 1;
+    if (ss.resource_id == 0) {
+        scanout = &g->scanout[ss.scanout_id];
+        if (scanout->resource_id) {
+            res = virtio_gpu_find_resource(g, scanout->resource_id);
+            if (res) {
+                res->scanout_bitmask &= ~(1 << ss.scanout_id);
+            }
+        }
+        if (ss.scanout_id == 0 ||
+            ss.scanout_id >= g->conf.max_outputs) {
+            qemu_log_mask(LOG_GUEST_ERROR,
+                          "%s: illegal scanout id specified %d",
+                          __func__, ss.scanout_id);
+            cmd->error = VIRTIO_GPU_RESP_ERR_INVALID_SCANOUT_ID;
+            return;
+        }
+        dpy_gfx_replace_surface(g->scanout[ss.scanout_id].con, NULL);
+        scanout->ds = NULL;
+        scanout->width = 0;
+        scanout->height = 0;
+        return;
+    }
+
+    /* create a surface for this scanout */
+    if (ss.scanout_id >= VIRTIO_GPU_MAX_SCANOUT ||
+        ss.scanout_id >= g->conf.max_outputs) {
+        qemu_log_mask(LOG_GUEST_ERROR, "%s: illegal scanout id specified %d",
+                      __func__, ss.scanout_id);
+        cmd->error = VIRTIO_GPU_RESP_ERR_INVALID_SCANOUT_ID;
+        return;
+    }
+
+    res = virtio_gpu_find_resource(g, ss.resource_id);
+    if (!res) {
+        qemu_log_mask(LOG_GUEST_ERROR, "%s: illegal resource specified %d\n",
+                      __func__, ss.resource_id);
+        cmd->error = VIRTIO_GPU_RESP_ERR_INVALID_RESOURCE_ID;
+        return;
+    }
+
+    if (ss.r.x > res->width ||
+        ss.r.y > res->height ||
+        ss.r.width > res->width ||
+        ss.r.height > res->height ||
+        ss.r.x + ss.r.width > res->width ||
+        ss.r.y + ss.r.height > res->height) {
+        qemu_log_mask(LOG_GUEST_ERROR, "%s: illegal scanout %d bounds for"
+                      " resource %d, (%d,%d)+%d,%d vs %d %d\n",
+                      __func__, ss.scanout_id, ss.resource_id, ss.r.x, ss.r.y,
+                      ss.r.width, ss.r.height, res->width, res->height);
+        cmd->error = VIRTIO_GPU_RESP_ERR_INVALID_PARAMETER;
+        return;
+    }
+
+    scanout = &g->scanout[ss.scanout_id];
+
+    format = pixman_image_get_format(res->image);
+    bpp = (PIXMAN_FORMAT_BPP(format) + 7) / 8;
+    offset = (ss.r.x * bpp) + ss.r.y * pixman_image_get_stride(res->image);
+    if (!scanout->ds || surface_data(scanout->ds)
+        != ((uint8_t *)pixman_image_get_data(res->image) + offset) ||
+        scanout->width != ss.r.width ||
+        scanout->height != ss.r.height) {
+        /* realloc the surface ptr */
+        scanout->ds = qemu_create_displaysurface_from
+            (ss.r.width, ss.r.height, format,
+             pixman_image_get_stride(res->image),
+             (uint8_t *)pixman_image_get_data(res->image) + offset);
+        if (!scanout->ds) {
+            cmd->error = VIRTIO_GPU_RESP_ERR_UNSPEC;
+            return;
+        }
+        dpy_gfx_replace_surface(g->scanout[ss.scanout_id].con, scanout->ds);
+    }
+
+    res->scanout_bitmask |= (1 << ss.scanout_id);
+    scanout->resource_id = ss.resource_id;
+    scanout->x = ss.r.x;
+    scanout->y = ss.r.y;
+    scanout->width = ss.r.width;
+    scanout->height = ss.r.height;
+}
+
+int virtio_gpu_create_mapping_iov(struct virtio_gpu_resource_attach_backing *ab,
+                                  struct virtio_gpu_ctrl_command *cmd,
+                                  struct iovec **iov)
+{
+    struct virtio_gpu_mem_entry *ents;
+    size_t esize, s;
+    int i;
+
+    if (ab->nr_entries > 16384) {
+        qemu_log_mask(LOG_GUEST_ERROR,
+                      "%s: nr_entries is too big (%d > 1024)\n",
+                      __func__, ab->nr_entries);
+        return -1;
+    }
+
+    esize = sizeof(*ents) * ab->nr_entries;
+    ents = g_malloc(esize);
+    s = iov_to_buf(cmd->elem.out_sg, cmd->elem.out_num,
+                   sizeof(*ab), ents, esize);
+    if (s != esize) {
+        qemu_log_mask(LOG_GUEST_ERROR,
+                      "%s: command data size incorrect %zu vs %zu\n",
+                      __func__, s, esize);
+        g_free(ents);
+        return -1;
+    }
+
+    *iov = g_malloc0(sizeof(struct iovec) * ab->nr_entries);
+    for (i = 0; i < ab->nr_entries; i++) {
+        hwaddr len = ents[i].length;
+        (*iov)[i].iov_len = ents[i].length;
+        (*iov)[i].iov_base = cpu_physical_memory_map(ents[i].addr, &len, 1);
+        if (!(*iov)[i].iov_base || len != ents[i].length) {
+            qemu_log_mask(LOG_GUEST_ERROR, "%s: failed to map MMIO memory for"
+                          " resource %d element %d\n",
+                          __func__, ab->resource_id, i);
+            virtio_gpu_cleanup_mapping_iov(*iov, i);
+            g_free(ents);
+            g_free(*iov);
+            *iov = NULL;
+            return -1;
+        }
+    }
+    g_free(ents);
+    return 0;
+}
+
+void virtio_gpu_cleanup_mapping_iov(struct iovec *iov, uint32_t count)
+{
+    int i;
+
+    for (i = 0; i < count; i++) {
+        cpu_physical_memory_unmap(iov[i].iov_base, iov[i].iov_len, 1,
+                                  iov[i].iov_len);
+    }
+}
+
+static void virtio_gpu_cleanup_mapping(struct virtio_gpu_simple_resource *res)
+{
+    virtio_gpu_cleanup_mapping_iov(res->iov, res->iov_cnt);
+    g_free(res->iov);
+    res->iov = NULL;
+    res->iov_cnt = 0;
+}
+
+static void
+virtio_gpu_resource_attach_backing(VirtIOGPU *g,
+                                   struct virtio_gpu_ctrl_command *cmd)
+{
+    struct virtio_gpu_simple_resource *res;
+    struct virtio_gpu_resource_attach_backing ab;
+    int ret;
+
+    VIRTIO_GPU_FILL_CMD(ab);
+    trace_virtio_gpu_cmd_res_back_attach(ab.resource_id);
+
+    res = virtio_gpu_find_resource(g, ab.resource_id);
+    if (!res) {
+        qemu_log_mask(LOG_GUEST_ERROR, "%s: illegal resource specified %d\n",
+                      __func__, ab.resource_id);
+        cmd->error = VIRTIO_GPU_RESP_ERR_INVALID_RESOURCE_ID;
+        return;
+    }
+
+    ret = virtio_gpu_create_mapping_iov(&ab, cmd, &res->iov);
+    if (ret != 0) {
+        cmd->error = VIRTIO_GPU_RESP_ERR_UNSPEC;
+        return;
+    }
+
+    res->iov_cnt = ab.nr_entries;
+}
+
+static void
+virtio_gpu_resource_detach_backing(VirtIOGPU *g,
+                                   struct virtio_gpu_ctrl_command *cmd)
+{
+    struct virtio_gpu_simple_resource *res;
+    struct virtio_gpu_resource_detach_backing detach;
+
+    VIRTIO_GPU_FILL_CMD(detach);
+    trace_virtio_gpu_cmd_res_back_detach(detach.resource_id);
+
+    res = virtio_gpu_find_resource(g, detach.resource_id);
+    if (!res || !res->iov) {
+        qemu_log_mask(LOG_GUEST_ERROR, "%s: illegal resource specified %d\n",
+                      __func__, detach.resource_id);
+        cmd->error = VIRTIO_GPU_RESP_ERR_INVALID_RESOURCE_ID;
+        return;
+    }
+    virtio_gpu_cleanup_mapping(res);
+}
+
+static void virtio_gpu_simple_process_cmd(VirtIOGPU *g,
+                                          struct virtio_gpu_ctrl_command *cmd)
+{
+    VIRTIO_GPU_FILL_CMD(cmd->cmd_hdr);
+
+    switch (cmd->cmd_hdr.type) {
+    case VIRTIO_GPU_CMD_GET_DISPLAY_INFO:
+        virtio_gpu_get_display_info(g, cmd);
+        break;
+    case VIRTIO_GPU_CMD_RESOURCE_CREATE_2D:
+        virtio_gpu_resource_create_2d(g, cmd);
+        break;
+    case VIRTIO_GPU_CMD_RESOURCE_UNREF:
+        virtio_gpu_resource_unref(g, cmd);
+        break;
+    case VIRTIO_GPU_CMD_RESOURCE_FLUSH:
+        virtio_gpu_resource_flush(g, cmd);
+        break;
+    case VIRTIO_GPU_CMD_TRANSFER_TO_HOST_2D:
+        virtio_gpu_transfer_to_host_2d(g, cmd);
+        break;
+    case VIRTIO_GPU_CMD_SET_SCANOUT:
+        virtio_gpu_set_scanout(g, cmd);
+        break;
+    case VIRTIO_GPU_CMD_RESOURCE_ATTACH_BACKING:
+        virtio_gpu_resource_attach_backing(g, cmd);
+        break;
+    case VIRTIO_GPU_CMD_RESOURCE_DETACH_BACKING:
+        virtio_gpu_resource_detach_backing(g, cmd);
+        break;
+    default:
+        cmd->error = VIRTIO_GPU_RESP_ERR_UNSPEC;
+        break;
+    }
+    if (!cmd->finished) {
+        virtio_gpu_ctrl_response_nodata(g, cmd, cmd->error ? cmd->error :
+                                        VIRTIO_GPU_RESP_OK_NODATA);
+    }
+}
+
+static void virtio_gpu_handle_ctrl_cb(VirtIODevice *vdev, VirtQueue *vq)
+{
+    VirtIOGPU *g = VIRTIO_GPU(vdev);
+    qemu_bh_schedule(g->ctrl_bh);
+}
+
+static void virtio_gpu_handle_cursor_cb(VirtIODevice *vdev, VirtQueue *vq)
+{
+    VirtIOGPU *g = VIRTIO_GPU(vdev);
+    qemu_bh_schedule(g->cursor_bh);
+}
+
+static void virtio_gpu_handle_ctrl(VirtIODevice *vdev, VirtQueue *vq)
+{
+    VirtIOGPU *g = VIRTIO_GPU(vdev);
+    struct virtio_gpu_ctrl_command *cmd;
+
+    if (!virtio_queue_ready(vq)) {
+        return;
+    }
+
+    cmd = g_new(struct virtio_gpu_ctrl_command, 1);
+    while (virtqueue_pop(vq, &cmd->elem)) {
+        cmd->vq = vq;
+        cmd->error = 0;
+        cmd->finished = false;
+        g->stats.requests++;
+
+        virtio_gpu_simple_process_cmd(g, cmd);
+        if (!cmd->finished) {
+            QTAILQ_INSERT_TAIL(&g->fenceq, cmd, next);
+            g->stats.inflight++;
+            if (g->stats.max_inflight < g->stats.inflight) {
+                g->stats.max_inflight = g->stats.inflight;
+            }
+            fprintf(stderr, "inflight: %3d (+)\r", g->stats.inflight);
+            cmd = g_new(struct virtio_gpu_ctrl_command, 1);
+        }
+    }
+    g_free(cmd);
+}
+
+static void virtio_gpu_ctrl_bh(void *opaque)
+{
+    VirtIOGPU *g = opaque;
+    virtio_gpu_handle_ctrl(&g->parent_obj, g->ctrl_vq);
+}
+
+static void virtio_gpu_handle_cursor(VirtIODevice *vdev, VirtQueue *vq)
+{
+    VirtIOGPU *g = VIRTIO_GPU(vdev);
+    VirtQueueElement elem;
+    size_t s;
+    struct virtio_gpu_update_cursor cursor_info;
+
+    if (!virtio_queue_ready(vq)) {
+        return;
+    }
+    while (virtqueue_pop(vq, &elem)) {
+        s = iov_to_buf(elem.out_sg, elem.out_num, 0,
+                       &cursor_info, sizeof(cursor_info));
+        if (s != sizeof(cursor_info)) {
+            qemu_log_mask(LOG_GUEST_ERROR,
+                          "%s: cursor size incorrect %zu vs %zu\n",
+                          __func__, s, sizeof(cursor_info));
+        } else {
+            update_cursor(g, &cursor_info);
+        }
+        virtqueue_push(vq, &elem, 0);
+        virtio_notify(vdev, vq);
+    }
+}
+
+static void virtio_gpu_cursor_bh(void *opaque)
+{
+    VirtIOGPU *g = opaque;
+    virtio_gpu_handle_cursor(&g->parent_obj, g->cursor_vq);
+}
+
+static void virtio_gpu_invalidate_display(void *opaque)
+{
+}
+
+static void virtio_gpu_update_display(void *opaque)
+{
+}
+
+static void virtio_gpu_text_update(void *opaque, console_ch_t *chardata)
+{
+}
+
+static int virtio_gpu_ui_info(void *opaque, uint32_t idx, QemuUIInfo *info)
+{
+    VirtIOGPU *g = opaque;
+
+    if (idx > g->conf.max_outputs) {
+        return -1;
+    }
+
+    g->req_state[idx].x = info->xoff;
+    g->req_state[idx].y = info->yoff;
+    g->req_state[idx].width = info->width;
+    g->req_state[idx].height = info->height;
+
+    if (info->width && info->height) {
+        g->enabled_output_bitmask |= (1 << idx);
+    } else {
+        g->enabled_output_bitmask &= ~(1 << idx);
+    }
+
+    /* send event to guest */
+    virtio_gpu_notify_event(g, VIRTIO_GPU_EVENT_DISPLAY);
+    return 0;
+}
+
+const GraphicHwOps virtio_gpu_ops = {
+    .invalidate = virtio_gpu_invalidate_display,
+    .gfx_update = virtio_gpu_update_display,
+    .text_update = virtio_gpu_text_update,
+    .ui_info = virtio_gpu_ui_info,
+};
+
+static void virtio_gpu_device_realize(DeviceState *qdev, Error **errp)
+{
+    VirtIODevice *vdev = VIRTIO_DEVICE(qdev);
+    VirtIOGPU *g = VIRTIO_GPU(qdev);
+    int i;
+
+    g->config_size = sizeof(struct virtio_gpu_config);
+    g->virtio_config.num_scanouts = g->conf.max_outputs;
+    virtio_init(VIRTIO_DEVICE(g), "virtio-gpu", VIRTIO_ID_GPU,
+                g->config_size);
+
+    g->req_state[0].width = 1024;
+    g->req_state[0].height = 768;
+
+    g->ctrl_vq   = virtio_add_queue(vdev, 64, virtio_gpu_handle_ctrl_cb);
+    g->cursor_vq = virtio_add_queue(vdev, 16, virtio_gpu_handle_cursor_cb);
+
+    g->ctrl_bh = qemu_bh_new(virtio_gpu_ctrl_bh, g);
+    g->cursor_bh = qemu_bh_new(virtio_gpu_cursor_bh, g);
+    QTAILQ_INIT(&g->reslist);
+    QTAILQ_INIT(&g->fenceq);
+
+    g->enabled_output_bitmask = 1;
+    g->qdev = qdev;
+
+    for (i = 0; i < g->conf.max_outputs; i++) {
+        g->scanout[i].con =
+            graphic_console_init(DEVICE(g), i, &virtio_gpu_ops, g);
+        if (i > 0) {
+            dpy_gfx_replace_surface(g->scanout[i].con, NULL);
+        }
+    }
+}
+
+static void virtio_gpu_instance_init(Object *obj)
+{
+}
+
+static void virtio_gpu_reset(VirtIODevice *vdev)
+{
+    VirtIOGPU *g = VIRTIO_GPU(vdev);
+    struct virtio_gpu_simple_resource *res, *tmp;
+    int i;
+
+    g->enable = 0;
+
+    QTAILQ_FOREACH_SAFE(res, &g->reslist, next, tmp) {
+        virtio_gpu_resource_destroy(g, res);
+    }
+    for (i = 0; i < g->conf.max_outputs; i++) {
+#if 0
+        g->req_state[i].x = 0;
+        g->req_state[i].y = 0;
+        if (i == 0) {
+            g->req_state[0].width = 1024;
+            g->req_state[0].height = 768;
+        } else {
+            g->req_state[i].width = 0;
+            g->req_state[i].height = 0;
+        }
+#endif
+        g->scanout[i].resource_id = 0;
+        g->scanout[i].width = 0;
+        g->scanout[i].height = 0;
+        g->scanout[i].x = 0;
+        g->scanout[i].y = 0;
+        g->scanout[i].ds = NULL;
+    }
+    g->enabled_output_bitmask = 1;
+}
+
+static Property virtio_gpu_properties[] = {
+    DEFINE_VIRTIO_GPU_PROPERTIES(VirtIOGPU, conf),
+    DEFINE_PROP_END_OF_LIST(),
+};
+
+static void virtio_gpu_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+    VirtioDeviceClass *vdc = VIRTIO_DEVICE_CLASS(klass);
+
+    vdc->realize = virtio_gpu_device_realize;
+    vdc->get_config = virtio_gpu_get_config;
+    vdc->set_config = virtio_gpu_set_config;
+    vdc->get_features = virtio_gpu_get_features;
+
+    vdc->reset = virtio_gpu_reset;
+
+    dc->props = virtio_gpu_properties;
+}
+
+static const TypeInfo virtio_gpu_info = {
+    .name = TYPE_VIRTIO_GPU,
+    .parent = TYPE_VIRTIO_DEVICE,
+    .instance_size = sizeof(VirtIOGPU),
+    .instance_init = virtio_gpu_instance_init,
+    .class_init = virtio_gpu_class_init,
+};
+
+static void virtio_register_types(void)
+{
+    type_register_static(&virtio_gpu_info);
+}
+
+type_init(virtio_register_types)
+
+QEMU_BUILD_BUG_ON(sizeof(struct virtio_gpu_ctrl_hdr)                != 24);
+QEMU_BUILD_BUG_ON(sizeof(struct virtio_gpu_update_cursor)           != 56);
+QEMU_BUILD_BUG_ON(sizeof(struct virtio_gpu_resource_unref)          != 32);
+QEMU_BUILD_BUG_ON(sizeof(struct virtio_gpu_resource_create_2d)      != 40);
+QEMU_BUILD_BUG_ON(sizeof(struct virtio_gpu_set_scanout)             != 48);
+QEMU_BUILD_BUG_ON(sizeof(struct virtio_gpu_resource_flush)          != 48);
+QEMU_BUILD_BUG_ON(sizeof(struct virtio_gpu_transfer_to_host_2d)     != 56);
+QEMU_BUILD_BUG_ON(sizeof(struct virtio_gpu_mem_entry)               != 16);
+QEMU_BUILD_BUG_ON(sizeof(struct virtio_gpu_resource_attach_backing) != 32);
+QEMU_BUILD_BUG_ON(sizeof(struct virtio_gpu_resource_detach_backing) != 32);
+QEMU_BUILD_BUG_ON(sizeof(struct virtio_gpu_resp_display_info)       != 408);
diff --git a/hw/dma/rc4030.c b/hw/dma/rc4030.c
index af26632..3efa6de 100644
--- a/hw/dma/rc4030.c
+++ b/hw/dma/rc4030.c
@@ -1,7 +1,7 @@
 /*
  * QEMU JAZZ RC4030 chipset
  *
- * Copyright (c) 2007-2009 Herve Poussineau
+ * Copyright (c) 2007-2013 Hervé Poussineau
  *
  * Permission is hereby granted, free of charge, to any person obtaining a copy
  * of this software and associated documentation files (the "Software"), to deal
@@ -24,29 +24,16 @@
 
 #include "hw/hw.h"
 #include "hw/mips/mips.h"
+#include "hw/sysbus.h"
 #include "qemu/timer.h"
-
-/********************************************************/
-/* debug rc4030 */
-
-//#define DEBUG_RC4030
-//#define DEBUG_RC4030_DMA
-
-#ifdef DEBUG_RC4030
-#define DPRINTF(fmt, ...) \
-do { printf("rc4030: " fmt , ## __VA_ARGS__); } while (0)
-static const char* irq_names[] = { "parallel", "floppy", "sound", "video",
-            "network", "scsi", "keyboard", "mouse", "serial0", "serial1" };
-#else
-#define DPRINTF(fmt, ...)
-#endif
-
-#define RC4030_ERROR(fmt, ...) \
-do { fprintf(stderr, "rc4030 ERROR: %s: " fmt, __func__ , ## __VA_ARGS__); } while (0)
+#include "exec/address-spaces.h"
+#include "trace.h"
 
 /********************************************************/
 /* rc4030 emulation                                     */
 
+#define MAX_TL_ENTRIES 512
+
 typedef struct dma_pagetable_entry {
     int32_t frame;
     int32_t owner;
@@ -63,8 +50,14 @@
 #define DMA_FLAG_MEM_INTR   0x0200
 #define DMA_FLAG_ADDR_INTR  0x0400
 
+#define TYPE_RC4030 "rc4030"
+#define RC4030(obj) \
+    OBJECT_CHECK(rc4030State, (obj), TYPE_RC4030)
+
 typedef struct rc4030State
 {
+    SysBusDevice parent;
+
     uint32_t config; /* 0x0000: RC4030 config register */
     uint32_t revision; /* 0x0008: RC4030 Revision register */
     uint32_t invalid_address_register; /* 0x0010: Invalid Address register */
@@ -83,7 +76,7 @@
     uint32_t cache_bmask; /* 0x0058: I/O Cache Byte Mask */
 
     uint32_t nmi_interrupt; /* 0x0200: interrupt source */
-    uint32_t offset210;
+    uint32_t memory_refresh_rate; /* 0x0210: memory refresh rate */
     uint32_t nvram_protect; /* 0x0220: NV ram protect register */
     uint32_t rem_speed[16];
     uint32_t imr_jazz; /* Local bus int enable mask */
@@ -96,6 +89,16 @@
     qemu_irq timer_irq;
     qemu_irq jazz_bus_irq;
 
+    /* biggest translation table */
+    MemoryRegion dma_tt;
+    /* translation table memory region alias, added to system RAM */
+    MemoryRegion dma_tt_alias;
+    /* whole DMA memory region, root of DMA address space */
+    MemoryRegion dma_mr;
+    /* translation table entry aliases, added to DMA memory region */
+    MemoryRegion dma_mrs[MAX_TL_ENTRIES];
+    AddressSpace dma_as;
+
     MemoryRegion iomem_chipset;
     MemoryRegion iomem_jazzio;
 } rc4030State;
@@ -112,7 +115,7 @@
 }
 
 /* called for accesses to rc4030 */
-static uint32_t rc4030_readl(void *opaque, hwaddr addr)
+static uint64_t rc4030_read(void *opaque, hwaddr addr, unsigned int size)
 {
     rc4030State *s = opaque;
     uint32_t val;
@@ -220,9 +223,9 @@
     case 0x0208:
         val = 0;
         break;
-    /* Offset 0x0210 */
+    /* Memory refresh rate */
     case 0x0210:
-        val = s->offset210;
+        val = s->memory_refresh_rate;
         break;
     /* NV ram protect register */
     case 0x0220:
@@ -238,39 +241,117 @@
         val = 7; /* FIXME: should be read from EISA controller */
         break;
     default:
-        RC4030_ERROR("invalid read [" TARGET_FMT_plx "]\n", addr);
+        qemu_log_mask(LOG_GUEST_ERROR,
+                      "rc4030: invalid read at 0x%x", (int)addr);
         val = 0;
         break;
     }
 
     if ((addr & ~3) != 0x230) {
-        DPRINTF("read 0x%02x at " TARGET_FMT_plx "\n", val, addr);
+        trace_rc4030_read(addr, val);
     }
 
     return val;
 }
 
-static uint32_t rc4030_readw(void *opaque, hwaddr addr)
+static void rc4030_dma_as_update_one(rc4030State *s, int index, uint32_t frame)
 {
-    uint32_t v = rc4030_readl(opaque, addr & ~0x3);
-    if (addr & 0x2)
-        return v >> 16;
-    else
-        return v & 0xffff;
+    if (index < MAX_TL_ENTRIES) {
+        memory_region_set_enabled(&s->dma_mrs[index], false);
+    }
+
+    if (!frame) {
+        return;
+    }
+
+    if (index >= MAX_TL_ENTRIES) {
+        qemu_log_mask(LOG_UNIMP,
+                      "rc4030: trying to use too high "
+                      "translation table entry %d (max allowed=%d)",
+                      index, MAX_TL_ENTRIES);
+        return;
+    }
+    memory_region_set_alias_offset(&s->dma_mrs[index], frame);
+    memory_region_set_enabled(&s->dma_mrs[index], true);
 }
 
-static uint32_t rc4030_readb(void *opaque, hwaddr addr)
-{
-    uint32_t v = rc4030_readl(opaque, addr & ~0x3);
-    return (v >> (8 * (addr & 0x3))) & 0xff;
-}
-
-static void rc4030_writel(void *opaque, hwaddr addr, uint32_t val)
+static void rc4030_dma_tt_write(void *opaque, hwaddr addr, uint64_t data,
+                                unsigned int size)
 {
     rc4030State *s = opaque;
+
+    /* write memory */
+    memcpy(memory_region_get_ram_ptr(&s->dma_tt) + addr, &data, size);
+
+    /* update dma address space (only if frame field has been written) */
+    if (addr % sizeof(dma_pagetable_entry) == 0) {
+        int index = addr / sizeof(dma_pagetable_entry);
+        memory_region_transaction_begin();
+        rc4030_dma_as_update_one(s, index, (uint32_t)data);
+        memory_region_transaction_commit();
+    }
+}
+
+static const MemoryRegionOps rc4030_dma_tt_ops = {
+    .write = rc4030_dma_tt_write,
+    .impl.min_access_size = 4,
+    .impl.max_access_size = 4,
+};
+
+static void rc4030_dma_tt_update(rc4030State *s, uint32_t new_tl_base,
+                                 uint32_t new_tl_limit)
+{
+    int entries, i;
+    dma_pagetable_entry *dma_tl_contents;
+
+    if (s->dma_tl_limit) {
+        /* write old dma tl table to physical memory */
+        memory_region_del_subregion(get_system_memory(), &s->dma_tt_alias);
+        cpu_physical_memory_write(s->dma_tl_limit & 0x7fffffff,
+                                  memory_region_get_ram_ptr(&s->dma_tt),
+                                  memory_region_size(&s->dma_tt_alias));
+    }
+    object_unparent(OBJECT(&s->dma_tt_alias));
+
+    s->dma_tl_base = new_tl_base;
+    s->dma_tl_limit = new_tl_limit;
+    new_tl_base &= 0x7fffffff;
+
+    if (s->dma_tl_limit) {
+        uint64_t dma_tt_size;
+        if (s->dma_tl_limit <= memory_region_size(&s->dma_tt)) {
+            dma_tt_size = s->dma_tl_limit;
+        } else {
+            dma_tt_size = memory_region_size(&s->dma_tt);
+        }
+        memory_region_init_alias(&s->dma_tt_alias, OBJECT(s),
+                                 "dma-table-alias",
+                                 &s->dma_tt, 0, dma_tt_size);
+        dma_tl_contents = memory_region_get_ram_ptr(&s->dma_tt);
+        cpu_physical_memory_read(new_tl_base, dma_tl_contents, dma_tt_size);
+
+        memory_region_transaction_begin();
+        entries = dma_tt_size / sizeof(dma_pagetable_entry);
+        for (i = 0; i < entries; i++) {
+            rc4030_dma_as_update_one(s, i, dma_tl_contents[i].frame);
+        }
+        memory_region_add_subregion(get_system_memory(), new_tl_base,
+                                    &s->dma_tt_alias);
+        memory_region_transaction_commit();
+    } else {
+        memory_region_init(&s->dma_tt_alias, OBJECT(s),
+                           "dma-table-alias", 0);
+    }
+}
+
+static void rc4030_write(void *opaque, hwaddr addr, uint64_t data,
+                         unsigned int size)
+{
+    rc4030State *s = opaque;
+    uint32_t val = data;
     addr &= 0x3fff;
 
-    DPRINTF("write 0x%02x at " TARGET_FMT_plx "\n", val, addr);
+    trace_rc4030_write(addr, val);
 
     switch (addr & ~0x3) {
     /* Global config register */
@@ -279,11 +360,11 @@
         break;
     /* DMA transl. table base */
     case 0x0018:
-        s->dma_tl_base = val;
+        rc4030_dma_tt_update(s, val, s->dma_tl_limit);
         break;
     /* DMA transl. table limit */
     case 0x0020:
-        s->dma_tl_limit = val;
+        rc4030_dma_tt_update(s, s->dma_tl_base, val);
         break;
     /* DMA transl. table invalidated */
     case 0x0028:
@@ -371,9 +452,9 @@
             s->dma_regs[entry][idx] = val;
         }
         break;
-    /* Offset 0x0210 */
+    /* Memory refresh rate */
     case 0x0210:
-        s->offset210 = val;
+        s->memory_refresh_rate = val;
         break;
     /* Interval timer reload */
     case 0x0228:
@@ -385,48 +466,18 @@
     case 0x0238:
         break;
     default:
-        RC4030_ERROR("invalid write of 0x%02x at [" TARGET_FMT_plx "]\n", val, addr);
+        qemu_log_mask(LOG_GUEST_ERROR,
+                      "rc4030: invalid write of 0x%02x at 0x%x",
+                      val, (int)addr);
         break;
     }
 }
 
-static void rc4030_writew(void *opaque, hwaddr addr, uint32_t val)
-{
-    uint32_t old_val = rc4030_readl(opaque, addr & ~0x3);
-
-    if (addr & 0x2)
-        val = (val << 16) | (old_val & 0x0000ffff);
-    else
-        val = val | (old_val & 0xffff0000);
-    rc4030_writel(opaque, addr & ~0x3, val);
-}
-
-static void rc4030_writeb(void *opaque, hwaddr addr, uint32_t val)
-{
-    uint32_t old_val = rc4030_readl(opaque, addr & ~0x3);
-
-    switch (addr & 3) {
-    case 0:
-        val = val | (old_val & 0xffffff00);
-        break;
-    case 1:
-        val = (val << 8) | (old_val & 0xffff00ff);
-        break;
-    case 2:
-        val = (val << 16) | (old_val & 0xff00ffff);
-        break;
-    case 3:
-        val = (val << 24) | (old_val & 0x00ffffff);
-        break;
-    }
-    rc4030_writel(opaque, addr & ~0x3, val);
-}
-
 static const MemoryRegionOps rc4030_ops = {
-    .old_mmio = {
-        .read = { rc4030_readb, rc4030_readw, rc4030_readl, },
-        .write = { rc4030_writeb, rc4030_writew, rc4030_writel, },
-    },
+    .read = rc4030_read,
+    .write = rc4030_write,
+    .impl.min_access_size = 4,
+    .impl.max_access_size = 4,
     .endianness = DEVICE_NATIVE_ENDIAN,
 };
 
@@ -436,22 +487,6 @@
 
     pending = s->isr_jazz & s->imr_jazz;
 
-#ifdef DEBUG_RC4030
-    if (s->isr_jazz != 0) {
-        uint32_t irq = 0;
-        DPRINTF("pending irqs:");
-        for (irq = 0; irq < ARRAY_SIZE(irq_names); irq++) {
-            if (s->isr_jazz & (1 << irq)) {
-                printf(" %s", irq_names[irq]);
-                if (!(s->imr_jazz & (1 << irq))) {
-                    printf("(ignored)");
-                }
-            }
-        }
-        printf("\n");
-    }
-#endif
-
     if (pending != 0)
         qemu_irq_raise(s->jazz_bus_irq);
     else
@@ -479,7 +514,7 @@
     qemu_irq_raise(s->timer_irq);
 }
 
-static uint32_t jazzio_readw(void *opaque, hwaddr addr)
+static uint64_t jazzio_read(void *opaque, hwaddr addr, unsigned int size)
 {
     rc4030State *s = opaque;
     uint32_t val;
@@ -494,7 +529,6 @@
         irq = 0;
         while (pending) {
             if (pending & 1) {
-                DPRINTF("returning irq %s\n", irq_names[irq]);
                 val = (irq + 1) << 2;
                 break;
             }
@@ -508,36 +542,25 @@
         val = s->imr_jazz;
         break;
     default:
-        RC4030_ERROR("(jazz io controller) invalid read [" TARGET_FMT_plx "]\n", addr);
+        qemu_log_mask(LOG_GUEST_ERROR,
+                      "rc4030/jazzio: invalid read at 0x%x", (int)addr);
         val = 0;
+        break;
     }
 
-    DPRINTF("(jazz io controller) read 0x%04x at " TARGET_FMT_plx "\n", val, addr);
+    trace_jazzio_read(addr, val);
 
     return val;
 }
 
-static uint32_t jazzio_readb(void *opaque, hwaddr addr)
-{
-    uint32_t v;
-    v = jazzio_readw(opaque, addr & ~0x1);
-    return (v >> (8 * (addr & 0x1))) & 0xff;
-}
-
-static uint32_t jazzio_readl(void *opaque, hwaddr addr)
-{
-    uint32_t v;
-    v = jazzio_readw(opaque, addr);
-    v |= jazzio_readw(opaque, addr + 2) << 16;
-    return v;
-}
-
-static void jazzio_writew(void *opaque, hwaddr addr, uint32_t val)
+static void jazzio_write(void *opaque, hwaddr addr, uint64_t data,
+                         unsigned int size)
 {
     rc4030State *s = opaque;
+    uint32_t val = data;
     addr &= 0xfff;
 
-    DPRINTF("(jazz io controller) write 0x%04x at " TARGET_FMT_plx "\n", val, addr);
+    trace_jazzio_write(addr, val);
 
     switch (addr) {
     /* Local bus int enable mask */
@@ -546,43 +569,24 @@
         update_jazz_irq(s);
         break;
     default:
-        RC4030_ERROR("(jazz io controller) invalid write of 0x%04x at [" TARGET_FMT_plx "]\n", val, addr);
+        qemu_log_mask(LOG_GUEST_ERROR,
+                      "rc4030/jazzio: invalid write of 0x%02x at 0x%x",
+                      val, (int)addr);
         break;
     }
 }
 
-static void jazzio_writeb(void *opaque, hwaddr addr, uint32_t val)
-{
-    uint32_t old_val = jazzio_readw(opaque, addr & ~0x1);
-
-    switch (addr & 1) {
-    case 0:
-        val = val | (old_val & 0xff00);
-        break;
-    case 1:
-        val = (val << 8) | (old_val & 0x00ff);
-        break;
-    }
-    jazzio_writew(opaque, addr & ~0x1, val);
-}
-
-static void jazzio_writel(void *opaque, hwaddr addr, uint32_t val)
-{
-    jazzio_writew(opaque, addr, val & 0xffff);
-    jazzio_writew(opaque, addr + 2, (val >> 16) & 0xffff);
-}
-
 static const MemoryRegionOps jazzio_ops = {
-    .old_mmio = {
-        .read = { jazzio_readb, jazzio_readw, jazzio_readl, },
-        .write = { jazzio_writeb, jazzio_writew, jazzio_writel, },
-    },
+    .read = jazzio_read,
+    .write = jazzio_write,
+    .impl.min_access_size = 2,
+    .impl.max_access_size = 2,
     .endianness = DEVICE_NATIVE_ENDIAN,
 };
 
-static void rc4030_reset(void *opaque)
+static void rc4030_reset(DeviceState *dev)
 {
-    rc4030State *s = opaque;
+    rc4030State *s = RC4030(dev);
     int i;
 
     s->config = 0x410; /* some boards seem to accept 0x104 too */
@@ -590,14 +594,14 @@
     s->invalid_address_register = 0;
 
     memset(s->dma_regs, 0, sizeof(s->dma_regs));
-    s->dma_tl_base = s->dma_tl_limit = 0;
+    rc4030_dma_tt_update(s, 0, 0);
 
     s->remote_failed_address = s->memory_failed_address = 0;
     s->cache_maint = 0;
     s->cache_ptag = s->cache_ltag = 0;
     s->cache_bmask = 0;
 
-    s->offset210 = 0x18186;
+    s->memory_refresh_rate = 0x18186;
     s->nvram_protect = 7;
     for (i = 0; i < 15; i++)
         s->rem_speed[i] = 7;
@@ -631,7 +635,7 @@
     s->cache_ptag = qemu_get_be32(f);
     s->cache_ltag = qemu_get_be32(f);
     s->cache_bmask = qemu_get_be32(f);
-    s->offset210 = qemu_get_be32(f);
+    s->memory_refresh_rate = qemu_get_be32(f);
     s->nvram_protect = qemu_get_be32(f);
     for (i = 0; i < 15; i++)
         s->rem_speed[i] = qemu_get_be32(f);
@@ -663,7 +667,7 @@
     qemu_put_be32(f, s->cache_ptag);
     qemu_put_be32(f, s->cache_ltag);
     qemu_put_be32(f, s->cache_bmask);
-    qemu_put_be32(f, s->offset210);
+    qemu_put_be32(f, s->memory_refresh_rate);
     qemu_put_be32(f, s->nvram_protect);
     for (i = 0; i < 15; i++)
         qemu_put_be32(f, s->rem_speed[i]);
@@ -672,44 +676,6 @@
     qemu_put_be32(f, s->itr);
 }
 
-void rc4030_dma_memory_rw(void *opaque, hwaddr addr, uint8_t *buf, int len, int is_write)
-{
-    rc4030State *s = opaque;
-    hwaddr entry_addr;
-    hwaddr phys_addr;
-    dma_pagetable_entry entry;
-    int index;
-    int ncpy, i;
-
-    i = 0;
-    for (;;) {
-        if (i == len) {
-            break;
-        }
-
-        ncpy = DMA_PAGESIZE - (addr & (DMA_PAGESIZE - 1));
-        if (ncpy > len - i)
-            ncpy = len - i;
-
-        /* Get DMA translation table entry */
-        index = addr / DMA_PAGESIZE;
-        if (index >= s->dma_tl_limit / sizeof(dma_pagetable_entry)) {
-            break;
-        }
-        entry_addr = s->dma_tl_base + index * sizeof(dma_pagetable_entry);
-        /* XXX: not sure. should we really use only lowest bits? */
-        entry_addr &= 0x7fffffff;
-        cpu_physical_memory_read(entry_addr, &entry, sizeof(entry));
-
-        /* Read/write data at right place */
-        phys_addr = entry.frame + (addr & (DMA_PAGESIZE - 1));
-        cpu_physical_memory_rw(phys_addr, &buf[i], ncpy, is_write);
-
-        i += ncpy;
-        addr += ncpy;
-    }
-}
-
 static void rc4030_do_dma(void *opaque, int n, uint8_t *buf, int len, int is_write)
 {
     rc4030State *s = opaque;
@@ -733,32 +699,11 @@
     dma_addr = s->dma_regs[n][DMA_REG_ADDRESS];
 
     /* Read/write data at right place */
-    rc4030_dma_memory_rw(opaque, dma_addr, buf, len, is_write);
+    address_space_rw(&s->dma_as, dma_addr, MEMTXATTRS_UNSPECIFIED,
+                     buf, len, is_write);
 
     s->dma_regs[n][DMA_REG_ENABLE] |= DMA_FLAG_TC_INTR;
     s->dma_regs[n][DMA_REG_COUNT] -= len;
-
-#ifdef DEBUG_RC4030_DMA
-    {
-        int i, j;
-        printf("rc4030 dma: Copying %d bytes %s host %p\n",
-            len, is_write ? "from" : "to", buf);
-        for (i = 0; i < len; i += 16) {
-            int n = 16;
-            if (n > len - i) {
-                n = len - i;
-            }
-            for (j = 0; j < n; j++)
-                printf("%02x ", buf[i + j]);
-            while (j++ < 16)
-                printf("   ");
-            printf("| ");
-            for (j = 0; j < n; j++)
-                printf("%c", isprint(buf[i + j]) ? buf[i + j] : '.');
-            printf("\n");
-        }
-    }
-#endif
 }
 
 struct rc4030DMAState {
@@ -795,31 +740,102 @@
     return s;
 }
 
-void *rc4030_init(qemu_irq timer, qemu_irq jazz_bus,
-                  qemu_irq **irqs, rc4030_dma **dmas,
-                  MemoryRegion *sysmem)
+static void rc4030_initfn(Object *obj)
 {
-    rc4030State *s;
+    DeviceState *dev = DEVICE(obj);
+    rc4030State *s = RC4030(obj);
+    SysBusDevice *sysbus = SYS_BUS_DEVICE(obj);
 
-    s = g_malloc0(sizeof(rc4030State));
+    qdev_init_gpio_in(dev, rc4030_irq_jazz_request, 16);
 
-    *irqs = qemu_allocate_irqs(rc4030_irq_jazz_request, s, 16);
-    *dmas = rc4030_allocate_dmas(s, 4);
+    sysbus_init_irq(sysbus, &s->timer_irq);
+    sysbus_init_irq(sysbus, &s->jazz_bus_irq);
 
-    s->periodic_timer = timer_new_ns(QEMU_CLOCK_VIRTUAL, rc4030_periodic_timer, s);
-    s->timer_irq = timer;
-    s->jazz_bus_irq = jazz_bus;
-
-    qemu_register_reset(rc4030_reset, s);
     register_savevm(NULL, "rc4030", 0, 2, rc4030_save, rc4030_load, s);
-    rc4030_reset(s);
+
+    sysbus_init_mmio(sysbus, &s->iomem_chipset);
+    sysbus_init_mmio(sysbus, &s->iomem_jazzio);
+}
+
+static void rc4030_realize(DeviceState *dev, Error **errp)
+{
+    rc4030State *s = RC4030(dev);
+    Object *o = OBJECT(dev);
+    int i;
+
+    s->periodic_timer = timer_new_ns(QEMU_CLOCK_VIRTUAL,
+                                     rc4030_periodic_timer, s);
 
     memory_region_init_io(&s->iomem_chipset, NULL, &rc4030_ops, s,
                           "rc4030.chipset", 0x300);
-    memory_region_add_subregion(sysmem, 0x80000000, &s->iomem_chipset);
     memory_region_init_io(&s->iomem_jazzio, NULL, &jazzio_ops, s,
                           "rc4030.jazzio", 0x00001000);
-    memory_region_add_subregion(sysmem, 0xf0000000, &s->iomem_jazzio);
 
-    return s;
+    memory_region_init_rom_device(&s->dma_tt, o,
+                                  &rc4030_dma_tt_ops, s, "dma-table",
+                                  MAX_TL_ENTRIES * sizeof(dma_pagetable_entry),
+                                  NULL);
+    memory_region_init(&s->dma_tt_alias, o, "dma-table-alias", 0);
+    memory_region_init(&s->dma_mr, o, "dma", INT32_MAX);
+    for (i = 0; i < MAX_TL_ENTRIES; ++i) {
+        memory_region_init_alias(&s->dma_mrs[i], o, "dma-alias",
+                                 get_system_memory(), 0, DMA_PAGESIZE);
+        memory_region_set_enabled(&s->dma_mrs[i], false);
+        memory_region_add_subregion(&s->dma_mr, i * DMA_PAGESIZE,
+                                    &s->dma_mrs[i]);
+    }
+    address_space_init(&s->dma_as, &s->dma_mr, "rc4030-dma");
+}
+
+static void rc4030_unrealize(DeviceState *dev, Error **errp)
+{
+    rc4030State *s = RC4030(dev);
+    int i;
+
+    timer_free(s->periodic_timer);
+
+    address_space_destroy(&s->dma_as);
+    object_unparent(OBJECT(&s->dma_tt));
+    object_unparent(OBJECT(&s->dma_tt_alias));
+    object_unparent(OBJECT(&s->dma_mr));
+    for (i = 0; i < MAX_TL_ENTRIES; ++i) {
+        memory_region_del_subregion(&s->dma_mr, &s->dma_mrs[i]);
+        object_unparent(OBJECT(&s->dma_mrs[i]));
+    }
+}
+
+static void rc4030_class_init(ObjectClass *klass, void *class_data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+
+    dc->realize = rc4030_realize;
+    dc->unrealize = rc4030_unrealize;
+    dc->reset = rc4030_reset;
+}
+
+static const TypeInfo rc4030_info = {
+    .name = TYPE_RC4030,
+    .parent = TYPE_SYS_BUS_DEVICE,
+    .instance_size = sizeof(rc4030State),
+    .instance_init = rc4030_initfn,
+    .class_init = rc4030_class_init,
+};
+
+static void rc4030_register_types(void)
+{
+    type_register_static(&rc4030_info);
+}
+
+type_init(rc4030_register_types)
+
+DeviceState *rc4030_init(rc4030_dma **dmas, MemoryRegion **dma_mr)
+{
+    DeviceState *dev;
+
+    dev = qdev_create(NULL, TYPE_RC4030);
+    qdev_init_nofail(dev);
+
+    *dmas = rc4030_allocate_dmas(dev, 4);
+    *dma_mr = &RC4030(dev)->dma_mr;
+    return dev;
 }
diff --git a/hw/i386/acpi-build.c b/hw/i386/acpi-build.c
index 5593e41..b71e942 100644
--- a/hw/i386/acpi-build.c
+++ b/hw/i386/acpi-build.c
@@ -833,7 +833,7 @@
              * Work-around for old bioses
              * that do not support multiple root buses
              */
-            if (range_base || range_base > range_limit) {
+            if (range_base && range_base <= range_limit) {
                 aml_append(crs,
                            aml_word_io(AML_MIN_FIXED, AML_MAX_FIXED,
                                        AML_POS_DECODE, AML_ENTIRE_RANGE,
@@ -854,7 +854,7 @@
              * Work-around for old bioses
              * that do not support multiple root buses
              */
-            if (range_base || range_base > range_limit) {
+            if (range_base && range_base <= range_limit) {
                 aml_append(crs,
                            aml_dword_memory(AML_POS_DECODE, AML_MIN_FIXED,
                                             AML_MAX_FIXED, AML_NON_CACHEABLE,
@@ -865,7 +865,7 @@
                                             0,
                                             range_limit - range_base + 1));
                 crs_range_insert(mem_ranges, range_base, range_limit);
-          }
+            }
 
             range_base =
                 pci_bridge_get_base(dev, PCI_BASE_ADDRESS_MEM_PREFETCH);
@@ -876,7 +876,7 @@
              * Work-around for old bioses
              * that do not support multiple root buses
              */
-            if (range_base || range_base > range_limit) {
+            if (range_base && range_base <= range_limit) {
                 aml_append(crs,
                            aml_dword_memory(AML_POS_DECODE, AML_MIN_FIXED,
                                             AML_MAX_FIXED, AML_NON_CACHEABLE,
@@ -945,9 +945,8 @@
 
             scope = aml_scope("\\_SB");
             dev = aml_device("PC%.02X", bus_num);
-            aml_append(dev,
-                       aml_name_decl("_UID", aml_string("PC%.02X", bus_num)));
-            aml_append(dev, aml_name_decl("_HID", aml_string("PNP0A03")));
+            aml_append(dev, aml_name_decl("_UID", aml_int(bus_num)));
+            aml_append(dev, aml_name_decl("_HID", aml_eisaid("PNP0A03")));
             aml_append(dev, aml_name_decl("_BBN", aml_int(bus_num)));
 
             if (numa_node != NUMA_NODE_UNASSIGNED) {
diff --git a/hw/i386/pc_piix.c b/hw/i386/pc_piix.c
index 5253e6d..e142f75 100644
--- a/hw/i386/pc_piix.c
+++ b/hw/i386/pc_piix.c
@@ -52,6 +52,7 @@
 #ifdef CONFIG_XEN
 #  include <xen/hvm/hvm_info_table.h>
 #endif
+#include "migration/migration.h"
 
 #define MAX_IDE_BUS 2
 
@@ -305,6 +306,7 @@
 
 static void pc_compat_2_3(MachineState *machine)
 {
+    savevm_skip_section_footers();
 }
 
 static void pc_compat_2_2(MachineState *machine)
diff --git a/hw/i386/pc_q35.c b/hw/i386/pc_q35.c
index 110dfb7..b68263d 100644
--- a/hw/i386/pc_q35.c
+++ b/hw/i386/pc_q35.c
@@ -45,6 +45,7 @@
 #include "hw/usb.h"
 #include "hw/cpu/icc_bus.h"
 #include "qemu/error-report.h"
+#include "migration/migration.h"
 
 /* ICH9 AHCI has 6 ports */
 #define MAX_SATA_PORTS     6
@@ -289,6 +290,7 @@
 
 static void pc_compat_2_3(MachineState *machine)
 {
+    savevm_skip_section_footers();
 }
 
 static void pc_compat_2_2(MachineState *machine)
diff --git a/hw/ide/core.c b/hw/ide/core.c
index fcb9080..1efd98a 100644
--- a/hw/ide/core.c
+++ b/hw/ide/core.c
@@ -2561,6 +2561,7 @@
     .name ="ide_drive/atapi/gesn_state",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = ide_atapi_gesn_needed,
     .fields = (VMStateField[]) {
         VMSTATE_BOOL(events.new_media, IDEState),
         VMSTATE_BOOL(events.eject_request, IDEState),
@@ -2572,6 +2573,7 @@
     .name = "ide_drive/tray_state",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = ide_tray_state_needed,
     .fields = (VMStateField[]) {
         VMSTATE_BOOL(tray_open, IDEState),
         VMSTATE_BOOL(tray_locked, IDEState),
@@ -2585,6 +2587,7 @@
     .minimum_version_id = 1,
     .pre_save = ide_drive_pio_pre_save,
     .post_load = ide_drive_pio_post_load,
+    .needed = ide_drive_pio_state_needed,
     .fields = (VMStateField[]) {
         VMSTATE_INT32(req_nb_sectors, IDEState),
         VMSTATE_VARRAY_INT32(io_buffer, IDEState, io_buffer_total_len, 1,
@@ -2626,19 +2629,11 @@
         VMSTATE_UINT8_V(cdrom_changed, IDEState, 3),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection []) {
-        {
-            .vmsd = &vmstate_ide_drive_pio_state,
-            .needed = ide_drive_pio_state_needed,
-        }, {
-            .vmsd = &vmstate_ide_tray_state,
-            .needed = ide_tray_state_needed,
-        }, {
-            .vmsd = &vmstate_ide_atapi_gesn_state,
-            .needed = ide_atapi_gesn_needed,
-        }, {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_ide_drive_pio_state,
+        &vmstate_ide_tray_state,
+        &vmstate_ide_atapi_gesn_state,
+        NULL
     }
 };
 
@@ -2646,6 +2641,7 @@
     .name ="ide_bus/error",
     .version_id = 2,
     .minimum_version_id = 1,
+    .needed = ide_error_needed,
     .fields = (VMStateField[]) {
         VMSTATE_INT32(error_status, IDEBus),
         VMSTATE_INT64_V(retry_sector_num, IDEBus, 2),
@@ -2664,13 +2660,9 @@
         VMSTATE_UINT8(unit, IDEBus),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection []) {
-        {
-            .vmsd = &vmstate_ide_error_status,
-            .needed = ide_error_needed,
-        }, {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_ide_error_status,
+        NULL
     }
 };
 
diff --git a/hw/ide/pci.c b/hw/ide/pci.c
index 4b5e32d..4afd0cf 100644
--- a/hw/ide/pci.c
+++ b/hw/ide/pci.c
@@ -350,6 +350,7 @@
     .name = "ide bmdma_current",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = ide_bmdma_current_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT32(cur_addr, BMDMAState),
         VMSTATE_UINT32(cur_prd_last, BMDMAState),
@@ -363,6 +364,7 @@
     .name ="ide bmdma/status",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = ide_bmdma_status_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT8(status, BMDMAState),
         VMSTATE_END_OF_LIST()
@@ -383,16 +385,10 @@
         VMSTATE_UINT8(migration_retry_unit, BMDMAState),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection []) {
-        {
-            .vmsd = &vmstate_bmdma_current,
-            .needed = ide_bmdma_current_needed,
-        }, {
-            .vmsd = &vmstate_bmdma_status,
-            .needed = ide_bmdma_status_needed,
-        }, {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_bmdma_current,
+        &vmstate_bmdma_status,
+        NULL
     }
 };
 
diff --git a/hw/input/pckbd.c b/hw/input/pckbd.c
index 9b9a7d7..ddac69d 100644
--- a/hw/input/pckbd.c
+++ b/hw/input/pckbd.c
@@ -391,23 +391,24 @@
     return 0;
 }
 
-static const VMStateDescription vmstate_kbd_outport = {
-    .name = "pckbd_outport",
-    .version_id = 1,
-    .minimum_version_id = 1,
-    .post_load = kbd_outport_post_load,
-    .fields = (VMStateField[]) {
-        VMSTATE_UINT8(outport, KBDState),
-        VMSTATE_END_OF_LIST()
-    }
-};
-
 static bool kbd_outport_needed(void *opaque)
 {
     KBDState *s = opaque;
     return s->outport != kbd_outport_default(s);
 }
 
+static const VMStateDescription vmstate_kbd_outport = {
+    .name = "pckbd_outport",
+    .version_id = 1,
+    .minimum_version_id = 1,
+    .post_load = kbd_outport_post_load,
+    .needed = kbd_outport_needed,
+    .fields = (VMStateField[]) {
+        VMSTATE_UINT8(outport, KBDState),
+        VMSTATE_END_OF_LIST()
+    }
+};
+
 static int kbd_post_load(void *opaque, int version_id)
 {
     KBDState *s = opaque;
@@ -430,12 +431,9 @@
         VMSTATE_UINT8(pending, KBDState),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_kbd_outport,
-            .needed = kbd_outport_needed,
-        },
-        VMSTATE_END_OF_LIST()
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_kbd_outport,
+        NULL
     }
 };
 
diff --git a/hw/input/ps2.c b/hw/input/ps2.c
index 4baeea2..fdbe565 100644
--- a/hw/input/ps2.c
+++ b/hw/input/ps2.c
@@ -677,6 +677,7 @@
     .version_id = 3,
     .minimum_version_id = 2,
     .post_load = ps2_kbd_ledstate_post_load,
+    .needed = ps2_keyboard_ledstate_needed,
     .fields = (VMStateField[]) {
         VMSTATE_INT32(ledstate, PS2KbdState),
         VMSTATE_END_OF_LIST()
@@ -717,13 +718,9 @@
         VMSTATE_INT32_V(scancode_set, PS2KbdState,3),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection []) {
-        {
-            .vmsd = &vmstate_ps2_keyboard_ledstate,
-            .needed = ps2_keyboard_ledstate_needed,
-        }, {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_ps2_keyboard_ledstate,
+        NULL
     }
 };
 
diff --git a/hw/intc/apic_common.c b/hw/intc/apic_common.c
index d595d63..0032b97 100644
--- a/hw/intc/apic_common.c
+++ b/hw/intc/apic_common.c
@@ -369,6 +369,7 @@
     .name = "apic_sipi",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = apic_common_sipi_needed,
     .fields = (VMStateField[]) {
         VMSTATE_INT32(sipi_vector, APICCommonState),
         VMSTATE_INT32(wait_for_sipi, APICCommonState),
@@ -408,12 +409,9 @@
                       APICCommonState), /* open-coded timer state */
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_apic_common_sipi,
-            .needed = apic_common_sipi_needed,
-        },
-        VMSTATE_END_OF_LIST()
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_apic_common_sipi,
+        NULL
     }
 };
 
diff --git a/hw/isa/lpc_ich9.c b/hw/isa/lpc_ich9.c
index 71a9f7a..b3e0b1f 100644
--- a/hw/isa/lpc_ich9.c
+++ b/hw/isa/lpc_ich9.c
@@ -650,6 +650,7 @@
     .name = "ICH9LPC/rst_cnt",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = ich9_rst_cnt_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT8(rst_cnt, ICH9LPCState),
         VMSTATE_END_OF_LIST()
@@ -669,12 +670,9 @@
         VMSTATE_UINT32(sci_level, ICH9LPCState),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_ich9_rst_cnt,
-            .needed = ich9_rst_cnt_needed
-        },
-        { 0 }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_ich9_rst_cnt,
+        NULL
     }
 };
 
diff --git a/hw/mips/Makefile.objs b/hw/mips/Makefile.objs
index 0a652f8..9633f3a 100644
--- a/hw/mips/Makefile.objs
+++ b/hw/mips/Makefile.objs
@@ -1,4 +1,5 @@
-obj-y += mips_r4k.o mips_jazz.o mips_malta.o mips_mipssim.o
+obj-y += mips_r4k.o mips_malta.o mips_mipssim.o
 obj-y += addr.o cputimer.o mips_int.o
+obj-$(CONFIG_JAZZ) += mips_jazz.o
 obj-$(CONFIG_FULONG) += mips_fulong2e.o
 obj-y += gt64xxx_pci.o
diff --git a/hw/mips/mips_jazz.c b/hw/mips/mips_jazz.c
index 2c153e0..9d60633 100644
--- a/hw/mips/mips_jazz.c
+++ b/hw/mips/mips_jazz.c
@@ -135,16 +135,16 @@
     MIPSCPU *cpu;
     CPUClass *cc;
     CPUMIPSState *env;
-    qemu_irq *rc4030, *i8259;
+    qemu_irq *i8259;
     rc4030_dma *dmas;
-    void* rc4030_opaque;
+    MemoryRegion *rc4030_dma_mr;
     MemoryRegion *isa_mem = g_new(MemoryRegion, 1);
     MemoryRegion *isa_io = g_new(MemoryRegion, 1);
     MemoryRegion *rtc = g_new(MemoryRegion, 1);
     MemoryRegion *i8042 = g_new(MemoryRegion, 1);
     MemoryRegion *dma_dummy = g_new(MemoryRegion, 1);
     NICInfo *nd;
-    DeviceState *dev;
+    DeviceState *dev, *rc4030;
     SysBusDevice *sysbus;
     ISABus *isa_bus;
     ISADevice *pit;
@@ -157,12 +157,7 @@
 
     /* init CPUs */
     if (cpu_model == NULL) {
-#ifdef TARGET_MIPS64
         cpu_model = "R4000";
-#else
-        /* FIXME: All wrong, this maybe should be R3000 for the older JAZZs. */
-        cpu_model = "24Kf";
-#endif
     }
     cpu = cpu_mips_init(cpu_model);
     if (cpu == NULL) {
@@ -218,8 +213,14 @@
     cpu_mips_clock_init(env);
 
     /* Chipset */
-    rc4030_opaque = rc4030_init(env->irq[6], env->irq[3], &rc4030, &dmas,
-                                address_space);
+    rc4030 = rc4030_init(&dmas, &rc4030_dma_mr);
+    sysbus = SYS_BUS_DEVICE(rc4030);
+    sysbus_connect_irq(sysbus, 0, env->irq[6]);
+    sysbus_connect_irq(sysbus, 1, env->irq[3]);
+    memory_region_add_subregion(address_space, 0x80000000,
+                                sysbus_mmio_get_region(sysbus, 0));
+    memory_region_add_subregion(address_space, 0xf0000000,
+                                sysbus_mmio_get_region(sysbus, 1));
     memory_region_init_io(dma_dummy, NULL, &dma_dummy_ops, NULL, "dummy_dma", 0x1000);
     memory_region_add_subregion(address_space, 0x8000d000, dma_dummy);
 
@@ -246,7 +247,7 @@
         sysbus = SYS_BUS_DEVICE(dev);
         sysbus_mmio_map(sysbus, 0, 0x60080000);
         sysbus_mmio_map(sysbus, 1, 0x40000000);
-        sysbus_connect_irq(sysbus, 0, rc4030[3]);
+        sysbus_connect_irq(sysbus, 0, qdev_get_gpio_in(rc4030, 3));
         {
             /* Simple ROM, so user doesn't have to provide one */
             MemoryRegion *rom_mr = g_new(MemoryRegion, 1);
@@ -272,8 +273,17 @@
         if (!nd->model)
             nd->model = g_strdup("dp83932");
         if (strcmp(nd->model, "dp83932") == 0) {
-            dp83932_init(nd, 0x80001000, 2, get_system_memory(), rc4030[4],
-                         rc4030_opaque, rc4030_dma_memory_rw);
+            qemu_check_nic_model(nd, "dp83932");
+
+            dev = qdev_create(NULL, "dp8393x");
+            qdev_set_nic_properties(dev, nd);
+            qdev_prop_set_uint8(dev, "it_shift", 2);
+            qdev_prop_set_ptr(dev, "dma_mr", rc4030_dma_mr);
+            qdev_init_nofail(dev);
+            sysbus = SYS_BUS_DEVICE(dev);
+            sysbus_mmio_map(sysbus, 0, 0x80001000);
+            sysbus_mmio_map(sysbus, 1, 0x8000b000);
+            sysbus_connect_irq(sysbus, 0, qdev_get_gpio_in(rc4030, 4));
             break;
         } else if (is_help_option(nd->model)) {
             fprintf(stderr, "qemu: Supported NICs: dp83932\n");
@@ -287,7 +297,7 @@
     /* SCSI adapter */
     esp_init(0x80002000, 0,
              rc4030_dma_read, rc4030_dma_write, dmas[0],
-             rc4030[5], &esp_reset, &dma_enable);
+             qdev_get_gpio_in(rc4030, 5), &esp_reset, &dma_enable);
 
     /* Floppy */
     if (drive_get_max_bus(IF_FLOPPY) >= MAX_FD) {
@@ -297,7 +307,7 @@
     for (n = 0; n < MAX_FD; n++) {
         fds[n] = drive_get(IF_FLOPPY, 0, n);
     }
-    fdctrl_init_sysbus(rc4030[1], 0, 0x80003000, fds);
+    fdctrl_init_sysbus(qdev_get_gpio_in(rc4030, 1), 0, 0x80003000, fds);
 
     /* Real time clock */
     rtc_init(isa_bus, 1980, NULL);
@@ -305,23 +315,26 @@
     memory_region_add_subregion(address_space, 0x80004000, rtc);
 
     /* Keyboard (i8042) */
-    i8042_mm_init(rc4030[6], rc4030[7], i8042, 0x1000, 0x1);
+    i8042_mm_init(qdev_get_gpio_in(rc4030, 6), qdev_get_gpio_in(rc4030, 7),
+                  i8042, 0x1000, 0x1);
     memory_region_add_subregion(address_space, 0x80005000, i8042);
 
     /* Serial ports */
     if (serial_hds[0]) {
-        serial_mm_init(address_space, 0x80006000, 0, rc4030[8], 8000000/16,
+        serial_mm_init(address_space, 0x80006000, 0,
+                       qdev_get_gpio_in(rc4030, 8), 8000000/16,
                        serial_hds[0], DEVICE_NATIVE_ENDIAN);
     }
     if (serial_hds[1]) {
-        serial_mm_init(address_space, 0x80007000, 0, rc4030[9], 8000000/16,
+        serial_mm_init(address_space, 0x80007000, 0,
+                       qdev_get_gpio_in(rc4030, 9), 8000000/16,
                        serial_hds[1], DEVICE_NATIVE_ENDIAN);
     }
 
     /* Parallel port */
     if (parallel_hds[0])
-        parallel_mm_init(address_space, 0x80008000, 0, rc4030[0],
-                         parallel_hds[0]);
+        parallel_mm_init(address_space, 0x80008000, 0,
+                         qdev_get_gpio_in(rc4030, 0), parallel_hds[0]);
 
     /* FIXME: missing Jazz sound at 0x8000c000, rc4030[2] */
 
diff --git a/hw/mips/mips_malta.c b/hw/mips/mips_malta.c
index 5140882..786a8f0 100644
--- a/hw/mips/mips_malta.c
+++ b/hw/mips/mips_malta.c
@@ -97,7 +97,7 @@
 static ISADevice *pit;
 
 static struct _loaderparams {
-    int ram_size;
+    int ram_size, ram_low_size;
     const char *kernel_filename;
     const char *kernel_cmdline;
     const char *initrd_filename;
@@ -641,8 +641,8 @@
     stl_p(p++, 0x34a50000 | (ENVP_ADDR & 0xffff));               /* ori a1, a1, low(ENVP_ADDR) */
     stl_p(p++, 0x3c060000 | (((ENVP_ADDR + 8) >> 16) & 0xffff)); /* lui a2, high(ENVP_ADDR + 8) */
     stl_p(p++, 0x34c60000 | ((ENVP_ADDR + 8) & 0xffff));         /* ori a2, a2, low(ENVP_ADDR + 8) */
-    stl_p(p++, 0x3c070000 | (loaderparams.ram_size >> 16));     /* lui a3, high(ram_size) */
-    stl_p(p++, 0x34e70000 | (loaderparams.ram_size & 0xffff));  /* ori a3, a3, low(ram_size) */
+    stl_p(p++, 0x3c070000 | (loaderparams.ram_low_size >> 16));     /* lui a3, high(ram_low_size) */
+    stl_p(p++, 0x34e70000 | (loaderparams.ram_low_size & 0xffff));  /* ori a3, a3, low(ram_low_size) */
 
     /* Load BAR registers as done by YAMON */
     stl_p(p++, 0x3c09b400);                                      /* lui t1, 0xb400 */
@@ -851,8 +851,10 @@
     }
 
     prom_set(prom_buf, prom_index++, "memsize");
-    prom_set(prom_buf, prom_index++, "%i",
-             MIN(loaderparams.ram_size, 256 << 20));
+    prom_set(prom_buf, prom_index++, "%u", loaderparams.ram_low_size);
+
+    prom_set(prom_buf, prom_index++, "ememsize");
+    prom_set(prom_buf, prom_index++, "%u", loaderparams.ram_size);
 
     prom_set(prom_buf, prom_index++, "modetty0");
     prom_set(prom_buf, prom_index++, "38400n8r");
@@ -1054,7 +1056,8 @@
         }
 
         /* Write a small bootloader to the flash location. */
-        loaderparams.ram_size = ram_low_size;
+        loaderparams.ram_size = ram_size;
+        loaderparams.ram_low_size = ram_low_size;
         loaderparams.kernel_filename = kernel_filename;
         loaderparams.kernel_cmdline = kernel_cmdline;
         loaderparams.initrd_filename = initrd_filename;
diff --git a/hw/net/dp8393x.c b/hw/net/dp8393x.c
index 7ce13d2..cd889bc 100644
--- a/hw/net/dp8393x.c
+++ b/hw/net/dp8393x.c
@@ -17,20 +17,15 @@
  * with this program; if not, see <http://www.gnu.org/licenses/>.
  */
 
-#include "hw/hw.h"
-#include "qemu/timer.h"
+#include "hw/sysbus.h"
+#include "hw/devices.h"
 #include "net/net.h"
-#include "hw/mips/mips.h"
+#include "qemu/timer.h"
+#include <zlib.h>
 
 //#define DEBUG_SONIC
 
-/* Calculate CRCs properly on Rx packets */
-#define SONIC_CALCULATE_RXCRC
-
-#if defined(SONIC_CALCULATE_RXCRC)
-/* For crc32 */
-#include <zlib.h>
-#endif
+#define SONIC_PROM_SIZE 0x1000
 
 #ifdef DEBUG_SONIC
 #define DPRINTF(fmt, ...) \
@@ -145,9 +140,14 @@
 #define SONIC_ISR_PINT   0x0800
 #define SONIC_ISR_LCD    0x1000
 
+#define TYPE_DP8393X "dp8393x"
+#define DP8393X(obj) OBJECT_CHECK(dp8393xState, (obj), TYPE_DP8393X)
+
 typedef struct dp8393xState {
+    SysBusDevice parent_obj;
+
     /* Hardware */
-    int it_shift;
+    uint8_t it_shift;
     qemu_irq irq;
 #ifdef DEBUG_SONIC
     int irq_level;
@@ -156,8 +156,8 @@
     int64_t wt_last_update;
     NICConf conf;
     NICState *nic;
-    MemoryRegion *address_space;
     MemoryRegion mmio;
+    MemoryRegion prom;
 
     /* Registers */
     uint8_t cam[16][6];
@@ -168,8 +168,8 @@
     int loopback_packet;
 
     /* Memory access */
-    void (*memory_rw)(void *opaque, hwaddr addr, uint8_t *buf, int len, int is_write);
-    void* mem_opaque;
+    void *dma_mr;
+    AddressSpace as;
 } dp8393xState;
 
 static void dp8393x_update_irq(dp8393xState *s)
@@ -190,7 +190,7 @@
     qemu_set_irq(s->irq, level);
 }
 
-static void do_load_cam(dp8393xState *s)
+static void dp8393x_do_load_cam(dp8393xState *s)
 {
     uint16_t data[8];
     int width, size;
@@ -201,9 +201,9 @@
 
     while (s->regs[SONIC_CDC] & 0x1f) {
         /* Fill current entry */
-        s->memory_rw(s->mem_opaque,
+        address_space_rw(&s->as,
             (s->regs[SONIC_URRA] << 16) | s->regs[SONIC_CDP],
-            (uint8_t *)data, size, 0);
+            MEMTXATTRS_UNSPECIFIED, (uint8_t *)data, size, 0);
         s->cam[index][0] = data[1 * width] & 0xff;
         s->cam[index][1] = data[1 * width] >> 8;
         s->cam[index][2] = data[2 * width] & 0xff;
@@ -220,9 +220,9 @@
     }
 
     /* Read CAM enable */
-    s->memory_rw(s->mem_opaque,
+    address_space_rw(&s->as,
         (s->regs[SONIC_URRA] << 16) | s->regs[SONIC_CDP],
-        (uint8_t *)data, size, 0);
+        MEMTXATTRS_UNSPECIFIED, (uint8_t *)data, size, 0);
     s->regs[SONIC_CE] = data[0 * width];
     DPRINTF("load cam done. cam enable mask 0x%04x\n", s->regs[SONIC_CE]);
 
@@ -232,7 +232,7 @@
     dp8393x_update_irq(s);
 }
 
-static void do_read_rra(dp8393xState *s)
+static void dp8393x_do_read_rra(dp8393xState *s)
 {
     uint16_t data[8];
     int width, size;
@@ -240,9 +240,9 @@
     /* Read memory */
     width = (s->regs[SONIC_DCR] & SONIC_DCR_DW) ? 2 : 1;
     size = sizeof(uint16_t) * 4 * width;
-    s->memory_rw(s->mem_opaque,
+    address_space_rw(&s->as,
         (s->regs[SONIC_URRA] << 16) | s->regs[SONIC_RRP],
-        (uint8_t *)data, size, 0);
+        MEMTXATTRS_UNSPECIFIED, (uint8_t *)data, size, 0);
 
     /* Update SONIC registers */
     s->regs[SONIC_CRBA0] = data[0 * width];
@@ -272,7 +272,7 @@
     s->regs[SONIC_CR] &= ~SONIC_CR_RRRA;
 }
 
-static void do_software_reset(dp8393xState *s)
+static void dp8393x_do_software_reset(dp8393xState *s)
 {
     timer_del(s->watchdog);
 
@@ -280,7 +280,7 @@
     s->regs[SONIC_CR] |= SONIC_CR_RST | SONIC_CR_RXDIS;
 }
 
-static void set_next_tick(dp8393xState *s)
+static void dp8393x_set_next_tick(dp8393xState *s)
 {
     uint32_t ticks;
     int64_t delay;
@@ -296,7 +296,7 @@
     timer_mod(s->watchdog, s->wt_last_update + delay);
 }
 
-static void update_wt_regs(dp8393xState *s)
+static void dp8393x_update_wt_regs(dp8393xState *s)
 {
     int64_t elapsed;
     uint32_t val;
@@ -311,33 +311,33 @@
     val -= elapsed / 5000000;
     s->regs[SONIC_WT1] = (val >> 16) & 0xffff;
     s->regs[SONIC_WT0] = (val >> 0)  & 0xffff;
-    set_next_tick(s);
+    dp8393x_set_next_tick(s);
 
 }
 
-static void do_start_timer(dp8393xState *s)
+static void dp8393x_do_start_timer(dp8393xState *s)
 {
     s->regs[SONIC_CR] &= ~SONIC_CR_STP;
-    set_next_tick(s);
+    dp8393x_set_next_tick(s);
 }
 
-static void do_stop_timer(dp8393xState *s)
+static void dp8393x_do_stop_timer(dp8393xState *s)
 {
     s->regs[SONIC_CR] &= ~SONIC_CR_ST;
-    update_wt_regs(s);
+    dp8393x_update_wt_regs(s);
 }
 
-static void do_receiver_enable(dp8393xState *s)
+static void dp8393x_do_receiver_enable(dp8393xState *s)
 {
     s->regs[SONIC_CR] &= ~SONIC_CR_RXDIS;
 }
 
-static void do_receiver_disable(dp8393xState *s)
+static void dp8393x_do_receiver_disable(dp8393xState *s)
 {
     s->regs[SONIC_CR] &= ~SONIC_CR_RXEN;
 }
 
-static void do_transmit_packets(dp8393xState *s)
+static void dp8393x_do_transmit_packets(dp8393xState *s)
 {
     NetClientState *nc = qemu_get_queue(s->nic);
     uint16_t data[12];
@@ -353,9 +353,9 @@
                 (s->regs[SONIC_UTDA] << 16) | s->regs[SONIC_CTDA]);
         size = sizeof(uint16_t) * 6 * width;
         s->regs[SONIC_TTDA] = s->regs[SONIC_CTDA];
-        s->memory_rw(s->mem_opaque,
+        address_space_rw(&s->as,
             ((s->regs[SONIC_UTDA] << 16) | s->regs[SONIC_TTDA]) + sizeof(uint16_t) * width,
-            (uint8_t *)data, size, 0);
+            MEMTXATTRS_UNSPECIFIED, (uint8_t *)data, size, 0);
         tx_len = 0;
 
         /* Update registers */
@@ -379,18 +379,18 @@
             if (tx_len + len > sizeof(s->tx_buffer)) {
                 len = sizeof(s->tx_buffer) - tx_len;
             }
-            s->memory_rw(s->mem_opaque,
+            address_space_rw(&s->as,
                 (s->regs[SONIC_TSA1] << 16) | s->regs[SONIC_TSA0],
-                &s->tx_buffer[tx_len], len, 0);
+                MEMTXATTRS_UNSPECIFIED, &s->tx_buffer[tx_len], len, 0);
             tx_len += len;
 
             i++;
             if (i != s->regs[SONIC_TFC]) {
                 /* Read next fragment details */
                 size = sizeof(uint16_t) * 3 * width;
-                s->memory_rw(s->mem_opaque,
+                address_space_rw(&s->as,
                     ((s->regs[SONIC_UTDA] << 16) | s->regs[SONIC_TTDA]) + sizeof(uint16_t) * (4 + 3 * i) * width,
-                    (uint8_t *)data, size, 0);
+                    MEMTXATTRS_UNSPECIFIED, (uint8_t *)data, size, 0);
                 s->regs[SONIC_TSA0] = data[0 * width];
                 s->regs[SONIC_TSA1] = data[1 * width];
                 s->regs[SONIC_TFS] = data[2 * width];
@@ -422,16 +422,16 @@
         /* Write status */
         data[0 * width] = s->regs[SONIC_TCR] & 0x0fff; /* status */
         size = sizeof(uint16_t) * width;
-        s->memory_rw(s->mem_opaque,
+        address_space_rw(&s->as,
             (s->regs[SONIC_UTDA] << 16) | s->regs[SONIC_TTDA],
-            (uint8_t *)data, size, 1);
+            MEMTXATTRS_UNSPECIFIED, (uint8_t *)data, size, 1);
 
         if (!(s->regs[SONIC_CR] & SONIC_CR_HTX)) {
             /* Read footer of packet */
             size = sizeof(uint16_t) * width;
-            s->memory_rw(s->mem_opaque,
+            address_space_rw(&s->as,
                 ((s->regs[SONIC_UTDA] << 16) | s->regs[SONIC_TTDA]) + sizeof(uint16_t) * (4 + 3 * s->regs[SONIC_TFC]) * width,
-                (uint8_t *)data, size, 0);
+                MEMTXATTRS_UNSPECIFIED, (uint8_t *)data, size, 0);
             s->regs[SONIC_CTDA] = data[0 * width] & ~0x1;
             if (data[0 * width] & 0x1) {
                 /* EOL detected */
@@ -446,12 +446,12 @@
     dp8393x_update_irq(s);
 }
 
-static void do_halt_transmission(dp8393xState *s)
+static void dp8393x_do_halt_transmission(dp8393xState *s)
 {
     /* Nothing to do */
 }
 
-static void do_command(dp8393xState *s, uint16_t command)
+static void dp8393x_do_command(dp8393xState *s, uint16_t command)
 {
     if ((s->regs[SONIC_CR] & SONIC_CR_RST) && !(command & SONIC_CR_RST)) {
         s->regs[SONIC_CR] &= ~SONIC_CR_RST;
@@ -461,34 +461,36 @@
     s->regs[SONIC_CR] |= (command & SONIC_CR_MASK);
 
     if (command & SONIC_CR_HTX)
-        do_halt_transmission(s);
+        dp8393x_do_halt_transmission(s);
     if (command & SONIC_CR_TXP)
-        do_transmit_packets(s);
+        dp8393x_do_transmit_packets(s);
     if (command & SONIC_CR_RXDIS)
-        do_receiver_disable(s);
+        dp8393x_do_receiver_disable(s);
     if (command & SONIC_CR_RXEN)
-        do_receiver_enable(s);
+        dp8393x_do_receiver_enable(s);
     if (command & SONIC_CR_STP)
-        do_stop_timer(s);
+        dp8393x_do_stop_timer(s);
     if (command & SONIC_CR_ST)
-        do_start_timer(s);
+        dp8393x_do_start_timer(s);
     if (command & SONIC_CR_RST)
-        do_software_reset(s);
+        dp8393x_do_software_reset(s);
     if (command & SONIC_CR_RRRA)
-        do_read_rra(s);
+        dp8393x_do_read_rra(s);
     if (command & SONIC_CR_LCAM)
-        do_load_cam(s);
+        dp8393x_do_load_cam(s);
 }
 
-static uint16_t read_register(dp8393xState *s, int reg)
+static uint64_t dp8393x_read(void *opaque, hwaddr addr, unsigned int size)
 {
+    dp8393xState *s = opaque;
+    int reg = addr >> s->it_shift;
     uint16_t val = 0;
 
     switch (reg) {
         /* Update data before reading it */
         case SONIC_WT0:
         case SONIC_WT1:
-            update_wt_regs(s);
+            dp8393x_update_wt_regs(s);
             val = s->regs[reg];
             break;
         /* Accept read to some registers only when in reset mode */
@@ -510,14 +512,18 @@
     return val;
 }
 
-static void write_register(dp8393xState *s, int reg, uint16_t val)
+static void dp8393x_write(void *opaque, hwaddr addr, uint64_t data,
+                          unsigned int size)
 {
-    DPRINTF("write 0x%04x to reg %s\n", val, reg_names[reg]);
+    dp8393xState *s = opaque;
+    int reg = addr >> s->it_shift;
+
+    DPRINTF("write 0x%04x to reg %s\n", (uint16_t)data, reg_names[reg]);
 
     switch (reg) {
         /* Command register */
         case SONIC_CR:
-            do_command(s, val);
+            dp8393x_do_command(s, data);
             break;
         /* Prevent write to read-only registers */
         case SONIC_CAP2:
@@ -530,37 +536,37 @@
         /* Accept write to some registers only when in reset mode */
         case SONIC_DCR:
             if (s->regs[SONIC_CR] & SONIC_CR_RST) {
-                s->regs[reg] = val & 0xbfff;
+                s->regs[reg] = data & 0xbfff;
             } else {
                 DPRINTF("writing to DCR invalid\n");
             }
             break;
         case SONIC_DCR2:
             if (s->regs[SONIC_CR] & SONIC_CR_RST) {
-                s->regs[reg] = val & 0xf017;
+                s->regs[reg] = data & 0xf017;
             } else {
                 DPRINTF("writing to DCR2 invalid\n");
             }
             break;
         /* 12 lower bytes are Read Only */
         case SONIC_TCR:
-            s->regs[reg] = val & 0xf000;
+            s->regs[reg] = data & 0xf000;
             break;
         /* 9 lower bytes are Read Only */
         case SONIC_RCR:
-            s->regs[reg] = val & 0xffe0;
+            s->regs[reg] = data & 0xffe0;
             break;
         /* Ignore most significant bit */
         case SONIC_IMR:
-            s->regs[reg] = val & 0x7fff;
+            s->regs[reg] = data & 0x7fff;
             dp8393x_update_irq(s);
             break;
         /* Clear bits by writing 1 to them */
         case SONIC_ISR:
-            val &= s->regs[reg];
-            s->regs[reg] &= ~val;
-            if (val & SONIC_ISR_RBE) {
-                do_read_rra(s);
+            data &= s->regs[reg];
+            s->regs[reg] &= ~data;
+            if (data & SONIC_ISR_RBE) {
+                dp8393x_do_read_rra(s);
             }
             dp8393x_update_irq(s);
             break;
@@ -569,24 +575,32 @@
         case SONIC_REA:
         case SONIC_RRP:
         case SONIC_RWP:
-            s->regs[reg] = val & 0xfffe;
+            s->regs[reg] = data & 0xfffe;
             break;
         /* Invert written value for some registers */
         case SONIC_CRCT:
         case SONIC_FAET:
         case SONIC_MPT:
-            s->regs[reg] = val ^ 0xffff;
+            s->regs[reg] = data ^ 0xffff;
             break;
         /* All other registers have no special contrainst */
         default:
-            s->regs[reg] = val;
+            s->regs[reg] = data;
     }
 
     if (reg == SONIC_WT0 || reg == SONIC_WT1) {
-        set_next_tick(s);
+        dp8393x_set_next_tick(s);
     }
 }
 
+static const MemoryRegionOps dp8393x_ops = {
+    .read = dp8393x_read,
+    .write = dp8393x_write,
+    .impl.min_access_size = 2,
+    .impl.max_access_size = 2,
+    .endianness = DEVICE_NATIVE_ENDIAN,
+};
+
 static void dp8393x_watchdog(void *opaque)
 {
     dp8393xState *s = opaque;
@@ -597,84 +611,14 @@
 
     s->regs[SONIC_WT1] = 0xffff;
     s->regs[SONIC_WT0] = 0xffff;
-    set_next_tick(s);
+    dp8393x_set_next_tick(s);
 
     /* Signal underflow */
     s->regs[SONIC_ISR] |= SONIC_ISR_TC;
     dp8393x_update_irq(s);
 }
 
-static uint32_t dp8393x_readw(void *opaque, hwaddr addr)
-{
-    dp8393xState *s = opaque;
-    int reg;
-
-    if ((addr & ((1 << s->it_shift) - 1)) != 0) {
-        return 0;
-    }
-
-    reg = addr >> s->it_shift;
-    return read_register(s, reg);
-}
-
-static uint32_t dp8393x_readb(void *opaque, hwaddr addr)
-{
-    uint16_t v = dp8393x_readw(opaque, addr & ~0x1);
-    return (v >> (8 * (addr & 0x1))) & 0xff;
-}
-
-static uint32_t dp8393x_readl(void *opaque, hwaddr addr)
-{
-    uint32_t v;
-    v = dp8393x_readw(opaque, addr);
-    v |= dp8393x_readw(opaque, addr + 2) << 16;
-    return v;
-}
-
-static void dp8393x_writew(void *opaque, hwaddr addr, uint32_t val)
-{
-    dp8393xState *s = opaque;
-    int reg;
-
-    if ((addr & ((1 << s->it_shift) - 1)) != 0) {
-        return;
-    }
-
-    reg = addr >> s->it_shift;
-
-    write_register(s, reg, (uint16_t)val);
-}
-
-static void dp8393x_writeb(void *opaque, hwaddr addr, uint32_t val)
-{
-    uint16_t old_val = dp8393x_readw(opaque, addr & ~0x1);
-
-    switch (addr & 3) {
-    case 0:
-        val = val | (old_val & 0xff00);
-        break;
-    case 1:
-        val = (val << 8) | (old_val & 0x00ff);
-        break;
-    }
-    dp8393x_writew(opaque, addr & ~0x1, val);
-}
-
-static void dp8393x_writel(void *opaque, hwaddr addr, uint32_t val)
-{
-    dp8393x_writew(opaque, addr, val & 0xffff);
-    dp8393x_writew(opaque, addr + 2, (val >> 16) & 0xffff);
-}
-
-static const MemoryRegionOps dp8393x_ops = {
-    .old_mmio = {
-        .read = { dp8393x_readb, dp8393x_readw, dp8393x_readl, },
-        .write = { dp8393x_writeb, dp8393x_writew, dp8393x_writel, },
-    },
-    .endianness = DEVICE_NATIVE_ENDIAN,
-};
-
-static int nic_can_receive(NetClientState *nc)
+static int dp8393x_can_receive(NetClientState *nc)
 {
     dp8393xState *s = qemu_get_nic_opaque(nc);
 
@@ -685,7 +629,8 @@
     return 1;
 }
 
-static int receive_filter(dp8393xState *s, const uint8_t * buf, int size)
+static int dp8393x_receive_filter(dp8393xState *s, const uint8_t * buf,
+                                  int size)
 {
     static const uint8_t bcast[] = {0xff, 0xff, 0xff, 0xff, 0xff, 0xff};
     int i;
@@ -723,7 +668,8 @@
     return -1;
 }
 
-static ssize_t nic_receive(NetClientState *nc, const uint8_t * buf, size_t size)
+static ssize_t dp8393x_receive(NetClientState *nc, const uint8_t * buf,
+                               size_t size)
 {
     dp8393xState *s = qemu_get_nic_opaque(nc);
     uint16_t data[10];
@@ -737,7 +683,7 @@
     s->regs[SONIC_RCR] &= ~(SONIC_RCR_PRX | SONIC_RCR_LBK | SONIC_RCR_FAER |
         SONIC_RCR_CRCR | SONIC_RCR_LPKT | SONIC_RCR_BC | SONIC_RCR_MC);
 
-    packet_type = receive_filter(s, buf, size);
+    packet_type = dp8393x_receive_filter(s, buf, size);
     if (packet_type < 0) {
         DPRINTF("packet not for netcard\n");
         return -1;
@@ -750,7 +696,8 @@
         /* Are we still in resource exhaustion? */
         size = sizeof(uint16_t) * 1 * width;
         address = ((s->regs[SONIC_URDA] << 16) | s->regs[SONIC_CRDA]) + sizeof(uint16_t) * 5 * width;
-        s->memory_rw(s->mem_opaque, address, (uint8_t*)data, size, 0);
+        address_space_rw(&s->as, address, MEMTXATTRS_UNSPECIFIED,
+                         (uint8_t *)data, size, 0);
         if (data[0 * width] & 0x1) {
             /* Still EOL ; stop reception */
             return -1;
@@ -764,18 +711,16 @@
     s->regs[SONIC_TRBA0] = s->regs[SONIC_CRBA0];
 
     /* Calculate the ethernet checksum */
-#ifdef SONIC_CALCULATE_RXCRC
     checksum = cpu_to_le32(crc32(0, buf, rx_len));
-#else
-    checksum = 0;
-#endif
 
     /* Put packet into RBA */
     DPRINTF("Receive packet at %08x\n", (s->regs[SONIC_CRBA1] << 16) | s->regs[SONIC_CRBA0]);
     address = (s->regs[SONIC_CRBA1] << 16) | s->regs[SONIC_CRBA0];
-    s->memory_rw(s->mem_opaque, address, (uint8_t*)buf, rx_len, 1);
+    address_space_rw(&s->as, address,
+        MEMTXATTRS_UNSPECIFIED, (uint8_t *)buf, rx_len, 1);
     address += rx_len;
-    s->memory_rw(s->mem_opaque, address, (uint8_t*)&checksum, 4, 1);
+    address_space_rw(&s->as, address,
+        MEMTXATTRS_UNSPECIFIED, (uint8_t *)&checksum, 4, 1);
     rx_len += 4;
     s->regs[SONIC_CRBA1] = address >> 16;
     s->regs[SONIC_CRBA0] = address & 0xffff;
@@ -803,29 +748,30 @@
     data[3 * width] = s->regs[SONIC_TRBA1]; /* pkt_ptr1 */
     data[4 * width] = s->regs[SONIC_RSC]; /* seq_no */
     size = sizeof(uint16_t) * 5 * width;
-    s->memory_rw(s->mem_opaque, (s->regs[SONIC_URDA] << 16) | s->regs[SONIC_CRDA], (uint8_t *)data, size, 1);
+    address_space_rw(&s->as, (s->regs[SONIC_URDA] << 16) | s->regs[SONIC_CRDA],
+        MEMTXATTRS_UNSPECIFIED, (uint8_t *)data, size, 1);
 
     /* Move to next descriptor */
     size = sizeof(uint16_t) * width;
-    s->memory_rw(s->mem_opaque,
+    address_space_rw(&s->as,
         ((s->regs[SONIC_URDA] << 16) | s->regs[SONIC_CRDA]) + sizeof(uint16_t) * 5 * width,
-        (uint8_t *)data, size, 0);
+        MEMTXATTRS_UNSPECIFIED, (uint8_t *)data, size, 0);
     s->regs[SONIC_LLFA] = data[0 * width];
     if (s->regs[SONIC_LLFA] & 0x1) {
         /* EOL detected */
         s->regs[SONIC_ISR] |= SONIC_ISR_RDE;
     } else {
         data[0 * width] = 0; /* in_use */
-        s->memory_rw(s->mem_opaque,
+        address_space_rw(&s->as,
             ((s->regs[SONIC_URDA] << 16) | s->regs[SONIC_CRDA]) + sizeof(uint16_t) * 6 * width,
-            (uint8_t *)data, size, 1);
+            MEMTXATTRS_UNSPECIFIED, (uint8_t *)data, sizeof(uint16_t), 1);
         s->regs[SONIC_CRDA] = s->regs[SONIC_LLFA];
         s->regs[SONIC_ISR] |= SONIC_ISR_PKTRX;
         s->regs[SONIC_RSC] = (s->regs[SONIC_RSC] & 0xff00) | (((s->regs[SONIC_RSC] & 0x00ff) + 1) & 0x00ff);
 
         if (s->regs[SONIC_RCR] & SONIC_RCR_LPKT) {
             /* Read next RRA */
-            do_read_rra(s);
+            dp8393x_do_read_rra(s);
         }
     }
 
@@ -835,11 +781,12 @@
     return size;
 }
 
-static void nic_reset(void *opaque)
+static void dp8393x_reset(DeviceState *dev)
 {
-    dp8393xState *s = opaque;
+    dp8393xState *s = DP8393X(dev);
     timer_del(s->watchdog);
 
+    memset(s->regs, 0, sizeof(s->regs));
     s->regs[SONIC_CR] = SONIC_CR_RST | SONIC_CR_STP | SONIC_CR_RXDIS;
     s->regs[SONIC_DCR] &= ~(SONIC_DCR_EXBUS | SONIC_DCR_LBR);
     s->regs[SONIC_RCR] &= ~(SONIC_RCR_LB0 | SONIC_RCR_LB1 | SONIC_RCR_BRD | SONIC_RCR_RNT);
@@ -862,39 +809,91 @@
 static NetClientInfo net_dp83932_info = {
     .type = NET_CLIENT_OPTIONS_KIND_NIC,
     .size = sizeof(NICState),
-    .can_receive = nic_can_receive,
-    .receive = nic_receive,
+    .can_receive = dp8393x_can_receive,
+    .receive = dp8393x_receive,
 };
 
-void dp83932_init(NICInfo *nd, hwaddr base, int it_shift,
-                  MemoryRegion *address_space,
-                  qemu_irq irq, void* mem_opaque,
-                  void (*memory_rw)(void *opaque, hwaddr addr, uint8_t *buf, int len, int is_write))
+static void dp8393x_instance_init(Object *obj)
 {
-    dp8393xState *s;
+    SysBusDevice *sbd = SYS_BUS_DEVICE(obj);
+    dp8393xState *s = DP8393X(obj);
 
-    qemu_check_nic_model(nd, "dp83932");
+    sysbus_init_mmio(sbd, &s->mmio);
+    sysbus_init_mmio(sbd, &s->prom);
+    sysbus_init_irq(sbd, &s->irq);
+}
 
-    s = g_malloc0(sizeof(dp8393xState));
+static void dp8393x_realize(DeviceState *dev, Error **errp)
+{
+    dp8393xState *s = DP8393X(dev);
+    int i, checksum;
+    uint8_t *prom;
 
-    s->address_space = address_space;
-    s->mem_opaque = mem_opaque;
-    s->memory_rw = memory_rw;
-    s->it_shift = it_shift;
-    s->irq = irq;
+    address_space_init(&s->as, s->dma_mr, "dp8393x");
+    memory_region_init_io(&s->mmio, OBJECT(dev), &dp8393x_ops, s,
+                          "dp8393x-regs", 0x40 << s->it_shift);
+
+    s->nic = qemu_new_nic(&net_dp83932_info, &s->conf,
+                          object_get_typename(OBJECT(dev)), dev->id, s);
+    qemu_format_nic_info_str(qemu_get_queue(s->nic), s->conf.macaddr.a);
+
     s->watchdog = timer_new_ns(QEMU_CLOCK_VIRTUAL, dp8393x_watchdog, s);
     s->regs[SONIC_SR] = 0x0004; /* only revision recognized by Linux */
 
-    s->conf.macaddr = nd->macaddr;
-    s->conf.peers.ncs[0] = nd->netdev;
-
-    s->nic = qemu_new_nic(&net_dp83932_info, &s->conf, nd->model, nd->name, s);
-
-    qemu_format_nic_info_str(qemu_get_queue(s->nic), s->conf.macaddr.a);
-    qemu_register_reset(nic_reset, s);
-    nic_reset(s);
-
-    memory_region_init_io(&s->mmio, NULL, &dp8393x_ops, s,
-                          "dp8393x", 0x40 << it_shift);
-    memory_region_add_subregion(address_space, base, &s->mmio);
+    memory_region_init_rom_device(&s->prom, OBJECT(dev), NULL, NULL,
+                                  "dp8393x-prom", SONIC_PROM_SIZE, NULL);
+    prom = memory_region_get_ram_ptr(&s->prom);
+    checksum = 0;
+    for (i = 0; i < 6; i++) {
+        prom[i] = s->conf.macaddr.a[i];
+        checksum += prom[i];
+        if (checksum > 0xff) {
+            checksum = (checksum + 1) & 0xff;
+        }
+    }
+    prom[7] = 0xff - checksum;
 }
+
+static const VMStateDescription vmstate_dp8393x = {
+    .name = "dp8393x",
+    .version_id = 0,
+    .minimum_version_id = 0,
+    .fields = (VMStateField []) {
+        VMSTATE_BUFFER_UNSAFE(cam, dp8393xState, 0, 16 * 6),
+        VMSTATE_UINT16_ARRAY(regs, dp8393xState, 0x40),
+        VMSTATE_END_OF_LIST()
+    }
+};
+
+static Property dp8393x_properties[] = {
+    DEFINE_NIC_PROPERTIES(dp8393xState, conf),
+    DEFINE_PROP_PTR("dma_mr", dp8393xState, dma_mr),
+    DEFINE_PROP_UINT8("it_shift", dp8393xState, it_shift, 0),
+    DEFINE_PROP_END_OF_LIST(),
+};
+
+static void dp8393x_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+
+    set_bit(DEVICE_CATEGORY_NETWORK, dc->categories);
+    dc->realize = dp8393x_realize;
+    dc->reset = dp8393x_reset;
+    dc->vmsd = &vmstate_dp8393x;
+    dc->props = dp8393x_properties;
+}
+
+static const TypeInfo dp8393x_info = {
+    .name          = TYPE_DP8393X,
+    .parent        = TYPE_SYS_BUS_DEVICE,
+    .instance_size = sizeof(dp8393xState),
+    .instance_init = dp8393x_instance_init,
+    .class_init    = dp8393x_class_init,
+};
+
+static void dp8393x_register_types(void)
+{
+    type_register_static(&dp8393x_info);
+}
+
+type_init(dp8393x_register_types)
diff --git a/hw/net/e1000.c b/hw/net/e1000.c
index 091d61a..bab8e2a 100644
--- a/hw/net/e1000.c
+++ b/hw/net/e1000.c
@@ -1370,6 +1370,7 @@
     .name = "e1000/mit_state",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = e1000_mit_state_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT32(mac_reg[RDTR], E1000State),
         VMSTATE_UINT32(mac_reg[RADV], E1000State),
@@ -1457,13 +1458,9 @@
         VMSTATE_UINT32_SUB_ARRAY(mac_reg, E1000State, VFTA, 128),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_e1000_mit_state,
-            .needed = e1000_mit_state_needed,
-        }, {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_e1000_mit_state,
+        NULL
     }
 };
 
diff --git a/hw/net/pcnet.c b/hw/net/pcnet.c
index bdfd38f..68b9981 100644
--- a/hw/net/pcnet.c
+++ b/hw/net/pcnet.c
@@ -1241,6 +1241,14 @@
         }
 
         bcnt = 4096 - GET_FIELD(tmd.length, TMDL, BCNT);
+
+        /* if multi-tmd packet outsizes s->buffer then skip it silently.
+           Note: this is not what real hw does */
+        if (s->xmit_pos + bcnt > sizeof(s->buffer)) {
+            s->xmit_pos = -1;
+            goto txdone;
+        }
+
         s->phys_mem_read(s->dma_opaque, PHYSADDR(s, tmd.tbadr),
                          s->buffer + s->xmit_pos, bcnt, CSR_BSWP(s));
         s->xmit_pos += bcnt;
diff --git a/hw/net/rtl8139.c b/hw/net/rtl8139.c
index f868108..e0db472 100644
--- a/hw/net/rtl8139.c
+++ b/hw/net/rtl8139.c
@@ -3240,6 +3240,7 @@
     .name = "rtl8139/hotplug_ready",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = rtl8139_hotplug_ready_needed,
     .fields = (VMStateField[]) {
         VMSTATE_END_OF_LIST()
     }
@@ -3335,13 +3336,9 @@
         VMSTATE_UINT32_V(cplus_enabled, RTL8139State, 4),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection []) {
-        {
-            .vmsd = &vmstate_rtl8139_hotplug_ready,
-            .needed = rtl8139_hotplug_ready_needed,
-        }, {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_rtl8139_hotplug_ready,
+        NULL
     }
 };
 
diff --git a/hw/net/vhost_net.c b/hw/net/vhost_net.c
index 426b23e..1c55517 100644
--- a/hw/net/vhost_net.c
+++ b/hw/net/vhost_net.c
@@ -52,6 +52,7 @@
     VIRTIO_RING_F_INDIRECT_DESC,
     VIRTIO_RING_F_EVENT_IDX,
     VIRTIO_NET_F_MRG_RXBUF,
+    VIRTIO_F_VERSION_1,
     VHOST_INVALID_FEATURE_BIT
 };
 
@@ -62,6 +63,7 @@
     VIRTIO_RING_F_EVENT_IDX,
 
     VIRTIO_F_ANY_LAYOUT,
+    VIRTIO_F_VERSION_1,
     VIRTIO_NET_F_CSUM,
     VIRTIO_NET_F_GUEST_CSUM,
     VIRTIO_NET_F_GSO,
@@ -107,13 +109,13 @@
     return feature_bits;
 }
 
-unsigned vhost_net_get_features(struct vhost_net *net, unsigned features)
+uint64_t vhost_net_get_features(struct vhost_net *net, uint64_t features)
 {
     return vhost_get_features(&net->dev, vhost_net_get_feature_bits(net),
             features);
 }
 
-void vhost_net_ack_features(struct vhost_net *net, unsigned features)
+void vhost_net_ack_features(struct vhost_net *net, uint64_t features)
 {
     net->dev.acked_features = net->dev.backend_features;
     vhost_ack_features(&net->dev, vhost_net_get_feature_bits(net), features);
@@ -147,7 +149,7 @@
             goto fail;
         }
         net->dev.backend_features = qemu_has_vnet_hdr(options->net_backend)
-            ? 0 : (1 << VHOST_NET_F_VIRTIO_NET_HDR);
+            ? 0 : (1ULL << VHOST_NET_F_VIRTIO_NET_HDR);
         net->backend = r;
     } else {
         net->dev.backend_features = 0;
@@ -167,7 +169,7 @@
     if (backend_kernel) {
         if (!qemu_has_vnet_hdr_len(options->net_backend,
                                sizeof(struct virtio_net_hdr_mrg_rxbuf))) {
-            net->dev.features &= ~(1 << VIRTIO_NET_F_MRG_RXBUF);
+            net->dev.features &= ~(1ULL << VIRTIO_NET_F_MRG_RXBUF);
         }
         if (~net->dev.features & net->dev.backend_features) {
             fprintf(stderr, "vhost lacks feature mask %" PRIu64
@@ -431,11 +433,11 @@
 {
 }
 
-unsigned vhost_net_get_features(struct vhost_net *net, unsigned features)
+uint64_t vhost_net_get_features(struct vhost_net *net, uint64_t features)
 {
     return features;
 }
-void vhost_net_ack_features(struct vhost_net *net, unsigned features)
+void vhost_net_ack_features(struct vhost_net *net, uint64_t features)
 {
 }
 
diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
index 0d3bf0f..9281aa1 100644
--- a/hw/net/virtio-net.c
+++ b/hw/net/virtio-net.c
@@ -87,6 +87,7 @@
     memcpy(&netcfg, config, n->config_size);
 
     if (!virtio_has_feature(vdev, VIRTIO_NET_F_CTRL_MAC_ADDR) &&
+        !virtio_has_feature(vdev, VIRTIO_F_VERSION_1) &&
         memcmp(netcfg.mac, n->mac, ETH_ALEN)) {
         memcpy(n->mac, netcfg.mac, ETH_ALEN);
         qemu_format_nic_info_str(qemu_get_queue(n->nic), n->mac);
@@ -366,15 +367,21 @@
     return n->has_ufo;
 }
 
-static void virtio_net_set_mrg_rx_bufs(VirtIONet *n, int mergeable_rx_bufs)
+static void virtio_net_set_mrg_rx_bufs(VirtIONet *n, int mergeable_rx_bufs,
+                                       int version_1)
 {
     int i;
     NetClientState *nc;
 
     n->mergeable_rx_bufs = mergeable_rx_bufs;
 
-    n->guest_hdr_len = n->mergeable_rx_bufs ?
-        sizeof(struct virtio_net_hdr_mrg_rxbuf) : sizeof(struct virtio_net_hdr);
+    if (version_1) {
+        n->guest_hdr_len = sizeof(struct virtio_net_hdr_mrg_rxbuf);
+    } else {
+        n->guest_hdr_len = n->mergeable_rx_bufs ?
+            sizeof(struct virtio_net_hdr_mrg_rxbuf) :
+            sizeof(struct virtio_net_hdr);
+    }
 
     for (i = 0; i < n->max_queues; i++) {
         nc = qemu_get_subqueue(n->nic, i);
@@ -463,6 +470,7 @@
     }
 
     if (!get_vhost_net(nc->peer)) {
+        virtio_add_feature(&features, VIRTIO_F_VERSION_1);
         return features;
     }
     return vhost_net_get_features(get_vhost_net(nc->peer), features);
@@ -521,7 +529,9 @@
 
     virtio_net_set_mrg_rx_bufs(n,
                                __virtio_has_feature(features,
-                                                    VIRTIO_NET_F_MRG_RXBUF));
+                                                    VIRTIO_NET_F_MRG_RXBUF),
+                               __virtio_has_feature(features,
+                                                    VIRTIO_F_VERSION_1));
 
     if (n->has_vnet_hdr) {
         n->curr_guest_offloads =
@@ -1374,7 +1384,8 @@
     qemu_get_buffer(f, n->mac, ETH_ALEN);
     n->vqs[0].tx_waiting = qemu_get_be32(f);
 
-    virtio_net_set_mrg_rx_bufs(n, qemu_get_be32(f));
+    virtio_net_set_mrg_rx_bufs(n, qemu_get_be32(f),
+                               virtio_has_feature(vdev, VIRTIO_F_VERSION_1));
 
     if (version_id >= 3)
         n->status = qemu_get_be16(f);
@@ -1626,7 +1637,7 @@
 
     n->vqs[0].tx_waiting = 0;
     n->tx_burst = n->net_conf.txburst;
-    virtio_net_set_mrg_rx_bufs(n, 0);
+    virtio_net_set_mrg_rx_bufs(n, 0, 0);
     n->promisc = 1; /* for compatibility */
 
     n->mac_table.macs = g_malloc0(MAC_TABLE_ENTRIES * ETH_ALEN);
@@ -1696,10 +1707,50 @@
 }
 
 static Property virtio_net_properties[] = {
-    DEFINE_VIRTIO_NET_FEATURES(VirtIONet, host_features),
+    DEFINE_PROP_BIT("any_layout", VirtIONet, host_features,
+                    VIRTIO_F_ANY_LAYOUT, true),
+    DEFINE_PROP_BIT("csum", VirtIONet, host_features, VIRTIO_NET_F_CSUM, true),
+    DEFINE_PROP_BIT("guest_csum", VirtIONet, host_features,
+                    VIRTIO_NET_F_GUEST_CSUM, true),
+    DEFINE_PROP_BIT("gso", VirtIONet, host_features, VIRTIO_NET_F_GSO, true),
+    DEFINE_PROP_BIT("guest_tso4", VirtIONet, host_features,
+                    VIRTIO_NET_F_GUEST_TSO4, true),
+    DEFINE_PROP_BIT("guest_tso6", VirtIONet, host_features,
+                    VIRTIO_NET_F_GUEST_TSO6, true),
+    DEFINE_PROP_BIT("guest_ecn", VirtIONet, host_features,
+                    VIRTIO_NET_F_GUEST_ECN, true),
+    DEFINE_PROP_BIT("guest_ufo", VirtIONet, host_features,
+                    VIRTIO_NET_F_GUEST_UFO, true),
+    DEFINE_PROP_BIT("guest_announce", VirtIONet, host_features,
+                    VIRTIO_NET_F_GUEST_ANNOUNCE, true),
+    DEFINE_PROP_BIT("host_tso4", VirtIONet, host_features,
+                    VIRTIO_NET_F_HOST_TSO4, true),
+    DEFINE_PROP_BIT("host_tso6", VirtIONet, host_features,
+                    VIRTIO_NET_F_HOST_TSO6, true),
+    DEFINE_PROP_BIT("host_ecn", VirtIONet, host_features,
+                    VIRTIO_NET_F_HOST_ECN, true),
+    DEFINE_PROP_BIT("host_ufo", VirtIONet, host_features,
+                    VIRTIO_NET_F_HOST_UFO, true),
+    DEFINE_PROP_BIT("mrg_rxbuf", VirtIONet, host_features,
+                    VIRTIO_NET_F_MRG_RXBUF, true),
+    DEFINE_PROP_BIT("status", VirtIONet, host_features,
+                    VIRTIO_NET_F_STATUS, true),
+    DEFINE_PROP_BIT("ctrl_vq", VirtIONet, host_features,
+                    VIRTIO_NET_F_CTRL_VQ, true),
+    DEFINE_PROP_BIT("ctrl_rx", VirtIONet, host_features,
+                    VIRTIO_NET_F_CTRL_RX, true),
+    DEFINE_PROP_BIT("ctrl_vlan", VirtIONet, host_features,
+                    VIRTIO_NET_F_CTRL_VLAN, true),
+    DEFINE_PROP_BIT("ctrl_rx_extra", VirtIONet, host_features,
+                    VIRTIO_NET_F_CTRL_RX_EXTRA, true),
+    DEFINE_PROP_BIT("ctrl_mac_addr", VirtIONet, host_features,
+                    VIRTIO_NET_F_CTRL_MAC_ADDR, true),
+    DEFINE_PROP_BIT("ctrl_guest_offloads", VirtIONet, host_features,
+                    VIRTIO_NET_F_CTRL_GUEST_OFFLOADS, true),
+    DEFINE_PROP_BIT("mq", VirtIONet, host_features, VIRTIO_NET_F_MQ, false),
     DEFINE_NIC_PROPERTIES(VirtIONet, nic_conf),
     DEFINE_PROP_UINT32("x-txtimer", VirtIONet, net_conf.txtimer,
-                                               TX_TIMER_INTERVAL),
+                       TX_TIMER_INTERVAL),
     DEFINE_PROP_INT32("x-txburst", VirtIONet, net_conf.txburst, TX_BURST),
     DEFINE_PROP_STRING("tx", VirtIONet, net_conf.tx),
     DEFINE_PROP_END_OF_LIST(),
diff --git a/hw/net/vmxnet3.c b/hw/net/vmxnet3.c
index dfb328d..8bcdf3e 100644
--- a/hw/net/vmxnet3.c
+++ b/hw/net/vmxnet3.c
@@ -2226,6 +2226,7 @@
     .version_id = 1,
     .minimum_version_id = 1,
     .pre_load = vmxnet3_mcast_list_pre_load,
+    .needed = vmxnet3_mc_list_needed,
     .fields = (VMStateField[]) {
         VMSTATE_VBUFFER_UINT32(mcast_list, VMXNET3State, 0, NULL, 0,
             mcast_list_buff_size),
@@ -2470,14 +2471,9 @@
 
             VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmxstate_vmxnet3_mcast_list,
-            .needed = vmxnet3_mc_list_needed
-        },
-        {
-            /* empty element. */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmxstate_vmxnet3_mcast_list,
+        NULL
     }
 };
 
diff --git a/hw/nvram/fw_cfg.c b/hw/nvram/fw_cfg.c
index 68eff77..88481b7 100644
--- a/hw/nvram/fw_cfg.c
+++ b/hw/nvram/fw_cfg.c
@@ -46,7 +46,6 @@
     uint32_t len;
     uint8_t *data;
     void *callback_opaque;
-    FWCfgCallback callback;
     FWCfgReadCallback read_callback;
 } FWCfgEntry;
 
@@ -232,19 +231,7 @@
 
 static void fw_cfg_write(FWCfgState *s, uint8_t value)
 {
-    int arch = !!(s->cur_entry & FW_CFG_ARCH_LOCAL);
-    FWCfgEntry *e = &s->entries[arch][s->cur_entry & FW_CFG_ENTRY_MASK];
-
-    trace_fw_cfg_write(s, value);
-
-    if (s->cur_entry & FW_CFG_WRITE_CHANNEL && e->callback &&
-        s->cur_offset < e->len) {
-        e->data[s->cur_offset++] = value;
-        if (s->cur_offset == e->len) {
-            e->callback(e->callback_opaque, e->data);
-            s->cur_offset = 0;
-        }
-    }
+    /* nothing, write support removed in QEMU v2.4+ */
 }
 
 static int fw_cfg_select(FWCfgState *s, uint16_t key)
@@ -436,6 +423,7 @@
     key &= FW_CFG_ENTRY_MASK;
 
     assert(key < FW_CFG_MAX_ENTRY && len < UINT32_MAX);
+    assert(s->entries[arch][key].data == NULL); /* avoid key conflict */
 
     s->entries[arch][key].data = data;
     s->entries[arch][key].len = (uint32_t)len;
@@ -458,7 +446,6 @@
     s->entries[arch][key].data = data;
     s->entries[arch][key].len = len;
     s->entries[arch][key].callback_opaque = NULL;
-    s->entries[arch][key].callback = NULL;
 
     return ptr;
 }
@@ -484,6 +471,16 @@
     fw_cfg_add_bytes(s, key, copy, sizeof(value));
 }
 
+void fw_cfg_modify_i16(FWCfgState *s, uint16_t key, uint16_t value)
+{
+    uint16_t *copy, *old;
+
+    copy = g_malloc(sizeof(value));
+    *copy = cpu_to_le16(value);
+    old = fw_cfg_modify_bytes_read(s, key, copy, sizeof(value));
+    g_free(old);
+}
+
 void fw_cfg_add_i32(FWCfgState *s, uint16_t key, uint32_t value)
 {
     uint32_t *copy;
@@ -502,23 +499,6 @@
     fw_cfg_add_bytes(s, key, copy, sizeof(value));
 }
 
-void fw_cfg_add_callback(FWCfgState *s, uint16_t key, FWCfgCallback callback,
-                         void *callback_opaque, void *data, size_t len)
-{
-    int arch = !!(key & FW_CFG_ARCH_LOCAL);
-
-    assert(key & FW_CFG_WRITE_CHANNEL);
-
-    key &= FW_CFG_ENTRY_MASK;
-
-    assert(key < FW_CFG_MAX_ENTRY && len <= UINT32_MAX);
-
-    s->entries[arch][key].data = data;
-    s->entries[arch][key].len = (uint32_t)len;
-    s->entries[arch][key].callback_opaque = callback_opaque;
-    s->entries[arch][key].callback = callback;
-}
-
 void fw_cfg_add_file_callback(FWCfgState *s,  const char *filename,
                               FWCfgReadCallback callback, void *callback_opaque,
                               void *data, size_t len)
@@ -535,18 +515,19 @@
     index = be32_to_cpu(s->files->count);
     assert(index < FW_CFG_FILE_SLOTS);
 
-    fw_cfg_add_bytes_read_callback(s, FW_CFG_FILE_FIRST + index,
-                                   callback, callback_opaque, data, len);
-
     pstrcpy(s->files->f[index].name, sizeof(s->files->f[index].name),
             filename);
     for (i = 0; i < index; i++) {
         if (strcmp(s->files->f[index].name, s->files->f[i].name) == 0) {
-            trace_fw_cfg_add_file_dupe(s, s->files->f[index].name);
-            return;
+            error_report("duplicate fw_cfg file name: %s",
+                         s->files->f[index].name);
+            exit(1);
         }
     }
 
+    fw_cfg_add_bytes_read_callback(s, FW_CFG_FILE_FIRST + index,
+                                   callback, callback_opaque, data, len);
+
     s->files->f[index].size   = cpu_to_be32(len);
     s->files->f[index].select = cpu_to_be16(FW_CFG_FILE_FIRST + index);
     trace_fw_cfg_add_file(s, index, s->files->f[index].name, len);
diff --git a/hw/pci-host/piix.c b/hw/pci-host/piix.c
index f1712b8..ed2424c 100644
--- a/hw/pci-host/piix.c
+++ b/hw/pci-host/piix.c
@@ -582,6 +582,7 @@
     .name = "PIIX3/rcr",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = piix3_rcr_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT8(rcr, PIIX3State),
         VMSTATE_END_OF_LIST()
@@ -600,12 +601,9 @@
                               PIIX_NUM_PIRQS, 3),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_piix3_rcr,
-            .needed = piix3_rcr_needed,
-        },
-        { 0 }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_piix3_rcr,
+        NULL
     }
 };
 
diff --git a/hw/ppc/mac_newworld.c b/hw/ppc/mac_newworld.c
index a365bf9..0f3e341 100644
--- a/hw/ppc/mac_newworld.c
+++ b/hw/ppc/mac_newworld.c
@@ -119,7 +119,7 @@
 static void fw_cfg_boot_set(void *opaque, const char *boot_device,
                             Error **errp)
 {
-    fw_cfg_add_i16(opaque, FW_CFG_BOOT_DEVICE, boot_device[0]);
+    fw_cfg_modify_i16(opaque, FW_CFG_BOOT_DEVICE, boot_device[0]);
 }
 
 static uint64_t translate_kernel_address(void *opaque, uint64_t addr)
diff --git a/hw/ppc/mac_oldworld.c b/hw/ppc/mac_oldworld.c
index f26133d..99879dd 100644
--- a/hw/ppc/mac_oldworld.c
+++ b/hw/ppc/mac_oldworld.c
@@ -52,7 +52,7 @@
 static void fw_cfg_boot_set(void *opaque, const char *boot_device,
                             Error **errp)
 {
-    fw_cfg_add_i16(opaque, FW_CFG_BOOT_DEVICE, boot_device[0]);
+    fw_cfg_modify_i16(opaque, FW_CFG_BOOT_DEVICE, boot_device[0]);
 }
 
 static uint64_t translate_kernel_address(void *opaque, uint64_t addr)
diff --git a/hw/s390x/virtio-ccw.c b/hw/s390x/virtio-ccw.c
index ab5fc7f..b7a88d6 100644
--- a/hw/s390x/virtio-ccw.c
+++ b/hw/s390x/virtio-ccw.c
@@ -498,15 +498,19 @@
             if (!(status & VIRTIO_CONFIG_S_DRIVER_OK)) {
                 virtio_ccw_stop_ioeventfd(dev);
             }
-            virtio_set_status(vdev, status);
-            if (vdev->status == 0) {
-                virtio_reset(vdev);
+            if (virtio_set_status(vdev, status) == 0) {
+                if (vdev->status == 0) {
+                    virtio_reset(vdev);
+                }
+                if (status & VIRTIO_CONFIG_S_DRIVER_OK) {
+                    virtio_ccw_start_ioeventfd(dev);
+                }
+                sch->curr_status.scsw.count = ccw.count - sizeof(status);
+                ret = 0;
+            } else {
+                /* Trigger a command reject. */
+                ret = -ENOSYS;
             }
-            if (status & VIRTIO_CONFIG_S_DRIVER_OK) {
-                virtio_ccw_start_ioeventfd(dev);
-            }
-            sch->curr_status.scsw.count = ccw.count - sizeof(status);
-            ret = 0;
         }
         break;
     case CCW_CMD_SET_IND:
diff --git a/hw/scsi/scsi-bus.c b/hw/scsi/scsi-bus.c
index bd2c0e4..f50b2f0 100644
--- a/hw/scsi/scsi-bus.c
+++ b/hw/scsi/scsi-bus.c
@@ -1968,6 +1968,7 @@
     .name = "SCSIDevice/sense",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = scsi_sense_state_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT8_SUB_ARRAY(sense, SCSIDevice,
                                 SCSI_SENSE_BUF_SIZE_OLD,
@@ -1998,13 +1999,9 @@
         },
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection []) {
-        {
-            .vmsd = &vmstate_scsi_sense_state,
-            .needed = scsi_sense_state_needed,
-        }, {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_scsi_sense_state,
+        NULL
     }
 };
 
diff --git a/hw/scsi/vhost-scsi.c b/hw/scsi/vhost-scsi.c
index 9c76486..1941aa1 100644
--- a/hw/scsi/vhost-scsi.c
+++ b/hw/scsi/vhost-scsi.c
@@ -294,7 +294,14 @@
 }
 
 static Property vhost_scsi_properties[] = {
-    DEFINE_VHOST_SCSI_PROPERTIES(VHostSCSI, parent_obj.conf),
+    DEFINE_PROP_STRING("vhostfd", VHostSCSI, parent_obj.conf.vhostfd),
+    DEFINE_PROP_STRING("wwpn", VHostSCSI, parent_obj.conf.wwpn),
+    DEFINE_PROP_UINT32("boot_tpgt", VHostSCSI, parent_obj.conf.boot_tpgt, 0),
+    DEFINE_PROP_UINT32("num_queues", VHostSCSI, parent_obj.conf.num_queues, 1),
+    DEFINE_PROP_UINT32("max_sectors", VHostSCSI, parent_obj.conf.max_sectors,
+                                                 0xFFFF),
+    DEFINE_PROP_UINT32("cmd_per_lun", VHostSCSI, parent_obj.conf.cmd_per_lun,
+                                                 128),
     DEFINE_PROP_END_OF_LIST(),
 };
 
diff --git a/hw/scsi/virtio-scsi.c b/hw/scsi/virtio-scsi.c
index b0dee29..f7d3c7c 100644
--- a/hw/scsi/virtio-scsi.c
+++ b/hw/scsi/virtio-scsi.c
@@ -948,8 +948,17 @@
 }
 
 static Property virtio_scsi_properties[] = {
-    DEFINE_VIRTIO_SCSI_PROPERTIES(VirtIOSCSI, parent_obj.conf),
-    DEFINE_VIRTIO_SCSI_FEATURES(VirtIOSCSI, host_features),
+    DEFINE_PROP_UINT32("num_queues", VirtIOSCSI, parent_obj.conf.num_queues, 1),
+    DEFINE_PROP_UINT32("max_sectors", VirtIOSCSI, parent_obj.conf.max_sectors,
+                                                  0xFFFF),
+    DEFINE_PROP_UINT32("cmd_per_lun", VirtIOSCSI, parent_obj.conf.cmd_per_lun,
+                                                  128),
+    DEFINE_PROP_BIT("any_layout", VirtIOSCSI, host_features,
+                                              VIRTIO_F_ANY_LAYOUT, true),
+    DEFINE_PROP_BIT("hotplug", VirtIOSCSI, host_features,
+                                           VIRTIO_SCSI_F_HOTPLUG, true),
+    DEFINE_PROP_BIT("param_change", VirtIOSCSI, host_features,
+                                                VIRTIO_SCSI_F_CHANGE, true),
     DEFINE_PROP_END_OF_LIST(),
 };
 
diff --git a/hw/sparc/sun4m.c b/hw/sparc/sun4m.c
index 8a3599c..68ac4d8 100644
--- a/hw/sparc/sun4m.c
+++ b/hw/sparc/sun4m.c
@@ -124,7 +124,7 @@
 static void fw_cfg_boot_set(void *opaque, const char *boot_device,
                             Error **errp)
 {
-    fw_cfg_add_i16(opaque, FW_CFG_BOOT_DEVICE, boot_device[0]);
+    fw_cfg_modify_i16(opaque, FW_CFG_BOOT_DEVICE, boot_device[0]);
 }
 
 static void nvram_init(Nvram *nvram, uint8_t *macaddr,
diff --git a/hw/sparc64/sun4u.c b/hw/sparc64/sun4u.c
index 6f34e87..30cfa0e 100644
--- a/hw/sparc64/sun4u.c
+++ b/hw/sparc64/sun4u.c
@@ -127,7 +127,7 @@
 static void fw_cfg_boot_set(void *opaque, const char *boot_device,
                             Error **errp)
 {
-    fw_cfg_add_i16(opaque, FW_CFG_BOOT_DEVICE, boot_device[0]);
+    fw_cfg_modify_i16(opaque, FW_CFG_BOOT_DEVICE, boot_device[0]);
 }
 
 static int sun4u_NVRAM_set_params(Nvram *nvram, uint16_t NVRAM_size,
diff --git a/hw/timer/hpet.c b/hw/timer/hpet.c
index b6b8a20..b50071e 100644
--- a/hw/timer/hpet.c
+++ b/hw/timer/hpet.c
@@ -283,6 +283,7 @@
     .name = "hpet/rtc_irq_level",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = hpet_rtc_irq_level_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT8(rtc_irq_level, HPETState),
         VMSTATE_END_OF_LIST()
@@ -322,13 +323,9 @@
                                     vmstate_hpet_timer, HPETTimer),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_hpet_rtc_irq_level,
-            .needed = hpet_rtc_irq_level_needed,
-        }, {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_hpet_rtc_irq_level,
+        NULL
     }
 };
 
diff --git a/hw/timer/mc146818rtc.c b/hw/timer/mc146818rtc.c
index f2b77fa..3204825 100644
--- a/hw/timer/mc146818rtc.c
+++ b/hw/timer/mc146818rtc.c
@@ -733,22 +733,23 @@
     return 0;
 }
 
-static const VMStateDescription vmstate_rtc_irq_reinject_on_ack_count = {
-    .name = "mc146818rtc/irq_reinject_on_ack_count",
-    .version_id = 1,
-    .minimum_version_id = 1,
-    .fields = (VMStateField[]) {
-        VMSTATE_UINT16(irq_reinject_on_ack_count, RTCState),
-        VMSTATE_END_OF_LIST()
-    }
-};
-
 static bool rtc_irq_reinject_on_ack_count_needed(void *opaque)
 {
     RTCState *s = (RTCState *)opaque;
     return s->irq_reinject_on_ack_count != 0;
 }
 
+static const VMStateDescription vmstate_rtc_irq_reinject_on_ack_count = {
+    .name = "mc146818rtc/irq_reinject_on_ack_count",
+    .version_id = 1,
+    .minimum_version_id = 1,
+    .needed = rtc_irq_reinject_on_ack_count_needed,
+    .fields = (VMStateField[]) {
+        VMSTATE_UINT16(irq_reinject_on_ack_count, RTCState),
+        VMSTATE_END_OF_LIST()
+    }
+};
+
 static const VMStateDescription vmstate_rtc = {
     .name = "mc146818rtc",
     .version_id = 3,
@@ -770,13 +771,9 @@
         VMSTATE_UINT64_V(next_alarm_time, RTCState, 3),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_rtc_irq_reinject_on_ack_count,
-            .needed = rtc_irq_reinject_on_ack_count_needed,
-        }, {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_rtc_irq_reinject_on_ack_count,
+        NULL
     }
 };
 
diff --git a/hw/usb/hcd-ohci.c b/hw/usb/hcd-ohci.c
index 1a22c9c..7d65818 100644
--- a/hw/usb/hcd-ohci.c
+++ b/hw/usb/hcd-ohci.c
@@ -2034,6 +2034,7 @@
     .version_id = 1,
     .minimum_version_id = 1,
     .pre_load = ohci_eof_timer_pre_load,
+    .needed = ohci_eof_timer_needed,
     .fields = (VMStateField[]) {
         VMSTATE_TIMER_PTR(eof_timer, OHCIState),
         VMSTATE_END_OF_LIST()
@@ -2081,13 +2082,9 @@
         VMSTATE_BOOL(async_complete, OHCIState),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection []) {
-        {
-            .vmsd = &vmstate_ohci_eof_timer,
-            .needed = ohci_eof_timer_needed,
-        } , {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_ohci_eof_timer,
+        NULL
     }
 };
 
diff --git a/hw/usb/redirect.c b/hw/usb/redirect.c
index 242a654..6b4218c 100644
--- a/hw/usb/redirect.c
+++ b/hw/usb/redirect.c
@@ -2257,16 +2257,6 @@
 
 
 /* For endp_data migration */
-static const VMStateDescription usbredir_bulk_receiving_vmstate = {
-    .name = "usb-redir-ep/bulk-receiving",
-    .version_id = 1,
-    .minimum_version_id = 1,
-    .fields = (VMStateField[]) {
-        VMSTATE_UINT8(bulk_receiving_started, struct endp_data),
-        VMSTATE_END_OF_LIST()
-    }
-};
-
 static bool usbredir_bulk_receiving_needed(void *priv)
 {
     struct endp_data *endp = priv;
@@ -2274,12 +2264,13 @@
     return endp->bulk_receiving_started;
 }
 
-static const VMStateDescription usbredir_stream_vmstate = {
-    .name = "usb-redir-ep/stream-state",
+static const VMStateDescription usbredir_bulk_receiving_vmstate = {
+    .name = "usb-redir-ep/bulk-receiving",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = usbredir_bulk_receiving_needed,
     .fields = (VMStateField[]) {
-        VMSTATE_UINT32(max_streams, struct endp_data),
+        VMSTATE_UINT8(bulk_receiving_started, struct endp_data),
         VMSTATE_END_OF_LIST()
     }
 };
@@ -2291,6 +2282,17 @@
     return endp->max_streams;
 }
 
+static const VMStateDescription usbredir_stream_vmstate = {
+    .name = "usb-redir-ep/stream-state",
+    .version_id = 1,
+    .minimum_version_id = 1,
+    .needed = usbredir_stream_needed,
+    .fields = (VMStateField[]) {
+        VMSTATE_UINT32(max_streams, struct endp_data),
+        VMSTATE_END_OF_LIST()
+    }
+};
+
 static const VMStateDescription usbredir_ep_vmstate = {
     .name = "usb-redir-ep",
     .version_id = 1,
@@ -2318,16 +2320,10 @@
         VMSTATE_INT32(bufpq_target_size, struct endp_data),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &usbredir_bulk_receiving_vmstate,
-            .needed = usbredir_bulk_receiving_needed,
-        }, {
-            .vmsd = &usbredir_stream_vmstate,
-            .needed = usbredir_stream_needed,
-        }, {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &usbredir_bulk_receiving_vmstate,
+        &usbredir_stream_vmstate,
+        NULL
     }
 };
 
diff --git a/hw/vfio/Makefile.objs b/hw/vfio/Makefile.objs
index e31f30e..d540c9d 100644
--- a/hw/vfio/Makefile.objs
+++ b/hw/vfio/Makefile.objs
@@ -1,4 +1,6 @@
 ifeq ($(CONFIG_LINUX), y)
 obj-$(CONFIG_SOFTMMU) += common.o
 obj-$(CONFIG_PCI) += pci.o
+obj-$(CONFIG_SOFTMMU) += platform.o
+obj-$(CONFIG_SOFTMMU) += calxeda-xgmac.o
 endif
diff --git a/hw/vfio/calxeda-xgmac.c b/hw/vfio/calxeda-xgmac.c
new file mode 100644
index 0000000..eb914f0
--- /dev/null
+++ b/hw/vfio/calxeda-xgmac.c
@@ -0,0 +1,55 @@
+/*
+ * calxeda xgmac VFIO device
+ *
+ * Copyright Linaro Limited, 2014
+ *
+ * Authors:
+ *  Eric Auger <eric.auger@linaro.org>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.  See
+ * the COPYING file in the top-level directory.
+ *
+ */
+
+#include "hw/vfio/vfio-calxeda-xgmac.h"
+
+static void calxeda_xgmac_realize(DeviceState *dev, Error **errp)
+{
+    VFIOPlatformDevice *vdev = VFIO_PLATFORM_DEVICE(dev);
+    VFIOCalxedaXgmacDeviceClass *k = VFIO_CALXEDA_XGMAC_DEVICE_GET_CLASS(dev);
+
+    vdev->compat = g_strdup("calxeda,hb-xgmac");
+
+    k->parent_realize(dev, errp);
+}
+
+static const VMStateDescription vfio_platform_calxeda_xgmac_vmstate = {
+    .name = TYPE_VFIO_CALXEDA_XGMAC,
+    .unmigratable = 1,
+};
+
+static void vfio_calxeda_xgmac_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+    VFIOCalxedaXgmacDeviceClass *vcxc =
+        VFIO_CALXEDA_XGMAC_DEVICE_CLASS(klass);
+    vcxc->parent_realize = dc->realize;
+    dc->realize = calxeda_xgmac_realize;
+    dc->desc = "VFIO Calxeda XGMAC";
+    dc->vmsd = &vfio_platform_calxeda_xgmac_vmstate;
+}
+
+static const TypeInfo vfio_calxeda_xgmac_dev_info = {
+    .name = TYPE_VFIO_CALXEDA_XGMAC,
+    .parent = TYPE_VFIO_PLATFORM,
+    .instance_size = sizeof(VFIOCalxedaXgmacDevice),
+    .class_init = vfio_calxeda_xgmac_class_init,
+    .class_size = sizeof(VFIOCalxedaXgmacDeviceClass),
+};
+
+static void register_calxeda_xgmac_dev_type(void)
+{
+    type_register_static(&vfio_calxeda_xgmac_dev_info);
+}
+
+type_init(register_calxeda_xgmac_dev_type)
diff --git a/hw/vfio/platform.c b/hw/vfio/platform.c
new file mode 100644
index 0000000..9382bb7
--- /dev/null
+++ b/hw/vfio/platform.c
@@ -0,0 +1,615 @@
+/*
+ * vfio based device assignment support - platform devices
+ *
+ * Copyright Linaro Limited, 2014
+ *
+ * Authors:
+ *  Kim Phillips <kim.phillips@linaro.org>
+ *  Eric Auger <eric.auger@linaro.org>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.  See
+ * the COPYING file in the top-level directory.
+ *
+ * Based on vfio based PCI device assignment support:
+ *  Copyright Red Hat, Inc. 2012
+ */
+
+#include <linux/vfio.h>
+#include <sys/ioctl.h>
+
+#include "hw/vfio/vfio-platform.h"
+#include "qemu/error-report.h"
+#include "qemu/range.h"
+#include "sysemu/sysemu.h"
+#include "exec/memory.h"
+#include "qemu/queue.h"
+#include "hw/sysbus.h"
+#include "trace.h"
+#include "hw/platform-bus.h"
+
+/*
+ * Functions used whatever the injection method
+ */
+
+/**
+ * vfio_init_intp - allocate, initialize the IRQ struct pointer
+ * and add it into the list of IRQs
+ * @vbasedev: the VFIO device handle
+ * @info: irq info struct retrieved from VFIO driver
+ */
+static VFIOINTp *vfio_init_intp(VFIODevice *vbasedev,
+                                struct vfio_irq_info info)
+{
+    int ret;
+    VFIOPlatformDevice *vdev =
+        container_of(vbasedev, VFIOPlatformDevice, vbasedev);
+    SysBusDevice *sbdev = SYS_BUS_DEVICE(vdev);
+    VFIOINTp *intp;
+
+    intp = g_malloc0(sizeof(*intp));
+    intp->vdev = vdev;
+    intp->pin = info.index;
+    intp->flags = info.flags;
+    intp->state = VFIO_IRQ_INACTIVE;
+
+    sysbus_init_irq(sbdev, &intp->qemuirq);
+
+    /* Get an eventfd for trigger */
+    ret = event_notifier_init(&intp->interrupt, 0);
+    if (ret) {
+        g_free(intp);
+        error_report("vfio: Error: trigger event_notifier_init failed ");
+        return NULL;
+    }
+
+    QLIST_INSERT_HEAD(&vdev->intp_list, intp, next);
+    return intp;
+}
+
+/**
+ * vfio_set_trigger_eventfd - set VFIO eventfd handling
+ *
+ * @intp: IRQ struct handle
+ * @handler: handler to be called on eventfd signaling
+ *
+ * Setup VFIO signaling and attach an optional user-side handler
+ * to the eventfd
+ */
+static int vfio_set_trigger_eventfd(VFIOINTp *intp,
+                                    eventfd_user_side_handler_t handler)
+{
+    VFIODevice *vbasedev = &intp->vdev->vbasedev;
+    struct vfio_irq_set *irq_set;
+    int argsz, ret;
+    int32_t *pfd;
+
+    argsz = sizeof(*irq_set) + sizeof(*pfd);
+    irq_set = g_malloc0(argsz);
+    irq_set->argsz = argsz;
+    irq_set->flags = VFIO_IRQ_SET_DATA_EVENTFD | VFIO_IRQ_SET_ACTION_TRIGGER;
+    irq_set->index = intp->pin;
+    irq_set->start = 0;
+    irq_set->count = 1;
+    pfd = (int32_t *)&irq_set->data;
+    *pfd = event_notifier_get_fd(&intp->interrupt);
+    qemu_set_fd_handler(*pfd, (IOHandler *)handler, NULL, intp);
+    ret = ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, irq_set);
+    g_free(irq_set);
+    if (ret < 0) {
+        error_report("vfio: Failed to set trigger eventfd: %m");
+        qemu_set_fd_handler(*pfd, NULL, NULL, NULL);
+    }
+    return ret;
+}
+
+/*
+ * Functions only used when eventfds are handled on user-side
+ * ie. without irqfd
+ */
+
+/**
+ * vfio_mmap_set_enabled - enable/disable the fast path mode
+ * @vdev: the VFIO platform device
+ * @enabled: the target mmap state
+ *
+ * enabled = true ~ fast path = MMIO region is mmaped (no KVM TRAP);
+ * enabled = false ~ slow path = MMIO region is trapped and region callbacks
+ * are called; slow path enables to trap the device IRQ status register reset
+*/
+
+static void vfio_mmap_set_enabled(VFIOPlatformDevice *vdev, bool enabled)
+{
+    int i;
+
+    trace_vfio_platform_mmap_set_enabled(enabled);
+
+    for (i = 0; i < vdev->vbasedev.num_regions; i++) {
+        VFIORegion *region = vdev->regions[i];
+
+        memory_region_set_enabled(&region->mmap_mem, enabled);
+    }
+}
+
+/**
+ * vfio_intp_mmap_enable - timer function, restores the fast path
+ * if there is no more active IRQ
+ * @opaque: actually points to the VFIO platform device
+ *
+ * Called on mmap timer timout, this function checks whether the
+ * IRQ is still active and if not, restores the fast path.
+ * by construction a single eventfd is handled at a time.
+ * if the IRQ is still active, the timer is re-programmed.
+ */
+static void vfio_intp_mmap_enable(void *opaque)
+{
+    VFIOINTp *tmp;
+    VFIOPlatformDevice *vdev = (VFIOPlatformDevice *)opaque;
+
+    qemu_mutex_lock(&vdev->intp_mutex);
+    QLIST_FOREACH(tmp, &vdev->intp_list, next) {
+        if (tmp->state == VFIO_IRQ_ACTIVE) {
+            trace_vfio_platform_intp_mmap_enable(tmp->pin);
+            /* re-program the timer to check active status later */
+            timer_mod(vdev->mmap_timer,
+                      qemu_clock_get_ms(QEMU_CLOCK_VIRTUAL) +
+                          vdev->mmap_timeout);
+            qemu_mutex_unlock(&vdev->intp_mutex);
+            return;
+        }
+    }
+    vfio_mmap_set_enabled(vdev, true);
+    qemu_mutex_unlock(&vdev->intp_mutex);
+}
+
+/**
+ * vfio_intp_inject_pending_lockheld - Injects a pending IRQ
+ * @opaque: opaque pointer, in practice the VFIOINTp handle
+ *
+ * The function is called on a previous IRQ completion, from
+ * vfio_platform_eoi, while the intp_mutex is locked.
+ * Also in such situation, the slow path already is set and
+ * the mmap timer was already programmed.
+ */
+static void vfio_intp_inject_pending_lockheld(VFIOINTp *intp)
+{
+    trace_vfio_platform_intp_inject_pending_lockheld(intp->pin,
+                              event_notifier_get_fd(&intp->interrupt));
+
+    intp->state = VFIO_IRQ_ACTIVE;
+
+    /* trigger the virtual IRQ */
+    qemu_set_irq(intp->qemuirq, 1);
+}
+
+/**
+ * vfio_intp_interrupt - The user-side eventfd handler
+ * @opaque: opaque pointer which in practice is the VFIOINTp handle
+ *
+ * the function is entered in event handler context:
+ * the vIRQ is injected into the guest if there is no other active
+ * or pending IRQ.
+ */
+static void vfio_intp_interrupt(VFIOINTp *intp)
+{
+    int ret;
+    VFIOINTp *tmp;
+    VFIOPlatformDevice *vdev = intp->vdev;
+    bool delay_handling = false;
+
+    qemu_mutex_lock(&vdev->intp_mutex);
+    if (intp->state == VFIO_IRQ_INACTIVE) {
+        QLIST_FOREACH(tmp, &vdev->intp_list, next) {
+            if (tmp->state == VFIO_IRQ_ACTIVE ||
+                tmp->state == VFIO_IRQ_PENDING) {
+                delay_handling = true;
+                break;
+            }
+        }
+    }
+    if (delay_handling) {
+        /*
+         * the new IRQ gets a pending status and is pushed in
+         * the pending queue
+         */
+        intp->state = VFIO_IRQ_PENDING;
+        trace_vfio_intp_interrupt_set_pending(intp->pin);
+        QSIMPLEQ_INSERT_TAIL(&vdev->pending_intp_queue,
+                             intp, pqnext);
+        ret = event_notifier_test_and_clear(&intp->interrupt);
+        qemu_mutex_unlock(&vdev->intp_mutex);
+        return;
+    }
+
+    trace_vfio_platform_intp_interrupt(intp->pin,
+                              event_notifier_get_fd(&intp->interrupt));
+
+    ret = event_notifier_test_and_clear(&intp->interrupt);
+    if (!ret) {
+        error_report("Error when clearing fd=%d (ret = %d)\n",
+                     event_notifier_get_fd(&intp->interrupt), ret);
+    }
+
+    intp->state = VFIO_IRQ_ACTIVE;
+
+    /* sets slow path */
+    vfio_mmap_set_enabled(vdev, false);
+
+    /* trigger the virtual IRQ */
+    qemu_set_irq(intp->qemuirq, 1);
+
+    /*
+     * Schedule the mmap timer which will restore fastpath when no IRQ
+     * is active anymore
+     */
+    if (vdev->mmap_timeout) {
+        timer_mod(vdev->mmap_timer,
+                  qemu_clock_get_ms(QEMU_CLOCK_VIRTUAL) +
+                      vdev->mmap_timeout);
+    }
+    qemu_mutex_unlock(&vdev->intp_mutex);
+}
+
+/**
+ * vfio_platform_eoi - IRQ completion routine
+ * @vbasedev: the VFIO device handle
+ *
+ * De-asserts the active virtual IRQ and unmasks the physical IRQ
+ * (effective for level sensitive IRQ auto-masked by the  VFIO driver).
+ * Then it handles next pending IRQ if any.
+ * eoi function is called on the first access to any MMIO region
+ * after an IRQ was triggered, trapped since slow path was set.
+ * It is assumed this access corresponds to the IRQ status
+ * register reset. With such a mechanism, a single IRQ can be
+ * handled at a time since there is no way to know which IRQ
+ * was completed by the guest (we would need additional details
+ * about the IRQ status register mask).
+ */
+static void vfio_platform_eoi(VFIODevice *vbasedev)
+{
+    VFIOINTp *intp;
+    VFIOPlatformDevice *vdev =
+        container_of(vbasedev, VFIOPlatformDevice, vbasedev);
+
+    qemu_mutex_lock(&vdev->intp_mutex);
+    QLIST_FOREACH(intp, &vdev->intp_list, next) {
+        if (intp->state == VFIO_IRQ_ACTIVE) {
+            trace_vfio_platform_eoi(intp->pin,
+                                event_notifier_get_fd(&intp->interrupt));
+            intp->state = VFIO_IRQ_INACTIVE;
+
+            /* deassert the virtual IRQ */
+            qemu_set_irq(intp->qemuirq, 0);
+
+            if (intp->flags & VFIO_IRQ_INFO_AUTOMASKED) {
+                /* unmasks the physical level-sensitive IRQ */
+                vfio_unmask_single_irqindex(vbasedev, intp->pin);
+            }
+
+            /* a single IRQ can be active at a time */
+            break;
+        }
+    }
+    /* in case there are pending IRQs, handle the first one */
+    if (!QSIMPLEQ_EMPTY(&vdev->pending_intp_queue)) {
+        intp = QSIMPLEQ_FIRST(&vdev->pending_intp_queue);
+        vfio_intp_inject_pending_lockheld(intp);
+        QSIMPLEQ_REMOVE_HEAD(&vdev->pending_intp_queue, pqnext);
+    }
+    qemu_mutex_unlock(&vdev->intp_mutex);
+}
+
+/**
+ * vfio_start_eventfd_injection - starts the virtual IRQ injection using
+ * user-side handled eventfds
+ * @intp: the IRQ struct pointer
+ */
+
+static int vfio_start_eventfd_injection(VFIOINTp *intp)
+{
+    int ret;
+
+    ret = vfio_set_trigger_eventfd(intp, vfio_intp_interrupt);
+    if (ret) {
+        error_report("vfio: Error: Failed to pass IRQ fd to the driver: %m");
+    }
+    return ret;
+}
+
+/* VFIO skeleton */
+
+static void vfio_platform_compute_needs_reset(VFIODevice *vbasedev)
+{
+    vbasedev->needs_reset = true;
+}
+
+/* not implemented yet */
+static int vfio_platform_hot_reset_multi(VFIODevice *vbasedev)
+{
+    return -1;
+}
+
+/**
+ * vfio_populate_device - Allocate and populate MMIO region
+ * and IRQ structs according to driver returned information
+ * @vbasedev: the VFIO device handle
+ *
+ */
+static int vfio_populate_device(VFIODevice *vbasedev)
+{
+    VFIOINTp *intp, *tmp;
+    int i, ret = -1;
+    VFIOPlatformDevice *vdev =
+        container_of(vbasedev, VFIOPlatformDevice, vbasedev);
+
+    if (!(vbasedev->flags & VFIO_DEVICE_FLAGS_PLATFORM)) {
+        error_report("vfio: Um, this isn't a platform device");
+        return ret;
+    }
+
+    vdev->regions = g_new0(VFIORegion *, vbasedev->num_regions);
+
+    for (i = 0; i < vbasedev->num_regions; i++) {
+        struct vfio_region_info reg_info = { .argsz = sizeof(reg_info) };
+        VFIORegion *ptr;
+
+        vdev->regions[i] = g_malloc0(sizeof(VFIORegion));
+        ptr = vdev->regions[i];
+        reg_info.index = i;
+        ret = ioctl(vbasedev->fd, VFIO_DEVICE_GET_REGION_INFO, &reg_info);
+        if (ret) {
+            error_report("vfio: Error getting region %d info: %m", i);
+            goto reg_error;
+        }
+        ptr->flags = reg_info.flags;
+        ptr->size = reg_info.size;
+        ptr->fd_offset = reg_info.offset;
+        ptr->nr = i;
+        ptr->vbasedev = vbasedev;
+
+        trace_vfio_platform_populate_regions(ptr->nr,
+                            (unsigned long)ptr->flags,
+                            (unsigned long)ptr->size,
+                            ptr->vbasedev->fd,
+                            (unsigned long)ptr->fd_offset);
+    }
+
+    vdev->mmap_timer = timer_new_ms(QEMU_CLOCK_VIRTUAL,
+                                    vfio_intp_mmap_enable, vdev);
+
+    QSIMPLEQ_INIT(&vdev->pending_intp_queue);
+
+    for (i = 0; i < vbasedev->num_irqs; i++) {
+        struct vfio_irq_info irq = { .argsz = sizeof(irq) };
+
+        irq.index = i;
+        ret = ioctl(vbasedev->fd, VFIO_DEVICE_GET_IRQ_INFO, &irq);
+        if (ret) {
+            error_printf("vfio: error getting device %s irq info",
+                         vbasedev->name);
+            goto irq_err;
+        } else {
+            trace_vfio_platform_populate_interrupts(irq.index,
+                                                    irq.count,
+                                                    irq.flags);
+            intp = vfio_init_intp(vbasedev, irq);
+            if (!intp) {
+                error_report("vfio: Error installing IRQ %d up", i);
+                goto irq_err;
+            }
+        }
+    }
+    return 0;
+irq_err:
+    timer_del(vdev->mmap_timer);
+    QLIST_FOREACH_SAFE(intp, &vdev->intp_list, next, tmp) {
+        QLIST_REMOVE(intp, next);
+        g_free(intp);
+    }
+reg_error:
+    for (i = 0; i < vbasedev->num_regions; i++) {
+        g_free(vdev->regions[i]);
+    }
+    g_free(vdev->regions);
+    return ret;
+}
+
+/* specialized functions for VFIO Platform devices */
+static VFIODeviceOps vfio_platform_ops = {
+    .vfio_compute_needs_reset = vfio_platform_compute_needs_reset,
+    .vfio_hot_reset_multi = vfio_platform_hot_reset_multi,
+    .vfio_eoi = vfio_platform_eoi,
+};
+
+/**
+ * vfio_base_device_init - perform preliminary VFIO setup
+ * @vbasedev: the VFIO device handle
+ *
+ * Implement the VFIO command sequence that allows to discover
+ * assigned device resources: group extraction, device
+ * fd retrieval, resource query.
+ * Precondition: the device name must be initialized
+ */
+static int vfio_base_device_init(VFIODevice *vbasedev)
+{
+    VFIOGroup *group;
+    VFIODevice *vbasedev_iter;
+    char path[PATH_MAX], iommu_group_path[PATH_MAX], *group_name;
+    ssize_t len;
+    struct stat st;
+    int groupid;
+    int ret;
+
+    /* name must be set prior to the call */
+    if (!vbasedev->name || strchr(vbasedev->name, '/')) {
+        return -EINVAL;
+    }
+
+    /* Check that the host device exists */
+    g_snprintf(path, sizeof(path), "/sys/bus/platform/devices/%s/",
+               vbasedev->name);
+
+    if (stat(path, &st) < 0) {
+        error_report("vfio: error: no such host device: %s", path);
+        return -errno;
+    }
+
+    g_strlcat(path, "iommu_group", sizeof(path));
+    len = readlink(path, iommu_group_path, sizeof(iommu_group_path));
+    if (len < 0 || len >= sizeof(iommu_group_path)) {
+        error_report("vfio: error no iommu_group for device");
+        return len < 0 ? -errno : -ENAMETOOLONG;
+    }
+
+    iommu_group_path[len] = 0;
+    group_name = basename(iommu_group_path);
+
+    if (sscanf(group_name, "%d", &groupid) != 1) {
+        error_report("vfio: error reading %s: %m", path);
+        return -errno;
+    }
+
+    trace_vfio_platform_base_device_init(vbasedev->name, groupid);
+
+    group = vfio_get_group(groupid, &address_space_memory);
+    if (!group) {
+        error_report("vfio: failed to get group %d", groupid);
+        return -ENOENT;
+    }
+
+    g_snprintf(path, sizeof(path), "%s", vbasedev->name);
+
+    QLIST_FOREACH(vbasedev_iter, &group->device_list, next) {
+        if (strcmp(vbasedev_iter->name, vbasedev->name) == 0) {
+            error_report("vfio: error: device %s is already attached", path);
+            vfio_put_group(group);
+            return -EBUSY;
+        }
+    }
+    ret = vfio_get_device(group, path, vbasedev);
+    if (ret) {
+        error_report("vfio: failed to get device %s", path);
+        vfio_put_group(group);
+        return ret;
+    }
+
+    ret = vfio_populate_device(vbasedev);
+    if (ret) {
+        error_report("vfio: failed to populate device %s", path);
+        vfio_put_group(group);
+    }
+
+    return ret;
+}
+
+/**
+ * vfio_map_region - initialize the 2 memory regions for a given
+ * MMIO region index
+ * @vdev: the VFIO platform device handle
+ * @nr: the index of the region
+ *
+ * Init the top memory region and the mmapped memory region beneath
+ * VFIOPlatformDevice is used since VFIODevice is not a QOM Object
+ * and could not be passed to memory region functions
+*/
+static void vfio_map_region(VFIOPlatformDevice *vdev, int nr)
+{
+    VFIORegion *region = vdev->regions[nr];
+    uint64_t size = region->size;
+    char name[64];
+
+    if (!size) {
+        return;
+    }
+
+    g_snprintf(name, sizeof(name), "VFIO %s region %d",
+               vdev->vbasedev.name, nr);
+
+    /* A "slow" read/write mapping underlies all regions */
+    memory_region_init_io(&region->mem, OBJECT(vdev), &vfio_region_ops,
+                          region, name, size);
+
+    g_strlcat(name, " mmap", sizeof(name));
+
+    if (vfio_mmap_region(OBJECT(vdev), region, &region->mem,
+                         &region->mmap_mem, &region->mmap, size, 0, name)) {
+        error_report("%s unsupported. Performance may be slow", name);
+    }
+}
+
+/**
+ * vfio_platform_realize  - the device realize function
+ * @dev: device state pointer
+ * @errp: error
+ *
+ * initialize the device, its memory regions and IRQ structures
+ * IRQ are started separately
+ */
+static void vfio_platform_realize(DeviceState *dev, Error **errp)
+{
+    VFIOPlatformDevice *vdev = VFIO_PLATFORM_DEVICE(dev);
+    SysBusDevice *sbdev = SYS_BUS_DEVICE(dev);
+    VFIODevice *vbasedev = &vdev->vbasedev;
+    VFIOINTp *intp;
+    int i, ret;
+
+    vbasedev->type = VFIO_DEVICE_TYPE_PLATFORM;
+    vbasedev->ops = &vfio_platform_ops;
+
+    trace_vfio_platform_realize(vbasedev->name, vdev->compat);
+
+    ret = vfio_base_device_init(vbasedev);
+    if (ret) {
+        error_setg(errp, "vfio: vfio_base_device_init failed for %s",
+                   vbasedev->name);
+        return;
+    }
+
+    for (i = 0; i < vbasedev->num_regions; i++) {
+        vfio_map_region(vdev, i);
+        sysbus_init_mmio(sbdev, &vdev->regions[i]->mem);
+    }
+
+    QLIST_FOREACH(intp, &vdev->intp_list, next) {
+        vfio_start_eventfd_injection(intp);
+    }
+}
+
+static const VMStateDescription vfio_platform_vmstate = {
+    .name = TYPE_VFIO_PLATFORM,
+    .unmigratable = 1,
+};
+
+static Property vfio_platform_dev_properties[] = {
+    DEFINE_PROP_STRING("host", VFIOPlatformDevice, vbasedev.name),
+    DEFINE_PROP_BOOL("x-mmap", VFIOPlatformDevice, vbasedev.allow_mmap, true),
+    DEFINE_PROP_UINT32("mmap-timeout-ms", VFIOPlatformDevice,
+                       mmap_timeout, 1100),
+    DEFINE_PROP_END_OF_LIST(),
+};
+
+static void vfio_platform_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+
+    dc->realize = vfio_platform_realize;
+    dc->props = vfio_platform_dev_properties;
+    dc->vmsd = &vfio_platform_vmstate;
+    dc->desc = "VFIO-based platform device assignment";
+    set_bit(DEVICE_CATEGORY_MISC, dc->categories);
+}
+
+static const TypeInfo vfio_platform_dev_info = {
+    .name = TYPE_VFIO_PLATFORM,
+    .parent = TYPE_SYS_BUS_DEVICE,
+    .instance_size = sizeof(VFIOPlatformDevice),
+    .class_init = vfio_platform_class_init,
+    .class_size = sizeof(VFIOPlatformDeviceClass),
+    .abstract   = true,
+};
+
+static void register_vfio_platform_dev_type(void)
+{
+    type_register_static(&vfio_platform_dev_info);
+}
+
+type_init(register_vfio_platform_dev_type)
diff --git a/hw/virtio/dataplane/vring.c b/hw/virtio/dataplane/vring.c
index e378733..3589185 100644
--- a/hw/virtio/dataplane/vring.c
+++ b/hw/virtio/dataplane/vring.c
@@ -157,15 +157,18 @@
 }
 
 
-static int get_desc(Vring *vring, VirtQueueElement *elem,
+static int get_desc(VirtIODevice *vdev, Vring *vring, VirtQueueElement *elem,
                     struct vring_desc *desc)
 {
     unsigned *num;
     struct iovec *iov;
     hwaddr *addr;
     MemoryRegion *mr;
+    int is_write = virtio_tswap16(vdev, desc->flags) & VRING_DESC_F_WRITE;
+    uint32_t len = virtio_tswap32(vdev, desc->len);
+    uint64_t desc_addr = virtio_tswap64(vdev, desc->addr);
 
-    if (desc->flags & VRING_DESC_F_WRITE) {
+    if (is_write) {
         num = &elem->in_num;
         iov = &elem->in_sg[*num];
         addr = &elem->in_addr[*num];
@@ -189,18 +192,17 @@
     }
 
     /* TODO handle non-contiguous memory across region boundaries */
-    iov->iov_base = vring_map(&mr, desc->addr, desc->len,
-                              desc->flags & VRING_DESC_F_WRITE);
+    iov->iov_base = vring_map(&mr, desc_addr, len, is_write);
     if (!iov->iov_base) {
         error_report("Failed to map descriptor addr %#" PRIx64 " len %u",
-                     (uint64_t)desc->addr, desc->len);
+                     (uint64_t)desc_addr, len);
         return -EFAULT;
     }
 
     /* The MemoryRegion is looked up again and unref'ed later, leave the
      * ref in place.  */
-    iov->iov_len = desc->len;
-    *addr = desc->addr;
+    iov->iov_len = len;
+    *addr = desc_addr;
     *num += 1;
     return 0;
 }
@@ -222,21 +224,23 @@
     struct vring_desc desc;
     unsigned int i = 0, count, found = 0;
     int ret;
+    uint32_t len = virtio_tswap32(vdev, indirect->len);
+    uint64_t addr = virtio_tswap64(vdev, indirect->addr);
 
     /* Sanity check */
-    if (unlikely(indirect->len % sizeof(desc))) {
+    if (unlikely(len % sizeof(desc))) {
         error_report("Invalid length in indirect descriptor: "
                      "len %#x not multiple of %#zx",
-                     indirect->len, sizeof(desc));
+                     len, sizeof(desc));
         vring->broken = true;
         return -EFAULT;
     }
 
-    count = indirect->len / sizeof(desc);
+    count = len / sizeof(desc);
     /* Buffers are chained via a 16 bit next field, so
      * we can have at most 2^16 of these. */
     if (unlikely(count > USHRT_MAX + 1)) {
-        error_report("Indirect buffer length too big: %d", indirect->len);
+        error_report("Indirect buffer length too big: %d", len);
         vring->broken = true;
         return -EFAULT;
     }
@@ -247,12 +251,12 @@
 
         /* Translate indirect descriptor */
         desc_ptr = vring_map(&mr,
-                             indirect->addr + found * sizeof(desc),
+                             addr + found * sizeof(desc),
                              sizeof(desc), false);
         if (!desc_ptr) {
             error_report("Failed to map indirect descriptor "
                          "addr %#" PRIx64 " len %zu",
-                         (uint64_t)indirect->addr + found * sizeof(desc),
+                         (uint64_t)addr + found * sizeof(desc),
                          sizeof(desc));
             vring->broken = true;
             return -EFAULT;
@@ -270,19 +274,20 @@
             return -EFAULT;
         }
 
-        if (unlikely(desc.flags & VRING_DESC_F_INDIRECT)) {
+        if (unlikely(virtio_tswap16(vdev, desc.flags)
+                     & VRING_DESC_F_INDIRECT)) {
             error_report("Nested indirect descriptor");
             vring->broken = true;
             return -EFAULT;
         }
 
-        ret = get_desc(vring, elem, &desc);
+        ret = get_desc(vdev, vring, elem, &desc);
         if (ret < 0) {
             vring->broken |= (ret == -EFAULT);
             return ret;
         }
-        i = desc.next;
-    } while (desc.flags & VRING_DESC_F_NEXT);
+        i = virtio_tswap16(vdev, desc.next);
+    } while (virtio_tswap16(vdev, desc.flags) & VRING_DESC_F_NEXT);
     return 0;
 }
 
@@ -383,7 +388,7 @@
         /* Ensure descriptor is loaded before accessing fields */
         barrier();
 
-        if (desc.flags & VRING_DESC_F_INDIRECT) {
+        if (virtio_tswap16(vdev, desc.flags) & VRING_DESC_F_INDIRECT) {
             ret = get_indirect(vdev, vring, elem, &desc);
             if (ret < 0) {
                 goto out;
@@ -391,13 +396,13 @@
             continue;
         }
 
-        ret = get_desc(vring, elem, &desc);
+        ret = get_desc(vdev, vring, elem, &desc);
         if (ret < 0) {
             goto out;
         }
 
-        i = desc.next;
-    } while (desc.flags & VRING_DESC_F_NEXT);
+        i = virtio_tswap16(vdev, desc.next);
+    } while (virtio_tswap16(vdev, desc.flags) & VRING_DESC_F_NEXT);
 
     /* On success, increment avail index. */
     vring->last_avail_idx++;
diff --git a/hw/virtio/vhost.c b/hw/virtio/vhost.c
index a7fe3c5..2d6c27a 100644
--- a/hw/virtio/vhost.c
+++ b/hw/virtio/vhost.c
@@ -591,7 +591,7 @@
     uint64_t features = dev->acked_features;
     int r;
     if (enable_log) {
-        features |= 0x1 << VHOST_F_LOG_ALL;
+        features |= 0x1ULL << VHOST_F_LOG_ALL;
     }
     r = dev->vhost_ops->vhost_call(dev, VHOST_SET_FEATURES, &features);
     return r < 0 ? -errno : 0;
@@ -902,7 +902,7 @@
         .priority = 10
     };
     hdev->migration_blocker = NULL;
-    if (!(hdev->features & (0x1 << VHOST_F_LOG_ALL))) {
+    if (!(hdev->features & (0x1ULL << VHOST_F_LOG_ALL))) {
         error_setg(&hdev->migration_blocker,
                    "Migration disabled: vhost lacks VHOST_F_LOG_ALL feature.");
         migrate_add_blocker(hdev->migration_blocker);
@@ -1045,12 +1045,12 @@
     assert(r >= 0);
 }
 
-unsigned vhost_get_features(struct vhost_dev *hdev, const int *feature_bits,
-        unsigned features)
+uint64_t vhost_get_features(struct vhost_dev *hdev, const int *feature_bits,
+                            uint64_t features)
 {
     const int *bit = feature_bits;
     while (*bit != VHOST_INVALID_FEATURE_BIT) {
-        unsigned bit_mask = (1 << *bit);
+        uint64_t bit_mask = (1ULL << *bit);
         if (!(hdev->features & bit_mask)) {
             features &= ~bit_mask;
         }
@@ -1060,11 +1060,11 @@
 }
 
 void vhost_ack_features(struct vhost_dev *hdev, const int *feature_bits,
-        unsigned features)
+                        uint64_t features)
 {
     const int *bit = feature_bits;
     while (*bit != VHOST_INVALID_FEATURE_BIT) {
-        unsigned bit_mask = (1 << *bit);
+        uint64_t bit_mask = (1ULL << *bit);
         if (features & bit_mask) {
             hdev->acked_features |= bit_mask;
         }
@@ -1114,9 +1114,7 @@
 
     return 0;
 fail_log:
-    if (hdev->log_size) {
-        vhost_log_put(hdev, false);
-    }
+    vhost_log_put(hdev, false);
 fail_vq:
     while (--i >= 0) {
         vhost_virtqueue_stop(hdev,
diff --git a/hw/virtio/virtio-balloon.c b/hw/virtio/virtio-balloon.c
index f915c7b..78bc14f 100644
--- a/hw/virtio/virtio-balloon.c
+++ b/hw/virtio/virtio-balloon.c
@@ -312,7 +312,7 @@
 
 static uint64_t virtio_balloon_get_features(VirtIODevice *vdev, uint64_t f)
 {
-    f |= (1 << VIRTIO_BALLOON_F_STATS_VQ);
+    virtio_add_feature(&f, VIRTIO_BALLOON_F_STATS_VQ);
     return f;
 }
 
diff --git a/hw/virtio/virtio-mmio.c b/hw/virtio/virtio-mmio.c
index c8f7294..18660b0 100644
--- a/hw/virtio/virtio-mmio.c
+++ b/hw/virtio/virtio-mmio.c
@@ -333,8 +333,11 @@
     case VIRTIO_MMIO_QUEUENUM:
         DPRINTF("mmio_queue write %d max %d\n", (int)value, VIRTQUEUE_MAX_SIZE);
         virtio_queue_set_num(vdev, vdev->queue_sel, value);
+        /* Note: only call this function for legacy devices */
+        virtio_queue_update_rings(vdev, vdev->queue_sel);
         break;
     case VIRTIO_MMIO_QUEUEALIGN:
+        /* Note: this is only valid for legacy devices */
         virtio_queue_set_align(vdev, vdev->queue_sel, value);
         break;
     case VIRTIO_MMIO_QUEUEPFN:
diff --git a/hw/virtio/virtio-pci.c b/hw/virtio/virtio-pci.c
index 6d4f64e..d7cf34c 100644
--- a/hw/virtio/virtio-pci.c
+++ b/hw/virtio/virtio-pci.c
@@ -24,6 +24,7 @@
 #include "hw/virtio/virtio-serial.h"
 #include "hw/virtio/virtio-scsi.h"
 #include "hw/virtio/virtio-balloon.h"
+#include "hw/virtio/virtio-input.h"
 #include "hw/pci/pci.h"
 #include "qemu/error-report.h"
 #include "hw/pci/msi.h"
@@ -38,6 +39,8 @@
 
 #define VIRTIO_PCI_REGION_SIZE(dev)     VIRTIO_PCI_CONFIG_OFF(msix_present(dev))
 
+#undef VIRTIO_PCI_CONFIG
+
 /* The remaining space is defined by each driver as the per-driver
  * configuration space */
 #define VIRTIO_PCI_CONFIG_SIZE(dev)     VIRTIO_PCI_CONFIG_OFF(msix_enabled(dev))
@@ -133,12 +136,21 @@
     return 0;
 }
 
+#define QEMU_VIRTIO_PCI_QUEUE_MEM_MULT 0x1000
+
 static int virtio_pci_set_host_notifier_internal(VirtIOPCIProxy *proxy,
                                                  int n, bool assign, bool set_handler)
 {
     VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
     VirtQueue *vq = virtio_get_queue(vdev, n);
     EventNotifier *notifier = virtio_queue_get_host_notifier(vq);
+    bool legacy = !(proxy->flags & VIRTIO_PCI_FLAG_DISABLE_LEGACY);
+    bool modern = !(proxy->flags & VIRTIO_PCI_FLAG_DISABLE_MODERN);
+    MemoryRegion *modern_mr = &proxy->notify.mr;
+    MemoryRegion *legacy_mr = &proxy->bar;
+    hwaddr modern_addr = QEMU_VIRTIO_PCI_QUEUE_MEM_MULT *
+                         virtio_get_queue_index(vq);
+    hwaddr legacy_addr = VIRTIO_PCI_QUEUE_NOTIFY;
     int r = 0;
 
     if (assign) {
@@ -149,11 +161,23 @@
             return r;
         }
         virtio_queue_set_host_notifier_fd_handler(vq, true, set_handler);
-        memory_region_add_eventfd(&proxy->bar, VIRTIO_PCI_QUEUE_NOTIFY, 2,
-                                  true, n, notifier);
+        if (modern) {
+            memory_region_add_eventfd(modern_mr, modern_addr, 2,
+                                      true, n, notifier);
+        }
+        if (legacy) {
+            memory_region_add_eventfd(legacy_mr, legacy_addr, 2,
+                                      true, n, notifier);
+        }
     } else {
-        memory_region_del_eventfd(&proxy->bar, VIRTIO_PCI_QUEUE_NOTIFY, 2,
-                                  true, n, notifier);
+        if (modern) {
+            memory_region_del_eventfd(modern_mr, modern_addr, 2,
+                                      true, n, notifier);
+        }
+        if (legacy) {
+            memory_region_del_eventfd(legacy_mr, legacy_addr, 2,
+                                      true, n, notifier);
+        }
         virtio_queue_set_host_notifier_fd_handler(vq, false, false);
         event_notifier_cleanup(notifier);
     }
@@ -918,11 +942,359 @@
     return proxy->nvectors;
 }
 
+static void virtio_pci_add_mem_cap(VirtIOPCIProxy *proxy,
+                                   struct virtio_pci_cap *cap)
+{
+    PCIDevice *dev = &proxy->pci_dev;
+    int offset;
+
+    offset = pci_add_capability(dev, PCI_CAP_ID_VNDR, 0, cap->cap_len);
+    assert(offset > 0);
+
+    assert(cap->cap_len >= sizeof *cap);
+    memcpy(dev->config + offset + PCI_CAP_FLAGS, &cap->cap_len,
+           cap->cap_len - PCI_CAP_FLAGS);
+}
+
+static uint64_t virtio_pci_common_read(void *opaque, hwaddr addr,
+                                       unsigned size)
+{
+    VirtIOPCIProxy *proxy = opaque;
+    VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
+    uint32_t val = 0;
+    int i;
+
+    switch (addr) {
+    case VIRTIO_PCI_COMMON_DFSELECT:
+        val = proxy->dfselect;
+        break;
+    case VIRTIO_PCI_COMMON_DF:
+        if (proxy->dfselect <= 1) {
+            val = vdev->host_features >> (32 * proxy->dfselect);
+        }
+        break;
+    case VIRTIO_PCI_COMMON_GFSELECT:
+        val = proxy->gfselect;
+        break;
+    case VIRTIO_PCI_COMMON_GF:
+        if (proxy->gfselect <= ARRAY_SIZE(proxy->guest_features)) {
+            val = proxy->guest_features[proxy->gfselect];
+        }
+        break;
+    case VIRTIO_PCI_COMMON_MSIX:
+        val = vdev->config_vector;
+        break;
+    case VIRTIO_PCI_COMMON_NUMQ:
+        for (i = 0; i < VIRTIO_QUEUE_MAX; ++i) {
+            if (virtio_queue_get_num(vdev, i)) {
+                val = i + 1;
+            }
+        }
+        break;
+    case VIRTIO_PCI_COMMON_STATUS:
+        val = vdev->status;
+        break;
+    case VIRTIO_PCI_COMMON_CFGGENERATION:
+        val = vdev->generation;
+        break;
+    case VIRTIO_PCI_COMMON_Q_SELECT:
+        val = vdev->queue_sel;
+        break;
+    case VIRTIO_PCI_COMMON_Q_SIZE:
+        val = virtio_queue_get_num(vdev, vdev->queue_sel);
+        break;
+    case VIRTIO_PCI_COMMON_Q_MSIX:
+        val = virtio_queue_vector(vdev, vdev->queue_sel);
+        break;
+    case VIRTIO_PCI_COMMON_Q_ENABLE:
+        val = proxy->vqs[vdev->queue_sel].enabled;
+        break;
+    case VIRTIO_PCI_COMMON_Q_NOFF:
+        /* Simply map queues in order */
+        val = vdev->queue_sel;
+        break;
+    case VIRTIO_PCI_COMMON_Q_DESCLO:
+        val = proxy->vqs[vdev->queue_sel].desc[0];
+        break;
+    case VIRTIO_PCI_COMMON_Q_DESCHI:
+        val = proxy->vqs[vdev->queue_sel].desc[1];
+        break;
+    case VIRTIO_PCI_COMMON_Q_AVAILLO:
+        val = proxy->vqs[vdev->queue_sel].avail[0];
+        break;
+    case VIRTIO_PCI_COMMON_Q_AVAILHI:
+        val = proxy->vqs[vdev->queue_sel].avail[1];
+        break;
+    case VIRTIO_PCI_COMMON_Q_USEDLO:
+        val = proxy->vqs[vdev->queue_sel].used[0];
+        break;
+    case VIRTIO_PCI_COMMON_Q_USEDHI:
+        val = proxy->vqs[vdev->queue_sel].used[1];
+        break;
+    default:
+        val = 0;
+    }
+
+    return val;
+}
+
+static void virtio_pci_common_write(void *opaque, hwaddr addr,
+                                    uint64_t val, unsigned size)
+{
+    VirtIOPCIProxy *proxy = opaque;
+    VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
+
+    switch (addr) {
+    case VIRTIO_PCI_COMMON_DFSELECT:
+        proxy->dfselect = val;
+        break;
+    case VIRTIO_PCI_COMMON_GFSELECT:
+        proxy->gfselect = val;
+        break;
+    case VIRTIO_PCI_COMMON_GF:
+        if (proxy->gfselect <= ARRAY_SIZE(proxy->guest_features)) {
+            proxy->guest_features[proxy->gfselect] = val;
+            virtio_set_features(vdev,
+                                (((uint64_t)proxy->guest_features[1]) << 32) |
+                                proxy->guest_features[0]);
+        }
+        break;
+    case VIRTIO_PCI_COMMON_MSIX:
+        msix_vector_unuse(&proxy->pci_dev, vdev->config_vector);
+        /* Make it possible for guest to discover an error took place. */
+        if (msix_vector_use(&proxy->pci_dev, val) < 0) {
+            val = VIRTIO_NO_VECTOR;
+        }
+        vdev->config_vector = val;
+        break;
+    case VIRTIO_PCI_COMMON_STATUS:
+        if (!(val & VIRTIO_CONFIG_S_DRIVER_OK)) {
+            virtio_pci_stop_ioeventfd(proxy);
+        }
+
+        virtio_set_status(vdev, val & 0xFF);
+
+        if (val & VIRTIO_CONFIG_S_DRIVER_OK) {
+            virtio_pci_start_ioeventfd(proxy);
+        }
+
+        if (vdev->status == 0) {
+            virtio_reset(vdev);
+            msix_unuse_all_vectors(&proxy->pci_dev);
+        }
+
+        break;
+    case VIRTIO_PCI_COMMON_Q_SELECT:
+        if (val < VIRTIO_QUEUE_MAX) {
+            vdev->queue_sel = val;
+        }
+        break;
+    case VIRTIO_PCI_COMMON_Q_SIZE:
+        proxy->vqs[vdev->queue_sel].num = val;
+        break;
+    case VIRTIO_PCI_COMMON_Q_MSIX:
+        msix_vector_unuse(&proxy->pci_dev,
+                          virtio_queue_vector(vdev, vdev->queue_sel));
+        /* Make it possible for guest to discover an error took place. */
+        if (msix_vector_use(&proxy->pci_dev, val) < 0) {
+            val = VIRTIO_NO_VECTOR;
+        }
+        virtio_queue_set_vector(vdev, vdev->queue_sel, val);
+        break;
+    case VIRTIO_PCI_COMMON_Q_ENABLE:
+        /* TODO: need a way to put num back on reset. */
+        virtio_queue_set_num(vdev, vdev->queue_sel,
+                             proxy->vqs[vdev->queue_sel].num);
+        virtio_queue_set_rings(vdev, vdev->queue_sel,
+                       ((uint64_t)proxy->vqs[vdev->queue_sel].desc[1]) << 32 |
+                       proxy->vqs[vdev->queue_sel].desc[0],
+                       ((uint64_t)proxy->vqs[vdev->queue_sel].avail[1]) << 32 |
+                       proxy->vqs[vdev->queue_sel].avail[0],
+                       ((uint64_t)proxy->vqs[vdev->queue_sel].used[1]) << 32 |
+                       proxy->vqs[vdev->queue_sel].used[0]);
+        break;
+    case VIRTIO_PCI_COMMON_Q_DESCLO:
+        proxy->vqs[vdev->queue_sel].desc[0] = val;
+        break;
+    case VIRTIO_PCI_COMMON_Q_DESCHI:
+        proxy->vqs[vdev->queue_sel].desc[1] = val;
+        break;
+    case VIRTIO_PCI_COMMON_Q_AVAILLO:
+        proxy->vqs[vdev->queue_sel].avail[0] = val;
+        break;
+    case VIRTIO_PCI_COMMON_Q_AVAILHI:
+        proxy->vqs[vdev->queue_sel].avail[1] = val;
+        break;
+    case VIRTIO_PCI_COMMON_Q_USEDLO:
+        proxy->vqs[vdev->queue_sel].used[0] = val;
+        break;
+    case VIRTIO_PCI_COMMON_Q_USEDHI:
+        proxy->vqs[vdev->queue_sel].used[1] = val;
+        break;
+    default:
+        break;
+    }
+}
+
+
+static uint64_t virtio_pci_notify_read(void *opaque, hwaddr addr,
+                                       unsigned size)
+{
+    return 0;
+}
+
+static void virtio_pci_notify_write(void *opaque, hwaddr addr,
+                                    uint64_t val, unsigned size)
+{
+    VirtIODevice *vdev = opaque;
+    unsigned queue = addr / QEMU_VIRTIO_PCI_QUEUE_MEM_MULT;
+
+    if (queue < VIRTIO_QUEUE_MAX) {
+        virtio_queue_notify(vdev, queue);
+    }
+}
+
+static uint64_t virtio_pci_isr_read(void *opaque, hwaddr addr,
+                                    unsigned size)
+{
+    VirtIOPCIProxy *proxy = opaque;
+    VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
+    uint64_t val = vdev->isr;
+
+    vdev->isr = 0;
+    pci_irq_deassert(&proxy->pci_dev);
+
+    return val;
+}
+
+static void virtio_pci_isr_write(void *opaque, hwaddr addr,
+                                 uint64_t val, unsigned size)
+{
+}
+
+static uint64_t virtio_pci_device_read(void *opaque, hwaddr addr,
+                                       unsigned size)
+{
+    VirtIODevice *vdev = opaque;
+    uint64_t val = 0;
+
+    switch (size) {
+    case 1:
+        val = virtio_config_modern_readb(vdev, addr);
+        break;
+    case 2:
+        val = virtio_config_modern_readw(vdev, addr);
+        break;
+    case 4:
+        val = virtio_config_modern_readl(vdev, addr);
+        break;
+    }
+    return val;
+}
+
+static void virtio_pci_device_write(void *opaque, hwaddr addr,
+                                    uint64_t val, unsigned size)
+{
+    VirtIODevice *vdev = opaque;
+    switch (size) {
+    case 1:
+        virtio_config_modern_writeb(vdev, addr, val);
+        break;
+    case 2:
+        virtio_config_modern_writew(vdev, addr, val);
+        break;
+    case 4:
+        virtio_config_modern_writel(vdev, addr, val);
+        break;
+    }
+}
+
+static void virtio_pci_modern_regions_init(VirtIOPCIProxy *proxy)
+{
+    static const MemoryRegionOps common_ops = {
+        .read = virtio_pci_common_read,
+        .write = virtio_pci_common_write,
+        .impl = {
+            .min_access_size = 1,
+            .max_access_size = 4,
+        },
+        .endianness = DEVICE_LITTLE_ENDIAN,
+    };
+    static const MemoryRegionOps isr_ops = {
+        .read = virtio_pci_isr_read,
+        .write = virtio_pci_isr_write,
+        .impl = {
+            .min_access_size = 1,
+            .max_access_size = 4,
+        },
+        .endianness = DEVICE_LITTLE_ENDIAN,
+    };
+    static const MemoryRegionOps device_ops = {
+        .read = virtio_pci_device_read,
+        .write = virtio_pci_device_write,
+        .impl = {
+            .min_access_size = 1,
+            .max_access_size = 4,
+        },
+        .endianness = DEVICE_LITTLE_ENDIAN,
+    };
+    static const MemoryRegionOps notify_ops = {
+        .read = virtio_pci_notify_read,
+        .write = virtio_pci_notify_write,
+        .impl = {
+            .min_access_size = 1,
+            .max_access_size = 4,
+        },
+        .endianness = DEVICE_LITTLE_ENDIAN,
+    };
+
+    memory_region_init_io(&proxy->common.mr, OBJECT(proxy),
+                          &common_ops,
+                          proxy,
+                          "virtio-pci-common",
+                          proxy->common.size);
+
+    memory_region_init_io(&proxy->isr.mr, OBJECT(proxy),
+                          &isr_ops,
+                          proxy,
+                          "virtio-pci-isr",
+                          proxy->isr.size);
+
+    memory_region_init_io(&proxy->device.mr, OBJECT(proxy),
+                          &device_ops,
+                          virtio_bus_get_device(&proxy->bus),
+                          "virtio-pci-device",
+                          proxy->device.size);
+
+    memory_region_init_io(&proxy->notify.mr, OBJECT(proxy),
+                          &notify_ops,
+                          virtio_bus_get_device(&proxy->bus),
+                          "virtio-pci-notify",
+                          proxy->notify.size);
+}
+
+static void virtio_pci_modern_region_map(VirtIOPCIProxy *proxy,
+                                         VirtIOPCIRegion *region,
+                                         struct virtio_pci_cap *cap)
+{
+    memory_region_add_subregion(&proxy->modern_bar,
+                                region->offset,
+                                &region->mr);
+
+    cap->cfg_type = region->type;
+    cap->bar = proxy->modern_mem_bar;
+    cap->offset = cpu_to_le32(region->offset);
+    cap->length = cpu_to_le32(region->size);
+    virtio_pci_add_mem_cap(proxy, cap);
+}
+
 /* This is called by virtio-bus just after the device is plugged. */
 static void virtio_pci_device_plugged(DeviceState *d, Error **errp)
 {
     VirtIOPCIProxy *proxy = VIRTIO_PCI(d);
     VirtioBusState *bus = &proxy->bus;
+    bool legacy = !(proxy->flags & VIRTIO_PCI_FLAG_DISABLE_LEGACY);
+    bool modern = !(proxy->flags & VIRTIO_PCI_FLAG_DISABLE_MODERN);
     uint8_t *config;
     uint32_t size;
     VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
@@ -931,13 +1303,51 @@
     if (proxy->class_code) {
         pci_config_set_class(config, proxy->class_code);
     }
-    pci_set_word(config + PCI_SUBSYSTEM_VENDOR_ID,
-                 pci_get_word(config + PCI_VENDOR_ID));
-    pci_set_word(config + PCI_SUBSYSTEM_ID, virtio_bus_get_vdev_id(bus));
+
+    if (legacy) {
+        /* legacy and transitional */
+        pci_set_word(config + PCI_SUBSYSTEM_VENDOR_ID,
+                     pci_get_word(config + PCI_VENDOR_ID));
+        pci_set_word(config + PCI_SUBSYSTEM_ID, virtio_bus_get_vdev_id(bus));
+    } else {
+        /* pure virtio-1.0 */
+        pci_set_word(config + PCI_VENDOR_ID,
+                     PCI_VENDOR_ID_REDHAT_QUMRANET);
+        pci_set_word(config + PCI_DEVICE_ID,
+                     0x1040 + virtio_bus_get_vdev_id(bus));
+        pci_config_set_revision(config, 1);
+    }
     config[PCI_INTERRUPT_PIN] = 1;
 
+
+    if (modern) {
+        struct virtio_pci_cap cap = {
+            .cap_len = sizeof cap,
+        };
+        struct virtio_pci_notify_cap notify = {
+            .cap.cap_len = sizeof notify,
+            .notify_off_multiplier =
+                cpu_to_le32(QEMU_VIRTIO_PCI_QUEUE_MEM_MULT),
+        };
+
+        /* TODO: add io access for speed */
+
+        virtio_add_feature(&vdev->host_features, VIRTIO_F_VERSION_1);
+        virtio_pci_modern_regions_init(proxy);
+        virtio_pci_modern_region_map(proxy, &proxy->common, &cap);
+        virtio_pci_modern_region_map(proxy, &proxy->isr, &cap);
+        virtio_pci_modern_region_map(proxy, &proxy->device, &cap);
+        virtio_pci_modern_region_map(proxy, &proxy->notify, &notify.cap);
+        pci_register_bar(&proxy->pci_dev, proxy->modern_mem_bar,
+                         PCI_BASE_ADDRESS_SPACE_MEMORY |
+                         PCI_BASE_ADDRESS_MEM_PREFETCH |
+                         PCI_BASE_ADDRESS_MEM_TYPE_64,
+                         &proxy->modern_bar);
+    }
+
     if (proxy->nvectors &&
-        msix_init_exclusive_bar(&proxy->pci_dev, proxy->nvectors, 1)) {
+        msix_init_exclusive_bar(&proxy->pci_dev, proxy->nvectors,
+                                proxy->msix_bar)) {
         error_report("unable to init msix vectors to %" PRIu32,
                      proxy->nvectors);
         proxy->nvectors = 0;
@@ -945,16 +1355,20 @@
 
     proxy->pci_dev.config_write = virtio_write_config;
 
-    size = VIRTIO_PCI_REGION_SIZE(&proxy->pci_dev)
-         + virtio_bus_get_vdev_config_len(bus);
-    if (size & (size - 1)) {
-        size = 1 << qemu_fls(size);
-    }
+    if (legacy) {
+        size = VIRTIO_PCI_REGION_SIZE(&proxy->pci_dev)
+            + virtio_bus_get_vdev_config_len(bus);
+        if (size & (size - 1)) {
+            size = 1 << qemu_fls(size);
+        }
 
-    memory_region_init_io(&proxy->bar, OBJECT(proxy), &virtio_pci_config_ops,
-                          proxy, "virtio-pci", size);
-    pci_register_bar(&proxy->pci_dev, 0, PCI_BASE_ADDRESS_SPACE_IO,
-                     &proxy->bar);
+        memory_region_init_io(&proxy->bar, OBJECT(proxy),
+                              &virtio_pci_config_ops,
+                              proxy, "virtio-pci", size);
+
+        pci_register_bar(&proxy->pci_dev, proxy->legacy_io_bar,
+                         PCI_BASE_ADDRESS_SPACE_IO, &proxy->bar);
+    }
 
     if (!kvm_has_many_ioeventfds()) {
         proxy->flags &= ~VIRTIO_PCI_FLAG_USE_IOEVENTFD;
@@ -972,12 +1386,47 @@
 
 static void virtio_pci_realize(PCIDevice *pci_dev, Error **errp)
 {
-    VirtIOPCIProxy *dev = VIRTIO_PCI(pci_dev);
+    VirtIOPCIProxy *proxy = VIRTIO_PCI(pci_dev);
     VirtioPCIClass *k = VIRTIO_PCI_GET_CLASS(pci_dev);
 
-    virtio_pci_bus_new(&dev->bus, sizeof(dev->bus), dev);
+    /*
+     * virtio pci bar layout used by default.
+     * subclasses can re-arrange things if needed.
+     *
+     *   region 0   --  virtio legacy io bar
+     *   region 1   --  msi-x bar
+     *   region 4+5 --  virtio modern memory (64bit) bar
+     *
+     */
+    proxy->legacy_io_bar  = 0;
+    proxy->msix_bar       = 1;
+    proxy->modern_mem_bar = 4;
+
+    proxy->common.offset = 0x0;
+    proxy->common.size = 0x1000;
+    proxy->common.type = VIRTIO_PCI_CAP_COMMON_CFG;
+
+    proxy->isr.offset = 0x1000;
+    proxy->isr.size = 0x1000;
+    proxy->isr.type = VIRTIO_PCI_CAP_ISR_CFG;
+
+    proxy->device.offset = 0x2000;
+    proxy->device.size = 0x1000;
+    proxy->device.type = VIRTIO_PCI_CAP_DEVICE_CFG;
+
+    proxy->notify.offset = 0x3000;
+    proxy->notify.size =
+        QEMU_VIRTIO_PCI_QUEUE_MEM_MULT * VIRTIO_QUEUE_MAX;
+    proxy->notify.type = VIRTIO_PCI_CAP_NOTIFY_CFG;
+
+    /* subclasses can enforce modern, so do this unconditionally */
+    memory_region_init(&proxy->modern_bar, OBJECT(proxy), "virtio-pci",
+                       2 * QEMU_VIRTIO_PCI_QUEUE_MEM_MULT *
+                       VIRTIO_QUEUE_MAX);
+
+    virtio_pci_bus_new(&proxy->bus, sizeof(proxy->bus), proxy);
     if (k->realize) {
-        k->realize(dev, errp);
+        k->realize(proxy, errp);
     }
 }
 
@@ -998,6 +1447,10 @@
 static Property virtio_pci_properties[] = {
     DEFINE_PROP_BIT("virtio-pci-bus-master-bug-migration", VirtIOPCIProxy, flags,
                     VIRTIO_PCI_FLAG_BUS_MASTER_BUG_MIGRATION_BIT, false),
+    DEFINE_PROP_BIT("disable-legacy", VirtIOPCIProxy, flags,
+                    VIRTIO_PCI_FLAG_DISABLE_LEGACY_BIT, false),
+    DEFINE_PROP_BIT("disable-modern", VirtIOPCIProxy, flags,
+                    VIRTIO_PCI_FLAG_DISABLE_MODERN_BIT, true),
     DEFINE_PROP_END_OF_LIST(),
 };
 
@@ -1445,6 +1898,120 @@
     .class_init    = virtio_rng_pci_class_init,
 };
 
+/* virtio-input-pci */
+
+static Property virtio_input_hid_pci_properties[] = {
+    DEFINE_VIRTIO_INPUT_PROPERTIES(VirtIOInputPCI, vdev.input),
+    DEFINE_PROP_UINT32("vectors", VirtIOPCIProxy, nvectors, 2),
+    DEFINE_PROP_END_OF_LIST(),
+};
+
+static void virtio_input_pci_realize(VirtIOPCIProxy *vpci_dev, Error **errp)
+{
+    VirtIOInputPCI *vinput = VIRTIO_INPUT_PCI(vpci_dev);
+    DeviceState *vdev = DEVICE(&vinput->vdev);
+
+    qdev_set_parent_bus(vdev, BUS(&vpci_dev->bus));
+    /* force virtio-1.0 */
+    vpci_dev->flags &= ~VIRTIO_PCI_FLAG_DISABLE_MODERN;
+    vpci_dev->flags |= VIRTIO_PCI_FLAG_DISABLE_LEGACY;
+    object_property_set_bool(OBJECT(vdev), true, "realized", errp);
+}
+
+static void virtio_input_pci_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+    VirtioPCIClass *k = VIRTIO_PCI_CLASS(klass);
+    PCIDeviceClass *pcidev_k = PCI_DEVICE_CLASS(klass);
+
+    k->realize = virtio_input_pci_realize;
+    set_bit(DEVICE_CATEGORY_INPUT, dc->categories);
+
+    pcidev_k->class_id = PCI_CLASS_INPUT_OTHER;
+}
+
+static void virtio_input_hid_pci_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+
+    dc->props = virtio_input_hid_pci_properties;
+}
+
+static void virtio_input_hid_kbd_pci_class_init(ObjectClass *klass, void *data)
+{
+    PCIDeviceClass *pcidev_k = PCI_DEVICE_CLASS(klass);
+
+    pcidev_k->class_id = PCI_CLASS_INPUT_KEYBOARD;
+}
+
+static void virtio_input_hid_mouse_pci_class_init(ObjectClass *klass,
+                                                  void *data)
+{
+    PCIDeviceClass *pcidev_k = PCI_DEVICE_CLASS(klass);
+
+    pcidev_k->class_id = PCI_CLASS_INPUT_MOUSE;
+}
+
+static void virtio_keyboard_initfn(Object *obj)
+{
+    VirtIOInputHIDPCI *dev = VIRTIO_INPUT_HID_PCI(obj);
+    object_initialize(&dev->vdev, sizeof(dev->vdev), TYPE_VIRTIO_KEYBOARD);
+    object_property_add_child(obj, "virtio-backend", OBJECT(&dev->vdev), NULL);
+}
+
+static void virtio_mouse_initfn(Object *obj)
+{
+    VirtIOInputHIDPCI *dev = VIRTIO_INPUT_HID_PCI(obj);
+    object_initialize(&dev->vdev, sizeof(dev->vdev), TYPE_VIRTIO_MOUSE);
+    object_property_add_child(obj, "virtio-backend", OBJECT(&dev->vdev), NULL);
+}
+
+static void virtio_tablet_initfn(Object *obj)
+{
+    VirtIOInputHIDPCI *dev = VIRTIO_INPUT_HID_PCI(obj);
+    object_initialize(&dev->vdev, sizeof(dev->vdev), TYPE_VIRTIO_TABLET);
+    object_property_add_child(obj, "virtio-backend", OBJECT(&dev->vdev), NULL);
+}
+
+static const TypeInfo virtio_input_pci_info = {
+    .name          = TYPE_VIRTIO_INPUT_PCI,
+    .parent        = TYPE_VIRTIO_PCI,
+    .instance_size = sizeof(VirtIOInputPCI),
+    .class_init    = virtio_input_pci_class_init,
+    .abstract      = true,
+};
+
+static const TypeInfo virtio_input_hid_pci_info = {
+    .name          = TYPE_VIRTIO_INPUT_HID_PCI,
+    .parent        = TYPE_VIRTIO_INPUT_PCI,
+    .instance_size = sizeof(VirtIOInputHIDPCI),
+    .class_init    = virtio_input_hid_pci_class_init,
+    .abstract      = true,
+};
+
+static const TypeInfo virtio_keyboard_pci_info = {
+    .name          = TYPE_VIRTIO_KEYBOARD_PCI,
+    .parent        = TYPE_VIRTIO_INPUT_HID_PCI,
+    .class_init    = virtio_input_hid_kbd_pci_class_init,
+    .instance_size = sizeof(VirtIOInputHIDPCI),
+    .instance_init = virtio_keyboard_initfn,
+};
+
+static const TypeInfo virtio_mouse_pci_info = {
+    .name          = TYPE_VIRTIO_MOUSE_PCI,
+    .parent        = TYPE_VIRTIO_INPUT_HID_PCI,
+    .class_init    = virtio_input_hid_mouse_pci_class_init,
+    .instance_size = sizeof(VirtIOInputHIDPCI),
+    .instance_init = virtio_mouse_initfn,
+};
+
+static const TypeInfo virtio_tablet_pci_info = {
+    .name          = TYPE_VIRTIO_TABLET_PCI,
+    .parent        = TYPE_VIRTIO_INPUT_HID_PCI,
+    .instance_size = sizeof(VirtIOInputHIDPCI),
+    .instance_init = virtio_tablet_initfn,
+};
+
 /* virtio-pci-bus */
 
 static void virtio_pci_bus_new(VirtioBusState *bus, size_t bus_size,
@@ -1486,6 +2053,11 @@
 static void virtio_pci_register_types(void)
 {
     type_register_static(&virtio_rng_pci_info);
+    type_register_static(&virtio_input_pci_info);
+    type_register_static(&virtio_input_hid_pci_info);
+    type_register_static(&virtio_keyboard_pci_info);
+    type_register_static(&virtio_mouse_pci_info);
+    type_register_static(&virtio_tablet_pci_info);
     type_register_static(&virtio_pci_bus_info);
     type_register_static(&virtio_pci_info);
 #ifdef CONFIG_VIRTFS
diff --git a/hw/virtio/virtio-pci.h b/hw/virtio/virtio-pci.h
index de39468..d962125 100644
--- a/hw/virtio/virtio-pci.h
+++ b/hw/virtio/virtio-pci.h
@@ -24,6 +24,7 @@
 #include "hw/virtio/virtio-balloon.h"
 #include "hw/virtio/virtio-bus.h"
 #include "hw/virtio/virtio-9p.h"
+#include "hw/virtio/virtio-input.h"
 #ifdef CONFIG_VIRTFS
 #include "hw/9pfs/virtio-9p.h"
 #endif
@@ -39,6 +40,8 @@
 typedef struct VirtIONetPCI VirtIONetPCI;
 typedef struct VHostSCSIPCI VHostSCSIPCI;
 typedef struct VirtIORngPCI VirtIORngPCI;
+typedef struct VirtIOInputPCI VirtIOInputPCI;
+typedef struct VirtIOInputHIDPCI VirtIOInputHIDPCI;
 
 /* virtio-pci-bus */
 
@@ -63,6 +66,12 @@
 #define VIRTIO_PCI_FLAG_USE_IOEVENTFD_BIT 1
 #define VIRTIO_PCI_FLAG_USE_IOEVENTFD   (1 << VIRTIO_PCI_FLAG_USE_IOEVENTFD_BIT)
 
+/* virtio version flags */
+#define VIRTIO_PCI_FLAG_DISABLE_LEGACY_BIT 2
+#define VIRTIO_PCI_FLAG_DISABLE_MODERN_BIT 3
+#define VIRTIO_PCI_FLAG_DISABLE_LEGACY (1 << VIRTIO_PCI_FLAG_DISABLE_LEGACY_BIT)
+#define VIRTIO_PCI_FLAG_DISABLE_MODERN (1 << VIRTIO_PCI_FLAG_DISABLE_MODERN_BIT)
+
 typedef struct {
     MSIMessage msg;
     int virq;
@@ -85,12 +94,38 @@
     void (*realize)(VirtIOPCIProxy *vpci_dev, Error **errp);
 } VirtioPCIClass;
 
+typedef struct VirtIOPCIRegion {
+    MemoryRegion mr;
+    uint32_t offset;
+    uint32_t size;
+    uint32_t type;
+} VirtIOPCIRegion;
+
 struct VirtIOPCIProxy {
     PCIDevice pci_dev;
     MemoryRegion bar;
+    VirtIOPCIRegion common;
+    VirtIOPCIRegion isr;
+    VirtIOPCIRegion device;
+    VirtIOPCIRegion notify;
+    MemoryRegion modern_bar;
+    uint32_t legacy_io_bar;
+    uint32_t msix_bar;
+    uint32_t modern_mem_bar;
     uint32_t flags;
     uint32_t class_code;
     uint32_t nvectors;
+    uint32_t dfselect;
+    uint32_t gfselect;
+    uint32_t guest_features[2];
+    struct {
+        uint16_t num;
+        bool enabled;
+        uint32_t desc[2];
+        uint32_t avail[2];
+        uint32_t used[2];
+    } vqs[VIRTIO_QUEUE_MAX];
+
     bool ioeventfd_disabled;
     bool ioeventfd_started;
     VirtIOIRQFD *vector_irqfd;
@@ -202,6 +237,30 @@
     VirtIORNG vdev;
 };
 
+/*
+ * virtio-input-pci: This extends VirtioPCIProxy.
+ */
+#define TYPE_VIRTIO_INPUT_PCI "virtio-input-pci"
+#define VIRTIO_INPUT_PCI(obj) \
+        OBJECT_CHECK(VirtIOInputPCI, (obj), TYPE_VIRTIO_INPUT_PCI)
+
+struct VirtIOInputPCI {
+    VirtIOPCIProxy parent_obj;
+    VirtIOInput vdev;
+};
+
+#define TYPE_VIRTIO_INPUT_HID_PCI "virtio-input-hid-pci"
+#define TYPE_VIRTIO_KEYBOARD_PCI  "virtio-keyboard-pci"
+#define TYPE_VIRTIO_MOUSE_PCI     "virtio-mouse-pci"
+#define TYPE_VIRTIO_TABLET_PCI    "virtio-tablet-pci"
+#define VIRTIO_INPUT_HID_PCI(obj) \
+        OBJECT_CHECK(VirtIOInputHIDPCI, (obj), TYPE_VIRTIO_INPUT_HID_PCI)
+
+struct VirtIOInputHIDPCI {
+    VirtIOPCIProxy parent_obj;
+    VirtIOInputHID vdev;
+};
+
 /* Virtio ABI version, if we increment this, we break the guest driver. */
 #define VIRTIO_PCI_ABI_VERSION          0
 
diff --git a/hw/virtio/virtio-rng.c b/hw/virtio/virtio-rng.c
index 420c39f..22b1d87 100644
--- a/hw/virtio/virtio-rng.c
+++ b/hw/virtio/virtio-rng.c
@@ -219,7 +219,13 @@
 }
 
 static Property virtio_rng_properties[] = {
-    DEFINE_VIRTIO_RNG_PROPERTIES(VirtIORNG, conf),
+    /* Set a default rate limit of 2^47 bytes per minute or roughly 2TB/s.  If
+     * you have an entropy source capable of generating more entropy than this
+     * and you can pass it through via virtio-rng, then hats off to you.  Until
+     * then, this is unlimited for all practical purposes.
+     */
+    DEFINE_PROP_UINT64("max-bytes", VirtIORNG, conf.max_bytes, INT64_MAX),
+    DEFINE_PROP_UINT32("period", VirtIORNG, conf.period_ms, 1 << 16),
     DEFINE_PROP_END_OF_LIST(),
 };
 
diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index 8ac6156..ee4e07c 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -69,7 +69,6 @@
 struct VirtQueue
 {
     VRing vring;
-    hwaddr pa;
     uint16_t last_avail_idx;
     /* Last used index value we have signalled on */
     uint16_t signalled_used;
@@ -93,15 +92,18 @@
 };
 
 /* virt queue functions */
-static void virtqueue_init(VirtQueue *vq)
+void virtio_queue_update_rings(VirtIODevice *vdev, int n)
 {
-    hwaddr pa = vq->pa;
+    VRing *vring = &vdev->vq[n].vring;
 
-    vq->vring.desc = pa;
-    vq->vring.avail = pa + vq->vring.num * sizeof(VRingDesc);
-    vq->vring.used = vring_align(vq->vring.avail +
-                                 offsetof(VRingAvail, ring[vq->vring.num]),
-                                 vq->vring.align);
+    if (!vring->desc) {
+        /* not yet setup -> nothing to do */
+        return;
+    }
+    vring->avail = vring->desc + vring->num * sizeof(VRingDesc);
+    vring->used = vring_align(vring->avail +
+                              offsetof(VRingAvail, ring[vring->num]),
+                              vring->align);
 }
 
 static inline uint64_t vring_desc_addr(VirtIODevice *vdev, hwaddr desc_pa,
@@ -542,15 +544,37 @@
     virtio_notify_vector(vdev, VIRTIO_NO_VECTOR);
 }
 
-void virtio_set_status(VirtIODevice *vdev, uint8_t val)
+static int virtio_validate_features(VirtIODevice *vdev)
+{
+    VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
+
+    if (k->validate_features) {
+        return k->validate_features(vdev);
+    } else {
+        return 0;
+    }
+}
+
+int virtio_set_status(VirtIODevice *vdev, uint8_t val)
 {
     VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
     trace_virtio_set_status(vdev, val);
 
+    if (virtio_has_feature(vdev, VIRTIO_F_VERSION_1)) {
+        if (!(vdev->status & VIRTIO_CONFIG_S_FEATURES_OK) &&
+            val & VIRTIO_CONFIG_S_FEATURES_OK) {
+            int ret = virtio_validate_features(vdev);
+
+            if (ret) {
+                return ret;
+            }
+        }
+    }
     if (k->set_status) {
         k->set_status(vdev, val);
     }
     vdev->status = val;
+    return 0;
 }
 
 bool target_words_bigendian(void);
@@ -605,7 +629,6 @@
         vdev->vq[i].vring.avail = 0;
         vdev->vq[i].vring.used = 0;
         vdev->vq[i].last_avail_idx = 0;
-        vdev->vq[i].pa = 0;
         virtio_queue_set_vector(vdev, i, VIRTIO_NO_VECTOR);
         vdev->vq[i].signalled_used = 0;
         vdev->vq[i].signalled_used_valid = false;
@@ -706,15 +729,119 @@
     }
 }
 
+uint32_t virtio_config_modern_readb(VirtIODevice *vdev, uint32_t addr)
+{
+    VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
+    uint8_t val;
+
+    if (addr + sizeof(val) > vdev->config_len) {
+        return (uint32_t)-1;
+    }
+
+    k->get_config(vdev, vdev->config);
+
+    val = ldub_p(vdev->config + addr);
+    return val;
+}
+
+uint32_t virtio_config_modern_readw(VirtIODevice *vdev, uint32_t addr)
+{
+    VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
+    uint16_t val;
+
+    if (addr + sizeof(val) > vdev->config_len) {
+        return (uint32_t)-1;
+    }
+
+    k->get_config(vdev, vdev->config);
+
+    val = lduw_le_p(vdev->config + addr);
+    return val;
+}
+
+uint32_t virtio_config_modern_readl(VirtIODevice *vdev, uint32_t addr)
+{
+    VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
+    uint32_t val;
+
+    if (addr + sizeof(val) > vdev->config_len) {
+        return (uint32_t)-1;
+    }
+
+    k->get_config(vdev, vdev->config);
+
+    val = ldl_le_p(vdev->config + addr);
+    return val;
+}
+
+void virtio_config_modern_writeb(VirtIODevice *vdev,
+                                 uint32_t addr, uint32_t data)
+{
+    VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
+    uint8_t val = data;
+
+    if (addr + sizeof(val) > vdev->config_len) {
+        return;
+    }
+
+    stb_p(vdev->config + addr, val);
+
+    if (k->set_config) {
+        k->set_config(vdev, vdev->config);
+    }
+}
+
+void virtio_config_modern_writew(VirtIODevice *vdev,
+                                 uint32_t addr, uint32_t data)
+{
+    VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
+    uint16_t val = data;
+
+    if (addr + sizeof(val) > vdev->config_len) {
+        return;
+    }
+
+    stw_le_p(vdev->config + addr, val);
+
+    if (k->set_config) {
+        k->set_config(vdev, vdev->config);
+    }
+}
+
+void virtio_config_modern_writel(VirtIODevice *vdev,
+                                 uint32_t addr, uint32_t data)
+{
+    VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
+    uint32_t val = data;
+
+    if (addr + sizeof(val) > vdev->config_len) {
+        return;
+    }
+
+    stl_le_p(vdev->config + addr, val);
+
+    if (k->set_config) {
+        k->set_config(vdev, vdev->config);
+    }
+}
+
 void virtio_queue_set_addr(VirtIODevice *vdev, int n, hwaddr addr)
 {
-    vdev->vq[n].pa = addr;
-    virtqueue_init(&vdev->vq[n]);
+    vdev->vq[n].vring.desc = addr;
+    virtio_queue_update_rings(vdev, n);
 }
 
 hwaddr virtio_queue_get_addr(VirtIODevice *vdev, int n)
 {
-    return vdev->vq[n].pa;
+    return vdev->vq[n].vring.desc;
+}
+
+void virtio_queue_set_rings(VirtIODevice *vdev, int n, hwaddr desc,
+                            hwaddr avail, hwaddr used)
+{
+    vdev->vq[n].vring.desc = desc;
+    vdev->vq[n].vring.avail = avail;
+    vdev->vq[n].vring.used = used;
 }
 
 void virtio_queue_set_num(VirtIODevice *vdev, int n, int num)
@@ -728,7 +855,6 @@
         return;
     }
     vdev->vq[n].vring.num = num;
-    virtqueue_init(&vdev->vq[n]);
 }
 
 VirtQueue *virtio_vector_first_queue(VirtIODevice *vdev, uint16_t vector)
@@ -771,6 +897,11 @@
     BusState *qbus = qdev_get_parent_bus(DEVICE(vdev));
     VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
 
+    /* virtio-1 compliant devices cannot change the alignment */
+    if (virtio_has_feature(vdev, VIRTIO_F_VERSION_1)) {
+        error_report("tried to modify queue alignment for virtio-1 device");
+        return;
+    }
     /* Check that the transport told us it was going to do this
      * (so a buggy transport will immediately assert rather than
      * silently failing to migrate this state)
@@ -778,7 +909,7 @@
     assert(k->has_variable_vring_alignment);
 
     vdev->vq[n].vring.align = align;
-    virtqueue_init(&vdev->vq[n]);
+    virtio_queue_update_rings(vdev, n);
 }
 
 void virtio_queue_notify_vq(VirtQueue *vq)
@@ -895,6 +1026,7 @@
         return;
 
     vdev->isr |= 0x03;
+    vdev->generation++;
     virtio_notify_vector(vdev, vdev->config_vector);
 }
 
@@ -903,7 +1035,11 @@
     VirtIODevice *vdev = opaque;
 
     assert(vdev->device_endian != VIRTIO_DEVICE_ENDIAN_UNKNOWN);
-    return vdev->device_endian != virtio_default_endian();
+    if (!virtio_has_feature(vdev, VIRTIO_F_VERSION_1)) {
+        return vdev->device_endian != virtio_default_endian();
+    }
+    /* Devices conforming to VIRTIO 1.0 or later are always LE. */
+    return vdev->device_endian != VIRTIO_DEVICE_ENDIAN_LITTLE;
 }
 
 static bool virtio_64bit_features_needed(void *opaque)
@@ -917,6 +1053,7 @@
     .name = "virtio/device_endian",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = &virtio_device_endian_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT8(device_endian, VirtIODevice),
         VMSTATE_END_OF_LIST()
@@ -927,6 +1064,7 @@
     .name = "virtio/64bit_features",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = &virtio_64bit_features_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64(guest_features, VirtIODevice),
         VMSTATE_END_OF_LIST()
@@ -941,16 +1079,10 @@
     .fields = (VMStateField[]) {
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_virtio_device_endian,
-            .needed = &virtio_device_endian_needed
-        },
-        {
-            .vmsd = &vmstate_virtio_64bit_features,
-            .needed = &virtio_64bit_features_needed
-        },
-        { 0 }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_virtio_device_endian,
+        &vmstate_virtio_64bit_features,
+        NULL
     }
 };
 
@@ -988,7 +1120,8 @@
         if (k->has_variable_vring_alignment) {
             qemu_put_be32(f, vdev->vq[i].vring.align);
         }
-        qemu_put_be64(f, vdev->vq[i].pa);
+        /* XXX virtio-1 devices */
+        qemu_put_be64(f, vdev->vq[i].vring.desc);
         qemu_put_be16s(f, &vdev->vq[i].last_avail_idx);
         if (k->save_queue) {
             k->save_queue(qbus->parent, i, f);
@@ -1003,7 +1136,7 @@
     vmstate_save_state(f, &vmstate_virtio, vdev, NULL);
 }
 
-int virtio_set_features(VirtIODevice *vdev, uint64_t val)
+static int virtio_set_features_nocheck(VirtIODevice *vdev, uint64_t val)
 {
     VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
     bool bad = (val & ~(vdev->host_features)) != 0;
@@ -1016,6 +1149,18 @@
     return bad ? -1 : 0;
 }
 
+int virtio_set_features(VirtIODevice *vdev, uint64_t val)
+{
+   /*
+     * The driver must not attempt to set features after feature negotiation
+     * has finished.
+     */
+    if (vdev->status & VIRTIO_CONFIG_S_FEATURES_OK) {
+        return -EINVAL;
+    }
+    return virtio_set_features_nocheck(vdev, val);
+}
+
 int virtio_load(VirtIODevice *vdev, QEMUFile *f, int version_id)
 {
     int i, ret;
@@ -1072,13 +1217,14 @@
         if (k->has_variable_vring_alignment) {
             vdev->vq[i].vring.align = qemu_get_be32(f);
         }
-        vdev->vq[i].pa = qemu_get_be64(f);
+        vdev->vq[i].vring.desc = qemu_get_be64(f);
         qemu_get_be16s(f, &vdev->vq[i].last_avail_idx);
         vdev->vq[i].signalled_used_valid = false;
         vdev->vq[i].notification = true;
 
-        if (vdev->vq[i].pa) {
-            virtqueue_init(&vdev->vq[i]);
+        if (vdev->vq[i].vring.desc) {
+            /* XXX virtio-1 devices */
+            virtio_queue_update_rings(vdev, i);
         } else if (vdev->vq[i].last_avail_idx) {
             error_report("VQ %d address 0x0 "
                          "inconsistent with Host index 0x%x",
@@ -1118,14 +1264,14 @@
          * host_features.
          */
         uint64_t features64 = vdev->guest_features;
-        if (virtio_set_features(vdev, features64) < 0) {
+        if (virtio_set_features_nocheck(vdev, features64) < 0) {
             error_report("Features 0x%" PRIx64 " unsupported. "
                          "Allowed features: 0x%" PRIx64,
                          features64, vdev->host_features);
             return -1;
         }
     } else {
-        if (virtio_set_features(vdev, features) < 0) {
+        if (virtio_set_features_nocheck(vdev, features) < 0) {
             error_report("Features 0x%x unsupported. "
                          "Allowed features: 0x%" PRIx64,
                          features, vdev->host_features);
@@ -1134,7 +1280,7 @@
     }
 
     for (i = 0; i < num; i++) {
-        if (vdev->vq[i].pa) {
+        if (vdev->vq[i].vring.desc) {
             uint16_t nheads;
             nheads = vring_avail_idx(&vdev->vq[i]) - vdev->vq[i].last_avail_idx;
             /* Check it isn't doing strange things with descriptor numbers. */
diff --git a/include/exec/cpu-common.h b/include/exec/cpu-common.h
index 43428bd..de8a720 100644
--- a/include/exec/cpu-common.h
+++ b/include/exec/cpu-common.h
@@ -126,10 +126,10 @@
 extern struct MemoryRegion io_mem_rom;
 extern struct MemoryRegion io_mem_notdirty;
 
-typedef void (RAMBlockIterFunc)(void *host_addr,
+typedef int (RAMBlockIterFunc)(const char *block_name, void *host_addr,
     ram_addr_t offset, ram_addr_t length, void *opaque);
 
-void qemu_ram_foreach_block(RAMBlockIterFunc func, void *opaque);
+int qemu_ram_foreach_block(RAMBlockIterFunc func, void *opaque);
 
 #endif
 
diff --git a/include/exec/exec-all.h b/include/exec/exec-all.h
index 2f7a4f1..2573e8c 100644
--- a/include/exec/exec-all.h
+++ b/include/exec/exec-all.h
@@ -105,6 +105,8 @@
                              hwaddr paddr, MemTxAttrs attrs,
                              int prot, int mmu_idx, target_ulong size);
 void tb_invalidate_phys_addr(AddressSpace *as, hwaddr addr);
+void probe_write(CPUArchState *env, target_ulong addr, int mmu_idx,
+                 uintptr_t retaddr);
 #else
 static inline void tlb_flush_page(CPUState *cpu, target_ulong addr)
 {
diff --git a/include/hw/mips/mips.h b/include/hw/mips/mips.h
index 2a7a9c9..e0065ce 100644
--- a/include/hw/mips/mips.h
+++ b/include/hw/mips/mips.h
@@ -15,18 +15,9 @@
 
 /* rc4030.c */
 typedef struct rc4030DMAState *rc4030_dma;
-void rc4030_dma_memory_rw(void *opaque, hwaddr addr, uint8_t *buf, int len, int is_write);
 void rc4030_dma_read(void *dma, uint8_t *buf, int len);
 void rc4030_dma_write(void *dma, uint8_t *buf, int len);
 
-void *rc4030_init(qemu_irq timer, qemu_irq jazz_bus,
-                  qemu_irq **irqs, rc4030_dma **dmas,
-                  MemoryRegion *sysmem);
-
-/* dp8393x.c */
-void dp83932_init(NICInfo *nd, hwaddr base, int it_shift,
-                  MemoryRegion *address_space,
-                  qemu_irq irq, void* mem_opaque,
-                  void (*memory_rw)(void *opaque, hwaddr addr, uint8_t *buf, int len, int is_write));
+DeviceState *rc4030_init(rc4030_dma **dmas, MemoryRegion **dma_mr);
 
 #endif
diff --git a/include/hw/nvram/fw_cfg.h b/include/hw/nvram/fw_cfg.h
index 6d8a8ac..e60d3ca 100644
--- a/include/hw/nvram/fw_cfg.h
+++ b/include/hw/nvram/fw_cfg.h
@@ -67,10 +67,9 @@
 void fw_cfg_add_bytes(FWCfgState *s, uint16_t key, void *data, size_t len);
 void fw_cfg_add_string(FWCfgState *s, uint16_t key, const char *value);
 void fw_cfg_add_i16(FWCfgState *s, uint16_t key, uint16_t value);
+void fw_cfg_modify_i16(FWCfgState *s, uint16_t key, uint16_t value);
 void fw_cfg_add_i32(FWCfgState *s, uint16_t key, uint32_t value);
 void fw_cfg_add_i64(FWCfgState *s, uint16_t key, uint64_t value);
-void fw_cfg_add_callback(FWCfgState *s, uint16_t key, FWCfgCallback callback,
-                         void *callback_opaque, void *data, size_t len);
 void fw_cfg_add_file(FWCfgState *s, const char *filename, void *data,
                      size_t len);
 void fw_cfg_add_file_callback(FWCfgState *s, const char *filename,
diff --git a/include/hw/pci/pci_ids.h b/include/hw/pci/pci_ids.h
index c6de710..49c062b 100644
--- a/include/hw/pci/pci_ids.h
+++ b/include/hw/pci/pci_ids.h
@@ -47,6 +47,13 @@
 #define PCI_CLASS_COMMUNICATION_SERIAL   0x0700
 #define PCI_CLASS_COMMUNICATION_OTHER    0x0780
 
+#define PCI_CLASS_INPUT_KEYBOARD         0x0900
+#define PCI_CLASS_INPUT_PEN              0x0901
+#define PCI_CLASS_INPUT_MOUSE            0x0902
+#define PCI_CLASS_INPUT_SCANNER          0x0903
+#define PCI_CLASS_INPUT_GAMEPORT         0x0904
+#define PCI_CLASS_INPUT_OTHER            0x0980
+
 #define PCI_CLASS_PROCESSOR_CO           0x0b40
 #define PCI_CLASS_PROCESSOR_POWERPC      0x0b20
 
diff --git a/include/hw/vfio/vfio-calxeda-xgmac.h b/include/hw/vfio/vfio-calxeda-xgmac.h
new file mode 100644
index 0000000..f994775
--- /dev/null
+++ b/include/hw/vfio/vfio-calxeda-xgmac.h
@@ -0,0 +1,46 @@
+/*
+ * VFIO calxeda xgmac device
+ *
+ * Copyright Linaro Limited, 2014
+ *
+ * Authors:
+ *  Eric Auger <eric.auger@linaro.org>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.  See
+ * the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef HW_VFIO_VFIO_CALXEDA_XGMAC_H
+#define HW_VFIO_VFIO_CALXEDA_XGMAC_H
+
+#include "hw/vfio/vfio-platform.h"
+
+#define TYPE_VFIO_CALXEDA_XGMAC "vfio-calxeda-xgmac"
+
+/**
+ * This device exposes:
+ * - a single MMIO region corresponding to its register space
+ * - 3 IRQS (main and 2 power related IRQs)
+ */
+typedef struct VFIOCalxedaXgmacDevice {
+    VFIOPlatformDevice vdev;
+} VFIOCalxedaXgmacDevice;
+
+typedef struct VFIOCalxedaXgmacDeviceClass {
+    /*< private >*/
+    VFIOPlatformDeviceClass parent_class;
+    /*< public >*/
+    DeviceRealize parent_realize;
+} VFIOCalxedaXgmacDeviceClass;
+
+#define VFIO_CALXEDA_XGMAC_DEVICE(obj) \
+     OBJECT_CHECK(VFIOCalxedaXgmacDevice, (obj), TYPE_VFIO_CALXEDA_XGMAC)
+#define VFIO_CALXEDA_XGMAC_DEVICE_CLASS(klass) \
+     OBJECT_CLASS_CHECK(VFIOCalxedaXgmacDeviceClass, (klass), \
+                        TYPE_VFIO_CALXEDA_XGMAC)
+#define VFIO_CALXEDA_XGMAC_DEVICE_GET_CLASS(obj) \
+     OBJECT_GET_CLASS(VFIOCalxedaXgmacDeviceClass, (obj), \
+                      TYPE_VFIO_CALXEDA_XGMAC)
+
+#endif
diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h
index 0d1fb80..59a321d 100644
--- a/include/hw/vfio/vfio-common.h
+++ b/include/hw/vfio/vfio-common.h
@@ -42,6 +42,7 @@
 
 enum {
     VFIO_DEVICE_TYPE_PCI = 0,
+    VFIO_DEVICE_TYPE_PLATFORM = 1,
 };
 
 typedef struct VFIORegion {
diff --git a/include/hw/vfio/vfio-platform.h b/include/hw/vfio/vfio-platform.h
new file mode 100644
index 0000000..26b2ad6
--- /dev/null
+++ b/include/hw/vfio/vfio-platform.h
@@ -0,0 +1,75 @@
+/*
+ * vfio based device assignment support - platform devices
+ *
+ * Copyright Linaro Limited, 2014
+ *
+ * Authors:
+ *  Kim Phillips <kim.phillips@linaro.org>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.  See
+ * the COPYING file in the top-level directory.
+ *
+ * Based on vfio based PCI device assignment support:
+ *  Copyright Red Hat, Inc. 2012
+ */
+
+#ifndef HW_VFIO_VFIO_PLATFORM_H
+#define HW_VFIO_VFIO_PLATFORM_H
+
+#include "hw/sysbus.h"
+#include "hw/vfio/vfio-common.h"
+#include "qemu/event_notifier.h"
+#include "qemu/queue.h"
+#include "hw/irq.h"
+
+#define TYPE_VFIO_PLATFORM "vfio-platform"
+
+enum {
+    VFIO_IRQ_INACTIVE = 0,
+    VFIO_IRQ_PENDING = 1,
+    VFIO_IRQ_ACTIVE = 2,
+    /* VFIO_IRQ_ACTIVE_AND_PENDING cannot happen with VFIO */
+};
+
+typedef struct VFIOINTp {
+    QLIST_ENTRY(VFIOINTp) next; /* entry for IRQ list */
+    QSIMPLEQ_ENTRY(VFIOINTp) pqnext; /* entry for pending IRQ queue */
+    EventNotifier interrupt; /* eventfd triggered on interrupt */
+    EventNotifier unmask; /* eventfd for unmask on QEMU bypass */
+    qemu_irq qemuirq;
+    struct VFIOPlatformDevice *vdev; /* back pointer to device */
+    int state; /* inactive, pending, active */
+    uint8_t pin; /* index */
+    uint32_t flags; /* IRQ info flags */
+} VFIOINTp;
+
+/* function type for user side eventfd handler */
+typedef void (*eventfd_user_side_handler_t)(VFIOINTp *intp);
+
+typedef struct VFIOPlatformDevice {
+    SysBusDevice sbdev;
+    VFIODevice vbasedev; /* not a QOM object */
+    VFIORegion **regions;
+    QLIST_HEAD(, VFIOINTp) intp_list; /* list of IRQs */
+    /* queue of pending IRQs */
+    QSIMPLEQ_HEAD(pending_intp_queue, VFIOINTp) pending_intp_queue;
+    char *compat; /* compatibility string */
+    uint32_t mmap_timeout; /* delay to re-enable mmaps after interrupt */
+    QEMUTimer *mmap_timer; /* allows fast-path resume after IRQ hit */
+    QemuMutex intp_mutex; /* protect the intp_list IRQ state */
+} VFIOPlatformDevice;
+
+typedef struct VFIOPlatformDeviceClass {
+    /*< private >*/
+    SysBusDeviceClass parent_class;
+    /*< public >*/
+} VFIOPlatformDeviceClass;
+
+#define VFIO_PLATFORM_DEVICE(obj) \
+     OBJECT_CHECK(VFIOPlatformDevice, (obj), TYPE_VFIO_PLATFORM)
+#define VFIO_PLATFORM_DEVICE_CLASS(klass) \
+     OBJECT_CLASS_CHECK(VFIOPlatformDeviceClass, (klass), TYPE_VFIO_PLATFORM)
+#define VFIO_PLATFORM_DEVICE_GET_CLASS(obj) \
+     OBJECT_GET_CLASS(VFIOPlatformDeviceClass, (obj), TYPE_VFIO_PLATFORM)
+
+#endif /*HW_VFIO_VFIO_PLATFORM_H*/
diff --git a/include/hw/virtio/vhost-scsi.h b/include/hw/virtio/vhost-scsi.h
index dea0075..701bfee 100644
--- a/include/hw/virtio/vhost-scsi.h
+++ b/include/hw/virtio/vhost-scsi.h
@@ -66,13 +66,4 @@
     int lun;
 } VHostSCSI;
 
-#define DEFINE_VHOST_SCSI_PROPERTIES(_state, _conf_field) \
-    DEFINE_PROP_STRING("vhostfd", _state, _conf_field.vhostfd), \
-    DEFINE_PROP_STRING("wwpn", _state, _conf_field.wwpn), \
-    DEFINE_PROP_UINT32("boot_tpgt", _state, _conf_field.boot_tpgt, 0), \
-    DEFINE_PROP_UINT32("num_queues", _state, _conf_field.num_queues, 1), \
-    DEFINE_PROP_UINT32("max_sectors", _state, _conf_field.max_sectors, 0xFFFF), \
-    DEFINE_PROP_UINT32("cmd_per_lun", _state, _conf_field.cmd_per_lun, 128)
-
-
 #endif
diff --git a/include/hw/virtio/vhost.h b/include/hw/virtio/vhost.h
index 816a2e8..84f170e 100644
--- a/include/hw/virtio/vhost.h
+++ b/include/hw/virtio/vhost.h
@@ -78,8 +78,8 @@
  */
 void vhost_virtqueue_mask(struct vhost_dev *hdev, VirtIODevice *vdev, int n,
                           bool mask);
-unsigned vhost_get_features(struct vhost_dev *hdev, const int *feature_bits,
-        unsigned features);
+uint64_t vhost_get_features(struct vhost_dev *hdev, const int *feature_bits,
+                            uint64_t features);
 void vhost_ack_features(struct vhost_dev *hdev, const int *feature_bits,
-        unsigned features);
+                        uint64_t features);
 #endif
diff --git a/include/hw/virtio/virtio-access.h b/include/hw/virtio/virtio-access.h
index 46456fd..ee28c21 100644
--- a/include/hw/virtio/virtio-access.h
+++ b/include/hw/virtio/virtio-access.h
@@ -19,6 +19,10 @@
 
 static inline bool virtio_access_is_big_endian(VirtIODevice *vdev)
 {
+    if (virtio_has_feature(vdev, VIRTIO_F_VERSION_1)) {
+        /* Devices conforming to VIRTIO 1.0 or later are always LE. */
+        return false;
+    }
 #if defined(TARGET_IS_BIENDIAN)
     return virtio_is_big_endian(vdev);
 #elif defined(TARGET_WORDS_BIGENDIAN)
diff --git a/include/hw/virtio/virtio-balloon.h b/include/hw/virtio/virtio-balloon.h
index 4ab8f54..346a9fd 100644
--- a/include/hw/virtio/virtio-balloon.h
+++ b/include/hw/virtio/virtio-balloon.h
@@ -25,6 +25,12 @@
 
 typedef struct virtio_balloon_stat VirtIOBalloonStat;
 
+typedef struct virtio_balloon_stat_modern {
+       uint16_t tag;
+       uint8_t reserved[6];
+       uint64_t val;
+} VirtIOBalloonStatModern;
+
 typedef struct VirtIOBalloon {
     VirtIODevice parent_obj;
     VirtQueue *ivq, *dvq, *svq;
diff --git a/include/hw/virtio/virtio-gpu.h b/include/hw/virtio/virtio-gpu.h
new file mode 100644
index 0000000..b8c9244
--- /dev/null
+++ b/include/hw/virtio/virtio-gpu.h
@@ -0,0 +1,145 @@
+/*
+ * Virtio GPU Device
+ *
+ * Copyright Red Hat, Inc. 2013-2014
+ *
+ * Authors:
+ *     Dave Airlie <airlied@redhat.com>
+ *     Gerd Hoffmann <kraxel@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.
+ * See the COPYING file in the top-level directory.
+ */
+
+#ifndef _QEMU_VIRTIO_VGA_H
+#define _QEMU_VIRTIO_VGA_H
+
+#include "qemu/queue.h"
+#include "ui/qemu-pixman.h"
+#include "ui/console.h"
+#include "hw/virtio/virtio.h"
+#include "hw/pci/pci.h"
+
+#include "standard-headers/linux/virtio_gpu.h"
+#define TYPE_VIRTIO_GPU "virtio-gpu-device"
+#define VIRTIO_GPU(obj)                                        \
+        OBJECT_CHECK(VirtIOGPU, (obj), TYPE_VIRTIO_GPU)
+
+#define VIRTIO_ID_GPU 16
+
+#define VIRTIO_GPU_MAX_SCANOUT 4
+
+struct virtio_gpu_simple_resource {
+    uint32_t resource_id;
+    uint32_t width;
+    uint32_t height;
+    uint32_t format;
+    struct iovec *iov;
+    unsigned int iov_cnt;
+    uint32_t scanout_bitmask;
+    pixman_image_t *image;
+    QTAILQ_ENTRY(virtio_gpu_simple_resource) next;
+};
+
+struct virtio_gpu_scanout {
+    QemuConsole *con;
+    DisplaySurface *ds;
+    uint32_t width, height;
+    int x, y;
+    int invalidate;
+    uint32_t resource_id;
+    QEMUCursor *current_cursor;
+};
+
+struct virtio_gpu_requested_state {
+    uint32_t width, height;
+    int x, y;
+};
+
+struct virtio_gpu_conf {
+    uint32_t max_outputs;
+};
+
+struct virtio_gpu_ctrl_command {
+    VirtQueueElement elem;
+    VirtQueue *vq;
+    struct virtio_gpu_ctrl_hdr cmd_hdr;
+    uint32_t error;
+    bool finished;
+    QTAILQ_ENTRY(virtio_gpu_ctrl_command) next;
+};
+
+typedef struct VirtIOGPU {
+    VirtIODevice parent_obj;
+
+    QEMUBH *ctrl_bh;
+    QEMUBH *cursor_bh;
+    VirtQueue *ctrl_vq;
+    VirtQueue *cursor_vq;
+
+    int enable;
+
+    int config_size;
+    DeviceState *qdev;
+
+    QTAILQ_HEAD(, virtio_gpu_simple_resource) reslist;
+    QTAILQ_HEAD(, virtio_gpu_ctrl_command) fenceq;
+
+    struct virtio_gpu_scanout scanout[VIRTIO_GPU_MAX_SCANOUT];
+    struct virtio_gpu_requested_state req_state[VIRTIO_GPU_MAX_SCANOUT];
+
+    struct virtio_gpu_conf conf;
+    int enabled_output_bitmask;
+    struct virtio_gpu_config virtio_config;
+
+    QEMUTimer *fence_poll;
+    QEMUTimer *print_stats;
+
+    struct {
+        uint32_t inflight;
+        uint32_t max_inflight;
+        uint32_t requests;
+        uint32_t req_3d;
+        uint32_t bytes_3d;
+    } stats;
+} VirtIOGPU;
+
+extern const GraphicHwOps virtio_gpu_ops;
+
+/* to share between PCI and VGA */
+#define DEFINE_VIRTIO_GPU_PCI_PROPERTIES(_state)               \
+    DEFINE_PROP_BIT("ioeventfd", _state, flags,                \
+                    VIRTIO_PCI_FLAG_USE_IOEVENTFD_BIT, false), \
+    DEFINE_PROP_UINT32("vectors", _state, nvectors, 3)
+
+#define DEFINE_VIRTIO_GPU_PROPERTIES(_state, _conf_field)               \
+    DEFINE_PROP_UINT32("max_outputs", _state, _conf_field.max_outputs, 1)
+
+#define VIRTIO_GPU_FILL_CMD(out) do {                                   \
+        size_t s;                                                       \
+        s = iov_to_buf(cmd->elem.out_sg, cmd->elem.out_num, 0,          \
+                       &out, sizeof(out));                              \
+        if (s != sizeof(out)) {                                         \
+            qemu_log_mask(LOG_GUEST_ERROR,                              \
+                          "%s: command size incorrect %zu vs %zu\n",    \
+                          __func__, s, sizeof(out));                    \
+            return;                                                     \
+        }                                                               \
+    } while (0)
+
+/* virtio-gpu.c */
+void virtio_gpu_ctrl_response(VirtIOGPU *g,
+                              struct virtio_gpu_ctrl_command *cmd,
+                              struct virtio_gpu_ctrl_hdr *resp,
+                              size_t resp_len);
+void virtio_gpu_ctrl_response_nodata(VirtIOGPU *g,
+                                     struct virtio_gpu_ctrl_command *cmd,
+                                     enum virtio_gpu_ctrl_type type);
+void virtio_gpu_get_display_info(VirtIOGPU *g,
+                                 struct virtio_gpu_ctrl_command *cmd);
+int virtio_gpu_create_mapping_iov(struct virtio_gpu_resource_attach_backing *ab,
+                                  struct virtio_gpu_ctrl_command *cmd,
+                                  struct iovec **iov);
+void virtio_gpu_cleanup_mapping_iov(struct iovec *iov, uint32_t count);
+
+#endif
diff --git a/include/hw/virtio/virtio-net.h b/include/hw/virtio/virtio-net.h
index e0dbb41..280dacf 100644
--- a/include/hw/virtio/virtio-net.h
+++ b/include/hw/virtio/virtio-net.h
@@ -107,36 +107,7 @@
  * VIRTIO_NET_F_CTRL_GUEST_OFFLOADS feature bit.
  */
 #define VIRTIO_NET_CTRL_GUEST_OFFLOADS   5
- #define VIRTIO_NET_CTRL_GUEST_OFFLOADS_SET        0
-
-#define DEFINE_VIRTIO_NET_FEATURES(_state, _field) \
-        DEFINE_PROP_BIT("any_layout", _state, _field, VIRTIO_F_ANY_LAYOUT, true), \
-        DEFINE_PROP_BIT("csum", _state, _field, VIRTIO_NET_F_CSUM, true), \
-        DEFINE_PROP_BIT("guest_csum", _state, _field, VIRTIO_NET_F_GUEST_CSUM, true), \
-        DEFINE_PROP_BIT("gso", _state, _field, VIRTIO_NET_F_GSO, true), \
-        DEFINE_PROP_BIT("guest_tso4", _state, _field, VIRTIO_NET_F_GUEST_TSO4, true), \
-        DEFINE_PROP_BIT("guest_tso6", _state, _field, VIRTIO_NET_F_GUEST_TSO6, true), \
-        DEFINE_PROP_BIT("guest_ecn", _state, _field, VIRTIO_NET_F_GUEST_ECN, true), \
-        DEFINE_PROP_BIT("guest_ufo", _state, _field, VIRTIO_NET_F_GUEST_UFO, true), \
-        DEFINE_PROP_BIT("guest_announce", _state, _field, VIRTIO_NET_F_GUEST_ANNOUNCE, true), \
-        DEFINE_PROP_BIT("host_tso4", _state, _field, VIRTIO_NET_F_HOST_TSO4, true), \
-        DEFINE_PROP_BIT("host_tso6", _state, _field, VIRTIO_NET_F_HOST_TSO6, true), \
-        DEFINE_PROP_BIT("host_ecn", _state, _field, VIRTIO_NET_F_HOST_ECN, true), \
-        DEFINE_PROP_BIT("host_ufo", _state, _field, VIRTIO_NET_F_HOST_UFO, true), \
-        DEFINE_PROP_BIT("mrg_rxbuf", _state, _field, VIRTIO_NET_F_MRG_RXBUF, true), \
-        DEFINE_PROP_BIT("status", _state, _field, VIRTIO_NET_F_STATUS, true), \
-        DEFINE_PROP_BIT("ctrl_vq", _state, _field, VIRTIO_NET_F_CTRL_VQ, true), \
-        DEFINE_PROP_BIT("ctrl_rx", _state, _field, VIRTIO_NET_F_CTRL_RX, true), \
-        DEFINE_PROP_BIT("ctrl_vlan", _state, _field, VIRTIO_NET_F_CTRL_VLAN, true), \
-        DEFINE_PROP_BIT("ctrl_rx_extra", _state, _field, VIRTIO_NET_F_CTRL_RX_EXTRA, true), \
-        DEFINE_PROP_BIT("ctrl_mac_addr", _state, _field, VIRTIO_NET_F_CTRL_MAC_ADDR, true), \
-        DEFINE_PROP_BIT("ctrl_guest_offloads", _state, _field, VIRTIO_NET_F_CTRL_GUEST_OFFLOADS, true), \
-        DEFINE_PROP_BIT("mq", _state, _field, VIRTIO_NET_F_MQ, false)
-
-#define DEFINE_VIRTIO_NET_PROPERTIES(_state, _field)                           \
-    DEFINE_PROP_UINT32("x-txtimer", _state, _field.txtimer, TX_TIMER_INTERVAL),\
-    DEFINE_PROP_INT32("x-txburst", _state, _field.txburst, TX_BURST),          \
-    DEFINE_PROP_STRING("tx", _state, _field.tx)
+#define VIRTIO_NET_CTRL_GUEST_OFFLOADS_SET        0
 
 void virtio_net_set_netclient_name(VirtIONet *n, const char *name,
                                    const char *type);
diff --git a/include/hw/virtio/virtio-rng.h b/include/hw/virtio/virtio-rng.h
index 7702ff4..0316488 100644
--- a/include/hw/virtio/virtio-rng.h
+++ b/include/hw/virtio/virtio-rng.h
@@ -46,14 +46,4 @@
     int64_t quota_remaining;
 } VirtIORNG;
 
-/* Set a default rate limit of 2^47 bytes per minute or roughly 2TB/s.  If
-   you have an entropy source capable of generating more entropy than this
-   and you can pass it through via virtio-rng, then hats off to you.  Until
-   then, this is unlimited for all practical purposes.
-*/
-#define DEFINE_VIRTIO_RNG_PROPERTIES(_state, _conf_field)                    \
-        DEFINE_PROP_UINT64("max-bytes", _state, _conf_field.max_bytes,       \
-                           INT64_MAX),                                       \
-        DEFINE_PROP_UINT32("period", _state, _conf_field.period_ms, 1 << 16)
-
 #endif
diff --git a/include/hw/virtio/virtio-scsi.h b/include/hw/virtio/virtio-scsi.h
index b42e7f1..088fe9f 100644
--- a/include/hw/virtio/virtio-scsi.h
+++ b/include/hw/virtio/virtio-scsi.h
@@ -141,19 +141,6 @@
     } req;
 } VirtIOSCSIReq;
 
-#define DEFINE_VIRTIO_SCSI_PROPERTIES(_state, _conf_field)                     \
-    DEFINE_PROP_UINT32("num_queues", _state, _conf_field.num_queues, 1),       \
-    DEFINE_PROP_UINT32("max_sectors", _state, _conf_field.max_sectors, 0xFFFF),\
-    DEFINE_PROP_UINT32("cmd_per_lun", _state, _conf_field.cmd_per_lun, 128)
-
-#define DEFINE_VIRTIO_SCSI_FEATURES(_state, _feature_field)                    \
-    DEFINE_PROP_BIT("any_layout", _state, _feature_field,                      \
-                    VIRTIO_F_ANY_LAYOUT, true),                                \
-    DEFINE_PROP_BIT("hotplug", _state, _feature_field, VIRTIO_SCSI_F_HOTPLUG,  \
-                                                       true),                  \
-    DEFINE_PROP_BIT("param_change", _state, _feature_field,                    \
-                                            VIRTIO_SCSI_F_CHANGE, true)
-
 typedef void (*HandleOutput)(VirtIODevice *, VirtQueue *);
 
 void virtio_scsi_common_realize(DeviceState *dev, Error **errp,
diff --git a/include/hw/virtio/virtio-serial.h b/include/hw/virtio/virtio-serial.h
index 18d1bcc..527d0bf 100644
--- a/include/hw/virtio/virtio-serial.h
+++ b/include/hw/virtio/virtio-serial.h
@@ -221,7 +221,4 @@
 #define VIRTIO_SERIAL(obj) \
         OBJECT_CHECK(VirtIOSerial, (obj), TYPE_VIRTIO_SERIAL)
 
-#define DEFINE_VIRTIO_SERIAL_PROPERTIES(_state, _field) \
-        DEFINE_PROP_UINT32("max_ports", _state, _field.max_virtserial_ports, 31)
-
 #endif
diff --git a/include/hw/virtio/virtio.h b/include/hw/virtio/virtio.h
index 2bb7c1a..473fb75 100644
--- a/include/hw/virtio/virtio.h
+++ b/include/hw/virtio/virtio.h
@@ -78,6 +78,7 @@
     size_t config_len;
     void *config;
     uint16_t config_vector;
+    uint32_t generation;
     int nvectors;
     VirtQueue *vq;
     uint16_t device_id;
@@ -99,6 +100,7 @@
     uint64_t (*get_features)(VirtIODevice *vdev, uint64_t requested_features);
     uint64_t (*bad_features)(VirtIODevice *vdev);
     void (*set_features)(VirtIODevice *vdev, uint64_t val);
+    int (*validate_features)(VirtIODevice *vdev);
     void (*get_config)(VirtIODevice *vdev, uint8_t *config);
     void (*set_config)(VirtIODevice *vdev, const uint8_t *config);
     void (*reset)(VirtIODevice *vdev);
@@ -172,16 +174,28 @@
 void virtio_config_writeb(VirtIODevice *vdev, uint32_t addr, uint32_t data);
 void virtio_config_writew(VirtIODevice *vdev, uint32_t addr, uint32_t data);
 void virtio_config_writel(VirtIODevice *vdev, uint32_t addr, uint32_t data);
+uint32_t virtio_config_modern_readb(VirtIODevice *vdev, uint32_t addr);
+uint32_t virtio_config_modern_readw(VirtIODevice *vdev, uint32_t addr);
+uint32_t virtio_config_modern_readl(VirtIODevice *vdev, uint32_t addr);
+void virtio_config_modern_writeb(VirtIODevice *vdev,
+                                 uint32_t addr, uint32_t data);
+void virtio_config_modern_writew(VirtIODevice *vdev,
+                                 uint32_t addr, uint32_t data);
+void virtio_config_modern_writel(VirtIODevice *vdev,
+                                 uint32_t addr, uint32_t data);
 void virtio_queue_set_addr(VirtIODevice *vdev, int n, hwaddr addr);
 hwaddr virtio_queue_get_addr(VirtIODevice *vdev, int n);
 void virtio_queue_set_num(VirtIODevice *vdev, int n, int num);
 int virtio_queue_get_num(VirtIODevice *vdev, int n);
 int virtio_get_num_queues(VirtIODevice *vdev);
+void virtio_queue_set_rings(VirtIODevice *vdev, int n, hwaddr desc,
+                            hwaddr avail, hwaddr used);
+void virtio_queue_update_rings(VirtIODevice *vdev, int n);
 void virtio_queue_set_align(VirtIODevice *vdev, int n, int align);
 void virtio_queue_notify(VirtIODevice *vdev, int n);
 uint16_t virtio_queue_vector(VirtIODevice *vdev, int n);
 void virtio_queue_set_vector(VirtIODevice *vdev, int n, uint16_t vector);
-void virtio_set_status(VirtIODevice *vdev, uint8_t val);
+int virtio_set_status(VirtIODevice *vdev, uint8_t val);
 void virtio_reset(void *opaque);
 void virtio_update_irq(VirtIODevice *vdev);
 int virtio_set_features(VirtIODevice *vdev, uint64_t val);
@@ -252,7 +266,11 @@
 
 static inline bool virtio_is_big_endian(VirtIODevice *vdev)
 {
-    assert(vdev->device_endian != VIRTIO_DEVICE_ENDIAN_UNKNOWN);
-    return vdev->device_endian == VIRTIO_DEVICE_ENDIAN_BIG;
+    if (!virtio_has_feature(vdev, VIRTIO_F_VERSION_1)) {
+        assert(vdev->device_endian != VIRTIO_DEVICE_ENDIAN_UNKNOWN);
+        return vdev->device_endian == VIRTIO_DEVICE_ENDIAN_BIG;
+    }
+    /* Devices conforming to VIRTIO 1.0 or later are always LE. */
+    return false;
 }
 #endif
diff --git a/include/migration/migration.h b/include/migration/migration.h
index a6e025a..9387c8c 100644
--- a/include/migration/migration.h
+++ b/include/migration/migration.h
@@ -34,6 +34,7 @@
 #define QEMU_VM_SECTION_FULL         0x04
 #define QEMU_VM_SUBSECTION           0x05
 #define QEMU_VM_VMDESCRIPTION        0x06
+#define QEMU_VM_SECTION_FOOTER       0x7e
 
 struct MigrationParams {
     bool blk;
@@ -42,6 +43,20 @@
 
 typedef struct MigrationState MigrationState;
 
+typedef QLIST_HEAD(, LoadStateEntry) LoadStateEntry_Head;
+
+/* State for the incoming migration */
+struct MigrationIncomingState {
+    QEMUFile *file;
+
+    /* See savevm.c */
+    LoadStateEntry_Head loadvm_handlers;
+};
+
+MigrationIncomingState *migration_incoming_get_current(void);
+MigrationIncomingState *migration_incoming_state_new(QEMUFile *f);
+void migration_incoming_state_destroy(void);
+
 struct MigrationState
 {
     int64_t bandwidth_limit;
@@ -180,4 +195,6 @@
                              ram_addr_t offset, size_t size,
                              uint64_t *bytes_sent);
 
+void ram_mig_init(void);
+void savevm_skip_section_footers(void);
 #endif
diff --git a/include/migration/qemu-file.h b/include/migration/qemu-file.h
index a01c5b8..4f67d79 100644
--- a/include/migration/qemu-file.h
+++ b/include/migration/qemu-file.h
@@ -157,7 +157,7 @@
 void qemu_put_be16(QEMUFile *f, unsigned int v);
 void qemu_put_be32(QEMUFile *f, unsigned int v);
 void qemu_put_be64(QEMUFile *f, uint64_t v);
-int qemu_peek_buffer(QEMUFile *f, uint8_t *buf, int size, size_t offset);
+int qemu_peek_buffer(QEMUFile *f, uint8_t **buf, int size, size_t offset);
 int qemu_get_buffer(QEMUFile *f, uint8_t *buf, int size);
 ssize_t qemu_put_compression_data(QEMUFile *f, const uint8_t *p, size_t size,
                                   int level);
@@ -312,4 +312,7 @@
 {
     qemu_get_be64s(f, (uint64_t *)pv);
 }
+
+size_t qemu_get_counted_string(QEMUFile *f, char buf[256]);
+
 #endif
diff --git a/include/migration/vmstate.h b/include/migration/vmstate.h
index bc7616a..7153b1e 100644
--- a/include/migration/vmstate.h
+++ b/include/migration/vmstate.h
@@ -120,11 +120,6 @@
     bool (*field_exists)(void *opaque, int version_id);
 } VMStateField;
 
-typedef struct VMStateSubsection {
-    const VMStateDescription *vmsd;
-    bool (*needed)(void *opaque);
-} VMStateSubsection;
-
 struct VMStateDescription {
     const char *name;
     int unmigratable;
@@ -135,8 +130,9 @@
     int (*pre_load)(void *opaque);
     int (*post_load)(void *opaque, int version_id);
     void (*pre_save)(void *opaque);
+    bool (*needed)(void *opaque);
     VMStateField *fields;
-    const VMStateSubsection *subsections;
+    const VMStateDescription **subsections;
 };
 
 extern const VMStateDescription vmstate_dummy;
@@ -812,6 +808,8 @@
 
 #define SELF_ANNOUNCE_ROUNDS 5
 
+void loadvm_free_handlers(MigrationIncomingState *mis);
+
 int vmstate_load_state(QEMUFile *f, const VMStateDescription *vmsd,
                        void *opaque, int version_id);
 void vmstate_save_state(QEMUFile *f, const VMStateDescription *vmsd,
diff --git a/include/net/vhost_net.h b/include/net/vhost_net.h
index b1c18a3..9eb493e 100644
--- a/include/net/vhost_net.h
+++ b/include/net/vhost_net.h
@@ -22,8 +22,8 @@
 
 void vhost_net_cleanup(VHostNetState *net);
 
-unsigned vhost_net_get_features(VHostNetState *net, unsigned features);
-void vhost_net_ack_features(VHostNetState *net, unsigned features);
+uint64_t vhost_net_get_features(VHostNetState *net, uint64_t features);
+void vhost_net_ack_features(VHostNetState *net, uint64_t features);
 
 bool vhost_net_virtqueue_pending(VHostNetState *net, int n);
 void vhost_net_virtqueue_mask(VHostNetState *net, VirtIODevice *dev,
diff --git a/include/qemu/typedefs.h b/include/qemu/typedefs.h
index cde3314..6fdcbcd 100644
--- a/include/qemu/typedefs.h
+++ b/include/qemu/typedefs.h
@@ -31,6 +31,7 @@
 typedef struct I2SCodec I2SCodec;
 typedef struct ISABus ISABus;
 typedef struct ISADevice ISADevice;
+typedef struct LoadStateEntry LoadStateEntry;
 typedef struct MACAddr MACAddr;
 typedef struct MachineClass MachineClass;
 typedef struct MachineState MachineState;
@@ -38,6 +39,7 @@
 typedef struct MemoryMappingList MemoryMappingList;
 typedef struct MemoryRegion MemoryRegion;
 typedef struct MemoryRegionSection MemoryRegionSection;
+typedef struct MigrationIncomingState MigrationIncomingState;
 typedef struct MigrationParams MigrationParams;
 typedef struct Monitor Monitor;
 typedef struct MouseTransformInfo MouseTransformInfo;
diff --git a/include/standard-headers/linux/virtio_gpu.h b/include/standard-headers/linux/virtio_gpu.h
new file mode 100644
index 0000000..cfcfb46
--- /dev/null
+++ b/include/standard-headers/linux/virtio_gpu.h
@@ -0,0 +1,204 @@
+/*
+ * Virtio GPU Device
+ *
+ * Copyright Red Hat, Inc. 2013-2014
+ *
+ * Authors:
+ *     Dave Airlie <airlied@redhat.com>
+ *     Gerd Hoffmann <kraxel@redhat.com>
+ *
+ * This header is BSD licensed so anyone can use the definitions
+ * to implement compatible drivers/servers:
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of IBM nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL IBM OR
+ * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF
+ * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
+ * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+ * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
+ * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ */
+
+#ifndef VIRTIO_GPU_HW_H
+#define VIRTIO_GPU_HW_H
+
+enum virtio_gpu_ctrl_type {
+	VIRTIO_GPU_UNDEFINED = 0,
+
+	/* 2d commands */
+	VIRTIO_GPU_CMD_GET_DISPLAY_INFO = 0x0100,
+	VIRTIO_GPU_CMD_RESOURCE_CREATE_2D,
+	VIRTIO_GPU_CMD_RESOURCE_UNREF,
+	VIRTIO_GPU_CMD_SET_SCANOUT,
+	VIRTIO_GPU_CMD_RESOURCE_FLUSH,
+	VIRTIO_GPU_CMD_TRANSFER_TO_HOST_2D,
+	VIRTIO_GPU_CMD_RESOURCE_ATTACH_BACKING,
+	VIRTIO_GPU_CMD_RESOURCE_DETACH_BACKING,
+
+	/* cursor commands */
+	VIRTIO_GPU_CMD_UPDATE_CURSOR = 0x0300,
+	VIRTIO_GPU_CMD_MOVE_CURSOR,
+
+	/* success responses */
+	VIRTIO_GPU_RESP_OK_NODATA = 0x1100,
+	VIRTIO_GPU_RESP_OK_DISPLAY_INFO,
+
+	/* error responses */
+	VIRTIO_GPU_RESP_ERR_UNSPEC = 0x1200,
+	VIRTIO_GPU_RESP_ERR_OUT_OF_MEMORY,
+	VIRTIO_GPU_RESP_ERR_INVALID_SCANOUT_ID,
+	VIRTIO_GPU_RESP_ERR_INVALID_RESOURCE_ID,
+	VIRTIO_GPU_RESP_ERR_INVALID_CONTEXT_ID,
+	VIRTIO_GPU_RESP_ERR_INVALID_PARAMETER,
+};
+
+#define VIRTIO_GPU_FLAG_FENCE (1 << 0)
+
+struct virtio_gpu_ctrl_hdr {
+	uint32_t type;
+	uint32_t flags;
+	uint64_t fence_id;
+	uint32_t ctx_id;
+	uint32_t padding;
+};
+
+/* data passed in the cursor vq */
+
+struct virtio_gpu_cursor_pos {
+	uint32_t scanout_id;
+	uint32_t x;
+	uint32_t y;
+	uint32_t padding;
+};
+
+/* VIRTIO_GPU_CMD_UPDATE_CURSOR, VIRTIO_GPU_CMD_MOVE_CURSOR */
+struct virtio_gpu_update_cursor {
+	struct virtio_gpu_ctrl_hdr hdr;
+	struct virtio_gpu_cursor_pos pos;  /* update & move */
+	uint32_t resource_id;           /* update only */
+	uint32_t hot_x;                 /* update only */
+	uint32_t hot_y;                 /* update only */
+	uint32_t padding;
+};
+
+/* data passed in the control vq, 2d related */
+
+struct virtio_gpu_rect {
+	uint32_t x;
+	uint32_t y;
+	uint32_t width;
+	uint32_t height;
+};
+
+/* VIRTIO_GPU_CMD_RESOURCE_UNREF */
+struct virtio_gpu_resource_unref {
+	struct virtio_gpu_ctrl_hdr hdr;
+	uint32_t resource_id;
+	uint32_t padding;
+};
+
+/* VIRTIO_GPU_CMD_RESOURCE_CREATE_2D: create a 2d resource with a format */
+struct virtio_gpu_resource_create_2d {
+	struct virtio_gpu_ctrl_hdr hdr;
+	uint32_t resource_id;
+	uint32_t format;
+	uint32_t width;
+	uint32_t height;
+};
+
+/* VIRTIO_GPU_CMD_SET_SCANOUT */
+struct virtio_gpu_set_scanout {
+	struct virtio_gpu_ctrl_hdr hdr;
+	struct virtio_gpu_rect r;
+	uint32_t scanout_id;
+	uint32_t resource_id;
+};
+
+/* VIRTIO_GPU_CMD_RESOURCE_FLUSH */
+struct virtio_gpu_resource_flush {
+	struct virtio_gpu_ctrl_hdr hdr;
+	struct virtio_gpu_rect r;
+	uint32_t resource_id;
+	uint32_t padding;
+};
+
+/* VIRTIO_GPU_CMD_TRANSFER_TO_HOST_2D: simple transfer to_host */
+struct virtio_gpu_transfer_to_host_2d {
+	struct virtio_gpu_ctrl_hdr hdr;
+	struct virtio_gpu_rect r;
+	uint64_t offset;
+	uint32_t resource_id;
+	uint32_t padding;
+};
+
+struct virtio_gpu_mem_entry {
+	uint64_t addr;
+	uint32_t length;
+	uint32_t padding;
+};
+
+/* VIRTIO_GPU_CMD_RESOURCE_ATTACH_BACKING */
+struct virtio_gpu_resource_attach_backing {
+	struct virtio_gpu_ctrl_hdr hdr;
+	uint32_t resource_id;
+	uint32_t nr_entries;
+};
+
+/* VIRTIO_GPU_CMD_RESOURCE_DETACH_BACKING */
+struct virtio_gpu_resource_detach_backing {
+	struct virtio_gpu_ctrl_hdr hdr;
+	uint32_t resource_id;
+	uint32_t padding;
+};
+
+/* VIRTIO_GPU_RESP_OK_DISPLAY_INFO */
+#define VIRTIO_GPU_MAX_SCANOUTS 16
+struct virtio_gpu_resp_display_info {
+	struct virtio_gpu_ctrl_hdr hdr;
+	struct virtio_gpu_display_one {
+		struct virtio_gpu_rect r;
+		uint32_t enabled;
+		uint32_t flags;
+	} pmodes[VIRTIO_GPU_MAX_SCANOUTS];
+};
+
+#define VIRTIO_GPU_EVENT_DISPLAY (1 << 0)
+
+struct virtio_gpu_config {
+	uint32_t events_read;
+	uint32_t events_clear;
+	uint32_t num_scanouts;
+	uint32_t reserved;
+};
+
+/* simple formats for fbcon/X use */
+enum virtio_gpu_formats {
+	VIRTIO_GPU_FORMAT_B8G8R8A8_UNORM  = 1,
+	VIRTIO_GPU_FORMAT_B8G8R8X8_UNORM  = 2,
+	VIRTIO_GPU_FORMAT_A8R8G8B8_UNORM  = 3,
+	VIRTIO_GPU_FORMAT_X8R8G8B8_UNORM  = 4,
+
+	VIRTIO_GPU_FORMAT_R8G8B8A8_UNORM  = 67,
+	VIRTIO_GPU_FORMAT_X8B8G8R8_UNORM  = 68,
+
+	VIRTIO_GPU_FORMAT_A8B8G8R8_UNORM  = 121,
+	VIRTIO_GPU_FORMAT_R8G8B8X8_UNORM  = 134,
+};
+
+#endif
diff --git a/include/standard-headers/linux/virtio_ids.h b/include/standard-headers/linux/virtio_ids.h
index 5f60aa4..77925f5 100644
--- a/include/standard-headers/linux/virtio_ids.h
+++ b/include/standard-headers/linux/virtio_ids.h
@@ -39,6 +39,7 @@
 #define VIRTIO_ID_9P		9 /* 9p virtio console */
 #define VIRTIO_ID_RPROC_SERIAL 11 /* virtio remoteproc serial link */
 #define VIRTIO_ID_CAIF	       12 /* Virtio caif */
+#define VIRTIO_ID_GPU          16 /* virtio GPU */
 #define VIRTIO_ID_INPUT        18 /* virtio input */
 
 #endif /* _LINUX_VIRTIO_IDS_H */
diff --git a/include/sysemu/arch_init.h b/include/sysemu/arch_init.h
index 54b36c1..c38892f 100644
--- a/include/sysemu/arch_init.h
+++ b/include/sysemu/arch_init.h
@@ -30,7 +30,6 @@
 void select_soundhw(const char *optarg);
 void do_acpitable_option(const QemuOpts *opts);
 void do_smbios_option(QemuOpts *opts);
-void ram_mig_init(void);
 void cpudef_init(void);
 void audio_init(void);
 int kvm_available(void);
diff --git a/include/sysemu/sysemu.h b/include/sysemu/sysemu.h
index 853d90a..ef793f7 100644
--- a/include/sysemu/sysemu.h
+++ b/include/sysemu/sysemu.h
@@ -84,6 +84,7 @@
 bool qemu_savevm_state_blocked(Error **errp);
 void qemu_savevm_state_begin(QEMUFile *f,
                              const MigrationParams *params);
+void qemu_savevm_state_header(QEMUFile *f);
 int qemu_savevm_state_iterate(QEMUFile *f);
 void qemu_savevm_state_complete(QEMUFile *f);
 void qemu_savevm_state_cancel(void);
diff --git a/linux-headers/linux/virtio_pci.h b/linux-headers/linux/virtio_pci.h
new file mode 100644
index 0000000..92624e5
--- /dev/null
+++ b/linux-headers/linux/virtio_pci.h
@@ -0,0 +1,192 @@
+/*
+ * Virtio PCI driver
+ *
+ * This module allows virtio devices to be used over a virtual PCI device.
+ * This can be used with QEMU based VMMs like KVM or Xen.
+ *
+ * Copyright IBM Corp. 2007
+ *
+ * Authors:
+ *  Anthony Liguori  <aliguori@us.ibm.com>
+ *
+ * This header is BSD licensed so anyone can use the definitions to implement
+ * compatible drivers/servers.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of IBM nor the names of its contributors
+ *    may be used to endorse or promote products derived from this software
+ *    without specific prior written permission.
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS'' AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED.  IN NO EVENT SHALL IBM OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
+ * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
+ * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
+ * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
+ * SUCH DAMAGE.
+ */
+
+#ifndef _LINUX_VIRTIO_PCI_H
+#define _LINUX_VIRTIO_PCI_H
+
+#include <linux/types.h>
+
+#ifndef VIRTIO_PCI_NO_LEGACY
+
+/* A 32-bit r/o bitmask of the features supported by the host */
+#define VIRTIO_PCI_HOST_FEATURES	0
+
+/* A 32-bit r/w bitmask of features activated by the guest */
+#define VIRTIO_PCI_GUEST_FEATURES	4
+
+/* A 32-bit r/w PFN for the currently selected queue */
+#define VIRTIO_PCI_QUEUE_PFN		8
+
+/* A 16-bit r/o queue size for the currently selected queue */
+#define VIRTIO_PCI_QUEUE_NUM		12
+
+/* A 16-bit r/w queue selector */
+#define VIRTIO_PCI_QUEUE_SEL		14
+
+/* A 16-bit r/w queue notifier */
+#define VIRTIO_PCI_QUEUE_NOTIFY		16
+
+/* An 8-bit device status register.  */
+#define VIRTIO_PCI_STATUS		18
+
+/* An 8-bit r/o interrupt status register.  Reading the value will return the
+ * current contents of the ISR and will also clear it.  This is effectively
+ * a read-and-acknowledge. */
+#define VIRTIO_PCI_ISR			19
+
+/* MSI-X registers: only enabled if MSI-X is enabled. */
+/* A 16-bit vector for configuration changes. */
+#define VIRTIO_MSI_CONFIG_VECTOR        20
+/* A 16-bit vector for selected queue notifications. */
+#define VIRTIO_MSI_QUEUE_VECTOR         22
+
+/* The remaining space is defined by each driver as the per-driver
+ * configuration space */
+#define VIRTIO_PCI_CONFIG_OFF(msix_enabled)	((msix_enabled) ? 24 : 20)
+/* Deprecated: please use VIRTIO_PCI_CONFIG_OFF instead */
+#define VIRTIO_PCI_CONFIG(dev)	VIRTIO_PCI_CONFIG_OFF((dev)->msix_enabled)
+
+/* Virtio ABI version, this must match exactly */
+#define VIRTIO_PCI_ABI_VERSION		0
+
+/* How many bits to shift physical queue address written to QUEUE_PFN.
+ * 12 is historical, and due to x86 page size. */
+#define VIRTIO_PCI_QUEUE_ADDR_SHIFT	12
+
+/* The alignment to use between consumer and producer parts of vring.
+ * x86 pagesize again. */
+#define VIRTIO_PCI_VRING_ALIGN		4096
+
+#endif /* VIRTIO_PCI_NO_LEGACY */
+
+/* The bit of the ISR which indicates a device configuration change. */
+#define VIRTIO_PCI_ISR_CONFIG		0x2
+/* Vector value used to disable MSI for queue */
+#define VIRTIO_MSI_NO_VECTOR            0xffff
+
+#ifndef VIRTIO_PCI_NO_MODERN
+
+/* IDs for different capabilities.  Must all exist. */
+
+/* Common configuration */
+#define VIRTIO_PCI_CAP_COMMON_CFG	1
+/* Notifications */
+#define VIRTIO_PCI_CAP_NOTIFY_CFG	2
+/* ISR access */
+#define VIRTIO_PCI_CAP_ISR_CFG		3
+/* Device specific confiuration */
+#define VIRTIO_PCI_CAP_DEVICE_CFG	4
+
+/* This is the PCI capability header: */
+struct virtio_pci_cap {
+	__u8 cap_vndr;		/* Generic PCI field: PCI_CAP_ID_VNDR */
+	__u8 cap_next;		/* Generic PCI field: next ptr. */
+	__u8 cap_len;		/* Generic PCI field: capability length */
+	__u8 cfg_type;		/* Identifies the structure. */
+	__u8 bar;		/* Where to find it. */
+	__u8 padding[3];	/* Pad to full dword. */
+	__le32 offset;		/* Offset within bar. */
+	__le32 length;		/* Length of the structure, in bytes. */
+};
+
+struct virtio_pci_notify_cap {
+	struct virtio_pci_cap cap;
+	__le32 notify_off_multiplier;	/* Multiplier for queue_notify_off. */
+};
+
+/* Fields in VIRTIO_PCI_CAP_COMMON_CFG: */
+struct virtio_pci_common_cfg {
+	/* About the whole device. */
+	__le32 device_feature_select;	/* read-write */
+	__le32 device_feature;		/* read-only */
+	__le32 guest_feature_select;	/* read-write */
+	__le32 guest_feature;		/* read-write */
+	__le16 msix_config;		/* read-write */
+	__le16 num_queues;		/* read-only */
+	__u8 device_status;		/* read-write */
+	__u8 config_generation;		/* read-only */
+
+	/* About a specific virtqueue. */
+	__le16 queue_select;		/* read-write */
+	__le16 queue_size;		/* read-write, power of 2. */
+	__le16 queue_msix_vector;	/* read-write */
+	__le16 queue_enable;		/* read-write */
+	__le16 queue_notify_off;	/* read-only */
+	__le32 queue_desc_lo;		/* read-write */
+	__le32 queue_desc_hi;		/* read-write */
+	__le32 queue_avail_lo;		/* read-write */
+	__le32 queue_avail_hi;		/* read-write */
+	__le32 queue_used_lo;		/* read-write */
+	__le32 queue_used_hi;		/* read-write */
+};
+
+/* Macro versions of offsets for the Old Timers! */
+#define VIRTIO_PCI_CAP_VNDR		0
+#define VIRTIO_PCI_CAP_NEXT		1
+#define VIRTIO_PCI_CAP_LEN		2
+#define VIRTIO_PCI_CAP_CFG_TYPE		3
+#define VIRTIO_PCI_CAP_BAR		4
+#define VIRTIO_PCI_CAP_OFFSET		8
+#define VIRTIO_PCI_CAP_LENGTH		12
+
+#define VIRTIO_PCI_NOTIFY_CAP_MULT	16
+
+
+#define VIRTIO_PCI_COMMON_DFSELECT	0
+#define VIRTIO_PCI_COMMON_DF		4
+#define VIRTIO_PCI_COMMON_GFSELECT	8
+#define VIRTIO_PCI_COMMON_GF		12
+#define VIRTIO_PCI_COMMON_MSIX		16
+#define VIRTIO_PCI_COMMON_NUMQ		18
+#define VIRTIO_PCI_COMMON_STATUS	20
+#define VIRTIO_PCI_COMMON_CFGGENERATION	21
+#define VIRTIO_PCI_COMMON_Q_SELECT	22
+#define VIRTIO_PCI_COMMON_Q_SIZE	24
+#define VIRTIO_PCI_COMMON_Q_MSIX	26
+#define VIRTIO_PCI_COMMON_Q_ENABLE	28
+#define VIRTIO_PCI_COMMON_Q_NOFF	30
+#define VIRTIO_PCI_COMMON_Q_DESCLO	32
+#define VIRTIO_PCI_COMMON_Q_DESCHI	36
+#define VIRTIO_PCI_COMMON_Q_AVAILLO	40
+#define VIRTIO_PCI_COMMON_Q_AVAILHI	44
+#define VIRTIO_PCI_COMMON_Q_USEDLO	48
+#define VIRTIO_PCI_COMMON_Q_USEDHI	52
+
+#endif /* VIRTIO_PCI_NO_MODERN */
+
+#endif
diff --git a/migration/migration.c b/migration/migration.c
index 732d229..b04b457 100644
--- a/migration/migration.c
+++ b/migration/migration.c
@@ -53,6 +53,7 @@
    migrations at once.  For now we don't need to add
    dynamic creation of migration */
 
+/* For outgoing */
 MigrationState *migrate_get_current(void)
 {
     static MigrationState current_migration = {
@@ -71,6 +72,30 @@
     return &current_migration;
 }
 
+/* For incoming */
+static MigrationIncomingState *mis_current;
+
+MigrationIncomingState *migration_incoming_get_current(void)
+{
+    return mis_current;
+}
+
+MigrationIncomingState *migration_incoming_state_new(QEMUFile* f)
+{
+    mis_current = g_malloc0(sizeof(MigrationIncomingState));
+    mis_current->file = f;
+    QLIST_INIT(&mis_current->loadvm_handlers);
+
+    return mis_current;
+}
+
+void migration_incoming_state_destroy(void)
+{
+    loadvm_free_handlers(mis_current);
+    g_free(mis_current);
+    mis_current = NULL;
+}
+
 /*
  * Called on -incoming with a defer: uri.
  * The migration can be started later after any parameters have been
@@ -115,9 +140,14 @@
     Error *local_err = NULL;
     int ret;
 
+    migration_incoming_state_new(f);
+
     ret = qemu_loadvm_state(f);
+
     qemu_fclose(f);
     free_xbzrle_decoded_buf();
+    migration_incoming_state_destroy();
+
     if (ret < 0) {
         error_report("load of migration failed: %s", strerror(-ret));
         migrate_decompress_threads_join();
@@ -738,6 +768,7 @@
     int64_t start_time = initial_time;
     bool old_vm_running = false;
 
+    qemu_savevm_state_header(s->file);
     qemu_savevm_state_begin(s->file, &s->params);
 
     s->setup_time = qemu_clock_get_ms(QEMU_CLOCK_HOST) - setup_start;
@@ -838,9 +869,6 @@
 
 void migrate_fd_connect(MigrationState *s)
 {
-    s->state = MIGRATION_STATUS_SETUP;
-    trace_migrate_set_state(MIGRATION_STATUS_SETUP);
-
     /* This is a best 1st approximation. ns to ms */
     s->expected_downtime = max_downtime/1000000;
     s->cleanup_bh = qemu_bh_new(migrate_fd_cleanup, s);
diff --git a/migration/qemu-file.c b/migration/qemu-file.c
index 2750365..965a757 100644
--- a/migration/qemu-file.c
+++ b/migration/qemu-file.c
@@ -349,14 +349,14 @@
 }
 
 /*
- * Read 'size' bytes from file (at 'offset') into buf without moving the
- * pointer.
+ * Read 'size' bytes from file (at 'offset') without moving the
+ * pointer and set 'buf' to point to that data.
  *
  * It will return size bytes unless there was an error, in which case it will
  * return as many as it managed to read (assuming blocking fd's which
  * all current QEMUFile are)
  */
-int qemu_peek_buffer(QEMUFile *f, uint8_t *buf, int size, size_t offset)
+int qemu_peek_buffer(QEMUFile *f, uint8_t **buf, int size, size_t offset)
 {
     int pending;
     int index;
@@ -392,7 +392,7 @@
         size = pending;
     }
 
-    memcpy(buf, f->buf + index, size);
+    *buf = f->buf + index;
     return size;
 }
 
@@ -411,11 +411,13 @@
 
     while (pending > 0) {
         int res;
+        uint8_t *src;
 
-        res = qemu_peek_buffer(f, buf, MIN(pending, IO_BUF_SIZE), 0);
+        res = qemu_peek_buffer(f, &src, MIN(pending, IO_BUF_SIZE), 0);
         if (res == 0) {
             return done;
         }
+        memcpy(buf, src, res);
         qemu_file_skip(f, res);
         buf += res;
         pending -= res;
@@ -585,3 +587,20 @@
     }
     return len;
 }
+
+/*
+ * Get a string whose length is determined by a single preceding byte
+ * A preallocated 256 byte buffer must be passed in.
+ * Returns: len on success and a 0 terminated string in the buffer
+ *          else 0
+ *          (Note a 0 length string will return 0 either way)
+ */
+size_t qemu_get_counted_string(QEMUFile *f, char buf[256])
+{
+    size_t len = qemu_get_byte(f);
+    size_t res = qemu_get_buffer(f, (uint8_t *)buf, len);
+
+    buf[res] = 0;
+
+    return res == len ? res : 0;
+}
diff --git a/migration/ram.c b/migration/ram.c
new file mode 100644
index 0000000..57368e1
--- /dev/null
+++ b/migration/ram.c
@@ -0,0 +1,1628 @@
+/*
+ * QEMU System Emulator
+ *
+ * Copyright (c) 2003-2008 Fabrice Bellard
+ * Copyright (c) 2011-2015 Red Hat Inc
+ *
+ * Authors:
+ *  Juan Quintela <quintela@redhat.com>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+#include <stdint.h>
+#include <zlib.h>
+#include "qemu/bitops.h"
+#include "qemu/bitmap.h"
+#include "qemu/timer.h"
+#include "qemu/main-loop.h"
+#include "migration/migration.h"
+#include "exec/address-spaces.h"
+#include "migration/page_cache.h"
+#include "qemu/error-report.h"
+#include "trace.h"
+#include "exec/ram_addr.h"
+#include "qemu/rcu_queue.h"
+
+#ifdef DEBUG_MIGRATION_RAM
+#define DPRINTF(fmt, ...) \
+    do { fprintf(stdout, "migration_ram: " fmt, ## __VA_ARGS__); } while (0)
+#else
+#define DPRINTF(fmt, ...) \
+    do { } while (0)
+#endif
+
+static bool mig_throttle_on;
+static int dirty_rate_high_cnt;
+static void check_guest_throttling(void);
+
+static uint64_t bitmap_sync_count;
+
+/***********************************************************/
+/* ram save/restore */
+
+#define RAM_SAVE_FLAG_FULL     0x01 /* Obsolete, not used anymore */
+#define RAM_SAVE_FLAG_COMPRESS 0x02
+#define RAM_SAVE_FLAG_MEM_SIZE 0x04
+#define RAM_SAVE_FLAG_PAGE     0x08
+#define RAM_SAVE_FLAG_EOS      0x10
+#define RAM_SAVE_FLAG_CONTINUE 0x20
+#define RAM_SAVE_FLAG_XBZRLE   0x40
+/* 0x80 is reserved in migration.h start with 0x100 next */
+#define RAM_SAVE_FLAG_COMPRESS_PAGE    0x100
+
+static const uint8_t ZERO_TARGET_PAGE[TARGET_PAGE_SIZE];
+
+static inline bool is_zero_range(uint8_t *p, uint64_t size)
+{
+    return buffer_find_nonzero_offset(p, size) == size;
+}
+
+/* struct contains XBZRLE cache and a static page
+   used by the compression */
+static struct {
+    /* buffer used for XBZRLE encoding */
+    uint8_t *encoded_buf;
+    /* buffer for storing page content */
+    uint8_t *current_buf;
+    /* Cache for XBZRLE, Protected by lock. */
+    PageCache *cache;
+    QemuMutex lock;
+} XBZRLE;
+
+/* buffer used for XBZRLE decoding */
+static uint8_t *xbzrle_decoded_buf;
+
+static void XBZRLE_cache_lock(void)
+{
+    if (migrate_use_xbzrle())
+        qemu_mutex_lock(&XBZRLE.lock);
+}
+
+static void XBZRLE_cache_unlock(void)
+{
+    if (migrate_use_xbzrle())
+        qemu_mutex_unlock(&XBZRLE.lock);
+}
+
+/*
+ * called from qmp_migrate_set_cache_size in main thread, possibly while
+ * a migration is in progress.
+ * A running migration maybe using the cache and might finish during this
+ * call, hence changes to the cache are protected by XBZRLE.lock().
+ */
+int64_t xbzrle_cache_resize(int64_t new_size)
+{
+    PageCache *new_cache;
+    int64_t ret;
+
+    if (new_size < TARGET_PAGE_SIZE) {
+        return -1;
+    }
+
+    XBZRLE_cache_lock();
+
+    if (XBZRLE.cache != NULL) {
+        if (pow2floor(new_size) == migrate_xbzrle_cache_size()) {
+            goto out_new_size;
+        }
+        new_cache = cache_init(new_size / TARGET_PAGE_SIZE,
+                                        TARGET_PAGE_SIZE);
+        if (!new_cache) {
+            error_report("Error creating cache");
+            ret = -1;
+            goto out;
+        }
+
+        cache_fini(XBZRLE.cache);
+        XBZRLE.cache = new_cache;
+    }
+
+out_new_size:
+    ret = pow2floor(new_size);
+out:
+    XBZRLE_cache_unlock();
+    return ret;
+}
+
+/* accounting for migration statistics */
+typedef struct AccountingInfo {
+    uint64_t dup_pages;
+    uint64_t skipped_pages;
+    uint64_t norm_pages;
+    uint64_t iterations;
+    uint64_t xbzrle_bytes;
+    uint64_t xbzrle_pages;
+    uint64_t xbzrle_cache_miss;
+    double xbzrle_cache_miss_rate;
+    uint64_t xbzrle_overflows;
+} AccountingInfo;
+
+static AccountingInfo acct_info;
+
+static void acct_clear(void)
+{
+    memset(&acct_info, 0, sizeof(acct_info));
+}
+
+uint64_t dup_mig_bytes_transferred(void)
+{
+    return acct_info.dup_pages * TARGET_PAGE_SIZE;
+}
+
+uint64_t dup_mig_pages_transferred(void)
+{
+    return acct_info.dup_pages;
+}
+
+uint64_t skipped_mig_bytes_transferred(void)
+{
+    return acct_info.skipped_pages * TARGET_PAGE_SIZE;
+}
+
+uint64_t skipped_mig_pages_transferred(void)
+{
+    return acct_info.skipped_pages;
+}
+
+uint64_t norm_mig_bytes_transferred(void)
+{
+    return acct_info.norm_pages * TARGET_PAGE_SIZE;
+}
+
+uint64_t norm_mig_pages_transferred(void)
+{
+    return acct_info.norm_pages;
+}
+
+uint64_t xbzrle_mig_bytes_transferred(void)
+{
+    return acct_info.xbzrle_bytes;
+}
+
+uint64_t xbzrle_mig_pages_transferred(void)
+{
+    return acct_info.xbzrle_pages;
+}
+
+uint64_t xbzrle_mig_pages_cache_miss(void)
+{
+    return acct_info.xbzrle_cache_miss;
+}
+
+double xbzrle_mig_cache_miss_rate(void)
+{
+    return acct_info.xbzrle_cache_miss_rate;
+}
+
+uint64_t xbzrle_mig_pages_overflow(void)
+{
+    return acct_info.xbzrle_overflows;
+}
+
+/* This is the last block that we have visited serching for dirty pages
+ */
+static RAMBlock *last_seen_block;
+/* This is the last block from where we have sent data */
+static RAMBlock *last_sent_block;
+static ram_addr_t last_offset;
+static unsigned long *migration_bitmap;
+static uint64_t migration_dirty_pages;
+static uint32_t last_version;
+static bool ram_bulk_stage;
+
+struct CompressParam {
+    bool start;
+    bool done;
+    QEMUFile *file;
+    QemuMutex mutex;
+    QemuCond cond;
+    RAMBlock *block;
+    ram_addr_t offset;
+};
+typedef struct CompressParam CompressParam;
+
+struct DecompressParam {
+    bool start;
+    QemuMutex mutex;
+    QemuCond cond;
+    void *des;
+    uint8 *compbuf;
+    int len;
+};
+typedef struct DecompressParam DecompressParam;
+
+static CompressParam *comp_param;
+static QemuThread *compress_threads;
+/* comp_done_cond is used to wake up the migration thread when
+ * one of the compression threads has finished the compression.
+ * comp_done_lock is used to co-work with comp_done_cond.
+ */
+static QemuMutex *comp_done_lock;
+static QemuCond *comp_done_cond;
+/* The empty QEMUFileOps will be used by file in CompressParam */
+static const QEMUFileOps empty_ops = { };
+
+static bool compression_switch;
+static bool quit_comp_thread;
+static bool quit_decomp_thread;
+static DecompressParam *decomp_param;
+static QemuThread *decompress_threads;
+static uint8_t *compressed_data_buf;
+
+static int do_compress_ram_page(CompressParam *param);
+
+static void *do_data_compress(void *opaque)
+{
+    CompressParam *param = opaque;
+
+    while (!quit_comp_thread) {
+        qemu_mutex_lock(&param->mutex);
+        /* Re-check the quit_comp_thread in case of
+         * terminate_compression_threads is called just before
+         * qemu_mutex_lock(&param->mutex) and after
+         * while(!quit_comp_thread), re-check it here can make
+         * sure the compression thread terminate as expected.
+         */
+        while (!param->start && !quit_comp_thread) {
+            qemu_cond_wait(&param->cond, &param->mutex);
+        }
+        if (!quit_comp_thread) {
+            do_compress_ram_page(param);
+        }
+        param->start = false;
+        qemu_mutex_unlock(&param->mutex);
+
+        qemu_mutex_lock(comp_done_lock);
+        param->done = true;
+        qemu_cond_signal(comp_done_cond);
+        qemu_mutex_unlock(comp_done_lock);
+    }
+
+    return NULL;
+}
+
+static inline void terminate_compression_threads(void)
+{
+    int idx, thread_count;
+
+    thread_count = migrate_compress_threads();
+    quit_comp_thread = true;
+    for (idx = 0; idx < thread_count; idx++) {
+        qemu_mutex_lock(&comp_param[idx].mutex);
+        qemu_cond_signal(&comp_param[idx].cond);
+        qemu_mutex_unlock(&comp_param[idx].mutex);
+    }
+}
+
+void migrate_compress_threads_join(void)
+{
+    int i, thread_count;
+
+    if (!migrate_use_compression()) {
+        return;
+    }
+    terminate_compression_threads();
+    thread_count = migrate_compress_threads();
+    for (i = 0; i < thread_count; i++) {
+        qemu_thread_join(compress_threads + i);
+        qemu_fclose(comp_param[i].file);
+        qemu_mutex_destroy(&comp_param[i].mutex);
+        qemu_cond_destroy(&comp_param[i].cond);
+    }
+    qemu_mutex_destroy(comp_done_lock);
+    qemu_cond_destroy(comp_done_cond);
+    g_free(compress_threads);
+    g_free(comp_param);
+    g_free(comp_done_cond);
+    g_free(comp_done_lock);
+    compress_threads = NULL;
+    comp_param = NULL;
+    comp_done_cond = NULL;
+    comp_done_lock = NULL;
+}
+
+void migrate_compress_threads_create(void)
+{
+    int i, thread_count;
+
+    if (!migrate_use_compression()) {
+        return;
+    }
+    quit_comp_thread = false;
+    compression_switch = true;
+    thread_count = migrate_compress_threads();
+    compress_threads = g_new0(QemuThread, thread_count);
+    comp_param = g_new0(CompressParam, thread_count);
+    comp_done_cond = g_new0(QemuCond, 1);
+    comp_done_lock = g_new0(QemuMutex, 1);
+    qemu_cond_init(comp_done_cond);
+    qemu_mutex_init(comp_done_lock);
+    for (i = 0; i < thread_count; i++) {
+        /* com_param[i].file is just used as a dummy buffer to save data, set
+         * it's ops to empty.
+         */
+        comp_param[i].file = qemu_fopen_ops(NULL, &empty_ops);
+        comp_param[i].done = true;
+        qemu_mutex_init(&comp_param[i].mutex);
+        qemu_cond_init(&comp_param[i].cond);
+        qemu_thread_create(compress_threads + i, "compress",
+                           do_data_compress, comp_param + i,
+                           QEMU_THREAD_JOINABLE);
+    }
+}
+
+/**
+ * save_page_header: Write page header to wire
+ *
+ * If this is the 1st block, it also writes the block identification
+ *
+ * Returns: Number of bytes written
+ *
+ * @f: QEMUFile where to send the data
+ * @block: block that contains the page we want to send
+ * @offset: offset inside the block for the page
+ *          in the lower bits, it contains flags
+ */
+static size_t save_page_header(QEMUFile *f, RAMBlock *block, ram_addr_t offset)
+{
+    size_t size;
+
+    qemu_put_be64(f, offset);
+    size = 8;
+
+    if (!(offset & RAM_SAVE_FLAG_CONTINUE)) {
+        qemu_put_byte(f, strlen(block->idstr));
+        qemu_put_buffer(f, (uint8_t *)block->idstr,
+                        strlen(block->idstr));
+        size += 1 + strlen(block->idstr);
+    }
+    return size;
+}
+
+/* Update the xbzrle cache to reflect a page that's been sent as all 0.
+ * The important thing is that a stale (not-yet-0'd) page be replaced
+ * by the new data.
+ * As a bonus, if the page wasn't in the cache it gets added so that
+ * when a small write is made into the 0'd page it gets XBZRLE sent
+ */
+static void xbzrle_cache_zero_page(ram_addr_t current_addr)
+{
+    if (ram_bulk_stage || !migrate_use_xbzrle()) {
+        return;
+    }
+
+    /* We don't care if this fails to allocate a new cache page
+     * as long as it updated an old one */
+    cache_insert(XBZRLE.cache, current_addr, ZERO_TARGET_PAGE,
+                 bitmap_sync_count);
+}
+
+#define ENCODING_FLAG_XBZRLE 0x1
+
+/**
+ * save_xbzrle_page: compress and send current page
+ *
+ * Returns: 1 means that we wrote the page
+ *          0 means that page is identical to the one already sent
+ *          -1 means that xbzrle would be longer than normal
+ *
+ * @f: QEMUFile where to send the data
+ * @current_data:
+ * @current_addr:
+ * @block: block that contains the page we want to send
+ * @offset: offset inside the block for the page
+ * @last_stage: if we are at the completion stage
+ * @bytes_transferred: increase it with the number of transferred bytes
+ */
+static int save_xbzrle_page(QEMUFile *f, uint8_t **current_data,
+                            ram_addr_t current_addr, RAMBlock *block,
+                            ram_addr_t offset, bool last_stage,
+                            uint64_t *bytes_transferred)
+{
+    int encoded_len = 0, bytes_xbzrle;
+    uint8_t *prev_cached_page;
+
+    if (!cache_is_cached(XBZRLE.cache, current_addr, bitmap_sync_count)) {
+        acct_info.xbzrle_cache_miss++;
+        if (!last_stage) {
+            if (cache_insert(XBZRLE.cache, current_addr, *current_data,
+                             bitmap_sync_count) == -1) {
+                return -1;
+            } else {
+                /* update *current_data when the page has been
+                   inserted into cache */
+                *current_data = get_cached_data(XBZRLE.cache, current_addr);
+            }
+        }
+        return -1;
+    }
+
+    prev_cached_page = get_cached_data(XBZRLE.cache, current_addr);
+
+    /* save current buffer into memory */
+    memcpy(XBZRLE.current_buf, *current_data, TARGET_PAGE_SIZE);
+
+    /* XBZRLE encoding (if there is no overflow) */
+    encoded_len = xbzrle_encode_buffer(prev_cached_page, XBZRLE.current_buf,
+                                       TARGET_PAGE_SIZE, XBZRLE.encoded_buf,
+                                       TARGET_PAGE_SIZE);
+    if (encoded_len == 0) {
+        DPRINTF("Skipping unmodified page\n");
+        return 0;
+    } else if (encoded_len == -1) {
+        DPRINTF("Overflow\n");
+        acct_info.xbzrle_overflows++;
+        /* update data in the cache */
+        if (!last_stage) {
+            memcpy(prev_cached_page, *current_data, TARGET_PAGE_SIZE);
+            *current_data = prev_cached_page;
+        }
+        return -1;
+    }
+
+    /* we need to update the data in the cache, in order to get the same data */
+    if (!last_stage) {
+        memcpy(prev_cached_page, XBZRLE.current_buf, TARGET_PAGE_SIZE);
+    }
+
+    /* Send XBZRLE based compressed page */
+    bytes_xbzrle = save_page_header(f, block, offset | RAM_SAVE_FLAG_XBZRLE);
+    qemu_put_byte(f, ENCODING_FLAG_XBZRLE);
+    qemu_put_be16(f, encoded_len);
+    qemu_put_buffer(f, XBZRLE.encoded_buf, encoded_len);
+    bytes_xbzrle += encoded_len + 1 + 2;
+    acct_info.xbzrle_pages++;
+    acct_info.xbzrle_bytes += bytes_xbzrle;
+    *bytes_transferred += bytes_xbzrle;
+
+    return 1;
+}
+
+static inline
+ram_addr_t migration_bitmap_find_and_reset_dirty(MemoryRegion *mr,
+                                                 ram_addr_t start)
+{
+    unsigned long base = mr->ram_addr >> TARGET_PAGE_BITS;
+    unsigned long nr = base + (start >> TARGET_PAGE_BITS);
+    uint64_t mr_size = TARGET_PAGE_ALIGN(memory_region_size(mr));
+    unsigned long size = base + (mr_size >> TARGET_PAGE_BITS);
+
+    unsigned long next;
+
+    if (ram_bulk_stage && nr > base) {
+        next = nr + 1;
+    } else {
+        next = find_next_bit(migration_bitmap, size, nr);
+    }
+
+    if (next < size) {
+        clear_bit(next, migration_bitmap);
+        migration_dirty_pages--;
+    }
+    return (next - base) << TARGET_PAGE_BITS;
+}
+
+static void migration_bitmap_sync_range(ram_addr_t start, ram_addr_t length)
+{
+    migration_dirty_pages +=
+        cpu_physical_memory_sync_dirty_bitmap(migration_bitmap, start, length);
+}
+
+
+/* Fix me: there are too many global variables used in migration process. */
+static int64_t start_time;
+static int64_t bytes_xfer_prev;
+static int64_t num_dirty_pages_period;
+static uint64_t xbzrle_cache_miss_prev;
+static uint64_t iterations_prev;
+
+static void migration_bitmap_sync_init(void)
+{
+    start_time = 0;
+    bytes_xfer_prev = 0;
+    num_dirty_pages_period = 0;
+    xbzrle_cache_miss_prev = 0;
+    iterations_prev = 0;
+}
+
+/* Called with iothread lock held, to protect ram_list.dirty_memory[] */
+static void migration_bitmap_sync(void)
+{
+    RAMBlock *block;
+    uint64_t num_dirty_pages_init = migration_dirty_pages;
+    MigrationState *s = migrate_get_current();
+    int64_t end_time;
+    int64_t bytes_xfer_now;
+
+    bitmap_sync_count++;
+
+    if (!bytes_xfer_prev) {
+        bytes_xfer_prev = ram_bytes_transferred();
+    }
+
+    if (!start_time) {
+        start_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME);
+    }
+
+    trace_migration_bitmap_sync_start();
+    address_space_sync_dirty_bitmap(&address_space_memory);
+
+    rcu_read_lock();
+    QLIST_FOREACH_RCU(block, &ram_list.blocks, next) {
+        migration_bitmap_sync_range(block->mr->ram_addr, block->used_length);
+    }
+    rcu_read_unlock();
+
+    trace_migration_bitmap_sync_end(migration_dirty_pages
+                                    - num_dirty_pages_init);
+    num_dirty_pages_period += migration_dirty_pages - num_dirty_pages_init;
+    end_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME);
+
+    /* more than 1 second = 1000 millisecons */
+    if (end_time > start_time + 1000) {
+        if (migrate_auto_converge()) {
+            /* The following detection logic can be refined later. For now:
+               Check to see if the dirtied bytes is 50% more than the approx.
+               amount of bytes that just got transferred since the last time we
+               were in this routine. If that happens >N times (for now N==4)
+               we turn on the throttle down logic */
+            bytes_xfer_now = ram_bytes_transferred();
+            if (s->dirty_pages_rate &&
+               (num_dirty_pages_period * TARGET_PAGE_SIZE >
+                   (bytes_xfer_now - bytes_xfer_prev)/2) &&
+               (dirty_rate_high_cnt++ > 4)) {
+                    trace_migration_throttle();
+                    mig_throttle_on = true;
+                    dirty_rate_high_cnt = 0;
+             }
+             bytes_xfer_prev = bytes_xfer_now;
+        } else {
+             mig_throttle_on = false;
+        }
+        if (migrate_use_xbzrle()) {
+            if (iterations_prev != acct_info.iterations) {
+                acct_info.xbzrle_cache_miss_rate =
+                   (double)(acct_info.xbzrle_cache_miss -
+                            xbzrle_cache_miss_prev) /
+                   (acct_info.iterations - iterations_prev);
+            }
+            iterations_prev = acct_info.iterations;
+            xbzrle_cache_miss_prev = acct_info.xbzrle_cache_miss;
+        }
+        s->dirty_pages_rate = num_dirty_pages_period * 1000
+            / (end_time - start_time);
+        s->dirty_bytes_rate = s->dirty_pages_rate * TARGET_PAGE_SIZE;
+        start_time = end_time;
+        num_dirty_pages_period = 0;
+    }
+    s->dirty_sync_count = bitmap_sync_count;
+}
+
+/**
+ * save_zero_page: Send the zero page to the stream
+ *
+ * Returns: Number of pages written.
+ *
+ * @f: QEMUFile where to send the data
+ * @block: block that contains the page we want to send
+ * @offset: offset inside the block for the page
+ * @p: pointer to the page
+ * @bytes_transferred: increase it with the number of transferred bytes
+ */
+static int save_zero_page(QEMUFile *f, RAMBlock *block, ram_addr_t offset,
+                          uint8_t *p, uint64_t *bytes_transferred)
+{
+    int pages = -1;
+
+    if (is_zero_range(p, TARGET_PAGE_SIZE)) {
+        acct_info.dup_pages++;
+        *bytes_transferred += save_page_header(f, block,
+                                               offset | RAM_SAVE_FLAG_COMPRESS);
+        qemu_put_byte(f, 0);
+        *bytes_transferred += 1;
+        pages = 1;
+    }
+
+    return pages;
+}
+
+/**
+ * ram_save_page: Send the given page to the stream
+ *
+ * Returns: Number of pages written.
+ *
+ * @f: QEMUFile where to send the data
+ * @block: block that contains the page we want to send
+ * @offset: offset inside the block for the page
+ * @last_stage: if we are at the completion stage
+ * @bytes_transferred: increase it with the number of transferred bytes
+ */
+static int ram_save_page(QEMUFile *f, RAMBlock* block, ram_addr_t offset,
+                         bool last_stage, uint64_t *bytes_transferred)
+{
+    int pages = -1;
+    uint64_t bytes_xmit;
+    ram_addr_t current_addr;
+    MemoryRegion *mr = block->mr;
+    uint8_t *p;
+    int ret;
+    bool send_async = true;
+
+    p = memory_region_get_ram_ptr(mr) + offset;
+
+    /* In doubt sent page as normal */
+    bytes_xmit = 0;
+    ret = ram_control_save_page(f, block->offset,
+                           offset, TARGET_PAGE_SIZE, &bytes_xmit);
+    if (bytes_xmit) {
+        *bytes_transferred += bytes_xmit;
+        pages = 1;
+    }
+
+    XBZRLE_cache_lock();
+
+    current_addr = block->offset + offset;
+
+    if (block == last_sent_block) {
+        offset |= RAM_SAVE_FLAG_CONTINUE;
+    }
+    if (ret != RAM_SAVE_CONTROL_NOT_SUPP) {
+        if (ret != RAM_SAVE_CONTROL_DELAYED) {
+            if (bytes_xmit > 0) {
+                acct_info.norm_pages++;
+            } else if (bytes_xmit == 0) {
+                acct_info.dup_pages++;
+            }
+        }
+    } else {
+        pages = save_zero_page(f, block, offset, p, bytes_transferred);
+        if (pages > 0) {
+            /* Must let xbzrle know, otherwise a previous (now 0'd) cached
+             * page would be stale
+             */
+            xbzrle_cache_zero_page(current_addr);
+        } else if (!ram_bulk_stage && migrate_use_xbzrle()) {
+            pages = save_xbzrle_page(f, &p, current_addr, block,
+                                     offset, last_stage, bytes_transferred);
+            if (!last_stage) {
+                /* Can't send this cached data async, since the cache page
+                 * might get updated before it gets to the wire
+                 */
+                send_async = false;
+            }
+        }
+    }
+
+    /* XBZRLE overflow or normal page */
+    if (pages == -1) {
+        *bytes_transferred += save_page_header(f, block,
+                                               offset | RAM_SAVE_FLAG_PAGE);
+        if (send_async) {
+            qemu_put_buffer_async(f, p, TARGET_PAGE_SIZE);
+        } else {
+            qemu_put_buffer(f, p, TARGET_PAGE_SIZE);
+        }
+        *bytes_transferred += TARGET_PAGE_SIZE;
+        pages = 1;
+        acct_info.norm_pages++;
+    }
+
+    XBZRLE_cache_unlock();
+
+    return pages;
+}
+
+static int do_compress_ram_page(CompressParam *param)
+{
+    int bytes_sent, blen;
+    uint8_t *p;
+    RAMBlock *block = param->block;
+    ram_addr_t offset = param->offset;
+
+    p = memory_region_get_ram_ptr(block->mr) + (offset & TARGET_PAGE_MASK);
+
+    bytes_sent = save_page_header(param->file, block, offset |
+                                  RAM_SAVE_FLAG_COMPRESS_PAGE);
+    blen = qemu_put_compression_data(param->file, p, TARGET_PAGE_SIZE,
+                                     migrate_compress_level());
+    bytes_sent += blen;
+
+    return bytes_sent;
+}
+
+static inline void start_compression(CompressParam *param)
+{
+    param->done = false;
+    qemu_mutex_lock(&param->mutex);
+    param->start = true;
+    qemu_cond_signal(&param->cond);
+    qemu_mutex_unlock(&param->mutex);
+}
+
+static inline void start_decompression(DecompressParam *param)
+{
+    qemu_mutex_lock(&param->mutex);
+    param->start = true;
+    qemu_cond_signal(&param->cond);
+    qemu_mutex_unlock(&param->mutex);
+}
+
+static uint64_t bytes_transferred;
+
+static void flush_compressed_data(QEMUFile *f)
+{
+    int idx, len, thread_count;
+
+    if (!migrate_use_compression()) {
+        return;
+    }
+    thread_count = migrate_compress_threads();
+    for (idx = 0; idx < thread_count; idx++) {
+        if (!comp_param[idx].done) {
+            qemu_mutex_lock(comp_done_lock);
+            while (!comp_param[idx].done && !quit_comp_thread) {
+                qemu_cond_wait(comp_done_cond, comp_done_lock);
+            }
+            qemu_mutex_unlock(comp_done_lock);
+        }
+        if (!quit_comp_thread) {
+            len = qemu_put_qemu_file(f, comp_param[idx].file);
+            bytes_transferred += len;
+        }
+    }
+}
+
+static inline void set_compress_params(CompressParam *param, RAMBlock *block,
+                                       ram_addr_t offset)
+{
+    param->block = block;
+    param->offset = offset;
+}
+
+static int compress_page_with_multi_thread(QEMUFile *f, RAMBlock *block,
+                                           ram_addr_t offset,
+                                           uint64_t *bytes_transferred)
+{
+    int idx, thread_count, bytes_xmit = -1, pages = -1;
+
+    thread_count = migrate_compress_threads();
+    qemu_mutex_lock(comp_done_lock);
+    while (true) {
+        for (idx = 0; idx < thread_count; idx++) {
+            if (comp_param[idx].done) {
+                bytes_xmit = qemu_put_qemu_file(f, comp_param[idx].file);
+                set_compress_params(&comp_param[idx], block, offset);
+                start_compression(&comp_param[idx]);
+                pages = 1;
+                acct_info.norm_pages++;
+                *bytes_transferred += bytes_xmit;
+                break;
+            }
+        }
+        if (pages > 0) {
+            break;
+        } else {
+            qemu_cond_wait(comp_done_cond, comp_done_lock);
+        }
+    }
+    qemu_mutex_unlock(comp_done_lock);
+
+    return pages;
+}
+
+/**
+ * ram_save_compressed_page: compress the given page and send it to the stream
+ *
+ * Returns: Number of pages written.
+ *
+ * @f: QEMUFile where to send the data
+ * @block: block that contains the page we want to send
+ * @offset: offset inside the block for the page
+ * @last_stage: if we are at the completion stage
+ * @bytes_transferred: increase it with the number of transferred bytes
+ */
+static int ram_save_compressed_page(QEMUFile *f, RAMBlock *block,
+                                    ram_addr_t offset, bool last_stage,
+                                    uint64_t *bytes_transferred)
+{
+    int pages = -1;
+    uint64_t bytes_xmit;
+    MemoryRegion *mr = block->mr;
+    uint8_t *p;
+    int ret;
+
+    p = memory_region_get_ram_ptr(mr) + offset;
+
+    bytes_xmit = 0;
+    ret = ram_control_save_page(f, block->offset,
+                                offset, TARGET_PAGE_SIZE, &bytes_xmit);
+    if (bytes_xmit) {
+        *bytes_transferred += bytes_xmit;
+        pages = 1;
+    }
+    if (block == last_sent_block) {
+        offset |= RAM_SAVE_FLAG_CONTINUE;
+    }
+    if (ret != RAM_SAVE_CONTROL_NOT_SUPP) {
+        if (ret != RAM_SAVE_CONTROL_DELAYED) {
+            if (bytes_xmit > 0) {
+                acct_info.norm_pages++;
+            } else if (bytes_xmit == 0) {
+                acct_info.dup_pages++;
+            }
+        }
+    } else {
+        /* When starting the process of a new block, the first page of
+         * the block should be sent out before other pages in the same
+         * block, and all the pages in last block should have been sent
+         * out, keeping this order is important, because the 'cont' flag
+         * is used to avoid resending the block name.
+         */
+        if (block != last_sent_block) {
+            flush_compressed_data(f);
+            pages = save_zero_page(f, block, offset, p, bytes_transferred);
+            if (pages == -1) {
+                set_compress_params(&comp_param[0], block, offset);
+                /* Use the qemu thread to compress the data to make sure the
+                 * first page is sent out before other pages
+                 */
+                bytes_xmit = do_compress_ram_page(&comp_param[0]);
+                acct_info.norm_pages++;
+                qemu_put_qemu_file(f, comp_param[0].file);
+                *bytes_transferred += bytes_xmit;
+                pages = 1;
+            }
+        } else {
+            pages = save_zero_page(f, block, offset, p, bytes_transferred);
+            if (pages == -1) {
+                pages = compress_page_with_multi_thread(f, block, offset,
+                                                        bytes_transferred);
+            }
+        }
+    }
+
+    return pages;
+}
+
+/**
+ * ram_find_and_save_block: Finds a dirty page and sends it to f
+ *
+ * Called within an RCU critical section.
+ *
+ * Returns:  The number of pages written
+ *           0 means no dirty pages
+ *
+ * @f: QEMUFile where to send the data
+ * @last_stage: if we are at the completion stage
+ * @bytes_transferred: increase it with the number of transferred bytes
+ */
+
+static int ram_find_and_save_block(QEMUFile *f, bool last_stage,
+                                   uint64_t *bytes_transferred)
+{
+    RAMBlock *block = last_seen_block;
+    ram_addr_t offset = last_offset;
+    bool complete_round = false;
+    int pages = 0;
+    MemoryRegion *mr;
+
+    if (!block)
+        block = QLIST_FIRST_RCU(&ram_list.blocks);
+
+    while (true) {
+        mr = block->mr;
+        offset = migration_bitmap_find_and_reset_dirty(mr, offset);
+        if (complete_round && block == last_seen_block &&
+            offset >= last_offset) {
+            break;
+        }
+        if (offset >= block->used_length) {
+            offset = 0;
+            block = QLIST_NEXT_RCU(block, next);
+            if (!block) {
+                block = QLIST_FIRST_RCU(&ram_list.blocks);
+                complete_round = true;
+                ram_bulk_stage = false;
+                if (migrate_use_xbzrle()) {
+                    /* If xbzrle is on, stop using the data compression at this
+                     * point. In theory, xbzrle can do better than compression.
+                     */
+                    flush_compressed_data(f);
+                    compression_switch = false;
+                }
+            }
+        } else {
+            if (compression_switch && migrate_use_compression()) {
+                pages = ram_save_compressed_page(f, block, offset, last_stage,
+                                                 bytes_transferred);
+            } else {
+                pages = ram_save_page(f, block, offset, last_stage,
+                                      bytes_transferred);
+            }
+
+            /* if page is unmodified, continue to the next */
+            if (pages > 0) {
+                last_sent_block = block;
+                break;
+            }
+        }
+    }
+
+    last_seen_block = block;
+    last_offset = offset;
+
+    return pages;
+}
+
+void acct_update_position(QEMUFile *f, size_t size, bool zero)
+{
+    uint64_t pages = size / TARGET_PAGE_SIZE;
+    if (zero) {
+        acct_info.dup_pages += pages;
+    } else {
+        acct_info.norm_pages += pages;
+        bytes_transferred += size;
+        qemu_update_position(f, size);
+    }
+}
+
+static ram_addr_t ram_save_remaining(void)
+{
+    return migration_dirty_pages;
+}
+
+uint64_t ram_bytes_remaining(void)
+{
+    return ram_save_remaining() * TARGET_PAGE_SIZE;
+}
+
+uint64_t ram_bytes_transferred(void)
+{
+    return bytes_transferred;
+}
+
+uint64_t ram_bytes_total(void)
+{
+    RAMBlock *block;
+    uint64_t total = 0;
+
+    rcu_read_lock();
+    QLIST_FOREACH_RCU(block, &ram_list.blocks, next)
+        total += block->used_length;
+    rcu_read_unlock();
+    return total;
+}
+
+void free_xbzrle_decoded_buf(void)
+{
+    g_free(xbzrle_decoded_buf);
+    xbzrle_decoded_buf = NULL;
+}
+
+static void migration_end(void)
+{
+    if (migration_bitmap) {
+        memory_global_dirty_log_stop();
+        g_free(migration_bitmap);
+        migration_bitmap = NULL;
+    }
+
+    XBZRLE_cache_lock();
+    if (XBZRLE.cache) {
+        cache_fini(XBZRLE.cache);
+        g_free(XBZRLE.encoded_buf);
+        g_free(XBZRLE.current_buf);
+        XBZRLE.cache = NULL;
+        XBZRLE.encoded_buf = NULL;
+        XBZRLE.current_buf = NULL;
+    }
+    XBZRLE_cache_unlock();
+}
+
+static void ram_migration_cancel(void *opaque)
+{
+    migration_end();
+}
+
+static void reset_ram_globals(void)
+{
+    last_seen_block = NULL;
+    last_sent_block = NULL;
+    last_offset = 0;
+    last_version = ram_list.version;
+    ram_bulk_stage = true;
+}
+
+#define MAX_WAIT 50 /* ms, half buffered_file limit */
+
+
+/* Each of ram_save_setup, ram_save_iterate and ram_save_complete has
+ * long-running RCU critical section.  When rcu-reclaims in the code
+ * start to become numerous it will be necessary to reduce the
+ * granularity of these critical sections.
+ */
+
+static int ram_save_setup(QEMUFile *f, void *opaque)
+{
+    RAMBlock *block;
+    int64_t ram_bitmap_pages; /* Size of bitmap in pages, including gaps */
+
+    mig_throttle_on = false;
+    dirty_rate_high_cnt = 0;
+    bitmap_sync_count = 0;
+    migration_bitmap_sync_init();
+
+    if (migrate_use_xbzrle()) {
+        XBZRLE_cache_lock();
+        XBZRLE.cache = cache_init(migrate_xbzrle_cache_size() /
+                                  TARGET_PAGE_SIZE,
+                                  TARGET_PAGE_SIZE);
+        if (!XBZRLE.cache) {
+            XBZRLE_cache_unlock();
+            error_report("Error creating cache");
+            return -1;
+        }
+        XBZRLE_cache_unlock();
+
+        /* We prefer not to abort if there is no memory */
+        XBZRLE.encoded_buf = g_try_malloc0(TARGET_PAGE_SIZE);
+        if (!XBZRLE.encoded_buf) {
+            error_report("Error allocating encoded_buf");
+            return -1;
+        }
+
+        XBZRLE.current_buf = g_try_malloc(TARGET_PAGE_SIZE);
+        if (!XBZRLE.current_buf) {
+            error_report("Error allocating current_buf");
+            g_free(XBZRLE.encoded_buf);
+            XBZRLE.encoded_buf = NULL;
+            return -1;
+        }
+
+        acct_clear();
+    }
+
+    /* iothread lock needed for ram_list.dirty_memory[] */
+    qemu_mutex_lock_iothread();
+    qemu_mutex_lock_ramlist();
+    rcu_read_lock();
+    bytes_transferred = 0;
+    reset_ram_globals();
+
+    ram_bitmap_pages = last_ram_offset() >> TARGET_PAGE_BITS;
+    migration_bitmap = bitmap_new(ram_bitmap_pages);
+    bitmap_set(migration_bitmap, 0, ram_bitmap_pages);
+
+    /*
+     * Count the total number of pages used by ram blocks not including any
+     * gaps due to alignment or unplugs.
+     */
+    migration_dirty_pages = ram_bytes_total() >> TARGET_PAGE_BITS;
+
+    memory_global_dirty_log_start();
+    migration_bitmap_sync();
+    qemu_mutex_unlock_ramlist();
+    qemu_mutex_unlock_iothread();
+
+    qemu_put_be64(f, ram_bytes_total() | RAM_SAVE_FLAG_MEM_SIZE);
+
+    QLIST_FOREACH_RCU(block, &ram_list.blocks, next) {
+        qemu_put_byte(f, strlen(block->idstr));
+        qemu_put_buffer(f, (uint8_t *)block->idstr, strlen(block->idstr));
+        qemu_put_be64(f, block->used_length);
+    }
+
+    rcu_read_unlock();
+
+    ram_control_before_iterate(f, RAM_CONTROL_SETUP);
+    ram_control_after_iterate(f, RAM_CONTROL_SETUP);
+
+    qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
+
+    return 0;
+}
+
+static int ram_save_iterate(QEMUFile *f, void *opaque)
+{
+    int ret;
+    int i;
+    int64_t t0;
+    int pages_sent = 0;
+
+    rcu_read_lock();
+    if (ram_list.version != last_version) {
+        reset_ram_globals();
+    }
+
+    /* Read version before ram_list.blocks */
+    smp_rmb();
+
+    ram_control_before_iterate(f, RAM_CONTROL_ROUND);
+
+    t0 = qemu_clock_get_ns(QEMU_CLOCK_REALTIME);
+    i = 0;
+    while ((ret = qemu_file_rate_limit(f)) == 0) {
+        int pages;
+
+        pages = ram_find_and_save_block(f, false, &bytes_transferred);
+        /* no more pages to sent */
+        if (pages == 0) {
+            break;
+        }
+        pages_sent += pages;
+        acct_info.iterations++;
+        check_guest_throttling();
+        /* we want to check in the 1st loop, just in case it was the 1st time
+           and we had to sync the dirty bitmap.
+           qemu_get_clock_ns() is a bit expensive, so we only check each some
+           iterations
+        */
+        if ((i & 63) == 0) {
+            uint64_t t1 = (qemu_clock_get_ns(QEMU_CLOCK_REALTIME) - t0) / 1000000;
+            if (t1 > MAX_WAIT) {
+                DPRINTF("big wait: %" PRIu64 " milliseconds, %d iterations\n",
+                        t1, i);
+                break;
+            }
+        }
+        i++;
+    }
+    flush_compressed_data(f);
+    rcu_read_unlock();
+
+    /*
+     * Must occur before EOS (or any QEMUFile operation)
+     * because of RDMA protocol.
+     */
+    ram_control_after_iterate(f, RAM_CONTROL_ROUND);
+
+    qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
+    bytes_transferred += 8;
+
+    ret = qemu_file_get_error(f);
+    if (ret < 0) {
+        return ret;
+    }
+
+    return pages_sent;
+}
+
+/* Called with iothread lock */
+static int ram_save_complete(QEMUFile *f, void *opaque)
+{
+    rcu_read_lock();
+
+    migration_bitmap_sync();
+
+    ram_control_before_iterate(f, RAM_CONTROL_FINISH);
+
+    /* try transferring iterative blocks of memory */
+
+    /* flush all remaining blocks regardless of rate limiting */
+    while (true) {
+        int pages;
+
+        pages = ram_find_and_save_block(f, true, &bytes_transferred);
+        /* no more blocks to sent */
+        if (pages == 0) {
+            break;
+        }
+    }
+
+    flush_compressed_data(f);
+    ram_control_after_iterate(f, RAM_CONTROL_FINISH);
+    migration_end();
+
+    rcu_read_unlock();
+    qemu_put_be64(f, RAM_SAVE_FLAG_EOS);
+
+    return 0;
+}
+
+static uint64_t ram_save_pending(QEMUFile *f, void *opaque, uint64_t max_size)
+{
+    uint64_t remaining_size;
+
+    remaining_size = ram_save_remaining() * TARGET_PAGE_SIZE;
+
+    if (remaining_size < max_size) {
+        qemu_mutex_lock_iothread();
+        rcu_read_lock();
+        migration_bitmap_sync();
+        rcu_read_unlock();
+        qemu_mutex_unlock_iothread();
+        remaining_size = ram_save_remaining() * TARGET_PAGE_SIZE;
+    }
+    return remaining_size;
+}
+
+static int load_xbzrle(QEMUFile *f, ram_addr_t addr, void *host)
+{
+    unsigned int xh_len;
+    int xh_flags;
+
+    if (!xbzrle_decoded_buf) {
+        xbzrle_decoded_buf = g_malloc(TARGET_PAGE_SIZE);
+    }
+
+    /* extract RLE header */
+    xh_flags = qemu_get_byte(f);
+    xh_len = qemu_get_be16(f);
+
+    if (xh_flags != ENCODING_FLAG_XBZRLE) {
+        error_report("Failed to load XBZRLE page - wrong compression!");
+        return -1;
+    }
+
+    if (xh_len > TARGET_PAGE_SIZE) {
+        error_report("Failed to load XBZRLE page - len overflow!");
+        return -1;
+    }
+    /* load data and decode */
+    qemu_get_buffer(f, xbzrle_decoded_buf, xh_len);
+
+    /* decode RLE */
+    if (xbzrle_decode_buffer(xbzrle_decoded_buf, xh_len, host,
+                             TARGET_PAGE_SIZE) == -1) {
+        error_report("Failed to load XBZRLE page - decode error!");
+        return -1;
+    }
+
+    return 0;
+}
+
+/* Must be called from within a rcu critical section.
+ * Returns a pointer from within the RCU-protected ram_list.
+ */
+static inline void *host_from_stream_offset(QEMUFile *f,
+                                            ram_addr_t offset,
+                                            int flags)
+{
+    static RAMBlock *block = NULL;
+    char id[256];
+    uint8_t len;
+
+    if (flags & RAM_SAVE_FLAG_CONTINUE) {
+        if (!block || block->max_length <= offset) {
+            error_report("Ack, bad migration stream!");
+            return NULL;
+        }
+
+        return memory_region_get_ram_ptr(block->mr) + offset;
+    }
+
+    len = qemu_get_byte(f);
+    qemu_get_buffer(f, (uint8_t *)id, len);
+    id[len] = 0;
+
+    QLIST_FOREACH_RCU(block, &ram_list.blocks, next) {
+        if (!strncmp(id, block->idstr, sizeof(id)) &&
+            block->max_length > offset) {
+            return memory_region_get_ram_ptr(block->mr) + offset;
+        }
+    }
+
+    error_report("Can't find block %s!", id);
+    return NULL;
+}
+
+/*
+ * If a page (or a whole RDMA chunk) has been
+ * determined to be zero, then zap it.
+ */
+void ram_handle_compressed(void *host, uint8_t ch, uint64_t size)
+{
+    if (ch != 0 || !is_zero_range(host, size)) {
+        memset(host, ch, size);
+    }
+}
+
+static void *do_data_decompress(void *opaque)
+{
+    DecompressParam *param = opaque;
+    unsigned long pagesize;
+
+    while (!quit_decomp_thread) {
+        qemu_mutex_lock(&param->mutex);
+        while (!param->start && !quit_decomp_thread) {
+            qemu_cond_wait(&param->cond, &param->mutex);
+            pagesize = TARGET_PAGE_SIZE;
+            if (!quit_decomp_thread) {
+                /* uncompress() will return failed in some case, especially
+                 * when the page is dirted when doing the compression, it's
+                 * not a problem because the dirty page will be retransferred
+                 * and uncompress() won't break the data in other pages.
+                 */
+                uncompress((Bytef *)param->des, &pagesize,
+                           (const Bytef *)param->compbuf, param->len);
+            }
+            param->start = false;
+        }
+        qemu_mutex_unlock(&param->mutex);
+    }
+
+    return NULL;
+}
+
+void migrate_decompress_threads_create(void)
+{
+    int i, thread_count;
+
+    thread_count = migrate_decompress_threads();
+    decompress_threads = g_new0(QemuThread, thread_count);
+    decomp_param = g_new0(DecompressParam, thread_count);
+    compressed_data_buf = g_malloc0(compressBound(TARGET_PAGE_SIZE));
+    quit_decomp_thread = false;
+    for (i = 0; i < thread_count; i++) {
+        qemu_mutex_init(&decomp_param[i].mutex);
+        qemu_cond_init(&decomp_param[i].cond);
+        decomp_param[i].compbuf = g_malloc0(compressBound(TARGET_PAGE_SIZE));
+        qemu_thread_create(decompress_threads + i, "decompress",
+                           do_data_decompress, decomp_param + i,
+                           QEMU_THREAD_JOINABLE);
+    }
+}
+
+void migrate_decompress_threads_join(void)
+{
+    int i, thread_count;
+
+    quit_decomp_thread = true;
+    thread_count = migrate_decompress_threads();
+    for (i = 0; i < thread_count; i++) {
+        qemu_mutex_lock(&decomp_param[i].mutex);
+        qemu_cond_signal(&decomp_param[i].cond);
+        qemu_mutex_unlock(&decomp_param[i].mutex);
+    }
+    for (i = 0; i < thread_count; i++) {
+        qemu_thread_join(decompress_threads + i);
+        qemu_mutex_destroy(&decomp_param[i].mutex);
+        qemu_cond_destroy(&decomp_param[i].cond);
+        g_free(decomp_param[i].compbuf);
+    }
+    g_free(decompress_threads);
+    g_free(decomp_param);
+    g_free(compressed_data_buf);
+    decompress_threads = NULL;
+    decomp_param = NULL;
+    compressed_data_buf = NULL;
+}
+
+static void decompress_data_with_multi_threads(uint8_t *compbuf,
+                                               void *host, int len)
+{
+    int idx, thread_count;
+
+    thread_count = migrate_decompress_threads();
+    while (true) {
+        for (idx = 0; idx < thread_count; idx++) {
+            if (!decomp_param[idx].start) {
+                memcpy(decomp_param[idx].compbuf, compbuf, len);
+                decomp_param[idx].des = host;
+                decomp_param[idx].len = len;
+                start_decompression(&decomp_param[idx]);
+                break;
+            }
+        }
+        if (idx < thread_count) {
+            break;
+        }
+    }
+}
+
+static int ram_load(QEMUFile *f, void *opaque, int version_id)
+{
+    int flags = 0, ret = 0;
+    static uint64_t seq_iter;
+    int len = 0;
+
+    seq_iter++;
+
+    if (version_id != 4) {
+        ret = -EINVAL;
+    }
+
+    /* This RCU critical section can be very long running.
+     * When RCU reclaims in the code start to become numerous,
+     * it will be necessary to reduce the granularity of this
+     * critical section.
+     */
+    rcu_read_lock();
+    while (!ret && !(flags & RAM_SAVE_FLAG_EOS)) {
+        ram_addr_t addr, total_ram_bytes;
+        void *host;
+        uint8_t ch;
+
+        addr = qemu_get_be64(f);
+        flags = addr & ~TARGET_PAGE_MASK;
+        addr &= TARGET_PAGE_MASK;
+
+        switch (flags & ~RAM_SAVE_FLAG_CONTINUE) {
+        case RAM_SAVE_FLAG_MEM_SIZE:
+            /* Synchronize RAM block list */
+            total_ram_bytes = addr;
+            while (!ret && total_ram_bytes) {
+                RAMBlock *block;
+                char id[256];
+                ram_addr_t length;
+
+                len = qemu_get_byte(f);
+                qemu_get_buffer(f, (uint8_t *)id, len);
+                id[len] = 0;
+                length = qemu_get_be64(f);
+
+                QLIST_FOREACH_RCU(block, &ram_list.blocks, next) {
+                    if (!strncmp(id, block->idstr, sizeof(id))) {
+                        if (length != block->used_length) {
+                            Error *local_err = NULL;
+
+                            ret = qemu_ram_resize(block->offset, length, &local_err);
+                            if (local_err) {
+                                error_report_err(local_err);
+                            }
+                        }
+                        break;
+                    }
+                }
+
+                if (!block) {
+                    error_report("Unknown ramblock \"%s\", cannot "
+                                 "accept migration", id);
+                    ret = -EINVAL;
+                }
+
+                total_ram_bytes -= length;
+            }
+            break;
+        case RAM_SAVE_FLAG_COMPRESS:
+            host = host_from_stream_offset(f, addr, flags);
+            if (!host) {
+                error_report("Illegal RAM offset " RAM_ADDR_FMT, addr);
+                ret = -EINVAL;
+                break;
+            }
+            ch = qemu_get_byte(f);
+            ram_handle_compressed(host, ch, TARGET_PAGE_SIZE);
+            break;
+        case RAM_SAVE_FLAG_PAGE:
+            host = host_from_stream_offset(f, addr, flags);
+            if (!host) {
+                error_report("Illegal RAM offset " RAM_ADDR_FMT, addr);
+                ret = -EINVAL;
+                break;
+            }
+            qemu_get_buffer(f, host, TARGET_PAGE_SIZE);
+            break;
+        case RAM_SAVE_FLAG_COMPRESS_PAGE:
+            host = host_from_stream_offset(f, addr, flags);
+            if (!host) {
+                error_report("Invalid RAM offset " RAM_ADDR_FMT, addr);
+                ret = -EINVAL;
+                break;
+            }
+
+            len = qemu_get_be32(f);
+            if (len < 0 || len > compressBound(TARGET_PAGE_SIZE)) {
+                error_report("Invalid compressed data length: %d", len);
+                ret = -EINVAL;
+                break;
+            }
+            qemu_get_buffer(f, compressed_data_buf, len);
+            decompress_data_with_multi_threads(compressed_data_buf, host, len);
+            break;
+        case RAM_SAVE_FLAG_XBZRLE:
+            host = host_from_stream_offset(f, addr, flags);
+            if (!host) {
+                error_report("Illegal RAM offset " RAM_ADDR_FMT, addr);
+                ret = -EINVAL;
+                break;
+            }
+            if (load_xbzrle(f, addr, host) < 0) {
+                error_report("Failed to decompress XBZRLE page at "
+                             RAM_ADDR_FMT, addr);
+                ret = -EINVAL;
+                break;
+            }
+            break;
+        case RAM_SAVE_FLAG_EOS:
+            /* normal exit */
+            break;
+        default:
+            if (flags & RAM_SAVE_FLAG_HOOK) {
+                ram_control_load_hook(f, flags);
+            } else {
+                error_report("Unknown combination of migration flags: %#x",
+                             flags);
+                ret = -EINVAL;
+            }
+        }
+        if (!ret) {
+            ret = qemu_file_get_error(f);
+        }
+    }
+
+    rcu_read_unlock();
+    DPRINTF("Completed load of VM with exit code %d seq iteration "
+            "%" PRIu64 "\n", ret, seq_iter);
+    return ret;
+}
+
+static SaveVMHandlers savevm_ram_handlers = {
+    .save_live_setup = ram_save_setup,
+    .save_live_iterate = ram_save_iterate,
+    .save_live_complete = ram_save_complete,
+    .save_live_pending = ram_save_pending,
+    .load_state = ram_load,
+    .cancel = ram_migration_cancel,
+};
+
+void ram_mig_init(void)
+{
+    qemu_mutex_init(&XBZRLE.lock);
+    register_savevm_live(NULL, "ram", 0, 4, &savevm_ram_handlers, NULL);
+}
+/* Stub function that's gets run on the vcpu when its brought out of the
+   VM to run inside qemu via async_run_on_cpu()*/
+
+static void mig_sleep_cpu(void *opq)
+{
+    qemu_mutex_unlock_iothread();
+    g_usleep(30*1000);
+    qemu_mutex_lock_iothread();
+}
+
+/* To reduce the dirty rate explicitly disallow the VCPUs from spending
+   much time in the VM. The migration thread will try to catchup.
+   Workload will experience a performance drop.
+*/
+static void mig_throttle_guest_down(void)
+{
+    CPUState *cpu;
+
+    qemu_mutex_lock_iothread();
+    CPU_FOREACH(cpu) {
+        async_run_on_cpu(cpu, mig_sleep_cpu, NULL);
+    }
+    qemu_mutex_unlock_iothread();
+}
+
+static void check_guest_throttling(void)
+{
+    static int64_t t0;
+    int64_t        t1;
+
+    if (!mig_throttle_on) {
+        return;
+    }
+
+    if (!t0)  {
+        t0 = qemu_clock_get_ns(QEMU_CLOCK_REALTIME);
+        return;
+    }
+
+    t1 = qemu_clock_get_ns(QEMU_CLOCK_REALTIME);
+
+    /* If it has been more than 40 ms since the last time the guest
+     * was throttled then do it again.
+     */
+    if (40 < (t1-t0)/1000000) {
+        mig_throttle_guest_down();
+        t0 = t1;
+    }
+}
diff --git a/migration/rdma.c b/migration/rdma.c
index 77e3444..48b3e64 100644
--- a/migration/rdma.c
+++ b/migration/rdma.c
@@ -236,13 +236,13 @@
  * corresponding RDMALocalBlock with
  * the information needed to perform the actual RDMA.
  */
-typedef struct QEMU_PACKED RDMARemoteBlock {
+typedef struct QEMU_PACKED RDMADestBlock {
     uint64_t remote_host_addr;
     uint64_t offset;
     uint64_t length;
     uint32_t remote_rkey;
     uint32_t padding;
-} RDMARemoteBlock;
+} RDMADestBlock;
 
 static uint64_t htonll(uint64_t v)
 {
@@ -258,20 +258,20 @@
     return ((uint64_t)ntohl(u.lv[0]) << 32) | (uint64_t) ntohl(u.lv[1]);
 }
 
-static void remote_block_to_network(RDMARemoteBlock *rb)
+static void dest_block_to_network(RDMADestBlock *db)
 {
-    rb->remote_host_addr = htonll(rb->remote_host_addr);
-    rb->offset = htonll(rb->offset);
-    rb->length = htonll(rb->length);
-    rb->remote_rkey = htonl(rb->remote_rkey);
+    db->remote_host_addr = htonll(db->remote_host_addr);
+    db->offset = htonll(db->offset);
+    db->length = htonll(db->length);
+    db->remote_rkey = htonl(db->remote_rkey);
 }
 
-static void network_to_remote_block(RDMARemoteBlock *rb)
+static void network_to_dest_block(RDMADestBlock *db)
 {
-    rb->remote_host_addr = ntohll(rb->remote_host_addr);
-    rb->offset = ntohll(rb->offset);
-    rb->length = ntohll(rb->length);
-    rb->remote_rkey = ntohl(rb->remote_rkey);
+    db->remote_host_addr = ntohll(db->remote_host_addr);
+    db->offset = ntohll(db->offset);
+    db->length = ntohll(db->length);
+    db->remote_rkey = ntohl(db->remote_rkey);
 }
 
 /*
@@ -350,7 +350,7 @@
      * Description of ram blocks used throughout the code.
      */
     RDMALocalBlocks local_ram_blocks;
-    RDMARemoteBlock *block;
+    RDMADestBlock  *dest_blocks;
 
     /*
      * Migration on *destination* started.
@@ -570,10 +570,10 @@
  * in advanced before the migration starts. This tells us where the RAM blocks
  * are so that we can register them individually.
  */
-static void qemu_rdma_init_one_block(void *host_addr,
+static int qemu_rdma_init_one_block(const char *block_name, void *host_addr,
     ram_addr_t block_offset, ram_addr_t length, void *opaque)
 {
-    rdma_add_block(opaque, host_addr, block_offset, length);
+    return rdma_add_block(opaque, host_addr, block_offset, length);
 }
 
 /*
@@ -590,7 +590,7 @@
     memset(local, 0, sizeof *local);
     qemu_ram_foreach_block(qemu_rdma_init_one_block, rdma);
     trace_qemu_rdma_init_ram_blocks(local->nb_blocks);
-    rdma->block = (RDMARemoteBlock *) g_malloc0(sizeof(RDMARemoteBlock) *
+    rdma->dest_blocks = (RDMADestBlock *) g_malloc0(sizeof(RDMADestBlock) *
                         rdma->local_ram_blocks.nb_blocks);
     local->init = true;
     return 0;
@@ -790,6 +790,13 @@
 
         for (x = 0; x < num_devices; x++) {
             verbs = ibv_open_device(dev_list[x]);
+            if (!verbs) {
+                if (errno == EPERM) {
+                    continue;
+                } else {
+                    return -EINVAL;
+                }
+            }
 
             if (ibv_query_port(verbs, 1, &port_attr)) {
                 ibv_close_device(verbs);
@@ -2177,8 +2184,8 @@
         rdma->connected = false;
     }
 
-    g_free(rdma->block);
-    rdma->block = NULL;
+    g_free(rdma->dest_blocks);
+    rdma->dest_blocks = NULL;
 
     for (idx = 0; idx < RDMA_WRID_MAX; idx++) {
         if (rdma->wr_data[idx].control_mr) {
@@ -2445,7 +2452,6 @@
 
     if (host_port) {
         rdma = g_malloc0(sizeof(RDMAContext));
-        memset(rdma, 0, sizeof(RDMAContext));
         rdma->current_index = -1;
         rdma->current_chunk = -1;
 
@@ -2967,25 +2973,25 @@
              * their "local" descriptions with what was sent.
              */
             for (i = 0; i < local->nb_blocks; i++) {
-                rdma->block[i].remote_host_addr =
+                rdma->dest_blocks[i].remote_host_addr =
                     (uintptr_t)(local->block[i].local_host_addr);
 
                 if (rdma->pin_all) {
-                    rdma->block[i].remote_rkey = local->block[i].mr->rkey;
+                    rdma->dest_blocks[i].remote_rkey = local->block[i].mr->rkey;
                 }
 
-                rdma->block[i].offset = local->block[i].offset;
-                rdma->block[i].length = local->block[i].length;
+                rdma->dest_blocks[i].offset = local->block[i].offset;
+                rdma->dest_blocks[i].length = local->block[i].length;
 
-                remote_block_to_network(&rdma->block[i]);
+                dest_block_to_network(&rdma->dest_blocks[i]);
             }
 
             blocks.len = rdma->local_ram_blocks.nb_blocks
-                                                * sizeof(RDMARemoteBlock);
+                                                * sizeof(RDMADestBlock);
 
 
             ret = qemu_rdma_post_send_control(rdma,
-                                        (uint8_t *) rdma->block, &blocks);
+                                        (uint8_t *) rdma->dest_blocks, &blocks);
 
             if (ret < 0) {
                 error_report("rdma migration: error sending remote info");
@@ -3141,7 +3147,7 @@
     if (flags == RAM_CONTROL_SETUP) {
         RDMAControlHeader resp = {.type = RDMA_CONTROL_RAM_BLOCKS_RESULT };
         RDMALocalBlocks *local = &rdma->local_ram_blocks;
-        int reg_result_idx, i, j, nb_remote_blocks;
+        int reg_result_idx, i, j, nb_dest_blocks;
 
         head.type = RDMA_CONTROL_RAM_BLOCKS_REQUEST;
         trace_qemu_rdma_registration_stop_ram();
@@ -3162,7 +3168,7 @@
             return ret;
         }
 
-        nb_remote_blocks = resp.len / sizeof(RDMARemoteBlock);
+        nb_dest_blocks = resp.len / sizeof(RDMADestBlock);
 
         /*
          * The protocol uses two different sets of rkeys (mutually exclusive):
@@ -3176,7 +3182,7 @@
          * and then propagates the remote ram block descriptions to his local copy.
          */
 
-        if (local->nb_blocks != nb_remote_blocks) {
+        if (local->nb_blocks != nb_dest_blocks) {
             ERROR(errp, "ram blocks mismatch #1! "
                         "Your QEMU command line parameters are probably "
                         "not identical on both the source and destination.");
@@ -3184,26 +3190,26 @@
         }
 
         qemu_rdma_move_header(rdma, reg_result_idx, &resp);
-        memcpy(rdma->block,
+        memcpy(rdma->dest_blocks,
             rdma->wr_data[reg_result_idx].control_curr, resp.len);
-        for (i = 0; i < nb_remote_blocks; i++) {
-            network_to_remote_block(&rdma->block[i]);
+        for (i = 0; i < nb_dest_blocks; i++) {
+            network_to_dest_block(&rdma->dest_blocks[i]);
 
             /* search local ram blocks */
             for (j = 0; j < local->nb_blocks; j++) {
-                if (rdma->block[i].offset != local->block[j].offset) {
+                if (rdma->dest_blocks[i].offset != local->block[j].offset) {
                     continue;
                 }
 
-                if (rdma->block[i].length != local->block[j].length) {
+                if (rdma->dest_blocks[i].length != local->block[j].length) {
                     ERROR(errp, "ram blocks mismatch #2! "
                         "Your QEMU command line parameters are probably "
                         "not identical on both the source and destination.");
                     return -EINVAL;
                 }
                 local->block[j].remote_host_addr =
-                        rdma->block[i].remote_host_addr;
-                local->block[j].remote_rkey = rdma->block[i].remote_rkey;
+                        rdma->dest_blocks[i].remote_host_addr;
+                local->block[j].remote_rkey = rdma->dest_blocks[i].remote_rkey;
                 break;
             }
 
diff --git a/savevm.c b/migration/savevm.c
similarity index 87%
rename from savevm.c
rename to migration/savevm.c
index 3b0e222..2091882 100644
--- a/savevm.c
+++ b/migration/savevm.c
@@ -2,6 +2,10 @@
  * QEMU System Emulator
  *
  * Copyright (c) 2003-2008 Fabrice Bellard
+ * Copyright (c) 2009-2015 Red Hat Inc
+ *
+ * Authors:
+ *  Juan Quintela <quintela@redhat.com>
  *
  * Permission is hereby granted, free of charge, to any person obtaining a copy
  * of this software and associated documentation files (the "Software"), to deal
@@ -51,6 +55,8 @@
 #define ARP_PTYPE_IP 0x0800
 #define ARP_OP_REQUEST_REV 0x3
 
+static bool skip_section_footers;
+
 static int announce_self_create(uint8_t *buf,
                                 uint8_t *mac_addr)
 {
@@ -235,10 +241,15 @@
     int is_ram;
 } SaveStateEntry;
 
+typedef struct SaveState {
+    QTAILQ_HEAD(, SaveStateEntry) handlers;
+    int global_section_id;
+} SaveState;
 
-static QTAILQ_HEAD(savevm_handlers, SaveStateEntry) savevm_handlers =
-    QTAILQ_HEAD_INITIALIZER(savevm_handlers);
-static int global_section_id;
+static SaveState savevm_state = {
+    .handlers = QTAILQ_HEAD_INITIALIZER(savevm_state.handlers),
+    .global_section_id = 0,
+};
 
 static void dump_vmstate_vmsd(FILE *out_file,
                               const VMStateDescription *vmsd, int indent,
@@ -263,11 +274,11 @@
 }
 
 static void dump_vmstate_vmss(FILE *out_file,
-                              const VMStateSubsection *subsection,
+                              const VMStateDescription **subsection,
                               int indent)
 {
-    if (subsection->vmsd != NULL) {
-        dump_vmstate_vmsd(out_file, subsection->vmsd, indent, true);
+    if (*subsection != NULL) {
+        dump_vmstate_vmsd(out_file, *subsection, indent, true);
     }
 }
 
@@ -308,12 +319,12 @@
         fprintf(out_file, "\n%*s]", indent, "");
     }
     if (vmsd->subsections != NULL) {
-        const VMStateSubsection *subsection = vmsd->subsections;
+        const VMStateDescription **subsection = vmsd->subsections;
         bool first;
 
         fprintf(out_file, ",\n%*s\"Subsections\": [\n", indent, "");
         first = true;
-        while (subsection->vmsd != NULL) {
+        while (*subsection != NULL) {
             if (!first) {
                 fprintf(out_file, ",\n");
             }
@@ -383,7 +394,7 @@
     SaveStateEntry *se;
     int instance_id = 0;
 
-    QTAILQ_FOREACH(se, &savevm_handlers, entry) {
+    QTAILQ_FOREACH(se, &savevm_state.handlers, entry) {
         if (strcmp(idstr, se->idstr) == 0
             && instance_id <= se->instance_id) {
             instance_id = se->instance_id + 1;
@@ -397,7 +408,7 @@
     SaveStateEntry *se;
     int instance_id = 0;
 
-    QTAILQ_FOREACH(se, &savevm_handlers, entry) {
+    QTAILQ_FOREACH(se, &savevm_state.handlers, entry) {
         if (!se->compat) {
             continue;
         }
@@ -425,7 +436,7 @@
 
     se = g_malloc0(sizeof(SaveStateEntry));
     se->version_id = version_id;
-    se->section_id = global_section_id++;
+    se->section_id = savevm_state.global_section_id++;
     se->ops = ops;
     se->opaque = opaque;
     se->vmsd = NULL;
@@ -457,7 +468,7 @@
     }
     assert(!se->compat || se->instance_id == 0);
     /* add at the end of list */
-    QTAILQ_INSERT_TAIL(&savevm_handlers, se, entry);
+    QTAILQ_INSERT_TAIL(&savevm_state.handlers, se, entry);
     return 0;
 }
 
@@ -491,9 +502,9 @@
     }
     pstrcat(id, sizeof(id), idstr);
 
-    QTAILQ_FOREACH_SAFE(se, &savevm_handlers, entry, new_se) {
+    QTAILQ_FOREACH_SAFE(se, &savevm_state.handlers, entry, new_se) {
         if (strcmp(se->idstr, id) == 0 && se->opaque == opaque) {
-            QTAILQ_REMOVE(&savevm_handlers, se, entry);
+            QTAILQ_REMOVE(&savevm_state.handlers, se, entry);
             if (se->compat) {
                 g_free(se->compat);
             }
@@ -515,7 +526,7 @@
 
     se = g_malloc0(sizeof(SaveStateEntry));
     se->version_id = vmsd->version_id;
-    se->section_id = global_section_id++;
+    se->section_id = savevm_state.global_section_id++;
     se->opaque = opaque;
     se->vmsd = vmsd;
     se->alias_id = alias_id;
@@ -543,7 +554,7 @@
     }
     assert(!se->compat || se->instance_id == 0);
     /* add at the end of list */
-    QTAILQ_INSERT_TAIL(&savevm_handlers, se, entry);
+    QTAILQ_INSERT_TAIL(&savevm_state.handlers, se, entry);
     return 0;
 }
 
@@ -552,9 +563,9 @@
 {
     SaveStateEntry *se, *new_se;
 
-    QTAILQ_FOREACH_SAFE(se, &savevm_handlers, entry, new_se) {
+    QTAILQ_FOREACH_SAFE(se, &savevm_state.handlers, entry, new_se) {
         if (se->vmsd == vmsd && se->opaque == opaque) {
-            QTAILQ_REMOVE(&savevm_handlers, se, entry);
+            QTAILQ_REMOVE(&savevm_state.handlers, se, entry);
             if (se->compat) {
                 g_free(se->compat);
             }
@@ -602,11 +613,84 @@
     vmstate_save_state(f, se->vmsd, se->opaque, vmdesc);
 }
 
+void savevm_skip_section_footers(void)
+{
+    skip_section_footers = true;
+}
+
+/*
+ * Write the header for device section (QEMU_VM_SECTION START/END/PART/FULL)
+ */
+static void save_section_header(QEMUFile *f, SaveStateEntry *se,
+                                uint8_t section_type)
+{
+    qemu_put_byte(f, section_type);
+    qemu_put_be32(f, se->section_id);
+
+    if (section_type == QEMU_VM_SECTION_FULL ||
+        section_type == QEMU_VM_SECTION_START) {
+        /* ID string */
+        size_t len = strlen(se->idstr);
+        qemu_put_byte(f, len);
+        qemu_put_buffer(f, (uint8_t *)se->idstr, len);
+
+        qemu_put_be32(f, se->instance_id);
+        qemu_put_be32(f, se->version_id);
+    }
+}
+
+/*
+ * Write a footer onto device sections that catches cases misformatted device
+ * sections.
+ */
+static void save_section_footer(QEMUFile *f, SaveStateEntry *se)
+{
+    if (!skip_section_footers) {
+        qemu_put_byte(f, QEMU_VM_SECTION_FOOTER);
+        qemu_put_be32(f, se->section_id);
+    }
+}
+
+/*
+ * Read a footer off the wire and check that it matches the expected section
+ *
+ * Returns: true if the footer was good
+ *          false if there is a problem (and calls error_report to say why)
+ */
+static bool check_section_footer(QEMUFile *f, SaveStateEntry *se)
+{
+    uint8_t read_mark;
+    uint32_t read_section_id;
+
+    if (skip_section_footers) {
+        /* No footer to check */
+        return true;
+    }
+
+    read_mark = qemu_get_byte(f);
+
+    if (read_mark != QEMU_VM_SECTION_FOOTER) {
+        error_report("Missing section footer for %s", se->idstr);
+        return false;
+    }
+
+    read_section_id = qemu_get_be32(f);
+    if (read_section_id != se->section_id) {
+        error_report("Mismatched section id in footer for %s -"
+                     " read 0x%x expected 0x%x",
+                     se->idstr, read_section_id, se->section_id);
+        return false;
+    }
+
+    /* All good */
+    return true;
+}
+
 bool qemu_savevm_state_blocked(Error **errp)
 {
     SaveStateEntry *se;
 
-    QTAILQ_FOREACH(se, &savevm_handlers, entry) {
+    QTAILQ_FOREACH(se, &savevm_state.handlers, entry) {
         if (se->vmsd && se->vmsd->unmigratable) {
             error_setg(errp, "State blocked by non-migratable device '%s'",
                        se->idstr);
@@ -616,6 +700,13 @@
     return false;
 }
 
+void qemu_savevm_state_header(QEMUFile *f)
+{
+    trace_savevm_state_header();
+    qemu_put_be32(f, QEMU_VM_FILE_MAGIC);
+    qemu_put_be32(f, QEMU_VM_FILE_VERSION);
+}
+
 void qemu_savevm_state_begin(QEMUFile *f,
                              const MigrationParams *params)
 {
@@ -623,19 +714,14 @@
     int ret;
 
     trace_savevm_state_begin();
-    QTAILQ_FOREACH(se, &savevm_handlers, entry) {
+    QTAILQ_FOREACH(se, &savevm_state.handlers, entry) {
         if (!se->ops || !se->ops->set_params) {
             continue;
         }
         se->ops->set_params(params, se->opaque);
     }
 
-    qemu_put_be32(f, QEMU_VM_FILE_MAGIC);
-    qemu_put_be32(f, QEMU_VM_FILE_VERSION);
-
-    QTAILQ_FOREACH(se, &savevm_handlers, entry) {
-        int len;
-
+    QTAILQ_FOREACH(se, &savevm_state.handlers, entry) {
         if (!se->ops || !se->ops->save_live_setup) {
             continue;
         }
@@ -644,19 +730,10 @@
                 continue;
             }
         }
-        /* Section type */
-        qemu_put_byte(f, QEMU_VM_SECTION_START);
-        qemu_put_be32(f, se->section_id);
-
-        /* ID string */
-        len = strlen(se->idstr);
-        qemu_put_byte(f, len);
-        qemu_put_buffer(f, (uint8_t *)se->idstr, len);
-
-        qemu_put_be32(f, se->instance_id);
-        qemu_put_be32(f, se->version_id);
+        save_section_header(f, se, QEMU_VM_SECTION_START);
 
         ret = se->ops->save_live_setup(f, se->opaque);
+        save_section_footer(f, se);
         if (ret < 0) {
             qemu_file_set_error(f, ret);
             break;
@@ -676,7 +753,7 @@
     int ret = 1;
 
     trace_savevm_state_iterate();
-    QTAILQ_FOREACH(se, &savevm_handlers, entry) {
+    QTAILQ_FOREACH(se, &savevm_state.handlers, entry) {
         if (!se->ops || !se->ops->save_live_iterate) {
             continue;
         }
@@ -689,12 +766,12 @@
             return 0;
         }
         trace_savevm_section_start(se->idstr, se->section_id);
-        /* Section type */
-        qemu_put_byte(f, QEMU_VM_SECTION_PART);
-        qemu_put_be32(f, se->section_id);
+
+        save_section_header(f, se, QEMU_VM_SECTION_PART);
 
         ret = se->ops->save_live_iterate(f, se->opaque);
         trace_savevm_section_end(se->idstr, se->section_id, ret);
+        save_section_footer(f, se);
 
         if (ret < 0) {
             qemu_file_set_error(f, ret);
@@ -727,7 +804,7 @@
 
     cpu_synchronize_all_states();
 
-    QTAILQ_FOREACH(se, &savevm_handlers, entry) {
+    QTAILQ_FOREACH(se, &savevm_state.handlers, entry) {
         if (!se->ops || !se->ops->save_live_complete) {
             continue;
         }
@@ -737,12 +814,12 @@
             }
         }
         trace_savevm_section_start(se->idstr, se->section_id);
-        /* Section type */
-        qemu_put_byte(f, QEMU_VM_SECTION_END);
-        qemu_put_be32(f, se->section_id);
+
+        save_section_header(f, se, QEMU_VM_SECTION_END);
 
         ret = se->ops->save_live_complete(f, se->opaque);
         trace_savevm_section_end(se->idstr, se->section_id, ret);
+        save_section_footer(f, se);
         if (ret < 0) {
             qemu_file_set_error(f, ret);
             return;
@@ -752,8 +829,7 @@
     vmdesc = qjson_new();
     json_prop_int(vmdesc, "page_size", TARGET_PAGE_SIZE);
     json_start_array(vmdesc, "devices");
-    QTAILQ_FOREACH(se, &savevm_handlers, entry) {
-        int len;
+    QTAILQ_FOREACH(se, &savevm_state.handlers, entry) {
 
         if ((!se->ops || !se->ops->save_state) && !se->vmsd) {
             continue;
@@ -764,22 +840,13 @@
         json_prop_str(vmdesc, "name", se->idstr);
         json_prop_int(vmdesc, "instance_id", se->instance_id);
 
-        /* Section type */
-        qemu_put_byte(f, QEMU_VM_SECTION_FULL);
-        qemu_put_be32(f, se->section_id);
-
-        /* ID string */
-        len = strlen(se->idstr);
-        qemu_put_byte(f, len);
-        qemu_put_buffer(f, (uint8_t *)se->idstr, len);
-
-        qemu_put_be32(f, se->instance_id);
-        qemu_put_be32(f, se->version_id);
+        save_section_header(f, se, QEMU_VM_SECTION_FULL);
 
         vmstate_save(f, se, vmdesc);
 
         json_end_object(vmdesc);
         trace_savevm_section_end(se->idstr, se->section_id, 0);
+        save_section_footer(f, se);
     }
 
     qemu_put_byte(f, QEMU_VM_EOF);
@@ -803,7 +870,7 @@
     SaveStateEntry *se;
     uint64_t ret = 0;
 
-    QTAILQ_FOREACH(se, &savevm_handlers, entry) {
+    QTAILQ_FOREACH(se, &savevm_state.handlers, entry) {
         if (!se->ops || !se->ops->save_live_pending) {
             continue;
         }
@@ -822,7 +889,7 @@
     SaveStateEntry *se;
 
     trace_savevm_state_cancel();
-    QTAILQ_FOREACH(se, &savevm_handlers, entry) {
+    QTAILQ_FOREACH(se, &savevm_state.handlers, entry) {
         if (se->ops && se->ops->cancel) {
             se->ops->cancel(se->opaque);
         }
@@ -842,6 +909,7 @@
     }
 
     qemu_mutex_unlock_iothread();
+    qemu_savevm_state_header(f);
     qemu_savevm_state_begin(f, &params);
     qemu_mutex_lock_iothread();
 
@@ -872,9 +940,7 @@
 
     cpu_synchronize_all_states();
 
-    QTAILQ_FOREACH(se, &savevm_handlers, entry) {
-        int len;
-
+    QTAILQ_FOREACH(se, &savevm_state.handlers, entry) {
         if (se->is_ram) {
             continue;
         }
@@ -882,19 +948,11 @@
             continue;
         }
 
-        /* Section type */
-        qemu_put_byte(f, QEMU_VM_SECTION_FULL);
-        qemu_put_be32(f, se->section_id);
-
-        /* ID string */
-        len = strlen(se->idstr);
-        qemu_put_byte(f, len);
-        qemu_put_buffer(f, (uint8_t *)se->idstr, len);
-
-        qemu_put_be32(f, se->instance_id);
-        qemu_put_be32(f, se->version_id);
+        save_section_header(f, se, QEMU_VM_SECTION_FULL);
 
         vmstate_save(f, se, NULL);
+
+        save_section_footer(f, se);
     }
 
     qemu_put_byte(f, QEMU_VM_EOF);
@@ -906,7 +964,7 @@
 {
     SaveStateEntry *se;
 
-    QTAILQ_FOREACH(se, &savevm_handlers, entry) {
+    QTAILQ_FOREACH(se, &savevm_state.handlers, entry) {
         if (!strcmp(se->idstr, idstr) &&
             (instance_id == se->instance_id ||
              instance_id == se->alias_id))
@@ -922,18 +980,26 @@
     return NULL;
 }
 
-typedef struct LoadStateEntry {
+struct LoadStateEntry {
     QLIST_ENTRY(LoadStateEntry) entry;
     SaveStateEntry *se;
     int section_id;
     int version_id;
-} LoadStateEntry;
+};
+
+void loadvm_free_handlers(MigrationIncomingState *mis)
+{
+    LoadStateEntry *le, *new_le;
+
+    QLIST_FOREACH_SAFE(le, &mis->loadvm_handlers, entry, new_le) {
+        QLIST_REMOVE(le, entry);
+        g_free(le);
+    }
+}
 
 int qemu_loadvm_state(QEMUFile *f)
 {
-    QLIST_HEAD(, LoadStateEntry) loadvm_handlers =
-        QLIST_HEAD_INITIALIZER(loadvm_handlers);
-    LoadStateEntry *le, *new_le;
+    MigrationIncomingState *mis = migration_incoming_get_current();
     Error *local_err = NULL;
     uint8_t section_type;
     unsigned int v;
@@ -964,8 +1030,8 @@
     while ((section_type = qemu_get_byte(f)) != QEMU_VM_EOF) {
         uint32_t instance_id, version_id, section_id;
         SaveStateEntry *se;
-        char idstr[257];
-        int len;
+        LoadStateEntry *le;
+        char idstr[256];
 
         trace_qemu_loadvm_state_section(section_type);
         switch (section_type) {
@@ -973,9 +1039,11 @@
         case QEMU_VM_SECTION_FULL:
             /* Read section start */
             section_id = qemu_get_be32(f);
-            len = qemu_get_byte(f);
-            qemu_get_buffer(f, (uint8_t *)idstr, len);
-            idstr[len] = 0;
+            if (!qemu_get_counted_string(f, idstr)) {
+                error_report("Unable to read ID string for section %u",
+                            section_id);
+                return -EINVAL;
+            }
             instance_id = qemu_get_be32(f);
             version_id = qemu_get_be32(f);
 
@@ -1004,7 +1072,7 @@
             le->se = se;
             le->section_id = section_id;
             le->version_id = version_id;
-            QLIST_INSERT_HEAD(&loadvm_handlers, le, entry);
+            QLIST_INSERT_HEAD(&mis->loadvm_handlers, le, entry);
 
             ret = vmstate_load(f, le->se, le->version_id);
             if (ret < 0) {
@@ -1012,13 +1080,17 @@
                              " device '%s'", instance_id, idstr);
                 goto out;
             }
+            if (!check_section_footer(f, le->se)) {
+                ret = -EINVAL;
+                goto out;
+            }
             break;
         case QEMU_VM_SECTION_PART:
         case QEMU_VM_SECTION_END:
             section_id = qemu_get_be32(f);
 
             trace_qemu_loadvm_state_section_partend(section_id);
-            QLIST_FOREACH(le, &loadvm_handlers, entry) {
+            QLIST_FOREACH(le, &mis->loadvm_handlers, entry) {
                 if (le->section_id == section_id) {
                     break;
                 }
@@ -1035,6 +1107,10 @@
                              section_id, le->se->idstr);
                 goto out;
             }
+            if (!check_section_footer(f, le->se)) {
+                ret = -EINVAL;
+                goto out;
+            }
             break;
         default:
             error_report("Unknown savevm section type %d", section_type);
@@ -1066,11 +1142,6 @@
     ret = 0;
 
 out:
-    QLIST_FOREACH_SAFE(le, &loadvm_handlers, entry, new_le) {
-        QLIST_REMOVE(le, entry);
-        g_free(le);
-    }
-
     if (ret == 0) {
         /* We may not have a VMDESC section, so ignore relative errors */
         ret = file_error_after_eof;
@@ -1314,9 +1385,11 @@
     }
 
     qemu_system_reset(VMRESET_SILENT);
+    migration_incoming_state_new(f);
     ret = qemu_loadvm_state(f);
 
     qemu_fclose(f);
+    migration_incoming_state_destroy();
     if (ret < 0) {
         error_report("Error %d while loading VM state", ret);
         return ret;
diff --git a/migration/vmstate.c b/migration/vmstate.c
index e5388f0..6138d1a 100644
--- a/migration/vmstate.c
+++ b/migration/vmstate.c
@@ -341,11 +341,11 @@
 }
 
 static const VMStateDescription *
-    vmstate_get_subsection(const VMStateSubsection *sub, char *idstr)
+vmstate_get_subsection(const VMStateDescription **sub, char *idstr)
 {
-    while (sub && sub->needed) {
-        if (strcmp(idstr, sub->vmsd->name) == 0) {
-            return sub->vmsd;
+    while (sub && *sub && (*sub)->needed) {
+        if (strcmp(idstr, (*sub)->name) == 0) {
+            return *sub;
         }
         sub++;
     }
@@ -358,7 +358,7 @@
     trace_vmstate_subsection_load(vmsd->name);
 
     while (qemu_peek_byte(f, 0) == QEMU_VM_SUBSECTION) {
-        char idstr[256];
+        char idstr[256], *idstr_ret;
         int ret;
         uint8_t version_id, len, size;
         const VMStateDescription *sub_vmsd;
@@ -369,11 +369,12 @@
             trace_vmstate_subsection_load_bad(vmsd->name, "(short)");
             return 0;
         }
-        size = qemu_peek_buffer(f, (uint8_t *)idstr, len, 2);
+        size = qemu_peek_buffer(f, (uint8_t **)&idstr_ret, len, 2);
         if (size != len) {
             trace_vmstate_subsection_load_bad(vmsd->name, "(peek fail)");
             return 0;
         }
+        memcpy(idstr, idstr_ret, size);
         idstr[size] = 0;
 
         if (strncmp(vmsd->name, idstr, strlen(vmsd->name)) != 0) {
@@ -405,12 +406,12 @@
 static void vmstate_subsection_save(QEMUFile *f, const VMStateDescription *vmsd,
                                     void *opaque, QJSON *vmdesc)
 {
-    const VMStateSubsection *sub = vmsd->subsections;
+    const VMStateDescription **sub = vmsd->subsections;
     bool subsection_found = false;
 
-    while (sub && sub->needed) {
-        if (sub->needed(opaque)) {
-            const VMStateDescription *vmsd = sub->vmsd;
+    while (sub && *sub && (*sub)->needed) {
+        if ((*sub)->needed(opaque)) {
+            const VMStateDescription *vmsd = *sub;
             uint8_t len;
 
             if (vmdesc) {
diff --git a/qemu-options.hx b/qemu-options.hx
index 4be98f7..1d281f6 100644
--- a/qemu-options.hx
+++ b/qemu-options.hx
@@ -2686,6 +2686,17 @@
 @table @option
 ETEXI
 
+DEF("fw_cfg", HAS_ARG, QEMU_OPTION_fwcfg,
+    "-fw_cfg [name=]<name>,file=<file>\n"
+    "                add named fw_cfg entry from file\n",
+    QEMU_ARCH_ALL)
+STEXI
+@item -fw_cfg [name=]@var{name},file=@var{file}
+@findex -fw_cfg
+Add named fw_cfg entry from file. @var{name} determines the name of
+the entry in the fw_cfg file directory exposed to the guest.
+ETEXI
+
 DEF("serial", HAS_ARG, QEMU_OPTION_serial, \
     "-serial dev     redirect the serial port to char device 'dev'\n",
     QEMU_ARCH_ALL)
diff --git a/scripts/analyze-migration.py b/scripts/analyze-migration.py
index 0c8b22f..f6894be 100755
--- a/scripts/analyze-migration.py
+++ b/scripts/analyze-migration.py
@@ -474,6 +474,7 @@
     QEMU_VM_SECTION_FULL  = 0x04
     QEMU_VM_SUBSECTION    = 0x05
     QEMU_VM_VMDESCRIPTION = 0x06
+    QEMU_VM_SECTION_FOOTER= 0x7e
 
     def __init__(self, filename):
         self.section_classes = { ( 'ram', 0 ) : [ RamSection, None ],
@@ -526,6 +527,10 @@
             elif section_type == self.QEMU_VM_SECTION_PART or section_type == self.QEMU_VM_SECTION_END:
                 section_id = file.read32()
                 self.sections[section_id].read()
+            elif section_type == self.QEMU_VM_SECTION_FOOTER:
+                read_section_id = file.read32()
+                if read_section_id != section_id:
+                    raise Exception("Mismatched section footer: %x vs %x" % (read_section_id, section_id))
             else:
                 raise Exception("Unknown section type: %d" % section_type)
         file.close()
diff --git a/softmmu_template.h b/softmmu_template.h
index 39f571b..d42d89d 100644
--- a/softmmu_template.h
+++ b/softmmu_template.h
@@ -548,6 +548,28 @@
     helper_te_st_name(env, addr, val, oi, GETRA());
 }
 
+#if DATA_SIZE == 1
+/* Probe for whether the specified guest write access is permitted.
+ * If it is not permitted then an exception will be taken in the same
+ * way as if this were a real write access (and we will not return).
+ * Otherwise the function will return, and there will be a valid
+ * entry in the TLB for this access.
+ */
+void probe_write(CPUArchState *env, target_ulong addr, int mmu_idx,
+                 uintptr_t retaddr)
+{
+    int index = (addr >> TARGET_PAGE_BITS) & (CPU_TLB_SIZE - 1);
+    target_ulong tlb_addr = env->tlb_table[mmu_idx][index].addr_write;
+
+    if ((addr & TARGET_PAGE_MASK)
+        != (tlb_addr & (TARGET_PAGE_MASK | TLB_INVALID_MASK))) {
+        /* TLB entry is for a different page */
+        if (!VICTIM_TLB_HIT(addr_write)) {
+            tlb_fill(ENV_GET_CPU(env), addr, MMU_DATA_STORE, mmu_idx, retaddr);
+        }
+    }
+}
+#endif
 #endif /* !defined(SOFTMMU_CODE_ACCESS) */
 
 #undef READ_ACCESS_TYPE
diff --git a/target-arm/machine.c b/target-arm/machine.c
index 9446e5a..36365a5 100644
--- a/target-arm/machine.c
+++ b/target-arm/machine.c
@@ -40,6 +40,7 @@
     .name = "cpu/vfp",
     .version_id = 3,
     .minimum_version_id = 3,
+    .needed = vfp_needed,
     .fields = (VMStateField[]) {
         VMSTATE_FLOAT64_ARRAY(env.vfp.regs, ARMCPU, 64),
         /* The xregs array is a little awkward because element 1 (FPSCR)
@@ -72,6 +73,7 @@
     .name = "cpu/iwmmxt",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = iwmmxt_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64_ARRAY(env.iwmmxt.regs, ARMCPU, 16),
         VMSTATE_UINT32_ARRAY(env.iwmmxt.cregs, ARMCPU, 16),
@@ -91,6 +93,7 @@
     .name = "cpu/m",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = m_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT32(env.v7m.other_sp, ARMCPU),
         VMSTATE_UINT32(env.v7m.vecbase, ARMCPU),
@@ -114,6 +117,7 @@
     .name = "cpu/thumb2ee",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = thumb2ee_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT32(env.teecr, ARMCPU),
         VMSTATE_UINT32(env.teehbr, ARMCPU),
@@ -282,21 +286,11 @@
         VMSTATE_BOOL(powered_off, ARMCPU),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_vfp,
-            .needed = vfp_needed,
-        } , {
-            .vmsd = &vmstate_iwmmxt,
-            .needed = iwmmxt_needed,
-        } , {
-            .vmsd = &vmstate_m,
-            .needed = m_needed,
-        } , {
-            .vmsd = &vmstate_thumb2ee,
-            .needed = thumb2ee_needed,
-        } , {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_vfp,
+        &vmstate_iwmmxt,
+        &vmstate_m,
+        &vmstate_thumb2ee,
+        NULL
     }
 };
diff --git a/target-i386/machine.c b/target-i386/machine.c
index 69d86cb..a0df64b 100644
--- a/target-i386/machine.c
+++ b/target-i386/machine.c
@@ -403,6 +403,7 @@
     .name = "cpu/steal_time_msr",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = steal_time_msr_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64(env.steal_time_msr, X86CPU),
         VMSTATE_END_OF_LIST()
@@ -413,6 +414,7 @@
     .name = "cpu/async_pf_msr",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = async_pf_msr_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64(env.async_pf_en_msr, X86CPU),
         VMSTATE_END_OF_LIST()
@@ -423,6 +425,7 @@
     .name = "cpu/async_pv_eoi_msr",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = pv_eoi_msr_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64(env.pv_eoi_en_msr, X86CPU),
         VMSTATE_END_OF_LIST()
@@ -441,6 +444,7 @@
     .name = "cpu/fpop_ip_dp",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = fpop_ip_dp_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT16(env.fpop, X86CPU),
         VMSTATE_UINT64(env.fpip, X86CPU),
@@ -461,6 +465,7 @@
     .name = "cpu/msr_tsc_adjust",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = tsc_adjust_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64(env.tsc_adjust, X86CPU),
         VMSTATE_END_OF_LIST()
@@ -479,6 +484,7 @@
     .name = "cpu/msr_tscdeadline",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = tscdeadline_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64(env.tsc_deadline, X86CPU),
         VMSTATE_END_OF_LIST()
@@ -505,6 +511,7 @@
     .name = "cpu/msr_ia32_misc_enable",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = misc_enable_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64(env.msr_ia32_misc_enable, X86CPU),
         VMSTATE_END_OF_LIST()
@@ -515,6 +522,7 @@
     .name = "cpu/msr_ia32_feature_control",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = feature_control_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64(env.msr_ia32_feature_control, X86CPU),
         VMSTATE_END_OF_LIST()
@@ -549,6 +557,7 @@
     .name = "cpu/msr_architectural_pmu",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = pmu_enable_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64(env.msr_fixed_ctr_ctrl, X86CPU),
         VMSTATE_UINT64(env.msr_global_ctrl, X86CPU),
@@ -584,6 +593,7 @@
     .name = "cpu/mpx",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = mpx_needed,
     .fields = (VMStateField[]) {
         VMSTATE_BND_REGS(env.bnd_regs, X86CPU, 4),
         VMSTATE_UINT64(env.bndcs_regs.cfgu, X86CPU),
@@ -605,6 +615,7 @@
     .name = "cpu/msr_hyperv_hypercall",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = hyperv_hypercall_enable_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64(env.msr_hv_guest_os_id, X86CPU),
         VMSTATE_UINT64(env.msr_hv_hypercall, X86CPU),
@@ -624,6 +635,7 @@
     .name = "cpu/msr_hyperv_vapic",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = hyperv_vapic_enable_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64(env.msr_hv_vapic, X86CPU),
         VMSTATE_END_OF_LIST()
@@ -642,6 +654,7 @@
     .name = "cpu/msr_hyperv_time",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = hyperv_time_enable_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64(env.msr_hv_tsc, X86CPU),
         VMSTATE_END_OF_LIST()
@@ -683,6 +696,7 @@
     .name = "cpu/avx512",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = avx512_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64_ARRAY(env.opmask_regs, X86CPU, NB_OPMASK_REGS),
         VMSTATE_ZMMH_REGS_VARS(env.xmm_regs, X86CPU, 0),
@@ -705,6 +719,7 @@
     .name = "cpu/xss",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = xss_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64(env.xss, X86CPU),
         VMSTATE_END_OF_LIST()
@@ -813,54 +828,22 @@
         VMSTATE_END_OF_LIST()
         /* The above list is not sorted /wrt version numbers, watch out! */
     },
-    .subsections = (VMStateSubsection []) {
-        {
-            .vmsd = &vmstate_async_pf_msr,
-            .needed = async_pf_msr_needed,
-        } , {
-            .vmsd = &vmstate_pv_eoi_msr,
-            .needed = pv_eoi_msr_needed,
-        } , {
-            .vmsd = &vmstate_steal_time_msr,
-            .needed = steal_time_msr_needed,
-        } , {
-            .vmsd = &vmstate_fpop_ip_dp,
-            .needed = fpop_ip_dp_needed,
-        }, {
-            .vmsd = &vmstate_msr_tsc_adjust,
-            .needed = tsc_adjust_needed,
-        }, {
-            .vmsd = &vmstate_msr_tscdeadline,
-            .needed = tscdeadline_needed,
-        }, {
-            .vmsd = &vmstate_msr_ia32_misc_enable,
-            .needed = misc_enable_needed,
-        }, {
-            .vmsd = &vmstate_msr_ia32_feature_control,
-            .needed = feature_control_needed,
-        }, {
-            .vmsd = &vmstate_msr_architectural_pmu,
-            .needed = pmu_enable_needed,
-        } , {
-            .vmsd = &vmstate_mpx,
-            .needed = mpx_needed,
-        }, {
-            .vmsd = &vmstate_msr_hypercall_hypercall,
-            .needed = hyperv_hypercall_enable_needed,
-        }, {
-            .vmsd = &vmstate_msr_hyperv_vapic,
-            .needed = hyperv_vapic_enable_needed,
-        }, {
-            .vmsd = &vmstate_msr_hyperv_time,
-            .needed = hyperv_time_enable_needed,
-        }, {
-            .vmsd = &vmstate_avx512,
-            .needed = avx512_needed,
-         }, {
-            .vmsd = &vmstate_xss,
-            .needed = xss_needed,
-        } , {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_async_pf_msr,
+        &vmstate_pv_eoi_msr,
+        &vmstate_steal_time_msr,
+        &vmstate_fpop_ip_dp,
+        &vmstate_msr_tsc_adjust,
+        &vmstate_msr_tscdeadline,
+        &vmstate_msr_ia32_misc_enable,
+        &vmstate_msr_ia32_feature_control,
+        &vmstate_msr_architectural_pmu,
+        &vmstate_mpx,
+        &vmstate_msr_hypercall_hypercall,
+        &vmstate_msr_hyperv_vapic,
+        &vmstate_msr_hyperv_time,
+        &vmstate_avx512,
+        &vmstate_xss,
+        NULL
     }
 };
diff --git a/target-mips/cpu.h b/target-mips/cpu.h
index f9d2b4c..474a0e3 100644
--- a/target-mips/cpu.h
+++ b/target-mips/cpu.h
@@ -34,7 +34,7 @@
     uint_fast16_t RI0:1;
     uint_fast16_t RI1:1;
     uint_fast16_t EHINV:1;
-    target_ulong PFN[2];
+    uint64_t PFN[2];
 };
 
 #if !defined(CONFIG_USER_ONLY)
@@ -100,6 +100,7 @@
     float_status fp_status;
     /* fpu implementation/revision register (fir) */
     uint32_t fcr0;
+#define FCR0_FREP 29
 #define FCR0_UFRP 28
 #define FCR0_F64 22
 #define FCR0_L 21
@@ -223,8 +224,14 @@
 
     uint32_t SEGBITS;
     uint32_t PABITS;
+#if defined(TARGET_MIPS64)
+# define PABITS_BASE 36
+#else
+# define PABITS_BASE 32
+#endif
     target_ulong SEGMask;
-    target_ulong PAMask;
+    uint64_t PAMask;
+#define PAMASK_BASE ((1ULL << PABITS_BASE) - 1)
 
     int32_t msair;
 #define MSAIR_ProcID    8
@@ -272,8 +279,8 @@
 #define CP0VPEOpt_DWX2	2
 #define CP0VPEOpt_DWX1	1
 #define CP0VPEOpt_DWX0	0
-    target_ulong CP0_EntryLo0;
-    target_ulong CP0_EntryLo1;
+    uint64_t CP0_EntryLo0;
+    uint64_t CP0_EntryLo1;
 #if defined(TARGET_MIPS64)
 # define CP0EnLo_RI 63
 # define CP0EnLo_XI 62
@@ -288,6 +295,7 @@
     int32_t CP0_PageGrain;
 #define CP0PG_RIE 31
 #define CP0PG_XIE 30
+#define CP0PG_ELPA 29
 #define CP0PG_IEC 27
     int32_t CP0_Wired;
     int32_t CP0_SRSConf0_rw_bitmask;
@@ -462,17 +470,21 @@
 #define CP0C5_CV         29
 #define CP0C5_EVA        28
 #define CP0C5_MSAEn      27
+#define CP0C5_UFE        9
+#define CP0C5_FRE        8
 #define CP0C5_SBRI       6
+#define CP0C5_MVH        5
+#define CP0C5_LLB        4
 #define CP0C5_UFR        2
 #define CP0C5_NFExists   0
     int32_t CP0_Config6;
     int32_t CP0_Config7;
     /* XXX: Maybe make LLAddr per-TC? */
-    target_ulong lladdr;
+    uint64_t lladdr;
     target_ulong llval;
     target_ulong llnewval;
     target_ulong llreg;
-    target_ulong CP0_LLAddr_rw_bitmask;
+    uint64_t CP0_LLAddr_rw_bitmask;
     int CP0_LLAddr_shift;
     target_ulong CP0_WatchLo[8];
     int32_t CP0_WatchHi[8];
@@ -499,7 +511,7 @@
 #define CP0DB_DSS  0
     target_ulong CP0_DEPC;
     int32_t CP0_Performance0;
-    int32_t CP0_TagLo;
+    uint64_t CP0_TagLo;
     int32_t CP0_DataLo;
     int32_t CP0_TagHi;
     int32_t CP0_DataHi;
@@ -514,7 +526,7 @@
 #define EXCP_INST_NOTAVAIL 0x2 /* No valid instruction word for BadInstr */
     uint32_t hflags;    /* CPU State */
     /* TMASK defines different execution modes */
-#define MIPS_HFLAG_TMASK  0x15807FF
+#define MIPS_HFLAG_TMASK  0x75807FF
 #define MIPS_HFLAG_MODE   0x00007 /* execution modes                    */
     /* The KSU flags must be the lowest bits in hflags. The flag order
        must be the same as defined for CP0 Status. This allows to use
@@ -561,6 +573,8 @@
 #define MIPS_HFLAG_SBRI  0x400000 /* R6 SDBBP causes RI excpt. in user mode */
 #define MIPS_HFLAG_FBNSLOT 0x800000 /* Forbidden slot                   */
 #define MIPS_HFLAG_MSA   0x1000000
+#define MIPS_HFLAG_FRE   0x2000000 /* FRE enabled */
+#define MIPS_HFLAG_ELPA  0x4000000
     target_ulong btarget;        /* Jump / branch target               */
     target_ulong bcond;          /* Branch condition (if needed)       */
 
@@ -796,6 +810,15 @@
     set_flush_inputs_to_zero(flush_to_zero, status);
 }
 
+static inline void restore_pamask(CPUMIPSState *env)
+{
+    if (env->hflags & MIPS_HFLAG_ELPA) {
+        env->PAMask = (1ULL << env->PABITS) - 1;
+    } else {
+        env->PAMask = PAMASK_BASE;
+    }
+}
+
 static inline void cpu_get_tb_cpu_state(CPUMIPSState *env, target_ulong *pc,
                                         target_ulong *cs_base, int *flags)
 {
@@ -843,7 +866,8 @@
     env->hflags &= ~(MIPS_HFLAG_COP1X | MIPS_HFLAG_64 | MIPS_HFLAG_CP0 |
                      MIPS_HFLAG_F64 | MIPS_HFLAG_FPU | MIPS_HFLAG_KSU |
                      MIPS_HFLAG_AWRAP | MIPS_HFLAG_DSP | MIPS_HFLAG_DSPR2 |
-                     MIPS_HFLAG_SBRI | MIPS_HFLAG_MSA);
+                     MIPS_HFLAG_SBRI | MIPS_HFLAG_MSA | MIPS_HFLAG_FRE |
+                     MIPS_HFLAG_ELPA);
     if (!(env->CP0_Status & (1 << CP0St_EXL)) &&
         !(env->CP0_Status & (1 << CP0St_ERL)) &&
         !(env->hflags & MIPS_HFLAG_DM)) {
@@ -924,6 +948,16 @@
             env->hflags |= MIPS_HFLAG_MSA;
         }
     }
+    if (env->active_fpu.fcr0 & (1 << FCR0_FREP)) {
+        if (env->CP0_Config5 & (1 << CP0C5_FRE)) {
+            env->hflags |= MIPS_HFLAG_FRE;
+        }
+    }
+    if (env->CP0_Config3 & (1 << CP0C3_LPA)) {
+        if (env->CP0_PageGrain & (1 << CP0PG_ELPA)) {
+            env->hflags |= MIPS_HFLAG_ELPA;
+        }
+    }
 }
 
 #ifndef CONFIG_USER_ONLY
diff --git a/target-mips/helper.h b/target-mips/helper.h
index 3bd0b02..8df98c7 100644
--- a/target-mips/helper.h
+++ b/target-mips/helper.h
@@ -348,6 +348,7 @@
 DEF_HELPER_1(di, tl, env)
 DEF_HELPER_1(ei, tl, env)
 DEF_HELPER_1(eret, void, env)
+DEF_HELPER_1(eretnc, void, env)
 DEF_HELPER_1(deret, void, env)
 #endif /* !CONFIG_USER_ONLY */
 DEF_HELPER_1(rdhwr_cpunum, tl, env)
@@ -931,5 +932,11 @@
 DEF_HELPER_4(msa_ffint_s_df, void, env, i32, i32, i32)
 DEF_HELPER_4(msa_ffint_u_df, void, env, i32, i32, i32)
 
-DEF_HELPER_5(msa_ld_df, void, env, i32, i32, i32, s32)
-DEF_HELPER_5(msa_st_df, void, env, i32, i32, i32, s32)
+#define MSALDST_PROTO(type)                         \
+DEF_HELPER_3(msa_ld_ ## type, void, env, i32, tl)   \
+DEF_HELPER_3(msa_st_ ## type, void, env, i32, tl)
+MSALDST_PROTO(b)
+MSALDST_PROTO(h)
+MSALDST_PROTO(w)
+MSALDST_PROTO(d)
+#undef MSALDST_PROTO
diff --git a/target-mips/machine.c b/target-mips/machine.c
index 7d1fa32..8fa755c 100644
--- a/target-mips/machine.c
+++ b/target-mips/machine.c
@@ -10,6 +10,7 @@
     restore_fp_status(env);
     restore_msa_fp_status(env);
     compute_hflags(env);
+    restore_pamask(env);
 
     return 0;
 }
@@ -142,8 +143,8 @@
     v->RI0 = (flags >> 13) & 1;
     v->XI1 = (flags >> 12) & 1;
     v->XI0 = (flags >> 11) & 1;
-    qemu_get_betls(f, &v->PFN[0]);
-    qemu_get_betls(f, &v->PFN[1]);
+    qemu_get_be64s(f, &v->PFN[0]);
+    qemu_get_be64s(f, &v->PFN[1]);
 
     return 0;
 }
@@ -169,8 +170,8 @@
     qemu_put_be32s(f, &v->PageMask);
     qemu_put_8s(f, &v->ASID);
     qemu_put_be16s(f, &flags);
-    qemu_put_betls(f, &v->PFN[0]);
-    qemu_put_betls(f, &v->PFN[1]);
+    qemu_put_be64s(f, &v->PFN[0]);
+    qemu_put_be64s(f, &v->PFN[1]);
 }
 
 const VMStateInfo vmstate_info_tlb = {
@@ -201,8 +202,8 @@
 
 const VMStateDescription vmstate_mips_cpu = {
     .name = "cpu",
-    .version_id = 6,
-    .minimum_version_id = 6,
+    .version_id = 7,
+    .minimum_version_id = 7,
     .post_load = cpu_post_load,
     .fields = (VMStateField[]) {
         /* Active TC */
@@ -237,8 +238,8 @@
         VMSTATE_UINTTL(env.CP0_VPESchedule, MIPSCPU),
         VMSTATE_UINTTL(env.CP0_VPEScheFBack, MIPSCPU),
         VMSTATE_INT32(env.CP0_VPEOpt, MIPSCPU),
-        VMSTATE_UINTTL(env.CP0_EntryLo0, MIPSCPU),
-        VMSTATE_UINTTL(env.CP0_EntryLo1, MIPSCPU),
+        VMSTATE_UINT64(env.CP0_EntryLo0, MIPSCPU),
+        VMSTATE_UINT64(env.CP0_EntryLo1, MIPSCPU),
         VMSTATE_UINTTL(env.CP0_Context, MIPSCPU),
         VMSTATE_INT32(env.CP0_PageMask, MIPSCPU),
         VMSTATE_INT32(env.CP0_PageGrain, MIPSCPU),
@@ -269,7 +270,7 @@
         VMSTATE_INT32(env.CP0_Config3, MIPSCPU),
         VMSTATE_INT32(env.CP0_Config6, MIPSCPU),
         VMSTATE_INT32(env.CP0_Config7, MIPSCPU),
-        VMSTATE_UINTTL(env.lladdr, MIPSCPU),
+        VMSTATE_UINT64(env.lladdr, MIPSCPU),
         VMSTATE_UINTTL_ARRAY(env.CP0_WatchLo, MIPSCPU, 8),
         VMSTATE_INT32_ARRAY(env.CP0_WatchHi, MIPSCPU, 8),
         VMSTATE_UINTTL(env.CP0_XContext, MIPSCPU),
@@ -277,7 +278,7 @@
         VMSTATE_INT32(env.CP0_Debug, MIPSCPU),
         VMSTATE_UINTTL(env.CP0_DEPC, MIPSCPU),
         VMSTATE_INT32(env.CP0_Performance0, MIPSCPU),
-        VMSTATE_INT32(env.CP0_TagLo, MIPSCPU),
+        VMSTATE_UINT64(env.CP0_TagLo, MIPSCPU),
         VMSTATE_INT32(env.CP0_DataLo, MIPSCPU),
         VMSTATE_INT32(env.CP0_TagHi, MIPSCPU),
         VMSTATE_INT32(env.CP0_DataHi, MIPSCPU),
diff --git a/target-mips/mips-defs.h b/target-mips/mips-defs.h
index 1784227..20aa87c 100644
--- a/target-mips/mips-defs.h
+++ b/target-mips/mips-defs.h
@@ -10,11 +10,11 @@
 
 #if defined(TARGET_MIPS64)
 #define TARGET_LONG_BITS 64
-#define TARGET_PHYS_ADDR_SPACE_BITS 36
+#define TARGET_PHYS_ADDR_SPACE_BITS 48
 #define TARGET_VIRT_ADDR_SPACE_BITS 42
 #else
 #define TARGET_LONG_BITS 32
-#define TARGET_PHYS_ADDR_SPACE_BITS 36
+#define TARGET_PHYS_ADDR_SPACE_BITS 40
 #define TARGET_VIRT_ADDR_SPACE_BITS 32
 #endif
 
diff --git a/target-mips/op_helper.c b/target-mips/op_helper.c
index 73a8e45..2a9ddff 100644
--- a/target-mips/op_helper.c
+++ b/target-mips/op_helper.c
@@ -90,10 +90,10 @@
     }                                                                   \
 }
 #endif
-HELPER_LD(lbu, ldub, uint8_t)
-HELPER_LD(lhu, lduw, uint16_t)
 HELPER_LD(lw, ldl, int32_t)
+#if defined(TARGET_MIPS64)
 HELPER_LD(ld, ldq, int64_t)
+#endif
 #undef HELPER_LD
 
 #if defined(CONFIG_USER_ONLY)
@@ -118,9 +118,10 @@
 }
 #endif
 HELPER_ST(sb, stb, uint8_t)
-HELPER_ST(sh, stw, uint16_t)
 HELPER_ST(sw, stl, uint32_t)
+#if defined(TARGET_MIPS64)
 HELPER_ST(sd, stq, uint64_t)
+#endif
 #undef HELPER_ST
 
 target_ulong helper_clo (target_ulong arg1)
@@ -1067,19 +1068,23 @@
     env->CP0_VPEOpt = arg1 & 0x0000ffff;
 }
 
+#define MTC0_ENTRYLO_MASK(env) ((env->PAMask >> 6) & 0x3FFFFFFF)
+
 void helper_mtc0_entrylo0(CPUMIPSState *env, target_ulong arg1)
 {
-    /* Large physaddr (PABITS) not implemented */
     /* 1k pages not implemented */
     target_ulong rxi = arg1 & (env->CP0_PageGrain & (3u << CP0PG_XIE));
-    env->CP0_EntryLo0 = (arg1 & 0x3FFFFFFF) | (rxi << (CP0EnLo_XI - 30));
+    env->CP0_EntryLo0 = (arg1 & MTC0_ENTRYLO_MASK(env))
+                        | (rxi << (CP0EnLo_XI - 30));
 }
 
 #if defined(TARGET_MIPS64)
+#define DMTC0_ENTRYLO_MASK(env) (env->PAMask >> 6)
+
 void helper_dmtc0_entrylo0(CPUMIPSState *env, uint64_t arg1)
 {
     uint64_t rxi = arg1 & ((env->CP0_PageGrain & (3ull << CP0PG_XIE)) << 32);
-    env->CP0_EntryLo0 = (arg1 & 0x3FFFFFFF) | rxi;
+    env->CP0_EntryLo0 = (arg1 & DMTC0_ENTRYLO_MASK(env)) | rxi;
 }
 #endif
 
@@ -1245,17 +1250,17 @@
 
 void helper_mtc0_entrylo1(CPUMIPSState *env, target_ulong arg1)
 {
-    /* Large physaddr (PABITS) not implemented */
     /* 1k pages not implemented */
     target_ulong rxi = arg1 & (env->CP0_PageGrain & (3u << CP0PG_XIE));
-    env->CP0_EntryLo1 = (arg1 & 0x3FFFFFFF) | (rxi << (CP0EnLo_XI - 30));
+    env->CP0_EntryLo1 = (arg1 & MTC0_ENTRYLO_MASK(env))
+                        | (rxi << (CP0EnLo_XI - 30));
 }
 
 #if defined(TARGET_MIPS64)
 void helper_dmtc0_entrylo1(CPUMIPSState *env, uint64_t arg1)
 {
     uint64_t rxi = arg1 & ((env->CP0_PageGrain & (3ull << CP0PG_XIE)) << 32);
-    env->CP0_EntryLo1 = (arg1 & 0x3FFFFFFF) | rxi;
+    env->CP0_EntryLo1 = (arg1 & DMTC0_ENTRYLO_MASK(env)) | rxi;
 }
 #endif
 
@@ -1278,10 +1283,11 @@
 void helper_mtc0_pagegrain(CPUMIPSState *env, target_ulong arg1)
 {
     /* SmartMIPS not implemented */
-    /* Large physaddr (PABITS) not implemented */
     /* 1k pages not implemented */
     env->CP0_PageGrain = (arg1 & env->CP0_PageGrain_rw_bitmask) |
                          (env->CP0_PageGrain & ~env->CP0_PageGrain_rw_bitmask);
+    compute_hflags(env);
+    restore_pamask(env);
 }
 
 void helper_mtc0_wired(CPUMIPSState *env, target_ulong arg1)
@@ -1825,6 +1831,16 @@
     }
 }
 
+static inline uint64_t get_tlb_pfn_from_entrylo(uint64_t entrylo)
+{
+#if defined(TARGET_MIPS64)
+    return extract64(entrylo, 6, 54);
+#else
+    return extract64(entrylo, 6, 24) | /* PFN */
+           (extract64(entrylo, 32, 32) << 24); /* PFNX */
+#endif
+}
+
 static void r4k_fill_tlb(CPUMIPSState *env, int idx)
 {
     r4k_tlb_t *tlb;
@@ -1848,13 +1864,13 @@
     tlb->C0 = (env->CP0_EntryLo0 >> 3) & 0x7;
     tlb->XI0 = (env->CP0_EntryLo0 >> CP0EnLo_XI) & 1;
     tlb->RI0 = (env->CP0_EntryLo0 >> CP0EnLo_RI) & 1;
-    tlb->PFN[0] = (env->CP0_EntryLo0 >> 6) << 12;
+    tlb->PFN[0] = get_tlb_pfn_from_entrylo(env->CP0_EntryLo0) << 12;
     tlb->V1 = (env->CP0_EntryLo1 & 2) != 0;
     tlb->D1 = (env->CP0_EntryLo1 & 4) != 0;
     tlb->C1 = (env->CP0_EntryLo1 >> 3) & 0x7;
     tlb->XI1 = (env->CP0_EntryLo1 >> CP0EnLo_XI) & 1;
     tlb->RI1 = (env->CP0_EntryLo1 >> CP0EnLo_RI) & 1;
-    tlb->PFN[1] = (env->CP0_EntryLo1 >> 6) << 12;
+    tlb->PFN[1] = get_tlb_pfn_from_entrylo(env->CP0_EntryLo1) << 12;
 }
 
 void r4k_helper_tlbinv(CPUMIPSState *env)
@@ -1971,6 +1987,16 @@
     }
 }
 
+static inline uint64_t get_entrylo_pfn_from_tlb(uint64_t tlb_pfn)
+{
+#if defined(TARGET_MIPS64)
+    return tlb_pfn << 6;
+#else
+    return (extract64(tlb_pfn, 0, 24) << 6) | /* PFN */
+           (extract64(tlb_pfn, 24, 32) << 32); /* PFNX */
+#endif
+}
+
 void r4k_helper_tlbr(CPUMIPSState *env)
 {
     r4k_tlb_t *tlb;
@@ -1996,13 +2022,13 @@
         env->CP0_EntryHi = tlb->VPN | tlb->ASID;
         env->CP0_PageMask = tlb->PageMask;
         env->CP0_EntryLo0 = tlb->G | (tlb->V0 << 1) | (tlb->D0 << 2) |
-                        ((target_ulong)tlb->RI0 << CP0EnLo_RI) |
-                        ((target_ulong)tlb->XI0 << CP0EnLo_XI) |
-                        (tlb->C0 << 3) | (tlb->PFN[0] >> 6);
+                        ((uint64_t)tlb->RI0 << CP0EnLo_RI) |
+                        ((uint64_t)tlb->XI0 << CP0EnLo_XI) | (tlb->C0 << 3) |
+                        get_entrylo_pfn_from_tlb(tlb->PFN[0] >> 12);
         env->CP0_EntryLo1 = tlb->G | (tlb->V1 << 1) | (tlb->D1 << 2) |
-                        ((target_ulong)tlb->RI1 << CP0EnLo_RI) |
-                        ((target_ulong)tlb->XI1 << CP0EnLo_XI) |
-                        (tlb->C1 << 3) | (tlb->PFN[1] >> 6);
+                        ((uint64_t)tlb->RI1 << CP0EnLo_RI) |
+                        ((uint64_t)tlb->XI1 << CP0EnLo_XI) | (tlb->C1 << 3) |
+                        get_entrylo_pfn_from_tlb(tlb->PFN[1] >> 12);
     }
 }
 
@@ -2098,7 +2124,7 @@
     }
 }
 
-void helper_eret(CPUMIPSState *env)
+static inline void exception_return(CPUMIPSState *env)
 {
     debug_pre_eret(env);
     if (env->CP0_Status & (1 << CP0St_ERL)) {
@@ -2110,9 +2136,19 @@
     }
     compute_hflags(env);
     debug_post_eret(env);
+}
+
+void helper_eret(CPUMIPSState *env)
+{
+    exception_return(env);
     env->lladdr = 1;
 }
 
+void helper_eretnc(CPUMIPSState *env)
+{
+    exception_return(env);
+}
+
 void helper_deret(CPUMIPSState *env)
 {
     debug_pre_eret(env);
@@ -2303,6 +2339,16 @@
             }
         }
         break;
+    case 5:
+        /* FRE Support - read Config5.FRE bit */
+        if (env->active_fpu.fcr0 & (1 << FCR0_FREP)) {
+            if (env->CP0_Config5 & (1 << CP0C5_UFE)) {
+                arg1 = (env->CP0_Config5 >> CP0C5_FRE) & 1;
+            } else {
+                helper_raise_exception(env, EXCP_RI);
+            }
+        }
+        break;
     case 25:
         arg1 = ((env->active_fpu.fcr31 >> 24) & 0xfe) | ((env->active_fpu.fcr31 >> 23) & 0x1);
         break;
@@ -2347,6 +2393,30 @@
             helper_raise_exception(env, EXCP_RI);
         }
         break;
+    case 5:
+        /* FRE Support - clear Config5.FRE bit */
+        if (!((env->active_fpu.fcr0 & (1 << FCR0_FREP)) && (rt == 0))) {
+            return;
+        }
+        if (env->CP0_Config5 & (1 << CP0C5_UFE)) {
+            env->CP0_Config5 &= ~(1 << CP0C5_FRE);
+            compute_hflags(env);
+        } else {
+            helper_raise_exception(env, EXCP_RI);
+        }
+        break;
+    case 6:
+        /* FRE Support - set Config5.FRE bit */
+        if (!((env->active_fpu.fcr0 & (1 << FCR0_FREP)) && (rt == 0))) {
+            return;
+        }
+        if (env->CP0_Config5 & (1 << CP0C5_UFE)) {
+            env->CP0_Config5 |= (1 << CP0C5_FRE);
+            compute_hflags(env);
+        } else {
+            helper_raise_exception(env, EXCP_RI);
+        }
+        break;
     case 25:
         if ((env->insn_flags & ISA_MIPS32R6) || (arg1 & 0xffffff00)) {
             return;
@@ -3558,72 +3628,82 @@
 /* Element-by-element access macros */
 #define DF_ELEMENTS(df) (MSA_WRLEN / DF_BITS(df))
 
-void helper_msa_ld_df(CPUMIPSState *env, uint32_t df, uint32_t wd, uint32_t rs,
-                     int32_t s10)
-{
-    wr_t *pwd = &(env->active_fpu.fpr[wd].wr);
-    target_ulong addr = env->active_tc.gpr[rs] + (s10 << df);
-    int i;
+#if !defined(CONFIG_USER_ONLY)
+#define MEMOP_IDX(DF)                                           \
+        TCGMemOpIdx oi = make_memop_idx(MO_TE | DF | MO_UNALN,  \
+                                        cpu_mmu_index(env));
+#else
+#define MEMOP_IDX(DF)
+#endif
 
-    switch (df) {
-    case DF_BYTE:
-        for (i = 0; i < DF_ELEMENTS(DF_BYTE); i++) {
-            pwd->b[i] = do_lbu(env, addr + (i << DF_BYTE),
-                                env->hflags & MIPS_HFLAG_KSU);
-        }
-        break;
-    case DF_HALF:
-        for (i = 0; i < DF_ELEMENTS(DF_HALF); i++) {
-            pwd->h[i] = do_lhu(env, addr + (i << DF_HALF),
-                                env->hflags & MIPS_HFLAG_KSU);
-        }
-        break;
-    case DF_WORD:
-        for (i = 0; i < DF_ELEMENTS(DF_WORD); i++) {
-            pwd->w[i] = do_lw(env, addr + (i << DF_WORD),
-                                env->hflags & MIPS_HFLAG_KSU);
-        }
-        break;
-    case DF_DOUBLE:
-        for (i = 0; i < DF_ELEMENTS(DF_DOUBLE); i++) {
-            pwd->d[i] = do_ld(env, addr + (i << DF_DOUBLE),
-                                env->hflags & MIPS_HFLAG_KSU);
-        }
-        break;
-    }
+#define MSA_LD_DF(DF, TYPE, LD_INSN, ...)                               \
+void helper_msa_ld_ ## TYPE(CPUMIPSState *env, uint32_t wd,             \
+                            target_ulong addr)                          \
+{                                                                       \
+    wr_t *pwd = &(env->active_fpu.fpr[wd].wr);                          \
+    wr_t wx;                                                            \
+    int i;                                                              \
+    MEMOP_IDX(DF)                                                       \
+    for (i = 0; i < DF_ELEMENTS(DF); i++) {                             \
+        wx.TYPE[i] = LD_INSN(env, addr + (i << DF), ##__VA_ARGS__);     \
+    }                                                                   \
+    memcpy(pwd, &wx, sizeof(wr_t));                                     \
 }
 
-void helper_msa_st_df(CPUMIPSState *env, uint32_t df, uint32_t wd, uint32_t rs,
-                     int32_t s10)
-{
-    wr_t *pwd = &(env->active_fpu.fpr[wd].wr);
-    target_ulong addr = env->active_tc.gpr[rs] + (s10 << df);
-    int i;
+#if !defined(CONFIG_USER_ONLY)
+MSA_LD_DF(DF_BYTE,   b, helper_ret_ldub_mmu, oi, GETRA())
+MSA_LD_DF(DF_HALF,   h, helper_ret_lduw_mmu, oi, GETRA())
+MSA_LD_DF(DF_WORD,   w, helper_ret_ldul_mmu, oi, GETRA())
+MSA_LD_DF(DF_DOUBLE, d, helper_ret_ldq_mmu,  oi, GETRA())
+#else
+MSA_LD_DF(DF_BYTE,   b, cpu_ldub_data)
+MSA_LD_DF(DF_HALF,   h, cpu_lduw_data)
+MSA_LD_DF(DF_WORD,   w, cpu_ldl_data)
+MSA_LD_DF(DF_DOUBLE, d, cpu_ldq_data)
+#endif
 
-    switch (df) {
-    case DF_BYTE:
-        for (i = 0; i < DF_ELEMENTS(DF_BYTE); i++) {
-            do_sb(env, addr + (i << DF_BYTE), pwd->b[i],
-                    env->hflags & MIPS_HFLAG_KSU);
-        }
-        break;
-    case DF_HALF:
-        for (i = 0; i < DF_ELEMENTS(DF_HALF); i++) {
-            do_sh(env, addr + (i << DF_HALF), pwd->h[i],
-                    env->hflags & MIPS_HFLAG_KSU);
-        }
-        break;
-    case DF_WORD:
-        for (i = 0; i < DF_ELEMENTS(DF_WORD); i++) {
-            do_sw(env, addr + (i << DF_WORD), pwd->w[i],
-                    env->hflags & MIPS_HFLAG_KSU);
-        }
-        break;
-    case DF_DOUBLE:
-        for (i = 0; i < DF_ELEMENTS(DF_DOUBLE); i++) {
-            do_sd(env, addr + (i << DF_DOUBLE), pwd->d[i],
-                    env->hflags & MIPS_HFLAG_KSU);
-        }
-        break;
+#define MSA_PAGESPAN(x) \
+        ((((x) & ~TARGET_PAGE_MASK) + MSA_WRLEN/8 - 1) >= TARGET_PAGE_SIZE)
+
+static inline void ensure_writable_pages(CPUMIPSState *env,
+                                         target_ulong addr,
+                                         int mmu_idx,
+                                         uintptr_t retaddr)
+{
+#if !defined(CONFIG_USER_ONLY)
+    target_ulong page_addr;
+    if (unlikely(MSA_PAGESPAN(addr))) {
+        /* first page */
+        probe_write(env, addr, mmu_idx, retaddr);
+        /* second page */
+        page_addr = (addr & TARGET_PAGE_MASK) + TARGET_PAGE_SIZE;
+        probe_write(env, page_addr, mmu_idx, retaddr);
     }
+#endif
 }
+
+#define MSA_ST_DF(DF, TYPE, ST_INSN, ...)                               \
+void helper_msa_st_ ## TYPE(CPUMIPSState *env, uint32_t wd,             \
+                            target_ulong addr)                          \
+{                                                                       \
+    wr_t *pwd = &(env->active_fpu.fpr[wd].wr);                          \
+    int mmu_idx = cpu_mmu_index(env);                                   \
+    int i;                                                              \
+    MEMOP_IDX(DF)                                                       \
+    ensure_writable_pages(env, addr, mmu_idx, GETRA());                 \
+    for (i = 0; i < DF_ELEMENTS(DF); i++) {                             \
+        ST_INSN(env, addr + (i << DF), pwd->TYPE[i], ##__VA_ARGS__);    \
+    }                                                                   \
+}
+
+#if !defined(CONFIG_USER_ONLY)
+MSA_ST_DF(DF_BYTE,   b, helper_ret_stb_mmu, oi, GETRA())
+MSA_ST_DF(DF_HALF,   h, helper_ret_stw_mmu, oi, GETRA())
+MSA_ST_DF(DF_WORD,   w, helper_ret_stl_mmu, oi, GETRA())
+MSA_ST_DF(DF_DOUBLE, d, helper_ret_stq_mmu, oi, GETRA())
+#else
+MSA_ST_DF(DF_BYTE,   b, cpu_stb_data)
+MSA_ST_DF(DF_HALF,   h, cpu_stw_data)
+MSA_ST_DF(DF_WORD,   w, cpu_stl_data)
+MSA_ST_DF(DF_DOUBLE, d, cpu_stq_data)
+#endif
diff --git a/target-mips/translate.c b/target-mips/translate.c
index fd063a2..1d128ee 100644
--- a/target-mips/translate.c
+++ b/target-mips/translate.c
@@ -868,8 +868,10 @@
 enum {
     OPC_MFC0     = (0x00 << 21) | OPC_CP0,
     OPC_DMFC0    = (0x01 << 21) | OPC_CP0,
+    OPC_MFHC0    = (0x02 << 21) | OPC_CP0,
     OPC_MTC0     = (0x04 << 21) | OPC_CP0,
     OPC_DMTC0    = (0x05 << 21) | OPC_CP0,
+    OPC_MTHC0    = (0x06 << 21) | OPC_CP0,
     OPC_MFTR     = (0x08 << 21) | OPC_CP0,
     OPC_RDPGPR   = (0x0A << 21) | OPC_CP0,
     OPC_MFMC0    = (0x0B << 21) | OPC_CP0,
@@ -1414,6 +1416,7 @@
     int32_t CP0_Config1;
     /* Routine used to access memory */
     int mem_idx;
+    TCGMemOp default_tcg_memop_mask;
     uint32_t hflags, saved_hflags;
     int bstate;
     target_ulong btarget;
@@ -1423,6 +1426,9 @@
     int ie;
     bool bi;
     bool bp;
+    uint64_t PAMask;
+    bool mvh;
+    int CP0_LLAddr_shift;
 } DisasContext;
 
 enum {
@@ -1557,15 +1563,80 @@
     }
 }
 
-/* Floating point register moves. */
-static void gen_load_fpr32(TCGv_i32 t, int reg)
+/* Tests */
+static inline void gen_save_pc(target_ulong pc)
 {
+    tcg_gen_movi_tl(cpu_PC, pc);
+}
+
+static inline void save_cpu_state(DisasContext *ctx, int do_save_pc)
+{
+    LOG_DISAS("hflags %08x saved %08x\n", ctx->hflags, ctx->saved_hflags);
+    if (do_save_pc && ctx->pc != ctx->saved_pc) {
+        gen_save_pc(ctx->pc);
+        ctx->saved_pc = ctx->pc;
+    }
+    if (ctx->hflags != ctx->saved_hflags) {
+        tcg_gen_movi_i32(hflags, ctx->hflags);
+        ctx->saved_hflags = ctx->hflags;
+        switch (ctx->hflags & MIPS_HFLAG_BMASK_BASE) {
+        case MIPS_HFLAG_BR:
+            break;
+        case MIPS_HFLAG_BC:
+        case MIPS_HFLAG_BL:
+        case MIPS_HFLAG_B:
+            tcg_gen_movi_tl(btarget, ctx->btarget);
+            break;
+        }
+    }
+}
+
+static inline void restore_cpu_state(CPUMIPSState *env, DisasContext *ctx)
+{
+    ctx->saved_hflags = ctx->hflags;
+    switch (ctx->hflags & MIPS_HFLAG_BMASK_BASE) {
+    case MIPS_HFLAG_BR:
+        break;
+    case MIPS_HFLAG_BC:
+    case MIPS_HFLAG_BL:
+    case MIPS_HFLAG_B:
+        ctx->btarget = env->btarget;
+        break;
+    }
+}
+
+static inline void generate_exception_err(DisasContext *ctx, int excp, int err)
+{
+    TCGv_i32 texcp = tcg_const_i32(excp);
+    TCGv_i32 terr = tcg_const_i32(err);
+    save_cpu_state(ctx, 1);
+    gen_helper_raise_exception_err(cpu_env, texcp, terr);
+    tcg_temp_free_i32(terr);
+    tcg_temp_free_i32(texcp);
+}
+
+static inline void generate_exception(DisasContext *ctx, int excp)
+{
+    save_cpu_state(ctx, 1);
+    gen_helper_0e0i(raise_exception, excp);
+}
+
+/* Floating point register moves. */
+static void gen_load_fpr32(DisasContext *ctx, TCGv_i32 t, int reg)
+{
+    if (ctx->hflags & MIPS_HFLAG_FRE) {
+        generate_exception(ctx, EXCP_RI);
+    }
     tcg_gen_trunc_i64_i32(t, fpu_f64[reg]);
 }
 
-static void gen_store_fpr32(TCGv_i32 t, int reg)
+static void gen_store_fpr32(DisasContext *ctx, TCGv_i32 t, int reg)
 {
-    TCGv_i64 t64 = tcg_temp_new_i64();
+    TCGv_i64 t64;
+    if (ctx->hflags & MIPS_HFLAG_FRE) {
+        generate_exception(ctx, EXCP_RI);
+    }
+    t64 = tcg_temp_new_i64();
     tcg_gen_extu_i32_i64(t64, t);
     tcg_gen_deposit_i64(fpu_f64[reg], fpu_f64[reg], t64, 0, 32);
     tcg_temp_free_i64(t64);
@@ -1579,7 +1650,7 @@
         tcg_gen_trunc_i64_i32(t, t64);
         tcg_temp_free_i64(t64);
     } else {
-        gen_load_fpr32(t, reg | 1);
+        gen_load_fpr32(ctx, t, reg | 1);
     }
 }
 
@@ -1591,7 +1662,7 @@
         tcg_gen_deposit_i64(fpu_f64[reg], fpu_f64[reg], t64, 32, 32);
         tcg_temp_free_i64(t64);
     } else {
-        gen_store_fpr32(t, reg | 1);
+        gen_store_fpr32(ctx, t, reg | 1);
     }
 }
 
@@ -1626,66 +1697,6 @@
         return 23;
 }
 
-/* Tests */
-static inline void gen_save_pc(target_ulong pc)
-{
-    tcg_gen_movi_tl(cpu_PC, pc);
-}
-
-static inline void save_cpu_state (DisasContext *ctx, int do_save_pc)
-{
-    LOG_DISAS("hflags %08x saved %08x\n", ctx->hflags, ctx->saved_hflags);
-    if (do_save_pc && ctx->pc != ctx->saved_pc) {
-        gen_save_pc(ctx->pc);
-        ctx->saved_pc = ctx->pc;
-    }
-    if (ctx->hflags != ctx->saved_hflags) {
-        tcg_gen_movi_i32(hflags, ctx->hflags);
-        ctx->saved_hflags = ctx->hflags;
-        switch (ctx->hflags & MIPS_HFLAG_BMASK_BASE) {
-        case MIPS_HFLAG_BR:
-            break;
-        case MIPS_HFLAG_BC:
-        case MIPS_HFLAG_BL:
-        case MIPS_HFLAG_B:
-            tcg_gen_movi_tl(btarget, ctx->btarget);
-            break;
-        }
-    }
-}
-
-static inline void restore_cpu_state (CPUMIPSState *env, DisasContext *ctx)
-{
-    ctx->saved_hflags = ctx->hflags;
-    switch (ctx->hflags & MIPS_HFLAG_BMASK_BASE) {
-    case MIPS_HFLAG_BR:
-        break;
-    case MIPS_HFLAG_BC:
-    case MIPS_HFLAG_BL:
-    case MIPS_HFLAG_B:
-        ctx->btarget = env->btarget;
-        break;
-    }
-}
-
-static inline void
-generate_exception_err (DisasContext *ctx, int excp, int err)
-{
-    TCGv_i32 texcp = tcg_const_i32(excp);
-    TCGv_i32 terr = tcg_const_i32(err);
-    save_cpu_state(ctx, 1);
-    gen_helper_raise_exception_err(cpu_env, texcp, terr);
-    tcg_temp_free_i32(terr);
-    tcg_temp_free_i32(texcp);
-}
-
-static inline void
-generate_exception (DisasContext *ctx, int excp)
-{
-    save_cpu_state(ctx, 1);
-    gen_helper_0e0i(raise_exception, excp);
-}
-
 /* Addresses computation */
 static inline void gen_op_addr_add (DisasContext *ctx, TCGv ret, TCGv arg0, TCGv arg1)
 {
@@ -1815,11 +1826,20 @@
 }
 #endif
 
+#ifndef CONFIG_USER_ONLY
+static inline void check_mvh(DisasContext *ctx)
+{
+    if (unlikely(!ctx->mvh)) {
+        generate_exception(ctx, EXCP_RI);
+    }
+}
+#endif
+
 /* Define small wrappers for gen_load_fpr* so that we have a uniform
    calling interface for 32 and 64-bit FPRs.  No sense in changing
    all callers for gen_load_fpr32 when we need the CTX parameter for
    this one use.  */
-#define gen_ldcmp_fpr32(ctx, x, y) gen_load_fpr32(x, y)
+#define gen_ldcmp_fpr32(ctx, x, y) gen_load_fpr32(ctx, x, y)
 #define gen_ldcmp_fpr64(ctx, x, y) gen_load_fpr64(ctx, x, y)
 #define FOP_CONDS(type, abs, fmt, ifmt, bits)                                 \
 static inline void gen_cmp ## type ## _ ## fmt(DisasContext *ctx, int n,      \
@@ -1963,7 +1983,7 @@
 }
 
 FOP_CONDNS(d, FMT_D, 64, gen_store_fpr64(ctx, fp0, fd))
-FOP_CONDNS(s, FMT_S, 32, gen_store_fpr32(fp0, fd))
+FOP_CONDNS(s, FMT_S, 32, gen_store_fpr32(ctx, fp0, fd))
 #undef FOP_CONDNS
 #undef gen_ldcmp_fpr32
 #undef gen_ldcmp_fpr64
@@ -2081,12 +2101,14 @@
     switch (opc) {
 #if defined(TARGET_MIPS64)
     case OPC_LWU:
-        tcg_gen_qemu_ld_tl(t0, t0, ctx->mem_idx, MO_TEUL);
+        tcg_gen_qemu_ld_tl(t0, t0, ctx->mem_idx, MO_TEUL |
+                           ctx->default_tcg_memop_mask);
         gen_store_gpr(t0, rt);
         opn = "lwu";
         break;
     case OPC_LD:
-        tcg_gen_qemu_ld_tl(t0, t0, ctx->mem_idx, MO_TEQ);
+        tcg_gen_qemu_ld_tl(t0, t0, ctx->mem_idx, MO_TEQ |
+                           ctx->default_tcg_memop_mask);
         gen_store_gpr(t0, rt);
         opn = "ld";
         break;
@@ -2157,17 +2179,20 @@
         opn = "lwpc";
         break;
     case OPC_LW:
-        tcg_gen_qemu_ld_tl(t0, t0, ctx->mem_idx, MO_TESL);
+        tcg_gen_qemu_ld_tl(t0, t0, ctx->mem_idx, MO_TESL |
+                           ctx->default_tcg_memop_mask);
         gen_store_gpr(t0, rt);
         opn = "lw";
         break;
     case OPC_LH:
-        tcg_gen_qemu_ld_tl(t0, t0, ctx->mem_idx, MO_TESW);
+        tcg_gen_qemu_ld_tl(t0, t0, ctx->mem_idx, MO_TESW |
+                           ctx->default_tcg_memop_mask);
         gen_store_gpr(t0, rt);
         opn = "lh";
         break;
     case OPC_LHU:
-        tcg_gen_qemu_ld_tl(t0, t0, ctx->mem_idx, MO_TEUW);
+        tcg_gen_qemu_ld_tl(t0, t0, ctx->mem_idx, MO_TEUW |
+                           ctx->default_tcg_memop_mask);
         gen_store_gpr(t0, rt);
         opn = "lhu";
         break;
@@ -2251,7 +2276,8 @@
     switch (opc) {
 #if defined(TARGET_MIPS64)
     case OPC_SD:
-        tcg_gen_qemu_st_tl(t1, t0, ctx->mem_idx, MO_TEQ);
+        tcg_gen_qemu_st_tl(t1, t0, ctx->mem_idx, MO_TEQ |
+                           ctx->default_tcg_memop_mask);
         opn = "sd";
         break;
     case OPC_SDL:
@@ -2266,11 +2292,13 @@
         break;
 #endif
     case OPC_SW:
-        tcg_gen_qemu_st_tl(t1, t0, ctx->mem_idx, MO_TEUL);
+        tcg_gen_qemu_st_tl(t1, t0, ctx->mem_idx, MO_TEUL |
+                           ctx->default_tcg_memop_mask);
         opn = "sw";
         break;
     case OPC_SH:
-        tcg_gen_qemu_st_tl(t1, t0, ctx->mem_idx, MO_TEUW);
+        tcg_gen_qemu_st_tl(t1, t0, ctx->mem_idx, MO_TEUW |
+                           ctx->default_tcg_memop_mask);
         opn = "sh";
         break;
     case OPC_SB:
@@ -2347,8 +2375,9 @@
     case OPC_LWC1:
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
-            tcg_gen_qemu_ld_i32(fp0, t0, ctx->mem_idx, MO_TESL);
-            gen_store_fpr32(fp0, ft);
+            tcg_gen_qemu_ld_i32(fp0, t0, ctx->mem_idx, MO_TESL |
+                                ctx->default_tcg_memop_mask);
+            gen_store_fpr32(ctx, fp0, ft);
             tcg_temp_free_i32(fp0);
         }
         opn = "lwc1";
@@ -2356,8 +2385,9 @@
     case OPC_SWC1:
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
-            gen_load_fpr32(fp0, ft);
-            tcg_gen_qemu_st_i32(fp0, t0, ctx->mem_idx, MO_TEUL);
+            gen_load_fpr32(ctx, fp0, ft);
+            tcg_gen_qemu_st_i32(fp0, t0, ctx->mem_idx, MO_TEUL |
+                                ctx->default_tcg_memop_mask);
             tcg_temp_free_i32(fp0);
         }
         opn = "swc1";
@@ -2365,7 +2395,8 @@
     case OPC_LDC1:
         {
             TCGv_i64 fp0 = tcg_temp_new_i64();
-            tcg_gen_qemu_ld_i64(fp0, t0, ctx->mem_idx, MO_TEQ);
+            tcg_gen_qemu_ld_i64(fp0, t0, ctx->mem_idx, MO_TEQ |
+                                ctx->default_tcg_memop_mask);
             gen_store_fpr64(ctx, fp0, ft);
             tcg_temp_free_i64(fp0);
         }
@@ -2375,7 +2406,8 @@
         {
             TCGv_i64 fp0 = tcg_temp_new_i64();
             gen_load_fpr64(ctx, fp0, ft);
-            tcg_gen_qemu_st_i64(fp0, t0, ctx->mem_idx, MO_TEQ);
+            tcg_gen_qemu_st_i64(fp0, t0, ctx->mem_idx, MO_TEQ |
+                                ctx->default_tcg_memop_mask);
             tcg_temp_free_i64(fp0);
         }
         opn = "sdc1";
@@ -4815,6 +4847,69 @@
 
 #ifndef CONFIG_USER_ONLY
 /* CP0 (MMU and control) */
+static inline void gen_move_low32(TCGv ret, TCGv_i64 arg)
+{
+#if defined(TARGET_MIPS64)
+    tcg_gen_ext32s_tl(ret, arg);
+#else
+    tcg_gen_trunc_i64_tl(ret, arg);
+#endif
+}
+
+static inline void gen_mthc0_entrylo(TCGv arg, target_ulong off)
+{
+    TCGv_i64 t0 = tcg_temp_new_i64();
+    TCGv_i64 t1 = tcg_temp_new_i64();
+
+    tcg_gen_ext_tl_i64(t0, arg);
+    tcg_gen_ld_i64(t1, cpu_env, off);
+#if defined(TARGET_MIPS64)
+    tcg_gen_deposit_i64(t1, t1, t0, 30, 32);
+#else
+    tcg_gen_concat32_i64(t1, t1, t0);
+#endif
+    tcg_gen_st_i64(t1, cpu_env, off);
+    tcg_temp_free_i64(t1);
+    tcg_temp_free_i64(t0);
+}
+
+static inline void gen_mthc0_store64(TCGv arg, target_ulong off)
+{
+    TCGv_i64 t0 = tcg_temp_new_i64();
+    TCGv_i64 t1 = tcg_temp_new_i64();
+
+    tcg_gen_ext_tl_i64(t0, arg);
+    tcg_gen_ld_i64(t1, cpu_env, off);
+    tcg_gen_concat32_i64(t1, t1, t0);
+    tcg_gen_st_i64(t1, cpu_env, off);
+    tcg_temp_free_i64(t1);
+    tcg_temp_free_i64(t0);
+}
+
+static inline void gen_mfhc0_entrylo(TCGv arg, target_ulong off)
+{
+    TCGv_i64 t0 = tcg_temp_new_i64();
+
+    tcg_gen_ld_i64(t0, cpu_env, off);
+#if defined(TARGET_MIPS64)
+    tcg_gen_shri_i64(t0, t0, 30);
+#else
+    tcg_gen_shri_i64(t0, t0, 32);
+#endif
+    gen_move_low32(arg, t0);
+    tcg_temp_free_i64(t0);
+}
+
+static inline void gen_mfhc0_load64(TCGv arg, target_ulong off, int shift)
+{
+    TCGv_i64 t0 = tcg_temp_new_i64();
+
+    tcg_gen_ld_i64(t0, cpu_env, off);
+    tcg_gen_shri_i64(t0, t0, 32 + shift);
+    gen_move_low32(arg, t0);
+    tcg_temp_free_i64(t0);
+}
+
 static inline void gen_mfc0_load32 (TCGv arg, target_ulong off)
 {
     TCGv_i32 t0 = tcg_temp_new_i32();
@@ -4845,6 +4940,140 @@
     tcg_gen_st_tl(arg, cpu_env, off);
 }
 
+static void gen_mfhc0(DisasContext *ctx, TCGv arg, int reg, int sel)
+{
+    const char *rn = "invalid";
+
+    if (!(ctx->hflags & MIPS_HFLAG_ELPA)) {
+        goto mfhc0_read_zero;
+    }
+
+    switch (reg) {
+    case 2:
+        switch (sel) {
+        case 0:
+            gen_mfhc0_entrylo(arg, offsetof(CPUMIPSState, CP0_EntryLo0));
+            rn = "EntryLo0";
+            break;
+        default:
+            goto mfhc0_read_zero;
+        }
+        break;
+    case 3:
+        switch (sel) {
+        case 0:
+            gen_mfhc0_entrylo(arg, offsetof(CPUMIPSState, CP0_EntryLo1));
+            rn = "EntryLo1";
+            break;
+        default:
+            goto mfhc0_read_zero;
+        }
+        break;
+    case 17:
+        switch (sel) {
+        case 0:
+            gen_mfhc0_load64(arg, offsetof(CPUMIPSState, lladdr),
+                             ctx->CP0_LLAddr_shift);
+            rn = "LLAddr";
+            break;
+        default:
+            goto mfhc0_read_zero;
+        }
+        break;
+    case 28:
+        switch (sel) {
+        case 0:
+        case 2:
+        case 4:
+        case 6:
+            gen_mfhc0_load64(arg, offsetof(CPUMIPSState, CP0_TagLo), 0);
+            rn = "TagLo";
+            break;
+        default:
+            goto mfhc0_read_zero;
+        }
+        break;
+    default:
+        goto mfhc0_read_zero;
+    }
+
+    (void)rn; /* avoid a compiler warning */
+    LOG_DISAS("mfhc0 %s (reg %d sel %d)\n", rn, reg, sel);
+    return;
+
+mfhc0_read_zero:
+    LOG_DISAS("mfhc0 %s (reg %d sel %d)\n", rn, reg, sel);
+    tcg_gen_movi_tl(arg, 0);
+}
+
+static void gen_mthc0(DisasContext *ctx, TCGv arg, int reg, int sel)
+{
+    const char *rn = "invalid";
+    uint64_t mask = ctx->PAMask >> 36;
+
+    if (!(ctx->hflags & MIPS_HFLAG_ELPA)) {
+        goto mthc0_nop;
+    }
+
+    switch (reg) {
+    case 2:
+        switch (sel) {
+        case 0:
+            tcg_gen_andi_tl(arg, arg, mask);
+            gen_mthc0_entrylo(arg, offsetof(CPUMIPSState, CP0_EntryLo0));
+            rn = "EntryLo0";
+            break;
+        default:
+            goto mthc0_nop;
+        }
+        break;
+    case 3:
+        switch (sel) {
+        case 0:
+            tcg_gen_andi_tl(arg, arg, mask);
+            gen_mthc0_entrylo(arg, offsetof(CPUMIPSState, CP0_EntryLo1));
+            rn = "EntryLo1";
+            break;
+        default:
+            goto mthc0_nop;
+        }
+        break;
+    case 17:
+        switch (sel) {
+        case 0:
+            /* LLAddr is read-only (the only exception is bit 0 if LLB is
+               supported); the CP0_LLAddr_rw_bitmask does not seem to be
+               relevant for modern MIPS cores supporting MTHC0, therefore
+               treating MTHC0 to LLAddr as NOP. */
+            rn = "LLAddr";
+            break;
+        default:
+            goto mthc0_nop;
+        }
+        break;
+    case 28:
+        switch (sel) {
+        case 0:
+        case 2:
+        case 4:
+        case 6:
+            tcg_gen_andi_tl(arg, arg, mask);
+            gen_mthc0_store64(arg, offsetof(CPUMIPSState, CP0_TagLo));
+            rn = "TagLo";
+            break;
+        default:
+            goto mthc0_nop;
+        }
+        break;
+    default:
+        goto mthc0_nop;
+    }
+
+    (void)rn; /* avoid a compiler warning */
+mthc0_nop:
+    LOG_DISAS("mthc0 %s (reg %d sel %d)\n", rn, reg, sel);
+}
+
 static inline void gen_mfc0_unimplemented(DisasContext *ctx, TCGv arg)
 {
     if (ctx->insn_flags & ISA_MIPS32R6) {
@@ -4943,17 +5172,20 @@
     case 2:
         switch (sel) {
         case 0:
-            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUMIPSState, CP0_EntryLo0));
+            {
+                TCGv_i64 tmp = tcg_temp_new_i64();
+                tcg_gen_ld_i64(tmp, cpu_env,
+                               offsetof(CPUMIPSState, CP0_EntryLo0));
 #if defined(TARGET_MIPS64)
-            if (ctx->rxi) {
-                TCGv tmp = tcg_temp_new();
-                tcg_gen_andi_tl(tmp, arg, (3ull << CP0EnLo_XI));
-                tcg_gen_shri_tl(tmp, tmp, 32);
-                tcg_gen_or_tl(arg, arg, tmp);
-                tcg_temp_free(tmp);
-            }
+                if (ctx->rxi) {
+                    /* Move RI/XI fields to bits 31:30 */
+                    tcg_gen_shri_tl(arg, tmp, CP0EnLo_XI);
+                    tcg_gen_deposit_tl(tmp, tmp, arg, 30, 2);
+                }
 #endif
-            tcg_gen_ext32s_tl(arg, arg);
+                gen_move_low32(arg, tmp);
+                tcg_temp_free_i64(tmp);
+            }
             rn = "EntryLo0";
             break;
         case 1:
@@ -4998,17 +5230,20 @@
     case 3:
         switch (sel) {
         case 0:
-            tcg_gen_ld_tl(arg, cpu_env, offsetof(CPUMIPSState, CP0_EntryLo1));
+            {
+                TCGv_i64 tmp = tcg_temp_new_i64();
+                tcg_gen_ld_i64(tmp, cpu_env,
+                               offsetof(CPUMIPSState, CP0_EntryLo1));
 #if defined(TARGET_MIPS64)
-            if (ctx->rxi) {
-                TCGv tmp = tcg_temp_new();
-                tcg_gen_andi_tl(tmp, arg, (3ull << CP0EnLo_XI));
-                tcg_gen_shri_tl(tmp, tmp, 32);
-                tcg_gen_or_tl(arg, arg, tmp);
-                tcg_temp_free(tmp);
-            }
+                if (ctx->rxi) {
+                    /* Move RI/XI fields to bits 31:30 */
+                    tcg_gen_shri_tl(arg, tmp, CP0EnLo_XI);
+                    tcg_gen_deposit_tl(tmp, tmp, arg, 30, 2);
+                }
 #endif
-            tcg_gen_ext32s_tl(arg, arg);
+                gen_move_low32(arg, tmp);
+                tcg_temp_free_i64(tmp);
+            }
             rn = "EntryLo1";
             break;
         default:
@@ -5418,7 +5653,12 @@
         case 2:
         case 4:
         case 6:
-            gen_mfc0_load32(arg, offsetof(CPUMIPSState, CP0_TagLo));
+            {
+                TCGv_i64 tmp = tcg_temp_new_i64();
+                tcg_gen_ld_i64(tmp, cpu_env, offsetof(CPUMIPSState, CP0_TagLo));
+                gen_move_low32(arg, tmp);
+                tcg_temp_free_i64(tmp);
+            }
             rn = "TagLo";
             break;
         case 1:
@@ -5661,6 +5901,7 @@
             check_insn(ctx, ISA_MIPS32R2);
             gen_helper_mtc0_pagegrain(cpu_env, arg);
             rn = "PageGrain";
+            ctx->bstate = BS_STOP;
             break;
         default:
             goto cp0_unimplemented;
@@ -7557,7 +7798,7 @@
         if (h == 0) {
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, rt);
+            gen_load_fpr32(ctx, fp0, rt);
             tcg_gen_ext_i32_tl(t0, fp0);
             tcg_temp_free_i32(fp0);
         } else {
@@ -7756,7 +7997,7 @@
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
             tcg_gen_trunc_tl_i32(fp0, t0);
-            gen_store_fpr32(fp0, rd);
+            gen_store_fpr32(ctx, fp0, rd);
             tcg_temp_free_i32(fp0);
         } else {
             TCGv_i32 fp0 = tcg_temp_new_i32();
@@ -7841,6 +8082,25 @@
         opn = "dmtc0";
         break;
 #endif
+    case OPC_MFHC0:
+        check_mvh(ctx);
+        if (rt == 0) {
+            /* Treat as NOP. */
+            return;
+        }
+        gen_mfhc0(ctx, cpu_gpr[rt], rd, ctx->opcode & 0x7);
+        opn = "mfhc0";
+        break;
+    case OPC_MTHC0:
+        check_mvh(ctx);
+        {
+            TCGv t0 = tcg_temp_new();
+            gen_load_gpr(t0, rt);
+            gen_mthc0(ctx, t0, rd, ctx->opcode & 0x7);
+            tcg_temp_free(t0);
+        }
+        opn = "mthc0";
+        break;
     case OPC_MFTR:
         check_insn(ctx, ASE_MT);
         if (rd == 0) {
@@ -7899,16 +8159,26 @@
             goto die;
         gen_helper_tlbr(cpu_env);
         break;
-    case OPC_ERET:
-        opn = "eret";
-        check_insn(ctx, ISA_MIPS2);
+    case OPC_ERET: /* OPC_ERETNC */
         if ((ctx->insn_flags & ISA_MIPS32R6) &&
             (ctx->hflags & MIPS_HFLAG_BMASK)) {
             MIPS_DEBUG("CTI in delay / forbidden slot");
             goto die;
+        } else {
+            int bit_shift = (ctx->hflags & MIPS_HFLAG_M16) ? 16 : 6;
+            if (ctx->opcode & (1 << bit_shift)) {
+                /* OPC_ERETNC */
+                opn = "eretnc";
+                check_insn(ctx, ISA_MIPS32R5);
+                gen_helper_eretnc(cpu_env);
+            } else {
+                /* OPC_ERET */
+                opn = "eret";
+                check_insn(ctx, ISA_MIPS2);
+                gen_helper_eret(cpu_env);
+            }
+            ctx->bstate = BS_EXCP;
         }
-        gen_helper_eret(cpu_env);
-        ctx->bstate = BS_EXCP;
         break;
     case OPC_DERET:
         opn = "deret";
@@ -8346,7 +8616,7 @@
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             tcg_gen_ext_i32_tl(t0, fp0);
             tcg_temp_free_i32(fp0);
         }
@@ -8359,7 +8629,7 @@
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
             tcg_gen_trunc_tl_i32(fp0, t0);
-            gen_store_fpr32(fp0, fs);
+            gen_store_fpr32(ctx, fp0, fs);
             tcg_temp_free_i32(fp0);
         }
         opn = "mtc1";
@@ -8457,7 +8727,8 @@
     gen_set_label(l1);
 }
 
-static inline void gen_movcf_s (int fs, int fd, int cc, int tf)
+static inline void gen_movcf_s(DisasContext *ctx, int fs, int fd, int cc,
+                               int tf)
 {
     int cond;
     TCGv_i32 t0 = tcg_temp_new_i32();
@@ -8470,8 +8741,8 @@
 
     tcg_gen_andi_i32(t0, fpu_fcr31, 1 << get_fp_bit(cc));
     tcg_gen_brcondi_i32(cond, t0, 0, l1);
-    gen_load_fpr32(t0, fs);
-    gen_store_fpr32(t0, fd);
+    gen_load_fpr32(ctx, t0, fs);
+    gen_store_fpr32(ctx, t0, fd);
     gen_set_label(l1);
     tcg_temp_free_i32(t0);
 }
@@ -8513,8 +8784,8 @@
 
     tcg_gen_andi_i32(t0, fpu_fcr31, 1 << get_fp_bit(cc));
     tcg_gen_brcondi_i32(cond, t0, 0, l1);
-    gen_load_fpr32(t0, fs);
-    gen_store_fpr32(t0, fd);
+    gen_load_fpr32(ctx, t0, fs);
+    gen_store_fpr32(ctx, t0, fd);
     gen_set_label(l1);
 
     tcg_gen_andi_i32(t0, fpu_fcr31, 1 << get_fp_bit(cc+1));
@@ -8532,9 +8803,9 @@
     TCGv_i32 fp0 = tcg_temp_new_i32();
     TCGv_i32 fp1 = tcg_temp_new_i32();
     TCGv_i32 fp2 = tcg_temp_new_i32();
-    gen_load_fpr32(fp0, fd);
-    gen_load_fpr32(fp1, ft);
-    gen_load_fpr32(fp2, fs);
+    gen_load_fpr32(ctx, fp0, fd);
+    gen_load_fpr32(ctx, fp1, ft);
+    gen_load_fpr32(ctx, fp2, fs);
 
     switch (op1) {
     case OPC_SEL_S:
@@ -8555,7 +8826,7 @@
         break;
     }
 
-    gen_store_fpr32(fp0, fd);
+    gen_store_fpr32(ctx, fp0, fd);
     tcg_temp_free_i32(fp2);
     tcg_temp_free_i32(fp1);
     tcg_temp_free_i32(fp0);
@@ -8648,11 +8919,11 @@
             TCGv_i32 fp0 = tcg_temp_new_i32();
             TCGv_i32 fp1 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
-            gen_load_fpr32(fp1, ft);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_load_fpr32(ctx, fp1, ft);
             gen_helper_float_add_s(fp0, cpu_env, fp0, fp1);
             tcg_temp_free_i32(fp1);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "add.s";
@@ -8663,11 +8934,11 @@
             TCGv_i32 fp0 = tcg_temp_new_i32();
             TCGv_i32 fp1 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
-            gen_load_fpr32(fp1, ft);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_load_fpr32(ctx, fp1, ft);
             gen_helper_float_sub_s(fp0, cpu_env, fp0, fp1);
             tcg_temp_free_i32(fp1);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "sub.s";
@@ -8678,11 +8949,11 @@
             TCGv_i32 fp0 = tcg_temp_new_i32();
             TCGv_i32 fp1 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
-            gen_load_fpr32(fp1, ft);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_load_fpr32(ctx, fp1, ft);
             gen_helper_float_mul_s(fp0, cpu_env, fp0, fp1);
             tcg_temp_free_i32(fp1);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "mul.s";
@@ -8693,11 +8964,11 @@
             TCGv_i32 fp0 = tcg_temp_new_i32();
             TCGv_i32 fp1 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
-            gen_load_fpr32(fp1, ft);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_load_fpr32(ctx, fp1, ft);
             gen_helper_float_div_s(fp0, cpu_env, fp0, fp1);
             tcg_temp_free_i32(fp1);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "div.s";
@@ -8707,9 +8978,9 @@
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             gen_helper_float_sqrt_s(fp0, cpu_env, fp0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "sqrt.s";
@@ -8718,9 +8989,9 @@
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             gen_helper_float_abs_s(fp0, fp0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "abs.s";
@@ -8729,8 +9000,8 @@
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
-            gen_store_fpr32(fp0, fd);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "mov.s";
@@ -8739,9 +9010,9 @@
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             gen_helper_float_chs_s(fp0, fp0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "neg.s";
@@ -8752,7 +9023,7 @@
             TCGv_i32 fp32 = tcg_temp_new_i32();
             TCGv_i64 fp64 = tcg_temp_new_i64();
 
-            gen_load_fpr32(fp32, fs);
+            gen_load_fpr32(ctx, fp32, fs);
             gen_helper_float_roundl_s(fp64, cpu_env, fp32);
             tcg_temp_free_i32(fp32);
             gen_store_fpr64(ctx, fp64, fd);
@@ -8766,7 +9037,7 @@
             TCGv_i32 fp32 = tcg_temp_new_i32();
             TCGv_i64 fp64 = tcg_temp_new_i64();
 
-            gen_load_fpr32(fp32, fs);
+            gen_load_fpr32(ctx, fp32, fs);
             gen_helper_float_truncl_s(fp64, cpu_env, fp32);
             tcg_temp_free_i32(fp32);
             gen_store_fpr64(ctx, fp64, fd);
@@ -8780,7 +9051,7 @@
             TCGv_i32 fp32 = tcg_temp_new_i32();
             TCGv_i64 fp64 = tcg_temp_new_i64();
 
-            gen_load_fpr32(fp32, fs);
+            gen_load_fpr32(ctx, fp32, fs);
             gen_helper_float_ceill_s(fp64, cpu_env, fp32);
             tcg_temp_free_i32(fp32);
             gen_store_fpr64(ctx, fp64, fd);
@@ -8794,7 +9065,7 @@
             TCGv_i32 fp32 = tcg_temp_new_i32();
             TCGv_i64 fp64 = tcg_temp_new_i64();
 
-            gen_load_fpr32(fp32, fs);
+            gen_load_fpr32(ctx, fp32, fs);
             gen_helper_float_floorl_s(fp64, cpu_env, fp32);
             tcg_temp_free_i32(fp32);
             gen_store_fpr64(ctx, fp64, fd);
@@ -8806,9 +9077,9 @@
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             gen_helper_float_roundw_s(fp0, cpu_env, fp0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "round.w.s";
@@ -8817,9 +9088,9 @@
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             gen_helper_float_truncw_s(fp0, cpu_env, fp0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "trunc.w.s";
@@ -8828,9 +9099,9 @@
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             gen_helper_float_ceilw_s(fp0, cpu_env, fp0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "ceil.w.s";
@@ -8839,9 +9110,9 @@
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             gen_helper_float_floorw_s(fp0, cpu_env, fp0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "floor.w.s";
@@ -8863,7 +9134,7 @@
         break;
     case OPC_MOVCF_S:
         check_insn_opc_removed(ctx, ISA_MIPS32R6);
-        gen_movcf_s(fs, fd, (ft >> 2) & 0x7, ft & 0x1);
+        gen_movcf_s(ctx, fs, fd, (ft >> 2) & 0x7, ft & 0x1);
         opn = "movcf.s";
         break;
     case OPC_MOVZ_S:
@@ -8876,8 +9147,8 @@
                 tcg_gen_brcondi_tl(TCG_COND_NE, cpu_gpr[ft], 0, l1);
             }
             fp0 = tcg_temp_new_i32();
-            gen_load_fpr32(fp0, fs);
-            gen_store_fpr32(fp0, fd);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
             gen_set_label(l1);
         }
@@ -8892,8 +9163,8 @@
             if (ft != 0) {
                 tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_gpr[ft], 0, l1);
                 fp0 = tcg_temp_new_i32();
-                gen_load_fpr32(fp0, fs);
-                gen_store_fpr32(fp0, fd);
+                gen_load_fpr32(ctx, fp0, fs);
+                gen_store_fpr32(ctx, fp0, fd);
                 tcg_temp_free_i32(fp0);
                 gen_set_label(l1);
             }
@@ -8905,9 +9176,9 @@
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             gen_helper_float_recip_s(fp0, cpu_env, fp0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "recip.s";
@@ -8917,9 +9188,9 @@
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             gen_helper_float_rsqrt_s(fp0, cpu_env, fp0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "rsqrt.s";
@@ -8930,11 +9201,11 @@
             TCGv_i32 fp0 = tcg_temp_new_i32();
             TCGv_i32 fp1 = tcg_temp_new_i32();
             TCGv_i32 fp2 = tcg_temp_new_i32();
-            gen_load_fpr32(fp0, fs);
-            gen_load_fpr32(fp1, ft);
-            gen_load_fpr32(fp2, fd);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_load_fpr32(ctx, fp1, ft);
+            gen_load_fpr32(ctx, fp2, fd);
             gen_helper_float_maddf_s(fp2, cpu_env, fp0, fp1, fp2);
-            gen_store_fpr32(fp2, fd);
+            gen_store_fpr32(ctx, fp2, fd);
             tcg_temp_free_i32(fp2);
             tcg_temp_free_i32(fp1);
             tcg_temp_free_i32(fp0);
@@ -8947,11 +9218,11 @@
             TCGv_i32 fp0 = tcg_temp_new_i32();
             TCGv_i32 fp1 = tcg_temp_new_i32();
             TCGv_i32 fp2 = tcg_temp_new_i32();
-            gen_load_fpr32(fp0, fs);
-            gen_load_fpr32(fp1, ft);
-            gen_load_fpr32(fp2, fd);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_load_fpr32(ctx, fp1, ft);
+            gen_load_fpr32(ctx, fp2, fd);
             gen_helper_float_msubf_s(fp2, cpu_env, fp0, fp1, fp2);
-            gen_store_fpr32(fp2, fd);
+            gen_store_fpr32(ctx, fp2, fd);
             tcg_temp_free_i32(fp2);
             tcg_temp_free_i32(fp1);
             tcg_temp_free_i32(fp0);
@@ -8962,9 +9233,9 @@
         check_insn(ctx, ISA_MIPS32R6);
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             gen_helper_float_rint_s(fp0, cpu_env, fp0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
             opn = "rint.s";
         }
@@ -8973,9 +9244,9 @@
         check_insn(ctx, ISA_MIPS32R6);
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             gen_helper_float_class_s(fp0, fp0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
             opn = "class.s";
         }
@@ -8986,10 +9257,10 @@
             TCGv_i32 fp0 = tcg_temp_new_i32();
             TCGv_i32 fp1 = tcg_temp_new_i32();
             TCGv_i32 fp2 = tcg_temp_new_i32();
-            gen_load_fpr32(fp0, fs);
-            gen_load_fpr32(fp1, ft);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_load_fpr32(ctx, fp1, ft);
             gen_helper_float_min_s(fp2, cpu_env, fp0, fp1);
-            gen_store_fpr32(fp2, fd);
+            gen_store_fpr32(ctx, fp2, fd);
             tcg_temp_free_i32(fp2);
             tcg_temp_free_i32(fp1);
             tcg_temp_free_i32(fp0);
@@ -9001,11 +9272,11 @@
                 TCGv_i32 fp0 = tcg_temp_new_i32();
                 TCGv_i32 fp1 = tcg_temp_new_i32();
 
-                gen_load_fpr32(fp0, fs);
-                gen_load_fpr32(fp1, ft);
+                gen_load_fpr32(ctx, fp0, fs);
+                gen_load_fpr32(ctx, fp1, ft);
                 gen_helper_float_recip2_s(fp0, cpu_env, fp0, fp1);
                 tcg_temp_free_i32(fp1);
-                gen_store_fpr32(fp0, fd);
+                gen_store_fpr32(ctx, fp0, fd);
                 tcg_temp_free_i32(fp0);
             }
             opn = "recip2.s";
@@ -9017,10 +9288,10 @@
             TCGv_i32 fp0 = tcg_temp_new_i32();
             TCGv_i32 fp1 = tcg_temp_new_i32();
             TCGv_i32 fp2 = tcg_temp_new_i32();
-            gen_load_fpr32(fp0, fs);
-            gen_load_fpr32(fp1, ft);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_load_fpr32(ctx, fp1, ft);
             gen_helper_float_mina_s(fp2, cpu_env, fp0, fp1);
-            gen_store_fpr32(fp2, fd);
+            gen_store_fpr32(ctx, fp2, fd);
             tcg_temp_free_i32(fp2);
             tcg_temp_free_i32(fp1);
             tcg_temp_free_i32(fp0);
@@ -9031,9 +9302,9 @@
             {
                 TCGv_i32 fp0 = tcg_temp_new_i32();
 
-                gen_load_fpr32(fp0, fs);
+                gen_load_fpr32(ctx, fp0, fs);
                 gen_helper_float_recip1_s(fp0, cpu_env, fp0);
-                gen_store_fpr32(fp0, fd);
+                gen_store_fpr32(ctx, fp0, fd);
                 tcg_temp_free_i32(fp0);
             }
             opn = "recip1.s";
@@ -9044,10 +9315,10 @@
             /* OPC_MAX_S */
             TCGv_i32 fp0 = tcg_temp_new_i32();
             TCGv_i32 fp1 = tcg_temp_new_i32();
-            gen_load_fpr32(fp0, fs);
-            gen_load_fpr32(fp1, ft);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_load_fpr32(ctx, fp1, ft);
             gen_helper_float_max_s(fp1, cpu_env, fp0, fp1);
-            gen_store_fpr32(fp1, fd);
+            gen_store_fpr32(ctx, fp1, fd);
             tcg_temp_free_i32(fp1);
             tcg_temp_free_i32(fp0);
             opn = "max.s";
@@ -9057,9 +9328,9 @@
             {
                 TCGv_i32 fp0 = tcg_temp_new_i32();
 
-                gen_load_fpr32(fp0, fs);
+                gen_load_fpr32(ctx, fp0, fs);
                 gen_helper_float_rsqrt1_s(fp0, cpu_env, fp0);
-                gen_store_fpr32(fp0, fd);
+                gen_store_fpr32(ctx, fp0, fd);
                 tcg_temp_free_i32(fp0);
             }
             opn = "rsqrt1.s";
@@ -9070,10 +9341,10 @@
             /* OPC_MAXA_S */
             TCGv_i32 fp0 = tcg_temp_new_i32();
             TCGv_i32 fp1 = tcg_temp_new_i32();
-            gen_load_fpr32(fp0, fs);
-            gen_load_fpr32(fp1, ft);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_load_fpr32(ctx, fp1, ft);
             gen_helper_float_maxa_s(fp1, cpu_env, fp0, fp1);
-            gen_store_fpr32(fp1, fd);
+            gen_store_fpr32(ctx, fp1, fd);
             tcg_temp_free_i32(fp1);
             tcg_temp_free_i32(fp0);
             opn = "maxa.s";
@@ -9084,11 +9355,11 @@
                 TCGv_i32 fp0 = tcg_temp_new_i32();
                 TCGv_i32 fp1 = tcg_temp_new_i32();
 
-                gen_load_fpr32(fp0, fs);
-                gen_load_fpr32(fp1, ft);
+                gen_load_fpr32(ctx, fp0, fs);
+                gen_load_fpr32(ctx, fp1, ft);
                 gen_helper_float_rsqrt2_s(fp0, cpu_env, fp0, fp1);
                 tcg_temp_free_i32(fp1);
-                gen_store_fpr32(fp0, fd);
+                gen_store_fpr32(ctx, fp0, fd);
                 tcg_temp_free_i32(fp0);
             }
             opn = "rsqrt2.s";
@@ -9100,7 +9371,7 @@
             TCGv_i32 fp32 = tcg_temp_new_i32();
             TCGv_i64 fp64 = tcg_temp_new_i64();
 
-            gen_load_fpr32(fp32, fs);
+            gen_load_fpr32(ctx, fp32, fs);
             gen_helper_float_cvtd_s(fp64, cpu_env, fp32);
             tcg_temp_free_i32(fp32);
             gen_store_fpr64(ctx, fp64, fd);
@@ -9112,9 +9383,9 @@
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             gen_helper_float_cvtw_s(fp0, cpu_env, fp0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "cvt.w.s";
@@ -9125,7 +9396,7 @@
             TCGv_i32 fp32 = tcg_temp_new_i32();
             TCGv_i64 fp64 = tcg_temp_new_i64();
 
-            gen_load_fpr32(fp32, fs);
+            gen_load_fpr32(ctx, fp32, fs);
             gen_helper_float_cvtl_s(fp64, cpu_env, fp32);
             tcg_temp_free_i32(fp32);
             gen_store_fpr64(ctx, fp64, fd);
@@ -9141,8 +9412,8 @@
             TCGv_i32 fp32_0 = tcg_temp_new_i32();
             TCGv_i32 fp32_1 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp32_0, fs);
-            gen_load_fpr32(fp32_1, ft);
+            gen_load_fpr32(ctx, fp32_0, fs);
+            gen_load_fpr32(ctx, fp32_1, ft);
             tcg_gen_concat_i32_i64(fp64, fp32_1, fp32_0);
             tcg_temp_free_i32(fp32_1);
             tcg_temp_free_i32(fp32_0);
@@ -9344,7 +9615,7 @@
             gen_load_fpr64(ctx, fp64, fs);
             gen_helper_float_roundw_d(fp32, cpu_env, fp64);
             tcg_temp_free_i64(fp64);
-            gen_store_fpr32(fp32, fd);
+            gen_store_fpr32(ctx, fp32, fd);
             tcg_temp_free_i32(fp32);
         }
         opn = "round.w.d";
@@ -9358,7 +9629,7 @@
             gen_load_fpr64(ctx, fp64, fs);
             gen_helper_float_truncw_d(fp32, cpu_env, fp64);
             tcg_temp_free_i64(fp64);
-            gen_store_fpr32(fp32, fd);
+            gen_store_fpr32(ctx, fp32, fd);
             tcg_temp_free_i32(fp32);
         }
         opn = "trunc.w.d";
@@ -9372,7 +9643,7 @@
             gen_load_fpr64(ctx, fp64, fs);
             gen_helper_float_ceilw_d(fp32, cpu_env, fp64);
             tcg_temp_free_i64(fp64);
-            gen_store_fpr32(fp32, fd);
+            gen_store_fpr32(ctx, fp32, fd);
             tcg_temp_free_i32(fp32);
         }
         opn = "ceil.w.d";
@@ -9386,7 +9657,7 @@
             gen_load_fpr64(ctx, fp64, fs);
             gen_helper_float_floorw_d(fp32, cpu_env, fp64);
             tcg_temp_free_i64(fp64);
-            gen_store_fpr32(fp32, fd);
+            gen_store_fpr32(ctx, fp32, fd);
             tcg_temp_free_i32(fp32);
         }
         opn = "floor.w.d";
@@ -9669,7 +9940,7 @@
             gen_load_fpr64(ctx, fp64, fs);
             gen_helper_float_cvts_d(fp32, cpu_env, fp64);
             tcg_temp_free_i64(fp64);
-            gen_store_fpr32(fp32, fd);
+            gen_store_fpr32(ctx, fp32, fd);
             tcg_temp_free_i32(fp32);
         }
         opn = "cvt.s.d";
@@ -9683,7 +9954,7 @@
             gen_load_fpr64(ctx, fp64, fs);
             gen_helper_float_cvtw_d(fp32, cpu_env, fp64);
             tcg_temp_free_i64(fp64);
-            gen_store_fpr32(fp32, fd);
+            gen_store_fpr32(ctx, fp32, fd);
             tcg_temp_free_i32(fp32);
         }
         opn = "cvt.w.d";
@@ -9704,9 +9975,9 @@
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             gen_helper_float_cvts_w(fp0, cpu_env, fp0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "cvt.s.w";
@@ -9717,7 +9988,7 @@
             TCGv_i32 fp32 = tcg_temp_new_i32();
             TCGv_i64 fp64 = tcg_temp_new_i64();
 
-            gen_load_fpr32(fp32, fs);
+            gen_load_fpr32(ctx, fp32, fs);
             gen_helper_float_cvtd_w(fp64, cpu_env, fp32);
             tcg_temp_free_i32(fp32);
             gen_store_fpr64(ctx, fp64, fd);
@@ -9734,7 +10005,7 @@
             gen_load_fpr64(ctx, fp64, fs);
             gen_helper_float_cvts_l(fp32, cpu_env, fp64);
             tcg_temp_free_i64(fp64);
-            gen_store_fpr32(fp32, fd);
+            gen_store_fpr32(ctx, fp32, fd);
             tcg_temp_free_i32(fp32);
         }
         opn = "cvt.s.l";
@@ -9973,7 +10244,7 @@
 
             gen_load_fpr32h(ctx, fp0, fs);
             gen_helper_float_cvts_pu(fp0, cpu_env, fp0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "cvt.s.pu";
@@ -9995,9 +10266,9 @@
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             gen_helper_float_cvts_pl(fp0, cpu_env, fp0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "cvt.s.pl";
@@ -10008,10 +10279,10 @@
             TCGv_i32 fp0 = tcg_temp_new_i32();
             TCGv_i32 fp1 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
-            gen_load_fpr32(fp1, ft);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_load_fpr32(ctx, fp1, ft);
             gen_store_fpr32h(ctx, fp0, fd);
-            gen_store_fpr32(fp1, fd);
+            gen_store_fpr32(ctx, fp1, fd);
             tcg_temp_free_i32(fp0);
             tcg_temp_free_i32(fp1);
         }
@@ -10023,9 +10294,9 @@
             TCGv_i32 fp0 = tcg_temp_new_i32();
             TCGv_i32 fp1 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             gen_load_fpr32h(ctx, fp1, ft);
-            gen_store_fpr32(fp1, fd);
+            gen_store_fpr32(ctx, fp1, fd);
             gen_store_fpr32h(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
             tcg_temp_free_i32(fp1);
@@ -10039,8 +10310,8 @@
             TCGv_i32 fp1 = tcg_temp_new_i32();
 
             gen_load_fpr32h(ctx, fp0, fs);
-            gen_load_fpr32(fp1, ft);
-            gen_store_fpr32(fp1, fd);
+            gen_load_fpr32(ctx, fp1, ft);
+            gen_store_fpr32(ctx, fp1, fd);
             gen_store_fpr32h(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
             tcg_temp_free_i32(fp1);
@@ -10055,7 +10326,7 @@
 
             gen_load_fpr32h(ctx, fp0, fs);
             gen_load_fpr32h(ctx, fp1, ft);
-            gen_store_fpr32(fp1, fd);
+            gen_store_fpr32(ctx, fp1, fd);
             gen_store_fpr32h(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
             tcg_temp_free_i32(fp1);
@@ -10130,7 +10401,7 @@
 
             tcg_gen_qemu_ld_tl(t0, t0, ctx->mem_idx, MO_TESL);
             tcg_gen_trunc_tl_i32(fp0, t0);
-            gen_store_fpr32(fp0, fd);
+            gen_store_fpr32(ctx, fp0, fd);
             tcg_temp_free_i32(fp0);
         }
         opn = "lwxc1";
@@ -10162,7 +10433,7 @@
         check_cop1x(ctx);
         {
             TCGv_i32 fp0 = tcg_temp_new_i32();
-            gen_load_fpr32(fp0, fs);
+            gen_load_fpr32(ctx, fp0, fs);
             tcg_gen_qemu_st_i32(fp0, t0, ctx->mem_idx, MO_TEUL);
             tcg_temp_free_i32(fp0);
         }
@@ -10219,23 +10490,23 @@
             tcg_gen_andi_tl(t0, t0, 0x7);
 
             tcg_gen_brcondi_tl(TCG_COND_NE, t0, 0, l1);
-            gen_load_fpr32(fp, fs);
+            gen_load_fpr32(ctx, fp, fs);
             gen_load_fpr32h(ctx, fph, fs);
-            gen_store_fpr32(fp, fd);
+            gen_store_fpr32(ctx, fp, fd);
             gen_store_fpr32h(ctx, fph, fd);
             tcg_gen_br(l2);
             gen_set_label(l1);
             tcg_gen_brcondi_tl(TCG_COND_NE, t0, 4, l2);
             tcg_temp_free(t0);
 #ifdef TARGET_WORDS_BIGENDIAN
-            gen_load_fpr32(fp, fs);
+            gen_load_fpr32(ctx, fp, fs);
             gen_load_fpr32h(ctx, fph, ft);
             gen_store_fpr32h(ctx, fp, fd);
-            gen_store_fpr32(fph, fd);
+            gen_store_fpr32(ctx, fph, fd);
 #else
             gen_load_fpr32h(ctx, fph, fs);
-            gen_load_fpr32(fp, ft);
-            gen_store_fpr32(fph, fd);
+            gen_load_fpr32(ctx, fp, ft);
+            gen_store_fpr32(ctx, fph, fd);
             gen_store_fpr32h(ctx, fp, fd);
 #endif
             gen_set_label(l2);
@@ -10251,13 +10522,13 @@
             TCGv_i32 fp1 = tcg_temp_new_i32();
             TCGv_i32 fp2 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
-            gen_load_fpr32(fp1, ft);
-            gen_load_fpr32(fp2, fr);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_load_fpr32(ctx, fp1, ft);
+            gen_load_fpr32(ctx, fp2, fr);
             gen_helper_float_madd_s(fp2, cpu_env, fp0, fp1, fp2);
             tcg_temp_free_i32(fp0);
             tcg_temp_free_i32(fp1);
-            gen_store_fpr32(fp2, fd);
+            gen_store_fpr32(ctx, fp2, fd);
             tcg_temp_free_i32(fp2);
         }
         opn = "madd.s";
@@ -10306,13 +10577,13 @@
             TCGv_i32 fp1 = tcg_temp_new_i32();
             TCGv_i32 fp2 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
-            gen_load_fpr32(fp1, ft);
-            gen_load_fpr32(fp2, fr);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_load_fpr32(ctx, fp1, ft);
+            gen_load_fpr32(ctx, fp2, fr);
             gen_helper_float_msub_s(fp2, cpu_env, fp0, fp1, fp2);
             tcg_temp_free_i32(fp0);
             tcg_temp_free_i32(fp1);
-            gen_store_fpr32(fp2, fd);
+            gen_store_fpr32(ctx, fp2, fd);
             tcg_temp_free_i32(fp2);
         }
         opn = "msub.s";
@@ -10361,13 +10632,13 @@
             TCGv_i32 fp1 = tcg_temp_new_i32();
             TCGv_i32 fp2 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
-            gen_load_fpr32(fp1, ft);
-            gen_load_fpr32(fp2, fr);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_load_fpr32(ctx, fp1, ft);
+            gen_load_fpr32(ctx, fp2, fr);
             gen_helper_float_nmadd_s(fp2, cpu_env, fp0, fp1, fp2);
             tcg_temp_free_i32(fp0);
             tcg_temp_free_i32(fp1);
-            gen_store_fpr32(fp2, fd);
+            gen_store_fpr32(ctx, fp2, fd);
             tcg_temp_free_i32(fp2);
         }
         opn = "nmadd.s";
@@ -10416,13 +10687,13 @@
             TCGv_i32 fp1 = tcg_temp_new_i32();
             TCGv_i32 fp2 = tcg_temp_new_i32();
 
-            gen_load_fpr32(fp0, fs);
-            gen_load_fpr32(fp1, ft);
-            gen_load_fpr32(fp2, fr);
+            gen_load_fpr32(ctx, fp0, fs);
+            gen_load_fpr32(ctx, fp1, ft);
+            gen_load_fpr32(ctx, fp2, fr);
             gen_helper_float_nmsub_s(fp2, cpu_env, fp0, fp1, fp2);
             tcg_temp_free_i32(fp0);
             tcg_temp_free_i32(fp1);
-            gen_store_fpr32(fp2, fd);
+            gen_store_fpr32(ctx, fp2, fd);
             tcg_temp_free_i32(fp2);
         }
         opn = "nmsub.s";
@@ -13502,7 +13773,7 @@
                 case MOVF_FMT:
                     switch (fmt) {
                     case FMT_SDPS_S:
-                        gen_movcf_s(rs, rt, cc, 0);
+                        gen_movcf_s(ctx, rs, rt, cc, 0);
                         break;
                     case FMT_SDPS_D:
                         gen_movcf_d(ctx, rs, rt, cc, 0);
@@ -13517,7 +13788,7 @@
                 case MOVT_FMT:
                     switch (fmt) {
                     case FMT_SDPS_S:
-                        gen_movcf_s(rs, rt, cc, 1);
+                        gen_movcf_s(ctx, rs, rt, cc, 1);
                         break;
                     case FMT_SDPS_D:
                         gen_movcf_d(ctx, rs, rt, cc, 1);
@@ -18404,32 +18675,39 @@
             uint8_t wd = (ctx->opcode >> 6) & 0x1f;
             uint8_t df = (ctx->opcode >> 0) & 0x3;
 
-            TCGv_i32 tdf = tcg_const_i32(df);
             TCGv_i32 twd = tcg_const_i32(wd);
-            TCGv_i32 trs = tcg_const_i32(rs);
-            TCGv_i32 ts10 = tcg_const_i32(s10);
+            TCGv taddr = tcg_temp_new();
+            gen_base_offset_addr(ctx, taddr, rs, s10 << df);
 
             switch (MASK_MSA_MINOR(opcode)) {
             case OPC_LD_B:
+                gen_helper_msa_ld_b(cpu_env, twd, taddr);
+                break;
             case OPC_LD_H:
+                gen_helper_msa_ld_h(cpu_env, twd, taddr);
+                break;
             case OPC_LD_W:
+                gen_helper_msa_ld_w(cpu_env, twd, taddr);
+                break;
             case OPC_LD_D:
-                save_cpu_state(ctx, 1);
-                gen_helper_msa_ld_df(cpu_env, tdf, twd, trs, ts10);
+                gen_helper_msa_ld_d(cpu_env, twd, taddr);
                 break;
             case OPC_ST_B:
+                gen_helper_msa_st_b(cpu_env, twd, taddr);
+                break;
             case OPC_ST_H:
+                gen_helper_msa_st_h(cpu_env, twd, taddr);
+                break;
             case OPC_ST_W:
+                gen_helper_msa_st_w(cpu_env, twd, taddr);
+                break;
             case OPC_ST_D:
-                save_cpu_state(ctx, 1);
-                gen_helper_msa_st_df(cpu_env, tdf, twd, trs, ts10);
+                gen_helper_msa_st_d(cpu_env, twd, taddr);
                 break;
             }
 
             tcg_temp_free_i32(twd);
-            tcg_temp_free_i32(tdf);
-            tcg_temp_free_i32(trs);
-            tcg_temp_free_i32(ts10);
+            tcg_temp_free(taddr);
         }
         break;
     default:
@@ -18564,6 +18842,8 @@
         case OPC_MTC0:
         case OPC_MFTR:
         case OPC_MTTR:
+        case OPC_MFHC0:
+        case OPC_MTHC0:
 #if defined(TARGET_MIPS64)
         case OPC_DMFC0:
         case OPC_DMTC0:
@@ -19134,6 +19414,9 @@
     ctx.ie = (env->CP0_Config4 >> CP0C4_IE) & 3;
     ctx.bi = (env->CP0_Config3 >> CP0C3_BI) & 1;
     ctx.bp = (env->CP0_Config3 >> CP0C3_BP) & 1;
+    ctx.PAMask = env->PAMask;
+    ctx.mvh = (env->CP0_Config5 >> CP0C5_MVH) & 1;
+    ctx.CP0_LLAddr_shift = env->CP0_LLAddr_shift;
     /* Restore delay slot state from the tb context.  */
     ctx.hflags = (uint32_t)tb->flags; /* FIXME: maybe use 64 bits here? */
     ctx.ulri = (env->CP0_Config3 >> CP0C3_ULRI) & 1;
@@ -19143,6 +19426,8 @@
 #else
         ctx.mem_idx = ctx.hflags & MIPS_HFLAG_KSU;
 #endif
+    ctx.default_tcg_memop_mask = (ctx.insn_flags & ISA_MIPS32R6) ?
+                                 MO_UNALN : MO_ALIGN;
     num_insns = 0;
     max_insns = tb->cflags & CF_COUNT_MASK;
     if (max_insns == 0)
@@ -19385,7 +19670,8 @@
 
     cpu_fprintf(f, "CP0 Status  0x%08x Cause   0x%08x EPC    0x" TARGET_FMT_lx "\n",
                 env->CP0_Status, env->CP0_Cause, env->CP0_EPC);
-    cpu_fprintf(f, "    Config0 0x%08x Config1 0x%08x LLAddr 0x" TARGET_FMT_lx "\n",
+    cpu_fprintf(f, "    Config0 0x%08x Config1 0x%08x LLAddr 0x%016"
+                PRIx64 "\n",
                 env->CP0_Config0, env->CP0_Config1, env->lladdr);
     cpu_fprintf(f, "    Config2 0x%08x Config3 0x%08x\n",
                 env->CP0_Config2, env->CP0_Config3);
@@ -19519,7 +19805,6 @@
     }
 #endif
     env->PABITS = env->cpu_model->PABITS;
-    env->PAMask = (target_ulong)((1ULL << env->cpu_model->PABITS) - 1);
     env->CP0_SRSConf0_rw_bitmask = env->cpu_model->CP0_SRSConf0_rw_bitmask;
     env->CP0_SRSConf0 = env->cpu_model->CP0_SRSConf0;
     env->CP0_SRSConf1_rw_bitmask = env->cpu_model->CP0_SRSConf1_rw_bitmask;
@@ -19640,6 +19925,7 @@
     compute_hflags(env);
     restore_rounding_mode(env);
     restore_flush_mode(env);
+    restore_pamask(env);
     cs->exception_index = EXCP_NONE;
 }
 
diff --git a/target-mips/translate_init.c b/target-mips/translate_init.c
index 85a65e7..30605da 100644
--- a/target-mips/translate_init.c
+++ b/target-mips/translate_init.c
@@ -400,10 +400,12 @@
                        (0 << CP0C1_DS) | (3 << CP0C1_DL) | (1 << CP0C1_DA) |
                        (1 << CP0C1_CA),
         .CP0_Config2 = MIPS_CONFIG2,
-        .CP0_Config3 = MIPS_CONFIG3 | (1U << CP0C3_M) | (1 << CP0C3_MSAP),
+        .CP0_Config3 = MIPS_CONFIG3 | (1U << CP0C3_M) | (1 << CP0C3_MSAP) |
+                       (1 << CP0C3_LPA),
         .CP0_Config4 = MIPS_CONFIG4 | (1U << CP0C4_M),
         .CP0_Config4_rw_bitmask = 0,
-        .CP0_Config5 = MIPS_CONFIG5 | (1 << CP0C5_UFR),
+        .CP0_Config5 = MIPS_CONFIG5 | (1 << CP0C5_UFR) | (1 << CP0C5_LLB) |
+                       (1 << CP0C5_MVH),
         .CP0_Config5_rw_bitmask = (0 << CP0C5_M) | (1 << CP0C5_K) |
                                   (1 << CP0C5_CV) | (0 << CP0C5_EVA) |
                                   (1 << CP0C5_MSAEn) | (1 << CP0C5_UFR) |
@@ -413,11 +415,12 @@
         .SYNCI_Step = 32,
         .CCRes = 2,
         .CP0_Status_rw_bitmask = 0x3778FF1F,
+        .CP0_PageGrain_rw_bitmask = (1 << CP0PG_ELPA),
         .CP1_fcr0 = (1 << FCR0_UFRP) | (1 << FCR0_F64) | (1 << FCR0_L) |
                     (1 << FCR0_W) | (1 << FCR0_D) | (1 << FCR0_S) |
                     (0x93 << FCR0_PRID),
         .SEGBITS = 32,
-        .PABITS = 32,
+        .PABITS = 40,
         .insn_flags = CPU_MIPS32R5 | ASE_MIPS16 | ASE_MSA,
         .mmu_type = MMU_TYPE_R4000,
     },
@@ -553,9 +556,6 @@
                     (1 << FCR0_L) | (1 << FCR0_W) | (1 << FCR0_D) |
                     (1 << FCR0_S) | (0x00 << FCR0_PRID) | (0x0 << FCR0_REV),
         .SEGBITS = 42,
-        /* The architectural limit is 59, but we have hardcoded 36 bit
-           in some places...
-        .PABITS = 59, */ /* the architectural limit */
         .PABITS = 36,
         .insn_flags = CPU_MIPS64R2 | ASE_MIPS3D,
         .mmu_type = MMU_TYPE_R4000,
@@ -607,7 +607,7 @@
     },
     {
         /* A generic CPU supporting MIPS64 Release 6 ISA.
-           FIXME: Support IEEE 754-2008 FP and misaligned memory accesses.
+           FIXME: Support IEEE 754-2008 FP.
                   Eventually this should be replaced by a real CPU model. */
         .name = "MIPS64R6-generic",
         .CP0_PRid = 0x00010000,
@@ -619,10 +619,13 @@
                        (0 << CP0C1_PC) | (1 << CP0C1_WR) | (1 << CP0C1_EP),
         .CP0_Config2 = MIPS_CONFIG2,
         .CP0_Config3 = MIPS_CONFIG3 | (1 << CP0C3_RXI) | (1 << CP0C3_BP) |
-                       (1 << CP0C3_BI) | (1 << CP0C3_ULRI) | (1U << CP0C3_M),
+                       (1 << CP0C3_BI) | (1 << CP0C3_ULRI) | (1 << CP0C3_LPA) |
+                       (1U << CP0C3_M),
         .CP0_Config4 = MIPS_CONFIG4 | (0xfc << CP0C4_KScrExist) |
                        (3 << CP0C4_IE) | (1 << CP0C4_M),
-        .CP0_Config5_rw_bitmask = (1 << CP0C5_SBRI),
+        .CP0_Config5 = MIPS_CONFIG5 | (1 << CP0C5_LLB),
+        .CP0_Config5_rw_bitmask = (1 << CP0C5_SBRI) | (1 << CP0C5_FRE) |
+                                  (1 << CP0C5_UFE),
         .CP0_LLAddr_rw_bitmask = 0,
         .CP0_LLAddr_shift = 0,
         .SYNCI_Step = 32,
@@ -630,15 +633,12 @@
         .CP0_Status_rw_bitmask = 0x30D8FFFF,
         .CP0_PageGrain = (1 << CP0PG_IEC) | (1 << CP0PG_XIE) |
                          (1U << CP0PG_RIE),
-        .CP0_PageGrain_rw_bitmask = 0,
-        .CP1_fcr0 = (1 << FCR0_F64) | (1 << FCR0_L) | (1 << FCR0_W) |
-                    (1 << FCR0_D) | (1 << FCR0_S) | (0x00 << FCR0_PRID) |
-                    (0x0 << FCR0_REV),
+        .CP0_PageGrain_rw_bitmask = (1 << CP0PG_ELPA),
+        .CP1_fcr0 = (1 << FCR0_FREP) | (1 << FCR0_F64) | (1 << FCR0_L) |
+                    (1 << FCR0_W) | (1 << FCR0_D) | (1 << FCR0_S) |
+                    (0x00 << FCR0_PRID) | (0x0 << FCR0_REV),
         .SEGBITS = 42,
-        /* The architectural limit is 59, but we have hardcoded 36 bit
-           in some places...
-        .PABITS = 59, */ /* the architectural limit */
-        .PABITS = 36,
+        .PABITS = 48,
         .insn_flags = CPU_MIPS64R6,
         .mmu_type = MMU_TYPE_R4000,
     },
@@ -701,9 +701,6 @@
                     (1 << FCR0_L) | (1 << FCR0_W) | (1 << FCR0_D) |
                     (1 << FCR0_S) | (0x00 << FCR0_PRID) | (0x0 << FCR0_REV),
         .SEGBITS = 42,
-        /* The architectural limit is 59, but we have hardcoded 36 bit
-           in some places...
-        .PABITS = 59, */ /* the architectural limit */
         .PABITS = 36,
         .insn_flags = CPU_MIPS64R2 | ASE_DSP | ASE_DSPR2,
         .mmu_type = MMU_TYPE_R4000,
diff --git a/target-ppc/machine.c b/target-ppc/machine.c
index d875211..f4ac761 100644
--- a/target-ppc/machine.c
+++ b/target-ppc/machine.c
@@ -213,6 +213,7 @@
     .name = "cpu/fpu",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = fpu_needed,
     .fields = (VMStateField[]) {
         VMSTATE_FLOAT64_ARRAY(env.fpr, PowerPCCPU, 32),
         VMSTATE_UINTTL(env.fpscr, PowerPCCPU),
@@ -231,6 +232,7 @@
     .name = "cpu/altivec",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = altivec_needed,
     .fields = (VMStateField[]) {
         VMSTATE_AVR_ARRAY(env.avr, PowerPCCPU, 32),
         VMSTATE_UINT32(env.vscr, PowerPCCPU),
@@ -249,6 +251,7 @@
     .name = "cpu/vsx",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = vsx_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64_ARRAY(env.vsr, PowerPCCPU, 32),
         VMSTATE_END_OF_LIST()
@@ -269,6 +272,7 @@
     .version_id = 1,
     .minimum_version_id = 1,
     .minimum_version_id_old = 1,
+    .needed = tm_needed,
     .fields      = (VMStateField []) {
         VMSTATE_UINTTL_ARRAY(env.tm_gpr, PowerPCCPU, 32),
         VMSTATE_AVR_ARRAY(env.tm_vsr, PowerPCCPU, 64),
@@ -302,6 +306,7 @@
     .name = "cpu/sr",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = sr_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINTTL_ARRAY(env.sr, PowerPCCPU, 32),
         VMSTATE_END_OF_LIST()
@@ -351,6 +356,7 @@
     .name = "cpu/slb",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = slb_needed,
     .fields = (VMStateField[]) {
         VMSTATE_INT32_EQUAL(env.slb_nr, PowerPCCPU),
         VMSTATE_SLB_ARRAY(env.slb, PowerPCCPU, MAX_SLB_ENTRIES),
@@ -383,6 +389,7 @@
     .name = "cpu/tlb6xx",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = tlb6xx_needed,
     .fields = (VMStateField[]) {
         VMSTATE_INT32_EQUAL(env.nb_tlb, PowerPCCPU),
         VMSTATE_STRUCT_VARRAY_POINTER_INT32(env.tlb.tlb6, PowerPCCPU,
@@ -429,6 +436,7 @@
     .name = "cpu/pbr403",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = pbr403_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINTTL_ARRAY(env.pb, PowerPCCPU, 4),
         VMSTATE_END_OF_LIST()
@@ -439,6 +447,7 @@
     .name = "cpu/tlb6xx",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = tlbemb_needed,
     .fields = (VMStateField[]) {
         VMSTATE_INT32_EQUAL(env.nb_tlb, PowerPCCPU),
         VMSTATE_STRUCT_VARRAY_POINTER_INT32(env.tlb.tlbe, PowerPCCPU,
@@ -448,13 +457,9 @@
         /* 403 protection registers */
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection []) {
-        {
-            .vmsd = &vmstate_pbr403,
-            .needed = pbr403_needed,
-        } , {
-            /* empty */
-        }
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_pbr403,
+        NULL
     }
 };
 
@@ -483,6 +488,7 @@
     .name = "cpu/tlbmas",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = tlbmas_needed,
     .fields = (VMStateField[]) {
         VMSTATE_INT32_EQUAL(env.nb_tlb, PowerPCCPU),
         VMSTATE_STRUCT_VARRAY_POINTER_INT32(env.tlb.tlbm, PowerPCCPU,
@@ -533,38 +539,18 @@
         VMSTATE_UINT32_EQUAL(env.nb_BATs, PowerPCCPU),
         VMSTATE_END_OF_LIST()
     },
-    .subsections = (VMStateSubsection []) {
-        {
-            .vmsd = &vmstate_fpu,
-            .needed = fpu_needed,
-        } , {
-            .vmsd = &vmstate_altivec,
-            .needed = altivec_needed,
-        } , {
-            .vmsd = &vmstate_vsx,
-            .needed = vsx_needed,
-        } , {
-            .vmsd = &vmstate_sr,
-            .needed = sr_needed,
-        } , {
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_fpu,
+        &vmstate_altivec,
+        &vmstate_vsx,
+        &vmstate_sr,
 #ifdef TARGET_PPC64
-            .vmsd = &vmstate_tm,
-            .needed = tm_needed,
-        } , {
-            .vmsd = &vmstate_slb,
-            .needed = slb_needed,
-        } , {
+        &vmstate_tm,
+        &vmstate_slb,
 #endif /* TARGET_PPC64 */
-            .vmsd = &vmstate_tlb6xx,
-            .needed = tlb6xx_needed,
-        } , {
-            .vmsd = &vmstate_tlbemb,
-            .needed = tlbemb_needed,
-        } , {
-            .vmsd = &vmstate_tlbmas,
-            .needed = tlbmas_needed,
-        } , {
-            /* empty */
-        }
+        &vmstate_tlb6xx,
+        &vmstate_tlbemb,
+        &vmstate_tlbmas,
+        NULL
     }
 };
diff --git a/target-s390x/machine.c b/target-s390x/machine.c
index 0044749..b76fb08 100644
--- a/target-s390x/machine.c
+++ b/target-s390x/machine.c
@@ -42,10 +42,17 @@
     }
 }
 
+static inline bool fpu_needed(void *opaque)
+{
+    /* This looks odd, but we might want to NOT transfer fprs in the future */
+    return true;
+}
+
 const VMStateDescription vmstate_fpu = {
     .name = "cpu/fpu",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = fpu_needed,
     .fields = (VMStateField[]) {
         VMSTATE_UINT64(env.vregs[0][0].ll, S390CPU),
         VMSTATE_UINT64(env.vregs[1][0].ll, S390CPU),
@@ -68,16 +75,11 @@
     }
 };
 
-static inline bool fpu_needed(void *opaque)
-{
-    /* This looks odd, but we might want to NOT transfer fprs in the future */
-    return true;
-}
-
 const VMStateDescription vmstate_vregs = {
     .name = "cpu/vregs",
     .version_id = 1,
     .minimum_version_id = 1,
+    .needed = vregs_needed,
     .fields = (VMStateField[]) {
         /* vregs[0][0] -> vregs[15][0] and fregs are overlays */
         VMSTATE_UINT64(env.vregs[16][0].ll, S390CPU),
@@ -159,16 +161,10 @@
         VMSTATE_VBUFFER_UINT32(irqstate, S390CPU, 4, NULL, 0,
                                irqstate_saved_size),
         VMSTATE_END_OF_LIST()
-     },
-    .subsections = (VMStateSubsection[]) {
-        {
-            .vmsd = &vmstate_fpu,
-            .needed = fpu_needed,
-        } , {
-            .vmsd = &vmstate_vregs,
-            .needed = vregs_needed,
-        } , {
-            /* empty */
-        }
+    },
+    .subsections = (const VMStateDescription*[]) {
+        &vmstate_fpu,
+        &vmstate_vregs,
+        NULL
     },
 };
diff --git a/tests/bios-tables-test.c b/tests/bios-tables-test.c
index 7e85dc4..0de1742 100644
--- a/tests/bios-tables-test.c
+++ b/tests/bios-tables-test.c
@@ -599,7 +599,45 @@
     free_test_data(&exp_data);
 }
 
-static void test_smbios_ep_address(test_data *data)
+static bool smbios_ep_table_ok(test_data *data)
+{
+    struct smbios_entry_point *ep_table = &data->smbios_ep_table;
+    uint32_t addr = data->smbios_ep_addr;
+
+    ACPI_READ_ARRAY(ep_table->anchor_string, addr);
+    if (memcmp(ep_table->anchor_string, "_SM_", 4)) {
+        return false;
+    }
+    ACPI_READ_FIELD(ep_table->checksum, addr);
+    ACPI_READ_FIELD(ep_table->length, addr);
+    ACPI_READ_FIELD(ep_table->smbios_major_version, addr);
+    ACPI_READ_FIELD(ep_table->smbios_minor_version, addr);
+    ACPI_READ_FIELD(ep_table->max_structure_size, addr);
+    ACPI_READ_FIELD(ep_table->entry_point_revision, addr);
+    ACPI_READ_ARRAY(ep_table->formatted_area, addr);
+    ACPI_READ_ARRAY(ep_table->intermediate_anchor_string, addr);
+    if (memcmp(ep_table->intermediate_anchor_string, "_DMI_", 5)) {
+        return false;
+    }
+    ACPI_READ_FIELD(ep_table->intermediate_checksum, addr);
+    ACPI_READ_FIELD(ep_table->structure_table_length, addr);
+    if (ep_table->structure_table_length == 0) {
+        return false;
+    }
+    ACPI_READ_FIELD(ep_table->structure_table_address, addr);
+    ACPI_READ_FIELD(ep_table->number_of_structures, addr);
+    if (ep_table->number_of_structures == 0) {
+        return false;
+    }
+    ACPI_READ_FIELD(ep_table->smbios_bcd_revision, addr);
+    if (acpi_checksum((uint8_t *)ep_table, sizeof *ep_table) ||
+        acpi_checksum((uint8_t *)ep_table + 0x10, sizeof *ep_table - 0x10)) {
+        return false;
+    }
+    return true;
+}
+
+static void test_smbios_entry_point(test_data *data)
 {
     uint32_t off;
 
@@ -613,40 +651,15 @@
         }
 
         if (!memcmp(sig, "_SM_", sizeof sig)) {
-            break;
+            /* signature match, but is this a valid entry point? */
+            data->smbios_ep_addr = off;
+            if (smbios_ep_table_ok(data)) {
+                break;
+            }
         }
     }
 
     g_assert_cmphex(off, <, 0x100000);
-    data->smbios_ep_addr = off;
-}
-
-static void test_smbios_ep_table(test_data *data)
-{
-    struct smbios_entry_point *ep_table = &data->smbios_ep_table;
-    uint32_t addr = data->smbios_ep_addr;
-
-    ACPI_READ_ARRAY(ep_table->anchor_string, addr);
-    g_assert(!memcmp(ep_table->anchor_string, "_SM_", 4));
-    ACPI_READ_FIELD(ep_table->checksum, addr);
-    ACPI_READ_FIELD(ep_table->length, addr);
-    ACPI_READ_FIELD(ep_table->smbios_major_version, addr);
-    ACPI_READ_FIELD(ep_table->smbios_minor_version, addr);
-    ACPI_READ_FIELD(ep_table->max_structure_size, addr);
-    ACPI_READ_FIELD(ep_table->entry_point_revision, addr);
-    ACPI_READ_ARRAY(ep_table->formatted_area, addr);
-    ACPI_READ_ARRAY(ep_table->intermediate_anchor_string, addr);
-    g_assert(!memcmp(ep_table->intermediate_anchor_string, "_DMI_", 5));
-    ACPI_READ_FIELD(ep_table->intermediate_checksum, addr);
-    ACPI_READ_FIELD(ep_table->structure_table_length, addr);
-    g_assert_cmpuint(ep_table->structure_table_length, >, 0);
-    ACPI_READ_FIELD(ep_table->structure_table_address, addr);
-    ACPI_READ_FIELD(ep_table->number_of_structures, addr);
-    g_assert_cmpuint(ep_table->number_of_structures, >, 0);
-    ACPI_READ_FIELD(ep_table->smbios_bcd_revision, addr);
-    g_assert(!acpi_checksum((uint8_t *)ep_table, sizeof *ep_table));
-    g_assert(!acpi_checksum((uint8_t *)ep_table + 0x10,
-                            sizeof *ep_table - 0x10));
 }
 
 static inline bool smbios_single_instance(uint8_t type)
@@ -767,8 +780,7 @@
         }
     }
 
-    test_smbios_ep_address(data);
-    test_smbios_ep_table(data);
+    test_smbios_entry_point(data);
     test_smbios_structs(data);
 
     qtest_quit(global_qtest);
diff --git a/tests/endianness-test.c b/tests/endianness-test.c
index 92e17d2..26ee734 100644
--- a/tests/endianness-test.c
+++ b/tests/endianness-test.c
@@ -31,8 +31,6 @@
 
 static const TestCase test_cases[] = {
     { "i386", "pc", -1 },
-    { "mips", "magnum", 0x90000000, .bswap = true },
-    { "mips", "pica61", 0x90000000, .bswap = true },
     { "mips", "mips", 0x14000000, .bswap = true },
     { "mips", "malta", 0x10000000, .bswap = true },
     { "mips64", "magnum", 0x90000000, .bswap = true },
diff --git a/trace-events b/trace-events
index a589650..6060d36 100644
--- a/trace-events
+++ b/trace-events
@@ -193,10 +193,8 @@
 ecc_diag_mem_readb(uint64_t addr, uint32_t ret) "Read diagnostic %"PRId64"= %02x"
 
 # hw/nvram/fw_cfg.c
-fw_cfg_write(void *s, uint8_t value) "%p %d"
 fw_cfg_select(void *s, uint16_t key, int ret) "%p key %d = %d"
 fw_cfg_read(void *s, uint8_t ret) "%p = %d"
-fw_cfg_add_file_dupe(void *s, char *name) "%p %s"
 fw_cfg_add_file(void *s, int index, char *name, size_t len) "%p #%d: %s (%zd bytes)"
 
 # hw/block/hd-geometry.c
@@ -282,6 +280,12 @@
 slavio_timer_mem_writel_mode_invalid(void) "not system timer"
 slavio_timer_mem_writel_invalid(uint64_t addr) "invalid write address %"PRIx64
 
+# hw/dma/rc4030.c
+jazzio_read(uint64_t addr, uint32_t ret) "read reg[0x%"PRIx64"] = 0x%x"
+jazzio_write(uint64_t addr, uint32_t val) "write reg[0x%"PRIx64"] = 0x%x"
+rc4030_read(uint64_t addr, uint32_t ret) "read reg[0x%"PRIx64"] = 0x%x"
+rc4030_write(uint64_t addr, uint32_t val) "write reg[0x%"PRIx64"] = 0x%x"
+
 # hw/dma/sparc32_dma.c
 ledma_memory_read(uint64_t addr) "DMA read addr 0x%"PRIx64
 ledma_memory_write(uint64_t addr) "DMA write addr 0x%"PRIx64
@@ -1167,13 +1171,28 @@
 vmware_scratch_write(uint32_t index, uint32_t value) "index %d, value 0x%x"
 vmware_setmode(uint32_t w, uint32_t h, uint32_t bpp) "%dx%d @ %d bpp"
 
-# savevm.c
+# hw/display/virtio-gpu.c
+virtio_gpu_cmd_get_display_info(void) ""
+virtio_gpu_cmd_get_caps(void) ""
+virtio_gpu_cmd_set_scanout(uint32_t id, uint32_t res, uint32_t w, uint32_t h, uint32_t x, uint32_t y) "id %d, res 0x%x, w %d, h %d, x %d, y %d"
+virtio_gpu_cmd_res_create_2d(uint32_t res, uint32_t fmt, uint32_t w, uint32_t h) "res 0x%x, fmt 0x%x, w %d, h %d"
+virtio_gpu_cmd_res_create_3d(uint32_t res, uint32_t fmt, uint32_t w, uint32_t h, uint32_t d) "res 0x%x, fmt 0x%x, w %d, h %d, d %d"
+virtio_gpu_cmd_res_unref(uint32_t res) "res 0x%x"
+virtio_gpu_cmd_res_back_attach(uint32_t res) "res 0x%x"
+virtio_gpu_cmd_res_back_detach(uint32_t res) "res 0x%x"
+virtio_gpu_cmd_res_xfer_toh_2d(uint32_t res) "res 0x%x"
+virtio_gpu_cmd_res_flush(uint32_t res, uint32_t w, uint32_t h, uint32_t x, uint32_t y) "res 0x%x, w %d, h %d, x %d, y %d"
+virtio_gpu_fence_ctrl(uint64_t fence, uint32_t type) "fence 0x%" PRIx64 ", type 0x%x"
+virtio_gpu_fence_resp(uint64_t fence) "fence 0x%" PRIx64
+
+# migration/savevm.c
 qemu_loadvm_state_section(unsigned int section_type) "%d"
 qemu_loadvm_state_section_partend(uint32_t section_id) "%u"
 qemu_loadvm_state_section_startfull(uint32_t section_id, const char *idstr, uint32_t instance_id, uint32_t version_id) "%u(%s) %u %u"
 savevm_section_start(const char *id, unsigned int section_id) "%s, section_id %u"
 savevm_section_end(const char *id, unsigned int section_id, int ret) "%s, section_id %u -> %d"
 savevm_state_begin(void) ""
+savevm_state_header(void) ""
 savevm_state_iterate(void) ""
 savevm_state_complete(void) ""
 savevm_state_cancel(void) ""
@@ -1193,7 +1212,7 @@
 # qemu-file.c
 qemu_file_fclose(void) ""
 
-# arch_init.c
+# migration/ram.c
 migration_bitmap_sync_start(void) ""
 migration_bitmap_sync_end(uint64_t dirty_pages) "dirty_pages %" PRIu64""
 migration_throttle(void) ""
@@ -1564,6 +1583,18 @@
 vfio_get_device(const char * name, unsigned int flags, unsigned int num_regions, unsigned int num_irqs) "Device %s flags: %u, regions: %u, irqs: %u"
 vfio_put_base_device(int fd) "close vdev->fd=%d"
 
+# hw/vfio/platform.c
+vfio_platform_populate_regions(int region_index, unsigned long flag, unsigned long size, int fd, unsigned long offset) "- region %d flags = 0x%lx, size = 0x%lx, fd= %d, offset = 0x%lx"
+vfio_platform_base_device_init(char *name, int groupid) "%s belongs to group #%d"
+vfio_platform_realize(char *name, char *compat) "vfio device %s, compat = %s"
+vfio_platform_eoi(int pin, int fd) "EOI IRQ pin %d (fd=%d)"
+vfio_platform_mmap_set_enabled(bool enabled) "fast path = %d"
+vfio_platform_intp_mmap_enable(int pin) "IRQ #%d still active, stay in slow path"
+vfio_platform_intp_interrupt(int pin, int fd) "Inject IRQ #%d (fd = %d)"
+vfio_platform_intp_inject_pending_lockheld(int pin, int fd) "Inject pending IRQ #%d (fd = %d)"
+vfio_platform_populate_interrupts(int pin, int count, int flags) "- IRQ index %d: count %d, flags=0x%x"
+vfio_intp_interrupt_set_pending(int index) "irq %d is set PENDING"
+
 #hw/acpi/memory_hotplug.c
 mhp_acpi_invalid_slot_selected(uint32_t slot) "0x%"PRIx32
 mhp_acpi_ejecting_invalid_slot(uint32_t slot) "0x%"PRIx32
diff --git a/ui/gtk.c b/ui/gtk.c
index 126326a..df2a79e 100644
--- a/ui/gtk.c
+++ b/ui/gtk.c
@@ -1917,12 +1917,19 @@
 #endif
 }
 
+static gboolean gtkinit;
+
 void gtk_display_init(DisplayState *ds, bool full_screen, bool grab_on_hover)
 {
     GtkDisplayState *s = g_malloc0(sizeof(*s));
     char *filename;
     GdkDisplay *window_display;
 
+    if (!gtkinit) {
+        fprintf(stderr, "gtk initialization failed\n");
+        exit(1);
+    }
+
     s->window = gtk_window_new(GTK_WINDOW_TOPLEVEL);
 #if GTK_CHECK_VERSION(3, 2, 0)
     s->vbox = gtk_box_new(GTK_ORIENTATION_VERTICAL, 0);
@@ -2003,7 +2010,11 @@
 
 void early_gtk_display_init(int opengl)
 {
-    gtk_init(NULL, NULL);
+    gtkinit = gtk_init_check(NULL, NULL);
+    if (!gtkinit) {
+        /* don't exit yet, that'll break -help */
+        return;
+    }
 
     switch (opengl) {
     case -1: /* default */
diff --git a/ui/sdl2.c b/ui/sdl2.c
index 2d60179..5cb75aa 100644
--- a/ui/sdl2.c
+++ b/ui/sdl2.c
@@ -521,6 +521,10 @@
 {
     struct sdl2_console *scon = get_scon_from_window(ev->window.windowID);
 
+    if (!scon) {
+        return;
+    }
+
     switch (ev->window.event) {
     case SDL_WINDOWEVENT_RESIZED:
         {
diff --git a/ui/spice-display.c b/ui/spice-display.c
index 9c63132..cc4a6ce 100644
--- a/ui/spice-display.c
+++ b/ui/spice-display.c
@@ -199,7 +199,7 @@
     static const int blksize = 32;
     int blocks = (surface_width(ssd->ds) + blksize - 1) / blksize;
     int dirty_top[blocks];
-    int y, yoff, x, xoff, blk, bw;
+    int y, yoff1, yoff2, x, xoff, blk, bw;
     int bpp = surface_bytes_per_pixel(ssd->ds);
     uint8_t *guest, *mirror;
 
@@ -214,13 +214,14 @@
     guest = surface_data(ssd->ds);
     mirror = (void *)pixman_image_get_data(ssd->mirror);
     for (y = ssd->dirty.top; y < ssd->dirty.bottom; y++) {
-        yoff = y * surface_stride(ssd->ds);
+        yoff1 = y * surface_stride(ssd->ds);
+        yoff2 = y * pixman_image_get_stride(ssd->mirror);
         for (x = ssd->dirty.left; x < ssd->dirty.right; x += blksize) {
             xoff = x * bpp;
             blk = x / blksize;
             bw = MIN(blksize, ssd->dirty.right - x);
-            if (memcmp(guest + yoff + xoff,
-                       mirror + yoff + xoff,
+            if (memcmp(guest + yoff1 + xoff,
+                       mirror + yoff2 + xoff,
                        bw * bpp) == 0) {
                 if (dirty_top[blk] != -1) {
                     QXLRect update = {
@@ -660,7 +661,10 @@
 {
     SimpleSpiceDisplay *ssd = container_of(sin, SimpleSpiceDisplay, qxl);
     QemuUIInfo info;
-    int rc;
+
+    if (!dpy_ui_info_supported(ssd->dcl.con)) {
+        return 0; /* == not supported by guest */
+    }
 
     if (!mc) {
         return 1;
@@ -675,14 +679,10 @@
         info.width  = mc->monitors[0].width;
         info.height = mc->monitors[0].height;
     }
-    rc = dpy_set_ui_info(ssd->dcl.con, &info);
-    dprint(1, "%s/%d: size %dx%d, rc %d   <---   ==========================\n",
-           __func__, ssd->qxl.id, info.width, info.height, rc);
-    if (rc != 0) {
-        return 0; /* == not supported by guest */
-    } else {
-        return 1;
-    }
+    dpy_set_ui_info(ssd->dcl.con, &info);
+    dprint(1, "%s/%d: size %dx%d\n", __func__, ssd->qxl.id,
+           info.width, info.height);
+    return 1;
 }
 
 static const QXLInterface dpy_interface = {
diff --git a/util/qemu-config.c b/util/qemu-config.c
index 35adfda..6cfdd72 100644
--- a/util/qemu-config.c
+++ b/util/qemu-config.c
@@ -6,7 +6,7 @@
 #include "qapi/error.h"
 #include "qmp-commands.h"
 
-static QemuOptsList *vm_config_groups[32];
+static QemuOptsList *vm_config_groups[48];
 static QemuOptsList *drive_config_groups[4];
 
 static QemuOptsList *find_list(QemuOptsList **lists, const char *group,
diff --git a/vl.c b/vl.c
index d4b2d03..9542095 100644
--- a/vl.c
+++ b/vl.c
@@ -492,6 +492,25 @@
     },
 };
 
+static QemuOptsList qemu_fw_cfg_opts = {
+    .name = "fw_cfg",
+    .implied_opt_name = "name",
+    .head = QTAILQ_HEAD_INITIALIZER(qemu_fw_cfg_opts.head),
+    .desc = {
+        {
+            .name = "name",
+            .type = QEMU_OPT_STRING,
+            .help = "Sets the fw_cfg name of the blob to be inserted",
+        }, {
+            .name = "file",
+            .type = QEMU_OPT_STRING,
+            .help = "Sets the name of the file from which\n"
+                    "the fw_cfg blob will be loaded",
+        },
+        { /* end of list */ }
+    },
+};
+
 /**
  * Get machine options
  *
@@ -2127,6 +2146,38 @@
     return NULL;
 }
 
+static int parse_fw_cfg(void *opaque, QemuOpts *opts, Error **errp)
+{
+    gchar *buf;
+    size_t size;
+    const char *name, *file;
+
+    if (opaque == NULL) {
+        error_report("fw_cfg device not available");
+        return -1;
+    }
+    name = qemu_opt_get(opts, "name");
+    file = qemu_opt_get(opts, "file");
+    if (name == NULL || *name == '\0' || file == NULL || *file == '\0') {
+        error_report("invalid argument value");
+        return -1;
+    }
+    if (strlen(name) > FW_CFG_MAX_FILE_PATH - 1) {
+        error_report("name too long (max. %d char)", FW_CFG_MAX_FILE_PATH - 1);
+        return -1;
+    }
+    if (strncmp(name, "opt/", 4) != 0) {
+        error_report("WARNING: externally provided fw_cfg item names "
+                     "should be prefixed with \"opt/\"!");
+    }
+    if (!g_file_get_contents(file, &buf, &size, NULL)) {
+        error_report("can't load %s", file);
+        return -1;
+    }
+    fw_cfg_add_file((FWCfgState *)opaque, name, buf, size);
+    return 0;
+}
+
 static int device_help_func(void *opaque, QemuOpts *opts, Error **errp)
 {
     return qdev_device_help(opts);
@@ -2822,6 +2873,7 @@
     qemu_add_opts(&qemu_numa_opts);
     qemu_add_opts(&qemu_icount_opts);
     qemu_add_opts(&qemu_semihosting_config_opts);
+    qemu_add_opts(&qemu_fw_cfg_opts);
 
     runstate_init();
 
@@ -3438,6 +3490,12 @@
                 }
                 do_smbios_option(opts);
                 break;
+            case QEMU_OPTION_fwcfg:
+                opts = qemu_opts_parse(qemu_find_opts("fw_cfg"), optarg, 1);
+                if (opts == NULL) {
+                    exit(1);
+                }
+                break;
             case QEMU_OPTION_enable_kvm:
                 olist = qemu_find_opts("machine");
                 qemu_opts_parse(olist, "accel=kvm", 0);
@@ -4274,6 +4332,11 @@
 
     numa_post_machine_init();
 
+    if (qemu_opts_foreach(qemu_find_opts("fw_cfg"),
+                          parse_fw_cfg, fw_cfg_find(), NULL) != 0) {
+        exit(1);
+    }
+
     /* init USB devices */
     if (usb_enabled()) {
         if (foreach_device_config(DEV_USB, usb_parse) < 0)