This allows the groveler to correctly handle cstruct definitions that include arrays

 * build all branches
 * test more implementations (but only SBCL and CCL are mandatory)
 * libffi-dev is now installed by cl-travis
 * disable TEST-STATIC-PROGRAM because it fails due to sbcl.o not being
build with -fPIE
 * print unexpected failures at the end of the test run

It remains compatible with existing code, but accepts additional formats for bindings:
1) (name slot-name) - just like WITH-SLOTS;
2) (name :pointer slot-name) - a pointer version.

This reverts commit b3e6b3f2.

Osicat fails to compile. lp#1922428

Shinmera authored Aug 21, 2020 and Stelian Ionescu committed Oct 15, 2020

This includes size, ssize, ptrdiff, offset, intptr, and uintptr

This reverts commit c5d140fe.

This is useful for very large enumerations of which we're only interested in a subset, or enumerations that allow for arbitrary client or vendor extensions that we cannot be aware of. Erroring on a value that we ostensibly do not care about in such a situation is not desirable, and instead simply passing through the integer allows handling the interesting cases easily in a CASE or something similar.

CFFI now always translates aggregate struct values to lisp property
lists.  Earlier struct values which were defined by types introduced
by defctype were translated as foreign pointers.

One consequence of this change is that defcvar to a foreign struct
will always produce an immutable lisp list.  To modify the struct the
user first has to get at the pointer to the struct via
FORIEGN-SYMBOL-VALUE or GET-VAR-POINTER.  To modify nested structs the
user has to get at the pointer to the field via FORIEGN-SLOT-POINTER,
- or use the the (:POINTER VAR) syntax of WITH-FOREIGN-SLOTS.

* src/early-types.lisp: (defctype): use translatable-foreign-type
instead of enhanced-foreign-type.

* tests/structs.lisp: New tests: (misnamed) STRUCT-VALUES.FSBV.1
STRUCT-VALUES.FSBV.1 Updated tests: STRUCT.ALIGNMENT.1
STRUCT.ALIGNMENT.2 STRUCT.ALIGNMENT.3 STRUCT.ALIGNMENT.4
STRUCT.ALIGNMENT.5 STRUCT.ALIGNMENT.6 STRUCT.ALIGNMENT.7
STRUCT.NESTED-SETF STRUCT.ALIGNMENT.8

Rename test ARRAY.ROUND-TRIP to ARRAY.FOREIGN-TO-LISP.BASIC
Add test ARRAY.FOREIGN-TO-LISP.ADJUSTABLE
Add test ARRAY.FOREIGN-TO-LISP.DISPLACED
Add test ARRAY.FOREIGN-TO-LISP.SPECIALIZED

Implementation detail for test ARRAY.FOREIGN-TO-LISP.SPECIALIZED:

15.1.2.2 of the Holy Standard states that the only truly portable array
specializations are for bits (bit-vectors) and characters (strings). Since
char-codes are implementation-dependent, it would be tricky to write a
portable test for them without generating characters at runtime. So, for a
truly portable test, we are only left with bits, which are luckily numeric,
and equal to (UNSIGNED-BYTE 1).

This is why this test is so terribly wasteful, spending a whole byte for a
single bit - CFFI has no capabilities for dealing with single bits, and this
test is only meant to check correctness of the :ELEMENT-TYPE argument to
MAKE-ARRAY. In actual use cases of specialized FOREIGN-ARRAY-TO-LISP, capable
implementations will be able to make specialized arrays of types that are
commonly optimized for and/or representable in hardware, such as
(UNSIGNED-BYTE 8) on x86 architectures.

Implements lp bug #1740800.

Daniel Kochmański authored May 31, 2017 and Luís Oliveira committed Oct 05, 2017

enum: even if it's just a cosmetic difference between :int and :uint
in ECL for inline calls it does matter - and another-boolean enum is
of type :uint (no negative values).

foreign-alloc.10: foreign-free may return no values, wrap result
in (values #0).

unlock callback.bff.* tests for ECL which doesn't choke on them as was
suggested in the comments.

add defcfun.varargs.nostdlib and funcall.varargs.nostdlib tests. They
use variadic function defined in libtest.c (hence - not declared
anywhere) `sum_double_arbitrary'.

`sprintf' is a good vararg function example, but if
implementation (like ECL) has it the declaration already included, it
may lead some tests to pass which wouldn't otherwise (and vice-versa)
for static ffi.

Daniel Kochmański authored Oct 01, 2017 and Luís Oliveira committed Oct 05, 2017

This function takes variable number of arguments and we have
abstraction for that.

... and avoid catching non-simple-errors coming from
%LOAD-FOREIGN-LIBRARY since those are unexpected and we're
specifically set up to report SIMPLE-ERRORs.

Fixes lp#1720626.

Francois-Rene Rideau authored Dec 21, 2016 and Attila Lendvai committed Jan 29, 2017

Add support for getting parameters from new SBCL sbcl.mk.
Add support for getting parameters from CLISP makevars.
Add support for getting parameters for ECL and MKCL.

Refactor static linking to use the above.
On SBCL, this relies on the new feature :sb-linkable-runtime
(from 1.3.14-20-g402a8fab6).
On ECL and MKCL, have static-{image,program}-op be pass-through synonyms
for the usual variants.
Remove deceitful incomplete support for static linking on CMUCL.

Add shell test for static linking, to better work with SBCL (uses cl-launch).

Add class c-file for C files to compile, and o-file for precompiled objects.
They can be either dynamically loaded into the current process,
or statically linked into a new runtime.

Upgrade backward compatibility bundle from asdf 3.1.6 to 3.2.0.
Stop manually curating a minimal set of changes,
just adopt a new bundle.lisp wholesale.

E.g., ONE is defined as a constant in a DEFBITFIELD so it can't be used
as an argument name in CHECK-ENUMS.

INCOMPATIBLE CHANGE: bitfields in FOREIGN-BITFIELD-SYMBOLS were
treating (field 0) entries as a valid bitfield member. This commit
changes that, the zero'th bit is a mask of 1, so 0 entries will not
treated as bitfields entries (only mere enum entries).

Once fixed, also added follow-typedefs into their accessors.

NOTE: the new ENSURE-PARSED-BASE-TYPE function is more premissive
than PARSE-TYPE because it doesn't error on already parsed types.
This shouldn't be a problem, but there might be some assumptions
i'm not aware of.

In translate-objects-ret for cffi-libffi, dereference the pointer
returned from the foreign function first, then call
translate-objects. This fixes an error reported by Attila Lendvai in
e54b7abb when typedef is used. Test he contributed (now called
fsbv.return-value-typedef) now passes.

Also added a new failing test called fsbv.wfo

The function #'expand-to-foreign-dyn-indirect, used by cffi-libffi, did
not create the correct expansion when there was an enumerated argument,
because two translations are required, one from the symbol to an
integer, and the second to make the integer indirect, passing a pointer
to libffi. This remedies that problem by defining a method for
expand-to-foreign-dyn-indirect specific to foreign-enum that recursively
calls expand-to-foreign-dyn-indirect twice. A test, fsbv.8, has been
created.