guile/libguile/continuations.c

524 lines
14 KiB
C
Raw Normal View History

/* Copyright (C) 1995,1996,1998,2000,2001,2004, 2006, 2008, 2009, 2010, 2011, 2012, 2013, 2014 Free Software Foundation, Inc.
*
* This 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 3 of
* the License, or (at your option) any later version.
*
* This 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 this library; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
* 02110-1301 USA
*/
#ifdef HAVE_CONFIG_H
# include <config.h>
#endif
#include "libguile/_scm.h"
the dynamic stack is really a stack now, instead of a list * libguile/dynstack.h: * libguile/dynstack.c: New files, implementing the dynamic stack as a true stack instead of a linked list. This lowers the cost of dynwinds: frames, winders, prompts, with-fluids, and dynamic-wind. For the most part, we allocate these items directly on the stack. * libguile/dynwinds.h: * libguile/dynwinds.c: Adapt all manipulators of the wind stack to use interfaces from dynstack.c. Remove heap-allocated winder and frame object types. (scm_dowinds, scm_i_dowinds): Remove these. The first was exported, but it was not a public interface. * libguile/continuations.c: * libguile/continuations.h (scm_t_contregs): Continuation objects reference scm_t_dynstack* values now. Adapt to the new interfaces. * libguile/control.c: * libguile/control.h: There is no longer a scm_tc7_prompt kind of object that can be allocated on the heap. Instead, the prompt flags, key, and registers are pushed on the dynwind stack. (The registers are still on the heap.) Also, since the vm_cont will reference the dynwinds, make the partial continuation stub take just one extra arg, instead of storing the intwinds separately in the object table. * libguile/fluids.c: * libguile/fluids.h: No more with-fluids objects; instead, the fluids go on the dynstack. The values still have to be on the heap, though. (scm_prepare_fluids, scm_swap_fluids): New internal functions, replacing scm_i_make_with_fluids and scm_i_swap_with_fluids. * libguile/print.c: Remove prompt and with-fluids printers. * libguile/tags.h: Revert prompt and with-fluids tc7 values to what they were before they were allocated. * libguile/vm-i-system.c (partial_cont_call): Just pop the vmcont, the intwinds will not be passed as a second arg. Rewind the dynamic stack from within the VM, so that any rewinder sees valid prompt entries. (call_cc, tail_call_cc): Adapt to pass the dynstack to scm_i_vm_capture_stack. (prompt, wind, unwind, wind_fluids, unwind_fluids): Adapt to the new interfaces. * libguile/vm.h (scm_i_capture_current_stack): Rename from scm_i_vm_capture_continuation. (scm_i_vm_capture_stack): Take a dynstack as an argument. * libguile/vm.c (vm_reinstate_partial_continuation): Don't wind here, as that could result in winders seeing invalid prompts. * libguile/eval.c: * libguile/root.c: * libguile/stacks.c: * libguile/threads.c: * libguile/threads.h: * libguile/throw.c: Adapt other users of dynwinds to use the dynstack.
2012-03-03 17:01:16 +01:00
#include <assert.h>
2001-01-26 17:30:54 +00:00
#include <string.h>
#include <stdio.h>
2001-01-26 17:30:54 +00:00
#include "libguile/async.h"
#include "libguile/debug.h"
#include "libguile/root.h"
#include "libguile/stackchk.h"
#include "libguile/smob.h"
#include "libguile/ports.h"
the dynamic stack is really a stack now, instead of a list * libguile/dynstack.h: * libguile/dynstack.c: New files, implementing the dynamic stack as a true stack instead of a linked list. This lowers the cost of dynwinds: frames, winders, prompts, with-fluids, and dynamic-wind. For the most part, we allocate these items directly on the stack. * libguile/dynwinds.h: * libguile/dynwinds.c: Adapt all manipulators of the wind stack to use interfaces from dynstack.c. Remove heap-allocated winder and frame object types. (scm_dowinds, scm_i_dowinds): Remove these. The first was exported, but it was not a public interface. * libguile/continuations.c: * libguile/continuations.h (scm_t_contregs): Continuation objects reference scm_t_dynstack* values now. Adapt to the new interfaces. * libguile/control.c: * libguile/control.h: There is no longer a scm_tc7_prompt kind of object that can be allocated on the heap. Instead, the prompt flags, key, and registers are pushed on the dynwind stack. (The registers are still on the heap.) Also, since the vm_cont will reference the dynwinds, make the partial continuation stub take just one extra arg, instead of storing the intwinds separately in the object table. * libguile/fluids.c: * libguile/fluids.h: No more with-fluids objects; instead, the fluids go on the dynstack. The values still have to be on the heap, though. (scm_prepare_fluids, scm_swap_fluids): New internal functions, replacing scm_i_make_with_fluids and scm_i_swap_with_fluids. * libguile/print.c: Remove prompt and with-fluids printers. * libguile/tags.h: Revert prompt and with-fluids tc7 values to what they were before they were allocated. * libguile/vm-i-system.c (partial_cont_call): Just pop the vmcont, the intwinds will not be passed as a second arg. Rewind the dynamic stack from within the VM, so that any rewinder sees valid prompt entries. (call_cc, tail_call_cc): Adapt to pass the dynstack to scm_i_vm_capture_stack. (prompt, wind, unwind, wind_fluids, unwind_fluids): Adapt to the new interfaces. * libguile/vm.h (scm_i_capture_current_stack): Rename from scm_i_vm_capture_continuation. (scm_i_vm_capture_stack): Take a dynstack as an argument. * libguile/vm.c (vm_reinstate_partial_continuation): Don't wind here, as that could result in winders seeing invalid prompts. * libguile/eval.c: * libguile/root.c: * libguile/stacks.c: * libguile/threads.c: * libguile/threads.h: * libguile/throw.c: Adapt other users of dynwinds to use the dynstack.
2012-03-03 17:01:16 +01:00
#include "libguile/dynstack.h"
2005-03-02 20:42:01 +00:00
#include "libguile/eval.h"
#include "libguile/vm.h"
#include "libguile/instructions.h"
2001-03-04 17:09:34 +00:00
#include "libguile/validate.h"
#include "libguile/continuations.h"
static scm_t_bits tc16_continuation;
#define SCM_CONTREGSP(x) SCM_TYP16_PREDICATE (tc16_continuation, x)
#define SCM_CONTREGS(x) ((scm_t_contregs *) SCM_SMOB_DATA_1 (x))
#define SCM_CONTINUATION_LENGTH(x) (SCM_CONTREGS (x)->num_stack_items)
#define SCM_SET_CONTINUATION_LENGTH(x, n)\
(SCM_CONTREGS (x)->num_stack_items = (n))
#define SCM_JMPBUF(x) ((SCM_CONTREGS (x))->jmpbuf)
#define SCM_CONTINUATION_ROOT(x) ((SCM_CONTREGS (x))->root)
#define SCM_DFRAME(x) ((SCM_CONTREGS (x))->dframe)
/* scm_i_make_continuation will return a procedure whose code will
reinstate the continuation. Here, as in gsubr.c, we define the form
of that trampoline function.
*/
static const scm_t_uint32 continuation_stub_code[] =
{
SCM_PACK_OP_24 (continuation_call, 0)
};
static SCM
make_continuation_trampoline (SCM contregs)
{
SCM ret;
scm_t_bits nfree = 1;
scm_t_bits flags = SCM_F_PROGRAM_IS_CONTINUATION;
ret = scm_words (scm_tc7_program | (nfree << 16) | flags, nfree + 2);
SCM_SET_CELL_WORD_1 (ret, continuation_stub_code);
SCM_PROGRAM_FREE_VARIABLE_SET (ret, 0, contregs);
return ret;
}
/* {Continuations}
*/
2000-12-08 17:32:56 +00:00
static int
continuation_print (SCM obj, SCM port, scm_print_state *state SCM_UNUSED)
{
2001-06-14 19:50:43 +00:00
scm_t_contregs *continuation = SCM_CONTREGS (obj);
scm_puts_unlocked ("#<continuation ", port);
scm_intprint (continuation->num_stack_items, 10, port);
scm_puts_unlocked (" @ ", port);
scm_uintprint (SCM_SMOB_DATA_1 (obj), 16, port);
scm_putc_unlocked ('>', port);
return 1;
}
/* James Clark came up with this neat one instruction fix for
* continuations on the SPARC. It flushes the register windows so
* that all the state of the process is contained in the stack.
*/
#if defined (sparc) || defined (__sparc__) || defined (__sparc)
# define SCM_FLUSH_REGISTER_WINDOWS asm("ta 3")
#else
# define SCM_FLUSH_REGISTER_WINDOWS /* empty */
#endif
/* this may return more than once: the first time with the escape
continuations return multiple values on the stack * libguile/vm.h (struct scm_vm_cont): Instead of saving the "IP", save "RA" and "MVRA". That is, save singly-valued and multiply-valued return addresses, so that we can return multiple values on the stack. (scm_i_vm_reinstate_continuation): Remove. * libguile/vm.c (vm_capture_continuation): Rename from capture_vm_cont, and change the prototype so we can capture the RA and MVRA, and so that tail calls to call/cc can capture a continuation without the call/cc application frame. (vm_return_to_continuation): Rename from reinstate_vm_cont, and take arguments to return to the continuation. Handles returning to single or multiple-value RA. (scm_i_vm_capture_continuation): Change to invoke vm_capture_continuation. Kept around for the benefit of make-stack. * libguile/vm-i-system.c (continuation-call): Handle reinstatement of the VM stack, with arguments. (call/cc, tail-call/cc): Adapt to new vm_capture_continuation prototype. tail-call/cc captures tail continuations. * libguile/stacks.c (scm_make_stack): Update for scm_vm_cont structure change. * libguile/continuations.h (struct scm_contregs): Remove throw_value member, which was used to return a value to a continuation. (scm_i_check_continuation): New internal function, checks that a continuation may be reinstated. (scm_i_reinstate_continuation): Replaces scm_i_continuation_call; just reinstates the C stack. (scm_i_contregs_vm, scm_i_contregs_vm_cont): New internal accessors. * libguile/continuations.c (scm_i_make_continuation): Return SCM_UNDEFINED if we are returning again. (grow_stack, copy_stack_and_call, scm_dynthrow): Remove extra arg, as vm opcodes handle value returns. (copy_stack): No need to instate VM continuation. (scm_i_reinstate_continuation): Adapt.
2010-02-08 22:59:25 +01:00
procedure, then subsequently with SCM_UNDEFINED (the vals already having been
placed on the VM stack). */
#define FUNC_NAME "scm_i_make_continuation"
SCM
scm_i_make_continuation (int *first, struct scm_vm *vp, SCM vm_cont)
{
2005-03-02 20:42:01 +00:00
scm_i_thread *thread = SCM_I_CURRENT_THREAD;
SCM cont;
2001-06-14 19:50:43 +00:00
scm_t_contregs *continuation;
long stack_size;
SCM_STACKITEM * src;
SCM_FLUSH_REGISTER_WINDOWS;
2005-03-02 20:42:01 +00:00
stack_size = scm_stack_size (thread->continuation_base);
continuation = scm_gc_malloc (sizeof (scm_t_contregs)
+ (stack_size - 1) * sizeof (SCM_STACKITEM),
"continuation");
continuation->num_stack_items = stack_size;
2005-03-02 20:42:01 +00:00
continuation->root = thread->continuation_root;
src = thread->continuation_base;
#if ! SCM_STACK_GROWS_UP
src -= stack_size;
#endif
continuation->offset = continuation->stack - src;
memcpy (continuation->stack, src, sizeof (SCM_STACKITEM) * stack_size);
continuation->vp = vp;
continuation->vm_cont = vm_cont;
SCM_NEWSMOB (cont, tc16_continuation, continuation);
*first = !SCM_I_SETJMP (continuation->jmpbuf);
Fix continuation problems on IA64. * Specific problems in IA64 make check ** test-unwind Representation of the relevant dynamic context: non-rewindable catch frame make cont. o----o-----a----------b-------------c \ \ call cont. o-----o-----------d A continuation is captured at (c), with a non-rewindable frame in the dynamic context at (b). If a rewind through that frame was attempted, Guile would throw to the catch at (a). Then the context unwinds back past (a), then winds forwards again, and the captured continuation is called at (d). We should end up at the catch at (a). On ia64, we get an "illegal instruction". The problem is that Guile does not restore the ia64 register backing store (RBS) stack (which is saved off when the continuation is captured) until all the unwinding and rewinding is done. Therefore, when the rewind code (scm_i_dowinds) hits the non-rewindable frame at (b), the RBS stack hasn't yet been restored. The throw finds the jmp_buf (for the catch at (a)) correctly from the dynamic context, and jumps back to (a), but the RBS stack is invalid, hence the illegal instruction. This could be fixed by restoring the RBS stack earlier, at the same point (copy_stack) where the normal stack is restored. But that causes a problem in the next test... ** continuations.test The dynamic context diagram for this case is similar: non-rewindable catch frame make cont. a----x-----o----------b-------------c \ \ call cont. o-------d The only significant difference is that the catch point (a) is upstream of where the dynamic context forks. This means that the RBS stack at (d) already contains the correct RBS contents for throwing back to (a), so it doesn't matter whether the RBS stack that was saved off with the continuation gets restored. This test passes with the Guile 1.8.4 code, but fails (with an "illegal instruction") when the code is changed to restore the RBS stack earlier as described above. The problem now is that the RBS stack is being restored _too_ early; specifically when there is still stuff to do that relies on the old RBS contents. When a continuation is called, the sequence of relevant events is: (1) Grow the (normal) stack until it is bigger than the (normal) stack saved off in the continuation. (scm_dynthrow, grow_stack) (2) scm_i_dowinds calls itself recursively, such that (2.1) for each rewind (from (x) to (c)) that will be needed, another frame is added to the stack (both normal and RBS), with local variables specifying the required rewind; the rewinds don't actually happen yet, they will happen when the stack unwinds again through these frames (2.2) required unwinds - back from where the continuation was called (d) to the fork point (x) - are done immediately. (3) The normal (i.e. non-RBS) stack that was stored in the continuation is restored (i.e. copied on top of the actual stack). Note that this doesn't overwrite the frames that were added in (2.1), because the growth in (1) ensures that the added frames are beyond the end of the restored stack. (4) ? Restore the RBS stack here too ? (5) Return (from copy_stack) through the (2.1) frames, which means that the rewinds now happen. (6) setcontext (or longjmp) to the context (c) where the continuation was captured. The trouble is that step (1) does not create space in the RBS stack in the same kind of way that it does for the normal stack. Therefore, if the saved (in the continuation) RBS stack is big enough, it can overwrite the RBS of the (2.1) frames that still need to complete. This causes an illegal instruction when we return through those frames and try to perform the rewinds. * Fix The key to the fix is that the saved RBS stack only needs to be restored at some point before the next setcontext call, and that doing it as close to the setcontext call as possible will avoid bad interactions with the pre-setcontext stack. Therefore we do the restoration at the last possible point, immediately before the next setcontext call. The situation is complicated by there being two ways that the next setcontext call can happen. - If the unwinding and rewinding is all successful, the next setcontext will be the one from step (6) above. This is the "normal" continuation invocation case. - If one of the rewinds throws an error, the next setcontext will come from the throw implementation code. (And the one in step (6) will never happen.) This is the rewind error case. In the rewind error case, the code calling setcontext knows nothing about the continuation. So to cover both cases, we: - copy (in step (4) above) the address and length of the continuation's saved RBS stack to the current thread state (SCM_I_CURRENT_THREAD) - modify all setcontext callers so that they check the current thread state for a saved RBS stack, and restore it if so before calling setcontext. * Notes ** I think rewinders cannot rely on using any stack data Unless it can be guaranteed that the data won't go into a register. I'm not 100% sure about this, but I think it follows from the fact that the RBS stack is not restored until after the rewinds have happened. Note that this isn't a regression caused by the current fix. In Guile 1.8.4, the RBS stack was restored _after_ the rewinds, and this is still the case now. ** Most setcontext calls for `throw' don't need to change the RBS stack In the absence of continuation invocation, the setcontext call in the throw implementation code always sets context to a place higher up the same stack (both normal and RBS), hence no stack restoration is needed. * Other changes ** Using setcontext for all non-local jumps (for __ia64__) Along the way, I read a claim somewhere that setcontext was more reliable than longjmp, in cases where the stack has been manipulated. I don't now have any reason to believe this, but it seems reasonable anyway to leave the __ia64__ code using getcontext/setcontext, instead of setjmp/longjmp. (I think the only possible argument against this would be performance - if getcontext was significantly slower than setjmp. It that proves to be the case, we should revisit this.) ** Capping RBS base for non-main threads Somewhere else along the way, I hit a problem in GC, involving the RBS stack of a non-main thread. The problem was, in SCM_MARK_BACKING_STORE, that scm_ia64_register_backing_store_base was returning a value that was massively greater than the value of scm_ia64_ar_bsp, leading to a seg fault. This is because the implementation of scm_ia64_register_backing_store_base is only valid for the main thread. I couldn't find a neat way of getting the true RBS base of a non-main thread, but one idea is simply to call scm_ia64_ar_bsp when guilifying a thread, and use the value returned as an upper bound for that thread's RBS base. (Note that the RBS stack grows upwards.) (Were it not for scm_init_guile, we could be much more definitive about this. We could take the value of scm_ia64_ar_bsp as a definitive base address for the part of the RBS stack that Guile cares about. We could also then discard scm_ia64_register_backing_store_base.)
2008-05-08 00:29:53 +01:00
if (*first)
{
Fix continuation problems on IA64. * Specific problems in IA64 make check ** test-unwind Representation of the relevant dynamic context: non-rewindable catch frame make cont. o----o-----a----------b-------------c \ \ call cont. o-----o-----------d A continuation is captured at (c), with a non-rewindable frame in the dynamic context at (b). If a rewind through that frame was attempted, Guile would throw to the catch at (a). Then the context unwinds back past (a), then winds forwards again, and the captured continuation is called at (d). We should end up at the catch at (a). On ia64, we get an "illegal instruction". The problem is that Guile does not restore the ia64 register backing store (RBS) stack (which is saved off when the continuation is captured) until all the unwinding and rewinding is done. Therefore, when the rewind code (scm_i_dowinds) hits the non-rewindable frame at (b), the RBS stack hasn't yet been restored. The throw finds the jmp_buf (for the catch at (a)) correctly from the dynamic context, and jumps back to (a), but the RBS stack is invalid, hence the illegal instruction. This could be fixed by restoring the RBS stack earlier, at the same point (copy_stack) where the normal stack is restored. But that causes a problem in the next test... ** continuations.test The dynamic context diagram for this case is similar: non-rewindable catch frame make cont. a----x-----o----------b-------------c \ \ call cont. o-------d The only significant difference is that the catch point (a) is upstream of where the dynamic context forks. This means that the RBS stack at (d) already contains the correct RBS contents for throwing back to (a), so it doesn't matter whether the RBS stack that was saved off with the continuation gets restored. This test passes with the Guile 1.8.4 code, but fails (with an "illegal instruction") when the code is changed to restore the RBS stack earlier as described above. The problem now is that the RBS stack is being restored _too_ early; specifically when there is still stuff to do that relies on the old RBS contents. When a continuation is called, the sequence of relevant events is: (1) Grow the (normal) stack until it is bigger than the (normal) stack saved off in the continuation. (scm_dynthrow, grow_stack) (2) scm_i_dowinds calls itself recursively, such that (2.1) for each rewind (from (x) to (c)) that will be needed, another frame is added to the stack (both normal and RBS), with local variables specifying the required rewind; the rewinds don't actually happen yet, they will happen when the stack unwinds again through these frames (2.2) required unwinds - back from where the continuation was called (d) to the fork point (x) - are done immediately. (3) The normal (i.e. non-RBS) stack that was stored in the continuation is restored (i.e. copied on top of the actual stack). Note that this doesn't overwrite the frames that were added in (2.1), because the growth in (1) ensures that the added frames are beyond the end of the restored stack. (4) ? Restore the RBS stack here too ? (5) Return (from copy_stack) through the (2.1) frames, which means that the rewinds now happen. (6) setcontext (or longjmp) to the context (c) where the continuation was captured. The trouble is that step (1) does not create space in the RBS stack in the same kind of way that it does for the normal stack. Therefore, if the saved (in the continuation) RBS stack is big enough, it can overwrite the RBS of the (2.1) frames that still need to complete. This causes an illegal instruction when we return through those frames and try to perform the rewinds. * Fix The key to the fix is that the saved RBS stack only needs to be restored at some point before the next setcontext call, and that doing it as close to the setcontext call as possible will avoid bad interactions with the pre-setcontext stack. Therefore we do the restoration at the last possible point, immediately before the next setcontext call. The situation is complicated by there being two ways that the next setcontext call can happen. - If the unwinding and rewinding is all successful, the next setcontext will be the one from step (6) above. This is the "normal" continuation invocation case. - If one of the rewinds throws an error, the next setcontext will come from the throw implementation code. (And the one in step (6) will never happen.) This is the rewind error case. In the rewind error case, the code calling setcontext knows nothing about the continuation. So to cover both cases, we: - copy (in step (4) above) the address and length of the continuation's saved RBS stack to the current thread state (SCM_I_CURRENT_THREAD) - modify all setcontext callers so that they check the current thread state for a saved RBS stack, and restore it if so before calling setcontext. * Notes ** I think rewinders cannot rely on using any stack data Unless it can be guaranteed that the data won't go into a register. I'm not 100% sure about this, but I think it follows from the fact that the RBS stack is not restored until after the rewinds have happened. Note that this isn't a regression caused by the current fix. In Guile 1.8.4, the RBS stack was restored _after_ the rewinds, and this is still the case now. ** Most setcontext calls for `throw' don't need to change the RBS stack In the absence of continuation invocation, the setcontext call in the throw implementation code always sets context to a place higher up the same stack (both normal and RBS), hence no stack restoration is needed. * Other changes ** Using setcontext for all non-local jumps (for __ia64__) Along the way, I read a claim somewhere that setcontext was more reliable than longjmp, in cases where the stack has been manipulated. I don't now have any reason to believe this, but it seems reasonable anyway to leave the __ia64__ code using getcontext/setcontext, instead of setjmp/longjmp. (I think the only possible argument against this would be performance - if getcontext was significantly slower than setjmp. It that proves to be the case, we should revisit this.) ** Capping RBS base for non-main threads Somewhere else along the way, I hit a problem in GC, involving the RBS stack of a non-main thread. The problem was, in SCM_MARK_BACKING_STORE, that scm_ia64_register_backing_store_base was returning a value that was massively greater than the value of scm_ia64_ar_bsp, leading to a seg fault. This is because the implementation of scm_ia64_register_backing_store_base is only valid for the main thread. I couldn't find a neat way of getting the true RBS base of a non-main thread, but one idea is simply to call scm_ia64_ar_bsp when guilifying a thread, and use the value returned as an upper bound for that thread's RBS base. (Note that the RBS stack grows upwards.) (Were it not for scm_init_guile, we could be much more definitive about this. We could take the value of scm_ia64_ar_bsp as a definitive base address for the part of the RBS stack that Guile cares about. We could also then discard scm_ia64_register_backing_store_base.)
2008-05-08 00:29:53 +01:00
#ifdef __ia64__
continuation->backing_store_size =
Fix continuation problems on IA64. * Specific problems in IA64 make check ** test-unwind Representation of the relevant dynamic context: non-rewindable catch frame make cont. o----o-----a----------b-------------c \ \ call cont. o-----o-----------d A continuation is captured at (c), with a non-rewindable frame in the dynamic context at (b). If a rewind through that frame was attempted, Guile would throw to the catch at (a). Then the context unwinds back past (a), then winds forwards again, and the captured continuation is called at (d). We should end up at the catch at (a). On ia64, we get an "illegal instruction". The problem is that Guile does not restore the ia64 register backing store (RBS) stack (which is saved off when the continuation is captured) until all the unwinding and rewinding is done. Therefore, when the rewind code (scm_i_dowinds) hits the non-rewindable frame at (b), the RBS stack hasn't yet been restored. The throw finds the jmp_buf (for the catch at (a)) correctly from the dynamic context, and jumps back to (a), but the RBS stack is invalid, hence the illegal instruction. This could be fixed by restoring the RBS stack earlier, at the same point (copy_stack) where the normal stack is restored. But that causes a problem in the next test... ** continuations.test The dynamic context diagram for this case is similar: non-rewindable catch frame make cont. a----x-----o----------b-------------c \ \ call cont. o-------d The only significant difference is that the catch point (a) is upstream of where the dynamic context forks. This means that the RBS stack at (d) already contains the correct RBS contents for throwing back to (a), so it doesn't matter whether the RBS stack that was saved off with the continuation gets restored. This test passes with the Guile 1.8.4 code, but fails (with an "illegal instruction") when the code is changed to restore the RBS stack earlier as described above. The problem now is that the RBS stack is being restored _too_ early; specifically when there is still stuff to do that relies on the old RBS contents. When a continuation is called, the sequence of relevant events is: (1) Grow the (normal) stack until it is bigger than the (normal) stack saved off in the continuation. (scm_dynthrow, grow_stack) (2) scm_i_dowinds calls itself recursively, such that (2.1) for each rewind (from (x) to (c)) that will be needed, another frame is added to the stack (both normal and RBS), with local variables specifying the required rewind; the rewinds don't actually happen yet, they will happen when the stack unwinds again through these frames (2.2) required unwinds - back from where the continuation was called (d) to the fork point (x) - are done immediately. (3) The normal (i.e. non-RBS) stack that was stored in the continuation is restored (i.e. copied on top of the actual stack). Note that this doesn't overwrite the frames that were added in (2.1), because the growth in (1) ensures that the added frames are beyond the end of the restored stack. (4) ? Restore the RBS stack here too ? (5) Return (from copy_stack) through the (2.1) frames, which means that the rewinds now happen. (6) setcontext (or longjmp) to the context (c) where the continuation was captured. The trouble is that step (1) does not create space in the RBS stack in the same kind of way that it does for the normal stack. Therefore, if the saved (in the continuation) RBS stack is big enough, it can overwrite the RBS of the (2.1) frames that still need to complete. This causes an illegal instruction when we return through those frames and try to perform the rewinds. * Fix The key to the fix is that the saved RBS stack only needs to be restored at some point before the next setcontext call, and that doing it as close to the setcontext call as possible will avoid bad interactions with the pre-setcontext stack. Therefore we do the restoration at the last possible point, immediately before the next setcontext call. The situation is complicated by there being two ways that the next setcontext call can happen. - If the unwinding and rewinding is all successful, the next setcontext will be the one from step (6) above. This is the "normal" continuation invocation case. - If one of the rewinds throws an error, the next setcontext will come from the throw implementation code. (And the one in step (6) will never happen.) This is the rewind error case. In the rewind error case, the code calling setcontext knows nothing about the continuation. So to cover both cases, we: - copy (in step (4) above) the address and length of the continuation's saved RBS stack to the current thread state (SCM_I_CURRENT_THREAD) - modify all setcontext callers so that they check the current thread state for a saved RBS stack, and restore it if so before calling setcontext. * Notes ** I think rewinders cannot rely on using any stack data Unless it can be guaranteed that the data won't go into a register. I'm not 100% sure about this, but I think it follows from the fact that the RBS stack is not restored until after the rewinds have happened. Note that this isn't a regression caused by the current fix. In Guile 1.8.4, the RBS stack was restored _after_ the rewinds, and this is still the case now. ** Most setcontext calls for `throw' don't need to change the RBS stack In the absence of continuation invocation, the setcontext call in the throw implementation code always sets context to a place higher up the same stack (both normal and RBS), hence no stack restoration is needed. * Other changes ** Using setcontext for all non-local jumps (for __ia64__) Along the way, I read a claim somewhere that setcontext was more reliable than longjmp, in cases where the stack has been manipulated. I don't now have any reason to believe this, but it seems reasonable anyway to leave the __ia64__ code using getcontext/setcontext, instead of setjmp/longjmp. (I think the only possible argument against this would be performance - if getcontext was significantly slower than setjmp. It that proves to be the case, we should revisit this.) ** Capping RBS base for non-main threads Somewhere else along the way, I hit a problem in GC, involving the RBS stack of a non-main thread. The problem was, in SCM_MARK_BACKING_STORE, that scm_ia64_register_backing_store_base was returning a value that was massively greater than the value of scm_ia64_ar_bsp, leading to a seg fault. This is because the implementation of scm_ia64_register_backing_store_base is only valid for the main thread. I couldn't find a neat way of getting the true RBS base of a non-main thread, but one idea is simply to call scm_ia64_ar_bsp when guilifying a thread, and use the value returned as an upper bound for that thread's RBS base. (Note that the RBS stack grows upwards.) (Were it not for scm_init_guile, we could be much more definitive about this. We could take the value of scm_ia64_ar_bsp as a definitive base address for the part of the RBS stack that Guile cares about. We could also then discard scm_ia64_register_backing_store_base.)
2008-05-08 00:29:53 +01:00
(char *) scm_ia64_ar_bsp(&continuation->jmpbuf.ctx)
-
Fix continuation problems on IA64. * Specific problems in IA64 make check ** test-unwind Representation of the relevant dynamic context: non-rewindable catch frame make cont. o----o-----a----------b-------------c \ \ call cont. o-----o-----------d A continuation is captured at (c), with a non-rewindable frame in the dynamic context at (b). If a rewind through that frame was attempted, Guile would throw to the catch at (a). Then the context unwinds back past (a), then winds forwards again, and the captured continuation is called at (d). We should end up at the catch at (a). On ia64, we get an "illegal instruction". The problem is that Guile does not restore the ia64 register backing store (RBS) stack (which is saved off when the continuation is captured) until all the unwinding and rewinding is done. Therefore, when the rewind code (scm_i_dowinds) hits the non-rewindable frame at (b), the RBS stack hasn't yet been restored. The throw finds the jmp_buf (for the catch at (a)) correctly from the dynamic context, and jumps back to (a), but the RBS stack is invalid, hence the illegal instruction. This could be fixed by restoring the RBS stack earlier, at the same point (copy_stack) where the normal stack is restored. But that causes a problem in the next test... ** continuations.test The dynamic context diagram for this case is similar: non-rewindable catch frame make cont. a----x-----o----------b-------------c \ \ call cont. o-------d The only significant difference is that the catch point (a) is upstream of where the dynamic context forks. This means that the RBS stack at (d) already contains the correct RBS contents for throwing back to (a), so it doesn't matter whether the RBS stack that was saved off with the continuation gets restored. This test passes with the Guile 1.8.4 code, but fails (with an "illegal instruction") when the code is changed to restore the RBS stack earlier as described above. The problem now is that the RBS stack is being restored _too_ early; specifically when there is still stuff to do that relies on the old RBS contents. When a continuation is called, the sequence of relevant events is: (1) Grow the (normal) stack until it is bigger than the (normal) stack saved off in the continuation. (scm_dynthrow, grow_stack) (2) scm_i_dowinds calls itself recursively, such that (2.1) for each rewind (from (x) to (c)) that will be needed, another frame is added to the stack (both normal and RBS), with local variables specifying the required rewind; the rewinds don't actually happen yet, they will happen when the stack unwinds again through these frames (2.2) required unwinds - back from where the continuation was called (d) to the fork point (x) - are done immediately. (3) The normal (i.e. non-RBS) stack that was stored in the continuation is restored (i.e. copied on top of the actual stack). Note that this doesn't overwrite the frames that were added in (2.1), because the growth in (1) ensures that the added frames are beyond the end of the restored stack. (4) ? Restore the RBS stack here too ? (5) Return (from copy_stack) through the (2.1) frames, which means that the rewinds now happen. (6) setcontext (or longjmp) to the context (c) where the continuation was captured. The trouble is that step (1) does not create space in the RBS stack in the same kind of way that it does for the normal stack. Therefore, if the saved (in the continuation) RBS stack is big enough, it can overwrite the RBS of the (2.1) frames that still need to complete. This causes an illegal instruction when we return through those frames and try to perform the rewinds. * Fix The key to the fix is that the saved RBS stack only needs to be restored at some point before the next setcontext call, and that doing it as close to the setcontext call as possible will avoid bad interactions with the pre-setcontext stack. Therefore we do the restoration at the last possible point, immediately before the next setcontext call. The situation is complicated by there being two ways that the next setcontext call can happen. - If the unwinding and rewinding is all successful, the next setcontext will be the one from step (6) above. This is the "normal" continuation invocation case. - If one of the rewinds throws an error, the next setcontext will come from the throw implementation code. (And the one in step (6) will never happen.) This is the rewind error case. In the rewind error case, the code calling setcontext knows nothing about the continuation. So to cover both cases, we: - copy (in step (4) above) the address and length of the continuation's saved RBS stack to the current thread state (SCM_I_CURRENT_THREAD) - modify all setcontext callers so that they check the current thread state for a saved RBS stack, and restore it if so before calling setcontext. * Notes ** I think rewinders cannot rely on using any stack data Unless it can be guaranteed that the data won't go into a register. I'm not 100% sure about this, but I think it follows from the fact that the RBS stack is not restored until after the rewinds have happened. Note that this isn't a regression caused by the current fix. In Guile 1.8.4, the RBS stack was restored _after_ the rewinds, and this is still the case now. ** Most setcontext calls for `throw' don't need to change the RBS stack In the absence of continuation invocation, the setcontext call in the throw implementation code always sets context to a place higher up the same stack (both normal and RBS), hence no stack restoration is needed. * Other changes ** Using setcontext for all non-local jumps (for __ia64__) Along the way, I read a claim somewhere that setcontext was more reliable than longjmp, in cases where the stack has been manipulated. I don't now have any reason to believe this, but it seems reasonable anyway to leave the __ia64__ code using getcontext/setcontext, instead of setjmp/longjmp. (I think the only possible argument against this would be performance - if getcontext was significantly slower than setjmp. It that proves to be the case, we should revisit this.) ** Capping RBS base for non-main threads Somewhere else along the way, I hit a problem in GC, involving the RBS stack of a non-main thread. The problem was, in SCM_MARK_BACKING_STORE, that scm_ia64_register_backing_store_base was returning a value that was massively greater than the value of scm_ia64_ar_bsp, leading to a seg fault. This is because the implementation of scm_ia64_register_backing_store_base is only valid for the main thread. I couldn't find a neat way of getting the true RBS base of a non-main thread, but one idea is simply to call scm_ia64_ar_bsp when guilifying a thread, and use the value returned as an upper bound for that thread's RBS base. (Note that the RBS stack grows upwards.) (Were it not for scm_init_guile, we could be much more definitive about this. We could take the value of scm_ia64_ar_bsp as a definitive base address for the part of the RBS stack that Guile cares about. We could also then discard scm_ia64_register_backing_store_base.)
2008-05-08 00:29:53 +01:00
(char *) thread->register_backing_store_base;
continuation->backing_store = NULL;
continuation->backing_store =
scm_gc_malloc (continuation->backing_store_size,
"continuation backing store");
memcpy (continuation->backing_store,
Fix continuation problems on IA64. * Specific problems in IA64 make check ** test-unwind Representation of the relevant dynamic context: non-rewindable catch frame make cont. o----o-----a----------b-------------c \ \ call cont. o-----o-----------d A continuation is captured at (c), with a non-rewindable frame in the dynamic context at (b). If a rewind through that frame was attempted, Guile would throw to the catch at (a). Then the context unwinds back past (a), then winds forwards again, and the captured continuation is called at (d). We should end up at the catch at (a). On ia64, we get an "illegal instruction". The problem is that Guile does not restore the ia64 register backing store (RBS) stack (which is saved off when the continuation is captured) until all the unwinding and rewinding is done. Therefore, when the rewind code (scm_i_dowinds) hits the non-rewindable frame at (b), the RBS stack hasn't yet been restored. The throw finds the jmp_buf (for the catch at (a)) correctly from the dynamic context, and jumps back to (a), but the RBS stack is invalid, hence the illegal instruction. This could be fixed by restoring the RBS stack earlier, at the same point (copy_stack) where the normal stack is restored. But that causes a problem in the next test... ** continuations.test The dynamic context diagram for this case is similar: non-rewindable catch frame make cont. a----x-----o----------b-------------c \ \ call cont. o-------d The only significant difference is that the catch point (a) is upstream of where the dynamic context forks. This means that the RBS stack at (d) already contains the correct RBS contents for throwing back to (a), so it doesn't matter whether the RBS stack that was saved off with the continuation gets restored. This test passes with the Guile 1.8.4 code, but fails (with an "illegal instruction") when the code is changed to restore the RBS stack earlier as described above. The problem now is that the RBS stack is being restored _too_ early; specifically when there is still stuff to do that relies on the old RBS contents. When a continuation is called, the sequence of relevant events is: (1) Grow the (normal) stack until it is bigger than the (normal) stack saved off in the continuation. (scm_dynthrow, grow_stack) (2) scm_i_dowinds calls itself recursively, such that (2.1) for each rewind (from (x) to (c)) that will be needed, another frame is added to the stack (both normal and RBS), with local variables specifying the required rewind; the rewinds don't actually happen yet, they will happen when the stack unwinds again through these frames (2.2) required unwinds - back from where the continuation was called (d) to the fork point (x) - are done immediately. (3) The normal (i.e. non-RBS) stack that was stored in the continuation is restored (i.e. copied on top of the actual stack). Note that this doesn't overwrite the frames that were added in (2.1), because the growth in (1) ensures that the added frames are beyond the end of the restored stack. (4) ? Restore the RBS stack here too ? (5) Return (from copy_stack) through the (2.1) frames, which means that the rewinds now happen. (6) setcontext (or longjmp) to the context (c) where the continuation was captured. The trouble is that step (1) does not create space in the RBS stack in the same kind of way that it does for the normal stack. Therefore, if the saved (in the continuation) RBS stack is big enough, it can overwrite the RBS of the (2.1) frames that still need to complete. This causes an illegal instruction when we return through those frames and try to perform the rewinds. * Fix The key to the fix is that the saved RBS stack only needs to be restored at some point before the next setcontext call, and that doing it as close to the setcontext call as possible will avoid bad interactions with the pre-setcontext stack. Therefore we do the restoration at the last possible point, immediately before the next setcontext call. The situation is complicated by there being two ways that the next setcontext call can happen. - If the unwinding and rewinding is all successful, the next setcontext will be the one from step (6) above. This is the "normal" continuation invocation case. - If one of the rewinds throws an error, the next setcontext will come from the throw implementation code. (And the one in step (6) will never happen.) This is the rewind error case. In the rewind error case, the code calling setcontext knows nothing about the continuation. So to cover both cases, we: - copy (in step (4) above) the address and length of the continuation's saved RBS stack to the current thread state (SCM_I_CURRENT_THREAD) - modify all setcontext callers so that they check the current thread state for a saved RBS stack, and restore it if so before calling setcontext. * Notes ** I think rewinders cannot rely on using any stack data Unless it can be guaranteed that the data won't go into a register. I'm not 100% sure about this, but I think it follows from the fact that the RBS stack is not restored until after the rewinds have happened. Note that this isn't a regression caused by the current fix. In Guile 1.8.4, the RBS stack was restored _after_ the rewinds, and this is still the case now. ** Most setcontext calls for `throw' don't need to change the RBS stack In the absence of continuation invocation, the setcontext call in the throw implementation code always sets context to a place higher up the same stack (both normal and RBS), hence no stack restoration is needed. * Other changes ** Using setcontext for all non-local jumps (for __ia64__) Along the way, I read a claim somewhere that setcontext was more reliable than longjmp, in cases where the stack has been manipulated. I don't now have any reason to believe this, but it seems reasonable anyway to leave the __ia64__ code using getcontext/setcontext, instead of setjmp/longjmp. (I think the only possible argument against this would be performance - if getcontext was significantly slower than setjmp. It that proves to be the case, we should revisit this.) ** Capping RBS base for non-main threads Somewhere else along the way, I hit a problem in GC, involving the RBS stack of a non-main thread. The problem was, in SCM_MARK_BACKING_STORE, that scm_ia64_register_backing_store_base was returning a value that was massively greater than the value of scm_ia64_ar_bsp, leading to a seg fault. This is because the implementation of scm_ia64_register_backing_store_base is only valid for the main thread. I couldn't find a neat way of getting the true RBS base of a non-main thread, but one idea is simply to call scm_ia64_ar_bsp when guilifying a thread, and use the value returned as an upper bound for that thread's RBS base. (Note that the RBS stack grows upwards.) (Were it not for scm_init_guile, we could be much more definitive about this. We could take the value of scm_ia64_ar_bsp as a definitive base address for the part of the RBS stack that Guile cares about. We could also then discard scm_ia64_register_backing_store_base.)
2008-05-08 00:29:53 +01:00
(void *) thread->register_backing_store_base,
continuation->backing_store_size);
Fix continuation problems on IA64. * Specific problems in IA64 make check ** test-unwind Representation of the relevant dynamic context: non-rewindable catch frame make cont. o----o-----a----------b-------------c \ \ call cont. o-----o-----------d A continuation is captured at (c), with a non-rewindable frame in the dynamic context at (b). If a rewind through that frame was attempted, Guile would throw to the catch at (a). Then the context unwinds back past (a), then winds forwards again, and the captured continuation is called at (d). We should end up at the catch at (a). On ia64, we get an "illegal instruction". The problem is that Guile does not restore the ia64 register backing store (RBS) stack (which is saved off when the continuation is captured) until all the unwinding and rewinding is done. Therefore, when the rewind code (scm_i_dowinds) hits the non-rewindable frame at (b), the RBS stack hasn't yet been restored. The throw finds the jmp_buf (for the catch at (a)) correctly from the dynamic context, and jumps back to (a), but the RBS stack is invalid, hence the illegal instruction. This could be fixed by restoring the RBS stack earlier, at the same point (copy_stack) where the normal stack is restored. But that causes a problem in the next test... ** continuations.test The dynamic context diagram for this case is similar: non-rewindable catch frame make cont. a----x-----o----------b-------------c \ \ call cont. o-------d The only significant difference is that the catch point (a) is upstream of where the dynamic context forks. This means that the RBS stack at (d) already contains the correct RBS contents for throwing back to (a), so it doesn't matter whether the RBS stack that was saved off with the continuation gets restored. This test passes with the Guile 1.8.4 code, but fails (with an "illegal instruction") when the code is changed to restore the RBS stack earlier as described above. The problem now is that the RBS stack is being restored _too_ early; specifically when there is still stuff to do that relies on the old RBS contents. When a continuation is called, the sequence of relevant events is: (1) Grow the (normal) stack until it is bigger than the (normal) stack saved off in the continuation. (scm_dynthrow, grow_stack) (2) scm_i_dowinds calls itself recursively, such that (2.1) for each rewind (from (x) to (c)) that will be needed, another frame is added to the stack (both normal and RBS), with local variables specifying the required rewind; the rewinds don't actually happen yet, they will happen when the stack unwinds again through these frames (2.2) required unwinds - back from where the continuation was called (d) to the fork point (x) - are done immediately. (3) The normal (i.e. non-RBS) stack that was stored in the continuation is restored (i.e. copied on top of the actual stack). Note that this doesn't overwrite the frames that were added in (2.1), because the growth in (1) ensures that the added frames are beyond the end of the restored stack. (4) ? Restore the RBS stack here too ? (5) Return (from copy_stack) through the (2.1) frames, which means that the rewinds now happen. (6) setcontext (or longjmp) to the context (c) where the continuation was captured. The trouble is that step (1) does not create space in the RBS stack in the same kind of way that it does for the normal stack. Therefore, if the saved (in the continuation) RBS stack is big enough, it can overwrite the RBS of the (2.1) frames that still need to complete. This causes an illegal instruction when we return through those frames and try to perform the rewinds. * Fix The key to the fix is that the saved RBS stack only needs to be restored at some point before the next setcontext call, and that doing it as close to the setcontext call as possible will avoid bad interactions with the pre-setcontext stack. Therefore we do the restoration at the last possible point, immediately before the next setcontext call. The situation is complicated by there being two ways that the next setcontext call can happen. - If the unwinding and rewinding is all successful, the next setcontext will be the one from step (6) above. This is the "normal" continuation invocation case. - If one of the rewinds throws an error, the next setcontext will come from the throw implementation code. (And the one in step (6) will never happen.) This is the rewind error case. In the rewind error case, the code calling setcontext knows nothing about the continuation. So to cover both cases, we: - copy (in step (4) above) the address and length of the continuation's saved RBS stack to the current thread state (SCM_I_CURRENT_THREAD) - modify all setcontext callers so that they check the current thread state for a saved RBS stack, and restore it if so before calling setcontext. * Notes ** I think rewinders cannot rely on using any stack data Unless it can be guaranteed that the data won't go into a register. I'm not 100% sure about this, but I think it follows from the fact that the RBS stack is not restored until after the rewinds have happened. Note that this isn't a regression caused by the current fix. In Guile 1.8.4, the RBS stack was restored _after_ the rewinds, and this is still the case now. ** Most setcontext calls for `throw' don't need to change the RBS stack In the absence of continuation invocation, the setcontext call in the throw implementation code always sets context to a place higher up the same stack (both normal and RBS), hence no stack restoration is needed. * Other changes ** Using setcontext for all non-local jumps (for __ia64__) Along the way, I read a claim somewhere that setcontext was more reliable than longjmp, in cases where the stack has been manipulated. I don't now have any reason to believe this, but it seems reasonable anyway to leave the __ia64__ code using getcontext/setcontext, instead of setjmp/longjmp. (I think the only possible argument against this would be performance - if getcontext was significantly slower than setjmp. It that proves to be the case, we should revisit this.) ** Capping RBS base for non-main threads Somewhere else along the way, I hit a problem in GC, involving the RBS stack of a non-main thread. The problem was, in SCM_MARK_BACKING_STORE, that scm_ia64_register_backing_store_base was returning a value that was massively greater than the value of scm_ia64_ar_bsp, leading to a seg fault. This is because the implementation of scm_ia64_register_backing_store_base is only valid for the main thread. I couldn't find a neat way of getting the true RBS base of a non-main thread, but one idea is simply to call scm_ia64_ar_bsp when guilifying a thread, and use the value returned as an upper bound for that thread's RBS base. (Note that the RBS stack grows upwards.) (Were it not for scm_init_guile, we could be much more definitive about this. We could take the value of scm_ia64_ar_bsp as a definitive base address for the part of the RBS stack that Guile cares about. We could also then discard scm_ia64_register_backing_store_base.)
2008-05-08 00:29:53 +01:00
#endif /* __ia64__ */
return make_continuation_trampoline (cont);
}
else
{
scm_gc_after_nonlocal_exit ();
return SCM_UNDEFINED;
}
}
#undef FUNC_NAME
int
scm_i_continuation_to_frame (SCM continuation, struct scm_frame *frame)
{
SCM contregs;
scm_t_contregs *cont;
contregs = SCM_PROGRAM_FREE_VARIABLE_REF (continuation, 0);
cont = SCM_CONTREGS (contregs);
if (scm_is_true (cont->vm_cont))
{
struct scm_vm_cont *data = SCM_VM_CONT_DATA (cont->vm_cont);
frame->stack_holder = data;
frame->fp_offset = (data->fp + data->reloc) - data->stack_base;
frame->sp_offset = (data->sp + data->reloc) - data->stack_base;
frame->ip = data->ra;
return 1;
}
else
return 0;
}
struct scm_vm *
scm_i_contregs_vp (SCM contregs)
continuations return multiple values on the stack * libguile/vm.h (struct scm_vm_cont): Instead of saving the "IP", save "RA" and "MVRA". That is, save singly-valued and multiply-valued return addresses, so that we can return multiple values on the stack. (scm_i_vm_reinstate_continuation): Remove. * libguile/vm.c (vm_capture_continuation): Rename from capture_vm_cont, and change the prototype so we can capture the RA and MVRA, and so that tail calls to call/cc can capture a continuation without the call/cc application frame. (vm_return_to_continuation): Rename from reinstate_vm_cont, and take arguments to return to the continuation. Handles returning to single or multiple-value RA. (scm_i_vm_capture_continuation): Change to invoke vm_capture_continuation. Kept around for the benefit of make-stack. * libguile/vm-i-system.c (continuation-call): Handle reinstatement of the VM stack, with arguments. (call/cc, tail-call/cc): Adapt to new vm_capture_continuation prototype. tail-call/cc captures tail continuations. * libguile/stacks.c (scm_make_stack): Update for scm_vm_cont structure change. * libguile/continuations.h (struct scm_contregs): Remove throw_value member, which was used to return a value to a continuation. (scm_i_check_continuation): New internal function, checks that a continuation may be reinstated. (scm_i_reinstate_continuation): Replaces scm_i_continuation_call; just reinstates the C stack. (scm_i_contregs_vm, scm_i_contregs_vm_cont): New internal accessors. * libguile/continuations.c (scm_i_make_continuation): Return SCM_UNDEFINED if we are returning again. (grow_stack, copy_stack_and_call, scm_dynthrow): Remove extra arg, as vm opcodes handle value returns. (copy_stack): No need to instate VM continuation. (scm_i_reinstate_continuation): Adapt.
2010-02-08 22:59:25 +01:00
{
return SCM_CONTREGS (contregs)->vp;
continuations return multiple values on the stack * libguile/vm.h (struct scm_vm_cont): Instead of saving the "IP", save "RA" and "MVRA". That is, save singly-valued and multiply-valued return addresses, so that we can return multiple values on the stack. (scm_i_vm_reinstate_continuation): Remove. * libguile/vm.c (vm_capture_continuation): Rename from capture_vm_cont, and change the prototype so we can capture the RA and MVRA, and so that tail calls to call/cc can capture a continuation without the call/cc application frame. (vm_return_to_continuation): Rename from reinstate_vm_cont, and take arguments to return to the continuation. Handles returning to single or multiple-value RA. (scm_i_vm_capture_continuation): Change to invoke vm_capture_continuation. Kept around for the benefit of make-stack. * libguile/vm-i-system.c (continuation-call): Handle reinstatement of the VM stack, with arguments. (call/cc, tail-call/cc): Adapt to new vm_capture_continuation prototype. tail-call/cc captures tail continuations. * libguile/stacks.c (scm_make_stack): Update for scm_vm_cont structure change. * libguile/continuations.h (struct scm_contregs): Remove throw_value member, which was used to return a value to a continuation. (scm_i_check_continuation): New internal function, checks that a continuation may be reinstated. (scm_i_reinstate_continuation): Replaces scm_i_continuation_call; just reinstates the C stack. (scm_i_contregs_vm, scm_i_contregs_vm_cont): New internal accessors. * libguile/continuations.c (scm_i_make_continuation): Return SCM_UNDEFINED if we are returning again. (grow_stack, copy_stack_and_call, scm_dynthrow): Remove extra arg, as vm opcodes handle value returns. (copy_stack): No need to instate VM continuation. (scm_i_reinstate_continuation): Adapt.
2010-02-08 22:59:25 +01:00
}
SCM
scm_i_contregs_vm_cont (SCM contregs)
{
return SCM_CONTREGS (contregs)->vm_cont;
}
/* {Apply}
*/
/* Invoking a continuation proceeds as follows:
*
* - the stack is made large enough for the called continuation
* - the old windchain is unwound down to the branching point
* - the continuation stack is copied into place
* - the windchain is rewound up to the continuation's context
* - the continuation is invoked via longjmp (or setcontext)
*
* This order is important so that unwind and rewind handlers are run
* with their correct stack.
*/
continuations return multiple values on the stack * libguile/vm.h (struct scm_vm_cont): Instead of saving the "IP", save "RA" and "MVRA". That is, save singly-valued and multiply-valued return addresses, so that we can return multiple values on the stack. (scm_i_vm_reinstate_continuation): Remove. * libguile/vm.c (vm_capture_continuation): Rename from capture_vm_cont, and change the prototype so we can capture the RA and MVRA, and so that tail calls to call/cc can capture a continuation without the call/cc application frame. (vm_return_to_continuation): Rename from reinstate_vm_cont, and take arguments to return to the continuation. Handles returning to single or multiple-value RA. (scm_i_vm_capture_continuation): Change to invoke vm_capture_continuation. Kept around for the benefit of make-stack. * libguile/vm-i-system.c (continuation-call): Handle reinstatement of the VM stack, with arguments. (call/cc, tail-call/cc): Adapt to new vm_capture_continuation prototype. tail-call/cc captures tail continuations. * libguile/stacks.c (scm_make_stack): Update for scm_vm_cont structure change. * libguile/continuations.h (struct scm_contregs): Remove throw_value member, which was used to return a value to a continuation. (scm_i_check_continuation): New internal function, checks that a continuation may be reinstated. (scm_i_reinstate_continuation): Replaces scm_i_continuation_call; just reinstates the C stack. (scm_i_contregs_vm, scm_i_contregs_vm_cont): New internal accessors. * libguile/continuations.c (scm_i_make_continuation): Return SCM_UNDEFINED if we are returning again. (grow_stack, copy_stack_and_call, scm_dynthrow): Remove extra arg, as vm opcodes handle value returns. (copy_stack): No need to instate VM continuation. (scm_i_reinstate_continuation): Adapt.
2010-02-08 22:59:25 +01:00
static void scm_dynthrow (SCM);
/* Grow the stack by a fixed amount to provide space to copy in the
* continuation. Possibly this function has to be called several times
* recursively before enough space is available. Make sure the compiler does
* not optimize the growth array away by storing it's address into a global
* variable.
*/
static scm_t_bits scm_i_dummy;
static void
continuations return multiple values on the stack * libguile/vm.h (struct scm_vm_cont): Instead of saving the "IP", save "RA" and "MVRA". That is, save singly-valued and multiply-valued return addresses, so that we can return multiple values on the stack. (scm_i_vm_reinstate_continuation): Remove. * libguile/vm.c (vm_capture_continuation): Rename from capture_vm_cont, and change the prototype so we can capture the RA and MVRA, and so that tail calls to call/cc can capture a continuation without the call/cc application frame. (vm_return_to_continuation): Rename from reinstate_vm_cont, and take arguments to return to the continuation. Handles returning to single or multiple-value RA. (scm_i_vm_capture_continuation): Change to invoke vm_capture_continuation. Kept around for the benefit of make-stack. * libguile/vm-i-system.c (continuation-call): Handle reinstatement of the VM stack, with arguments. (call/cc, tail-call/cc): Adapt to new vm_capture_continuation prototype. tail-call/cc captures tail continuations. * libguile/stacks.c (scm_make_stack): Update for scm_vm_cont structure change. * libguile/continuations.h (struct scm_contregs): Remove throw_value member, which was used to return a value to a continuation. (scm_i_check_continuation): New internal function, checks that a continuation may be reinstated. (scm_i_reinstate_continuation): Replaces scm_i_continuation_call; just reinstates the C stack. (scm_i_contregs_vm, scm_i_contregs_vm_cont): New internal accessors. * libguile/continuations.c (scm_i_make_continuation): Return SCM_UNDEFINED if we are returning again. (grow_stack, copy_stack_and_call, scm_dynthrow): Remove extra arg, as vm opcodes handle value returns. (copy_stack): No need to instate VM continuation. (scm_i_reinstate_continuation): Adapt.
2010-02-08 22:59:25 +01:00
grow_stack (SCM cont)
{
2001-06-14 19:50:43 +00:00
scm_t_bits growth[100];
2001-06-14 19:50:43 +00:00
scm_i_dummy = (scm_t_bits) growth;
continuations return multiple values on the stack * libguile/vm.h (struct scm_vm_cont): Instead of saving the "IP", save "RA" and "MVRA". That is, save singly-valued and multiply-valued return addresses, so that we can return multiple values on the stack. (scm_i_vm_reinstate_continuation): Remove. * libguile/vm.c (vm_capture_continuation): Rename from capture_vm_cont, and change the prototype so we can capture the RA and MVRA, and so that tail calls to call/cc can capture a continuation without the call/cc application frame. (vm_return_to_continuation): Rename from reinstate_vm_cont, and take arguments to return to the continuation. Handles returning to single or multiple-value RA. (scm_i_vm_capture_continuation): Change to invoke vm_capture_continuation. Kept around for the benefit of make-stack. * libguile/vm-i-system.c (continuation-call): Handle reinstatement of the VM stack, with arguments. (call/cc, tail-call/cc): Adapt to new vm_capture_continuation prototype. tail-call/cc captures tail continuations. * libguile/stacks.c (scm_make_stack): Update for scm_vm_cont structure change. * libguile/continuations.h (struct scm_contregs): Remove throw_value member, which was used to return a value to a continuation. (scm_i_check_continuation): New internal function, checks that a continuation may be reinstated. (scm_i_reinstate_continuation): Replaces scm_i_continuation_call; just reinstates the C stack. (scm_i_contregs_vm, scm_i_contregs_vm_cont): New internal accessors. * libguile/continuations.c (scm_i_make_continuation): Return SCM_UNDEFINED if we are returning again. (grow_stack, copy_stack_and_call, scm_dynthrow): Remove extra arg, as vm opcodes handle value returns. (copy_stack): No need to instate VM continuation. (scm_i_reinstate_continuation): Adapt.
2010-02-08 22:59:25 +01:00
scm_dynthrow (cont);
}
/* Copy the continuation stack into the current stack. Calling functions from
* within this function is safe, since only stack frames below this function's
* own frame are overwritten. Thus, memcpy can be used for best performance.
*/
static void
continuations return multiple values on the stack * libguile/vm.h (struct scm_vm_cont): Instead of saving the "IP", save "RA" and "MVRA". That is, save singly-valued and multiply-valued return addresses, so that we can return multiple values on the stack. (scm_i_vm_reinstate_continuation): Remove. * libguile/vm.c (vm_capture_continuation): Rename from capture_vm_cont, and change the prototype so we can capture the RA and MVRA, and so that tail calls to call/cc can capture a continuation without the call/cc application frame. (vm_return_to_continuation): Rename from reinstate_vm_cont, and take arguments to return to the continuation. Handles returning to single or multiple-value RA. (scm_i_vm_capture_continuation): Change to invoke vm_capture_continuation. Kept around for the benefit of make-stack. * libguile/vm-i-system.c (continuation-call): Handle reinstatement of the VM stack, with arguments. (call/cc, tail-call/cc): Adapt to new vm_capture_continuation prototype. tail-call/cc captures tail continuations. * libguile/stacks.c (scm_make_stack): Update for scm_vm_cont structure change. * libguile/continuations.h (struct scm_contregs): Remove throw_value member, which was used to return a value to a continuation. (scm_i_check_continuation): New internal function, checks that a continuation may be reinstated. (scm_i_reinstate_continuation): Replaces scm_i_continuation_call; just reinstates the C stack. (scm_i_contregs_vm, scm_i_contregs_vm_cont): New internal accessors. * libguile/continuations.c (scm_i_make_continuation): Return SCM_UNDEFINED if we are returning again. (grow_stack, copy_stack_and_call, scm_dynthrow): Remove extra arg, as vm opcodes handle value returns. (copy_stack): No need to instate VM continuation. (scm_i_reinstate_continuation): Adapt.
2010-02-08 22:59:25 +01:00
copy_stack_and_call (scm_t_contregs *continuation,
SCM_STACKITEM * dst)
{
the dynamic stack is really a stack now, instead of a list * libguile/dynstack.h: * libguile/dynstack.c: New files, implementing the dynamic stack as a true stack instead of a linked list. This lowers the cost of dynwinds: frames, winders, prompts, with-fluids, and dynamic-wind. For the most part, we allocate these items directly on the stack. * libguile/dynwinds.h: * libguile/dynwinds.c: Adapt all manipulators of the wind stack to use interfaces from dynstack.c. Remove heap-allocated winder and frame object types. (scm_dowinds, scm_i_dowinds): Remove these. The first was exported, but it was not a public interface. * libguile/continuations.c: * libguile/continuations.h (scm_t_contregs): Continuation objects reference scm_t_dynstack* values now. Adapt to the new interfaces. * libguile/control.c: * libguile/control.h: There is no longer a scm_tc7_prompt kind of object that can be allocated on the heap. Instead, the prompt flags, key, and registers are pushed on the dynwind stack. (The registers are still on the heap.) Also, since the vm_cont will reference the dynwinds, make the partial continuation stub take just one extra arg, instead of storing the intwinds separately in the object table. * libguile/fluids.c: * libguile/fluids.h: No more with-fluids objects; instead, the fluids go on the dynstack. The values still have to be on the heap, though. (scm_prepare_fluids, scm_swap_fluids): New internal functions, replacing scm_i_make_with_fluids and scm_i_swap_with_fluids. * libguile/print.c: Remove prompt and with-fluids printers. * libguile/tags.h: Revert prompt and with-fluids tc7 values to what they were before they were allocated. * libguile/vm-i-system.c (partial_cont_call): Just pop the vmcont, the intwinds will not be passed as a second arg. Rewind the dynamic stack from within the VM, so that any rewinder sees valid prompt entries. (call_cc, tail_call_cc): Adapt to pass the dynstack to scm_i_vm_capture_stack. (prompt, wind, unwind, wind_fluids, unwind_fluids): Adapt to the new interfaces. * libguile/vm.h (scm_i_capture_current_stack): Rename from scm_i_vm_capture_continuation. (scm_i_vm_capture_stack): Take a dynstack as an argument. * libguile/vm.c (vm_reinstate_partial_continuation): Don't wind here, as that could result in winders seeing invalid prompts. * libguile/eval.c: * libguile/root.c: * libguile/stacks.c: * libguile/threads.c: * libguile/threads.h: * libguile/throw.c: Adapt other users of dynwinds to use the dynstack.
2012-03-03 17:01:16 +01:00
scm_t_dynstack *dynstack;
scm_t_bits *joint;
scm_i_thread *thread = SCM_I_CURRENT_THREAD;
dynstack = SCM_VM_CONT_DATA (continuation->vm_cont)->dynstack;
joint = scm_dynstack_unwind_fork (&thread->dynstack, dynstack);
memcpy (dst, continuation->stack,
sizeof (SCM_STACKITEM) * continuation->num_stack_items);
#ifdef __ia64__
thread->pending_rbs_continuation = continuation;
#endif
the dynamic stack is really a stack now, instead of a list * libguile/dynstack.h: * libguile/dynstack.c: New files, implementing the dynamic stack as a true stack instead of a linked list. This lowers the cost of dynwinds: frames, winders, prompts, with-fluids, and dynamic-wind. For the most part, we allocate these items directly on the stack. * libguile/dynwinds.h: * libguile/dynwinds.c: Adapt all manipulators of the wind stack to use interfaces from dynstack.c. Remove heap-allocated winder and frame object types. (scm_dowinds, scm_i_dowinds): Remove these. The first was exported, but it was not a public interface. * libguile/continuations.c: * libguile/continuations.h (scm_t_contregs): Continuation objects reference scm_t_dynstack* values now. Adapt to the new interfaces. * libguile/control.c: * libguile/control.h: There is no longer a scm_tc7_prompt kind of object that can be allocated on the heap. Instead, the prompt flags, key, and registers are pushed on the dynwind stack. (The registers are still on the heap.) Also, since the vm_cont will reference the dynwinds, make the partial continuation stub take just one extra arg, instead of storing the intwinds separately in the object table. * libguile/fluids.c: * libguile/fluids.h: No more with-fluids objects; instead, the fluids go on the dynstack. The values still have to be on the heap, though. (scm_prepare_fluids, scm_swap_fluids): New internal functions, replacing scm_i_make_with_fluids and scm_i_swap_with_fluids. * libguile/print.c: Remove prompt and with-fluids printers. * libguile/tags.h: Revert prompt and with-fluids tc7 values to what they were before they were allocated. * libguile/vm-i-system.c (partial_cont_call): Just pop the vmcont, the intwinds will not be passed as a second arg. Rewind the dynamic stack from within the VM, so that any rewinder sees valid prompt entries. (call_cc, tail_call_cc): Adapt to pass the dynstack to scm_i_vm_capture_stack. (prompt, wind, unwind, wind_fluids, unwind_fluids): Adapt to the new interfaces. * libguile/vm.h (scm_i_capture_current_stack): Rename from scm_i_vm_capture_continuation. (scm_i_vm_capture_stack): Take a dynstack as an argument. * libguile/vm.c (vm_reinstate_partial_continuation): Don't wind here, as that could result in winders seeing invalid prompts. * libguile/eval.c: * libguile/root.c: * libguile/stacks.c: * libguile/threads.c: * libguile/threads.h: * libguile/throw.c: Adapt other users of dynwinds to use the dynstack.
2012-03-03 17:01:16 +01:00
scm_dynstack_wind (&thread->dynstack, joint);
SCM_I_LONGJMP (continuation->jmpbuf, 1);
}
Fix continuation problems on IA64. * Specific problems in IA64 make check ** test-unwind Representation of the relevant dynamic context: non-rewindable catch frame make cont. o----o-----a----------b-------------c \ \ call cont. o-----o-----------d A continuation is captured at (c), with a non-rewindable frame in the dynamic context at (b). If a rewind through that frame was attempted, Guile would throw to the catch at (a). Then the context unwinds back past (a), then winds forwards again, and the captured continuation is called at (d). We should end up at the catch at (a). On ia64, we get an "illegal instruction". The problem is that Guile does not restore the ia64 register backing store (RBS) stack (which is saved off when the continuation is captured) until all the unwinding and rewinding is done. Therefore, when the rewind code (scm_i_dowinds) hits the non-rewindable frame at (b), the RBS stack hasn't yet been restored. The throw finds the jmp_buf (for the catch at (a)) correctly from the dynamic context, and jumps back to (a), but the RBS stack is invalid, hence the illegal instruction. This could be fixed by restoring the RBS stack earlier, at the same point (copy_stack) where the normal stack is restored. But that causes a problem in the next test... ** continuations.test The dynamic context diagram for this case is similar: non-rewindable catch frame make cont. a----x-----o----------b-------------c \ \ call cont. o-------d The only significant difference is that the catch point (a) is upstream of where the dynamic context forks. This means that the RBS stack at (d) already contains the correct RBS contents for throwing back to (a), so it doesn't matter whether the RBS stack that was saved off with the continuation gets restored. This test passes with the Guile 1.8.4 code, but fails (with an "illegal instruction") when the code is changed to restore the RBS stack earlier as described above. The problem now is that the RBS stack is being restored _too_ early; specifically when there is still stuff to do that relies on the old RBS contents. When a continuation is called, the sequence of relevant events is: (1) Grow the (normal) stack until it is bigger than the (normal) stack saved off in the continuation. (scm_dynthrow, grow_stack) (2) scm_i_dowinds calls itself recursively, such that (2.1) for each rewind (from (x) to (c)) that will be needed, another frame is added to the stack (both normal and RBS), with local variables specifying the required rewind; the rewinds don't actually happen yet, they will happen when the stack unwinds again through these frames (2.2) required unwinds - back from where the continuation was called (d) to the fork point (x) - are done immediately. (3) The normal (i.e. non-RBS) stack that was stored in the continuation is restored (i.e. copied on top of the actual stack). Note that this doesn't overwrite the frames that were added in (2.1), because the growth in (1) ensures that the added frames are beyond the end of the restored stack. (4) ? Restore the RBS stack here too ? (5) Return (from copy_stack) through the (2.1) frames, which means that the rewinds now happen. (6) setcontext (or longjmp) to the context (c) where the continuation was captured. The trouble is that step (1) does not create space in the RBS stack in the same kind of way that it does for the normal stack. Therefore, if the saved (in the continuation) RBS stack is big enough, it can overwrite the RBS of the (2.1) frames that still need to complete. This causes an illegal instruction when we return through those frames and try to perform the rewinds. * Fix The key to the fix is that the saved RBS stack only needs to be restored at some point before the next setcontext call, and that doing it as close to the setcontext call as possible will avoid bad interactions with the pre-setcontext stack. Therefore we do the restoration at the last possible point, immediately before the next setcontext call. The situation is complicated by there being two ways that the next setcontext call can happen. - If the unwinding and rewinding is all successful, the next setcontext will be the one from step (6) above. This is the "normal" continuation invocation case. - If one of the rewinds throws an error, the next setcontext will come from the throw implementation code. (And the one in step (6) will never happen.) This is the rewind error case. In the rewind error case, the code calling setcontext knows nothing about the continuation. So to cover both cases, we: - copy (in step (4) above) the address and length of the continuation's saved RBS stack to the current thread state (SCM_I_CURRENT_THREAD) - modify all setcontext callers so that they check the current thread state for a saved RBS stack, and restore it if so before calling setcontext. * Notes ** I think rewinders cannot rely on using any stack data Unless it can be guaranteed that the data won't go into a register. I'm not 100% sure about this, but I think it follows from the fact that the RBS stack is not restored until after the rewinds have happened. Note that this isn't a regression caused by the current fix. In Guile 1.8.4, the RBS stack was restored _after_ the rewinds, and this is still the case now. ** Most setcontext calls for `throw' don't need to change the RBS stack In the absence of continuation invocation, the setcontext call in the throw implementation code always sets context to a place higher up the same stack (both normal and RBS), hence no stack restoration is needed. * Other changes ** Using setcontext for all non-local jumps (for __ia64__) Along the way, I read a claim somewhere that setcontext was more reliable than longjmp, in cases where the stack has been manipulated. I don't now have any reason to believe this, but it seems reasonable anyway to leave the __ia64__ code using getcontext/setcontext, instead of setjmp/longjmp. (I think the only possible argument against this would be performance - if getcontext was significantly slower than setjmp. It that proves to be the case, we should revisit this.) ** Capping RBS base for non-main threads Somewhere else along the way, I hit a problem in GC, involving the RBS stack of a non-main thread. The problem was, in SCM_MARK_BACKING_STORE, that scm_ia64_register_backing_store_base was returning a value that was massively greater than the value of scm_ia64_ar_bsp, leading to a seg fault. This is because the implementation of scm_ia64_register_backing_store_base is only valid for the main thread. I couldn't find a neat way of getting the true RBS base of a non-main thread, but one idea is simply to call scm_ia64_ar_bsp when guilifying a thread, and use the value returned as an upper bound for that thread's RBS base. (Note that the RBS stack grows upwards.) (Were it not for scm_init_guile, we could be much more definitive about this. We could take the value of scm_ia64_ar_bsp as a definitive base address for the part of the RBS stack that Guile cares about. We could also then discard scm_ia64_register_backing_store_base.)
2008-05-08 00:29:53 +01:00
#ifdef __ia64__
void
scm_ia64_longjmp (scm_i_jmp_buf *JB, int VAL)
Fix continuation problems on IA64. * Specific problems in IA64 make check ** test-unwind Representation of the relevant dynamic context: non-rewindable catch frame make cont. o----o-----a----------b-------------c \ \ call cont. o-----o-----------d A continuation is captured at (c), with a non-rewindable frame in the dynamic context at (b). If a rewind through that frame was attempted, Guile would throw to the catch at (a). Then the context unwinds back past (a), then winds forwards again, and the captured continuation is called at (d). We should end up at the catch at (a). On ia64, we get an "illegal instruction". The problem is that Guile does not restore the ia64 register backing store (RBS) stack (which is saved off when the continuation is captured) until all the unwinding and rewinding is done. Therefore, when the rewind code (scm_i_dowinds) hits the non-rewindable frame at (b), the RBS stack hasn't yet been restored. The throw finds the jmp_buf (for the catch at (a)) correctly from the dynamic context, and jumps back to (a), but the RBS stack is invalid, hence the illegal instruction. This could be fixed by restoring the RBS stack earlier, at the same point (copy_stack) where the normal stack is restored. But that causes a problem in the next test... ** continuations.test The dynamic context diagram for this case is similar: non-rewindable catch frame make cont. a----x-----o----------b-------------c \ \ call cont. o-------d The only significant difference is that the catch point (a) is upstream of where the dynamic context forks. This means that the RBS stack at (d) already contains the correct RBS contents for throwing back to (a), so it doesn't matter whether the RBS stack that was saved off with the continuation gets restored. This test passes with the Guile 1.8.4 code, but fails (with an "illegal instruction") when the code is changed to restore the RBS stack earlier as described above. The problem now is that the RBS stack is being restored _too_ early; specifically when there is still stuff to do that relies on the old RBS contents. When a continuation is called, the sequence of relevant events is: (1) Grow the (normal) stack until it is bigger than the (normal) stack saved off in the continuation. (scm_dynthrow, grow_stack) (2) scm_i_dowinds calls itself recursively, such that (2.1) for each rewind (from (x) to (c)) that will be needed, another frame is added to the stack (both normal and RBS), with local variables specifying the required rewind; the rewinds don't actually happen yet, they will happen when the stack unwinds again through these frames (2.2) required unwinds - back from where the continuation was called (d) to the fork point (x) - are done immediately. (3) The normal (i.e. non-RBS) stack that was stored in the continuation is restored (i.e. copied on top of the actual stack). Note that this doesn't overwrite the frames that were added in (2.1), because the growth in (1) ensures that the added frames are beyond the end of the restored stack. (4) ? Restore the RBS stack here too ? (5) Return (from copy_stack) through the (2.1) frames, which means that the rewinds now happen. (6) setcontext (or longjmp) to the context (c) where the continuation was captured. The trouble is that step (1) does not create space in the RBS stack in the same kind of way that it does for the normal stack. Therefore, if the saved (in the continuation) RBS stack is big enough, it can overwrite the RBS of the (2.1) frames that still need to complete. This causes an illegal instruction when we return through those frames and try to perform the rewinds. * Fix The key to the fix is that the saved RBS stack only needs to be restored at some point before the next setcontext call, and that doing it as close to the setcontext call as possible will avoid bad interactions with the pre-setcontext stack. Therefore we do the restoration at the last possible point, immediately before the next setcontext call. The situation is complicated by there being two ways that the next setcontext call can happen. - If the unwinding and rewinding is all successful, the next setcontext will be the one from step (6) above. This is the "normal" continuation invocation case. - If one of the rewinds throws an error, the next setcontext will come from the throw implementation code. (And the one in step (6) will never happen.) This is the rewind error case. In the rewind error case, the code calling setcontext knows nothing about the continuation. So to cover both cases, we: - copy (in step (4) above) the address and length of the continuation's saved RBS stack to the current thread state (SCM_I_CURRENT_THREAD) - modify all setcontext callers so that they check the current thread state for a saved RBS stack, and restore it if so before calling setcontext. * Notes ** I think rewinders cannot rely on using any stack data Unless it can be guaranteed that the data won't go into a register. I'm not 100% sure about this, but I think it follows from the fact that the RBS stack is not restored until after the rewinds have happened. Note that this isn't a regression caused by the current fix. In Guile 1.8.4, the RBS stack was restored _after_ the rewinds, and this is still the case now. ** Most setcontext calls for `throw' don't need to change the RBS stack In the absence of continuation invocation, the setcontext call in the throw implementation code always sets context to a place higher up the same stack (both normal and RBS), hence no stack restoration is needed. * Other changes ** Using setcontext for all non-local jumps (for __ia64__) Along the way, I read a claim somewhere that setcontext was more reliable than longjmp, in cases where the stack has been manipulated. I don't now have any reason to believe this, but it seems reasonable anyway to leave the __ia64__ code using getcontext/setcontext, instead of setjmp/longjmp. (I think the only possible argument against this would be performance - if getcontext was significantly slower than setjmp. It that proves to be the case, we should revisit this.) ** Capping RBS base for non-main threads Somewhere else along the way, I hit a problem in GC, involving the RBS stack of a non-main thread. The problem was, in SCM_MARK_BACKING_STORE, that scm_ia64_register_backing_store_base was returning a value that was massively greater than the value of scm_ia64_ar_bsp, leading to a seg fault. This is because the implementation of scm_ia64_register_backing_store_base is only valid for the main thread. I couldn't find a neat way of getting the true RBS base of a non-main thread, but one idea is simply to call scm_ia64_ar_bsp when guilifying a thread, and use the value returned as an upper bound for that thread's RBS base. (Note that the RBS stack grows upwards.) (Were it not for scm_init_guile, we could be much more definitive about this. We could take the value of scm_ia64_ar_bsp as a definitive base address for the part of the RBS stack that Guile cares about. We could also then discard scm_ia64_register_backing_store_base.)
2008-05-08 00:29:53 +01:00
{
scm_i_thread *t = SCM_I_CURRENT_THREAD;
if (t->pending_rbs_continuation)
{
memcpy (t->register_backing_store_base,
t->pending_rbs_continuation->backing_store,
t->pending_rbs_continuation->backing_store_size);
t->pending_rbs_continuation = NULL;
}
setcontext (&JB->ctx);
}
#endif
/* Call grow_stack until the stack space is large enough, then, as the current
* stack frame might get overwritten, let copy_stack_and_call perform the
* actual copying and continuation calling.
*/
static void
continuations return multiple values on the stack * libguile/vm.h (struct scm_vm_cont): Instead of saving the "IP", save "RA" and "MVRA". That is, save singly-valued and multiply-valued return addresses, so that we can return multiple values on the stack. (scm_i_vm_reinstate_continuation): Remove. * libguile/vm.c (vm_capture_continuation): Rename from capture_vm_cont, and change the prototype so we can capture the RA and MVRA, and so that tail calls to call/cc can capture a continuation without the call/cc application frame. (vm_return_to_continuation): Rename from reinstate_vm_cont, and take arguments to return to the continuation. Handles returning to single or multiple-value RA. (scm_i_vm_capture_continuation): Change to invoke vm_capture_continuation. Kept around for the benefit of make-stack. * libguile/vm-i-system.c (continuation-call): Handle reinstatement of the VM stack, with arguments. (call/cc, tail-call/cc): Adapt to new vm_capture_continuation prototype. tail-call/cc captures tail continuations. * libguile/stacks.c (scm_make_stack): Update for scm_vm_cont structure change. * libguile/continuations.h (struct scm_contregs): Remove throw_value member, which was used to return a value to a continuation. (scm_i_check_continuation): New internal function, checks that a continuation may be reinstated. (scm_i_reinstate_continuation): Replaces scm_i_continuation_call; just reinstates the C stack. (scm_i_contregs_vm, scm_i_contregs_vm_cont): New internal accessors. * libguile/continuations.c (scm_i_make_continuation): Return SCM_UNDEFINED if we are returning again. (grow_stack, copy_stack_and_call, scm_dynthrow): Remove extra arg, as vm opcodes handle value returns. (copy_stack): No need to instate VM continuation. (scm_i_reinstate_continuation): Adapt.
2010-02-08 22:59:25 +01:00
scm_dynthrow (SCM cont)
{
2005-03-02 20:42:01 +00:00
scm_i_thread *thread = SCM_I_CURRENT_THREAD;
2001-06-14 19:50:43 +00:00
scm_t_contregs *continuation = SCM_CONTREGS (cont);
2005-03-02 20:42:01 +00:00
SCM_STACKITEM *dst = thread->continuation_base;
SCM_STACKITEM stack_top_element;
if (thread->critical_section_level)
{
fprintf (stderr, "continuation invoked from within critical section.\n");
abort ();
}
#if SCM_STACK_GROWS_UP
if (dst + continuation->num_stack_items >= &stack_top_element)
continuations return multiple values on the stack * libguile/vm.h (struct scm_vm_cont): Instead of saving the "IP", save "RA" and "MVRA". That is, save singly-valued and multiply-valued return addresses, so that we can return multiple values on the stack. (scm_i_vm_reinstate_continuation): Remove. * libguile/vm.c (vm_capture_continuation): Rename from capture_vm_cont, and change the prototype so we can capture the RA and MVRA, and so that tail calls to call/cc can capture a continuation without the call/cc application frame. (vm_return_to_continuation): Rename from reinstate_vm_cont, and take arguments to return to the continuation. Handles returning to single or multiple-value RA. (scm_i_vm_capture_continuation): Change to invoke vm_capture_continuation. Kept around for the benefit of make-stack. * libguile/vm-i-system.c (continuation-call): Handle reinstatement of the VM stack, with arguments. (call/cc, tail-call/cc): Adapt to new vm_capture_continuation prototype. tail-call/cc captures tail continuations. * libguile/stacks.c (scm_make_stack): Update for scm_vm_cont structure change. * libguile/continuations.h (struct scm_contregs): Remove throw_value member, which was used to return a value to a continuation. (scm_i_check_continuation): New internal function, checks that a continuation may be reinstated. (scm_i_reinstate_continuation): Replaces scm_i_continuation_call; just reinstates the C stack. (scm_i_contregs_vm, scm_i_contregs_vm_cont): New internal accessors. * libguile/continuations.c (scm_i_make_continuation): Return SCM_UNDEFINED if we are returning again. (grow_stack, copy_stack_and_call, scm_dynthrow): Remove extra arg, as vm opcodes handle value returns. (copy_stack): No need to instate VM continuation. (scm_i_reinstate_continuation): Adapt.
2010-02-08 22:59:25 +01:00
grow_stack (cont);
#else
dst -= continuation->num_stack_items;
2002-08-04 00:17:18 +00:00
if (dst <= &stack_top_element)
continuations return multiple values on the stack * libguile/vm.h (struct scm_vm_cont): Instead of saving the "IP", save "RA" and "MVRA". That is, save singly-valued and multiply-valued return addresses, so that we can return multiple values on the stack. (scm_i_vm_reinstate_continuation): Remove. * libguile/vm.c (vm_capture_continuation): Rename from capture_vm_cont, and change the prototype so we can capture the RA and MVRA, and so that tail calls to call/cc can capture a continuation without the call/cc application frame. (vm_return_to_continuation): Rename from reinstate_vm_cont, and take arguments to return to the continuation. Handles returning to single or multiple-value RA. (scm_i_vm_capture_continuation): Change to invoke vm_capture_continuation. Kept around for the benefit of make-stack. * libguile/vm-i-system.c (continuation-call): Handle reinstatement of the VM stack, with arguments. (call/cc, tail-call/cc): Adapt to new vm_capture_continuation prototype. tail-call/cc captures tail continuations. * libguile/stacks.c (scm_make_stack): Update for scm_vm_cont structure change. * libguile/continuations.h (struct scm_contregs): Remove throw_value member, which was used to return a value to a continuation. (scm_i_check_continuation): New internal function, checks that a continuation may be reinstated. (scm_i_reinstate_continuation): Replaces scm_i_continuation_call; just reinstates the C stack. (scm_i_contregs_vm, scm_i_contregs_vm_cont): New internal accessors. * libguile/continuations.c (scm_i_make_continuation): Return SCM_UNDEFINED if we are returning again. (grow_stack, copy_stack_and_call, scm_dynthrow): Remove extra arg, as vm opcodes handle value returns. (copy_stack): No need to instate VM continuation. (scm_i_reinstate_continuation): Adapt.
2010-02-08 22:59:25 +01:00
grow_stack (cont);
#endif /* def SCM_STACK_GROWS_UP */
SCM_FLUSH_REGISTER_WINDOWS;
continuations return multiple values on the stack * libguile/vm.h (struct scm_vm_cont): Instead of saving the "IP", save "RA" and "MVRA". That is, save singly-valued and multiply-valued return addresses, so that we can return multiple values on the stack. (scm_i_vm_reinstate_continuation): Remove. * libguile/vm.c (vm_capture_continuation): Rename from capture_vm_cont, and change the prototype so we can capture the RA and MVRA, and so that tail calls to call/cc can capture a continuation without the call/cc application frame. (vm_return_to_continuation): Rename from reinstate_vm_cont, and take arguments to return to the continuation. Handles returning to single or multiple-value RA. (scm_i_vm_capture_continuation): Change to invoke vm_capture_continuation. Kept around for the benefit of make-stack. * libguile/vm-i-system.c (continuation-call): Handle reinstatement of the VM stack, with arguments. (call/cc, tail-call/cc): Adapt to new vm_capture_continuation prototype. tail-call/cc captures tail continuations. * libguile/stacks.c (scm_make_stack): Update for scm_vm_cont structure change. * libguile/continuations.h (struct scm_contregs): Remove throw_value member, which was used to return a value to a continuation. (scm_i_check_continuation): New internal function, checks that a continuation may be reinstated. (scm_i_reinstate_continuation): Replaces scm_i_continuation_call; just reinstates the C stack. (scm_i_contregs_vm, scm_i_contregs_vm_cont): New internal accessors. * libguile/continuations.c (scm_i_make_continuation): Return SCM_UNDEFINED if we are returning again. (grow_stack, copy_stack_and_call, scm_dynthrow): Remove extra arg, as vm opcodes handle value returns. (copy_stack): No need to instate VM continuation. (scm_i_reinstate_continuation): Adapt.
2010-02-08 22:59:25 +01:00
copy_stack_and_call (continuation, dst);
}
2001-03-04 17:09:34 +00:00
void
continuations return multiple values on the stack * libguile/vm.h (struct scm_vm_cont): Instead of saving the "IP", save "RA" and "MVRA". That is, save singly-valued and multiply-valued return addresses, so that we can return multiple values on the stack. (scm_i_vm_reinstate_continuation): Remove. * libguile/vm.c (vm_capture_continuation): Rename from capture_vm_cont, and change the prototype so we can capture the RA and MVRA, and so that tail calls to call/cc can capture a continuation without the call/cc application frame. (vm_return_to_continuation): Rename from reinstate_vm_cont, and take arguments to return to the continuation. Handles returning to single or multiple-value RA. (scm_i_vm_capture_continuation): Change to invoke vm_capture_continuation. Kept around for the benefit of make-stack. * libguile/vm-i-system.c (continuation-call): Handle reinstatement of the VM stack, with arguments. (call/cc, tail-call/cc): Adapt to new vm_capture_continuation prototype. tail-call/cc captures tail continuations. * libguile/stacks.c (scm_make_stack): Update for scm_vm_cont structure change. * libguile/continuations.h (struct scm_contregs): Remove throw_value member, which was used to return a value to a continuation. (scm_i_check_continuation): New internal function, checks that a continuation may be reinstated. (scm_i_reinstate_continuation): Replaces scm_i_continuation_call; just reinstates the C stack. (scm_i_contregs_vm, scm_i_contregs_vm_cont): New internal accessors. * libguile/continuations.c (scm_i_make_continuation): Return SCM_UNDEFINED if we are returning again. (grow_stack, copy_stack_and_call, scm_dynthrow): Remove extra arg, as vm opcodes handle value returns. (copy_stack): No need to instate VM continuation. (scm_i_reinstate_continuation): Adapt.
2010-02-08 22:59:25 +01:00
scm_i_check_continuation (SCM cont)
{
2005-03-02 20:42:01 +00:00
scm_i_thread *thread = SCM_I_CURRENT_THREAD;
2001-06-14 19:50:43 +00:00
scm_t_contregs *continuation = SCM_CONTREGS (cont);
if (!scm_is_eq (continuation->root, thread->continuation_root))
scm_misc_error
("%continuation-call",
"invoking continuation would cross continuation barrier: ~A",
scm_list_1 (cont));
continuations return multiple values on the stack * libguile/vm.h (struct scm_vm_cont): Instead of saving the "IP", save "RA" and "MVRA". That is, save singly-valued and multiply-valued return addresses, so that we can return multiple values on the stack. (scm_i_vm_reinstate_continuation): Remove. * libguile/vm.c (vm_capture_continuation): Rename from capture_vm_cont, and change the prototype so we can capture the RA and MVRA, and so that tail calls to call/cc can capture a continuation without the call/cc application frame. (vm_return_to_continuation): Rename from reinstate_vm_cont, and take arguments to return to the continuation. Handles returning to single or multiple-value RA. (scm_i_vm_capture_continuation): Change to invoke vm_capture_continuation. Kept around for the benefit of make-stack. * libguile/vm-i-system.c (continuation-call): Handle reinstatement of the VM stack, with arguments. (call/cc, tail-call/cc): Adapt to new vm_capture_continuation prototype. tail-call/cc captures tail continuations. * libguile/stacks.c (scm_make_stack): Update for scm_vm_cont structure change. * libguile/continuations.h (struct scm_contregs): Remove throw_value member, which was used to return a value to a continuation. (scm_i_check_continuation): New internal function, checks that a continuation may be reinstated. (scm_i_reinstate_continuation): Replaces scm_i_continuation_call; just reinstates the C stack. (scm_i_contregs_vm, scm_i_contregs_vm_cont): New internal accessors. * libguile/continuations.c (scm_i_make_continuation): Return SCM_UNDEFINED if we are returning again. (grow_stack, copy_stack_and_call, scm_dynthrow): Remove extra arg, as vm opcodes handle value returns. (copy_stack): No need to instate VM continuation. (scm_i_reinstate_continuation): Adapt.
2010-02-08 22:59:25 +01:00
}
void
scm_i_reinstate_continuation (SCM cont)
{
scm_dynthrow (cont);
}
2005-03-02 20:42:01 +00:00
SCM
scm_i_with_continuation_barrier (scm_t_catch_body body,
void *body_data,
scm_t_catch_handler handler,
* throw.h (scm_c_catch, scm_c_with_throw_handler, scm_catch_with_pre_unwind_handler, scm_with_throw_handler): New. * throw.c (SCM_JBPREUNWIND, SCM_SETJBPREUNWIND): New. (struct pre_unwind_data): New, replaces struct lazy_catch. (scm_c_catch): New, replaces scm_internal_catch as the primary catch API for C code; adds pre-unwind handler support. (scm_internal_catch): Now just a wrapper for scm_c_catch, for back compatibility. (tc16_pre_unwind_data, pre_unwind_data_print, make_pre_unwind_data, SCM_PRE_UNWIND_DATA_P): Renamed from "lazy_catch" equivalents. (scm_c_with_throw_handler): New, replaces scm_internal_lazy_catch as the primary C API for a "lazy" catch. (scm_internal_lazy_catch): Now just a wrapper for scm_c_with_throw_handler, for back compatibility. (scm_catch_with_pre_unwind_handler): Renamed from scm_catch; adds pre-unwind handler support. (scm_catch): Now just a wrapper for scm_catch_with_pre_unwind_handler, for back compatibility. (scm_with_throw_handler): New. (scm_lazy_catch): Update comment to say that the handler can return, and what happens if it does. (toggle_pre_unwind_running): New. (scm_ithrow): When identifying the throw target, take running flags into account. In general, change naming of things from "lazy_catch" to "pre_unwind". When throwing to a throw handler, don't unwind the dynamic context first. Add dynwind framing to manage the running flag of a throw handler. If a lazy catch or throw handler returns, rethrow the same exception again. Add pre-unwind support to the normal catch case (SCM_JMPBUFP). * root.c (scm_internal_cwdr): Add NULL args to scm_i_with_continuation_barrier call. * dynwind.c: Change comment mentioning lazy-catch to mention pre-unwind data and throw handler also. * continuations.h (scm_i_with_continuation_barrier): Add pre-unwind handler args. * continuations.c (scm_i_with_continuation_barrier): Add pre-unwind handler args, and pass on to scm_c_catch (changed from scm_internal_catch). (c_handler): Remove scm_handle_by_message_noexit call. (scm_c_with_continuation_barrier): Call scm_i_with_continuation_barrier with scm_handle_by_message_noexit as the pre-unwind handler. (scm_handler): Remove scm_handle_by_message_noexit call. (s_scm_with_continuation_barrier): Call scm_i_with_continuation_barrier with scm_handle_by_message_noexit as the pre-unwind handler.
2006-02-04 14:36:24 +00:00
void *handler_data,
scm_t_catch_handler pre_unwind_handler,
void *pre_unwind_handler_data)
2005-03-02 20:42:01 +00:00
{
SCM_STACKITEM stack_item;
scm_i_thread *thread = SCM_I_CURRENT_THREAD;
SCM old_controot;
SCM_STACKITEM *old_contbase;
SCM result;
/* Establish a fresh continuation root.
*/
old_controot = thread->continuation_root;
old_contbase = thread->continuation_base;
thread->continuation_root = scm_cons (thread->handle, old_controot);
thread->continuation_base = &stack_item;
/* Call FUNC inside a catch all. This is now guaranteed to return
directly and exactly once.
*/
* throw.h (scm_c_catch, scm_c_with_throw_handler, scm_catch_with_pre_unwind_handler, scm_with_throw_handler): New. * throw.c (SCM_JBPREUNWIND, SCM_SETJBPREUNWIND): New. (struct pre_unwind_data): New, replaces struct lazy_catch. (scm_c_catch): New, replaces scm_internal_catch as the primary catch API for C code; adds pre-unwind handler support. (scm_internal_catch): Now just a wrapper for scm_c_catch, for back compatibility. (tc16_pre_unwind_data, pre_unwind_data_print, make_pre_unwind_data, SCM_PRE_UNWIND_DATA_P): Renamed from "lazy_catch" equivalents. (scm_c_with_throw_handler): New, replaces scm_internal_lazy_catch as the primary C API for a "lazy" catch. (scm_internal_lazy_catch): Now just a wrapper for scm_c_with_throw_handler, for back compatibility. (scm_catch_with_pre_unwind_handler): Renamed from scm_catch; adds pre-unwind handler support. (scm_catch): Now just a wrapper for scm_catch_with_pre_unwind_handler, for back compatibility. (scm_with_throw_handler): New. (scm_lazy_catch): Update comment to say that the handler can return, and what happens if it does. (toggle_pre_unwind_running): New. (scm_ithrow): When identifying the throw target, take running flags into account. In general, change naming of things from "lazy_catch" to "pre_unwind". When throwing to a throw handler, don't unwind the dynamic context first. Add dynwind framing to manage the running flag of a throw handler. If a lazy catch or throw handler returns, rethrow the same exception again. Add pre-unwind support to the normal catch case (SCM_JMPBUFP). * root.c (scm_internal_cwdr): Add NULL args to scm_i_with_continuation_barrier call. * dynwind.c: Change comment mentioning lazy-catch to mention pre-unwind data and throw handler also. * continuations.h (scm_i_with_continuation_barrier): Add pre-unwind handler args. * continuations.c (scm_i_with_continuation_barrier): Add pre-unwind handler args, and pass on to scm_c_catch (changed from scm_internal_catch). (c_handler): Remove scm_handle_by_message_noexit call. (scm_c_with_continuation_barrier): Call scm_i_with_continuation_barrier with scm_handle_by_message_noexit as the pre-unwind handler. (scm_handler): Remove scm_handle_by_message_noexit call. (s_scm_with_continuation_barrier): Call scm_i_with_continuation_barrier with scm_handle_by_message_noexit as the pre-unwind handler.
2006-02-04 14:36:24 +00:00
result = scm_c_catch (SCM_BOOL_T,
body, body_data,
handler, handler_data,
pre_unwind_handler, pre_unwind_handler_data);
2005-03-02 20:42:01 +00:00
/* Return to old continuation root.
*/
thread->continuation_base = old_contbase;
thread->continuation_root = old_controot;
return result;
}
static int
should_print_backtrace (SCM tag, SCM stack)
{
return SCM_BACKTRACE_P
&& scm_is_true (stack)
&& scm_initialized_p
/* It's generally not useful to print backtraces for errors reading
or expanding code in these fallback catch statements. */
&& !scm_is_eq (tag, scm_from_latin1_symbol ("read-error"))
&& !scm_is_eq (tag, scm_from_latin1_symbol ("syntax-error"));
}
static void
print_exception_and_backtrace (SCM port, SCM tag, SCM args)
{
SCM stack, frame;
/* We get here via a throw to a catch-all. In that case there is the
throw frame active, and this catch closure, so narrow by two
frames. */
stack = scm_make_stack (SCM_BOOL_T, scm_list_1 (scm_from_int (2)));
frame = scm_is_true (stack) ? scm_stack_ref (stack, SCM_INUM0) : SCM_BOOL_F;
if (should_print_backtrace (tag, stack))
{
scm_puts_unlocked ("Backtrace:\n", port);
scm_display_backtrace_with_highlights (stack, port,
SCM_BOOL_F, SCM_BOOL_F,
SCM_EOL);
scm_newline (port);
}
scm_print_exception (port, frame, tag, args);
}
2005-03-02 20:42:01 +00:00
struct c_data {
void *(*func) (void *);
void *data;
void *result;
};
static SCM
c_body (void *d)
{
struct c_data *data = (struct c_data *)d;
data->result = data->func (data->data);
return SCM_UNSPECIFIED;
}
static SCM
c_handler (void *d, SCM tag, SCM args)
{
struct c_data *data;
/* If TAG is `quit', exit() the process. */
if (scm_is_eq (tag, scm_from_latin1_symbol ("quit")))
exit (scm_exit_status (args));
data = (struct c_data *)d;
2005-03-02 20:42:01 +00:00
data->result = NULL;
return SCM_UNSPECIFIED;
}
static SCM
pre_unwind_handler (void *error_port, SCM tag, SCM args)
{
/* Print the exception unless TAG is `quit'. */
if (!scm_is_eq (tag, scm_from_latin1_symbol ("quit")))
print_exception_and_backtrace (SCM_PACK_POINTER (error_port), tag, args);
return SCM_UNSPECIFIED;
}
2005-03-02 20:42:01 +00:00
void *
scm_c_with_continuation_barrier (void *(*func) (void *), void *data)
{
struct c_data c_data;
c_data.func = func;
c_data.data = data;
scm_i_with_continuation_barrier (c_body, &c_data,
* throw.h (scm_c_catch, scm_c_with_throw_handler, scm_catch_with_pre_unwind_handler, scm_with_throw_handler): New. * throw.c (SCM_JBPREUNWIND, SCM_SETJBPREUNWIND): New. (struct pre_unwind_data): New, replaces struct lazy_catch. (scm_c_catch): New, replaces scm_internal_catch as the primary catch API for C code; adds pre-unwind handler support. (scm_internal_catch): Now just a wrapper for scm_c_catch, for back compatibility. (tc16_pre_unwind_data, pre_unwind_data_print, make_pre_unwind_data, SCM_PRE_UNWIND_DATA_P): Renamed from "lazy_catch" equivalents. (scm_c_with_throw_handler): New, replaces scm_internal_lazy_catch as the primary C API for a "lazy" catch. (scm_internal_lazy_catch): Now just a wrapper for scm_c_with_throw_handler, for back compatibility. (scm_catch_with_pre_unwind_handler): Renamed from scm_catch; adds pre-unwind handler support. (scm_catch): Now just a wrapper for scm_catch_with_pre_unwind_handler, for back compatibility. (scm_with_throw_handler): New. (scm_lazy_catch): Update comment to say that the handler can return, and what happens if it does. (toggle_pre_unwind_running): New. (scm_ithrow): When identifying the throw target, take running flags into account. In general, change naming of things from "lazy_catch" to "pre_unwind". When throwing to a throw handler, don't unwind the dynamic context first. Add dynwind framing to manage the running flag of a throw handler. If a lazy catch or throw handler returns, rethrow the same exception again. Add pre-unwind support to the normal catch case (SCM_JMPBUFP). * root.c (scm_internal_cwdr): Add NULL args to scm_i_with_continuation_barrier call. * dynwind.c: Change comment mentioning lazy-catch to mention pre-unwind data and throw handler also. * continuations.h (scm_i_with_continuation_barrier): Add pre-unwind handler args. * continuations.c (scm_i_with_continuation_barrier): Add pre-unwind handler args, and pass on to scm_c_catch (changed from scm_internal_catch). (c_handler): Remove scm_handle_by_message_noexit call. (scm_c_with_continuation_barrier): Call scm_i_with_continuation_barrier with scm_handle_by_message_noexit as the pre-unwind handler. (scm_handler): Remove scm_handle_by_message_noexit call. (s_scm_with_continuation_barrier): Call scm_i_with_continuation_barrier with scm_handle_by_message_noexit as the pre-unwind handler.
2006-02-04 14:36:24 +00:00
c_handler, &c_data,
pre_unwind_handler,
SCM_UNPACK_POINTER (scm_current_error_port ()));
2005-03-02 20:42:01 +00:00
return c_data.result;
}
struct scm_data {
SCM proc;
};
static SCM
scm_body (void *d)
{
struct scm_data *data = (struct scm_data *)d;
return scm_call_0 (data->proc);
}
static SCM
scm_handler (void *d, SCM tag, SCM args)
{
/* Print a message. Note that if TAG is `quit', this will exit() the
process. */
scm_handle_by_message_noexit (NULL, tag, args);
2005-03-02 20:42:01 +00:00
return SCM_BOOL_F;
}
SCM_DEFINE (scm_with_continuation_barrier, "with-continuation-barrier", 1,0,0,
(SCM proc),
2005-03-04 17:55:49 +00:00
"Call @var{proc} and return its result. Do not allow the invocation of\n"
"continuations that would leave or enter the dynamic extent of the call\n"
"to @code{with-continuation-barrier}. Such an attempt causes an error\n"
"to be signaled.\n"
"\n"
"Throws (such as errors) that are not caught from within @var{proc} are\n"
"caught by @code{with-continuation-barrier}. In that case, a short\n"
"message is printed to the current error port and @code{#f} is returned.\n"
"\n"
"Thus, @code{with-continuation-barrier} returns exactly once.\n")
2005-03-02 20:42:01 +00:00
#define FUNC_NAME s_scm_with_continuation_barrier
{
struct scm_data scm_data;
scm_data.proc = proc;
return scm_i_with_continuation_barrier (scm_body, &scm_data,
* throw.h (scm_c_catch, scm_c_with_throw_handler, scm_catch_with_pre_unwind_handler, scm_with_throw_handler): New. * throw.c (SCM_JBPREUNWIND, SCM_SETJBPREUNWIND): New. (struct pre_unwind_data): New, replaces struct lazy_catch. (scm_c_catch): New, replaces scm_internal_catch as the primary catch API for C code; adds pre-unwind handler support. (scm_internal_catch): Now just a wrapper for scm_c_catch, for back compatibility. (tc16_pre_unwind_data, pre_unwind_data_print, make_pre_unwind_data, SCM_PRE_UNWIND_DATA_P): Renamed from "lazy_catch" equivalents. (scm_c_with_throw_handler): New, replaces scm_internal_lazy_catch as the primary C API for a "lazy" catch. (scm_internal_lazy_catch): Now just a wrapper for scm_c_with_throw_handler, for back compatibility. (scm_catch_with_pre_unwind_handler): Renamed from scm_catch; adds pre-unwind handler support. (scm_catch): Now just a wrapper for scm_catch_with_pre_unwind_handler, for back compatibility. (scm_with_throw_handler): New. (scm_lazy_catch): Update comment to say that the handler can return, and what happens if it does. (toggle_pre_unwind_running): New. (scm_ithrow): When identifying the throw target, take running flags into account. In general, change naming of things from "lazy_catch" to "pre_unwind". When throwing to a throw handler, don't unwind the dynamic context first. Add dynwind framing to manage the running flag of a throw handler. If a lazy catch or throw handler returns, rethrow the same exception again. Add pre-unwind support to the normal catch case (SCM_JMPBUFP). * root.c (scm_internal_cwdr): Add NULL args to scm_i_with_continuation_barrier call. * dynwind.c: Change comment mentioning lazy-catch to mention pre-unwind data and throw handler also. * continuations.h (scm_i_with_continuation_barrier): Add pre-unwind handler args. * continuations.c (scm_i_with_continuation_barrier): Add pre-unwind handler args, and pass on to scm_c_catch (changed from scm_internal_catch). (c_handler): Remove scm_handle_by_message_noexit call. (scm_c_with_continuation_barrier): Call scm_i_with_continuation_barrier with scm_handle_by_message_noexit as the pre-unwind handler. (scm_handler): Remove scm_handle_by_message_noexit call. (s_scm_with_continuation_barrier): Call scm_i_with_continuation_barrier with scm_handle_by_message_noexit as the pre-unwind handler.
2006-02-04 14:36:24 +00:00
scm_handler, &scm_data,
pre_unwind_handler,
SCM_UNPACK_POINTER (scm_current_error_port ()));
2005-03-02 20:42:01 +00:00
}
#undef FUNC_NAME
2001-03-04 17:09:34 +00:00
void
scm_init_continuations ()
{
tc16_continuation = scm_make_smob_type ("continuation", 0);
scm_set_smob_print (tc16_continuation, continuation_print);
#include "libguile/continuations.x"
}
/*
Local Variables:
c-file-style: "gnu"
End:
*/