- Devel - libvirt List Archives

[libvirt] (no subject)
by Luis Costa 08 Mar '10

08 Mar '10

-- Luís Carlos Moreira da Costa Eclipse RAP, RCP, GMF, OSGI, Spring-DM and Pentaho Developer Regional Communities/Brazil http://wiki.eclipse.org/Regional_Communities/Brazil Java, write once, run everywhere™

1 0

[libvirt] Fix locking in qemudDomainMemoryStats
by Adam Litke 08 Mar '10

08 Mar '10

When adding domainMemoryStats API support for the qemu driver, I didn't follow the locking rules exactly. The job condition must be held when executing monitor commands. This corrects the segfaults I was seeing when calling domainMemoryStats in a multi-threaded environment. If this patch is accepted, I would also consider it a candidate for the 0.7.6 stable patch stream (if such a thing exists). Thanks. Signed-off-by: Adam Litke <agl(a)us.ibm.com> diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index b8b7916..a9023da 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -7526,6 +7526,9 @@ qemudDomainMemoryStats (virDomainPtr dom, goto cleanup; } + if (qemuDomainObjBeginJob(vm) < 0) + goto cleanup; + if (virDomainObjIsActive(vm)) { qemuDomainObjPrivatePtr priv = vm->privateData; qemuDomainObjEnterMonitor(vm); @@ -7536,6 +7539,9 @@ qemudDomainMemoryStats (virDomainPtr dom, "%s", _("domain is not running")); } + if (qemuDomainObjEndJob(vm) == 0) + vm = NULL; + cleanup: if (vm) virDomainObjUnlock(vm); -- Thanks, Adam

3 2

[libvirt] Unable to start a VM with USB device
by Laurent Léonard 08 Mar '10

08 Mar '10

Hi, I just packaged libvirt 0.7.7 for Debian. But I can't start a VM with an USB device attached: # virsh start test-vm error: Failed to start domain test-vm error: unable to set security context '107:113' on '/dev/bus/usb/000/000': No such file or directory I know there were changes about USB passthrough so I suppose it's a bug. Thank you, -- Laurent Léonard

3 4

[libvirt] [PATCHv2] Eliminate large stack buffer in doTunnelSendAll
by Laine Stump 08 Mar '10

08 Mar '10

doTunnelSendAll function (used by QEMU migration) uses a 64k buffer on the stack, which could be problematic. This patch replaces that with a buffer from the heap. While in the neighborhood, this patch also improves error reporting in the case that saferead fails - previously, virStreamAbort() was called (resetting errno) before reporting the error. It's been changed to report the error first. --- Eric noticed that I was inconsistent in the ordering of cleanup vs. error reporting, so I fixed that in this version. src/qemu/qemu_driver.c | 18 +++++++++++++++--- 1 files changed, 15 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index bb3edde..5a18938 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -8399,19 +8399,28 @@ cleanup: } +#define TUNNEL_SEND_BUF_SIZE 65536 + static int doTunnelSendAll(virStreamPtr st, int sock) { - char buffer[65536]; - int nbytes = sizeof(buffer); + char *buffer; + int nbytes = TUNNEL_SEND_BUF_SIZE; + + if (VIR_ALLOC_N(buffer, TUNNEL_SEND_BUF_SIZE) < 0) { + virReportOOMError(); + virStreamAbort(st); + return -1; + } /* XXX should honour the 'resource' parameter here */ for (;;) { nbytes = saferead(sock, buffer, nbytes); if (nbytes < 0) { - virStreamAbort(st); virReportSystemError(errno, "%s", _("tunnelled migration failed to read from qemu")); + virStreamAbort(st); + VIR_FREE(buffer); return -1; } else if (nbytes == 0) @@ -8421,10 +8430,13 @@ static int doTunnelSendAll(virStreamPtr st, if (virStreamSend(st, buffer, nbytes) < 0) { qemuReportError(VIR_ERR_OPERATION_FAILED, "%s", _("Failed to write migration data to remote libvirtd")); + VIR_FREE(buffer); return -1; } } + VIR_FREE(buffer); + if (virStreamFinish(st) < 0) /* virStreamFinish set the error for us */ return -1; -- 1.6.6.1

2 2

[libvirt] [PATCH 0/5] build: use C99 varargs
by Eric Blake 08 Mar '10

08 Mar '10

3 7

[libvirt] [PATCH] Get thread and socket information in virsh nodeinfo.
by Chris Lalancette 08 Mar '10

08 Mar '10

The current code for "nodeinfo" is pretty naive about socket and thread information. To determine the sockets, it just takes the number of cpus and divides by the number of cores. For the thread count, it always sets it to 1. With more recent Intel machines, however, hyperthreading is again an option, meaning that these heuristics no longer work and give bogus numbers. This patch goes through /sys to get the additional information so we properly report it. Note that I had to edit the tests not to report on socket and thread counts, since these are determined dynamically now. Signed-off-by: Chris Lalancette <clalance(a)redhat.com> --- src/nodeinfo.c | 140 +++++++++++++++++++++++++++++-- tests/nodeinfodata/linux-nodeinfo-1.txt | 2 +- tests/nodeinfodata/linux-nodeinfo-2.txt | 2 +- tests/nodeinfodata/linux-nodeinfo-3.txt | 2 +- tests/nodeinfodata/linux-nodeinfo-4.txt | 2 +- tests/nodeinfodata/linux-nodeinfo-5.txt | 2 +- tests/nodeinfodata/linux-nodeinfo-6.txt | 2 +- tests/nodeinfotest.c | 5 +- 8 files changed, 139 insertions(+), 18 deletions(-) diff --git a/src/nodeinfo.c b/src/nodeinfo.c index 2d44609..18efd28 100644 --- a/src/nodeinfo.c +++ b/src/nodeinfo.c @@ -28,6 +28,7 @@ #include <stdlib.h> #include <stdint.h> #include <errno.h> +#include <dirent.h> #if HAVE_NUMACTL # define NUMA_VERSION1_COMPATIBILITY 1 @@ -55,22 +56,117 @@ #ifdef __linux__ #define CPUINFO_PATH "/proc/cpuinfo" +#define CPU_SYS_PATH "/sys/devices/system/cpu" -/* NB, these are not static as we need to call them from testsuite */ +/* NB, this is not static as we need to call it from the testsuite */ int linuxNodeInfoCPUPopulate(virConnectPtr conn, FILE *cpuinfo, virNodeInfoPtr nodeinfo); -int linuxNodeInfoCPUPopulate(virConnectPtr conn, FILE *cpuinfo, virNodeInfoPtr nodeinfo) { +static int popcnt(char x) +{ + char count; + for (count = 0; x; count++) + x &= x-1; + return count; +} + +static unsigned long count_thread_siblings(int cpu) +{ + unsigned long ret = 0; + char *path = NULL; + FILE *pathfp = NULL; + char str[1024]; + int i; + + if (virAsprintf(&path, "%s/cpu%d/topology/thread_siblings", CPU_SYS_PATH, + cpu) < 0) { + virReportOOMError(); + return 0; + } + + pathfp = fopen(path, "r"); + if (pathfp == NULL) { + virReportSystemError(errno, _("cannot open %s"), path); + VIR_FREE(path); + return 0; + } + + if (fgets(str, sizeof(str), pathfp) == NULL) { + virReportSystemError(errno, _("cannot read from %s"), path); + goto cleanup; + } + + i = 0; + while (str[i] != '\0') { + if (str[i] != '\n' && str[i] != ',') + ret += popcnt(str[i] - '0'); + i++; + } + +cleanup: + fclose(pathfp); + VIR_FREE(path); + + return ret; +} + +static int parse_socket(int cpu) +{ + char *path = NULL; + FILE *pathfp; + char socket_str[1024]; + char *tmp; + int socket; + + if (virAsprintf(&path, "%s/cpu%d/topology/physical_package_id", + CPU_SYS_PATH, cpu) < 0) { + virReportOOMError(); + return -1; + } + + pathfp = fopen(path, "r"); + if (pathfp == NULL) { + virReportSystemError(errno, _("cannot open %s"), path); + goto cleanup; + } + + if (fgets(socket_str, sizeof(socket_str), pathfp) == NULL) { + virReportSystemError(errno, _("cannot read from %s"), path); + goto cleanup; + } + if (virStrToLong_i(socket_str, &tmp, 10, &socket) < 0) { + nodeReportError(NULL, VIR_ERR_INTERNAL_ERROR, + _("could not convert '%s' to an integer"), + socket_str); + goto cleanup; + } + +cleanup: + fclose(pathfp); + VIR_FREE(path); + + return socket; +} + +int linuxNodeInfoCPUPopulate(virConnectPtr conn, FILE *cpuinfo, + virNodeInfoPtr nodeinfo) +{ char line[1024]; + DIR *cpudir = NULL; + struct dirent *cpudirent = NULL; + int cpu; + unsigned long cur_threads; + int socket; + unsigned long long socket_mask = 0; nodeinfo->cpus = 0; nodeinfo->mhz = 0; - nodeinfo->nodes = nodeinfo->sockets = nodeinfo->cores = nodeinfo->threads = 1; + nodeinfo->nodes = nodeinfo->cores = 1; /* NB: It is impossible to fill our nodes, since cpuinfo * has not knowledge of NUMA nodes */ - /* XXX hyperthreads */ + /* NOTE: hyperthreads are ignored here; they are parsed out of /sys */ while (fgets(line, sizeof(line), cpuinfo) != NULL) { char *buf = line; if (STRPREFIX(buf, "processor")) { /* aka a single logical CPU */ @@ -122,12 +218,38 @@ int linuxNodeInfoCPUPopulate(virConnectPtr conn, FILE *cpuinfo, virNodeInfoPtr n return -1; } - /* - * Can't reliably count sockets from proc metadata, so - * infer it based on total CPUs vs cores. - * XXX hyperthreads + /* OK, we've parsed what we can out of /proc/cpuinfo. Get the socket + * and thread information from /sys */ - nodeinfo->sockets = nodeinfo->cpus / nodeinfo->cores; + cpudir = opendir(CPU_SYS_PATH); + if (cpudir == NULL) { + virReportSystemError(errno, _("cannot opendir %s"), CPU_SYS_PATH); + return -1; + } + while ((cpudirent = readdir(cpudir))) { + if (sscanf(cpudirent->d_name, "cpu%d", &cpu) != 1) + continue; + + socket = parse_socket(cpu); + if (socket < 0) { + closedir(cpudir); + return -1; + } + if (!(socket_mask & (1 << socket))) { + socket_mask |= (1 << socket); + nodeinfo->sockets++; + } + + cur_threads = count_thread_siblings(cpu); + if (cur_threads == 0) { + closedir(cpudir); + return -1; + } + if (cur_threads > nodeinfo->threads) + nodeinfo->threads = cur_threads; + } + + closedir(cpudir); return 0; } diff --git a/tests/nodeinfodata/linux-nodeinfo-1.txt b/tests/nodeinfodata/linux-nodeinfo-1.txt index e52e20a..09e2946 100644 --- a/tests/nodeinfodata/linux-nodeinfo-1.txt +++ b/tests/nodeinfodata/linux-nodeinfo-1.txt @@ -1 +1 @@ -CPUs: 2, MHz: 2800, Nodes: 1, Sockets: 1, Cores: 2, Threads: 1 +CPUs: 2, MHz: 2800, Nodes: 1, Cores: 2 diff --git a/tests/nodeinfodata/linux-nodeinfo-2.txt b/tests/nodeinfodata/linux-nodeinfo-2.txt index 12e819b..e4eea94 100644 --- a/tests/nodeinfodata/linux-nodeinfo-2.txt +++ b/tests/nodeinfodata/linux-nodeinfo-2.txt @@ -1 +1 @@ -CPUs: 2, MHz: 2211, Nodes: 1, Sockets: 1, Cores: 2, Threads: 1 +CPUs: 2, MHz: 2211, Nodes: 1, Cores: 2 diff --git a/tests/nodeinfodata/linux-nodeinfo-3.txt b/tests/nodeinfodata/linux-nodeinfo-3.txt index d285781..17d4d8e 100644 --- a/tests/nodeinfodata/linux-nodeinfo-3.txt +++ b/tests/nodeinfodata/linux-nodeinfo-3.txt @@ -1 +1 @@ -CPUs: 4, MHz: 1595, Nodes: 1, Sockets: 2, Cores: 2, Threads: 1 +CPUs: 4, MHz: 1595, Nodes: 1, Cores: 2 diff --git a/tests/nodeinfodata/linux-nodeinfo-4.txt b/tests/nodeinfodata/linux-nodeinfo-4.txt index 991d4f9..5a5c919 100644 --- a/tests/nodeinfodata/linux-nodeinfo-4.txt +++ b/tests/nodeinfodata/linux-nodeinfo-4.txt @@ -1 +1 @@ -CPUs: 4, MHz: 1000, Nodes: 1, Sockets: 1, Cores: 4, Threads: 1 +CPUs: 4, MHz: 1000, Nodes: 1, Cores: 4 diff --git a/tests/nodeinfodata/linux-nodeinfo-5.txt b/tests/nodeinfodata/linux-nodeinfo-5.txt index dce7ada..54abb5d 100644 --- a/tests/nodeinfodata/linux-nodeinfo-5.txt +++ b/tests/nodeinfodata/linux-nodeinfo-5.txt @@ -1 +1 @@ -CPUs: 4, MHz: 2814, Nodes: 1, Sockets: 2, Cores: 2, Threads: 1 +CPUs: 4, MHz: 2814, Nodes: 1, Cores: 2 diff --git a/tests/nodeinfodata/linux-nodeinfo-6.txt b/tests/nodeinfodata/linux-nodeinfo-6.txt index 75cdaa9..f89e35e 100644 --- a/tests/nodeinfodata/linux-nodeinfo-6.txt +++ b/tests/nodeinfodata/linux-nodeinfo-6.txt @@ -1 +1 @@ -CPUs: 4, MHz: 1000, Nodes: 1, Sockets: 2, Cores: 2, Threads: 1 +CPUs: 4, MHz: 1000, Nodes: 1, Cores: 2 diff --git a/tests/nodeinfotest.c b/tests/nodeinfotest.c index 4cb248a..b3b91ad 100644 --- a/tests/nodeinfotest.c +++ b/tests/nodeinfotest.c @@ -47,9 +47,8 @@ static int linuxTestCompareFiles(const char *cpuinfofile, const char *outputfile fclose(cpuinfo); snprintf(actualData, MAX_FILE, - "CPUs: %u, MHz: %u, Nodes: %u, Sockets: %u, Cores: %u, Threads: %u\n", - nodeinfo.cpus, nodeinfo.mhz, nodeinfo.nodes, nodeinfo.sockets, - nodeinfo.cores, nodeinfo.threads); + "CPUs: %u, MHz: %u, Nodes: %u, Cores: %u\n", + nodeinfo.cpus, nodeinfo.mhz, nodeinfo.nodes, nodeinfo.cores); if (STRNEQ(actualData, expectData)) { if (getenv("DEBUG_TESTS")) { -- 1.6.6.1

3 3

[libvirt] pxe not working with libvirt 0.7.7
by Ruben Kerkhof 08 Mar '10

08 Mar '10

Hi all, Maybe this is a known issue, but I had pxe boot working with libvirt 0.7.1 This config: <domain type='kvm'> <name>pxetest</name> <vcpu>1</vcpu> <memory>196608</memory> <os> <type>hvm</type> <boot dev="network" /> </os> <devices> <console type='pty' /> <interface type='bridge'> <source bridge='br1' /> </interface> </devices> </domain> Creates the following output with libvirt 0.7.7: [root@localhost data]# virsh create --console pxetest.xml error: Failed to create domain from pxetest.xml error: internal error Process exited while reading console log output: char device redirected to /dev/pts/0 Cannot boot from non-existent NIC Here's the log: LC_ALL=C PATH=/sbin:/usr/sbin:/bin:/usr/bin /usr/bin/qemu-kvm -S -M fedora-13 -enable-kvm -m 192 -smp 1,sockets=1,cores=1,threads=1 -name pxetest -uuid 0722f833-059e-a172-0d4c-bd2ffe106959 -nographic -nodefaults -chardev socket,id=monitor,path=/var/lib/libvirt/qemu/pxetest.monitor,server,nowait -mon chardev=monitor,mode=readline -rtc base=utc -no-acpi -boot n -device rtl8139,vlan=0,id=net0,mac=52:54:00:9f:b0:bb,bus=pci.0,addr=0x4 -net tap,fd=20,vlan=0,name=hostnet0 -chardev pty,id=serial0 -device isa-serial,chardev=serial0 -usb -device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x3 char device redirected to /dev/pts/0 Cannot boot from non-existent NIC libvirt 0.7.1 nicely presents the pxe menu on the console. Here's the qemu cmd under libvirt 0.7.1: LC_ALL=C PATH=/sbin:/usr/sbin:/bin:/usr/bin /usr/bin/qemu-kvm -S -M fedora-13 -m 192 -smp 1 -name pxetest -uuid 73a5f203-9417-c362-9d22-eb7da3cdbfde -nographic -monitor unix:/var/lib/libvirt/qemu/pxetest.monitor,server,nowait -no-acpi -boot n -net nic,macaddr=52:54:00:84:1e:ca,vlan=0,name=nic.0 -net tap,fd=18,vlan=0,name=tap.0 -serial pty -parallel none -usb char device redirected to /dev/pts/0 Regards, Ruben Kerkhof

1 0

[libvirt] [PATCH] Eliminate large stack buffer in doTunnelSendAll
by Laine Stump 06 Mar '10

06 Mar '10

doTunnelSendAll function (used by QEMU migration) uses a 64k buffer on the stack, which could be problematic. This patch replaces that with a buffer from the heap. While in the neighborhood, this patch also improves error reporting in the case that saferead fails - previously, virStreamAbort() was called (resetting errno) before reporting the error. It's been changed to report the error first. --- src/qemu/qemu_driver.c | 18 +++++++++++++++--- 1 files changed, 15 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index bb3edde..463c716 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -8399,19 +8399,28 @@ cleanup: } +#define TUNNEL_SEND_BUF_SIZE 65536 + static int doTunnelSendAll(virStreamPtr st, int sock) { - char buffer[65536]; - int nbytes = sizeof(buffer); + char *buffer; + int nbytes = TUNNEL_SEND_BUF_SIZE; + + if (VIR_ALLOC_N(buffer, TUNNEL_SEND_BUF_SIZE) < 0) { + virStreamAbort(st); + virReportOOMError(); + return -1; + } /* XXX should honour the 'resource' parameter here */ for (;;) { nbytes = saferead(sock, buffer, nbytes); if (nbytes < 0) { - virStreamAbort(st); virReportSystemError(errno, "%s", _("tunnelled migration failed to read from qemu")); + virStreamAbort(st); + VIR_FREE(buffer); return -1; } else if (nbytes == 0) @@ -8421,10 +8430,13 @@ static int doTunnelSendAll(virStreamPtr st, if (virStreamSend(st, buffer, nbytes) < 0) { qemuReportError(VIR_ERR_OPERATION_FAILED, "%s", _("Failed to write migration data to remote libvirtd")); + VIR_FREE(buffer); return -1; } } + VIR_FREE(buffer); + if (virStreamFinish(st) < 0) /* virStreamFinish set the error for us */ return -1; -- 1.6.6.1

2 2

Re: [libvirt] [virt-devel] RFC: Modelling timers / clocks & tick policies in libvirt
by Zachary Amsden 06 Mar '10

06 Mar '10

On 03/05/2010 04:27 AM, Daniel P. Berrange wrote: > This mail describes how I'm suggesting libvirt addresses Dor's RFE to > timers/tick policies in libvirt > > https://bugzilla.redhat.com/show_bug.cgi?id=557285 > > Thanks to those who've answered my previous mail / irc questions. Please > point out any mistakes I made in understanding / modelling this problem > > Daniel > > > Virtual machine timer management in libvirt > =========================================== > > On PC hardware there are a number of terrible timers / clock sources available > to operating systems > > * PIT > Timer with periodic interrupts > > * RTC > > Time of Day clock, continuous running > Timer with periodic interrupts > > * Local APIC Timer > Timer with periodic interrupts > > * ACPI Timer > Timer with periodic interrupts > > * TSC > Read via rdtsc instruction. No interrupts > Never, ever tell anyone this, but it is possible to generate interrupts from the TSC. Doing so is a million orders of perverse because of the problems you list: > Unreliable on some hardware. eg changes frequency. > Not synced between cores > Different HZ across hosts > > * HPET > Multiple timers with periodic interrupts > Can replace PIT/RTC timers > > They all generally suck in real hardware, and this gets worse in virtual machines. > Many different approaches to making them suck less in VMWare, Xen& KVM, but there > are some reasonably common concepts.... > HPET doesn't suck. > > Virtual timekeeping problems > ---------------------------- > > Three primary problems / areas to deal with: > > * Time of day clock (RTC) > > - Initialized to UTC/Localtime/Timezone/UTC+offset > - Two modes of operation: > 1. Guest wallclock: only runs when guest is executing. ie stopped across save/restore, etc > 2. Host wallclock: runs continuously with host wall time. > Windows wants RTC in localtime, Linux can do either but RTC in UTC is easiest to maintain. Guest / host wall clock is a general consideration for all timers, obviously RTC is supposed to be a wallclock, but there is a question: do we allow the guest to see host time exactly or do we hide it? > * Interrupt timers > > - Ticks can not always be delivered on time > > Policies to deal with "missed" ticks: > > 1. Deliver at normal rate without catchup > 2. Deliver at higher rate to catch up > 3. Merge into 1 tick& deliver asap > > 4. Discard all missed ticks > The issue is actually more complex than just these policies. A naive implementation of the policy leads to a guest DOS of the host. We actually have such a bug, and it demands a policy which merges ticks over a certain threshold and does not deliver ASAP. It's tricky and complex to fix because it means our notion of timers for the guest is wrong, and we need to introduce a higher order scheduling behaviour. In general, there isn't much we can tune here, but what we can tune is whether the other counters (RTC / HPET / TSC / ACPI) stay in sync with ticks delivered. It's not perfect or completely well defined because the tick can't actually be delivered until a fairly complex set of hardware rules is obeyed. This may not be apparent now, because it gets worse as we implement more hardware support for NMIs and SMIs. An ideal solution would sync the other counters when the tick is generated, not when it is injected. However, this leads us back to the DOS attack. There are also problems with SMP timing here (which CPU gets timer interrupts can change, and are they broadcast?). These problems are made worse because we don't gang schedule. > * TSC > - rdtsc instruction can be exposed to guests in two ways > > 1. Trap + emulate (slow, but more reliable) > 2. Native (fast, but possibly unreliable) > > Optionally also expose a 'rdtscp' instruction > > Possiblly set a fixed HZ independant of host. > There is also 3) a mixed approach; trap and emulate only when required, allow native access and offset appropriately at each exit; and 4) a SMP safe approach; trap and emulate always, and interlock SMP access to the clock so it is globally consistent 5) a secure approach; trap and emulate always and hide host time. This precludes the possibility of SMP, as timing differences can be observed since we don't gang schedule. This obviously has implications for the other timers. So this variable is not a simple boolean, but a multi-choice. > VMWare timekeeping > ------------------ > > * All timers run in "apparant time" ie track guest wallclock > * Missed tick policy is to deliver at higher rate to catchup > * TSC can be switched between native/emulate (virtual_rdtsc=TRUE|FALSE) > * TSC can have hardcoded HZ in emulate mode (apparantHZ=VALUE) > * RTC time of day is synced to host at startup (rtc.diffFromUTC or rtc.startTime) > * VMWare tools reset guest TOD if it gets out of sync > There is also lateness hiding; (timeTracker.hideLateness); adjust TSC to compensate for lateness of injected interrupts (it's the slightly buggy counter compensation at each tick I mention above). > Xen timekeeping > --------------- > > * Virtual platform timer (VPT) used as source for other timers > * VPT has 4 modes > > 0: delay_for_missed_ticks > > Missed ticks are delivered when next scheduled, at the normal > rate. RTC runs in guest wallclock, so is delayed. No catchup is > attempted > > 1: no_delay_for_missed_ticks > > Missed ticks are delivered when next scheduled, at the normal > rate. RTC runs in host wallclock, so is not delayed. > > 2: no_missed_ticks_pending > > Missed ticks are discarded& next tick is delivered normally. RTC > runs in host wallclock. > > > 3: one_missed_tick_pending > > Missed interrupts are collapsed into a single late tick. RTC > run in host wallclock. > > * HPET > > Optionally enabled > > * TSC. Can run in 4 modes > > - auto: emulate if host TSC is unstable. native with invariant TSC > - native: always native regardless of host TSC stability > - emulate: trap + emulate regardless of host TSC invariant > - pvrdtsc: native, requiring invariant TSC. Also exposes rdtscp instruction > TSC is complex enough without RDTSCP. Let's consider rdtscp as a host optimization for vendors of hardware with buggy clocks who want fast gettimeofday system calls. We already are compensating to try to keep virtual TSC in sync on KVM and probably don't need this mode. > > KVM timekeeping > --------------- > > * PIT: can be in kernel, or userspace (userspace deprecated for KVM) > > Default tick policies differ for both impls > > - Userspace: Default: missed ticks are delivered when next scheduled at normal rate > > -tdf flag enable tick reinjection to catchup > > - Kernel: Default: Missed ticks are delivered at higher rate to catch up > > -no-kvm-pit-reinjection to disable tick reinjection catchup > > * RTC > > TOD clock can run in host or guest wallclock (clock=host|guest) > > Default: missed ticks are delivered when next scheduled at normal rate > > -rtc-td-hack or -clock driftfix=slew: missed ticks are delivered at a > higher rate to catchup > > * TSC > > - Always runs native. > > * HPET > > - Optionally enabled/disabled > > > Mapping in libvirt XML > ---------------------- > > Currently supports setting Time of Day clock via > > <clock offset="utc"/> > > Always sync to UTC > > <clock offset="localtime"/> > > Always sync to host timezone > > <clock offset="timezone" timezone='Europe/Paris'/> > > Sync to arbitrary timezone > > <clock offset="variable" adjustment='123456'/> > > Sync to UTC + arbitrary offset > > > > Proposal to model all timers policies as sub-elements of this<clock/> > In general we wil allow zero or more<timer/> elements following the > syntax: > > <timer name='platform|pit|rtc|hpet|tsc' > wallclock='host|guest' > tickpolicy='none|catchup|merge|discard' > frequency='123' > mode='auto|native|emulate|paravirt' > present='yes|no' /> > > Meaning of 'name': > > Names map to regular PC timers / clocks. 'Platform' refers to the > (optional) master virtual clock source that may be used to drive > policy of "other" clocks (eg used in Xen, which clocks are controlled > by the platform clock is to be undefined because it has varied in > Xen over time). > > Meaning of 'tickpolicy': > > none: continue to deliver at normal rate (ie ticks are delayed) > catchup: deliver at higher rate to catchup > merge: ticks merged into 1 single tick > discard: all missed ticks are discarded > > Meaning of 'wallclock': > > Only valid for name='rtc' or 'platform' > > host: RTC wallclock always tracks host time > guest: RTC wallclock always tracks host time > > Meaning of 'frequency': > > Set a fixed frequency in HZ. > > NB: Only relevant for TSC. All other timers are fixed (PIT, RTC), or > fully guest controlled frequency (HPET) > Actually, the guest doesn't control the HPET base frequency, only the divider. I think. I'd need to recheck the spec. > Meaning of 'mode': > > Control how the clock is exposed to guest. > > auto: native if safe, otherwise emulate > native: always native > emulate: always emulate > paravirt: native + paravirtualize > > NB: Only relevant for TSC. All other timers are always emulated. > auto, native, emulate can map nicely for us, but it would be good to have an smp safe mode. (A secure mode is more of a global setting for all timers). > Meaing of 'present': > > Used to override default set of timers visible to the guest. eg to > enable or disable the HPET > > > > Mapping to VMWare > ----------------- > > eg with guest config showing > > diffFromUTC='123456' > apparentHZ='123456' > virtual_rdtsc=False > > libvirt XML gets: > > <clock mode='variable' adjustment='123456'> > <timer name='tsc' frequency='123456' mode='native'/> > </clock> > > > Mapping to Xen > -------------- > > eg with guest config showing > > timer_mode=3 > hpet=1 > tsc_mode=2 > localtime=1 > > <clock mode='localtime'> > <timer name='platform' tickpolicy='merge' wallclock='host'/> > <timer name='hpet'/> > <timer name='tsc' mode='native'/> > </clock> > > > Mapping to KVM > -------------- > > eg with guest ARGV showing > > -no-kvm-pit-reinjection > -clock base=localtime,clock=guest,driftfix=slew > -no-hpet > > > <clock mode='localtime'> > <timer name='rtc' tickpolicy='catchup' wallclock='guest'/> > <timer name='pit' tickpolicy='none'/> > <timer name='hpet' present='no'/> > </clock> > > > > Further reading > --------------- > > VMWare has the best doc: > > http://www.vmware.com/pdf/vmware_timekeeping.pdf > > Xen: > > Docs on 'tsc_mode' at > > $SOURCETREE/docs/misc/tscmode.txt > > Docs for 'timer_mode' in the source code only: > > xen/include/public/hvm/params.h > > KVM: > > No docs at all. Guess from -help descriptions, reading source code& asking > clever people about it :-) > Let me propose an XML mapping a bit later today. I haven't had coffee yet, and we know what that can do. Zach

2 2

[libvirt] [PATCH 0/1] contributor guidelines
by David Allan 05 Mar '10

05 Mar '10

Here's a patch to consolidate the contributor guidelines into one place on the website and add a section on the use of goto. There was a small amount of content that was in HACKING but not in hacking.html.in, which I added. I also added Dan's suggested goto labels for common cases to the section on goto, and I removed the contents of HACKING and replaced them with a link to the website. Dave David Allan (1): Consolidating contributor guidelines HACKING | 395 +------------------------------------------------- docs/hacking.html.in | 125 ++++++++++++++++- 2 files changed, 125 insertions(+), 395 deletions(-)

2 2