Classifications

• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06F—ELECTRIC DIGITAL DATA PROCESSING
  • G06F13/00—Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
  • G06F13/14—Handling requests for interconnection or transfer
  • G06F13/20—Handling requests for interconnection or transfer for access to input/output bus
  • G06F13/24—Handling requests for interconnection or transfer for access to input/output bus using interrupt
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06F—ELECTRIC DIGITAL DATA PROCESSING
  • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06F—ELECTRIC DIGITAL DATA PROCESSING
  • G06F13/00—Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
  • G06F13/14—Handling requests for interconnection or transfer
  • G06F13/20—Handling requests for interconnection or transfer for access to input/output bus
  • G06F13/28—Handling requests for interconnection or transfer for access to input/output bus using burst mode transfer, e.g. direct memory access DMA, cycle steal

Definitions

• Disclosed embodiments are directed to techniques for handling interrupts in processors.
• exemplary embodiments are directed to systems and methods for reducing interrupt latency in two level interrupt controllers configured for multithreaded processors.
• Processing systems commonly support interrupt mechanisms, wherein an interrupt may asynchronously halt or suspend a current execution thread or instruction stream of a processor, such that the interrupt may be serviced.
• An interrupt may be generated from various sources, including on-chip or off-chip external devices.
• Interrupts may also be generated internally within a processor or CPU, such as from one or more threads in a multi-threaded processor.
• an interrupt service routine may be executed by the processor receiving the interrupt.
• Each interrupt may include a particular ISR associated with the interrupt.
• an interrupt controller is commonly used to handle the tasks of receiving the interrupts, prioritizing among several outstanding interrupts, and tracking the status of pending interrupts such that a processor's availability to process new interrupts may be ascertained.
• a vectored interrupt controller VIC is known in the art to track vectored addresses associated with each interrupt, such that the VIC may be enabled to provide the processor servicing the interrupt with the associated ISR.
• priority levels may be dynamically or statically assigned to the threads, in order for an interrupt controller to determine which thread should be interrupted in order to service an interrupt.
• a first level or LI interrupt controller may be configured, for example, to handle interrupts related to a processor core, such as a multi-threaded processor.
• a second level or L2 interrupt controller may be configured, for example, to handle interrupts from external devices or interrupts on a global scale.
• the L2 interrupt controller may be in communication with the LI interrupt controller through system buses such as AHB/AXI to direct interrupts accordingly from the L2 to the LI interrupt controller.
• Two-level interrupt controllers, such as LI and L2 interrupt controllers may find several other applications in processing systems, as will be recognized by one of ordinary skill in the art.
• L2 interrupt controller 102 may communicate an interrupt over bus 108 to LI interrupt controller 104, which may be attached to core 106. As shown, core 106 is in direct communication with only LI interrupt controller 104, and not L2 interrupt controller 102. Initially, LI interrupt controller 104 may receive a first interrupt from L2 interrupt controller 102. Thereafter, the processing of subsequent interrupts may be handled in one of two ways, for example, based on processor resources.
• core 106 may provide an indication to L2 interrupt controller 102, through LI interrupt controller 104, that the core 106 is ready for a new interrupt.
• L2 interrupt controller 102 may then send the processor core a second interrupt if a second interrupt is pending at L2 interrupt controller 102.
• the first interrupt may be serviced by a first thread of the multi-threaded processor, and a second thread may be in WAIT state, and available to process a second interrupt.
• the multi-threaded processor may provide an indication L2 interrupt controller 102, for example, via LI interrupt controller 104, that L2 interrupt controller 102 may send a second interrupt, immediately after the processor core receives the first interrupt.
• core 106 may provide indication to L2 interrupt controller 102 to defer sending any new requests until a later point in time or until further notice.
• core 106 may be configured as a multi-threaded processor, all the threads may be busy, and the real-time operating system (RTOS) associated with the processor core may require a time delay in order to determine which thread to interrupt.
• RTOS may determine which hardware thread is running a software thread with the least priority, and designate that thread as the lowest priority software thread, such that LI interrupt controller 104 may direct a second interrupt from L2 interrupt controller 102 to the lowest priority software thread.
• the determination of the lowest priority software thread may incur significant time delay, and correspondingly, the rate at which interrupts can be processed suffers degradation.
• information regarding interrupts such as vectored addresses associated with the interrupt' s ISR, is communicated between L2 interrupt controller 102 and LI interrupt controller 104 over an Advanced Microcontroller Bus Architecture High Performance Bus (AHB).
• AHB Advanced Microcontroller Bus Architecture High Performance Bus
• the process associated with reading the AHB in order to retrieve the above information may add a significant delay to the interrupt latency, and thus further impact the rate at which interrupts are processed.
• Exemplary embodiments of the invention are directed to systems and method for reducing interrupt latency in two level interrupt controllers configured for multithreaded processors.
• an exemplary embodiment is directed to a method for reducing interrupt latency time, comprising: communicating a first interrupt and a first vector identifier from a second interrupt controller to a first interrupt controller; processing the first interrupt and the first vector identifier at the first interrupt controller; sending the processed interrupt from the first interrupt controller to a thread in a core; determining when the core is ready to receive a second interrupt; and sending an instruction from the core to the second interrupt controller indicating the core is ready to receive the second interrupt.
• Another exemplary embodiment is directed to a multi-threaded processor, comprising: a core, a level two interrupt controller, and a line coupling the core to the level two interrupt controller, wherein the core is configured to indicate to the level two interrupt controller, via the line, that the core is ready to receive a level two interrupt.
• Yet another exemplary embodiment is directed to a processing system configured for reduced interrupt latency, the processing system comprising: a first interrupt controller coupled to a core; means for communicating a first interrupt and a first vector identifier from a second interrupt controller to the first interrupt controller; means for processing the first interrupt and the first vector identifier at the first interrupt controller; means for sending the processed interrupt to a thread in the core; means for determining when the core is ready to receive a second interrupt; and means for sending an instruction from the core to the second interrupt controller indicating the core is ready to receive the second interrupt.
• Another exemplary embodiment is directed to a non-transitory computer-readable storage medium comprising code, which, when executed by a processor, causes the processor to perform operations for reducing interrupt latency time, the non-transitory computer-readable storage medium comprising: code for communicating a first interrupt and a first vector identifier from a second interrupt controller to a first interrupt controller; code for processing the first interrupt and the first vector identifier at the first interrupt controller; code for sending the processed interrupt from the first interrupt controller to a thread in a core; code for determining when the core is ready to receive a second interrupt; and code for sending an instruction from the core to the second interrupt controller indicating the core is ready to receive the second interrupt.
• FIG. 1 illustrates a conventional two-level interrupt controller interface to a core.
• FIG. 2 illustrates a two-level interrupt controller interface to a multi-threaded processor, configured according to exemplary embodiments.
• FIG. 3 illustrates a flow chart detailing a method for configuring a two-level interrupt controller interface to a multi-threaded processor according to exemplary embodiments.
• FIG. 4 illustrates an exemplary wireless communication system 400 in which an embodiment of the disclosure may be advantageously employed.
• exemplary embodiments are directed to low latency interrupt controllers configured for high rate of interrupt processing. More particularly, embodiments may include two-level low latency interrupt controllers which may be interfaced to multi-threaded processor cores.
• system 200 comprising an LI interrupt controller 204, which may be configured for direct interfacing with multi-threaded processor 206.
• LI interrupt controller 204 may handle interrupts directed to the multithreaded processor from one or more devices, including L2 interrupt controller 202.
• both LI and L2 interrupt controllers 204 and 202 may be a Vectored Interrupt Controller, such as previously described. Accordingly, the L2 interrupt controller 202 may be configured to send interrupts along with vector addresses associated with ISRs of the interrupts. According to a non-limiting illustration, the L2 interrupt controller 202 may support up to 1024 low-latency interrupts. The 1024 low latency interrupts may be prioritized by L2 interrupt controller 202. LI interrupt controller 204 may be a VIC with 32 register entries [31:0], such that register entry [31] of LI interrupt controller 204 may correspond to L2 interrupt controller 202.
• the remaining 31 register entries of LI interrupt controller 204 may be reserved for interrupts from other sources, including interrupts that are generated internally from the two or more threads of multi-threaded processor 206 for thread-to- thread signaling, or for legacy applications.
• all interrupts from external devices may first be received at L2 interrupt controller 202 and thereafter may be directed to LI interrupt controller 204.
• a first interrupt may be received by L2 interrupt controller 202 from an external device (not shown).
• the first interrupt may be communicated to LI interrupt controller 204 over bus 208, along with corresponding vector ID on bus 210.
• a global vector ID register (not shown) may also be updated with the vector ID, wherein the global vector ID register may be accessible through a control register (CR) transfer instruction.
• the global vector ID register may assist in tracking interrupts.
• the global VID register may be configured to track which specific L2 interrupt was sent to LI interrupt controller 204.
• embodiments may deviate from conventional techniques in several ways in order to expedite the handling of subsequent requests. While conventional techniques such as illustrated in FIG. 1, relied on hardware solutions to inform L2 interrupt controller 102 over network 110 regarding the availability of core 106 for accepting a new request, embodiments may include software routines configured for monitoring the readiness of multi-threaded processor 206. For example, a software routine may efficiently track the state of processes running on the two or more threads of multi-threaded processor 206. The software routine may determine whether one or more threads may be in a WAIT state to immediately accept a new interrupt, or if a low priority process running on a thread may be interrupted to immediately service a new interrupt.
• embodiments may include a single instruction to accomplish both: informing LI interrupt controller 204 that the first interrupt has been accepted for processing and also informing L2 interrupt controller 202 that multi-threaded processor 206 is now ready for accepting a new interrupt. It will be recognized that these embodiments are not restrained by a conventional two-level interrupt framework, such as illustrated in FIG. 1, wherein core 106 may communicate its readiness to L2 interrupt controller 102 for a new interrupt only through network 110. On the other hand, embodiments may communicate the readiness of multi-threaded processor 206 directly to L2 interrupt controller 202, through a dedicated hardware line, such as line 212 in FIG. 2.
• some embodiments may also include a dedicated hardware port in multithreaded processor 206, such that L2 interrupt controller 202 may directly communicate interrupts (e.g. "New interrupt" in FIG. 2) to multi-threaded processor 206 via a dedicated hardware line, such as line 214.
• interrupts e.g. "New interrupt” in FIG. 2
• line 214 a dedicated hardware line
• Exemplary embodiments may include a Clear Interrupt Auto Disable (CIAD) register.
• CIC Clear Interrupt Auto Disable
• a CIAD instruction may be used to ensure that the same interrupt is not taken more than once. For example, multi-threaded processor 206 may automatically set the CIAD register as soon as a first interrupt is taken. Once it is determined, for example by a software routine as discussed previously, that multi-threaded processor 206 is ready for accepting a new interrupt, the software routine may trigger a CIAD instruction to be issued. The CIAD instruction may then clear the CIAD register to activate the interrupt line such that a new interrupt may be taken on the same line.
• the CIAD instruction may be issued by multi-threaded processor 206 to both clear a status indicating that the first interrupt is pending at LI interrupt controller 204 and also inform L2 interrupt controller 202 that multi-threaded processor 206 is ready to receive another interrupt and accompanying vector ID over buses 208 and 210 respectively.
• the CIAD instruction may be associated with register entry [31] of LI interrupt controller 204, which as previously described, may be dedicated for interrupts from L2 interrupt controller 202.
• the CIAD instruction may provide an effective handshake mechanism between L2 interrupt controller 202, LI interrupt controller 204 and multi-threaded processor 206.
• the process such as, a software routine, may be started to generate the CIAD instruction, for expediting the handling of subsequent interrupts. It will be understood that embodiments may also implement the above processes with dedicated hardware or a combination of hardware and software, without being limited to a software routine as discussed in the aforementioned example.
• embodiments may include implementations wherein register entry [31] of LI interrupt controller 204 may be programmed to capture information on a rising edge of a clock.
• L2 interrupt controller 202 may be either edge triggered or level sensitive, and send interrupts and corresponding vector IDs to LI interrupt controller 204 over buses 208 and 210 asynchronously.
• LI interrupt controller 204 By configuring LI interrupt controller 204 as edge triggered, the interrupts may be synchronized to a clock corresponding to multi-threaded processor 206.
• edge triggered configurations may enable improved communication protocols between LI interrupt controller 204 and L2 interrupt controller 202.
• multi-threaded processor 206 may begin the transition to a state of readiness to accept a new interrupt, for example, by clearing register entry [31] of LI interrupt controller 204.
• embodiments may significantly improve the rate of processing interrupts and also decrease the latency of interrupt processing.
• an embodiment can include a method for reducing interrupt latency time, comprising: coupling a first interrupt controller, such as LI interrupt controller 204, to a core, such as multi-threaded processor 206 (Block 302); communicating a first interrupt and a first vector identifier from a second interrupt controller, such as L2 interrupt controller 202 to the first interrupt controller, for example over buses 208 and 210 respectively (Block 304); processing the first interrupt and the first vector identifier at the first interrupt controller (Block 306); sending the processed interrupt to a thread in the core (Block 308); determining when the core is ready to receive a second interrupt (Block 310); and sending an instruction from the core to the second interrupt controller indicating the core is ready to receive the second interrupt (Block 312).
• a first interrupt controller such as LI interrupt controller 204
• a core such as multi-threaded processor 206
• a software module may reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
• An exemplary storage medium is coupled to the processor such that the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium may be integral to the processor.
• an embodiment of the invention can include a computer readable media embodying a method for reducing interrupt latency in a two-level interrupt controller interface to a multi-threaded processor. Accordingly, the invention is not limited to illustrated examples and any means for performing the functionality described herein are included in embodiments of the invention.
• FIG. 4 a block diagram of a particular illustrative embodiment of a wireless device that includes a multi-core processor configured according to exemplary embodiments is depicted and generally designated 400.
• the device 400 includes a digital signal processor (DSP) 464 which may include system 200 of FIG. 2.
• DSP digital signal processor
• FIG. 4 also shows display controller 426 that is coupled to DSP 464 and to display 428.
• Coder/decoder (CODEC) 434 e.g., an audio and/or voice CODEC
• Other components, such as wireless controller 440 (which may include a modem) are also illustrated.
• Speaker 436 and microphone 438 can be coupled to CODEC 434.
• wireless controller 440 can be coupled to wireless antenna 442.
• DSP 464, display controller 426, memory 432, CODEC 434, and wireless controller 440 are included in a system-in- package or system-on-chip device 422.
• input device 430 and power supply 444 are coupled to the system-on-chip device 422.
• display 428, input device 430, speaker 436, microphone 438, wireless antenna 442, and power supply 444 are external to the system-on-chip device 422.
• each of display 428, input device 430, speaker 436, microphone 438, wireless antenna 442, and power supply 444 can be coupled to a component of the system-on-chip device 422, such as an interface or a controller.
• FIG. 4 depicts a wireless communications device
• DSP 464 and memory 432 may also be integrated into a set-top box, a music player, a video player, an entertainment unit, a navigation device, a personal digital assistant (PDA), a fixed location data unit, or a computer.
• a processor e.g., DSP 464 may also be integrated into such a device.
• the foregoing disclosed devices and methods are typically designed and are configured into GDSII and GERBER computer files, stored on a computer readable media. These files are in turn provided to fabrication handlers who fabricate devices based on these files. The resulting products are semiconductor wafers that are then cut into semiconductor die and packaged into a semiconductor chip. The chips are then employed in devices described above.

Landscapes

• Engineering & Computer Science (AREA)
• Theoretical Computer Science (AREA)
• Physics & Mathematics (AREA)
• General Engineering & Computer Science (AREA)
• General Physics & Mathematics (AREA)
• Bus Control (AREA)
• Human Computer Interaction (AREA)

Priority Applications (4)

Applications Claiming Priority (2)

Publications (2)

Family

ID=47172881

Family Applications (1)

Country Status (6)

Families Citing this family (9)

Family Cites Families (41)

• 2011
  • 2011-10-04 US US13/252,670 patent/US8972642B2/en active Active
• 2012
  • 2012-10-04 WO PCT/US2012/058780 patent/WO2013052684A2/en not_active Ceased
  • 2012-10-04 KR KR1020147012326A patent/KR101563576B1/ko not_active Expired - Fee Related
  • 2012-10-04 CN CN201280048379.2A patent/CN103874990B/zh active Active
  • 2012-10-04 JP JP2014534728A patent/JP5847949B2/ja not_active Expired - Fee Related
  • 2012-10-04 EP EP12784381.1A patent/EP2764442B1/en not_active Not-in-force
• 2015
  • 2015-11-25 JP JP2015229471A patent/JP6153991B2/ja not_active Expired - Fee Related

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events