ObjFW  History of src/macros.h of merge-in:ea967a4

This should be more efficient on some architectures.

As a nice side effect, this is also what Linux does in some userspace
APIs, so after this change, these macros can be used for this as well
(and Linux interestingly does not provide similar macros in the
userspace headers). file: [3c8f4714cc] check-in: [96f946d7ad] user: js, branch: trunk, size: 25972 [annotate] [blame] [check-ins using] [diff]

Changing the alignment of OFVector4D globally (as was done previously)
would have technically been an ABI break in extremely rare cases.
However, since we only need the alignment for the methods added after
1.0, it's better to only have the alignment there and get back to full
ABI compatibility with 1.0. file: [d86f5f8b91] check-in: [1af54eb2c2] user: js, branch: trunk, size: 19548 [annotate] [blame] [check-ins using] [diff]

Now that we use SSE in some inline assembly, it needs to be at least 16.
However, _Alignof(max_align_t) can be just 8. file: [1147a05908] check-in: [c80075f6e4] user: js, branch: trunk, size: 19543 [annotate] [blame] [check-ins using] [diff]

This requires the vectors to be 16 byte aligned. In order to achieve
this, the OFVector4D type is changed to have an alignment of 16 bytes.
However, this does *not* break ABI because the only method actually
requiring 16 byte alignment is -[transformVectors:count:], which was not
in ObjFW 1.0. Hence binaries compiled for ObjFW 1.0 have no 16 byte
alignment for OFVector4D, but also cannot ever call into any code that
needs it. (-[transformedVector:] calls into -[transformVectors:count:],
but creates a properly aligned copy that it passes.) file: [c4d1f32dd0] check-in: [7f304f573b] user: js, branch: trunk, size: 19412 [annotate] [blame] [check-ins using] [diff]

Nothing breaks, but just in case we ever have an Objective-C compiler
that is not GCC-compatible. file: [8e5d433333] check-in: [f2891af000] user: js, branch: trunk, size: 19157 [annotate] [blame] [check-ins using] [diff]

It's just a less awkward name and also what the BSDs use. file: [4191bc0c4c] check-in: [b7a4e61b88] user: js, branch: trunk, size: 18471 [annotate] [blame] [check-ins using] [diff]

This uses the same format as OFEnsure(). file: [07933aec14] check-in: [c86b8f10f0] user: js, branch: trunk, size: 18478 [annotate] [blame] [check-ins using] [diff]

GCC now supports the attribute, but still complains if the method is not
implemented, so always implement it with a method that throws. file: [60a41b82a1] check-in: [e53c0160a8] user: js, branch: trunk, size: 18304 [annotate] [blame] [check-ins using] [diff]

All necessary changes to adjust for the changes made in trunk are
included in the merge commit. file: [e163f96b10] check-in: [3c88df0ce4] user: js, branch: amiga-library, size: 18267 [annotate] [blame] [check-ins using] [diff]

Duplicating defines for every CPU architecture is not very useful. file: [365b468ef6] check-in: [c859e2230c] user: js, branch: trunk, size: 21137 [annotate] [blame] [check-ins using] [diff]

The internal hash is still 32 bit in most places, but this way, it is at
least not baked into the API and ABI and can be upgraded later, should
that ever be necessary. file: [43e7dc4d8f] check-in: [aeb403a1ed] user: js, branch: trunk, size: 21501 [annotate] [blame] [check-ins using] [diff]

__COUNTER__ does not exist in GCC 4.2, and Apple GCC 4.2 is still the
newest compiler available for macOS 10.5. file: [0f3365fd01] check-in: [ef614a225d] user: js, branch: trunk, size: 21485 [annotate] [blame] [check-ins using] [diff]

This makes old versions of Clang happy which don't know about
-Wobjc-designated-initializers. file: [98097ba1eb] check-in: [94c7e3e8f3] user: js, branch: tagged-pointers, size: 21490 [annotate] [blame] [check-ins using] [diff]

When arc4random() is unavailable, either random() or rand() is used and
both need to be seeded. If of_random() is a macro, it needs to be
(re)seeded every time, as it's unknown whether it has already been
seeded. As it is seeded with gettimeofday() due to the lack of a better
initial seed, this means every call returns the first state for the
current time, which is very predictable. random() and rand() are both
not cryptographic, but this should at least make it a little bit better
now. file: [41e131467f] check-in: [c1fe4b2b77] user: js, branch: trunk, size: 21483 [annotate] [blame] [check-ins using] [diff]

This first needs to be enabled in Clang for the ObjFW runtime. file: [3627a6abf3] check-in: [96b35d19c5] user: js, branch: trunk, size: 22002 [annotate] [blame] [check-ins using] [diff]

These are based on runtime/amiga-{library,glue,end}.m. file: [f9a566d40b] check-in: [f6ab72da6e] user: js, branch: amiga-library, size: 21962 [annotate] [blame] [check-ins using] [diff]

This allows ABI stability with the fragile ABI. file: [fff954ce5f] check-in: [496695d778] user: js, branch: trunk, size: 21194 [annotate] [blame] [check-ins using] [diff]

This is more compatible with the various conventions used on various
systems. file: [e1abebe584] check-in: [79fe29dbf4] user: js, branch: trunk, size: 20869 [annotate] [blame] [check-ins using] [diff]

This does *NOT* provide cryptographically secure randomness! file: [3f9a87cf9e] check-in: [a2a06d372a] user: js, branch: trunk, size: 20670 [annotate] [blame] [check-ins using] [diff]

These were way too simplistic, as paths differ in more than just that on
various operating systems. file: [9420a38fbf] check-in: [156b850131] user: js, branch: trunk, size: 20135 [annotate] [blame] [check-ins using] [diff]

This also renames ObjFW-RT to ObjFW_RT, as hyphens should not be in
framework names. file: [6b06f3d8b7] check-in: [0f42b0eafd] user: js, branch: trunk, size: 20442 [annotate] [blame] [check-ins using] [diff]

GCC does not support it. file: [58ba402fca] check-in: [f12043ee28] user: js, branch: trunk, size: 20087 [annotate] [blame] [check-ins using] [diff]

This improves Swift compatibility. file: [059976bd73] check-in: [a2e849c681] user: js, branch: trunk, size: 19936 [annotate] [blame] [check-ins using] [diff]

This also adds -[OFURL fileSystemRepresentation] to correctly convert a
URL back to a path in the local file system. file: [40aa2d8e24] check-in: [b83d1414b1] user: js, branch: trunk, size: 19709 [annotate] [blame] [check-ins using] [diff]

__GCC_VERSION__ is reserved and could be used by GCC in the future. file: [62a0be4bea] check-in: [d6fa7f9bf6] user: js, branch: 0.90, size: 16542 [annotate] [blame] [check-ins using] [diff]

__GCC_VERSION__ is reserved and could be used by GCC in the future. file: [52d18b1a04] check-in: [60214356c2] user: js, branch: trunk, size: 19871 [annotate] [blame] [check-ins using] [diff]

Casts are now written like types in variable declarations. file: [4c5e15ca2a] check-in: [4af49a13c3] user: js, branch: trunk, size: 15609 [annotate] [blame] [check-ins using] [diff]

Forgot to add 2017, even though I already did quite some changes in
2017. file: [7f5d9116f6] check-in: [44f45c2e35] user: js, branch: trunk, size: 14992 [annotate] [blame] [check-ins using] [diff]

For many of those, it's not actually required to declare them dllexport.
Declaring them at all is enough to prevent Clang from using its implicit
declaration that is dllimport. file: [e92edaa214] check-in: [f57765b5c6] user: js, branch: trunk, size: 14986 [annotate] [blame] [check-ins using] [diff]

These are independent of the locale and work on the ASCII character set.

Unlike the C ones, these are 8-bit safe, meaning if a character > 0x7F
is passed, is{alpha,alnum} returns false and to{upper,lower} returns the
original character. file: [49ab80b017] check-in: [d9eb7b50b3] user: js, branch: trunk, size: 15181 [annotate] [blame] [check-ins using] [diff]

This is for the ObjFW runtime - for the Apple runtime, this did already
exist. file: [a0cafa1d2d] check-in: [d9c9a3d0bc] user: js, branch: trunk, size: 14052 [annotate] [blame] [check-ins using] [diff]

Multiline preprocessor directives now always indent the continued lines
by 4 spaces. The only exception are defines that define to code blocks:
Those are still indented by a tab in order to align with the rest of the
code. file: [96b0f326a5] check-in: [543eb96c02] user: js, branch: trunk, size: 13651 [annotate] [blame] [check-ins using] [diff]

Newer versions of Clang have these as built-in defines, even when ARC is
disabled. file: [c42b7644a0] check-in: [ea5a19118a] user: js, branch: 0.8, size: 15108 [annotate] [blame] [check-ins using] [diff]

This was not limited to Android, but also happens with newer versions of
Clang on all platforms. file: [207b43dc6a] check-in: [826c9570d3] user: js, branch: trunk, size: 13613 [annotate] [blame] [check-ins using] [diff]

This has only been tested on Android so far. file: [f5907cb050] check-in: [7e6bad326c] user: js, branch: trunk, size: 13683 [annotate] [blame] [check-ins using] [diff]

It seems Clang has these as built-in defines when cross-compiling for
Android, even when ARC is disabled. file: [9f039167db] check-in: [3d11f9ee8f] user: js, branch: trunk, size: 13659 [annotate] [blame] [check-ins using] [diff]

While at it, also update the mail address. file: [e7b0f05f6d] check-in: [cec0f072f8] user: js, branch: 0.8, size: 14928 [annotate] [blame] [check-ins using] [diff]

While at it, also update the mail address. file: [3aac42d28b] check-in: [2a27cf3000] user: js, branch: trunk, size: 13409 [annotate] [blame] [check-ins using] [diff]

Now that we require GCC >= 4.6, those features are always available. file: [bcbd739063] check-in: [be8db49635] user: js, branch: trunk, size: 13405 [annotate] [blame] [check-ins using] [diff]

This increases the required GCC version from 4.0 to 4.6 (exception:
Apple GCC, which already supports this with >= 4.0 starting with OS X
10.5). Since even GCC 4.6 is really old by now, there is no point in
still supporting something even older and making the code ugly because
of that. While some hardware and OS support was dropped from GCC 4.6
compared to GCC 4.0, there is nothing in there that would be an
interesting target with the exception of BeOS maybe - but a port to BeOS
can also be achieved using the Haiku support. The other dropped OSes are
mostly old versions of OSes while newer ones are still being supported
(and those newer versions of those OSes still support the same
hardware). file: [d6041f701d] check-in: [48980f2297] user: js, branch: trunk, size: 13786 [annotate] [blame] [check-ins using] [diff]

Now that Clang switched to use _Nullable and _Nonnull instead of
__nullable / __nonnull, there is no longer a conflict with glibc, which
means we can just define _Nullable / _Nonnull to nothing if they are not
understood by the compiler (which did not work with __nullable /
__nonnull due to this conflict).

This also defines _Null_unspecified to nothing if unavailable. file: [9c39f051f0] check-in: [d526d938d7] user: js, branch: trunk, size: 14074 [annotate] [blame] [check-ins using] [diff]

This is necessary as the Wii does not allow picking a random free port,
and thus we need to track which ports are used. file: [a7d0bdefbc] check-in: [ed4549ddd3] user: js, branch: trunk, size: 15283 [annotate] [blame] [check-ins using] [diff]

Same for _Nonnull / __nonnull.

Apple implemented it as __nullable / __nonnull in their version of Clang
and sent that for upstreaming. My inital comments about this being a
potential problem and to better use _Nullable / _Nonnull had been
rejected as being too late, so I assumed it stayed that way. But
apparently, it has been changed to _Nullable / _Nonnull before the 3.7.0
release. file: [a5359bc24a] check-in: [2fb22c605b] user: js, branch: trunk, size: 14924 [annotate] [blame] [check-ins using] [diff]

This is necessary to make GCC (>= 4.6 and Apple GCC) happy as they don't
like the list of property attributes ending in a comma. file: [67aef400e2] check-in: [80e1a9655b] user: js, branch: trunk, size: 14737 [annotate] [blame] [check-ins using] [diff]

Using __nonnull directly doesn't work on systems using glibc, as glibc
defines __nonnull as a parameterized define. While this does not fix the
problem of Clang introducing __nonnull even though it conflicts with
glibc, this at least means it's possible again to compile things with
versions of Clang that don't support __nonnull on systems with glibc. file: [6f2726fb11] check-in: [b7097a67b6] user: js, branch: trunk, size: 14648 [annotate] [blame] [check-ins using] [diff]

OF_NO_RETURN_FUNC is defined when __attribute__((__noreturn__)) is
supported for functions, whereas OF_NO_RETURN is only defined when it is
supported for both functions and methods. file: [d141131552] check-in: [ffb39d1df6] user: js, branch: trunk, size: 13962 [annotate] [blame] [check-ins using] [diff]

It causes just too much trouble: It breaks with some old versions of
Clang, where noreturn does not work correctly, and OS X headers break if
noreturn is defined, requiring an ugly and fragile workaround. It's just
not worth the trouble it causes, as the same functionality is available
through __attribute__((__noreturn__)). file: [d03270f705] check-in: [716b1fc2b3] user: js, branch: trunk, size: 13788 [annotate] [blame] [check-ins using] [diff]

This means that OFObject.h imports macros.h now, making it unnecessary
to manually import macros.h in almost every file. And while at it, also
import autorelease.h in OFObject.h, so that this doesn't need to be
manually imported in almost every file as well. file: [605fc19a18] check-in: [13ee56edf3] user: js, branch: trunk, size: 14324 [annotate] [blame] [check-ins using] [diff]

static_asserts can be used in places where asserts won't work and take
two parameters, while asserts only take one. file: [9ad949e286] check-in: [61369b9d26] user: js, branch: trunk, size: 10165 [annotate] [blame] [check-ins using] [diff]

Not using -std=gnu11 means _GNU_SOURCE does not get defined anymore,
therefore this commit also adds the required feature defines for glibc.

Additionally, this adds of_strdup in macros.h, as strdup is an
extension. file: [3858acab50] check-in: [2f5af58573] user: js, branch: trunk, size: 10044 [annotate] [blame] [check-ins using] [diff]

When the compiler supports C11, _Alignof(max_align_t) is used; otherwise
the __BIGGEST_ALIGNMENT__ define is tried with a fallback to 16. file: [d51c5fe23d] check-in: [fdfb2be253] user: js, branch: trunk, size: 9805 [annotate] [blame] [check-ins using] [diff]

The initial reason to do this in macros.h was that it would be
theoretically possible to do a universal build targeting OS X and iOS,
but in practice, they are too different as they use a different SDK,
therefore this does not make much sense. file: [abc861d5c4] check-in: [8602025532] user: js, branch: trunk, size: 9431 [annotate] [blame] [check-ins using] [diff]

The port was much easier than expected. The only changes that were
required were defining the missing lrint() and useconds_t, the rest was
optional (like using backslashes for paths, even though slashes work
fine). file: [dcc14d0966] check-in: [a0b63a8df4] user: js, branch: trunk, size: 9541 [annotate] [blame] [check-ins using] [diff]

The reason is that the symbol nanosleep is available now on Win32, but
is part of the pthreads wrapper, which we don't use. file: [1065db52e2] check-in: [bd2389b6b8] user: js, branch: trunk, size: 9512 [annotate] [blame] [check-ins using] [diff]

Currently, no methods returning structs are supported, see the comment
in OFObject. file: [1a66d76e58] check-in: [27daad1290] user: js, branch: trunk, size: 9563 [annotate] [blame] [check-ins using] [diff]

Checking at configure time can potentially break universal builds. file: [efb84c3744] check-in: [bbabe17c4f] user: js, branch: trunk, size: 9462 [annotate] [blame] [check-ins using] [diff]

This makes the map table use the full range of the hash and not only the
least significant bits, making it harder for an attacker to find
collisions. file: [19f9d4c0fa] check-in: [7cebeaa808] user: js, branch: trunk, size: 9163 [annotate] [blame] [check-ins using] [diff]

By rotating the hash by a random number of bits, an attacker needs to
find collisions on the full 32 bits of the hash and not only on the
lower n bits that are actually used by the map table, as an attacker
can't know which bits are actually used for the map table. file: [6b2bc955be] check-in: [4857107479] user: js, branch: trunk, size: 9039 [annotate] [blame] [check-ins using] [diff]

This prevents DoSing hashtables by creating conflicts deliberately. file: [419a2f32d2] check-in: [f60e4012b7] user: js, branch: trunk, size: 8880 [annotate] [blame] [check-ins using] [diff]

This makes it easier to use runtime functions in portable code. file: [f956fa931a] check-in: [120caad331] user: js, branch: trunk, size: 8869 [annotate] [blame] [check-ins using] [diff]

This makes it clear that multiple evaluation of parameters is possible. file: [1ed7809f51] check-in: [1cb8fee5c3] user: js, branch: trunk, size: 8945 [annotate] [blame] [check-ins using] [diff]

This is to be used when the condition includes a statement that needs to
be executed, as with assert, this would not be executed anymore when
NDEBUG is defined. file: [f51fb05aa2] check-in: [417b213d41] user: js, branch: trunk, size: 8605 [annotate] [blame] [check-ins using] [diff]

* runtime: (76 commits)
runtime: Add exception handling.
... file: [cf167a3135] check-in: [4c4fdb3429] user: js, branch: trunk, size: 8381 [annotate] [blame] [check-ins using] [diff]

No more abusing a poor BOOL... file: [ef3c33f2fd] check-in: [d989794cc2] user: js, branch: trunk, size: 8729 [annotate] [blame] [check-ins using] [diff]

Also removes unnecessary OF_HASH_ADD_INT* macros (except
OF_HASH_ADD_INT32 which was often used to hash hashes and is now called
OF_HASH_ADD_HASH). file: [fee9f03105] check-in: [f3db074478] user: js, branch: trunk, size: 8601 [annotate] [blame] [check-ins using] [diff]

This way, we also don't need -std=gnu99 in objfw-config anymore.

Fixes ObjC++, for which -std=gnu99 is invalid and where there is no
restrict. file: [7481fee224] check-in: [77fe5572fe] user: js, branch: trunk, size: 8727 [annotate] [blame] [check-ins using] [diff]

Does compile with --disable-threads --disable-shared, but does not link
yet. file: [957228cdfe] check-in: [d230ac577d] user: js, branch: trunk, size: 5698 [annotate] [blame] [check-ins using] [diff]