xref: /csrg-svn/lib/libc/vax/stdio.old/doprnt.c (revision 18326)
1 /* @(#)doprnt.c	4.6 (Berkeley) 03/13/85 */
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 	.align	1
50 __doprnt:
51 	.word	0xfc0			# uses r11-r6
52 	jbr	doit
53 
54 strfoo:
55 	clrl r4					# fix interrupt race
56 	jbr strok				# and try again
57 strout2:		# enter here to force out r2; r0,r1 must be set
58 	# do some tricks with line buffering (_IOLBF) first
59 	movl	fdesc,r3
60 	jbc	$7,16(r3),0f		# not line buffered (unbuffered)
61 	addl3	12(r3),8(r3),r4		# fdesc->_base+fdesc->_bufsiz
62 	cmpl	4(r3),r4		# buffer full?
63 	jgeq	0f			#  yes
64 	cmpl	r2,$10			# c == '\n'?
65 	jeql	0f			#  yes
66 	movb	r2,*4(r3)		# line buffered and not buffer full
67 	incl	4(r3)			#  and not newline
68 	clrl	(r3)			#  just stuff it and fix _cnt
69 	incl	nchar			# count the char
70 	jbr	strout			# skip __flsbuf
71 0:	pushr	$3				# save input descriptor
72 	pushl fdesc				# FILE
73 	pushl r2				# the char
74 	calls $2,__flsbuf		# please empty the buffer and handle 1 char
75 	tstl r0					# successful?
76 	jgeq strm1				# yes
77 	jbcs $31,nchar,strm1	# turn on sign bit of nchar to signify error
78 strm1:
79 	incl nchar				# count the char
80 	popr $3					# get input descriptor back
81 strout:			# enter via bsb with (r0,r1)=input descriptor
82 	movab strtab,r3			# table address
83 	movq *fdesc,r4			# output descriptor
84 	jbs $31,r4,strfoo		# negative count is a no no
85 strok:
86 	addl2 r0,nchar			# we intend to move this many chars
87 /******* Start bogus movtuc workaround  *****/
88 	clrl r2
89 	tstl    r0
90 	bleq    movdon
91 movlp:
92 	tstl    r4
93 	bleq    movdon
94 	movzbl  (r1)+,r3
95 	tstb    strtab[r3]
96 	bneq    1f
97 	mnegl   $1,r2
98 	decl    r1
99 	brb     movdon
100 1:
101 	movb    r3,(r5)+
102 	decl    r4
103 	sobgtr  r0,movlp
104   /******* End bogus movtuc workaround ***
105 	movtuc r0,(r1),$0,(r3),r4,(r5)
106 	movpsl r2                       /*  squirrel away condition codes */
107   /******* End equally bogus movtuc ****/
108 movdon: movq r4,*fdesc                  /*  update output descriptor */
109 	subl2 r0,nchar			# some chars not moved
110 	jbs $vbit,r2,stresc		# terminated by escape?
111 	sobgeq r0,strmore		# no; but out buffer might be full
112 stresc:
113 	rsb
114 strmore:
115 	movzbl (r1)+,r2			# one char
116 	tstb strtab[r2]			# translate
117 	jneq strout2			# bad guy in disguise (outbuf is full)
118 	incl r0				# fix the length
119 	decl r1				# and the addr
120 	movl $1<vbit,r2			# fake condition codes
121 	rsb
122 
123 errdone:
124 	jbcs $31,nchar,prdone	# set error bit
125 prdone:
126 	movl nchar,r0
127 	ret
128 
129 doit:
130 	movab -256(sp),sp		# work space
131 	movl 4(ap),r11			# addr of format string
132 	movl 12(ap),fdesc		# output FILE ptr
133 	movl 8(ap),ap			# addr of first arg
134 	clrl nchar				# number of chars transferred
135 loop:
136 	movzwl $65535,r0		# pseudo length
137 	movl r11,r1				# fmt addr
138 		# comet sucks.
139 	movq *fdesc,r4
140 	subl3 r1,r5,r2
141 	jlss lp1
142 	cmpl r0,r2
143 	jleq lp1
144 	movl r2,r0
145 lp1:
146 		#
147 	bsbw strout				# copy to output, stop at null or percent
148 	movl r1,r11				# new fmt
149 	jbc $vbit,r2,loop		# if no escape, then very long fmt
150 	tstb (r11)+				# escape; null or percent?
151 	jeql prdone				# null means end of fmt
152 
153 	movl sp,r5			# reset output buffer pointer
154 	clrq r9				# width; flags
155 	clrq r6				# lrafx,llafx
156 longorunsg:				# we can ignore both of these distinctions
157 short:
158 L4a:
159 	movzbl (r11)+,r0		# so capital letters can tail merge
160 L4:	caseb r0,$' ,$'x-' 		# format char
161 L5:
162 	.word space-L5			# space
163 	.word fmtbad-L5			# !
164 	.word fmtbad-L5			# "
165 	.word sharp-L5			# #
166 	.word fmtbad-L5			# $
167 	.word fmtbad-L5			# %
168 	.word fmtbad-L5			# &
169 	.word fmtbad-L5			# '
170 	.word fmtbad-L5			# (
171 	.word fmtbad-L5			# )
172 	.word indir-L5			# *
173 	.word plus-L5			# +
174 	.word fmtbad-L5			# ,
175 	.word minus-L5			# -
176 	.word dot-L5			# .
177 	.word fmtbad-L5			# /
178 	.word gnum0-L5			# 0
179 	.word gnum-L5			# 1
180 	.word gnum-L5			# 2
181 	.word gnum-L5			# 3
182 	.word gnum-L5			# 4
183 	.word gnum-L5			# 5
184 	.word gnum-L5			# 6
185 	.word gnum-L5			# 7
186 	.word gnum-L5			# 8
187 	.word gnum-L5			# 9
188 	.word fmtbad-L5			# :
189 	.word fmtbad-L5			# ;
190 	.word fmtbad-L5			# <
191 	.word fmtbad-L5			# =
192 	.word fmtbad-L5			# >
193 	.word fmtbad-L5			# ?
194 	.word fmtbad-L5			# @
195 	.word fmtbad-L5			# A
196 	.word fmtbad-L5			# B
197 	.word fmtbad-L5			# C
198 	.word decimal-L5		# D
199 	.word capital-L5		# E
200 	.word fmtbad-L5			# F
201 	.word capital-L5		# G
202 	.word fmtbad-L5			# H
203 	.word fmtbad-L5			# I
204 	.word fmtbad-L5			# J
205 	.word fmtbad-L5			# K
206 	.word fmtbad-L5			# L
207 	.word fmtbad-L5			# M
208 	.word fmtbad-L5			# N
209 	.word octal-L5			# O
210 	.word fmtbad-L5			# P
211 	.word fmtbad-L5			# Q
212 	.word fmtbad-L5			# R
213 	.word fmtbad-L5			# S
214 	.word fmtbad-L5			# T
215 	.word unsigned-L5		# U
216 	.word fmtbad-L5			# V
217 	.word fmtbad-L5			# W
218 	.word capital-L5		# X
219 	.word fmtbad-L5			# Y
220 	.word fmtbad-L5			# Z
221 	.word fmtbad-L5			# [
222 	.word fmtbad-L5			# \
223 	.word fmtbad-L5			# ]
224 	.word fmtbad-L5			# ^
225 	.word fmtbad-L5			# _
226 	.word fmtbad-L5			# `
227 	.word fmtbad-L5			# a
228 	.word fmtbad-L5			# b
229 	.word charac-L5			# c
230 	.word decimal-L5		# d
231 	.word scien-L5			# e
232 	.word float-L5			# f
233 	.word general-L5		# g
234 	.word short-L5			# h
235 	.word fmtbad-L5			# i
236 	.word fmtbad-L5			# j
237 	.word fmtbad-L5			# k
238 	.word longorunsg-L5		# l
239 	.word fmtbad-L5			# m
240 	.word fmtbad-L5			# n
241 	.word octal-L5			# o
242 	.word fmtbad-L5			# p
243 	.word fmtbad-L5			# q
244 	.word fmtbad-L5			# r
245 	.word string-L5			# s
246 	.word fmtbad-L5			# t
247 	.word unsigned-L5		# u
248 	.word fmtbad-L5			# v
249 	.word fmtbad-L5			# w
250 	.word hex-L5			# x
251 fmtbad:
252 	movb r0,(r5)+			# print the unfound character
253 	jeql errdone			# dumb users who end the format with a %
254 	jbr prbuf
255 capital:
256 	bisl2 $1<caps,flags		# note that it was capitalized
257 	xorb2 $'a^'A,r0			# make it small
258 	jbr L4					# and try again
259 
260 string:
261 	movl ndigit,r0
262 	jbs $prec,flags,L20		# max length was specified
263 	mnegl $1,r0			# default max length
264 L20:	movl (ap)+,r2			# addr first byte
265 	locc $0,r0,(r2)			# find the zero at the end
266 	movl r1,r5			# addr last byte +1
267 	movl r2,r1			# addr first byte
268 	jbr prstr
269 
270 htab:	.byte	'0,'1,'2,'3,'4,'5,'6,'7,'8,'9,'a,'b,'c,'d,'e,'f
271 Htab:	.byte	'0,'1,'2,'3,'4,'5,'6,'7,'8,'9,'A,'B,'C,'D,'E,'F
272 
273 octal:
274 	movl $30,r2			# init position
275 	movl $3,r3			# field width
276 	movab htab,llafx	# translate table
277 	jbr L10
278 
279 hex:
280 	movl $28,r2			# init position
281 	movl $4,r3			# field width
282 	movab htab,llafx	# translate table
283 	jbc $caps,flags,L10
284 	movab Htab,llafx
285 L10:	mnegl r3,r6			# increment
286 	clrl r1
287 	addl2 $4,r5			# room for left affix (2) and slop [forced sign?]
288 	movl (ap)+,r0			# fetch arg
289 L11:	extzv r2,r3,r0,r1		# pull out a digit
290 	movb (llafx)[r1],(r5)+		# convert to character
291 L12:	acbl $0,r6,r2,L11		# continue until done
292 	clrq r6				# lrafx, llafx
293 	clrb (r5)			# flag end
294 	skpc $'0,$11,4(sp)		# skip over leading zeroes
295 	jbc $numsgn,flags,prn3	# easy if no left affix
296 	tstl -4(ap)				# original value
297 	jeql prn3			# no affix on 0, for some reason
298 	cmpl r3,$4			# were we doing hex or octal?
299 	jneq L12a			# octal
300 	movb $'x,r0
301 	jbc $caps,flags,L12b
302 	movb $'X,r0
303 L12b:	movb r0,-(r1)
304 	movl $2,llafx		# leading 0x for hex is an affix
305 L12a:	movb $'0,-(r1)	# leading zero for octal is a digit, not an affix
306 	jbr prn3			# omit sign (plus, blank) massaging
307 
308 unsigned:
309 lunsigned:
310 	bicl2 $1<plssgn|1<blank,flags	# omit sign (plus, blank) massaging
311 	extzv $1,$31,(ap),r0		# right shift logical 1 bit
312 	cvtlp r0,$10,(sp)		# convert [n/2] to packed
313 	movp $10,(sp),8(sp)		# copy packed
314 	addp4 $10,8(sp),$10,(sp)	# 2*[n/2] in packed, at (sp)
315 	blbc (ap)+,L14			# n was even
316 	addp4 $1,pone,$10,(sp)		# n was odd
317 	jbr L14
318 
319 patdec:					# editpc pattern for decimal printing
320 	.byte 0xAA			# eo$float 10
321 	.byte 0x01			# eo$end_float
322 	.byte 0				# eo$end
323 
324 decimal:
325 	cvtlp (ap)+,$10,(sp)		# 10 digits max
326 	jgeq L14
327 	incl llafx			# minus sign is a left affix
328 L14:	editpc $10,(sp),patdec,8(sp)	# ascii at 8(sp); r5=end+1
329 	skpc $' ,$11,8(sp)		# skip leading blanks; r1=first
330 
331 prnum:			# r1=addr first byte, r5=addr last byte +1, llafx=size of signs
332 				# -1(r1) vacant, for forced sign
333 	tstl llafx
334 	jneq prn3			# already some left affix, dont fuss
335 	jbc $plssgn,flags,prn2
336 	movb $'+,-(r1)		# needs a plus sign
337 	jbr prn4
338 prn2:	jbc $blank,flags,prn3
339 	movb $' ,-(r1)		# needs a blank sign
340 prn4:	incl llafx
341 prn3:	jbs $prec,flags,prn1
342 	movl $1,ndigit		# default precision is 1
343 prn1:	subl3 r1,r5,lrafx	# raw width
344 	subl2 llafx,lrafx	# number of digits
345 	subl2 lrafx,ndigit	# number of leading zeroes needed
346 	jleq prstr			# none
347 	addl2 llafx,r1		# where current digits start
348 	pushl r1			# movcx gobbles registers
349 		# check bounds on users who say %.300d
350 	movab 32(r5)[ndigit],r2
351 	subl2 fp,r2
352 	jlss prn5
353 	subl2 r2,ndigit
354 prn5:
355 		#
356 	movc3 lrafx,(r1),(r1)[ndigit]	# make room in middle
357 	movc5 $0,(r1),$ch.zer,ndigit,*(sp)	# '0 fill
358 	subl3 llafx,(sp)+,r1	# first byte addr
359 	addl3 lrafx,r3,r5	# last byte addr +1
360 
361 prstr:			# r1=addr first byte; r5=addr last byte +1
362 				# width=minimum width; llafx=len. left affix
363 				# ndigit=<avail>
364 	subl3 r1,r5,ndigit		# raw width
365 	subl3 ndigit,width,r0	# pad length
366 	jleq padlno				# in particular, no left padding
367 	jbs $minsgn,flags,padlno
368 			# extension for %0 flag causing left zero padding to field width
369 	jbs $zfill,flags,padlz
370 			# this bsbb needed even if %0 flag extension is removed
371 	bsbb padb				# blank pad on left
372 	jbr padnlz
373 padlz:
374 	movl llafx,r0
375 	jleq padnlx				# left zero pad requires left affix first
376 	subl2 r0,ndigit			# part of total length will be transferred
377 	subl2 r0,width			# and will account for part of minimum width
378 	bsbw strout				# left affix
379 padnlx:
380 	subl3 ndigit,width,r0	# pad length
381 	bsbb padz				# zero pad on left
382 padnlz:
383 			# end of extension for left zero padding
384 padlno:			# remaining: root, possible right padding
385 	subl2 ndigit,width		# root reduces minimum width
386 	movl ndigit,r0			# root length
387 p1:	bsbw strout				# transfer to output buffer
388 p3:	jbc $vbit,r2,padnpct	# percent sign (or null byte via %c) ?
389 	decl r0					# yes; adjust count
390 	movzbl (r1)+,r2			# fetch byte
391 	movq *fdesc,r4			# output buffer descriptor
392 	sobgeq r4,p2			# room at the out [inn] ?
393 	bsbw strout2			# no; force it, then try rest
394 	jbr p3					# here we go 'round the mullberry bush, ...
395 p2:	movb r2,(r5)+			# hand-deposit the percent or null
396 	incl nchar				# count it
397 	movq r4,*fdesc			# store output descriptor
398 	jbr p1					# what an expensive hiccup!
399 padnpct:
400 	movl width,r0	# size of pad
401 	jleq loop
402 	bsbb padb
403 	jbr loop
404 
405 padz:
406 	movb $'0,r2
407 	jbr pad
408 padb:
409 	movb $' ,r2
410 pad:
411 	subl2 r0,width			# pad width decreases minimum width
412 	pushl r1				# save non-pad addr
413 	movl r0,llafx			# remember width of pad
414 	subl2 r0,sp				# allocate
415 	movc5 $0,(r0),r2,llafx,(sp)	# create pad string
416 	movl llafx,r0			# length
417 	movl sp,r1				# addr
418 	bsbw strout
419 	addl2 llafx,sp			# deallocate
420 	movl (sp)+,r1			# recover non-pad addr
421 	rsb
422 
423 pone:	.byte	0x1C			# packed 1
424 
425 charac:
426 	movl (ap)+,r0		# word containing the char
427 	movb r0,(r5)+		# one byte, that's all
428 
429 prbuf:
430 	movl sp,r1			# addr first byte
431 	jbr prstr
432 
433 space:	bisl2 $1<blank,flags		# constant width e fmt, no plus sign
434 	jbr L4a
435 sharp:	bisl2 $1<numsgn,flags		# 'self identifying', please
436 	jbr L4a
437 plus:	bisl2 $1<plssgn,flags		# always print sign for floats
438 	jbr L4a
439 minus:	bisl2 $1<minsgn,flags		# left justification, please
440 	jbr L4a
441 gnum0:	jbs $ndfnd,flags,gnum
442 	jbs $prec,flags,gnump		# ignore when reading precision
443 	bisl2 $1<zfill,flags		# leading zero fill, please
444 gnum:	jbs $prec,flags,gnump
445 	moval (width)[width],width	# width *= 5;
446 	movaw -ch.zer(r0)[width],width	# width = 2*witdh + r0 - '0';
447 	jbr gnumd
448 gnump:	moval (ndigit)[ndigit],ndigit	# ndigit *= 5;
449 	movaw -ch.zer(r0)[ndigit],ndigit # ndigit = 2*ndigit + r0 - '0';
450 gnumd:	bisl2 $1<ndfnd,flags		# digit seen
451 	jbr L4a
452 dot:	clrl ndigit			# start on the precision
453 	bisl2 $1<prec,flags
454 	bicl2 $1<ndfnd,flags
455 	jbr L4a
456 indir:
457 	jbs $prec,flags,in1
458 	movl (ap)+,width		# width specified by parameter
459 	jgeq gnumd
460 	xorl2 $1<minsgn,flags		# parameterized left adjustment
461 	mnegl width,width
462 	jbr gnumd
463 in1:
464 	movl (ap)+,ndigit		# precision specified by paratmeter
465 	jgeq gnumd
466 	mnegl ndigit,ndigit
467 	jbr gnumd
468 
469 float:
470 	jbs $prec,flags,float1
471 	movl $6,ndigit			# default # digits to right of decpt.
472 float1:	bsbw fltcvt
473 	addl3 exp,ndigit,r7
474 	movl r7,r6			# for later "underflow" checking
475 	bgeq fxplrd
476 	clrl r7				# poor programmer planning
477 fxplrd:	cmpl r7,$31			# expressible in packed decimal?
478 	bleq fnarro			# yes
479 	movl $31,r7
480 fnarro:	subl3 $17,r7,r0			# where to round
481 	ashp r0,$17,(sp),$5,r7,16(sp)	# do it
482 	bvc fnovfl
483 		# band-aid for microcode error (spurious overflow)
484 	#	clrl r0				# assume even length result
485 	#	jlbc r7,fleven			# right
486 	#	movl $4,r0			# odd length result
487 	#fleven:	cmpv r0,$4,16(sp),$0		# top digit zero iff true overflow
488 	#	bneq fnovfl
489 		# end band-aid
490 	aobleq $0,r6,fnovfl		# if "underflow" then jump
491 	movl r7,r0
492 	incl exp
493 	incl r7
494 	ashp r0,$1,pone,$0,r7,16(sp)
495 	ashl $-1,r7,r0			# displ to last byte
496 	bisb2 sign,16(sp)[r0]		# insert sign
497 fnovfl:
498 	movab 16(sp),r1		# packed source
499 	movl r7,r6		# packed length
500 	pushab prnum	# goto prnum after fall-through call to fedit
501 
502 
503 	# enter via bsb
504 	#	r1=addr of packed source
505 	#	   16(r1) used to unpack source
506 	#	   48(r1) used to construct pattern to unpack source
507 	#	   48(r1) used to hold result
508 	#	r6=length of packed source (destroyed)
509 	#	exp=# digits to left of decimal point (destroyed)
510 	#	ndigit=# digits to right of decimal point (destroyed)
511 	#	sign=1 if negative, 0 otherwise
512 	# stack will be used for work space for pattern and unpacked source
513 	# exits with
514 	#	r1=addr of punctuated result
515 	#	r5=addr of last byte +1
516 	#	llafx=1 if minus sign inserted, 0 otherwise
517 fedit:
518 	pushab 48(r1)			# save result addr
519 	movab 48(r1),r3			# pattern addr
520 	movb $0x03,(r3)+		# eo$set_signif
521 	movc5 $0,(r1),$0x91,r6,(r3)	# eo$move 1
522 	clrb (r3)				# eo$end
523 	editpc r6,(r1),48(r1),16(r1)	# unpack 'em all
524 	subl3 r6,r5,r1			# addr unpacked source
525 	movl (sp),r3			# punctuated output placed here
526 	clrl llafx
527 	jlbc sign,f1
528 	movb $'-,(r3)+		# negative
529 	incl llafx
530 f1:	movl exp,r0
531 	jgtr f2
532 	movb $'0,(r3)+		# must have digit before decimal point
533 	jbr f3
534 f2:	cmpl r0,r6			# limit on packed length
535 	jleq f4
536 	movl r6,r0
537 f4:	subl2 r0,r6			# eat some digits
538 	subl2 r0,exp		# from the exponent
539 	movc3 r0,(r1),(r3)	# (most of the) digits to left of decimal point
540 	movl exp,r0			# need any more?
541 	jleq f3
542 	movc5 $0,(r1),$'0,r0,(r3)	# '0 fill
543 f3:	movl ndigit,r0		# # digits to right of decimal point
544 	jgtr f5
545 	jbs $numsgn,flags,f5	# no decimal point unless forced
546 	jbcs $dpflag,flags,f6	# no decimal point
547 f5:	movb $'.,(r3)+		# the decimal point
548 f6:	mnegl exp,r0		# "leading" zeroes to right of decimal point
549 	jleq f9
550 	cmpl r0,ndigit		# cant exceed this many
551 	jleq fa
552 	movl ndigit,r0
553 fa:	subl2 r0,ndigit
554 	movc5 $0,(r1),$'0,r0,(r3)
555 f9:	movl ndigit,r0
556 	cmpl r0,r6			# limit on packed length
557 	jleq f7
558 	movl r6,r0
559 f7:	subl2 r0,ndigit		# eat some digits from the fraction
560 	movc3 r0,(r1),(r3)	# (most of the) digits to right of decimal point
561 	movl ndigit,r0			# need any more?
562 	jleq f8
563 		# check bounds on users who say %.300f
564 	movab 32(r3)[r0],r2
565 	subl2 fp,r2
566 	jlss fb
567 	subl2 r2,r0			# truncate, willy-nilly
568 	movl r0,ndigit		# and no more digits later, either
569 fb:
570 		#
571 	subl2 r0,ndigit		# eat some digits from the fraction
572 	movc5 $0,(r1),$'0,r0,(r3)	# '0 fill
573 f8:	movl r3,r5			# addr last byte +1
574 	popr $1<1			# [movl (sp)+,r1] addr first byte
575 	rsb
576 
577 patexp:	.byte	0x03			# eo$set_signif
578 	.byte	0x44,'e			# eo$insert 'e
579 	.byte	0x42,'+			# eo$load_plus '+
580 	.byte	0x04			# eo$store_sign
581 	.byte	0x92			# eo$move 2
582 	.byte	0			# eo$end
583 
584 scien:
585 	incl ndigit
586 	jbs $prec,flags,L23
587 	movl $7,ndigit
588 L23:	bsbw fltcvt			# get packed digits
589 	movl ndigit,r7
590 	cmpl r7,$31				# expressible in packed decimal?
591 	jleq snarro				# yes
592 	movl $31,r7
593 snarro:	subl3 $17,r7,r0		# rounding position
594 	ashp r0,$17,(sp),$5,r7,16(sp) # shift and round
595 	bvc snovfl
596 		# band-aid for microcode error (spurious overflow)
597 	#	clrl r0				# assume even length result
598 	#	jlbc ndigit,sceven		# right
599 	#	movl $4,r0			# odd length result
600 	#sceven:	cmpv r0,$4,16(sp),$0		# top digit zero iff true overflow
601 	#	bneq snovfl
602 		# end band-aid
603 	incl exp			# rounding overflowed to 100...
604 	subl3 $1,r7,r0
605 	ashp r0,$1,pone,$0,r7,16(sp)
606 	ashl $-1,r7,r0		# displ to last byte
607 	bisb2 sign,16(sp)[r0]		# insert sign
608 snovfl:
609 	jbs $gflag,flags,gfmt		# %g format
610 	movab 16(sp),r1
611 	bsbb eedit
612 eexp:
613 	movl r1,r6		# save fwa from destruction by cvtlp
614 	subl3 $1,sexp,r0	# 1P exponent
615 	cvtlp r0,$2,(sp)	# packed
616 	editpc $2,(sp),patexp,(r5)
617 	movl r6,r1		# fwa
618 	jbc $caps,flags,prnum
619 	xorb2 $'e^'E,-4(r5)
620 	jbr prnum
621 
622 eedit:
623 	movl r7,r6		# packed length
624 	decl ndigit		# 1 digit before decimal point
625 	movl exp,sexp	# save from destruction
626 	movl $1,exp		# and pretend
627 	jbr fedit
628 
629 gfmt:
630 	addl3 $3,exp,r0		# exp is 1 more than e
631 	jlss gfmte		# (e+1)+3<0, e+4<=-1, e<=-5
632 	subl2 $3,r0		# exp [==(e+1)]
633 	cmpl r0,ndigit
634 	jgtr gfmte		# e+1>n, e>=n
635 gfmtf:
636 	movl r7,r6
637 	subl2 r0,ndigit		# n-e-1
638 	movab 16(sp),r1
639 	bsbw fedit
640 g1:	jbs $numsgn,flags,g2
641 	jbs $dpflag,flags,g2	# dont strip if no decimal point
642 g3:	cmpb -(r5),$'0		# strip trailing zeroes
643 	jeql g3
644 	cmpb (r5),$'.		# and trailing decimal point
645 	jeql g2
646 	incl r5
647 g2:	jbc $gflag,flags,eexp
648 	jbr prnum
649 gfmte:
650 	movab 16(sp),r1		# packed source
651 	bsbw eedit
652 	jbsc $gflag,flags,g1	# gflag now means "use %f" [hence no exponent]
653 
654 general:
655 	jbs $prec,flags,gn1
656 	movl $6,ndigit		# default precision is 6 significant digits
657 gn1:	tstl ndigit		# cannot allow precision of 0
658 	jgtr gn2
659 	movl $1,ndigit		# change 0 to 1, willy-nilly
660 gn2:	jbcs $gflag,flags,L23
661 	jbr L23			# safety net
662 
663 	# convert double-floating at (ap) to 17-digit packed at (sp),
664 	# set 'sign' and 'exp', advance ap.
665 fltcvt:
666 	clrb sign
667 	movd (ap)+,r5
668 	jeql fzero
669 	bgtr fpos
670 	mnegd r5,r5
671 	incb sign
672 fpos:
673 	extzv $7,$8,r5,r2		# exponent of 2
674 	movab -0200(r2),r2		# unbias
675 	mull2 $59,r2			# 59/196: 3rd convergent continued frac of log10(2)
676 	jlss eneg
677 	movab 196(r2),r2
678 eneg:
679 	movab -98(r2),r2
680 	divl2 $196,r2
681 	bsbw expten
682 	cmpd r0,r5
683 	bgtr ceil
684 	incl r2
685 ceil:	movl r2,exp
686 	mnegl r2,r2
687 	cmpl r2,$29			# 10^(29+9) is all we can handle
688 	bleq getman
689 	muld2 ten16,r5
690 	subl2 $16,r2
691 getman:	addl2 $9,r2			# -ceil(log10(x)) + 9
692 	jsb expten
693 	emodd r0,r4,r5,r0,r5		# (r0+r4)*r5; r0=int, r5=frac
694 fz1:	cvtlp r0,$9,16(sp)		# leading 9 digits
695 	ashp $8,$9,16(sp),$0,$17,4(sp)	# as top 9 of 17
696 	emodd ten8,$0,r5,r0,r5
697 	cvtlp r0,$8,16(sp)		# trailing 8 digits
698 		# if precision >= 17, must round here
699 	movl ndigit,r7			# so figure out what precision is
700 	pushab scien
701 	cmpl (sp)+,(sp)
702 	jleq gm1			# who called us?
703 	addl2 exp,r7			# float; adjust for exponent
704 gm1:	cmpl r7,$17
705 	jlss gm2
706 	cmpd r5,$0d0.5			# must round here; check fraction
707 	jlss gm2
708 	bisb2 $0x10,8+4(sp)		# increment l.s. digit
709 gm2:		# end of "round here" code
710 	addp4 $8,16(sp),$17,4(sp)	# combine leading and trailing
711 	bisb2 sign,12(sp)		# and insert sign
712 	rsb
713 fzero:	clrl r0
714 	movl $1,exp		# 0.000e+00 and 0.000 rather than 0.000e-01 and .000
715 	jbr fz1
716 
717 	.align 2
718 lsb: .long 0x00010000		# lsb in the crazy floating-point format
719 
720 	# return 10^r2 as a double float in r0||r1 and 8 extra bits of precision in r4
721 	# preserve r2, r5||r6
722 expten:
723 	movd $0d1.0,r0			# begin computing 10^exp10
724 	clrl r4				# bit counter
725 	movad ten1,r3			# table address
726 	tstl r2
727 	bgeq e10lp
728 	mnegl r2,r2			# get absolute value
729 	jbss $6,r2,e10lp		# flag as negative
730 e10lp:	jbc r4,r2,el1			# want this power?
731 	muld2 (r3),r0			# yes
732 el1:	addl2 $8,r3			# advance to next power
733 	aobleq $5,r4,e10lp		# through 10^32
734 	jbcc $6,r2,el2			# correct for negative exponent
735 	divd3 r0,$0d1.0,r0		# by taking reciprocal
736 	cmpl $28,r2
737 	jneq enm28
738 	addl2 lsb,r1			# 10**-28 needs lsb incremented
739 enm28:	mnegl r2,r2			# original exponent of 10
740 el2:	addl3 $5*8,r2,r3		# negative bit positions are illegal?
741 	jbc r3,xlsbh-5,eoklsb
742 	subl2 lsb,r1			# lsb was too high
743 eoklsb:
744 	movzbl xprec[r2],r4		# 8 extra bits
745 	rsb
746 
747 	# powers of ten
748 	.align	2
749 ten1:	.word	0x4220,0,0,0
750 ten2:	.word	0x43c8,0,0,0
751 ten4:	.word	0x471c,0x4000,0,0
752 ten8:	.word	0x4dbe,0xbc20,0,0
753 ten16:	.word	0x5b0e,0x1bc9,0xbf04,0
754 ten32:	.word	0x759d,0xc5ad,0xa82b,0x70b6
755 
756 	# whether lsb is too high or not
757 	.byte 1:0,1:0,1:0,1:0,1:1,1:0,1:1,1:0	# -40 thru -33
758 	.byte 1:0,1:1,1:0,1:0,1:0,1:0,1:1,1:0	# -32 thru -25
759 	.byte 1:0,1:0,1:1,1:1,1:1,1:1,1:0,1:0	# -24 thru -17
760 	.byte 1:0,1:1,1:0,1:0,1:1,1:1,1:1,1:1	# -16 thru -9
761 	.byte 1:1,1:1,1:1,1:0,1:0,1:0,1:0,1:1	# -8  thru -1
762 xlsbh:
763 	.byte 1:0,1:0,1:0,1:0,1:0,1:0,1:0,1:0	# 0 thru 7
764 	.byte 1:0,1:0,1:0,1:0,1:0,1:0,1:0,1:0	# 8 thru 15
765 	.byte 1:0,1:0,1:0,1:0,1:0,1:0,1:0,1:0	# 16 thru 23
766 	.byte 1:0,1:1,1:1,1:0,1:1,1:1,1:1,1:1	# 24 thru 31
767 	.byte 1:1,1:1,1:1,1:1,1:1,1:1,1:1    	# 32 thru 38
768 
769 	# bytes of extra precision
770 	.byte           0x56,0x76,0xd3,0x88,0xb5,0x62	# -38 thru -33
771 	.byte 0xba,0xf5,0x32,0x3e,0x0e,0x48,0xdb,0x51	# -32 thru -25
772 	.byte 0x53,0x27,0xb1,0xef,0xeb,0xa5,0x07,0x49	# -24 thru -17
773 	.byte 0x5b,0xd9,0x0f,0x13,0xcd,0xff,0xbf,0x97	# -16 thru -9
774 	.byte 0xfd,0xbc,0xb6,0x23,0x2c,0x3b,0x0a,0xcd	# -8  thru -1
775 xprec:
776 	.byte 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00	# 0  thru 7
777 	.byte 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00	# 8  thru 15
778 	.byte 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00	# 16 thru 23
779 	.byte 0x00,0xa0,0xc8,0x3a,0x84,0xe4,0xdc,0x92	# 24 thru 31
780 	.byte 0x9b,0x00,0xc0,0x58,0xae,0x18,0xef     	# 32 thru 38
781