Red Hat, Inc.

Red Hat Israel, Ltd.

Marcelo Wormsbecker Tosatti

Henri Han Van Riel

Paolo Bonzini

Karen Lee Noel

Avi Kivity

Ronen Hod

Luiz Capitulino

Henri Van Riel

Processor frequencies can be managed. For example, a computing device can determine (i) a first estimate of an operating characteristic of a processor in using a first pair of frequencies to perform a task, and (ii) a second estimate of the operating characteristic of the processor in using a second pair of frequencies to perform the task. The computing device can select the first pair of frequencies based on determining that the first estimate is closer to a target operating-characteristic of the processor while performing the task than the second estimate. Based on selecting the first pair of frequencies, the computing device can set a clock rate of the processor to a lower frequency in the first pair of frequencies while performing the task. The computing device can also set the clock rate of the processor to a higher frequency in the first pair of frequencies while performing the task.

Managing processor frequencies

A system includes a hypervisor, virtual machine(s), and virtual processor(s). The virtual processor(s) includes a real time virtual processor and a housekeeping virtual processor. The housekeeping virtual processor is configured to initiate a non-preemptible task and responsive to initiating the non-preemptible task, increment a counter value from an initial counter value. The housekeeping virtual processor is also configured to complete the non-preemptible task and responsive to completing the non-preemptible task, decrement the counter value towards the initial counter value. Responsive to the counter value reaching the initial counter value, the virtual housekeeping processor is configured to check a flag set by the hypervisor. Responsive to checking the flag, the housekeeping process is configured to execute a hypercall to yield a processor associated with the housekeeping virtual processor.

Housekeeping virtual processor overcommit for real time virtualization

A system and methods are disclosed for configuring a virtualized system for real-time response from virtual machines (VMs). In accordance with one embodiment, a VM request is received, where the request specifies a number N of virtual processors for executing real-time tasks, and where N is a positive integer. In response to the VM request, a plurality of physical processors are partitioned into a first subset dedicated to real-time tasks and a second subset dedicated to non-real-time tasks, the first subset having at least N physical processors, and one or more commands are submitted to one or both of a hypervisor and a guest OS of a first VM to configure the first VM to have N virtual processors dedicated to real-time tasks and a set of additional virtual processors dedicated to non-real-time tasks.

Configuration of virtual machines in view of response time constraints

Power can be conserved by managing the frequencies at which a processor operates to process data packets. For example, a pair of frequencies can be selected from among at least three candidate frequencies based on a power consumption associated with the pair of frequencies being less than another power consumption associated with another pair of frequencies selected from the at least three candidate frequencies. A clock rate of the processor can be set to a lower frequency in the selected pair of frequencies based on a number of data packets in a data buffer being below a threshold number of data packets. The clock rate of the processor can be set to a higher frequency in the selected pair of frequencies based on the number of data packets in the data buffer meeting or exceeding the threshold number of data packets.

Optimized extended context management for virtual machines is disclosed. An example system includes a memory, processor, VM, and kernel, which executes the VM using guest space, stores a guest state in an extended state register, which the VM has access to, receives an instruction, saves the guest state, and determines whether the kernel is configured to handle the process using kernel space. If so, the process is handled using kernel space without using user space, and if not, a user state in the extended state register is restored, the process is handled using user space, an updated user state is stored in the extended state register, the updated user state is saved, and the guest state is restored in the extended state register. Finally, execution of the VM using the guest space is resumed.

Optimized extended context management for virtual machines

An application associated with a virtual processor running on a physical processor reads a first value of a counter and a second value of the counter. The counter is indicative of a migration status of the virtual processor with respect to the physical processor. Responsive to determining that the first value of the counter does not equal the second value of the counter, the application ascertains whether a value of a hardware parameter associated with the physical processor has changed during a time interval. The migration status indicates a count of the number of times the virtual processor has migrated a first physical processor to a second physical processor. The application determines the validity of a value of a performance monitoring unit derived from the hardware parameter in view of the application ascertaining whether the value of the hardware parameter has changed during the time interval.

Paravirtualized migration counter for migrating a virtual CPU to a different physical CPU

An application associated with a processor reads a first value of a counter and a second value of the counter. The counter is indicative of a migration status of the application with respect to the processor. Responsive to determining that the first value of the counter does not equal the second value of the counter, the application ascertains whether a value of a hardware parameter associated with the processor has changed during a time interval. The migration status indicates a count of the number of times the application has migrated from one processor to another processor. The application determines the validity of a value of a performance monitoring unit derived from the hardware parameter in view of the application ascertaining whether the value of the hardware parameter has changed during the time interval.

Paravirtualized migration counter

A virtual machine disk page detector running on a computing device detects guest disk cache usage. The detector detects a request from a virtual operating system to read an object into a virtual memory page from a virtual disk, maintains a record of a page identifier and a corresponding virtual disk address in a guest cache data structure, and modifies a protection identifier of the virtual memory page to indicate that the virtual memory page is protected.

Detection of guest disk cache

In one embodiment, a mechanism for out-of-synch virtual machine memory management optimization is disclosed. In one embodiment, a method for out-of-synch virtual machine memory management optimization includes receiving a memory management unit (MMU) synchronization event issued from a VM virtualized by a VM monitor (VMM) of a host server device, and synchronizing one or more unsynchronized page tables (PTs) of a shadow PT hierarchy maintained by the VMM with one or more corresponding guest PTs of a guest PT hierarchy maintained by the VM, wherein the one or more unsynchronized PTs include an unlimited number of unsynchronized PTs in a visible address space of the shadow PT hierarchy that is determined by a current CR3 register address of the shadow PT hierarchy.

Growing community of inventors

Curitiba, Brazil

Marcelo Wormsbecker Tosatti