Update for 0.12.4 release

Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Workaround for broken OSS_GETVERSION on FreeBSD, part two
2010-05-04 09:17:19 -05:00 · 2010-04-25 12:20:26 +02:00 · 2010-04-25 12:20:10 +02:00 · 2010-04-25 12:20:01 +02:00 · 2010-04-25 12:19:48 +02:00 · 2010-04-25 12:19:39 +02:00
90 changed files with 1447 additions and 740 deletions
@@ -1,3 +1,143 @@
+version 0.12.4
+ - Workaround for broken OSS_GETVERSION on FreeBSD, part two (Juergen Lock)
+ - oss: fix fragment setting (malc)
+ - oss: issue OSS_GETVERSION ioctl only when needed (malc)
+ - oss: refactor code around policy setting (malc)
+ - oss: workaround for cases when OSS_GETVERSION is not defined (malc)
+ - block: Free iovec arrays allocated by multiwrite_merge() (Stefan Hajnoczi)
+ - lsi: fix segfault in lsi_command_complete (Gerd Hoffmann)
+ - lsi: pass lsi_request to lsi_reselect (Gerd Hoffmann)
+ - lsi: move dma_len+dma_buf into lsi_request (Gerd Hoffmann)
+ - lsi: move current_dev into lsi_request (Gerd Hoffmann)
+ - lsi: have lsi_request for the whole life time of the request. (Gerd Hoffmann)
+ - lsi: use QTAILQ for lsi_queue (Gerd Hoffmann)
+ - tcp/mips: Change TCG_AREG0 (fp -> s0) (Stefan Weil)
+ - sh_pci: fix memory and I/O access (Aurelien Jarno)
+ - Fix incoming migration with iothread (Marcelo Tosatti)
+ - Fix SIGFPE for vnc display of width/height = 1 (Chris Webb)
+ - net: remove broken net_set_boot_mask() boot device validation (Eduardo Habkost)
+ - qcow2: Remove request from in-flight list after error (Kevin Wolf)
+ - qcow2: Don't ignore immediate read/write failures (Kevin Wolf)
+ - block: Fix multiwrite memory leak in error case (Kevin Wolf)
+ - block: Fix error code in multiwrite for immediate failures (Kevin Wolf)
+ - block: Fix multiwrite error handling (Kevin Wolf)
+ - scsi-disk: fix buffer overflow (Gerd Hoffmann)
+ - qcow2: Rewrite alloc_refcount_block/grow_refcount_table (Kevin Wolf)
+ - qcow2: Factor next_refcount_table_size out (Kevin Wolf)
+ - block: avoid creating too large iovecs in multiwrite_merge (Christoph Hellwig)
+ - json-parser: Fix segfault on malformed input (Kevin Wolf)
+ - linux-user: switch default ppc64 CPU to 970fx from 970 (Aurelien Jarno)
+ - target-sh4: MMU: fix store queue addresses (Aurelien Jarno)
+ - target-sh4: MMU: fix ITLB priviledge check (Aurelien Jarno)
+ - target-sh4: MMU: fix mem_idx computation (Aurelien Jarno)
+ - sh7750: handle MMUCR TI bit (Aurelien Jarno)
+ - UHCI spurious interrut fix (Paul Brook)
+ - tcg/mips: fix branch offset during retranslation (Aurelien Jarno)
+ - tcg/arm: correctly save/restore registers in prologue/epilogue (Aurelien Jarno)
+ - workaround for cmd646 bmdma register access while no dma is active (Igor V. Kovalenko)
+ - Fix corner case in chardev udp: parameter (Jan Kiszka)
+ - Don't set default monitor when there is a mux'ed one (Jan Kiszka)
+ - spelling typo (compatibilty) in hw/fw_cfg.c (Vagrant Cascadian)
+ - fdc: fix drive property handling. (Gerd Hoffmann)
+ - target-i386: fix commit c22549204a6edc431e8e4358e61bd56386ff6957 (TeLeMan)
+ - target-i386: fix SIB decoding with index = 4 (Aurelien Jarno)
+ - Fix segfault with ram_size > 4095M without kvm (Ryan Harper)
+ - target-i386: Fix long jumps/calls in long mode with REX.W set (malc)
+ - target-i386: fix lddqu SSE instruction (Aurelien Jarno)
+ - qemu-char.c: drop debug printfs from qemu_chr_parse_compat (Jan Kiszka)
+ - fix undefined shifts by >32 (Paolo Bonzini)
+ - Fix qemu -net user,hostfwd= example (Aurelien Jarno)
+
+version 0.12.3
+  - kvm: Fix eflags corruption in kvm mode (Jan Kiszka)
+  - qcow2: Fix access after end of array (Kevin Wolf)
+  - ide save/restore pio/atapi cmd transfer fields and io buffer (Marcelo Tosatti)
+  - net: Monitor command set_link finds only VLAN clients, fix (Markus Armbruster)
+  - net: info network shows only VLAN clients, fix (Markus Armbruster)
+  - net: net_check_clients() checks only VLAN clients, fix (Markus Armbruster)
+  - net: Fix bogus "Warning: vlan 0 with no nics" with -device (Markus Armbruster)
+  - net: net_check_clients() runs too early to see -device, fix (Markus Armbruster)
+  - net: Remove unused net_client_uninit() (Markus Armbruster)
+  - don't dereference NULL after failed strdup (Jim Meyering)
+  - virtio-net: fix network stall under load (Tom Lendacky)
+  - json: fix PRId64 on Win32 (Roy Tam)
+  - fix inet_parse typo (Marcelo Tosatti)
+  - iothread: fix vcpu stop with smp tcg (Marcelo Tosatti)
+  - segfault due to buffer overrun in usb-serial (David S. Ahern)
+  - qcow2: Fix signedness bugs (Kevin Wolf)
+  - Do not ignore error, if open file failed (-serial /dev/tty) (Evgeniy Dushistov)
+  - pc-bios: update to newer version of (stable) seabios (Anthony Liguori)
+  - target-mips: fix ROTR and DROTR by zero (Aurelien Jarno)
+  - target-mips: fix CpU exception for coprocessor 0 (Nathan Froyd)
+  - tcg/mips: fix crash in tcg_out_qemu_ld() (Aurelien Jarno)
+  - target-mips: don't call cpu_loop_exit() from helper.c (Aurelien Jarno)
+  - virtio-blk: Fix error cases which ignored rerror/werror (Kevin Wolf)
+  - virtio-blk: Fix restart after read error (Kevin Wolf)
+  - virtio_blk: Factor virtio_blk_handle_request out (Kevin Wolf)
+  - cirrus: Properly re-register cirrus_linear_io_addr on vram unmap (Jan Kiszka)
+  - qcow2: Don't ignore qcow2_alloc_clusters return value (Kevin Wolf)
+  - qcow2: Don't ignore update_refcount return value (Kevin Wolf)
+  - qcow2: Allow updating no refcounts (Kevin Wolf)
+  - qcow2: Improve error handling in update_refcount (Kevin Wolf)
+  - qcow2: Fix error handling in grow_refcount_table (Kevin Wolf)
+  - block: Return original error codes in bdrv_pread/write (Kevin Wolf)
+  - qcow2: Return 0/-errno in qcow2_alloc_cluster_offset (Kevin Wolf)
+  - qcow2: Return 0/-errno in get_cluster_table (Kevin Wolf)
+  - qcow2: Fix error handling in qcow_save_vmstate (Kevin Wolf)
+  - qcow2: Fix error handling in qcow2_grow_l1_table (Kevin Wolf)
+  - win32/sdl: Fix toggle full screen (Herve Poussineau)
+  - win32: pair qemu_memalign() with qemu_vfree() (Herve Poussineau)
+  - vnc_refresh: calling vnc_update_client might free vs (Stefano Stabellini)
+  - Musicpal: Fix descriptor walk in eth_send (Jan Kiszka)
+  - Musicpal: Fix wm8750 I2C address (Jan Kiszka)
+  - fix savevm command without id or tag (Marcelo Tosatti)
+  - reduce number of reinjects on ACK (Gleb Natapov)
+  - QMP: Fix asynchronous events delivery (Luiz Capitulino)
+  - Documentation: Add missing documentation for qdev related command line options (Stefan Weil)
+  - pc: add driver version compat properties (Gerd Hoffmann)
+  - scsi: device version property (Gerd Hoffmann)
+  - ide: device version property (Gerd Hoffmann)
+  - QMP: Emit asynchronous events on all QMP monitors (Adam Litke)
+  - Fix QEMU_WARN_UNUSED_RESULT (Kevin Wolf)
+
+version 0.12.2:
+  - Qemu's internal TFTP server breaks lock-step-iness of TFTP (Milan Plzik)
+  - osdep.c: Fix accept4 fallback (Kevin Wolf)
+  - pc: add rombar to compat properties for pc-0.10 and pc-0.11 (Gerd Hoffmann)
+  - pci: allow loading roms via fw_cfg. (Gerd Hoffmann)
+  - roms: rework rom loading via fw (Gerd Hoffmann)
+  - fw_cfg: rom loader tweaks. (Gerd Hoffmann)
+  - roms: minor fixes and cleanups. (Gerd Hoffmann)
+  - pc: add machine type for 0.12 (Gerd Hoffmann)
+  - loader: more ignores for rom intended to be loaded by the bios (Aurelien Jarno)
+  - vnc_refresh: return if vd->timer is NULL (Stefano Stabellini)
+  - QMP: Don't free async event's 'data' (Luiz Capitulino)
+  - Handle TFTP ERROR from client (Thomas Horsten)
+  - dmg: fix ->open failure (Christoph Hellwig)
+  - virtio-pci: thinko fix (Michael S. Tsirkin)
+  - pc-bios: Update README (SeaBIOS) (Stefan Weil)
+  - vmware_vga: Check cursor dimensions passed from guest to avoid buffer overflow (Roland Dreier)
+  - remove pending exception on vcpu reset. (Gleb Natapov)
+  - Fix CPU topology initialization (Jiri Denemark)
+  - MCE: Fix bug of IA32_MCG_STATUS after system reset (Huang Ying)
+  - linuxboot: fix gdt address calculation (Avi Kivity)
+  - QMP: Drop wrong assert() (Luiz Capitulino)
+  - vnc: Fix artifacts in hextile decoding (Anthony Liguori)
+  - target-i386: Fix "call im" on x86_64 when executing 32-bit code (Aurelien Jarno)
+  - Add missing newline at the end of options list (Michael Tokarev)
+  - Don't load options roms intended to be loaded by the bios in qemu (Avi Kivity)
+  - USB: Improve usbdevice error messages (Scott Tsai)
+  - cpu-all.h: fix cpu_get_real_ticks() #ifdef (Aurelien Jarno)
+  - alpha: fix compile (Blue Swirl)
+  - user_only: compile everything with -fpie (Kirill A. Shutemov)
+  - fdc/sparc32: don't hang on detection under OBP (Artyom Tarasenko)
+  - scsi-disk: Inquiry with allocation length of CDB < 36 (v4) (Artyom Tarasenko)
+  - e1000: fix init values for command register (Michael S. Tsirkin)
+
+version 0.12.1:
+  - loader: fix rom loading at address 0 (fixes target-arm) (Aurelien Jarno)
+  - loader: fix rom_copy (fixes multiboot) (Kevin Wolf)
+
 version 0.12.0:

  - Update to SeaBIOS 0.5.0
@@ -2,10 +2,15 @@

 include ../config-host.mak
 include $(SRC_PATH)/rules.mak
+-include config.mak

 .PHONY: all

-VPATH=$(SRC_PATH)
+# Do not take %.o from $(SRC_PATH), only %.c and %.h
+# All %.o for user targets should be built with -fpie, when
+# configured with --enable-user-pie, so we don't want to
+# take %.o from $(SRC_PATH), since they built without -fpie
+vpath %.c %.h $(SRC_PATH)

 QEMU_CFLAGS+=-I..

@@ -1 +1 @@
-0.12.0
+0.12.4
@@ -38,6 +38,10 @@
 #define AUDIO_CAP "oss"
 #include "audio_int.h"

+#if defined OSS_GETVERSION && defined SNDCTL_DSP_POLICY
+#define USE_DSP_POLICY
+#endif
+
 typedef struct OSSVoiceOut {
    HWVoiceOut hw;
    void *pcm_buf;
@@ -236,14 +240,39 @@ static void oss_dump_info (struct oss_params *req, struct oss_params *obt)
 }
 #endif

+#ifdef USE_DSP_POLICY
+static int oss_get_version (int fd, int *version, const char *typ)
+{
+    if (ioctl (fd, OSS_GETVERSION, &version)) {
+#if defined(__FreeBSD__) || defined(__FreeBSD_kernel__)
+        /*
+         * Looks like atm (20100109) FreeBSD knows OSS_GETVERSION
+         * since 7.x, but currently only on the mixer device (or in
+         * the Linuxolator), and in the native version that part of
+         * the code is in fact never reached so the ioctl fails anyway.
+         * Until this is fixed, just check the errno and if its what
+         * FreeBSD's sound drivers return atm assume they are new enough.
+         */
+        if (errno == EINVAL) {
+            *version = 0x040000;
+            return 0;
+        }
+#endif
+        oss_logerr2 (errno, typ, "Failed to get OSS version\n");
+        return -1;
+    }
+    return 0;
+}
+#endif
+
 static int oss_open (int in, struct oss_params *req,
                     struct oss_params *obt, int *pfd)
 {
    int fd;
-    int version;
    int oflags = conf.exclusive ? O_EXCL : 0;
    audio_buf_info abinfo;
    int fmt, freq, nchannels;
+    int setfragment = 1;
    const char *dspname = in ? conf.devpath_in : conf.devpath_out;
    const char *typ = in ? "ADC" : "DAC";

@@ -281,27 +310,30 @@ static int oss_open (int in, struct oss_params *req,
        goto err;
    }

-    if (ioctl (fd, OSS_GETVERSION, &version)) {
-        oss_logerr2 (errno, typ, "Failed to get OSS version\n");
-        version = 0;
-    }
+#ifdef USE_DSP_POLICY
+    if (conf.policy >= 0) {
+        int version;

-    if (conf.debug) {
-        dolog ("OSS version = %#x\n", version);
-    }
+        if (!oss_get_version (fd, &version, typ)) {
+            if (conf.debug) {
+                dolog ("OSS version = %#x\n", version);
+            }

-#ifdef SNDCTL_DSP_POLICY
-    if (conf.policy >= 0 && version >= 0x040000) {
-        int policy = conf.policy;
-        if (ioctl (fd, SNDCTL_DSP_POLICY, &policy)) {
-            oss_logerr2 (errno, typ, "Failed to set timing policy to %d\n",
-                         conf.policy);
-            goto err;
+            if (version >= 0x040000) {
+                int policy = conf.policy;
+                if (ioctl (fd, SNDCTL_DSP_POLICY, &policy)) {
+                    oss_logerr2 (errno, typ,
+                                 "Failed to set timing policy to %d\n",
+                                 conf.policy);
+                    goto err;
+                }
+                setfragment = 0;
+            }
        }
    }
-    else
 #endif
-    {
+
+    if (setfragment) {
        int mmmmssss = (req->nfrags << 16) | ctz32 (req->fragsize);
        if (ioctl (fd, SNDCTL_DSP_SETFRAGMENT, &mmmmssss)) {
            oss_logerr2 (errno, typ, "Failed to set buffer length (%d, %d)\n",
@@ -857,7 +889,7 @@ static struct audio_option oss_options[] = {
        .valp  = &conf.exclusive,
        .descr = "Open device in exclusive mode (vmix wont work)"
    },
-#ifdef SNDCTL_DSP_POLICY
+#ifdef USE_DSP_POLICY
    {
        .name  = "POLICY",
        .tag   = AUD_OPT_INT,
@@ -691,6 +691,7 @@ int bdrv_pread(BlockDriverState *bs, int64_t offset,
    uint8_t tmp_buf[BDRV_SECTOR_SIZE];
    int len, nb_sectors, count;
    int64_t sector_num;
+    int ret;

    count = count1;
    /* first read to align to sector start */
@@ -699,8 +700,8 @@ int bdrv_pread(BlockDriverState *bs, int64_t offset,
        len = count;
    sector_num = offset >> BDRV_SECTOR_BITS;
    if (len > 0) {
-        if (bdrv_read(bs, sector_num, tmp_buf, 1) < 0)
-            return -EIO;
+        if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
+            return ret;
        memcpy(buf, tmp_buf + (offset & (BDRV_SECTOR_SIZE - 1)), len);
        count -= len;
        if (count == 0)
@@ -712,8 +713,8 @@ int bdrv_pread(BlockDriverState *bs, int64_t offset,
    /* read the sectors "in place" */
    nb_sectors = count >> BDRV_SECTOR_BITS;
    if (nb_sectors > 0) {
-        if (bdrv_read(bs, sector_num, buf, nb_sectors) < 0)
-            return -EIO;
+        if ((ret = bdrv_read(bs, sector_num, buf, nb_sectors)) < 0)
+            return ret;
        sector_num += nb_sectors;
        len = nb_sectors << BDRV_SECTOR_BITS;
        buf += len;
@@ -722,8 +723,8 @@ int bdrv_pread(BlockDriverState *bs, int64_t offset,

    /* add data from the last sector */
    if (count > 0) {
-        if (bdrv_read(bs, sector_num, tmp_buf, 1) < 0)
-            return -EIO;
+        if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
+            return ret;
        memcpy(buf, tmp_buf, count);
    }
    return count1;
@@ -735,6 +736,7 @@ int bdrv_pwrite(BlockDriverState *bs, int64_t offset,
    uint8_t tmp_buf[BDRV_SECTOR_SIZE];
    int len, nb_sectors, count;
    int64_t sector_num;
+    int ret;

    count = count1;
    /* first write to align to sector start */
@@ -743,11 +745,11 @@ int bdrv_pwrite(BlockDriverState *bs, int64_t offset,
        len = count;
    sector_num = offset >> BDRV_SECTOR_BITS;
    if (len > 0) {
-        if (bdrv_read(bs, sector_num, tmp_buf, 1) < 0)
-            return -EIO;
+        if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
+            return ret;
        memcpy(tmp_buf + (offset & (BDRV_SECTOR_SIZE - 1)), buf, len);
-        if (bdrv_write(bs, sector_num, tmp_buf, 1) < 0)
-            return -EIO;
+        if ((ret = bdrv_write(bs, sector_num, tmp_buf, 1)) < 0)
+            return ret;
        count -= len;
        if (count == 0)
            return count1;
@@ -758,8 +760,8 @@ int bdrv_pwrite(BlockDriverState *bs, int64_t offset,
    /* write the sectors "in place" */
    nb_sectors = count >> BDRV_SECTOR_BITS;
    if (nb_sectors > 0) {
-        if (bdrv_write(bs, sector_num, buf, nb_sectors) < 0)
-            return -EIO;
+        if ((ret = bdrv_write(bs, sector_num, buf, nb_sectors)) < 0)
+            return ret;
        sector_num += nb_sectors;
        len = nb_sectors << BDRV_SECTOR_BITS;
        buf += len;
@@ -768,11 +770,11 @@ int bdrv_pwrite(BlockDriverState *bs, int64_t offset,

    /* add data from the last sector */
    if (count > 0) {
-        if (bdrv_read(bs, sector_num, tmp_buf, 1) < 0)
-            return -EIO;
+        if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
+            return ret;
        memcpy(tmp_buf, buf, count);
-        if (bdrv_write(bs, sector_num, tmp_buf, 1) < 0)
-            return -EIO;
+        if ((ret = bdrv_write(bs, sector_num, tmp_buf, 1)) < 0)
+            return ret;
    }
    return count1;
 }
@@ -1606,8 +1608,11 @@ static void multiwrite_user_cb(MultiwriteCB *mcb)

    for (i = 0; i < mcb->num_callbacks; i++) {
        mcb->callbacks[i].cb(mcb->callbacks[i].opaque, mcb->error);
+        if (mcb->callbacks[i].free_qiov) {
+            qemu_iovec_destroy(mcb->callbacks[i].free_qiov);
+        }
        qemu_free(mcb->callbacks[i].free_qiov);
-        qemu_free(mcb->callbacks[i].free_buf);
+        qemu_vfree(mcb->callbacks[i].free_buf);
    }
 }

@@ -1615,7 +1620,7 @@ static void multiwrite_cb(void *opaque, int ret)
 {
    MultiwriteCB *mcb = opaque;

-    if (ret < 0) {
+    if (ret < 0 && !mcb->error) {
        mcb->error = ret;
        multiwrite_user_cb(mcb);
    }
@@ -1667,6 +1672,10 @@ static int multiwrite_merge(BlockDriverState *bs, BlockRequest *reqs,
            merge = bs->drv->bdrv_merge_requests(bs, &reqs[outidx], &reqs[i]);
        }

+        if (reqs[outidx].qiov->niov + reqs[i].qiov->niov + 1 > IOV_MAX) {
+            merge = 0;
+        }
+
        if (merge) {
            size_t size;
            QEMUIOVector *qiov = qemu_mallocz(sizeof(*qiov));
@@ -1752,10 +1761,11 @@ int bdrv_aio_multiwrite(BlockDriverState *bs, BlockRequest *reqs, int num_reqs)
            // submitted yet. Otherwise we'll wait for the submitted AIOs to
            // complete and report the error in the callback.
            if (mcb->num_requests == 0) {
-                reqs[i].error = EIO;
+                reqs[i].error = -EIO;
                goto fail;
            } else {
-                mcb->error = EIO;
+                mcb->num_requests++;
+                multiwrite_cb(mcb, -EIO);
                break;
            }
        } else {
@@ -309,7 +309,7 @@ static int curl_open(BlockDriverState *bs, const char *filename, int flags)

    static int inited = 0;

-    file = strdup(filename);
+    file = qemu_strdup(filename);
    s->readahead_size = READ_AHEAD_SIZE;

    /* Parse a trailing ":readahead=#:" param, if present. */
@@ -90,24 +90,21 @@ static int dmg_open(BlockDriverState *bs, const char *filename, int flags)

    /* read offset of info blocks */
    if(lseek(s->fd,-0x1d8,SEEK_END)<0) {
-dmg_close:
-	close(s->fd);
-	/* open raw instead */
-	bs->drv=bdrv_find_format("raw");
-	return bs->drv->bdrv_open(bs, filename, flags);
+        goto fail;
    }
+
    info_begin=read_off(s->fd);
    if(info_begin==0)
-	goto dmg_close;
+	goto fail;
    if(lseek(s->fd,info_begin,SEEK_SET)<0)
-	goto dmg_close;
+	goto fail;
    if(read_uint32(s->fd)!=0x100)
-	goto dmg_close;
+	goto fail;
    if((count = read_uint32(s->fd))==0)
-	goto dmg_close;
+	goto fail;
    info_end = info_begin+count;
    if(lseek(s->fd,0xf8,SEEK_CUR)<0)
-	goto dmg_close;
+	goto fail;

    /* read offsets */
    last_in_offset = last_out_offset = 0;
@@ -116,14 +113,14 @@ dmg_close:

 	count = read_uint32(s->fd);
 	if(count==0)
-	    goto dmg_close;
+	    goto fail;
 	type = read_uint32(s->fd);
 	if(type!=0x6d697368 || count<244)
 	    lseek(s->fd,count-4,SEEK_CUR);
 	else {
 	    int new_size, chunk_count;
 	    if(lseek(s->fd,200,SEEK_CUR)<0)
-	        goto dmg_close;
+	        goto fail;
 	    chunk_count = (count-204)/40;
 	    new_size = sizeof(uint64_t) * (s->n_chunks + chunk_count);
 	    s->types = qemu_realloc(s->types, new_size/2);
@@ -142,7 +139,7 @@ dmg_close:
 		    chunk_count--;
 		    i--;
 		    if(lseek(s->fd,36,SEEK_CUR)<0)
-			goto dmg_close;
+			goto fail;
 		    continue;
 		}
 		read_uint32(s->fd);
@@ -163,11 +160,14 @@ dmg_close:
    s->compressed_chunk = qemu_malloc(max_compressed_size+1);
    s->uncompressed_chunk = qemu_malloc(512*max_sectors_per_chunk);
    if(inflateInit(&s->zstream) != Z_OK)
-	goto dmg_close;
+	goto fail;

    s->current_chunk = s->n_chunks;

    return 0;
+fail:
+    close(s->fd);
+    return -1;
 }

 static inline int is_sector_in_chunk(BDRVDMGState* s,
@@ -33,7 +33,7 @@ int qcow2_grow_l1_table(BlockDriverState *bs, int min_size)
    BDRVQcowState *s = bs->opaque;
    int new_l1_size, new_l1_size2, ret, i;
    uint64_t *new_l1_table;
-    uint64_t new_l1_table_offset;
+    int64_t new_l1_table_offset;
    uint8_t data[12];

    new_l1_size = s->l1_size;
@@ -55,6 +55,10 @@ int qcow2_grow_l1_table(BlockDriverState *bs, int min_size)

    /* write new table (align to cluster) */
    new_l1_table_offset = qcow2_alloc_clusters(bs, new_l1_size2);
+    if (new_l1_table_offset < 0) {
+        qemu_free(new_l1_table);
+        return new_l1_table_offset;
+    }

    for(i = 0; i < s->l1_size; i++)
        new_l1_table[i] = cpu_to_be64(new_l1_table[i]);
@@ -67,9 +71,10 @@ int qcow2_grow_l1_table(BlockDriverState *bs, int min_size)
    /* set new table */
    cpu_to_be32w((uint32_t*)data, new_l1_size);
    cpu_to_be64w((uint64_t*)(data + 4), new_l1_table_offset);
-    if (bdrv_pwrite(s->hd, offsetof(QCowHeader, l1_size), data,
-                sizeof(data)) != sizeof(data))
+    ret = bdrv_pwrite(s->hd, offsetof(QCowHeader, l1_size), data,sizeof(data));
+    if (ret != sizeof(data)) {
        goto fail;
+    }
    qemu_free(s->l1_table);
    qcow2_free_clusters(bs, s->l1_table_offset, s->l1_size * sizeof(uint64_t));
    s->l1_table_offset = new_l1_table_offset;
@@ -77,8 +82,9 @@ int qcow2_grow_l1_table(BlockDriverState *bs, int min_size)
    s->l1_size = new_l1_size;
    return 0;
 fail:
-    qemu_free(s->l1_table);
-    return -EIO;
+    qemu_free(new_l1_table);
+    qcow2_free_clusters(bs, new_l1_table_offset, new_l1_size2);
+    return ret < 0 ? ret : -EIO;
 }

 void qcow2_l2_cache_reset(BlockDriverState *bs)
@@ -213,13 +219,17 @@ static uint64_t *l2_allocate(BlockDriverState *bs, int l1_index)
    BDRVQcowState *s = bs->opaque;
    int min_index;
    uint64_t old_l2_offset;
-    uint64_t *l2_table, l2_offset;
+    uint64_t *l2_table;
+    int64_t l2_offset;

    old_l2_offset = s->l1_table[l1_index];

    /* allocate a new l2 entry */

    l2_offset = qcow2_alloc_clusters(bs, s->l2_size * sizeof(uint64_t));
+    if (l2_offset < 0) {
+        return NULL;
+    }

    /* update the L1 entry */

@@ -479,8 +489,8 @@ out:
 * the l2 table offset in the qcow2 file and the cluster index
 * in the l2 table are given to the caller.
 *
+ * Returns 0 on success, -errno in failure case
 */
-
 static int get_cluster_table(BlockDriverState *bs, uint64_t offset,
                             uint64_t **new_l2_table,
                             uint64_t *new_l2_offset,
@@ -496,8 +506,9 @@ static int get_cluster_table(BlockDriverState *bs, uint64_t offset,
    l1_index = offset >> (s->l2_bits + s->cluster_bits);
    if (l1_index >= s->l1_size) {
        ret = qcow2_grow_l1_table(bs, l1_index + 1);
-        if (ret < 0)
-            return 0;
+        if (ret < 0) {
+            return ret;
+        }
    }
    l2_offset = s->l1_table[l1_index];

@@ -507,14 +518,16 @@ static int get_cluster_table(BlockDriverState *bs, uint64_t offset,
        /* load the l2 table in memory */
        l2_offset &= ~QCOW_OFLAG_COPIED;
        l2_table = l2_load(bs, l2_offset);
-        if (l2_table == NULL)
-            return 0;
+        if (l2_table == NULL) {
+            return -EIO;
+        }
    } else {
        if (l2_offset)
            qcow2_free_clusters(bs, l2_offset, s->l2_size * sizeof(uint64_t));
        l2_table = l2_allocate(bs, l1_index);
-        if (l2_table == NULL)
-            return 0;
+        if (l2_table == NULL) {
+            return -EIO;
+        }
        l2_offset = s->l1_table[l1_index] & ~QCOW_OFLAG_COPIED;
    }

@@ -526,7 +539,7 @@ static int get_cluster_table(BlockDriverState *bs, uint64_t offset,
    *new_l2_offset = l2_offset;
    *new_l2_index = l2_index;

-    return 1;
+    return 0;
 }

 /*
@@ -548,12 +561,14 @@ uint64_t qcow2_alloc_compressed_cluster_offset(BlockDriverState *bs,
 {
    BDRVQcowState *s = bs->opaque;
    int l2_index, ret;
-    uint64_t l2_offset, *l2_table, cluster_offset;
+    uint64_t l2_offset, *l2_table;
+    int64_t cluster_offset;
    int nb_csectors;

    ret = get_cluster_table(bs, offset, &l2_table, &l2_offset, &l2_index);
-    if (ret == 0)
+    if (ret < 0) {
        return 0;
+    }

    cluster_offset = be64_to_cpu(l2_table[l2_index]);
    if (cluster_offset & QCOW_OFLAG_COPIED)
@@ -563,6 +578,10 @@ uint64_t qcow2_alloc_compressed_cluster_offset(BlockDriverState *bs,
        qcow2_free_any_clusters(bs, cluster_offset, 1);

    cluster_offset = qcow2_alloc_bytes(bs, compressed_size);
+    if (cluster_offset < 0) {
+        return 0;
+    }
+
    nb_csectors = ((cluster_offset + compressed_size - 1) >> 9) -
                  (cluster_offset >> 9);

@@ -605,12 +624,12 @@ static int write_l2_entries(BDRVQcowState *s, uint64_t *l2_table,
    return 0;
 }

-int qcow2_alloc_cluster_link_l2(BlockDriverState *bs, uint64_t cluster_offset,
-    QCowL2Meta *m)
+int qcow2_alloc_cluster_link_l2(BlockDriverState *bs, QCowL2Meta *m)
 {
    BDRVQcowState *s = bs->opaque;
    int i, j = 0, l2_index, ret;
    uint64_t *old_cluster, start_sect, l2_offset, *l2_table;
+    uint64_t cluster_offset = m->cluster_offset;

    if (m->nb_clusters == 0)
        return 0;
@@ -633,10 +652,11 @@ int qcow2_alloc_cluster_link_l2(BlockDriverState *bs, uint64_t cluster_offset,
            goto err;
    }

-    ret = -EIO;
    /* update L2 table */
-    if (!get_cluster_table(bs, m->offset, &l2_table, &l2_offset, &l2_index))
+    ret = get_cluster_table(bs, m->offset, &l2_table, &l2_offset, &l2_index);
+    if (ret < 0) {
        goto err;
+    }

    for (i = 0; i < m->nb_clusters; i++) {
        /* if two concurrent writes happen to the same unallocated cluster
@@ -670,30 +690,36 @@ err:
 /*
 * alloc_cluster_offset
 *
- * For a given offset of the disk image, return cluster offset in
- * qcow2 file.
- *
+ * For a given offset of the disk image, return cluster offset in qcow2 file.
 * If the offset is not found, allocate a new cluster.
 *
- * Return the cluster offset if successful,
- * Return 0, otherwise.
+ * If the cluster was already allocated, m->nb_clusters is set to 0,
+ * m->depends_on is set to NULL and the other fields in m are meaningless.
 *
+ * If the cluster is newly allocated, m->nb_clusters is set to the number of
+ * contiguous clusters that have been allocated. This may be 0 if the request
+ * conflict with another write request in flight; in this case, m->depends_on
+ * is set and the remaining fields of m are meaningless.
+ *
+ * If m->nb_clusters is non-zero, the other fields of m are valid and contain
+ * information about the first allocated cluster.
+ *
+ * Return 0 on success and -errno in error cases
 */
-
-uint64_t qcow2_alloc_cluster_offset(BlockDriverState *bs,
-                                    uint64_t offset,
-                                    int n_start, int n_end,
-                                    int *num, QCowL2Meta *m)
+int qcow2_alloc_cluster_offset(BlockDriverState *bs, uint64_t offset,
+    int n_start, int n_end, int *num, QCowL2Meta *m)
 {
    BDRVQcowState *s = bs->opaque;
    int l2_index, ret;
-    uint64_t l2_offset, *l2_table, cluster_offset;
+    uint64_t l2_offset, *l2_table;
+    int64_t cluster_offset;
    unsigned int nb_clusters, i = 0;
    QCowL2Meta *old_alloc;

    ret = get_cluster_table(bs, offset, &l2_table, &l2_offset, &l2_index);
-    if (ret == 0)
-        return 0;
+    if (ret < 0) {
+        return ret;
+    }

    nb_clusters = size_to_clusters(s, n_end << 9);

@@ -709,6 +735,7 @@ uint64_t qcow2_alloc_cluster_offset(BlockDriverState *bs,

        cluster_offset &= ~QCOW_OFLAG_COPIED;
        m->nb_clusters = 0;
+        m->depends_on = NULL;

        goto out;
    }
@@ -723,12 +750,15 @@ uint64_t qcow2_alloc_cluster_offset(BlockDriverState *bs,
    while (i < nb_clusters) {
        i += count_contiguous_clusters(nb_clusters - i, s->cluster_size,
                &l2_table[l2_index], i, 0);
-
-        if(be64_to_cpu(l2_table[l2_index + i]))
+        if ((i >= nb_clusters) || be64_to_cpu(l2_table[l2_index + i])) {
            break;
+        }

        i += count_contiguous_free_clusters(nb_clusters - i,
                &l2_table[l2_index + i]);
+        if (i >= nb_clusters) {
+            break;
+        }

        cluster_offset = be64_to_cpu(l2_table[l2_index + i]);

@@ -736,6 +766,7 @@ uint64_t qcow2_alloc_cluster_offset(BlockDriverState *bs,
                (cluster_offset & QCOW_OFLAG_COMPRESSED))
            break;
    }
+    assert(i <= nb_clusters);
    nb_clusters = i;

    /*
@@ -779,6 +810,10 @@ uint64_t qcow2_alloc_cluster_offset(BlockDriverState *bs,
    /* allocate a new cluster */

    cluster_offset = qcow2_alloc_clusters(bs, nb_clusters * s->cluster_size);
+    if (cluster_offset < 0) {
+        QLIST_REMOVE(m, next_in_flight);
+        return cluster_offset;
+    }

    /* save info needed for meta data update */
    m->offset = offset;
@@ -787,10 +822,11 @@ uint64_t qcow2_alloc_cluster_offset(BlockDriverState *bs,

 out:
    m->nb_available = MIN(nb_clusters << (s->cluster_bits - 9), n_end);
+    m->cluster_offset = cluster_offset;

    *num = m->nb_available - n_start;

-    return cluster_offset;
+    return 0;
 }

 static int decompress_buffer(uint8_t *out_buf, int out_buf_size,
@@ -27,7 +27,7 @@
 #include "block/qcow2.h"

 static int64_t alloc_clusters_noref(BlockDriverState *bs, int64_t size);
-static int update_refcount(BlockDriverState *bs,
+static int QEMU_WARN_UNUSED_RESULT update_refcount(BlockDriverState *bs,
                            int64_t offset, int64_t length,
                            int addend);

@@ -123,121 +123,266 @@ static int get_refcount(BlockDriverState *bs, int64_t cluster_index)
    return be16_to_cpu(s->refcount_block_cache[block_index]);
 }

-static int grow_refcount_table(BlockDriverState *bs, int min_size)
+/*
+ * Rounds the refcount table size up to avoid growing the table for each single
+ * refcount block that is allocated.
+ */
+static unsigned int next_refcount_table_size(BDRVQcowState *s,
+    unsigned int min_size)
 {
-    BDRVQcowState *s = bs->opaque;
-    int new_table_size, new_table_size2, refcount_table_clusters, i, ret;
-    uint64_t *new_table;
-    int64_t table_offset;
-    uint8_t data[12];
-    int old_table_size;
-    int64_t old_table_offset;
+    unsigned int min_clusters = (min_size >> (s->cluster_bits - 3)) + 1;
+    unsigned int refcount_table_clusters =
+        MAX(1, s->refcount_table_size >> (s->cluster_bits - 3));

-    if (min_size <= s->refcount_table_size)
-        return 0;
-    /* compute new table size */
-    refcount_table_clusters = s->refcount_table_size >> (s->cluster_bits - 3);
-    for(;;) {
-        if (refcount_table_clusters == 0) {
-            refcount_table_clusters = 1;
-        } else {
-            refcount_table_clusters = (refcount_table_clusters * 3 + 1) / 2;
-        }
-        new_table_size = refcount_table_clusters << (s->cluster_bits - 3);
-        if (min_size <= new_table_size)
-            break;
+    while (min_clusters > refcount_table_clusters) {
+        refcount_table_clusters = (refcount_table_clusters * 3 + 1) / 2;
    }
-#ifdef DEBUG_ALLOC2
-    printf("grow_refcount_table from %d to %d\n",
-           s->refcount_table_size,
-           new_table_size);
-#endif
-    new_table_size2 = new_table_size * sizeof(uint64_t);
-    new_table = qemu_mallocz(new_table_size2);
-    memcpy(new_table, s->refcount_table,
-           s->refcount_table_size * sizeof(uint64_t));
-    for(i = 0; i < s->refcount_table_size; i++)
-        cpu_to_be64s(&new_table[i]);
-    /* Note: we cannot update the refcount now to avoid recursion */
-    table_offset = alloc_clusters_noref(bs, new_table_size2);
-    ret = bdrv_pwrite(s->hd, table_offset, new_table, new_table_size2);
-    if (ret != new_table_size2)
-        goto fail;
-    for(i = 0; i < s->refcount_table_size; i++)
-        be64_to_cpus(&new_table[i]);

-    cpu_to_be64w((uint64_t*)data, table_offset);
-    cpu_to_be32w((uint32_t*)(data + 8), refcount_table_clusters);
-    if (bdrv_pwrite(s->hd, offsetof(QCowHeader, refcount_table_offset),
-                    data, sizeof(data)) != sizeof(data))
-        goto fail;
-    qemu_free(s->refcount_table);
-    old_table_offset = s->refcount_table_offset;
-    old_table_size = s->refcount_table_size;
-    s->refcount_table = new_table;
-    s->refcount_table_size = new_table_size;
-    s->refcount_table_offset = table_offset;
-
-    update_refcount(bs, table_offset, new_table_size2, 1);
-    qcow2_free_clusters(bs, old_table_offset, old_table_size * sizeof(uint64_t));
-    return 0;
- fail:
-    qemu_free(new_table);
-    return -EIO;
+    return refcount_table_clusters << (s->cluster_bits - 3);
 }


+/* Checks if two offsets are described by the same refcount block */
+static int in_same_refcount_block(BDRVQcowState *s, uint64_t offset_a,
+    uint64_t offset_b)
+{
+    uint64_t block_a = offset_a >> (2 * s->cluster_bits - REFCOUNT_SHIFT);
+    uint64_t block_b = offset_b >> (2 * s->cluster_bits - REFCOUNT_SHIFT);
+
+    return (block_a == block_b);
+}
+
+/*
+ * Loads a refcount block. If it doesn't exist yet, it is allocated first
+ * (including growing the refcount table if needed).
+ *
+ * Returns the offset of the refcount block on success or -errno in error case
+ */
 static int64_t alloc_refcount_block(BlockDriverState *bs, int64_t cluster_index)
 {
    BDRVQcowState *s = bs->opaque;
-    int64_t offset, refcount_block_offset;
    unsigned int refcount_table_index;
    int ret;
-    uint64_t data64;
-    int cache = cache_refcount_updates;

-    /* Find L1 index and grow refcount table if needed */
+    /* Find the refcount block for the given cluster */
    refcount_table_index = cluster_index >> (s->cluster_bits - REFCOUNT_SHIFT);
-    if (refcount_table_index >= s->refcount_table_size) {
-        ret = grow_refcount_table(bs, refcount_table_index + 1);
-        if (ret < 0)
+
+    if (refcount_table_index < s->refcount_table_size) {
+
+        uint64_t refcount_block_offset =
+            s->refcount_table[refcount_table_index];
+
+        /* If it's already there, we're done */
+        if (refcount_block_offset) {
+            if (refcount_block_offset != s->refcount_block_cache_offset) {
+                ret = load_refcount_block(bs, refcount_block_offset);
+                if (ret < 0) {
+                    return ret;
+                }
+            }
+            return refcount_block_offset;
+        }
+    }
+
+    /*
+     * If we came here, we need to allocate something. Something is at least
+     * a cluster for the new refcount block. It may also include a new refcount
+     * table if the old refcount table is too small.
+     *
+     * Note that allocating clusters here needs some special care:
+     *
+     * - We can't use the normal qcow2_alloc_clusters(), it would try to
+     *   increase the refcount and very likely we would end up with an endless
+     *   recursion. Instead we must place the refcount blocks in a way that
+     *   they can describe them themselves.
+     *
+     * - We need to consider that at this point we are inside update_refcounts
+     *   and doing the initial refcount increase. This means that some clusters
+     *   have already been allocated by the caller, but their refcount isn't
+     *   accurate yet. free_cluster_index tells us where this allocation ends
+     *   as long as we don't overwrite it by freeing clusters.
+     *
+     * - alloc_clusters_noref and qcow2_free_clusters may load a different
+     *   refcount block into the cache
+     */
+
+    if (cache_refcount_updates) {
+        ret = write_refcount_block(s);
+        if (ret < 0) {
            return ret;
+        }
    }

-    /* Load or allocate the refcount block */
-    refcount_block_offset = s->refcount_table[refcount_table_index];
-    if (!refcount_block_offset) {
-        if (cache_refcount_updates) {
-            write_refcount_block(s);
-            cache_refcount_updates = 0;
-        }
-        /* create a new refcount block */
-        /* Note: we cannot update the refcount now to avoid recursion */
-        offset = alloc_clusters_noref(bs, s->cluster_size);
-        memset(s->refcount_block_cache, 0, s->cluster_size);
-        ret = bdrv_pwrite(s->hd, offset, s->refcount_block_cache, s->cluster_size);
-        if (ret != s->cluster_size)
-            return -EINVAL;
-        s->refcount_table[refcount_table_index] = offset;
-        data64 = cpu_to_be64(offset);
-        ret = bdrv_pwrite(s->hd, s->refcount_table_offset +
-                          refcount_table_index * sizeof(uint64_t),
-                          &data64, sizeof(data64));
-        if (ret != sizeof(data64))
-            return -EINVAL;
+    /* Allocate the refcount block itself and mark it as used */
+    uint64_t new_block = alloc_clusters_noref(bs, s->cluster_size);
+    memset(s->refcount_block_cache, 0, s->cluster_size);
+    s->refcount_block_cache_offset = new_block;

-        refcount_block_offset = offset;
-        s->refcount_block_cache_offset = offset;
-        update_refcount(bs, offset, s->cluster_size, 1);
-        cache_refcount_updates = cache;
+#ifdef DEBUG_ALLOC2
+    fprintf(stderr, "qcow2: Allocate refcount block %d for %" PRIx64
+        " at %" PRIx64 "\n",
+        refcount_table_index, cluster_index << s->cluster_bits, new_block);
+#endif
+
+    if (in_same_refcount_block(s, new_block, cluster_index << s->cluster_bits)) {
+        /* The block describes itself, need to update the cache */
+        int block_index = (new_block >> s->cluster_bits) &
+            ((1 << (s->cluster_bits - REFCOUNT_SHIFT)) - 1);
+        s->refcount_block_cache[block_index] = cpu_to_be16(1);
    } else {
-        if (refcount_block_offset != s->refcount_block_cache_offset) {
-            if (load_refcount_block(bs, refcount_block_offset) < 0)
-                return -EIO;
+        /* Described somewhere else. This can recurse at most twice before we
+         * arrive at a block that describes itself. */
+        ret = update_refcount(bs, new_block, s->cluster_size, 1);
+        if (ret < 0) {
+            goto fail_block;
        }
    }

-    return refcount_block_offset;
+    /* Now the new refcount block needs to be written to disk */
+    ret = bdrv_pwrite(s->hd, new_block, s->refcount_block_cache,
+        s->cluster_size);
+    if (ret < 0) {
+        goto fail_block;
+    }
+
+    /* If the refcount table is big enough, just hook the block up there */
+    if (refcount_table_index < s->refcount_table_size) {
+        uint64_t data64 = cpu_to_be64(new_block);
+        ret = bdrv_pwrite(s->hd,
+            s->refcount_table_offset + refcount_table_index * sizeof(uint64_t),
+            &data64, sizeof(data64));
+        if (ret < 0) {
+            goto fail_block;
+        }
+
+        s->refcount_table[refcount_table_index] = new_block;
+        return new_block;
+    }
+
+    /*
+     * If we come here, we need to grow the refcount table. Again, a new
+     * refcount table needs some space and we can't simply allocate to avoid
+     * endless recursion.
+     *
+     * Therefore let's grab new refcount blocks at the end of the image, which
+     * will describe themselves and the new refcount table. This way we can
+     * reference them only in the new table and do the switch to the new
+     * refcount table at once without producing an inconsistent state in
+     * between.
+     */
+    /* Calculate the number of refcount blocks needed so far */
+    uint64_t refcount_block_clusters = 1 << (s->cluster_bits - REFCOUNT_SHIFT);
+    uint64_t blocks_used = (s->free_cluster_index +
+        refcount_block_clusters - 1) / refcount_block_clusters;
+
+    /* And now we need at least one block more for the new metadata */
+    uint64_t table_size = next_refcount_table_size(s, blocks_used + 1);
+    uint64_t last_table_size;
+    uint64_t blocks_clusters;
+    do {
+        uint64_t table_clusters = size_to_clusters(s, table_size);
+        blocks_clusters = 1 +
+            ((table_clusters + refcount_block_clusters - 1)
+            / refcount_block_clusters);
+        uint64_t meta_clusters = table_clusters + blocks_clusters;
+
+        last_table_size = table_size;
+        table_size = next_refcount_table_size(s, blocks_used +
+            ((meta_clusters + refcount_block_clusters - 1)
+            / refcount_block_clusters));
+
+    } while (last_table_size != table_size);
+
+#ifdef DEBUG_ALLOC2
+    fprintf(stderr, "qcow2: Grow refcount table %" PRId32 " => %" PRId64 "\n",
+        s->refcount_table_size, table_size);
+#endif
+
+    /* Create the new refcount table and blocks */
+    uint64_t meta_offset = (blocks_used * refcount_block_clusters) *
+        s->cluster_size;
+    uint64_t table_offset = meta_offset + blocks_clusters * s->cluster_size;
+    uint16_t *new_blocks = qemu_mallocz(blocks_clusters * s->cluster_size);
+    uint64_t *new_table = qemu_mallocz(table_size * sizeof(uint64_t));
+
+    assert(meta_offset >= (s->free_cluster_index * s->cluster_size));
+
+    /* Fill the new refcount table */
+    memcpy(new_table, s->refcount_table,
+        s->refcount_table_size * sizeof(uint64_t));
+    new_table[refcount_table_index] = new_block;
+
+    int i;
+    for (i = 0; i < blocks_clusters; i++) {
+        new_table[blocks_used + i] = meta_offset + (i * s->cluster_size);
+    }
+
+    /* Fill the refcount blocks */
+    uint64_t table_clusters = size_to_clusters(s, table_size * sizeof(uint64_t));
+    int block = 0;
+    for (i = 0; i < table_clusters + blocks_clusters; i++) {
+        new_blocks[block++] = cpu_to_be16(1);
+    }
+
+    /* Write refcount blocks to disk */
+    ret = bdrv_pwrite(s->hd, meta_offset, new_blocks,
+        blocks_clusters * s->cluster_size);
+    qemu_free(new_blocks);
+    if (ret < 0) {
+        goto fail_table;
+    }
+
+    /* Write refcount table to disk */
+    for(i = 0; i < table_size; i++) {
+        cpu_to_be64s(&new_table[i]);
+    }
+
+    ret = bdrv_pwrite(s->hd, table_offset, new_table,
+        table_size * sizeof(uint64_t));
+    if (ret < 0) {
+        goto fail_table;
+    }
+
+    for(i = 0; i < table_size; i++) {
+        cpu_to_be64s(&new_table[i]);
+    }
+
+    /* Hook up the new refcount table in the qcow2 header */
+    uint8_t data[12];
+    cpu_to_be64w((uint64_t*)data, table_offset);
+    cpu_to_be32w((uint32_t*)(data + 8), table_clusters);
+    ret = bdrv_pwrite(s->hd, offsetof(QCowHeader, refcount_table_offset),
+        data, sizeof(data));
+    if (ret < 0) {
+        goto fail_table;
+    }
+
+    /* And switch it in memory */
+    uint64_t old_table_offset = s->refcount_table_offset;
+    uint64_t old_table_size = s->refcount_table_size;
+
+    qemu_free(s->refcount_table);
+    s->refcount_table = new_table;
+    s->refcount_table_size = table_size;
+    s->refcount_table_offset = table_offset;
+
+    /* Free old table. Remember, we must not change free_cluster_index */
+    uint64_t old_free_cluster_index = s->free_cluster_index;
+    qcow2_free_clusters(bs, old_table_offset, old_table_size * sizeof(uint64_t));
+    s->free_cluster_index = old_free_cluster_index;
+
+    ret = load_refcount_block(bs, new_block);
+    if (ret < 0) {
+        goto fail_block;
+    }
+
+    return new_block;
+
+fail_table:
+    qemu_free(new_table);
+fail_block:
+    s->refcount_block_cache_offset = 0;
+    return ret;
 }

 #define REFCOUNTS_PER_SECTOR (512 >> REFCOUNT_SHIFT)
@@ -266,22 +411,26 @@ static int write_refcount_block_entries(BDRVQcowState *s,
 }

 /* XXX: cache several refcount block clusters ? */
-static int update_refcount(BlockDriverState *bs,
-                            int64_t offset, int64_t length,
-                            int addend)
+static int QEMU_WARN_UNUSED_RESULT update_refcount(BlockDriverState *bs,
+    int64_t offset, int64_t length, int addend)
 {
    BDRVQcowState *s = bs->opaque;
    int64_t start, last, cluster_offset;
    int64_t refcount_block_offset = 0;
    int64_t table_index = -1, old_table_index;
    int first_index = -1, last_index = -1;
+    int ret;

 #ifdef DEBUG_ALLOC2
    printf("update_refcount: offset=%" PRId64 " size=%" PRId64 " addend=%d\n",
           offset, length, addend);
 #endif
-    if (length <= 0)
+    if (length < 0) {
        return -EINVAL;
+    } else if (length == 0) {
+        return 0;
+    }
+
    start = offset & ~(s->cluster_size - 1);
    last = (offset + length - 1) & ~(s->cluster_size - 1);
    for(cluster_offset = start; cluster_offset <= last;
@@ -289,6 +438,7 @@ static int update_refcount(BlockDriverState *bs,
    {
        int block_index, refcount;
        int64_t cluster_index = cluster_offset >> s->cluster_bits;
+        int64_t new_block;

        /* Only write refcount block to disk when we are done with it */
        old_table_index = table_index;
@@ -306,10 +456,12 @@ static int update_refcount(BlockDriverState *bs,
        }

        /* Load the refcount block and allocate it if needed */
-        refcount_block_offset = alloc_refcount_block(bs, cluster_index);
-        if (refcount_block_offset < 0) {
-            return refcount_block_offset;
+        new_block = alloc_refcount_block(bs, cluster_index);
+        if (new_block < 0) {
+            ret = new_block;
+            goto fail;
        }
+        refcount_block_offset = new_block;

        /* we can update the count and save it */
        block_index = cluster_index &
@@ -323,24 +475,38 @@ static int update_refcount(BlockDriverState *bs,

        refcount = be16_to_cpu(s->refcount_block_cache[block_index]);
        refcount += addend;
-        if (refcount < 0 || refcount > 0xffff)
-            return -EINVAL;
+        if (refcount < 0 || refcount > 0xffff) {
+            ret = -EINVAL;
+            goto fail;
+        }
        if (refcount == 0 && cluster_index < s->free_cluster_index) {
            s->free_cluster_index = cluster_index;
        }
        s->refcount_block_cache[block_index] = cpu_to_be16(refcount);
    }

+    ret = 0;
+fail:
+
    /* Write last changed block to disk */
    if (refcount_block_offset != 0) {
        if (write_refcount_block_entries(s, refcount_block_offset,
            first_index, last_index) < 0)
        {
-            return -EIO;
+            return ret < 0 ? ret : -EIO;
        }
    }

-    return 0;
+    /*
+     * Try do undo any updates if an error is returned (This may succeed in
+     * some cases like ENOSPC for allocating a new refcount block)
+     */
+    if (ret < 0) {
+        int dummy;
+        dummy = update_refcount(bs, offset, cluster_offset - offset, -addend);
+    }
+
+    return ret;
 }

 /* addend must be 1 or -1 */
@@ -390,9 +556,13 @@ retry:
 int64_t qcow2_alloc_clusters(BlockDriverState *bs, int64_t size)
 {
    int64_t offset;
+    int ret;

    offset = alloc_clusters_noref(bs, size);
-    update_refcount(bs, offset, size, 1);
+    ret = update_refcount(bs, offset, size, 1);
+    if (ret < 0) {
+        return ret;
+    }
    return offset;
 }

@@ -407,6 +577,9 @@ int64_t qcow2_alloc_bytes(BlockDriverState *bs, int size)
    assert(size > 0 && size <= s->cluster_size);
    if (s->free_byte_offset == 0) {
        s->free_byte_offset = qcow2_alloc_clusters(bs, s->cluster_size);
+        if (s->free_byte_offset < 0) {
+            return s->free_byte_offset;
+        }
    }
 redo:
    free_in_cluster = s->cluster_size -
@@ -422,6 +595,9 @@ int64_t qcow2_alloc_bytes(BlockDriverState *bs, int size)
            update_cluster_refcount(bs, offset >> s->cluster_bits, 1);
    } else {
        offset = qcow2_alloc_clusters(bs, s->cluster_size);
+        if (offset < 0) {
+            return offset;
+        }
        cluster_offset = s->free_byte_offset & ~(s->cluster_size - 1);
        if ((cluster_offset + s->cluster_size) == offset) {
            /* we are lucky: contiguous data */
@@ -439,7 +615,13 @@ int64_t qcow2_alloc_bytes(BlockDriverState *bs, int size)
 void qcow2_free_clusters(BlockDriverState *bs,
                          int64_t offset, int64_t size)
 {
-    update_refcount(bs, offset, size, -1);
+    int ret;
+
+    ret = update_refcount(bs, offset, size, -1);
+    if (ret < 0) {
+        fprintf(stderr, "qcow2_free_clusters failed: %s\n", strerror(-ret));
+        abort();
+    }
 }

 /*
@@ -549,9 +731,15 @@ int qcow2_update_snapshot_refcount(BlockDriverState *bs,
                    if (offset & QCOW_OFLAG_COMPRESSED) {
                        nb_csectors = ((offset >> s->csize_shift) &
                                       s->csize_mask) + 1;
-                        if (addend != 0)
-                            update_refcount(bs, (offset & s->cluster_offset_mask) & ~511,
-                                            nb_csectors * 512, addend);
+                        if (addend != 0) {
+                            int ret;
+                            ret = update_refcount(bs,
+                                (offset & s->cluster_offset_mask) & ~511,
+                                nb_csectors * 512, addend);
+                            if (ret < 0) {
+                                goto fail;
+                            }
+                        }
                        /* compressed clusters are never modified */
                        refcount = 2;
                    } else {
@@ -139,6 +139,9 @@ static int qcow_write_snapshots(BlockDriverState *bs)

    snapshots_offset = qcow2_alloc_clusters(bs, snapshots_size);
    offset = snapshots_offset;
+    if (offset < 0) {
+        return offset;
+    }

    for(i = 0; i < s->nb_snapshots; i++) {
        sn = s->snapshots + i;
@@ -235,6 +238,7 @@ int qcow2_snapshot_create(BlockDriverState *bs, QEMUSnapshotInfo *sn_info)
    QCowSnapshot *snapshots1, sn1, *sn = &sn1;
    int i, ret;
    uint64_t *l1_table = NULL;
+    int64_t l1_table_offset;

    memset(sn, 0, sizeof(*sn));

@@ -263,7 +267,12 @@ int qcow2_snapshot_create(BlockDriverState *bs, QEMUSnapshotInfo *sn_info)
        goto fail;

    /* create the L1 table of the snapshot */
-    sn->l1_table_offset = qcow2_alloc_clusters(bs, s->l1_size * sizeof(uint64_t));
+    l1_table_offset = qcow2_alloc_clusters(bs, s->l1_size * sizeof(uint64_t));
+    if (l1_table_offset < 0) {
+        goto fail;
+    }
+
+    sn->l1_table_offset = l1_table_offset;
    sn->l1_size = s->l1_size;

    if (s->l1_size != 0) {
@@ -467,8 +467,10 @@ static void qcow_aio_read_cb(void *opaque, int ret)
        acb->hd_aiocb = bdrv_aio_readv(s->hd,
                            (acb->cluster_offset >> 9) + index_in_cluster,
                            &acb->hd_qiov, acb->n, qcow_aio_read_cb, acb);
-        if (acb->hd_aiocb == NULL)
+        if (acb->hd_aiocb == NULL) {
+            ret = -EIO;
            goto done;
+        }
    }

    return;
@@ -561,7 +563,7 @@ static void qcow_aio_write_cb(void *opaque, int ret)
    acb->hd_aiocb = NULL;

    if (ret >= 0) {
-        ret = qcow2_alloc_cluster_link_l2(bs, acb->cluster_offset, &acb->l2meta);
+        ret = qcow2_alloc_cluster_link_l2(bs, &acb->l2meta);
    }

    run_dependent_requests(&acb->l2meta);
@@ -585,21 +587,23 @@ static void qcow_aio_write_cb(void *opaque, int ret)
        n_end > QCOW_MAX_CRYPT_CLUSTERS * s->cluster_sectors)
        n_end = QCOW_MAX_CRYPT_CLUSTERS * s->cluster_sectors;

-    acb->cluster_offset = qcow2_alloc_cluster_offset(bs, acb->sector_num << 9,
-                                          index_in_cluster,
-                                          n_end, &acb->n, &acb->l2meta);
+    ret = qcow2_alloc_cluster_offset(bs, acb->sector_num << 9,
+        index_in_cluster, n_end, &acb->n, &acb->l2meta);
+    if (ret < 0) {
+        goto done;
+    }
+
+    acb->cluster_offset = acb->l2meta.cluster_offset;

    /* Need to wait for another request? If so, we are done for now. */
-    if (!acb->cluster_offset && acb->l2meta.depends_on != NULL) {
+    if (acb->l2meta.nb_clusters == 0 && acb->l2meta.depends_on != NULL) {
        QLIST_INSERT_HEAD(&acb->l2meta.depends_on->dependent_requests,
            acb, next_depend);
        return;
    }

-    if (!acb->cluster_offset || (acb->cluster_offset & 511) != 0) {
-        ret = -EIO;
-        goto done;
-    }
+    assert((acb->cluster_offset & 511) == 0);
+
    if (s->crypt_method) {
        if (!acb->cluster_data) {
            acb->cluster_data = qemu_mallocz(QCOW_MAX_CRYPT_CLUSTERS *
@@ -618,11 +622,17 @@ static void qcow_aio_write_cb(void *opaque, int ret)
                                    (acb->cluster_offset >> 9) + index_in_cluster,
                                    &acb->hd_qiov, acb->n,
                                    qcow_aio_write_cb, acb);
-    if (acb->hd_aiocb == NULL)
-        goto done;
+    if (acb->hd_aiocb == NULL) {
+        ret = -EIO;
+        goto fail;
+    }

    return;

+fail:
+    if (acb->l2meta.nb_clusters != 0) {
+        QLIST_REMOVE(&acb->l2meta, next_in_flight);
+    }
 done:
    if (acb->qiov->niov > 1)
        qemu_vfree(acb->orig_buf);
@@ -683,27 +693,27 @@ static int get_bits_from_size(size_t size)
 static int preallocate(BlockDriverState *bs)
 {
    BDRVQcowState *s = bs->opaque;
-    uint64_t cluster_offset = 0;
    uint64_t nb_sectors;
    uint64_t offset;
    int num;
+    int ret;
    QCowL2Meta meta;

    nb_sectors = bdrv_getlength(bs) >> 9;
    offset = 0;
    QLIST_INIT(&meta.dependent_requests);
+    meta.cluster_offset = 0;

    while (nb_sectors) {
        num = MIN(nb_sectors, INT_MAX >> 9);
-        cluster_offset = qcow2_alloc_cluster_offset(bs, offset, 0, num, &num,
-            &meta);
+        ret = qcow2_alloc_cluster_offset(bs, offset, 0, num, &num, &meta);

-        if (cluster_offset == 0) {
+        if (ret < 0) {
            return -1;
        }

-        if (qcow2_alloc_cluster_link_l2(bs, cluster_offset, &meta) < 0) {
-            qcow2_free_any_clusters(bs, cluster_offset, meta.nb_clusters);
+        if (qcow2_alloc_cluster_link_l2(bs, &meta) < 0) {
+            qcow2_free_any_clusters(bs, meta.cluster_offset, meta.nb_clusters);
            return -1;
        }

@@ -722,10 +732,10 @@ static int preallocate(BlockDriverState *bs)
     * all of the allocated clusters (otherwise we get failing reads after
     * EOF). Extend the image to the last allocated sector.
     */
-    if (cluster_offset != 0) {
+    if (meta.cluster_offset != 0) {
        uint8_t buf[512];
        memset(buf, 0, 512);
-        bdrv_write(s->hd, (cluster_offset >> 9) + num - 1, buf, 1);
+        bdrv_write(s->hd, (meta.cluster_offset >> 9) + num - 1, buf, 1);
    }

    return 0;
@@ -1056,12 +1066,13 @@ static int qcow_save_vmstate(BlockDriverState *bs, const uint8_t *buf,
 {
    BDRVQcowState *s = bs->opaque;
    int growable = bs->growable;
+    int ret;

    bs->growable = 1;
-    bdrv_pwrite(bs, qcow_vm_state_offset(s) + pos, buf, size);
+    ret = bdrv_pwrite(bs, qcow_vm_state_offset(s) + pos, buf, size);
    bs->growable = growable;

-    return size;
+    return ret;
 }

 static int qcow_load_vmstate(BlockDriverState *bs, uint8_t *buf,
@@ -135,6 +135,7 @@ struct QCowAIOCB;
 typedef struct QCowL2Meta
 {
    uint64_t offset;
+    uint64_t cluster_offset;
    int n_start;
    int nb_available;
    int nb_clusters;
@@ -191,16 +192,13 @@ void qcow2_encrypt_sectors(BDRVQcowState *s, int64_t sector_num,

 uint64_t qcow2_get_cluster_offset(BlockDriverState *bs, uint64_t offset,
    int *num);
-uint64_t qcow2_alloc_cluster_offset(BlockDriverState *bs,
-                              uint64_t offset,
-                              int n_start, int n_end,
-                              int *num, QCowL2Meta *m);
+int qcow2_alloc_cluster_offset(BlockDriverState *bs, uint64_t offset,
+    int n_start, int n_end, int *num, QCowL2Meta *m);
 uint64_t qcow2_alloc_compressed_cluster_offset(BlockDriverState *bs,
                                         uint64_t offset,
                                         int compressed_size);

-int qcow2_alloc_cluster_link_l2(BlockDriverState *bs, uint64_t cluster_offset,
-    QCowL2Meta *m);
+int qcow2_alloc_cluster_link_l2(BlockDriverState *bs, QCowL2Meta *m);

 /* qcow2-snapshot.c functions */
 int qcow2_snapshot_create(BlockDriverState *bs, QEMUSnapshotInfo *sn_info);
@@ -595,7 +595,7 @@ static void raw_close(BlockDriverState *bs)
        close(s->fd);
        s->fd = -1;
        if (s->aligned_buf != NULL)
-            qemu_free(s->aligned_buf);
+            qemu_vfree(s->aligned_buf);
    }
 }

@@ -882,7 +882,7 @@ static int init_directories(BDRVVVFATState* s,
    mapping->dir_index = 0;
    mapping->info.dir.parent_mapping_index = -1;
    mapping->first_mapping_index = -1;
-    mapping->path = strdup(dirname);
+    mapping->path = qemu_strdup(dirname);
    i = strlen(mapping->path);
    if (i > 0 && mapping->path[i - 1] == '/')
 	mapping->path[i - 1] = '\0';
@@ -1632,10 +1632,10 @@ static uint32_t get_cluster_count_for_direntry(BDRVVVFATState* s,

 	    /* rename */
 	    if (strcmp(basename, basename2))
-		schedule_rename(s, cluster_num, strdup(path));
+		schedule_rename(s, cluster_num, qemu_strdup(path));
 	} else if (is_file(direntry))
 	    /* new file */
-	    schedule_new_file(s, strdup(path), cluster_num);
+	    schedule_new_file(s, qemu_strdup(path), cluster_num);
 	else {
 	    assert(0);
 	    return 0;
@@ -1752,10 +1752,10 @@ static int check_directory_consistency(BDRVVVFATState *s,
 	mapping->mode &= ~MODE_DELETED;

 	if (strcmp(basename, basename2))
-	    schedule_rename(s, cluster_num, strdup(path));
+	    schedule_rename(s, cluster_num, qemu_strdup(path));
    } else
 	/* new directory */
-	schedule_mkdir(s, cluster_num, strdup(path));
+	schedule_mkdir(s, cluster_num, qemu_strdup(path));

    lfn_init(&lfn);
    do {
@@ -2652,3 +2652,6 @@ d=libuser
 mkdir -p $d
 rm -f $d/Makefile
 ln -s $source_path/Makefile.user $d/Makefile
+if test "$static" = "no" -a "$user_pie" = "yes" ; then
+  echo "QEMU_CFLAGS+=-fpie" > $d/config.mak
+fi
@@ -1017,7 +1017,8 @@ static inline int64_t cpu_get_real_ticks (void)
 #endif
 }

-#elif (defined(__mips_isa_rev) && __mips_isa_rev >= 2) || defined(__linux__)
+#elif defined(__mips__) && \
+      ((defined(__mips_isa_rev) && __mips_isa_rev >= 2) || defined(__linux__))
 /*
 * binutils wants to use rdhwr only on mips32r2
 * but as linux kernel emulate it, it's fine
@@ -232,11 +232,13 @@ int cpu_exec(CPUState *env1)

    env_to_regs();
 #if defined(TARGET_I386)
-    /* put eflags in CPU temporary format */
-    CC_SRC = env->eflags & (CC_O | CC_S | CC_Z | CC_A | CC_P | CC_C);
-    DF = 1 - (2 * ((env->eflags >> 10) & 1));
-    CC_OP = CC_OP_EFLAGS;
-    env->eflags &= ~(DF_MASK | CC_O | CC_S | CC_Z | CC_A | CC_P | CC_C);
+    if (!kvm_enabled()) {
+        /* put eflags in CPU temporary format */
+        CC_SRC = env->eflags & (CC_O | CC_S | CC_Z | CC_A | CC_P | CC_C);
+        DF = 1 - (2 * ((env->eflags >> 10) & 1));
+        CC_OP = CC_OP_EFLAGS;
+        env->eflags &= ~(DF_MASK | CC_O | CC_S | CC_Z | CC_A | CC_P | CC_C);
+    }
 #elif defined(TARGET_SPARC)
 #elif defined(TARGET_M68K)
    env->cc_op = CC_OP_FLAGS;
@@ -2,3 +2,4 @@

 CONFIG_USB_OHCI=y
 CONFIG_PTIMER=y
+CONFIG_ISA_MMIO=y
@@ -2,3 +2,4 @@

 CONFIG_USB_OHCI=y
 CONFIG_PTIMER=y
+CONFIG_ISA_MMIO=y
@@ -69,9 +69,9 @@ extern int printf(const char *, ...);
 #define AREG1 "r14"
 #define AREG2 "r15"
 #elif defined(__mips__)
-#define AREG0 "fp"
-#define AREG1 "s0"
-#define AREG2 "s1"
+#define AREG0 "s0"
+#define AREG1 "s1"
+#define AREG2 "fp"
 #elif defined(__sparc__)
 #ifdef CONFIG_SOLARIS
 #define AREG0 "g2"
@@ -3307,7 +3307,7 @@ void cpu_physical_memory_unmap(void *buffer, target_phys_addr_t len,
    if (is_write) {
        cpu_physical_memory_write(bounce.addr, bounce.buffer, access_len);
    }
-    qemu_free(bounce.buffer);
+    qemu_vfree(bounce.buffer);
    bounce.buffer = NULL;
    cpu_notify_map_clients();
 }
@@ -2585,9 +2585,11 @@ static void map_linear_vram(CirrusVGAState *s)

 static void unmap_linear_vram(CirrusVGAState *s)
 {
-    if (s->vga.map_addr && s->vga.lfb_addr && s->vga.lfb_end)
+    if (s->vga.map_addr && s->vga.lfb_addr && s->vga.lfb_end) {
        s->vga.map_addr = s->vga.map_end = 0;
-
+         cpu_register_physical_memory(s->vga.lfb_addr, s->vga.vram_size,
+                                      s->cirrus_linear_io_addr);
+    }
    cpu_register_physical_memory(isa_mem_base + 0xa0000, 0x20000,
                                 s->vga.vga_io_memory);
 }
@@ -935,6 +935,12 @@ static uint32_t fdctrl_read_main_status (fdctrl_t *fdctrl)
    fdctrl->dsr &= ~FD_DSR_PWRDOWN;
    fdctrl->dor |= FD_DOR_nRESET;

+    /* Sparc mutation */
+    if (fdctrl->sun4m) {
+        retval |= FD_MSR_DIO;
+        fdctrl_reset_irq(fdctrl);
+    };
+
    FLOPPY_DPRINTF("main status register: 0x%02x\n", retval);

    return retval;
@@ -1854,8 +1860,12 @@ fdctrl_t *fdctrl_init_isa(DriveInfo **fds)
    ISADevice *dev;

    dev = isa_create("isa-fdc");
-    qdev_prop_set_drive(&dev->qdev, "driveA", fds[0]);
-    qdev_prop_set_drive(&dev->qdev, "driveB", fds[1]);
+    if (fds[0]) {
+        qdev_prop_set_drive(&dev->qdev, "driveA", fds[0]);
+    }
+    if (fds[1]) {
+        qdev_prop_set_drive(&dev->qdev, "driveB", fds[1]);
+    }
    if (qdev_init(&dev->qdev) < 0)
        return NULL;
    return &(DO_UPCAST(fdctrl_isabus_t, busdev, dev)->state);
@@ -1873,8 +1883,12 @@ fdctrl_t *fdctrl_init_sysbus(qemu_irq irq, int dma_chann,
    sys = DO_UPCAST(fdctrl_sysbus_t, busdev.qdev, dev);
    fdctrl = &sys->state;
    fdctrl->dma_chann = dma_chann; /* FIXME */
-    qdev_prop_set_drive(dev, "driveA", fds[0]);
-    qdev_prop_set_drive(dev, "driveB", fds[1]);
+    if (fds[0]) {
+        qdev_prop_set_drive(dev, "driveA", fds[0]);
+    }
+    if (fds[1]) {
+        qdev_prop_set_drive(dev, "driveB", fds[1]);
+    }
    qdev_init_nofail(dev);
    sysbus_connect_irq(&sys->busdev, 0, irq);
    sysbus_mmio_map(&sys->busdev, 0, mmio_base);
@@ -1890,7 +1904,9 @@ fdctrl_t *sun4m_fdctrl_init (qemu_irq irq, target_phys_addr_t io_base,
    fdctrl_t *fdctrl;

    dev = qdev_create(NULL, "SUNW,fdtwo");
-    qdev_prop_set_drive(dev, "drive", fds[0]);
+    if (fds[0]) {
+        qdev_prop_set_drive(dev, "drive", fds[0]);
+    }
    qdev_init_nofail(dev);
    sys = DO_UPCAST(fdctrl_sysbus_t, busdev.qdev, dev);
    fdctrl = &sys->state;
@@ -179,7 +179,7 @@ static int get_uint32_as_uint16(QEMUFile *f, void *pv, size_t size)

 static void put_unused(QEMUFile *f, void *pv, size_t size)
 {
-    fprintf(stderr, "uint32_as_uint16 is only used for backward compatibilty.\n");
+    fprintf(stderr, "uint32_as_uint16 is only used for backward compatibility.\n");
    fprintf(stderr, "This functions shouldn't be called.\n");
 }

@@ -278,7 +278,7 @@ int fw_cfg_add_file(FWCfgState *s,  const char *dir, const char *filename,
                    uint8_t *data, uint32_t len)
 {
    const char *basename;
-    int index;
+    int i, index;

    if (!s->files) {
        int dsize = sizeof(uint32_t) + sizeof(FWCfgFile) * FW_CFG_FILE_SLOTS;
@@ -300,13 +300,17 @@ int fw_cfg_add_file(FWCfgState *s,  const char *dir, const char *filename,
    } else {
        basename = filename;
    }
-    if (dir) {
-        snprintf(s->files->f[index].name, sizeof(s->files->f[index].name),
-                 "%s/%s", dir, basename);
-    } else {
-        snprintf(s->files->f[index].name, sizeof(s->files->f[index].name),
-                 "%s", basename);
+
+    snprintf(s->files->f[index].name, sizeof(s->files->f[index].name),
+             "%s/%s", dir, basename);
+    for (i = 0; i < index; i++) {
+        if (strcmp(s->files->f[index].name, s->files->f[i].name) == 0) {
+            FW_CFG_DPRINTF("%s: skip duplicate: %s\n", __FUNCTION__,
+                           s->files->f[index].name);
+            return 1;
+        }
    }
+
    s->files->f[index].size   = cpu_to_be32(len);
    s->files->f[index].select = cpu_to_be16(FW_CFG_FILE_FIRST + index);
    FW_CFG_DPRINTF("%s: #%d: %s (%d bytes)\n", __FUNCTION__,
@@ -70,11 +70,7 @@ static void ide_map(PCIDevice *pci_dev, int region_num,

 static PCIIDEState *pci_from_bm(BMDMAState *bm)
 {
-    if (bm->unit == 0) {
-        return container_of(bm, PCIIDEState, bmdma[0]);
-    } else {
-        return container_of(bm, PCIIDEState, bmdma[1]);
-    }
+    return bm->pci_dev;
 }

 static uint32_t bmdma_readb(void *opaque, uint32_t addr)
@@ -145,6 +141,7 @@ static void bmdma_map(PCIDevice *pci_dev, int region_num,
        BMDMAState *bm = &d->bmdma[i];
        d->bus[i].bmdma = bm;
        bm->bus = d->bus+i;
+        bm->pci_dev = d;
        qemu_add_vm_change_state_handler(ide_dma_restart_cb, bm);

        register_ioport_write(addr, 1, 1, bmdma_cmd_writeb, bm);
@@ -115,7 +115,7 @@ static void ide_identify(IDEState *s)
    put_le16(p + 20, 3); /* XXX: retired, remove ? */
    put_le16(p + 21, 512); /* cache size in sectors */
    put_le16(p + 22, 4); /* ecc bytes */
-    padstr((char *)(p + 23), QEMU_VERSION, 8); /* firmware version */
+    padstr((char *)(p + 23), s->version, 8); /* firmware version */
    padstr((char *)(p + 27), "QEMU HARDDISK", 40); /* model */
 #if MAX_MULT_SECTORS > 1
    put_le16(p + 47, 0x8000 | MAX_MULT_SECTORS);
@@ -186,7 +186,7 @@ static void ide_atapi_identify(IDEState *s)
    put_le16(p + 20, 3); /* buffer type */
    put_le16(p + 21, 512); /* cache size in sectors */
    put_le16(p + 22, 4); /* ecc bytes */
-    padstr((char *)(p + 23), QEMU_VERSION, 8); /* firmware version */
+    padstr((char *)(p + 23), s->version, 8); /* firmware version */
    padstr((char *)(p + 27), "QEMU DVD-ROM", 40); /* model */
    put_le16(p + 48, 1); /* dword I/O (XXX: should not be set on CDROM) */
 #ifdef USE_DMA_CDROM
@@ -238,7 +238,7 @@ static void ide_cfata_identify(IDEState *s)
    put_le16(p + 8, s->nb_sectors);		/* Sectors per card */
    padstr((char *)(p + 10), s->drive_serial_str, 20); /* serial number */
    put_le16(p + 22, 0x0004);			/* ECC bytes */
-    padstr((char *) (p + 23), QEMU_VERSION, 8);	/* Firmware Revision */
+    padstr((char *) (p + 23), s->version, 8);	/* Firmware Revision */
    padstr((char *) (p + 27), "QEMU MICRODRIVE", 40);/* Model number */
 #if MAX_MULT_SECTORS > 1
    put_le16(p + 47, 0x8000 | MAX_MULT_SECTORS);
@@ -1591,7 +1591,7 @@ static void ide_atapi_cmd(IDEState *s)
        buf[7] = 0; /* reserved */
        padstr8(buf + 8, 8, "QEMU");
        padstr8(buf + 16, 16, "QEMU DVD-ROM");
-        padstr8(buf + 32, 4, QEMU_VERSION);
+        padstr8(buf + 32, 4, s->version);
        ide_atapi_cmd_reply(s, 36, max_len);
        break;
    case GPCMD_GET_CONFIGURATION:
@@ -2590,7 +2590,7 @@ void ide_bus_reset(IDEBus *bus)
    ide_clear_hob(bus);
 }

-void ide_init_drive(IDEState *s, DriveInfo *dinfo)
+void ide_init_drive(IDEState *s, DriveInfo *dinfo, const char *version)
 {
    int cylinders, heads, secs;
    uint64_t nb_sectors;
@@ -2619,6 +2619,11 @@ void ide_init_drive(IDEState *s, DriveInfo *dinfo)
    if (strlen(s->drive_serial_str) == 0)
        snprintf(s->drive_serial_str, sizeof(s->drive_serial_str),
                 "QM%05d", s->drive_serial);
+    if (version) {
+        pstrcpy(s->version, sizeof(s->version), version);
+    } else {
+        pstrcpy(s->version, sizeof(s->version), QEMU_VERSION);
+    }
    ide_reset(s);
 }

@@ -2635,13 +2640,14 @@ void ide_init2(IDEBus *bus, DriveInfo *hd0, DriveInfo *hd1,
        s->unit = i;
        s->drive_serial = drive_serial++;
        s->io_buffer = qemu_blockalign(s->bs, IDE_DMA_BUF_SECTORS*512 + 4);
+        s->io_buffer_total_len = IDE_DMA_BUF_SECTORS*512 + 4;
        s->smart_selftest_data = qemu_blockalign(s->bs, 512);
        s->sector_write_timer = qemu_new_timer(vm_clock,
                                               ide_sector_write_timer_cb, s);
        if (i == 0)
-            ide_init_drive(s, hd0);
+            ide_init_drive(s, hd0, NULL);
        if (i == 1)
-            ide_init_drive(s, hd1);
+            ide_init_drive(s, hd1, NULL);
    }
    bus->irq = irq;
 }
@@ -2669,6 +2675,25 @@ static bool is_identify_set(void *opaque, int version_id)
    return s->identify_set != 0;
 }

+static EndTransferFunc* transfer_end_table[] = {
+        ide_sector_read,
+        ide_sector_write,
+        ide_transfer_stop,
+        ide_atapi_cmd_reply_end,
+        ide_atapi_cmd,
+};
+
+static int transfer_end_table_idx(EndTransferFunc *fn)
+{
+    int i;
+
+    for (i = 0; i < ARRAY_SIZE(transfer_end_table); i++)
+        if (transfer_end_table[i] == fn)
+            return i;
+
+    return -1;
+}
+
 static int ide_drive_post_load(void *opaque, int version_id)
 {
    IDEState *s = opaque;
@@ -2679,14 +2704,42 @@ static int ide_drive_post_load(void *opaque, int version_id)
            s->cdrom_changed = 1;
        }
    }
+
+    if (s->cur_io_buffer_len) {
+        s->end_transfer_func = transfer_end_table[s->end_transfer_fn_idx];
+        s->data_ptr = s->io_buffer + s->cur_io_buffer_offset;
+        s->data_end = s->data_ptr + s->cur_io_buffer_len;
+    }
+        
    return 0;
 }

+static void ide_drive_pre_save(void *opaque)
+{
+    IDEState *s = opaque;
+
+    s->cur_io_buffer_len = 0;
+
+    if (!(s->status & DRQ_STAT))
+        return;
+
+    s->cur_io_buffer_offset = s->data_ptr - s->io_buffer;
+    s->cur_io_buffer_len = s->data_end - s->data_ptr;
+
+    s->end_transfer_fn_idx = transfer_end_table_idx(s->end_transfer_func);
+    if (s->end_transfer_fn_idx == -1) {
+        fprintf(stderr, "%s: invalid end_transfer_func for DRQ_STAT\n",
+                        __func__);
+        s->end_transfer_fn_idx = 2;
+    }
+}
+
 const VMStateDescription vmstate_ide_drive = {
    .name = "ide_drive",
-    .version_id = 3,
+    .version_id = 4,
    .minimum_version_id = 0,
    .minimum_version_id_old = 0,
+    .pre_save = ide_drive_pre_save,
    .post_load = ide_drive_post_load,
    .fields      = (VMStateField []) {
        VMSTATE_INT32(mult_sectors, IDEState),
@@ -2709,7 +2762,14 @@ const VMStateDescription vmstate_ide_drive = {
        VMSTATE_UINT8(sense_key, IDEState),
        VMSTATE_UINT8(asc, IDEState),
        VMSTATE_UINT8_V(cdrom_changed, IDEState, 3),
-        /* XXX: if a transfer is pending, we do not save it yet */
+        VMSTATE_INT32_V(req_nb_sectors, IDEState, 4),
+        VMSTATE_VARRAY_INT32(io_buffer, IDEState, io_buffer_total_len, 4,
+			     vmstate_info_uint8, uint8_t),
+        VMSTATE_INT32_V(cur_io_buffer_offset, IDEState, 4),
+        VMSTATE_INT32_V(cur_io_buffer_len, IDEState, 4),
+        VMSTATE_UINT8_V(end_transfer_fn_idx, IDEState, 4),
+        VMSTATE_INT32_V(elementary_transfer_size, IDEState, 4),
+        VMSTATE_INT32_V(packet_transfer_size, IDEState, 4),
        VMSTATE_END_OF_LIST()
    }
 };
@@ -397,6 +397,7 @@ struct IDEState {
    /* set for lba48 access */
    uint8_t lba48;
    BlockDriverState *bs;
+    char version[9];
    /* ATAPI specific */
    uint8_t sense_key;
    uint8_t asc;
@@ -416,6 +417,11 @@ struct IDEState {
    uint8_t *data_ptr;
    uint8_t *data_end;
    uint8_t *io_buffer;
+    /* PIO save/restore */
+    int32_t io_buffer_total_len;
+    int cur_io_buffer_offset;
+    int cur_io_buffer_len;
+    uint8_t end_transfer_fn_idx;
    QEMUTimer *sector_write_timer; /* only used for win2k install hack */
    uint32_t irq_count; /* counts IRQs when using win2k install hack */
    /* CF-ATA extended error */
@@ -449,6 +455,7 @@ struct IDEDevice {
    DeviceState qdev;
    uint32_t unit;
    DriveInfo *dinfo;
+    char *version;
 };

 typedef int (*ide_qdev_initfn)(IDEDevice *dev);
@@ -474,6 +481,7 @@ struct BMDMAState {
    uint8_t status;
    uint32_t addr;

+    struct PCIIDEState *pci_dev;
    IDEBus *bus;
    /* current transfer state */
    uint32_t cur_addr;
@@ -548,7 +556,7 @@ uint32_t ide_data_readw(void *opaque, uint32_t addr);
 void ide_data_writel(void *opaque, uint32_t addr, uint32_t val);
 uint32_t ide_data_readl(void *opaque, uint32_t addr);

-void ide_init_drive(IDEState *s, DriveInfo *dinfo);
+void ide_init_drive(IDEState *s, DriveInfo *dinfo, const char *version);
 void ide_init2(IDEBus *bus, DriveInfo *hd0, DriveInfo *hd1,
               qemu_irq irq);
 void ide_init_ioport(IDEBus *bus, int iobase, int iobase2);
@@ -78,6 +78,7 @@ static void bmdma_map(PCIDevice *pci_dev, int region_num,
        BMDMAState *bm = &d->bmdma[i];
        d->bus[i].bmdma = bm;
        bm->bus = d->bus+i;
+        bm->pci_dev = d;
        qemu_add_vm_change_state_handler(ide_dma_restart_cb, bm);

        register_ioport_write(addr, 1, 1, bmdma_cmd_writeb, bm);
@@ -99,7 +99,7 @@ typedef struct IDEDrive {
 static int ide_drive_initfn(IDEDevice *dev)
 {
    IDEBus *bus = DO_UPCAST(IDEBus, qbus, dev->qdev.parent_bus);
-    ide_init_drive(bus->ifs + dev->unit, dev->dinfo);
+    ide_init_drive(bus->ifs + dev->unit, dev->dinfo, dev->version);
    return 0;
 }

@@ -110,6 +110,7 @@ static IDEDeviceInfo ide_drive_info = {
    .qdev.props = (Property[]) {
        DEFINE_PROP_UINT32("unit", IDEDrive, dev.unit, -1),
        DEFINE_PROP_DRIVE("drive", IDEDrive, dev.dinfo),
+        DEFINE_PROP_STRING("ver",  IDEDrive, dev.version),
        DEFINE_PROP_END_OF_LIST(),
    }
 };
@@ -535,6 +535,7 @@ struct Rom {
    QTAILQ_ENTRY(Rom) next;
 };

+static FWCfgState *fw_cfg;
 static QTAILQ_HEAD(, Rom) roms = QTAILQ_HEAD_INITIALIZER(roms);
 int rom_enable_driver_roms;

@@ -556,7 +557,7 @@ static void rom_insert(Rom *rom)
    QTAILQ_INSERT_TAIL(&roms, rom, next);
 }

-int rom_add_file(const char *file, const char *fw_dir, const char *fw_file,
+int rom_add_file(const char *file, const char *fw_dir,
                 target_phys_addr_t addr)
 {
    Rom *rom;
@@ -576,8 +577,10 @@ int rom_add_file(const char *file, const char *fw_dir, const char *fw_file,
        goto err;
    }

-    rom->fw_dir  = fw_dir  ? qemu_strdup(fw_dir)  : NULL;
-    rom->fw_file = fw_file ? qemu_strdup(fw_file) : NULL;
+    if (fw_dir) {
+        rom->fw_dir  = qemu_strdup(fw_dir);
+        rom->fw_file = qemu_strdup(file);
+    }
    rom->addr    = addr;
    rom->romsize = lseek(fd, 0, SEEK_END);
    rom->data    = qemu_mallocz(rom->romsize);
@@ -590,6 +593,8 @@ int rom_add_file(const char *file, const char *fw_dir, const char *fw_file,
    }
    close(fd);
    rom_insert(rom);
+    if (rom->fw_file && fw_cfg)
+        fw_cfg_add_file(fw_cfg, rom->fw_dir, rom->fw_file, rom->data, rom->romsize);
    return 0;

 err:
@@ -621,14 +626,14 @@ int rom_add_vga(const char *file)
 {
    if (!rom_enable_driver_roms)
        return 0;
-    return rom_add_file(file, "vgaroms", file, 0);
+    return rom_add_file(file, "vgaroms", 0);
 }

 int rom_add_option(const char *file)
 {
    if (!rom_enable_driver_roms)
        return 0;
-    return rom_add_file(file, "genroms", file, 0);
+    return rom_add_file(file, "genroms", 0);
 }

 static void rom_reset(void *unused)
@@ -636,10 +641,12 @@ static void rom_reset(void *unused)
    Rom *rom;

    QTAILQ_FOREACH(rom, &roms, next) {
-        if (rom->addr == 0)
+        if (rom->fw_file) {
            continue;
-        if (rom->data == NULL)
+        }
+        if (rom->data == NULL) {
            continue;
+        }
        cpu_physical_memory_write_rom(rom->addr, rom->data, rom->romsize);
        if (rom->isrom) {
            /* rom needs to be written only once */
@@ -656,8 +663,9 @@ int rom_load_all(void)
    Rom *rom;

    QTAILQ_FOREACH(rom, &roms, next) {
-        if (rom->addr == 0)
+        if (rom->fw_file) {
            continue;
+        }
        if (addr > rom->addr) {
            fprintf(stderr, "rom: requested regions overlap "
                    "(rom %s. free=0x" TARGET_FMT_plx
@@ -676,16 +684,9 @@ int rom_load_all(void)
    return 0;
 }

-int rom_load_fw(void *fw_cfg)
+void rom_set_fw(void *f)
 {
-    Rom *rom;
-
-    QTAILQ_FOREACH(rom, &roms, next) {
-        if (!rom->fw_file)
-            continue;
-        fw_cfg_add_file(fw_cfg, rom->fw_dir, rom->fw_file, rom->data, rom->romsize);
-    }
-    return 0;
+    fw_cfg = f;
 }

 static Rom *find_rom(target_phys_addr_t addr)
@@ -693,17 +694,25 @@ static Rom *find_rom(target_phys_addr_t addr)
    Rom *rom;

    QTAILQ_FOREACH(rom, &roms, next) {
-        if (rom->addr == 0)
+        if (rom->fw_file) {
            continue;
-        if (rom->addr > addr)
+        }
+        if (rom->addr > addr) {
            continue;
-        if (rom->addr + rom->romsize < addr)
+        }
+        if (rom->addr + rom->romsize < addr) {
            continue;
+        }
        return rom;
    }
    return NULL;
 }

+/*
+ * Copies memory from registered ROMs to dest. Any memory that is contained in
+ * a ROM between addr and addr + size is copied. Note that this can involve
+ * multiple ROMs, which need not start at addr and need not end at addr + size.
+ */
 int rom_copy(uint8_t *dest, target_phys_addr_t addr, size_t size)
 {
    target_phys_addr_t end = addr + size;
@@ -712,16 +721,18 @@ int rom_copy(uint8_t *dest, target_phys_addr_t addr, size_t size)
    Rom *rom;

    QTAILQ_FOREACH(rom, &roms, next) {
-        if (rom->addr == 0)
+        if (rom->fw_file) {
            continue;
-        if (rom->addr > addr)
+        }
+        if (rom->addr + rom->romsize < addr) {
            continue;
-        if (rom->addr + rom->romsize < addr)
-            continue;
-        if (rom->addr > end)
+        }
+        if (rom->addr > end) {
            break;
-        if (!rom->data)
+        }
+        if (!rom->data) {
            continue;
+        }

        d = dest + (rom->addr - addr);
        s = rom->data;
@@ -757,17 +768,16 @@ void do_info_roms(Monitor *mon)
    Rom *rom;

    QTAILQ_FOREACH(rom, &roms, next) {
-        if (rom->addr) {
+        if (!rom->fw_file) {
            monitor_printf(mon, "addr=" TARGET_FMT_plx
                           " size=0x%06zx mem=%s name=\"%s\" \n",
                           rom->addr, rom->romsize,
                           rom->isrom ? "rom" : "ram",
                           rom->name);
        } else {
-            monitor_printf(mon, "fw=%s%s%s"
+            monitor_printf(mon, "fw=%s/%s"
                           " size=0x%06zx name=\"%s\" \n",
-                           rom->fw_dir ? rom->fw_dir : "",
-                           rom->fw_dir ? "/" : "",
+                           rom->fw_dir,
                           rom->fw_file,
                           rom->romsize,
                           rom->name);
@@ -19,18 +19,19 @@ void pstrcpy_targphys(const char *name,
                      target_phys_addr_t dest, int buf_size,
                      const char *source);

-int rom_add_file(const char *file, const char *fw_dir, const char *fw_file,
+
+int rom_add_file(const char *file, const char *fw_dir,
                 target_phys_addr_t addr);
 int rom_add_blob(const char *name, const void *blob, size_t len,
                 target_phys_addr_t addr);
 int rom_load_all(void);
-int rom_load_fw(void *fw_cfg);
+void rom_set_fw(void *f);
 int rom_copy(uint8_t *dest, target_phys_addr_t addr, size_t size);
 void *rom_ptr(target_phys_addr_t addr);
 void do_info_roms(Monitor *mon);

 #define rom_add_file_fixed(_f, _a)              \
-    rom_add_file(_f, NULL, NULL, _a)
+    rom_add_file(_f, NULL, _a)
 #define rom_add_blob_fixed(_f, _b, _l, _a)      \
    rom_add_blob(_f, _b, _l, _a)

@@ -173,11 +173,15 @@ do { fprintf(stderr, "lsi_scsi: error: " fmt , ## __VA_ARGS__);} while (0)
 /* Flag set if this is a tagged command.  */
 #define LSI_TAG_VALID     (1 << 16)

-typedef struct {
+typedef struct lsi_request {
    uint32_t tag;
+    SCSIDevice *dev;
+    uint32_t dma_len;
+    uint8_t *dma_buf;
    uint32_t pending;
    int out;
-} lsi_queue;
+    QTAILQ_ENTRY(lsi_request) next;
+} lsi_request;

 typedef struct {
    PCIDevice dev;
@@ -198,16 +202,13 @@ typedef struct {
     * 3 if a DMA operation is in progress.  */
    int waiting;
    SCSIBus bus;
-    SCSIDevice *current_dev;
+    SCSIDevice *select_dev;
    int current_lun;
    /* The tag is a combination of the device ID and the SCSI tag.  */
-    uint32_t current_tag;
-    uint32_t current_dma_len;
+    uint32_t select_tag;
    int command_complete;
-    uint8_t *dma_buf;
-    lsi_queue *queue;
-    int queue_len;
-    int active_commands;
+    QTAILQ_HEAD(, lsi_request) queue;
+    lsi_request *current;

    uint32_t dsa;
    uint32_t temp;
@@ -370,7 +371,7 @@ static int lsi_dma_64bit(LSIState *s)
 static uint8_t lsi_reg_readb(LSIState *s, int offset);
 static void lsi_reg_writeb(LSIState *s, int offset, uint8_t val);
 static void lsi_execute_script(LSIState *s);
-static void lsi_reselect(LSIState *s, uint32_t tag);
+static void lsi_reselect(LSIState *s, lsi_request *p);

 static inline uint32_t read_dword(LSIState *s, uint32_t addr)
 {
@@ -391,9 +392,9 @@ static void lsi_stop_script(LSIState *s)

 static void lsi_update_irq(LSIState *s)
 {
-    int i;
    int level;
    static int last_level;
+    lsi_request *p;

    /* It's unclear whether the DIP/SIP bits should be cleared when the
       Interrupt Status Registers are cleared or when istat0 is read.
@@ -427,9 +428,9 @@ static void lsi_update_irq(LSIState *s)
    if (!level && lsi_irq_on_rsl(s) && !(s->scntl1 & LSI_SCNTL1_CON)) {
        DPRINTF("Handled IRQs & disconnected, looking for pending "
                "processes\n");
-        for (i = 0; i < s->active_commands; i++) {
-            if (s->queue[i].pending) {
-                lsi_reselect(s, s->queue[i].tag);
+        QTAILQ_FOREACH(p, &s->queue, next) {
+            if (p->pending) {
+                lsi_reselect(s, p);
                break;
            }
        }
@@ -508,15 +509,16 @@ static void lsi_do_dma(LSIState *s, int out)
    uint32_t count;
    target_phys_addr_t addr;

-    if (!s->current_dma_len) {
+    assert(s->current);
+    if (!s->current->dma_len) {
        /* Wait until data is available.  */
        DPRINTF("DMA no data available\n");
        return;
    }

    count = s->dbc;
-    if (count > s->current_dma_len)
-        count = s->current_dma_len;
+    if (count > s->current->dma_len)
+        count = s->current->dma_len;

    addr = s->dnad;
    /* both 40 and Table Indirect 64-bit DMAs store upper bits in dnad64 */
@@ -532,29 +534,29 @@ static void lsi_do_dma(LSIState *s, int out)
    s->dnad += count;
    s->dbc -= count;

-    if (s->dma_buf == NULL) {
-        s->dma_buf = s->current_dev->info->get_buf(s->current_dev,
-                                                   s->current_tag);
+    if (s->current->dma_buf == NULL) {
+        s->current->dma_buf = s->current->dev->info->get_buf(s->current->dev,
+                                                             s->current->tag);
    }

    /* ??? Set SFBR to first data byte.  */
    if (out) {
-        cpu_physical_memory_read(addr, s->dma_buf, count);
+        cpu_physical_memory_read(addr, s->current->dma_buf, count);
    } else {
-        cpu_physical_memory_write(addr, s->dma_buf, count);
+        cpu_physical_memory_write(addr, s->current->dma_buf, count);
    }
-    s->current_dma_len -= count;
-    if (s->current_dma_len == 0) {
-        s->dma_buf = NULL;
+    s->current->dma_len -= count;
+    if (s->current->dma_len == 0) {
+        s->current->dma_buf = NULL;
        if (out) {
            /* Write the data.  */
-            s->current_dev->info->write_data(s->current_dev, s->current_tag);
+            s->current->dev->info->write_data(s->current->dev, s->current->tag);
        } else {
            /* Request any remaining data.  */
-            s->current_dev->info->read_data(s->current_dev, s->current_tag);
+            s->current->dev->info->read_data(s->current->dev, s->current->tag);
        }
    } else {
-        s->dma_buf += count;
+        s->current->dma_buf += count;
        lsi_resume_script(s);
    }
 }
@@ -563,15 +565,14 @@ static void lsi_do_dma(LSIState *s, int out)
 /* Add a command to the queue.  */
 static void lsi_queue_command(LSIState *s)
 {
-    lsi_queue *p;
+    lsi_request *p = s->current;

    DPRINTF("Queueing tag=0x%x\n", s->current_tag);
-    if (s->queue_len == s->active_commands) {
-        s->queue_len++;
-        s->queue = qemu_realloc(s->queue, s->queue_len * sizeof(lsi_queue));
-    }
-    p = &s->queue[s->active_commands++];
-    p->tag = s->current_tag;
+    assert(s->current != NULL);
+    assert(s->current->dma_len == 0);
+    QTAILQ_INSERT_TAIL(&s->queue, s->current, next);
+    s->current = NULL;
+
    p->pending = 0;
    p->out = (s->sstat1 & PHASE_MASK) == PHASE_DO;
 }
@@ -588,45 +589,29 @@ static void lsi_add_msg_byte(LSIState *s, uint8_t data)
 }

 /* Perform reselection to continue a command.  */
-static void lsi_reselect(LSIState *s, uint32_t tag)
+static void lsi_reselect(LSIState *s, lsi_request *p)
 {
-    lsi_queue *p;
-    int n;
    int id;

-    p = NULL;
-    for (n = 0; n < s->active_commands; n++) {
-        p = &s->queue[n];
-        if (p->tag == tag)
-            break;
-    }
-    if (n == s->active_commands) {
-        BADF("Reselected non-existant command tag=0x%x\n", tag);
-        return;
-    }
-    id = (tag >> 8) & 0xf;
+    assert(s->current == NULL);
+    QTAILQ_REMOVE(&s->queue, p, next);
+    s->current = p;
+
+    id = (p->tag >> 8) & 0xf;
    s->ssid = id | 0x80;
    /* LSI53C700 Family Compatibility, see LSI53C895A 4-73 */
    if (!s->dcntl & LSI_DCNTL_COM) {
        s->sfbr = 1 << (id & 0x7);
    }
    DPRINTF("Reselected target %d\n", id);
-    s->current_dev = s->bus.devs[id];
-    s->current_tag = tag;
    s->scntl1 |= LSI_SCNTL1_CON;
    lsi_set_phase(s, PHASE_MI);
    s->msg_action = p->out ? 2 : 3;
-    s->current_dma_len = p->pending;
-    s->dma_buf = NULL;
+    s->current->dma_len = p->pending;
    lsi_add_msg_byte(s, 0x80);
-    if (s->current_tag & LSI_TAG_VALID) {
+    if (s->current->tag & LSI_TAG_VALID) {
        lsi_add_msg_byte(s, 0x20);
-        lsi_add_msg_byte(s, tag & 0xff);
-    }
-
-    s->active_commands--;
-    if (n != s->active_commands) {
-        s->queue[n] = s->queue[s->active_commands];
+        lsi_add_msg_byte(s, p->tag & 0xff);
    }

    if (lsi_irq_on_rsl(s)) {
@@ -638,10 +623,9 @@ static void lsi_reselect(LSIState *s, uint32_t tag)
   the device was reselected, nonzero if the IO is deferred.  */
 static int lsi_queue_tag(LSIState *s, uint32_t tag, uint32_t arg)
 {
-    lsi_queue *p;
-    int i;
-    for (i = 0; i < s->active_commands; i++) {
-        p = &s->queue[i];
+    lsi_request *p;
+
+    QTAILQ_FOREACH(p, &s->queue, next) {
        if (p->tag == tag) {
            if (p->pending) {
                BADF("Multiple IO pending for tag %d\n", tag);
@@ -656,10 +640,10 @@ static int lsi_queue_tag(LSIState *s, uint32_t tag, uint32_t arg)
                (lsi_irq_on_rsl(s) && !(s->scntl1 & LSI_SCNTL1_CON) &&
                 !(s->istat0 & (LSI_ISTAT0_SIP | LSI_ISTAT0_DIP)))) {
                /* Reselect device.  */
-                lsi_reselect(s, tag);
+                lsi_reselect(s, p);
                return 0;
            } else {
-               DPRINTF("Queueing IO tag=0x%x\n", tag);
+                DPRINTF("Queueing IO tag=0x%x\n", tag);
                p->pending = arg;
                return 1;
            }
@@ -687,11 +671,15 @@ static void lsi_command_complete(SCSIBus *bus, int reason, uint32_t tag,
        } else {
            lsi_set_phase(s, PHASE_ST);
        }
+
+        qemu_free(s->current);
+        s->current = NULL;
+
        lsi_resume_script(s);
        return;
    }

-    if (s->waiting == 1 || tag != s->current_tag ||
+    if (s->waiting == 1 || !s->current || tag != s->current->tag ||
        (lsi_irq_on_rsl(s) && !(s->scntl1 & LSI_SCNTL1_CON))) {
        if (lsi_queue_tag(s, tag, arg))
            return;
@@ -699,7 +687,7 @@ static void lsi_command_complete(SCSIBus *bus, int reason, uint32_t tag,

    /* host adapter (re)connected */
    DPRINTF("Data ready tag=0x%x len=%d\n", tag, arg);
-    s->current_dma_len = arg;
+    s->current->dma_len = arg;
    s->command_complete = 1;
    if (!s->waiting)
        return;
@@ -721,14 +709,20 @@ static void lsi_do_command(LSIState *s)
    cpu_physical_memory_read(s->dnad, buf, s->dbc);
    s->sfbr = buf[0];
    s->command_complete = 0;
-    n = s->current_dev->info->send_command(s->current_dev, s->current_tag, buf,
-                                           s->current_lun);
+
+    assert(s->current == NULL);
+    s->current = qemu_mallocz(sizeof(lsi_request));
+    s->current->tag = s->select_tag;
+    s->current->dev = s->select_dev;
+
+    n = s->current->dev->info->send_command(s->current->dev, s->current->tag, buf,
+                                            s->current_lun);
    if (n > 0) {
        lsi_set_phase(s, PHASE_DI);
-        s->current_dev->info->read_data(s->current_dev, s->current_tag);
+        s->current->dev->info->read_data(s->current->dev, s->current->tag);
    } else if (n < 0) {
        lsi_set_phase(s, PHASE_DO);
-        s->current_dev->info->write_data(s->current_dev, s->current_tag);
+        s->current->dev->info->write_data(s->current->dev, s->current->tag);
    }

    if (!s->command_complete) {
@@ -851,16 +845,16 @@ static void lsi_do_msgout(LSIState *s)
            }
            break;
        case 0x20: /* SIMPLE queue */
-            s->current_tag |= lsi_get_msgbyte(s) | LSI_TAG_VALID;
+            s->select_tag |= lsi_get_msgbyte(s) | LSI_TAG_VALID;
            DPRINTF("SIMPLE queue tag=0x%x\n", s->current_tag & 0xff);
            break;
        case 0x21: /* HEAD of queue */
            BADF("HEAD queue not implemented\n");
-            s->current_tag |= lsi_get_msgbyte(s) | LSI_TAG_VALID;
+            s->select_tag |= lsi_get_msgbyte(s) | LSI_TAG_VALID;
            break;
        case 0x22: /* ORDERED queue */
            BADF("ORDERED queue not implemented\n");
-            s->current_tag |= lsi_get_msgbyte(s) | LSI_TAG_VALID;
+            s->select_tag |= lsi_get_msgbyte(s) | LSI_TAG_VALID;
            break;
        default:
            if ((msg & 0x80) == 0) {
@@ -905,17 +899,17 @@ static void lsi_memcpy(LSIState *s, uint32_t dest, uint32_t src, int count)

 static void lsi_wait_reselect(LSIState *s)
 {
-    int i;
+    lsi_request *p;
+
    DPRINTF("Wait Reselect\n");
-    if (s->current_dma_len)
-        BADF("Reselect with pending DMA\n");
-    for (i = 0; i < s->active_commands; i++) {
-        if (s->queue[i].pending) {
-            lsi_reselect(s, s->queue[i].tag);
+
+    QTAILQ_FOREACH(p, &s->queue, next) {
+        if (p->pending) {
+            lsi_reselect(s, p);
            break;
        }
    }
-    if (s->current_dma_len == 0) {
+    if (s->current == NULL) {
        s->waiting = 1;
    }
 }
@@ -1093,8 +1087,8 @@ again:
                /* ??? Linux drivers compain when this is set.  Maybe
                   it only applies in low-level mode (unimplemented).
                lsi_script_scsi_interrupt(s, LSI_SIST0_CMP, 0); */
-                s->current_dev = s->bus.devs[id];
-                s->current_tag = id << 8;
+                s->select_dev = s->bus.devs[id];
+                s->select_tag = id << 8;
                s->scntl1 |= LSI_SCNTL1_CON;
                if (insn & (1 << 3)) {
                    s->socl |= LSI_SOCL_ATN;
@@ -2006,9 +2000,11 @@ static void lsi_pre_save(void *opaque)
 {
    LSIState *s = opaque;

-    assert(s->dma_buf == NULL);
-    assert(s->current_dma_len == 0);
-    assert(s->active_commands == 0);
+    if (s->current) {
+        assert(s->current->dma_buf == NULL);
+        assert(s->current->dma_len == 0);
+    }
+    assert(QTAILQ_EMPTY(&s->queue));
 }

 static const VMStateDescription vmstate_lsi_scsi = {
@@ -2101,8 +2097,6 @@ static int lsi_scsi_uninit(PCIDevice *d)
    cpu_unregister_io_memory(s->mmio_io_addr);
    cpu_unregister_io_memory(s->ram_io_addr);

-    qemu_free(s->queue);
-
    return 0;
 }

@@ -2138,9 +2132,7 @@ static int lsi_scsi_init(PCIDevice *dev)
                           PCI_BASE_ADDRESS_SPACE_MEMORY, lsi_mmio_mapfunc);
    pci_register_bar((struct PCIDevice *)s, 2, 0x2000,
                           PCI_BASE_ADDRESS_SPACE_MEMORY, lsi_ram_mapfunc);
-    s->queue = qemu_malloc(sizeof(lsi_queue));
-    s->queue_len = 1;
-    s->active_commands = 0;
+    QTAILQ_INIT(&s->queue);

    lsi_soft_reset(s);

@@ -30,7 +30,7 @@

 //#define DEBUG_CMOS

-#define RTC_REINJECT_ON_ACK_COUNT 1000
+#define RTC_REINJECT_ON_ACK_COUNT 20

 #define RTC_SECONDS             0
 #define RTC_SECONDS_ALARM       1
@@ -67,7 +67,7 @@
 #define MP_AUDIO_IRQ            30

 /* Wolfson 8750 I2C address */
-#define MP_WM_ADDR              0x34
+#define MP_WM_ADDR              0x1A

 /* Ethernet register offsets */
 #define MP_ETH_SMIR             0x010
@@ -238,14 +238,13 @@ static void eth_send(mv88w8618_eth_state *s, int queue_index)
 {
    uint32_t desc_addr = s->tx_queue[queue_index];
    mv88w8618_tx_desc desc;
+    uint32_t next_desc;
    uint8_t buf[2048];
    int len;

-    if (!desc_addr) {
-        return;
-    }
    do {
        eth_tx_desc_get(desc_addr, &desc);
+        next_desc = desc.next;
        if (desc.cmdstat & MP_ETH_TX_OWN) {
            len = desc.bytes;
            if (len < 2048) {
@@ -256,7 +255,7 @@ static void eth_send(mv88w8618_eth_state *s, int queue_index)
            s->icr |= 1 << (MP_ETH_IRQ_TXLO_BIT - queue_index);
            eth_tx_desc_put(desc_addr, &desc);
        }
-        desc_addr = desc.next;
+        desc_addr = next_desc;
    } while (desc_addr != s->tx_queue[queue_index]);
 }

@@ -1090,6 +1090,7 @@ static void pc_init1(ram_addr_t ram_size,
                                 bios_size, bios_offset | IO_MEM_ROM);

    fw_cfg = bochs_bios_init();
+    rom_set_fw(fw_cfg);

    if (linux_boot) {
        load_linux(fw_cfg, kernel_filename, initrd_filename, kernel_cmdline, below_4g_mem_size);
@@ -1250,8 +1251,6 @@ static void pc_init1(ram_addr_t ram_size,
            }
        }
    }
-
-    rom_load_fw(fw_cfg);
 }

 static void pc_init_pci(ram_addr_t ram_size,
@@ -1289,7 +1288,7 @@ void cmos_set_s3_resume(void)
 }

 static QEMUMachine pc_machine = {
-    .name = "pc-0.11",
+    .name = "pc-0.12",
    .alias = "pc",
    .desc = "Standard PC",
    .init = pc_init_pci,
@@ -1297,6 +1296,33 @@ static QEMUMachine pc_machine = {
    .is_default = 1,
 };

+static QEMUMachine pc_machine_v0_11 = {
+    .name = "pc-0.11",
+    .desc = "Standard PC, qemu 0.11",
+    .init = pc_init_pci,
+    .max_cpus = 255,
+    .compat_props = (GlobalProperty[]) {
+        {
+            .driver   = "virtio-blk-pci",
+            .property = "vectors",
+            .value    = stringify(0),
+        },{
+            .driver   = "ide-drive",
+            .property = "ver",
+            .value    = "0.11",
+        },{
+            .driver   = "scsi-disk",
+            .property = "ver",
+            .value    = "0.11",
+        },{
+            .driver   = "PCI",
+            .property = "rombar",
+            .value    = stringify(0),
+        },
+        { /* end of list */ }
+    }
+};
+
 static QEMUMachine pc_machine_v0_10 = {
    .name = "pc-0.10",
    .desc = "Standard PC, qemu 0.10",
@@ -1319,6 +1345,18 @@ static QEMUMachine pc_machine_v0_10 = {
            .driver   = "virtio-blk-pci",
            .property = "vectors",
            .value    = stringify(0),
+        },{
+            .driver   = "ide-drive",
+            .property = "ver",
+            .value    = "0.10",
+        },{
+            .driver   = "scsi-disk",
+            .property = "ver",
+            .value    = "0.10",
+        },{
+            .driver   = "PCI",
+            .property = "rombar",
+            .value    = stringify(0),
        },
        { /* end of list */ }
    },
@@ -1334,6 +1372,7 @@ static QEMUMachine isapc_machine = {
 static void pc_machine_init(void)
 {
    qemu_register_machine(&pc_machine);
+    qemu_register_machine(&pc_machine_v0_11);
    qemu_register_machine(&pc_machine_v0_10);
    qemu_register_machine(&isapc_machine);
 }
@@ -64,6 +64,7 @@ static struct BusInfo pci_bus_info = {
    .props      = (Property[]) {
        DEFINE_PROP_PCI_DEVFN("addr", PCIDevice, devfn, -1),
        DEFINE_PROP_STRING("romfile", PCIDevice, romfile),
+        DEFINE_PROP_UINT32("rombar",  PCIDevice, rom_bar, 1),
        DEFINE_PROP_END_OF_LIST()
    }
 };
@@ -1464,6 +1465,20 @@ static int pci_add_option_rom(PCIDevice *pdev)
    if (strlen(pdev->romfile) == 0)
        return 0;

+    if (!pdev->rom_bar) {
+        /*
+         * Load rom via fw_cfg instead of creating a rom bar,
+         * for 0.11 compatibility.
+         */
+        int class = pci_get_word(pdev->config + PCI_CLASS_DEVICE);
+        if (class == 0x0300) {
+            rom_add_vga(pdev->romfile);
+        } else {
+            rom_add_option(pdev->romfile);
+        }
+        return 0;
+    }
+
    path = qemu_find_file(QEMU_FILE_TYPE_BIOS, pdev->romfile);
    if (path == NULL) {
        path = qemu_strdup(pdev->romfile);
@@ -246,6 +246,7 @@ struct PCIDevice {
    /* Location of option rom */
    char *romfile;
    ram_addr_t rom_offset;
+    uint32_t rom_bar;
 };

 PCIDevice *pci_register_device(PCIBus *bus, const char *name,
@@ -5,6 +5,12 @@
 * Based on code by Fabrice Bellard
 *
 * Written by Paul Brook
+ * Modifications:
+ *  2009-Dec-12 Artyom Tarasenko : implemented stamdard inquiry for the case
+ *                                 when the allocation length of CDB is smaller
+ *                                 than 36.
+ *  2009-Oct-13 Artyom Tarasenko : implemented the block descriptor in the
+ *                                 MODE SENSE response.
 *
 * This code is licenced under the LGPL.
 *
@@ -59,6 +65,7 @@ struct SCSIDiskState
    int cluster_size;
    uint64_t max_lba;
    QEMUBH *bh;
+    char *version;
 };

 static SCSIDiskReq *scsi_new_request(SCSIDevice *d, uint32_t tag, uint32_t lun)
@@ -74,7 +81,7 @@ static SCSIDiskReq *scsi_new_request(SCSIDevice *d, uint32_t tag, uint32_t lun)

 static void scsi_remove_request(SCSIDiskReq *r)
 {
-    qemu_free(r->iov.iov_base);
+    qemu_vfree(r->iov.iov_base);
    scsi_req_free(&r->req);
 }

@@ -309,6 +316,7 @@ static uint8_t *scsi_get_buf(SCSIDevice *d, uint32_t tag)
 static int scsi_disk_emulate_inquiry(SCSIRequest *req, uint8_t *outbuf)
 {
    BlockDriverState *bdrv = req->dev->dinfo->bdrv;
+    SCSIDiskState *s = DO_UPCAST(SCSIDiskState, qdev, req->dev);
    int buflen = 0;

    if (req->cmd.buf[1] & 0x2) {
@@ -406,11 +414,6 @@ static int scsi_disk_emulate_inquiry(SCSIRequest *req, uint8_t *outbuf)
        return -1;
    }

-    if (req->cmd.xfer < 36) {
-        BADF("Error: Inquiry (STANDARD) buffer size %zd "
-             "is less than 36 (TODO: only 5 required)\n", req->cmd.xfer);
-    }
-
    buflen = req->cmd.xfer;
    if (buflen > SCSI_MAX_INQUIRY_LEN)
        buflen = SCSI_MAX_INQUIRY_LEN;
@@ -431,12 +434,22 @@ static int scsi_disk_emulate_inquiry(SCSIRequest *req, uint8_t *outbuf)
        memcpy(&outbuf[16], "QEMU HARDDISK   ", 16);
    }
    memcpy(&outbuf[8], "QEMU    ", 8);
-    memcpy(&outbuf[32], QEMU_VERSION, 4);
+    memset(&outbuf[32], 0, 4);
+    memcpy(&outbuf[32], s->version ? s->version : QEMU_VERSION,
+           MIN(4, strlen(s->version ? s->version : QEMU_VERSION)));
    /* Identify device as SCSI-3 rev 1.
       Some later commands are also implemented. */
    outbuf[2] = 3;
    outbuf[3] = 2; /* Format 2 */
-    outbuf[4] = buflen - 5; /* Additional Length = (Len - 1) - 4 */
+
+    if (buflen > 36) {
+        outbuf[4] = buflen - 5; /* Additional Length = (Len - 1) - 4 */
+    } else {
+        /* If the allocation length of CDB is too small,
+               the additional length is not adjusted */
+        outbuf[4] = 36 - 5;
+    }
+
    /* Sync data transfer and TCQ.  */
    outbuf[7] = 0x10 | (req->bus->tcq ? 0x02 : 0);
    return buflen;
@@ -1020,6 +1033,7 @@ static SCSIDeviceInfo scsi_disk_info = {
    .get_buf      = scsi_get_buf,
    .qdev.props   = (Property[]) {
        DEFINE_PROP_DRIVE("drive", SCSIDiskState, qdev.dinfo),
+        DEFINE_PROP_STRING("ver",  SCSIDiskState, version),
        DEFINE_PROP_END_OF_LIST(),
    },
 };
@@ -396,8 +396,11 @@ static void sh7750_mem_writel(void *opaque, target_phys_addr_t addr,
 	portb_changed(s, temp);
 	return;
    case SH7750_MMUCR_A7:
-	s->cpu->mmucr = mem_value;
-	return;
+        if (mem_value & MMUCR_TI) {
+            cpu_sh4_invalidate_tlb(s->cpu);
+        }
+        s->cpu->mmucr = mem_value & ~MMUCR_TI;
+        return;
    case SH7750_PTEH_A7:
        /* If asid changes, clear all registered tlb entries. */
 	if ((s->cpu->pteh & 0xff) != (mem_value & 0xff))
@@ -47,10 +47,15 @@ static void sh_pci_reg_write (void *p, target_phys_addr_t addr, uint32_t val)
        pcic->par = val;
        break;
    case 0x1c4:
-        pcic->mbr = val;
+        pcic->mbr = val & 0xff000001;
        break;
    case 0x1c8:
-        pcic->iobr = val;
+        if ((val & 0xfffc0000) != (pcic->iobr & 0xfffc0000)) {
+            cpu_register_physical_memory(pcic->iobr & 0xfffc0000, 0x40000,
+                                         IO_MEM_UNASSIGNED);
+            pcic->iobr = val & 0xfffc0001;
+            isa_mmio_init(pcic->iobr & 0xfffc0000, 0x40000);
+        }
        break;
    case 0x220:
        pci_data_write(pcic->bus, pcic->par, val, 4);
@@ -66,89 +71,16 @@ static uint32_t sh_pci_reg_read (void *p, target_phys_addr_t addr)
        return le32_to_cpup((uint32_t*)(pcic->dev->config + addr));
    case 0x1c0:
        return pcic->par;
+    case 0x1c4:
+        return pcic->mbr;
+    case 0x1c8:
+        return pcic->iobr;
    case 0x220:
        return pci_data_read(pcic->bus, pcic->par, 4);
    }
    return 0;
 }

-static void sh_pci_data_write (SHPCIC *pcic, target_phys_addr_t addr,
-                               uint32_t val, int size)
-{
-    pci_data_write(pcic->bus, addr + pcic->mbr, val, size);
-}
-
-static uint32_t sh_pci_mem_read (SHPCIC *pcic, target_phys_addr_t addr,
-                                 int size)
-{
-    return pci_data_read(pcic->bus, addr + pcic->mbr, size);
-}
-
-static void sh_pci_writeb (void *p, target_phys_addr_t addr, uint32_t val)
-{
-    sh_pci_data_write(p, addr, val, 1);
-}
-
-static void sh_pci_writew (void *p, target_phys_addr_t addr, uint32_t val)
-{
-    sh_pci_data_write(p, addr, val, 2);
-}
-
-static void sh_pci_writel (void *p, target_phys_addr_t addr, uint32_t val)
-{
-    sh_pci_data_write(p, addr, val, 4);
-}
-
-static uint32_t sh_pci_readb (void *p, target_phys_addr_t addr)
-{
-    return sh_pci_mem_read(p, addr, 1);
-}
-
-static uint32_t sh_pci_readw (void *p, target_phys_addr_t addr)
-{
-    return sh_pci_mem_read(p, addr, 2);
-}
-
-static uint32_t sh_pci_readl (void *p, target_phys_addr_t addr)
-{
-    return sh_pci_mem_read(p, addr, 4);
-}
-
-static int sh_pci_addr2port(SHPCIC *pcic, target_phys_addr_t addr)
-{
-    return addr + pcic->iobr;
-}
-
-static void sh_pci_outb (void *p, target_phys_addr_t addr, uint32_t val)
-{
-    cpu_outb(sh_pci_addr2port(p, addr), val);
-}
-
-static void sh_pci_outw (void *p, target_phys_addr_t addr, uint32_t val)
-{
-    cpu_outw(sh_pci_addr2port(p, addr), val);
-}
-
-static void sh_pci_outl (void *p, target_phys_addr_t addr, uint32_t val)
-{
-    cpu_outl(sh_pci_addr2port(p, addr), val);
-}
-
-static uint32_t sh_pci_inb (void *p, target_phys_addr_t addr)
-{
-    return cpu_inb(sh_pci_addr2port(p, addr));
-}
-
-static uint32_t sh_pci_inw (void *p, target_phys_addr_t addr)
-{
-    return cpu_inw(sh_pci_addr2port(p, addr));
-}
-
-static uint32_t sh_pci_inl (void *p, target_phys_addr_t addr)
-{
-    return cpu_inl(sh_pci_addr2port(p, addr));
-}
-
 typedef struct {
    CPUReadMemoryFunc * const r[3];
    CPUWriteMemoryFunc * const w[3];
@@ -159,21 +91,11 @@ static MemOp sh_pci_reg = {
    { NULL, NULL, sh_pci_reg_write },
 };

-static MemOp sh_pci_mem = {
-    { sh_pci_readb, sh_pci_readw, sh_pci_readl },
-    { sh_pci_writeb, sh_pci_writew, sh_pci_writel },
-};
-
-static MemOp sh_pci_iop = {
-    { sh_pci_inb, sh_pci_inw, sh_pci_inl },
-    { sh_pci_outb, sh_pci_outw, sh_pci_outl },
-};
-
 PCIBus *sh_pci_register_bus(pci_set_irq_fn set_irq, pci_map_irq_fn map_irq,
                            void *opaque, int devfn_min, int nirq)
 {
    SHPCIC *p;
-    int mem, reg, iop;
+    int reg;

    p = qemu_mallocz(sizeof(SHPCIC));
    p->bus = pci_register_bus(NULL, "pci",
@@ -182,14 +104,11 @@ PCIBus *sh_pci_register_bus(pci_set_irq_fn set_irq, pci_map_irq_fn map_irq,
    p->dev = pci_register_device(p->bus, "SH PCIC", sizeof(PCIDevice),
                                 -1, NULL, NULL);
    reg = cpu_register_io_memory(sh_pci_reg.r, sh_pci_reg.w, p);
-    iop = cpu_register_io_memory(sh_pci_iop.r, sh_pci_iop.w, p);
-    mem = cpu_register_io_memory(sh_pci_mem.r, sh_pci_mem.w, p);
    cpu_register_physical_memory(0x1e200000, 0x224, reg);
-    cpu_register_physical_memory(0x1e240000, 0x40000, iop);
-    cpu_register_physical_memory(0x1d000000, 0x1000000, mem);
    cpu_register_physical_memory(0xfe200000, 0x224, reg);
-    cpu_register_physical_memory(0xfe240000, 0x40000, iop);
-    cpu_register_physical_memory(0xfd000000, 0x1000000, mem);
+
+    p->iobr = 0xfe240000;
+    isa_mmio_init(p->iobr, 0x40000);

    pci_config_set_vendor_id(p->dev->config, PCI_VENDOR_ID_HITACHI);
    pci_config_set_device_id(p->dev->config, PCI_DEVICE_ID_HITACHI_SH7751R);
@@ -1492,7 +1492,7 @@ static USBDevice *usb_net_init(const char *cmdline)

    dev = usb_create(NULL /* FIXME */, "usb-net");
    qdev_set_nic_properties(&dev->qdev, &nd_table[idx]);
-    qdev_init(&dev->qdev);
+    qdev_init_nofail(&dev->qdev);
    return dev;
 }

@@ -497,12 +497,28 @@ static int usb_serial_can_read(void *opaque)
 static void usb_serial_read(void *opaque, const uint8_t *buf, int size)
 {
    USBSerialState *s = opaque;
-    int first_size = RECV_BUF - s->recv_ptr;
-    if (first_size > size)
-        first_size = size;
-    memcpy(s->recv_buf + s->recv_ptr + s->recv_used, buf, first_size);
-    if (size > first_size)
-        memcpy(s->recv_buf, buf + first_size, size - first_size);
+    int first_size, start;
+
+    /* room in the buffer? */
+    if (size > (RECV_BUF - s->recv_used))
+        size = RECV_BUF - s->recv_used;
+
+    start = s->recv_ptr + s->recv_used;
+    if (start < RECV_BUF) {
+        /* copy data to end of buffer */
+        first_size = RECV_BUF - start;
+        if (first_size > size)
+            first_size = size;
+
+        memcpy(s->recv_buf + start, buf, first_size);
+
+        /* wrap around to front if needed */
+        if (size > first_size)
+            memcpy(s->recv_buf, buf + first_size, size - first_size);
+    } else {
+        start -= RECV_BUF;
+        memcpy(s->recv_buf + start, buf, size);
+    }
    s->recv_used += size;
 }

@@ -583,7 +599,7 @@ static USBDevice *usb_serial_init(const char *filename)
        qdev_prop_set_uint16(&dev->qdev, "vendorid", vendorid);
    if (productid)
        qdev_prop_set_uint16(&dev->qdev, "productid", productid);
-    qdev_init(&dev->qdev);
+    qdev_init_nofail(&dev->qdev);

    return dev;
 }
@@ -599,7 +615,7 @@ static USBDevice *usb_braille_init(const char *unused)

    dev = usb_create(NULL /* FIXME */, "usb-braille");
    qdev_prop_set_chr(&dev->qdev, "chardev", cdrv);
-    qdev_init(&dev->qdev);
+    qdev_init_nofail(&dev->qdev);

    return dev;
 }
@@ -677,9 +677,6 @@ static int uhci_complete_td(UHCIState *s, UHCI_TD *td, UHCIAsync *async, uint32_

    ret = async->packet.len;

-    if (td->ctrl & TD_CTRL_IOC)
-        *int_mask |= 0x01;
-
    if (td->ctrl & TD_CTRL_IOS)
        td->ctrl &= ~TD_CTRL_ACTIVE;

@@ -693,6 +690,8 @@ static int uhci_complete_td(UHCIState *s, UHCI_TD *td, UHCIAsync *async, uint32_
       here.  The docs are somewhat unclear, but win2k relies on this
       behavior.  */
    td->ctrl &= ~(TD_CTRL_ACTIVE | TD_CTRL_NAK);
+    if (td->ctrl & TD_CTRL_IOC)
+        *int_mask |= 0x01;

    if (pid == USB_TOKEN_IN) {
        if (len > max_len) {
@@ -750,6 +749,8 @@ out:
        if (err == 0) {
            td->ctrl &= ~TD_CTRL_ACTIVE;
            s->status |= UHCI_STS_USBERR;
+            if (td->ctrl & TD_CTRL_IOC)
+                *int_mask |= 0x01;
            uhci_update_irq(s);
        }
    }
@@ -272,7 +272,7 @@ static void do_multiwrite(BlockDriverState *bs, BlockRequest *blkreq,
    if (ret != 0) {
        for (i = 0; i < num_writes; i++) {
            if (blkreq[i].error) {
-                virtio_blk_req_complete(blkreq[i].opaque, VIRTIO_BLK_S_IOERR);
+                virtio_blk_rw_complete(blkreq[i].opaque, -EIO);
            }
        }
    }
@@ -316,7 +316,46 @@ static void virtio_blk_handle_read(VirtIOBlockReq *req)
    acb = bdrv_aio_readv(req->dev->bs, req->out->sector, &req->qiov,
                         req->qiov.size / 512, virtio_blk_rw_complete, req);
    if (!acb) {
-        virtio_blk_req_complete(req, VIRTIO_BLK_S_IOERR);
+        virtio_blk_rw_complete(req, -EIO);
+    }
+}
+
+typedef struct MultiReqBuffer {
+    BlockRequest        blkreq[32];
+    int                 num_writes;
+    BlockDriverState    *old_bs;
+} MultiReqBuffer;
+
+static void virtio_blk_handle_request(VirtIOBlockReq *req,
+    MultiReqBuffer *mrb)
+{
+    if (req->elem.out_num < 1 || req->elem.in_num < 1) {
+        fprintf(stderr, "virtio-blk missing headers\n");
+        exit(1);
+    }
+
+    if (req->elem.out_sg[0].iov_len < sizeof(*req->out) ||
+        req->elem.in_sg[req->elem.in_num - 1].iov_len < sizeof(*req->in)) {
+        fprintf(stderr, "virtio-blk header not in correct element\n");
+        exit(1);
+    }
+
+    req->out = (void *)req->elem.out_sg[0].iov_base;
+    req->in = (void *)req->elem.in_sg[req->elem.in_num - 1].iov_base;
+
+    if (req->out->type & VIRTIO_BLK_T_FLUSH) {
+        virtio_blk_handle_flush(req);
+    } else if (req->out->type & VIRTIO_BLK_T_SCSI_CMD) {
+        virtio_blk_handle_scsi(req);
+    } else if (req->out->type & VIRTIO_BLK_T_OUT) {
+        qemu_iovec_init_external(&req->qiov, &req->elem.out_sg[1],
+                                 req->elem.out_num - 1);
+        virtio_blk_handle_write(mrb->blkreq, &mrb->num_writes,
+            req, &mrb->old_bs);
+    } else {
+        qemu_iovec_init_external(&req->qiov, &req->elem.in_sg[0],
+                                 req->elem.in_num - 1);
+        virtio_blk_handle_read(req);
    }
 }

@@ -324,42 +363,17 @@ static void virtio_blk_handle_output(VirtIODevice *vdev, VirtQueue *vq)
 {
    VirtIOBlock *s = to_virtio_blk(vdev);
    VirtIOBlockReq *req;
-    BlockRequest blkreq[32];
-    int num_writes = 0;
-    BlockDriverState *old_bs = NULL;
+    MultiReqBuffer mrb = {
+        .num_writes = 0,
+        .old_bs = NULL,
+    };

    while ((req = virtio_blk_get_request(s))) {
-        if (req->elem.out_num < 1 || req->elem.in_num < 1) {
-            fprintf(stderr, "virtio-blk missing headers\n");
-            exit(1);
-        }
-
-        if (req->elem.out_sg[0].iov_len < sizeof(*req->out) ||
-            req->elem.in_sg[req->elem.in_num - 1].iov_len < sizeof(*req->in)) {
-            fprintf(stderr, "virtio-blk header not in correct element\n");
-            exit(1);
-        }
-
-        req->out = (void *)req->elem.out_sg[0].iov_base;
-        req->in = (void *)req->elem.in_sg[req->elem.in_num - 1].iov_base;
-
-        if (req->out->type & VIRTIO_BLK_T_FLUSH) {
-            virtio_blk_handle_flush(req);
-        } else if (req->out->type & VIRTIO_BLK_T_SCSI_CMD) {
-            virtio_blk_handle_scsi(req);
-        } else if (req->out->type & VIRTIO_BLK_T_OUT) {
-            qemu_iovec_init_external(&req->qiov, &req->elem.out_sg[1],
-                                     req->elem.out_num - 1);
-            virtio_blk_handle_write(blkreq, &num_writes, req, &old_bs);
-        } else {
-            qemu_iovec_init_external(&req->qiov, &req->elem.in_sg[0],
-                                     req->elem.in_num - 1);
-            virtio_blk_handle_read(req);
-        }
+        virtio_blk_handle_request(req, &mrb);
    }

-    if (num_writes > 0) {
-        do_multiwrite(old_bs, blkreq, num_writes);
+    if (mrb.num_writes > 0) {
+        do_multiwrite(mrb.old_bs, mrb.blkreq, mrb.num_writes);
    }

    /*
@@ -373,6 +387,10 @@ static void virtio_blk_dma_restart_bh(void *opaque)
 {
    VirtIOBlock *s = opaque;
    VirtIOBlockReq *req = s->rq;
+    MultiReqBuffer mrb = {
+        .num_writes = 0,
+        .old_bs = NULL,
+    };

    qemu_bh_delete(s->bh);
    s->bh = NULL;
@@ -380,10 +398,13 @@ static void virtio_blk_dma_restart_bh(void *opaque)
    s->rq = NULL;

    while (req) {
-        bdrv_aio_writev(req->dev->bs, req->out->sector, &req->qiov,
-            req->qiov.size / 512, virtio_blk_rw_complete, req);
+        virtio_blk_handle_request(req, &mrb);
        req = req->next;
    }
+
+    if (mrb.num_writes > 0) {
+        do_multiwrite(mrb.old_bs, mrb.blkreq, mrb.num_writes);
+    }
 }

 static void virtio_blk_dma_restart_cb(void *opaque, int running, int reason)
@@ -384,7 +384,15 @@ static int virtio_net_has_buffers(VirtIONet *n, int bufsize)
        (n->mergeable_rx_bufs &&
         !virtqueue_avail_bytes(n->rx_vq, bufsize, 0))) {
        virtio_queue_set_notification(n->rx_vq, 1);
-        return 0;
+
+        /* To avoid a race condition where the guest has made some buffers
+         * available after the above check but before notification was
+         * enabled, check for available buffers again.
+         */
+        if (virtio_queue_empty(n->rx_vq) ||
+            (n->mergeable_rx_bufs &&
+             !virtqueue_avail_bytes(n->rx_vq, bufsize, 0)))
+            return 0;
    }

    virtio_queue_set_notification(n->rx_vq, 0);
@@ -372,7 +372,7 @@ static void virtio_write_config(PCIDevice *pci_dev, uint32_t address,

    if (PCI_COMMAND == address) {
        if (!(val & PCI_COMMAND_MASTER)) {
-            proxy->vdev->status &= !VIRTIO_CONFIG_S_DRIVER_OK;
+            proxy->vdev->status &= ~VIRTIO_CONFIG_S_DRIVER_OK;
        }
    }

@@ -562,6 +562,13 @@ static void vmsvga_fifo_run(struct vmsvga_state_s *s)
            cursor.height = y = vmsvga_fifo_read(s);
            vmsvga_fifo_read(s);
            cursor.bpp = vmsvga_fifo_read(s);
+
+	    if (SVGA_BITMAP_SIZE(x, y) > sizeof cursor.mask ||
+		SVGA_PIXMAP_SIZE(x, y, cursor.bpp) > sizeof cursor.image) {
+		    args = SVGA_BITMAP_SIZE(x, y) + SVGA_PIXMAP_SIZE(x, y, cursor.bpp);
+		    goto badcmd;
+	    }
+
            for (args = 0; args < SVGA_BITMAP_SIZE(x, y); args ++)
                cursor.mask[args] = vmsvga_fifo_read_raw(s);
            for (args = 0; args < SVGA_PIXMAP_SIZE(x, y, cursor.bpp); args ++)
@@ -54,6 +54,9 @@ enum json_lexer_state {
    IN_ESCAPE,
    IN_ESCAPE_L,
    IN_ESCAPE_LL,
+    IN_ESCAPE_I,
+    IN_ESCAPE_I6,
+    IN_ESCAPE_I64,
    IN_ESCAPE_DONE,
    IN_WHITESPACE,
    IN_OPERATOR_DONE,
@@ -223,6 +226,18 @@ static const uint8_t json_lexer[][256] =  {
        ['l'] = IN_ESCAPE_LL,
    },

+    [IN_ESCAPE_I64] = {
+        ['d'] = IN_ESCAPE_DONE,
+    },
+
+    [IN_ESCAPE_I6] = {
+        ['4'] = IN_ESCAPE_I64,
+    },
+
+    [IN_ESCAPE_I] = {
+        ['6'] = IN_ESCAPE_I6,
+    },
+
    [IN_ESCAPE] = {
        ['d'] = IN_ESCAPE_DONE,
        ['i'] = IN_ESCAPE_DONE,
@@ -230,6 +245,7 @@ static const uint8_t json_lexer[][256] =  {
        ['s'] = IN_ESCAPE_DONE,
        ['f'] = IN_ESCAPE_DONE,
        ['l'] = IN_ESCAPE_L,
+        ['I'] = IN_ESCAPE_I,
    },

    /* top level rule */
@@ -266,7 +266,7 @@ static int parse_pair(JSONParserContext *ctxt, QDict *dict, QList **tokens, va_l

    peek = qlist_peek(working);
    key = parse_value(ctxt, &working, ap);
-    if (qobject_type(key) != QTYPE_QSTRING) {
+    if (!key || qobject_type(key) != QTYPE_QSTRING) {
        parse_error(ctxt, peek, "key is not a string in object");
        goto out;
    }
@@ -476,7 +476,8 @@ static QObject *parse_escape(JSONParserContext *ctxt, QList **tokens, va_list *a
        obj = QOBJECT(qint_from_int(va_arg(*ap, int)));
    } else if (token_is_escape(token, "%ld")) {
        obj = QOBJECT(qint_from_int(va_arg(*ap, long)));
-    } else if (token_is_escape(token, "%lld")) {
+    } else if (token_is_escape(token, "%lld") ||
+               token_is_escape(token, "%I64d")) {
        obj = QOBJECT(qint_from_int(va_arg(*ap, long long)));
    } else if (token_is_escape(token, "%s")) {
        obj = QOBJECT(qstring_from_str(va_arg(*ap, const char *)));
@@ -2684,7 +2684,7 @@ int main(int argc, char **argv, char **envp)
 #endif
 #elif defined(TARGET_PPC)
 #ifdef TARGET_PPC64
-        cpu_model = "970";
+        cpu_model = "970fx";
 #else
        cpu_model = "750";
 #endif
@@ -283,7 +283,6 @@ static void monitor_protocol_emitter(Monitor *mon, QObject *data)
    if (!monitor_has_error(mon)) {
        /* success response */
        if (data) {
-            assert(qobject_type(data) == QTYPE_QDICT);
            qobject_incref(data);
            qdict_put_obj(qmp, "return", data);
        } else {
@@ -335,13 +334,10 @@ void monitor_protocol_event(MonitorEvent event, QObject *data)
 {
    QDict *qmp;
    const char *event_name;
-    Monitor *mon = cur_mon;
+    Monitor *mon;

    assert(event < QEVENT_MAX);

-    if (!monitor_ctrl_mode(mon))
-        return;
-
    switch (event) {
        case QEVENT_DEBUG:
            event_name = "DEBUG";
@@ -366,10 +362,16 @@ void monitor_protocol_event(MonitorEvent event, QObject *data)
    qmp = qdict_new();
    timestamp_put(qmp);
    qdict_put(qmp, "event", qstring_from_str(event_name));
-    if (data)
+    if (data) {
+        qobject_incref(data);
        qdict_put_obj(qmp, "data", data);
+    }

-    monitor_json_emitter(mon, QOBJECT(qmp));
+    QLIST_FOREACH(mon, &mon_list, entry) {
+        if (monitor_ctrl_mode(mon)) {
+            monitor_json_emitter(mon, QOBJECT(qmp));
+        }
+    }
    QDECREF(qmp);
 }

@@ -96,7 +96,7 @@ int parse_host_src_port(struct sockaddr_in *haddr,
                        struct sockaddr_in *saddr,
                        const char *input_str)
 {
-    char *str = strdup(input_str);
+    char *str = qemu_strdup(input_str);
    char *host_str = str;
    char *src_str;
    const char *src_str2;
@@ -812,9 +812,6 @@ static int net_init_nic(QemuOpts *opts,
    }

    nd->used = 1;
-    if (vlan) {
-        nd->vlan->nb_guest_devs++;
-    }
    nb_nics++;

    return idx;
@@ -1128,20 +1125,6 @@ int net_client_init(Monitor *mon, QemuOpts *opts, int is_netdev)
    return -1;
 }

-void net_client_uninit(NICInfo *nd)
-{
-    if (nd->vlan) {
-        nd->vlan->nb_guest_devs--;
-    }
-    nb_nics--;
-
-    qemu_free(nd->model);
-    qemu_free(nd->name);
-    qemu_free(nd->devaddr);
-
-    nd->used = 0;
-}
-
 static int net_host_check_device(const char *device)
 {
    int i;
@@ -1204,39 +1187,26 @@ void net_host_device_remove(Monitor *mon, const QDict *qdict)
    qemu_del_vlan_client(vc);
 }

-void net_set_boot_mask(int net_boot_mask)
-{
-    int i;
-
-    /* Only the first four NICs may be bootable */
-    net_boot_mask = net_boot_mask & 0xF;
-
-    for (i = 0; i < nb_nics; i++) {
-        if (net_boot_mask & (1 << i)) {
-            nd_table[i].bootable = 1;
-            net_boot_mask &= ~(1 << i);
-        }
-    }
-
-    if (net_boot_mask) {
-        fprintf(stderr, "Cannot boot from non-existent NIC\n");
-        exit(1);
-    }
-}
-
 void do_info_network(Monitor *mon)
 {
    VLANState *vlan;
+    VLANClientState *vc;

    QTAILQ_FOREACH(vlan, &vlans, next) {
-        VLANClientState *vc;
-
        monitor_printf(mon, "VLAN %d devices:\n", vlan->id);

        QTAILQ_FOREACH(vc, &vlan->clients, next) {
            monitor_printf(mon, "  %s: %s\n", vc->name, vc->info_str);
        }
    }
+    monitor_printf(mon, "Devices not on any VLAN:\n");
+    QTAILQ_FOREACH(vc, &non_vlan_clients, next) {
+        monitor_printf(mon, "  %s: %s", vc->name, vc->info_str);
+        if (vc->peer) {
+            monitor_printf(mon, " peer=%s", vc->peer->name);
+        }
+        monitor_printf(mon, "\n");
+    }
 }

 void do_set_link(Monitor *mon, const QDict *qdict)
@@ -1253,6 +1223,7 @@ void do_set_link(Monitor *mon, const QDict *qdict)
            }
        }
    }
+    vc = qemu_find_netdev(name);
 done:

    if (!vc) {
@@ -1289,20 +1260,41 @@ void net_cleanup(void)
    }
 }

-static void net_check_clients(void)
+void net_check_clients(void)
 {
    VLANState *vlan;
+    VLANClientState *vc;
+    int has_nic, has_host_dev;

    QTAILQ_FOREACH(vlan, &vlans, next) {
-        if (vlan->nb_guest_devs == 0 && vlan->nb_host_devs == 0)
-            continue;
-        if (vlan->nb_guest_devs == 0)
+        QTAILQ_FOREACH(vc, &vlan->clients, next) {
+            switch (vc->info->type) {
+            case NET_CLIENT_TYPE_NIC:
+                has_nic = 1;
+                break;
+            case NET_CLIENT_TYPE_SLIRP:
+            case NET_CLIENT_TYPE_TAP:
+            case NET_CLIENT_TYPE_SOCKET:
+            case NET_CLIENT_TYPE_VDE:
+                has_host_dev = 1;
+                break;
+            default: ;
+            }
+        }
+        if (has_host_dev && !has_nic)
            fprintf(stderr, "Warning: vlan %d with no nics\n", vlan->id);
-        if (vlan->nb_host_devs == 0)
+        if (has_nic && !has_host_dev)
            fprintf(stderr,
                    "Warning: vlan %d is not connected to host network\n",
                    vlan->id);
    }
+    QTAILQ_FOREACH(vc, &non_vlan_clients, next) {
+        if (!vc->peer) {
+            fprintf(stderr, "Warning: %s %s has no peer\n",
+                    vc->info->type == NET_CLIENT_TYPE_NIC ? "nic" : "netdev",
+                    vc->name);
+        }
+    }
 }

 static int net_init_client(QemuOpts *opts, void *dummy)
@@ -1337,8 +1329,6 @@ int net_init_clients(void)
        return -1;
    }

-    net_check_clients();
-
    return 0;
 }

@@ -76,7 +76,6 @@ struct VLANState {
    int id;
    QTAILQ_HEAD(, VLANClientState) clients;
    QTAILQ_ENTRY(VLANState) next;
-    unsigned int nb_guest_devs, nb_host_devs;
    NetQueue *send_queue;
 };

@@ -160,11 +159,10 @@ extern const char *legacy_tftp_prefix;
 extern const char *legacy_bootp_filename;

 int net_client_init(Monitor *mon, QemuOpts *opts, int is_netdev);
-void net_client_uninit(NICInfo *nd);
 int net_client_parse(QemuOptsList *opts_list, const char *str);
 int net_init_clients(void);
+void net_check_clients(void);
 void net_cleanup(void);
-void net_set_boot_mask(int boot_mask);
 void net_host_device_add(Monitor *mon, const QDict *qdict);
 void net_host_device_remove(Monitor *mon, const QDict *qdict);

@@ -728,10 +728,6 @@ int net_init_slirp(QemuOpts *opts,
        qemu_free(config);
    }

-    if (ret != -1 && vlan) {
-        vlan->nb_host_devs++;
-    }
-
    qemu_free(vnet);

    return ret;
@@ -569,9 +569,5 @@ int net_init_socket(QemuOpts *opts,
        return -1;
    }

-    if (vlan) {
-        vlan->nb_host_devs++;
-    }
-
    return 0;
 }
@@ -714,10 +714,6 @@ int net_init_tap(QemuOpts *opts, Monitor *mon, const char *name, VLANState *vlan
        return -1;
    }

-    if (vlan) {
-        vlan->nb_host_devs++;
-    }
-
    return 0;
 }

@@ -441,9 +441,5 @@ int net_init_tap(QemuOpts *opts, Monitor *mon, const char *name, VLANState *vlan
        }
    }

-    if (vlan) {
-        vlan->nb_host_devs++;
-    }
-
    return 0;
 }
@@ -127,9 +127,5 @@ int net_init_vde(QemuOpts *opts, Monitor *mon, const char *name, VLANState *vlan
        return -1;
    }

-    if (vlan) {
-        vlan->nb_host_devs++;
-    }
-
    return 0;
 }
@@ -306,7 +306,7 @@ int qemu_accept(int s, struct sockaddr *addr, socklen_t *addrlen)

 #ifdef CONFIG_ACCEPT4
    ret = accept4(s, addr, addrlen, SOCK_CLOEXEC);
-    if (ret != -1 || errno != EINVAL) {
+    if (ret != -1 || errno != ENOSYS) {
        return ret;
    }
 #endif
@@ -1,4 +1,5 @@
- The PC BIOS comes from the Bochs project (http://bochs.sourceforge.net/).
+- SeaBIOS (bios.bin) is the successor of pc bios.
+  See http://www.seabios.org/ for more information.

 - The VGA BIOS and the Cirrus VGA BIOS come from the LGPL VGA bios
  project (http://www.nongnu.org/vgabios/).
@@ -86,9 +86,10 @@ copy_kernel:
 	/* Now create the GDT descriptor */
 	movw		$((3 * 8) - 1), -16(%bp)
 	mov		%cs, %eax
+	movzwl		%ax, %eax
 	shl		$4, %eax
-	addl		$gdt, %ebx
-	movl		%ebx, -14(%bp)
+	addl		$gdt, %eax
+	movl		%eax, -14(%bp)

 	/* And load the GDT */
 	data32 lgdt	-16(%bp)
@@ -1180,6 +1180,9 @@ static CharDriverState *qemu_chr_open_tty(QemuOpts *opts)
    int fd;

    TFR(fd = open(filename, O_RDWR | O_NONBLOCK));
+    if (fd < 0) {
+        return NULL;
+    }
    tty_serial_init(fd, 115200, 'N', 8, 1);
    chr = qemu_chr_open_fd(fd, fd);
    if (!chr) {
@@ -2314,8 +2317,7 @@ QemuOpts *qemu_chr_parse_compat(const char *label, const char *filename)
        qemu_opt_set(opts, "backend", "udp");
        if (sscanf(p, "%64[^:]:%32[^@,]%n", host, port, &pos) < 2) {
            host[0] = 0;
-            if (sscanf(p, ":%32[^,]%n", port, &pos) < 1) {
-                fprintf(stderr, "udp #1\n");
+            if (sscanf(p, ":%32[^@,]%n", port, &pos) < 1) {
                goto fail;
            }
        }
@@ -2326,7 +2328,6 @@ QemuOpts *qemu_chr_parse_compat(const char *label, const char *filename)
            if (sscanf(p, "%64[^:]:%32[^,]%n", host, port, &pos) < 2) {
                host[0] = 0;
                if (sscanf(p, ":%32[^,]%n", port, &pos) < 1) {
-                    fprintf(stderr, "udp #2\n");
                    goto fail;
                }
            }
@@ -2354,7 +2355,6 @@ QemuOpts *qemu_chr_parse_compat(const char *label, const char *filename)
    }

 fail:
-    fprintf(stderr, "%s: fail on \"%s\"\n", __FUNCTION__, filename);
    qemu_opts_del(opts);
    return NULL;
 }
@@ -2,6 +2,8 @@
 #ifndef QEMU_COMMON_H
 #define QEMU_COMMON_H

+#include "config-host.h"
+
 #define QEMU_NORETURN __attribute__ ((__noreturn__))
 #ifdef CONFIG_GCC_ATTRIBUTE_WARN_UNUSED_RESULT
 #define QEMU_WARN_UNUSED_RESULT __attribute__((warn_unused_result))
@@ -29,7 +31,6 @@
 #include <fcntl.h>
 #include <sys/stat.h>
 #include <assert.h>
-#include "config-host.h"

 #ifndef O_LARGEFILE
 #define O_LARGEFILE 0
@@ -53,6 +54,10 @@ struct iovec {
    void *iov_base;
    size_t iov_len;
 };
+/*
+ * Use the same value as Linux for now.
+ */
+#define IOV_MAX		1024
 #else
 #include <sys/uio.h>
 #endif
@@ -467,7 +467,7 @@ int main(int argc, char **argv)
            }
        }
    } while (persistent || nb_fds > 1);
-    qemu_free(data);
+    qemu_vfree(data);

    close(sharing_fds[0]);
    bdrv_close(bs);
@@ -404,6 +404,12 @@ ETEXI

 DEF("device", HAS_ARG, QEMU_OPTION_device,
    "-device driver[,options]  add device\n")
+STEXI
+@item -device @var{driver}[,@var{option}[,...]]
+Add device @var{driver}. Depending on the device type,
+@var{option} (typically @var{key}=@var{value}) may be useful.
+ETEXI
+
 DEF("name", HAS_ARG, QEMU_OPTION_name,
    "-name string1[,process=string2]    set the name of the guest\n"
    "            string1 sets the window title and string2 the process name (on Linux)\n")
@@ -962,7 +968,7 @@ the guest, use the following:

@example
 # on the host
-qemu -net user,hostfwd=tcp:5555::23 [...]
+qemu -net user,hostfwd=tcp::5555-:23 [...]
 telnet localhost 5555
@end example

@@ -1893,6 +1899,8 @@ ETEXI
 DEF("nodefaults", 0, QEMU_OPTION_nodefaults, \
    "-nodefaults     don't create default devices.\n")
 STEXI
+@item -nodefaults
+Don't create default devices.
 ETEXI

 #ifndef _WIN32
@@ -1915,10 +1923,6 @@ Immediately before starting guest execution, drop root privileges, switching
 to the specified user.
 ETEXI

-STEXI
-@end table
-ETEXI
-
 #if defined(TARGET_SPARC) || defined(TARGET_PPC)
 DEF("prom-env", HAS_ARG, QEMU_OPTION_prom_env,
    "-prom-env variable=value\n"
@@ -1934,6 +1938,19 @@ DEF("old-param", 0, QEMU_OPTION_old_param,
 #endif
 DEF("readconfig", HAS_ARG, QEMU_OPTION_readconfig,
    "-readconfig <file>\n")
+STEXI
+@item -readconfig @var{file}
+Read device configuration from @var{file}.
+ETEXI
 DEF("writeconfig", HAS_ARG, QEMU_OPTION_writeconfig,
    "-writeconfig <file>\n"
-    "                read/write config file")
+    "                read/write config file\n")
+STEXI
+@item -writeconfig @var{file}
+Write device configuration to @var{file}.
+ETEXI
+
+HXCOMM This is the last statement. Insert new options before this line!
+STEXI
+@end table
+ETEXI
@@ -424,7 +424,7 @@ static int inet_parse(QemuOpts *opts, const char *str)
                    __FUNCTION__, str);
            return -1;
        }
-        qemu_opt_set(opts, "ipv6", "yes");
+        qemu_opt_set(opts, "ipv6", "on");
    } else if (qemu_isdigit(str[0])) {
        /* IPv4 addr */
        if (2 != sscanf(str,"%64[0-9.]:%32[^,]%n",addr,port,&pos)) {
@@ -432,7 +432,7 @@ static int inet_parse(QemuOpts *opts, const char *str)
                    __FUNCTION__, str);
            return -1;
        }
-        qemu_opt_set(opts, "ipv4", "yes");
+        qemu_opt_set(opts, "ipv4", "on");
    } else {
        /* hostname */
        if (2 != sscanf(str,"%64[^:]:%32[^,]%n",addr,port,&pos)) {
@@ -450,9 +450,9 @@ static int inet_parse(QemuOpts *opts, const char *str)
    if (h)
        qemu_opt_set(opts, "to", h+4);
    if (strstr(optstr, ",ipv4"))
-        qemu_opt_set(opts, "ipv4", "yes");
+        qemu_opt_set(opts, "ipv4", "on");
    if (strstr(optstr, ",ipv6"))
-        qemu_opt_set(opts, "ipv6", "yes");
+        qemu_opt_set(opts, "ipv6", "on");
    return 0;
 }

@@ -1696,7 +1696,7 @@ void do_savevm(Monitor *mon, const QDict *qdict)
    sn->vm_clock_nsec = qemu_get_clock(vm_clock);

    /* Delete old snapshots of the same name */
-    if (del_existing_snapshots(mon, name) < 0) {
+    if (name && del_existing_snapshots(mon, name) < 0) {
        goto the_end;
    }

@@ -525,6 +525,7 @@ static void sdl_send_mouse_event(int dx, int dy, int dz, int x, int y, int state
 static void toggle_full_screen(DisplayState *ds)
 {
    gui_fullscreen = !gui_fullscreen;
+    do_sdl_resize(real_screen->w, real_screen->h, real_screen->format->BitsPerPixel);
    if (gui_fullscreen) {
        scaling_active = 0;
        gui_saved_grab = gui_grab;
@@ -179,7 +179,7 @@ fork_exec(struct socket *so, const char *ex, int do_pty)
 		   close(s);

 		i = 0;
-		bptr = strdup(ex); /* No need to free() this */
+		bptr = qemu_strdup(ex); /* No need to free() this */
 		if (do_pty == 1) {
 			/* Setup "slirp.telnetd -x" */
 			argv[i++] = "slirp.telnetd";
@@ -264,6 +264,12 @@ static void tftp_handle_rrq(Slirp *slirp, struct tftp_t *tp, int pktlen)
  size_t prefix_len;
  char *req_fname;

+  /* check if a session already exists and if so terminate it */
+  s = tftp_session_find(slirp, tp);
+  if (s >= 0) {
+    tftp_session_terminate(&slirp->tftp_sessions[s]);
+  }
+
  s = tftp_session_allocate(slirp, tp);

  if (s < 0) {
@@ -362,6 +368,7 @@ static void tftp_handle_rrq(Slirp *slirp, struct tftp_t *tp, int pktlen)
 	  }

 	  tftp_send_oack(spt, "tsize", tsize, tp);
+	  return;
      }
  }

@@ -385,6 +392,19 @@ static void tftp_handle_ack(Slirp *slirp, struct tftp_t *tp, int pktlen)
  }
 }

+static void tftp_handle_error(Slirp *slirp, struct tftp_t *tp, int pktlen)
+{
+  int s;
+
+  s = tftp_session_find(slirp, tp);
+
+  if (s < 0) {
+    return;
+  }
+
+  tftp_session_terminate(&slirp->tftp_sessions[s]);
+}
+
 void tftp_input(struct mbuf *m)
 {
  struct tftp_t *tp = (struct tftp_t *)m->m_data;
@@ -397,5 +417,9 @@ void tftp_input(struct mbuf *m)
  case TFTP_ACK:
    tftp_handle_ack(m->slirp, tp, m->m_len);
    break;
+
+  case TFTP_ERROR:
+    tftp_handle_error(m->slirp, tp, m->m_len);
+    break;
  }
 }
@@ -46,10 +46,10 @@ uint64_t helper_load_fpcr (void)
 {
    uint64_t ret = 0;
 #ifdef CONFIG_SOFTFLOAT
-    ret |= env->fp_status.float_exception_flags << 52;
+    ret |= (uint64_t)env->fp_status.float_exception_flags << 52;
    if (env->fp_status.float_exception_flags)
        ret |= 1ULL << 63;
-    env->ipr[IPR_EXC_SUM] &= ~0x3E:
+    env->ipr[IPR_EXC_SUM] &= ~0x3E;
    env->ipr[IPR_EXC_SUM] |= env->fp_status.float_exception_flags << 1;
 #endif
    switch (env->fp_status.float_rounding_mode) {
@@ -617,6 +617,8 @@ void cpu_reset(CPUX86State *env)
    env->dr[7] = DR7_FIXED_1;
    cpu_breakpoint_remove_all(env, BP_CPU);
    cpu_watchpoint_remove_all(env, BP_CPU);
+
+    env->mcg_status = 0;
 }

 void cpu_x86_close(CPUX86State *env)
@@ -227,6 +227,7 @@ int kvm_arch_init_vcpu(CPUState *env)

 void kvm_arch_reset_vcpu(CPUState *env)
 {
+    env->exception_injected = -1;
    env->interrupt_injected = -1;
    env->nmi_injected = 0;
    env->nmi_pending = 0;
@@ -2047,8 +2047,8 @@ static void gen_lea_modrm(DisasContext *s, int modrm, int *reg_ptr, int *offset_
                gen_op_movl_A0_im(disp);
            }
        }
-        /* XXX: index == 4 is always invalid */
-        if (havesib && (index != 4 || scale != 0)) {
+        /* index == 4 means no index */
+        if (havesib && (index != 4)) {
 #ifdef TARGET_X86_64
            if (s->aflag == 2) {
                gen_op_addq_A0_reg_sN(scale, index);
@@ -3165,12 +3165,17 @@ static void gen_sse(DisasContext *s, int b, target_ulong pc_start, int rex_r)
        case 0x1e7: /* movntdq */
        case 0x02b: /* movntps */
        case 0x12b: /* movntps */
-        case 0x3f0: /* lddqu */
            if (mod == 3)
                goto illegal_op;
            gen_lea_modrm(s, modrm, &reg_addr, &offset_addr);
            gen_sto_env_A0(s->mem_index, offsetof(CPUX86State,xmm_regs[reg]));
            break;
+        case 0x3f0: /* lddqu */
+            if (mod == 3)
+                goto illegal_op;
+            gen_lea_modrm(s, modrm, &reg_addr, &offset_addr);
+            gen_ldo_env_A0(s->mem_index, offsetof(CPUX86State,xmm_regs[reg]));
+            break;
        case 0x22b: /* movntss */
        case 0x32b: /* movntsd */
            if (mod == 3)
@@ -4591,9 +4596,7 @@ static target_ulong disas_insn(DisasContext *s, target_ulong pc_start)
                /* operand size for jumps is 64 bit */
                ot = OT_QUAD;
            } else if (op == 3 || op == 5) {
-                /* for call calls, the operand is 16 or 32 bit, even
-                   in long mode */
-                ot = dflag ? OT_LONG : OT_WORD;
+                ot = dflag ? OT_LONG + (rex_w == 1) : OT_WORD;
            } else if (op == 6) {
                /* default push size is 64 bit */
                ot = dflag ? OT_QUAD : OT_WORD;
@@ -6259,6 +6262,8 @@ static target_ulong disas_insn(DisasContext *s, target_ulong pc_start)
            tval += next_eip;
            if (s->dflag == 0)
                tval &= 0xffff;
+            else if(!CODE64(s))
+                tval &= 0xffffffff;
            gen_movtl_T0_im(next_eip);
            gen_push_T0(s);
            gen_jmp(s, tval);
@@ -590,8 +590,8 @@ int cpu_mips_handle_mmu_fault (CPUState *env, target_ulong address, int rw,
 #define cpu_handle_mmu_fault cpu_mips_handle_mmu_fault
 void do_interrupt (CPUState *env);
 void r4k_invalidate_tlb (CPUState *env, int idx, int use_extra);
-target_phys_addr_t do_translate_address (CPUState *env, target_ulong address,
-		                         int rw);
+target_phys_addr_t cpu_mips_translate_address (CPUState *env, target_ulong address,
+		                               int rw);

 static inline void cpu_pc_from_tb(CPUState *env, TranslationBlock *tb)
 {
@@ -311,7 +311,7 @@ int cpu_mips_handle_mmu_fault (CPUState *env, target_ulong address, int rw,
 }

 #if !defined(CONFIG_USER_ONLY)
-target_phys_addr_t do_translate_address(CPUState *env, target_ulong address, int rw)
+target_phys_addr_t cpu_mips_translate_address(CPUState *env, target_ulong address, int rw)
 {
    target_phys_addr_t physical;
    int prot;
@@ -326,10 +326,10 @@ target_phys_addr_t do_translate_address(CPUState *env, target_ulong address, int
                               address, rw, access_type);
    if (ret != TLBRET_MATCH) {
        raise_mmu_exception(env, address, rw, ret);
-        cpu_loop_exit();
+        return -1LL;
+    } else {
+        return physical;
    }
-
-    return physical;
 }
 #endif

@@ -276,10 +276,24 @@ void helper_dmultu (target_ulong arg1, target_ulong arg2)
 #endif

 #ifndef CONFIG_USER_ONLY
+
+static inline target_phys_addr_t do_translate_address(target_ulong address, int rw)
+{
+    target_phys_addr_t lladdr;
+
+    lladdr = cpu_mips_translate_address(env, address, rw);
+
+    if (lladdr == -1LL) {
+        cpu_loop_exit();
+    } else {
+        return lladdr;
+    }
+}
+
 #define HELPER_LD_ATOMIC(name, insn)                                          \
 target_ulong helper_##name(target_ulong arg, int mem_idx)                     \
 {                                                                             \
-    env->lladdr = do_translate_address(env, arg, 0);                          \
+    env->lladdr = do_translate_address(arg, 0);                               \
    env->llval = do_##insn(arg, mem_idx);                                     \
    return env->llval;                                                        \
 }
@@ -298,7 +312,7 @@ target_ulong helper_##name(target_ulong arg1, target_ulong arg2, int mem_idx) \
        env->CP0_BadVAddr = arg2;                                             \
        helper_raise_exception(EXCP_AdES);                                    \
    }                                                                         \
-    if (do_translate_address(env, arg2, 1) == env->lladdr) {                  \
+    if (do_translate_address(arg2, 1) == env->lladdr) {                       \
        tmp = do_##ld_insn(arg2, mem_idx);                                    \
        if (tmp == env->llval) {                                              \
            do_##st_insn(arg2, arg1, mem_idx);                                \
@@ -821,7 +821,7 @@ static inline void gen_op_addr_add (DisasContext *ctx, TCGv ret, TCGv arg0, TCGv
 static inline void check_cp0_enabled(DisasContext *ctx)
 {
    if (unlikely(!(ctx->hflags & MIPS_HFLAG_CP0)))
-        generate_exception_err(ctx, EXCP_CpU, 1);
+        generate_exception_err(ctx, EXCP_CpU, 0);
 }

 static inline void check_cp1_enabled(DisasContext *ctx)
@@ -1451,6 +1451,8 @@ static void gen_shift_imm(CPUState *env, DisasContext *ctx, uint32_t opc,
                    tcg_gen_rotri_i32(t1, t1, uimm);
                    tcg_gen_ext_i32_tl(cpu_gpr[rt], t1);
                    tcg_temp_free_i32(t1);
+                } else {
+                    tcg_gen_ext32s_tl(cpu_gpr[rt], t0);
                }
                opn = "rotr";
            } else {
@@ -1489,6 +1491,8 @@ static void gen_shift_imm(CPUState *env, DisasContext *ctx, uint32_t opc,
            if (env->insn_flags & ISA_MIPS32R2) {
                if (uimm != 0) {
                    tcg_gen_rotri_tl(cpu_gpr[rt], t0, uimm);
+                } else {
+                    tcg_gen_mov_tl(cpu_gpr[rt], t0);
                }
                opn = "drotr";
            } else {
@@ -167,6 +167,7 @@ int cpu_sh4_handle_mmu_fault(CPUSH4State * env, target_ulong address, int rw,
 void do_interrupt(CPUSH4State * env);

 void sh4_cpu_list(FILE *f, int (*cpu_fprintf)(FILE *f, const char *fmt, ...));
+void cpu_sh4_invalidate_tlb(CPUSH4State *s);
 void cpu_sh4_write_mmaped_utlb_addr(CPUSH4State *s, target_phys_addr_t addr,
 				    uint32_t mem_value);

@@ -222,6 +223,7 @@ enum {
 /* MMU control register */
 #define MMUCR    0x1F000010
 #define MMUCR_AT (1<<0)
+#define MMUCR_TI (1<<2)
 #define MMUCR_SV (1<<8)
 #define MMUCR_URC_BITS (6)
 #define MMUCR_URC_OFFSET (10)
@@ -377,7 +377,7 @@ static int get_mmu_address(CPUState * env, target_ulong * physical,
 	n = find_itlb_entry(env, address, use_asid, 1);
 	if (n >= 0) {
 	    matching = &env->itlb[n];
-	    if ((env->sr & SR_MD) & !(matching->pr & 2))
+	    if (!(env->sr & SR_MD) && !(matching->pr & 2))
 		n = MMU_ITLB_VIOLATION;
 	    else
 		*prot = PAGE_READ;
@@ -430,7 +430,7 @@ static int get_physical_address(CPUState * env, target_ulong * physical,
    if ((address >= 0x80000000 && address < 0xc0000000) ||
 	address >= 0xe0000000) {
 	if (!(env->sr & SR_MD)
-	    && (address < 0xe0000000 || address > 0xe4000000)) {
+	    && (address < 0xe0000000 || address >= 0xe4000000)) {
 	    /* Unauthorized access in user mode (only store queues are available) */
 	    fprintf(stderr, "Unauthorized access\n");
 	    if (rw == 0)
@@ -574,6 +574,24 @@ void cpu_load_tlb(CPUSH4State * env)
    entry->tc   = (uint8_t)cpu_ptea_tc(env->ptea);
 }

+ void cpu_sh4_invalidate_tlb(CPUSH4State *s)
+{
+    int i;
+
+    /* UTLB */
+    for (i = 0; i < UTLB_SIZE; i++) {
+        tlb_t * entry = &s->utlb[i];
+        entry->v = 0;
+    }
+    /* ITLB */
+    for (i = 0; i < UTLB_SIZE; i++) {
+        tlb_t * entry = &s->utlb[i];
+        entry->v = 0;
+    }
+
+    tlb_flush(s, 1);
+}
+
 void cpu_sh4_write_mmaped_utlb_addr(CPUSH4State *s, target_phys_addr_t addr,
 				    uint32_t mem_value)
 {
@@ -1905,7 +1905,7 @@ gen_intermediate_code_internal(CPUState * env, TranslationBlock * tb,
    ctx.bstate = BS_NONE;
    ctx.sr = env->sr;
    ctx.fpscr = env->fpscr;
-    ctx.memidx = (env->sr & SR_MD) ? 1 : 0;
+    ctx.memidx = (env->sr & SR_MD) == 0 ? 1 : 0;
    /* We don't know if the delayed pc came from a dynamic or static branch,
       so assume it is a dynamic branch.  */
    ctx.delayed_pc = -1; /* use delayed pc from env pointer */
@@ -1697,12 +1697,15 @@ static inline void tcg_out_movi(TCGContext *s, TCGType type,

 void tcg_target_qemu_prologue(TCGContext *s)
 {
-    /* stmdb sp!, { r9 - r11, lr } */
-    tcg_out32(s, (COND_AL << 28) | 0x092d4e00);
+    /* Theoretically there is no need to save r12, but an
+       even number of registers to be saved as per EABI */
+
+    /* stmdb sp!, { r4 - r12, lr } */
+    tcg_out32(s, (COND_AL << 28) | 0x092d5ff0);

    tcg_out_bx(s, COND_AL, TCG_REG_R0);
    tb_ret_addr = s->code_ptr;

-    /* ldmia sp!, { r9 - r11, pc } */
-    tcg_out32(s, (COND_AL << 28) | 0x08bd8e00);
+    /* ldmia sp!, { r4 - r12, pc } */
+    tcg_out32(s, (COND_AL << 28) | 0x08bd9ff0);
 }
@@ -339,6 +339,17 @@ static inline void tcg_out_opc_imm(TCGContext *s, int opc, int rt, int rs, int i
    tcg_out32(s, inst);
 }

+/*
+ * Type branch
+ */
+static inline void tcg_out_opc_br(TCGContext *s, int opc, int rt, int rs)
+{
+    /* We need to keep the offset unchanged for retranslation */
+    uint16_t offset = (uint16_t)(*(uint32_t *) &s->code_ptr);
+
+    tcg_out_opc_imm(s, opc, rt, rs, offset);
+}
+
 /*
 * Type sa
 */
@@ -469,42 +480,42 @@ static void tcg_out_brcond(TCGContext *s, int cond, int arg1,

    switch (cond) {
    case TCG_COND_EQ:
-        tcg_out_opc_imm(s, OPC_BEQ, arg1, arg2, 0);
+        tcg_out_opc_br(s, OPC_BEQ, arg1, arg2);
        break;
    case TCG_COND_NE:
-        tcg_out_opc_imm(s, OPC_BNE, arg1, arg2, 0);
+        tcg_out_opc_br(s, OPC_BNE, arg1, arg2);
        break;
    case TCG_COND_LT:
        tcg_out_opc_reg(s, OPC_SLT, TCG_REG_AT, arg1, arg2);
-        tcg_out_opc_imm(s, OPC_BNE, TCG_REG_AT, TCG_REG_ZERO, 0);
+        tcg_out_opc_br(s, OPC_BNE, TCG_REG_AT, TCG_REG_ZERO);
        break;
    case TCG_COND_LTU:
        tcg_out_opc_reg(s, OPC_SLTU, TCG_REG_AT, arg1, arg2);
-        tcg_out_opc_imm(s, OPC_BNE, TCG_REG_AT, TCG_REG_ZERO, 0);
+        tcg_out_opc_br(s, OPC_BNE, TCG_REG_AT, TCG_REG_ZERO);
        break;
    case TCG_COND_GE:
        tcg_out_opc_reg(s, OPC_SLT, TCG_REG_AT, arg1, arg2);
-        tcg_out_opc_imm(s, OPC_BEQ, TCG_REG_AT, TCG_REG_ZERO, 0);
+        tcg_out_opc_br(s, OPC_BEQ, TCG_REG_AT, TCG_REG_ZERO);
        break;
    case TCG_COND_GEU:
        tcg_out_opc_reg(s, OPC_SLTU, TCG_REG_AT, arg1, arg2);
-        tcg_out_opc_imm(s, OPC_BEQ, TCG_REG_AT, TCG_REG_ZERO, 0);
+        tcg_out_opc_br(s, OPC_BEQ, TCG_REG_AT, TCG_REG_ZERO);
        break;
    case TCG_COND_LE:
        tcg_out_opc_reg(s, OPC_SLT, TCG_REG_AT, arg2, arg1);
-        tcg_out_opc_imm(s, OPC_BEQ, TCG_REG_AT, TCG_REG_ZERO, 0);
+        tcg_out_opc_br(s, OPC_BEQ, TCG_REG_AT, TCG_REG_ZERO);
        break;
    case TCG_COND_LEU:
        tcg_out_opc_reg(s, OPC_SLTU, TCG_REG_AT, arg2, arg1);
-        tcg_out_opc_imm(s, OPC_BEQ, TCG_REG_AT, TCG_REG_ZERO, 0);
+        tcg_out_opc_br(s, OPC_BEQ, TCG_REG_AT, TCG_REG_ZERO);
        break;
    case TCG_COND_GT:
        tcg_out_opc_reg(s, OPC_SLT, TCG_REG_AT, arg2, arg1);
-        tcg_out_opc_imm(s, OPC_BNE, TCG_REG_AT, TCG_REG_ZERO, 0);
+        tcg_out_opc_br(s, OPC_BNE, TCG_REG_AT, TCG_REG_ZERO);
        break;
    case TCG_COND_GTU:
        tcg_out_opc_reg(s, OPC_SLTU, TCG_REG_AT, arg2, arg1);
-        tcg_out_opc_imm(s, OPC_BNE, TCG_REG_AT, TCG_REG_ZERO, 0);
+        tcg_out_opc_br(s, OPC_BNE, TCG_REG_AT, TCG_REG_ZERO);
        break;
    default:
        tcg_abort();
@@ -553,7 +564,7 @@ static void tcg_out_brcond2(TCGContext *s, int cond, int arg1,
    }

    label_ptr = s->code_ptr;
-    tcg_out_opc_imm(s, OPC_BNE, arg2, arg4, 0);
+    tcg_out_opc_br(s, OPC_BNE, arg2, arg4);
    tcg_out_nop(s);

    switch(cond) {
@@ -670,20 +681,20 @@ static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args,

 # if TARGET_LONG_BITS == 64
    label3_ptr = s->code_ptr;
-    tcg_out_opc_imm(s, OPC_BNE, TCG_REG_T0, TCG_REG_AT, 0);
+    tcg_out_opc_br(s, OPC_BNE, TCG_REG_T0, TCG_REG_AT);
    tcg_out_nop(s);

    tcg_out_opc_imm(s, OPC_LW, TCG_REG_AT, TCG_REG_A0,
                    offsetof(CPUState, tlb_table[mem_index][0].addr_read) + addr_memh);

    label1_ptr = s->code_ptr;
-    tcg_out_opc_imm(s, OPC_BEQ, addr_regh, TCG_REG_AT, 0);
+    tcg_out_opc_br(s, OPC_BEQ, addr_regh, TCG_REG_AT);
    tcg_out_nop(s);

    reloc_pc16(label3_ptr, (tcg_target_long) s->code_ptr);
 # else
    label1_ptr = s->code_ptr;
-    tcg_out_opc_imm(s, OPC_BEQ, TCG_REG_T0, TCG_REG_AT, 0);
+    tcg_out_opc_br(s, OPC_BEQ, TCG_REG_T0, TCG_REG_AT);
    tcg_out_nop(s);
 # endif

@@ -725,15 +736,15 @@ static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args,
    }

    label2_ptr = s->code_ptr;
-    tcg_out_opc_imm(s, OPC_BEQ, TCG_REG_ZERO, TCG_REG_ZERO, 0);
+    tcg_out_opc_br(s, OPC_BEQ, TCG_REG_ZERO, TCG_REG_ZERO);
    tcg_out_nop(s);

    /* label1: fast path */
    reloc_pc16(label1_ptr, (tcg_target_long) s->code_ptr);

-    tcg_out_opc_imm(s, OPC_LW, TCG_REG_V0, TCG_REG_A0,
+    tcg_out_opc_imm(s, OPC_LW, TCG_REG_A0, TCG_REG_A0,
                    offsetof(CPUState, tlb_table[mem_index][0].addend) + addr_meml);
-    tcg_out_opc_reg(s, OPC_ADDU, TCG_REG_V0, TCG_REG_V0, addr_regl);
+    tcg_out_opc_reg(s, OPC_ADDU, TCG_REG_V0, TCG_REG_A0, addr_regl);

    addr_reg1 = TCG_REG_V0;
 #endif
@@ -857,20 +868,20 @@ static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args,

 # if TARGET_LONG_BITS == 64
    label3_ptr = s->code_ptr;
-    tcg_out_opc_imm(s, OPC_BNE, TCG_REG_T0, TCG_REG_AT, 0);
+    tcg_out_opc_br(s, OPC_BNE, TCG_REG_T0, TCG_REG_AT);
    tcg_out_nop(s);

    tcg_out_opc_imm(s, OPC_LW, TCG_REG_AT, TCG_REG_A0,
                    offsetof(CPUState, tlb_table[mem_index][0].addr_write) + addr_memh);

    label1_ptr = s->code_ptr;
-    tcg_out_opc_imm(s, OPC_BEQ, addr_regh, TCG_REG_AT, 0);
+    tcg_out_opc_br(s, OPC_BEQ, addr_regh, TCG_REG_AT);
    tcg_out_nop(s);

    reloc_pc16(label3_ptr, (tcg_target_long) s->code_ptr);
 # else
    label1_ptr = s->code_ptr;
-    tcg_out_opc_imm(s, OPC_BEQ, TCG_REG_T0, TCG_REG_AT, 0);
+    tcg_out_opc_br(s, OPC_BEQ, TCG_REG_T0, TCG_REG_AT);
    tcg_out_nop(s);
 # endif

@@ -911,7 +922,7 @@ static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args,
    tcg_out_nop(s);

    label2_ptr = s->code_ptr;
-    tcg_out_opc_imm(s, OPC_BEQ, TCG_REG_ZERO, TCG_REG_ZERO, 0);
+    tcg_out_opc_br(s, OPC_BEQ, TCG_REG_ZERO, TCG_REG_ZERO);
    tcg_out_nop(s);

    /* label1: fast path */
@@ -1262,7 +1273,9 @@ static const TCGTargetOpDef mips_op_defs[] = {
 };

 static int tcg_target_callee_save_regs[] = {
+#if 0 /* used for the global env (TCG_AREG0), so no need to save */
    TCG_REG_S0,
+#endif
    TCG_REG_S1,
    TCG_REG_S2,
    TCG_REG_S3,
@@ -1271,8 +1284,7 @@ static int tcg_target_callee_save_regs[] = {
    TCG_REG_S6,
    TCG_REG_S7,
    TCG_REG_GP,
-    /* TCG_REG_FP, */ /* currently used for the global env, so np
-                         need to save */
+    TCG_REG_FP,
    TCG_REG_RA,       /* should be last for ABI compliance */
 };

@@ -92,9 +92,9 @@ enum {
 #undef TCG_TARGET_HAS_ext16u_i32   /* andi rt, rs, 0xffff */

 /* Note: must be synced with dyngen-exec.h */
-#define TCG_AREG0 TCG_REG_FP
-#define TCG_AREG1 TCG_REG_S0
-#define TCG_AREG2 TCG_REG_S1
+#define TCG_AREG0 TCG_REG_S0
+#define TCG_AREG1 TCG_REG_S1
+#define TCG_AREG2 TCG_REG_FP

 #include <sys/cachectl.h>

@@ -1036,7 +1036,7 @@ USBDevice *usb_host_device_open(const char *devname)
    qdev_prop_set_uint32(&dev->qdev, "hostaddr",  filter.addr);
    qdev_prop_set_uint32(&dev->qdev, "vendorid",  filter.vendor_id);
    qdev_prop_set_uint32(&dev->qdev, "productid", filter.product_id);
-    qdev_init(&dev->qdev);
+    qdev_init_nofail(&dev->qdev);
    return dev;

 fail:
@@ -2573,9 +2573,9 @@ static void numa_add(const char *optarg)
                        fprintf(stderr,
                            "only 63 CPUs in NUMA mode supported.\n");
                    }
-                    value = (1 << (endvalue + 1)) - (1 << value);
+                    value = (2ULL << endvalue) - (1ULL << value);
                } else {
-                    value = 1 << value;
+                    value = 1ULL << value;
                }
            }
            node_cpumask[nodenr] = value;
@@ -2689,17 +2689,28 @@ static int usb_device_del(const char *devname)

 static int usb_parse(const char *cmdline)
 {
-    return usb_device_add(cmdline, 0);
+    int r;
+    r = usb_device_add(cmdline, 0);
+    if (r < 0) {
+        fprintf(stderr, "qemu: could not add USB device '%s'\n", cmdline);
+    }
+    return r;
 }

 void do_usb_add(Monitor *mon, const QDict *qdict)
 {
-    usb_device_add(qdict_get_str(qdict, "devname"), 1);
+    const char *devname = qdict_get_str(qdict, "devname");
+    if (usb_device_add(devname, 1) < 0) {
+        qemu_error("could not add USB device '%s'\n", devname);
+    }
 }

 void do_usb_del(Monitor *mon, const QDict *qdict)
 {
-    usb_device_del(qdict_get_str(qdict, "devname"));
+    const char *devname = qdict_get_str(qdict, "devname");
+    if (usb_device_del(devname) < 0) {
+        qemu_error("could not delete USB device '%s'\n", devname);
+    }
 }

 /***********************************************************/
@@ -3460,6 +3471,8 @@ static int cpu_can_run(CPUState *env)
        return 0;
    if (env->stopped)
        return 0;
+    if (!vm_running)
+        return 0;
    return 1;
 }

@@ -3473,10 +3486,10 @@ void qemu_init_vcpu(void *_env)
 {
    CPUState *env = _env;

-    if (kvm_enabled())
-        kvm_init_vcpu(env);
    env->nr_cores = smp_cores;
    env->nr_threads = smp_threads;
+    if (kvm_enabled())
+        kvm_init_vcpu(env);
    return;
 }

@@ -3802,12 +3815,12 @@ void qemu_init_vcpu(void *_env)
 {
    CPUState *env = _env;

+    env->nr_cores = smp_cores;
+    env->nr_threads = smp_threads;
    if (kvm_enabled())
        kvm_start_vcpu(env);
    else
        tcg_init_vcpu(env);
-    env->nr_cores = smp_cores;
-    env->nr_threads = smp_threads;
 }

 void qemu_notify_event(void)
@@ -4031,14 +4044,15 @@ static void tcg_cpu_exec(void)
    for (; next_cpu != NULL; next_cpu = next_cpu->next_cpu) {
        CPUState *env = cur_cpu = next_cpu;

-        if (!vm_running)
-            break;
        if (timer_alarm_pending) {
            timer_alarm_pending = 0;
            break;
        }
        if (cpu_can_run(env))
            ret = qemu_cpu_exec(env);
+        else if (env->stop)
+            break;
+
        if (ret == EXCP_DEBUG) {
            gdb_set_stop_cpu(env);
            debug_requested = 1;
@@ -4839,7 +4853,7 @@ int main(int argc, char **argv, char **envp)
    const char *gdbstub_dev = NULL;
    uint32_t boot_devices_bitmap = 0;
    int i;
-    int snapshot, linux_boot, net_boot;
+    int snapshot, linux_boot;
    const char *initrd_filename;
    const char *kernel_filename, *kernel_cmdline;
    char boot_devices[33] = "cad"; /* default to HD->floppy->CD-ROM */
@@ -5356,6 +5370,9 @@ int main(int argc, char **argv, char **envp)
            case QEMU_OPTION_serial:
                add_device_config(DEV_SERIAL, optarg);
                default_serial = 0;
+                if (strncmp(optarg, "mon:", 4) == 0) {
+                    default_monitor = 0;
+                }
                break;
            case QEMU_OPTION_watchdog:
                if (watchdog) {
@@ -5374,10 +5391,16 @@ int main(int argc, char **argv, char **envp)
            case QEMU_OPTION_virtiocon:
                add_device_config(DEV_VIRTCON, optarg);
                default_virtcon = 0;
+                if (strncmp(optarg, "mon:", 4) == 0) {
+                    default_monitor = 0;
+                }
                break;
            case QEMU_OPTION_parallel:
                add_device_config(DEV_PARALLEL, optarg);
                default_parallel = 0;
+                if (strncmp(optarg, "mon:", 4) == 0) {
+                    default_monitor = 0;
+                }
                break;
 	    case QEMU_OPTION_loadvm:
 		loadvm = optarg;
@@ -5780,6 +5803,12 @@ int main(int argc, char **argv, char **envp)
            fprintf(stderr, "failed to initialize KVM\n");
            exit(1);
        }
+    } else {
+        /* without kvm enabled, we can only support 4095 MB RAM */
+        if (ram_size > (4095UL << 20)) {
+            fprintf(stderr, "qemu: without kvm support at most 4095 MB RAM can be simulated\n");
+            exit(1);
+        }
    }

    if (qemu_init_main_loop()) {
@@ -5823,9 +5852,6 @@ int main(int argc, char **argv, char **envp)
        exit(1);
    }

-    net_boot = (boot_devices_bitmap >> ('n' - 'a')) & 0xF;
-    net_set_boot_mask(net_boot);
-
    /* init the bluetooth world */
    if (foreach_device_config(DEV_BT, bt_parse))
        exit(1);
@@ -5957,6 +5983,8 @@ int main(int argc, char **argv, char **envp)
    if (qemu_opts_foreach(&qemu_device_opts, device_init_func, NULL, 1) != 0)
        exit(1);

+    net_check_clients();
+
    if (!display_state)
        dumb_display_init();
    /* just use the first displaystate for the moment */
@@ -1421,8 +1421,10 @@ static void pointer_event(VncState *vs, int button_mask, int x, int y)
        dz = 1;

    if (vs->absolute) {
-        kbd_mouse_event(x * 0x7FFF / (ds_get_width(vs->ds) - 1),
-                        y * 0x7FFF / (ds_get_height(vs->ds) - 1),
+        kbd_mouse_event(ds_get_width(vs->ds) > 1 ?
+                          x * 0x7FFF / (ds_get_width(vs->ds) - 1) : 0x4000,
+                        ds_get_height(vs->ds) > 1 ?
+                          y * 0x7FFF / (ds_get_height(vs->ds) - 1) : 0x4000,
                        dz, buttons);
    } else if (vnc_has_feature(vs, VNC_FEATURE_POINTER_TYPE_CHANGE)) {
        x -= 0x7FFF;
@@ -2293,7 +2295,7 @@ static int vnc_refresh_server_surface(VncDisplay *vd)
 static void vnc_refresh(void *opaque)
 {
    VncDisplay *vd = opaque;
-    VncState *vs = NULL;
+    VncState *vs = NULL, *vn = NULL;
    int has_dirty = 0, rects = 0;

    vga_hw_update();
@@ -2302,9 +2304,15 @@ static void vnc_refresh(void *opaque)

    vs = vd->clients;
    while (vs != NULL) {
+        vn = vs->next;
        rects += vnc_update_client(vs, has_dirty);
-        vs = vs->next;
+        /* vs might be free()ed here */
+        vs = vn;
    }
+    /* vd->timer could be NULL now if the last client disconnected,
+     * in this case don't update the timer */
+    if (vd->timer == NULL)
+        return;

    if (has_dirty && rects) {
        vd->timer_interval /= 2;
@@ -165,6 +165,8 @@ static void CONCAT(send_hextile_tile_, NAME)(VncState *vs,
 	    irow += ds_get_linesize(vs->ds) / sizeof(pixel_t);
 	}

+	/* A SubrectsColoured subtile invalidates the foreground color */
+	*has_fg = 0;
 	if (n_data > (w * h * sizeof(pixel_t))) {
 	    n_colors = 4;
 	    flags = 0x01;
@@ -1 +1 @@
 .12.0
 .12.4