November 2020 - Devel - libvirt List Archives

[libvirt PATCH v2] Do not disable incompatible-pointer-types-discards-qualifiers
by Martin Kletzander 16 Nov '20

16 Nov '20

This reverts commit b3710e9a2af402a2b620de570b062294e11190eb. That check is very valuable for our code, but it causes issue with glib >= 2.67.0 when building with clang. The reason is a combination of two commits in glib, firstly fdda405b6b1b which adds a g_atomic_pointer_{set,get} variants that enforce stricter type checking (by removing an extra cast) for compilers that support __typeof__, and commit dce24dc4492d which effectively enabled the new variant of glib's atomic code for clang. This will not be necessary when glib's issue #600 [0] (8 years old) is fixed. Thankfully, MR #1719 [1], which is supposed to deal with this issue was opened 3 weeks ago, so there is a slight sliver of hope. [0] https://gitlab.gnome.org/GNOME/glib/-/issues/600 [1] https://gitlab.gnome.org/GNOME/glib/-/merge_requests/1719 Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com> --- meson.build | 3 --- src/util/glibcompat.h | 24 ++++++++++++++++++++++++ 2 files changed, 24 insertions(+), 3 deletions(-) diff --git a/meson.build b/meson.build index cecaad199d4c..04646e3a078c 100644 --- a/meson.build +++ b/meson.build @@ -405,9 +405,6 @@ cc_flags += [ # so use this Clang-specific arg to keep it quiet '-Wno-typedef-redefinition', - # Clang complains about casts in G_DEFINE_TYPE(...) - '-Wno-incompatible-pointer-types-discards-qualifiers', - # We don't use -Wc++-compat so we have to enable it explicitly '-Wjump-misses-init', diff --git a/src/util/glibcompat.h b/src/util/glibcompat.h index 6f50a76f3c2b..f4867f3eadbc 100644 --- a/src/util/glibcompat.h +++ b/src/util/glibcompat.h @@ -20,6 +20,30 @@ #include <glib.h> #include <glib/gstdio.h> +#include <glib-object.h> + +#ifdef __clang__ +/* + * Clang detects (valid) issue in G_DEFINE_TYPE and derivatives starting with + * glib >= 2.67.0. See https://gitlab.gnome.org/GNOME/glib/-/issues/600 + * + * For that we need to disable the one check that produces an error in our + * builds when using any G_DEFINE_TYPE* macro. Thankfully all those macros end + * up using _G_DEFINE_TYPE_EXTENDED_BEGIN. Because with that we can redefine + * this one macro to cover all use cases. The macro is defined the same way it + * is defined in glib (with a very low probability of being changed thanks to a + * comment above it). + */ +# undef _G_DEFINE_TYPE_EXTENDED_BEGIN + +# define _G_DEFINE_TYPE_EXTENDED_BEGIN(TypeName, type_name, TYPE_PARENT, flags) \ + _Pragma("GCC diagnostic push") \ + _Pragma("GCC diagnostic ignored \"-Wincompatible-pointer-types-discards-qualifiers\"") \ + _G_DEFINE_TYPE_EXTENDED_BEGIN_PRE(TypeName, type_name, TYPE_PARENT) \ + _G_DEFINE_TYPE_EXTENDED_BEGIN_REGISTER(TypeName, type_name, TYPE_PARENT, flags) \ + _Pragma("GCC diagnostic pop") + +#endif /* __clang__ */ gchar * vir_g_canonicalize_filename(const gchar *filename, const gchar *relative_to); -- 2.29.2

2 1

[PATCH 1/1] qemu_tmp.c: fix 'shortName' leak
by Daniel Henrique Barboza 16 Nov '20

16 Nov '20

This is a Coverity fix pointed out by John in IRC. This code was introduced in 19d74fdf0eb, when the TPM Proxy device for for ppc64 was introduced. This will leak in case we have 2 TPMs in the same domain, a possible scenario with the protected Ultravisor execution in PowerPC guests. Fixes: 19d74fdf0eb5d2e89e80ceedea736425160ffccb Reported-by: John Ferlan <jferlan(a)redhat.com> Signed-off-by: Daniel Henrique Barboza <danielhb413(a)gmail.com> --- src/qemu/qemu_tpm.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu_tpm.c b/src/qemu/qemu_tpm.c index 8adb0e42b8..872be16570 100644 --- a/src/qemu/qemu_tpm.c +++ b/src/qemu/qemu_tpm.c @@ -841,10 +841,11 @@ qemuExtTPMStop(virQEMUDriverPtr driver, virDomainObjPtr vm) { g_autoptr(virQEMUDriverConfig) cfg = virQEMUDriverGetConfig(driver); - g_autofree char *shortName = NULL; size_t i; for (i = 0; i < vm->def->ntpms; i++) { + g_autofree char *shortName = NULL; + if (vm->def->tpms[i]->type != VIR_DOMAIN_TPM_TYPE_EMULATOR) continue; @@ -866,12 +867,13 @@ qemuExtTPMSetupCgroup(virQEMUDriverPtr driver, virCgroupPtr cgroup) { g_autoptr(virQEMUDriverConfig) cfg = virQEMUDriverGetConfig(driver); - g_autofree char *shortName = NULL; int rc; pid_t pid; size_t i; for (i = 0; i < def->ntpms; i++) { + g_autofree char *shortName = NULL; + if (def->tpms[i]->type != VIR_DOMAIN_TPM_TYPE_EMULATOR) continue; -- 2.26.2

2 1

[libvirt PATCH 0/4] Misc version updates
by Daniel P. Berrangé 16 Nov '20

16 Nov '20

In both yajl and wireshark we can bump our min version per our platform support matrix and thus eliminate some historical cruft. Daniel P. Berrangé (4): meson: assume pkg-config support for yajl meson: bump min wireshark to 2.6.0 rpm: remove version checks for wireshark rpm: enable wireshark build for RHEL-8 and later libvirt.spec.in | 8 +++---- meson.build | 63 ++++++++++++++----------------------------------- 2 files changed, 22 insertions(+), 49 deletions(-) -- 2.28.0

2 5

[libvirt PATCH] Do not disable incompatible-pointer-types-discards-qualifiers
by Martin Kletzander 16 Nov '20

16 Nov '20

This reverts commit b3710e9a2af402a2b620de570b062294e11190eb. That check is very valuable for our code, but it causes issue with glib >= 2.67.0 when building with clang. The reason is a combination of two commits in glib, firstly fdda405b6b1b which adds a g_atomic_pointer_{set,get} variants that enforce stricter type checking (by removing an extra cast) for compilers that support __typeof__, and commit dce24dc4492d which effectively enabled the new variant of glib's atomic code for clang. This will not be necessary when glib's issue #600 [0] (8 years old) is fixed. Thankfully, MR #1719 [1], which is supposed to deal with this issue was opened 3 weeks ago, so there is a slight sliver of hope. [0] https://gitlab.gnome.org/GNOME/glib/-/issues/600 [1] https://gitlab.gnome.org/GNOME/glib/-/merge_requests/1719 Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com> --- meson.build | 3 --- src/qemu/qemu_domain.c | 7 +++++++ src/util/vireventthread.c | 6 ++++++ src/util/viridentity.c | 6 ++++++ src/util/virobject.c | 6 ++++++ 5 files changed, 25 insertions(+), 3 deletions(-) diff --git a/meson.build b/meson.build index cecaad199d4c..04646e3a078c 100644 --- a/meson.build +++ b/meson.build @@ -405,9 +405,6 @@ cc_flags += [ # so use this Clang-specific arg to keep it quiet '-Wno-typedef-redefinition', - # Clang complains about casts in G_DEFINE_TYPE(...) - '-Wno-incompatible-pointer-types-discards-qualifiers', - # We don't use -Wc++-compat so we have to enable it explicitly '-Wjump-misses-init', diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index 2158080a56ad..3349476cceae 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -330,7 +330,14 @@ struct _qemuDomainLogContext { virLogManagerPtr manager; }; +#pragma clang diagnostic push +/* Workaround for glib's issue #600 on clang */ +#pragma clang diagnostic ignored "-Wincompatible-pointer-types-discards-qualifiers" + G_DEFINE_TYPE(qemuDomainLogContext, qemu_domain_log_context, G_TYPE_OBJECT); + +#pragma clang diagnostic pop + static virClassPtr qemuDomainSaveCookieClass; static void qemuDomainLogContextFinalize(GObject *obj); diff --git a/src/util/vireventthread.c b/src/util/vireventthread.c index 8342f420f666..4e202c94d0f0 100644 --- a/src/util/vireventthread.c +++ b/src/util/vireventthread.c @@ -32,8 +32,14 @@ struct _virEventThread { GMainLoop *loop; }; +#pragma clang diagnostic push +/* Workaround for glib's issue #600 on clang */ +#pragma clang diagnostic ignored "-Wincompatible-pointer-types-discards-qualifiers" + G_DEFINE_TYPE(virEventThread, vir_event_thread, G_TYPE_OBJECT) +#pragma clang diagnostic pop + #define VIR_FROM_THIS VIR_FROM_EVENT static void diff --git a/src/util/viridentity.c b/src/util/viridentity.c index 2cb9042a846a..43f447f629fe 100644 --- a/src/util/viridentity.c +++ b/src/util/viridentity.c @@ -50,8 +50,14 @@ struct _virIdentity { virTypedParameterPtr params; }; +#pragma clang diagnostic push +/* Workaround for glib's issue #600 on clang */ +#pragma clang diagnostic ignored "-Wincompatible-pointer-types-discards-qualifiers" + G_DEFINE_TYPE(virIdentity, vir_identity, G_TYPE_OBJECT) +#pragma clang diagnostic pop + static virThreadLocal virIdentityCurrent; static void virIdentityFinalize(GObject *obj); diff --git a/src/util/virobject.c b/src/util/virobject.c index a6305354c35e..8d1f26618334 100644 --- a/src/util/virobject.c +++ b/src/util/virobject.c @@ -53,8 +53,14 @@ struct _virObjectPrivate { }; +#pragma clang diagnostic push +/* Workaround for glib's issue #600 on clang */ +#pragma clang diagnostic ignored "-Wincompatible-pointer-types-discards-qualifiers" + G_DEFINE_TYPE_WITH_PRIVATE(virObject, vir_object, G_TYPE_OBJECT) +#pragma clang diagnostic pop + #define VIR_OBJECT_NOTVALID(obj) (!obj || !VIR_IS_OBJECT(obj)) #define VIR_OBJECT_USAGE_PRINT_WARNING(anyobj, objclass) \ -- 2.29.2

4 10

[libvirt PATCH] news: Fix vertical and horizontal spacing
by Andrea Bolognani 16 Nov '20

16 Nov '20

Signed-off-by: Andrea Bolognani <abologna(a)redhat.com> --- Pushed as trivial. NEWS.rst | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/NEWS.rst b/NEWS.rst index f4fc407b77..98303432f0 100644 --- a/NEWS.rst +++ b/NEWS.rst @@ -33,6 +33,7 @@ v6.10.0 (unreleased) This drops support for Windows Server 2008R2 and 2012. The earliest supported version is now Windows 2012R2. + v6.9.0 (2020-11-02) =================== @@ -165,17 +166,17 @@ v6.8.0 (2020-10-01) * qemu: Preserve qcow2 cluster size after external snapshots - The new overlay image which is installed on top of the current chain when - taking an external snapshot now preserves the cluser size of the original - top image to preserve any performance tuning done on the original image. + The new overlay image which is installed on top of the current chain when + taking an external snapshot now preserves the cluser size of the original + top image to preserve any performance tuning done on the original image. * **Bug fixes** * qemu: Various (i)SCSI backed hostdev fixes - (i)SCSI backed hostdevs now work again with an arbitrarily long - user-specified device alias and also honor the 'readonly' property after a - recent rewrite. + (i)SCSI backed hostdevs now work again with an arbitrarily long + user-specified device alias and also honor the 'readonly' property after a + recent rewrite. * **Removed features** -- 2.26.2

1 0

[PATCH] News: Several apparmor improvements at v6.7.0[v3]
by jgao 16 Nov '20

16 Nov '20

From: Jianan Gao <jgao(a)redhat.com> Add news about apparmor about the improvements. Signed-off-by: Jianan Gao <jgao(a)redhat.com> --- Clarified news content and revised commit information --- NEWS.rst | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/NEWS.rst b/NEWS.rst index 905deba5a8..fd3254d47f 100644 --- a/NEWS.rst +++ b/NEWS.rst @@ -224,6 +224,11 @@ v6.7.0 (2020-09-01) forbidden and no size auto-alignment will be made. Instead, libvirt will suggest an aligned round up size for the user. + * apparmor: Several improvements + + Add support for virtiofs filesystem and allow qemu load old shared objects after + upgrade. + * **Bug fixes** * virdevmapper: Deal with kernels without DM support -- 2.21.3

2 1

Issues 14 and 90 Contribution Details
by Dustan B Helm 15 Nov '20

15 Nov '20

Hey Libvirt team. We are looking into issue 14 <https://gitlab.com/libvirt/libvirt/-/issues/14> and issue 90 <https://gitlab.com/libvirt/libvirt/-/issues/90> to make possible contributions to one or both of them. However, before we start preparing patches for either of these issues, we have some clarifying questions about the scope of the necessary changes. Issue 14: Mostly we wanted to know what was wrong with Guannan's previous attempt at patching this. Was it incorrect, incomplete, or otherwise lacking? We also were confused on how Peter's usb-audio commit actually addressed an issue. It seems like most of the changes were documentation and constant and only a few changes changed significant amounts of code. We understand that QEMU is the program actually handling the sound card, but how do Peter's changes actually hook up the sound card with QEMU? Would the changes needed for the usb-net device be the same as Peter's usb-audio changes except in different sections of each file, or would the changes need to be made elsewhere? We saw that Peter and Guannan both had changes to tests/qemuhelptest.c and src/qemu/qemu_capabilities.c (to name a few). Issue 90:In order to implement a new network disk protocol, we believe there should be changes to the parameters in virsh/attach-disk, but we also know that there will likely be changes elsewhere in the codebase to allow support for a new protocol type. Ex: diskSourceNetworkProtocol definitions in docs/schemas/domaincommon.rng. Is there documentation available on what files/sections of the codebase handle supporting network disk protocols, so we can find more reference examples and begin to make changes in the appropriate locations? Thanks in advance, Dustan Helm, Ryan Gahagan, Barrett Schonefeld

1 0

[PATCH] News: Several apparmor improvements at v6.7.0
by jgao 14 Nov '20

14 Nov '20

From: jgao <jgao(a)redhat.com> Add news about apparmor about the improvements. Signed-off-by: jgao <jgao(a)redhat.com> --- NEWS.rst | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/NEWS.rst b/NEWS.rst index 905deba5a8..30a4f1065a 100644 --- a/NEWS.rst +++ b/NEWS.rst @@ -224,6 +224,11 @@ v6.7.0 (2020-09-01) forbidden and no size auto-alignment will be made. Instead, libvirt will suggest an aligned round up size for the user. + * apparmor: Several improvements + + Add support for virtiofs filesystem, allowing qemu load old shared objects after + upgrades. + * **Bug fixes** * virdevmapper: Deal with kernels without DM support -- 2.21.3

4 3

[PATCH] util: remove ATTRIBUTE_NONNULL from virDirClose declaration
by Laine Stump 13 Nov '20

13 Nov '20

Before commit 24d8968c, virDirClose took a DIR**, and that was never NULL, so its declaration included ATTRIBUTE_NONNULL(1). Since that commit, virDirClose takes a DIR*, and it may be NULL (e.g. if the DIR* is initialized to NULL and was never closed). Even though virDirClose() is currently only called implicitly (as the cleanup for a g_autoptr(DIR)), and (as I've just newly learned) the autocleanup function g_autoptr will only be called if the pointer in question is non-null (see the definition of _GLIB_AUTOPTR_CLEAR_FUNC_NAME in /usr/include/glib-2.0/glib/gmacros.h), it does still cause Coverity to complain that it *could* be called with a NULL, and it's also possible that in the future someone might add code that explicitly calls virDirClose. To eliminate the Coverity complaints, and protect against the hypothetical future where someone both explicitly calls virDirClose() with a potentially NULL value, *and* re-enables the nonnull directive when not building with Coverity (disabled by commit eefb881) this patch removes the ATTRIBUTE_NONNULL(1) from the declaration of virDirClose(). Fixes: 24d8968cd0a718af4badbbc858b1b449fea7205a Details-Research-by: Dan Berrange <berrange(a)redhat.com> Signed-off-by: Laine Stump <laine(a)redhat.com> --- src/util/virfile.h | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/src/util/virfile.h b/src/util/virfile.h index ba31a78e58..32505826ee 100644 --- a/src/util/virfile.h +++ b/src/util/virfile.h @@ -269,8 +269,7 @@ int virDirOpenQuiet(DIR **dirp, const char *dirname) ATTRIBUTE_NONNULL(1) ATTRIBUTE_NONNULL(2) G_GNUC_WARN_UNUSED_RESULT; int virDirRead(DIR *dirp, struct dirent **ent, const char *dirname) ATTRIBUTE_NONNULL(1) ATTRIBUTE_NONNULL(2) G_GNUC_WARN_UNUSED_RESULT; -void virDirClose(DIR *dirp) - ATTRIBUTE_NONNULL(1); +void virDirClose(DIR *dirp); G_DEFINE_AUTOPTR_CLEANUP_FUNC(DIR, virDirClose); int virFileMakePath(const char *path) G_GNUC_WARN_UNUSED_RESULT; -- 2.28.0

1 1

[PATCH 1/1] qemu_driver.c: do not redefine 'event' in qemuDomainDefineXMLFlags()
by Daniel Henrique Barboza 13 Nov '20

13 Nov '20

A bad merge while rebasing 74b2834333a caused the @event variable to be defined twice, inside the 'cleanup' label, causing coverity errors. This code was originally moved outside of the label by commit 773c7c43611a. Delete the unintended code in the 'cleanup' label. Fixes: 74b2834333ab3bf500f870e0a6d4e8309379d96a Reported-by: John Ferlan <jferlan(a)redhat.com> Signed-off-by: Daniel Henrique Barboza <danielhb413(a)gmail.com> --- src/qemu/qemu_driver.c | 9 --------- 1 file changed, 9 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 78ccf96df5..ac38edf009 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -6748,15 +6748,6 @@ qemuDomainDefineXMLFlags(virConnectPtr conn, } } - event = virDomainEventLifecycleNewFromObj(vm, - VIR_DOMAIN_EVENT_DEFINED, - !oldDef ? - VIR_DOMAIN_EVENT_DEFINED_ADDED : - VIR_DOMAIN_EVENT_DEFINED_UPDATED); - - VIR_INFO("Creating domain '%s'", vm->def->name); - dom = virGetDomain(conn, vm->def->name, vm->def->uuid, vm->def->id); - virDomainObjEndAPI(&vm); virObjectEventStateQueue(driver->domainEventState, event); return dom; -- 2.26.2

2 1