cherry-pick some bugfix patches from open source community

cherry-pick patchs list:
2ab8dba5    qemuDomainGetUnplugTimeout: Add G_GNUC_NO_INLINE 
423664a6    virNetDevSwitchdevFeature: Make failure to get 'family_id' non-fatal 
ca616274    virNetDevGetFamilyId: Change signature 
67b973b5    qemuDomainDefPostParse: Fail if unable to fill machine type 
67e19fc9    qemu: Revoke access to mirror on failed blockcopy 
93b15ba0    qemu: fix hang in p2p + xbzrle compression + parallel migration 
a13ac587    util: fix iteration in virSocketAddrResolveService 
88011ed2    libxl: fix crash when initializing driver

Signed-off-by: AlexChen <alex.chen@huawei.com>

libvirt.spec

@@ -99,7 +99,7 @@
 Summary: Library providing a simple virtualization API
 Name: libvirt
 Version: 6.2.0
-Release: 12
+Release: 13
 License: LGPLv2+
 URL: https://libvirt.org/
 
@@ -156,6 +156,14 @@ Patch0045: libvirt-leaseshelper-Report-more-errors.patch
 Patch0046: libvirt-qemuBuildMemoryBackendProps-Use-boolean-type-for-pme.patch
 Patch0047: libvirt-qemu-format-ramfb-attribute-for-mediated-devices.patch
 Patch0048: libvirt-resctrl-Do-not-open-directory-for-writing.patch
+Patch0049: libxl-fix-crash-when-initializing-driver.patch
+Patch0050: util-fix-iteration-in-virSocketAddrResolveService.patch
+Patch0051: qemu-fix-hang-in-p2p-xbzrle-compression-parallel-mig.patch
+Patch0052: qemu-Revoke-access-to-mirror-on-failed-blockcopy.patch
+Patch0053: qemuDomainDefPostParse-Fail-if-unable-to-fill-machin.patch
+Patch0054: virNetDevGetFamilyId-Change-signature.patch
+Patch0055: virNetDevSwitchdevFeature-Make-failure-to-get-family.patch
+Patch0056: qemuDomainGetUnplugTimeout-Add-G_GNUC_NO_INLINE.patch
 
 Requires: libvirt-daemon = %{version}-%{release}
 Requires: libvirt-daemon-config-network = %{version}-%{release}
@@ -1888,6 +1896,8 @@ exit 0
 
 
 %changelog
+* Tue Sep 22 2020 AlexChen <alex.chen@huawei.com> - 6.2.0-13
+- bugfix: cherry-pick some bugfix patches from opensource community
 * Tue Sep 22 2020 Hao Wang <wanghao232@huawei.com> - 6.2.0-12
 - backport upstream patches
 * Tue Sep 22 2020 Zeyu Jin <jinzeyu@huawei.com> - 6.2.0-11

libxl-fix-crash-when-initializing-driver.patch

@@ -0,0 +1,47 @@
+From 0e95f7b912055cf254b71b0b02dcb0acf7da3870 Mon Sep 17 00:00:00 2001
+From: Jim Fehlig <jfehlig@suse.com>
+Date: Fri, 3 Apr 2020 15:51:48 -0600
+Subject: [PATCH 1/8] libxl: fix crash when initializing driver
+MIME-Version: 1.0
+Content-Type: text/plain; charset=UTF-8
+Content-Transfer-Encoding: 8bit
+
+Commit 54a401af478 split out DriverConfigInit from DriverConfigNew, but
+then called it a bit late from libxlStateInitialize. The cfg is used in
+libxlDriverConfigLoadFile and when uninitialized results in a crash.
+Calling DriverConfigInit immediately after DriverConfigNew fixes the
+crash.
+
+Signed-off-by: Jim Fehlig <jfehlig@suse.com>
+Reviewed-by: Erik Skultety <eskultet@redhat.com>
+Reviewed-by: Ján Tomko <jtomko@redhat.com>
+Signed-off-by: AlexChen <alex.chen@huawei.com>
+---
+ src/libxl/libxl_driver.c | 6 +++---
+ 1 file changed, 3 insertions(+), 3 deletions(-)
+
+diff --git a/src/libxl/libxl_driver.c b/src/libxl/libxl_driver.c
+index 7ec4fcc3d1..980984b199 100644
+--- a/src/libxl/libxl_driver.c
++++ b/src/libxl/libxl_driver.c
+@@ -702,14 +702,14 @@ libxlStateInitialize(bool privileged,
+     if (!(cfg = libxlDriverConfigNew()))
+         goto error;
+ 
++    if (libxlDriverConfigInit(cfg) < 0)
++        goto error;
++
+     driverConf = g_strdup_printf("%s/libxl.conf", cfg->configBaseDir);
+ 
+     if (libxlDriverConfigLoadFile(cfg, driverConf) < 0)
+         goto error;
+ 
+-    if (libxlDriverConfigInit(cfg) < 0)
+-        goto error;
+-
+     /* Register the callbacks providing access to libvirt's event loop */
+     libxl_osevent_register_hooks(cfg->ctx, &libxl_osevent_callbacks, cfg->ctx);
+ 
+-- 
+2.23.0
+

qemu-Revoke-access-to-mirror-on-failed-blockcopy.patch

@@ -0,0 +1,71 @@
+From 19d06bcb3c64662d5c131a035146d2caa8c5a437 Mon Sep 17 00:00:00 2001
+From: Michal Privoznik <mprivozn@redhat.com>
+Date: Tue, 14 Apr 2020 11:18:02 +0200
+Subject: [PATCH 4/8] qemu: Revoke access to mirror on failed blockcopy
+
+When preparing to do a blockcopy, the mirror image is modified so
+that QEMU can access it. For instance, the mirror has seclabels
+set, if it is a NVMe disk it is detached from the host and so on.
+And usually, the restore is done upon successful finish of the
+blockcopy operation. But, if something fails then we need to
+explicitly revoke the access to the mirror image (and thus
+reattach NVMe disk back to the host).
+
+Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1822538
+
+Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
+Reviewed-by: Pavel Mores <pmores@redhat.com>
+Signed-off-by: AlexChen <alex.chen@huawei.com>
+---
+ src/qemu/qemu_driver.c | 21 +++++++++++++--------
+ 1 file changed, 13 insertions(+), 8 deletions(-)
+
+diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c
+index 8bc5368b2f..7162875492 100644
+--- a/src/qemu/qemu_driver.c
++++ b/src/qemu/qemu_driver.c
+@@ -17972,6 +17972,7 @@ qemuDomainBlockCopyCommon(virDomainObjPtr vm,
+     virDomainDiskDefPtr disk = NULL;
+     int ret = -1;
+     bool need_unlink = false;
++    bool need_revoke = false;
+     g_autoptr(virQEMUDriverConfig) cfg = virQEMUDriverGetConfig(driver);
+     const char *format = NULL;
+     bool mirror_reuse = !!(flags & VIR_DOMAIN_BLOCK_COPY_REUSE_EXT);
+@@ -18146,6 +18147,7 @@ qemuDomainBlockCopyCommon(virDomainObjPtr vm,
+ 
+     if (qemuDomainStorageSourceChainAccessAllow(driver, vm, mirror) < 0)
+         goto endjob;
++    need_revoke = true;
+ 
+     if (blockdev) {
+         if (mirror_reuse) {
+@@ -18254,14 +18256,17 @@ qemuDomainBlockCopyCommon(virDomainObjPtr vm,
+ 
+  endjob:
+     if (ret < 0 &&
+-        virDomainObjIsActive(vm) &&
+-        (data || crdata)) {
+-        qemuDomainObjEnterMonitor(driver, vm);
+-        if (data)
+-            qemuBlockStorageSourceChainDetach(priv->mon, data);
+-        if (crdata)
+-            qemuBlockStorageSourceAttachRollback(priv->mon, crdata->srcdata[0]);
+-        ignore_value(qemuDomainObjExitMonitor(driver, vm));
++        virDomainObjIsActive(vm)) {
++        if (data || crdata) {
++            qemuDomainObjEnterMonitor(driver, vm);
++            if (data)
++                qemuBlockStorageSourceChainDetach(priv->mon, data);
++            if (crdata)
++                qemuBlockStorageSourceAttachRollback(priv->mon, crdata->srcdata[0]);
++            ignore_value(qemuDomainObjExitMonitor(driver, vm));
++        }
++        if (need_revoke)
++            qemuDomainStorageSourceChainAccessRevoke(driver, vm, mirror);
+     }
+     if (need_unlink && virStorageFileUnlink(mirror) < 0)
+         VIR_WARN("%s", _("unable to remove just-created copy target"));
+-- 
+2.23.0
+

qemu-fix-hang-in-p2p-xbzrle-compression-parallel-mig.patch

@@ -0,0 +1,47 @@
+From f15935ea3b510bf0633dc2219a132412a7a8e96c Mon Sep 17 00:00:00 2001
+From: Lin Ma <lma@suse.com>
+Date: Thu, 16 Apr 2020 12:44:51 +0800
+Subject: [PATCH 3/8] qemu: fix hang in p2p + xbzrle compression + parallel
+ migration
+
+When we do parallel migration, The multifd-channels migration parameter
+needs to be set on the destination side as well before incoming migration
+URI, unless we accept the default number of connections(2).
+
+Usually, This can be correctly handled by libvirtd. But in this case if
+we use p2p + xbzrle compression without parameter '--comp-xbzrle-cache',
+qemuMigrationParamsDump returns too early, The corresponding migration
+parameter will not be set on the destination side, It results QEMU hangs.
+
+Reproducer:
+virsh migrate --live --p2p --comp-methods xbzrle \
+--parallel --parallel-connections 3 GUEST qemu+ssh://dsthost/system
+
+or
+
+virsh migrate --live --p2p --compressed \
+--parallel --parallel-connections 3 GUEST qemu+ssh://dsthost/system
+
+Signed-off-by: Lin Ma <lma@suse.com>
+Message-Id: <20200416044451.21134-1-lma@suse.com>
+Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
+Signed-off-by: AlexChen <alex.chen@huawei.com>
+---
+ src/qemu/qemu_migration_params.c | 1 -
+ 1 file changed, 1 deletion(-)
+
+diff --git a/src/qemu/qemu_migration_params.c b/src/qemu/qemu_migration_params.c
+index dd5e2ce1b6..810199370f 100644
+--- a/src/qemu/qemu_migration_params.c
++++ b/src/qemu/qemu_migration_params.c
+@@ -630,7 +630,6 @@ qemuMigrationParamsDump(qemuMigrationParamsPtr migParams,
+     if (migParams->compMethods == 1ULL << QEMU_MIGRATION_COMPRESS_XBZRLE &&
+         !migParams->params[QEMU_MIGRATION_PARAM_XBZRLE_CACHE_SIZE].set) {
+         *flags |= VIR_MIGRATE_COMPRESSED;
+-        return 0;
+     }
+ 
+     for (i = 0; i < QEMU_MIGRATION_COMPRESS_LAST; ++i) {
+-- 
+2.23.0
+

qemuDomainDefPostParse-Fail-if-unable-to-fill-machin.patch

@@ -0,0 +1,46 @@
+From d9190851ed9a28e86d5baa3a9872eb6246270cd1 Mon Sep 17 00:00:00 2001
+From: Michal Privoznik <mprivozn@redhat.com>
+Date: Thu, 16 Apr 2020 14:18:28 +0200
+Subject: [PATCH 5/8] qemuDomainDefPostParse: Fail if unable to fill machine
+ type
+
+Previously, we used virCapabilitiesDomainDataLookup() to fill
+machine type in post parse callback if none was provided in the
+domain XML. If machine type couldn't be filled in an error was
+reported. After 4a4132b4625 we've changed it to
+virQEMUCapsGetPreferredMachine() which returns NULL, but we no
+longer report an error and proceed with the post parse callbacks
+processing. This may lead to a crash because the code later on
+assumes def->os.machine is not NULL.
+
+Fixes: 4a4132b4625
+
+Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
+Reviewed-by: Pavel Mores <pmores@redhat.com>
+Signed-off-by: AlexChen <alex.chen@huawei.com>
+---
+ src/qemu/qemu_domain.c | 8 ++++++++
+ 1 file changed, 8 insertions(+)
+
+diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c
+index b1a46478ed..36e6343995 100644
+--- a/src/qemu/qemu_domain.c
++++ b/src/qemu/qemu_domain.c
+@@ -4971,6 +4971,14 @@ qemuDomainDefPostParse(virDomainDefPtr def,
+     if (!def->os.machine) {
+         const char *machine = virQEMUCapsGetPreferredMachine(qemuCaps,
+                                                              def->virtType);
++        if (!machine) {
++            virReportError(VIR_ERR_INVALID_ARG,
++                           _("could not get preferred machine for %s type=%s"),
++                           def->emulator,
++                           virDomainVirtTypeToString(def->virtType));
++            return -1;
++        }
++
+         def->os.machine = g_strdup(machine);
+     }
+ 
+-- 
+2.23.0
+

qemuDomainGetUnplugTimeout-Add-G_GNUC_NO_INLINE.patch

@@ -0,0 +1,32 @@
+From 9494102bc1560c1870c8ba603e936fa81beaa28c Mon Sep 17 00:00:00 2001
+From: Peter Krempa <pkrempa@redhat.com>
+Date: Thu, 23 Apr 2020 10:30:13 +0200
+Subject: [PATCH 8/8] qemuDomainGetUnplugTimeout: Add G_GNUC_NO_INLINE
+
+The function is mocked in qemuhotplugmock.so. Recent clang versions
+decided to inline it so the mock stopped working resulting in
+qemuhotplugtest wasting 15 seconds waiting for timeouts.
+
+Signed-off-by: Peter Krempa <pkrempa@redhat.com>
+Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
+Signed-off-by: AlexChen <alex.chen@huawei.com>
+---
+ src/qemu/qemu_hotplug.c | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c
+index 14654a17d7..60d0729f1e 100644
+--- a/src/qemu/qemu_hotplug.c
++++ b/src/qemu/qemu_hotplug.c
+@@ -5122,7 +5122,7 @@ qemuDomainResetDeviceRemoval(virDomainObjPtr vm)
+ }
+ 
+ 
+-unsigned long long
++unsigned long long G_GNUC_NO_INLINE
+ qemuDomainGetUnplugTimeout(virDomainObjPtr vm)
+ {
+     if (qemuDomainIsPSeries(vm->def))
+-- 
+2.23.0
+

util-fix-iteration-in-virSocketAddrResolveService.patch

@@ -0,0 +1,30 @@
+From 9056229f9dff923406d97f0ac295bd502eaa5265 Mon Sep 17 00:00:00 2001
+From: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
+Date: Mon, 13 Apr 2020 16:48:43 +0300
+Subject: [PATCH 2/8] util: fix iteration in virSocketAddrResolveService
+
+getaddrinfo returns linked list. Fix iteration accordingly.
+
+Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
+Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
+Signed-off-by: AlexChen <alex.chen@huawei.com>
+---
+ src/util/virsocketaddr.c | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/src/util/virsocketaddr.c b/src/util/virsocketaddr.c
+index 4cad329d15..4c9f124e88 100644
+--- a/src/util/virsocketaddr.c
++++ b/src/util/virsocketaddr.c
+@@ -271,7 +271,7 @@ int virSocketAddrResolveService(const char *service)
+             port = ntohs(in.sin6_port);
+             goto cleanup;
+         }
+-        tmp++;
++        tmp = tmp->ai_next;
+     }
+ 
+     virReportError(VIR_ERR_SYSTEM_ERROR,
+-- 
+2.23.0
+

virNetDevGetFamilyId-Change-signature.patch

@@ -0,0 +1,90 @@
+From 56635fe00f44f8c83ba059432d42c093be4e9ed8 Mon Sep 17 00:00:00 2001
+From: Michal Privoznik <mprivozn@redhat.com>
+Date: Sun, 19 Apr 2020 08:26:04 +0200
+Subject: [PATCH 6/8] virNetDevGetFamilyId: Change signature
+MIME-Version: 1.0
+Content-Type: text/plain; charset=UTF-8
+Content-Transfer-Encoding: 8bit
+
+Introduced in v3.8.0-rc1~96, the virNetDevGetFamilyId() gets
+netlink family ID for passed family name (even though it's used
+only for getting "devlink" ID). Nevertheless, the function
+returns 0 on an error or if no family ID was found. This makes it
+harder for a caller to distinguish these two. Change the retval
+so that a negative value is returned upon error, zero is no ID
+found (but no error encountered) and a positive value is returned
+on successful translation.
+
+Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
+Reviewed-by: Ján Tomko <jtomko@redhat.com>
+Signed-off-by: AlexChen <alex.chen@huawei.com>
+---
+ src/util/virnetdev.c | 23 +++++++++++++++--------
+ 1 file changed, 15 insertions(+), 8 deletions(-)
+
+diff --git a/src/util/virnetdev.c b/src/util/virnetdev.c
+index b465bdac2e..3431aaf6a9 100644
+--- a/src/util/virnetdev.c
++++ b/src/util/virnetdev.c
+@@ -3057,11 +3057,15 @@ virNetDevGetEthtoolFeatures(virBitmapPtr bitmap,
+  * This function supplies the devlink family id
+  *
+  * @family_name: the name of the family to query
++ * @family_id: family ID
+  *
+- * Returns family id or 0 on failure.
++ * Returns: 0 if no family was found,
++ *          1 if family was found (@family_id is set),
++ *         -1 otherwise
+  */
+-static uint32_t
+-virNetDevGetFamilyId(const char *family_name)
++static int
++virNetDevGetFamilyId(const char *family_name,
++                     uint32_t *family_id)
+ {
+     struct nl_msg *nl_msg = NULL;
+     struct nlmsghdr *resp = NULL;
+@@ -3072,7 +3076,7 @@ virNetDevGetFamilyId(const char *family_name)
+     };
+     struct nlattr *tb[CTRL_ATTR_MAX + 1] = {NULL, };
+     unsigned int recvbuflen;
+-    uint32_t family_id = 0;
++    int ret = -1;
+ 
+     if (!(nl_msg = nlmsg_alloc_simple(GENL_ID_CTRL,
+                                       NLM_F_REQUEST | NLM_F_ACK))) {
+@@ -3098,15 +3102,18 @@ virNetDevGetFamilyId(const char *family_name)
+         goto cleanup;
+     }
+ 
+-    if (tb[CTRL_ATTR_FAMILY_ID] == NULL)
++    if (tb[CTRL_ATTR_FAMILY_ID] == NULL) {
++        ret = 0;
+         goto cleanup;
++    }
+ 
+-    family_id = *(uint32_t *)RTA_DATA(tb[CTRL_ATTR_FAMILY_ID]);
++    *family_id = *(uint32_t *)RTA_DATA(tb[CTRL_ATTR_FAMILY_ID]);
++    ret = 1;
+ 
+  cleanup:
+     nlmsg_free(nl_msg);
+     VIR_FREE(resp);
+-    return family_id;
++    return ret;
+ }
+ 
+ 
+@@ -3140,7 +3147,7 @@ virNetDevSwitchdevFeature(const char *ifname,
+     int ret = -1;
+     uint32_t family_id;
+ 
+-    if ((family_id = virNetDevGetFamilyId(DEVLINK_GENL_NAME)) <= 0)
++    if (virNetDevGetFamilyId(DEVLINK_GENL_NAME, &family_id) <= 0)
+         return ret;
+ 
+     if ((is_vf = virNetDevIsVirtualFunction(ifname)) < 0)
+-- 
+2.23.0
+

virNetDevSwitchdevFeature-Make-failure-to-get-family.patch

@@ -0,0 +1,57 @@
+From 7581a0bbd552d63d654a88f0ad2a68e99fa8b54a Mon Sep 17 00:00:00 2001
+From: Michal Privoznik <mprivozn@redhat.com>
+Date: Sun, 19 Apr 2020 07:25:34 +0200
+Subject: [PATCH 7/8] virNetDevSwitchdevFeature: Make failure to get
+ 'family_id' non-fatal
+MIME-Version: 1.0
+Content-Type: text/plain; charset=UTF-8
+Content-Transfer-Encoding: 8bit
+
+I've just got a new machine and I'm still converging on the
+kernel config. Anyway, since I don't have enabled any of SRIO-V
+drivers, my kernel doesn't have NET_DEVLINK enabled (i.e.
+virNetDevGetFamilyId() returns 0). But this makes nodedev driver
+ignore all interfaces, because when enumerating all devices via
+udev, the control reaches virNetDevSwitchdevFeature() eventually
+and subsequently virNetDevGetFamilyId() which 'fails'. Well, it's
+not really a failure - the virNetDevSwitchdevFeature() stub
+simply returns 0.
+
+Also, move the call a few lines below, just around the place
+where it's needed.
+
+Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
+Reviewed-by: Ján Tomko <jtomko@redhat.com>
+Signed-off-by: AlexChen <alex.chen@huawei.com>
+---
+ src/util/virnetdev.c | 7 ++++---
+ 1 file changed, 4 insertions(+), 3 deletions(-)
+
+diff --git a/src/util/virnetdev.c b/src/util/virnetdev.c
+index 3431aaf6a9..ff86aa1fc9 100644
+--- a/src/util/virnetdev.c
++++ b/src/util/virnetdev.c
+@@ -3146,9 +3146,7 @@ virNetDevSwitchdevFeature(const char *ifname,
+     int is_vf = -1;
+     int ret = -1;
+     uint32_t family_id;
+-
+-    if (virNetDevGetFamilyId(DEVLINK_GENL_NAME, &family_id) <= 0)
+-        return ret;
++    int rv;
+ 
+     if ((is_vf = virNetDevIsVirtualFunction(ifname)) < 0)
+         return ret;
+@@ -3168,6 +3166,9 @@ virNetDevSwitchdevFeature(const char *ifname,
+         goto cleanup;
+     }
+ 
++    if ((rv = virNetDevGetFamilyId(DEVLINK_GENL_NAME, &family_id)) <= 0)
++        return rv;
++
+     if (!(nl_msg = nlmsg_alloc_simple(family_id,
+                                       NLM_F_REQUEST | NLM_F_ACK))) {
+         virReportOOMError();
+-- 
+2.23.0
+