search for: hypervisor_sched_op

Hi, I am trying to understand Hypervisor_sched_op function and I have few question regarding this - 1. I understand that a guest VM can use the above function with "yield", in order to relinquish CPU time to other guests with running tasks. a) So when the guest VM makes this call does it save/persist any scheduling information about it...

Hi, I am trying to understand Hypervisor_sched_op function and I have few question regarding this - 1. I understand that a guest VM can use the above function with "yield", in order to relinquish CPU time to other guests with running tasks. a) So when the guest VM makes this call does it save/persist any scheduling information about it...

Keir, The patch below is necessary to get ia64 building on current xen-unstable.hg. Thanks, Alex Signed-off-by: Alex Williamson <alex.williamson@hp.com> --- diff -r 3983e4f1b054 xen/arch/ia64/Rules.mk --- a/xen/arch/ia64/Rules.mk Sun Mar 12 10:03:33 2006 +0100 +++ b/xen/arch/ia64/Rules.mk Mon Mar 13 09:36:01 2006 -0700 @@ -12,7 +12,7 @@ CPPFLAGS += -I$(BASEDIR)/include -I$(BA

...es(struct cpuinfo_x86 *c) > } > } > > +static void xen_pin_vcpu(int cpu) > +{ > + static bool disable_pinning; > + struct sched_pin_override pin_override; > + int ret; > + > + if (disable_pinning) > + return; > + > + pin_override.pcpu = cpu; > + ret = HYPERVISOR_sched_op(SCHEDOP_pin_override, &pin_override); /* Ignore errors when removing override. */ > + if (cpu < 0) > + return; > + > + switch (ret) { > + case -ENOSYS: > + pr_warn("The kernel tried to call a function on physical cpu %d, but Xen isn't\n" > + "sup...

...es(struct cpuinfo_x86 *c) > } > } > > +static void xen_pin_vcpu(int cpu) > +{ > + static bool disable_pinning; > + struct sched_pin_override pin_override; > + int ret; > + > + if (disable_pinning) > + return; > + > + pin_override.pcpu = cpu; > + ret = HYPERVISOR_sched_op(SCHEDOP_pin_override, &pin_override); /* Ignore errors when removing override. */ > + if (cpu < 0) > + return; > + > + switch (ret) { > + case -ENOSYS: > + pr_warn("The kernel tried to call a function on physical cpu %d, but Xen isn't\n" > + "sup...

...arch specific sub-ops: + * * XENMEM_add_to_physmap + * * XENMEM_add_to_physmap_range + * + * HYPERVISOR_domctl + * All generic sub-operations, with the exception of: + * * XEN_DOMCTL_iomem_permission (not yet implemented) + * * XEN_DOMCTL_irq_permission (not yet implemented) + * + * HYPERVISOR_sched_op + * All generic sub-operations, with the exception of: + * * SCHEDOP_block -- prefer wfi hardware instruction + * + * HYPERVISOR_console_io + * All generic sub-operations + * + * HYPERVISOR_xen_version + * All generic sub-operations + * + * HYPERVISOR_event_channel_op + * All generic...

.../xen/enlighten.c @@ -1885,6 +1885,45 @@ static void xen_set_cpu_features(struct cpuinfo_x86 *c) } } +static void xen_pin_vcpu(int cpu) +{ + static bool disable_pinning; + struct sched_pin_override pin_override; + int ret; + + if (disable_pinning) + return; + + pin_override.pcpu = cpu; + ret = HYPERVISOR_sched_op(SCHEDOP_pin_override, &pin_override); + if (cpu < 0) + return; + + switch (ret) { + case -ENOSYS: + pr_warn("The kernel tried to call a function on physical cpu %d, but Xen isn't\n" + "supporting this. In case of problems you might consider vcpu pinning.\n", + c...

...xen/events.c b/drivers/xen/events.c index c3290bc..81e00ee 100644 --- a/drivers/xen/events.c +++ b/drivers/xen/events.c @@ -772,7 +772,7 @@ void xen_poll_irq(int irq) poll.nr_ports = 1; poll.timeout = 0; - poll.ports = &evtchn; + set_xen_guest_handle(poll.ports, &evtchn); if (HYPERVISOR_sched_op(SCHEDOP_poll, &poll) != 0) BUG(); -- 1.6.0.rc0.42.g186458

...as/mini-os/arch/x86/time.c Thu May 08 11:40:05 2008 +0100 @@ -201,10 +201,12 @@ void block_domain(s_time_t until) { struct timeval tv; gettimeofday(&tv, NULL); + ASSERT(irqs_disabled()); if(monotonic_clock() < until) { HYPERVISOR_set_timer_op(until); HYPERVISOR_sched_op(SCHEDOP_block, 0); + local_irq_disable(); } } _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel

...xen/events.c b/drivers/xen/events.c index c3290bc..81e00ee 100644 --- a/drivers/xen/events.c +++ b/drivers/xen/events.c @@ -772,7 +772,7 @@ void xen_poll_irq(int irq) poll.nr_ports = 1; poll.timeout = 0; - poll.ports = &evtchn; + set_xen_guest_handle(poll.ports, &evtchn); if (HYPERVISOR_sched_op(SCHEDOP_poll, &poll) != 0) BUG(); -- 1.6.0.rc0.42.g186458

...xen/events.c b/drivers/xen/events.c index c3290bc..81e00ee 100644 --- a/drivers/xen/events.c +++ b/drivers/xen/events.c @@ -772,7 +772,7 @@ void xen_poll_irq(int irq) poll.nr_ports = 1; poll.timeout = 0; - poll.ports = &evtchn; + set_xen_guest_handle(poll.ports, &evtchn); if (HYPERVISOR_sched_op(SCHEDOP_poll, &poll) != 0) BUG(); -- 1.6.0.rc0.42.g186458

...gt; >> +static void xen_pin_vcpu(int cpu) >> +{ >> + static bool disable_pinning; >> + struct sched_pin_override pin_override; >> + int ret; >> + >> + if (disable_pinning) >> + return; >> + >> + pin_override.pcpu = cpu; >> + ret = HYPERVISOR_sched_op(SCHEDOP_pin_override, &pin_override); > > /* Ignore errors when removing override. */ Okay. >> + if (cpu < 0) >> + return; >> + >> + switch (ret) { >> + case -ENOSYS: >> + pr_warn("The kernel tried to call a function on physical cpu %d, bu...

-- Juergen Gross Principal Developer IP SW OS6 Telephone: +49 (0) 89 636 47950 Fujitsu Siemens Computers e-mail: juergen.gross@fujitsu-siemens.com Otto-Hahn-Ring 6 Internet: www.fujitsu-siemens.com D-81739 Muenchen Company details: www.fujitsu-siemens.com/imprint.html _______________________________________________

Some hardware (e.g. Dell Studio laptops) require special functions to be called on physical cpu 0 in order to avoid occasional hangs. When running as dom0 under Xen this could be achieved only via special boot parameters (vcpu pinning) limiting the hypervisor in it's scheduling decisions. This patch series is adding a generic function to be able to temporarily pin a (virtual) cpu to a

Some hardware (e.g. Dell Studio laptops) require special functions to be called on physical cpu 0 in order to avoid occasional hangs. When running as dom0 under Xen this could be achieved only via special boot parameters (vcpu pinning) limiting the hypervisor in it's scheduling decisions. This patch series is adding a generic function to be able to temporarily pin a (virtual) cpu to a

...ep = 0; @@ -391,9 +405,11 @@ fbfront_update(fb_dev, x - 16, y - 16, 33, 33); } } else if (kbdevent.key.keycode == KEY_Q) { - struct sched_shutdown sched_shutdown = { .reason = SHUTDOWN_poweroff }; - HYPERVISOR_sched_op(SCHEDOP_shutdown, &sched_shutdown); - do_exit(); + shutdown_reason = SHUTDOWN_poweroff; + wmb(); + do_shutdown = 1; + wmb(); + wake_up(&shutdown_queue); }...

...> /* + * Guest Scheduler Operations + * + * The SCHEDOP interface provides mechanisms for a guest to interact + * with the scheduler, including yield, blocking and shutting itself + * down. + */ + +/* * The prototype for this hypercall is: - * long sched_op_new(int cmd, void *arg) + * long HYPERVISOR_sched_op(enum sched_op cmd, void *arg, ...) + * * @cmd == SCHEDOP_??? (scheduler operation). * @arg == Operation-specific extra argument(s), as described below. + * ... == Additional Operation-specific extra arguments, described below. * - * **NOTE**: - * Versions of Xen prior to 3.0.2 provide only t...

...> /* + * Guest Scheduler Operations + * + * The SCHEDOP interface provides mechanisms for a guest to interact + * with the scheduler, including yield, blocking and shutting itself + * down. + */ + +/* * The prototype for this hypercall is: - * long sched_op_new(int cmd, void *arg) + * long HYPERVISOR_sched_op(enum sched_op cmd, void *arg, ...) + * * @cmd == SCHEDOP_??? (scheduler operation). * @arg == Operation-specific extra argument(s), as described below. + * ... == Additional Operation-specific extra arguments, described below. * - * **NOTE**: - * Versions of Xen prior to 3.0.2 provide only t...

...pu_var(cpu_state) = CPU_DEAD; - - /* We shouldn''t have to disable interrupts while dead, but - * some interrupts just don''t seem to go away, and this makes - * it "work" for testing purposes. */ - /* Death loop */ - while (__get_cpu_var(cpu_state) != CPU_UP_PREPARE) - HYPERVISOR_sched_op(SCHEDOP_yield, 0); - - local_irq_disable(); - __flush_tlb_all(); - cpu_set(smp_processor_id(), cpu_online_map); + HYPERVISOR_vcpu_op(VCPUOP_down, smp_processor_id(), NULL); local_irq_enable(); } #else _______________________________________________ Xen-devel mailing list Xen-devel@lists.xens...

This patch fixes the bug that "xm reboot <domid>" can not reboot vmx domain. In xm reboot process, control panel will wait until DomU notified it. Unlike DomU, vmx domain can not passively do that, since it runs unmodified OS. This patch adds the missing logic in control panel. Any comment? Signed-off-by: Xu Dan <dan.d.xu@intel.com > Yu Ke