- Devel - libvirt List Archives

[libvirt] [PULL 0/4] Python queue, 2019-07-01
by Eduardo Habkost 04 Jul '19

04 Jul '19

The following changes since commit 7d0e02405fc02a181319b1ab8681d2f72246b7c6: Merge remote-tracking branch 'remotes/vivier2/tags/trivial-patches-pull-request' into staging (2019-07-01 17:40:32 +0100) are available in the Git repository at: git://github.com/ehabkost/qemu.git tags/python-next-pull-request for you to fetch changes up to e5abf59eae5990296c243202f95f801495c52e76: Deprecate Python 2 support (2019-07-01 19:02:10 -0300) ---------------------------------------------------------------- Python queue, 2019-07-01 * Deprecate Python 2 support (Eduardo Habkost) * qemu/__init__.py refactor (John Snow) * make qmp-shell work with python3 (Igor Mammedov) ---------------------------------------------------------------- Eduardo Habkost (1): Deprecate Python 2 support Igor Mammedov (1): qmp: make qmp-shell work with python3 John Snow (2): python/qemu: split QEMUMachine out from underneath __init__.py machine.py: minor delinting configure | 8 + scripts/render_block_graph.py | 2 +- python/qemu/__init__.py | 502 +------------------- python/qemu/machine.py | 531 ++++++++++++++++++++++ python/qemu/qtest.py | 2 +- qemu-deprecated.texi | 8 + scripts/device-crash-test | 2 +- scripts/qmp/qmp-shell | 5 +- tests/acceptance/avocado_qemu/__init__.py | 2 +- tests/acceptance/virtio_version.py | 2 +- tests/migration/guestperf/engine.py | 24 +- tests/qemu-iotests/235 | 2 +- tests/vm/basevm.py | 3 +- 13 files changed, 572 insertions(+), 521 deletions(-) create mode 100644 python/qemu/machine.py -- 2.18.0.rc1.1.g3f1ff2140

2 5

[libvirt] [RFC] test_driver: Thoughts on the implementation of some FS-related APIs
by Ilias Stamatis 04 Jul '19

04 Jul '19

Hello, I was thinking about how to implement the following APIs in the test driver: -virDomainFSFreeze -virDomainFSThaw -virDomainFSTrim The first two are conceptually paired. They both get a mountpoints argument. The QEMU driver (which is the only driver implementing this currently) ignores this completely. However, I thought that we can allow it and pretend that the only available mountpoints are "/" and "/boot" (like in a default Fedora installation). The following questions arise though: - Should we keep some kind of temporary or permanent state on the test driver about this? Because it makes sense for virDomainFSThaw to succeed only on mountpoints for which previously virDomainFSFreeze has been called. If so, *where* exactly should we keep this state? - In case a non-existing mountpoint is passed to either of the first 2 APIs. Should we fail? If so, a new error code should be introduced for this. Also for the cases above where e.g. Thaw is called on a fs that isn't frozen probably yet another different error code must be used. Should we add all these new error codes? Or maybe all the above is more complicated than necessary and it's fine if we ignore mountpoints completely like the QEMU driver does and make all 3 APIs above to always succeed no matter what in the test driver? Thanks, Ilias

2 3

[libvirt] [PATCH] remote: mention libssh in error message
by Pino Toscano 04 Jul '19

04 Jul '19

Mention libssh as possible transport in the error message of an unrecognized transport. https://bugzilla.redhat.com/show_bug.cgi?id=1727013 Signed-off-by: Pino Toscano <ptoscano(a)redhat.com> --- src/remote/remote_driver.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/remote/remote_driver.c b/src/remote/remote_driver.c index df58b23c8c..97d9d5fa88 100644 --- a/src/remote/remote_driver.c +++ b/src/remote/remote_driver.c @@ -825,7 +825,7 @@ doRemoteOpen(virConnectPtr conn, } else { virReportError(VIR_ERR_INVALID_ARG, "%s", _("remote_open: transport in URL not recognised " - "(should be tls|unix|ssh|ext|tcp|libssh2)")); + "(should be tls|unix|ssh|ext|tcp|libssh2|libssh)")); return VIR_DRV_OPEN_ERROR; } } -- 2.21.0

2 1

[libvirt] [PATCH] Revert "lxc: Try harder to stop/reboot containers"
by Daniel P. Berrangé 04 Jul '19

04 Jul '19

This reverts commit 14b6a1854fb4c02c5fb2f51679f8ff099f28f53c. If virLXCDomainSetRunlevel returns -1 this indicates a serious error / failure that must be propagated to the caller. We must not carry on with other shutdown methods in this case. If virLXCDomainSetRunlevel return 0, this indicates that no initctl was found and it is thus reasonable to fallback to sending SIGTERM. The commit being reverted is broken because it would fallback to SIGTERM when virLXCDomainSetRunlevel returns -1, and would not fallback when virLXCDomainSetRunlevel returns 0. ie it did the exact opposite of what was required. Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- src/lxc/lxc_driver.c | 40 ++++++++++++++++++++++------------------ 1 file changed, 22 insertions(+), 18 deletions(-) diff --git a/src/lxc/lxc_driver.c b/src/lxc/lxc_driver.c index 9db2a02dee..01af5f6db8 100644 --- a/src/lxc/lxc_driver.c +++ b/src/lxc/lxc_driver.c @@ -3262,7 +3262,7 @@ lxcDomainShutdownFlags(virDomainPtr dom, virLXCDomainObjPrivatePtr priv; virDomainObjPtr vm; int ret = -1; - int rc = -1; + int rc; virCheckFlags(VIR_DOMAIN_SHUTDOWN_INITCTL | VIR_DOMAIN_SHUTDOWN_SIGNAL, -1); @@ -3291,17 +3291,19 @@ lxcDomainShutdownFlags(virDomainPtr dom, (flags & VIR_DOMAIN_SHUTDOWN_INITCTL)) { int command = VIR_INITCTL_RUNLEVEL_POWEROFF; - if ((rc = virLXCDomainSetRunlevel(vm, command)) < 0) { - if (flags != 0 && - (flags & VIR_DOMAIN_SHUTDOWN_INITCTL)) { - virReportError(VIR_ERR_OPERATION_UNSUPPORTED, "%s", - _("Container does not provide an initctl pipe")); - goto endjob; - } + if ((rc = virLXCDomainSetRunlevel(vm, command)) < 0) + goto endjob; + if (rc == 0 && flags != 0 && + ((flags & ~VIR_DOMAIN_SHUTDOWN_INITCTL) == 0)) { + virReportError(VIR_ERR_OPERATION_UNSUPPORTED, "%s", + _("Container does not provide an initctl pipe")); + goto endjob; } + } else { + rc = 0; } - if (rc < 0 && + if (rc == 0 && (flags == 0 || (flags & VIR_DOMAIN_SHUTDOWN_SIGNAL))) { if (kill(priv->initpid, SIGTERM) < 0 && @@ -3338,7 +3340,7 @@ lxcDomainReboot(virDomainPtr dom, virLXCDomainObjPrivatePtr priv; virDomainObjPtr vm; int ret = -1; - int rc = -1; + int rc; virCheckFlags(VIR_DOMAIN_REBOOT_INITCTL | VIR_DOMAIN_REBOOT_SIGNAL, -1); @@ -3367,17 +3369,19 @@ lxcDomainReboot(virDomainPtr dom, (flags & VIR_DOMAIN_REBOOT_INITCTL)) { int command = VIR_INITCTL_RUNLEVEL_REBOOT; - if ((rc = virLXCDomainSetRunlevel(vm, command)) < 0) { - if (flags != 0 && - (flags & VIR_DOMAIN_REBOOT_INITCTL)) { - virReportError(VIR_ERR_OPERATION_UNSUPPORTED, "%s", - _("Container does not provide an initctl pipe")); - goto endjob; - } + if ((rc = virLXCDomainSetRunlevel(vm, command)) < 0) + goto endjob; + if (rc == 0 && flags != 0 && + ((flags & ~VIR_DOMAIN_SHUTDOWN_INITCTL) == 0)) { + virReportError(VIR_ERR_OPERATION_UNSUPPORTED, "%s", + _("Container does not provide an initctl pipe")); + goto endjob; } + } else { + rc = 0; } - if (rc < 0 && + if (rc == 0 && (flags == 0 || (flags & VIR_DOMAIN_REBOOT_SIGNAL))) { if (kill(priv->initpid, SIGHUP) < 0 && -- 2.21.0

1 0

[libvirt] [PATCH] lxc: Fix reboot via initctl
by Lubomir Rintel 04 Jul '19

04 Jul '19

virInitctlSetRunLevel() returns 0 only if ended up doing nothing, 1 if it actually succeeded. Let's check for the error condition. Without this, a successful reboot would be treated as a failure and the LXC driver will proceed sending a TERM signal instead, effectively cancelling the shutdown. Signed-off-by: Lubomir Rintel <lkundrak(a)v3.sk> --- src/lxc/lxc_domain.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/src/lxc/lxc_domain.c b/src/lxc/lxc_domain.c index 51a9fd36eb..1dc2d0d4cf 100644 --- a/src/lxc/lxc_domain.c +++ b/src/lxc/lxc_domain.c @@ -457,7 +457,9 @@ lxcDomainInitctlCallback(pid_t pid ATTRIBUTE_UNUSED, data->st[i].st_ino == cont_sb.st_ino) continue; - return virInitctlSetRunLevel(fifo, data->runlevel); + if (virInitctlSetRunLevel(fifo, data->runlevel) == -1) + return -1; + return 0; } /* If no usable fifo was found then declare success. Caller -- 2.21.0

2 1

[libvirt] UDP broadcasts vs. nat Masquerading issue
by Nikolai Zhubr 04 Jul '19

04 Jul '19

Hi all, I'm observing an issue that as soon as libvirt starts, UPD broadcasts flowing through physical network (and intended for other services, unrelated to any virtualization) get broken. Specifically, windows neighbourhood browsing through samba's nmbd starts suffering badly (Samba is running on this same box). At the moment I'm running a quite outdated version 1.2.9 of libvirt, but because other than this issue it does its job pretty well I'd first consider some patching/backporting rather than totally replacing it with a new one. Anyway, I first need to better understand what is going on and what is wrong with it. This could also be related somewhat to https://www.redhat.com/archives/libvir-list/2013-September/msg01311.html but I suppose it is not exactly that thing, and besides, my version does already include a fix for the broadcasts issue mentioned in the msg01311. I've already figured the source of trouble is anyway related to these rules added: -A POSTROUTING -o br0 -j MASQUERADE -A POSTROUTING -o enp0s25 -j MASQUERADE -A POSTROUTING -o virbr2_nic -j MASQUERADE -A POSTROUTING -o vnet0 -j MASQUERADE Here, virbr2_nic and vnet0 are used by libvirt for arranging NAT-mode network configurations for VMs, unrelated to normal network stuff, so it looks ok. However, br0 (with enp0s25 in it) is a main interface of this host with primary ip address. And enp0s25 is a physical nic of this host, and it is used for all sorts of regular (unrelated to virtualization) communications as well. Also, br0 is used for attaching some bridge-mode (as opposed to NAT-mode) VMs managed by libvirt, but bridge mode is not supposed to employ address translation anyway. So, clearly, libvirt somehow chooses to set up masquerading for literally all existing network interfaces here (except lo), but I can't see a reason for the first two rules in the list above. Furthermore, they corrupt UDP broadcats coming from outside and reaching this host (through enp0s25/br0) such that source address gets replaced by this hosts primary address (as per masquerading). I've verified this by arranging a hand-crafted UDP listener and printing the respective source addresses as seen by normal userspace. Obviously Samba server can not work correctly under such conditions. Now I've discovered that I can "eliminate" the problem by e.g.: 1. Removing "-A POSTROUTING -o br0 -j MASQUERADE" (manual brute force) 2. Inserting "-A POSTROUTING -s 192.168.0.0/24 -d 192.168.0.255/32 -j ACCEPT" (Of course correcting rules by hand is not a solution, just a test) So question is, how the correct rules should ideally look like? And, is this issue known/fixed in most current libvirt? Thank you, Regards, Nikolai

2 1

[libvirt] [PATCH] remote: remove the only use of LIBVIRTD_CONFIGURATION_FILE
by Fabiano Fidêncio 04 Jul '19

04 Jul '19

86fbce56f27e removed the constant, but didn't actually adjust the only place where the constant was used. Signed-off-by: Fabiano Fidêncio <fidencio(a)redhat.com> --- This patch fixes a build breakage. It's either this, or reverting 86fbce56f27e. --- src/remote/remote_daemon.c | 4 ---- 1 file changed, 4 deletions(-) diff --git a/src/remote/remote_daemon.c b/src/remote/remote_daemon.c index fdc9e4333a..3cd7b16163 100644 --- a/src/remote/remote_daemon.c +++ b/src/remote/remote_daemon.c @@ -899,9 +899,6 @@ daemonUsage(const char *argv0, bool privileged) _("\n" " Default paths:\n" "\n" - " Configuration file (unless overridden by -f):\n" - " %s\n" - "\n" " Sockets:\n" " %s\n" " %s\n" @@ -914,7 +911,6 @@ daemonUsage(const char *argv0, bool privileged) " PID file (unless overridden by -p):\n" " %s/run/libvirtd.pid\n" "\n"), - LIBVIRTD_CONFIGURATION_FILE, LIBVIRTD_PRIV_UNIX_SOCKET, LIBVIRTD_PRIV_UNIX_SOCKET_RO, LIBVIRT_CACERT, -- 2.21.0

5 5

[libvirt] [PATCH] Revert "remote: remove unused constant for libvirtd config file"
by Peter Krempa 04 Jul '19

04 Jul '19

The constant is not unused and breaks the build of the remote driver. This reverts commit 86fbce56f27e06b34d63879c9f634a58a1e0a04a. Signed-off-by: Peter Krempa <pkrempa(a)redhat.com> --- Pushed under the build-breaker rule. src/remote/remote_driver.h | 1 + 1 file changed, 1 insertion(+) diff --git a/src/remote/remote_driver.h b/src/remote/remote_driver.h index 132e478ef3..8c7da6b000 100644 --- a/src/remote/remote_driver.h +++ b/src/remote/remote_driver.h @@ -34,6 +34,7 @@ unsigned long remoteVersion(void); #define LIBVIRTD_PRIV_UNIX_SOCKET LOCALSTATEDIR "/run/libvirt/libvirt-sock" #define LIBVIRTD_PRIV_UNIX_SOCKET_RO LOCALSTATEDIR "/run/libvirt/libvirt-sock-ro" #define LIBVIRTD_USER_UNIX_SOCKET "libvirt-sock" +#define LIBVIRTD_CONFIGURATION_FILE SYSCONFDIR "/libvirt/libvirtd.conf" /* Defaults for PKI directory. */ #define LIBVIRT_PKI_DIR SYSCONFDIR "/pki" -- 2.21.0

2 1

[libvirt] [PATCH python 0/2] spec: fedora tweaks
by Cole Robinson 03 Jul '19

03 Jul '19

Some fedora version tweaks Cole Robinson (2): spec: Unconditionally build python2 on Fedora spec: Bump minimum supported Fedora version to 29 libvirt-python.spec.in | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) -- 2.21.0

2 3

[libvirt] [PATCH] remote: remove unused constant for libvirtd config file
by Daniel P. Berrangé 03 Jul '19

03 Jul '19

The LIBVIRTD_CONFIGURATION_FILE constant was introduced in commit b7c42619e69acd8416be88b8aa3b4161bc813f8a Author: Richard W.M. Jones <rjones(a)redhat.com> Date: Mon Jun 11 11:43:41 2007 +0000 Mon Jun 11 12:41:00 BST 2007 Richard W.M. Jones <rjones(a)redhat.com> and then never used ! Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- Pushed under the trivial rule src/remote/remote_driver.h | 1 - 1 file changed, 1 deletion(-) diff --git a/src/remote/remote_driver.h b/src/remote/remote_driver.h index 8c7da6b000..132e478ef3 100644 --- a/src/remote/remote_driver.h +++ b/src/remote/remote_driver.h @@ -34,7 +34,6 @@ unsigned long remoteVersion(void); #define LIBVIRTD_PRIV_UNIX_SOCKET LOCALSTATEDIR "/run/libvirt/libvirt-sock" #define LIBVIRTD_PRIV_UNIX_SOCKET_RO LOCALSTATEDIR "/run/libvirt/libvirt-sock-ro" #define LIBVIRTD_USER_UNIX_SOCKET "libvirt-sock" -#define LIBVIRTD_CONFIGURATION_FILE SYSCONFDIR "/libvirt/libvirtd.conf" /* Defaults for PKI directory. */ #define LIBVIRT_PKI_DIR SYSCONFDIR "/pki" -- 2.21.0

3 2