* libguile/vm-i-system.c (goto/args): On a tail call to a different
procedure, init the locals to valid scheme values. Shouldn't matter for
well-compiled scheme, but inspecting uninitialized locals could give
garbage, or badly-compiled code could cause a crash.
* module/language/Makefile.am (NOCOMP_SOURCES): For the moment, don't
compile compile-ghil.scm. I need to fix this.
* module/language/ecmascript/compile-ghil.scm (load-toplevel): Sigh, and
disable stack checking in the evaluator too. Grr.
(comp): Implement (unnamed) break and continue.
* module/language/ecmascript/parse.scm (parse-ecmascript): Fix var
statements in `for' -- though it still doesn't work.
* libguile/vm-i-system.c (drop, return): Declare drop and return as
popping one arg from the stack.
* module/language/ghil/compile-glil.scm:
* module/language/glil/compile-assembly.scm (make-meta): Adjust so that
we declare 'drop and 'return calls as popping one arg from the stack.
* module/language/ecmascript/compile-ghil.scm (comp, comp-body): Flesh
out a bit more. Most significantly, scoping within functions obeys
javascript semantics better, modulo bits about with() forms.
* module/language/ecmascript/impl.scm: Define some runtime helper
routines.
* module/language/Makefile.am (SOURCES): Add impl.scm.
* module/language/ecmascript/parse.scm (parse-ecmascript): Minor tweaks.
* module/language/ecmascript/tokenize.scm (read-identifier): Identifiers
now read as symbols, not strings.
* libguile/goops.c (scm_make): In the pre-inst `make', default
`procedure' to #f, and read a `make-procedure' instead of
`compile-env'.
* libguile/goops.h (scm_si_make_procedure): This instead of
scm_si_compile_env.
* module/oop/goops.scm (make-method): Remove this unused function. Users
should use (make <method> ...) directly.
(method): Capture `make-procedure' instead of `procedure' in the case
that the body calls a next-method. Allows for the kind of
"recompilation" that we were using before, but with closures instead of
re-entering the compiler. Type-specific compilation is still
interesting, but probably should be implemented in another way.
(initialize): Default #:procedure to #f, and
s/compile-env/make-procedure/.
* module/oop/goops/compile.scm (code-table-lookup): Just return the
cmethod, not the entry -- since the entry is now just (append types
cmethod).
(compile-make-procedure): New procedure, returns a form that, when
evaluated/compiled, will yield a procedure of one argument, the
next-method. When called with a next-method, the procedure returns an
actual method implementation. compile-make-procedure returns #f if the
body doesn't call next-method.
(compile-method): Unify to always return procedures. Much cleaner and
*much* faster in the compiled case. In the interpreted case, there
might be a slight slowdown, but if there is one it should be slight.
* module/oop/goops/dispatch.scm (method-cache-install!): Adapt to removal
of compute-entry-with-cmethod.
* module/language/ghil/compile-glil.scm (codegen): Record source location
for offset 0 into a lambda, if we can.
* module/language/scheme/compile-ghil.scm (translate-1)
(define-scheme-translator): In the retrans procedures, propagate the
location information from the enclosing expression if the subexpression
has no location information. Gives source information to many more
expressions.
(location): Just propagate the source properties as they are, the
glil->assembly compiler will interpret them.
* module/language/glil.scm (<glil>): Change glil-source to take "props"
and not "loc", as it's the source properties that we're interested in.
* module/language/glil/compile-assembly.scm (limn-sources): New function,
takes a list of addr-source property pairs and "compresses" them for
serialization to disk.
(glil->assembly): Limn the sources before writing them to disk. Avoid
non-tail recursion when determining total byte length of code.
* module/system/vm/program.scm (source:file, source:line, source:column):
Update for new source representation.
(program-source): Export.
(write-program): Nicer pretty-printing of anonymous procedures.
* libguile/backtrace.c (display_backtrace_get_file_line): Update for the
new VM source representation.
* libguile/programs.h:
* libguile/programs.c (scm_program_sources): Update for the new
serialized source representation, where the filename is not in the
stream unless it changes.
(scm_program_source): New exported function, looks up the source for a
given ip offset.
(scm_c_program_source): Update to return the last source information
that was <= the given IP, because we only serialize source info when it
changes.
* libguile/stacks.c (scm_make_stack): Instead of aborting when we misread
the number of stack frames, just print a warning. I'd like to figure
out what these cases are, exactly.
* module/language/scheme/compile-ghil.scm (lambda): Reindent the lambda
transformer.
* module/system/base/compile.scm (call-with-compile-error-catch): Write
the expression instead of displaying it.
(call-with-output-file/atomic): Don't actually redirect output to this
port, as it's not necessary -- the language-printer should respect the
port that we pass.
* libguile/frames.c (vm_frame_print): Add a frame printer.
* libguile/stacks.c (stack_depth, read_frames): Only switch the VM stack
for boot program dframes.
* libguile/vm-engine.c (VM_NAME): Push one debug frame per invocation,
unconditionally. (If we push them at all, of course.)
* libguile/programs.h (SCM_F_PROGRAM_IS_BOOT, SCM_PROGRAM_IS_BOOT): Flags
for determining if a program is a boot program. It turns out that our
heuristics e.g. in stacks.c would catch non-boot programs, like
programs that end with (goto/args 1), because the 1 is the same byte as
`halt'. That took a while to find...
* libguile/stacks.c (stack_depth, read_frames): Use the new boot prog
macros.
(scm_make_stack): Assert that we read the number of frames that we said
we would.
* libguile/vm.c (really_make_boot_program): Mark boot programs
appropriately.
* libguile/vm-engine.c (VM_PUSH_DEBUG_FRAMES): New knob, if true we much
with the scm_i_last_debug_frame when entering the VM, because sometimes
the evaluator doesn't do it for us.
(VM_ENGINE): Plug through debug frame fondling. Now, program exit comes
back to the main text. Rename err_args to finish_args, and reuse for
the return value.
* libguile/vm-engine.h (PUSH_LIST):
* libguile/vm-i-loader.c:
* libguile/vm-i-scheme.c:
* libguile/vm-i-system.c: Update for finish_args.
(halt): goto vm_done, now, instead of returning directly.
* libguile/eval.c (scm_call_0, scm_call_1, scm_call_2, scm_call_3)
(scm_call_4): Special-case compiled procedures here, to avoid consing.
* libguile/vm.h:
* libguile/vm.c (scm_c_vm_run): Take a SCM after all.
(scm_vm_apply, scm_load_compiled_with_vm): Adapt to vm_run change.
* libguile/programs.c (scm_make_program): Add a comment.
* libguile/vm-engine.h (INIT_ARGS): Add a couple of UNLIKELY notes.
* libguile/vm-i-system.c (make-closure): Inline the call to
scm_make_program, which avoids some redundant checks.
* 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.
* libguile/throw.c (scm_c_catch): Stash away the current vm's regs, and
restore them if there's a nonlocal exit. There is a terrible case we
have to handle if we catch from when the vm smob type isn't registered
but the throw has the vm registered, but I think we handle this fine.
* libguile/vm-engine.c (vm_run):
* libguile/vm-i-system.c (halt): Don't make a dynwind context, so that
entering the VM doesn't cons at all, except for the arg list. Maybe we
can fix that bit too.
* libguile/vm.c (vm_reset_stack): Remove, as there is no more dynwind.
(make_vm): Return #f if the tc16 hasn't yet been registered.
* libguile/instructions.c (fetch_instruction_table)
(scm_lookup_instruction_by_name): Rework so we lazily load instructions
into an array keyed by opcode, and a hash table keyed by symbolic name.
Much faster, in this hot spot of compilation.
* libguile/vm-engine.c (vm_run): Use malloc instead of scm_gc_malloc,
given that we aren't ever going to free this thing.
* libguile/vm-expand.h (VM_DEFINE_FUNCTION, VM_DEFINE_LOADER): Rework to
always be aliases to VM_DEFINE_INSTRUCTION.
(VM_DEFINE_INSTRUCTION): In the table case, update to work with
fetch_instruction_table().
* libguile/objects.c (scm_apply_generic): Inline the case when the
generic is a program.
* libguile/vm-i-system.c (return): Tick when functions return.
* module/oop/goops.scm (object-eqv?, object-equal?): Remove these
historical methods.
* libguile/goops.c (scm_smob_class): Statically allocate it.
(create_smob_classes): Don't malloc(3) `scm_smob_class'.
* libguile/goops.h (scm_smob_class): Update declaration.
* libguile/smob.c (scm_make_smob_type, scm_set_smob_apply): When
checking whether GOOPS is initialized, check whether the first element
of SCM_SMOB_CLASS is non-zero.
* libguile/goops.c (create_smob_classes): Refer to
`SCM_I_MAX_SMOB_TYPE_COUNT' rather than 255 (which is wrong) or 256.
* libguile/smob.c (MAX_SMOB_COUNT): Alias for `SCM_I_MAX_SMOB_TYPE_COUNT'.
* libguile/smob.h (SCM_I_MAX_SMOB_TYPE_COUNT): New macro.
* module/system/vm/program.scm:
* libguile/programs.h:
* libguile/programs.c (scm_program_bindings, scm_program_bindings)
(scm_program_properties, scm_program_name): Unfortunately, implement
more procs in C, so that C can use them more easily.
* libguile/debug.c (scm_procedure_name): Dispatch to scm_program_name as
appropriate.
* libguile/vm-i-system.c (call): Tick in a call.
* libguile/vm-i-system.c (vector): Don't cons up a list just to make a
vector. Saves a couple percent in total cell allocation when loading
syncase. Probably not worth it, but foo!
* libguile/objcodes.c (scm_bytecode_to_objcode): Check that the length of
the vector matches the length embedded in the bytecode.
* libguile/programs.c (scm_program_meta): Call through to
scm_objcode_meta, instead of looking in the object table. Avoids
consing up a program+objcode slice for the meta until the meta is
actually called.
* libguile/vm-i-loader.c (load-program): Step past the metadata too.
* module/language/glil/compile-assembly.scm (make-meta): Just return the
load-program form, or #f.
(assoc-ref-or-acons, object-index-and-alist, make-object-table): Don't
write the meta into the object table.
(glil->assembly): Instead write the meta into the load-program form.
* libguile/objcodes.c (make_objcode_by_mmap, scm_c_make_objcode_slice):
Verify the lengths with the meta-length.
(scm_objcode_meta): New procedure, for getting at the meta-info of an
objcode.
(scm_objcode_to_bytecode):
(scm_write_objcode): Write bytecode with the metadata too.
* module/system/vm/objcode.scm: Export object-meta.
* module/language/assembly.scm (byte-length):
* module/language/assembly/compile-bytecode.scm (write-bytecode):
* module/language/assembly/decompile-bytecode.scm (decode-load-program):
* module/language/assembly/disassemble.scm (disassemble-load-program):
* module/language/glil/compile-assembly.scm (glil->assembly):
* test-suite/tests/asm-to-bytecode.test ("compiler"): Change to
load-program format to have meta-or-#f instead of meta-length, so that
we can serialize the meta as objcode without a load-program byte. Add a
test for writing out the meta.
* module/system/vm/Makefile.am:
* module/system/vm/conv.scm:
* module/system/vm/disasm.scm: Remove these modules, as their
functionality is now in (language ...).
* libguile/objcodes.h:
* libguile/objcodes.c:
* module/system/vm/objcode.scm: Rename objcode->u8vector to
objcode->bytecode.
* module/system/vm/frame.scm:
* module/language/bytecode/spec.scm: Fix for objcode->bytecode.
* scripts/disassemble:
* testsuite/run-vm-tests.scm: Fix for (system vm disasm) removal.
* module/system/repl/command.scm: Use the right disassembler.
* 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.
Reported by Roland Haeder. The declaration and definition of
scm_pthread_cond_timedwait were using possibly different types for the
third arg.
* THANKS: Added Roland Haeder.
* libguile/threads.h (scm_pthread_cond_timedwait): Use scm_t_timespec
for third arg rather than struct timespec, for consistency with the
function implementation.
* 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.
For each thread that goes into Guile mode, Guile pushes a cleanup
function, scm_leave_guile_cleanup, whose purpose is to execute
`scm_leave_guile ()' if the thread is terminated while in Guile mode.
The problem is that there are various places - like
scm_pthread_cond_wait, scm_without_guile and scm_std_select - where
the thread temporarily leaves Guile mode (which means unlocking the
heap mutex), and the cleanup function is still in place. Therefore if
the thread is terminated at these places, the cleanup function ends up
trying to unlock a mutex (the heap mutex) which isn't actually locked.
* libguile/threads.h (scm_i_thread): New heap_mutex_locked_by_self field.
* libguile/threads.c (scm_enter_guile): Set heap_mutex_locked_by_self.
(scm_leave_guile): Only unlock if heap_mutex_locked_by_self is 1.
(guilify_self_1): Initialize heap_mutex_locked_by_self.
(scm_i_thread_sleep_for_gc): Remove incorrect use of t->held_mutex
here.
* libguile/vm.h (struct scm_vm_cont):
* libguile/vm.c (capture_vm_cont, reinstate_vm_cont): Change so we just
store the registers as they are, with the reloc.
(vm_cont_mark): Only mark the stack if it has elements on it, otherwise
we get a bogus fp.
* libguile/stacks.c (scm_make_stack): Update for change to vm
continuations.
Because of how Guile saves and restores continuations (by copying the
stack), and how it uses alloca to create space for debug information
on the stack, we must have an alloca() that really does use the stack,
and not one that uses the heap.
To do this, we use the Gnulib "alloca-opt" module instead of "alloca".
This commit also updates the Gnulib sources from the current Gnulib
Git repository.
* benchmark/measure.scm (measure): Fix bitrot.
* libguile/vm.c (VM_ENABLE_STACK_NULLING): Undefine this, as it hasn't
caught any errors in quite a while.
