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
|
/*
** Machine code management.
** Copyright (C) 2005-2015 Mike Pall. See Copyright Notice in luajit.h
*/
#define lj_mcode_c
#define LUA_CORE
#include "lj_obj.h"
#if LJ_HASJIT
#include "lj_gc.h"
#include "lj_err.h"
#include "lj_jit.h"
#include "lj_mcode.h"
#include "lj_trace.h"
#include "lj_dispatch.h"
#endif
#if LJ_HASJIT || LJ_HASFFI
#include "lj_vm.h"
#endif
/* -- OS-specific functions ----------------------------------------------- */
#if LJ_HASJIT || LJ_HASFFI
/* Define this if you want to run LuaJIT with Valgrind. */
#ifdef LUAJIT_USE_VALGRIND
#include <valgrind/valgrind.h>
#endif
#if LJ_TARGET_IOS
void sys_icache_invalidate(void *start, size_t len);
#endif
/* Synchronize data/instruction cache. */
void lj_mcode_sync(void *start, void *end)
{
#ifdef LUAJIT_USE_VALGRIND
VALGRIND_DISCARD_TRANSLATIONS(start, (char *)end-(char *)start);
#endif
#if LJ_TARGET_X86ORX64
UNUSED(start); UNUSED(end);
#elif LJ_TARGET_IOS
sys_icache_invalidate(start, (char *)end-(char *)start);
#elif LJ_TARGET_PPC
lj_vm_cachesync(start, end);
#elif defined(__GNUC__)
__clear_cache(start, end);
#else
#error "Missing builtin to flush instruction cache"
#endif
}
#endif
#if LJ_HASJIT
#if LJ_TARGET_WINDOWS
#define WIN32_LEAN_AND_MEAN
#include <windows.h>
#define MCPROT_RW PAGE_READWRITE
#define MCPROT_RX PAGE_EXECUTE_READ
#define MCPROT_RWX PAGE_EXECUTE_READWRITE
static void *mcode_alloc_at(jit_State *J, uintptr_t hint, size_t sz, DWORD prot)
{
void *p = VirtualAlloc((void *)hint, sz,
MEM_RESERVE|MEM_COMMIT|MEM_TOP_DOWN, prot);
if (!p && !hint)
lj_trace_err(J, LJ_TRERR_MCODEAL);
return p;
}
static void mcode_free(jit_State *J, void *p, size_t sz)
{
UNUSED(J); UNUSED(sz);
VirtualFree(p, 0, MEM_RELEASE);
}
static int mcode_setprot(void *p, size_t sz, DWORD prot)
{
DWORD oprot;
return !VirtualProtect(p, sz, prot, &oprot);
}
#elif LJ_TARGET_POSIX
#include <sys/mman.h>
#ifndef MAP_ANONYMOUS
#define MAP_ANONYMOUS MAP_ANON
#endif
#define MCPROT_RW (PROT_READ|PROT_WRITE)
#define MCPROT_RX (PROT_READ|PROT_EXEC)
#define MCPROT_RWX (PROT_READ|PROT_WRITE|PROT_EXEC)
static void *mcode_alloc_at(jit_State *J, uintptr_t hint, size_t sz, int prot)
{
void *p = mmap((void *)hint, sz, prot, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0);
if (p == MAP_FAILED) {
if (!hint) lj_trace_err(J, LJ_TRERR_MCODEAL);
p = NULL;
}
return p;
}
static void mcode_free(jit_State *J, void *p, size_t sz)
{
UNUSED(J);
munmap(p, sz);
}
static int mcode_setprot(void *p, size_t sz, int prot)
{
return mprotect(p, sz, prot);
}
#elif LJ_64
#error "Missing OS support for explicit placement of executable memory"
#else
/* Fallback allocator. This will fail if memory is not executable by default. */
#define LUAJIT_UNPROTECT_MCODE
#define MCPROT_RW 0
#define MCPROT_RX 0
#define MCPROT_RWX 0
static void *mcode_alloc_at(jit_State *J, uintptr_t hint, size_t sz, int prot)
{
UNUSED(hint); UNUSED(prot);
return lj_mem_new(J->L, sz);
}
static void mcode_free(jit_State *J, void *p, size_t sz)
{
lj_mem_free(J2G(J), p, sz);
}
#endif
/* -- MCode area protection ----------------------------------------------- */
/* Define this ONLY if page protection twiddling becomes a bottleneck. */
#ifdef LUAJIT_UNPROTECT_MCODE
/* It's generally considered to be a potential security risk to have
** pages with simultaneous write *and* execute access in a process.
**
** Do not even think about using this mode for server processes or
** apps handling untrusted external data (such as a browser).
**
** The security risk is not in LuaJIT itself -- but if an adversary finds
** any *other* flaw in your C application logic, then any RWX memory page
** simplifies writing an exploit considerably.
*/
#define MCPROT_GEN MCPROT_RWX
#define MCPROT_RUN MCPROT_RWX
static void mcode_protect(jit_State *J, int prot)
{
UNUSED(J); UNUSED(prot);
}
#else
/* This is the default behaviour and much safer:
**
** Most of the time the memory pages holding machine code are executable,
** but NONE of them is writable.
**
** The current memory area is marked read-write (but NOT executable) only
** during the short time window while the assembler generates machine code.
*/
#define MCPROT_GEN MCPROT_RW
#define MCPROT_RUN MCPROT_RX
/* Protection twiddling failed. Probably due to kernel security. */
static LJ_NOINLINE void mcode_protfail(jit_State *J)
{
lua_CFunction panic = J2G(J)->panic;
if (panic) {
lua_State *L = J->L;
setstrV(L, L->top++, lj_err_str(L, LJ_ERR_JITPROT));
panic(L);
}
}
/* Change protection of MCode area. */
static void mcode_protect(jit_State *J, int prot)
{
if (J->mcprot != prot) {
if (LJ_UNLIKELY(mcode_setprot(J->mcarea, J->szmcarea, prot)))
mcode_protfail(J);
J->mcprot = prot;
}
}
#endif
/* -- MCode area allocation ----------------------------------------------- */
#if LJ_TARGET_X64
#define mcode_validptr(p) ((p) && (uintptr_t)(p) < (uintptr_t)1<<47)
#else
#define mcode_validptr(p) ((p) && (uintptr_t)(p) < 0xffff0000)
#endif
#ifdef LJ_TARGET_JUMPRANGE
/* Get memory within relative jump distance of our code in 64 bit mode. */
static void *mcode_alloc(jit_State *J, size_t sz)
{
/* Target an address in the static assembler code (64K aligned).
** Try addresses within a distance of target-range/2+1MB..target+range/2-1MB.
** Use half the jump range so every address in the range can reach any other.
*/
#if LJ_TARGET_MIPS
/* Use the middle of the 256MB-aligned region. */
uintptr_t target = ((uintptr_t)(void *)lj_vm_exit_handler & 0xf0000000u) +
0x08000000u;
#else
uintptr_t target = (uintptr_t)(void *)lj_vm_exit_handler & ~(uintptr_t)0xffff;
#endif
const uintptr_t range = (1u << (LJ_TARGET_JUMPRANGE-1)) - (1u << 21);
/* First try a contiguous area below the last one. */
uintptr_t hint = J->mcarea ? (uintptr_t)J->mcarea - sz : 0;
int i;
for (i = 0; i < 32; i++) { /* 32 attempts ought to be enough ... */
if (mcode_validptr(hint)) {
void *p = mcode_alloc_at(J, hint, sz, MCPROT_GEN);
if (mcode_validptr(p) &&
((uintptr_t)p + sz - target < range || target - (uintptr_t)p < range))
return p;
if (p) mcode_free(J, p, sz); /* Free badly placed area. */
}
/* Next try probing pseudo-random addresses. */
do {
hint = (0x78fb ^ LJ_PRNG_BITS(J, 15)) << 16; /* 64K aligned. */
} while (!(hint + sz < range));
hint = target + hint - (range>>1);
}
lj_trace_err(J, LJ_TRERR_MCODEAL); /* Give up. OS probably ignores hints? */
return NULL;
}
#else
/* All memory addresses are reachable by relative jumps. */
static void *mcode_alloc(jit_State *J, size_t sz)
{
#ifdef __OpenBSD__
/* Allow better executable memory allocation for OpenBSD W^X mode. */
void *p = mcode_alloc_at(J, 0, sz, MCPROT_RUN);
if (p && mcode_setprot(p, sz, MCPROT_GEN)) {
mcode_free(J, p, sz);
return NULL;
}
return p;
#else
return mcode_alloc_at(J, 0, sz, MCPROT_GEN);
#endif
}
#endif
/* -- MCode area management ----------------------------------------------- */
/* Linked list of MCode areas. */
typedef struct MCLink {
MCode *next; /* Next area. */
size_t size; /* Size of current area. */
} MCLink;
/* Allocate a new MCode area. */
static void mcode_allocarea(jit_State *J)
{
MCode *oldarea = J->mcarea;
size_t sz = (size_t)J->param[JIT_P_sizemcode] << 10;
sz = (sz + LJ_PAGESIZE-1) & ~(size_t)(LJ_PAGESIZE - 1);
J->mcarea = (MCode *)mcode_alloc(J, sz);
J->szmcarea = sz;
J->mcprot = MCPROT_GEN;
J->mctop = (MCode *)((char *)J->mcarea + J->szmcarea);
J->mcbot = (MCode *)((char *)J->mcarea + sizeof(MCLink));
((MCLink *)J->mcarea)->next = oldarea;
((MCLink *)J->mcarea)->size = sz;
J->szallmcarea += sz;
}
/* Free all MCode areas. */
void lj_mcode_free(jit_State *J)
{
MCode *mc = J->mcarea;
J->mcarea = NULL;
J->szallmcarea = 0;
while (mc) {
MCode *next = ((MCLink *)mc)->next;
mcode_free(J, mc, ((MCLink *)mc)->size);
mc = next;
}
}
/* -- MCode transactions -------------------------------------------------- */
/* Reserve the remainder of the current MCode area. */
MCode *lj_mcode_reserve(jit_State *J, MCode **lim)
{
if (!J->mcarea)
mcode_allocarea(J);
else
mcode_protect(J, MCPROT_GEN);
*lim = J->mcbot;
return J->mctop;
}
/* Commit the top part of the current MCode area. */
void lj_mcode_commit(jit_State *J, MCode *top)
{
J->mctop = top;
mcode_protect(J, MCPROT_RUN);
}
/* Abort the reservation. */
void lj_mcode_abort(jit_State *J)
{
if (J->mcarea)
mcode_protect(J, MCPROT_RUN);
}
/* Set/reset protection to allow patching of MCode areas. */
MCode *lj_mcode_patch(jit_State *J, MCode *ptr, int finish)
{
#ifdef LUAJIT_UNPROTECT_MCODE
UNUSED(J); UNUSED(ptr); UNUSED(finish);
return NULL;
#else
if (finish) {
if (J->mcarea == ptr)
mcode_protect(J, MCPROT_RUN);
else if (LJ_UNLIKELY(mcode_setprot(ptr, ((MCLink *)ptr)->size, MCPROT_RUN)))
mcode_protfail(J);
return NULL;
} else {
MCode *mc = J->mcarea;
/* Try current area first to use the protection cache. */
if (ptr >= mc && ptr < (MCode *)((char *)mc + J->szmcarea)) {
mcode_protect(J, MCPROT_GEN);
return mc;
}
/* Otherwise search through the list of MCode areas. */
for (;;) {
mc = ((MCLink *)mc)->next;
lua_assert(mc != NULL);
if (ptr >= mc && ptr < (MCode *)((char *)mc + ((MCLink *)mc)->size)) {
if (LJ_UNLIKELY(mcode_setprot(mc, ((MCLink *)mc)->size, MCPROT_GEN)))
mcode_protfail(J);
return mc;
}
}
}
#endif
}
/* Limit of MCode reservation reached. */
void lj_mcode_limiterr(jit_State *J, size_t need)
{
size_t sizemcode, maxmcode;
lj_mcode_abort(J);
sizemcode = (size_t)J->param[JIT_P_sizemcode] << 10;
sizemcode = (sizemcode + LJ_PAGESIZE-1) & ~(size_t)(LJ_PAGESIZE - 1);
maxmcode = (size_t)J->param[JIT_P_maxmcode] << 10;
if ((size_t)need > sizemcode)
lj_trace_err(J, LJ_TRERR_MCODEOV); /* Too long for any area. */
if (J->szallmcarea + sizemcode > maxmcode)
lj_trace_err(J, LJ_TRERR_MCODEAL);
mcode_allocarea(J);
lj_trace_err(J, LJ_TRERR_MCODELM); /* Retry with new area. */
}
#endif
|