• Top
    • Documentation
    • Books
    • Boolean-reasoning
    • Projects
    • Debugging
    • Community
    • Std
    • Proof-automation
    • Macro-libraries
    • ACL2
    • Interfacing-tools
    • Hardware-verification
    • Software-verification
      • Kestrel-books
        • Crypto-hdwallet
        • Apt
        • Error-checking
        • Fty-extensions
        • Isar
        • Kestrel-utilities
        • Set
        • C
          • Syntax-for-tools
          • Atc
          • Transformation-tools
          • Language
            • Abstract-syntax
            • Integer-ranges
            • Implementation-environments
              • Integer-format-templates
                • Integer-format
                • Uinteger-sinteger-bit-roles-wfp
                • Sinteger-format
                • Uinteger-format
                • Sinteger-format->min
                  • Integer-format-inc-sign-tcnpnt
                  • Sinteger-bit-role
                  • Sinteger-bit-roles-wfp
                  • Uinteger-bit-role
                  • Uinteger-bit-roles-wfp
                  • Uinteger+sinteger-format
                  • Sinteger-bit-roles-value-count
                  • Uinteger-bit-roles-value-count
                  • Sinteger-bit-roles-inc-n-and-sign
                  • Uinteger-bit-roles-inc-n
                  • Sinteger-format-inc-sign-tcnpnt
                  • Sinteger-bit-roles-inc-n
                  • Uinteger-bit-roles-exponents
                  • Sinteger-bit-roles-exponents
                  • Sinteger-format->max
                  • Integer-format->bit-size
                  • Uinteger-format->max
                  • Sinteger-bit-roles-sign-count
                  • Integer-format->signed-min
                  • Uinteger-format-inc-npnt
                  • Integer-format->unsigned-max
                  • Integer-format->signed-max
                  • Uinteger-bit-role-list
                  • Sinteger-bit-role-list
                  • Uinteger-sinteger-bit-roles-wfp-of-inc-n-and-sign
                • Ienv
                • Integer-formats
                • Ienv-ushort-rangep
                • Ienv-ulong-rangep
                • Ienv-ullong-rangep
                • Ienv-uint-rangep
                • Ienv-uchar-rangep
                • Ienv-sshort-rangep
                • Ienv-slong-rangep
                • Ienv-sllong-rangep
                • Ienv-sint-rangep
                • Ienv-schar-rangep
                • Ienv-char-rangep
                • Ienv->schar-min
                • Ienv->schar-max
                • Ienv->char-size
                • Ienv->char-min
                • Ienv->char-max
                • Ienv->uchar-max
                • Ienv->short-bit-size
                • Ienv->long-bit-size
                • Ienv->llong-bit-size
                • Ienv->int-bit-size
                • Ienv->short-byte-size
                • Ienv->long-byte-size
                • Ienv->llong-byte-size
                • Ienv->int-byte-size
                • Versions
                • Ienv->ushort-max
                • Ienv->ullong-max
                • Ienv->sshort-min
                • Ienv->sllong-min
                • Ienv->sllong-max
                • Ienv->bool-byte-size
                • Ienv->bool-bit-size
                • Ienv->ulong-max
                • Ienv->uint-max
                • Ienv->sshort-max
                • Ienv->slong-min
                • Ienv->slong-max
                • Ienv->sint-min
                • Ienv->sint-max
                • Schar-formats
                • Char-formats
                • Uchar-formats
                • Signed-formats
                • Bool-formats
              • Dynamic-semantics
              • Static-semantics
              • Grammar
              • Types
              • Integer-formats-definitions
              • Computation-states
              • Portable-ascii-identifiers
              • Values
              • Integer-operations
              • Object-designators
              • Operations
              • Errors
              • Tag-environments
              • Function-environments
              • Character-sets
              • Flexible-array-member-removal
              • Arithmetic-operations
              • Pointer-operations
              • Real-operations
              • Array-operations
              • Scalar-operations
              • Structure-operations
            • Representation
            • Insertion-sort
            • Pack
          • Soft
          • Bv
          • Imp-language
          • Ethereum
          • Event-macros
          • Java
          • Riscv
          • Bitcoin
          • Zcash
          • Yul
          • ACL2-programming-language
          • Prime-fields
          • Json
          • Syntheto
          • File-io-light
          • Cryptography
          • Number-theory
          • Axe
          • Lists-light
          • Builtins
          • Solidity
          • Helpers
          • Htclient
          • Typed-lists-light
          • Arithmetic-light
        • X86isa
        • Axe
        • Execloader
      • Math
      • Testing-utilities
    • Integer-format-templates

    Sinteger-format->min

    The ACL2 integer value of the minimum value representable in a signed integer format.

    Signature
    (sinteger-format->min format) → min
    Arguments
    format — Guard (sinteger-formatp format).
    Returns
    min — Type (integerp min).

    This is determined by the number M of value bits, the signed format, and possibly the trap representations [C17:6.2.6.2/2]. If the signed format is either sign and magnitude or ones' complement, the minimum value is the negation of the maximum value, i.e. - (2^M - 1). If the signed format is two's complement, there are two possibilities: if the representation with sign bit 1 and all value bits 0 is a trap representation, the minimum value is - (2^M - 1); otherwise, it is - 2^M. As explained in sinteger-format, currently we do not have a detailed model of trap representations; as a placeholder, for now we regard that representation to be a trap one iff the traps component of sinteger-format is not nil.

    Since M <= T - 1, where T is the total number of bits, and where the 1 accounts for the sign bit, the minimum value cannot be below - 2^(T-1).

    Definitions and Theorems

    Function: sinteger-format->min

    (defun sinteger-format->min (format)
      (declare (xargs :guard (sinteger-formatp format)))
      (if
       (and (equal (signed-format-kind (sinteger-format->signed format))
                   :twos-complement)
            (not (sinteger-format->traps format)))
       (- (expt 2
                (sinteger-bit-roles-value-count
                     (sinteger-format->bits format))))
       (- (1- (expt 2
                    (sinteger-bit-roles-value-count
                         (sinteger-format->bits format)))))))

    Theorem: integerp-of-sinteger-format->min

    (defthm integerp-of-sinteger-format->min
      (b* ((min (sinteger-format->min format)))
        (integerp min))
      :rule-classes :rewrite)

    Theorem: sinteger-format->min-type-prescription

    (defthm sinteger-format->min-type-prescription
      (b* ((common-lisp::?min (sinteger-format->min format)))
        (and (integerp min) (< min 0)))
      :rule-classes :type-prescription)

    Theorem: sinteger-format->min-lower-bound

    (defthm sinteger-format->min-lower-bound
      (b* ((common-lisp::?min (sinteger-format->min format)))
        (>= min
            (- (expt 2
                     (1- (len (sinteger-format->bits format)))))))
      :rule-classes :linear)

    Theorem: sinteger-format->min-of-sinteger-format-fix-format

    (defthm sinteger-format->min-of-sinteger-format-fix-format
      (equal (sinteger-format->min (sinteger-format-fix format))
             (sinteger-format->min format)))

    Theorem: sinteger-format->min-sinteger-format-equiv-congruence-on-format

    (defthm
        sinteger-format->min-sinteger-format-equiv-congruence-on-format
      (implies (sinteger-format-equiv format format-equiv)
               (equal (sinteger-format->min format)
                      (sinteger-format->min format-equiv)))
      :rule-classes :congruence)