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
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
|
/* Copyright (C) 2002-2021 Free Software Foundation, Inc.
This file is part of the GNU C Library.
Contributed by Ulrich Drepper <drepper@redhat.com>, 2002.
The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
The GNU C Library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with the GNU C Library; if not, see
<https://www.gnu.org/licenses/>. */
#include <assert.h>
#include <errno.h>
#include <signal.h>
#include <stdint.h>
#include <string.h>
#include <unistd.h>
#include <sys/mman.h>
#include <sys/param.h>
#include <dl-sysdep.h>
#include <dl-tls.h>
#include <tls.h>
#include <list.h>
#include <lowlevellock.h>
#include <futex-internal.h>
#include <kernel-features.h>
#ifndef NEED_SEPARATE_REGISTER_STACK
/* Most architectures have exactly one stack pointer. Some have more. */
# define STACK_VARIABLES void *stackaddr = NULL
/* How to pass the values to the 'create_thread' function. */
# define STACK_VARIABLES_ARGS stackaddr
/* How to declare function which gets there parameters. */
# define STACK_VARIABLES_PARMS void *stackaddr
/* How to declare allocate_stack. */
# define ALLOCATE_STACK_PARMS void **stack
/* This is how the function is called. We do it this way to allow
other variants of the function to have more parameters. */
# define ALLOCATE_STACK(attr, pd) allocate_stack (attr, pd, &stackaddr)
#else
/* We need two stacks. The kernel will place them but we have to tell
the kernel about the size of the reserved address space. */
# define STACK_VARIABLES void *stackaddr = NULL; size_t stacksize = 0
/* How to pass the values to the 'create_thread' function. */
# define STACK_VARIABLES_ARGS stackaddr, stacksize
/* How to declare function which gets there parameters. */
# define STACK_VARIABLES_PARMS void *stackaddr, size_t stacksize
/* How to declare allocate_stack. */
# define ALLOCATE_STACK_PARMS void **stack, size_t *stacksize
/* This is how the function is called. We do it this way to allow
other variants of the function to have more parameters. */
# define ALLOCATE_STACK(attr, pd) \
allocate_stack (attr, pd, &stackaddr, &stacksize)
#endif
/* Default alignment of stack. */
#ifndef STACK_ALIGN
# define STACK_ALIGN __alignof__ (long double)
#endif
/* Default value for minimal stack size after allocating thread
descriptor and guard. */
#ifndef MINIMAL_REST_STACK
# define MINIMAL_REST_STACK 4096
#endif
/* Newer kernels have the MAP_STACK flag to indicate a mapping is used for
a stack. Use it when possible. */
#ifndef MAP_STACK
# define MAP_STACK 0
#endif
/* This yields the pointer that TLS support code calls the thread pointer. */
#if TLS_TCB_AT_TP
# define TLS_TPADJ(pd) (pd)
#elif TLS_DTV_AT_TP
# define TLS_TPADJ(pd) ((struct pthread *)((char *) (pd) + TLS_PRE_TCB_SIZE))
#endif
/* Cache handling for not-yet free stacks. */
/* Maximum size in kB of cache. */
static size_t stack_cache_maxsize = 40 * 1024 * 1024; /* 40MiBi by default. */
/* Check whether the stack is still used or not. */
#define FREE_P(descr) ((descr)->tid <= 0)
static void
stack_list_del (list_t *elem)
{
GL (dl_in_flight_stack) = (uintptr_t) elem;
atomic_write_barrier ();
list_del (elem);
atomic_write_barrier ();
GL (dl_in_flight_stack) = 0;
}
static void
stack_list_add (list_t *elem, list_t *list)
{
GL (dl_in_flight_stack) = (uintptr_t) elem | 1;
atomic_write_barrier ();
list_add (elem, list);
atomic_write_barrier ();
GL (dl_in_flight_stack) = 0;
}
/* We create a double linked list of all cache entries. Double linked
because this allows removing entries from the end. */
/* Get a stack frame from the cache. We have to match by size since
some blocks might be too small or far too large. */
static struct pthread *
get_cached_stack (size_t *sizep, void **memp)
{
size_t size = *sizep;
struct pthread *result = NULL;
list_t *entry;
lll_lock (GL (dl_stack_cache_lock), LLL_PRIVATE);
/* Search the cache for a matching entry. We search for the
smallest stack which has at least the required size. Note that
in normal situations the size of all allocated stacks is the
same. As the very least there are only a few different sizes.
Therefore this loop will exit early most of the time with an
exact match. */
list_for_each (entry, &GL (dl_stack_cache))
{
struct pthread *curr;
curr = list_entry (entry, struct pthread, list);
if (FREE_P (curr) && curr->stackblock_size >= size)
{
if (curr->stackblock_size == size)
{
result = curr;
break;
}
if (result == NULL
|| result->stackblock_size > curr->stackblock_size)
result = curr;
}
}
if (__builtin_expect (result == NULL, 0)
/* Make sure the size difference is not too excessive. In that
case we do not use the block. */
|| __builtin_expect (result->stackblock_size > 4 * size, 0))
{
/* Release the lock. */
lll_unlock (GL (dl_stack_cache_lock), LLL_PRIVATE);
return NULL;
}
/* Don't allow setxid until cloned. */
result->setxid_futex = -1;
/* Dequeue the entry. */
stack_list_del (&result->list);
/* And add to the list of stacks in use. */
stack_list_add (&result->list, &GL (dl_stack_used));
/* And decrease the cache size. */
GL (dl_stack_cache_actsize) -= result->stackblock_size;
/* Release the lock early. */
lll_unlock (GL (dl_stack_cache_lock), LLL_PRIVATE);
/* Report size and location of the stack to the caller. */
*sizep = result->stackblock_size;
*memp = result->stackblock;
/* Cancellation handling is back to the default. */
result->cancelhandling = 0;
result->cleanup = NULL;
/* No pending event. */
result->nextevent = NULL;
result->tls_state = (struct tls_internal_t) { 0 };
/* Clear the DTV. */
dtv_t *dtv = GET_DTV (TLS_TPADJ (result));
for (size_t cnt = 0; cnt < dtv[-1].counter; ++cnt)
free (dtv[1 + cnt].pointer.to_free);
memset (dtv, '\0', (dtv[-1].counter + 1) * sizeof (dtv_t));
/* Re-initialize the TLS. */
_dl_allocate_tls_init (TLS_TPADJ (result));
return result;
}
/* Free stacks until cache size is lower than LIMIT. */
static void
free_stacks (size_t limit)
{
/* We reduce the size of the cache. Remove the last entries until
the size is below the limit. */
list_t *entry;
list_t *prev;
/* Search from the end of the list. */
list_for_each_prev_safe (entry, prev, &GL (dl_stack_cache))
{
struct pthread *curr;
curr = list_entry (entry, struct pthread, list);
if (FREE_P (curr))
{
/* Unlink the block. */
stack_list_del (entry);
/* Account for the freed memory. */
GL (dl_stack_cache_actsize) -= curr->stackblock_size;
/* Free the memory associated with the ELF TLS. */
_dl_deallocate_tls (TLS_TPADJ (curr), false);
/* Remove this block. This should never fail. If it does
something is really wrong. */
if (__munmap (curr->stackblock, curr->stackblock_size) != 0)
abort ();
/* Maybe we have freed enough. */
if (GL (dl_stack_cache_actsize) <= limit)
break;
}
}
}
/* Free all the stacks on cleanup. */
void
__nptl_stacks_freeres (void)
{
free_stacks (0);
}
/* Add a stack frame which is not used anymore to the stack. Must be
called with the cache lock held. */
static inline void
__attribute ((always_inline))
queue_stack (struct pthread *stack)
{
/* We unconditionally add the stack to the list. The memory may
still be in use but it will not be reused until the kernel marks
the stack as not used anymore. */
stack_list_add (&stack->list, &GL (dl_stack_cache));
GL (dl_stack_cache_actsize) += stack->stackblock_size;
if (__glibc_unlikely (GL (dl_stack_cache_actsize) > stack_cache_maxsize))
free_stacks (stack_cache_maxsize);
}
/* Return the guard page position on allocated stack. */
static inline char *
__attribute ((always_inline))
guard_position (void *mem, size_t size, size_t guardsize, struct pthread *pd,
size_t pagesize_m1)
{
#ifdef NEED_SEPARATE_REGISTER_STACK
return mem + (((size - guardsize) / 2) & ~pagesize_m1);
#elif _STACK_GROWS_DOWN
return mem;
#elif _STACK_GROWS_UP
return (char *) (((uintptr_t) pd - guardsize) & ~pagesize_m1);
#endif
}
/* Based on stack allocated with PROT_NONE, setup the required portions with
'prot' flags based on the guard page position. */
static inline int
setup_stack_prot (char *mem, size_t size, char *guard, size_t guardsize,
const int prot)
{
char *guardend = guard + guardsize;
#if _STACK_GROWS_DOWN && !defined(NEED_SEPARATE_REGISTER_STACK)
/* As defined at guard_position, for architectures with downward stack
the guard page is always at start of the allocated area. */
if (__mprotect (guardend, size - guardsize, prot) != 0)
return errno;
#else
size_t mprots1 = (uintptr_t) guard - (uintptr_t) mem;
if (__mprotect (mem, mprots1, prot) != 0)
return errno;
size_t mprots2 = ((uintptr_t) mem + size) - (uintptr_t) guardend;
if (__mprotect (guardend, mprots2, prot) != 0)
return errno;
#endif
return 0;
}
/* Mark the memory of the stack as usable to the kernel. It frees everything
except for the space used for the TCB itself. */
static __always_inline void
advise_stack_range (void *mem, size_t size, uintptr_t pd, size_t guardsize)
{
uintptr_t sp = (uintptr_t) CURRENT_STACK_FRAME;
size_t pagesize_m1 = __getpagesize () - 1;
#if _STACK_GROWS_DOWN && !defined(NEED_SEPARATE_REGISTER_STACK)
size_t freesize = (sp - (uintptr_t) mem) & ~pagesize_m1;
assert (freesize < size);
if (freesize > PTHREAD_STACK_MIN)
__madvise (mem, freesize - PTHREAD_STACK_MIN, MADV_DONTNEED);
#else
/* Page aligned start of memory to free (higher than or equal
to current sp plus the minimum stack size). */
uintptr_t freeblock = (sp + PTHREAD_STACK_MIN + pagesize_m1) & ~pagesize_m1;
uintptr_t free_end = (pd - guardsize) & ~pagesize_m1;
if (free_end > freeblock)
{
size_t freesize = free_end - freeblock;
assert (freesize < size);
__madvise ((void*) freeblock, freesize, MADV_DONTNEED);
}
#endif
}
/* Returns a usable stack for a new thread either by allocating a
new stack or reusing a cached stack of sufficient size.
ATTR must be non-NULL and point to a valid pthread_attr.
PDP must be non-NULL. */
static int
allocate_stack (const struct pthread_attr *attr, struct pthread **pdp,
ALLOCATE_STACK_PARMS)
{
struct pthread *pd;
size_t size;
size_t pagesize_m1 = __getpagesize () - 1;
assert (powerof2 (pagesize_m1 + 1));
assert (TCB_ALIGNMENT >= STACK_ALIGN);
/* Get the stack size from the attribute if it is set. Otherwise we
use the default we determined at start time. */
if (attr->stacksize != 0)
size = attr->stacksize;
else
{
lll_lock (__default_pthread_attr_lock, LLL_PRIVATE);
size = __default_pthread_attr.internal.stacksize;
lll_unlock (__default_pthread_attr_lock, LLL_PRIVATE);
}
/* Get memory for the stack. */
if (__glibc_unlikely (attr->flags & ATTR_FLAG_STACKADDR))
{
uintptr_t adj;
char *stackaddr = (char *) attr->stackaddr;
/* Assume the same layout as the _STACK_GROWS_DOWN case, with struct
pthread at the top of the stack block. Later we adjust the guard
location and stack address to match the _STACK_GROWS_UP case. */
if (_STACK_GROWS_UP)
stackaddr += attr->stacksize;
/* If the user also specified the size of the stack make sure it
is large enough. */
if (attr->stacksize != 0
&& attr->stacksize < (__static_tls_size + MINIMAL_REST_STACK))
return EINVAL;
/* Adjust stack size for alignment of the TLS block. */
#if TLS_TCB_AT_TP
adj = ((uintptr_t) stackaddr - TLS_TCB_SIZE)
& __static_tls_align_m1;
assert (size > adj + TLS_TCB_SIZE);
#elif TLS_DTV_AT_TP
adj = ((uintptr_t) stackaddr - __static_tls_size)
& __static_tls_align_m1;
assert (size > adj);
#endif
/* The user provided some memory. Let's hope it matches the
size... We do not allocate guard pages if the user provided
the stack. It is the user's responsibility to do this if it
is wanted. */
#if TLS_TCB_AT_TP
pd = (struct pthread *) ((uintptr_t) stackaddr
- TLS_TCB_SIZE - adj);
#elif TLS_DTV_AT_TP
pd = (struct pthread *) (((uintptr_t) stackaddr
- __static_tls_size - adj)
- TLS_PRE_TCB_SIZE);
#endif
/* The user provided stack memory needs to be cleared. */
memset (pd, '\0', sizeof (struct pthread));
/* The first TSD block is included in the TCB. */
pd->specific[0] = pd->specific_1stblock;
/* Remember the stack-related values. */
pd->stackblock = (char *) stackaddr - size;
pd->stackblock_size = size;
/* This is a user-provided stack. It will not be queued in the
stack cache nor will the memory (except the TLS memory) be freed. */
pd->user_stack = true;
/* This is at least the second thread. */
pd->header.multiple_threads = 1;
#ifndef TLS_MULTIPLE_THREADS_IN_TCB
__libc_multiple_threads = 1;
#endif
#ifdef NEED_DL_SYSINFO
SETUP_THREAD_SYSINFO (pd);
#endif
/* Don't allow setxid until cloned. */
pd->setxid_futex = -1;
/* Allocate the DTV for this thread. */
if (_dl_allocate_tls (TLS_TPADJ (pd)) == NULL)
{
/* Something went wrong. */
assert (errno == ENOMEM);
return errno;
}
/* Prepare to modify global data. */
lll_lock (GL (dl_stack_cache_lock), LLL_PRIVATE);
/* And add to the list of stacks in use. */
list_add (&pd->list, &GL (dl_stack_user));
lll_unlock (GL (dl_stack_cache_lock), LLL_PRIVATE);
}
else
{
/* Allocate some anonymous memory. If possible use the cache. */
size_t guardsize;
size_t reported_guardsize;
size_t reqsize;
void *mem;
const int prot = (PROT_READ | PROT_WRITE
| ((GL(dl_stack_flags) & PF_X) ? PROT_EXEC : 0));
/* Adjust the stack size for alignment. */
size &= ~__static_tls_align_m1;
assert (size != 0);
/* Make sure the size of the stack is enough for the guard and
eventually the thread descriptor. On some targets there is
a minimum guard size requirement, ARCH_MIN_GUARD_SIZE, so
internally enforce it (unless the guard was disabled), but
report the original guard size for backward compatibility:
before POSIX 2008 the guardsize was specified to be one page
by default which is observable via pthread_attr_getguardsize
and pthread_getattr_np. */
guardsize = (attr->guardsize + pagesize_m1) & ~pagesize_m1;
reported_guardsize = guardsize;
if (guardsize > 0 && guardsize < ARCH_MIN_GUARD_SIZE)
guardsize = ARCH_MIN_GUARD_SIZE;
if (guardsize < attr->guardsize || size + guardsize < guardsize)
/* Arithmetic overflow. */
return EINVAL;
size += guardsize;
if (__builtin_expect (size < ((guardsize + __static_tls_size
+ MINIMAL_REST_STACK + pagesize_m1)
& ~pagesize_m1),
0))
/* The stack is too small (or the guard too large). */
return EINVAL;
/* Try to get a stack from the cache. */
reqsize = size;
pd = get_cached_stack (&size, &mem);
if (pd == NULL)
{
/* If a guard page is required, avoid committing memory by first
allocate with PROT_NONE and then reserve with required permission
excluding the guard page. */
mem = __mmap (NULL, size, (guardsize == 0) ? prot : PROT_NONE,
MAP_PRIVATE | MAP_ANONYMOUS | MAP_STACK, -1, 0);
if (__glibc_unlikely (mem == MAP_FAILED))
return errno;
/* SIZE is guaranteed to be greater than zero.
So we can never get a null pointer back from mmap. */
assert (mem != NULL);
/* Place the thread descriptor at the end of the stack. */
#if TLS_TCB_AT_TP
pd = (struct pthread *) ((((uintptr_t) mem + size)
- TLS_TCB_SIZE)
& ~__static_tls_align_m1);
#elif TLS_DTV_AT_TP
pd = (struct pthread *) ((((uintptr_t) mem + size
- __static_tls_size)
& ~__static_tls_align_m1)
- TLS_PRE_TCB_SIZE);
#endif
/* Now mprotect the required region excluding the guard area. */
if (__glibc_likely (guardsize > 0))
{
char *guard = guard_position (mem, size, guardsize, pd,
pagesize_m1);
if (setup_stack_prot (mem, size, guard, guardsize, prot) != 0)
{
__munmap (mem, size);
return errno;
}
}
/* Remember the stack-related values. */
pd->stackblock = mem;
pd->stackblock_size = size;
/* Update guardsize for newly allocated guardsize to avoid
an mprotect in guard resize below. */
pd->guardsize = guardsize;
/* We allocated the first block thread-specific data array.
This address will not change for the lifetime of this
descriptor. */
pd->specific[0] = pd->specific_1stblock;
/* This is at least the second thread. */
pd->header.multiple_threads = 1;
#ifndef TLS_MULTIPLE_THREADS_IN_TCB
__libc_multiple_threads = 1;
#endif
#ifdef NEED_DL_SYSINFO
SETUP_THREAD_SYSINFO (pd);
#endif
/* Don't allow setxid until cloned. */
pd->setxid_futex = -1;
/* Allocate the DTV for this thread. */
if (_dl_allocate_tls (TLS_TPADJ (pd)) == NULL)
{
/* Something went wrong. */
assert (errno == ENOMEM);
/* Free the stack memory we just allocated. */
(void) __munmap (mem, size);
return errno;
}
/* Prepare to modify global data. */
lll_lock (GL (dl_stack_cache_lock), LLL_PRIVATE);
/* And add to the list of stacks in use. */
stack_list_add (&pd->list, &GL (dl_stack_used));
lll_unlock (GL (dl_stack_cache_lock), LLL_PRIVATE);
/* There might have been a race. Another thread might have
caused the stacks to get exec permission while this new
stack was prepared. Detect if this was possible and
change the permission if necessary. */
if (__builtin_expect ((GL(dl_stack_flags) & PF_X) != 0
&& (prot & PROT_EXEC) == 0, 0))
{
int err = __nptl_change_stack_perm (pd);
if (err != 0)
{
/* Free the stack memory we just allocated. */
(void) __munmap (mem, size);
return err;
}
}
/* Note that all of the stack and the thread descriptor is
zeroed. This means we do not have to initialize fields
with initial value zero. This is specifically true for
the 'tid' field which is always set back to zero once the
stack is not used anymore and for the 'guardsize' field
which will be read next. */
}
/* Create or resize the guard area if necessary. */
if (__glibc_unlikely (guardsize > pd->guardsize))
{
char *guard = guard_position (mem, size, guardsize, pd,
pagesize_m1);
if (__mprotect (guard, guardsize, PROT_NONE) != 0)
{
mprot_error:
lll_lock (GL (dl_stack_cache_lock), LLL_PRIVATE);
/* Remove the thread from the list. */
stack_list_del (&pd->list);
lll_unlock (GL (dl_stack_cache_lock), LLL_PRIVATE);
/* Get rid of the TLS block we allocated. */
_dl_deallocate_tls (TLS_TPADJ (pd), false);
/* Free the stack memory regardless of whether the size
of the cache is over the limit or not. If this piece
of memory caused problems we better do not use it
anymore. Uh, and we ignore possible errors. There
is nothing we could do. */
(void) __munmap (mem, size);
return errno;
}
pd->guardsize = guardsize;
}
else if (__builtin_expect (pd->guardsize - guardsize > size - reqsize,
0))
{
/* The old guard area is too large. */
#ifdef NEED_SEPARATE_REGISTER_STACK
char *guard = mem + (((size - guardsize) / 2) & ~pagesize_m1);
char *oldguard = mem + (((size - pd->guardsize) / 2) & ~pagesize_m1);
if (oldguard < guard
&& __mprotect (oldguard, guard - oldguard, prot) != 0)
goto mprot_error;
if (__mprotect (guard + guardsize,
oldguard + pd->guardsize - guard - guardsize,
prot) != 0)
goto mprot_error;
#elif _STACK_GROWS_DOWN
if (__mprotect ((char *) mem + guardsize, pd->guardsize - guardsize,
prot) != 0)
goto mprot_error;
#elif _STACK_GROWS_UP
char *new_guard = (char *)(((uintptr_t) pd - guardsize)
& ~pagesize_m1);
char *old_guard = (char *)(((uintptr_t) pd - pd->guardsize)
& ~pagesize_m1);
/* The guard size difference might be > 0, but once rounded
to the nearest page the size difference might be zero. */
if (new_guard > old_guard
&& __mprotect (old_guard, new_guard - old_guard, prot) != 0)
goto mprot_error;
#endif
pd->guardsize = guardsize;
}
/* The pthread_getattr_np() calls need to get passed the size
requested in the attribute, regardless of how large the
actually used guardsize is. */
pd->reported_guardsize = reported_guardsize;
}
/* Initialize the lock. We have to do this unconditionally since the
stillborn thread could be canceled while the lock is taken. */
pd->lock = LLL_LOCK_INITIALIZER;
/* The robust mutex lists also need to be initialized
unconditionally because the cleanup for the previous stack owner
might have happened in the kernel. */
pd->robust_head.futex_offset = (offsetof (pthread_mutex_t, __data.__lock)
- offsetof (pthread_mutex_t,
__data.__list.__next));
pd->robust_head.list_op_pending = NULL;
#if __PTHREAD_MUTEX_HAVE_PREV
pd->robust_prev = &pd->robust_head;
#endif
pd->robust_head.list = &pd->robust_head;
/* We place the thread descriptor at the end of the stack. */
*pdp = pd;
#if _STACK_GROWS_DOWN
void *stacktop;
# if TLS_TCB_AT_TP
/* The stack begins before the TCB and the static TLS block. */
stacktop = ((char *) (pd + 1) - __static_tls_size);
# elif TLS_DTV_AT_TP
stacktop = (char *) (pd - 1);
# endif
# ifdef NEED_SEPARATE_REGISTER_STACK
*stack = pd->stackblock;
*stacksize = stacktop - *stack;
# else
*stack = stacktop;
# endif
#else
*stack = pd->stackblock;
#endif
return 0;
}
void
__deallocate_stack (struct pthread *pd)
{
lll_lock (GL (dl_stack_cache_lock), LLL_PRIVATE);
/* Remove the thread from the list of threads with user defined
stacks. */
stack_list_del (&pd->list);
/* Not much to do. Just free the mmap()ed memory. Note that we do
not reset the 'used' flag in the 'tid' field. This is done by
the kernel. If no thread has been created yet this field is
still zero. */
if (__glibc_likely (! pd->user_stack))
(void) queue_stack (pd);
else
/* Free the memory associated with the ELF TLS. */
_dl_deallocate_tls (TLS_TPADJ (pd), false);
lll_unlock (GL (dl_stack_cache_lock), LLL_PRIVATE);
}
/* In case of a fork() call the memory allocation in the child will be
the same but only one thread is running. All stacks except that of
the one running thread are not used anymore. We have to recycle
them. */
void
__reclaim_stacks (void)
{
struct pthread *self = (struct pthread *) THREAD_SELF;
/* No locking necessary. The caller is the only stack in use. But
we have to be aware that we might have interrupted a list
operation. */
if (GL (dl_in_flight_stack) != 0)
{
bool add_p = GL (dl_in_flight_stack) & 1;
list_t *elem = (list_t *) (GL (dl_in_flight_stack) & ~(uintptr_t) 1);
if (add_p)
{
/* We always add at the beginning of the list. So in this case we
only need to check the beginning of these lists to see if the
pointers at the head of the list are inconsistent. */
list_t *l = NULL;
if (GL (dl_stack_used).next->prev != &GL (dl_stack_used))
l = &GL (dl_stack_used);
else if (GL (dl_stack_cache).next->prev != &GL (dl_stack_cache))
l = &GL (dl_stack_cache);
if (l != NULL)
{
assert (l->next->prev == elem);
elem->next = l->next;
elem->prev = l;
l->next = elem;
}
}
else
{
/* We can simply always replay the delete operation. */
elem->next->prev = elem->prev;
elem->prev->next = elem->next;
}
}
/* Mark all stacks except the still running one as free. */
list_t *runp;
list_for_each (runp, &GL (dl_stack_used))
{
struct pthread *curp = list_entry (runp, struct pthread, list);
if (curp != self)
{
/* This marks the stack as free. */
curp->tid = 0;
/* Account for the size of the stack. */
GL (dl_stack_cache_actsize) += curp->stackblock_size;
if (curp->specific_used)
{
/* Clear the thread-specific data. */
memset (curp->specific_1stblock, '\0',
sizeof (curp->specific_1stblock));
curp->specific_used = false;
for (size_t cnt = 1; cnt < PTHREAD_KEY_1STLEVEL_SIZE; ++cnt)
if (curp->specific[cnt] != NULL)
{
memset (curp->specific[cnt], '\0',
sizeof (curp->specific_1stblock));
/* We have allocated the block which we do not
free here so re-set the bit. */
curp->specific_used = true;
}
}
}
}
/* Add the stack of all running threads to the cache. */
list_splice (&GL (dl_stack_used), &GL (dl_stack_cache));
/* Remove the entry for the current thread to from the cache list
and add it to the list of running threads. Which of the two
lists is decided by the user_stack flag. */
stack_list_del (&self->list);
/* Re-initialize the lists for all the threads. */
INIT_LIST_HEAD (&GL (dl_stack_used));
INIT_LIST_HEAD (&GL (dl_stack_user));
if (__glibc_unlikely (THREAD_GETMEM (self, user_stack)))
list_add (&self->list, &GL (dl_stack_user));
else
list_add (&self->list, &GL (dl_stack_used));
/* There is one thread running. */
__nptl_nthreads = 1;
GL (dl_in_flight_stack) = 0;
/* Initialize locks. */
GL (dl_stack_cache_lock) = LLL_LOCK_INITIALIZER;
__default_pthread_attr_lock = LLL_LOCK_INITIALIZER;
}
|