From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:38048) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WYayk-0007EJ-0g for qemu-devel@nongnu.org; Fri, 11 Apr 2014 08:54:52 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1WYayd-00069y-R7 for qemu-devel@nongnu.org; Fri, 11 Apr 2014 08:54:45 -0400 Received: from mx1.redhat.com ([209.132.183.28]:20266) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WYayd-00069t-GM for qemu-devel@nongnu.org; Fri, 11 Apr 2014 08:54:39 -0400 Received: from int-mx12.intmail.prod.int.phx2.redhat.com (int-mx12.intmail.prod.int.phx2.redhat.com [10.5.11.25]) by mx1.redhat.com (8.14.4/8.14.4) with ESMTP id s3BCsckH019596 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK) for ; Fri, 11 Apr 2014 08:54:38 -0400 Date: Fri, 11 Apr 2014 14:54:30 +0200 From: Kevin Wolf Message-ID: <20140411125430.GJ4038@noname.redhat.com> References: <1396961442-24046-1-git-send-email-mreitz@redhat.com> <1396961442-24046-4-git-send-email-mreitz@redhat.com> <20140408151458.GE6262@noname.str.redhat.com> <5346AB8F.7090606@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <5346AB8F.7090606@redhat.com> Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PATCH v2 3/6] qemu-img: Implement commit like QMP List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Max Reitz Cc: qemu-devel@nongnu.org, Stefan Hajnoczi Am 10.04.2014 um 16:32 hat Max Reitz geschrieben: > On 08.04.2014 17:14, Kevin Wolf wrote: > >Am 08.04.2014 um 14:50 hat Max Reitz geschrieben: > >>qemu-img should use QMP commands whenever possible in order to ensure > >>feature completeness of both online and offline image operations. As > >>qemu-img itself has no access to QMP (since this would basically requ= ire > >>just everything being linked into qemu-img), imitate QMP's > >>implementation of block-commit by using commit_active_start() and the= n > >>waiting for the block job to finish. > >Leaves us with the HMP commit command that uses the old bdrv_commit() > >function. I wonder if we can get rid of it by letting the HMP command > >stop the VM, do a live commit, and then restart the VM. > > > >>This new implementation does not empty the snapshot image, as opposed= to > >>the old implementation using bdrv_commit(). However, as QMP's > >>block-commit apparently never did this and as qcow2 (which is probabl= y > >>qemu's standard image format) does not even implement the required > >>function (bdrv_make_empty()), it does not seem necessary. > >In fact, I think since qcow2 has discard support it would actually be > >possible to write a sensible implementation of bdrv_make_empty(). That= 's > >a separate feature, though, and can go in a different patch series. > > > >>Signed-off-by: Max Reitz > >>--- > >> block/Makefile.objs | 2 +- > >> qemu-img.c | 70 ++++++++++++++++++++++++++++++++++++++----= ----------- > >> 2 files changed, 52 insertions(+), 20 deletions(-) > >> > >>diff --git a/block/Makefile.objs b/block/Makefile.objs > >>index fd88c03..2c37e80 100644 > >>--- a/block/Makefile.objs > >>+++ b/block/Makefile.objs > >>@@ -9,6 +9,7 @@ block-obj-y +=3D snapshot.o qapi.o > >> block-obj-$(CONFIG_WIN32) +=3D raw-win32.o win32-aio.o > >> block-obj-$(CONFIG_POSIX) +=3D raw-posix.o > >> block-obj-$(CONFIG_LINUX_AIO) +=3D linux-aio.o > >>+block-obj-y +=3D mirror.o > >> ifeq ($(CONFIG_POSIX),y) > >> block-obj-y +=3D nbd.o nbd-client.o sheepdog.o > >>@@ -22,7 +23,6 @@ endif > >> common-obj-y +=3D stream.o > >> common-obj-y +=3D commit.o > >>-common-obj-y +=3D mirror.o > >> common-obj-y +=3D backup.o > >> iscsi.o-cflags :=3D $(LIBISCSI_CFLAGS) > >>diff --git a/qemu-img.c b/qemu-img.c > >>index 8455994..e86911f 100644 > >>--- a/qemu-img.c > >>+++ b/qemu-img.c > >>@@ -30,6 +30,7 @@ > >> #include "qemu/osdep.h" > >> #include "sysemu/sysemu.h" > >> #include "block/block_int.h" > >>+#include "block/blockjob.h" > >> #include "block/qapi.h" > >> #include > >>@@ -682,12 +683,37 @@ fail: > >> return ret; > >> } > >>+static void dummy_block_job_cb(void *opaque, int ret) > >>+{ > >>+} > >Why don't we need to check the return value? >=20 > I didn't check it, because it was not called =E2=80=93 which apparently > didn't sound strange to me at all. I checked and the much more > interesting fact is that I assumed block_job_complete() would > actually complete the block job without any further need for > aio_poll(); but it doesn't. I'll fix both things (calling aio_poll() > until the CB is called and checking the return value). There is a block_job_cancel_sync(). Should we add a new block_job_complete_sync() instead of adding the polling logic to qemu-img? Kevin