Patent Grant
11249776
This application claims priority from European patent application number EP19157095.1, filed Feb. 14, 2019, which is hereby incorporated herein by reference in its entirety.
One or more aspects of the present disclosure relate, in general, to interrupt processing within a computer system, and in particular, to handling interrupts generated by bus connected modules in a multiprocessor computer system.
Interrupts are used to signal to a processor that an event needs the attention of the processor. For example, hardware devices, e.g., hardware devices connected with the processors via a bus, use interrupts to communicate that they require attention from an operating system. In case the receiving processor is currently performing some activities, the receiving processor may in response to receiving an interrupt signal suspend its current activities, save its state, and handle the interrupt, e.g., by executing an interrupt handler. The interruption of the processors' current activities resulting from the receipt is only temporary. After having handled the interrupt, the processor may resume its suspended activities. Thus, interrupts may allow for a performance improvement by eliminating unproductive waiting time of a processor in polling loops, waiting for external events.
In multiprocessor computer systems, interrupt routing efficiency issues may arise. The challenge is to forward interrupt signals sent by hardware devices, like, e.g., bus connected modules, to a processor of the multiple processors assigned for use by the operating system in an efficient way. This may be particularly challenging in case the interrupt is used to communicate with a guest operating system on a virtual machine. A hypervisor or virtual machine monitor (VMM) creates and runs one or more virtual machines, i.e., guest machines. A virtual machine provides a guest operating system executed on the same machine with a virtual operating platform, while hiding the physical characteristics of the underlying platform. Using multiple virtual machines allows to run multiple operating systems in parallel. Since being executed on a virtual operating platform, the guest operating system's view of the processors may in general differ from an underlying, e.g., physical view of the processors. The guest operating system uses virtual processor IDs to identify processors, which in general do not coincide with underlying logical processor IDs. The hypervisor which manages the execution of the guest operating system defines a mapping between underlying logical processor IDs and virtual processor IDs used by the guest operating system. However, this mapping and the selection of processors scheduled for use by the guest operating system are not static, but may be changed by the hypervisor, while the guest operating system is running, without the knowledge of the guest operating system.
Typically, this challenge is solved by forwarding interrupt signals using broadcasting. When using the broadcasting, the interrupt signal is successively forwarded between the multiple processors until a processor suitable for handling the interrupt signal is met. However, in case of multiple processors, the probability that a processor receiving the broadcasted interrupt signal first is indeed suitable for handling the interrupt signal may be rather low. Furthermore, being suitable for handling the interrupt signal not necessarily means that the respective processor is the best choice for handling the interrupt.
Shortcomings of the prior art are overcome and additional advantages are provided through the provision of a computer program product for providing an interrupt signal to a guest operating system executed using one or more processors of a plurality of processors of a computer system assigned for usage by the guest operating system. The computer program product includes at least one computer readable storage medium readable by at least one processing circuit and storing instructions for performing a method. The method includes receiving, by a bus attachment device from a bus connected module of a plurality of bus connected modules operationally coupled to the plurality of processors via the bus attachment device, an interrupt signal with an interrupt target ID. The interrupt target ID identifies one processor of the plurality of processors assigned for usage by the guest operating system as a target processor for handling the interrupt signal. The bus attachment device retrieves a first copy of an interrupt table entry assigned to the interrupt target ID from an interrupt table stored in a memory operationally connected with the bus attachment device. The first copy of the interrupt table entry comprises a first copy of a running indicator to indicate whether the target processor identified by the interrupt target ID is scheduled for usage by the guest operating system. The bus attachment device checks, using the first copy of the running indicator, whether the target processor is scheduled for usage by the guest operating system. Based on the target processor being scheduled for usage by the guest operating system, the bus attachment device translates the interrupt target ID to a logical processor ID and forwards the interrupt signal to the target processor to handle. The forwarding uses the logical processor ID resulting from the translating to address the target processor directly. Based on the target processor not being scheduled for usage by the guest operating system, the bus attachment device forwards the interrupt signal to handle to the plurality of processors using broadcasting.
Methods and systems relating to one or more aspects are also described and claimed herein.
Additional features and advantages are realized through the techniques described herein. Other embodiments and aspects are described in detail herein and are considered a part of the claimed aspects.
In the following, embodiments of aspects of the invention are explained in greater detail, by way of example only, making reference to the drawings in which:
The descriptions of the various embodiments of aspects of the present invention will be presented for purposes of illustration, but are not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of aspects of the described embodiments. The terminology used herein was chosen to best explain the principles of the embodiments, the practical application or technical improvement over technologies found in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Embodiments may have the beneficial effect of enabling the bus attachment device to address the target processor directly. Thus, the interrupt signal may be targeted by the issuing bus connected module selecting a target processor ID to a particular processor, i.e. target processor, of a multiprocessor computer system. For example, a processor may be selected as a target processor for the interrupt signal, which has performed activities related to the interrupt before. Handling the interrupt signal by the same processor as the respective activities may result in a performance advantage, since in case the same processor is also handling the interrupt signal, the data (e.g., all the data) in context with this interrupt may already be available to the processor and/or stored in a local cache enabling a fast access for the respective processor without requiring extensive cache traffic.
Thus, a broadcasting of the interrupt signal may be avoided for which there is no guarantee that the processor which will handle the interrupt in the end is best suited for this task from a performance point of view, like cache traffic minimization. Rather than presenting the interrupt signal to all processors, each processor trying to handle it and one processor wins, the interrupt signal may be provided directly to a target processor increasing the efficiency of the interrupt signal handling.
An interrupt mechanism may be implemented using directed interrupts. The bus attachment device may be enabled to directly address a target processor using a logical processor ID of the same processor, when forwarding an interrupt signal for handling to its target processor defined by the issuing bus connected module. Translating interrupt target IDs to logical processor IDs by the bus connected device may further ensure that from a point of view of the guest operating system the same processor is addressed, even though the mapping between interrupt target IDs and logical processor IDs or the selection of processors scheduled for use by the guest operating system may be changed by the hypervisor.
Embodiments may have the beneficial effect of preventing of interrupts targeted at processors not running, i.e., not scheduled for usage by the guest operating system. Embodiments may have the beneficial effect of supporting the hypervisor to reschedule processors.
A running indicator indicates whether the target processor identified by the interrupt target ID received with an interrupt signal is scheduled for usage by the guest operating system. The running indicator may for example be implemented in a form of a running bit, i.e., a single bit indicating whether a processor assigned with the respective bit is running, i.e., is scheduled for usage by the guest operating system. Thus, an enabled running bit may tell the bus attachment device that the target processor is currently scheduled, while an unenabled running bit may tell the bus attachment device that the target processor is currently not scheduled. In case the target processor is not running, the bus attachment device may send a fallback broadcast interrupt request right way, without trying to address one of the processors directly.
In accordance with an embodiment, the interrupt signal is received in a form of a message signaled interrupt comprising the interrupt target ID of the target processor. Using message signaled interrupts (MSI) is a method for a bus connected module, such as a Peripheral Component Interconnect (PCI) or Peripheral Component Interconnect express (PCIe) function, to generate a central processing unit (CPU) interrupt in order to notify a guest operating system using the respective central processing unit of the occurrence of an event or the presence of some status. MSI provides an in-band method of signaling an interrupt, using special in-band messages, thereby avoiding a need for dedicated paths separated from a main data path to send such control information, such as dedicated interruption pins on each device. MSI rather relies on exchanging special messages indicating interrupts through the main data path. When a bus connected module is configured to use MSI, the respective module requests an interrupt by performing an MSI write operation of a specified number of bytes of data to a special address. The combination of this special address, i.e., MSI address, and a unique data value, i.e. MSI data, is termed an MSI vector.
Modern PCIe standard adapters have the capability to present multiple interrupts. MSI-X, for example, permits a bus connected module to allocate up to 2048 interrupts. Thus, targeting individual interrupts to different processors, such as in a high-speed networking application relying on multiprocessor systems, is enabled. MSI-X allows to assign a multitude of interrupts, each with an individual MSI address and MSI data value.
For transmitting an interrupt signal, an MSI-X message may be used. The required content of the MSI-X message may be determined using an MSI-X data table. The MSI-X data table local to the bus connected module, i.e. PCIe adapter/function, may be indexed by a number assigned to each interrupt signal, also referred to as an interrupt request (IRQ). The MSI-X data table content is under control of the guest operating system and may be set to the operating system by guidance of hardware and/or firmware. A single PCIe adapter may comprise multiple PCIe functions, each of which may have an independent MSI-X data table. This may, for example, be the case for a single root input/output virtualization (SR-IOV) or multi-function devices.
An interrupt target ID, like, e.g., a virtual processor ID, may be directly encoded as part of a message sent by the bus connected module, like, e.g., an MSI-X message, comprising the interrupt signal. The message, e.g., an MSI-X message, may comprise a requestor ID, i.e. an ID of the bus connected module, the aforementioned interrupt target ID, a DIBV (directed interrupt signal vector) or an AIBV (interrupt signal vector) index, an MSI address and MSI data. An MSI-X message may provide 64 bits for the MSI address and 32 bits for the data. A bus connected module may request an interrupt using MSI by performing an MSI write operation of a specific MSI data value to a special MSI address.
The device table is, for instance, a shared table which may be fully indexed by the requestor ID (RID) of the interrupt requestor, i.e., the bus connected module. The bus attachment device remaps and posts the interrupt, i.e., the bus attachment device translates the interrupt target ID and uses the same to directly address the target processor.
A guest operating system may use virtual processor IDs to identify processors in a multiprocessor computer system. Thus, the guest operating system's view of processors may not be identical to a view of an underlying system using logical processor IDs. Bus connected modules providing resources used by a guest operating system may use virtual processor IDs as the resources for communicating with the guest operating system, like, e.g., an MSI-X data table, may be under the control of the guest operating system. As an alternative to a virtual processor ID, any other ID may be defined for the bus connected module to address the processors.
The interrupt is presented to the guest operating system or other software executed thereon, such as other programs, etc. As used herein, the term operating system includes operating system device drivers.
As used herein, the term bus connected module may comprise any type of bus connected module. According to embodiments, the module may be a hardware module, like, e.g., a storage function, processing module, network module, cryptographic module, PCI/PCIe adapter, other type of input/output module, etc. According to other embodiments, the module may be a software module, i.e., a function, like, e.g., a storage function, processing function, network function, cryptographic function, PCI/PCIe adapter, other type of input/output function, etc. Thus, in the examples presented herein, module is used interchangeably with function, e.g., PCI/PCIe function, and adapter, e.g., PCI/PCIe adapter, unless otherwise noted.
Embodiments may have the benefit that an interrupt signal routing mechanism, e.g., MSI-X message routing mechanism, is provided which allows it to keep bus connected modules, e.g., PCIe adapters and functions, as well as device drivers used to operate or control the bus connected modules, unchanged. Furthermore, the hypervisor may be kept from intercepting the underlying architecture for implementing communication between bus connected modules and a guest operating system, e.g., PCIe MSI-X architecture. In other words, changes to the interrupt signal routing mechanism may be implemented outside the hypervisor and bus connected modules.
In accordance with an embodiment, the running indicator is implemented as a single bit. Embodiments may have the beneficial effect that the running indicator is provided in a form with minimum memory space and is fast and efficiently processable.
In accordance with an embodiment, the first copy of the interrupt table entry further comprises a first mapping of the received interrupt target ID to a first one of the logical processor IDs, the bus attachment device using the first copy of the interrupt table entry for the translating of the received interrupt target ID to the logical processor ID of the target processor.
Embodiments may have the beneficial effect of providing an interrupt table (IRT) comprising interrupt table entries (IRTE), each entry providing a mapping of an interrupt target ID to a logical processor ID. Thus, the entries may define a unique assignment for each interrupt target ID to a logical processor ID. According to embodiments, the interrupt target ID may be provided in a form of a virtual processor ID. According to embodiments, the interrupt target ID may be any other ID used by the guest operating system to identify individual processors used.
According to embodiments, the IRT is provided in the memory for use by the bus attachment device to map interrupt target IDs to logical processor IDs. According to embodiments, the IRT may be provided in a single location. An address indicator, e.g., a pointer, indicating a memory address of the IRT may be provided. The address indicator may, for example, be provided by an entry of a device table fetched from the memory by the bus attachment device. Embodiments may have the beneficial effect that no large mapping tables have to be stored in the bus attachment device. The interrupt table used for mapping may rather be stored in the memory and accessed by the bus attachment device, e.g., if necessary. Thus, the bus attachment device may only have to handle a working copy of one or more interrupt table entries for each interrupt signal to be forwarded. The number of interrupt table entries may, for instance, be small, e.g. one.
According to embodiments, the IRT or individual IRTE may be updated upon a rescheduling of the processors. According to embodiments, the IRT may be stored in an internal section of the memory, i.e., an HSA (hardware system area).
In accordance with an embodiment, the first copy of the interrupt table entry further comprises an interrupt blocking indicator indicating whether the target processor identified by the interrupt target ID is currently blocked from receiving interrupt signals. The method further comprises checking by the bus attachment device using the interrupt blocking indicator whether the target processor is blocked from receiving interrupt signals. If the target processor is unblocked, continuing with the forwarding of the interrupt signal, else blocking by the bus attachment device the interrupt signal from being forwarded to the target processor for handling.
According to embodiments, a direct interrupt blocking indicator is introduced in the interrupt entries of an interrupt table in memory. The direct interrupt blocking indicator may be implemented in a form of a single bit, i.e., a dIBPIA bit.
According to embodiments, the IRTE is fetched from the memory, and a running indicator is checked in order to determine whether the target processor is scheduled. In case the target processor is scheduled, a direct interrupt blocking indicator is enabled, in order to block the target processor from receiving a further interrupt signal, while handling the current interrupt signal. A further interrupt signal may otherwise interfere with the handling of the current interrupt signal. In order to ensure that the target processor has not been rescheduled in the meanwhile, the IRTE is re-fetched and the current running indicator is checked again in order to determine whether the target processor is still scheduled. In case the target processor is still scheduled, the interrupt signal may be forwarded to the target processor using a logical processor ID of the target processor to address the target processor directly. In addition, it may be checked whether the logical processor ID of the target processor provided by the IRTE for the received interrupt target ID is still the same.
In accordance with an embodiment, the method further comprises forwarding by the bus attachment device the interrupt signal for handling to the remaining processors of the plurality of processors using broadcasting.
In accordance with an embodiment, the method further comprises checking by an interrupt handler of the guest operating system whether any interrupts addressed to the target processor are pending for handling by the target processor. If no interrupts addressed to the target processor are pending for handling by the target processor, changing by the guest operating system the interrupt blocking indicator in the interrupt table entry assigned to the target processor to indicate the target processor is unblocked. Embodiments may have the beneficial effect of keeping track whether individual target processors are blocked.
In accordance with an embodiment, the method further comprises if the target processor is unblocked, changing by the bus attachment device the interrupt blocking indicator in the interrupt table entry assigned to the interrupt target ID to indicate the first logical processor ID being blocked. The changing is performed before the forwarding of the interrupt signal to the target processor for handling.
In accordance with an embodiment, the method further comprises retrieving by the bus attachment device after the changing of the interrupt blocking indicator a second copy of the interrupt table entry assigned to the received interrupt target ID; and checking the second copy of the interrupt table entry by the bus attachment device to exclude a predefined type of change of the second copy of the interrupt table relative to the first copy of the interrupt table entry. A successful exclusion of the predefined type of change being, e.g., required for the forwarding of the interrupt signal to the target processor for handling. Embodiments may have the beneficial effect of taking into account potential intermediate changes.
In accordance with an embodiment, the predefined type of change is a change of the first mapping of the received interrupt target ID relative to a second mapping of the received interrupt target ID to a second one of the logical processor IDs comprised by the second copy of the interrupt table entry. If the second mapping comprises a change relative to the first mapping, the interrupt signal is forwarded by the bus attachment device for handling to the plurality of processors using broadcasting.
In accordance with an embodiment, the predefined type of change is a change of the first copy of the running indicator relative to a second copy of the running indicator comprised by the second copy of the interrupt table entry. If the second copy of the running indicator comprises a change relative to the first copy of the running indicator (e.g., bit), the second running indicator indicates the target processor not being scheduled for usage by the operating system, and the interrupt signal is forwarded by the bus attachment device for handling to the plurality of processors using broadcasting.
According to embodiments, a double fetch of the IRTE may be performed, in order to prevent the interrupt signal from being sent to a processor which, e.g., has been deactivated in the meantime. According to embodiments, after forwarding the interrupt signal to the processor identified by the logical processor ID resulting from the translation of the interrupt target ID using a first copy of the IRTE, a second copy of the same IRTE may be fetched to check whether in the meantime any changes of the IRTE took place. In case the IRTE has been updated in the meantime, there is a risk that the interrupt signal has been forwarded to a deactivated processor. Therefore, the second copy of the IRTE may be used to translate the interrupt target ID again and forward the interrupt signal to the processor identified by the logical processor ID resulting from the second translation. According to an alternative embodiment, the complete method starting with the fetching of the first copy of the IRTE may be repeated in case the second copy of the IRTE does not match the first copy. For example, a third copy of the IRTE may be fetched replacing the first copy of the IRTE or the second copy of the IRTE may replace the first copy of the IRTE and a third copy of the IRTE may be fetched to implement a double fetch scheme also for the partial repetition of the method. This scheme may be repeated until a match is achieved. According to further alternative embodiments, the interrupt signal may be forwarded using broadcasting in case the second copy of the IRTE does not match the first copy. According to embodiments, the bus attachment device participates in the memory-cache-coherence protocol and detects alternations on the IRTE, e.g., cache snooping, by the same mechanism CPUs may detect cache-line-alternations.
Embodiments may have the beneficial effect of avoiding cache purges which may have an inefficient scaling. The double fetch may be global or specific to an IRTE, i.e., the whole entry may be subject to the double fetching or restricted to specific information comprised by the respective entry.
According to embodiments, the condition resulting from the time required for translating the interrupt target ID and forwarding the interrupt signal to the target processor, until it reaches the processor may be caught by a checking logic on a CPU checking whether the receiving processor is still the correct target processor. For the check, an interrupt target ID and/or logical partition ID received with the interrupt request may be compared with a current interrupt target ID and/or logical partition ID assigned to the receiving processor as a reference. In case of a match, the receiving processor directly addressed using the logical processor ID resulting from the translation using the copy of the IRTE is indeed the correct target processor. Thus, the information provided by the copy of the IRTE has been up-to-date. In case of a mismatch, the copy of the IRTE has not been up-to-date and the receiving processor is not the target processor anymore. In case of a mismatch, the interrupt signal may, e.g., be forwarded to the target operating system using broadcasting.
According to embodiments, there may be three entities operating in parallel, i.e. the bus attachment device and the target processor processing the interrupt signal, as well as the hypervisor which may change the assignments between interrupt target IDs and logical processor IDs. According to embodiments, there may be no central synchronization point in physically distributed systems besides memory which provides a virtual appearance of such a system at a latency cost. Embodiments using a double fetch scheme may have the beneficial effect of providing a method which is optimized for speed vs preventing double deliveries or even misses of interrupt requests.
In view of the interrupt signal, the following actions may be performed: A1) Reading a first copy of the IRTE, A2) Sending an interrupt request to a processor directly addressed, and A3) Reading a second copy of the IRTE. Simultaneously, the following sequence regarding a change of assignment between an interrupt target ID and a logical processor ID may occur: B1) Activate additional processor with additional logical processor ID and deactivate previous processor with previous logical processor ID, and B2) Update IRTE with additional logical processor ID, i.e., replacing the previous logical processor ID by additional logical processor ID.
In certain error cases, a processor, e.g., the target processor, may be reset to a checkpoint and lose intermediate information. In order to regain the lost information, the processor may scan IRTE entries (e.g., all IRTE entries) for this specific processor, i.e. assigned to its logical processor ID, and deliver direct interrupt requests as indicated by the pending direct interrupt indicators, e.g., dPIA bits, present in memory which does not get impacted by a processor recovery.
The pending direct interrupt indicator comprised by the IRTE, e.g., IRTE.dPIA bit, may be used as a master copy, i.e., a single point of truth, if an interrupt signal should be presented. In order to simplify processor recovery, a pending direct interrupt indicator in the processor may be used, e.g., as a shadow-copy of the IRTE.dPIA bit to keep direct interrupts pending on the processor.
In case the memory has a strictly ordered nature, e.g., only the following sequences may be possible in view of steps A1, A2, and B1: alternative 1 with A1→A3→B1 and alternative 2 with A1→B1→→A3. In case of alternative 1, the first and the second copy of the IRTE may match. Thus, the interrupt signal may be forwarded to the previous processor instead of the current target processor. The previous processor may see a mismatch, regarding the interrupt target ID and/or logical partition ID and initiate a broadcast of the received interrupt signal. In case of alternative 2, the bus attachment device may see a mismatch between the first and the second copy of the IRTE. In response to the mismatch, the bus attachment device may broadcast the interrupt signal. Due to the broadcasting, the interrupt signal may be received by the additional processor which sees a hit and directly processes the received interrupt request. Embodiments may have the beneficial effect of closing the timing window through an over-initiative-approach.
In accordance with an embodiment, the method further comprises retrieving by the bus attachment device a copy of a device table entry from a device table stored in the memory. The device table entry comprises a direct signaling indicator indicating whether the target processor is to be addressed directly. If the direct signaling indicator indicates a direct forwarding of the interrupt signal, the forwarding of the interrupt signal using a logical processor ID of the target processor to address the target processor directly is executed; else forwarding by the bus attachment device the interrupt signal for handling to the plurality of processors using broadcasting.
Embodiments may have the beneficial effect of controlling with the direct signaling indicator whether an interrupt signal is forwarded using direct addressing or broadcasting. Using the direct signaling indicator for each bus connected module an individual predefined selection may be provided of whether for interrupt signals received from this bus connected module a direct addressing is to be performed or a broadcasting.
In accordance with an embodiment, the device table entry further comprises an interrupt table address indicator indicating a memory address of the interrupt table, and the bus attachment device uses the memory address of the interrupt table for the retrieving of the first copy of the interrupt table entry. Embodiments may have the beneficial effect of providing the bus attachment device efficiently with a mapping of interrupt target IDs to logical processor ID as defined by a hypervisor of the system.
In accordance with an embodiment, the memory further comprises an interrupt summary vector, and the device table entry further comprises an interrupt summary vector address indicator indicating a memory address of the interrupt summary vector. The interrupt summary vector comprises an interrupt summary indicator per bus connected module, and each interrupt summary indicator is assigned to a bus connected module indicating whether there is an interrupt signal issued by the respective bus connected module to be handled, wherein the method further comprises using by the bus attachment device the indicated memory address of the interrupt summary vector to update the interrupt summary indicator assigned to the bus connected module from which the interrupt signal is received such that the updated interrupt summary indicator indicates that there is an interrupt signal issued by the respective bus connected module to be handled.
Embodiments may have the beneficial effect of monitoring and recording from which bus connected modules there are interrupt signals to be handled. This information may be particularly helpful in case a broadcast has to be performed, e.g., as a fallback in case a direct addressing fails or is not available.
In accordance with an embodiment, the memory further comprises a directed interrupt summary vector, and the device table entry further comprises a directed interrupt summary vector address indicator indicating a memory address of the directed interrupt summary vector. The directed interrupt summary vector comprises a directed interrupt summary indicator per interrupt target ID, and each directed interrupt summary indicator is assigned to an interrupt target ID indicating whether there is an interrupt signal addressed to the respective interrupt target ID to be handled. The method further comprises using by the bus attachment device the indicated memory address of the directed interrupt summary vector to update the interrupt summary indicator assigned to the target processor ID to which the received interrupt signal is addressed such that the updated interrupt summary indicator indicates that there is an interrupt signal addressed to the respective interrupt target ID to be handled.
When an interrupt cannot be delivered directly, e.g., because the hypervisor has not scheduled the target processor, the guest operating system may benefit by delivering the interrupt with the originally intended affinity, i.e. information for which processor the interrupt was intended, using broadcast. In this case, the bus attachment device may set a bit specifying the target processor in the DISB after setting the DIBV and before delivering a broadcast interrupt request to the guest operating system. If a guest operating system receives a broadcast interrupt request, it may therefore identify by scanning and disabling direct interrupt summary indicators in the DISB, e.g., scanning and resetting direct interrupt summary bits, which target processors have an interrupt signal pending as signaled in a DIBV. The guest operating system may thus be enabled to decide, whether the interrupt signal is handled by the current processor which received the broadcast or further forwarded to the original target processor.
In accordance with an embodiment, the memory further comprises one or more interrupt signal vectors, and the device table entry further comprises an interrupt signal vector address indicator indicating a memory address of an interrupt signal vector of the one or more interrupt signal vectors. Each of the interrupt signal vectors comprises one or more signal indicators, and each interrupt signal indicator is assigned to a bus connected module of the one or more bus connected modules and an interrupt target ID indicating whether an interrupt signal has been received from the respective bus connected module addressed to the respective interrupt target ID. The method further comprises: using by the bus attachment device the indicated memory address of the interrupt signal vector to select the interrupt signal indicator assigned to the bus connected module which issued the received interrupt signal and to the interrupt target ID to which the received interrupt signal is addressed; and updating the selected interrupt signal indicator such that the selected interrupt signal indicator indicates that there is an interrupt signal issued by the respective bus connected module and addressed to the respective interrupt target ID to be handled.
Embodiments may have the beneficial effect of providing a monitoring and recording regarding individual bus connected modules for which target processors they have issued interrupt signals.
In accordance with an embodiment, each interrupt signal vector comprises an interrupt signal indicator per interrupt target ID assigned to the respective interrupt target ID, and each of the interrupt signal vectors is assigned to an individual bus connected module with the interrupt signal indicators of the respective interrupt signal vector being further assigned to the respective individual bus connected module.
In accordance with an embodiment, each interrupt signal vector comprises an interrupt signal indicator per bus connected module assigned to the respective bus connected module, and each of the interrupt signal vectors is assigned to an individual target processor ID with the interrupt signal indicators of the respective interrupt signal vector being further assigned to the respective target processor ID.
Thus, the interrupt signal vectors may be implemented, in one embodiment, as directed interrupt signal vectors ordered depending on the target processor ID, i.e., optimized for taking track of directed interrupts. In other words, an order criterium is the target processor IDs rather than requestor IDs identifying the issuing bus connected modules. Each directed interrupt signal vector may comprise one or more directed interrupt signal indicators depending on the number of bus connected modules.
An ordering of interrupt signal indicators, e.g., in a form of interrupt signaling bits, indicating an individual interrupt signal has been received, e.g., in a form of an MSI-X message, sequentially within a contiguous area of memory, like a cache line, for an individual bus connected module, like, e.g., a PCIe function, may thus be avoided. Enabling and/or disabling an interrupt signal indicator, e.g., by setting and/or resetting an interrupt signaling bit, in one embodiment, requires the respective contiguous area of memory to be moved to one of the processors to change the respective interrupt signal indicator accordingly.
It may be intended that a processor handles, e.g., all indicators for which it is responsible from a guest operating system perspective, i.e., in particular all indicators assigned to the respective processor. This may enable a performance advantage, since in case each processor is handling, e.g., all data assigned to the same a likelihood that data required in this context is provided to the processor and/or stored in a local cache may be high enabling a fast access to the respective data for the processor without requiring extensive cache traffic.
However, each processor trying to handle, e.g., all indicators for which it is responsible may nevertheless lead to a high cache traffic between the processors, as each processor is to write, e.g., all cache lines for, e.g., all functions, since the indicators assigned to each individual processor may be distributed over, e.g., all contiguous areas, such as cache lines.
The interrupt signaling indicators may be reordered in a form of directed interrupt signaling vectors such that, e.g., all interrupt signaling indicators assigned to the same interrupt target ID are combined in the same contiguous area of memory, e.g., cache line. Thus, a processor intending to handle indicators assigned to the respective processor, i.e., interrupt target ID, may only have to load a single contiguous area of memory. Thus, a contiguous area per interrupt target ID is used rather than a contiguous area per bus connected module. Each processor may only need to scan and update a single contiguous area of memory, e.g., a cache line for, e.g., all interrupt signals received from, e.g., all available bus connected modules targeted to that specific processor as a target processor identified by the interrupt target ID.
According to embodiments, an offset may be applied by the hypervisor for a guest operating system to align bits to different offsets.
In accordance with an embodiment, the device table entry further comprises a logical partition ID identifying a logical partition to which the guest operating system is assigned, and the forwarding of the interrupt signal by the bus attachment device further comprises forwarding with the interrupt signal the logical partition ID. Embodiments may have the beneficial effect of enabling the receiving processer to check to which guest operating system an interrupt signal is addressed.
In accordance with an embodiment, the method further comprises retrieving by the bus attachment device an interrupt subclass ID identifying an interrupt subclass to which the received interrupt signal is assigned, and the forwarding of the interrupt signal by the bus attachment device further comprises forwarding with the interrupt signal the interrupt subclass ID.
According to embodiments, the instructions provided on the computer readable non-transitory medium for execution by a processing circuit are configured for performing any of the embodiments of the method for providing an interrupt signal to a guest operating system, as described herein.
According to embodiments, the computer system is further configured to execute any of the embodiments of the method for providing an interrupt signal to a guest operating system, as described herein.
The bus connected modules 120 may, e.g., be provided in a form of Peripheral Component Interconnect express (PCIe) modules, also referred to as a PCIe adapter or PCIe functions provided by a PCIe adapter. A PCIe function 120 may issue a request which is sent to a bus attachment device 110, e.g., a PCI Host Bridge (PHB), also referred to a PCI Bridge Unit (PBU). The bus attachment device 110 receives the requests from the bus connected modules 120. The requests may, for example, comprise an input/output address that is used to perform a direct memory access (DMA) to memory 140 by the bus attachment device 110 or an input/output address that indicates an interrupt signal, e.g., a message signaled interrupt (MSI).
The virtual machines may use a V=V model, in which the memory assigned to a virtual machine is backed by virtual memory, instead of real memory. Thus, each virtual machine has a virtual linear memory space. The physical resources are owned by hypervisor 200, such as a VM hypervisor, and the shared physical resources are dispatched by the hypervisor to the guest operating systems, as needed, to meet their processing demands. The V=V virtual machine model assumes that the interactions between the guest operating systems and the physical shared machine resources are controlled by the VM hypervisor, since the large number of guests may preclude the hypervisor from simply partitioning and assigning the hardware resources to the configured guest.
The processors 130 are assignable to virtual machines 202 by the hypervisor 200. A virtual machine 202 may, e.g., be assigned with one or more logical processors. Each of the logical processors may represent all or a share of the physical processors 130 that may be dynamically allocated to the virtual machine 202 by the hypervisor 200. Virtual machines 202 are managed by a hypervisor 200. The hypervisor 200 may, for example, be implemented in firmware running on the processors 130 or may be a part of an operating system executed on the computer system 100. The hypervisor 200 may, for example, be a VM hypervisor, such as z/VM® offered by International Business Machines Corporation, Armonk, N.Y.
A bus connected module 120 addressing processors 130 used by a first/second level guest operating system may use a target processor ID in a form of the first/second virtual processor IDs vCPU 224, 226 or alternative IDs derived from the first/second virtual processor IDs vCPU 224, 226.
The memory furthermore includes, for instance, a device table (DT) 144 with a device table entry (DTE) 146 for each bus connected module 120. Upon receipt of an interrupt signal, e.g., an MSI-X write message with an interrupt target ID identifying a target processor for the interrupt request and a requestor ID identifying an origin of the interrupt request in a form of the bus connected module 120, the bus attachment device 110 fetches the DTE 146 assigned to the requesting bus connected module 120. The DTE 146 may indicate whether a directed addressing of the target processor is enabled for the requesting bus connected module 120, e.g., using a direct signaling indicator (e.g., a dIRQ bit). The bus attachment device updates an entry of a directed interrupt signal vector (DIBV) 162, as well as a directed interrupt summary vector (DISB) 160 in order to keep track for which of the processors 130 an interrupt signal has been received. The DISB 160 may comprise one entry per interrupt target ID, indicating whether there is an interrupt signal to be handled for this processor 130 from any bus connected module 120. Each DIBV 162 is assigned to one of the interrupt target IDs, i.e. processors 130, and may comprise one or more entries. Each entry is assigned to one of the bus connected modules 120. Thus, the DIBV indicates from which bus connected modules there are interrupt signals for a specific processor 130 to be handled. This may have the advantage that in order to check whether there is any interrupt signal or from which bus connected module 120 there are interrupt signals for a specific processer to be handled. Only a signal entry, e.g., bit, or only a signal vector, e.g. bit vector, has to be read from the memory 140. According to alternative embodiments, interrupt signal vectors (AIBV) and an interrupt summary vector (AISB) may be used. The AIBVs, as well as the entries of the AISB, are each assigned to a specific bus connected module 120.
The bus attachment device 110 uses, for instance, an entry (IRTE) 152 of an interrupt table (IRT) 150 stored in memory 140 to translate the interrupt target ID (IT_ID) to a logical processor ID (1CPU) and forwarding the received interrupt signal to the target processor using the logical processor ID to address the target processor directly. For the translation, the bus attachment device 110 fetches a copy 114 of an entry (IRTE) 152. The copy may either be fetched from a local cache or from memory 140 using an address of the interrupt table 150 (IRT@) provided by the copy of the DTE 146. The IRTE 152 provides a mapping of the interrupt target ID to a logical processor ID which is used by the bus attachment device 110 to address the target processor directly in case of a directed interrupt forwarding.
The IRTE 152 in addition provides a running indicator 154 indicating whether the target processor identified by the interrupt target ID is scheduled, i.e. running, at all. In case the target processor is not scheduled, a broadcast may be initiated in order to enable a timely interrupt handling.
Each processor includes, for instance, firmware, e.g., millicode 132, to receive and process direct interrupt signals. Firmware may further comprise, e.g., microcode and/or macrocode of the processors 130. It may comprise hardware-level instructions and/or data structures used in implementation of higher-level machine code. According to embodiments, it may comprise proprietary code that may be delivered as microcode that includes trusted software or microcode specific to the underlying hardware and controls operating system access to the system hardware. Furthermore, the firmware of the processors 130 may comprise a check logic 134 to check whether the receiving processor is identical with the target processor according to the interrupt target ID which is forwarded by the bus attachment device 110 to the receiving processor 130. In case the receiving processor 130 is not the target processor, i.e., in case of a mismatch of the received interrupt target ID and a reference interrupt target ID of the receiving processor 130, the interrupt signal is broadcasted to the logical partition in order to find a processor for handling the interrupt signal.
In step 340, the bus attachment device fetches a copy of an entry of a device table stored in the memory. The device table entry (DTE) provides address indicators of one or more vectors or vector entries to be updated in order to indicate that an interrupt signal has been received for the target processor. An address indicator for a vector entry may, e.g., comprise an address of the vector in the memory, as well as an offset within the vector. Furthermore, the DTE may provide a direct signaling indicator indicating whether the target processor is to be addressed directly by the bus attachment device using the interrupt target ID provided with the interrupt signal. Furthermore, the DTE may provide a logical partition ID, also referred to as a zone ID, and an interrupt subclass ID. The respective copy of the device table entry may be fetched from a cache or from memory.
Referring to
In step 360, the bus attachment device checks the direct signaling indicator (e.g., dIRQ bit) provided with the interrupt signal. In case the direct signaling indicator indicates no direct signaling, the bus attachment device forwards the interrupt signal in step 370 by broadcasting using the zone identifier and the interrupt subclass identifier, in order to provide the interrupt signal to a processor used by the guest operating system. In case the direct signaling indicator indicates direct signaling, the bus attachment device further checks in step 362 whether a running indicator comprised by the copy of the IRTE indicates that the target processor identified by the interrupt target ID is running.
In case the target processor is not running, the bus attachment device in step 364 sends a broadcast interrupt as a fallback using, e.g., a logical partition ID and/or an interrupt subclass ID to identify a processor suitable for handling the interrupt. In case no suitable processor matching the logical partition ID and/or the interrupt subclass ID is found, the hypervisor, i.e. a processor assigned for usage by the hypervisor, may receive the interrupt request instead of a processor assigned to the guest operating system. The hypervisor may decide on broadcasting the interrupt request again if one or more processors assigned to the guest operating system get scheduled. On an entry of a processor assigned to the operating system, the hypervisor may check for direct interrupt pending indicators, e.g., dPIA bits, to be presented to the entering processor. According to embodiments, the hypervisor may, e.g., selectively reschedule, i.e., wake up, the target processor.
In case the target processor is running, in step 380, the bus attachment device translates the interrupt target ID provided with the interrupt signal to a logical processor ID of a processor assigned for usage by the guest operating system. For the translation, the bus attachment device may use a mapping table comprised by the bus attachment device. The bus attachment device may comprise a mapping table or sub-table per zone, i.e. logical partition. In step 390, the bus attachment device forwards the interrupt signal to the target processor using the logical processor ID to address the respective processor directly, i.e., sending a direct message. The direct message may further comprise the zone ID and/or interrupt subclass ID.
In step 396, firmware, e.g., millicode, of the target processor receives the interrupt. In response, the firmware may interrupt its activities, e.g., program execution, and switches to execute an interrupt handler of the guest operating system. The interrupt may be presented to the guest operating system with a direct signaling indication. In case a checking logic is implemented on the receiving processor, a check may be performed checking whether the received interrupt target ID and/or logical partition ID matches an interrupt target ID and/or logical partition currently assigned to the receiving processor and accessible for the checking logic. In case of a mismatch, the receiving firmware may initiate a broadcast and broadcast the received interrupt request to the remaining processors using the logical partition ID and/or an interrupt subclass ID to identify a valid target processor for handling the interrupt.
In step 413a, it is checked whether a running indicator comprised by the IRTE is enabled. In case the running indicator is unenabled, the interrupt signal may be forwarded by the bus attachment device using broadcasting in step 413b. In case the running indicator is enabled, the bus attachment device continues in step 414 with translating the interrupt target ID to a logical processor ID using the fetched copy of the IRTE. In step 416, a message forwarding the interrupt signal to the target processor is sent using the logical processor ID to address the target processor directly. The message comprises, for instance, the interrupt target ID, a logical partition ID and an interrupt subclass ID. In step 418, the processor receives the message. In step 419, the processor checks whether the interrupt target ID and/or logical partition ID match the current interrupt target ID and/or logical partition ID provided as references for the check. In case of a match, the processor presents the interrupt request to the guest operating system in step 420. In case of a mismatch, the processor broadcasts the interrupt request to the other processors in step 422. Then, the processor continues its activities until the next interrupt message is received.
Furthermore, an example IRTE 152 is depicted in
In step 708, the PBU reads an entry of a device table (DT) assigned to the VF. Entries of DT stored in a hardware system area (HSA) of the memory are illustrated as rows of the table. The entry of DT may comprise an address of an interrupt table (IRT@), as well as a directed signaling bit (S) indicating whether directed signaling is to be performed. The PBU uses IRT@ to fetch an entry of the IRT assigned to the vCPU from the HSA comprising a running bit (R) indicating whether the vCPU is running, a directed interrupt blocking bit (dIBPIA) indicating whether the vCPU is blocked from receiving interrupts, as well as a directed interrupt pending bit (dPIA) indicating whether an interrupt directed to the vCPU is pending. At an earlier point in time in step 700, a start interpretive execution instruction (SW-entry) has been issued initiating a state change of the target processor from hypervisor mode to guest mode. In step 701, R is set to 1 in the IRTE assigned to the target processor, as well as a logical processor ID (TrgtPU #) of the target processor is provided. Then, the method ends with 702. For firmware and hardware, TrgtPU # refers to, e.g., a physical ID of a processing unit (1 physical PU), while for zOS and a logical partition (LPAR), TrgtPU # refers to, e.g., a logical ID of a processing unit (logical PU).
In step 710, the PBU sets a bit in the DIBV assigned to the vCPU using DIBV-Idx from MSI-X in order to indicate that there is an interrupt signal targeted to the vCPU from the VF. In step 712, the PBU checks whether the IRTE is blocked, i.e. whether IRTE.dIBIA==1. In case the IRTE assigned to the vCPU is blocked, and thus, the vCPU is blocked from receiving further interrupts, the method ends with step 714. In case the IRTE is unblocked, the method continues with step 716 in which it is checked by the PBU whether the vCPU is running, i.e. whether R is set in the IRTE.
If R is set, the method continues with step 718 executing a directed addressing. In step 718, d1BPIA and dPIA are set to 1 in the IRTE indicating that the vCPU is currently blocked from receiving interrupt signals and that an interrupt addressed to vCPU is pending. In step 720, it is checked whether the IRTE, more precisely the state of R and/or TrgtPU # of the IRTE, are changed compared to the IRTE in step 708. Thus, a double fetch scheme is implemented reading IRTE twice in order to ensure that no relevant changes have taken place between the readings, e.g., due to a SW-entry of another guest like shown in step 722 (
In step 722, a SIE-entry instruction for another guest is executed on the target processor. In step 724, the other guest reads the IRTE of a previous guest and issues in step 726 an atomic reset command for R, i.e. setting R=0 and indicating that the vCPU is not running anymore. Furthermore, dPIA is read from the IRTE. In step 728, it is checked whether dPIA is set (IRTE.dPIA==1) indicating that an interrupt for the vCPU is still pending. If no interrupt is pending, the method ends with step 730. If an interrupt is still pending, in step 732, a pending interrupt indicator PU.dPIA is reset on the target PU, as well as IRTE.dPIA of IRTE and a broadcast for the pending interrupt is initiated. Thus, in case a relevant change of IRTE is determined in step 720, the interrupt is broadcasted.
In case no relevant change of the IRTE is determined in step 720 (
In step 748, the operating system interrupt handler (OS IO-Irpt) is called and in step 750 reads and resets the DIBV bit set in step 710. In step 752, it is looped over, e.g., all DIBV bits of the DIBV assigned to the target PU, i.e. directed PU. Thus, e.g., all interrupts for the target PU may successively be handled. In case, e.g., all DIBV bits have been handled, the target PU is unblocked in step 754 (SIC.OC17) by resetting IRTE.dIBPIA. Furthermore, the DIBV is re-read in order to determine in step 756 whether meanwhile another DIBV bit has been set. If this is the case, the respective interrupt is handled, else the method ends with step 758.
If the result of the check in step 716 (
In step 786, the operating system interrupt handler (OS IO-Irpt) is called and in step 788 reads and resets the DISB bit set in step 760. In steps 790 and 792, it is signaled to the corresponding directed PU that the interrupt is handled. In step 794, it is looped over, e.g., all DISB bits of the DISB array, each bit assigned to another PU. The DISB summarizes, e.g., all interrupts to be handled by broadcasting. The interrupts are ordered to the PU to which they are targeted. Thus, e.g., all interrupts to be handled by broadcasting may successively be handled by the PU. In case, e.g., all DISB bits have been handled, the PU is unblocked in step 796 (SIC.OC1) by resetting IBPIA. Furthermore, DISB is re-read in order to determine in step 798 whether meanwhile another DISB bit has been set. If this is the case, the respective interrupt is handled, else the method ends with step 799.
Various embodiments of aspects of the invention provide a method of providing an interrupt signal to a guest operating system executed using one or more processors of a plurality of processors of a computer system assigned for usage by the guest operating system, as well as a computer system and a computer program product as described by the subject matter of the independent claims. Embodiments are described in the dependent claims. Embodiments of the present invention can be freely combined with each other if they are not mutually exclusive.
One aspect of the invention relates to a method of providing an interrupt signal to a guest operating system executed using one or more processors of a plurality of processors of a computer system assigned for usage by the guest operating system. The computer system further comprises one or more bus connected modules being operationally connected with the plurality of processors via a bus and a bus attachment device. The computer system further comprises a memory operationally connected with the bus attachment device. Each processor of the plurality of processors is assigned with a logical processor ID used by the bus attachment device to address the respective processor, and each processor of the plurality of processors assigned for usage by the guest operating system is further assigned with an interrupt target ID used by the guest operating system and the one or more bus connected modules to address the respective processor. The method comprises: receiving by the bus attachment device an interrupt signal from one of the bus connected modules with an interrupt target ID, the interrupt target ID identifying one of the processors assigned for usage by the guest operating system as a target processor for handling the interrupt signal; retrieving by the bus attachment device a first copy of an interrupt table entry assigned to the received interrupt target ID from an interrupt table stored in the memory, the first copy of the interrupt table entry comprising a first copy of a running indicator indicating whether the target processor identified by the interrupt target ID is scheduled for usage by the guest operating system; checking by the bus attachment device using the first copy of the running indicator whether the target processor is scheduled for usage by the guest operating system; if the target processor is scheduled for usage by the guest operating system, translating by the bus attachment device the received interrupt target ID to a logical processor ID and forwarding the interrupt signal to the target processor for handling using the logical processor ID resulting from the translation to address the target processor directly; else forwarding by the bus attachment device the interrupt signal for handling to the plurality of processors using broadcasting.
Another aspect of the invention relates to a computer system for providing an interrupt signal to a guest operating system executed using one or more processors of a plurality of processors of the computer system assigned for usage by the guest operating system. The computer system further comprises one or more bus connected modules being operationally connected with the plurality of processors via a bus and a bus attachment device. The computer system further comprises a memory operationally connected with the bus attachment device. Each processor of the plurality of processors is assigned with a logical processor ID used by the bus attachment device to address the respective processor, and each processor of the plurality of processors assigned for usage by the guest operating system is further assigned with an interrupt target ID used by the guest operating system and the one or more bus connected modules to address the respective processor. The computer system is configured to perform a method comprising: receiving by the bus attachment device an interrupt signal from one of the bus connected modules with an interrupt target ID, the interrupt target ID identifying one of the processors assigned for usage by the guest operating system as a target processor for handling the interrupt signal; retrieving by the bus attachment device a first copy of an interrupt table entry assigned to the received interrupt target ID from an interrupt table stored in the memory, the first copy of the interrupt table entry comprising a running indicator indicating whether the target processor identified by the interrupt target ID is scheduled for usage by the guest operating system; checking by the bus attachment device using the running indicator whether the target processor is scheduled for usage by the guest operating system; if the target processor is scheduled for usage by the guest operating system, translating by the bus attachment device the received interrupt target ID to a logical processor ID and forwarding the interrupt signal to the target processor for handling using the logical processor ID resulting from the translation to address the target processor directly; else forwarding by the bus attachment device the interrupt signal for handling to the plurality of processors using broadcasting.
Another aspect of the invention relates to a computer program product for providing an interrupt signal to a guest operating system executed using one or more processors of a plurality of processors of a computer system assigned for usage by the guest operating system. The computer system further comprises one or more bus connected modules being operationally connected with the plurality of processors via a bus and a bus attachment device. The computer system further comprises a memory operationally connected with the bus attachment device. Each processor of the plurality of processors is assigned with a logical processor ID used by the bus attachment device to address the respective processor, and each processor of the plurality of processors assigned for usage by the guest operating system is further assigned with an interrupt target ID used by the guest operating system and the one or more bus connected modules to address the respective processor. The computer program product comprises a computer readable non-transitory medium readable by a processing circuit and storing instructions for execution by the processing circuit for performing a method comprising: receiving by the bus attachment device an interrupt signal from one of the bus connected modules with an interrupt target ID, the interrupt target ID identifying one of the processors assigned for usage by the guest operating system as a target processor for handling the interrupt signal; retrieving by the bus attachment device a first copy of an interrupt table entry assigned to the received interrupt target ID from an interrupt table stored in the memory, the first copy of the interrupt table entry comprising a running indicator indicating whether the target processor identified by the interrupt target ID is scheduled for usage by the guest operating system; checking by the bus attachment device using the running indicator whether the target processor is scheduled for usage by the guest operating system; if the target processor is scheduled for usage by the guest operating system, translating by the bus attachment device the received interrupt target ID to a logical processor ID and forwarding the interrupt signal to the target processor for handling using the logical processor ID resulting from the translation to address the target processor directly; else forwarding by the bus attachment device the interrupt signal for handling to the plurality of processors using broadcasting.
In one or more embodiments, the guest operating system may, for example, be implemented using a pageable storage mode guest. A pageable guest, e.g., in a z/Architecture® hardware architecture, may be interpretively executed via a Start Interpretive Execution (SIE) instruction, at level 2 of interpretation. For instance, a logical partition (LPAR) hypervisor executes the SIE instruction to begin a logical partition in physical, fixed memory. An operating system in that logical partition, e.g., z/VM®, may issue the SIE instruction to execute its guests (virtual) machines in its virtual storage. Thus, the LPAR hypervisor may use level-1 SIE and the z/VM® hypervisor may use level-2 SIE.
According to embodiments, a computer system is a System z server offered by International Business Machines Corporation. System z® is based on the z/Architecture® hardware architecture offered by International Business Machines Corporation. Details regarding the z/Architecture® hardware architecture are described in an IBM publication entitled, “z/Architecture Principles of Operation,” IBM Publication No. SA22-7832-11, Aug. 25, 2017, which is hereby incorporated herein by reference in its entirety. IBM, System z and z/Architecture are trademarks or registered trademarks of International Business Machines Corporation in at least one jurisdiction. Other names used herein may be registered trademarks, trademarks or product names of International Business Machines Corporation or other companies.
According to embodiments, computer systems of other architectures may implement and use one or more aspects of the present invention. As examples, servers other than System z® servers, such as Power Systems servers or other servers offered by International Business Machines Corporation, or servers of other companies implement, use and/or benefit from one or more aspects of the present invention. Power is a trademark or a registered trademark of International Business Machines Corporation in at least one jurisdiction. Further, although in the examples herein, the bus connected modules and the bus attachment device are considered a part of a server, in other embodiments, they do not have to necessarily be considered a part of the server, but may simply be considered as being coupled to system memory and/or other components of a computer system. The computer system need not be a server. Further, although the bus connected modules may be PCIe, one or more aspects of the present invention are usable with other bus connected modules. PCIe adapter and PCIe functions are just examples. Further, one or more aspects of the present invention may be applicable to interrupt schemes other than PCI MSI and PCI MSI-X. Yet further, although examples are described in which bits are set, in other embodiments, bytes or other types of indicators may be set. Moreover, the DTE and other structures may include more, less and/or different information.
Further, other types of computer systems may benefit from one or more aspects of the present invention. As an example, a data processing system suitable for storing and/or executing program code is usable that includes at least two processors coupled directly or indirectly to memory elements through a system bus. The memory elements include, for instance, local memory employed during actual execution of the program code, bulk storage, and cache memory which provide temporary storage of at least some program code in order to reduce the number of times code is to be retrieved from bulk storage during execution.
Input/Output or I/O devices including, but not limited to, keyboards, displays, pointing devices, DASD (Direct Access Storage Device), tape, CDs (Compact Discs), DVDs (Digital Versatile Discs), thumb drives and other memory media, etc., may be coupled to the system either directly or through intervening I/O controllers. Network adapters may also be coupled to the system to enable the data processing system to become coupled to other data processing systems or remote printers or storage devices through intervening private or public networks. Modems, cable modems, and Ethernet cards are just a few of the available types of network adapters.
Referring to
A computer system may comprise information in local or main storage, as well as addressing, protection, and reference and change recording. Some aspects of addressing include the format of addresses, the concept of address spaces, the various types of addresses, and the manner in which one type of address is translated to another type of address. Some of main storage includes permanently assigned storage locations. Main storage provides the system with directly addressable fast-access storage of data. Both data and programs are to be loaded into main storage, e.g., from input devices, before they may be processed.
Main storage may include one or more smaller, faster-access buffer storages, sometimes called caches. A cache may be physically associated with a CPU or an I/O processor. The effects, except on performance, of the physical construction and use of distinct storage media may generally not be observable by programs executed.
Separate caches may be maintained for instructions and for data operands. Information within a cache may be maintained in contiguous bytes on an integral boundary called a cache block or cache line. A model may provide an EXTRACT CACHE ATTRIBUTE instruction which returns the size of a cache line in bytes. A model may also provide PREFETCH DATA and PREFETCH DATA RELATIVE LONG instructions which effects the prefetching of storage into the data or instruction cache or the releasing of data from the cache.
Storage may be viewed as a long horizontal string of bits. For most operations, accesses to storage may proceed in a left-to-right sequence. The string of bits is subdivided into units of eight bits. An eight-bit unit is called a byte, which is the basic building block of all information formats. Each byte location in storage may be identified by a unique nonnegative integer, which is the address of that byte location, also referred to as the byte address. Adjacent byte locations may have consecutive addresses, starting with 0 on the left and proceeding in a left-to-right sequence. Addresses are unsigned binary integers and may e.g. be 24, 31, or 64 bits.
Information is transmitted between memory and CPUs one byte, or a group of bytes, at a time. Unless otherwise specified, in, for instance, the z/Architecture® hardware architecture, a group of bytes in memory is addressed by the leftmost byte of the group. The number of bytes in the group is either implied or explicitly specified by the operation to be performed. When used in a CPU operation, a group of bytes is called a field. Within each group of bytes, in, for instance, the z/Architecture® hardware architecture, bits are numbered in a left-to-right sequence. In the z/Architecture® hardware architecture, the leftmost bits are sometimes referred to as the “high-order” bits and the rightmost bits as the “low-order” bits. Bit numbers are not storage addresses, however. Only bytes may be addressable. To operate on individual bits of a byte in storage, the entire byte may be accessed. The bits in a byte may be numbered 0 through 7, from left to right in, e.g., the z/Architecture hardware architecture. The bits in an address may be numbered 8-31 or 40-63 for 24-bit addresses, or 1-31 or 33-63 for 31-bit addresses; they are numbered 0-63 for 64-bit addresses. Within any other fixed-length format of multiple bytes, the bits making up the format may be consecutively numbered starting from 0. For purposes of error detection, and in, e.g., for correction, one or more check bits may be transmitted with each byte or with a group of bytes. Such check bits are generated automatically by the machine and cannot be directly controlled by the program. Storage capacities are expressed in number of bytes. When the length of a storage-operand field is implied by the operation code of an instruction, the field is said to have a fixed length, which may be one, two, four, eight, or sixteen bytes. Larger fields may be implied for some instructions. When the length of a storage-operand field is not implied but is stated explicitly, the field is said to have a variable length. Variable-length operands may vary in length by increments of one byte or with some instructions, in multiples of two bytes or other multiples. When information is placed in storage, the contents of only those byte locations are replaced that are included in the designated field, even though the width of the physical path to storage may be greater than the length of the field being stored.
Certain units of information are to be on an integral boundary in storage. A boundary is called integral for a unit of information when its storage address is a multiple of the length of the unit in bytes. Special names are given to fields of 2, 4, 8, and 16 bytes on an integral boundary. A halfword is a group of two consecutive bytes on a two-byte boundary and is the basic building block of instructions. A word is a group of four consecutive bytes on a four-byte boundary. A doubleword is a group of eight consecutive bytes on an eight-byte boundary. A quadword is a group of 16 consecutive bytes on a 16-byte boundary. When storage addresses designate halfwords, words, doublewords, and quadwords, the binary representation of the address contains one, two, three, or four rightmost zero bits, respectively. Instructions are to be on two-byte integral boundaries. The storage operands of most instructions do not have boundary-alignment requirements.
On devices that implement separate caches for instructions and data operands, a significant delay may be experienced if the program stores into a cache line from which instructions are subsequently fetched, regardless of whether the store alters the instructions that are subsequently fetched.
In one embodiment, aspects of the invention may be practiced by software, sometimes referred to licensed internal code, firmware, micro-code, milli-code, pico-code and the like, any of which would be consistent with aspects of the present invention. Referring to
Software program code may comprise an operating system which controls the function and interaction of the various computer components and one or more application programs. Program code may be paged from storage media device 811 to the relatively higher-speed computer storage 802 where it is available for processing by processor 801. Well known techniques and methods for embodying software program code in memory, on physical media, and/or distributing software code via networks may be used. Program code, when created and stored on a tangible medium, including but not limited to electronic memory modules (RAM), flash memory, Compact Discs (CDs), DVDs, Magnetic Tape, may be referred to as a “computer program product”. The computer program product medium may be readable by a processing circuit in, e.g., a computer system for execution by the processing circuit.
The system 821 may communicate with other computers or networks of computers by way of a network adapter capable of communicating 828 with a network 829. Example network adapters are communications channels, token ring, Ethernet or modems. Alternatively, the system 821 may communicate using a wireless interface, such as a cellular digital packet data (CDPD) card. The system 821 may be associated with such other computers in a Local Area Network (LAN) or a Wide Area Network (WAN), or the system 821 may be a client in a client/server arrangement with another computer, etc.
Still referring to
Referring concurrently to
Alternatively, the programming code may be embodied in the memory 825, and accessed by the processor 826 using the processor bus. Such programming code may include an operating system 831 which controls the function and interaction of the various computer components and one or more application programs 832. Program code may be paged from storage media 827 to high-speed memory 825, where it is available for processing by the processor 826. Well known techniques and methods for embodying software programming code in memory, on physical media, and/or distributing software code via networks may be used.
The cache that is most readily available to a processor, i.e., which may be faster and smaller than other caches of the processor, is the lowest cache, also referred to as the L1 or level one cache, and main memory is the highest-level cache, also referred to as Ln, e.g., L3, if there are n, e.g., n=3, levels. The lowest-level cache may be divided into an instruction cache, also referred to as I-cache, holding machine readable instructions to be executed and a data cache, also referred to as D-Cache, holding data operands.
Referring to
A program counter (instruction counter) 861 keeps track of the address of the current instruction to be executed. A program counter in a z/Architecture® processor is 64 bits and may be truncated to 31 or 24 bits to support prior addressing limits. A program counter may be embodied in a program status word (PSW) of a computer such that it persists during context switching. Thus, a program in progress, having a program counter value, may be interrupted by, for example, the operating system resulting in a context switch from the program environment to the operating system environment. The PSW of the program maintains the program counter value while the program is not active, and the program counter in the PSW of the operating system is used while the operating system is executing. The program counter may be incremented by an amount equal to the number of bytes of the current instruction. Reduced Instruction Set Computing (RISC) instructions may be fixed length, while Complex Instruction Set Computing (CISC) instructions may be variable length. Instructions of the IBM z/Architecture® hardware architecture are, e.g., CISC instructions having a length of 2, 4 or 6 bytes. The Program counter 861 may be modified by either a context switch operation or a branch taken operation of a branch instruction for example. In a context switch operation, the current program counter value is saved in the program status word along with other state information about the program being executed, such as condition codes, and a new program counter value is loaded pointing to an instruction of a new program module to be executed. A branch taken operation may be performed in order to permit the program to make decisions or loop within the program by loading the result of the branch instruction into the program counter 861.
An instruction fetch unit 855 may be employed to fetch instructions on behalf of the processor 826. The fetch unit either fetches “next sequential instructions,” target instructions of branch taken instructions, or first instructions of a program following a context switch. Modern instruction fetch units may employ prefetch techniques to speculatively prefetch instructions based on the likelihood that the prefetched instructions might be used. For example, a fetch unit may fetch 16 bytes of an instruction that includes the next sequential instruction and additional bytes of further sequential instructions.
The fetched instructions may then be executed by the processor 826. According to embodiments, the fetched instruction(s) may be passed to a dispatch unit 856 of the fetch unit. The dispatch unit decodes the instruction(s) and forwards information about the decoded instruction(s) to appropriate units 857, 858, 860. An execution unit 857 may receive information about decoded arithmetic instructions from the instruction fetch unit 855 and may perform arithmetic operations on operands according to the opcode of the instruction. Operands may be provided to the execution unit 857 either from memory 825, architected registers 859 or from an immediate field of the instruction being executed. Results of the execution, when stored, may be stored either in memory 825, registers 859 or in other machine hardware, such as control registers, PSW registers and the like.
A processor 826 may have dynamic address translation (DAT) 862 for translating program addresses, e.g., virtual addresses to, e.g., real addresses, and processor 826 may comprise one or more units 857, 858, 860 for executing the function of the instruction. Referring to
An ADD instruction for example may be executed in an execution unit 857 having arithmetic and logical functionality while a floating-point instruction for example would be executed in a floating-point execution having specialized floating point capability. An execution unit operates on operands identified by an instruction by performing an opcode defined function on the operands. For example, an ADD instruction may be executed by an execution unit 857 on operands found in two registers 859 identified by register fields of the instruction.
The execution unit 857 performs the arithmetic addition on two operands and stores the result in a third operand where the third operand may be a third register or one of the two source registers. The execution unit utilizes, e.g., an Arithmetic Logic Unit (ALU) 866 that is capable of performing a variety of logical functions such as Shift, Rotate, And, Or and XOR, as well as a variety of algebraic functions including any of add, subtract, multiply, divide. Some ALUs 866 are designed for scalar operations and some for floating point. Data may be big endian, where the least significant byte is at the highest byte address, or little endian, where the least significant byte is at the lowest byte address, depending on architecture. The IBM z/Architecture® hardware architecture is big endian. Signed fields may be sign and magnitude, 1's complement or 2's complement depending on architecture. A 2's complement number may be advantageous in that the ALU does not need to design a subtract capability since either a negative value or a positive value in 2's complement requires only an addition within the ALU. Numbers may be described in shorthand, where a 12-bit field defines an address of a 4,096-byte block and described as a 4 Kbyte (Kilo-byte) block, for example.
Referring to
The execution of a group of instructions may be interrupted for a variety of reasons including a context switch initiated by an operating system, a program exception or error causing a context switch, an I/O interrupt signal causing a context switch or a multi-threading activity of a plurality of programs in a multi-threaded environment, for example. A context switch action saves state information about a currently executing program and then loads state information about another program being invoked. State information may be saved in hardware registers or in memory, for example. State information comprises, e.g., a program counter value pointing to a next instruction to be executed, condition codes, memory translation information and architected register content. A context switch activity may be exercised by hardware circuits, application programs, operating system programs or firmware code, like e.g. microcode, pico-code or licensed internal code (LIC), alone or in combination.
A processor accesses operands according to instruction defined methods. The instruction may provide an immediate operand using the value of a portion of the instruction, may provide one or more register fields explicitly pointing to either general purpose registers or special purpose registers, like e.g. floating-point registers, for example. The instruction may utilize implied registers identified by an opcode field as operands. The instruction may utilize memory locations for operands. A memory location of an operand may be provided by a register, an immediate field, or a combination of registers and immediate field as exemplified by the z/Architecture® hardware architecture long displacement facility wherein the instruction defines a base register, an index register and an immediate field, i.e., displacement field, that are added together to provide the address of the operand in memory for example. Location herein may imply a location in main memory unless otherwise indicated.
Referring to
Addresses that an application program “sees” are often referred to as virtual addresses. Virtual addresses are sometimes also referred to as “logical addresses” and “effective addresses”. These virtual addresses are virtual in that they are redirected to physical memory location by one of a variety of dynamic address translation (DAT) technologies including, but not limited to, simply prefixing a virtual address with an offset value, translating the virtual address via one or more translation tables, the translation tables comprising at least a segment table and a page table alone or in combination, the segment table having an entry pointing to the page table. In the z/Architecture® hardware architecture, a hierarchy of translation is provided including a region first table, a region second table, a region third table, a segment table and an optional page table. The performance of the address translation is often improved by utilizing a translation lookaside buffer (TLB) which comprises entries mapping a virtual address to an associated physical memory location. The entries are created when the DAT translates a virtual address using the translation tables. Subsequent use of the virtual address may then utilize the entry of the fast TLB rather than the slow sequential translation table accesses. TLB content may be managed by a variety of replacement algorithms including Least Recently used (LRU).
Each processor of a multiprocessor system has the responsibility to keep shared resources, such as I/O, caches, TLBs and memory, interlocked for coherency. So-called “snoop” technologies may be utilized in maintaining cache coherency. In a snoop environment, each cache line may be marked as being in any one of a shared state, an exclusive state, a changed state, an invalid state and the like in order to facilitate sharing.
I/O units 854 (
Further, other types of computer systems may benefit from one or more aspects of the present invention. As an example, the computer system may comprise an emulator, e.g., software or other emulation mechanisms, in which a particular architecture including, for example, instruction execution, architected functions, such as address translation, and architected registers, is emulated or in which a subset thereof is emulated, e.g., on a native computer system having a processor and memory. In such an environment, one or more emulation functions of the emulator may implement one or more aspects of the present invention, even though a computer executing the emulator may have a different architecture than the capabilities being emulated. For example, in emulation mode, the specific instruction or operation being emulated may be decoded, and an appropriate emulation function may be built to implement the individual instruction or operation.
In an emulation environment, a host computer may for example comprise a memory to store instructions and data, an instruction fetch unit to fetch instructions from memory and to optionally provide local buffering for the fetched instruction, an instruction decode unit to receive the fetched instructions and to determine the type of instructions that have been fetched, and an instruction execution unit to execute the instructions. Execution may include loading data into a register from memory, storing data back to memory from a register, and/or performing some type of arithmetic or logical operation, as determined by the decode unit. For example, each unit may be implemented in software. Operations being performed by the units may be implemented as one or more subroutines within emulator software.
More particularly, in a mainframe, architected machine instructions are used by programmers, like e.g. “C” programmers, e.g., by way of a compiler application. These instructions stored in the storage medium may be executed natively in a z/Architecture® IBM® Server, or alternatively in machines executing other architectures. They may be emulated in the existing and in future IBM® mainframe servers and on other machines of IBM® (e.g., Power Systems servers and System x® Servers). They may be executed in machines running Linux on a wide variety of machines using hardware manufactured by IBM®, Intel®, AMD™, and others. Besides execution on that hardware under a z/Architecture® hardware architecture, Linux may be used as well as machines which use emulation by Hercules, UMX, or FSI (Fundamental Software, Inc), where generally execution is in an emulation mode. In emulation mode, emulation software is executed by a native processor to emulate the architecture of an emulated processor.
A native processor may execute emulation software comprising either firmware or a native operating system to perform emulation of the emulated processor. The emulation software is responsible for fetching and executing instructions of the emulated processor architecture. The emulation software maintains an emulated program counter to keep track of instruction boundaries. The emulation software may fetch one or more emulated machine instructions at a time and convert the one or more emulated machine instructions to a corresponding group of native machine instructions for execution by the native processor. These converted instructions may be cached such that a faster conversion may be accomplished. Notwithstanding, the emulation software is to maintain the architecture rules of the emulated processor architecture so as to assure operating systems and applications written for the emulated processor operate correctly. Furthermore, the emulation software is to provide resources identified by the emulated processor architecture including, but not limited to, control registers, general purpose registers, floating point registers, dynamic address translation function including segment tables and page tables for example, interrupt mechanisms, context switch mechanisms, Time of Day (TOD) clocks and architected interfaces to I/O subsystems such that an operating system or an application program designed to run on the emulated processor, may be run on the native processor having the emulation software.
A specific instruction being emulated is decoded, and a subroutine is called to perform the function of the individual instruction. An emulation software function emulating a function of an emulated processor is implemented, for example, in a “C” subroutine or driver, or some other method of providing a driver for the specific hardware.
In
It is understood that one or more of the aforementioned embodiments of the invention may be combined as long as the combined embodiments are not mutually exclusive. Ordinal numbers, like e.g. ‘first’ and ‘second’, are used herein to indicate different elements assigned with the same name, but do not necessarily establish any order of the respective elements.
Aspects of the present invention are described herein with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer readable program instructions.
Aspects of the present invention may be a system, a method, and/or a computer program product at any possible technical detail level of integration. The computer program product may include a computer readable storage medium (or media) having computer readable program instructions thereon for causing a processor to carry out aspects of the present invention.
The computer readable storage medium can be a tangible device that can retain and store instructions for use by an instruction execution device. The computer readable storage medium may be, for example, but is not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. A non-exhaustive list of more specific examples of the computer readable storage medium includes the following: a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanically encoded device such as punch-cards or raised structures in a groove having instructions recorded thereon, and any suitable combination of the foregoing. A computer readable storage medium, as used herein, is not to be construed as being transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission media (e.g., light pulses passing through a fiber-optic cable), or electrical signals transmitted through a wire.
Computer readable program instructions described herein can be downloaded to respective computing/processing devices from a computer readable storage medium or to an external computer or external storage device via a network, for example, the Internet, a local area network, a wide area network and/or a wireless network. The network may comprise copper transmission cables, optical transmission fibers, wireless transmission, routers, firewalls, switches, gateway computers and/or edge servers. A network adapter card or network interface in each computing/processing device receives computer readable program instructions from the network and forwards the computer readable program instructions for storage in a computer readable storage medium within the respective computing/processing device.
Computer readable program instructions for carrying out operations of the present invention may be assembler instructions, instruction-set-architecture (ISA) instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, state-setting data, configuration data for integrated circuitry, or either source code or object code written in any combination of one or more programming languages, including an object oriented programming language such as Smalltalk, C++, or the like, and procedural programming languages, such as the “C” programming language or similar programming languages. The computer readable program instructions may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet Service Provider). In some embodiments, electronic circuitry including, for example, programmable logic circuitry, field-programmable gate arrays (FPGA), or programmable logic arrays (PLA) may execute the computer readable program instructions by utilizing state information of the computer readable program instructions to personalize the electronic circuitry, in order to perform aspects of the present invention.
Aspects of the present invention are described herein with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer readable program instructions.
These computer readable program instructions may be provided to a processor of a computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks. These computer readable program instructions may also be stored in a computer readable storage medium that can direct a computer, a programmable data processing apparatus, and/or other devices to function in a particular manner, such that the computer readable storage medium having instructions stored therein comprises an article of manufacture including instructions which implement aspects of the function/act specified in the flowchart and/or block diagram block or blocks.
The computer readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer implemented process, such that the instructions which execute on the computer, other programmable apparatus, or other device implement the functions/acts specified in the flowchart and/or block diagram block or blocks.
The flowchart and block diagrams in the Figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the blocks may occur out of the order noted in the Figures. For example, two blocks shown in succession may, in fact, be accomplished as one step, executed concurrently, substantially concurrently, in a partially or wholly temporally overlapping manner, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems that perform the specified functions or acts or carry out combinations of special purpose hardware and computer instructions.
Possible combinations of features described herein may be the following:
In one aspect, a method for providing an interrupt signal to a guest operating system executed using one or more processors of a plurality of processors of a computer system assigned for usage by the guest operating system is provided. The computer system further comprises one or more bus connected modules being operationally connected with the plurality of processors via a bus and a bus attachment device. The computer system further comprises a memory operationally connected with the bus attachment device. Each processor of the plurality of processors is assigned with a logical processor ID used by the bus attachment device to address the respective processor, and each processor of the plurality of processors assigned for usage by the guest operating system is further assigned with an interrupt target ID used by the guest operating system and the one or more bus connected modules to address the respective processor. The method comprises: receiving by the bus attachment device an interrupt signal from one of the bus connected modules with an interrupt target ID, the interrupt target ID identifying one of the processors assigned for usage by the guest operating system as a target processor for handling the interrupt signal; retrieving by the bus attachment device a first copy of an interrupt table entry assigned to the received interrupt target ID from an interrupt table stored in the memory, the first copy of the interrupt table entry comprising a first copy of a running indicator indicating whether the target processor identified by the interrupt target ID is scheduled for usage by the guest operating system; checking by the bus attachment device using the first copy of the running indicator whether the target processor is scheduled for usage by the guest operating system; and if the target processor is scheduled for usage by the guest operating system translating by the bus attachment device the received interrupt target ID to a logical processor ID and forwarding the interrupt signal to the target processor for handling using the logical processor ID resulting from the translation to address the target processor directly, else forwarding by the bus attachment device the interrupt signal for handling to the plurality of processors using broadcasting.
In one embodiment, the running indicator is implemented as a single bit.
In one embodiment, the interrupt signal is received in a form of a message signaled interrupt comprising the interrupt target ID of the target processor.
In one embodiment, the first copy of the interrupt table entry further comprises a first mapping of the received interrupt target ID to a first one of the logical processor IDs. The bus attachment device uses the first copy of the interrupt table entry for the translating of the received interrupt target ID to the logical processor ID of the target processor.
In one embodiment, the first copy of the interrupt table entry further comprises an interrupt blocking indicator indicating whether the target processor identified by the interrupt target ID is currently blocked from receiving interrupt signals. The method further comprises: checking by the bus attachment device using the interrupt blocking indicator whether the target processor is blocked from receiving interrupt signals; and if the target processor is unblocked, continuing with the forwarding of the interrupt signal, else blocking by the bus attachment device the interrupt signal from being forwarded to the target processor for handling.
In one embodiment, the method comprises forwarding by the bus attachment device the interrupt signal for handling to the remaining processors of the plurality of processors using broadcasting.
In one embodiment, the method further comprises checking by an interrupt handler of the guest operating system whether any interrupts addressed to the target processor are pending for handling by the target processor. If no interrupts addressed to the target processor being pending for handling by the target processor, changing by the guest operating system the interrupt blocking indicator in the interrupt table entry assigned to the target processor to indicate the target processor being unblocked.
In one embodiment, the method further comprises if the target processor is unblocked, changing by the bus attachment device the interrupt blocking indicator in the interrupt table entry assigned to the interrupt target ID to indicate the first logical processor ID being blocked. The changing is performed before the forwarding of the interrupt signal to the target processor for handling.
In one embodiment, the method further comprises retrieving by the bus attachment device after the changing of the interrupt blocking indicator a second copy of the interrupt table entry assigned to the received interrupt target ID, and checking the second copy of the interrupt table entry by the bus attachment device to exclude a predefined type of change of the second copy of the interrupt table relative to the first copy of the interrupt table entry. A successful exclusion of the predefined type of change being required for the forwarding of the interrupt signal to the target processor for handling.
In one embodiment, the predefined type of change is a change of the first mapping of the received interrupt target ID relative to a second mapping of the received interrupt target ID to a second one of the logical processor IDs comprised by the second copy of the interrupt table entry. In case the second mapping comprises a change relative to the first mapping, the interrupt signal is forwarded by the bus attachment device for handling to the plurality of processors using broadcasting.
In one embodiment, the predefined type of change is a change of the first copy of the running indicator relative to a second copy of the running indicator comprised by the second copy of the interrupt table entry. In case the second copy of the running indicator comprises a change relative to the first copy of the running indicator (e.g., bit), the second running indicator is indicating that the target processor is not being scheduled for usage by the operating system. The interrupt signal is forwarded by the bus attachment device for handling to the plurality of processors using broadcasting.
In one embodiment, the method further comprises retrieving by the bus attachment device a copy of a device table entry from a device table stored in the memory, the device table entry comprising a direct signaling indicator indicating whether the target processor is to be addressed directly. If the direct signaling indicator indicates a direct forwarding of the interrupt signal, the forwarding of the interrupt signal using logical processor ID of the target processor to address the target processor directly is executed, else the method comprises forwarding by the bus attachment device the interrupt signal for handling to the plurality of processors using broadcasting.
In one embodiment, the device table entry further comprises an interrupt table address indicator indicating a memory address of the interrupt table. The bus attachment device uses the memory address of the interrupt table for the retrieving of the first copy of the interrupt table entry.
In one embodiment, the memory further comprises an interrupt summary vector, and the device table entry further comprises an interrupt summary vector address indicator indicating a memory address of the interrupt summary vector, the interrupt summary vector comprising an interrupt summary indicator per bus connected module, each interrupt summary indicator being assigned to a bus connected module indicating whether there is an interrupt signal issued by the respective bus connected module to be handled. The method further comprises using by the bus attachment device the indicated memory address of the interrupt summary vector to update the interrupt summary indicator assigned to the bus connected module from which the interrupt signal is received such that the updated interrupt summary indicator indicates that there is an interrupt signal issued by the respective bus connected module to be handled.
In one embodiment, the memory further comprises a directed interrupt summary vector, and the device table entry further comprises a directed interrupt summary vector address indicator indicating a memory address of the directed interrupt summary vector, the directed interrupt summary vector comprising a directed interrupt summary indicator per interrupt target ID, each directed interrupt summary indicator being assigned to an interrupt target ID indicating whether there is an interrupt signal addressed to the respective interrupt target ID to be handled. The method further comprises using by the bus attachment device the indicated memory address of the directed interrupt summary vector to update the interrupt summary indicator assigned to the target processor ID to which the received interrupt signal is addressed such that the updated interrupt summary indicator indicates that there is an interrupt signal addressed to the respective interrupt target ID to be handled.
In one embodiment, the memory further comprises one or more interrupt signal vectors, and the device table entry further comprises an interrupt signal vector address indicator indicating a memory address of an interrupt signal vector of the one or more interrupt signal vectors, each of the interrupt signal vectors comprising one or more signal indicators, each interrupt signal indicator being assigned to a bus connected module of the one or more bus connected modules and an interrupt target ID indicating whether an interrupt signal has been received from the respective bus connected module addressed to the respective interrupt target ID. The method further comprises: using by the bus attachment device the indicated memory address of the interrupt signal vector to select the interrupt signal indicator assigned to the bus connected module which issued the received interrupt signal and to the interrupt target ID to which the received interrupt signal is addressed; and updating the selected interrupt signal indicator such that the selected interrupt signal indicator indicates that there is an interrupt signal issued by the respective bus connected module and addressed to the respective interrupt target ID to be handled.
In one embodiment, the interrupt signal vectors each comprises an interrupt signal indicator per interrupt target ID assigned to the respective interrupt target ID, each of the interrupt signal vectors being assigned to an individual bus connected module with the interrupt signal indicators of the respective interrupt signal vector being further assigned to the respective individual bus connected module.
In one embodiment, the interrupt signal vectors each comprises an interrupt signal indicator per bus connected module assigned to the respective bus connected module, each of the interrupt signal vectors being assigned to an individual target processor ID with the interrupt signal indicators of the respective interrupt signal vector being further assigned to the respective target processor ID.
In one embodiment, the device table entry further comprises a logical partition ID identifying a logical partition to which the guest operating system is assigned, and the forwarding of the interrupt signal by the bus attachment device further comprises forwarding with the interrupt signal the logical partition ID.
In one embodiment, the method further comprises retrieving by the bus attachment device an interrupt subclass ID identifying an interrupt subclass to which the received interrupt signal is assigned, and the forwarding of the interrupt signal by the bus attachment device further comprises forwarding with the interrupt signal the interrupt subclass ID.
In one aspect, a computer system for providing an interrupt signal to a guest operating system executed using one or more processors of a plurality of processors of the computer system assigned for usage by the guest operating system is provided. The computer system further comprises one or more bus connected modules being operationally connected with the plurality of processors via a bus and a bus attachment device. The computer system further comprises a memory operationally connected with the bus attachment device. Each processor of the plurality of processors is assigned with a logical processor ID used by the bus attachment device to address the respective processor, and each processor of the plurality of processors assigned for usage by the guest operating system is further assigned with an interrupt target ID used by the guest operating system and the one or more bus connected modules to address the respective processor. The computer system is configured to perform a method comprising: receiving by the bus attachment device an interrupt signal from one of the bus connected modules with an interrupt target ID, the interrupt target ID identifying one of the processors assigned for usage by the guest operating system as a target processor for handling the interrupt signal; retrieving by the bus attachment device a first copy of an interrupt table entry assigned to the received interrupt target ID from an interrupt table stored in the memory, the first copy of the interrupt table entry comprising a running indicator indicating whether the target processor identified by the interrupt target ID is scheduled for usage by the guest operating system; checking by the bus attachment device using the running indicator whether the target processor is scheduled for usage by the guest operating system; and if the target processor is scheduled for usage by the guest operating system, translating by the bus attachment device the received interrupt target ID to a logical processor ID and forwarding the interrupt signal to the target processor for handling using the logical processor ID resulting from the translation to address the target processor directly, else forwarding by the bus attachment device the interrupt signal for handling to the plurality of processors using broadcasting.
In one aspect, a computer program product for providing an interrupt signal to a guest operating system executed using one or more processors of a plurality of processors of a computer system assigned for usage by the guest operating system is provided. The computer system further comprises one or more bus connected modules being operationally connected with the plurality of processors via a bus and a bus attachment device. The computer system further comprises a memory operationally connected with the bus attachment device. Each processor of the plurality of processors is assigned with a logical processor ID used by the bus attachment device to address the respective processor, and each processor of the plurality of processors assigned for usage by the guest operating system is further assigned with an interrupt target ID used by the guest operating system and the one or more bus connected modules to address the respective processor. The computer program product comprises a computer readable non-transitory medium readable by a processing circuit and storing instructions for execution by the processing circuit for performing a method comprising: receiving by the bus attachment device an interrupt signal from one of the bus connected modules with an interrupt target ID, the interrupt target ID identifying one of the processors assigned for usage by the guest operating system as a target processor for handling the interrupt signal; retrieving by the bus attachment device a first copy of an interrupt table entry assigned to the received interrupt target ID from an interrupt table stored in the memory, the first copy of the interrupt table entry comprising a first copy of a running indicator indicating whether the target processor identified by the interrupt target ID is scheduled for usage by the guest operating system; checking by the bus attachment device using the first copy of the running indicator whether the target processor is scheduled for usage by the guest operating system; and if the target processor is scheduled for usage by the guest operating system, translating by the bus attachment device the received interrupt target ID to a logical processor ID and forwarding the interrupt signal to the target processor for handling using the logical processor ID resulting from the translation to address the target processor directly, else forwarding by the bus attachment device the interrupt signal for handling to the plurality of processors using broadcasting.
| Number | Date | Country | Kind |
|---|---|---|---|
| 19157095 | Feb 2019 | EP | regional |
| Number | Name | Date | Kind |
|---|---|---|---|
| 5577255 | Roemer | Nov 1996 | A |
| 6460105 | Jones | Oct 2002 | B1 |
| 7281075 | Armstrong | Oct 2007 | B2 |
| 7386640 | Lee et al. | Jun 2008 | B2 |
| 7447820 | Raj | Nov 2008 | B2 |
| 8453143 | Mahalingam | May 2013 | B2 |
| 8601194 | Tsirkin | Dec 2013 | B2 |
| 8612659 | Serebrin | Dec 2013 | B1 |
| 9032127 | Fischer | May 2015 | B2 |
| 9043521 | Tiruvallur | May 2015 | B2 |
| 9116869 | Madukkarumukumana | Aug 2015 | B2 |
| 9378162 | Jebson | Jun 2016 | B2 |
| 9465760 | Egi | Oct 2016 | B2 |
| 9507740 | Feehrer | Nov 2016 | B2 |
| 9697029 | Tsirkin | Jul 2017 | B2 |
| 9772868 | Tu | Sep 2017 | B2 |
| 9830286 | Tsirkin | Nov 2017 | B2 |
| 9910699 | Sankaran | Mar 2018 | B2 |
| 9910700 | Prosch | Mar 2018 | B2 |
| 9940291 | Bacha | Apr 2018 | B2 |
| 9952987 | Guddeti et al. | Apr 2018 | B2 |
| 10078603 | Tsirkin | Sep 2018 | B2 |
| 10282327 | Farrell et al. | May 2019 | B2 |
| 10713095 | Yoo | Jul 2020 | B2 |
| 10922111 | Raisch | Feb 2021 | B2 |
| 11016800 | Kraemer | May 2021 | B2 |
| 20030200250 | Klick | Oct 2003 | A1 |
| 20030204655 | Schmisseur | Oct 2003 | A1 |
| 20050080982 | Vasilevsky | Apr 2005 | A1 |
| 20050125580 | Madukkarumukumana | Jun 2005 | A1 |
| 20060233203 | Iwamura | Oct 2006 | A1 |
| 20080162762 | Neiger | Jul 2008 | A1 |
| 20090183180 | Nelson | Jul 2009 | A1 |
| 20100095039 | Marietta | Apr 2010 | A1 |
| 20110047309 | Brinkman | Feb 2011 | A1 |
| 20110087814 | Liu | Apr 2011 | A1 |
| 20110197003 | Serebrin et al. | Aug 2011 | A1 |
| 20110273309 | Zhang | Nov 2011 | A1 |
| 20110320664 | Belmar et al. | Dec 2011 | A1 |
| 20110320840 | Nayar | Dec 2011 | A1 |
| 20120099521 | Ryu | Apr 2012 | A1 |
| 20120131248 | Abali | May 2012 | A1 |
| 20130080673 | Feehrer | Mar 2013 | A1 |
| 20130104127 | Domsch | Apr 2013 | A1 |
| 20130159579 | Neiger et al. | Jun 2013 | A1 |
| 20130339327 | Belmar | Dec 2013 | A1 |
| 20140047150 | Marietta | Feb 2014 | A1 |
| 20140157397 | Dalal | Jun 2014 | A1 |
| 20140223060 | Tsirkin et al. | Aug 2014 | A1 |
| 20150006783 | Chew | Jan 2015 | A1 |
| 20150312116 | Taheri | Oct 2015 | A1 |
| 20160117190 | Sankaran et al. | Apr 2016 | A1 |
| 20160147679 | Guddeti | May 2016 | A1 |
| 20180129619 | Neiger et al. | May 2018 | A1 |
| 20180349307 | Auernhammer | Jun 2018 | A1 |
| 20180356964 | Morris | Dec 2018 | A1 |
| 20190163519 | Freeman | May 2019 | A1 |
| 20200264911 | Raisch | Aug 2020 | A1 |
| 20200264912 | Raisch | Aug 2020 | A1 |
| 20200264917 | Kraemer | Aug 2020 | A1 |
| 20200264992 | Raisch | Aug 2020 | A1 |
| 20200264993 | Nerz | Aug 2020 | A1 |
| 20200264994 | Raisch | Aug 2020 | A1 |
| 20200264995 | Raisch | Aug 2020 | A1 |
| 20210055945 | Raisch | Feb 2021 | A1 |
| 20210117229 | Kraemer | Apr 2021 | A1 |
| Number | Date | Country |
|---|---|---|
| 200636485 | Oct 2006 | TW |
| 201310345 | Mar 2013 | TW |
| 201430572 | Aug 2014 | TW |
| 1537822 | Nov 2016 | TW |
| 1570563 | Nov 2017 | TW |
| 201830259 | Aug 2018 | TW |
| Entry |
|---|
| IBM, “z/Architecture—Principles of Operation,” IBM Publication No. SA22-7832-12, Thirteenth Edition, Sep. 2019, pp. 1-2000. |
| Mell, Peter and Tim Grance, “The NIST Definition of Cloud Computing,” National Institute of Standards and Technology, Information Technology Laboratory, Special Publication 800-145, Sep. 2011, pp. 1-7. |
| Intel 64 and IA-32 Architectures Software Developer's Manual—vol. 3A; System Programming Guide, Part 1, Sep. 2016, Chapter 10 (10-1-10-50) (+ cover pages). |
| Williams, Chris, “Intel's Broadwell Xeon E5-2600 v4 chips: So What's in it for you, smartie-pants coders,” Mar. 2016, pp. 1-9. |
| Raj, Ashok, “Recent Enhancements in Intel Virtualization Technology for Directed I/O (Intel VT-D),” Sep. 2018, pp. 1-7. |
| Raisch, Christoph et al., “Directed Interrupt Virtualization,” U.S. Appl. No. 16/789,504, filed Feb. 13, 2020, pp. 1-83. |
| Kraemer, Marco et al., “Directed Interrupt Virtualization With Interrupt Table,” U.S. Appl. No. 16/789,515, filed Feb. 13, 2020, pp. 1-103. |
| Nerz, Bernd et al., “Directed Interrupt for Multilevel Virtualization,” U.S. Appl. No. 16/789,532, filed Feb. 13, 2020, pp. 1-109. |
| Raisch, Christoph et al., “Directed Interrupt for Multilevel Virtualization With Interrupt Table,” U.S. Appl. No. 16/789,556, filed Feb. 13, 2020, pp. 1-106. |
| Raisch, Christoph et al., “Directed Interrupt Virtualization With Blocking Indicator,” U.S. Appl. No. 16/789,567, filed Feb. 13, 2020, pp. 1-99. |
| Raisch, Christoph et al., “Interrupt Signaling for Directed Interrupt Virtualization,” U.S. Appl. No. 16/789,581, filed Feb. 13, 2020, pp. 1-101. |
| Raisch, Christoph et al., “Directed Interrupt Virtualization With Fallback,” U.S. Appl. No. 16/789,600, filed Feb. 13, 2020, pp. 1-103. |
| List of IBM Patents or Patent Applications Treated As Related, May 13, 2020, 2 pages. |
| Raisch, Christoph et al., “Interrupt Signaling For Directed Interrupt Virtualization,” U.S. Appl. No. 17/091,498, filed Nov. 6, 2020, pp. 1-98. |
| Kraemer, Marco et al., “Directed Interrupt Virtualization with Interrupt Table,” U.S. Appl. No. 17/137,755, filed Dec. 30, 2020, pp. 1-103. |
| List of IBM Patents or Patent Applications Treated As Related, Feb. 19, 2021, 2 pages. |
| Nerz, Bernd et al., “Directed Interrupt for Multilevel Virtualization,” U.S. Appl. No. 17/358,172, filed Jun. 26, 2021, pp. 1-109. |
| Raisch, Christoph et al., “Directed Interrupt for Multilevel Virtualization With Interrupt Table,” U.S. Appl. No. 17/482,192, filed Sep. 22, 2021, pp. 1-101. |
| Kraemer, Marco et al., “Directed Interrupt Virtualization With Running Indicator,” U.S. Appl. No. 17/482,514, filed Sep. 22, 2021, pp. 1-105. |
| List of IBM Patents or Patent Applications Treated as Related, Oct. 12, 2021, 2 pages. |
| Raisch, Christoph et al., “Directed Interrupt Virtualization With Fallback,” U.S. Appl. No. 17/506,368, filed Oct. 20, 2021, pp. 1-103. |
| List of IBM Patents or Patent Applications Treated as Related, Oct. 26, 2021, 2 pages. |
| Examination Report for AU Application No. 2020222167, dated Dec. 8, 2021 (3 pages). |
| Number | Date | Country | |
|---|---|---|---|
| 20200264910 A1 | Aug 2020 | US |
