console: Remove unused QEMU_BIG_ENDIAN_FLAG

If we need to, we should use the pixman formats instead but for
now this is unused except in commented out code so take it out
to avoid further confusion about surface endianness.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

.gitignore

@@ -11,6 +11,10 @@
 /trace/generated-tracers.dtrace
 /trace/generated-events.h
 /trace/generated-events.c
+/trace/generated-helpers-wrappers.h
+/trace/generated-helpers.h
+/trace/generated-helpers.c
+/trace/generated-tcg-tracers.h
 /trace/generated-ust-provider.h
 /trace/generated-ust.c
 /libcacard/trace/generated-tracers.c
@@ -26,6 +30,7 @@
 /qapi-generated
 /qapi-types.[ch]
 /qapi-visit.[ch]
+/qapi-event.[ch]
 /qmp-commands.h
 /qmp-marshal.c
 /qemu-doc.html

CODING_STYLE

@@ -91,3 +91,17 @@ Mixed declarations (interleaving statements and declarations within blocks)
 are not allowed; declarations should be at the beginning of blocks.  In other
 words, the code should not generate warnings if using GCC's
 -Wdeclaration-after-statement option.
+
+6. Conditional statements
+
+When comparing a variable for (in)equality with a constant, list the
+constant on the right, as in:
+
+if (a == 1) {
+    /* Reads like: "If a equals 1" */
+    do_something();
+}
+
+Rationale: Yoda conditions (as in 'if (1 == a)') are awkward to read.
+Besides, good compilers already warn users when '==' is mis-typed as '=',
+even when the constant is on the right.

MAINTAINERS

@@ -161,6 +161,12 @@ S: Maintained
 F: target-xtensa/
 F: hw/xtensa/
 
+TriCore
+M: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
+S: Maintained
+F: target-tricore/
+F: hw/tricore/
+
 Guest CPU Cores (KVM):
 ----------------------
 
@@ -563,6 +569,7 @@ Devices
 -------
 IDE
 M: Kevin Wolf <kwolf@redhat.com>
+M: Stefan Hajnoczi <stefanha@redhat.com>
 S: Odd Fixes
 F: include/hw/ide.h
 F: hw/ide/
@@ -613,7 +620,7 @@ USB
 M: Gerd Hoffmann <kraxel@redhat.com>
 S: Maintained
 F: hw/usb/*
-F: tests/usb-hcd-ehci-test.c
+F: tests/usb-*-test.c
 
 VFIO
 M: Alex Williamson <alex.williamson@redhat.com>
@@ -677,6 +684,12 @@ S: Maintained
 F: hw/*/xilinx_*
 F: include/hw/xilinx.h
 
+Vmware
+M: Dmitry Fleytman <dmitry@daynix.com>
+S: Maintained
+F: hw/net/vmxnet*
+F: hw/scsi/vmw_pvscsi*
+
 Subsystems
 ----------
 Audio
@@ -853,7 +866,7 @@ S: Odd Fixes
 F: scripts/checkpatch.pl
 
 Seccomp
-M: Eduardo Otubo <otubo@linux.vnet.ibm.com>
+M: Eduardo Otubo <eduardo.otubo@profitbricks.com>
 S: Supported
 F: qemu-seccomp.c
 F: include/sysemu/seccomp.h
@@ -999,3 +1012,9 @@ SSH
 M: Richard W.M. Jones <rjones@redhat.com>
 S: Supported
 F: block/ssh.c
+
+ARCHIPELAGO
+M: Chrysostomos Nanakos <cnanakos@grnet.gr>
+M: Chrysostomos Nanakos <chris@include.gr>
+S: Maintained
+F: block/archipelago.c

Makefile

@@ -57,6 +57,12 @@ GENERATED_HEADERS += trace/generated-tracers-dtrace.h
 endif
 GENERATED_SOURCES += trace/generated-tracers.c
 
+GENERATED_HEADERS += trace/generated-tcg-tracers.h
+
+GENERATED_HEADERS += trace/generated-helpers-wrappers.h
+GENERATED_HEADERS += trace/generated-helpers.h
+GENERATED_SOURCES += trace/generated-helpers.c
+
 ifeq ($(findstring ust,$(TRACE_BACKENDS)),ust)
 GENERATED_HEADERS += trace/generated-ust-provider.h
 GENERATED_SOURCES += trace/generated-ust.c
@@ -202,7 +208,7 @@ Makefile: $(version-obj-y) $(version-lobj-y)
 # Build libraries
 
 libqemustub.a: $(stub-obj-y)
-libqemuutil.a: $(util-obj-y) qapi-types.o qapi-visit.o qapi-event.o
+libqemuutil.a: $(util-obj-y)
 
 block-modules = $(foreach o,$(block-obj-m),"$(basename $(subst /,-,$o))",) NULL
 util/module.o-cflags = -D'CONFIG_BLOCK_MODULES=$(block-modules)'
@@ -248,7 +254,7 @@ $(SRC_PATH)/qga/qapi-schema.json $(SRC_PATH)/scripts/qapi-commands.py $(qapi-py)
 
 qapi-modules = $(SRC_PATH)/qapi-schema.json $(SRC_PATH)/qapi/common.json \
                $(SRC_PATH)/qapi/block.json $(SRC_PATH)/qapi/block-core.json \
-               $(SRC_PATH)/qapi-event.json
+               $(SRC_PATH)/qapi/event.json
 
 qapi-types.c qapi-types.h :\
 $(qapi-modules) $(SRC_PATH)/scripts/qapi-types.py $(qapi-py)
@@ -344,7 +350,8 @@ multiboot.bin linuxboot.bin kvmvapic.bin \
 s390-zipl.rom \
 s390-ccw.img \
 spapr-rtas.bin slof.bin \
-palcode-clipper
+palcode-clipper \
+u-boot.e500
 else
 BLOBS=
 endif

Makefile.objs

@@ -1,7 +1,7 @@
 #######################################################################
 # Common libraries for tools and emulators
 stub-obj-y = stubs/
-util-obj-y = util/ qobject/ qapi/ trace/
+util-obj-y = util/ qobject/ qapi/ qapi-types.o qapi-visit.o qapi-event.o
 
 #######################################################################
 # block-obj-y is code used by both qemu system emulation and qemu-img
@@ -12,7 +12,6 @@ block-obj-y += main-loop.o iohandler.o qemu-timer.o
 block-obj-$(CONFIG_POSIX) += aio-posix.o
 block-obj-$(CONFIG_WIN32) += aio-win32.o
 block-obj-y += block/
-block-obj-y += qapi-types.o qapi-visit.o qapi-event.o
 block-obj-y += qemu-io-cmds.o
 
 block-obj-y += qemu-coroutine.o qemu-coroutine-lock.o qemu-coroutine-io.o
@@ -88,11 +87,6 @@ common-obj-y += qmp-marshal.o
 common-obj-y += qmp.o hmp.o
 endif
 
-######################################################################
-# some qapi visitors are used by both system and user emulation:
-
-common-obj-y += qapi-visit.o qapi-types.o
-
 #######################################################################
 # Target-independent parts used in system and user emulation
 common-obj-y += qemu-log.o
@@ -106,10 +100,15 @@ common-obj-y += disas/
 version-obj-$(CONFIG_WIN32) += $(BUILD_DIR)/version.o
 version-lobj-$(CONFIG_WIN32) += $(BUILD_DIR)/version.lo
 
+######################################################################
+# tracing
+util-obj-y +=  trace/
+target-obj-y += trace/
+
 ######################################################################
 # guest agent
 
 # FIXME: a few definitions from qapi-types.o/qapi-visit.o are needed
 # by libqemuutil.a.  These should be moved to a separate .json schema.
-qga-obj-y = qga/ qapi-types.o qapi-visit.o
+qga-obj-y = qga/
 qga-vss-dll-obj-y = qga/

Makefile.target

@@ -38,7 +38,7 @@ config-target.h: config-target.h-timestamp
 config-target.h-timestamp: config-target.mak
 
 ifdef CONFIG_TRACE_SYSTEMTAP
-stap: $(QEMU_PROG).stp-installed $(QEMU_PROG).stp
+stap: $(QEMU_PROG).stp-installed $(QEMU_PROG).stp $(QEMU_PROG)-simpletrace.stp
 
 ifdef CONFIG_USER_ONLY
 TARGET_TYPE=user
@@ -64,6 +64,13 @@ $(QEMU_PROG).stp: $(SRC_PATH)/trace-events
 		--target-type=$(TARGET_TYPE) \
 		< $< > $@,"  GEN   $(TARGET_DIR)$(QEMU_PROG).stp")
 
+$(QEMU_PROG)-simpletrace.stp: $(SRC_PATH)/trace-events
+	$(call quiet-command,$(TRACETOOL) \
+		--format=simpletrace-stap \
+		--backends=$(TRACE_BACKENDS) \
+		--probe-prefix=qemu.$(TARGET_TYPE).$(TARGET_NAME) \
+		< $< > $@,"  GEN   $(TARGET_DIR)$(QEMU_PROG)-simpletrace.stp")
+
 else
 stap:
 endif
@@ -152,21 +159,22 @@ endif # CONFIG_SOFTMMU
 dummy := $(call unnest-vars,,obj-y)
 all-obj-y := $(obj-y)
 
+target-obj-y :=
 block-obj-y :=
 common-obj-y :=
 include $(SRC_PATH)/Makefile.objs
+dummy := $(call unnest-vars,,target-obj-y)
+target-obj-y-save := $(target-obj-y)
 dummy := $(call unnest-vars,.., \
                block-obj-y \
                block-obj-m \
                common-obj-y \
                common-obj-m)
+target-obj-y := $(target-obj-y-save)
 all-obj-y += $(common-obj-y)
+all-obj-y += $(target-obj-y)
 all-obj-$(CONFIG_SOFTMMU) += $(block-obj-y)
 
-ifndef CONFIG_HAIKU
-LIBS+=-lm
-endif
-
 # build either PROG or PROGW
 $(QEMU_PROG_BUILD): $(all-obj-y) ../libqemuutil.a ../libqemustub.a
 	$(call LINK,$^)
@@ -195,6 +203,7 @@ endif
 ifdef CONFIG_TRACE_SYSTEMTAP
 	$(INSTALL_DIR) "$(DESTDIR)$(qemu_datadir)/../systemtap/tapset"
 	$(INSTALL_DATA) $(QEMU_PROG).stp-installed "$(DESTDIR)$(qemu_datadir)/../systemtap/tapset/$(QEMU_PROG).stp"
+	$(INSTALL_DATA) $(QEMU_PROG)-simpletrace.stp "$(DESTDIR)$(qemu_datadir)/../systemtap/tapset/$(QEMU_PROG)-simpletrace.stp"
 endif
 
 GENERATED_HEADERS += config-target.h

VERSION

@@ -1 +1 @@
-2.0.50
+2.1.50

aio-posix.c

@@ -100,6 +100,11 @@ void aio_set_event_notifier(AioContext *ctx,
                        (IOHandler *)io_read, NULL, notifier);
 }
 
+bool aio_prepare(AioContext *ctx)
+{
+    return false;
+}
+
 bool aio_pending(AioContext *ctx)
 {
     AioHandler *node;
@@ -119,13 +124,22 @@ bool aio_pending(AioContext *ctx)
     return false;
 }
 
-static bool aio_dispatch(AioContext *ctx)
+bool aio_dispatch(AioContext *ctx)
 {
     AioHandler *node;
     bool progress = false;
 
     /*
-     * We have to walk very carefully in case qemu_aio_set_fd_handler is
+     * If there are callbacks left that have been queued, we need to call them.
+     * Do not call select in this case, because it is possible that the caller
+     * does not need a complete flush (as is the case for aio_poll loops).
+     */
+    if (aio_bh_poll(ctx)) {
+        progress = true;
+    }
+
+    /*
+     * We have to walk very carefully in case aio_set_fd_handler is
      * called while we're walking.
      */
     node = QLIST_FIRST(&ctx->aio_handlers);
@@ -175,28 +189,24 @@ static bool aio_dispatch(AioContext *ctx)
 bool aio_poll(AioContext *ctx, bool blocking)
 {
     AioHandler *node;
+    bool was_dispatching;
     int ret;
     bool progress;
 
+    was_dispatching = ctx->dispatching;
     progress = false;
 
-    /*
-     * If there are callbacks left that have been queued, we need to call them.
-     * Do not call select in this case, because it is possible that the caller
-     * does not need a complete flush (as is the case for qemu_aio_wait loops).
+    /* aio_notify can avoid the expensive event_notifier_set if
+     * everything (file descriptors, bottom halves, timers) will
+     * be re-evaluated before the next blocking poll().  This is
+     * already true when aio_poll is called with blocking == false;
+     * if blocking == true, it is only true after poll() returns.
+     *
+     * If we're in a nested event loop, ctx->dispatching might be true.
+     * In that case we can restore it just before returning, but we
+     * have to clear it now.
      */
-    if (aio_bh_poll(ctx)) {
-        blocking = false;
-        progress = true;
-    }
-
-    if (aio_dispatch(ctx)) {
-        progress = true;
-    }
-
-    if (progress && !blocking) {
-        return true;
-    }
+    aio_set_dispatching(ctx, !blocking);
 
     ctx->walking_handlers++;
 
@@ -220,7 +230,7 @@ bool aio_poll(AioContext *ctx, bool blocking)
     /* wait until next event */
     ret = qemu_poll_ns((GPollFD *)ctx->pollfds->data,
                          ctx->pollfds->len,
-                         blocking ? timerlistgroup_deadline_ns(&ctx->tlg) : 0);
+                         blocking ? aio_compute_timeout(ctx) : 0);
 
     /* if we have any readable fds, dispatch event */
     if (ret > 0) {
@@ -234,9 +244,11 @@ bool aio_poll(AioContext *ctx, bool blocking)
     }
 
     /* Run dispatch even if there were no readable fds to run timers */
+    aio_set_dispatching(ctx, true);
     if (aio_dispatch(ctx)) {
         progress = true;
     }
 
+    aio_set_dispatching(ctx, was_dispatching);
     return progress;
 }

aio-win32.c

@@ -22,12 +22,80 @@
 
 struct AioHandler {
     EventNotifier *e;
+    IOHandler *io_read;
+    IOHandler *io_write;
     EventNotifierHandler *io_notify;
     GPollFD pfd;
     int deleted;
+    void *opaque;
     QLIST_ENTRY(AioHandler) node;
 };
 
+void aio_set_fd_handler(AioContext *ctx,
+                        int fd,
+                        IOHandler *io_read,
+                        IOHandler *io_write,
+                        void *opaque)
+{
+    /* fd is a SOCKET in our case */
+    AioHandler *node;
+
+    QLIST_FOREACH(node, &ctx->aio_handlers, node) {
+        if (node->pfd.fd == fd && !node->deleted) {
+            break;
+        }
+    }
+
+    /* Are we deleting the fd handler? */
+    if (!io_read && !io_write) {
+        if (node) {
+            /* If the lock is held, just mark the node as deleted */
+            if (ctx->walking_handlers) {
+                node->deleted = 1;
+                node->pfd.revents = 0;
+            } else {
+                /* Otherwise, delete it for real.  We can't just mark it as
+                 * deleted because deleted nodes are only cleaned up after
+                 * releasing the walking_handlers lock.
+                 */
+                QLIST_REMOVE(node, node);
+                g_free(node);
+            }
+        }
+    } else {
+        HANDLE event;
+
+        if (node == NULL) {
+            /* Alloc and insert if it's not already there */
+            node = g_malloc0(sizeof(AioHandler));
+            node->pfd.fd = fd;
+            QLIST_INSERT_HEAD(&ctx->aio_handlers, node, node);
+        }
+
+        node->pfd.events = 0;
+        if (node->io_read) {
+            node->pfd.events |= G_IO_IN;
+        }
+        if (node->io_write) {
+            node->pfd.events |= G_IO_OUT;
+        }
+
+        node->e = &ctx->notifier;
+
+        /* Update handler with latest information */
+        node->opaque = opaque;
+        node->io_read = io_read;
+        node->io_write = io_write;
+
+        event = event_notifier_get_handle(&ctx->notifier);
+        WSAEventSelect(node->pfd.fd, event,
+                       FD_READ | FD_ACCEPT | FD_CLOSE |
+                       FD_CONNECT | FD_WRITE | FD_OOB);
+    }
+
+    aio_notify(ctx);
+}
+
 void aio_set_event_notifier(AioContext *ctx,
                             EventNotifier *e,
                             EventNotifierHandler *io_notify)
@@ -76,6 +144,43 @@ void aio_set_event_notifier(AioContext *ctx,
     aio_notify(ctx);
 }
 
+bool aio_prepare(AioContext *ctx)
+{
+    static struct timeval tv0;
+    AioHandler *node;
+    bool have_select_revents = false;
+    fd_set rfds, wfds;
+
+    /* fill fd sets */
+    FD_ZERO(&rfds);
+    FD_ZERO(&wfds);
+    QLIST_FOREACH(node, &ctx->aio_handlers, node) {
+        if (node->io_read) {
+            FD_SET ((SOCKET)node->pfd.fd, &rfds);
+        }
+        if (node->io_write) {
+            FD_SET ((SOCKET)node->pfd.fd, &wfds);
+        }
+    }
+
+    if (select(0, &rfds, &wfds, NULL, &tv0) > 0) {
+        QLIST_FOREACH(node, &ctx->aio_handlers, node) {
+            node->pfd.revents = 0;
+            if (FD_ISSET(node->pfd.fd, &rfds)) {
+                node->pfd.revents |= G_IO_IN;
+                have_select_revents = true;
+            }
+
+            if (FD_ISSET(node->pfd.fd, &wfds)) {
+                node->pfd.revents |= G_IO_OUT;
+                have_select_revents = true;
+            }
+        }
+    }
+
+    return have_select_revents;
+}
+
 bool aio_pending(AioContext *ctx)
 {
     AioHandler *node;
@@ -84,47 +189,37 @@ bool aio_pending(AioContext *ctx)
         if (node->pfd.revents && node->io_notify) {
             return true;
         }
+
+        if ((node->pfd.revents & G_IO_IN) && node->io_read) {
+            return true;
+        }
+        if ((node->pfd.revents & G_IO_OUT) && node->io_write) {
+            return true;
+        }
     }
 
     return false;
 }
 
-bool aio_poll(AioContext *ctx, bool blocking)
+static bool aio_dispatch_handlers(AioContext *ctx, HANDLE event)
 {
     AioHandler *node;
-    HANDLE events[MAXIMUM_WAIT_OBJECTS + 1];
-    bool progress;
-    int count;
-    int timeout;
-
-    progress = false;
+    bool progress = false;
 
     /*
-     * If there are callbacks left that have been queued, we need to call then.
-     * Do not call select in this case, because it is possible that the caller
-     * does not need a complete flush (as is the case for qemu_aio_wait loops).
-     */
-    if (aio_bh_poll(ctx)) {
-        blocking = false;
-        progress = true;
-    }
-
-    /* Run timers */
-    progress |= timerlistgroup_run_timers(&ctx->tlg);
-
-    /*
-     * Then dispatch any pending callbacks from the GSource.
-     *
-     * We have to walk very carefully in case qemu_aio_set_fd_handler is
+     * We have to walk very carefully in case aio_set_fd_handler is
      * called while we're walking.
      */
     node = QLIST_FIRST(&ctx->aio_handlers);
     while (node) {
         AioHandler *tmp;
+        int revents = node->pfd.revents;
 
         ctx->walking_handlers++;
 
-        if (node->pfd.revents && node->io_notify) {
+        if (!node->deleted &&
+            (revents || event_notifier_get_handle(node->e) == event) &&
+            node->io_notify) {
             node->pfd.revents = 0;
             node->io_notify(node->e);
 
@@ -134,6 +229,28 @@ bool aio_poll(AioContext *ctx, bool blocking)
             }
         }
 
+        if (!node->deleted &&
+            (node->io_read || node->io_write)) {
+            node->pfd.revents = 0;
+            if ((revents & G_IO_IN) && node->io_read) {
+                node->io_read(node->opaque);
+                progress = true;
+            }
+            if ((revents & G_IO_OUT) && node->io_write) {
+                node->io_write(node->opaque);
+                progress = true;
+            }
+
+            /* if the next select() will return an event, we have progressed */
+            if (event == event_notifier_get_handle(&ctx->notifier)) {
+                WSANETWORKEVENTS ev;
+                WSAEnumNetworkEvents(node->pfd.fd, event, &ev);
+                if (ev.lNetworkEvents) {
+                    progress = true;
+                }
+            }
+        }
+
         tmp = node;
         node = QLIST_NEXT(node, node);
 
@@ -145,10 +262,47 @@ bool aio_poll(AioContext *ctx, bool blocking)
         }
     }
 
-    if (progress && !blocking) {
-        return true;
+    return progress;
+}
+
+bool aio_dispatch(AioContext *ctx)
+{
+    bool progress;
+
+    progress = aio_bh_poll(ctx);
+    progress |= aio_dispatch_handlers(ctx, INVALID_HANDLE_VALUE);
+    progress |= timerlistgroup_run_timers(&ctx->tlg);
+    return progress;
+}
+
+bool aio_poll(AioContext *ctx, bool blocking)
+{
+    AioHandler *node;
+    HANDLE events[MAXIMUM_WAIT_OBJECTS + 1];
+    bool was_dispatching, progress, have_select_revents, first;
+    int count;
+    int timeout;
+
+    if (aio_prepare(ctx)) {
+        blocking = false;
+        have_select_revents = true;
     }
 
+    was_dispatching = ctx->dispatching;
+    progress = false;
+
+    /* aio_notify can avoid the expensive event_notifier_set if
+     * everything (file descriptors, bottom halves, timers) will
+     * be re-evaluated before the next blocking poll().  This is
+     * already true when aio_poll is called with blocking == false;
+     * if blocking == true, it is only true after poll() returns.
+     *
+     * If we're in a nested event loop, ctx->dispatching might be true.
+     * In that case we can restore it just before returning, but we
+     * have to clear it now.
+     */
+    aio_set_dispatching(ctx, !blocking);
+
     ctx->walking_handlers++;
 
     /* fill fd sets */
@@ -160,64 +314,42 @@ bool aio_poll(AioContext *ctx, bool blocking)
     }
 
     ctx->walking_handlers--;
+    first = true;
 
     /* wait until next event */
     while (count > 0) {
+        HANDLE event;
         int ret;
 
-        timeout = blocking ?
-            qemu_timeout_ns_to_ms(timerlistgroup_deadline_ns(&ctx->tlg)) : 0;
+        timeout = blocking
+            ? qemu_timeout_ns_to_ms(aio_compute_timeout(ctx)) : 0;
         ret = WaitForMultipleObjects(count, events, FALSE, timeout);
+        aio_set_dispatching(ctx, true);
+
+        if (first && aio_bh_poll(ctx)) {
+            progress = true;
+        }
+        first = false;
 
         /* if we have any signaled events, dispatch event */
-        if ((DWORD) (ret - WAIT_OBJECT_0) >= count) {
+        event = NULL;
+        if ((DWORD) (ret - WAIT_OBJECT_0) < count) {
+            event = events[ret - WAIT_OBJECT_0];
+        } else if (!have_select_revents) {
             break;
         }
 
+        have_select_revents = false;
         blocking = false;
 
-        /* we have to walk very carefully in case
-         * qemu_aio_set_fd_handler is called while we're walking */
-        node = QLIST_FIRST(&ctx->aio_handlers);
-        while (node) {
-            AioHandler *tmp;
-
-            ctx->walking_handlers++;
-
-            if (!node->deleted &&
-                event_notifier_get_handle(node->e) == events[ret - WAIT_OBJECT_0] &&
-                node->io_notify) {
-                node->io_notify(node->e);
-
-                /* aio_notify() does not count as progress */
-                if (node->e != &ctx->notifier) {
-                    progress = true;
-                }
-            }
-
-            tmp = node;
-            node = QLIST_NEXT(node, node);
-
-            ctx->walking_handlers--;
-
-            if (!ctx->walking_handlers && tmp->deleted) {
-                QLIST_REMOVE(tmp, node);
-                g_free(tmp);
-            }
-        }
+        progress |= aio_dispatch_handlers(ctx, event);
 
         /* Try again, but only call each handler once.  */
         events[ret - WAIT_OBJECT_0] = events[--count];
     }
 
-    if (blocking) {
-        /* Run the timers a second time. We do this because otherwise aio_wait
-         * will not note progress - and will stop a drain early - if we have
-         * a timer that was not ready to run entering g_poll but is ready
-         * after g_poll. This will only do anything if a timer has expired.
-         */
-        progress |= timerlistgroup_run_timers(&ctx->tlg);
-    }
+    progress |= timerlistgroup_run_timers(&ctx->tlg);
 
+    aio_set_dispatching(ctx, was_dispatching);
     return progress;
 }

arch_init.c

@@ -104,6 +104,8 @@ int graphic_depth = 32;
 #define QEMU_ARCH QEMU_ARCH_XTENSA
 #elif defined(TARGET_UNICORE32)
 #define QEMU_ARCH QEMU_ARCH_UNICORE32
+#elif defined(TARGET_TRICORE)
+#define QEMU_ARCH QEMU_ARCH_TRICORE
 #endif
 
 const uint32_t arch_type = QEMU_ARCH;
@@ -1072,8 +1074,8 @@ static int ram_load(QEMUFile *f, void *opaque, int version_id)
                 QTAILQ_FOREACH(block, &ram_list.blocks, next) {
                     if (!strncmp(id, block->idstr, sizeof(id))) {
                         if (block->length != length) {
-                            error_report("Length mismatch: %s: " RAM_ADDR_FMT
-                                         " in != " RAM_ADDR_FMT, id, length,
+                            error_report("Length mismatch: %s: 0x" RAM_ADDR_FMT
+                                         " in != 0x" RAM_ADDR_FMT, id, length,
                                          block->length);
                             ret =  -EINVAL;
                         }

async.c

@@ -26,6 +26,7 @@
 #include "block/aio.h"
 #include "block/thread-pool.h"
 #include "qemu/main-loop.h"
+#include "qemu/atomic.h"
 
 /***********************************************************/
 /* bottom halves (can be seen as timers which expire ASAP) */
@@ -151,39 +152,48 @@ void qemu_bh_delete(QEMUBH *bh)
     bh->deleted = 1;
 }
 
-static gboolean
-aio_ctx_prepare(GSource *source, gint    *timeout)
+int64_t
+aio_compute_timeout(AioContext *ctx)
 {
-    AioContext *ctx = (AioContext *) source;
+    int64_t deadline;
+    int timeout = -1;
     QEMUBH *bh;
-    int deadline;
 
-    /* We assume there is no timeout already supplied */
-    *timeout = -1;
     for (bh = ctx->first_bh; bh; bh = bh->next) {
         if (!bh->deleted && bh->scheduled) {
             if (bh->idle) {
                 /* idle bottom halves will be polled at least
                  * every 10ms */
-                *timeout = 10;
+                timeout = 10000000;
             } else {
                 /* non-idle bottom halves will be executed
                  * immediately */
-                *timeout = 0;
-                return true;
+                return 0;
             }
         }
     }
 
-    deadline = qemu_timeout_ns_to_ms(timerlistgroup_deadline_ns(&ctx->tlg));
+    deadline = timerlistgroup_deadline_ns(&ctx->tlg);
     if (deadline == 0) {
-        *timeout = 0;
-        return true;
+        return 0;
     } else {
-        *timeout = qemu_soonest_timeout(*timeout, deadline);
+        return qemu_soonest_timeout(timeout, deadline);
+    }
+}
+
+static gboolean
+aio_ctx_prepare(GSource *source, gint    *timeout)
+{
+    AioContext *ctx = (AioContext *) source;
+
+    /* We assume there is no timeout already supplied */
+    *timeout = qemu_timeout_ns_to_ms(aio_compute_timeout(ctx));
+
+    if (aio_prepare(ctx)) {
+        *timeout = 0;
     }
 
-    return false;
+    return *timeout == 0;
 }
 
 static gboolean
@@ -208,7 +218,7 @@ aio_ctx_dispatch(GSource     *source,
     AioContext *ctx = (AioContext *) source;
 
     assert(callback == NULL);
-    aio_poll(ctx, false);
+    aio_dispatch(ctx);
     return true;
 }
 
@@ -247,9 +257,25 @@ ThreadPool *aio_get_thread_pool(AioContext *ctx)
     return ctx->thread_pool;
 }
 
+void aio_set_dispatching(AioContext *ctx, bool dispatching)
+{
+    ctx->dispatching = dispatching;
+    if (!dispatching) {
+        /* Write ctx->dispatching before reading e.g. bh->scheduled.
+         * Optimization: this is only needed when we're entering the "unsafe"
+         * phase where other threads must call event_notifier_set.
+         */
+        smp_mb();
+    }
+}
+
 void aio_notify(AioContext *ctx)
 {
-    event_notifier_set(&ctx->notifier);
+    /* Write e.g. bh->scheduled before reading ctx->dispatching.  */
+    smp_mb();
+    if (!ctx->dispatching) {
+        event_notifier_set(&ctx->notifier);
+    }
 }
 
 static void aio_timerlist_notify(void *opaque)

backends/Makefile.objs

@@ -1,7 +1,7 @@
 common-obj-y += rng.o rng-egd.o
 common-obj-$(CONFIG_POSIX) += rng-random.o
 
-common-obj-y += msmouse.o
+common-obj-y += msmouse.o testdev.o
 common-obj-$(CONFIG_BRLAPI) += baum.o
 baum.o-cflags := $(SDL_CFLAGS)
 

backends/hostmem.c

@@ -257,15 +257,6 @@ static void host_memory_backend_init(Object *obj)
                         host_memory_backend_set_policy, NULL, NULL, NULL);
 }
 
-static void host_memory_backend_finalize(Object *obj)
-{
-    HostMemoryBackend *backend = MEMORY_BACKEND(obj);
-
-    if (memory_region_size(&backend->mr)) {
-        memory_region_destroy(&backend->mr);
-    }
-}
-
 MemoryRegion *
 host_memory_backend_get_memory(HostMemoryBackend *backend, Error **errp)
 {
@@ -304,7 +295,7 @@ host_memory_backend_memory_complete(UserCreatable *uc, Error **errp)
         /* ensure policy won't be ignored in case memory is preallocated
          * before mbind(). note: MPOL_MF_STRICT is ignored on hugepages so
          * this doesn't catch hugepage case. */
-        unsigned flags = MPOL_MF_STRICT;
+        unsigned flags = MPOL_MF_STRICT | MPOL_MF_MOVE;
 
         /* check for invalid host-nodes and policies and give more verbose
          * error messages than mbind(). */
@@ -360,7 +351,6 @@ static const TypeInfo host_memory_backend_info = {
     .class_init = host_memory_backend_class_init,
     .instance_size = sizeof(HostMemoryBackend),
     .instance_init = host_memory_backend_init,
-    .instance_finalize = host_memory_backend_finalize,
     .interfaces = (InterfaceInfo[]) {
         { TYPE_USER_CREATABLE },
         { }

backends/testdev.c

@@ -0,0 +1,131 @@
+/*
+ * QEMU Char Device for testsuite control
+ *
+ * Copyright (c) 2014 Red Hat, Inc.
+ *
+ * Author: Paolo Bonzini <pbonzini@redhat.com>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+#include "qemu-common.h"
+#include "sysemu/char.h"
+
+#define BUF_SIZE 32
+
+typedef struct {
+    CharDriverState *chr;
+    uint8_t in_buf[32];
+    int in_buf_used;
+} TestdevCharState;
+
+/* Try to interpret a whole incoming packet */
+static int testdev_eat_packet(TestdevCharState *testdev)
+{
+    const uint8_t *cur = testdev->in_buf;
+    int len = testdev->in_buf_used;
+    uint8_t c;
+    int arg;
+
+#define EAT(c) do { \
+    if (!len--) {   \
+        return 0;   \
+    }               \
+    c = *cur++;     \
+} while (0)
+
+    EAT(c);
+
+    while (isspace(c)) {
+        EAT(c);
+    }
+
+    arg = 0;
+    while (isdigit(c)) {
+        arg = arg * 10 + c - '0';
+        EAT(c);
+    }
+
+    while (isspace(c)) {
+        EAT(c);
+    }
+
+    switch (c) {
+    case 'q':
+        exit((arg << 1) | 1);
+        break;
+    default:
+        break;
+    }
+    return cur - testdev->in_buf;
+}
+
+/* The other end is writing some data.  Store it and try to interpret */
+static int testdev_write(CharDriverState *chr, const uint8_t *buf, int len)
+{
+    TestdevCharState *testdev = chr->opaque;
+    int tocopy, eaten, orig_len = len;
+
+    while (len) {
+        /* Complete our buffer as much as possible */
+        tocopy = MIN(len, BUF_SIZE - testdev->in_buf_used);
+
+        memcpy(testdev->in_buf + testdev->in_buf_used, buf, tocopy);
+        testdev->in_buf_used += tocopy;
+        buf += tocopy;
+        len -= tocopy;
+
+        /* Interpret it as much as possible */
+        while (testdev->in_buf_used > 0 &&
+               (eaten = testdev_eat_packet(testdev)) > 0) {
+            memmove(testdev->in_buf, testdev->in_buf + eaten,
+                    testdev->in_buf_used - eaten);
+            testdev->in_buf_used -= eaten;
+        }
+    }
+    return orig_len;
+}
+
+static void testdev_close(struct CharDriverState *chr)
+{
+    TestdevCharState *testdev = chr->opaque;
+
+    g_free(testdev);
+}
+
+CharDriverState *chr_testdev_init(void)
+{
+    TestdevCharState *testdev;
+    CharDriverState *chr;
+
+    testdev = g_malloc0(sizeof(TestdevCharState));
+    testdev->chr = chr = g_malloc0(sizeof(CharDriverState));
+
+    chr->opaque = testdev;
+    chr->chr_write = testdev_write;
+    chr->chr_close = testdev_close;
+
+    return chr;
+}
+
+static void register_types(void)
+{
+    register_char_driver_qapi("testdev", CHARDEV_BACKEND_KIND_TESTDEV, NULL);
+}
+
+type_init(register_types);

block-migration.c

@@ -186,7 +186,7 @@ static int bmds_aio_inflight(BlkMigDevState *bmds, int64_t sector)
 {
     int64_t chunk = sector / (int64_t)BDRV_SECTORS_PER_DIRTY_CHUNK;
 
-    if ((sector << BDRV_SECTOR_BITS) < bdrv_getlength(bmds->bs)) {
+    if (sector < bdrv_nb_sectors(bmds->bs)) {
         return !!(bmds->aio_bitmap[chunk / (sizeof(unsigned long) * 8)] &
             (1UL << (chunk % (sizeof(unsigned long) * 8))));
     } else {
@@ -223,8 +223,7 @@ static void alloc_aio_bitmap(BlkMigDevState *bmds)
     BlockDriverState *bs = bmds->bs;
     int64_t bitmap_size;
 
-    bitmap_size = (bdrv_getlength(bs) >> BDRV_SECTOR_BITS) +
-            BDRV_SECTORS_PER_DIRTY_CHUNK * 8 - 1;
+    bitmap_size = bdrv_nb_sectors(bs) + BDRV_SECTORS_PER_DIRTY_CHUNK * 8 - 1;
     bitmap_size /= BDRV_SECTORS_PER_DIRTY_CHUNK * 8;
 
     bmds->aio_bitmap = g_malloc0(bitmap_size);
@@ -284,7 +283,7 @@ static int mig_save_device_bulk(QEMUFile *f, BlkMigDevState *bmds)
         nr_sectors = total_sectors - cur_sector;
     }
 
-    blk = g_malloc(sizeof(BlkMigBlock));
+    blk = g_new(BlkMigBlock, 1);
     blk->buf = g_malloc(BLOCK_SIZE);
     blk->bmds = bmds;
     blk->sector = cur_sector;
@@ -350,12 +349,12 @@ static void init_blk_migration_it(void *opaque, BlockDriverState *bs)
     int64_t sectors;
 
     if (!bdrv_is_read_only(bs)) {
-        sectors = bdrv_getlength(bs) >> BDRV_SECTOR_BITS;
+        sectors = bdrv_nb_sectors(bs);
         if (sectors <= 0) {
             return;
         }
 
-        bmds = g_malloc0(sizeof(BlkMigDevState));
+        bmds = g_new0(BlkMigDevState, 1);
         bmds->bs = bs;
         bmds->bulk_completed = 0;
         bmds->total_sectors = sectors;
@@ -466,7 +465,7 @@ static int mig_save_device_dirty(QEMUFile *f, BlkMigDevState *bmds,
             } else {
                 nr_sectors = BDRV_SECTORS_PER_DIRTY_CHUNK;
             }
-            blk = g_malloc(sizeof(BlkMigBlock));
+            blk = g_new(BlkMigBlock, 1);
             blk->buf = g_malloc(BLOCK_SIZE);
             blk->bmds = bmds;
             blk->sector = sector;
@@ -799,7 +798,7 @@ static int block_load(QEMUFile *f, void *opaque, int version_id)
 
             if (bs != bs_prev) {
                 bs_prev = bs;
-                total_sectors = bdrv_getlength(bs) >> BDRV_SECTOR_BITS;
+                total_sectors = bdrv_nb_sectors(bs);
                 if (total_sectors <= 0) {
                     error_report("Error getting length of block device %s",
                                  device_name);
@@ -861,7 +860,7 @@ static bool block_is_active(void *opaque)
     return block_mig_state.blk_enable == 1;
 }
 
-SaveVMHandlers savevm_block_handlers = {
+static SaveVMHandlers savevm_block_handlers = {
     .set_params = block_set_params,
     .save_live_setup = block_save_setup,
     .save_live_iterate = block_save_iterate,

block/Makefile.objs

@@ -10,15 +10,14 @@ block-obj-$(CONFIG_WIN32) += raw-win32.o win32-aio.o
 block-obj-$(CONFIG_POSIX) += raw-posix.o
 block-obj-$(CONFIG_LINUX_AIO) += linux-aio.o
 
-ifeq ($(CONFIG_POSIX),y)
 block-obj-y += nbd.o nbd-client.o sheepdog.o
 block-obj-$(CONFIG_LIBISCSI) += iscsi.o
 block-obj-$(CONFIG_LIBNFS) += nfs.o
 block-obj-$(CONFIG_CURL) += curl.o
 block-obj-$(CONFIG_RBD) += rbd.o
 block-obj-$(CONFIG_GLUSTERFS) += gluster.o
+block-obj-$(CONFIG_ARCHIPELAGO) += archipelago.o
 block-obj-$(CONFIG_LIBSSH2) += ssh.o
-endif
 
 common-obj-y += stream.o
 common-obj-y += commit.o
@@ -35,5 +34,6 @@ gluster.o-cflags   := $(GLUSTERFS_CFLAGS)
 gluster.o-libs     := $(GLUSTERFS_LIBS)
 ssh.o-cflags       := $(LIBSSH2_CFLAGS)
 ssh.o-libs         := $(LIBSSH2_LIBS)
+archipelago.o-libs := $(ARCHIPELAGO_LIBS)
 qcow.o-libs        := -lz
 linux-aio.o-libs   := -laio

block/backup.c

@@ -307,7 +307,7 @@ static void coroutine_fn backup_run(void *opaque)
                                 BACKUP_SECTORS_PER_CLUSTER - i, &n);
                     i += n;
 
-                    if (alloced == 1) {
+                    if (alloced == 1 || n == 0) {
                         break;
                     }
                 }

block/blkdebug.c

@@ -26,6 +26,10 @@
 #include "qemu/config-file.h"
 #include "block/block_int.h"
 #include "qemu/module.h"
+#include "qapi/qmp/qbool.h"
+#include "qapi/qmp/qdict.h"
+#include "qapi/qmp/qint.h"
+#include "qapi/qmp/qstring.h"
 
 typedef struct BDRVBlkdebugState {
     int state;
@@ -449,6 +453,10 @@ static void error_callback_bh(void *opaque)
 static void blkdebug_aio_cancel(BlockDriverAIOCB *blockacb)
 {
     BlkdebugAIOCB *acb = container_of(blockacb, BlkdebugAIOCB, common);
+    if (acb->bh) {
+        qemu_bh_delete(acb->bh);
+        acb->bh = NULL;
+    }
     qemu_aio_release(acb);
 }
 
@@ -522,6 +530,25 @@ static BlockDriverAIOCB *blkdebug_aio_writev(BlockDriverState *bs,
     return bdrv_aio_writev(bs->file, sector_num, qiov, nb_sectors, cb, opaque);
 }
 
+static BlockDriverAIOCB *blkdebug_aio_flush(BlockDriverState *bs,
+    BlockDriverCompletionFunc *cb, void *opaque)
+{
+    BDRVBlkdebugState *s = bs->opaque;
+    BlkdebugRule *rule = NULL;
+
+    QSIMPLEQ_FOREACH(rule, &s->active_rules, active_next) {
+        if (rule->options.inject.sector == -1) {
+            break;
+        }
+    }
+
+    if (rule && rule->options.inject.error) {
+        return inject_error(bs, cb, opaque, rule);
+    }
+
+    return bdrv_aio_flush(bs->file, cb, opaque);
+}
+
 
 static void blkdebug_close(BlockDriverState *bs)
 {
@@ -687,6 +714,98 @@ static int64_t blkdebug_getlength(BlockDriverState *bs)
     return bdrv_getlength(bs->file);
 }
 
+static void blkdebug_refresh_filename(BlockDriverState *bs)
+{
+    BDRVBlkdebugState *s = bs->opaque;
+    struct BlkdebugRule *rule;
+    QDict *opts;
+    QList *inject_error_list = NULL, *set_state_list = NULL;
+    QList *suspend_list = NULL;
+    int event;
+
+    if (!bs->file->full_open_options) {
+        /* The config file cannot be recreated, so creating a plain filename
+         * is impossible */
+        return;
+    }
+
+    opts = qdict_new();
+    qdict_put_obj(opts, "driver", QOBJECT(qstring_from_str("blkdebug")));
+
+    QINCREF(bs->file->full_open_options);
+    qdict_put_obj(opts, "image", QOBJECT(bs->file->full_open_options));
+
+    for (event = 0; event < BLKDBG_EVENT_MAX; event++) {
+        QLIST_FOREACH(rule, &s->rules[event], next) {
+            if (rule->action == ACTION_INJECT_ERROR) {
+                QDict *inject_error = qdict_new();
+
+                qdict_put_obj(inject_error, "event", QOBJECT(qstring_from_str(
+                              BlkdebugEvent_lookup[rule->event])));
+                qdict_put_obj(inject_error, "state",
+                              QOBJECT(qint_from_int(rule->state)));
+                qdict_put_obj(inject_error, "errno", QOBJECT(qint_from_int(
+                              rule->options.inject.error)));
+                qdict_put_obj(inject_error, "sector", QOBJECT(qint_from_int(
+                              rule->options.inject.sector)));
+                qdict_put_obj(inject_error, "once", QOBJECT(qbool_from_int(
+                              rule->options.inject.once)));
+                qdict_put_obj(inject_error, "immediately",
+                              QOBJECT(qbool_from_int(
+                              rule->options.inject.immediately)));
+
+                if (!inject_error_list) {
+                    inject_error_list = qlist_new();
+                }
+
+                qlist_append_obj(inject_error_list, QOBJECT(inject_error));
+            } else if (rule->action == ACTION_SET_STATE) {
+                QDict *set_state = qdict_new();
+
+                qdict_put_obj(set_state, "event", QOBJECT(qstring_from_str(
+                              BlkdebugEvent_lookup[rule->event])));
+                qdict_put_obj(set_state, "state",
+                              QOBJECT(qint_from_int(rule->state)));
+                qdict_put_obj(set_state, "new_state", QOBJECT(qint_from_int(
+                              rule->options.set_state.new_state)));
+
+                if (!set_state_list) {
+                    set_state_list = qlist_new();
+                }
+
+                qlist_append_obj(set_state_list, QOBJECT(set_state));
+            } else if (rule->action == ACTION_SUSPEND) {
+                QDict *suspend = qdict_new();
+
+                qdict_put_obj(suspend, "event", QOBJECT(qstring_from_str(
+                              BlkdebugEvent_lookup[rule->event])));
+                qdict_put_obj(suspend, "state",
+                              QOBJECT(qint_from_int(rule->state)));
+                qdict_put_obj(suspend, "tag", QOBJECT(qstring_from_str(
+                              rule->options.suspend.tag)));
+
+                if (!suspend_list) {
+                    suspend_list = qlist_new();
+                }
+
+                qlist_append_obj(suspend_list, QOBJECT(suspend));
+            }
+        }
+    }
+
+    if (inject_error_list) {
+        qdict_put_obj(opts, "inject-error", QOBJECT(inject_error_list));
+    }
+    if (set_state_list) {
+        qdict_put_obj(opts, "set-state", QOBJECT(set_state_list));
+    }
+    if (suspend_list) {
+        qdict_put_obj(opts, "suspend", QOBJECT(suspend_list));
+    }
+
+    bs->full_open_options = opts;
+}
+
 static BlockDriver bdrv_blkdebug = {
     .format_name            = "blkdebug",
     .protocol_name          = "blkdebug",
@@ -696,9 +815,11 @@ static BlockDriver bdrv_blkdebug = {
     .bdrv_file_open         = blkdebug_open,
     .bdrv_close             = blkdebug_close,
     .bdrv_getlength         = blkdebug_getlength,
+    .bdrv_refresh_filename  = blkdebug_refresh_filename,
 
     .bdrv_aio_readv         = blkdebug_aio_readv,
     .bdrv_aio_writev        = blkdebug_aio_writev,
+    .bdrv_aio_flush         = blkdebug_aio_flush,
 
     .bdrv_debug_event           = blkdebug_debug_event,
     .bdrv_debug_breakpoint      = blkdebug_debug_breakpoint,

block/blkverify.c

@@ -10,6 +10,8 @@
 #include <stdarg.h>
 #include "qemu/sockets.h" /* for EINPROGRESS on Windows */
 #include "block/block_int.h"
+#include "qapi/qmp/qdict.h"
+#include "qapi/qmp/qstring.h"
 
 typedef struct {
     BlockDriverState *test_file;
@@ -156,6 +158,7 @@ static int blkverify_open(BlockDriverState *bs, QDict *options, int flags,
 
     ret = 0;
 fail:
+    qemu_opts_del(opts);
     return ret;
 }
 
@@ -320,6 +323,32 @@ static void blkverify_attach_aio_context(BlockDriverState *bs,
     bdrv_attach_aio_context(s->test_file, new_context);
 }
 
+static void blkverify_refresh_filename(BlockDriverState *bs)
+{
+    BDRVBlkverifyState *s = bs->opaque;
+
+    /* bs->file has already been refreshed */
+    bdrv_refresh_filename(s->test_file);
+
+    if (bs->file->full_open_options && s->test_file->full_open_options) {
+        QDict *opts = qdict_new();
+        qdict_put_obj(opts, "driver", QOBJECT(qstring_from_str("blkverify")));
+
+        QINCREF(bs->file->full_open_options);
+        qdict_put_obj(opts, "raw", QOBJECT(bs->file->full_open_options));
+        QINCREF(s->test_file->full_open_options);
+        qdict_put_obj(opts, "test", QOBJECT(s->test_file->full_open_options));
+
+        bs->full_open_options = opts;
+    }
+
+    if (bs->file->exact_filename[0] && s->test_file->exact_filename[0]) {
+        snprintf(bs->exact_filename, sizeof(bs->exact_filename),
+                 "blkverify:%s:%s",
+                 bs->file->exact_filename, s->test_file->exact_filename);
+    }
+}
+
 static BlockDriver bdrv_blkverify = {
     .format_name                      = "blkverify",
     .protocol_name                    = "blkverify",
@@ -329,6 +358,7 @@ static BlockDriver bdrv_blkverify = {
     .bdrv_file_open                   = blkverify_open,
     .bdrv_close                       = blkverify_close,
     .bdrv_getlength                   = blkverify_getlength,
+    .bdrv_refresh_filename            = blkverify_refresh_filename,
 
     .bdrv_aio_readv                   = blkverify_aio_readv,
     .bdrv_aio_writev                  = blkverify_aio_writev,

block/bochs.c

@@ -131,7 +131,11 @@ static int bochs_open(BlockDriverState *bs, QDict *options, int flags,
         return -EFBIG;
     }
 
-    s->catalog_bitmap = g_malloc(s->catalog_size * 4);
+    s->catalog_bitmap = g_try_new(uint32_t, s->catalog_size);
+    if (s->catalog_size && s->catalog_bitmap == NULL) {
+        error_setg(errp, "Could not allocate memory for catalog");
+        return -ENOMEM;
+    }
 
     ret = bdrv_pread(bs->file, le32_to_cpu(bochs.header), s->catalog_bitmap,
                      s->catalog_size * 4);

block/cloop.c

@@ -116,7 +116,12 @@ static int cloop_open(BlockDriverState *bs, QDict *options, int flags,
                    "try increasing block size");
         return -EINVAL;
     }
-    s->offsets = g_malloc(offsets_size);
+
+    s->offsets = g_try_malloc(offsets_size);
+    if (s->offsets == NULL) {
+        error_setg(errp, "Could not allocate offsets table");
+        return -ENOMEM;
+    }
 
     ret = bdrv_pread(bs->file, 128 + 4 + 4, s->offsets, offsets_size);
     if (ret < 0) {
@@ -158,8 +163,20 @@ static int cloop_open(BlockDriverState *bs, QDict *options, int flags,
     }
 
     /* initialize zlib engine */
-    s->compressed_block = g_malloc(max_compressed_block_size + 1);
-    s->uncompressed_block = g_malloc(s->block_size);
+    s->compressed_block = g_try_malloc(max_compressed_block_size + 1);
+    if (s->compressed_block == NULL) {
+        error_setg(errp, "Could not allocate compressed_block");
+        ret = -ENOMEM;
+        goto fail;
+    }
+
+    s->uncompressed_block = g_try_malloc(s->block_size);
+    if (s->uncompressed_block == NULL) {
+        error_setg(errp, "Could not allocate uncompressed_block");
+        ret = -ENOMEM;
+        goto fail;
+    }
+
     if (inflateInit(&s->zstream) != Z_OK) {
         ret = -EINVAL;
         goto fail;

block/commit.c

@@ -37,6 +37,7 @@ typedef struct CommitBlockJob {
     BlockdevOnError on_error;
     int base_flags;
     int orig_overlay_flags;
+    char *backing_file_str;
 } CommitBlockJob;
 
 static int coroutine_fn commit_populate(BlockDriverState *bs,
@@ -141,7 +142,7 @@ wait:
 
     if (!block_job_is_cancelled(&s->common) && sector_num == end) {
         /* success */
-        ret = bdrv_drop_intermediate(active, top, base);
+        ret = bdrv_drop_intermediate(active, top, base, s->backing_file_str);
     }
 
 exit_free_buf:
@@ -158,7 +159,7 @@ exit_restore_reopen:
     if (overlay_bs && s->orig_overlay_flags != bdrv_get_flags(overlay_bs)) {
         bdrv_reopen(overlay_bs, s->orig_overlay_flags, NULL);
     }
-
+    g_free(s->backing_file_str);
     block_job_completed(&s->common, ret);
 }
 
@@ -182,7 +183,7 @@ static const BlockJobDriver commit_job_driver = {
 void commit_start(BlockDriverState *bs, BlockDriverState *base,
                   BlockDriverState *top, int64_t speed,
                   BlockdevOnError on_error, BlockDriverCompletionFunc *cb,
-                  void *opaque, Error **errp)
+                  void *opaque, const char *backing_file_str, Error **errp)
 {
     CommitBlockJob *s;
     BlockReopenQueue *reopen_queue = NULL;
@@ -244,6 +245,8 @@ void commit_start(BlockDriverState *bs, BlockDriverState *base,
     s->base_flags          = orig_base_flags;
     s->orig_overlay_flags  = orig_overlay_flags;
 
+    s->backing_file_str = g_strdup(backing_file_str);
+
     s->on_error = on_error;
     s->common.co = qemu_coroutine_create(commit_run);
 

block/cow.c

@@ -332,7 +332,7 @@ static int cow_create(const char *filename, QemuOpts *opts, Error **errp)
     char *image_filename = NULL;
     Error *local_err = NULL;
     int ret;
-    BlockDriverState *cow_bs;
+    BlockDriverState *cow_bs = NULL;
 
     /* Read out options */
     image_sectors = qemu_opt_get_size_del(opts, BLOCK_OPT_SIZE, 0) / 512;
@@ -344,7 +344,6 @@ static int cow_create(const char *filename, QemuOpts *opts, Error **errp)
         goto exit;
     }
 
-    cow_bs = NULL;
     ret = bdrv_open(&cow_bs, filename, NULL, NULL,
                     BDRV_O_RDWR | BDRV_O_PROTOCOL, NULL, &local_err);
     if (ret < 0) {
@@ -383,7 +382,9 @@ static int cow_create(const char *filename, QemuOpts *opts, Error **errp)
 
 exit:
     g_free(image_filename);
-    bdrv_unref(cow_bs);
+    if (cow_bs) {
+        bdrv_unref(cow_bs);
+    }
     return ret;
 }
 
@@ -414,6 +415,7 @@ static BlockDriver bdrv_cow = {
     .bdrv_close     = cow_close,
     .bdrv_create    = cow_create,
     .bdrv_has_zero_init     = bdrv_has_zero_init_1,
+    .supports_backing       = true,
 
     .bdrv_read              = cow_co_read,
     .bdrv_write             = cow_co_write,

block/curl.c

@@ -26,7 +26,7 @@
 #include "qapi/qmp/qbool.h"
 #include <curl/curl.h>
 
-// #define DEBUG
+// #define DEBUG_CURL
 // #define DEBUG_VERBOSE
 
 #ifdef DEBUG_CURL
@@ -63,6 +63,7 @@ static CURLMcode __curl_multi_socket_action(CURLM *multi_handle,
 #define CURL_NUM_ACB    8
 #define SECTOR_SIZE     512
 #define READ_AHEAD_DEFAULT (256 * 1024)
+#define CURL_TIMEOUT_DEFAULT 5
 
 #define FIND_RET_NONE   0
 #define FIND_RET_OK     1
@@ -71,6 +72,8 @@ static CURLMcode __curl_multi_socket_action(CURLM *multi_handle,
 #define CURL_BLOCK_OPT_URL       "url"
 #define CURL_BLOCK_OPT_READAHEAD "readahead"
 #define CURL_BLOCK_OPT_SSLVERIFY "sslverify"
+#define CURL_BLOCK_OPT_TIMEOUT "timeout"
+#define CURL_BLOCK_OPT_COOKIE    "cookie"
 
 struct BDRVCURLState;
 
@@ -109,6 +112,8 @@ typedef struct BDRVCURLState {
     char *url;
     size_t readahead_size;
     bool sslverify;
+    int timeout;
+    char *cookie;
     bool accept_range;
     AioContext *aio_context;
 } BDRVCURLState;
@@ -352,7 +357,7 @@ static void curl_multi_timeout_do(void *arg)
 #endif
 }
 
-static CURLState *curl_init_state(BDRVCURLState *s)
+static CURLState *curl_init_state(BlockDriverState *bs, BDRVCURLState *s)
 {
     CURLState *state = NULL;
     int i, j;
@@ -370,7 +375,7 @@ static CURLState *curl_init_state(BDRVCURLState *s)
             break;
         }
         if (!state) {
-            aio_poll(state->s->aio_context, true);
+            aio_poll(bdrv_get_aio_context(bs), true);
         }
     } while(!state);
 
@@ -382,7 +387,10 @@ static CURLState *curl_init_state(BDRVCURLState *s)
         curl_easy_setopt(state->curl, CURLOPT_URL, s->url);
         curl_easy_setopt(state->curl, CURLOPT_SSL_VERIFYPEER,
                          (long) s->sslverify);
-        curl_easy_setopt(state->curl, CURLOPT_TIMEOUT, 5);
+        if (s->cookie) {
+            curl_easy_setopt(state->curl, CURLOPT_COOKIE, s->cookie);
+        }
+        curl_easy_setopt(state->curl, CURLOPT_TIMEOUT, s->timeout);
         curl_easy_setopt(state->curl, CURLOPT_WRITEFUNCTION,
                          (void *)curl_read_cb);
         curl_easy_setopt(state->curl, CURLOPT_WRITEDATA, (void *)state);
@@ -489,6 +497,16 @@ static QemuOptsList runtime_opts = {
             .type = QEMU_OPT_BOOL,
             .help = "Verify SSL certificate"
         },
+        {
+            .name = CURL_BLOCK_OPT_TIMEOUT,
+            .type = QEMU_OPT_NUMBER,
+            .help = "Curl timeout"
+        },
+        {
+            .name = CURL_BLOCK_OPT_COOKIE,
+            .type = QEMU_OPT_STRING,
+            .help = "Pass the cookie or list of cookies with each request"
+        },
         { /* end of list */ }
     },
 };
@@ -501,6 +519,7 @@ static int curl_open(BlockDriverState *bs, QDict *options, int flags,
     QemuOpts *opts;
     Error *local_err = NULL;
     const char *file;
+    const char *cookie;
     double d;
 
     static int inited = 0;
@@ -525,8 +544,14 @@ static int curl_open(BlockDriverState *bs, QDict *options, int flags,
         goto out_noclean;
     }
 
+    s->timeout = qemu_opt_get_number(opts, CURL_BLOCK_OPT_TIMEOUT,
+                                     CURL_TIMEOUT_DEFAULT);
+
     s->sslverify = qemu_opt_get_bool(opts, CURL_BLOCK_OPT_SSLVERIFY, true);
 
+    cookie = qemu_opt_get(opts, CURL_BLOCK_OPT_COOKIE);
+    s->cookie = g_strdup(cookie);
+
     file = qemu_opt_get(opts, CURL_BLOCK_OPT_URL);
     if (file == NULL) {
         error_setg(errp, "curl block driver requires an 'url' option");
@@ -541,7 +566,7 @@ static int curl_open(BlockDriverState *bs, QDict *options, int flags,
     DPRINTF("CURL: Opening %s\n", file);
     s->aio_context = bdrv_get_aio_context(bs);
     s->url = g_strdup(file);
-    state = curl_init_state(s);
+    state = curl_init_state(bs, s);
     if (!state)
         goto out_noclean;
 
@@ -582,6 +607,7 @@ out:
     curl_easy_cleanup(state->curl);
     state->curl = NULL;
 out_noclean:
+    g_free(s->cookie);
     g_free(s->url);
     qemu_opts_del(opts);
     return -EINVAL;
@@ -625,7 +651,7 @@ static void curl_readv_bh_cb(void *p)
     }
 
     // No cache found, so let's start a new request
-    state = curl_init_state(s);
+    state = curl_init_state(acb->common.bs, s);
     if (!state) {
         acb->common.cb(acb->common.opaque, -EIO);
         qemu_aio_release(acb);
@@ -640,7 +666,13 @@ static void curl_readv_bh_cb(void *p)
     state->buf_start = start;
     state->buf_len = acb->end + s->readahead_size;
     end = MIN(start + state->buf_len, s->len) - 1;
-    state->orig_buf = g_malloc(state->buf_len);
+    state->orig_buf = g_try_malloc(state->buf_len);
+    if (state->buf_len && state->orig_buf == NULL) {
+        curl_clean_state(state);
+        acb->common.cb(acb->common.opaque, -ENOMEM);
+        qemu_aio_release(acb);
+        return;
+    }
     state->acb[0] = acb;
 
     snprintf(state->range, 127, "%zd-%zd", start, end);
@@ -678,6 +710,7 @@ static void curl_close(BlockDriverState *bs)
     DPRINTF("CURL: Close\n");
     curl_detach_aio_context(bs);
 
+    g_free(s->cookie);
     g_free(s->url);
 }
 

block/dmg.c

@@ -284,8 +284,15 @@ static int dmg_open(BlockDriverState *bs, QDict *options, int flags,
     }
 
     /* initialize zlib engine */
-    s->compressed_chunk = g_malloc(max_compressed_size + 1);
-    s->uncompressed_chunk = g_malloc(512 * max_sectors_per_chunk);
+    s->compressed_chunk = qemu_try_blockalign(bs->file,
+                                              max_compressed_size + 1);
+    s->uncompressed_chunk = qemu_try_blockalign(bs->file,
+                                                512 * max_sectors_per_chunk);
+    if (s->compressed_chunk == NULL || s->uncompressed_chunk == NULL) {
+        ret = -ENOMEM;
+        goto fail;
+    }
+
     if (inflateInit(&s->zstream) != Z_OK) {
         ret = -EINVAL;
         goto fail;
@@ -302,8 +309,8 @@ fail:
     g_free(s->lengths);
     g_free(s->sectors);
     g_free(s->sectorcounts);
-    g_free(s->compressed_chunk);
-    g_free(s->uncompressed_chunk);
+    qemu_vfree(s->compressed_chunk);
+    qemu_vfree(s->uncompressed_chunk);
     return ret;
 }
 
@@ -426,8 +433,8 @@ static void dmg_close(BlockDriverState *bs)
     g_free(s->lengths);
     g_free(s->sectors);
     g_free(s->sectorcounts);
-    g_free(s->compressed_chunk);
-    g_free(s->uncompressed_chunk);
+    qemu_vfree(s->compressed_chunk);
+    qemu_vfree(s->uncompressed_chunk);
 
     inflateEnd(&s->zstream);
 }

block/gluster.c

@@ -291,7 +291,7 @@ static int qemu_gluster_open(BlockDriverState *bs,  QDict *options,
     BDRVGlusterState *s = bs->opaque;
     int open_flags = 0;
     int ret = 0;
-    GlusterConf *gconf = g_malloc0(sizeof(GlusterConf));
+    GlusterConf *gconf = g_new0(GlusterConf, 1);
     QemuOpts *opts;
     Error *local_err = NULL;
     const char *filename;
@@ -351,12 +351,12 @@ static int qemu_gluster_reopen_prepare(BDRVReopenState *state,
     assert(state != NULL);
     assert(state->bs != NULL);
 
-    state->opaque = g_malloc0(sizeof(BDRVGlusterReopenState));
+    state->opaque = g_new0(BDRVGlusterReopenState, 1);
     reop_s = state->opaque;
 
     qemu_gluster_parse_flags(state->flags, &open_flags);
 
-    gconf = g_malloc0(sizeof(GlusterConf));
+    gconf = g_new0(GlusterConf, 1);
 
     reop_s->glfs = qemu_gluster_init(gconf, state->bs->filename, errp);
     if (reop_s->glfs == NULL) {
@@ -486,7 +486,7 @@ static int qemu_gluster_create(const char *filename,
     int prealloc = 0;
     int64_t total_size = 0;
     char *tmp = NULL;
-    GlusterConf *gconf = g_malloc0(sizeof(GlusterConf));
+    GlusterConf *gconf = g_new0(GlusterConf, 1);
 
     glfs = qemu_gluster_init(gconf, filename, errp);
     if (!glfs) {

block/iscsi.c

@@ -893,7 +893,10 @@ coroutine_fn iscsi_co_write_zeroes(BlockDriverState *bs, int64_t sector_num,
     nb_blocks = sector_qemu2lun(nb_sectors, iscsilun);
 
     if (iscsilun->zeroblock == NULL) {
-        iscsilun->zeroblock = g_malloc0(iscsilun->block_size);
+        iscsilun->zeroblock = g_try_malloc0(iscsilun->block_size);
+        if (iscsilun->zeroblock == NULL) {
+            return -ENOMEM;
+        }
     }
 
     iscsi_co_init_iscsitask(iscsilun, &iTask);
@@ -1450,7 +1453,7 @@ static void iscsi_close(BlockDriverState *bs)
     memset(iscsilun, 0, sizeof(IscsiLun));
 }
 
-static int iscsi_refresh_limits(BlockDriverState *bs)
+static void iscsi_refresh_limits(BlockDriverState *bs, Error **errp)
 {
     IscsiLun *iscsilun = bs->opaque;
 
@@ -1475,7 +1478,6 @@ static int iscsi_refresh_limits(BlockDriverState *bs)
     }
     bs->bl.opt_transfer_length = sector_lun2qemu(iscsilun->bl.opt_xfer_len,
                                                  iscsilun);
-    return 0;
 }
 
 /* Since iscsi_open() ignores bdrv_flags, there is nothing to do here in
@@ -1510,7 +1512,8 @@ static int iscsi_truncate(BlockDriverState *bs, int64_t offset)
     if (iscsilun->allocationmap != NULL) {
         g_free(iscsilun->allocationmap);
         iscsilun->allocationmap =
-            bitmap_new(DIV_ROUND_UP(bs->total_sectors,
+            bitmap_new(DIV_ROUND_UP(sector_lun2qemu(iscsilun->num_blocks,
+                                                    iscsilun),
                                     iscsilun->cluster_sectors));
     }
 
@@ -1530,7 +1533,7 @@ static int iscsi_create(const char *filename, QemuOpts *opts, Error **errp)
     /* Read out options */
     total_size =
         qemu_opt_get_size_del(opts, BLOCK_OPT_SIZE, 0) / BDRV_SECTOR_SIZE;
-    bs->opaque = g_malloc0(sizeof(struct IscsiLun));
+    bs->opaque = g_new0(struct IscsiLun, 1);
     iscsilun = bs->opaque;
 
     bs_options = qdict_new();

block/linux-aio.c

@@ -25,6 +25,8 @@
  */
 #define MAX_EVENTS 128
 
+#define MAX_QUEUED_IO  128
+
 struct qemu_laiocb {
     BlockDriverAIOCB common;
     struct qemu_laio_state *ctx;
@@ -36,9 +38,25 @@ struct qemu_laiocb {
     QLIST_ENTRY(qemu_laiocb) node;
 };
 
+typedef struct {
+    struct iocb *iocbs[MAX_QUEUED_IO];
+    int plugged;
+    unsigned int size;
+    unsigned int idx;
+} LaioQueue;
+
 struct qemu_laio_state {
     io_context_t ctx;
     EventNotifier e;
+
+    /* io queue for submit at batch */
+    LaioQueue io_q;
+
+    /* I/O completion processing */
+    QEMUBH *completion_bh;
+    struct io_event events[MAX_EVENTS];
+    int event_idx;
+    int event_max;
 };
 
 static inline ssize_t io_event_ret(struct io_event *ev)
@@ -74,27 +92,58 @@ static void qemu_laio_process_completion(struct qemu_laio_state *s,
     qemu_aio_release(laiocb);
 }
 
+/* The completion BH fetches completed I/O requests and invokes their
+ * callbacks.
+ *
+ * The function is somewhat tricky because it supports nested event loops, for
+ * example when a request callback invokes aio_poll().  In order to do this,
+ * the completion events array and index are kept in qemu_laio_state.  The BH
+ * reschedules itself as long as there are completions pending so it will
+ * either be called again in a nested event loop or will be called after all
+ * events have been completed.  When there are no events left to complete, the
+ * BH returns without rescheduling.
+ */
+static void qemu_laio_completion_bh(void *opaque)
+{
+    struct qemu_laio_state *s = opaque;
+
+    /* Fetch more completion events when empty */
+    if (s->event_idx == s->event_max) {
+        do {
+            struct timespec ts = { 0 };
+            s->event_max = io_getevents(s->ctx, MAX_EVENTS, MAX_EVENTS,
+                                        s->events, &ts);
+        } while (s->event_max == -EINTR);
+
+        s->event_idx = 0;
+        if (s->event_max <= 0) {
+            s->event_max = 0;
+            return; /* no more events */
+        }
+    }
+
+    /* Reschedule so nested event loops see currently pending completions */
+    qemu_bh_schedule(s->completion_bh);
+
+    /* Process completion events */
+    while (s->event_idx < s->event_max) {
+        struct iocb *iocb = s->events[s->event_idx].obj;
+        struct qemu_laiocb *laiocb =
+                container_of(iocb, struct qemu_laiocb, iocb);
+
+        laiocb->ret = io_event_ret(&s->events[s->event_idx]);
+        s->event_idx++;
+
+        qemu_laio_process_completion(s, laiocb);
+    }
+}
+
 static void qemu_laio_completion_cb(EventNotifier *e)
 {
     struct qemu_laio_state *s = container_of(e, struct qemu_laio_state, e);
 
-    while (event_notifier_test_and_clear(&s->e)) {
-        struct io_event events[MAX_EVENTS];
-        struct timespec ts = { 0 };
-        int nevents, i;
-
-        do {
-            nevents = io_getevents(s->ctx, MAX_EVENTS, MAX_EVENTS, events, &ts);
-        } while (nevents == -EINTR);
-
-        for (i = 0; i < nevents; i++) {
-            struct iocb *iocb = events[i].obj;
-            struct qemu_laiocb *laiocb =
-                    container_of(iocb, struct qemu_laiocb, iocb);
-
-            laiocb->ret = io_event_ret(&events[i]);
-            qemu_laio_process_completion(s, laiocb);
-        }
+    if (event_notifier_test_and_clear(&s->e)) {
+        qemu_bh_schedule(s->completion_bh);
     }
 }
 
@@ -135,6 +184,79 @@ static const AIOCBInfo laio_aiocb_info = {
     .cancel             = laio_cancel,
 };
 
+static void ioq_init(LaioQueue *io_q)
+{
+    io_q->size = MAX_QUEUED_IO;
+    io_q->idx = 0;
+    io_q->plugged = 0;
+}
+
+static int ioq_submit(struct qemu_laio_state *s)
+{
+    int ret, i = 0;
+    int len = s->io_q.idx;
+
+    do {
+        ret = io_submit(s->ctx, len, s->io_q.iocbs);
+    } while (i++ < 3 && ret == -EAGAIN);
+
+    /* empty io queue */
+    s->io_q.idx = 0;
+
+    if (ret < 0) {
+        i = 0;
+    } else {
+        i = ret;
+    }
+
+    for (; i < len; i++) {
+        struct qemu_laiocb *laiocb =
+            container_of(s->io_q.iocbs[i], struct qemu_laiocb, iocb);
+
+        laiocb->ret = (ret < 0) ? ret : -EIO;
+        qemu_laio_process_completion(s, laiocb);
+    }
+    return ret;
+}
+
+static void ioq_enqueue(struct qemu_laio_state *s, struct iocb *iocb)
+{
+    unsigned int idx = s->io_q.idx;
+
+    s->io_q.iocbs[idx++] = iocb;
+    s->io_q.idx = idx;
+
+    /* submit immediately if queue is full */
+    if (idx == s->io_q.size) {
+        ioq_submit(s);
+    }
+}
+
+void laio_io_plug(BlockDriverState *bs, void *aio_ctx)
+{
+    struct qemu_laio_state *s = aio_ctx;
+
+    s->io_q.plugged++;
+}
+
+int laio_io_unplug(BlockDriverState *bs, void *aio_ctx, bool unplug)
+{
+    struct qemu_laio_state *s = aio_ctx;
+    int ret = 0;
+
+    assert(s->io_q.plugged > 0 || !unplug);
+
+    if (unplug && --s->io_q.plugged > 0) {
+        return 0;
+    }
+
+    if (s->io_q.idx > 0) {
+        ret = ioq_submit(s);
+    }
+
+    return ret;
+}
+
 BlockDriverAIOCB *laio_submit(BlockDriverState *bs, void *aio_ctx, int fd,
         int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
         BlockDriverCompletionFunc *cb, void *opaque, int type)
@@ -168,8 +290,13 @@ BlockDriverAIOCB *laio_submit(BlockDriverState *bs, void *aio_ctx, int fd,
     }
     io_set_eventfd(&laiocb->iocb, event_notifier_get_fd(&s->e));
 
-    if (io_submit(s->ctx, 1, &iocbs) < 0)
-        goto out_free_aiocb;
+    if (!s->io_q.plugged) {
+        if (io_submit(s->ctx, 1, &iocbs) < 0) {
+            goto out_free_aiocb;
+        }
+    } else {
+        ioq_enqueue(s, iocbs);
+    }
     return &laiocb->common;
 
 out_free_aiocb:
@@ -182,12 +309,14 @@ void laio_detach_aio_context(void *s_, AioContext *old_context)
     struct qemu_laio_state *s = s_;
 
     aio_set_event_notifier(old_context, &s->e, NULL);
+    qemu_bh_delete(s->completion_bh);
 }
 
 void laio_attach_aio_context(void *s_, AioContext *new_context)
 {
     struct qemu_laio_state *s = s_;
 
+    s->completion_bh = aio_bh_new(new_context, qemu_laio_completion_bh, s);
     aio_set_event_notifier(new_context, &s->e, qemu_laio_completion_cb);
 }
 
@@ -204,6 +333,8 @@ void *laio_init(void)
         goto out_close_efd;
     }
 
+    ioq_init(&s->io_q);
+
     return s;
 
 out_close_efd:
@@ -218,5 +349,10 @@ void laio_cleanup(void *s_)
     struct qemu_laio_state *s = s_;
 
     event_notifier_cleanup(&s->e);
+
+    if (io_destroy(s->ctx) != 0) {
+        fprintf(stderr, "%s: destroy AIO context %p failed\n",
+                        __func__, &s->ctx);
+    }
     g_free(s);
 }

block/mirror.c

@@ -32,6 +32,12 @@ typedef struct MirrorBlockJob {
     RateLimit limit;
     BlockDriverState *target;
     BlockDriverState *base;
+    /* The name of the graph node to replace */
+    char *replaces;
+    /* The BDS to replace */
+    BlockDriverState *to_replace;
+    /* Used to block operations on the drive-mirror-replace target */
+    Error *replace_blocker;
     bool is_none_mode;
     BlockdevOnError on_source_error, on_target_error;
     bool synced;
@@ -151,7 +157,7 @@ static uint64_t coroutine_fn mirror_iteration(MirrorBlockJob *s)
     BlockDriverState *source = s->common.bs;
     int nb_sectors, sectors_per_chunk, nb_chunks;
     int64_t end, sector_num, next_chunk, next_sector, hbitmap_next_sector;
-    uint64_t delay_ns;
+    uint64_t delay_ns = 0;
     MirrorOp *op;
 
     s->sector_num = hbitmap_iter_next(&s->hbi);
@@ -241,8 +247,6 @@ static uint64_t coroutine_fn mirror_iteration(MirrorBlockJob *s)
         next_chunk += added_chunks;
         if (!s->synced && s->common.speed) {
             delay_ns = ratelimit_calculate_delay(&s->limit, added_sectors);
-        } else {
-            delay_ns = 0;
         }
     } while (delay_ns == 0 && next_sector < end);
 
@@ -259,9 +263,11 @@ static uint64_t coroutine_fn mirror_iteration(MirrorBlockJob *s)
     next_sector = sector_num;
     while (nb_chunks-- > 0) {
         MirrorBuffer *buf = QSIMPLEQ_FIRST(&s->buf_free);
+        size_t remaining = (nb_sectors * BDRV_SECTOR_SIZE) - op->qiov.size;
+
         QSIMPLEQ_REMOVE_HEAD(&s->buf_free, next);
         s->buf_free_count--;
-        qemu_iovec_add(&op->qiov, buf, s->granularity);
+        qemu_iovec_add(&op->qiov, buf, MIN(s->granularity, remaining));
 
         /* Advance the HBitmapIter in parallel, so that we do not examine
          * the same sector twice.
@@ -324,9 +330,18 @@ static void coroutine_fn mirror_run(void *opaque)
     }
 
     s->common.len = bdrv_getlength(bs);
-    if (s->common.len <= 0) {
+    if (s->common.len < 0) {
         ret = s->common.len;
         goto immediate_exit;
+    } else if (s->common.len == 0) {
+        /* Report BLOCK_JOB_READY and wait for complete. */
+        block_job_event_ready(&s->common);
+        s->synced = true;
+        while (!block_job_is_cancelled(&s->common) && !s->should_complete) {
+            block_job_yield(&s->common);
+        }
+        s->common.cancelled = false;
+        goto immediate_exit;
     }
 
     length = DIV_ROUND_UP(s->common.len, s->granularity);
@@ -350,7 +365,12 @@ static void coroutine_fn mirror_run(void *opaque)
     }
 
     end = s->common.len >> BDRV_SECTOR_BITS;
-    s->buf = qemu_blockalign(bs, s->buf_size);
+    s->buf = qemu_try_blockalign(bs, s->buf_size);
+    if (s->buf == NULL) {
+        ret = -ENOMEM;
+        goto immediate_exit;
+    }
+
     sectors_per_chunk = s->granularity >> BDRV_SECTOR_BITS;
     mirror_free_init(s);
 
@@ -491,10 +511,14 @@ immediate_exit:
     bdrv_release_dirty_bitmap(bs, s->dirty_bitmap);
     bdrv_iostatus_disable(s->target);
     if (s->should_complete && ret == 0) {
-        if (bdrv_get_flags(s->target) != bdrv_get_flags(s->common.bs)) {
-            bdrv_reopen(s->target, bdrv_get_flags(s->common.bs), NULL);
+        BlockDriverState *to_replace = s->common.bs;
+        if (s->to_replace) {
+            to_replace = s->to_replace;
         }
-        bdrv_swap(s->target, s->common.bs);
+        if (bdrv_get_flags(s->target) != bdrv_get_flags(to_replace)) {
+            bdrv_reopen(s->target, bdrv_get_flags(to_replace), NULL);
+        }
+        bdrv_swap(s->target, to_replace);
         if (s->common.driver->job_type == BLOCK_JOB_TYPE_COMMIT) {
             /* drop the bs loop chain formed by the swap: break the loop then
              * trigger the unref from the top one */
@@ -503,6 +527,12 @@ immediate_exit:
             bdrv_unref(p);
         }
     }
+    if (s->to_replace) {
+        bdrv_op_unblock_all(s->to_replace, s->replace_blocker);
+        error_free(s->replace_blocker);
+        bdrv_unref(s->to_replace);
+    }
+    g_free(s->replaces);
     bdrv_unref(s->target);
     block_job_completed(&s->common, ret);
 }
@@ -541,6 +571,20 @@ static void mirror_complete(BlockJob *job, Error **errp)
         return;
     }
 
+    /* check the target bs is not blocked and block all operations on it */
+    if (s->replaces) {
+        s->to_replace = check_to_replace_node(s->replaces, &local_err);
+        if (!s->to_replace) {
+            error_propagate(errp, local_err);
+            return;
+        }
+
+        error_setg(&s->replace_blocker,
+                   "block device is in use by block-job-complete");
+        bdrv_op_block_all(s->to_replace, s->replace_blocker);
+        bdrv_ref(s->to_replace);
+    }
+
     s->should_complete = true;
     block_job_resume(job);
 }
@@ -563,14 +607,15 @@ static const BlockJobDriver commit_active_job_driver = {
 };
 
 static void mirror_start_job(BlockDriverState *bs, BlockDriverState *target,
-                            int64_t speed, int64_t granularity,
-                            int64_t buf_size,
-                            BlockdevOnError on_source_error,
-                            BlockdevOnError on_target_error,
-                            BlockDriverCompletionFunc *cb,
-                            void *opaque, Error **errp,
-                            const BlockJobDriver *driver,
-                            bool is_none_mode, BlockDriverState *base)
+                             const char *replaces,
+                             int64_t speed, int64_t granularity,
+                             int64_t buf_size,
+                             BlockdevOnError on_source_error,
+                             BlockdevOnError on_target_error,
+                             BlockDriverCompletionFunc *cb,
+                             void *opaque, Error **errp,
+                             const BlockJobDriver *driver,
+                             bool is_none_mode, BlockDriverState *base)
 {
     MirrorBlockJob *s;
 
@@ -601,6 +646,7 @@ static void mirror_start_job(BlockDriverState *bs, BlockDriverState *target,
         return;
     }
 
+    s->replaces = g_strdup(replaces);
     s->on_source_error = on_source_error;
     s->on_target_error = on_target_error;
     s->target = target;
@@ -622,6 +668,7 @@ static void mirror_start_job(BlockDriverState *bs, BlockDriverState *target,
 }
 
 void mirror_start(BlockDriverState *bs, BlockDriverState *target,
+                  const char *replaces,
                   int64_t speed, int64_t granularity, int64_t buf_size,
                   MirrorSyncMode mode, BlockdevOnError on_source_error,
                   BlockdevOnError on_target_error,
@@ -633,7 +680,8 @@ void mirror_start(BlockDriverState *bs, BlockDriverState *target,
 
     is_none_mode = mode == MIRROR_SYNC_MODE_NONE;
     base = mode == MIRROR_SYNC_MODE_TOP ? bs->backing_hd : NULL;
-    mirror_start_job(bs, target, speed, granularity, buf_size,
+    mirror_start_job(bs, target, replaces,
+                     speed, granularity, buf_size,
                      on_source_error, on_target_error, cb, opaque, errp,
                      &mirror_job_driver, is_none_mode, base);
 }
@@ -681,7 +729,7 @@ void commit_active_start(BlockDriverState *bs, BlockDriverState *base,
     }
 
     bdrv_ref(base);
-    mirror_start_job(bs, base, speed, 0, 0,
+    mirror_start_job(bs, base, NULL, speed, 0, 0,
                      on_error, on_error, cb, opaque, &local_err,
                      &commit_active_job_driver, false, base);
     if (local_err) {

block/nbd.c

@@ -31,8 +31,10 @@
 #include "block/block_int.h"
 #include "qemu/module.h"
 #include "qemu/sockets.h"
+#include "qapi/qmp/qdict.h"
 #include "qapi/qmp/qjson.h"
 #include "qapi/qmp/qint.h"
+#include "qapi/qmp/qstring.h"
 
 #include <sys/types.h>
 #include <unistd.h>
@@ -338,6 +340,37 @@ static void nbd_attach_aio_context(BlockDriverState *bs,
     nbd_client_session_attach_aio_context(&s->client, new_context);
 }
 
+static void nbd_refresh_filename(BlockDriverState *bs)
+{
+    BDRVNBDState *s = bs->opaque;
+    QDict *opts = qdict_new();
+    const char *path   = qemu_opt_get(s->socket_opts, "path");
+    const char *host   = qemu_opt_get(s->socket_opts, "host");
+    const char *port   = qemu_opt_get(s->socket_opts, "port");
+    const char *export = qemu_opt_get(s->socket_opts, "export");
+
+    qdict_put_obj(opts, "driver", QOBJECT(qstring_from_str("nbd")));
+
+    if (path) {
+        snprintf(bs->exact_filename, sizeof(bs->exact_filename),
+                 "nbd+unix:%s", path);
+        qdict_put_obj(opts, "path", QOBJECT(qstring_from_str(path)));
+    } else if (export) {
+        snprintf(bs->exact_filename, sizeof(bs->exact_filename),
+                 "nbd:%s:%s/%s", host, port, export);
+        qdict_put_obj(opts, "host",   QOBJECT(qstring_from_str(host)));
+        qdict_put_obj(opts, "port",   QOBJECT(qstring_from_str(port)));
+        qdict_put_obj(opts, "export", QOBJECT(qstring_from_str(export)));
+    } else {
+        snprintf(bs->exact_filename, sizeof(bs->exact_filename),
+                 "nbd:%s:%s", host, port);
+        qdict_put_obj(opts, "host", QOBJECT(qstring_from_str(host)));
+        qdict_put_obj(opts, "port", QOBJECT(qstring_from_str(port)));
+    }
+
+    bs->full_open_options = opts;
+}
+
 static BlockDriver bdrv_nbd = {
     .format_name                = "nbd",
     .protocol_name              = "nbd",
@@ -352,6 +385,7 @@ static BlockDriver bdrv_nbd = {
     .bdrv_getlength             = nbd_getlength,
     .bdrv_detach_aio_context    = nbd_detach_aio_context,
     .bdrv_attach_aio_context    = nbd_attach_aio_context,
+    .bdrv_refresh_filename      = nbd_refresh_filename,
 };
 
 static BlockDriver bdrv_nbd_tcp = {
@@ -368,6 +402,7 @@ static BlockDriver bdrv_nbd_tcp = {
     .bdrv_getlength             = nbd_getlength,
     .bdrv_detach_aio_context    = nbd_detach_aio_context,
     .bdrv_attach_aio_context    = nbd_attach_aio_context,
+    .bdrv_refresh_filename      = nbd_refresh_filename,
 };
 
 static BlockDriver bdrv_nbd_unix = {
@@ -384,6 +419,7 @@ static BlockDriver bdrv_nbd_unix = {
     .bdrv_getlength             = nbd_getlength,
     .bdrv_detach_aio_context    = nbd_detach_aio_context,
     .bdrv_attach_aio_context    = nbd_attach_aio_context,
+    .bdrv_refresh_filename      = nbd_refresh_filename,
 };
 
 static void bdrv_nbd_init(void)

block/nfs.c

@@ -172,7 +172,11 @@ static int coroutine_fn nfs_co_writev(BlockDriverState *bs,
 
     nfs_co_init_task(client, &task);
 
-    buf = g_malloc(nb_sectors * BDRV_SECTOR_SIZE);
+    buf = g_try_malloc(nb_sectors * BDRV_SECTOR_SIZE);
+    if (nb_sectors && buf == NULL) {
+        return -ENOMEM;
+    }
+
     qemu_iovec_to_buf(iov, 0, buf, nb_sectors * BDRV_SECTOR_SIZE);
 
     if (nfs_pwrite_async(client->context, client->fh,
@@ -304,17 +308,27 @@ static int64_t nfs_client_open(NFSClient *client, const char *filename,
 
     qp = query_params_parse(uri->query);
     for (i = 0; i < qp->n; i++) {
+        unsigned long long val;
         if (!qp->p[i].value) {
             error_setg(errp, "Value for NFS parameter expected: %s",
                        qp->p[i].name);
             goto fail;
         }
-        if (!strncmp(qp->p[i].name, "uid", 3)) {
-            nfs_set_uid(client->context, atoi(qp->p[i].value));
-        } else if (!strncmp(qp->p[i].name, "gid", 3)) {
-            nfs_set_gid(client->context, atoi(qp->p[i].value));
-        } else if (!strncmp(qp->p[i].name, "tcp-syncnt", 10)) {
-            nfs_set_tcp_syncnt(client->context, atoi(qp->p[i].value));
+        if (parse_uint_full(qp->p[i].value, &val, 0)) {
+            error_setg(errp, "Illegal value for NFS parameter: %s",
+                       qp->p[i].name);
+            goto fail;
+        }
+        if (!strcmp(qp->p[i].name, "uid")) {
+            nfs_set_uid(client->context, val);
+        } else if (!strcmp(qp->p[i].name, "gid")) {
+            nfs_set_gid(client->context, val);
+        } else if (!strcmp(qp->p[i].name, "tcp-syncnt")) {
+            nfs_set_tcp_syncnt(client->context, val);
+#ifdef LIBNFS_FEATURE_READAHEAD
+        } else if (!strcmp(qp->p[i].name, "readahead")) {
+            nfs_set_readahead(client->context, val);
+#endif
         } else {
             error_setg(errp, "Unknown NFS parameter name: %s",
                        qp->p[i].name);
@@ -379,23 +393,27 @@ static int nfs_file_open(BlockDriverState *bs, QDict *options, int flags,
     qemu_opts_absorb_qdict(opts, options, &local_err);
     if (local_err) {
         error_propagate(errp, local_err);
-        return -EINVAL;
+        ret = -EINVAL;
+        goto out;
     }
     ret = nfs_client_open(client, qemu_opt_get(opts, "filename"),
                           (flags & BDRV_O_RDWR) ? O_RDWR : O_RDONLY,
                           errp);
     if (ret < 0) {
-        return ret;
+        goto out;
     }
     bs->total_sectors = ret;
-    return 0;
+    ret = 0;
+out:
+    qemu_opts_del(opts);
+    return ret;
 }
 
 static int nfs_file_create(const char *url, QemuOpts *opts, Error **errp)
 {
     int ret = 0;
     int64_t total_size = 0;
-    NFSClient *client = g_malloc0(sizeof(NFSClient));
+    NFSClient *client = g_new0(NFSClient, 1);
 
     client->aio_context = qemu_get_aio_context();
 

block/parallels.c

@@ -30,6 +30,7 @@
 /**************************************************************/
 
 #define HEADER_MAGIC "WithoutFreeSpace"
+#define HEADER_MAGIC2 "WithouFreSpacExt"
 #define HEADER_VERSION 2
 #define HEADER_SIZE 64
 
@@ -41,8 +42,10 @@ struct parallels_header {
     uint32_t cylinders;
     uint32_t tracks;
     uint32_t catalog_entries;
-    uint32_t nb_sectors;
-    char padding[24];
+    uint64_t nb_sectors;
+    uint32_t inuse;
+    uint32_t data_off;
+    char padding[12];
 } QEMU_PACKED;
 
 typedef struct BDRVParallelsState {
@@ -52,6 +55,8 @@ typedef struct BDRVParallelsState {
     unsigned int catalog_size;
 
     unsigned int tracks;
+
+    unsigned int off_multiplier;
 } BDRVParallelsState;
 
 static int parallels_probe(const uint8_t *buf, int buf_size, const char *filename)
@@ -59,11 +64,12 @@ static int parallels_probe(const uint8_t *buf, int buf_size, const char *filenam
     const struct parallels_header *ph = (const void *)buf;
 
     if (buf_size < HEADER_SIZE)
-	return 0;
+        return 0;
 
-    if (!memcmp(ph->magic, HEADER_MAGIC, 16) &&
-	(le32_to_cpu(ph->version) == HEADER_VERSION))
-	return 100;
+    if ((!memcmp(ph->magic, HEADER_MAGIC, 16) ||
+        !memcmp(ph->magic, HEADER_MAGIC2, 16)) &&
+        (le32_to_cpu(ph->version) == HEADER_VERSION))
+        return 100;
 
     return 0;
 }
@@ -83,14 +89,19 @@ static int parallels_open(BlockDriverState *bs, QDict *options, int flags,
         goto fail;
     }
 
-    if (memcmp(ph.magic, HEADER_MAGIC, 16) ||
-        (le32_to_cpu(ph.version) != HEADER_VERSION)) {
-        error_setg(errp, "Image not in Parallels format");
-        ret = -EINVAL;
-        goto fail;
-    }
+    bs->total_sectors = le64_to_cpu(ph.nb_sectors);
 
-    bs->total_sectors = le32_to_cpu(ph.nb_sectors);
+    if (le32_to_cpu(ph.version) != HEADER_VERSION) {
+        goto fail_format;
+    }
+    if (!memcmp(ph.magic, HEADER_MAGIC, 16)) {
+        s->off_multiplier = 1;
+        bs->total_sectors = 0xffffffff & bs->total_sectors;
+    } else if (!memcmp(ph.magic, HEADER_MAGIC2, 16)) {
+        s->off_multiplier = le32_to_cpu(ph.tracks);
+    } else {
+        goto fail_format;
+    }
 
     s->tracks = le32_to_cpu(ph.tracks);
     if (s->tracks == 0) {
@@ -98,6 +109,11 @@ static int parallels_open(BlockDriverState *bs, QDict *options, int flags,
         ret = -EINVAL;
         goto fail;
     }
+    if (s->tracks > INT32_MAX/513) {
+        error_setg(errp, "Invalid image: Too big cluster");
+        ret = -EFBIG;
+        goto fail;
+    }
 
     s->catalog_size = le32_to_cpu(ph.catalog_entries);
     if (s->catalog_size > INT_MAX / 4) {
@@ -105,7 +121,11 @@ static int parallels_open(BlockDriverState *bs, QDict *options, int flags,
         ret = -EFBIG;
         goto fail;
     }
-    s->catalog_bitmap = g_malloc(s->catalog_size * 4);
+    s->catalog_bitmap = g_try_new(uint32_t, s->catalog_size);
+    if (s->catalog_size && s->catalog_bitmap == NULL) {
+        ret = -ENOMEM;
+        goto fail;
+    }
 
     ret = bdrv_pread(bs->file, 64, s->catalog_bitmap, s->catalog_size * 4);
     if (ret < 0) {
@@ -113,11 +133,14 @@ static int parallels_open(BlockDriverState *bs, QDict *options, int flags,
     }
 
     for (i = 0; i < s->catalog_size; i++)
-	le32_to_cpus(&s->catalog_bitmap[i]);
+        le32_to_cpus(&s->catalog_bitmap[i]);
 
     qemu_co_mutex_init(&s->lock);
     return 0;
 
+fail_format:
+    error_setg(errp, "Image not in Parallels format");
+    ret = -EINVAL;
 fail:
     g_free(s->catalog_bitmap);
     return ret;
@@ -133,8 +156,9 @@ static int64_t seek_to_sector(BlockDriverState *bs, int64_t sector_num)
 
     /* not allocated */
     if ((index > s->catalog_size) || (s->catalog_bitmap[index] == 0))
-	return -1;
-    return (uint64_t)(s->catalog_bitmap[index] + offset) * 512;
+        return -1;
+    return
+        ((uint64_t)s->catalog_bitmap[index] * s->off_multiplier + offset) * 512;
 }
 
 static int parallels_read(BlockDriverState *bs, int64_t sector_num,

block/qapi.c

@@ -28,6 +28,13 @@
 #include "qapi-visit.h"
 #include "qapi/qmp-output-visitor.h"
 #include "qapi/qmp/types.h"
+#ifdef __linux__
+#include <linux/fs.h>
+#include <sys/ioctl.h>
+#ifndef FS_NOCOW_FL
+#define FS_NOCOW_FL                     0x00800000 /* Do not cow file */
+#endif
+#endif
 
 BlockDeviceInfo *bdrv_block_device_info(BlockDriverState *bs)
 {
@@ -165,19 +172,28 @@ void bdrv_query_image_info(BlockDriverState *bs,
                            ImageInfo **p_info,
                            Error **errp)
 {
-    uint64_t total_sectors;
+    int64_t size;
     const char *backing_filename;
     char backing_filename2[1024];
     BlockDriverInfo bdi;
     int ret;
     Error *err = NULL;
-    ImageInfo *info = g_new0(ImageInfo, 1);
+    ImageInfo *info;
+#ifdef __linux__
+    int fd, attr;
+#endif
 
-    bdrv_get_geometry(bs, &total_sectors);
+    size = bdrv_getlength(bs);
+    if (size < 0) {
+        error_setg_errno(errp, -size, "Can't get size of device '%s'",
+                         bdrv_get_device_name(bs));
+        return;
+    }
 
+    info = g_new0(ImageInfo, 1);
     info->filename        = g_strdup(bs->filename);
     info->format          = g_strdup(bdrv_get_format_name(bs));
-    info->virtual_size    = total_sectors * 512;
+    info->virtual_size    = size;
     info->actual_size     = bdrv_get_allocated_file_size(bs);
     info->has_actual_size = info->actual_size >= 0;
     if (bdrv_is_encrypted(bs)) {
@@ -195,6 +211,18 @@ void bdrv_query_image_info(BlockDriverState *bs,
     info->format_specific     = bdrv_get_specific_info(bs);
     info->has_format_specific = info->format_specific != NULL;
 
+#ifdef __linux__
+    /* get NOCOW info */
+    fd = qemu_open(bs->filename, O_RDONLY | O_NONBLOCK);
+    if (fd >= 0) {
+        if (ioctl(fd, FS_IOC_GETFLAGS, &attr) == 0 && (attr & FS_NOCOW_FL)) {
+            info->has_nocow = true;
+            info->nocow = true;
+        }
+        qemu_close(fd);
+    }
+#endif
+
     backing_filename = bs->backing_file;
     if (backing_filename[0] != '\0') {
         info->backing_filename = g_strdup(backing_filename);
@@ -293,7 +321,7 @@ void bdrv_query_info(BlockDriverState *bs,
     qapi_free_BlockInfo(info);
 }
 
-BlockStats *bdrv_query_stats(const BlockDriverState *bs)
+static BlockStats *bdrv_query_stats(const BlockDriverState *bs)
 {
     BlockStats *s;
 
@@ -360,7 +388,11 @@ BlockStatsList *qmp_query_blockstats(Error **errp)
 
      while ((bs = bdrv_next(bs))) {
         BlockStatsList *info = g_malloc0(sizeof(*info));
+        AioContext *ctx = bdrv_get_aio_context(bs);
+
+        aio_context_acquire(ctx);
         info->value = bdrv_query_stats(bs);
+        aio_context_release(ctx);
 
         *p_next = info;
         p_next = &info->next;
@@ -621,4 +653,8 @@ void bdrv_image_info_dump(fprintf_function func_fprintf, void *f,
         func_fprintf(f, "Format specific information:\n");
         bdrv_image_info_specific_dump(func_fprintf, f, info->format_specific);
     }
+
+    if (info->has_nocow && info->nocow) {
+        func_fprintf(f, "NOCOW flag: set\n");
+    }
 }

block/qcow.c

@@ -182,7 +182,12 @@ static int qcow_open(BlockDriverState *bs, QDict *options, int flags,
     }
 
     s->l1_table_offset = header.l1_table_offset;
-    s->l1_table = g_malloc(s->l1_size * sizeof(uint64_t));
+    s->l1_table = g_try_new(uint64_t, s->l1_size);
+    if (s->l1_table == NULL) {
+        error_setg(errp, "Could not allocate memory for L1 table");
+        ret = -ENOMEM;
+        goto fail;
+    }
 
     ret = bdrv_pread(bs->file, s->l1_table_offset, s->l1_table,
                s->l1_size * sizeof(uint64_t));
@@ -193,8 +198,16 @@ static int qcow_open(BlockDriverState *bs, QDict *options, int flags,
     for(i = 0;i < s->l1_size; i++) {
         be64_to_cpus(&s->l1_table[i]);
     }
-    /* alloc L2 cache */
-    s->l2_cache = g_malloc(s->l2_size * L2_CACHE_SIZE * sizeof(uint64_t));
+
+    /* alloc L2 cache (max. 64k * 16 * 8 = 8 MB) */
+    s->l2_cache =
+        qemu_try_blockalign(bs->file,
+                            s->l2_size * L2_CACHE_SIZE * sizeof(uint64_t));
+    if (s->l2_cache == NULL) {
+        error_setg(errp, "Could not allocate L2 table cache");
+        ret = -ENOMEM;
+        goto fail;
+    }
     s->cluster_cache = g_malloc(s->cluster_size);
     s->cluster_data = g_malloc(s->cluster_size);
     s->cluster_cache_offset = -1;
@@ -226,7 +239,7 @@ static int qcow_open(BlockDriverState *bs, QDict *options, int flags,
 
  fail:
     g_free(s->l1_table);
-    g_free(s->l2_cache);
+    qemu_vfree(s->l2_cache);
     g_free(s->cluster_cache);
     g_free(s->cluster_data);
     return ret;
@@ -517,7 +530,10 @@ static coroutine_fn int qcow_co_readv(BlockDriverState *bs, int64_t sector_num,
     void *orig_buf;
 
     if (qiov->niov > 1) {
-        buf = orig_buf = qemu_blockalign(bs, qiov->size);
+        buf = orig_buf = qemu_try_blockalign(bs, qiov->size);
+        if (buf == NULL) {
+            return -ENOMEM;
+        }
     } else {
         orig_buf = NULL;
         buf = (uint8_t *)qiov->iov->iov_base;
@@ -619,7 +635,10 @@ static coroutine_fn int qcow_co_writev(BlockDriverState *bs, int64_t sector_num,
     s->cluster_cache_offset = -1; /* disable compressed cache */
 
     if (qiov->niov > 1) {
-        buf = orig_buf = qemu_blockalign(bs, qiov->size);
+        buf = orig_buf = qemu_try_blockalign(bs, qiov->size);
+        if (buf == NULL) {
+            return -ENOMEM;
+        }
         qemu_iovec_to_buf(qiov, 0, buf, qiov->size);
     } else {
         orig_buf = NULL;
@@ -685,7 +704,7 @@ static void qcow_close(BlockDriverState *bs)
     BDRVQcowState *s = bs->opaque;
 
     g_free(s->l1_table);
-    g_free(s->l2_cache);
+    qemu_vfree(s->l2_cache);
     g_free(s->cluster_cache);
     g_free(s->cluster_data);
 
@@ -941,6 +960,7 @@ static BlockDriver bdrv_qcow = {
     .bdrv_reopen_prepare    = qcow_reopen_prepare,
     .bdrv_create            = qcow_create,
     .bdrv_has_zero_init     = bdrv_has_zero_init_1,
+    .supports_backing       = true,
 
     .bdrv_co_readv          = qcow_co_readv,
     .bdrv_co_writev         = qcow_co_writev,

block/qcow2-cache.c

@@ -48,15 +48,31 @@ Qcow2Cache *qcow2_cache_create(BlockDriverState *bs, int num_tables)
     Qcow2Cache *c;
     int i;
 
-    c = g_malloc0(sizeof(*c));
+    c = g_new0(Qcow2Cache, 1);
     c->size = num_tables;
-    c->entries = g_malloc0(sizeof(*c->entries) * num_tables);
+    c->entries = g_try_new0(Qcow2CachedTable, num_tables);
+    if (!c->entries) {
+        goto fail;
+    }
 
     for (i = 0; i < c->size; i++) {
-        c->entries[i].table = qemu_blockalign(bs, s->cluster_size);
+        c->entries[i].table = qemu_try_blockalign(bs->file, s->cluster_size);
+        if (c->entries[i].table == NULL) {
+            goto fail;
+        }
     }
 
     return c;
+
+fail:
+    if (c->entries) {
+        for (i = 0; i < c->size; i++) {
+            qemu_vfree(c->entries[i].table);
+        }
+    }
+    g_free(c->entries);
+    g_free(c);
+    return NULL;
 }
 
 int qcow2_cache_destroy(BlockDriverState* bs, Qcow2Cache *c)

block/qcow2-cluster.c

@@ -72,14 +72,20 @@ int qcow2_grow_l1_table(BlockDriverState *bs, uint64_t min_size,
 #endif
 
     new_l1_size2 = sizeof(uint64_t) * new_l1_size;
-    new_l1_table = g_malloc0(align_offset(new_l1_size2, 512));
+    new_l1_table = qemu_try_blockalign(bs->file,
+                                       align_offset(new_l1_size2, 512));
+    if (new_l1_table == NULL) {
+        return -ENOMEM;
+    }
+    memset(new_l1_table, 0, align_offset(new_l1_size2, 512));
+
     memcpy(new_l1_table, s->l1_table, s->l1_size * sizeof(uint64_t));
 
     /* write new table (align to cluster) */
     BLKDBG_EVENT(bs->file, BLKDBG_L1_GROW_ALLOC_TABLE);
     new_l1_table_offset = qcow2_alloc_clusters(bs, new_l1_size2);
     if (new_l1_table_offset < 0) {
-        g_free(new_l1_table);
+        qemu_vfree(new_l1_table);
         return new_l1_table_offset;
     }
 
@@ -113,7 +119,7 @@ int qcow2_grow_l1_table(BlockDriverState *bs, uint64_t min_size,
     if (ret < 0) {
         goto fail;
     }
-    g_free(s->l1_table);
+    qemu_vfree(s->l1_table);
     old_l1_table_offset = s->l1_table_offset;
     s->l1_table_offset = new_l1_table_offset;
     s->l1_table = new_l1_table;
@@ -123,7 +129,7 @@ int qcow2_grow_l1_table(BlockDriverState *bs, uint64_t min_size,
                         QCOW2_DISCARD_OTHER);
     return 0;
  fail:
-    g_free(new_l1_table);
+    qemu_vfree(new_l1_table);
     qcow2_free_clusters(bs, new_l1_table_offset, new_l1_size2,
                         QCOW2_DISCARD_OTHER);
     return ret;
@@ -372,7 +378,10 @@ static int coroutine_fn copy_sectors(BlockDriverState *bs,
     }
 
     iov.iov_len = n * BDRV_SECTOR_SIZE;
-    iov.iov_base = qemu_blockalign(bs, iov.iov_len);
+    iov.iov_base = qemu_try_blockalign(bs, iov.iov_len);
+    if (iov.iov_base == NULL) {
+        return -ENOMEM;
+    }
 
     qemu_iovec_init_external(&qiov, &iov, 1);
 
@@ -702,7 +711,11 @@ int qcow2_alloc_cluster_link_l2(BlockDriverState *bs, QCowL2Meta *m)
     trace_qcow2_cluster_link_l2(qemu_coroutine_self(), m->nb_clusters);
     assert(m->nb_clusters > 0);
 
-    old_cluster = g_malloc(m->nb_clusters * sizeof(uint64_t));
+    old_cluster = g_try_new(uint64_t, m->nb_clusters);
+    if (old_cluster == NULL) {
+        ret = -ENOMEM;
+        goto err;
+    }
 
     /* copy content of unmodified sectors */
     ret = perform_cow(bs, m, &m->cow_start);
@@ -1106,6 +1119,17 @@ static int handle_alloc(BlockDriverState *bs, uint64_t guest_offset,
         return 0;
     }
 
+    /* !*host_offset would overwrite the image header and is reserved for "no
+     * host offset preferred". If 0 was a valid host offset, it'd trigger the
+     * following overlap check; do that now to avoid having an invalid value in
+     * *host_offset. */
+    if (!alloc_cluster_offset) {
+        ret = qcow2_pre_write_overlap_check(bs, 0, alloc_cluster_offset,
+                                            nb_clusters * s->cluster_size);
+        assert(ret < 0);
+        goto fail;
+    }
+
     /*
      * Save info needed for meta data update.
      *
@@ -1562,7 +1586,10 @@ static int expand_zero_clusters_in_l1(BlockDriverState *bs, uint64_t *l1_table,
     if (!is_active_l1) {
         /* inactive L2 tables require a buffer to be stored in when loading
          * them from disk */
-        l2_table = qemu_blockalign(bs, s->cluster_size);
+        l2_table = qemu_try_blockalign(bs->file, s->cluster_size);
+        if (l2_table == NULL) {
+            return -ENOMEM;
+        }
     }
 
     for (i = 0; i < l1_size; i++) {
@@ -1740,7 +1767,11 @@ int qcow2_expand_zero_clusters(BlockDriverState *bs)
 
     nb_clusters = size_to_clusters(s, bs->file->total_sectors *
                                    BDRV_SECTOR_SIZE);
-    expanded_clusters = g_malloc0((nb_clusters + 7) / 8);
+    expanded_clusters = g_try_malloc0((nb_clusters + 7) / 8);
+    if (expanded_clusters == NULL) {
+        ret = -ENOMEM;
+        goto fail;
+    }
 
     ret = expand_zero_clusters_in_l1(bs, s->l1_table, s->l1_size,
                                      &expanded_clusters, &nb_clusters);

block/qcow2-refcount.c

@@ -46,19 +46,25 @@ int qcow2_refcount_init(BlockDriverState *bs)
 
     assert(s->refcount_table_size <= INT_MAX / sizeof(uint64_t));
     refcount_table_size2 = s->refcount_table_size * sizeof(uint64_t);
-    s->refcount_table = g_malloc(refcount_table_size2);
+    s->refcount_table = g_try_malloc(refcount_table_size2);
+
     if (s->refcount_table_size > 0) {
+        if (s->refcount_table == NULL) {
+            ret = -ENOMEM;
+            goto fail;
+        }
         BLKDBG_EVENT(bs->file, BLKDBG_REFTABLE_LOAD);
         ret = bdrv_pread(bs->file, s->refcount_table_offset,
                          s->refcount_table, refcount_table_size2);
-        if (ret != refcount_table_size2)
+        if (ret < 0) {
             goto fail;
+        }
         for(i = 0; i < s->refcount_table_size; i++)
             be64_to_cpus(&s->refcount_table[i]);
     }
     return 0;
  fail:
-    return -ENOMEM;
+    return ret;
 }
 
 void qcow2_refcount_close(BlockDriverState *bs)
@@ -344,8 +350,14 @@ static int alloc_refcount_block(BlockDriverState *bs,
     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 = g_malloc0(blocks_clusters * s->cluster_size);
-    uint64_t *new_table = g_malloc0(table_size * sizeof(uint64_t));
+    uint64_t *new_table = g_try_new0(uint64_t, table_size);
+    uint16_t *new_blocks = g_try_malloc0(blocks_clusters * s->cluster_size);
+
+    assert(table_size > 0 && blocks_clusters > 0);
+    if (new_table == NULL || new_blocks == NULL) {
+        ret = -ENOMEM;
+        goto fail_table;
+    }
 
     /* Fill the new refcount table */
     memcpy(new_table, s->refcount_table,
@@ -369,6 +381,7 @@ static int alloc_refcount_block(BlockDriverState *bs,
     ret = bdrv_pwrite_sync(bs->file, meta_offset, new_blocks,
         blocks_clusters * s->cluster_size);
     g_free(new_blocks);
+    new_blocks = NULL;
     if (ret < 0) {
         goto fail_table;
     }
@@ -424,6 +437,7 @@ static int alloc_refcount_block(BlockDriverState *bs,
     return -EAGAIN;
 
 fail_table:
+    g_free(new_blocks);
     g_free(new_table);
 fail_block:
     if (*refcount_block != NULL) {
@@ -847,7 +861,8 @@ int qcow2_update_snapshot_refcount(BlockDriverState *bs,
     int64_t l1_table_offset, int l1_size, int addend)
 {
     BDRVQcowState *s = bs->opaque;
-    uint64_t *l1_table, *l2_table, l2_offset, offset, l1_size2, l1_allocated;
+    uint64_t *l1_table, *l2_table, l2_offset, offset, l1_size2;
+    bool l1_allocated = false;
     int64_t old_offset, old_l2_offset;
     int i, j, l1_modified = 0, nb_csectors, refcount;
     int ret;
@@ -862,8 +877,12 @@ int qcow2_update_snapshot_refcount(BlockDriverState *bs,
      * l1_table_offset when it is the current s->l1_table_offset! Be careful
      * when changing this! */
     if (l1_table_offset != s->l1_table_offset) {
-        l1_table = g_malloc0(align_offset(l1_size2, 512));
-        l1_allocated = 1;
+        l1_table = g_try_malloc0(align_offset(l1_size2, 512));
+        if (l1_size2 && l1_table == NULL) {
+            ret = -ENOMEM;
+            goto fail;
+        }
+        l1_allocated = true;
 
         ret = bdrv_pread(bs->file, l1_table_offset, l1_table, l1_size2);
         if (ret < 0) {
@@ -875,7 +894,7 @@ int qcow2_update_snapshot_refcount(BlockDriverState *bs,
     } else {
         assert(l1_size == s->l1_size);
         l1_table = s->l1_table;
-        l1_allocated = 0;
+        l1_allocated = false;
     }
 
     for(i = 0; i < l1_size; i++) {
@@ -1197,7 +1216,11 @@ static int check_refcounts_l1(BlockDriverState *bs,
     if (l1_size2 == 0) {
         l1_table = NULL;
     } else {
-        l1_table = g_malloc(l1_size2);
+        l1_table = g_try_malloc(l1_size2);
+        if (l1_table == NULL) {
+            ret = -ENOMEM;
+            goto fail;
+        }
         if (bdrv_pread(bs->file, l1_table_offset,
                        l1_table, l1_size2) != l1_size2)
             goto fail;
@@ -1501,7 +1524,11 @@ int qcow2_check_refcounts(BlockDriverState *bs, BdrvCheckResult *res,
         return -EFBIG;
     }
 
-    refcount_table = g_malloc0(nb_clusters * sizeof(uint16_t));
+    refcount_table = g_try_new0(uint16_t, nb_clusters);
+    if (nb_clusters && refcount_table == NULL) {
+        res->check_errors++;
+        return -ENOMEM;
+    }
 
     res->bfi.total_clusters =
         size_to_clusters(s, bs->total_sectors * BDRV_SECTOR_SIZE);
@@ -1578,8 +1605,8 @@ int qcow2_check_refcounts(BlockDriverState *bs, BdrvCheckResult *res,
                         /* increase refcount_table size if necessary */
                         int old_nb_clusters = nb_clusters;
                         nb_clusters = (new_offset >> s->cluster_bits) + 1;
-                        refcount_table = g_realloc(refcount_table,
-                                nb_clusters * sizeof(uint16_t));
+                        refcount_table = g_renew(uint16_t, refcount_table,
+                                                 nb_clusters);
                         memset(&refcount_table[old_nb_clusters], 0, (nb_clusters
                                 - old_nb_clusters) * sizeof(uint16_t));
                     }
@@ -1753,9 +1780,13 @@ int qcow2_check_metadata_overlap(BlockDriverState *bs, int ign, int64_t offset,
             uint64_t l1_ofs = s->snapshots[i].l1_table_offset;
             uint32_t l1_sz  = s->snapshots[i].l1_size;
             uint64_t l1_sz2 = l1_sz * sizeof(uint64_t);
-            uint64_t *l1 = g_malloc(l1_sz2);
+            uint64_t *l1 = g_try_malloc(l1_sz2);
             int ret;
 
+            if (l1_sz2 && l1 == NULL) {
+                return -ENOMEM;
+            }
+
             ret = bdrv_pread(bs->file, l1_ofs, l1, l1_sz2);
             if (ret < 0) {
                 g_free(l1);

block/qcow2-snapshot.c

@@ -58,7 +58,7 @@ int qcow2_read_snapshots(BlockDriverState *bs)
     }
 
     offset = s->snapshots_offset;
-    s->snapshots = g_malloc0(s->nb_snapshots * sizeof(QCowSnapshot));
+    s->snapshots = g_new0(QCowSnapshot, s->nb_snapshots);
 
     for(i = 0; i < s->nb_snapshots; i++) {
         /* Read statically sized part of the snapshot header */
@@ -381,7 +381,12 @@ int qcow2_snapshot_create(BlockDriverState *bs, QEMUSnapshotInfo *sn_info)
     sn->l1_table_offset = l1_table_offset;
     sn->l1_size = s->l1_size;
 
-    l1_table = g_malloc(s->l1_size * sizeof(uint64_t));
+    l1_table = g_try_new(uint64_t, s->l1_size);
+    if (s->l1_size && l1_table == NULL) {
+        ret = -ENOMEM;
+        goto fail;
+    }
+
     for(i = 0; i < s->l1_size; i++) {
         l1_table[i] = cpu_to_be64(s->l1_table[i]);
     }
@@ -412,7 +417,7 @@ int qcow2_snapshot_create(BlockDriverState *bs, QEMUSnapshotInfo *sn_info)
     }
 
     /* Append the new snapshot to the snapshot list */
-    new_snapshot_list = g_malloc((s->nb_snapshots + 1) * sizeof(QCowSnapshot));
+    new_snapshot_list = g_new(QCowSnapshot, s->nb_snapshots + 1);
     if (s->snapshots) {
         memcpy(new_snapshot_list, s->snapshots,
                s->nb_snapshots * sizeof(QCowSnapshot));
@@ -499,7 +504,11 @@ int qcow2_snapshot_goto(BlockDriverState *bs, const char *snapshot_id)
      * Decrease the refcount referenced by the old one only when the L1
      * table is overwritten.
      */
-    sn_l1_table = g_malloc0(cur_l1_bytes);
+    sn_l1_table = g_try_malloc0(cur_l1_bytes);
+    if (cur_l1_bytes && sn_l1_table == NULL) {
+        ret = -ENOMEM;
+        goto fail;
+    }
 
     ret = bdrv_pread(bs->file, sn->l1_table_offset, sn_l1_table, sn_l1_bytes);
     if (ret < 0) {
@@ -652,7 +661,7 @@ int qcow2_snapshot_list(BlockDriverState *bs, QEMUSnapshotInfo **psn_tab)
         return s->nb_snapshots;
     }
 
-    sn_tab = g_malloc0(s->nb_snapshots * sizeof(QEMUSnapshotInfo));
+    sn_tab = g_new0(QEMUSnapshotInfo, s->nb_snapshots);
     for(i = 0; i < s->nb_snapshots; i++) {
         sn_info = sn_tab + i;
         sn = s->snapshots + i;
@@ -698,17 +707,21 @@ int qcow2_snapshot_load_tmp(BlockDriverState *bs,
         return -EFBIG;
     }
     new_l1_bytes = sn->l1_size * sizeof(uint64_t);
-    new_l1_table = g_malloc0(align_offset(new_l1_bytes, 512));
+    new_l1_table = qemu_try_blockalign(bs->file,
+                                       align_offset(new_l1_bytes, 512));
+    if (new_l1_table == NULL) {
+        return -ENOMEM;
+    }
 
     ret = bdrv_pread(bs->file, sn->l1_table_offset, new_l1_table, new_l1_bytes);
     if (ret < 0) {
         error_setg(errp, "Failed to read l1 table for snapshot");
-        g_free(new_l1_table);
+        qemu_vfree(new_l1_table);
         return ret;
     }
 
     /* Switch the L1 table */
-    g_free(s->l1_table);
+    qemu_vfree(s->l1_table);
 
     s->l1_size = sn->l1_size;
     s->l1_table_offset = sn->l1_table_offset;

block/qcow2.c

@@ -210,20 +210,31 @@ static void GCC_FMT_ATTR(3, 4) report_unsupported(BlockDriverState *bs,
 static void report_unsupported_feature(BlockDriverState *bs,
     Error **errp, Qcow2Feature *table, uint64_t mask)
 {
+    char *features = g_strdup("");
+    char *old;
+
     while (table && table->name[0] != '\0') {
         if (table->type == QCOW2_FEAT_TYPE_INCOMPATIBLE) {
-            if (mask & (1 << table->bit)) {
-                report_unsupported(bs, errp, "%.46s", table->name);
-                mask &= ~(1 << table->bit);
+            if (mask & (1ULL << table->bit)) {
+                old = features;
+                features = g_strdup_printf("%s%s%.46s", old, *old ? ", " : "",
+                                           table->name);
+                g_free(old);
+                mask &= ~(1ULL << table->bit);
             }
         }
         table++;
     }
 
     if (mask) {
-        report_unsupported(bs, errp, "Unknown incompatible feature: %" PRIx64,
-                           mask);
+        old = features;
+        features = g_strdup_printf("%s%sUnknown incompatible feature: %" PRIx64,
+                                   old, *old ? ", " : "", mask);
+        g_free(old);
     }
+
+    report_unsupported(bs, errp, "%s", features);
+    g_free(features);
 }
 
 /*
@@ -431,6 +442,22 @@ static QemuOptsList qcow2_runtime_opts = {
             .type = QEMU_OPT_BOOL,
             .help = "Check for unintended writes into an inactive L2 table",
         },
+        {
+            .name = QCOW2_OPT_CACHE_SIZE,
+            .type = QEMU_OPT_SIZE,
+            .help = "Maximum combined metadata (L2 tables and refcount blocks) "
+                    "cache size",
+        },
+        {
+            .name = QCOW2_OPT_L2_CACHE_SIZE,
+            .type = QEMU_OPT_SIZE,
+            .help = "Maximum L2 table cache size",
+        },
+        {
+            .name = QCOW2_OPT_REFCOUNT_CACHE_SIZE,
+            .type = QEMU_OPT_SIZE,
+            .help = "Maximum refcount block cache size",
+        },
         { /* end of list */ }
     },
 };
@@ -446,6 +473,61 @@ static const char *overlap_bool_option_names[QCOW2_OL_MAX_BITNR] = {
     [QCOW2_OL_INACTIVE_L2_BITNR]    = QCOW2_OPT_OVERLAP_INACTIVE_L2,
 };
 
+static void read_cache_sizes(QemuOpts *opts, uint64_t *l2_cache_size,
+                             uint64_t *refcount_cache_size, Error **errp)
+{
+    uint64_t combined_cache_size;
+    bool l2_cache_size_set, refcount_cache_size_set, combined_cache_size_set;
+
+    combined_cache_size_set = qemu_opt_get(opts, QCOW2_OPT_CACHE_SIZE);
+    l2_cache_size_set = qemu_opt_get(opts, QCOW2_OPT_L2_CACHE_SIZE);
+    refcount_cache_size_set = qemu_opt_get(opts, QCOW2_OPT_REFCOUNT_CACHE_SIZE);
+
+    combined_cache_size = qemu_opt_get_size(opts, QCOW2_OPT_CACHE_SIZE, 0);
+    *l2_cache_size = qemu_opt_get_size(opts, QCOW2_OPT_L2_CACHE_SIZE, 0);
+    *refcount_cache_size = qemu_opt_get_size(opts,
+                                             QCOW2_OPT_REFCOUNT_CACHE_SIZE, 0);
+
+    if (combined_cache_size_set) {
+        if (l2_cache_size_set && refcount_cache_size_set) {
+            error_setg(errp, QCOW2_OPT_CACHE_SIZE ", " QCOW2_OPT_L2_CACHE_SIZE
+                       " and " QCOW2_OPT_REFCOUNT_CACHE_SIZE " may not be set "
+                       "the same time");
+            return;
+        } else if (*l2_cache_size > combined_cache_size) {
+            error_setg(errp, QCOW2_OPT_L2_CACHE_SIZE " may not exceed "
+                       QCOW2_OPT_CACHE_SIZE);
+            return;
+        } else if (*refcount_cache_size > combined_cache_size) {
+            error_setg(errp, QCOW2_OPT_REFCOUNT_CACHE_SIZE " may not exceed "
+                       QCOW2_OPT_CACHE_SIZE);
+            return;
+        }
+
+        if (l2_cache_size_set) {
+            *refcount_cache_size = combined_cache_size - *l2_cache_size;
+        } else if (refcount_cache_size_set) {
+            *l2_cache_size = combined_cache_size - *refcount_cache_size;
+        } else {
+            *refcount_cache_size = combined_cache_size
+                                 / (DEFAULT_L2_REFCOUNT_SIZE_RATIO + 1);
+            *l2_cache_size = combined_cache_size - *refcount_cache_size;
+        }
+    } else {
+        if (!l2_cache_size_set && !refcount_cache_size_set) {
+            *l2_cache_size = DEFAULT_L2_CACHE_BYTE_SIZE;
+            *refcount_cache_size = *l2_cache_size
+                                 / DEFAULT_L2_REFCOUNT_SIZE_RATIO;
+        } else if (!l2_cache_size_set) {
+            *l2_cache_size = *refcount_cache_size
+                           * DEFAULT_L2_REFCOUNT_SIZE_RATIO;
+        } else if (!refcount_cache_size_set) {
+            *refcount_cache_size = *l2_cache_size
+                                 / DEFAULT_L2_REFCOUNT_SIZE_RATIO;
+        }
+    }
+}
+
 static int qcow2_open(BlockDriverState *bs, QDict *options, int flags,
                       Error **errp)
 {
@@ -459,6 +541,7 @@ static int qcow2_open(BlockDriverState *bs, QDict *options, int flags,
     uint64_t l1_vm_state_index;
     const char *opt_overlap_check;
     int overlap_check_template = 0;
+    uint64_t l2_cache_size, refcount_cache_size;
 
     ret = bdrv_pread(bs->file, 0, &header, sizeof(header));
     if (ret < 0) {
@@ -677,8 +760,13 @@ static int qcow2_open(BlockDriverState *bs, QDict *options, int flags,
 
 
     if (s->l1_size > 0) {
-        s->l1_table = g_malloc0(
+        s->l1_table = qemu_try_blockalign(bs->file,
             align_offset(s->l1_size * sizeof(uint64_t), 512));
+        if (s->l1_table == NULL) {
+            error_setg(errp, "Could not allocate L1 table");
+            ret = -ENOMEM;
+            goto fail;
+        }
         ret = bdrv_pread(bs->file, s->l1_table_offset, s->l1_table,
                          s->l1_size * sizeof(uint64_t));
         if (ret < 0) {
@@ -690,14 +778,61 @@ static int qcow2_open(BlockDriverState *bs, QDict *options, int flags,
         }
     }
 
+    /* get L2 table/refcount block cache size from command line options */
+    opts = qemu_opts_create(&qcow2_runtime_opts, NULL, 0, &error_abort);
+    qemu_opts_absorb_qdict(opts, options, &local_err);
+    if (local_err) {
+        error_propagate(errp, local_err);
+        ret = -EINVAL;
+        goto fail;
+    }
+
+    read_cache_sizes(opts, &l2_cache_size, &refcount_cache_size, &local_err);
+    if (local_err) {
+        error_propagate(errp, local_err);
+        ret = -EINVAL;
+        goto fail;
+    }
+
+    l2_cache_size /= s->cluster_size;
+    if (l2_cache_size < MIN_L2_CACHE_SIZE) {
+        l2_cache_size = MIN_L2_CACHE_SIZE;
+    }
+    if (l2_cache_size > INT_MAX) {
+        error_setg(errp, "L2 cache size too big");
+        ret = -EINVAL;
+        goto fail;
+    }
+
+    refcount_cache_size /= s->cluster_size;
+    if (refcount_cache_size < MIN_REFCOUNT_CACHE_SIZE) {
+        refcount_cache_size = MIN_REFCOUNT_CACHE_SIZE;
+    }
+    if (refcount_cache_size > INT_MAX) {
+        error_setg(errp, "Refcount cache size too big");
+        ret = -EINVAL;
+        goto fail;
+    }
+
     /* alloc L2 table/refcount block cache */
-    s->l2_table_cache = qcow2_cache_create(bs, L2_CACHE_SIZE);
-    s->refcount_block_cache = qcow2_cache_create(bs, REFCOUNT_CACHE_SIZE);
+    s->l2_table_cache = qcow2_cache_create(bs, l2_cache_size);
+    s->refcount_block_cache = qcow2_cache_create(bs, refcount_cache_size);
+    if (s->l2_table_cache == NULL || s->refcount_block_cache == NULL) {
+        error_setg(errp, "Could not allocate metadata caches");
+        ret = -ENOMEM;
+        goto fail;
+    }
 
     s->cluster_cache = g_malloc(s->cluster_size);
     /* one more sector for decompressed data alignment */
-    s->cluster_data = qemu_blockalign(bs, QCOW_MAX_CRYPT_CLUSTERS * s->cluster_size
-                                  + 512);
+    s->cluster_data = qemu_try_blockalign(bs->file, QCOW_MAX_CRYPT_CLUSTERS
+                                                    * s->cluster_size + 512);
+    if (s->cluster_data == NULL) {
+        error_setg(errp, "Could not allocate temporary cluster buffer");
+        ret = -ENOMEM;
+        goto fail;
+    }
+
     s->cluster_cache_offset = -1;
     s->flags = flags;
 
@@ -771,14 +906,6 @@ static int qcow2_open(BlockDriverState *bs, QDict *options, int flags,
     }
 
     /* Enable lazy_refcounts according to image and command line options */
-    opts = qemu_opts_create(&qcow2_runtime_opts, NULL, 0, &error_abort);
-    qemu_opts_absorb_qdict(opts, options, &local_err);
-    if (local_err) {
-        error_propagate(errp, local_err);
-        ret = -EINVAL;
-        goto fail;
-    }
-
     s->use_lazy_refcounts = qemu_opt_get_bool(opts, QCOW2_OPT_LAZY_REFCOUNTS,
         (s->compatible_features & QCOW2_COMPAT_LAZY_REFCOUNTS));
 
@@ -841,7 +968,7 @@ static int qcow2_open(BlockDriverState *bs, QDict *options, int flags,
     cleanup_unknown_header_ext(bs);
     qcow2_free_snapshots(bs);
     qcow2_refcount_close(bs);
-    g_free(s->l1_table);
+    qemu_vfree(s->l1_table);
     /* else pre-write overlap checks in cache_destroy may crash */
     s->l1_table = NULL;
     if (s->l2_table_cache) {
@@ -855,13 +982,11 @@ static int qcow2_open(BlockDriverState *bs, QDict *options, int flags,
     return ret;
 }
 
-static int qcow2_refresh_limits(BlockDriverState *bs)
+static void qcow2_refresh_limits(BlockDriverState *bs, Error **errp)
 {
     BDRVQcowState *s = bs->opaque;
 
     bs->bl.write_zeroes_alignment = s->cluster_sectors;
-
-    return 0;
 }
 
 static int qcow2_set_key(BlockDriverState *bs, const char *key)
@@ -1020,11 +1145,20 @@ static coroutine_fn int qcow2_co_readv(BlockDriverState *bs, int64_t sector_num,
                 n1 = qcow2_backing_read1(bs->backing_hd, &hd_qiov,
                     sector_num, cur_nr_sectors);
                 if (n1 > 0) {
+                    QEMUIOVector local_qiov;
+
+                    qemu_iovec_init(&local_qiov, hd_qiov.niov);
+                    qemu_iovec_concat(&local_qiov, &hd_qiov, 0,
+                                      n1 * BDRV_SECTOR_SIZE);
+
                     BLKDBG_EVENT(bs->file, BLKDBG_READ_BACKING_AIO);
                     qemu_co_mutex_unlock(&s->lock);
                     ret = bdrv_co_readv(bs->backing_hd, sector_num,
-                                        n1, &hd_qiov);
+                                        n1, &local_qiov);
                     qemu_co_mutex_lock(&s->lock);
+
+                    qemu_iovec_destroy(&local_qiov);
+
                     if (ret < 0) {
                         goto fail;
                     }
@@ -1064,7 +1198,12 @@ static coroutine_fn int qcow2_co_readv(BlockDriverState *bs, int64_t sector_num,
                  */
                 if (!cluster_data) {
                     cluster_data =
-                        qemu_blockalign(bs, QCOW_MAX_CRYPT_CLUSTERS * s->cluster_size);
+                        qemu_try_blockalign(bs->file, QCOW_MAX_CRYPT_CLUSTERS
+                                                      * s->cluster_size);
+                    if (cluster_data == NULL) {
+                        ret = -ENOMEM;
+                        goto fail;
+                    }
                 }
 
                 assert(cur_nr_sectors <=
@@ -1164,8 +1303,13 @@ static coroutine_fn int qcow2_co_writev(BlockDriverState *bs,
 
         if (s->crypt_method) {
             if (!cluster_data) {
-                cluster_data = qemu_blockalign(bs, QCOW_MAX_CRYPT_CLUSTERS *
-                                                 s->cluster_size);
+                cluster_data = qemu_try_blockalign(bs->file,
+                                                   QCOW_MAX_CRYPT_CLUSTERS
+                                                   * s->cluster_size);
+                if (cluster_data == NULL) {
+                    ret = -ENOMEM;
+                    goto fail;
+                }
             }
 
             assert(hd_qiov.size <=
@@ -1252,7 +1396,7 @@ fail:
 static void qcow2_close(BlockDriverState *bs)
 {
     BDRVQcowState *s = bs->opaque;
-    g_free(s->l1_table);
+    qemu_vfree(s->l1_table);
     /* else pre-write overlap checks in cache_destroy may crash */
     s->l1_table = NULL;
 
@@ -1539,7 +1683,7 @@ static int preallocate(BlockDriverState *bs)
     int ret;
     QCowL2Meta *meta;
 
-    nb_sectors = bdrv_getlength(bs) >> BDRV_SECTOR_BITS;
+    nb_sectors = bdrv_nb_sectors(bs);
     offset = 0;
 
     while (nb_sectors) {
@@ -1929,7 +2073,6 @@ static int qcow2_write_compressed(BlockDriverState *bs, int64_t sector_num,
         /* align end of file to a sector boundary to ease reading with
            sector based I/Os */
         cluster_offset = bdrv_getlength(bs->file);
-        cluster_offset = (cluster_offset + 511) & ~511;
         bdrv_truncate(bs->file, cluster_offset);
         return 0;
     }
@@ -2418,6 +2561,7 @@ static BlockDriver bdrv_qcow2 = {
     .bdrv_save_vmstate    = qcow2_save_vmstate,
     .bdrv_load_vmstate    = qcow2_load_vmstate,
 
+    .supports_backing           = true,
     .bdrv_change_backing_file   = qcow2_change_backing_file,
 
     .bdrv_refresh_limits        = qcow2_refresh_limits,

block/qcow2.h

@@ -64,10 +64,16 @@
 #define MIN_CLUSTER_BITS 9
 #define MAX_CLUSTER_BITS 21
 
-#define L2_CACHE_SIZE 16
+#define MIN_L2_CACHE_SIZE 1 /* cluster */
 
 /* Must be at least 4 to cover all cases of refcount table growth */
-#define REFCOUNT_CACHE_SIZE 4
+#define MIN_REFCOUNT_CACHE_SIZE 4 /* clusters */
+
+#define DEFAULT_L2_CACHE_BYTE_SIZE 1048576 /* bytes */
+
+/* The refblock cache needs only a fourth of the L2 cache size to cover as many
+ * clusters */
+#define DEFAULT_L2_REFCOUNT_SIZE_RATIO 4
 
 #define DEFAULT_CLUSTER_SIZE 65536
 
@@ -85,6 +91,9 @@
 #define QCOW2_OPT_OVERLAP_SNAPSHOT_TABLE "overlap-check.snapshot-table"
 #define QCOW2_OPT_OVERLAP_INACTIVE_L1 "overlap-check.inactive-l1"
 #define QCOW2_OPT_OVERLAP_INACTIVE_L2 "overlap-check.inactive-l2"
+#define QCOW2_OPT_CACHE_SIZE "cache-size"
+#define QCOW2_OPT_L2_CACHE_SIZE "l2-cache-size"
+#define QCOW2_OPT_REFCOUNT_CACHE_SIZE "refcount-cache-size"
 
 typedef struct QCowHeader {
     uint32_t magic;

block/qed-check.c

@@ -227,8 +227,10 @@ int qed_check(BDRVQEDState *s, BdrvCheckResult *result, bool fix)
     };
     int ret;
 
-    check.used_clusters = g_malloc0(((check.nclusters + 31) / 32) *
-                                       sizeof(check.used_clusters[0]));
+    check.used_clusters = g_try_new0(uint32_t, (check.nclusters + 31) / 32);
+    if (check.nclusters && check.used_clusters == NULL) {
+        return -ENOMEM;
+    }
 
     check.result->bfi.total_clusters =
         (s->header.image_size + s->header.cluster_size - 1) /

block/qed.c

@@ -528,13 +528,11 @@ out:
     return ret;
 }
 
-static int bdrv_qed_refresh_limits(BlockDriverState *bs)
+static void bdrv_qed_refresh_limits(BlockDriverState *bs, Error **errp)
 {
     BDRVQEDState *s = bs->opaque;
 
     bs->bl.write_zeroes_alignment = s->header.cluster_size >> BDRV_SECTOR_BITS;
-
-    return 0;
 }
 
 /* We have nothing to do for QED reopen, stubs just return
@@ -567,7 +565,7 @@ static void bdrv_qed_close(BlockDriverState *bs)
 static int qed_create(const char *filename, uint32_t cluster_size,
                       uint64_t image_size, uint32_t table_size,
                       const char *backing_file, const char *backing_fmt,
-                      Error **errp)
+                      QemuOpts *opts, Error **errp)
 {
     QEDHeader header = {
         .magic = QED_MAGIC,
@@ -586,7 +584,7 @@ static int qed_create(const char *filename, uint32_t cluster_size,
     int ret = 0;
     BlockDriverState *bs;
 
-    ret = bdrv_create_file(filename, NULL, &local_err);
+    ret = bdrv_create_file(filename, opts, &local_err);
     if (ret < 0) {
         error_propagate(errp, local_err);
         return ret;
@@ -682,7 +680,7 @@ static int bdrv_qed_create(const char *filename, QemuOpts *opts, Error **errp)
     }
 
     ret = qed_create(filename, cluster_size, image_size, table_size,
-                     backing_file, backing_fmt, errp);
+                     backing_file, backing_fmt, opts, errp);
 
 finish:
     g_free(backing_file);
@@ -761,17 +759,19 @@ static BDRVQEDState *acb_to_s(QEDAIOCB *acb)
 /**
  * Read from the backing file or zero-fill if no backing file
  *
- * @s:          QED state
- * @pos:        Byte position in device
- * @qiov:       Destination I/O vector
- * @cb:         Completion function
- * @opaque:     User data for completion function
+ * @s:              QED state
+ * @pos:            Byte position in device
+ * @qiov:           Destination I/O vector
+ * @backing_qiov:   Possibly shortened copy of qiov, to be allocated here
+ * @cb:             Completion function
+ * @opaque:         User data for completion function
  *
  * This function reads qiov->size bytes starting at pos from the backing file.
  * If there is no backing file then zeroes are read.
  */
 static void qed_read_backing_file(BDRVQEDState *s, uint64_t pos,
                                   QEMUIOVector *qiov,
+                                  QEMUIOVector **backing_qiov,
                                   BlockDriverCompletionFunc *cb, void *opaque)
 {
     uint64_t backing_length = 0;
@@ -804,15 +804,21 @@ static void qed_read_backing_file(BDRVQEDState *s, uint64_t pos,
     /* If the read straddles the end of the backing file, shorten it */
     size = MIN((uint64_t)backing_length - pos, qiov->size);
 
+    assert(*backing_qiov == NULL);
+    *backing_qiov = g_new(QEMUIOVector, 1);
+    qemu_iovec_init(*backing_qiov, qiov->niov);
+    qemu_iovec_concat(*backing_qiov, qiov, 0, size);
+
     BLKDBG_EVENT(s->bs->file, BLKDBG_READ_BACKING_AIO);
     bdrv_aio_readv(s->bs->backing_hd, pos / BDRV_SECTOR_SIZE,
-                   qiov, size / BDRV_SECTOR_SIZE, cb, opaque);
+                   *backing_qiov, size / BDRV_SECTOR_SIZE, cb, opaque);
 }
 
 typedef struct {
     GenericCB gencb;
     BDRVQEDState *s;
     QEMUIOVector qiov;
+    QEMUIOVector *backing_qiov;
     struct iovec iov;
     uint64_t offset;
 } CopyFromBackingFileCB;
@@ -829,6 +835,12 @@ static void qed_copy_from_backing_file_write(void *opaque, int ret)
     CopyFromBackingFileCB *copy_cb = opaque;
     BDRVQEDState *s = copy_cb->s;
 
+    if (copy_cb->backing_qiov) {
+        qemu_iovec_destroy(copy_cb->backing_qiov);
+        g_free(copy_cb->backing_qiov);
+        copy_cb->backing_qiov = NULL;
+    }
+
     if (ret) {
         qed_copy_from_backing_file_cb(copy_cb, ret);
         return;
@@ -866,11 +878,12 @@ static void qed_copy_from_backing_file(BDRVQEDState *s, uint64_t pos,
     copy_cb = gencb_alloc(sizeof(*copy_cb), cb, opaque);
     copy_cb->s = s;
     copy_cb->offset = offset;
+    copy_cb->backing_qiov = NULL;
     copy_cb->iov.iov_base = qemu_blockalign(s->bs, len);
     copy_cb->iov.iov_len = len;
     qemu_iovec_init_external(&copy_cb->qiov, &copy_cb->iov, 1);
 
-    qed_read_backing_file(s, pos, &copy_cb->qiov,
+    qed_read_backing_file(s, pos, &copy_cb->qiov, &copy_cb->backing_qiov,
                           qed_copy_from_backing_file_write, copy_cb);
 }
 
@@ -1227,7 +1240,11 @@ static void qed_aio_write_inplace(QEDAIOCB *acb, uint64_t offset, size_t len)
         struct iovec *iov = acb->qiov->iov;
 
         if (!iov->iov_base) {
-            iov->iov_base = qemu_blockalign(acb->common.bs, iov->iov_len);
+            iov->iov_base = qemu_try_blockalign(acb->common.bs, iov->iov_len);
+            if (iov->iov_base == NULL) {
+                qed_aio_complete(acb, -ENOMEM);
+                return;
+            }
             memset(iov->iov_base, 0, iov->iov_len);
         }
     }
@@ -1313,7 +1330,7 @@ static void qed_aio_read_data(void *opaque, int ret,
         return;
     } else if (ret != QED_CLUSTER_FOUND) {
         qed_read_backing_file(s, acb->cur_pos, &acb->cur_qiov,
-                              qed_aio_next_io, acb);
+                              &acb->backing_qiov, qed_aio_next_io, acb);
         return;
     }
 
@@ -1339,6 +1356,12 @@ static void qed_aio_next_io(void *opaque, int ret)
 
     trace_qed_aio_next_io(s, acb, ret, acb->cur_pos + acb->cur_qiov.size);
 
+    if (acb->backing_qiov) {
+        qemu_iovec_destroy(acb->backing_qiov);
+        g_free(acb->backing_qiov);
+        acb->backing_qiov = NULL;
+    }
+
     /* Handle I/O error */
     if (ret) {
         qed_aio_complete(acb, ret);
@@ -1378,6 +1401,7 @@ static BlockDriverAIOCB *qed_aio_setup(BlockDriverState *bs,
     acb->qiov_offset = 0;
     acb->cur_pos = (uint64_t)sector_num * BDRV_SECTOR_SIZE;
     acb->end_pos = acb->cur_pos + nb_sectors * BDRV_SECTOR_SIZE;
+    acb->backing_qiov = NULL;
     acb->request.l2_table = NULL;
     qemu_iovec_init(&acb->cur_qiov, qiov->niov);
 
@@ -1652,6 +1676,7 @@ static BlockDriver bdrv_qed = {
     .format_name              = "qed",
     .instance_size            = sizeof(BDRVQEDState),
     .create_opts              = &qed_create_opts,
+    .supports_backing         = true,
 
     .bdrv_probe               = bdrv_qed_probe,
     .bdrv_rebind              = bdrv_qed_rebind,

block/qed.h

@@ -142,6 +142,7 @@ typedef struct QEDAIOCB {
 
     /* Current cluster scatter-gather list */
     QEMUIOVector cur_qiov;
+    QEMUIOVector *backing_qiov;
     uint64_t cur_pos;               /* position on block device, in bytes */
     uint64_t cur_cluster;           /* cluster offset in image file */
     unsigned int cur_nclusters;     /* number of clusters being accessed */

block/quorum.c

@@ -16,13 +16,20 @@
 #include <gnutls/gnutls.h>
 #include <gnutls/crypto.h>
 #include "block/block_int.h"
+#include "qapi/qmp/qbool.h"
+#include "qapi/qmp/qdict.h"
+#include "qapi/qmp/qint.h"
 #include "qapi/qmp/qjson.h"
+#include "qapi/qmp/qlist.h"
+#include "qapi/qmp/qstring.h"
 #include "qapi-event.h"
 
 #define HASH_LENGTH 32
 
 #define QUORUM_OPT_VOTE_THRESHOLD "vote-threshold"
 #define QUORUM_OPT_BLKVERIFY      "blkverify"
+#define QUORUM_OPT_REWRITE        "rewrite-corrupted"
+#define QUORUM_OPT_READ_PATTERN   "read-pattern"
 
 /* This union holds a vote hash value */
 typedef union QuorumVoteValue {
@@ -70,6 +77,11 @@ typedef struct BDRVQuorumState {
                             * It is useful to debug other block drivers by
                             * comparing them with a reference one.
                             */
+    bool rewrite_corrupted;/* true if the driver must rewrite-on-read corrupted
+                            * block if Quorum is reached.
+                            */
+
+    QuorumReadPattern read_pattern;
 } BDRVQuorumState;
 
 typedef struct QuorumAIOCB QuorumAIOCB;
@@ -105,13 +117,18 @@ struct QuorumAIOCB {
     int count;                  /* number of completed AIOCB */
     int success_count;          /* number of successfully completed AIOCB */
 
+    int rewrite_count;          /* number of replica to rewrite: count down to
+                                 * zero once writes are fired
+                                 */
+
     QuorumVotes votes;
 
     bool is_read;
     int vote_ret;
+    int child_iter;             /* which child to read in fifo pattern */
 };
 
-static void quorum_vote(QuorumAIOCB *acb);
+static bool quorum_vote(QuorumAIOCB *acb);
 
 static void quorum_aio_cancel(BlockDriverAIOCB *blockacb)
 {
@@ -135,7 +152,6 @@ static AIOCBInfo quorum_aiocb_info = {
 
 static void quorum_aio_finalize(QuorumAIOCB *acb)
 {
-    BDRVQuorumState *s = acb->common.bs->opaque;
     int i, ret = 0;
 
     if (acb->vote_ret) {
@@ -145,7 +161,8 @@ static void quorum_aio_finalize(QuorumAIOCB *acb)
     acb->common.cb(acb->common.opaque, ret);
 
     if (acb->is_read) {
-        for (i = 0; i < s->num_children; i++) {
+        /* on the quorum case acb->child_iter == s->num_children - 1 */
+        for (i = 0; i <= acb->child_iter; i++) {
             qemu_vfree(acb->qcrs[i].buf);
             qemu_iovec_destroy(&acb->qcrs[i].qiov);
         }
@@ -183,6 +200,7 @@ static QuorumAIOCB *quorum_aio_get(BDRVQuorumState *s,
     acb->qcrs = g_new0(QuorumChildRequest, s->num_children);
     acb->count = 0;
     acb->success_count = 0;
+    acb->rewrite_count = 0;
     acb->votes.compare = quorum_sha256_compare;
     QLIST_INIT(&acb->votes.vote_list);
     acb->is_read = false;
@@ -232,11 +250,57 @@ static bool quorum_has_too_much_io_failed(QuorumAIOCB *acb)
     return false;
 }
 
+static void quorum_rewrite_aio_cb(void *opaque, int ret)
+{
+    QuorumAIOCB *acb = opaque;
+
+    /* one less rewrite to do */
+    acb->rewrite_count--;
+
+    /* wait until all rewrite callbacks have completed */
+    if (acb->rewrite_count) {
+        return;
+    }
+
+    quorum_aio_finalize(acb);
+}
+
+static BlockDriverAIOCB *read_fifo_child(QuorumAIOCB *acb);
+
+static void quorum_copy_qiov(QEMUIOVector *dest, QEMUIOVector *source)
+{
+    int i;
+    assert(dest->niov == source->niov);
+    assert(dest->size == source->size);
+    for (i = 0; i < source->niov; i++) {
+        assert(dest->iov[i].iov_len == source->iov[i].iov_len);
+        memcpy(dest->iov[i].iov_base,
+               source->iov[i].iov_base,
+               source->iov[i].iov_len);
+    }
+}
+
 static void quorum_aio_cb(void *opaque, int ret)
 {
     QuorumChildRequest *sacb = opaque;
     QuorumAIOCB *acb = sacb->parent;
     BDRVQuorumState *s = acb->common.bs->opaque;
+    bool rewrite = false;
+
+    if (acb->is_read && s->read_pattern == QUORUM_READ_PATTERN_FIFO) {
+        /* We try to read next child in FIFO order if we fail to read */
+        if (ret < 0 && ++acb->child_iter < s->num_children) {
+            read_fifo_child(acb);
+            return;
+        }
+
+        if (ret == 0) {
+            quorum_copy_qiov(acb->qiov, &acb->qcrs[acb->child_iter].qiov);
+        }
+        acb->vote_ret = ret;
+        quorum_aio_finalize(acb);
+        return;
+    }
 
     sacb->ret = ret;
     acb->count++;
@@ -253,12 +317,15 @@ static void quorum_aio_cb(void *opaque, int ret)
 
     /* Do the vote on read */
     if (acb->is_read) {
-        quorum_vote(acb);
+        rewrite = quorum_vote(acb);
     } else {
         quorum_has_too_much_io_failed(acb);
     }
 
-    quorum_aio_finalize(acb);
+    /* if no rewrite is done the code will finish right away */
+    if (!rewrite) {
+        quorum_aio_finalize(acb);
+    }
 }
 
 static void quorum_report_bad_versions(BDRVQuorumState *s,
@@ -278,17 +345,41 @@ static void quorum_report_bad_versions(BDRVQuorumState *s,
     }
 }
 
-static void quorum_copy_qiov(QEMUIOVector *dest, QEMUIOVector *source)
+static bool quorum_rewrite_bad_versions(BDRVQuorumState *s, QuorumAIOCB *acb,
+                                        QuorumVoteValue *value)
 {
-    int i;
-    assert(dest->niov == source->niov);
-    assert(dest->size == source->size);
-    for (i = 0; i < source->niov; i++) {
-        assert(dest->iov[i].iov_len == source->iov[i].iov_len);
-        memcpy(dest->iov[i].iov_base,
-               source->iov[i].iov_base,
-               source->iov[i].iov_len);
+    QuorumVoteVersion *version;
+    QuorumVoteItem *item;
+    int count = 0;
+
+    /* first count the number of bad versions: done first to avoid concurrency
+     * issues.
+     */
+    QLIST_FOREACH(version, &acb->votes.vote_list, next) {
+        if (acb->votes.compare(&version->value, value)) {
+            continue;
+        }
+        QLIST_FOREACH(item, &version->items, next) {
+            count++;
+        }
     }
+
+    /* quorum_rewrite_aio_cb will count down this to zero */
+    acb->rewrite_count = count;
+
+    /* now fire the correcting rewrites */
+    QLIST_FOREACH(version, &acb->votes.vote_list, next) {
+        if (acb->votes.compare(&version->value, value)) {
+            continue;
+        }
+        QLIST_FOREACH(item, &version->items, next) {
+            bdrv_aio_writev(s->bs[item->index], acb->sector_num, acb->qiov,
+                            acb->nb_sectors, quorum_rewrite_aio_cb, acb);
+        }
+    }
+
+    /* return true if any rewrite is done else false */
+    return count;
 }
 
 static void quorum_count_vote(QuorumVotes *votes,
@@ -468,16 +559,17 @@ static int quorum_vote_error(QuorumAIOCB *acb)
     return ret;
 }
 
-static void quorum_vote(QuorumAIOCB *acb)
+static bool quorum_vote(QuorumAIOCB *acb)
 {
     bool quorum = true;
+    bool rewrite = false;
     int i, j, ret;
     QuorumVoteValue hash;
     BDRVQuorumState *s = acb->common.bs->opaque;
     QuorumVoteVersion *winner;
 
     if (quorum_has_too_much_io_failed(acb)) {
-        return;
+        return false;
     }
 
     /* get the index of the first successful read */
@@ -505,7 +597,7 @@ static void quorum_vote(QuorumAIOCB *acb)
     /* Every successful read agrees */
     if (quorum) {
         quorum_copy_qiov(acb->qiov, &acb->qcrs[i].qiov);
-        return;
+        return false;
     }
 
     /* compute hashes for each successful read, also store indexes */
@@ -538,37 +630,71 @@ static void quorum_vote(QuorumAIOCB *acb)
     /* some versions are bad print them */
     quorum_report_bad_versions(s, acb, &winner->value);
 
+    /* corruption correction is enabled */
+    if (s->rewrite_corrupted) {
+        rewrite = quorum_rewrite_bad_versions(s, acb, &winner->value);
+    }
+
 free_exit:
     /* free lists */
     quorum_free_vote_list(&acb->votes);
+    return rewrite;
+}
+
+static BlockDriverAIOCB *read_quorum_children(QuorumAIOCB *acb)
+{
+    BDRVQuorumState *s = acb->common.bs->opaque;
+    int i;
+
+    for (i = 0; i < s->num_children; i++) {
+        acb->qcrs[i].buf = qemu_blockalign(s->bs[i], acb->qiov->size);
+        qemu_iovec_init(&acb->qcrs[i].qiov, acb->qiov->niov);
+        qemu_iovec_clone(&acb->qcrs[i].qiov, acb->qiov, acb->qcrs[i].buf);
+    }
+
+    for (i = 0; i < s->num_children; i++) {
+        bdrv_aio_readv(s->bs[i], acb->sector_num, &acb->qcrs[i].qiov,
+                       acb->nb_sectors, quorum_aio_cb, &acb->qcrs[i]);
+    }
+
+    return &acb->common;
+}
+
+static BlockDriverAIOCB *read_fifo_child(QuorumAIOCB *acb)
+{
+    BDRVQuorumState *s = acb->common.bs->opaque;
+
+    acb->qcrs[acb->child_iter].buf = qemu_blockalign(s->bs[acb->child_iter],
+                                                     acb->qiov->size);
+    qemu_iovec_init(&acb->qcrs[acb->child_iter].qiov, acb->qiov->niov);
+    qemu_iovec_clone(&acb->qcrs[acb->child_iter].qiov, acb->qiov,
+                     acb->qcrs[acb->child_iter].buf);
+    bdrv_aio_readv(s->bs[acb->child_iter], acb->sector_num,
+                   &acb->qcrs[acb->child_iter].qiov, acb->nb_sectors,
+                   quorum_aio_cb, &acb->qcrs[acb->child_iter]);
+
+    return &acb->common;
 }
 
 static BlockDriverAIOCB *quorum_aio_readv(BlockDriverState *bs,
-                                         int64_t sector_num,
-                                         QEMUIOVector *qiov,
-                                         int nb_sectors,
-                                         BlockDriverCompletionFunc *cb,
-                                         void *opaque)
+                                          int64_t sector_num,
+                                          QEMUIOVector *qiov,
+                                          int nb_sectors,
+                                          BlockDriverCompletionFunc *cb,
+                                          void *opaque)
 {
     BDRVQuorumState *s = bs->opaque;
     QuorumAIOCB *acb = quorum_aio_get(s, bs, qiov, sector_num,
                                       nb_sectors, cb, opaque);
-    int i;
-
     acb->is_read = true;
 
-    for (i = 0; i < s->num_children; i++) {
-        acb->qcrs[i].buf = qemu_blockalign(s->bs[i], qiov->size);
-        qemu_iovec_init(&acb->qcrs[i].qiov, qiov->niov);
-        qemu_iovec_clone(&acb->qcrs[i].qiov, qiov, acb->qcrs[i].buf);
+    if (s->read_pattern == QUORUM_READ_PATTERN_QUORUM) {
+        acb->child_iter = s->num_children - 1;
+        return read_quorum_children(acb);
     }
 
-    for (i = 0; i < s->num_children; i++) {
-        bdrv_aio_readv(s->bs[i], sector_num, &acb->qcrs[i].qiov, nb_sectors,
-                       quorum_aio_cb, &acb->qcrs[i]);
-    }
-
-    return &acb->common;
+    acb->child_iter = 0;
+    return read_fifo_child(acb);
 }
 
 static BlockDriverAIOCB *quorum_aio_writev(BlockDriverState *bs,
@@ -705,16 +831,44 @@ static QemuOptsList quorum_runtime_opts = {
             .type = QEMU_OPT_BOOL,
             .help = "Trigger block verify mode if set",
         },
+        {
+            .name = QUORUM_OPT_REWRITE,
+            .type = QEMU_OPT_BOOL,
+            .help = "Rewrite corrupted block on read quorum",
+        },
+        {
+            .name = QUORUM_OPT_READ_PATTERN,
+            .type = QEMU_OPT_STRING,
+            .help = "Allowed pattern: quorum, fifo. Quorum is default",
+        },
         { /* end of list */ }
     },
 };
 
+static int parse_read_pattern(const char *opt)
+{
+    int i;
+
+    if (!opt) {
+        /* Set quorum as default */
+        return QUORUM_READ_PATTERN_QUORUM;
+    }
+
+    for (i = 0; i < QUORUM_READ_PATTERN_MAX; i++) {
+        if (!strcmp(opt, QuorumReadPattern_lookup[i])) {
+            return i;
+        }
+    }
+
+    return -EINVAL;
+}
+
 static int quorum_open(BlockDriverState *bs, QDict *options, int flags,
                        Error **errp)
 {
     BDRVQuorumState *s = bs->opaque;
     Error *local_err = NULL;
-    QemuOpts *opts;
+    QemuOpts *opts = NULL;
     bool *opened;
     QDict *sub = NULL;
     QList *list = NULL;
@@ -750,20 +904,37 @@ static int quorum_open(BlockDriverState *bs, QDict *options, int flags,
     }
 
     s->threshold = qemu_opt_get_number(opts, QUORUM_OPT_VOTE_THRESHOLD, 0);
-
-    /* and validate it against s->num_children */
-    ret = quorum_valid_threshold(s->threshold, s->num_children, &local_err);
+    ret = parse_read_pattern(qemu_opt_get(opts, QUORUM_OPT_READ_PATTERN));
     if (ret < 0) {
+        error_setg(&local_err, "Please set read-pattern as fifo or quorum");
         goto exit;
     }
+    s->read_pattern = ret;
 
-    /* is the driver in blkverify mode */
-    if (qemu_opt_get_bool(opts, QUORUM_OPT_BLKVERIFY, false) &&
-        s->num_children == 2 && s->threshold == 2) {
-        s->is_blkverify = true;
-    } else if (qemu_opt_get_bool(opts, QUORUM_OPT_BLKVERIFY, false)) {
-        fprintf(stderr, "blkverify mode is set by setting blkverify=on "
-                "and using two files with vote_threshold=2\n");
+    if (s->read_pattern == QUORUM_READ_PATTERN_QUORUM) {
+        /* and validate it against s->num_children */
+        ret = quorum_valid_threshold(s->threshold, s->num_children, &local_err);
+        if (ret < 0) {
+            goto exit;
+        }
+
+        /* is the driver in blkverify mode */
+        if (qemu_opt_get_bool(opts, QUORUM_OPT_BLKVERIFY, false) &&
+            s->num_children == 2 && s->threshold == 2) {
+            s->is_blkverify = true;
+        } else if (qemu_opt_get_bool(opts, QUORUM_OPT_BLKVERIFY, false)) {
+            fprintf(stderr, "blkverify mode is set by setting blkverify=on "
+                    "and using two files with vote_threshold=2\n");
+        }
+
+        s->rewrite_corrupted = qemu_opt_get_bool(opts, QUORUM_OPT_REWRITE,
+                                                 false);
+        if (s->rewrite_corrupted && s->is_blkverify) {
+            error_setg(&local_err,
+                       "rewrite-corrupted=on cannot be used with blkverify=on");
+            ret = -EINVAL;
+            goto exit;
+        }
     }
 
     /* allocate the children BlockDriverState array */
@@ -818,6 +989,7 @@ close_exit:
     g_free(s->bs);
     g_free(opened);
 exit:
+    qemu_opts_del(opts);
     /* propagate error */
     if (local_err) {
         error_propagate(errp, local_err);
@@ -860,6 +1032,39 @@ static void quorum_attach_aio_context(BlockDriverState *bs,
     }
 }
 
+static void quorum_refresh_filename(BlockDriverState *bs)
+{
+    BDRVQuorumState *s = bs->opaque;
+    QDict *opts;
+    QList *children;
+    int i;
+
+    for (i = 0; i < s->num_children; i++) {
+        bdrv_refresh_filename(s->bs[i]);
+        if (!s->bs[i]->full_open_options) {
+            return;
+        }
+    }
+
+    children = qlist_new();
+    for (i = 0; i < s->num_children; i++) {
+        QINCREF(s->bs[i]->full_open_options);
+        qlist_append_obj(children, QOBJECT(s->bs[i]->full_open_options));
+    }
+
+    opts = qdict_new();
+    qdict_put_obj(opts, "driver", QOBJECT(qstring_from_str("quorum")));
+    qdict_put_obj(opts, QUORUM_OPT_VOTE_THRESHOLD,
+                  QOBJECT(qint_from_int(s->threshold)));
+    qdict_put_obj(opts, QUORUM_OPT_BLKVERIFY,
+                  QOBJECT(qbool_from_int(s->is_blkverify)));
+    qdict_put_obj(opts, QUORUM_OPT_REWRITE,
+                  QOBJECT(qbool_from_int(s->rewrite_corrupted)));
+    qdict_put_obj(opts, "children", QOBJECT(children));
+
+    bs->full_open_options = opts;
+}
+
 static BlockDriver bdrv_quorum = {
     .format_name                        = "quorum",
     .protocol_name                      = "quorum",
@@ -868,6 +1073,7 @@ static BlockDriver bdrv_quorum = {
 
     .bdrv_file_open                     = quorum_open,
     .bdrv_close                         = quorum_close,
+    .bdrv_refresh_filename              = quorum_refresh_filename,
 
     .bdrv_co_flush_to_disk              = quorum_co_flush,
 

block/raw-aio.h

@@ -40,6 +40,8 @@ BlockDriverAIOCB *laio_submit(BlockDriverState *bs, void *aio_ctx, int fd,
         BlockDriverCompletionFunc *cb, void *opaque, int type);
 void laio_detach_aio_context(void *s, AioContext *old_context);
 void laio_attach_aio_context(void *s, AioContext *new_context);
+void laio_io_plug(BlockDriverState *bs, void *aio_ctx);
+int laio_io_unplug(BlockDriverState *bs, void *aio_ctx, bool unplug);
 #endif
 
 #ifdef _WIN32

block/raw-posix.c

@@ -55,6 +55,9 @@
 #include <linux/cdrom.h>
 #include <linux/fd.h>
 #include <linux/fs.h>
+#ifndef FS_NOCOW_FL
+#define FS_NOCOW_FL                     0x00800000 /* Do not cow file */
+#endif
 #endif
 #ifdef CONFIG_FIEMAP
 #include <linux/fiemap.h>
@@ -218,7 +221,7 @@ static int raw_normalize_devicepath(const char **filename)
 }
 #endif
 
-static void raw_probe_alignment(BlockDriverState *bs)
+static void raw_probe_alignment(BlockDriverState *bs, int fd, Error **errp)
 {
     BDRVRawState *s = bs->opaque;
     char *buf;
@@ -237,24 +240,24 @@ static void raw_probe_alignment(BlockDriverState *bs)
     s->buf_align = 0;
 
 #ifdef BLKSSZGET
-    if (ioctl(s->fd, BLKSSZGET, &sector_size) >= 0) {
+    if (ioctl(fd, BLKSSZGET, &sector_size) >= 0) {
         bs->request_alignment = sector_size;
     }
 #endif
 #ifdef DKIOCGETBLOCKSIZE
-    if (ioctl(s->fd, DKIOCGETBLOCKSIZE, &sector_size) >= 0) {
+    if (ioctl(fd, DKIOCGETBLOCKSIZE, &sector_size) >= 0) {
         bs->request_alignment = sector_size;
     }
 #endif
 #ifdef DIOCGSECTORSIZE
-    if (ioctl(s->fd, DIOCGSECTORSIZE, &sector_size) >= 0) {
+    if (ioctl(fd, DIOCGSECTORSIZE, &sector_size) >= 0) {
         bs->request_alignment = sector_size;
     }
 #endif
 #ifdef CONFIG_XFS
     if (s->is_xfs) {
         struct dioattr da;
-        if (xfsctl(NULL, s->fd, XFS_IOC_DIOINFO, &da) >= 0) {
+        if (xfsctl(NULL, fd, XFS_IOC_DIOINFO, &da) >= 0) {
             bs->request_alignment = da.d_miniosz;
             /* The kernel returns wrong information for d_mem */
             /* s->buf_align = da.d_mem; */
@@ -267,7 +270,7 @@ static void raw_probe_alignment(BlockDriverState *bs)
         size_t align;
         buf = qemu_memalign(MAX_BLOCKSIZE, 2 * MAX_BLOCKSIZE);
         for (align = 512; align <= MAX_BLOCKSIZE; align <<= 1) {
-            if (pread(s->fd, buf + align, MAX_BLOCKSIZE, 0) >= 0) {
+            if (pread(fd, buf + align, MAX_BLOCKSIZE, 0) >= 0) {
                 s->buf_align = align;
                 break;
             }
@@ -279,13 +282,18 @@ static void raw_probe_alignment(BlockDriverState *bs)
         size_t align;
         buf = qemu_memalign(s->buf_align, MAX_BLOCKSIZE);
         for (align = 512; align <= MAX_BLOCKSIZE; align <<= 1) {
-            if (pread(s->fd, buf, align, 0) >= 0) {
+            if (pread(fd, buf, align, 0) >= 0) {
                 bs->request_alignment = align;
                 break;
             }
         }
         qemu_vfree(buf);
     }
+
+    if (!s->buf_align || !bs->request_alignment) {
+        error_setg(errp, "Could not find working O_DIRECT alignment. "
+                         "Try cache.direct=off.");
+    }
 }
 
 static void raw_parse_flags(int bdrv_flags, int *open_flags)
@@ -502,13 +510,14 @@ static int raw_reopen_prepare(BDRVReopenState *state,
     BDRVRawState *s;
     BDRVRawReopenState *raw_s;
     int ret = 0;
+    Error *local_err = NULL;
 
     assert(state != NULL);
     assert(state->bs != NULL);
 
     s = state->bs->opaque;
 
-    state->opaque = g_malloc0(sizeof(BDRVRawReopenState));
+    state->opaque = g_new0(BDRVRawReopenState, 1);
     raw_s = state->opaque;
 
 #ifdef CONFIG_LINUX_AIO
@@ -574,6 +583,19 @@ static int raw_reopen_prepare(BDRVReopenState *state,
             ret = -1;
         }
     }
+
+    /* Fail already reopen_prepare() if we can't get a working O_DIRECT
+     * alignment with the new fd. */
+    if (raw_s->fd != -1) {
+        raw_probe_alignment(state->bs, raw_s->fd, &local_err);
+        if (local_err) {
+            qemu_close(raw_s->fd);
+            raw_s->fd = -1;
+            error_propagate(errp, local_err);
+            ret = -EINVAL;
+        }
+    }
+
     return ret;
 }
 
@@ -612,14 +634,12 @@ static void raw_reopen_abort(BDRVReopenState *state)
     state->opaque = NULL;
 }
 
-static int raw_refresh_limits(BlockDriverState *bs)
+static void raw_refresh_limits(BlockDriverState *bs, Error **errp)
 {
     BDRVRawState *s = bs->opaque;
 
-    raw_probe_alignment(bs);
+    raw_probe_alignment(bs, s->fd, errp);
     bs->bl.opt_mem_alignment = s->buf_align;
-
-    return 0;
 }
 
 static ssize_t handle_aiocb_ioctl(RawPosixAIOData *aiocb)
@@ -727,6 +747,15 @@ static ssize_t handle_aiocb_rw_linear(RawPosixAIOData *aiocb, char *buf)
         }
         if (len == -1 && errno == EINTR) {
             continue;
+        } else if (len == -1 && errno == EINVAL &&
+                   (aiocb->bs->open_flags & BDRV_O_NOCACHE) &&
+                   !(aiocb->aio_type & QEMU_AIO_WRITE) &&
+                   offset > 0) {
+            /* O_DIRECT pread() may fail with EINVAL when offset is unaligned
+             * after a short read.  Assume that O_DIRECT short reads only occur
+             * at EOF.  Therefore this is a short read, not an I/O error.
+             */
+            break;
         } else if (len == -1) {
             offset = -errno;
             break;
@@ -778,7 +807,11 @@ static ssize_t handle_aiocb_rw(RawPosixAIOData *aiocb)
      * Ok, we have to do it the hard way, copy all segments into
      * a single aligned buffer.
      */
-    buf = qemu_blockalign(aiocb->bs, aiocb->aio_nbytes);
+    buf = qemu_try_blockalign(aiocb->bs, aiocb->aio_nbytes);
+    if (buf == NULL) {
+        return -ENOMEM;
+    }
+
     if (aiocb->aio_type & QEMU_AIO_WRITE) {
         char *p = buf;
         int i;
@@ -787,6 +820,7 @@ static ssize_t handle_aiocb_rw(RawPosixAIOData *aiocb)
             memcpy(p, aiocb->aio_iov[i].iov_base, aiocb->aio_iov[i].iov_len);
             p += aiocb->aio_iov[i].iov_len;
         }
+        assert(p - buf == aiocb->aio_nbytes);
     }
 
     nbytes = handle_aiocb_rw_linear(aiocb, buf);
@@ -801,9 +835,11 @@ static ssize_t handle_aiocb_rw(RawPosixAIOData *aiocb)
                 copy = aiocb->aio_iov[i].iov_len;
             }
             memcpy(aiocb->aio_iov[i].iov_base, p, copy);
+            assert(count >= copy);
             p     += copy;
             count -= copy;
         }
+        assert(count == 0);
     }
     qemu_vfree(buf);
 
@@ -990,12 +1026,14 @@ static int paio_submit_co(BlockDriverState *bs, int fd,
     acb->aio_type = type;
     acb->aio_fildes = fd;
 
+    acb->aio_nbytes = nb_sectors * BDRV_SECTOR_SIZE;
+    acb->aio_offset = sector_num * BDRV_SECTOR_SIZE;
+
     if (qiov) {
         acb->aio_iov = qiov->iov;
         acb->aio_niov = qiov->niov;
+        assert(qiov->size == acb->aio_nbytes);
     }
-    acb->aio_nbytes = nb_sectors * 512;
-    acb->aio_offset = sector_num * 512;
 
     trace_paio_submit_co(sector_num, nb_sectors, type);
     pool = aio_get_thread_pool(bdrv_get_aio_context(bs));
@@ -1013,12 +1051,14 @@ static BlockDriverAIOCB *paio_submit(BlockDriverState *bs, int fd,
     acb->aio_type = type;
     acb->aio_fildes = fd;
 
+    acb->aio_nbytes = nb_sectors * BDRV_SECTOR_SIZE;
+    acb->aio_offset = sector_num * BDRV_SECTOR_SIZE;
+
     if (qiov) {
         acb->aio_iov = qiov->iov;
         acb->aio_niov = qiov->niov;
+        assert(qiov->size == acb->aio_nbytes);
     }
-    acb->aio_nbytes = nb_sectors * 512;
-    acb->aio_offset = sector_num * 512;
 
     trace_paio_submit(acb, opaque, sector_num, nb_sectors, type);
     pool = aio_get_thread_pool(bdrv_get_aio_context(bs));
@@ -1054,6 +1094,36 @@ static BlockDriverAIOCB *raw_aio_submit(BlockDriverState *bs,
                        cb, opaque, type);
 }
 
+static void raw_aio_plug(BlockDriverState *bs)
+{
+#ifdef CONFIG_LINUX_AIO
+    BDRVRawState *s = bs->opaque;
+    if (s->use_aio) {
+        laio_io_plug(bs, s->aio_ctx);
+    }
+#endif
+}
+
+static void raw_aio_unplug(BlockDriverState *bs)
+{
+#ifdef CONFIG_LINUX_AIO
+    BDRVRawState *s = bs->opaque;
+    if (s->use_aio) {
+        laio_io_unplug(bs, s->aio_ctx, true);
+    }
+#endif
+}
+
+static void raw_aio_flush_io_queue(BlockDriverState *bs)
+{
+#ifdef CONFIG_LINUX_AIO
+    BDRVRawState *s = bs->opaque;
+    if (s->use_aio) {
+        laio_io_unplug(bs, s->aio_ctx, false);
+    }
+#endif
+}
+
 static BlockDriverAIOCB *raw_aio_readv(BlockDriverState *bs,
         int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
         BlockDriverCompletionFunc *cb, void *opaque)
@@ -1130,12 +1200,12 @@ static int64_t raw_getlength(BlockDriverState *bs)
     struct stat st;
 
     if (fstat(fd, &st))
-        return -1;
+        return -errno;
     if (S_ISCHR(st.st_mode) || S_ISBLK(st.st_mode)) {
         struct disklabel dl;
 
         if (ioctl(fd, DIOCGDINFO, &dl))
-            return -1;
+            return -errno;
         return (uint64_t)dl.d_secsize *
             dl.d_partitions[DISKPART(st.st_rdev)].p_size;
     } else
@@ -1149,7 +1219,7 @@ static int64_t raw_getlength(BlockDriverState *bs)
     struct stat st;
 
     if (fstat(fd, &st))
-        return -1;
+        return -errno;
     if (S_ISCHR(st.st_mode) || S_ISBLK(st.st_mode)) {
         struct dkwedge_info dkw;
 
@@ -1159,7 +1229,7 @@ static int64_t raw_getlength(BlockDriverState *bs)
             struct disklabel dl;
 
             if (ioctl(fd, DIOCGDINFO, &dl))
-                return -1;
+                return -errno;
             return (uint64_t)dl.d_secsize *
                 dl.d_partitions[DISKPART(st.st_rdev)].p_size;
         }
@@ -1172,6 +1242,7 @@ static int64_t raw_getlength(BlockDriverState *bs)
     BDRVRawState *s = bs->opaque;
     struct dk_minfo minfo;
     int ret;
+    int64_t size;
 
     ret = fd_open(bs);
     if (ret < 0) {
@@ -1190,7 +1261,11 @@ static int64_t raw_getlength(BlockDriverState *bs)
      * There are reports that lseek on some devices fails, but
      * irc discussion said that contingency on contingency was overkill.
      */
-    return lseek(s->fd, 0, SEEK_END);
+    size = lseek(s->fd, 0, SEEK_END);
+    if (size < 0) {
+        return -errno;
+    }
+    return size;
 }
 #elif defined(CONFIG_BSD)
 static int64_t raw_getlength(BlockDriverState *bs)
@@ -1228,6 +1303,9 @@ again:
         size = LLONG_MAX;
 #else
         size = lseek(fd, 0LL, SEEK_END);
+        if (size < 0) {
+            return -errno;
+        }
 #endif
 #if defined(__FreeBSD__) || defined(__FreeBSD_kernel__)
         switch(s->type) {
@@ -1244,6 +1322,9 @@ again:
 #endif
     } else {
         size = lseek(fd, 0, SEEK_END);
+        if (size < 0) {
+            return -errno;
+        }
     }
     return size;
 }
@@ -1252,13 +1333,18 @@ static int64_t raw_getlength(BlockDriverState *bs)
 {
     BDRVRawState *s = bs->opaque;
     int ret;
+    int64_t size;
 
     ret = fd_open(bs);
     if (ret < 0) {
         return ret;
     }
 
-    return lseek(s->fd, 0, SEEK_END);
+    size = lseek(s->fd, 0, SEEK_END);
+    if (size < 0) {
+        return -errno;
+    }
+    return size;
 }
 #endif
 
@@ -1278,12 +1364,14 @@ static int raw_create(const char *filename, QemuOpts *opts, Error **errp)
     int fd;
     int result = 0;
     int64_t total_size = 0;
+    bool nocow = false;
 
     strstart(filename, "file:", &filename);
 
     /* Read out options */
     total_size =
         qemu_opt_get_size_del(opts, BLOCK_OPT_SIZE, 0) / BDRV_SECTOR_SIZE;
+    nocow = qemu_opt_get_bool(opts, BLOCK_OPT_NOCOW, false);
 
     fd = qemu_open(filename, O_WRONLY | O_CREAT | O_TRUNC | O_BINARY,
                    0644);
@@ -1291,6 +1379,21 @@ static int raw_create(const char *filename, QemuOpts *opts, Error **errp)
         result = -errno;
         error_setg_errno(errp, -result, "Could not create file");
     } else {
+        if (nocow) {
+#ifdef __linux__
+            /* Set NOCOW flag to solve performance issue on fs like btrfs.
+             * This is an optimisation. The FS_IOC_SETFLAGS ioctl return value
+             * will be ignored since any failure of this operation should not
+             * block the left work.
+             */
+            int attr;
+            if (ioctl(fd, FS_IOC_GETFLAGS, &attr) == 0) {
+                attr |= FS_NOCOW_FL;
+                ioctl(fd, FS_IOC_SETFLAGS, &attr);
+            }
+#endif
+        }
+
         if (ftruncate(fd, total_size * BDRV_SECTOR_SIZE) != 0) {
             result = -errno;
             error_setg_errno(errp, -result, "Could not resize file");
@@ -1477,6 +1580,11 @@ static QemuOptsList raw_create_opts = {
             .type = QEMU_OPT_SIZE,
             .help = "Virtual disk size"
         },
+        {
+            .name = BLOCK_OPT_NOCOW,
+            .type = QEMU_OPT_BOOL,
+            .help = "Turn off copy-on-write (valid only on btrfs)"
+        },
         { /* end of list */ }
     }
 };
@@ -1503,6 +1611,9 @@ static BlockDriver bdrv_file = {
     .bdrv_aio_flush = raw_aio_flush,
     .bdrv_aio_discard = raw_aio_discard,
     .bdrv_refresh_limits = raw_refresh_limits,
+    .bdrv_io_plug = raw_aio_plug,
+    .bdrv_io_unplug = raw_aio_unplug,
+    .bdrv_flush_io_queue = raw_aio_flush_io_queue,
 
     .bdrv_truncate = raw_truncate,
     .bdrv_getlength = raw_getlength,
@@ -1902,6 +2013,9 @@ static BlockDriver bdrv_host_device = {
     .bdrv_aio_flush	= raw_aio_flush,
     .bdrv_aio_discard   = hdev_aio_discard,
     .bdrv_refresh_limits = raw_refresh_limits,
+    .bdrv_io_plug = raw_aio_plug,
+    .bdrv_io_unplug = raw_aio_unplug,
+    .bdrv_flush_io_queue = raw_aio_flush_io_queue,
 
     .bdrv_truncate      = raw_truncate,
     .bdrv_getlength	= raw_getlength,
@@ -2047,6 +2161,9 @@ static BlockDriver bdrv_host_floppy = {
     .bdrv_aio_writev    = raw_aio_writev,
     .bdrv_aio_flush	= raw_aio_flush,
     .bdrv_refresh_limits = raw_refresh_limits,
+    .bdrv_io_plug = raw_aio_plug,
+    .bdrv_io_unplug = raw_aio_unplug,
+    .bdrv_flush_io_queue = raw_aio_flush_io_queue,
 
     .bdrv_truncate      = raw_truncate,
     .bdrv_getlength      = raw_getlength,
@@ -2175,6 +2292,9 @@ static BlockDriver bdrv_host_cdrom = {
     .bdrv_aio_writev    = raw_aio_writev,
     .bdrv_aio_flush	= raw_aio_flush,
     .bdrv_refresh_limits = raw_refresh_limits,
+    .bdrv_io_plug = raw_aio_plug,
+    .bdrv_io_unplug = raw_aio_unplug,
+    .bdrv_flush_io_queue = raw_aio_flush_io_queue,
 
     .bdrv_truncate      = raw_truncate,
     .bdrv_getlength      = raw_getlength,
@@ -2309,6 +2429,9 @@ static BlockDriver bdrv_host_cdrom = {
     .bdrv_aio_writev    = raw_aio_writev,
     .bdrv_aio_flush	= raw_aio_flush,
     .bdrv_refresh_limits = raw_refresh_limits,
+    .bdrv_io_plug = raw_aio_plug,
+    .bdrv_io_unplug = raw_aio_unplug,
+    .bdrv_flush_io_queue = raw_aio_flush_io_queue,
 
     .bdrv_truncate      = raw_truncate,
     .bdrv_getlength      = raw_getlength,

block/raw_bsd.c

@@ -94,10 +94,9 @@ static int raw_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
     return bdrv_get_info(bs->file, bdi);
 }
 
-static int raw_refresh_limits(BlockDriverState *bs)
+static void raw_refresh_limits(BlockDriverState *bs, Error **errp)
 {
     bs->bl = bs->file->bl;
-    return 0;
 }
 
 static int raw_truncate(BlockDriverState *bs, int64_t offset)

block/rbd.c

@@ -617,7 +617,7 @@ static BlockDriverAIOCB *rbd_start_aio(BlockDriverState *bs,
                                        RBDAIOCmd cmd)
 {
     RBDAIOCB *acb;
-    RADOSCB *rcb;
+    RADOSCB *rcb = NULL;
     rbd_completion_t c;
     int64_t off, size;
     char *buf;
@@ -631,7 +631,10 @@ static BlockDriverAIOCB *rbd_start_aio(BlockDriverState *bs,
     if (cmd == RBD_AIO_DISCARD || cmd == RBD_AIO_FLUSH) {
         acb->bounce = NULL;
     } else {
-        acb->bounce = qemu_blockalign(bs, qiov->size);
+        acb->bounce = qemu_try_blockalign(bs, qiov->size);
+        if (acb->bounce == NULL) {
+            goto failed;
+        }
     }
     acb->ret = 0;
     acb->error = 0;
@@ -649,7 +652,7 @@ static BlockDriverAIOCB *rbd_start_aio(BlockDriverState *bs,
     off = sector_num * BDRV_SECTOR_SIZE;
     size = nb_sectors * BDRV_SECTOR_SIZE;
 
-    rcb = g_malloc(sizeof(RADOSCB));
+    rcb = g_new(RADOSCB, 1);
     rcb->done = 0;
     rcb->acb = acb;
     rcb->buf = buf;
@@ -859,7 +862,7 @@ static int qemu_rbd_snap_list(BlockDriverState *bs,
     int max_snaps = RBD_MAX_SNAPS;
 
     do {
-        snaps = g_malloc(sizeof(*snaps) * max_snaps);
+        snaps = g_new(rbd_snap_info_t, max_snaps);
         snap_count = rbd_snap_list(s->image, snaps, &max_snaps);
         if (snap_count <= 0) {
             g_free(snaps);
@@ -870,7 +873,7 @@ static int qemu_rbd_snap_list(BlockDriverState *bs,
         goto done;
     }
 
-    sn_tab = g_malloc0(snap_count * sizeof(QEMUSnapshotInfo));
+    sn_tab = g_new0(QEMUSnapshotInfo, snap_count);
 
     for (i = 0; i < snap_count; i++) {
         const char *snap_name = snaps[i].name;

block/sheepdog.c

@@ -103,6 +103,9 @@
 #define SD_INODE_SIZE (sizeof(SheepdogInode))
 #define CURRENT_VDI_ID 0
 
+#define LOCK_TYPE_NORMAL 0
+#define LOCK_TYPE_SHARED 1      /* for iSCSI multipath */
+
 typedef struct SheepdogReq {
     uint8_t proto_ver;
     uint8_t opcode;
@@ -166,7 +169,8 @@ typedef struct SheepdogVdiReq {
     uint8_t copy_policy;
     uint8_t reserved[2];
     uint32_t snapid;
-    uint32_t pad[3];
+    uint32_t type;
+    uint32_t pad[2];
 } SheepdogVdiReq;
 
 typedef struct SheepdogVdiRsp {
@@ -712,7 +716,6 @@ static void coroutine_fn send_pending_req(BDRVSheepdogState *s, uint64_t oid)
 
 static coroutine_fn void reconnect_to_sdog(void *opaque)
 {
-    Error *local_err = NULL;
     BDRVSheepdogState *s = opaque;
     AIOReq *aio_req, *next;
 
@@ -727,6 +730,7 @@ static coroutine_fn void reconnect_to_sdog(void *opaque)
 
     /* Try to reconnect the sheepdog server every one second. */
     while (s->fd < 0) {
+        Error *local_err = NULL;
         s->fd = get_sheep_fd(s, &local_err);
         if (s->fd < 0) {
             DPRINTF("Wait for connection to be established\n");
@@ -1090,6 +1094,7 @@ static int find_vdi_name(BDRVSheepdogState *s, const char *filename,
     memset(&hdr, 0, sizeof(hdr));
     if (lock) {
         hdr.opcode = SD_OP_LOCK_VDI;
+        hdr.type = LOCK_TYPE_NORMAL;
     } else {
         hdr.opcode = SD_OP_GET_VDI_INFO;
     }
@@ -1110,6 +1115,8 @@ static int find_vdi_name(BDRVSheepdogState *s, const char *filename,
                    sd_strerror(rsp->result), filename, snapid, tag);
         if (rsp->result == SD_RES_NO_VDI) {
             ret = -ENOENT;
+        } else if (rsp->result == SD_RES_VDI_LOCKED) {
+            ret = -EBUSY;
         } else {
             ret = -EIO;
         }
@@ -1682,7 +1689,7 @@ static int sd_create(const char *filename, QemuOpts *opts,
     uint32_t snapid;
     bool prealloc = false;
 
-    s = g_malloc0(sizeof(BDRVSheepdogState));
+    s = g_new0(BDRVSheepdogState, 1);
 
     memset(tag, 0, sizeof(tag));
     if (strstr(filename, "://")) {
@@ -1793,6 +1800,7 @@ static void sd_close(BlockDriverState *bs)
     memset(&hdr, 0, sizeof(hdr));
 
     hdr.opcode = SD_OP_RELEASE_VDI;
+    hdr.type = LOCK_TYPE_NORMAL;
     hdr.base_vdi_id = s->inode.vdi_id;
     wlen = strlen(s->name) + 1;
     hdr.data_length = wlen;
@@ -2273,7 +2281,7 @@ static int sd_snapshot_goto(BlockDriverState *bs, const char *snapshot_id)
     uint32_t snapid = 0;
     int ret = 0;
 
-    old_s = g_malloc(sizeof(BDRVSheepdogState));
+    old_s = g_new(BDRVSheepdogState, 1);
 
     memcpy(old_s, s, sizeof(BDRVSheepdogState));
 
@@ -2357,7 +2365,7 @@ static int sd_snapshot_list(BlockDriverState *bs, QEMUSnapshotInfo **psn_tab)
         goto out;
     }
 
-    sn_tab = g_malloc0(nr * sizeof(*sn_tab));
+    sn_tab = g_new0(QEMUSnapshotInfo, nr);
 
     /* calculate a vdi id with hash function */
     hval = fnv_64a_buf(s->name, strlen(s->name), FNV1A_64_INIT);

block/stream.c

@@ -32,7 +32,7 @@ typedef struct StreamBlockJob {
     RateLimit limit;
     BlockDriverState *base;
     BlockdevOnError on_error;
-    char backing_file_id[1024];
+    char *backing_file_str;
 } StreamBlockJob;
 
 static int coroutine_fn stream_populate(BlockDriverState *bs,
@@ -76,7 +76,7 @@ static void close_unused_images(BlockDriverState *top, BlockDriverState *base,
         bdrv_unref(unused);
     }
 
-    bdrv_refresh_limits(top);
+    bdrv_refresh_limits(top, NULL);
 }
 
 static void coroutine_fn stream_run(void *opaque)
@@ -186,7 +186,7 @@ wait:
     if (!block_job_is_cancelled(&s->common) && sector_num == end && ret == 0) {
         const char *base_id = NULL, *base_fmt = NULL;
         if (base) {
-            base_id = s->backing_file_id;
+            base_id = s->backing_file_str;
             if (base->drv) {
                 base_fmt = base->drv->format_name;
             }
@@ -196,6 +196,7 @@ wait:
     }
 
     qemu_vfree(buf);
+    g_free(s->backing_file_str);
     block_job_completed(&s->common, ret);
 }
 
@@ -217,7 +218,7 @@ static const BlockJobDriver stream_job_driver = {
 };
 
 void stream_start(BlockDriverState *bs, BlockDriverState *base,
-                  const char *base_id, int64_t speed,
+                  const char *backing_file_str, int64_t speed,
                   BlockdevOnError on_error,
                   BlockDriverCompletionFunc *cb,
                   void *opaque, Error **errp)
@@ -237,9 +238,7 @@ void stream_start(BlockDriverState *bs, BlockDriverState *base,
     }
 
     s->base = base;
-    if (base_id) {
-        pstrcpy(s->backing_file_id, sizeof(s->backing_file_id), base_id);
-    }
+    s->backing_file_str = g_strdup(backing_file_str);
 
     s->on_error = on_error;
     s->common.co = qemu_coroutine_create(stream_run);

block/vdi.c

@@ -292,7 +292,12 @@ static int vdi_check(BlockDriverState *bs, BdrvCheckResult *res,
         return -ENOTSUP;
     }
 
-    bmap = g_malloc(s->header.blocks_in_image * sizeof(uint32_t));
+    bmap = g_try_new(uint32_t, s->header.blocks_in_image);
+    if (s->header.blocks_in_image && bmap == NULL) {
+        res->check_errors++;
+        return -ENOMEM;
+    }
+
     memset(bmap, 0xff, s->header.blocks_in_image * sizeof(uint32_t));
 
     /* Check block map and value of blocks_allocated. */
@@ -350,23 +355,23 @@ static int vdi_make_empty(BlockDriverState *bs)
 static int vdi_probe(const uint8_t *buf, int buf_size, const char *filename)
 {
     const VdiHeader *header = (const VdiHeader *)buf;
-    int result = 0;
+    int ret = 0;
 
     logout("\n");
 
     if (buf_size < sizeof(*header)) {
         /* Header too small, no VDI. */
     } else if (le32_to_cpu(header->signature) == VDI_SIGNATURE) {
-        result = 100;
+        ret = 100;
     }
 
-    if (result == 0) {
+    if (ret == 0) {
         logout("no vdi image\n");
     } else {
         logout("%s", header->text);
     }
 
-    return result;
+    return ret;
 }
 
 static int vdi_open(BlockDriverState *bs, QDict *options, int flags,
@@ -471,7 +476,12 @@ static int vdi_open(BlockDriverState *bs, QDict *options, int flags,
 
     bmap_size = header.blocks_in_image * sizeof(uint32_t);
     bmap_size = (bmap_size + SECTOR_SIZE - 1) / SECTOR_SIZE;
-    s->bmap = g_malloc(bmap_size * SECTOR_SIZE);
+    s->bmap = qemu_try_blockalign(bs->file, bmap_size * SECTOR_SIZE);
+    if (s->bmap == NULL) {
+        ret = -ENOMEM;
+        goto fail;
+    }
+
     ret = bdrv_read(bs->file, s->bmap_sector, (uint8_t *)s->bmap, bmap_size);
     if (ret < 0) {
         goto fail_free_bmap;
@@ -486,7 +496,7 @@ static int vdi_open(BlockDriverState *bs, QDict *options, int flags,
     return 0;
 
  fail_free_bmap:
-    g_free(s->bmap);
+    qemu_vfree(s->bmap);
 
  fail:
     return ret;
@@ -674,8 +684,7 @@ static int vdi_co_write(BlockDriverState *bs,
 
 static int vdi_create(const char *filename, QemuOpts *opts, Error **errp)
 {
-    int fd;
-    int result = 0;
+    int ret = 0;
     uint64_t bytes = 0;
     uint32_t blocks;
     size_t block_size = DEFAULT_CLUSTER_SIZE;
@@ -683,6 +692,10 @@ static int vdi_create(const char *filename, QemuOpts *opts, Error **errp)
     VdiHeader header;
     size_t i;
     size_t bmap_size;
+    int64_t offset = 0;
+    Error *local_err = NULL;
+    BlockDriverState *bs = NULL;
+    uint32_t *bmap = NULL;
 
     logout("\n");
 
@@ -701,18 +714,22 @@ static int vdi_create(const char *filename, QemuOpts *opts, Error **errp)
 #endif
 
     if (bytes > VDI_DISK_SIZE_MAX) {
-        result = -ENOTSUP;
+        ret = -ENOTSUP;
         error_setg(errp, "Unsupported VDI image size (size is 0x%" PRIx64
                           ", max supported is 0x%" PRIx64 ")",
                           bytes, VDI_DISK_SIZE_MAX);
         goto exit;
     }
 
-    fd = qemu_open(filename,
-                   O_WRONLY | O_CREAT | O_TRUNC | O_BINARY | O_LARGEFILE,
-                   0644);
-    if (fd < 0) {
-        result = -errno;
+    ret = bdrv_create_file(filename, opts, &local_err);
+    if (ret < 0) {
+        error_propagate(errp, local_err);
+        goto exit;
+    }
+    ret = bdrv_open(&bs, filename, NULL, NULL, BDRV_O_RDWR | BDRV_O_PROTOCOL,
+                    NULL, &local_err);
+    if (ret < 0) {
+        error_propagate(errp, local_err);
         goto exit;
     }
 
@@ -745,13 +762,20 @@ static int vdi_create(const char *filename, QemuOpts *opts, Error **errp)
     vdi_header_print(&header);
 #endif
     vdi_header_to_le(&header);
-    if (write(fd, &header, sizeof(header)) < 0) {
-        result = -errno;
-        goto close_and_exit;
+    ret = bdrv_pwrite_sync(bs, offset, &header, sizeof(header));
+    if (ret < 0) {
+        error_setg(errp, "Error writing header to %s", filename);
+        goto exit;
     }
+    offset += sizeof(header);
 
     if (bmap_size > 0) {
-        uint32_t *bmap = g_malloc0(bmap_size);
+        bmap = g_try_malloc0(bmap_size);
+        if (bmap == NULL) {
+            ret = -ENOMEM;
+            error_setg(errp, "Could not allocate bmap");
+            goto exit;
+        }
         for (i = 0; i < blocks; i++) {
             if (image_type == VDI_TYPE_STATIC) {
                 bmap[i] = i;
@@ -759,35 +783,33 @@ static int vdi_create(const char *filename, QemuOpts *opts, Error **errp)
                 bmap[i] = VDI_UNALLOCATED;
             }
         }
-        if (write(fd, bmap, bmap_size) < 0) {
-            result = -errno;
-            g_free(bmap);
-            goto close_and_exit;
+        ret = bdrv_pwrite_sync(bs, offset, bmap, bmap_size);
+        if (ret < 0) {
+            error_setg(errp, "Error writing bmap to %s", filename);
+            goto exit;
         }
-        g_free(bmap);
+        offset += bmap_size;
     }
 
     if (image_type == VDI_TYPE_STATIC) {
-        if (ftruncate(fd, sizeof(header) + bmap_size + blocks * block_size)) {
-            result = -errno;
-            goto close_and_exit;
+        ret = bdrv_truncate(bs, offset + blocks * block_size);
+        if (ret < 0) {
+            error_setg(errp, "Failed to statically allocate %s", filename);
+            goto exit;
         }
     }
 
-close_and_exit:
-    if ((close(fd) < 0) && !result) {
-        result = -errno;
-    }
-
 exit:
-    return result;
+    bdrv_unref(bs);
+    g_free(bmap);
+    return ret;
 }
 
 static void vdi_close(BlockDriverState *bs)
 {
     BDRVVdiState *s = bs->opaque;
 
-    g_free(s->bmap);
+    qemu_vfree(s->bmap);
 
     migrate_del_blocker(s->migration_blocker);
     error_free(s->migration_blocker);
@@ -818,6 +840,11 @@ static QemuOptsList vdi_create_opts = {
             .def_value_str = "off"
         },
 #endif
+        {
+            .name = BLOCK_OPT_NOCOW,
+            .type = QEMU_OPT_BOOL,
+            .help = "Turn off copy-on-write (valid only on btrfs)"
+        },
         /* TODO: An additional option to set UUID values might be useful. */
         { /* end of list */ }
     }

block/vhdx-endian.c

@@ -82,8 +82,6 @@ void vhdx_log_desc_le_import(VHDXLogDescriptor *d)
     assert(d != NULL);
 
     le32_to_cpus(&d->signature);
-    le32_to_cpus(&d->trailing_bytes);
-    le64_to_cpus(&d->leading_bytes);
     le64_to_cpus(&d->file_offset);
     le64_to_cpus(&d->sequence_number);
 }
@@ -99,6 +97,15 @@ void vhdx_log_desc_le_export(VHDXLogDescriptor *d)
     cpu_to_le64s(&d->sequence_number);
 }
 
+void vhdx_log_data_le_import(VHDXLogDataSector *d)
+{
+    assert(d != NULL);
+
+    le32_to_cpus(&d->data_signature);
+    le32_to_cpus(&d->sequence_high);
+    le32_to_cpus(&d->sequence_low);
+}
+
 void vhdx_log_data_le_export(VHDXLogDataSector *d)
 {
     assert(d != NULL);

block/vhdx-log.c

@@ -84,6 +84,7 @@ static int vhdx_log_peek_hdr(BlockDriverState *bs, VHDXLogEntries *log,
     if (ret < 0) {
         goto exit;
     }
+    vhdx_log_entry_hdr_le_import(hdr);
 
 exit:
     return ret;
@@ -211,7 +212,7 @@ static bool vhdx_log_hdr_is_valid(VHDXLogEntries *log, VHDXLogEntryHeader *hdr,
 {
     int valid = false;
 
-    if (memcmp(&hdr->signature, "loge", 4)) {
+    if (hdr->signature != VHDX_LOG_SIGNATURE) {
         goto exit;
     }
 
@@ -275,12 +276,12 @@ static bool vhdx_log_desc_is_valid(VHDXLogDescriptor *desc,
         goto exit;
     }
 
-    if (!memcmp(&desc->signature, "zero", 4)) {
+    if (desc->signature == VHDX_LOG_ZERO_SIGNATURE) {
         if (desc->zero_length % VHDX_LOG_SECTOR_SIZE == 0) {
             /* valid */
             ret = true;
         }
-    } else if (!memcmp(&desc->signature, "desc", 4)) {
+    } else if (desc->signature == VHDX_LOG_DESC_SIGNATURE) {
             /* valid */
             ret = true;
     }
@@ -327,13 +328,15 @@ static int vhdx_compute_desc_sectors(uint32_t desc_cnt)
  * passed into this function. Each descriptor will also be validated,
  * and error returned if any are invalid. */
 static int vhdx_log_read_desc(BlockDriverState *bs, BDRVVHDXState *s,
-                              VHDXLogEntries *log, VHDXLogDescEntries **buffer)
+                              VHDXLogEntries *log, VHDXLogDescEntries **buffer,
+                              bool convert_endian)
 {
     int ret = 0;
     uint32_t desc_sectors;
     uint32_t sectors_read;
     VHDXLogEntryHeader hdr;
     VHDXLogDescEntries *desc_entries = NULL;
+    VHDXLogDescriptor desc;
     int i;
 
     assert(*buffer == NULL);
@@ -342,14 +345,19 @@ static int vhdx_log_read_desc(BlockDriverState *bs, BDRVVHDXState *s,
     if (ret < 0) {
         goto exit;
     }
-    vhdx_log_entry_hdr_le_import(&hdr);
+
     if (vhdx_log_hdr_is_valid(log, &hdr, s) == false) {
         ret = -EINVAL;
         goto exit;
     }
 
     desc_sectors = vhdx_compute_desc_sectors(hdr.descriptor_count);
-    desc_entries = qemu_blockalign(bs, desc_sectors * VHDX_LOG_SECTOR_SIZE);
+    desc_entries = qemu_try_blockalign(bs->file,
+                                       desc_sectors * VHDX_LOG_SECTOR_SIZE);
+    if (desc_entries == NULL) {
+        ret = -ENOMEM;
+        goto exit;
+    }
 
     ret = vhdx_log_read_sectors(bs, log, &sectors_read, desc_entries,
                                 desc_sectors, false);
@@ -363,12 +371,19 @@ static int vhdx_log_read_desc(BlockDriverState *bs, BDRVVHDXState *s,
 
     /* put in proper endianness, and validate each desc */
     for (i = 0; i < hdr.descriptor_count; i++) {
-        vhdx_log_desc_le_import(&desc_entries->desc[i]);
-        if (vhdx_log_desc_is_valid(&desc_entries->desc[i], &hdr) == false) {
+        desc = desc_entries->desc[i];
+        vhdx_log_desc_le_import(&desc);
+        if (convert_endian) {
+            desc_entries->desc[i] = desc;
+        }
+        if (vhdx_log_desc_is_valid(&desc, &hdr) == false) {
             ret = -EINVAL;
             goto free_and_exit;
         }
     }
+    if (convert_endian) {
+        desc_entries->hdr = hdr;
+    }
 
     *buffer = desc_entries;
     goto exit;
@@ -403,7 +418,7 @@ static int vhdx_log_flush_desc(BlockDriverState *bs, VHDXLogDescriptor *desc,
 
     buffer = qemu_blockalign(bs, VHDX_LOG_SECTOR_SIZE);
 
-    if (!memcmp(&desc->signature, "desc", 4)) {
+    if (desc->signature == VHDX_LOG_DESC_SIGNATURE) {
         /* data sector */
         if (data == NULL) {
             ret = -EFAULT;
@@ -431,10 +446,15 @@ static int vhdx_log_flush_desc(BlockDriverState *bs, VHDXLogDescriptor *desc,
 
         memcpy(buffer+offset, &desc->trailing_bytes, 4);
 
-    } else if (!memcmp(&desc->signature, "zero", 4)) {
+    } else if (desc->signature == VHDX_LOG_ZERO_SIGNATURE) {
         /* write 'count' sectors of sector */
         memset(buffer, 0, VHDX_LOG_SECTOR_SIZE);
         count = desc->zero_length / VHDX_LOG_SECTOR_SIZE;
+    } else {
+        error_report("Invalid VHDX log descriptor entry signature 0x%" PRIx32,
+                      desc->signature);
+        ret = -EINVAL;
+        goto exit;
     }
 
     file_offset = desc->file_offset;
@@ -493,13 +513,13 @@ static int vhdx_log_flush(BlockDriverState *bs, BDRVVHDXState *s,
             goto exit;
         }
 
-        ret = vhdx_log_read_desc(bs, s, &logs->log, &desc_entries);
+        ret = vhdx_log_read_desc(bs, s, &logs->log, &desc_entries, true);
         if (ret < 0) {
             goto exit;
         }
 
         for (i = 0; i < desc_entries->hdr.descriptor_count; i++) {
-            if (!memcmp(&desc_entries->desc[i].signature, "desc", 4)) {
+            if (desc_entries->desc[i].signature == VHDX_LOG_DESC_SIGNATURE) {
                 /* data sector, so read a sector to flush */
                 ret = vhdx_log_read_sectors(bs, &logs->log, &sectors_read,
                                             data, 1, false);
@@ -510,6 +530,7 @@ static int vhdx_log_flush(BlockDriverState *bs, BDRVVHDXState *s,
                     ret = -EINVAL;
                     goto exit;
                 }
+                vhdx_log_data_le_import(data);
             }
 
             ret = vhdx_log_flush_desc(bs, &desc_entries->desc[i], data);
@@ -558,9 +579,6 @@ static int vhdx_validate_log_entry(BlockDriverState *bs, BDRVVHDXState *s,
         goto inc_and_exit;
     }
 
-    vhdx_log_entry_hdr_le_import(&hdr);
-
-
     if (vhdx_log_hdr_is_valid(log, &hdr, s) == false) {
         goto inc_and_exit;
     }
@@ -573,13 +591,13 @@ static int vhdx_validate_log_entry(BlockDriverState *bs, BDRVVHDXState *s,
 
     desc_sectors = vhdx_compute_desc_sectors(hdr.descriptor_count);
 
-    /* Read desc sectors, and calculate log checksum */
+    /* Read all log sectors, and calculate log checksum */
 
     total_sectors = hdr.entry_length / VHDX_LOG_SECTOR_SIZE;
 
 
     /* read_desc() will increment the read idx */
-    ret = vhdx_log_read_desc(bs, s, log, &desc_buffer);
+    ret = vhdx_log_read_desc(bs, s, log, &desc_buffer, false);
     if (ret < 0) {
         goto free_and_exit;
     }
@@ -602,7 +620,7 @@ static int vhdx_validate_log_entry(BlockDriverState *bs, BDRVVHDXState *s,
         }
     }
     crc ^= 0xffffffff;
-    if (crc != desc_buffer->hdr.checksum) {
+    if (crc != hdr.checksum) {
         goto free_and_exit;
     }
 
@@ -905,7 +923,7 @@ static int vhdx_log_write(BlockDriverState *bs, BDRVVHDXState *s,
     buffer = qemu_blockalign(bs, total_length);
     memcpy(buffer, &new_hdr, sizeof(new_hdr));
 
-    new_desc = (VHDXLogDescriptor *) (buffer + sizeof(new_hdr));
+    new_desc = buffer + sizeof(new_hdr);
     data_sector = buffer + (desc_sectors * VHDX_LOG_SECTOR_SIZE);
     data_tmp = data;
 
@@ -962,7 +980,6 @@ static int vhdx_log_write(BlockDriverState *bs, BDRVVHDXState *s,
      * last data sector */
     vhdx_update_checksum(buffer, total_length,
                          offsetof(VHDXLogEntryHeader, checksum));
-    cpu_to_le32s((uint32_t *)(buffer + 4));
 
     /* now write to the log */
     ret = vhdx_log_write_sectors(bs, &s->log, &sectors_written, buffer,

block/vhdx.c

@@ -135,10 +135,8 @@ typedef struct VHDXSectorInfo {
  * buf: buffer pointer
  * size: size of buffer (must be > crc_offset+4)
  *
- * Note: The resulting checksum is in the CPU endianness, not necessarily
- *       in the file format endianness (LE).  Any header export to disk should
- *       make sure that vhdx_header_le_export() is used to convert to the
- *       correct endianness
+ * Note: The buffer should have all multi-byte data in little-endian format,
+ *       and the resulting checksum is in little endian format.
  */
 uint32_t vhdx_update_checksum(uint8_t *buf, size_t size, int crc_offset)
 {
@@ -149,6 +147,7 @@ uint32_t vhdx_update_checksum(uint8_t *buf, size_t size, int crc_offset)
 
     memset(buf + crc_offset, 0, sizeof(crc));
     crc =  crc32c(0xffffffff, buf, size);
+    cpu_to_le32s(&crc);
     memcpy(buf + crc_offset, &crc, sizeof(crc));
 
     return crc;
@@ -300,7 +299,7 @@ static int vhdx_write_header(BlockDriverState *bs_file, VHDXHeader *hdr,
 {
     uint8_t *buffer = NULL;
     int ret;
-    VHDXHeader header_le;
+    VHDXHeader *header_le;
 
     assert(bs_file != NULL);
     assert(hdr != NULL);
@@ -321,11 +320,12 @@ static int vhdx_write_header(BlockDriverState *bs_file, VHDXHeader *hdr,
     }
 
     /* overwrite the actual VHDXHeader portion */
-    memcpy(buffer, hdr, sizeof(VHDXHeader));
-    hdr->checksum = vhdx_update_checksum(buffer, VHDX_HEADER_SIZE,
-                                         offsetof(VHDXHeader, checksum));
-    vhdx_header_le_export(hdr, &header_le);
-    ret = bdrv_pwrite_sync(bs_file, offset, &header_le, sizeof(VHDXHeader));
+    header_le = (VHDXHeader *)buffer;
+    memcpy(header_le, hdr, sizeof(VHDXHeader));
+    vhdx_header_le_export(hdr, header_le);
+    vhdx_update_checksum(buffer, VHDX_HEADER_SIZE,
+                         offsetof(VHDXHeader, checksum));
+    ret = bdrv_pwrite_sync(bs_file, offset, header_le, sizeof(VHDXHeader));
 
 exit:
     qemu_vfree(buffer);
@@ -432,13 +432,14 @@ static void vhdx_parse_header(BlockDriverState *bs, BDRVVHDXState *s,
     }
     /* copy over just the relevant portion that we need */
     memcpy(header1, buffer, sizeof(VHDXHeader));
-    vhdx_header_le_import(header1);
 
-    if (vhdx_checksum_is_valid(buffer, VHDX_HEADER_SIZE, 4) &&
-        !memcmp(&header1->signature, "head", 4)             &&
-        header1->version == 1) {
-        h1_seq = header1->sequence_number;
-        h1_valid = true;
+    if (vhdx_checksum_is_valid(buffer, VHDX_HEADER_SIZE, 4)) {
+        vhdx_header_le_import(header1);
+        if (header1->signature == VHDX_HEADER_SIGNATURE &&
+            header1->version == 1) {
+            h1_seq = header1->sequence_number;
+            h1_valid = true;
+        }
     }
 
     ret = bdrv_pread(bs->file, VHDX_HEADER2_OFFSET, buffer, VHDX_HEADER_SIZE);
@@ -447,13 +448,14 @@ static void vhdx_parse_header(BlockDriverState *bs, BDRVVHDXState *s,
     }
     /* copy over just the relevant portion that we need */
     memcpy(header2, buffer, sizeof(VHDXHeader));
-    vhdx_header_le_import(header2);
 
-    if (vhdx_checksum_is_valid(buffer, VHDX_HEADER_SIZE, 4) &&
-        !memcmp(&header2->signature, "head", 4)             &&
-        header2->version == 1) {
-        h2_seq = header2->sequence_number;
-        h2_valid = true;
+    if (vhdx_checksum_is_valid(buffer, VHDX_HEADER_SIZE, 4)) {
+        vhdx_header_le_import(header2);
+        if (header2->signature == VHDX_HEADER_SIGNATURE &&
+            header2->version == 1) {
+            h2_seq = header2->sequence_number;
+            h2_valid = true;
+        }
     }
 
     /* If there is only 1 valid header (or no valid headers), we
@@ -519,15 +521,21 @@ static int vhdx_open_region_tables(BlockDriverState *bs, BDRVVHDXState *s)
         goto fail;
     }
     memcpy(&s->rt, buffer, sizeof(s->rt));
-    vhdx_region_header_le_import(&s->rt);
     offset += sizeof(s->rt);
 
-    if (!vhdx_checksum_is_valid(buffer, VHDX_HEADER_BLOCK_SIZE, 4) ||
-        memcmp(&s->rt.signature, "regi", 4)) {
+    if (!vhdx_checksum_is_valid(buffer, VHDX_HEADER_BLOCK_SIZE, 4)) {
         ret = -EINVAL;
         goto fail;
     }
 
+    vhdx_region_header_le_import(&s->rt);
+
+    if (s->rt.signature != VHDX_REGION_SIGNATURE) {
+        ret = -EINVAL;
+        goto fail;
+    }
+
+
     /* Per spec, maximum region table entry count is 2047 */
     if (s->rt.entry_count > 2047) {
         ret = -EINVAL;
@@ -630,7 +638,7 @@ static int vhdx_parse_metadata(BlockDriverState *bs, BDRVVHDXState *s)
 
     vhdx_metadata_header_le_import(&s->metadata_hdr);
 
-    if (memcmp(&s->metadata_hdr.signature, "metadata", 8)) {
+    if (s->metadata_hdr.signature != VHDX_METADATA_SIGNATURE) {
         ret = -EINVAL;
         goto exit;
     }
@@ -950,7 +958,11 @@ static int vhdx_open(BlockDriverState *bs, QDict *options, int flags,
     }
 
     /* s->bat is freed in vhdx_close() */
-    s->bat = qemu_blockalign(bs, s->bat_rt.length);
+    s->bat = qemu_try_blockalign(bs->file, s->bat_rt.length);
+    if (s->bat == NULL) {
+        ret = -ENOMEM;
+        goto fail;
+    }
 
     ret = bdrv_pread(bs->file, s->bat_offset, s->bat, s->bat_rt.length);
     if (ret < 0) {
@@ -1369,7 +1381,7 @@ static int vhdx_create_new_headers(BlockDriverState *bs, uint64_t image_size,
     int ret = 0;
     VHDXHeader *hdr = NULL;
 
-    hdr = g_malloc0(sizeof(VHDXHeader));
+    hdr = g_new0(VHDXHeader, 1);
 
     hdr->signature       = VHDX_HEADER_SIGNATURE;
     hdr->sequence_number = g_random_int();
@@ -1540,7 +1552,8 @@ exit:
  */
 static int vhdx_create_bat(BlockDriverState *bs, BDRVVHDXState *s,
                            uint64_t image_size, VHDXImageType type,
-                           bool use_zero_blocks, VHDXRegionTableEntry *rt_bat)
+                           bool use_zero_blocks, uint64_t file_offset,
+                           uint32_t length)
 {
     int ret = 0;
     uint64_t data_file_offset;
@@ -1555,7 +1568,7 @@ static int vhdx_create_bat(BlockDriverState *bs, BDRVVHDXState *s,
     /* this gives a data start after BAT/bitmap entries, and well
      * past any metadata entries (with a 4 MB buffer for future
      * expansion */
-    data_file_offset = rt_bat->file_offset + rt_bat->length + 5 * MiB;
+    data_file_offset = file_offset + length + 5 * MiB;
     total_sectors = image_size >> s->logical_sector_size_bits;
 
     if (type == VHDX_TYPE_DYNAMIC) {
@@ -1579,7 +1592,11 @@ static int vhdx_create_bat(BlockDriverState *bs, BDRVVHDXState *s,
                 use_zero_blocks ||
                 bdrv_has_zero_init(bs) == 0) {
         /* for a fixed file, the default BAT entry is not zero */
-        s->bat = g_malloc0(rt_bat->length);
+        s->bat = g_try_malloc0(length);
+        if (length && s->bat != NULL) {
+            ret = -ENOMEM;
+            goto exit;
+        }
         block_state = type == VHDX_TYPE_FIXED ? PAYLOAD_BLOCK_FULLY_PRESENT :
                                                 PAYLOAD_BLOCK_NOT_PRESENT;
         block_state = use_zero_blocks ? PAYLOAD_BLOCK_ZERO : block_state;
@@ -1594,7 +1611,7 @@ static int vhdx_create_bat(BlockDriverState *bs, BDRVVHDXState *s,
             cpu_to_le64s(&s->bat[sinfo.bat_idx]);
             sector_num += s->sectors_per_block;
         }
-        ret = bdrv_pwrite(bs, rt_bat->file_offset, s->bat, rt_bat->length);
+        ret = bdrv_pwrite(bs, file_offset, s->bat, length);
         if (ret < 0) {
             goto exit;
         }
@@ -1626,6 +1643,8 @@ static int vhdx_create_new_region_table(BlockDriverState *bs,
     int ret = 0;
     uint32_t offset = 0;
     void *buffer = NULL;
+    uint64_t bat_file_offset;
+    uint32_t bat_length;
     BDRVVHDXState *s = NULL;
     VHDXRegionTableHeader *region_table;
     VHDXRegionTableEntry *rt_bat;
@@ -1635,7 +1654,7 @@ static int vhdx_create_new_region_table(BlockDriverState *bs,
 
     /* Populate enough of the BDRVVHDXState to be able to use the
      * pre-existing BAT calculation, translation, and update functions */
-    s = g_malloc0(sizeof(BDRVVHDXState));
+    s = g_new0(BDRVVHDXState, 1);
 
     s->chunk_ratio = (VHDX_MAX_SECTORS_PER_BLOCK) *
                      (uint64_t) sector_size / (uint64_t) block_size;
@@ -1674,19 +1693,26 @@ static int vhdx_create_new_region_table(BlockDriverState *bs,
     rt_metadata->length      = 1 * MiB; /* min size, and more than enough */
     *metadata_offset = rt_metadata->file_offset;
 
+    bat_file_offset = rt_bat->file_offset;
+    bat_length = rt_bat->length;
+
+    vhdx_region_header_le_export(region_table);
+    vhdx_region_entry_le_export(rt_bat);
+    vhdx_region_entry_le_export(rt_metadata);
+
     vhdx_update_checksum(buffer, VHDX_HEADER_BLOCK_SIZE,
                          offsetof(VHDXRegionTableHeader, checksum));
 
 
     /* The region table gives us the data we need to create the BAT,
      * so do that now */
-    ret = vhdx_create_bat(bs, s, image_size, type, use_zero_blocks, rt_bat);
+    ret = vhdx_create_bat(bs, s, image_size, type, use_zero_blocks,
+                          bat_file_offset, bat_length);
+    if (ret < 0) {
+        goto exit;
+    }
 
     /* Now write out the region headers to disk */
-    vhdx_region_header_le_export(region_table);
-    vhdx_region_entry_le_export(rt_bat);
-    vhdx_region_entry_le_export(rt_metadata);
-
     ret = bdrv_pwrite(bs, VHDX_REGION_TABLE_OFFSET, buffer,
                       VHDX_HEADER_BLOCK_SIZE);
     if (ret < 0) {

block/vhdx.h

@@ -435,6 +435,7 @@ void vhdx_header_le_import(VHDXHeader *h);
 void vhdx_header_le_export(VHDXHeader *orig_h, VHDXHeader *new_h);
 void vhdx_log_desc_le_import(VHDXLogDescriptor *d);
 void vhdx_log_desc_le_export(VHDXLogDescriptor *d);
+void vhdx_log_data_le_import(VHDXLogDataSector *d);
 void vhdx_log_data_le_export(VHDXLogDataSector *d);
 void vhdx_log_entry_hdr_le_import(VHDXLogEntryHeader *hdr);
 void vhdx_log_entry_hdr_le_export(VHDXLogEntryHeader *hdr);

block/vmdk.c

@@ -106,6 +106,7 @@ typedef struct VmdkExtent {
     uint32_t l2_cache_counts[L2_CACHE_SIZE];
 
     int64_t cluster_sectors;
+    int64_t next_cluster_sector;
     char *type;
 } VmdkExtent;
 
@@ -124,7 +125,6 @@ typedef struct BDRVVmdkState {
 } BDRVVmdkState;
 
 typedef struct VmdkMetaData {
-    uint32_t offset;
     unsigned int l1_index;
     unsigned int l2_index;
     unsigned int l2_offset;
@@ -233,7 +233,7 @@ static void vmdk_free_last_extent(BlockDriverState *bs)
         return;
     }
     s->num_extents--;
-    s->extents = g_realloc(s->extents, s->num_extents * sizeof(VmdkExtent));
+    s->extents = g_renew(VmdkExtent, s->extents, s->num_extents);
 }
 
 static uint32_t vmdk_read_cid(BlockDriverState *bs, int parent)
@@ -397,6 +397,7 @@ static int vmdk_add_extent(BlockDriverState *bs,
 {
     VmdkExtent *extent;
     BDRVVmdkState *s = bs->opaque;
+    int64_t nb_sectors;
 
     if (cluster_sectors > 0x200000) {
         /* 0x200000 * 512Bytes = 1GB for one cluster is unrealistic */
@@ -412,8 +413,12 @@ static int vmdk_add_extent(BlockDriverState *bs,
         return -EFBIG;
     }
 
-    s->extents = g_realloc(s->extents,
-                              (s->num_extents + 1) * sizeof(VmdkExtent));
+    nb_sectors = bdrv_nb_sectors(file);
+    if (nb_sectors < 0) {
+        return nb_sectors;
+    }
+
+    s->extents = g_renew(VmdkExtent, s->extents, s->num_extents + 1);
     extent = &s->extents[s->num_extents];
     s->num_extents++;
 
@@ -427,6 +432,7 @@ static int vmdk_add_extent(BlockDriverState *bs,
     extent->l1_entry_sectors = l2_size * cluster_sectors;
     extent->l2_size = l2_size;
     extent->cluster_sectors = flat ? sectors : cluster_sectors;
+    extent->next_cluster_sector = ROUND_UP(nb_sectors, cluster_sectors);
 
     if (s->num_extents > 1) {
         extent->end_sector = (*(extent - 1)).end_sector + extent->sectors;
@@ -448,7 +454,11 @@ static int vmdk_init_tables(BlockDriverState *bs, VmdkExtent *extent,
 
     /* read the L1 table */
     l1_size = extent->l1_size * sizeof(uint32_t);
-    extent->l1_table = g_malloc(l1_size);
+    extent->l1_table = g_try_malloc(l1_size);
+    if (l1_size && extent->l1_table == NULL) {
+        return -ENOMEM;
+    }
+
     ret = bdrv_pread(extent->file,
                      extent->l1_table_offset,
                      extent->l1_table,
@@ -464,7 +474,11 @@ static int vmdk_init_tables(BlockDriverState *bs, VmdkExtent *extent,
     }
 
     if (extent->l1_backup_table_offset) {
-        extent->l1_backup_table = g_malloc(l1_size);
+        extent->l1_backup_table = g_try_malloc(l1_size);
+        if (l1_size && extent->l1_backup_table == NULL) {
+            ret = -ENOMEM;
+            goto fail_l1;
+        }
         ret = bdrv_pread(extent->file,
                          extent->l1_backup_table_offset,
                          extent->l1_backup_table,
@@ -481,7 +495,7 @@ static int vmdk_init_tables(BlockDriverState *bs, VmdkExtent *extent,
     }
 
     extent->l2_cache =
-        g_malloc(extent->l2_size * L2_CACHE_SIZE * sizeof(uint32_t));
+        g_new(uint32_t, extent->l2_size * L2_CACHE_SIZE);
     return 0;
  fail_l1b:
     g_free(extent->l1_backup_table);
@@ -669,8 +683,7 @@ static int vmdk_open_vmdk4(BlockDriverState *bs,
     if (le32_to_cpu(header.flags) & VMDK4_FLAG_RGD) {
         l1_backup_offset = le64_to_cpu(header.rgd_offset) << 9;
     }
-    if (bdrv_getlength(file) <
-            le64_to_cpu(header.grain_offset) * BDRV_SECTOR_SIZE) {
+    if (bdrv_nb_sectors(file) < le64_to_cpu(header.grain_offset)) {
         error_setg(errp, "File truncated, expecting at least %" PRId64 " bytes",
                    (int64_t)(le64_to_cpu(header.grain_offset)
                              * BDRV_SECTOR_SIZE));
@@ -938,7 +951,7 @@ fail:
 }
 
 
-static int vmdk_refresh_limits(BlockDriverState *bs)
+static void vmdk_refresh_limits(BlockDriverState *bs, Error **errp)
 {
     BDRVVmdkState *s = bs->opaque;
     int i;
@@ -950,61 +963,99 @@ static int vmdk_refresh_limits(BlockDriverState *bs)
                     s->extents[i].cluster_sectors);
         }
     }
-
-    return 0;
 }
 
+/**
+ * get_whole_cluster
+ *
+ * Copy backing file's cluster that covers @sector_num, otherwise write zero,
+ * to the cluster at @cluster_sector_num.
+ *
+ * If @skip_start_sector < @skip_end_sector, the relative range
+ * [@skip_start_sector, @skip_end_sector) is not copied or written, and leave
+ * it for call to write user data in the request.
+ */
 static int get_whole_cluster(BlockDriverState *bs,
-                VmdkExtent *extent,
-                uint64_t cluster_offset,
-                uint64_t offset,
-                bool allocate)
+                             VmdkExtent *extent,
+                             uint64_t cluster_sector_num,
+                             uint64_t sector_num,
+                             uint64_t skip_start_sector,
+                             uint64_t skip_end_sector)
 {
     int ret = VMDK_OK;
-    uint8_t *whole_grain = NULL;
+    int64_t cluster_bytes;
+    uint8_t *whole_grain;
 
+    /* For COW, align request sector_num to cluster start */
+    sector_num = QEMU_ALIGN_DOWN(sector_num, extent->cluster_sectors);
+    cluster_bytes = extent->cluster_sectors << BDRV_SECTOR_BITS;
+    whole_grain = qemu_blockalign(bs, cluster_bytes);
+
+    if (!bs->backing_hd) {
+        memset(whole_grain, 0,  skip_start_sector << BDRV_SECTOR_BITS);
+        memset(whole_grain + (skip_end_sector << BDRV_SECTOR_BITS), 0,
+               cluster_bytes - (skip_end_sector << BDRV_SECTOR_BITS));
+    }
+
+    assert(skip_end_sector <= extent->cluster_sectors);
     /* we will be here if it's first write on non-exist grain(cluster).
      * try to read from parent image, if exist */
-    if (bs->backing_hd) {
-        whole_grain =
-            qemu_blockalign(bs, extent->cluster_sectors << BDRV_SECTOR_BITS);
-        if (!vmdk_is_cid_valid(bs)) {
-            ret = VMDK_ERROR;
-            goto exit;
-        }
+    if (bs->backing_hd && !vmdk_is_cid_valid(bs)) {
+        ret = VMDK_ERROR;
+        goto exit;
+    }
 
-        /* floor offset to cluster */
-        offset -= offset % (extent->cluster_sectors * 512);
-        ret = bdrv_read(bs->backing_hd, offset >> 9, whole_grain,
-                extent->cluster_sectors);
-        if (ret < 0) {
-            ret = VMDK_ERROR;
-            goto exit;
+    /* Read backing data before skip range */
+    if (skip_start_sector > 0) {
+        if (bs->backing_hd) {
+            ret = bdrv_read(bs->backing_hd, sector_num,
+                            whole_grain, skip_start_sector);
+            if (ret < 0) {
+                ret = VMDK_ERROR;
+                goto exit;
+            }
         }
-
-        /* Write grain only into the active image */
-        ret = bdrv_write(extent->file, cluster_offset, whole_grain,
-                extent->cluster_sectors);
+        ret = bdrv_write(extent->file, cluster_sector_num, whole_grain,
+                         skip_start_sector);
         if (ret < 0) {
             ret = VMDK_ERROR;
             goto exit;
         }
     }
+    /* Read backing data after skip range */
+    if (skip_end_sector < extent->cluster_sectors) {
+        if (bs->backing_hd) {
+            ret = bdrv_read(bs->backing_hd, sector_num + skip_end_sector,
+                            whole_grain + (skip_end_sector << BDRV_SECTOR_BITS),
+                            extent->cluster_sectors - skip_end_sector);
+            if (ret < 0) {
+                ret = VMDK_ERROR;
+                goto exit;
+            }
+        }
+        ret = bdrv_write(extent->file, cluster_sector_num + skip_end_sector,
+                         whole_grain + (skip_end_sector << BDRV_SECTOR_BITS),
+                         extent->cluster_sectors - skip_end_sector);
+        if (ret < 0) {
+            ret = VMDK_ERROR;
+            goto exit;
+        }
+    }
+
 exit:
     qemu_vfree(whole_grain);
     return ret;
 }
 
-static int vmdk_L2update(VmdkExtent *extent, VmdkMetaData *m_data)
+static int vmdk_L2update(VmdkExtent *extent, VmdkMetaData *m_data,
+                         uint32_t offset)
 {
-    uint32_t offset;
-    QEMU_BUILD_BUG_ON(sizeof(offset) != sizeof(m_data->offset));
-    offset = cpu_to_le32(m_data->offset);
+    offset = cpu_to_le32(offset);
     /* update L2 table */
     if (bdrv_pwrite_sync(
                 extent->file,
                 ((int64_t)m_data->l2_offset * 512)
-                    + (m_data->l2_index * sizeof(m_data->offset)),
+                    + (m_data->l2_index * sizeof(offset)),
                 &offset, sizeof(offset)) < 0) {
         return VMDK_ERROR;
     }
@@ -1014,7 +1065,7 @@ static int vmdk_L2update(VmdkExtent *extent, VmdkMetaData *m_data)
         if (bdrv_pwrite_sync(
                     extent->file,
                     ((int64_t)m_data->l2_offset * 512)
-                        + (m_data->l2_index * sizeof(m_data->offset)),
+                        + (m_data->l2_index * sizeof(offset)),
                     &offset, sizeof(offset)) < 0) {
             return VMDK_ERROR;
         }
@@ -1026,17 +1077,41 @@ static int vmdk_L2update(VmdkExtent *extent, VmdkMetaData *m_data)
     return VMDK_OK;
 }
 
+/**
+ * get_cluster_offset
+ *
+ * Look up cluster offset in extent file by sector number, and store in
+ * @cluster_offset.
+ *
+ * For flat extents, the start offset as parsed from the description file is
+ * returned.
+ *
+ * For sparse extents, look up in L1, L2 table. If allocate is true, return an
+ * offset for a new cluster and update L2 cache. If there is a backing file,
+ * COW is done before returning; otherwise, zeroes are written to the allocated
+ * cluster. Both COW and zero writing skips the sector range
+ * [@skip_start_sector, @skip_end_sector) passed in by caller, because caller
+ * has new data to write there.
+ *
+ * Returns: VMDK_OK if cluster exists and mapped in the image.
+ *          VMDK_UNALLOC if cluster is not mapped and @allocate is false.
+ *          VMDK_ERROR if failed.
+ */
 static int get_cluster_offset(BlockDriverState *bs,
-                                    VmdkExtent *extent,
-                                    VmdkMetaData *m_data,
-                                    uint64_t offset,
-                                    int allocate,
-                                    uint64_t *cluster_offset)
+                              VmdkExtent *extent,
+                              VmdkMetaData *m_data,
+                              uint64_t offset,
+                              bool allocate,
+                              uint64_t *cluster_offset,
+                              uint64_t skip_start_sector,
+                              uint64_t skip_end_sector)
 {
     unsigned int l1_index, l2_offset, l2_index;
     int min_index, i, j;
     uint32_t min_count, *l2_table;
     bool zeroed = false;
+    int64_t ret;
+    int32_t cluster_sector;
 
     if (m_data) {
         m_data->valid = 0;
@@ -1090,52 +1165,41 @@ static int get_cluster_offset(BlockDriverState *bs,
     extent->l2_cache_counts[min_index] = 1;
  found:
     l2_index = ((offset >> 9) / extent->cluster_sectors) % extent->l2_size;
-    *cluster_offset = le32_to_cpu(l2_table[l2_index]);
+    cluster_sector = le32_to_cpu(l2_table[l2_index]);
 
     if (m_data) {
         m_data->valid = 1;
         m_data->l1_index = l1_index;
         m_data->l2_index = l2_index;
-        m_data->offset = *cluster_offset;
         m_data->l2_offset = l2_offset;
         m_data->l2_cache_entry = &l2_table[l2_index];
     }
-    if (extent->has_zero_grain && *cluster_offset == VMDK_GTE_ZEROED) {
+    if (extent->has_zero_grain && cluster_sector == VMDK_GTE_ZEROED) {
         zeroed = true;
     }
 
-    if (!*cluster_offset || zeroed) {
+    if (!cluster_sector || zeroed) {
         if (!allocate) {
             return zeroed ? VMDK_ZEROED : VMDK_UNALLOC;
         }
 
-        /* Avoid the L2 tables update for the images that have snapshots. */
-        *cluster_offset = bdrv_getlength(extent->file);
-        if (!extent->compressed) {
-            bdrv_truncate(
-                extent->file,
-                *cluster_offset + (extent->cluster_sectors << 9)
-            );
-        }
-
-        *cluster_offset >>= 9;
-        l2_table[l2_index] = cpu_to_le32(*cluster_offset);
+        cluster_sector = extent->next_cluster_sector;
+        extent->next_cluster_sector += extent->cluster_sectors;
 
         /* First of all we write grain itself, to avoid race condition
          * that may to corrupt the image.
          * This problem may occur because of insufficient space on host disk
          * or inappropriate VM shutdown.
          */
-        if (get_whole_cluster(
-                bs, extent, *cluster_offset, offset, allocate) == -1) {
-            return VMDK_ERROR;
-        }
-
-        if (m_data) {
-            m_data->offset = *cluster_offset;
+        ret = get_whole_cluster(bs, extent,
+                                cluster_sector,
+                                offset >> BDRV_SECTOR_BITS,
+                                skip_start_sector, skip_end_sector);
+        if (ret) {
+            return ret;
         }
     }
-    *cluster_offset <<= 9;
+    *cluster_offset = cluster_sector << BDRV_SECTOR_BITS;
     return VMDK_OK;
 }
 
@@ -1170,7 +1234,8 @@ static int64_t coroutine_fn vmdk_co_get_block_status(BlockDriverState *bs,
     }
     qemu_co_mutex_lock(&s->lock);
     ret = get_cluster_offset(bs, extent, NULL,
-                            sector_num * 512, 0, &offset);
+                             sector_num * 512, false, &offset,
+                             0, 0);
     qemu_co_mutex_unlock(&s->lock);
 
     switch (ret) {
@@ -1323,9 +1388,9 @@ static int vmdk_read(BlockDriverState *bs, int64_t sector_num,
         if (!extent) {
             return -EIO;
         }
-        ret = get_cluster_offset(
-                            bs, extent, NULL,
-                            sector_num << 9, 0, &cluster_offset);
+        ret = get_cluster_offset(bs, extent, NULL,
+                                 sector_num << 9, false, &cluster_offset,
+                                 0, 0);
         extent_begin_sector = extent->end_sector - extent->sectors;
         extent_relative_sector_num = sector_num - extent_begin_sector;
         index_in_cluster = extent_relative_sector_num % extent->cluster_sectors;
@@ -1406,12 +1471,17 @@ static int vmdk_write(BlockDriverState *bs, int64_t sector_num,
         if (!extent) {
             return -EIO;
         }
-        ret = get_cluster_offset(
-                                bs,
-                                extent,
-                                &m_data,
-                                sector_num << 9, !extent->compressed,
-                                &cluster_offset);
+        extent_begin_sector = extent->end_sector - extent->sectors;
+        extent_relative_sector_num = sector_num - extent_begin_sector;
+        index_in_cluster = extent_relative_sector_num % extent->cluster_sectors;
+        n = extent->cluster_sectors - index_in_cluster;
+        if (n > nb_sectors) {
+            n = nb_sectors;
+        }
+        ret = get_cluster_offset(bs, extent, &m_data, sector_num << 9,
+                                 !(extent->compressed || zeroed),
+                                 &cluster_offset,
+                                 index_in_cluster, index_in_cluster + n);
         if (extent->compressed) {
             if (ret == VMDK_OK) {
                 /* Refuse write to allocated cluster for streamOptimized */
@@ -1420,24 +1490,13 @@ static int vmdk_write(BlockDriverState *bs, int64_t sector_num,
                 return -EIO;
             } else {
                 /* allocate */
-                ret = get_cluster_offset(
-                                        bs,
-                                        extent,
-                                        &m_data,
-                                        sector_num << 9, 1,
-                                        &cluster_offset);
+                ret = get_cluster_offset(bs, extent, &m_data, sector_num << 9,
+                                         true, &cluster_offset, 0, 0);
             }
         }
         if (ret == VMDK_ERROR) {
             return -EINVAL;
         }
-        extent_begin_sector = extent->end_sector - extent->sectors;
-        extent_relative_sector_num = sector_num - extent_begin_sector;
-        index_in_cluster = extent_relative_sector_num % extent->cluster_sectors;
-        n = extent->cluster_sectors - index_in_cluster;
-        if (n > nb_sectors) {
-            n = nb_sectors;
-        }
         if (zeroed) {
             /* Do zeroed write, buf is ignored */
             if (extent->has_zero_grain &&
@@ -1445,9 +1504,9 @@ static int vmdk_write(BlockDriverState *bs, int64_t sector_num,
                     n >= extent->cluster_sectors) {
                 n = extent->cluster_sectors;
                 if (!zero_dry_run) {
-                    m_data.offset = VMDK_GTE_ZEROED;
                     /* update L2 tables */
-                    if (vmdk_L2update(extent, &m_data) != VMDK_OK) {
+                    if (vmdk_L2update(extent, &m_data, VMDK_GTE_ZEROED)
+                            != VMDK_OK) {
                         return -EIO;
                     }
                 }
@@ -1463,7 +1522,9 @@ static int vmdk_write(BlockDriverState *bs, int64_t sector_num,
             }
             if (m_data.valid) {
                 /* update L2 tables */
-                if (vmdk_L2update(extent, &m_data) != VMDK_OK) {
+                if (vmdk_L2update(extent, &m_data,
+                                  cluster_offset >> BDRV_SECTOR_BITS)
+                        != VMDK_OK) {
                     return -EIO;
                 }
             }
@@ -1529,7 +1590,7 @@ static int coroutine_fn vmdk_co_write_zeroes(BlockDriverState *bs,
 
 static int vmdk_create_extent(const char *filename, int64_t filesize,
                               bool flat, bool compress, bool zeroed_grain,
-                              Error **errp)
+                              QemuOpts *opts, Error **errp)
 {
     int ret, i;
     BlockDriverState *bs = NULL;
@@ -1539,7 +1600,7 @@ static int vmdk_create_extent(const char *filename, int64_t filesize,
     uint32_t *gd_buf = NULL;
     int gd_buf_size;
 
-    ret = bdrv_create_file(filename, NULL, &local_err);
+    ret = bdrv_create_file(filename, opts, &local_err);
     if (ret < 0) {
         error_propagate(errp, local_err);
         goto exit;
@@ -1845,7 +1906,7 @@ static int vmdk_create(const char *filename, QemuOpts *opts, Error **errp)
                 path, desc_filename);
 
         if (vmdk_create_extent(ext_filename, size,
-                               flat, compress, zeroed_grain, errp)) {
+                               flat, compress, zeroed_grain, opts, errp)) {
             ret = -EINVAL;
             goto exit;
         }
@@ -2001,7 +2062,7 @@ static int vmdk_check(BlockDriverState *bs, BdrvCheckResult *result,
     BDRVVmdkState *s = bs->opaque;
     VmdkExtent *extent = NULL;
     int64_t sector_num = 0;
-    int64_t total_sectors = bdrv_getlength(bs) / BDRV_SECTOR_SIZE;
+    int64_t total_sectors = bdrv_nb_sectors(bs);
     int ret;
     uint64_t cluster_offset;
 
@@ -2022,7 +2083,7 @@ static int vmdk_check(BlockDriverState *bs, BdrvCheckResult *result,
         }
         ret = get_cluster_offset(bs, extent, NULL,
                                  sector_num << BDRV_SECTOR_BITS,
-                                 0, &cluster_offset);
+                                 false, &cluster_offset, 0, 0);
         if (ret == VMDK_ERROR) {
             fprintf(stderr,
                     "ERROR: could not get cluster_offset for sector %"
@@ -2180,6 +2241,7 @@ static BlockDriver bdrv_vmdk = {
     .bdrv_detach_aio_context      = vmdk_detach_aio_context,
     .bdrv_attach_aio_context      = vmdk_attach_aio_context,
 
+    .supports_backing             = true,
     .create_opts                  = &vmdk_create_opts,
 };
 

block/vpc.c

@@ -269,7 +269,11 @@ static int vpc_open(BlockDriverState *bs, QDict *options, int flags,
             goto fail;
         }
 
-        s->pagetable = qemu_blockalign(bs, s->max_table_entries * 4);
+        s->pagetable = qemu_try_blockalign(bs->file, s->max_table_entries * 4);
+        if (s->pagetable == NULL) {
+            ret = -ENOMEM;
+            goto fail;
+        }
 
         s->bat_offset = be64_to_cpu(dyndisk_header->table_offset);
 
@@ -649,39 +653,41 @@ static int calculate_geometry(int64_t total_sectors, uint16_t* cyls,
     return 0;
 }
 
-static int create_dynamic_disk(int fd, uint8_t *buf, int64_t total_sectors)
+static int create_dynamic_disk(BlockDriverState *bs, uint8_t *buf,
+                               int64_t total_sectors)
 {
     VHDDynDiskHeader *dyndisk_header =
         (VHDDynDiskHeader *) buf;
     size_t block_size, num_bat_entries;
     int i;
-    int ret = -EIO;
+    int ret;
+    int64_t offset = 0;
 
     // Write the footer (twice: at the beginning and at the end)
     block_size = 0x200000;
     num_bat_entries = (total_sectors + block_size / 512) / (block_size / 512);
 
-    if (write(fd, buf, HEADER_SIZE) != HEADER_SIZE) {
+    ret = bdrv_pwrite_sync(bs, offset, buf, HEADER_SIZE);
+    if (ret) {
         goto fail;
     }
 
-    if (lseek(fd, 1536 + ((num_bat_entries * 4 + 511) & ~511), SEEK_SET) < 0) {
-        goto fail;
-    }
-    if (write(fd, buf, HEADER_SIZE) != HEADER_SIZE) {
+    offset = 1536 + ((num_bat_entries * 4 + 511) & ~511);
+    ret = bdrv_pwrite_sync(bs, offset, buf, HEADER_SIZE);
+    if (ret < 0) {
         goto fail;
     }
 
     // Write the initial BAT
-    if (lseek(fd, 3 * 512, SEEK_SET) < 0) {
-        goto fail;
-    }
+    offset = 3 * 512;
 
     memset(buf, 0xFF, 512);
     for (i = 0; i < (num_bat_entries * 4 + 511) / 512; i++) {
-        if (write(fd, buf, 512) != 512) {
+        ret = bdrv_pwrite_sync(bs, offset, buf, 512);
+        if (ret < 0) {
             goto fail;
         }
+        offset += 512;
     }
 
     // Prepare the Dynamic Disk Header
@@ -702,39 +708,35 @@ static int create_dynamic_disk(int fd, uint8_t *buf, int64_t total_sectors)
     dyndisk_header->checksum = be32_to_cpu(vpc_checksum(buf, 1024));
 
     // Write the header
-    if (lseek(fd, 512, SEEK_SET) < 0) {
-        goto fail;
-    }
+    offset = 512;
 
-    if (write(fd, buf, 1024) != 1024) {
+    ret = bdrv_pwrite_sync(bs, offset, buf, 1024);
+    if (ret < 0) {
         goto fail;
     }
-    ret = 0;
 
  fail:
     return ret;
 }
 
-static int create_fixed_disk(int fd, uint8_t *buf, int64_t total_size)
+static int create_fixed_disk(BlockDriverState *bs, uint8_t *buf,
+                             int64_t total_size)
 {
-    int ret = -EIO;
+    int ret;
 
     /* Add footer to total size */
-    total_size += 512;
-    if (ftruncate(fd, total_size) != 0) {
-        ret = -errno;
-        goto fail;
-    }
-    if (lseek(fd, -512, SEEK_END) < 0) {
-        goto fail;
-    }
-    if (write(fd, buf, HEADER_SIZE) != HEADER_SIZE) {
-        goto fail;
+    total_size += HEADER_SIZE;
+
+    ret = bdrv_truncate(bs, total_size);
+    if (ret < 0) {
+        return ret;
     }
 
-    ret = 0;
+    ret = bdrv_pwrite_sync(bs, total_size - HEADER_SIZE, buf, HEADER_SIZE);
+    if (ret < 0) {
+        return ret;
+    }
 
- fail:
     return ret;
 }
 
@@ -743,7 +745,7 @@ static int vpc_create(const char *filename, QemuOpts *opts, Error **errp)
     uint8_t buf[1024];
     VHDFooter *footer = (VHDFooter *) buf;
     char *disk_type_param;
-    int fd, i;
+    int i;
     uint16_t cyls = 0;
     uint8_t heads = 0;
     uint8_t secs_per_cyl = 0;
@@ -751,6 +753,8 @@ static int vpc_create(const char *filename, QemuOpts *opts, Error **errp)
     int64_t total_size;
     int disk_type;
     int ret = -EIO;
+    Error *local_err = NULL;
+    BlockDriverState *bs = NULL;
 
     /* Read out options */
     total_size = qemu_opt_get_size_del(opts, BLOCK_OPT_SIZE, 0);
@@ -768,10 +772,15 @@ static int vpc_create(const char *filename, QemuOpts *opts, Error **errp)
         disk_type = VHD_DYNAMIC;
     }
 
-    /* Create the file */
-    fd = qemu_open(filename, O_WRONLY | O_CREAT | O_TRUNC | O_BINARY, 0644);
-    if (fd < 0) {
-        ret = -EIO;
+    ret = bdrv_create_file(filename, opts, &local_err);
+    if (ret < 0) {
+        error_propagate(errp, local_err);
+        goto out;
+    }
+    ret = bdrv_open(&bs, filename, NULL, NULL, BDRV_O_RDWR | BDRV_O_PROTOCOL,
+                    NULL, &local_err);
+    if (ret < 0) {
+        error_propagate(errp, local_err);
         goto out;
     }
 
@@ -786,7 +795,7 @@ static int vpc_create(const char *filename, QemuOpts *opts, Error **errp)
                                &secs_per_cyl))
         {
             ret = -EFBIG;
-            goto fail;
+            goto out;
         }
     }
 
@@ -832,14 +841,13 @@ static int vpc_create(const char *filename, QemuOpts *opts, Error **errp)
     footer->checksum = be32_to_cpu(vpc_checksum(buf, HEADER_SIZE));
 
     if (disk_type == VHD_DYNAMIC) {
-        ret = create_dynamic_disk(fd, buf, total_sectors);
+        ret = create_dynamic_disk(bs, buf, total_sectors);
     } else {
-        ret = create_fixed_disk(fd, buf, total_size);
+        ret = create_fixed_disk(bs, buf, total_size);
     }
 
-fail:
-    qemu_close(fd);
 out:
+    bdrv_unref(bs);
     g_free(disk_type_param);
     return ret;
 }
@@ -884,6 +892,11 @@ static QemuOptsList vpc_create_opts = {
                 "Type of virtual hard disk format. Supported formats are "
                 "{dynamic (default) | fixed} "
         },
+        {
+            .name = BLOCK_OPT_NOCOW,
+            .type = QEMU_OPT_BOOL,
+            .help = "Turn off copy-on-write (valid only on btrfs)"
+        },
         { /* end of list */ }
     }
 };

block/vvfat.c

@@ -52,10 +52,6 @@
 
 #define DLOG(a) a
 
-#undef stderr
-#define stderr STDERR
-FILE* stderr = NULL;
-
 static void checkpoint(void);
 
 #ifdef __MINGW32__
@@ -732,7 +728,7 @@ static int read_directory(BDRVVVFATState* s, int mapping_index)
 	if(first_cluster == 0 && (is_dotdot || is_dot))
 	    continue;
 
-	buffer=(char*)g_malloc(length);
+	buffer = g_malloc(length);
 	snprintf(buffer,length,"%s/%s",dirname,entry->d_name);
 
 	if(stat(buffer,&st)<0) {
@@ -767,7 +763,7 @@ static int read_directory(BDRVVVFATState* s, int mapping_index)
 
 	/* create mapping for this file */
 	if(!is_dot && !is_dotdot && (S_ISDIR(st.st_mode) || st.st_size)) {
-	    s->current_mapping=(mapping_t*)array_get_next(&(s->mapping));
+	    s->current_mapping = array_get_next(&(s->mapping));
 	    s->current_mapping->begin=0;
 	    s->current_mapping->end=st.st_size;
 	    /*
@@ -811,12 +807,12 @@ static int read_directory(BDRVVVFATState* s, int mapping_index)
     }
 
      /* reget the mapping, since s->mapping was possibly realloc()ed */
-    mapping = (mapping_t*)array_get(&(s->mapping), mapping_index);
+    mapping = array_get(&(s->mapping), mapping_index);
     first_cluster += (s->directory.next - mapping->info.dir.first_dir_index)
 	* 0x20 / s->cluster_size;
     mapping->end = first_cluster;
 
-    direntry = (direntry_t*)array_get(&(s->directory), mapping->dir_index);
+    direntry = array_get(&(s->directory), mapping->dir_index);
     set_begin_of_direntry(direntry, mapping->begin);
 
     return 0;
@@ -1082,11 +1078,6 @@ static int vvfat_open(BlockDriverState *bs, QDict *options, int flags,
     vvv = s;
 #endif
 
-DLOG(if (stderr == NULL) {
-    stderr = fopen("vvfat.log", "a");
-    setbuf(stderr, NULL);
-})
-
     opts = qemu_opts_create(&runtime_opts, NULL, 0, &error_abort);
     qemu_opts_absorb_qdict(opts, options, &local_err);
     if (local_err) {
@@ -2950,7 +2941,7 @@ static int enable_write_target(BDRVVVFATState *s, Error **errp)
 
     bdrv_set_backing_hd(s->bs, bdrv_new("", &error_abort));
     s->bs->backing_hd->drv = &vvfat_write_target;
-    s->bs->backing_hd->opaque = g_malloc(sizeof(void*));
+    s->bs->backing_hd->opaque = g_new(void *, 1);
     *(void**)s->bs->backing_hd->opaque = s;
 
     return 0;

block/win32-aio.c

@@ -139,7 +139,10 @@ BlockDriverAIOCB *win32_aio_submit(BlockDriverState *bs,
     waiocb->is_read = (type == QEMU_AIO_READ);
 
     if (qiov->niov > 1) {
-        waiocb->buf = qemu_blockalign(bs, qiov->size);
+        waiocb->buf = qemu_try_blockalign(bs, qiov->size);
+        if (waiocb->buf == NULL) {
+            goto out;
+        }
         if (type & QEMU_AIO_WRITE) {
             iov_to_buf(qiov->iov, qiov->niov, 0, waiocb->buf, qiov->size);
         }
@@ -168,6 +171,7 @@ BlockDriverAIOCB *win32_aio_submit(BlockDriverState *bs,
 
 out_dec_count:
     aio->count--;
+out:
     qemu_aio_release(waiocb);
     return NULL;
 }

blockdev-nbd.c

@@ -28,6 +28,7 @@ static void nbd_accept(void *opaque)
 
     int fd = accept(server_fd, (struct sockaddr *)&addr, &addr_len);
     if (fd >= 0 && !nbd_client_new(NULL, fd, nbd_client_put)) {
+        shutdown(fd, 2);
         close(fd);
     }
 }
@@ -91,6 +92,10 @@ void qmp_nbd_server_add(const char *device, bool has_writable, bool writable,
         error_set(errp, QERR_DEVICE_NOT_FOUND, device);
         return;
     }
+    if (!bdrv_is_inserted(bs)) {
+        error_set(errp, QERR_DEVICE_HAS_NO_MEDIUM, device);
+        return;
+    }
 
     if (!has_writable) {
         writable = false;
@@ -103,7 +108,7 @@ void qmp_nbd_server_add(const char *device, bool has_writable, bool writable,
 
     nbd_export_set_name(exp, device);
 
-    n = g_malloc0(sizeof(NBDCloseNotifier));
+    n = g_new0(NBDCloseNotifier, 1);
     n->n.notify = nbd_close_notifier;
     n->exp = exp;
     bdrv_add_close_notifier(bs, &n->n);

blockdev.c

@@ -1094,7 +1094,7 @@ SnapshotInfo *qmp_blockdev_snapshot_delete_internal_sync(const char *device,
         return NULL;
     }
 
-    info = g_malloc0(sizeof(SnapshotInfo));
+    info = g_new0(SnapshotInfo, 1);
     info->id = g_strdup(sn.id_str);
     info->name = g_strdup(sn.name);
     info->date_nsec = sn.date_nsec;
@@ -1757,6 +1757,7 @@ int do_drive_del(Monitor *mon, const QDict *qdict, QObject **ret_data)
 {
     const char *id = qdict_get_str(qdict, "id");
     BlockDriverState *bs;
+    AioContext *aio_context;
     Error *local_err = NULL;
 
     bs = bdrv_find(id);
@@ -1764,9 +1765,14 @@ int do_drive_del(Monitor *mon, const QDict *qdict, QObject **ret_data)
         error_report("Device '%s' not found", id);
         return -1;
     }
+
+    aio_context = bdrv_get_aio_context(bs);
+    aio_context_acquire(aio_context);
+
     if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_DRIVE_DEL, &local_err)) {
         error_report("%s", error_get_pretty(local_err));
         error_free(local_err);
+        aio_context_release(aio_context);
         return -1;
     }
 
@@ -1790,6 +1796,7 @@ int do_drive_del(Monitor *mon, const QDict *qdict, QObject **ret_data)
         drive_del(drive_get_by_blockdev(bs));
     }
 
+    aio_context_release(aio_context);
     return 0;
 }
 
@@ -1799,6 +1806,7 @@ void qmp_block_resize(bool has_device, const char *device,
 {
     Error *local_err = NULL;
     BlockDriverState *bs;
+    AioContext *aio_context;
     int ret;
 
     bs = bdrv_lookup_bs(has_device ? device : NULL,
@@ -1809,14 +1817,22 @@ void qmp_block_resize(bool has_device, const char *device,
         return;
     }
 
+    aio_context = bdrv_get_aio_context(bs);
+    aio_context_acquire(aio_context);
+
     if (!bdrv_is_first_non_filter(bs)) {
         error_set(errp, QERR_FEATURE_DISABLED, "resize");
-        return;
+        goto out;
     }
 
     if (size < 0) {
         error_set(errp, QERR_INVALID_PARAMETER_VALUE, "size", "a >0 size");
-        return;
+        goto out;
+    }
+
+    if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_RESIZE, NULL)) {
+        error_set(errp, QERR_DEVICE_IN_USE, device);
+        goto out;
     }
 
     /* complete all in-flight operations before resizing the device */
@@ -1842,6 +1858,9 @@ void qmp_block_resize(bool has_device, const char *device,
         error_setg_errno(errp, -ret, "Could not resize");
         break;
     }
+
+out:
+    aio_context_release(aio_context);
 }
 
 static void block_job_cb(void *opaque, int ret)
@@ -1866,14 +1885,17 @@ static void block_job_cb(void *opaque, int ret)
     bdrv_put_ref_bh_schedule(bs);
 }
 
-void qmp_block_stream(const char *device, bool has_base,
-                      const char *base, bool has_speed, int64_t speed,
+void qmp_block_stream(const char *device,
+                      bool has_base, const char *base,
+                      bool has_backing_file, const char *backing_file,
+                      bool has_speed, int64_t speed,
                       bool has_on_error, BlockdevOnError on_error,
                       Error **errp)
 {
     BlockDriverState *bs;
     BlockDriverState *base_bs = NULL;
     Error *local_err = NULL;
+    const char *base_name = NULL;
 
     if (!has_on_error) {
         on_error = BLOCKDEV_ON_ERROR_REPORT;
@@ -1889,15 +1911,27 @@ void qmp_block_stream(const char *device, bool has_base,
         return;
     }
 
-    if (base) {
+    if (has_base) {
         base_bs = bdrv_find_backing_image(bs, base);
         if (base_bs == NULL) {
             error_set(errp, QERR_BASE_NOT_FOUND, base);
             return;
         }
+        base_name = base;
     }
 
-    stream_start(bs, base_bs, base, has_speed ? speed : 0,
+    /* if we are streaming the entire chain, the result will have no backing
+     * file, and specifying one is therefore an error */
+    if (base_bs == NULL && has_backing_file) {
+        error_setg(errp, "backing file specified, but streaming the "
+                         "entire chain");
+        return;
+    }
+
+    /* backing_file string overrides base bs filename */
+    base_name = has_backing_file ? backing_file : base_name;
+
+    stream_start(bs, base_bs, base_name, has_speed ? speed : 0,
                  on_error, block_job_cb, bs, &local_err);
     if (local_err) {
         error_propagate(errp, local_err);
@@ -1908,7 +1942,9 @@ void qmp_block_stream(const char *device, bool has_base,
 }
 
 void qmp_block_commit(const char *device,
-                      bool has_base, const char *base, const char *top,
+                      bool has_base, const char *base,
+                      bool has_top, const char *top,
+                      bool has_backing_file, const char *backing_file,
                       bool has_speed, int64_t speed,
                       Error **errp)
 {
@@ -1927,6 +1963,11 @@ void qmp_block_commit(const char *device,
     /* drain all i/o before commits */
     bdrv_drain_all();
 
+    /* Important Note:
+     *  libvirt relies on the DeviceNotFound error class in order to probe for
+     *  live commit feature versions; for this to work, we must make sure to
+     *  perform the device lookup before any generic errors that may occur in a
+     *  scenario in which all optional arguments are omitted. */
     bs = bdrv_find(device);
     if (!bs) {
         error_set(errp, QERR_DEVICE_NOT_FOUND, device);
@@ -1940,7 +1981,7 @@ void qmp_block_commit(const char *device,
     /* default top_bs is the active layer */
     top_bs = bs;
 
-    if (top) {
+    if (has_top && top) {
         if (strcmp(bs->filename, top) != 0) {
             top_bs = bdrv_find_backing_image(bs, top);
         }
@@ -1962,12 +2003,23 @@ void qmp_block_commit(const char *device,
         return;
     }
 
+    /* Do not allow attempts to commit an image into itself */
+    if (top_bs == base_bs) {
+        error_setg(errp, "cannot commit an image into itself");
+        return;
+    }
+
     if (top_bs == bs) {
+        if (has_backing_file) {
+            error_setg(errp, "'backing-file' specified,"
+                             " but 'top' is the active layer");
+            return;
+        }
         commit_active_start(bs, base_bs, speed, on_error, block_job_cb,
                             bs, &local_err);
     } else {
         commit_start(bs, base_bs, top_bs, speed, on_error, block_job_cb, bs,
-                    &local_err);
+                     has_backing_file ? backing_file : NULL, &local_err);
     }
     if (local_err != NULL) {
         error_propagate(errp, local_err);
@@ -2094,6 +2146,8 @@ BlockDeviceInfoList *qmp_query_named_block_nodes(Error **errp)
 
 void qmp_drive_mirror(const char *device, const char *target,
                       bool has_format, const char *format,
+                      bool has_node_name, const char *node_name,
+                      bool has_replaces, const char *replaces,
                       enum MirrorSyncMode sync,
                       bool has_mode, enum NewImageMode mode,
                       bool has_speed, int64_t speed,
@@ -2107,6 +2161,7 @@ void qmp_drive_mirror(const char *device, const char *target,
     BlockDriverState *source, *target_bs;
     BlockDriver *drv = NULL;
     Error *local_err = NULL;
+    QDict *options = NULL;
     int flags;
     int64_t size;
     int ret;
@@ -2131,11 +2186,12 @@ void qmp_drive_mirror(const char *device, const char *target,
     }
 
     if (granularity != 0 && (granularity < 512 || granularity > 1048576 * 64)) {
-        error_set(errp, QERR_INVALID_PARAMETER, device);
+        error_set(errp, QERR_INVALID_PARAMETER_VALUE, "granularity",
+                  "a value in range [512B, 64MB]");
         return;
     }
     if (granularity & (granularity - 1)) {
-        error_set(errp, QERR_INVALID_PARAMETER, device);
+        error_set(errp, QERR_INVALID_PARAMETER_VALUE, "granularity", "power of 2");
         return;
     }
 
@@ -2180,6 +2236,29 @@ void qmp_drive_mirror(const char *device, const char *target,
         return;
     }
 
+    if (has_replaces) {
+        BlockDriverState *to_replace_bs;
+
+        if (!has_node_name) {
+            error_setg(errp, "a node-name must be provided when replacing a"
+                             " named node of the graph");
+            return;
+        }
+
+        to_replace_bs = check_to_replace_node(replaces, &local_err);
+
+        if (!to_replace_bs) {
+            error_propagate(errp, local_err);
+            return;
+        }
+
+        if (size != bdrv_getlength(to_replace_bs)) {
+            error_setg(errp, "cannot replace image with a mirror image of "
+                             "different size");
+            return;
+        }
+    }
+
     if ((sync == MIRROR_SYNC_MODE_FULL || !source)
         && mode != NEW_IMAGE_MODE_EXISTING)
     {
@@ -2208,18 +2287,28 @@ void qmp_drive_mirror(const char *device, const char *target,
         return;
     }
 
+    if (has_node_name) {
+        options = qdict_new();
+        qdict_put(options, "node-name", qstring_from_str(node_name));
+    }
+
     /* Mirroring takes care of copy-on-write using the source's backing
      * file.
      */
     target_bs = NULL;
-    ret = bdrv_open(&target_bs, target, NULL, NULL, flags | BDRV_O_NO_BACKING,
-                    drv, &local_err);
+    ret = bdrv_open(&target_bs, target, NULL, options,
+                    flags | BDRV_O_NO_BACKING, drv, &local_err);
     if (ret < 0) {
         error_propagate(errp, local_err);
         return;
     }
 
-    mirror_start(bs, target_bs, speed, granularity, buf_size, sync,
+    /* pass the node name to replace to mirror start since it's loose coupling
+     * and will allow to check whether the node still exist at mirror completion
+     */
+    mirror_start(bs, target_bs,
+                 has_replaces ? replaces : NULL,
+                 speed, granularity, buf_size, sync,
                  on_source_error, on_target_error,
                  block_job_cb, bs, &local_err);
     if (local_err != NULL) {
@@ -2314,6 +2403,85 @@ void qmp_block_job_complete(const char *device, Error **errp)
     block_job_complete(job, errp);
 }
 
+void qmp_change_backing_file(const char *device,
+                             const char *image_node_name,
+                             const char *backing_file,
+                             Error **errp)
+{
+    BlockDriverState *bs = NULL;
+    BlockDriverState *image_bs = NULL;
+    Error *local_err = NULL;
+    bool ro;
+    int open_flags;
+    int ret;
+
+    /* find the top layer BDS of the chain */
+    bs = bdrv_find(device);
+    if (!bs) {
+        error_set(errp, QERR_DEVICE_NOT_FOUND, device);
+        return;
+    }
+
+    image_bs = bdrv_lookup_bs(NULL, image_node_name, &local_err);
+    if (local_err) {
+        error_propagate(errp, local_err);
+        return;
+    }
+
+    if (!image_bs) {
+        error_setg(errp, "image file not found");
+        return;
+    }
+
+    if (bdrv_find_base(image_bs) == image_bs) {
+        error_setg(errp, "not allowing backing file change on an image "
+                         "without a backing file");
+        return;
+    }
+
+    /* even though we are not necessarily operating on bs, we need it to
+     * determine if block ops are currently prohibited on the chain */
+    if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_CHANGE, errp)) {
+        return;
+    }
+
+    /* final sanity check */
+    if (!bdrv_chain_contains(bs, image_bs)) {
+        error_setg(errp, "'%s' and image file are not in the same chain",
+                   device);
+        return;
+    }
+
+    /* if not r/w, reopen to make r/w */
+    open_flags = image_bs->open_flags;
+    ro = bdrv_is_read_only(image_bs);
+
+    if (ro) {
+        bdrv_reopen(image_bs, open_flags | BDRV_O_RDWR, &local_err);
+        if (local_err) {
+            error_propagate(errp, local_err);
+            return;
+        }
+    }
+
+    ret = bdrv_change_backing_file(image_bs, backing_file,
+                               image_bs->drv ? image_bs->drv->format_name : "");
+
+    if (ret < 0) {
+        error_setg_errno(errp, -ret, "Could not change backing file to '%s'",
+                         backing_file);
+        /* don't exit here, so we can try to restore open flags if
+         * appropriate */
+    }
+
+    if (ro) {
+        bdrv_reopen(image_bs, open_flags, &local_err);
+        if (local_err) {
+            error_propagate(errp, local_err); /* will preserve prior errp */
+        }
+    }
+}
+
 void qmp_blockdev_add(BlockdevOptions *options, Error **errp)
 {
     QmpOutputVisitor *ov = qmp_output_visitor_new();

blockjob.c

@@ -187,7 +187,7 @@ int block_job_cancel_sync(BlockJob *job)
     job->opaque = &data;
     block_job_cancel(job);
     while (data.ret == -EINPROGRESS) {
-        qemu_aio_wait();
+        aio_poll(bdrv_get_aio_context(bs), true);
     }
     return (data.cancelled && data.ret == 0) ? -ECANCELED : data.ret;
 }
@@ -205,11 +205,25 @@ void block_job_sleep_ns(BlockJob *job, QEMUClockType type, int64_t ns)
     if (block_job_is_paused(job)) {
         qemu_coroutine_yield();
     } else {
-        co_sleep_ns(type, ns);
+        co_aio_sleep_ns(bdrv_get_aio_context(job->bs), type, ns);
     }
     job->busy = true;
 }
 
+void block_job_yield(BlockJob *job)
+{
+    assert(job->busy);
+
+    /* Check cancellation *before* setting busy = false, too!  */
+    if (block_job_is_cancelled(job)) {
+        return;
+    }
+
+    job->busy = false;
+    qemu_coroutine_yield();
+    job->busy = true;
+}
+
 BlockJobInfo *block_job_query(BlockJob *job)
 {
     BlockJobInfo *info = g_new0(BlockJobInfo, 1);
@@ -256,7 +270,11 @@ void block_job_event_completed(BlockJob *job, const char *msg)
 
 void block_job_event_ready(BlockJob *job)
 {
-    qapi_event_send_block_job_ready(bdrv_get_device_name(job->bs), &error_abort);
+    qapi_event_send_block_job_ready(job->driver->job_type,
+                                    bdrv_get_device_name(job->bs),
+                                    job->len,
+                                    job->offset,
+                                    job->speed, &error_abort);
 }
 
 BlockErrorAction block_job_error_action(BlockJob *job, BlockDriverState *bs,
@@ -282,7 +300,7 @@ BlockErrorAction block_job_error_action(BlockJob *job, BlockDriverState *bs,
     default:
         abort();
     }
-    qapi_event_send_block_job_error(bdrv_get_device_name(bs),
+    qapi_event_send_block_job_error(bdrv_get_device_name(job->bs),
                                     is_read ? IO_OPERATION_TYPE_READ :
                                     IO_OPERATION_TYPE_WRITE,
                                     action, &error_abort);

configure

@@ -326,6 +326,7 @@ seccomp=""
 glusterfs=""
 glusterfs_discard="no"
 glusterfs_zerofill="no"
+archipelago=""
 virtio_blk_data_plane=""
 gtk=""
 gtkabi=""
@@ -1087,6 +1088,10 @@ for opt do
   ;;
   --enable-glusterfs) glusterfs="yes"
   ;;
+  --disable-archipelago) archipelago="no"
+  ;;
+  --enable-archipelago) archipelago="yes"
+  ;;
   --disable-virtio-blk-data-plane) virtio_blk_data_plane="no"
   ;;
   --enable-virtio-blk-data-plane) virtio_blk_data_plane="yes"
@@ -1382,6 +1387,8 @@ Advanced options (experts only):
   --enable-coroutine-pool  enable coroutine freelist (better performance)
   --enable-glusterfs       enable GlusterFS backend
   --disable-glusterfs      disable GlusterFS backend
+  --enable-archipelago     enable Archipelago backend
+  --disable-archipelago    disable Archipelago backend
   --enable-gcov            enable test coverage analysis with gcov
   --gcov=GCOV              use specified gcov [$gcov_tool]
   --disable-tpm            disable TPM support
@@ -1489,8 +1496,9 @@ for flag in $gcc_flags; do
     fi
 done
 
-if test "$stack_protector" != "no" ; then
+if test "$stack_protector" != "no"; then
   gcc_flags="-fstack-protector-strong -fstack-protector-all"
+  sp_on=0
   for flag in $gcc_flags; do
     # We need to check both a compile and a link, since some compiler
     # setups fail only on a .c->.o compile and some only at link time
@@ -1498,9 +1506,15 @@ if test "$stack_protector" != "no" ; then
        compile_prog "-Werror $flag" ""; then
       QEMU_CFLAGS="$QEMU_CFLAGS $flag"
       LIBTOOLFLAGS="$LIBTOOLFLAGS -Wc,$flag"
+      sp_on=1
       break
     fi
   done
+  if test "$stack_protector" = yes; then
+    if test $sp_on = 0; then
+      error_exit "Stack protector not supported"
+    fi
+  fi
 fi
 
 # Workaround for http://gcc.gnu.org/PR55489.  Happens with -fPIE/-fPIC and
@@ -1711,6 +1725,20 @@ else
     echo big/little test failed
 fi
 
+##########################################
+# L2TPV3 probe
+
+cat > $TMPC <<EOF
+#include <sys/socket.h>
+#include <linux/ip.h>
+int main(void) { return sizeof(struct mmsghdr); }
+EOF
+if compile_prog "" "" ; then
+  l2tpv3=yes
+else
+  l2tpv3=no
+fi
+
 ##########################################
 # pkg-config probe
 
@@ -3051,6 +3079,33 @@ EOF
   fi
 fi
 
+
+##########################################
+# archipelago probe
+if test "$archipelago" != "no" ; then
+    cat > $TMPC <<EOF
+#include <stdio.h>
+#include <xseg/xseg.h>
+#include <xseg/protocol.h>
+int main(void) {
+    xseg_initialize();
+    return 0;
+}
+EOF
+    archipelago_libs=-lxseg
+    if compile_prog "" "$archipelago_libs"; then
+        archipelago="yes"
+        libs_tools="$archipelago_libs $libs_tools"
+        libs_softmmu="$archipelago_libs $libs_softmmu"
+    else
+      if test "$archipelago" = "yes" ; then
+        feature_not_found "Archipelago backend support" "Install libxseg devel"
+      fi
+      archipelago="no"
+    fi
+fi
+
+
 ##########################################
 # glusterfs probe
 if test "$glusterfs" != "no" ; then
@@ -3066,7 +3121,8 @@ if test "$glusterfs" != "no" ; then
     fi
   else
     if test "$glusterfs" = "yes" ; then
-      feature_not_found "GlusterFS backend support" "Install glusterfs-api devel"
+      feature_not_found "GlusterFS backend support" \
+          "Install glusterfs-api devel >= 3"
     fi
     glusterfs="no"
   fi
@@ -3400,6 +3456,37 @@ if compile_prog "" "" ; then
   sendfile=yes
 fi
 
+# check for timerfd support (glibc 2.8 and newer)
+timerfd=no
+cat > $TMPC << EOF
+#include <sys/timerfd.h>
+
+int main(void)
+{
+    return(timerfd_create(CLOCK_REALTIME, 0));
+}
+EOF
+if compile_prog "" "" ; then
+  timerfd=yes
+fi
+
+# check for setns and unshare support
+setns=no
+cat > $TMPC << EOF
+#include <sched.h>
+
+int main(void)
+{
+    int ret;
+    ret = setns(0, 0);
+    ret = unshare(0);
+    return ret;
+}
+EOF
+if compile_prog "" "" ; then
+  setns=yes
+fi
+
 # Check if tools are available to build documentation.
 if test "$docs" != "no" ; then
   if has makeinfo && has pod2man; then
@@ -3453,7 +3540,7 @@ fi
 # Do we need libm
 cat > $TMPC << EOF
 #include <math.h>
-int main(void) { return isnan(sin(0.0)); }
+int main(int argc, char **argv) { return isnan(sin((double)argc)); }
 EOF
 if compile_prog "" "" ; then
   :
@@ -3511,7 +3598,8 @@ EOF
     spice_server_version=$($pkg_config --modversion spice-server)
   else
     if test "$spice" = "yes" ; then
-      feature_not_found "spice" "Install spice-server and spice-protocol devel"
+      feature_not_found "spice" \
+          "Install spice-server(>=0.12.0) and spice-protocol(>=0.12.3) devel"
     fi
     spice="no"
   fi
@@ -3542,7 +3630,7 @@ EOF
         smartcard_nss="yes"
     else
         if test "$smartcard_nss" = "yes"; then
-            feature_not_found "nss"
+            feature_not_found "nss" "Install nss devel >= 3.12.8"
         fi
         smartcard_nss="no"
     fi
@@ -3558,7 +3646,7 @@ if test "$libusb" != "no" ; then
         libs_softmmu="$libs_softmmu $libusb_libs"
     else
         if test "$libusb" = "yes"; then
-            feature_not_found "libusb" "Install libusb devel"
+            feature_not_found "libusb" "Install libusb devel >= 1.0.13"
         fi
         libusb="no"
     fi
@@ -3983,7 +4071,7 @@ if test "$libnfs" != "no" ; then
     LIBS="$LIBS $libnfs_libs"
   else
     if test "$libnfs" = "yes" ; then
-      feature_not_found "libnfs"
+      feature_not_found "libnfs" "Install libnfs devel >= 1.9.3"
     fi
     libnfs="no"
   fi
@@ -4230,6 +4318,7 @@ echo "seccomp support   $seccomp"
 echo "coroutine backend $coroutine"
 echo "coroutine pool    $coroutine_pool"
 echo "GlusterFS support $glusterfs"
+echo "Archipelago support $archipelago"
 echo "virtio-blk-data-plane $virtio_blk_data_plane"
 echo "gcov              $gcov_tool"
 echo "gcov enabled      $gcov"
@@ -4343,6 +4432,9 @@ fi
 if test "$netmap" = "yes" ; then
   echo "CONFIG_NETMAP=y" >> $config_host_mak
 fi
+if test "$l2tpv3" = "yes" ; then
+  echo "CONFIG_L2TPV3=y" >> $config_host_mak
+fi
 if test "$cap_ng" = "yes" ; then
   echo "CONFIG_LIBCAP=y" >> $config_host_mak
 fi
@@ -4463,6 +4555,12 @@ fi
 if test "$sendfile" = "yes" ; then
   echo "CONFIG_SENDFILE=y" >> $config_host_mak
 fi
+if test "$timerfd" = "yes" ; then
+  echo "CONFIG_TIMERFD=y" >> $config_host_mak
+fi
+if test "$setns" = "yes" ; then
+  echo "CONFIG_SETNS=y" >> $config_host_mak
+fi
 if test "$inotify" = "yes" ; then
   echo "CONFIG_INOTIFY=y" >> $config_host_mak
 fi
@@ -4665,6 +4763,11 @@ if test "$glusterfs_zerofill" = "yes" ; then
   echo "CONFIG_GLUSTERFS_ZEROFILL=y" >> $config_host_mak
 fi
 
+if test "$archipelago" = "yes" ; then
+  echo "CONFIG_ARCHIPELAGO=m" >> $config_host_mak
+  echo "ARCHIPELAGO_LIBS=$archipelago_libs" >> $config_host_mak
+fi
+
 if test "$libssh2" = "yes" ; then
   echo "CONFIG_LIBSSH2=m" >> $config_host_mak
   echo "LIBSSH2_CFLAGS=$libssh2_cflags" >> $config_host_mak
@@ -4942,6 +5045,8 @@ case "$target_name" in
     TARGET_BASE_ARCH=mips
     echo "TARGET_ABI_MIPSN64=y" >> $config_target_mak
   ;;
+  tricore)
+  ;;
   moxie)
   ;;
   or32)
@@ -4990,6 +5095,7 @@ case "$target_name" in
     echo "TARGET_ABI32=y" >> $config_target_mak
   ;;
   s390x)
+    gdb_xml_files="s390x-core64.xml s390-acr.xml s390-fpr.xml"
   ;;
   unicore32)
   ;;
@@ -5269,8 +5375,16 @@ for rom in seabios vgabios ; do
     echo "LD=$ld" >> $config_mak
 done
 
-if test "$docs" = "yes" ; then
-  mkdir -p QMP
+# set up qemu-iotests in this build directory
+iotests_common_env="tests/qemu-iotests/common.env"
+iotests_check="tests/qemu-iotests/check"
+
+echo "# Automatically generated by configure - do not modify" > "$iotests_common_env"
+echo >> "$iotests_common_env"
+echo "export PYTHON='$python'" >> "$iotests_common_env"
+
+if [ ! -e "$iotests_check" ]; then
+    symlink "$source_path/$iotests_check" "$iotests_check"
 fi
 
 # Save the configure command line for later reuse.

coroutine-win32.c

@@ -36,8 +36,17 @@ typedef struct
 static __thread CoroutineWin32 leader;
 static __thread Coroutine *current;
 
-CoroutineAction qemu_coroutine_switch(Coroutine *from_, Coroutine *to_,
-                                      CoroutineAction action)
+/* This function is marked noinline to prevent GCC from inlining it
+ * into coroutine_trampoline(). If we allow it to do that then it
+ * hoists the code to get the address of the TLS variable "current"
+ * out of the while() loop. This is an invalid transformation because
+ * the SwitchToFiber() call may be called when running thread A but
+ * return in thread B, and so we might be in a different thread
+ * context each time round the loop.
+ */
+CoroutineAction __attribute__((noinline))
+qemu_coroutine_switch(Coroutine *from_, Coroutine *to_,
+                      CoroutineAction action)
 {
     CoroutineWin32 *from = DO_UPCAST(CoroutineWin32, base, from_);
     CoroutineWin32 *to = DO_UPCAST(CoroutineWin32, base, to_);

cpu-exec.c

@@ -18,10 +18,114 @@
  */
 #include "config.h"
 #include "cpu.h"
+#include "trace.h"
 #include "disas/disas.h"
 #include "tcg.h"
 #include "qemu/atomic.h"
 #include "sysemu/qtest.h"
+#include "qemu/timer.h"
+
+/* -icount align implementation. */
+
+typedef struct SyncClocks {
+    int64_t diff_clk;
+    int64_t last_cpu_icount;
+    int64_t realtime_clock;
+} SyncClocks;
+
+#if !defined(CONFIG_USER_ONLY)
+/* Allow the guest to have a max 3ms advance.
+ * The difference between the 2 clocks could therefore
+ * oscillate around 0.
+ */
+#define VM_CLOCK_ADVANCE 3000000
+#define THRESHOLD_REDUCE 1.5
+#define MAX_DELAY_PRINT_RATE 2000000000LL
+#define MAX_NB_PRINTS 100
+
+static void align_clocks(SyncClocks *sc, const CPUState *cpu)
+{
+    int64_t cpu_icount;
+
+    if (!icount_align_option) {
+        return;
+    }
+
+    cpu_icount = cpu->icount_extra + cpu->icount_decr.u16.low;
+    sc->diff_clk += cpu_icount_to_ns(sc->last_cpu_icount - cpu_icount);
+    sc->last_cpu_icount = cpu_icount;
+
+    if (sc->diff_clk > VM_CLOCK_ADVANCE) {
+#ifndef _WIN32
+        struct timespec sleep_delay, rem_delay;
+        sleep_delay.tv_sec = sc->diff_clk / 1000000000LL;
+        sleep_delay.tv_nsec = sc->diff_clk % 1000000000LL;
+        if (nanosleep(&sleep_delay, &rem_delay) < 0) {
+            sc->diff_clk -= (sleep_delay.tv_sec - rem_delay.tv_sec) * 1000000000LL;
+            sc->diff_clk -= sleep_delay.tv_nsec - rem_delay.tv_nsec;
+        } else {
+            sc->diff_clk = 0;
+        }
+#else
+        Sleep(sc->diff_clk / SCALE_MS);
+        sc->diff_clk = 0;
+#endif
+    }
+}
+
+static void print_delay(const SyncClocks *sc)
+{
+    static float threshold_delay;
+    static int64_t last_realtime_clock;
+    static int nb_prints;
+
+    if (icount_align_option &&
+        sc->realtime_clock - last_realtime_clock >= MAX_DELAY_PRINT_RATE &&
+        nb_prints < MAX_NB_PRINTS) {
+        if ((-sc->diff_clk / (float)1000000000LL > threshold_delay) ||
+            (-sc->diff_clk / (float)1000000000LL <
+             (threshold_delay - THRESHOLD_REDUCE))) {
+            threshold_delay = (-sc->diff_clk / 1000000000LL) + 1;
+            printf("Warning: The guest is now late by %.1f to %.1f seconds\n",
+                   threshold_delay - 1,
+                   threshold_delay);
+            nb_prints++;
+            last_realtime_clock = sc->realtime_clock;
+        }
+    }
+}
+
+static void init_delay_params(SyncClocks *sc,
+                              const CPUState *cpu)
+{
+    if (!icount_align_option) {
+        return;
+    }
+    sc->realtime_clock = qemu_clock_get_ns(QEMU_CLOCK_REALTIME);
+    sc->diff_clk = qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL) -
+                   sc->realtime_clock +
+                   cpu_get_clock_offset();
+    sc->last_cpu_icount = cpu->icount_extra + cpu->icount_decr.u16.low;
+    if (sc->diff_clk < max_delay) {
+        max_delay = sc->diff_clk;
+    }
+    if (sc->diff_clk > max_advance) {
+        max_advance = sc->diff_clk;
+    }
+
+    /* Print every 2s max if the guest is late. We limit the number
+       of printed messages to NB_PRINT_MAX(currently 100) */
+    print_delay(sc);
+}
+#else
+static void align_clocks(SyncClocks *sc, const CPUState *cpu)
+{
+}
+
+static void init_delay_params(SyncClocks *sc, const CPUState *cpu)
+{
+}
+#endif /* CONFIG USER ONLY */
 
 void cpu_loop_exit(CPUState *cpu)
 {
@@ -65,6 +169,9 @@ static inline tcg_target_ulong cpu_tb_exec(CPUState *cpu, uint8_t *tb_ptr)
 #endif /* DEBUG_DISAS */
 
     next_tb = tcg_qemu_tb_exec(env, tb_ptr);
+    trace_exec_tb_exit((void *) (next_tb & ~TB_EXIT_MASK),
+                       next_tb & TB_EXIT_MASK);
+
     if ((next_tb & TB_EXIT_MASK) > TB_EXIT_IDX1) {
         /* We didn't start executing this TB (eg because the instruction
          * counter hit zero); we must restore the guest PC to the address
@@ -105,6 +212,7 @@ static void cpu_exec_nocache(CPUArchState *env, int max_cycles,
                      max_cycles);
     cpu->current_tb = tb;
     /* execute the generated code */
+    trace_exec_tb_nocache(tb, tb->pc);
     cpu_tb_exec(cpu, tb->tc_ptr);
     cpu->current_tb = NULL;
     tb_phys_invalidate(tb, -1);
@@ -227,6 +335,8 @@ int cpu_exec(CPUArchState *env)
     TranslationBlock *tb;
     uint8_t *tc_ptr;
     uintptr_t next_tb;
+    SyncClocks sc;
+
     /* This must be volatile so it is not trashed by longjmp() */
     volatile bool have_tb_lock = false;
 
@@ -277,12 +387,20 @@ int cpu_exec(CPUArchState *env)
 #elif defined(TARGET_CRIS)
 #elif defined(TARGET_S390X)
 #elif defined(TARGET_XTENSA)
+#elif defined(TARGET_TRICORE)
     /* XXXXX */
 #else
 #error unsupported target CPU
 #endif
     cpu->exception_index = -1;
 
+    /* Calculate difference between guest clock and host clock.
+     * This delay includes the delay of the last cycle, so
+     * what we have to do is sleep until it is 0. As for the
+     * advance/delay we gain here, we try to fix it next time.
+     */
+    init_delay_params(&sc, cpu);
+
     /* prepare setjmp context for exception handling */
     for(;;) {
         if (sigsetjmp(cpu->jmp_env, 0) == 0) {
@@ -327,7 +445,8 @@ int cpu_exec(CPUArchState *env)
                     }
 #if defined(TARGET_ARM) || defined(TARGET_SPARC) || defined(TARGET_MIPS) || \
     defined(TARGET_PPC) || defined(TARGET_ALPHA) || defined(TARGET_CRIS) || \
-    defined(TARGET_MICROBLAZE) || defined(TARGET_LM32) || defined(TARGET_UNICORE32)
+    defined(TARGET_MICROBLAZE) || defined(TARGET_LM32) ||                   \
+    defined(TARGET_UNICORE32) || defined(TARGET_TRICORE)
                     if (interrupt_request & CPU_INTERRUPT_HALT) {
                         cpu->interrupt_request &= ~CPU_INTERRUPT_HALT;
                         cpu->halted = 1;
@@ -443,6 +562,12 @@ int cpu_exec(CPUArchState *env)
                         cc->do_interrupt(cpu);
                         next_tb = 0;
                     }
+#elif defined(TARGET_TRICORE)
+                    if ((interrupt_request & CPU_INTERRUPT_HARD)) {
+                        cc->do_interrupt(cpu);
+                        next_tb = 0;
+                    }
+
 #elif defined(TARGET_OPENRISC)
                     {
                         int idx = -1;
@@ -637,6 +762,7 @@ int cpu_exec(CPUArchState *env)
                 cpu->current_tb = tb;
                 barrier();
                 if (likely(!cpu->exit_request)) {
+                    trace_exec_tb(tb, tb->pc);
                     tc_ptr = tb->tc_ptr;
                     /* execute the generated code */
                     next_tb = cpu_tb_exec(cpu, tc_ptr);
@@ -672,6 +798,7 @@ int cpu_exec(CPUArchState *env)
                             if (insns_left > 0) {
                                 /* Execute remaining instructions.  */
                                 cpu_exec_nocache(env, insns_left, tb);
+                                align_clocks(&sc, cpu);
                             }
                             cpu->exception_index = EXCP_INTERRUPT;
                             next_tb = 0;
@@ -684,6 +811,9 @@ int cpu_exec(CPUArchState *env)
                     }
                 }
                 cpu->current_tb = NULL;
+                /* Try to align the host and virtual clocks
+                   if the guest is in advance */
+                align_clocks(&sc, cpu);
                 /* reset soft MMU for next block (it can currently
                    only be set by a memory fault) */
             } /* for(;;) */
@@ -724,6 +854,7 @@ int cpu_exec(CPUArchState *env)
               | env->cc_dest | (env->cc_x << 4);
 #elif defined(TARGET_MICROBLAZE)
 #elif defined(TARGET_MIPS)
+#elif defined(TARGET_TRICORE)
 #elif defined(TARGET_MOXIE)
 #elif defined(TARGET_OPENRISC)
 #elif defined(TARGET_SH4)

cpus.c

@@ -40,6 +40,7 @@
 #include "qemu/bitmap.h"
 #include "qemu/seqlock.h"
 #include "qapi-event.h"
+#include "hw/nmi.h"
 
 #ifndef _WIN32
 #include "qemu/compatfd.h"
@@ -64,6 +65,8 @@
 #endif /* CONFIG_LINUX */
 
 static CPUState *next_cpu;
+int64_t max_delay;
+int64_t max_advance;
 
 bool cpu_is_stopped(CPUState *cpu)
 {
@@ -102,17 +105,12 @@ static bool all_cpu_threads_idle(void)
 
 /* Protected by TimersState seqlock */
 
-/* Compensate for varying guest execution speed.  */
-static int64_t qemu_icount_bias;
-static int64_t vm_clock_warp_start;
+static int64_t vm_clock_warp_start = -1;
 /* Conversion factor from emulated instructions to virtual clock ticks.  */
 static int icount_time_shift;
 /* Arbitrarily pick 1MIPS as the minimum allowable speed.  */
 #define MAX_ICOUNT_SHIFT 10
 
-/* Only written by TCG thread */
-static int64_t qemu_icount;
-
 static QEMUTimer *icount_rt_timer;
 static QEMUTimer *icount_vm_timer;
 static QEMUTimer *icount_warp_timer;
@@ -129,6 +127,11 @@ typedef struct TimersState {
     int64_t cpu_clock_offset;
     int32_t cpu_ticks_enabled;
     int64_t dummy;
+
+    /* Compensate for varying guest execution speed.  */
+    int64_t qemu_icount_bias;
+    /* Only written by TCG thread */
+    int64_t qemu_icount;
 } TimersState;
 
 static TimersState timers_state;
@@ -139,14 +142,14 @@ static int64_t cpu_get_icount_locked(void)
     int64_t icount;
     CPUState *cpu = current_cpu;
 
-    icount = qemu_icount;
+    icount = timers_state.qemu_icount;
     if (cpu) {
         if (!cpu_can_do_io(cpu)) {
             fprintf(stderr, "Bad clock read\n");
         }
         icount -= (cpu->icount_decr.u16.low + cpu->icount_extra);
     }
-    return qemu_icount_bias + (icount << icount_time_shift);
+    return timers_state.qemu_icount_bias + cpu_icount_to_ns(icount);
 }
 
 int64_t cpu_get_icount(void)
@@ -162,6 +165,11 @@ int64_t cpu_get_icount(void)
     return icount;
 }
 
+int64_t cpu_icount_to_ns(int64_t icount)
+{
+    return icount << icount_time_shift;
+}
+
 /* return the host CPU cycle counter and handle stop/restart */
 /* Caller must hold the BQL */
 int64_t cpu_get_ticks(void)
@@ -214,6 +222,23 @@ int64_t cpu_get_clock(void)
     return ti;
 }
 
+/* return the offset between the host clock and virtual CPU clock */
+int64_t cpu_get_clock_offset(void)
+{
+    int64_t ti;
+    unsigned start;
+
+    do {
+        start = seqlock_read_begin(&timers_state.vm_clock_seqlock);
+        ti = timers_state.cpu_clock_offset;
+        if (!timers_state.cpu_ticks_enabled) {
+            ti -= get_clock();
+        }
+    } while (seqlock_read_retry(&timers_state.vm_clock_seqlock, start));
+
+    return -ti;
+}
+
 /* enable cpu_get_ticks()
  * Caller must hold BQL which server as mutex for vm_clock_seqlock.
  */
@@ -284,7 +309,8 @@ static void icount_adjust(void)
         icount_time_shift++;
     }
     last_delta = delta;
-    qemu_icount_bias = cur_icount - (qemu_icount << icount_time_shift);
+    timers_state.qemu_icount_bias = cur_icount
+                              - (timers_state.qemu_icount << icount_time_shift);
     seqlock_write_unlock(&timers_state.vm_clock_seqlock);
 }
 
@@ -333,7 +359,7 @@ static void icount_warp_rt(void *opaque)
             int64_t delta = cur_time - cur_icount;
             warp_delta = MIN(warp_delta, delta);
         }
-        qemu_icount_bias += warp_delta;
+        timers_state.qemu_icount_bias += warp_delta;
     }
     vm_clock_warp_start = -1;
     seqlock_write_unlock(&timers_state.vm_clock_seqlock);
@@ -351,7 +377,7 @@ void qtest_clock_warp(int64_t dest)
         int64_t deadline = qemu_clock_deadline_ns_all(QEMU_CLOCK_VIRTUAL);
         int64_t warp = qemu_soonest_timeout(dest - clock, deadline);
         seqlock_write_lock(&timers_state.vm_clock_seqlock);
-        qemu_icount_bias += warp;
+        timers_state.qemu_icount_bias += warp;
         seqlock_write_unlock(&timers_state.vm_clock_seqlock);
 
         qemu_clock_run_timers(QEMU_CLOCK_VIRTUAL);
@@ -428,6 +454,25 @@ void qemu_clock_warp(QEMUClockType type)
     }
 }
 
+static bool icount_state_needed(void *opaque)
+{
+    return use_icount;
+}
+
+/*
+ * This is a subsection for icount migration.
+ */
+static const VMStateDescription icount_vmstate_timers = {
+    .name = "timer/icount",
+    .version_id = 1,
+    .minimum_version_id = 1,
+    .fields = (VMStateField[]) {
+        VMSTATE_INT64(qemu_icount_bias, TimersState),
+        VMSTATE_INT64(qemu_icount, TimersState),
+        VMSTATE_END_OF_LIST()
+    }
+};
+
 static const VMStateDescription vmstate_timers = {
     .name = "timer",
     .version_id = 2,
@@ -437,23 +482,44 @@ static const VMStateDescription vmstate_timers = {
         VMSTATE_INT64(dummy, TimersState),
         VMSTATE_INT64_V(cpu_clock_offset, TimersState, 2),
         VMSTATE_END_OF_LIST()
+    },
+    .subsections = (VMStateSubsection[]) {
+        {
+            .vmsd = &icount_vmstate_timers,
+            .needed = icount_state_needed,
+        }, {
+            /* empty */
+        }
     }
 };
 
-void configure_icount(const char *option)
+void configure_icount(QemuOpts *opts, Error **errp)
 {
+    const char *option;
+    char *rem_str = NULL;
+
     seqlock_init(&timers_state.vm_clock_seqlock, NULL);
     vmstate_register(NULL, 0, &vmstate_timers, &timers_state);
+    option = qemu_opt_get(opts, "shift");
     if (!option) {
+        if (qemu_opt_get(opts, "align") != NULL) {
+            error_setg(errp, "Please specify shift option when using align");
+        }
         return;
     }
-
+    icount_align_option = qemu_opt_get_bool(opts, "align", false);
     icount_warp_timer = timer_new_ns(QEMU_CLOCK_REALTIME,
                                           icount_warp_rt, NULL);
     if (strcmp(option, "auto") != 0) {
-        icount_time_shift = strtol(option, NULL, 0);
+        errno = 0;
+        icount_time_shift = strtol(option, &rem_str, 0);
+        if (errno != 0 || *rem_str != '\0' || !strlen(option)) {
+            error_setg(errp, "icount: Invalid shift value");
+        }
         use_icount = 1;
         return;
+    } else if (icount_align_option) {
+        error_setg(errp, "shift=auto and align=on are incompatible");
     }
 
     use_icount = 2;
@@ -1250,7 +1316,8 @@ static int tcg_cpu_exec(CPUArchState *env)
         int64_t count;
         int64_t deadline;
         int decr;
-        qemu_icount -= (cpu->icount_decr.u16.low + cpu->icount_extra);
+        timers_state.qemu_icount -= (cpu->icount_decr.u16.low
+                                    + cpu->icount_extra);
         cpu->icount_decr.u16.low = 0;
         cpu->icount_extra = 0;
         deadline = qemu_clock_deadline_ns_all(QEMU_CLOCK_VIRTUAL);
@@ -1265,7 +1332,7 @@ static int tcg_cpu_exec(CPUArchState *env)
         }
 
         count = qemu_icount_round(deadline);
-        qemu_icount += count;
+        timers_state.qemu_icount += count;
         decr = (count > 0xffff) ? 0xffff : count;
         count -= decr;
         cpu->icount_decr.u16.low = decr;
@@ -1278,7 +1345,8 @@ static int tcg_cpu_exec(CPUArchState *env)
     if (use_icount) {
         /* Fold pending instructions back into the
            instruction counter, and clear the interrupt flag.  */
-        qemu_icount -= (cpu->icount_decr.u16.low + cpu->icount_extra);
+        timers_state.qemu_icount -= (cpu->icount_decr.u16.low
+                        + cpu->icount_extra);
         cpu->icount_decr.u32 = 0;
         cpu->icount_extra = 0;
     }
@@ -1342,6 +1410,9 @@ CpuInfoList *qmp_query_cpus(Error **errp)
 #elif defined(TARGET_MIPS)
         MIPSCPU *mips_cpu = MIPS_CPU(cpu);
         CPUMIPSState *env = &mips_cpu->env;
+#elif defined(TARGET_TRICORE)
+        TriCoreCPU *tricore_cpu = TRICORE_CPU(cpu);
+        CPUTriCoreState *env = &tricore_cpu->env;
 #endif
 
         cpu_synchronize_state(cpu);
@@ -1366,6 +1437,9 @@ CpuInfoList *qmp_query_cpus(Error **errp)
 #elif defined(TARGET_MIPS)
         info->value->has_PC = true;
         info->value->PC = env->active_tc.PC;
+#elif defined(TARGET_TRICORE)
+        info->value->has_PC = true;
+        info->value->PC = env->PC;
 #endif
 
         /* XXX: waiting for the qapi to support GSList */
@@ -1469,21 +1543,24 @@ void qmp_inject_nmi(Error **errp)
             apic_deliver_nmi(cpu->apic_state);
         }
     }
-#elif defined(TARGET_S390X)
-    CPUState *cs;
-    S390CPU *cpu;
-
-    CPU_FOREACH(cs) {
-        cpu = S390_CPU(cs);
-        if (cpu->env.cpu_num == monitor_get_cpu_index()) {
-            if (s390_cpu_restart(S390_CPU(cs)) == -1) {
-                error_set(errp, QERR_UNSUPPORTED);
-                return;
-            }
-            break;
-        }
-    }
 #else
-    error_set(errp, QERR_UNSUPPORTED);
+    nmi_monitor_handle(monitor_get_cpu_index(), errp);
 #endif
 }
+
+void dump_drift_info(FILE *f, fprintf_function cpu_fprintf)
+{
+    if (!use_icount) {
+        return;
+    }
+
+    cpu_fprintf(f, "Host - Guest clock  %"PRIi64" ms\n",
+                (cpu_get_clock() - cpu_get_icount())/SCALE_MS);
+    if (icount_align_option) {
+        cpu_fprintf(f, "Max guest delay     %"PRIi64" ms\n", -max_delay/SCALE_MS);
+        cpu_fprintf(f, "Max guest advance   %"PRIi64" ms\n", max_advance/SCALE_MS);
+    } else {
+        cpu_fprintf(f, "Max guest delay     NA\n");
+        cpu_fprintf(f, "Max guest advance   NA\n");
+    }
+}

cputlb.c

@@ -60,8 +60,10 @@ void tlb_flush(CPUState *cpu, int flush_global)
     cpu->current_tb = NULL;
 
     memset(env->tlb_table, -1, sizeof(env->tlb_table));
+    memset(env->tlb_v_table, -1, sizeof(env->tlb_v_table));
     memset(cpu->tb_jmp_cache, 0, sizeof(cpu->tb_jmp_cache));
 
+    env->vtlb_index = 0;
     env->tlb_flush_addr = -1;
     env->tlb_flush_mask = 0;
     tlb_flush_count++;
@@ -108,6 +110,14 @@ void tlb_flush_page(CPUState *cpu, target_ulong addr)
         tlb_flush_entry(&env->tlb_table[mmu_idx][i], addr);
     }
 
+    /* check whether there are entries that need to be flushed in the vtlb */
+    for (mmu_idx = 0; mmu_idx < NB_MMU_MODES; mmu_idx++) {
+        int k;
+        for (k = 0; k < CPU_VTLB_SIZE; k++) {
+            tlb_flush_entry(&env->tlb_v_table[mmu_idx][k], addr);
+        }
+    }
+
     tb_flush_jmp_cache(cpu, addr);
 }
 
@@ -172,6 +182,11 @@ void cpu_tlb_reset_dirty_all(ram_addr_t start1, ram_addr_t length)
                 tlb_reset_dirty_range(&env->tlb_table[mmu_idx][i],
                                       start1, length);
             }
+
+            for (i = 0; i < CPU_VTLB_SIZE; i++) {
+                tlb_reset_dirty_range(&env->tlb_v_table[mmu_idx][i],
+                                      start1, length);
+            }
         }
     }
 }
@@ -195,6 +210,13 @@ void tlb_set_dirty(CPUArchState *env, target_ulong vaddr)
     for (mmu_idx = 0; mmu_idx < NB_MMU_MODES; mmu_idx++) {
         tlb_set_dirty1(&env->tlb_table[mmu_idx][i], vaddr);
     }
+
+    for (mmu_idx = 0; mmu_idx < NB_MMU_MODES; mmu_idx++) {
+        int k;
+        for (k = 0; k < CPU_VTLB_SIZE; k++) {
+            tlb_set_dirty1(&env->tlb_v_table[mmu_idx][k], vaddr);
+        }
+    }
 }
 
 /* Our TLB does not support large pages, so remember the area covered by
@@ -235,6 +257,7 @@ void tlb_set_page(CPUState *cpu, target_ulong vaddr,
     uintptr_t addend;
     CPUTLBEntry *te;
     hwaddr iotlb, xlat, sz;
+    unsigned vidx = env->vtlb_index++ % CPU_VTLB_SIZE;
 
     assert(size >= TARGET_PAGE_SIZE);
     if (size != TARGET_PAGE_SIZE) {
@@ -267,8 +290,14 @@ void tlb_set_page(CPUState *cpu, target_ulong vaddr,
                                             prot, &address);
 
     index = (vaddr >> TARGET_PAGE_BITS) & (CPU_TLB_SIZE - 1);
-    env->iotlb[mmu_idx][index] = iotlb - vaddr;
     te = &env->tlb_table[mmu_idx][index];
+
+    /* do not discard the translation in te, evict it into a victim tlb */
+    env->tlb_v_table[mmu_idx][vidx] = *te;
+    env->iotlb_v[mmu_idx][vidx] = env->iotlb[mmu_idx][index];
+
+    /* refill the tlb */
+    env->iotlb[mmu_idx][index] = iotlb - vaddr;
     te->addend = addend - vaddr;
     if (prot & PAGE_READ) {
         te->addr_read = address;

device_tree.c

@@ -20,6 +20,7 @@
 
 #include "config.h"
 #include "qemu-common.h"
+#include "qemu/error-report.h"
 #include "sysemu/device_tree.h"
 #include "sysemu/sysemu.h"
 #include "hw/loader.h"
@@ -59,13 +60,13 @@ void *create_device_tree(int *sizep)
     }
     ret = fdt_open_into(fdt, fdt, *sizep);
     if (ret) {
-        fprintf(stderr, "Unable to copy device tree in memory\n");
+        error_report("Unable to copy device tree in memory");
         exit(1);
     }
 
     return fdt;
 fail:
-    fprintf(stderr, "%s Couldn't create dt: %s\n", __func__, fdt_strerror(ret));
+    error_report("%s Couldn't create dt: %s", __func__, fdt_strerror(ret));
     exit(1);
 }
 
@@ -79,8 +80,8 @@ void *load_device_tree(const char *filename_path, int *sizep)
     *sizep = 0;
     dt_size = get_image_size(filename_path);
     if (dt_size < 0) {
-        printf("Unable to get size of device tree file '%s'\n",
-            filename_path);
+        error_report("Unable to get size of device tree file '%s'",
+                     filename_path);
         goto fail;
     }
 
@@ -92,21 +93,21 @@ void *load_device_tree(const char *filename_path, int *sizep)
 
     dt_file_load_size = load_image(filename_path, fdt);
     if (dt_file_load_size < 0) {
-        printf("Unable to open device tree file '%s'\n",
-               filename_path);
+        error_report("Unable to open device tree file '%s'",
+                     filename_path);
         goto fail;
     }
 
     ret = fdt_open_into(fdt, fdt, dt_size);
     if (ret) {
-        printf("Unable to copy device tree in memory\n");
+        error_report("Unable to copy device tree in memory");
         goto fail;
     }
 
     /* Check sanity of device tree */
     if (fdt_check_header(fdt)) {
-        printf ("Device tree file loaded into memory is invalid: %s\n",
-            filename_path);
+        error_report("Device tree file loaded into memory is invalid: %s",
+                     filename_path);
         goto fail;
     }
     *sizep = dt_size;
@@ -123,8 +124,8 @@ static int findnode_nofail(void *fdt, const char *node_path)
 
     offset = fdt_path_offset(fdt, node_path);
     if (offset < 0) {
-        fprintf(stderr, "%s Couldn't find node %s: %s\n", __func__, node_path,
-                fdt_strerror(offset));
+        error_report("%s Couldn't find node %s: %s", __func__, node_path,
+                     fdt_strerror(offset));
         exit(1);
     }
 
@@ -138,8 +139,8 @@ int qemu_fdt_setprop(void *fdt, const char *node_path,
 
     r = fdt_setprop(fdt, findnode_nofail(fdt, node_path), property, val, size);
     if (r < 0) {
-        fprintf(stderr, "%s: Couldn't set %s/%s: %s\n", __func__, node_path,
-                property, fdt_strerror(r));
+        error_report("%s: Couldn't set %s/%s: %s", __func__, node_path,
+                     property, fdt_strerror(r));
         exit(1);
     }
 
@@ -153,8 +154,8 @@ int qemu_fdt_setprop_cell(void *fdt, const char *node_path,
 
     r = fdt_setprop_cell(fdt, findnode_nofail(fdt, node_path), property, val);
     if (r < 0) {
-        fprintf(stderr, "%s: Couldn't set %s/%s = %#08x: %s\n", __func__,
-                node_path, property, val, fdt_strerror(r));
+        error_report("%s: Couldn't set %s/%s = %#08x: %s", __func__,
+                     node_path, property, val, fdt_strerror(r));
         exit(1);
     }
 
@@ -175,8 +176,8 @@ int qemu_fdt_setprop_string(void *fdt, const char *node_path,
 
     r = fdt_setprop_string(fdt, findnode_nofail(fdt, node_path), property, string);
     if (r < 0) {
-        fprintf(stderr, "%s: Couldn't set %s/%s = %s: %s\n", __func__,
-                node_path, property, string, fdt_strerror(r));
+        error_report("%s: Couldn't set %s/%s = %s: %s", __func__,
+                     node_path, property, string, fdt_strerror(r));
         exit(1);
     }
 
@@ -193,8 +194,8 @@ const void *qemu_fdt_getprop(void *fdt, const char *node_path,
     }
     r = fdt_getprop(fdt, findnode_nofail(fdt, node_path), property, lenp);
     if (!r) {
-        fprintf(stderr, "%s: Couldn't get %s/%s: %s\n", __func__,
-                node_path, property, fdt_strerror(*lenp));
+        error_report("%s: Couldn't get %s/%s: %s", __func__,
+                     node_path, property, fdt_strerror(*lenp));
         exit(1);
     }
     return r;
@@ -206,8 +207,8 @@ uint32_t qemu_fdt_getprop_cell(void *fdt, const char *node_path,
     int len;
     const uint32_t *p = qemu_fdt_getprop(fdt, node_path, property, &len);
     if (len != 4) {
-        fprintf(stderr, "%s: %s/%s not 4 bytes long (not a cell?)\n",
-                __func__, node_path, property);
+        error_report("%s: %s/%s not 4 bytes long (not a cell?)",
+                     __func__, node_path, property);
         exit(1);
     }
     return be32_to_cpu(*p);
@@ -219,8 +220,8 @@ uint32_t qemu_fdt_get_phandle(void *fdt, const char *path)
 
     r = fdt_get_phandle(fdt, findnode_nofail(fdt, path));
     if (r == 0) {
-        fprintf(stderr, "%s: Couldn't get phandle for %s: %s\n", __func__,
-                path, fdt_strerror(r));
+        error_report("%s: Couldn't get phandle for %s: %s", __func__,
+                     path, fdt_strerror(r));
         exit(1);
     }
 
@@ -265,8 +266,8 @@ int qemu_fdt_nop_node(void *fdt, const char *node_path)
 
     r = fdt_nop_node(fdt, findnode_nofail(fdt, node_path));
     if (r < 0) {
-        fprintf(stderr, "%s: Couldn't nop node %s: %s\n", __func__, node_path,
-                fdt_strerror(r));
+        error_report("%s: Couldn't nop node %s: %s", __func__, node_path,
+                     fdt_strerror(r));
         exit(1);
     }
 
@@ -294,8 +295,8 @@ int qemu_fdt_add_subnode(void *fdt, const char *name)
 
     retval = fdt_add_subnode(fdt, parent, basename);
     if (retval < 0) {
-        fprintf(stderr, "FDT: Failed to create subnode %s: %s\n", name,
-                fdt_strerror(retval));
+        error_report("FDT: Failed to create subnode %s: %s", name,
+                     fdt_strerror(retval));
         exit(1);
     }
 

disas/libvixl/Makefile.objs

@@ -3,6 +3,6 @@ libvixl_OBJS = utils.o \
                a64/decoder-a64.o \
                a64/disasm-a64.o
 
-$(addprefix $(obj)/,$(libvixl_OBJS)): QEMU_CFLAGS += -I$(SRC_PATH)/disas/libvixl
+$(addprefix $(obj)/,$(libvixl_OBJS)): QEMU_CFLAGS := -I$(SRC_PATH)/disas/libvixl $(QEMU_CFLAGS)
 
 common-obj-$(CONFIG_ARM_A64_DIS) += $(libvixl_OBJS)

disas/libvixl/README

@@ -2,7 +2,7 @@
 The code in this directory is a subset of libvixl:
  https://github.com/armvixl/vixl
 (specifically, it is the set of files needed for disassembly only,
-taken from libvixl 1.1).
+taken from libvixl 1.5).
 Bugfixes should preferably be sent upstream initially.
 
 The disassembler does not currently support the entire A64 instruction

disas/libvixl/a64/assembler-a64.h

@@ -28,6 +28,7 @@
 #define VIXL_A64_ASSEMBLER_A64_H_
 
 #include <list>
+#include <stack>
 
 #include "globals.h"
 #include "utils.h"
@@ -574,34 +575,107 @@ class MemOperand {
 
 class Label {
  public:
-  Label() : is_bound_(false), link_(NULL), target_(NULL) {}
+  Label() : location_(kLocationUnbound) {}
   ~Label() {
     // If the label has been linked to, it needs to be bound to a target.
     VIXL_ASSERT(!IsLinked() || IsBound());
   }
 
-  inline Instruction* link() const { return link_; }
-  inline Instruction* target() const { return target_; }
-
-  inline bool IsBound() const { return is_bound_; }
-  inline bool IsLinked() const { return link_ != NULL; }
-
-  inline void set_link(Instruction* new_link) { link_ = new_link; }
-
-  static const int kEndOfChain = 0;
+  inline bool IsBound() const { return location_ >= 0; }
+  inline bool IsLinked() const { return !links_.empty(); }
 
  private:
-  // Indicates if the label has been bound, ie its location is fixed.
-  bool is_bound_;
-  // Branches instructions branching to this label form a chained list, with
-  // their offset indicating where the next instruction is located.
-  // link_ points to the latest branch instruction generated branching to this
-  // branch.
-  // If link_ is not NULL, the label has been linked to.
-  Instruction* link_;
-  // The label location.
-  Instruction* target_;
+  // The list of linked instructions is stored in a stack-like structure. We
+  // don't use std::stack directly because it's slow for the common case where
+  // only one or two instructions refer to a label, and labels themselves are
+  // short-lived. This class behaves like std::stack, but the first few links
+  // are preallocated (configured by kPreallocatedLinks).
+  //
+  // If more than N links are required, this falls back to std::stack.
+  class LinksStack {
+   public:
+    LinksStack() : size_(0), links_extended_(NULL) {}
+    ~LinksStack() {
+      delete links_extended_;
+    }
 
+    size_t size() const {
+      return size_;
+    }
+
+    bool empty() const {
+      return size_ == 0;
+    }
+
+    void push(ptrdiff_t value) {
+      if (size_ < kPreallocatedLinks) {
+        links_[size_] = value;
+      } else {
+        if (links_extended_ == NULL) {
+          links_extended_ = new std::stack<ptrdiff_t>();
+        }
+        VIXL_ASSERT(size_ == (links_extended_->size() + kPreallocatedLinks));
+        links_extended_->push(value);
+      }
+      size_++;
+    }
+
+    ptrdiff_t top() const {
+      return (size_ <= kPreallocatedLinks) ? links_[size_ - 1]
+                                           : links_extended_->top();
+    }
+
+    void pop() {
+      size_--;
+      if (size_ >= kPreallocatedLinks) {
+        links_extended_->pop();
+        VIXL_ASSERT(size_ == (links_extended_->size() + kPreallocatedLinks));
+      }
+    }
+
+   private:
+    static const size_t kPreallocatedLinks = 4;
+
+    size_t size_;
+    ptrdiff_t links_[kPreallocatedLinks];
+    std::stack<ptrdiff_t> * links_extended_;
+  };
+
+  inline ptrdiff_t location() const { return location_; }
+
+  inline void Bind(ptrdiff_t location) {
+    // Labels can only be bound once.
+    VIXL_ASSERT(!IsBound());
+    location_ = location;
+  }
+
+  inline void AddLink(ptrdiff_t instruction) {
+    // If a label is bound, the assembler already has the information it needs
+    // to write the instruction, so there is no need to add it to links_.
+    VIXL_ASSERT(!IsBound());
+    links_.push(instruction);
+  }
+
+  inline ptrdiff_t GetAndRemoveNextLink() {
+    VIXL_ASSERT(IsLinked());
+    ptrdiff_t link = links_.top();
+    links_.pop();
+    return link;
+  }
+
+  // The offsets of the instructions that have linked to this label.
+  LinksStack links_;
+  // The label location.
+  ptrdiff_t location_;
+
+  static const ptrdiff_t kLocationUnbound = -1;
+
+  // It is not safe to copy labels, so disable the copy constructor by declaring
+  // it private (without an implementation).
+  Label(const Label&);
+
+  // The Assembler class is responsible for binding and linking labels, since
+  // the stored offsets need to be consistent with the Assembler's buffer.
   friend class Assembler;
 };
 
@@ -635,10 +709,49 @@ class Literal {
 };
 
 
+// Control whether or not position-independent code should be emitted.
+enum PositionIndependentCodeOption {
+  // All code generated will be position-independent; all branches and
+  // references to labels generated with the Label class will use PC-relative
+  // addressing.
+  PositionIndependentCode,
+
+  // Allow VIXL to generate code that refers to absolute addresses. With this
+  // option, it will not be possible to copy the code buffer and run it from a
+  // different address; code must be generated in its final location.
+  PositionDependentCode,
+
+  // Allow VIXL to assume that the bottom 12 bits of the address will be
+  // constant, but that the top 48 bits may change. This allows `adrp` to
+  // function in systems which copy code between pages, but otherwise maintain
+  // 4KB page alignment.
+  PageOffsetDependentCode
+};
+
+
+// Control how scaled- and unscaled-offset loads and stores are generated.
+enum LoadStoreScalingOption {
+  // Prefer scaled-immediate-offset instructions, but emit unscaled-offset,
+  // register-offset, pre-index or post-index instructions if necessary.
+  PreferScaledOffset,
+
+  // Prefer unscaled-immediate-offset instructions, but emit scaled-offset,
+  // register-offset, pre-index or post-index instructions if necessary.
+  PreferUnscaledOffset,
+
+  // Require scaled-immediate-offset instructions.
+  RequireScaledOffset,
+
+  // Require unscaled-immediate-offset instructions.
+  RequireUnscaledOffset
+};
+
+
 // Assembler.
 class Assembler {
  public:
-  Assembler(byte* buffer, unsigned buffer_size);
+  Assembler(byte* buffer, unsigned buffer_size,
+            PositionIndependentCodeOption pic = PositionIndependentCode);
 
   // The destructor asserts that one of the following is true:
   //  * The Assembler object has not been used.
@@ -662,12 +775,15 @@ class Assembler {
   // Label.
   // Bind a label to the current PC.
   void bind(Label* label);
-  int UpdateAndGetByteOffsetTo(Label* label);
-  inline int UpdateAndGetInstructionOffsetTo(Label* label) {
-    VIXL_ASSERT(Label::kEndOfChain == 0);
-    return UpdateAndGetByteOffsetTo(label) >> kInstructionSizeLog2;
-  }
 
+  // Return the address of a bound label.
+  template <typename T>
+  inline T GetLabelAddress(const Label * label) {
+    VIXL_ASSERT(label->IsBound());
+    VIXL_STATIC_ASSERT(sizeof(T) >= sizeof(uintptr_t));
+    VIXL_STATIC_ASSERT(sizeof(*buffer_) == 1);
+    return reinterpret_cast<T>(buffer_ + label->location());
+  }
 
   // Instruction set functions.
 
@@ -733,6 +849,12 @@ class Assembler {
   // Calculate the address of a PC offset.
   void adr(const Register& rd, int imm21);
 
+  // Calculate the page address of a label.
+  void adrp(const Register& rd, Label* label);
+
+  // Calculate the page address of a PC offset.
+  void adrp(const Register& rd, int imm21);
+
   // Data Processing instructions.
   // Add.
   void add(const Register& rd,
@@ -1112,31 +1234,76 @@ class Assembler {
 
   // Memory instructions.
   // Load integer or FP register.
-  void ldr(const CPURegister& rt, const MemOperand& src);
+  void ldr(const CPURegister& rt, const MemOperand& src,
+           LoadStoreScalingOption option = PreferScaledOffset);
 
   // Store integer or FP register.
-  void str(const CPURegister& rt, const MemOperand& dst);
+  void str(const CPURegister& rt, const MemOperand& dst,
+           LoadStoreScalingOption option = PreferScaledOffset);
 
   // Load word with sign extension.
-  void ldrsw(const Register& rt, const MemOperand& src);
+  void ldrsw(const Register& rt, const MemOperand& src,
+             LoadStoreScalingOption option = PreferScaledOffset);
 
   // Load byte.
-  void ldrb(const Register& rt, const MemOperand& src);
+  void ldrb(const Register& rt, const MemOperand& src,
+            LoadStoreScalingOption option = PreferScaledOffset);
 
   // Store byte.
-  void strb(const Register& rt, const MemOperand& dst);
+  void strb(const Register& rt, const MemOperand& dst,
+            LoadStoreScalingOption option = PreferScaledOffset);
 
   // Load byte with sign extension.
-  void ldrsb(const Register& rt, const MemOperand& src);
+  void ldrsb(const Register& rt, const MemOperand& src,
+             LoadStoreScalingOption option = PreferScaledOffset);
 
   // Load half-word.
-  void ldrh(const Register& rt, const MemOperand& src);
+  void ldrh(const Register& rt, const MemOperand& src,
+            LoadStoreScalingOption option = PreferScaledOffset);
 
   // Store half-word.
-  void strh(const Register& rt, const MemOperand& dst);
+  void strh(const Register& rt, const MemOperand& dst,
+            LoadStoreScalingOption option = PreferScaledOffset);
 
   // Load half-word with sign extension.
-  void ldrsh(const Register& rt, const MemOperand& src);
+  void ldrsh(const Register& rt, const MemOperand& src,
+             LoadStoreScalingOption option = PreferScaledOffset);
+
+  // Load integer or FP register (with unscaled offset).
+  void ldur(const CPURegister& rt, const MemOperand& src,
+            LoadStoreScalingOption option = PreferUnscaledOffset);
+
+  // Store integer or FP register (with unscaled offset).
+  void stur(const CPURegister& rt, const MemOperand& src,
+            LoadStoreScalingOption option = PreferUnscaledOffset);
+
+  // Load word with sign extension.
+  void ldursw(const Register& rt, const MemOperand& src,
+              LoadStoreScalingOption option = PreferUnscaledOffset);
+
+  // Load byte (with unscaled offset).
+  void ldurb(const Register& rt, const MemOperand& src,
+             LoadStoreScalingOption option = PreferUnscaledOffset);
+
+  // Store byte (with unscaled offset).
+  void sturb(const Register& rt, const MemOperand& dst,
+             LoadStoreScalingOption option = PreferUnscaledOffset);
+
+  // Load byte with sign extension (and unscaled offset).
+  void ldursb(const Register& rt, const MemOperand& src,
+              LoadStoreScalingOption option = PreferUnscaledOffset);
+
+  // Load half-word (with unscaled offset).
+  void ldurh(const Register& rt, const MemOperand& src,
+             LoadStoreScalingOption option = PreferUnscaledOffset);
+
+  // Store half-word (with unscaled offset).
+  void sturh(const Register& rt, const MemOperand& dst,
+             LoadStoreScalingOption option = PreferUnscaledOffset);
+
+  // Load half-word with sign extension (and unscaled offset).
+  void ldursh(const Register& rt, const MemOperand& src,
+              LoadStoreScalingOption option = PreferUnscaledOffset);
 
   // Load integer or FP register pair.
   void ldp(const CPURegister& rt, const CPURegister& rt2,
@@ -1166,6 +1333,79 @@ class Assembler {
   // Load single precision floating point literal to FP register.
   void ldr(const FPRegister& ft, float imm);
 
+  // Store exclusive byte.
+  void stxrb(const Register& rs, const Register& rt, const MemOperand& dst);
+
+  // Store exclusive half-word.
+  void stxrh(const Register& rs, const Register& rt, const MemOperand& dst);
+
+  // Store exclusive register.
+  void stxr(const Register& rs, const Register& rt, const MemOperand& dst);
+
+  // Load exclusive byte.
+  void ldxrb(const Register& rt, const MemOperand& src);
+
+  // Load exclusive half-word.
+  void ldxrh(const Register& rt, const MemOperand& src);
+
+  // Load exclusive register.
+  void ldxr(const Register& rt, const MemOperand& src);
+
+  // Store exclusive register pair.
+  void stxp(const Register& rs,
+            const Register& rt,
+            const Register& rt2,
+            const MemOperand& dst);
+
+  // Load exclusive register pair.
+  void ldxp(const Register& rt, const Register& rt2, const MemOperand& src);
+
+  // Store-release exclusive byte.
+  void stlxrb(const Register& rs, const Register& rt, const MemOperand& dst);
+
+  // Store-release exclusive half-word.
+  void stlxrh(const Register& rs, const Register& rt, const MemOperand& dst);
+
+  // Store-release exclusive register.
+  void stlxr(const Register& rs, const Register& rt, const MemOperand& dst);
+
+  // Load-acquire exclusive byte.
+  void ldaxrb(const Register& rt, const MemOperand& src);
+
+  // Load-acquire exclusive half-word.
+  void ldaxrh(const Register& rt, const MemOperand& src);
+
+  // Load-acquire exclusive register.
+  void ldaxr(const Register& rt, const MemOperand& src);
+
+  // Store-release exclusive register pair.
+  void stlxp(const Register& rs,
+             const Register& rt,
+             const Register& rt2,
+             const MemOperand& dst);
+
+  // Load-acquire exclusive register pair.
+  void ldaxp(const Register& rt, const Register& rt2, const MemOperand& src);
+
+  // Store-release byte.
+  void stlrb(const Register& rt, const MemOperand& dst);
+
+  // Store-release half-word.
+  void stlrh(const Register& rt, const MemOperand& dst);
+
+  // Store-release register.
+  void stlr(const Register& rt, const MemOperand& dst);
+
+  // Load-acquire byte.
+  void ldarb(const Register& rt, const MemOperand& src);
+
+  // Load-acquire half-word.
+  void ldarh(const Register& rt, const MemOperand& src);
+
+  // Load-acquire register.
+  void ldar(const Register& rt, const MemOperand& src);
+
+
   // Move instructions. The default shift of -1 indicates that the move
   // instruction will calculate an appropriate 16-bit immediate and left shift
   // that is equal to the 64-bit immediate argument. If an explicit left shift
@@ -1214,6 +1454,9 @@ class Assembler {
   // System hint.
   void hint(SystemHint code);
 
+  // Clear exclusive monitor.
+  void clrex(int imm4 = 0xf);
+
   // Data memory barrier.
   void dmb(BarrierDomain domain, BarrierType type);
 
@@ -1429,6 +1672,11 @@ class Assembler {
     return rt2.code() << Rt2_offset;
   }
 
+  static Instr Rs(CPURegister rs) {
+    VIXL_ASSERT(rs.code() != kSPRegInternalCode);
+    return rs.code() << Rs_offset;
+  }
+
   // These encoding functions allow the stack pointer to be encoded, and
   // disallow the zero register.
   static Instr RdSP(Register rd) {
@@ -1619,6 +1867,11 @@ class Assembler {
     return imm7 << ImmHint_offset;
   }
 
+  static Instr CRm(int imm4) {
+    VIXL_ASSERT(is_uint4(imm4));
+    return imm4 << CRm_offset;
+  }
+
   static Instr ImmBarrierDomain(int imm2) {
     VIXL_ASSERT(is_uint2(imm2));
     return imm2 << ImmBarrierDomain_offset;
@@ -1660,16 +1913,20 @@ class Assembler {
   }
 
   // Size of the code generated in bytes
-  uint64_t SizeOfCodeGenerated() const {
+  size_t SizeOfCodeGenerated() const {
     VIXL_ASSERT((pc_ >= buffer_) && (pc_ < (buffer_ + buffer_size_)));
     return pc_ - buffer_;
   }
 
   // Size of the code generated since label to the current position.
-  uint64_t SizeOfCodeGeneratedSince(Label* label) const {
+  size_t SizeOfCodeGeneratedSince(Label* label) const {
+    size_t pc_offset = SizeOfCodeGenerated();
+
     VIXL_ASSERT(label->IsBound());
-    VIXL_ASSERT((pc_ >= label->target()) && (pc_ < (buffer_ + buffer_size_)));
-    return pc_ - label->target();
+    VIXL_ASSERT(pc_offset >= static_cast<size_t>(label->location()));
+    VIXL_ASSERT(pc_offset < buffer_size_);
+
+    return pc_offset - label->location();
   }
 
 
@@ -1693,6 +1950,15 @@ class Assembler {
   void EmitLiteralPool(LiteralPoolEmitOption option = NoJumpRequired);
   size_t LiteralPoolSize();
 
+  inline PositionIndependentCodeOption pic() {
+    return pic_;
+  }
+
+  inline bool AllowPageOffsetDependentCode() {
+    return (pic() == PageOffsetDependentCode) ||
+           (pic() == PositionDependentCode);
+  }
+
  protected:
   inline const Register& AppropriateZeroRegFor(const CPURegister& reg) const {
     return reg.Is64Bits() ? xzr : wzr;
@@ -1701,7 +1967,8 @@ class Assembler {
 
   void LoadStore(const CPURegister& rt,
                  const MemOperand& addr,
-                 LoadStoreOp op);
+                 LoadStoreOp op,
+                 LoadStoreScalingOption option = PreferScaledOffset);
   static bool IsImmLSUnscaled(ptrdiff_t offset);
   static bool IsImmLSScaled(ptrdiff_t offset, LSDataSize size);
 
@@ -1717,9 +1984,9 @@ class Assembler {
                         LogicalOp op);
   static bool IsImmLogical(uint64_t value,
                            unsigned width,
-                           unsigned* n,
-                           unsigned* imm_s,
-                           unsigned* imm_r);
+                           unsigned* n = NULL,
+                           unsigned* imm_s = NULL,
+                           unsigned* imm_r = NULL);
 
   void ConditionalCompare(const Register& rn,
                           const Operand& operand,
@@ -1823,6 +2090,17 @@ class Assembler {
 
   void RecordLiteral(int64_t imm, unsigned size);
 
+  // Link the current (not-yet-emitted) instruction to the specified label, then
+  // return an offset to be encoded in the instruction. If the label is not yet
+  // bound, an offset of 0 is returned.
+  ptrdiff_t LinkAndGetByteOffsetTo(Label * label);
+  ptrdiff_t LinkAndGetInstructionOffsetTo(Label * label);
+  ptrdiff_t LinkAndGetPageOffsetTo(Label * label);
+
+  // A common implementation for the LinkAndGet<Type>OffsetTo helpers.
+  template <int element_size>
+  ptrdiff_t LinkAndGetOffsetTo(Label* label);
+
   // Emit the instruction at pc_.
   void Emit(Instr instruction) {
     VIXL_STATIC_ASSERT(sizeof(*pc_) == 1);
@@ -1864,12 +2142,15 @@ class Assembler {
   // The buffer into which code and relocation info are generated.
   Instruction* buffer_;
   // Buffer size, in bytes.
-  unsigned buffer_size_;
+  size_t buffer_size_;
   Instruction* pc_;
   std::list<Literal*> literals_;
   Instruction* next_literal_pool_check_;
   unsigned literal_pool_monitor_;
 
+  PositionIndependentCodeOption pic_;
+
+  friend class Label;
   friend class BlockLiteralPoolScope;
 
 #ifdef DEBUG

disas/libvixl/a64/constants-a64.h

@@ -46,13 +46,13 @@ R(24) R(25) R(26) R(27) R(28) R(29) R(30) R(31)
 
 #define INSTRUCTION_FIELDS_LIST(V_)                                            \
 /* Register fields */                                                          \
-V_(Rd, 4, 0, Bits)                        /* Destination register.     */      \
-V_(Rn, 9, 5, Bits)                        /* First source register.    */      \
-V_(Rm, 20, 16, Bits)                      /* Second source register.   */      \
-V_(Ra, 14, 10, Bits)                      /* Third source register.    */      \
-V_(Rt, 4, 0, Bits)                        /* Load dest / store source. */      \
-V_(Rt2, 14, 10, Bits)                     /* Load second dest /        */      \
-                                         /* store second source.      */       \
+V_(Rd, 4, 0, Bits)                        /* Destination register.        */   \
+V_(Rn, 9, 5, Bits)                        /* First source register.       */   \
+V_(Rm, 20, 16, Bits)                      /* Second source register.      */   \
+V_(Ra, 14, 10, Bits)                      /* Third source register.       */   \
+V_(Rt, 4, 0, Bits)                        /* Load/store register.         */   \
+V_(Rt2, 14, 10, Bits)                     /* Load/store second register.  */   \
+V_(Rs, 20, 16, Bits)                      /* Exclusive access status.     */   \
 V_(PrefetchMode, 4, 0, Bits)                                                   \
                                                                                \
 /* Common bits */                                                              \
@@ -126,6 +126,13 @@ V_(SysOp1, 18, 16, Bits)                                                       \
 V_(SysOp2, 7, 5, Bits)                                                         \
 V_(CRn, 15, 12, Bits)                                                          \
 V_(CRm, 11, 8, Bits)                                                           \
+                                                                               \
+/* Load-/store-exclusive */                                                    \
+V_(LdStXLoad, 22, 22, Bits)                                                    \
+V_(LdStXNotExclusive, 23, 23, Bits)                                            \
+V_(LdStXAcquireRelease, 15, 15, Bits)                                          \
+V_(LdStXSizeLog2, 31, 30, Bits)                                                \
+V_(LdStXPair, 21, 21, Bits)                                                    \
 
 
 #define SYSTEM_REGISTER_FIELDS_LIST(V_, M_)                                    \
@@ -585,6 +592,13 @@ enum MemBarrierOp {
   ISB             = MemBarrierFixed | 0x00000040
 };
 
+enum SystemExclusiveMonitorOp {
+  SystemExclusiveMonitorFixed = 0xD503305F,
+  SystemExclusiveMonitorFMask = 0xFFFFF0FF,
+  SystemExclusiveMonitorMask  = 0xFFFFF0FF,
+  CLREX                       = SystemExclusiveMonitorFixed
+};
+
 // Any load or store.
 enum LoadStoreAnyOp {
   LoadStoreAnyFMask = 0x0a000000,
@@ -702,7 +716,7 @@ enum LoadStoreUnscaledOffsetOp {
 
 // Load/store (post, pre, offset and unsigned.)
 enum LoadStoreOp {
-  LoadStoreOpMask   = 0xC4C00000,
+  LoadStoreOpMask = 0xC4C00000,
   #define LOAD_STORE(A, B, C, D)  \
   A##B##_##C = D
   LOAD_STORE_OP_LIST(LOAD_STORE),
@@ -756,6 +770,44 @@ enum LoadStoreRegisterOffset {
   #undef LOAD_STORE_REGISTER_OFFSET
 };
 
+enum LoadStoreExclusive {
+  LoadStoreExclusiveFixed = 0x08000000,
+  LoadStoreExclusiveFMask = 0x3F000000,
+  LoadStoreExclusiveMask  = 0xFFE08000,
+  STXRB_w  = LoadStoreExclusiveFixed | 0x00000000,
+  STXRH_w  = LoadStoreExclusiveFixed | 0x40000000,
+  STXR_w   = LoadStoreExclusiveFixed | 0x80000000,
+  STXR_x   = LoadStoreExclusiveFixed | 0xC0000000,
+  LDXRB_w  = LoadStoreExclusiveFixed | 0x00400000,
+  LDXRH_w  = LoadStoreExclusiveFixed | 0x40400000,
+  LDXR_w   = LoadStoreExclusiveFixed | 0x80400000,
+  LDXR_x   = LoadStoreExclusiveFixed | 0xC0400000,
+  STXP_w   = LoadStoreExclusiveFixed | 0x80200000,
+  STXP_x   = LoadStoreExclusiveFixed | 0xC0200000,
+  LDXP_w   = LoadStoreExclusiveFixed | 0x80600000,
+  LDXP_x   = LoadStoreExclusiveFixed | 0xC0600000,
+  STLXRB_w = LoadStoreExclusiveFixed | 0x00008000,
+  STLXRH_w = LoadStoreExclusiveFixed | 0x40008000,
+  STLXR_w  = LoadStoreExclusiveFixed | 0x80008000,
+  STLXR_x  = LoadStoreExclusiveFixed | 0xC0008000,
+  LDAXRB_w = LoadStoreExclusiveFixed | 0x00408000,
+  LDAXRH_w = LoadStoreExclusiveFixed | 0x40408000,
+  LDAXR_w  = LoadStoreExclusiveFixed | 0x80408000,
+  LDAXR_x  = LoadStoreExclusiveFixed | 0xC0408000,
+  STLXP_w  = LoadStoreExclusiveFixed | 0x80208000,
+  STLXP_x  = LoadStoreExclusiveFixed | 0xC0208000,
+  LDAXP_w  = LoadStoreExclusiveFixed | 0x80608000,
+  LDAXP_x  = LoadStoreExclusiveFixed | 0xC0608000,
+  STLRB_w  = LoadStoreExclusiveFixed | 0x00808000,
+  STLRH_w  = LoadStoreExclusiveFixed | 0x40808000,
+  STLR_w   = LoadStoreExclusiveFixed | 0x80808000,
+  STLR_x   = LoadStoreExclusiveFixed | 0xC0808000,
+  LDARB_w  = LoadStoreExclusiveFixed | 0x00C08000,
+  LDARH_w  = LoadStoreExclusiveFixed | 0x40C08000,
+  LDAR_w   = LoadStoreExclusiveFixed | 0x80C08000,
+  LDAR_x   = LoadStoreExclusiveFixed | 0xC0C08000
+};
+
 // Conditional compare.
 enum ConditionalCompareOp {
   ConditionalCompareMask = 0x60000000,

disas/libvixl/a64/cpu-a64.h

@@ -28,6 +28,7 @@
 #define VIXL_CPU_A64_H
 
 #include "globals.h"
+#include "instructions-a64.h"
 
 namespace vixl {
 
@@ -42,6 +43,32 @@ class CPU {
   // safely run.
   static void EnsureIAndDCacheCoherency(void *address, size_t length);
 
+  // Handle tagged pointers.
+  template <typename T>
+  static T SetPointerTag(T pointer, uint64_t tag) {
+    VIXL_ASSERT(is_uintn(kAddressTagWidth, tag));
+
+    // Use C-style casts to get static_cast behaviour for integral types (T),
+    // and reinterpret_cast behaviour for other types.
+
+    uint64_t raw = (uint64_t)pointer;
+    VIXL_STATIC_ASSERT(sizeof(pointer) == sizeof(raw));
+
+    raw = (raw & ~kAddressTagMask) | (tag << kAddressTagOffset);
+    return (T)raw;
+  }
+
+  template <typename T>
+  static uint64_t GetPointerTag(T pointer) {
+    // Use C-style casts to get static_cast behaviour for integral types (T),
+    // and reinterpret_cast behaviour for other types.
+
+    uint64_t raw = (uint64_t)pointer;
+    VIXL_STATIC_ASSERT(sizeof(pointer) == sizeof(raw));
+
+    return (raw & kAddressTagMask) >> kAddressTagOffset;
+  }
+
  private:
   // Return the content of the cache type register.
   static uint32_t GetCacheType();

disas/libvixl/a64/decoder-a64.cc

@@ -171,9 +171,9 @@ void Decoder::DecodePCRelAddressing(Instruction* instr) {
 
 void Decoder::DecodeBranchSystemException(Instruction* instr) {
   VIXL_ASSERT((instr->Bits(27, 24) == 0x4) ||
-         (instr->Bits(27, 24) == 0x5) ||
-         (instr->Bits(27, 24) == 0x6) ||
-         (instr->Bits(27, 24) == 0x7) );
+              (instr->Bits(27, 24) == 0x5) ||
+              (instr->Bits(27, 24) == 0x6) ||
+              (instr->Bits(27, 24) == 0x7) );
 
   switch (instr->Bits(31, 29)) {
     case 0:
@@ -272,16 +272,15 @@ void Decoder::DecodeBranchSystemException(Instruction* instr) {
 
 void Decoder::DecodeLoadStore(Instruction* instr) {
   VIXL_ASSERT((instr->Bits(27, 24) == 0x8) ||
-         (instr->Bits(27, 24) == 0x9) ||
-         (instr->Bits(27, 24) == 0xC) ||
-         (instr->Bits(27, 24) == 0xD) );
+              (instr->Bits(27, 24) == 0x9) ||
+              (instr->Bits(27, 24) == 0xC) ||
+              (instr->Bits(27, 24) == 0xD) );
 
   if (instr->Bit(24) == 0) {
     if (instr->Bit(28) == 0) {
       if (instr->Bit(29) == 0) {
         if (instr->Bit(26) == 0) {
-          // TODO: VisitLoadStoreExclusive.
-          VisitUnimplemented(instr);
+          VisitLoadStoreExclusive(instr);
         } else {
           DecodeAdvSIMDLoadStore(instr);
         }

disas/libvixl/a64/decoder-a64.h

@@ -59,6 +59,7 @@
   V(LoadStorePreIndex)              \
   V(LoadStoreRegisterOffset)        \
   V(LoadStoreUnsignedOffset)        \
+  V(LoadStoreExclusive)             \
   V(LogicalShifted)                 \
   V(AddSubShifted)                  \
   V(AddSubExtended)                 \

disas/libvixl/a64/disasm-a64.cc

@@ -24,6 +24,7 @@
 // OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 // OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 
+#include <cstdlib>
 #include "a64/disasm-a64.h"
 
 namespace vixl {
@@ -529,7 +530,7 @@ void Disassembler::VisitExtract(Instruction* instr) {
 void Disassembler::VisitPCRelAddressing(Instruction* instr) {
   switch (instr->Mask(PCRelAddressingMask)) {
     case ADR: Format(instr, "adr", "'Xd, 'AddrPCRelByte"); break;
-    // ADRP is not implemented.
+    case ADRP: Format(instr, "adrp", "'Xd, 'AddrPCRelPage"); break;
     default: Format(instr, "unimplemented", "(PCRelAddressing)");
   }
 }
@@ -943,6 +944,49 @@ void Disassembler::VisitLoadStorePairNonTemporal(Instruction* instr) {
 }
 
 
+void Disassembler::VisitLoadStoreExclusive(Instruction* instr) {
+  const char *mnemonic = "unimplemented";
+  const char *form;
+
+  switch (instr->Mask(LoadStoreExclusiveMask)) {
+    case STXRB_w: mnemonic = "stxrb"; form = "'Ws, 'Wt, ['Xns]"; break;
+    case STXRH_w: mnemonic = "stxrh"; form = "'Ws, 'Wt, ['Xns]"; break;
+    case STXR_w: mnemonic = "stxr"; form = "'Ws, 'Wt, ['Xns]"; break;
+    case STXR_x: mnemonic = "stxr"; form = "'Ws, 'Xt, ['Xns]"; break;
+    case LDXRB_w: mnemonic = "ldxrb"; form = "'Wt, ['Xns]"; break;
+    case LDXRH_w: mnemonic = "ldxrh"; form = "'Wt, ['Xns]"; break;
+    case LDXR_w: mnemonic = "ldxr"; form = "'Wt, ['Xns]"; break;
+    case LDXR_x: mnemonic = "ldxr"; form = "'Xt, ['Xns]"; break;
+    case STXP_w: mnemonic = "stxp"; form = "'Ws, 'Wt, 'Wt2, ['Xns]"; break;
+    case STXP_x: mnemonic = "stxp"; form = "'Ws, 'Xt, 'Xt2, ['Xns]"; break;
+    case LDXP_w: mnemonic = "ldxp"; form = "'Wt, 'Wt2, ['Xns]"; break;
+    case LDXP_x: mnemonic = "ldxp"; form = "'Xt, 'Xt2, ['Xns]"; break;
+    case STLXRB_w: mnemonic = "stlxrb"; form = "'Ws, 'Wt, ['Xns]"; break;
+    case STLXRH_w: mnemonic = "stlxrh"; form = "'Ws, 'Wt, ['Xns]"; break;
+    case STLXR_w: mnemonic = "stlxr"; form = "'Ws, 'Wt, ['Xns]"; break;
+    case STLXR_x: mnemonic = "stlxr"; form = "'Ws, 'Xt, ['Xns]"; break;
+    case LDAXRB_w: mnemonic = "ldaxrb"; form = "'Wt, ['Xns]"; break;
+    case LDAXRH_w: mnemonic = "ldaxrh"; form = "'Wt, ['Xns]"; break;
+    case LDAXR_w: mnemonic = "ldaxr"; form = "'Wt, ['Xns]"; break;
+    case LDAXR_x: mnemonic = "ldaxr"; form = "'Xt, ['Xns]"; break;
+    case STLXP_w: mnemonic = "stlxp"; form = "'Ws, 'Wt, 'Wt2, ['Xns]"; break;
+    case STLXP_x: mnemonic = "stlxp"; form = "'Ws, 'Xt, 'Xt2, ['Xns]"; break;
+    case LDAXP_w: mnemonic = "ldaxp"; form = "'Wt, 'Wt2, ['Xns]"; break;
+    case LDAXP_x: mnemonic = "ldaxp"; form = "'Xt, 'Xt2, ['Xns]"; break;
+    case STLRB_w: mnemonic = "stlrb"; form = "'Wt, ['Xns]"; break;
+    case STLRH_w: mnemonic = "stlrh"; form = "'Wt, ['Xns]"; break;
+    case STLR_w: mnemonic = "stlr"; form = "'Wt, ['Xns]"; break;
+    case STLR_x: mnemonic = "stlr"; form = "'Xt, ['Xns]"; break;
+    case LDARB_w: mnemonic = "ldarb"; form = "'Wt, ['Xns]"; break;
+    case LDARH_w: mnemonic = "ldarh"; form = "'Wt, ['Xns]"; break;
+    case LDAR_w: mnemonic = "ldar"; form = "'Wt, ['Xns]"; break;
+    case LDAR_x: mnemonic = "ldar"; form = "'Xt, ['Xns]"; break;
+    default: form = "(LoadStoreExclusive)";
+  }
+  Format(instr, mnemonic, form);
+}
+
+
 void Disassembler::VisitFPCompare(Instruction* instr) {
   const char *mnemonic = "unimplemented";
   const char *form = "'Fn, 'Fm";
@@ -1162,7 +1206,15 @@ void Disassembler::VisitSystem(Instruction* instr) {
   const char *mnemonic = "unimplemented";
   const char *form = "(System)";
 
-  if (instr->Mask(SystemSysRegFMask) == SystemSysRegFixed) {
+  if (instr->Mask(SystemExclusiveMonitorFMask) == SystemExclusiveMonitorFixed) {
+    switch (instr->Mask(SystemExclusiveMonitorMask)) {
+      case CLREX: {
+        mnemonic = "clrex";
+        form = (instr->CRm() == 0xf) ? NULL : "'IX";
+        break;
+      }
+    }
+  } else if (instr->Mask(SystemSysRegFMask) == SystemSysRegFixed) {
     switch (instr->Mask(SystemSysRegMask)) {
       case MRS: {
         mnemonic = "mrs";
@@ -1184,7 +1236,6 @@ void Disassembler::VisitSystem(Instruction* instr) {
       }
     }
   } else if (instr->Mask(SystemHintFMask) == SystemHintFixed) {
-    VIXL_ASSERT(instr->Mask(SystemHintMask) == HINT);
     switch (instr->ImmHint()) {
       case NOP: {
         mnemonic = "nop";
@@ -1312,6 +1363,7 @@ int Disassembler::SubstituteRegisterField(Instruction* instr,
     case 'n': reg_num = instr->Rn(); break;
     case 'm': reg_num = instr->Rm(); break;
     case 'a': reg_num = instr->Ra(); break;
+    case 's': reg_num = instr->Rs(); break;
     case 't': {
       if (format[2] == '2') {
         reg_num = instr->Rt2();
@@ -1369,7 +1421,7 @@ int Disassembler::SubstituteImmediateField(Instruction* instr,
         VIXL_ASSERT(format[5] == 'L');
         AppendToOutput("#0x%" PRIx64, instr->ImmMoveWide());
         if (instr->ShiftMoveWide() > 0) {
-          AppendToOutput(", lsl #%d", 16 * instr->ShiftMoveWide());
+          AppendToOutput(", lsl #%" PRId64, 16 * instr->ShiftMoveWide());
         }
       }
       return 8;
@@ -1418,7 +1470,7 @@ int Disassembler::SubstituteImmediateField(Instruction* instr,
     }
     case 'F': {  // IFPSingle, IFPDouble or IFPFBits.
       if (format[3] == 'F') {  // IFPFbits.
-        AppendToOutput("#%d", 64 - instr->FPScale());
+        AppendToOutput("#%" PRId64, 64 - instr->FPScale());
         return 8;
       } else {
         AppendToOutput("#0x%" PRIx64 " (%.4f)", instr->ImmFP(),
@@ -1439,25 +1491,29 @@ int Disassembler::SubstituteImmediateField(Instruction* instr,
       return 5;
     }
     case 'P': {  // IP - Conditional compare.
-      AppendToOutput("#%d", instr->ImmCondCmp());
+      AppendToOutput("#%" PRId64, instr->ImmCondCmp());
       return 2;
     }
     case 'B': {  // Bitfields.
       return SubstituteBitfieldImmediateField(instr, format);
     }
     case 'E': {  // IExtract.
-      AppendToOutput("#%d", instr->ImmS());
+      AppendToOutput("#%" PRId64, instr->ImmS());
       return 8;
     }
     case 'S': {  // IS - Test and branch bit.
-      AppendToOutput("#%d", (instr->ImmTestBranchBit5() << 5) |
-                            instr->ImmTestBranchBit40());
+      AppendToOutput("#%" PRId64, (instr->ImmTestBranchBit5() << 5) |
+                                  instr->ImmTestBranchBit40());
       return 2;
     }
     case 'D': {  // IDebug - HLT and BRK instructions.
-      AppendToOutput("#0x%x", instr->ImmException());
+      AppendToOutput("#0x%" PRIx64, instr->ImmException());
       return 6;
     }
+    case 'X': {  // IX - CLREX instruction.
+      AppendToOutput("#0x%" PRIx64, instr->CRm());
+      return 2;
+    }
     default: {
       VIXL_UNIMPLEMENTED();
       return 0;
@@ -1564,21 +1620,20 @@ int Disassembler::SubstituteConditionField(Instruction* instr,
 
 int Disassembler::SubstitutePCRelAddressField(Instruction* instr,
                                               const char* format) {
-  USE(format);
-  VIXL_ASSERT(strncmp(format, "AddrPCRel", 9) == 0);
+  VIXL_ASSERT((strcmp(format, "AddrPCRelByte") == 0) ||   // Used by `adr`.
+              (strcmp(format, "AddrPCRelPage") == 0));    // Used by `adrp`.
 
-  int offset = instr->ImmPCRel();
+  int64_t offset = instr->ImmPCRel();
+  Instruction * base = instr;
 
-  // Only ADR (AddrPCRelByte) is supported.
-  VIXL_ASSERT(strcmp(format, "AddrPCRelByte") == 0);
-
-  char sign = '+';
-  if (offset < 0) {
-    offset = -offset;
-    sign = '-';
+  if (format[9] == 'P') {
+    offset *= kPageSize;
+    base = AlignDown(base, kPageSize);
   }
-  VIXL_STATIC_ASSERT(sizeof(*instr) == 1);
-  AppendToOutput("#%c0x%x (addr %p)", sign, offset, instr + offset);
+
+  char sign = (offset < 0) ? '-' : '+';
+  void * target = reinterpret_cast<void *>(base + offset);
+  AppendToOutput("#%c0x%" PRIx64 " (addr %p)", sign, std::abs(offset), target);
   return 13;
 }
 
@@ -1606,7 +1661,8 @@ int Disassembler::SubstituteBranchTargetField(Instruction* instr,
     sign = '-';
   }
   VIXL_STATIC_ASSERT(sizeof(*instr) == 1);
-  AppendToOutput("#%c0x%" PRIx64 " (addr %p)", sign, offset, instr + offset);
+  void * address = reinterpret_cast<void *>(instr + offset);
+  AppendToOutput("#%c0x%" PRIx64 " (addr %p)", sign, offset, address);
   return 8;
 }
 
@@ -1626,12 +1682,12 @@ int Disassembler::SubstituteExtendField(Instruction* instr,
       (((instr->ExtendMode() == UXTW) && (instr->SixtyFourBits() == 0)) ||
        (instr->ExtendMode() == UXTX))) {
     if (instr->ImmExtendShift() > 0) {
-      AppendToOutput(", lsl #%d", instr->ImmExtendShift());
+      AppendToOutput(", lsl #%" PRId64, instr->ImmExtendShift());
     }
   } else {
     AppendToOutput(", %s", extend_mode[instr->ExtendMode()]);
     if (instr->ImmExtendShift() > 0) {
-      AppendToOutput(" #%d", instr->ImmExtendShift());
+      AppendToOutput(" #%" PRId64, instr->ImmExtendShift());
     }
   }
   return 3;
@@ -1660,7 +1716,7 @@ int Disassembler::SubstituteLSRegOffsetField(Instruction* instr,
   if (!((ext == UXTX) && (shift == 0))) {
     AppendToOutput(", %s", extend_mode[ext]);
     if (shift != 0) {
-      AppendToOutput(" #%d", instr->SizeLS());
+      AppendToOutput(" #%" PRId64, instr->SizeLS());
     }
   }
   return 9;

disas/libvixl/a64/disasm-a64.h

@@ -85,7 +85,7 @@ class Disassembler: public DecoderVisitor {
   bool IsMovzMovnImm(unsigned reg_size, uint64_t value);
 
   void ResetOutput();
-  void AppendToOutput(const char* string, ...);
+  void AppendToOutput(const char* string, ...) PRINTF_CHECK(2, 3);
 
   char* buffer_;
   uint32_t buffer_pos_;

disas/libvixl/a64/instructions-a64.cc

@@ -149,17 +149,24 @@ LSDataSize CalcLSPairDataSize(LoadStorePairOp op) {
 
 
 Instruction* Instruction::ImmPCOffsetTarget() {
+  Instruction * base = this;
   ptrdiff_t offset;
   if (IsPCRelAddressing()) {
-    // PC-relative addressing. Only ADR is supported.
+    // ADR and ADRP.
     offset = ImmPCRel();
+    if (Mask(PCRelAddressingMask) == ADRP) {
+      base = AlignDown(base, kPageSize);
+      offset *= kPageSize;
+    } else {
+      VIXL_ASSERT(Mask(PCRelAddressingMask) == ADR);
+    }
   } else {
     // All PC-relative branches.
     VIXL_ASSERT(BranchType() != UnknownBranchType);
     // Relative branch offsets are instruction-size-aligned.
     offset = ImmBranch() << kInstructionSizeLog2;
   }
-  return this + offset;
+  return base + offset;
 }
 
 
@@ -185,10 +192,16 @@ void Instruction::SetImmPCOffsetTarget(Instruction* target) {
 
 
 void Instruction::SetPCRelImmTarget(Instruction* target) {
-  // ADRP is not supported, so 'this' must point to an ADR instruction.
-  VIXL_ASSERT(Mask(PCRelAddressingMask) == ADR);
-
-  Instr imm = Assembler::ImmPCRelAddress(target - this);
+  int32_t imm21;
+  if ((Mask(PCRelAddressingMask) == ADR)) {
+    imm21 = target - this;
+  } else {
+    VIXL_ASSERT(Mask(PCRelAddressingMask) == ADRP);
+    uintptr_t this_page = reinterpret_cast<uintptr_t>(this) / kPageSize;
+    uintptr_t target_page = reinterpret_cast<uintptr_t>(target) / kPageSize;
+    imm21 = target_page - this_page;
+  }
+  Instr imm = Assembler::ImmPCRelAddress(imm21);
 
   SetInstructionBits(Mask(~ImmPCRel_mask) | imm);
 }

disas/libvixl/a64/instructions-a64.h

@@ -41,6 +41,10 @@ const unsigned kLiteralEntrySize = 4;
 const unsigned kLiteralEntrySizeLog2 = 2;
 const unsigned kMaxLoadLiteralRange = 1 * MBytes;
 
+// This is the nominal page size (as used by the adrp instruction); the actual
+// size of the memory pages allocated by the kernel is likely to differ.
+const unsigned kPageSize = 4 * KBytes;
+
 const unsigned kWRegSize = 32;
 const unsigned kWRegSizeLog2 = 5;
 const unsigned kWRegSizeInBytes = kWRegSize / 8;
@@ -79,6 +83,12 @@ const unsigned kZeroRegCode = 31;
 const unsigned kSPRegInternalCode = 63;
 const unsigned kRegCodeMask = 0x1f;
 
+const unsigned kAddressTagOffset = 56;
+const unsigned kAddressTagWidth = 8;
+const uint64_t kAddressTagMask =
+    ((UINT64_C(1) << kAddressTagWidth) - 1) << kAddressTagOffset;
+VIXL_STATIC_ASSERT(kAddressTagMask == UINT64_C(0xff00000000000000));
+
 // AArch64 floating-point specifics. These match IEEE-754.
 const unsigned kDoubleMantissaBits = 52;
 const unsigned kDoubleExponentBits = 11;

disas/libvixl/platform.h

@@ -28,14 +28,10 @@
 #define PLATFORM_H
 
 // Define platform specific functionalities.
+#include <signal.h>
 
 namespace vixl {
-#ifdef USE_SIMULATOR
-// Currently we assume running the simulator implies running on x86 hardware.
-inline void HostBreakpoint() { asm("int3"); }
-#else
-inline void HostBreakpoint() { asm("brk"); }
-#endif
+inline void HostBreakpoint() { raise(SIGINT); }
 }  // namespace vixl
 
 #endif

disas/libvixl/utils.cc

@@ -124,4 +124,14 @@ int CountSetBits(uint64_t value, int width) {
 
   return value;
 }
+
+
+uint64_t LowestSetBit(uint64_t value) {
+  return value & -value;
+}
+
+
+bool IsPowerOf2(int64_t value) {
+  return (value != 0) && ((value & (value - 1)) == 0);
+}
 }  // namespace vixl

disas/libvixl/utils.h

@@ -33,6 +33,14 @@
 
 namespace vixl {
 
+// Macros for compile-time format checking.
+#if defined(__GNUC__)
+#define PRINTF_CHECK(format_index, varargs_index) \
+  __attribute__((format(printf, format_index, varargs_index)))
+#else
+#define PRINTF_CHECK(format_index, varargs_index)
+#endif
+
 // Check number width.
 inline bool is_intn(unsigned n, int64_t x) {
   VIXL_ASSERT((0 < n) && (n < 64));
@@ -155,6 +163,8 @@ int CountLeadingZeros(uint64_t value, int width);
 int CountLeadingSignBits(int64_t value, int width);
 int CountTrailingZeros(uint64_t value, int width);
 int CountSetBits(uint64_t value, int width);
+uint64_t LowestSetBit(uint64_t value);
+bool IsPowerOf2(int64_t value);
 
 // Pointer alignment
 // TODO: rename/refactor to make it specific to instructions.
@@ -167,21 +177,31 @@ bool IsWordAligned(T pointer) {
 // Increment a pointer until it has the specified alignment.
 template<class T>
 T AlignUp(T pointer, size_t alignment) {
-  VIXL_STATIC_ASSERT(sizeof(pointer) == sizeof(uintptr_t));
-  uintptr_t pointer_raw = reinterpret_cast<uintptr_t>(pointer);
+  // Use C-style casts to get static_cast behaviour for integral types (T), and
+  // reinterpret_cast behaviour for other types.
+
+  uintptr_t pointer_raw = (uintptr_t)pointer;
+  VIXL_STATIC_ASSERT(sizeof(pointer) == sizeof(pointer_raw));
+
   size_t align_step = (alignment - pointer_raw) % alignment;
   VIXL_ASSERT((pointer_raw + align_step) % alignment == 0);
-  return reinterpret_cast<T>(pointer_raw + align_step);
+
+  return (T)(pointer_raw + align_step);
 }
 
 // Decrement a pointer until it has the specified alignment.
 template<class T>
 T AlignDown(T pointer, size_t alignment) {
-  VIXL_STATIC_ASSERT(sizeof(pointer) == sizeof(uintptr_t));
-  uintptr_t pointer_raw = reinterpret_cast<uintptr_t>(pointer);
+  // Use C-style casts to get static_cast behaviour for integral types (T), and
+  // reinterpret_cast behaviour for other types.
+
+  uintptr_t pointer_raw = (uintptr_t)pointer;
+  VIXL_STATIC_ASSERT(sizeof(pointer) == sizeof(pointer_raw));
+
   size_t align_step = pointer_raw % alignment;
   VIXL_ASSERT((pointer_raw - align_step) % alignment == 0);
-  return reinterpret_cast<T>(pointer_raw - align_step);
+
+  return (T)(pointer_raw - align_step);
 }
 
 

dma-helpers.c

@@ -170,6 +170,10 @@ static void dma_bdrv_cb(void *opaque, int ret)
         return;
     }
 
+    if (dbs->iov.size & ~BDRV_SECTOR_MASK) {
+        qemu_iovec_discard_back(&dbs->iov, dbs->iov.size & ~BDRV_SECTOR_MASK);
+    }
+
     dbs->acb = dbs->io_func(dbs->bs, dbs->sector_num, &dbs->iov,
                             dbs->iov.size / 512, dma_bdrv_cb, dbs);
     assert(dbs->acb);

docs/aio_notify.promela

@@ -0,0 +1,104 @@
+/*
+ * This model describes the interaction between aio_set_dispatching()
+ * and aio_notify().
+ *
+ * Author: Paolo Bonzini <pbonzini@redhat.com>
+ *
+ * This file is in the public domain.  If you really want a license,
+ * the WTFPL will do.
+ *
+ * To simulate it:
+ *     spin -p docs/aio_notify.promela
+ *
+ * To verify it:
+ *     spin -a docs/aio_notify.promela
+ *     gcc -O2 pan.c
+ *     ./a.out -a
+ */
+
+#define MAX   4
+#define LAST  (1 << (MAX - 1))
+#define FINAL ((LAST << 1) - 1)
+
+bool dispatching;
+bool event;
+
+int req, done;
+
+active proctype waiter()
+{
+     int fetch, blocking;
+
+     do
+        :: done != FINAL -> {
+            // Computing "blocking" is separate from execution of the
+            // "bottom half"
+            blocking = (req == 0);
+
+            // This is our "bottom half"
+            atomic { fetch = req; req = 0; }
+            done = done | fetch;
+
+            // Wait for a nudge from the other side
+            do
+                :: event == 1 -> { event = 0; break; }
+                :: !blocking  -> break;
+            od;
+
+            dispatching = 1;
+
+            // If you are simulating this model, you may want to add
+            // something like this here:
+            //
+            //      int foo; foo++; foo++; foo++;
+            //
+            // This only wastes some time and makes it more likely
+            // that the notifier process hits the "fast path".
+
+            dispatching = 0;
+        }
+        :: else -> break;
+    od
+}
+
+active proctype notifier()
+{
+    int next = 1;
+    int sets = 0;
+
+    do
+        :: next <= LAST -> {
+            // generate a request
+            req = req | next;
+            next = next << 1;
+
+            // aio_notify
+            if
+                :: dispatching == 0 -> sets++; event = 1;
+                :: else             -> skip;
+            fi;
+
+            // Test both synchronous and asynchronous delivery
+            if
+                :: 1 -> do
+                            :: req == 0 -> break;
+                        od;
+                :: 1 -> skip;
+            fi;
+        }
+        :: else -> break;
+    od;
+    printf("Skipped %d event_notifier_set\n", MAX - sets);
+}
+
+#define p (done == FINAL)
+
+never  {
+    do
+        :: 1                      // after an arbitrarily long prefix
+        :: p -> break             // p becomes true
+    od;
+    do
+        :: !p -> accept: break    // it then must remains true forever after
+    od
+}

docs/image-fuzzer.txt

@@ -0,0 +1,238 @@
+# Specification for the fuzz testing tool
+#
+# Copyright (C) 2014 Maria Kustova <maria.k@catit.be>
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+
+Image fuzzer
+============
+
+Description
+-----------
+
+The goal of the image fuzzer is to catch crashes of qemu-io/qemu-img
+by providing to them randomly corrupted images.
+Test images are generated from scratch and have valid inner structure with some
+elements, e.g. L1/L2 tables, having random invalid values.
+
+
+Test runner
+-----------
+
+The test runner generates test images, executes tests utilizing generated
+images, indicates their results and collects all test related artifacts (logs,
+core dumps, test images, backing files).
+The test means execution of all available commands under test with the same
+generated test image.
+By default, the test runner generates new tests and executes them until
+keyboard interruption. But if a test seed is specified via the '--seed' runner
+parameter, then only one test with this seed will be executed, after its finish
+the runner will exit.
+
+The runner uses an external image fuzzer to generate test images. An image
+generator should be specified as a mandatory parameter of the test runner.
+Details about interactions between the runner and fuzzers see "Module
+interfaces".
+
+The runner activates generation of core dumps during test executions, but it
+assumes that core dumps will be generated in the current working directory.
+For comprehensive test results, please, set up your test environment
+properly.
+
+Paths to binaries under test (SUTs) qemu-img and qemu-io are retrieved from
+environment variables. If the environment check fails the runner will
+use SUTs installed in system paths.
+qemu-img is required for creation of backing files, so it's mandatory to set
+the related environment variable if it's not installed in the system path.
+For details about environment variables see qemu-iotests/check.
+
+The runner accepts a JSON array of fields expected to be fuzzed via the
+'--config' argument, e.g.
+
+       '[["feature_name_table"], ["header", "l1_table_offset"]]'
+
+Each sublist can have one or two strings defining image structure elements.
+In the latter case a parent element should be placed on the first position,
+and a field name on the second one.
+
+The runner accepts a list of commands under test as a JSON array via
+the '--command' argument. Each command is a list containing a SUT and all its
+arguments, e.g.
+
+       runner.py -c '[["qemu-io", "$test_img", "-c", "write $off $len"]]'
+     /tmp/test ../qcow2
+
+For variable arguments next aliases can be used:
+    - $test_img for a fuzzed img
+    - $off for an offset in the fuzzed image
+    - $len for a data size
+
+Values for last two aliases will be generated based on a size of a virtual
+disk of the generated image.
+In case when no commands are specified the runner will execute commands from
+the default list:
+    - qemu-img check
+    - qemu-img info
+    - qemu-img convert
+    - qemu-io -c read
+    - qemu-io -c write
+    - qemu-io -c aio_read
+    - qemu-io -c aio_write
+    - qemu-io -c flush
+    - qemu-io -c discard
+    - qemu-io -c truncate
+
+
+Qcow2 image generator
+---------------------
+
+The 'qcow2' generator is a Python package providing 'create_image' method as
+a single public API. See details in 'Test runner/image fuzzer' chapter of
+'Module interfaces'.
+
+Qcow2 contains two submodules: fuzz.py and layout.py.
+
+'fuzz.py' contains all fuzzing functions, one per image field. It's assumed
+that after code analysis every field will have own constraints for its value.
+For now only universal potentially dangerous values are used, e.g. type limits
+for integers or unsafe symbols as '%s' for strings. For bitmasks random amount
+of bits are set to ones. All fuzzed values are checked on non-equality to the
+current valid value of the field. In case of equality the value will be
+regenerated.
+
+'layout.py' creates a random valid image, fuzzes a random subset of the image
+fields by 'fuzz.py' module and writes a fuzzed image to the file specified.
+If a fuzzer configuration is specified, then it has the next interpretation:
+
+    1. If a list contains a parent image element only, then some random portion
+    of fields of this element will be fuzzed every test.
+    The same behavior is applied for the entire image if no configuration is
+    used. This case is useful for the test specialization.
+
+    2. If a list contains a parent element and a field name, then a field
+    will be always fuzzed for every test. This case is useful for regression
+    testing.
+
+For now only header fields, header extensions and L1/L2 tables are generated.
+
+Module interfaces
+-----------------
+
+* Test runner/image fuzzer
+
+The runner calls an image generator specifying the path to a test image file,
+path to a backing file and its format and a fuzzer configuration.
+An image generator is expected to provide a
+
+   'create_image(test_img_path, backing_file_path=None,
+                 backing_file_format=None, fuzz_config=None)'
+
+method that creates a test image, writes it to the specified file and returns
+the size of the virtual disk.
+The file should be created if it doesn't exist or overwritten otherwise.
+fuzz_config has a form of a list of lists. Every sublist can have one
+or two elements: first element is a name of a parent image element, second one
+if exists is a name of a field in this element.
+Example,
+        [['header', 'l1_table_offset'],
+         ['header', 'nb_snapshots'],
+         ['feature_name_table']]
+
+Random seed is set by the runner at every test execution for the regression
+purpose, so an image generator is not recommended to modify it internally.
+
+
+Overall fuzzer requirements
+===========================
+
+Input data:
+----------
+
+ - image template (generator)
+ - work directory
+ - action vector (optional)
+ - seed (optional)
+ - SUT and its arguments (optional)
+
+
+Fuzzer requirements:
+-------------------
+
+1.  Should be able to inject random data
+2.  Should be able to select a random value from the manually pregenerated
+    vector (boundary values, e.g. max/min cluster size)
+3.  Image template should describe a general structure invariant for all
+    test images (image format description)
+4.  Image template should be autonomous and other fuzzer parts should not
+    rely on it
+5.  Image template should contain reference rules (not only block+size
+    description)
+6.  Should generate the test image with the correct structure based on an image
+    template
+7.  Should accept a seed as an argument (for regression purpose)
+8.  Should generate a seed if it is not specified as an input parameter.
+9.  The same seed should generate the same image for the same action vector,
+    specified or generated.
+10. Should accept a vector of actions as an argument (for test reproducing and
+    for test case specification, e.g. group of tests for header structure,
+    group of test for snapshots, etc)
+11. Action vector should be randomly generated from the pool of available
+    actions, if it is not specified as an input parameter
+12. Pool of actions should be defined automatically based on an image template
+13. Should accept a SUT and its call parameters as an argument or select them
+    randomly otherwise. As far as it's expected to be rarely changed, the list
+    of all possible test commands can be available in the test runner
+    internally.
+14. Should support an external cancellation of a test run
+15. Seed should be logged (for regression purpose)
+16. All files related to a test result should be collected: a test image,
+    SUT logs, fuzzer logs and crash dumps
+17. Should be compatible with python version 2.4-2.7
+18. Usage of external libraries should be limited as much as possible.
+
+
+Image formats:
+-------------
+
+Main target image format is qcow2, but support of image templates should
+provide an ability to add any other image format.
+
+
+Effectiveness:
+-------------
+
+The fuzzer can be controlled via template, seed and action vector;
+it makes the fuzzer itself invariant to an image format and test logic.
+It should be able to perform rather complex and precise tests, that can be
+specified via an action vector. Otherwise, knowledge about an image structure
+allows the fuzzer to generate the pool of all available areas can be fuzzed
+and randomly select some of them and so compose its own action vector.
+Also complexity of a template defines complexity of the fuzzer, so its
+functionality can be varied from simple model-independent fuzzing to smart
+model-based one.
+
+
+Glossary:
+--------
+
+Action vector is a sequence of structure elements retrieved from an image
+format, each of them will be fuzzed for the test image. It's a subset of
+elements of the action pool. Example: header, refcount table, etc.
+Action pool is all available elements of an image structure that generated
+automatically from an image template.
+Image template is a formal description of an image structure and relations
+between image blocks.
+Test image is an output image of the fuzzer defined by the current seed and
+action vector.

docs/memory.txt

@@ -74,11 +74,16 @@ Region lifecycle
 ----------------
 
 A region is created by one of the constructor functions (memory_region_init*())
-and destroyed by the destructor (memory_region_destroy()).  In between,
-a region can be added to an address space by using memory_region_add_subregion()
-and removed using memory_region_del_subregion().  Region attributes may be
-changed at any point; they take effect once the region becomes exposed to the
-guest.
+and attached to an object.  It is then destroyed by object_unparent() or simply
+when the parent object dies.
+
+In between, a region can be added to an address space
+by using memory_region_add_subregion() and removed using
+memory_region_del_subregion().  Destroying the region implicitly
+removes the region from the address space.
+
+Region attributes may be changed at any point; they take effect once
+the region becomes exposed to the guest.
 
 Overlapping regions and priority
 --------------------------------

docs/multiple-iothreads.txt

@@ -0,0 +1,134 @@
+Copyright (c) 2014 Red Hat Inc.
+
+This work is licensed under the terms of the GNU GPL, version 2 or later.  See
+the COPYING file in the top-level directory.
+
+
+This document explains the IOThread feature and how to write code that runs
+outside the QEMU global mutex.
+
+The main loop and IOThreads
+---------------------------
+QEMU is an event-driven program that can do several things at once using an
+event loop.  The VNC server and the QMP monitor are both processed from the
+same event loop, which monitors their file descriptors until they become
+readable and then invokes a callback.
+
+The default event loop is called the main loop (see main-loop.c).  It is
+possible to create additional event loop threads using -object
+iothread,id=my-iothread.
+
+Side note: The main loop and IOThread are both event loops but their code is
+not shared completely.  Sometimes it is useful to remember that although they
+are conceptually similar they are currently not interchangeable.
+
+Why IOThreads are useful
+------------------------
+IOThreads allow the user to control the placement of work.  The main loop is a
+scalability bottleneck on hosts with many CPUs.  Work can be spread across
+several IOThreads instead of just one main loop.  When set up correctly this
+can improve I/O latency and reduce jitter seen by the guest.
+
+The main loop is also deeply associated with the QEMU global mutex, which is a
+scalability bottleneck in itself.  vCPU threads and the main loop use the QEMU
+global mutex to serialize execution of QEMU code.  This mutex is necessary
+because a lot of QEMU's code historically was not thread-safe.
+
+The fact that all I/O processing is done in a single main loop and that the
+QEMU global mutex is contended by all vCPU threads and the main loop explain
+why it is desirable to place work into IOThreads.
+
+The experimental virtio-blk data-plane implementation has been benchmarked and
+shows these effects:
+ftp://public.dhe.ibm.com/linux/pdfs/KVM_Virtualized_IO_Performance_Paper.pdf
+
+How to program for IOThreads
+----------------------------
+The main difference between legacy code and new code that can run in an
+IOThread is dealing explicitly with the event loop object, AioContext
+(see include/block/aio.h).  Code that only works in the main loop
+implicitly uses the main loop's AioContext.  Code that supports running
+in IOThreads must be aware of its AioContext.
+
+AioContext supports the following services:
+ * File descriptor monitoring (read/write/error on POSIX hosts)
+ * Event notifiers (inter-thread signalling)
+ * Timers
+ * Bottom Halves (BH) deferred callbacks
+
+There are several old APIs that use the main loop AioContext:
+ * LEGACY qemu_aio_set_fd_handler() - monitor a file descriptor
+ * LEGACY qemu_aio_set_event_notifier() - monitor an event notifier
+ * LEGACY timer_new_ms() - create a timer
+ * LEGACY qemu_bh_new() - create a BH
+ * LEGACY qemu_aio_wait() - run an event loop iteration
+
+Since they implicitly work on the main loop they cannot be used in code that
+runs in an IOThread.  They might cause a crash or deadlock if called from an
+IOThread since the QEMU global mutex is not held.
+
+Instead, use the AioContext functions directly (see include/block/aio.h):
+ * aio_set_fd_handler() - monitor a file descriptor
+ * aio_set_event_notifier() - monitor an event notifier
+ * aio_timer_new() - create a timer
+ * aio_bh_new() - create a BH
+ * aio_poll() - run an event loop iteration
+
+The AioContext can be obtained from the IOThread using
+iothread_get_aio_context() or for the main loop using qemu_get_aio_context().
+Code that takes an AioContext argument works both in IOThreads or the main
+loop, depending on which AioContext instance the caller passes in.
+
+How to synchronize with an IOThread
+-----------------------------------
+AioContext is not thread-safe so some rules must be followed when using file
+descriptors, event notifiers, timers, or BHs across threads:
+
+1. AioContext functions can be called safely from file descriptor, event
+notifier, timer, or BH callbacks invoked by the AioContext.  No locking is
+necessary.
+
+2. Other threads wishing to access the AioContext must use
+aio_context_acquire()/aio_context_release() for mutual exclusion.  Once the
+context is acquired no other thread can access it or run event loop iterations
+in this AioContext.
+
+aio_context_acquire()/aio_context_release() calls may be nested.  This
+means you can call them if you're not sure whether #1 applies.
+
+There is currently no lock ordering rule if a thread needs to acquire multiple
+AioContexts simultaneously.  Therefore, it is only safe for code holding the
+QEMU global mutex to acquire other AioContexts.
+
+Side note: the best way to schedule a function call across threads is to create
+a BH in the target AioContext beforehand and then call qemu_bh_schedule().  No
+acquire/release or locking is needed for the qemu_bh_schedule() call.  But be
+sure to acquire the AioContext for aio_bh_new() if necessary.
+
+The relationship between AioContext and the block layer
+-------------------------------------------------------
+The AioContext originates from the QEMU block layer because it provides a
+scoped way of running event loop iterations until all work is done.  This
+feature is used to complete all in-flight block I/O requests (see
+bdrv_drain_all()).  Nowadays AioContext is a generic event loop that can be
+used by any QEMU subsystem.
+
+The block layer has support for AioContext integrated.  Each BlockDriverState
+is associated with an AioContext using bdrv_set_aio_context() and
+bdrv_get_aio_context().  This allows block layer code to process I/O inside the
+right AioContext.  Other subsystems may wish to follow a similar approach.
+
+Block layer code must therefore expect to run in an IOThread and avoid using
+old APIs that implicitly use the main loop.  See the "How to program for
+IOThreads" above for information on how to do that.
+
+If main loop code such as a QMP function wishes to access a BlockDriverState it
+must first call aio_context_acquire(bdrv_get_aio_context(bs)) to ensure the
+IOThread does not run in parallel.
+
+Long-running jobs (usually in the form of coroutines) are best scheduled in the
+BlockDriverState's AioContext to avoid the need to acquire/release around each
+bdrv_*() call.  Be aware that there is currently no mechanism to get notified
+when bdrv_set_aio_context() moves this BlockDriverState to a different
+AioContext (see bdrv_detach_aio_context()/bdrv_attach_aio_context()), so you
+may need to add this if you want to support long-running jobs.

docs/qapi-code-gen.txt

@@ -218,10 +218,10 @@ An example command is:
 === Events ===
 
 Events are defined with the keyword 'event'.  When 'data' is also specified,
-additional info will be carried on.  Finally there will be C API generated
-in qapi-event.h; when called by QEMU code, a message with timestamp will
-be emitted on the wire.  If timestamp is -1, it means failure to retrieve host
-time.
+additional info will be included in the event.  Finally there will be C API
+generated in qapi-event.h; when called by QEMU code, a message with timestamp
+will be emitted on the wire.  If timestamp is -1, it means failure to retrieve
+host time.
 
 An example event is:
 

docs/qmp/qmp-events.txt

@@ -0,0 +1,627 @@
+                   QEMU Machine Protocol Events
+                   ============================
+
+ACPI_DEVICE_OST
+---------------
+
+Emitted when guest executes ACPI _OST method.
+
+ - data: ACPIOSTInfo type as described in qapi-schema.json
+
+{ "event": "ACPI_DEVICE_OST",
+     "data": { "device": "d1", "slot": "0", "slot-type": "DIMM", "source": 1, "status": 0 } }
+
+BALLOON_CHANGE
+--------------
+
+Emitted when the guest changes the actual BALLOON level. This
+value is equivalent to the 'actual' field return by the
+'query-balloon' command
+
+Data:
+
+- "actual": actual level of the guest memory balloon in bytes (json-number)
+
+Example:
+
+{ "event": "BALLOON_CHANGE",
+    "data": { "actual": 944766976 },
+    "timestamp": { "seconds": 1267020223, "microseconds": 435656 } }
+
+BLOCK_IMAGE_CORRUPTED
+---------------------
+
+Emitted when a disk image is being marked corrupt.
+
+Data:
+
+- "device": Device name (json-string)
+- "msg":    Informative message (e.g., reason for the corruption) (json-string)
+- "offset": If the corruption resulted from an image access, this is the access
+            offset into the image (json-int)
+- "size":   If the corruption resulted from an image access, this is the access
+            size (json-int)
+
+Example:
+
+{ "event": "BLOCK_IMAGE_CORRUPTED",
+    "data": { "device": "ide0-hd0",
+        "msg": "Prevented active L1 table overwrite", "offset": 196608,
+        "size": 65536 },
+    "timestamp": { "seconds": 1378126126, "microseconds": 966463 } }
+
+BLOCK_IO_ERROR
+--------------
+
+Emitted when a disk I/O error occurs.
+
+Data:
+
+- "device": device name (json-string)
+- "operation": I/O operation (json-string, "read" or "write")
+- "action": action that has been taken, it's one of the following (json-string):
+    "ignore": error has been ignored
+    "report": error has been reported to the device
+    "stop": the VM is going to stop because of the error
+
+Example:
+
+{ "event": "BLOCK_IO_ERROR",
+    "data": { "device": "ide0-hd1",
+              "operation": "write",
+              "action": "stop" },
+    "timestamp": { "seconds": 1265044230, "microseconds": 450486 } }
+
+Note: If action is "stop", a STOP event will eventually follow the
+BLOCK_IO_ERROR event.
+
+BLOCK_JOB_CANCELLED
+-------------------
+
+Emitted when a block job has been cancelled.
+
+Data:
+
+- "type":     Job type (json-string; "stream" for image streaming
+                                     "commit" for block commit)
+- "device":   Device name (json-string)
+- "len":      Maximum progress value (json-int)
+- "offset":   Current progress value (json-int)
+              On success this is equal to len.
+              On failure this is less than len.
+- "speed":    Rate limit, bytes per second (json-int)
+
+Example:
+
+{ "event": "BLOCK_JOB_CANCELLED",
+     "data": { "type": "stream", "device": "virtio-disk0",
+               "len": 10737418240, "offset": 134217728,
+               "speed": 0 },
+     "timestamp": { "seconds": 1267061043, "microseconds": 959568 } }
+
+BLOCK_JOB_COMPLETED
+-------------------
+
+Emitted when a block job has completed.
+
+Data:
+
+- "type":     Job type (json-string; "stream" for image streaming
+                                     "commit" for block commit)
+- "device":   Device name (json-string)
+- "len":      Maximum progress value (json-int)
+- "offset":   Current progress value (json-int)
+              On success this is equal to len.
+              On failure this is less than len.
+- "speed":    Rate limit, bytes per second (json-int)
+- "error":    Error message (json-string, optional)
+              Only present on failure.  This field contains a human-readable
+              error message.  There are no semantics other than that streaming
+              has failed and clients should not try to interpret the error
+              string.
+
+Example:
+
+{ "event": "BLOCK_JOB_COMPLETED",
+     "data": { "type": "stream", "device": "virtio-disk0",
+               "len": 10737418240, "offset": 10737418240,
+               "speed": 0 },
+     "timestamp": { "seconds": 1267061043, "microseconds": 959568 } }
+
+BLOCK_JOB_ERROR
+---------------
+
+Emitted when a block job encounters an error.
+
+Data:
+
+- "device": device name (json-string)
+- "operation": I/O operation (json-string, "read" or "write")
+- "action": action that has been taken, it's one of the following (json-string):
+    "ignore": error has been ignored, the job may fail later
+    "report": error will be reported and the job canceled
+    "stop": error caused job to be paused
+
+Example:
+
+{ "event": "BLOCK_JOB_ERROR",
+    "data": { "device": "ide0-hd1",
+              "operation": "write",
+              "action": "stop" },
+    "timestamp": { "seconds": 1265044230, "microseconds": 450486 } }
+
+BLOCK_JOB_READY
+---------------
+
+Emitted when a block job is ready to complete.
+
+Data:
+
+- "type":     Job type (json-string; "stream" for image streaming
+                                     "commit" for block commit)
+- "device":   Device name (json-string)
+- "len":      Maximum progress value (json-int)
+- "offset":   Current progress value (json-int)
+              On success this is equal to len.
+              On failure this is less than len.
+- "speed":    Rate limit, bytes per second (json-int)
+
+Example:
+
+{ "event": "BLOCK_JOB_READY",
+    "data": { "device": "drive0", "type": "mirror", "speed": 0,
+              "len": 2097152, "offset": 2097152 }
+    "timestamp": { "seconds": 1265044230, "microseconds": 450486 } }
+
+Note: The "ready to complete" status is always reset by a BLOCK_JOB_ERROR
+event.
+
+DEVICE_DELETED
+--------------
+
+Emitted whenever the device removal completion is acknowledged
+by the guest.
+At this point, it's safe to reuse the specified device ID.
+Device removal can be initiated by the guest or by HMP/QMP commands.
+
+Data:
+
+- "device": device name (json-string, optional)
+- "path": device path (json-string)
+
+{ "event": "DEVICE_DELETED",
+  "data": { "device": "virtio-net-pci-0",
+            "path": "/machine/peripheral/virtio-net-pci-0" },
+  "timestamp": { "seconds": 1265044230, "microseconds": 450486 } }
+
+DEVICE_TRAY_MOVED
+-----------------
+
+It's emitted whenever the tray of a removable device is moved by the guest
+or by HMP/QMP commands.
+
+Data:
+
+- "device": device name (json-string)
+- "tray-open": true if the tray has been opened or false if it has been closed
+               (json-bool)
+
+{ "event": "DEVICE_TRAY_MOVED",
+  "data": { "device": "ide1-cd0",
+            "tray-open": true
+  },
+  "timestamp": { "seconds": 1265044230, "microseconds": 450486 } }
+
+GUEST_PANICKED
+--------------
+
+Emitted when guest OS panic is detected.
+
+Data:
+
+- "action": Action that has been taken (json-string, currently always "pause").
+
+Example:
+
+{ "event": "GUEST_PANICKED",
+     "data": { "action": "pause" } }
+
+NIC_RX_FILTER_CHANGED
+---------------------
+
+The event is emitted once until the query command is executed,
+the first event will always be emitted.
+
+Data:
+
+- "name": net client name (json-string)
+- "path": device path (json-string)
+
+{ "event": "NIC_RX_FILTER_CHANGED",
+  "data": { "name": "vnet0",
+            "path": "/machine/peripheral/vnet0/virtio-backend" },
+  "timestamp": { "seconds": 1368697518, "microseconds": 326866 } }
+}
+
+POWERDOWN
+---------
+
+Emitted when the Virtual Machine is powered down through the power
+control system, such as via ACPI.
+
+Data: None.
+
+Example:
+
+{ "event": "POWERDOWN",
+    "timestamp": { "seconds": 1267040730, "microseconds": 682951 } }
+
+QUORUM_FAILURE
+--------------
+
+Emitted by the Quorum block driver if it fails to establish a quorum.
+
+Data:
+
+- "reference":     device name if defined else node name.
+- "sector-num":    Number of the first sector of the failed read operation.
+- "sectors-count": Failed read operation sector count.
+
+Example:
+
+{ "event": "QUORUM_FAILURE",
+     "data": { "reference": "usr1", "sector-num": 345435, "sectors-count": 5 },
+     "timestamp": { "seconds": 1344522075, "microseconds": 745528 } }
+
+QUORUM_REPORT_BAD
+-----------------
+
+Emitted to report a corruption of a Quorum file.
+
+Data:
+
+- "error":         Error message (json-string, optional)
+                   Only present on failure.  This field contains a human-readable
+                   error message.  There are no semantics other than that the
+                   block layer reported an error and clients should not try to
+                   interpret the error string.
+- "node-name":     The graph node name of the block driver state.
+- "sector-num":    Number of the first sector of the failed read operation.
+- "sectors-count": Failed read operation sector count.
+
+Example:
+
+{ "event": "QUORUM_REPORT_BAD",
+     "data": { "node-name": "1.raw", "sector-num": 345435, "sectors-count": 5 },
+     "timestamp": { "seconds": 1344522075, "microseconds": 745528 } }
+
+RESET
+-----
+
+Emitted when the Virtual Machine is reset.
+
+Data: None.
+
+Example:
+
+{ "event": "RESET",
+    "timestamp": { "seconds": 1267041653, "microseconds": 9518 } }
+
+RESUME
+------
+
+Emitted when the Virtual Machine resumes execution.
+
+Data: None.
+
+Example:
+
+{ "event": "RESUME",
+    "timestamp": { "seconds": 1271770767, "microseconds": 582542 } }
+
+RTC_CHANGE
+----------
+
+Emitted when the guest changes the RTC time.
+
+Data:
+
+- "offset": Offset between base RTC clock (as specified by -rtc base), and
+new RTC clock value (json-number)
+
+Example:
+
+{ "event": "RTC_CHANGE",
+    "data": { "offset": 78 },
+    "timestamp": { "seconds": 1267020223, "microseconds": 435656 } }
+
+SHUTDOWN
+--------
+
+Emitted when the Virtual Machine has shut down, indicating that qemu
+is about to exit.
+
+Data: None.
+
+Example:
+
+{ "event": "SHUTDOWN",
+    "timestamp": { "seconds": 1267040730, "microseconds": 682951 } }
+
+Note: If the command-line option "-no-shutdown" has been specified, a STOP
+event will eventually follow the SHUTDOWN event.
+
+SPICE_CONNECTED
+---------------
+
+Emitted when a SPICE client connects.
+
+Data:
+
+- "server": Server information (json-object)
+  - "host": IP address (json-string)
+  - "port": port number (json-string)
+  - "family": address family (json-string, "ipv4" or "ipv6")
+- "client": Client information (json-object)
+  - "host": IP address (json-string)
+  - "port": port number (json-string)
+  - "family": address family (json-string, "ipv4" or "ipv6")
+
+Example:
+
+{ "timestamp": {"seconds": 1290688046, "microseconds": 388707},
+  "event": "SPICE_CONNECTED",
+  "data": {
+    "server": { "port": "5920", "family": "ipv4", "host": "127.0.0.1"},
+    "client": {"port": "52873", "family": "ipv4", "host": "127.0.0.1"}
+}}
+
+SPICE_DISCONNECTED
+------------------
+
+Emitted when a SPICE client disconnects.
+
+Data:
+
+- "server": Server information (json-object)
+  - "host": IP address (json-string)
+  - "port": port number (json-string)
+  - "family": address family (json-string, "ipv4" or "ipv6")
+- "client": Client information (json-object)
+  - "host": IP address (json-string)
+  - "port": port number (json-string)
+  - "family": address family (json-string, "ipv4" or "ipv6")
+
+Example:
+
+{ "timestamp": {"seconds": 1290688046, "microseconds": 388707},
+  "event": "SPICE_DISCONNECTED",
+  "data": {
+    "server": { "port": "5920", "family": "ipv4", "host": "127.0.0.1"},
+    "client": {"port": "52873", "family": "ipv4", "host": "127.0.0.1"}
+}}
+
+SPICE_INITIALIZED
+-----------------
+
+Emitted after initial handshake and authentication takes place (if any)
+and the SPICE channel is up and running
+
+Data:
+
+- "server": Server information (json-object)
+  - "host": IP address (json-string)
+  - "port": port number (json-string)
+  - "family": address family (json-string, "ipv4" or "ipv6")
+  - "auth": authentication method (json-string, optional)
+- "client": Client information (json-object)
+  - "host": IP address (json-string)
+  - "port": port number (json-string)
+  - "family": address family (json-string, "ipv4" or "ipv6")
+  - "connection-id": spice connection id.  All channels with the same id
+                     belong to the same spice session (json-int)
+  - "channel-type": channel type.  "1" is the main control channel, filter for
+                    this one if you want track spice sessions only (json-int)
+  - "channel-id": channel id.  Usually "0", might be different needed when
+                  multiple channels of the same type exist, such as multiple
+                  display channels in a multihead setup (json-int)
+  - "tls": whevener the channel is encrypted (json-bool)
+
+Example:
+
+{ "timestamp": {"seconds": 1290688046, "microseconds": 417172},
+  "event": "SPICE_INITIALIZED",
+  "data": {"server": {"auth": "spice", "port": "5921",
+                      "family": "ipv4", "host": "127.0.0.1"},
+           "client": {"port": "49004", "family": "ipv4", "channel-type": 3,
+                      "connection-id": 1804289383, "host": "127.0.0.1",
+                      "channel-id": 0, "tls": true}
+}}
+
+SPICE_MIGRATE_COMPLETED
+-----------------------
+
+Emitted when SPICE migration has completed
+
+Data: None.
+
+Example:
+
+{ "timestamp": {"seconds": 1290688046, "microseconds": 417172},
+  "event": "SPICE_MIGRATE_COMPLETED" }
+
+
+STOP
+----
+
+Emitted when the Virtual Machine is stopped.
+
+Data: None.
+
+Example:
+
+{ "event": "STOP",
+    "timestamp": { "seconds": 1267041730, "microseconds": 281295 } }
+
+SUSPEND
+-------
+
+Emitted when guest enters S3 state.
+
+Data: None.
+
+Example:
+
+{ "event": "SUSPEND",
+     "timestamp": { "seconds": 1344456160, "microseconds": 309119 } }
+
+SUSPEND_DISK
+------------
+
+Emitted when the guest makes a request to enter S4 state.
+
+Data: None.
+
+Example:
+
+{ "event": "SUSPEND_DISK",
+     "timestamp": { "seconds": 1344456160, "microseconds": 309119 } }
+
+Note: QEMU shuts down when entering S4 state.
+
+VNC_CONNECTED
+-------------
+
+Emitted when a VNC client establishes a connection.
+
+Data:
+
+- "server": Server information (json-object)
+  - "host": IP address (json-string)
+  - "service": port number (json-string)
+  - "family": address family (json-string, "ipv4" or "ipv6")
+  - "auth": authentication method (json-string, optional)
+- "client": Client information (json-object)
+  - "host": IP address (json-string)
+  - "service": port number (json-string)
+  - "family": address family (json-string, "ipv4" or "ipv6")
+
+Example:
+
+{ "event": "VNC_CONNECTED",
+    "data": {
+        "server": { "auth": "sasl", "family": "ipv4",
+                    "service": "5901", "host": "0.0.0.0" },
+        "client": { "family": "ipv4", "service": "58425",
+                    "host": "127.0.0.1" } },
+    "timestamp": { "seconds": 1262976601, "microseconds": 975795 } }
+
+
+Note: This event is emitted before any authentication takes place, thus
+the authentication ID is not provided.
+
+VNC_DISCONNECTED
+----------------
+
+Emitted when the connection is closed.
+
+Data:
+
+- "server": Server information (json-object)
+  - "host": IP address (json-string)
+  - "service": port number (json-string)
+  - "family": address family (json-string, "ipv4" or "ipv6")
+  - "auth": authentication method (json-string, optional)
+- "client": Client information (json-object)
+  - "host": IP address (json-string)
+  - "service": port number (json-string)
+  - "family": address family (json-string, "ipv4" or "ipv6")
+  - "x509_dname": TLS dname (json-string, optional)
+  - "sasl_username": SASL username (json-string, optional)
+
+Example:
+
+{ "event": "VNC_DISCONNECTED",
+    "data": {
+        "server": { "auth": "sasl", "family": "ipv4",
+                    "service": "5901", "host": "0.0.0.0" },
+        "client": { "family": "ipv4", "service": "58425",
+                    "host": "127.0.0.1", "sasl_username": "luiz" } },
+    "timestamp": { "seconds": 1262976601, "microseconds": 975795 } }
+
+VNC_INITIALIZED
+---------------
+
+Emitted after authentication takes place (if any) and the VNC session is
+made active.
+
+Data:
+
+- "server": Server information (json-object)
+  - "host": IP address (json-string)
+  - "service": port number (json-string)
+  - "family": address family (json-string, "ipv4" or "ipv6")
+  - "auth": authentication method (json-string, optional)
+- "client": Client information (json-object)
+  - "host": IP address (json-string)
+  - "service": port number (json-string)
+  - "family": address family (json-string, "ipv4" or "ipv6")
+  - "x509_dname": TLS dname (json-string, optional)
+  - "sasl_username": SASL username (json-string, optional)
+
+Example:
+
+{ "event": "VNC_INITIALIZED",
+    "data": {
+        "server": { "auth": "sasl", "family": "ipv4",
+                    "service": "5901", "host": "0.0.0.0"},
+        "client": { "family": "ipv4", "service": "46089",
+                    "host": "127.0.0.1", "sasl_username": "luiz" } },
+        "timestamp": { "seconds": 1263475302, "microseconds": 150772 } }
+
+VSERPORT_CHANGE
+---------------
+
+Emitted when the guest opens or closes a virtio-serial port.
+
+Data:
+
+- "id": device identifier of the virtio-serial port (json-string)
+- "open": true if the guest has opened the virtio-serial port (json-bool)
+
+Example:
+
+{ "event": "VSERPORT_CHANGE",
+    "data": { "id": "channel0", "open": true },
+    "timestamp": { "seconds": 1401385907, "microseconds": 422329 } }
+
+WAKEUP
+------
+
+Emitted when the guest has woken up from S3 and is running.
+
+Data: None.
+
+Example:
+
+{ "event": "WAKEUP",
+     "timestamp": { "seconds": 1344522075, "microseconds": 745528 } }
+
+WATCHDOG
+--------
+
+Emitted when the watchdog device's timer is expired.
+
+Data:
+
+- "action": Action that has been taken, it's one of the following (json-string):
+            "reset", "shutdown", "poweroff", "pause", "debug", or "none"
+
+Example:
+
+{ "event": "WATCHDOG",
+     "data": { "action": "reset" },
+     "timestamp": { "seconds": 1267061043, "microseconds": 959568 } }
+
+Note: If action is "reset", "shutdown", or "pause" the WATCHDOG event is
+followed respectively by the RESET, SHUTDOWN, or STOP events.

docs/specs/qcow2.txt

@@ -135,12 +135,12 @@ be stored. Each extension has a structure like the following:
 Unless stated otherwise, each header extension type shall appear at most once
 in the same image.
 
-The remaining space between the end of the header extension area and the end of
-the first cluster can be used for the backing file name. It is not allowed to
-store other data here, so that an implementation can safely modify the header
-and add extensions without harming data of compatible features that it
-doesn't support. Compatible features that need space for additional data can
-use a header extension.
+If the image has a backing file then the backing file name should be stored in
+the remaining space between the end of the header extension area and the end of
+the first cluster. It is not allowed to store other data here, so that an
+implementation can safely modify the header and add extensions without harming
+data of compatible features that it doesn't support. Compatible features that
+need space for additional data can use a header extension.
 
 
 == Feature name table ==

docs/specs/vhost-user.txt

@@ -78,14 +78,14 @@ Depending on the request type, payload can be:
    Padding: 32-bit
 
    A region is:
-   ---------------------------------------
-   | guest address | size | user address |
-   ---------------------------------------
+   -----------------------------------------------------
+   | guest address | size | user address | mmap offset |
+   -----------------------------------------------------
 
    Guest address: a 64-bit guest address of the region
    Size: a 64-bit size
    User address: a 64-bit user address
-
+   mmap offset: 64-bit offset where region starts in the mapped memory
 
 In QEMU the vhost-user message is implemented with the following struct:
 
@@ -132,7 +132,7 @@ Message types
 
  * VHOST_USER_GET_FEATURES
 
-      Id: 2
+      Id: 1
       Equivalent ioctl: VHOST_GET_FEATURES
       Master payload: N/A
       Slave payload: u64
@@ -141,7 +141,7 @@ Message types
 
  * VHOST_USER_SET_FEATURES
 
-      Id: 3
+      Id: 2
       Ioctl: VHOST_SET_FEATURES
       Master payload: u64
 
@@ -149,7 +149,7 @@ Message types
 
  * VHOST_USER_SET_OWNER
 
-      Id: 4
+      Id: 3
       Equivalent ioctl: VHOST_SET_OWNER
       Master payload: N/A
 
@@ -159,7 +159,7 @@ Message types
 
  * VHOST_USER_RESET_OWNER
 
-      Id: 5
+      Id: 4
       Equivalent ioctl: VHOST_RESET_OWNER
       Master payload: N/A
 
@@ -168,7 +168,7 @@ Message types
 
  * VHOST_USER_SET_MEM_TABLE
 
-      Id: 6
+      Id: 5
       Equivalent ioctl: VHOST_SET_MEM_TABLE
       Master payload: memory regions description
 
@@ -179,7 +179,7 @@ Message types
 
  * VHOST_USER_SET_LOG_BASE
 
-      Id: 7
+      Id: 6
       Equivalent ioctl: VHOST_SET_LOG_BASE
       Master payload: u64
 
@@ -187,7 +187,7 @@ Message types
 
  * VHOST_USER_SET_LOG_FD
 
-      Id: 8
+      Id: 7
       Equivalent ioctl: VHOST_SET_LOG_FD
       Master payload: N/A
 
@@ -195,7 +195,7 @@ Message types
 
  * VHOST_USER_SET_VRING_NUM
 
-      Id: 9
+      Id: 8
       Equivalent ioctl: VHOST_SET_VRING_NUM
       Master payload: vring state description
 
@@ -203,7 +203,7 @@ Message types
 
  * VHOST_USER_SET_VRING_ADDR
 
-      Id: 10
+      Id: 9
       Equivalent ioctl: VHOST_SET_VRING_ADDR
       Master payload: vring address description
       Slave payload: N/A
@@ -212,7 +212,7 @@ Message types
 
  * VHOST_USER_SET_VRING_BASE
 
-      Id: 11
+      Id: 10
       Equivalent ioctl: VHOST_SET_VRING_BASE
       Master payload: vring state description
 
@@ -220,7 +220,7 @@ Message types
 
  * VHOST_USER_GET_VRING_BASE
 
-      Id: 12
+      Id: 11
       Equivalent ioctl: VHOST_USER_GET_VRING_BASE
       Master payload: vring state description
       Slave payload: vring state description
@@ -229,7 +229,7 @@ Message types
 
  * VHOST_USER_SET_VRING_KICK
 
-      Id: 13
+      Id: 12
       Equivalent ioctl: VHOST_SET_VRING_KICK
       Master payload: u64
 
@@ -242,7 +242,7 @@ Message types
 
  * VHOST_USER_SET_VRING_CALL
 
-      Id: 14
+      Id: 13
       Equivalent ioctl: VHOST_SET_VRING_CALL
       Master payload: u64
 
@@ -255,7 +255,7 @@ Message types
 
  * VHOST_USER_SET_VRING_ERR
 
-      Id: 15
+      Id: 14
       Equivalent ioctl: VHOST_SET_VRING_ERR
       Master payload: u64
 

docs/tracing.txt

@@ -307,3 +307,43 @@ guard such computations and avoid its compilation when the event is disabled:
 You can check both if the event has been disabled and is dynamically enabled at
 the same time using the 'trace_event_get_state' routine (see header
 "trace/control.h" for more information).
+
+=== "tcg" ===
+
+Guest code generated by TCG can be traced by defining an event with the "tcg"
+event property. Internally, this property generates two events:
+"<eventname>_trans" to trace the event at translation time, and
+"<eventname>_exec" to trace the event at execution time.
+
+Instead of using these two events, you should instead use the function
+"trace_<eventname>_tcg" during translation (TCG code generation). This function
+will automatically call "trace_<eventname>_trans", and will generate the
+necessary TCG code to call "trace_<eventname>_exec" during guest code execution.
+
+Events with the "tcg" property can be declared in the "trace-events" file with a
+mix of native and TCG types, and "trace_<eventname>_tcg" will gracefully forward
+them to the "<eventname>_trans" and "<eventname>_exec" events. Since TCG values
+are not known at translation time, these are ignored by the "<eventname>_trans"
+event. Because of this, the entry in the "trace-events" file needs two printing
+formats (separated by a comma):
+
+    tcg foo(uint8_t a1, TCGv_i32 a2) "a1=%d", "a1=%d a2=%d"
+
+For example:
+
+    #include "trace-tcg.h"
+    
+    void some_disassembly_func (...)
+    {
+        uint8_t a1 = ...;
+        TCGv_i32 a2 = ...;
+        trace_foo_tcg(a1, a2);
+    }
+
+This will immediately call:
+
+    void trace_foo_trans(uint8_t a1);
+
+and will generate the TCG code to call:
+
+    void trace_foo(uint8_t a1, uint32_t a2);

dump.c

@@ -71,18 +71,14 @@ uint64_t cpu_to_dump64(DumpState *s, uint64_t val)
 
 static int dump_cleanup(DumpState *s)
 {
-    int ret = 0;
-
     guest_phys_blocks_free(&s->guest_phys_blocks);
     memory_mapping_list_free(&s->list);
-    if (s->fd != -1) {
-        close(s->fd);
-    }
+    close(s->fd);
     if (s->resume) {
         vm_start();
     }
 
-    return ret;
+    return 0;
 }
 
 static void dump_error(DumpState *s, const char *reason)
@@ -1499,6 +1495,8 @@ static int dump_init(DumpState *s, int fd, bool has_format,
     s->begin = begin;
     s->length = length;
 
+    memory_mapping_list_init(&s->list);
+
     guest_phys_blocks_init(&s->guest_phys_blocks);
     guest_phys_blocks_append(&s->guest_phys_blocks);
 
@@ -1526,7 +1524,6 @@ static int dump_init(DumpState *s, int fd, bool has_format,
     }
 
     /* get memory mapping */
-    memory_mapping_list_init(&s->list);
     if (paging) {
         qemu_get_guest_memory_mapping(&s->list, &s->guest_phys_blocks, &err);
         if (err != NULL) {
@@ -1622,12 +1619,7 @@ static int dump_init(DumpState *s, int fd, bool has_format,
     return 0;
 
 cleanup:
-    guest_phys_blocks_free(&s->guest_phys_blocks);
-
-    if (s->resume) {
-        vm_start();
-    }
-
+    dump_cleanup(s);
     return -1;
 }
 

exec.c

@@ -373,7 +373,7 @@ MemoryRegion *address_space_translate(AddressSpace *as, hwaddr addr,
             break;
         }
 
-        iotlb = mr->iommu_ops->translate(mr, addr);
+        iotlb = mr->iommu_ops->translate(mr, addr, is_write);
         addr = ((iotlb.translated_addr & ~iotlb.addr_mask)
                 | (addr & iotlb.addr_mask));
         len = MIN(len, (addr | iotlb.addr_mask) - addr + 1);
@@ -883,7 +883,7 @@ static void phys_section_destroy(MemoryRegion *mr)
 
     if (mr->subpage) {
         subpage_t *subpage = container_of(mr, subpage_t, iomem);
-        memory_region_destroy(&subpage->iomem);
+        object_unref(OBJECT(&subpage->iomem));
         g_free(subpage);
     }
 }
@@ -1044,7 +1044,7 @@ static void *file_ram_alloc(RAMBlock *block,
     }
 
     /* Make name safe to use with mkstemp by replacing '/' with '_'. */
-    sanitized_name = g_strdup(block->mr->name);
+    sanitized_name = g_strdup(memory_region_name(block->mr));
     for (c = sanitized_name; *c != '\0'; c++) {
         if (*c == '/')
             *c = '_';
@@ -1242,7 +1242,7 @@ static ram_addr_t ram_block_add(RAMBlock *new_block)
             new_block->host = phys_mem_alloc(new_block->length);
             if (!new_block->host) {
                 fprintf(stderr, "Cannot set up guest memory '%s': %s\n",
-                        new_block->mr->name, strerror(errno));
+                        memory_region_name(new_block->mr), strerror(errno));
                 exit(1);
             }
             memory_try_enable_merging(new_block->host, new_block->length);
@@ -1456,6 +1456,13 @@ int qemu_get_ram_fd(ram_addr_t addr)
     return block->fd;
 }
 
+void *qemu_get_ram_block_host_ptr(ram_addr_t addr)
+{
+    RAMBlock *block = qemu_get_ram_block(addr);
+
+    return block->host;
+}
+
 /* Return a host pointer to ram allocated with qemu_ram_alloc.
    With the exception of the softmmu code in this file, this should
    only be used for local memory (e.g. video ram) that the device owns,
@@ -1561,8 +1568,7 @@ static void notdirty_mem_write(void *opaque, hwaddr ram_addr,
     default:
         abort();
     }
-    cpu_physical_memory_set_dirty_flag(ram_addr, DIRTY_MEMORY_MIGRATION);
-    cpu_physical_memory_set_dirty_flag(ram_addr, DIRTY_MEMORY_VGA);
+    cpu_physical_memory_set_dirty_range_nocode(ram_addr, size);
     /* we remove the notdirty callback only if the code has been
        flushed */
     if (!cpu_physical_memory_is_clean(ram_addr)) {
@@ -1761,7 +1767,7 @@ static subpage_t *subpage_init(AddressSpace *as, hwaddr base)
     mmio->as = as;
     mmio->base = base;
     memory_region_init_io(&mmio->iomem, NULL, &subpage_ops, mmio,
-                          "subpage", TARGET_PAGE_SIZE);
+                          NULL, TARGET_PAGE_SIZE);
     mmio->iomem.subpage = true;
 #if defined(DEBUG_SUBPAGE)
     printf("%s: %p base " TARGET_FMT_plx " len %08x\n", __func__,
@@ -1794,13 +1800,13 @@ MemoryRegion *iotlb_to_region(AddressSpace *as, hwaddr index)
 
 static void io_mem_init(void)
 {
-    memory_region_init_io(&io_mem_rom, NULL, &unassigned_mem_ops, NULL, "rom", UINT64_MAX);
+    memory_region_init_io(&io_mem_rom, NULL, &unassigned_mem_ops, NULL, NULL, UINT64_MAX);
     memory_region_init_io(&io_mem_unassigned, NULL, &unassigned_mem_ops, NULL,
-                          "unassigned", UINT64_MAX);
+                          NULL, UINT64_MAX);
     memory_region_init_io(&io_mem_notdirty, NULL, &notdirty_mem_ops, NULL,
-                          "notdirty", UINT64_MAX);
+                          NULL, UINT64_MAX);
     memory_region_init_io(&io_mem_watch, NULL, &watch_mem_ops, NULL,
-                          "watch", UINT64_MAX);
+                          NULL, UINT64_MAX);
 }
 
 static void mem_begin(MemoryListener *listener)
@@ -1971,8 +1977,7 @@ static void invalidate_and_set_dirty(hwaddr addr,
         /* invalidate code */
         tb_invalidate_phys_page_range(addr, addr + length, 0);
         /* set dirty bit */
-        cpu_physical_memory_set_dirty_flag(addr, DIRTY_MEMORY_VGA);
-        cpu_physical_memory_set_dirty_flag(addr, DIRTY_MEMORY_MIGRATION);
+        cpu_physical_memory_set_dirty_range_nocode(addr, length);
     }
     xen_modified_memory(addr, length);
 }
@@ -2328,15 +2333,7 @@ void address_space_unmap(AddressSpace *as, void *buffer, hwaddr len,
         mr = qemu_ram_addr_from_host(buffer, &addr1);
         assert(mr != NULL);
         if (is_write) {
-            while (access_len) {
-                unsigned l;
-                l = TARGET_PAGE_SIZE;
-                if (l > access_len)
-                    l = access_len;
-                invalidate_and_set_dirty(addr1, l);
-                addr1 += l;
-                access_len -= l;
-            }
+            invalidate_and_set_dirty(addr1, access_len);
         }
         if (xen_enabled()) {
             xen_invalidate_map_cache_entry(buffer);
@@ -2574,9 +2571,7 @@ void stl_phys_notdirty(AddressSpace *as, hwaddr addr, uint32_t val)
                 /* invalidate code */
                 tb_invalidate_phys_page_range(addr1, addr1 + 4, 0);
                 /* set dirty bit */
-                cpu_physical_memory_set_dirty_flag(addr1,
-                                                   DIRTY_MEMORY_MIGRATION);
-                cpu_physical_memory_set_dirty_flag(addr1, DIRTY_MEMORY_VGA);
+                cpu_physical_memory_set_dirty_range_nocode(addr1, 4);
             }
         }
     }
@@ -2752,14 +2747,12 @@ int cpu_memory_rw_debug(CPUState *cpu, target_ulong addr,
 }
 #endif
 
-#if !defined(CONFIG_USER_ONLY)
-
 /*
  * A helper function for the _utterly broken_ virtio device model to find out if
  * it's running on a big endian machine. Don't do this at home kids!
  */
-bool virtio_is_big_endian(void);
-bool virtio_is_big_endian(void)
+bool target_words_bigendian(void);
+bool target_words_bigendian(void)
 {
 #if defined(TARGET_WORDS_BIGENDIAN)
     return true;
@@ -2768,8 +2761,6 @@ bool virtio_is_big_endian(void)
 #endif
 }
 
-#endif
-
 #ifndef CONFIG_USER_ONLY
 bool cpu_physical_memory_is_io(hwaddr phys_addr)
 {

gdb-xml/s390-acr.xml

@@ -0,0 +1,26 @@
+<?xml version="1.0"?>
+<!-- Copyright (C) 2010-2014 Free Software Foundation, Inc.
+
+     Copying and distribution of this file, with or without modification,
+     are permitted in any medium without royalty provided the copyright
+     notice and this notice are preserved.  -->
+
+<!DOCTYPE feature SYSTEM "gdb-target.dtd">
+<feature name="org.gnu.gdb.s390.acr">
+  <reg name="acr0" bitsize="32" type="uint32" group="access"/>
+  <reg name="acr1" bitsize="32" type="uint32" group="access"/>
+  <reg name="acr2" bitsize="32" type="uint32" group="access"/>
+  <reg name="acr3" bitsize="32" type="uint32" group="access"/>
+  <reg name="acr4" bitsize="32" type="uint32" group="access"/>
+  <reg name="acr5" bitsize="32" type="uint32" group="access"/>
+  <reg name="acr6" bitsize="32" type="uint32" group="access"/>
+  <reg name="acr7" bitsize="32" type="uint32" group="access"/>
+  <reg name="acr8" bitsize="32" type="uint32" group="access"/>
+  <reg name="acr9" bitsize="32" type="uint32" group="access"/>
+  <reg name="acr10" bitsize="32" type="uint32" group="access"/>
+  <reg name="acr11" bitsize="32" type="uint32" group="access"/>
+  <reg name="acr12" bitsize="32" type="uint32" group="access"/>
+  <reg name="acr13" bitsize="32" type="uint32" group="access"/>
+  <reg name="acr14" bitsize="32" type="uint32" group="access"/>
+  <reg name="acr15" bitsize="32" type="uint32" group="access"/>
+</feature>

gdb-xml/s390-fpr.xml

@@ -0,0 +1,27 @@
+<?xml version="1.0"?>
+<!-- Copyright (C) 2010-2014 Free Software Foundation, Inc.
+
+     Copying and distribution of this file, with or without modification,
+     are permitted in any medium without royalty provided the copyright
+     notice and this notice are preserved.  -->
+
+<!DOCTYPE feature SYSTEM "gdb-target.dtd">
+<feature name="org.gnu.gdb.s390.fpr">
+  <reg name="fpc" bitsize="32" type="uint32" group="float"/>
+  <reg name="f0" bitsize="64" type="ieee_double" group="float"/>
+  <reg name="f1" bitsize="64" type="ieee_double" group="float"/>
+  <reg name="f2" bitsize="64" type="ieee_double" group="float"/>
+  <reg name="f3" bitsize="64" type="ieee_double" group="float"/>
+  <reg name="f4" bitsize="64" type="ieee_double" group="float"/>
+  <reg name="f5" bitsize="64" type="ieee_double" group="float"/>
+  <reg name="f6" bitsize="64" type="ieee_double" group="float"/>
+  <reg name="f7" bitsize="64" type="ieee_double" group="float"/>
+  <reg name="f8" bitsize="64" type="ieee_double" group="float"/>
+  <reg name="f9" bitsize="64" type="ieee_double" group="float"/>
+  <reg name="f10" bitsize="64" type="ieee_double" group="float"/>
+  <reg name="f11" bitsize="64" type="ieee_double" group="float"/>
+  <reg name="f12" bitsize="64" type="ieee_double" group="float"/>
+  <reg name="f13" bitsize="64" type="ieee_double" group="float"/>
+  <reg name="f14" bitsize="64" type="ieee_double" group="float"/>
+  <reg name="f15" bitsize="64" type="ieee_double" group="float"/>
+</feature>

gdb-xml/s390x-core64.xml

@@ -0,0 +1,28 @@
+<?xml version="1.0"?>
+<!-- Copyright (C) 2010-2014 Free Software Foundation, Inc.
+
+     Copying and distribution of this file, with or without modification,
+     are permitted in any medium without royalty provided the copyright
+     notice and this notice are preserved.  -->
+
+<!DOCTYPE feature SYSTEM "gdb-target.dtd">
+<feature name="org.gnu.gdb.s390.core">
+  <reg name="pswm" bitsize="64" type="uint64" group="psw"/>
+  <reg name="pswa" bitsize="64" type="uint64" group="psw"/>
+  <reg name="r0" bitsize="64" type="uint64" group="general"/>
+  <reg name="r1" bitsize="64" type="uint64" group="general"/>
+  <reg name="r2" bitsize="64" type="uint64" group="general"/>
+  <reg name="r3" bitsize="64" type="uint64" group="general"/>
+  <reg name="r4" bitsize="64" type="uint64" group="general"/>
+  <reg name="r5" bitsize="64" type="uint64" group="general"/>
+  <reg name="r6" bitsize="64" type="uint64" group="general"/>
+  <reg name="r7" bitsize="64" type="uint64" group="general"/>
+  <reg name="r8" bitsize="64" type="uint64" group="general"/>
+  <reg name="r9" bitsize="64" type="uint64" group="general"/>
+  <reg name="r10" bitsize="64" type="uint64" group="general"/>
+  <reg name="r11" bitsize="64" type="uint64" group="general"/>
+  <reg name="r12" bitsize="64" type="uint64" group="general"/>
+  <reg name="r13" bitsize="64" type="uint64" group="general"/>
+  <reg name="r14" bitsize="64" type="uint64" group="general"/>
+  <reg name="r15" bitsize="64" type="uint64" group="general"/>
+</feature>