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
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
|
/*
* from: vector.s, 386BSD 0.1 unknown origin
* $Id: vector.s,v 1.27 1997/04/27 21:18:59 fsmp Exp $
*/
/*
* modified for PC98 by Kakefuda
*/
#include "opt_auto_eoi.h"
#include "opt_smp.h"
#if defined(SMP)
#include <machine/smpasm.h> /* this includes <machine/apic.h> */
#endif /* SMP */
#include <i386/isa/icu.h>
#ifdef PC98
#include <pc98/pc98/pc98.h>
#else
#include <i386/isa/isa.h>
#endif
#ifdef PC98
#define ICU_IMR_OFFSET 2 /* IO_ICU{1,2} + 2 */
#else
#define ICU_IMR_OFFSET 1 /* IO_ICU{1,2} + 1 */
#endif
#if defined(SMP)
#define GET_MPLOCK call _get_mplock
#define REL_MPLOCK call _rel_mplock
#else
#define GET_MPLOCK /* NOP get Kernel Mutex */
#define REL_MPLOCK /* NOP release mutex */
#endif /* SMP */
#if defined(APIC_IO)
#define REDTBL_IDX(irq_num) (0x10 + ((irq_num) * 2))
#define IRQ_BIT(irq_num) (1 << (irq_num))
#define ENABLE_APIC \
movl _apic_base, %eax ; \
movl $0, APIC_EOI(%eax)
#define ENABLE_ICU1 ENABLE_APIC
#define ENABLE_ICU1_AND_2 ENABLE_APIC
#define MASK_IRQ(irq_num,icu) \
orl $IRQ_BIT(irq_num),_imen ; /* set the mask bit */ \
movl _io_apic_base,%ecx ; /* io apic addr */ \
movl $REDTBL_IDX(irq_num),(%ecx) ; /* write the index */ \
movl IOAPIC_WINDOW(%ecx),%eax ; /* current value */ \
orl $IOART_INTMASK,%eax ; /* set the mask */ \
movl %eax,IOAPIC_WINDOW(%ecx) ; /* new value */
#define UNMASK_IRQ(irq_num,icu) \
andl $~IRQ_BIT(irq_num),_imen ; /* clear mask bit */ \
movl _io_apic_base,%ecx ; /* io apic addr */ \
movl $REDTBL_IDX(irq_num),(%ecx) ; /* write the index */ \
movl IOAPIC_WINDOW(%ecx),%eax ; /* current value */ \
andl $~IOART_INTMASK,%eax ; /* clear the mask */ \
movl %eax,IOAPIC_WINDOW(%ecx) ; /* new value */
#define TEST_IRQ(irq_num,reg) \
testl $IRQ_BIT(irq_num),%eax
#define SET_IPENDING(irq_num) \
orl $IRQ_BIT(irq_num),_ipending
/*
* 'lazy masking' code submitted by: Bruce Evans <bde@zeta.org.au>
*/
#define MAYBE_MASK_IRQ(irq_num,icu) \
testl $IRQ_BIT(irq_num),iactive ; /* lazy masking */ \
je 1f ; /* NOT currently active */ \
MASK_IRQ(irq_num,icu) ; \
ENABLE_APIC ; \
SET_IPENDING(irq_num) ; \
REL_MPLOCK ; /* SMP release global lock */ \
popl %es ; \
popl %ds ; \
popal ; \
addl $4+4,%esp ; \
iret ; \
; \
ALIGN_TEXT ; \
1: ; \
orl $IRQ_BIT(irq_num),iactive
#define MAYBE_UNMASK_IRQ(irq_num,icu) \
andl $~IRQ_BIT(irq_num),iactive ; \
testl $IRQ_BIT(irq_num),_imen ; \
je 3f ; \
UNMASK_IRQ(irq_num,icu) ; \
3:
#else /* APIC_IO */
#define MASK_IRQ(irq_num,icu) \
movb _imen + IRQ_BYTE(irq_num),%al ; \
orb $IRQ_BIT(irq_num),%al ; \
movb %al,_imen + IRQ_BYTE(irq_num) ; \
outb %al,$icu+ICU_IMR_OFFSET
#define UNMASK_IRQ(irq_num,icu) \
movb _imen + IRQ_BYTE(irq_num),%al ; \
andb $~IRQ_BIT(irq_num),%al ; \
movb %al,_imen + IRQ_BYTE(irq_num) ; \
outb %al,$icu+ICU_IMR_OFFSET
#define TEST_IRQ(irq_num,reg) \
testb $IRQ_BIT(irq_num),%reg
#define SET_IPENDING(irq_num) \
orb $IRQ_BIT(irq_num),_ipending + IRQ_BYTE(irq_num)
#define ICU_EOI 0x20 /* XXX - define elsewhere */
#define IRQ_BIT(irq_num) (1 << ((irq_num) % 8))
#define IRQ_BYTE(irq_num) ((irq_num) / 8)
#ifdef AUTO_EOI_1
#define ENABLE_ICU1 /* use auto-EOI to reduce i/o */
#define OUTB_ICU1
#else
#define ENABLE_ICU1 \
movb $ICU_EOI,%al ; /* as soon as possible send EOI ... */ \
OUTB_ICU1 /* ... to clear in service bit */
#define OUTB_ICU1 \
outb %al,$IO_ICU1
#endif
#ifdef AUTO_EOI_2
/*
* The data sheet says no auto-EOI on slave, but it sometimes works.
*/
#define ENABLE_ICU1_AND_2 ENABLE_ICU1
#else
#define ENABLE_ICU1_AND_2 \
movb $ICU_EOI,%al ; /* as above */ \
outb %al,$IO_ICU2 ; /* but do second icu first ... */ \
OUTB_ICU1 /* ... then first icu (if !AUTO_EOI_1) */
#endif
#define MAYBE_MASK_IRQ(irq_num,icu) \
MASK_IRQ(irq_num,icu)
#define MAYBE_UNMASK_IRQ(irq_num,icu) \
UNMASK_IRQ(irq_num,icu)
#endif /* APIC_IO */
#ifdef FAST_INTR_HANDLER_USES_ES
#define ACTUALLY_PUSHED 1
#define MAYBE_MOVW_AX_ES movl %ax,%es
#define MAYBE_POPL_ES popl %es
#define MAYBE_PUSHL_ES pushl %es
#else
/*
* We can usually skip loading %es for fastintr handlers. %es should
* only be used for string instructions, and fastintr handlers shouldn't
* do anything slow enough to justify using a string instruction.
*/
#define ACTUALLY_PUSHED 0
#define MAYBE_MOVW_AX_ES
#define MAYBE_POPL_ES
#define MAYBE_PUSHL_ES
#endif
/*
* Macros for interrupt interrupt entry, call to handler, and exit.
*
* XXX - the interrupt frame is set up to look like a trap frame. This is
* usually a waste of time. The only interrupt handlers that want a frame
* are the clock handler (it wants a clock frame), the npx handler (it's
* easier to do right all in assembler). The interrupt return routine
* needs a trap frame for rare AST's (it could easily convert the frame).
* The direct costs of setting up a trap frame are two pushl's (error
* code and trap number), an addl to get rid of these, and pushing and
* popping the call-saved regs %esi, %edi and %ebp twice, The indirect
* costs are making the driver interface nonuniform so unpending of
* interrupts is more complicated and slower (call_driver(unit) would
* be easier than ensuring an interrupt frame for all handlers. Finally,
* there are some struct copies in the npx handler and maybe in the clock
* handler that could be avoided by working more with pointers to frames
* instead of frames.
*
* XXX - should we do a cld on every system entry to avoid the requirement
* for scattered cld's?
*
* Coding notes for *.s:
*
* If possible, avoid operations that involve an operand size override.
* Word-sized operations might be smaller, but the operand size override
* makes them slower on on 486's and no faster on 386's unless perhaps
* the instruction pipeline is depleted. E.g.,
*
* Use movl to seg regs instead of the equivalent but more descriptive
* movw - gas generates an irelevant (slower) operand size override.
*
* Use movl to ordinary regs in preference to movw and especially
* in preference to movz[bw]l. Use unsigned (long) variables with the
* top bits clear instead of unsigned short variables to provide more
* opportunities for movl.
*
* If possible, use byte-sized operations. They are smaller and no slower.
*
* Use (%reg) instead of 0(%reg) - gas generates larger code for the latter.
*
* If the interrupt frame is made more flexible, INTR can push %eax first
* and decide the ipending case with less overhead, e.g., by avoiding
* loading segregs.
*/
#define FAST_INTR(irq_num, vec_name, enable_icus) \
.text ; \
SUPERALIGN_TEXT ; \
IDTVEC(vec_name) ; \
pushl %eax ; /* save only call-used registers */ \
pushl %ecx ; \
pushl %edx ; \
pushl %ds ; \
MAYBE_PUSHL_ES ; \
movl $KDSEL,%eax ; \
movl %ax,%ds ; \
MAYBE_MOVW_AX_ES ; \
FAKE_MCOUNT((4+ACTUALLY_PUSHED)*4(%esp)) ; \
GET_MPLOCK ; /* SMP Spin lock */ \
pushl _intr_unit + (irq_num) * 4 ; \
call *_intr_handler + (irq_num) * 4 ; /* do the work ASAP */ \
enable_icus ; /* (re)enable ASAP (helps edge trigger?) */ \
addl $4,%esp ; \
incl _cnt+V_INTR ; /* book-keeping can wait */ \
movl _intr_countp + (irq_num) * 4,%eax ; \
incl (%eax) ; \
movl _cpl,%eax ; /* are we unmasking pending HWIs or SWIs? */ \
notl %eax ; \
andl _ipending,%eax ; \
jne 2f ; /* yes, maybe handle them */ \
1: ; \
MEXITCOUNT ; \
REL_MPLOCK ; /* SMP release global lock */ \
MAYBE_POPL_ES ; \
popl %ds ; \
popl %edx ; \
popl %ecx ; \
popl %eax ; \
iret ; \
; \
ALIGN_TEXT ; \
2: ; \
cmpb $3,_intr_nesting_level ; /* is there enough stack? */ \
jae 1b ; /* no, return */ \
movl _cpl,%eax ; \
/* XXX next line is probably unnecessary now. */ \
movl $HWI_MASK|SWI_MASK,_cpl ; /* limit nesting ... */ \
incb _intr_nesting_level ; /* ... really limit it ... */ \
sti ; /* ... to do this as early as possible */ \
MAYBE_POPL_ES ; /* discard most of thin frame ... */ \
popl %ecx ; /* ... original %ds ... */ \
popl %edx ; \
xchgl %eax,4(%esp) ; /* orig %eax; save cpl */ \
pushal ; /* build fat frame (grrr) ... */ \
pushl %ecx ; /* ... actually %ds ... */ \
pushl %es ; \
movl $KDSEL,%eax ; \
movl %ax,%es ; \
movl (2+8+0)*4(%esp),%ecx ; /* ... %ecx from thin frame ... */ \
movl %ecx,(2+6)*4(%esp) ; /* ... to fat frame ... */ \
movl (2+8+1)*4(%esp),%eax ; /* ... cpl from thin frame */ \
pushl %eax ; \
subl $4,%esp ; /* junk for unit number */ \
MEXITCOUNT ; \
jmp _doreti
#define INTR(irq_num, vec_name, icu, enable_icus, reg) \
.text ; \
SUPERALIGN_TEXT ; \
IDTVEC(vec_name) ; \
pushl $0 ; /* dummy error code */ \
pushl $0 ; /* dummy trap type */ \
pushal ; \
pushl %ds ; /* save our data and extra segments ... */ \
pushl %es ; \
movl $KDSEL,%eax ; /* ... and reload with kernel's own ... */ \
movl %ax,%ds ; /* ... early for obsolete reasons */ \
movl %ax,%es ; \
GET_MPLOCK ; /* SMP Spin lock */ \
MAYBE_MASK_IRQ(irq_num,icu) ; \
enable_icus ; \
movl _cpl,%eax ; \
TEST_IRQ(irq_num,reg) ; \
jne 2f ; \
incb _intr_nesting_level ; \
__CONCAT(Xresume,irq_num): ; \
FAKE_MCOUNT(12*4(%esp)) ; /* XXX late to avoid double count */ \
incl _cnt+V_INTR ; /* tally interrupts */ \
movl _intr_countp + (irq_num) * 4,%eax ; \
incl (%eax) ; \
movl _cpl,%eax ; \
pushl %eax ; \
pushl _intr_unit + (irq_num) * 4 ; \
orl _intr_mask + (irq_num) * 4,%eax ; \
movl %eax,_cpl ; \
sti ; \
call *_intr_handler + (irq_num) * 4 ; \
cli ; /* must unmask _imen and icu atomically */ \
MAYBE_UNMASK_IRQ(irq_num,icu) ; \
sti ; /* XXX _doreti repeats the cli/sti */ \
MEXITCOUNT ; \
/* We could usually avoid the following jmp by inlining some of */ \
/* _doreti, but it's probably better to use less cache. */ \
jmp _doreti ; \
; \
ALIGN_TEXT ; \
2: ; \
/* XXX skip mcounting here to avoid double count */ \
SET_IPENDING(irq_num) ; \
REL_MPLOCK ; /* SMP release global lock */ \
popl %es ; \
popl %ds ; \
popal ; \
addl $4+4,%esp ; \
iret
#if defined(APIC_IO)
.text
SUPERALIGN_TEXT
.globl _Xinvltlb
_Xinvltlb:
pushl %eax
movl %cr3, %eax
movl %eax, %cr3
ss
movl _apic_base, %eax
ss
movl $0, APIC_EOI(%eax)
popl %eax
iret
#endif /* APIC_IO */
MCOUNT_LABEL(bintr)
FAST_INTR(0,fastintr0, ENABLE_ICU1)
FAST_INTR(1,fastintr1, ENABLE_ICU1)
FAST_INTR(2,fastintr2, ENABLE_ICU1)
FAST_INTR(3,fastintr3, ENABLE_ICU1)
FAST_INTR(4,fastintr4, ENABLE_ICU1)
FAST_INTR(5,fastintr5, ENABLE_ICU1)
FAST_INTR(6,fastintr6, ENABLE_ICU1)
FAST_INTR(7,fastintr7, ENABLE_ICU1)
FAST_INTR(8,fastintr8, ENABLE_ICU1_AND_2)
FAST_INTR(9,fastintr9, ENABLE_ICU1_AND_2)
FAST_INTR(10,fastintr10, ENABLE_ICU1_AND_2)
FAST_INTR(11,fastintr11, ENABLE_ICU1_AND_2)
FAST_INTR(12,fastintr12, ENABLE_ICU1_AND_2)
FAST_INTR(13,fastintr13, ENABLE_ICU1_AND_2)
FAST_INTR(14,fastintr14, ENABLE_ICU1_AND_2)
FAST_INTR(15,fastintr15, ENABLE_ICU1_AND_2)
#if defined(APIC_IO)
FAST_INTR(16,fastintr16, ENABLE_ICU1_AND_2)
FAST_INTR(17,fastintr17, ENABLE_ICU1_AND_2)
FAST_INTR(18,fastintr18, ENABLE_ICU1_AND_2)
FAST_INTR(19,fastintr19, ENABLE_ICU1_AND_2)
FAST_INTR(20,fastintr20, ENABLE_ICU1_AND_2)
FAST_INTR(21,fastintr21, ENABLE_ICU1_AND_2)
FAST_INTR(22,fastintr22, ENABLE_ICU1_AND_2)
FAST_INTR(23,fastintr23, ENABLE_ICU1_AND_2)
#endif /* APIC_IO */
INTR(0,intr0, IO_ICU1, ENABLE_ICU1, al)
INTR(1,intr1, IO_ICU1, ENABLE_ICU1, al)
INTR(2,intr2, IO_ICU1, ENABLE_ICU1, al)
INTR(3,intr3, IO_ICU1, ENABLE_ICU1, al)
INTR(4,intr4, IO_ICU1, ENABLE_ICU1, al)
INTR(5,intr5, IO_ICU1, ENABLE_ICU1, al)
INTR(6,intr6, IO_ICU1, ENABLE_ICU1, al)
INTR(7,intr7, IO_ICU1, ENABLE_ICU1, al)
INTR(8,intr8, IO_ICU2, ENABLE_ICU1_AND_2, ah)
INTR(9,intr9, IO_ICU2, ENABLE_ICU1_AND_2, ah)
INTR(10,intr10, IO_ICU2, ENABLE_ICU1_AND_2, ah)
INTR(11,intr11, IO_ICU2, ENABLE_ICU1_AND_2, ah)
INTR(12,intr12, IO_ICU2, ENABLE_ICU1_AND_2, ah)
INTR(13,intr13, IO_ICU2, ENABLE_ICU1_AND_2, ah)
INTR(14,intr14, IO_ICU2, ENABLE_ICU1_AND_2, ah)
INTR(15,intr15, IO_ICU2, ENABLE_ICU1_AND_2, ah)
#if defined(APIC_IO)
INTR(16,intr16, IO_ICU2, ENABLE_ICU1_AND_2, ah)
INTR(17,intr17, IO_ICU2, ENABLE_ICU1_AND_2, ah)
INTR(18,intr18, IO_ICU2, ENABLE_ICU1_AND_2, ah)
INTR(19,intr19, IO_ICU2, ENABLE_ICU1_AND_2, ah)
INTR(20,intr20, IO_ICU2, ENABLE_ICU1_AND_2, ah)
INTR(21,intr21, IO_ICU2, ENABLE_ICU1_AND_2, ah)
INTR(22,intr22, IO_ICU2, ENABLE_ICU1_AND_2, ah)
INTR(23,intr23, IO_ICU2, ENABLE_ICU1_AND_2, ah)
#endif /* APIC_IO */
MCOUNT_LABEL(eintr)
.data
ihandlers: /* addresses of interrupt handlers */
/* actually resumption addresses for HWI's */
.long Xresume0, Xresume1, Xresume2, Xresume3
.long Xresume4, Xresume5, Xresume6, Xresume7
.long Xresume8, Xresume9, Xresume10, Xresume11
.long Xresume12, Xresume13, Xresume14, Xresume15
#if defined(APIC_IO)
.long Xresume16, Xresume17, Xresume18, Xresume19
.long Xresume20, Xresume21, Xresume22, Xresume23
.long 0, 0, 0, 0, swi_tty, swi_net, _softclock, swi_ast
#else
.long swi_tty, swi_net, 0, 0, 0, 0, 0, 0
.long 0, 0, 0, 0, 0, 0, _softclock, swi_ast
#endif /* APIC_IO */
imasks: /* masks for interrupt handlers */
.space NHWI*4 /* padding; HWI masks are elsewhere */
#if defined(APIC_IO)
.long 0, 0, 0, 0 /* padding */
.long SWI_TTY_MASK, SWI_NET_MASK, SWI_CLOCK_MASK, SWI_AST_MASK
#else
.long SWI_TTY_MASK, SWI_NET_MASK, 0, 0, 0, 0, 0, 0
.long 0, 0, 0, 0, 0, 0, SWI_CLOCK_MASK, SWI_AST_MASK
#endif /* APIC_IO */
.globl _intr_nesting_level
_intr_nesting_level:
.byte 0
.space 3
#if defined(APIC_IO)
.globl _ivectors
_ivectors:
.long _Xintr0, _Xintr1, _Xintr2, _Xintr3
.long _Xintr4, _Xintr5, _Xintr6, _Xintr7
.long _Xintr8, _Xintr9, _Xintr10, _Xintr11
.long _Xintr12, _Xintr13, _Xintr14, _Xintr15
.long _Xintr16, _Xintr17, _Xintr18, _Xintr19
.long _Xintr20, _Xintr21, _Xintr22, _Xintr23
/* active flag for lazy masking */
iactive:
.long 0
#endif /* APIC_IO */
/*
* Interrupt counters and names. The format of these and the label names
* must agree with what vmstat expects. The tables are indexed by device
* ids so that we don't have to move the names around as devices are
* attached.
*/
#include "vector.h"
.globl _intrcnt, _eintrcnt
_intrcnt:
.space (NR_DEVICES + ICU_LEN) * 4
_eintrcnt:
.globl _intrnames, _eintrnames
_intrnames:
.ascii DEVICE_NAMES
.asciz "stray irq0"
.asciz "stray irq1"
.asciz "stray irq2"
.asciz "stray irq3"
.asciz "stray irq4"
.asciz "stray irq5"
.asciz "stray irq6"
.asciz "stray irq7"
.asciz "stray irq8"
.asciz "stray irq9"
.asciz "stray irq10"
.asciz "stray irq11"
.asciz "stray irq12"
.asciz "stray irq13"
.asciz "stray irq14"
.asciz "stray irq15"
#if defined(APIC_IO)
.asciz "stray irq16"
.asciz "stray irq17"
.asciz "stray irq18"
.asciz "stray irq19"
.asciz "stray irq20"
.asciz "stray irq21"
.asciz "stray irq22"
.asciz "stray irq23"
#endif /* APIC_IO */
_eintrnames:
.text
|