- Devel - libvirt List Archives

[libvirt] RFC: btrfs storage pool using subvolumes & snapshots
by Oskari Saarenmaa 30 Aug '13

30 Aug '13

I wrote an experimental btrfs storage pool which uses subvolumes (and optionally snapshots) as storage volumes in LXC domains. The code is available at https://github.com/saaros/libvirt/compare/btrfs-storage but it's still missing some features like quotas for the subvolumes (currently the capacity definition for volumes is ignored) and doesn't have any documentation so far. Sample usage: mkdir /virtual; mkfs.btrfs /dev/vdb; mount -t btrfs /dev/vdb /virtual virsh pool-create-as testpool btrfs --target /virtual virsh vol-create-as testpool vanilla 0 echo vanilla > /virtual/vanilla/test virsh vol-create-as testpool test 0 --backing-vol vanilla cat /virtual/test/test btrfs subvolume list /virtual Does this look like a useful feature and does it make sense to implement it as a new storage pool type, or should it be merged into an existing one? I looked at the existing ones and couldn't really figure out how to make it fit nicely in any of them. As far as I could tell none of the existing storage pools or volumes offered a way to create a new copy-on-write volume for easy use in LXC domains using the libvirt API. With the new btrfs pool I was able to replace KVM domains using qcow2 volumes with LXC + btrfs with very little changes to the application code. Cheers, Oskari

3 2

[libvirt] [PATCH] random: don't mix RAND_MAX with random_r
by Eric Blake 30 Aug '13

30 Aug '13

FreeBSD 10 recently changed their definition of RAND_MAX, to try and cover the fact that their evenly distributed results really are a smaller range than a full power of 2. As a result, I did some investigation, and learned: 1. POSIX requires random() to be evenly distributed across exactly 31 bits. glibc also guarantees this for rand(), but the two are unrelated, and POSIX only associates RAND_MAX with rand(). Avoiding RAND_MAX altogether thus avoids a build failure on FreeBSD 10. 2. Concatenating random bits from a PRNG will NOT provide uniform coverage over the larger value UNLESS the period of the original PRNG is at least as large as the number of bits being concatenated. Simple example: suppose that RAND_MAX were 1 with a period of 2**1 (which means that the PRNG merely alternates between 0 and 1). Concatenating two successive rand() calls would then invariably result in 01 or 10, which is a rather non-uniform distribution (00 and 11 are impossible) and an even worse period (2**0, since our second attempt will get the same number as our first attempt). But a RAND_MAX of 1 with a period of 2**2 (alternating between 0, 1, 1, 0) provides sane coverage of all four values, if properly tempered. (Back-to-back calls would still only see half the values if we don't do some tempering). We therefore want to guarantee a period of at least 2**64, preferably larger (as a tempering factor); POSIX only makes this guarantee for random() with 256 bytes of info. * src/util/virrandom.c (virRandomBits): Use constants that are accurate for the PRNG we are using, not an unrelated PRNG. (randomState): Ensure the period of our PRNG exceeds our usage. Signed-off-by: Eric Blake <eblake(a)redhat.com> --- src/util/virrandom.c | 27 ++++++++++++++++++--------- 1 file changed, 18 insertions(+), 9 deletions(-) diff --git a/src/util/virrandom.c b/src/util/virrandom.c index c233732..0e368b0 100644 --- a/src/util/virrandom.c +++ b/src/util/virrandom.c @@ -1,5 +1,5 @@ /* - * Copyright (C) 2012 Red Hat, Inc. + * Copyright (C) 2012-2013 Red Hat, Inc. * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public @@ -36,7 +36,7 @@ #define VIR_FROM_THIS VIR_FROM_NONE -static char randomState[128]; +static char randomState[256]; static struct random_data randomData; static virMutex randomLock; @@ -70,0 +70,0 @@ virRandomOnceInit(void) VIR_ONCE_GLOBAL_INIT(virRandom) -/* The algorithm of virRandomBits requires that RAND_MAX == 2^n-1 for - * some n; gnulib's random_r meets this property. */ -verify(((RAND_MAX + 1U) & RAND_MAX) == 0); +/* The algorithm of virRandomBits relies on gnulib's guarantee that + * random_r() matches the POSIX requirements on random() of being + * evenly distributed among exactly [0, 2**31) (that is, we always get + * exactly 31 bits). While this happens to be the value of RAND_MAX + * on glibc, note that POSIX only requires RAND_MAX to be tied to the + * weaker rand(), so there are platforms where RAND_MAX is smaller + * than the range of random_r(). For the results to be evenly + * distributed among up to 64 bits, we also rely on the period of + * random_r() to be at least 2**64. */ +enum { + RANDOM_BITS_PER_ITER = 31, + RANDOM_BITS_MASK = (1U << RANDOM_BITS_PER_ITER) - 1, +}; /** * virRandomBits: @@ -85,7 +95,6 @@ verify(((RAND_MAX + 1U) & RAND_MAX) == 0); */ uint64_t virRandomBits(int nbits) { - int bits_per_iter = count_one_bits(RAND_MAX); uint64_t ret = 0; int32_t bits; @@ -98,10 +107,10 @@ uint64_t virRandomBits(int nbits) virMutexLock(&randomLock); - while (nbits > bits_per_iter) { + while (nbits > RANDOM_BITS_PER_ITER) { random_r(&randomData, &bits); - ret = (ret << bits_per_iter) | (bits & RAND_MAX); - nbits -= bits_per_iter; + ret = (ret << RANDOM_BITS_PER_ITER) | (bits & RANDOM_BITS_MASK); + nbits -= RANDOM_BITS_PER_ITER; } random_r(&randomData, &bits); -- 1.8.3.1

1 1

[libvirt] FreeBSD, no gcc present libvirt build issue
by Jason Helfman 29 Aug '13

29 Aug '13

Hello, There is an initiative at FreeBSD in removing gcc from base system and using CLANG. That being said, we are trying to resolve issues of ports that can't build without gcc. Libvirt fell into this group, and are curious if anyone has any ideas on why this breakage would occur. Here is a link to the buildlog: http://pb2.nyi.freebsd.org/bulk/nogcc-default/2013-08-26_21h09m25s/logs/err… Thanks very much! -jgh -- Jason Helfman | FreeBSD Committer jgh(a)FreeBSD.org | http://people.freebsd.org/~jgh | The Power to Serve

3 10

[libvirt] [PATCH 00/12] Ensure array bounds checking is present on all RPC calls
by Daniel P. Berrange 29 Aug '13

29 Aug '13

From: "Daniel P. Berrange" <berrange(a)redhat.com> Missing bounds checking on array parameters is a security issue for libvirtd, since it allows a client to make libvirtd allocate unbounded memory. Missing bounds checking on array return values is not a security issue, but it is a robustness issue. If the RPC stream somehow got corrupted, the client could end up allocating unbounded memory. The first patch in this series fixes the security flaw introduced in version 1.1.0, and indentified during discussion of this patch: https://www.redhat.com/archives/libvir-list/2013-August/msg00787.html The remaining patches address the robustness issues, and add a test suite to prevent this flaw recurring. Daniel P. Berrange (12): Add bounds checking on virDomainMigrate*Params RPC calls (CVE-2013-4292) Add bounds checking on virDomainGetJobStats RPC call Add bounds checking on virDomain{SnapshotListAllChildren,ListAllSnapshots} RPC calls Add bounds checking on virConnectListAllDomains RPC call Add bounds checking on virConnectListAllStoragePools RPC call Add bounds checking on virStoragePoolListAllVolumes RPC call Add bounds checking on virConnectListAllNetworks RPC call Add bounds checking on virConnectListAllInterfaces RPC call Add bounds checking on virConnectListAllNodeDevices RPC call Add bounds checking on virConnectListAllNWFilters RPC call Add bounds checking on virConnectListAllSecrets RPC call Prohibit unbounded arrays in XDR protocols cfg.mk | 6 ++ daemon/remote.c | 119 +++++++++++++++++++++++++++++++++++++++ src/remote/remote_driver.c | 130 +++++++++++++++++++++++++++++++++++++++++-- src/remote/remote_protocol.x | 108 ++++++++++++++++++----------------- 4 files changed, 304 insertions(+), 59 deletions(-) -- 1.8.3.1

3 17

[libvirt] [PATCH] security: provide supplemental groups even when parsing label (CVE-2013-4291)
by Eric Blake 29 Aug '13

29 Aug '13

Commit 29fe5d7 (released in 1.1.1) introduced a latent problem for any caller of virSecurityManagerSetProcessLabel and where the domain already had a uid:gid label to be parsed. Such a setup would collect the list of supplementary groups during virSecurityManagerPreFork, but then ignores that information, and thus fails to call setgroups() to adjust the supplementary groups of the process. Upstream does not use virSecurityManagerSetProcessLabel for qemu (it uses virSecurityManagerSetChildProcessLabel instead), so this problem remained latent until backporting the initial commit into v0.10.2-maint (commit c061ff5, released in 0.10.2.7), where virSecurityManagerSetChildProcessLabel has not been backported. As a result of using a different code path in the backport, attempts to start a qemu domain that runs as qemu:qemu will end up with supplementary groups unchanged from the libvirtd parent process, rather than the desired supplementary groups of the qemu user. This can lead to failure to start a domain (typical Fedora setup assigns user 107 'qemu' to both group 107 'qemu' and group 36 'kvm', so a disk image that is only readable under kvm group rights is locked out). Worse, it is a security hole (the qemu process will inherit supplemental group rights from the parent libvirtd process, which means it has access rights to files owned by group 0 even when such files should not normally be visible to user qemu). https://bugzilla.redhat.com/show_bug.cgi?id=964359 first demonstrated the problem of broken permissions in qemu when using a build based on 0.10.2. LXC does not use the DAC security driver, so it is not vulnerable at this time. Still, it is better to plug the latent hole on the master branch first, before cherry-picking it to the only vulnerable branch v0.10.2-maint. * src/security/security_dac.c (virSecurityDACGetIds): Always populate groups and ngroups, rather than only when no label is parsed. Signed-off-by: Eric Blake <eblake(a)redhat.com> --- src/security/security_dac.c | 16 ++++++---------- 1 file changed, 6 insertions(+), 10 deletions(-) diff --git a/src/security/security_dac.c b/src/security/security_dac.c index 8cbb083..6876bd5 100644 --- a/src/security/security_dac.c +++ b/src/security/security_dac.c @@ -115,13 +115,13 @@ virSecurityDACGetIds(virDomainDefPtr def, virSecurityDACDataPtr priv, return -1; } - if ((ret = virSecurityDACParseIds(def, uidPtr, gidPtr)) <= 0) { - if (groups) - *groups = NULL; - if (ngroups) - *ngroups = 0; + if (groups) + *groups = priv ? priv->groups : NULL; + if (ngroups) + *ngroups = priv ? priv->ngroups : 0; + + if ((ret = virSecurityDACParseIds(def, uidPtr, gidPtr)) <= 0) return ret; - } if (!priv) { virReportError(VIR_ERR_INTERNAL_ERROR, @@ -134,10 +134,6 @@ virSecurityDACGetIds(virDomainDefPtr def, virSecurityDACDataPtr priv, *uidPtr = priv->user; if (gidPtr) *gidPtr = priv->group; - if (groups) - *groups = priv->groups; - if (ngroups) - *ngroups = priv->ngroups; return 0; } -- 1.8.3.1

2 2

[libvirt] [PATCH] Fix a PyList usage mistake
by Guan Qiang 29 Aug '13

29 Aug '13

From: Guan Qiang <hzguanqiang(a)corp.netease.com> Fix PyList usage mistake in Function libvirt_lxc_virDomainLxcOpenNamespace. https://bugzilla.redhat.com/show_bug.cgi?id=1002383 --- python/libvirt-lxc-override.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/python/libvirt-lxc-override.c b/python/libvirt-lxc-override.c index 73685df..fa7e963 100644 --- a/python/libvirt-lxc-override.c +++ b/python/libvirt-lxc-override.c @@ -80,7 +80,7 @@ libvirt_lxc_virDomainLxcOpenNamespace(PyObject *self ATTRIBUTE_UNUSED, if (c_retval < 0) return VIR_PY_NONE; - py_retval = PyList_New(c_retval); + py_retval = PyList_New(0); for (i = 0; i < c_retval; i++) { PyObject *item = NULL; -- 1.7.10.4

2 1

[libvirt] [PATCHv2] virNetDevVethCreate: assign container if name based on parent if name
by Oskari Saarenmaa 29 Aug '13

29 Aug '13

Replace the loop trying to find a free veth interface name for the container by assigning the container if name to parent name + 'p' by default. Interface name selection logic is susceptible to race conditions, so try to select just one name by default and use that as a template for the second name. The parent name can also be overriden in domain configuration. Signed-off-by: Oskari Saarenmaa <os(a)ohmu.fi> --- v2: generate first name as before (if it wasn't given by the caller), and use the parent if name as a template for the container if name so that the caller doesn't have to select two names (which is not possible at the moment.) src/util/virnetdevveth.c | 12 +++--------- 1 file changed, 3 insertions(+), 9 deletions(-) diff --git a/src/util/virnetdevveth.c b/src/util/virnetdevveth.c index 039767f..91e2829 100644 --- a/src/util/virnetdevveth.c +++ b/src/util/virnetdevveth.c @@ -114,20 +114,14 @@ int virNetDevVethCreate(char** veth1, char** veth2) } argv[3] = *veth1; - while (*veth2 == NULL) { - if ((vethDev = virNetDevVethGetFreeName(veth2, vethDev)) < 0) { + if (*veth2 == NULL) { + /* Append a 'p' to veth1 if name */ + if (virAsprintf(veth2, "%sp", *veth1) < 0) { if (veth1_alloc) VIR_FREE(*veth1); goto cleanup; } - /* Just make sure they didn't accidentally get same name */ - if (STREQ(*veth1, *veth2)) { - vethDev++; - VIR_FREE(*veth2); - continue; - } - VIR_DEBUG("Assigned guest: %s", *veth2); veth2_alloc = true; } -- 1.8.3.1

2 3

[libvirt] [PATCH] autogen.sh: Correctly detect .git as a file
by Michal Privoznik 29 Aug '13

29 Aug '13

One of my previous patches 5cfe0d37cd0be tried to handle the case when libvirt is a submodule of another project. In that case, the .git is just a link to the parent .git directory (which the autogen.sh script didn't count on). The fix was missing 'test' though. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- autogen.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/autogen.sh b/autogen.sh index f7c2022..5aa1990 100755 --- a/autogen.sh +++ b/autogen.sh @@ -64,7 +64,7 @@ bootstrap_hash() # like to run 'git clean -x -f po' to fix it; but only ./bootstrap regenerates # the required file po/Makevars. # Only run bootstrap from a git checkout, never from a tarball. -if test -d .git || -f .git; then +if test -d .git || test -f .git; then curr_status=.git-module-status t= if test "$no_git"; then t=no-git -- 1.8.1.5

2 1

[libvirt] Question about how to contribute patches
by hzguanqiang＠corp.netease.com 29 Aug '13

29 Aug '13

Hi exports, I tried to contribute a patch to fix a bug, following the contributor guidelines on http://libvirt.org/hacking.html. After I do the 'git send-email' operation and edit the email content, It returns: Who should the emails appear to be from? [Guan Qiang <hzguanqiang(a)corp.netease.com>] Emails will be sent from: Guan Qiang <hzguanqiang(a)corp.netease.com> Message-ID to be used as In-Reply-To for the first email? (mbox) Adding cc: Guan Qiang <hzguanqiang(a)corp.netease.com> from line 'From: Guan Qiang <hzguanqiang(a)corp.netease.com>' From: Guan Qiang <hzguanqiang(a)corp.netease.com> To: libvir-list(a)redhat.com Cc: Guan Qiang <hzguanqiang(a)corp.netease.com> Subject: [PATCH]Fix a PyList usage mistake Date: Thu, 29 Aug 2013 13:41:38 +0800 Message-Id: <1377754899-29698-1-git-send-email-hzguanqiang(a)corp.netease.com> X-Mailer: git-send-email 1.7.10.4 The Cc list above has been expanded by additional addresses found in the patch commit message. By default send-email prompts before sending whenever this occurs. This behavior is controlled by the sendemail.confirm configuration setting. For additional information, run 'git send-email --help'. To retain the current behavior, but squelch this message, run 'git config --global sendemail.confirm auto'. Send this email? ([y]es|[n]o|[q]uit|[a]ll): y OK. Log says: Sendmail: /usr/sbin/sendmail -i libvir-list(a)redhat.com hzguanqiang(a)corp.netease.com From: Guan Qiang <hzguanqiang(a)corp.netease.com> To: libvir-list(a)redhat.com Cc: Guan Qiang <hzguanqiang(a)corp.netease.com> Subject: [PATCH]Fix a PyList usage mistake Date: Thu, 29 Aug 2013 13:41:38 +0800 Message-Id: <1377754899-29698-1-git-send-email-hzguanqiang(a)corp.netease.com> X-Mailer: git-send-email 1.7.10.4 Result: OK (mbox) Adding cc: Guan Qiang <hzguanqiang(a)corp.netease.com> from line 'From: Guan Qiang <hzguanqiang(a)corp.netease.com>' From: Guan Qiang <hzguanqiang(a)corp.netease.com> To: libvir-list(a)redhat.com Cc: Guan Qiang <hzguanqiang(a)corp.netease.com> Subject: [PATCH] Fix a PyList usage mistake Date: Thu, 29 Aug 2013 13:41:39 +0800 Message-Id: <1377754899-29698-2-git-send-email-hzguanqiang(a)corp.netease.com> X-Mailer: git-send-email 1.7.10.4 In-Reply-To: <1377754899-29698-1-git-send-email-hzguanqiang(a)corp.netease.com> References: <1377754899-29698-1-git-send-email-hzguanqiang(a)corp.netease.com> Send this email? ([y]es|[n]o|[q]uit|[a]ll): y OK. Log says: Sendmail: /usr/sbin/sendmail -i libvir-list(a)redhat.com hzguanqiang(a)corp.netease.com From: Guan Qiang <hzguanqiang(a)corp.netease.com> To: libvir-list(a)redhat.com Cc: Guan Qiang <hzguanqiang(a)corp.netease.com> Subject: [PATCH] Fix a PyList usage mistake Date: Thu, 29 Aug 2013 13:41:39 +0800 Message-Id: <1377754899-29698-2-git-send-email-hzguanqiang(a)corp.netease.com> X-Mailer: git-send-email 1.7.10.4 In-Reply-To: <1377754899-29698-1-git-send-email-hzguanqiang(a)corp.netease.com> References: <1377754899-29698-1-git-send-email-hzguanqiang(a)corp.netease.com> Result: OK Does this mean everything is done? Or do I still need to do other thing? Since I do this operation, It had passed three hours and I still can't find my patch email on 'libvir-list'. Really appreciate your advice, Thanks. ------------------ Best regards! GuanQiang 2013-08-29

4 4

[libvirt] [PATCH] virNetDevVethCreate: assign names based on mac address by default
by Oskari Saarenmaa 29 Aug '13

29 Aug '13

Interface names do not have to be numerical (or veth + number) and trying to assign them to that format is susceptible to race conditions. Instead, assign the parent interface name according to the mac address (the last three bytes) if no name was given by the caller and use the parent interface name plus 'p' for the container name. Signed-off-by: Oskari Saarenmaa <os(a)ohmu.fi> --- src/lxc/lxc_process.c | 2 +- src/util/virnetdevveth.c | 81 +++++++++--------------------------------------- src/util/virnetdevveth.h | 6 ++-- 3 files changed, 19 insertions(+), 70 deletions(-) diff --git a/src/lxc/lxc_process.c b/src/lxc/lxc_process.c index 4835bd5..956bbdb 100644 --- a/src/lxc/lxc_process.c +++ b/src/lxc/lxc_process.c @@ -243,7 +243,7 @@ char *virLXCProcessSetupInterfaceBridged(virConnectPtr conn, VIR_DEBUG("calling vethCreate()"); parentVeth = net->ifname; - if (virNetDevVethCreate(&parentVeth, &containerVeth) < 0) + if (virNetDevVethCreate(&parentVeth, &containerVeth, &net->mac) < 0) goto cleanup; VIR_DEBUG("parentVeth: %s, containerVeth: %s", parentVeth, containerVeth); diff --git a/src/util/virnetdevveth.c b/src/util/virnetdevveth.c index 039767f..c0e3e93 100644 --- a/src/util/virnetdevveth.c +++ b/src/util/virnetdevveth.c @@ -38,96 +38,46 @@ /* Functions */ /** - * virNetDevVethGetFreeName: - * @veth: pointer to store returned name for veth device - * @startDev: device number to start at (x in vethx) - * - * Looks in /sys/class/net/ to find the first available veth device - * name. - * - * Returns non-negative device number on success or -1 in case of error - */ -static int virNetDevVethGetFreeName(char **veth, int startDev) -{ - int devNum = startDev-1; - char *path = NULL; - - VIR_DEBUG("Find free from veth%d", startDev); - do { - VIR_FREE(path); - ++devNum; - if (virAsprintf(&path, "/sys/class/net/veth%d/", devNum) < 0) - return -1; - VIR_DEBUG("Probe %s", path); - } while (virFileExists(path)); - VIR_FREE(path); - - if (virAsprintf(veth, "veth%d", devNum) < 0) - return -1; - - return devNum; -} - -/** * virNetDevVethCreate: * @veth1: pointer to name for parent end of veth pair * @veth2: pointer to return name for container end of veth pair + * @mac: mac address of the device * * Creates a veth device pair using the ip command: * ip link add veth1 type veth peer name veth2 - * If veth1 points to NULL on entry, it will be a valid interface on - * return. veth2 should point to NULL on entry. * - * NOTE: If veth1 and veth2 names are not specified, ip will auto assign - * names. There seems to be two problems here - - * 1) There doesn't seem to be a way to determine the names of the - * devices that it creates. They show up in ip link show and - * under /sys/class/net/ however there is no guarantee that they - * are the devices that this process just created. - * 2) Once one of the veth devices is moved to another namespace, it - * is no longer visible in the parent namespace. This seems to - * confuse the name assignment causing it to fail with File exists. - * Because of these issues, this function currently allocates names - * prior to using the ip command, and returns any allocated names - * to the caller. + * If veth1 or veth2 points to NULL on entry, they will be a valid interface + * on return. The name is generated based on the mac address given. * * Returns 0 on success or -1 in case of error */ -int virNetDevVethCreate(char** veth1, char** veth2) +int virNetDevVethCreate(char** veth1, char** veth2, const virMacAddrPtr mac) { - int rc = -1; const char *argv[] = { "ip", "link", "add", NULL, "type", "veth", "peer", "name", NULL, NULL }; - int vethDev = 0; bool veth1_alloc = false; bool veth2_alloc = false; VIR_DEBUG("Host: %s guest: %s", NULLSTR(*veth1), NULLSTR(*veth2)); if (*veth1 == NULL) { - if ((vethDev = virNetDevVethGetFreeName(veth1, vethDev)) < 0) - goto cleanup; + /* Use the last three bytes of the mac address as our if name */ + if (virAsprintf(veth1, "veth%02x%02x%02x", + mac->addr[3], mac->addr[4], mac->addr[5]) < 0) + return -1; VIR_DEBUG("Assigned host: %s", *veth1); veth1_alloc = true; - vethDev++; } argv[3] = *veth1; - while (*veth2 == NULL) { - if ((vethDev = virNetDevVethGetFreeName(veth2, vethDev)) < 0) { + if (*veth2 == NULL) { + /* Append a 'p' to veth1 if name */ + if (virAsprintf(veth2, "%sp", *veth1) < 0) { if (veth1_alloc) VIR_FREE(*veth1); - goto cleanup; - } - - /* Just make sure they didn't accidentally get same name */ - if (STREQ(*veth1, *veth2)) { - vethDev++; - VIR_FREE(*veth2); - continue; + return -1; } - VIR_DEBUG("Assigned guest: %s", *veth2); veth2_alloc = true; } @@ -139,13 +89,10 @@ int virNetDevVethCreate(char** veth1, char** veth2) VIR_FREE(*veth1); if (veth2_alloc) VIR_FREE(*veth2); - goto cleanup; + return -1; } - rc = 0; - -cleanup: - return rc; + return 0; } /** diff --git a/src/util/virnetdevveth.h b/src/util/virnetdevveth.h index 4c220c1..175fbfa 100644 --- a/src/util/virnetdevveth.h +++ b/src/util/virnetdevveth.h @@ -25,10 +25,12 @@ # define __VIR_NETDEV_VETH_H__ # include "internal.h" +# include "virmacaddr.h" /* Function declarations */ -int virNetDevVethCreate(char **veth1, char **veth2) - ATTRIBUTE_NONNULL(1) ATTRIBUTE_NONNULL(2) ATTRIBUTE_RETURN_CHECK; +int virNetDevVethCreate(char **veth1, char **veth2, const virMacAddrPtr mac) + ATTRIBUTE_NONNULL(1) ATTRIBUTE_NONNULL(2) ATTRIBUTE_NONNULL(3) + ATTRIBUTE_RETURN_CHECK; int virNetDevVethDelete(const char *veth) ATTRIBUTE_NONNULL(1) ATTRIBUTE_RETURN_CHECK; -- 1.8.3.1

4 7

[libvirt] [PATCH] bridge_driver: Introduce networkObjFromNetwork
by Michal Privoznik 29 Aug '13

29 Aug '13

Similarly to qemu_driver.c, we can join often repeating code of looking up network into one function: networkObjFromNetwork. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/network/bridge_driver.c | 76 ++++++++++++++++++--------------------------- 1 file changed, 31 insertions(+), 45 deletions(-) diff --git a/src/network/bridge_driver.c b/src/network/bridge_driver.c index 62ae0b7..3a8be90 100644 --- a/src/network/bridge_driver.c +++ b/src/network/bridge_driver.c @@ -112,6 +112,27 @@ static int networkUnplugBandwidth(virNetworkObjPtr net, static virNetworkDriverStatePtr driverState = NULL; +static virNetworkObjPtr +networkObjFromNetwork(virNetworkPtr net) +{ + virNetworkDriverStatePtr driver = net->conn->networkPrivateData; + virNetworkObjPtr network; + char uuidstr[VIR_UUID_STRING_BUFLEN]; + + networkDriverLock(driver); + network = virNetworkFindByUUID(&driver->networks, net->uuid); + networkDriverUnlock(driver); + + if (!network) { + virUUIDFormat(net->uuid, uuidstr); + virReportError(VIR_ERR_NO_NETWORK, + _("no network with matching uuid '%s' (%s)"), + uuidstr, net->name); + } + + return network; +} + static char * networkDnsmasqLeaseFileNameDefault(const char *netname) { @@ -2260,20 +2281,14 @@ cleanup: static int networkIsActive(virNetworkPtr net) { - virNetworkDriverStatePtr driver = net->conn->networkPrivateData; virNetworkObjPtr obj; int ret = -1; - networkDriverLock(driver); - obj = virNetworkFindByUUID(&driver->networks, net->uuid); - networkDriverUnlock(driver); - if (!obj) { - virReportError(VIR_ERR_NO_NETWORK, NULL); - goto cleanup; - } + if (!(obj = networkObjFromNetwork(net))) + return ret; if (virNetworkIsActiveEnsureACL(net->conn, obj->def) < 0) goto cleanup; ret = virNetworkObjIsActive(obj); @@ -2285,20 +2300,14 @@ cleanup: static int networkIsPersistent(virNetworkPtr net) { - virNetworkDriverStatePtr driver = net->conn->networkPrivateData; virNetworkObjPtr obj; int ret = -1; - networkDriverLock(driver); - obj = virNetworkFindByUUID(&driver->networks, net->uuid); - networkDriverUnlock(driver); - if (!obj) { - virReportError(VIR_ERR_NO_NETWORK, NULL); - goto cleanup; - } + if (!(obj = networkObjFromNetwork(net))) + return ret; if (virNetworkIsPersistentEnsureACL(net->conn, obj->def) < 0) goto cleanup; ret = obj->persistent; @@ -2832,30 +2841,22 @@ cleanup: static char *networkGetXMLDesc(virNetworkPtr net, unsigned int flags) { - virNetworkDriverStatePtr driver = net->conn->networkPrivateData; virNetworkObjPtr network; virNetworkDefPtr def; char *ret = NULL; virCheckFlags(VIR_NETWORK_XML_INACTIVE, NULL); - networkDriverLock(driver); - network = virNetworkFindByUUID(&driver->networks, net->uuid); - networkDriverUnlock(driver); - - if (!network) { - virReportError(VIR_ERR_NO_NETWORK, - "%s", _("no network with matching uuid")); - goto cleanup; - } + if (!(network = networkObjFromNetwork(net))) + return ret; if (virNetworkGetXMLDescEnsureACL(net->conn, network->def) < 0) goto cleanup; if ((flags & VIR_NETWORK_XML_INACTIVE) && network->newDef) def = network->newDef; else def = network->def; ret = virNetworkDefFormat(def, flags); @@ -2866,29 +2867,21 @@ cleanup: } static char *networkGetBridgeName(virNetworkPtr net) { - virNetworkDriverStatePtr driver = net->conn->networkPrivateData; virNetworkObjPtr network; char *bridge = NULL; - networkDriverLock(driver); - network = virNetworkFindByUUID(&driver->networks, net->uuid); - networkDriverUnlock(driver); - - if (!network) { - virReportError(VIR_ERR_NO_NETWORK, - "%s", _("no network with matching id")); - goto cleanup; - } + if (!(network = networkObjFromNetwork(net))) + return bridge; if (virNetworkGetBridgeNameEnsureACL(net->conn, network->def) < 0) goto cleanup; if (!(network->def->bridge)) { virReportError(VIR_ERR_INTERNAL_ERROR, _("network '%s' does not have a bridge name."), network->def->name); goto cleanup; } ignore_value(VIR_STRDUP(bridge, network->def->bridge)); @@ -2900,22 +2893,15 @@ cleanup: static int networkGetAutostart(virNetworkPtr net, int *autostart) { - virNetworkDriverStatePtr driver = net->conn->networkPrivateData; virNetworkObjPtr network; int ret = -1; - networkDriverLock(driver); - network = virNetworkFindByUUID(&driver->networks, net->uuid); - networkDriverUnlock(driver); - if (!network) { - virReportError(VIR_ERR_NO_NETWORK, - "%s", _("no network with matching uuid")); - goto cleanup; - } + if (!(network = networkObjFromNetwork(net))) + return ret; if (virNetworkGetAutostartEnsureACL(net->conn, network->def) < 0) goto cleanup; *autostart = network->autostart; ret = 0; -- 1.8.1.5

2 1

[libvirt] [PATCHv2 0/2] Fix crash of libvirtd with <interaface type=network and hostdev forwarding
by Peter Krempa 29 Aug '13

29 Aug '13

Version 2 contains a fix of more callers in qemu and a whitespace cleanup. Peter Krempa (2): qemu: Remove hostdev entry when freeing the depending network entry qemu_hotplug: Fix whitespace around addition in argument src/conf/domain_conf.c | 26 ++++++++++++++++++-------- src/conf/domain_conf.h | 1 + src/libvirt_private.syms | 1 + src/qemu/qemu_hotplug.c | 9 +++++++-- src/qemu/qemu_process.c | 2 ++ 5 files changed, 29 insertions(+), 10 deletions(-) -- 1.8.3.2

2 4

[libvirt] [PATCH 0/2] netcf: use only a single netcf instance for all connections
by Laine Stump 28 Aug '13

28 Aug '13

There is a good description in 2/2. I'll be out of reach for the next several days, so unable to push even if it gets ACKed before the freeze. I have done a rudimentary test and it seems to work, but if someone with a suitable setup wants to do a stress test with [many] connections, that would be very useful. Laine Stump (2): rename "struct interface_driver" to virNetcfDriverState netcf driver: use a single netcf handle for all connections src/interface/interface_backend_netcf.c | 209 ++++++++++++++++++++++---------- 1 file changed, 144 insertions(+), 65 deletions(-) -- 1.7.11.7

2 4

[libvirt] [PATCH] virsh: detect programming errors with option parsing
by Eric Blake 28 Aug '13

28 Aug '13

Noticed while reviewing another patch that had an accidental mismatch due to refactoring. An audit of the code showed that very few callers of vshCommandOpt were expecting a return of -2, indicating programmer error, and of those that DID check, they just propagated that status to yet another caller that did not check. Fix this by making the code blatantly warn the programmer, rather than silently ignoring it and possibly doing the wrong thing downstream. I know that we frown on assert()/abort() inside libvirtd (libraries should NEVER kill the program that linked them), but as virsh is an app rather than the library, and as this is not the first use of assert() in virsh, I think this approach is okay. * tools/virsh.h (vshCommandOpt): Drop declaration. * tools/virsh.c (vshCommandOpt): Make static, and add a parameter. Abort on programmer errors rather than making callers repeat that logic. (vshCommandOptInt, vshCommandOptUInt, vshCommandOptUL) (vshCommandOptString, vshCommandOptStringReq) (vshCommandOptLongLong, vshCommandOptULongLong) (vshCommandOptBool): Adjust callers. Signed-off-by: Eric Blake <eblake(a)redhat.com> --- In response to my observation on Don's email: https://www.redhat.com/archives/libvir-list/2013-August/msg00769.html tools/virsh.c | 97 +++++++++++++++++++++-------------------------------------- tools/virsh.h | 5 +-- 2 files changed, 35 insertions(+), 67 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index f65dc79..4c4e92c 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -25,6 +25,7 @@ #include <config.h> #include "virsh.h" +#include <assert.h> #include <stdio.h> #include <stdlib.h> #include <string.h> @@ -1340,39 +1341,46 @@ vshCommandFree(vshCmd *cmd) * @cmd: parsed command line to search * @name: option name to search for * @opt: result of the search + * @needData: true if option must be non-boolean * * Look up an option passed to CMD by NAME. Returns 1 with *OPT set * to the option if found, 0 with *OPT set to NULL if the name is * valid and the option is not required, -1 with *OPT set to NULL if - * the option is required but not present, and -2 if NAME is not valid - * (-2 indicates a programming error). No error messages are issued. + * the option is required but not present, and abort if NAME is not valid + * (which indicates a programming error). No error messages are issued + * if a value is returned. */ -int -vshCommandOpt(const vshCmd *cmd, const char *name, vshCmdOpt **opt) +static int +vshCommandOpt(const vshCmd *cmd, const char *name, vshCmdOpt **opt, + bool needData) { vshCmdOpt *candidate = cmd->opts; const vshCmdOptDef *valid = cmd->def->opts; + int ret = 0; + + /* See if option is valid and/or required. */ + *opt = NULL; + while (valid) { + assert(valid->name); + if (STREQ(name, valid->name)) + break; + valid++; + } + if (needData && valid->type == VSH_OT_BOOL) + abort(); + if (valid->flags & VSH_OFLAG_REQ) + ret = -1; /* See if option is present on command line. */ while (candidate) { if (STREQ(candidate->def->name, name)) { *opt = candidate; - return 1; + ret = 1; + break; } candidate = candidate->next; } - - /* Option not present, see if command requires it. */ - *opt = NULL; - while (valid) { - if (!valid->name) - break; - if (STREQ(name, valid->name)) - return (valid->flags & VSH_OFLAG_REQ) == 0 ? 0 : -1; - valid++; - } - /* If we got here, the name is unknown. */ - return -2; + return ret; } /** @@ -1393,14 +1401,9 @@ vshCommandOptInt(const vshCmd *cmd, const char *name, int *value) vshCmdOpt *arg; int ret; - ret = vshCommandOpt(cmd, name, &arg); + ret = vshCommandOpt(cmd, name, &arg, true); if (ret <= 0) return ret; - if (!arg->data) { - /* only possible on bool, but if name is bool, this is a - * programming bug */ - return -2; - } if (virStrToLong_i(arg->data, NULL, 10, value) < 0) return -1; @@ -1423,14 +1426,9 @@ vshCommandOptUInt(const vshCmd *cmd, const char *name, unsigned int *value) vshCmdOpt *arg; int ret; - ret = vshCommandOpt(cmd, name, &arg); + ret = vshCommandOpt(cmd, name, &arg, true); if (ret <= 0) return ret; - if (!arg->data) { - /* only possible on bool, but if name is bool, this is a - * programming bug */ - return -2; - } if (virStrToLong_ui(arg->data, NULL, 10, value) < 0) return -1; @@ -1453,14 +1451,9 @@ vshCommandOptUL(const vshCmd *cmd, const char *name, unsigned long *value) vshCmdOpt *arg; int ret; - ret = vshCommandOpt(cmd, name, &arg); + ret = vshCommandOpt(cmd, name, &arg, true); if (ret <= 0) return ret; - if (!arg->data) { - /* only possible on bool, but if name is bool, this is a - * programming bug */ - return -2; - } if (virStrToLong_ul(arg->data, NULL, 10, value) < 0) return -1; @@ -1485,14 +1478,9 @@ vshCommandOptString(const vshCmd *cmd, const char *name, const char **value) vshCmdOpt *arg; int ret; - ret = vshCommandOpt(cmd, name, &arg); + ret = vshCommandOpt(cmd, name, &arg, true); if (ret <= 0) return ret; - if (!arg->data) { - /* only possible on bool, but if name is bool, this is a - * programming bug */ - return -2; - } if (!*arg->data && !(arg->def->flags & VSH_OFLAG_EMPTY_OK)) { return -1; @@ -1527,21 +1515,14 @@ vshCommandOptStringReq(vshControl *ctl, /* clear out the value */ *value = NULL; - ret = vshCommandOpt(cmd, name, &arg); + ret = vshCommandOpt(cmd, name, &arg, true); /* option is not required and not present */ if (ret == 0) return 0; /* this should not be propagated here, just to be sure */ if (ret == -1) error = N_("Mandatory option not present"); - - if (ret == -2) - error = N_("Programming error: Invalid option name"); - - if (!arg->data) - error = N_("Programming error: Requested option is a boolean"); - - if (arg->data && !*arg->data && !(arg->def->flags & VSH_OFLAG_EMPTY_OK)) + else if (!*arg->data && !(arg->def->flags & VSH_OFLAG_EMPTY_OK)) error = N_("Option argument is empty"); if (error) { @@ -1569,14 +1550,9 @@ vshCommandOptLongLong(const vshCmd *cmd, const char *name, vshCmdOpt *arg; int ret; - ret = vshCommandOpt(cmd, name, &arg); + ret = vshCommandOpt(cmd, name, &arg, true); if (ret <= 0) return ret; - if (!arg->data) { - /* only possible on bool, but if name is bool, this is a - * programming bug */ - return -2; - } if (virStrToLong_ll(arg->data, NULL, 10, value) < 0) return -1; @@ -1599,14 +1575,9 @@ vshCommandOptULongLong(const vshCmd *cmd, const char *name, vshCmdOpt *arg; int ret; - ret = vshCommandOpt(cmd, name, &arg); + ret = vshCommandOpt(cmd, name, &arg, true); if (ret <= 0) return ret; - if (!arg->data) { - /* only possible on bool, but if name is bool, this is a - * programming bug */ - return -2; - } if (virStrToLong_ull(arg->data, NULL, 10, value) < 0) return -1; @@ -1659,7 +1630,7 @@ vshCommandOptBool(const vshCmd *cmd, const char *name) { vshCmdOpt *dummy; - return vshCommandOpt(cmd, name, &dummy) == 1; + return vshCommandOpt(cmd, name, &dummy, false) == 1; } /** diff --git a/tools/virsh.h b/tools/virsh.h index a407428..382e00a 100644 --- a/tools/virsh.h +++ b/tools/virsh.h @@ -1,7 +1,7 @@ /* * virsh.h: a shell to exercise the libvirt API * - * Copyright (C) 2005, 2007-2012 Red Hat, Inc. + * Copyright (C) 2005, 2007-2013 Red Hat, Inc. * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public @@ -261,9 +261,6 @@ bool vshCmddefHelp(vshControl *ctl, const char *name); const vshCmdGrp *vshCmdGrpSearch(const char *grpname); bool vshCmdGrpHelp(vshControl *ctl, const char *name); -int vshCommandOpt(const vshCmd *cmd, const char *name, vshCmdOpt **opt) - ATTRIBUTE_NONNULL(1) ATTRIBUTE_NONNULL(2) ATTRIBUTE_NONNULL(3) - ATTRIBUTE_RETURN_CHECK; int vshCommandOptInt(const vshCmd *cmd, const char *name, int *value) ATTRIBUTE_NONNULL(3) ATTRIBUTE_RETURN_CHECK; int vshCommandOptUInt(const vshCmd *cmd, const char *name, -- 1.8.3.1

2 2

Re: [libvirt] Information needed regarding libvirt error
by Martin Kletzander 28 Aug '13

28 Aug '13

[cc'ing libvir-list] On 08/26/2013 05:59 PM, arun abhinay wrote: > Hi, > > Can you please consider this is an gentle reminder for my below request. > > Thanks > Abhinay > > On Fri, Aug 23, 2013 at 1:01 PM, arun abhinay <arun.abhinay(a)gmail.com>wrote: > >> Hi Martin/Peter, >> Hi, firstly, feel free to write to the list, not just particular people. There are many more people on the list that may help you. Not mentioning that those people you write might not be available, don't have time etc. There was a similar question to the one you've asked, you could have checked that out before as well and reply on it. There's also a patch fixing cpu-stats command, check it that helps. >> In reference to >> http://comments.gmane.org/gmane.comp.emulators.libvirt/62451 >> >> We have created a VM on a qemu-kvm emulator using the domain.xml as >> mentioned in attachment. The starting of the VM is sucessful but we are >> facing issue while executing the command "cpu-stats <domain>". While >> analysis we found below below log in libvirtd.log >> >> 2013-08-22 12:45:44.326+0000: 17554: warning : qemuSetupCgroupForVcpu:953 >> : Unable to get vcpus' pids >> This basically means that you have old qemu that doesn't support (I'm guesing the command now) "query-vcpus" QMP command. The problem probably lies somewhere else, even though my change might have caused the other part to behave differently. >> The complete logs while starting VM and while executing the "cpu-stats" >> command are attached with this mail for reference. Can you please check and >> provide us some inputs or pointers to get "cpu-stats" command working. >> >> I am new to using libvirt. Your inputs would be of great help to us. >> >> Libvrit and qemu-kvm verisons used are mentioned below >> libvirt-1.0.6-1 >> qemu-kvm-0.14.1-1 >> >> <<CLA_new-0.xml >> --> Domain.xml used to define VM >> << libvirt_testing_starting_VG.log >> --> Log while staring VM >> <<libvirt_testing_cpu_stats.log >> ---> Log while executing command >> cpu_stats. >> According to the logs, it looks like it's completely the same like a question already asked... Wanting to post the link here, I've found that even the machine name is the same, so it shouldn't be a problem for you to find it. I'm looking at the issue now, so I'll see what I can do, but maybe in the meantime sombody else will reply. Martin

2 5

[libvirt] [libvirt-glib] Add missing symbols to .sym files
by Christophe Fergeau 28 Aug '13

28 Aug '13

These symbols are in public headers, but were not listed in the corresponding .sym file, causing them to be unavailable from the resulting shared library. I would have preferred not to export gvir_config_object_new_from_xml() at all, but since the similar gvir_config_object_new() is already exported, I've chosen to export it as well. --- libvirt-gconfig/libvirt-gconfig.sym | 11 +++++++++++ libvirt-gobject/libvirt-gobject.sym | 4 ++++ 2 files changed, 15 insertions(+) diff --git a/libvirt-gconfig/libvirt-gconfig.sym b/libvirt-gconfig/libvirt-gconfig.sym index be05ce4..72eafc1 100644 --- a/libvirt-gconfig/libvirt-gconfig.sym +++ b/libvirt-gconfig/libvirt-gconfig.sym @@ -599,4 +599,15 @@ LIBVIRT_GCONFIG_0.1.7 { gvir_config_domain_device_get_alias; } LIBVIRT_GCONFIG_0.1.6; +LIBVIRT_GCONFIG_0.1.8 { +global: + gvir_config_domain_get_uuid; + gvir_config_domain_set_uuid; + + gvir_config_domain_graphics_rdp_set_multi_user; + gvir_config_domain_graphics_rdp_set_replace_user; + + gvir_config_object_new_from_xml; +} LIBVIRT_GCONFIG_0.1.7; + # .... define new API here using predicted next version number .... diff --git a/libvirt-gobject/libvirt-gobject.sym b/libvirt-gobject/libvirt-gobject.sym index 9522167..edc876d 100644 --- a/libvirt-gobject/libvirt-gobject.sym +++ b/libvirt-gobject/libvirt-gobject.sym @@ -232,4 +232,8 @@ LIBVIRT_GOBJECT_0.1.5 { gvir_connection_open_read_only_finish; } LIBVIRT_GOBJECT_0.1.4; +LIBVIRT_GOBJECT_0.1.8 { + gvir_connection_restore_domain_from_file_finish; +} LIBVIRT_GOBJECT_0.1.5; + # .... define new API here using predicted next version number .... -- 1.8.3.1

3 2

[libvirt] [PATCH] virt-sanlock-cleanup; Fix augtool usage
by Jiri Denemark 28 Aug '13

28 Aug '13

Surprisingly, augtool get (or print) returns "path = value" while we are only interested in the value. We need to remove the "path = " part from the augtool's output. The following is an example of the augtool command as used in virt-sanlock-cleanup script: $ augtool get /files/etc/libvirt/qemu-sanlock.conf/disk_lease_dir /files/etc/libvirt/qemu-sanlock.conf/disk_lease_dir = /var/lib/libvirt/sanlock --- tools/virt-sanlock-cleanup.in | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/tools/virt-sanlock-cleanup.in b/tools/virt-sanlock-cleanup.in index e0e8083..9855f42 100644 --- a/tools/virt-sanlock-cleanup.in +++ b/tools/virt-sanlock-cleanup.in @@ -26,7 +26,8 @@ fi LOCKSPACE="__LIBVIRT__DISKS__" -LOCKDIR=`augtool print '/files@sysconfdir@/libvirt/qemu-sanlock.conf/disk_lease_dir'` +LOCKDIR=`augtool get '/files@sysconfdir@/libvirt/qemu-sanlock.conf/disk_lease_dir'` +LOCKDIR=${LOCKDIR#* = } if test $? != 0 || test "x$LOCKDIR" = "x" ; then LOCKDIR="@localstatedir@/lib/libvirt/sanlock" fi -- 1.8.3.2

2 3

[libvirt] [PATCH] virsh: Fix debugging
by Martin Kletzander 28 Aug '13

28 Aug '13

Commit a0b6a36f "fixed" what abfff210 broke (URI precedence), but there was still one more thing missing to fix. When using virsh parameters to setup debugging, those weren't honored, because at the time debugging was initializing, arguments weren't parsed yet. To make ewerything work as expected, we need to initialize the debugging twice, once before debugging (so we can debug option parsing properly) and then again after these options are parsed. Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com> --- tools/virsh.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/tools/virsh.c b/tools/virsh.c index 2ea44a6..ac77156 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -2333,6 +2333,10 @@ vshInitDebug(vshControl *ctl) static bool vshInit(vshControl *ctl) { + /* Since we have the commandline arguments parsed, we need to + * re-initialize all the debugging to make it work properly */ + vshInitDebug(ctl); + if (ctl->conn) return false; -- 1.8.3.2

4 6

[libvirt] [PATCH] qemu: Remove hostdev entry when freeing the depending network entry
by Peter Krempa 28 Aug '13

28 Aug '13

When using a <interface type="network"> that points to a network with hostdev forwarding mode a hostdev alias is created for the network. This allias is inserted into the hostdev list, but is backed with a part of the network object that it is connected to. When a VM is being stopped qemuProcessStop() calls networkReleaseActualDevice() which eventually frees the memory for the hostdev object. Afterwards when the domain definition is being freed by virDomainDefFree() an invalid pointer is accessed by virDomainHostdevDefFree() and may cause a crash of the daemon. This patch removes the entry in the hostdev list before freeing the depending memory to avoid this issue. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1000973 --- src/qemu/qemu_process.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 128618b..2a69c8d 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -4241,6 +4241,9 @@ void qemuProcessStop(virQEMUDriverPtr driver, def = vm->def; for (i = 0; i < def->nnets; i++) { virDomainNetDefPtr net = def->nets[i]; + virDomainHostdevDefPtr hostdev = virDomainNetGetActualHostdev(net); + int hostdev_index; + if (virDomainNetGetActualType(net) == VIR_DOMAIN_NET_TYPE_DIRECT) { ignore_value(virNetDevMacVLanDeleteWithVPortProfile( net->ifname, &net->mac, @@ -4259,6 +4262,11 @@ void qemuProcessStop(virQEMUDriverPtr driver, virDomainNetGetActualBridgeName(net), net->ifname)); + if (hostdev) { + if ((hostdev_index = virDomainHostdevFind(def, hostdev, NULL)) > 0) + virDomainHostdevRemove(def, hostdev_index); + } + networkReleaseActualDevice(net); } -- 1.8.3.2

2 3

[libvirt] [PATCH] virsh-pool.c: Don't jump over variable declaration
by Michal Privoznik 28 Aug '13

28 Aug '13

Since 785ff34bf8 we are using the outputStr variable in cleanup label. However, there is a possibility to jump to the label before the variable has been declared: virsh-pool.c: In function 'cmdPoolList': virsh-pool.c:1121:25: error: jump skips variable initialization [-Werror=jump-misses-init] goto asprintf_failure; ^ virsh-pool.c:1308:1: note: label 'asprintf_failure' defined here asprintf_failure: ^ virsh-pool.c:1267:11: note: 'outputStr' declared here char *outputStr = NULL; Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- Pushed under buil-breaker rule. tools/virsh-pool.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/virsh-pool.c b/tools/virsh-pool.c index 3d3c61e..c771226 100644 --- a/tools/virsh-pool.c +++ b/tools/virsh-pool.c @@ -964,6 +964,7 @@ cmdPoolList(vshControl *ctl, const vshCmd *cmd ATTRIBUTE_UNUSED) const char *type = NULL; bool details = vshCommandOptBool(cmd, "details"); bool inactive, all; + char *outputStr = NULL; inactive = vshCommandOptBool(cmd, "inactive"); all = vshCommandOptBool(cmd, "all"); @@ -1264,7 +1265,6 @@ cmdPoolList(vshControl *ctl, const vshCmd *cmd ATTRIBUTE_UNUSED) /* Create the output template. Each column is sized according to * the longest string. */ - char *outputStr = NULL; ret = virAsprintf(&outputStr, "%%-%lus %%-%lus %%-%lus %%-%lus %%%lus %%%lus %%%lus\n", (unsigned long) nameStrLength, -- 1.8.1.5

1 0

[libvirt] [PATCH v3 0/2] Add tests for network update XML parsing
by Ján Tomko 28 Aug '13

28 Aug '13

v1: https://www.redhat.com/archives/libvir-list/2013-July/msg01971.html v2: https://www.redhat.com/archives/libvir-list/2013-August/msg00035.html remove the spaces from other elements instead of adding one v3: rebase because of 4f595ba adding forwardPlainNames Ján Tomko (2): Remove the space before the slash in network XML Test network update XML parsing src/conf/network_conf.c | 26 +- src/conf/network_conf.h | 9 + src/libvirt_private.syms | 1 + src/network/default.xml | 4 +- src/test/test_driver.c | 4 +- tests/Makefile.am | 9 +- tests/networkxml2confdata/dhcp6-nat-network.xml | 20 +- tests/networkxml2confdata/dhcp6-network.xml | 14 +- .../dhcp6host-routed-network.xml | 16 +- tests/networkxml2confdata/isolated-network.xml | 4 +- .../networkxml2confdata/nat-network-dns-hosts.xml | 2 +- .../nat-network-dns-srv-record-minimal.xml | 10 +- .../nat-network-dns-srv-record.xml | 10 +- .../nat-network-dns-txt-record.xml | 10 +- tests/networkxml2confdata/nat-network.xml | 8 +- tests/networkxml2confdata/netboot-network.xml | 8 +- .../networkxml2confdata/netboot-proxy-network.xml | 6 +- tests/networkxml2confdata/routed-network.xml | 2 +- tests/networkxml2xmlin/bandwidth-network.xml | 4 +- .../networkxml2xmlin/dhcp6host-routed-network.xml | 16 +- tests/networkxml2xmlin/empty-allow-ipv6.xml | 2 +- tests/networkxml2xmlin/isolated-network.xml | 4 +- tests/networkxml2xmlin/nat-network-dns-hosts.xml | 2 +- .../nat-network-dns-srv-record-minimal.xml | 10 +- .../nat-network-dns-srv-record.xml | 10 +- .../nat-network-dns-srv-records.xml | 27 ++ .../nat-network-dns-txt-record.xml | 10 +- tests/networkxml2xmlin/nat-network.xml | 8 +- tests/networkxml2xmlin/netboot-network.xml | 6 +- tests/networkxml2xmlin/netboot-proxy-network.xml | 4 +- tests/networkxml2xmlin/routed-network.xml | 2 +- tests/networkxml2xmlout/bandwidth-network.xml | 4 +- .../networkxml2xmlout/dhcp6host-routed-network.xml | 16 +- tests/networkxml2xmlout/empty-allow-ipv6.xml | 2 +- tests/networkxml2xmlout/host-bridge-net.xml | 2 +- tests/networkxml2xmlout/isolated-network.xml | 4 +- .../nat-network-dns-forward-plain.xml | 2 +- tests/networkxml2xmlout/nat-network-dns-hosts.xml | 2 +- .../nat-network-dns-srv-record-minimal.xml | 10 +- .../nat-network-dns-srv-record.xml | 10 +- .../nat-network-dns-srv-records.xml | 27 ++ .../nat-network-dns-txt-record.xml | 10 +- tests/networkxml2xmlout/nat-network.xml | 8 +- tests/networkxml2xmlout/netboot-network.xml | 8 +- tests/networkxml2xmlout/netboot-proxy-network.xml | 6 +- tests/networkxml2xmlout/routed-network.xml | 2 +- tests/networkxml2xmltest.c | 3 + .../networkxml2xmlupdatein/dhcp-range-existing.xml | 1 + tests/networkxml2xmlupdatein/dhcp-range.xml | 1 + .../dns-host-gateway-incomplete.xml | 3 + tests/networkxml2xmlupdatein/dns-host-pudding.xml | 3 + .../dns-txt-record-example.xml | 1 + .../dns-txt-record-snowman.xml | 1 + tests/networkxml2xmlupdatein/host-existing.xml | 1 + tests/networkxml2xmlupdatein/host-incomplete.xml | 1 + .../networkxml2xmlupdatein/host-new-incomplete.xml | 1 + tests/networkxml2xmlupdatein/host-new.xml | 1 + tests/networkxml2xmlupdatein/host-updated.xml | 1 + tests/networkxml2xmlupdatein/interface-eth1.xml | 1 + tests/networkxml2xmlupdatein/interface-eth47.xml | 1 + .../networkxml2xmlupdatein/portgroup-alice-new.xml | 10 + tests/networkxml2xmlupdatein/portgroup-alison.xml | 11 + tests/networkxml2xmlupdatein/srv-record-donkey.xml | 1 + .../networkxml2xmlupdatein/srv-record-invalid.xml | 1 + .../networkxml2xmlupdatein/srv-record-protocol.xml | 1 + .../networkxml2xmlupdatein/srv-record-service.xml | 1 + tests/networkxml2xmlupdatein/srv-record.xml | 1 + .../networkxml2xmlupdatein/unparsable-dns-host.xml | 1 + .../dhcp6host-routed-network-another-range.xml | 27 ++ .../dhcp6host-routed-network-range.xml | 27 ++ .../nat-network-dns-more-hosts.xml | 19 ++ .../nat-network-dns-srv-record.xml | 26 ++ .../nat-network-dns-srv-records.xml | 27 ++ .../nat-network-dns-txt-none.xml | 23 ++ .../nat-network-dns-txt-records.xml | 27 ++ .../nat-network-forward-ifaces.xml | 27 ++ .../nat-network-host-updated.xml | 23 ++ .../networkxml2xmlupdateout/nat-network-hosts.xml | 24 ++ .../nat-network-no-forward-ifaces.xml | 24 ++ .../nat-network-no-hosts.xml | 10 + .../nat-network-no-range.xml | 22 ++ .../nat-network-one-host.xml | 22 ++ tests/networkxml2xmlupdateout/nat-network.xml | 23 ++ .../openvswitch-net-modified.xml | 33 ++ .../openvswitch-net-more-portgroups.xml | 44 +++ .../openvswitch-net-without-alice.xml | 23 ++ tests/networkxml2xmlupdatetest.c | 372 +++++++++++++++++++++ 87 files changed, 1096 insertions(+), 155 deletions(-) create mode 100644 tests/networkxml2xmlin/nat-network-dns-srv-records.xml create mode 100644 tests/networkxml2xmlout/nat-network-dns-srv-records.xml create mode 100644 tests/networkxml2xmlupdatein/dhcp-range-existing.xml create mode 100644 tests/networkxml2xmlupdatein/dhcp-range.xml create mode 100644 tests/networkxml2xmlupdatein/dns-host-gateway-incomplete.xml create mode 100644 tests/networkxml2xmlupdatein/dns-host-pudding.xml create mode 100644 tests/networkxml2xmlupdatein/dns-txt-record-example.xml create mode 100644 tests/networkxml2xmlupdatein/dns-txt-record-snowman.xml create mode 100644 tests/networkxml2xmlupdatein/host-existing.xml create mode 100644 tests/networkxml2xmlupdatein/host-incomplete.xml create mode 100644 tests/networkxml2xmlupdatein/host-new-incomplete.xml create mode 100644 tests/networkxml2xmlupdatein/host-new.xml create mode 100644 tests/networkxml2xmlupdatein/host-updated.xml create mode 100644 tests/networkxml2xmlupdatein/interface-eth1.xml create mode 100644 tests/networkxml2xmlupdatein/interface-eth47.xml create mode 100644 tests/networkxml2xmlupdatein/portgroup-alice-new.xml create mode 100644 tests/networkxml2xmlupdatein/portgroup-alison.xml create mode 100644 tests/networkxml2xmlupdatein/srv-record-donkey.xml create mode 100644 tests/networkxml2xmlupdatein/srv-record-invalid.xml create mode 100644 tests/networkxml2xmlupdatein/srv-record-protocol.xml create mode 100644 tests/networkxml2xmlupdatein/srv-record-service.xml create mode 100644 tests/networkxml2xmlupdatein/srv-record.xml create mode 100644 tests/networkxml2xmlupdatein/unparsable-dns-host.xml create mode 100644 tests/networkxml2xmlupdateout/dhcp6host-routed-network-another-range.xml create mode 100644 tests/networkxml2xmlupdateout/dhcp6host-routed-network-range.xml create mode 100644 tests/networkxml2xmlupdateout/nat-network-dns-more-hosts.xml create mode 100644 tests/networkxml2xmlupdateout/nat-network-dns-srv-record.xml create mode 100644 tests/networkxml2xmlupdateout/nat-network-dns-srv-records.xml create mode 100644 tests/networkxml2xmlupdateout/nat-network-dns-txt-none.xml create mode 100644 tests/networkxml2xmlupdateout/nat-network-dns-txt-records.xml create mode 100644 tests/networkxml2xmlupdateout/nat-network-forward-ifaces.xml create mode 100644 tests/networkxml2xmlupdateout/nat-network-host-updated.xml create mode 100644 tests/networkxml2xmlupdateout/nat-network-hosts.xml create mode 100644 tests/networkxml2xmlupdateout/nat-network-no-forward-ifaces.xml create mode 100644 tests/networkxml2xmlupdateout/nat-network-no-hosts.xml create mode 100644 tests/networkxml2xmlupdateout/nat-network-no-range.xml create mode 100644 tests/networkxml2xmlupdateout/nat-network-one-host.xml create mode 100644 tests/networkxml2xmlupdateout/nat-network.xml create mode 100644 tests/networkxml2xmlupdateout/openvswitch-net-modified.xml create mode 100644 tests/networkxml2xmlupdateout/openvswitch-net-more-portgroups.xml create mode 100644 tests/networkxml2xmlupdateout/openvswitch-net-without-alice.xml create mode 100644 tests/networkxml2xmlupdatetest.c -- 1.8.1.5

2 5

[libvirt] [PATCH 0/4] Plug some virsh leaks
by Ján Tomko 28 Aug '13

28 Aug '13

https://bugzilla.redhat.com/show_bug.cgi?id=1001536 Ján Tomko (4): virsh: free messages after logging them to a file virsh: free the list from ListAll APIs even for 0 items virsh: free the formatting string when listing pool details virsh: free the caps list properly if one of them is invalid tools/virsh-interface.c | 2 +- tools/virsh-network.c | 2 +- tools/virsh-nodedev.c | 6 +++--- tools/virsh-nwfilter.c | 2 +- tools/virsh-pool.c | 3 ++- tools/virsh-secret.c | 2 +- tools/virsh.c | 1 + 7 files changed, 10 insertions(+), 8 deletions(-) -- 1.8.1.5

2 6

[libvirt] [PATCH] qemu: Add missing VIR_DOMAIN_BLOCK_COMMIT_DELETE flags
by Alex Jia 28 Aug '13

28 Aug '13

The flag "VIR_DOMAIN_BLOCK_COMMIT_DELETE" is missed by qemuDomainBlockCommit(), and then will hit error "unsupported flags (0x2) in function qemuDomainBlockCommit" if users run 'virsh blockcommit' with '--delete' option. RHBZ: https://bugzilla.redhat.com/show_bug.cgi?id=1001475 Signed-off-by: Alex Jia <ajia(a)redhat.com> --- src/qemu/qemu_driver.c | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index ed29373..8863124 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -14444,7 +14444,8 @@ qemuDomainBlockCommit(virDomainPtr dom, const char *path, const char *base, const char *base_canon = NULL; bool clean_access = false; - virCheckFlags(VIR_DOMAIN_BLOCK_COMMIT_SHALLOW, -1); + virCheckFlags(VIR_DOMAIN_BLOCK_COMMIT_SHALLOW | + VIR_DOMAIN_BLOCK_COMMIT_DELETE, -1); if (!(vm = qemuDomObjFromDomain(dom))) goto cleanup; -- 1.7.1

4 8

[libvirt] How about realize domainSetBlockIoTune by cgroup?
by hzguanqiang＠corp.netease.com 28 Aug '13

28 Aug '13

Hello experts, It seems that blkiotune of kvm domain is set through qemu now, and We found there are some throttle shaking problem when using this interface. As you can do the same thing through blkio cgroup, and this sounds to show better performance, Can we realize domainSetBlockIoTune by cgroup just like interface domainSetBlkioParameters ? Or Add a new interface to support setting disk IOtune by blkio cgroup? Thanks ------------------ Best regards! GuanQiang 2013-08-27

3 6

[libvirt] [PATCH] Use loop-control to allocate loopback device.
by Ian Main 27 Aug '13

27 Aug '13

This patch changes virFileLoopDeviceOpen() to use the new loop-control device to allocate a new loop device. If this behavior is unsupported or an error occurs while trying to do this it falls back to the previous method of searching /dev for a free device. With this patch you can start as many image based LXC domains as you like (well almost). Fixes bug https://bugzilla.redhat.com/show_bug.cgi?id=995543 Signed-off-by: Ian Main <imain(a)redhat.com> --- configure.ac | 12 +++++++++++ src/util/virfile.c | 61 +++++++++++++++++++++++++++++++++++++++++++++++++++++- 2 files changed, 72 insertions(+), 1 deletion(-) diff --git a/configure.ac b/configure.ac index ac8cfa1..10cd872 100644 --- a/configure.ac +++ b/configure.ac @@ -913,6 +913,18 @@ if test "$with_lxc" = "yes" || test "$with_lxc" = "check"; then AC_MSG_ERROR([Required kernel features for LXC were not found]) fi ]) + AC_LINK_IFELSE([AC_LANG_PROGRAM( + [[ + #include <sched.h> + #include <linux/loop.h> + #include <sys/epoll.h> + ]], [[ + unshare(!(LOOP_CTL_GET_FREE)); + ]])], [ + AC_DEFINE([HAVE_DECL_LOOP_CTL_GET_FREE], [1], + [Define to 1 if you have the declaration of `LOOP_CTL_GET_FREE', + and to 0 if you don't.]) + ]) fi if test "$with_lxc" = "yes" ; then AC_DEFINE_UNQUOTED([WITH_LXC], 1, [whether LXC driver is enabled]) diff --git a/src/util/virfile.c b/src/util/virfile.c index 2b07ac9..f70f9fe 100644 --- a/src/util/virfile.c +++ b/src/util/virfile.c @@ -528,7 +528,47 @@ int virFileUpdatePerm(const char *path, #if defined(__linux__) && HAVE_DECL_LO_FLAGS_AUTOCLEAR -static int virFileLoopDeviceOpen(char **dev_name) + +#if HAVE_DECL_LOOP_CTL_GET_FREE +static int virFileLoopDeviceOpenLoopCtl(char **dev_name) +{ + int fd = -1; + int devnr; + char *looppath = NULL; + + VIR_DEBUG("Opening loop-control device"); + if ((fd = open("/dev/loop-control", O_RDWR)) < 0) { + virReportSystemError(errno, "%s", + _("Unable to open /dev/loop-control")); + return -1; + } + + if ((devnr = ioctl(fd, LOOP_CTL_GET_FREE)) < 0) { + virReportSystemError(errno, "%s", + _("Unable to get free loop device via ioctl")); + close(fd); + return -1; + } + close(fd); + + VIR_DEBUG("Found free loop device number %i", devnr); + + if (virAsprintf(&looppath, "/dev/loop%i", devnr) < 0) + return -1; + + if ((fd = open(looppath, O_RDWR)) < 0) { + virReportSystemError(errno, + _("Unable to open %s"), looppath); + VIR_FREE(looppath); + return -1; + } + + *dev_name = looppath; + return fd; +} +#endif /* HAVE_DECL_LOOP_CTL_GET_FREE */ + +static int virFileLoopDeviceOpenSearch(char **dev_name) { int fd = -1; DIR *dh = NULL; @@ -601,6 +641,25 @@ cleanup: return fd; } +static int virFileLoopDeviceOpen(char **dev_name) +{ + int loop_fd; + +#ifdef HAVE_DECL_LOOP_CTL_GET_FREE + loop_fd = virFileLoopDeviceOpenLoopCtl(dev_name); + VIR_DEBUG("Return from loop-control got fd %d\n", loop_fd); + if (loop_fd < 0) { + VIR_WARN("loop-control allocation failed, trying search technique."); + } else { + return loop_fd; + } +#endif /* HAVE_DECL_LOOP_CTL_GET_FREE */ + + /* Without the loop control device we just use the old technique. */ + loop_fd = virFileLoopDeviceOpenSearch(dev_name); + + return loop_fd; +} int virFileLoopDeviceAssociate(const char *file, char **dev) -- 1.8.1.4

2 3