Series comparison

The following changes since commit 5dae13cd71f0755a1395b5a4cde635b8a6ee3f58:

Merge remote-tracking branch 'remotes/rth/tags/pull-or-20170214' into staging (2017-02-14 09:55:48 +0000)

are available in the git repository at:

git://github.com/stefanha/qemu.git tags/block-pull-request

for you to fetch changes up to decc18f33adecb1316437a47fff0cf0a7665906a:

coroutine-lock: make CoRwlock thread-safe and fair (2017-02-16 17:17:34 +0000)

----------------------------------------------------------------

Paolo Bonzini (24):
  block: move AioContext, QEMUTimer, main-loop to libqemuutil
  aio: introduce aio_co_schedule and aio_co_wake
  block-backend: allow blk_prw from coroutine context
  test-thread-pool: use generic AioContext infrastructure
  io: add methods to set I/O handlers on AioContext
  io: make qio_channel_yield aware of AioContexts
  nbd: convert to use qio_channel_yield
  coroutine-lock: reschedule coroutine on the AioContext it was running
    on
  blkdebug: reschedule coroutine on the AioContext it is running on
  qed: introduce qed_aio_start_io and qed_aio_next_io_cb
  aio: push aio_context_acquire/release down to dispatching
  block: explicitly acquire aiocontext in timers that need it
  block: explicitly acquire aiocontext in callbacks that need it
  block: explicitly acquire aiocontext in bottom halves that need it
  block: explicitly acquire aiocontext in aio callbacks that need it
  aio-posix: partially inline aio_dispatch into aio_poll
  async: remove unnecessary inc/dec pairs
  block: document fields protected by AioContext lock
  coroutine-lock: make CoMutex thread-safe
  coroutine-lock: add limited spinning to CoMutex
  test-aio-multithread: add performance comparison with thread-based
    mutexes
  coroutine-lock: place CoMutex before CoQueue in header
  coroutine-lock: add mutex argument to CoQueue APIs
  coroutine-lock: make CoRwlock thread-safe and fair

-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

AioContext is fairly self contained, the only dependency is QEMUTimer but
that in turn doesn't need anything else.  So move them out of block-obj-y
to avoid introducing a dependency from io/ to block-obj-y.

main-loop and its dependency iohandler also need to be moved, because
later in this series io/ will call iohandler_get_aio_context.

[Changed copyright "the QEMU team" to "other QEMU contributors" as
suggested by Daniel Berrange and agreed by Paolo.
--Stefan]

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 20170213135235.12274-2-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 Makefile.objs                       |  4 ---
 stubs/Makefile.objs                 |  1 +
 tests/Makefile.include              | 11 ++++----
 util/Makefile.objs                  |  6 +++-
 block/io.c                          | 29 -------------------
 stubs/linux-aio.c                   | 32 +++++++++++++++++++++
 stubs/set-fd-handler.c              | 11 --------
 aio-posix.c => util/aio-posix.c     |  2 +-
 aio-win32.c => util/aio-win32.c     |  0
 util/aiocb.c                        | 55 +++++++++++++++++++++++++++++++++++++
 async.c => util/async.c             |  3 +-
 iohandler.c => util/iohandler.c     |  0
 main-loop.c => util/main-loop.c     |  0
 qemu-timer.c => util/qemu-timer.c   |  0
 thread-pool.c => util/thread-pool.c |  2 +-
 trace-events                        | 11 --------
 util/trace-events                   | 11 ++++++++
 17 files changed, 114 insertions(+), 64 deletions(-)
 create mode 100644 stubs/linux-aio.c
 rename aio-posix.c => util/aio-posix.c (99%)
 rename aio-win32.c => util/aio-win32.c (100%)
 create mode 100644 util/aiocb.c
 rename async.c => util/async.c (99%)
 rename iohandler.c => util/iohandler.c (100%)
 rename main-loop.c => util/main-loop.c (100%)
 rename qemu-timer.c => util/qemu-timer.c (100%)
 rename thread-pool.c => util/thread-pool.c (99%)

diff --git a/Makefile.objs b/Makefile.objs
index XXXXXXX..XXXXXXX 100644
--- a/Makefile.objs
+++ b/Makefile.objs
@@ -XXX,XX +XXX,XX @@ chardev-obj-y = chardev/
 #######################################################################
 # block-obj-y is code used by both qemu system emulation and qemu-img
 
-block-obj-y = async.o thread-pool.o
 block-obj-y += nbd/
 block-obj-y += block.o blockjob.o
-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 += qemu-io-cmds.o
 block-obj-$(CONFIG_REPLICATION) += replication.o
diff --git a/stubs/Makefile.objs b/stubs/Makefile.objs
index XXXXXXX..XXXXXXX 100644
--- a/stubs/Makefile.objs
+++ b/stubs/Makefile.objs
@@ -XXX,XX +XXX,XX @@ stub-obj-y += get-vm-name.o
 stub-obj-y += iothread.o
 stub-obj-y += iothread-lock.o
 stub-obj-y += is-daemonized.o
+stub-obj-$(CONFIG_LINUX_AIO) += linux-aio.o
 stub-obj-y += machine-init-done.o
 stub-obj-y += migr-blocker.o
 stub-obj-y += monitor.o
diff --git a/tests/Makefile.include b/tests/Makefile.include
index XXXXXXX..XXXXXXX 100644
--- a/tests/Makefile.include
+++ b/tests/Makefile.include
@@ -XXX,XX +XXX,XX @@ check-unit-y += tests/test-visitor-serialization$(EXESUF)
 check-unit-y += tests/test-iov$(EXESUF)
 gcov-files-test-iov-y = util/iov.c
 check-unit-y += tests/test-aio$(EXESUF)
+gcov-files-test-aio-y = util/async.c util/qemu-timer.o
+gcov-files-test-aio-$(CONFIG_WIN32) += util/aio-win32.c
+gcov-files-test-aio-$(CONFIG_POSIX) += util/aio-posix.c
 check-unit-y += tests/test-throttle$(EXESUF)
 gcov-files-test-aio-$(CONFIG_WIN32) = aio-win32.c
 gcov-files-test-aio-$(CONFIG_POSIX) = aio-posix.c
@@ -XXX,XX +XXX,XX @@ tests/check-qjson$(EXESUF): tests/check-qjson.o $(test-util-obj-y)
 tests/check-qom-interface$(EXESUF): tests/check-qom-interface.o $(test-qom-obj-y)
 tests/check-qom-proplist$(EXESUF): tests/check-qom-proplist.o $(test-qom-obj-y)
 
-tests/test-char$(EXESUF): tests/test-char.o qemu-timer.o \
-	$(test-util-obj-y) $(qtest-obj-y) $(test-block-obj-y) $(chardev-obj-y)
+tests/test-char$(EXESUF): tests/test-char.o $(test-util-obj-y) $(qtest-obj-y) $(test-io-obj-y) $(chardev-obj-y)
 tests/test-coroutine$(EXESUF): tests/test-coroutine.o $(test-block-obj-y)
 tests/test-aio$(EXESUF): tests/test-aio.o $(test-block-obj-y)
 tests/test-throttle$(EXESUF): tests/test-throttle.o $(test-block-obj-y)
@@ -XXX,XX +XXX,XX @@ tests/test-vmstate$(EXESUF): tests/test-vmstate.o \
 	migration/vmstate.o migration/qemu-file.o \
         migration/qemu-file-channel.o migration/qjson.o \
 	$(test-io-obj-y)
-tests/test-timed-average$(EXESUF): tests/test-timed-average.o qemu-timer.o \
-	$(test-util-obj-y)
+tests/test-timed-average$(EXESUF): tests/test-timed-average.o $(test-util-obj-y)
 tests/test-base64$(EXESUF): tests/test-base64.o \
 	libqemuutil.a libqemustub.a
 tests/ptimer-test$(EXESUF): tests/ptimer-test.o tests/ptimer-test-stubs.o hw/core/ptimer.o libqemustub.a
@@ -XXX,XX +XXX,XX @@ tests/usb-hcd-ehci-test$(EXESUF): tests/usb-hcd-ehci-test.o $(libqos-usb-obj-y)
 tests/usb-hcd-xhci-test$(EXESUF): tests/usb-hcd-xhci-test.o $(libqos-usb-obj-y)
 tests/pc-cpu-test$(EXESUF): tests/pc-cpu-test.o
 tests/postcopy-test$(EXESUF): tests/postcopy-test.o
-tests/vhost-user-test$(EXESUF): tests/vhost-user-test.o qemu-timer.o \
+tests/vhost-user-test$(EXESUF): tests/vhost-user-test.o $(test-util-obj-y) \
 	$(qtest-obj-y) $(test-io-obj-y) $(libqos-virtio-obj-y) $(libqos-pc-obj-y) \
 	$(chardev-obj-y)
 tests/qemu-iotests/socket_scm_helper$(EXESUF): tests/qemu-iotests/socket_scm_helper.o
diff --git a/util/Makefile.objs b/util/Makefile.objs
index XXXXXXX..XXXXXXX 100644
--- a/util/Makefile.objs
+++ b/util/Makefile.objs
@@ -XXX,XX +XXX,XX @@
 util-obj-y = osdep.o cutils.o unicode.o qemu-timer-common.o
 util-obj-y += bufferiszero.o
 util-obj-y += lockcnt.o
+util-obj-y += aiocb.o async.o thread-pool.o qemu-timer.o
+util-obj-y += main-loop.o iohandler.o
+util-obj-$(CONFIG_POSIX) += aio-posix.o
 util-obj-$(CONFIG_POSIX) += compatfd.o
 util-obj-$(CONFIG_POSIX) += event_notifier-posix.o
 util-obj-$(CONFIG_POSIX) += mmap-alloc.o
 util-obj-$(CONFIG_POSIX) += oslib-posix.o
 util-obj-$(CONFIG_POSIX) += qemu-openpty.o
 util-obj-$(CONFIG_POSIX) += qemu-thread-posix.o
-util-obj-$(CONFIG_WIN32) += event_notifier-win32.o
 util-obj-$(CONFIG_POSIX) += memfd.o
+util-obj-$(CONFIG_WIN32) += aio-win32.o
+util-obj-$(CONFIG_WIN32) += event_notifier-win32.o
 util-obj-$(CONFIG_WIN32) += oslib-win32.o
 util-obj-$(CONFIG_WIN32) += qemu-thread-win32.o
 util-obj-y += envlist.o path.o module.o
diff --git a/block/io.c b/block/io.c
index XXXXXXX..XXXXXXX 100644
--- a/block/io.c
+++ b/block/io.c
@@ -XXX,XX +XXX,XX @@ BlockAIOCB *bdrv_aio_flush(BlockDriverState *bs,
     return &acb->common;
 }
 
-void *qemu_aio_get(const AIOCBInfo *aiocb_info, BlockDriverState *bs,
-                   BlockCompletionFunc *cb, void *opaque)
-{
-    BlockAIOCB *acb;
-
-    acb = g_malloc(aiocb_info->aiocb_size);
-    acb->aiocb_info = aiocb_info;
-    acb->bs = bs;
-    acb->cb = cb;
-    acb->opaque = opaque;
-    acb->refcnt = 1;
-    return acb;
-}
-
-void qemu_aio_ref(void *p)
-{
-    BlockAIOCB *acb = p;
-    acb->refcnt++;
-}
-
-void qemu_aio_unref(void *p)
-{
-    BlockAIOCB *acb = p;
-    assert(acb->refcnt > 0);
-    if (--acb->refcnt == 0) {
-        g_free(acb);
-    }
-}
-
 /**************************************************************/
 /* Coroutine block device emulation */
 
diff --git a/stubs/linux-aio.c b/stubs/linux-aio.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/stubs/linux-aio.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * Linux native AIO support.
+ *
+ * Copyright (C) 2009 IBM, Corp.
+ * Copyright (C) 2009 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.
+ */
+#include "qemu/osdep.h"
+#include "block/aio.h"
+#include "block/raw-aio.h"
+
+void laio_detach_aio_context(LinuxAioState *s, AioContext *old_context)
+{
+    abort();
+}
+
+void laio_attach_aio_context(LinuxAioState *s, AioContext *new_context)
+{
+    abort();
+}
+
+LinuxAioState *laio_init(void)
+{
+    abort();
+}
+
+void laio_cleanup(LinuxAioState *s)
+{
+    abort();
+}
diff --git a/stubs/set-fd-handler.c b/stubs/set-fd-handler.c
index XXXXXXX..XXXXXXX 100644
--- a/stubs/set-fd-handler.c
+++ b/stubs/set-fd-handler.c
@@ -XXX,XX +XXX,XX @@ void qemu_set_fd_handler(int fd,
 {
     abort();
 }
-
-void aio_set_fd_handler(AioContext *ctx,
-                        int fd,
-                        bool is_external,
-                        IOHandler *io_read,
-                        IOHandler *io_write,
-                        AioPollFn *io_poll,
-                        void *opaque)
-{
-    abort();
-}
diff --git a/aio-posix.c b/util/aio-posix.c
similarity index 99%
rename from aio-posix.c
rename to util/aio-posix.c
index XXXXXXX..XXXXXXX 100644
--- a/aio-posix.c
+++ b/util/aio-posix.c
@@ -XXX,XX +XXX,XX @@
 #include "qemu/rcu_queue.h"
 #include "qemu/sockets.h"
 #include "qemu/cutils.h"
-#include "trace-root.h"
+#include "trace.h"
 #ifdef CONFIG_EPOLL_CREATE1
 #include <sys/epoll.h>
 #endif
diff --git a/aio-win32.c b/util/aio-win32.c
similarity index 100%
rename from aio-win32.c
rename to util/aio-win32.c
diff --git a/util/aiocb.c b/util/aiocb.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/util/aiocb.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * BlockAIOCB allocation
+ *
+ * Copyright (c) 2003-2017 Fabrice Bellard and other QEMU contributors
+ *
+ * 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/osdep.h"
+#include "block/aio.h"
+
+void *qemu_aio_get(const AIOCBInfo *aiocb_info, BlockDriverState *bs,
+                   BlockCompletionFunc *cb, void *opaque)
+{
+    BlockAIOCB *acb;
+
+    acb = g_malloc(aiocb_info->aiocb_size);
+    acb->aiocb_info = aiocb_info;
+    acb->bs = bs;
+    acb->cb = cb;
+    acb->opaque = opaque;
+    acb->refcnt = 1;
+    return acb;
+}
+
+void qemu_aio_ref(void *p)
+{
+    BlockAIOCB *acb = p;
+    acb->refcnt++;
+}
+
+void qemu_aio_unref(void *p)
+{
+    BlockAIOCB *acb = p;
+    assert(acb->refcnt > 0);
+    if (--acb->refcnt == 0) {
+        g_free(acb);
+    }
+}
diff --git a/async.c b/util/async.c
similarity index 99%
rename from async.c
rename to util/async.c
index XXXXXXX..XXXXXXX 100644
--- a/async.c
+++ b/util/async.c
@@ -XXX,XX +XXX,XX @@
 /*
- * QEMU System Emulator
+ * Data plane event loop
  *
  * Copyright (c) 2003-2008 Fabrice Bellard
+ * Copyright (c) 2009-2017 QEMU contributors
  *
  * Permission is hereby granted, free of charge, to any person obtaining a copy
  * of this software and associated documentation files (the "Software"), to deal
diff --git a/iohandler.c b/util/iohandler.c
similarity index 100%
rename from iohandler.c
rename to util/iohandler.c
diff --git a/main-loop.c b/util/main-loop.c
similarity index 100%
rename from main-loop.c
rename to util/main-loop.c
diff --git a/qemu-timer.c b/util/qemu-timer.c
similarity index 100%
rename from qemu-timer.c
rename to util/qemu-timer.c
diff --git a/thread-pool.c b/util/thread-pool.c
similarity index 99%
rename from thread-pool.c
rename to util/thread-pool.c
index XXXXXXX..XXXXXXX 100644
--- a/thread-pool.c
+++ b/util/thread-pool.c
@@ -XXX,XX +XXX,XX @@
 #include "qemu/queue.h"
 #include "qemu/thread.h"
 #include "qemu/coroutine.h"
-#include "trace-root.h"
+#include "trace.h"
 #include "block/thread-pool.h"
 #include "qemu/main-loop.h"
 
diff --git a/trace-events b/trace-events
index XXXXXXX..XXXXXXX 100644
--- a/trace-events
+++ b/trace-events
@@ -XXX,XX +XXX,XX @@
 #
 # The <format-string> should be a sprintf()-compatible format string.
 
-# aio-posix.c
-run_poll_handlers_begin(void *ctx, int64_t max_ns) "ctx %p max_ns %"PRId64
-run_poll_handlers_end(void *ctx, bool progress) "ctx %p progress %d"
-poll_shrink(void *ctx, int64_t old, int64_t new) "ctx %p old %"PRId64" new %"PRId64
-poll_grow(void *ctx, int64_t old, int64_t new) "ctx %p old %"PRId64" new %"PRId64
-
-# thread-pool.c
-thread_pool_submit(void *pool, void *req, void *opaque) "pool %p req %p opaque %p"
-thread_pool_complete(void *pool, void *req, void *opaque, int ret) "pool %p req %p opaque %p ret %d"
-thread_pool_cancel(void *req, void *opaque) "req %p opaque %p"
-
 # ioport.c
 cpu_in(unsigned int addr, char size, unsigned int val) "addr %#x(%c) value %u"
 cpu_out(unsigned int addr, char size, unsigned int val) "addr %#x(%c) value %u"
diff --git a/util/trace-events b/util/trace-events
index XXXXXXX..XXXXXXX 100644
--- a/util/trace-events
+++ b/util/trace-events
@@ -XXX,XX +XXX,XX @@
 # See docs/tracing.txt for syntax documentation.
 
+# util/aio-posix.c
+run_poll_handlers_begin(void *ctx, int64_t max_ns) "ctx %p max_ns %"PRId64
+run_poll_handlers_end(void *ctx, bool progress) "ctx %p progress %d"
+poll_shrink(void *ctx, int64_t old, int64_t new) "ctx %p old %"PRId64" new %"PRId64
+poll_grow(void *ctx, int64_t old, int64_t new) "ctx %p old %"PRId64" new %"PRId64
+
+# util/thread-pool.c
+thread_pool_submit(void *pool, void *req, void *opaque) "pool %p req %p opaque %p"
+thread_pool_complete(void *pool, void *req, void *opaque, int ret) "pool %p req %p opaque %p ret %d"
+thread_pool_cancel(void *req, void *opaque) "req %p opaque %p"
+
 # util/buffer.c
 buffer_resize(const char *buf, size_t olen, size_t len) "%s: old %zd, new %zd"
 buffer_move_empty(const char *buf, size_t len, const char *from) "%s: %zd bytes from %s"
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

aio_co_wake provides the infrastructure to start a coroutine on a "home"
AioContext.  It will be used by CoMutex and CoQueue, so that coroutines
don't jump from one context to another when they go to sleep on a
mutex or waitqueue.  However, it can also be used as a more efficient
alternative to one-shot bottom halves, and saves the effort of tracking
which AioContext a coroutine is running on.

aio_co_schedule is the part of aio_co_wake that starts a coroutine
on a remove AioContext, but it is also useful to implement e.g.
bdrv_set_aio_context callbacks.

The implementation of aio_co_schedule is based on a lock-free
multiple-producer, single-consumer queue.  The multiple producers use
cmpxchg to add to a LIFO stack.  The consumer (a per-AioContext bottom
half) grabs all items added so far, inverts the list to make it FIFO,
and goes through it one item at a time until it's empty.  The data
structure was inspired by OSv, which uses it in the very code we'll
"port" to QEMU for the thread-safe CoMutex.

Most of the new code is really tests.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 20170213135235.12274-3-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 tests/Makefile.include       |   8 +-
 include/block/aio.h          |  32 +++++++
 include/qemu/coroutine_int.h |  11 ++-
 tests/iothread.h             |  25 +++++
 tests/iothread.c             |  91 ++++++++++++++++++
 tests/test-aio-multithread.c | 213 +++++++++++++++++++++++++++++++++++++++++++
 util/async.c                 |  65 +++++++++++++
 util/qemu-coroutine.c        |   8 ++
 util/trace-events            |   4 +
 9 files changed, 453 insertions(+), 4 deletions(-)
 create mode 100644 tests/iothread.h
 create mode 100644 tests/iothread.c
 create mode 100644 tests/test-aio-multithread.c

diff --git a/tests/Makefile.include b/tests/Makefile.include
index XXXXXXX..XXXXXXX 100644
--- a/tests/Makefile.include
+++ b/tests/Makefile.include
@@ -XXX,XX +XXX,XX @@ check-unit-y += tests/test-aio$(EXESUF)
 gcov-files-test-aio-y = util/async.c util/qemu-timer.o
 gcov-files-test-aio-$(CONFIG_WIN32) += util/aio-win32.c
 gcov-files-test-aio-$(CONFIG_POSIX) += util/aio-posix.c
+check-unit-y += tests/test-aio-multithread$(EXESUF)
+gcov-files-test-aio-multithread-y = $(gcov-files-test-aio-y)
+gcov-files-test-aio-multithread-y += util/qemu-coroutine.c tests/iothread.c
 check-unit-y += tests/test-throttle$(EXESUF)
-gcov-files-test-aio-$(CONFIG_WIN32) = aio-win32.c
-gcov-files-test-aio-$(CONFIG_POSIX) = aio-posix.c
 check-unit-y += tests/test-thread-pool$(EXESUF)
 gcov-files-test-thread-pool-y = thread-pool.c
 gcov-files-test-hbitmap-y = util/hbitmap.c
@@ -XXX,XX +XXX,XX @@ test-qapi-obj-y = tests/test-qapi-visit.o tests/test-qapi-types.o \
 	$(test-qom-obj-y)
 test-crypto-obj-y = $(crypto-obj-y) $(test-qom-obj-y)
 test-io-obj-y = $(io-obj-y) $(test-crypto-obj-y)
-test-block-obj-y = $(block-obj-y) $(test-io-obj-y)
+test-block-obj-y = $(block-obj-y) $(test-io-obj-y) tests/iothread.o
 
 tests/check-qint$(EXESUF): tests/check-qint.o $(test-util-obj-y)
 tests/check-qstring$(EXESUF): tests/check-qstring.o $(test-util-obj-y)
@@ -XXX,XX +XXX,XX @@ tests/check-qom-proplist$(EXESUF): tests/check-qom-proplist.o $(test-qom-obj-y)
 tests/test-char$(EXESUF): tests/test-char.o $(test-util-obj-y) $(qtest-obj-y) $(test-io-obj-y) $(chardev-obj-y)
 tests/test-coroutine$(EXESUF): tests/test-coroutine.o $(test-block-obj-y)
 tests/test-aio$(EXESUF): tests/test-aio.o $(test-block-obj-y)
+tests/test-aio-multithread$(EXESUF): tests/test-aio-multithread.o $(test-block-obj-y)
 tests/test-throttle$(EXESUF): tests/test-throttle.o $(test-block-obj-y)
 tests/test-blockjob$(EXESUF): tests/test-blockjob.o $(test-block-obj-y) $(test-util-obj-y)
 tests/test-blockjob-txn$(EXESUF): tests/test-blockjob-txn.o $(test-block-obj-y) $(test-util-obj-y)
diff --git a/include/block/aio.h b/include/block/aio.h
index XXXXXXX..XXXXXXX 100644
--- a/include/block/aio.h
+++ b/include/block/aio.h
@@ -XXX,XX +XXX,XX @@ typedef void QEMUBHFunc(void *opaque);
 typedef bool AioPollFn(void *opaque);
 typedef void IOHandler(void *opaque);
 
+struct Coroutine;
 struct ThreadPool;
 struct LinuxAioState;
 
@@ -XXX,XX +XXX,XX @@ struct AioContext {
     bool notified;
     EventNotifier notifier;
 
+    QSLIST_HEAD(, Coroutine) scheduled_coroutines;
+    QEMUBH *co_schedule_bh;
+
     /* Thread pool for performing work and receiving completion callbacks.
      * Has its own locking.
      */
@@ -XXX,XX +XXX,XX @@ static inline bool aio_node_check(AioContext *ctx, bool is_external)
 }
 
 /**
+ * aio_co_schedule:
+ * @ctx: the aio context
+ * @co: the coroutine
+ *
+ * Start a coroutine on a remote AioContext.
+ *
+ * The coroutine must not be entered by anyone else while aio_co_schedule()
+ * is active.  In addition the coroutine must have yielded unless ctx
+ * is the context in which the coroutine is running (i.e. the value of
+ * qemu_get_current_aio_context() from the coroutine itself).
+ */
+void aio_co_schedule(AioContext *ctx, struct Coroutine *co);
+
+/**
+ * aio_co_wake:
+ * @co: the coroutine
+ *
+ * Restart a coroutine on the AioContext where it was running last, thus
+ * preventing coroutines from jumping from one context to another when they
+ * go to sleep.
+ *
+ * aio_co_wake may be executed either in coroutine or non-coroutine
+ * context.  The coroutine must not be entered by anyone else while
+ * aio_co_wake() is active.
+ */
+void aio_co_wake(struct Coroutine *co);
+
+/**
  * Return the AioContext whose event loop runs in the current thread.
  *
  * If called from an IOThread this will be the IOThread's AioContext.  If
diff --git a/include/qemu/coroutine_int.h b/include/qemu/coroutine_int.h
index XXXXXXX..XXXXXXX 100644
--- a/include/qemu/coroutine_int.h
+++ b/include/qemu/coroutine_int.h
@@ -XXX,XX +XXX,XX @@ struct Coroutine {
     CoroutineEntry *entry;
     void *entry_arg;
     Coroutine *caller;
+
+    /* Only used when the coroutine has terminated.  */
     QSLIST_ENTRY(Coroutine) pool_next;
+
     size_t locks_held;
 
-    /* Coroutines that should be woken up when we yield or terminate */
+    /* Coroutines that should be woken up when we yield or terminate.
+     * Only used when the coroutine is running.
+     */
     QSIMPLEQ_HEAD(, Coroutine) co_queue_wakeup;
+
+    /* Only used when the coroutine has yielded.  */
+    AioContext *ctx;
     QSIMPLEQ_ENTRY(Coroutine) co_queue_next;
+    QSLIST_ENTRY(Coroutine) co_scheduled_next;
 };
 
 Coroutine *qemu_coroutine_new(void);
diff --git a/tests/iothread.h b/tests/iothread.h
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/tests/iothread.h
@@ -XXX,XX +XXX,XX @@
+/*
+ * Event loop thread implementation for unit tests
+ *
+ * Copyright Red Hat Inc., 2013, 2016
+ *
+ * Authors:
+ *  Stefan Hajnoczi   <stefanha@redhat.com>
+ *  Paolo Bonzini     <pbonzini@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+#ifndef TEST_IOTHREAD_H
+#define TEST_IOTHREAD_H
+
+#include "block/aio.h"
+#include "qemu/thread.h"
+
+typedef struct IOThread IOThread;
+
+IOThread *iothread_new(void);
+void iothread_join(IOThread *iothread);
+AioContext *iothread_get_aio_context(IOThread *iothread);
+
+#endif
diff --git a/tests/iothread.c b/tests/iothread.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/tests/iothread.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * Event loop thread implementation for unit tests
+ *
+ * Copyright Red Hat Inc., 2013, 2016
+ *
+ * Authors:
+ *  Stefan Hajnoczi   <stefanha@redhat.com>
+ *  Paolo Bonzini     <pbonzini@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qapi/error.h"
+#include "block/aio.h"
+#include "qemu/main-loop.h"
+#include "qemu/rcu.h"
+#include "iothread.h"
+
+struct IOThread {
+    AioContext *ctx;
+
+    QemuThread thread;
+    QemuMutex init_done_lock;
+    QemuCond init_done_cond;    /* is thread initialization done? */
+    bool stopping;
+};
+
+static __thread IOThread *my_iothread;
+
+AioContext *qemu_get_current_aio_context(void)
+{
+    return my_iothread ? my_iothread->ctx : qemu_get_aio_context();
+}
+
+static void *iothread_run(void *opaque)
+{
+    IOThread *iothread = opaque;
+
+    rcu_register_thread();
+
+    my_iothread = iothread;
+    qemu_mutex_lock(&iothread->init_done_lock);
+    iothread->ctx = aio_context_new(&error_abort);
+    qemu_cond_signal(&iothread->init_done_cond);
+    qemu_mutex_unlock(&iothread->init_done_lock);
+
+    while (!atomic_read(&iothread->stopping)) {
+        aio_poll(iothread->ctx, true);
+    }
+
+    rcu_unregister_thread();
+    return NULL;
+}
+
+void iothread_join(IOThread *iothread)
+{
+    iothread->stopping = true;
+    aio_notify(iothread->ctx);
+    qemu_thread_join(&iothread->thread);
+    qemu_cond_destroy(&iothread->init_done_cond);
+    qemu_mutex_destroy(&iothread->init_done_lock);
+    aio_context_unref(iothread->ctx);
+    g_free(iothread);
+}
+
+IOThread *iothread_new(void)
+{
+    IOThread *iothread = g_new0(IOThread, 1);
+
+    qemu_mutex_init(&iothread->init_done_lock);
+    qemu_cond_init(&iothread->init_done_cond);
+    qemu_thread_create(&iothread->thread, NULL, iothread_run,
+                       iothread, QEMU_THREAD_JOINABLE);
+
+    /* Wait for initialization to complete */
+    qemu_mutex_lock(&iothread->init_done_lock);
+    while (iothread->ctx == NULL) {
+        qemu_cond_wait(&iothread->init_done_cond,
+                       &iothread->init_done_lock);
+    }
+    qemu_mutex_unlock(&iothread->init_done_lock);
+    return iothread;
+}
+
+AioContext *iothread_get_aio_context(IOThread *iothread)
+{
+    return iothread->ctx;
+}
diff --git a/tests/test-aio-multithread.c b/tests/test-aio-multithread.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/tests/test-aio-multithread.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * AioContext multithreading tests
+ *
+ * Copyright Red Hat, Inc. 2016
+ *
+ * Authors:
+ *  Paolo Bonzini    <pbonzini@redhat.com>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2 or later.
+ * See the COPYING.LIB file in the top-level directory.
+ */
+
+#include "qemu/osdep.h"
+#include <glib.h>
+#include "block/aio.h"
+#include "qapi/error.h"
+#include "qemu/coroutine.h"
+#include "qemu/thread.h"
+#include "qemu/error-report.h"
+#include "iothread.h"
+
+/* AioContext management */
+
+#define NUM_CONTEXTS 5
+
+static IOThread *threads[NUM_CONTEXTS];
+static AioContext *ctx[NUM_CONTEXTS];
+static __thread int id = -1;
+
+static QemuEvent done_event;
+
+/* Run a function synchronously on a remote iothread. */
+
+typedef struct CtxRunData {
+    QEMUBHFunc *cb;
+    void *arg;
+} CtxRunData;
+
+static void ctx_run_bh_cb(void *opaque)
+{
+    CtxRunData *data = opaque;
+
+    data->cb(data->arg);
+    qemu_event_set(&done_event);
+}
+
+static void ctx_run(int i, QEMUBHFunc *cb, void *opaque)
+{
+    CtxRunData data = {
+        .cb = cb,
+        .arg = opaque
+    };
+
+    qemu_event_reset(&done_event);
+    aio_bh_schedule_oneshot(ctx[i], ctx_run_bh_cb, &data);
+    qemu_event_wait(&done_event);
+}
+
+/* Starting the iothreads. */
+
+static void set_id_cb(void *opaque)
+{
+    int *i = opaque;
+
+    id = *i;
+}
+
+static void create_aio_contexts(void)
+{
+    int i;
+
+    for (i = 0; i < NUM_CONTEXTS; i++) {
+        threads[i] = iothread_new();
+        ctx[i] = iothread_get_aio_context(threads[i]);
+    }
+
+    qemu_event_init(&done_event, false);
+    for (i = 0; i < NUM_CONTEXTS; i++) {
+        ctx_run(i, set_id_cb, &i);
+    }
+}
+
+/* Stopping the iothreads. */
+
+static void join_aio_contexts(void)
+{
+    int i;
+
+    for (i = 0; i < NUM_CONTEXTS; i++) {
+        aio_context_ref(ctx[i]);
+    }
+    for (i = 0; i < NUM_CONTEXTS; i++) {
+        iothread_join(threads[i]);
+    }
+    for (i = 0; i < NUM_CONTEXTS; i++) {
+        aio_context_unref(ctx[i]);
+    }
+    qemu_event_destroy(&done_event);
+}
+
+/* Basic test for the stuff above. */
+
+static void test_lifecycle(void)
+{
+    create_aio_contexts();
+    join_aio_contexts();
+}
+
+/* aio_co_schedule test.  */
+
+static Coroutine *to_schedule[NUM_CONTEXTS];
+
+static bool now_stopping;
+
+static int count_retry;
+static int count_here;
+static int count_other;
+
+static bool schedule_next(int n)
+{
+    Coroutine *co;
+
+    co = atomic_xchg(&to_schedule[n], NULL);
+    if (!co) {
+        atomic_inc(&count_retry);
+        return false;
+    }
+
+    if (n == id) {
+        atomic_inc(&count_here);
+    } else {
+        atomic_inc(&count_other);
+    }
+
+    aio_co_schedule(ctx[n], co);
+    return true;
+}
+
+static void finish_cb(void *opaque)
+{
+    schedule_next(id);
+}
+
+static coroutine_fn void test_multi_co_schedule_entry(void *opaque)
+{
+    g_assert(to_schedule[id] == NULL);
+    atomic_mb_set(&to_schedule[id], qemu_coroutine_self());
+
+    while (!atomic_mb_read(&now_stopping)) {
+        int n;
+
+        n = g_test_rand_int_range(0, NUM_CONTEXTS);
+        schedule_next(n);
+        qemu_coroutine_yield();
+
+        g_assert(to_schedule[id] == NULL);
+        atomic_mb_set(&to_schedule[id], qemu_coroutine_self());
+    }
+}
+
+
+static void test_multi_co_schedule(int seconds)
+{
+    int i;
+
+    count_here = count_other = count_retry = 0;
+    now_stopping = false;
+
+    create_aio_contexts();
+    for (i = 0; i < NUM_CONTEXTS; i++) {
+        Coroutine *co1 = qemu_coroutine_create(test_multi_co_schedule_entry, NULL);
+        aio_co_schedule(ctx[i], co1);
+    }
+
+    g_usleep(seconds * 1000000);
+
+    atomic_mb_set(&now_stopping, true);
+    for (i = 0; i < NUM_CONTEXTS; i++) {
+        ctx_run(i, finish_cb, NULL);
+        to_schedule[i] = NULL;
+    }
+
+    join_aio_contexts();
+    g_test_message("scheduled %d, queued %d, retry %d, total %d\n",
+                  count_other, count_here, count_retry,
+                  count_here + count_other + count_retry);
+}
+
+static void test_multi_co_schedule_1(void)
+{
+    test_multi_co_schedule(1);
+}
+
+static void test_multi_co_schedule_10(void)
+{
+    test_multi_co_schedule(10);
+}
+
+/* End of tests.  */
+
+int main(int argc, char **argv)
+{
+    init_clocks();
+
+    g_test_init(&argc, &argv, NULL);
+    g_test_add_func("/aio/multi/lifecycle", test_lifecycle);
+    if (g_test_quick()) {
+        g_test_add_func("/aio/multi/schedule", test_multi_co_schedule_1);
+    } else {
+        g_test_add_func("/aio/multi/schedule", test_multi_co_schedule_10);
+    }
+    return g_test_run();
+}
diff --git a/util/async.c b/util/async.c
index XXXXXXX..XXXXXXX 100644
--- a/util/async.c
+++ b/util/async.c
@@ -XXX,XX +XXX,XX @@
 #include "qemu/main-loop.h"
 #include "qemu/atomic.h"
 #include "block/raw-aio.h"
+#include "qemu/coroutine_int.h"
+#include "trace.h"
 
 /***********************************************************/
 /* bottom halves (can be seen as timers which expire ASAP) */
@@ -XXX,XX +XXX,XX @@ aio_ctx_finalize(GSource     *source)
     }
 #endif
 
+    assert(QSLIST_EMPTY(&ctx->scheduled_coroutines));
+    qemu_bh_delete(ctx->co_schedule_bh);
+
     qemu_lockcnt_lock(&ctx->list_lock);
     assert(!qemu_lockcnt_count(&ctx->list_lock));
     while (ctx->first_bh) {
@@ -XXX,XX +XXX,XX @@ static bool event_notifier_poll(void *opaque)
     return atomic_read(&ctx->notified);
 }
 
+static void co_schedule_bh_cb(void *opaque)
+{
+    AioContext *ctx = opaque;
+    QSLIST_HEAD(, Coroutine) straight, reversed;
+
+    QSLIST_MOVE_ATOMIC(&reversed, &ctx->scheduled_coroutines);
+    QSLIST_INIT(&straight);
+
+    while (!QSLIST_EMPTY(&reversed)) {
+        Coroutine *co = QSLIST_FIRST(&reversed);
+        QSLIST_REMOVE_HEAD(&reversed, co_scheduled_next);
+        QSLIST_INSERT_HEAD(&straight, co, co_scheduled_next);
+    }
+
+    while (!QSLIST_EMPTY(&straight)) {
+        Coroutine *co = QSLIST_FIRST(&straight);
+        QSLIST_REMOVE_HEAD(&straight, co_scheduled_next);
+        trace_aio_co_schedule_bh_cb(ctx, co);
+        qemu_coroutine_enter(co);
+    }
+}
+
 AioContext *aio_context_new(Error **errp)
 {
     int ret;
@@ -XXX,XX +XXX,XX @@ AioContext *aio_context_new(Error **errp)
     }
     g_source_set_can_recurse(&ctx->source, true);
     qemu_lockcnt_init(&ctx->list_lock);
+
+    ctx->co_schedule_bh = aio_bh_new(ctx, co_schedule_bh_cb, ctx);
+    QSLIST_INIT(&ctx->scheduled_coroutines);
+
     aio_set_event_notifier(ctx, &ctx->notifier,
                            false,
                            (EventNotifierHandler *)
@@ -XXX,XX +XXX,XX @@ fail:
     return NULL;
 }
 
+void aio_co_schedule(AioContext *ctx, Coroutine *co)
+{
+    trace_aio_co_schedule(ctx, co);
+    QSLIST_INSERT_HEAD_ATOMIC(&ctx->scheduled_coroutines,
+                              co, co_scheduled_next);
+    qemu_bh_schedule(ctx->co_schedule_bh);
+}
+
+void aio_co_wake(struct Coroutine *co)
+{
+    AioContext *ctx;
+
+    /* Read coroutine before co->ctx.  Matches smp_wmb in
+     * qemu_coroutine_enter.
+     */
+    smp_read_barrier_depends();
+    ctx = atomic_read(&co->ctx);
+
+    if (ctx != qemu_get_current_aio_context()) {
+        aio_co_schedule(ctx, co);
+        return;
+    }
+
+    if (qemu_in_coroutine()) {
+        Coroutine *self = qemu_coroutine_self();
+        assert(self != co);
+        QSIMPLEQ_INSERT_TAIL(&self->co_queue_wakeup, co, co_queue_next);
+    } else {
+        aio_context_acquire(ctx);
+        qemu_coroutine_enter(co);
+        aio_context_release(ctx);
+    }
+}
+
 void aio_context_ref(AioContext *ctx)
 {
     g_source_ref(&ctx->source);
diff --git a/util/qemu-coroutine.c b/util/qemu-coroutine.c
index XXXXXXX..XXXXXXX 100644
--- a/util/qemu-coroutine.c
+++ b/util/qemu-coroutine.c
@@ -XXX,XX +XXX,XX @@
 #include "qemu/atomic.h"
 #include "qemu/coroutine.h"
 #include "qemu/coroutine_int.h"
+#include "block/aio.h"
 
 enum {
     POOL_BATCH_SIZE = 64,
@@ -XXX,XX +XXX,XX @@ void qemu_coroutine_enter(Coroutine *co)
     }
 
     co->caller = self;
+    co->ctx = qemu_get_current_aio_context();
+
+    /* Store co->ctx before anything that stores co.  Matches
+     * barrier in aio_co_wake.
+     */
+    smp_wmb();
+
     ret = qemu_coroutine_switch(self, co, COROUTINE_ENTER);
 
     qemu_co_queue_run_restart(co);
diff --git a/util/trace-events b/util/trace-events
index XXXXXXX..XXXXXXX 100644
--- a/util/trace-events
+++ b/util/trace-events
@@ -XXX,XX +XXX,XX @@ run_poll_handlers_end(void *ctx, bool progress) "ctx %p progress %d"
 poll_shrink(void *ctx, int64_t old, int64_t new) "ctx %p old %"PRId64" new %"PRId64
 poll_grow(void *ctx, int64_t old, int64_t new) "ctx %p old %"PRId64" new %"PRId64
 
+# util/async.c
+aio_co_schedule(void *ctx, void *co) "ctx %p co %p"
+aio_co_schedule_bh_cb(void *ctx, void *co) "ctx %p co %p"
+
 # util/thread-pool.c
 thread_pool_submit(void *pool, void *req, void *opaque) "pool %p req %p opaque %p"
 thread_pool_complete(void *pool, void *req, void *opaque, int ret) "pool %p req %p opaque %p ret %d"
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

qcow2_create2 calls this.  Do not run a nested event loop, as that
breaks when aio_co_wake tries to queue the coroutine on the co_queue_wakeup
list of the currently running one.

Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 20170213135235.12274-4-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/block-backend.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

diff --git a/block/block-backend.c b/block/block-backend.c
index XXXXXXX..XXXXXXX 100644
--- a/block/block-backend.c
+++ b/block/block-backend.c
@@ -XXX,XX +XXX,XX @@ static int blk_prw(BlockBackend *blk, int64_t offset, uint8_t *buf,
 {
     QEMUIOVector qiov;
     struct iovec iov;
-    Coroutine *co;
     BlkRwCo rwco;
 
     iov = (struct iovec) {
@@ -XXX,XX +XXX,XX @@ static int blk_prw(BlockBackend *blk, int64_t offset, uint8_t *buf,
         .ret    = NOT_DONE,
     };
 
-    co = qemu_coroutine_create(co_entry, &rwco);
-    qemu_coroutine_enter(co);
-    BDRV_POLL_WHILE(blk_bs(blk), rwco.ret == NOT_DONE);
+    if (qemu_in_coroutine()) {
+        /* Fast-path if already in coroutine context */
+        co_entry(&rwco);
+    } else {
+        Coroutine *co = qemu_coroutine_create(co_entry, &rwco);
+        qemu_coroutine_enter(co);
+        BDRV_POLL_WHILE(blk_bs(blk), rwco.ret == NOT_DONE);
+    }
 
     return rwco.ret;
 }
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

Once the thread pool starts using aio_co_wake, it will also need
qemu_get_current_aio_context().  Make test-thread-pool create
an AioContext with qemu_init_main_loop, so that stubs/iothread.c
and tests/iothread.c can provide the rest.

Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 20170213135235.12274-5-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 tests/test-thread-pool.c | 12 +++---------
 1 file changed, 3 insertions(+), 9 deletions(-)

diff --git a/tests/test-thread-pool.c b/tests/test-thread-pool.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/test-thread-pool.c
+++ b/tests/test-thread-pool.c
@@ -XXX,XX +XXX,XX @@
 #include "qapi/error.h"
 #include "qemu/timer.h"
 #include "qemu/error-report.h"
+#include "qemu/main-loop.h"
 
 static AioContext *ctx;
 static ThreadPool *pool;
@@ -XXX,XX +XXX,XX @@ static void test_cancel_async(void)
 int main(int argc, char **argv)
 {
     int ret;
-    Error *local_error = NULL;
 
-    init_clocks();
-
-    ctx = aio_context_new(&local_error);
-    if (!ctx) {
-        error_reportf_err(local_error, "Failed to create AIO Context: ");
-        exit(1);
-    }
+    qemu_init_main_loop(&error_abort);
+    ctx = qemu_get_current_aio_context();
     pool = aio_get_thread_pool(ctx);
 
     g_test_init(&argc, &argv, NULL);
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
 
     ret = g_test_run();
 
-    aio_context_unref(ctx);
     return ret;
 }
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

This is in preparation for making qio_channel_yield work on
AioContexts other than the main one.

Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 20170213135235.12274-6-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/io/channel.h | 25 +++++++++++++++++++++++++
 io/channel-command.c | 13 +++++++++++++
 io/channel-file.c    | 11 +++++++++++
 io/channel-socket.c  | 16 +++++++++++-----
 io/channel-tls.c     | 12 ++++++++++++
 io/channel-watch.c   |  6 ++++++
 io/channel.c         | 11 +++++++++++
 7 files changed, 89 insertions(+), 5 deletions(-)

diff --git a/include/io/channel.h b/include/io/channel.h
index XXXXXXX..XXXXXXX 100644
--- a/include/io/channel.h
+++ b/include/io/channel.h
@@ -XXX,XX +XXX,XX @@
 
 #include "qemu-common.h"
 #include "qom/object.h"
+#include "block/aio.h"
 
 #define TYPE_QIO_CHANNEL "qio-channel"
 #define QIO_CHANNEL(obj)                                    \
@@ -XXX,XX +XXX,XX @@ struct QIOChannelClass {
                      off_t offset,
                      int whence,
                      Error **errp);
+    void (*io_set_aio_fd_handler)(QIOChannel *ioc,
+                                  AioContext *ctx,
+                                  IOHandler *io_read,
+                                  IOHandler *io_write,
+                                  void *opaque);
 };
 
 /* General I/O handling functions */
@@ -XXX,XX +XXX,XX @@ void qio_channel_yield(QIOChannel *ioc,
 void qio_channel_wait(QIOChannel *ioc,
                       GIOCondition condition);
 
+/**
+ * qio_channel_set_aio_fd_handler:
+ * @ioc: the channel object
+ * @ctx: the AioContext to set the handlers on
+ * @io_read: the read handler
+ * @io_write: the write handler
+ * @opaque: the opaque value passed to the handler
+ *
+ * This is used internally by qio_channel_yield().  It can
+ * be used by channel implementations to forward the handlers
+ * to another channel (e.g. from #QIOChannelTLS to the
+ * underlying socket).
+ */
+void qio_channel_set_aio_fd_handler(QIOChannel *ioc,
+                                    AioContext *ctx,
+                                    IOHandler *io_read,
+                                    IOHandler *io_write,
+                                    void *opaque);
+
 #endif /* QIO_CHANNEL_H */
diff --git a/io/channel-command.c b/io/channel-command.c
index XXXXXXX..XXXXXXX 100644
--- a/io/channel-command.c
+++ b/io/channel-command.c
@@ -XXX,XX +XXX,XX @@ static int qio_channel_command_close(QIOChannel *ioc,
 }
 
 
+static void qio_channel_command_set_aio_fd_handler(QIOChannel *ioc,
+                                                   AioContext *ctx,
+                                                   IOHandler *io_read,
+                                                   IOHandler *io_write,
+                                                   void *opaque)
+{
+    QIOChannelCommand *cioc = QIO_CHANNEL_COMMAND(ioc);
+    aio_set_fd_handler(ctx, cioc->readfd, false, io_read, NULL, NULL, opaque);
+    aio_set_fd_handler(ctx, cioc->writefd, false, NULL, io_write, NULL, opaque);
+}
+
+
 static GSource *qio_channel_command_create_watch(QIOChannel *ioc,
                                                  GIOCondition condition)
 {
@@ -XXX,XX +XXX,XX @@ static void qio_channel_command_class_init(ObjectClass *klass,
     ioc_klass->io_set_blocking = qio_channel_command_set_blocking;
     ioc_klass->io_close = qio_channel_command_close;
     ioc_klass->io_create_watch = qio_channel_command_create_watch;
+    ioc_klass->io_set_aio_fd_handler = qio_channel_command_set_aio_fd_handler;
 }
 
 static const TypeInfo qio_channel_command_info = {
diff --git a/io/channel-file.c b/io/channel-file.c
index XXXXXXX..XXXXXXX 100644
--- a/io/channel-file.c
+++ b/io/channel-file.c
@@ -XXX,XX +XXX,XX @@ static int qio_channel_file_close(QIOChannel *ioc,
 }
 
 
+static void qio_channel_file_set_aio_fd_handler(QIOChannel *ioc,
+                                                AioContext *ctx,
+                                                IOHandler *io_read,
+                                                IOHandler *io_write,
+                                                void *opaque)
+{
+    QIOChannelFile *fioc = QIO_CHANNEL_FILE(ioc);
+    aio_set_fd_handler(ctx, fioc->fd, false, io_read, io_write, NULL, opaque);
+}
+
 static GSource *qio_channel_file_create_watch(QIOChannel *ioc,
                                               GIOCondition condition)
 {
@@ -XXX,XX +XXX,XX @@ static void qio_channel_file_class_init(ObjectClass *klass,
     ioc_klass->io_seek = qio_channel_file_seek;
     ioc_klass->io_close = qio_channel_file_close;
     ioc_klass->io_create_watch = qio_channel_file_create_watch;
+    ioc_klass->io_set_aio_fd_handler = qio_channel_file_set_aio_fd_handler;
 }
 
 static const TypeInfo qio_channel_file_info = {
diff --git a/io/channel-socket.c b/io/channel-socket.c
index XXXXXXX..XXXXXXX 100644
--- a/io/channel-socket.c
+++ b/io/channel-socket.c
@@ -XXX,XX +XXX,XX @@ qio_channel_socket_set_blocking(QIOChannel *ioc,
         qemu_set_block(sioc->fd);
     } else {
         qemu_set_nonblock(sioc->fd);
-#ifdef WIN32
-        WSAEventSelect(sioc->fd, ioc->event,
-                       FD_READ | FD_ACCEPT | FD_CLOSE |
-                       FD_CONNECT | FD_WRITE | FD_OOB);
-#endif
     }
     return 0;
 }
@@ -XXX,XX +XXX,XX @@ qio_channel_socket_shutdown(QIOChannel *ioc,
     return 0;
 }
 
+static void qio_channel_socket_set_aio_fd_handler(QIOChannel *ioc,
+                                                  AioContext *ctx,
+                                                  IOHandler *io_read,
+                                                  IOHandler *io_write,
+                                                  void *opaque)
+{
+    QIOChannelSocket *sioc = QIO_CHANNEL_SOCKET(ioc);
+    aio_set_fd_handler(ctx, sioc->fd, false, io_read, io_write, NULL, opaque);
+}
+
 static GSource *qio_channel_socket_create_watch(QIOChannel *ioc,
                                                 GIOCondition condition)
 {
@@ -XXX,XX +XXX,XX @@ static void qio_channel_socket_class_init(ObjectClass *klass,
     ioc_klass->io_set_cork = qio_channel_socket_set_cork;
     ioc_klass->io_set_delay = qio_channel_socket_set_delay;
     ioc_klass->io_create_watch = qio_channel_socket_create_watch;
+    ioc_klass->io_set_aio_fd_handler = qio_channel_socket_set_aio_fd_handler;
 }
 
 static const TypeInfo qio_channel_socket_info = {
diff --git a/io/channel-tls.c b/io/channel-tls.c
index XXXXXXX..XXXXXXX 100644
--- a/io/channel-tls.c
+++ b/io/channel-tls.c
@@ -XXX,XX +XXX,XX @@ static int qio_channel_tls_close(QIOChannel *ioc,
     return qio_channel_close(tioc->master, errp);
 }
 
+static void qio_channel_tls_set_aio_fd_handler(QIOChannel *ioc,
+                                               AioContext *ctx,
+                                               IOHandler *io_read,
+                                               IOHandler *io_write,
+                                               void *opaque)
+{
+    QIOChannelTLS *tioc = QIO_CHANNEL_TLS(ioc);
+
+    qio_channel_set_aio_fd_handler(tioc->master, ctx, io_read, io_write, opaque);
+}
+
 static GSource *qio_channel_tls_create_watch(QIOChannel *ioc,
                                              GIOCondition condition)
 {
@@ -XXX,XX +XXX,XX @@ static void qio_channel_tls_class_init(ObjectClass *klass,
     ioc_klass->io_close = qio_channel_tls_close;
     ioc_klass->io_shutdown = qio_channel_tls_shutdown;
     ioc_klass->io_create_watch = qio_channel_tls_create_watch;
+    ioc_klass->io_set_aio_fd_handler = qio_channel_tls_set_aio_fd_handler;
 }
 
 static const TypeInfo qio_channel_tls_info = {
diff --git a/io/channel-watch.c b/io/channel-watch.c
index XXXXXXX..XXXXXXX 100644
--- a/io/channel-watch.c
+++ b/io/channel-watch.c
@@ -XXX,XX +XXX,XX @@ GSource *qio_channel_create_socket_watch(QIOChannel *ioc,
     GSource *source;
     QIOChannelSocketSource *ssource;
 
+#ifdef WIN32
+    WSAEventSelect(socket, ioc->event,
+                   FD_READ | FD_ACCEPT | FD_CLOSE |
+                   FD_CONNECT | FD_WRITE | FD_OOB);
+#endif
+
     source = g_source_new(&qio_channel_socket_source_funcs,
                           sizeof(QIOChannelSocketSource));
     ssource = (QIOChannelSocketSource *)source;
diff --git a/io/channel.c b/io/channel.c
index XXXXXXX..XXXXXXX 100644
--- a/io/channel.c
+++ b/io/channel.c
@@ -XXX,XX +XXX,XX @@ GSource *qio_channel_create_watch(QIOChannel *ioc,
 }
 
 
+void qio_channel_set_aio_fd_handler(QIOChannel *ioc,
+                                    AioContext *ctx,
+                                    IOHandler *io_read,
+                                    IOHandler *io_write,
+                                    void *opaque)
+{
+    QIOChannelClass *klass = QIO_CHANNEL_GET_CLASS(ioc);
+
+    klass->io_set_aio_fd_handler(ioc, ctx, io_read, io_write, opaque);
+}
+
 guint qio_channel_add_watch(QIOChannel *ioc,
                             GIOCondition condition,
                             QIOChannelFunc func,
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

Support separate coroutines for reading and writing, and place the
read/write handlers on the AioContext that the QIOChannel is registered
with.

Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 20170213135235.12274-7-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/io/channel.h | 47 ++++++++++++++++++++++++++--
 io/channel.c         | 86 +++++++++++++++++++++++++++++++++++++++-------------
 2 files changed, 109 insertions(+), 24 deletions(-)

diff --git a/include/io/channel.h b/include/io/channel.h
index XXXXXXX..XXXXXXX 100644
--- a/include/io/channel.h
+++ b/include/io/channel.h
@@ -XXX,XX +XXX,XX @@
 
 #include "qemu-common.h"
 #include "qom/object.h"
+#include "qemu/coroutine.h"
 #include "block/aio.h"
 
 #define TYPE_QIO_CHANNEL "qio-channel"
@@ -XXX,XX +XXX,XX @@ struct QIOChannel {
     Object parent;
     unsigned int features; /* bitmask of QIOChannelFeatures */
     char *name;
+    AioContext *ctx;
+    Coroutine *read_coroutine;
+    Coroutine *write_coroutine;
 #ifdef _WIN32
     HANDLE event; /* For use with GSource on Win32 */
 #endif
@@ -XXX,XX +XXX,XX @@ guint qio_channel_add_watch(QIOChannel *ioc,
 
 
 /**
+ * qio_channel_attach_aio_context:
+ * @ioc: the channel object
+ * @ctx: the #AioContext to set the handlers on
+ *
+ * Request that qio_channel_yield() sets I/O handlers on
+ * the given #AioContext.  If @ctx is %NULL, qio_channel_yield()
+ * uses QEMU's main thread event loop.
+ *
+ * You can move a #QIOChannel from one #AioContext to another even if
+ * I/O handlers are set for a coroutine.  However, #QIOChannel provides
+ * no synchronization between the calls to qio_channel_yield() and
+ * qio_channel_attach_aio_context().
+ *
+ * Therefore you should first call qio_channel_detach_aio_context()
+ * to ensure that the coroutine is not entered concurrently.  Then,
+ * while the coroutine has yielded, call qio_channel_attach_aio_context(),
+ * and then aio_co_schedule() to place the coroutine on the new
+ * #AioContext.  The calls to qio_channel_detach_aio_context()
+ * and qio_channel_attach_aio_context() should be protected with
+ * aio_context_acquire() and aio_context_release().
+ */
+void qio_channel_attach_aio_context(QIOChannel *ioc,
+                                    AioContext *ctx);
+
+/**
+ * qio_channel_detach_aio_context:
+ * @ioc: the channel object
+ *
+ * Disable any I/O handlers set by qio_channel_yield().  With the
+ * help of aio_co_schedule(), this allows moving a coroutine that was
+ * paused by qio_channel_yield() to another context.
+ */
+void qio_channel_detach_aio_context(QIOChannel *ioc);
+
+/**
  * qio_channel_yield:
  * @ioc: the channel object
  * @condition: the I/O condition to wait for
  *
- * Yields execution from the current coroutine until
- * the condition indicated by @condition becomes
- * available.
+ * Yields execution from the current coroutine until the condition
+ * indicated by @condition becomes available.  @condition must
+ * be either %G_IO_IN or %G_IO_OUT; it cannot contain both.  In
+ * addition, no two coroutine can be waiting on the same condition
+ * and channel at the same time.
  *
  * This must only be called from coroutine context
  */
diff --git a/io/channel.c b/io/channel.c
index XXXXXXX..XXXXXXX 100644
--- a/io/channel.c
+++ b/io/channel.c
@@ -XXX,XX +XXX,XX @@
 #include "qemu/osdep.h"
 #include "io/channel.h"
 #include "qapi/error.h"
-#include "qemu/coroutine.h"
+#include "qemu/main-loop.h"
 
 bool qio_channel_has_feature(QIOChannel *ioc,
                              QIOChannelFeature feature)
@@ -XXX,XX +XXX,XX @@ off_t qio_channel_io_seek(QIOChannel *ioc,
 }
 
 
-typedef struct QIOChannelYieldData QIOChannelYieldData;
-struct QIOChannelYieldData {
-    QIOChannel *ioc;
-    Coroutine *co;
-};
+static void qio_channel_set_aio_fd_handlers(QIOChannel *ioc);
 
+static void qio_channel_restart_read(void *opaque)
+{
+    QIOChannel *ioc = opaque;
+    Coroutine *co = ioc->read_coroutine;
+
+    ioc->read_coroutine = NULL;
+    qio_channel_set_aio_fd_handlers(ioc);
+    aio_co_wake(co);
+}
 
-static gboolean qio_channel_yield_enter(QIOChannel *ioc,
-                                        GIOCondition condition,
-                                        gpointer opaque)
+static void qio_channel_restart_write(void *opaque)
 {
-    QIOChannelYieldData *data = opaque;
-    qemu_coroutine_enter(data->co);
-    return FALSE;
+    QIOChannel *ioc = opaque;
+    Coroutine *co = ioc->write_coroutine;
+
+    ioc->write_coroutine = NULL;
+    qio_channel_set_aio_fd_handlers(ioc);
+    aio_co_wake(co);
 }
 
+static void qio_channel_set_aio_fd_handlers(QIOChannel *ioc)
+{
+    IOHandler *rd_handler = NULL, *wr_handler = NULL;
+    AioContext *ctx;
+
+    if (ioc->read_coroutine) {
+        rd_handler = qio_channel_restart_read;
+    }
+    if (ioc->write_coroutine) {
+        wr_handler = qio_channel_restart_write;
+    }
+
+    ctx = ioc->ctx ? ioc->ctx : iohandler_get_aio_context();
+    qio_channel_set_aio_fd_handler(ioc, ctx, rd_handler, wr_handler, ioc);
+}
+
+void qio_channel_attach_aio_context(QIOChannel *ioc,
+                                    AioContext *ctx)
+{
+    AioContext *old_ctx;
+    if (ioc->ctx == ctx) {
+        return;
+    }
+
+    old_ctx = ioc->ctx ? ioc->ctx : iohandler_get_aio_context();
+    qio_channel_set_aio_fd_handler(ioc, old_ctx, NULL, NULL, NULL);
+    ioc->ctx = ctx;
+    qio_channel_set_aio_fd_handlers(ioc);
+}
+
+void qio_channel_detach_aio_context(QIOChannel *ioc)
+{
+    ioc->read_coroutine = NULL;
+    ioc->write_coroutine = NULL;
+    qio_channel_set_aio_fd_handlers(ioc);
+    ioc->ctx = NULL;
+}
 
 void coroutine_fn qio_channel_yield(QIOChannel *ioc,
                                     GIOCondition condition)
 {
-    QIOChannelYieldData data;
-
     assert(qemu_in_coroutine());
-    data.ioc = ioc;
-    data.co = qemu_coroutine_self();
-    qio_channel_add_watch(ioc,
-                          condition,
-                          qio_channel_yield_enter,
-                          &data,
-                          NULL);
+    if (condition == G_IO_IN) {
+        assert(!ioc->read_coroutine);
+        ioc->read_coroutine = qemu_coroutine_self();
+    } else if (condition == G_IO_OUT) {
+        assert(!ioc->write_coroutine);
+        ioc->write_coroutine = qemu_coroutine_self();
+    } else {
+        abort();
+    }
+    qio_channel_set_aio_fd_handlers(ioc);
     qemu_coroutine_yield();
 }
 
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

In the client, read the reply headers from a coroutine, switching the
read side between the "read header" coroutine and the I/O coroutine that
reads the body of the reply.

In the server, if the server can read more requests it will create a new
"read request" coroutine as soon as a request has been read.  Otherwise,
the new coroutine is created in nbd_request_put.

diff --git a/block/nbd-client.h b/block/nbd-client.h
index XXXXXXX..XXXXXXX 100644
--- a/block/nbd-client.h
+++ b/block/nbd-client.h
@@ -XXX,XX +XXX,XX @@ typedef struct NBDClientSession {
 
     CoMutex send_mutex;
     CoQueue free_sema;
-    Coroutine *send_coroutine;
+    Coroutine *read_reply_co;
     int in_flight;
 
     Coroutine *recv_coroutine[MAX_NBD_REQUESTS];
diff --git a/block/nbd-client.c b/block/nbd-client.c
index XXXXXXX..XXXXXXX 100644
--- a/block/nbd-client.c
+++ b/block/nbd-client.c
@@ -XXX,XX +XXX,XX @@
 #define HANDLE_TO_INDEX(bs, handle) ((handle) ^ ((uint64_t)(intptr_t)bs))
 #define INDEX_TO_HANDLE(bs, index)  ((index)  ^ ((uint64_t)(intptr_t)bs))
 
-static void nbd_recv_coroutines_enter_all(NBDClientSession *s)
+static void nbd_recv_coroutines_enter_all(BlockDriverState *bs)
 {
+    NBDClientSession *s = nbd_get_client_session(bs);
     int i;
 
     for (i = 0; i < MAX_NBD_REQUESTS; i++) {
@@ -XXX,XX +XXX,XX @@ static void nbd_recv_coroutines_enter_all(NBDClientSession *s)
             qemu_coroutine_enter(s->recv_coroutine[i]);
         }
     }
+    BDRV_POLL_WHILE(bs, s->read_reply_co);
 }
 
 static void nbd_teardown_connection(BlockDriverState *bs)
@@ -XXX,XX +XXX,XX @@ static void nbd_teardown_connection(BlockDriverState *bs)
     qio_channel_shutdown(client->ioc,
                          QIO_CHANNEL_SHUTDOWN_BOTH,
                          NULL);
-    nbd_recv_coroutines_enter_all(client);
+    nbd_recv_coroutines_enter_all(bs);
 
     nbd_client_detach_aio_context(bs);
     object_unref(OBJECT(client->sioc));
@@ -XXX,XX +XXX,XX @@ static void nbd_teardown_connection(BlockDriverState *bs)
     client->ioc = NULL;
 }
 
-static void nbd_reply_ready(void *opaque)
+static coroutine_fn void nbd_read_reply_entry(void *opaque)
 {
-    BlockDriverState *bs = opaque;
-    NBDClientSession *s = nbd_get_client_session(bs);
+    NBDClientSession *s = opaque;
     uint64_t i;
     int ret;
 
-    if (!s->ioc) { /* Already closed */
-        return;
-    }
-
-    if (s->reply.handle == 0) {
-        /* No reply already in flight.  Fetch a header.  It is possible
-         * that another thread has done the same thing in parallel, so
-         * the socket is not readable anymore.
-         */
+    for (;;) {
+        assert(s->reply.handle == 0);
         ret = nbd_receive_reply(s->ioc, &s->reply);
-        if (ret == -EAGAIN) {
-            return;
-        }
         if (ret < 0) {
-            s->reply.handle = 0;
-            goto fail;
+            break;
         }
-    }
 
-    /* There's no need for a mutex on the receive side, because the
-     * handler acts as a synchronization point and ensures that only
-     * one coroutine is called until the reply finishes.  */
-    i = HANDLE_TO_INDEX(s, s->reply.handle);
-    if (i >= MAX_NBD_REQUESTS) {
-        goto fail;
-    }
+        /* There's no need for a mutex on the receive side, because the
+         * handler acts as a synchronization point and ensures that only
+         * one coroutine is called until the reply finishes.
+         */
+        i = HANDLE_TO_INDEX(s, s->reply.handle);
+        if (i >= MAX_NBD_REQUESTS || !s->recv_coroutine[i]) {
+            break;
+        }
 
-    if (s->recv_coroutine[i]) {
-        qemu_coroutine_enter(s->recv_coroutine[i]);
-        return;
+        /* We're woken up by the recv_coroutine itself.  Note that there
+         * is no race between yielding and reentering read_reply_co.  This
+         * is because:
+         *
+         * - if recv_coroutine[i] runs on the same AioContext, it is only
+         *   entered after we yield
+         *
+         * - if recv_coroutine[i] runs on a different AioContext, reentering
+         *   read_reply_co happens through a bottom half, which can only
+         *   run after we yield.
+         */
+        aio_co_wake(s->recv_coroutine[i]);
+        qemu_coroutine_yield();
     }
-
-fail:
-    nbd_teardown_connection(bs);
-}
-
-static void nbd_restart_write(void *opaque)
-{
-    BlockDriverState *bs = opaque;
-
-    qemu_coroutine_enter(nbd_get_client_session(bs)->send_coroutine);
+    s->read_reply_co = NULL;
 }
 
 static int nbd_co_send_request(BlockDriverState *bs,
@@ -XXX,XX +XXX,XX @@ static int nbd_co_send_request(BlockDriverState *bs,
                                QEMUIOVector *qiov)
 {
     NBDClientSession *s = nbd_get_client_session(bs);
-    AioContext *aio_context;
     int rc, ret, i;
 
     qemu_co_mutex_lock(&s->send_mutex);
@@ -XXX,XX +XXX,XX @@ static int nbd_co_send_request(BlockDriverState *bs,
         return -EPIPE;
     }
 
-    s->send_coroutine = qemu_coroutine_self();
-    aio_context = bdrv_get_aio_context(bs);
-
-    aio_set_fd_handler(aio_context, s->sioc->fd, false,
-                       nbd_reply_ready, nbd_restart_write, NULL, bs);
     if (qiov) {
         qio_channel_set_cork(s->ioc, true);
         rc = nbd_send_request(s->ioc, request);
@@ -XXX,XX +XXX,XX @@ static int nbd_co_send_request(BlockDriverState *bs,
     } else {
         rc = nbd_send_request(s->ioc, request);
     }
-    aio_set_fd_handler(aio_context, s->sioc->fd, false,
-                       nbd_reply_ready, NULL, NULL, bs);
-    s->send_coroutine = NULL;
     qemu_co_mutex_unlock(&s->send_mutex);
     return rc;
 }
@@ -XXX,XX +XXX,XX @@ static void nbd_co_receive_reply(NBDClientSession *s,
 {
     int ret;
 
-    /* Wait until we're woken up by the read handler.  TODO: perhaps
-     * peek at the next reply and avoid yielding if it's ours?  */
+    /* Wait until we're woken up by nbd_read_reply_entry.  */
     qemu_coroutine_yield();
     *reply = s->reply;
     if (reply->handle != request->handle ||
@@ -XXX,XX +XXX,XX @@ static void nbd_coroutine_start(NBDClientSession *s,
     /* s->recv_coroutine[i] is set as soon as we get the send_lock.  */
 }
 
-static void nbd_coroutine_end(NBDClientSession *s,
+static void nbd_coroutine_end(BlockDriverState *bs,
                               NBDRequest *request)
 {
+    NBDClientSession *s = nbd_get_client_session(bs);
     int i = HANDLE_TO_INDEX(s, request->handle);
+
     s->recv_coroutine[i] = NULL;
-    if (s->in_flight-- == MAX_NBD_REQUESTS) {
-        qemu_co_queue_next(&s->free_sema);
+    s->in_flight--;
+    qemu_co_queue_next(&s->free_sema);
+
+    /* Kick the read_reply_co to get the next reply.  */
+    if (s->read_reply_co) {
+        aio_co_wake(s->read_reply_co);
     }
 }
 
@@ -XXX,XX +XXX,XX @@ int nbd_client_co_preadv(BlockDriverState *bs, uint64_t offset,
     } else {
         nbd_co_receive_reply(client, &request, &reply, qiov);
     }
-    nbd_coroutine_end(client, &request);
+    nbd_coroutine_end(bs, &request);
     return -reply.error;
 }
 
@@ -XXX,XX +XXX,XX @@ int nbd_client_co_pwritev(BlockDriverState *bs, uint64_t offset,
     } else {
         nbd_co_receive_reply(client, &request, &reply, NULL);
     }
-    nbd_coroutine_end(client, &request);
+    nbd_coroutine_end(bs, &request);
     return -reply.error;
 }
 
@@ -XXX,XX +XXX,XX @@ int nbd_client_co_pwrite_zeroes(BlockDriverState *bs, int64_t offset,
     } else {
         nbd_co_receive_reply(client, &request, &reply, NULL);
     }
-    nbd_coroutine_end(client, &request);
+    nbd_coroutine_end(bs, &request);
     return -reply.error;
 }
 
@@ -XXX,XX +XXX,XX @@ int nbd_client_co_flush(BlockDriverState *bs)
     } else {
         nbd_co_receive_reply(client, &request, &reply, NULL);
     }
-    nbd_coroutine_end(client, &request);
+    nbd_coroutine_end(bs, &request);
     return -reply.error;
 }
 
@@ -XXX,XX +XXX,XX @@ int nbd_client_co_pdiscard(BlockDriverState *bs, int64_t offset, int count)
     } else {
         nbd_co_receive_reply(client, &request, &reply, NULL);
     }
-    nbd_coroutine_end(client, &request);
+    nbd_coroutine_end(bs, &request);
     return -reply.error;
 
 }
 
 void nbd_client_detach_aio_context(BlockDriverState *bs)
 {
-    aio_set_fd_handler(bdrv_get_aio_context(bs),
-                       nbd_get_client_session(bs)->sioc->fd,
-                       false, NULL, NULL, NULL, NULL);
+    NBDClientSession *client = nbd_get_client_session(bs);
+    qio_channel_detach_aio_context(QIO_CHANNEL(client->sioc));
 }
 
 void nbd_client_attach_aio_context(BlockDriverState *bs,
                                    AioContext *new_context)
 {
-    aio_set_fd_handler(new_context, nbd_get_client_session(bs)->sioc->fd,
-                       false, nbd_reply_ready, NULL, NULL, bs);
+    NBDClientSession *client = nbd_get_client_session(bs);
+    qio_channel_attach_aio_context(QIO_CHANNEL(client->sioc), new_context);
+    aio_co_schedule(new_context, client->read_reply_co);
 }
 
 void nbd_client_close(BlockDriverState *bs)
@@ -XXX,XX +XXX,XX @@ int nbd_client_init(BlockDriverState *bs,
     /* Now that we're connected, set the socket to be non-blocking and
      * kick the reply mechanism.  */
     qio_channel_set_blocking(QIO_CHANNEL(sioc), false, NULL);
-
+    client->read_reply_co = qemu_coroutine_create(nbd_read_reply_entry, client);
     nbd_client_attach_aio_context(bs, bdrv_get_aio_context(bs));
 
     logout("Established connection with NBD server\n");
diff --git a/nbd/client.c b/nbd/client.c
index XXXXXXX..XXXXXXX 100644
--- a/nbd/client.c
+++ b/nbd/client.c
@@ -XXX,XX +XXX,XX @@ ssize_t nbd_receive_reply(QIOChannel *ioc, NBDReply *reply)
     ssize_t ret;
 
     ret = read_sync(ioc, buf, sizeof(buf));
-    if (ret < 0) {
+    if (ret <= 0) {
         return ret;
     }
 
diff --git a/nbd/common.c b/nbd/common.c
index XXXXXXX..XXXXXXX 100644
--- a/nbd/common.c
+++ b/nbd/common.c
@@ -XXX,XX +XXX,XX @@ ssize_t nbd_wr_syncv(QIOChannel *ioc,
         }
         if (len == QIO_CHANNEL_ERR_BLOCK) {
             if (qemu_in_coroutine()) {
-                /* XXX figure out if we can create a variant on
-                 * qio_channel_yield() that works with AIO contexts
-                 * and consider using that in this branch */
-                qemu_coroutine_yield();
-            } else if (done) {
-                /* XXX this is needed by nbd_reply_ready.  */
-                qio_channel_wait(ioc,
-                                 do_read ? G_IO_IN : G_IO_OUT);
+                qio_channel_yield(ioc, do_read ? G_IO_IN : G_IO_OUT);
             } else {
                 return -EAGAIN;
             }
diff --git a/nbd/server.c b/nbd/server.c
index XXXXXXX..XXXXXXX 100644
--- a/nbd/server.c
+++ b/nbd/server.c
@@ -XXX,XX +XXX,XX @@ struct NBDClient {
     CoMutex send_lock;
     Coroutine *send_coroutine;
 
-    bool can_read;
-
     QTAILQ_ENTRY(NBDClient) next;
     int nb_requests;
     bool closing;
@@ -XXX,XX +XXX,XX @@ struct NBDClient {
 
 /* That's all folks */
 
-static void nbd_set_handlers(NBDClient *client);
-static void nbd_unset_handlers(NBDClient *client);
-static void nbd_update_can_read(NBDClient *client);
+static void nbd_client_receive_next_request(NBDClient *client);
 
 static gboolean nbd_negotiate_continue(QIOChannel *ioc,
                                        GIOCondition condition,
@@ -XXX,XX +XXX,XX @@ void nbd_client_put(NBDClient *client)
          */
         assert(client->closing);
 
-        nbd_unset_handlers(client);
+        qio_channel_detach_aio_context(client->ioc);
         object_unref(OBJECT(client->sioc));
         object_unref(OBJECT(client->ioc));
         if (client->tlscreds) {
@@ -XXX,XX +XXX,XX @@ static NBDRequestData *nbd_request_get(NBDClient *client)
 
     assert(client->nb_requests <= MAX_NBD_REQUESTS - 1);
     client->nb_requests++;
-    nbd_update_can_read(client);
 
     req = g_new0(NBDRequestData, 1);
     nbd_client_get(client);
@@ -XXX,XX +XXX,XX @@ static void nbd_request_put(NBDRequestData *req)
     g_free(req);
 
     client->nb_requests--;
-    nbd_update_can_read(client);
+    nbd_client_receive_next_request(client);
+
     nbd_client_put(client);
 }
 
@@ -XXX,XX +XXX,XX @@ static void blk_aio_attached(AioContext *ctx, void *opaque)
     exp->ctx = ctx;
 
     QTAILQ_FOREACH(client, &exp->clients, next) {
-        nbd_set_handlers(client);
+        qio_channel_attach_aio_context(client->ioc, ctx);
+        if (client->recv_coroutine) {
+            aio_co_schedule(ctx, client->recv_coroutine);
+        }
+        if (client->send_coroutine) {
+            aio_co_schedule(ctx, client->send_coroutine);
+        }
     }
 }
 
@@ -XXX,XX +XXX,XX @@ static void blk_aio_detach(void *opaque)
     TRACE("Export %s: Detaching clients from AIO context %p\n", exp->name, exp->ctx);
 
     QTAILQ_FOREACH(client, &exp->clients, next) {
-        nbd_unset_handlers(client);
+        qio_channel_detach_aio_context(client->ioc);
     }
 
     exp->ctx = NULL;
@@ -XXX,XX +XXX,XX @@ static ssize_t nbd_co_send_reply(NBDRequestData *req, NBDReply *reply,
     g_assert(qemu_in_coroutine());
     qemu_co_mutex_lock(&client->send_lock);
     client->send_coroutine = qemu_coroutine_self();
-    nbd_set_handlers(client);
 
     if (!len) {
         rc = nbd_send_reply(client->ioc, reply);
@@ -XXX,XX +XXX,XX @@ static ssize_t nbd_co_send_reply(NBDRequestData *req, NBDReply *reply,
     }
 
     client->send_coroutine = NULL;
-    nbd_set_handlers(client);
     qemu_co_mutex_unlock(&client->send_lock);
     return rc;
 }
@@ -XXX,XX +XXX,XX @@ static ssize_t nbd_co_receive_request(NBDRequestData *req,
     ssize_t rc;
 
     g_assert(qemu_in_coroutine());
-    client->recv_coroutine = qemu_coroutine_self();
-    nbd_update_can_read(client);
-
+    assert(client->recv_coroutine == qemu_coroutine_self());
     rc = nbd_receive_request(client->ioc, request);
     if (rc < 0) {
         if (rc != -EAGAIN) {
@@ -XXX,XX +XXX,XX @@ static ssize_t nbd_co_receive_request(NBDRequestData *req,
 
 out:
     client->recv_coroutine = NULL;
-    nbd_update_can_read(client);
+    nbd_client_receive_next_request(client);
 
     return rc;
 }
 
-static void nbd_trip(void *opaque)
+/* Owns a reference to the NBDClient passed as opaque.  */
+static coroutine_fn void nbd_trip(void *opaque)
 {
     NBDClient *client = opaque;
     NBDExport *exp = client->exp;
     NBDRequestData *req;
-    NBDRequest request;
+    NBDRequest request = { 0 };    /* GCC thinks it can be used uninitialized */
     NBDReply reply;
     ssize_t ret;
     int flags;
 
     TRACE("Reading request.");
     if (client->closing) {
+        nbd_client_put(client);
         return;
     }
 
@@ -XXX,XX +XXX,XX @@ static void nbd_trip(void *opaque)
 
 done:
     nbd_request_put(req);
+    nbd_client_put(client);
     return;
 
 out:
     nbd_request_put(req);
     client_close(client);
+    nbd_client_put(client);
 }
 
-static void nbd_read(void *opaque)
+static void nbd_client_receive_next_request(NBDClient *client)
 {
-    NBDClient *client = opaque;
-
-    if (client->recv_coroutine) {
-        qemu_coroutine_enter(client->recv_coroutine);
-    } else {
-        qemu_coroutine_enter(qemu_coroutine_create(nbd_trip, client));
-    }
-}
-
-static void nbd_restart_write(void *opaque)
-{
-    NBDClient *client = opaque;
-
-    qemu_coroutine_enter(client->send_coroutine);
-}
-
-static void nbd_set_handlers(NBDClient *client)
-{
-    if (client->exp && client->exp->ctx) {
-        aio_set_fd_handler(client->exp->ctx, client->sioc->fd, true,
-                           client->can_read ? nbd_read : NULL,
-                           client->send_coroutine ? nbd_restart_write : NULL,
-                           NULL, client);
-    }
-}
-
-static void nbd_unset_handlers(NBDClient *client)
-{
-    if (client->exp && client->exp->ctx) {
-        aio_set_fd_handler(client->exp->ctx, client->sioc->fd, true, NULL,
-                           NULL, NULL, NULL);
-    }
-}
-
-static void nbd_update_can_read(NBDClient *client)
-{
-    bool can_read = client->recv_coroutine ||
-                    client->nb_requests < MAX_NBD_REQUESTS;
-
-    if (can_read != client->can_read) {
-        client->can_read = can_read;
-        nbd_set_handlers(client);
-
-        /* There is no need to invoke aio_notify(), since aio_set_fd_handler()
-         * in nbd_set_handlers() will have taken care of that */
+    if (!client->recv_coroutine && client->nb_requests < MAX_NBD_REQUESTS) {
+        nbd_client_get(client);
+        client->recv_coroutine = qemu_coroutine_create(nbd_trip, client);
+        aio_co_schedule(client->exp->ctx, client->recv_coroutine);
     }
 }
 
@@ -XXX,XX +XXX,XX @@ static coroutine_fn void nbd_co_client_start(void *opaque)
         goto out;
     }
     qemu_co_mutex_init(&client->send_lock);
-    nbd_set_handlers(client);
 
     if (exp) {
         QTAILQ_INSERT_TAIL(&exp->clients, client, next);
     }
+
+    nbd_client_receive_next_request(client);
+
 out:
     g_free(data);
 }
@@ -XXX,XX +XXX,XX @@ void nbd_client_new(NBDExport *exp,
     object_ref(OBJECT(client->sioc));
     client->ioc = QIO_CHANNEL(sioc);
     object_ref(OBJECT(client->ioc));
-    client->can_read = true;
     client->close = close_fn;
 
     data->client = client;
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

As a small step towards the introduction of multiqueue, we want
coroutines to remain on the same AioContext that started them,
unless they are moved explicitly with e.g. aio_co_schedule.  This patch
avoids that coroutines switch AioContext when they use a CoMutex.
For now it does not make much of a difference, because the CoMutex
is not thread-safe and the AioContext itself is used to protect the
CoMutex from concurrent access.  However, this is going to change.

diff --git a/util/qemu-coroutine-lock.c b/util/qemu-coroutine-lock.c
index XXXXXXX..XXXXXXX 100644
--- a/util/qemu-coroutine-lock.c
+++ b/util/qemu-coroutine-lock.c
@@ -XXX,XX +XXX,XX @@
 #include "qemu/coroutine.h"
 #include "qemu/coroutine_int.h"
 #include "qemu/queue.h"
+#include "block/aio.h"
 #include "trace.h"
 
 void qemu_co_queue_init(CoQueue *queue)
@@ -XXX,XX +XXX,XX @@ void qemu_co_queue_run_restart(Coroutine *co)
 
 static bool qemu_co_queue_do_restart(CoQueue *queue, bool single)
 {
-    Coroutine *self = qemu_coroutine_self();
     Coroutine *next;
 
     if (QSIMPLEQ_EMPTY(&queue->entries)) {
@@ -XXX,XX +XXX,XX @@ static bool qemu_co_queue_do_restart(CoQueue *queue, bool single)
 
     while ((next = QSIMPLEQ_FIRST(&queue->entries)) != NULL) {
         QSIMPLEQ_REMOVE_HEAD(&queue->entries, co_queue_next);
-        QSIMPLEQ_INSERT_TAIL(&self->co_queue_wakeup, next, co_queue_next);
-        trace_qemu_co_queue_next(next);
+        aio_co_wake(next);
         if (single) {
             break;
         }
diff --git a/util/trace-events b/util/trace-events
index XXXXXXX..XXXXXXX 100644
--- a/util/trace-events
+++ b/util/trace-events
@@ -XXX,XX +XXX,XX @@ qemu_coroutine_terminate(void *co) "self %p"
 
 # util/qemu-coroutine-lock.c
 qemu_co_queue_run_restart(void *co) "co %p"
-qemu_co_queue_next(void *nxt) "next %p"
 qemu_co_mutex_lock_entry(void *mutex, void *self) "mutex %p self %p"
 qemu_co_mutex_lock_return(void *mutex, void *self) "mutex %p self %p"
 qemu_co_mutex_unlock_entry(void *mutex, void *self) "mutex %p self %p"
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

Keep the coroutine on the same AioContext.  Without this change,
there would be a race between yielding the coroutine and reentering it.
While the race cannot happen now, because the code only runs from a single
AioContext, this will change with multiqueue support in the block layer.

While doing the change, replace custom bottom half with aio_co_schedule.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 20170213135235.12274-10-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/blkdebug.c | 9 +--------
 1 file changed, 1 insertion(+), 8 deletions(-)

diff --git a/block/blkdebug.c b/block/blkdebug.c
index XXXXXXX..XXXXXXX 100644
--- a/block/blkdebug.c
+++ b/block/blkdebug.c
@@ -XXX,XX +XXX,XX @@ out:
     return ret;
 }
 
-static void error_callback_bh(void *opaque)
-{
-    Coroutine *co = opaque;
-    qemu_coroutine_enter(co);
-}
-
 static int inject_error(BlockDriverState *bs, BlkdebugRule *rule)
 {
     BDRVBlkdebugState *s = bs->opaque;
@@ -XXX,XX +XXX,XX @@ static int inject_error(BlockDriverState *bs, BlkdebugRule *rule)
     }
 
     if (!immediately) {
-        aio_bh_schedule_oneshot(bdrv_get_aio_context(bs), error_callback_bh,
-                                qemu_coroutine_self());
+        aio_co_schedule(qemu_get_current_aio_context(), qemu_coroutine_self());
         qemu_coroutine_yield();
     }
 
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

qed_aio_start_io and qed_aio_next_io will not have to acquire/release
the AioContext, while qed_aio_next_io_cb will.  Split the functionality
and gain a little type-safety in the process.

diff --git a/block/qed.c b/block/qed.c
index XXXXXXX..XXXXXXX 100644
--- a/block/qed.c
+++ b/block/qed.c
@@ -XXX,XX +XXX,XX @@ static CachedL2Table *qed_new_l2_table(BDRVQEDState *s)
     return l2_table;
 }
 
-static void qed_aio_next_io(void *opaque, int ret);
+static void qed_aio_next_io(QEDAIOCB *acb, int ret);
+
+static void qed_aio_start_io(QEDAIOCB *acb)
+{
+    qed_aio_next_io(acb, 0);
+}
+
+static void qed_aio_next_io_cb(void *opaque, int ret)
+{
+    QEDAIOCB *acb = opaque;
+
+    qed_aio_next_io(acb, ret);
+}
 
 static void qed_plug_allocating_write_reqs(BDRVQEDState *s)
 {
@@ -XXX,XX +XXX,XX @@ static void qed_unplug_allocating_write_reqs(BDRVQEDState *s)
 
     acb = QSIMPLEQ_FIRST(&s->allocating_write_reqs);
     if (acb) {
-        qed_aio_next_io(acb, 0);
+        qed_aio_start_io(acb);
     }
 }
 
@@ -XXX,XX +XXX,XX @@ static void qed_aio_complete(QEDAIOCB *acb, int ret)
         QSIMPLEQ_REMOVE_HEAD(&s->allocating_write_reqs, next);
         acb = QSIMPLEQ_FIRST(&s->allocating_write_reqs);
         if (acb) {
-            qed_aio_next_io(acb, 0);
+            qed_aio_start_io(acb);
         } else if (s->header.features & QED_F_NEED_CHECK) {
             qed_start_need_check_timer(s);
         }
@@ -XXX,XX +XXX,XX @@ static void qed_commit_l2_update(void *opaque, int ret)
     acb->request.l2_table = qed_find_l2_cache_entry(&s->l2_cache, l2_offset);
     assert(acb->request.l2_table != NULL);
 
-    qed_aio_next_io(opaque, ret);
+    qed_aio_next_io(acb, ret);
 }
 
 /**
@@ -XXX,XX +XXX,XX @@ static void qed_aio_write_l2_update(QEDAIOCB *acb, int ret, uint64_t offset)
     if (need_alloc) {
         /* Write out the whole new L2 table */
         qed_write_l2_table(s, &acb->request, 0, s->table_nelems, true,
-                            qed_aio_write_l1_update, acb);
+                           qed_aio_write_l1_update, acb);
     } else {
         /* Write out only the updated part of the L2 table */
         qed_write_l2_table(s, &acb->request, index, acb->cur_nclusters, false,
-                            qed_aio_next_io, acb);
+                           qed_aio_next_io_cb, acb);
     }
     return;
 
@@ -XXX,XX +XXX,XX @@ static void qed_aio_write_main(void *opaque, int ret)
     }
 
     if (acb->find_cluster_ret == QED_CLUSTER_FOUND) {
-        next_fn = qed_aio_next_io;
+        next_fn = qed_aio_next_io_cb;
     } else {
         if (s->bs->backing) {
             next_fn = qed_aio_write_flush_before_l2_update;
@@ -XXX,XX +XXX,XX @@ static void qed_aio_write_alloc(QEDAIOCB *acb, size_t len)
     if (acb->flags & QED_AIOCB_ZERO) {
         /* Skip ahead if the clusters are already zero */
         if (acb->find_cluster_ret == QED_CLUSTER_ZERO) {
-            qed_aio_next_io(acb, 0);
+            qed_aio_start_io(acb);
             return;
         }
 
@@ -XXX,XX +XXX,XX @@ static void qed_aio_read_data(void *opaque, int ret,
     /* Handle zero cluster and backing file reads */
     if (ret == QED_CLUSTER_ZERO) {
         qemu_iovec_memset(&acb->cur_qiov, 0, 0, acb->cur_qiov.size);
-        qed_aio_next_io(acb, 0);
+        qed_aio_start_io(acb);
         return;
     } else if (ret != QED_CLUSTER_FOUND) {
         qed_read_backing_file(s, acb->cur_pos, &acb->cur_qiov,
-                              &acb->backing_qiov, qed_aio_next_io, acb);
+                              &acb->backing_qiov, qed_aio_next_io_cb, acb);
         return;
     }
 
     BLKDBG_EVENT(bs->file, BLKDBG_READ_AIO);
     bdrv_aio_readv(bs->file, offset / BDRV_SECTOR_SIZE,
                    &acb->cur_qiov, acb->cur_qiov.size / BDRV_SECTOR_SIZE,
-                   qed_aio_next_io, acb);
+                   qed_aio_next_io_cb, acb);
     return;
 
 err:
@@ -XXX,XX +XXX,XX @@ err:
 /**
  * Begin next I/O or complete the request
  */
-static void qed_aio_next_io(void *opaque, int ret)
+static void qed_aio_next_io(QEDAIOCB *acb, int ret)
 {
-    QEDAIOCB *acb = opaque;
     BDRVQEDState *s = acb_to_s(acb);
     QEDFindClusterFunc *io_fn = (acb->flags & QED_AIOCB_WRITE) ?
                                 qed_aio_write_data : qed_aio_read_data;
@@ -XXX,XX +XXX,XX @@ static BlockAIOCB *qed_aio_setup(BlockDriverState *bs,
     qemu_iovec_init(&acb->cur_qiov, qiov->niov);
 
     /* Start request */
-    qed_aio_next_io(acb, 0);
+    qed_aio_start_io(acb);
     return &acb->common;
 }
 
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

The AioContext data structures are now protected by list_lock and/or
they are walked with FOREACH_RCU primitives.  There is no need anymore
to acquire the AioContext for the entire duration of aio_dispatch.
Instead, just acquire it before and after invoking the callbacks.
The next step is then to push it further down.

diff --git a/util/aio-posix.c b/util/aio-posix.c
index XXXXXXX..XXXXXXX 100644
--- a/util/aio-posix.c
+++ b/util/aio-posix.c
@@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx)
             (revents & (G_IO_IN | G_IO_HUP | G_IO_ERR)) &&
             aio_node_check(ctx, node->is_external) &&
             node->io_read) {
+            aio_context_acquire(ctx);
             node->io_read(node->opaque);
+            aio_context_release(ctx);
 
             /* aio_notify() does not count as progress */
             if (node->opaque != &ctx->notifier) {
@@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx)
             (revents & (G_IO_OUT | G_IO_ERR)) &&
             aio_node_check(ctx, node->is_external) &&
             node->io_write) {
+            aio_context_acquire(ctx);
             node->io_write(node->opaque);
+            aio_context_release(ctx);
             progress = true;
         }
 
@@ -XXX,XX +XXX,XX @@ bool aio_dispatch(AioContext *ctx, bool dispatch_fds)
     }
 
     /* Run our timers */
+    aio_context_acquire(ctx);
     progress |= timerlistgroup_run_timers(&ctx->tlg);
+    aio_context_release(ctx);
 
     return progress;
 }
@@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking)
     int64_t timeout;
     int64_t start = 0;
 
-    aio_context_acquire(ctx);
-    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
@@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking)
         start = qemu_clock_get_ns(QEMU_CLOCK_REALTIME);
     }
 
-    if (try_poll_mode(ctx, blocking)) {
-        progress = true;
-    } else {
+    aio_context_acquire(ctx);
+    progress = try_poll_mode(ctx, blocking);
+    aio_context_release(ctx);
+
+    if (!progress) {
         assert(npfd == 0);
 
         /* fill pollfds */
@@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking)
         timeout = blocking ? aio_compute_timeout(ctx) : 0;
 
         /* wait until next event */
-        if (timeout) {
-            aio_context_release(ctx);
-        }
         if (aio_epoll_check_poll(ctx, pollfds, npfd, timeout)) {
             AioHandler epoll_handler;
 
@@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking)
         } else  {
             ret = qemu_poll_ns(pollfds, npfd, timeout);
         }
-        if (timeout) {
-            aio_context_acquire(ctx);
-        }
     }
 
     if (blocking) {
@@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking)
         progress = true;
     }
 
-    aio_context_release(ctx);
-
     return progress;
 }
 
diff --git a/util/aio-win32.c b/util/aio-win32.c
index XXXXXXX..XXXXXXX 100644
--- a/util/aio-win32.c
+++ b/util/aio-win32.c
@@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx, HANDLE event)
             (revents || event_notifier_get_handle(node->e) == event) &&
             node->io_notify) {
             node->pfd.revents = 0;
+            aio_context_acquire(ctx);
             node->io_notify(node->e);
+            aio_context_release(ctx);
 
             /* aio_notify() does not count as progress */
             if (node->e != &ctx->notifier) {
@@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx, HANDLE event)
             (node->io_read || node->io_write)) {
             node->pfd.revents = 0;
             if ((revents & G_IO_IN) && node->io_read) {
+                aio_context_acquire(ctx);
                 node->io_read(node->opaque);
+                aio_context_release(ctx);
                 progress = true;
             }
             if ((revents & G_IO_OUT) && node->io_write) {
+                aio_context_acquire(ctx);
                 node->io_write(node->opaque);
+                aio_context_release(ctx);
                 progress = true;
             }
 
@@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking)
     int count;
     int timeout;
 
-    aio_context_acquire(ctx);
     progress = false;
 
     /* aio_notify can avoid the expensive event_notifier_set if
@@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking)
 
         timeout = blocking && !have_select_revents
             ? qemu_timeout_ns_to_ms(aio_compute_timeout(ctx)) : 0;
-        if (timeout) {
-            aio_context_release(ctx);
-        }
         ret = WaitForMultipleObjects(count, events, FALSE, timeout);
         if (blocking) {
             assert(first);
             atomic_sub(&ctx->notify_me, 2);
         }
-        if (timeout) {
-            aio_context_acquire(ctx);
-        }
 
         if (first) {
             aio_notify_accept(ctx);
@@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking)
         progress |= aio_dispatch_handlers(ctx, event);
     } while (count > 0);
 
+    aio_context_acquire(ctx);
     progress |= timerlistgroup_run_timers(&ctx->tlg);
-
     aio_context_release(ctx);
     return progress;
 }
diff --git a/util/async.c b/util/async.c
index XXXXXXX..XXXXXXX 100644
--- a/util/async.c
+++ b/util/async.c
@@ -XXX,XX +XXX,XX @@ int aio_bh_poll(AioContext *ctx)
                 ret = 1;
             }
             bh->idle = 0;
+            aio_context_acquire(ctx);
             aio_bh_call(bh);
+            aio_context_release(ctx);
         }
         if (bh->deleted) {
             deleted = true;
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 20170213135235.12274-13-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/qed.h                 |  3 +++
 block/curl.c                |  2 ++
 block/io.c                  |  5 +++++
 block/iscsi.c               |  8 ++++++--
 block/null.c                |  4 ++++
 block/qed.c                 | 12 ++++++++++++
 block/throttle-groups.c     |  2 ++
 util/aio-posix.c            |  2 --
 util/aio-win32.c            |  2 --
 util/qemu-coroutine-sleep.c |  2 +-
 10 files changed, 35 insertions(+), 7 deletions(-)

diff --git a/block/qed.h b/block/qed.h
index XXXXXXX..XXXXXXX 100644
--- a/block/qed.h
+++ b/block/qed.h
@@ -XXX,XX +XXX,XX @@ enum {
  */
 typedef void QEDFindClusterFunc(void *opaque, int ret, uint64_t offset, size_t len);
 
+void qed_acquire(BDRVQEDState *s);
+void qed_release(BDRVQEDState *s);
+
 /**
  * Generic callback for chaining async callbacks
  */
diff --git a/block/curl.c b/block/curl.c
index XXXXXXX..XXXXXXX 100644
--- a/block/curl.c
+++ b/block/curl.c
@@ -XXX,XX +XXX,XX @@ static void curl_multi_timeout_do(void *arg)
         return;
     }
 
+    aio_context_acquire(s->aio_context);
     curl_multi_socket_action(s->multi, CURL_SOCKET_TIMEOUT, 0, &running);
 
     curl_multi_check_completion(s);
+    aio_context_release(s->aio_context);
 #else
     abort();
 #endif
diff --git a/block/io.c b/block/io.c
index XXXXXXX..XXXXXXX 100644
--- a/block/io.c
+++ b/block/io.c
@@ -XXX,XX +XXX,XX @@ void bdrv_aio_cancel(BlockAIOCB *acb)
         if (acb->aiocb_info->get_aio_context) {
             aio_poll(acb->aiocb_info->get_aio_context(acb), true);
         } else if (acb->bs) {
+            /* qemu_aio_ref and qemu_aio_unref are not thread-safe, so
+             * assert that we're not using an I/O thread.  Thread-safe
+             * code should use bdrv_aio_cancel_async exclusively.
+             */
+            assert(bdrv_get_aio_context(acb->bs) == qemu_get_aio_context());
             aio_poll(bdrv_get_aio_context(acb->bs), true);
         } else {
             abort();
diff --git a/block/iscsi.c b/block/iscsi.c
index XXXXXXX..XXXXXXX 100644
--- a/block/iscsi.c
+++ b/block/iscsi.c
@@ -XXX,XX +XXX,XX @@ static void iscsi_retry_timer_expired(void *opaque)
     struct IscsiTask *iTask = opaque;
     iTask->complete = 1;
     if (iTask->co) {
-        qemu_coroutine_enter(iTask->co);
+        aio_co_wake(iTask->co);
     }
 }
 
@@ -XXX,XX +XXX,XX @@ static void iscsi_nop_timed_event(void *opaque)
 {
     IscsiLun *iscsilun = opaque;
 
+    aio_context_acquire(iscsilun->aio_context);
     if (iscsi_get_nops_in_flight(iscsilun->iscsi) >= MAX_NOP_FAILURES) {
         error_report("iSCSI: NOP timeout. Reconnecting...");
         iscsilun->request_timed_out = true;
     } else if (iscsi_nop_out_async(iscsilun->iscsi, NULL, NULL, 0, NULL) != 0) {
         error_report("iSCSI: failed to sent NOP-Out. Disabling NOP messages.");
-        return;
+        goto out;
     }
 
     timer_mod(iscsilun->nop_timer, qemu_clock_get_ms(QEMU_CLOCK_REALTIME) + NOP_INTERVAL);
     iscsi_set_events(iscsilun);
+
+out:
+    aio_context_release(iscsilun->aio_context);
 }
 
 static void iscsi_readcapacity_sync(IscsiLun *iscsilun, Error **errp)
diff --git a/block/null.c b/block/null.c
index XXXXXXX..XXXXXXX 100644
--- a/block/null.c
+++ b/block/null.c
@@ -XXX,XX +XXX,XX @@ static void null_bh_cb(void *opaque)
 static void null_timer_cb(void *opaque)
 {
     NullAIOCB *acb = opaque;
+    AioContext *ctx = bdrv_get_aio_context(acb->common.bs);
+
+    aio_context_acquire(ctx);
     acb->common.cb(acb->common.opaque, 0);
+    aio_context_release(ctx);
     timer_deinit(&acb->timer);
     qemu_aio_unref(acb);
 }
diff --git a/block/qed.c b/block/qed.c
index XXXXXXX..XXXXXXX 100644
--- a/block/qed.c
+++ b/block/qed.c
@@ -XXX,XX +XXX,XX @@ static void qed_need_check_timer_cb(void *opaque)
 
     trace_qed_need_check_timer_cb(s);
 
+    qed_acquire(s);
     qed_plug_allocating_write_reqs(s);
 
     /* Ensure writes are on disk before clearing flag */
     bdrv_aio_flush(s->bs->file->bs, qed_clear_need_check, s);
+    qed_release(s);
+}
+
+void qed_acquire(BDRVQEDState *s)
+{
+    aio_context_acquire(bdrv_get_aio_context(s->bs));
+}
+
+void qed_release(BDRVQEDState *s)
+{
+    aio_context_release(bdrv_get_aio_context(s->bs));
 }
 
 static void qed_start_need_check_timer(BDRVQEDState *s)
diff --git a/block/throttle-groups.c b/block/throttle-groups.c
index XXXXXXX..XXXXXXX 100644
--- a/block/throttle-groups.c
+++ b/block/throttle-groups.c
@@ -XXX,XX +XXX,XX @@ static void timer_cb(BlockBackend *blk, bool is_write)
     qemu_mutex_unlock(&tg->lock);
 
     /* Run the request that was waiting for this timer */
+    aio_context_acquire(blk_get_aio_context(blk));
     empty_queue = !qemu_co_enter_next(&blkp->throttled_reqs[is_write]);
+    aio_context_release(blk_get_aio_context(blk));
 
     /* If the request queue was empty then we have to take care of
      * scheduling the next one */
diff --git a/util/aio-posix.c b/util/aio-posix.c
index XXXXXXX..XXXXXXX 100644
--- a/util/aio-posix.c
+++ b/util/aio-posix.c
@@ -XXX,XX +XXX,XX @@ bool aio_dispatch(AioContext *ctx, bool dispatch_fds)
     }
 
     /* Run our timers */
-    aio_context_acquire(ctx);
     progress |= timerlistgroup_run_timers(&ctx->tlg);
-    aio_context_release(ctx);
 
     return progress;
 }
diff --git a/util/aio-win32.c b/util/aio-win32.c
index XXXXXXX..XXXXXXX 100644
--- a/util/aio-win32.c
+++ b/util/aio-win32.c
@@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking)
         progress |= aio_dispatch_handlers(ctx, event);
     } while (count > 0);
 
-    aio_context_acquire(ctx);
     progress |= timerlistgroup_run_timers(&ctx->tlg);
-    aio_context_release(ctx);
     return progress;
 }
 
diff --git a/util/qemu-coroutine-sleep.c b/util/qemu-coroutine-sleep.c
index XXXXXXX..XXXXXXX 100644
--- a/util/qemu-coroutine-sleep.c
+++ b/util/qemu-coroutine-sleep.c
@@ -XXX,XX +XXX,XX @@ static void co_sleep_cb(void *opaque)
 {
     CoSleepCB *sleep_cb = opaque;
 
-    qemu_coroutine_enter(sleep_cb->co);
+    aio_co_wake(sleep_cb->co);
 }
 
 void coroutine_fn co_aio_sleep_ns(AioContext *ctx, QEMUClockType type,
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

This covers both file descriptor callbacks and polling callbacks,
since they execute related code.

Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 20170213135235.12274-14-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/curl.c          | 16 +++++++++++++---
 block/iscsi.c         |  4 ++++
 block/linux-aio.c     |  4 ++++
 block/nfs.c           |  6 ++++++
 block/sheepdog.c      | 29 +++++++++++++++--------------
 block/ssh.c           | 29 +++++++++--------------------
 block/win32-aio.c     | 10 ++++++----
 hw/block/virtio-blk.c |  5 ++++-
 hw/scsi/virtio-scsi.c |  6 ++++++
 util/aio-posix.c      |  7 -------
 util/aio-win32.c      |  6 ------
 11 files changed, 67 insertions(+), 55 deletions(-)

diff --git a/block/curl.c b/block/curl.c
index XXXXXXX..XXXXXXX 100644
--- a/block/curl.c
+++ b/block/curl.c
@@ -XXX,XX +XXX,XX @@ static void curl_multi_check_completion(BDRVCURLState *s)
     }
 }
 
-static void curl_multi_do(void *arg)
+static void curl_multi_do_locked(CURLState *s)
 {
-    CURLState *s = (CURLState *)arg;
     CURLSocket *socket, *next_socket;
     int running;
     int r;
@@ -XXX,XX +XXX,XX @@ static void curl_multi_do(void *arg)
     }
 }
 
+static void curl_multi_do(void *arg)
+{
+    CURLState *s = (CURLState *)arg;
+
+    aio_context_acquire(s->s->aio_context);
+    curl_multi_do_locked(s);
+    aio_context_release(s->s->aio_context);
+}
+
 static void curl_multi_read(void *arg)
 {
     CURLState *s = (CURLState *)arg;
 
-    curl_multi_do(arg);
+    aio_context_acquire(s->s->aio_context);
+    curl_multi_do_locked(s);
     curl_multi_check_completion(s->s);
+    aio_context_release(s->s->aio_context);
 }
 
 static void curl_multi_timeout_do(void *arg)
diff --git a/block/iscsi.c b/block/iscsi.c
index XXXXXXX..XXXXXXX 100644
--- a/block/iscsi.c
+++ b/block/iscsi.c
@@ -XXX,XX +XXX,XX @@ iscsi_process_read(void *arg)
     IscsiLun *iscsilun = arg;
     struct iscsi_context *iscsi = iscsilun->iscsi;
 
+    aio_context_acquire(iscsilun->aio_context);
     iscsi_service(iscsi, POLLIN);
     iscsi_set_events(iscsilun);
+    aio_context_release(iscsilun->aio_context);
 }
 
 static void
@@ -XXX,XX +XXX,XX @@ iscsi_process_write(void *arg)
     IscsiLun *iscsilun = arg;
     struct iscsi_context *iscsi = iscsilun->iscsi;
 
+    aio_context_acquire(iscsilun->aio_context);
     iscsi_service(iscsi, POLLOUT);
     iscsi_set_events(iscsilun);
+    aio_context_release(iscsilun->aio_context);
 }
 
 static int64_t sector_lun2qemu(int64_t sector, IscsiLun *iscsilun)
diff --git a/block/linux-aio.c b/block/linux-aio.c
index XXXXXXX..XXXXXXX 100644
--- a/block/linux-aio.c
+++ b/block/linux-aio.c
@@ -XXX,XX +XXX,XX @@ static void qemu_laio_completion_cb(EventNotifier *e)
     LinuxAioState *s = container_of(e, LinuxAioState, e);
 
     if (event_notifier_test_and_clear(&s->e)) {
+        aio_context_acquire(s->aio_context);
         qemu_laio_process_completions_and_submit(s);
+        aio_context_release(s->aio_context);
     }
 }
 
@@ -XXX,XX +XXX,XX @@ static bool qemu_laio_poll_cb(void *opaque)
         return false;
     }
 
+    aio_context_acquire(s->aio_context);
     qemu_laio_process_completions_and_submit(s);
+    aio_context_release(s->aio_context);
     return true;
 }
 
diff --git a/block/nfs.c b/block/nfs.c
index XXXXXXX..XXXXXXX 100644
--- a/block/nfs.c
+++ b/block/nfs.c
@@ -XXX,XX +XXX,XX @@ static void nfs_set_events(NFSClient *client)
 static void nfs_process_read(void *arg)
 {
     NFSClient *client = arg;
+
+    aio_context_acquire(client->aio_context);
     nfs_service(client->context, POLLIN);
     nfs_set_events(client);
+    aio_context_release(client->aio_context);
 }
 
 static void nfs_process_write(void *arg)
 {
     NFSClient *client = arg;
+
+    aio_context_acquire(client->aio_context);
     nfs_service(client->context, POLLOUT);
     nfs_set_events(client);
+    aio_context_release(client->aio_context);
 }
 
 static void nfs_co_init_task(BlockDriverState *bs, NFSRPC *task)
diff --git a/block/sheepdog.c b/block/sheepdog.c
index XXXXXXX..XXXXXXX 100644
--- a/block/sheepdog.c
+++ b/block/sheepdog.c
@@ -XXX,XX +XXX,XX @@ static coroutine_fn int send_co_req(int sockfd, SheepdogReq *hdr, void *data,
     return ret;
 }
 
-static void restart_co_req(void *opaque)
-{
-    Coroutine *co = opaque;
-
-    qemu_coroutine_enter(co);
-}
-
 typedef struct SheepdogReqCo {
     int sockfd;
     BlockDriverState *bs;
@@ -XXX,XX +XXX,XX @@ typedef struct SheepdogReqCo {
     unsigned int *rlen;
     int ret;
     bool finished;
+    Coroutine *co;
 } SheepdogReqCo;
 
+static void restart_co_req(void *opaque)
+{
+    SheepdogReqCo *srco = opaque;
+
+    aio_co_wake(srco->co);
+}
+
 static coroutine_fn void do_co_req(void *opaque)
 {
     int ret;
-    Coroutine *co;
     SheepdogReqCo *srco = opaque;
     int sockfd = srco->sockfd;
     SheepdogReq *hdr = srco->hdr;
@@ -XXX,XX +XXX,XX @@ static coroutine_fn void do_co_req(void *opaque)
     unsigned int *wlen = srco->wlen;
     unsigned int *rlen = srco->rlen;
 
-    co = qemu_coroutine_self();
+    srco->co = qemu_coroutine_self();
     aio_set_fd_handler(srco->aio_context, sockfd, false,
-                       NULL, restart_co_req, NULL, co);
+                       NULL, restart_co_req, NULL, srco);
 
     ret = send_co_req(sockfd, hdr, data, wlen);
     if (ret < 0) {
@@ -XXX,XX +XXX,XX @@ static coroutine_fn void do_co_req(void *opaque)
     }
 
     aio_set_fd_handler(srco->aio_context, sockfd, false,
-                       restart_co_req, NULL, NULL, co);
+                       restart_co_req, NULL, NULL, srco);
 
     ret = qemu_co_recv(sockfd, hdr, sizeof(*hdr));
     if (ret != sizeof(*hdr)) {
@@ -XXX,XX +XXX,XX @@ out:
     aio_set_fd_handler(srco->aio_context, sockfd, false,
                        NULL, NULL, NULL, NULL);
 
+    srco->co = NULL;
     srco->ret = ret;
     srco->finished = true;
     if (srco->bs) {
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn aio_read_response(void *opaque)
          * We've finished all requests which belong to the AIOCB, so
          * we can switch back to sd_co_readv/writev now.
          */
-        qemu_coroutine_enter(acb->coroutine);
+        aio_co_wake(acb->coroutine);
     }
 
     return;
@@ -XXX,XX +XXX,XX @@ static void co_read_response(void *opaque)
         s->co_recv = qemu_coroutine_create(aio_read_response, opaque);
     }
 
-    qemu_coroutine_enter(s->co_recv);
+    aio_co_wake(s->co_recv);
 }
 
 static void co_write_request(void *opaque)
 {
     BDRVSheepdogState *s = opaque;
 
-    qemu_coroutine_enter(s->co_send);
+    aio_co_wake(s->co_send);
 }
 
 /*
diff --git a/block/ssh.c b/block/ssh.c
index XXXXXXX..XXXXXXX 100644
--- a/block/ssh.c
+++ b/block/ssh.c
@@ -XXX,XX +XXX,XX @@ static void restart_coroutine(void *opaque)
 
     DPRINTF("co=%p", co);
 
-    qemu_coroutine_enter(co);
+    aio_co_wake(co);
 }
 
-static coroutine_fn void set_fd_handler(BDRVSSHState *s, BlockDriverState *bs)
+/* A non-blocking call returned EAGAIN, so yield, ensuring the
+ * handlers are set up so that we'll be rescheduled when there is an
+ * interesting event on the socket.
+ */
+static coroutine_fn void co_yield(BDRVSSHState *s, BlockDriverState *bs)
 {
     int r;
     IOHandler *rd_handler = NULL, *wr_handler = NULL;
@@ -XXX,XX +XXX,XX @@ static coroutine_fn void set_fd_handler(BDRVSSHState *s, BlockDriverState *bs)
 
     aio_set_fd_handler(bdrv_get_aio_context(bs), s->sock,
                        false, rd_handler, wr_handler, NULL, co);
-}
-
-static coroutine_fn void clear_fd_handler(BDRVSSHState *s,
-                                          BlockDriverState *bs)
-{
-    DPRINTF("s->sock=%d", s->sock);
-    aio_set_fd_handler(bdrv_get_aio_context(bs), s->sock,
-                       false, NULL, NULL, NULL, NULL);
-}
-
-/* A non-blocking call returned EAGAIN, so yield, ensuring the
- * handlers are set up so that we'll be rescheduled when there is an
- * interesting event on the socket.
- */
-static coroutine_fn void co_yield(BDRVSSHState *s, BlockDriverState *bs)
-{
-    set_fd_handler(s, bs);
     qemu_coroutine_yield();
-    clear_fd_handler(s, bs);
+    DPRINTF("s->sock=%d - back", s->sock);
+    aio_set_fd_handler(bdrv_get_aio_context(bs), s->sock, false,
+                       NULL, NULL, NULL, NULL);
 }
 
 /* SFTP has a function `libssh2_sftp_seek64' which seeks to a position
diff --git a/block/win32-aio.c b/block/win32-aio.c
index XXXXXXX..XXXXXXX 100644
--- a/block/win32-aio.c
+++ b/block/win32-aio.c
@@ -XXX,XX +XXX,XX @@ struct QEMUWin32AIOState {
     HANDLE hIOCP;
     EventNotifier e;
     int count;
-    bool is_aio_context_attached;
+    AioContext *aio_ctx;
 };
 
 typedef struct QEMUWin32AIOCB {
@@ -XXX,XX +XXX,XX @@ static void win32_aio_process_completion(QEMUWin32AIOState *s,
     }
 
 
+    aio_context_acquire(s->aio_ctx);
     waiocb->common.cb(waiocb->common.opaque, ret);
+    aio_context_release(s->aio_ctx);
     qemu_aio_unref(waiocb);
 }
 
@@ -XXX,XX +XXX,XX @@ void win32_aio_detach_aio_context(QEMUWin32AIOState *aio,
                                   AioContext *old_context)
 {
     aio_set_event_notifier(old_context, &aio->e, false, NULL, NULL);
-    aio->is_aio_context_attached = false;
+    aio->aio_ctx = NULL;
 }
 
 void win32_aio_attach_aio_context(QEMUWin32AIOState *aio,
                                   AioContext *new_context)
 {
-    aio->is_aio_context_attached = true;
+    aio->aio_ctx = new_context;
     aio_set_event_notifier(new_context, &aio->e, false,
                            win32_aio_completion_cb, NULL);
 }
@@ -XXX,XX +XXX,XX @@ out_free_state:
 
 void win32_aio_cleanup(QEMUWin32AIOState *aio)
 {
-    assert(!aio->is_aio_context_attached);
+    assert(!aio->aio_ctx);
     CloseHandle(aio->hIOCP);
     event_notifier_cleanup(&aio->e);
     g_free(aio);
diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/block/virtio-blk.c
+++ b/hw/block/virtio-blk.c
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_ioctl_complete(void *opaque, int status)
 {
     VirtIOBlockIoctlReq *ioctl_req = opaque;
     VirtIOBlockReq *req = ioctl_req->req;
-    VirtIODevice *vdev = VIRTIO_DEVICE(req->dev);
+    VirtIOBlock *s = req->dev;
+    VirtIODevice *vdev = VIRTIO_DEVICE(s);
     struct virtio_scsi_inhdr *scsi;
     struct sg_io_hdr *hdr;
 
@@ -XXX,XX +XXX,XX @@ void virtio_blk_handle_vq(VirtIOBlock *s, VirtQueue *vq)
     VirtIOBlockReq *req;
     MultiReqBuffer mrb = {};
 
+    aio_context_acquire(blk_get_aio_context(s->blk));
     blk_io_plug(s->blk);
 
     do {
@@ -XXX,XX +XXX,XX @@ void virtio_blk_handle_vq(VirtIOBlock *s, VirtQueue *vq)
     }
 
     blk_io_unplug(s->blk);
+    aio_context_release(blk_get_aio_context(s->blk));
 }
 
 static void virtio_blk_handle_output(VirtIODevice *vdev, VirtQueue *vq)
diff --git a/hw/scsi/virtio-scsi.c b/hw/scsi/virtio-scsi.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/scsi/virtio-scsi.c
+++ b/hw/scsi/virtio-scsi.c
@@ -XXX,XX +XXX,XX @@ void virtio_scsi_handle_ctrl_vq(VirtIOSCSI *s, VirtQueue *vq)
 {
     VirtIOSCSIReq *req;
 
+    virtio_scsi_acquire(s);
     while ((req = virtio_scsi_pop_req(s, vq))) {
         virtio_scsi_handle_ctrl_req(s, req);
     }
+    virtio_scsi_release(s);
 }
 
 static void virtio_scsi_handle_ctrl(VirtIODevice *vdev, VirtQueue *vq)
@@ -XXX,XX +XXX,XX @@ void virtio_scsi_handle_cmd_vq(VirtIOSCSI *s, VirtQueue *vq)
 
     QTAILQ_HEAD(, VirtIOSCSIReq) reqs = QTAILQ_HEAD_INITIALIZER(reqs);
 
+    virtio_scsi_acquire(s);
     do {
         virtio_queue_set_notification(vq, 0);
 
@@ -XXX,XX +XXX,XX @@ void virtio_scsi_handle_cmd_vq(VirtIOSCSI *s, VirtQueue *vq)
     QTAILQ_FOREACH_SAFE(req, &reqs, next, next) {
         virtio_scsi_handle_cmd_req_submit(s, req);
     }
+    virtio_scsi_release(s);
 }
 
 static void virtio_scsi_handle_cmd(VirtIODevice *vdev, VirtQueue *vq)
@@ -XXX,XX +XXX,XX @@ out:
 
 void virtio_scsi_handle_event_vq(VirtIOSCSI *s, VirtQueue *vq)
 {
+    virtio_scsi_acquire(s);
     if (s->events_dropped) {
         virtio_scsi_push_event(s, NULL, VIRTIO_SCSI_T_NO_EVENT, 0);
     }
+    virtio_scsi_release(s);
 }
 
 static void virtio_scsi_handle_event(VirtIODevice *vdev, VirtQueue *vq)
diff --git a/util/aio-posix.c b/util/aio-posix.c
index XXXXXXX..XXXXXXX 100644
--- a/util/aio-posix.c
+++ b/util/aio-posix.c
@@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx)
             (revents & (G_IO_IN | G_IO_HUP | G_IO_ERR)) &&
             aio_node_check(ctx, node->is_external) &&
             node->io_read) {
-            aio_context_acquire(ctx);
             node->io_read(node->opaque);
-            aio_context_release(ctx);
 
             /* aio_notify() does not count as progress */
             if (node->opaque != &ctx->notifier) {
@@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx)
             (revents & (G_IO_OUT | G_IO_ERR)) &&
             aio_node_check(ctx, node->is_external) &&
             node->io_write) {
-            aio_context_acquire(ctx);
             node->io_write(node->opaque);
-            aio_context_release(ctx);
             progress = true;
         }
 
@@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking)
         start = qemu_clock_get_ns(QEMU_CLOCK_REALTIME);
     }
 
-    aio_context_acquire(ctx);
     progress = try_poll_mode(ctx, blocking);
-    aio_context_release(ctx);
-
     if (!progress) {
         assert(npfd == 0);
 
diff --git a/util/aio-win32.c b/util/aio-win32.c
index XXXXXXX..XXXXXXX 100644
--- a/util/aio-win32.c
+++ b/util/aio-win32.c
@@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx, HANDLE event)
             (revents || event_notifier_get_handle(node->e) == event) &&
             node->io_notify) {
             node->pfd.revents = 0;
-            aio_context_acquire(ctx);
             node->io_notify(node->e);
-            aio_context_release(ctx);
 
             /* aio_notify() does not count as progress */
             if (node->e != &ctx->notifier) {
@@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx, HANDLE event)
             (node->io_read || node->io_write)) {
             node->pfd.revents = 0;
             if ((revents & G_IO_IN) && node->io_read) {
-                aio_context_acquire(ctx);
                 node->io_read(node->opaque);
-                aio_context_release(ctx);
                 progress = true;
             }
             if ((revents & G_IO_OUT) && node->io_write) {
-                aio_context_acquire(ctx);
                 node->io_write(node->opaque);
-                aio_context_release(ctx);
                 progress = true;
             }
 
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 20170213135235.12274-15-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/archipelago.c   |  3 +++
 block/blkreplay.c     |  2 +-
 block/block-backend.c |  6 ++++++
 block/curl.c          | 26 ++++++++++++++++++--------
 block/gluster.c       |  9 +--------
 block/io.c            |  6 +++++-
 block/iscsi.c         |  6 +++++-
 block/linux-aio.c     | 15 +++++++++------
 block/nfs.c           |  3 ++-
 block/null.c          |  4 ++++
 block/qed.c           |  3 +++
 block/rbd.c           |  4 ++++
 dma-helpers.c         |  2 ++
 hw/block/virtio-blk.c |  2 ++
 hw/scsi/scsi-bus.c    |  2 ++
 util/async.c          |  4 ++--
 util/thread-pool.c    |  2 ++
 17 files changed, 71 insertions(+), 28 deletions(-)

diff --git a/block/archipelago.c b/block/archipelago.c
index XXXXXXX..XXXXXXX 100644
--- a/block/archipelago.c
+++ b/block/archipelago.c
@@ -XXX,XX +XXX,XX @@ static void qemu_archipelago_complete_aio(void *opaque)
 {
     AIORequestData *reqdata = (AIORequestData *) opaque;
     ArchipelagoAIOCB *aio_cb = (ArchipelagoAIOCB *) reqdata->aio_cb;
+    AioContext *ctx = bdrv_get_aio_context(aio_cb->common.bs);
 
+    aio_context_acquire(ctx);
     aio_cb->common.cb(aio_cb->common.opaque, aio_cb->ret);
+    aio_context_release(ctx);
     aio_cb->status = 0;
 
     qemu_aio_unref(aio_cb);
diff --git a/block/blkreplay.c b/block/blkreplay.c
index XXXXXXX..XXXXXXX 100755
--- a/block/blkreplay.c
+++ b/block/blkreplay.c
@@ -XXX,XX +XXX,XX @@ static int64_t blkreplay_getlength(BlockDriverState *bs)
 static void blkreplay_bh_cb(void *opaque)
 {
     Request *req = opaque;
-    qemu_coroutine_enter(req->co);
+    aio_co_wake(req->co);
     qemu_bh_delete(req->bh);
     g_free(req);
 }
diff --git a/block/block-backend.c b/block/block-backend.c
index XXXXXXX..XXXXXXX 100644
--- a/block/block-backend.c
+++ b/block/block-backend.c
@@ -XXX,XX +XXX,XX @@ int blk_make_zero(BlockBackend *blk, BdrvRequestFlags flags)
 static void error_callback_bh(void *opaque)
 {
     struct BlockBackendAIOCB *acb = opaque;
+    AioContext *ctx = bdrv_get_aio_context(acb->common.bs);
 
     bdrv_dec_in_flight(acb->common.bs);
+    aio_context_acquire(ctx);
     acb->common.cb(acb->common.opaque, acb->ret);
+    aio_context_release(ctx);
     qemu_aio_unref(acb);
 }
 
@@ -XXX,XX +XXX,XX @@ static void blk_aio_complete(BlkAioEmAIOCB *acb)
 static void blk_aio_complete_bh(void *opaque)
 {
     BlkAioEmAIOCB *acb = opaque;
+    AioContext *ctx = bdrv_get_aio_context(acb->common.bs);
 
     assert(acb->has_returned);
+    aio_context_acquire(ctx);
     blk_aio_complete(acb);
+    aio_context_release(ctx);
 }
 
 static BlockAIOCB *blk_aio_prwv(BlockBackend *blk, int64_t offset, int bytes,
diff --git a/block/curl.c b/block/curl.c
index XXXXXXX..XXXXXXX 100644
--- a/block/curl.c
+++ b/block/curl.c
@@ -XXX,XX +XXX,XX @@ static void curl_readv_bh_cb(void *p)
 {
     CURLState *state;
     int running;
+    int ret = -EINPROGRESS;
 
     CURLAIOCB *acb = p;
-    BDRVCURLState *s = acb->common.bs->opaque;
+    BlockDriverState *bs = acb->common.bs;
+    BDRVCURLState *s = bs->opaque;
+    AioContext *ctx = bdrv_get_aio_context(bs);
 
     size_t start = acb->sector_num * BDRV_SECTOR_SIZE;
     size_t end;
 
+    aio_context_acquire(ctx);
+
     // In case we have the requested data already (e.g. read-ahead),
     // we can just call the callback and be done.
     switch (curl_find_buf(s, start, acb->nb_sectors * BDRV_SECTOR_SIZE, acb)) {
@@ -XXX,XX +XXX,XX @@ static void curl_readv_bh_cb(void *p)
             qemu_aio_unref(acb);
             // fall through
         case FIND_RET_WAIT:
-            return;
+            goto out;
         default:
             break;
     }
@@ -XXX,XX +XXX,XX @@ static void curl_readv_bh_cb(void *p)
     // No cache found, so let's start a new request
     state = curl_init_state(acb->common.bs, s);
     if (!state) {
-        acb->common.cb(acb->common.opaque, -EIO);
-        qemu_aio_unref(acb);
-        return;
+        ret = -EIO;
+        goto out;
     }
 
     acb->start = 0;
@@ -XXX,XX +XXX,XX @@ static void curl_readv_bh_cb(void *p)
     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_unref(acb);
-        return;
+        ret = -ENOMEM;
+        goto out;
     }
     state->acb[0] = acb;
 
@@ -XXX,XX +XXX,XX @@ static void curl_readv_bh_cb(void *p)
 
     /* Tell curl it needs to kick things off */
     curl_multi_socket_action(s->multi, CURL_SOCKET_TIMEOUT, 0, &running);
+
+out:
+    if (ret != -EINPROGRESS) {
+        acb->common.cb(acb->common.opaque, ret);
+        qemu_aio_unref(acb);
+    }
+    aio_context_release(ctx);
 }
 
 static BlockAIOCB *curl_aio_readv(BlockDriverState *bs,
diff --git a/block/gluster.c b/block/gluster.c
index XXXXXXX..XXXXXXX 100644
--- a/block/gluster.c
+++ b/block/gluster.c
@@ -XXX,XX +XXX,XX @@ static struct glfs *qemu_gluster_init(BlockdevOptionsGluster *gconf,
     return qemu_gluster_glfs_init(gconf, errp);
 }
 
-static void qemu_gluster_complete_aio(void *opaque)
-{
-    GlusterAIOCB *acb = (GlusterAIOCB *)opaque;
-
-    qemu_coroutine_enter(acb->coroutine);
-}
-
 /*
  * AIO callback routine called from GlusterFS thread.
  */
@@ -XXX,XX +XXX,XX @@ static void gluster_finish_aiocb(struct glfs_fd *fd, ssize_t ret, void *arg)
         acb->ret = -EIO; /* Partial read/write - fail it */
     }
 
-    aio_bh_schedule_oneshot(acb->aio_context, qemu_gluster_complete_aio, acb);
+    aio_co_schedule(acb->aio_context, acb->coroutine);
 }
 
 static void qemu_gluster_parse_flags(int bdrv_flags, int *open_flags)
diff --git a/block/io.c b/block/io.c
index XXXXXXX..XXXXXXX 100644
--- a/block/io.c
+++ b/block/io.c
@@ -XXX,XX +XXX,XX @@ static void bdrv_co_drain_bh_cb(void *opaque)
     bdrv_dec_in_flight(bs);
     bdrv_drained_begin(bs);
     data->done = true;
-    qemu_coroutine_enter(co);
+    aio_co_wake(co);
 }
 
 static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs)
@@ -XXX,XX +XXX,XX @@ static void bdrv_co_complete(BlockAIOCBCoroutine *acb)
 static void bdrv_co_em_bh(void *opaque)
 {
     BlockAIOCBCoroutine *acb = opaque;
+    BlockDriverState *bs = acb->common.bs;
+    AioContext *ctx = bdrv_get_aio_context(bs);
 
     assert(!acb->need_bh);
+    aio_context_acquire(ctx);
     bdrv_co_complete(acb);
+    aio_context_release(ctx);
 }
 
 static void bdrv_co_maybe_schedule_bh(BlockAIOCBCoroutine *acb)
diff --git a/block/iscsi.c b/block/iscsi.c
index XXXXXXX..XXXXXXX 100644
--- a/block/iscsi.c
+++ b/block/iscsi.c
@@ -XXX,XX +XXX,XX @@ static void
 iscsi_bh_cb(void *p)
 {
     IscsiAIOCB *acb = p;
+    AioContext *ctx = bdrv_get_aio_context(acb->common.bs);
 
     qemu_bh_delete(acb->bh);
 
     g_free(acb->buf);
     acb->buf = NULL;
 
+    aio_context_acquire(ctx);
     acb->common.cb(acb->common.opaque, acb->status);
+    aio_context_release(ctx);
 
     if (acb->task != NULL) {
         scsi_free_scsi_task(acb->task);
@@ -XXX,XX +XXX,XX @@ iscsi_schedule_bh(IscsiAIOCB *acb)
 static void iscsi_co_generic_bh_cb(void *opaque)
 {
     struct IscsiTask *iTask = opaque;
+
     iTask->complete = 1;
-    qemu_coroutine_enter(iTask->co);
+    aio_co_wake(iTask->co);
 }
 
 static void iscsi_retry_timer_expired(void *opaque)
diff --git a/block/linux-aio.c b/block/linux-aio.c
index XXXXXXX..XXXXXXX 100644
--- a/block/linux-aio.c
+++ b/block/linux-aio.c
@@ -XXX,XX +XXX,XX @@ struct LinuxAioState {
     io_context_t ctx;
     EventNotifier e;
 
-    /* io queue for submit at batch */
+    /* io queue for submit at batch.  Protected by AioContext lock. */
     LaioQueue io_q;
 
-    /* I/O completion processing */
+    /* I/O completion processing.  Only runs in I/O thread.  */
     QEMUBH *completion_bh;
     int event_idx;
     int event_max;
@@ -XXX,XX +XXX,XX @@ static inline ssize_t io_event_ret(struct io_event *ev)
  */
 static void qemu_laio_process_completion(struct qemu_laiocb *laiocb)
 {
+    LinuxAioState *s = laiocb->ctx;
     int ret;
 
     ret = laiocb->ret;
@@ -XXX,XX +XXX,XX @@ static void qemu_laio_process_completion(struct qemu_laiocb *laiocb)
     }
 
     laiocb->ret = ret;
+    aio_context_acquire(s->aio_context);
     if (laiocb->co) {
         /* If the coroutine is already entered it must be in ioq_submit() and
          * will notice laio->ret has been filled in when it eventually runs
@@ -XXX,XX +XXX,XX @@ static void qemu_laio_process_completion(struct qemu_laiocb *laiocb)
         laiocb->common.cb(laiocb->common.opaque, ret);
         qemu_aio_unref(laiocb);
     }
+    aio_context_release(s->aio_context);
 }
 
 /**
@@ -XXX,XX +XXX,XX @@ static void qemu_laio_process_completions(LinuxAioState *s)
 static void qemu_laio_process_completions_and_submit(LinuxAioState *s)
 {
     qemu_laio_process_completions(s);
+
+    aio_context_acquire(s->aio_context);
     if (!s->io_q.plugged && !QSIMPLEQ_EMPTY(&s->io_q.pending)) {
         ioq_submit(s);
     }
+    aio_context_release(s->aio_context);
 }
 
 static void qemu_laio_completion_bh(void *opaque)
@@ -XXX,XX +XXX,XX @@ static void qemu_laio_completion_cb(EventNotifier *e)
     LinuxAioState *s = container_of(e, LinuxAioState, e);
 
     if (event_notifier_test_and_clear(&s->e)) {
-        aio_context_acquire(s->aio_context);
         qemu_laio_process_completions_and_submit(s);
-        aio_context_release(s->aio_context);
     }
 }
 
@@ -XXX,XX +XXX,XX @@ static bool qemu_laio_poll_cb(void *opaque)
         return false;
     }
 
-    aio_context_acquire(s->aio_context);
     qemu_laio_process_completions_and_submit(s);
-    aio_context_release(s->aio_context);
     return true;
 }
 
@@ -XXX,XX +XXX,XX @@ void laio_detach_aio_context(LinuxAioState *s, AioContext *old_context)
 {
     aio_set_event_notifier(old_context, &s->e, false, NULL, NULL);
     qemu_bh_delete(s->completion_bh);
+    s->aio_context = NULL;
 }
 
 void laio_attach_aio_context(LinuxAioState *s, AioContext *new_context)
diff --git a/block/nfs.c b/block/nfs.c
index XXXXXXX..XXXXXXX 100644
--- a/block/nfs.c
+++ b/block/nfs.c
@@ -XXX,XX +XXX,XX @@ static void nfs_co_init_task(BlockDriverState *bs, NFSRPC *task)
 static void nfs_co_generic_bh_cb(void *opaque)
 {
     NFSRPC *task = opaque;
+
     task->complete = 1;
-    qemu_coroutine_enter(task->co);
+    aio_co_wake(task->co);
 }
 
 static void
diff --git a/block/null.c b/block/null.c
index XXXXXXX..XXXXXXX 100644
--- a/block/null.c
+++ b/block/null.c
@@ -XXX,XX +XXX,XX @@ static const AIOCBInfo null_aiocb_info = {
 static void null_bh_cb(void *opaque)
 {
     NullAIOCB *acb = opaque;
+    AioContext *ctx = bdrv_get_aio_context(acb->common.bs);
+
+    aio_context_acquire(ctx);
     acb->common.cb(acb->common.opaque, 0);
+    aio_context_release(ctx);
     qemu_aio_unref(acb);
 }
 
diff --git a/block/qed.c b/block/qed.c
index XXXXXXX..XXXXXXX 100644
--- a/block/qed.c
+++ b/block/qed.c
@@ -XXX,XX +XXX,XX @@ static void qed_update_l2_table(BDRVQEDState *s, QEDTable *table, int index,
 static void qed_aio_complete_bh(void *opaque)
 {
     QEDAIOCB *acb = opaque;
+    BDRVQEDState *s = acb_to_s(acb);
     BlockCompletionFunc *cb = acb->common.cb;
     void *user_opaque = acb->common.opaque;
     int ret = acb->bh_ret;
@@ -XXX,XX +XXX,XX @@ static void qed_aio_complete_bh(void *opaque)
     qemu_aio_unref(acb);
 
     /* Invoke callback */
+    qed_acquire(s);
     cb(user_opaque, ret);
+    qed_release(s);
 }
 
 static void qed_aio_complete(QEDAIOCB *acb, int ret)
diff --git a/block/rbd.c b/block/rbd.c
index XXXXXXX..XXXXXXX 100644
--- a/block/rbd.c
+++ b/block/rbd.c
@@ -XXX,XX +XXX,XX @@ shutdown:
 static void qemu_rbd_complete_aio(RADOSCB *rcb)
 {
     RBDAIOCB *acb = rcb->acb;
+    AioContext *ctx = bdrv_get_aio_context(acb->common.bs);
     int64_t r;
 
     r = rcb->ret;
@@ -XXX,XX +XXX,XX @@ static void qemu_rbd_complete_aio(RADOSCB *rcb)
         qemu_iovec_from_buf(acb->qiov, 0, acb->bounce, acb->qiov->size);
     }
     qemu_vfree(acb->bounce);
+
+    aio_context_acquire(ctx);
     acb->common.cb(acb->common.opaque, (acb->ret > 0 ? 0 : acb->ret));
+    aio_context_release(ctx);
 
     qemu_aio_unref(acb);
 }
diff --git a/dma-helpers.c b/dma-helpers.c
index XXXXXXX..XXXXXXX 100644
--- a/dma-helpers.c
+++ b/dma-helpers.c
@@ -XXX,XX +XXX,XX @@ static void dma_blk_cb(void *opaque, int ret)
                                 QEMU_ALIGN_DOWN(dbs->iov.size, dbs->align));
     }
 
+    aio_context_acquire(dbs->ctx);
     dbs->acb = dbs->io_func(dbs->offset, &dbs->iov,
                             dma_blk_cb, dbs, dbs->io_func_opaque);
+    aio_context_release(dbs->ctx);
     assert(dbs->acb);
 }
 
diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/block/virtio-blk.c
+++ b/hw/block/virtio-blk.c
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_dma_restart_bh(void *opaque)
 
     s->rq = NULL;
 
+    aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
     while (req) {
         VirtIOBlockReq *next = req->next;
         if (virtio_blk_handle_request(req, &mrb)) {
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_dma_restart_bh(void *opaque)
     if (mrb.num_reqs) {
         virtio_blk_submit_multireq(s->blk, &mrb);
     }
+    aio_context_release(blk_get_aio_context(s->conf.conf.blk));
 }
 
 static void virtio_blk_dma_restart_cb(void *opaque, int running,
diff --git a/hw/scsi/scsi-bus.c b/hw/scsi/scsi-bus.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/scsi/scsi-bus.c
+++ b/hw/scsi/scsi-bus.c
@@ -XXX,XX +XXX,XX @@ static void scsi_dma_restart_bh(void *opaque)
     qemu_bh_delete(s->bh);
     s->bh = NULL;
 
+    aio_context_acquire(blk_get_aio_context(s->conf.blk));
     QTAILQ_FOREACH_SAFE(req, &s->requests, next, next) {
         scsi_req_ref(req);
         if (req->retry) {
@@ -XXX,XX +XXX,XX @@ static void scsi_dma_restart_bh(void *opaque)
         }
         scsi_req_unref(req);
     }
+    aio_context_release(blk_get_aio_context(s->conf.blk));
 }
 
 void scsi_req_retry(SCSIRequest *req)
diff --git a/util/async.c b/util/async.c
index XXXXXXX..XXXXXXX 100644
--- a/util/async.c
+++ b/util/async.c
@@ -XXX,XX +XXX,XX @@ int aio_bh_poll(AioContext *ctx)
                 ret = 1;
             }
             bh->idle = 0;
-            aio_context_acquire(ctx);
             aio_bh_call(bh);
-            aio_context_release(ctx);
         }
         if (bh->deleted) {
             deleted = true;
@@ -XXX,XX +XXX,XX @@ static void co_schedule_bh_cb(void *opaque)
         Coroutine *co = QSLIST_FIRST(&straight);
         QSLIST_REMOVE_HEAD(&straight, co_scheduled_next);
         trace_aio_co_schedule_bh_cb(ctx, co);
+        aio_context_acquire(ctx);
         qemu_coroutine_enter(co);
+        aio_context_release(ctx);
     }
 }
 
diff --git a/util/thread-pool.c b/util/thread-pool.c
index XXXXXXX..XXXXXXX 100644
--- a/util/thread-pool.c
+++ b/util/thread-pool.c
@@ -XXX,XX +XXX,XX @@ static void thread_pool_completion_bh(void *opaque)
     ThreadPool *pool = opaque;
     ThreadPoolElement *elem, *next;
 
+    aio_context_acquire(pool->ctx);
 restart:
     QLIST_FOREACH_SAFE(elem, &pool->head, all, next) {
         if (elem->state != THREAD_DONE) {
@@ -XXX,XX +XXX,XX @@ restart:
             qemu_aio_unref(elem);
         }
     }
+    aio_context_release(pool->ctx);
 }
 
 static void thread_pool_cancel(BlockAIOCB *acb)
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 20170213135235.12274-16-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 block/archipelago.c    |  3 ---
 block/block-backend.c  |  7 -------
 block/curl.c           |  2 +-
 block/io.c             |  6 +-----
 block/iscsi.c          |  3 ---
 block/linux-aio.c      |  5 +----
 block/mirror.c         | 12 +++++++++---
 block/null.c           |  8 --------
 block/qed-cluster.c    |  2 ++
 block/qed-table.c      | 12 ++++++++++--
 block/qed.c            |  4 ++--
 block/rbd.c            |  4 ----
 block/win32-aio.c      |  3 ---
 hw/block/virtio-blk.c  | 12 +++++++++++-
 hw/scsi/scsi-disk.c    | 15 +++++++++++++++
 hw/scsi/scsi-generic.c | 20 +++++++++++++++++---
 util/thread-pool.c     |  4 +++-
 17 files changed, 72 insertions(+), 50 deletions(-)

diff --git a/block/archipelago.c b/block/archipelago.c
index XXXXXXX..XXXXXXX 100644
--- a/block/archipelago.c
+++ b/block/archipelago.c
@@ -XXX,XX +XXX,XX @@ static void qemu_archipelago_complete_aio(void *opaque)
 {
     AIORequestData *reqdata = (AIORequestData *) opaque;
     ArchipelagoAIOCB *aio_cb = (ArchipelagoAIOCB *) reqdata->aio_cb;
-    AioContext *ctx = bdrv_get_aio_context(aio_cb->common.bs);
 
-    aio_context_acquire(ctx);
     aio_cb->common.cb(aio_cb->common.opaque, aio_cb->ret);
-    aio_context_release(ctx);
     aio_cb->status = 0;
 
     qemu_aio_unref(aio_cb);
diff --git a/block/block-backend.c b/block/block-backend.c
index XXXXXXX..XXXXXXX 100644
--- a/block/block-backend.c
+++ b/block/block-backend.c
@@ -XXX,XX +XXX,XX @@ int blk_make_zero(BlockBackend *blk, BdrvRequestFlags flags)
 static void error_callback_bh(void *opaque)
 {
     struct BlockBackendAIOCB *acb = opaque;
-    AioContext *ctx = bdrv_get_aio_context(acb->common.bs);
 
     bdrv_dec_in_flight(acb->common.bs);
-    aio_context_acquire(ctx);
     acb->common.cb(acb->common.opaque, acb->ret);
-    aio_context_release(ctx);
     qemu_aio_unref(acb);
 }
 
@@ -XXX,XX +XXX,XX @@ static void blk_aio_complete(BlkAioEmAIOCB *acb)
 static void blk_aio_complete_bh(void *opaque)
 {
     BlkAioEmAIOCB *acb = opaque;
-    AioContext *ctx = bdrv_get_aio_context(acb->common.bs);
-
     assert(acb->has_returned);
-    aio_context_acquire(ctx);
     blk_aio_complete(acb);
-    aio_context_release(ctx);
 }
 
 static BlockAIOCB *blk_aio_prwv(BlockBackend *blk, int64_t offset, int bytes,
diff --git a/block/curl.c b/block/curl.c
index XXXXXXX..XXXXXXX 100644
--- a/block/curl.c
+++ b/block/curl.c
@@ -XXX,XX +XXX,XX @@ static void curl_readv_bh_cb(void *p)
     curl_multi_socket_action(s->multi, CURL_SOCKET_TIMEOUT, 0, &running);
 
 out:
+    aio_context_release(ctx);
     if (ret != -EINPROGRESS) {
         acb->common.cb(acb->common.opaque, ret);
         qemu_aio_unref(acb);
     }
-    aio_context_release(ctx);
 }
 
 static BlockAIOCB *curl_aio_readv(BlockDriverState *bs,
diff --git a/block/io.c b/block/io.c
index XXXXXXX..XXXXXXX 100644
--- a/block/io.c
+++ b/block/io.c
@@ -XXX,XX +XXX,XX @@ static void bdrv_co_io_em_complete(void *opaque, int ret)
     CoroutineIOCompletion *co = opaque;
 
     co->ret = ret;
-    qemu_coroutine_enter(co->coroutine);
+    aio_co_wake(co->coroutine);
 }
 
 static int coroutine_fn bdrv_driver_preadv(BlockDriverState *bs,
@@ -XXX,XX +XXX,XX @@ static void bdrv_co_complete(BlockAIOCBCoroutine *acb)
 static void bdrv_co_em_bh(void *opaque)
 {
     BlockAIOCBCoroutine *acb = opaque;
-    BlockDriverState *bs = acb->common.bs;
-    AioContext *ctx = bdrv_get_aio_context(bs);
 
     assert(!acb->need_bh);
-    aio_context_acquire(ctx);
     bdrv_co_complete(acb);
-    aio_context_release(ctx);
 }
 
 static void bdrv_co_maybe_schedule_bh(BlockAIOCBCoroutine *acb)
diff --git a/block/iscsi.c b/block/iscsi.c
index XXXXXXX..XXXXXXX 100644
--- a/block/iscsi.c
+++ b/block/iscsi.c
@@ -XXX,XX +XXX,XX @@ static void
 iscsi_bh_cb(void *p)
 {
     IscsiAIOCB *acb = p;
-    AioContext *ctx = bdrv_get_aio_context(acb->common.bs);
 
     qemu_bh_delete(acb->bh);
 
     g_free(acb->buf);
     acb->buf = NULL;
 
-    aio_context_acquire(ctx);
     acb->common.cb(acb->common.opaque, acb->status);
-    aio_context_release(ctx);
 
     if (acb->task != NULL) {
         scsi_free_scsi_task(acb->task);
diff --git a/block/linux-aio.c b/block/linux-aio.c
index XXXXXXX..XXXXXXX 100644
--- a/block/linux-aio.c
+++ b/block/linux-aio.c
@@ -XXX,XX +XXX,XX @@ static inline ssize_t io_event_ret(struct io_event *ev)
  */
 static void qemu_laio_process_completion(struct qemu_laiocb *laiocb)
 {
-    LinuxAioState *s = laiocb->ctx;
     int ret;
 
     ret = laiocb->ret;
@@ -XXX,XX +XXX,XX @@ static void qemu_laio_process_completion(struct qemu_laiocb *laiocb)
     }
 
     laiocb->ret = ret;
-    aio_context_acquire(s->aio_context);
     if (laiocb->co) {
         /* If the coroutine is already entered it must be in ioq_submit() and
          * will notice laio->ret has been filled in when it eventually runs
@@ -XXX,XX +XXX,XX @@ static void qemu_laio_process_completion(struct qemu_laiocb *laiocb)
          * that!
          */
         if (!qemu_coroutine_entered(laiocb->co)) {
-            qemu_coroutine_enter(laiocb->co);
+            aio_co_wake(laiocb->co);
         }
     } else {
         laiocb->common.cb(laiocb->common.opaque, ret);
         qemu_aio_unref(laiocb);
     }
-    aio_context_release(s->aio_context);
 }
 
 /**
diff --git a/block/mirror.c b/block/mirror.c
index XXXXXXX..XXXXXXX 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -XXX,XX +XXX,XX @@ static void mirror_write_complete(void *opaque, int ret)
 {
     MirrorOp *op = opaque;
     MirrorBlockJob *s = op->s;
+
+    aio_context_acquire(blk_get_aio_context(s->common.blk));
     if (ret < 0) {
         BlockErrorAction action;
 
@@ -XXX,XX +XXX,XX @@ static void mirror_write_complete(void *opaque, int ret)
         }
     }
     mirror_iteration_done(op, ret);
+    aio_context_release(blk_get_aio_context(s->common.blk));
 }
 
 static void mirror_read_complete(void *opaque, int ret)
 {
     MirrorOp *op = opaque;
     MirrorBlockJob *s = op->s;
+
+    aio_context_acquire(blk_get_aio_context(s->common.blk));
     if (ret < 0) {
         BlockErrorAction action;
 
@@ -XXX,XX +XXX,XX @@ static void mirror_read_complete(void *opaque, int ret)
         }
 
         mirror_iteration_done(op, ret);
-        return;
+    } else {
+        blk_aio_pwritev(s->target, op->sector_num * BDRV_SECTOR_SIZE, &op->qiov,
+                        0, mirror_write_complete, op);
     }
-    blk_aio_pwritev(s->target, op->sector_num * BDRV_SECTOR_SIZE, &op->qiov,
-                    0, mirror_write_complete, op);
+    aio_context_release(blk_get_aio_context(s->common.blk));
 }
 
 static inline void mirror_clip_sectors(MirrorBlockJob *s,
diff --git a/block/null.c b/block/null.c
index XXXXXXX..XXXXXXX 100644
--- a/block/null.c
+++ b/block/null.c
@@ -XXX,XX +XXX,XX @@ static const AIOCBInfo null_aiocb_info = {
 static void null_bh_cb(void *opaque)
 {
     NullAIOCB *acb = opaque;
-    AioContext *ctx = bdrv_get_aio_context(acb->common.bs);
-
-    aio_context_acquire(ctx);
     acb->common.cb(acb->common.opaque, 0);
-    aio_context_release(ctx);
     qemu_aio_unref(acb);
 }
 
 static void null_timer_cb(void *opaque)
 {
     NullAIOCB *acb = opaque;
-    AioContext *ctx = bdrv_get_aio_context(acb->common.bs);
-
-    aio_context_acquire(ctx);
     acb->common.cb(acb->common.opaque, 0);
-    aio_context_release(ctx);
     timer_deinit(&acb->timer);
     qemu_aio_unref(acb);
 }
diff --git a/block/qed-cluster.c b/block/qed-cluster.c
index XXXXXXX..XXXXXXX 100644
--- a/block/qed-cluster.c
+++ b/block/qed-cluster.c
@@ -XXX,XX +XXX,XX @@ static void qed_find_cluster_cb(void *opaque, int ret)
     unsigned int index;
     unsigned int n;
 
+    qed_acquire(s);
     if (ret) {
         goto out;
     }
@@ -XXX,XX +XXX,XX @@ static void qed_find_cluster_cb(void *opaque, int ret)
 
 out:
     find_cluster_cb->cb(find_cluster_cb->opaque, ret, offset, len);
+    qed_release(s);
     g_free(find_cluster_cb);
 }
 
diff --git a/block/qed-table.c b/block/qed-table.c
index XXXXXXX..XXXXXXX 100644
--- a/block/qed-table.c
+++ b/block/qed-table.c
@@ -XXX,XX +XXX,XX @@ static void qed_read_table_cb(void *opaque, int ret)
 {
     QEDReadTableCB *read_table_cb = opaque;
     QEDTable *table = read_table_cb->table;
+    BDRVQEDState *s = read_table_cb->s;
     int noffsets = read_table_cb->qiov.size / sizeof(uint64_t);
     int i;
 
@@ -XXX,XX +XXX,XX @@ static void qed_read_table_cb(void *opaque, int ret)
     }
 
     /* Byteswap offsets */
+    qed_acquire(s);
     for (i = 0; i < noffsets; i++) {
         table->offsets[i] = le64_to_cpu(table->offsets[i]);
     }
+    qed_release(s);
 
 out:
     /* Completion */
-    trace_qed_read_table_cb(read_table_cb->s, read_table_cb->table, ret);
+    trace_qed_read_table_cb(s, read_table_cb->table, ret);
     gencb_complete(&read_table_cb->gencb, ret);
 }
 
@@ -XXX,XX +XXX,XX @@ typedef struct {
 static void qed_write_table_cb(void *opaque, int ret)
 {
     QEDWriteTableCB *write_table_cb = opaque;
+    BDRVQEDState *s = write_table_cb->s;
 
-    trace_qed_write_table_cb(write_table_cb->s,
+    trace_qed_write_table_cb(s,
                              write_table_cb->orig_table,
                              write_table_cb->flush,
                              ret);
@@ -XXX,XX +XXX,XX @@ static void qed_write_table_cb(void *opaque, int ret)
     if (write_table_cb->flush) {
         /* We still need to flush first */
         write_table_cb->flush = false;
+        qed_acquire(s);
         bdrv_aio_flush(write_table_cb->s->bs, qed_write_table_cb,
                        write_table_cb);
+        qed_release(s);
         return;
     }
 
@@ -XXX,XX +XXX,XX @@ static void qed_read_l2_table_cb(void *opaque, int ret)
     CachedL2Table *l2_table = request->l2_table;
     uint64_t l2_offset = read_l2_table_cb->l2_offset;
 
+    qed_acquire(s);
     if (ret) {
         /* can't trust loaded L2 table anymore */
         qed_unref_l2_cache_entry(l2_table);
@@ -XXX,XX +XXX,XX @@ static void qed_read_l2_table_cb(void *opaque, int ret)
         request->l2_table = qed_find_l2_cache_entry(&s->l2_cache, l2_offset);
         assert(request->l2_table != NULL);
     }
+    qed_release(s);
 
     gencb_complete(&read_l2_table_cb->gencb, ret);
 }
diff --git a/block/qed.c b/block/qed.c
index XXXXXXX..XXXXXXX 100644
--- a/block/qed.c
+++ b/block/qed.c
@@ -XXX,XX +XXX,XX @@ static void qed_is_allocated_cb(void *opaque, int ret, uint64_t offset, size_t l
     }
 
     if (cb->co) {
-        qemu_coroutine_enter(cb->co);
+        aio_co_wake(cb->co);
     }
 }
 
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn qed_co_pwrite_zeroes_cb(void *opaque, int ret)
     cb->done = true;
     cb->ret = ret;
     if (cb->co) {
-        qemu_coroutine_enter(cb->co);
+        aio_co_wake(cb->co);
     }
 }
 
diff --git a/block/rbd.c b/block/rbd.c
index XXXXXXX..XXXXXXX 100644
--- a/block/rbd.c
+++ b/block/rbd.c
@@ -XXX,XX +XXX,XX @@ shutdown:
 static void qemu_rbd_complete_aio(RADOSCB *rcb)
 {
     RBDAIOCB *acb = rcb->acb;
-    AioContext *ctx = bdrv_get_aio_context(acb->common.bs);
     int64_t r;
 
     r = rcb->ret;
@@ -XXX,XX +XXX,XX @@ static void qemu_rbd_complete_aio(RADOSCB *rcb)
         qemu_iovec_from_buf(acb->qiov, 0, acb->bounce, acb->qiov->size);
     }
     qemu_vfree(acb->bounce);
-
-    aio_context_acquire(ctx);
     acb->common.cb(acb->common.opaque, (acb->ret > 0 ? 0 : acb->ret));
-    aio_context_release(ctx);
 
     qemu_aio_unref(acb);
 }
diff --git a/block/win32-aio.c b/block/win32-aio.c
index XXXXXXX..XXXXXXX 100644
--- a/block/win32-aio.c
+++ b/block/win32-aio.c
@@ -XXX,XX +XXX,XX @@ static void win32_aio_process_completion(QEMUWin32AIOState *s,
         qemu_vfree(waiocb->buf);
     }
 
-
-    aio_context_acquire(s->aio_ctx);
     waiocb->common.cb(waiocb->common.opaque, ret);
-    aio_context_release(s->aio_ctx);
     qemu_aio_unref(waiocb);
 }
 
diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/block/virtio-blk.c
+++ b/hw/block/virtio-blk.c
@@ -XXX,XX +XXX,XX @@ static int virtio_blk_handle_rw_error(VirtIOBlockReq *req, int error,
 static void virtio_blk_rw_complete(void *opaque, int ret)
 {
     VirtIOBlockReq *next = opaque;
+    VirtIOBlock *s = next->dev;
 
+    aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
     while (next) {
         VirtIOBlockReq *req = next;
         next = req->mr_next;
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_rw_complete(void *opaque, int ret)
         block_acct_done(blk_get_stats(req->dev->blk), &req->acct);
         virtio_blk_free_request(req);
     }
+    aio_context_release(blk_get_aio_context(s->conf.conf.blk));
 }
 
 static void virtio_blk_flush_complete(void *opaque, int ret)
 {
     VirtIOBlockReq *req = opaque;
+    VirtIOBlock *s = req->dev;
 
+    aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
     if (ret) {
         if (virtio_blk_handle_rw_error(req, -ret, 0)) {
-            return;
+            goto out;
         }
     }
 
     virtio_blk_req_complete(req, VIRTIO_BLK_S_OK);
     block_acct_done(blk_get_stats(req->dev->blk), &req->acct);
     virtio_blk_free_request(req);
+
+out:
+    aio_context_release(blk_get_aio_context(s->conf.conf.blk));
 }
 
 #ifdef __linux__
@@ -XXX,XX +XXX,XX @@ static void virtio_blk_ioctl_complete(void *opaque, int status)
     virtio_stl_p(vdev, &scsi->data_len, hdr->dxfer_len);
 
 out:
+    aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
     virtio_blk_req_complete(req, status);
     virtio_blk_free_request(req);
+    aio_context_release(blk_get_aio_context(s->conf.conf.blk));
     g_free(ioctl_req);
 }
 
diff --git a/hw/scsi/scsi-disk.c b/hw/scsi/scsi-disk.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/scsi/scsi-disk.c
+++ b/hw/scsi/scsi-disk.c
@@ -XXX,XX +XXX,XX @@ static void scsi_aio_complete(void *opaque, int ret)
 
     assert(r->req.aiocb != NULL);
     r->req.aiocb = NULL;
+    aio_context_acquire(blk_get_aio_context(s->qdev.conf.blk));
     if (scsi_disk_req_check_error(r, ret, true)) {
         goto done;
     }
@@ -XXX,XX +XXX,XX @@ static void scsi_aio_complete(void *opaque, int ret)
     scsi_req_complete(&r->req, GOOD);
 
 done:
+    aio_context_release(blk_get_aio_context(s->qdev.conf.blk));
     scsi_req_unref(&r->req);
 }
 
@@ -XXX,XX +XXX,XX @@ static void scsi_dma_complete(void *opaque, int ret)
     assert(r->req.aiocb != NULL);
     r->req.aiocb = NULL;
 
+    aio_context_acquire(blk_get_aio_context(s->qdev.conf.blk));
     if (ret < 0) {
         block_acct_failed(blk_get_stats(s->qdev.conf.blk), &r->acct);
     } else {
         block_acct_done(blk_get_stats(s->qdev.conf.blk), &r->acct);
     }
     scsi_dma_complete_noio(r, ret);
+    aio_context_release(blk_get_aio_context(s->qdev.conf.blk));
 }
 
 static void scsi_read_complete(void * opaque, int ret)
@@ -XXX,XX +XXX,XX @@ static void scsi_read_complete(void * opaque, int ret)
 
     assert(r->req.aiocb != NULL);
     r->req.aiocb = NULL;
+    aio_context_acquire(blk_get_aio_context(s->qdev.conf.blk));
     if (scsi_disk_req_check_error(r, ret, true)) {
         goto done;
     }
@@ -XXX,XX +XXX,XX @@ static void scsi_read_complete(void * opaque, int ret)
 
 done:
     scsi_req_unref(&r->req);
+    aio_context_release(blk_get_aio_context(s->qdev.conf.blk));
 }
 
 /* Actually issue a read to the block device.  */
@@ -XXX,XX +XXX,XX @@ static void scsi_do_read_cb(void *opaque, int ret)
     assert (r->req.aiocb != NULL);
     r->req.aiocb = NULL;
 
+    aio_context_acquire(blk_get_aio_context(s->qdev.conf.blk));
     if (ret < 0) {
         block_acct_failed(blk_get_stats(s->qdev.conf.blk), &r->acct);
     } else {
         block_acct_done(blk_get_stats(s->qdev.conf.blk), &r->acct);
     }
     scsi_do_read(opaque, ret);
+    aio_context_release(blk_get_aio_context(s->qdev.conf.blk));
 }
 
 /* Read more data from scsi device into buffer.  */
@@ -XXX,XX +XXX,XX @@ static void scsi_write_complete(void * opaque, int ret)
     assert (r->req.aiocb != NULL);
     r->req.aiocb = NULL;
 
+    aio_context_acquire(blk_get_aio_context(s->qdev.conf.blk));
     if (ret < 0) {
         block_acct_failed(blk_get_stats(s->qdev.conf.blk), &r->acct);
     } else {
         block_acct_done(blk_get_stats(s->qdev.conf.blk), &r->acct);
     }
     scsi_write_complete_noio(r, ret);
+    aio_context_release(blk_get_aio_context(s->qdev.conf.blk));
 }
 
 static void scsi_write_data(SCSIRequest *req)
@@ -XXX,XX +XXX,XX @@ static void scsi_unmap_complete(void *opaque, int ret)
 {
     UnmapCBData *data = opaque;
     SCSIDiskReq *r = data->r;
+    SCSIDiskState *s = DO_UPCAST(SCSIDiskState, qdev, r->req.dev);
 
     assert(r->req.aiocb != NULL);
     r->req.aiocb = NULL;
 
+    aio_context_acquire(blk_get_aio_context(s->qdev.conf.blk));
     scsi_unmap_complete_noio(data, ret);
+    aio_context_release(blk_get_aio_context(s->qdev.conf.blk));
 }
 
 static void scsi_disk_emulate_unmap(SCSIDiskReq *r, uint8_t *inbuf)
@@ -XXX,XX +XXX,XX @@ static void scsi_write_same_complete(void *opaque, int ret)
 
     assert(r->req.aiocb != NULL);
     r->req.aiocb = NULL;
+    aio_context_acquire(blk_get_aio_context(s->qdev.conf.blk));
     if (scsi_disk_req_check_error(r, ret, true)) {
         goto done;
     }
@@ -XXX,XX +XXX,XX @@ done:
     scsi_req_unref(&r->req);
     qemu_vfree(data->iov.iov_base);
     g_free(data);
+    aio_context_release(blk_get_aio_context(s->qdev.conf.blk));
 }
 
 static void scsi_disk_emulate_write_same(SCSIDiskReq *r, uint8_t *inbuf)
diff --git a/hw/scsi/scsi-generic.c b/hw/scsi/scsi-generic.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/scsi/scsi-generic.c
+++ b/hw/scsi/scsi-generic.c
@@ -XXX,XX +XXX,XX @@ done:
 static void scsi_command_complete(void *opaque, int ret)
 {
     SCSIGenericReq *r = (SCSIGenericReq *)opaque;
+    SCSIDevice *s = r->req.dev;
 
     assert(r->req.aiocb != NULL);
     r->req.aiocb = NULL;
+
+    aio_context_acquire(blk_get_aio_context(s->conf.blk));
     scsi_command_complete_noio(r, ret);
+    aio_context_release(blk_get_aio_context(s->conf.blk));
 }
 
 static int execute_command(BlockBackend *blk,
@@ -XXX,XX +XXX,XX @@ static void scsi_read_complete(void * opaque, int ret)
     assert(r->req.aiocb != NULL);
     r->req.aiocb = NULL;
 
+    aio_context_acquire(blk_get_aio_context(s->conf.blk));
+
     if (ret || r->req.io_canceled) {
         scsi_command_complete_noio(r, ret);
-        return;
+        goto done;
     }
 
     len = r->io_header.dxfer_len - r->io_header.resid;
@@ -XXX,XX +XXX,XX @@ static void scsi_read_complete(void * opaque, int ret)
     r->len = -1;
     if (len == 0) {
         scsi_command_complete_noio(r, 0);
-        return;
+        goto done;
     }
 
     /* Snoop READ CAPACITY output to set the blocksize.  */
@@ -XXX,XX +XXX,XX @@ static void scsi_read_complete(void * opaque, int ret)
     }
     scsi_req_data(&r->req, len);
     scsi_req_unref(&r->req);
+
+done:
+    aio_context_release(blk_get_aio_context(s->conf.blk));
 }
 
 /* Read more data from scsi device into buffer.  */
@@ -XXX,XX +XXX,XX @@ static void scsi_write_complete(void * opaque, int ret)
     assert(r->req.aiocb != NULL);
     r->req.aiocb = NULL;
 
+    aio_context_acquire(blk_get_aio_context(s->conf.blk));
+
     if (ret || r->req.io_canceled) {
         scsi_command_complete_noio(r, ret);
-        return;
+        goto done;
     }
 
     if (r->req.cmd.buf[0] == MODE_SELECT && r->req.cmd.buf[4] == 12 &&
@@ -XXX,XX +XXX,XX @@ static void scsi_write_complete(void * opaque, int ret)
     }
 
     scsi_command_complete_noio(r, ret);
+
+done:
+    aio_context_release(blk_get_aio_context(s->conf.blk));
 }
 
 /* Write data to a scsi device.  Returns nonzero on failure.
diff --git a/util/thread-pool.c b/util/thread-pool.c
index XXXXXXX..XXXXXXX 100644
--- a/util/thread-pool.c
+++ b/util/thread-pool.c
@@ -XXX,XX +XXX,XX @@ restart:
              */
             qemu_bh_schedule(pool->completion_bh);
 
+            aio_context_release(pool->ctx);
             elem->common.cb(elem->common.opaque, elem->ret);
+            aio_context_acquire(pool->ctx);
             qemu_aio_unref(elem);
             goto restart;
         } else {
@@ -XXX,XX +XXX,XX @@ static void thread_pool_co_cb(void *opaque, int ret)
     ThreadPoolCo *co = opaque;
 
     co->ret = ret;
-    qemu_coroutine_enter(co->co);
+    aio_co_wake(co->co);
 }
 
 int coroutine_fn thread_pool_submit_co(ThreadPool *pool, ThreadPoolFunc *func,
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

This patch prepares for the removal of unnecessary lockcnt inc/dec pairs.
Extract the dispatching loop for file descriptor handlers into a new
function aio_dispatch_handlers, and then inline aio_dispatch into
aio_poll.

aio_dispatch can now become void.

diff --git a/include/block/aio.h b/include/block/aio.h
index XXXXXXX..XXXXXXX 100644
--- a/include/block/aio.h
+++ b/include/block/aio.h
@@ -XXX,XX +XXX,XX @@ bool aio_pending(AioContext *ctx);
 /* Dispatch any pending callbacks from the GSource attached to the AioContext.
  *
  * This is used internally in the implementation of the GSource.
- *
- * @dispatch_fds: true to process fds, false to skip them
- *                (can be used as an optimization by callers that know there
- *                are no fds ready)
  */
-bool aio_dispatch(AioContext *ctx, bool dispatch_fds);
+void aio_dispatch(AioContext *ctx);
 
 /* Progress in completing AIO work to occur.  This can issue new pending
  * aio as a result of executing I/O completion or bh callbacks.
diff --git a/util/aio-posix.c b/util/aio-posix.c
index XXXXXXX..XXXXXXX 100644
--- a/util/aio-posix.c
+++ b/util/aio-posix.c
@@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx)
     AioHandler *node, *tmp;
     bool progress = false;
 
-    /*
-     * We have to walk very carefully in case aio_set_fd_handler is
-     * called while we're walking.
-     */
-    qemu_lockcnt_inc(&ctx->list_lock);
-
     QLIST_FOREACH_SAFE_RCU(node, &ctx->aio_handlers, node, tmp) {
         int revents;
 
@@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx)
         }
     }
 
-    qemu_lockcnt_dec(&ctx->list_lock);
     return progress;
 }
 
-/*
- * Note that dispatch_fds == false has the side-effect of post-poning the
- * freeing of deleted handlers.
- */
-bool aio_dispatch(AioContext *ctx, bool dispatch_fds)
+void aio_dispatch(AioContext *ctx)
 {
-    bool progress;
+    aio_bh_poll(ctx);
 
-    /*
-     * 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).
-     */
-    progress = aio_bh_poll(ctx);
+    qemu_lockcnt_inc(&ctx->list_lock);
+    aio_dispatch_handlers(ctx);
+    qemu_lockcnt_dec(&ctx->list_lock);
 
-    if (dispatch_fds) {
-        progress |= aio_dispatch_handlers(ctx);
-    }
-
-    /* Run our timers */
-    progress |= timerlistgroup_run_timers(&ctx->tlg);
-
-    return progress;
+    timerlistgroup_run_timers(&ctx->tlg);
 }
 
 /* These thread-local variables are used only in a small part of aio_poll
@@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking)
     npfd = 0;
     qemu_lockcnt_dec(&ctx->list_lock);
 
-    /* Run dispatch even if there were no readable fds to run timers */
-    if (aio_dispatch(ctx, ret > 0)) {
-        progress = true;
+    progress |= aio_bh_poll(ctx);
+
+    if (ret > 0) {
+        qemu_lockcnt_inc(&ctx->list_lock);
+        progress |= aio_dispatch_handlers(ctx);
+        qemu_lockcnt_dec(&ctx->list_lock);
     }
 
+    progress |= timerlistgroup_run_timers(&ctx->tlg);
+
     return progress;
 }
 
diff --git a/util/aio-win32.c b/util/aio-win32.c
index XXXXXXX..XXXXXXX 100644
--- a/util/aio-win32.c
+++ b/util/aio-win32.c
@@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx, HANDLE event)
     return progress;
 }
 
-bool aio_dispatch(AioContext *ctx, bool dispatch_fds)
+void aio_dispatch(AioContext *ctx)
 {
-    bool progress;
-
-    progress = aio_bh_poll(ctx);
-    if (dispatch_fds) {
-        progress |= aio_dispatch_handlers(ctx, INVALID_HANDLE_VALUE);
-    }
-    progress |= timerlistgroup_run_timers(&ctx->tlg);
-    return progress;
+    aio_bh_poll(ctx);
+    aio_dispatch_handlers(ctx, INVALID_HANDLE_VALUE);
+    timerlistgroup_run_timers(&ctx->tlg);
 }
 
 bool aio_poll(AioContext *ctx, bool blocking)
diff --git a/util/async.c b/util/async.c
index XXXXXXX..XXXXXXX 100644
--- a/util/async.c
+++ b/util/async.c
@@ -XXX,XX +XXX,XX @@ aio_ctx_dispatch(GSource     *source,
     AioContext *ctx = (AioContext *) source;
 
     assert(callback == NULL);
-    aio_dispatch(ctx, true);
+    aio_dispatch(ctx);
     return true;
 }
 
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

Pull the increment/decrement pair out of aio_bh_poll and into the
callers.

diff --git a/util/aio-posix.c b/util/aio-posix.c
index XXXXXXX..XXXXXXX 100644
--- a/util/aio-posix.c
+++ b/util/aio-posix.c
@@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx)
 
 void aio_dispatch(AioContext *ctx)
 {
+    qemu_lockcnt_inc(&ctx->list_lock);
     aio_bh_poll(ctx);
-
-    qemu_lockcnt_inc(&ctx->list_lock);
     aio_dispatch_handlers(ctx);
     qemu_lockcnt_dec(&ctx->list_lock);
 
@@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking)
     }
 
     npfd = 0;
-    qemu_lockcnt_dec(&ctx->list_lock);
 
     progress |= aio_bh_poll(ctx);
 
     if (ret > 0) {
-        qemu_lockcnt_inc(&ctx->list_lock);
         progress |= aio_dispatch_handlers(ctx);
-        qemu_lockcnt_dec(&ctx->list_lock);
     }
 
+    qemu_lockcnt_dec(&ctx->list_lock);
+
     progress |= timerlistgroup_run_timers(&ctx->tlg);
 
     return progress;
diff --git a/util/aio-win32.c b/util/aio-win32.c
index XXXXXXX..XXXXXXX 100644
--- a/util/aio-win32.c
+++ b/util/aio-win32.c
@@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx, HANDLE event)
     bool progress = false;
     AioHandler *tmp;
 
-    qemu_lockcnt_inc(&ctx->list_lock);
-
     /*
      * We have to walk very carefully in case aio_set_fd_handler is
      * called while we're walking.
@@ -XXX,XX +XXX,XX @@ static bool aio_dispatch_handlers(AioContext *ctx, HANDLE event)
         }
     }
 
-    qemu_lockcnt_dec(&ctx->list_lock);
     return progress;
 }
 
 void aio_dispatch(AioContext *ctx)
 {
+    qemu_lockcnt_inc(&ctx->list_lock);
     aio_bh_poll(ctx);
     aio_dispatch_handlers(ctx, INVALID_HANDLE_VALUE);
+    qemu_lockcnt_dec(&ctx->list_lock);
     timerlistgroup_run_timers(&ctx->tlg);
 }
 
@@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking)
         }
     }
 
-    qemu_lockcnt_dec(&ctx->list_lock);
     first = true;
 
     /* ctx->notifier is always registered.  */
@@ -XXX,XX +XXX,XX @@ bool aio_poll(AioContext *ctx, bool blocking)
         progress |= aio_dispatch_handlers(ctx, event);
     } while (count > 0);
 
+    qemu_lockcnt_dec(&ctx->list_lock);
+
     progress |= timerlistgroup_run_timers(&ctx->tlg);
     return progress;
 }
diff --git a/util/async.c b/util/async.c
index XXXXXXX..XXXXXXX 100644
--- a/util/async.c
+++ b/util/async.c
@@ -XXX,XX +XXX,XX @@ void aio_bh_call(QEMUBH *bh)
     bh->cb(bh->opaque);
 }
 
-/* Multiple occurrences of aio_bh_poll cannot be called concurrently */
+/* Multiple occurrences of aio_bh_poll cannot be called concurrently.
+ * The count in ctx->list_lock is incremented before the call, and is
+ * not affected by the call.
+ */
 int aio_bh_poll(AioContext *ctx)
 {
     QEMUBH *bh, **bhp, *next;
     int ret;
     bool deleted = false;
 
-    qemu_lockcnt_inc(&ctx->list_lock);
-
     ret = 0;
     for (bh = atomic_rcu_read(&ctx->first_bh); bh; bh = next) {
         next = atomic_rcu_read(&bh->next);
@@ -XXX,XX +XXX,XX @@ int aio_bh_poll(AioContext *ctx)
 
     /* remove deleted bhs */
     if (!deleted) {
-        qemu_lockcnt_dec(&ctx->list_lock);
         return ret;
     }
 
-    if (qemu_lockcnt_dec_and_lock(&ctx->list_lock)) {
+    if (qemu_lockcnt_dec_if_lock(&ctx->list_lock)) {
         bhp = &ctx->first_bh;
         while (*bhp) {
             bh = *bhp;
@@ -XXX,XX +XXX,XX @@ int aio_bh_poll(AioContext *ctx)
                 bhp = &bh->next;
             }
         }
-        qemu_lockcnt_unlock(&ctx->list_lock);
+        qemu_lockcnt_inc_and_unlock(&ctx->list_lock);
     }
     return ret;
 }
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

diff --git a/include/block/block_int.h b/include/block/block_int.h
index XXXXXXX..XXXXXXX 100644
--- a/include/block/block_int.h
+++ b/include/block/block_int.h
@@ -XXX,XX +XXX,XX @@ struct BdrvChild {
  * copied as well.
  */
 struct BlockDriverState {
-    int64_t total_sectors; /* if we are reading a disk image, give its
-                              size in sectors */
+    /* Protected by big QEMU lock or read-only after opening.  No special
+     * locking needed during I/O...
+     */
     int open_flags; /* flags used to open the file, re-used for re-open */
     bool read_only; /* if true, the media is read only */
     bool encrypted; /* if true, the media is encrypted */
@@ -XXX,XX +XXX,XX @@ struct BlockDriverState {
     bool sg;        /* if true, the device is a /dev/sg* */
     bool probed;    /* if true, format was probed rather than specified */
 
-    int copy_on_read; /* if nonzero, copy read backing sectors into image.
-                         note this is a reference count */
-
-    CoQueue flush_queue;            /* Serializing flush queue */
-    bool active_flush_req;          /* Flush request in flight? */
-    unsigned int write_gen;         /* Current data generation */
-    unsigned int flushed_gen;       /* Flushed write generation */
-
     BlockDriver *drv; /* NULL means no media */
     void *opaque;
 
@@ -XXX,XX +XXX,XX @@ struct BlockDriverState {
     BdrvChild *backing;
     BdrvChild *file;
 
-    /* Callback before write request is processed */
-    NotifierWithReturnList before_write_notifiers;
-
-    /* number of in-flight requests; overall and serialising */
-    unsigned int in_flight;
-    unsigned int serialising_in_flight;
-
-    bool wakeup;
-
-    /* Offset after the highest byte written to */
-    uint64_t wr_highest_offset;
-
     /* I/O Limits */
     BlockLimits bl;
 
@@ -XXX,XX +XXX,XX @@ struct BlockDriverState {
     QTAILQ_ENTRY(BlockDriverState) bs_list;
     /* element of the list of monitor-owned BDS */
     QTAILQ_ENTRY(BlockDriverState) monitor_list;
-    QLIST_HEAD(, BdrvDirtyBitmap) dirty_bitmaps;
     int refcnt;
 
-    QLIST_HEAD(, BdrvTrackedRequest) tracked_requests;
-
     /* operation blockers */
     QLIST_HEAD(, BdrvOpBlocker) op_blockers[BLOCK_OP_TYPE_MAX];
 
@@ -XXX,XX +XXX,XX @@ struct BlockDriverState {
     /* The error object in use for blocking operations on backing_hd */
     Error *backing_blocker;
 
+    /* Protected by AioContext lock */
+
+    /* If true, copy read backing sectors into image.  Can be >1 if more
+     * than one client has requested copy-on-read.
+     */
+    int copy_on_read;
+
+    /* If we are reading a disk image, give its size in sectors.
+     * Generally read-only; it is written to by load_vmstate and save_vmstate,
+     * but the block layer is quiescent during those.
+     */
+    int64_t total_sectors;
+
+    /* Callback before write request is processed */
+    NotifierWithReturnList before_write_notifiers;
+
+    /* number of in-flight requests; overall and serialising */
+    unsigned int in_flight;
+    unsigned int serialising_in_flight;
+
+    bool wakeup;
+
+    /* Offset after the highest byte written to */
+    uint64_t wr_highest_offset;
+
     /* threshold limit for writes, in bytes. "High water mark". */
     uint64_t write_threshold_offset;
     NotifierWithReturn write_threshold_notifier;
@@ -XXX,XX +XXX,XX @@ struct BlockDriverState {
     /* counter for nested bdrv_io_plug */
     unsigned io_plugged;
 
+    QLIST_HEAD(, BdrvTrackedRequest) tracked_requests;
+    CoQueue flush_queue;                  /* Serializing flush queue */
+    bool active_flush_req;                /* Flush request in flight? */
+    unsigned int write_gen;               /* Current data generation */
+    unsigned int flushed_gen;             /* Flushed write generation */
+
+    QLIST_HEAD(, BdrvDirtyBitmap) dirty_bitmaps;
+
+    /* do we need to tell the quest if we have a volatile write cache? */
+    int enable_write_cache;
+
     int quiesce_counter;
 };
 
diff --git a/include/sysemu/block-backend.h b/include/sysemu/block-backend.h
index XXXXXXX..XXXXXXX 100644
--- a/include/sysemu/block-backend.h
+++ b/include/sysemu/block-backend.h
@@ -XXX,XX +XXX,XX @@ typedef struct BlockDevOps {
  * fields that must be public. This is in particular for QLIST_ENTRY() and
  * friends so that BlockBackends can be kept in lists outside block-backend.c */
 typedef struct BlockBackendPublic {
-    /* I/O throttling.
-     * throttle_state tells us if this BlockBackend has I/O limits configured.
-     * io_limits_disabled tells us if they are currently being enforced */
+    /* I/O throttling has its own locking, but also some fields are
+     * protected by the AioContext lock.
+     */
+
+    /* Protected by AioContext lock.  */
     CoQueue      throttled_reqs[2];
+
+    /* Nonzero if the I/O limits are currently being ignored; generally
+     * it is zero.  */
     unsigned int io_limits_disabled;
 
     /* The following fields are protected by the ThrottleGroup lock.
-     * See the ThrottleGroup documentation for details. */
+     * See the ThrottleGroup documentation for details.
+     * throttle_state tells us if I/O limits are configured. */
     ThrottleState *throttle_state;
     ThrottleTimers throttle_timers;
     unsigned       pending_reqs[2];
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

This uses the lock-free mutex described in the paper '"Blocking without
Locking", or LFTHREADS: A lock-free thread library' by Gidenstam and
Papatriantafilou.  The same technique is used in OSv, and in fact
the code is essentially a conversion to C of OSv's code.

[Added missing coroutine_fn in tests/test-aio-multithread.c.
--Stefan]

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 20170213181244.16297-2-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/qemu/coroutine.h     |  17 ++++-
 tests/test-aio-multithread.c |  86 ++++++++++++++++++++++++
 util/qemu-coroutine-lock.c   | 155 ++++++++++++++++++++++++++++++++++++++++---
 util/trace-events            |   1 +
 4 files changed, 246 insertions(+), 13 deletions(-)

diff --git a/include/qemu/coroutine.h b/include/qemu/coroutine.h
index XXXXXXX..XXXXXXX 100644
--- a/include/qemu/coroutine.h
+++ b/include/qemu/coroutine.h
@@ -XXX,XX +XXX,XX @@ bool qemu_co_queue_empty(CoQueue *queue);
 /**
  * Provides a mutex that can be used to synchronise coroutines
  */
+struct CoWaitRecord;
 typedef struct CoMutex {
-    bool locked;
+    /* Count of pending lockers; 0 for a free mutex, 1 for an
+     * uncontended mutex.
+     */
+    unsigned locked;
+
+    /* A queue of waiters.  Elements are added atomically in front of
+     * from_push.  to_pop is only populated, and popped from, by whoever
+     * is in charge of the next wakeup.  This can be an unlocker or,
+     * through the handoff protocol, a locker that is about to go to sleep.
+     */
+    QSLIST_HEAD(, CoWaitRecord) from_push, to_pop;
+
+    unsigned handoff, sequence;
+
     Coroutine *holder;
-    CoQueue queue;
 } CoMutex;
 
 /**
diff --git a/tests/test-aio-multithread.c b/tests/test-aio-multithread.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/test-aio-multithread.c
+++ b/tests/test-aio-multithread.c
@@ -XXX,XX +XXX,XX @@ static void test_multi_co_schedule_10(void)
     test_multi_co_schedule(10);
 }
 
+/* CoMutex thread-safety.  */
+
+static uint32_t atomic_counter;
+static uint32_t running;
+static uint32_t counter;
+static CoMutex comutex;
+
+static void coroutine_fn test_multi_co_mutex_entry(void *opaque)
+{
+    while (!atomic_mb_read(&now_stopping)) {
+        qemu_co_mutex_lock(&comutex);
+        counter++;
+        qemu_co_mutex_unlock(&comutex);
+
+        /* Increase atomic_counter *after* releasing the mutex.  Otherwise
+         * there is a chance (it happens about 1 in 3 runs) that the iothread
+         * exits before the coroutine is woken up, causing a spurious
+         * assertion failure.
+         */
+        atomic_inc(&atomic_counter);
+    }
+    atomic_dec(&running);
+}
+
+static void test_multi_co_mutex(int threads, int seconds)
+{
+    int i;
+
+    qemu_co_mutex_init(&comutex);
+    counter = 0;
+    atomic_counter = 0;
+    now_stopping = false;
+
+    create_aio_contexts();
+    assert(threads <= NUM_CONTEXTS);
+    running = threads;
+    for (i = 0; i < threads; i++) {
+        Coroutine *co1 = qemu_coroutine_create(test_multi_co_mutex_entry, NULL);
+        aio_co_schedule(ctx[i], co1);
+    }
+
+    g_usleep(seconds * 1000000);
+
+    atomic_mb_set(&now_stopping, true);
+    while (running > 0) {
+        g_usleep(100000);
+    }
+
+    join_aio_contexts();
+    g_test_message("%d iterations/second\n", counter / seconds);
+    g_assert_cmpint(counter, ==, atomic_counter);
+}
+
+/* Testing with NUM_CONTEXTS threads focuses on the queue.  The mutex however
+ * is too contended (and the threads spend too much time in aio_poll)
+ * to actually stress the handoff protocol.
+ */
+static void test_multi_co_mutex_1(void)
+{
+    test_multi_co_mutex(NUM_CONTEXTS, 1);
+}
+
+static void test_multi_co_mutex_10(void)
+{
+    test_multi_co_mutex(NUM_CONTEXTS, 10);
+}
+
+/* Testing with fewer threads stresses the handoff protocol too.  Still, the
+ * case where the locker _can_ pick up a handoff is very rare, happening
+ * about 10 times in 1 million, so increase the runtime a bit compared to
+ * other "quick" testcases that only run for 1 second.
+ */
+static void test_multi_co_mutex_2_3(void)
+{
+    test_multi_co_mutex(2, 3);
+}
+
+static void test_multi_co_mutex_2_30(void)
+{
+    test_multi_co_mutex(2, 30);
+}
+
 /* End of tests.  */
 
 int main(int argc, char **argv)
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
     g_test_add_func("/aio/multi/lifecycle", test_lifecycle);
     if (g_test_quick()) {
         g_test_add_func("/aio/multi/schedule", test_multi_co_schedule_1);
+        g_test_add_func("/aio/multi/mutex/contended", test_multi_co_mutex_1);
+        g_test_add_func("/aio/multi/mutex/handoff", test_multi_co_mutex_2_3);
     } else {
         g_test_add_func("/aio/multi/schedule", test_multi_co_schedule_10);
+        g_test_add_func("/aio/multi/mutex/contended", test_multi_co_mutex_10);
+        g_test_add_func("/aio/multi/mutex/handoff", test_multi_co_mutex_2_30);
     }
     return g_test_run();
 }
diff --git a/util/qemu-coroutine-lock.c b/util/qemu-coroutine-lock.c
index XXXXXXX..XXXXXXX 100644
--- a/util/qemu-coroutine-lock.c
+++ b/util/qemu-coroutine-lock.c
@@ -XXX,XX +XXX,XX @@
  * 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.
+ *
+ * The lock-free mutex implementation is based on OSv
+ * (core/lfmutex.cc, include/lockfree/mutex.hh).
+ * Copyright (C) 2013 Cloudius Systems, Ltd.
  */
 
 #include "qemu/osdep.h"
@@ -XXX,XX +XXX,XX @@ bool qemu_co_queue_empty(CoQueue *queue)
     return QSIMPLEQ_FIRST(&queue->entries) == NULL;
 }
 
+/* The wait records are handled with a multiple-producer, single-consumer
+ * lock-free queue.  There cannot be two concurrent pop_waiter() calls
+ * because pop_waiter() can only be called while mutex->handoff is zero.
+ * This can happen in three cases:
+ * - in qemu_co_mutex_unlock, before the hand-off protocol has started.
+ *   In this case, qemu_co_mutex_lock will see mutex->handoff == 0 and
+ *   not take part in the handoff.
+ * - in qemu_co_mutex_lock, if it steals the hand-off responsibility from
+ *   qemu_co_mutex_unlock.  In this case, qemu_co_mutex_unlock will fail
+ *   the cmpxchg (it will see either 0 or the next sequence value) and
+ *   exit.  The next hand-off cannot begin until qemu_co_mutex_lock has
+ *   woken up someone.
+ * - in qemu_co_mutex_unlock, if it takes the hand-off token itself.
+ *   In this case another iteration starts with mutex->handoff == 0;
+ *   a concurrent qemu_co_mutex_lock will fail the cmpxchg, and
+ *   qemu_co_mutex_unlock will go back to case (1).
+ *
+ * The following functions manage this queue.
+ */
+typedef struct CoWaitRecord {
+    Coroutine *co;
+    QSLIST_ENTRY(CoWaitRecord) next;
+} CoWaitRecord;
+
+static void push_waiter(CoMutex *mutex, CoWaitRecord *w)
+{
+    w->co = qemu_coroutine_self();
+    QSLIST_INSERT_HEAD_ATOMIC(&mutex->from_push, w, next);
+}
+
+static void move_waiters(CoMutex *mutex)
+{
+    QSLIST_HEAD(, CoWaitRecord) reversed;
+    QSLIST_MOVE_ATOMIC(&reversed, &mutex->from_push);
+    while (!QSLIST_EMPTY(&reversed)) {
+        CoWaitRecord *w = QSLIST_FIRST(&reversed);
+        QSLIST_REMOVE_HEAD(&reversed, next);
+        QSLIST_INSERT_HEAD(&mutex->to_pop, w, next);
+    }
+}
+
+static CoWaitRecord *pop_waiter(CoMutex *mutex)
+{
+    CoWaitRecord *w;
+
+    if (QSLIST_EMPTY(&mutex->to_pop)) {
+        move_waiters(mutex);
+        if (QSLIST_EMPTY(&mutex->to_pop)) {
+            return NULL;
+        }
+    }
+    w = QSLIST_FIRST(&mutex->to_pop);
+    QSLIST_REMOVE_HEAD(&mutex->to_pop, next);
+    return w;
+}
+
+static bool has_waiters(CoMutex *mutex)
+{
+    return QSLIST_EMPTY(&mutex->to_pop) || QSLIST_EMPTY(&mutex->from_push);
+}
+
 void qemu_co_mutex_init(CoMutex *mutex)
 {
     memset(mutex, 0, sizeof(*mutex));
-    qemu_co_queue_init(&mutex->queue);
 }
 
-void coroutine_fn qemu_co_mutex_lock(CoMutex *mutex)
+static void coroutine_fn qemu_co_mutex_lock_slowpath(CoMutex *mutex)
 {
     Coroutine *self = qemu_coroutine_self();
+    CoWaitRecord w;
+    unsigned old_handoff;
 
     trace_qemu_co_mutex_lock_entry(mutex, self);
+    w.co = self;
+    push_waiter(mutex, &w);
 
-    while (mutex->locked) {
-        qemu_co_queue_wait(&mutex->queue);
+    /* This is the "Responsibility Hand-Off" protocol; a lock() picks from
+     * a concurrent unlock() the responsibility of waking somebody up.
+     */
+    old_handoff = atomic_mb_read(&mutex->handoff);
+    if (old_handoff &&
+        has_waiters(mutex) &&
+        atomic_cmpxchg(&mutex->handoff, old_handoff, 0) == old_handoff) {
+        /* There can be no concurrent pops, because there can be only
+         * one active handoff at a time.
+         */
+        CoWaitRecord *to_wake = pop_waiter(mutex);
+        Coroutine *co = to_wake->co;
+        if (co == self) {
+            /* We got the lock ourselves!  */
+            assert(to_wake == &w);
+            return;
+        }
+
+        aio_co_wake(co);
     }
 
-    mutex->locked = true;
-    mutex->holder = self;
-    self->locks_held++;
-
+    qemu_coroutine_yield();
     trace_qemu_co_mutex_lock_return(mutex, self);
 }
 
+void coroutine_fn qemu_co_mutex_lock(CoMutex *mutex)
+{
+    Coroutine *self = qemu_coroutine_self();
+
+    if (atomic_fetch_inc(&mutex->locked) == 0) {
+        /* Uncontended.  */
+        trace_qemu_co_mutex_lock_uncontended(mutex, self);
+    } else {
+        qemu_co_mutex_lock_slowpath(mutex);
+    }
+    mutex->holder = self;
+    self->locks_held++;
+}
+
 void coroutine_fn qemu_co_mutex_unlock(CoMutex *mutex)
 {
     Coroutine *self = qemu_coroutine_self();
 
     trace_qemu_co_mutex_unlock_entry(mutex, self);
 
-    assert(mutex->locked == true);
+    assert(mutex->locked);
     assert(mutex->holder == self);
     assert(qemu_in_coroutine());
 
-    mutex->locked = false;
     mutex->holder = NULL;
     self->locks_held--;
-    qemu_co_queue_next(&mutex->queue);
+    if (atomic_fetch_dec(&mutex->locked) == 1) {
+        /* No waiting qemu_co_mutex_lock().  Pfew, that was easy!  */
+        return;
+    }
+
+    for (;;) {
+        CoWaitRecord *to_wake = pop_waiter(mutex);
+        unsigned our_handoff;
+
+        if (to_wake) {
+            Coroutine *co = to_wake->co;
+            aio_co_wake(co);
+            break;
+        }
+
+        /* Some concurrent lock() is in progress (we know this because
+         * mutex->locked was >1) but it hasn't yet put itself on the wait
+         * queue.  Pick a sequence number for the handoff protocol (not 0).
+         */
+        if (++mutex->sequence == 0) {
+            mutex->sequence = 1;
+        }
+
+        our_handoff = mutex->sequence;
+        atomic_mb_set(&mutex->handoff, our_handoff);
+        if (!has_waiters(mutex)) {
+            /* The concurrent lock has not added itself yet, so it
+             * will be able to pick our handoff.
+             */
+            break;
+        }
+
+        /* Try to do the handoff protocol ourselves; if somebody else has
+         * already taken it, however, we're done and they're responsible.
+         */
+        if (atomic_cmpxchg(&mutex->handoff, our_handoff, 0) != our_handoff) {
+            break;
+        }
+    }
 
     trace_qemu_co_mutex_unlock_return(mutex, self);
 }
diff --git a/util/trace-events b/util/trace-events
index XXXXXXX..XXXXXXX 100644
--- a/util/trace-events
+++ b/util/trace-events
@@ -XXX,XX +XXX,XX @@ qemu_coroutine_terminate(void *co) "self %p"
 
 # util/qemu-coroutine-lock.c
 qemu_co_queue_run_restart(void *co) "co %p"
+qemu_co_mutex_lock_uncontended(void *mutex, void *self) "mutex %p self %p"
 qemu_co_mutex_lock_entry(void *mutex, void *self) "mutex %p self %p"
 qemu_co_mutex_lock_return(void *mutex, void *self) "mutex %p self %p"
 qemu_co_mutex_unlock_entry(void *mutex, void *self) "mutex %p self %p"
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

Running a very small critical section on pthread_mutex_t and CoMutex
shows that pthread_mutex_t is much faster because it doesn't actually
go to sleep.  What happens is that the critical section is shorter
than the latency of entering the kernel and thus FUTEX_WAIT always
fails.  With CoMutex there is no such latency but you still want to
avoid wait and wakeup.  So introduce it artificially.

This only works with one waiters; because CoMutex is fair, it will
always have more waits and wakeups than a pthread_mutex_t.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 20170213181244.16297-3-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/qemu/coroutine.h   |  5 +++++
 util/qemu-coroutine-lock.c | 51 ++++++++++++++++++++++++++++++++++++++++------
 util/qemu-coroutine.c      |  2 +-
 3 files changed, 51 insertions(+), 7 deletions(-)

diff --git a/include/qemu/coroutine.h b/include/qemu/coroutine.h
index XXXXXXX..XXXXXXX 100644
--- a/include/qemu/coroutine.h
+++ b/include/qemu/coroutine.h
@@ -XXX,XX +XXX,XX @@ typedef struct CoMutex {
      */
     unsigned locked;
 
+    /* Context that is holding the lock.  Useful to avoid spinning
+     * when two coroutines on the same AioContext try to get the lock. :)
+     */
+    AioContext *ctx;
+
     /* A queue of waiters.  Elements are added atomically in front of
      * from_push.  to_pop is only populated, and popped from, by whoever
      * is in charge of the next wakeup.  This can be an unlocker or,
diff --git a/util/qemu-coroutine-lock.c b/util/qemu-coroutine-lock.c
index XXXXXXX..XXXXXXX 100644
--- a/util/qemu-coroutine-lock.c
+++ b/util/qemu-coroutine-lock.c
@@ -XXX,XX +XXX,XX @@
 #include "qemu-common.h"
 #include "qemu/coroutine.h"
 #include "qemu/coroutine_int.h"
+#include "qemu/processor.h"
 #include "qemu/queue.h"
 #include "block/aio.h"
 #include "trace.h"
@@ -XXX,XX +XXX,XX @@ void qemu_co_mutex_init(CoMutex *mutex)
     memset(mutex, 0, sizeof(*mutex));
 }
 
-static void coroutine_fn qemu_co_mutex_lock_slowpath(CoMutex *mutex)
+static void coroutine_fn qemu_co_mutex_wake(CoMutex *mutex, Coroutine *co)
+{
+    /* Read co before co->ctx; pairs with smp_wmb() in
+     * qemu_coroutine_enter().
+     */
+    smp_read_barrier_depends();
+    mutex->ctx = co->ctx;
+    aio_co_wake(co);
+}
+
+static void coroutine_fn qemu_co_mutex_lock_slowpath(AioContext *ctx,
+                                                     CoMutex *mutex)
 {
     Coroutine *self = qemu_coroutine_self();
     CoWaitRecord w;
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn qemu_co_mutex_lock_slowpath(CoMutex *mutex)
         if (co == self) {
             /* We got the lock ourselves!  */
             assert(to_wake == &w);
+            mutex->ctx = ctx;
             return;
         }
 
-        aio_co_wake(co);
+        qemu_co_mutex_wake(mutex, co);
     }
 
     qemu_coroutine_yield();
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn qemu_co_mutex_lock_slowpath(CoMutex *mutex)
 
 void coroutine_fn qemu_co_mutex_lock(CoMutex *mutex)
 {
+    AioContext *ctx = qemu_get_current_aio_context();
     Coroutine *self = qemu_coroutine_self();
+    int waiters, i;
 
-    if (atomic_fetch_inc(&mutex->locked) == 0) {
+    /* Running a very small critical section on pthread_mutex_t and CoMutex
+     * shows that pthread_mutex_t is much faster because it doesn't actually
+     * go to sleep.  What happens is that the critical section is shorter
+     * than the latency of entering the kernel and thus FUTEX_WAIT always
+     * fails.  With CoMutex there is no such latency but you still want to
+     * avoid wait and wakeup.  So introduce it artificially.
+     */
+    i = 0;
+retry_fast_path:
+    waiters = atomic_cmpxchg(&mutex->locked, 0, 1);
+    if (waiters != 0) {
+        while (waiters == 1 && ++i < 1000) {
+            if (atomic_read(&mutex->ctx) == ctx) {
+                break;
+            }
+            if (atomic_read(&mutex->locked) == 0) {
+                goto retry_fast_path;
+            }
+            cpu_relax();
+        }
+        waiters = atomic_fetch_inc(&mutex->locked);
+    }
+
+    if (waiters == 0) {
         /* Uncontended.  */
         trace_qemu_co_mutex_lock_uncontended(mutex, self);
+        mutex->ctx = ctx;
     } else {
-        qemu_co_mutex_lock_slowpath(mutex);
+        qemu_co_mutex_lock_slowpath(ctx, mutex);
     }
     mutex->holder = self;
     self->locks_held++;
@@ -XXX,XX +XXX,XX @@ void coroutine_fn qemu_co_mutex_unlock(CoMutex *mutex)
     assert(mutex->holder == self);
     assert(qemu_in_coroutine());
 
+    mutex->ctx = NULL;
     mutex->holder = NULL;
     self->locks_held--;
     if (atomic_fetch_dec(&mutex->locked) == 1) {
@@ -XXX,XX +XXX,XX @@ void coroutine_fn qemu_co_mutex_unlock(CoMutex *mutex)
         unsigned our_handoff;
 
         if (to_wake) {
-            Coroutine *co = to_wake->co;
-            aio_co_wake(co);
+            qemu_co_mutex_wake(mutex, to_wake->co);
             break;
         }
 
diff --git a/util/qemu-coroutine.c b/util/qemu-coroutine.c
index XXXXXXX..XXXXXXX 100644
--- a/util/qemu-coroutine.c
+++ b/util/qemu-coroutine.c
@@ -XXX,XX +XXX,XX @@ void qemu_coroutine_enter(Coroutine *co)
     co->ctx = qemu_get_current_aio_context();
 
     /* Store co->ctx before anything that stores co.  Matches
-     * barrier in aio_co_wake.
+     * barrier in aio_co_wake and qemu_co_mutex_wake.
      */
     smp_wmb();
 
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

Add two implementations of the same benchmark as the previous patch,
but using pthreads.  One uses a normal QemuMutex, the other is Linux
only and implements a fair mutex based on MCS locks and futexes.
This shows that the slower performance of the 5-thread case is due to
the fairness of CoMutex, rather than to coroutines.  If fairness does
not matter, as is the case with two threads, CoMutex can actually be
faster than pthreads.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 20170213181244.16297-4-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 tests/test-aio-multithread.c | 164 +++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 164 insertions(+)

diff --git a/tests/test-aio-multithread.c b/tests/test-aio-multithread.c
index XXXXXXX..XXXXXXX 100644
--- a/tests/test-aio-multithread.c
+++ b/tests/test-aio-multithread.c
@@ -XXX,XX +XXX,XX @@ static void test_multi_co_mutex_2_30(void)
     test_multi_co_mutex(2, 30);
 }
 
+/* Same test with fair mutexes, for performance comparison.  */
+
+#ifdef CONFIG_LINUX
+#include "qemu/futex.h"
+
+/* The nodes for the mutex reside in this structure (on which we try to avoid
+ * false sharing).  The head of the mutex is in the "mutex_head" variable.
+ */
+static struct {
+    int next, locked;
+    int padding[14];
+} nodes[NUM_CONTEXTS] __attribute__((__aligned__(64)));
+
+static int mutex_head = -1;
+
+static void mcs_mutex_lock(void)
+{
+    int prev;
+
+    nodes[id].next = -1;
+    nodes[id].locked = 1;
+    prev = atomic_xchg(&mutex_head, id);
+    if (prev != -1) {
+        atomic_set(&nodes[prev].next, id);
+        qemu_futex_wait(&nodes[id].locked, 1);
+    }
+}
+
+static void mcs_mutex_unlock(void)
+{
+    int next;
+    if (nodes[id].next == -1) {
+        if (atomic_read(&mutex_head) == id &&
+            atomic_cmpxchg(&mutex_head, id, -1) == id) {
+            /* Last item in the list, exit.  */
+            return;
+        }
+        while (atomic_read(&nodes[id].next) == -1) {
+            /* mcs_mutex_lock did the xchg, but has not updated
+             * nodes[prev].next yet.
+             */
+        }
+    }
+
+    /* Wake up the next in line.  */
+    next = nodes[id].next;
+    nodes[next].locked = 0;
+    qemu_futex_wake(&nodes[next].locked, 1);
+}
+
+static void test_multi_fair_mutex_entry(void *opaque)
+{
+    while (!atomic_mb_read(&now_stopping)) {
+        mcs_mutex_lock();
+        counter++;
+        mcs_mutex_unlock();
+        atomic_inc(&atomic_counter);
+    }
+    atomic_dec(&running);
+}
+
+static void test_multi_fair_mutex(int threads, int seconds)
+{
+    int i;
+
+    assert(mutex_head == -1);
+    counter = 0;
+    atomic_counter = 0;
+    now_stopping = false;
+
+    create_aio_contexts();
+    assert(threads <= NUM_CONTEXTS);
+    running = threads;
+    for (i = 0; i < threads; i++) {
+        Coroutine *co1 = qemu_coroutine_create(test_multi_fair_mutex_entry, NULL);
+        aio_co_schedule(ctx[i], co1);
+    }
+
+    g_usleep(seconds * 1000000);
+
+    atomic_mb_set(&now_stopping, true);
+    while (running > 0) {
+        g_usleep(100000);
+    }
+
+    join_aio_contexts();
+    g_test_message("%d iterations/second\n", counter / seconds);
+    g_assert_cmpint(counter, ==, atomic_counter);
+}
+
+static void test_multi_fair_mutex_1(void)
+{
+    test_multi_fair_mutex(NUM_CONTEXTS, 1);
+}
+
+static void test_multi_fair_mutex_10(void)
+{
+    test_multi_fair_mutex(NUM_CONTEXTS, 10);
+}
+#endif
+
+/* Same test with pthread mutexes, for performance comparison and
+ * portability.  */
+
+static QemuMutex mutex;
+
+static void test_multi_mutex_entry(void *opaque)
+{
+    while (!atomic_mb_read(&now_stopping)) {
+        qemu_mutex_lock(&mutex);
+        counter++;
+        qemu_mutex_unlock(&mutex);
+        atomic_inc(&atomic_counter);
+    }
+    atomic_dec(&running);
+}
+
+static void test_multi_mutex(int threads, int seconds)
+{
+    int i;
+
+    qemu_mutex_init(&mutex);
+    counter = 0;
+    atomic_counter = 0;
+    now_stopping = false;
+
+    create_aio_contexts();
+    assert(threads <= NUM_CONTEXTS);
+    running = threads;
+    for (i = 0; i < threads; i++) {
+        Coroutine *co1 = qemu_coroutine_create(test_multi_mutex_entry, NULL);
+        aio_co_schedule(ctx[i], co1);
+    }
+
+    g_usleep(seconds * 1000000);
+
+    atomic_mb_set(&now_stopping, true);
+    while (running > 0) {
+        g_usleep(100000);
+    }
+
+    join_aio_contexts();
+    g_test_message("%d iterations/second\n", counter / seconds);
+    g_assert_cmpint(counter, ==, atomic_counter);
+}
+
+static void test_multi_mutex_1(void)
+{
+    test_multi_mutex(NUM_CONTEXTS, 1);
+}
+
+static void test_multi_mutex_10(void)
+{
+    test_multi_mutex(NUM_CONTEXTS, 10);
+}
+
 /* End of tests.  */
 
 int main(int argc, char **argv)
@@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv)
         g_test_add_func("/aio/multi/schedule", test_multi_co_schedule_1);
         g_test_add_func("/aio/multi/mutex/contended", test_multi_co_mutex_1);
         g_test_add_func("/aio/multi/mutex/handoff", test_multi_co_mutex_2_3);
+#ifdef CONFIG_LINUX
+        g_test_add_func("/aio/multi/mutex/mcs", test_multi_fair_mutex_1);
+#endif
+        g_test_add_func("/aio/multi/mutex/pthread", test_multi_mutex_1);
     } else {
         g_test_add_func("/aio/multi/schedule", test_multi_co_schedule_10);
         g_test_add_func("/aio/multi/mutex/contended", test_multi_co_mutex_10);
         g_test_add_func("/aio/multi/mutex/handoff", test_multi_co_mutex_2_30);
+#ifdef CONFIG_LINUX
+        g_test_add_func("/aio/multi/mutex/mcs", test_multi_fair_mutex_10);
+#endif
+        g_test_add_func("/aio/multi/mutex/pthread", test_multi_mutex_10);
     }
     return g_test_run();
 }
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

This will avoid forward references in the next patch.  It is also
more logical because CoQueue is not anymore the basic primitive.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 20170213181244.16297-5-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/qemu/coroutine.h | 89 ++++++++++++++++++++++++------------------------
 1 file changed, 44 insertions(+), 45 deletions(-)

diff --git a/include/qemu/coroutine.h b/include/qemu/coroutine.h
index XXXXXXX..XXXXXXX 100644
--- a/include/qemu/coroutine.h
+++ b/include/qemu/coroutine.h
@@ -XXX,XX +XXX,XX @@ bool qemu_in_coroutine(void);
  */
 bool qemu_coroutine_entered(Coroutine *co);
 
-
-/**
- * CoQueues are a mechanism to queue coroutines in order to continue executing
- * them later. They provide the fundamental primitives on which coroutine locks
- * are built.
- */
-typedef struct CoQueue {
-    QSIMPLEQ_HEAD(, Coroutine) entries;
-} CoQueue;
-
-/**
- * Initialise a CoQueue. This must be called before any other operation is used
- * on the CoQueue.
- */
-void qemu_co_queue_init(CoQueue *queue);
-
-/**
- * Adds the current coroutine to the CoQueue and transfers control to the
- * caller of the coroutine.
- */
-void coroutine_fn qemu_co_queue_wait(CoQueue *queue);
-
-/**
- * Restarts the next coroutine in the CoQueue and removes it from the queue.
- *
- * Returns true if a coroutine was restarted, false if the queue is empty.
- */
-bool coroutine_fn qemu_co_queue_next(CoQueue *queue);
-
-/**
- * Restarts all coroutines in the CoQueue and leaves the queue empty.
- */
-void coroutine_fn qemu_co_queue_restart_all(CoQueue *queue);
-
-/**
- * Enter the next coroutine in the queue
- */
-bool qemu_co_enter_next(CoQueue *queue);
-
-/**
- * Checks if the CoQueue is empty.
- */
-bool qemu_co_queue_empty(CoQueue *queue);
-
-
 /**
  * Provides a mutex that can be used to synchronise coroutines
  */
@@ -XXX,XX +XXX,XX @@ void coroutine_fn qemu_co_mutex_lock(CoMutex *mutex);
  */
 void coroutine_fn qemu_co_mutex_unlock(CoMutex *mutex);
 
+
+/**
+ * CoQueues are a mechanism to queue coroutines in order to continue executing
+ * them later.
+ */
+typedef struct CoQueue {
+    QSIMPLEQ_HEAD(, Coroutine) entries;
+} CoQueue;
+
+/**
+ * Initialise a CoQueue. This must be called before any other operation is used
+ * on the CoQueue.
+ */
+void qemu_co_queue_init(CoQueue *queue);
+
+/**
+ * Adds the current coroutine to the CoQueue and transfers control to the
+ * caller of the coroutine.
+ */
+void coroutine_fn qemu_co_queue_wait(CoQueue *queue);
+
+/**
+ * Restarts the next coroutine in the CoQueue and removes it from the queue.
+ *
+ * Returns true if a coroutine was restarted, false if the queue is empty.
+ */
+bool coroutine_fn qemu_co_queue_next(CoQueue *queue);
+
+/**
+ * Restarts all coroutines in the CoQueue and leaves the queue empty.
+ */
+void coroutine_fn qemu_co_queue_restart_all(CoQueue *queue);
+
+/**
+ * Enter the next coroutine in the queue
+ */
+bool qemu_co_enter_next(CoQueue *queue);
+
+/**
+ * Checks if the CoQueue is empty.
+ */
+bool qemu_co_queue_empty(CoQueue *queue);
+
+
 typedef struct CoRwlock {
     bool writer;
     int reader;
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

All that CoQueue needs in order to become thread-safe is help
from an external mutex.  Add this to the API.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 20170213181244.16297-6-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/qemu/coroutine.h   |  8 +++++---
 block/backup.c             |  2 +-
 block/io.c                 |  4 ++--
 block/nbd-client.c         |  2 +-
 block/qcow2-cluster.c      |  4 +---
 block/sheepdog.c           |  2 +-
 block/throttle-groups.c    |  2 +-
 hw/9pfs/9p.c               |  2 +-
 util/qemu-coroutine-lock.c | 24 +++++++++++++++++++++---
 9 files changed, 34 insertions(+), 16 deletions(-)

diff --git a/include/qemu/coroutine.h b/include/qemu/coroutine.h
index XXXXXXX..XXXXXXX 100644
--- a/include/qemu/coroutine.h
+++ b/include/qemu/coroutine.h
@@ -XXX,XX +XXX,XX @@ void coroutine_fn qemu_co_mutex_unlock(CoMutex *mutex);
 
 /**
  * CoQueues are a mechanism to queue coroutines in order to continue executing
- * them later.
+ * them later.  They are similar to condition variables, but they need help
+ * from an external mutex in order to maintain thread-safety.
  */
 typedef struct CoQueue {
     QSIMPLEQ_HEAD(, Coroutine) entries;
@@ -XXX,XX +XXX,XX @@ void qemu_co_queue_init(CoQueue *queue);
 
 /**
  * Adds the current coroutine to the CoQueue and transfers control to the
- * caller of the coroutine.
+ * caller of the coroutine.  The mutex is unlocked during the wait and
+ * locked again afterwards.
  */
-void coroutine_fn qemu_co_queue_wait(CoQueue *queue);
+void coroutine_fn qemu_co_queue_wait(CoQueue *queue, CoMutex *mutex);
 
 /**
  * Restarts the next coroutine in the CoQueue and removes it from the queue.
diff --git a/block/backup.c b/block/backup.c
index XXXXXXX..XXXXXXX 100644
--- a/block/backup.c
+++ b/block/backup.c
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn wait_for_overlapping_requests(BackupBlockJob *job,
         retry = false;
         QLIST_FOREACH(req, &job->inflight_reqs, list) {
             if (end > req->start && start < req->end) {
-                qemu_co_queue_wait(&req->wait_queue);
+                qemu_co_queue_wait(&req->wait_queue, NULL);
                 retry = true;
                 break;
             }
diff --git a/block/io.c b/block/io.c
index XXXXXXX..XXXXXXX 100644
--- a/block/io.c
+++ b/block/io.c
@@ -XXX,XX +XXX,XX @@ static bool coroutine_fn wait_serialising_requests(BdrvTrackedRequest *self)
                  * (instead of producing a deadlock in the former case). */
                 if (!req->waiting_for) {
                     self->waiting_for = req;
-                    qemu_co_queue_wait(&req->wait_queue);
+                    qemu_co_queue_wait(&req->wait_queue, NULL);
                     self->waiting_for = NULL;
                     retry = true;
                     waited = true;
@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_flush(BlockDriverState *bs)
 
     /* Wait until any previous flushes are completed */
     while (bs->active_flush_req) {
-        qemu_co_queue_wait(&bs->flush_queue);
+        qemu_co_queue_wait(&bs->flush_queue, NULL);
     }
 
     bs->active_flush_req = true;
diff --git a/block/nbd-client.c b/block/nbd-client.c
index XXXXXXX..XXXXXXX 100644
--- a/block/nbd-client.c
+++ b/block/nbd-client.c
@@ -XXX,XX +XXX,XX @@ static void nbd_coroutine_start(NBDClientSession *s,
     /* Poor man semaphore.  The free_sema is locked when no other request
      * can be accepted, and unlocked after receiving one reply.  */
     if (s->in_flight == MAX_NBD_REQUESTS) {
-        qemu_co_queue_wait(&s->free_sema);
+        qemu_co_queue_wait(&s->free_sema, NULL);
         assert(s->in_flight < MAX_NBD_REQUESTS);
     }
     s->in_flight++;
diff --git a/block/qcow2-cluster.c b/block/qcow2-cluster.c
index XXXXXXX..XXXXXXX 100644
--- a/block/qcow2-cluster.c
+++ b/block/qcow2-cluster.c
@@ -XXX,XX +XXX,XX @@ static int handle_dependencies(BlockDriverState *bs, uint64_t guest_offset,
             if (bytes == 0) {
                 /* Wait for the dependency to complete. We need to recheck
                  * the free/allocated clusters when we continue. */
-                qemu_co_mutex_unlock(&s->lock);
-                qemu_co_queue_wait(&old_alloc->dependent_requests);
-                qemu_co_mutex_lock(&s->lock);
+                qemu_co_queue_wait(&old_alloc->dependent_requests, &s->lock);
                 return -EAGAIN;
             }
         }
diff --git a/block/sheepdog.c b/block/sheepdog.c
index XXXXXXX..XXXXXXX 100644
--- a/block/sheepdog.c
+++ b/block/sheepdog.c
@@ -XXX,XX +XXX,XX @@ static void wait_for_overlapping_aiocb(BDRVSheepdogState *s, SheepdogAIOCB *acb)
 retry:
     QLIST_FOREACH(cb, &s->inflight_aiocb_head, aiocb_siblings) {
         if (AIOCBOverlapping(acb, cb)) {
-            qemu_co_queue_wait(&s->overlapping_queue);
+            qemu_co_queue_wait(&s->overlapping_queue, NULL);
             goto retry;
         }
     }
diff --git a/block/throttle-groups.c b/block/throttle-groups.c
index XXXXXXX..XXXXXXX 100644
--- a/block/throttle-groups.c
+++ b/block/throttle-groups.c
@@ -XXX,XX +XXX,XX @@ void coroutine_fn throttle_group_co_io_limits_intercept(BlockBackend *blk,
     if (must_wait || blkp->pending_reqs[is_write]) {
         blkp->pending_reqs[is_write]++;
         qemu_mutex_unlock(&tg->lock);
-        qemu_co_queue_wait(&blkp->throttled_reqs[is_write]);
+        qemu_co_queue_wait(&blkp->throttled_reqs[is_write], NULL);
         qemu_mutex_lock(&tg->lock);
         blkp->pending_reqs[is_write]--;
     }
diff --git a/hw/9pfs/9p.c b/hw/9pfs/9p.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/9pfs/9p.c
+++ b/hw/9pfs/9p.c
@@ -XXX,XX +XXX,XX @@ static void coroutine_fn v9fs_flush(void *opaque)
         /*
          * Wait for pdu to complete.
          */
-        qemu_co_queue_wait(&cancel_pdu->complete);
+        qemu_co_queue_wait(&cancel_pdu->complete, NULL);
         cancel_pdu->cancelled = 0;
         pdu_free(cancel_pdu);
     }
diff --git a/util/qemu-coroutine-lock.c b/util/qemu-coroutine-lock.c
index XXXXXXX..XXXXXXX 100644
--- a/util/qemu-coroutine-lock.c
+++ b/util/qemu-coroutine-lock.c
@@ -XXX,XX +XXX,XX @@ void qemu_co_queue_init(CoQueue *queue)
     QSIMPLEQ_INIT(&queue->entries);
 }
 
-void coroutine_fn qemu_co_queue_wait(CoQueue *queue)
+void coroutine_fn qemu_co_queue_wait(CoQueue *queue, CoMutex *mutex)
 {
     Coroutine *self = qemu_coroutine_self();
     QSIMPLEQ_INSERT_TAIL(&queue->entries, self, co_queue_next);
+
+    if (mutex) {
+        qemu_co_mutex_unlock(mutex);
+    }
+
+    /* There is no race condition here.  Other threads will call
+     * aio_co_schedule on our AioContext, which can reenter this
+     * coroutine but only after this yield and after the main loop
+     * has gone through the next iteration.
+     */
     qemu_coroutine_yield();
     assert(qemu_in_coroutine());
+
+    /* TODO: OSv implements wait morphing here, where the wakeup
+     * primitive automatically places the woken coroutine on the
+     * mutex's queue.  This avoids the thundering herd effect.
+     */
+    if (mutex) {
+        qemu_co_mutex_lock(mutex);
+    }
 }
 
 /**
@@ -XXX,XX +XXX,XX @@ void qemu_co_rwlock_rdlock(CoRwlock *lock)
     Coroutine *self = qemu_coroutine_self();
 
     while (lock->writer) {
-        qemu_co_queue_wait(&lock->queue);
+        qemu_co_queue_wait(&lock->queue, NULL);
     }
     lock->reader++;
     self->locks_held++;
@@ -XXX,XX +XXX,XX @@ void qemu_co_rwlock_wrlock(CoRwlock *lock)
     Coroutine *self = qemu_coroutine_self();
 
     while (lock->writer || lock->reader) {
-        qemu_co_queue_wait(&lock->queue);
+        qemu_co_queue_wait(&lock->queue, NULL);
     }
     lock->writer = true;
     self->locks_held++;
-- 
2.9.3

From: Paolo Bonzini <pbonzini@redhat.com>

This adds a CoMutex around the existing CoQueue.  Because the write-side
can just take CoMutex, the old "writer" field is not necessary anymore.
Instead of removing it altogether, count the number of pending writers
during a read-side critical section and forbid further readers from
entering.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 20170213181244.16297-7-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/qemu/coroutine.h   |  3 ++-
 util/qemu-coroutine-lock.c | 35 ++++++++++++++++++++++++-----------
 2 files changed, 26 insertions(+), 12 deletions(-)

diff --git a/include/qemu/coroutine.h b/include/qemu/coroutine.h
index XXXXXXX..XXXXXXX 100644
--- a/include/qemu/coroutine.h
+++ b/include/qemu/coroutine.h
@@ -XXX,XX +XXX,XX @@ bool qemu_co_queue_empty(CoQueue *queue);
 
 
 typedef struct CoRwlock {
-    bool writer;
+    int pending_writer;
     int reader;
+    CoMutex mutex;
     CoQueue queue;
 } CoRwlock;
 
diff --git a/util/qemu-coroutine-lock.c b/util/qemu-coroutine-lock.c
index XXXXXXX..XXXXXXX 100644
--- a/util/qemu-coroutine-lock.c
+++ b/util/qemu-coroutine-lock.c
@@ -XXX,XX +XXX,XX @@ void qemu_co_rwlock_init(CoRwlock *lock)
 {
     memset(lock, 0, sizeof(*lock));
     qemu_co_queue_init(&lock->queue);
+    qemu_co_mutex_init(&lock->mutex);
 }
 
 void qemu_co_rwlock_rdlock(CoRwlock *lock)
 {
     Coroutine *self = qemu_coroutine_self();
 
-    while (lock->writer) {
-        qemu_co_queue_wait(&lock->queue, NULL);
+    qemu_co_mutex_lock(&lock->mutex);
+    /* For fairness, wait if a writer is in line.  */
+    while (lock->pending_writer) {
+        qemu_co_queue_wait(&lock->queue, &lock->mutex);
     }
     lock->reader++;
+    qemu_co_mutex_unlock(&lock->mutex);
+
+    /* The rest of the read-side critical section is run without the mutex.  */
     self->locks_held++;
 }
 
@@ -XXX,XX +XXX,XX @@ void qemu_co_rwlock_unlock(CoRwlock *lock)
     Coroutine *self = qemu_coroutine_self();
 
     assert(qemu_in_coroutine());
-    if (lock->writer) {
-        lock->writer = false;
+    if (!lock->reader) {
+        /* The critical section started in qemu_co_rwlock_wrlock.  */
         qemu_co_queue_restart_all(&lock->queue);
     } else {
+        self->locks_held--;
+
+        qemu_co_mutex_lock(&lock->mutex);
         lock->reader--;
         assert(lock->reader >= 0);
         /* Wakeup only one waiting writer */
@@ -XXX,XX +XXX,XX @@ void qemu_co_rwlock_unlock(CoRwlock *lock)
             qemu_co_queue_next(&lock->queue);
         }
     }
-    self->locks_held--;
+    qemu_co_mutex_unlock(&lock->mutex);
 }
 
 void qemu_co_rwlock_wrlock(CoRwlock *lock)
 {
-    Coroutine *self = qemu_coroutine_self();
-
-    while (lock->writer || lock->reader) {
-        qemu_co_queue_wait(&lock->queue, NULL);
+    qemu_co_mutex_lock(&lock->mutex);
+    lock->pending_writer++;
+    while (lock->reader) {
+        qemu_co_queue_wait(&lock->queue, &lock->mutex);
     }
-    lock->writer = true;
-    self->locks_held++;
+    lock->pending_writer--;
+
+    /* The rest of the write-side critical section is run with
+     * the mutex taken, so that lock->reader remains zero.
+     * There is no need to update self->locks_held.
+     */
 }
-- 
2.9.3

The following changes since commit 1214d55d1c41fbab3a9973a05085b8760647e411:

Merge remote-tracking branch 'remotes/nvme/tags/nvme-next-pull-request' into staging (2021-02-09 13:24:37 +0000)

are available in the Git repository at:

https://gitlab.com/stefanha/qemu.git tags/block-pull-request

for you to fetch changes up to eb847c42296497978942f738cd41dc29a35a49b2:

docs: fix Parallels Image "dirty bitmap" section (2021-02-10 09:23:28 +0000)

----------------------------------------------------------------
Pull request

v4:
 * Add PCI_EXPRESS Kconfig dependency to fix s390x in "multi-process: setup PCI
   host bridge for remote device" [Philippe and Thomas]

----------------------------------------------------------------

Denis V. Lunev (1):
  docs: fix Parallels Image "dirty bitmap" section

Elena Ufimtseva (8):
  multi-process: add configure and usage information
  io: add qio_channel_writev_full_all helper
  io: add qio_channel_readv_full_all_eof & qio_channel_readv_full_all
    helpers
  multi-process: define MPQemuMsg format and transmission functions
  multi-process: introduce proxy object
  multi-process: add proxy communication functions
  multi-process: Forward PCI config space acceses to the remote process
  multi-process: perform device reset in the remote process

Jagannathan Raman (11):
  memory: alloc RAM from file at offset
  multi-process: Add config option for multi-process QEMU
  multi-process: setup PCI host bridge for remote device
  multi-process: setup a machine object for remote device process
  multi-process: Initialize message handler in remote device
  multi-process: Associate fd of a PCIDevice with its object
  multi-process: setup memory manager for remote device
  multi-process: PCI BAR read/write handling for proxy & remote
    endpoints
  multi-process: Synchronize remote memory
  multi-process: create IOHUB object to handle irq
  multi-process: Retrieve PCI info from remote process

John G Johnson (1):
  multi-process: add the concept description to
    docs/devel/qemu-multiprocess

Stefan Hajnoczi (6):
  .github: point Repo Lockdown bot to GitLab repo
  gitmodules: use GitLab repos instead of qemu.org
  gitlab-ci: remove redundant GitLab repo URL command
  docs: update README to use GitLab repo URLs
  pc-bios: update mirror URLs to GitLab
  get_maintainer: update repo URL to GitLab

MAINTAINERS                               |  24 +
 README.rst                                |   4 +-
 docs/devel/index.rst                      |   1 +
 docs/devel/multi-process.rst              | 966 ++++++++++++++++++++++
 docs/system/index.rst                     |   1 +
 docs/system/multi-process.rst             |  64 ++
 docs/interop/parallels.txt                |   2 +-
 configure                                 |  10 +
 meson.build                               |   5 +-
 hw/remote/trace.h                         |   1 +
 include/exec/memory.h                     |   2 +
 include/exec/ram_addr.h                   |   4 +-
 include/hw/pci-host/remote.h              |  30 +
 include/hw/pci/pci_ids.h                  |   3 +
 include/hw/remote/iohub.h                 |  42 +
 include/hw/remote/machine.h               |  38 +
 include/hw/remote/memory.h                |  19 +
 include/hw/remote/mpqemu-link.h           |  99 +++
 include/hw/remote/proxy-memory-listener.h |  28 +
 include/hw/remote/proxy.h                 |  48 ++
 include/io/channel.h                      |  78 ++
 include/qemu/mmap-alloc.h                 |   4 +-
 include/sysemu/iothread.h                 |   6 +
 backends/hostmem-memfd.c                  |   2 +-
 hw/misc/ivshmem.c                         |   3 +-
 hw/pci-host/remote.c                      |  75 ++
 hw/remote/iohub.c                         | 119 +++
 hw/remote/machine.c                       |  80 ++
 hw/remote/memory.c                        |  65 ++
 hw/remote/message.c                       | 230 ++++++
 hw/remote/mpqemu-link.c                   | 267 ++++++
 hw/remote/proxy-memory-listener.c         | 227 +++++
 hw/remote/proxy.c                         | 379 +++++++++
 hw/remote/remote-obj.c                    | 203 +++++
 io/channel.c                              | 116 ++-
 iothread.c                                |   6 +
 softmmu/memory.c                          |   3 +-
 softmmu/physmem.c                         |  12 +-
 util/mmap-alloc.c                         |   8 +-
 util/oslib-posix.c                        |   2 +-
 .github/lockdown.yml                      |   8 +-
 .gitlab-ci.yml                            |   1 -
 .gitmodules                               |  44 +-
 Kconfig.host                              |   4 +
 hw/Kconfig                                |   1 +
 hw/meson.build                            |   1 +
 hw/pci-host/Kconfig                       |   3 +
 hw/pci-host/meson.build                   |   1 +
 hw/remote/Kconfig                         |   4 +
 hw/remote/meson.build                     |  13 +
 hw/remote/trace-events                    |   4 +
 pc-bios/README                            |   4 +-
 scripts/get_maintainer.pl                 |   2 +-
 53 files changed, 3296 insertions(+), 70 deletions(-)
 create mode 100644 docs/devel/multi-process.rst
 create mode 100644 docs/system/multi-process.rst
 create mode 100644 hw/remote/trace.h
 create mode 100644 include/hw/pci-host/remote.h
 create mode 100644 include/hw/remote/iohub.h
 create mode 100644 include/hw/remote/machine.h
 create mode 100644 include/hw/remote/memory.h
 create mode 100644 include/hw/remote/mpqemu-link.h
 create mode 100644 include/hw/remote/proxy-memory-listener.h
 create mode 100644 include/hw/remote/proxy.h
 create mode 100644 hw/pci-host/remote.c
 create mode 100644 hw/remote/iohub.c
 create mode 100644 hw/remote/machine.c
 create mode 100644 hw/remote/memory.c
 create mode 100644 hw/remote/message.c
 create mode 100644 hw/remote/mpqemu-link.c
 create mode 100644 hw/remote/proxy-memory-listener.c
 create mode 100644 hw/remote/proxy.c
 create mode 100644 hw/remote/remote-obj.c
 create mode 100644 hw/remote/Kconfig
 create mode 100644 hw/remote/meson.build
 create mode 100644 hw/remote/trace-events

-- 
2.29.2

Use the GitLab repo URL as the main repo location in order to reduce
load on qemu.org.

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-id: 20210111115017.156802-2-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 .github/lockdown.yml | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/.github/lockdown.yml b/.github/lockdown.yml
index XXXXXXX..XXXXXXX 100644
--- a/.github/lockdown.yml
+++ b/.github/lockdown.yml
@@ -XXX,XX +XXX,XX @@ issues:
   comment: |
     Thank you for your interest in the QEMU project.
 
-    This repository is a read-only mirror of the project's master
-    repostories hosted on https://git.qemu.org/git/qemu.git.
+    This repository is a read-only mirror of the project's repostories hosted
+    at https://gitlab.com/qemu-project/qemu.git.
     The project does not process issues filed on GitHub.
 
     The project issues are tracked on Launchpad:
@@ -XXX,XX +XXX,XX @@ pulls:
   comment: |
     Thank you for your interest in the QEMU project.
 
-    This repository is a read-only mirror of the project's master
-    repostories hosted on https://git.qemu.org/git/qemu.git.
+    This repository is a read-only mirror of the project's repostories hosted
+    on https://gitlab.com/qemu-project/qemu.git.
     The project does not process merge requests filed on GitHub.
 
     QEMU welcomes contributions of code (either fixing bugs or adding new
-- 
2.29.2

qemu.org is running out of bandwidth and the QEMU project is moving
towards a gating CI on GitLab. Use the GitLab repos instead of qemu.org
(they will become mirrors).

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: 20210111115017.156802-3-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 .gitmodules | 44 ++++++++++++++++++++++----------------------
 1 file changed, 22 insertions(+), 22 deletions(-)

diff --git a/.gitmodules b/.gitmodules
index XXXXXXX..XXXXXXX 100644
--- a/.gitmodules
+++ b/.gitmodules
@@ -XXX,XX +XXX,XX @@
 [submodule "roms/seabios"]
 	path = roms/seabios
-	url = https://git.qemu.org/git/seabios.git/
+	url = https://gitlab.com/qemu-project/seabios.git/
 [submodule "roms/SLOF"]
 	path = roms/SLOF
-	url = https://git.qemu.org/git/SLOF.git
+	url = https://gitlab.com/qemu-project/SLOF.git
 [submodule "roms/ipxe"]
 	path = roms/ipxe
-	url = https://git.qemu.org/git/ipxe.git
+	url = https://gitlab.com/qemu-project/ipxe.git
 [submodule "roms/openbios"]
 	path = roms/openbios
-	url = https://git.qemu.org/git/openbios.git
+	url = https://gitlab.com/qemu-project/openbios.git
 [submodule "roms/qemu-palcode"]
 	path = roms/qemu-palcode
-	url = https://git.qemu.org/git/qemu-palcode.git
+	url = https://gitlab.com/qemu-project/qemu-palcode.git
 [submodule "roms/sgabios"]
 	path = roms/sgabios
-	url = https://git.qemu.org/git/sgabios.git
+	url = https://gitlab.com/qemu-project/sgabios.git
 [submodule "dtc"]
 	path = dtc
-	url = https://git.qemu.org/git/dtc.git
+	url = https://gitlab.com/qemu-project/dtc.git
 [submodule "roms/u-boot"]
 	path = roms/u-boot
-	url = https://git.qemu.org/git/u-boot.git
+	url = https://gitlab.com/qemu-project/u-boot.git
 [submodule "roms/skiboot"]
 	path = roms/skiboot
-	url = https://git.qemu.org/git/skiboot.git
+	url = https://gitlab.com/qemu-project/skiboot.git
 [submodule "roms/QemuMacDrivers"]
 	path = roms/QemuMacDrivers
-	url = https://git.qemu.org/git/QemuMacDrivers.git
+	url = https://gitlab.com/qemu-project/QemuMacDrivers.git
 [submodule "ui/keycodemapdb"]
 	path = ui/keycodemapdb
-	url = https://git.qemu.org/git/keycodemapdb.git
+	url = https://gitlab.com/qemu-project/keycodemapdb.git
 [submodule "capstone"]
 	path = capstone
-	url = https://git.qemu.org/git/capstone.git
+	url = https://gitlab.com/qemu-project/capstone.git
 [submodule "roms/seabios-hppa"]
 	path = roms/seabios-hppa
-	url = https://git.qemu.org/git/seabios-hppa.git
+	url = https://gitlab.com/qemu-project/seabios-hppa.git
 [submodule "roms/u-boot-sam460ex"]
 	path = roms/u-boot-sam460ex
-	url = https://git.qemu.org/git/u-boot-sam460ex.git
+	url = https://gitlab.com/qemu-project/u-boot-sam460ex.git
 [submodule "tests/fp/berkeley-testfloat-3"]
 	path = tests/fp/berkeley-testfloat-3
-	url = https://git.qemu.org/git/berkeley-testfloat-3.git
+	url = https://gitlab.com/qemu-project/berkeley-testfloat-3.git
 [submodule "tests/fp/berkeley-softfloat-3"]
 	path = tests/fp/berkeley-softfloat-3
-	url = https://git.qemu.org/git/berkeley-softfloat-3.git
+	url = https://gitlab.com/qemu-project/berkeley-softfloat-3.git
 [submodule "roms/edk2"]
 	path = roms/edk2
-	url = https://git.qemu.org/git/edk2.git
+	url = https://gitlab.com/qemu-project/edk2.git
 [submodule "slirp"]
 	path = slirp
-	url = https://git.qemu.org/git/libslirp.git
+	url = https://gitlab.com/qemu-project/libslirp.git
 [submodule "roms/opensbi"]
 	path = roms/opensbi
-	url = 	https://git.qemu.org/git/opensbi.git
+	url = 	https://gitlab.com/qemu-project/opensbi.git
 [submodule "roms/qboot"]
 	path = roms/qboot
-	url = https://git.qemu.org/git/qboot.git
+	url = https://gitlab.com/qemu-project/qboot.git
 [submodule "meson"]
 	path = meson
-	url = https://git.qemu.org/git/meson.git
+	url = https://gitlab.com/qemu-project/meson.git
 [submodule "roms/vbootrom"]
 	path = roms/vbootrom
-	url = https://git.qemu.org/git/vbootrom.git
+	url = https://gitlab.com/qemu-project/vbootrom.git
-- 
2.29.2

qemu.org is running out of bandwidth and the QEMU project is moving
towards a gating CI on GitLab. Use the GitLab repos instead of qemu.org
(they will become mirrors).

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: 20210111115017.156802-5-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 README.rst | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/README.rst b/README.rst
index XXXXXXX..XXXXXXX 100644
--- a/README.rst
+++ b/README.rst
@@ -XXX,XX +XXX,XX @@ The QEMU source code is maintained under the GIT version control system.
 
 .. code-block:: shell
 
-   git clone https://git.qemu.org/git/qemu.git
+   git clone https://gitlab.com/qemu-project/qemu.git
 
 When submitting patches, one common approach is to use 'git
 format-patch' and/or 'git send-email' to format & send the mail to the
@@ -XXX,XX +XXX,XX @@ The QEMU website is also maintained under source control.
 
 .. code-block:: shell
 
-  git clone https://git.qemu.org/git/qemu-web.git
+  git clone https://gitlab.com/qemu-project/qemu-web.git
 
 * `<https://www.qemu.org/2017/02/04/the-new-qemu-website-is-up/>`_
 
-- 
2.29.2

qemu.org is running out of bandwidth and the QEMU project is moving
towards a gating CI on GitLab. Use the GitLab repos instead of qemu.org
(they will become mirrors).

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: 20210111115017.156802-6-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 pc-bios/README | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/pc-bios/README b/pc-bios/README
index XXXXXXX..XXXXXXX 100644
--- a/pc-bios/README
+++ b/pc-bios/README
@@ -XXX,XX +XXX,XX @@
   legacy x86 software to communicate with an attached serial console as
   if a video card were attached.  The master sources reside in a subversion
   repository at http://sgabios.googlecode.com/svn/trunk.  A git mirror is
-  available at https://git.qemu.org/git/sgabios.git.
+  available at https://gitlab.com/qemu-project/sgabios.git.
 
 - The PXE roms come from the iPXE project. Built with BANNER_TIME 0.
   Sources available at http://ipxe.org.  Vendor:Device ID -> ROM mapping:
@@ -XXX,XX +XXX,XX @@
 
 - The u-boot binary for e500 comes from the upstream denx u-boot project where
   it was compiled using the qemu-ppce500 target.
-  A git mirror is available at: https://git.qemu.org/git/u-boot.git
+  A git mirror is available at: https://gitlab.com/qemu-project/u-boot.git
   The hash used to compile the current version is: 2072e72
 
 - Skiboot (https://github.com/open-power/skiboot/) is an OPAL
-- 
2.29.2

qemu.org is running out of bandwidth and the QEMU project is moving
towards a gating CI on GitLab. Use the GitLab repos instead of qemu.org
(they will become mirrors).

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: 20210111115017.156802-7-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 scripts/get_maintainer.pl | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/scripts/get_maintainer.pl b/scripts/get_maintainer.pl
index XXXXXXX..XXXXXXX 100755
--- a/scripts/get_maintainer.pl
+++ b/scripts/get_maintainer.pl
@@ -XXX,XX +XXX,XX @@ sub vcs_exists {
 	warn("$P: No supported VCS found.  Add --nogit to options?\n");
 	warn("Using a git repository produces better results.\n");
 	warn("Try latest git repository using:\n");
-	warn("git clone https://git.qemu.org/git/qemu.git\n");
+	warn("git clone https://gitlab.com/qemu-project/qemu.git\n");
 	$printed_novcs = 1;
     }
     return 0;
-- 
2.29.2

From: John G Johnson <john.g.johnson@oracle.com>

Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 02a68adef99f5df6a380bf8fd7b90948777e411c.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 MAINTAINERS                  |   7 +
 docs/devel/index.rst         |   1 +
 docs/devel/multi-process.rst | 966 +++++++++++++++++++++++++++++++++++
 3 files changed, 974 insertions(+)
 create mode 100644 docs/devel/multi-process.rst

diff --git a/MAINTAINERS b/MAINTAINERS
index XXXXXXX..XXXXXXX 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -XXX,XX +XXX,XX @@ S: Maintained
 F: hw/semihosting/
 F: include/hw/semihosting/
 
+Multi-process QEMU
+M: Elena Ufimtseva <elena.ufimtseva@oracle.com>
+M: Jagannathan Raman <jag.raman@oracle.com>
+M: John G Johnson <john.g.johnson@oracle.com>
+S: Maintained
+F: docs/devel/multi-process.rst
+
 Build and test automation
 -------------------------
 Build and test automation
diff --git a/docs/devel/index.rst b/docs/devel/index.rst
index XXXXXXX..XXXXXXX 100644
--- a/docs/devel/index.rst
+++ b/docs/devel/index.rst
@@ -XXX,XX +XXX,XX @@ Contents:
    clocks
    qom
    block-coroutine-wrapper
+   multi-process
diff --git a/docs/devel/multi-process.rst b/docs/devel/multi-process.rst
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/docs/devel/multi-process.rst
@@ -XXX,XX +XXX,XX @@
+This is the design document for multi-process QEMU. It does not
+necessarily reflect the status of the current implementation, which
+may lack features or be considerably different from what is described
+in this document. This document is still useful as a description of
+the goals and general direction of this feature.
+
+Please refer to the following wiki for latest details:
+https://wiki.qemu.org/Features/MultiProcessQEMU
+
+Multi-process QEMU
+===================
+
+QEMU is often used as the hypervisor for virtual machines running in the
+Oracle cloud. Since one of the advantages of cloud computing is the
+ability to run many VMs from different tenants in the same cloud
+infrastructure, a guest that compromised its hypervisor could
+potentially use the hypervisor's access privileges to access data it is
+not authorized for.
+
+QEMU can be susceptible to security attacks because it is a large,
+monolithic program that provides many features to the VMs it services.
+Many of these features can be configured out of QEMU, but even a reduced
+configuration QEMU has a large amount of code a guest can potentially
+attack. Separating QEMU reduces the attack surface by aiding to
+limit each component in the system to only access the resources that
+it needs to perform its job.
+
+QEMU services
+-------------
+
+QEMU can be broadly described as providing three main services. One is a
+VM control point, where VMs can be created, migrated, re-configured, and
+destroyed. A second is to emulate the CPU instructions within the VM,
+often accelerated by HW virtualization features such as Intel's VT
+extensions. Finally, it provides IO services to the VM by emulating HW
+IO devices, such as disk and network devices.
+
+A multi-process QEMU
+~~~~~~~~~~~~~~~~~~~~
+
+A multi-process QEMU involves separating QEMU services into separate
+host processes. Each of these processes can be given only the privileges
+it needs to provide its service, e.g., a disk service could be given
+access only to the disk images it provides, and not be allowed to
+access other files, or any network devices. An attacker who compromised
+this service would not be able to use this exploit to access files or
+devices beyond what the disk service was given access to.
+
+A QEMU control process would remain, but in multi-process mode, will
+have no direct interfaces to the VM. During VM execution, it would still
+provide the user interface to hot-plug devices or live migrate the VM.
+
+A first step in creating a multi-process QEMU is to separate IO services
+from the main QEMU program, which would continue to provide CPU
+emulation. i.e., the control process would also be the CPU emulation
+process. In a later phase, CPU emulation could be separated from the
+control process.
+
+Separating IO services
+----------------------
+
+Separating IO services into individual host processes is a good place to
+begin for a couple of reasons. One is the sheer number of IO devices QEMU
+can emulate provides a large surface of interfaces which could potentially
+be exploited, and, indeed, have been a source of exploits in the past.
+Another is the modular nature of QEMU device emulation code provides
+interface points where the QEMU functions that perform device emulation
+can be separated from the QEMU functions that manage the emulation of
+guest CPU instructions. The devices emulated in the separate process are
+referred to as remote devices.
+
+QEMU device emulation
+~~~~~~~~~~~~~~~~~~~~~
+
+QEMU uses an object oriented SW architecture for device emulation code.
+Configured objects are all compiled into the QEMU binary, then objects
+are instantiated by name when used by the guest VM. For example, the
+code to emulate a device named "foo" is always present in QEMU, but its
+instantiation code is only run when the device is included in the target
+VM. (e.g., via the QEMU command line as *-device foo*)
+
+The object model is hierarchical, so device emulation code names its
+parent object (such as "pci-device" for a PCI device) and QEMU will
+instantiate a parent object before calling the device's instantiation
+code.
+
+Current separation models
+~~~~~~~~~~~~~~~~~~~~~~~~~
+
+In order to separate the device emulation code from the CPU emulation
+code, the device object code must run in a different process. There are
+a couple of existing QEMU features that can run emulation code
+separately from the main QEMU process. These are examined below.
+
+vhost user model
+^^^^^^^^^^^^^^^^
+
+Virtio guest device drivers can be connected to vhost user applications
+in order to perform their IO operations. This model uses special virtio
+device drivers in the guest and vhost user device objects in QEMU, but
+once the QEMU vhost user code has configured the vhost user application,
+mission-mode IO is performed by the application. The vhost user
+application is a daemon process that can be contacted via a known UNIX
+domain socket.
+
+vhost socket
+''''''''''''
+
+As mentioned above, one of the tasks of the vhost device object within
+QEMU is to contact the vhost application and send it configuration
+information about this device instance. As part of the configuration
+process, the application can also be sent other file descriptors over
+the socket, which then can be used by the vhost user application in
+various ways, some of which are described below.
+
+vhost MMIO store acceleration
+'''''''''''''''''''''''''''''
+
+VMs are often run using HW virtualization features via the KVM kernel
+driver. This driver allows QEMU to accelerate the emulation of guest CPU
+instructions by running the guest in a virtual HW mode. When the guest
+executes instructions that cannot be executed by virtual HW mode,
+execution returns to the KVM driver so it can inform QEMU to emulate the
+instructions in SW.
+
+One of the events that can cause a return to QEMU is when a guest device
+driver accesses an IO location. QEMU then dispatches the memory
+operation to the corresponding QEMU device object. In the case of a
+vhost user device, the memory operation would need to be sent over a
+socket to the vhost application. This path is accelerated by the QEMU
+virtio code by setting up an eventfd file descriptor that the vhost
+application can directly receive MMIO store notifications from the KVM
+driver, instead of needing them to be sent to the QEMU process first.
+
+vhost interrupt acceleration
+''''''''''''''''''''''''''''
+
+Another optimization used by the vhost application is the ability to
+directly inject interrupts into the VM via the KVM driver, again,
+bypassing the need to send the interrupt back to the QEMU process first.
+The QEMU virtio setup code configures the KVM driver with an eventfd
+that triggers the device interrupt in the guest when the eventfd is
+written. This irqfd file descriptor is then passed to the vhost user
+application program.
+
+vhost access to guest memory
+''''''''''''''''''''''''''''
+
+The vhost application is also allowed to directly access guest memory,
+instead of needing to send the data as messages to QEMU. This is also
+done with file descriptors sent to the vhost user application by QEMU.
+These descriptors can be passed to ``mmap()`` by the vhost application
+to map the guest address space into the vhost application.
+
+IOMMUs introduce another level of complexity, since the address given to
+the guest virtio device to DMA to or from is not a guest physical
+address. This case is handled by having vhost code within QEMU register
+as a listener for IOMMU mapping changes. The vhost application maintains
+a cache of IOMMMU translations: sending translation requests back to
+QEMU on cache misses, and in turn receiving flush requests from QEMU
+when mappings are purged.
+
+applicability to device separation
+''''''''''''''''''''''''''''''''''
+
+Much of the vhost model can be re-used by separated device emulation. In
+particular, the ideas of using a socket between QEMU and the device
+emulation application, using a file descriptor to inject interrupts into
+the VM via KVM, and allowing the application to ``mmap()`` the guest
+should be re used.
+
+There are, however, some notable differences between how a vhost
+application works and the needs of separated device emulation. The most
+basic is that vhost uses custom virtio device drivers which always
+trigger IO with MMIO stores. A separated device emulation model must
+work with existing IO device models and guest device drivers. MMIO loads
+break vhost store acceleration since they are synchronous - guest
+progress cannot continue until the load has been emulated. By contrast,
+stores are asynchronous, the guest can continue after the store event
+has been sent to the vhost application.
+
+Another difference is that in the vhost user model, a single daemon can
+support multiple QEMU instances. This is contrary to the security regime
+desired, in which the emulation application should only be allowed to
+access the files or devices the VM it's running on behalf of can access.
+#### qemu-io model
+
+Qemu-io is a test harness used to test changes to the QEMU block backend
+object code. (e.g., the code that implements disk images for disk driver
+emulation) Qemu-io is not a device emulation application per se, but it
+does compile the QEMU block objects into a separate binary from the main
+QEMU one. This could be useful for disk device emulation, since its
+emulation applications will need to include the QEMU block objects.
+
+New separation model based on proxy objects
+-------------------------------------------
+
+A different model based on proxy objects in the QEMU program
+communicating with remote emulation programs could provide separation
+while minimizing the changes needed to the device emulation code. The
+rest of this section is a discussion of how a proxy object model would
+work.
+
+Remote emulation processes
+~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The remote emulation process will run the QEMU object hierarchy without
+modification. The device emulation objects will be also be based on the
+QEMU code, because for anything but the simplest device, it would not be
+a tractable to re-implement both the object model and the many device
+backends that QEMU has.
+
+The processes will communicate with the QEMU process over UNIX domain
+sockets. The processes can be executed either as standalone processes,
+or be executed by QEMU. In both cases, the host backends the emulation
+processes will provide are specified on its command line, as they would
+be for QEMU. For example:
+
+::
+
+    disk-proc -blockdev driver=file,node-name=file0,filename=disk-file0  \
+    -blockdev driver=qcow2,node-name=drive0,file=file0
+
+would indicate process *disk-proc* uses a qcow2 emulated disk named
+*file0* as its backend.
+
+Emulation processes may emulate more than one guest controller. A common
+configuration might be to put all controllers of the same device class
+(e.g., disk, network, etc.) in a single process, so that all backends of
+the same type can be managed by a single QMP monitor.
+
+communication with QEMU
+^^^^^^^^^^^^^^^^^^^^^^^
+
+The first argument to the remote emulation process will be a Unix domain
+socket that connects with the Proxy object. This is a required argument.
+
+::
+
+    disk-proc <socket number> <backend list>
+
+remote process QMP monitor
+^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+Remote emulation processes can be monitored via QMP, similar to QEMU
+itself. The QMP monitor socket is specified the same as for a QEMU
+process:
+
+::
+
+    disk-proc -qmp unix:/tmp/disk-mon,server
+
+can be monitored over the UNIX socket path */tmp/disk-mon*.
+
+QEMU command line
+~~~~~~~~~~~~~~~~~
+
+Each remote device emulated in a remote process on the host is
+represented as a *-device* of type *pci-proxy-dev*. A socket
+sub-option to this option specifies the Unix socket that connects
+to the remote process. An *id* sub-option is required, and it should
+be the same id as used in the remote process.
+
+::
+
+    qemu-system-x86_64 ... -device pci-proxy-dev,id=lsi0,socket=3
+
+can be used to add a device emulated in a remote process
+
+
+QEMU management of remote processes
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+QEMU is not aware of the type of type of the remote PCI device. It is
+a pass through device as far as QEMU is concerned.
+
+communication with emulation process
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+primary channel
+'''''''''''''''
+
+The primary channel (referred to as com in the code) is used to bootstrap
+the remote process. It is also used to pass on device-agnostic commands
+like reset.
+
+per-device channels
+'''''''''''''''''''
+
+Each remote device communicates with QEMU using a dedicated communication
+channel. The proxy object sets up this channel using the primary
+channel during its initialization.
+
+QEMU device proxy objects
+~~~~~~~~~~~~~~~~~~~~~~~~~
+
+QEMU has an object model based on sub-classes inherited from the
+"object" super-class. The sub-classes that are of interest here are the
+"device" and "bus" sub-classes whose child sub-classes make up the
+device tree of a QEMU emulated system.
+
+The proxy object model will use device proxy objects to replace the
+device emulation code within the QEMU process. These objects will live
+in the same place in the object and bus hierarchies as the objects they
+replace. i.e., the proxy object for an LSI SCSI controller will be a
+sub-class of the "pci-device" class, and will have the same PCI bus
+parent and the same SCSI bus child objects as the LSI controller object
+it replaces.
+
+It is worth noting that the same proxy object is used to mediate with
+all types of remote PCI devices.
+
+object initialization
+^^^^^^^^^^^^^^^^^^^^^
+
+The Proxy device objects are initialized in the exact same manner in
+which any other QEMU device would be initialized.
+
+In addition, the Proxy objects perform the following two tasks:
+- Parses the "socket" sub option and connects to the remote process
+using this channel
+- Uses the "id" sub-option to connect to the emulated device on the
+separate process
+
+class\_init
+'''''''''''
+
+The ``class_init()`` method of a proxy object will, in general behave
+similarly to the object it replaces, including setting any static
+properties and methods needed by the proxy.
+
+instance\_init / realize
+''''''''''''''''''''''''
+
+The ``instance_init()`` and ``realize()`` functions would only need to
+perform tasks related to being a proxy, such are registering its own
+MMIO handlers, or creating a child bus that other proxy devices can be
+attached to later.
+
+Other tasks will be device-specific. For example, PCI device objects
+will initialize the PCI config space in order to make a valid PCI device
+tree within the QEMU process.
+
+address space registration
+^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+Most devices are driven by guest device driver accesses to IO addresses
+or ports. The QEMU device emulation code uses QEMU's memory region
+function calls (such as ``memory_region_init_io()``) to add callback
+functions that QEMU will invoke when the guest accesses the device's
+areas of the IO address space. When a guest driver does access the
+device, the VM will exit HW virtualization mode and return to QEMU,
+which will then lookup and execute the corresponding callback function.
+
+A proxy object would need to mirror the memory region calls the actual
+device emulator would perform in its initialization code, but with its
+own callbacks. When invoked by QEMU as a result of a guest IO operation,
+they will forward the operation to the device emulation process.
+
+PCI config space
+^^^^^^^^^^^^^^^^
+
+PCI devices also have a configuration space that can be accessed by the
+guest driver. Guest accesses to this space is not handled by the device
+emulation object, but by its PCI parent object. Much of this space is
+read-only, but certain registers (especially BAR and MSI-related ones)
+need to be propagated to the emulation process.
+
+PCI parent proxy
+''''''''''''''''
+
+One way to propagate guest PCI config accesses is to create a
+"pci-device-proxy" class that can serve as the parent of a PCI device
+proxy object. This class's parent would be "pci-device" and it would
+override the PCI parent's ``config_read()`` and ``config_write()``
+methods with ones that forward these operations to the emulation
+program.
+
+interrupt receipt
+^^^^^^^^^^^^^^^^^
+
+A proxy for a device that generates interrupts will need to create a
+socket to receive interrupt indications from the emulation process. An
+incoming interrupt indication would then be sent up to its bus parent to
+be injected into the guest. For example, a PCI device object may use
+``pci_set_irq()``.
+
+live migration
+^^^^^^^^^^^^^^
+
+The proxy will register to save and restore any *vmstate* it needs over
+a live migration event. The device proxy does not need to manage the
+remote device's *vmstate*; that will be handled by the remote process
+proxy (see below).
+
+QEMU remote device operation
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Generic device operations, such as DMA, will be performed by the remote
+process proxy by sending messages to the remote process.
+
+DMA operations
+^^^^^^^^^^^^^^
+
+DMA operations would be handled much like vhost applications do. One of
+the initial messages sent to the emulation process is a guest memory
+table. Each entry in this table consists of a file descriptor and size
+that the emulation process can ``mmap()`` to directly access guest
+memory, similar to ``vhost_user_set_mem_table()``. Note guest memory
+must be backed by file descriptors, such as when QEMU is given the
+*-mem-path* command line option.
+
+IOMMU operations
+^^^^^^^^^^^^^^^^
+
+When the emulated system includes an IOMMU, the remote process proxy in
+QEMU will need to create a socket for IOMMU requests from the emulation
+process. It will handle those requests with an
+``address_space_get_iotlb_entry()`` call. In order to handle IOMMU
+unmaps, the remote process proxy will also register as a listener on the
+device's DMA address space. When an IOMMU memory region is created
+within the DMA address space, an IOMMU notifier for unmaps will be added
+to the memory region that will forward unmaps to the emulation process
+over the IOMMU socket.
+
+device hot-plug via QMP
+^^^^^^^^^^^^^^^^^^^^^^^
+
+An QMP "device\_add" command can add a device emulated by a remote
+process. It will also have "rid" option to the command, just as the
+*-device* command line option does. The remote process may either be one
+started at QEMU startup, or be one added by the "add-process" QMP
+command described above. In either case, the remote process proxy will
+forward the new device's JSON description to the corresponding emulation
+process.
+
+live migration
+^^^^^^^^^^^^^^
+
+The remote process proxy will also register for live migration
+notifications with ``vmstate_register()``. When called to save state,
+the proxy will send the remote process a secondary socket file
+descriptor to save the remote process's device *vmstate* over. The
+incoming byte stream length and data will be saved as the proxy's
+*vmstate*. When the proxy is resumed on its new host, this *vmstate*
+will be extracted, and a secondary socket file descriptor will be sent
+to the new remote process through which it receives the *vmstate* in
+order to restore the devices there.
+
+device emulation in remote process
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The parts of QEMU that the emulation program will need include the
+object model; the memory emulation objects; the device emulation objects
+of the targeted device, and any dependent devices; and, the device's
+backends. It will also need code to setup the machine environment,
+handle requests from the QEMU process, and route machine-level requests
+(such as interrupts or IOMMU mappings) back to the QEMU process.
+
+initialization
+^^^^^^^^^^^^^^
+
+The process initialization sequence will follow the same sequence
+followed by QEMU. It will first initialize the backend objects, then
+device emulation objects. The JSON descriptions sent by the QEMU process
+will drive which objects need to be created.
+
+-  address spaces
+
+Before the device objects are created, the initial address spaces and
+memory regions must be configured with ``memory_map_init()``. This
+creates a RAM memory region object (*system\_memory*) and an IO memory
+region object (*system\_io*).
+
+-  RAM
+
+RAM memory region creation will follow how ``pc_memory_init()`` creates
+them, but must use ``memory_region_init_ram_from_fd()`` instead of
+``memory_region_allocate_system_memory()``. The file descriptors needed
+will be supplied by the guest memory table from above. Those RAM regions
+would then be added to the *system\_memory* memory region with
+``memory_region_add_subregion()``.
+
+-  PCI
+
+IO initialization will be driven by the JSON descriptions sent from the
+QEMU process. For a PCI device, a PCI bus will need to be created with
+``pci_root_bus_new()``, and a PCI memory region will need to be created
+and added to the *system\_memory* memory region with
+``memory_region_add_subregion_overlap()``. The overlap version is
+required for architectures where PCI memory overlaps with RAM memory.
+
+MMIO handling
+^^^^^^^^^^^^^
+
+The device emulation objects will use ``memory_region_init_io()`` to
+install their MMIO handlers, and ``pci_register_bar()`` to associate
+those handlers with a PCI BAR, as they do within QEMU currently.
+
+In order to use ``address_space_rw()`` in the emulation process to
+handle MMIO requests from QEMU, the PCI physical addresses must be the
+same in the QEMU process and the device emulation process. In order to
+accomplish that, guest BAR programming must also be forwarded from QEMU
+to the emulation process.
+
+interrupt injection
+^^^^^^^^^^^^^^^^^^^
+
+When device emulation wants to inject an interrupt into the VM, the
+request climbs the device's bus object hierarchy until the point where a
+bus object knows how to signal the interrupt to the guest. The details
+depend on the type of interrupt being raised.
+
+-  PCI pin interrupts
+
+On x86 systems, there is an emulated IOAPIC object attached to the root
+PCI bus object, and the root PCI object forwards interrupt requests to
+it. The IOAPIC object, in turn, calls the KVM driver to inject the
+corresponding interrupt into the VM. The simplest way to handle this in
+an emulation process would be to setup the root PCI bus driver (via
+``pci_bus_irqs()``) to send a interrupt request back to the QEMU
+process, and have the device proxy object reflect it up the PCI tree
+there.
+
+-  PCI MSI/X interrupts
+
+PCI MSI/X interrupts are implemented in HW as DMA writes to a
+CPU-specific PCI address. In QEMU on x86, a KVM APIC object receives
+these DMA writes, then calls into the KVM driver to inject the interrupt
+into the VM. A simple emulation process implementation would be to send
+the MSI DMA address from QEMU as a message at initialization, then
+install an address space handler at that address which forwards the MSI
+message back to QEMU.
+
+DMA operations
+^^^^^^^^^^^^^^
+
+When a emulation object wants to DMA into or out of guest memory, it
+first must use dma\_memory\_map() to convert the DMA address to a local
+virtual address. The emulation process memory region objects setup above
+will be used to translate the DMA address to a local virtual address the
+device emulation code can access.
+
+IOMMU
+^^^^^
+
+When an IOMMU is in use in QEMU, DMA translation uses IOMMU memory
+regions to translate the DMA address to a guest physical address before
+that physical address can be translated to a local virtual address. The
+emulation process will need similar functionality.
+
+-  IOTLB cache
+
+The emulation process will maintain a cache of recent IOMMU translations
+(the IOTLB). When the translate() callback of an IOMMU memory region is
+invoked, the IOTLB cache will be searched for an entry that will map the
+DMA address to a guest PA. On a cache miss, a message will be sent back
+to QEMU requesting the corresponding translation entry, which be both be
+used to return a guest address and be added to the cache.
+
+-  IOTLB purge
+
+The IOMMU emulation will also need to act on unmap requests from QEMU.
+These happen when the guest IOMMU driver purges an entry from the
+guest's translation table.
+
+live migration
+^^^^^^^^^^^^^^
+
+When a remote process receives a live migration indication from QEMU, it
+will set up a channel using the received file descriptor with
+``qio_channel_socket_new_fd()``. This channel will be used to create a
+*QEMUfile* that can be passed to ``qemu_save_device_state()`` to send
+the process's device state back to QEMU. This method will be reversed on
+restore - the channel will be passed to ``qemu_loadvm_state()`` to
+restore the device state.
+
+Accelerating device emulation
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The messages that are required to be sent between QEMU and the emulation
+process can add considerable latency to IO operations. The optimizations
+described below attempt to ameliorate this effect by allowing the
+emulation process to communicate directly with the kernel KVM driver.
+The KVM file descriptors created would be passed to the emulation process
+via initialization messages, much like the guest memory table is done.
+#### MMIO acceleration
+
+Vhost user applications can receive guest virtio driver stores directly
+from KVM. The issue with the eventfd mechanism used by vhost user is
+that it does not pass any data with the event indication, so it cannot
+handle guest loads or guest stores that carry store data. This concept
+could, however, be expanded to cover more cases.
+
+The expanded idea would require a new type of KVM device:
+*KVM\_DEV\_TYPE\_USER*. This device has two file descriptors: a master
+descriptor that QEMU can use for configuration, and a slave descriptor
+that the emulation process can use to receive MMIO notifications. QEMU
+would create both descriptors using the KVM driver, and pass the slave
+descriptor to the emulation process via an initialization message.
+
+data structures
+^^^^^^^^^^^^^^^
+
+-  guest physical range
+
+The guest physical range structure describes the address range that a
+device will respond to. It includes the base and length of the range, as
+well as which bus the range resides on (e.g., on an x86machine, it can
+specify whether the range refers to memory or IO addresses).
+
+A device can have multiple physical address ranges it responds to (e.g.,
+a PCI device can have multiple BARs), so the structure will also include
+an enumerated identifier to specify which of the device's ranges is
+being referred to.
+
++--------+----------------------------+
+| Name   | Description                |
++========+============================+
+| addr   | range base address         |
++--------+----------------------------+
+| len    | range length               |
++--------+----------------------------+
+| bus    | addr type (memory or IO)   |
++--------+----------------------------+
+| id     | range ID (e.g., PCI BAR)   |
++--------+----------------------------+
+
+-  MMIO request structure
+
+This structure describes an MMIO operation. It includes which guest
+physical range the MMIO was within, the offset within that range, the
+MMIO type (e.g., load or store), and its length and data. It also
+includes a sequence number that can be used to reply to the MMIO, and
+the CPU that issued the MMIO.
+
++----------+------------------------+
+| Name     | Description            |
++==========+========================+
+| rid      | range MMIO is within   |
++----------+------------------------+
+| offset   | offset withing *rid*   |
++----------+------------------------+
+| type     | e.g., load or store    |
++----------+------------------------+
+| len      | MMIO length            |
++----------+------------------------+
+| data     | store data             |
++----------+------------------------+
+| seq      | sequence ID            |
++----------+------------------------+
+
+-  MMIO request queues
+
+MMIO request queues are FIFO arrays of MMIO request structures. There
+are two queues: pending queue is for MMIOs that haven't been read by the
+emulation program, and the sent queue is for MMIOs that haven't been
+acknowledged. The main use of the second queue is to validate MMIO
+replies from the emulation program.
+
+-  scoreboard
+
+Each CPU in the VM is emulated in QEMU by a separate thread, so multiple
+MMIOs may be waiting to be consumed by an emulation program and multiple
+threads may be waiting for MMIO replies. The scoreboard would contain a
+wait queue and sequence number for the per-CPU threads, allowing them to
+be individually woken when the MMIO reply is received from the emulation
+program. It also tracks the number of posted MMIO stores to the device
+that haven't been replied to, in order to satisfy the PCI constraint
+that a load to a device will not complete until all previous stores to
+that device have been completed.
+
+-  device shadow memory
+
+Some MMIO loads do not have device side-effects. These MMIOs can be
+completed without sending a MMIO request to the emulation program if the
+emulation program shares a shadow image of the device's memory image
+with the KVM driver.
+
+The emulation program will ask the KVM driver to allocate memory for the
+shadow image, and will then use ``mmap()`` to directly access it. The
+emulation program can control KVM access to the shadow image by sending
+KVM an access map telling it which areas of the image have no
+side-effects (and can be completed immediately), and which require a
+MMIO request to the emulation program. The access map can also inform
+the KVM drive which size accesses are allowed to the image.
+
+master descriptor
+^^^^^^^^^^^^^^^^^
+
+The master descriptor is used by QEMU to configure the new KVM device.
+The descriptor would be returned by the KVM driver when QEMU issues a
+*KVM\_CREATE\_DEVICE* ``ioctl()`` with a *KVM\_DEV\_TYPE\_USER* type.
+
+KVM\_DEV\_TYPE\_USER device ops
+
+
+The *KVM\_DEV\_TYPE\_USER* operations vector will be registered by a
+``kvm_register_device_ops()`` call when the KVM system in initialized by
+``kvm_init()``. These device ops are called by the KVM driver when QEMU
+executes certain ``ioctl()`` operations on its KVM file descriptor. They
+include:
+
+-  create
+
+This routine is called when QEMU issues a *KVM\_CREATE\_DEVICE*
+``ioctl()`` on its per-VM file descriptor. It will allocate and
+initialize a KVM user device specific data structure, and assign the
+*kvm\_device* private field to it.
+
+-  ioctl
+
+This routine is invoked when QEMU issues an ``ioctl()`` on the master
+descriptor. The ``ioctl()`` commands supported are defined by the KVM
+device type. *KVM\_DEV\_TYPE\_USER* ones will need several commands:
+
+*KVM\_DEV\_USER\_SLAVE\_FD* creates the slave file descriptor that will
+be passed to the device emulation program. Only one slave can be created
+by each master descriptor. The file operations performed by this
+descriptor are described below.
+
+The *KVM\_DEV\_USER\_PA\_RANGE* command configures a guest physical
+address range that the slave descriptor will receive MMIO notifications
+for. The range is specified by a guest physical range structure
+argument. For buses that assign addresses to devices dynamically, this
+command can be executed while the guest is running, such as the case
+when a guest changes a device's PCI BAR registers.
+
+*KVM\_DEV\_USER\_PA\_RANGE* will use ``kvm_io_bus_register_dev()`` to
+register *kvm\_io\_device\_ops* callbacks to be invoked when the guest
+performs a MMIO operation within the range. When a range is changed,
+``kvm_io_bus_unregister_dev()`` is used to remove the previous
+instantiation.
+
+*KVM\_DEV\_USER\_TIMEOUT* will configure a timeout value that specifies
+how long KVM will wait for the emulation process to respond to a MMIO
+indication.
+
+-  destroy
+
+This routine is called when the VM instance is destroyed. It will need
+to destroy the slave descriptor; and free any memory allocated by the
+driver, as well as the *kvm\_device* structure itself.
+
+slave descriptor
+^^^^^^^^^^^^^^^^
+
+The slave descriptor will have its own file operations vector, which
+responds to system calls on the descriptor performed by the device
+emulation program.
+
+-  read
+
+A read returns any pending MMIO requests from the KVM driver as MMIO
+request structures. Multiple structures can be returned if there are
+multiple MMIO operations pending. The MMIO requests are moved from the
+pending queue to the sent queue, and if there are threads waiting for
+space in the pending to add new MMIO operations, they will be woken
+here.
+
+-  write
+
+A write also consists of a set of MMIO requests. They are compared to
+the MMIO requests in the sent queue. Matches are removed from the sent
+queue, and any threads waiting for the reply are woken. If a store is
+removed, then the number of posted stores in the per-CPU scoreboard is
+decremented. When the number is zero, and a non side-effect load was
+waiting for posted stores to complete, the load is continued.
+
+-  ioctl
+
+There are several ioctl()s that can be performed on the slave
+descriptor.
+
+A *KVM\_DEV\_USER\_SHADOW\_SIZE* ``ioctl()`` causes the KVM driver to
+allocate memory for the shadow image. This memory can later be
+``mmap()``\ ed by the emulation process to share the emulation's view of
+device memory with the KVM driver.
+
+A *KVM\_DEV\_USER\_SHADOW\_CTRL* ``ioctl()`` controls access to the
+shadow image. It will send the KVM driver a shadow control map, which
+specifies which areas of the image can complete guest loads without
+sending the load request to the emulation program. It will also specify
+the size of load operations that are allowed.
+
+-  poll
+
+An emulation program will use the ``poll()`` call with a *POLLIN* flag
+to determine if there are MMIO requests waiting to be read. It will
+return if the pending MMIO request queue is not empty.
+
+-  mmap
+
+This call allows the emulation program to directly access the shadow
+image allocated by the KVM driver. As device emulation updates device
+memory, changes with no side-effects will be reflected in the shadow,
+and the KVM driver can satisfy guest loads from the shadow image without
+needing to wait for the emulation program.
+
+kvm\_io\_device ops
+^^^^^^^^^^^^^^^^^^^
+
+Each KVM per-CPU thread can handle MMIO operation on behalf of the guest
+VM. KVM will use the MMIO's guest physical address to search for a
+matching *kvm\_io\_device* to see if the MMIO can be handled by the KVM
+driver instead of exiting back to QEMU. If a match is found, the
+corresponding callback will be invoked.
+
+-  read
+
+This callback is invoked when the guest performs a load to the device.
+Loads with side-effects must be handled synchronously, with the KVM
+driver putting the QEMU thread to sleep waiting for the emulation
+process reply before re-starting the guest. Loads that do not have
+side-effects may be optimized by satisfying them from the shadow image,
+if there are no outstanding stores to the device by this CPU. PCI memory
+ordering demands that a load cannot complete before all older stores to
+the same device have been completed.
+
+-  write
+
+Stores can be handled asynchronously unless the pending MMIO request
+queue is full. In this case, the QEMU thread must sleep waiting for
+space in the queue. Stores will increment the number of posted stores in
+the per-CPU scoreboard, in order to implement the PCI ordering
+constraint above.
+
+interrupt acceleration
+^^^^^^^^^^^^^^^^^^^^^^
+
+This performance optimization would work much like a vhost user
+application does, where the QEMU process sets up *eventfds* that cause
+the device's corresponding interrupt to be triggered by the KVM driver.
+These irq file descriptors are sent to the emulation process at
+initialization, and are used when the emulation code raises a device
+interrupt.
+
+intx acceleration
+'''''''''''''''''
+
+Traditional PCI pin interrupts are level based, so, in addition to an
+irq file descriptor, a re-sampling file descriptor needs to be sent to
+the emulation program. This second file descriptor allows multiple
+devices sharing an irq to be notified when the interrupt has been
+acknowledged by the guest, so they can re-trigger the interrupt if their
+device has not de-asserted its interrupt.
+
+intx irq descriptor
+
+
+The irq descriptors are created by the proxy object
+``using event_notifier_init()`` to create the irq and re-sampling
+*eventds*, and ``kvm_vm_ioctl(KVM_IRQFD)`` to bind them to an interrupt.
+The interrupt route can be found with
+``pci_device_route_intx_to_irq()``.
+
+intx routing changes
+
+
+Intx routing can be changed when the guest programs the APIC the device
+pin is connected to. The proxy object in QEMU will use
+``pci_device_set_intx_routing_notifier()`` to be informed of any guest
+changes to the route. This handler will broadly follow the VFIO
+interrupt logic to change the route: de-assigning the existing irq
+descriptor from its route, then assigning it the new route. (see
+``vfio_intx_update()``)
+
+MSI/X acceleration
+''''''''''''''''''
+
+MSI/X interrupts are sent as DMA transactions to the host. The interrupt
+data contains a vector that is programmed by the guest, A device may have
+multiple MSI interrupts associated with it, so multiple irq descriptors
+may need to be sent to the emulation program.
+
+MSI/X irq descriptor
+
+
+This case will also follow the VFIO example. For each MSI/X interrupt,
+an *eventfd* is created, a virtual interrupt is allocated by
+``kvm_irqchip_add_msi_route()``, and the virtual interrupt is bound to
+the eventfd with ``kvm_irqchip_add_irqfd_notifier()``.
+
+MSI/X config space changes
+
+
+The guest may dynamically update several MSI-related tables in the
+device's PCI config space. These include per-MSI interrupt enables and
+vector data. Additionally, MSIX tables exist in device memory space, not
+config space. Much like the BAR case above, the proxy object must look
+at guest config space programming to keep the MSI interrupt state
+consistent between QEMU and the emulation program.
+
+--------------
+
+Disaggregated CPU emulation
+---------------------------
+
+After IO services have been disaggregated, a second phase would be to
+separate a process to handle CPU instruction emulation from the main
+QEMU control function. There are no object separation points for this
+code, so the first task would be to create one.
+
+Host access controls
+--------------------
+
+Separating QEMU relies on the host OS's access restriction mechanisms to
+enforce that the differing processes can only access the objects they
+are entitled to. There are a couple types of mechanisms usually provided
+by general purpose OSs.
+
+Discretionary access control
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Discretionary access control allows each user to control who can access
+their files. In Linux, this type of control is usually too coarse for
+QEMU separation, since it only provides three separate access controls:
+one for the same user ID, the second for users IDs with the same group
+ID, and the third for all other user IDs. Each device instance would
+need a separate user ID to provide access control, which is likely to be
+unwieldy for dynamically created VMs.
+
+Mandatory access control
+~~~~~~~~~~~~~~~~~~~~~~~~
+
+Mandatory access control allows the OS to add an additional set of
+controls on top of discretionary access for the OS to control. It also
+adds other attributes to processes and files such as types, roles, and
+categories, and can establish rules for how processes and files can
+interact.
+
+Type enforcement
+^^^^^^^^^^^^^^^^
+
+Type enforcement assigns a *type* attribute to processes and files, and
+allows rules to be written on what operations a process with a given
+type can perform on a file with a given type. QEMU separation could take
+advantage of type enforcement by running the emulation processes with
+different types, both from the main QEMU process, and from the emulation
+processes of different classes of devices.
+
+For example, guest disk images and disk emulation processes could have
+types separate from the main QEMU process and non-disk emulation
+processes, and the type rules could prevent processes other than disk
+emulation ones from accessing guest disk images. Similarly, network
+emulation processes can have a type separate from the main QEMU process
+and non-network emulation process, and only that type can access the
+host tun/tap device used to provide guest networking.
+
+Category enforcement
+^^^^^^^^^^^^^^^^^^^^
+
+Category enforcement assigns a set of numbers within a given range to
+the process or file. The process is granted access to the file if the
+process's set is a superset of the file's set. This enforcement can be
+used to separate multiple instances of devices in the same class.
+
+For example, if there are multiple disk devices provides to a guest,
+each device emulation process could be provisioned with a separate
+category. The different device emulation processes would not be able to
+access each other's backing disk images.
+
+Alternatively, categories could be used in lieu of the type enforcement
+scheme described above. In this scenario, different categories would be
+used to prevent device emulation processes in different classes from
+accessing resources assigned to other classes.
-- 
2.29.2

From: Elena Ufimtseva <elena.ufimtseva@oracle.com>

Adds documentation explaining the command-line arguments needed
to use multi-process.

[Move orphan docs/multi-process.rst document into docs/system/ and add
it to index.rst to prevent Sphinx "document isn't included in any
toctree" error.
--Stefan]

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 MAINTAINERS                   |  1 +
 docs/system/index.rst         |  1 +
 docs/system/multi-process.rst | 64 +++++++++++++++++++++++++++++++++++
 3 files changed, 66 insertions(+)
 create mode 100644 docs/system/multi-process.rst

diff --git a/MAINTAINERS b/MAINTAINERS
index XXXXXXX..XXXXXXX 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -XXX,XX +XXX,XX @@ M: Jagannathan Raman <jag.raman@oracle.com>
 M: John G Johnson <john.g.johnson@oracle.com>
 S: Maintained
 F: docs/devel/multi-process.rst
+F: docs/system/multi-process.rst
 
 Build and test automation
 -------------------------
diff --git a/docs/system/index.rst b/docs/system/index.rst
index XXXXXXX..XXXXXXX 100644
--- a/docs/system/index.rst
+++ b/docs/system/index.rst
@@ -XXX,XX +XXX,XX @@ Contents:
    pr-manager
    targets
    security
+   multi-process
    deprecated
    removed-features
    build-platforms
diff --git a/docs/system/multi-process.rst b/docs/system/multi-process.rst
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/docs/system/multi-process.rst
@@ -XXX,XX +XXX,XX @@
+Multi-process QEMU
+==================
+
+This document describes how to configure and use multi-process qemu.
+For the design document refer to docs/devel/qemu-multiprocess.
+
+1) Configuration
+----------------
+
+multi-process is enabled by default for targets that enable KVM
+
+
+2) Usage
+--------
+
+Multi-process QEMU requires an orchestrator to launch.
+
+Following is a description of command-line used to launch mpqemu.
+
+* Orchestrator:
+
+  - The Orchestrator creates a unix socketpair
+
+  - It launches the remote process and passes one of the
+    sockets to it via command-line.
+
+  - It then launches QEMU and specifies the other socket as an option
+    to the Proxy device object
+
+* Remote Process:
+
+  - QEMU can enter remote process mode by using the "remote" machine
+    option.
+
+  - The orchestrator creates a "remote-object" with details about
+    the device and the file descriptor for the device
+
+  - The remaining options are no different from how one launches QEMU with
+    devices.
+
+  - Example command-line for the remote process is as follows:
+
+      /usr/bin/qemu-system-x86_64                                        \
+      -machine x-remote                                                  \
+      -device lsi53c895a,id=lsi0                                         \
+      -drive id=drive_image2,file=/build/ol7-nvme-test-1.qcow2           \
+      -device scsi-hd,id=drive2,drive=drive_image2,bus=lsi0.0,scsi-id=0  \
+      -object x-remote-object,id=robj1,devid=lsi1,fd=4,
+
+* QEMU:
+
+  - Since parts of the RAM are shared between QEMU & remote process, a
+    memory-backend-memfd is required to facilitate this, as follows:
+
+    -object memory-backend-memfd,id=mem,size=2G
+
+  - A "x-pci-proxy-dev" device is created for each of the PCI devices emulated
+    in the remote process. A "socket" sub-option specifies the other end of
+    unix channel created by orchestrator. The "id" sub-option must be specified
+    and should be the same as the "id" specified for the remote PCI device
+
+  - Example commandline for QEMU is as follows:
+
+      -device x-pci-proxy-dev,id=lsi0,socket=3
-- 
2.29.2

From: Jagannathan Raman <jag.raman@oracle.com>

Allow RAM MemoryRegion to be created from an offset in a file, instead
of allocating at offset of 0 by default. This is needed to synchronize
RAM between QEMU & remote process.

Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 609996697ad8617e3b01df38accc5c208c24d74e.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/exec/memory.h     |  2 ++
 include/exec/ram_addr.h   |  4 ++--
 include/qemu/mmap-alloc.h |  4 +++-
 backends/hostmem-memfd.c  |  2 +-
 hw/misc/ivshmem.c         |  3 ++-
 softmmu/memory.c          |  3 ++-
 softmmu/physmem.c         | 12 +++++++-----
 util/mmap-alloc.c         |  8 +++++---
 util/oslib-posix.c        |  2 +-
 9 files changed, 25 insertions(+), 15 deletions(-)

diff --git a/include/exec/memory.h b/include/exec/memory.h
index XXXXXXX..XXXXXXX 100644
--- a/include/exec/memory.h
+++ b/include/exec/memory.h
@@ -XXX,XX +XXX,XX @@ void memory_region_init_ram_from_file(MemoryRegion *mr,
  * @size: size of the region.
  * @share: %true if memory must be mmaped with the MAP_SHARED flag
  * @fd: the fd to mmap.
+ * @offset: offset within the file referenced by fd
  * @errp: pointer to Error*, to store an error if it happens.
  *
  * Note that this function does not do anything to cause the data in the
@@ -XXX,XX +XXX,XX @@ void memory_region_init_ram_from_fd(MemoryRegion *mr,
                                     uint64_t size,
                                     bool share,
                                     int fd,
+                                    ram_addr_t offset,
                                     Error **errp);
 #endif
 
diff --git a/include/exec/ram_addr.h b/include/exec/ram_addr.h
index XXXXXXX..XXXXXXX 100644
--- a/include/exec/ram_addr.h
+++ b/include/exec/ram_addr.h
@@ -XXX,XX +XXX,XX @@ RAMBlock *qemu_ram_alloc_from_file(ram_addr_t size, MemoryRegion *mr,
                                    uint32_t ram_flags, const char *mem_path,
                                    bool readonly, Error **errp);
 RAMBlock *qemu_ram_alloc_from_fd(ram_addr_t size, MemoryRegion *mr,
-                                 uint32_t ram_flags, int fd, bool readonly,
-                                 Error **errp);
+                                 uint32_t ram_flags, int fd, off_t offset,
+                                 bool readonly, Error **errp);
 
 RAMBlock *qemu_ram_alloc_from_ptr(ram_addr_t size, void *host,
                                   MemoryRegion *mr, Error **errp);
diff --git a/include/qemu/mmap-alloc.h b/include/qemu/mmap-alloc.h
index XXXXXXX..XXXXXXX 100644
--- a/include/qemu/mmap-alloc.h
+++ b/include/qemu/mmap-alloc.h
@@ -XXX,XX +XXX,XX @@ size_t qemu_mempath_getpagesize(const char *mem_path);
  *  @readonly: true for a read-only mapping, false for read/write.
  *  @shared: map has RAM_SHARED flag.
  *  @is_pmem: map has RAM_PMEM flag.
+ *  @map_offset: map starts at offset of map_offset from the start of fd
  *
  * Return:
  *  On success, return a pointer to the mapped area.
@@ -XXX,XX +XXX,XX @@ void *qemu_ram_mmap(int fd,
                     size_t align,
                     bool readonly,
                     bool shared,
-                    bool is_pmem);
+                    bool is_pmem,
+                    off_t map_offset);
 
 void qemu_ram_munmap(int fd, void *ptr, size_t size);
 
diff --git a/backends/hostmem-memfd.c b/backends/hostmem-memfd.c
index XXXXXXX..XXXXXXX 100644
--- a/backends/hostmem-memfd.c
+++ b/backends/hostmem-memfd.c
@@ -XXX,XX +XXX,XX @@ memfd_backend_memory_alloc(HostMemoryBackend *backend, Error **errp)
     name = host_memory_backend_get_name(backend);
     memory_region_init_ram_from_fd(&backend->mr, OBJECT(backend),
                                    name, backend->size,
-                                   backend->share, fd, errp);
+                                   backend->share, fd, 0, errp);
     g_free(name);
 }
 
diff --git a/hw/misc/ivshmem.c b/hw/misc/ivshmem.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/misc/ivshmem.c
+++ b/hw/misc/ivshmem.c
@@ -XXX,XX +XXX,XX @@ static void process_msg_shmem(IVShmemState *s, int fd, Error **errp)
 
     /* mmap the region and map into the BAR2 */
     memory_region_init_ram_from_fd(&s->server_bar2, OBJECT(s),
-                                   "ivshmem.bar2", size, true, fd, &local_err);
+                                   "ivshmem.bar2", size, true, fd, 0,
+                                   &local_err);
     if (local_err) {
         error_propagate(errp, local_err);
         return;
diff --git a/softmmu/memory.c b/softmmu/memory.c
index XXXXXXX..XXXXXXX 100644
--- a/softmmu/memory.c
+++ b/softmmu/memory.c
@@ -XXX,XX +XXX,XX @@ void memory_region_init_ram_from_fd(MemoryRegion *mr,
                                     uint64_t size,
                                     bool share,
                                     int fd,
+                                    ram_addr_t offset,
                                     Error **errp)
 {
     Error *err = NULL;
@@ -XXX,XX +XXX,XX @@ void memory_region_init_ram_from_fd(MemoryRegion *mr,
     mr->destructor = memory_region_destructor_ram;
     mr->ram_block = qemu_ram_alloc_from_fd(size, mr,
                                            share ? RAM_SHARED : 0,
-                                           fd, false, &err);
+                                           fd, offset, false, &err);
     if (err) {
         mr->size = int128_zero();
         object_unparent(OBJECT(mr));
diff --git a/softmmu/physmem.c b/softmmu/physmem.c
index XXXXXXX..XXXXXXX 100644
--- a/softmmu/physmem.c
+++ b/softmmu/physmem.c
@@ -XXX,XX +XXX,XX @@ static void *file_ram_alloc(RAMBlock *block,
                             int fd,
                             bool readonly,
                             bool truncate,
+                            off_t offset,
                             Error **errp)
 {
     void *area;
@@ -XXX,XX +XXX,XX @@ static void *file_ram_alloc(RAMBlock *block,
     }
 
     area = qemu_ram_mmap(fd, memory, block->mr->align, readonly,
-                         block->flags & RAM_SHARED, block->flags & RAM_PMEM);
+                         block->flags & RAM_SHARED, block->flags & RAM_PMEM,
+                         offset);
     if (area == MAP_FAILED) {
         error_setg_errno(errp, errno,
                          "unable to map backing store for guest RAM");
@@ -XXX,XX +XXX,XX @@ static void ram_block_add(RAMBlock *new_block, Error **errp, bool shared)
 
 #ifdef CONFIG_POSIX
 RAMBlock *qemu_ram_alloc_from_fd(ram_addr_t size, MemoryRegion *mr,
-                                 uint32_t ram_flags, int fd, bool readonly,
-                                 Error **errp)
+                                 uint32_t ram_flags, int fd, off_t offset,
+                                 bool readonly, Error **errp)
 {
     RAMBlock *new_block;
     Error *local_err = NULL;
@@ -XXX,XX +XXX,XX @@ RAMBlock *qemu_ram_alloc_from_fd(ram_addr_t size, MemoryRegion *mr,
     new_block->max_length = size;
     new_block->flags = ram_flags;
     new_block->host = file_ram_alloc(new_block, size, fd, readonly,
-                                     !file_size, errp);
+                                     !file_size, offset, errp);
     if (!new_block->host) {
         g_free(new_block);
         return NULL;
@@ -XXX,XX +XXX,XX @@ RAMBlock *qemu_ram_alloc_from_file(ram_addr_t size, MemoryRegion *mr,
         return NULL;
     }
 
-    block = qemu_ram_alloc_from_fd(size, mr, ram_flags, fd, readonly, errp);
+    block = qemu_ram_alloc_from_fd(size, mr, ram_flags, fd, 0, readonly, errp);
     if (!block) {
         if (created) {
             unlink(mem_path);
diff --git a/util/mmap-alloc.c b/util/mmap-alloc.c
index XXXXXXX..XXXXXXX 100644
--- a/util/mmap-alloc.c
+++ b/util/mmap-alloc.c
@@ -XXX,XX +XXX,XX @@ void *qemu_ram_mmap(int fd,
                     size_t align,
                     bool readonly,
                     bool shared,
-                    bool is_pmem)
+                    bool is_pmem,
+                    off_t map_offset)
 {
     int prot;
     int flags;
@@ -XXX,XX +XXX,XX @@ void *qemu_ram_mmap(int fd,
 
     prot = PROT_READ | (readonly ? 0 : PROT_WRITE);
 
-    ptr = mmap(guardptr + offset, size, prot, flags | map_sync_flags, fd, 0);
+    ptr = mmap(guardptr + offset, size, prot,
+               flags | map_sync_flags, fd, map_offset);
 
     if (ptr == MAP_FAILED && map_sync_flags) {
         if (errno == ENOTSUP) {
@@ -XXX,XX +XXX,XX @@ void *qemu_ram_mmap(int fd,
          * if map failed with MAP_SHARED_VALIDATE | MAP_SYNC,
          * we will remove these flags to handle compatibility.
          */
-        ptr = mmap(guardptr + offset, size, prot, flags, fd, 0);
+        ptr = mmap(guardptr + offset, size, prot, flags, fd, map_offset);
     }
 
     if (ptr == MAP_FAILED) {
diff --git a/util/oslib-posix.c b/util/oslib-posix.c
index XXXXXXX..XXXXXXX 100644
--- a/util/oslib-posix.c
+++ b/util/oslib-posix.c
@@ -XXX,XX +XXX,XX @@ void *qemu_memalign(size_t alignment, size_t size)
 void *qemu_anon_ram_alloc(size_t size, uint64_t *alignment, bool shared)
 {
     size_t align = QEMU_VMALLOC_ALIGN;
-    void *ptr = qemu_ram_mmap(-1, size, align, false, shared, false);
+    void *ptr = qemu_ram_mmap(-1, size, align, false, shared, false, 0);
 
     if (ptr == MAP_FAILED) {
         return NULL;
-- 
2.29.2

From: Jagannathan Raman <jag.raman@oracle.com>

Add configuration options to enable or disable multiprocess QEMU code

diff --git a/configure b/configure
index XXXXXXX..XXXXXXX 100755
--- a/configure
+++ b/configure
@@ -XXX,XX +XXX,XX @@ skip_meson=no
 gettext="auto"
 fuse="auto"
 fuse_lseek="auto"
+multiprocess="no"
 
 malloc_trim="auto"
 
@@ -XXX,XX +XXX,XX @@ Linux)
   linux="yes"
   linux_user="yes"
   vhost_user=${default_feature:-yes}
+  multiprocess=${default_feature:-yes}
 ;;
 esac
 
@@ -XXX,XX +XXX,XX @@ for opt do
   ;;
   --disable-fuse-lseek) fuse_lseek="disabled"
   ;;
+  --enable-multiprocess) multiprocess="yes"
+  ;;
+  --disable-multiprocess) multiprocess="no"
+  ;;
   *)
       echo "ERROR: unknown option $opt"
       echo "Try '$0 --help' for more information"
@@ -XXX,XX +XXX,XX @@ disabled with --disable-FEATURE, default is enabled if available
   libdaxctl       libdaxctl support
   fuse            FUSE block device export
   fuse-lseek      SEEK_HOLE/SEEK_DATA support for FUSE exports
+  multiprocess    Multiprocess QEMU support
 
 NOTE: The object files are built at the place where configure is launched
 EOF
@@ -XXX,XX +XXX,XX @@ fi
 if test "$have_mlockall" = "yes" ; then
   echo "HAVE_MLOCKALL=y" >> $config_host_mak
 fi
+if test "$multiprocess" = "yes" ; then
+  echo "CONFIG_MULTIPROCESS_ALLOWED=y" >> $config_host_mak
+fi
 if test "$fuzzing" = "yes" ; then
   # If LIB_FUZZING_ENGINE is set, assume we are running on OSS-Fuzz, and the
   # needed CFLAGS have already been provided
diff --git a/meson.build b/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/meson.build
+++ b/meson.build
@@ -XXX,XX +XXX,XX @@ host_kconfig = \
   ('CONFIG_VHOST_KERNEL' in config_host ? ['CONFIG_VHOST_KERNEL=y'] : []) + \
   (have_virtfs ? ['CONFIG_VIRTFS=y'] : []) + \
   ('CONFIG_LINUX' in config_host ? ['CONFIG_LINUX=y'] : []) + \
-  ('CONFIG_PVRDMA' in config_host ? ['CONFIG_PVRDMA=y'] : [])
+  ('CONFIG_PVRDMA' in config_host ? ['CONFIG_PVRDMA=y'] : []) + \
+  ('CONFIG_MULTIPROCESS_ALLOWED' in config_host ? ['CONFIG_MULTIPROCESS_ALLOWED=y'] : [])
 
 ignored = [ 'TARGET_XML_FILES', 'TARGET_ABI_DIR', 'TARGET_ARCH' ]
 
@@ -XXX,XX +XXX,XX @@ summary_info += {'libpmem support':   config_host.has_key('CONFIG_LIBPMEM')}
 summary_info += {'libdaxctl support': config_host.has_key('CONFIG_LIBDAXCTL')}
 summary_info += {'libudev':           libudev.found()}
 summary_info += {'FUSE lseek':        fuse_lseek.found()}
+summary_info += {'Multiprocess QEMU': config_host.has_key('CONFIG_MULTIPROCESS_ALLOWED')}
 summary(summary_info, bool_yn: true, section: 'Dependencies')
 
 if not supported_cpus.contains(cpu)
diff --git a/Kconfig.host b/Kconfig.host
index XXXXXXX..XXXXXXX 100644
--- a/Kconfig.host
+++ b/Kconfig.host
@@ -XXX,XX +XXX,XX @@ config VIRTFS
 
 config PVRDMA
     bool
+
+config MULTIPROCESS_ALLOWED
+    bool
+    imply MULTIPROCESS
diff --git a/hw/Kconfig b/hw/Kconfig
index XXXXXXX..XXXXXXX 100644
--- a/hw/Kconfig
+++ b/hw/Kconfig
@@ -XXX,XX +XXX,XX @@ source pci-host/Kconfig
 source pcmcia/Kconfig
 source pci/Kconfig
 source rdma/Kconfig
+source remote/Kconfig
 source rtc/Kconfig
 source scsi/Kconfig
 source sd/Kconfig
diff --git a/hw/remote/Kconfig b/hw/remote/Kconfig
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/remote/Kconfig
@@ -XXX,XX +XXX,XX @@
+config MULTIPROCESS
+    bool
+    depends on PCI && KVM
-- 
2.29.2

From: Jagannathan Raman <jag.raman@oracle.com>

PCI host bridge is setup for the remote device process. It is
implemented using remote-pcihost object. It is an extension of the PCI
host bridge setup by QEMU.
Remote-pcihost configures a PCI bus which could be used by the remote
PCI device to latch on to.

[Added PCI_EXPRESS condition in hw/remote/Kconfig since remote-pcihost
needs PCIe. This solves "make check" failure on s390x. Fix suggested by
Philippe Mathieu-Daudé <philmd@redhat.com> and Thomas Huth
<thuth@redhat.com>.
--Stefan]

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 MAINTAINERS                  |  2 +
 include/hw/pci-host/remote.h | 29 ++++++++++++++
 hw/pci-host/remote.c         | 75 ++++++++++++++++++++++++++++++++++++
 hw/pci-host/Kconfig          |  3 ++
 hw/pci-host/meson.build      |  1 +
 hw/remote/Kconfig            |  3 +-
 6 files changed, 112 insertions(+), 1 deletion(-)
 create mode 100644 include/hw/pci-host/remote.h
 create mode 100644 hw/pci-host/remote.c

diff --git a/MAINTAINERS b/MAINTAINERS
index XXXXXXX..XXXXXXX 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -XXX,XX +XXX,XX @@ M: John G Johnson <john.g.johnson@oracle.com>
 S: Maintained
 F: docs/devel/multi-process.rst
 F: docs/system/multi-process.rst
+F: hw/pci-host/remote.c
+F: include/hw/pci-host/remote.h
 
 Build and test automation
 -------------------------
diff --git a/include/hw/pci-host/remote.h b/include/hw/pci-host/remote.h
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/include/hw/pci-host/remote.h
@@ -XXX,XX +XXX,XX @@
+/*
+ * PCI Host for remote device
+ *
+ * Copyright © 2018, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef REMOTE_PCIHOST_H
+#define REMOTE_PCIHOST_H
+
+#include "exec/memory.h"
+#include "hw/pci/pcie_host.h"
+
+#define TYPE_REMOTE_PCIHOST "remote-pcihost"
+OBJECT_DECLARE_SIMPLE_TYPE(RemotePCIHost, REMOTE_PCIHOST)
+
+struct RemotePCIHost {
+    /*< private >*/
+    PCIExpressHost parent_obj;
+    /*< public >*/
+
+    MemoryRegion *mr_pci_mem;
+    MemoryRegion *mr_sys_io;
+};
+
+#endif
diff --git a/hw/pci-host/remote.c b/hw/pci-host/remote.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/pci-host/remote.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * Remote PCI host device
+ *
+ * Unlike PCI host devices that model physical hardware, the purpose
+ * of this PCI host is to host multi-process QEMU devices.
+ *
+ * Multi-process QEMU extends the PCI host of a QEMU machine into a
+ * remote process. Any PCI device attached to the remote process is
+ * visible in the QEMU guest. This allows existing QEMU device models
+ * to be reused in the remote process.
+ *
+ * This PCI host is purely a container for PCI devices. It's fake in the
+ * sense that the guest never sees this PCI host and has no way of
+ * accessing it. Its job is just to provide the environment that QEMU
+ * PCI device models need when running in a remote process.
+ *
+ * Copyright © 2018, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+
+#include "hw/pci/pci.h"
+#include "hw/pci/pci_host.h"
+#include "hw/pci/pcie_host.h"
+#include "hw/qdev-properties.h"
+#include "hw/pci-host/remote.h"
+#include "exec/memory.h"
+
+static const char *remote_pcihost_root_bus_path(PCIHostState *host_bridge,
+                                                PCIBus *rootbus)
+{
+    return "0000:00";
+}
+
+static void remote_pcihost_realize(DeviceState *dev, Error **errp)
+{
+    PCIHostState *pci = PCI_HOST_BRIDGE(dev);
+    RemotePCIHost *s = REMOTE_PCIHOST(dev);
+
+    pci->bus = pci_root_bus_new(DEVICE(s), "remote-pci",
+                                s->mr_pci_mem, s->mr_sys_io,
+                                0, TYPE_PCIE_BUS);
+}
+
+static void remote_pcihost_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+    PCIHostBridgeClass *hc = PCI_HOST_BRIDGE_CLASS(klass);
+
+    hc->root_bus_path = remote_pcihost_root_bus_path;
+    dc->realize = remote_pcihost_realize;
+
+    dc->user_creatable = false;
+    set_bit(DEVICE_CATEGORY_BRIDGE, dc->categories);
+    dc->fw_name = "pci";
+}
+
+static const TypeInfo remote_pcihost_info = {
+    .name = TYPE_REMOTE_PCIHOST,
+    .parent = TYPE_PCIE_HOST_BRIDGE,
+    .instance_size = sizeof(RemotePCIHost),
+    .class_init = remote_pcihost_class_init,
+};
+
+static void remote_pcihost_register(void)
+{
+    type_register_static(&remote_pcihost_info);
+}
+
+type_init(remote_pcihost_register)
diff --git a/hw/pci-host/Kconfig b/hw/pci-host/Kconfig
index XXXXXXX..XXXXXXX 100644
--- a/hw/pci-host/Kconfig
+++ b/hw/pci-host/Kconfig
@@ -XXX,XX +XXX,XX @@ config PCI_POWERNV
     select PCI_EXPRESS
     select MSI_NONBROKEN
     select PCIE_PORT
+
+config REMOTE_PCIHOST
+    bool
diff --git a/hw/pci-host/meson.build b/hw/pci-host/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/hw/pci-host/meson.build
+++ b/hw/pci-host/meson.build
@@ -XXX,XX +XXX,XX @@ pci_ss.add(when: 'CONFIG_PCI_EXPRESS_XILINX', if_true: files('xilinx-pcie.c'))
 pci_ss.add(when: 'CONFIG_PCI_I440FX', if_true: files('i440fx.c'))
 pci_ss.add(when: 'CONFIG_PCI_SABRE', if_true: files('sabre.c'))
 pci_ss.add(when: 'CONFIG_XEN_IGD_PASSTHROUGH', if_true: files('xen_igd_pt.c'))
+pci_ss.add(when: 'CONFIG_REMOTE_PCIHOST', if_true: files('remote.c'))
 
 # PPC devices
 pci_ss.add(when: 'CONFIG_PREP_PCI', if_true: files('prep.c'))
diff --git a/hw/remote/Kconfig b/hw/remote/Kconfig
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/Kconfig
+++ b/hw/remote/Kconfig
@@ -XXX,XX +XXX,XX @@
 config MULTIPROCESS
     bool
-    depends on PCI && KVM
+    depends on PCI && PCI_EXPRESS && KVM
+    select REMOTE_PCIHOST
-- 
2.29.2

From: Jagannathan Raman <jag.raman@oracle.com>

x-remote-machine object sets up various subsystems of the remote
device process. Instantiate PCI host bridge object and initialize RAM, IO &
PCI memory regions.

Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: c537f38d17f90453ca610c6b70cf3480274e0ba1.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 MAINTAINERS                  |  2 ++
 include/hw/pci-host/remote.h |  1 +
 include/hw/remote/machine.h  | 27 ++++++++++++++
 hw/remote/machine.c          | 70 ++++++++++++++++++++++++++++++++++++
 hw/meson.build               |  1 +
 hw/remote/meson.build        |  5 +++
 6 files changed, 106 insertions(+)
 create mode 100644 include/hw/remote/machine.h
 create mode 100644 hw/remote/machine.c
 create mode 100644 hw/remote/meson.build

diff --git a/MAINTAINERS b/MAINTAINERS
index XXXXXXX..XXXXXXX 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -XXX,XX +XXX,XX @@ F: docs/devel/multi-process.rst
 F: docs/system/multi-process.rst
 F: hw/pci-host/remote.c
 F: include/hw/pci-host/remote.h
+F: hw/remote/machine.c
+F: include/hw/remote/machine.h
 
 Build and test automation
 -------------------------
diff --git a/include/hw/pci-host/remote.h b/include/hw/pci-host/remote.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/pci-host/remote.h
+++ b/include/hw/pci-host/remote.h
@@ -XXX,XX +XXX,XX @@ struct RemotePCIHost {
 
     MemoryRegion *mr_pci_mem;
     MemoryRegion *mr_sys_io;
+    MemoryRegion *mr_sys_mem;
 };
 
 #endif
diff --git a/include/hw/remote/machine.h b/include/hw/remote/machine.h
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/include/hw/remote/machine.h
@@ -XXX,XX +XXX,XX @@
+/*
+ * Remote machine configuration
+ *
+ * Copyright © 2018, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef REMOTE_MACHINE_H
+#define REMOTE_MACHINE_H
+
+#include "qom/object.h"
+#include "hw/boards.h"
+#include "hw/pci-host/remote.h"
+
+struct RemoteMachineState {
+    MachineState parent_obj;
+
+    RemotePCIHost *host;
+};
+
+#define TYPE_REMOTE_MACHINE "x-remote-machine"
+OBJECT_DECLARE_SIMPLE_TYPE(RemoteMachineState, REMOTE_MACHINE)
+
+#endif
diff --git a/hw/remote/machine.c b/hw/remote/machine.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/remote/machine.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * Machine for remote device
+ *
+ *  This machine type is used by the remote device process in multi-process
+ *  QEMU. QEMU device models depend on parent busses, interrupt controllers,
+ *  memory regions, etc. The remote machine type offers this environment so
+ *  that QEMU device models can be used as remote devices.
+ *
+ * Copyright © 2018, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+
+#include "hw/remote/machine.h"
+#include "exec/address-spaces.h"
+#include "exec/memory.h"
+#include "qapi/error.h"
+
+static void remote_machine_init(MachineState *machine)
+{
+    MemoryRegion *system_memory, *system_io, *pci_memory;
+    RemoteMachineState *s = REMOTE_MACHINE(machine);
+    RemotePCIHost *rem_host;
+
+    system_memory = get_system_memory();
+    system_io = get_system_io();
+
+    pci_memory = g_new(MemoryRegion, 1);
+    memory_region_init(pci_memory, NULL, "pci", UINT64_MAX);
+
+    rem_host = REMOTE_PCIHOST(qdev_new(TYPE_REMOTE_PCIHOST));
+
+    rem_host->mr_pci_mem = pci_memory;
+    rem_host->mr_sys_mem = system_memory;
+    rem_host->mr_sys_io = system_io;
+
+    s->host = rem_host;
+
+    object_property_add_child(OBJECT(s), "remote-pcihost", OBJECT(rem_host));
+    memory_region_add_subregion_overlap(system_memory, 0x0, pci_memory, -1);
+
+    qdev_realize(DEVICE(rem_host), sysbus_get_default(), &error_fatal);
+}
+
+static void remote_machine_class_init(ObjectClass *oc, void *data)
+{
+    MachineClass *mc = MACHINE_CLASS(oc);
+
+    mc->init = remote_machine_init;
+    mc->desc = "Experimental remote machine";
+}
+
+static const TypeInfo remote_machine = {
+    .name = TYPE_REMOTE_MACHINE,
+    .parent = TYPE_MACHINE,
+    .instance_size = sizeof(RemoteMachineState),
+    .class_init = remote_machine_class_init,
+};
+
+static void remote_machine_register_types(void)
+{
+    type_register_static(&remote_machine);
+}
+
+type_init(remote_machine_register_types);
diff --git a/hw/meson.build b/hw/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/hw/meson.build
+++ b/hw/meson.build
@@ -XXX,XX +XXX,XX @@ subdir('moxie')
 subdir('nios2')
 subdir('openrisc')
 subdir('ppc')
+subdir('remote')
 subdir('riscv')
 subdir('rx')
 subdir('s390x')
diff --git a/hw/remote/meson.build b/hw/remote/meson.build
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/remote/meson.build
@@ -XXX,XX +XXX,XX @@
+remote_ss = ss.source_set()
+
+remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('machine.c'))
+
+softmmu_ss.add_all(when: 'CONFIG_MULTIPROCESS', if_true: remote_ss)
-- 
2.29.2

From: Elena Ufimtseva <elena.ufimtseva@oracle.com>

Adds qio_channel_writev_full_all() to transmit both data and FDs.
Refactors existing code to use this helper.

Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Daniel P. Berrangé <berrange@redhat.com>
Message-id: 480fbf1fe4152495d60596c9b665124549b426a5.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/io/channel.h | 25 +++++++++++++++++++++++++
 io/channel.c         | 15 ++++++++++++++-
 2 files changed, 39 insertions(+), 1 deletion(-)

From: Elena Ufimtseva <elena.ufimtseva@oracle.com>

Adds qio_channel_readv_full_all_eof() and qio_channel_readv_full_all()
to read both data and FDs. Refactors existing code to use these helpers.

Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Acked-by: Daniel P. Berrangé <berrange@redhat.com>
Message-id: b059c4cc0fb741e794d644c144cc21372cad877d.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/io/channel.h |  53 +++++++++++++++++++++++
 io/channel.c         | 101 ++++++++++++++++++++++++++++++++++---------
 2 files changed, 134 insertions(+), 20 deletions(-)

diff --git a/include/io/channel.h b/include/io/channel.h
index XXXXXXX..XXXXXXX 100644
--- a/include/io/channel.h
+++ b/include/io/channel.h
@@ -XXX,XX +XXX,XX @@ void qio_channel_set_aio_fd_handler(QIOChannel *ioc,
                                     IOHandler *io_write,
                                     void *opaque);
 
+/**
+ * qio_channel_readv_full_all_eof:
+ * @ioc: the channel object
+ * @iov: the array of memory regions to read data to
+ * @niov: the length of the @iov array
+ * @fds: an array of file handles to read
+ * @nfds: number of file handles in @fds
+ * @errp: pointer to a NULL-initialized error object
+ *
+ *
+ * Performs same function as qio_channel_readv_all_eof.
+ * Additionally, attempts to read file descriptors shared
+ * over the channel. The function will wait for all
+ * requested data to be read, yielding from the current
+ * coroutine if required. data refers to both file
+ * descriptors and the iovs.
+ *
+ * Returns: 1 if all bytes were read, 0 if end-of-file
+ *          occurs without data, or -1 on error
+ */
+
+int qio_channel_readv_full_all_eof(QIOChannel *ioc,
+                                   const struct iovec *iov,
+                                   size_t niov,
+                                   int **fds, size_t *nfds,
+                                   Error **errp);
+
+/**
+ * qio_channel_readv_full_all:
+ * @ioc: the channel object
+ * @iov: the array of memory regions to read data to
+ * @niov: the length of the @iov array
+ * @fds: an array of file handles to read
+ * @nfds: number of file handles in @fds
+ * @errp: pointer to a NULL-initialized error object
+ *
+ *
+ * Performs same function as qio_channel_readv_all_eof.
+ * Additionally, attempts to read file descriptors shared
+ * over the channel. The function will wait for all
+ * requested data to be read, yielding from the current
+ * coroutine if required. data refers to both file
+ * descriptors and the iovs.
+ *
+ * Returns: 0 if all bytes were read, or -1 on error
+ */
+
+int qio_channel_readv_full_all(QIOChannel *ioc,
+                               const struct iovec *iov,
+                               size_t niov,
+                               int **fds, size_t *nfds,
+                               Error **errp);
+
 /**
  * qio_channel_writev_full_all:
  * @ioc: the channel object
diff --git a/io/channel.c b/io/channel.c
index XXXXXXX..XXXXXXX 100644
--- a/io/channel.c
+++ b/io/channel.c
@@ -XXX,XX +XXX,XX @@ int qio_channel_readv_all_eof(QIOChannel *ioc,
                               const struct iovec *iov,
                               size_t niov,
                               Error **errp)
+{
+    return qio_channel_readv_full_all_eof(ioc, iov, niov, NULL, NULL, errp);
+}
+
+int qio_channel_readv_all(QIOChannel *ioc,
+                          const struct iovec *iov,
+                          size_t niov,
+                          Error **errp)
+{
+    return qio_channel_readv_full_all(ioc, iov, niov, NULL, NULL, errp);
+}
+
+int qio_channel_readv_full_all_eof(QIOChannel *ioc,
+                                   const struct iovec *iov,
+                                   size_t niov,
+                                   int **fds, size_t *nfds,
+                                   Error **errp)
 {
     int ret = -1;
     struct iovec *local_iov = g_new(struct iovec, niov);
     struct iovec *local_iov_head = local_iov;
     unsigned int nlocal_iov = niov;
+    int **local_fds = fds;
+    size_t *local_nfds = nfds;
     bool partial = false;
 
+    if (nfds) {
+        *nfds = 0;
+    }
+
+    if (fds) {
+        *fds = NULL;
+    }
+
     nlocal_iov = iov_copy(local_iov, nlocal_iov,
                           iov, niov,
                           0, iov_size(iov, niov));
 
-    while (nlocal_iov > 0) {
+    while ((nlocal_iov > 0) || local_fds) {
         ssize_t len;
-        len = qio_channel_readv(ioc, local_iov, nlocal_iov, errp);
+        len = qio_channel_readv_full(ioc, local_iov, nlocal_iov, local_fds,
+                                     local_nfds, errp);
         if (len == QIO_CHANNEL_ERR_BLOCK) {
             if (qemu_in_coroutine()) {
                 qio_channel_yield(ioc, G_IO_IN);
@@ -XXX,XX +XXX,XX @@ int qio_channel_readv_all_eof(QIOChannel *ioc,
                 qio_channel_wait(ioc, G_IO_IN);
             }
             continue;
-        } else if (len < 0) {
-            goto cleanup;
-        } else if (len == 0) {
-            if (partial) {
-                error_setg(errp,
-                           "Unexpected end-of-file before all bytes were read");
-            } else {
+        }
+
+        if (len == 0) {
+            if (local_nfds && *local_nfds) {
+                /*
+                 * Got some FDs, but no data yet. This isn't an EOF
+                 * scenario (yet), so carry on to try to read data
+                 * on next loop iteration
+                 */
+                goto next_iter;
+            } else if (!partial) {
+                /* No fds and no data - EOF before any data read */
                 ret = 0;
+                goto cleanup;
+            } else {
+                len = -1;
+                error_setg(errp,
+                           "Unexpected end-of-file before all data were read");
+                /* Fallthrough into len < 0 handling */
+            }
+        }
+
+        if (len < 0) {
+            /* Close any FDs we previously received */
+            if (nfds && fds) {
+                size_t i;
+                for (i = 0; i < (*nfds); i++) {
+                    close((*fds)[i]);
+                }
+                g_free(*fds);
+                *fds = NULL;
+                *nfds = 0;
             }
             goto cleanup;
         }
 
+        if (nlocal_iov) {
+            iov_discard_front(&local_iov, &nlocal_iov, len);
+        }
+
+next_iter:
         partial = true;
-        iov_discard_front(&local_iov, &nlocal_iov, len);
+        local_fds = NULL;
+        local_nfds = NULL;
     }
 
     ret = 1;
@@ -XXX,XX +XXX,XX @@ int qio_channel_readv_all_eof(QIOChannel *ioc,
     return ret;
 }
 
-int qio_channel_readv_all(QIOChannel *ioc,
-                          const struct iovec *iov,
-                          size_t niov,
-                          Error **errp)
+int qio_channel_readv_full_all(QIOChannel *ioc,
+                               const struct iovec *iov,
+                               size_t niov,
+                               int **fds, size_t *nfds,
+                               Error **errp)
 {
-    int ret = qio_channel_readv_all_eof(ioc, iov, niov, errp);
+    int ret = qio_channel_readv_full_all_eof(ioc, iov, niov, fds, nfds, errp);
 
     if (ret == 0) {
-        ret = -1;
-        error_setg(errp,
-                   "Unexpected end-of-file before all bytes were read");
-    } else if (ret == 1) {
-        ret = 0;
+        error_prepend(errp,
+                      "Unexpected end-of-file before all data were read.");
+        return -1;
     }
+    if (ret == 1) {
+        return 0;
+    }
+
     return ret;
 }
 
-- 
2.29.2

From: Elena Ufimtseva <elena.ufimtseva@oracle.com>

Defines MPQemuMsg, which is the message that is sent to the remote
process. This message is sent over QIOChannel and is used to
command the remote process to perform various tasks.
Define transmission functions used by proxy and by remote.

[Replace struct iovec send[2] = {0} with {} to make clang happy as
suggested by Peter Maydell <peter.maydell@linaro.org>.
--Stefan]

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 MAINTAINERS                     |   2 +
 meson.build                     |   1 +
 hw/remote/trace.h               |   1 +
 include/hw/remote/mpqemu-link.h |  63 ++++++++++
 include/sysemu/iothread.h       |   6 +
 hw/remote/mpqemu-link.c         | 205 ++++++++++++++++++++++++++++++++
 iothread.c                      |   6 +
 hw/remote/meson.build           |   1 +
 hw/remote/trace-events          |   4 +
 9 files changed, 289 insertions(+)
 create mode 100644 hw/remote/trace.h
 create mode 100644 include/hw/remote/mpqemu-link.h
 create mode 100644 hw/remote/mpqemu-link.c
 create mode 100644 hw/remote/trace-events

diff --git a/MAINTAINERS b/MAINTAINERS
index XXXXXXX..XXXXXXX 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -XXX,XX +XXX,XX @@ F: hw/pci-host/remote.c
 F: include/hw/pci-host/remote.h
 F: hw/remote/machine.c
 F: include/hw/remote/machine.h
+F: hw/remote/mpqemu-link.c
+F: include/hw/remote/mpqemu-link.h
 
 Build and test automation
 -------------------------
diff --git a/meson.build b/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/meson.build
+++ b/meson.build
@@ -XXX,XX +XXX,XX @@ if have_system
     'net',
     'softmmu',
     'ui',
+    'hw/remote',
   ]
 endif
 if have_system or have_user
diff --git a/hw/remote/trace.h b/hw/remote/trace.h
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/remote/trace.h
@@ -0,0 +1 @@
+#include "trace/trace-hw_remote.h"
diff --git a/include/hw/remote/mpqemu-link.h b/include/hw/remote/mpqemu-link.h
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/include/hw/remote/mpqemu-link.h
@@ -XXX,XX +XXX,XX @@
+/*
+ * Communication channel between QEMU and remote device process
+ *
+ * Copyright © 2018, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef MPQEMU_LINK_H
+#define MPQEMU_LINK_H
+
+#include "qom/object.h"
+#include "qemu/thread.h"
+#include "io/channel.h"
+
+#define REMOTE_MAX_FDS 8
+
+#define MPQEMU_MSG_HDR_SIZE offsetof(MPQemuMsg, data.u64)
+
+/**
+ * MPQemuCmd:
+ *
+ * MPQemuCmd enum type to specify the command to be executed on the remote
+ * device.
+ *
+ * This uses a private protocol between QEMU and the remote process. vfio-user
+ * protocol would supersede this in the future.
+ *
+ */
+typedef enum {
+    MPQEMU_CMD_MAX,
+} MPQemuCmd;
+
+/**
+ * MPQemuMsg:
+ * @cmd: The remote command
+ * @size: Size of the data to be shared
+ * @data: Structured data
+ * @fds: File descriptors to be shared with remote device
+ *
+ * MPQemuMsg Format of the message sent to the remote device from QEMU.
+ *
+ */
+typedef struct {
+    int cmd;
+    size_t size;
+
+    union {
+        uint64_t u64;
+    } data;
+
+    int fds[REMOTE_MAX_FDS];
+    int num_fds;
+} MPQemuMsg;
+
+bool mpqemu_msg_send(MPQemuMsg *msg, QIOChannel *ioc, Error **errp);
+bool mpqemu_msg_recv(MPQemuMsg *msg, QIOChannel *ioc, Error **errp);
+
+bool mpqemu_msg_valid(MPQemuMsg *msg);
+
+#endif
diff --git a/include/sysemu/iothread.h b/include/sysemu/iothread.h
index XXXXXXX..XXXXXXX 100644
--- a/include/sysemu/iothread.h
+++ b/include/sysemu/iothread.h
@@ -XXX,XX +XXX,XX @@ IOThread *iothread_create(const char *id, Error **errp);
 void iothread_stop(IOThread *iothread);
 void iothread_destroy(IOThread *iothread);
 
+/*
+ * Returns true if executing withing IOThread context,
+ * false otherwise.
+ */
+bool qemu_in_iothread(void);
+
 #endif /* IOTHREAD_H */
diff --git a/hw/remote/mpqemu-link.c b/hw/remote/mpqemu-link.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/remote/mpqemu-link.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * Communication channel between QEMU and remote device process
+ *
+ * Copyright © 2018, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+
+#include "qemu/module.h"
+#include "hw/remote/mpqemu-link.h"
+#include "qapi/error.h"
+#include "qemu/iov.h"
+#include "qemu/error-report.h"
+#include "qemu/main-loop.h"
+#include "io/channel.h"
+#include "sysemu/iothread.h"
+#include "trace.h"
+
+/*
+ * Send message over the ioc QIOChannel.
+ * This function is safe to call from:
+ * - main loop in co-routine context. Will block the main loop if not in
+ *   co-routine context;
+ * - vCPU thread with no co-routine context and if the channel is not part
+ *   of the main loop handling;
+ * - IOThread within co-routine context, outside of co-routine context
+ *   will block IOThread;
+ * Returns true if no errors were encountered, false otherwise.
+ */
+bool mpqemu_msg_send(MPQemuMsg *msg, QIOChannel *ioc, Error **errp)
+{
+    ERRP_GUARD();
+    bool iolock = qemu_mutex_iothread_locked();
+    bool iothread = qemu_in_iothread();
+    struct iovec send[2] = {};
+    int *fds = NULL;
+    size_t nfds = 0;
+    bool ret = false;
+
+    send[0].iov_base = msg;
+    send[0].iov_len = MPQEMU_MSG_HDR_SIZE;
+
+    send[1].iov_base = (void *)&msg->data;
+    send[1].iov_len = msg->size;
+
+    if (msg->num_fds) {
+        nfds = msg->num_fds;
+        fds = msg->fds;
+    }
+
+    /*
+     * Dont use in IOThread out of co-routine context as
+     * it will block IOThread.
+     */
+    assert(qemu_in_coroutine() || !iothread);
+
+    /*
+     * Skip unlocking/locking iothread lock when the IOThread is running
+     * in co-routine context. Co-routine context is asserted above
+     * for IOThread case.
+     * Also skip lock handling while in a co-routine in the main context.
+     */
+    if (iolock && !iothread && !qemu_in_coroutine()) {
+        qemu_mutex_unlock_iothread();
+    }
+
+    if (!qio_channel_writev_full_all(ioc, send, G_N_ELEMENTS(send),
+                                    fds, nfds, errp)) {
+        ret = true;
+    } else {
+        trace_mpqemu_send_io_error(msg->cmd, msg->size, nfds);
+    }
+
+    if (iolock && !iothread && !qemu_in_coroutine()) {
+        /* See above comment why skip locking here. */
+        qemu_mutex_lock_iothread();
+    }
+
+    return ret;
+}
+
+/*
+ * Read message from the ioc QIOChannel.
+ * This function is safe to call from:
+ * - From main loop in co-routine context. Will block the main loop if not in
+ *   co-routine context;
+ * - From vCPU thread with no co-routine context and if the channel is not part
+ *   of the main loop handling;
+ * - From IOThread within co-routine context, outside of co-routine context
+ *   will block IOThread;
+ */
+static ssize_t mpqemu_read(QIOChannel *ioc, void *buf, size_t len, int **fds,
+                           size_t *nfds, Error **errp)
+{
+    ERRP_GUARD();
+    struct iovec iov = { .iov_base = buf, .iov_len = len };
+    bool iolock = qemu_mutex_iothread_locked();
+    bool iothread = qemu_in_iothread();
+    int ret = -1;
+
+    /*
+     * Dont use in IOThread out of co-routine context as
+     * it will block IOThread.
+     */
+    assert(qemu_in_coroutine() || !iothread);
+
+    if (iolock && !iothread && !qemu_in_coroutine()) {
+        qemu_mutex_unlock_iothread();
+    }
+
+    ret = qio_channel_readv_full_all_eof(ioc, &iov, 1, fds, nfds, errp);
+
+    if (iolock && !iothread && !qemu_in_coroutine()) {
+        qemu_mutex_lock_iothread();
+    }
+
+    return (ret <= 0) ? ret : iov.iov_len;
+}
+
+bool mpqemu_msg_recv(MPQemuMsg *msg, QIOChannel *ioc, Error **errp)
+{
+    ERRP_GUARD();
+    g_autofree int *fds = NULL;
+    size_t nfds = 0;
+    ssize_t len;
+    bool ret = false;
+
+    len = mpqemu_read(ioc, msg, MPQEMU_MSG_HDR_SIZE, &fds, &nfds, errp);
+    if (len <= 0) {
+        goto fail;
+    } else if (len != MPQEMU_MSG_HDR_SIZE) {
+        error_setg(errp, "Message header corrupted");
+        goto fail;
+    }
+
+    if (msg->size > sizeof(msg->data)) {
+        error_setg(errp, "Invalid size for message");
+        goto fail;
+    }
+
+    if (!msg->size) {
+        goto copy_fds;
+    }
+
+    len = mpqemu_read(ioc, &msg->data, msg->size, NULL, NULL, errp);
+    if (len <= 0) {
+        goto fail;
+    }
+    if (len != msg->size) {
+        error_setg(errp, "Unable to read full message");
+        goto fail;
+    }
+
+copy_fds:
+    msg->num_fds = nfds;
+    if (nfds > G_N_ELEMENTS(msg->fds)) {
+        error_setg(errp,
+                   "Overflow error: received %zu fds, more than max of %d fds",
+                   nfds, REMOTE_MAX_FDS);
+        goto fail;
+    }
+    if (nfds) {
+        memcpy(msg->fds, fds, nfds * sizeof(int));
+    }
+
+    ret = true;
+
+fail:
+    if (*errp) {
+        trace_mpqemu_recv_io_error(msg->cmd, msg->size, nfds);
+    }
+    while (*errp && nfds) {
+        close(fds[nfds - 1]);
+        nfds--;
+    }
+
+    return ret;
+}
+
+bool mpqemu_msg_valid(MPQemuMsg *msg)
+{
+    if (msg->cmd >= MPQEMU_CMD_MAX && msg->cmd < 0) {
+        return false;
+    }
+
+    /* Verify FDs. */
+    if (msg->num_fds >= REMOTE_MAX_FDS) {
+        return false;
+    }
+
+    if (msg->num_fds > 0) {
+        for (int i = 0; i < msg->num_fds; i++) {
+            if (fcntl(msg->fds[i], F_GETFL) == -1) {
+                return false;
+            }
+        }
+    }
+
+    return true;
+}
diff --git a/iothread.c b/iothread.c
index XXXXXXX..XXXXXXX 100644
--- a/iothread.c
+++ b/iothread.c
@@ -XXX,XX +XXX,XX @@ IOThread *iothread_by_id(const char *id)
 {
     return IOTHREAD(object_resolve_path_type(id, TYPE_IOTHREAD, NULL));
 }
+
+bool qemu_in_iothread(void)
+{
+    return qemu_get_current_aio_context() == qemu_get_aio_context() ?
+                    false : true;
+}
diff --git a/hw/remote/meson.build b/hw/remote/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/meson.build
+++ b/hw/remote/meson.build
@@ -XXX,XX +XXX,XX @@
 remote_ss = ss.source_set()
 
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('machine.c'))
+remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('mpqemu-link.c'))
 
 softmmu_ss.add_all(when: 'CONFIG_MULTIPROCESS', if_true: remote_ss)
diff --git a/hw/remote/trace-events b/hw/remote/trace-events
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/remote/trace-events
@@ -XXX,XX +XXX,XX @@
+# multi-process trace events
+
+mpqemu_send_io_error(int cmd, int size, int nfds) "send command %d size %d, %d file descriptors to remote process"
+mpqemu_recv_io_error(int cmd, int size, int nfds) "failed to receive %d size %d, %d file descriptors to remote process"
-- 
2.29.2

From: Jagannathan Raman <jag.raman@oracle.com>

Initializes the message handler function in the remote process. It is
called whenever there's an event pending on QIOChannel that registers
this function.

Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 99d38d8b93753a6409ac2340e858858cda59ab1b.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 MAINTAINERS                 |  1 +
 include/hw/remote/machine.h |  9 ++++++
 hw/remote/message.c         | 57 +++++++++++++++++++++++++++++++++++++
 hw/remote/meson.build       |  1 +
 4 files changed, 68 insertions(+)
 create mode 100644 hw/remote/message.c

diff --git a/MAINTAINERS b/MAINTAINERS
index XXXXXXX..XXXXXXX 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -XXX,XX +XXX,XX @@ F: hw/remote/machine.c
 F: include/hw/remote/machine.h
 F: hw/remote/mpqemu-link.c
 F: include/hw/remote/mpqemu-link.h
+F: hw/remote/message.c
 
 Build and test automation
 -------------------------
diff --git a/include/hw/remote/machine.h b/include/hw/remote/machine.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/remote/machine.h
+++ b/include/hw/remote/machine.h
@@ -XXX,XX +XXX,XX @@
 #include "qom/object.h"
 #include "hw/boards.h"
 #include "hw/pci-host/remote.h"
+#include "io/channel.h"
 
 struct RemoteMachineState {
     MachineState parent_obj;
@@ -XXX,XX +XXX,XX @@ struct RemoteMachineState {
     RemotePCIHost *host;
 };
 
+/* Used to pass to co-routine device and ioc. */
+typedef struct RemoteCommDev {
+    PCIDevice *dev;
+    QIOChannel *ioc;
+} RemoteCommDev;
+
 #define TYPE_REMOTE_MACHINE "x-remote-machine"
 OBJECT_DECLARE_SIMPLE_TYPE(RemoteMachineState, REMOTE_MACHINE)
 
+void coroutine_fn mpqemu_remote_msg_loop_co(void *data);
+
 #endif
diff --git a/hw/remote/message.c b/hw/remote/message.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/remote/message.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * Copyright © 2020, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL-v2, version 2 or later.
+ *
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+
+#include "hw/remote/machine.h"
+#include "io/channel.h"
+#include "hw/remote/mpqemu-link.h"
+#include "qapi/error.h"
+#include "sysemu/runstate.h"
+
+void coroutine_fn mpqemu_remote_msg_loop_co(void *data)
+{
+    g_autofree RemoteCommDev *com = (RemoteCommDev *)data;
+    PCIDevice *pci_dev = NULL;
+    Error *local_err = NULL;
+
+    assert(com->ioc);
+
+    pci_dev = com->dev;
+    for (; !local_err;) {
+        MPQemuMsg msg = {0};
+
+        if (!mpqemu_msg_recv(&msg, com->ioc, &local_err)) {
+            break;
+        }
+
+        if (!mpqemu_msg_valid(&msg)) {
+            error_setg(&local_err, "Received invalid message from proxy"
+                                   "in remote process pid="FMT_pid"",
+                                   getpid());
+            break;
+        }
+
+        switch (msg.cmd) {
+        default:
+            error_setg(&local_err,
+                       "Unknown command (%d) received for device %s"
+                       " (pid="FMT_pid")",
+                       msg.cmd, DEVICE(pci_dev)->id, getpid());
+        }
+    }
+
+    if (local_err) {
+        error_report_err(local_err);
+        qemu_system_shutdown_request(SHUTDOWN_CAUSE_HOST_ERROR);
+    } else {
+        qemu_system_shutdown_request(SHUTDOWN_CAUSE_GUEST_SHUTDOWN);
+    }
+}
diff --git a/hw/remote/meson.build b/hw/remote/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/meson.build
+++ b/hw/remote/meson.build
@@ -XXX,XX +XXX,XX @@ remote_ss = ss.source_set()
 
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('machine.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('mpqemu-link.c'))
+remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('message.c'))
 
 softmmu_ss.add_all(when: 'CONFIG_MULTIPROCESS', if_true: remote_ss)
-- 
2.29.2

From: Jagannathan Raman <jag.raman@oracle.com>

Associate the file descriptor for a PCIDevice in remote process with
DeviceState object.

Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: f405a2ed5d7518b87bea7c59cfdf334d67e5ee51.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 MAINTAINERS            |   1 +
 hw/remote/remote-obj.c | 203 +++++++++++++++++++++++++++++++++++++++++
 hw/remote/meson.build  |   1 +
 3 files changed, 205 insertions(+)
 create mode 100644 hw/remote/remote-obj.c

diff --git a/MAINTAINERS b/MAINTAINERS
index XXXXXXX..XXXXXXX 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -XXX,XX +XXX,XX @@ F: include/hw/remote/machine.h
 F: hw/remote/mpqemu-link.c
 F: include/hw/remote/mpqemu-link.h
 F: hw/remote/message.c
+F: hw/remote/remote-obj.c
 
 Build and test automation
 -------------------------
diff --git a/hw/remote/remote-obj.c b/hw/remote/remote-obj.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/remote/remote-obj.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * Copyright © 2020, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL-v2, version 2 or later.
+ *
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+
+#include "qemu/error-report.h"
+#include "qemu/notify.h"
+#include "qom/object_interfaces.h"
+#include "hw/qdev-core.h"
+#include "io/channel.h"
+#include "hw/qdev-core.h"
+#include "hw/remote/machine.h"
+#include "io/channel-util.h"
+#include "qapi/error.h"
+#include "sysemu/sysemu.h"
+#include "hw/pci/pci.h"
+#include "qemu/sockets.h"
+#include "monitor/monitor.h"
+
+#define TYPE_REMOTE_OBJECT "x-remote-object"
+OBJECT_DECLARE_TYPE(RemoteObject, RemoteObjectClass, REMOTE_OBJECT)
+
+struct RemoteObjectClass {
+    ObjectClass parent_class;
+
+    unsigned int nr_devs;
+    unsigned int max_devs;
+};
+
+struct RemoteObject {
+    /* private */
+    Object parent;
+
+    Notifier machine_done;
+
+    int32_t fd;
+    char *devid;
+
+    QIOChannel *ioc;
+
+    DeviceState *dev;
+    DeviceListener listener;
+};
+
+static void remote_object_set_fd(Object *obj, const char *str, Error **errp)
+{
+    RemoteObject *o = REMOTE_OBJECT(obj);
+    int fd = -1;
+
+    fd = monitor_fd_param(monitor_cur(), str, errp);
+    if (fd == -1) {
+        error_prepend(errp, "Could not parse remote object fd %s:", str);
+        return;
+    }
+
+    if (!fd_is_socket(fd)) {
+        error_setg(errp, "File descriptor '%s' is not a socket", str);
+        close(fd);
+        return;
+    }
+
+    o->fd = fd;
+}
+
+static void remote_object_set_devid(Object *obj, const char *str, Error **errp)
+{
+    RemoteObject *o = REMOTE_OBJECT(obj);
+
+    g_free(o->devid);
+
+    o->devid = g_strdup(str);
+}
+
+static void remote_object_unrealize_listener(DeviceListener *listener,
+                                             DeviceState *dev)
+{
+    RemoteObject *o = container_of(listener, RemoteObject, listener);
+
+    if (o->dev == dev) {
+        object_unref(OBJECT(o));
+    }
+}
+
+static void remote_object_machine_done(Notifier *notifier, void *data)
+{
+    RemoteObject *o = container_of(notifier, RemoteObject, machine_done);
+    DeviceState *dev = NULL;
+    QIOChannel *ioc = NULL;
+    Coroutine *co = NULL;
+    RemoteCommDev *comdev = NULL;
+    Error *err = NULL;
+
+    dev = qdev_find_recursive(sysbus_get_default(), o->devid);
+    if (!dev || !object_dynamic_cast(OBJECT(dev), TYPE_PCI_DEVICE)) {
+        error_report("%s is not a PCI device", o->devid);
+        return;
+    }
+
+    ioc = qio_channel_new_fd(o->fd, &err);
+    if (!ioc) {
+        error_report_err(err);
+        return;
+    }
+    qio_channel_set_blocking(ioc, false, NULL);
+
+    o->dev = dev;
+
+    o->listener.unrealize = remote_object_unrealize_listener;
+    device_listener_register(&o->listener);
+
+    /* co-routine should free this. */
+    comdev = g_new0(RemoteCommDev, 1);
+    *comdev = (RemoteCommDev) {
+        .ioc = ioc,
+        .dev = PCI_DEVICE(dev),
+    };
+
+    co = qemu_coroutine_create(mpqemu_remote_msg_loop_co, comdev);
+    qemu_coroutine_enter(co);
+}
+
+static void remote_object_init(Object *obj)
+{
+    RemoteObjectClass *k = REMOTE_OBJECT_GET_CLASS(obj);
+    RemoteObject *o = REMOTE_OBJECT(obj);
+
+    if (k->nr_devs >= k->max_devs) {
+        error_report("Reached maximum number of devices: %u", k->max_devs);
+        return;
+    }
+
+    o->ioc = NULL;
+    o->fd = -1;
+    o->devid = NULL;
+
+    k->nr_devs++;
+
+    o->machine_done.notify = remote_object_machine_done;
+    qemu_add_machine_init_done_notifier(&o->machine_done);
+}
+
+static void remote_object_finalize(Object *obj)
+{
+    RemoteObjectClass *k = REMOTE_OBJECT_GET_CLASS(obj);
+    RemoteObject *o = REMOTE_OBJECT(obj);
+
+    device_listener_unregister(&o->listener);
+
+    if (o->ioc) {
+        qio_channel_shutdown(o->ioc, QIO_CHANNEL_SHUTDOWN_BOTH, NULL);
+        qio_channel_close(o->ioc, NULL);
+    }
+
+    object_unref(OBJECT(o->ioc));
+
+    k->nr_devs--;
+    g_free(o->devid);
+}
+
+static void remote_object_class_init(ObjectClass *klass, void *data)
+{
+    RemoteObjectClass *k = REMOTE_OBJECT_CLASS(klass);
+
+    /*
+     * Limit number of supported devices to 1. This is done to avoid devices
+     * from one VM accessing the RAM of another VM. This is done until we
+     * start using separate address spaces for individual devices.
+     */
+    k->max_devs = 1;
+    k->nr_devs = 0;
+
+    object_class_property_add_str(klass, "fd", NULL, remote_object_set_fd);
+    object_class_property_add_str(klass, "devid", NULL,
+                                  remote_object_set_devid);
+}
+
+static const TypeInfo remote_object_info = {
+    .name = TYPE_REMOTE_OBJECT,
+    .parent = TYPE_OBJECT,
+    .instance_size = sizeof(RemoteObject),
+    .instance_init = remote_object_init,
+    .instance_finalize = remote_object_finalize,
+    .class_size = sizeof(RemoteObjectClass),
+    .class_init = remote_object_class_init,
+    .interfaces = (InterfaceInfo[]) {
+        { TYPE_USER_CREATABLE },
+        { }
+    }
+};
+
+static void register_types(void)
+{
+    type_register_static(&remote_object_info);
+}
+
+type_init(register_types);
diff --git a/hw/remote/meson.build b/hw/remote/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/meson.build
+++ b/hw/remote/meson.build
@@ -XXX,XX +XXX,XX @@ remote_ss = ss.source_set()
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('machine.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('mpqemu-link.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('message.c'))
+remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('remote-obj.c'))
 
 softmmu_ss.add_all(when: 'CONFIG_MULTIPROCESS', if_true: remote_ss)
-- 
2.29.2

From: Jagannathan Raman <jag.raman@oracle.com>

SyncSysMemMsg message format is defined. It is used to send
file descriptors of the RAM regions to remote device.
RAM on the remote device is configured with a set of file descriptors.
Old RAM regions are deleted and new regions, each with an fd, is
added to the RAM.

Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 7d2d1831d812e85f681e7a8ab99e032cf4704689.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 MAINTAINERS                     |  2 +
 include/hw/remote/memory.h      | 19 ++++++++++
 include/hw/remote/mpqemu-link.h | 10 +++++
 hw/remote/memory.c              | 65 +++++++++++++++++++++++++++++++++
 hw/remote/mpqemu-link.c         | 11 ++++++
 hw/remote/meson.build           |  2 +
 6 files changed, 109 insertions(+)
 create mode 100644 include/hw/remote/memory.h
 create mode 100644 hw/remote/memory.c

diff --git a/MAINTAINERS b/MAINTAINERS
index XXXXXXX..XXXXXXX 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -XXX,XX +XXX,XX @@ F: hw/remote/mpqemu-link.c
 F: include/hw/remote/mpqemu-link.h
 F: hw/remote/message.c
 F: hw/remote/remote-obj.c
+F: include/hw/remote/memory.h
+F: hw/remote/memory.c
 
 Build and test automation
 -------------------------
diff --git a/include/hw/remote/memory.h b/include/hw/remote/memory.h
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/include/hw/remote/memory.h
@@ -XXX,XX +XXX,XX @@
+/*
+ * Memory manager for remote device
+ *
+ * Copyright © 2018, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef REMOTE_MEMORY_H
+#define REMOTE_MEMORY_H
+
+#include "exec/hwaddr.h"
+#include "hw/remote/mpqemu-link.h"
+
+void remote_sysmem_reconfig(MPQemuMsg *msg, Error **errp);
+
+#endif
diff --git a/include/hw/remote/mpqemu-link.h b/include/hw/remote/mpqemu-link.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/remote/mpqemu-link.h
+++ b/include/hw/remote/mpqemu-link.h
@@ -XXX,XX +XXX,XX @@
 #include "qom/object.h"
 #include "qemu/thread.h"
 #include "io/channel.h"
+#include "exec/hwaddr.h"
 
 #define REMOTE_MAX_FDS 8
 
@@ -XXX,XX +XXX,XX @@
  *
  */
 typedef enum {
+    MPQEMU_CMD_SYNC_SYSMEM,
     MPQEMU_CMD_MAX,
 } MPQemuCmd;
 
+typedef struct {
+    hwaddr gpas[REMOTE_MAX_FDS];
+    uint64_t sizes[REMOTE_MAX_FDS];
+    off_t offsets[REMOTE_MAX_FDS];
+} SyncSysmemMsg;
+
 /**
  * MPQemuMsg:
  * @cmd: The remote command
@@ -XXX,XX +XXX,XX @@ typedef enum {
  * MPQemuMsg Format of the message sent to the remote device from QEMU.
  *
  */
+
 typedef struct {
     int cmd;
     size_t size;
 
     union {
         uint64_t u64;
+        SyncSysmemMsg sync_sysmem;
     } data;
 
     int fds[REMOTE_MAX_FDS];
diff --git a/hw/remote/memory.c b/hw/remote/memory.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/remote/memory.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * Memory manager for remote device
+ *
+ * Copyright © 2018, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+
+#include "hw/remote/memory.h"
+#include "exec/address-spaces.h"
+#include "exec/ram_addr.h"
+#include "qapi/error.h"
+
+static void remote_sysmem_reset(void)
+{
+    MemoryRegion *sysmem, *subregion, *next;
+
+    sysmem = get_system_memory();
+
+    QTAILQ_FOREACH_SAFE(subregion, &sysmem->subregions, subregions_link, next) {
+        if (subregion->ram) {
+            memory_region_del_subregion(sysmem, subregion);
+            object_unparent(OBJECT(subregion));
+        }
+    }
+}
+
+void remote_sysmem_reconfig(MPQemuMsg *msg, Error **errp)
+{
+    ERRP_GUARD();
+    SyncSysmemMsg *sysmem_info = &msg->data.sync_sysmem;
+    MemoryRegion *sysmem, *subregion;
+    static unsigned int suffix;
+    int region;
+
+    sysmem = get_system_memory();
+
+    remote_sysmem_reset();
+
+    for (region = 0; region < msg->num_fds; region++) {
+        g_autofree char *name;
+        subregion = g_new(MemoryRegion, 1);
+        name = g_strdup_printf("remote-mem-%u", suffix++);
+        memory_region_init_ram_from_fd(subregion, NULL,
+                                       name, sysmem_info->sizes[region],
+                                       true, msg->fds[region],
+                                       sysmem_info->offsets[region],
+                                       errp);
+
+        if (*errp) {
+            g_free(subregion);
+            remote_sysmem_reset();
+            return;
+        }
+
+        memory_region_add_subregion(sysmem, sysmem_info->gpas[region],
+                                    subregion);
+
+    }
+}
diff --git a/hw/remote/mpqemu-link.c b/hw/remote/mpqemu-link.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/mpqemu-link.c
+++ b/hw/remote/mpqemu-link.c
@@ -XXX,XX +XXX,XX @@ bool mpqemu_msg_valid(MPQemuMsg *msg)
         }
     }
 
+     /* Verify message specific fields. */
+    switch (msg->cmd) {
+    case MPQEMU_CMD_SYNC_SYSMEM:
+        if (msg->num_fds == 0 || msg->size != sizeof(SyncSysmemMsg)) {
+            return false;
+        }
+        break;
+    default:
+        break;
+    }
+
     return true;
 }
diff --git a/hw/remote/meson.build b/hw/remote/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/meson.build
+++ b/hw/remote/meson.build
@@ -XXX,XX +XXX,XX @@ remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('mpqemu-link.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('message.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('remote-obj.c'))
 
+specific_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('memory.c'))
+
 softmmu_ss.add_all(when: 'CONFIG_MULTIPROCESS', if_true: remote_ss)
-- 
2.29.2

From: Elena Ufimtseva <elena.ufimtseva@oracle.com>

Defines a PCI Device proxy object as a child of TYPE_PCI_DEVICE.

Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: b5186ebfedf8e557044d09a768846c59230ad3a7.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 MAINTAINERS               |  2 +
 include/hw/remote/proxy.h | 33 +++++++++++++
 hw/remote/proxy.c         | 99 +++++++++++++++++++++++++++++++++++++++
 hw/remote/meson.build     |  1 +
 4 files changed, 135 insertions(+)
 create mode 100644 include/hw/remote/proxy.h
 create mode 100644 hw/remote/proxy.c

diff --git a/MAINTAINERS b/MAINTAINERS
index XXXXXXX..XXXXXXX 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -XXX,XX +XXX,XX @@ F: hw/remote/message.c
 F: hw/remote/remote-obj.c
 F: include/hw/remote/memory.h
 F: hw/remote/memory.c
+F: hw/remote/proxy.c
+F: include/hw/remote/proxy.h
 
 Build and test automation
 -------------------------
diff --git a/include/hw/remote/proxy.h b/include/hw/remote/proxy.h
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/include/hw/remote/proxy.h
@@ -XXX,XX +XXX,XX @@
+/*
+ * Copyright © 2018, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef PROXY_H
+#define PROXY_H
+
+#include "hw/pci/pci.h"
+#include "io/channel.h"
+
+#define TYPE_PCI_PROXY_DEV "x-pci-proxy-dev"
+OBJECT_DECLARE_SIMPLE_TYPE(PCIProxyDev, PCI_PROXY_DEV)
+
+struct PCIProxyDev {
+    PCIDevice parent_dev;
+    char *fd;
+
+    /*
+     * Mutex used to protect the QIOChannel fd from
+     * the concurrent access by the VCPUs since proxy
+     * blocks while awaiting for the replies from the
+     * process remote.
+     */
+    QemuMutex io_mutex;
+    QIOChannel *ioc;
+    Error *migration_blocker;
+};
+
+#endif /* PROXY_H */
diff --git a/hw/remote/proxy.c b/hw/remote/proxy.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/remote/proxy.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * Copyright © 2018, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+
+#include "hw/remote/proxy.h"
+#include "hw/pci/pci.h"
+#include "qapi/error.h"
+#include "io/channel-util.h"
+#include "hw/qdev-properties.h"
+#include "monitor/monitor.h"
+#include "migration/blocker.h"
+#include "qemu/sockets.h"
+
+static void pci_proxy_dev_realize(PCIDevice *device, Error **errp)
+{
+    ERRP_GUARD();
+    PCIProxyDev *dev = PCI_PROXY_DEV(device);
+    int fd;
+
+    if (!dev->fd) {
+        error_setg(errp, "fd parameter not specified for %s",
+                   DEVICE(device)->id);
+        return;
+    }
+
+    fd = monitor_fd_param(monitor_cur(), dev->fd, errp);
+    if (fd == -1) {
+        error_prepend(errp, "proxy: unable to parse fd %s: ", dev->fd);
+        return;
+    }
+
+    if (!fd_is_socket(fd)) {
+        error_setg(errp, "proxy: fd %d is not a socket", fd);
+        close(fd);
+        return;
+    }
+
+    dev->ioc = qio_channel_new_fd(fd, errp);
+
+    error_setg(&dev->migration_blocker, "%s does not support migration",
+               TYPE_PCI_PROXY_DEV);
+    migrate_add_blocker(dev->migration_blocker, errp);
+
+    qemu_mutex_init(&dev->io_mutex);
+    qio_channel_set_blocking(dev->ioc, true, NULL);
+}
+
+static void pci_proxy_dev_exit(PCIDevice *pdev)
+{
+    PCIProxyDev *dev = PCI_PROXY_DEV(pdev);
+
+    if (dev->ioc) {
+        qio_channel_close(dev->ioc, NULL);
+    }
+
+    migrate_del_blocker(dev->migration_blocker);
+
+    error_free(dev->migration_blocker);
+}
+
+static Property proxy_properties[] = {
+    DEFINE_PROP_STRING("fd", PCIProxyDev, fd),
+    DEFINE_PROP_END_OF_LIST(),
+};
+
+static void pci_proxy_dev_class_init(ObjectClass *klass, void *data)
+{
+    DeviceClass *dc = DEVICE_CLASS(klass);
+    PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
+
+    k->realize = pci_proxy_dev_realize;
+    k->exit = pci_proxy_dev_exit;
+    device_class_set_props(dc, proxy_properties);
+}
+
+static const TypeInfo pci_proxy_dev_type_info = {
+    .name          = TYPE_PCI_PROXY_DEV,
+    .parent        = TYPE_PCI_DEVICE,
+    .instance_size = sizeof(PCIProxyDev),
+    .class_init    = pci_proxy_dev_class_init,
+    .interfaces = (InterfaceInfo[]) {
+        { INTERFACE_CONVENTIONAL_PCI_DEVICE },
+        { },
+    },
+};
+
+static void pci_proxy_dev_register_types(void)
+{
+    type_register_static(&pci_proxy_dev_type_info);
+}
+
+type_init(pci_proxy_dev_register_types)
diff --git a/hw/remote/meson.build b/hw/remote/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/meson.build
+++ b/hw/remote/meson.build
@@ -XXX,XX +XXX,XX @@ remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('machine.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('mpqemu-link.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('message.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('remote-obj.c'))
+remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('proxy.c'))
 
 specific_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('memory.c'))
 
-- 
2.29.2

From: Elena Ufimtseva <elena.ufimtseva@oracle.com>

Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: d54edb4176361eed86b903e8f27058363b6c83b3.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/hw/remote/mpqemu-link.h |  4 ++++
 hw/remote/mpqemu-link.c         | 34 +++++++++++++++++++++++++++++++++
 2 files changed, 38 insertions(+)

diff --git a/include/hw/remote/mpqemu-link.h b/include/hw/remote/mpqemu-link.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/remote/mpqemu-link.h
+++ b/include/hw/remote/mpqemu-link.h
@@ -XXX,XX +XXX,XX @@
 #include "qemu/thread.h"
 #include "io/channel.h"
 #include "exec/hwaddr.h"
+#include "io/channel-socket.h"
+#include "hw/remote/proxy.h"
 
 #define REMOTE_MAX_FDS 8
 
@@ -XXX,XX +XXX,XX @@ typedef struct {
 bool mpqemu_msg_send(MPQemuMsg *msg, QIOChannel *ioc, Error **errp);
 bool mpqemu_msg_recv(MPQemuMsg *msg, QIOChannel *ioc, Error **errp);
 
+uint64_t mpqemu_msg_send_and_await_reply(MPQemuMsg *msg, PCIProxyDev *pdev,
+                                         Error **errp);
 bool mpqemu_msg_valid(MPQemuMsg *msg);
 
 #endif
diff --git a/hw/remote/mpqemu-link.c b/hw/remote/mpqemu-link.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/mpqemu-link.c
+++ b/hw/remote/mpqemu-link.c
@@ -XXX,XX +XXX,XX @@ fail:
     return ret;
 }
 
+/*
+ * Send msg and wait for a reply with command code RET_MSG.
+ * Returns the message received of size u64 or UINT64_MAX
+ * on error.
+ * Called from VCPU thread in non-coroutine context.
+ * Used by the Proxy object to communicate to remote processes.
+ */
+uint64_t mpqemu_msg_send_and_await_reply(MPQemuMsg *msg, PCIProxyDev *pdev,
+                                         Error **errp)
+{
+    ERRP_GUARD();
+    MPQemuMsg msg_reply = {0};
+    uint64_t ret = UINT64_MAX;
+
+    assert(!qemu_in_coroutine());
+
+    QEMU_LOCK_GUARD(&pdev->io_mutex);
+    if (!mpqemu_msg_send(msg, pdev->ioc, errp)) {
+        return ret;
+    }
+
+    if (!mpqemu_msg_recv(&msg_reply, pdev->ioc, errp)) {
+        return ret;
+    }
+
+    if (!mpqemu_msg_valid(&msg_reply)) {
+        error_setg(errp, "ERROR: Invalid reply received for command %d",
+                         msg->cmd);
+        return ret;
+    }
+
+    return msg_reply.data.u64;
+}
+
 bool mpqemu_msg_valid(MPQemuMsg *msg)
 {
     if (msg->cmd >= MPQEMU_CMD_MAX && msg->cmd < 0) {
-- 
2.29.2

From: Elena Ufimtseva <elena.ufimtseva@oracle.com>

The Proxy Object sends the PCI config space accesses as messages
to the remote process over the communication channel

Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: d3c94f4618813234655356c60e6f0d0362ff42d6.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/hw/remote/mpqemu-link.h | 10 ++++++
 hw/remote/message.c             | 60 +++++++++++++++++++++++++++++++++
 hw/remote/mpqemu-link.c         |  8 ++++-
 hw/remote/proxy.c               | 55 ++++++++++++++++++++++++++++++
 4 files changed, 132 insertions(+), 1 deletion(-)

diff --git a/include/hw/remote/mpqemu-link.h b/include/hw/remote/mpqemu-link.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/remote/mpqemu-link.h
+++ b/include/hw/remote/mpqemu-link.h
@@ -XXX,XX +XXX,XX @@
  */
 typedef enum {
     MPQEMU_CMD_SYNC_SYSMEM,
+    MPQEMU_CMD_RET,
+    MPQEMU_CMD_PCI_CFGWRITE,
+    MPQEMU_CMD_PCI_CFGREAD,
     MPQEMU_CMD_MAX,
 } MPQemuCmd;
 
@@ -XXX,XX +XXX,XX @@ typedef struct {
     off_t offsets[REMOTE_MAX_FDS];
 } SyncSysmemMsg;
 
+typedef struct {
+    uint32_t addr;
+    uint32_t val;
+    int len;
+} PciConfDataMsg;
+
 /**
  * MPQemuMsg:
  * @cmd: The remote command
@@ -XXX,XX +XXX,XX @@ typedef struct {
 
     union {
         uint64_t u64;
+        PciConfDataMsg pci_conf_data;
         SyncSysmemMsg sync_sysmem;
     } data;
 
diff --git a/hw/remote/message.c b/hw/remote/message.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/message.c
+++ b/hw/remote/message.c
@@ -XXX,XX +XXX,XX @@
 #include "hw/remote/mpqemu-link.h"
 #include "qapi/error.h"
 #include "sysemu/runstate.h"
+#include "hw/pci/pci.h"
+
+static void process_config_write(QIOChannel *ioc, PCIDevice *dev,
+                                 MPQemuMsg *msg, Error **errp);
+static void process_config_read(QIOChannel *ioc, PCIDevice *dev,
+                                MPQemuMsg *msg, Error **errp);
 
 void coroutine_fn mpqemu_remote_msg_loop_co(void *data)
 {
@@ -XXX,XX +XXX,XX @@ void coroutine_fn mpqemu_remote_msg_loop_co(void *data)
         }
 
         switch (msg.cmd) {
+        case MPQEMU_CMD_PCI_CFGWRITE:
+            process_config_write(com->ioc, pci_dev, &msg, &local_err);
+            break;
+        case MPQEMU_CMD_PCI_CFGREAD:
+            process_config_read(com->ioc, pci_dev, &msg, &local_err);
+            break;
         default:
             error_setg(&local_err,
                        "Unknown command (%d) received for device %s"
@@ -XXX,XX +XXX,XX @@ void coroutine_fn mpqemu_remote_msg_loop_co(void *data)
         qemu_system_shutdown_request(SHUTDOWN_CAUSE_GUEST_SHUTDOWN);
     }
 }
+
+static void process_config_write(QIOChannel *ioc, PCIDevice *dev,
+                                 MPQemuMsg *msg, Error **errp)
+{
+    ERRP_GUARD();
+    PciConfDataMsg *conf = (PciConfDataMsg *)&msg->data.pci_conf_data;
+    MPQemuMsg ret = { 0 };
+
+    if ((conf->addr + sizeof(conf->val)) > pci_config_size(dev)) {
+        error_setg(errp, "Bad address for PCI config write, pid "FMT_pid".",
+                   getpid());
+        ret.data.u64 = UINT64_MAX;
+    } else {
+        pci_default_write_config(dev, conf->addr, conf->val, conf->len);
+    }
+
+    ret.cmd = MPQEMU_CMD_RET;
+    ret.size = sizeof(ret.data.u64);
+
+    if (!mpqemu_msg_send(&ret, ioc, NULL)) {
+        error_prepend(errp, "Error returning code to proxy, pid "FMT_pid": ",
+                      getpid());
+    }
+}
+
+static void process_config_read(QIOChannel *ioc, PCIDevice *dev,
+                                MPQemuMsg *msg, Error **errp)
+{
+    ERRP_GUARD();
+    PciConfDataMsg *conf = (PciConfDataMsg *)&msg->data.pci_conf_data;
+    MPQemuMsg ret = { 0 };
+
+    if ((conf->addr + sizeof(conf->val)) > pci_config_size(dev)) {
+        error_setg(errp, "Bad address for PCI config read, pid "FMT_pid".",
+                   getpid());
+        ret.data.u64 = UINT64_MAX;
+    } else {
+        ret.data.u64 = pci_default_read_config(dev, conf->addr, conf->len);
+    }
+
+    ret.cmd = MPQEMU_CMD_RET;
+    ret.size = sizeof(ret.data.u64);
+
+    if (!mpqemu_msg_send(&ret, ioc, NULL)) {
+        error_prepend(errp, "Error returning code to proxy, pid "FMT_pid": ",
+                      getpid());
+    }
+}
diff --git a/hw/remote/mpqemu-link.c b/hw/remote/mpqemu-link.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/mpqemu-link.c
+++ b/hw/remote/mpqemu-link.c
@@ -XXX,XX +XXX,XX @@ uint64_t mpqemu_msg_send_and_await_reply(MPQemuMsg *msg, PCIProxyDev *pdev,
         return ret;
     }
 
-    if (!mpqemu_msg_valid(&msg_reply)) {
+    if (!mpqemu_msg_valid(&msg_reply) || msg_reply.cmd != MPQEMU_CMD_RET) {
         error_setg(errp, "ERROR: Invalid reply received for command %d",
                          msg->cmd);
         return ret;
@@ -XXX,XX +XXX,XX @@ bool mpqemu_msg_valid(MPQemuMsg *msg)
             return false;
         }
         break;
+    case MPQEMU_CMD_PCI_CFGWRITE:
+    case MPQEMU_CMD_PCI_CFGREAD:
+        if (msg->size != sizeof(PciConfDataMsg)) {
+            return false;
+        }
+        break;
     default:
         break;
     }
diff --git a/hw/remote/proxy.c b/hw/remote/proxy.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/proxy.c
+++ b/hw/remote/proxy.c
@@ -XXX,XX +XXX,XX @@
 #include "monitor/monitor.h"
 #include "migration/blocker.h"
 #include "qemu/sockets.h"
+#include "hw/remote/mpqemu-link.h"
+#include "qemu/error-report.h"
 
 static void pci_proxy_dev_realize(PCIDevice *device, Error **errp)
 {
@@ -XXX,XX +XXX,XX @@ static void pci_proxy_dev_exit(PCIDevice *pdev)
     error_free(dev->migration_blocker);
 }
 
+static void config_op_send(PCIProxyDev *pdev, uint32_t addr, uint32_t *val,
+                           int len, unsigned int op)
+{
+    MPQemuMsg msg = { 0 };
+    uint64_t ret = -EINVAL;
+    Error *local_err = NULL;
+
+    msg.cmd = op;
+    msg.data.pci_conf_data.addr = addr;
+    msg.data.pci_conf_data.val = (op == MPQEMU_CMD_PCI_CFGWRITE) ? *val : 0;
+    msg.data.pci_conf_data.len = len;
+    msg.size = sizeof(PciConfDataMsg);
+
+    ret = mpqemu_msg_send_and_await_reply(&msg, pdev, &local_err);
+    if (local_err) {
+        error_report_err(local_err);
+    }
+
+    if (ret == UINT64_MAX) {
+        error_report("Failed to perform PCI config %s operation",
+                     (op == MPQEMU_CMD_PCI_CFGREAD) ? "READ" : "WRITE");
+    }
+
+    if (op == MPQEMU_CMD_PCI_CFGREAD) {
+        *val = (uint32_t)ret;
+    }
+}
+
+static uint32_t pci_proxy_read_config(PCIDevice *d, uint32_t addr, int len)
+{
+    uint32_t val;
+
+    config_op_send(PCI_PROXY_DEV(d), addr, &val, len, MPQEMU_CMD_PCI_CFGREAD);
+
+    return val;
+}
+
+static void pci_proxy_write_config(PCIDevice *d, uint32_t addr, uint32_t val,
+                                   int len)
+{
+    /*
+     * Some of the functions access the copy of remote device's PCI config
+     * space which is cached in the proxy device. Therefore, maintain
+     * it updated.
+     */
+    pci_default_write_config(d, addr, val, len);
+
+    config_op_send(PCI_PROXY_DEV(d), addr, &val, len, MPQEMU_CMD_PCI_CFGWRITE);
+}
+
 static Property proxy_properties[] = {
     DEFINE_PROP_STRING("fd", PCIProxyDev, fd),
     DEFINE_PROP_END_OF_LIST(),
@@ -XXX,XX +XXX,XX @@ static void pci_proxy_dev_class_init(ObjectClass *klass, void *data)
 
     k->realize = pci_proxy_dev_realize;
     k->exit = pci_proxy_dev_exit;
+    k->config_read = pci_proxy_read_config;
+    k->config_write = pci_proxy_write_config;
+
     device_class_set_props(dc, proxy_properties);
 }
 
-- 
2.29.2

From: Jagannathan Raman <jag.raman@oracle.com>

Proxy device object implements handler for PCI BAR writes and reads.
The handler uses BAR_WRITE/BAR_READ message to communicate to the
remote process with the BAR address and value to be written/read.
The remote process implements handler for BAR_WRITE/BAR_READ
message.

Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: a8b76714a9688be5552c4c92d089bc9e8a4707ff.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/hw/remote/mpqemu-link.h | 10 ++++
 include/hw/remote/proxy.h       |  9 ++++
 hw/remote/message.c             | 83 +++++++++++++++++++++++++++++++++
 hw/remote/mpqemu-link.c         |  6 +++
 hw/remote/proxy.c               | 60 ++++++++++++++++++++++++
 5 files changed, 168 insertions(+)

diff --git a/include/hw/remote/mpqemu-link.h b/include/hw/remote/mpqemu-link.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/remote/mpqemu-link.h
+++ b/include/hw/remote/mpqemu-link.h
@@ -XXX,XX +XXX,XX @@ typedef enum {
     MPQEMU_CMD_RET,
     MPQEMU_CMD_PCI_CFGWRITE,
     MPQEMU_CMD_PCI_CFGREAD,
+    MPQEMU_CMD_BAR_WRITE,
+    MPQEMU_CMD_BAR_READ,
     MPQEMU_CMD_MAX,
 } MPQemuCmd;
 
@@ -XXX,XX +XXX,XX @@ typedef struct {
     int len;
 } PciConfDataMsg;
 
+typedef struct {
+    hwaddr addr;
+    uint64_t val;
+    unsigned size;
+    bool memory;
+} BarAccessMsg;
+
 /**
  * MPQemuMsg:
  * @cmd: The remote command
@@ -XXX,XX +XXX,XX @@ typedef struct {
         uint64_t u64;
         PciConfDataMsg pci_conf_data;
         SyncSysmemMsg sync_sysmem;
+        BarAccessMsg bar_access;
     } data;
 
     int fds[REMOTE_MAX_FDS];
diff --git a/include/hw/remote/proxy.h b/include/hw/remote/proxy.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/remote/proxy.h
+++ b/include/hw/remote/proxy.h
@@ -XXX,XX +XXX,XX @@
 #define TYPE_PCI_PROXY_DEV "x-pci-proxy-dev"
 OBJECT_DECLARE_SIMPLE_TYPE(PCIProxyDev, PCI_PROXY_DEV)
 
+typedef struct ProxyMemoryRegion {
+    PCIProxyDev *dev;
+    MemoryRegion mr;
+    bool memory;
+    bool present;
+    uint8_t type;
+} ProxyMemoryRegion;
+
 struct PCIProxyDev {
     PCIDevice parent_dev;
     char *fd;
@@ -XXX,XX +XXX,XX @@ struct PCIProxyDev {
     QemuMutex io_mutex;
     QIOChannel *ioc;
     Error *migration_blocker;
+    ProxyMemoryRegion region[PCI_NUM_REGIONS];
 };
 
 #endif /* PROXY_H */
diff --git a/hw/remote/message.c b/hw/remote/message.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/message.c
+++ b/hw/remote/message.c
@@ -XXX,XX +XXX,XX @@
 #include "qapi/error.h"
 #include "sysemu/runstate.h"
 #include "hw/pci/pci.h"
+#include "exec/memattrs.h"
 
 static void process_config_write(QIOChannel *ioc, PCIDevice *dev,
                                  MPQemuMsg *msg, Error **errp);
 static void process_config_read(QIOChannel *ioc, PCIDevice *dev,
                                 MPQemuMsg *msg, Error **errp);
+static void process_bar_write(QIOChannel *ioc, MPQemuMsg *msg, Error **errp);
+static void process_bar_read(QIOChannel *ioc, MPQemuMsg *msg, Error **errp);
 
 void coroutine_fn mpqemu_remote_msg_loop_co(void *data)
 {
@@ -XXX,XX +XXX,XX @@ void coroutine_fn mpqemu_remote_msg_loop_co(void *data)
         case MPQEMU_CMD_PCI_CFGREAD:
             process_config_read(com->ioc, pci_dev, &msg, &local_err);
             break;
+        case MPQEMU_CMD_BAR_WRITE:
+            process_bar_write(com->ioc, &msg, &local_err);
+            break;
+        case MPQEMU_CMD_BAR_READ:
+            process_bar_read(com->ioc, &msg, &local_err);
+            break;
         default:
             error_setg(&local_err,
                        "Unknown command (%d) received for device %s"
@@ -XXX,XX +XXX,XX @@ static void process_config_read(QIOChannel *ioc, PCIDevice *dev,
                       getpid());
     }
 }
+
+static void process_bar_write(QIOChannel *ioc, MPQemuMsg *msg, Error **errp)
+{
+    ERRP_GUARD();
+    BarAccessMsg *bar_access = &msg->data.bar_access;
+    AddressSpace *as =
+        bar_access->memory ? &address_space_memory : &address_space_io;
+    MPQemuMsg ret = { 0 };
+    MemTxResult res;
+    uint64_t val;
+
+    if (!is_power_of_2(bar_access->size) ||
+       (bar_access->size > sizeof(uint64_t))) {
+        ret.data.u64 = UINT64_MAX;
+        goto fail;
+    }
+
+    val = cpu_to_le64(bar_access->val);
+
+    res = address_space_rw(as, bar_access->addr, MEMTXATTRS_UNSPECIFIED,
+                           (void *)&val, bar_access->size, true);
+
+    if (res != MEMTX_OK) {
+        error_setg(errp, "Bad address %"PRIx64" for mem write, pid "FMT_pid".",
+                   bar_access->addr, getpid());
+        ret.data.u64 = -1;
+    }
+
+fail:
+    ret.cmd = MPQEMU_CMD_RET;
+    ret.size = sizeof(ret.data.u64);
+
+    if (!mpqemu_msg_send(&ret, ioc, NULL)) {
+        error_prepend(errp, "Error returning code to proxy, pid "FMT_pid": ",
+                      getpid());
+    }
+}
+
+static void process_bar_read(QIOChannel *ioc, MPQemuMsg *msg, Error **errp)
+{
+    ERRP_GUARD();
+    BarAccessMsg *bar_access = &msg->data.bar_access;
+    MPQemuMsg ret = { 0 };
+    AddressSpace *as;
+    MemTxResult res;
+    uint64_t val = 0;
+
+    as = bar_access->memory ? &address_space_memory : &address_space_io;
+
+    if (!is_power_of_2(bar_access->size) ||
+       (bar_access->size > sizeof(uint64_t))) {
+        val = UINT64_MAX;
+        goto fail;
+    }
+
+    res = address_space_rw(as, bar_access->addr, MEMTXATTRS_UNSPECIFIED,
+                           (void *)&val, bar_access->size, false);
+
+    if (res != MEMTX_OK) {
+        error_setg(errp, "Bad address %"PRIx64" for mem read, pid "FMT_pid".",
+                   bar_access->addr, getpid());
+        val = UINT64_MAX;
+    }
+
+fail:
+    ret.cmd = MPQEMU_CMD_RET;
+    ret.data.u64 = le64_to_cpu(val);
+    ret.size = sizeof(ret.data.u64);
+
+    if (!mpqemu_msg_send(&ret, ioc, NULL)) {
+        error_prepend(errp, "Error returning code to proxy, pid "FMT_pid": ",
+                      getpid());
+    }
+}
diff --git a/hw/remote/mpqemu-link.c b/hw/remote/mpqemu-link.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/mpqemu-link.c
+++ b/hw/remote/mpqemu-link.c
@@ -XXX,XX +XXX,XX @@ bool mpqemu_msg_valid(MPQemuMsg *msg)
             return false;
         }
         break;
+    case MPQEMU_CMD_BAR_WRITE:
+    case MPQEMU_CMD_BAR_READ:
+        if ((msg->size != sizeof(BarAccessMsg)) || (msg->num_fds != 0)) {
+            return false;
+        }
+        break;
     default:
         break;
     }
diff --git a/hw/remote/proxy.c b/hw/remote/proxy.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/proxy.c
+++ b/hw/remote/proxy.c
@@ -XXX,XX +XXX,XX @@ static void pci_proxy_dev_register_types(void)
 }
 
 type_init(pci_proxy_dev_register_types)
+
+static void send_bar_access_msg(PCIProxyDev *pdev, MemoryRegion *mr,
+                                bool write, hwaddr addr, uint64_t *val,
+                                unsigned size, bool memory)
+{
+    MPQemuMsg msg = { 0 };
+    long ret = -EINVAL;
+    Error *local_err = NULL;
+
+    msg.size = sizeof(BarAccessMsg);
+    msg.data.bar_access.addr = mr->addr + addr;
+    msg.data.bar_access.size = size;
+    msg.data.bar_access.memory = memory;
+
+    if (write) {
+        msg.cmd = MPQEMU_CMD_BAR_WRITE;
+        msg.data.bar_access.val = *val;
+    } else {
+        msg.cmd = MPQEMU_CMD_BAR_READ;
+    }
+
+    ret = mpqemu_msg_send_and_await_reply(&msg, pdev, &local_err);
+    if (local_err) {
+        error_report_err(local_err);
+    }
+
+    if (!write) {
+        *val = ret;
+    }
+}
+
+static void proxy_bar_write(void *opaque, hwaddr addr, uint64_t val,
+                            unsigned size)
+{
+    ProxyMemoryRegion *pmr = opaque;
+
+    send_bar_access_msg(pmr->dev, &pmr->mr, true, addr, &val, size,
+                        pmr->memory);
+}
+
+static uint64_t proxy_bar_read(void *opaque, hwaddr addr, unsigned size)
+{
+    ProxyMemoryRegion *pmr = opaque;
+    uint64_t val;
+
+    send_bar_access_msg(pmr->dev, &pmr->mr, false, addr, &val, size,
+                        pmr->memory);
+
+    return val;
+}
+
+const MemoryRegionOps proxy_mr_ops = {
+    .read = proxy_bar_read,
+    .write = proxy_bar_write,
+    .endianness = DEVICE_NATIVE_ENDIAN,
+    .impl = {
+        .min_access_size = 1,
+        .max_access_size = 8,
+    },
+};
-- 
2.29.2

From: Jagannathan Raman <jag.raman@oracle.com>

Add ProxyMemoryListener object which is used to keep the view of the RAM
in sync between QEMU and remote process.
A MemoryListener is registered for system-memory AddressSpace. The
listener sends SYNC_SYSMEM message to the remote process when memory
listener commits the changes to memory, the remote process receives
the message and processes it in the handler for SYNC_SYSMEM message.

Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 04fe4e6a9ca90d4f11ab6f59be7652f5b086a071.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 MAINTAINERS                               |   2 +
 include/hw/remote/proxy-memory-listener.h |  28 +++
 include/hw/remote/proxy.h                 |   2 +
 hw/remote/message.c                       |   4 +
 hw/remote/proxy-memory-listener.c         | 227 ++++++++++++++++++++++
 hw/remote/proxy.c                         |   6 +
 hw/remote/meson.build                     |   1 +
 7 files changed, 270 insertions(+)
 create mode 100644 include/hw/remote/proxy-memory-listener.h
 create mode 100644 hw/remote/proxy-memory-listener.c

diff --git a/MAINTAINERS b/MAINTAINERS
index XXXXXXX..XXXXXXX 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -XXX,XX +XXX,XX @@ F: include/hw/remote/memory.h
 F: hw/remote/memory.c
 F: hw/remote/proxy.c
 F: include/hw/remote/proxy.h
+F: hw/remote/proxy-memory-listener.c
+F: include/hw/remote/proxy-memory-listener.h
 
 Build and test automation
 -------------------------
diff --git a/include/hw/remote/proxy-memory-listener.h b/include/hw/remote/proxy-memory-listener.h
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/include/hw/remote/proxy-memory-listener.h
@@ -XXX,XX +XXX,XX @@
+/*
+ * Copyright © 2018, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef PROXY_MEMORY_LISTENER_H
+#define PROXY_MEMORY_LISTENER_H
+
+#include "exec/memory.h"
+#include "io/channel.h"
+
+typedef struct ProxyMemoryListener {
+    MemoryListener listener;
+
+    int n_mr_sections;
+    MemoryRegionSection *mr_sections;
+
+    QIOChannel *ioc;
+} ProxyMemoryListener;
+
+void proxy_memory_listener_configure(ProxyMemoryListener *proxy_listener,
+                                     QIOChannel *ioc);
+void proxy_memory_listener_deconfigure(ProxyMemoryListener *proxy_listener);
+
+#endif
diff --git a/include/hw/remote/proxy.h b/include/hw/remote/proxy.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/remote/proxy.h
+++ b/include/hw/remote/proxy.h
@@ -XXX,XX +XXX,XX @@
 
 #include "hw/pci/pci.h"
 #include "io/channel.h"
+#include "hw/remote/proxy-memory-listener.h"
 
 #define TYPE_PCI_PROXY_DEV "x-pci-proxy-dev"
 OBJECT_DECLARE_SIMPLE_TYPE(PCIProxyDev, PCI_PROXY_DEV)
@@ -XXX,XX +XXX,XX @@ struct PCIProxyDev {
     QemuMutex io_mutex;
     QIOChannel *ioc;
     Error *migration_blocker;
+    ProxyMemoryListener proxy_listener;
     ProxyMemoryRegion region[PCI_NUM_REGIONS];
 };
 
diff --git a/hw/remote/message.c b/hw/remote/message.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/message.c
+++ b/hw/remote/message.c
@@ -XXX,XX +XXX,XX @@
 #include "sysemu/runstate.h"
 #include "hw/pci/pci.h"
 #include "exec/memattrs.h"
+#include "hw/remote/memory.h"
 
 static void process_config_write(QIOChannel *ioc, PCIDevice *dev,
                                  MPQemuMsg *msg, Error **errp);
@@ -XXX,XX +XXX,XX @@ void coroutine_fn mpqemu_remote_msg_loop_co(void *data)
         case MPQEMU_CMD_BAR_READ:
             process_bar_read(com->ioc, &msg, &local_err);
             break;
+        case MPQEMU_CMD_SYNC_SYSMEM:
+            remote_sysmem_reconfig(&msg, &local_err);
+            break;
         default:
             error_setg(&local_err,
                        "Unknown command (%d) received for device %s"
diff --git a/hw/remote/proxy-memory-listener.c b/hw/remote/proxy-memory-listener.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/remote/proxy-memory-listener.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * Copyright © 2018, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+
+#include "qemu/compiler.h"
+#include "qemu/int128.h"
+#include "qemu/range.h"
+#include "exec/memory.h"
+#include "exec/cpu-common.h"
+#include "cpu.h"
+#include "exec/ram_addr.h"
+#include "exec/address-spaces.h"
+#include "qapi/error.h"
+#include "hw/remote/mpqemu-link.h"
+#include "hw/remote/proxy-memory-listener.h"
+
+/*
+ * TODO: get_fd_from_hostaddr(), proxy_mrs_can_merge() and
+ * proxy_memory_listener_commit() defined below perform tasks similar to the
+ * functions defined in vhost-user.c. These functions are good candidates
+ * for refactoring.
+ *
+ */
+
+static void proxy_memory_listener_reset(MemoryListener *listener)
+{
+    ProxyMemoryListener *proxy_listener = container_of(listener,
+                                                       ProxyMemoryListener,
+                                                       listener);
+    int mrs;
+
+    for (mrs = 0; mrs < proxy_listener->n_mr_sections; mrs++) {
+        memory_region_unref(proxy_listener->mr_sections[mrs].mr);
+    }
+
+    g_free(proxy_listener->mr_sections);
+    proxy_listener->mr_sections = NULL;
+    proxy_listener->n_mr_sections = 0;
+}
+
+static int get_fd_from_hostaddr(uint64_t host, ram_addr_t *offset)
+{
+    MemoryRegion *mr;
+    ram_addr_t off;
+
+    /**
+     * Assumes that the host address is a valid address as it's
+     * coming from the MemoryListener system. In the case host
+     * address is not valid, the following call would return
+     * the default subregion of "system_memory" region, and
+     * not NULL. So it's not possible to check for NULL here.
+     */
+    mr = memory_region_from_host((void *)(uintptr_t)host, &off);
+
+    if (offset) {
+        *offset = off;
+    }
+
+    return memory_region_get_fd(mr);
+}
+
+static bool proxy_mrs_can_merge(uint64_t host, uint64_t prev_host, size_t size)
+{
+    if (((prev_host + size) != host)) {
+        return false;
+    }
+
+    if (get_fd_from_hostaddr(host, NULL) !=
+            get_fd_from_hostaddr(prev_host, NULL)) {
+        return false;
+    }
+
+    return true;
+}
+
+static bool try_merge(ProxyMemoryListener *proxy_listener,
+                      MemoryRegionSection *section)
+{
+    uint64_t mrs_size, mrs_gpa, mrs_page;
+    MemoryRegionSection *prev_sec;
+    bool merged = false;
+    uintptr_t mrs_host;
+    RAMBlock *mrs_rb;
+
+    if (!proxy_listener->n_mr_sections) {
+        return false;
+    }
+
+    mrs_rb = section->mr->ram_block;
+    mrs_page = (uint64_t)qemu_ram_pagesize(mrs_rb);
+    mrs_size = int128_get64(section->size);
+    mrs_gpa = section->offset_within_address_space;
+    mrs_host = (uintptr_t)memory_region_get_ram_ptr(section->mr) +
+               section->offset_within_region;
+
+    if (get_fd_from_hostaddr(mrs_host, NULL) < 0) {
+        return true;
+    }
+
+    mrs_host = mrs_host & ~(mrs_page - 1);
+    mrs_gpa = mrs_gpa & ~(mrs_page - 1);
+    mrs_size = ROUND_UP(mrs_size, mrs_page);
+
+    prev_sec = proxy_listener->mr_sections +
+               (proxy_listener->n_mr_sections - 1);
+    uint64_t prev_gpa_start = prev_sec->offset_within_address_space;
+    uint64_t prev_size = int128_get64(prev_sec->size);
+    uint64_t prev_gpa_end   = range_get_last(prev_gpa_start, prev_size);
+    uint64_t prev_host_start =
+        (uintptr_t)memory_region_get_ram_ptr(prev_sec->mr) +
+        prev_sec->offset_within_region;
+    uint64_t prev_host_end = range_get_last(prev_host_start, prev_size);
+
+    if (mrs_gpa <= (prev_gpa_end + 1)) {
+        g_assert(mrs_gpa > prev_gpa_start);
+
+        if ((section->mr == prev_sec->mr) &&
+            proxy_mrs_can_merge(mrs_host, prev_host_start,
+                                (mrs_gpa - prev_gpa_start))) {
+            uint64_t max_end = MAX(prev_host_end, mrs_host + mrs_size);
+            merged = true;
+            prev_sec->offset_within_address_space =
+                MIN(prev_gpa_start, mrs_gpa);
+            prev_sec->offset_within_region =
+                MIN(prev_host_start, mrs_host) -
+                (uintptr_t)memory_region_get_ram_ptr(prev_sec->mr);
+            prev_sec->size = int128_make64(max_end - MIN(prev_host_start,
+                                                         mrs_host));
+        }
+    }
+
+    return merged;
+}
+
+static void proxy_memory_listener_region_addnop(MemoryListener *listener,
+                                                MemoryRegionSection *section)
+{
+    ProxyMemoryListener *proxy_listener = container_of(listener,
+                                                       ProxyMemoryListener,
+                                                       listener);
+
+    if (!memory_region_is_ram(section->mr) ||
+            memory_region_is_rom(section->mr)) {
+        return;
+    }
+
+    if (try_merge(proxy_listener, section)) {
+        return;
+    }
+
+    ++proxy_listener->n_mr_sections;
+    proxy_listener->mr_sections = g_renew(MemoryRegionSection,
+                                          proxy_listener->mr_sections,
+                                          proxy_listener->n_mr_sections);
+    proxy_listener->mr_sections[proxy_listener->n_mr_sections - 1] = *section;
+    proxy_listener->mr_sections[proxy_listener->n_mr_sections - 1].fv = NULL;
+    memory_region_ref(section->mr);
+}
+
+static void proxy_memory_listener_commit(MemoryListener *listener)
+{
+    ProxyMemoryListener *proxy_listener = container_of(listener,
+                                                       ProxyMemoryListener,
+                                                       listener);
+    MPQemuMsg msg;
+    MemoryRegionSection *section;
+    ram_addr_t offset;
+    uintptr_t host_addr;
+    int region;
+    Error *local_err = NULL;
+
+    memset(&msg, 0, sizeof(MPQemuMsg));
+
+    msg.cmd = MPQEMU_CMD_SYNC_SYSMEM;
+    msg.num_fds = proxy_listener->n_mr_sections;
+    msg.size = sizeof(SyncSysmemMsg);
+    if (msg.num_fds > REMOTE_MAX_FDS) {
+        error_report("Number of fds is more than %d", REMOTE_MAX_FDS);
+        return;
+    }
+
+    for (region = 0; region < proxy_listener->n_mr_sections; region++) {
+        section = &proxy_listener->mr_sections[region];
+        msg.data.sync_sysmem.gpas[region] =
+            section->offset_within_address_space;
+        msg.data.sync_sysmem.sizes[region] = int128_get64(section->size);
+        host_addr = (uintptr_t)memory_region_get_ram_ptr(section->mr) +
+                    section->offset_within_region;
+        msg.fds[region] = get_fd_from_hostaddr(host_addr, &offset);
+        msg.data.sync_sysmem.offsets[region] = offset;
+    }
+    if (!mpqemu_msg_send(&msg, proxy_listener->ioc, &local_err)) {
+        error_report_err(local_err);
+    }
+}
+
+void proxy_memory_listener_deconfigure(ProxyMemoryListener *proxy_listener)
+{
+    memory_listener_unregister(&proxy_listener->listener);
+
+    proxy_memory_listener_reset(&proxy_listener->listener);
+}
+
+void proxy_memory_listener_configure(ProxyMemoryListener *proxy_listener,
+                                     QIOChannel *ioc)
+{
+    proxy_listener->n_mr_sections = 0;
+    proxy_listener->mr_sections = NULL;
+
+    proxy_listener->ioc = ioc;
+
+    proxy_listener->listener.begin = proxy_memory_listener_reset;
+    proxy_listener->listener.commit = proxy_memory_listener_commit;
+    proxy_listener->listener.region_add = proxy_memory_listener_region_addnop;
+    proxy_listener->listener.region_nop = proxy_memory_listener_region_addnop;
+    proxy_listener->listener.priority = 10;
+
+    memory_listener_register(&proxy_listener->listener,
+                             &address_space_memory);
+}
diff --git a/hw/remote/proxy.c b/hw/remote/proxy.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/proxy.c
+++ b/hw/remote/proxy.c
@@ -XXX,XX +XXX,XX @@
 #include "qemu/sockets.h"
 #include "hw/remote/mpqemu-link.h"
 #include "qemu/error-report.h"
+#include "hw/remote/proxy-memory-listener.h"
+#include "qom/object.h"
 
 static void pci_proxy_dev_realize(PCIDevice *device, Error **errp)
 {
@@ -XXX,XX +XXX,XX @@ static void pci_proxy_dev_realize(PCIDevice *device, Error **errp)
 
     qemu_mutex_init(&dev->io_mutex);
     qio_channel_set_blocking(dev->ioc, true, NULL);
+
+    proxy_memory_listener_configure(&dev->proxy_listener, dev->ioc);
 }
 
 static void pci_proxy_dev_exit(PCIDevice *pdev)
@@ -XXX,XX +XXX,XX @@ static void pci_proxy_dev_exit(PCIDevice *pdev)
     migrate_del_blocker(dev->migration_blocker);
 
     error_free(dev->migration_blocker);
+
+    proxy_memory_listener_deconfigure(&dev->proxy_listener);
 }
 
 static void config_op_send(PCIProxyDev *pdev, uint32_t addr, uint32_t *val,
diff --git a/hw/remote/meson.build b/hw/remote/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/meson.build
+++ b/hw/remote/meson.build
@@ -XXX,XX +XXX,XX @@ remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('remote-obj.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('proxy.c'))
 
 specific_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('memory.c'))
+specific_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('proxy-memory-listener.c'))
 
 softmmu_ss.add_all(when: 'CONFIG_MULTIPROCESS', if_true: remote_ss)
-- 
2.29.2

From: Jagannathan Raman <jag.raman@oracle.com>

IOHUB object is added to manage PCI IRQs. It uses KVM_IRQFD
ioctl to create irqfd to injecting PCI interrupts to the guest.
IOHUB object forwards the irqfd to the remote process. Remote process
uses this fd to directly send interrupts to the guest, bypassing QEMU.

Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 51d5c3d54e28a68b002e3875c59599c9f5a424a1.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 MAINTAINERS                     |   2 +
 include/hw/pci/pci_ids.h        |   3 +
 include/hw/remote/iohub.h       |  42 +++++++++++
 include/hw/remote/machine.h     |   2 +
 include/hw/remote/mpqemu-link.h |   1 +
 include/hw/remote/proxy.h       |   4 ++
 hw/remote/iohub.c               | 119 ++++++++++++++++++++++++++++++++
 hw/remote/machine.c             |  10 +++
 hw/remote/message.c             |   4 ++
 hw/remote/mpqemu-link.c         |   5 ++
 hw/remote/proxy.c               |  56 +++++++++++++++
 hw/remote/meson.build           |   1 +
 12 files changed, 249 insertions(+)
 create mode 100644 include/hw/remote/iohub.h
 create mode 100644 hw/remote/iohub.c

diff --git a/MAINTAINERS b/MAINTAINERS
index XXXXXXX..XXXXXXX 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -XXX,XX +XXX,XX @@ F: hw/remote/proxy.c
 F: include/hw/remote/proxy.h
 F: hw/remote/proxy-memory-listener.c
 F: include/hw/remote/proxy-memory-listener.h
+F: hw/remote/iohub.c
+F: include/hw/remote/iohub.h
 
 Build and test automation
 -------------------------
diff --git a/include/hw/pci/pci_ids.h b/include/hw/pci/pci_ids.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/pci/pci_ids.h
+++ b/include/hw/pci/pci_ids.h
@@ -XXX,XX +XXX,XX @@
 #define PCI_DEVICE_ID_SUN_SIMBA          0x5000
 #define PCI_DEVICE_ID_SUN_SABRE          0xa000
 
+#define PCI_VENDOR_ID_ORACLE             0x108e
+#define PCI_DEVICE_ID_REMOTE_IOHUB       0xb000
+
 #define PCI_VENDOR_ID_CMD                0x1095
 #define PCI_DEVICE_ID_CMD_646            0x0646
 
diff --git a/include/hw/remote/iohub.h b/include/hw/remote/iohub.h
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/include/hw/remote/iohub.h
@@ -XXX,XX +XXX,XX @@
+/*
+ * IO Hub for remote device
+ *
+ * Copyright © 2018, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#ifndef REMOTE_IOHUB_H
+#define REMOTE_IOHUB_H
+
+#include "hw/pci/pci.h"
+#include "qemu/event_notifier.h"
+#include "qemu/thread-posix.h"
+#include "hw/remote/mpqemu-link.h"
+
+#define REMOTE_IOHUB_NB_PIRQS    PCI_DEVFN_MAX
+
+typedef struct ResampleToken {
+    void *iohub;
+    int pirq;
+} ResampleToken;
+
+typedef struct RemoteIOHubState {
+    PCIDevice d;
+    EventNotifier irqfds[REMOTE_IOHUB_NB_PIRQS];
+    EventNotifier resamplefds[REMOTE_IOHUB_NB_PIRQS];
+    unsigned int irq_level[REMOTE_IOHUB_NB_PIRQS];
+    ResampleToken token[REMOTE_IOHUB_NB_PIRQS];
+    QemuMutex irq_level_lock[REMOTE_IOHUB_NB_PIRQS];
+} RemoteIOHubState;
+
+int remote_iohub_map_irq(PCIDevice *pci_dev, int intx);
+void remote_iohub_set_irq(void *opaque, int pirq, int level);
+void process_set_irqfd_msg(PCIDevice *pci_dev, MPQemuMsg *msg);
+
+void remote_iohub_init(RemoteIOHubState *iohub);
+void remote_iohub_finalize(RemoteIOHubState *iohub);
+
+#endif
diff --git a/include/hw/remote/machine.h b/include/hw/remote/machine.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/remote/machine.h
+++ b/include/hw/remote/machine.h
@@ -XXX,XX +XXX,XX @@
 #include "hw/boards.h"
 #include "hw/pci-host/remote.h"
 #include "io/channel.h"
+#include "hw/remote/iohub.h"
 
 struct RemoteMachineState {
     MachineState parent_obj;
 
     RemotePCIHost *host;
+    RemoteIOHubState iohub;
 };
 
 /* Used to pass to co-routine device and ioc. */
diff --git a/include/hw/remote/mpqemu-link.h b/include/hw/remote/mpqemu-link.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/remote/mpqemu-link.h
+++ b/include/hw/remote/mpqemu-link.h
@@ -XXX,XX +XXX,XX @@ typedef enum {
     MPQEMU_CMD_PCI_CFGREAD,
     MPQEMU_CMD_BAR_WRITE,
     MPQEMU_CMD_BAR_READ,
+    MPQEMU_CMD_SET_IRQFD,
     MPQEMU_CMD_MAX,
 } MPQemuCmd;
 
diff --git a/include/hw/remote/proxy.h b/include/hw/remote/proxy.h
index XXXXXXX..XXXXXXX 100644
--- a/include/hw/remote/proxy.h
+++ b/include/hw/remote/proxy.h
@@ -XXX,XX +XXX,XX @@
 #include "hw/pci/pci.h"
 #include "io/channel.h"
 #include "hw/remote/proxy-memory-listener.h"
+#include "qemu/event_notifier.h"
 
 #define TYPE_PCI_PROXY_DEV "x-pci-proxy-dev"
 OBJECT_DECLARE_SIMPLE_TYPE(PCIProxyDev, PCI_PROXY_DEV)
@@ -XXX,XX +XXX,XX @@ struct PCIProxyDev {
     QIOChannel *ioc;
     Error *migration_blocker;
     ProxyMemoryListener proxy_listener;
+    int virq;
+    EventNotifier intr;
+    EventNotifier resample;
     ProxyMemoryRegion region[PCI_NUM_REGIONS];
 };
 
diff --git a/hw/remote/iohub.c b/hw/remote/iohub.c
new file mode 100644
index XXXXXXX..XXXXXXX
--- /dev/null
+++ b/hw/remote/iohub.c
@@ -XXX,XX +XXX,XX @@
+/*
+ * Remote IO Hub
+ *
+ * Copyright © 2018, 2021 Oracle and/or its affiliates.
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ *
+ */
+
+#include "qemu/osdep.h"
+#include "qemu-common.h"
+
+#include "hw/pci/pci.h"
+#include "hw/pci/pci_ids.h"
+#include "hw/pci/pci_bus.h"
+#include "qemu/thread.h"
+#include "hw/boards.h"
+#include "hw/remote/machine.h"
+#include "hw/remote/iohub.h"
+#include "qemu/main-loop.h"
+
+void remote_iohub_init(RemoteIOHubState *iohub)
+{
+    int pirq;
+
+    memset(&iohub->irqfds, 0, sizeof(iohub->irqfds));
+    memset(&iohub->resamplefds, 0, sizeof(iohub->resamplefds));
+
+    for (pirq = 0; pirq < REMOTE_IOHUB_NB_PIRQS; pirq++) {
+        qemu_mutex_init(&iohub->irq_level_lock[pirq]);
+        iohub->irq_level[pirq] = 0;
+        event_notifier_init_fd(&iohub->irqfds[pirq], -1);
+        event_notifier_init_fd(&iohub->resamplefds[pirq], -1);
+    }
+}
+
+void remote_iohub_finalize(RemoteIOHubState *iohub)
+{
+    int pirq;
+
+    for (pirq = 0; pirq < REMOTE_IOHUB_NB_PIRQS; pirq++) {
+        qemu_set_fd_handler(event_notifier_get_fd(&iohub->resamplefds[pirq]),
+                            NULL, NULL, NULL);
+        event_notifier_cleanup(&iohub->irqfds[pirq]);
+        event_notifier_cleanup(&iohub->resamplefds[pirq]);
+        qemu_mutex_destroy(&iohub->irq_level_lock[pirq]);
+    }
+}
+
+int remote_iohub_map_irq(PCIDevice *pci_dev, int intx)
+{
+    return pci_dev->devfn;
+}
+
+void remote_iohub_set_irq(void *opaque, int pirq, int level)
+{
+    RemoteIOHubState *iohub = opaque;
+
+    assert(pirq >= 0);
+    assert(pirq < PCI_DEVFN_MAX);
+
+    QEMU_LOCK_GUARD(&iohub->irq_level_lock[pirq]);
+
+    if (level) {
+        if (++iohub->irq_level[pirq] == 1) {
+            event_notifier_set(&iohub->irqfds[pirq]);
+        }
+    } else if (iohub->irq_level[pirq] > 0) {
+        iohub->irq_level[pirq]--;
+    }
+}
+
+static void intr_resample_handler(void *opaque)
+{
+    ResampleToken *token = opaque;
+    RemoteIOHubState *iohub = token->iohub;
+    int pirq, s;
+
+    pirq = token->pirq;
+
+    s = event_notifier_test_and_clear(&iohub->resamplefds[pirq]);
+
+    assert(s >= 0);
+
+    QEMU_LOCK_GUARD(&iohub->irq_level_lock[pirq]);
+
+    if (iohub->irq_level[pirq]) {
+        event_notifier_set(&iohub->irqfds[pirq]);
+    }
+}
+
+void process_set_irqfd_msg(PCIDevice *pci_dev, MPQemuMsg *msg)
+{
+    RemoteMachineState *machine = REMOTE_MACHINE(current_machine);
+    RemoteIOHubState *iohub = &machine->iohub;
+    int pirq, intx;
+
+    intx = pci_get_byte(pci_dev->config + PCI_INTERRUPT_PIN) - 1;
+
+    pirq = remote_iohub_map_irq(pci_dev, intx);
+
+    if (event_notifier_get_fd(&iohub->irqfds[pirq]) != -1) {
+        qemu_set_fd_handler(event_notifier_get_fd(&iohub->resamplefds[pirq]),
+                            NULL, NULL, NULL);
+        event_notifier_cleanup(&iohub->irqfds[pirq]);
+        event_notifier_cleanup(&iohub->resamplefds[pirq]);
+        memset(&iohub->token[pirq], 0, sizeof(ResampleToken));
+    }
+
+    event_notifier_init_fd(&iohub->irqfds[pirq], msg->fds[0]);
+    event_notifier_init_fd(&iohub->resamplefds[pirq], msg->fds[1]);
+
+    iohub->token[pirq].iohub = iohub;
+    iohub->token[pirq].pirq = pirq;
+
+    qemu_set_fd_handler(msg->fds[1], intr_resample_handler, NULL,
+                        &iohub->token[pirq]);
+}
diff --git a/hw/remote/machine.c b/hw/remote/machine.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/machine.c
+++ b/hw/remote/machine.c
@@ -XXX,XX +XXX,XX @@
 #include "exec/address-spaces.h"
 #include "exec/memory.h"
 #include "qapi/error.h"
+#include "hw/pci/pci_host.h"
+#include "hw/remote/iohub.h"
 
 static void remote_machine_init(MachineState *machine)
 {
     MemoryRegion *system_memory, *system_io, *pci_memory;
     RemoteMachineState *s = REMOTE_MACHINE(machine);
     RemotePCIHost *rem_host;
+    PCIHostState *pci_host;
 
     system_memory = get_system_memory();
     system_io = get_system_io();
@@ -XXX,XX +XXX,XX @@ static void remote_machine_init(MachineState *machine)
     memory_region_add_subregion_overlap(system_memory, 0x0, pci_memory, -1);
 
     qdev_realize(DEVICE(rem_host), sysbus_get_default(), &error_fatal);
+
+    pci_host = PCI_HOST_BRIDGE(rem_host);
+
+    remote_iohub_init(&s->iohub);
+
+    pci_bus_irqs(pci_host->bus, remote_iohub_set_irq, remote_iohub_map_irq,
+                 &s->iohub, REMOTE_IOHUB_NB_PIRQS);
 }
 
 static void remote_machine_class_init(ObjectClass *oc, void *data)
diff --git a/hw/remote/message.c b/hw/remote/message.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/message.c
+++ b/hw/remote/message.c
@@ -XXX,XX +XXX,XX @@
 #include "hw/pci/pci.h"
 #include "exec/memattrs.h"
 #include "hw/remote/memory.h"
+#include "hw/remote/iohub.h"
 
 static void process_config_write(QIOChannel *ioc, PCIDevice *dev,
                                  MPQemuMsg *msg, Error **errp);
@@ -XXX,XX +XXX,XX @@ void coroutine_fn mpqemu_remote_msg_loop_co(void *data)
         case MPQEMU_CMD_SYNC_SYSMEM:
             remote_sysmem_reconfig(&msg, &local_err);
             break;
+        case MPQEMU_CMD_SET_IRQFD:
+            process_set_irqfd_msg(pci_dev, &msg);
+            break;
         default:
             error_setg(&local_err,
                        "Unknown command (%d) received for device %s"
diff --git a/hw/remote/mpqemu-link.c b/hw/remote/mpqemu-link.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/mpqemu-link.c
+++ b/hw/remote/mpqemu-link.c
@@ -XXX,XX +XXX,XX @@ bool mpqemu_msg_valid(MPQemuMsg *msg)
             return false;
         }
         break;
+    case MPQEMU_CMD_SET_IRQFD:
+        if (msg->size || (msg->num_fds != 2)) {
+            return false;
+        }
+        break;
     default:
         break;
     }
diff --git a/hw/remote/proxy.c b/hw/remote/proxy.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/proxy.c
+++ b/hw/remote/proxy.c
@@ -XXX,XX +XXX,XX @@
 #include "qemu/error-report.h"
 #include "hw/remote/proxy-memory-listener.h"
 #include "qom/object.h"
+#include "qemu/event_notifier.h"
+#include "sysemu/kvm.h"
+#include "util/event_notifier-posix.c"
+
+static void proxy_intx_update(PCIDevice *pci_dev)
+{
+    PCIProxyDev *dev = PCI_PROXY_DEV(pci_dev);
+    PCIINTxRoute route;
+    int pin = pci_get_byte(pci_dev->config + PCI_INTERRUPT_PIN) - 1;
+
+    if (dev->virq != -1) {
+        kvm_irqchip_remove_irqfd_notifier_gsi(kvm_state, &dev->intr, dev->virq);
+        dev->virq = -1;
+    }
+
+    route = pci_device_route_intx_to_irq(pci_dev, pin);
+
+    dev->virq = route.irq;
+
+    if (dev->virq != -1) {
+        kvm_irqchip_add_irqfd_notifier_gsi(kvm_state, &dev->intr,
+                                           &dev->resample, dev->virq);
+    }
+}
+
+static void setup_irqfd(PCIProxyDev *dev)
+{
+    PCIDevice *pci_dev = PCI_DEVICE(dev);
+    MPQemuMsg msg;
+    Error *local_err = NULL;
+
+    event_notifier_init(&dev->intr, 0);
+    event_notifier_init(&dev->resample, 0);
+
+    memset(&msg, 0, sizeof(MPQemuMsg));
+    msg.cmd = MPQEMU_CMD_SET_IRQFD;
+    msg.num_fds = 2;
+    msg.fds[0] = event_notifier_get_fd(&dev->intr);
+    msg.fds[1] = event_notifier_get_fd(&dev->resample);
+    msg.size = 0;
+
+    if (!mpqemu_msg_send(&msg, dev->ioc, &local_err)) {
+        error_report_err(local_err);
+    }
+
+    dev->virq = -1;
+
+    proxy_intx_update(pci_dev);
+
+    pci_device_set_intx_routing_notifier(pci_dev, proxy_intx_update);
+}
 
 static void pci_proxy_dev_realize(PCIDevice *device, Error **errp)
 {
@@ -XXX,XX +XXX,XX @@ static void pci_proxy_dev_realize(PCIDevice *device, Error **errp)
     qio_channel_set_blocking(dev->ioc, true, NULL);
 
     proxy_memory_listener_configure(&dev->proxy_listener, dev->ioc);
+
+    setup_irqfd(dev);
 }
 
 static void pci_proxy_dev_exit(PCIDevice *pdev)
@@ -XXX,XX +XXX,XX @@ static void pci_proxy_dev_exit(PCIDevice *pdev)
     error_free(dev->migration_blocker);
 
     proxy_memory_listener_deconfigure(&dev->proxy_listener);
+
+    event_notifier_cleanup(&dev->intr);
+    event_notifier_cleanup(&dev->resample);
 }
 
 static void config_op_send(PCIProxyDev *pdev, uint32_t addr, uint32_t *val,
diff --git a/hw/remote/meson.build b/hw/remote/meson.build
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/meson.build
+++ b/hw/remote/meson.build
@@ -XXX,XX +XXX,XX @@ remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('mpqemu-link.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('message.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('remote-obj.c'))
 remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('proxy.c'))
+remote_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('iohub.c'))
 
 specific_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('memory.c'))
 specific_ss.add(when: 'CONFIG_MULTIPROCESS', if_true: files('proxy-memory-listener.c'))
-- 
2.29.2

From: Jagannathan Raman <jag.raman@oracle.com>

Retrieve PCI configuration info about the remote device and
configure the Proxy PCI object based on the returned information

Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 85ee367bbb993aa23699b44cfedd83b4ea6d5221.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 hw/remote/proxy.c | 84 +++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 84 insertions(+)

diff --git a/hw/remote/proxy.c b/hw/remote/proxy.c
index XXXXXXX..XXXXXXX 100644
--- a/hw/remote/proxy.c
+++ b/hw/remote/proxy.c
@@ -XXX,XX +XXX,XX @@
 #include "sysemu/kvm.h"
 #include "util/event_notifier-posix.c"
 
+static void probe_pci_info(PCIDevice *dev, Error **errp);
+
 static void proxy_intx_update(PCIDevice *pci_dev)
 {
     PCIProxyDev *dev = PCI_PROXY_DEV(pci_dev);
@@ -XXX,XX +XXX,XX @@ static void pci_proxy_dev_realize(PCIDevice *device, Error **errp)
 {
     ERRP_GUARD();
     PCIProxyDev *dev = PCI_PROXY_DEV(device);
+    uint8_t *pci_conf = device->config;
     int fd;
 
     if (!dev->fd) {
@@ -XXX,XX +XXX,XX @@ static void pci_proxy_dev_realize(PCIDevice *device, Error **errp)
     qemu_mutex_init(&dev->io_mutex);
     qio_channel_set_blocking(dev->ioc, true, NULL);
 
+    pci_conf[PCI_LATENCY_TIMER] = 0xff;
+    pci_conf[PCI_INTERRUPT_PIN] = 0x01;
+
     proxy_memory_listener_configure(&dev->proxy_listener, dev->ioc);
 
     setup_irqfd(dev);
+
+    probe_pci_info(PCI_DEVICE(dev), errp);
 }
 
 static void pci_proxy_dev_exit(PCIDevice *pdev)
@@ -XXX,XX +XXX,XX @@ const MemoryRegionOps proxy_mr_ops = {
         .max_access_size = 8,
     },
 };
+
+static void probe_pci_info(PCIDevice *dev, Error **errp)
+{
+    PCIDeviceClass *pc = PCI_DEVICE_GET_CLASS(dev);
+    uint32_t orig_val, new_val, base_class, val;
+    PCIProxyDev *pdev = PCI_PROXY_DEV(dev);
+    DeviceClass *dc = DEVICE_CLASS(pc);
+    uint8_t type;
+    int i, size;
+
+    config_op_send(pdev, PCI_VENDOR_ID, &val, 2, MPQEMU_CMD_PCI_CFGREAD);
+    pc->vendor_id = (uint16_t)val;
+
+    config_op_send(pdev, PCI_DEVICE_ID, &val, 2, MPQEMU_CMD_PCI_CFGREAD);
+    pc->device_id = (uint16_t)val;
+
+    config_op_send(pdev, PCI_CLASS_DEVICE, &val, 2, MPQEMU_CMD_PCI_CFGREAD);
+    pc->class_id = (uint16_t)val;
+
+    config_op_send(pdev, PCI_SUBSYSTEM_ID, &val, 2, MPQEMU_CMD_PCI_CFGREAD);
+    pc->subsystem_id = (uint16_t)val;
+
+    base_class = pc->class_id >> 4;
+    switch (base_class) {
+    case PCI_BASE_CLASS_BRIDGE:
+        set_bit(DEVICE_CATEGORY_BRIDGE, dc->categories);
+        break;
+    case PCI_BASE_CLASS_STORAGE:
+        set_bit(DEVICE_CATEGORY_STORAGE, dc->categories);
+        break;
+    case PCI_BASE_CLASS_NETWORK:
+        set_bit(DEVICE_CATEGORY_NETWORK, dc->categories);
+        break;
+    case PCI_BASE_CLASS_INPUT:
+        set_bit(DEVICE_CATEGORY_INPUT, dc->categories);
+        break;
+    case PCI_BASE_CLASS_DISPLAY:
+        set_bit(DEVICE_CATEGORY_DISPLAY, dc->categories);
+        break;
+    case PCI_BASE_CLASS_PROCESSOR:
+        set_bit(DEVICE_CATEGORY_CPU, dc->categories);
+        break;
+    default:
+        set_bit(DEVICE_CATEGORY_MISC, dc->categories);
+        break;
+    }
+
+    for (i = 0; i < PCI_NUM_REGIONS; i++) {
+        config_op_send(pdev, PCI_BASE_ADDRESS_0 + (4 * i), &orig_val, 4,
+                       MPQEMU_CMD_PCI_CFGREAD);
+        new_val = 0xffffffff;
+        config_op_send(pdev, PCI_BASE_ADDRESS_0 + (4 * i), &new_val, 4,
+                       MPQEMU_CMD_PCI_CFGWRITE);
+        config_op_send(pdev, PCI_BASE_ADDRESS_0 + (4 * i), &new_val, 4,
+                       MPQEMU_CMD_PCI_CFGREAD);
+        size = (~(new_val & 0xFFFFFFF0)) + 1;
+        config_op_send(pdev, PCI_BASE_ADDRESS_0 + (4 * i), &orig_val, 4,
+                       MPQEMU_CMD_PCI_CFGWRITE);
+        type = (new_val & 0x1) ?
+                   PCI_BASE_ADDRESS_SPACE_IO : PCI_BASE_ADDRESS_SPACE_MEMORY;
+
+        if (size) {
+            g_autofree char *name;
+            pdev->region[i].dev = pdev;
+            pdev->region[i].present = true;
+            if (type == PCI_BASE_ADDRESS_SPACE_MEMORY) {
+                pdev->region[i].memory = true;
+            }
+            name = g_strdup_printf("bar-region-%d", i);
+            memory_region_init_io(&pdev->region[i].mr, OBJECT(pdev),
+                                  &proxy_mr_ops, &pdev->region[i],
+                                  name, size);
+            pci_register_bar(dev, i, type, &pdev->region[i].mr);
+        }
+    }
+}
-- 
2.29.2

From: Elena Ufimtseva <elena.ufimtseva@oracle.com>

Perform device reset in the remote process when QEMU performs
device reset. This is required to reset the internal state
(like registers, etc...) of emulated devices

Signed-off-by: Elena Ufimtseva <elena.ufimtseva@oracle.com>
Signed-off-by: John G Johnson <john.g.johnson@oracle.com>
Signed-off-by: Jagannathan Raman <jag.raman@oracle.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 7cb220a51f565dc0817bd76e2f540e89c2d2b850.1611938319.git.jag.raman@oracle.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 include/hw/remote/mpqemu-link.h |  1 +
 hw/remote/message.c             | 22 ++++++++++++++++++++++
 hw/remote/proxy.c               | 19 +++++++++++++++++++
 3 files changed, 42 insertions(+)

From: "Denis V. Lunev" <den@openvz.org>

Original specification says that l1 table size if 64 * l1_size, which
is obviously wrong. The size of the l1 entry is 64 _bits_, not bytes.
Thus 64 is to be replaces with 8 as specification says about bytes.

There is also minor tweak, field name is renamed from l1 to l1_table,
which matches with the later text.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-id: 20210128171313.2210947-1-den@openvz.org
CC: Stefan Hajnoczi <stefanha@redhat.com>
CC: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

[Replace the original commit message "docs: fix mistake in dirty bitmap
feature description" as suggested by Eric Blake.
--Stefan]

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 docs/interop/parallels.txt | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/interop/parallels.txt b/docs/interop/parallels.txt
index XXXXXXX..XXXXXXX 100644
--- a/docs/interop/parallels.txt
+++ b/docs/interop/parallels.txt
@@ -XXX,XX +XXX,XX @@ of its data area are:
   28 - 31:    l1_size
               The number of entries in the L1 table of the bitmap.
 
-  variable:   l1 (64 * l1_size bytes)
+  variable:   l1_table (8 * l1_size bytes)
               L1 offset table (in bytes)
 
 A dirty bitmap is stored using a one-level structure for the mapping to host
-- 
2.29.2