We want object_delete to disappear, and we will do this one class at a
time. Inline it for the qdev case, which we will tackle first.
Signed-off-by: Paolo Bonzini pbonz...@redhat.com
---
hw/qdev.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/hw/qdev.c b/hw/qdev.c
Similarly, a bus holds a reference back to the device, and this will
prevent the device from going away as soon as this reference is counted
properly. To avoid this, move the unrealization of devices to the
unparent callback. This includes recursively unparenting all the buses
and (after the
The data plane thread needs to map guest physical addresses to host
pointers. Normally this is done with cpu_physical_memory_map() but the
function assumes the global mutex is held. The data plane thread does
not touch the global mutex and therefore needs a thread-safe memory
mapping mechanism.
The virtio-blk-data-plane cannot access memory using the usual QEMU
functions since it executes outside the global mutex and the memory APIs
are this time are not thread-safe.
This patch introduces a virtqueue module based on the kernel's vhost
vring code. The trick is that we map guest memory
virtio-blk-data-plane is a subset implementation of virtio-blk. It only
handles read, write, and flush requests. It does this using a dedicated
thread that executes an epoll(2)-based event loop and processes I/O
using Linux AIO.
This approach performs very well but can be used for raw image
The qemu_iovec_concat() function copies a subset of a QEMUIOVector. The
new qemu_iovec_concat_iov() function does the same for a iov/cnt pair.
It is easy to define qemu_iovec_concat() in terms of
qemu_iovec_concat_iov(). The existing code is mostly unchanged, except
for the assertion src-size =
The virtio-blk-data-plane feature is easy to integrate into
hw/virtio-blk.c. The data plane can be started and stopped similar to
vhost-net.
Users can take advantage of the virtio-blk-data-plane feature using the
new -device virtio-blk-pci,x-data-plane=on property.
The x-data-plane name was
On Wed, Dec 05, 2012 at 05:21:50PM -0200, Eduardo Habkost wrote:
On Wed, Dec 05, 2012 at 11:52:29AM -0600, mdroth wrote:
On Tue, Dec 04, 2012 at 05:34:42PM -0200, Eduardo Habkost wrote:
[...]
diff --git a/qapi/string-input-visitor.c b/qapi/string-input-visitor.c
index 497eb9a..74fe395
The iov_discard() function removes data from the front or back of the
vector. This is useful when peeling off header/footer structs.
Signed-off-by: Stefan Hajnoczi stefa...@redhat.com
---
iov.c | 41 +
iov.h | 13 +
2 files changed, 54
Now that the unparent callbacks are complete, we can correctly account
more missing references.
Signed-off-by: Paolo Bonzini pbonz...@redhat.com
---
hw/qdev.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/hw/qdev.c b/hw/qdev.c
index d7f1545..62b6c14 100644
--- a/hw/qdev.c
+++
Outside the safety of the global mutex we need to poll on file
descriptors. I found epoll(2) is a convenient way to do that, although
other options could replace this module in the future (such as an
AioContext-based loop or glib's GMainLoop).
One important feature of this small event loop
Each device has a reference through the BusChild. This reference
was not accounted for, add it now.
Signed-off-by: Paolo Bonzini pbonz...@redhat.com
---
hw/qdev.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/hw/qdev.c b/hw/qdev.c
index e758131..87dfcb5 100644
--- a/hw/qdev.c
+++
The virtio-blk-data-plane feature only works with Linux AIO. Therefore
add a ./configure option and necessary checks to implement this
dependency.
Signed-off-by: Stefan Hajnoczi stefa...@redhat.com
---
configure | 21 +
1 file changed, 21 insertions(+)
diff --git
Oops, I forgot. The contents are OK but 'git am' didn't like the patch. patch
had to use fuzz. This may need to be rebased to latest master.
-Eric
-Original Message-
From: qemu-devel-bounces+ericj=mips@nongnu.org [mailto:qemu-devel-
bounces+ericj=mips@nongnu.org] On Behalf
Signed-off-by: Stefan Hajnoczi stefa...@redhat.com
---
tests/test-iov.c | 129 +++
1 file changed, 129 insertions(+)
diff --git a/tests/test-iov.c b/tests/test-iov.c
index cbe7a89..7997fb5 100644
--- a/tests/test-iov.c
+++ b/tests/test-iov.c
@@
qdev_free and qbus_free have to do unparent+unref, because nobody else
drops the initial reference (the one included by object_initialize)
before them.
For device_init_func and do_device_add, this is trivially correct,
since the DeviceState goes out of scope.
For qdev_create, qdev_try_create and
This series makes the ref_count field of device and bus objects actually
match the number of references that the objects have. Once this is done,
qdev_free and qbus_free are equivalent to simply object_unparent, and
object_delete can go.
Patches 1-3 fix some warts in the last minute patches that
This is now unused.
Signed-off-by: Paolo Bonzini pbonz...@redhat.com
---
include/qemu/object.h | 17 ++---
qom/object.c | 7 ---
2 files changed, 2 insertions(+), 22 deletions(-)
diff --git a/include/qemu/object.h b/include/qemu/object.h
index ed1f47f..c5169ee 100644
This series adds the -device virtio-blk-pci,x-data-plane=on property that
enables a high performance I/O codepath. A dedicated thread is used to process
virtio-blk requests outside the global mutex and without going through the QEMU
block layer.
Khoa Huynh k...@us.ibm.com reported an increase
A device will never be finalized as long as it has a reference from
other devices that sit on its buses. To ensure that the references
go away, deassociate a bus from its children in the unparent callback
for the bus.
Signed-off-by: Paolo Bonzini pbonz...@redhat.com
---
hw/qdev.c | 37
The IOQueue has a pool of iocb structs and a function to add new
read/write requests. Multiple requests can be added before calling the
submit function to actually tell the host kernel to begin I/O. This
allows callers to batch requests and submit them in one go.
The actual I/O is performed
The raw_get_aio_fd() function allows virtio-blk-data-plane to get the
file descriptor of a raw image file with Linux AIO enabled. This
interface is really a layering violation that can be resolved once the
block layer is able to run outside the global mutex - at that point
virtio-blk-data-plane
After the previous patch, creation of heap-allocated buses can
use qbus_init and does not need to hack the free callback.
Signed-off-by: Paolo Bonzini pbonz...@redhat.com
---
hw/pci.c| 11 +--
hw/pci.h| 5 -
hw/sysbus.c | 6 ++
3 files changed, 7 insertions(+), 15
CPUs are never added to the composition tree, so delete is achieved
simply by removing the last references to them.
Signed-off-by: Paolo Bonzini pbonz...@redhat.com
---
linux-user/syscall.c | 2 +-
target-i386/helper.c | 4 ++--
target-sparc/cpu.c | 2 +-
3 files changed, 4 insertions(+), 4
Avoid that the object disappears after it's deleted from the QOM
composition tree, in case that was the only reference to it.
Signed-off-by: Paolo Bonzini pbonz...@redhat.com
---
qom/object.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/qom/object.c b/qom/object.c
index 0739aa2..ecdf164
On 2012-12-05 10:38, Peter Maydell wrote:
I don't think we should block this patch on that general
cleanup, though. All the sign extensions in target-mips/translate.c
are done in the double-shift way, so this is consistent with
the existing code.
Fair enough. The original can have my
Rebased to commit 16c6c80ac3a772b42a87b77dfdf0fdac7c607b0e pull from
git://repo.or.cz/qemu/rth.git s390-reorg
The only substantive change from the last ping is to put my name in
the maintainer slot for the code.
r~
Richard Henderson (152):
tcg: Add TCGV_IS_UNUSED_*
target-s390:
hey Eric,
the patch does not have to be rebased since no changes have been made to
these files since I created the patch.
You can try it on clean master with:
wget -O shilo.diff http://patchwork.ozlabs.org/patch/203744/mbox
git am ./shilo.diff
You probably have the previous (not yet committed)
On 06.12.2012, at 00:11, Richard Henderson wrote:
Rebased to commit 16c6c80ac3a772b42a87b77dfdf0fdac7c607b0e pull from
git://repo.or.cz/qemu/rth.git s390-reorg
The only substantive change from the last ping is to put my name in
the maintainer slot for the code.
Acked-by: Alexander Graf
Sorry mail wrap issue.
-Original Message-
From: Jovanovic, Petar
Sent: Wednesday, December 05, 2012 3:31 PM
To: Johnson, Eric; qemu-devel@nongnu.org
Cc: blauwir...@gmail.com; rth7...@gmail.com; afaer...@suse.de;
aurel...@aurel32.net
Subject: RE: [Qemu-devel] [PATCH v2] target-mips:
The select() do not masked signal in os_host_main_loop_wait.
For example, qemu-timer.c gives SIGALRM regularly.
Even if a buffer of select is empty,
I think that it is a problem that this select() does exit by this SIGALRM.
--
E.Furukawa
-Original Message-
From: Stefan Hajnoczi
在 2012-12-05三的 09:53 +0100，Jan Kiszka写道：
On 2012-12-05 01:51, li guang wrote:
在 2012-12-04二的 11:26 +，Peter Maydell写道：
On 4 December 2012 11:11, Jan Kiszka jan.kis...@siemens.com wrote:
On 2012-12-04 11:23, Peter Maydell wrote:
Doesn't this break the use of this function in
在 2012-12-06四的 10:08 +0800，li guang写道：
在 2012-12-05三的 09:53 +0100，Jan Kiszka写道：
On 2012-12-05 01:51, li guang wrote:
在 2012-12-04二的 11:26 +，Peter Maydell写道：
On 4 December 2012 11:11, Jan Kiszka jan.kis...@siemens.com wrote:
On 2012-12-04 11:23, Peter Maydell wrote:
Doesn't this
Public bug reported:
This seems inexplicable but has been verified in a number of ways.
Given the following domain XML (from libvirt):
devices
emulator/usr/bin/kvm/emulator
disk type='block' device='disk'
driver name='qemu' type='raw' cache='none'/
source
try to make breakpoint related functions
more cleaner.
originally, these functions implicit use
0,1,2,3 as name of breakpoint, really
hard to understand, so name them readable,
and also refactor the statement to
check these types for clean reason.
v2 changes:
1. split 'breakpoint_enabled' into
Signed-off-by: liguang lig.f...@cn.fujitsu.com
---
target-i386/cpu.h |7 +++
1 files changed, 7 insertions(+), 0 deletions(-)
diff --git a/target-i386/cpu.h b/target-i386/cpu.h
index 90ef1ff..29245d1 100644
--- a/target-i386/cpu.h
+++ b/target-i386/cpu.h
@@ -231,6 +231,13 @@
#define
Signed-off-by: liguang lig.f...@cn.fujitsu.com
---
target-i386/cpu.h | 15 +--
1 files changed, 13 insertions(+), 2 deletions(-)
diff --git a/target-i386/cpu.h b/target-i386/cpu.h
index 29245d1..3646128 100644
--- a/target-i386/cpu.h
+++ b/target-i386/cpu.h
@@ -996,9 +996,20 @@ int
Signed-off-by: liguang lig.f...@cn.fujitsu.com
---
target-i386/helper.c | 74 +---
target-i386/machine.c |5 ++-
target-i386/misc_helper.c |4 +-
target-i386/seg_helper.c |6 ++--
4 files changed, 57 insertions(+), 32 deletions(-)
On 2012/12/05 20:49, Luiz Capitulino wrote:
On Wed, 05 Dec 2012 17:41:08 +0900
Tomoki Sekiyama tomoki.sekiyama...@hitachi.com wrote:
On 2012/12/05 1:51, Luiz Capitulino wrote:
On Fri, 30 Nov 2012 22:33:57 +0900
Tomoki Sekiyama tomoki.sekiyama...@hitachi.com wrote:
Adds sample hook scripts
I tried the head today.
Qemu crashes in the same way as before.
Hi,
And qemu error output is:
qemu: /home/akorolev/qemu-kvm/exec.c:2255: register_subpage: Assertion
`existing-mr-subpage || existing-mr == io_mem_unassigned' failed.
Guest OS is Centos 5.5 and log is pretty boring, as qemu
-Original Message-
From: Alexander Graf [mailto:ag...@suse.de]
Sent: Sunday, December 02, 2012 7:20 PM
To: Yin Olivia-R63875
Cc: qemu-devel@nongnu.org; qemu-...@nongnu.org
Subject: Re: [Qemu-ppc] [RFC PATCH v7 1/4] use image_file_reset to reload
initrd
Missing patch description
The current model of loader copy rom blobs and kept in memory until a reset
occurs and waste host memory.
This serial of patches uses private reset handlers to load uimage/initrd/vmlinux
from hard disk on reset, which could make loader framework more dynamic and
reduce the memory consumption of
Signed-off-by: Olivia Yin hong-hua@freescale.com
---
hw/loader.c | 24
hw/loader.h |6 ++
2 files changed, 30 insertions(+), 0 deletions(-)
diff --git a/hw/loader.c b/hw/loader.c
index ba01ca6..f62aa7c 100644
--- a/hw/loader.c
+++ b/hw/loader.c
@@ -86,6
---
hw/loader.c | 57 +
hw/loader.h |8
2 files changed, 53 insertions(+), 12 deletions(-)
diff --git a/hw/loader.c b/hw/loader.c
index f62aa7c..9a65ed3 100644
--- a/hw/loader.c
+++ b/hw/loader.c
@@ -457,15 +457,15 @@ static
---
hw/elf_ops.h | 45 -
hw/loader.c | 11 +++
hw/loader.h |9 +
3 files changed, 60 insertions(+), 5 deletions(-)
diff --git a/hw/elf_ops.h b/hw/elf_ops.h
index 531a425..eddaa5a 100644
--- a/hw/elf_ops.h
+++ b/hw/elf_ops.h
@@
The previous patches reload uimage/initrd/vmlinux reset. Other firmware images
stay in the memory and don't need reload on reset.
So rom_reset() could be removed.
Signed-off-by: Olivia Yin hong-hua@freescale.com
---
hw/loader.c | 24 +++-
1 files changed, 3
On Wed, Dec 5, 2012 at 1:50 AM, Jan Kiszka jan.kis...@siemens.com wrote:
On 2012-12-05 06:34, Cam Macdonell wrote:
static void ivshmem_write_config(PCIDevice *pci_dev, uint32_t address,
uint32_t val, int len)
{
+bool is_enabled, was_enabled =
Michael S. Tsirkin m...@redhat.com writes:
Add sanity check to address the following concern:
On Wed, Dec 05, 2012 at 09:47:22AM +1030, Rusty Russell wrote:
All we need is the index of the request; the rest can be re-read from
the ring.
The terminology I used here was loose, indeed.
We need
On Thu, Dec 6, 2012 at 1:10 PM, Cam Macdonell c...@cs.ualberta.ca wrote:
On Wed, Dec 5, 2012 at 1:50 AM, Jan Kiszka jan.kis...@siemens.com wrote:
On 2012-12-05 06:34, Cam Macdonell wrote:
static void ivshmem_write_config(PCIDevice *pci_dev, uint32_t address,
From: Liu Ping Fan pingf...@linux.vnet.ibm.com
This logic has been integrated into pci core, so remove it.
Signed-off-by: Liu Ping Fan pingf...@linux.vnet.ibm.com
Signed-off-by: Cam Macdonell c...@cs.ualberta.ca
---
hw/ivshmem.c |1 -
1 files changed, 0 insertions(+), 1 deletions(-)
diff
From: Liu Ping Fan pingf...@linux.vnet.ibm.com
Using irqfd, so we can avoid switch between kernel and user when
VMs interrupts each other.
Signed-off-by: Liu Ping Fan pingf...@linux.vnet.ibm.com
Signed-off-by: Cam Macdonell c...@cs.ualberta.ca
---
hw/ivshmem.c | 54
The new functions are opts_accepts_any() and find_desc_by_name(), which
are also going to be used by qemu_opts_validate() (see next commit).
This also makes opt_set() slightly more readable.
Signed-off-by: Luiz Capitulino lcapitul...@redhat.com
Signed-off-by: Dong Xu Wang
While id is NULL, qemu_opts_create can not fail, so ignore
errors is fine.
Signed-off-by: Dong Xu Wang wdon...@linux.vnet.ibm.com
---
qemu-option.c |9 +
qemu-option.h |1 +
2 files changed, 10 insertions(+), 0 deletions(-)
diff --git a/qemu-option.c b/qemu-option.c
index
Use opts_accepts_any() and find_desc_by_name().
Signed-off-by: Luiz Capitulino lcapitul...@redhat.com
Signed-off-by: Dong Xu Wang wdon...@linux.vnet.ibm.com
---
qemu-option.c | 14 +++---
1 files changed, 3 insertions(+), 11 deletions(-)
diff --git a/qemu-option.c b/qemu-option.c
Patch 1-3 are from Luiz, added Markus's comments, discussion could be found
here:
http://lists.nongnu.org/archive/html/qemu-devel/2012-07/msg02716.html
Patch 3 was changed according Paolo's comments.
Patch 4-5: because qemu_opts_create can not fail while id is null, so create
function
We will use qemu_opts_create_nofail function, it can make code
more readable.
Signed-off-by: Dong Xu Wang wdon...@linux.vnet.ibm.com
---
blockdev.c |2 +-
hw/watchdog.c |2 +-
qemu-config.c |4 ++--
qemu-img.c |2 +-
qemu-sockets.c | 16
vl.c
qemu_opts_print has no user now, so can re-write the function safely.
qemu_opts_print will be used while using qemu-img create, it will
produce the same output as previous code.
The behavior of this function has changed:
1. Print every possible option, whether a value has been set or not.
2.
It will set opt-str in qemu_opt_set_bool, without opt-str, there
will be some potential bugs.
These are uses of opt-str, and what happens when it isn't set:
* qemu_opt_get(): returns NULL, which means not set. Bug can bite
when value isn't the default value.
* qemu_opt_parse(): passes NULL
It will introduce a new file format: add-cow.
The add-cow file format makes it possible to perform copy-on-write on top of
a raw disk image. When we know that no backing file clusters remain visible
(e.g. we have streamed the entire image and copied all data from the backing
file), then it is
This patch will use qemu-iotests to test add-cow file format.
Signed-off-by: Dong Xu Wang wdon...@linux.vnet.ibm.com
---
tests/qemu-iotests/017 |2 +-
tests/qemu-iotests/020 |2 +-
tests/qemu-iotests/common|6 ++
tests/qemu-iotests/common.rc | 15 ++-
add-cow file format core code. It use block-cache.c as cache code.
It lacks of snapshot_blkdev support.
v16-v17:
1) Use stringify.
v15-v16:
1) Judge if opts is null in add_cow_create function.
Signed-off-by: Dong Xu Wang wdon...@linux.vnet.ibm.com
---
block/Makefile.objs |1 +
We will re-use qcow2-cache as block layer common cache code,
so change its name and made some changes, define a struct named
BlockTableType, pass BlockTableType and table size parameters to
block cache initialization function.
Signed-off-by: Dong Xu Wang wdon...@linux.vnet.ibm.com
---
Public bug reported:
QEMU 1.3.0 on OpenBSD now crashes with an error as shown below and the
command line params do not seem to matter.
assertion QLIST_EMPTY(bs-tracked_requests) failed: file block.c,
line 1220, function bdrv_drain_all
#1 0x030d1bce24aa in abort () at
Signed-off-by: Dong Xu Wang wdon...@linux.vnet.ibm.com
---
qemu-option.c | 22 ++
qemu-option.h |1 +
2 files changed, 23 insertions(+), 0 deletions(-)
diff --git a/qemu-option.c b/qemu-option.c
index 1303188..94557cf 100644
--- a/qemu-option.c
+++ b/qemu-option.c
@@
This patch will create 4 functions, count_opts_list, append_opts_list,
free_opts_list and print_opts_list, they will used in following commits.
v6-v7):
1) Fix typo: enouth-enough.
v5-v6):
1) allocate enough space in append_opts_list function.
Signed-off-by: Dong Xu Wang
Signed-off-by: Dong Xu Wang wdon...@linux.vnet.ibm.com
---
qemu-option.c | 285 -
qemu-option.h | 32 ---
2 files changed, 0 insertions(+), 317 deletions(-)
diff --git a/qemu-option.c b/qemu-option.c
index fb912c1..250a094 100644
---
Any suggestion? Or new design idea for this?
Thanks
Pingfan
On Sun, Nov 25, 2012 at 10:02 AM, Liu Ping Fan qemul...@gmail.com wrote:
From: Liu Ping Fan pingf...@linux.vnet.ibm.com
v1:
https://lists.gnu.org/archive/html/qemu-devel/2012-07/msg03312.html
v2:
We will use path_has_protocol outside block.c, so just make it public.
Signed-off-by: Dong Xu Wang wdon...@linux.vnet.ibm.com
Reviewed-by: Michael Roth mdr...@linux.vnet.ibm.com
---
block.c |2 +-
block.h |1 +
2 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/block.c
Il 05/12/2012 21:47, Stefan Hajnoczi ha scritto:
+
+/* Block until pending requests have completed
+ *
+ * The vring continues to be serviced so ensure no new requests will be added
+ * to avoid races.
+ */
+void virtio_blk_data_plane_drain(VirtIOBlockDataPlane *s)
+{
+
This patch will use QemuOpts related functions in block layer, add
a member bdrv_create_options to BlockDriver struct, it will return
a QemuOptsList pointer, which includes the image format's create
options.
And create options's primary consumer is block creating related functions,
so modify them
Document for add-cow format, the usage and spec of add-cow are introduced.
Signed-off-by: Dong Xu Wang wdon...@linux.vnet.ibm.com
---
docs/specs/add-cow.txt | 154
1 files changed, 154 insertions(+), 0 deletions(-)
create mode 100644
Make qed_read_string function to a common interface, so move it to block.c.
Signed-off-by: Dong Xu Wang wdon...@linux.vnet.ibm.com
---
block.c | 27 +++
block.h |2 ++
block/qed.c | 34 --
3 files changed, 33 insertions(+),
I've been looking at handling of msix masking
in qemu. It looks like all of virtio,vfio and
device assignment implemented their own
similar but slightly different thing.
So I am inclined to move this handling to common
code in msix.c, adding irqfd support right there.
While doing this rework, one
On Thu, Dec 06, 2012 at 04:33:06PM +1030, Rusty Russell wrote:
Michael S. Tsirkin m...@redhat.com writes:
Add sanity check to address the following concern:
On Wed, Dec 05, 2012 at 09:47:22AM +1030, Rusty Russell wrote:
All we need is the index of the request; the rest can be re-read from
101 - 174 of 174 matches
Mail list logo
