1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
|
@c Copyright 1991, 1992, 1993, 1994, 1995, 1996, 2002
@c Free Software Foundation, Inc.
@c This is part of the GAS manual.
@c For copying conditions, see the file as.texinfo.
@ifset GENERIC
@page
@node i960-Dependent
@chapter Intel 80960 Dependent Features
@end ifset
@ifclear GENERIC
@node Machine Dependencies
@chapter Intel 80960 Dependent Features
@end ifclear
@cindex i960 support
@menu
* Options-i960:: i960 Command-line Options
* Floating Point-i960:: Floating Point
* Directives-i960:: i960 Machine Directives
* Opcodes for i960:: i960 Opcodes
@end menu
@c FIXME! Add Syntax sec with discussion of bitfields here, at least so
@c long as they're not turned on for other machines than 960.
@node Options-i960
@section i960 Command-line Options
@cindex i960 options
@cindex options, i960
@table @code
@cindex i960 architecture options
@cindex architecture options, i960
@cindex @code{-A} options, i960
@item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
Select the 80960 architecture. Instructions or features not supported
by the selected architecture cause fatal errors.
@samp{-ACA} is equivalent to @samp{-ACA_A}; @samp{-AKC} is equivalent to
@samp{-AMC}. Synonyms are provided for compatibility with other tools.
If you do not specify any of these options, @code{@value{AS}} generates code
for any instruction or feature that is supported by @emph{some} version of the
960 (even if this means mixing architectures!). In principle,
@code{@value{AS}} attempts to deduce the minimal sufficient processor type if
none is specified; depending on the object code format, the processor type may
be recorded in the object file. If it is critical that the @code{@value{AS}}
output match a specific architecture, specify that architecture explicitly.
@cindex @code{-b} option, i960
@cindex branch recording, i960
@cindex i960 branch recording
@item -b
Add code to collect information about conditional branches taken, for
later optimization using branch prediction bits. (The conditional branch
instructions have branch prediction bits in the CA, CB, and CC
architectures.) If @var{BR} represents a conditional branch instruction,
the following represents the code generated by the assembler when
@samp{-b} is specified:
@smallexample
call @var{increment routine}
.word 0 # pre-counter
Label: @var{BR}
call @var{increment routine}
.word 0 # post-counter
@end smallexample
The counter following a branch records the number of times that branch
was @emph{not} taken; the differenc between the two counters is the
number of times the branch @emph{was} taken.
@cindex @code{gbr960}, i960 postprocessor
@cindex branch statistics table, i960
A table of every such @code{Label} is also generated, so that the
external postprocessor @code{gbr960} (supplied by Intel) can locate all
the counters. This table is always labeled @samp{__BRANCH_TABLE__};
this is a local symbol to permit collecting statistics for many separate
object files. The table is word aligned, and begins with a two-word
header. The first word, initialized to 0, is used in maintaining linked
lists of branch tables. The second word is a count of the number of
entries in the table, which follow immediately: each is a word, pointing
to one of the labels illustrated above.
@c TEXI2ROFF-KILL
@ifinfo
@c END TEXI2ROFF-KILL
@example
+------------+------------+------------+ ... +------------+
| | | | | |
| *NEXT | COUNT: N | *BRLAB 1 | | *BRLAB N |
| | | | | |
+------------+------------+------------+ ... +------------+
__BRANCH_TABLE__ layout
@end example
@c TEXI2ROFF-KILL
@end ifinfo
@need 2000
@tex
\vskip 1pc
\line{\leftskip=0pt\hskip\tableindent
\boxit{2cm}{\tt *NEXT}\boxit{2cm}{\tt COUNT: \it N}\boxit{2cm}{\tt
*BRLAB 1}\ibox{1cm}{\quad\dots}\boxit{2cm}{\tt *BRLAB \it N}\hfil}
\centerline{\it {\tt \_\_BRANCH\_TABLE\_\_} layout}
@end tex
@c END TEXI2ROFF-KILL
The first word of the header is used to locate multiple branch tables,
since each object file may contain one. Normally the links are
maintained with a call to an initialization routine, placed at the
beginning of each function in the file. The @sc{gnu} C compiler
generates these calls automatically when you give it a @samp{-b} option.
For further details, see the documentation of @samp{gbr960}.
@cindex @code{-no-relax} option, i960
@item -no-relax
Normally, Compare-and-Branch instructions with targets that require
displacements greater than 13 bits (or that have external targets) are
replaced with the corresponding compare (or @samp{chkbit}) and branch
instructions. You can use the @samp{-no-relax} option to specify that
@code{@value{AS}} should generate errors instead, if the target displacement
is larger than 13 bits.
This option does not affect the Compare-and-Jump instructions; the code
emitted for them is @emph{always} adjusted when necessary (depending on
displacement size), regardless of whether you use @samp{-no-relax}.
@end table
@node Floating Point-i960
@section Floating Point
@cindex floating point, i960 (@sc{ieee})
@cindex i960 floating point (@sc{ieee})
@code{@value{AS}} generates @sc{ieee} floating-point numbers for the directives
@samp{.float}, @samp{.double}, @samp{.extended}, and @samp{.single}.
@node Directives-i960
@section i960 Machine Directives
@cindex machine directives, i960
@cindex i960 machine directives
@table @code
@cindex @code{bss} directive, i960
@item .bss @var{symbol}, @var{length}, @var{align}
Reserve @var{length} bytes in the bss section for a local @var{symbol},
aligned to the power of two specified by @var{align}. @var{length} and
@var{align} must be positive absolute expressions. This directive
differs from @samp{.lcomm} only in that it permits you to specify
an alignment. @xref{Lcomm,,@code{.lcomm}}.
@end table
@table @code
@cindex @code{extended} directive, i960
@item .extended @var{flonums}
@code{.extended} expects zero or more flonums, separated by commas; for
each flonum, @samp{.extended} emits an @sc{ieee} extended-format (80-bit)
floating-point number.
@cindex @code{leafproc} directive, i960
@item .leafproc @var{call-lab}, @var{bal-lab}
You can use the @samp{.leafproc} directive in conjunction with the
optimized @code{callj} instruction to enable faster calls of leaf
procedures. If a procedure is known to call no other procedures, you
may define an entry point that skips procedure prolog code (and that does
not depend on system-supplied saved context), and declare it as the
@var{bal-lab} using @samp{.leafproc}. If the procedure also has an
entry point that goes through the normal prolog, you can specify that
entry point as @var{call-lab}.
A @samp{.leafproc} declaration is meant for use in conjunction with the
optimized call instruction @samp{callj}; the directive records the data
needed later to choose between converting the @samp{callj} into a
@code{bal} or a @code{call}.
@var{call-lab} is optional; if only one argument is present, or if the
two arguments are identical, the single argument is assumed to be the
@code{bal} entry point.
@cindex @code{sysproc} directive, i960
@item .sysproc @var{name}, @var{index}
The @samp{.sysproc} directive defines a name for a system procedure.
After you define it using @samp{.sysproc}, you can use @var{name} to
refer to the system procedure identified by @var{index} when calling
procedures with the optimized call instruction @samp{callj}.
Both arguments are required; @var{index} must be between 0 and 31
(inclusive).
@end table
@node Opcodes for i960
@section i960 Opcodes
@cindex opcodes, i960
@cindex i960 opcodes
All Intel 960 machine instructions are supported;
@pxref{Options-i960,,i960 Command-line Options} for a discussion of
selecting the instruction subset for a particular 960
architecture.@refill
Some opcodes are processed beyond simply emitting a single corresponding
instruction: @samp{callj}, and Compare-and-Branch or Compare-and-Jump
instructions with target displacements larger than 13 bits.
@menu
* callj-i960:: @code{callj}
* Compare-and-branch-i960:: Compare-and-Branch
@end menu
@node callj-i960
@subsection @code{callj}
@cindex @code{callj}, i960 pseudo-opcode
@cindex i960 @code{callj} pseudo-opcode
You can write @code{callj} to have the assembler or the linker determine
the most appropriate form of subroutine call: @samp{call},
@samp{bal}, or @samp{calls}. If the assembly source contains
enough information---a @samp{.leafproc} or @samp{.sysproc} directive
defining the operand---then @code{@value{AS}} translates the
@code{callj}; if not, it simply emits the @code{callj}, leaving it
for the linker to resolve.
@node Compare-and-branch-i960
@subsection Compare-and-Branch
@cindex i960 compare/branch instructions
@cindex compare/branch instructions, i960
The 960 architectures provide combined Compare-and-Branch instructions
that permit you to store the branch target in the lower 13 bits of the
instruction word itself. However, if you specify a branch target far
enough away that its address won't fit in 13 bits, the assembler can
either issue an error, or convert your Compare-and-Branch instruction
into separate instructions to do the compare and the branch.
@cindex compare and jump expansions, i960
@cindex i960 compare and jump expansions
Whether @code{@value{AS}} gives an error or expands the instruction depends
on two choices you can make: whether you use the @samp{-no-relax} option,
and whether you use a ``Compare and Branch'' instruction or a ``Compare
and Jump'' instruction. The ``Jump'' instructions are @emph{always}
expanded if necessary; the ``Branch'' instructions are expanded when
necessary @emph{unless} you specify @code{-no-relax}---in which case
@code{@value{AS}} gives an error instead.
These are the Compare-and-Branch instructions, their ``Jump'' variants,
and the instruction pairs they may expand into:
@c TEXI2ROFF-KILL
@ifinfo
@c END TEXI2ROFF-KILL
@example
Compare and
Branch Jump Expanded to
------ ------ ------------
bbc chkbit; bno
bbs chkbit; bo
cmpibe cmpije cmpi; be
cmpibg cmpijg cmpi; bg
cmpibge cmpijge cmpi; bge
cmpibl cmpijl cmpi; bl
cmpible cmpijle cmpi; ble
cmpibno cmpijno cmpi; bno
cmpibne cmpijne cmpi; bne
cmpibo cmpijo cmpi; bo
cmpobe cmpoje cmpo; be
cmpobg cmpojg cmpo; bg
cmpobge cmpojge cmpo; bge
cmpobl cmpojl cmpo; bl
cmpoble cmpojle cmpo; ble
cmpobne cmpojne cmpo; bne
@end example
@c TEXI2ROFF-KILL
@end ifinfo
@tex
\hskip\tableindent
\halign{\hfil {\tt #}\quad&\hfil {\tt #}\qquad&{\tt #}\hfil\cr
\omit{\hfil\it Compare and\hfil}\span\omit&\cr
{\it Branch}&{\it Jump}&{\it Expanded to}\cr
bbc& & chkbit; bno\cr
bbs& & chkbit; bo\cr
cmpibe& cmpije& cmpi; be\cr
cmpibg& cmpijg& cmpi; bg\cr
cmpibge& cmpijge& cmpi; bge\cr
cmpibl& cmpijl& cmpi; bl\cr
cmpible& cmpijle& cmpi; ble\cr
cmpibno& cmpijno& cmpi; bno\cr
cmpibne& cmpijne& cmpi; bne\cr
cmpibo& cmpijo& cmpi; bo\cr
cmpobe& cmpoje& cmpo; be\cr
cmpobg& cmpojg& cmpo; bg\cr
cmpobge& cmpojge& cmpo; bge\cr
cmpobl& cmpojl& cmpo; bl\cr
cmpoble& cmpojle& cmpo; ble\cr
cmpobne& cmpojne& cmpo; bne\cr}
@end tex
@c END TEXI2ROFF-KILL
|