* module/language/cps/dfg.scm (reverse-post-order): Fold-all-conts is
now a required arg.
(analyze-control-flow): Reverse CFA adds forward-reachable
continuations to the numbering.
* module/language/cps/dfg.scm (constant-needs-allocation?): Use of a
constant in a $values expression of any arity does not cause slot
allocation.
* module/language/cps/compile-bytecode.scm (compile-fun): Allow $values
with a constant value to be compiled in test context. Really we
should fold these in a previous pass!
* module/ice-9/psyntax.scm (binding-type): Update the header comment
to mention the new 'ellipsis' binding type.
(macros-only-env): Preserve ellipsis bindings.
(ellipsis?): Add 'r' and 'mod' as arguments. Search the lexical
environment for an ellipsis binding, and use it.
(gen-syntax): Adapt to the additional arguments of 'ellipsis?'.
(with-ellipsis): New core syntax.
(convert-pattern): Add unary 'ellipsis?' procedure as an argument.
(gen-clause): Adapt to the additional arguments of 'ellipsis?'.
Pass unary 'ellipsis?' procedure to 'convert-pattern'.
(syntax-case): Adapt to the additional arguments of 'ellipsis?'.
(syntax-local-binding): Support new 'ellipsis' binding type.
(syntax-rules): Add support for a custom ellipsis identifier as
the first operand, as per R7RS. Collect common code within new
local procedure 'expand-syntax-rules'.
* module/ice-9/psyntax-pp.scm: Regenerate.
* module/ice-9/local-eval.scm (analyze-identifiers): Add support for
'ellipsis' binding type.
* doc/ref/api-macros.texi (Syntax Rules): Add docs for R7RS custom
ellipsis syntax. Use @dots{}.
(Syntax Case): Add docs for 'with-ellipsis'. Use @dots{}.
(Syntax Transformer Helpers): Update to include new 'ellipsis'
binding type.
* test-suite/tests/syntax.test: Add tests.
* module/ice-9/psyntax.scm (expand-top-sequence): When defining a
toplevel variable, use the value of the same-named imported
_variable_, if any. Do _not_ use an existing syntactic binding.
* module/ice-9/psyntax-pp.scm: Regenerate.
Reported by Jordy Dickinson <jordy.dickinson@gmail.com>.
Fixes <http://bugs.gnu.org/15691>.
* module/system/vm/traps.scm (frame-matcher): Extract procedure when
proc is an applicable struct.
Fixes <http://bugs.gnu.org/15533>.
* module/language/tree-il/peval.scm (peval): Final list argument to
`apply' should not be inlined if it is mutable.
* test-suite/tests/peval.test ("partial evaluation"): Add test.
* meta/guile.m4: GUILE_PROGS now takes an optional argument to specify a
required Guile version. By default, it requires Guile >= 2.0. A micro
version can also be specified (e.g. GUILE_PROGS([2.0.10])).
Signed-off-by: Ludovic Courtès <ludo@gnu.org>
Reported by taylanbayirli@gmail.com (Taylan Ulrich B.).
* module/ice-9/psyntax.scm (cvt*): Use 'syntax-case' to destructure
the pattern tail, instead of 'pair?', 'car', and 'cdr'.
(gen-clause): When checking for errors, check for misplaced ellipsis
before duplicate pattern variables, to improve the error message in
case of multiple misplaced ellipses.
* module/ice-9/psyntax-pp.scm: Regenerate.
* test-suite/tests/syntax.test: Add tests.
* module/system/vm/linker.scm (relocate-section-header):
(write-linker-object): Sections that are not loadable should not have
their sh_addr fields set. Fix.
* module/language/cps/slot-allocation.scm (allocate-slots): Don't
allocate slots to unused results of function calls. This can allow us
to avoid consing a rest list for call-with-values with an ignored rest
parameter, and can improve the parallel move code.
* module/language/cps/compile-bytecode.scm (compile-fun): Adapt to avoid
emitting bind-rest in values context if the rest arg is unused.
* doc/ref/api-macros.texi (Macro Expansion): New section.
* doc/ref/api-procedures.texi (Compiled Procedures): Beginnings of a
revision. Not finished.
* doc/ref/compiler.texi (Compiling to the Virtual Machine): Beginnings
of a revision. CPS and bytecode are not done yet.
* lib/jit_x86-cpu.c: Use the emms instruction before
calling any function. This is particularly important
when using c99 complex functions as it can easily
overflow the x87 stack due to the way lightning uses
the x87 stack as a flat register file.
