* move stage2.cpp code into zig0.cpp for simplicity
* add -ftime-report and some more CLI options to stage2
* stage2 compites the llvm cpu features string
* classifyFileExt understands more file extensions
* correction to generateBuiltinZigSource using the wrong allocator
(thanks dbandstra!)
* stage2 is now able to build hello.zig into hello.o using stage1 as a
library however it fails linking due to missing compiler-rt
* remove dead code
* simplify zig0 builtin.zig source
* fix not resolving builtin.zig source path causing duplicate imports
* fix stage1.h not being valid C code
* fix stage2.h not being valid C code
Deleted 16,000+ lines of c++ code, including:
* an implementation of blake hashing
* the cache hash system
* compiler.cpp
* all the linking code, and everything having to do with building
glibc, musl, and mingw-w64
* much of the stage1 compiler internals got slimmed down since it
now assumes it is always outputting an object file.
More stuff:
* stage1 is now built with a different strategy: we have a tiny
zig0.cpp which is a slimmed down version of what stage1 main.cpp used
to be. Its only purpose is to build stage2 zig code into an object
file, which is then linked by the host build system (cmake) into
stage1. zig0.cpp uses the same C API that stage2 now has access to,
so that stage2 zig code can call into stage1 c++ code.
- stage1.h is
- stage2.h is
- stage1.zig is the main entry point for the Zig/C++
hybrid compiler. It has the functions exported from Zig, called
in C++, and bindings for the functions exported from C++, called
from Zig.
* removed the memory profiling instrumentation from stage1.
Abandon ship!
* Re-added the sections to the README about how to build stage2 and
stage3.
* stage2 now knows as a comptime boolean whether it is being compiled
as part of stage1 or as stage2.
- TODO use this flag to call into stage1 for compiling zig code.
* introduce -fdll-export-fns and -fno-dll-export-fns and clarify
its relationship to link_mode (static/dynamic)
* implement depending on LLVM to detect native target cpu features when
LLVM extensions are enabled and zig lacks CPU feature detection for
that target architecture.
* C importing is broken, will need some stage2 support to function
again.
Without this, building from source caused:
CommandLine Error: Option 'mc-relax-all' registered more than once!
LLVM ERROR: inconsistency in registered CommandLine options
This is due to LLVM static libs compiled in multiple times. But without
the LLVM static libs on the linker line, it caused undefined symbol
linker errors.
So our hands are tied. Homebrew users will have to specify
`-DZIG_PREFER_CLANG_CPP_DYLIB`.
* don't unconditionally pass -lz3 for mingw builds. If mingw builds
require this then the llvm-config executable should put it as part of
--system-libs. If there is a bug and it does not do that, and we need
a workaround, then the workaround should be an explicit cmake option.
* don't link libstage2.a against -lntdll. This causes zig to set
`builtin.link_mode == .Dynamic` and include the TLS definitions,
which then collide with the mingw-w64 symbols. This should probably
be addressed separately, but for now this solves the problem and
there is no reason to link a static library against a DLL.
* Findllvm.cmake no longer treats the libraries as "optional" and will
emit a cmake error if one is not found. Additionally, the
not-required LLVM library LLVMTableGen is omitted.
Make fallthrough an error when compiler supports it. This requires a new
macro that is defined with such compilers to be used as a statement, at
all fallthrough sites:
switch (...) {
case 0:
...
ZIG_FALLTHROUGH;
case 1:
...
break;
default:
...
break;
}
If we ever move to C++17 as minimal requirement, then the macro can be
replaced with `[[fallthrough]];` at statement sites.
When the build.zig logic to build libzigstage2 was converted to a cmake
command, it neglected to use baseline CPU features rather than compiling
with native features.
This adds a hard coded flag `-mcpu=baseline` which can be used to detect
the native target, but mark it as non-native so that it does not get the
CPU features specific to the host used to compile libzigstage2.
Full `-mcpu` support is happening in the upcoming pull request #4509,
and so this "quick fix" will be cleaned up in that branch, before it is
merged to master.
closes#4506
when both `--cache on` and `--output-dir` parameters
are provided. This prevents re-linking `zig` with every
`make` even when `libzigstage2.a` was unchanged.
Rather than `zig0 build ...` the build now does
`zig0 build-lib ...`, avoiding the requirement of linking the build
script, and thus avoiding the requirement of finding native libc,
for systems where libc is the system ABI.
* libc_installation.cpp is deleted.
src-self-hosted/libc_installation.zig is now used for both stage1 and
stage2 compilers.
* (breaking) move `std.fs.File.access` to `std.fs.Dir.access`. The API
now encourages use with an open directory handle.
* Add `std.os.faccessat` and related functions.
* Deprecate the "C" suffix naming convention for null-terminated
parameters. "C" should be used when it is related to libc. However
null-terminated parameters often have to do with the native system
ABI rather than libc. "Z" suffix is the new convention. For example,
`std.os.openC` is deprecated in favor of `std.os.openZ`.
* Add `std.mem.dupeZ` for using an allocator to copy memory and add a
null terminator.
* Remove dead struct field `std.ChildProcess.llnode`.
* Introduce `std.event.Batch`. This API allows expressing concurrency
without forcing code to be async. It requires no Allocator and does
not introduce any failure conditions. However it is not thread-safe.
* There is now an ongoing experiment to transition away from
`std.event.Group` in favor of `std.event.Batch`.
* `std.os.execvpeC` calls `getenvZ` rather than `getenv`. This is
slightly more efficient on most systems, and works around a
limitation of `getenv` lack of integration with libc.
* (breaking) `std.os.AccessError` gains `FileBusy`, `SymLinkLoop`, and
`ReadOnlyFileSystem`. Previously these error codes were all reported
as `PermissionDenied`.
* Add `std.Target.isDragonFlyBSD`.
* stage2: access to the windows_sdk functions is done with a manually
maintained .zig binding file instead of `@cImport`.
* Update src-self-hosted/libc_installation.zig with all the
improvements that stage1 has seen to src/libc_installation.cpp until
now. In addition, it now takes advantage of Batch so that evented I/O
mode takes advantage of concurrency, but it still works in blocking
I/O mode, which is how it is used in stage1.
See discussion here for context:
c6df5deb34 (comments)
Michael - I appreciate what you did here, making the configure script
work better for people in practice. When it was checking the build type
against a whitelist, I think it was worth it. However, now that we are
supporting systems which use non-standard cmake build modes, I don't
think this case-mismatch detection thing is worth it. It's starting to
get to the point where it's a lot of complication for very little
benefit. Besides, cmake is not case sensitive. If we support
non-standard build modes, then we would need to support a hypothetical
build mode of `release` (lower case).
So let's just remove this and rely on people to use the build system
correctly (like they will have to do when building any cmake project
from source).
Various maintainers pass custom build types and we don't need to check
those. We are interested only in checking and diagnosing common errors
for Zig project supported types.
Check is now limited to look for case-mismatch only on the well-known
values { Debug, Release, RelWithDebInfo, MinSizeRel }.
- split util_base.hpp from util.hpp
- new namespaces: `mem` and `heap`
- new `mem::Allocator` interface
- new `heap::CAllocator` impl with global `heap::c_allocator`
- new `heap::ArenaAllocator` impl
- new `mem::TypeInfo` extracts names without RTTI
- name extraction is enabled w/ ZIG_ENABLE_MEM_PROFILE=1
- new `mem::List` takes explicit `Allocator&` parameter
- new `mem::HashMap` takes explicit `Allocator&` parameter
- add Codegen.pass1_arena and use for all `ZigValue` allocs
- deinit Codegen.pass1_arena early in `zig_llvm_emit_output()`
- zig CMakeLists.txt CMAKE_BUILD_TYPE string comparisons are case-sensitive
- cmake itself is unclear about how tolerant it is for case-mismatches
- add CMAKE_BUILD_TYPE guard in CMakeLists.txt to force exact matches
- `make` or `ninja` will not build but not install
- `make install` or `ninja install` will build __and__ install
Only for build system generator Visual Studio, specify the following
to disable installation of lib files:
ZIG_SKIP_INSTALL_LIB_FILES=ON
During build an empty .o on macOS/Xcode emits warning:
ranlib: file: zig_cpp/libcompiler.a(memory_profiling.cpp.o) has no symbols
ranlib: file: zig_cpp/libcompiler.a(memory_profiling.cpp.o) has no symbols
This commit adds -fdump-analysis which creates
a `$NAME-analysis.json` file with all of the finished
semantic analysis that the stage1 compiler produced.
It contains types, packages, declarations, and files.
This is an initial implementation; some data will be
missing. However it's easy to improve the implementation,
which is in `src/dump_analysis.cpp`.
The next step for #21 will be to create Zig code which parses
this json file and creates user-facing HTML documentation.
This feature has other uses, however; for example, it could
be used for IDE integration features until the self-hosted
compiler is available.
This moves the installation of shipped source files from large
CMakeLists.txt lists to zig build recursive directory installation.
On my computer a cmake `make install` takes 2.4 seconds even when it has
to do nothing, and prints a lot of unnecessary lines to stdout that say
"up-to-date: [some file it is installing]".
After this commit, the default output of `make` is down to 1
second, and it does not print any junk to stdout. Further, a `make
install` is no longer required and `make` is sufficient.
This closes#2874.
It also closes#2585. `make` now always invokes `zig build` for
installing files and libuserland.a, and zig's own caching system makes
that go fast.
This is the beginning of supporting minimum GLIBC version as part of the
target. See #2509 for the motivation.
The dummy libc zig files are removed. A future commit will build them
on-the-fly, using the generated text files generated by the new tool,
which are checked into source control and distributed along with zig.
These generated text files are, together, 142KB (20KB gzipped).
Compare that to a naive bundling of the .abilist files, which would be
2.2MiB (375KB gzipped).
This is based on glibc 2.29.
Not all of the softfloat library is being built....
Vector support is very buggy at the moment, but should work when the bugs are fixed.
(as I had the same code working with another vector function, that hasn't been merged yet).
* introduce wasm32-freestanding-musl .h files to fix
conflicts with stddef.h and errno.h
* fix an issue with zig build system regarding installation of
webassembly libraries
* add implementations to zig's libc:
- strcmp
- strncmp
- strerror
- strlen
See #514
- wip for #2046
- clang .d output must be created with `clang -MV` switch
- implemented in Zig
- hybridized for zig stage0 and stage1
- zig test src-self-hosted/dep_tokenizer.zig
* rename std/special/builtin.zig to std/special/c.zig
not to be confused with @import("builtin") which is entirely
different, this is zig's multi-target libc implementation.
* WebAssembly: build-exe is for executables which have a main().
build-lib is for building libraries of functions to use from,
for example, a web browser environment.
- for now pass --export-all for libraries when there are any
C objects because we have no way to detect the list of exports
when compiling C code.
- stop passing --no-entry for executables. if you want --no-entry
then use build-lib.
* make the "musl" ABI the default ABI for wasm32-freestanding.
* zig provides libc for wasm32-freestanding-musl.
Previously, `zig fmt` on the stage1 compiler (which is what we currently
ship) would perform what equates to `zig run std/special/fmt_runner.zig`
Now, `zig fmt` is implemented with the hybrid zig/C++ strategy outlined
by #1964.
This means Zig no longer has to ship some of the stage2 .zig files, and
there is no longer a delay when running `zig fmt` for the first time.
On macOS building with Xcode/clang the linker complains loudly when
symbol visibility is inconsistent. This option syncs visibilty setting
of both LLVM and Zig.
Previously libuserland was being built for the native target,
which could depend on native CPU features such as AVX. The
CI infrastructure is intending to create binaries that are
widely compatible and do not make use of specific CPU features.
There could be something to gain from enabling native CPU features
sometimes, by introducing a new cmake configuration option, but
since it's planned to eventually ship self-hosted rather than stage1,
I don't think it really matters.
closes#2348
This modifies the build process of Zig to put all of the source files
into libcompiler.a, except main.cpp and userland.cpp.
Next, the build process links main.cpp, userland.cpp, and libcompiler.a
into zig1. userland.cpp is a shim for functions that will later be
replaced with self-hosted implementations.
Next, the build process uses zig1 to build src-self-hosted/stage1.zig
into libuserland.a, which does not depend on any of the things that
are shimmed in userland.cpp, such as translate-c.
Finally, the build process re-links main.cpp and libcompiler.a, except
with libuserland.a instead of userland.cpp. Now the shims are replaced
with .zig code. This provides all of the Zig standard library to the
stage1 C++ compiler, and enables us to move certain things to userland,
such as translate-c.
As a proof of concept I have made the `zig zen` command use text defined
in userland. I added `zig translate-c-2` which is a work-in-progress
reimplementation of translate-c in userland, which currently calls
`std.debug.panic("unimplemented")` and you can see the stack trace makes
it all the way back into the C++ main() function (Thanks LemonBoy for
improving that!).
This could potentially let us move other things into userland, such as
hashing algorithms, the entire cache system, .d file parsing, pretty
much anything that libuserland.a itself doesn't need to depend on.
This can also let us have `zig fmt` in stage1 without the overhead
of child process execution, and without the initial compilation delay
before it gets cached.
See #1964
Does NOT look at the locale the way the C functions do.
int isalnum(int c);
int isalpha(int c);
int iscntrl(int c);
int isdigit(int c);
int isgraph(int c);
int islower(int c);
int isprint(int c);
int ispunct(int c);
int isspace(int c);
int isupper(int c);
int isxdigit(int c);
int isascii(int c);
int isblank(int c);
int toupper(int c);
int tolower(int c);
Tested to match glibc (when using C locale) with this program:
const c = @cImport({
// See https://github.com/ziglang/zig/issues/515
@cDefine("_NO_CRT_STDIO_INLINE", "1");
@cInclude("stdio.h");
@cInclude("string.h");
@cInclude("ctype.h");
});
const std = @import("std");
const ascii = std.ascii;
const abort = std.os.abort;
export fn main(argc: c_int, argv: **u8) c_int {
var i: u8 = undefined;
i = 0;
while (true) {
if (ascii.isAlNum(i) != (c.isalnum(i) > 0)) { abort(); }
if (ascii.isAlpha(i) != (c.isalpha(i) > 0)) { abort(); }
if (ascii.isCtrl(i) != (c.iscntrl(i) > 0)) { abort(); }
if (ascii.isDigit(i) != (c.isdigit(i) > 0)) { abort(); }
if (ascii.isGraph(i) != (c.isgraph(i) > 0)) { abort(); }
if (ascii.isLower(i) != (c.islower(i) > 0)) { abort(); }
if (ascii.isPrint(i) != (c.isprint(i) > 0)) { abort(); }
if (ascii.isPunct(i) != (c.ispunct(i) > 0)) { abort(); }
if (ascii.isSpace(i) != (c.isspace(i) > 0)) { abort(); }
if (ascii.isUpper(i) != (c.isupper(i) > 0)) { abort(); }
if (ascii.isXDigit(i) != (c.isxdigit(i) > 0)) { abort(); }
if (i == 255) { break; }
i += 1;
}
_ = c.printf(c"Success!\n");
return 0;
}
Unlike the other glibc source code checked into the repo, `csu/init.c`
did not have a license clause that allowed linking without restrictions.
`_IO_stdin_used` is the only symbol in the file and appears to be a 20
year old compatibility shim for the glibc 2.0 ABI. Obsolete in 2.1.
closes#2024
there's a new cli option `--main-pkg-path` which you can use to choose
a different root package directory besides the one inferred from the
root source file
and a corresponding build.zig API:
foo.setMainPkgPath(path)
* better libc detection
This introduces a new command `zig libc` which prints
the various paths of libc files. It outputs them to stdout
in a simple text file format that it is capable of parsing.
You can use `zig libc libc.txt` to validate a file.
These arguments are gone:
--libc-lib-dir [path] directory where libc crt1.o resides
--libc-static-lib-dir [path] directory where libc crtbegin.o resides
--msvc-lib-dir [path] (windows) directory where vcruntime.lib resides
--kernel32-lib-dir [path] (windows) directory where kernel32.lib resides
Instead we have this argument:
--libc [file] Provide a file which specifies libc paths
This is used to pass a libc text file (which can be generated with
`zig libc`). So it is easier to manage multiple cross compilation
environments.
`--cache on` now works when linking against libc.
`ZigTarget` now has a bool field `is_native`
Better error messaging when you try to link against libc or use
`@cImport` but the various paths cannot be found. It should also be
faster.
* save native_libc.txt in zig-cache
This avoids having to detect libc at runtime on every invocation.
Mostly picking the same paths as FreeBSD.
We need a little special handling for crt files, as netbsd uses its
own (and not GCC's) for those, with slightly different names.
This is not intended to be the long-term implementation as it doesn't
provide various properties that we eventually will want (e.g.
round-tripping, denormal support). It also uses f64 internally so the
wider f128 will be inaccurate.
Previously, std.debug.assert would `@panic` in test builds,
if the assertion failed. Now, it's always `unreachable`.
This makes release mode test builds more accurately test
the actual code that will be run.
However this requires tests to call `std.testing.expect`
rather than `std.debug.assert` to make sure output is correct.
Here is the explanation of when to use either one, copied from
the assert doc comments:
Inside a test block, it is best to use the `std.testing` module
rather than assert, because assert may not detect a test failure
in ReleaseFast and ReleaseSafe mode. Outside of a test block, assert
is the correct function to use.
closes#1304
this should actually improve CI times a bit too
See the description at the top of std/os/startup.zig (deleted in this
commit) for a more detailed understanding of what this commit does.
And add std.math.f128_* constants.
The routines are:
__fixdfdi, __fixdfsi, __fixdfti,
__fixsfdi, __fixsfsi, __fixsfti,
__fixtfdi, __fixtfsi, __fixtfti.
These all call fixint which is a generic zig function that does the
conversion:
pub fn fixint(comptime fp_t: type, comptime fixint_t: type, a: fp_t) fixint_t
There are also a set tests:
__fixdfdi_test, __fixdfsi_test, __fixdfti_test,
__fixsfdi_test, __fixsfsi_test, __fixsfti_test,
__fixtfdi_test, __fixtfsi_test, __fixtfti_test.
Relevant #764
dwarf debug info is modified to use this instead of std.os.File
directly to make it easier for bare metal projects to take advantage
of debug info parsing
* add __multi3 compiler rt function. See #1290
* compiler rt includes ARM functions for thumb and aarch64 and
other sub-arches left out. See #1526
* support C ABI for returning structs on ARM. see #1481
init-lib creates a working static library with tests, and
init-exe creates a working hello world with a `run` target.
both now have test coverage with the new "cli tests" file.
closes#1035
These are translated from [monocypher](https://monocypher.org/) which
has fairly competitive performance while remaining quite simple.
Initial performance comparision:
Zig:
Poly1305: 1423 MiB/s
X25519: 8671 exchanges per second
Monocypher:
Poly1305: 1567 MiB/s
X25519: 10539 exchanges per second
There is room for improvement and no real effort has been made at all in
optimization beyond a direct translation.
* add std.event.RwLock and std.event.RwLocked
* std.debug.warn does its printing locked
* add std.Mutex, however it's currently implemented as a spinlock
* rename std.event.Group.cancelAll to std.event.Group.deinit and change
the docs and assumptions.
* add std.HashMap.clone
