2 * interrupt-handling magic
6 * This software is part of the SBCL system. See the README file for
9 * This software is derived from the CMU CL system, which was
10 * written at Carnegie Mellon University and released into the
11 * public domain. The software is in the public domain and is
12 * provided with absolutely no warranty. See the COPYING and CREDITS
13 * files for more information.
17 /* As far as I can tell, what's going on here is:
19 * In the case of most signals, when Lisp asks us to handle the
20 * signal, the outermost handler (the one actually passed to UNIX) is
21 * either interrupt_handle_now(..) or maybe_now_maybe_later(..).
22 * In that case, the Lisp-level handler is stored in interrupt_handlers[..]
23 * and interrupt_low_level_handlers[..] is cleared.
25 * However, some signals need special handling, e.g.
27 * o the SIGSEGV (for e.g. Linux) or SIGBUS (for e.g. FreeBSD) used by the
28 * garbage collector to detect violations of write protection,
29 * because some cases of such signals (e.g. GC-related violations of
30 * write protection) are handled at C level and never passed on to
31 * Lisp. For such signals, we still store any Lisp-level handler
32 * in interrupt_handlers[..], but for the outermost handle we use
33 * the value from interrupt_low_level_handlers[..], instead of the
34 * ordinary interrupt_handle_now(..) or interrupt_handle_later(..).
36 * o the SIGTRAP (Linux/Alpha) which Lisp code uses to handle breakpoints,
37 * pseudo-atomic sections, and some classes of error (e.g. "function
38 * not defined"). This never goes anywhere near the Lisp handlers at all.
39 * See runtime/alpha-arch.c and code/signal.lisp
41 * - WHN 20000728, dan 20010128 */
49 #include <sys/types.h>
50 #ifndef LISP_FEATURE_WIN32
58 #include "interrupt.h"
66 #include "pseudo-atomic.h"
67 #include "genesis/fdefn.h"
68 #include "genesis/simple-fun.h"
69 #include "genesis/cons.h"
71 /* When we catch an internal error, should we pass it back to Lisp to
72 * be handled in a high-level way? (Early in cold init, the answer is
73 * 'no', because Lisp is still too brain-dead to handle anything.
74 * After sufficient initialization has been completed, the answer
76 boolean internal_errors_enabled = 0;
78 #ifndef LISP_FEATURE_WIN32
80 void (*interrupt_low_level_handlers[NSIG]) (int, siginfo_t*, os_context_t*);
82 union interrupt_handler interrupt_handlers[NSIG];
84 /* Under Linux on some architectures, we appear to have to restore the
85 * FPU control word from the context, as after the signal is delivered
86 * we appear to have a null FPU control word. */
87 #if defined(RESTORE_FP_CONTROL_FROM_CONTEXT)
88 #define RESTORE_FP_CONTROL_WORD(context,void_context) \
89 os_context_t *context = arch_os_get_context(&void_context); \
90 os_restore_fp_control(context);
92 #define RESTORE_FP_CONTROL_WORD(context,void_context) \
93 os_context_t *context = arch_os_get_context(&void_context);
96 /* Foreign code may want to start some threads on its own.
97 * Non-targetted, truly asynchronous signals can be delivered to
98 * basically any thread, but invoking Lisp handlers in such foregign
99 * threads is really bad, so let's resignal it.
101 * This should at least bring attention to the problem, but it cannot
102 * work for SIGSEGV and similar. It is good enough for timers, and
103 * maybe all deferrables. */
105 #ifdef LISP_FEATURE_SB_THREAD
107 add_handled_signals(sigset_t *sigset)
110 for(i = 1; i < NSIG; i++) {
111 if (!(ARE_SAME_HANDLER(interrupt_low_level_handlers[i], SIG_DFL)) ||
112 !(ARE_SAME_HANDLER(interrupt_handlers[i].c, SIG_DFL))) {
113 sigaddset(sigset, i);
118 void block_signals(sigset_t *what, sigset_t *where, sigset_t *old);
122 maybe_resignal_to_lisp_thread(int signal, os_context_t *context)
124 #ifdef LISP_FEATURE_SB_THREAD
125 if (!pthread_getspecific(lisp_thread)) {
126 if (!(sigismember(&deferrable_sigset,signal))) {
127 corruption_warning_and_maybe_lose
128 ("Received signal %d in non-lisp thread %lu, resignalling to a lisp thread.",
134 sigemptyset(&sigset);
135 add_handled_signals(&sigset);
136 block_signals(&sigset, 0, 0);
137 block_signals(&sigset, os_context_sigmask_addr(context), 0);
138 kill(getpid(), signal);
146 /* These are to be used in signal handlers. Currently all handlers are
147 * called from one of:
149 * interrupt_handle_now_handler
150 * maybe_now_maybe_later
151 * unblock_me_trampoline
152 * low_level_handle_now_handler
153 * low_level_maybe_now_maybe_later
154 * low_level_unblock_me_trampoline
156 * This gives us a single point of control (or six) over errno, fp
157 * control word, and fixing up signal context on sparc.
159 * The SPARC/Linux platform doesn't quite do signals the way we want
160 * them done. The third argument in the handler isn't filled in by the
161 * kernel properly, so we fix it up ourselves in the
162 * arch_os_get_context(..) function. -- CSR, 2002-07-23
164 #define SAVE_ERRNO(signal,context,void_context) \
166 int _saved_errno = errno; \
167 RESTORE_FP_CONTROL_WORD(context,void_context); \
168 if (!maybe_resignal_to_lisp_thread(signal, context)) \
171 #define RESTORE_ERRNO \
173 errno = _saved_errno; \
176 static void run_deferred_handler(struct interrupt_data *data,
177 os_context_t *context);
178 #ifndef LISP_FEATURE_WIN32
179 static void store_signal_data_for_later (struct interrupt_data *data,
180 void *handler, int signal,
182 os_context_t *context);
185 /* Generic signal related utilities. */
188 get_current_sigmask(sigset_t *sigset)
190 /* Get the current sigmask, by blocking the empty set. */
191 thread_sigmask(SIG_BLOCK, 0, sigset);
195 block_signals(sigset_t *what, sigset_t *where, sigset_t *old)
200 sigcopyset(old, where);
201 for(i = 1; i < NSIG; i++) {
202 if (sigismember(what, i))
206 thread_sigmask(SIG_BLOCK, what, old);
211 unblock_signals(sigset_t *what, sigset_t *where, sigset_t *old)
216 sigcopyset(old, where);
217 for(i = 1; i < NSIG; i++) {
218 if (sigismember(what, i))
222 thread_sigmask(SIG_UNBLOCK, what, old);
227 print_sigset(sigset_t *sigset)
230 for(i = 1; i < NSIG; i++) {
231 if (sigismember(sigset, i))
232 fprintf(stderr, "Signal %d masked\n", i);
236 /* Return 1 is all signals is sigset2 are masked in sigset, return 0
237 * if all re unmasked else die. Passing NULL for sigset is a shorthand
238 * for the current sigmask. */
240 all_signals_blocked_p(sigset_t *sigset, sigset_t *sigset2,
243 #if !defined(LISP_FEATURE_WIN32)
245 boolean has_blocked = 0, has_unblocked = 0;
248 get_current_sigmask(¤t);
251 for(i = 1; i < NSIG; i++) {
252 if (sigismember(sigset2, i)) {
253 if (sigismember(sigset, i))
259 if (has_blocked && has_unblocked) {
260 print_sigset(sigset);
261 lose("some %s signals blocked, some unblocked\n", name);
271 /* Deferrables, blockables, gc signals. */
274 sigaddset_deferrable(sigset_t *s)
276 sigaddset(s, SIGHUP);
277 sigaddset(s, SIGINT);
278 sigaddset(s, SIGTERM);
279 sigaddset(s, SIGQUIT);
280 sigaddset(s, SIGPIPE);
281 sigaddset(s, SIGALRM);
282 sigaddset(s, SIGURG);
283 sigaddset(s, SIGTSTP);
284 sigaddset(s, SIGCHLD);
286 #ifndef LISP_FEATURE_HPUX
287 sigaddset(s, SIGXCPU);
288 sigaddset(s, SIGXFSZ);
290 sigaddset(s, SIGVTALRM);
291 sigaddset(s, SIGPROF);
292 sigaddset(s, SIGWINCH);
296 sigaddset_blockable(sigset_t *sigset)
298 sigaddset_deferrable(sigset);
299 sigaddset_gc(sigset);
303 sigaddset_gc(sigset_t *sigset)
305 #ifdef LISP_FEATURE_SB_THREAD
306 sigaddset(sigset,SIG_STOP_FOR_GC);
310 /* initialized in interrupt_init */
311 sigset_t deferrable_sigset;
312 sigset_t blockable_sigset;
317 #if !defined(LISP_FEATURE_WIN32)
319 deferrables_blocked_p(sigset_t *sigset)
321 return all_signals_blocked_p(sigset, &deferrable_sigset, "deferrable");
326 check_deferrables_unblocked_or_lose(sigset_t *sigset)
328 #if !defined(LISP_FEATURE_WIN32)
329 if (deferrables_blocked_p(sigset))
330 lose("deferrables blocked\n");
335 check_deferrables_blocked_or_lose(sigset_t *sigset)
337 #if !defined(LISP_FEATURE_WIN32)
338 if (!deferrables_blocked_p(sigset))
339 lose("deferrables unblocked\n");
343 #if !defined(LISP_FEATURE_WIN32)
345 blockables_blocked_p(sigset_t *sigset)
347 return all_signals_blocked_p(sigset, &blockable_sigset, "blockable");
352 check_blockables_unblocked_or_lose(sigset_t *sigset)
354 #if !defined(LISP_FEATURE_WIN32)
355 if (blockables_blocked_p(sigset))
356 lose("blockables blocked\n");
361 check_blockables_blocked_or_lose(sigset_t *sigset)
363 #if !defined(LISP_FEATURE_WIN32)
364 if (!blockables_blocked_p(sigset))
365 lose("blockables unblocked\n");
369 #if !defined(LISP_FEATURE_WIN32)
371 gc_signals_blocked_p(sigset_t *sigset)
373 return all_signals_blocked_p(sigset, &gc_sigset, "gc");
378 check_gc_signals_unblocked_or_lose(sigset_t *sigset)
380 #if !defined(LISP_FEATURE_WIN32)
381 if (gc_signals_blocked_p(sigset))
382 lose("gc signals blocked\n");
387 check_gc_signals_blocked_or_lose(sigset_t *sigset)
389 #if !defined(LISP_FEATURE_WIN32)
390 if (!gc_signals_blocked_p(sigset))
391 lose("gc signals unblocked\n");
396 block_deferrable_signals(sigset_t *where, sigset_t *old)
398 #ifndef LISP_FEATURE_WIN32
399 block_signals(&deferrable_sigset, where, old);
404 block_blockable_signals(sigset_t *where, sigset_t *old)
406 #ifndef LISP_FEATURE_WIN32
407 block_signals(&blockable_sigset, where, old);
412 block_gc_signals(sigset_t *where, sigset_t *old)
414 #ifndef LISP_FEATURE_WIN32
415 block_signals(&gc_sigset, where, old);
420 unblock_deferrable_signals(sigset_t *where, sigset_t *old)
422 #ifndef LISP_FEATURE_WIN32
423 if (interrupt_handler_pending_p())
424 lose("unblock_deferrable_signals: losing proposition\n");
425 check_gc_signals_unblocked_or_lose(where);
426 unblock_signals(&deferrable_sigset, where, old);
431 unblock_blockable_signals(sigset_t *where, sigset_t *old)
433 #ifndef LISP_FEATURE_WIN32
434 unblock_signals(&blockable_sigset, where, old);
439 unblock_gc_signals(sigset_t *where, sigset_t *old)
441 #ifndef LISP_FEATURE_WIN32
442 unblock_signals(&gc_sigset, where, old);
447 unblock_signals_in_context_and_maybe_warn(os_context_t *context)
449 #ifndef LISP_FEATURE_WIN32
450 sigset_t *sigset = os_context_sigmask_addr(context);
451 if (all_signals_blocked_p(sigset, &gc_sigset, "gc")) {
452 corruption_warning_and_maybe_lose(
453 "Enabling blocked gc signals to allow returning to Lisp without risking\n\
454 gc deadlocks. Since GC signals are only blocked in signal handlers when \n\
455 they are not safe to interrupt at all, this is a pretty severe occurrence.\n");
456 unblock_gc_signals(sigset, 0);
458 if (!interrupt_handler_pending_p()) {
459 unblock_deferrable_signals(sigset, 0);
466 check_interrupts_enabled_or_lose(os_context_t *context)
468 struct thread *thread=arch_os_get_current_thread();
469 if (SymbolValue(INTERRUPTS_ENABLED,thread) == NIL)
470 lose("interrupts not enabled\n");
471 if (arch_pseudo_atomic_atomic(context))
472 lose ("in pseudo atomic section\n");
475 /* Save sigset (or the current sigmask if 0) if there is no pending
476 * handler, because that means that deferabbles are already blocked.
477 * The purpose is to avoid losing the pending gc signal if a
478 * deferrable interrupt async unwinds between clearing the pseudo
479 * atomic and trapping to GC.*/
481 maybe_save_gc_mask_and_block_deferrables(sigset_t *sigset)
483 #ifndef LISP_FEATURE_WIN32
484 struct thread *thread = arch_os_get_current_thread();
485 struct interrupt_data *data = thread->interrupt_data;
487 /* Obviously, this function is called when signals may not be
488 * blocked. Let's make sure we are not interrupted. */
489 block_blockable_signals(0, &oldset);
490 #ifndef LISP_FEATURE_SB_THREAD
491 /* With threads a SIG_STOP_FOR_GC and a normal GC may also want to
493 if (data->gc_blocked_deferrables)
494 lose("gc_blocked_deferrables already true\n");
496 if ((!data->pending_handler) &&
497 (!data->gc_blocked_deferrables)) {
498 FSHOW_SIGNAL((stderr,"/setting gc_blocked_deferrables\n"));
499 data->gc_blocked_deferrables = 1;
501 /* This is the sigmask of some context. */
502 sigcopyset(&data->pending_mask, sigset);
503 sigaddset_deferrable(sigset);
504 thread_sigmask(SIG_SETMASK,&oldset,0);
507 /* Operating on the current sigmask. Save oldset and
508 * unblock gc signals. In the end, this is equivalent to
509 * blocking the deferrables. */
510 sigcopyset(&data->pending_mask, &oldset);
511 thread_sigmask(SIG_UNBLOCK, &gc_sigset, 0);
515 thread_sigmask(SIG_SETMASK,&oldset,0);
519 /* Are we leaving WITH-GCING and already running with interrupts
520 * enabled, without the protection of *GC-INHIBIT* T and there is gc
521 * (or stop for gc) pending, but we haven't trapped yet? */
523 in_leaving_without_gcing_race_p(struct thread *thread)
525 return ((SymbolValue(IN_WITHOUT_GCING,thread) != NIL) &&
526 (SymbolValue(INTERRUPTS_ENABLED,thread) != NIL) &&
527 (SymbolValue(GC_INHIBIT,thread) == NIL) &&
528 ((SymbolValue(GC_PENDING,thread) != NIL)
529 #if defined(LISP_FEATURE_SB_THREAD)
530 || (SymbolValue(STOP_FOR_GC_PENDING,thread) != NIL)
535 /* Check our baroque invariants. */
537 check_interrupt_context_or_lose(os_context_t *context)
539 #ifndef LISP_FEATURE_WIN32
540 struct thread *thread = arch_os_get_current_thread();
541 struct interrupt_data *data = thread->interrupt_data;
542 int interrupt_deferred_p = (data->pending_handler != 0);
543 int interrupt_pending = (SymbolValue(INTERRUPT_PENDING,thread) != NIL);
544 sigset_t *sigset = os_context_sigmask_addr(context);
545 /* On PPC pseudo_atomic_interrupted is cleared when coming out of
546 * handle_allocation_trap. */
547 #if defined(LISP_FEATURE_GENCGC) && !defined(LISP_FEATURE_PPC)
548 int interrupts_enabled = (SymbolValue(INTERRUPTS_ENABLED,thread) != NIL);
549 int gc_inhibit = (SymbolValue(GC_INHIBIT,thread) != NIL);
550 int gc_pending = (SymbolValue(GC_PENDING,thread) == T);
551 int pseudo_atomic_interrupted = get_pseudo_atomic_interrupted(thread);
552 int in_race_p = in_leaving_without_gcing_race_p(thread);
553 /* In the time window between leaving the *INTERRUPTS-ENABLED* NIL
554 * section and trapping, a SIG_STOP_FOR_GC would see the next
555 * check fail, for this reason sig_stop_for_gc handler does not
556 * call this function. */
557 if (interrupt_deferred_p) {
558 if (!(!interrupts_enabled || pseudo_atomic_interrupted || in_race_p))
559 lose("Stray deferred interrupt.\n");
562 if (!(pseudo_atomic_interrupted || gc_inhibit || in_race_p))
563 lose("GC_PENDING, but why?\n");
564 #if defined(LISP_FEATURE_SB_THREAD)
566 int stop_for_gc_pending =
567 (SymbolValue(STOP_FOR_GC_PENDING,thread) != NIL);
568 if (stop_for_gc_pending)
569 if (!(pseudo_atomic_interrupted || gc_inhibit || in_race_p))
570 lose("STOP_FOR_GC_PENDING, but why?\n");
571 if (pseudo_atomic_interrupted)
572 if (!(gc_pending || stop_for_gc_pending || interrupt_deferred_p))
573 lose("pseudo_atomic_interrupted, but why?\n");
576 if (pseudo_atomic_interrupted)
577 if (!(gc_pending || interrupt_deferred_p))
578 lose("pseudo_atomic_interrupted, but why?\n");
581 if (interrupt_pending && !interrupt_deferred_p)
582 lose("INTERRUPT_PENDING but not pending handler.\n");
583 if ((data->gc_blocked_deferrables) && interrupt_pending)
584 lose("gc_blocked_deferrables and interrupt pending\n.");
585 if (data->gc_blocked_deferrables)
586 check_deferrables_blocked_or_lose(sigset);
587 if (interrupt_pending || interrupt_deferred_p ||
588 data->gc_blocked_deferrables)
589 check_deferrables_blocked_or_lose(sigset);
591 check_deferrables_unblocked_or_lose(sigset);
592 /* If deferrables are unblocked then we are open to signals
593 * that run lisp code. */
594 check_gc_signals_unblocked_or_lose(sigset);
600 * utility routines used by various signal handlers
604 build_fake_control_stack_frames(struct thread *th,os_context_t *context)
606 #ifndef LISP_FEATURE_C_STACK_IS_CONTROL_STACK
610 /* Build a fake stack frame or frames */
612 current_control_frame_pointer =
613 (lispobj *)(unsigned long)
614 (*os_context_register_addr(context, reg_CSP));
615 if ((lispobj *)(unsigned long)
616 (*os_context_register_addr(context, reg_CFP))
617 == current_control_frame_pointer) {
618 /* There is a small window during call where the callee's
619 * frame isn't built yet. */
620 if (lowtag_of(*os_context_register_addr(context, reg_CODE))
621 == FUN_POINTER_LOWTAG) {
622 /* We have called, but not built the new frame, so
623 * build it for them. */
624 current_control_frame_pointer[0] =
625 *os_context_register_addr(context, reg_OCFP);
626 current_control_frame_pointer[1] =
627 *os_context_register_addr(context, reg_LRA);
628 current_control_frame_pointer += 8;
629 /* Build our frame on top of it. */
630 oldcont = (lispobj)(*os_context_register_addr(context, reg_CFP));
633 /* We haven't yet called, build our frame as if the
634 * partial frame wasn't there. */
635 oldcont = (lispobj)(*os_context_register_addr(context, reg_OCFP));
638 /* We can't tell whether we are still in the caller if it had to
639 * allocate a stack frame due to stack arguments. */
640 /* This observation provoked some past CMUCL maintainer to ask
641 * "Can anything strange happen during return?" */
644 oldcont = (lispobj)(*os_context_register_addr(context, reg_CFP));
647 current_control_stack_pointer = current_control_frame_pointer + 8;
649 current_control_frame_pointer[0] = oldcont;
650 current_control_frame_pointer[1] = NIL;
651 current_control_frame_pointer[2] =
652 (lispobj)(*os_context_register_addr(context, reg_CODE));
656 /* Stores the context for gc to scavange and builds fake stack
659 fake_foreign_function_call(os_context_t *context)
662 struct thread *thread=arch_os_get_current_thread();
664 /* context_index incrementing must not be interrupted */
665 check_blockables_blocked_or_lose(0);
667 /* Get current Lisp state from context. */
669 dynamic_space_free_pointer =
670 (lispobj *)(unsigned long)
671 (*os_context_register_addr(context, reg_ALLOC));
672 /* fprintf(stderr,"dynamic_space_free_pointer: %p\n", */
673 /* dynamic_space_free_pointer); */
674 #if defined(LISP_FEATURE_ALPHA) || defined(LISP_FEATURE_MIPS)
675 if ((long)dynamic_space_free_pointer & 1) {
676 lose("dead in fake_foreign_function_call, context = %x\n", context);
679 /* why doesnt PPC and SPARC do something like this: */
680 #if defined(LISP_FEATURE_HPPA)
681 if ((long)dynamic_space_free_pointer & 4) {
682 lose("dead in fake_foreign_function_call, context = %x, d_s_f_p = %x\n", context, dynamic_space_free_pointer);
687 current_binding_stack_pointer =
688 (lispobj *)(unsigned long)
689 (*os_context_register_addr(context, reg_BSP));
692 build_fake_control_stack_frames(thread,context);
694 /* Do dynamic binding of the active interrupt context index
695 * and save the context in the context array. */
697 fixnum_value(SymbolValue(FREE_INTERRUPT_CONTEXT_INDEX,thread));
699 if (context_index >= MAX_INTERRUPTS) {
700 lose("maximum interrupt nesting depth (%d) exceeded\n", MAX_INTERRUPTS);
703 bind_variable(FREE_INTERRUPT_CONTEXT_INDEX,
704 make_fixnum(context_index + 1),thread);
706 thread->interrupt_contexts[context_index] = context;
708 #ifdef FOREIGN_FUNCTION_CALL_FLAG
709 foreign_function_call_active = 1;
713 /* blocks all blockable signals. If you are calling from a signal handler,
714 * the usual signal mask will be restored from the context when the handler
715 * finishes. Otherwise, be careful */
717 undo_fake_foreign_function_call(os_context_t *context)
719 struct thread *thread=arch_os_get_current_thread();
720 /* Block all blockable signals. */
721 block_blockable_signals(0, 0);
723 #ifdef FOREIGN_FUNCTION_CALL_FLAG
724 foreign_function_call_active = 0;
727 /* Undo dynamic binding of FREE_INTERRUPT_CONTEXT_INDEX */
731 /* Put the dynamic space free pointer back into the context. */
732 *os_context_register_addr(context, reg_ALLOC) =
733 (unsigned long) dynamic_space_free_pointer
734 | (*os_context_register_addr(context, reg_ALLOC)
737 ((unsigned long)(*os_context_register_addr(context, reg_ALLOC))
739 | ((unsigned long) dynamic_space_free_pointer & LOWTAG_MASK);
744 /* a handler for the signal caused by execution of a trap opcode
745 * signalling an internal error */
747 interrupt_internal_error(os_context_t *context, boolean continuable)
751 fake_foreign_function_call(context);
753 if (!internal_errors_enabled) {
754 describe_internal_error(context);
755 /* There's no good way to recover from an internal error
756 * before the Lisp error handling mechanism is set up. */
757 lose("internal error too early in init, can't recover\n");
760 /* Allocate the SAP object while the interrupts are still
762 unblock_gc_signals(0, 0);
763 context_sap = alloc_sap(context);
765 #ifndef LISP_FEATURE_WIN32
766 thread_sigmask(SIG_SETMASK, os_context_sigmask_addr(context), 0);
769 #if defined(LISP_FEATURE_LINUX) && defined(LISP_FEATURE_MIPS)
770 /* Workaround for blocked SIGTRAP. */
773 sigemptyset(&newset);
774 sigaddset(&newset, SIGTRAP);
775 thread_sigmask(SIG_UNBLOCK, &newset, 0);
779 SHOW("in interrupt_internal_error");
781 /* Display some rudimentary debugging information about the
782 * error, so that even if the Lisp error handler gets badly
783 * confused, we have a chance to determine what's going on. */
784 describe_internal_error(context);
786 funcall2(StaticSymbolFunction(INTERNAL_ERROR), context_sap,
787 continuable ? T : NIL);
789 undo_fake_foreign_function_call(context); /* blocks signals again */
791 arch_skip_instruction(context);
795 interrupt_handler_pending_p(void)
797 struct thread *thread = arch_os_get_current_thread();
798 struct interrupt_data *data = thread->interrupt_data;
799 return (data->pending_handler != 0);
803 interrupt_handle_pending(os_context_t *context)
805 /* There are three ways we can get here. First, if an interrupt
806 * occurs within pseudo-atomic, it will be deferred, and we'll
807 * trap to here at the end of the pseudo-atomic block. Second, if
808 * the GC (in alloc()) decides that a GC is required, it will set
809 * *GC-PENDING* and pseudo-atomic-interrupted if not *GC-INHIBIT*,
810 * and alloc() is always called from within pseudo-atomic, and
811 * thus we end up here again. Third, when calling GC-ON or at the
812 * end of a WITHOUT-GCING, MAYBE-HANDLE-PENDING-GC will trap to
813 * here if there is a pending GC. Fourth, ahem, at the end of
814 * WITHOUT-INTERRUPTS (bar complications with nesting). */
816 /* Win32 only needs to handle the GC cases (for now?) */
818 struct thread *thread = arch_os_get_current_thread();
819 struct interrupt_data *data = thread->interrupt_data;
821 if (arch_pseudo_atomic_atomic(context)) {
822 lose("Handling pending interrupt in pseudo atomic.");
825 FSHOW_SIGNAL((stderr, "/entering interrupt_handle_pending\n"));
827 check_blockables_blocked_or_lose(0);
829 /* If GC/SIG_STOP_FOR_GC struck during PA and there was no pending
830 * handler, then the pending mask was saved and
831 * gc_blocked_deferrables set. Hence, there can be no pending
832 * handler and it's safe to restore the pending mask.
834 * Note, that if gc_blocked_deferrables is false we may still have
835 * to GC. In this case, we are coming out of a WITHOUT-GCING or a
836 * pseudo atomic was interrupt be a deferrable first. */
837 if (data->gc_blocked_deferrables) {
838 if (data->pending_handler)
839 lose("GC blocked deferrables but still got a pending handler.");
840 if (SymbolValue(GC_INHIBIT,thread)!=NIL)
841 lose("GC blocked deferrables while GC is inhibited.");
842 /* Restore the saved signal mask from the original signal (the
843 * one that interrupted us during the critical section) into
844 * the os_context for the signal we're currently in the
845 * handler for. This should ensure that when we return from
846 * the handler the blocked signals are unblocked. */
847 #ifndef LISP_FEATURE_WIN32
848 sigcopyset(os_context_sigmask_addr(context), &data->pending_mask);
850 data->gc_blocked_deferrables = 0;
853 if (SymbolValue(GC_INHIBIT,thread)==NIL) {
854 void *original_pending_handler = data->pending_handler;
856 #ifdef LISP_FEATURE_SB_THREAD
857 if (SymbolValue(STOP_FOR_GC_PENDING,thread) != NIL) {
858 /* STOP_FOR_GC_PENDING and GC_PENDING are cleared by
859 * the signal handler if it actually stops us. */
860 arch_clear_pseudo_atomic_interrupted(context);
861 sig_stop_for_gc_handler(SIG_STOP_FOR_GC,NULL,context);
864 /* Test for T and not for != NIL since the value :IN-PROGRESS
865 * is used in SUB-GC as part of the mechanism to supress
867 if (SymbolValue(GC_PENDING,thread) == T) {
869 /* Two reasons for doing this. First, if there is a
870 * pending handler we don't want to run. Second, we are
871 * going to clear pseudo atomic interrupted to avoid
872 * spurious trapping on every allocation in SUB_GC and
873 * having a pending handler with interrupts enabled and
874 * without pseudo atomic interrupted breaks an
876 if (data->pending_handler) {
877 bind_variable(ALLOW_WITH_INTERRUPTS, NIL, thread);
878 bind_variable(INTERRUPTS_ENABLED, NIL, thread);
881 arch_clear_pseudo_atomic_interrupted(context);
883 /* GC_PENDING is cleared in SUB-GC, or if another thread
884 * is doing a gc already we will get a SIG_STOP_FOR_GC and
885 * that will clear it.
887 * If there is a pending handler or gc was triggerred in a
888 * signal handler then maybe_gc won't run POST_GC and will
889 * return normally. */
890 if (!maybe_gc(context))
891 lose("GC not inhibited but maybe_gc did not GC.");
893 if (data->pending_handler) {
897 } else if (SymbolValue(GC_PENDING,thread) != NIL) {
898 /* It's not NIL or T so GC_PENDING is :IN-PROGRESS. If
899 * GC-PENDING is not NIL then we cannot trap on pseudo
900 * atomic due to GC (see if(GC_PENDING) logic in
901 * cheneygc.c an gengcgc.c), plus there is a outer
902 * WITHOUT-INTERRUPTS SUB_GC, so how did we end up
904 lose("Trapping to run pending handler while GC in progress.");
907 check_blockables_blocked_or_lose(0);
909 /* No GC shall be lost. If SUB_GC triggers another GC then
910 * that should be handled on the spot. */
911 if (SymbolValue(GC_PENDING,thread) != NIL)
912 lose("GC_PENDING after doing gc.");
913 #ifdef LISP_FEATURE_SB_THREAD
914 if (SymbolValue(STOP_FOR_GC_PENDING,thread) != NIL)
915 lose("STOP_FOR_GC_PENDING after doing gc.");
917 /* Check two things. First, that gc does not clobber a handler
918 * that's already pending. Second, that there is no interrupt
919 * lossage: if original_pending_handler was NULL then even if
920 * an interrupt arrived during GC (POST-GC, really) it was
922 if (original_pending_handler != data->pending_handler)
923 lose("pending handler changed in gc: %x -> %d.",
924 original_pending_handler, data->pending_handler);
927 #ifndef LISP_FEATURE_WIN32
928 /* There may be no pending handler, because it was only a gc that
929 * had to be executed or because Lisp is a bit too eager to call
930 * DO-PENDING-INTERRUPT. */
931 if ((SymbolValue(INTERRUPTS_ENABLED,thread) != NIL) &&
932 (data->pending_handler)) {
933 /* No matter how we ended up here, clear both
934 * INTERRUPT_PENDING and pseudo atomic interrupted. It's safe
935 * because we checked above that there is no GC pending. */
936 SetSymbolValue(INTERRUPT_PENDING, NIL, thread);
937 arch_clear_pseudo_atomic_interrupted(context);
938 /* Restore the sigmask in the context. */
939 sigcopyset(os_context_sigmask_addr(context), &data->pending_mask);
940 run_deferred_handler(data, context);
943 #ifdef LISP_FEATURE_GENCGC
944 if (get_pseudo_atomic_interrupted(thread))
945 lose("pseudo_atomic_interrupted after interrupt_handle_pending\n");
947 /* It is possible that the end of this function was reached
948 * without never actually doing anything, the tests in Lisp for
949 * when to call receive-pending-interrupt are not exact. */
950 FSHOW_SIGNAL((stderr, "/exiting interrupt_handle_pending\n"));
955 interrupt_handle_now(int signal, siginfo_t *info, os_context_t *context)
957 #ifdef FOREIGN_FUNCTION_CALL_FLAG
958 boolean were_in_lisp;
960 union interrupt_handler handler;
962 check_blockables_blocked_or_lose(0);
964 #ifndef LISP_FEATURE_WIN32
965 if (sigismember(&deferrable_sigset,signal))
966 check_interrupts_enabled_or_lose(context);
969 handler = interrupt_handlers[signal];
971 if (ARE_SAME_HANDLER(handler.c, SIG_IGN)) {
975 #ifdef FOREIGN_FUNCTION_CALL_FLAG
976 were_in_lisp = !foreign_function_call_active;
980 fake_foreign_function_call(context);
983 FSHOW_SIGNAL((stderr,
984 "/entering interrupt_handle_now(%d, info, context)\n",
987 if (ARE_SAME_HANDLER(handler.c, SIG_DFL)) {
989 /* This can happen if someone tries to ignore or default one
990 * of the signals we need for runtime support, and the runtime
991 * support decides to pass on it. */
992 lose("no handler for signal %d in interrupt_handle_now(..)\n", signal);
994 } else if (lowtag_of(handler.lisp) == FUN_POINTER_LOWTAG) {
995 /* Once we've decided what to do about contexts in a
996 * return-elsewhere world (the original context will no longer
997 * be available; should we copy it or was nobody using it anyway?)
998 * then we should convert this to return-elsewhere */
1000 /* CMUCL comment said "Allocate the SAPs while the interrupts
1001 * are still disabled.". I (dan, 2003.08.21) assume this is
1002 * because we're not in pseudoatomic and allocation shouldn't
1003 * be interrupted. In which case it's no longer an issue as
1004 * all our allocation from C now goes through a PA wrapper,
1005 * but still, doesn't hurt.
1007 * Yeah, but non-gencgc platforms don't really wrap allocation
1008 * in PA. MG - 2005-08-29 */
1010 lispobj info_sap, context_sap;
1011 /* Leave deferrable signals blocked, the handler itself will
1012 * allow signals again when it sees fit. */
1013 unblock_gc_signals(0, 0);
1014 context_sap = alloc_sap(context);
1015 info_sap = alloc_sap(info);
1017 FSHOW_SIGNAL((stderr,"/calling Lisp-level handler\n"));
1019 funcall3(handler.lisp,
1020 make_fixnum(signal),
1024 /* This cannot happen in sane circumstances. */
1026 FSHOW_SIGNAL((stderr,"/calling C-level handler\n"));
1028 #ifndef LISP_FEATURE_WIN32
1029 /* Allow signals again. */
1030 thread_sigmask(SIG_SETMASK, os_context_sigmask_addr(context), 0);
1032 (*handler.c)(signal, info, context);
1035 #ifdef FOREIGN_FUNCTION_CALL_FLAG
1039 undo_fake_foreign_function_call(context); /* block signals again */
1042 FSHOW_SIGNAL((stderr,
1043 "/returning from interrupt_handle_now(%d, info, context)\n",
1047 /* This is called at the end of a critical section if the indications
1048 * are that some signal was deferred during the section. Note that as
1049 * far as C or the kernel is concerned we dealt with the signal
1050 * already; we're just doing the Lisp-level processing now that we
1053 run_deferred_handler(struct interrupt_data *data, os_context_t *context)
1055 /* The pending_handler may enable interrupts and then another
1056 * interrupt may hit, overwrite interrupt_data, so reset the
1057 * pending handler before calling it. Trust the handler to finish
1058 * with the siginfo before enabling interrupts. */
1059 void (*pending_handler) (int, siginfo_t*, os_context_t*) =
1060 data->pending_handler;
1062 data->pending_handler=0;
1063 FSHOW_SIGNAL((stderr, "/running deferred handler %p\n", pending_handler));
1064 (*pending_handler)(data->pending_signal,&(data->pending_info), context);
1067 #ifndef LISP_FEATURE_WIN32
1069 maybe_defer_handler(void *handler, struct interrupt_data *data,
1070 int signal, siginfo_t *info, os_context_t *context)
1072 struct thread *thread=arch_os_get_current_thread();
1074 check_blockables_blocked_or_lose(0);
1076 if (SymbolValue(INTERRUPT_PENDING,thread) != NIL)
1077 lose("interrupt already pending\n");
1078 if (thread->interrupt_data->pending_handler)
1079 lose("there is a pending handler already (PA)\n");
1080 if (data->gc_blocked_deferrables)
1081 lose("maybe_defer_handler: gc_blocked_deferrables true\n");
1082 check_interrupt_context_or_lose(context);
1083 /* If interrupts are disabled then INTERRUPT_PENDING is set and
1084 * not PSEDUO_ATOMIC_INTERRUPTED. This is important for a pseudo
1085 * atomic section inside a WITHOUT-INTERRUPTS.
1087 * Also, if in_leaving_without_gcing_race_p then
1088 * interrupt_handle_pending is going to be called soon, so
1089 * stashing the signal away is safe.
1091 if ((SymbolValue(INTERRUPTS_ENABLED,thread) == NIL) ||
1092 in_leaving_without_gcing_race_p(thread)) {
1093 FSHOW_SIGNAL((stderr,
1094 "/maybe_defer_handler(%x,%d): deferred (RACE=%d)\n",
1095 (unsigned int)handler,signal,
1096 in_leaving_without_gcing_race_p(thread)));
1097 store_signal_data_for_later(data,handler,signal,info,context);
1098 SetSymbolValue(INTERRUPT_PENDING, T,thread);
1099 check_interrupt_context_or_lose(context);
1102 /* a slightly confusing test. arch_pseudo_atomic_atomic() doesn't
1103 * actually use its argument for anything on x86, so this branch
1104 * may succeed even when context is null (gencgc alloc()) */
1105 if (arch_pseudo_atomic_atomic(context)) {
1106 FSHOW_SIGNAL((stderr,
1107 "/maybe_defer_handler(%x,%d): deferred(PA)\n",
1108 (unsigned int)handler,signal));
1109 store_signal_data_for_later(data,handler,signal,info,context);
1110 arch_set_pseudo_atomic_interrupted(context);
1111 check_interrupt_context_or_lose(context);
1114 FSHOW_SIGNAL((stderr,
1115 "/maybe_defer_handler(%x,%d): not deferred\n",
1116 (unsigned int)handler,signal));
1121 store_signal_data_for_later (struct interrupt_data *data, void *handler,
1123 siginfo_t *info, os_context_t *context)
1125 if (data->pending_handler)
1126 lose("tried to overwrite pending interrupt handler %x with %x\n",
1127 data->pending_handler, handler);
1129 lose("tried to defer null interrupt handler\n");
1130 data->pending_handler = handler;
1131 data->pending_signal = signal;
1133 memcpy(&(data->pending_info), info, sizeof(siginfo_t));
1135 FSHOW_SIGNAL((stderr, "/store_signal_data_for_later: signal: %d\n",
1139 lose("Null context");
1141 /* the signal mask in the context (from before we were
1142 * interrupted) is copied to be restored when run_deferred_handler
1143 * happens. Then the usually-blocked signals are added to the mask
1144 * in the context so that we are running with blocked signals when
1145 * the handler returns */
1146 sigcopyset(&(data->pending_mask),os_context_sigmask_addr(context));
1147 sigaddset_deferrable(os_context_sigmask_addr(context));
1151 maybe_now_maybe_later(int signal, siginfo_t *info, void *void_context)
1153 SAVE_ERRNO(signal,context,void_context);
1154 struct thread *thread = arch_os_get_current_thread();
1155 struct interrupt_data *data = thread->interrupt_data;
1156 if(!maybe_defer_handler(interrupt_handle_now,data,signal,info,context))
1157 interrupt_handle_now(signal, info, context);
1162 low_level_interrupt_handle_now(int signal, siginfo_t *info,
1163 os_context_t *context)
1165 /* No FP control fixage needed, caller has done that. */
1166 check_blockables_blocked_or_lose(0);
1167 check_interrupts_enabled_or_lose(context);
1168 (*interrupt_low_level_handlers[signal])(signal, info, context);
1169 /* No Darwin context fixage needed, caller does that. */
1173 low_level_maybe_now_maybe_later(int signal, siginfo_t *info, void *void_context)
1175 SAVE_ERRNO(signal,context,void_context);
1176 struct thread *thread = arch_os_get_current_thread();
1177 struct interrupt_data *data = thread->interrupt_data;
1179 if(!maybe_defer_handler(low_level_interrupt_handle_now,data,
1180 signal,info,context))
1181 low_level_interrupt_handle_now(signal, info, context);
1186 #ifdef LISP_FEATURE_SB_THREAD
1188 /* This function must not cons, because that may trigger a GC. */
1190 sig_stop_for_gc_handler(int signal, siginfo_t *info, os_context_t *context)
1192 struct thread *thread=arch_os_get_current_thread();
1194 /* Test for GC_INHIBIT _first_, else we'd trap on every single
1195 * pseudo atomic until gc is finally allowed. */
1196 if (SymbolValue(GC_INHIBIT,thread) != NIL) {
1197 FSHOW_SIGNAL((stderr, "sig_stop_for_gc deferred (*GC-INHIBIT*)\n"));
1198 SetSymbolValue(STOP_FOR_GC_PENDING,T,thread);
1200 } else if (arch_pseudo_atomic_atomic(context)) {
1201 FSHOW_SIGNAL((stderr,"sig_stop_for_gc deferred (PA)\n"));
1202 SetSymbolValue(STOP_FOR_GC_PENDING,T,thread);
1203 arch_set_pseudo_atomic_interrupted(context);
1204 maybe_save_gc_mask_and_block_deferrables
1205 (os_context_sigmask_addr(context));
1209 FSHOW_SIGNAL((stderr, "/sig_stop_for_gc_handler\n"));
1211 /* Not PA and GC not inhibited -- we can stop now. */
1213 /* need the context stored so it can have registers scavenged */
1214 fake_foreign_function_call(context);
1216 /* Not pending anymore. */
1217 SetSymbolValue(GC_PENDING,NIL,thread);
1218 SetSymbolValue(STOP_FOR_GC_PENDING,NIL,thread);
1220 /* Consider this: in a PA section GC is requested: GC_PENDING,
1221 * pseudo_atomic_interrupted and gc_blocked_deferrables are set,
1222 * deferrables are blocked then pseudo_atomic_atomic is cleared,
1223 * but a SIG_STOP_FOR_GC arrives before trapping to
1224 * interrupt_handle_pending. Here, GC_PENDING is cleared but
1225 * pseudo_atomic_interrupted is not and we go on running with
1226 * pseudo_atomic_interrupted but without a pending interrupt or
1227 * GC. GC_BLOCKED_DEFERRABLES is also left at 1. So let's tidy it
1229 if (thread->interrupt_data->gc_blocked_deferrables) {
1230 FSHOW_SIGNAL((stderr,"cleaning up after gc_blocked_deferrables\n"));
1231 clear_pseudo_atomic_interrupted(thread);
1232 sigcopyset(os_context_sigmask_addr(context),
1233 &thread->interrupt_data->pending_mask);
1234 thread->interrupt_data->gc_blocked_deferrables = 0;
1237 if(thread_state(thread)!=STATE_RUNNING) {
1238 lose("sig_stop_for_gc_handler: wrong thread state: %ld\n",
1239 fixnum_value(thread->state));
1242 set_thread_state(thread,STATE_SUSPENDED);
1243 FSHOW_SIGNAL((stderr,"suspended\n"));
1245 wait_for_thread_state_change(thread, STATE_SUSPENDED);
1246 FSHOW_SIGNAL((stderr,"resumed\n"));
1248 if(thread_state(thread)!=STATE_RUNNING) {
1249 lose("sig_stop_for_gc_handler: wrong thread state on wakeup: %ld\n",
1250 fixnum_value(thread_state(thread)));
1253 undo_fake_foreign_function_call(context);
1259 interrupt_handle_now_handler(int signal, siginfo_t *info, void *void_context)
1261 SAVE_ERRNO(signal,context,void_context);
1262 #ifndef LISP_FEATURE_WIN32
1263 if ((signal == SIGILL) || (signal == SIGBUS)
1264 #ifndef LISP_FEATURE_LINUX
1265 || (signal == SIGEMT)
1268 corruption_warning_and_maybe_lose("Signal %d recieved", signal);
1270 interrupt_handle_now(signal, info, context);
1274 /* manipulate the signal context and stack such that when the handler
1275 * returns, it will call function instead of whatever it was doing
1279 #if (defined(LISP_FEATURE_X86) || defined(LISP_FEATURE_X86_64))
1280 extern int *context_eflags_addr(os_context_t *context);
1283 extern lispobj call_into_lisp(lispobj fun, lispobj *args, int nargs);
1284 extern void post_signal_tramp(void);
1285 extern void call_into_lisp_tramp(void);
1287 arrange_return_to_lisp_function(os_context_t *context, lispobj function)
1289 #ifndef LISP_FEATURE_WIN32
1290 check_gc_signals_unblocked_or_lose
1291 (os_context_sigmask_addr(context));
1293 #if !(defined(LISP_FEATURE_X86) || defined(LISP_FEATURE_X86_64))
1294 void * fun=native_pointer(function);
1295 void *code = &(((struct simple_fun *) fun)->code);
1298 /* Build a stack frame showing `interrupted' so that the
1299 * user's backtrace makes (as much) sense (as usual) */
1301 /* fp state is saved and restored by call_into_lisp */
1302 /* FIXME: errno is not restored, but since current uses of this
1303 * function only call Lisp code that signals an error, it's not
1304 * much of a problem. In other words, running out of the control
1305 * stack between a syscall and (GET-ERRNO) may clobber errno if
1306 * something fails during signalling or in the handler. But I
1307 * can't see what can go wrong as long as there is no CONTINUE
1308 * like restart on them. */
1309 #ifdef LISP_FEATURE_X86
1310 /* Suppose the existence of some function that saved all
1311 * registers, called call_into_lisp, then restored GP registers and
1312 * returned. It would look something like this:
1320 pushl {address of function to call}
1321 call 0x8058db0 <call_into_lisp>
1328 * What we do here is set up the stack that call_into_lisp would
1329 * expect to see if it had been called by this code, and frob the
1330 * signal context so that signal return goes directly to call_into_lisp,
1331 * and when that function (and the lisp function it invoked) returns,
1332 * it returns to the second half of this imaginary function which
1333 * restores all registers and returns to C
1335 * For this to work, the latter part of the imaginary function
1336 * must obviously exist in reality. That would be post_signal_tramp
1339 u32 *sp=(u32 *)*os_context_register_addr(context,reg_ESP);
1341 #if defined(LISP_FEATURE_DARWIN)
1342 u32 *register_save_area = (u32 *)os_validate(0, 0x40);
1344 FSHOW_SIGNAL((stderr, "/arrange_return_to_lisp_function: preparing to go to function %x, sp: %x\n", function, sp));
1345 FSHOW_SIGNAL((stderr, "/arrange_return_to_lisp_function: context: %x, &context %x\n", context, &context));
1347 /* 1. os_validate (malloc/mmap) register_save_block
1348 * 2. copy register state into register_save_block
1349 * 3. put a pointer to register_save_block in a register in the context
1350 * 4. set the context's EIP to point to a trampoline which:
1351 * a. builds the fake stack frame from the block
1352 * b. frees the block
1353 * c. calls the function
1356 *register_save_area = *os_context_pc_addr(context);
1357 *(register_save_area + 1) = function;
1358 *(register_save_area + 2) = *os_context_register_addr(context,reg_EDI);
1359 *(register_save_area + 3) = *os_context_register_addr(context,reg_ESI);
1360 *(register_save_area + 4) = *os_context_register_addr(context,reg_EDX);
1361 *(register_save_area + 5) = *os_context_register_addr(context,reg_ECX);
1362 *(register_save_area + 6) = *os_context_register_addr(context,reg_EBX);
1363 *(register_save_area + 7) = *os_context_register_addr(context,reg_EAX);
1364 *(register_save_area + 8) = *context_eflags_addr(context);
1366 *os_context_pc_addr(context) =
1367 (os_context_register_t) call_into_lisp_tramp;
1368 *os_context_register_addr(context,reg_ECX) =
1369 (os_context_register_t) register_save_area;
1372 /* return address for call_into_lisp: */
1373 *(sp-15) = (u32)post_signal_tramp;
1374 *(sp-14) = function; /* args for call_into_lisp : function*/
1375 *(sp-13) = 0; /* arg array */
1376 *(sp-12) = 0; /* no. args */
1377 /* this order matches that used in POPAD */
1378 *(sp-11)=*os_context_register_addr(context,reg_EDI);
1379 *(sp-10)=*os_context_register_addr(context,reg_ESI);
1381 *(sp-9)=*os_context_register_addr(context,reg_ESP)-8;
1382 /* POPAD ignores the value of ESP: */
1384 *(sp-7)=*os_context_register_addr(context,reg_EBX);
1386 *(sp-6)=*os_context_register_addr(context,reg_EDX);
1387 *(sp-5)=*os_context_register_addr(context,reg_ECX);
1388 *(sp-4)=*os_context_register_addr(context,reg_EAX);
1389 *(sp-3)=*context_eflags_addr(context);
1390 *(sp-2)=*os_context_register_addr(context,reg_EBP);
1391 *(sp-1)=*os_context_pc_addr(context);
1395 #elif defined(LISP_FEATURE_X86_64)
1396 u64 *sp=(u64 *)*os_context_register_addr(context,reg_RSP);
1398 /* return address for call_into_lisp: */
1399 *(sp-18) = (u64)post_signal_tramp;
1401 *(sp-17)=*os_context_register_addr(context,reg_R15);
1402 *(sp-16)=*os_context_register_addr(context,reg_R14);
1403 *(sp-15)=*os_context_register_addr(context,reg_R13);
1404 *(sp-14)=*os_context_register_addr(context,reg_R12);
1405 *(sp-13)=*os_context_register_addr(context,reg_R11);
1406 *(sp-12)=*os_context_register_addr(context,reg_R10);
1407 *(sp-11)=*os_context_register_addr(context,reg_R9);
1408 *(sp-10)=*os_context_register_addr(context,reg_R8);
1409 *(sp-9)=*os_context_register_addr(context,reg_RDI);
1410 *(sp-8)=*os_context_register_addr(context,reg_RSI);
1411 /* skip RBP and RSP */
1412 *(sp-7)=*os_context_register_addr(context,reg_RBX);
1413 *(sp-6)=*os_context_register_addr(context,reg_RDX);
1414 *(sp-5)=*os_context_register_addr(context,reg_RCX);
1415 *(sp-4)=*os_context_register_addr(context,reg_RAX);
1416 *(sp-3)=*context_eflags_addr(context);
1417 *(sp-2)=*os_context_register_addr(context,reg_RBP);
1418 *(sp-1)=*os_context_pc_addr(context);
1420 *os_context_register_addr(context,reg_RDI) =
1421 (os_context_register_t)function; /* function */
1422 *os_context_register_addr(context,reg_RSI) = 0; /* arg. array */
1423 *os_context_register_addr(context,reg_RDX) = 0; /* no. args */
1425 struct thread *th=arch_os_get_current_thread();
1426 build_fake_control_stack_frames(th,context);
1429 #ifdef LISP_FEATURE_X86
1431 #if !defined(LISP_FEATURE_DARWIN)
1432 *os_context_pc_addr(context) = (os_context_register_t)call_into_lisp;
1433 *os_context_register_addr(context,reg_ECX) = 0;
1434 *os_context_register_addr(context,reg_EBP) = (os_context_register_t)(sp-2);
1436 *os_context_register_addr(context,reg_UESP) =
1437 (os_context_register_t)(sp-15);
1439 *os_context_register_addr(context,reg_ESP) = (os_context_register_t)(sp-15);
1440 #endif /* __NETBSD__ */
1441 #endif /* LISP_FEATURE_DARWIN */
1443 #elif defined(LISP_FEATURE_X86_64)
1444 *os_context_pc_addr(context) = (os_context_register_t)call_into_lisp;
1445 *os_context_register_addr(context,reg_RCX) = 0;
1446 *os_context_register_addr(context,reg_RBP) = (os_context_register_t)(sp-2);
1447 *os_context_register_addr(context,reg_RSP) = (os_context_register_t)(sp-18);
1449 /* this much of the calling convention is common to all
1451 *os_context_pc_addr(context) = (os_context_register_t)(unsigned long)code;
1452 *os_context_register_addr(context,reg_NARGS) = 0;
1453 *os_context_register_addr(context,reg_LIP) =
1454 (os_context_register_t)(unsigned long)code;
1455 *os_context_register_addr(context,reg_CFP) =
1456 (os_context_register_t)(unsigned long)current_control_frame_pointer;
1458 #ifdef ARCH_HAS_NPC_REGISTER
1459 *os_context_npc_addr(context) =
1460 4 + *os_context_pc_addr(context);
1462 #ifdef LISP_FEATURE_SPARC
1463 *os_context_register_addr(context,reg_CODE) =
1464 (os_context_register_t)(fun + FUN_POINTER_LOWTAG);
1466 FSHOW((stderr, "/arranged return to Lisp function (0x%lx)\n",
1470 /* KLUDGE: Theoretically the approach we use for undefined alien
1471 * variables should work for functions as well, but on PPC/Darwin
1472 * we get bus error at bogus addresses instead, hence this workaround,
1473 * that has the added benefit of automatically discriminating between
1474 * functions and variables.
1477 undefined_alien_function(void)
1479 funcall0(StaticSymbolFunction(UNDEFINED_ALIEN_FUNCTION_ERROR));
1483 handle_guard_page_triggered(os_context_t *context,os_vm_address_t addr)
1485 struct thread *th=arch_os_get_current_thread();
1487 if(addr >= CONTROL_STACK_HARD_GUARD_PAGE(th) &&
1488 addr < CONTROL_STACK_HARD_GUARD_PAGE(th) + os_vm_page_size) {
1489 lose("Control stack exhausted");
1491 else if(addr >= CONTROL_STACK_GUARD_PAGE(th) &&
1492 addr < CONTROL_STACK_GUARD_PAGE(th) + os_vm_page_size) {
1493 /* We hit the end of the control stack: disable guard page
1494 * protection so the error handler has some headroom, protect the
1495 * previous page so that we can catch returns from the guard page
1496 * and restore it. */
1497 protect_control_stack_guard_page(0, NULL);
1498 protect_control_stack_return_guard_page(1, NULL);
1499 fprintf(stderr, "INFO: Control stack guard page unprotected\n");
1501 #ifdef LISP_FEATURE_C_STACK_IS_CONTROL_STACK
1502 /* For the unfortunate case, when the control stack is
1503 * exhausted in a signal handler. */
1504 unblock_signals_in_context_and_maybe_warn(context);
1506 arrange_return_to_lisp_function
1507 (context, StaticSymbolFunction(CONTROL_STACK_EXHAUSTED_ERROR));
1510 else if(addr >= CONTROL_STACK_RETURN_GUARD_PAGE(th) &&
1511 addr < CONTROL_STACK_RETURN_GUARD_PAGE(th) + os_vm_page_size) {
1512 /* We're returning from the guard page: reprotect it, and
1513 * unprotect this one. This works even if we somehow missed
1514 * the return-guard-page, and hit it on our way to new
1515 * exhaustion instead. */
1516 protect_control_stack_guard_page(1, NULL);
1517 protect_control_stack_return_guard_page(0, NULL);
1518 fprintf(stderr, "INFO: Control stack guard page reprotected\n");
1521 else if(addr >= BINDING_STACK_HARD_GUARD_PAGE(th) &&
1522 addr < BINDING_STACK_HARD_GUARD_PAGE(th) + os_vm_page_size) {
1523 lose("Binding stack exhausted");
1525 else if(addr >= BINDING_STACK_GUARD_PAGE(th) &&
1526 addr < BINDING_STACK_GUARD_PAGE(th) + os_vm_page_size) {
1527 protect_binding_stack_guard_page(0, NULL);
1528 protect_binding_stack_return_guard_page(1, NULL);
1529 fprintf(stderr, "INFO: Binding stack guard page unprotected\n");
1531 /* For the unfortunate case, when the binding stack is
1532 * exhausted in a signal handler. */
1533 unblock_signals_in_context_and_maybe_warn(context);
1534 arrange_return_to_lisp_function
1535 (context, StaticSymbolFunction(BINDING_STACK_EXHAUSTED_ERROR));
1538 else if(addr >= BINDING_STACK_RETURN_GUARD_PAGE(th) &&
1539 addr < BINDING_STACK_RETURN_GUARD_PAGE(th) + os_vm_page_size) {
1540 protect_binding_stack_guard_page(1, NULL);
1541 protect_binding_stack_return_guard_page(0, NULL);
1542 fprintf(stderr, "INFO: Binding stack guard page reprotected\n");
1545 else if(addr >= ALIEN_STACK_HARD_GUARD_PAGE(th) &&
1546 addr < ALIEN_STACK_HARD_GUARD_PAGE(th) + os_vm_page_size) {
1547 lose("Alien stack exhausted");
1549 else if(addr >= ALIEN_STACK_GUARD_PAGE(th) &&
1550 addr < ALIEN_STACK_GUARD_PAGE(th) + os_vm_page_size) {
1551 protect_alien_stack_guard_page(0, NULL);
1552 protect_alien_stack_return_guard_page(1, NULL);
1553 fprintf(stderr, "INFO: Alien stack guard page unprotected\n");
1555 /* For the unfortunate case, when the alien stack is
1556 * exhausted in a signal handler. */
1557 unblock_signals_in_context_and_maybe_warn(context);
1558 arrange_return_to_lisp_function
1559 (context, StaticSymbolFunction(ALIEN_STACK_EXHAUSTED_ERROR));
1562 else if(addr >= ALIEN_STACK_RETURN_GUARD_PAGE(th) &&
1563 addr < ALIEN_STACK_RETURN_GUARD_PAGE(th) + os_vm_page_size) {
1564 protect_alien_stack_guard_page(1, NULL);
1565 protect_alien_stack_return_guard_page(0, NULL);
1566 fprintf(stderr, "INFO: Alien stack guard page reprotected\n");
1569 else if (addr >= undefined_alien_address &&
1570 addr < undefined_alien_address + os_vm_page_size) {
1571 arrange_return_to_lisp_function
1572 (context, StaticSymbolFunction(UNDEFINED_ALIEN_VARIABLE_ERROR));
1579 * noise to install handlers
1582 #ifndef LISP_FEATURE_WIN32
1583 /* In Linux 2.4 synchronous signals (sigtrap & co) can be delivered if
1584 * they are blocked, in Linux 2.6 the default handler is invoked
1585 * instead that usually coredumps. One might hastily think that adding
1586 * SA_NODEFER helps, but until ~2.6.13 if SA_NODEFER is specified then
1587 * the whole sa_mask is ignored and instead of not adding the signal
1588 * in question to the mask. That means if it's not blockable the
1589 * signal must be unblocked at the beginning of signal handlers.
1591 * It turns out that NetBSD's SA_NODEFER doesn't DTRT in a different
1592 * way: if SA_NODEFER is set and the signal is in sa_mask, the signal
1593 * will be unblocked in the sigmask during the signal handler. -- RMK
1596 static volatile int sigaction_nodefer_works = -1;
1598 #define SA_NODEFER_TEST_BLOCK_SIGNAL SIGABRT
1599 #define SA_NODEFER_TEST_KILL_SIGNAL SIGUSR1
1602 sigaction_nodefer_test_handler(int signal, siginfo_t *info, void *void_context)
1606 get_current_sigmask(¤t);
1607 /* There should be exactly two blocked signals: the two we added
1608 * to sa_mask when setting up the handler. NetBSD doesn't block
1609 * the signal we're handling when SA_NODEFER is set; Linux before
1610 * 2.6.13 or so also doesn't block the other signal when
1611 * SA_NODEFER is set. */
1612 for(i = 1; i < NSIG; i++)
1613 if (sigismember(¤t, i) !=
1614 (((i == SA_NODEFER_TEST_BLOCK_SIGNAL) || (i == signal)) ? 1 : 0)) {
1615 FSHOW_SIGNAL((stderr, "SA_NODEFER doesn't work, signal %d\n", i));
1616 sigaction_nodefer_works = 0;
1618 if (sigaction_nodefer_works == -1)
1619 sigaction_nodefer_works = 1;
1623 see_if_sigaction_nodefer_works(void)
1625 struct sigaction sa, old_sa;
1627 sa.sa_flags = SA_SIGINFO | SA_NODEFER;
1628 sa.sa_sigaction = sigaction_nodefer_test_handler;
1629 sigemptyset(&sa.sa_mask);
1630 sigaddset(&sa.sa_mask, SA_NODEFER_TEST_BLOCK_SIGNAL);
1631 sigaddset(&sa.sa_mask, SA_NODEFER_TEST_KILL_SIGNAL);
1632 sigaction(SA_NODEFER_TEST_KILL_SIGNAL, &sa, &old_sa);
1633 /* Make sure no signals are blocked. */
1636 sigemptyset(&empty);
1637 thread_sigmask(SIG_SETMASK, &empty, 0);
1639 kill(getpid(), SA_NODEFER_TEST_KILL_SIGNAL);
1640 while (sigaction_nodefer_works == -1);
1641 sigaction(SA_NODEFER_TEST_KILL_SIGNAL, &old_sa, NULL);
1644 #undef SA_NODEFER_TEST_BLOCK_SIGNAL
1645 #undef SA_NODEFER_TEST_KILL_SIGNAL
1648 unblock_me_trampoline(int signal, siginfo_t *info, void *void_context)
1650 SAVE_ERRNO(signal,context,void_context);
1653 sigemptyset(&unblock);
1654 sigaddset(&unblock, signal);
1655 thread_sigmask(SIG_UNBLOCK, &unblock, 0);
1656 interrupt_handle_now(signal, info, context);
1661 low_level_unblock_me_trampoline(int signal, siginfo_t *info, void *void_context)
1663 SAVE_ERRNO(signal,context,void_context);
1666 sigemptyset(&unblock);
1667 sigaddset(&unblock, signal);
1668 thread_sigmask(SIG_UNBLOCK, &unblock, 0);
1669 (*interrupt_low_level_handlers[signal])(signal, info, context);
1674 low_level_handle_now_handler(int signal, siginfo_t *info, void *void_context)
1676 SAVE_ERRNO(signal,context,void_context);
1677 (*interrupt_low_level_handlers[signal])(signal, info, context);
1682 undoably_install_low_level_interrupt_handler (int signal,
1683 interrupt_handler_t handler)
1685 struct sigaction sa;
1687 if (0 > signal || signal >= NSIG) {
1688 lose("bad signal number %d\n", signal);
1691 if (ARE_SAME_HANDLER(handler, SIG_DFL))
1692 sa.sa_sigaction = (void (*)(int, siginfo_t*, void*))handler;
1693 else if (sigismember(&deferrable_sigset,signal))
1694 sa.sa_sigaction = low_level_maybe_now_maybe_later;
1695 else if (!sigaction_nodefer_works &&
1696 !sigismember(&blockable_sigset, signal))
1697 sa.sa_sigaction = low_level_unblock_me_trampoline;
1699 sa.sa_sigaction = low_level_handle_now_handler;
1701 sigcopyset(&sa.sa_mask, &blockable_sigset);
1702 sa.sa_flags = SA_SIGINFO | SA_RESTART
1703 | (sigaction_nodefer_works ? SA_NODEFER : 0);
1704 #ifdef LISP_FEATURE_C_STACK_IS_CONTROL_STACK
1705 if((signal==SIG_MEMORY_FAULT))
1706 sa.sa_flags |= SA_ONSTACK;
1709 sigaction(signal, &sa, NULL);
1710 interrupt_low_level_handlers[signal] =
1711 (ARE_SAME_HANDLER(handler, SIG_DFL) ? 0 : handler);
1715 /* This is called from Lisp. */
1717 install_handler(int signal, void handler(int, siginfo_t*, os_context_t*))
1719 #ifndef LISP_FEATURE_WIN32
1720 struct sigaction sa;
1722 union interrupt_handler oldhandler;
1724 FSHOW((stderr, "/entering POSIX install_handler(%d, ..)\n", signal));
1726 block_blockable_signals(0, &old);
1728 FSHOW((stderr, "/interrupt_low_level_handlers[signal]=%x\n",
1729 (unsigned int)interrupt_low_level_handlers[signal]));
1730 if (interrupt_low_level_handlers[signal]==0) {
1731 if (ARE_SAME_HANDLER(handler, SIG_DFL) ||
1732 ARE_SAME_HANDLER(handler, SIG_IGN))
1733 sa.sa_sigaction = (void (*)(int, siginfo_t*, void*))handler;
1734 else if (sigismember(&deferrable_sigset, signal))
1735 sa.sa_sigaction = maybe_now_maybe_later;
1736 else if (!sigaction_nodefer_works &&
1737 !sigismember(&blockable_sigset, signal))
1738 sa.sa_sigaction = unblock_me_trampoline;
1740 sa.sa_sigaction = interrupt_handle_now_handler;
1742 sigcopyset(&sa.sa_mask, &blockable_sigset);
1743 sa.sa_flags = SA_SIGINFO | SA_RESTART |
1744 (sigaction_nodefer_works ? SA_NODEFER : 0);
1745 sigaction(signal, &sa, NULL);
1748 oldhandler = interrupt_handlers[signal];
1749 interrupt_handlers[signal].c = handler;
1751 thread_sigmask(SIG_SETMASK, &old, 0);
1753 FSHOW((stderr, "/leaving POSIX install_handler(%d, ..)\n", signal));
1755 return (unsigned long)oldhandler.lisp;
1757 /* Probably-wrong Win32 hack */
1762 /* This must not go through lisp as it's allowed anytime, even when on
1765 sigabrt_handler(int signal, siginfo_t *info, os_context_t *context)
1767 lose("SIGABRT received.\n");
1771 interrupt_init(void)
1773 #ifndef LISP_FEATURE_WIN32
1775 SHOW("entering interrupt_init()");
1776 see_if_sigaction_nodefer_works();
1777 sigemptyset(&deferrable_sigset);
1778 sigemptyset(&blockable_sigset);
1779 sigemptyset(&gc_sigset);
1780 sigaddset_deferrable(&deferrable_sigset);
1781 sigaddset_blockable(&blockable_sigset);
1782 sigaddset_gc(&gc_sigset);
1784 /* Set up high level handler information. */
1785 for (i = 0; i < NSIG; i++) {
1786 interrupt_handlers[i].c =
1787 /* (The cast here blasts away the distinction between
1788 * SA_SIGACTION-style three-argument handlers and
1789 * signal(..)-style one-argument handlers, which is OK
1790 * because it works to call the 1-argument form where the
1791 * 3-argument form is expected.) */
1792 (void (*)(int, siginfo_t*, os_context_t*))SIG_DFL;
1794 undoably_install_low_level_interrupt_handler(SIGABRT, sigabrt_handler);
1795 SHOW("returning from interrupt_init()");
1799 #ifndef LISP_FEATURE_WIN32
1801 siginfo_code(siginfo_t *info)
1803 return info->si_code;
1805 os_vm_address_t current_memory_fault_address;
1808 lisp_memory_fault_error(os_context_t *context, os_vm_address_t addr)
1810 /* FIXME: This is lossy: if we get another memory fault (eg. from
1811 * another thread) before lisp has read this, we lose the information.
1812 * However, since this is mostly informative, we'll live with that for
1813 * now -- some address is better then no address in this case.
1815 current_memory_fault_address = addr;
1816 /* To allow debugging memory faults in signal handlers and such. */
1817 corruption_warning_and_maybe_lose("Memory fault at %x (pc=%p, sp=%p)",
1819 *os_context_pc_addr(context),
1820 #ifdef ARCH_HAS_STACK_POINTER
1821 *os_context_sp_addr(context)
1826 unblock_signals_in_context_and_maybe_warn(context);
1827 #ifdef LISP_FEATURE_C_STACK_IS_CONTROL_STACK
1828 arrange_return_to_lisp_function(context,
1829 StaticSymbolFunction(MEMORY_FAULT_ERROR));
1831 funcall0(StaticSymbolFunction(MEMORY_FAULT_ERROR));
1837 unhandled_trap_error(os_context_t *context)
1839 lispobj context_sap;
1840 fake_foreign_function_call(context);
1841 unblock_gc_signals(0, 0);
1842 context_sap = alloc_sap(context);
1843 #ifndef LISP_FEATURE_WIN32
1844 thread_sigmask(SIG_SETMASK, os_context_sigmask_addr(context), 0);
1846 funcall1(StaticSymbolFunction(UNHANDLED_TRAP_ERROR), context_sap);
1847 lose("UNHANDLED-TRAP-ERROR fell through");
1850 /* Common logic for trapping instructions. How we actually handle each
1851 * case is highly architecture dependent, but the overall shape is
1854 handle_trap(os_context_t *context, int trap)
1857 case trap_PendingInterrupt:
1858 FSHOW((stderr, "/<trap pending interrupt>\n"));
1859 arch_skip_instruction(context);
1860 interrupt_handle_pending(context);
1864 FSHOW((stderr, "/<trap error/cerror %d>\n", trap));
1865 interrupt_internal_error(context, trap==trap_Cerror);
1867 case trap_Breakpoint:
1868 arch_handle_breakpoint(context);
1870 case trap_FunEndBreakpoint:
1871 arch_handle_fun_end_breakpoint(context);
1873 #ifdef trap_AfterBreakpoint
1874 case trap_AfterBreakpoint:
1875 arch_handle_after_breakpoint(context);
1878 #ifdef trap_SingleStepAround
1879 case trap_SingleStepAround:
1880 case trap_SingleStepBefore:
1881 arch_handle_single_step_trap(context, trap);
1885 fake_foreign_function_call(context);
1886 lose("%%PRIMITIVE HALT called; the party is over.\n");
1888 unhandled_trap_error(context);