EP3850486A1 - Procédé et circuit de multiplexage temporel d'accès concurrents à une ressource informatique - Google Patents
Procédé et circuit de multiplexage temporel d'accès concurrents à une ressource informatiqueInfo
- Publication number
- EP3850486A1 EP3850486A1 EP19813627.7A EP19813627A EP3850486A1 EP 3850486 A1 EP3850486 A1 EP 3850486A1 EP 19813627 A EP19813627 A EP 19813627A EP 3850486 A1 EP3850486 A1 EP 3850486A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- processing
- request
- time
- critical program
- access
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/52—Program synchronisation; Mutual exclusion, e.g. by means of semaphores
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/52—Program synchronisation; Mutual exclusion, e.g. by means of semaphores
- G06F9/526—Mutual exclusion algorithms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
Definitions
- the field of the invention is that of real-time computer systems for which the execution time of tasks, and in particular the execution time in the worst case (called WCET for "Worst Case Execution Time”), must be known to ensure validation and guarantee safety.
- the invention more particularly aims to improve the rate of use of a resource shared by computer programs executing in parallel while guaranteeing the estimation of the WCET of these programs by means of a time-division multiplex arbitration policy. for access to the shared resource.
- Real-time systems must react reliably, which implies both being certain of the result produced by their programs but also knowing the time they take to execute.
- the worst-case execution times are therefore fundamental data for the validation and safety of such real-time systems, and even more in the context of autonomous real-time systems (robotics, autonomous car, GPS) for which safety is essential.
- time-division multiplexing TDM
- time-division multiplexing TDM
- time slots each allocated to a predefined program for exclusive access to the shared resource.
- the access time to the resource shared by a program can then be easily limited (it corresponds to the maximum delay until reaching the next time slot allocated to the program), the bandwidth offered to a program is independent of the other programs and the WCET programs can thus be determined.
- the access time to the resource shared by a program depends on the scheduling of the time slots allocated to it.
- this scheduling is generally static since it is carried out during the design of the system, before the programs are executed. It generally includes the assignment of a sequence of time slots to different programs, this sequence repeating periodically (we speak of TDM period).
- the time slots of a TDM period are only used by the programs if they have a request to access the shared resource to be sent. When this is not the case, these time slots are therefore unused.
- a basic TDM arbitration policy is therefore said to be idle because these time slots are not recovered by the other programs in order to reduce their access time to the shared resource.
- TMDds dynamic TDM with slack counters
- a processing margin counter (“slack counter” in English) is associated with each critical program. This counter indicates the number of time slots separating the effective termination of the processing of an access request from its processing deadline. This counter is used to delay, if necessary, the deadline for processing the next access request from the critical program.
- the arbitration carried out at the start of each time slot then consists in selecting the pending access request having the closest processing deadline, independently of the critical program to which this time slot is allocated. If this TDMds arbitration policy makes it possible to improve the use of the shared resource, it nevertheless remains room for improvement.
- the length of the time slots must be greater than or equal to the maximum time taken by a request, sent in isolation, to be processed by the shared resource.
- the shared resource is a memory
- the access time of a write request is less than the access time of a read request, since in this second case the memory must send the requested data to destination of the requesting program.
- the access time to a memory can depend on the history of the requests emitted by the other programs. An upper bound for processing requests must therefore be identified to determine the length of the time slots. However, the processing of most requests will by definition end before this terminal, so that the TDMds arbitration policy remains idle.
- the object of the invention is to propose an improved arbitration policy which makes it possible to optimize the use of the shared resource while guaranteeing the temporal predictability of the processing of requests for access to the shared resource originating from critical programs.
- the invention provides a computer implemented method of arbitration between computer programs seeking to access concurrently a shared resource by each issuing an access request.
- the method realizes a time division multiple access according to which the time is divided into time slots each allocated to a critical program for access to the shared resource, each time slot comprising a plurality of time units.
- a processing margin is associated with each critical program in order to delay a processing deadline for an access request sent by the critical program.
- the method comprises, at each time unit, a step of selecting an access request from one or more pending access requests and a step of determining an authorization for immediate processing of the selected access request .
- the determination step comprises for a time unit not corresponding to the start of a time slot, when the critical program to which the next time slot is allocated has not issued the selected request, the authorization for immediate processing of the selected request if the processing margin of the critical program to which the next time slot is allocated is greater at a threshold.
- the determination step comprises, for a time unit not corresponding to the start of a time slot, when the critical program to which the next time slot is allocated has not sent the selected request and when the processing margin of the critical program to which the next time slot is allocated is below the threshold, putting the selected request on hold;
- the threshold corresponds to the time difference between the time unit at which the determination step is carried out and the time unit corresponding to the start of the next time slot;
- the determination step comprises, for a time unit not corresponding to the start of a time slot, when the critical program to which the next time slot is allocated has sent the selected request, authorizing the immediate processing of the selected request ;
- the determination step comprises, for a time unit corresponding to the start of a time slot, authorizing the immediate processing of the selected request;
- the selection step comprises, in the presence of an access request awaiting a critical program and an access request awaiting a non-critical program, the selection of the access request of the non-critical program if the deadline for processing the access request of the critical program is later than the end of the next time slot;
- the processing margin of a critical program is updated at each termination of processing of an access request from the critical program to correspond to the number of time units separating the termination of processing from the request for access and the deadline for processing the access request; - the deadline for processing a request to access a critical program corresponds to the end of a time slot allocated to the critical program which is the first to start after a date corresponding to a date of issue of the request access delayed by adding the processing margin of the critical program;
- Figures la and lb are diagrams illustrating the principle of arbitration carried out in the invention to authorize or not an immediate access to the shared resource
- FIG. 2 is a state machine representing a sequence of steps of the method according to the invention.
- FIG. 3 is a timing diagram showing an example of arbitration operated by the method according to the invention.
- FIG. 4 is a diagram of a circuit configured to implement the method according to the invention.
- a computer implemented method of arbitrating between computer programs seeking concurrent access to a shared resource by each issuing an access request is typically implemented on a multi-core hardware architecture where each core CO, Cl, Cm is configured to execute a program capable of requiring access to the shared resource during its execution.
- Each core can have a private cache memory and the shared resource is for example a shared memory MM.
- a hardware implementation configured to implement this method takes the form of a time division multiplexing circuit CMT interposed between the cores and the shared resource.
- the method according to the invention realizes a multiple access with distribution in time according to which the time is divided into time slots (slots in English) each allocated to a critical program for an access to the shared resource, each time slot comprising a plurality of time units (a time unit is for example a clock cycle).
- FIGS. 1a, 1b and 3 show an example of such a time division into slots of identical sizes alternately allocated to a first critical program A (the slots having an odd number) and to a second critical program B (the slots with an even number).
- a critical program is for example a program subject to time constraints and for which it is desired to be able to limit the execution time in the worst case (WCET).
- the method according to the invention associates a processing deadline with each access request sent by a critical program. This deadline corresponds to the moment at which the processing of the critical program request must be completed at the latest and must be respected.
- the method also relates to access requests sent by programs not subject to time constraints, i.e. non-critical programs, and for which no notion of processing deadline is necessary.
- the method according to the invention also exploits a processing margin associated with each critical program, this processing margin being used to possibly delay a processing deadline for an access request sent by the critical program, in particular when the program critic has accumulated sufficient processing margin.
- this processing margin being used to possibly delay a processing deadline for an access request sent by the critical program, in particular when the program critic has accumulated sufficient processing margin.
- arbitration decisions are made at the start of each time slot and are only based on requests awaiting processing. It is then possible to delay a request from a critical program, depending on the deadline for processing this request.
- the processing margin is not used during arbitration decisions and has only an effect on the calculation of the processing times for requests.
- the invention is based on the fact of recognizing that a processing margin is also valid when a critical program has not (yet) issued a request for access to the resource.
- the shared resource can then start processing any of the pending requests if this lower bound is later than the end of the next time slot. This ensures that the memory can partially process the request in the current time slot, while completing this processing in the next time slot without violating the behavior in the worst case.
- the invention thus proposes to exploit the processing margins, not at the start of each time slot, but at each time unit.
- the processing margin associated with the critical program to which the next time slot is allocated is consulted to determine whether a request can be immediately processed by the shared resource.
- the purpose of this consultation is to check whether an overflow of the processing time of the request on the next time slot does not call into question the respect of the processing deadline associated with a possible request which would be issued by the critical program to which this next time slot is allocated.
- the processing margin of a critical program is updated at each termination of processing of an access request from the critical program to correspond to the number of time units separating the termination of the processing of the access request and the deadline for processing the access request.
- the deadline for processing a request for access to a critical program corresponds to the end of the time slot allocated to the critical program which is the first to start after a date corresponding to a date of issue of the request for access delayed by adding the processing margin of the critical program.
- FIGS. 1a and 1b show an example in which at a time unit corresponding to time t, access to the resource is released and it is checked whether a pending request can be processed.
- This time unit t does not correspond to the start of the time slot allocated to the critical program B which starts at to.
- the processing margin MAY of the critical program A to which the next time slot is allocated (slot n ° 3) is large enough to allow delaying the processing time of a request which would be issued by the program A between t and ti. Instead of this deadline corresponding to the end 11 of the next time slot allocated to program A, it is delayed to correspond to the end El of time slot No. 5 allocated to program A which follows a date corresponding to a date of issue of the access request delayed by adding the processing margin of the critical program MAY. This amounts to determining that the processing margin MAI is greater than the number of time units separating t and ti.
- FIG. 1b shows a state machine representative of an embodiment of the method according to the invention.
- the reference “Rac” designates the issue of a request for access to the shared resource.
- an “NC-AR” arbitration of non-critical requests is carried out according to a policy of arbitration (for example fixed priority or first come first served) to come and select one.
- the selection includes the choice, among the critical program requests, of the request presenting the closest processing deadline.
- the result of the selection is therefore either the critical request presenting the nearest processing deadline, or the non-critical request resulting from the “NC-AR” arbitration.
- the access request of the user is selected. non-critical program if the processing deadline for the critical program is later than the end date of the next time slot.
- the threshold can correspond, as we have seen above, to the time difference between the time unit t at which the determination step is carried out and the time unit ti corresponding to the start of the next time slot. Alternatively, the threshold can correspond to the size of a time slot.
- the “MAJ MTs” update of the processing margin of this critical program is updated.
- the processing margin of a critical program is updated at each termination of processing of an access request from the critical program to correspond to the number of time units separating the termination of processing from the access request and the deadline for processing the access request.
- FIG. 3 shows an example of implementation of this method with two critical programs A and B, a non-critical program c and slots having a size of 8 time units.
- the current value of the processing margin is indicated by the exponent of each request (for example the request Ai of critical program A has a processing margin of 10 D, or 10 time units).
- a first access request Ao from the critical program A is firstly sent during the slot n ° l allocated to the program A.
- the processing margin of A is zero and its processing deadline is fixed at the end of the next slot allocated to A (slot n ° 3).
- the processing margin for B owner of the next slot # 2 is zero and it is not possible to perform immediate processing of Ao before the start of the next slot # 2.
- the processing of Ao lasts six units of time and A has therefore accumulated a processing margin of 10 units of time.
- a first access request Bo of the critical program B is issued during the slot n ° 2.
- the next slot n ° 3 is allocated to the program A which has accumulated enough processing margin so that the immediate processing of the Bo request is authorized without risk that the overflowing of the processing of Bo on the niche 3 will jeopardize compliance with the processing deadline associated with a possible request sent by the critical program A to which slot No. 3 is allocated.
- Program B sees its processing margin increase to 10 time units.
- a first access request co from the non-critical program c and a second access request Bi from the critical program B are simultaneously sent.
- the co request of the non-critical program is selected because the critical program B does not own the next slot # 5, and is immediately processed because the critical program A owner of the next slot # 5 has accumulated enough processing margin (in this case 10 time units).
- the request Bi is processed immediately since program B owns the next slot n ° 6.
- a second access request Ai of the critical program A and a second access request ci of the non-critical program c are sent.
- the latter sees its processing margin increase to 7 time units.
- the request Ai is selected and immediately processed because program A owns the next niche 7.
- the processing of the request Ai ends during the slot n ° 6, and the processing margin of the program A increases to 11 time units.
- the second access request ci of the non-critical program c is selected because the deadline for processing request B 2 is after the end of time slot 7, then immediately processed because the processing margin of program A owning the next time slot 7 is sufficient.
- the request B 2 is immediately processed because the program B owns the next slot n ° 8.
- the latter sees its processing margin increase to 5 units of time.
- a third access request A 2 of the critical program A is then sent which, in view of the processing margin which then benefits the program A, has a processing deadline at the end of slot n ° ll.
- program A owns the next slot n ° 9 and its request A 2 can be immediately processed.
- FIG. 4 shows a possible implementation of a time division multiplexing circuit CMT configured to implement the method according to the invention.
- the circuit CMT is interposed between cores C0, Cl, Cm each executing a program and a shared resource.
- a processing margin is associated with each core executing a critical program.
- a counter can for this be implemented in the form of a register storing a number of time units not used by the heart. This counter is part of the execution context of the critical program, and its value is therefore saved / restored at each preemption / resumption of program execution.
- the CMT circuit includes an AR arbitration unit and a MUX multiplexer controlled by the arbitration unit to achieve selective access of one of the cores to the shared resource.
- the CMT circuit also includes a processing deadline calculation unit DC0, DC1, DCm associated with each of the cores. This unit is configured to receive an access request R0, RI, Rm sent by a program executed on the core with which it is associated and to calculate the deadline for processing the request when it is sent by a critical program.
- the unit DC0, DC1, DCm can include dedicated registers to store the calculated due date as well as the current processing margin.
- the unit DC0, DC1, DCm simply passes the request to a unit NC.AR in charge of carrying out an arbitration between non-critical requests according to any arbitration policy. It can come to store a predefined value in the register dedicated to the deadline, for example the maximum value that can be taken by a deadline.
- the arbitration circuit CA also includes a Min.extractor unit configured to receive processing deadline calculation units DC0, DC1, DCm the value stored in their register dedicated to the deadline, determining which is the smallest value and provide the arbitration unit AR with an identifier of the request having the closest deadline with respect to the current instant.
- a Min.extractor unit configured to receive processing deadline calculation units DC0, DC1, DCm the value stored in their register dedicated to the deadline, determining which is the smallest value and provide the arbitration unit AR with an identifier of the request having the closest deadline with respect to the current instant.
- the arbitration circuit CA also includes a unit SL configured to receive processing deadline calculation units DC0, DC1, DCm the value stored in their register dedicated to the margin and provide the arbitration unit AR with the margin of processing corresponding to the critical program to which the next time slot is allocated.
- the AR arbitration unit selects a request (from the Min.extractor unit or the NC.AR unit) and checks whether it can be processed at once. To do this, if the time unit corresponds to the start of a time slot, the request is processed. Otherwise, the arbitration unit determines whether the processing of the request may extend to the next time slot. As we saw previously, such an overflow on the following time slot is authorized if one or the other of the following two conditions is met:
- the next slot is allocated to the program sending the selected request.
- the program to which the next slot is allocated has accumulated sufficient margin to allow an overflow of the selected request in its slot.
- a sufficient margin corresponds to the size of a niche.
- the cores can perform preemptive multitasking, that is to say distribute processor time between different programs.
- a program running on one of the cores can thus be preempted in favor of another program.
- the running program is then interrupted to allow the other program to run.
- a time slot is actually associated with each heart. This window is allocated to it when the heart executes a critical program.
- the slot can be exploited for processing other requests coming from other cores.
- a critical program wishes to preempt a non-critical program (according to the scheduling of preemptive multitasking) and this non-critical program is awaiting processing of a request for access to the shared resource, this request must be processed so that the non-critical program can actually be preempted.
- the latter in order to limit the waiting time for processing the non-critical request, the latter inherits the criticality of the critical program.
- the heart then becomes critical and the time slot associated with it is again dedicated to it for processing the request.
- the calculation of the deadline is therefore based on the next occurrence of the time slot which is again dedicated to it.
- the current value of the processing margin of the critical program preempting the non-critical program can be added to determine this deadline.
- the method according to the invention can thus include a preemption step for the benefit of a critical program of a first program having issued an access request to the shared resource awaiting processing.
- This preemption step includes the calculation, possibly taking into account the processing margin of the critical program, of a processing deadline for said request corresponding to the end of a time slot allocated to the critical program.
- the time division multiplexing circuit CMT can carry out such management of the preemptions in the following manner.
- the pre-emptive core notifies the due date calculation unit associated with it. If a non-critical program A is preempted in favor of a critical program B, the presence of a request from the non-critical program A in the unit NC.AR is checked. If such a request exists, then this request is withdrawn from the unit NC.AR and is considered to be critical in the rest of the steps of the method previously described while potentially considering the processing margin of the critical program B for the calculation of the processing deadline now associated with program A. If a critical program A is pre-empted in favor of a critical program B, similarly, the processing margin of critical program B can be used to calculate the new processing deadline for program A .
- the invention is not limited to the method described above but also extends to a time division multiplexing circuit for concurrent access to a shared resource required by computer programs, configured to implement the method, and in particular to the CMT circuit. previously described.
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multi Processors (AREA)
- Bus Control (AREA)
Abstract
Description
Claims
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
FR1860117A FR3087982B1 (fr) | 2018-10-31 | 2018-10-31 | Procede et circuit de multiplexage temporel d'acces concurrents a une ressource informatique |
PCT/FR2019/052513 WO2020089542A1 (fr) | 2018-10-31 | 2019-10-22 | Procédé et circuit de multiplexage temporel d'accès concurrents à une ressource informatique |
Publications (1)
Publication Number | Publication Date |
---|---|
EP3850486A1 true EP3850486A1 (fr) | 2021-07-21 |
Family
ID=65861387
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
EP19813627.7A Pending EP3850486A1 (fr) | 2018-10-31 | 2019-10-22 | Procédé et circuit de multiplexage temporel d'accès concurrents à une ressource informatique |
Country Status (4)
Country | Link |
---|---|
US (1) | US20210397488A1 (fr) |
EP (1) | EP3850486A1 (fr) |
FR (1) | FR3087982B1 (fr) |
WO (1) | WO2020089542A1 (fr) |
Family Cites Families (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6061361A (en) * | 1997-06-19 | 2000-05-09 | Advanced Micro Devices, Inc. | Time multiplexed scheme for deadlock resolution in distributed arbitration |
EP1550953A1 (fr) * | 2003-12-29 | 2005-07-06 | CNX S.p.A. | Méthode et dispositif implémentant le multiplexage temporel d'accès à une mémoire RAM à deux ports de plusieurs sources de données disposant d'horloges indépendantes. |
US7603503B1 (en) * | 2006-09-18 | 2009-10-13 | Nvidia Corporation | Efficiency based arbiter |
US20090217280A1 (en) * | 2008-02-21 | 2009-08-27 | Honeywell International Inc. | Shared-Resource Time Partitioning in a Multi-Core System |
US8695002B2 (en) * | 2009-10-20 | 2014-04-08 | Lantiq Deutschland Gmbh | Multi-threaded processors and multi-processor systems comprising shared resources |
US9189379B2 (en) * | 2013-02-06 | 2015-11-17 | Avago Technologies General Ip (Singapore) Pte. Ltd. | Buffer for managing data samples in a read channel |
US20160283272A1 (en) * | 2015-03-25 | 2016-09-29 | Intel Corporation | Shared resource access control method and apparatus |
US10908955B2 (en) * | 2018-03-22 | 2021-02-02 | Honeywell International Inc. | Systems and methods for variable rate limiting of shared resource access |
CN110556390A (zh) * | 2018-05-31 | 2019-12-10 | 松下知识产权经营株式会社 | 摄像装置 |
FR3082029B1 (fr) * | 2018-06-05 | 2020-07-10 | Thales | Controleur de partage de ressources d'une plate-forme informatique et procede associe de partage des ressources |
-
2018
- 2018-10-31 FR FR1860117A patent/FR3087982B1/fr active Active
-
2019
- 2019-10-22 US US17/289,270 patent/US20210397488A1/en active Pending
- 2019-10-22 WO PCT/FR2019/052513 patent/WO2020089542A1/fr unknown
- 2019-10-22 EP EP19813627.7A patent/EP3850486A1/fr active Pending
Also Published As
Publication number | Publication date |
---|---|
US20210397488A1 (en) | 2021-12-23 |
FR3087982A1 (fr) | 2020-05-01 |
FR3087982B1 (fr) | 2020-12-04 |
WO2020089542A1 (fr) | 2020-05-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11606261B2 (en) | Data service overload detection and mitigation | |
WO2014193438A1 (fr) | Ordonnancement de fils d'execution sensible a la priorite efficace | |
US20070277232A1 (en) | Ike daemon self-adjusting negotiation throttle | |
US11150999B2 (en) | Method, device, and computer program product for scheduling backup jobs | |
US11330047B2 (en) | Work-load management in a client-server infrastructure | |
WO2022183802A1 (fr) | Procédé, appareil et dispositif d'équilibrage de charge, support de stockage et produit programme informatique | |
EP2203817A2 (fr) | Procede de gestion des preemptions dans un systeme d'exploitation en temps reel | |
FR3047821A1 (fr) | Procede et dispositif de gestion d'un appareil de commande | |
FR2990782A1 (fr) | Procede de gestion d'une execution de taches dans un systeme informatique | |
EP3660677B1 (fr) | Procédé et dispositif de surveillance d'application(s) logicielle(s) avec période temporelle tampon précédant une section réservée pour un ensemble de ressource(s) partagée(s), programme d'ordinateur et système avionique associés | |
EP3850486A1 (fr) | Procédé et circuit de multiplexage temporel d'accès concurrents à une ressource informatique | |
JP4791315B2 (ja) | 業務管理装置及び業務管理方法及びプログラム | |
Jolai et al. | A preemptive discrete-time priority buffer system with partial buffer sharing | |
WO2012038000A1 (fr) | Procede de gestion de taches dans un microprocesseur ou un ensemble de microprocesseurs | |
US11126472B2 (en) | System and method for managing shared computer resources | |
FR2980611A1 (fr) | Circuit pour planifier le deroulement d'un traitement de donnees | |
EP1841151A1 (fr) | Système de commutation de message | |
US12019623B2 (en) | Data transfer resiliency during bulk to streaming transition | |
WO2021210094A1 (fr) | Dispositif de gestion de microservices, procédé de gestion de microservices et programme | |
US8805899B2 (en) | Data channel utilization management in multi-user system | |
EP4113297A1 (fr) | Procédé de gestion des travaux dans un système informatique et système associé | |
US10300386B1 (en) | Content item instance scaling based on wait time | |
WO2023052730A1 (fr) | Procédé de supervision et superviseur du pilotage d'un traitement parallèle distribué dans un réseau de communication, dispositif de fourniture de services et programme le mettant en œuvre | |
WO2005010633A2 (fr) | Procédé de gestion de l'exécution d'au moins un programme sur plusieurs calculateurs | |
EP3807757A1 (fr) | Procédé d'accéleration de l'exécution d'un programme à chemin unique par exécution en parallèle de séquences conditionnellement concurrentes |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: UNKNOWN |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE |
|
PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE |
|
17P | Request for examination filed |
Effective date: 20210413 |
|
AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR |
|
RAP1 | Party data changed (applicant data changed or rights of an application transferred) |
Owner name: INSTITUT MINES-TELECOM Owner name: COMMISSARIAT A L'ENERGIE ATOMIQUE ET AUX ENERGIES ALTERNATIVES |
|
DAV | Request for validation of the european patent (deleted) | ||
DAX | Request for extension of the european patent (deleted) | ||
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: EXAMINATION IS IN PROGRESS |
|
17Q | First examination report despatched |
Effective date: 20230222 |