(19)
(11)EP 3 369 028 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
20.01.2021 Bulletin 2021/03

(21)Application number: 15791266.8

(22)Date of filing:  29.10.2015
(51)Int. Cl.: 
G06F 21/51  (2013.01)
G06F 21/56  (2013.01)
G06F 21/54  (2013.01)
G06F 21/12  (2013.01)
(86)International application number:
PCT/EP2015/075165
(87)International publication number:
WO 2017/071763 (04.05.2017 Gazette  2017/18)

(54)

CHECKING A SECURITY VALUE CALCULATED FOR A PART OF A PROGRAM CODE

PRÜFEN EINES FÜR EINEN TEIL EINES PROGRAMMCODES BERECHNETEN SICHERHEITSWERTS

VÉRIFICATION D'UNE VALEUR DE SÉCURITÉ CALCULÉE POUR UNE PARTIE D'UN CODE DE PROGRAMME


(84)Designated Contracting States:
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

(43)Date of publication of application:
05.09.2018 Bulletin 2018/36

(73)Proprietor: Hewlett-Packard Development Company, L.P.
Spring TX 77389 (US)

(72)Inventors:
  • BALDWIN, Adrian
    Stoke Gifford Bristol (GB)
  • GRIFFIN, Jonathan
    Stoke Gifford Bristol (GB)

(74)Representative: Haseltine Lake Kempner LLP 
Redcliff Quay 120 Redcliff Street
Bristol BS1 6HU
Bristol BS1 6HU (GB)


(56)References cited: : 
US-A1- 2006 047 955
US-A1- 2009 193 211
US-A1- 2014 189 340
US-A1- 2007 283 140
US-A1- 2010 107 246
  
      
    Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


    Description

    BACKGROUND



    [0001] A program code causes a computer system to run an application or agent when being executed by a processor of the computing system. The program code may be subject to illegitimate modifications. US2010/107246A1 discloses related art.

    BRIEF DESCRIPTION OF THE DRAWINGS



    [0002] 

    FIG. 1 is a schematic illustration of an example guard network.

    FIG. 2 shows an example memory layout of the guarded program code 100 of FIG. 1.

    FIG. 3 shows an example apparatus to execute a guarded program code.

    FIG. 4 is a schematic representation of an example validation program including additional functionality.

    FIG. 5 illustrates an example computing system for implementing the technique described herein.

    FIG. 6 is a flow diagram of an example method for verifying the correct execution of a program code.


    DETAILED DESCRIPTION



    [0003] The following detailed description refers to the accompanying drawings. Wherever possible, the same reference numbers are used in the drawings and the following description to refer to the same or similar parts. While several examples are described in the following, modifications, adaptations, and other implementations are possible. Accordingly, the following detailed description does not limit the disclosed examples. Instead, the proper scope of the disclosed examples may be defined by the appended claims.

    [0004] Mechanisms for protecting program codes may be embedded into the program code or may be part of the program code. Tamperproof techniques aim at making it hard to subvert a piece of the program code as it runs. This may be used for copyright protection to ensure that license checks cannot be removed from a program code. An example of such a mechanism for protecting a program code involves adding or embedding into the program code a code block for monitoring a desired part of the program code and to take some action in case tampering of the monitored part of the program code. The program block may also be referred to as execution unit or guard. The program code may include a plurality of functions, which are to be monitored, and some guards may be provided in the program code to perform a check on the functions. Such a mechanism may also be referred to as a guard network.

    [0005] FIG. 1 is a schematic illustration of an example guard network. FIG. 1 shows a schematic representation of a program code 100 having a plurality of code blocks 1001 to 1007. The program code 100 includes five security-sensitive regions represented by the code blocks 1001,1003 and 1005 to 1007. The security-sensitive regions perform the functions F1 to F5, which are monitored so as to ensure for a correct working of the program code when being run on a computing system. To monitor the security-sensitive regions of the program code a plurality of guards 1021 to 1026 are added to the program code. A first guard 102, has been added to the code block 1001 implementing a first function F1. Likewise, guards 1022 to 1025 are added to the code blocks implementing the functions F2 to F5. Code block 1002, which does not implement a security-sensitive function, includes include another guard 1026. The guards 1021 to 1026 include a first part C1 to C6, which is provided to run a check on the memory for the functions F1 to F5 that it is being pointed to. The check may be the calculation of a first security value, e.g., a checksum or a crypto hash. The check may include retrieving the address of the function in a memory in which the code blocks are stored and to calculate the checksum or crypto hash over the memory portion containing the part of the program code to be monitored.

    [0006] The guards 1021 to 1026 include a second part G1 to G6 which receives the results of the check, e.g. the calculated checksum, and determines whether the result is correct or not. For example, the first security value may be comparing it to a precomputed value, also referred to as a second security value. The second security value may be a precomputed checksum to which the calculated checksum is compared. In case of a match of the first and second security values program execution is continued normally, and in case of a mismatch between the first and second security values the guard may trigger security-related actions. A mismatch between the first and second security values may also be referred to as a detection event, and the security-related action be the logging the detection events for later evaluation. In other instances, the program code may be made unusable, for example by halting its execution or causing an eventual crash that may be hard to trace back to the guard. Another example of a security-related action may be a restoration of the piece of damaged or tampered code, which caused the mismatch, before executing it. The restoration may include overwriting a tampered part of the program code with a clean copy of the part of the program code.

    [0007] In the example shown in FIG. 1 the security-sensitive functions F1 to F5 are shown. In accordance with other examples more or less security-relevant functions may be included in the program code. In addition, program code blocks 1022 and 1024 are shown which do not include any security-relevant sections that are to be monitored. In other examples there may be more or less non-security-relevant code blocks in the program code. In the example of FIG. 1, the guards 1021 to 1025 are shown to be in the same code block as the functions F1 to F5. In accordance with other examples the guards 1021 to 1025 may be provided in parts of the program code which are separate or distant from the parts where the functions are placed, see e.g. code block 1002 including the guard 1026 for monitoring function F1 remote from the part of program the code where the function F1 is included.

    [0008] In the example of FIG. 1, guard 1021 runs a checksum calculation on the part of the memory holding the function F2 and receives the result of the checksum calculation, which is then evaluated or checked by the guard 1021 to see whether it matches a precomputed checksum. In a similar way guard 1022 monitors function F4, guard 1023 monitors function F5, guard 1024 monitors function F3, and guard 1025 monitors function F1. In accordance with examples the guards may also monitor other guards. The guard 1023 monitors guard 1026 and calculates a checksum using the memory part in which the guard 1026 is held so as to determine, using the returned checksum, whether the guard 1026 has been tampered or not. Guard 1026, in the depicted example, is used to monitor function F1.

    [0009] The guards depicted in the example guard network of FIG. 1 may cause different security-related actions in case of a mismatch of the first and second security values. For example, some guards may log suspicious events indicated by the mismatch, cause a repair of the monitored memory portion using reference data, or cause a halt of the execution of the function or the entire program code. The guards may also use combinations of the just-mentioned actions.

    [0010] The program code 100 shown in FIG. 1 may be stored in a memory of a computing system, and FIG. 2 shows an example memory layout of the guarded program code 100. The code blocks 1001 to 1007 are arranged subsequently in the memory layout and the guards are schematically represented by the hashed sections of the code blocks. The arrows represent the checks performed by the guards on the other memory parts.

    [0011] In accordance with examples, the program code shown in FIG. 1 and FIG. 2 may be an agent application downloaded onto a client computer system, for example a laptop, to help manage the computer system, for example for performing security management and monitoring tasks. Such an application is likely to be targeted by malware infecting the computing system in order to avoid detection and removal. The above-described guard network or guarded program code provides a security feature that allows to check and report on the correct working of the agent or program code running on the client computer system, and it allows for verifying the correct execution of a long running process by providing the possibility to carry out runtime checks as the agent or a program code is executed.

    [0012] FIG. 3 is a schematic representation of an example apparatus to execute, in a first processing resource 200, the program code 100 including the plurality of guards 1021 to 1026. When executing the program code 100 in the first processing resource 200 a guard, also referred to as execution unit, calculates a first security value for a part of the program code 100. The first security value may be calculated using different types of calculations. A variety of functions may be used with varying complexity from simple checksum calculations or additions through to cryptographic hash calculations. The apparatus comprises a second processing resource 202 separate from the first processing resource 100. The second processing resource 202 runs a validation program 204. The validation program 204 receives the first security value, as is schematically depicted by arrow 206. The validation program 204 checks the first security value against a second security value that is calculated from a corresponding part of a reference copy 208 of the program code 100 to obtain a check result which is returned to the guard, as is schematically represented by arrow 210. The first and second processing resources may be provided with a fast inter-domain communication mechanism, for example using a shared memory, for communicating information, like the first security value and the check result. In accordance with other examples additional information may be communicated. The guard receiving the check result 210 performs a security-related action in response to a check result 210 indicating a mismatch between the first security value and the second security value. In accordance with examples, the security-related action includes logging the mismatch event, halting execution of the program code or of a part of the program code and/or repairing part of the program code using, for example, the reference copy 208.

    [0013] In accordance with examples, the second security value may be calculated remote from the validation program 204, for example by an external unit of a computing system generating second security values for code blocks of the program code using the reference copy of the program code 100. The pre-calculated security values, for example checksums, may be stored in the second processing environment 202 together with the validation program 204. To validate the first security value 206 a corresponding second security value may be accessed and compared to the first security value 206 provide the check result 210.

    [0014] In accordance with other examples, the second processing resource 202 may store the reference copy 208 of the program code 100, and the validation program 204 may include a validation routine 212, which receives the first security value 206 and calculates the second security value using the reference copy 208 of the program code 100 stored at the second processing resource 202. The guards provided in the program code 100 may perform different checks or security calculations for different parts of the program code to be monitored, for example some of the guards may perform a checksum calculation, and others may perform a crypto hash value calculation. In other words, the guards 1021 to 1026 may perform the same or different types of calculations for obtaining the first security value. In examples allowing for the use of different types of calculations to be performed by the guards 1021 to 1026, when forwarding the first security value 206 to the validation program 204, also a type identifier 214 is forwarded to the validation program 204. The validation routine 212 of the validation program 204 determines, using the received type identifier 214, which type of calculation has been used by the guard in the program code 100 to calculate the first security value 206, and uses the identified type of calculation and the reference copy of the program code to calculate the second security value. In other words, it is determined what kind of check or what type of calculation has been applied to a part of the program code 100 and the same check or type of calculation is applied to the corresponding part of the reference copy 208 held in the second processing resource.

    [0015] The first processing resource may be a main or first operating system, and the second processing resource may be a second operating system that are run on a single client system including a processor and memory, wherein the physical resources for the main and second operating systems 200, 202, for example the memory, are kept separate by the processor. The program code 100 may implement a management agent MA running in the main operating system 200, and the validation program 200 may implement a protected agent PA running in the second operating system 202. The management agent may be an application program downloaded onto the client computer system, for example a laptop computer, a desktop computer, a PDA or a smartphone. The management agent MA may help managing the client computer system or applications executed by the client computer system, for example for performing security management, configuration actions, security actions and monitoring tasks. The example apparatus of FIG. 3 allows detecting tampering or corruption of the management agent MA using the tamperproof program code 100. The first processing resource and the second processing resource may be co-located on the same physical system, e.g. a single computer including a processor on which the first processing resource and the second processing are executed, The computer may include more than one processor. In accordance with examples the first processing resource and the second processing resource may work off-line.

    [0016] The technique described herein takes advantage of the separate processing environments for securely running the management agent MA free from interference of malware. In accordance with examples, the main operating system 200 and the second operating system 202 may be implemented on a single client system using virtualization technologies using, e.g., a secure hypervisor, which implements a multiple processing resource environment for running the program code 100 and the validation program 204 separate from each other in the separate operating systems. In accordance with examples other mechanisms for running secure processes may be employed, such as a checking process run within an area protected by TrustZone. In other words, examples of the technique described herein use an alternative processing resource, for example the second operating system 202, to verify the correct execution of an agent program which is implemented by a tamperproof program code including an embedded network of guards within a piece of the program code to check that selected functions have not been changed. The second processing resource is used to perform the validation or verification of the result of the check performed separate from the actual program code so that the result of the check is no longer hidden in the program code 100, which makes it hard to subvert the management agents program code 100. In accordance with examples, additional functionality may be added to the checks, e.g. randomness, as is described below.

    [0017] In accordance with examples, the validation program 204 may retain a log of a plurality of check results obtained during execution of the program code 100, e.g. the management agent MA. The results may be stored to provide record traces of checks performed by the guarded program code, as is indicated by block 216 in FIG. 3. A trace of the execution of the management agent may be a count of checks performed successfully or unsuccessfully or may include frequency information indicating how often a specific function was checked and the results of the check.

    [0018] In the examples described above the validation program 200 does not access first processing resource, also referred to as first processing domain. For example, the validation program 200 does not access the memory of the main operating system 200. The validation is performed in the second operating system 202, also referred to as second processing domain, and, therefore, has no knowledge about the operation of the main operating system or the organization of the memory used by the main operating system. In accordance with other examples the second processing resource 202 may see into and change the memory contents within the main operating system 200.

    [0019] In accordance with examples, a system is provided including the apparatus and, as is shown in FIG. 3, a management server 218. The management server 218 is connected via a secure channel 220 to the validation program 204. The secure channel 220 may be established over the internet. The management server 218 may communicate with both the management agent MA and the protected agent PA to receive data from or to issue commands to the management agent MA and the protected agent PA. For example, the results of runtime checks, which may be recorded and stored by the validation program 204, are communicated to the management server 218 to allow detecting potential problems with the management agent MA from a remote position. For example, a system administrator may monitor a plurality of management agents running on different machines using the management agent MA, and examples of the technique described herein give confidence that the management agent is running correctly. The check information obtained by the validation program may be communicated to the management server 218 using a heartbeat protocol. The management server 218 may receive information about the check result or a log of a plurality of check results for performing an analysis. In accordance with an example, the management system 218 my send message to the validation program in response to the received resource information about the check result or in response to the analysis of the log of the plurality of check results, like a message causing the validation program to change the amount of ignore check signals send to the execution unit, or causing the validation program to change the program code to modify the execution unit as is described below.

    [0020] FIG. 4 is a schematic representation of an example validation program 204 including additional functionality that may be implemented, for example in the apparatus of FIG. 3. Adding the subsequently described functionality to the validation program 204 may make the overall system more robust against tampering. FIG. 4 shows schematically the outline of the validation program 204 including the validation routine 212 and the reference copy 208. The validation program 204 may include additional routines, e.g. for adding randomness to the calculation/check to be performed on the parts of the program code 100.

    [0021] For example, prior to performing a check the management agent MA issues a request 222 to the validation program 204, In response to the request 222, the validation program 204 may call some of the subsequently described routines of the validation program 204. In response to the request 222, the validation program 204 may call routine 224a to randomly select the function in the management agent MA to be monitored, as is indicated by block 226. The routine 224a outputs to the management agent MA an indication 228 which of the functions F1 to F5 or which of the guards 1021 to 1026 is to be evaluated by calculating the first security value. The routine 224a may include a table 230 holding the functions/guards to be monitored. The table 230 may be created and randomized at an initialization stage. The function/guard to be monitored may be selected by accessing the table 230, for example, by going through the table using the table index. The randomization may be achieved by creating the table at the initialization stage with a random distribution of the functions/guards in the table. This example may be used to ensure that security calculations or checks cover the entire program code 100 even when the some of the functions F1 to F5 are called a majority of the time, while others are not called that often. To cover the entire program code 100, the validation program 204 may include a balance function 232 which selects the functions/guards in accordance with a preset pattern so as to ensure that all functions are evaluated/checked independent of how often a function is actually called by the program code 100. The pattern may cause that each of the functions/guard is evaluated/checked with the same frequency. In accordance with other examples, functions/guards determined to be more security-sensitive than others may be checked more often when compared to functions/guards being considered less security-relevant.

    [0022] In accordance with another example, in response to the request 222, the validation program 204 may call a subroutine 224b which references a check table 234. The check table 234 indicates which kind of checks or calculations are to be carried out by the guards provided in the management agent. The check or the type of calculation 236 to be used for calculating the first security value is output to the management agent. The subroutine 224b may randomly select from the check routine table the check or the type of calculation 236 to be applied. The subroutine 224b may customize each of the check routines selected and may forward the check routine to be performed by the guards of the program code 100.

    [0023] In accordance with other examples, the checks or calculations performed by the guards may use an initialization vector. For adding randomness the validation program 204 may include a subroutine 224c which, in response to the receipt of the request 222, causes the generation of a random initialization vector 238, which is returned to the management agent. The random initialization vector 238 may be used for randomly initializing a the check routine calculating the first security values so that the first security value cannot be predicted, For example, if the check is a hash then the initialization vector would cause hash(iv::code) to be computed rather than just the hash(code), where :: is a concatenate operation. The result cannot be predicted. But both parts of the check know the answer.

    [0024] The routines 224a to 224c adding randomness to the check routines to be performed make it harder to prepare answers to checks or check routines that compare correctly with the reference copy and associated checks that the validation program performs.

    [0025] In accordance with another example, the validation program 204 may include routine 240 which uses time information about the performance of the checks performed. A guard which is about to start the calculation of the first security value may send an indication 242 of the start. The indication 242 is received by the routine 240 which measures the time between receipt of the indication 224 and receipt of the first security value 206 at the validation program 204. In case the response from the management agent, namely the receipt of the first security value, is slow in relation to the check routine applied to the program code part, this may be marked to be suspicious event. For example, dependent on the type of calculation used to calculate the first security value, e.g. the complexity of the calculation, a threshold may be set, and in case the time measured by routine 240 exceeds the threshold, it may be determined that there is a suspicious event and that the program code is to be monitored more closely.

    [0026] In accordance with other examples, the validation program 204 may include a routine 244 to analyze the performance of the management agent MA. For example, a statistical analysis or heuristics may be used to evaluate the check results 210. The threat environment may be set in response to a message from the management server 218 or may be determined through the analysis of the checks performed. If a low threat environment is indicated, in response to the request 222 the routine 244 may return an ignore check signal 246 for some of the checks. The ignore check signal 246 may prevent the execution unit from performing the check, e.g. calculating the first security value. If a change to a high threat environment is indicated, the amount of ignore messages 246 may be reduced. For example, in case the number of suspicious events rises, for example due to the fact that the number of mismatches between the first and second security values has risen above a predefined threshold or in case the time measurement in routine 240 indicated suspicious events, the number of ignore messages 246 may be reduced. This allows dynamically dealing with performance versus risk tradeoffs.

    [0027] In accordance with other examples, instead of building all checks into the program code, detours-style hooks may be added to the functions in the program code as a part of the checking function to add a dynamic nature to the checks, which is a technique malware uses to hook into code, but in accordance with examples is used to dynamically hook in additional checks. For example, the validation program may access a memory associated with the first processing resource, and change the program code to modify the execution unit, or to replace the execution unit by a new execution unit, or to add to the program code a new execution unit.

    [0028] In accordance with examples in which the second processing resource 202 may see into and change the memory contents within the main operating system 200, the validation program 204 may call routine 248 to introspect the memory of the main operating system 200. This may be used to perform randomization by placing different check addresses, initialization vectors and check code directly into the management agent.

    [0029] In accordance with examples, some or all of the above described routines 224a to 224c, 240, 244, 248 may be implemented by the validation program 204.

    [0030] In accordance with examples, the technique described herein also supports plugin extensions to the management agent MA. For example, in case of using tables indicative of the functions to be monitored in the program code 100, like in subroutine 224a in FIG. 4, in case of installing a plugin into the management agent MA, the table 230 may be updated or extended to include the additional functions to be monitored in the plugin. In addition, when adding a plugin a certified, for example, signed copy of the plugin may also be given to the validation program 204 as a reference copy so as to allow for the verification of the generated security values. Installing the plugin which is combined with checking functionality, allows the plugins to also be covered within the checking network or guard network.

    [0031] FIG. 5 illustrates an example computing system for implementing the technique described herein. The computing device 300 includes a processor 302 and a memory 304. A bus 306 may be provided for interconnecting the elements of the computing device 300. In accordance with examples, the memory 204 includes a virtual machine monitor 308, like a hypervisor, defining virtual machines VM1 and VM2 on which the first and second processing resources 200, 202, e.g. the main and second operating systems, are installed. The computing device 300 may also include input/output devices 310, for example a keyboard, a cursor control, etc., a communication interface 312, for example a network interface card, a modem, etc., and a storage device 314, for example a hard drive or a compact disc reader.

    [0032] FIG. 6 is a flow diagram of an example method for verifying the correct execution of a program code. At 400, the program code 100 including the embedded execution units 1021 to 1026 is executed by a first processing resource 200, for example by a first or main operating system installed on a virtual machine, like VM1 in FIG. 5. At 402, the validation program 204 is run, for example by the second operating system 202 installed on a second virtual machine, like VM2 in FIG. 5. During execution of the program code 100, at 404, a first security value for a part of the program code 100 is calculated, for example by one of the guards or execution units 1021 to 1026. The validation program 204 receives the first security value at 406 and, at 408, checks the first security value against the second security value calculated from a corresponding part of a reference copy of the program code to obtain a check result. At 410, the check result is returned to the guard or execution unit that performed the calculation of the first security value and, in case it is determined at 412 that there is no match between the first and second security values, the execution unit or guard performs a security-related action at 414. In accordance with examples, the security-related action at 414 may cause a logging of the result as indicated at 416 and the process then may return to 404 where a calculation for checking another part of a program is performed. In accordance with other examples, the security-related action at 414 may cause a halt of the program as indicated at 418. In case it is determined at 412 that the first and second security values match, the method returns to 404 and a new calculation or security value, for example for another part of the program code is carried out, thereby monitoring the integrity of the program code 100 during execution thereof.

    [0033] Although some aspects of the techniques described herein have been described in the context of an apparatus, these aspects may also represent a description of the corresponding method, where a block or device corresponds to a method block or a feature of a method block. Analogously, aspects described in the context of a method also represent a description of a corresponding block or item or feature of a corresponding apparatus.

    [0034] Examples described herein may be realized in the form of hardware, machine readable instructions or a combination of hardware and machine readable instructions. Any such machine readable instructions may be stored in the form of volatile or nonvolatile storage, for example, a storage device such as a ROM, whether erasable or rewritable or not, or in the form of a volatile memory, for example, RAM, memory chips device or integrated circuits or an optically or magnetically readable medium, for example, a CD, DVD, magnetic disc or magnetic tape. The storage devices and storage media are examples of machine readable storage that is suitable for storing a program or programs that, when executed, implement examples described herein.

    [0035] All of the features disclosed in this specification, including any accompanying claims, abstract and drawings, and/or all of the method or process so disclosed may be combined in any combination, except combinations where at least some of the features are mutually exclusive. Each feature disclosed in this specification, including any accompanying claims, abstract and drawings, may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise. Thus, unless expressly stated otherwise, each feature disclosed is one example of a generic series of equivalent or similar features.


    Claims

    1. An apparatus, comprising:
    a first processing resource (200) to execute a program code, the program code including a monitoring block, the monitoring block, during execution of the program code, to:

    issue a check request to a validation program (204),

    on receipt of an ignore check signal provided by the validation program as a response to the check request, decline to calculate a first security value for a part of the program code. and

    otherwise, calculate the first security value and send the first security value to the validation program; and

    a second processing resource (202) separate from the first processing resource, the second processing resource to run the validation program, the validation program to:

    send an ignore check signal to the monitoring block in response to the check request, or

    receive the first security value, check the first security value against a second security value calculated from a corresponding part of a reference copy of the program code to obtain a check result, and return the check result to the monitoring block,

    wherein the monitoring block is to perform a security-related action in response to a check result indicating a mismatch between the first security value and the second security value; and

    wherein an amount of ignore check signals sent to the monitoring block is determined dependent on a threat environment, set by a management server or determined through the analysis of previous check results.


     
    2. The apparatus of claim 1, wherein
    the second processing resource (202) is to store the reference copy of the program code; and
    the validation program (204) is to calculate the second security value using the reference copy of the program code stored at the second processing resource.
     
    3. The apparatus of claim 2, wherein
    the validation program (204) is to receive a type identifier, wherein the type identifier identifies a type of calculation used to calculate the first security value, and wherein the validation program is to calculate the second security value using the identified type of calculation and the reference copy of the program code.
     
    4. The apparatus of claim 1, wherein
    the validation program (204) is to retain a log of a plurality of check results obtained during execution of the program code.
     
    5. The apparatus of claim 1, wherein
    the monitoring block is to issue a check request to the validation program; and
    the validation program (204) is to indicate to the monitoring block a type of calculation to be performed on the part of the program code to calculate the first security value,
    wherein the validation program is to randomly select the type of calculation from a plurality of types of calculation.
     
    6. The apparatus of claim 1, wherein
    the monitoring block is to calculate the first security value using a type of calculation using an initialization vector;
    the monitoring block is to issue a check request to the validation program (204); and
    the validation program is to provide a random initialization vector to the monitoring block.
     
    7. The apparatus of claim 1, wherein
    the monitoring block is to issue a check request to the validation program (204); and
    the validation program is to randomly select from the program code the part for which the first security value is calculated by the monitoring block.
     
    8. The apparatus of claim 1, wherein
    the monitoring block is to issue a check request to the validation program (204), and to calculate the first security value responsive to a reply from the validation program;
    the validation program is to measure a time between sending the reply to the monitoring block and receiving the first security value, and to indicate a suspicious event in case the time measured exceeds a predefined threshold, the predefined threshold being set dependent on a type of calculation used to calculate the first security value.
     
    9. The apparatus of claim 1, wherein
    the security-related action includes logging the mismatch event, halting execution of the program code or repairing the part of the program code using the reference copy.
     
    10. The apparatus of claim 1, wherein
    the validation program (204) is to access a memory associated with the first processing resource, and
    the validation program is to change the program code to modify the monitoring block, or to replace the monitoring block by a new monitoring block, or to add to the program code a new monitoring block.
     
    11. The apparatus of claim 1, wherein
    the first processing resource (200) includes a first operating system;
    the second processing resource (202) includes a second operating system; and
    a processor is to run a virtual machine monitor (308), wherein the virtual machine monitor is to run the first operating system and the second operating system on the processor.
     
    12. A system, comprising:
    an apparatus comprising:
    a first processing resource (200) to execute a program code, the program code including a monitoring block, the monitoring block, during execution of the program code, to:

    issue a check request to a validation program (204),

    on receipt of an ignore check signal provided by the validation program as a response to the check request, decline to calculate a first security value for a part of the program code, and

    otherwise, calculate the first security value and send the first security value to the validation program; and

    a second processing resource (202) separate from the first processing resource, the second processing resource to run the validation program, the validation program to:

    send an ignore check signal to the monitoring block in response to the check request, or

    receive the first security value, check the first security value against a second security value calculated from a corresponding part of a reference copy of the program code to obtain a check result, and return the check result to the monitoring block,

    wherein the monitoring block is to perform a security-related action in response to a check result indicating a mismatch between the first security value and the second security value, and

    wherein an amount of ignore check signals sent to the monitoring block is determined dependent on a threat environment, set by a management server or determined through the analysis of previous check results; and

    a server (218) connected to the second processing resource of the apparatus via a secure channel (220), the server to receive from the second processing resource information about the check result or a log of a plurality of check results for performing an analysis.


     
    13. The system of claim 12, wherein
    the server (218) is to send a message to the validation program (204) in response to the received resource information about the check result or in response to the analysis of the log of the plurality of check results.
     
    14. A method, comprising:

    executing, by a first processing resource (200), a program code, the program code including a monitoring block;

    running, by a second processing resource (202) separate from the first processing resource, a validation program (204);

    during execution of the program code,

    issuing, by the monitoring block, a check request to the validation program;

    on receipt of an ignore check signal provided by the validation program as a response to the check request, refraining, by the monitoring block, from calculating a first security value for a part of the program code,

    otherwise, calculating, by the monitoring block, the first security value and send the first security value to the validation program;

    sending, by the validation program, an ignore check signal to the monitoring block in response to the check request, or

    receiving, by the validation program, the first security value;

    checking, by the validation program, the first security value against a second security value calculated from a corresponding part of a reference copy of the program code to obtain a check result;

    returning, by the validation program, the check result to the monitoring block; and

    performing, by the monitoring block, a security-related action in response to a check result indicating a mismatch between the first security value and the second security value;

    wherein an amount of ignore check signals sent to the monitoring block is determined dependent on a threat environment, set by a management server or determined through the analysis of previous check results.


     


    Ansprüche

    1. Einrichtung, die Folgendes umfasst:
    eine erste Verarbeitungsressource (200), um einen Programmcode auszuführen, wobei der Programmcode einen Überwachungsblock einschließt, wobei der Überwachungsblock während der Ausführung des Programmcodes zu Folgendem dient:

    Ausgeben einer Überprüfungsanforderung an ein Validierungsprogramm (204),

    auf Empfang eines Signals zum Ignorieren der Überprüfung, das durch das Validierungsprogramm als Antwort auf die Überprüfungsanforderung bereitgestellt wird, Ablehnen eines Berechnens eines erstens Sicherheitswerts für einen Teil des Programmcodes, und

    andernfalls Berechnen des ersten Sicherheitswerts und Senden des ersten Sicherheitswerts an das Validierungsprogramm; und

    eine zweite Verarbeitungsressource (202), die von der ersten Verarbeitungsressource getrennt ist, wobei die zweite Verarbeitungsressource das Validierungsprogramm ablaufen lassen soll, wobei das Validierungsprogramm zu Folgendem dient:

    Senden eines Signals zum Ignorieren der Überprüfung an den Überwachungsblock als Reaktion auf die Überprüfungsanforderung oder

    Empfangen des ersten Sicherheitswerts, Überprüfen des ersten Sicherheitswerts durch Vergleich mit einem zweiten Sicherheitswert, der aus einem entsprechenden Teil einer Referenzkopie des Programmcodes berechnet wurde, um ein Überprüfungsergebnis zu erhalten, und Zurückgeben des Überprüfungsergebnisses an den Überwachungsblock,

    wobei der Überwachungsblock als Reaktion auf ein Überprüfungsergebnis durchführen soll, das eine Diskrepanz zwischen dem ersten Sicherheitswert und dem zweiten Sicherheitswert eine sicherheitsbezogene Aktion anzeigt; und

    wobei eine Anzahl von Signalen zum Ignorieren der Überprüfung, die an den Überwachungsblock gesendet wurden, in Abhängigkeit von einer Bedrohungsumgebung bestimmt wird, die durch einen Verwaltungsserver festgelegt oder durch die Analyse vorheriger Überprüfungsergebnisse bestimmt wird.


     
    2. Einrichtung nach Anspruch 1, wobei

    die zweite Verarbeitungsressource (202) die Referenzkopie des Programmcodes speichern soll; und

    das Validierungsprogramm (204) den zweiten Sicherheitswert unter Verwendung der Referenzkopie des Programmcodes, der an der zweiten Verarbeitungsressource gespeichert ist, berechnen soll.


     
    3. Einrichtung nach Anspruch 2, wobei
    das Validierungsprogramm (204) einen Typenbezeichner empfangen soll, wobei der Typenbezeichner eine Berechnungsart identifiziert, die verwendet wird, um den ersten Sicherheitswert zu berechnen, und wobei das Validierungsprogramm den zweiten Sicherheitswert unter Verwendung der identifizierten Berechnungsart und der Referenzkopie des Programmcodes berechnen soll.
     
    4. Einrichtung nach Anspruch 1, wobei
    das Validierungsprogramm (204) ein Protokoll mehrerer Überprüfungsergebnisse beibehalten soll, die während der Ausführung des Programmcodes erhalten werden.
     
    5. Einrichtung nach Anspruch 1, wobei

    der Überwachungsblock eine Überprüfungsanforderung an das Validierungsprogramm ausgeben soll; und

    das Validierungsprogramm (204) dem Überwachungsblock eine Berechnungsart anzeigen soll, die seitens des Programmcodes durchgeführt werden soll, um den ersten Sicherheitswert zu berechnen, wobei das Validierungsprogramm die Berechnungsart zufällig aus mehreren Berechnungsarten auswählen soll.


     
    6. Einrichtung nach Anspruch 1, wobei

    der Überwachungsblock den ersten Sicherheitswert unter Verwendung einer Berechnungsart unter Verwendung eines Initialisierungsvektors berechnen soll;

    der Überwachungsblock eine Überprüfungsanforderung an das Validierungsprogramm (204) ausgeben soll; und

    das Validierungsprogramm dem Überwachungsblock einen zufälligen Initialisierungsvektor bereitstellen soll.


     
    7. Einrichtung nach Anspruch 1, wobei

    der Überwachungsblock eine Überprüfungsanforderung an das Validierungsprogramm (204) ausgeben soll; und

    das Validierungsprogramm aus dem Programmcode zufällig den Teil auswählt, für den der erste Sicherheitswert durch den Überwachungsblock berechnet wird.


     
    8. Einrichtung nach Anspruch 1, wobei

    der Überwachungsblock eine Überprüfungsanforderung an das Validierungsprogramm (204) ausgeben und als Reaktion auf eine Antwort von dem Validierungsprogramm den ersten Sicherheitswert berechnen soll;

    das Validierungsprogramm eine Zeit zwischen dem Senden der Antwort an den Überwachungsblock und dem Empfangen des ersten Sicherheitswerts messen und ein verdächtiges Ereignis anzeigen soll, falls die gemessene Zeit einen vordefinierten Schwellenwert überschreitet, wobei der vordefinierte Schwellenwert abhängig von einer Berechnungsart festgelegt wird, die verwendet wird, um den ersten Sicherheitswert zu berechnen.


     
    9. Einrichtung nach Anspruch 1, wobei
    die sicherheitsbezogene Aktion ein Protokollieren des Diskrepanzereignisses, ein Anhalten der Ausführung des Programmcodes oder ein Reparieren des Teils des Programmcodes unter Verwendung der Referenzkopie einschließt.
     
    10. Einrichtung nach Anspruch 1, wobei

    das Validierungsprogramm (204) auf einen Speicher zugreifen soll, der der ersten Verarbeitungsressource zugeordnet ist, und

    das Validierungsprogramm den Programmcode ändern soll, um den Überwachungsblock zu modifizieren oder den Überwachungsblock durch einen neuen Überwachungsblock zu ersetzen oder um dem Programmcode einen neuen Überwachungsblock hinzuzufügen.


     
    11. Einrichtung nach Anspruch 1, wobei

    die erste Verarbeitungsressource (200) ein erstes Betriebssystem einschließt;

    die zweite Verarbeitungsressource (202) ein zweites Betriebssystem einschließt; und

    ein Prozessor einen Monitor einer virtuellen Maschine (308) laufen lassen soll, wobei der Monitor einer virtuellen Maschine das erste Betriebssystem und das zweite Betriebssystem auf dem Prozessor laufen lassen soll.


     
    12. System, das Folgendes umfasst:
    eine Einrichtung, die Folgendes umfasst:
    eine erste Verarbeitungsressource (200), um einen Programmcode auszuführen, wobei der Programmcode einen Überwachungsblock einschließt, wobei der Überwachungsblock während der Ausführung des Programmcodes zu Folgendem dient:

    Ausgeben einer Überprüfungsanforderung an ein Validierungsprogramm (204),

    auf Empfang eines Signals zum Ignorieren der Überprüfung, das durch das Validierungsprogramm als Antwort auf die Überprüfungsanforderung bereitgestellt wird, Ablehnen eines Berechnens eines erstens Sicherheitswerts für einen Teil des Programmcodes und

    andernfalls Berechnen des ersten Sicherheitswerts und Senden des ersten Sicherheitswerts an das Validierungsprogramm; und

    eine zweite Verarbeitungsressource (202), die von der ersten Verarbeitungsressource getrennt ist, wobei die zweite Verarbeitungsressource das Validierungsprogramm ablaufen lassen soll, wobei das Validierungsprogramm zu Folgendem dient:

    Senden eines Signals zum Ignorieren der Überprüfung an den Überwachungsblock als Reaktion auf die Überprüfungsanforderung oder

    Empfangen des ersten Sicherheitswerts, Überprüfen des ersten Sicherheitswerts durch Vergleich mit einem zweiten Sicherheitswert, der aus einem entsprechenden Teil einer Referenzkopie des Programmcodes berechnet wurde, um ein Überprüfungsergebnis zu erhalten, und Zurückgeben des Überprüfungsergebnisses an den Überwachungsblock,

    wobei der Überwachungsblock eine sicherheitsbezogene Aktion als Reaktion auf ein Überprüfungsergebnis durchführen soll, das eine Diskrepanz zwischen dem ersten Sicherheitswert und dem zweiten Sicherheitswert anzeigt, und

    wobei eine Anzahl von Signalen zum Ignorieren der Überprüfung, die an den Überwachungsblock gesendet wurden, in Abhängigkeit von einer Bedrohungsumgebung bestimmt wird, die durch einen Verwaltungsserver festgelegt oder durch die Analyse vorheriger Überprüfungsergebnisse bestimmt wird; und

    einen Server (218), der über einen sicheren Kanal (220) mit der zweiten Verarbeitungsressource der Einrichtung verbunden ist, wobei der Server von der zweiten Verarbeitungsressource Informationen über das Überprüfungsergebnis oder ein Protokoll mehrerer Überprüfungsergebnisse zum Durchführen einer Analyse empfängt.


     
    13. System nach Anspruch 12, wobei
    der Server (218) als Reaktion auf die empfangenen Ressourceninformationen über das Überprüfungsergebnis oder als Reaktion auf die Analyse des Protokolls der mehreren Überprüfungsergebnisse eine Nachricht an das Validierungsprogramm (204) senden soll.
     
    14. Verfahren, das Folgendes umfasst:

    Ausführen, durch eine erste Verarbeitungsressource (200), eines Programmcodes, wobei der Programmcode einen Überwachungsblock einschließt;

    Laufenlassen, durch eine zweite Verarbeitungsressource (202), die von der ersten Verarbeitungsressource getrennt ist, eines Validierungsprogramms (204);

    während der Ausführung des Programmcodes,

    Ausgeben, durch den Überwachungsblock, einer Überprüfungsanforderung an das Validierungsprogramm;

    auf Empfang eines durch das Validierungsprogramm als Antwort auf die Überprüfungsanforderung bereitgestellten Signals zum Ignorieren der Überprüfung, Unterlassen, durch den Überwachungsblock, des Berechnens eines ersten Sicherheitswerts für einen Teil des Programmcodes,

    andernfalls Berechnen, durch den Überwachungsblock, des ersten Sicherheitswerts und Senden des ersten Sicherheitswerts an das Validierungsprogramm;

    Senden, durch das Validierungsprogramm, eines Signals zum Ignorieren der Überprüfung an den Überwachungsblock als Reaktion auf die Überprüfungsanforderung oder

    Empfangen, durch das Validierungsprogramm, des ersten Sicherheitswerts;

    Überprüfen, durch das Validierungsprogramm, des ersten Sicherheitswerts durch Vergleich mit einem zweiten Sicherheitswert, der aus einem entsprechenden Teil einer Referenzkopie des Programmcodes berechnet wurde, um ein Überprüfungsergebnis zu erhalten;

    Zurückgeben, durch das Validierungsprogramm, des Überprüfungsergebnisses an den Überwachungsblock; und

    Durchführen, durch den Überwachungsblock, einer sicherheitsbezogenen Aktion als Reaktion auf ein Überprüfungsergebnis, das eine Diskrepanz zwischen dem ersten Sicherheitswert und dem zweiten Sicherheitswert anzeigt;

    wobei eine Anzahl von Signalen zum Ignorieren der Überprüfung, die an den Überwachungsblock gesendet wurden, in Abhängigkeit von einer Bedrohungsumgebung bestimmt wird, die durch einen Verwaltungsserver festgelegt oder durch die Analyse vorheriger Überprüfungsergebnisse bestimmt wird.


     


    Revendications

    1. Appareil, comprenant :
    une première ressource de traitement (200) pour exécuter un code de programme, le code de programme comportant un bloc de surveillance, le bloc de surveillance, pendant l'exécution du code de programme, étant destiné à :

    émettre une demande de vérification à un programme de validation (204),

    à la réception d'un signal d'ignorance de vérification fourni par le programme de validation en réponse à la demande de vérification, refuser de calculer une première valeur de sécurité pour une partie du code de programme, et

    sinon, calculer la première valeur de sécurité et envoyer la première valeur de sécurité au programme de validation ; et

    une seconde ressource de traitement (202) séparée de la première ressource de traitement, la seconde ressource de traitement devant exécuter le programme de validation, le programme de validation étant destiné à :

    envoyer un signal d'ignorance de vérification au bloc de surveillance en réponse à la demande de vérification, ou

    recevoir la première valeur de sécurité, vérifier la première valeur de sécurité par rapport à une seconde valeur de sécurité calculée à partir d'une partie correspondante d'une copie de référence du code de programme pour obtenir un résultat de vérification, et renvoyer le résultat de vérification au bloc de surveillance,

    dans lequel le bloc de surveillance doit effectuer une action liée à la sécurité en réponse à un résultat de vérification indiquant une non-correspondance entre la première valeur de sécurité et la seconde valeur de sécurité ; et

    dans lequel une quantité de signaux d'ignorance de vérification envoyés au bloc de surveillance est déterminée en fonction d'un environnement de menace, définie par un serveur de gestion ou déterminée par l'analyse des résultats de vérification précédents.


     
    2. Appareil selon la revendication 1, dans lequel
    la seconde ressource de traitement (202) doit stocker la copie de référence du code de programme ; et
    le programme de validation (204) doit calculer la seconde valeur de sécurité à l'aide de la copie de référence du code de programme stocké au niveau de la seconde ressource de traitement.
     
    3. Appareil selon la revendication 2, dans lequel
    le programme de validation (204) doit recevoir un identifiant de type, l'identifiant de type identifiant un type de calcul utilisé pour calculer la première valeur de sécurité, et le programme de validation devant calculer la seconde valeur de sécurité à l'aide du type de calcul identifié et de la copie de référence du code de programme.
     
    4. Appareil selon la revendication 1, dans lequel
    le programme de validation (204) doit conserver un journal d'une pluralité de résultats de vérification obtenus pendant l'exécution du code de programme.
     
    5. Appareil selon la revendication 1, dans lequel
    le bloc de surveillance doit émettre une demande de vérification au programme de validation ; et
    le programme de validation (204) doit indiquer au bloc de surveillance un type de calcul à effectuer sur la partie du code de programme pour calculer la première valeur de sécurité,
    le programme de validation devant sélectionner de manière aléatoire le type de calcul parmi une pluralité de types de calcul.
     
    6. Appareil selon la revendication 1, dans lequel
    le bloc de surveillance doit calculer la première valeur de sécurité à l'aide d'un type de calcul utilisant un vecteur d'initialisation ;
    le bloc de surveillance doit émettre une demande de vérification au programme de validation (204) ; et
    le programme de validation doit fournir un vecteur d'initialisation aléatoire au bloc de surveillance.
     
    7. Appareil selon la revendication 1, dans lequel
    le bloc de surveillance doit émettre une demande de vérification au programme de validation (204) ; et
    le programme de validation doit sélectionner de manière aléatoire depuis le code de programme la partie pour laquelle la première valeur de sécurité est calculée par le bloc de surveillance.
     
    8. Appareil selon la revendication 1, dans lequel
    le bloc de surveillance doit émettre une demande de vérification au programme de validation (204), et calculer la première valeur de sécurité en réponse à une réponse du programme de validation ;
    le programme de validation doit mesurer un temps entre l'envoi de la réponse au bloc de surveillance et la réception de la première valeur de sécurité, et indiquer un événement suspect au cas où le temps mesuré dépasse un seuil prédéfini, le seuil prédéfini étant fixé en fonction d'un type de calcul utilisé pour calculer la première valeur de sécurité.
     
    9. Appareil selon la revendication 1, dans lequel
    l'action liée à la sécurité comporte la journalisation de l'événement de non-correspondance, l'arrêt de l'exécution du code de programme ou la réparation de la partie du code de programme à l'aide de la copie de référence.
     
    10. Appareil selon la revendication 1, dans lequel
    le programme de validation (204) doit accéder à une mémoire associée à la première ressource de traitement, et
    le programme de validation doit changer le code de programme pour modifier le bloc de surveillance, ou remplacer le bloc de surveillance par un nouveau bloc de surveillance, ou ajouter au code de programme un nouveau bloc de surveillance.
     
    11. Appareil selon la revendication 1, dans lequel
    la première ressource de traitement (200) comporte un premier système d'exploitation ;
    la seconde ressource de traitement (202) comporte un second système d'exploitation ; et
    un processeur doit exécuter un moniteur de machine virtuelle (308), le moniteur de machine virtuelle devant exécuter le premier système d'exploitation et le second système d'exploitation sur le processeur.
     
    12. Système, comprenant :
    un appareil comprenant :
    une première ressource de traitement (200) pour exécuter un code de programme, le code de programme comportant un bloc de surveillance, le bloc de surveillance, pendant l'exécution du code de programme, étant destiné à :

    émettre une demande de vérification à un programme de validation (204),

    à la réception d'un signal d'ignorance de vérification fourni par le programme de validation en réponse à la demande de vérification, refuser de calculer une première valeur de sécurité pour une partie du code de programme, et

    sinon, calculer la première valeur de sécurité et envoyer la première valeur de sécurité au programme de validation ; et

    une seconde ressource de traitement (202) séparée de la première ressource de traitement, la seconde ressource de traitement devant exécuter le programme de validation, le programme de validation étant destiné à :

    envoyer un signal d'ignorance de vérification au bloc de surveillance en réponse à la demande de vérification, ou

    recevoir la première valeur de sécurité, vérifier la première valeur de sécurité par rapport à une seconde valeur de sécurité calculée à partir d'une partie correspondante d'une copie de référence du code de programme pour obtenir un résultat de vérification, et renvoyer le résultat de vérification au bloc de surveillance,

    dans lequel le bloc de surveillance doit effectuer une action liée à la sécurité en réponse à un résultat de vérification indiquant une non-correspondance entre la première valeur de sécurité et la seconde valeur de sécurité, et

    dans lequel une quantité de signaux d'ignorance de vérification envoyés au bloc de surveillance sont déterminés en fonction d'un environnement de menace, définis par un serveur de gestion ou déterminés par l'analyse des résultats de vérification précédents ; et

    un serveur (218) connecté à la seconde ressource de traitement de l'appareil par l'intermédiaire d'un canal sécurisé (220), le serveur devant recevoir à partir de la seconde ressource de traitement des informations concernant le résultat de vérification ou un journal d'une pluralité de résultats de vérification destinés à la réalisation d'une analyse.


     
    13. Système selon la revendication 12, dans lequel
    le serveur (218) doit envoyer un message au programme de validation (204) en réponse aux informations de ressource reçues concernant le résultat de vérification ou en réponse à l'analyse du journal de la pluralité de résultats de vérification.
     
    14. Procédé comprenant :

    l'exécution, par une première ressource de traitement (200), d'un code de programme, le code de programme comportant un bloc de surveillance ;

    l'exécution, par une seconde ressource de traitement (202) séparée de la première ressource de traitement, d'un programme de validation (204) ;

    lors de l'exécution du code de programme,

    l'émission, par le bloc de surveillance, d'une demande de vérification au programme de validation ;

    à la réception d'un signal d'ignorance de vérification fourni par le programme de validation en réponse à la demande de vérification, l'abstention, par le bloc de surveillance, du fait de calculer une première valeur de sécurité pour une partie du code de programme,

    sinon, le calcul, par le bloc de surveillance, de la première valeur de sécurité et l'envoi de la première valeur de sécurité au programme de validation ;

    l'envoi, par le programme de validation, d'un signal d'ignorance de vérification au bloc de surveillance en réponse à la demande de vérification, ou la réception, par le programme de validation, de la première valeur de sécurité ;

    la vérification, par le programme de validation, de la première valeur de sécurité par rapport à une seconde valeur de sécurité calculée à partir d'une partie correspondante d'une copie de référence du code de programme pour obtenir un résultat de vérification ;

    le renvoi, par le programme de validation, du résultat de vérification au bloc de surveillance ; et

    la réalisation, par le bloc de surveillance, d'une action liée à la sécurité en réponse à un résultat de vérification indiquant une non-correspondance entre la première valeur de sécurité et la seconde valeur de sécurité ;

    dans lequel une quantité de signaux d'ignorance de vérification envoyés au bloc de surveillance est déterminée en fonction d'un environnement de menace, définie par un serveur de gestion ou déterminée par l'analyse des résultats de vérification précédents.


     




    Drawing





















    REFERENCES CITED IN THE DESCRIPTION



    This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

    Patent documents cited in the description