Bug Summary

File:internal/floatscan.c
Location:line 158, column 3
Description:Value stored to 'j' is never read

Annotated Source Code

1#include <stdint.h>
2#include <stdio.h>
3#include <math.h>
4#include <float.h>
5#include <limits.h>
6#include <errno(*__errno_location()).h>
7#include <ctype.h>
8
9#include "shgetc.h"
10#include "floatscan.h"
11
12#if LDBL_MANT_DIG64 == 53 && LDBL_MAX_EXP16384 == 1024
13
14#define LD_B1B_DIG3 2
15#define LD_B1B_MAX18, 446744073, 709551615 9007199, 254740991
16#define KMAX2048 128
17
18#elif LDBL_MANT_DIG64 == 64 && LDBL_MAX_EXP16384 == 16384
19
20#define LD_B1B_DIG3 3
21#define LD_B1B_MAX18, 446744073, 709551615 18, 446744073, 709551615
22#define KMAX2048 2048
23
24#elif LDBL_MANT_DIG64 == 113 && LDBL_MAX_EXP16384 == 16384
25
26#define LD_B1B_DIG3 4
27#define LD_B1B_MAX18, 446744073, 709551615 10384593, 717069655, 257060992, 658440191
28#define KMAX2048 2048
29
30#else
31#error Unsupported long double representation
32#endif
33
34#define MASK(2048 -1) (KMAX2048-1)
35
36#define CONCAT2(x,y)xy x ## y
37#define CONCAT(x,y)xy CONCAT2(x,y)xy
38
39static long long scanexp(FILE *f, int pok)
40{
41 int c;
42 int x;
43 long long y;
44 int neg = 0;
45
46 c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f));
47 if (c=='+' || c=='-') {
48 neg = (c=='-');
49 c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f));
50 if (c-'0'>=10U && pok) shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
51 }
52 if (c-'0'>=10U) {
53 shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
54 return LLONG_MIN(-0x7fffffffffffffffLL -1);
55 }
56 for (x=0; c-'0'<10U && x<INT_MAX0x7fffffff/10; c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f)))
57 x = 10*x + c-'0';
58 for (y=x; c-'0'<10U && y<LLONG_MAX0x7fffffffffffffffLL/100; c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f)))
59 y = 10*y + c-'0';
60 for (; c-'0'<10U; c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f)));
61 shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
62 return neg ? -y : y;
63}
64
65
66static long double decfloat(FILE *f, int c, int bits, int emin, int sign, int pok)
67{
68 uint32_t x[KMAX2048];
69 static const uint32_t th[] = { LD_B1B_MAX18, 446744073, 709551615 };
70 int i, j, k, a, z;
71 long long lrp=0, dc=0;
72 long long e10=0;
73 int lnz = 0;
74 int gotdig = 0, gotrad = 0;
75 int rp;
76 int e2;
77 int emax = -emin-bits+3;
78 int denormal = 0;
79 long double y;
80 long double frac=0;
81 long double bias=0;
82 static const int p10s[] = { 10, 100, 1000, 10000,
83 100000, 1000000, 10000000, 100000000 };
84
85 j=0;
86 k=0;
87
88 /* Don't let leading zeros consume buffer space */
89 for (; c=='0'; c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f))) gotdig=1;
90 if (c=='.') {
91 gotrad = 1;
92 for (c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f)); c=='0'; c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f))) gotdig=1, lrp--;
93 }
94
95 x[0] = 0;
96 for (; c-'0'<10U || c=='.'; c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f))) {
97 if (c == '.') {
98 if (gotrad) break;
99 gotrad = 1;
100 lrp = dc;
101 } else if (k < KMAX2048-3) {
102 dc++;
103 if (c!='0') lnz = dc;
104 if (j) x[k] = x[k]*10 + c-'0';
105 else x[k] = c-'0';
106 if (++j==9) {
107 k++;
108 j=0;
109 }
110 gotdig=1;
111 } else {
112 dc++;
113 if (c!='0') x[KMAX2048-4] |= 1;
114 }
115 }
116 if (!gotrad) lrp=dc;
117
118 if (gotdig && (c|32)=='e') {
119 e10 = scanexp(f, pok);
120 if (e10 == LLONG_MIN(-0x7fffffffffffffffLL -1)) {
121 if (pok) {
122 shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
123 } else {
124 shlim(f, 0)__shlim((f), (0));
125 return 0;
126 }
127 e10 = 0;
128 }
129 lrp += e10;
130 } else if (c>=0) {
131 shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
132 }
133 if (!gotdig) {
134 errno(*__errno_location()) = EINVAL22;
135 shlim(f, 0)__shlim((f), (0));
136 return 0;
137 }
138
139 /* Handle zero specially to avoid nasty special cases later */
140 if (!x[0]) return sign * 0.0;
141
142 /* Optimize small integers (w/no exponent) and over/under-flow */
143 if (lrp==dc && dc<10 && (bits>30 || x[0]>>bits==0))
144 return sign * (long double)x[0];
145 if (lrp > -emin/2) {
146 errno(*__errno_location()) = ERANGE34;
147 return sign * LDBL_MAX1.1897314953572317650e+4932L * LDBL_MAX1.1897314953572317650e+4932L;
148 }
149 if (lrp < emin-2*LDBL_MANT_DIG64) {
150 errno(*__errno_location()) = ERANGE34;
151 return sign * LDBL_MIN3.3621031431120935063e-4932L * LDBL_MIN3.3621031431120935063e-4932L;
152 }
153
154 /* Align incomplete final B1B digit */
155 if (j) {
156 for (; j<9; j++) x[k]*=10;
157 k++;
158 j=0;
Value stored to 'j' is never read
159 }
160
161 a = 0;
162 z = k;
163 e2 = 0;
164 rp = lrp;
165
166 /* Optimize small to mid-size integers (even in exp. notation) */
167 if (lnz<9 && lnz<=rp && rp < 18) {
168 if (rp == 9) return sign * (long double)x[0];
169 if (rp < 9) return sign * (long double)x[0] / p10s[8-rp];
170 int bitlim = bits-3*(int)(rp-9);
171 if (bitlim>30 || x[0]>>bitlim==0)
172 return sign * (long double)x[0] * p10s[rp-10];
173 }
174
175 /* Align radix point to B1B digit boundary */
176 if (rp % 9) {
177 int rpm9 = rp>=0 ? rp%9 : rp%9+9;
178 int p10 = p10s[8-rpm9];
179 uint32_t carry = 0;
180 for (k=a; k!=z; k++) {
181 uint32_t tmp = x[k] % p10;
182 x[k] = x[k]/p10 + carry;
183 carry = 1000000000/p10 * tmp;
184 if (k==a && !x[k]) {
185 a = (a+1 & MASK(2048 -1));
186 rp -= 9;
187 }
188 }
189 if (carry) x[z++] = carry;
190 rp += 9-rpm9;
191 }
192
193 /* Upscale until desired number of bits are left of radix point */
194 while (rp < 9*LD_B1B_DIG3 || (rp == 9*LD_B1B_DIG3 && x[a]<th[0])) {
195 uint32_t carry = 0;
196 e2 -= 29;
197 for (k=(z-1 & MASK(2048 -1)); ; k=(k-1 & MASK(2048 -1))) {
198 uint64_t tmp = ((uint64_t)x[k] << 29) + carry;
199 if (tmp > 1000000000) {
200 carry = tmp / 1000000000;
201 x[k] = tmp % 1000000000;
202 } else {
203 carry = 0;
204 x[k] = tmp;
205 }
206 if (k==(z-1 & MASK(2048 -1)) && k!=a && !x[k]) z = k;
207 if (k==a) break;
208 }
209 if (carry) {
210 rp += 9;
211 a = (a-1 & MASK(2048 -1));
212 if (a == z) {
213 z = (z-1 & MASK(2048 -1));
214 x[z-1 & MASK(2048 -1)] |= x[z];
215 }
216 x[a] = carry;
217 }
218 }
219
220 /* Downscale until exactly number of bits are left of radix point */
221 for (;;) {
222 uint32_t carry = 0;
223 int sh = 1;
224 for (i=0; i<LD_B1B_DIG3; i++) {
225 k = (a+i & MASK(2048 -1));
226 if (k == z || x[k] < th[i]) {
227 i=LD_B1B_DIG3;
228 break;
229 }
230 if (x[a+i & MASK(2048 -1)] > th[i]) break;
231 }
232 if (i==LD_B1B_DIG3 && rp==9*LD_B1B_DIG3) break;
233 /* FIXME: find a way to compute optimal sh */
234 if (rp > 9+9*LD_B1B_DIG3) sh = 9;
235 e2 += sh;
236 for (k=a; k!=z; k=(k+1 & MASK(2048 -1))) {
237 uint32_t tmp = x[k] & (1<<sh)-1;
238 x[k] = (x[k]>>sh) + carry;
239 carry = (1000000000>>sh) * tmp;
240 if (k==a && !x[k]) {
241 a = (a+1 & MASK(2048 -1));
242 i--;
243 rp -= 9;
244 }
245 }
246 if (carry) {
247 if ((z+1 & MASK(2048 -1)) != a) {
248 x[z] = carry;
249 z = (z+1 & MASK(2048 -1));
250 } else x[z-1 & MASK(2048 -1)] |= 1;
251 }
252 }
253
254 /* Assemble desired bits into floating point variable */
255 for (y=i=0; i<LD_B1B_DIG3; i++) {
256 if ((a+i & MASK(2048 -1))==z) x[(z=(z+1 & MASK(2048 -1)))-1] = 0;
257 y = 1000000000.0L * y + x[a+i & MASK(2048 -1)];
258 }
259
260 y *= sign;
261
262 /* Limit precision for denormal results */
263 if (bits > LDBL_MANT_DIG64+e2-emin) {
264 bits = LDBL_MANT_DIG64+e2-emin;
265 if (bits<0) bits=0;
266 denormal = 1;
267 }
268
269 /* Calculate bias term to force rounding, move out lower bits */
270 if (bits < LDBL_MANT_DIG64) {
271 bias = copysignl(scalbn(1, 2*LDBL_MANT_DIG64-bits-1), y);
272 frac = fmodl(y, scalbn(1, LDBL_MANT_DIG64-bits));
273 y -= frac;
274 y += bias;
275 }
276
277 /* Process tail of decimal input so it can affect rounding */
278 if ((a+i & MASK(2048 -1)) != z) {
279 uint32_t t = x[a+i & MASK(2048 -1)];
280 if (t < 500000000 && (t || (a+i+1 & MASK(2048 -1)) != z))
281 frac += 0.25*sign;
282 else if (t > 500000000)
283 frac += 0.75*sign;
284 else if (t == 500000000) {
285 if ((a+i+1 & MASK(2048 -1)) == z)
286 frac += 0.5*sign;
287 else
288 frac += 0.75*sign;
289 }
290 if (LDBL_MANT_DIG64-bits >= 2 && !fmodl(frac, 1))
291 frac++;
292 }
293
294 y += frac;
295 y -= bias;
296
297 if ((e2+LDBL_MANT_DIG64 & INT_MAX0x7fffffff) > emax-5) {
298 if (fabs(y) >= CONCAT(0x1p, LDBL_MANT_DIG)0x1p64) {
299 if (denormal && bits==LDBL_MANT_DIG64+e2-emin)
300 denormal = 0;
301 y *= 0.5;
302 e2++;
303 }
304 if (e2+LDBL_MANT_DIG64>emax || (denormal && frac))
305 errno(*__errno_location()) = ERANGE34;
306 }
307
308 return scalbnl(y, e2);
309}
310
311static long double hexfloat(FILE *f, int bits, int emin, int sign, int pok)
312{
313 uint32_t x = 0;
314 long double y = 0;
315 long double scale = 1;
316 long double bias = 0;
317 int gottail = 0, gotrad = 0, gotdig = 0;
318 long long rp = 0;
319 long long dc = 0;
320 long long e2 = 0;
321 int d;
322 int c;
323
324 c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f));
325
326 /* Skip leading zeros */
327 for (; c=='0'; c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f))) gotdig = 1;
328
329 if (c=='.') {
330 gotrad = 1;
331 c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f));
332 /* Count zeros after the radix point before significand */
333 for (rp=0; c=='0'; c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f)), rp--) gotdig = 1;
334 }
335
336 for (; c-'0'<10U || (c|32)-'a'<6U || c=='.'; c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f))) {
337 if (c=='.') {
338 if (gotrad) break;
339 rp = dc;
340 gotrad = 1;
341 } else {
342 gotdig = 1;
343 if (c > '9') d = (c|32)+10-'a';
344 else d = c-'0';
345 if (dc<8) {
346 x = x*16 + d;
347 } else if (dc < LDBL_MANT_DIG64/4+1) {
348 y += d*(scale/=16);
349 } else if (d && !gottail) {
350 y += 0.5*scale;
351 gottail = 1;
352 }
353 dc++;
354 }
355 }
356 if (!gotdig) {
357 shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
358 if (pok) {
359 shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
360 if (gotrad) shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
361 } else {
362 shlim(f, 0)__shlim((f), (0));
363 }
364 return sign * 0.0;
365 }
366 if (!gotrad) rp = dc;
367 while (dc<8) x *= 16, dc++;
368 if ((c|32)=='p') {
369 e2 = scanexp(f, pok);
370 if (e2 == LLONG_MIN(-0x7fffffffffffffffLL -1)) {
371 if (pok) {
372 shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
373 } else {
374 shlim(f, 0)__shlim((f), (0));
375 return 0;
376 }
377 e2 = 0;
378 }
379 } else {
380 shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
381 }
382 e2 += 4*rp - 32;
383
384 if (!x) return sign * 0.0;
385 if (e2 > -emin) {
386 errno(*__errno_location()) = ERANGE34;
387 return sign * LDBL_MAX1.1897314953572317650e+4932L * LDBL_MAX1.1897314953572317650e+4932L;
388 }
389 if (e2 < emin-2*LDBL_MANT_DIG64) {
390 errno(*__errno_location()) = ERANGE34;
391 return sign * LDBL_MIN3.3621031431120935063e-4932L * LDBL_MIN3.3621031431120935063e-4932L;
392 }
393
394 while (x < 0x80000000) {
395 if (y>=0.5) {
396 x += x + 1;
397 y += y - 1;
398 } else {
399 x += x;
400 y += y;
401 }
402 e2--;
403 }
404
405 if (bits > 32+e2-emin) {
406 bits = 32+e2-emin;
407 if (bits<0) bits=0;
408 }
409
410 if (bits < LDBL_MANT_DIG64)
411 bias = copysignl(scalbn(1, 32+LDBL_MANT_DIG64-bits-1), sign);
412
413 if (bits<32 && y && !(x&1)) x++, y=0;
414
415 y = bias + sign*(long double)x + sign*y;
416 y -= bias;
417
418 if (!y) errno(*__errno_location()) = ERANGE34;
419
420 return scalbnl(y, e2);
421}
422
423long double __floatscan(FILE *f, int prec, int pok)
424{
425 int sign = 1;
426 size_t i;
427 int bits;
428 int emin;
429 int c;
430
431 switch (prec) {
432 case 0:
433 bits = FLT_MANT_DIG24;
434 emin = FLT_MIN_EXP(-125)-bits;
435 break;
436 case 1:
437 bits = DBL_MANT_DIG53;
438 emin = DBL_MIN_EXP(-1021)-bits;
439 break;
440 case 2:
441 bits = LDBL_MANT_DIG64;
442 emin = LDBL_MIN_EXP(-16381)-bits;
443 break;
444 default:
445 return 0;
446 }
447
448 while (isspace((c=shgetc(f)))__isspace((c=(((f)->rpos < (f)->shend) ? *(f)->rpos
++ : __shgetc(f)))));
449
450 if (c=='+' || c=='-') {
451 sign -= 2*(c=='-');
452 c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f));
453 }
454
455 for (i=0; i<8 && (c|32)=="infinity"[i]; i++)
456 if (i<7) c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f));
457 if (i==3 || i==8 || (i>3 && pok)) {
458 if (i!=8) {
459 shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
460 if (pok) for (; i>3; i--) shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
461 }
462 return sign * INFINITY__builtin_inff();
463 }
464 if (!i) for (i=0; i<3 && (c|32)=="nan"[i]; i++)
465 if (i<2) c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f));
466 if (i==3) {
467 if (shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f)) != '(') {
468 shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
469 return NAN__builtin_nanf("");
470 }
471 for (i=1; ; i++) {
472 c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f));
473 if (c-'0'<10U || c-'A'<26U || c-'a'<26U || c=='_')
474 continue;
475 if (c==')') return NAN__builtin_nanf("");
476 shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
477 if (!pok) {
478 errno(*__errno_location()) = EINVAL22;
479 shlim(f, 0)__shlim((f), (0));
480 return 0;
481 }
482 while (i--) shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
483 return NAN__builtin_nanf("");
484 }
485 return NAN__builtin_nanf("");
486 }
487
488 if (i) {
489 shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
490 errno(*__errno_location()) = EINVAL22;
491 shlim(f, 0)__shlim((f), (0));
492 return 0;
493 }
494
495 if (c=='0') {
496 c = shgetc(f)(((f)->rpos < (f)->shend) ? *(f)->rpos++ : __shgetc
(f));
497 if ((c|32) == 'x')
498 return hexfloat(f, bits, emin, sign, pok);
499 shunget(f)((f)->shend ? (void)(f)->rpos-- : (void)0);
500 c = '0';
501 }
502
503 return decfloat(f, c, bits, emin, sign, pok);
504}