* module/language/cps/optimize.scm (cps-optimizations):
* module/language/tree-il/optimize.scm (tree-il-optimizations):
* module/system/base/optimize.scm (available-optimizations): Invert the
dependency tree to hold the names and default optimization levels in a
central place instead of in the optimizers. It moves definitions
farther from uses, but it avoids us having to load the CPS optimizer
if we don't need it, which improves bootstrap times.
* module/language/tree-il/compile-bytecode.scm (compile-closure): Make
it so that for-tail is actually tail-recursive. Likewise improve tail
recursion for the other helpers.
* module/language/tree-il/compile-bytecode.scm (emit-box-set!): Fix to
reference by SCM, not word.
(emit-box-ref): New helper.
(emit-cached-module-box, emit-cached-toplevel-box, emit-toplevel-box):
Add bound? arg. Before these could produce #f instead of a variable,
and unbound variable errors weren't any good as they didn't have the
variable name.
(compile-closure): Use more box-ref and box-set!. Pass bound? arg to
the helpers.
* module/language/cps/reify-primitives.scm (reify-lookup):
* module/language/tree-il/compile-cps.scm (toplevel-box): Instead of
checking that the result of module-variable is a variable, and
possibly checking that it's bound, we just call intrinsics that throw
exceptions if the variable isn't bound. This reduces useless inlining
that can't inform CPS optimizations, as they are tangled in diamond
control flow.
* module/language/tree-il/spec.scm (join): Use match rather than
pmatch.
(tree-il): Declare compiler to bytecode.
(choose-compiler): New implementation. Note, the baseline compiler
probably doesn't even work!!!
* module/system/base/compile.scm (next-pass): Invoke the language's
compiler chooser if there is more than one compiler.
(compute-compiler): Ensure from and to are languages.
* module/system/base/language.scm (<language>): Add compiler-chooser
field.
* module/language/brainfuck/spec.scm (choose-compiler, brainfuck):
Define a compiler chooser.
* module/language/cps/compile-bytecode.scm (compile-bytecode):
* module/language/tree-il/compile-bytecode.scm (compile-bytecode):
* module/language/tree-il/compile-cps.scm (compile-cps): Rely on
compiler to lower incoming term already.
* module/language/tree-il/optimize.scm (make-lowerer): New procedure.
* module/system/base/compile.scm (compute-lowerer): New procedure,
replaceing add-default-optimizations.
(compute-compiler): Lower before running compiler.
* module/system/base/language.scm (<language>): Change
optimizations-for-level field to "lowerer".
* module/scripts/compile.scm (%options, compile): Parse -O0, -O1 and so
on to #:optimization-level instead of expanding to all the
optimization flags.
* module/language/cps/optimize.scm (lower-cps): Move here from
compile-bytecode.scm.
(make-cps-lowerer): New function.
* module/language/cps/spec.scm (cps): Declare lowerer.
* module/language/tree-il/analyze.scm (make-analyzer): Expect an int for
optimization level.
* module/scripts/compile.scm (%options, show-warning-help): No more
-Wnone / Wall; use -W0 or -W9 instead.
* module/system/base/compile.scm (level-validator): Validate small int.
(compute-analyzer, add-default-optimizations): Likewise.
* test-suite/tests/optargs.test (without-compiler-warnings):
* test-suite/tests/tree-il.test (call-with-warnings): Parameterize level
to 0, not #f.
* bootstrap/Makefile.am (GUILE_WARNINGS): Use -W0, not -Wnone.
* module/system/base/compile.scm (compute-analyzer): Compute analyzer to
run on expressions before the compiler runs.
(add-default-optimizations): Flesh out; still a stub.a
(read-and-compile, compile, compile-and-load, compile-file): Default
warning and optimization levels.
(default-warning-level): New parameter, defaulting to 1.
(default-optimization-level): New parameter, defaulting to 2.
Currently unused.
* module/system/base/language.scm (<language>): Add
optimizations-for-level and analyzer fields.
* module/language/tree-il/compile-bytecode.scm (compile-bytecode):
* module/language/tree-il/compile-cps.scm (optimize-tree-il): No need to
run warnings passes here; compilers infrastructure will run them.
* module/language/tree-il/spec.scm (tree-il): Define make-analyzer as
analyzer.
* module/language/tree-il/analyze.scm (make-analyzer): New exported
procedure.
(%warning-passes): New private variable.
* .dir-locals.el: Add with-test-prefix/c&e indent mode.
* test-suite/tests/cross-compilation.test:
* test-suite/tests/optargs.test:
* test-suite/tests/tree-il.test: Adjust to disable default warnings.
* module/language/tree-il/compile-bytecode.scm (canonicalize): Don't add
an extra false? around predicates.
(compile-closure): Fix predicate comparison instructions.
* module/language/tree-il/compile-bytecode.scm (compile-closure): We can
emit the precise move sequence and just do a reset-frame once, so go
ahead and do that.
* module/language/tree-il/compile-bytecode.scm (compile-closure):
Provide names for locals, including the closure. Fix emission of
primitives with immediate args.
* module/language/tree-il/primitives.scm (call-with-prompt): Only pass
"values handlers" as handler: lambdas with only req and rest args, and
only one clause.
* module/language/tree-il/compile-cps.scm (canonicalize): Remove
eta-conversion pass here.
* test-suite/tests/peval.test ("partial evaluation"): Adapt test.
The old name was wonky and had bad argument order.
* NEWS: Add entry.
* doc/ref/api-data.texi (Bit Vectors): Update.
* libguile/bitvectors.h:
* libguile/bitvectors.c (scm_bitvector_position): New function.
* libguile/deprecated.h:
* libguile/deprecated.c (scm_bit_position): Deprecate.
* module/ice-9/sandbox.scm (bitvector-bindings): Replace bit-position
with bitvector-position.
* module/language/cps/intset.scm (bitvector->intset): Use
bitvector-position.
* module/system/vm/frame.scm (available-bindings): Use
bitvector-position.
* test-suite/tests/bitvectors.test ("bitvector-position"): Add test.
* module/language/cps/specialize-numbers.scm (preserve-eq?): New
helper.
(sigbits-union): Use the new helper. Fixes bugs.gnu.org/38486.
Thanks to Zack Marvel for the bug report and Matt Wette for tracking
it down.
Fixes <https://bugs.gnu.org/39509>.
Reported by Klaus Stehle <klaus.stehle@uni-tuebingen.de>.
* module/language/tree-il/primitives.scm (error): Remove extra "?"
argument when the first argument is not a constant.
* test-suite/tests/tree-il.test ("primitives")["error"]: New test
prefix.
* module/language/tree-il/letrectify.scm (compute-procedures-without-identity):
(letrectify): Only eta-expand lambda references that appear outside
the operator position more than once. This should restore peoples'
expectations that (eqv? f f) without penalizing optimization.
* module/language/tree-il/peval.scm (peval): Fix arity check for type
confusion (empty value of "rest" in this context was (), not #f). The
effect was that we'd silently allow extra arguments to inlined calls.
Thanks to Christopher Lam for the report! Fixes#38617.
* test-suite/tests/peval.test ("partial evaluation"): Add a test.
* module/language/cps/closure-conversion.scm (convert-one):
Strongly-connected components of letrec bindings that do not share a
closure may include member functions that have a single free variable,
or even no free variables as a result of free variable pruning.
Handle this case instead of erroring out. Thanks to Stefan Israelsson
Tampe for the report.
* module/language/cps/effects-analysis.scm (&header): New memory kind,
for the fixed parts of objects. Distinguishing init-only memory
allows us to determine that vector-set! doesn't stomple
vector-length.
(annotation->memory-kind*): New helper, mapping references to fixed
offsets to &header. Use for scm-ref/immediate et al.
* module/language/cps/types.scm (div-result-range): It is possible for a
max value to be less than a minimum. In this bug from zig:
(define (benchmark x)
(let loop ((count 0)
(sum 0))
(if (= count 10)
(exact->inexact (/ sum 10)))
(loop (+ count 1) x)))
Here the first iteration gets peeled, and thus the first "if" can't be
true, because "count" is zero. However on the true branch of the if,
range inference produces bogus ranges -- notably, the variable bound
to 10 is inferred to have a min of 10 and a max of 0. This is fine,
because it's unreachable; but that then infects the division, because
the same variable bound to 10 is used there, resulting in division by
zero.
