687bd92f9c
self-hosted: generate zig IR for simple function
...
no tests for this yet. I think the quickest path to testing will be
creating the .o files and linking with libc, executing, and then
comparing output.
2018-07-12 15:12:44 -04:00
30c4add85a
std.event.Future: workaround in tests for llvm coro memory
...
See #1194
2018-07-11 20:17:47 -04:00
9751a0ae04
std.atomic: use spinlocks
...
the lock-free data structures all had ABA problems and
std.atomic.Stack had a possibility to load an unmapped memory address.
2018-07-11 19:38:01 -04:00
9bdcd2a495
add std.event.Future
...
This is like a promise, but it's for multiple getters, and
uses an event loop.
2018-07-11 16:00:06 -04:00
3f30897fdc
add compile error for disallowed types in extern structs
...
closes #1218
2018-07-11 14:08:56 -04:00
9b054e73f6
Add generic comparator generator functions for sorting
...
- Copy-by-value instead of pointer where appropriate
- Clean up old zig fmt issues
2018-07-11 18:44:30 +12:00
574e31f0a0
self-hosted: first passing test
...
* introduce std.atomic.Int
* add src-self-hosted/test.zig which is tested by the main test suite
- it fully utilizes the multithreaded async/await event loop so the
tests should Go Fast
* `stage2/bin/zig build-obj test.zig` is able to spit out an error if 2 exported
functions collide
* ability for `zig test` to accept `--object` and `--assembly`
arguments
* std.build: TestStep supports addLibPath and addObjectFile
2018-07-10 20:18:43 -04:00
8fba0a6ae8
introduce std.event.Group for making parallel async calls
2018-07-10 15:17:01 -04:00
0ce6934e26
allow var args calls to async functions
2018-07-10 11:44:47 -04:00
b5cfbfd84e
fix regression from b6eb4048
2018-07-09 23:41:28 -04:00
1b82a9defc
enable basic event loop test
2018-07-09 22:41:16 -04:00
b6eb404831
organize std.event into directories
2018-07-09 22:22:44 -04:00
ccef60a640
Merge pull request #1198 from ziglang/m-n-threading
...
M:N threading
2018-07-09 22:06:47 -04:00
c89aac85c4
better workaround for guaranteeing memory in coroutine frame
...
See #1194
2018-07-09 21:21:59 -04:00
a2834d48b9
Update throughput_test.zig. ( #1211 )
2018-07-09 17:21:20 -04:00
1a1534ecb5
fix regression on macos
2018-07-09 17:16:06 -04:00
3f4d0ecd7e
Merge remote-tracking branch 'origin/master' into m-n-threading
2018-07-09 17:14:42 -04:00
9462852433
std.event.Loop multithreading for windows using IOCP
2018-07-09 16:49:46 -04:00
caa0085057
implement std.os.cpuCount for windows
2018-07-09 13:19:11 -04:00
05f1ea33d2
ZIG_DEBUG_COLOR=1 overrides tty detection for runtime stack traces
2018-07-09 12:12:37 -04:00
42ba06133a
std.Hashmap - don't use catch unreachable in tests
2018-07-09 10:44:06 -04:00
a0c564d762
zig fmt
2018-07-09 01:23:47 -04:00
3ba451778f
fix regressions on linux
2018-07-09 01:22:36 -04:00
82e9190d09
Update zig.parser benchmark program
2018-07-09 17:14:04 +12:00
04d3da4bd1
std.os.cpuCount implementation for macos
2018-07-09 01:08:33 -04:00
50d70d5f49
tests passing with kqueue on macos
2018-07-08 02:46:10 -04:00
410b4d9bdf
builder.addBuildOption
2018-07-08 00:00:05 -04:00
ced3aae3b2
cleaner output from zig build when there are compile errors
2018-07-07 20:31:50 -04:00
c15a6fa9d0
add std.os.cpuCount and have std.event.Loop use it for thread pool size
2018-07-07 01:23:18 -04:00
57f36c4201
std.event.Loop: use EPOLLONESHOT to save 1 syscall
...
when a thread pool worker accepts a coroutine to resume
2018-07-07 00:32:19 -04:00
eb326e1553
M:N threading
...
* add std.atomic.QueueMpsc.isEmpty
* make std.debug.global_allocator thread-safe
* std.event.Loop: now you have to choose between
- initSingleThreaded
- initMultiThreaded
* std.event.Loop multiplexes coroutines onto kernel threads
* Remove std.event.Loop.stop. Instead the event loop run() function
returns once there are no pending coroutines.
* fix crash in ir.cpp for calling methods under some conditions
* small progress self-hosted compiler, analyzing top level declarations
* Introduce std.event.Lock for synchronizing coroutines
* introduce std.event.Locked(T) for data that only 1 coroutine should
modify at once.
* make the self hosted compiler use multi threaded event loop
* make std.heap.DirectAllocator thread-safe
See #174
TODO:
* call sched_getaffinity instead of hard coding thread pool size 4
* support for Windows and MacOS
* #1194
* #1197
2018-07-07 00:32:19 -04:00
28821b5f31
Fixed last commit compiler error
2018-07-04 11:35:29 +02:00
4f32b86142
Allow allocation of any 0 sized type (not just void)
2018-07-04 11:29:02 +02:00
1eda86e1ad
Clean up outstanding compiler_rt todos
2018-07-03 13:22:12 +12:00
bd282d6cca
Merge pull request #1176 from bnoordhuis/f16-std
...
improve std.math f16 support
2018-07-02 16:03:25 -04:00
96a6bc57d2
modify std.event.Loop to work for windows and macos
2018-07-02 14:38:11 -04:00
a3f55aaf34
add event loop Channel abstraction
...
This is akin to channels in Go, except:
* implemented in userland
* they are lock-free and thread-safe
* they integrate with the userland event loop
The self hosted compiler is changed to use a channel for events,
and made to stay alive, watching files and performing builds when
things change, however the main.zig file exits after 1 build.
Note that nothing is actually built yet, it just parses the input
and then declares that the build succeeded.
Next items to do:
* add windows and macos support for std.event.Loop
* improve the event loop stop() operation
* make the event loop multiplex coroutines onto kernel threads
* watch source file for updates, and provide AST diffs
(at least list the top level declaration changes)
* top level declaration analysis
2018-07-02 14:38:11 -04:00
887c97742f
Alignment fix and allow rudimentary f128 float printing
2018-06-30 21:58:59 +12:00
9f48b2ab48
compiler_rt: Remove wrapping add/sub operators where unneeded
...
Closes #495 .
2018-06-30 21:58:59 +12:00
814a34f263
compiler_rt: Add floattitf/floattidf/floattisf
2018-06-30 21:58:59 +12:00
e19fc4a0a3
compiler_rt: Add missing exports
2018-06-30 21:58:59 +12:00
cb7bdc2da1
compiler_rt: Add floatuntitf
2018-06-30 21:58:59 +12:00
61ebfe6603
compiler_rt: Add floatunditf and floatunsitf
2018-06-30 21:58:59 +12:00
c32b2e45ef
compiler_rt: Add floatuntisf
2018-06-30 21:58:59 +12:00
379950f81d
compiler_rt: Add trunc f128 narrowing functions
2018-06-30 21:58:59 +12:00
30cfc0ab2c
test std.math f16 sqrt support
...
refs #1122
2018-06-30 01:58:17 +02:00
be36179064
add std.math f16 signbit support
...
refs #1122
2018-06-30 01:58:17 +02:00
ca444e6191
add std.math f16 copysign support
...
refs #1122
2018-06-30 01:58:17 +02:00
d293f1a0ed
add std.math f16 floor support
...
refs #1122
2018-06-30 01:58:17 +02:00
1abc925292
add std.math f16 fabs support
...
refs #1122
2018-06-30 01:58:17 +02:00
f36b095b5f
add std.math f16 isnormal support
...
refs #1122
2018-06-30 01:58:17 +02:00
30b75ae353
add std.math f16 isfinite support
...
refs #1122
2018-06-30 01:58:17 +02:00
a36d7b6131
add std.math f16 inf support
...
refs #1122
2018-06-30 01:58:17 +02:00
27b02413dc
add std.math f16 nan support
...
refs #1122
2018-06-30 01:58:17 +02:00
61df5bc142
add std.math f16 constants
...
refs #1122
2018-06-30 01:58:17 +02:00
0874a5ba77
std.atomic.queue - document limitation and add MPSC queue
2018-06-29 14:45:42 -04:00
3e94347e61
Fix up some std.rand syntax #1161 ( #1162 )
...
* Fix old syntax in rand
Ziggurat somehow did not get updated to latest syntax
* Fix broken float casts
f32 float casts somehow not updated to latest syntax
2018-06-27 12:30:15 -04:00
fd75e73ee9
add f16 type
...
Add support for half-precision floating point operations.
Introduce `__extendhfsf2` and `__truncsfhf2` in std/special/compiler_rt.
Add `__gnu_h2f_ieee` and `__gnu_f2h_ieee` as aliases that are used in
Windows builds.
The logic in std/special/compiler_rt/extendXfYf2.zig has been reworked
and can now operate on 16 bits floating point types.
`extendXfYf2()` and `truncXfYf2()` are marked `inline` to work around
a not entirely understood stack alignment issue on Windows when calling
the f16 versions of the builtins.
closes #1122
2018-06-27 16:20:04 +02:00
3290e72833
std.zig.ast: fix incorrect impl of FnProto.firstToken
...
closes #1151
2018-06-25 11:54:10 -04:00
5f38d6e2e9
add casting docs, __extenddftf2, and __extendsftf2
2018-06-21 14:44:35 -04:00
f50c0c664f
Add float repr bit extraction functions
2018-06-21 01:45:12 -04:00
85f928f8bf
remove std.mem.Allocator.construct and other fixups
2018-06-20 17:33:29 -04:00
e891f9cd9d
zig fmt
2018-06-20 17:16:27 -04:00
6bd8610063
std.mem.Allocator.construct: improve formatting;
2018-06-21 01:40:25 +09:00
4b46af4810
std.mem.Allocator.construct: remove deprecation warning;
2018-06-21 01:39:48 +09:00
71db8df548
std: update stdlib to match updated allocator create signature; ref #733
2018-06-21 00:40:21 +09:00
457c0f0a7e
std.mem: remove allocator create in favor of construct; ref #733
2018-06-21 00:39:19 +09:00
c7804277bf
`@floatToInt` now has safety-checked undefined behavior
...
when the integer part does not fit in the destination integer type
* Also fix incorrect safety triggered for integer casting an
`i32` to a `u7`. closes #1138
* adds compiler-rt function: `__floatuntidf`
2018-06-19 16:06:10 -04:00
a3ddd0826b
remove enum to/from int casting syntax; add `@enumToInt`/`@intToEnum`
...
see #1061
2018-06-19 03:50:38 -04:00
626b73e8be
remove error to/from int casting syntax; add `@errorToInt`/`@intToError`
...
See #1061
2018-06-18 18:48:29 -04:00
a430853a48
standard library fixes
2018-06-18 17:43:01 -04:00
1aafbae5be
remove []u8 casting syntax. add `@bytesToSlice` and `@sliceToBytes`
...
See #1061
2018-06-18 17:25:29 -04:00
4210f1f6a0
remove bool to int syntax. add @boolToInt
...
add missing docs
See #1061
2018-06-18 03:07:16 -04:00
906ed059ce
update std.DynLib to use @intCast
2018-06-18 00:52:55 -04:00
7912061226
remove integer and float casting syntax
...
* add `@intCast`
* add `@floatCast`
* add `@floatToInt`
* add `@intToFloat`
See #1061
2018-06-17 02:57:07 -04:00
06a26f0965
std.Complex: use better arg passing convention and fix a TODO
2018-06-16 21:32:53 -04:00
751518787a
Merge pull request #1109 from ziglang/pass-by-non-copying-value
...
allow passing by non-copying value
2018-06-16 21:13:10 -04:00
3ee4d23ebd
posix read can return error.IsDir
2018-06-16 19:54:16 -04:00
e311cd562b
don't automatically take pointer when passing by non-copying value
...
this commit does not have all tests passing
2018-06-16 19:37:00 -04:00
65d04cbeb4
std.DynLib: open the fd with CLOEXEC
2018-06-16 17:27:45 -04:00
48de57d824
add basic std lib code for loading dynamic libraries
...
this is going to only work for very basic libraries;
I plan to slowly add more features over time to support more
complicated libraries
2018-06-16 17:01:23 -04:00
cdf1e366f9
fix build on windows, broken by previous commit
2018-06-14 16:36:07 -04:00
6943cefebf
std.os.path.dirname: return null instead of empty slice
...
for when there is no directory component. Makes it harder
to write bugs.
closes #1017
2018-06-14 16:15:32 -04:00
2219cc0612
Merge pull request #1105 from ziglang/i128-compiler-rt
...
Add i128 compiler-rt div/mul support
2018-06-14 10:59:38 -04:00
a369d69c51
Add windows x86_64 i128 abi workaround
2018-06-14 21:18:36 +12:00
fc87f6e417
fix race condition bug in test harness of std.atomic
2018-06-13 11:57:57 -04:00
e1f56c9af6
std.zig.ast: add test for iterate
...
closes #1101
2018-06-13 11:48:06 -04:00
9110140514
Add i128 compiler-rt div/mul support
2018-06-13 22:25:04 +12:00
fdd9cf0928
better debugging for CI failures of std.atomic
2018-06-12 15:14:32 -04:00
7580e39b38
zig fmt
2018-06-12 02:18:11 -04:00
3dd9af9948
implement std.os.Dir for windows
...
improve std.os.File.access so that it does not depend on shlwapi.dll
closes #1084
2018-06-12 01:57:09 -04:00
854f90aa30
Added C string support to fmt by using "{s}". The format string requirement is for saftey. ( #1092 )
2018-06-10 12:57:21 -04:00
dc8bda7e02
Add arbitrary-precision integer to std
...
A few notes on the implementation:
- Any unsigned power of two integer type less than 64 bits in size is supported
as a Limb type.
- The algorithms used are kept simple for the moment. More complicated
algorithms are generally only more useful as integer sizes increase a
lot and I don't expect our current usage to be used for this purpose
just yet.
- All branches (practically) have been covered by tests.
See 986a2b3243/benchCloses #364 .
2018-06-10 18:24:34 +12:00
0a95b0f1ff
std.zig: update syntax for orelse keyword
2018-06-10 01:18:31 -04:00
77678b2cbc
breaking syntax change: orelse keyword instead of ?? ( #1096 )
...
use the `zig-fmt-optional-default` branch to have zig fmt
automatically do the changes.
closes #1023
2018-06-10 01:13:51 -04:00
ec1b6f6673
breaking syntax change: ??x to x.? ( #1095 )
...
See #1023
This also renames Nullable/Maybe to Optional
2018-06-09 23:42:14 -04:00
7a9635555b
std/fmt: Use lowercase k for kilo in base 1000 ( #1090 )
2018-06-09 12:05:58 -04:00
fc6446702e
clean up std.ArrayList
...
* add `std.debug.assertError`
* `std.ArrayList` update everything to follow `self` convention
* rename `std.ArrayList.set` to `std.ArrayList.setOrError`
* add `std.ArrayList.set` which asserts
Before 1.0.0 we might remove some of this API, because you can use
`toSlice()` for everything, but it's ok to add these functions as
an experiment before then.
2018-06-09 12:03:11 -04:00
e0092ee4a5
add set function to arraylist
...
so you can set a value without growing the underlying buffer,
with range safety checks
2018-06-09 11:42:41 -04:00
1a9d2f3aae
Fix error handling in Buffer::fromOwnedSlice ( #1082 )
2018-06-08 19:24:48 -04:00
ffb089a9f5
Fix json parser comma after empty object case
2018-06-08 17:43:13 +12:00
b11c5d8f82
fix std.os.windows.PathFileExists specified in the wrong DLL ( #1066 )
...
closes #1054
2018-06-06 15:36:47 -04:00
4fc601895b
Fix const-ness of buffer in replaceContents method ( #1065 )
2018-06-06 14:09:47 -04:00
e7f141b376
Add json.TokenStream ( #1062 )
...
This hides some of the low-level parsing details from the
StreamingParser. These don't need to be known when parsing a complete
slice at once (which is we can usually do).
Also, remove `Json` from Parser names. The namespace `json` is sufficient.
2018-06-06 11:24:36 -04:00
212449bc23
Fix Log2Int type construction
...
The following case for example, would previously fail:
const a = u24(1) << Log2Int(u24)(22);
2018-06-06 22:41:55 +12:00
d3693dca73
Pointer Reform: update @typeInfo
...
* add assertion for trying to do @typeInfo on global error set
* remove TypeInfo.Slice
* add TypeInfo.Pointer.Size with possible values
- One
- Many
- Slice
See #770
2018-06-06 00:39:39 -04:00
bd13e757e7
disable deref syntax for unknown length pointers
...
See #770
2018-06-05 23:26:43 -04:00
652f4bdf62
disallow unknown-length pointer to opaque
...
This also means that translate-c has to detect when a pointer to
opaque is happening, and use `*` instead of `[*]`.
See #1059
2018-06-05 18:03:21 -04:00
02cb220faf
Renamed "(int/float literal)" to "comptime_int/float"
2018-06-05 11:14:43 +02:00
e53b683bd3
Pointer Reform: proper slicing and indexing ( #1053 )
...
* enable slicing for single-item ptr to arrays
* disable slicing for other single-item pointers
* enable indexing for single-item ptr to arrays
* disable indexing for other single-item pointers
see #770
closes #386
2018-06-04 22:11:14 -04:00
d21a1922eb
support `zig fmt: off` and `zig fmt: on` between top level decls
...
closes #1030
closes #1033
2018-06-04 12:15:02 -04:00
11e7e03139
Merge branch 'zig-custom-format' of https://github.com/tgschultz/zig into tgschultz-zig-custom-format
...
I removed the code that checks for type signature and type.
A function named `format` is enough for zig to give it a try.
2018-06-04 11:06:55 -04:00
96164ce613
disallow single-item pointer indexing
...
add pointer arithmetic for unknown length pointer
2018-06-04 01:39:57 -04:00
f06bce5dda
introduce [*] for unknown length pointers
...
See #770
Currently it does not have any different behavior than `*`
but it is now recommended to use `[*]` for unknown length
pointers to be future-proof.
Instead of [ * ] being separate tokens as the proposal
suggested, this commit implements `[*]` as a single token.
2018-06-02 04:04:23 -04:00
7b386ea242
fix build file template
...
See #1035
2018-06-01 21:51:54 -04:00
08693411d2
fix typo ( #1034 )
2018-06-01 12:23:07 -04:00
e29d12d821
fix incorrect address-of syntax on windows
2018-06-01 01:29:34 -04:00
5f38a01ede
run zig fmt
2018-06-01 01:22:35 -04:00
4d13ab07de
std.zig: update to new pointer syntax
2018-06-01 01:19:26 -04:00
2f614c42fe
ir: rip out special logic for using addr-of instruction for types
...
See #588
2018-05-31 18:23:39 -04:00
fcbb7426fa
use * for pointer type instead of &
...
See #770
To help automatically translate code, see the
zig-fmt-pointer-reform-2 branch.
This will convert all & into *. Due to the syntax
ambiguity (which is why we are making this change),
even address-of & will turn into *, so you'll have
to manually fix thes instances. You will be guaranteed
to get compile errors for them - expected 'type', found 'foo'
2018-05-31 17:28:07 -04:00
a05acaf9fd
Add --color CLI option to zig fmt
...
It doesn't actually do terminal color yet because we need to add
cross platform terminal color abstractions. But it toggles between
the single line error reporting and the multiline error reporting.
See #1026
2018-05-30 18:26:09 -04:00
ea58f4a5a9
run zig fmt on the codebase
2018-05-30 16:09:11 -04:00
b082cd4580
zig fmt: field access does not cause spaces for slicing
...
See #1003
2018-05-30 16:08:40 -04:00
84b1842026
zig fmt: space after fn in fn prototypes
...
See #1003
2018-05-30 15:50:01 -04:00
93b51b0e40
spaces around slice operator if operands are infix
...
See #1003
2018-05-30 15:33:58 -04:00
2c96f19fd3
std.zig.render returns bool of whether anything changed
...
zig fmt only renames files and prints to stdout for files which changed
2018-05-30 14:58:27 -04:00
940a854448
Fix MacOS CI Timer test failing...?
2018-05-30 13:38:41 -05:00
fb001f5e90
Fixed character handling
2018-05-30 12:18:24 -05:00
8938c16f38
Formatting
2018-05-30 10:41:48 -05:00
4e1d0a59fa
Minor typo
2018-05-30 10:24:27 -05:00
8fc52a94f4
Added custom formatter support, refactored fmt.format
2018-05-30 10:18:11 -05:00
b0eebfa560
fix syntax of std/json_test.zig
...
See #663
2018-05-29 18:10:36 -04:00
d172e3f3bb
fix AtomicFile for relative paths
...
closes #1017
2018-05-29 17:38:50 -04:00
0c16cd2d0e
run zig fmt on the codebase
...
See #1003
2018-05-29 04:23:38 -04:00
cdf30c31ea
zig fmt: fix implementation of firstToken() for fn call
2018-05-29 03:47:27 -04:00
cd325e408e
zig fmt: fix extra space after comma before multi line string
2018-05-29 03:33:03 -04:00
eda6898c5b
zig fmt: handle if and while indentation better
2018-05-29 03:15:12 -04:00
530d175422
zig fmt: fix spacing when moving doc comment on var decls
2018-05-28 23:41:09 -04:00
0d1b47362c
zig fmt: if-else with comment before else
2018-05-28 22:41:05 -04:00
77ec81b035
zig fmt: respect line breaks in if-else
2018-05-28 22:22:01 -04:00
71badebd08
zig fmt: respect line breaks after infix operators
2018-05-28 21:28:32 -04:00
354ab1c5c8
zig fmt: render fn decl with trailing comma 1 line per param
2018-05-28 21:18:41 -04:00
530da36352
zig fmt: fix enum decl with no trailing comma
...
See #1003
2018-05-28 17:43:17 -04:00
6c1fda3f99
zig fmt: fix switch body indent
2018-05-28 17:09:55 -04:00
fd13a75785
zig fmt: allow same line struct literal with no trailing comma
...
See #1003
2018-05-28 17:00:04 -04:00
122a74724c
zig fmt: use simple newlines rather than empty comments to hint
...
now the first row of an array literal is the hint to zig fmt
for how long each row should be.
See #1003
2018-05-28 16:23:33 -04:00
3fed10883b
zig fmt: array literals with no trailing comma all on one line
2018-05-26 23:25:04 -04:00
afdfbc0367
zig fmt: delete empty comments that do nothing
2018-05-26 23:25:04 -04:00
b184ae5ca5
run zig fmt on some of the codebase
...
See #1003
2018-05-26 23:00:29 -04:00
Andrew Kelley
Marc Tiehuis
wilsonk
Josh Wolfe
Jimmi HC
Ben Noordhuis
tgschultz
kristopher tate
marleck55
Arthur Elliott
isaachier