- Devel - libvirt List Archives

[libvirt] [PATCH] qemu: Add some debugging at domain startup
by Cole Robinson 12 Mar '10

12 Mar '10

Signed-off-by: Cole Robinson <crobinso(a)redhat.com> --- src/qemu/qemu_driver.c | 24 +++++++++++++++++++++++- 1 files changed, 23 insertions(+), 1 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index f8ab545..26b5600 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -2695,12 +2695,15 @@ static int qemudStartVMDaemon(virConnectPtr conn, FD_ZERO(&keepfd); + DEBUG0("Beginning VM startup process"); + if (virDomainObjIsActive(vm)) { qemuReportError(VIR_ERR_OPERATION_INVALID, "%s", _("VM is already active")); return -1; } + DEBUG0("Generating domain security label (if required)"); /* If you are using a SecurityDriver with dynamic labelling, then generate a security label for isolation */ if (driver->securityDriver && @@ -2708,17 +2711,21 @@ static int qemudStartVMDaemon(virConnectPtr conn, driver->securityDriver->domainGenSecurityLabel(vm) < 0) return -1; + DEBUG0("Generating setting domain security labels (if required)"); if (driver->securityDriver && driver->securityDriver->domainSetSecurityAllLabel && driver->securityDriver->domainSetSecurityAllLabel(vm) < 0) goto cleanup; - /* Ensure no historical cgroup for this VM is lieing around bogus settings */ + /* Ensure no historical cgroup for this VM is lieing around bogus + * settings */ + DEBUG0("Removing old cgroup (if required)"); qemuRemoveCgroup(driver, vm, 1); if ((vm->def->ngraphics == 1) && vm->def->graphics[0]->type == VIR_DOMAIN_GRAPHICS_TYPE_VNC && vm->def->graphics[0]->data.vnc.autoport) { + DEBUG0("Determining VNC port"); int port = qemudNextFreeVNCPort(driver); if (port < 0) { qemuReportError(VIR_ERR_INTERNAL_ERROR, @@ -2735,6 +2742,7 @@ static int qemudStartVMDaemon(virConnectPtr conn, goto cleanup; } + DEBUG0("Creating domain log file"); if ((logfile = qemudLogFD(driver, vm->def->name)) < 0) goto cleanup; @@ -2751,14 +2759,17 @@ static int qemudStartVMDaemon(virConnectPtr conn, goto cleanup; } + DEBUG0("Determing emulator version"); if (qemudExtractVersionInfo(emulator, NULL, &qemuCmdFlags) < 0) goto cleanup; + DEBUG0("Setting up domain cgroup (if required)"); if (qemuSetupCgroup(driver, vm) < 0) goto cleanup; + DEBUG0("Preparing host devices"); if (qemuPrepareHostDevices(driver, vm->def) < 0) goto cleanup; @@ -2767,6 +2778,7 @@ static int qemudStartVMDaemon(virConnectPtr conn, goto cleanup; } + DEBUG0("Preparing monitor state"); if (qemuPrepareMonitorChr(driver, priv->monConfig, vm->def->name) < 0) goto cleanup; @@ -2798,6 +2810,7 @@ static int qemudStartVMDaemon(virConnectPtr conn, * use in hotplug */ if (qemuCmdFlags & QEMUD_CMD_FLAG_DEVICE) { + DEBUG0("Assigning domain PCI addresses"); /* Populate cache with current addresses */ if (priv->pciaddrs) { qemuDomainPCIAddressSetFree(priv->pciaddrs); @@ -2816,6 +2829,7 @@ static int qemudStartVMDaemon(virConnectPtr conn, priv->persistentAddrs = 0; } + DEBUG0("Building emulator command line"); vm->def->id = driver->nextvmid++; if (qemudBuildCommandLine(conn, driver, vm->def, priv->monConfig, priv->monJSON, qemuCmdFlags, &argv, &progenv, @@ -2899,25 +2913,31 @@ static int qemudStartVMDaemon(virConnectPtr conn, if (ret == -1) /* The VM failed to start */ goto cleanup; + DEBUG0("Waiting for monitor to show up"); if (qemudWaitForMonitor(driver, vm, pos) < 0) goto abort; + DEBUG0("Detecting VCPU PIDs"); if (qemuDetectVcpuPIDs(driver, vm) < 0) goto abort; + DEBUG0("Setting CPU affinity"); if (qemudInitCpuAffinity(vm) < 0) goto abort; + DEBUG0("Setting any required VM passwords"); if (qemuInitPasswords(conn, driver, vm, qemuCmdFlags) < 0) goto abort; /* If we have -device, then addresses are assigned explicitly. * If not, then we have to detect dynamic ones here */ if (!(qemuCmdFlags & QEMUD_CMD_FLAG_DEVICE)) { + DEBUG0("Determining domain device PCI addresses"); if (qemuInitPCIAddresses(driver, vm) < 0) goto abort; } + DEBUG0("Setting initial memory amount"); qemuDomainObjEnterMonitorWithDriver(driver, vm); if (qemuMonitorSetBalloon(priv->mon, vm->def->memory) < 0) { qemuDomainObjExitMonitorWithDriver(driver, vm); @@ -2925,6 +2945,7 @@ static int qemudStartVMDaemon(virConnectPtr conn, } if (migrateFrom == NULL) { + DEBUG0("Starting domain CPUs"); /* Allow the CPUS to start executing */ if (qemuMonitorStartCPUs(priv->mon, conn) < 0) { if (virGetLastError() == NULL) @@ -2937,6 +2958,7 @@ static int qemudStartVMDaemon(virConnectPtr conn, qemuDomainObjExitMonitorWithDriver(driver, vm); + DEBUG0("Writing domain status to disk"); if (virDomainSaveStatus(driver->caps, driver->stateDir, vm) < 0) goto abort; -- 1.6.6

2 1

Re: [libvirt] Request to mailing list libvir-list rejected
by Eric Blake 12 Mar '10

12 Mar '10

[in general, technical lists tend to frown on the practice of top-posting] On 03/10/2010 02:49 PM, Dev.Atom wrote: > Sorry, I'm not used to use mailing list > > I think the relevants part are these functions : > > int virFileOperation(const char *path, int openflags, mode_t mode, > uid_t uid, gid_t gid, > virFileOperationHook hook, void *hookdata, > unsigned int flags) > __attribute__((__warn_unused_result__)); > > int virDirCreate(const char *path, mode_t mode, uid_t uid, gid_t gid, > unsigned int flags) > __attribute__((__warn_unused_result__)); Yes, my request was not to dump the entire (750k) preprocessor output to the list, but to just the relevant portion of the output to make sure there weren't any macros interfering with the parse. > > Matthias say that he has patches which will be ready in the week, and I > can wait for these patches Matthias probably already hit it on the head - gnulib can guarantee that uid_t is defined in spite of mingw not providing it, but I haven't personally checked whether we are using this aspect of gnulib yet, and I will defer to see Matthias' patches. -- Eric Blake eblake(a)redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org

2 2

[libvirt] [PATCH] Only use the numa functions when they are available.
by Chris Lalancette 12 Mar '10

12 Mar '10

Signed-off-by: Chris Lalancette <clalance(a)redhat.com> --- src/nodeinfo.c | 8 +++++--- 1 files changed, 5 insertions(+), 3 deletions(-) diff --git a/src/nodeinfo.c b/src/nodeinfo.c index 8d7e055..bf57517 100644 --- a/src/nodeinfo.c +++ b/src/nodeinfo.c @@ -160,10 +160,12 @@ int linuxNodeInfoCPUPopulate(virConnectPtr conn, FILE *cpuinfo, nodeinfo->cpus = 0; nodeinfo->mhz = 0; nodeinfo->cores = 1; - if (numa_available() < 0) - nodeinfo->nodes = 1; - else + + nodeinfo->nodes = 1; +#if HAVE_NUMACTL + if (numa_available() >= 0) nodeinfo->nodes = numa_max_node() + 1; +#endif /* NB: It is impossible to fill our nodes, since cpuinfo * has no knowledge of NUMA nodes */ -- 1.6.6.1

2 1

[libvirt] [PATCH] Make nodeGetInfo report the correct number of NUMA nodes.
by Chris Lalancette 12 Mar '10

12 Mar '10

The nodeGetInfo code was always assuming that machine had a single NUMA node, which is not correct. The good news is that libnuma gives us this information pretty easily, so let's properly report it. NOTE: With recent hardware starting to support CPU hot-add and hot-remove, both this code and the nodeCapsInitNUMA() code are quickly going to become obsolete. We'll have to think of a more dynamic solution for dealing with NUMA nodes and CPUs that can come and go at will. Signed-off-by: Chris Lalancette <clalance(a)redhat.com> --- src/nodeinfo.c | 6 +++++- 1 files changed, 5 insertions(+), 1 deletions(-) diff --git a/src/nodeinfo.c b/src/nodeinfo.c index 0748602..8d7e055 100644 --- a/src/nodeinfo.c +++ b/src/nodeinfo.c @@ -159,7 +159,11 @@ int linuxNodeInfoCPUPopulate(virConnectPtr conn, FILE *cpuinfo, nodeinfo->cpus = 0; nodeinfo->mhz = 0; - nodeinfo->nodes = nodeinfo->cores = 1; + nodeinfo->cores = 1; + if (numa_available() < 0) + nodeinfo->nodes = 1; + else + nodeinfo->nodes = numa_max_node() + 1; /* NB: It is impossible to fill our nodes, since cpuinfo * has no knowledge of NUMA nodes */ -- 1.6.6.1

3 5

[libvirt] [PATCH] Allow a user-settable number of VNC autoports.
by Chris Lalancette 12 Mar '10

12 Mar '10

Currently we have a hard-coded maximum of 100 VNC autoports that the qemu driver can use. This may not be enough for running large numbers of guests on larger machines. However, we don't necessarily necessarily want to make it an open-ended number; that could lead to Denial of Service. Allow a user-settable option to control the number of autoports we will assign. The default is still 100, but now users can increase that if they wish. Signed-off-by: Chris Lalancette <clalance(a)redhat.com> --- src/qemu/libvirtd_qemu.aug | 3 +++ src/qemu/qemu.conf | 4 ++++ src/qemu/qemu_conf.c | 6 ++++++ src/qemu/qemu_conf.h | 2 ++ src/qemu/qemu_driver.c | 5 +++-- src/qemu/test_libvirtd_qemu.aug | 14 ++++++++++++++ 6 files changed, 32 insertions(+), 2 deletions(-) diff --git a/src/qemu/libvirtd_qemu.aug b/src/qemu/libvirtd_qemu.aug index 5bd60b3..0dd89ae 100644 --- a/src/qemu/libvirtd_qemu.aug +++ b/src/qemu/libvirtd_qemu.aug @@ -13,11 +13,13 @@ module Libvirtd_qemu = let str_val = del /\"/ "\"" . store /[^\"]*/ . del /\"/ "\"" let bool_val = store /0|1/ + let int_val = store /[0-9]+/ let str_array_element = [ seq "el" . str_val ] . del /[ \t\n]*/ "" let str_array_val = counter "el" . array_start . ( str_array_element . ( array_sep . str_array_element ) * ) ? . array_end let str_entry (kw:string) = [ key kw . value_sep . str_val ] let bool_entry (kw:string) = [ key kw . value_sep . bool_val ] + let int_entry (kw:string) = [ key kw . value_sep . int_val ] let str_array_entry (kw:string) = [ key kw . value_sep . str_array_val ] @@ -29,6 +31,7 @@ module Libvirtd_qemu = | str_entry "vnc_password" | bool_entry "vnc_sasl" | str_entry "vnc_sasl_dir" + | int_entry "vnc_max_connections" | str_entry "security_driver" | str_entry "user" | str_entry "group" diff --git a/src/qemu/qemu.conf b/src/qemu/qemu.conf index 3da332f..edcf083 100644 --- a/src/qemu/qemu.conf +++ b/src/qemu/qemu.conf @@ -79,6 +79,10 @@ # vnc_sasl_dir = "/some/directory/sasl2" +# The maximum number of VNC connections we will allow. Once +# libvirtd has created this many qemu guests with VNC ports, +# no more guests can be started. The default is 100 guests. +# vnc_max_connections = 100 # The default security driver is SELinux. If SELinux is disabled diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c index 40ca221..02186fd 100644 --- a/src/qemu/qemu_conf.c +++ b/src/qemu/qemu_conf.c @@ -113,6 +113,8 @@ int qemudLoadDriverConfig(struct qemud_driver *driver, return -1; } + driver->vncMaxConnections = 100; + #ifdef HAVE_MNTENT_H /* For privileged driver, try and find hugepage mount automatically. * Non-privileged driver requires admin to create a dir for the @@ -211,6 +213,10 @@ int qemudLoadDriverConfig(struct qemud_driver *driver, } } + p = virConfGetValue (conf, "vnc_max_connections"); + CHECK_TYPE ("vnc_max_connections", VIR_CONF_LONG); + if (p) driver->vncMaxConnections = p->l; + p = virConfGetValue (conf, "user"); CHECK_TYPE ("user", VIR_CONF_STRING); if (!(user = strdup(p && p->str ? p->str : QEMU_USER))) { diff --git a/src/qemu/qemu_conf.h b/src/qemu/qemu_conf.h index 6a9de5e..e0c4a9b 100644 --- a/src/qemu/qemu_conf.h +++ b/src/qemu/qemu_conf.h @@ -122,6 +122,8 @@ struct qemud_driver { char *vncListen; char *vncPassword; char *vncSASLdir; + unsigned int vncMaxConnections; + char *hugetlbfs_mount; char *hugepage_path; diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 49983dd..8324075 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -2147,10 +2147,11 @@ qemuInitPCIAddresses(struct qemud_driver *driver, return ret; } -static int qemudNextFreeVNCPort(struct qemud_driver *driver ATTRIBUTE_UNUSED) { +static int qemudNextFreeVNCPort(struct qemud_driver *driver) +{ int i; - for (i = 5900 ; i < 6000 ; i++) { + for (i = 5900 ; i < (5900 + driver->vncMaxConnections) ; i++) { int fd; int reuse = 1; struct sockaddr_in addr; diff --git a/src/qemu/test_libvirtd_qemu.aug b/src/qemu/test_libvirtd_qemu.aug index 2feedc0..3c2190a 100644 --- a/src/qemu/test_libvirtd_qemu.aug +++ b/src/qemu/test_libvirtd_qemu.aug @@ -80,6 +80,13 @@ vnc_sasl = 1 # vnc_sasl_dir = \"/some/directory/sasl2\" + +# The maximum number of VNC connections we will allow. Once +# libvirtd has created this many qemu guests with VNC ports, +# no more guests can be started. The default is 100 guests. +# +vnc_max_connections = 100 + security_driver = \"selinux\" user = \"root\" @@ -180,6 +187,13 @@ relaxed_acs_check = 1 { "#comment" = "" } { "vnc_sasl_dir" = "/some/directory/sasl2" } { "#empty" } +{ "#empty" } +{ "#comment" = "The maximum number of VNC connections we will allow. Once" } +{ "#comment" = "libvirtd has created this many qemu guests with VNC ports," } +{ "#comment" = "no more guests can be started. The default is 100 guests." } +{ "#comment" = "" } +{ "vnc_max_connections" = "100" } +{ "#empty" } { "security_driver" = "selinux" } { "#empty" } { "user" = "root" } -- 1.6.6.1

2 1

[libvirt] virsh stucked in virDomainCreate
by Anthony 12 Mar '10

12 Mar '10

Hi, I'm having an issue when trying to start a (previously created) vbox item : virsh gets stucked in virDomainCreate and I see no error message (LIBVIRT_DEBUG=1). Is there any mean to get more information on what is happening ? Moreover, now that virsh has stalled, running "connect vbox:///session" on another virsh instance gets an error from "secret driver". Is this normal, as if I'm only supposed to have 1 connection at a time ? Thanks Regards Anthony

1 0

[libvirt] cpu values
by Tavares, John 12 Mar '10

12 Mar '10

I am trying to compare the results that I am getting on a RHEL 5.3 server using xm list (using libvirt.so.0.3.3) against both my Dom0 and my Linux DomU on the same server to see if the cpu values match to what the kernel is reporting in /proc/stat. Here is an example of what I am seeing on both: >From Dom0: $ sudo /usr/sbin/xm list;cat /proc/stat | grep cpu Name ID Mem(MiB) VCPUs State Time(s) Domain-0 0 746 1 r----- 73722.3 rhel5_01 1 255 1 -b---- 18558.4 cpu 412542 1477862 522101 1111350509 347838 0 262 487234 cpu0 412542 1477862 522101 1111350509 347838 0 262 487234 >From DomU (rhelt_01): $ cat /proc/stat | grep cpu cpu 138184 148634 246128 924538039 467802 133 178 339822 cpu0 138184 148634 246128 924538039 467802 133 178 339822 The results of the two do not appear close. Should the cpu times match, is this a bug, or does Time mean something different than I think it does?? Thanks. John

2 3

[libvirt] [PATCH] Fix virsh command 'cd'.
by Chris Lalancette 12 Mar '10

12 Mar '10

cmdCd was returning a 0 on success and -1 on error, when the rest of the code expected a TRUE on success and a FALSE on error. Fix the discrepancy. Signed-off-by: Chris Lalancette <clalance(a)redhat.com> --- tools/virsh.c | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index a47edd5..c6e3f2a 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -7469,7 +7469,7 @@ cmdCd(vshControl *ctl, const vshCmd *cmd ATTRIBUTE_UNUSED) if (!ctl->imode) { vshError(ctl, "%s", _("cd: command valid only in interactive mode")); - return -1; + return FALSE; } dir = vshCommandOptString(cmd, "dir", &found); @@ -7482,10 +7482,10 @@ cmdCd(vshControl *ctl, const vshCmd *cmd ATTRIBUTE_UNUSED) if (chdir (dir) == -1) { vshError(ctl, _("cd: %s: %s"), strerror(errno), dir); - return -1; + return FALSE; } - return 0; + return TRUE; } #endif -- 1.6.6.1

3 2

[libvirt] [PATCH] Fix crash in virsh after bogus command.
by Chris Lalancette 12 Mar '10

12 Mar '10

If you ran virsh in interactive mode and ran a command that virsh could not parse, it would then SEGV on subsequent commands. The problem is that we are freeing the vshCmd structure in the syntaxError label at the end of vshCommandParse, but forgetting to set ctl->cmd to NULL. This means that on the next command, we would try to free the same structure again, leading to badness. Make sure to set ctl->cmd to NULL after freeing it. Signed-off-by: Chris Lalancette <clalance(a)redhat.com> --- tools/virsh.c | 4 +++- 1 files changed, 3 insertions(+), 1 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index c6e3f2a..eeaddbc 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -8643,8 +8643,10 @@ vshCommandParse(vshControl *ctl, char *cmdstr) return TRUE; syntaxError: - if (ctl->cmd) + if (ctl->cmd) { vshCommandFree(ctl->cmd); + ctl->cmd = NULL; + } if (first) vshCommandOptFree(first); VIR_FREE(tkdata); -- 1.6.6.1

3 2

[libvirt] [PATCH] Fix compiler warnings in virsh.c
by Laine Stump 12 Mar '10

12 Mar '10

No functional change. These all generated compiler warnings which, for some reason weren't converted to errors by --enable-compiler-warnings=error. * tools/virsh.c: * change return type frmo int to void on two functions that don't return a value. * remove unused variables/labels from two functions * eliminate non-literal format strings * typecast char* into xmlChar* when calling xmlParseBalancedChunkMemory --- tools/virsh.c | 24 ++++++++++-------------- 1 files changed, 10 insertions(+), 14 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index a47edd5..33f3647 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -421,7 +421,7 @@ static void vshCatchDisconnect(int sig, siginfo_t * siginfo, * Catch SIGPIPE signals which may arise when disconnection * from libvirtd occurs */ -static int +static void vshSetupSignals(void) { struct sigaction sig_action; @@ -435,10 +435,10 @@ vshSetupSignals(void) { /* * vshReconnect: * - * Reconnect after an + * Reconnect after a disconnect from libvirtd * */ -static int +static void vshReconnect(vshControl *ctl) { if (ctl->conn != NULL) virConnectClose(ctl->conn); @@ -1896,9 +1896,7 @@ cmdDomjobinfo(vshControl *ctl, const vshCmd *cmd) { virDomainJobInfo info; virDomainPtr dom; - int ret = TRUE, autostart; - unsigned int id; - char *str, uuid[VIR_UUID_STRING_BUFLEN]; + int ret = TRUE; if (!vshConnectionUsability(ctl, ctl->conn, TRUE)) return FALSE; @@ -1980,8 +1978,6 @@ cmdDomjobabort(vshControl *ctl, const vshCmd *cmd) { virDomainPtr dom; int ret = TRUE; - unsigned int id; - char *str, uuid[VIR_UUID_STRING_BUFLEN]; if (!vshConnectionUsability(ctl, ctl->conn, TRUE)) return FALSE; @@ -1992,7 +1988,6 @@ cmdDomjobabort(vshControl *ctl, const vshCmd *cmd) if (virDomainAbortJob(dom) < 0) ret = FALSE; -cleanup: virDomainFree(dom); return ret; } @@ -6705,7 +6700,7 @@ cmdAttachInterface(vshControl *ctl, const vshCmd *cmd) } if (ret != 0) { - vshError(ctl, _("Failed to attach interface")); + vshError(ctl, "%s", _("Failed to attach interface")); ret = FALSE; } else { vshPrint(ctl, "%s", _("Interface attached successfully\n")); @@ -6834,7 +6829,7 @@ cmdDetachInterface(vshControl *ctl, const vshCmd *cmd) } if (ret != 0) { - vshError(ctl, _("Failed to detach interface")); + vshError(ctl, "%s", _("Failed to detach interface")); ret = FALSE; } else { vshPrint(ctl, "%s", _("Interface detached successfully\n")); @@ -7007,7 +7002,7 @@ cmdAttachDisk(vshControl *ctl, const vshCmd *cmd) } if (ret != 0) { - vshError(ctl, _("Failed to attach disk")); + vshError(ctl, "%s", _("Failed to attach disk")); ret = FALSE; } else { vshPrint(ctl, "%s", _("Disk attached successfully\n")); @@ -7128,7 +7123,7 @@ cmdDetachDisk(vshControl *ctl, const vshCmd *cmd) } if (ret != 0) { - vshError(ctl, _("Failed to detach disk")); + vshError(ctl, "%s", _("Failed to detach disk")); ret = FALSE; } else { vshPrint(ctl, "%s", _("Disk detached successfully\n")); @@ -7257,7 +7252,8 @@ cmdCPUBaseline(vshControl *ctl, const vshCmd *cmd) if (doc == NULL) goto no_memory; - res = xmlParseBalancedChunkMemory(doc, NULL, NULL, 0, buffer, &node_list); + res = xmlParseBalancedChunkMemory(doc, NULL, NULL, 0, + (const xmlChar *)buffer, &node_list); if (res != 0) { vshError(ctl, _("Failed to parse XML fragment %s"), from); ret = FALSE; -- 1.7.0.1

2 1