September 2026 - Devel - libvirt List Archives

[PATCH v2 0/4] resctrl: support whole-process allocations and monitoring
by Jedrzej Wasiukiewicz 22 Sep '26

22 Sep '26

Resctrl support in libvirt currently associates allocations and monitors with a set of vCPUs. This does not account for resource consumption by QEMU IOThreads, worker threads and emulator threads. Allow cachetune, memorytune, energytune and their monitor elements to omit the vcpus attribute and apply to the whole QEMU process. Assigning the emulator process to the resctrl group before it creates threads lets all subsequently created threads inherit the same configuration. Whole-process and per-vCPU allocations cannot be mixed within a domain. A monitor without vcpus inherits its enclosing allocation's scope. Thus, it monitors the whole process under a whole-process allocation and the allocation's vCPU set under a per-vCPU allocation. This series was previously posted as: https://lists.libvirt.org/archives/list/devel@lists.libvirt.org/thread/QFXZ… Changes since v1: - Rebased onto current master. - Split the independent resctrl domstats fixes into a separate series. - Added Christopher's Reviewed-by. - Dropped the NEWS patch. Jedrzej Wasiukiewicz (4): conf: allow omitting vcpus in cachetune/memorytune/energytune conf: implement whole-process resctrl scope qemu: assign whole-process resctrl groups at domain start qemu: omit vcpus for whole-process resctrl monitor stats docs/formatdomain.rst | 57 +++-- src/conf/domain_conf.c | 231 ++++++++++++------ src/conf/domain_conf.h | 2 + src/conf/schemas/domaincommon.rng | 48 ++-- src/qemu/qemu_driver.c | 24 +- src/qemu/qemu_process.c | 40 ++- src/util/virresctrl.c | 10 +- .../cachetune-monitor-empty-vcpus.xml | 30 +++ .../cachetune-monitor-inherit-alloc.xml | 30 +++ .../cachetune-wholeprocess-duplicate.xml | 32 +++ ...chetune-wholeprocess-monitor-duplicate.xml | 31 +++ .../cachetune-wholeprocess-monitors.xml | 31 +++ .../energytune-colliding-monitor.xml | 30 +++ .../energytune-wholeprocess.xml | 29 +++ .../memorytune-wholeprocess.xml | 29 +++ .../resctrl-wholeprocess-alloc-monitor.xml | 32 +++ .../resctrl-wholeprocess-layering.xml | 32 +++ .../resctrl-wholeprocess-monitors.xml | 33 +++ .../cachetune-monitor-inherit-alloc.xml | 30 +++ tests/genericxml2xmltest.c | 11 + 20 files changed, 662 insertions(+), 130 deletions(-) create mode 100644 tests/genericxml2xmlindata/cachetune-monitor-empty-vcpus.xml create mode 100644 tests/genericxml2xmlindata/cachetune-monitor-inherit-alloc.xml create mode 100644 tests/genericxml2xmlindata/cachetune-wholeprocess-duplicate.xml create mode 100644 tests/genericxml2xmlindata/cachetune-wholeprocess-monitor-duplicate.xml create mode 100644 tests/genericxml2xmlindata/cachetune-wholeprocess-monitors.xml create mode 100644 tests/genericxml2xmlindata/energytune-colliding-monitor.xml create mode 100644 tests/genericxml2xmlindata/energytune-wholeprocess.xml create mode 100644 tests/genericxml2xmlindata/memorytune-wholeprocess.xml create mode 100644 tests/genericxml2xmlindata/resctrl-wholeprocess-alloc-monitor.xml create mode 100644 tests/genericxml2xmlindata/resctrl-wholeprocess-layering.xml create mode 100644 tests/genericxml2xmlindata/resctrl-wholeprocess-monitors.xml create mode 100644 tests/genericxml2xmloutdata/cachetune-monitor-inherit-alloc.xml -- 2.43.0 --------------------------------------------------------------------- Intel Technology Poland sp. z o.o. ul. Slowackiego 173 | 80-298 Gdansk | Sad Rejonowy Gdansk Polnoc | VII Wydzial Gospodarczy Krajowego Rejestru Sadowego - KRS 101882 | NIP 957-07-52-316 | Kapital zakladowy 200.000 PLN. Spolka oswiadcza, ze posiada status duzego przedsiebiorcy w rozumieniu ustawy z dnia 8 marca 2013 r. o przeciwdzialaniu nadmiernym opoznieniom w transakcjach handlowych. Ta wiadomosc wraz z zalacznikami jest przeznaczona dla okreslonego adresata i moze zawierac informacje poufne. W razie przypadkowego otrzymania tej wiadomosci, prosimy o powiadomienie nadawcy oraz trwale jej usuniecie; jakiekolwiek przegladanie lub rozpowszechnianie jest zabronione. This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). If you are not the intended recipient, please contact the sender and delete all copies; any review or distribution by others is strictly prohibited.

2 9

[PATCH] qemu: support 64-bit latency histogram boundaries
by Davide Marro 21 Sep '26

21 Sep '26

From: Davide Marro <dmarro(a)redhat.com> QEMU uses uint64 values for block latency histogram boundaries, while libvirt currently stores and parses them as unsigned int values. This limits histogram boundaries to UINT32_MAX nanoseconds and prevents using larger values supported by QEMU. Use unsigned long long for latency histogram boundaries throughout the domain configuration and QEMU monitor code. Update the XML schema and parser accordingly, and serialize the values using the 64-bit JSON number helper. Extend the XML and QEMU monitor tests with a 5000000000 nanosecond boundary to verify values above UINT32_MAX are preserved. Resolves: https://gitlab.com/libvirt/libvirt/-/work_items/919 Signed-off-by: Davide Marro <dmarro(a)redhat.com> --- src/conf/domain_conf.c | 12 ++--- src/conf/domain_conf.h | 10 ++-- src/conf/schemas/domaincommon.rng | 2 +- src/qemu/qemu_monitor.c | 10 ++-- src/qemu/qemu_monitor.h | 10 ++-- src/qemu/qemu_monitor_json.c | 14 +++--- src/qemu/qemu_monitor_json.h | 10 ++-- tests/qemumonitorjsontest.c | 46 ++++++++++++++++--- ...isk-statistics-intervals.x86_64-latest.xml | 1 + .../disk-statistics-intervals.xml | 1 + 10 files changed, 75 insertions(+), 41 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 1327793b0a..94a1f8bd8e 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -8427,7 +8427,7 @@ virDomainDiskDefDriverParseXMLHistogramOne(virDomainDiskDef *def, xmlNodePtr cur) { g_autofree char *histogram_type = NULL; - unsigned int **histogram_config = NULL; + unsigned long long **histogram_config = NULL; g_autoptr(GPtrArray) binNodes = virXMLNodeGetSubelementList(cur, "bin"); size_t nbins = 0; size_t i; @@ -8463,12 +8463,12 @@ virDomainDiskDefDriverParseXMLHistogramOne(virDomainDiskDef *def, return -1; } - *histogram_config = g_new0(unsigned int, binNodes->len + 1); + *histogram_config = g_new0(unsigned long long, binNodes->len + 1); for (i = 0; i < binNodes->len; i++) { - unsigned int val; + unsigned long long val; - if (virXMLPropUInt(g_ptr_array_index(binNodes, i), + if (virXMLPropULongLong(g_ptr_array_index(binNodes, i), "start", 10, VIR_XML_PROP_REQUIRED, &val) < 0) @@ -24571,7 +24571,7 @@ virDomainDiskDefFormatThrottleFilters(virBuffer *buf, static void virDomainDiskDefFormatDriverHistogram(virBuffer *buf, const char *type, - unsigned int *bins) + unsigned long long *bins) { g_auto(virBuffer) histogramAttrBuf = VIR_BUFFER_INITIALIZER; g_auto(virBuffer) histogramChildBuf = VIR_BUFFER_INIT_CHILD(buf); @@ -24586,7 +24586,7 @@ virDomainDiskDefFormatDriverHistogram(virBuffer *buf, virBufferAddLit(&histogramChildBuf, "<bin start='0'/>\n"); for (; *bins > 0; bins++) - virBufferAsprintf(&histogramChildBuf, "<bin start='%u'/>\n", *bins); + virBufferAsprintf(&histogramChildBuf, "<bin start='%llu'/>\n", *bins); virXMLFormatElement(buf, "latency-histogram", &histogramAttrBuf, &histogramChildBuf); } diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index f7cbe7bf1d..eadb5a49a4 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -599,11 +599,11 @@ struct _virDomainDiskDef { unsigned int *statistics; /* Optional, zero terminated list of intervals to collect statistics for */ /* optional zero terminated lists of bin boundaries for latency histograms */ - unsigned int *histogram_boundaries; - unsigned int *histogram_boundaries_read; - unsigned int *histogram_boundaries_write; - unsigned int *histogram_boundaries_zone; - unsigned int *histogram_boundaries_flush; + unsigned long long *histogram_boundaries; + unsigned long long *histogram_boundaries_read; + unsigned long long *histogram_boundaries_write; + unsigned long long *histogram_boundaries_zone; + unsigned long long *histogram_boundaries_flush; virDomainDiskDetectZeroes detect_zeroes; virTristateSwitch discard_no_unref; diff --git a/src/conf/schemas/domaincommon.rng b/src/conf/schemas/domaincommon.rng index 0c0a3597a9..d1b10c1c5f 100644 --- a/src/conf/schemas/domaincommon.rng +++ b/src/conf/schemas/domaincommon.rng @@ -2845,7 +2845,7 @@ <oneOrMore> <element name='bin'> <attribute name='start'> - <ref name="unsignedInt"/> + <ref name="unsignedLong"/> </attribute> </element> </oneOrMore> diff --git a/src/qemu/qemu_monitor.c b/src/qemu/qemu_monitor.c index 21df06ebd8..2c86cd1971 100644 --- a/src/qemu/qemu_monitor.c +++ b/src/qemu/qemu_monitor.c @@ -4558,11 +4558,11 @@ qemuMonitorBlockdevSetActive(qemuMonitor *mon, int qemuMonitorBlockLatencyHistogramSet(qemuMonitor *mon, const char *id, - unsigned int *boundaries, - unsigned int *boundaries_read, - unsigned int *boundaries_write, - unsigned int *boundaries_zone, - unsigned int *boundaries_flush) + unsigned long long *boundaries, + unsigned long long *boundaries_read, + unsigned long long *boundaries_write, + unsigned long long *boundaries_zone, + unsigned long long *boundaries_flush) { QEMU_CHECK_MONITOR(mon); VIR_DEBUG("id='%s'", id); diff --git a/src/qemu/qemu_monitor.h b/src/qemu/qemu_monitor.h index e560618bf4..e9b6a25f0c 100644 --- a/src/qemu/qemu_monitor.h +++ b/src/qemu/qemu_monitor.h @@ -1988,11 +1988,11 @@ qemuMonitorBlockdevSetActive(qemuMonitor *mon, int qemuMonitorBlockLatencyHistogramSet(qemuMonitor *mon, const char *id, - unsigned int *boundaries, - unsigned int *boundaries_read, - unsigned int *boundaries_write, - unsigned int *boundaries_zone, - unsigned int *boundaries_flush); + unsigned long long *boundaries, + unsigned long long *boundaries_read, + unsigned long long *boundaries_write, + unsigned long long *boundaries_zone, + unsigned long long *boundaries_flush); int qemuMonitorAnnounceSelf(qemuMonitor *mon, diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index 7e3249cd47..8e88dc9442 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -9176,7 +9176,7 @@ qemuMonitorJSONBlockdevSetActive(qemuMonitor *mon, static virJSONValue * -qemuMonitorJSONBlockLatencyHistogramBoundary(unsigned int *bound) +qemuMonitorJSONBlockLatencyHistogramBoundary(unsigned long long *bound) { g_autoptr(virJSONValue) ret = virJSONValueNewArray(); @@ -9184,7 +9184,7 @@ qemuMonitorJSONBlockLatencyHistogramBoundary(unsigned int *bound) return NULL; for (; *bound > 0; bound++) { - g_autoptr(virJSONValue) val = virJSONValueNewNumberUint(*bound); + g_autoptr(virJSONValue) val = virJSONValueNewNumberUlong(*bound); /* the only error is if the first argument is not an array */ ignore_value(virJSONValueArrayAppend(ret, &val)); @@ -9197,11 +9197,11 @@ qemuMonitorJSONBlockLatencyHistogramBoundary(unsigned int *bound) int qemuMonitorJSONBlockLatencyHistogramSet(qemuMonitor *mon, const char *id, - unsigned int *boundaries, - unsigned int *boundaries_read, - unsigned int *boundaries_write, - unsigned int *boundaries_zone, - unsigned int *boundaries_flush) + unsigned long long *boundaries, + unsigned long long *boundaries_read, + unsigned long long *boundaries_write, + unsigned long long *boundaries_zone, + unsigned long long *boundaries_flush) { g_autoptr(virJSONValue) cmd = NULL; g_autoptr(virJSONValue) reply = NULL; diff --git a/src/qemu/qemu_monitor_json.h b/src/qemu/qemu_monitor_json.h index 3b7b04a11d..63eded7ecb 100644 --- a/src/qemu/qemu_monitor_json.h +++ b/src/qemu/qemu_monitor_json.h @@ -818,11 +818,11 @@ qemuMonitorJSONBlockdevSetActive(qemuMonitor *mon, int qemuMonitorJSONBlockLatencyHistogramSet(qemuMonitor *mon, const char *id, - unsigned int *boundaries, - unsigned int *boundaries_read, - unsigned int *boundaries_write, - unsigned int *boundaries_zone, - unsigned int *boundaries_flush); + unsigned long long *boundaries, + unsigned long long *boundaries_read, + unsigned long long *boundaries_write, + unsigned long long *boundaries_zone, + unsigned long long *boundaries_flush); int qemuMonitorJSONAnnounceSelf(qemuMonitor *mon, diff --git a/tests/qemumonitorjsontest.c b/tests/qemumonitorjsontest.c index 70f03afb81..a963029c69 100644 --- a/tests/qemumonitorjsontest.c +++ b/tests/qemumonitorjsontest.c @@ -1130,13 +1130,45 @@ GEN_TEST_FUNC(qemuMonitorJSONSetAction, GEN_TEST_FUNC(qemuMonitorJSONSetLaunchSecurityState, "sev_secret_header", "sev_secret", 0, true) -unsigned int testHistogramBoundaries[] = {10, 30, 50, 0}; -GEN_TEST_FUNC(qemuMonitorJSONBlockLatencyHistogramSet, "devid", - testHistogramBoundaries, - testHistogramBoundaries, - testHistogramBoundaries, - testHistogramBoundaries, - testHistogramBoundaries) +static int +testQemuMonitorJSONqemuMonitorJSONBlockLatencyHistogramSet(const void *opaque) +{ + const testQemuMonitorJSONSimpleFuncData *data = opaque; + virDomainXMLOption *xmlopt = data->xmlopt; + unsigned long long boundaries[] = {10, 30, 5000000000ULL, 0}; + g_autoptr(qemuMonitorTest) test = NULL; + + if (!(test = qemuMonitorTestNewSchema(xmlopt, data->schema))) + return -1; + + if (qemuMonitorTestAddItemVerbatim(test, + "{" + "\"execute\":\"block-latency-histogram-set\"," + "\"arguments\":{" + "\"id\":\"devid\"," + "\"boundaries\":[10,30,5000000000]," + "\"boundaries-read\":[10,30,5000000000]," + "\"boundaries-write\":[10,30,5000000000]," + "\"boundaries-zap\":[10,30,5000000000]," + "\"boundaries-flush\":[10,30,5000000000]" + "}," + "\"id\":\"libvirt-1\"" + "}", + NULL, + "{\"return\":{}}") < 0) + return -1; + + if (qemuMonitorJSONBlockLatencyHistogramSet(qemuMonitorTestGetMonitor(test), + "devid", + boundaries, + boundaries, + boundaries, + boundaries, + boundaries) < 0) + return -1; + + return 0; +} static int testQemuMonitorJSONqemuMonitorJSONNBDServerStart(const void *opaque) diff --git a/tests/qemuxmlconfdata/disk-statistics-intervals.x86_64-latest.xml b/tests/qemuxmlconfdata/disk-statistics-intervals.x86_64-latest.xml index d02f954073..d245eedb5b 100644 --- a/tests/qemuxmlconfdata/disk-statistics-intervals.x86_64-latest.xml +++ b/tests/qemuxmlconfdata/disk-statistics-intervals.x86_64-latest.xml @@ -26,6 +26,7 @@ <bin start='0'/> <bin start='20'/> <bin start='30'/> + <bin start='5000000000'/> </latency-histogram> </statistics> </driver> diff --git a/tests/qemuxmlconfdata/disk-statistics-intervals.xml b/tests/qemuxmlconfdata/disk-statistics-intervals.xml index 5f9e9470d7..8b1e7481b2 100644 --- a/tests/qemuxmlconfdata/disk-statistics-intervals.xml +++ b/tests/qemuxmlconfdata/disk-statistics-intervals.xml @@ -23,6 +23,7 @@ <bin start='0'/> <bin start='20'/> <bin start='30'/> + <bin start='5000000000'/> </latency-histogram> </statistics> </driver> -- 2.54.0 (Apple Git-157)

1 0

[PATCH] virbitmap: Avoid UB with zero sized bitmaps
by Michal Privoznik 21 Sep '26

21 Sep '26

From: Michal Privoznik <mprivozn(a)redhat.com> Ever since of v6.9.0-rc1~384 we allow zero sized bitmaps. If that's the case then the map is kept unallocated (NULL). Unfortunately, some virBitmap APIs call memcpy()/memset() which doesn't allow NULL pointers. Introduce checks to these APIs (virBitmapNewCopy(), virBitmapSetAll() and virBitmapClearAll()). While at it, also introduce a test case to virbitmaptest to catch UB in our CI. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/util/virbitmap.c | 9 ++++++++- tests/virbitmaptest.c | 27 +++++++++++++++++++++++++++ 2 files changed, 35 insertions(+), 1 deletion(-) diff --git a/src/util/virbitmap.c b/src/util/virbitmap.c index 78545501b7..b9b737a087 100644 --- a/src/util/virbitmap.c +++ b/src/util/virbitmap.c @@ -587,7 +587,8 @@ virBitmapNewCopy(virBitmap *src) { virBitmap *dst = virBitmapNew(src->nbits); - memcpy(dst->map, src->map, dst->map_len * sizeof(src->map[0])); + if (G_LIKELY(dst->map_len > 0)) + memcpy(dst->map, src->map, dst->map_len * sizeof(src->map[0])); return dst; } @@ -770,6 +771,9 @@ virBitmapClearTail(virBitmap *bitmap) */ void virBitmapSetAll(virBitmap *bitmap) { + if (G_UNLIKELY(bitmap->map_len == 0)) + return; + memset(bitmap->map, 0xff, bitmap->map_len * (VIR_BITMAP_BITS_PER_UNIT / CHAR_BIT)); @@ -786,6 +790,9 @@ void virBitmapSetAll(virBitmap *bitmap) void virBitmapClearAll(virBitmap *bitmap) { + if (G_UNLIKELY(bitmap->map_len == 0)) + return; + memset(bitmap->map, 0, bitmap->map_len * (VIR_BITMAP_BITS_PER_UNIT / CHAR_BIT)); } diff --git a/tests/virbitmaptest.c b/tests/virbitmaptest.c index 709c62ab54..549956ad04 100644 --- a/tests/virbitmaptest.c +++ b/tests/virbitmaptest.c @@ -768,6 +768,30 @@ test17(const void *opaque G_GNUC_UNUSED) } +static int +test18(const void *opaque G_GNUC_UNUSED) +{ + g_autoptr(virBitmap) map = virBitmapNew(0); + g_autoptr(virBitmap) copy = virBitmapNewCopy(map); + + /* Creating an empty bitmap means underlying array of bits is + * unallocated (NULL). For some virBitmap APIs this may lead + * to UB. The whole point of this test is to call those and + * rely on clang's UB sanitizer to catch such errors at + * runtime. */ + + virBitmapSetAll(map); + virBitmapClearAll(copy); + + if (!virBitmapEqual(map, copy)) { + fprintf(stderr, "empty maps don't equal\n"); + return -1; + } + + return 0; +} + + #define TESTBINARYOP(A, B, RES, FUNC) \ testBinaryOpData.a = A; \ testBinaryOpData.b = B; \ @@ -847,6 +871,9 @@ mymain(void) if (virTestRun("test17", test17, NULL) < 0) ret = -1; + if (virTestRun("test18", test18, NULL) < 0) + ret = -1; + return ret == 0 ? EXIT_SUCCESS : EXIT_FAILURE; } -- 2.55.0

2 1

[PATCH] virBitmapNewCopy: skip copying zero-sized bitmaps
by Roman Bogorodskiy 21 Sep '26

21 Sep '26

The Ubuntu clang CI build currently fails on genericxml2xmltest with: ../src/util/virbitmap.c:590:12: runtime error: null pointer passed as argument 1, which is declared to never be null /usr/include/string.h:48:28: note: nonnull attribute specified here Which is happening on memcpy() in virBitmapNewCopy(). This happens when copying a zero-sized bitmap. Fix by calling memcpy() only when the source and destination bitmaps are not zero-sized. Signed-off-by: Roman Bogorodskiy <bogorodskiy(a)gmail.com> --- src/util/virbitmap.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/src/util/virbitmap.c b/src/util/virbitmap.c index 78545501b7..b75ba61a99 100644 --- a/src/util/virbitmap.c +++ b/src/util/virbitmap.c @@ -587,7 +587,8 @@ virBitmapNewCopy(virBitmap *src) { virBitmap *dst = virBitmapNew(src->nbits); - memcpy(dst->map, src->map, dst->map_len * sizeof(src->map[0])); + if (dst->map && src->map) + memcpy(dst->map, src->map, dst->map_len * sizeof(src->map[0])); return dst; } -- 2.55.0

2 1

[PATCH v2 0/2] qemu: Support backup and fsync optional attributes in SWTPM
by Arun Menon 19 Sep '26

19 Sep '26

In the upcoming release of swtpm, 2 new options, backup and fsync will be added to the directory backend [1] and [2]. This allows the user to backup and fsync the state file and its parent directory. It lowers the probability of TPM state file corruption in case of a power loss. Although this has not been added to the stable swtpm release yet, Libvirt will require some changes to incorporate these options in the XML config. Following is the sample xml configuration snip to enable backup and fsync: <tpm model='tpm-crb'> <backend type='emulator' version='2.0' backup='yes' fsync='yes'/> </tpm> Tested with swtpm commit #8ecf2c1 and libtpms commit #18e42c8d: $ ps aux | grep swtpm arun 33881 0.0 0.0 24080 10188 ? S 00:09 0:00 /usr/local/bin/swtpm socket --ctrl type=unixio,path=/run/user/1000/libvirt/qemu/run/swtpm/5-guest-vm-swtpm.sock,mode=0600 --tpmstate dir=/home/arun/.config/libvirt/qemu/swtpm/0a64239e-6311-4e53-8903-d1d49608a751/tpm2,mode=0600,lock,backup,fsync --log file=/home/arun/.cache/libvirt/qemu/log/guest-vm-swtpm.log --terminate --tpm2 $ ls -lrt $HOME/.config/libvirt/qemu/swtpm/0a64239e-6311-4e53-8903-d1d49608a751/tpm2 total 16 -rw-------. 1 arun arun 6807 Sep 7 00:14 tpm2-00.permall.bak -rw-------. 1 arun arun 6807 Sep 7 01:02 tpm2-00.permall [1] https://github.com/stefanberger/swtpm/commit/186f72a3a995fc68957ffc0b1438bb… [2] https://github.com/stefanberger/swtpm/commit/795c1973c913da59925fa6c2d688c5… Changes since v1: - Use TriStateBool type for both backup and sync in virDomainTPMDefParseXML() - Refactor qemuTPMVirCommandSwtpmAddTPMState() to use virBuffer API and return errors on failure. Arun Menon (2): conf: Add backup and fsync optional swtpm attributes qemu: tpm: Add backup and fsync options to the swtpm command docs/formatdomain.rst | 15 +++++ src/conf/domain_conf.c | 22 ++++++++ src/conf/domain_conf.h | 2 + src/conf/schemas/domaincommon.rng | 10 ++++ src/qemu/qemu_tpm.c | 55 ++++++++++++++----- src/util/virtpm.c | 2 + src/util/virtpm.h | 2 + .../qemuxmlconfdata/tpm-emulator-tpm2-enc.xml | 2 +- 8 files changed, 96 insertions(+), 14 deletions(-) -- 2.54.0

1 2

Plans for 12.8.0 release (freeze on 2026-09-24)
by Jiri Denemark 18 Sep '26

18 Sep '26

We are getting close to 12.8.0 release of libvirt. To aim for the release on Thursday 01 Oct I suggest entering the freeze on Thursday 24 Sep and tagging RC2 on Tuesday 29 Sep. I hope this works for everyone. Jirka

1 0

[PATCH] vsh: Remove 'disconnected' set-only global variable
by Peter Krempa 18 Sep '26

18 Sep '26

From: Peter Krempa <pkrempa(a)redhat.com> The instance of the 'disconnected' global variable in vsh.c is written-only, so can be removed. Caught by new clang: clang -Itools/libvirt_shell.a.p -Itools -I../tools -Iinclude -I../include -Isrc -I../src -Isrc/util -I../src/util -I. -I.. -I/usr/include/libxml2 -I/usr/include/glib-2.0 -I/usr/lib64/glib-2.0/include -I/usr/include/sysprof-6 -I/usr/include/gio-unix-2.0 -I/usr/include/libmount -I/usr/include/blkid -fdiagnostics-color=always -D_FILE_OFFSET_BITS=64 -Wall -Winvalid-pch -Wextra -std=gnu99 -O2 -g @/builds/libvirt/libvirt/build/c-warnings.txt -fPIC -D_DEFAULT_SOURCE -D_XOPEN_SOURCE=600 -D_FUNCTION_DEF -pthread -DWITH_GZFILEOP -MD -MQ tools/libvirt_shell.a.p/vsh.c.o -MF tools/libvirt_shell.a.p/vsh.c.o.d -o tools/libvirt_shell.a.p/vsh.c.o -c ../tools/vsh.c ../tools/vsh.c:242:12: error: variable 'disconnected' set but not used [-Werror,-Wunused-but-set-global] 242 | static int disconnected; /* we may have been disconnected */ | ^ 1 error generated. This instance exists since the 'vsh' module was split (834c5720e44). 'virsh' contains it's own instance. Signed-off-by: Peter Krempa <pkrempa(a)redhat.com> --- tools/vsh.c | 16 ---------------- 1 file changed, 16 deletions(-) diff --git a/tools/vsh.c b/tools/vsh.c index 667cc1a0d2..d1540810ee 100644 --- a/tools/vsh.c +++ b/tools/vsh.c @@ -236,12 +236,6 @@ vshReportError(vshControl *ctl) vshResetLibvirtError(); } -/* - * Detection of disconnections and automatic reconnection support - */ -static int disconnected; /* we may have been disconnected */ - - /* vshCmddefSearch: * @cmdname: name of command to find * @@ -1392,16 +1386,6 @@ vshCommandRun(vshControl *ctl, after = g_get_real_time(); - /* try to automatically catch disconnections */ - if (ret != EXIT_SUCCESS && - ((last_error != NULL) && - (((last_error->code == VIR_ERR_SYSTEM_ERROR) && - (last_error->domain == VIR_FROM_REMOTE)) || - (last_error->code == VIR_ERR_RPC) || - (last_error->code == VIR_ERR_NO_CONNECT) || - (last_error->code == VIR_ERR_INVALID_CONN)))) - disconnected++; - if (ret != EXIT_SUCCESS) vshReportError(ctl); -- 2.55.0

2 1

[PATCH] qemu: allow large VMs without IOMMU for x86
by Anton Gerasimov 17 Sep '26

17 Sep '26

Extended Destination ID is an alternative to interrupt remapping in IOMMU enabled in QEMU by default when kernel-irqchip=split is used. Signed-off-by: Anton Gerasimov <agerasimov(a)nebius.com> --- src/qemu/qemu_validate.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_validate.c b/src/qemu/qemu_validate.c index e946f55351..c8af25828b 100644 --- a/src/qemu/qemu_validate.c +++ b/src/qemu/qemu_validate.c @@ -954,10 +954,12 @@ qemuValidateDomainVCpuTopology(const virDomainDef *def, virQEMUCaps *qemuCaps) QEMU_MAX_VCPUS_WITHOUT_X2APIC); return -1; } - if (!def->iommus || (def->iommus[0]->eim != VIR_TRISTATE_SWITCH_ON && - def->iommus[0]->xtsup != VIR_TRISTATE_SWITCH_ON)) { + if ((def->features[VIR_DOMAIN_FEATURE_IOAPIC] != VIR_DOMAIN_IOAPIC_QEMU) && + (!def->iommus || + (def->iommus[0]->eim != VIR_TRISTATE_SWITCH_ON && + def->iommus[0]->xtsup != VIR_TRISTATE_SWITCH_ON))) { virReportError(VIR_ERR_CONFIG_UNSUPPORTED, - _("more than %1$d vCPUs require EIM or XTSup mode enabled on the iommu device"), + _("more than %1$d vCPUs require either IOAPIC in QEMU or an iommu device with EIM or XTSup mode enabled"), QEMU_MAX_VCPUS_WITHOUT_X2APIC); return -1; } -- 2.43.0

1 0

[PATCH] bhyve: remove transient domains on shutdown
by Roman Bogorodskiy 17 Sep '26

17 Sep '26

Currently, transient domains are removed when destroyed. But the shutdown event ends up handled by the monitor which does not remove the domain. Update bhyveMonitorIO() to remove the domain if it's transient and not active. Closes: https://gitlab.com/libvirt/libvirt/-/work_items/922 Signed-off-by: Roman Bogorodskiy <bogorodskiy(a)gmail.com> --- Related TCK change: https://gitlab.com/libvirt/libvirt-tck/-/merge_requests/94 src/bhyve/bhyve_monitor.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/src/bhyve/bhyve_monitor.c b/src/bhyve/bhyve_monitor.c index 379e588070..dab07ff95f 100644 --- a/src/bhyve/bhyve_monitor.c +++ b/src/bhyve/bhyve_monitor.c @@ -186,6 +186,9 @@ bhyveMonitorIO(int watch, int kq, int events G_GNUC_UNUSED, void *opaque) } } + if (!virDomainObjIsActive(vm) && !vm->persistent) + virDomainObjListRemove(driver->domains, vm); + cleanup: virObjectUnlock(vm); } -- 2.55.0

2 1

[PATCH v1 1/1] qemu: add support for zero-page-detection migration parameter
by Tejus GK 17 Sep '26

17 Sep '26

QEMU supports a migration parameter to determine whether zero-page checking should be performed and if yes, where it should be performed (i.e., in the main migration thread or multifd threads). Add support for this parameter in libvirt. Signed-off-by: Tejus GK <tejus.gk(a)nutanix.com> --- include/libvirt/libvirt-domain.h | 11 +++++++++++ src/qemu/qemu_migration.c | 1 + src/qemu/qemu_migration_params.c | 8 ++++++++ src/qemu/qemu_migration_params.h | 1 + tools/virsh-domain.c | 13 +++++++++++++ 5 files changed, 34 insertions(+) diff --git a/include/libvirt/libvirt-domain.h b/include/libvirt/libvirt-domain.h index 99208137c3..5a4875770b 100644 --- a/include/libvirt/libvirt-domain.h +++ b/include/libvirt/libvirt-domain.h @@ -1468,6 +1468,17 @@ typedef enum { */ # define VIR_MIGRATE_PARAM_TLS_DESTINATION "tls.destination" +/** + * VIR_MIGRATE_PARAM_ZERO_PAGE_DETECTION: + * + * virDomainMigrate* params field: zero page detection method as + * VIR_TYPED_PARAM_STRING. Supported values are 'none', 'legacy', and + * 'multifd'. + * + * Since: 12.8.0 + */ +# define VIR_MIGRATE_PARAM_ZERO_PAGE_DETECTION "zero_page_detection" + /* Domain migration. */ virDomainPtr virDomainMigrate (virDomainPtr domain, virConnectPtr dconn, unsigned long flags, const char *dname, diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c index 4a43ab83b0..f25506e590 100644 --- a/src/qemu/qemu_migration.c +++ b/src/qemu/qemu_migration.c @@ -110,6 +110,7 @@ const virTypedParamValidationTemplate qemuMigrationParametersValidation[] = { { VIR_MIGRATE_PARAM_TLS_DESTINATION, VIR_TYPED_PARAM_STRING }, { VIR_MIGRATE_PARAM_DISKS_URI, VIR_TYPED_PARAM_STRING }, { VIR_MIGRATE_PARAM_BANDWIDTH_AVAIL_SWITCHOVER, VIR_TYPED_PARAM_ULLONG }, + { VIR_MIGRATE_PARAM_ZERO_PAGE_DETECTION, VIR_TYPED_PARAM_STRING}, { "", 0 } }; diff --git a/src/qemu/qemu_migration_params.c b/src/qemu/qemu_migration_params.c index dd47516742..7064140379 100644 --- a/src/qemu/qemu_migration_params.c +++ b/src/qemu/qemu_migration_params.c @@ -131,6 +131,7 @@ VIR_ENUM_IMPL(qemuMigrationParam, "multifd-zstd-level", "avail-switchover-bandwidth", "direct-io", + "zero-page-detection", ); typedef struct _qemuMigrationParamsAlwaysOnItem qemuMigrationParamsAlwaysOnItem; @@ -274,6 +275,10 @@ static const qemuMigrationParamsTPMapItem qemuMigrationParamsTPMap[] = { .unit = 1024 * 1024, /* MiB/s */ .param = QEMU_MIGRATION_PARAM_AVAIL_SWITCHOVER_BANDWIDTH, .party = QEMU_MIGRATION_SOURCE}, + + {.typedParam = VIR_MIGRATE_PARAM_ZERO_PAGE_DETECTION, + .param = QEMU_MIGRATION_PARAM_ZERO_PAGE_DETECTION, + .party = QEMU_MIGRATION_SOURCE}, }; static const qemuMigrationParamInfoItem qemuMigrationParamInfo[] = { @@ -332,6 +337,9 @@ static const qemuMigrationParamInfoItem qemuMigrationParamInfo[] = { [QEMU_MIGRATION_PARAM_DIRECT_IO] = { .type = QEMU_MIGRATION_PARAM_TYPE_BOOL, }, + [QEMU_MIGRATION_PARAM_ZERO_PAGE_DETECTION] = { + .type = QEMU_MIGRATION_PARAM_TYPE_STRING, + }, }; G_STATIC_ASSERT(G_N_ELEMENTS(qemuMigrationParamInfo) == QEMU_MIGRATION_PARAM_LAST); diff --git a/src/qemu/qemu_migration_params.h b/src/qemu/qemu_migration_params.h index b7a829b85a..131fe82fc0 100644 --- a/src/qemu/qemu_migration_params.h +++ b/src/qemu/qemu_migration_params.h @@ -67,6 +67,7 @@ typedef enum { QEMU_MIGRATION_PARAM_MULTIFD_ZSTD_LEVEL, QEMU_MIGRATION_PARAM_AVAIL_SWITCHOVER_BANDWIDTH, QEMU_MIGRATION_PARAM_DIRECT_IO, + QEMU_MIGRATION_PARAM_ZERO_PAGE_DETECTION, QEMU_MIGRATION_PARAM_LAST } qemuMigrationParam; diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c index 2c64f19a19..519d8427bd 100644 --- a/tools/virsh-domain.c +++ b/tools/virsh-domain.c @@ -11432,6 +11432,10 @@ static const vshCmdOptDef opts_migrate[] = { .type = VSH_OT_INT, .help = N_("bandwidth (in MiB/s) available for the final phase of migration") }, + {.name = "zero-page-detection", + .type = VSH_OT_STRING, + .help = N_("zero page detection method for migration") + }, {.name = NULL} }; @@ -11733,6 +11737,15 @@ doMigrate(void *opaque) goto save_error; } + if ((rv = vshCommandOptString(ctl, cmd, "zero-page-detection", &opt)) < 0) { + goto out; + } else if (rv > 0) { + if (virTypedParamsAddString(&params, &nparams, &maxparams, + VIR_MIGRATE_PARAM_ZERO_PAGE_DETECTION, + opt) < 0) + goto save_error; + } + if ((rv = vshCommandOptInt(ctl, cmd, "comp-zlib-level", &intOpt)) < 0) { goto out; } else if (rv > 0) { -- 2.43.7

1 0