摘要:
A device controller can have multiple device performance states (DPS), which represent different levels of performance vs. power consumption during operation. The device controller can include a DPS status register that can be read by a processor, to indicate the current DPS, and a DPS control register that can be written by the processor, to change the current DPS to a desired DPS. The controller may also have a processor performance state (PPS) status register which can be used to affect the desired choice of DPS based on the performance state of the processor. Each of the registers can be accessed by the device driver for that device controller. The DPS of multiple devices can be coordinated to achieve an improved system-level reduction in power consumption, while maintaining sufficient operational capability.
摘要:
A selectable control over multiple clock frequency/voltage level combinations that can be activated in a processor. A table can be placed in hardware that defines multiple combinations of CPU clock frequency and CPU operating voltage. By placing the table in hardware, it can be assured that all the various combinations will work for the particular processor device. Software can then be used to select a combination from this table, to control the actual frequency/voltage combination that is being implemented at a given time. This allows dynamic control over the power/performance tradeoff, so that the system can see maximum power savings consistent with acceptable performance, as operating and environmental considerations continue to change the most desirable selections.
摘要:
An embodiment of a system for avoiding race conditions when using edge-triggered interrupts includes a processor that asserts an interrupt pending signal in response to the receipt of an edge-triggered interrupt. A power management device receives the interrupt pending signal. If the processor is in a low power state when it asserts the interrupt pending signal, then the power management device causes the processor to enter a high power state to allow the processor to service the pending interrupt.
摘要:
A high performance Local Bus Peripheral Interface (LBPI) for a computer local bus and its high performance peripheral interface(s), using a pipelined architecture to increase the use of the available data transfer bandwidth. To accomplish the above, the LBPI, which is coupled between the computer local bus and the peripheral interface(s), is provided a pipelined architecture which includes a Read Ahead Buffer, a Read Ahead Counter, a Data Out Latch, and a Controlling State Machine with a Configuration Register. In one embodiment, the LBPI can be selectably configured to couple on the host side to either a VL bus or PCI bus. Efficiency of Read-Ahead operations is further enhanced by maintaining a countdown of the number of words of a data sector already transferred and/or "snooping" the peripheral device commands from the computer to intelligently predict the occurrence of subsequent read data transfers commands. The Controlling State Machine also "snoops" the peripheral device commands to maintain its record of the operating parameters of the peripheral devices and also keeps track of which of the devices is currently active. In one embodiment, the LBPI supports DMA and PIO data transfers on the peripheral side. In another embodiment, the LBPI translates memory data transfers into IO data transfers to improve efficiency of IO data transfers. A DMA Timeout Counter is used during DMA mode data transfer operations to prevent the system from indefinitely waiting for an appropriate DMA Request Signal from a selected peripheral. During a DMA mode data transfer operation, forced interrupts may be generated and transmitted to the host in order to emulate a PIO mode data transfer operation. During a DMA mode data transfer operation, an imposed status or "Fake 3F6" register is utilized to transmit status information to the host system.
摘要:
A method and apparatus for performing logical attachments and detachments in a hot-plug-in data bus is described. A hot-plug-in data bus may utilize pull-down resistors to keep bus signals near a low voltage level when bus units are physically detached. Active pull-up resistors may then raise the bus signals away from ground when the bus units are physically attached via cabling or other forms of interconnection. The pull-up resistors may be switched away from the pull-up voltage source, which allows the remaining pull-down resistors to pull the bus signals down to the voltage levels corresponding to physical detachment of the cabling.
摘要:
A high performance Local Bus Peripheral Interface (LBPI) for a computer local bus and its high performance peripheral interface(s) uses a pipelined architecture to increase the use of the available data transfer bandwidth. In one embodiment, the LBPI can be selectably configured to couple on the host side to either a VL bus or PCI bus. The LBPI maintains a countdown of the number of words of a data sector already transferred and/or "snoops" the peripheral device commands from the computer to predict the occurrence of subsequent read data transfers commands. The Controlling State Machine also "snoops" the peripheral device commands to maintain its record of the operating parameters of the peripheral devices and also keeps track of which of the devices is currently active. In one embodiment, the LBPI supports DMA and PIO data transfers on the peripheral side. In another embodiment, the LBPI translates memory data transfers into IO data transfers to improve efficiency of IO data transfers. A DMA Timeout Counter is used during DMA mode data transfer operations to prevent the system from indefinitely waiting for an appropriate DMA Request Signal from a selected peripheral. During a DMA mode data transfer operation, forced interrupts may be generated and transmitted to the host in order to emulate a PIO mode data transfer operation. During a DMA mode data transfer operation, an imposed status or "Fake 3F6" register is utilized to transmit status information to the host system.
摘要:
For one disclosed embodiment, a lower limit for a power consumption device may be identified. Performance of the power consumption device may be reduced in response to a determination that a temperature corresponding to the power consumption device exceeds a threshold. Performance reduction may be limited based on the lower limit. Other embodiments are also disclosed.
摘要:
An audio noise mitigation approach. For one aspect, a first voltage associated with a first power management state is provided. A signal responsive to an indication associated with at least a first type of periodic exit event is received and responsive to the signal, a transition to a second voltage associated with a second state is initiated, a rate of the transition to the second voltage being slower than a similar voltage transition initiated in response to a non-periodic exit event.
摘要:
Systems and methods of managing power consumption provide for placing a processor in a non-snoopable state while a storage interface associated with the processor is enabled for bus mastering. In one embodiment, the bus mastering results in traffic between the storage interface and a storage device, where the traffic is monitored and the processor is placed a snoopable state when traffic is moving, and in the non-snoopable idle state if the traffic ceases for a period of time.
摘要:
In one embodiment, a data processing system includes, but is not limited to, a processor, a memory coupled to the processor, and a universal serial bus (USB) controller coupled to the processor and the memory. The USB controller includes a local memory to cache at least one activity descriptor of at least a portion of a periodic schedule having multiple frames stored in the main memory. The USB controller defers to service an active USB device described by one of the activity descriptors until a corresponding frame is scheduled to be serviced subsequently. Other methods and apparatuses are also described.