[Qemu-devel] [PATCH for-2.7? 0/3] Two coroutine tweaks

qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed

* [Qemu-devel] [PATCH for-2.7? 0/3] Two coroutine tweaks
@ 2016-07-04 17:09 Paolo Bonzini
  2016-07-04 17:09 ` [Qemu-devel] [PATCH 1/3] coroutine: use QSIMPLEQ instead of QTAILQ Paolo Bonzini
                   ` (4 more replies)
  0 siblings, 5 replies; 6+ messages in thread
From: Paolo Bonzini @ 2016-07-04 17:09 UTC (permalink / raw)
  To: qemu-devel; +Cc: kwolf, famz, qemu-block

Hi,

I meant to send these close to soft freeze because the final patch is
prone to conflicts, but it turns out I didn't send them.  Still they're
pretty mechanical so I'm shooting them out now.  Any thoughts?

Thanks,

Paolo

Paolo Bonzini (3):
  coroutine: use QSIMPLEQ instead of QTAILQ
  test-coroutine: prepare for the next patch
  coroutine: move entry argument to qemu_coroutine_create

 block.c                      |  4 +--
 block/backup.c               |  4 +--
 block/blkdebug.c             |  4 +--
 block/blkreplay.c            |  2 +-
 block/block-backend.c        |  8 +++---
 block/commit.c               |  4 +--
 block/gluster.c              |  2 +-
 block/io.c                   | 45 +++++++++++++++---------------
 block/iscsi.c                |  4 +--
 block/linux-aio.c            |  2 +-
 block/mirror.c               |  6 ++--
 block/nbd-client.c           |  6 ++--
 block/nfs.c                  |  2 +-
 block/qed.c                  |  4 +--
 block/sheepdog.c             | 14 +++++-----
 block/ssh.c                  |  2 +-
 block/stream.c               |  4 +--
 block/vmdk.c                 |  4 +--
 blockjob.c                   |  2 +-
 hw/9pfs/9p.c                 |  4 +--
 hw/9pfs/coth.c               |  4 +--
 include/qemu/coroutine.h     | 10 +++----
 include/qemu/coroutine_int.h |  4 +--
 include/qemu/main-loop.h     |  4 +--
 io/channel.c                 |  2 +-
 migration/migration.c        |  4 +--
 nbd/server.c                 | 12 ++++----
 qemu-io-cmds.c               |  4 +--
 tests/test-blockjob-txn.c    |  4 +--
 tests/test-coroutine.c       | 65 ++++++++++++++++++++++----------------------
 tests/test-thread-pool.c     |  4 +--
 thread-pool.c                |  2 +-
 util/qemu-coroutine-io.c     |  2 +-
 util/qemu-coroutine-lock.c   | 26 +++++++++---------
 util/qemu-coroutine-sleep.c  |  2 +-
 util/qemu-coroutine.c        | 10 +++----
 36 files changed, 143 insertions(+), 143 deletions(-)

-- 
1.8.3.1

^ permalink raw reply	[flat|nested] 6+ messages in thread

* [Qemu-devel] [PATCH 1/3] coroutine: use QSIMPLEQ instead of QTAILQ
  2016-07-04 17:09 [Qemu-devel] [PATCH for-2.7? 0/3] Two coroutine tweaks Paolo Bonzini
@ 2016-07-04 17:09 ` Paolo Bonzini
  2016-07-04 17:10 ` [Qemu-devel] [PATCH 2/3] test-coroutine: prepare for the next patch Paolo Bonzini
                   ` (3 subsequent siblings)
  4 siblings, 0 replies; 6+ messages in thread
From: Paolo Bonzini @ 2016-07-04 17:09 UTC (permalink / raw)
  To: qemu-devel; +Cc: kwolf, famz, qemu-block

CoQueue do not need to remove any element but the head of the list;
processing is always strictly FIFO.  Therefore, the simpler singly-linked
QSIMPLEQ can be used instead.

Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 include/qemu/coroutine.h     |  2 +-
 include/qemu/coroutine_int.h |  4 ++--
 util/qemu-coroutine-lock.c   | 22 +++++++++++-----------
 util/qemu-coroutine.c        |  2 +-
 4 files changed, 15 insertions(+), 15 deletions(-)

diff --git a/include/qemu/coroutine.h b/include/qemu/coroutine.h
index 305fe76..63ae7fe 100644
--- a/include/qemu/coroutine.h
+++ b/include/qemu/coroutine.h
@@ -102,7 +102,7 @@ bool qemu_in_coroutine(void);
  * are built.
  */
 typedef struct CoQueue {
-    QTAILQ_HEAD(, Coroutine) entries;
+    QSIMPLEQ_HEAD(, Coroutine) entries;
 } CoQueue;
 
 /**
diff --git a/include/qemu/coroutine_int.h b/include/qemu/coroutine_int.h
index 42d6838..581a7f5 100644
--- a/include/qemu/coroutine_int.h
+++ b/include/qemu/coroutine_int.h
@@ -41,8 +41,8 @@ struct Coroutine {
     QSLIST_ENTRY(Coroutine) pool_next;
 
     /* Coroutines that should be woken up when we yield or terminate */
-    QTAILQ_HEAD(, Coroutine) co_queue_wakeup;
-    QTAILQ_ENTRY(Coroutine) co_queue_next;
+    QSIMPLEQ_HEAD(, Coroutine) co_queue_wakeup;
+    QSIMPLEQ_ENTRY(Coroutine) co_queue_next;
 };
 
 Coroutine *qemu_coroutine_new(void);
diff --git a/util/qemu-coroutine-lock.c b/util/qemu-coroutine-lock.c
index da37ca7..cf53693 100644
--- a/util/qemu-coroutine-lock.c
+++ b/util/qemu-coroutine-lock.c
@@ -31,13 +31,13 @@
 
 void qemu_co_queue_init(CoQueue *queue)
 {
-    QTAILQ_INIT(&queue->entries);
+    QSIMPLEQ_INIT(&queue->entries);
 }
 
 void coroutine_fn qemu_co_queue_wait(CoQueue *queue)
 {
     Coroutine *self = qemu_coroutine_self();
-    QTAILQ_INSERT_TAIL(&queue->entries, self, co_queue_next);
+    QSIMPLEQ_INSERT_TAIL(&queue->entries, self, co_queue_next);
     qemu_coroutine_yield();
     assert(qemu_in_coroutine());
 }
@@ -55,8 +55,8 @@ void qemu_co_queue_run_restart(Coroutine *co)
     Coroutine *next;
 
     trace_qemu_co_queue_run_restart(co);
-    while ((next = QTAILQ_FIRST(&co->co_queue_wakeup))) {
-        QTAILQ_REMOVE(&co->co_queue_wakeup, next, co_queue_next);
+    while ((next = QSIMPLEQ_FIRST(&co->co_queue_wakeup))) {
+        QSIMPLEQ_REMOVE_HEAD(&co->co_queue_wakeup, co_queue_next);
         qemu_coroutine_enter(next, NULL);
     }
 }
@@ -66,13 +66,13 @@ static bool qemu_co_queue_do_restart(CoQueue *queue, bool single)
     Coroutine *self = qemu_coroutine_self();
     Coroutine *next;
 
-    if (QTAILQ_EMPTY(&queue->entries)) {
+    if (QSIMPLEQ_EMPTY(&queue->entries)) {
         return false;
     }
 
-    while ((next = QTAILQ_FIRST(&queue->entries)) != NULL) {
-        QTAILQ_REMOVE(&queue->entries, next, co_queue_next);
-        QTAILQ_INSERT_TAIL(&self->co_queue_wakeup, next, co_queue_next);
+    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);
         if (single) {
             break;
@@ -97,19 +97,19 @@ bool qemu_co_enter_next(CoQueue *queue)
 {
     Coroutine *next;
 
-    next = QTAILQ_FIRST(&queue->entries);
+    next = QSIMPLEQ_FIRST(&queue->entries);
     if (!next) {
         return false;
     }
 
-    QTAILQ_REMOVE(&queue->entries, next, co_queue_next);
+    QSIMPLEQ_REMOVE_HEAD(&queue->entries, co_queue_next);
     qemu_coroutine_enter(next, NULL);
     return true;
 }
 
 bool qemu_co_queue_empty(CoQueue *queue)
 {
-    return QTAILQ_FIRST(&queue->entries) == NULL;
+    return QSIMPLEQ_FIRST(&queue->entries) == NULL;
 }
 
 void qemu_co_mutex_init(CoMutex *mutex)
diff --git a/util/qemu-coroutine.c b/util/qemu-coroutine.c
index 5816702..b7cb636 100644
--- a/util/qemu-coroutine.c
+++ b/util/qemu-coroutine.c
@@ -76,7 +76,7 @@ Coroutine *qemu_coroutine_create(CoroutineEntry *entry)
     }
 
     co->entry = entry;
-    QTAILQ_INIT(&co->co_queue_wakeup);
+    QSIMPLEQ_INIT(&co->co_queue_wakeup);
     return co;
 }
 
-- 
1.8.3.1

^ permalink raw reply related	[flat|nested] 6+ messages in thread

* [Qemu-devel] [PATCH 2/3] test-coroutine: prepare for the next patch
  2016-07-04 17:09 [Qemu-devel] [PATCH for-2.7? 0/3] Two coroutine tweaks Paolo Bonzini
  2016-07-04 17:09 ` [Qemu-devel] [PATCH 1/3] coroutine: use QSIMPLEQ instead of QTAILQ Paolo Bonzini
@ 2016-07-04 17:10 ` Paolo Bonzini
  2016-07-04 17:10 ` [Qemu-devel] [PATCH 3/3] coroutine: move entry argument to qemu_coroutine_create Paolo Bonzini
                   ` (2 subsequent siblings)
  4 siblings, 0 replies; 6+ messages in thread
From: Paolo Bonzini @ 2016-07-04 17:10 UTC (permalink / raw)
  To: qemu-devel; +Cc: kwolf, famz, qemu-block

The next patch moves the coroutine argument from first-enter to
creation time.  In this case, coroutine has not been initialized
yet when the coroutine is created, so change to a pointer.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 tests/test-coroutine.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/tests/test-coroutine.c b/tests/test-coroutine.c
index 215b92e..5171174 100644
--- a/tests/test-coroutine.c
+++ b/tests/test-coroutine.c
@@ -40,7 +40,8 @@ static void test_in_coroutine(void)
 
 static void coroutine_fn verify_self(void *opaque)
 {
-    g_assert(qemu_coroutine_self() == opaque);
+    Coroutine **p_co = opaque;
+    g_assert(qemu_coroutine_self() == *p_co);
 }
 
 static void test_self(void)
@@ -48,7 +49,7 @@ static void test_self(void)
     Coroutine *coroutine;
 
     coroutine = qemu_coroutine_create(verify_self);
-    qemu_coroutine_enter(coroutine, coroutine);
+    qemu_coroutine_enter(coroutine, &coroutine);
 }
 
 /*
-- 
1.8.3.1

^ permalink raw reply related	[flat|nested] 6+ messages in thread

* [Qemu-devel] [PATCH 3/3] coroutine: move entry argument to qemu_coroutine_create
  2016-07-04 17:09 [Qemu-devel] [PATCH for-2.7? 0/3] Two coroutine tweaks Paolo Bonzini
  2016-07-04 17:09 ` [Qemu-devel] [PATCH 1/3] coroutine: use QSIMPLEQ instead of QTAILQ Paolo Bonzini
  2016-07-04 17:10 ` [Qemu-devel] [PATCH 2/3] test-coroutine: prepare for the next patch Paolo Bonzini
@ 2016-07-04 17:10 ` Paolo Bonzini
  2016-07-05  9:09 ` [Qemu-devel] [PATCH for-2.7? 0/3] Two coroutine tweaks Fam Zheng
  2016-07-08 15:50 ` Kevin Wolf
  4 siblings, 0 replies; 6+ messages in thread
From: Paolo Bonzini @ 2016-07-04 17:10 UTC (permalink / raw)
  To: qemu-devel; +Cc: kwolf, famz, qemu-block

In practice the entry argument is always known at creation time, and
it is confusing that sometimes qemu_coroutine_enter is used with a
non-NULL argument to re-enter a coroutine (this happens in
block/sheepdog.c and tests/test-coroutine.c).  So pass the opaque value
at creation time, for consistency with e.g. aio_bh_new.

Mostly done with the following semantic patch:

@ entry1 @
expression entry, arg, co;
@@
- co = qemu_coroutine_create(entry);
+ co = qemu_coroutine_create(entry, arg);
  ...
- qemu_coroutine_enter(co, arg);
+ qemu_coroutine_enter(co);

@ entry2 @
expression entry, arg;
identifier co;
@@
- Coroutine *co = qemu_coroutine_create(entry);
+ Coroutine *co = qemu_coroutine_create(entry, arg);
  ...
- qemu_coroutine_enter(co, arg);
+ qemu_coroutine_enter(co);

@ entry3 @
expression entry, arg;
@@
- qemu_coroutine_enter(qemu_coroutine_create(entry), arg);
+ qemu_coroutine_enter(qemu_coroutine_create(entry, arg));

@ reentry @
expression co;
@@
- qemu_coroutine_enter(co, NULL);
+ qemu_coroutine_enter(co);

except for the aforementioned few places where the semantic patch
stumbled (as expected) and for test_co_queue, which would otherwise
produce an uninitialized variable warning.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 block.c                     |  4 +--
 block/backup.c              |  4 +--
 block/blkdebug.c            |  4 +--
 block/blkreplay.c           |  2 +-
 block/block-backend.c       |  8 +++---
 block/commit.c              |  4 +--
 block/gluster.c             |  2 +-
 block/io.c                  | 45 ++++++++++++++++----------------
 block/iscsi.c               |  4 +--
 block/linux-aio.c           |  2 +-
 block/mirror.c              |  6 ++---
 block/nbd-client.c          |  6 ++---
 block/nfs.c                 |  2 +-
 block/qed.c                 |  4 +--
 block/sheepdog.c            | 14 +++++-----
 block/ssh.c                 |  2 +-
 block/stream.c              |  4 +--
 block/vmdk.c                |  4 +--
 blockjob.c                  |  2 +-
 hw/9pfs/9p.c                |  4 +--
 hw/9pfs/coth.c              |  4 +--
 include/qemu/coroutine.h    |  8 +++---
 include/qemu/main-loop.h    |  4 +--
 io/channel.c                |  2 +-
 migration/migration.c       |  4 +--
 nbd/server.c                | 12 ++++-----
 qemu-io-cmds.c              |  4 +--
 tests/test-blockjob-txn.c   |  4 +--
 tests/test-coroutine.c      | 62 ++++++++++++++++++++++-----------------------
 tests/test-thread-pool.c    |  4 +--
 thread-pool.c               |  2 +-
 util/qemu-coroutine-io.c    |  2 +-
 util/qemu-coroutine-lock.c  |  4 +--
 util/qemu-coroutine-sleep.c |  2 +-
 util/qemu-coroutine.c       |  8 +++---
 35 files changed, 126 insertions(+), 127 deletions(-)

diff --git a/block.c b/block.c
index f4648e9..cf62ae9 100644
--- a/block.c
+++ b/block.c
@@ -329,8 +329,8 @@ int bdrv_create(BlockDriver *drv, const char* filename,
         /* Fast-path if already in coroutine context */
         bdrv_create_co_entry(&cco);
     } else {
-        co = qemu_coroutine_create(bdrv_create_co_entry);
-        qemu_coroutine_enter(co, &cco);
+        co = qemu_coroutine_create(bdrv_create_co_entry, &cco);
+        qemu_coroutine_enter(co);
         while (cco.ret == NOT_DONE) {
             aio_poll(qemu_get_aio_context(), true);
         }
diff --git a/block/backup.c b/block/backup.c
index f87f8d5..462f8f5 100644
--- a/block/backup.c
+++ b/block/backup.c
@@ -575,9 +575,9 @@ void backup_start(BlockDriverState *bs, BlockDriverState *target,
 
     bdrv_op_block_all(target, job->common.blocker);
     job->common.len = len;
-    job->common.co = qemu_coroutine_create(backup_run);
+    job->common.co = qemu_coroutine_create(backup_run, job);
     block_job_txn_add_job(txn, &job->common);
-    qemu_coroutine_enter(job->common.co, job);
+    qemu_coroutine_enter(job->common.co);
     return;
 
  error:
diff --git a/block/blkdebug.c b/block/blkdebug.c
index 20d25bd..e5f02af 100644
--- a/block/blkdebug.c
+++ b/block/blkdebug.c
@@ -620,7 +620,7 @@ static int blkdebug_debug_resume(BlockDriverState *bs, const char *tag)
 
     QLIST_FOREACH_SAFE(r, &s->suspended_reqs, next, next) {
         if (!strcmp(r->tag, tag)) {
-            qemu_coroutine_enter(r->co, NULL);
+            qemu_coroutine_enter(r->co);
             return 0;
         }
     }
@@ -646,7 +646,7 @@ static int blkdebug_debug_remove_breakpoint(BlockDriverState *bs,
     }
     QLIST_FOREACH_SAFE(r, &s->suspended_reqs, next, r_next) {
         if (!strcmp(r->tag, tag)) {
-            qemu_coroutine_enter(r->co, NULL);
+            qemu_coroutine_enter(r->co);
             ret = 0;
         }
     }
diff --git a/block/blkreplay.c b/block/blkreplay.c
index 525c2d5..e0f6b95 100755
--- a/block/blkreplay.c
+++ b/block/blkreplay.c
@@ -65,7 +65,7 @@ static int64_t blkreplay_getlength(BlockDriverState *bs)
 static void blkreplay_bh_cb(void *opaque)
 {
     Request *req = opaque;
-    qemu_coroutine_enter(req->co, NULL);
+    qemu_coroutine_enter(req->co);
     qemu_bh_delete(req->bh);
     g_free(req);
 }
diff --git a/block/block-backend.c b/block/block-backend.c
index 34500e6..0e1b16c 100644
--- a/block/block-backend.c
+++ b/block/block-backend.c
@@ -836,8 +836,8 @@ static int blk_prw(BlockBackend *blk, int64_t offset, uint8_t *buf,
         .ret    = NOT_DONE,
     };
 
-    co = qemu_coroutine_create(co_entry);
-    qemu_coroutine_enter(co, &rwco);
+    co = qemu_coroutine_create(co_entry, &rwco);
+    qemu_coroutine_enter(co);
 
     aio_context = blk_get_aio_context(blk);
     while (rwco.ret == NOT_DONE) {
@@ -945,8 +945,8 @@ static BlockAIOCB *blk_aio_prwv(BlockBackend *blk, int64_t offset, int bytes,
     acb->bh = NULL;
     acb->has_returned = false;
 
-    co = qemu_coroutine_create(co_entry);
-    qemu_coroutine_enter(co, acb);
+    co = qemu_coroutine_create(co_entry, acb);
+    qemu_coroutine_enter(co);
 
     acb->has_returned = true;
     if (acb->rwco.ret != NOT_DONE) {
diff --git a/block/commit.c b/block/commit.c
index 444333b..e921b05 100644
--- a/block/commit.c
+++ b/block/commit.c
@@ -277,8 +277,8 @@ void commit_start(BlockDriverState *bs, BlockDriverState *base,
     s->backing_file_str = g_strdup(backing_file_str);
 
     s->on_error = on_error;
-    s->common.co = qemu_coroutine_create(commit_run);
+    s->common.co = qemu_coroutine_create(commit_run, s);
 
     trace_commit_start(bs, base, top, s, s->common.co, opaque);
-    qemu_coroutine_enter(s->common.co, s);
+    qemu_coroutine_enter(s->common.co);
 }
diff --git a/block/gluster.c b/block/gluster.c
index 16f7778..406c1e6 100644
--- a/block/gluster.c
+++ b/block/gluster.c
@@ -233,7 +233,7 @@ static void qemu_gluster_complete_aio(void *opaque)
 
     qemu_bh_delete(acb->bh);
     acb->bh = NULL;
-    qemu_coroutine_enter(acb->coroutine, NULL);
+    qemu_coroutine_enter(acb->coroutine);
 }
 
 /*
diff --git a/block/io.c b/block/io.c
index 7cf3645..6709254 100644
--- a/block/io.c
+++ b/block/io.c
@@ -199,7 +199,7 @@ static void bdrv_co_drain_bh_cb(void *opaque)
     qemu_bh_delete(data->bh);
     bdrv_drain_poll(data->bs);
     data->done = true;
-    qemu_coroutine_enter(co, NULL);
+    qemu_coroutine_enter(co);
 }
 
 static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs)
@@ -603,8 +603,8 @@ static int bdrv_prwv_co(BlockDriverState *bs, int64_t offset,
     } else {
         AioContext *aio_context = bdrv_get_aio_context(bs);
 
-        co = qemu_coroutine_create(bdrv_rw_co_entry);
-        qemu_coroutine_enter(co, &rwco);
+        co = qemu_coroutine_create(bdrv_rw_co_entry, &rwco);
+        qemu_coroutine_enter(co);
         while (rwco.ret == NOT_DONE) {
             aio_poll(aio_context, true);
         }
@@ -803,7 +803,7 @@ static void bdrv_co_io_em_complete(void *opaque, int ret)
     CoroutineIOCompletion *co = opaque;
 
     co->ret = ret;
-    qemu_coroutine_enter(co->coroutine, NULL);
+    qemu_coroutine_enter(co->coroutine);
 }
 
 static int coroutine_fn bdrv_driver_preadv(BlockDriverState *bs,
@@ -1743,8 +1743,9 @@ int64_t bdrv_get_block_status_above(BlockDriverState *bs,
     } else {
         AioContext *aio_context = bdrv_get_aio_context(bs);
 
-        co = qemu_coroutine_create(bdrv_get_block_status_above_co_entry);
-        qemu_coroutine_enter(co, &data);
+        co = qemu_coroutine_create(bdrv_get_block_status_above_co_entry,
+                                   &data);
+        qemu_coroutine_enter(co);
         while (!data.done) {
             aio_poll(aio_context, true);
         }
@@ -1892,9 +1893,9 @@ bdrv_rw_vmstate(BlockDriverState *bs, QEMUIOVector *qiov, int64_t pos,
             .is_read    = is_read,
             .ret        = -EINPROGRESS,
         };
-        Coroutine *co = qemu_coroutine_create(bdrv_co_rw_vmstate_entry);
+        Coroutine *co = qemu_coroutine_create(bdrv_co_rw_vmstate_entry, &data);
 
-        qemu_coroutine_enter(co, &data);
+        qemu_coroutine_enter(co);
         while (data.ret == -EINPROGRESS) {
             aio_poll(bdrv_get_aio_context(bs), true);
         }
@@ -2103,8 +2104,8 @@ static BlockAIOCB *bdrv_co_aio_rw_vector(BlockDriverState *bs,
     acb->req.flags = flags;
     acb->is_write = is_write;
 
-    co = qemu_coroutine_create(bdrv_co_do_rw);
-    qemu_coroutine_enter(co, acb);
+    co = qemu_coroutine_create(bdrv_co_do_rw, acb);
+    qemu_coroutine_enter(co);
 
     bdrv_co_maybe_schedule_bh(acb);
     return &acb->common;
@@ -2131,8 +2132,8 @@ BlockAIOCB *bdrv_aio_flush(BlockDriverState *bs,
     acb->need_bh = true;
     acb->req.error = -EINPROGRESS;
 
-    co = qemu_coroutine_create(bdrv_aio_flush_co_entry);
-    qemu_coroutine_enter(co, acb);
+    co = qemu_coroutine_create(bdrv_aio_flush_co_entry, acb);
+    qemu_coroutine_enter(co);
 
     bdrv_co_maybe_schedule_bh(acb);
     return &acb->common;
@@ -2161,8 +2162,8 @@ BlockAIOCB *bdrv_aio_discard(BlockDriverState *bs,
     acb->req.error = -EINPROGRESS;
     acb->req.sector = sector_num;
     acb->req.nb_sectors = nb_sectors;
-    co = qemu_coroutine_create(bdrv_aio_discard_co_entry);
-    qemu_coroutine_enter(co, acb);
+    co = qemu_coroutine_create(bdrv_aio_discard_co_entry, acb);
+    qemu_coroutine_enter(co);
 
     bdrv_co_maybe_schedule_bh(acb);
     return &acb->common;
@@ -2297,8 +2298,8 @@ int bdrv_flush(BlockDriverState *bs)
     } else {
         AioContext *aio_context = bdrv_get_aio_context(bs);
 
-        co = qemu_coroutine_create(bdrv_flush_co_entry);
-        qemu_coroutine_enter(co, &rwco);
+        co = qemu_coroutine_create(bdrv_flush_co_entry, &rwco);
+        qemu_coroutine_enter(co);
         while (rwco.ret == NOT_DONE) {
             aio_poll(aio_context, true);
         }
@@ -2424,8 +2425,8 @@ int bdrv_discard(BlockDriverState *bs, int64_t sector_num, int nb_sectors)
     } else {
         AioContext *aio_context = bdrv_get_aio_context(bs);
 
-        co = qemu_coroutine_create(bdrv_discard_co_entry);
-        qemu_coroutine_enter(co, &rwco);
+        co = qemu_coroutine_create(bdrv_discard_co_entry, &rwco);
+        qemu_coroutine_enter(co);
         while (rwco.ret == NOT_DONE) {
             aio_poll(aio_context, true);
         }
@@ -2487,9 +2488,9 @@ int bdrv_ioctl(BlockDriverState *bs, unsigned long int req, void *buf)
         /* Fast-path if already in coroutine context */
         bdrv_co_ioctl_entry(&data);
     } else {
-        Coroutine *co = qemu_coroutine_create(bdrv_co_ioctl_entry);
+        Coroutine *co = qemu_coroutine_create(bdrv_co_ioctl_entry, &data);
 
-        qemu_coroutine_enter(co, &data);
+        qemu_coroutine_enter(co);
         while (data.ret == -EINPROGRESS) {
             aio_poll(bdrv_get_aio_context(bs), true);
         }
@@ -2517,8 +2518,8 @@ BlockAIOCB *bdrv_aio_ioctl(BlockDriverState *bs,
     acb->req.error = -EINPROGRESS;
     acb->req.req = req;
     acb->req.buf = buf;
-    co = qemu_coroutine_create(bdrv_co_aio_ioctl_entry);
-    qemu_coroutine_enter(co, acb);
+    co = qemu_coroutine_create(bdrv_co_aio_ioctl_entry, acb);
+    qemu_coroutine_enter(co);
 
     bdrv_co_maybe_schedule_bh(acb);
     return &acb->common;
diff --git a/block/iscsi.c b/block/iscsi.c
index 9bb5ff6..70c02ee 100644
--- a/block/iscsi.c
+++ b/block/iscsi.c
@@ -153,7 +153,7 @@ static void iscsi_co_generic_bh_cb(void *opaque)
     struct IscsiTask *iTask = opaque;
     iTask->complete = 1;
     qemu_bh_delete(iTask->bh);
-    qemu_coroutine_enter(iTask->co, NULL);
+    qemu_coroutine_enter(iTask->co);
 }
 
 static void iscsi_retry_timer_expired(void *opaque)
@@ -161,7 +161,7 @@ static void iscsi_retry_timer_expired(void *opaque)
     struct IscsiTask *iTask = opaque;
     iTask->complete = 1;
     if (iTask->co) {
-        qemu_coroutine_enter(iTask->co, NULL);
+        qemu_coroutine_enter(iTask->co);
     }
 }
 
diff --git a/block/linux-aio.c b/block/linux-aio.c
index 3eb0a0e..0ed1c23 100644
--- a/block/linux-aio.c
+++ b/block/linux-aio.c
@@ -96,7 +96,7 @@ static void qemu_laio_process_completion(struct qemu_laiocb *laiocb)
 
     laiocb->ret = ret;
     if (laiocb->co) {
-        qemu_coroutine_enter(laiocb->co, NULL);
+        qemu_coroutine_enter(laiocb->co);
     } else {
         laiocb->common.cb(laiocb->common.opaque, ret);
         qemu_aio_unref(laiocb);
diff --git a/block/mirror.c b/block/mirror.c
index 8d96049..746bf06 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -121,7 +121,7 @@ static void mirror_iteration_done(MirrorOp *op, int ret)
     g_free(op);
 
     if (s->waiting_for_io) {
-        qemu_coroutine_enter(s->common.co, NULL);
+        qemu_coroutine_enter(s->common.co);
     }
 }
 
@@ -900,9 +900,9 @@ static void mirror_start_job(BlockDriverState *bs, BlockDriverState *target,
 
     bdrv_op_block_all(target, s->common.blocker);
 
-    s->common.co = qemu_coroutine_create(mirror_run);
+    s->common.co = qemu_coroutine_create(mirror_run, s);
     trace_mirror_start(bs, s, s->common.co, opaque);
-    qemu_coroutine_enter(s->common.co, s);
+    qemu_coroutine_enter(s->common.co);
 }
 
 void mirror_start(BlockDriverState *bs, BlockDriverState *target,
diff --git a/block/nbd-client.c b/block/nbd-client.c
index 4d13444..3067063 100644
--- a/block/nbd-client.c
+++ b/block/nbd-client.c
@@ -38,7 +38,7 @@ static void nbd_recv_coroutines_enter_all(NbdClientSession *s)
 
     for (i = 0; i < MAX_NBD_REQUESTS; i++) {
         if (s->recv_coroutine[i]) {
-            qemu_coroutine_enter(s->recv_coroutine[i], NULL);
+            qemu_coroutine_enter(s->recv_coroutine[i]);
         }
     }
 }
@@ -99,7 +99,7 @@ static void nbd_reply_ready(void *opaque)
     }
 
     if (s->recv_coroutine[i]) {
-        qemu_coroutine_enter(s->recv_coroutine[i], NULL);
+        qemu_coroutine_enter(s->recv_coroutine[i]);
         return;
     }
 
@@ -111,7 +111,7 @@ static void nbd_restart_write(void *opaque)
 {
     BlockDriverState *bs = opaque;
 
-    qemu_coroutine_enter(nbd_get_client_session(bs)->send_coroutine, NULL);
+    qemu_coroutine_enter(nbd_get_client_session(bs)->send_coroutine);
 }
 
 static int nbd_co_send_request(BlockDriverState *bs,
diff --git a/block/nfs.c b/block/nfs.c
index 15d6832..8602a44 100644
--- a/block/nfs.c
+++ b/block/nfs.c
@@ -104,7 +104,7 @@ static void nfs_co_generic_bh_cb(void *opaque)
     NFSRPC *task = opaque;
     task->complete = 1;
     qemu_bh_delete(task->bh);
-    qemu_coroutine_enter(task->co, NULL);
+    qemu_coroutine_enter(task->co);
 }
 
 static void
diff --git a/block/qed.c b/block/qed.c
index 1206806..b5a8a1f 100644
--- a/block/qed.c
+++ b/block/qed.c
@@ -708,7 +708,7 @@ static void qed_is_allocated_cb(void *opaque, int ret, uint64_t offset, size_t l
     }
 
     if (cb->co) {
-        qemu_coroutine_enter(cb->co, NULL);
+        qemu_coroutine_enter(cb->co);
     }
 }
 
@@ -1425,7 +1425,7 @@ 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, NULL);
+        qemu_coroutine_enter(cb->co);
     }
 }
 
diff --git a/block/sheepdog.c b/block/sheepdog.c
index ef5d044..e739c56 100644
--- a/block/sheepdog.c
+++ b/block/sheepdog.c
@@ -495,7 +495,7 @@ static inline void free_aio_req(BDRVSheepdogState *s, AIOReq *aio_req)
 
 static void coroutine_fn sd_finish_aiocb(SheepdogAIOCB *acb)
 {
-    qemu_coroutine_enter(acb->coroutine, NULL);
+    qemu_coroutine_enter(acb->coroutine);
     qemu_aio_unref(acb);
 }
 
@@ -636,7 +636,7 @@ static void restart_co_req(void *opaque)
 {
     Coroutine *co = opaque;
 
-    qemu_coroutine_enter(co, NULL);
+    qemu_coroutine_enter(co);
 }
 
 typedef struct SheepdogReqCo {
@@ -726,8 +726,8 @@ static int do_req(int sockfd, AioContext *aio_context, SheepdogReq *hdr,
     if (qemu_in_coroutine()) {
         do_co_req(&srco);
     } else {
-        co = qemu_coroutine_create(do_co_req);
-        qemu_coroutine_enter(co, &srco);
+        co = qemu_coroutine_create(do_co_req, &srco);
+        qemu_coroutine_enter(co);
         while (!srco.finished) {
             aio_poll(aio_context, true);
         }
@@ -925,17 +925,17 @@ static void co_read_response(void *opaque)
     BDRVSheepdogState *s = opaque;
 
     if (!s->co_recv) {
-        s->co_recv = qemu_coroutine_create(aio_read_response);
+        s->co_recv = qemu_coroutine_create(aio_read_response, opaque);
     }
 
-    qemu_coroutine_enter(s->co_recv, opaque);
+    qemu_coroutine_enter(s->co_recv);
 }
 
 static void co_write_request(void *opaque)
 {
     BDRVSheepdogState *s = opaque;
 
-    qemu_coroutine_enter(s->co_send, NULL);
+    qemu_coroutine_enter(s->co_send);
 }
 
 /*
diff --git a/block/ssh.c b/block/ssh.c
index 06928ed..bcbb0e4 100644
--- a/block/ssh.c
+++ b/block/ssh.c
@@ -777,7 +777,7 @@ static void restart_coroutine(void *opaque)
 
     DPRINTF("co=%p", co);
 
-    qemu_coroutine_enter(co, NULL);
+    qemu_coroutine_enter(co);
 }
 
 static coroutine_fn void set_fd_handler(BDRVSSHState *s, BlockDriverState *bs)
diff --git a/block/stream.c b/block/stream.c
index c0efbda..3418701 100644
--- a/block/stream.c
+++ b/block/stream.c
@@ -235,7 +235,7 @@ void stream_start(BlockDriverState *bs, BlockDriverState *base,
     s->backing_file_str = g_strdup(backing_file_str);
 
     s->on_error = on_error;
-    s->common.co = qemu_coroutine_create(stream_run);
+    s->common.co = qemu_coroutine_create(stream_run, s);
     trace_stream_start(bs, base, s, s->common.co, opaque);
-    qemu_coroutine_enter(s->common.co, s);
+    qemu_coroutine_enter(s->common.co);
 }
diff --git a/block/vmdk.c b/block/vmdk.c
index 2901692..7e5e87f 100644
--- a/block/vmdk.c
+++ b/block/vmdk.c
@@ -1690,8 +1690,8 @@ static int vmdk_write_compressed(BlockDriverState *bs,
             .nb_sectors = nb_sectors,
             .ret        = -EINPROGRESS,
         };
-        co = qemu_coroutine_create(vmdk_co_write_compressed);
-        qemu_coroutine_enter(co, &data);
+        co = qemu_coroutine_create(vmdk_co_write_compressed, &data);
+        qemu_coroutine_enter(co);
         while (data.ret == -EINPROGRESS) {
             aio_poll(aio_context, true);
         }
diff --git a/blockjob.c b/blockjob.c
index 205da9d..4ba5cbf 100644
--- a/blockjob.c
+++ b/blockjob.c
@@ -346,7 +346,7 @@ void block_job_resume(BlockJob *job)
 void block_job_enter(BlockJob *job)
 {
     if (job->co && !job->busy) {
-        qemu_coroutine_enter(job->co, NULL);
+        qemu_coroutine_enter(job->co);
     }
 }
 
diff --git a/hw/9pfs/9p.c b/hw/9pfs/9p.c
index 9acff92..b6b02b4 100644
--- a/hw/9pfs/9p.c
+++ b/hw/9pfs/9p.c
@@ -3278,8 +3278,8 @@ void pdu_submit(V9fsPDU *pdu)
     if (is_ro_export(&s->ctx) && !is_read_only_op(pdu)) {
         handler = v9fs_fs_ro;
     }
-    co = qemu_coroutine_create(handler);
-    qemu_coroutine_enter(co, pdu);
+    co = qemu_coroutine_create(handler, pdu);
+    qemu_coroutine_enter(co);
 }
 
 /* Returns 0 on success, 1 on failure. */
diff --git a/hw/9pfs/coth.c b/hw/9pfs/coth.c
index 9b1151b..89018de 100644
--- a/hw/9pfs/coth.c
+++ b/hw/9pfs/coth.c
@@ -22,14 +22,14 @@
 static void coroutine_enter_cb(void *opaque, int ret)
 {
     Coroutine *co = opaque;
-    qemu_coroutine_enter(co, NULL);
+    qemu_coroutine_enter(co);
 }
 
 /* Called from worker thread.  */
 static int coroutine_enter_func(void *arg)
 {
     Coroutine *co = arg;
-    qemu_coroutine_enter(co, NULL);
+    qemu_coroutine_enter(co);
     return 0;
 }
 
diff --git a/include/qemu/coroutine.h b/include/qemu/coroutine.h
index 63ae7fe..ac8d4c9 100644
--- a/include/qemu/coroutine.h
+++ b/include/qemu/coroutine.h
@@ -61,16 +61,14 @@ typedef void coroutine_fn CoroutineEntry(void *opaque);
  * Create a new coroutine
  *
  * Use qemu_coroutine_enter() to actually transfer control to the coroutine.
+ * The opaque argument is passed as the argument to the entry point.
  */
-Coroutine *qemu_coroutine_create(CoroutineEntry *entry);
+Coroutine *qemu_coroutine_create(CoroutineEntry *entry, void *opaque);
 
 /**
  * Transfer control to a coroutine
- *
- * The opaque argument is passed as the argument to the entry point when
- * entering the coroutine for the first time.  It is subsequently ignored.
  */
-void qemu_coroutine_enter(Coroutine *coroutine, void *opaque);
+void qemu_coroutine_enter(Coroutine *coroutine);
 
 /**
  * Transfer control back to a coroutine's caller
diff --git a/include/qemu/main-loop.h b/include/qemu/main-loop.h
index 19b5de3..10501aa 100644
--- a/include/qemu/main-loop.h
+++ b/include/qemu/main-loop.h
@@ -64,11 +64,11 @@ int qemu_init_main_loop(Error **errp);
  *
  *     void enter_co_bh(void *opaque) {
  *         QEMUCoroutine *co = opaque;
- *         qemu_coroutine_enter(co, NULL);
+ *         qemu_coroutine_enter(co);
  *     }
  *
  *     ...
- *     QEMUCoroutine *co = qemu_coroutine_create(coroutine_entry);
+ *     QEMUCoroutine *co = qemu_coroutine_create(coroutine_entry, NULL);
  *     QEMUBH *start_bh = qemu_bh_new(enter_co_bh, co);
  *     qemu_bh_schedule(start_bh);
  *     while (...) {
diff --git a/io/channel.c b/io/channel.c
index 692eb17..923c465 100644
--- a/io/channel.c
+++ b/io/channel.c
@@ -218,7 +218,7 @@ static gboolean qio_channel_yield_enter(QIOChannel *ioc,
                                         gpointer opaque)
 {
     QIOChannelYieldData *data = opaque;
-    qemu_coroutine_enter(data->co, NULL);
+    qemu_coroutine_enter(data->co);
     return FALSE;
 }
 
diff --git a/migration/migration.c b/migration/migration.c
index a560136..c4e0193 100644
--- a/migration/migration.c
+++ b/migration/migration.c
@@ -418,11 +418,11 @@ static void process_incoming_migration_co(void *opaque)
 
 void migration_fd_process_incoming(QEMUFile *f)
 {
-    Coroutine *co = qemu_coroutine_create(process_incoming_migration_co);
+    Coroutine *co = qemu_coroutine_create(process_incoming_migration_co, f);
 
     migrate_decompress_threads_create();
     qemu_file_set_blocking(f, false);
-    qemu_coroutine_enter(co, f);
+    qemu_coroutine_enter(co);
 }
 
 
diff --git a/nbd/server.c b/nbd/server.c
index a677e26..fbc82de 100644
--- a/nbd/server.c
+++ b/nbd/server.c
@@ -106,7 +106,7 @@ static gboolean nbd_negotiate_continue(QIOChannel *ioc,
                                        GIOCondition condition,
                                        void *opaque)
 {
-    qemu_coroutine_enter(opaque, NULL);
+    qemu_coroutine_enter(opaque);
     return TRUE;
 }
 
@@ -1230,9 +1230,9 @@ static void nbd_read(void *opaque)
     NBDClient *client = opaque;
 
     if (client->recv_coroutine) {
-        qemu_coroutine_enter(client->recv_coroutine, NULL);
+        qemu_coroutine_enter(client->recv_coroutine);
     } else {
-        qemu_coroutine_enter(qemu_coroutine_create(nbd_trip), client);
+        qemu_coroutine_enter(qemu_coroutine_create(nbd_trip, client));
     }
 }
 
@@ -1240,7 +1240,7 @@ static void nbd_restart_write(void *opaque)
 {
     NBDClient *client = opaque;
 
-    qemu_coroutine_enter(client->send_coroutine, NULL);
+    qemu_coroutine_enter(client->send_coroutine);
 }
 
 static void nbd_set_handlers(NBDClient *client)
@@ -1324,6 +1324,6 @@ void nbd_client_new(NBDExport *exp,
     client->close = close_fn;
 
     data->client = client;
-    data->co = qemu_coroutine_create(nbd_co_client_start);
-    qemu_coroutine_enter(data->co, data);
+    data->co = qemu_coroutine_create(nbd_co_client_start, data);
+    qemu_coroutine_enter(data->co);
 }
diff --git a/qemu-io-cmds.c b/qemu-io-cmds.c
index 09e879f..40fe2eb 100644
--- a/qemu-io-cmds.c
+++ b/qemu-io-cmds.c
@@ -483,8 +483,8 @@ static int do_co_pwrite_zeroes(BlockBackend *blk, int64_t offset,
         return -ERANGE;
     }
 
-    co = qemu_coroutine_create(co_pwrite_zeroes_entry);
-    qemu_coroutine_enter(co, &data);
+    co = qemu_coroutine_create(co_pwrite_zeroes_entry, &data);
+    qemu_coroutine_enter(co);
     while (!data.done) {
         aio_poll(blk_get_aio_context(blk), true);
     }
diff --git a/tests/test-blockjob-txn.c b/tests/test-blockjob-txn.c
index d3030f1..d3629f4 100644
--- a/tests/test-blockjob-txn.c
+++ b/tests/test-blockjob-txn.c
@@ -100,10 +100,10 @@ static BlockJob *test_block_job_start(unsigned int iterations,
     s->use_timer = use_timer;
     s->rc = rc;
     s->result = result;
-    s->common.co = qemu_coroutine_create(test_block_job_run);
+    s->common.co = qemu_coroutine_create(test_block_job_run, s);
     data->job = s;
     data->result = result;
-    qemu_coroutine_enter(s->common.co, s);
+    qemu_coroutine_enter(s->common.co);
     return &s->common;
 }
 
diff --git a/tests/test-coroutine.c b/tests/test-coroutine.c
index 5171174..ee5e06d 100644
--- a/tests/test-coroutine.c
+++ b/tests/test-coroutine.c
@@ -30,8 +30,8 @@ static void test_in_coroutine(void)
 
     g_assert(!qemu_in_coroutine());
 
-    coroutine = qemu_coroutine_create(verify_in_coroutine);
-    qemu_coroutine_enter(coroutine, NULL);
+    coroutine = qemu_coroutine_create(verify_in_coroutine, NULL);
+    qemu_coroutine_enter(coroutine);
 }
 
 /*
@@ -48,8 +48,8 @@ static void test_self(void)
 {
     Coroutine *coroutine;
 
-    coroutine = qemu_coroutine_create(verify_self);
-    qemu_coroutine_enter(coroutine, &coroutine);
+    coroutine = qemu_coroutine_create(verify_self, &coroutine);
+    qemu_coroutine_enter(coroutine);
 }
 
 /*
@@ -71,8 +71,8 @@ static void coroutine_fn nest(void *opaque)
     if (nd->n_enter < nd->max) {
         Coroutine *child;
 
-        child = qemu_coroutine_create(nest);
-        qemu_coroutine_enter(child, nd);
+        child = qemu_coroutine_create(nest, nd);
+        qemu_coroutine_enter(child);
     }
 
     nd->n_return++;
@@ -87,8 +87,8 @@ static void test_nesting(void)
         .max      = 128,
     };
 
-    root = qemu_coroutine_create(nest);
-    qemu_coroutine_enter(root, &nd);
+    root = qemu_coroutine_create(nest, &nd);
+    qemu_coroutine_enter(root);
 
     /* Must enter and return from max nesting level */
     g_assert_cmpint(nd.n_enter, ==, nd.max);
@@ -116,9 +116,9 @@ static void test_yield(void)
     bool done = false;
     int i = -1; /* one extra time to return from coroutine */
 
-    coroutine = qemu_coroutine_create(yield_5_times);
+    coroutine = qemu_coroutine_create(yield_5_times, &done);
     while (!done) {
-        qemu_coroutine_enter(coroutine, &done);
+        qemu_coroutine_enter(coroutine);
         i++;
     }
     g_assert_cmpint(i, ==, 5); /* coroutine must yield 5 times */
@@ -132,7 +132,7 @@ static void coroutine_fn c2_fn(void *opaque)
 static void coroutine_fn c1_fn(void *opaque)
 {
     Coroutine *c2 = opaque;
-    qemu_coroutine_enter(c2, NULL);
+    qemu_coroutine_enter(c2);
 }
 
 static void test_co_queue(void)
@@ -140,12 +140,12 @@ static void test_co_queue(void)
     Coroutine *c1;
     Coroutine *c2;
 
-    c1 = qemu_coroutine_create(c1_fn);
-    c2 = qemu_coroutine_create(c2_fn);
+    c2 = qemu_coroutine_create(c2_fn, NULL);
+    c1 = qemu_coroutine_create(c1_fn, c2);
 
-    qemu_coroutine_enter(c1, c2);
+    qemu_coroutine_enter(c1);
     memset(c1, 0xff, sizeof(Coroutine));
-    qemu_coroutine_enter(c2, NULL);
+    qemu_coroutine_enter(c2);
 }
 
 /*
@@ -165,14 +165,14 @@ static void test_lifecycle(void)
     bool done = false;
 
     /* Create, enter, and return from coroutine */
-    coroutine = qemu_coroutine_create(set_and_exit);
-    qemu_coroutine_enter(coroutine, &done);
+    coroutine = qemu_coroutine_create(set_and_exit, &done);
+    qemu_coroutine_enter(coroutine);
     g_assert(done); /* expect done to be true (first time) */
 
     /* Repeat to check that no state affects this test */
     done = false;
-    coroutine = qemu_coroutine_create(set_and_exit);
-    qemu_coroutine_enter(coroutine, &done);
+    coroutine = qemu_coroutine_create(set_and_exit, &done);
+    qemu_coroutine_enter(coroutine);
     g_assert(done); /* expect done to be true (second time) */
 }
 
@@ -206,12 +206,12 @@ static void do_order_test(void)
 {
     Coroutine *co;
 
-    co = qemu_coroutine_create(co_order_test);
+    co = qemu_coroutine_create(co_order_test, NULL);
     record_push(1, 1);
-    qemu_coroutine_enter(co, NULL);
+    qemu_coroutine_enter(co);
     record_push(1, 2);
     g_assert(!qemu_in_coroutine());
-    qemu_coroutine_enter(co, NULL);
+    qemu_coroutine_enter(co);
     record_push(1, 3);
     g_assert(!qemu_in_coroutine());
 }
@@ -248,8 +248,8 @@ static void perf_lifecycle(void)
 
     g_test_timer_start();
     for (i = 0; i < max; i++) {
-        coroutine = qemu_coroutine_create(empty_coroutine);
-        qemu_coroutine_enter(coroutine, NULL);
+        coroutine = qemu_coroutine_create(empty_coroutine, NULL);
+        qemu_coroutine_enter(coroutine);
     }
     duration = g_test_timer_elapsed();
 
@@ -272,8 +272,8 @@ static void perf_nesting(void)
             .n_return = 0,
             .max      = maxnesting,
         };
-        root = qemu_coroutine_create(nest);
-        qemu_coroutine_enter(root, &nd);
+        root = qemu_coroutine_create(nest, &nd);
+        qemu_coroutine_enter(root);
     }
     duration = g_test_timer_elapsed();
 
@@ -302,11 +302,11 @@ static void perf_yield(void)
 
     maxcycles = 100000000;
     i = maxcycles;
-    Coroutine *coroutine = qemu_coroutine_create(yield_loop);
+    Coroutine *coroutine = qemu_coroutine_create(yield_loop, &i);
 
     g_test_timer_start();
     while (i > 0) {
-        qemu_coroutine_enter(coroutine, &i);
+        qemu_coroutine_enter(coroutine);
     }
     duration = g_test_timer_elapsed();
 
@@ -352,9 +352,9 @@ static void perf_cost(void)
 
     g_test_timer_start();
     while (i++ < maxcycles) {
-        co = qemu_coroutine_create(perf_cost_func);
-        qemu_coroutine_enter(co, &i);
-        qemu_coroutine_enter(co, NULL);
+        co = qemu_coroutine_create(perf_cost_func, &i);
+        qemu_coroutine_enter(co);
+        qemu_coroutine_enter(co);
     }
     duration = g_test_timer_elapsed();
     ops = (long)(maxcycles / (duration * 1000));
diff --git a/tests/test-thread-pool.c b/tests/test-thread-pool.c
index b0e1f32..8dbf66a 100644
--- a/tests/test-thread-pool.c
+++ b/tests/test-thread-pool.c
@@ -91,9 +91,9 @@ static void co_test_cb(void *opaque)
 static void test_submit_co(void)
 {
     WorkerTestData data;
-    Coroutine *co = qemu_coroutine_create(co_test_cb);
+    Coroutine *co = qemu_coroutine_create(co_test_cb, &data);
 
-    qemu_coroutine_enter(co, &data);
+    qemu_coroutine_enter(co);
 
     /* Back here once the worker has started.  */
 
diff --git a/thread-pool.c b/thread-pool.c
index 03ba0b0..6fba913 100644
--- a/thread-pool.c
+++ b/thread-pool.c
@@ -267,7 +267,7 @@ static void thread_pool_co_cb(void *opaque, int ret)
     ThreadPoolCo *co = opaque;
 
     co->ret = ret;
-    qemu_coroutine_enter(co->co, NULL);
+    qemu_coroutine_enter(co->co);
 }
 
 int coroutine_fn thread_pool_submit_co(ThreadPool *pool, ThreadPoolFunc *func,
diff --git a/util/qemu-coroutine-io.c b/util/qemu-coroutine-io.c
index 91b9357..44a8969 100644
--- a/util/qemu-coroutine-io.c
+++ b/util/qemu-coroutine-io.c
@@ -75,7 +75,7 @@ static void fd_coroutine_enter(void *opaque)
 {
     FDYieldUntilData *data = opaque;
     qemu_set_fd_handler(data->fd, NULL, NULL, NULL);
-    qemu_coroutine_enter(data->co, NULL);
+    qemu_coroutine_enter(data->co);
 }
 
 void coroutine_fn yield_until_fd_readable(int fd)
diff --git a/util/qemu-coroutine-lock.c b/util/qemu-coroutine-lock.c
index cf53693..22aa9ab 100644
--- a/util/qemu-coroutine-lock.c
+++ b/util/qemu-coroutine-lock.c
@@ -57,7 +57,7 @@ void qemu_co_queue_run_restart(Coroutine *co)
     trace_qemu_co_queue_run_restart(co);
     while ((next = QSIMPLEQ_FIRST(&co->co_queue_wakeup))) {
         QSIMPLEQ_REMOVE_HEAD(&co->co_queue_wakeup, co_queue_next);
-        qemu_coroutine_enter(next, NULL);
+        qemu_coroutine_enter(next);
     }
 }
 
@@ -103,7 +103,7 @@ bool qemu_co_enter_next(CoQueue *queue)
     }
 
     QSIMPLEQ_REMOVE_HEAD(&queue->entries, co_queue_next);
-    qemu_coroutine_enter(next, NULL);
+    qemu_coroutine_enter(next);
     return true;
 }
 
diff --git a/util/qemu-coroutine-sleep.c b/util/qemu-coroutine-sleep.c
index 6966831..25de3ed 100644
--- a/util/qemu-coroutine-sleep.c
+++ b/util/qemu-coroutine-sleep.c
@@ -25,7 +25,7 @@ static void co_sleep_cb(void *opaque)
 {
     CoSleepCB *sleep_cb = opaque;
 
-    qemu_coroutine_enter(sleep_cb->co, NULL);
+    qemu_coroutine_enter(sleep_cb->co);
 }
 
 void coroutine_fn co_aio_sleep_ns(AioContext *ctx, QEMUClockType type,
diff --git a/util/qemu-coroutine.c b/util/qemu-coroutine.c
index b7cb636..89f21a9 100644
--- a/util/qemu-coroutine.c
+++ b/util/qemu-coroutine.c
@@ -42,7 +42,7 @@ static void coroutine_pool_cleanup(Notifier *n, void *value)
     }
 }
 
-Coroutine *qemu_coroutine_create(CoroutineEntry *entry)
+Coroutine *qemu_coroutine_create(CoroutineEntry *entry, void *opaque)
 {
     Coroutine *co = NULL;
 
@@ -76,6 +76,7 @@ Coroutine *qemu_coroutine_create(CoroutineEntry *entry)
     }
 
     co->entry = entry;
+    co->entry_arg = opaque;
     QSIMPLEQ_INIT(&co->co_queue_wakeup);
     return co;
 }
@@ -100,12 +101,12 @@ static void coroutine_delete(Coroutine *co)
     qemu_coroutine_delete(co);
 }
 
-void qemu_coroutine_enter(Coroutine *co, void *opaque)
+void qemu_coroutine_enter(Coroutine *co)
 {
     Coroutine *self = qemu_coroutine_self();
     CoroutineAction ret;
 
-    trace_qemu_coroutine_enter(self, co, opaque);
+    trace_qemu_coroutine_enter(self, co, co->entry_arg);
 
     if (co->caller) {
         fprintf(stderr, "Co-routine re-entered recursively\n");
@@ -113,7 +114,6 @@ void qemu_coroutine_enter(Coroutine *co, void *opaque)
     }
 
     co->caller = self;
-    co->entry_arg = opaque;
     ret = qemu_coroutine_switch(self, co, COROUTINE_ENTER);
 
     qemu_co_queue_run_restart(co);
-- 
1.8.3.1

^ permalink raw reply related	[flat|nested] 6+ messages in thread

* Re: [Qemu-devel] [PATCH for-2.7? 0/3] Two coroutine tweaks
  2016-07-04 17:09 [Qemu-devel] [PATCH for-2.7? 0/3] Two coroutine tweaks Paolo Bonzini
                   ` (2 preceding siblings ...)
  2016-07-04 17:10 ` [Qemu-devel] [PATCH 3/3] coroutine: move entry argument to qemu_coroutine_create Paolo Bonzini
@ 2016-07-05  9:09 ` Fam Zheng
  2016-07-08 15:50 ` Kevin Wolf
  4 siblings, 0 replies; 6+ messages in thread
From: Fam Zheng @ 2016-07-05  9:09 UTC (permalink / raw)
  To: Paolo Bonzini; +Cc: qemu-devel, kwolf, qemu-block

On Mon, 07/04 19:09, Paolo Bonzini wrote:
> Hi,
> 
> I meant to send these close to soft freeze because the final patch is
> prone to conflicts, but it turns out I didn't send them.  Still they're
> pretty mechanical so I'm shooting them out now.  Any thoughts?

I didn't review patch 3 line by line trusting coccinelle and the compiler,
otherwise the series looks good to me and the parameter moving is particularly
nice! Thanks,

Reviewed-by: Fam Zheng <famz@redhat.com>

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [Qemu-devel] [PATCH for-2.7? 0/3] Two coroutine tweaks
  2016-07-04 17:09 [Qemu-devel] [PATCH for-2.7? 0/3] Two coroutine tweaks Paolo Bonzini
                   ` (3 preceding siblings ...)
  2016-07-05  9:09 ` [Qemu-devel] [PATCH for-2.7? 0/3] Two coroutine tweaks Fam Zheng
@ 2016-07-08 15:50 ` Kevin Wolf
  4 siblings, 0 replies; 6+ messages in thread
From: Kevin Wolf @ 2016-07-08 15:50 UTC (permalink / raw)
  To: Paolo Bonzini; +Cc: qemu-devel, famz, qemu-block

Am 04.07.2016 um 19:09 hat Paolo Bonzini geschrieben:
> Hi,
> 
> I meant to send these close to soft freeze because the final patch is
> prone to conflicts, but it turns out I didn't send them.  Still they're
> pretty mechanical so I'm shooting them out now.  Any thoughts?

Yes, this makes a lot of sense.

Thanks, rebased (you were right about the conflicts) and applied to the
block branch.

Kevin

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2016-07-08 15:51 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2016-07-04 17:09 [Qemu-devel] [PATCH for-2.7? 0/3] Two coroutine tweaks Paolo Bonzini
2016-07-04 17:09 ` [Qemu-devel] [PATCH 1/3] coroutine: use QSIMPLEQ instead of QTAILQ Paolo Bonzini
2016-07-04 17:10 ` [Qemu-devel] [PATCH 2/3] test-coroutine: prepare for the next patch Paolo Bonzini
2016-07-04 17:10 ` [Qemu-devel] [PATCH 3/3] coroutine: move entry argument to qemu_coroutine_create Paolo Bonzini
2016-07-05  9:09 ` [Qemu-devel] [PATCH for-2.7? 0/3] Two coroutine tweaks Fam Zheng
2016-07-08 15:50 ` Kevin Wolf

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).