| Age | Commit message (Collapse) | Author |
|
closes #428
Thanks Marc Tiehuis for diagnosing this bug
|
|
previously we used the bigfloat abstraction to do all
compile-time float math. but runtime code and comptime code
are supposed to get the same result. so now if you add a
f32 to a f32 at compile time it does it with f32 math
instead of the bigfloat. float literals still get the
bigfloat math.
closes #424
|
|
|
|
* add u3, u4, u5, u6, u7 and i3, i4, i5, i6, i7
* shift operations shift amount parameter type is
integer with log2 bit width of other param
- This enforces not violating undefined behavior on
shift amount >= bit width with the type system
* clean up math.log, math.ln, math.log2, math.log10
closes #403
|
|
|
|
closes #418
also make clz, ctz return smaller integer bit widths
and use smaller integer bit widths for enum tag types
|
|
* add u128 and i128 integer types
* add f128 floating point type
* implement big integer multiplication (See #405)
|
|
Before:
* << is left shift, not allowed to shift 1 bits out
* <<% is left shift, allowed to shift 1 bits out
* >> is right shift, allowed to shift 1 bits out
After:
* << is left shift, allowed to shift 1 bits out
* >> is right shift, allowed to shift 1 bits out
* @shlExact is left shift, not allowed to shift 1 bits out
* @shrExact is right shift, not allowed to shift 1 bits out
Closes #413
|
|
closes #415
|
|
closes #407
The mangling strategy replaces bytes outside the alphabet
with "_xx_" where xx is the hex code of the byte.
|
|
|
|
|
|
See #393 for details
|
|
closes #390
|
|
also add bitCast builtin function. closes #387
|
|
|
|
Before, if you did something like:
```
const hi1 = "hi";
const hi2 = hi1;
```
This would create the "hi" data twice in the built object.
But since the value is const we don't have to duplicate the
data, now we take advantage of this fact.
closes #336
|
|
|
|
|
|
* skip installing std/rand_test.zig as it's not needed beyond running
the std lib tests
* add std.math.floor function
* add setFloatMode builtin function to choose between
builtin.FloatMode.Optimized (default) and builtin.FloatMode.Strict
(Optimized is equivalent to -ffast-math in gcc)
|
|
closes #373
|
|
...of special syntax.
partially reverts 41144a8566a6fbd779403f6b69424bb640c94a7f
closes #306
|
|
good riddance
|
|
to be consistent with other c primitive type names
|
|
* add `@divTrunc` and `@divFloor` functions
* add `@rem` and `@mod` functions
* add compile error for `/` and `%` with signed integers
* add `.bit_count` for float primitive types
closes #217
|
|
See #38
|
|
closes #288
|
|
|
|
See #226
Closes #220
|
|
closes #328
|
|
this reverts 5c04730534ea7933855429c5fc5dc7b22eba7bc2.
sadly the quality of the intel dialect in llvm's assembly
parser has many frustrating bugs, and generally has unfortunate
syntax.
the plan is to use AT&T for now since it at least works,
and eventually zig will have its own assembly parser for
x86 and it will be as close to NASM as possible.
|
|
See #298
|
|
fixes missing frame for unwrapping an error
closes #345
|
|
See #298
|
|
* add ability to add assembly files when building an exe, obj, or lib
* add implicit cast from `[N]T` to `?[]const T` (closes #343)
* remove link_exe and link_lib in favor of allowing build_exe and
build_lib support no root zig source file
|
|
they introduced complexity into the compiler and didn't
really help.
This reverts commit efa771af754281cd76a77ef22af107eb0a9aaf9a.
This reverts commit 8614397110595e267b7e4e1f558bfce619e60c02.
This reverts commit 13c6a58a61817ae6aae66dd1843385452fe65bd1.
|
|
once again this barely had an effect:
Before:
./build size: 1.3 MB
hello.zig size: 301 KB
full test: 1m31.253s
debug test: 19.607s
hello.zig timing:
Name Start End Duration Percent
Initialize 0.0000 0.0000 0.0000 0.0002
Semantic Analysis 0.0000 0.0431 0.0431 0.2262
Code Generation 0.0431 0.0660 0.0229 0.1201
LLVM Emit Object 0.0660 0.1765 0.1105 0.5795
Build Dependencies 0.1765 0.1890 0.0125 0.0655
LLVM Link 0.1890 0.1906 0.0016 0.0086
Generate .h 0.1906 0.1906 0.0000 0.0000
Total 0.0000 0.1906 0.1906 1.0000
After:
./build size: 1.3 MB
hello.zig size: 300 KB
full test: 1m31.882s
debug test: 19.569s
hello.zig timing:
Name Start End Duration Percent
Initialize 0.0000 0.0000 0.0000 0.0002
Semantic Analysis 0.0000 0.0425 0.0424 0.2228
Code Generation 0.0425 0.0661 0.0236 0.1239
LLVM Emit Object 0.0661 0.1762 0.1101 0.5782
Build Dependencies 0.1762 0.1888 0.0126 0.0664
LLVM Link 0.1888 0.1905 0.0016 0.0085
Generate .h 0.1905 0.1905 0.0000 0.0000
Total 0.0000 0.1905 0.1905 1.0000
|
|
move some boilerplate code having to do with overflow math safety
to functions.
Again the timing difference is not much:
Before:
./build size: 1.3 MB
hello.zig size: 308 KB
full test: 1m33.588s
debug test: 20.303s
hello.zig timing:
Name Start End Duration Percent
Initialize 0.0000 0.0000 0.0000 0.0002
Semantic Analysis 0.0000 0.0425 0.0425 0.2202
Code Generation 0.0425 0.0675 0.0250 0.1293
LLVM Emit Object 0.0675 0.1789 0.1114 0.5773
Build Dependencies 0.1789 0.1913 0.0124 0.0640
LLVM Link 0.1913 0.1931 0.0018 0.0091
Generate .h 0.1931 0.1931 0.0000 0.0000
Total 0.0000 0.1931 0.1931 1.0000
After:
./build size: 1.3 MB
hello.zig size: 301 KB
full test: 1m31.253s
debug test: 19.607s
hello.zig timing:
Name Start End Duration Percent
Initialize 0.0000 0.0000 0.0000 0.0002
Semantic Analysis 0.0000 0.0431 0.0431 0.2262
Code Generation 0.0431 0.0660 0.0229 0.1201
LLVM Emit Object 0.0660 0.1765 0.1105 0.5795
Build Dependencies 0.1765 0.1890 0.0125 0.0655
LLVM Link 0.1890 0.1906 0.0016 0.0086
Generate .h 0.1906 0.1906 0.0000 0.0000
Total 0.0000 0.1906 0.1906 1.0000
|
|
move some boilerplate code having to do with panicking
to a function.
Here's the timing difference. It's not much:
Before:
full test: 1m36.511s
debug test: 20.862s
hello.zig
Name Start End Duration Percent
Initialize 0.0000 0.0000 0.0000 0.0001
Semantic Analysis 0.0000 0.0421 0.0420 0.2109
Code Generation 0.0421 0.0620 0.0200 0.1003
LLVM Emit Object 0.0620 0.1852 0.1231 0.6180
Build Dependencies 0.1852 0.1974 0.0122 0.0615
LLVM Link 0.1974 0.1993 0.0018 0.0093
Generate .h 0.1993 0.1993 0.0000 0.0000
Total 0.0000 0.1993 0.1993 1.0000
After:
full test: 1m33.588s
debug test: 20.303s
hello.zig
Name Start End Duration Percent
Initialize 0.0000 0.0000 0.0000 0.0002
Semantic Analysis 0.0000 0.0425 0.0425 0.2202
Code Generation 0.0425 0.0675 0.0250 0.1293
LLVM Emit Object 0.0675 0.1789 0.1114 0.5773
Build Dependencies 0.1789 0.1913 0.0124 0.0640
LLVM Link 0.1913 0.1931 0.0018 0.0091
Generate .h 0.1931 0.1931 0.0000 0.0000
Total 0.0000 0.1931 0.1931 1.0000
|
|
pass --enable-timing-info to print a nice table like this:
```
Name Start End Duration Percent
Initialize 0.0000 0.0000 0.0000 0.0001
Semantic Analysis 0.0000 0.0421 0.0420 0.2109
Code Generation 0.0421 0.0620 0.0200 0.1003
LLVM Emit Object 0.0620 0.1852 0.1231 0.6180
Build Dependencies 0.1852 0.1974 0.0122 0.0615
LLVM Link 0.1974 0.1993 0.0018 0.0093
Generate .h 0.1993 0.1993 0.0000 0.0000
Total 0.0000 0.1993 0.1993 1.0000
```
|
|
|
|
closes #291
This changes the error message "return value ignored" to "expression value is ignored".
This is because this error also applies to {1;}, which has no function calls.
Also fix ignored expression values in std and test.
This caught a bug in debug.readAllocBytes where an early Eof error would have been missed.
See #219.
|
|
closes #268
|
|
closes #339
|
|
|
|
See #329
Supporting work:
* move std.cstr.Buffer0 to std.buffer.Buffer
* add build.zig to example/shared_library/ and add an automated test
for it
* add std.list.List.resizeDown
* improve std.os.makePath
- no longer recursive
- takes into account . and ..
* add std.os.path.isAbsolute
* add std.os.path.resolve
* reimplement std.os.path.dirname
- no longer requires an allocator
- handles edge cases correctly
|
|
|
|
rename self hosted tests to behavior tests
|
|
closes #30
|
|
closes #320
|