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
|
/* SPDX-License-Identifier: GPL-2.0-only */
/*
* Event entry/exit for Hexagon
*
* Copyright (c) 2010-2013, The Linux Foundation. All rights reserved.
*/
#include <asm/asm-offsets.h> /* assembly-safer versions of C defines */
#include <asm/mem-layout.h> /* sigh, except for page_offset */
#include <asm/hexagon_vm.h>
#include <asm/thread_info.h>
/*
* Entry into guest-mode Linux under Hexagon Virtual Machine.
* Stack pointer points to event record - build pt_regs on top of it,
* set up a plausible C stack frame, and dispatch to the C handler.
* On return, do vmrte virtual instruction with SP where we started.
*
* VM Spec 0.5 uses a trap to fetch HVM record now.
*/
/*
* Save full register state, while setting up thread_info struct
* pointer derived from kernel stack pointer in THREADINFO_REG
* register, putting prior thread_info.regs pointer in a callee-save
* register (R24, which had better not ever be assigned to THREADINFO_REG),
* and updating thread_info.regs to point to current stack frame,
* so as to support nested events in kernel mode.
*
* As this is common code, we set the pt_regs system call number
* to -1 for all events. It will be replaced with the system call
* number in the case where we decode a system call (trap0(#1)).
*/
#if CONFIG_HEXAGON_ARCH_VERSION < 4
#define save_pt_regs()\
memd(R0 + #_PT_R3130) = R31:30; \
{ memw(R0 + #_PT_R2928) = R28; \
R31 = memw(R0 + #_PT_ER_VMPSP); }\
{ memw(R0 + #(_PT_R2928 + 4)) = R31; \
R31 = ugp; } \
{ memd(R0 + #_PT_R2726) = R27:26; \
R30 = gp ; } \
memd(R0 + #_PT_R2524) = R25:24; \
memd(R0 + #_PT_R2322) = R23:22; \
memd(R0 + #_PT_R2120) = R21:20; \
memd(R0 + #_PT_R1918) = R19:18; \
memd(R0 + #_PT_R1716) = R17:16; \
memd(R0 + #_PT_R1514) = R15:14; \
memd(R0 + #_PT_R1312) = R13:12; \
{ memd(R0 + #_PT_R1110) = R11:10; \
R15 = lc0; } \
{ memd(R0 + #_PT_R0908) = R9:8; \
R14 = sa0; } \
{ memd(R0 + #_PT_R0706) = R7:6; \
R13 = lc1; } \
{ memd(R0 + #_PT_R0504) = R5:4; \
R12 = sa1; } \
{ memd(R0 + #_PT_GPUGP) = R31:30; \
R11 = m1; \
R2.H = #HI(_THREAD_SIZE); } \
{ memd(R0 + #_PT_LC0SA0) = R15:14; \
R10 = m0; \
R2.L = #LO(_THREAD_SIZE); } \
{ memd(R0 + #_PT_LC1SA1) = R13:12; \
R15 = p3:0; \
R2 = neg(R2); } \
{ memd(R0 + #_PT_M1M0) = R11:10; \
R14 = usr; \
R2 = and(R0,R2); } \
{ memd(R0 + #_PT_PREDSUSR) = R15:14; \
THREADINFO_REG = R2; } \
{ r24 = memw(THREADINFO_REG + #_THREAD_INFO_PT_REGS); \
memw(THREADINFO_REG + #_THREAD_INFO_PT_REGS) = R0; \
R2 = #-1; } \
{ memw(R0 + #_PT_SYSCALL_NR) = R2; \
R30 = #0; }
#else
/* V4+ */
/* the # ## # syntax inserts a literal ## */
#define save_pt_regs()\
{ memd(R0 + #_PT_R3130) = R31:30; \
R30 = memw(R0 + #_PT_ER_VMPSP); }\
{ memw(R0 + #_PT_R2928) = R28; \
memw(R0 + #(_PT_R2928 + 4)) = R30; }\
{ R31:30 = C11:10; \
memd(R0 + #_PT_R2726) = R27:26; \
memd(R0 + #_PT_R2524) = R25:24; }\
{ memd(R0 + #_PT_R2322) = R23:22; \
memd(R0 + #_PT_R2120) = R21:20; }\
{ memd(R0 + #_PT_R1918) = R19:18; \
memd(R0 + #_PT_R1716) = R17:16; }\
{ memd(R0 + #_PT_R1514) = R15:14; \
memd(R0 + #_PT_R1312) = R13:12; \
R17:16 = C13:12; }\
{ memd(R0 + #_PT_R1110) = R11:10; \
memd(R0 + #_PT_R0908) = R9:8; \
R15:14 = C1:0; } \
{ memd(R0 + #_PT_R0706) = R7:6; \
memd(R0 + #_PT_R0504) = R5:4; \
R13:12 = C3:2; } \
{ memd(R0 + #_PT_GPUGP) = R31:30; \
memd(R0 + #_PT_LC0SA0) = R15:14; \
R11:10 = C7:6; }\
{ THREADINFO_REG = and(R0, # ## #-_THREAD_SIZE); \
memd(R0 + #_PT_LC1SA1) = R13:12; \
R15 = p3:0; }\
{ memd(R0 + #_PT_M1M0) = R11:10; \
memw(R0 + #_PT_PREDSUSR + 4) = R15; }\
{ r24 = memw(THREADINFO_REG + #_THREAD_INFO_PT_REGS); \
memw(THREADINFO_REG + #_THREAD_INFO_PT_REGS) = R0; \
R2 = #-1; } \
{ memw(R0 + #_PT_SYSCALL_NR) = R2; \
memd(R0 + #_PT_CS1CS0) = R17:16; \
R30 = #0; }
#endif
/*
* Restore registers and thread_info.regs state. THREADINFO_REG
* is assumed to still be sane, and R24 to have been correctly
* preserved. Don't restore R29 (SP) until later.
*/
#if CONFIG_HEXAGON_ARCH_VERSION < 4
#define restore_pt_regs() \
{ memw(THREADINFO_REG + #_THREAD_INFO_PT_REGS) = R24; \
R15:14 = memd(R0 + #_PT_PREDSUSR); } \
{ R11:10 = memd(R0 + #_PT_M1M0); \
p3:0 = R15; } \
{ R13:12 = memd(R0 + #_PT_LC1SA1); \
usr = R14; } \
{ R15:14 = memd(R0 + #_PT_LC0SA0); \
m1 = R11; } \
{ R3:2 = memd(R0 + #_PT_R0302); \
m0 = R10; } \
{ R5:4 = memd(R0 + #_PT_R0504); \
lc1 = R13; } \
{ R7:6 = memd(R0 + #_PT_R0706); \
sa1 = R12; } \
{ R9:8 = memd(R0 + #_PT_R0908); \
lc0 = R15; } \
{ R11:10 = memd(R0 + #_PT_R1110); \
sa0 = R14; } \
{ R13:12 = memd(R0 + #_PT_R1312); \
R15:14 = memd(R0 + #_PT_R1514); } \
{ R17:16 = memd(R0 + #_PT_R1716); \
R19:18 = memd(R0 + #_PT_R1918); } \
{ R21:20 = memd(R0 + #_PT_R2120); \
R23:22 = memd(R0 + #_PT_R2322); } \
{ R25:24 = memd(R0 + #_PT_R2524); \
R27:26 = memd(R0 + #_PT_R2726); } \
R31:30 = memd(R0 + #_PT_GPUGP); \
{ R28 = memw(R0 + #_PT_R2928); \
ugp = R31; } \
{ R31:30 = memd(R0 + #_PT_R3130); \
gp = R30; }
#else
/* V4+ */
#define restore_pt_regs() \
{ memw(THREADINFO_REG + #_THREAD_INFO_PT_REGS) = R24; \
R15:14 = memd(R0 + #_PT_PREDSUSR); } \
{ R11:10 = memd(R0 + #_PT_M1M0); \
R13:12 = memd(R0 + #_PT_LC1SA1); \
p3:0 = R15; } \
{ R15:14 = memd(R0 + #_PT_LC0SA0); \
R3:2 = memd(R0 + #_PT_R0302); \
usr = R14; } \
{ R5:4 = memd(R0 + #_PT_R0504); \
R7:6 = memd(R0 + #_PT_R0706); \
C7:6 = R11:10; }\
{ R9:8 = memd(R0 + #_PT_R0908); \
R11:10 = memd(R0 + #_PT_R1110); \
C3:2 = R13:12; }\
{ R13:12 = memd(R0 + #_PT_R1312); \
R15:14 = memd(R0 + #_PT_R1514); \
C1:0 = R15:14; }\
{ R17:16 = memd(R0 + #_PT_R1716); \
R19:18 = memd(R0 + #_PT_R1918); } \
{ R21:20 = memd(R0 + #_PT_R2120); \
R23:22 = memd(R0 + #_PT_R2322); } \
{ R25:24 = memd(R0 + #_PT_R2524); \
R27:26 = memd(R0 + #_PT_R2726); } \
R31:30 = memd(R0 + #_PT_CS1CS0); \
{ C13:12 = R31:30; \
R31:30 = memd(R0 + #_PT_GPUGP) ; \
R28 = memw(R0 + #_PT_R2928); }\
{ C11:10 = R31:30; \
R31:30 = memd(R0 + #_PT_R3130); }
#endif
/*
* Clears off enough space for the rest of pt_regs; evrec is a part
* of pt_regs in HVM mode. Save R0/R1, set handler's address in R1.
* R0 is the address of pt_regs and is the parameter to save_pt_regs.
*/
/*
* Since the HVM isn't automagically pushing the EVREC onto the stack anymore,
* we'll subract the entire size out and then fill it in ourselves.
* Need to save off R0, R1, R2, R3 immediately.
*/
#if CONFIG_HEXAGON_ARCH_VERSION < 4
#define vm_event_entry(CHandler) \
{ \
R29 = add(R29, #-(_PT_REGS_SIZE)); \
memd(R29 + #(_PT_R0100 + -_PT_REGS_SIZE)) = R1:0; \
} \
{ \
memd(R29 +#_PT_R0302) = R3:2; \
} \
trap1(#HVM_TRAP1_VMGETREGS); \
{ \
memd(R29 + #_PT_ER_VMEL) = R1:0; \
R0 = R29; \
R1.L = #LO(CHandler); \
} \
{ \
memd(R29 + #_PT_ER_VMPSP) = R3:2; \
R1.H = #HI(CHandler); \
jump event_dispatch; \
}
#else
/* V4+ */
/* turn on I$ prefetch early */
/* the # ## # syntax inserts a literal ## */
#define vm_event_entry(CHandler) \
{ \
R29 = add(R29, #-(_PT_REGS_SIZE)); \
memd(R29 + #(_PT_R0100 + -_PT_REGS_SIZE)) = R1:0; \
memd(R29 + #(_PT_R0302 + -_PT_REGS_SIZE)) = R3:2; \
R0 = usr; \
} \
{ \
memw(R29 + #_PT_PREDSUSR) = R0; \
R0 = setbit(R0, #16); \
} \
usr = R0; \
R1:0 = G1:0; \
{ \
memd(R29 + #_PT_ER_VMEL) = R1:0; \
R1 = # ## #(CHandler); \
R3:2 = G3:2; \
} \
{ \
R0 = R29; \
memd(R29 + #_PT_ER_VMPSP) = R3:2; \
jump event_dispatch; \
}
#endif
.text
/*
* Do bulk save/restore in one place.
* Adds a jump to dispatch latency, but
* saves hundreds of bytes.
*/
event_dispatch:
save_pt_regs()
callr r1
/*
* Coming back from the C-world, our thread info pointer
* should be in the designated register (usually R19)
*
* If we were in kernel mode, we don't need to check scheduler
* or signals if CONFIG_PREEMPTION is not set. If set, then it has
* to jump to a need_resched kind of block.
* BTW, CONFIG_PREEMPTION is not supported yet.
*/
#ifdef CONFIG_PREEMPTION
R0 = #VM_INT_DISABLE
trap1(#HVM_TRAP1_VMSETIE)
#endif
/* "Nested control path" -- if the previous mode was kernel */
{
R0 = memw(R29 + #_PT_ER_VMEST);
R26.L = #LO(do_work_pending);
}
{
P0 = tstbit(R0, #HVM_VMEST_UM_SFT);
if (!P0.new) jump:nt restore_all;
R26.H = #HI(do_work_pending);
R0 = #VM_INT_DISABLE;
}
/*
* Check also the return from fork/system call, normally coming back from
* user mode
*
* R26 needs to have do_work_pending, and R0 should have VM_INT_DISABLE
*/
check_work_pending:
/* Disable interrupts while checking TIF */
trap1(#HVM_TRAP1_VMSETIE)
{
R0 = R29; /* regs should still be at top of stack */
R1 = memw(THREADINFO_REG + #_THREAD_INFO_FLAGS);
callr R26;
}
{
P0 = cmp.eq(R0, #0); if (!P0.new) jump:nt check_work_pending;
R0 = #VM_INT_DISABLE;
}
restore_all:
/*
* Disable interrupts, if they weren't already, before reg restore.
* R0 gets preloaded with #VM_INT_DISABLE before we get here.
*/
trap1(#HVM_TRAP1_VMSETIE)
/* do the setregs here for VM 0.5 */
/* R29 here should already be pointing at pt_regs */
{
R1:0 = memd(R29 + #_PT_ER_VMEL);
R3:2 = memd(R29 + #_PT_ER_VMPSP);
}
#if CONFIG_HEXAGON_ARCH_VERSION < 4
trap1(#HVM_TRAP1_VMSETREGS);
#else
G1:0 = R1:0;
G3:2 = R3:2;
#endif
R0 = R29
restore_pt_regs()
{
R1:0 = memd(R29 + #_PT_R0100);
R29 = add(R29, #_PT_REGS_SIZE);
}
trap1(#HVM_TRAP1_VMRTE)
/* Notreached */
.globl _K_enter_genex
_K_enter_genex:
vm_event_entry(do_genex)
.globl _K_enter_interrupt
_K_enter_interrupt:
vm_event_entry(arch_do_IRQ)
.globl _K_enter_trap0
_K_enter_trap0:
vm_event_entry(do_trap0)
.globl _K_enter_machcheck
_K_enter_machcheck:
vm_event_entry(do_machcheck)
.globl _K_enter_debug
_K_enter_debug:
vm_event_entry(do_debug_exception)
.globl ret_from_fork
ret_from_fork:
{
call schedule_tail
R26.H = #HI(do_work_pending);
}
{
P0 = cmp.eq(R24, #0);
R26.L = #LO(do_work_pending);
R0 = #VM_INT_DISABLE;
}
if (P0) jump check_work_pending
{
R0 = R25;
callr R24
}
{
jump check_work_pending
R0 = #VM_INT_DISABLE;
}
|