Official QEMU mirror. Please see https://www.qemu.org/contribute/ for how to submit changes to QEMU. Pull Requests are ignored. Please only use release tarballs from the QEMU website. http://www.qemu.org
Find a file
Kevin Wolf 984a32f17e file-posix: Support FUA writes
Until now, FUA was always emulated with a separate flush after the write
for file-posix. The overhead of processing a second request can reduce
performance significantly for a guest disk that has disabled the write
cache, especially if the host disk is already write through, too, and
the flush isn't actually doing anything.

Advertise support for REQ_FUA in write requests and implement it for
Linux AIO and io_uring using the RWF_DSYNC flag for write requests. The
thread pool still performs a separate fdatasync() call. This can be
improved later by using the pwritev2() syscall if available.

As an example, this is how fio numbers can be improved in some scenarios
with this patch (all using virtio-blk with cache=directsync on an nvme
block device for the VM, fio with ioengine=libaio,direct=1,sync=1):

                              | old           | with FUA support
------------------------------+---------------+-------------------
bs=4k, iodepth=1, numjobs=1   |  45.6k iops   |  56.1k iops
bs=4k, iodepth=1, numjobs=16  | 183.3k iops   | 236.0k iops
bs=4k, iodepth=16, numjobs=1  | 258.4k iops   | 311.1k iops

However, not all scenarios are clear wins. On another slower disk I saw
little to no improvment. In fact, in two corner case scenarios, I even
observed a regression, which I however consider acceptable:

1. On slow host disks in a write through cache mode, when the guest is
   using virtio-blk in a separate iothread so that polling can be
   enabled, and each completion is quickly followed up with a new
   request (so that polling gets it), it can happen that enabling FUA
   makes things slower - the additional very fast no-op flush we used to
   have gave the adaptive polling algorithm a success so that it kept
   polling. Without it, we only have the slow write request, which
   disables polling. This is a problem in the polling algorithm that
   will be fixed later in this series.

2. With a high queue depth, it can be beneficial to have flush requests
   for another reason: The optimisation in bdrv_co_flush() that flushes
   only once per write generation acts as a synchronisation mechanism
   that lets all requests complete at the same time. This can result in
   better batching and if the disk is very fast (I only saw this with a
   null_blk backend), this can make up for the overhead of the flush and
   improve throughput. In theory, we could optionally introduce a
   similar artificial latency in the normal completion path to achieve
   the same kind of completion batching. This is not implemented in this
   series.

Compatibility is not a concern for the kernel side of io_uring, it has
supported RWF_DSYNC from the start. However, io_uring_prep_writev2() is
not available before liburing 2.2.

Linux AIO started supporting it in Linux 4.13 and libaio 0.3.111. The
kernel is not a problem for any supported build platform, so it's not
necessary to add runtime checks. However, openSUSE is still stuck with
an older libaio version that would break the build.

We must detect the presence of the writev2 functions in the user space
libraries at build time to avoid build failures.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-ID: <20250307221634.71951-2-kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2025-03-13 17:44:55 +01:00
.github/workflows github: fix config mistake preventing repo lockdown commenting 2022-04-26 16:12:26 +01:00
.gitlab/issue_templates .gitlab/issue_templates: Move suggestions into comments 2022-12-15 15:19:24 +01:00
.gitlab-ci.d gitlab: add a new build_unit job to track build size 2025-03-10 10:30:01 +00:00
accel accel: Prefer cached CpuClass over CPU_GET_CLASS() macro 2025-03-09 17:00:47 +01:00
audio qapi: Move include/qapi/qmp/ to include/qobject/ 2025-02-10 15:33:16 +01:00
authz qapi: Move include/qapi/qmp/ to include/qobject/ 2025-02-10 15:33:16 +01:00
backends qdev: Rename PropertyInfo member @name to @type 2025-03-06 10:30:58 +01:00
block file-posix: Support FUA writes 2025-03-13 17:44:55 +01:00
bsd-user plugins/api: split out binary path/start/end/entry code 2025-03-10 10:30:01 +00:00
chardev chardev: use remoteAddr if the chardev is client 2025-03-05 09:38:09 +04:00
common-user plugins/api: split out binary path/start/end/entry code 2025-03-10 10:30:01 +00:00
configs hw/arm: Add NPCM8XX SoC 2025-02-20 15:22:22 +00:00
contrib plugins: add explicit dependency in functional tests 2025-03-10 10:30:00 +00:00
crypto crypto: Remove qcrypto_tls_session_get_handshake_status 2025-02-14 15:19:03 -03:00
disas disas: Prefer cached CpuClass over CPU_GET_CLASS() macro 2025-03-09 17:00:47 +01:00
docs aspeed queue: 2025-03-10 13:40:35 +08:00
dump qapi: Move include/qapi/qmp/ to include/qobject/ 2025-02-10 15:33:16 +01:00
ebpf ebpf: improve trace event coverage to all key operations 2024-10-28 14:37:25 +08:00
fpu fpu: Build only once 2025-02-25 15:32:57 +00:00
fsdev * pc: Add a description for the i8042 property 2024-10-04 19:28:37 +01:00
gdb-xml target/i386/gdbstub: Expose orig_ax 2024-10-13 10:05:51 -07:00
gdbstub gdbstub: Prefer cached CpuClass over CPU_GET_CLASS() macro 2025-03-09 17:00:47 +01:00
host/include qemu/atomic: Rename atomic128-ldst.h headers using .h.inc suffix 2025-03-08 09:35:06 -08:00
hw Xen queue: 2025-03-11 09:31:36 +08:00
include file-posix: Support FUA writes 2025-03-13 17:44:55 +01:00
io io: Add a read flag for relaxed EOF 2025-02-14 15:19:04 -03:00
libdecnumber libdecnumber/dpd/decimal64: Fix compiler warning from Clang 15 2022-11-11 09:13:52 +01:00
linux-headers linux-headers: Update to Linux v6.14-rc3 2025-03-04 15:42:54 +10:00
linux-user plugins/api: split out binary path/start/end/entry code 2025-03-10 10:30:01 +00:00
migration migration: Prioritize RDMA in ram_save_target_page() 2025-03-10 12:09:24 -03:00
monitor monitor/hmp-cmds.c: Clean up hmp_dumpdtb printf 2025-02-24 15:03:42 +00:00
nbd nbd/server: Support inactive nodes 2025-02-06 14:47:07 +01:00
net tap-linux: Open ipvtap and macvtap 2025-03-10 17:07:16 +08:00
pc-bios pc-bios: Add NPCM8XX vBootrom 2025-02-20 14:20:29 +00:00
plugins plugins/core: make a single build unit 2025-03-10 10:30:01 +00:00
po po: update Italian translation 2024-08-13 19:01:42 +02:00
python python: silence pylint raising-non-exception error 2024-11-25 11:03:14 +01:00
qapi NBD patches for 2025-03-05 2025-03-07 07:37:39 +08:00
qga qga: Don't daemonize before channel is initialized 2025-02-26 13:34:25 +02:00
qobject qapi: Move include/qapi/qmp/ to include/qobject/ 2025-02-10 15:33:16 +01:00
qom qom: Use command line syntax for default values in help 2025-02-25 10:49:23 +01:00
replay qapi: Move include/qapi/qmp/ to include/qobject/ 2025-02-10 15:33:16 +01:00
roms pc-bios: Add NPCM8XX vBootrom 2025-02-20 14:20:29 +00:00
rust rust: pl011: Allow NULL chardev argument to pl011_create() 2025-03-09 11:25:10 +01:00
scripts libvirt-ci: bump to latest for vulkan-tools 2025-03-10 10:30:00 +00:00
scsi qapi: Move include/qapi/qmp/ to include/qobject/ 2025-02-10 15:33:16 +01:00
semihosting semihosting/meson: Build config.o and console.o once 2025-01-17 10:44:03 +00:00
stats qapi: Move include/qapi/qmp/ to include/qobject/ 2025-02-10 15:33:16 +01:00
storage-daemon qapi: Move include/qapi/qmp/ to include/qobject/ 2025-02-10 15:33:16 +01:00
stubs stub: Remove monitor-fd.c 2025-02-25 16:18:11 +01:00
subprojects rust: subprojects: add libc crate 2025-02-25 10:49:23 +01:00
system migration: ram block cpr blockers 2025-03-10 12:09:24 -03:00
target target/xtensa: Move has_work() from CPUClass to SysemuCPUOps 2025-03-09 17:00:47 +01:00
tcg tcg: Remove TCG_TARGET_HAS_{br,set}cond2 from riscv and loongarch64 2025-02-18 08:29:03 -08:00
tests plugins/api: use qemu_target_page_mask() to get value 2025-03-10 10:30:01 +00:00
tools qemu-vmsr-helper: implement --verbose/-v 2024-07-31 13:15:06 +02:00
trace overall: Remove unnecessary g_strdup_printf() calls 2025-02-14 08:49:06 +03:00
ui ui/console-vc: implement DCH (delete) and ICH (insert) commands 2025-03-05 09:34:14 +04:00
util util/iov: Do not assert offset is in iov 2025-03-10 17:07:16 +08:00
.b4-config Add a b4 configuration file 2025-01-13 17:21:46 +01:00
.dir-locals.el Add .dir-locals.el file to configure emacs coding style 2015-10-08 19:46:01 +03:00
.editorconfig editorconfig: update for perl scripts 2025-01-17 10:45:38 +00:00
.exrc qemu: add .exrc 2012-09-07 09:02:44 +03:00
.gdbinit .gdbinit: load QEMU sub-commands when gdb starts 2017-06-07 14:38:45 +01:00
.git-blame-ignore-revs metadata: add .git-blame-ignore-revs 2023-04-04 15:56:44 +01:00
.gitattributes rust: patch bilge-impl to allow compilation with 1.63.0 2024-11-05 14:18:16 +01:00
.gitignore configure: rename --enable-pypi to --enable-download, control subprojects too 2023-06-06 16:30:01 +02:00
.gitlab-ci.yml docs: Document GitLab custom CI/CD variables 2021-07-29 07:56:01 +02:00
.gitmodules meson: subprojects: replace berkeley-{soft,test}float-3 with wraps 2023-06-06 16:30:01 +02:00
.gitpublish Add a git-publish configuration file 2018-03-05 09:03:17 +00:00
.mailmap MAINTAINERS: update email address for Leif Lindholm 2024-12-11 15:31:09 +00:00
.patchew.yml scripts/checkpatch: roll diff tweaking into checkpatch itself 2021-06-25 10:08:33 +01:00
.readthedocs.yml readthodocs: fully specify a build environment 2024-01-12 13:23:48 +00:00
.travis.yml Revert "Remove the unused sh4eb target" 2024-11-04 14:16:11 +01:00
block.c Block layer patches 2025-02-10 13:25:36 -05:00
blockdev-nbd.c qapi: merge common parts of NbdServerOptions and nbd-server-start data 2025-03-04 16:44:48 -06:00
blockdev.c Block layer patches 2025-02-10 13:25:36 -05:00
blockjob.c block: remove unused BLOCK_OP_TYPE_DATAPLANE 2025-02-06 14:51:10 +01:00
configure configure: reenable plugins by default for 32-bit hosts 2025-01-17 10:44:57 +00:00
COPYING COPYING: update from FSF 2008-10-12 17:54:42 +00:00
COPYING.LIB COPYING.LIB: Synchronize the LGPL 2.1 with the version from gnu.org 2019-01-30 11:01:22 +01:00
cpu-common.c cpus: Prefer cached CpuClass over CPU_GET_CLASS() macro 2025-03-09 17:00:47 +01:00
cpu-target.c cpus: Build cpu_exec_[un]realizefn() methods once 2025-03-09 17:00:47 +01:00
event-loop-base.c include: Rename sysemu/ -> system/ 2024-12-20 17:44:56 +01:00
gitdm.config contrib/gitdm: add group map for AMD 2023-03-22 15:08:26 +00:00
hmp-commands-info.hx hmp-commands-info.hx: Add missing info command for stats subcommand 2024-06-30 19:51:44 +03:00
hmp-commands.hx hmp/migration: Fix "migrate" command's documentation 2024-05-08 09:22:37 -03:00
iothread.c include: Rename sysemu/ -> system/ 2024-12-20 17:44:56 +01:00
job-qmp.c qapi job: Elide redundant has_FOO in generated C 2022-12-14 20:04:47 +01:00
job.c block: remove AioContext locking 2023-12-21 22:49:27 +01:00
Kconfig build-sys: Add rust feature option 2024-10-07 16:41:58 +02:00
Kconfig.host pvg: do not enable it on cross-architecture targets 2025-02-25 16:18:11 +01:00
LICENSE tcg/LICENSE: Remove out of date claim about TCG subdirectory licensing 2019-11-11 15:11:21 +01:00
MAINTAINERS MAINTAINERS: remove widely sanctioned entities 2025-03-10 10:30:01 +00:00
Makefile contrib/plugins: remove Makefile for contrib/plugins 2024-11-05 09:13:51 +00:00
meson.build file-posix: Support FUA writes 2025-03-13 17:44:55 +01:00
meson_options.txt meson.build: default to -gsplit-dwarf for debug info 2025-03-07 16:38:41 +01:00
module-common.c all: Clean up includes 2016-02-04 17:41:30 +00:00
os-posix.c os: add an ability to lock memory on_fault 2025-02-12 11:36:01 -05:00
os-win32.c include: Rename sysemu/ -> system/ 2024-12-20 17:44:56 +01:00
page-target.c include/exec: Move TARGET_PAGE_{SIZE,MASK,BITS} to target_page.h 2025-03-07 13:02:54 -08:00
page-vary-common.c Remove qemu-common.h include from most units 2022-04-06 14:31:55 +02:00
page-vary-target.c include/exec: Move TARGET_PAGE_{SIZE,MASK,BITS} to target_page.h 2025-03-07 13:02:54 -08:00
pythondeps.toml Require meson version 1.5.0 2024-10-07 16:41:57 +02:00
qemu-bridge-helper.c qemu-bridge-helper: relocate path to default ACL 2020-09-30 19:11:36 +02:00
qemu-edid.c qemu-edid: Restrict input parameter -d to avoid division by zero 2022-10-12 13:38:15 +02:00
qemu-img-cmds.hx docs/devel/docs: Document .hx file syntax 2024-01-15 17:12:22 +00:00
qemu-img.c qapi: Move include/qapi/qmp/ to include/qobject/ 2025-02-10 15:33:16 +01:00
qemu-io-cmds.c qapi: Move include/qapi/qmp/ to include/qobject/ 2025-02-10 15:33:16 +01:00
qemu-io.c qapi: Move include/qapi/qmp/ to include/qobject/ 2025-02-10 15:33:16 +01:00
qemu-keymap.c qemu-keymap: Release local allocation references 2024-10-03 17:26:05 +03:00
qemu-nbd.c nbd: Defer trace init until after daemonization 2025-03-05 13:00:22 -06:00
qemu-options.hx i386/pc: Support cache topology in -machine for PC machine 2025-02-25 16:18:12 +01:00
qemu.nsi license: Simplify GPL-2.0-or-later license descriptions 2024-09-20 10:11:59 +03:00
qemu.sasl sasl: remove comment about obsolete kerberos versions 2021-06-14 13:28:50 +01:00
README.rst README.rst: add the missing punctuations 2024-07-17 14:04:15 +03:00
replication.c replication: move include out of root directory 2021-05-26 14:49:46 +02:00
trace-events system/dma-helpers.c: Move trace events to system/trace-events 2024-11-19 14:14:13 +00:00
VERSION Open 10.0 development tree 2024-12-10 17:41:17 +00:00
version.rc configure: remove CONFIG_FILEVERSION and CONFIG_PRODUCTVERSION 2021-01-02 21:03:37 +01:00

===========
QEMU README
===========

QEMU is a generic and open source machine & userspace emulator and
virtualizer.

QEMU is capable of emulating a complete machine in software without any
need for hardware virtualization support. By using dynamic translation,
it achieves very good performance. QEMU can also integrate with the Xen
and KVM hypervisors to provide emulated hardware while allowing the
hypervisor to manage the CPU. With hypervisor support, QEMU can achieve
near native performance for CPUs. When QEMU emulates CPUs directly it is
capable of running operating systems made for one machine (e.g. an ARMv7
board) on a different machine (e.g. an x86_64 PC board).

QEMU is also capable of providing userspace API virtualization for Linux
and BSD kernel interfaces. This allows binaries compiled against one
architecture ABI (e.g. the Linux PPC64 ABI) to be run on a host using a
different architecture ABI (e.g. the Linux x86_64 ABI). This does not
involve any hardware emulation, simply CPU and syscall emulation.

QEMU aims to fit into a variety of use cases. It can be invoked directly
by users wishing to have full control over its behaviour and settings.
It also aims to facilitate integration into higher level management
layers, by providing a stable command line interface and monitor API.
It is commonly invoked indirectly via the libvirt library when using
open source applications such as oVirt, OpenStack and virt-manager.

QEMU as a whole is released under the GNU General Public License,
version 2. For full licensing details, consult the LICENSE file.


Documentation
=============

Documentation can be found hosted online at
`<https://www.qemu.org/documentation/>`_. The documentation for the
current development version that is available at
`<https://www.qemu.org/docs/master/>`_ is generated from the ``docs/``
folder in the source tree, and is built by `Sphinx
<https://www.sphinx-doc.org/en/master/>`_.


Building
========

QEMU is multi-platform software intended to be buildable on all modern
Linux platforms, OS-X, Win32 (via the Mingw64 toolchain) and a variety
of other UNIX targets. The simple steps to build QEMU are:


.. code-block:: shell

  mkdir build
  cd build
  ../configure
  make

Additional information can also be found online via the QEMU website:

* `<https://wiki.qemu.org/Hosts/Linux>`_
* `<https://wiki.qemu.org/Hosts/Mac>`_
* `<https://wiki.qemu.org/Hosts/W32>`_


Submitting patches
==================

The QEMU source code is maintained under the GIT version control system.

.. code-block:: shell

   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
qemu-devel@nongnu.org mailing list. All patches submitted must contain
a 'Signed-off-by' line from the author. Patches should follow the
guidelines set out in the `style section
<https://www.qemu.org/docs/master/devel/style.html>`_ of
the Developers Guide.

Additional information on submitting patches can be found online via
the QEMU website:

* `<https://wiki.qemu.org/Contribute/SubmitAPatch>`_
* `<https://wiki.qemu.org/Contribute/TrivialPatches>`_

The QEMU website is also maintained under source control.

.. code-block:: shell

  git clone https://gitlab.com/qemu-project/qemu-web.git

* `<https://www.qemu.org/2017/02/04/the-new-qemu-website-is-up/>`_

A 'git-publish' utility was created to make above process less
cumbersome, and is highly recommended for making regular contributions,
or even just for sending consecutive patch series revisions. It also
requires a working 'git send-email' setup, and by default doesn't
automate everything, so you may want to go through the above steps
manually for once.

For installation instructions, please go to:

*  `<https://github.com/stefanha/git-publish>`_

The workflow with 'git-publish' is:

.. code-block:: shell

  $ git checkout master -b my-feature
  $ # work on new commits, add your 'Signed-off-by' lines to each
  $ git publish

Your patch series will be sent and tagged as my-feature-v1 if you need to refer
back to it in the future.

Sending v2:

.. code-block:: shell

  $ git checkout my-feature # same topic branch
  $ # making changes to the commits (using 'git rebase', for example)
  $ git publish

Your patch series will be sent with 'v2' tag in the subject and the git tip
will be tagged as my-feature-v2.

Bug reporting
=============

The QEMU project uses GitLab issues to track bugs. Bugs
found when running code built from QEMU git or upstream released sources
should be reported via:

* `<https://gitlab.com/qemu-project/qemu/-/issues>`_

If using QEMU via an operating system vendor pre-built binary package, it
is preferable to report bugs to the vendor's own bug tracker first. If
the bug is also known to affect latest upstream code, it can also be
reported via GitLab.

For additional information on bug reporting consult:

* `<https://wiki.qemu.org/Contribute/ReportABug>`_


ChangeLog
=========

For version history and release notes, please visit
`<https://wiki.qemu.org/ChangeLog/>`_ or look at the git history for
more detailed information.


Contact
=======

The QEMU community can be contacted in a number of ways, with the two
main methods being email and IRC:

* `<mailto:qemu-devel@nongnu.org>`_
* `<https://lists.nongnu.org/mailman/listinfo/qemu-devel>`_
* #qemu on irc.oftc.net

Information on additional methods of contacting the community can be
found online via the QEMU website:

* `<https://wiki.qemu.org/Contribute/StartHere>`_