- Devel - libvirt List Archives

[PATCH 00/13] Big CI update (and some bug fixes)
by Michal Privoznik 07 May '24

07 May '24

There's a lot happening here, but I did not find a way to split this any better. It all started with me wanting to switch from Ubuntu 20.04 to 24.04 and resulted in this. Firstly, Ubuntu 24.04 started to complain about a few things: 1) ASAN wasn't working (some missing .a archive). Fix is pending: https://gitlab.com/libvirt/libvirt-ci/-/merge_requests/482 2) ASAN was complaining about some memleaks in qemuxmlconftest. When I tried to run the test locally under valgrind I found out I can't. This resulted in the first two patches. Mind you, after this, there are still some tests which are unable to run under valgrind (do NOT forget --trace-children=yes if the test binary uses mocks). Well, they may get fixes later. Firstly, I need to understand what really bother valgrind. 3) UBSAN complained about some undefined behavior. This was the hardest to make sense out of, for me. But part of the fix is bumping glib so yay! I guess. And, since I'm dropping AlmaLinux 8, we get to generate website on something newer. Finally, green pipeline with these patches applied: https://gitlab.com/MichalPrivoznik/libvirt/-/pipelines/1279666864 Michal Prívozník (13): domaincapsmock: Drop link time dependency on virQEMUCapsGet() qemuxml2argvmock: Drop link time dependency on qemuFDPassDirectNew() testutilsqemu: Don't leak struct testQemuArgs::vdpafds security: Fix return types of .probe callbacks meson: Disable -fsanitize=function gitlab-ci: Move website_job to Fedora 39 ci: Drop Fedora 38 ci: Drop Ubuntu 20.04 ci: Drop AlmaLinux 8 meson: Bump glib version to 2.58.0 ci: Introduce Fedora 40 ci: Introduce Ubuntu 24.04 ci: Introduce AlmaLinux 9 .gitlab-ci.yml | 6 +- .../{almalinux-8.sh => almalinux-9.sh} | 8 +- ci/buildenv/centos-stream-9.sh | 1 + ci/buildenv/debian-12-cross-aarch64.sh | 1 + ci/buildenv/debian-12-cross-armv6l.sh | 1 + ci/buildenv/debian-12-cross-armv7l.sh | 1 + ci/buildenv/debian-12-cross-i686.sh | 1 + ci/buildenv/debian-12-cross-mips64el.sh | 1 + ci/buildenv/debian-12-cross-mipsel.sh | 1 + ci/buildenv/debian-12-cross-ppc64le.sh | 1 + ci/buildenv/debian-12-cross-s390x.sh | 1 + ci/buildenv/debian-12.sh | 1 + ci/buildenv/debian-sid-cross-aarch64.sh | 1 + ci/buildenv/debian-sid-cross-armv6l.sh | 1 + ci/buildenv/debian-sid-cross-armv7l.sh | 1 + ci/buildenv/debian-sid-cross-i686.sh | 1 + ci/buildenv/debian-sid-cross-mips64el.sh | 1 + ci/buildenv/debian-sid-cross-ppc64le.sh | 1 + ci/buildenv/debian-sid-cross-s390x.sh | 1 + ci/buildenv/debian-sid.sh | 1 + ci/buildenv/fedora-39.sh | 1 + ...-mingw32.sh => fedora-40-cross-mingw32.sh} | 1 + ...-mingw64.sh => fedora-40-cross-mingw64.sh} | 1 + ci/buildenv/{fedora-38.sh => fedora-40.sh} | 1 + ci/buildenv/fedora-rawhide-cross-mingw32.sh | 1 + ci/buildenv/fedora-rawhide-cross-mingw64.sh | 1 + ci/buildenv/fedora-rawhide.sh | 1 + .../{ubuntu-2004.sh => ubuntu-2404.sh} | 8 +- ...ux-8.Dockerfile => almalinux-9.Dockerfile} | 10 +- ci/containers/centos-stream-9.Dockerfile | 1 + .../debian-12-cross-aarch64.Dockerfile | 1 + .../debian-12-cross-armv6l.Dockerfile | 1 + .../debian-12-cross-armv7l.Dockerfile | 1 + ci/containers/debian-12-cross-i686.Dockerfile | 1 + .../debian-12-cross-mips64el.Dockerfile | 1 + .../debian-12-cross-mipsel.Dockerfile | 1 + .../debian-12-cross-ppc64le.Dockerfile | 1 + .../debian-12-cross-s390x.Dockerfile | 1 + ci/containers/debian-12.Dockerfile | 1 + .../debian-sid-cross-aarch64.Dockerfile | 1 + .../debian-sid-cross-armv6l.Dockerfile | 1 + .../debian-sid-cross-armv7l.Dockerfile | 1 + .../debian-sid-cross-i686.Dockerfile | 1 + .../debian-sid-cross-mips64el.Dockerfile | 1 + .../debian-sid-cross-ppc64le.Dockerfile | 1 + .../debian-sid-cross-s390x.Dockerfile | 1 + ci/containers/debian-sid.Dockerfile | 1 + ci/containers/fedora-39.Dockerfile | 1 + ...ile => fedora-40-cross-mingw32.Dockerfile} | 3 +- ...ile => fedora-40-cross-mingw64.Dockerfile} | 3 +- ...ora-38.Dockerfile => fedora-40.Dockerfile} | 3 +- .../fedora-rawhide-cross-mingw32.Dockerfile | 1 + .../fedora-rawhide-cross-mingw64.Dockerfile | 1 + ci/containers/fedora-rawhide.Dockerfile | 1 + ...2004.Dockerfile => ubuntu-2404.Dockerfile} | 11 +- ci/gitlab/builds.yml | 113 ++++++++-------- ci/gitlab/containers.yml | 54 ++++---- ci/integration.yml | 24 ---- ci/lcitool/projects/libvirt.yml | 1 + ci/manifest.yml | 40 +++--- meson.build | 15 ++- src/libvirt_private.syms | 1 - src/security/security_apparmor.c | 2 +- src/security/security_selinux.c | 2 +- src/util/glibcompat.c | 125 ------------------ src/util/glibcompat.h | 10 -- tests/domaincapsmock.c | 13 +- tests/qemuxml2argvmock.c | 8 +- tests/testutilsqemu.c | 1 + 69 files changed, 203 insertions(+), 305 deletions(-) rename ci/buildenv/{almalinux-8.sh => almalinux-9.sh} (93%) rename ci/buildenv/{fedora-38-cross-mingw32.sh => fedora-40-cross-mingw32.sh} (98%) rename ci/buildenv/{fedora-38-cross-mingw64.sh => fedora-40-cross-mingw64.sh} (98%) rename ci/buildenv/{fedora-38.sh => fedora-40.sh} (99%) rename ci/buildenv/{ubuntu-2004.sh => ubuntu-2404.sh} (94%) rename ci/containers/{almalinux-8.Dockerfile => almalinux-9.Dockerfile} (91%) rename ci/containers/{fedora-38-cross-mingw32.Dockerfile => fedora-40-cross-mingw32.Dockerfile} (97%) rename ci/containers/{fedora-38-cross-mingw64.Dockerfile => fedora-40-cross-mingw64.Dockerfile} (97%) rename ci/containers/{fedora-38.Dockerfile => fedora-40.Dockerfile} (97%) rename ci/containers/{ubuntu-2004.Dockerfile => ubuntu-2404.Dockerfile} (93%) -- 2.43.2

4 33

[PATCH 00/12] fix many test failures in various build config scenarios
by Daniel P. Berrangé 07 May '24

07 May '24

I was annoyed that meson build --auto-features=disabled -Dtests=enabled will fail in unit tests. Rather than just fix that scenario, I went down the rabbit hole and tested (almost) every single minimal build option. By that I mean I looked at meson_options.txt and got a list of every option that has 'auto' as its default. I then attempted a minimal build with only that enabled. I quickly found some options have mandatory pre-requisite options, so I re-did the tests with such deps present. This uncovered many scenarios where we didn't use the right conditions for tests. Surprisingly (at first, but not in retrospect) it also found many places where the tests failed to depend on the earlier build artifacts. My test process was: opts=$(grep auto meson_options.txt | sed -e 's/option(//' -e 's/,.*//' -e "s/'//g") for opt in $opts do deps=`grep --before 1 "option('$opt'" meson_options.txt | grep dep: | head -1 | sed -e 's/# //' -e 's/dep:/-D/g' -e 's/$/=enabled/' -e 's/ -D/=enabled -D/g'` echo "Try test $opt with $deps" rm -rf build meson setup build --auto-features=disabled -Dtests=enabled $deps -D$opt=enabled 1> opt-$opt.log 2>&1 meson test -C build --no-suite syntax-check --print-errorlogs 1>> opt-$opt.log 2>&1 ; done done Then to check results grep '^Fail:' opt-*.log | grep -v 0 This is slow, but not as slow as you might think, because each build is very minimal, so we're not actually building more than 300-400 files each time, rather than 1000's, and we're not running as many tests either. Might be interesting to get extensive test into CI. Strictly speaking we care about the combinatorial expansion of meson_options.txt, but that would be insanely slow and is likely to be overkill. Daniel P. Berrangé (12): scripts/rpcgen: skip tests if tirpc is not present tests: fix tests when test driver is disabled meson: record which other options are a pre-requisite docs: ensure HTML/images are built before running reference tests src: ensure augeas test file is generated before running test tests: build 'virsh' before running virsh-auth test tests: build driver modules before virdrivermoduletest test: conditionalize 'virsh-auth' on test driver tests: always build securityselinuxhelper if libselinux is present test: drop bogus check for YAJL from libxl test/mock tests: don't run mdevctl test if lacking YAJL src/node_device: don't overwrite error messages docs/images/meson.build | 5 ++- docs/logos/meson.build | 5 ++- docs/meson.build | 1 + meson_options.txt | 20 +++++++++ scripts/rpcgen/meson.build | 2 +- src/meson.build | 6 ++- src/node_device/node_device_driver.c | 2 - tests/fchosttest.c | 9 ++-- tests/libxlmock.c | 4 +- tests/libxlxml2domconfigtest.c | 4 +- tests/meson.build | 67 ++++++++++++++++++---------- 11 files changed, 87 insertions(+), 38 deletions(-) -- 2.43.0

3 20

[PATCH 0/4] Introduce SSH proxy
by Michal Privoznik 07 May '24

07 May '24

*** BLURB HERE *** Michal Prívozník (4): datatypes: Declare g_autoptr cleanup functions for more public objects tools: Introduce SSH proxy docs: Document SSH proxy NEWS: Document SSH proxy feature NEWS.rst | 5 + docs/docs.rst | 3 + docs/meson.build | 1 + docs/nss.rst | 7 + docs/ssh-proxy.rst | 60 +++++ libvirt.spec.in | 27 +++ meson.build | 16 +- meson_options.txt | 2 + po/POTFILES | 1 + src/datatypes.h | 16 ++ tools/meson.build | 2 + tools/ssh-proxy/30-libvirt-ssh-proxy.conf.in | 10 + tools/ssh-proxy/meson.build | 25 ++ tools/ssh-proxy/ssh-proxy.c | 233 +++++++++++++++++++ 14 files changed, 407 insertions(+), 1 deletion(-) create mode 100644 docs/ssh-proxy.rst create mode 100644 tools/ssh-proxy/30-libvirt-ssh-proxy.conf.in create mode 100644 tools/ssh-proxy/meson.build create mode 100644 tools/ssh-proxy/ssh-proxy.c -- 2.43.2

3 8

[PATCH 0/2] interface: Refactor code and fix old listing API when interface is unbound
by Peter Krempa 07 May '24

07 May '24

Patch 1/2 removes a pointless helper in favor of making an almost identical function more universal, so that patch 2/2 then fixes all cases in one place. Peter Krempa (2): interface_udev: Replace udevNumOfInterfacesByStatus by udevListInterfacesByStatus udevListInterfacesByStatus: Don't try to return NULL names src/interface/interface_backend_udev.c | 84 +++++++++----------------- 1 file changed, 30 insertions(+), 54 deletions(-) -- 2.44.0

2 3

[PATCH] qemu_saveimage: add zstd to supported compression formats
by Adam Julis 06 May '24

06 May '24

Extend the list of supported formats, update and clarify comment in qemu.conf.in (removed misleading sentence about the order of compression format types). Signed-off-by: Adam Julis <ajulis(a)redhat.com> --- libvirt.spec.in | 1 + src/qemu/qemu.conf.in | 7 +++---- src/qemu/qemu_saveimage.c | 2 ++ 3 files changed, 6 insertions(+), 4 deletions(-) diff --git a/libvirt.spec.in b/libvirt.spec.in index 64018192b6..88c62f6d92 100644 --- a/libvirt.spec.in +++ b/libvirt.spec.in @@ -815,6 +815,7 @@ Requires: gzip Requires: bzip2 Requires: lzop Requires: xz +Requires: zstd Requires: systemd-container Requires: swtpm-tools %if %{with_numad} diff --git a/src/qemu/qemu.conf.in b/src/qemu/qemu.conf.in index f406df8749..6bc2140dcb 100644 --- a/src/qemu/qemu.conf.in +++ b/src/qemu/qemu.conf.in @@ -582,10 +582,9 @@ # memory from the domain is dumped out directly to a file. If you have # guests with a large amount of memory, however, this can take up quite # a bit of space. If you would like to compress the images while they -# are being saved to disk, you can also set "lzop", "gzip", "bzip2", or "xz" -# for save_image_format. Note that this means you slow down the process of -# saving a domain in order to save disk space; the list above is in descending -# order by performance and ascending order by compression ratio. +# are being saved to disk, you can also set "zstd", "lzop", "gzip", "bzip2", +# or "xz" for save_image_format. Note that this means you slow down the process +# of saving a domain in order to save disk space. # # save_image_format is used when you use 'virsh save' or 'virsh managedsave' # at scheduled saving, and it is an error if the specified save_image_format diff --git a/src/qemu/qemu_saveimage.c b/src/qemu/qemu_saveimage.c index 89112e3e44..018ab5a222 100644 --- a/src/qemu/qemu_saveimage.c +++ b/src/qemu/qemu_saveimage.c @@ -47,6 +47,7 @@ typedef enum { */ QEMU_SAVE_FORMAT_XZ = 3, QEMU_SAVE_FORMAT_LZOP = 4, + QEMU_SAVE_FORMAT_ZSTD = 5, /* Note: add new members only at the end. These values are used in the on-disk format. Do not change or re-use numbers. */ @@ -62,6 +63,7 @@ VIR_ENUM_IMPL(qemuSaveCompression, "bzip2", "xz", "lzop", + "zstd", ); static inline void -- 2.44.0

2 1

[PATCH v2 0/4] qemu: Substract isolcpus from all online affinity
by Michal Privoznik 06 May '24

06 May '24

v2 of: https://lists.libvirt.org/archives/list/devel@lists.libvirt.org/thread/4V7O… diff to v1: - Don't error out on systems where /sys/devices/system/cpu/isolated is unavailable. - Don't error out on systems where /sys/devices/system/cpu/isolated is empty. Both of these resulted in new patches. Michal Prívozník (4): virbitmap: Introduce virBitmapParseUnlimitedAllowEmpty() virfile: Introduce virFileReadValueBitmapAllowEmpty() virhostcpu: Introduce virHostCPUGetIsolated() qemu: Substract isolcpus from all online affinity src/libvirt_private.syms | 3 ++ src/qemu/qemu_process.c | 9 +++++ src/util/virbitmap.c | 40 +++++++++++++++++--- src/util/virbitmap.h | 3 ++ src/util/virfile.c | 81 ++++++++++++++++++++++++++++++---------- src/util/virfile.h | 2 + src/util/virhostcpu.c | 31 +++++++++++++++ src/util/virhostcpu.h | 1 + tests/virbitmaptest.c | 40 ++++++++++++++++++++ 9 files changed, 186 insertions(+), 24 deletions(-) -- 2.43.2

3 7

[PATCH] NEWS: document qemu: ras as a new feature
by Kristina Hanicova 06 May '24

06 May '24

Signed-off-by: Kristina Hanicova <khanicov(a)redhat.com> --- NEWS.rst | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/NEWS.rst b/NEWS.rst index 5a771b4b2f..d72c15bf10 100644 --- a/NEWS.rst +++ b/NEWS.rst @@ -17,6 +17,11 @@ v10.4.0 (unreleased) * **New features** + * qemu: Support for ras feature for virt machine type + + It is now possible to set on/off ``ras`` feature in the domain XML for virt + (Arm) machine type as ``<ras state='on'/>``. + * **Improvements** * **Bug fixes** -- 2.42.0

2 1

[PATCH] vsh: Don't init history in cmdComplete()
by Michal Privoznik 06 May '24

06 May '24

Recent rework of virshtest uncovered a subtle bug that was dormant in now vsh but before that even in monolithic virsh. In vsh.c there's this vshReadlineInit() function that's supposed to initialize readline library, i.e. set those global rl_* pointers. But it also initializes history library. Then, when virsh/virt-admin quits, vshReadlineDeinit() is called which writes history into a file (ensuring the parent directory exists). So far no problem. Problem arises when cmdComplete() is called (from a bash completer, for instance). It does not guard call to vshReadlineInit() with check for interactive shell (and it should not), but it sets ctl->historyfile which signals to vshReadlineDeinit() the history should be written. Now, no real history is written, because nothing was entered on the stdin, but the parent directory is created nevertheless. With recent movement in virshtest.c this means some test cases might create virsh history file which breaks our promise of not touching user's data in test suite. Resolves: https://bugs.gentoo.org/931109 Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- tools/vsh.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/tools/vsh.c b/tools/vsh.c index 58855f63ba..e74045c24e 100644 --- a/tools/vsh.c +++ b/tools/vsh.c @@ -2973,7 +2973,7 @@ vshReadlineInit(vshControl *ctl) const char *quote_characters = "\"'"; /* initialize readline stuff only once */ - if (ctl->historydir) + if (autoCompleteOpaque) return 0; /* Opaque data for autocomplete callbacks. */ @@ -2989,6 +2989,11 @@ vshReadlineInit(vshControl *ctl) rl_completer_quote_characters = quote_characters; rl_char_is_quoted_p = vshReadlineCharIsQuoted; + /* Stuff below is needed only for interactive mode. */ + if (!ctl->imode) { + return 0; + } + histsize_env = g_strdup_printf("%s_HISTSIZE", ctl->env_prefix); /* Limit the total size of the history buffer */ @@ -3149,7 +3154,7 @@ vshInit(vshControl *ctl, const vshCmdGrp *groups) cmdGroups = groups; if (vshInitDebug(ctl) < 0 || - (ctl->imode && vshReadlineInit(ctl) < 0)) + vshReadlineInit(ctl) < 0) return false; return true; @@ -3168,7 +3173,7 @@ vshInitReload(vshControl *ctl) if (ctl->imode) vshReadlineDeinit(ctl); - if (ctl->imode && vshReadlineInit(ctl) < 0) + if (vshReadlineInit(ctl) < 0) return false; return true; -- 2.43.2

3 2

[PATCH] hyperv: prevent potential NULL dereference
by Oleg Sviridov 06 May '24

06 May '24

Return value of a function 'virDomainChrDefNew' is dereferenced at hyperv_driver.c without checking for NULL, which can lead to NULL dereference immediatly after. Found by Linux Verification Center (linuxtesting.org) with SVACE. Signed-off-by: Oleg Sviridov <oleg.sviridov(a)red-soft.ru> --- src/hyperv/hyperv_driver.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/src/hyperv/hyperv_driver.c b/src/hyperv/hyperv_driver.c index 414274fdfd..7580c6a06c 100644 --- a/src/hyperv/hyperv_driver.c +++ b/src/hyperv/hyperv_driver.c @@ -1534,7 +1534,8 @@ hypervDomainDefParseSerial(virDomainDef *def, Msvm_ResourceAllocationSettingData continue; } - serial = virDomainChrDefNew(NULL); + if (!(serial = virDomainChrDefNew(NULL))) + return -1; serial->deviceType = VIR_DOMAIN_CHR_DEVICE_TYPE_SERIAL; serial->source->type = VIR_DOMAIN_CHR_TYPE_PIPE; -- 2.44.0

3 2

The Raptors defeated the Warriors
by n3x4pmgf＠nqmo.com 04 May '24

04 May '24

If The watch solutions feel mind-boggling, it may help to narrow your focus. Probably you merely want to know you’re finding something from a revered brand you could have faith in. https://www.newzealandescortspage.com

1 0