xref: /original-bsd/lib/libc/vax/stdio.old/doprnt.c (revision 6c57d260)
1 /* @(#)doprnt.c	4.3 (Berkeley) 03/22/81 */
2 	# C library -- conversions
3 
4 .globl	__doprnt
5 .globl	__flsbuf
6 
7 #define vbit 1
8 #define flags r10
9 #define ndfnd 0
10 #define prec 1
11 #define zfill 2
12 #define minsgn 3
13 #define plssgn 4
14 #define numsgn 5
15 #define caps 6
16 #define blank 7
17 #define gflag 8
18 #define dpflag 9
19 #define width r9
20 #define ndigit r8
21 #define llafx r7
22 #define lrafx r6
23 #define fdesc -4(fp)
24 #define exp -8(fp)
25 #define sexp -12(fp)
26 #define nchar -16(fp)
27 #define sign -17(fp)
28 	.set ch.zer,'0			# cpp doesn't like single appostrophes
29 
30 	.align 2
31 strtab:		# translate table for detecting null and percent
32 	.byte	0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15
33 	.byte	16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31
34 	.byte	' ,'!,'",'#,'$, 0,'&,'','(,'),'*,'+,',,'-,'.,'/
35 	.byte	'0,'1,'2,'3,'4,'5,'6,'7,'8,'9,':,';,'<,'=,'>,'?
36 	.byte	'@,'A,'B,'C,'D,'E,'F,'G,'H,'I,'J,'K,'L,'M,'N,'O
37 	.byte	'P,'Q,'R,'S,'T,'U,'V,'W,'X,'Y,'Z,'[,'\,'],'^,'_
38 	.byte	'`,'a,'b,'c,'d,'e,'f,'g,'h,'i,'j,'k,'l,'m,'n,'o
39 	.byte	'p,'q,'r,'s,'t,'u,'v,'w,'x,'y,'z,'{,'|,'},'~,127
40 	.byte	128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143
41 	.byte	144,145,146,147,148,149,150,151,152,153,154,155,156,157,158,159
42 	.byte	160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175
43 	.byte	176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191
44 	.byte	192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207
45 	.byte	208,209,210,211,212,213,214,215,216,217,218,219,220,221,222,223
46 	.byte	224,225,226,227,228,229,230,231,232,233,234,235,236,237,238,239
47 	.byte	240,241,242,243,244,245,246,247,248,249,250,251,252,253,254,255
48 
49 strfoo:
50 	clrl r4					# fix interrupt race
51 	jbr strok				# and try again
52 strmore:
53 	movzbl (r1)+,r2			# one char
54 	tstb strtab[r2]			# translate
55 	jeql stresc2			# bad guy in disguise (outbuf is full)
56 strout2:		# enter here to force out r2; r0,r1 must be set
57 	pushr $3				# save input descriptor
58 	pushl fdesc				# FILE
59 	pushl r2				# the char
60 	calls $2,__flsbuf		# please empty the buffer and handle 1 char
61 	tstl r0					# successful?
62 	jgeq strm1				# yes
63 	jbcs $31,nchar,strm1	# turn on sign bit of nchar to signify error
64 strm1:
65 	incl nchar				# count the char
66 	popr $3					# get input descriptor back
67 strout:			# enter via bsb with (r0,r1)=input descriptor
68 	movab strtab,r3			# table address
69 	movq *fdesc,r4			# output descriptor
70 	jbs $31,r4,strfoo		# negative count is a no no
71 strok:
72 	addl2 r0,nchar			# we intend to move this many chars
73 /******* Start bogus movtuc workaround  *****/
74 	clrl r2
75 	tstl    r0
76 	bleq    movdon
77 movlp:
78 	tstl    r4
79 	bleq    movdon
80 	movzbl  (r1)+,r3
81 	tstb    strtab[r3]
82 	bneq    1f
83 	mnegl   $1,r2
84 	decl    r1
85 	brb     movdon
86 1:
87 	movb    r3,(r5)+
88 	decl    r4
89 	sobgtr  r0,movlp
90   /******* End bogus movtuc workaround ***
91 	movtuc r0,(r1),$0,(r3),r4,(r5)
92 	movpsl r2                       /*  squirrel away condition codes */
93   /******* End equally bogus movtuc ****/
94 movdon: movq r4,*fdesc                  /*  update output descriptor */
95 	subl2 r0,nchar			# some chars not moved
96 	jbs $vbit,r2,stresc		# terminated by escape?
97 	sobgeq r0,strmore		# no; but out buffer might be full
98 stresc:
99 	rsb
100 stresc2:
101 	incl r0					# fix the length
102 	decl r1					# and the addr
103 	movl $1<vbit,r2			# fake condition codes
104 	rsb
105 
106 errdone:
107 	jbcs $31,nchar,prdone	# set error bit
108 prdone:
109 	movl nchar,r0
110 	ret
111 
112 	.align	1
113 __doprnt:
114 	.word	0xfc0			# uses r11-r6
115 	movab -256(sp),sp		# work space
116 	movl 4(ap),r11			# addr of format string
117 	movl 12(ap),fdesc		# output FILE ptr
118 	movl 8(ap),ap			# addr of first arg
119 	clrl nchar				# number of chars transferred
120 loop:
121 	movzwl $65535,r0		# pseudo length
122 	movl r11,r1				# fmt addr
123 		# comet sucks.
124 	movq *fdesc,r4
125 	subl3 r1,r5,r2
126 	jlss lp1
127 	cmpl r0,r2
128 	jleq lp1
129 	movl r2,r0
130 lp1:
131 		#
132 	bsbw strout				# copy to output, stop at null or percent
133 	movl r1,r11				# new fmt
134 	jbc $vbit,r2,loop		# if no escape, then very long fmt
135 	tstb (r11)+				# escape; null or percent?
136 	jeql prdone				# null means end of fmt
137 
138 	movl sp,r5			# reset output buffer pointer
139 	clrq r9				# width; flags
140 	clrq r6				# lrafx,llafx
141 longorunsg:				# we can ignore both of these distinctions
142 short:
143 L4a:
144 	movzbl (r11)+,r0		# so capital letters can tail merge
145 L4:	caseb r0,$' ,$'x-' 		# format char
146 L5:
147 	.word space-L5			# space
148 	.word fmtbad-L5			# !
149 	.word fmtbad-L5			# "
150 	.word sharp-L5			# #
151 	.word fmtbad-L5			# $
152 	.word fmtbad-L5			# %
153 	.word fmtbad-L5			# &
154 	.word fmtbad-L5			# '
155 	.word fmtbad-L5			# (
156 	.word fmtbad-L5			# )
157 	.word indir-L5			# *
158 	.word plus-L5			# +
159 	.word fmtbad-L5			# ,
160 	.word minus-L5			# -
161 	.word dot-L5			# .
162 	.word fmtbad-L5			# /
163 	.word gnum0-L5			# 0
164 	.word gnum-L5			# 1
165 	.word gnum-L5			# 2
166 	.word gnum-L5			# 3
167 	.word gnum-L5			# 4
168 	.word gnum-L5			# 5
169 	.word gnum-L5			# 6
170 	.word gnum-L5			# 7
171 	.word gnum-L5			# 8
172 	.word gnum-L5			# 9
173 	.word fmtbad-L5			# :
174 	.word fmtbad-L5			# ;
175 	.word fmtbad-L5			# <
176 	.word fmtbad-L5			# =
177 	.word fmtbad-L5			# >
178 	.word fmtbad-L5			# ?
179 	.word fmtbad-L5			# @
180 	.word fmtbad-L5			# A
181 	.word fmtbad-L5			# B
182 	.word fmtbad-L5			# C
183 	.word decimal-L5		# D
184 	.word capital-L5		# E
185 	.word fmtbad-L5			# F
186 	.word capital-L5		# G
187 	.word fmtbad-L5			# H
188 	.word fmtbad-L5			# I
189 	.word fmtbad-L5			# J
190 	.word fmtbad-L5			# K
191 	.word fmtbad-L5			# L
192 	.word fmtbad-L5			# M
193 	.word fmtbad-L5			# N
194 	.word octal-L5			# O
195 	.word fmtbad-L5			# P
196 	.word fmtbad-L5			# Q
197 	.word fmtbad-L5			# R
198 	.word fmtbad-L5			# S
199 	.word fmtbad-L5			# T
200 	.word unsigned-L5		# U
201 	.word fmtbad-L5			# V
202 	.word fmtbad-L5			# W
203 	.word hex-L5			# X
204 	.word fmtbad-L5			# Y
205 	.word fmtbad-L5			# Z
206 	.word fmtbad-L5			# [
207 	.word fmtbad-L5			# \
208 	.word fmtbad-L5			# ]
209 	.word fmtbad-L5			# ^
210 	.word fmtbad-L5			# _
211 	.word fmtbad-L5			# `
212 	.word fmtbad-L5			# a
213 	.word fmtbad-L5			# b
214 	.word charac-L5			# c
215 	.word decimal-L5		# d
216 	.word scien-L5			# e
217 	.word float-L5			# f
218 	.word general-L5		# g
219 	.word short-L5			# h
220 	.word fmtbad-L5			# i
221 	.word fmtbad-L5			# j
222 	.word fmtbad-L5			# k
223 	.word longorunsg-L5		# l
224 	.word fmtbad-L5			# m
225 	.word fmtbad-L5			# n
226 	.word octal-L5			# o
227 	.word fmtbad-L5			# p
228 	.word fmtbad-L5			# q
229 	.word fmtbad-L5			# r
230 	.word string-L5			# s
231 	.word fmtbad-L5			# t
232 	.word unsigned-L5		# u
233 	.word fmtbad-L5			# v
234 	.word fmtbad-L5			# w
235 	.word hex-L5			# x
236 fmtbad:
237 	movb r0,(r5)+			# print the unfound character
238 	jeql errdone			# dumb users who end the format with a %
239 	jbr prbuf
240 capital:
241 	bisl2 $1<caps,flags		# note that it was capitalized
242 	xorb2 $'a^'A,r0			# make it small
243 	jbr L4					# and try again
244 
245 string:
246 	movl ndigit,r0
247 	jbs $prec,flags,L20		# max length was specified
248 	mnegl $1,r0			# default max length
249 L20:	movl (ap)+,r2			# addr first byte
250 	locc $0,r0,(r2)			# find the zero at the end
251 	movl r1,r5			# addr last byte +1
252 	movl r2,r1			# addr first byte
253 	jbr prstr
254 
255 htab:	.byte	'0,'1,'2,'3,'4,'5,'6,'7,'8,'9,'a,'b,'c,'d,'e,'f
256 Htab:	.byte	'0,'1,'2,'3,'4,'5,'6,'7,'8,'9,'A,'B,'C,'D,'E,'F
257 
258 octal:
259 	movl $30,r2			# init position
260 	movl $3,r3			# field width
261 	movab htab,llafx	# translate table
262 	jbr L10
263 
264 hex:
265 	movl $28,r2			# init position
266 	movl $4,r3			# field width
267 	movab htab,llafx	# translate table
268 	jbc $caps,flags,L10
269 	movab Htab,llafx
270 L10:	mnegl r3,r6			# increment
271 	clrl r1
272 	addl2 $4,r5			# room for left affix (2) and slop [forced sign?]
273 	movl (ap)+,r0			# fetch arg
274 L11:	extzv r2,r3,r0,r1		# pull out a digit
275 	movb (llafx)[r1],(r5)+		# convert to character
276 L12:	acbl $0,r6,r2,L11		# continue until done
277 	clrq r6				# lrafx, llafx
278 	clrb (r5)			# flag end
279 	skpc $'0,$11,4(sp)		# skip over leading zeroes
280 	jbc $numsgn,flags,prn3	# easy if no left affix
281 	tstl -4(ap)				# original value
282 	jeql prn3			# no affix on 0, for some reason
283 	cmpl r3,$4			# were we doing hex or octal?
284 	jneq L12a			# octal
285 	movb $'x,r0
286 	jbc $caps,flags,L12b
287 	movb $'X,r0
288 L12b:	movb r0,-(r1)
289 	movl $2,llafx		# leading 0x for hex is an affix
290 L12a:	movb $'0,-(r1)	# leading zero for octal is a digit, not an affix
291 	jbr prn3			# omit sign (plus, blank) massaging
292 
293 unsigned:
294 lunsigned:
295 	bicl2 $1<plssgn|1<blank,flags	# omit sign (plus, blank) massaging
296 	extzv $1,$31,(ap),r0		# right shift logical 1 bit
297 	cvtlp r0,$10,(sp)		# convert [n/2] to packed
298 	movp $10,(sp),8(sp)		# copy packed
299 	addp4 $10,8(sp),$10,(sp)	# 2*[n/2] in packed, at (sp)
300 	blbc (ap)+,L14			# n was even
301 	addp4 $1,pone,$10,(sp)		# n was odd
302 	jbr L14
303 
304 patdec:					# editpc pattern for decimal printing
305 	.byte 0xAA			# eo$float 10
306 	.byte 0x01			# eo$end_float
307 	.byte 0				# eo$end
308 
309 decimal:
310 	cvtlp (ap)+,$10,(sp)		# 10 digits max
311 	jgeq L14
312 	incl llafx			# minus sign is a left affix
313 L14:	editpc $10,(sp),patdec,8(sp)	# ascii at 8(sp); r5=end+1
314 	skpc $' ,$11,8(sp)		# skip leading blanks; r1=first
315 
316 prnum:			# r1=addr first byte, r5=addr last byte +1, llafx=size of signs
317 				# -1(r1) vacant, for forced sign
318 	tstl llafx
319 	jneq prn3			# already some left affix, dont fuss
320 	jbc $plssgn,flags,prn2
321 	movb $'+,-(r1)		# needs a plus sign
322 	jbr prn4
323 prn2:	jbc $blank,flags,prn3
324 	movb $' ,-(r1)		# needs a blank sign
325 prn4:	incl llafx
326 prn3:	jbs $prec,flags,prn1
327 	movl $1,ndigit		# default precision is 1
328 prn1:	subl3 r1,r5,lrafx	# raw width
329 	subl2 llafx,lrafx	# number of digits
330 	subl2 lrafx,ndigit	# number of leading zeroes needed
331 	jleq prstr			# none
332 	addl2 llafx,r1		# where current digits start
333 	pushl r1			# movcx gobbles registers
334 		# check bounds on users who say %.300d
335 	movab 32(r5)[ndigit],r2
336 	subl2 fp,r2
337 	jlss prn5
338 	subl2 r2,ndigit
339 prn5:
340 		#
341 	movc3 lrafx,(r1),(r1)[ndigit]	# make room in middle
342 	movc5 $0,(r1),$ch.zer,ndigit,*(sp)	# '0 fill
343 	subl3 llafx,(sp)+,r1	# first byte addr
344 	addl3 lrafx,r3,r5	# last byte addr +1
345 
346 prstr:			# r1=addr first byte; r5=addr last byte +1
347 				# width=minimum width; llafx=len. left affix
348 				# ndigit=<avail>
349 	subl3 r1,r5,ndigit		# raw width
350 	subl3 ndigit,width,r0	# pad length
351 	jleq padlno				# in particular, no left padding
352 	jbs $minsgn,flags,padlno
353 			# extension for %0 flag causing left zero padding to field width
354 	jbs $zfill,flags,padlz
355 			# this bsbb needed even if %0 flag extension is removed
356 	bsbb padb				# blank pad on left
357 	jbr padnlz
358 padlz:
359 	movl llafx,r0
360 	jleq padnlx				# left zero pad requires left affix first
361 	subl2 r0,ndigit			# part of total length will be transferred
362 	subl2 r0,width			# and will account for part of minimum width
363 	bsbw strout				# left affix
364 padnlx:
365 	subl3 ndigit,width,r0	# pad length
366 	bsbb padz				# zero pad on left
367 padnlz:
368 			# end of extension for left zero padding
369 padlno:			# remaining: root, possible right padding
370 	subl2 ndigit,width		# root reduces minimum width
371 	movl ndigit,r0			# root length
372 p1:	bsbw strout				# transfer to output buffer
373 p3:	jbc $vbit,r2,padnpct	# percent sign (or null byte via %c) ?
374 	decl r0					# yes; adjust count
375 	movzbl (r1)+,r2			# fetch byte
376 	movq *fdesc,r4			# output buffer descriptor
377 	sobgeq r4,p2			# room at the out [inn] ?
378 	bsbw strout2			# no; force it, then try rest
379 	jbr p3					# here we go 'round the mullberry bush, ...
380 p2:	movb r2,(r5)+			# hand-deposit the percent or null
381 	incl nchar				# count it
382 	movq r4,*fdesc			# store output descriptor
383 	jbr p1					# what an expensive hiccup!
384 padnpct:
385 	movl width,r0	# size of pad
386 	jleq loop
387 	bsbb padb
388 	jbr loop
389 
390 padz:
391 	movb $'0,r2
392 	jbr pad
393 padb:
394 	movb $' ,r2
395 pad:
396 	subl2 r0,width			# pad width decreases minimum width
397 	pushl r1				# save non-pad addr
398 	movl r0,llafx			# remember width of pad
399 	subl2 r0,sp				# allocate
400 	movc5 $0,(r0),r2,llafx,(sp)	# create pad string
401 	movl llafx,r0			# length
402 	movl sp,r1				# addr
403 	bsbw strout
404 	addl2 llafx,sp			# deallocate
405 	movl (sp)+,r1			# recover non-pad addr
406 	rsb
407 
408 pone:	.byte	0x1C			# packed 1
409 
410 charac:
411 	movl (ap)+,r0		# word containing the char
412 	movb r0,(r5)+		# one byte, that's all
413 
414 prbuf:
415 	movl sp,r1			# addr first byte
416 	jbr prstr
417 
418 space:	bisl2 $1<blank,flags		# constant width e fmt, no plus sign
419 	jbr L4a
420 sharp:	bisl2 $1<numsgn,flags		# 'self identifying', please
421 	jbr L4a
422 plus:	bisl2 $1<plssgn,flags		# always print sign for floats
423 	jbr L4a
424 minus:	bisl2 $1<minsgn,flags		# left justification, please
425 	jbr L4a
426 gnum0:	jbs $ndfnd,flags,gnum
427 	jbs $prec,flags,gnump		# ignore when reading precision
428 	bisl2 $1<zfill,flags		# leading zero fill, please
429 gnum:	jbs $prec,flags,gnump
430 	moval (width)[width],width	# width *= 5;
431 	movaw -ch.zer(r0)[width],width	# width = 2*witdh + r0 - '0';
432 	jbr gnumd
433 gnump:	moval (ndigit)[ndigit],ndigit	# ndigit *= 5;
434 	movaw -ch.zer(r0)[ndigit],ndigit # ndigit = 2*ndigit + r0 - '0';
435 gnumd:	bisl2 $1<ndfnd,flags		# digit seen
436 	jbr L4a
437 dot:	clrl ndigit			# start on the precision
438 	bisl2 $1<prec,flags
439 	bicl2 $1<ndfnd,flags
440 	jbr L4a
441 indir:
442 	jbs $prec,flags,in1
443 	movl (ap)+,width		# width specified by parameter
444 	jgeq gnumd
445 	xorl2 $1<minsgn,flags		# parameterized left adjustment
446 	mnegl width,width
447 	jbr gnumd
448 in1:
449 	movl (ap)+,ndigit		# precision specified by paratmeter
450 	jgeq gnumd
451 	mnegl ndigit,ndigit
452 	jbr gnumd
453 
454 float:
455 	jbs $prec,flags,float1
456 	movl $6,ndigit			# default # digits to right of decpt.
457 float1:	bsbw fltcvt
458 	addl3 exp,ndigit,r7
459 	movl r7,r6			# for later "underflow" checking
460 	bgeq fxplrd
461 	clrl r7				# poor programmer planning
462 fxplrd:	cmpl r7,$31			# expressible in packed decimal?
463 	bleq fnarro			# yes
464 	movl $31,r7
465 fnarro:	subl3 $17,r7,r0			# where to round
466 	ashp r0,$17,(sp),$5,r7,16(sp)	# do it
467 	bvc fnovfl
468 		# band-aid for microcode error (spurious overflow)
469 	#	clrl r0				# assume even length result
470 	#	jlbc r7,fleven			# right
471 	#	movl $4,r0			# odd length result
472 	#fleven:	cmpv r0,$4,16(sp),$0		# top digit zero iff true overflow
473 	#	bneq fnovfl
474 		# end band-aid
475 	aobleq $0,r6,fnovfl		# if "underflow" then jump
476 	movl r7,r0
477 	incl exp
478 	incl r7
479 	ashp r0,$1,pone,$0,r7,16(sp)
480 	ashl $-1,r7,r0			# displ to last byte
481 	bisb2 sign,16(sp)[r0]		# insert sign
482 fnovfl:
483 	movab 16(sp),r1		# packed source
484 	movl r7,r6		# packed length
485 	pushab prnum	# goto prnum after fall-through call to fedit
486 
487 
488 	# enter via bsb
489 	#	r1=addr of packed source
490 	#	   16(r1) used to unpack source
491 	#	   48(r1) used to construct pattern to unpack source
492 	#	   48(r1) used to hold result
493 	#	r6=length of packed source (destroyed)
494 	#	exp=# digits to left of decimal point (destroyed)
495 	#	ndigit=# digits to right of decimal point (destroyed)
496 	#	sign=1 if negative, 0 otherwise
497 	# stack will be used for work space for pattern and unpacked source
498 	# exits with
499 	#	r1=addr of punctuated result
500 	#	r5=addr of last byte +1
501 	#	llafx=1 if minus sign inserted, 0 otherwise
502 fedit:
503 	pushab 48(r1)			# save result addr
504 	movab 48(r1),r3			# pattern addr
505 	movb $0x03,(r3)+		# eo$set_signif
506 	movc5 $0,(r1),$0x91,r6,(r3)	# eo$move 1
507 	clrb (r3)				# eo$end
508 	editpc r6,(r1),48(r1),16(r1)	# unpack 'em all
509 	subl3 r6,r5,r1			# addr unpacked source
510 	movl (sp),r3			# punctuated output placed here
511 	clrl llafx
512 	jlbc sign,f1
513 	movb $'-,(r3)+		# negative
514 	incl llafx
515 f1:	movl exp,r0
516 	jgtr f2
517 	movb $'0,(r3)+		# must have digit before decimal point
518 	jbr f3
519 f2:	cmpl r0,r6			# limit on packed length
520 	jleq f4
521 	movl r6,r0
522 f4:	subl2 r0,r6			# eat some digits
523 	subl2 r0,exp		# from the exponent
524 	movc3 r0,(r1),(r3)	# (most of the) digits to left of decimal point
525 	movl exp,r0			# need any more?
526 	jleq f3
527 	movc5 $0,(r1),$'0,r0,(r3)	# '0 fill
528 f3:	movl ndigit,r0		# # digits to right of decimal point
529 	jgtr f5
530 	jbs $numsgn,flags,f5	# no decimal point unless forced
531 	jbcs $dpflag,flags,f6	# no decimal point
532 f5:	movb $'.,(r3)+		# the decimal point
533 f6:	mnegl exp,r0		# "leading" zeroes to right of decimal point
534 	jleq f9
535 	cmpl r0,ndigit		# cant exceed this many
536 	jleq fa
537 	movl ndigit,r0
538 fa:	subl2 r0,ndigit
539 	movc5 $0,(r1),$'0,r0,(r3)
540 f9:	movl ndigit,r0
541 	cmpl r0,r6			# limit on packed length
542 	jleq f7
543 	movl r6,r0
544 f7:	subl2 r0,ndigit		# eat some digits from the fraction
545 	movc3 r0,(r1),(r3)	# (most of the) digits to right of decimal point
546 	movl ndigit,r0			# need any more?
547 	jleq f8
548 		# check bounds on users who say %.300f
549 	movab 32(r3)[r0],r2
550 	subl2 fp,r2
551 	jlss fb
552 	subl2 r2,r0			# truncate, willy-nilly
553 	movl r0,ndigit		# and no more digits later, either
554 fb:
555 		#
556 	subl2 r0,ndigit		# eat some digits from the fraction
557 	movc5 $0,(r1),$'0,r0,(r3)	# '0 fill
558 f8:	movl r3,r5			# addr last byte +1
559 	popr $1<1			# [movl (sp)+,r1] addr first byte
560 	rsb
561 
562 patexp:	.byte	0x03			# eo$set_signif
563 	.byte	0x44,'e			# eo$insert 'e
564 	.byte	0x42,'+			# eo$load_plus '+
565 	.byte	0x04			# eo$store_sign
566 	.byte	0x92			# eo$move 2
567 	.byte	0			# eo$end
568 
569 scien:
570 	incl ndigit
571 	jbs $prec,flags,L23
572 	movl $7,ndigit
573 L23:	bsbw fltcvt			# get packed digits
574 	movl ndigit,r7
575 	cmpl r7,$31				# expressible in packed decimal?
576 	jleq snarro				# yes
577 	movl $31,r7
578 snarro:	subl3 $17,r7,r0		# rounding position
579 	ashp r0,$17,(sp),$5,r7,16(sp) # shift and round
580 	bvc snovfl
581 		# band-aid for microcode error (spurious overflow)
582 	#	clrl r0				# assume even length result
583 	#	jlbc ndigit,sceven		# right
584 	#	movl $4,r0			# odd length result
585 	#sceven:	cmpv r0,$4,16(sp),$0		# top digit zero iff true overflow
586 	#	bneq snovfl
587 		# end band-aid
588 	incl exp			# rounding overflowed to 100...
589 	subl3 $1,r7,r0
590 	ashp r0,$1,pone,$0,r7,16(sp)
591 	ashl $-1,r7,r0		# displ to last byte
592 	bisb2 sign,16(sp)[r0]		# insert sign
593 snovfl:
594 	jbs $gflag,flags,gfmt		# %g format
595 	movab 16(sp),r1
596 	bsbb eedit
597 eexp:
598 	movl r1,r6		# save fwa from destruction by cvtlp
599 	subl3 $1,sexp,r0	# 1P exponent
600 	cvtlp r0,$2,(sp)	# packed
601 	editpc $2,(sp),patexp,(r5)
602 	movl r6,r1		# fwa
603 	jbc $caps,flags,prnum
604 	xorb2 $'e^'E,-4(r5)
605 	jbr prnum
606 
607 eedit:
608 	movl r7,r6		# packed length
609 	decl ndigit		# 1 digit before decimal point
610 	movl exp,sexp	# save from destruction
611 	movl $1,exp		# and pretend
612 	jbr fedit
613 
614 gfmt:
615 	addl3 $3,exp,r0		# exp is 1 more than e
616 	jlss gfmte		# (e+1)+3<0, e+4<=-1, e<=-5
617 	subl2 $3,r0		# exp [==(e+1)]
618 	cmpl r0,ndigit
619 	jgtr gfmte		# e+1>n, e>=n
620 gfmtf:
621 	movl r7,r6
622 	subl2 r0,ndigit		# n-e-1
623 	movab 16(sp),r1
624 	bsbw fedit
625 g1:	jbs $numsgn,flags,g2
626 	jbs $dpflag,flags,g2	# dont strip if no decimal point
627 g3:	cmpb -(r5),$'0		# strip trailing zeroes
628 	jeql g3
629 	cmpb (r5),$'.		# and trailing decimal point
630 	jeql g2
631 	incl r5
632 g2:	jbc $gflag,flags,eexp
633 	jbr prnum
634 gfmte:
635 	movab 16(sp),r1		# packed source
636 	bsbw eedit
637 	jbsc $gflag,flags,g1	# gflag now means "use %f" [hence no exponent]
638 
639 general:
640 	jbs $prec,flags,gn1
641 	movl $6,ndigit		# default precision is 6 significant digits
642 gn1:	tstl ndigit		# cannot allow precision of 0
643 	jgtr gn2
644 	movl $1,ndigit		# change 0 to 1, willy-nilly
645 gn2:	jbcs $gflag,flags,L23
646 	jbr L23			# safety net
647 
648 	# convert double-floating at (ap) to 17-digit packed at (sp),
649 	# set 'sign' and 'exp', advance ap.
650 fltcvt:
651 	clrb sign
652 	movd (ap)+,r5
653 	jeql fzero
654 	bgtr fpos
655 	mnegd r5,r5
656 	incb sign
657 fpos:
658 	extzv $7,$8,r5,r2		# exponent of 2
659 	movab -0200(r2),r2		# unbias
660 	mull2 $59,r2			# 59/196: 3rd convergent continued frac of log10(2)
661 	jlss eneg
662 	movab 196(r2),r2
663 eneg:
664 	movab -98(r2),r2
665 	divl2 $196,r2
666 	bsbw expten
667 	cmpd r0,r5
668 	bgtr ceil
669 	incl r2
670 ceil:	movl r2,exp
671 	mnegl r2,r2
672 	cmpl r2,$29			# 10^(29+9) is all we can handle
673 	bleq getman
674 	muld2 ten16,r5
675 	subl2 $16,r2
676 getman:	addl2 $9,r2			# -ceil(log10(x)) + 9
677 	jsb expten
678 	emodd r0,r4,r5,r0,r5		# (r0+r4)*r5; r0=int, r5=frac
679 fz1:	cvtlp r0,$9,16(sp)		# leading 9 digits
680 	ashp $8,$9,16(sp),$0,$17,4(sp)	# as top 9 of 17
681 	emodd ten8,$0,r5,r0,r5
682 	cvtlp r0,$8,16(sp)		# trailing 8 digits
683 		# if precision >= 17, must round here
684 	movl ndigit,r7			# so figure out what precision is
685 	pushab scien
686 	cmpl (sp)+,(sp)
687 	jleq gm1			# who called us?
688 	addl2 exp,r7			# float; adjust for exponent
689 gm1:	cmpl r7,$17
690 	jlss gm2
691 	cmpd r5,$0d0.5			# must round here; check fraction
692 	jlss gm2
693 	bisb2 $0x10,8+4(sp)		# increment l.s. digit
694 gm2:		# end of "round here" code
695 	addp4 $8,16(sp),$17,4(sp)	# combine leading and trailing
696 	bisb2 sign,12(sp)		# and insert sign
697 	rsb
698 fzero:	clrl r0
699 	movl $1,exp		# 0.000e+00 and 0.000 rather than 0.000e-01 and .000
700 	jbr fz1
701 
702 	.align 2
703 lsb: .long 0x00010000		# lsb in the crazy floating-point format
704 
705 	# return 10^r2 as a double float in r0||r1 and 8 extra bits of precision in r4
706 	# preserve r2, r5||r6
707 expten:
708 	movd $0d1.0,r0			# begin computing 10^exp10
709 	clrl r4				# bit counter
710 	movad ten1,r3			# table address
711 	tstl r2
712 	bgeq e10lp
713 	mnegl r2,r2			# get absolute value
714 	jbss $6,r2,e10lp		# flag as negative
715 e10lp:	jbc r4,r2,el1			# want this power?
716 	muld2 (r3),r0			# yes
717 el1:	addl2 $8,r3			# advance to next power
718 	aobleq $5,r4,e10lp		# through 10^32
719 	jbcc $6,r2,el2			# correct for negative exponent
720 	divd3 r0,$0d1.0,r0		# by taking reciprocal
721 	cmpl $28,r2
722 	jneq enm28
723 	addl2 lsb,r1			# 10**-28 needs lsb incremented
724 enm28:	mnegl r2,r2			# original exponent of 10
725 el2:	addl3 $5*8,r2,r3		# negative bit positions are illegal?
726 	jbc r3,xlsbh-5,eoklsb
727 	subl2 lsb,r1			# lsb was too high
728 eoklsb:
729 	movzbl xprec[r2],r4		# 8 extra bits
730 	rsb
731 
732 	# powers of ten
733 	.align	2
734 ten1:	.word	0x4220,0,0,0
735 ten2:	.word	0x43c8,0,0,0
736 ten4:	.word	0x471c,0x4000,0,0
737 ten8:	.word	0x4dbe,0xbc20,0,0
738 ten16:	.word	0x5b0e,0x1bc9,0xbf04,0
739 ten32:	.word	0x759d,0xc5ad,0xa82b,0x70b6
740 
741 	# whether lsb is too high or not
742 	.byte 1:0,1:0,1:0,1:0,1:1,1:0,1:1,1:0	# -40 thru -33
743 	.byte 1:0,1:1,1:0,1:0,1:0,1:0,1:1,1:0	# -32 thru -25
744 	.byte 1:0,1:0,1:1,1:1,1:1,1:1,1:0,1:0	# -24 thru -17
745 	.byte 1:0,1:1,1:0,1:0,1:1,1:1,1:1,1:1	# -16 thru -9
746 	.byte 1:1,1:1,1:1,1:0,1:0,1:0,1:0,1:1	# -8  thru -1
747 xlsbh:
748 	.byte 1:0,1:0,1:0,1:0,1:0,1:0,1:0,1:0	# 0 thru 7
749 	.byte 1:0,1:0,1:0,1:0,1:0,1:0,1:0,1:0	# 8 thru 15
750 	.byte 1:0,1:0,1:0,1:0,1:0,1:0,1:0,1:0	# 16 thru 23
751 	.byte 1:0,1:1,1:1,1:0,1:1,1:1,1:1,1:1	# 24 thru 31
752 	.byte 1:1,1:1,1:1,1:1,1:1,1:1,1:1    	# 32 thru 38
753 
754 	# bytes of extra precision
755 	.byte           0x56,0x76,0xd3,0x88,0xb5,0x62	# -38 thru -33
756 	.byte 0xba,0xf5,0x32,0x3e,0x0e,0x48,0xdb,0x51	# -32 thru -25
757 	.byte 0x53,0x27,0xb1,0xef,0xeb,0xa5,0x07,0x49	# -24 thru -17
758 	.byte 0x5b,0xd9,0x0f,0x13,0xcd,0xff,0xbf,0x97	# -16 thru -9
759 	.byte 0xfd,0xbc,0xb6,0x23,0x2c,0x3b,0x0a,0xcd	# -8  thru -1
760 xprec:
761 	.byte 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00	# 0  thru 7
762 	.byte 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00	# 8  thru 15
763 	.byte 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00	# 16 thru 23
764 	.byte 0x00,0xa0,0xc8,0x3a,0x84,0xe4,0xdc,0x92	# 24 thru 31
765 	.byte 0x9b,0x00,0xc0,0x58,0xae,0x18,0xef     	# 32 thru 38
766