- Devel - libvirt List Archives

[libvirt] [PATCH] build: avoid compiler warning
by Eric Blake 03 May '10

03 May '10

Necessary on cygwin, where uid_t and gid_t are 4-byte long rather than int, causing gcc -Wformat warnings. * src/util/util.c (virFileOperationNoFork, virDirCreateNoFork) (virFileOperation, virDirCreate, virGetUserEnt): Cast uid_t and gid_t before passing to printf. * .gitignore: Ignore cygwin executables. --- .gitignore | 1 + src/util/util.c | 21 ++++++++++++--------- 2 files changed, 13 insertions(+), 9 deletions(-) diff --git a/.gitignore b/.gitignore index 8c275f4..4faf384 100644 --- a/.gitignore +++ b/.gitignore @@ -1,6 +1,7 @@ *#*# *.#*# *.a +*.exe *.o *.orig *.rej diff --git a/src/util/util.c b/src/util/util.c index 3209185..0c5a22b 100644 --- a/src/util/util.c +++ b/src/util/util.c @@ -74,11 +74,14 @@ #include "util.h" #include "memory.h" #include "threads.h" +#include "verify.h" #ifndef NSIG # define NSIG 32 #endif +verify(sizeof (gid_t) <= sizeof (int) && sizeof (uid_t) <= sizeof (int)); + #define VIR_FROM_THIS VIR_FROM_NONE #define virUtilError(code, ...) \ @@ -1277,7 +1280,7 @@ static int virFileOperationNoFork(const char *path, int openflags, mode_t mode, && (fchown(fd, uid, gid) < 0)) { ret = errno; virReportSystemError(errno, _("cannot chown '%s' to (%u, %u)"), - path, uid, gid); + path, (int) uid, (int) gid); goto error; } if ((flags & VIR_FILE_OP_FORCE_PERMS) @@ -1328,7 +1331,7 @@ static int virDirCreateNoFork(const char *path, mode_t mode, uid_t uid, gid_t gi && (chown(path, uid, gid) < 0)) { ret = errno; virReportSystemError(errno, _("cannot chown '%s' to (%u, %u)"), - path, uid, gid); + path, (int) uid, (int) gid); goto error; } if ((flags & VIR_DIR_CREATE_FORCE_PERMS) @@ -1407,14 +1410,14 @@ parenterror: ret = errno; virReportSystemError(errno, _("cannot set gid %u creating '%s'"), - gid, path); + (int) gid, path); goto childerror; } if ((uid != 0) && (setuid(uid) != 0)) { ret = errno; virReportSystemError(errno, _("cannot set uid %u creating '%s'"), - uid, path); + (int) uid, path); goto childerror; } if ((fd = open(path, openflags, mode)) < 0) { @@ -1436,7 +1439,7 @@ parenterror: && (fchown(fd, -1, gid) < 0)) { ret = errno; virReportSystemError(errno, _("cannot chown '%s' to (%u, %u)"), - path, uid, gid); + path, (int) uid, (int) gid); goto childerror; } if ((flags & VIR_FILE_OP_FORCE_PERMS) @@ -1517,13 +1520,13 @@ parenterror: if ((gid != 0) && (setgid(gid) != 0)) { ret = errno; virReportSystemError(errno, _("cannot set gid %u creating '%s'"), - gid, path); + (int) gid, path); goto childerror; } if ((uid != 0) && (setuid(uid) != 0)) { ret = errno; virReportSystemError(errno, _("cannot set uid %u creating '%s'"), - uid, path); + (int) uid, path); goto childerror; } if (mkdir(path, mode) < 0) { @@ -1547,7 +1550,7 @@ parenterror: ret = errno; virReportSystemError(errno, _("cannot chown '%s' to group %u"), - path, gid); + path, (int) gid); goto childerror; } if ((flags & VIR_DIR_CREATE_FORCE_PERMS) @@ -2564,7 +2567,7 @@ static char *virGetUserEnt(uid_t uid, if (getpwuid_r(uid, &pwbuf, strbuf, strbuflen, &pw) != 0 || pw == NULL) { virReportSystemError(errno, _("Failed to find user record for uid '%d'"), - uid); + (int) uid); VIR_FREE(strbuf); return NULL; } -- 1.6.6.1

2 5

[libvirt] [PATCH] dnsmasqReload: avoid mingw link failure
by Eric Blake 03 May '10

03 May '10

* src/util/dnsmasq.c (dnsmasqReload): Mingw lacks kill, but is not running a dnsmasq daemon either. --- I'm not as familiar with the mingw cross-build setup, but I finally got enough pieces installed on my F-12 machine that this was the only remaining compilation failure that I ran into. Let me know if we need an alternate patch that solves dnsmasq more gracefully than just crippling one function. src/util/dnsmasq.c | 4 +++- 1 files changed, 3 insertions(+), 1 deletions(-) diff --git a/src/util/dnsmasq.c b/src/util/dnsmasq.c index 1cb5f21..d6cef40 100644 --- a/src/util/dnsmasq.c +++ b/src/util/dnsmasq.c @@ -328,14 +328,16 @@ dnsmasqDelete(const dnsmasqContext *ctx) * Reloads all the configurations associated to a context */ int -dnsmasqReload(pid_t pid) +dnsmasqReload(pid_t pid ATTRIBUTE_UNUSED) { +#ifndef __MINGW32__ if (kill(pid, SIGHUP) != 0) { virReportSystemError(errno, _("Failed to make dnsmasq (PID: %d) reload config files.\n"), pid); return -1; } +#endif /* __MINGW32__ */ return 0; } -- 1.6.6.1

2 5

[libvirt] [PATCH] nwfilter: pass info where request stems from to have rules applied
by Stefan Berger 03 May '10

03 May '10

For the decision on whether to instantiate the rules, the check for a pending IP address learn request is not sufficient since then only the thread could instantiate the rules. So, a boolean needs to be passed when the thread instantiates the filter rules late and the IP address learn request is still pending in order to override the check for the pending learn request. If the rules are to be updated while the thread is active, this will not be done immediately but the thread will do that later on. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com> --- src/nwfilter/nwfilter_gentech_driver.c | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-) Index: libvirt-acl/src/nwfilter/nwfilter_gentech_driver.c =================================================================== --- libvirt-acl.orig/src/nwfilter/nwfilter_gentech_driver.c +++ libvirt-acl/src/nwfilter/nwfilter_gentech_driver.c @@ -543,6 +543,8 @@ virNWFilterRuleInstancesToArray(int nEnt * @ifname: The name of the interface to apply the rules to * @vars: A map holding variable names and values used for instantiating * the filter and its subfilters. + * @forceWithPendingReq: Ignore the check whether a pending learn request + * is active; 'true' only when the rules are applied late * * Returns 0 on success, a value otherwise. * @@ -563,7 +565,8 @@ virNWFilterInstantiate(virConnectPtr con enum instCase useNewFilter, int *foundNewFilter, bool teardownOld, const unsigned char *macaddr, - virNWFilterDriverStatePtr driver) + virNWFilterDriverStatePtr driver, + bool forceWithPendingReq) { int rc; int j, nptrs; @@ -610,7 +613,8 @@ virNWFilterInstantiate(virConnectPtr con } else if (virHashSize(missing_vars->hashTable) > 1) { rc = 1; goto err_exit; - } else if (virNWFilterLookupLearnReq(ifindex) == NULL) { + } else if (!forceWithPendingReq && + virNWFilterLookupLearnReq(ifindex) != NULL) { goto err_exit; } @@ -688,7 +692,8 @@ __virNWFilterInstantiateFilter(virConnec const char *filtername, virNWFilterHashTablePtr filterparams, enum instCase useNewFilter, - virNWFilterDriverStatePtr driver) + virNWFilterDriverStatePtr driver, + bool forceWithPendingReq) { int rc; const char *drvname = EBIPTABLES_DRIVER_ID; @@ -789,7 +794,8 @@ __virNWFilterInstantiateFilter(virConnec useNewFilter, &foundNewFilter, teardownOld, macaddr, - driver); + driver, + forceWithPendingReq); virNWFilterHashTableFree(vars); @@ -830,7 +836,8 @@ _virNWFilterInstantiateFilter(virConnect net->filter, net->filterparams, useNewFilter, - conn->nwfilterPrivateData); + conn->nwfilterPrivateData, + false); } @@ -856,7 +863,8 @@ virNWFilterInstantiateFilterLate(virConn filtername, filterparams, INSTANTIATE_ALWAYS, - driver); + driver, + true); if (rc) { //something went wrong... 'DOWN' the interface if (ifaceCheck(false, ifname, NULL, ifindex) != 0 ||

2 1

[libvirt] [PATCH] cgroup: Fix possible memory leak in virCgroupMakeGroup
by Ryota Ozaki 03 May '10

03 May '10

* src/util/cgroup.c: free temporal path string before breaking loop --- src/util/cgroup.c | 4 +++- 1 files changed, 3 insertions(+), 1 deletions(-) diff --git a/src/util/cgroup.c b/src/util/cgroup.c index b649c3c..b8b2eb5 100644 --- a/src/util/cgroup.c +++ b/src/util/cgroup.c @@ -472,8 +472,10 @@ static int virCgroupMakeGroup(virCgroupPtr parent, virCgroupPtr group, int creat (i == VIR_CGROUP_CONTROLLER_CPUSET || STREQ(group->controllers[i].mountPoint, group->controllers[VIR_CGROUP_CONTROLLER_CPUSET].mountPoint))) { rc = virCgroupCpuSetInherit(parent, group); - if (rc != 0) + if (rc != 0) { + VIR_FREE(path); break; + } } } -- 1.6.5.2

2 1

[libvirt] [PATCH] dnsmasq.c: Fix OOM error reporting
by Matthias Bolte 03 May '10

03 May '10

Also do some indentation clean up. --- src/util/dnsmasq.c | 28 ++++++++++++++++------------ 1 files changed, 16 insertions(+), 12 deletions(-) diff --git a/src/util/dnsmasq.c b/src/util/dnsmasq.c index 1cb5f21..2d57884 100644 --- a/src/util/dnsmasq.c +++ b/src/util/dnsmasq.c @@ -84,12 +84,12 @@ hostsfileAdd(dnsmasqHostsfile *hostsfile, if (name) { if (virAsprintf(&hostsfile->hosts[hostsfile->nhosts].host, "%s,%s,%s", - mac, ip, name) < 0) { + mac, ip, name) < 0) { goto alloc_error; } } else { if (virAsprintf(&hostsfile->hosts[hostsfile->nhosts].host, "%s,%s", - mac, ip) < 0) { + mac, ip) < 0) { goto alloc_error; } } @@ -99,9 +99,8 @@ hostsfileAdd(dnsmasqHostsfile *hostsfile, return 0; alloc_error: - virReportSystemError(ENOMEM, - _("Failed to add dhcp host entry: mac=%s, ip=%s, name=%s\n"), - mac, ip, (name ? name : "(null)")); + virReportOOMError(); + return -1; } @@ -112,20 +111,23 @@ hostsfileNew(const char *name, int err; dnsmasqHostsfile *hostsfile; - if (VIR_ALLOC(hostsfile) < 0) + if (VIR_ALLOC(hostsfile) < 0) { + virReportOOMError(); return NULL; + } hostsfile->hosts = NULL; hostsfile->nhosts = 0; if (virAsprintf(&hostsfile->path, "%s/%s.%s", config_dir, name, - DNSMASQ_HOSTSFILE_SUFFIX) < 0) { + DNSMASQ_HOSTSFILE_SUFFIX) < 0) { + virReportOOMError(); goto error; } if ((err = virFileMakePath(config_dir))) { virReportSystemError(err, _("cannot create config directory '%s'"), - config_dir); + config_dir); goto error; } @@ -201,11 +203,11 @@ static int hostsfileSave(dnsmasqHostsfile *hostsfile) { int err = hostsfileWrite(hostsfile->path, hostsfile->hosts, - hostsfile->nhosts); + hostsfile->nhosts); if (err < 0) { virReportSystemError(err, _("cannot write config file '%s'"), - hostsfile->path); + hostsfile->path); return -1; } @@ -220,7 +222,7 @@ hostsfileDelete(dnsmasqHostsfile *hostsfile) if (unlink(hostsfile->path) < 0) { virReportSystemError(errno, _("cannot remove config file '%s'"), - hostsfile->path); + hostsfile->path); return -1; } @@ -240,8 +242,10 @@ dnsmasqContextNew(const char *network_name, { dnsmasqContext *ctx; - if (VIR_ALLOC(ctx) < 0) + if (VIR_ALLOC(ctx) < 0) { + virReportOOMError(); return NULL; + } if (!(ctx->hostsfile = hostsfileNew(network_name, config_dir))) goto error; -- 1.7.0.4

2 2

[libvirt] [PATCH] mingw: Fix symbol export
by Matthias Bolte 03 May '10

03 May '10

In commit 98fb83ce25f46e0236706fa6e0943032d921935f I changed the version script handling. But it seems that I didn't test this properly and broke it. The .def file is passed to the compiler directly, but it should get passed to the linker instead. Set VERSION_SCRIPT_FLAGS to -Wl, to pass the .def file correctly to the linker. This fixes the undefined symbol errors while linking virsh. --- configure.ac | 7 ++++--- 1 files changed, 4 insertions(+), 3 deletions(-) diff --git a/configure.ac b/configure.ac index f25e173..6ee5b90 100644 --- a/configure.ac +++ b/configure.ac @@ -78,7 +78,6 @@ AM_PROG_LD VERSION_SCRIPT_FLAGS=-Wl,--version-script= `$LD --help 2>&1 | grep -- --version-script >/dev/null` || \ VERSION_SCRIPT_FLAGS="-Wl,-M -Wl," -AC_SUBST(VERSION_SCRIPT_FLAGS) LIBVIRT_COMPILE_WARNINGS([maximum]) @@ -1824,10 +1823,11 @@ case "$host" in # from libvirt.syms and passes libvirt.def instead of libvirt.syms to the linker LIBVIRT_SYMBOL_FILE=libvirt.def # mingw's ld has the --version-script parameter, but it requires a .def file - # instead to work properly, therefore clear --version-script here + # instead to work properly, therefore clear --version-script here and use + # -Wl, to pass the .def file to the linker # cygwin's ld has the --version-script parameter too, but for some reason # it's working there as expected - VERSION_SCRIPT_FLAGS= + VERSION_SCRIPT_FLAGS="-Wl," ;; esac AC_SUBST([CYGWIN_EXTRA_LDFLAGS]) @@ -1836,6 +1836,7 @@ AC_SUBST([CYGWIN_EXTRA_PYTHON_LIBADD]) AC_SUBST([MINGW_EXTRA_LDFLAGS]) AC_SUBST([WIN32_EXTRA_CFLAGS]) AC_SUBST([LIBVIRT_SYMBOL_FILE]) +AC_SUBST([VERSION_SCRIPT_FLAGS]) dnl Look for windres to build a Windows icon resource. -- 1.7.0.4

2 2

[libvirt] [PATCH] Fix memory leaks in cmdInterfaceEdit and cmdNWFilterEdit.
by Laine Stump 03 May '10

03 May '10

This applies a fix to thos functions similar to that made to cmdEdit in 270895063d1cf86ab42fa220a8d090c24d58dbc0, thus fnixing a memory leak - if tmp is unlinked and NULLed early in the function, the memory used by tmp is never freed. Since we will always unlink tmp prior to freeing its memory at the end of the function, just remove the earlier code and let cleanup: do the cleanup. --- tools/virsh.c | 6 ------ 1 files changed, 0 insertions(+), 6 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index eb11a78..fb66a72 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -3248,9 +3248,6 @@ cmdInterfaceEdit (vshControl *ctl, const vshCmd *cmd) doc_edited = editReadBackFile (ctl, tmp); if (!doc_edited) goto cleanup; - unlink (tmp); - tmp = NULL; - /* Compare original XML with edited. Has it changed at all? */ if (STREQ (doc, doc_edited)) { vshPrint (ctl, _("Interface %s XML configuration not changed.\n"), @@ -4193,9 +4190,6 @@ cmdNWFilterEdit (vshControl *ctl, const vshCmd *cmd) doc_edited = editReadBackFile (ctl, tmp); if (!doc_edited) goto cleanup; - unlink (tmp); - tmp = NULL; - /* Compare original XML with edited. Has it changed at all? */ if (STREQ (doc, doc_edited)) { vshPrint (ctl, _("Network filter %s XML configuration not changed.\n"), -- 1.6.6.1

3 3

[libvirt] [PATCH] lxc: Fix failure on starting a domain that has multiple network interfaces
by Ryota Ozaki 03 May '10

03 May '10

[Error message] error: Failed to start domain lxc_test1 error: internal error Failed to create veth device pair: 512 The reason of the failure is that lxc driver unexpectedly re-uses an auto-assigned veth name and tries to create the created veth again. The failure will happen when a domain has multiple network interfaces and the names of those are not specified in XML. The patch fixes the problem by resetting buffers of veth names in every iteration of creating veth. * src/lxc/lxc_driver.c: prevent re-using auto-assigned veth name Reported by Kumar L Srikanth-B22348. --- src/lxc/lxc_driver.c | 5 +++-- 1 files changed, 3 insertions(+), 2 deletions(-) diff --git a/src/lxc/lxc_driver.c b/src/lxc/lxc_driver.c index 409b1cf..554bf66 100644 --- a/src/lxc/lxc_driver.c +++ b/src/lxc/lxc_driver.c @@ -786,14 +786,15 @@ static int lxcSetupInterfaces(virConnectPtr conn, { int rc = -1, i; char *bridge = NULL; - char parentVeth[PATH_MAX] = ""; - char containerVeth[PATH_MAX] = ""; brControl *brctl = NULL; if (brInit(&brctl) != 0) return -1; for (i = 0 ; i < def->nnets ; i++) { + char parentVeth[PATH_MAX] = ""; + char containerVeth[PATH_MAX] = ""; + switch (def->nets[i]->type) { case VIR_DOMAIN_NET_TYPE_NETWORK: { -- 1.6.5.2

2 1

[libvirt] Creating multiple network interfaces in libvirt Domain.
by Kumar L Srikanth-B22348 01 May '10

01 May '10

Hi, I want to create a Domain with two interfaces. I am using LXC hypervisor in the libvirt. My domain XML is shown below: <domain type='lxc' id='1'> <name>srikanth_vm2</name> <memory>500000</memory> <os> <type>exe</type> <init>/bin/sh</init> </os> <vcpu>1</vcpu> <clock offset='utc'/> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>destroy</on_crash> <devices> <emulator>/usr/libexec/libvirt_lxc</emulator> <filesystem type='mount'> <source dir='/root/lxc/fedora1'/> <target dir='/'/> </filesystem> <interface type='network'> <source network='default'/> </interface> <interface type='network'> <source network='default'/> </interface> <console type='pty' /> </devices> </domain> I am able to define the Domain. But, I am not able to start the Domain. While starting the domain, I am getting the following error: error: Failed to start domain srikanth_vm2 error: internal error Failed to create veth device pair: 512 Can you please let me know where I am going wrong? Regards, Srikanth.

2 2

[libvirt] [PATCH] Fix a virsh edit memory leak
by Chris Lalancette 01 May '10

01 May '10

When running virsh edit, we are unlinking and setting the tmp variable to NULL before going to the end of the function, meaning that we never free tmp. Since the exit to the function will always unlink and free tmp, just remove this bit of code and let it get done at the end. Signed-off-by: Chris Lalancette <clalance(a)redhat.com> --- tools/virsh.c | 3 --- 1 files changed, 0 insertions(+), 3 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 04b47a1..ab099da 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -8066,9 +8066,6 @@ cmdEdit (vshControl *ctl, const vshCmd *cmd) doc_edited = editReadBackFile (ctl, tmp); if (!doc_edited) goto cleanup; - unlink (tmp); - tmp = NULL; - /* Compare original XML with edited. Has it changed at all? */ if (STREQ (doc, doc_edited)) { vshPrint (ctl, _("Domain %s XML configuration not changed.\n"), -- 1.6.6.1

3 2