matherr(3M)
matherr - error-handling function
As shipped in IRIX 6.5. First release of IRIX 6.5.
NAME matherr - error-handling function SYNOPSIS #include <math.h> int matherr (struct exception *x); DESCRIPTION matherr is invoked by the following functions in libmx.a when errors are detected: acos acosf acosh asin asinf asinh atan atanf atan2 atan2f atanh cbrt cos cosf cosh coshf drem erf erfc exp expf expm1 expm1f fmod fmodf gamma hypot hypotf j0 j1 jn lgamma log logf log10 log10f log1p log1pf pow powf sin sinf sinh sinhf sqrt sqrtf tan tanf tanh tanhf y0 y1 yn Note that to use matherr, libmx.a needs to be linked with the program. Users may define their own routines for handling errors by including a function named matherr in their programs. matherr must be of the form described above. When an error occurs, a pointer to the exception structure x will be passed to the user-supplied matherr function. This structure, which is defined in the math.h header file, is as follows: struct exception { int type; char *name; double arg1, arg2, retval; }; The element type is an integer describing the type of error that has occurred, from the following list of constants (defined in the header file math.h): DOMAIN argument domain error SING argument singularity OVERFLOW overflow range error UNDERFLOW underflow range error TLOSS total loss of significance PLOSS partial loss of significance The element name points to a string containing the name of the function that incurred the error. The elements arg1 and arg2 are the arguments with which the function was invoked. retval is set to the default value that will be returned by the function unless the user's matherr sets it to a different value. NaN argument(s) to any of the above functions generates a call to matherr with a domain error. If the user's matherr function returns non-zero, no error message will be printed by the calling function, and errno will not be set. If matherr is not supplied by the user, the default error-handling procedures, described by the table below, will be invoked upon error. In every case, errno is set to EDOM or ERANGE and the program continues. _______________________________________________________________________________ Default Error Handling Procedures ______________________________________________________________________________ Types of Errors ______________________________________________________________________________ type DOMAIN SING OVERFLOW UNDERFLOW TLOSS PLOSS ______________________________________________________________________________ errno EDOM EDOM ERANGE ERANGE ERANGE ERANGE ______________________________________________________________________________ EXP, EXPF: - - M, H M, 0 - - ______________________________________________________________________________ EXPM1, EXPM1F: - - M, H M, 0 - - ______________________________________________________________________________ LOG, LOGF: LOG10, LOG10F: (arg < 0) M, N - - - - - (arg = 0) M, -H - - - - - ______________________________________________________________________________ LOG1P, LOG1PF: (arg < -1) M, N - - - - - (arg = -1) M, -H - - - - - ______________________________________________________________________________ POW, POWF: - - _H 0 - - neg ** non-int M, 0 - - - - - 0 ** non-pos M, H - - - - - ______________________________________________________________________________ SQRT, SQRTF: M, N - - - - - ______________________________________________________________________________ DREM, FMOD, FMODF: (arg1 = +/-inf M, N - - - - - or arg2 = 0) ______________________________________________________________________________ GAMMA, LGAMMA: - M, H M, H - - - ______________________________________________________________________________ J0, J1, JN: - - - - M, 0 - ______________________________________________________________________________ Y0, Y1, YN: (arg = 0) - - M, -H - - - (arg < 0) M, N - - - - - (arg >= 2**49) - - - - M, 0 - ______________________________________________________________________________ SINH, SINHF: - - _H - - - ______________________________________________________________________________ COSH, COSHF: - - H - - - ______________________________________________________________________________ HYPOT, HYPOTF: - - M, H - - - ______________________________________________________________________________ ASIN, ASINF: ACOS, ACOSF: M, N - - - - - ______________________________________________________________________________ ______________________________________________________________________________ | | | | | | | | SIN, COS, TAN: (arg >= 2**50) M, N - - - - - (arg = +/-inf) - - - - M, 0 - ______________________________________________________________________________ SINF, COSF, TANF: (arg >= 2**22) M, N - - - - - (arg = +/-inf) - - - - M, 0 - ______________________________________________________________________________ ATAN2, ATAN2F: M, 0 - - - - - ______________________________________________________________________________ ACOSH: M, N - - - - - ______________________________________________________________________________ ATANH: (|arg| > 1) M, N - - - - - ______________________________________________________________________________ | | | | | | | | ______________________________________________________________________________ Abbreviations M Message is printed H HUGE is returned -H -HUGE is returned _H HUGE or -HUGE is returned. 0 0 is returned. X arg1 is returned. N NaN is returned. ______________________________________________________________________________ | | EXAMPLE /* ** Example showing the use of libmx.a's matherr() to handle math errors. ** ** % cc -xansi my_matherr.c -lmx ** % a.out ** overflow range error in logf ** logf(0.0) is -inf. ** ** expf failed: Argument out of domain ** Abort (core dumped) */ #include <stdio.h> #include <math.h> #include <stdlib.h> #include <string.h> #include <errno.h> main() { float ff; ff = logf(0.0f); /* matherr() leaves error reporting to libmx.a */ printf("logf(0.0f) is %f.\n\n",ff); fflush(stdout); ff = 0.0f; ff = 0.0f/ff; ff = expf(ff); /* matherr() customized to force a core dump */ } int matherr( register struct exception *x ) { char *sterr; switch (x->type) { case DOMAIN: /* if it's a NaN, abort */ if ( isnan(x->arg1) ) { sterr = strerror(EDOM); fprintf(stderr, "%s failed: %s\n", x->name, sterr); abort(); } else if ( !strcmp(x->name, "sin") ) { x->retval = 0.0; /* return 0.0 */ fprintf(stderr, "domain error in sin\n"); return (1); /* suppress libmx.a error message and setting of errno */ } else return (0); /* libmx.a prints error message and sets errno */ break; default: return (0); /* libmx.a prints error message and sets errno */ } } NOTES Errors detected by functions in libm.a can be detected by examining the return value and/or the global variable errno. libmx.a is provided as an alternative error handling scheme for users who want to identify specific kinds of errors determined by libm functions without extensive modification of their source code. No extra overhead is incurred by programs using libmx.a if no error conditions are encountered in the libmx routines.