[Qemu-devel] [PATCH] tools: Use real async.c instead of stubs

[Qemu-devel] [PATCH] tools: Use real async.c instead of stubs

[Kevin Wolf]

It's wrong to call BHs directly, even in tools. The only operations that
schedule BHs are called in a loop that (indirectly) contains a call to
qemu_bh_poll anyway, so we're not losing the scheduled BHs: Tools either use
synchronous functions, which are guaranteed to have completed (including any
BHs) when they return; or if they use asynchronous functions, they need to call
qemu_aio_wait() or similar functions already today.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 Makefile.objs |    4 ++--
 qemu-tool.c   |   47 ++++-------------------------------------------
 2 files changed, 6 insertions(+), 45 deletions(-)

diff --git a/Makefile.objs b/Makefile.objs
index c144df1..fc369dc 100644
--- a/Makefile.objs
+++ b/Makefile.objs
@@ -13,7 +13,7 @@ oslib-obj-$(CONFIG_POSIX) += oslib-posix.o
 #######################################################################
 # block-obj-y is code used by both qemu system emulation and qemu-img
 
-block-obj-y = cutils.o cache-utils.o qemu-malloc.o qemu-option.o module.o
+block-obj-y = cutils.o cache-utils.o qemu-malloc.o qemu-option.o module.o async.o
 block-obj-y += nbd.o block.o aio.o aes.o qemu-config.o
 block-obj-$(CONFIG_POSIX) += posix-aio-compat.o
 block-obj-$(CONFIG_LINUX_AIO) += linux-aio.o
@@ -63,7 +63,7 @@ common-obj-y = $(block-obj-y) blockdev.o
 common-obj-y += $(net-obj-y)
 common-obj-y += $(qobject-obj-y)
 common-obj-$(CONFIG_LINUX) += $(fsdev-obj-$(CONFIG_LINUX))
-common-obj-y += readline.o console.o cursor.o async.o qemu-error.o
+common-obj-y += readline.o console.o cursor.o qemu-error.o
 common-obj-y += $(oslib-obj-y)
 common-obj-$(CONFIG_WIN32) += os-win32.o
 common-obj-$(CONFIG_POSIX) += os-posix.o
diff --git a/qemu-tool.c b/qemu-tool.c
index 392e1c9..d45840d 100644
--- a/qemu-tool.c
+++ b/qemu-tool.c
@@ -56,53 +56,10 @@ void monitor_print_filename(Monitor *mon, const char *filename)
 {
 }
 
-void async_context_push(void)
-{
-}
-
-void async_context_pop(void)
-{
-}
-
-int get_async_context_id(void)
-{
-    return 0;
-}
-
 void monitor_protocol_event(MonitorEvent event, QObject *data)
 {
 }
 
-QEMUBH *qemu_bh_new(QEMUBHFunc *cb, void *opaque)
-{
-    QEMUBH *bh;
-
-    bh = qemu_malloc(sizeof(*bh));
-    bh->cb = cb;
-    bh->opaque = opaque;
-
-    return bh;
-}
-
-int qemu_bh_poll(void)
-{
-    return 0;
-}
-
-void qemu_bh_schedule(QEMUBH *bh)
-{
-    bh->cb(bh->opaque);
-}
-
-void qemu_bh_cancel(QEMUBH *bh)
-{
-}
-
-void qemu_bh_delete(QEMUBH *bh)
-{
-    qemu_free(bh);
-}
-
 int qemu_set_fd_handler2(int fd,
                          IOCanReadHandler *fd_read_poll,
                          IOHandler *fd_read,
@@ -111,3 +68,7 @@ int qemu_set_fd_handler2(int fd,
 {
     return 0;
 }
+
+void qemu_notify_event(void)
+{
+}
-- 
1.7.2.3

[Qemu-devel] Re: [PATCH] tools: Use real async.c instead of stubs

[Stefan Hajnoczi]

On Wed, Feb 23, 2011 at 11:54 AM, Kevin Wolf <kwolf@redhat.com> wrote:
> It's wrong to call BHs directly, even in tools. The only operations that
> schedule BHs are called in a loop that (indirectly) contains a call to
> qemu_bh_poll anyway, so we're not losing the scheduled BHs: Tools either use
> synchronous functions, which are guaranteed to have completed (including any
> BHs) when they return; or if they use asynchronous functions, they need to call
> qemu_aio_wait() or similar functions already today.
>
> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
> ---
>  Makefile.objs |    4 ++--
>  qemu-tool.c   |   47 ++++-------------------------------------------
>  2 files changed, 6 insertions(+), 45 deletions(-)

We discussed this on IRC and I think it makes sense.  Review from
others would be appreciated.

Reviewed-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>