2011-01-26 21:44:12 +01:00
|
|
|
/* Copyright (C) 2001, 2009, 2010, 2011 Free Software Foundation, Inc.
|
2000-08-22 15:54:19 +00:00
|
|
|
*
|
2009-05-26 22:23:44 +02:00
|
|
|
* This library is free software; you can redistribute it and/or
|
2009-06-17 00:22:09 +01:00
|
|
|
* 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.
|
2000-08-22 15:54:19 +00:00
|
|
|
*
|
2009-06-17 00:22:09 +01:00
|
|
|
* This library is distributed in the hope that it will be useful, but
|
|
|
|
|
* WITHOUT ANY WARRANTY; without even the implied warranty of
|
2009-05-26 22:23:44 +02:00
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
|
|
|
* Lesser General Public License for more details.
|
2000-08-22 15:54:19 +00:00
|
|
|
*
|
2009-05-26 22:23:44 +02:00
|
|
|
* You should have received a copy of the GNU Lesser General Public
|
|
|
|
|
* License along with this library; if not, write to the Free Software
|
2009-06-17 00:22:09 +01:00
|
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
|
|
|
|
|
* 02110-1301 USA
|
2009-05-26 22:23:44 +02:00
|
|
|
*/
|
2000-08-22 15:54:19 +00:00
|
|
|
|
enable multiple vm engines (regular, debug, ...)
* libguile/vm-engine.c (VM_USE_HOOKS, VM_USE_CLOCK, VM_CHECK_EXTERNAL)
(VM_CHECK_OBJECT): Update to define these here, before including
vm-engine.h.
(vm_run): Change so that we can make different engines. Also, we take
an array of arguments, and the struct scm_vm directly, so as to avoid
any need to cons.
* libguile/vm-engine.h (CHECK_EXTERNAL, CHECK_OBJECT): Add some UNLIKELY
bits; don't seem to help.
* libguile/vm.c (vm_dispatch_hook): Change to not pass the VP. This needs
some love, and perhaps we revert to the old way.
(VM_ENGINE): Actually make two engines, vm_regular_engine and
vm_debug_engine. Probably there is room for improvement here. Actually
their speeds are the same at the moment.
(make_vm): Choose which engine to run; currently the debug engine by
default.
(scm_c_vm_run): A thin wrapper to invoke a VM without consing.
(scm_vm_apply): Use scm_c_vm_run.
(scm_load_compiled_with_vm): Use scm_c_vm_run.
2009-02-04 23:47:56 +01:00
|
|
|
/* This file is included in vm.c multiple times */
|
|
|
|
|
|
|
|
|
|
#if (VM_ENGINE == SCM_VM_REGULAR_ENGINE)
|
|
|
|
|
#define VM_USE_HOOKS 0 /* Various hooks */
|
VM tweaks
* libguile/vm-engine.c (VM_CHECK_OBJECT, VM_CHECK_FREE_VARIABLES): Set
to 0 for both engines. These are really internal debugging variables,
which don't affect user-visible features, provided that the compiler
is correct of course.
(VM_CHECK_UNDERFLOW): New var, also off by default: whether to check
for stack underflow when popping values.
(vm_engine): Don't declare object_count if we are not checking object
table accesses.
* libguile/vm-engine.h (CACHE_PROGRAM): Don't muck with object_count
if we are not checking object table accesses.
(CHECK_UNDERFLOW, PRE_CHECK_UNDERFLOW): Nop out if we are not checking
underflow.
(POP2, POP3): New macros which check for underflow before popping more
than one value.
* libguile/vm-i-loader.c (load_array):
* libguile/vm-i-scheme.c (set_car, set_cdr, vector_set, slot_set)
(BV_SET_WITH_ENDIANNESS, BV_FIXABLE_INT_SET, BV_INT_SET)
(BV_FLOAT_SET):
* libguile/vm-i-system.c (partial_cont_call, fix_closure, prompt)
(fluid_set): Use POP2 / POP3.
(local_set, long_local_set): Pop to locals instead of using values on
the stack then dropping; allows for underflow to be checked before the
value is accessed.
(BR): Don't NULLSTACK or DROP after the operation.
(br_if, br_if_not, br_if_eq, br_if_not_eq, br_if_null)
(br_if_not_null): Pop to locals before doing the compare and jump.
2011-05-05 14:04:23 +02:00
|
|
|
#define VM_CHECK_OBJECT 0 /* Check object table */
|
|
|
|
|
#define VM_CHECK_FREE_VARIABLES 0 /* Check free variable access */
|
|
|
|
|
#define VM_CHECK_UNDERFLOW 0 /* Check underflow when popping values */
|
enable multiple vm engines (regular, debug, ...)
* libguile/vm-engine.c (VM_USE_HOOKS, VM_USE_CLOCK, VM_CHECK_EXTERNAL)
(VM_CHECK_OBJECT): Update to define these here, before including
vm-engine.h.
(vm_run): Change so that we can make different engines. Also, we take
an array of arguments, and the struct scm_vm directly, so as to avoid
any need to cons.
* libguile/vm-engine.h (CHECK_EXTERNAL, CHECK_OBJECT): Add some UNLIKELY
bits; don't seem to help.
* libguile/vm.c (vm_dispatch_hook): Change to not pass the VP. This needs
some love, and perhaps we revert to the old way.
(VM_ENGINE): Actually make two engines, vm_regular_engine and
vm_debug_engine. Probably there is room for improvement here. Actually
their speeds are the same at the moment.
(make_vm): Choose which engine to run; currently the debug engine by
default.
(scm_c_vm_run): A thin wrapper to invoke a VM without consing.
(scm_vm_apply): Use scm_c_vm_run.
(scm_load_compiled_with_vm): Use scm_c_vm_run.
2009-02-04 23:47:56 +01:00
|
|
|
#elif (VM_ENGINE == SCM_VM_DEBUG_ENGINE)
|
|
|
|
|
#define VM_USE_HOOKS 1
|
VM tweaks
* libguile/vm-engine.c (VM_CHECK_OBJECT, VM_CHECK_FREE_VARIABLES): Set
to 0 for both engines. These are really internal debugging variables,
which don't affect user-visible features, provided that the compiler
is correct of course.
(VM_CHECK_UNDERFLOW): New var, also off by default: whether to check
for stack underflow when popping values.
(vm_engine): Don't declare object_count if we are not checking object
table accesses.
* libguile/vm-engine.h (CACHE_PROGRAM): Don't muck with object_count
if we are not checking object table accesses.
(CHECK_UNDERFLOW, PRE_CHECK_UNDERFLOW): Nop out if we are not checking
underflow.
(POP2, POP3): New macros which check for underflow before popping more
than one value.
* libguile/vm-i-loader.c (load_array):
* libguile/vm-i-scheme.c (set_car, set_cdr, vector_set, slot_set)
(BV_SET_WITH_ENDIANNESS, BV_FIXABLE_INT_SET, BV_INT_SET)
(BV_FLOAT_SET):
* libguile/vm-i-system.c (partial_cont_call, fix_closure, prompt)
(fluid_set): Use POP2 / POP3.
(local_set, long_local_set): Pop to locals instead of using values on
the stack then dropping; allows for underflow to be checked before the
value is accessed.
(BR): Don't NULLSTACK or DROP after the operation.
(br_if, br_if_not, br_if_eq, br_if_not_eq, br_if_null)
(br_if_not_null): Pop to locals before doing the compare and jump.
2011-05-05 14:04:23 +02:00
|
|
|
#define VM_CHECK_OBJECT 0
|
|
|
|
|
#define VM_CHECK_FREE_VARIABLES 0
|
|
|
|
|
#define VM_CHECK_UNDERFLOW 0 /* Check underflow when popping values */
|
enable multiple vm engines (regular, debug, ...)
* libguile/vm-engine.c (VM_USE_HOOKS, VM_USE_CLOCK, VM_CHECK_EXTERNAL)
(VM_CHECK_OBJECT): Update to define these here, before including
vm-engine.h.
(vm_run): Change so that we can make different engines. Also, we take
an array of arguments, and the struct scm_vm directly, so as to avoid
any need to cons.
* libguile/vm-engine.h (CHECK_EXTERNAL, CHECK_OBJECT): Add some UNLIKELY
bits; don't seem to help.
* libguile/vm.c (vm_dispatch_hook): Change to not pass the VP. This needs
some love, and perhaps we revert to the old way.
(VM_ENGINE): Actually make two engines, vm_regular_engine and
vm_debug_engine. Probably there is room for improvement here. Actually
their speeds are the same at the moment.
(make_vm): Choose which engine to run; currently the debug engine by
default.
(scm_c_vm_run): A thin wrapper to invoke a VM without consing.
(scm_vm_apply): Use scm_c_vm_run.
(scm_load_compiled_with_vm): Use scm_c_vm_run.
2009-02-04 23:47:56 +01:00
|
|
|
#else
|
|
|
|
|
#error unknown debug engine VM_ENGINE
|
|
|
|
|
#endif
|
2000-08-22 15:54:19 +00:00
|
|
|
|
2008-08-21 18:39:30 -07:00
|
|
|
#include "vm-engine.h"
|
2000-08-22 15:54:19 +00:00
|
|
|
|
2005-04-28 15:45:59 +00:00
|
|
|
|
2000-08-22 15:54:19 +00:00
|
|
|
static SCM
|
rework vm tracing
* libguile/vm-engine.c (VM_NAME): Engines take the VM itself (not the
vp), so they can pass the VM to hooks. No more hook args, we dispatch
without them.
* libguile/vm-engine.h (RUN_HOOK): Dispatch the hook if the trace level
is positive (instead of if the hook is there). Don't cache registers
on return from the dispatch.
* libguile/vm.h:
* libguile/vm.c (vm_dispatch_hook): Don't bother with a dynwind; instead
decrement the trace level when going into a hook, and if we have a
nonlocal exit, the trace level never gets incremented again. Worse is
better.
(make_vm, scm_vm_trace_level, scm_set_vm_trace_level_x): New concept,
trace level. If positive, we run the hooks, otherwise we don't. Should
work. Removed scm_vm_trace_frame, I don't think that was the right way
to do it.
* module/system/vm/vm.scm: Replace vm-trace-frame with vm-trace-level
and set-vm-trace-level!; the hooks actually get the frame as an
argument now.
2009-12-21 21:57:20 +01:00
|
|
|
VM_NAME (SCM vm, SCM program, SCM *argv, int nargs)
|
2000-08-22 15:54:19 +00:00
|
|
|
{
|
2001-04-01 05:03:41 +00:00
|
|
|
/* VM registers */
|
programs have their own tc7 now
* libguile/tags.h (scm_tc7_program):
* libguile/programs.h: Programs now have their own tc7 code. Fix up the
macros appropriately.
* libguile/programs.c: Remove smobby bits, leaving marking, printing,
and application for other parts of Guile.
* libguile/debug.c (scm_procedure_source):
* libguile/eval.c (scm_trampoline_0, scm_trampoline_1)
(scm_trampoline_2): Add cases for tc7_program.
* libguile/eval.i.c (CEVAL, SCM_APPLY):
* libguile/evalext.c (scm_self_evaluating_p):
* libguile/gc-card.c (scm_i_sweep_card, scm_i_tag_name):
* libguile/gc-mark.c (1):
* libguile/print.c (iprin1):
* libguile/procs.c (scm_procedure_p, scm_thunk_p)
* libguile/vm-i-system.c (make-closure): Adapt to new procedure
representation.
* libguile/procprop.c (scm_i_procedure_arity): Do the right thing for
programs.
* test-suite/tests/procprop.test ("procedure-arity"): Arity test now
succeeds.
* libguile/goops.c (scm_class_of): Programs now belong to the class
<procedure>, not a smob class.
* libguile/vm.h (struct vm, struct vm_cont):
* libguile/vm-engine.c (vm_engine):
* libguile/frames.h (SCM_FRAME_BYTE_CAST, struct vm_frame):
* libguile/frames.c (scm_c_make_vm_frame): Fix usages of scm_byte_t,
changing them to scm_t_uint8.
2009-08-20 14:27:38 +02:00
|
|
|
register scm_t_uint8 *ip IP_REG; /* instruction pointer */
|
2001-04-01 05:03:41 +00:00
|
|
|
register SCM *sp SP_REG; /* stack pointer */
|
|
|
|
|
register SCM *fp FP_REG; /* frame pointer */
|
rework vm tracing
* libguile/vm-engine.c (VM_NAME): Engines take the VM itself (not the
vp), so they can pass the VM to hooks. No more hook args, we dispatch
without them.
* libguile/vm-engine.h (RUN_HOOK): Dispatch the hook if the trace level
is positive (instead of if the hook is there). Don't cache registers
on return from the dispatch.
* libguile/vm.h:
* libguile/vm.c (vm_dispatch_hook): Don't bother with a dynwind; instead
decrement the trace level when going into a hook, and if we have a
nonlocal exit, the trace level never gets incremented again. Worse is
better.
(make_vm, scm_vm_trace_level, scm_set_vm_trace_level_x): New concept,
trace level. If positive, we run the hooks, otherwise we don't. Should
work. Removed scm_vm_trace_frame, I don't think that was the right way
to do it.
* module/system/vm/vm.scm: Replace vm-trace-frame with vm-trace-level
and set-vm-trace-level!; the hooks actually get the frame as an
argument now.
2009-12-21 21:57:20 +01:00
|
|
|
struct scm_vm *vp = SCM_VM_DATA (vm);
|
2000-08-22 15:54:19 +00:00
|
|
|
|
2000-08-22 19:02:22 +00:00
|
|
|
/* Cache variables */
|
static opcodes; refactor program/objcode division; use new assembly pipeline
* gdbinit: Untested attempts to get the stack fondling macros to deal
with the new program representation.
* libguile/frames.c (scm_vm_frame_arguments, scm_vm_frame_source)
(scm_vm_frame_local_ref, scm_vm_frame_local_set_x): SCM_PROGRAM_DATA is
a struct scm_objcode*.
* libguile/instructions.h:
* libguile/instructions.c: Hide the instruction table and the struct
scm_instruction structure; all access to instructions now goes through
procedures. This is because instructions are no longer in a packed
array indexed by opcode. Also, declare a mask that all instructions
should fit in.
* libguile/objcodes.h:
* libguile/objcodes.c: Rewrite so that object code directly maps its
arity and length from its bytecode. This makes it unnecessary to keep
this information in programs, allowing programs to be simple conses
between the code (objcodes) and data (the object table and the closure
variables).
* libguile/programs.c (scm_make_program): Rework so that make-program
takes objcode, an object table, and externals as arguments. It's much
clearer this way, and we avoid malloc().
* libguile/stacks.c (is_vm_bootstrap_frame): Update for program/objcode
changes.
* libguile/vm-engine.c (vm_run): Initialize the jump table on the first
run, with the opcodes declared in the instruction sources, and with bad
instructions raising an error instead of wandering off into the
Unknown.
* libguile/vm-engine.h (FETCH_LENGTH): Always represent lengths as 3
bytes. The old code was too error-prone.
(NEXT_JUMP): Mask the instruction with SCM_VM_INSTRUCTION_MASK.
(NEW_FRAME): Update for program/objcode changes.
* libguile/vm-expand.h (VM_DEFINE_FUNCTION, VM_DEFINE_INSTRUCTION)
(VM_DEFINE_LOADER): Update so that we explicitly specify opcodes, so
that we have a stable bytecode API.
* libguile/vm-i-loader.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(load-integer): Use an int instead of a long as the accumulator; still
need to revisit this code at some point, I think.
(load-program): Simplify, thankfully!! Just creates the objcode slice
and rolls with it.
* libguile/vm-i-scheme.c: Number the opcodes explicitly.
* libguile/vm-i-system.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(make-closure): Update for new program API.
* libguile/vm.c (vm_make_boot_program): Update for new program/objcode
API. Still a bit ugly.
(scm_load_compiled_with_vm): Update for new program/objcode API.
* module/language/assembly.scm (byte-length): Fix byte-length calculation
for loaders, and load-program.
(code-pack, code-unpack): Start to move things from (system vm conv)
here.
(object->code, code->object): More things from conv.scm.
* module/language/glil.scm (<glil-program>): Add a new field,
closure-level.
(make-glil-program, compute-closure-level): Calculate the "closure
level" when making a glil program. This is the maximum depth of
external binding refs in this closure.
(unparse-glil): Fix label serialization.
* module/language/glil/compile-assembly.scm (make-meta): Prepend #f for
the meta's object table, though maybe in the future we can avoid
creating assembly in the first place.
(assoc-ref-or-acons, object-index-and-alist): GRRR! Caught again by the
different sets of arguments to assoc and assoc-ref!
(glil->assembly): Attempt to make the <glil-program> case more
readable, and fix the bugs. Sorry I don't know how to comment this
change any more than this.
(glil->assembly): For <glil-module> serialize the whole key, not just
the name.
(dump-object): subprogram-code is already a list. Serialize integers as
strings, not u8vectors. Fix the order of lists and vectors.
* module/language/glil/spec.scm (glil): Switch orders, so we prefer glil
-> assembly -> objcode. Actually glil->objcode doesn't work any more,
needs to be removed I think.
* module/language/objcode/spec.scm (objcode->value):
s/objcode->program/make-program/.
* module/language/scheme/inline.scm: Add acons inline.
* module/system/vm/conv.scm (make-byte-decoder): Skip the first 8 bytes,
they are header. Handle subprograms properly. Still needs help though.
(decode-length): Lengths are always 3 bytes now.
* module/system/vm/disasm.scm: Superficial changes to keep things
working. I'd like to fix this better in the future.
* module/system/vm/frame.scm (bootstrap-frame?): Fixes for
program-bytecode.
* module/system/vm/program.scm: Export make-program. It's program-objcode
now, no more program-bytecode.
* module/system/vm/vm.scm (vm-load): Use make-program.
* test-suite/tests/asm-to-bytecode.test: New test, very minimal.
* module/system/vm/objcode.scm: Export word-size, byte-order, and
write-objcode.
2009-01-29 21:09:04 +01:00
|
|
|
struct scm_objcode *bp = NULL; /* program base pointer */
|
2001-04-01 05:03:41 +00:00
|
|
|
SCM *objects = NULL; /* constant objects */
|
VM tweaks
* libguile/vm-engine.c (VM_CHECK_OBJECT, VM_CHECK_FREE_VARIABLES): Set
to 0 for both engines. These are really internal debugging variables,
which don't affect user-visible features, provided that the compiler
is correct of course.
(VM_CHECK_UNDERFLOW): New var, also off by default: whether to check
for stack underflow when popping values.
(vm_engine): Don't declare object_count if we are not checking object
table accesses.
* libguile/vm-engine.h (CACHE_PROGRAM): Don't muck with object_count
if we are not checking object table accesses.
(CHECK_UNDERFLOW, PRE_CHECK_UNDERFLOW): Nop out if we are not checking
underflow.
(POP2, POP3): New macros which check for underflow before popping more
than one value.
* libguile/vm-i-loader.c (load_array):
* libguile/vm-i-scheme.c (set_car, set_cdr, vector_set, slot_set)
(BV_SET_WITH_ENDIANNESS, BV_FIXABLE_INT_SET, BV_INT_SET)
(BV_FLOAT_SET):
* libguile/vm-i-system.c (partial_cont_call, fix_closure, prompt)
(fluid_set): Use POP2 / POP3.
(local_set, long_local_set): Pop to locals instead of using values on
the stack then dropping; allows for underflow to be checked before the
value is accessed.
(BR): Don't NULLSTACK or DROP after the operation.
(br_if, br_if_not, br_if_eq, br_if_not_eq, br_if_null)
(br_if_not_null): Pop to locals before doing the compare and jump.
2011-05-05 14:04:23 +02:00
|
|
|
#if VM_CHECK_OBJECT
|
move module and meta inside programs' object tables
* libguile/programs.h (struct scm_program): Remove the module and meta
fields.
* libguile/programs.c (scm_c_make_program): Add a new argument, `objs'.
If it's a vector, we'll look for the module and the metadata in there,
instead of having them in the scm_program structure.
(scm_c_make_closure, program_mark, scm_program_meta)
(scm_c_program_source, scm_program_module): Adapt to the new program
representation.
* libguile/objcodes.c (scm_objcode_to_program): Pass #f as the object
table when making the program.
* libguile/vm-engine.h (CACHE_PROGRAM):
* libguile/vm-engine.c (vm_run): Rework to use the simple vector API for
getting the current object table. Call the helper,
vm_make_boot_program, to make the boot program.
* libguile/vm-i-loader.c (load-program): Set the current module and the
meta in the object vector, which we pass to scm_c_make_program.
* libguile/vm-i-system.c (toplevel-ref, toplevel-set): Adapt to the new
program representation.
* module/language/glil/compile-objcode.scm (codegen): Clarify.
2009-01-17 16:42:53 +01:00
|
|
|
size_t object_count = 0; /* length of OBJECTS */
|
VM tweaks
* libguile/vm-engine.c (VM_CHECK_OBJECT, VM_CHECK_FREE_VARIABLES): Set
to 0 for both engines. These are really internal debugging variables,
which don't affect user-visible features, provided that the compiler
is correct of course.
(VM_CHECK_UNDERFLOW): New var, also off by default: whether to check
for stack underflow when popping values.
(vm_engine): Don't declare object_count if we are not checking object
table accesses.
* libguile/vm-engine.h (CACHE_PROGRAM): Don't muck with object_count
if we are not checking object table accesses.
(CHECK_UNDERFLOW, PRE_CHECK_UNDERFLOW): Nop out if we are not checking
underflow.
(POP2, POP3): New macros which check for underflow before popping more
than one value.
* libguile/vm-i-loader.c (load_array):
* libguile/vm-i-scheme.c (set_car, set_cdr, vector_set, slot_set)
(BV_SET_WITH_ENDIANNESS, BV_FIXABLE_INT_SET, BV_INT_SET)
(BV_FLOAT_SET):
* libguile/vm-i-system.c (partial_cont_call, fix_closure, prompt)
(fluid_set): Use POP2 / POP3.
(local_set, long_local_set): Pop to locals instead of using values on
the stack then dropping; allows for underflow to be checked before the
value is accessed.
(BR): Don't NULLSTACK or DROP after the operation.
(br_if, br_if_not, br_if_eq, br_if_not_eq, br_if_null)
(br_if_not_null): Pop to locals before doing the compare and jump.
2011-05-05 14:04:23 +02:00
|
|
|
#endif
|
2001-04-06 00:17:39 +00:00
|
|
|
SCM *stack_limit = vp->stack_limit; /* stack limit address */
|
abort always dispatches to VM bytecode, to detect same-invocation aborts
* libguile/control.h:
* libguile/control.c (scm_c_make_prompt): Take an extra arg, a cookie.
Continuations will be rewindable only if the abort has the same cookie
as the prompt.
(scm_at_abort): Redefine from scm_abort, and instead of taking rest
args, take the abort values as a list directly. Also, don't allow
rewinding, because we won't support rewinding the C stack with
delimited continuations.
* libguile/eval.c (eval): Adapt to scm_c_make_prompt change.
* libguile/vm-engine.c (vm_engine): Use vp->cookie to get a unique value
corresponding to this VM invocation.
* libguile/vm-i-system.c (prompt): Pass the cookie to scm_c_make_prompt.
(abort): Take an additional tail arg.
* libguile/vm.c (vm_abort): Parse out the abort tail arg. This is for
the @abort case, or the (apply abort ...) case.
(make_vm): Initialize the cookie to 0.
* libguile/vm.h (struct scm_vm): Add cookie.
* module/ice-9/boot-9.scm (abort): Define here as a trampoline to
@abort. Needed to make sure that a call to abort dispatches to a VM
opcode, so the cookie will be the same.
* module/language/tree-il.scm (<tree-il>): Add a "tail" field to
<abort>, for the (apply abort ...) case, or (@abort tag args). Should
be #<const ()> in the normal case. Add support throughout.
* module/language/tree-il/analyze.scm (analyze-lexicals): Add abort-tail
support here too.
* module/language/tree-il/compile-glil.scm (flatten): Compile the tail
argument appropriately.
* module/language/tree-il/primitives.scm (*primitive-expand-table*): Fix
@abort and abort cases to pass the tail arg to make-abort.
2010-02-22 21:53:24 +01:00
|
|
|
|
2010-02-19 11:37:42 +01:00
|
|
|
SCM dynstate = SCM_I_CURRENT_THREAD->dynamic_state;
|
abort always dispatches to VM bytecode, to detect same-invocation aborts
* libguile/control.h:
* libguile/control.c (scm_c_make_prompt): Take an extra arg, a cookie.
Continuations will be rewindable only if the abort has the same cookie
as the prompt.
(scm_at_abort): Redefine from scm_abort, and instead of taking rest
args, take the abort values as a list directly. Also, don't allow
rewinding, because we won't support rewinding the C stack with
delimited continuations.
* libguile/eval.c (eval): Adapt to scm_c_make_prompt change.
* libguile/vm-engine.c (vm_engine): Use vp->cookie to get a unique value
corresponding to this VM invocation.
* libguile/vm-i-system.c (prompt): Pass the cookie to scm_c_make_prompt.
(abort): Take an additional tail arg.
* libguile/vm.c (vm_abort): Parse out the abort tail arg. This is for
the @abort case, or the (apply abort ...) case.
(make_vm): Initialize the cookie to 0.
* libguile/vm.h (struct scm_vm): Add cookie.
* module/ice-9/boot-9.scm (abort): Define here as a trampoline to
@abort. Needed to make sure that a call to abort dispatches to a VM
opcode, so the cookie will be the same.
* module/language/tree-il.scm (<tree-il>): Add a "tail" field to
<abort>, for the (apply abort ...) case, or (@abort tag args). Should
be #<const ()> in the normal case. Add support throughout.
* module/language/tree-il/analyze.scm (analyze-lexicals): Add abort-tail
support here too.
* module/language/tree-il/compile-glil.scm (flatten): Compile the tail
argument appropriately.
* module/language/tree-il/primitives.scm (*primitive-expand-table*): Fix
@abort and abort cases to pass the tail arg to make-abort.
2010-02-22 21:53:24 +01:00
|
|
|
scm_t_int64 vm_cookie = vp->cookie++;
|
2000-08-22 15:54:19 +00:00
|
|
|
|
2000-08-22 19:02:22 +00:00
|
|
|
/* Internal variables */
|
2008-09-15 00:04:34 +02:00
|
|
|
int nvalues = 0;
|
better error reporting from the vm
* libguile/vm-engine.c: Add func_name local, for error reporting.
(vm_error_apply_to_non_list): New error case.
(vm_error_wrong_type_arg): Remove this generic error case.
(vm_error_wrong_type_apply): Remove FUNC_NAME -- no sense in seeing
"vm-debug-engine" in the error report.
(vm_error_not_a_pair, vm_error_not_a_bytevector)
(vm_error_not_a_struct, vm_error_not_a_thunk): Use func_name instead
of FUNC_NAME, so we can indicate what caused the error.
* libguile/vm-i-scheme.c (VM_VALIDATE_CONS, car, cdr, set-car!)
(set-cdr!): Indicate provenance of errors.
(VM_VALIDATE_STRUCT, struct-vtable):
(VM_VALIDATE_BYTEVECTOR, BV_FIXABLE_INT_REF, BV_INT_REF)
(BV_FLOAT_REF, BV_FIXABLE_INT_SET, BV_INT_SET, BV_FLOAT_SET): Same.
* libguile/vm-i-system.c (apply, tail-apply): Use
vm_error_apply_to_non_list.
2010-07-15 12:44:15 +02:00
|
|
|
const char *func_name = NULL; /* used for error reporting */
|
2009-02-05 12:28:19 +01:00
|
|
|
SCM finish_args; /* used both for returns: both in error
|
|
|
|
|
and normal situations */
|
static opcodes; refactor program/objcode division; use new assembly pipeline
* gdbinit: Untested attempts to get the stack fondling macros to deal
with the new program representation.
* libguile/frames.c (scm_vm_frame_arguments, scm_vm_frame_source)
(scm_vm_frame_local_ref, scm_vm_frame_local_set_x): SCM_PROGRAM_DATA is
a struct scm_objcode*.
* libguile/instructions.h:
* libguile/instructions.c: Hide the instruction table and the struct
scm_instruction structure; all access to instructions now goes through
procedures. This is because instructions are no longer in a packed
array indexed by opcode. Also, declare a mask that all instructions
should fit in.
* libguile/objcodes.h:
* libguile/objcodes.c: Rewrite so that object code directly maps its
arity and length from its bytecode. This makes it unnecessary to keep
this information in programs, allowing programs to be simple conses
between the code (objcodes) and data (the object table and the closure
variables).
* libguile/programs.c (scm_make_program): Rework so that make-program
takes objcode, an object table, and externals as arguments. It's much
clearer this way, and we avoid malloc().
* libguile/stacks.c (is_vm_bootstrap_frame): Update for program/objcode
changes.
* libguile/vm-engine.c (vm_run): Initialize the jump table on the first
run, with the opcodes declared in the instruction sources, and with bad
instructions raising an error instead of wandering off into the
Unknown.
* libguile/vm-engine.h (FETCH_LENGTH): Always represent lengths as 3
bytes. The old code was too error-prone.
(NEXT_JUMP): Mask the instruction with SCM_VM_INSTRUCTION_MASK.
(NEW_FRAME): Update for program/objcode changes.
* libguile/vm-expand.h (VM_DEFINE_FUNCTION, VM_DEFINE_INSTRUCTION)
(VM_DEFINE_LOADER): Update so that we explicitly specify opcodes, so
that we have a stable bytecode API.
* libguile/vm-i-loader.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(load-integer): Use an int instead of a long as the accumulator; still
need to revisit this code at some point, I think.
(load-program): Simplify, thankfully!! Just creates the objcode slice
and rolls with it.
* libguile/vm-i-scheme.c: Number the opcodes explicitly.
* libguile/vm-i-system.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(make-closure): Update for new program API.
* libguile/vm.c (vm_make_boot_program): Update for new program/objcode
API. Still a bit ugly.
(scm_load_compiled_with_vm): Update for new program/objcode API.
* module/language/assembly.scm (byte-length): Fix byte-length calculation
for loaders, and load-program.
(code-pack, code-unpack): Start to move things from (system vm conv)
here.
(object->code, code->object): More things from conv.scm.
* module/language/glil.scm (<glil-program>): Add a new field,
closure-level.
(make-glil-program, compute-closure-level): Calculate the "closure
level" when making a glil program. This is the maximum depth of
external binding refs in this closure.
(unparse-glil): Fix label serialization.
* module/language/glil/compile-assembly.scm (make-meta): Prepend #f for
the meta's object table, though maybe in the future we can avoid
creating assembly in the first place.
(assoc-ref-or-acons, object-index-and-alist): GRRR! Caught again by the
different sets of arguments to assoc and assoc-ref!
(glil->assembly): Attempt to make the <glil-program> case more
readable, and fix the bugs. Sorry I don't know how to comment this
change any more than this.
(glil->assembly): For <glil-module> serialize the whole key, not just
the name.
(dump-object): subprogram-code is already a list. Serialize integers as
strings, not u8vectors. Fix the order of lists and vectors.
* module/language/glil/spec.scm (glil): Switch orders, so we prefer glil
-> assembly -> objcode. Actually glil->objcode doesn't work any more,
needs to be removed I think.
* module/language/objcode/spec.scm (objcode->value):
s/objcode->program/make-program/.
* module/language/scheme/inline.scm: Add acons inline.
* module/system/vm/conv.scm (make-byte-decoder): Skip the first 8 bytes,
they are header. Handle subprograms properly. Still needs help though.
(decode-length): Lengths are always 3 bytes now.
* module/system/vm/disasm.scm: Superficial changes to keep things
working. I'd like to fix this better in the future.
* module/system/vm/frame.scm (bootstrap-frame?): Fixes for
program-bytecode.
* module/system/vm/program.scm: Export make-program. It's program-objcode
now, no more program-bytecode.
* module/system/vm/vm.scm (vm-load): Use make-program.
* test-suite/tests/asm-to-bytecode.test: New test, very minimal.
* module/system/vm/objcode.scm: Export word-size, byte-order, and
write-objcode.
2009-01-29 21:09:04 +01:00
|
|
|
#ifdef HAVE_LABELS_AS_VALUES
|
|
|
|
|
static void **jump_table = NULL;
|
2009-02-05 12:28:19 +01:00
|
|
|
#endif
|
static opcodes; refactor program/objcode division; use new assembly pipeline
* gdbinit: Untested attempts to get the stack fondling macros to deal
with the new program representation.
* libguile/frames.c (scm_vm_frame_arguments, scm_vm_frame_source)
(scm_vm_frame_local_ref, scm_vm_frame_local_set_x): SCM_PROGRAM_DATA is
a struct scm_objcode*.
* libguile/instructions.h:
* libguile/instructions.c: Hide the instruction table and the struct
scm_instruction structure; all access to instructions now goes through
procedures. This is because instructions are no longer in a packed
array indexed by opcode. Also, declare a mask that all instructions
should fit in.
* libguile/objcodes.h:
* libguile/objcodes.c: Rewrite so that object code directly maps its
arity and length from its bytecode. This makes it unnecessary to keep
this information in programs, allowing programs to be simple conses
between the code (objcodes) and data (the object table and the closure
variables).
* libguile/programs.c (scm_make_program): Rework so that make-program
takes objcode, an object table, and externals as arguments. It's much
clearer this way, and we avoid malloc().
* libguile/stacks.c (is_vm_bootstrap_frame): Update for program/objcode
changes.
* libguile/vm-engine.c (vm_run): Initialize the jump table on the first
run, with the opcodes declared in the instruction sources, and with bad
instructions raising an error instead of wandering off into the
Unknown.
* libguile/vm-engine.h (FETCH_LENGTH): Always represent lengths as 3
bytes. The old code was too error-prone.
(NEXT_JUMP): Mask the instruction with SCM_VM_INSTRUCTION_MASK.
(NEW_FRAME): Update for program/objcode changes.
* libguile/vm-expand.h (VM_DEFINE_FUNCTION, VM_DEFINE_INSTRUCTION)
(VM_DEFINE_LOADER): Update so that we explicitly specify opcodes, so
that we have a stable bytecode API.
* libguile/vm-i-loader.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(load-integer): Use an int instead of a long as the accumulator; still
need to revisit this code at some point, I think.
(load-program): Simplify, thankfully!! Just creates the objcode slice
and rolls with it.
* libguile/vm-i-scheme.c: Number the opcodes explicitly.
* libguile/vm-i-system.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(make-closure): Update for new program API.
* libguile/vm.c (vm_make_boot_program): Update for new program/objcode
API. Still a bit ugly.
(scm_load_compiled_with_vm): Update for new program/objcode API.
* module/language/assembly.scm (byte-length): Fix byte-length calculation
for loaders, and load-program.
(code-pack, code-unpack): Start to move things from (system vm conv)
here.
(object->code, code->object): More things from conv.scm.
* module/language/glil.scm (<glil-program>): Add a new field,
closure-level.
(make-glil-program, compute-closure-level): Calculate the "closure
level" when making a glil program. This is the maximum depth of
external binding refs in this closure.
(unparse-glil): Fix label serialization.
* module/language/glil/compile-assembly.scm (make-meta): Prepend #f for
the meta's object table, though maybe in the future we can avoid
creating assembly in the first place.
(assoc-ref-or-acons, object-index-and-alist): GRRR! Caught again by the
different sets of arguments to assoc and assoc-ref!
(glil->assembly): Attempt to make the <glil-program> case more
readable, and fix the bugs. Sorry I don't know how to comment this
change any more than this.
(glil->assembly): For <glil-module> serialize the whole key, not just
the name.
(dump-object): subprogram-code is already a list. Serialize integers as
strings, not u8vectors. Fix the order of lists and vectors.
* module/language/glil/spec.scm (glil): Switch orders, so we prefer glil
-> assembly -> objcode. Actually glil->objcode doesn't work any more,
needs to be removed I think.
* module/language/objcode/spec.scm (objcode->value):
s/objcode->program/make-program/.
* module/language/scheme/inline.scm: Add acons inline.
* module/system/vm/conv.scm (make-byte-decoder): Skip the first 8 bytes,
they are header. Handle subprograms properly. Still needs help though.
(decode-length): Lengths are always 3 bytes now.
* module/system/vm/disasm.scm: Superficial changes to keep things
working. I'd like to fix this better in the future.
* module/system/vm/frame.scm (bootstrap-frame?): Fixes for
program-bytecode.
* module/system/vm/program.scm: Export make-program. It's program-objcode
now, no more program-bytecode.
* module/system/vm/vm.scm (vm-load): Use make-program.
* test-suite/tests/asm-to-bytecode.test: New test, very minimal.
* module/system/vm/objcode.scm: Export word-size, byte-order, and
write-objcode.
2009-01-29 21:09:04 +01:00
|
|
|
|
2009-02-05 12:28:19 +01:00
|
|
|
#ifdef HAVE_LABELS_AS_VALUES
|
static opcodes; refactor program/objcode division; use new assembly pipeline
* gdbinit: Untested attempts to get the stack fondling macros to deal
with the new program representation.
* libguile/frames.c (scm_vm_frame_arguments, scm_vm_frame_source)
(scm_vm_frame_local_ref, scm_vm_frame_local_set_x): SCM_PROGRAM_DATA is
a struct scm_objcode*.
* libguile/instructions.h:
* libguile/instructions.c: Hide the instruction table and the struct
scm_instruction structure; all access to instructions now goes through
procedures. This is because instructions are no longer in a packed
array indexed by opcode. Also, declare a mask that all instructions
should fit in.
* libguile/objcodes.h:
* libguile/objcodes.c: Rewrite so that object code directly maps its
arity and length from its bytecode. This makes it unnecessary to keep
this information in programs, allowing programs to be simple conses
between the code (objcodes) and data (the object table and the closure
variables).
* libguile/programs.c (scm_make_program): Rework so that make-program
takes objcode, an object table, and externals as arguments. It's much
clearer this way, and we avoid malloc().
* libguile/stacks.c (is_vm_bootstrap_frame): Update for program/objcode
changes.
* libguile/vm-engine.c (vm_run): Initialize the jump table on the first
run, with the opcodes declared in the instruction sources, and with bad
instructions raising an error instead of wandering off into the
Unknown.
* libguile/vm-engine.h (FETCH_LENGTH): Always represent lengths as 3
bytes. The old code was too error-prone.
(NEXT_JUMP): Mask the instruction with SCM_VM_INSTRUCTION_MASK.
(NEW_FRAME): Update for program/objcode changes.
* libguile/vm-expand.h (VM_DEFINE_FUNCTION, VM_DEFINE_INSTRUCTION)
(VM_DEFINE_LOADER): Update so that we explicitly specify opcodes, so
that we have a stable bytecode API.
* libguile/vm-i-loader.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(load-integer): Use an int instead of a long as the accumulator; still
need to revisit this code at some point, I think.
(load-program): Simplify, thankfully!! Just creates the objcode slice
and rolls with it.
* libguile/vm-i-scheme.c: Number the opcodes explicitly.
* libguile/vm-i-system.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(make-closure): Update for new program API.
* libguile/vm.c (vm_make_boot_program): Update for new program/objcode
API. Still a bit ugly.
(scm_load_compiled_with_vm): Update for new program/objcode API.
* module/language/assembly.scm (byte-length): Fix byte-length calculation
for loaders, and load-program.
(code-pack, code-unpack): Start to move things from (system vm conv)
here.
(object->code, code->object): More things from conv.scm.
* module/language/glil.scm (<glil-program>): Add a new field,
closure-level.
(make-glil-program, compute-closure-level): Calculate the "closure
level" when making a glil program. This is the maximum depth of
external binding refs in this closure.
(unparse-glil): Fix label serialization.
* module/language/glil/compile-assembly.scm (make-meta): Prepend #f for
the meta's object table, though maybe in the future we can avoid
creating assembly in the first place.
(assoc-ref-or-acons, object-index-and-alist): GRRR! Caught again by the
different sets of arguments to assoc and assoc-ref!
(glil->assembly): Attempt to make the <glil-program> case more
readable, and fix the bugs. Sorry I don't know how to comment this
change any more than this.
(glil->assembly): For <glil-module> serialize the whole key, not just
the name.
(dump-object): subprogram-code is already a list. Serialize integers as
strings, not u8vectors. Fix the order of lists and vectors.
* module/language/glil/spec.scm (glil): Switch orders, so we prefer glil
-> assembly -> objcode. Actually glil->objcode doesn't work any more,
needs to be removed I think.
* module/language/objcode/spec.scm (objcode->value):
s/objcode->program/make-program/.
* module/language/scheme/inline.scm: Add acons inline.
* module/system/vm/conv.scm (make-byte-decoder): Skip the first 8 bytes,
they are header. Handle subprograms properly. Still needs help though.
(decode-length): Lengths are always 3 bytes now.
* module/system/vm/disasm.scm: Superficial changes to keep things
working. I'd like to fix this better in the future.
* module/system/vm/frame.scm (bootstrap-frame?): Fixes for
program-bytecode.
* module/system/vm/program.scm: Export make-program. It's program-objcode
now, no more program-bytecode.
* module/system/vm/vm.scm (vm-load): Use make-program.
* test-suite/tests/asm-to-bytecode.test: New test, very minimal.
* module/system/vm/objcode.scm: Export word-size, byte-order, and
write-objcode.
2009-01-29 21:09:04 +01:00
|
|
|
if (SCM_UNLIKELY (!jump_table))
|
|
|
|
|
{
|
|
|
|
|
int i;
|
2009-02-03 22:36:02 +01:00
|
|
|
jump_table = malloc (SCM_VM_NUM_INSTRUCTIONS * sizeof(void*));
|
static opcodes; refactor program/objcode division; use new assembly pipeline
* gdbinit: Untested attempts to get the stack fondling macros to deal
with the new program representation.
* libguile/frames.c (scm_vm_frame_arguments, scm_vm_frame_source)
(scm_vm_frame_local_ref, scm_vm_frame_local_set_x): SCM_PROGRAM_DATA is
a struct scm_objcode*.
* libguile/instructions.h:
* libguile/instructions.c: Hide the instruction table and the struct
scm_instruction structure; all access to instructions now goes through
procedures. This is because instructions are no longer in a packed
array indexed by opcode. Also, declare a mask that all instructions
should fit in.
* libguile/objcodes.h:
* libguile/objcodes.c: Rewrite so that object code directly maps its
arity and length from its bytecode. This makes it unnecessary to keep
this information in programs, allowing programs to be simple conses
between the code (objcodes) and data (the object table and the closure
variables).
* libguile/programs.c (scm_make_program): Rework so that make-program
takes objcode, an object table, and externals as arguments. It's much
clearer this way, and we avoid malloc().
* libguile/stacks.c (is_vm_bootstrap_frame): Update for program/objcode
changes.
* libguile/vm-engine.c (vm_run): Initialize the jump table on the first
run, with the opcodes declared in the instruction sources, and with bad
instructions raising an error instead of wandering off into the
Unknown.
* libguile/vm-engine.h (FETCH_LENGTH): Always represent lengths as 3
bytes. The old code was too error-prone.
(NEXT_JUMP): Mask the instruction with SCM_VM_INSTRUCTION_MASK.
(NEW_FRAME): Update for program/objcode changes.
* libguile/vm-expand.h (VM_DEFINE_FUNCTION, VM_DEFINE_INSTRUCTION)
(VM_DEFINE_LOADER): Update so that we explicitly specify opcodes, so
that we have a stable bytecode API.
* libguile/vm-i-loader.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(load-integer): Use an int instead of a long as the accumulator; still
need to revisit this code at some point, I think.
(load-program): Simplify, thankfully!! Just creates the objcode slice
and rolls with it.
* libguile/vm-i-scheme.c: Number the opcodes explicitly.
* libguile/vm-i-system.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(make-closure): Update for new program API.
* libguile/vm.c (vm_make_boot_program): Update for new program/objcode
API. Still a bit ugly.
(scm_load_compiled_with_vm): Update for new program/objcode API.
* module/language/assembly.scm (byte-length): Fix byte-length calculation
for loaders, and load-program.
(code-pack, code-unpack): Start to move things from (system vm conv)
here.
(object->code, code->object): More things from conv.scm.
* module/language/glil.scm (<glil-program>): Add a new field,
closure-level.
(make-glil-program, compute-closure-level): Calculate the "closure
level" when making a glil program. This is the maximum depth of
external binding refs in this closure.
(unparse-glil): Fix label serialization.
* module/language/glil/compile-assembly.scm (make-meta): Prepend #f for
the meta's object table, though maybe in the future we can avoid
creating assembly in the first place.
(assoc-ref-or-acons, object-index-and-alist): GRRR! Caught again by the
different sets of arguments to assoc and assoc-ref!
(glil->assembly): Attempt to make the <glil-program> case more
readable, and fix the bugs. Sorry I don't know how to comment this
change any more than this.
(glil->assembly): For <glil-module> serialize the whole key, not just
the name.
(dump-object): subprogram-code is already a list. Serialize integers as
strings, not u8vectors. Fix the order of lists and vectors.
* module/language/glil/spec.scm (glil): Switch orders, so we prefer glil
-> assembly -> objcode. Actually glil->objcode doesn't work any more,
needs to be removed I think.
* module/language/objcode/spec.scm (objcode->value):
s/objcode->program/make-program/.
* module/language/scheme/inline.scm: Add acons inline.
* module/system/vm/conv.scm (make-byte-decoder): Skip the first 8 bytes,
they are header. Handle subprograms properly. Still needs help though.
(decode-length): Lengths are always 3 bytes now.
* module/system/vm/disasm.scm: Superficial changes to keep things
working. I'd like to fix this better in the future.
* module/system/vm/frame.scm (bootstrap-frame?): Fixes for
program-bytecode.
* module/system/vm/program.scm: Export make-program. It's program-objcode
now, no more program-bytecode.
* module/system/vm/vm.scm (vm-load): Use make-program.
* test-suite/tests/asm-to-bytecode.test: New test, very minimal.
* module/system/vm/objcode.scm: Export word-size, byte-order, and
write-objcode.
2009-01-29 21:09:04 +01:00
|
|
|
for (i = 0; i < SCM_VM_NUM_INSTRUCTIONS; i++)
|
|
|
|
|
jump_table[i] = &&vm_error_bad_instruction;
|
|
|
|
|
#define VM_INSTRUCTION_TO_LABEL 1
|
2009-03-16 23:39:29 +01:00
|
|
|
#include <libguile/vm-expand.h>
|
|
|
|
|
#include <libguile/vm-i-system.i>
|
|
|
|
|
#include <libguile/vm-i-scheme.i>
|
|
|
|
|
#include <libguile/vm-i-loader.i>
|
static opcodes; refactor program/objcode division; use new assembly pipeline
* gdbinit: Untested attempts to get the stack fondling macros to deal
with the new program representation.
* libguile/frames.c (scm_vm_frame_arguments, scm_vm_frame_source)
(scm_vm_frame_local_ref, scm_vm_frame_local_set_x): SCM_PROGRAM_DATA is
a struct scm_objcode*.
* libguile/instructions.h:
* libguile/instructions.c: Hide the instruction table and the struct
scm_instruction structure; all access to instructions now goes through
procedures. This is because instructions are no longer in a packed
array indexed by opcode. Also, declare a mask that all instructions
should fit in.
* libguile/objcodes.h:
* libguile/objcodes.c: Rewrite so that object code directly maps its
arity and length from its bytecode. This makes it unnecessary to keep
this information in programs, allowing programs to be simple conses
between the code (objcodes) and data (the object table and the closure
variables).
* libguile/programs.c (scm_make_program): Rework so that make-program
takes objcode, an object table, and externals as arguments. It's much
clearer this way, and we avoid malloc().
* libguile/stacks.c (is_vm_bootstrap_frame): Update for program/objcode
changes.
* libguile/vm-engine.c (vm_run): Initialize the jump table on the first
run, with the opcodes declared in the instruction sources, and with bad
instructions raising an error instead of wandering off into the
Unknown.
* libguile/vm-engine.h (FETCH_LENGTH): Always represent lengths as 3
bytes. The old code was too error-prone.
(NEXT_JUMP): Mask the instruction with SCM_VM_INSTRUCTION_MASK.
(NEW_FRAME): Update for program/objcode changes.
* libguile/vm-expand.h (VM_DEFINE_FUNCTION, VM_DEFINE_INSTRUCTION)
(VM_DEFINE_LOADER): Update so that we explicitly specify opcodes, so
that we have a stable bytecode API.
* libguile/vm-i-loader.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(load-integer): Use an int instead of a long as the accumulator; still
need to revisit this code at some point, I think.
(load-program): Simplify, thankfully!! Just creates the objcode slice
and rolls with it.
* libguile/vm-i-scheme.c: Number the opcodes explicitly.
* libguile/vm-i-system.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(make-closure): Update for new program API.
* libguile/vm.c (vm_make_boot_program): Update for new program/objcode
API. Still a bit ugly.
(scm_load_compiled_with_vm): Update for new program/objcode API.
* module/language/assembly.scm (byte-length): Fix byte-length calculation
for loaders, and load-program.
(code-pack, code-unpack): Start to move things from (system vm conv)
here.
(object->code, code->object): More things from conv.scm.
* module/language/glil.scm (<glil-program>): Add a new field,
closure-level.
(make-glil-program, compute-closure-level): Calculate the "closure
level" when making a glil program. This is the maximum depth of
external binding refs in this closure.
(unparse-glil): Fix label serialization.
* module/language/glil/compile-assembly.scm (make-meta): Prepend #f for
the meta's object table, though maybe in the future we can avoid
creating assembly in the first place.
(assoc-ref-or-acons, object-index-and-alist): GRRR! Caught again by the
different sets of arguments to assoc and assoc-ref!
(glil->assembly): Attempt to make the <glil-program> case more
readable, and fix the bugs. Sorry I don't know how to comment this
change any more than this.
(glil->assembly): For <glil-module> serialize the whole key, not just
the name.
(dump-object): subprogram-code is already a list. Serialize integers as
strings, not u8vectors. Fix the order of lists and vectors.
* module/language/glil/spec.scm (glil): Switch orders, so we prefer glil
-> assembly -> objcode. Actually glil->objcode doesn't work any more,
needs to be removed I think.
* module/language/objcode/spec.scm (objcode->value):
s/objcode->program/make-program/.
* module/language/scheme/inline.scm: Add acons inline.
* module/system/vm/conv.scm (make-byte-decoder): Skip the first 8 bytes,
they are header. Handle subprograms properly. Still needs help though.
(decode-length): Lengths are always 3 bytes now.
* module/system/vm/disasm.scm: Superficial changes to keep things
working. I'd like to fix this better in the future.
* module/system/vm/frame.scm (bootstrap-frame?): Fixes for
program-bytecode.
* module/system/vm/program.scm: Export make-program. It's program-objcode
now, no more program-bytecode.
* module/system/vm/vm.scm (vm-load): Use make-program.
* test-suite/tests/asm-to-bytecode.test: New test, very minimal.
* module/system/vm/objcode.scm: Export word-size, byte-order, and
write-objcode.
2009-01-29 21:09:04 +01:00
|
|
|
#undef VM_INSTRUCTION_TO_LABEL
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
2001-04-06 00:17:39 +00:00
|
|
|
/* Initialization */
|
|
|
|
|
{
|
2001-04-06 05:00:10 +00:00
|
|
|
SCM prog = program;
|
|
|
|
|
|
|
|
|
|
/* Boot program */
|
enable multiple vm engines (regular, debug, ...)
* libguile/vm-engine.c (VM_USE_HOOKS, VM_USE_CLOCK, VM_CHECK_EXTERNAL)
(VM_CHECK_OBJECT): Update to define these here, before including
vm-engine.h.
(vm_run): Change so that we can make different engines. Also, we take
an array of arguments, and the struct scm_vm directly, so as to avoid
any need to cons.
* libguile/vm-engine.h (CHECK_EXTERNAL, CHECK_OBJECT): Add some UNLIKELY
bits; don't seem to help.
* libguile/vm.c (vm_dispatch_hook): Change to not pass the VP. This needs
some love, and perhaps we revert to the old way.
(VM_ENGINE): Actually make two engines, vm_regular_engine and
vm_debug_engine. Probably there is room for improvement here. Actually
their speeds are the same at the moment.
(make_vm): Choose which engine to run; currently the debug engine by
default.
(scm_c_vm_run): A thin wrapper to invoke a VM without consing.
(scm_vm_apply): Use scm_c_vm_run.
(scm_load_compiled_with_vm): Use scm_c_vm_run.
2009-02-04 23:47:56 +01:00
|
|
|
program = vm_make_boot_program (nargs);
|
2000-08-22 15:54:19 +00:00
|
|
|
|
2001-04-06 00:17:39 +00:00
|
|
|
/* Initial frame */
|
|
|
|
|
CACHE_REGISTER ();
|
2009-08-15 14:13:08 +02:00
|
|
|
PUSH ((SCM)fp); /* dynamic link */
|
|
|
|
|
PUSH (0); /* mvra */
|
runtime and debugging support for callee-parsed procedure args
* libguile/objcodes.h: Bump for metadata format change.
* libguile/frames.h: Rework so we don't frob the program's nargs, nlocs,
etc at runtime. Instead we don't really know what's a local var, an
argument, or an intermediate value. It's a little unfortunate, but
this will allow for case-lambda, and eventually for good polymorphic
generic dispatch; and the nlocs etc can be heuristically
reconstructed. Such a reconstruction would be better done at the
Scheme level, though.
(SCM_FRAME_STACK_ADDRESS): New macro, the pointer to the base of the
stack elements (not counting the program).
(SCM_FRAME_UPPER_ADDRESS): Repurpose to be the address of the last
element in the bookkeeping part of the stack -- i.e. to point to the
return address.
* libguile/vm-engine.h:
* libguile/vm-i-system.c: Adapt to removal of stack_base. Though we
still detect stack-smashing underflow, we don't do so as precisely as
we did before, because now we only detect overwriting of the frame
metadata.
* libguile/vm-engine.c (vm_engine): Remove the stack_base variable. It
is unnecessary, and difficult to keep track of in the face of
case-lambda. Also fix miscommented "ra" and "mvra" pushes. Push the
vp->ip as the first ra...
* libguile/vm-i-system.c (halt): ...because here we can restore the
vp->ip instead of setting ip to 0. Allows us to introspect ips all
down the stack, including in recursive VM invocations.
* libguile/frames.h:
* libguile/frames.c (scm_vm_frame_stack): Removed, because it's getting
more difficult to tell what's an argument and what's a temporary stack
element.
(scm_vm_frame_num_locals): New accessor.
(scm_vm_frame_instruction_pointer): New accessor.
(scm_vm_frame_arguments): Defer to an implementation in Scheme.
(scm_vm_frame_num_locals scm_vm_frame_local_ref)
(scm_vm_frame_local_set_x): Since we can get not-yet-active frames on
the stack now, with our current calling convention, we have to add a
heuristic here to jump over those frames -- because frames have
pointers in them, not Scheme values.
* libguile/programs.h:
* libguile/programs.c (scm_program_arity): Remove, in favor of..
(scm_program_arities): ...this, which a list of arities, in a new
format, occupying a slot in the metadata.
* module/language/assembly/decompile-bytecode.scm (decode-load-program):
Fix mv-call decompilation.
* module/system/vm/frame.scm (vm-frame-bindings, vm-frame-binding-ref)
(vm-frame-binding-set!): New functions, to access bindings by name in
a frame.
(vm-frame-arguments): Function now implemented in Scheme. Commented
fairly extensively.
* module/system/vm/program.scm (program-bindings-by-index)
(program-bindings-for-ip): New accessors, parsing the program bindings
metadata into something more useful.
(program-arities, program-arguments): In a case-lambda world, we have
to assume that programs can have multiple arities. But it's tough to
detect this algorithmically; instead we're going to require that the
program metadata include information about the arities, and the parts
of the program that that metadata applies to.
(program-lambda-list): New accessor.
(write-program): Show multiple arities.
* module/language/glil/compile-assembly.scm (glil->assembly): Add
"arities" to the state of the compiler, and add arities entries as
appropriate.
2009-09-27 19:25:58 -04:00
|
|
|
PUSH ((SCM)ip); /* ra */
|
2001-04-06 05:00:10 +00:00
|
|
|
CACHE_PROGRAM ();
|
2001-04-07 09:39:38 +00:00
|
|
|
PUSH (program);
|
2009-08-15 14:13:08 +02:00
|
|
|
fp = sp + 1;
|
2009-12-14 23:11:47 +01:00
|
|
|
ip = SCM_C_OBJCODE_BASE (bp);
|
2009-08-15 14:15:56 +02:00
|
|
|
/* MV-call frame, function & arguments */
|
2011-03-15 23:33:32 +01:00
|
|
|
PUSH (0); /* dynamic link */
|
2009-08-15 14:15:56 +02:00
|
|
|
PUSH (0); /* mvra */
|
runtime and debugging support for callee-parsed procedure args
* libguile/objcodes.h: Bump for metadata format change.
* libguile/frames.h: Rework so we don't frob the program's nargs, nlocs,
etc at runtime. Instead we don't really know what's a local var, an
argument, or an intermediate value. It's a little unfortunate, but
this will allow for case-lambda, and eventually for good polymorphic
generic dispatch; and the nlocs etc can be heuristically
reconstructed. Such a reconstruction would be better done at the
Scheme level, though.
(SCM_FRAME_STACK_ADDRESS): New macro, the pointer to the base of the
stack elements (not counting the program).
(SCM_FRAME_UPPER_ADDRESS): Repurpose to be the address of the last
element in the bookkeeping part of the stack -- i.e. to point to the
return address.
* libguile/vm-engine.h:
* libguile/vm-i-system.c: Adapt to removal of stack_base. Though we
still detect stack-smashing underflow, we don't do so as precisely as
we did before, because now we only detect overwriting of the frame
metadata.
* libguile/vm-engine.c (vm_engine): Remove the stack_base variable. It
is unnecessary, and difficult to keep track of in the face of
case-lambda. Also fix miscommented "ra" and "mvra" pushes. Push the
vp->ip as the first ra...
* libguile/vm-i-system.c (halt): ...because here we can restore the
vp->ip instead of setting ip to 0. Allows us to introspect ips all
down the stack, including in recursive VM invocations.
* libguile/frames.h:
* libguile/frames.c (scm_vm_frame_stack): Removed, because it's getting
more difficult to tell what's an argument and what's a temporary stack
element.
(scm_vm_frame_num_locals): New accessor.
(scm_vm_frame_instruction_pointer): New accessor.
(scm_vm_frame_arguments): Defer to an implementation in Scheme.
(scm_vm_frame_num_locals scm_vm_frame_local_ref)
(scm_vm_frame_local_set_x): Since we can get not-yet-active frames on
the stack now, with our current calling convention, we have to add a
heuristic here to jump over those frames -- because frames have
pointers in them, not Scheme values.
* libguile/programs.h:
* libguile/programs.c (scm_program_arity): Remove, in favor of..
(scm_program_arities): ...this, which a list of arities, in a new
format, occupying a slot in the metadata.
* module/language/assembly/decompile-bytecode.scm (decode-load-program):
Fix mv-call decompilation.
* module/system/vm/frame.scm (vm-frame-bindings, vm-frame-binding-ref)
(vm-frame-binding-set!): New functions, to access bindings by name in
a frame.
(vm-frame-arguments): Function now implemented in Scheme. Commented
fairly extensively.
* module/system/vm/program.scm (program-bindings-by-index)
(program-bindings-for-ip): New accessors, parsing the program bindings
metadata into something more useful.
(program-arities, program-arguments): In a case-lambda world, we have
to assume that programs can have multiple arities. But it's tough to
detect this algorithmically; instead we're going to require that the
program metadata include information about the arities, and the parts
of the program that that metadata applies to.
(program-lambda-list): New accessor.
(write-program): Show multiple arities.
* module/language/glil/compile-assembly.scm (glil->assembly): Add
"arities" to the state of the compiler, and add arities entries as
appropriate.
2009-09-27 19:25:58 -04:00
|
|
|
PUSH (0); /* ra */
|
2001-04-07 09:39:38 +00:00
|
|
|
PUSH (prog);
|
enable multiple vm engines (regular, debug, ...)
* libguile/vm-engine.c (VM_USE_HOOKS, VM_USE_CLOCK, VM_CHECK_EXTERNAL)
(VM_CHECK_OBJECT): Update to define these here, before including
vm-engine.h.
(vm_run): Change so that we can make different engines. Also, we take
an array of arguments, and the struct scm_vm directly, so as to avoid
any need to cons.
* libguile/vm-engine.h (CHECK_EXTERNAL, CHECK_OBJECT): Add some UNLIKELY
bits; don't seem to help.
* libguile/vm.c (vm_dispatch_hook): Change to not pass the VP. This needs
some love, and perhaps we revert to the old way.
(VM_ENGINE): Actually make two engines, vm_regular_engine and
vm_debug_engine. Probably there is room for improvement here. Actually
their speeds are the same at the moment.
(make_vm): Choose which engine to run; currently the debug engine by
default.
(scm_c_vm_run): A thin wrapper to invoke a VM without consing.
(scm_vm_apply): Use scm_c_vm_run.
(scm_load_compiled_with_vm): Use scm_c_vm_run.
2009-02-04 23:47:56 +01:00
|
|
|
if (SCM_UNLIKELY (sp + nargs >= stack_limit))
|
|
|
|
|
goto vm_error_too_many_args;
|
|
|
|
|
while (nargs--)
|
|
|
|
|
PUSH (*argv++);
|
2001-04-06 00:17:39 +00:00
|
|
|
}
|
2000-08-22 15:54:19 +00:00
|
|
|
|
|
|
|
|
/* Let's go! */
|
static opcodes; refactor program/objcode division; use new assembly pipeline
* gdbinit: Untested attempts to get the stack fondling macros to deal
with the new program representation.
* libguile/frames.c (scm_vm_frame_arguments, scm_vm_frame_source)
(scm_vm_frame_local_ref, scm_vm_frame_local_set_x): SCM_PROGRAM_DATA is
a struct scm_objcode*.
* libguile/instructions.h:
* libguile/instructions.c: Hide the instruction table and the struct
scm_instruction structure; all access to instructions now goes through
procedures. This is because instructions are no longer in a packed
array indexed by opcode. Also, declare a mask that all instructions
should fit in.
* libguile/objcodes.h:
* libguile/objcodes.c: Rewrite so that object code directly maps its
arity and length from its bytecode. This makes it unnecessary to keep
this information in programs, allowing programs to be simple conses
between the code (objcodes) and data (the object table and the closure
variables).
* libguile/programs.c (scm_make_program): Rework so that make-program
takes objcode, an object table, and externals as arguments. It's much
clearer this way, and we avoid malloc().
* libguile/stacks.c (is_vm_bootstrap_frame): Update for program/objcode
changes.
* libguile/vm-engine.c (vm_run): Initialize the jump table on the first
run, with the opcodes declared in the instruction sources, and with bad
instructions raising an error instead of wandering off into the
Unknown.
* libguile/vm-engine.h (FETCH_LENGTH): Always represent lengths as 3
bytes. The old code was too error-prone.
(NEXT_JUMP): Mask the instruction with SCM_VM_INSTRUCTION_MASK.
(NEW_FRAME): Update for program/objcode changes.
* libguile/vm-expand.h (VM_DEFINE_FUNCTION, VM_DEFINE_INSTRUCTION)
(VM_DEFINE_LOADER): Update so that we explicitly specify opcodes, so
that we have a stable bytecode API.
* libguile/vm-i-loader.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(load-integer): Use an int instead of a long as the accumulator; still
need to revisit this code at some point, I think.
(load-program): Simplify, thankfully!! Just creates the objcode slice
and rolls with it.
* libguile/vm-i-scheme.c: Number the opcodes explicitly.
* libguile/vm-i-system.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(make-closure): Update for new program API.
* libguile/vm.c (vm_make_boot_program): Update for new program/objcode
API. Still a bit ugly.
(scm_load_compiled_with_vm): Update for new program/objcode API.
* module/language/assembly.scm (byte-length): Fix byte-length calculation
for loaders, and load-program.
(code-pack, code-unpack): Start to move things from (system vm conv)
here.
(object->code, code->object): More things from conv.scm.
* module/language/glil.scm (<glil-program>): Add a new field,
closure-level.
(make-glil-program, compute-closure-level): Calculate the "closure
level" when making a glil program. This is the maximum depth of
external binding refs in this closure.
(unparse-glil): Fix label serialization.
* module/language/glil/compile-assembly.scm (make-meta): Prepend #f for
the meta's object table, though maybe in the future we can avoid
creating assembly in the first place.
(assoc-ref-or-acons, object-index-and-alist): GRRR! Caught again by the
different sets of arguments to assoc and assoc-ref!
(glil->assembly): Attempt to make the <glil-program> case more
readable, and fix the bugs. Sorry I don't know how to comment this
change any more than this.
(glil->assembly): For <glil-module> serialize the whole key, not just
the name.
(dump-object): subprogram-code is already a list. Serialize integers as
strings, not u8vectors. Fix the order of lists and vectors.
* module/language/glil/spec.scm (glil): Switch orders, so we prefer glil
-> assembly -> objcode. Actually glil->objcode doesn't work any more,
needs to be removed I think.
* module/language/objcode/spec.scm (objcode->value):
s/objcode->program/make-program/.
* module/language/scheme/inline.scm: Add acons inline.
* module/system/vm/conv.scm (make-byte-decoder): Skip the first 8 bytes,
they are header. Handle subprograms properly. Still needs help though.
(decode-length): Lengths are always 3 bytes now.
* module/system/vm/disasm.scm: Superficial changes to keep things
working. I'd like to fix this better in the future.
* module/system/vm/frame.scm (bootstrap-frame?): Fixes for
program-bytecode.
* module/system/vm/program.scm: Export make-program. It's program-objcode
now, no more program-bytecode.
* module/system/vm/vm.scm (vm-load): Use make-program.
* test-suite/tests/asm-to-bytecode.test: New test, very minimal.
* module/system/vm/objcode.scm: Export word-size, byte-order, and
write-objcode.
2009-01-29 21:09:04 +01:00
|
|
|
NEXT;
|
2000-08-22 15:54:19 +00:00
|
|
|
|
|
|
|
|
#ifndef HAVE_LABELS_AS_VALUES
|
2001-04-01 05:03:41 +00:00
|
|
|
vm_start:
|
static opcodes; refactor program/objcode division; use new assembly pipeline
* gdbinit: Untested attempts to get the stack fondling macros to deal
with the new program representation.
* libguile/frames.c (scm_vm_frame_arguments, scm_vm_frame_source)
(scm_vm_frame_local_ref, scm_vm_frame_local_set_x): SCM_PROGRAM_DATA is
a struct scm_objcode*.
* libguile/instructions.h:
* libguile/instructions.c: Hide the instruction table and the struct
scm_instruction structure; all access to instructions now goes through
procedures. This is because instructions are no longer in a packed
array indexed by opcode. Also, declare a mask that all instructions
should fit in.
* libguile/objcodes.h:
* libguile/objcodes.c: Rewrite so that object code directly maps its
arity and length from its bytecode. This makes it unnecessary to keep
this information in programs, allowing programs to be simple conses
between the code (objcodes) and data (the object table and the closure
variables).
* libguile/programs.c (scm_make_program): Rework so that make-program
takes objcode, an object table, and externals as arguments. It's much
clearer this way, and we avoid malloc().
* libguile/stacks.c (is_vm_bootstrap_frame): Update for program/objcode
changes.
* libguile/vm-engine.c (vm_run): Initialize the jump table on the first
run, with the opcodes declared in the instruction sources, and with bad
instructions raising an error instead of wandering off into the
Unknown.
* libguile/vm-engine.h (FETCH_LENGTH): Always represent lengths as 3
bytes. The old code was too error-prone.
(NEXT_JUMP): Mask the instruction with SCM_VM_INSTRUCTION_MASK.
(NEW_FRAME): Update for program/objcode changes.
* libguile/vm-expand.h (VM_DEFINE_FUNCTION, VM_DEFINE_INSTRUCTION)
(VM_DEFINE_LOADER): Update so that we explicitly specify opcodes, so
that we have a stable bytecode API.
* libguile/vm-i-loader.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(load-integer): Use an int instead of a long as the accumulator; still
need to revisit this code at some point, I think.
(load-program): Simplify, thankfully!! Just creates the objcode slice
and rolls with it.
* libguile/vm-i-scheme.c: Number the opcodes explicitly.
* libguile/vm-i-system.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(make-closure): Update for new program API.
* libguile/vm.c (vm_make_boot_program): Update for new program/objcode
API. Still a bit ugly.
(scm_load_compiled_with_vm): Update for new program/objcode API.
* module/language/assembly.scm (byte-length): Fix byte-length calculation
for loaders, and load-program.
(code-pack, code-unpack): Start to move things from (system vm conv)
here.
(object->code, code->object): More things from conv.scm.
* module/language/glil.scm (<glil-program>): Add a new field,
closure-level.
(make-glil-program, compute-closure-level): Calculate the "closure
level" when making a glil program. This is the maximum depth of
external binding refs in this closure.
(unparse-glil): Fix label serialization.
* module/language/glil/compile-assembly.scm (make-meta): Prepend #f for
the meta's object table, though maybe in the future we can avoid
creating assembly in the first place.
(assoc-ref-or-acons, object-index-and-alist): GRRR! Caught again by the
different sets of arguments to assoc and assoc-ref!
(glil->assembly): Attempt to make the <glil-program> case more
readable, and fix the bugs. Sorry I don't know how to comment this
change any more than this.
(glil->assembly): For <glil-module> serialize the whole key, not just
the name.
(dump-object): subprogram-code is already a list. Serialize integers as
strings, not u8vectors. Fix the order of lists and vectors.
* module/language/glil/spec.scm (glil): Switch orders, so we prefer glil
-> assembly -> objcode. Actually glil->objcode doesn't work any more,
needs to be removed I think.
* module/language/objcode/spec.scm (objcode->value):
s/objcode->program/make-program/.
* module/language/scheme/inline.scm: Add acons inline.
* module/system/vm/conv.scm (make-byte-decoder): Skip the first 8 bytes,
they are header. Handle subprograms properly. Still needs help though.
(decode-length): Lengths are always 3 bytes now.
* module/system/vm/disasm.scm: Superficial changes to keep things
working. I'd like to fix this better in the future.
* module/system/vm/frame.scm (bootstrap-frame?): Fixes for
program-bytecode.
* module/system/vm/program.scm: Export make-program. It's program-objcode
now, no more program-bytecode.
* module/system/vm/vm.scm (vm-load): Use make-program.
* test-suite/tests/asm-to-bytecode.test: New test, very minimal.
* module/system/vm/objcode.scm: Export word-size, byte-order, and
write-objcode.
2009-01-29 21:09:04 +01:00
|
|
|
switch ((*ip++) & SCM_VM_INSTRUCTION_MASK) {
|
2000-08-22 15:54:19 +00:00
|
|
|
#endif
|
|
|
|
|
|
2008-08-21 18:39:30 -07:00
|
|
|
#include "vm-expand.h"
|
|
|
|
|
#include "vm-i-system.c"
|
|
|
|
|
#include "vm-i-scheme.c"
|
|
|
|
|
#include "vm-i-loader.c"
|
2000-08-22 15:54:19 +00:00
|
|
|
|
|
|
|
|
#ifndef HAVE_LABELS_AS_VALUES
|
static opcodes; refactor program/objcode division; use new assembly pipeline
* gdbinit: Untested attempts to get the stack fondling macros to deal
with the new program representation.
* libguile/frames.c (scm_vm_frame_arguments, scm_vm_frame_source)
(scm_vm_frame_local_ref, scm_vm_frame_local_set_x): SCM_PROGRAM_DATA is
a struct scm_objcode*.
* libguile/instructions.h:
* libguile/instructions.c: Hide the instruction table and the struct
scm_instruction structure; all access to instructions now goes through
procedures. This is because instructions are no longer in a packed
array indexed by opcode. Also, declare a mask that all instructions
should fit in.
* libguile/objcodes.h:
* libguile/objcodes.c: Rewrite so that object code directly maps its
arity and length from its bytecode. This makes it unnecessary to keep
this information in programs, allowing programs to be simple conses
between the code (objcodes) and data (the object table and the closure
variables).
* libguile/programs.c (scm_make_program): Rework so that make-program
takes objcode, an object table, and externals as arguments. It's much
clearer this way, and we avoid malloc().
* libguile/stacks.c (is_vm_bootstrap_frame): Update for program/objcode
changes.
* libguile/vm-engine.c (vm_run): Initialize the jump table on the first
run, with the opcodes declared in the instruction sources, and with bad
instructions raising an error instead of wandering off into the
Unknown.
* libguile/vm-engine.h (FETCH_LENGTH): Always represent lengths as 3
bytes. The old code was too error-prone.
(NEXT_JUMP): Mask the instruction with SCM_VM_INSTRUCTION_MASK.
(NEW_FRAME): Update for program/objcode changes.
* libguile/vm-expand.h (VM_DEFINE_FUNCTION, VM_DEFINE_INSTRUCTION)
(VM_DEFINE_LOADER): Update so that we explicitly specify opcodes, so
that we have a stable bytecode API.
* libguile/vm-i-loader.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(load-integer): Use an int instead of a long as the accumulator; still
need to revisit this code at some point, I think.
(load-program): Simplify, thankfully!! Just creates the objcode slice
and rolls with it.
* libguile/vm-i-scheme.c: Number the opcodes explicitly.
* libguile/vm-i-system.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(make-closure): Update for new program API.
* libguile/vm.c (vm_make_boot_program): Update for new program/objcode
API. Still a bit ugly.
(scm_load_compiled_with_vm): Update for new program/objcode API.
* module/language/assembly.scm (byte-length): Fix byte-length calculation
for loaders, and load-program.
(code-pack, code-unpack): Start to move things from (system vm conv)
here.
(object->code, code->object): More things from conv.scm.
* module/language/glil.scm (<glil-program>): Add a new field,
closure-level.
(make-glil-program, compute-closure-level): Calculate the "closure
level" when making a glil program. This is the maximum depth of
external binding refs in this closure.
(unparse-glil): Fix label serialization.
* module/language/glil/compile-assembly.scm (make-meta): Prepend #f for
the meta's object table, though maybe in the future we can avoid
creating assembly in the first place.
(assoc-ref-or-acons, object-index-and-alist): GRRR! Caught again by the
different sets of arguments to assoc and assoc-ref!
(glil->assembly): Attempt to make the <glil-program> case more
readable, and fix the bugs. Sorry I don't know how to comment this
change any more than this.
(glil->assembly): For <glil-module> serialize the whole key, not just
the name.
(dump-object): subprogram-code is already a list. Serialize integers as
strings, not u8vectors. Fix the order of lists and vectors.
* module/language/glil/spec.scm (glil): Switch orders, so we prefer glil
-> assembly -> objcode. Actually glil->objcode doesn't work any more,
needs to be removed I think.
* module/language/objcode/spec.scm (objcode->value):
s/objcode->program/make-program/.
* module/language/scheme/inline.scm: Add acons inline.
* module/system/vm/conv.scm (make-byte-decoder): Skip the first 8 bytes,
they are header. Handle subprograms properly. Still needs help though.
(decode-length): Lengths are always 3 bytes now.
* module/system/vm/disasm.scm: Superficial changes to keep things
working. I'd like to fix this better in the future.
* module/system/vm/frame.scm (bootstrap-frame?): Fixes for
program-bytecode.
* module/system/vm/program.scm: Export make-program. It's program-objcode
now, no more program-bytecode.
* module/system/vm/vm.scm (vm-load): Use make-program.
* test-suite/tests/asm-to-bytecode.test: New test, very minimal.
* module/system/vm/objcode.scm: Export word-size, byte-order, and
write-objcode.
2009-01-29 21:09:04 +01:00
|
|
|
default:
|
|
|
|
|
goto vm_error_bad_instruction;
|
2000-08-22 15:54:19 +00:00
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
2009-02-05 12:28:19 +01:00
|
|
|
|
|
|
|
|
vm_done:
|
|
|
|
|
SYNC_ALL ();
|
|
|
|
|
return finish_args;
|
|
|
|
|
|
2001-04-01 05:03:41 +00:00
|
|
|
/* Errors */
|
|
|
|
|
{
|
2009-02-05 12:28:19 +01:00
|
|
|
SCM err_msg;
|
|
|
|
|
|
2009-12-23 22:59:12 +01:00
|
|
|
/* FIXME: need to sync regs before allocating anything, in each case. */
|
|
|
|
|
|
static opcodes; refactor program/objcode division; use new assembly pipeline
* gdbinit: Untested attempts to get the stack fondling macros to deal
with the new program representation.
* libguile/frames.c (scm_vm_frame_arguments, scm_vm_frame_source)
(scm_vm_frame_local_ref, scm_vm_frame_local_set_x): SCM_PROGRAM_DATA is
a struct scm_objcode*.
* libguile/instructions.h:
* libguile/instructions.c: Hide the instruction table and the struct
scm_instruction structure; all access to instructions now goes through
procedures. This is because instructions are no longer in a packed
array indexed by opcode. Also, declare a mask that all instructions
should fit in.
* libguile/objcodes.h:
* libguile/objcodes.c: Rewrite so that object code directly maps its
arity and length from its bytecode. This makes it unnecessary to keep
this information in programs, allowing programs to be simple conses
between the code (objcodes) and data (the object table and the closure
variables).
* libguile/programs.c (scm_make_program): Rework so that make-program
takes objcode, an object table, and externals as arguments. It's much
clearer this way, and we avoid malloc().
* libguile/stacks.c (is_vm_bootstrap_frame): Update for program/objcode
changes.
* libguile/vm-engine.c (vm_run): Initialize the jump table on the first
run, with the opcodes declared in the instruction sources, and with bad
instructions raising an error instead of wandering off into the
Unknown.
* libguile/vm-engine.h (FETCH_LENGTH): Always represent lengths as 3
bytes. The old code was too error-prone.
(NEXT_JUMP): Mask the instruction with SCM_VM_INSTRUCTION_MASK.
(NEW_FRAME): Update for program/objcode changes.
* libguile/vm-expand.h (VM_DEFINE_FUNCTION, VM_DEFINE_INSTRUCTION)
(VM_DEFINE_LOADER): Update so that we explicitly specify opcodes, so
that we have a stable bytecode API.
* libguile/vm-i-loader.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(load-integer): Use an int instead of a long as the accumulator; still
need to revisit this code at some point, I think.
(load-program): Simplify, thankfully!! Just creates the objcode slice
and rolls with it.
* libguile/vm-i-scheme.c: Number the opcodes explicitly.
* libguile/vm-i-system.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(make-closure): Update for new program API.
* libguile/vm.c (vm_make_boot_program): Update for new program/objcode
API. Still a bit ugly.
(scm_load_compiled_with_vm): Update for new program/objcode API.
* module/language/assembly.scm (byte-length): Fix byte-length calculation
for loaders, and load-program.
(code-pack, code-unpack): Start to move things from (system vm conv)
here.
(object->code, code->object): More things from conv.scm.
* module/language/glil.scm (<glil-program>): Add a new field,
closure-level.
(make-glil-program, compute-closure-level): Calculate the "closure
level" when making a glil program. This is the maximum depth of
external binding refs in this closure.
(unparse-glil): Fix label serialization.
* module/language/glil/compile-assembly.scm (make-meta): Prepend #f for
the meta's object table, though maybe in the future we can avoid
creating assembly in the first place.
(assoc-ref-or-acons, object-index-and-alist): GRRR! Caught again by the
different sets of arguments to assoc and assoc-ref!
(glil->assembly): Attempt to make the <glil-program> case more
readable, and fix the bugs. Sorry I don't know how to comment this
change any more than this.
(glil->assembly): For <glil-module> serialize the whole key, not just
the name.
(dump-object): subprogram-code is already a list. Serialize integers as
strings, not u8vectors. Fix the order of lists and vectors.
* module/language/glil/spec.scm (glil): Switch orders, so we prefer glil
-> assembly -> objcode. Actually glil->objcode doesn't work any more,
needs to be removed I think.
* module/language/objcode/spec.scm (objcode->value):
s/objcode->program/make-program/.
* module/language/scheme/inline.scm: Add acons inline.
* module/system/vm/conv.scm (make-byte-decoder): Skip the first 8 bytes,
they are header. Handle subprograms properly. Still needs help though.
(decode-length): Lengths are always 3 bytes now.
* module/system/vm/disasm.scm: Superficial changes to keep things
working. I'd like to fix this better in the future.
* module/system/vm/frame.scm (bootstrap-frame?): Fixes for
program-bytecode.
* module/system/vm/program.scm: Export make-program. It's program-objcode
now, no more program-bytecode.
* module/system/vm/vm.scm (vm-load): Use make-program.
* test-suite/tests/asm-to-bytecode.test: New test, very minimal.
* module/system/vm/objcode.scm: Export word-size, byte-order, and
write-objcode.
2009-01-29 21:09:04 +01:00
|
|
|
vm_error_bad_instruction:
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("VM: Bad instruction: ~s");
|
2009-04-05 20:15:11 +02:00
|
|
|
finish_args = scm_list_1 (scm_from_uchar (ip[-1]));
|
static opcodes; refactor program/objcode division; use new assembly pipeline
* gdbinit: Untested attempts to get the stack fondling macros to deal
with the new program representation.
* libguile/frames.c (scm_vm_frame_arguments, scm_vm_frame_source)
(scm_vm_frame_local_ref, scm_vm_frame_local_set_x): SCM_PROGRAM_DATA is
a struct scm_objcode*.
* libguile/instructions.h:
* libguile/instructions.c: Hide the instruction table and the struct
scm_instruction structure; all access to instructions now goes through
procedures. This is because instructions are no longer in a packed
array indexed by opcode. Also, declare a mask that all instructions
should fit in.
* libguile/objcodes.h:
* libguile/objcodes.c: Rewrite so that object code directly maps its
arity and length from its bytecode. This makes it unnecessary to keep
this information in programs, allowing programs to be simple conses
between the code (objcodes) and data (the object table and the closure
variables).
* libguile/programs.c (scm_make_program): Rework so that make-program
takes objcode, an object table, and externals as arguments. It's much
clearer this way, and we avoid malloc().
* libguile/stacks.c (is_vm_bootstrap_frame): Update for program/objcode
changes.
* libguile/vm-engine.c (vm_run): Initialize the jump table on the first
run, with the opcodes declared in the instruction sources, and with bad
instructions raising an error instead of wandering off into the
Unknown.
* libguile/vm-engine.h (FETCH_LENGTH): Always represent lengths as 3
bytes. The old code was too error-prone.
(NEXT_JUMP): Mask the instruction with SCM_VM_INSTRUCTION_MASK.
(NEW_FRAME): Update for program/objcode changes.
* libguile/vm-expand.h (VM_DEFINE_FUNCTION, VM_DEFINE_INSTRUCTION)
(VM_DEFINE_LOADER): Update so that we explicitly specify opcodes, so
that we have a stable bytecode API.
* libguile/vm-i-loader.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(load-integer): Use an int instead of a long as the accumulator; still
need to revisit this code at some point, I think.
(load-program): Simplify, thankfully!! Just creates the objcode slice
and rolls with it.
* libguile/vm-i-scheme.c: Number the opcodes explicitly.
* libguile/vm-i-system.c: Update license to LGPLv2+. Explicitly declare
opcodes.
(make-closure): Update for new program API.
* libguile/vm.c (vm_make_boot_program): Update for new program/objcode
API. Still a bit ugly.
(scm_load_compiled_with_vm): Update for new program/objcode API.
* module/language/assembly.scm (byte-length): Fix byte-length calculation
for loaders, and load-program.
(code-pack, code-unpack): Start to move things from (system vm conv)
here.
(object->code, code->object): More things from conv.scm.
* module/language/glil.scm (<glil-program>): Add a new field,
closure-level.
(make-glil-program, compute-closure-level): Calculate the "closure
level" when making a glil program. This is the maximum depth of
external binding refs in this closure.
(unparse-glil): Fix label serialization.
* module/language/glil/compile-assembly.scm (make-meta): Prepend #f for
the meta's object table, though maybe in the future we can avoid
creating assembly in the first place.
(assoc-ref-or-acons, object-index-and-alist): GRRR! Caught again by the
different sets of arguments to assoc and assoc-ref!
(glil->assembly): Attempt to make the <glil-program> case more
readable, and fix the bugs. Sorry I don't know how to comment this
change any more than this.
(glil->assembly): For <glil-module> serialize the whole key, not just
the name.
(dump-object): subprogram-code is already a list. Serialize integers as
strings, not u8vectors. Fix the order of lists and vectors.
* module/language/glil/spec.scm (glil): Switch orders, so we prefer glil
-> assembly -> objcode. Actually glil->objcode doesn't work any more,
needs to be removed I think.
* module/language/objcode/spec.scm (objcode->value):
s/objcode->program/make-program/.
* module/language/scheme/inline.scm: Add acons inline.
* module/system/vm/conv.scm (make-byte-decoder): Skip the first 8 bytes,
they are header. Handle subprograms properly. Still needs help though.
(decode-length): Lengths are always 3 bytes now.
* module/system/vm/disasm.scm: Superficial changes to keep things
working. I'd like to fix this better in the future.
* module/system/vm/frame.scm (bootstrap-frame?): Fixes for
program-bytecode.
* module/system/vm/program.scm: Export make-program. It's program-objcode
now, no more program-bytecode.
* module/system/vm/vm.scm (vm-load): Use make-program.
* test-suite/tests/asm-to-bytecode.test: New test, very minimal.
* module/system/vm/objcode.scm: Export word-size, byte-order, and
write-objcode.
2009-01-29 21:09:04 +01:00
|
|
|
goto vm_error;
|
|
|
|
|
|
2001-04-01 05:03:41 +00:00
|
|
|
vm_error_unbound:
|
2010-12-07 12:26:07 +01:00
|
|
|
/* FINISH_ARGS should be the name of the unbound variable. */
|
|
|
|
|
SYNC_ALL ();
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("Unbound variable: ~s");
|
2010-12-07 12:26:07 +01:00
|
|
|
scm_error_scm (scm_misc_error_key, program, err_msg,
|
|
|
|
|
scm_list_1 (finish_args), SCM_BOOL_F);
|
2001-04-01 05:03:41 +00:00
|
|
|
goto vm_error;
|
|
|
|
|
|
2010-08-14 18:35:17 -04:00
|
|
|
vm_error_unbound_fluid:
|
|
|
|
|
SYNC_ALL ();
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("Unbound fluid: ~s");
|
2010-08-14 18:35:17 -04:00
|
|
|
scm_error_scm (scm_misc_error_key, program, err_msg,
|
|
|
|
|
scm_list_1 (finish_args), SCM_BOOL_F);
|
|
|
|
|
goto vm_error;
|
|
|
|
|
|
2011-01-27 10:49:51 +01:00
|
|
|
vm_error_not_a_variable:
|
|
|
|
|
SYNC_ALL ();
|
|
|
|
|
scm_error (scm_arg_type_key, func_name, "Not a variable: ~S",
|
|
|
|
|
scm_list_1 (finish_args), scm_list_1 (finish_args));
|
|
|
|
|
goto vm_error;
|
|
|
|
|
|
better error reporting from the vm
* libguile/vm-engine.c: Add func_name local, for error reporting.
(vm_error_apply_to_non_list): New error case.
(vm_error_wrong_type_arg): Remove this generic error case.
(vm_error_wrong_type_apply): Remove FUNC_NAME -- no sense in seeing
"vm-debug-engine" in the error report.
(vm_error_not_a_pair, vm_error_not_a_bytevector)
(vm_error_not_a_struct, vm_error_not_a_thunk): Use func_name instead
of FUNC_NAME, so we can indicate what caused the error.
* libguile/vm-i-scheme.c (VM_VALIDATE_CONS, car, cdr, set-car!)
(set-cdr!): Indicate provenance of errors.
(VM_VALIDATE_STRUCT, struct-vtable):
(VM_VALIDATE_BYTEVECTOR, BV_FIXABLE_INT_REF, BV_INT_REF)
(BV_FLOAT_REF, BV_FIXABLE_INT_SET, BV_INT_SET, BV_FLOAT_SET): Same.
* libguile/vm-i-system.c (apply, tail-apply): Use
vm_error_apply_to_non_list.
2010-07-15 12:44:15 +02:00
|
|
|
vm_error_apply_to_non_list:
|
2011-01-26 21:44:12 +01:00
|
|
|
SYNC_ALL ();
|
better error reporting from the vm
* libguile/vm-engine.c: Add func_name local, for error reporting.
(vm_error_apply_to_non_list): New error case.
(vm_error_wrong_type_arg): Remove this generic error case.
(vm_error_wrong_type_apply): Remove FUNC_NAME -- no sense in seeing
"vm-debug-engine" in the error report.
(vm_error_not_a_pair, vm_error_not_a_bytevector)
(vm_error_not_a_struct, vm_error_not_a_thunk): Use func_name instead
of FUNC_NAME, so we can indicate what caused the error.
* libguile/vm-i-scheme.c (VM_VALIDATE_CONS, car, cdr, set-car!)
(set-cdr!): Indicate provenance of errors.
(VM_VALIDATE_STRUCT, struct-vtable):
(VM_VALIDATE_BYTEVECTOR, BV_FIXABLE_INT_REF, BV_INT_REF)
(BV_FLOAT_REF, BV_FIXABLE_INT_SET, BV_INT_SET, BV_FLOAT_SET): Same.
* libguile/vm-i-system.c (apply, tail-apply): Use
vm_error_apply_to_non_list.
2010-07-15 12:44:15 +02:00
|
|
|
scm_error (scm_arg_type_key, "apply", "Apply to non-list: ~S",
|
2011-01-26 21:44:12 +01:00
|
|
|
scm_list_1 (finish_args), scm_list_1 (finish_args));
|
2001-04-07 23:32:30 +00:00
|
|
|
goto vm_error;
|
|
|
|
|
|
2009-10-17 17:23:09 +02:00
|
|
|
vm_error_kwargs_length_not_even:
|
2009-12-23 22:59:12 +01:00
|
|
|
SYNC_ALL ();
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("Odd length of keyword argument list");
|
2009-12-23 22:59:12 +01:00
|
|
|
scm_error_scm (sym_keyword_argument_error, program, err_msg,
|
|
|
|
|
SCM_EOL, SCM_BOOL_F);
|
2009-10-17 17:23:09 +02:00
|
|
|
|
|
|
|
|
vm_error_kwargs_invalid_keyword:
|
2009-12-23 22:59:12 +01:00
|
|
|
/* FIXME say which one it was */
|
|
|
|
|
SYNC_ALL ();
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("Invalid keyword");
|
2009-12-23 22:59:12 +01:00
|
|
|
scm_error_scm (sym_keyword_argument_error, program, err_msg,
|
|
|
|
|
SCM_EOL, SCM_BOOL_F);
|
2009-10-17 17:23:09 +02:00
|
|
|
|
|
|
|
|
vm_error_kwargs_unrecognized_keyword:
|
2009-12-23 22:59:12 +01:00
|
|
|
/* FIXME say which one it was */
|
|
|
|
|
SYNC_ALL ();
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("Unrecognized keyword");
|
2009-12-23 22:59:12 +01:00
|
|
|
scm_error_scm (sym_keyword_argument_error, program, err_msg,
|
|
|
|
|
SCM_EOL, SCM_BOOL_F);
|
2009-10-17 17:23:09 +02:00
|
|
|
|
enable multiple vm engines (regular, debug, ...)
* libguile/vm-engine.c (VM_USE_HOOKS, VM_USE_CLOCK, VM_CHECK_EXTERNAL)
(VM_CHECK_OBJECT): Update to define these here, before including
vm-engine.h.
(vm_run): Change so that we can make different engines. Also, we take
an array of arguments, and the struct scm_vm directly, so as to avoid
any need to cons.
* libguile/vm-engine.h (CHECK_EXTERNAL, CHECK_OBJECT): Add some UNLIKELY
bits; don't seem to help.
* libguile/vm.c (vm_dispatch_hook): Change to not pass the VP. This needs
some love, and perhaps we revert to the old way.
(VM_ENGINE): Actually make two engines, vm_regular_engine and
vm_debug_engine. Probably there is room for improvement here. Actually
their speeds are the same at the moment.
(make_vm): Choose which engine to run; currently the debug engine by
default.
(scm_c_vm_run): A thin wrapper to invoke a VM without consing.
(scm_vm_apply): Use scm_c_vm_run.
(scm_load_compiled_with_vm): Use scm_c_vm_run.
2009-02-04 23:47:56 +01:00
|
|
|
vm_error_too_many_args:
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("VM: Too many arguments");
|
2009-04-05 20:15:11 +02:00
|
|
|
finish_args = scm_list_1 (scm_from_int (nargs));
|
enable multiple vm engines (regular, debug, ...)
* libguile/vm-engine.c (VM_USE_HOOKS, VM_USE_CLOCK, VM_CHECK_EXTERNAL)
(VM_CHECK_OBJECT): Update to define these here, before including
vm-engine.h.
(vm_run): Change so that we can make different engines. Also, we take
an array of arguments, and the struct scm_vm directly, so as to avoid
any need to cons.
* libguile/vm-engine.h (CHECK_EXTERNAL, CHECK_OBJECT): Add some UNLIKELY
bits; don't seem to help.
* libguile/vm.c (vm_dispatch_hook): Change to not pass the VP. This needs
some love, and perhaps we revert to the old way.
(VM_ENGINE): Actually make two engines, vm_regular_engine and
vm_debug_engine. Probably there is room for improvement here. Actually
their speeds are the same at the moment.
(make_vm): Choose which engine to run; currently the debug engine by
default.
(scm_c_vm_run): A thin wrapper to invoke a VM without consing.
(scm_vm_apply): Use scm_c_vm_run.
(scm_load_compiled_with_vm): Use scm_c_vm_run.
2009-02-04 23:47:56 +01:00
|
|
|
goto vm_error;
|
|
|
|
|
|
2001-04-01 05:03:41 +00:00
|
|
|
vm_error_wrong_num_args:
|
2008-10-02 11:00:55 +02:00
|
|
|
/* nargs and program are valid */
|
2008-10-09 12:17:51 +02:00
|
|
|
SYNC_ALL ();
|
2008-10-02 11:00:55 +02:00
|
|
|
scm_wrong_num_args (program);
|
|
|
|
|
/* shouldn't get here */
|
2001-04-01 05:03:41 +00:00
|
|
|
goto vm_error;
|
|
|
|
|
|
|
|
|
|
vm_error_wrong_type_apply:
|
2009-08-20 12:46:36 +02:00
|
|
|
SYNC_ALL ();
|
better error reporting from the vm
* libguile/vm-engine.c: Add func_name local, for error reporting.
(vm_error_apply_to_non_list): New error case.
(vm_error_wrong_type_arg): Remove this generic error case.
(vm_error_wrong_type_apply): Remove FUNC_NAME -- no sense in seeing
"vm-debug-engine" in the error report.
(vm_error_not_a_pair, vm_error_not_a_bytevector)
(vm_error_not_a_struct, vm_error_not_a_thunk): Use func_name instead
of FUNC_NAME, so we can indicate what caused the error.
* libguile/vm-i-scheme.c (VM_VALIDATE_CONS, car, cdr, set-car!)
(set-cdr!): Indicate provenance of errors.
(VM_VALIDATE_STRUCT, struct-vtable):
(VM_VALIDATE_BYTEVECTOR, BV_FIXABLE_INT_REF, BV_INT_REF)
(BV_FLOAT_REF, BV_FIXABLE_INT_SET, BV_INT_SET, BV_FLOAT_SET): Same.
* libguile/vm-i-system.c (apply, tail-apply): Use
vm_error_apply_to_non_list.
2010-07-15 12:44:15 +02:00
|
|
|
scm_error (scm_arg_type_key, NULL, "Wrong type to apply: ~S",
|
really boot primitive-eval from scheme.
* libguile/eval.c (scm_primitive_eval, scm_c_primitive_eval):
(scm_init_eval): Rework so that scm_primitive_eval always calls out to
the primitive-eval variable. The previous definition is the default
value, which is probably overridden by scm_init_eval_in_scheme.
* libguile/init.c (scm_i_init_guile): Move ports and load-path up, so we
can debug when initing eval. Call scm_init_eval_in_scheme. Awesome.
* libguile/load.h:
* libguile/load.c (scm_init_eval_in_scheme): New procedure, loads up
ice-9/eval.scm to replace the primitive-eval definition, if everything
is there and up-to-date.
* libguile/modules.c (scm_module_transformer): Export to Scheme, so it's
there for eval.go.
* module/ice-9/boot-9.scm: No need to define module-transformer.
* module/ice-9/eval.scm (capture-env): Only reference the-root-module if
modules are booted.
(primitive-eval): Inline a definition for identity. Throw a more
standard error for "wrong number of arguments".
* module/ice-9/psyntax.scm (chi-install-global): The macro binding for a
syncase macro is now a pair: the transformer, and the module that was
current when the transformer was installed. The latter is used for
hygiene purposes, replacing the use of procedure-module, which didn't
work with the interpreter's shared-code closures.
(chi-macro): Adapt for the binding being a pair, and get the hygiene
from the cdr.
(eval-local-transformer): Adapt to new form of macro bindings.
* module/ice-9/psyntax-pp.scm: Regenerated.
* .gitignore: Ignore eval.go.stamp.
* module/Makefile.am: Reorder for fastest serial compilation, now that
there are no ordering constraints. I did a number of experiments here
and this seems to be the best; but the bulk of the time is compiling
psyntax-pp.scm with eval.scm. Not so great.
* libguile/vm-engine.c (vm-engine): Throw a more standard error for
"wrong type to apply".
* test-suite/tests/gc.test ("gc"): Remove a hack that shouldn't affect
the new evaluator, and throw in another (gc) for good measure.
* test-suite/tests/goops.test ("defining classes"):
* test-suite/tests/hooks.test (proc1): We can't currently check what the
arity is of a closure made by eval.scm -- or more accurately all
closures have 0 required args and no rest args. So punt for now.
* test-suite/tests/syntax.test ("letrec"): The scheme evaluator can't
check that a variable is unbound, currently; perhaps the full "fixing
letrec" expansion could fix this. But barring that, punt.
2009-12-01 23:54:25 +01:00
|
|
|
scm_list_1 (program), scm_list_1 (program));
|
2001-04-01 05:03:41 +00:00
|
|
|
goto vm_error;
|
|
|
|
|
|
2001-04-20 18:35:59 +00:00
|
|
|
vm_error_stack_overflow:
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("VM: Stack overflow");
|
2009-02-05 12:28:19 +01:00
|
|
|
finish_args = SCM_EOL;
|
2010-05-26 23:00:58 +02:00
|
|
|
if (stack_limit < vp->stack_base + vp->stack_size)
|
|
|
|
|
/* There are VM_STACK_RESERVE_SIZE bytes left. Make them available so
|
|
|
|
|
that `throw' below can run on this VM. */
|
|
|
|
|
vp->stack_limit = vp->stack_base + vp->stack_size;
|
2001-04-01 05:03:41 +00:00
|
|
|
goto vm_error;
|
|
|
|
|
|
2001-04-20 18:35:59 +00:00
|
|
|
vm_error_stack_underflow:
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("VM: Stack underflow");
|
2009-02-05 12:28:19 +01:00
|
|
|
finish_args = SCM_EOL;
|
2001-04-01 05:03:41 +00:00
|
|
|
goto vm_error;
|
|
|
|
|
|
2008-10-16 13:24:39 +02:00
|
|
|
vm_error_improper_list:
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("Expected a proper list, but got object with tail ~s");
|
2008-10-16 13:24:39 +02:00
|
|
|
goto vm_error;
|
|
|
|
|
|
2008-10-18 19:21:44 +02:00
|
|
|
vm_error_not_a_pair:
|
|
|
|
|
SYNC_ALL ();
|
better error reporting from the vm
* libguile/vm-engine.c: Add func_name local, for error reporting.
(vm_error_apply_to_non_list): New error case.
(vm_error_wrong_type_arg): Remove this generic error case.
(vm_error_wrong_type_apply): Remove FUNC_NAME -- no sense in seeing
"vm-debug-engine" in the error report.
(vm_error_not_a_pair, vm_error_not_a_bytevector)
(vm_error_not_a_struct, vm_error_not_a_thunk): Use func_name instead
of FUNC_NAME, so we can indicate what caused the error.
* libguile/vm-i-scheme.c (VM_VALIDATE_CONS, car, cdr, set-car!)
(set-cdr!): Indicate provenance of errors.
(VM_VALIDATE_STRUCT, struct-vtable):
(VM_VALIDATE_BYTEVECTOR, BV_FIXABLE_INT_REF, BV_INT_REF)
(BV_FLOAT_REF, BV_FIXABLE_INT_SET, BV_INT_SET, BV_FLOAT_SET): Same.
* libguile/vm-i-system.c (apply, tail-apply): Use
vm_error_apply_to_non_list.
2010-07-15 12:44:15 +02:00
|
|
|
scm_wrong_type_arg_msg (func_name, 1, finish_args, "pair");
|
2008-10-18 19:21:44 +02:00
|
|
|
/* shouldn't get here */
|
|
|
|
|
goto vm_error;
|
|
|
|
|
|
2009-06-24 23:49:11 +02:00
|
|
|
vm_error_not_a_bytevector:
|
|
|
|
|
SYNC_ALL ();
|
better error reporting from the vm
* libguile/vm-engine.c: Add func_name local, for error reporting.
(vm_error_apply_to_non_list): New error case.
(vm_error_wrong_type_arg): Remove this generic error case.
(vm_error_wrong_type_apply): Remove FUNC_NAME -- no sense in seeing
"vm-debug-engine" in the error report.
(vm_error_not_a_pair, vm_error_not_a_bytevector)
(vm_error_not_a_struct, vm_error_not_a_thunk): Use func_name instead
of FUNC_NAME, so we can indicate what caused the error.
* libguile/vm-i-scheme.c (VM_VALIDATE_CONS, car, cdr, set-car!)
(set-cdr!): Indicate provenance of errors.
(VM_VALIDATE_STRUCT, struct-vtable):
(VM_VALIDATE_BYTEVECTOR, BV_FIXABLE_INT_REF, BV_INT_REF)
(BV_FLOAT_REF, BV_FIXABLE_INT_SET, BV_INT_SET, BV_FLOAT_SET): Same.
* libguile/vm-i-system.c (apply, tail-apply): Use
vm_error_apply_to_non_list.
2010-07-15 12:44:15 +02:00
|
|
|
scm_wrong_type_arg_msg (func_name, 1, finish_args, "bytevector");
|
2009-06-24 23:49:11 +02:00
|
|
|
/* shouldn't get here */
|
|
|
|
|
goto vm_error;
|
|
|
|
|
|
2009-12-11 12:44:29 +01:00
|
|
|
vm_error_not_a_struct:
|
|
|
|
|
SYNC_ALL ();
|
better error reporting from the vm
* libguile/vm-engine.c: Add func_name local, for error reporting.
(vm_error_apply_to_non_list): New error case.
(vm_error_wrong_type_arg): Remove this generic error case.
(vm_error_wrong_type_apply): Remove FUNC_NAME -- no sense in seeing
"vm-debug-engine" in the error report.
(vm_error_not_a_pair, vm_error_not_a_bytevector)
(vm_error_not_a_struct, vm_error_not_a_thunk): Use func_name instead
of FUNC_NAME, so we can indicate what caused the error.
* libguile/vm-i-scheme.c (VM_VALIDATE_CONS, car, cdr, set-car!)
(set-cdr!): Indicate provenance of errors.
(VM_VALIDATE_STRUCT, struct-vtable):
(VM_VALIDATE_BYTEVECTOR, BV_FIXABLE_INT_REF, BV_INT_REF)
(BV_FLOAT_REF, BV_FIXABLE_INT_SET, BV_INT_SET, BV_FLOAT_SET): Same.
* libguile/vm-i-system.c (apply, tail-apply): Use
vm_error_apply_to_non_list.
2010-07-15 12:44:15 +02:00
|
|
|
scm_wrong_type_arg_msg (func_name, 1, finish_args, "struct");
|
2009-12-11 12:44:29 +01:00
|
|
|
/* shouldn't get here */
|
|
|
|
|
goto vm_error;
|
|
|
|
|
|
2010-01-30 15:45:37 +01:00
|
|
|
vm_error_not_a_thunk:
|
|
|
|
|
SYNC_ALL ();
|
better error reporting from the vm
* libguile/vm-engine.c: Add func_name local, for error reporting.
(vm_error_apply_to_non_list): New error case.
(vm_error_wrong_type_arg): Remove this generic error case.
(vm_error_wrong_type_apply): Remove FUNC_NAME -- no sense in seeing
"vm-debug-engine" in the error report.
(vm_error_not_a_pair, vm_error_not_a_bytevector)
(vm_error_not_a_struct, vm_error_not_a_thunk): Use func_name instead
of FUNC_NAME, so we can indicate what caused the error.
* libguile/vm-i-scheme.c (VM_VALIDATE_CONS, car, cdr, set-car!)
(set-cdr!): Indicate provenance of errors.
(VM_VALIDATE_STRUCT, struct-vtable):
(VM_VALIDATE_BYTEVECTOR, BV_FIXABLE_INT_REF, BV_INT_REF)
(BV_FLOAT_REF, BV_FIXABLE_INT_SET, BV_INT_SET, BV_FLOAT_SET): Same.
* libguile/vm-i-system.c (apply, tail-apply): Use
vm_error_apply_to_non_list.
2010-07-15 12:44:15 +02:00
|
|
|
scm_wrong_type_arg_msg ("dynamic-wind", 1, finish_args, "thunk");
|
2010-01-30 15:45:37 +01:00
|
|
|
/* shouldn't get here */
|
|
|
|
|
goto vm_error;
|
|
|
|
|
|
add multiple values support to the vm
* libguile/vm-engine.c (vm_run): The bootstrap program now uses mv_call,
so as to allow multiple values out of the VM. (It did before, because
multiple values were represented internally as single scm_values
objects, but now that values go on the stack, we need to note the boot
frame as accepting multiple values.)
(vm_error_no_values): New error, happens if you pass no values into a
single-value continuation. Passing more than one is OK though, it just
takes the first one.
* libguile/vm-i-system.c (halt): Assume that someone has pushed the
number of values onto the stack, and package up that number of values
as a scm_values() object, for communication with the interpreter.
(mv-call): New instruction, calls a procedure with a multiple-value
continuation, even handling calls out to the interpreter.
(return/values): New instruction, returns multiple values to the
continuation. If the continuation is single-valued, takes the first
value or errors if there are no values. Otherwise it returns to the
multiple-value return address, pushing the number of values on top of
the values.
* module/system/il/compile.scm (codegen): Compile <ghil-values> forms.
* module/system/il/ghil.scm (<ghil-values>) Add new GHIL data structure
and associated procedures.
* module/language/scheme/translate.scm (custom-transformer-table):
Compile (values .. ) forms into <ghil-values>.
2008-09-14 17:06:52 +02:00
|
|
|
vm_error_no_values:
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("Zero values returned to single-valued continuation");
|
2009-02-05 12:28:19 +01:00
|
|
|
finish_args = SCM_EOL;
|
add multiple values support to the vm
* libguile/vm-engine.c (vm_run): The bootstrap program now uses mv_call,
so as to allow multiple values out of the VM. (It did before, because
multiple values were represented internally as single scm_values
objects, but now that values go on the stack, we need to note the boot
frame as accepting multiple values.)
(vm_error_no_values): New error, happens if you pass no values into a
single-value continuation. Passing more than one is OK though, it just
takes the first one.
* libguile/vm-i-system.c (halt): Assume that someone has pushed the
number of values onto the stack, and package up that number of values
as a scm_values() object, for communication with the interpreter.
(mv-call): New instruction, calls a procedure with a multiple-value
continuation, even handling calls out to the interpreter.
(return/values): New instruction, returns multiple values to the
continuation. If the continuation is single-valued, takes the first
value or errors if there are no values. Otherwise it returns to the
multiple-value return address, pushing the number of values on top of
the values.
* module/system/il/compile.scm (codegen): Compile <ghil-values> forms.
* module/system/il/ghil.scm (<ghil-values>) Add new GHIL data structure
and associated procedures.
* module/language/scheme/translate.scm (custom-transformer-table):
Compile (values .. ) forms into <ghil-values>.
2008-09-14 17:06:52 +02:00
|
|
|
goto vm_error;
|
|
|
|
|
|
lambda-lifting for (lambda () ...) as consumer of call-with-values
* libguile/vm-engine.c (vm_run): Add new error case,
vm_error_not_enough_values.
* libguile/vm-i-system.c (goto/nargs, call/nargs): So, in these cases, if
we get too many values, we don't truncate the values like we do in the
single-value continuation case, or in the mvbind case. What to do? I
guess we either truncate them here, or only allow the correct number of
values. Dunno. Mark the code as a fixme.
(truncate-values): New instruction, for mv-bind: checks that the number
of values on the stack is compatible with the number of bindings we
have arranged for them, truncating if necessary.
* module/language/scheme/translate.scm (custom-transformer-table):
Compile receive as a primary form -- not so much because it is a
primary form, but more to test the mv-bind machinery. Also it's more
efficient, I think.
* module/system/il/compile.scm (lift-variables!): New helper, factored
out of `optimize'.
(optimize): Add a few more cases. Adapt `lambda' optimization, which
isn't much. I'm not happy with ghil as a mungeable language.
Add a case for call-with-values with the second argument is
a lambda: lift the lambda. Untested.
(codegen): Refactor the push-bindings! code. Compile mv-bind.
* module/system/il/ghil.scm (<ghil-mv-bind>): Add mv-bind construct,
along with its procedures.
* module/system/il/glil.scm (<glil-mv-bind>): Add mv-bind construct,
different from the high-level one. It makes sense in the source, I
think.
* module/system/vm/assemble.scm (codegen): Assemble glil-mv-bind by
pushing onto the bindings list, and actually push some code to truncate
the values.
2008-09-18 22:49:55 +02:00
|
|
|
vm_error_not_enough_values:
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("Too few values returned to continuation");
|
2009-02-05 12:28:19 +01:00
|
|
|
finish_args = SCM_EOL;
|
lambda-lifting for (lambda () ...) as consumer of call-with-values
* libguile/vm-engine.c (vm_run): Add new error case,
vm_error_not_enough_values.
* libguile/vm-i-system.c (goto/nargs, call/nargs): So, in these cases, if
we get too many values, we don't truncate the values like we do in the
single-value continuation case, or in the mvbind case. What to do? I
guess we either truncate them here, or only allow the correct number of
values. Dunno. Mark the code as a fixme.
(truncate-values): New instruction, for mv-bind: checks that the number
of values on the stack is compatible with the number of bindings we
have arranged for them, truncating if necessary.
* module/language/scheme/translate.scm (custom-transformer-table):
Compile receive as a primary form -- not so much because it is a
primary form, but more to test the mv-bind machinery. Also it's more
efficient, I think.
* module/system/il/compile.scm (lift-variables!): New helper, factored
out of `optimize'.
(optimize): Add a few more cases. Adapt `lambda' optimization, which
isn't much. I'm not happy with ghil as a mungeable language.
Add a case for call-with-values with the second argument is
a lambda: lift the lambda. Untested.
(codegen): Refactor the push-bindings! code. Compile mv-bind.
* module/system/il/ghil.scm (<ghil-mv-bind>): Add mv-bind construct,
along with its procedures.
* module/system/il/glil.scm (<glil-mv-bind>): Add mv-bind construct,
different from the high-level one. It makes sense in the source, I
think.
* module/system/vm/assemble.scm (codegen): Assemble glil-mv-bind by
pushing onto the bindings list, and actually push some code to truncate
the values.
2008-09-18 22:49:55 +02:00
|
|
|
goto vm_error;
|
|
|
|
|
|
2010-02-25 00:18:07 +01:00
|
|
|
vm_error_continuation_not_rewindable:
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("Unrewindable partial continuation");
|
2010-02-25 00:18:07 +01:00
|
|
|
finish_args = scm_cons (finish_args, SCM_EOL);
|
|
|
|
|
goto vm_error;
|
|
|
|
|
|
2009-08-12 16:33:49 +02:00
|
|
|
vm_error_bad_wide_string_length:
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("VM: Bad wide string length: ~S");
|
2009-08-12 16:33:49 +02:00
|
|
|
goto vm_error;
|
|
|
|
|
|
2009-11-17 23:40:51 +01:00
|
|
|
#ifdef VM_CHECK_IP
|
2001-04-20 18:35:59 +00:00
|
|
|
vm_error_invalid_address:
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("VM: Invalid program address");
|
2009-02-05 12:28:19 +01:00
|
|
|
finish_args = SCM_EOL;
|
2001-04-01 05:03:41 +00:00
|
|
|
goto vm_error;
|
2001-04-20 18:35:59 +00:00
|
|
|
#endif
|
|
|
|
|
|
2005-06-25 03:13:56 +00:00
|
|
|
#if VM_CHECK_OBJECT
|
|
|
|
|
vm_error_object:
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("VM: Invalid object table access");
|
2009-02-05 12:28:19 +01:00
|
|
|
finish_args = SCM_EOL;
|
2005-06-25 03:13:56 +00:00
|
|
|
goto vm_error;
|
|
|
|
|
#endif
|
|
|
|
|
|
2009-07-23 14:36:22 +02:00
|
|
|
#if VM_CHECK_FREE_VARIABLES
|
|
|
|
|
vm_error_free_variable:
|
2011-04-29 11:07:25 +02:00
|
|
|
err_msg = scm_from_latin1_string ("VM: Invalid free variable access");
|
2009-07-19 19:48:26 +02:00
|
|
|
finish_args = SCM_EOL;
|
|
|
|
|
goto vm_error;
|
|
|
|
|
#endif
|
|
|
|
|
|
2001-04-01 05:03:41 +00:00
|
|
|
vm_error:
|
|
|
|
|
SYNC_ALL ();
|
2005-06-25 06:57:20 +00:00
|
|
|
|
2009-04-05 20:15:11 +02:00
|
|
|
scm_ithrow (sym_vm_error, scm_list_3 (sym_vm_run, err_msg, finish_args),
|
|
|
|
|
1);
|
2001-04-01 05:03:41 +00:00
|
|
|
}
|
|
|
|
|
|
2000-08-22 15:54:19 +00:00
|
|
|
abort (); /* never reached */
|
|
|
|
|
}
|
enable multiple vm engines (regular, debug, ...)
* libguile/vm-engine.c (VM_USE_HOOKS, VM_USE_CLOCK, VM_CHECK_EXTERNAL)
(VM_CHECK_OBJECT): Update to define these here, before including
vm-engine.h.
(vm_run): Change so that we can make different engines. Also, we take
an array of arguments, and the struct scm_vm directly, so as to avoid
any need to cons.
* libguile/vm-engine.h (CHECK_EXTERNAL, CHECK_OBJECT): Add some UNLIKELY
bits; don't seem to help.
* libguile/vm.c (vm_dispatch_hook): Change to not pass the VP. This needs
some love, and perhaps we revert to the old way.
(VM_ENGINE): Actually make two engines, vm_regular_engine and
vm_debug_engine. Probably there is room for improvement here. Actually
their speeds are the same at the moment.
(make_vm): Choose which engine to run; currently the debug engine by
default.
(scm_c_vm_run): A thin wrapper to invoke a VM without consing.
(scm_vm_apply): Use scm_c_vm_run.
(scm_load_compiled_with_vm): Use scm_c_vm_run.
2009-02-04 23:47:56 +01:00
|
|
|
|
|
|
|
|
#undef VM_USE_HOOKS
|
|
|
|
|
#undef VM_CHECK_OBJECT
|
2009-07-23 14:36:22 +02:00
|
|
|
#undef VM_CHECK_FREE_VARIABLE
|
VM tweaks
* libguile/vm-engine.c (VM_CHECK_OBJECT, VM_CHECK_FREE_VARIABLES): Set
to 0 for both engines. These are really internal debugging variables,
which don't affect user-visible features, provided that the compiler
is correct of course.
(VM_CHECK_UNDERFLOW): New var, also off by default: whether to check
for stack underflow when popping values.
(vm_engine): Don't declare object_count if we are not checking object
table accesses.
* libguile/vm-engine.h (CACHE_PROGRAM): Don't muck with object_count
if we are not checking object table accesses.
(CHECK_UNDERFLOW, PRE_CHECK_UNDERFLOW): Nop out if we are not checking
underflow.
(POP2, POP3): New macros which check for underflow before popping more
than one value.
* libguile/vm-i-loader.c (load_array):
* libguile/vm-i-scheme.c (set_car, set_cdr, vector_set, slot_set)
(BV_SET_WITH_ENDIANNESS, BV_FIXABLE_INT_SET, BV_INT_SET)
(BV_FLOAT_SET):
* libguile/vm-i-system.c (partial_cont_call, fix_closure, prompt)
(fluid_set): Use POP2 / POP3.
(local_set, long_local_set): Pop to locals instead of using values on
the stack then dropping; allows for underflow to be checked before the
value is accessed.
(BR): Don't NULLSTACK or DROP after the operation.
(br_if, br_if_not, br_if_eq, br_if_not_eq, br_if_null)
(br_if_not_null): Pop to locals before doing the compare and jump.
2011-05-05 14:04:23 +02:00
|
|
|
#undef VM_CHECK_UNDERFLOW
|
2001-04-01 05:03:41 +00:00
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
Local Variables:
|
|
|
|
|
c-file-style: "gnu"
|
|
|
|
|
End:
|
|
|
|
|
*/
|