Lines Matching refs:stack
1624 The stack snapshot is taken immediately after the parent has
1747 try taking stack snapshots of it. */
2716 /* For each semaphore, we maintain a stack of SOs. When a 'post'
2720 SO), and the SO is pushed on the semaphore's stack.
2723 semaphore, we pop a SO off the semaphore's stack (which should be
2727 It may not be necessary to use a stack - perhaps a bag of SOs would
2780 /* xa is the stack for this semaphore. */
2786 return NULL; /* odd, the stack is empty */
2792 /* hmm, that's odd. No stack for this semaphore. */
2808 /* Empty out the semaphore's SO stack. This way of doing it is
2838 /* Empty out the semaphore's SO stack. This way of doing it is
2855 each of them, and push them all on the stack. */
2870 on on a stack of SOs associated with 'sem'. This is later used
2873 stack, they acquire dependencies on the posting thread
2900 the 'sem' from this semaphore's SO-stack, and do a strong recv
3130 /* and we don't need this any more. Perhaps a stack-allocated
3837 XArray* stack; /* of Lock* */
3851 stack = VG_(newXA)( HG_(zalloc), "hg.lddft.1", HG_(free), sizeof(Lock*) );
3854 (void) VG_(addToXA)( stack, &src );
3858 ssz = VG_(sizeXA)( stack );
3862 here = *(Lock**) VG_(indexXA)( stack, ssz-1 );
3863 VG_(dropTailXA)( stack, 1 );
3875 (void) VG_(addToXA)( stack, &succs_words[i] );
3879 VG_(deleteXA)( stack );
3964 but the stack traces that have established the required order
3977 However, we can report the current stack trace (where
3978 lk is being taken), and the stack trace where other was acquired:
4435 // mostly above) that we assume implies a stack reference.
4519 stack. In fact just skip stack refs that are within 4 pages
4521 filters out most stack references. */
5565 else if VG_BOOL_CLO(arg, "--check-stack-refs",
5582 " full: show both stack traces for a data race (can be very slow)\n"
5586 " --check-stack-refs=no|yes race-check reads and writes on the\n"
5587 " main stack and thread stacks? [yes]\n"
5635 //zz VG_(printf)(" hbefore: %'10lu stack high water mark\n",
