rpms/qemu-kvm
7
0
Fork 0
Code Issues Pull Requests Releases Activity

* Mon May 19 2025 Miroslav Rezanina <mrezanin@redhat.com> - 10.0.0-3

Browse Source 
- kvm-migration-postcopy-Spatial-locality-page-hint-for-pr.patch [RHEL-85635]
- kvm-meson-configure-add-valgrind-option-en-dis-able-valg.patch [RHEL-88457]
- kvm-distro-add-an-explicit-valgrind-devel-build-dep.patch [RHEL-88457]
- kvm-Allow-guest-get-load-QGA-command.patch [RHEL-91219]
- Resolves: RHEL-85635
  (Video stuck about 1 min after switchover phase when play one video during postcopy-preempt migration)
- Resolves: RHEL-88457
  (qemu inadvertantly built with valgrind coroutine stack debugging on x86_64)
- Resolves: RHEL-91219
  ([qemu-guest-agent] Enable 'guest-get-load' by default [RHEL-10])
This commit is contained in:
Miroslav Rezanina 2025-05-19 02:23:51 -04:00
parent 9f0b20db5e
commit 9c8774ec77
5 changed files with 439 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

71
kvm-docs-Don-t-define-duplicate-label-in-qemu-block-driv.patch Normal file
View File

@ -0,0 +1,71 @@
From c73cd0b322ad22a4ba47035ba87a0e83ff851b6a Mon Sep 17 00:00:00 2001
From: Peter Maydell <peter.maydell@linaro.org>
Date: Thu, 1 May 2025 10:31:26 +0100
Subject: [PATCH] docs: Don't define duplicate label in
qemu-block-drivers.rst.inc
Sphinx requires that labels within documents are unique across the
whole manual. This is because the "create a hyperlink" directive
specifies only the name of the label, not a filename+label. Some
Sphinx versions will warn about duplicate labels, but even if there
is no warning there is still an ambiguity and no guarantee that the
hyperlink will be created to the right target.
For QEMU this is awkward, because we have various .rst.inc fragments
which we include into multiple .rst files. If you define a label in
the .rst.inc file then it will be a duplicate label. We have mostly
worked around this by not putting labels into those .rst.inc files,
or by adding "insert a label" functionality into the hxtool extension
(see commit 1eeb432a953b0 "doc/sphinx/hxtool.py: add optional label
argument to SRST directive").
Unfortunately in commit 7f6314427e78 ("docs/devel: add a codebase
section") we accidentally added a duplicate label, because not all
Sphinx versions warn about the mistake.
In this case the link was only from the developer docs codebase
summary, so as the simplest fix for the stable branch, we drop
the link entirely.
Cc: qemu-stable@nongnu.org
Fixes: 1eeb432a953b0 "doc/sphinx/hxtool.py: add optional label argument to SRST directive"
Reported-by: Dario Faggioli <dfaggioli@suse.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Acked-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-id: 20250501093126.716667-1-peter.maydell@linaro.org
Signed-off-by: Miroslav Rezanina <mrezanin@redhat.com>
---
docs/devel/codebase.rst | 2 +-
docs/system/qemu-block-drivers.rst.inc | 2 --
2 files changed, 1 insertion(+), 3 deletions(-)
diff --git a/docs/devel/codebase.rst b/docs/devel/codebase.rst
index ef98578296..085da10a22 100644
--- a/docs/devel/codebase.rst
+++ b/docs/devel/codebase.rst
@@ -116,7 +116,7 @@ yet, so sometimes the source code is all you have.
* `monitor <https://gitlab.com/qemu-project/qemu/-/tree/master/monitor>`_:
`Monitor <QEMU monitor>` implementation (HMP & QMP).
* `nbd <https://gitlab.com/qemu-project/qemu/-/tree/master/nbd>`_:
- QEMU `NBD (Network Block Device) <nbd>` server.
+ QEMU NBD (Network Block Device) server.
* `net <https://gitlab.com/qemu-project/qemu/-/tree/master/net>`_:
Network (host) support.
* `pc-bios <https://gitlab.com/qemu-project/qemu/-/tree/master/pc-bios>`_:
diff --git a/docs/system/qemu-block-drivers.rst.inc b/docs/system/qemu-block-drivers.rst.inc
index cfe1acb78a..384e95ba76 100644
--- a/docs/system/qemu-block-drivers.rst.inc
+++ b/docs/system/qemu-block-drivers.rst.inc
@@ -500,8 +500,6 @@ What you should *never* do:
- expect it to work when loadvm'ing
- write to the FAT directory on the host system while accessing it with the guest system
-.. _nbd:
-
NBD access
~~~~~~~~~~
--
2.39.3

107
kvm-meson-configure-add-valgrind-option-en-dis-able-valg.patch Normal file
View File

@ -0,0 +1,107 @@
From 8227a9534bb09d202441b3e554da53815be66a28 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Fri, 25 Apr 2025 13:17:12 +0100
Subject: [PATCH 2/4] meson/configure: add 'valgrind' option & --{en,
dis}able-valgrind flag
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Daniel P. Berrangé <berrange@redhat.com>
RH-MergeRequest: 360: distro: add an explicit valgrind-devel build dep
RH-Jira: RHEL-88457
RH-Acked-by: Thomas Huth <thuth@redhat.com>
RH-Acked-by: Eric Blake <eblake@redhat.com>
RH-Commit: [1/2] 55ab738650e95ff0e951897001d9246a725ee699 (berrange/centos-src-qemu)
Currently valgrind debugging support for coroutine stacks is enabled
unconditionally when valgrind/valgrind.h is found. There is no way
to disable valgrind support if valgrind.h is present in the build env.
This is bad for distros, as an dependency far down the chain may cause
valgrind.h to become installed, inadvertently enabling QEMU's valgrind
debugging support. It also means if a distro wants valgrind support
there is no way to mandate this.
The solution is to add a 'valgrind' build feature to meson and thus
configure script.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250425121713.1913424-1-berrange@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
(cherry picked from commit 6b1c744ec0d66d6d568f9a156282153fc11a21cf)
---
meson.build | 13 ++++++++++++-
meson_options.txt | 2 ++
scripts/meson-buildoptions.sh | 3 +++
3 files changed, 17 insertions(+), 1 deletion(-)
diff --git a/meson.build b/meson.build
index 0607c1313b..dadd47d362 100644
--- a/meson.build
+++ b/meson.build
@@ -2617,7 +2617,17 @@ config_host_data.set('CONFIG_FSTRIM', qga_fstrim)
# has_header
config_host_data.set('CONFIG_EPOLL', cc.has_header('sys/epoll.h'))
config_host_data.set('CONFIG_LINUX_MAGIC_H', cc.has_header('linux/magic.h'))
-config_host_data.set('CONFIG_VALGRIND_H', cc.has_header('valgrind/valgrind.h'))
+valgrind = false
+if get_option('valgrind').allowed()
+ if cc.has_header('valgrind/valgrind.h')
+ valgrind = true
+ else
+ if get_option('valgrind').enabled()
+ error('valgrind requested but valgrind.h not found')
+ endif
+ endif
+endif
+config_host_data.set('CONFIG_VALGRIND_H', valgrind)
config_host_data.set('HAVE_BTRFS_H', cc.has_header('linux/btrfs.h'))
config_host_data.set('HAVE_DRM_H', cc.has_header('libdrm/drm.h'))
config_host_data.set('HAVE_OPENAT2_H', cc.has_header('linux/openat2.h'))
@@ -4856,6 +4866,7 @@ endif
if host_os == 'darwin'
summary_info += {'ParavirtualizedGraphics support': pvg}
endif
+summary_info += {'valgrind': valgrind}
summary(summary_info, bool_yn: true, section: 'Dependencies')
if host_arch == 'unknown'
diff --git a/meson_options.txt b/meson_options.txt
index ad6996178c..6592d9fb07 100644
--- a/meson_options.txt
+++ b/meson_options.txt
@@ -117,6 +117,8 @@ option('dbus_display', type: 'feature', value: 'auto',
description: '-display dbus support')
option('tpm', type : 'feature', value : 'auto',
description: 'TPM support')
+option('valgrind', type : 'feature', value: 'auto',
+ description: 'valgrind debug support for coroutine stacks')
# Do not enable it by default even for Mingw32, because it doesn't
# work on Wine.
diff --git a/scripts/meson-buildoptions.sh b/scripts/meson-buildoptions.sh
index e9edc8a919..541e1e7a2f 100644
--- a/scripts/meson-buildoptions.sh
+++ b/scripts/meson-buildoptions.sh
@@ -199,6 +199,7 @@ meson_options_help() {
printf "%s\n" ' u2f U2F emulation support'
printf "%s\n" ' uadk UADK Library support'
printf "%s\n" ' usb-redir libusbredir support'
+ printf "%s\n" ' valgrind valgrind debug support for coroutine stacks'
printf "%s\n" ' vde vde network backend support'
printf "%s\n" ' vdi vdi image format support'
printf "%s\n" ' vduse-blk-export'
@@ -528,6 +529,8 @@ _meson_option_parse() {
--disable-ubsan) printf "%s" -Dubsan=false ;;
--enable-usb-redir) printf "%s" -Dusb_redir=enabled ;;
--disable-usb-redir) printf "%s" -Dusb_redir=disabled ;;
+ --enable-valgrind) printf "%s" -Dvalgrind=enabled ;;
+ --disable-valgrind) printf "%s" -Dvalgrind=disabled ;;
--enable-vde) printf "%s" -Dvde=enabled ;;
--disable-vde) printf "%s" -Dvde=disabled ;;
--enable-vdi) printf "%s" -Dvdi=enabled ;;
--
2.39.3

234
kvm-migration-postcopy-Spatial-locality-page-hint-for-pr.patch Normal file
View File

@ -0,0 +1,234 @@
From 2e49bdf20390600286cd2596859feb400a932c44 Mon Sep 17 00:00:00 2001
From: Peter Xu <peterx@redhat.com>
Date: Thu, 24 Apr 2025 18:07:05 -0400
Subject: [PATCH 1/4] migration/postcopy: Spatial locality page hint for
preempt mode
RH-Author: Peter Xu <peterx@redhat.com>
RH-MergeRequest: 357: migration/postcopy: Spatial locality page hint for preempt mode
RH-Jira: RHEL-85635
RH-Acked-by: Juraj Marcin <None>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [1/1] 05f3befe60e1d3f2f8beaaf0e0e577e49a126dfe (peterx/qemu-kvm)
The preempt mode postcopy has been introduced for a while. From latency
POV, it should always win the vanilla postcopy.
However there's one thing missing when preempt mode is enabled right now,
which is the spatial locality hint when there're page requests from the
destination side.
In vanilla postcopy, as long as a page request was unqueued, it will update
the PSS of the precopy background stream, so that after a page request the
background thread will move the pages after whatever was requested. It's
pretty much a natural behavior when there's only one channel anyway, and
one scanner to send the pages.
Preempt mode didn't follow that, because preempt mode has its own channel
and its own PSS (which doesn't linearly scan the guest memory, but
dedicated to resolve page requested from destination). So the page request
process and the background migration process are completely separate.
This patch adds the hint explicitly for preempt mode. With that, whenever
the preempt mode receives a page request on the source, it will service the
remote page fault in the return path, then it'll provide a hint to the
background thread so that we'll start sending the pages right after the
requested ones in the background, assuming the follow up pages have a
higher chance to be accessed later.
NOTE: since the background migration thread and return path thread run
completely concurrently, it doesn't always mean the hint will be applied
every single time. For example, it's possible that the return path thread
receives multiple page requests in a row without the background thread
getting the chance to consume one. In such case, the preempt thread only
provide the hint if the previous hint has been consumed. After all,
there's no point queuing hints when we only have one linear scanner.
This could measureably improve the simple sequential memory access pattern
during postcopy (when preempt is on). For random accesses, I can measure a
slight increase of remote page fault latency from ~500us -> ~600us, that
could be a trade-off to have such hint mechanism, and after all that's
still greatly improved comparing to vanilla postcopy on random (~10ms).
The patch is verified by our QE team in a video streaming test case, to
reduce the pause of the video from ~1min to a few seconds when switching
over to postcopy with preempt mode.
Reported-by: Xiaohui Li <xiaohli@redhat.com>
Tested-by: Xiaohui Li <xiaohli@redhat.com>
Reviewed-by: Juraj Marcin <jmarcin@redhat.com>
Link: https://lore.kernel.org/r/20250424220705.195544-1-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
(cherry picked from commit 20d82622812d888478d04a2d0d8575d70eb5d749)
Signed-off-by: Peter Xu <peterx@redhat.com>
---
migration/ram.c | 97 ++++++++++++++++++++++++++++++++++++++++++++++++-
1 file changed, 96 insertions(+), 1 deletion(-)
diff --git a/migration/ram.c b/migration/ram.c
index 424df6d9f1..21d2f87ff1 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -91,6 +91,36 @@
XBZRLECacheStats xbzrle_counters;
+/*
+ * This structure locates a specific location of a guest page. In QEMU,
+ * it's described in a tuple of (ramblock, offset).
+ */
+struct PageLocation {
+ RAMBlock *block;
+ unsigned long offset;
+};
+typedef struct PageLocation PageLocation;
+
+/**
+ * PageLocationHint: describes a hint to a page location
+ *
+ * @valid set if the hint is vaild and to be consumed
+ * @location: the hint content
+ *
+ * In postcopy preempt mode, the urgent channel may provide hints to the
+ * background channel, so that QEMU source can try to migrate whatever is
+ * right after the requested urgent pages.
+ *
+ * This is based on the assumption that the VM (already running on the
+ * destination side) tends to access the memory with spatial locality.
+ * This is also the default behavior of vanilla postcopy (preempt off).
+ */
+struct PageLocationHint {
+ bool valid;
+ PageLocation location;
+};
+typedef struct PageLocationHint PageLocationHint;
+
/* used by the search for pages to send */
struct PageSearchStatus {
/* The migration channel used for a specific host page */
@@ -395,6 +425,13 @@ struct RAMState {
* RAM migration.
*/
unsigned int postcopy_bmap_sync_requested;
+ /*
+ * Page hint during postcopy when preempt mode is on. Return path
+ * thread sets it, while background migration thread consumes it.
+ *
+ * Protected by @bitmap_mutex.
+ */
+ PageLocationHint page_hint;
};
typedef struct RAMState RAMState;
@@ -2039,6 +2076,21 @@ static void pss_host_page_finish(PageSearchStatus *pss)
pss->host_page_start = pss->host_page_end = 0;
}
+static void ram_page_hint_update(RAMState *rs, PageSearchStatus *pss)
+{
+ PageLocationHint *hint = &rs->page_hint;
+
+ /* If there's a pending hint not consumed, don't bother */
+ if (hint->valid) {
+ return;
+ }
+
+ /* Provide a hint to the background stream otherwise */
+ hint->location.block = pss->block;
+ hint->location.offset = pss->page;
+ hint->valid = true;
+}
+
/*
* Send an urgent host page specified by `pss'. Need to be called with
* bitmap_mutex held.
@@ -2084,6 +2136,7 @@ out:
/* For urgent requests, flush immediately if sent */
if (sent) {
qemu_fflush(pss->pss_channel);
+ ram_page_hint_update(rs, pss);
}
return ret;
}
@@ -2171,6 +2224,30 @@ static int ram_save_host_page(RAMState *rs, PageSearchStatus *pss)
return (res < 0 ? res : pages);
}
+static bool ram_page_hint_valid(RAMState *rs)
+{
+ /* There's only page hint during postcopy preempt mode */
+ if (!postcopy_preempt_active()) {
+ return false;
+ }
+
+ return rs->page_hint.valid;
+}
+
+static void ram_page_hint_collect(RAMState *rs, RAMBlock **block,
+ unsigned long *page)
+{
+ PageLocationHint *hint = &rs->page_hint;
+
+ assert(hint->valid);
+
+ *block = hint->location.block;
+ *page = hint->location.offset;
+
+ /* Mark the hint consumed */
+ hint->valid = false;
+}
+
/**
* ram_find_and_save_block: finds a dirty page and sends it to f
*
@@ -2187,6 +2264,8 @@ static int ram_save_host_page(RAMState *rs, PageSearchStatus *pss)
static int ram_find_and_save_block(RAMState *rs)
{
PageSearchStatus *pss = &rs->pss[RAM_CHANNEL_PRECOPY];
+ unsigned long next_page;
+ RAMBlock *next_block;
int pages = 0;
/* No dirty page as there is zero RAM */
@@ -2206,7 +2285,14 @@ static int ram_find_and_save_block(RAMState *rs)
rs->last_page = 0;
}
- pss_init(pss, rs->last_seen_block, rs->last_page);
+ if (ram_page_hint_valid(rs)) {
+ ram_page_hint_collect(rs, &next_block, &next_page);
+ } else {
+ next_block = rs->last_seen_block;
+ next_page = rs->last_page;
+ }
+
+ pss_init(pss, next_block, next_page);
while (true){
if (!get_queued_page(rs, pss)) {
@@ -2339,6 +2425,13 @@ static void ram_save_cleanup(void *opaque)
ram_state_cleanup(rsp);
}
+static void ram_page_hint_reset(PageLocationHint *hint)
+{
+ hint->location.block = NULL;
+ hint->location.offset = 0;
+ hint->valid = false;
+}
+
static void ram_state_reset(RAMState *rs)
{
int i;
@@ -2351,6 +2444,8 @@ static void ram_state_reset(RAMState *rs)
rs->last_page = 0;
rs->last_version = ram_list.version;
rs->xbzrle_started = false;
+
+ ram_page_hint_reset(&rs->page_hint);
}
#define MAX_WAIT 50 /* ms, half buffered_file limit */
--
2.39.3

2
qemu-ga.sysconfig
View File

@ -13,7 +13,7 @@
#
# You can get the list of RPC commands using "qemu-ga --allow-rpcs='?'".
# There should be no spaces between commas and commands in the allow list.
FILTER_RPC_ARGS="--allow-rpcs=guest-sync-delimited,guest-sync,guest-ping,guest-get-time,guest-set-time,guest-info,guest-shutdown,guest-fsfreeze-status,guest-fsfreeze-freeze,guest-fsfreeze-freeze-list,guest-fsfreeze-thaw,guest-fstrim,guest-suspend-disk,guest-suspend-ram,guest-suspend-hybrid,guest-network-get-interfaces,guest-get-vcpus,guest-set-vcpus,guest-get-disks,guest-get-fsinfo,guest-set-user-password,guest-get-memory-blocks,guest-set-memory-blocks,guest-get-memory-block-info,guest-get-host-name,guest-get-users,guest-get-timezone,guest-get-osinfo,guest-get-devices,guest-ssh-get-authorized-keys,guest-ssh-add-authorized-keys,guest-ssh-remove-authorized-keys,guest-get-diskstats,guest-get-cpustats,guest-network-get-route"
FILTER_RPC_ARGS="--allow-rpcs=guest-sync-delimited,guest-sync,guest-ping,guest-get-time,guest-set-time,guest-info,guest-shutdown,guest-fsfreeze-status,guest-fsfreeze-freeze,guest-fsfreeze-freeze-list,guest-fsfreeze-thaw,guest-fstrim,guest-suspend-disk,guest-suspend-ram,guest-suspend-hybrid,guest-network-get-interfaces,guest-get-vcpus,guest-set-vcpus,guest-get-disks,guest-get-fsinfo,guest-set-user-password,guest-get-memory-blocks,guest-set-memory-blocks,guest-get-memory-block-info,guest-get-host-name,guest-get-users,guest-get-timezone,guest-get-osinfo,guest-get-devices,guest-ssh-get-authorized-keys,guest-ssh-add-authorized-keys,guest-ssh-remove-authorized-keys,guest-get-diskstats,guest-get-cpustats,guest-network-get-route,guest-get-load"
# Fsfreeze hook script specification.
#

27
qemu-kvm.spec
View File

@ -147,7 +147,7 @@ Obsoletes: %{name}-block-ssh <= %{epoch}:%{version} \
Summary: QEMU is a machine emulator and virtualizer
Name: qemu-kvm
Version: 10.0.0
Release: 2%{?rcrel}%{?dist}%{?cc_suffix}
Release: 3%{?rcrel}%{?dist}%{?cc_suffix}
# Epoch because we pushed a qemu-1.0 package. AIUI this can't ever be dropped
# Epoch 15 used for RHEL 8
# Epoch 17 used for RHEL 9 (due to release versioning offset in RHEL 8.5)
@ -196,6 +196,12 @@ Patch21: kvm-block-io-skip-head-tail-requests-on-EINVAL.patch
Patch22: kvm-file-posix-Fix-crash-on-discard_granularity-0.patch
# For RHEL-86056 - Enable 'vhost-user-gpu-pci' in qemu-kvm for RHIVOS
Patch23: kvm-Enable-vhost-user-gpu-pci-for-RHIVOS.patch
# For RHEL-85635 - Video stuck about 1 min after switchover phase when play one video during postcopy-preempt migration
Patch24: kvm-migration-postcopy-Spatial-locality-page-hint-for-pr.patch
# For RHEL-88457 - qemu inadvertantly built with valgrind coroutine stack debugging on x86_64
Patch25: kvm-meson-configure-add-valgrind-option-en-dis-able-valg.patch
# Fixing s390x build issues
Patch26: kvm-docs-Don-t-define-duplicate-label-in-qemu-block-driv.patch
%if %{have_clang}
BuildRequires: clang
@ -272,6 +278,9 @@ BuildRequires: libslirp-devel
BuildRequires: pulseaudio-libs-devel
BuildRequires: spice-protocol
BuildRequires: capstone-devel
%ifarch %{valgrind_arches}
BuildRequires: valgrind-devel
%endif
# Requires for qemu-kvm package
Requires: %{name}-core = %{epoch}:%{version}-%{release}
@ -653,6 +662,7 @@ ulimit -n 10240
--disable-ubsan \\\
--disable-usb-redir \\\
--disable-user \\\
--disable-valgrind \\\
--disable-vde \\\
--disable-vdi \\\
--disable-vduse-blk-export \\\
@ -777,6 +787,9 @@ run_configure \
--enable-tpm \
%if %{have_usbredir}
--enable-usb-redir \
%endif
%ifarch %{valgrind_arches}
--enable-valgrind \
%endif
--enable-vdi \
--enable-vhost-kernel \
@ -1271,6 +1284,18 @@ useradd -r -u 107 -g qemu -G kvm -d / -s /sbin/nologin \
%endif
%changelog
* Mon May 19 2025 Miroslav Rezanina <mrezanin@redhat.com> - 10.0.0-3
- kvm-migration-postcopy-Spatial-locality-page-hint-for-pr.patch [RHEL-85635]
- kvm-meson-configure-add-valgrind-option-en-dis-able-valg.patch [RHEL-88457]
- kvm-distro-add-an-explicit-valgrind-devel-build-dep.patch [RHEL-88457]
- kvm-Allow-guest-get-load-QGA-command.patch [RHEL-91219]
- Resolves: RHEL-85635
(Video stuck about 1 min after switchover phase when play one video during postcopy-preempt migration)
- Resolves: RHEL-88457
(qemu inadvertantly built with valgrind coroutine stack debugging on x86_64)
- Resolves: RHEL-91219
([qemu-guest-agent] Enable 'guest-get-load' by default [RHEL-10])
* Mon May 12 2025 Miroslav Rezanina <mrezanin@redhat.com> - 10.0.0-2
- kvm-file-posix-probe-discard-alignment-on-Linux-block-de.patch [RHEL-87642]
- kvm-block-io-skip-head-tail-requests-on-EINVAL.patch [RHEL-87642]