ref: 4e18c3987bb3f2940be37db6ccb3d7986902e3a1
dir: /posix/mp.h/
#pragma once #ifdef BITS64 typedef uint64_t mpdigit; #else typedef uint32_t mpdigit; #endif typedef union FPdbleword FPdbleword; union FPdbleword { double x; struct { #if BYTE_ORDER == LITTLE_ENDIAN uint32_t lo; uint32_t hi; #else uint32_t hi; uint32_t lo; #endif }; }; #define mpdighi (((mpdigit)1)<<(Dbits-1)) #define DIGITS(x) ((int)(x) >= -(Dbits-1) ? ((Dbits - 1 + (x))/Dbits) : 0) extern int dec16(uint8_t*, int, char*, int); extern int enc16(char*, int, uint8_t*, int); extern mpdigit dec16chr(int); extern int enc16chr(int); /* * the code assumes mpdigit to be at least an int * mpdigit must be an atomic type. mpdigit is defined * in the architecture specific u.h */ typedef struct mpint mpint; struct mpint { mpdigit *p; uint32_t size; /* allocated digits */ uint32_t top; /* significant digits */ int sign; /* +1 or -1 */ int flags; }; enum { MPstatic= 0x01, /* static constant */ MPnorm= 0x02, /* normalization status */ MPtimesafe= 0x04, /* request time invariant computation */ MPfield= 0x08, /* this mpint is a field modulus */ Dbytes= sizeof(mpdigit), /* bytes per digit */ Dbits= Dbytes*8 /* bits per digit */ }; /* allocation */ void mpsetminbits(int n); /* newly created mpint's get at least n bits */ mpint* mpnew(int n); /* create a new mpint with at least n bits */ void mpfree(mpint *b); void mpbits(mpint *b, int n); /* ensure that b has at least n bits */ mpint* mpnorm(mpint *b); /* dump leading zeros */ mpint* mpcopy(mpint *b); void mpassign(mpint *old, mpint *new); /* random bits */ mpint* mprand(int bits, void (*gen)(uint8_t*, int), mpint *b); /* return uniform random [0..n-1] */ mpint* mpnrand(mpint *n, void (*gen)(uint8_t*, int), mpint *b); /* conversion */ mpint* strtomp(const char*, char**, int, mpint*); /* ascii */ char* mptoa(mpint*, int, char*, int); mpint* letomp(uint8_t*, uint32_t, mpint*); /* byte array, little-endian */ int mptole(mpint*, uint8_t*, uint32_t, uint8_t**); void mptolel(mpint *b, uint8_t *p, int n); mpint* betomp(uint8_t*, uint32_t, mpint*); /* byte array, big-endian */ int mptobe(mpint*, uint8_t*, uint32_t, uint8_t**); void mptober(mpint *b, uint8_t *p, int n); uint32_t mptoui(mpint*); /* unsigned int */ mpint* uitomp(uint32_t, mpint*); int mptoi(mpint*); /* int */ mpint* itomp(int, mpint*); uint64_t mptouv(mpint*); /* unsigned int64_t */ mpint* uvtomp(uint64_t, mpint*); int64_t mptov(mpint*); /* int64_t */ mpint* vtomp(int64_t, mpint*); double mptod(mpint*); /* double */ mpint* dtomp(double, mpint*); /* divide 2 digits by one */ void mpdigdiv(mpdigit *dividend, mpdigit divisor, mpdigit *quotient); /* in the following, the result mpint may be */ /* the same as one of the inputs. */ void mpadd(mpint *b1, mpint *b2, mpint *sum); /* sum = b1+b2 */ void mpsub(mpint *b1, mpint *b2, mpint *diff); /* diff = b1-b2 */ void mpleft(mpint *b, int shift, mpint *res); /* res = b<<shift */ void mpright(mpint *b, int shift, mpint *res); /* res = b>>shift */ void mpmul(mpint *b1, mpint *b2, mpint *prod); /* prod = b1*b2 */ void mpexp(mpint *b, mpint *e, mpint *m, mpint *res); /* res = b**e mod m */ void mpmod(mpint *b, mpint *m, mpint *remainder); /* remainder = b mod m */ /* logical operations */ void mpand(mpint *b1, mpint *b2, mpint *res); void mpbic(mpint *b1, mpint *b2, mpint *res); void mpor(mpint *b1, mpint *b2, mpint *res); void mpnot(mpint *b, mpint *res); void mpxor(mpint *b1, mpint *b2, mpint *res); void mpasr(mpint *b, int shift, mpint *res); /* modular arithmetic, time invariant when 0≤b1≤m-1 and 0≤b2≤m-1 */ void mpmodadd(mpint *b1, mpint *b2, mpint *m, mpint *sum); /* sum = b1+b2 % m */ void mpmodsub(mpint *b1, mpint *b2, mpint *m, mpint *diff); /* diff = b1-b2 % m */ void mpmodmul(mpint *b1, mpint *b2, mpint *m, mpint *prod); /* prod = b1*b2 % m */ /* quotient = dividend/divisor, remainder = dividend % divisor */ void mpdiv(mpint *dividend, mpint *divisor, mpint *quotient, mpint *remainder); /* return neg, 0, pos as b1-b2 is neg, 0, pos */ int mpcmp(mpint *b1, mpint *b2); /* res = s != 0 ? b1 : b2 */ void mpsel(int s, mpint *b1, mpint *b2, mpint *res); /* extended gcd return d, x, and y, s.t. d = gcd(a,b) and ax+by = d */ void mpextendedgcd(mpint *a, mpint *b, mpint *d, mpint *x, mpint *y); /* res = b**-1 mod m */ void mpinvert(mpint *b, mpint *m, mpint *res); /* bit counting */ uint32_t mpsignif(mpint*); /* number of sigificant bits in mantissa */ uint32_t mplowbits0(mpint*); /* k, where n = 2**k * q for odd q */ /* well known constants */ extern mpint *mpzero, *mpone, *mptwo; /* sum[0:alen] = a[0:alen-1] + b[0:blen-1] */ /* prereq: alen >= blen, sum has room for alen+1 digits */ void mpvecadd(mpdigit *a, int alen, mpdigit *b, int blen, mpdigit *sum); /* diff[0:alen-1] = a[0:alen-1] - b[0:blen-1] */ /* prereq: alen >= blen, diff has room for alen digits */ void mpvecsub(mpdigit *a, int alen, mpdigit *b, int blen, mpdigit *diff); /* p[0:n] += m * b[0:n-1] */ /* prereq: p has room for n+1 digits */ void mpvecdigmuladd(mpdigit *b, int n, mpdigit m, mpdigit *p); /* p[0:n] -= m * b[0:n-1] */ /* prereq: p has room for n+1 digits */ int mpvecdigmulsub(mpdigit *b, int n, mpdigit m, mpdigit *p); /* p[0:alen+blen-1] = a[0:alen-1] * b[0:blen-1] */ /* prereq: alen >= blen, p has room for m*n digits */ void mpvecmul(mpdigit *a, int alen, mpdigit *b, int blen, mpdigit *p); void mpvectsmul(mpdigit *a, int alen, mpdigit *b, int blen, mpdigit *p); /* sign of a - b or zero if the same */ int mpveccmp(mpdigit *a, int alen, mpdigit *b, int blen); int mpvectscmp(mpdigit *a, int alen, mpdigit *b, int blen); /* divide the 2 digit dividend by the one digit divisor and stick in quotient */ /* we assume that the result is one digit - overflow is all 1's */ void mpdigdiv(mpdigit *dividend, mpdigit divisor, mpdigit *quotient); /* playing with magnitudes */ int mpmagcmp(mpint *b1, mpint *b2); void mpmagadd(mpint *b1, mpint *b2, mpint *sum); /* sum = b1+b2 */ void mpmagsub(mpint *b1, mpint *b2, mpint *sum); /* sum = b1+b2 */ /* chinese remainder theorem */ typedef struct CRTpre CRTpre; /* precomputed values for converting */ /* twixt residues and mpint */ typedef struct CRTres CRTres; /* residue form of an mpint */ struct CRTres { int n; /* number of residues */ mpint *r[1]; /* residues */ }; CRTpre* crtpre(int, mpint**); /* precompute conversion values */ CRTres* crtin(CRTpre*, mpint*); /* convert mpint to residues */ void crtout(CRTpre*, CRTres*, mpint*); /* convert residues to mpint */ void crtprefree(CRTpre*); void crtresfree(CRTres*); /* fast field arithmetic */ typedef struct Mfield Mfield; struct Mfield { mpint m; int (*reduce)(Mfield*, mpint*, mpint*); }; mpint *mpfield(mpint*); Mfield *gmfield(mpint*); Mfield *cnfield(mpint*);