Apparatus and method for implementing a unified hash algorithm pipeline

• US 7,684,563 B1
• Filed: 10/19/2004
• Issued: 03/23/2010
• Est. Priority Date: 12/12/2003
• Status: Active Grant

• Current Assignee: Oracle America, Inc. (Oracle Corporation)
• Original Assignee: Sun Microsystems Incorporated (Oracle Corporation)
• Inventors: Olson, Christopher H., Grohoski, Gregory F., Rarick, Leonard D.
• Primary Examiner(s): Zand; Kambiz
• Assistant Examiner(s): Turchen; James
• Application Number:
  US10/968,428
• Time in Patent Office: 1,981 Days
• Field of Search: 380/28, 712/1, 712/36, 712/206
• US Class Current: 380/28
• CPC Class Codes: G06F 9/3895 for complex operations, e.g. multidimensional or interleaved address generators, macros
• CPC Class Codes: H04L 2209/125 Parallelization or pipelining, e.g. for accelerating processing of cryptographic operations
• CPC Class Codes: H04L 9/0643 Hash functions, e.g. MD5, SHA, HMAC or f9 MAC
• CPC Class Codes: H04L 9/16 the keys or algorithms being changed during operation
• CPC Class Codes: H04L 9/50 using hash chains, e.g. blockchains or hash trees

Abstract

An apparatus and method for implementing a unified hash algorithm pipeline. In one embodiment, a cryptographic unit may include hash logic configured to compute a hash value of a data block according to a hash algorithm, where the hash algorithm is dynamically selectable from a plurality of hash algorithms, and where the hash logic comprises a plurality of pipeline stages each configured to compute a portion of the hash algorithm. The cryptographic unit may further include a word buffer configured to store the data block during computing by the hash logic.

Claims

1. A cryptographic unit, comprising:

hash logic configured to compute a hash value of a first data block according to a first hash algorithm, wherein said first hash algorithm is dynamically selectable from a plurality of hash algorithms, wherein said hash logic comprises pipelined adder logic having a plurality of pipeline stages, and wherein to compute said hash value of said first data block according to said first hash algorithm, said hash logic is further configured to perform multiple iterations within said pipelined adder logic, wherein for at least a given one of the iterations of the first hash algorithm, an input of the given iteration depends on an output of said pipelined adder logic produced by a preceding iteration; and

a word buffer configured to store said first data block during computing by said hash logic;

wherein said pipelined adder logic is configured such that during computation of said hash value of said first data block and after said multiple iterations begin, each of said plurality of pipeline stages is idle during one or more execution cycles;

wherein said hash logic is further configured to compute a hash value of a second data block according to a second hash algorithm, wherein to compute said hash value of said second data block according to said second hash algorithm, said hash logic is further configured to perform multiple iterations within said pipelined adder logic, wherein for at least a particular one of the iterations of the second hash algorithm, an input of the particular iteration depends on an output of said pipelined adder logic produced by a preceding iteration; and

wherein said hash logic is further configured to compute said hash value of said second data block during idle execution cycles of said plurality of pipeline stages that occur during computation of said hash value of said first data block, such that said hash logic is operable to concurrently compute hash values of said first data block and said second data block in an interleaved fashion.

2. The cryptographic unit as recited in claim 1, wherein said plurality of hash algorithms includes at least one of a hash algorithm compliant with a version of Message Digest 5 (MD5) hash functionality, a hash algorithm compliant with a version of Secure Hash Algorithm-1 (SHA-1) hash functionality, or a hash algorithm compliant with a version of Secure Hash Algorithm-256 (SHA-256) hash functionality.

3. The cryptographic unit as recited in claim 1, wherein said first data block is associated with a first one of a plurality of processor cores, and wherein said second data block is associated with a second one of said plurality of processor cores, such that to compute respective hash values of said first and said second data blocks, said first and said second processor cores are configured to share said hash logic.

4. The cryptographic unit as recited in claim 1, wherein said first hash algorithm is equivalent to said second hash algorithm.

5. The cryptographic unit as recited in claim 1, wherein a first one of said plurality of pipeline stages comprises a first adder configured to perform a first portion of said first hash algorithm, and wherein a second one of said plurality of pipeline stages comprises a second adder configured to perform a second portion of said first hash algorithm.

6. The cryptographic unit as recited in claim 5, wherein said first pipeline stage further comprises carry save adder logic coupled to an input of said first adder and rotator logic coupled to an output of said first adder, wherein an input of said second adder is coupled to receive said output of said first adder, and wherein said second pipeline stage further comprises a Boolean logic module coupled to said output of said second adder.

7. The cryptographic unit as recited in claim 5, wherein said first pipeline stage is configured to produce a first result during a first execution cycle, and wherein said second pipeline stage is configured to produce a second result dependent upon said first result during a second execution cycle.

8. The cryptographic unit as recited in claim 7, wherein said first pipeline stage is further configured to produce a third result dependent upon said first result during said second execution cycle.

9. The cryptographic unit as recited in claim 1, wherein said hash logic and said word buffer are fabricated as a single integrated circuit.

10. A processor, comprising:

the cryptographic unit as recited in claim 1; and

instruction fetch logic configured to issue a first instruction from one of a plurality of threads during one execution cycle and to issue a second instruction from another one of said plurality of threads during a successive execution cycle;

wherein said cryptographic unit is configured to execute independently of said instruction fetch logic.

11. A processor, comprising:

the cryptographic unit as recited in claim 1; and

instruction fetch logic configured to issue a plurality of instructions including a first and a second instruction, wherein said first instruction is issued from one of a plurality of threads during one execution cycle, and wherein said second instruction is issued from another one of said plurality of threads during a successive execution cycle;

wherein said cryptographic unit is configured to execute one of said plurality of instructions issued by said instruction fetch logic.

12. A method, comprising:

dynamically selecting a first hash algorithm from a plurality of hash algorithms;

storing a first data block in a word buffer;

computing a hash value of said first data block according to said first hash algorithm in pipelined adder logic having a plurality of pipeline stages, wherein computing said hash value of said first data block comprises performing multiple iterations within said pipelined adder logic, wherein for at least a given one of the iterations of the first hash algorithm, an input of the given iteration depends on an output of said pipelined adder logic produced by a preceding iteration, and wherein during computation of said hash value of said first data block and after said multiple iterations begin, each of said plurality of pipeline stages is idle during one or more execution cycles;

computing a hash value of a second data block according to a second hash algorithm, wherein computing said hash value of said second data block comprises performing multiple iterations within said pipelined adder logic, wherein for at least a particular one of the iterations of the second hash algorithm, an input of the particular iteration depends on an output of said pipelined adder logic produced by a preceding iteration;

wherein computing said hash value of said second data block occurs during idle execution cycles of said plurality of pipeline stages that occur during computation of said hash value of said first data block, such that computing of said hash values of said first data block and said second data block occurs concurrently in an interleaved fashion within said pipelined adder logic.

13. The method as recited in claim 12, wherein said plurality of hash algorithms includes at least one of a hash algorithm compliant with a version of Message Digest 5 (MD5) hash functionality, a hash algorithm compliant with a version of Secure Hash Algorithm-1 (SHA-1) hash functionality, or a hash algorithm compliant with a version of Secure Hash Algorithm-256 (SHA-256) hash functionality.

14. The method as recited in claim 12, further comprising:

a first one of a plurality of processor cores issuing an operation to compute said hash value of said first data block to a cryptographic unit comprising said pipelined adder logic; and

a second one of said plurality of processor cores issuing an operation to compute said hash value of said second data block to said cryptographic unit, such that during computation of said hash values of said first and said second data blocks, said first and said second processor cores share said cryptographic unit.

15. The method as recited in claim 12, wherein said first hash algorithm is equivalent to said second hash algorithm.

16. The method as recited in claim 12, wherein a first one of said plurality of pipeline stages comprises a first adder configured to perform a first portion of said first hash algorithm, and wherein a second one of said plurality of pipeline stages comprises a second adder configured to perform a second portion of said first hash algorithm.

17. The method as recited in claim 16, wherein said first pipeline stage further comprises carry save adder logic coupled to an input of said first adder and rotator logic coupled to an output of said first adder, wherein an input of said second adder is coupled to receive said output of said first adder, and wherein said second pipeline stage further comprises a Boolean logic module coupled to said output of said second adder.

18. The method as recited in claim 16, wherein said first pipeline stage is configured to produce a first result during a first execution cycle, and wherein said second pipeline stage is configured to produce a second result dependent upon said first result during a second execution cycle.

19. The method as recited in claim 18, wherein said first pipeline stage is further configured to produce a third result dependent upon said first result during said second execution cycle.

20. The method as recited in claim 12, further comprising:

issuing a first instruction from one of a plurality of threads during one execution cycle; and

issuing a second instruction from another one of said plurality of threads during a successive execution cycle;

wherein computing said hash value of said first data block is configured to operate independently of issuing instructions from said plurality of threads.

21. The method as recited in claim 12, further comprising issuing a plurality of instructions including a first instruction and a second instruction, wherein said first instruction is issued from one of a plurality of threads during one execution cycle, wherein said second instruction is issued from another one of said plurality of threads during a successive execution cycle, and wherein computing said hash value of said first data block occurs in response to issuing a particular one of said plurality of instructions.

22. A system, comprising:

a system memory; and

a processor coupled to said system memory and comprising a cryptographic unit, said cryptographic unit comprising;

hash logic configured to compute a hash value of a first data block according to a first hash algorithm, wherein said first hash algorithm is dynamically selectable from a plurality of hash algorithms, wherein said hash logic comprises pipelined adder logic having a plurality of pipeline stages, and wherein to compute said hash value of said first data block according to said first hash algorithm, said hash logic is further configured to perform multiple iterations within said pipelined adder logic, wherein for at least a given one of the iterations of the first hash algorithm, an input of the given iteration depends on an output of said pipelined adder logic produced by a preceding iteration; and

a word buffer configured to store said first data block during computing by said hash logic;

wherein said pipelined adder logic is configured such that during computation of said hash value of said first data block and after said multiple iterations begin, each of said plurality of pipeline stages is idle during one or more execution cycles; and

wherein said hash logic is further configured to compute a hash value of a second data block according to a second hash algorithm, wherein to compute said hash value of said second data block according to said second hash algorithm, said hash logic is further configured to perform multiple iterations within said pipelined adder logic, wherein for at least a particular one of the iterations of the second hash algorithm, an input of the particular iteration depends on an output of said pipelined adder logic produced by a preceding iteration;

wherein said hash logic is further configured to compute said hash value of said second data block during idle execution cycles of said plurality of pipeline stages that occur during computation of said hash value of said first data block, such that said hash logic is operable to concurrently compute hash values of said first data block and said second data block in an interleaved fashion.

23. The system as recited in claim 22, wherein said plurality of hash algorithms includes at least one of a hash algorithm compliant with a version of Message Digest 5 (MD5) hash functionality, a hash algorithm compliant with a version of Secure Hash Algorithm-1 (SHA-1) hash functionality, or a hash algorithm compliant with a version of Secure Hash Algorithm-256 (SHA-256) hash functionality.

24. The system as recited in claim 22, wherein said processor further comprises a plurality of processor cores, wherein said first data block is associated with a first one of said plurality of processor cores, and wherein said second data block is associated with a second one of said plurality of processor cores, such that to compute respective hash values of said first and said second data blocks, said first and said second processor cores are configured to share said hash logic.

25. The system as recited in claim 22, wherein said first hash algorithm is equivalent to said second hash algorithm.

26. The system as recited in claim 22, wherein a first one of said plurality of pipeline stages comprises a first adder configured to perform a first portion of said first hash algorithm, and wherein a second one of said plurality of pipeline stages comprises a second adder configured to perform a second portion of said first hash algorithm.

27. The system as recited in claim 26, wherein said first pipeline stage further comprises carry save adder logic coupled to an input of said first adder and rotator logic coupled to an output of said first adder, wherein an input of said second adder is coupled to receive said output of said first adder, and wherein said second pipeline stage further comprises a Boolean logic module coupled to said output of said second adder.

28. The system as recited in claim 26, wherein said first pipeline stage is configured to produce a first result during a first execution cycle, and wherein said second pipeline stage is configured to produce a second result dependent upon said first result during a second execution cycle.

29. The system as recited in claim 28, wherein said first pipeline stage is further configured to produce a third result dependent upon said first result during said second execution cycle.

30. The system as recited in claim 22, wherein said hash logic and said word buffer are fabricated as a single integrated circuit.

31. The system as recited in claim 22, wherein said processor further comprises instruction fetch logic configured to issue a first instruction from one of a plurality of threads during one execution cycle and to issue a second instruction from another one of said plurality of threads during a successive execution cycle, and wherein said cryptographic unit is configured to execute independently of said instruction fetch logic.

32. The system as recited in claim 22, wherein said processor further comprises instruction fetch logic configured to issue a plurality of instructions including a first and a second instruction, wherein said first instruction is issued from one of a plurality of threads during one execution cycle, wherein said second instruction is issued from another one of said plurality of threads during a successive execution cycle, and wherein said cryptographic unit is configured to execute one of said plurality of instructions issued by said instruction fetch logic.