January 2025 - Devel - libvirt List Archives

[libvirt PATCH v2 0/2] conf: Adjust hyperv tlbflush formatting
by Ján Tomko 07 Jan '25

07 Jan '25

Including the refactor. Ján Tomko (1): conf: refactor hyperv features formatting Martin Kletzander (1): conf: Adjust hyperv tlbflush formatting src/conf/domain_conf.c | 57 ++++++++++--------- .../qemuxmlconfdata/hyperv.x86_64-latest.xml | 5 +- 2 files changed, 33 insertions(+), 29 deletions(-) -- 2.47.0

2 3

[PATCH] security: apparmor: Remove hardcoded "libvirtd" profile name
by Jim Fehlig 07 Jan '25

07 Jan '25

The apparmor driver probe function checks for an active profile matching the full path of the running daemon binary. If not found, it checks for a profile named "libvirtd". This works fine when the running daemon is the old monolithic libvirtd, but fails with modular daemons. Remove the check for a hardcoded "libvirtd" profile and replace with the basename of the running daemon binary. Signed-off-by: Jim Fehlig <jfehlig(a)suse.com> --- src/security/security_apparmor.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/src/security/security_apparmor.c b/src/security/security_apparmor.c index c8e77c6cd2..eed0f265d6 100644 --- a/src/security/security_apparmor.c +++ b/src/security/security_apparmor.c @@ -239,7 +239,9 @@ use_apparmor(void) */ rc = profile_status(libvirt_daemon, 1); if (rc < 0) { - rc = profile_status("libvirtd", 1); + g_autofree char *basename = g_path_get_basename(libvirt_daemon); + + rc = profile_status(basename, 1); /* Error or unconfined should all result in -1 */ if (rc < 0) rc = -1; -- 2.43.0

2 1

[PATCH] conf: Adjust hyperv tlbflush formatting
by Martin Kletzander 07 Jan '25

07 Jan '25

Commi 247357cc292a added support for direct and extended modes for tlbflush, but forgot to do the formatting as well. Instead of rewriting the whole hyperv feature formatting to use yet another attribute and child buffers, just fix it in a way the other features are and leave the refactoring to later. One reason for that is that we should fix this ASAP since without this patch the features will be stripped when formatting the XML on the disk and lost after next daemon reload. Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com> --- src/conf/domain_conf.c | 20 ++++++++++++++++++- .../qemuxmlconfdata/hyperv.x86_64-latest.xml | 5 ++++- 2 files changed, 23 insertions(+), 2 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 785eb0e539b4..ae56b292ad1e 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -27972,7 +27972,6 @@ virDomainDefFormatFeatures(virBuffer *buf, case VIR_DOMAIN_HYPERV_RESET: case VIR_DOMAIN_HYPERV_FREQUENCIES: case VIR_DOMAIN_HYPERV_REENLIGHTENMENT: - case VIR_DOMAIN_HYPERV_TLBFLUSH: case VIR_DOMAIN_HYPERV_IPI: case VIR_DOMAIN_HYPERV_EVMCS: case VIR_DOMAIN_HYPERV_AVIC: @@ -28016,6 +28015,25 @@ virDomainDefFormatFeatures(virBuffer *buf, def->hyperv_vendor_id); break; + case VIR_DOMAIN_HYPERV_TLBFLUSH: + if (def->hyperv_features[j] != VIR_TRISTATE_SWITCH_ON) { + virBufferAddLit(&childBuf, "/>\n"); + break; + } + if (def->hyperv_tlbflush_direct == VIR_TRISTATE_SWITCH_ON || + def->hyperv_tlbflush_extended == VIR_TRISTATE_SWITCH_ON) { + virBufferAddLit(&childBuf, ">\n"); + virBufferAdjustIndent(&childBuf, 2); + if (def->hyperv_tlbflush_direct == VIR_TRISTATE_SWITCH_ON) + virBufferAddLit(&childBuf, "<direct state='on'/>\n"); + if (def->hyperv_tlbflush_extended == VIR_TRISTATE_SWITCH_ON) + virBufferAddLit(&childBuf, "<extended state='on'/>\n"); + virBufferAdjustIndent(&childBuf, -2); + virBufferAddLit(&childBuf, "</tlbflush>\n"); + } else { + virBufferAddLit(&childBuf, "/>\n"); + } + case VIR_DOMAIN_HYPERV_LAST: break; } diff --git a/tests/qemuxmlconfdata/hyperv.x86_64-latest.xml b/tests/qemuxmlconfdata/hyperv.x86_64-latest.xml index 36d9161fa811..49537188af3f 100644 --- a/tests/qemuxmlconfdata/hyperv.x86_64-latest.xml +++ b/tests/qemuxmlconfdata/hyperv.x86_64-latest.xml @@ -22,7 +22,10 @@ <vendor_id state='on' value='KVM Hv'/> <frequencies state='on'/> <reenlightenment state='on'/> - <tlbflush state='on'/> + <tlbflush state='on'> + <direct state='on'/> + <extended state='on'/> + </tlbflush> <ipi state='on'/> <evmcs state='on'/> <avic state='on'/> -- 2.47.1

2 1

[PATCH v2 1/3] Revert "qemu: explicit swtpm state locking"
by marcandre.lureau＠redhat.com 07 Jan '25

07 Jan '25

From: Marc-André Lureau <marcandre.lureau(a)redhat.com> This reverts commit bb5e26749fe5b5856a3541be2cbe147701e6e121. swtpm-setup doesn't have "tpmstate-lock", only swtpm. Signed-off-by: Marc-André Lureau <marcandre.lureau(a)redhat.com> --- src/qemu/qemu_tpm.c | 11 ++--------- src/util/virtpm.c | 1 - src/util/virtpm.h | 1 - tests/testutilsqemu.c | 1 - 4 files changed, 2 insertions(+), 12 deletions(-) diff --git a/src/qemu/qemu_tpm.c b/src/qemu/qemu_tpm.c index f5e0184e54..476e3dd224 100644 --- a/src/qemu/qemu_tpm.c +++ b/src/qemu/qemu_tpm.c @@ -344,23 +344,16 @@ static char * qemuTPMGetSwtpmSetupStateArg(const virDomainTPMSourceType source_type, const char *source_path) { - const char *lock = ",lock"; - - if (!virTPMSwtpmSetupCapsGet(VIR_TPM_SWTPM_SETUP_FEATURE_TPMSTATE_OPT_LOCK)) { - VIR_WARN("This swtpm version doesn't support explicit locking"); - lock = ""; - } - switch (source_type) { case VIR_DOMAIN_TPM_SOURCE_TYPE_FILE: /* the file:// prefix is supported since swtpm_setup 0.7.0 */ /* assume the capability check for swtpm is redundant. */ - return g_strdup_printf("file://%s%s", source_path, lock); + return g_strdup_printf("file://%s", source_path); case VIR_DOMAIN_TPM_SOURCE_TYPE_DIR: case VIR_DOMAIN_TPM_SOURCE_TYPE_DEFAULT: case VIR_DOMAIN_TPM_SOURCE_TYPE_LAST: default: - return g_strdup_printf("%s%s", source_path, lock); + return g_strdup_printf("%s", source_path); } } diff --git a/src/util/virtpm.c b/src/util/virtpm.c index 4016ad8fc4..f90839debe 100644 --- a/src/util/virtpm.c +++ b/src/util/virtpm.c @@ -53,7 +53,6 @@ VIR_ENUM_IMPL(virTPMSwtpmSetupFeature, "cmdarg-reconfigure-pcr-banks", "tpm-1.2", "tpm-2.0", - "tpmstate-opt-lock", "cmdarg-profile", ); diff --git a/src/util/virtpm.h b/src/util/virtpm.h index 03fb92629a..4119a903e5 100644 --- a/src/util/virtpm.h +++ b/src/util/virtpm.h @@ -45,7 +45,6 @@ typedef enum { VIR_TPM_SWTPM_SETUP_FEATURE_CMDARG_RECONFIGURE_PCR_BANKS, VIR_TPM_SWTPM_SETUP_FEATURE_TPM_1_2, VIR_TPM_SWTPM_SETUP_FEATURE_TPM_2_0, - VIR_TPM_SWTPM_SETUP_FEATURE_TPMSTATE_OPT_LOCK, VIR_TPM_SWTPM_SETUP_FEATURE_CMDARG_PROFILE, VIR_TPM_SWTPM_SETUP_FEATURE_LAST diff --git a/tests/testutilsqemu.c b/tests/testutilsqemu.c index 5caccbc6b4..abc425b9b7 100644 --- a/tests/testutilsqemu.c +++ b/tests/testutilsqemu.c @@ -71,7 +71,6 @@ virTPMSwtpmSetupCapsGet(virTPMSwtpmSetupFeature cap) case VIR_TPM_SWTPM_SETUP_FEATURE_CMDARG_CREATE_CONFIG_FILES: case VIR_TPM_SWTPM_SETUP_FEATURE_TPM12_NOT_NEED_ROOT: case VIR_TPM_SWTPM_SETUP_FEATURE_CMDARG_RECONFIGURE_PCR_BANKS: - case VIR_TPM_SWTPM_SETUP_FEATURE_TPMSTATE_OPT_LOCK: case VIR_TPM_SWTPM_SETUP_FEATURE_CMDARG_PROFILE: case VIR_TPM_SWTPM_SETUP_FEATURE_LAST: break; -- 2.47.0

4 6

[PATCH] apparmor: Allow running loongarch64 VMs on Debian 12
by Xianglai Li 06 Jan '25

06 Jan '25

Allows to load firmware in the qemu-efi-loongarch64 directory Allows the binary qemu-system-loongarch64 to be run This makes it impossible to run loongarch64 VMs when AppArmor is enabled Signed-off-by: Xianglai Li <lixianglai(a)loongson.cn> --- src/security/apparmor/libvirt-qemu.in | 1 + src/security/virt-aa-helper.c | 1 + 2 files changed, 2 insertions(+) diff --git a/src/security/apparmor/libvirt-qemu.in b/src/security/apparmor/libvirt-qemu.in index 694da26dea..c63077574e 100644 --- a/src/security/apparmor/libvirt-qemu.in +++ b/src/security/apparmor/libvirt-qemu.in @@ -144,6 +144,7 @@ /usr/bin/qemu-system-hppa rmix, /usr/bin/qemu-system-i386 rmix, /usr/bin/qemu-system-lm32 rmix, + /usr/bin/qemu-system-loongarch64 rmix, /usr/bin/qemu-system-m68k rmix, /usr/bin/qemu-system-microblaze rmix, /usr/bin/qemu-system-microblazeel rmix, diff --git a/src/security/virt-aa-helper.c b/src/security/virt-aa-helper.c index 1cf9d7ad3d..94a28bf331 100644 --- a/src/security/virt-aa-helper.c +++ b/src/security/virt-aa-helper.c @@ -481,6 +481,7 @@ valid_path(const char *path, const bool readonly) "/usr/share/AAVMF/", "/usr/share/qemu-efi/", /* for AAVMF images */ "/usr/share/qemu-efi-aarch64/", + "/usr/share/qemu-efi-loongarch64/", "/usr/share/qemu-efi-riscv64/", "/usr/share/qemu/", /* SUSE path for OVMF and AAVMF images */ "/usr/lib/u-boot/", -- 2.39.1

3 5

[PATCH] hyperv: Introduce and export 'facility' variable.
by Michal Privoznik 06 Jan '25

06 Jan '25

In its upstream commit [1] openwsman dropped 'facility' variable which is documented as: * all processes that use the libu must define a "facility" variable somewhere * to satisfy this external linkage reference. * * Such variable will be used as the syslog(3) facility argument. Well, prior to that commit, openwsman itself declared the variable (and set it to LOG_DAEMON). Now it's up to us. Yeah, the variable naming is terrible and also I we are not using libu directly, but apparently libwsman.so requires it anyway: $ objdump -T /usr/lib64/libwsman.so | grep facility 0000000000000000 D *UND* 0000000000000000 Base facility 1: https://github.com/Openwsman/openwsman/commit/d72c51f21b9c85a773b7955ac587d… Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/hyperv/hyperv_wmi.c | 3 +++ src/libvirt_hyperv.syms | 11 +++++++++++ src/meson.build | 6 ++++++ 3 files changed, 20 insertions(+) create mode 100644 src/libvirt_hyperv.syms diff --git a/src/hyperv/hyperv_wmi.c b/src/hyperv/hyperv_wmi.c index 0b82f1f131..040bcfec11 100644 --- a/src/hyperv/hyperv_wmi.c +++ b/src/hyperv/hyperv_wmi.c @@ -28,6 +28,7 @@ #include <wsman-soap.h> #include <wsman-xml.h> #include <wsman-xml-binding.h> +#include <u/syslog.h> #include "internal.h" #include "virerror.h" @@ -47,6 +48,8 @@ VIR_LOG_INIT("hyperv.hyperv_wmi"); +int facility = LOG_DAEMON; + int hypervGetWmiClassList(hypervPrivate *priv, hypervWmiClassInfo *wmiInfo, virBuffer *query, hypervObject **wmiClass) diff --git a/src/libvirt_hyperv.syms b/src/libvirt_hyperv.syms new file mode 100644 index 0000000000..60e98db473 --- /dev/null +++ b/src/libvirt_hyperv.syms @@ -0,0 +1,11 @@ +# +# HyperV-specific symbols +# + +# hyperv/hyperv_wmi.c +facility; + +# Let emacs know we want case-insensitive sorting +# Local Variables: +# sort-fold-case: t +# End: diff --git a/src/meson.build b/src/meson.build index b53ea2a71f..cce89fac27 100644 --- a/src/meson.build +++ b/src/meson.build @@ -125,6 +125,12 @@ else sym_files += 'libvirt_libssh2.syms' endif +if conf.has('WITH_HYPERV') + used_sym_files += 'libvirt_hyperv.syms' +else + sym_files += 'libvirt_hyperv.syms' +endif + # variables filled by subdirectories libvirt_libs = [] -- 2.45.2

2 1

[PATCH v2 0/3] po: handle translation of polkit policy file strings
by Daniel P. Berrangé 06 Jan '25

06 Jan '25

There was a proposal https://gitlab.com/libvirt/libvirt/-/merge_requests/387 to add translations for the polkit files. In reviewing this we came to the conclusion the approach was undesirable. After getting misled by a Debian/Ubuntu specific downstream only patch to polkit which auto-translated polkit files at runtime, this implements the manual approach of merging translations into the polkit files at build time. Changed in v2: - Fix introduction of 'polkitactionsdir' - Also introduce 'polkitrulesdir' - Fix syntax check rules for POFILES with generated polkit - Remove incorrect 'install: true' rule for policy.in file Daniel P. Berrangé (3): po: add its rules for translating polkit file strings meson: remove duplication of polkit dirs construction remote: apply translations to polkit files build-aux/syntax-check.mk | 5 +++-- meson.build | 6 ++++++ po/POTFILES | 2 ++ po/its/polkit.its | 8 ++++++++ po/its/polkit.loc | 6 ++++++ po/meson.build | 5 ++--- src/access/meson.build | 20 +++++++++++++++---- .../{libvirtd.policy => libvirtd.policy.in} | 0 src/remote/meson.build | 15 ++++++++------ 9 files changed, 52 insertions(+), 15 deletions(-) create mode 100644 po/its/polkit.its create mode 100644 po/its/polkit.loc rename src/remote/{libvirtd.policy => libvirtd.policy.in} (100%) -- 2.47.1

2 4

[PATCH] Remove the deprecated "-runas" command line option
by Thomas Huth 04 Jan '25

04 Jan '25

It has been marked as deprecated two releases ago, so it should be fine now to remove this command line option. Signed-off-by: Thomas Huth <thuth(a)redhat.com> --- docs/about/deprecated.rst | 6 ------ docs/about/removed-features.rst | 6 ++++++ system/vl.c | 9 --------- qemu-options.hx | 15 +-------------- 4 files changed, 7 insertions(+), 29 deletions(-) diff --git a/docs/about/deprecated.rst b/docs/about/deprecated.rst index d6809f94ea..63b46fd520 100644 --- a/docs/about/deprecated.rst +++ b/docs/about/deprecated.rst @@ -74,12 +74,6 @@ configurations (e.g. -smp drawers=1,books=1,clusters=1 for x86 PC machine) is marked deprecated since 9.0, users have to ensure that all the topology members described with -smp are supported by the target machine. -``-runas`` (since 9.1) -'''''''''''''''''''''' - -Use ``-run-with user=..`` instead. - - User-mode emulator command line arguments ----------------------------------------- diff --git a/docs/about/removed-features.rst b/docs/about/removed-features.rst index cb1388049a..c6616ce05e 100644 --- a/docs/about/removed-features.rst +++ b/docs/about/removed-features.rst @@ -555,6 +555,12 @@ to produce an odd effect (rotating input but not display output). But this was never intended or documented behaviour, so we have dropped the options along with the machine models they were intended for. +``-runas`` (removed in 10.0) +'''''''''''''''''''''''''''' + +Use ``-run-with user=..`` instead. + + User-mode emulator command line arguments ----------------------------------------- diff --git a/system/vl.c b/system/vl.c index 0843b7ab49..3c5bd36d7d 100644 --- a/system/vl.c +++ b/system/vl.c @@ -3611,15 +3611,6 @@ void qemu_init(int argc, char **argv) /* Nothing to be parsed here. Especially, do not error out below. */ break; #if defined(CONFIG_POSIX) - case QEMU_OPTION_runas: - warn_report("-runas is deprecated, use '-run-with user=...' instead"); - if (!os_set_runas(optarg)) { - error_report("User \"%s\" doesn't exist" - " (and is not <uid>:<gid>)", - optarg); - exit(1); - } - break; case QEMU_OPTION_daemonize: os_set_daemonize(true); break; diff --git a/qemu-options.hx b/qemu-options.hx index cc694d3b89..7090d59f6f 100644 --- a/qemu-options.hx +++ b/qemu-options.hx @@ -4976,19 +4976,6 @@ SRST ``-nodefaults`` option will disable all those default devices. ERST -#ifndef _WIN32 -DEF("runas", HAS_ARG, QEMU_OPTION_runas, \ - "-runas user change to user id user just before starting the VM\n" \ - " user can be numeric uid:gid instead\n", - QEMU_ARCH_ALL) -#endif -SRST -``-runas user`` - Immediately before starting guest execution, drop root privileges, - switching to the specified user. This option is deprecated, use - ``-run-with user=...`` instead. -ERST - DEF("prom-env", HAS_ARG, QEMU_OPTION_prom_env, "-prom-env variable=value\n" " set OpenBIOS nvram variables\n", @@ -5176,7 +5163,7 @@ SRST ``chroot=dir`` can be used for doing a chroot to the specified directory immediately before starting the guest execution. This is especially useful - in combination with -runas. + in combination with ``user=...``. ``user=username`` or ``user=uid:gid`` can be used to drop root privileges before starting guest execution. QEMU will use the ``setuid`` and ``setgid`` -- 2.47.1

2 1

[PATCH 0/2] po: handle translatin of polkit policy file strings
by Daniel P. Berrangé 04 Jan '25

04 Jan '25

There was a proposal https://gitlab.com/libvirt/libvirt/-/merge_requests/387 to add translations for the polkit files. In reviewing this we came to the conclusion the approach was undesirable. After getting misled by a Debian/Ubuntu specific downstream only patch to polkit which auto-translated polkit files at runtime, this implements the manual approach of merging translations into the polkit files at build time. Daniel P. Berrangé (2): po: add its rules for translating polkit file strings remote: apply translations to polkit files meson.build | 5 +++++ po/POTFILES | 2 ++ po/its/polkit.its | 8 ++++++++ po/its/polkit.loc | 6 ++++++ po/meson.build | 3 +-- src/access/meson.build | 18 ++++++++++++++---- .../{libvirtd.policy => libvirtd.policy.in} | 0 src/remote/meson.build | 13 ++++++++----- 8 files changed, 44 insertions(+), 11 deletions(-) create mode 100644 po/its/polkit.its create mode 100644 po/its/polkit.loc rename src/remote/{libvirtd.policy => libvirtd.policy.in} (100%) -- 2.46.0

2 4

[PATCH 00/19] hw/microblaze: Allow running cross-endian vCPUs
by Philippe Mathieu-Daudé 02 Jan '25

02 Jan '25

Make machines endianness-agnostic, allowing to run a big-endian vCPU on the little-endian 'qemu-system-microblazeel' binary, and a little endian one on the big-endian 'qemu-system-microblaze' binary. Tests added, following combinations covered: - little-endian vCPU using little-endian binary (in-tree) - little-endian vCPU using big-endian binary (new) - big-endian vCPU using little-endian binary (new) - big-endian vCPU using big-endian binary (in-tree) Deprecate untested big-endian machines, likely build on the big endian binary by mistake: - petalogix-ml605 - xlnx-zynqmp-pmu To make a target endian-agnostic we need to remove the MO_TE uses. In order to do that, we propagate the MemOp from earlier in the call stack, or we extract it from the vCPU env (on MicroBlaze the CPU endianness is exposed by the 'ENDI' bit). Note, since vCPU can run in any endianness, the MemoryRegionOps::endianness should not be DEVICE_NATIVE_ENDIAN anymore, because this definition expand to the binary endianness, swapping data regardless how the vcpu access it. See adjust_endianness() -> devend_memop(). Something to keep in mind, possibly requiring further work and optimizations (avoid double-swap). Next step: Look at unifying binaries. Please review, Phil. Philippe Mathieu-Daudé (19): target/microblaze: Rename CPU endianness property as 'little-endian' hw/microblaze: Deprecate big-endian petalogix-ml605 & xlnx-zynqmp-pmu hw/microblaze/s3adsp1800: Explicit CPU endianness hw/microblaze/s3adsp1800: Rename unimplemented MMIO region as xps_gpio hw/microblaze/s3adsp1800: Declare machine type using DEFINE_TYPES macro hw/microblaze: Fix MemoryRegionOps coding style hw/microblaze: Restrict MemoryRegionOps are implemented as 32-bit hw/microblaze: Propagate CPU endianness to microblaze_load_kernel() hw/intc/xilinx_intc: Only expect big-endian accesses hw/timer/xilinx_timer: Only expect big-endian accesses hw/timer/xilinx_timer: Allow down to 8-bit memory access hw/net/xilinx_ethlite: Only expect big-endian accesses target/microblaze: Explode MO_TExx -> MO_TE | MO_xx target/microblaze: Set MO_TE once in do_load() / do_store() target/microblaze: Introduce mo_endian() helper target/microblaze: Consider endianness while translating code hw/microblaze: Support various endianness for s3adsp1800 machines tests/functional: Explicit endianness of microblaze assets tests/functional: Add microblaze cross-endianness tests docs/about/deprecated.rst | 6 ++ .../devices/microblaze-softmmu/default.mak | 2 - .../devices/microblazeel-softmmu/default.mak | 5 +- hw/microblaze/boot.h | 4 +- target/microblaze/cpu.h | 7 ++ hw/char/xilinx_uartlite.c | 8 ++- hw/intc/xilinx_intc.c | 23 +++++-- hw/microblaze/boot.c | 8 +-- hw/microblaze/petalogix_ml605_mmu.c | 11 ++- hw/microblaze/petalogix_s3adsp1800_mmu.c | 67 +++++++++++++++++-- hw/microblaze/xlnx-zynqmp-pmu.c | 12 ++-- hw/net/xilinx_ethlite.c | 28 ++++++-- hw/timer/xilinx_timer.c | 15 +++-- target/microblaze/cpu.c | 2 +- target/microblaze/translate.c | 49 ++++++++------ .../functional/test_microblaze_s3adsp1800.py | 27 +++++++- .../test_microblazeel_s3adsp1800.py | 25 ++++++- 17 files changed, 236 insertions(+), 63 deletions(-) -- 2.45.2

7 71