mirror/guile
mirror/guile
1
Fork 0
mirror of https://git.savannah.gnu.org/git/guile.git synced 2025-05-30 17:00:23 +02:00
Code Activity
guile/libguile/__scm.h
Neil Jerram 346e4402a4 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.)
2008-05-12 23:24:28 +01:00

634 lines
20 KiB
C
Raw Blame History

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

/* classes: h_files */
#ifndef SCM___SCM_H
#define SCM___SCM_H
/* Copyright (C) 1995,1996,1998,1999,2000,2001,2002,2003, 2006, 2007, 2008 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 2.1 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
*/
/**********************************************************************
This file is Guile's central public header.
When included by other files, this file should preceed any include
other than __scm.h.
Under *NO* circumstances should new items be added to the global
namespace (via adding #define, typedef, or similar to this file) with
generic names. This usually means that any new names should be
prefixed by either SCM_ or GUILE_. i.e. do *not* #define HAVE_FOO or
SIZEOF_BAR. See configure.in, gen-scmconfig.h.in, and
gen-scmconfig.c for examples of how to properly handle this issue.
The main documentation is in gen-scmconfig.c.
"What's the difference between _scm.h and __scm.h?"
_scm.h is not installed; it's only visible to the libguile sources
themselves, and it includes config.h, the private config header.
__scm.h is installed, and is #included by <libguile.h>. If both
the client and libguile need some piece of information, and it
doesn't fit well into the header file for any particular module, it
should go in __scm.h. __scm.h includes scmconfig.h, the public
config header.
**********************************************************************/
/* What did the configure script discover about the outside world? */
#include "libguile/scmconfig.h"
/* {Compiler hints}
*
* The following macros are used to provide additional information for the
* compiler, which may help to do better error checking and code
* optimization. A second benefit of these macros is, that they also provide
* additional information to the developers.
*/
/* The macro SCM_NORETURN indicates that a function will never return.
* Examples:
* 1) int foo (char arg) SCM_NORETURN;
*/
#ifdef __GNUC__
#define SCM_NORETURN __attribute__ ((noreturn))
#else
#define SCM_NORETURN
#endif
/* The macro SCM_UNUSED indicates that a function, function argument or
* variable may potentially be unused.
* Examples:
* 1) static int unused_function (char arg) SCM_UNUSED;
* 2) int foo (char unused_argument SCM_UNUSED);
* 3) int unused_variable SCM_UNUSED;
*/
#ifdef __GNUC__
#define SCM_UNUSED __attribute__ ((unused))
#else
#define SCM_UNUSED
#endif
/* The SCM_EXPECT macros provide branch prediction hints to the compiler. To
* use only in places where the result of the expression under "normal"
* circumstances is known. */
#if defined(__GNUC__) && (__GNUC__ >= 3)
# define SCM_EXPECT __builtin_expect
#else
# define SCM_EXPECT(_expr, _value) (_expr)
#endif
#define SCM_LIKELY(_expr) SCM_EXPECT ((_expr), 1)
#define SCM_UNLIKELY(_expr) SCM_EXPECT ((_expr), 0)
/* {Supported Options}
*
* These may be defined or undefined.
*/
/* #define GUILE_DEBUG_FREELIST */
/* All the number support there is.
*/
#define BIGNUMS
/* GC should relinquish empty cons-pair arenas. */
/* cmm:FIXME look at this after done mangling the GC */
/* #define GC_FREE_SEGMENTS */
/* Provide a scheme-accessible count-down timer that
* generates a pseudo-interrupt.
*/
#define TICKS
/* Use engineering notation when converting numbers strings?
*/
#undef ENGNOT
/* {Unsupported Options}
*
* These must be defined as given here.
*/
#define CCLO
/* Guile Scheme supports the #f/() distinction; Guile Lisp won't. We
have horrible plans for their unification. */
#undef SICP
/* Random options (not yet supported or in final form). */
#define STACK_CHECKING
#undef NO_CEVAL_STACK_CHECKING
/* SCM_API is a macro prepended to all function and data definitions
which should be exported or imported in the resulting dynamic link
library (DLL) in the Win32 port. */
#if defined (SCM_IMPORT)
# define SCM_API __declspec (dllimport) extern
#elif defined (SCM_EXPORT) || defined (DLL_EXPORT)
# define SCM_API __declspec (dllexport) extern
#else
# define SCM_API extern
#endif
/* {Debugging Options}
*
* These compile time options determine whether to include code that is only
* useful for debugging guile itself or C level extensions to guile. The
* common prefix for all option macros of this kind is "SCM_DEBUG_". It is
* guaranteed that a macro named SCM_DEBUG_XXX is always defined (typically to
* either 0 or 1), i. e. there is no need to test for the undefined case.
* This allows to use these definitions comfortably within code, as in the
* following example:
* #define FOO do { if (SCM_DEBUG_XXX) bar(); else baz(); } while (0)
* Any sane compiler will remove the unused branch without any performance
* penalty for the resulting code.
*
* Note: Some SCM_DEBUG_XXX options are not settable at configure time.
* To change the value of such options you will have to edit this header
* file or give suitable options to make, like:
* make all CFLAGS="-DSCM_DEBUG_XXX=1 ..."
*/
/* The value of SCM_DEBUG determines the default for most of the not yet
* defined debugging options. This allows, for example, to enable most of the
* debugging options by simply defining SCM_DEBUG as 1.
*/
#ifndef SCM_DEBUG
#define SCM_DEBUG 0
#endif
/* If SCM_DEBUG_CELL_ACCESSES is set to 1, cell accesses will perform
* exhaustive parameter checking: It will be verified that cell parameters
* actually point to a valid heap cell. Note: If this option is enabled,
* guile will run about ten times slower than normally.
*/
#ifndef SCM_DEBUG_CELL_ACCESSES
#define SCM_DEBUG_CELL_ACCESSES SCM_DEBUG
#endif
/* If SCM_DEBUG_INTERRUPTS is set to 1, with every deferring and allowing of
* interrupts a consistency check will be performed.
*/
#ifndef SCM_DEBUG_INTERRUPTS
#define SCM_DEBUG_INTERRUPTS SCM_DEBUG
#endif
/* If SCM_DEBUG_PAIR_ACCESSES is set to 1, accesses to cons cells will be
* exhaustively checked. Note: If this option is enabled, guile will run
* slower than normally.
*/
#ifndef SCM_DEBUG_PAIR_ACCESSES
#define SCM_DEBUG_PAIR_ACCESSES SCM_DEBUG
#endif
/* If SCM_DEBUG_REST_ARGUMENT is set to 1, functions that take rest arguments
* will check whether the rest arguments are actually passed as a proper list.
* Otherwise, if SCM_DEBUG_REST_ARGUMENT is 0, functions that take rest
* arguments will take it for granted that these are passed as a proper list.
*/
#ifndef SCM_DEBUG_REST_ARGUMENT
#define SCM_DEBUG_REST_ARGUMENT SCM_DEBUG
#endif
/* The macro SCM_DEBUG_TYPING_STRICTNESS indicates what level of type checking
* shall be performed with respect to the use of the SCM datatype. The macro
* may be defined to one of the values 0, 1 and 2.
*
* A value of 0 means that there will be no compile time type checking, since
* the SCM datatype will be declared as an integral type. This setting should
* only be used on systems, where casting from integral types to pointers may
* lead to loss of bit information.
*
* A value of 1 means that there will an intermediate level of compile time
* type checking, since the SCM datatype will be declared as a pointer to an
* undefined struct. This setting is the default, since it does not cost
* anything in terms of performance or code size.
*
* A value of 2 provides a maximum level of compile time type checking since
* the SCM datatype will be declared as a struct. This setting should be used
* for _compile time_ type checking only, since the compiled result is likely
* to be quite inefficient. The right way to make use of this option is to do
* a 'make clean; make CFLAGS=-DSCM_DEBUG_TYPING_STRICTNESS=2', fix your
* errors, and then do 'make clean; make'.
*/
#ifndef SCM_DEBUG_TYPING_STRICTNESS
#define SCM_DEBUG_TYPING_STRICTNESS 1
#endif
/* If SCM_DEBUG_DEBUGGING_SUPPORT is set to 1, guile will provide a set of
* special functions that support debugging with a debugger like gdb or
* debugging of guile internals on the scheme level. The behaviour of guile
* is not changed by this macro, only the set of functions that are available
* will differ. All functions that are introduced this way have the prefix
* 'scm_dbg_' on the C level and the prefix 'dbg-' on the scheme level. This
* allows to easily determine the set of support functions, given that your
* debugger or repl provide automatic name completion. Note that these
* functions are intended to be used during interactive debugging sessions
* only. They are not considered part of guile's official API. They may
* change or disappear without notice or deprecation phase.
*/
#ifndef SCM_DEBUG_DEBUGGING_SUPPORT
#define SCM_DEBUG_DEBUGGING_SUPPORT SCM_DEBUG
#endif
/* {Feature Options}
*
* These compile time options determine whether code for certain features
* should be compiled into guile. The common prefix for all option macros
* of this kind is "SCM_ENABLE_". It is guaranteed that a macro named
* SCM_ENABLE_XXX is defined to be either 0 or 1, i. e. there is no need to
* test for the undefined case. This allows to use these definitions
* comfortably within code, as in the following example:
* #define FOO do { if (SCM_ENABLE_XXX) bar(); else baz(); } while (0)
* Any sane compiler will remove the unused branch without any performance
* penalty for the resulting code.
*
* Note: Some SCM_ENABLE_XXX options are not settable at configure time.
* To change the value of such options you will have to edit this header
* file or give suitable options to make, like:
* make all CFLAGS="-DSCM_ENABLE_XXX=1 ..."
*/
/* If SCM_ENABLE_DEPRECATED is set to 1, deprecated code will be included in
* guile, as well as some functions to issue run-time warnings about uses of
* deprecated functions.
*/
#ifndef SCM_ENABLE_DEPRECATED
#define SCM_ENABLE_DEPRECATED 0
#endif
/* {Architecture and compiler properties}
*
* Guile as of today can only work on systems which fulfill at least the
* following requirements:
*
* - scm_t_bits and SCM variables have at least 32 bits.
* Guile's type system is based on this assumption.
*
* - sizeof (scm_t_bits) >= sizeof (void*) and sizeof (SCM) >= sizeof (void*)
* Guile's type system is based on this assumption, since it must be
* possible to store pointers to cells on the heap in scm_t_bits and SCM
* variables.
*
* - sizeof (scm_t_bits) >= 4 and sizeof (scm_t_bits) is a power of 2.
* Guile's type system is based on this assumption. In particular, it is
* assumed that cells, i. e. pairs of scm_t_bits variables, are eight
* character aligned. This is because three bits of a scm_t_bits variable
* that is holding a pointer to a cell on the heap must be available for
* storing type data.
*
* - sizeof (scm_t_bits) <= sizeof (void*) and sizeof (SCM) <= sizeof (void*)
* In some parts of guile, scm_t_bits and SCM variables are passed to
* functions as void* arguments. Together with the requirement above, this
* requires a one-to-one correspondence between the size of a void* and the
* sizes of scm_t_bits and SCM variables.
*
* - numbers are encoded using two's complement.
* The implementation of the bitwise scheme level operations is based on
* this assumption.
*
* - ... add more
*/
#ifdef CHAR_BIT
# define SCM_CHAR_BIT CHAR_BIT
#else
# define SCM_CHAR_BIT 8
#endif
#ifdef LONG_BIT
# define SCM_LONG_BIT LONG_BIT
#else
# define SCM_LONG_BIT (SCM_CHAR_BIT * sizeof (long) / sizeof (char))
#endif
#ifdef UCHAR_MAX
# define SCM_CHAR_CODE_LIMIT (UCHAR_MAX + 1L)
#else
# define SCM_CHAR_CODE_LIMIT 256L
#endif
#define SCM_I_UTYPE_MAX(type) ((type)-1)
#define SCM_I_TYPE_MAX(type,umax) ((type)((umax)/2))
#define SCM_I_TYPE_MIN(type,umax) (-((type)((umax)/2))-1)
#define SCM_T_UINT8_MAX SCM_I_UTYPE_MAX(scm_t_uint8)
#define SCM_T_INT8_MIN SCM_I_TYPE_MIN(scm_t_int8,SCM_T_UINT8_MAX)
#define SCM_T_INT8_MAX SCM_I_TYPE_MAX(scm_t_int8,SCM_T_UINT8_MAX)
#define SCM_T_UINT16_MAX SCM_I_UTYPE_MAX(scm_t_uint16)
#define SCM_T_INT16_MIN SCM_I_TYPE_MIN(scm_t_int16,SCM_T_UINT16_MAX)
#define SCM_T_INT16_MAX SCM_I_TYPE_MAX(scm_t_int16,SCM_T_UINT16_MAX)
#define SCM_T_UINT32_MAX SCM_I_UTYPE_MAX(scm_t_uint32)
#define SCM_T_INT32_MIN SCM_I_TYPE_MIN(scm_t_int32,SCM_T_UINT32_MAX)
#define SCM_T_INT32_MAX SCM_I_TYPE_MAX(scm_t_int32,SCM_T_UINT32_MAX)
#if SCM_HAVE_T_INT64
#define SCM_T_UINT64_MAX SCM_I_UTYPE_MAX(scm_t_uint64)
#define SCM_T_INT64_MIN SCM_I_TYPE_MIN(scm_t_int64,SCM_T_UINT64_MAX)
#define SCM_T_INT64_MAX SCM_I_TYPE_MAX(scm_t_int64,SCM_T_UINT64_MAX)
#endif
#if SCM_SIZEOF_LONG_LONG
#define SCM_I_ULLONG_MAX SCM_I_UTYPE_MAX(unsigned long long)
#define SCM_I_LLONG_MIN SCM_I_TYPE_MIN(long long,SCM_I_ULLONG_MAX)
#define SCM_I_LLONG_MAX SCM_I_TYPE_MAX(long long,SCM_I_ULLONG_MAX)
#endif
#define SCM_T_UINTMAX_MAX SCM_I_UTYPE_MAX(scm_t_uintmax)
#define SCM_T_INTMAX_MIN SCM_I_TYPE_MIN(scm_t_intmax,SCM_T_UINTMAX_MAX)
#define SCM_T_INTMAX_MAX SCM_I_TYPE_MAX(scm_t_intmax,SCM_T_UINTMAX_MAX)
#define SCM_I_SIZE_MAX SCM_I_UTYPE_MAX(size_t)
#define SCM_I_SSIZE_MIN SCM_I_TYPE_MIN(ssize_t,SCM_I_SIZE_MAX)
#define SCM_I_SSIZE_MAX SCM_I_TYPE_MAX(ssize_t,SCM_I_SIZE_MAX)
#include "libguile/tags.h"
#ifdef vms
# ifndef CHEAP_CONTINUATIONS
typedef int jmp_buf[17];
extern int setjump(jmp_buf env);
extern int longjump(jmp_buf env, int ret);
# define setjmp setjump
# define longjmp longjump
# else
# include <setjmp.h>
# endif
#else /* ndef vms */
# ifdef _CRAY1
typedef int jmp_buf[112];
extern int setjump(jmp_buf env);
extern int longjump(jmp_buf env, int ret);
# define setjmp setjump
# define longjmp longjump
# else /* ndef _CRAY1 */
# if defined (__ia64__)
/* For IA64, emulate the setjmp API using getcontext. */
# include <signal.h>
# include <ucontext.h>
typedef struct {
ucontext_t ctx;
int fresh;
} jmp_buf;
# define setjmp(JB) \
( (JB).fresh = 1, \
getcontext (&((JB).ctx)), \
((JB).fresh ? ((JB).fresh = 0, 0) : 1) )
# define longjmp(JB,VAL) scm_ia64_longjmp (&(JB), VAL)
void scm_ia64_longjmp (jmp_buf *, int);
# else /* ndef __ia64__ */
# include <setjmp.h>
# endif /* ndef __ia64__ */
# endif /* ndef _CRAY1 */
#endif /* ndef vms */
/* 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
/* If stack is not longword aligned then
*/
/* #define SHORT_ALIGN */
#ifdef THINK_C
# define SHORT_ALIGN
#endif
#ifdef MSDOS
# define SHORT_ALIGN
#endif
#ifdef atarist
# define SHORT_ALIGN
#endif
#ifdef SHORT_ALIGN
typedef short SCM_STACKITEM;
#else
typedef long SCM_STACKITEM;
#endif
/* Cast pointer through (void *) in order to avoid compiler warnings
when strict aliasing is enabled */
#define SCM_STACK_PTR(ptr) ((SCM_STACKITEM *) (void *) (ptr))
#define SCM_ASYNC_TICK /*fixme* should change names */ \
do { \
if (SCM_I_CURRENT_THREAD->pending_asyncs) \
scm_async_click (); \
} while (0)
/* Anthony Green writes:
When the compiler sees...
DEFER_INTS;
[critical code here]
ALLOW_INTS;
...it doesn't actually promise to keep the critical code within the
boundries of the DEFER/ALLOW_INTS instructions. It may very well
schedule it outside of the magic defined in those macros.
However, GCC's volatile asm feature forms a barrier over which code is
never moved. So if you add...
asm ("");
...to each of the DEFER_INTS and ALLOW_INTS macros, the critical
code will always remain in place. asm's without inputs or outputs
are implicitly volatile. */
#ifdef __GNUC__
#define SCM_FENCE asm /* volatile */ ("")
#elif defined (__INTEL_COMPILER) && defined (__ia64)
#define SCM_FENCE __memory_barrier()
#else
#define SCM_FENCE
#endif
#define SCM_TICK \
do { \
SCM_ASYNC_TICK; \
SCM_THREAD_SWITCHING_CODE; \
} while (0)
/** SCM_ASSERT
**
**/
#ifdef SCM_RECKLESS
#define SCM_ASSERT(_cond, _arg, _pos, _subr)
#define SCM_ASSERT_TYPE(_cond, _arg, _pos, _subr, _msg)
#define SCM_ASRTGO(_cond, _label)
#else
#define SCM_ASSERT(_cond, _arg, _pos, _subr) \
do { if (SCM_UNLIKELY (!(_cond))) \
scm_wrong_type_arg (_subr, _pos, _arg); } while (0)
#define SCM_ASSERT_TYPE(_cond, _arg, _pos, _subr, _msg) \
do { if (SCM_UNLIKELY (!(_cond))) \
scm_wrong_type_arg_msg(_subr, _pos, _arg, _msg); } while (0)
#define SCM_ASRTGO(_cond, _label) \
do { if (SCM_UNLIKELY (!(_cond))) \
goto _label; } while (0)
#endif
/*
* SCM_WTA_DISPATCH
*/
/* Dirk:FIXME:: In all of the SCM_WTA_DISPATCH_* macros it is assumed that
* 'gf' is zero if uninitialized. It would be cleaner if some valid SCM value
* like SCM_BOOL_F or SCM_UNDEFINED was chosen.
*/
SCM_API SCM scm_call_generic_0 (SCM gf);
#define SCM_WTA_DISPATCH_0(gf, subr) \
return (SCM_UNPACK (gf) \
? scm_call_generic_0 ((gf)) \
: (scm_error_num_args_subr ((subr)), SCM_UNSPECIFIED))
#define SCM_GASSERT0(cond, gf, subr) \
if (SCM_UNLIKELY(!(cond))) \
SCM_WTA_DISPATCH_0((gf), (subr))
SCM_API SCM scm_call_generic_1 (SCM gf, SCM a1);
#define SCM_WTA_DISPATCH_1(gf, a1, pos, subr) \
return (SCM_UNPACK (gf) \
? scm_call_generic_1 ((gf), (a1)) \
: (scm_wrong_type_arg ((subr), (pos), (a1)), SCM_UNSPECIFIED))
#define SCM_GASSERT1(cond, gf, a1, pos, subr) \
if (SCM_UNLIKELY (!(cond))) \
SCM_WTA_DISPATCH_1((gf), (a1), (pos), (subr))
SCM_API SCM scm_call_generic_2 (SCM gf, SCM a1, SCM a2);
#define SCM_WTA_DISPATCH_2(gf, a1, a2, pos, subr) \
return (SCM_UNPACK (gf) \
? scm_call_generic_2 ((gf), (a1), (a2)) \
: (scm_wrong_type_arg ((subr), (pos), \
(pos) == SCM_ARG1 ? (a1) : (a2)), \
SCM_UNSPECIFIED))
#define SCM_GASSERT2(cond, gf, a1, a2, pos, subr) \
if (SCM_UNLIKELY (!(cond))) \
SCM_WTA_DISPATCH_2((gf), (a1), (a2), (pos), (subr))
SCM_API SCM scm_apply_generic (SCM gf, SCM args);
#define SCM_WTA_DISPATCH_n(gf, args, pos, subr) \
return (SCM_UNPACK (gf) \
? scm_apply_generic ((gf), (args)) \
: (scm_wrong_type_arg ((subr), (pos), \
scm_list_ref ((args), \
scm_from_int ((pos) - 1))), \
SCM_UNSPECIFIED))
#define SCM_GASSERTn(cond, gf, args, pos, subr) \
if (SCM_UNLIKELY (!(cond))) \
SCM_WTA_DISPATCH_n((gf), (args), (pos), (subr))
#ifndef SCM_MAGIC_SNARFER
/* Let these macros pass through if
we are snarfing; thus we can tell the
difference between the use of an actual
number vs. the use of one of these macros --
actual numbers in SCM_VALIDATE_* and SCM_ASSERT
constructs must match the formal argument name,
but using SCM_ARG* avoids the test */
#define SCM_ARGn 0
#define SCM_ARG1 1
#define SCM_ARG2 2
#define SCM_ARG3 3
#define SCM_ARG4 4
#define SCM_ARG5 5
#define SCM_ARG6 6
#define SCM_ARG7 7
#endif /* SCM_MAGIC_SNARFER */
/* SCM_EXIT_SUCCESS is the default code to return from SCM if no errors
* were encountered. SCM_EXIT_FAILURE is the default code to return from
* SCM if errors were encountered. The return code can be explicitly
* specified in a SCM program with (scm_quit <n>).
*/
#ifndef SCM_EXIT_SUCCESS
#ifdef vms
#define SCM_EXIT_SUCCESS 1
#else
#define SCM_EXIT_SUCCESS 0
#endif /* def vms */
#endif /* ndef SCM_EXIT_SUCCESS */
#ifndef SCM_EXIT_FAILURE
#ifdef vms
#define SCM_EXIT_FAILURE 2
#else
#define SCM_EXIT_FAILURE 1
#endif /* def vms */
#endif /* ndef SCM_EXIT_FAILURE */
/* Define SCM_C_INLINE_KEYWORD so that it can be used as a replacement
for the "inline" keyword, expanding to nothing when "inline" is not
available.
*/
#ifdef SCM_C_INLINE
#define SCM_C_INLINE_KEYWORD SCM_C_INLINE
#else
#define SCM_C_INLINE_KEYWORD
#endif
#endif /* SCM___SCM_H */
/*
Local Variables:
c-file-style: "gnu"
End:
*/
View git blame Copy permalink