If we inline a function which contains keyword arguments whose default
value depends on the presence of other keyword arguments as for
example in
(funcall (lambda (&key (a nil a-p) (b a-p)) ...))
where `b` depends on the presence of `a`, we need to set the
key-flag `a-p` immediately after scanning for the keyword and not at
the end after we have finished scanning all keywords as we did previously.
This reverts commit c5cb09f3cb and
84acbebd7f.
The solution as implemented does not work because we there is no clear
separation between assigning vv locations and later on writing them to
the created C file; we may have already written a vv location before
trying to add the same object again, resulting in miscompilation.
Precompiled headers may not work in every scenario (for example
compilation currently fails for the --with-cxx=yes configure option
due to precompiled headers). If we disable them by default, we are on
the safe side.
We need to compile the function body in the same environment in which
the function was defined. However the function arguments need to be
compiled in the current argument.
Fixes#577.
Unify handling of LAMBDA and LAMBDA-BLOCK in c1funcall and c1apply,
split off computation of let bindings and body in
optimize-funcall/apply-lambda into a separate function.
Preliminary work to fix inlining of local closures.
That makes lambda with a declaration si:function-block-name behave
consistently with ext:lambda-block (and in eval-macros
ext:lambda-block expands to have this declaration too to behave in
turn consistently with how the compiler treats ext:lambda-block).
We need to save env->nvalues before calling cl_set on any variable
because cl_set overwrites env->nvalues. Otherwise, we only get nil for
any variable after the first special one.
Fixes#591.
Improves compilation speed for single functions by about 40-50
percent. Precompiled headers are specific to the compiler version and
options in use. Due to this, we regenerate the header whenever the
compiler configuration changes.
Slot definitions are no longer a signature, but they are still needed
to update obsolete instances. Reader function name is also changed to
SI:INSTANCE-SLOTDS. SI:INSTANCE-SIG-SET name does not change, because
it sets both SLOTDS and the STAMP.
Function is rewritten in C in compiler.d to remove a dependency
between the bytecodes compiler and the clos module. It may be more
performant thanks to more precise type handing, however we use a list
instead of a hashtable, so it may be slower with lookup. To assess
that we should run some benchmarks against real code -- rewriting C
code to work with a hash table should be trivial.
clos::need-to-make-load-form-p is now si::need-to-make-load-form-p and
may be called from C code as si_need_to_make_load_form_p.
Init forms are deferred when possible. This change solves two problems:
- init forms using uninitialized constant boxes
- make forms not signaling an error when circular
Partial fix for #562 (we need to fix bytecodes compiler too).
When a local function calls a closure it has to be a closure too. Thus
when updating the closure type for a function f, we have to possibly
update also all functions referencing f.
Fixes#545.
According to the C99 standard, compilers are supposed to automatically
choose the correct type. However, for the C89 standard explicit suffixes
are needed for long longs. Guess what standard msvc follows...
Some of the compiler macros expand into FFI:C-INLINE forms, precluding
the compilation of statements like
(eval-when (:compile-toplevel :load-toplevel :execute)
(print "test")).
Commit b00e62f9d3 (now reverted)
attempted to fix this previously, but introduced other bugs (see #553).
Instead of the naive solution of evaluating everything
in (eval-when (:compile-toplevel)) directly with the bytecodes
compiler, the approach used in this commit is to be more specific and
ignore compiler macros during compile time evaluation (i.e. the output
of the compiler macro is used only in the emitted C code, but not as
input for CMP-EVAL).
Fixes#553.
All functions declared with si::c-export-fname must be either in
symbols_list.h or have a proclamation so that the compiler knows the
number of arguments that the function takes.
si::coerce-to-vector assumed that the to be coerced object had the
same length as that specified by the type. This lead to segmentation
faults even in safe code, for example in
(coerce '(a b c) '(vector * 4))
(coerce.error.3 test in ansi-tests)
Actually, si::coerce-to-vector had some checks for a correct length
previously, but they did not work correctly and were removed in commit
baaab01841.
This is a dead code which is not used in the compiler. It was meant
for providing entry points from Common Lisp code to ECL functions
written in C, but it was replaced by more robust machinery.
- normalize return-type from NIL to :void, ARRAY and '* in interpreted
dffi implementation -- it is already normalized in sffi
- remove invalid path where argument type was not a valid elementary
FFI type
when it was not c1-defcallback pushed result of add-object to
arg-type-constants and tried to pass the data as opaque
pointers. That said it could never work, because:
1. add-object could return a string (i.e for known symbols expanding
to ECL_SYM) and they were fed as elementary FFI type leading to
errors during compilation by C compiler (invalid enum type)
2. when ecl_make_foreign_data was called to pass opaque objects a
function FFI:SIZE-OF-FOREIGN-TYPE was called which resulted in
error (because return type is not a valid elementary FFI type
what this code path was meant to be)
Moreover we validate both return type and argument types during the
first compiler to fail as early as possible (previously only argument
types were validated early).
- some cosmetic fixes like indentation or redundant PROGN
- move the constant and a function foreign-elt-type-code to the top
- re-align the constant data
- fix the feature reader conditionals (they were misplaced)
- add reader conditionals for complex floats
