Home
last modified time | relevance | path

Searched refs:rounding (Results 1 – 25 of 85) sorted by relevance

1234

/trueos/lib/libc/softfloat/
HDtimesoftfloat.txt14 point routines. Each routine can be evaluated for every relevant rounding
15 mode, tininess mode, and/or rounding precision.
64 the four rounding modes, one after the other. If the rounding mode is not
66 some operations do not require rounding--only the nearest/even rounding mode
69 tininess detected before rounding and after rounding. For extended double-
70 precision operations affected by rounding precision control, `timesoftfloat'
71 also times the function for all three rounding precision modes, one after
72 the other. Evaluation of a function can be limited to a single rounding
73 mode, a single tininess mode, and/or a single rounding precision with
109 For extended double-precision functions affected by rounding precision
[all …]
HDsoftfloat.txt136 All four rounding modes prescribed by the IEC/IEEE Standard are implemented
137 for all operations that require rounding. The rounding mode is selected
140 `float_round_down', or `float_round_up'. The rounding mode is initialized
147 For extended double precision (`floatx80') only, the rounding precision
158 (`float64'), respectively. When rounding to reduced precision, additional
159 bits in the result significand beyond the rounding point are set to zero.
184 for underflow either before or after rounding. The choice is made by
187 Detecting tininess after rounding is better because it results in fewer
220 format are always exact and so require no rounding. Conversions from 32-bit
232 integer value, the operand is rounded according to the current rounding
[all …]
HDsoftfloat-history.txt45 detected before or after rounding.
/trueos/tools/test/testfloat/
HDsystemBugs.txt54 Errors found in floatx80_to_int32, rounding nearest_even:
60 Errors found in floatx80_to_int32, rounding to_zero:
67 Errors found in floatx80_to_int32, rounding down:
74 Errors found in floatx80_to_int32, rounding up:
87 zero when the rounding mode is nearest/even or up.
89 Errors found in floatx80_to_int32, rounding nearest_even:
100 Errors found in floatx80_to_int32, rounding up:
126 Errors found in floatx80_to_int32, rounding nearest_even:
129 Errors found in floatx80_to_int32, rounding to_zero:
131 Errors found in floatx80_to_int32, rounding up:
[all …]
HDtestfloat.txt106 plus the rounding of adds, subtracts, multiplies, and simple operations like
113 TestFloat is not remarkably good at testing difficult rounding cases for
128 For each operation, TestFloat can test all four rounding modes required
135 For machines that implement extended double precision with rounding
138 rounding precisions. The rounding precision can be set equivalent to single
197 TestFloat will ordinarily test a function for all four rounding modes, one
198 after the other. If the rounding mode is not supposed to have any affect
199 on the results--for instance, some operations do not require rounding--only
200 the nearest/even rounding mode is checked. For extended double-precision
201 operations affected by rounding precision control, TestFloat also tests all
[all …]
HDtestfloat-source.txt203 function for all of the relevant rounding modes and rounding precisions.
218 <float>_to_int32 (rounded according to rounding mode)
219 <float>_to_int64 (rounded according to rounding mode)
241 controlling the system's floating-point, including the rounding mode and
242 the rounding precision for extended double precision.
330 <float>_to_int32 (rounded according to rounding mode)
331 <float>_to_int64 (rounded according to rounding mode)
HDtestfloat-history.txt49 whether tininess should be detected before or after rounding.
/trueos/contrib/gdtoa/
HDstrtorf.c69 strtorf(s, sp, rounding, f) CONST char *s; char **sp; int rounding; float *f; in strtorf() argument
71 strtorf(CONST char *s, char **sp, int rounding, float *f)
81 if (rounding != FPI_Round_near) {
83 fpi1.rounding = rounding;
HDstrtord.c73 strtord_l(s, sp, rounding, d, locale) CONST char *s; char **sp; int rounding; in strtord_l() argument
76 strtord_l(CONST char *s, char **sp, int rounding, double *d, locale_t locale)
86 if (rounding != FPI_Round_near) {
88 fpi1.rounding = rounding;
HDstrtorxL.c88 strtorxL(s, sp, rounding, L) CONST char *s; char **sp; int rounding; void *L; in strtorxL() argument
90 strtorxL(CONST char *s, char **sp, int rounding, void *L)
100 if (rounding != FPI_Round_near) {
102 fpi1.rounding = rounding;
HDstrtorQ.c106 strtorQ_l(s, sp, rounding, L, locale) CONST char *s; char **sp; int rounding; in strtorQ_l() argument
109 strtorQ_l(CONST char *s, char **sp, int rounding, void *L, locale_t locale)
119 if (rounding != FPI_Round_near) {
121 fpi1.rounding = rounding;
HDstrtorx.c109 strtorx_l(s, sp, rounding, L, locale) CONST char *s; char **sp; int rounding; in strtorx_l() argument
112 strtorx_l(CONST char *s, char **sp, int rounding, void *L, locale_t locale)
122 if (rounding != FPI_Round_near) {
124 fpi1.rounding = rounding;
HDstrtordd.c175 strtordd(s, sp, rounding, dd) CONST char *s; char **sp; int rounding; double *dd; in strtordd() argument
177 strtordd(CONST char *s, char **sp, int rounding, double *dd)
191 if (rounding != FPI_Round_near) {
193 fpi1.rounding = rounding;
HDREADME21 http://cm.bell-labs.com/cm/cs/what/ampl/REFS/rounding.ps.gz
59 helper routines: one for round-nearest (or the current rounding
72 one with rounding direction specified:
127 the bottom of the exponent range. To get correct rounding, this
139 arithmetic with the default rounding mode (round-to-nearest) and, on
141 (53-bit) rounding precision. If the machine uses (the equivalent of)
150 rounding precision on systems (such as Intel IA32 systems) that may
151 suffer double rounding due to use of extended-precision registers.
153 one in strtod.c when the latter is run with 53-bit rounding precision.
194 style and conversions with direction of rounding specified (so that, if
[all …]
HDgethex.c157 switch(fpi->rounding) {
174 switch(fpi->rounding) {
269 switch (fpi->rounding) {
312 switch(fpi->rounding) {
HDchanges3 (fixed-point mode); fix rounding bug in these modes when the input
5 i.e., when the result, before rounding, would be empty but might
18 rounding, would be empty, but rounds to one digit: *decpt was low by
119 that caused one-bit rounding errors in certain denormal numbers, such
287 testbase program is unhappy with this new rounding, as it can
322 loop with a boundary case of directed rounding. Example input to
347 IEEE arithmetic (with 53-bit rounding precision on 80x87 chips).
373 denormal result should be zero, correct logic for rounding up when the
388 but does not require 53-bit rounding precision on Intel IA32 systems.
428 and with hex input, get the current rounding mode from fegetround().
[all …]
HDgdtoa_fltrnds.h17 fpi1.rounding = Rounding;
HDprintf.c089 int rounding;
93 enum { /* FPI.rounding values: same as FLT_ROUNDS */
520 fpi1 = *fpb.fpi; fpi1.rounding = Rounding; fpb.fpi = &fpi1;}
527 fpi1 = *fpb.fpi; fpi1.rounding = Rounding; fpb.fpi = &fpi1;}
582 int rounding = fpi->rounding;
584 if (rounding > FPI_Round_near && b->sign)
585 rounding = FPI_Round_up + FPI_Round_down - rounding;
586 if (rounding == FPI_Round_down)
587 rounding = FPI_Round_zero;
593 switch(rounding) {
HDgdtoa.h91 int rounding; member
/trueos/contrib/gcclibs/libdecnumber/
HDdecRound.c42 extern enum rounding __decGetRound (void);
45 static enum rounding __dfp_rounding_mode = DEC_ROUND_HALF_EVEN;
98 enum rounding
HDdecContext.h60 enum rounding enum
77 enum rounding round; /* rounding mode */
/trueos/lib/libc/gdtoa/
HD_ldtoa.c74 fpi.rounding ^= (fpi.rounding >> 1) & u.bits.sign; in __ldtoa()
/trueos/contrib/groff/src/devices/grops/
HDTODO14 Consider whether we ought to do rounding of graphical objects other
/trueos/sys/mips/mips/
HDfp.S668 and v0, a1, MIPS_FPU_ROUNDING_BITS # get rounding mode
727 and v0, a1, MIPS_FPU_ROUNDING_BITS # get rounding mode
791 and v0, a1, MIPS_FPU_ROUNDING_BITS # get rounding mode
885 and v0, a1, MIPS_FPU_ROUNDING_BITS # get rounding mode
1507 and v0, a1, MIPS_FPU_ROUNDING_BITS # get rounding mode
1517 addu t2, t2, 1 # add rounding bit
1521 li v0, GUARDBIT # load guard bit for rounding
1730 and v0, a1, MIPS_FPU_ROUNDING_BITS # get rounding mode
1740 addu t2, t2, 1 # add rounding bit
1746 li v0, GUARDBIT # load guard bit for rounding
[all …]
/trueos/contrib/llvm/lib/Target/R600/
HDAMDGPUInstrInfo.td65 // e is rounding error

1234