Compare commits

..

78 Commits

Author SHA1 Message Date
Christoffer Lerno
e293c435af 0.6.0: init_new/init_temp removed. LinkedList API rewritten. List "pop" and "remove" function now return Optionals. RingBuffer API rewritten. Allocator interface changed. Deprecated Allocator, DString and mem functions removed. "identity" functions are now constants for Matrix and Complex numbers. @default implementations for interfaces removed. any* => any, same for interfaces. Emit local/private globals as "private" in LLVM, following C "static". Updated enum syntax. Add support [rgba] properties in vectors. Improved checks of aliased "void". Subarray -> slice. Fix of llvm codegen enum check. Improved alignment handling. Add --output-dir #1155. Removed List/Object append. GenericList renamed AnyList. Remove unused "unwrap". Fixes to cond. Optimize output in dead branches. Better checking of operator methods. Disallow any from implementing dynamic methods. Check for operator mismatch. Remove unnecessary bitfield. Remove numbering in --list* commands Old style enum declaration for params/type, but now the type is optional. Add note on #1086. Allow making distinct types out of "void", "typeid", "anyfault" and faults. Remove system linker build options. "Try" expressions must be simple expressions. Add optimized build to Mac tests. Register int. assert(false) only allowed in unused branches or in tests. Compile time failed asserts is a compile time error. Remove current_block_is_target. Bug when assigning an optional from an optional. Remove unused emit_zstring. Simplify phi code. Remove unnecessary unreachable blocks and remove unnecessary current_block NULL assignments. Proper handling of '.' and Win32 '//server' paths. Add "no discard" to expression blocks with a return value. Detect "unsigned >= 0" as errors. Fix issue with distinct void as a member #1147. Improve callstack debug information #1184. Fix issue with absolute output-dir paths. Lambdas were not type checked thoroughly #1185. Fix compilation warning #1187. Request jump table using @jump for switches. Path normalization - fix possible null terminator out of bounds. Improved error messages on inlined macros.
Upgrade of mingw in CI. Fix problems using reflection on interface types #1203. Improved debug information on defer. $foreach doesn't create an implicit syntactic scope.
Error if `@if` depends on `@if`. Updated Linux stacktrace. Fix of default argument stacktrace. Allow linking libraries directly by file path. Improve inlining warning messages. Added `index_of_char_from`. Compiler crash using enum nameof from different module #1205. Removed unused fields in find_msvc. Use vswhere to find msvc. Update tests for LLVM 19
2024-06-12 10:14:26 +02:00
Christoffer Lerno
321c5ec756 Update mingw version and funding. 2024-05-27 12:03:45 +02:00
Christoffer Lerno
f04d93f9aa Fix flaw in bitstruct check. 2024-05-20 22:20:33 +02:00
Christoffer Lerno
9436efe554 Compiler crash on designated initializer for structs with bitstruct. 2024-05-20 14:42:09 +02:00
Christoffer Lerno
3acbf708d3 Fix location on foreach debug output. 2024-05-19 23:27:57 +02:00
Christoffer Lerno
92979984ea Fix mutex and wait signatures for Win32. 2024-05-18 22:24:45 +02:00
Christoffer Lerno
a16d41a1e1 Do not elide memory storage on variable for debug. 2024-05-17 19:51:35 +02:00
Christoffer Lerno
ff8b78fc99 Correct debug info on parameters without storage. 2024-05-17 16:26:12 +02:00
Christoffer Lerno
97c9bd7ce0 Assertion failed when casting argument to enum #1196 2024-05-16 16:07:55 +02:00
Christoffer Lerno
c40c93340d Compile time fmod evaluates to 0 #1195 2024-05-16 14:33:11 +02:00
Christoffer Lerno
094c105464 Union is not properly zero-initialized #1194 2024-05-16 11:28:04 +02:00
Christoffer Lerno
e36c696624 Update CI. 2024-05-15 21:49:18 +02:00
Christoffer Lerno
555a4ab4c5 Casting a slice address to its pointer type should not compile #1193. 2024-05-15 21:36:53 +02:00
Christoffer Lerno
7d8cc8776d Duplicate emit of expressions on negation would incorrectly compile negated macros. 2024-05-14 23:30:39 +02:00
Christoffer Lerno
960646ac8a Patch test. 2024-05-09 09:53:48 +02:00
Christoffer Lerno
ed9f15becf Foreach uses non-wrapping add/dec. 2024-05-08 23:05:12 +02:00
Christoffer Lerno
b09aa74f2f Generic modules parameterized with constants would sometimes get the wrong parameterized module name causing conversion errors #1192. 2024-05-04 23:34:37 +02:00
Christoffer Lerno
60805fd11d Bounds checking on length for foo[1:2] slicing #1191 2024-04-28 18:55:26 +02:00
Christoffer Lerno
89ecd4b33d Default to AVX on x64. 2024-04-26 19:29:32 +02:00
Christoffer Lerno
237f142a87 Default CPU actually defaults to a value instead of picking the native CPU. 2024-04-26 19:03:25 +02:00
Christoffer Lerno
a21647a1aa Do not default to native vector capability. 2024-04-26 18:45:14 +02:00
Christoffer Lerno
e9afe4ee25 Update CI script for mac. 2024-04-26 18:09:03 +02:00
Christoffer Lerno
acd067582a Update CI script. 2024-04-26 18:07:16 +02:00
Christoffer Lerno
82227e8901 Incorrect cast of bitstructs #1186 2024-04-26 17:39:30 +02:00
Christoffer Lerno
8b6735a6aa Allow recursive function definitions as long as they are pointers #1182. Add 'zstr' variants for string::new_format / string::tformat. 2024-04-16 19:42:32 +02:00
Christoffer Lerno
9ed8831500 Updated link 2024-04-14 23:07:48 +02:00
Christoffer Lerno
e7d726cc2c Fixup of scratch buffer code. 2024-04-09 14:27:52 +02:00
Christoffer Lerno
11a3dd26c8 Update mingw version. 2024-04-09 14:10:54 +02:00
Christoffer Lerno
04738586b9 Fix bug in scratch_buffer_printf. 2024-04-09 13:26:08 +02:00
cpiernikowski
18b4fce1ca Change return type of next_bool(random) from void to bool 2024-03-28 09:44:47 +01:00
Christoffer Lerno
3b9babe745 0.5.6 Add grammar for defer (catch err). 2024-03-26 09:36:45 +01:00
Brennan Cottrell
a4a85b7bbf Added print-input command line argument (#1175)
* added print-input command line argument
2024-03-26 09:33:47 +01:00
Christoffer Lerno
e8f0275d8e 0.5.6 Add defer (catch err) feature. 2024-03-25 11:35:16 +01:00
Christoffer Lerno
3251f58d46 Change version for MinGW 2024-03-25 09:40:52 +01:00
David
204fb211ac Fix x86_64 ABI small issue (#1174)
* Fix x86_64 ABI small issue Update tests for fix.

---------

Co-authored-by: Christoffer Lerno <christoffer@aegik.com>
2024-03-25 09:39:26 +01:00
Christoffer Lerno
6cade814e1 Update includes for FreeBSD. 2024-03-22 09:11:18 +01:00
Christoffer Lerno
eb2fbabbb1 Update version to 0.5.6 2024-03-19 16:15:13 +01:00
Christoffer Lerno
ee9c5db719 0.5.5 release. 2024-03-18 22:05:16 +01:00
Christoffer Lerno
d8af01dc46 Update release notes and change how versions are reported. 2024-03-18 11:51:31 +01:00
David Gonzalez Martin
5bead069f2 Fix aarch64 return type ABI bug 2024-03-17 20:06:53 +01:00
shv187
63e1345780 Add getModuleHandleA + W to win32 2024-03-16 09:47:42 +01:00
Christoffer Lerno
3df988d0b8 Allow String constants -> ichar*, and allow integer pointers to explicitly convert between unsigned signed. 2024-03-15 23:08:52 +01:00
Christoffer Lerno
656202dc0d Convert paths to backslash before running on Windows. 2024-03-15 20:09:06 +01:00
Christoffer Lerno
eec3253669 Mingw -> 18.1.1-3 2024-03-15 13:23:41 +01:00
Christoffer Lerno
d9423201b8 Change mingw version. 2024-03-15 13:20:05 +01:00
Christoffer Lerno
c6087bc369 Fix underlying type of llvm.used and update section. 2024-03-15 13:02:39 +01:00
Christoffer Lerno
b7077c7967 Fix Win32 with compile-run. 2024-03-14 12:05:01 +01:00
Christoffer Lerno
4acb07f1cb compile-run and run now returns the proper return code. 2024-03-14 11:55:55 +01:00
Christoffer Lerno
5207022a4a For MacOS, running with higher optimization would crash as initializers were removed. 2024-03-14 09:33:24 +01:00
Christoffer Lerno
1a25746343 Regression: no stacktrace. 2024-03-12 17:31:06 +01:00
Christoffer Lerno
0d7ceb625b Fixed link on msvc. 2024-03-12 10:24:48 +01:00
Christoffer Lerno
95fb5f904f New linker build option. "system-linker" deprecated and removed from project settings. 2024-03-12 10:09:02 +01:00
Christoffer Lerno
a0309855d7 Added @link attribute. 2024-03-11 18:10:40 +01:00
Christoffer Lerno
546754e803 'output' directory for projects was incorrect in templates. 2024-03-08 15:34:04 +01:00
Christoffer Lerno
86461909d3 Remove initial './' in Win32 paths when running a binary. 2024-03-04 17:04:57 +01:00
Christoffer Lerno
feebd2a733 Bug in time.add_seconds #1162. 2024-03-01 12:09:47 +01:00
Christoffer Lerno
75e7176675 Bitstruct cast to other bitstruct by way of underlying type would fail #1159. 2024-02-26 23:51:14 +01:00
Christoffer Lerno
bae5d9c7f8 Improved checks of aliased "void". 2024-02-26 18:45:55 +01:00
Christoffer Lerno
4ba033fc84 Fix of int.min incorrect behaviour #1154. 2024-02-26 18:13:40 +01:00
Christoffer Lerno
f0dd0e8f92 Fix of CT named arguments #1156. 2024-02-26 17:47:50 +01:00
Christoffer Lerno
7ea3d230bb 0.5.5 features (#1151)
0.5.5 Disallow multiple `_` in a row in digits, e.g. `1__000`. #1138. Fixed toposort example. Struct/union members now correctly rejects members without storage size #1147. `math::pow` will now correctly promote integer arguments. `math::pow` will now correctly promote integer arguments. Added `new_aligned` and `alloc_aligned` functions to prevent accidental under-alignment when allocating simd. Pointer difference would fail where alignment != size (structs etc) #1150. Add test that overalignment actually works for lists. Fixed array calculation for npot2 vectors. Use native aligned alloc on Windows and POSIX. Deprecates "offset". Simplification of the Allocator interface.
2024-02-22 17:13:51 +01:00
Christoffer Lerno
b7f4fd9074 Create FUNDING.yml 2024-02-20 09:27:21 +01:00
Christoffer Lerno
9a114b38d3 Updated retry and test.c3 examples. 2024-02-17 15:19:27 +01:00
Christoffer Lerno
bec1116f86 Fixes to scoped mem report. 2024-02-17 11:52:18 +01:00
Christoffer Lerno
d7cc37b951 Important fixes to project settings. 2024-02-17 11:39:08 +01:00
Christoffer Lerno
4ce62cf221 Fix of allocator::new. 2024-02-16 21:55:30 +01:00
Christoffer Lerno
1f052da0b9 Clock.c3 fix. 2024-02-16 21:31:49 +01:00
Christoffer Lerno
812dc0c292 Update memory test code. 2024-02-16 14:19:42 +01:00
Christoffer Lerno
798fe0dce9 Updated lex file. 2024-02-16 12:01:13 +01:00
Christoffer Lerno
3f6fe55f9a Grammar fix. 2024-02-15 23:18:11 +01:00
Christoffer Lerno
aee4aecfe7 Remove install_win_reqs.bat from releases. 2024-02-15 22:41:44 +01:00
Christoffer Lerno
748c737e8f 0.5.4: Hash variables accept designated initializers. @safemacro overrides the need for @ in macro names. Fixes to macro context evaluation. Updated allocator api. Removed install_win_reqs.bat. Deterministic @init for MacOS. Fixed temp memory issue with formatter. Support LLVM 19. Add support to compare bitstructs using == and !=. Support Windows .def files. Removed invalid grammar from grammar.y. Support compile time folding of &|^~ for bitstructs. output project setting now respected. Fix issue where constants were not properly constant folded. Add temp_push/pop. Aliased declarations caused errors when used in initializers. Fix export output. Fix of const ternary #1118. Fix of $$MODULE in nested macros #1117. Fix debug info on globals. out now correctly detects subscript[] use #1116. Lateral implicit imports removed. Default to '.' if no libdir is specified. Improved error messages for --lib. Fix raylib snake example. Overzealous local escape check corrected #1127. Improved yacc grammar #1128. --linker argument #1067. Fixes to the matrix operations #1130. Added GenericList. 2024-02-15 21:39:33 +01:00
Christoffer Lerno
c673101bbb Fix incorrect code in sample. 2024-02-14 09:16:36 +01:00
Poly2it
d66674655c Update compilation instructions for Void Linux 2024-02-13 16:45:51 +01:00
Christian Clauss
da292e41bd msvc_build_libraries.py: Remove unused import an f-strings with no placeholder
% `ruff ` # https://docs.astral.sh/ruff
```
 Error: msvc_build_libraries.py:9:8: F401 `os` imported but unused
Error: msvc_build_libraries.py:10:8: F401 `sys` imported but unused
Error: msvc_build_libraries.py:19:8: F401 `re` imported but unused
Error: msvc_build_libraries.py:179:3: F541 f-string without any placeholders
Error: msvc_build_libraries.py:180:3: F541 f-string without any placeholders
Error: msvc_build_libraries.py:182:3: F541 f-string without any placeholders
Error: Process completed with exit code 1.
```
2024-01-18 11:24:03 +01:00
Christoffer Lerno
deb4cc7c4b 0.5.3: Single-module not respected. Fix issue with compiler defined types. Fix optimization levels for projects. Use GEP i8 on offsets. Optimize foreach on len 1 arrays. Move panic blocks last. Fix generic module wildcard imports. Deprecate init_temp / init_new. Fix issue with macro vaarg and untyped lists. Fix extern const globals. 2024-01-14 15:34:54 +01:00
Christoffer Lerno
e91f6e268e 0.5.2: Allow trailing comma in calls and parameter declarations #1092. Fixes issue where single character filenames like 'a.c3' would be rejected. Improve error messages for incorrect user defined foreach. Fix bug with generics in generics. Fix to error with modified vector parameters. Crash with lhs vector inference. Fixes to priority queue. 2023-12-23 23:15:51 +01:00
Poly2it
2595ed5cc9 Add compilation instructions for Void Linux 2023-12-09 15:15:17 +01:00
474 changed files with 21660 additions and 16207 deletions

14
.github/FUNDING.yml vendored Normal file
View File

@@ -0,0 +1,14 @@
# These are supported funding model platforms
github: [c3lang]
patreon: # Replace with a single Patreon username
open_collective: # Replace with a single Open Collective username
ko_fi: c3lang
tidelift: # Replace with a single Tidelift platform-name/package-name e.g., npm/babel
community_bridge: # Replace with a single Community Bridge project-name e.g., cloud-foundry
liberapay: # Replace with a single Liberapay username
issuehunt: # Replace with a single IssueHunt username
otechie: # Replace with a single Otechie username
lfx_crowdfunding: # Replace with a single LFX Crowdfunding project-name e.g., cloud-foundry
polar: # Replace with a single Polar username
custom: # Replace with up to 4 custom sponsorship URLs e.g., ['link1', 'link2']

View File

@@ -2,7 +2,7 @@ name: CI
on:
push:
branches: [ master, dev, ci_testing ]
branches: [ master, dev, ci_testing, experiments ]
pull_request:
branches: [ master ]
@@ -33,9 +33,9 @@ jobs:
- name: Compile and run some examples
run: |
cd resources
..\build\${{ matrix.build_type }}\c3c.exe compile-run examples\hello_world_many.c3
..\build\${{ matrix.build_type }}\c3c.exe compile-run examples\time.c3
..\build\${{ matrix.build_type }}\c3c.exe compile-run examples\fannkuch-redux.c3
..\build\${{ matrix.build_type }}\c3c.exe compile-run --print-linking examples\hello_world_many.c3
..\build\${{ matrix.build_type }}\c3c.exe compile-run --print-linking examples\time.c3
..\build\${{ matrix.build_type }}\c3c.exe compile-run --print-linking examples\fannkuch-redux.c3
..\build\${{ matrix.build_type }}\c3c.exe compile-run examples\contextfree\boolerr.c3
..\build\${{ matrix.build_type }}\c3c.exe compile-run examples\ls.c3
..\build\${{ matrix.build_type }}\c3c.exe compile-run examples\load_world.c3
@@ -61,6 +61,14 @@ jobs:
run: |
build\${{ matrix.build_type }}\c3c.exe vendor-fetch raylib
- name: Try raylib
run: |
cd resources
..\build\${{ matrix.build_type }}\c3c.exe vendor-fetch raylib
..\build\${{ matrix.build_type }}\c3c.exe compile --lib raylib --wincrt=none examples\raylib\raylib_arkanoid.c3
..\build\${{ matrix.build_type }}\c3c.exe compile --lib raylib --wincrt=none examples\raylib\raylib_snake.c3
..\build\${{ matrix.build_type }}\c3c.exe compile --lib raylib --wincrt=none examples\raylib\raylib_tetris.c3
- name: run compiler tests
run: |
cd test
@@ -71,6 +79,11 @@ jobs:
cd test
..\build\${{ matrix.build_type }}\c3c.exe compile-test unit -O1
- name: Test python script
run: |
py msvc_build_libraries.py --accept-license
dir msvc_sdk
- name: upload artifacts
uses: actions/upload-artifact@v3
with:
@@ -99,8 +112,8 @@ jobs:
install: git binutils mingw-w64-x86_64-clang mingw-w64-x86_64-ninja mingw-w64-x86_64-cmake mingw-w64-x86_64-toolchain mingw-w64-x86_64-python
- shell: msys2 {0}
run: |
pacman --noconfirm -U https://mirror.msys2.org/mingw/mingw64/mingw-w64-x86_64-llvm-17.0.6-1-any.pkg.tar.zst
pacman --noconfirm -U https://mirror.msys2.org/mingw/mingw64/mingw-w64-x86_64-lld-17.0.6-1-any.pkg.tar.zst
pacman --noconfirm -U https://mirror.msys2.org/mingw/mingw64/mingw-w64-x86_64-llvm-18.1.6-1-any.pkg.tar.zst
pacman --noconfirm -U https://mirror.msys2.org/mingw/mingw64/mingw-w64-x86_64-lld-18.1.6-1-any.pkg.tar.zst
- name: CMake
run: |
cmake -B build -G Ninja -DCMAKE_C_COMPILER=clang -DCMAKE_BUILD_TYPE=${{ matrix.build_type }}
@@ -109,11 +122,11 @@ jobs:
- name: Compile and run some examples
run: |
cd resources
../build/c3c compile-run examples/hello_world_many.c3
../build/c3c compile-run examples/time.c3
../build/c3c compile-run examples/fannkuch-redux.c3
../build/c3c compile-run examples/contextfree/boolerr.c3
../build/c3c compile-run examples/load_world.c3
../build/c3c compile-run --print-linking examples/hello_world_many.c3
../build/c3c compile-run --print-linking examples/time.c3
../build/c3c compile-run --print-linking examples/fannkuch-redux.c3
../build/c3c compile-run --print-linking examples/contextfree/boolerr.c3
../build/c3c compile-run --print-linking examples/load_world.c3
../build/c3c compile --test -g -O0 --threads 1 --target macos-x64 examples/constants.c3
- name: Build testproject
@@ -193,7 +206,7 @@ jobs:
fail-fast: false
matrix:
build_type: [Release, Debug]
llvm_version: [15, 16, 17, 18]
llvm_version: [15, 16, 17, 18, 19]
steps:
- uses: actions/checkout@v4
@@ -207,7 +220,7 @@ jobs:
if [[ "${{matrix.llvm_version}}" < 16 ]]; then
sudo apt remove libllvm15
fi
if [[ "${{matrix.llvm_version}}" < 18 ]]; then
if [[ "${{matrix.llvm_version}}" < 19 ]]; then
sudo add-apt-repository "deb http://apt.llvm.org/focal/ llvm-toolchain-focal-${{matrix.llvm_version}} main"
sudo apt-get update
sudo apt-get install -y -t llvm-toolchain-focal-${{matrix.llvm_version}} libpolly-${{matrix.llvm_version}}-dev \
@@ -222,6 +235,7 @@ jobs:
libmlir-${{matrix.llvm_version}}-dev mlir-${{matrix.llvm_version}}-tools
fi
- name: CMake
if: matrix.llvm_version != 18
run: |
cmake -B build \
-G Ninja \
@@ -234,10 +248,24 @@ jobs:
-DCMAKE_DLLTOOL=llvm-dlltool-${{matrix.llvm_version}} \
-DC3_LLVM_VERSION=${{matrix.llvm_version}}
cmake --build build
- name: CMake18
if: matrix.llvm_version == 18
run: |
cmake -B build \
-G Ninja \
-DCMAKE_BUILD_TYPE=${{matrix.build_type}} \
-DCMAKE_C_COMPILER=clang-${{matrix.llvm_version}} \
-DCMAKE_CXX_COMPILER=clang++-${{matrix.llvm_version}} \
-DCMAKE_LINKER=lld-link-${{matrix.llvm_version}} \
-DCMAKE_OBJCOPY=llvm-objcopy-${{matrix.llvm_version}} \
-DCMAKE_STRIP=llvm-strip-${{matrix.llvm_version}} \
-DCMAKE_DLLTOOL=llvm-dlltool-${{matrix.llvm_version}} \
-DC3_LLVM_VERSION=18.1
cmake --build build
- name: Compile and run some examples
run: |
cd resources
cd resources
../build/c3c compile examples/base64.c3
../build/c3c compile examples/binarydigits.c3
../build/c3c compile examples/brainfk.c3
@@ -257,7 +285,7 @@ jobs:
../build/c3c compile examples/contextfree/dynscope.c3
../build/c3c compile examples/contextfree/guess_number.c3
../build/c3c compile examples/contextfree/multi.c3
../build/c3c compile examples/contextfree/cleanup.c3
../build/c3c compile examples/contextfree/cleanup.c3
../build/c3c compile-run examples/hello_world_many.c3
../build/c3c compile-run examples/time.c3
../build/c3c compile-run examples/fannkuch-redux.c3
@@ -265,7 +293,7 @@ jobs:
../build/c3c compile-run examples/load_world.c3
../build/c3c compile-run examples/process.c3
../build/c3c compile-run examples/ls.c3
../build/c3c compile-run --system-linker=no linux_stack.c3
../build/c3c compile-run --linker=builtin linux_stack.c3
../build/c3c compile-run linux_stack.c3
- name: Compile run unit tests
@@ -281,7 +309,7 @@ jobs:
- name: Build testproject direct linker
run: |
cd resources/testproject
../../build/c3c run --debug-log --system-linker=no
../../build/c3c run --debug-log --linker=builtin
- name: run compiler tests
run: |
@@ -372,7 +400,7 @@ jobs:
../build/c3c compile-run examples/factorial_macro.c3
../build/c3c compile-run examples/fasta.c3
../build/c3c compile-run examples/process.c3
../build/c3c compile-run --system-linker=no linux_stack.c3
../build/c3c compile-run --linker=builtin linux_stack.c3
../build/c3c compile-run linux_stack.c3
- name: Compile run unit tests
@@ -388,7 +416,7 @@ jobs:
- name: Build testproject direct linker
run: |
cd resources/testproject
../../build/c3c run --debug-log --system-linker=no
../../build/c3c run --debug-log --linker=builtin
- name: run compiler tests
run: |
@@ -418,13 +446,13 @@ jobs:
fail-fast: false
matrix:
build_type: [Release, Debug]
llvm_version: [15, 16]
llvm_version: [15, 16, 17]
steps:
- uses: actions/checkout@v4
- name: Download LLVM
run: |
brew install llvm@${{ matrix.llvm_version }} ninja curl
echo "/usr/local/opt/llvm@${{ matrix.llvm_version }}/bin" >> $GITHUB_PATH
echo "/opt/homebrew/opt/llvm@${{ matrix.llvm_version }}/bin" >> $GITHUB_PATH
TMP_PATH=$(xcrun --show-sdk-path)/user/include
echo "CPATH=$TMP_PATH" >> $GITHUB_ENV
@@ -446,6 +474,7 @@ jobs:
../build/c3c compile-run examples/contextfree/boolerr.c3
../build/c3c compile-run examples/process.c3
../build/c3c compile-run examples/load_world.c3
../build/c3c compile-run -O5 examples/load_world.c3
- name: Compile run unit tests
run: |
@@ -460,7 +489,7 @@ jobs:
- name: Build testproject direct linker
run: |
cd resources/testproject
../../build/c3c run --debug-log --system-linker=no
../../build/c3c run --debug-log --linker=builtin
- name: Build testproject lib
run: |
@@ -523,8 +552,6 @@ jobs:
- run: cp -r lib c3-windows-Debug
- run: cp msvc_build_libraries.py c3-windows-Release
- run: cp msvc_build_libraries.py c3-windows-Debug
- run: cp install_win_reqs.bat c3-windows-Release
- run: cp install_win_reqs.bat c3-windows-Debug
- run: zip -r c3-windows-Release.zip c3-windows-Release
- run: zip -r c3-windows-Debug.zip c3-windows-Debug

View File

@@ -75,7 +75,7 @@ if (NOT WIN32)
find_package(CURL)
endif()
if (NOT C3_LLVM_VERSION STREQUAL "auto")
if (${C3_LLVM_VERSION} VERSION_LESS 15 OR ${C3_LLVM_VERSION} VERSION_GREATER 18)
if (${C3_LLVM_VERSION} VERSION_LESS 15 OR ${C3_LLVM_VERSION} VERSION_GREATER 19)
message(FATAL_ERROR "LLVM ${C3_LLVM_VERSION} is not supported!")
endif()
endif()

View File

@@ -32,10 +32,10 @@ whole new language.
### Example code
The following code shows [generic modules](http://www.c3-lang.org/generics/) (more examples can be found at http://www.c3-lang.org/examples/).
The following code shows [generic modules](https://c3-lang.org/references/docs/generics/) (more examples can be found at https://c3-lang.org/references/docs/examples/).
```c++
module stack <Type>;
module stack (<Type>);
// Above: the parameterized type is applied to the entire module.
struct Stack
@@ -122,7 +122,7 @@ fn void main()
- No mandatory header files
- New semantic macro system
- Module based name spacing
- Subarrays (slices)
- Slices
- Compile time reflection
- Enhanced compile time execution
- Generics based on generic modules
@@ -212,6 +212,8 @@ More platforms will be supported in the future.
3. Unzip executable and standard lib.
4. Run `./c3c`.
(*Note that there is a known issue with debug symbol generation on MacOS 13, see issue #1086)
#### Installing on Arch Linux
There is an AUR package for the c3c compiler : [c3c-git](https://aur.archlinux.org/packages/c3c-git).
@@ -319,6 +321,21 @@ You should now have a `c3c` executable.
You can try it out by running some sample code: `./c3c compile ../resources/examples/hash.c3`
#### Compiling on Void Linux
1. As root, ensure that all project dependencies are installed: `xbps-install git cmake llvm15 lld-devel libcurl-devel ncurses-devel zlib-devel libzstd-devel libxml2-devel`
2. Clone the C3C repository: `git clone https://github.com/c3lang/c3c.git`
- If you only need the latest commit, you may want to make a shallow clone instead: `git clone https://github.com/c3lang/c3c.git --depth=1`
3. Enter the directory: `cd c3c`
4. Create a build directory: `mkdir build`
5. Enter the build directory: `cd build`
6. Create the CMake build cache: `cmake ..`
7. Build: `cmake --build .`
Your c3c executable should have compiled properly. You may want to test it: `./c3c compile ../resources/examples/hash.c3`
For a sytem-wide installation, run the following as root: `cmake --install .`
#### Compiling on other Linux / Unix variants
1. Install CMake.

View File

@@ -1,17 +0,0 @@
@echo off
set DOWNLOAD_URL=https://aka.ms/vs/17/release
mkdir tmp 2> NUL
if not exist "tmp\vs_buildtools.exe" (
bitsadmin /transfer /download /priority foreground %DOWNLOAD_URL%/vs_buildtools.exe %CD%\tmp\vs_buildtools.exe
)
echo Preparing Build Tools, please wait...
tmp\vs_BuildTools.exe --quiet --wait --layout tmp\ --add Microsoft.VisualStudio.Component.Windows10SDK.19041
echo Installing Build Tools, please wait...
tmp\vs_BuildTools.exe --quiet --wait --noweb --add Microsoft.VisualStudio.Component.Windows10SDK.19041
REM rmdir tmp /s /q

View File

@@ -124,7 +124,7 @@ macro @atomic_exec(#func, data, value, ordering) @local
case RELEASE: return #func(data, value, RELEASE);
case ACQUIRE_RELEASE: return #func(data, value, ACQUIRE_RELEASE);
case SEQ_CONSISTENT: return #func(data, value, SEQ_CONSISTENT);
default: assert(false, "Ordering may not be non-atomic or unordered.");
default: unreachable("Ordering may not be non-atomic or unordered.");
}
}

View File

@@ -9,7 +9,7 @@ macro @__atomic_compare_exchange_ordering_failure(ptr, expected, desired, $succe
case AtomicOrdering.RELAXED.ordinal: return $$compare_exchange(ptr, expected, desired, false, false, $success, AtomicOrdering.RELAXED.ordinal, $alignment);
case AtomicOrdering.ACQUIRE.ordinal: return $$compare_exchange(ptr, expected, desired, false, false, $success, AtomicOrdering.ACQUIRE.ordinal, $alignment);
case AtomicOrdering.SEQ_CONSISTENT.ordinal: return $$compare_exchange(ptr, expected, desired, false, false, $success, AtomicOrdering.SEQ_CONSISTENT.ordinal, $alignment);
default: assert(false, "Unrecognized failure ordering");
default: unreachable("Unrecognized failure ordering");
}
return 0;
}
@@ -23,12 +23,12 @@ macro @__atomic_compare_exchange_ordering_success(ptr, expected, desired, succes
case AtomicOrdering.RELEASE.ordinal: return @__atomic_compare_exchange_ordering_failure(ptr, expected, desired, AtomicOrdering.RELEASE.ordinal, failure, $alignment);
case AtomicOrdering.ACQUIRE_RELEASE.ordinal: return @__atomic_compare_exchange_ordering_failure(ptr, expected, desired, AtomicOrdering.ACQUIRE_RELEASE.ordinal, failure, $alignment);
case AtomicOrdering.SEQ_CONSISTENT.ordinal: return @__atomic_compare_exchange_ordering_failure(ptr, expected, desired, AtomicOrdering.SEQ_CONSISTENT.ordinal, failure, $alignment);
default: assert(false, "Unrecognized success ordering");
default: unreachable("Unrecognized success ordering");
}
return 0;
}
fn CInt __atomic_compare_exchange(CInt size, any* ptr, any* expected, any* desired, CInt success, CInt failure) @extern("__atomic_compare_exchange") @export
fn CInt __atomic_compare_exchange(CInt size, any ptr, any expected, any desired, CInt success, CInt failure) @extern("__atomic_compare_exchange") @export
{
switch (size)
{
@@ -57,7 +57,7 @@ fn CInt __atomic_compare_exchange(CInt size, any* ptr, any* expected, any* desir
nextcase;
$endif
default:
assert(false, "Unsuported size (%d) for atomic_compare_exchange", size);
unreachable("Unsuported size (%d) for atomic_compare_exchange", size);
}
return 0;
}

View File

@@ -0,0 +1,458 @@
// Copyright (c) 2024 Christoffer Lerno. All rights reserved.
// Use of self source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
module std::collections::anylist;
import std::io,std::math;
def AnyPredicate = fn bool(any value);
def AnyTest = fn bool(any type, any context);
struct AnyList (Printable)
{
usz size;
usz capacity;
Allocator allocator;
any* entries;
}
/**
* @param initial_capacity "The initial capacity to reserve"
* @param [&inout] allocator "The allocator to use, defaults to the heap allocator"
**/
fn AnyList* AnyList.new_init(&self, usz initial_capacity = 16, Allocator allocator = allocator::heap())
{
self.allocator = allocator;
self.size = 0;
if (initial_capacity > 0)
{
initial_capacity = math::next_power_of_2(initial_capacity);
self.entries = allocator::alloc_array(allocator, any, initial_capacity);
}
else
{
self.entries = null;
}
self.capacity = initial_capacity;
return self;
}
/**
* Initialize the list using the temp allocator.
*
* @param initial_capacity "The initial capacity to reserve"
**/
fn AnyList* AnyList.temp_init(&self, usz initial_capacity = 16)
{
return self.new_init(initial_capacity, allocator::temp()) @inline;
}
fn usz! AnyList.to_format(&self, Formatter* formatter) @dynamic
{
switch (self.size)
{
case 0:
return formatter.print("[]")!;
case 1:
return formatter.printf("[%s]", self.entries[0])!;
default:
usz n = formatter.print("[")!;
foreach (i, element : self.entries[:self.size])
{
if (i != 0) formatter.print(", ")!;
n += formatter.printf("%s", element)!;
}
n += formatter.print("]")!;
return n;
}
}
fn String AnyList.to_new_string(&self, Allocator allocator = allocator::heap()) @dynamic
{
return string::new_format("%s", *self, .allocator = allocator);
}
fn String AnyList.to_tstring(&self)
{
return string::tformat("%s", *self);
}
/**
* Push an element on the list by cloning it.
**/
macro void AnyList.push(&self, element)
{
if (!self.allocator) self.allocator = allocator::heap();
self.append_internal(allocator::clone(self.allocator, element));
}
fn void AnyList.append_internal(&self, any element) @local
{
self.ensure_capacity();
self.entries[self.size++] = element;
}
/**
* Free a retained element removed using *_retained.
**/
fn void AnyList.free_element(&self, any element) @inline
{
allocator::free(self.allocator, element.ptr);
}
/**
* Pop a value who's type is known. If the type is incorrect, this
* will still pop the element.
*
* @return! CastResult.TYPE_MISMATCH, IteratorResult.NO_MORE_ELEMENT
**/
macro AnyList.pop(&self, $Type)
{
if (!self.size) return IteratorResult.NO_MORE_ELEMENT?;
defer self.free_element(self.entries[self.size]);
return *anycast(self.entries[--self.size], $Type);
}
/**
* Pop the last value and allocate the copy using the given allocator.
* @return! IteratorResult.NO_MORE_ELEMENT
**/
fn any! AnyList.new_pop(&self, Allocator allocator = allocator::heap())
{
if (!self.size) return IteratorResult.NO_MORE_ELEMENT?;
defer self.free_element(self.entries[self.size]);
return allocator::clone_any(allocator, self.entries[--self.size]);
}
/**
* Pop the last value and allocate the copy using the temp allocator
* @return! IteratorResult.NO_MORE_ELEMENT
**/
fn any! AnyList.temp_pop(&self) => self.new_pop(allocator::temp());
/**
* Pop the last value. It must later be released using list.free_element()
* @return! IteratorResult.NO_MORE_ELEMENT
**/
fn any! AnyList.pop_retained(&self)
{
if (!self.size) return IteratorResult.NO_MORE_ELEMENT?;
return self.entries[--self.size];
}
fn void AnyList.clear(&self)
{
for (usz i = 0; i < self.size; i++)
{
self.free_element(self.entries[i]);
}
self.size = 0;
}
/**
* Same as pop() but pops the first value instead.
**/
macro AnyList.pop_first(&self, $Type)
{
if (!self.size) return IteratorResult.NO_MORE_ELEMENT?;
defer self.remove_at(0);
return *anycast(self.entries[0], $Type);
}
/**
* Same as pop_retained() but pops the first value instead.
**/
fn any! AnyList.pop_first_retained(&self)
{
if (!self.size) return IteratorResult.NO_MORE_ELEMENT?;
defer self.remove_at(0);
return self.entries[0];
}
/**
* Same as new_pop() but pops the first value instead.
**/
fn any! AnyList.new_pop_first(&self, Allocator allocator = allocator::heap())
{
if (!self.size) return IteratorResult.NO_MORE_ELEMENT?;
defer self.free_element(self.entries[self.size]);
defer self.remove_at(0);
return allocator::clone_any(allocator, self.entries[0]);
}
/**
* Same as temp_pop() but pops the first value instead.
**/
fn any! AnyList.temp_pop_first(&self) => self.new_pop_first(allocator::temp());
/**
* @require index < self.size
**/
fn void AnyList.remove_at(&self, usz index)
{
if (!--self.size || index == self.size) return;
self.free_element(self.entries[index]);
self.entries[index .. self.size - 1] = self.entries[index + 1 .. self.size];
}
fn void AnyList.add_all(&self, AnyList* other_list)
{
if (!other_list.size) return;
self.reserve(other_list.size);
foreach (value : other_list)
{
self.entries[self.size++] = allocator::clone_any(self.allocator, value);
}
}
/**
* Reverse the elements in a list.
**/
fn void AnyList.reverse(&self)
{
if (self.size < 2) return;
usz half = self.size / 2U;
usz end = self.size - 1;
for (usz i = 0; i < half; i++)
{
self.swap(i, end - i);
}
}
fn any[] AnyList.array_view(&self)
{
return self.entries[:self.size];
}
/**
* Push an element to the front of the list.
**/
macro void AnyList.push_front(&self, type)
{
self.insert_at(0, type);
}
/**
* @require index < self.size
**/
macro void AnyList.insert_at(&self, usz index, type) @local
{
any value = allocator::copy(self.allocator, type);
self.insert_at_internal(self, index, value);
}
/**
* @require index < self.size
**/
fn void AnyList.insert_at_internal(&self, usz index, any value) @local
{
self.ensure_capacity();
for (usz i = self.size; i > index; i--)
{
self.entries[i] = self.entries[i - 1];
}
self.size++;
self.entries[index] = value;
}
/**
* @require self.size > 0
**/
fn void AnyList.remove_last(&self)
{
self.free_element(self.entries[--self.size]);
}
/**
* @require self.size > 0
**/
fn void AnyList.remove_first(&self)
{
self.remove_at(0);
}
macro AnyList.first(&self, $Type)
{
return *anycast(self.first_any(), $Type);
}
fn any! AnyList.first_any(&self) @inline
{
return self.size ? self.entries[0] : IteratorResult.NO_MORE_ELEMENT?;
}
macro AnyList.last(&self, $Type)
{
return *anycast(self.last_any(), $Type);
}
fn any! AnyList.last_any(&self) @inline
{
return self.size ? self.entries[self.size - 1] : IteratorResult.NO_MORE_ELEMENT?;
}
fn bool AnyList.is_empty(&self) @inline
{
return !self.size;
}
fn usz AnyList.len(&self) @operator(len) @inline
{
return self.size;
}
/**
* @require index < self.size "Index out of range"
**/
macro AnyList.get(&self, usz index, $Type)
{
return *anycast(self.entries[index], $Type);
}
/**
* @require index < self.size "Index out of range"
**/
fn any AnyList.get_any(&self, usz index) @inline
{
return self.entries[index];
}
fn void AnyList.free(&self)
{
if (!self.allocator) return;
self.clear();
allocator::free(self.allocator, self.entries);
self.capacity = 0;
self.entries = null;
}
fn void AnyList.swap(&self, usz i, usz j)
{
any temp = self.entries[i];
self.entries[i] = self.entries[j];
self.entries[j] = temp;
}
/**
* @param filter "The function to determine if it should be removed or not"
* @return "the number of deleted elements"
**/
fn usz AnyList.remove_if(&self, AnyPredicate filter)
{
return self._remove_if(filter, false);
}
/**
* @param selection "The function to determine if it should be kept or not"
* @return "the number of deleted elements"
**/
fn usz AnyList.retain_if(&self, AnyPredicate selection)
{
return self._remove_if(selection, true);
}
macro usz AnyList._remove_if(&self, AnyPredicate filter, bool $invert) @local
{
usz size = self.size;
for (usz i = size, usz k = size; k > 0; k = i)
{
// Find last index of item to be deleted.
$if $invert:
while (i > 0 && !filter(&self.entries[i - 1])) i--;
$else
while (i > 0 && filter(&self.entries[i - 1])) i--;
$endif
// Remove the items from this index up to the one not to be deleted.
usz n = self.size - k;
for (usz j = i; j < k; j++) self.free_element(self.entries[j]);
self.entries[i:n] = self.entries[k:n];
self.size -= k - i;
// Find last index of item not to be deleted.
$if $invert:
while (i > 0 && filter(&self.entries[i - 1])) i--;
$else
while (i > 0 && !filter(&self.entries[i - 1])) i--;
$endif
}
return size - self.size;
}
fn usz AnyList.remove_using_test(&self, AnyTest filter, any context)
{
return self._remove_using_test(filter, false, context);
}
fn usz AnyList.retain_using_test(&self, AnyTest filter, any context)
{
return self._remove_using_test(filter, true, context);
}
macro usz AnyList._remove_using_test(&self, AnyTest filter, bool $invert, ctx) @local
{
usz size = self.size;
for (usz i = size, usz k = size; k > 0; k = i)
{
// Find last index of item to be deleted.
$if $invert:
while (i > 0 && !filter(&self.entries[i - 1], ctx)) i--;
$else
while (i > 0 && filter(&self.entries[i - 1], ctx)) i--;
$endif
// Remove the items from this index up to the one not to be deleted.
usz n = self.size - k;
for (usz j = i; j < k; j++) self.free_element(self.entries[j]);
self.entries[i:n] = self.entries[k:n];
self.size -= k - i;
// Find last index of item not to be deleted.
$if $invert:
while (i > 0 && filter(&self.entries[i - 1], ctx)) i--;
$else
while (i > 0 && !filter(&self.entries[i - 1], ctx)) i--;
$endif
}
return size - self.size;
}
/**
* Reserve at least min_capacity
**/
fn void AnyList.reserve(&self, usz min_capacity)
{
if (!min_capacity) return;
if (self.capacity >= min_capacity) return;
if (!self.allocator) self.allocator = allocator::heap();
min_capacity = math::next_power_of_2(min_capacity);
self.entries = allocator::realloc(self.allocator, self.entries, any.sizeof * min_capacity);
self.capacity = min_capacity;
}
macro any AnyList.@item_at(&self, usz index) @operator([])
{
return self.entries[index];
}
/**
* @require index <= self.size "Index out of range"
**/
macro void AnyList.set(&self, usz index, value)
{
if (index == self.size)
{
self.push(value);
return;
}
self.free_element(self.entries[index]);
self.entries[index] = allocator::copy(self.allocator, value);
}
fn void AnyList.ensure_capacity(&self, usz added = 1) @inline @private
{
usz new_size = self.size + added;
if (self.capacity >= new_size) return;
assert(new_size < usz.max / 2U);
usz new_capacity = self.capacity ? 2U * self.capacity : 16U;
while (new_capacity < new_size) new_capacity *= 2U;
self.reserve(new_capacity);
}

View File

@@ -86,15 +86,15 @@ struct GrowableBitSet
* @param initial_capacity
* @param [&inout] allocator "The allocator to use, defaults to the heap allocator"
**/
fn GrowableBitSet* GrowableBitSet.init_new(&self, usz initial_capacity = 1, Allocator* allocator = mem::heap())
fn GrowableBitSet* GrowableBitSet.new_init(&self, usz initial_capacity = 1, Allocator allocator = allocator::heap())
{
self.data.init_new(initial_capacity, allocator);
self.data.new_init(initial_capacity, allocator);
return self;
}
fn GrowableBitSet* GrowableBitSet.init_temp(&self, usz initial_capacity = 1)
fn GrowableBitSet* GrowableBitSet.temp_init(&self, usz initial_capacity = 1)
{
return self.init_new(initial_capacity, mem::temp()) @inline;
return self.new_init(initial_capacity, allocator::temp()) @inline;
}
fn void GrowableBitSet.free(&self)

View File

@@ -1,5 +1,5 @@
module std::collections::enummap(<Enum, ValueType>);
import std::io;
struct EnumMap (Printable)
{
ValueType[Enum.len] values;
@@ -25,7 +25,7 @@ fn usz! EnumMap.to_format(&self, Formatter* formatter) @dynamic
return n;
}
fn String EnumMap.to_new_string(&self, Allocator* allocator = mem::heap()) @dynamic
fn String EnumMap.to_new_string(&self, Allocator allocator = allocator::heap()) @dynamic
{
return string::new_format("%s", *self, .allocator = allocator);
}

View File

@@ -1,4 +1,4 @@
// Copyright (c) 2021 Christoffer Lerno. All rights reserved.
// Copyright (c) 2021-2024 Christoffer Lerno. All rights reserved.
// Use of self source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
@@ -6,8 +6,9 @@
* @require Enum.kindof == TypeKind.ENUM : "Only enums maybe be used with an enumset"
**/
module std::collections::enumset(<Enum>);
import std::io;
def EnumSetType = $typefrom(private::type_for_enum_elements(Enum.elements)) @private ;
def EnumSetType = $typefrom(private::type_for_enum_elements(Enum.elements)) @private;
const IS_CHAR_ARRAY = Enum.elements > 128;
distinct EnumSet (Printable) = EnumSetType;
@@ -140,7 +141,7 @@ fn usz! EnumSet.to_format(&set, Formatter* formatter) @dynamic
return n;
}
fn String EnumSet.to_new_string(&set, Allocator* allocator = mem::heap()) @dynamic
fn String EnumSet.to_new_string(&set, Allocator allocator = allocator::heap()) @dynamic
{
return string::new_format("%s", *set, .allocator = allocator);
}

View File

@@ -1,8 +1,10 @@
// Copyright (c) 2021 Christoffer Lerno. All rights reserved.
// Copyright (c) 2021-2024 Christoffer Lerno. All rights reserved.
// Use of self source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
module std::collections::linkedlist(<Type>);
const ELEMENT_IS_EQUATABLE = types::is_equatable_type(Type);
struct Node @private
{
Node *next;
@@ -12,35 +14,26 @@ struct Node @private
struct LinkedList
{
Allocator *allocator;
Allocator allocator;
usz size;
Node *_first;
Node *_last;
}
fn void LinkedList.push(&self, Type value)
{
self.link_first(value);
}
fn void LinkedList.push_last(&self, Type value)
{
self.link_last(value);
}
/**
* @param [&inout] allocator "The allocator to use, defaults to the heap allocator"
* @return "the initialized list"
**/
fn LinkedList* LinkedList.init_new(&self, Allocator* allocator = mem::heap())
fn LinkedList* LinkedList.new_init(&self, Allocator allocator = allocator::heap())
{
*self = { .allocator = allocator };
return self;
}
fn LinkedList* LinkedList.init_temp(&self)
fn LinkedList* LinkedList.temp_init(&self)
{
return self.init_new(mem::temp()) @inline;
return self.new_init(allocator::temp()) @inline;
}
/**
@@ -48,15 +41,16 @@ fn LinkedList* LinkedList.init_temp(&self)
**/
macro void LinkedList.free_node(&self, Node* node) @private
{
self.allocator.free(node);
}
macro Node* LinkedList.alloc_node(&self) @private
{
if (!self.allocator) self.allocator = mem::heap();
return self.allocator.new(Node);
allocator::free(self.allocator, node);
}
fn void LinkedList.link_first(&self, Type value) @private
macro Node* LinkedList.alloc_node(&self) @private
{
if (!self.allocator) self.allocator = allocator::heap();
return allocator::alloc(self.allocator, Node);
}
fn void LinkedList.push_front(&self, Type value)
{
Node *first = self._first;
Node *new_node = self.alloc_node();
@@ -73,7 +67,7 @@ fn void LinkedList.link_first(&self, Type value) @private
self.size++;
}
fn void LinkedList.link_last(&self, Type value) @private
fn void LinkedList.push(&self, Type value)
{
Node *last = self._last;
Node *new_node = self.alloc_node();
@@ -157,7 +151,7 @@ fn void LinkedList.set(&self, usz index, Type element)
/**
* @require index < self.size
**/
fn void LinkedList.remove(&self, usz index)
fn void LinkedList.remove_at(&self, usz index)
{
self.unlink(self.node_at_index(index));
}
@@ -165,14 +159,14 @@ fn void LinkedList.remove(&self, usz index)
/**
* @require index <= self.size
**/
fn void LinkedList.insert(&self, usz index, Type element)
fn void LinkedList.insert_at(&self, usz index, Type element)
{
switch (index)
{
case 0:
self.push(element);
self.push_front(element);
case self.size:
self.push_last(element);
self.push(element);
default:
self.link_before(self.node_at_index(index), element);
}
@@ -217,7 +211,53 @@ fn void LinkedList.unlink_first(&self) @private
self.size--;
}
fn bool LinkedList.remove_value(&self, Type t)
fn usz LinkedList.remove(&self, Type t) @if(ELEMENT_IS_EQUATABLE)
{
usz start = self.size;
Node* node = self._first;
while (node)
{
switch
{
case equals(node.value, t):
Node* next = node.next;
self.unlink(node);
node = next;
default:
node = node.next;
}
}
return start - self.size;
}
fn Type! LinkedList.pop(&self)
{
if (!self._last) return IteratorResult.NO_MORE_ELEMENT?;
defer self.unlink_last();
return self._last.value;
}
fn Type! LinkedList.pop_front(&self)
{
if (!self._first) return IteratorResult.NO_MORE_ELEMENT?;
defer self.unlink_first();
return self._first.value;
}
fn void! LinkedList.remove_last(&self) @maydiscard
{
if (!self._first) return IteratorResult.NO_MORE_ELEMENT?;
self.unlink_last();
}
fn void! LinkedList.remove_first(&self) @maydiscard
{
if (!self._first) return IteratorResult.NO_MORE_ELEMENT?;
self.unlink_first();
}
fn bool LinkedList.remove_first_match(&self, Type t) @if(ELEMENT_IS_EQUATABLE)
{
for (Node* node = self._first; node != null; node = node.next)
{
@@ -230,7 +270,7 @@ fn bool LinkedList.remove_value(&self, Type t)
return false;
}
fn bool LinkedList.remove_last_value(&self, Type t)
fn bool LinkedList.remove_last_match(&self, Type t) @if(ELEMENT_IS_EQUATABLE)
{
for (Node* node = self._last; node != null; node = node.prev)
{
@@ -242,26 +282,6 @@ fn bool LinkedList.remove_last_value(&self, Type t)
}
return false;
}
fn Type! LinkedList.pop(&self)
{
if (!self._first) return IteratorResult.NO_MORE_ELEMENT?;
defer self.unlink_first();
return self._first.value;
}
fn void! LinkedList.remove_last(&self)
{
if (!self._first) return IteratorResult.NO_MORE_ELEMENT?;
self.unlink_last();
}
fn void! LinkedList.remove_first(&self)
{
if (!self._first) return IteratorResult.NO_MORE_ELEMENT?;
self.unlink_first();
}
/**
* @require self._last
**/

View File

@@ -1,12 +1,11 @@
// Copyright (c) 2021 Christoffer Lerno. All rights reserved.
// Copyright (c) 2021-2024 Christoffer Lerno. All rights reserved.
// Use of self source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
module std::collections::list(<Type>);
import std::io;
import std::math;
import std::io,std::math;
def ElementPredicate = fn bool(Type *type);
def ElementTest = fn bool(Type *type, any* context);
def ElementTest = fn bool(Type *type, any context);
const ELEMENT_IS_EQUATABLE = types::is_equatable_type(Type);
const ELEMENT_IS_POINTER = Type.kindof == POINTER;
@@ -14,22 +13,23 @@ struct List (Printable)
{
usz size;
usz capacity;
Allocator *allocator;
Allocator allocator;
Type *entries;
}
/**
* @param initial_capacity "The initial capacity to reserve"
* @param [&inout] allocator "The allocator to use, defaults to the heap allocator"
**/
fn List* List.init_new(&self, usz initial_capacity = 16, Allocator* allocator = mem::heap())
fn List* List.new_init(&self, usz initial_capacity = 16, Allocator allocator = allocator::heap())
{
self.allocator = allocator;
self.size = 0;
if (initial_capacity > 0)
{
initial_capacity = math::next_power_of_2(initial_capacity);
self.entries = allocator.alloc_aligned(Type.sizeof * initial_capacity, .alignment = Type[1].alignof)!!;
self.entries = allocator::malloc_aligned(allocator, Type.sizeof * initial_capacity, .alignment = Type[1].alignof)!!;
}
else
{
@@ -44,15 +44,15 @@ fn List* List.init_new(&self, usz initial_capacity = 16, Allocator* allocator =
*
* @param initial_capacity "The initial capacity to reserve"
**/
fn List* List.init_temp(&self, usz initial_capacity = 16)
fn List* List.temp_init(&self, usz initial_capacity = 16)
{
return self.init_new(initial_capacity, mem::temp()) @inline;
return self.new_init(initial_capacity, allocator::temp()) @inline;
}
/**
* @require self.size == 0 "The List must be empty"
**/
fn void List.init_wrapping_array(&self, Type[] types, Allocator* allocator = mem::heap())
fn void List.init_wrapping_array(&self, Type[] types, Allocator allocator = allocator::heap())
{
self.allocator = allocator;
self.size = types.len;
@@ -80,7 +80,7 @@ fn usz! List.to_format(&self, Formatter* formatter) @dynamic
}
}
fn String List.to_new_string(&self, Allocator* allocator = mem::heap()) @dynamic
fn String List.to_new_string(&self, Allocator allocator = allocator::heap()) @dynamic
{
return string::new_format("%s", *self, .allocator = allocator);
}
@@ -91,21 +91,14 @@ fn String List.to_tstring(&self)
}
fn void List.push(&self, Type element) @inline
{
self.append(element);
}
fn void List.append(&self, Type element)
{
self.ensure_capacity();
self.entries[self.size++] = element;
self.entries[self.size++] = element;
}
/**
* @require self.size > 0
**/
fn Type List.pop(&self)
fn Type! List.pop(&self)
{
if (!self.size) return IteratorResult.NO_MORE_ELEMENT?;
return self.entries[--self.size];
}
@@ -117,11 +110,11 @@ fn void List.clear(&self)
/**
* @require self.size > 0
**/
fn Type List.pop_first(&self)
fn Type! List.pop_first(&self)
{
Type value = self.entries[0];
self.remove_at(0);
return value;
if (!self.size) return IteratorResult.NO_MORE_ELEMENT?;
defer self.remove_at(0);
return self.entries[0];
}
/**
@@ -129,11 +122,8 @@ fn Type List.pop_first(&self)
**/
fn void List.remove_at(&self, usz index)
{
for (usz i = index + 1; i < self.size; i++)
{
self.entries[i - 1] = self.entries[i];
}
self.size--;
if (!--self.size || index == self.size) return;
self.entries[index .. self.size - 1] = self.entries[index + 1 .. self.size];
}
fn void List.add_all(&self, List* other_list)
@@ -147,17 +137,17 @@ fn void List.add_all(&self, List* other_list)
}
fn Type[] List.to_new_array(&self, Allocator* allocator = mem::heap())
fn Type[] List.to_new_array(&self, Allocator allocator = allocator::heap())
{
if (!self.size) return Type[] {};
Type[] result = allocator.new_array(Type, self.size);
Type[] result = allocator::alloc_array(allocator, Type, self.size);
result[..] = self.entries[:self.size];
return result;
}
fn Type[] List.to_tarray(&self)
{
return self.to_new_array(mem::temp());
return self.to_new_array(allocator::temp());
}
/**
@@ -216,30 +206,28 @@ fn void List.set_at(&self, usz index, Type type)
self.entries[index] = type;
}
/**
* @require self.size > 0
**/
fn void List.remove_last(&self)
fn void! List.remove_last(&self) @maydiscard
{
if (!self.size) return IteratorResult.NO_MORE_ELEMENT?;
self.size--;
}
/**
* @require self.size > 0
**/
fn void List.remove_first(&self)
fn void! List.remove_first(&self) @maydiscard
{
if (!self.size) return IteratorResult.NO_MORE_ELEMENT?;
self.remove_at(0);
}
fn Type* List.first(&self)
fn Type! List.first(&self)
{
return self.size ? &self.entries[0] : null;
if (!self.size) return IteratorResult.NO_MORE_ELEMENT?;
return self.entries[0];
}
fn Type* List.last(&self)
fn Type! List.last(&self)
{
return self.size ? &self.entries[self.size - 1] : null;
if (!self.size) return IteratorResult.NO_MORE_ELEMENT?;
return self.entries[self.size - 1];
}
fn bool List.is_empty(&self) @inline
@@ -247,6 +235,11 @@ fn bool List.is_empty(&self) @inline
return !self.size;
}
fn usz List.byte_size(&self) @inline
{
return Type.sizeof * self.size;
}
fn usz List.len(&self) @operator(len) @inline
{
return self.size;
@@ -260,7 +253,7 @@ fn Type List.get(&self, usz index) @inline
fn void List.free(&self)
{
if (!self.allocator) return;
self.allocator.free_aligned(self.entries);
allocator::free_aligned(self.allocator, self.entries);
self.capacity = 0;
self.size = 0;
self.entries = null;
@@ -314,12 +307,12 @@ macro usz List._remove_if(&self, ElementPredicate filter, bool $invert) @local
return size - self.size;
}
fn usz List.remove_using_test(&self, ElementTest filter, any* context)
fn usz List.remove_using_test(&self, ElementTest filter, any context)
{
return self._remove_using_test(filter, false, context);
}
fn usz List.retain_using_test(&self, ElementTest filter, any* context)
fn usz List.retain_using_test(&self, ElementTest filter, any context)
{
return self._remove_using_test(filter, true, context);
}
@@ -356,9 +349,9 @@ fn void List.reserve(&self, usz min_capacity)
{
if (!min_capacity) return;
if (self.capacity >= min_capacity) return;
if (!self.allocator) self.allocator = mem::heap();
if (!self.allocator) self.allocator = allocator::heap();
min_capacity = math::next_power_of_2(min_capacity);
self.entries = self.allocator.realloc_aligned(self.entries, Type.sizeof * min_capacity, .alignment = Type[1].alignof) ?? null;
self.entries = allocator::realloc_aligned(self.allocator, self.entries, Type.sizeof * min_capacity, .alignment = Type[1].alignof) ?? null;
self.capacity = min_capacity;
}
@@ -436,18 +429,38 @@ fn bool List.contains(&self, Type value) @if(ELEMENT_IS_EQUATABLE)
}
/**
* @param [&inout] self "The list to remove elements from"
* @param value "The value to remove"
* @return "true if the value was found"
**/
fn bool List.remove_last_match(&self, Type value) @if(ELEMENT_IS_EQUATABLE)
{
return @ok(self.remove_at(self.rindex_of(value)));
}
/**
* @param [&inout] self "The list to remove elements from"
* @param value "The value to remove"
* @return "true if the value was found"
**/
fn bool List.remove_first_match(&self, Type value) @if(ELEMENT_IS_EQUATABLE)
{
return @ok(self.remove_at(self.index_of(value)));
}
/**
* @param [&inout] self "The list to remove elements from"
* @param value "The value to remove"
* @return "the number of deleted elements."
**/
fn usz List.remove(&self, Type value) @if(ELEMENT_IS_EQUATABLE)
fn usz List.remove_all_matches(&self, Type value) @if(ELEMENT_IS_EQUATABLE)
{
usz size = self.size;
for (usz i = size; i > 0; i--)
{
if (!equals(self.entries[i - 1], value)) continue;
for (usz j = i; j < size; j++)
for (usz j = i; j < self.size; j++)
{
self.entries[j - 1] = self.entries[j];
}
@@ -456,10 +469,10 @@ fn usz List.remove(&self, Type value) @if(ELEMENT_IS_EQUATABLE)
return size - self.size;
}
fn void List.remove_all(&self, List* other_list) @if(ELEMENT_IS_EQUATABLE)
fn void List.remove_all_from(&self, List* other_list) @if(ELEMENT_IS_EQUATABLE)
{
if (!other_list.size) return;
foreach (v : other_list) self.remove(v);
foreach (v : other_list) self.remove_all_matches(v);
}
/**

View File

@@ -13,7 +13,7 @@ const bool COPY_KEYS = types::implements_copy(Key);
struct HashMap
{
Entry*[] table;
Allocator* allocator;
Allocator allocator;
uint count; // Number of elements
uint threshold; // Resize limit
float load_factor;
@@ -23,30 +23,31 @@ struct HashMap
* @param [&inout] allocator "The allocator to use"
* @require capacity > 0 "The capacity must be 1 or higher"
* @require load_factor > 0.0 "The load factor must be higher than 0"
* @require !map.allocator "Map was already initialized"
* @require !self.allocator "Map was already initialized"
* @require capacity < MAXIMUM_CAPACITY "Capacity cannot exceed maximum"
**/
fn HashMap* HashMap.init_new(&map, uint capacity = DEFAULT_INITIAL_CAPACITY, float load_factor = DEFAULT_LOAD_FACTOR, Allocator* allocator = mem::heap())
fn HashMap* HashMap.new_init(&self, uint capacity = DEFAULT_INITIAL_CAPACITY, float load_factor = DEFAULT_LOAD_FACTOR, Allocator allocator = allocator::heap())
{
capacity = math::next_power_of_2(capacity);
map.allocator = allocator;
map.load_factor = load_factor;
map.threshold = (uint)(capacity * load_factor);
map.table = allocator.new_zero_array(Entry*, capacity);
return map;
self.allocator = allocator;
self.load_factor = load_factor;
self.threshold = (uint)(capacity * load_factor);
self.table = allocator::new_array(allocator, Entry*, capacity);
return self;
}
/**
* @require capacity > 0 "The capacity must be 1 or higher"
* @require load_factor > 0.0 "The load factor must be higher than 0"
* @require !map.allocator "Map was already initialized"
* @require !self.allocator "Map was already initialized"
* @require capacity < MAXIMUM_CAPACITY "Capacity cannot exceed maximum"
**/
fn HashMap* HashMap.init_temp(&map, uint capacity = DEFAULT_INITIAL_CAPACITY, float load_factor = DEFAULT_LOAD_FACTOR)
fn HashMap* HashMap.temp_init(&self, uint capacity = DEFAULT_INITIAL_CAPACITY, float load_factor = DEFAULT_LOAD_FACTOR)
{
return map.init_new(capacity, load_factor, mem::temp());
return self.new_init(capacity, load_factor, allocator::temp()) @inline;
}
/**
* Has this hash map been initialized yet?
*
@@ -62,9 +63,9 @@ fn bool HashMap.is_initialized(&map)
* @param [&inout] allocator "The allocator to use"
* @param [&in] other_map "The map to copy from."
**/
fn HashMap* HashMap.init_new_from_map(&self, HashMap* other_map, Allocator* allocator = mem::heap())
fn HashMap* HashMap.new_init_from_map(&self, HashMap* other_map, Allocator allocator = allocator::heap())
{
self.init_new(other_map.table.len, other_map.load_factor, allocator);
self.new_init(other_map.table.len, other_map.load_factor, allocator);
self.put_all_for_create(other_map);
return self;
}
@@ -72,9 +73,9 @@ fn HashMap* HashMap.init_new_from_map(&self, HashMap* other_map, Allocator* allo
/**
* @param [&in] other_map "The map to copy from."
**/
fn HashMap* HashMap.init_temp_from_map(&map, HashMap* other_map)
fn HashMap* HashMap.temp_init_from_map(&map, HashMap* other_map)
{
return map.init_new_from_map(other_map, mem::temp()) @inline;
return map.new_init_from_map(other_map, allocator::temp()) @inline;
}
fn bool HashMap.is_empty(&map) @inline
@@ -111,6 +112,7 @@ fn Entry*! HashMap.get_entry(&map, Key key)
/**
* Get the value or update and
* @require $assignable(#expr, Value)
**/
macro Value HashMap.@get_or_set(&map, Key key, Value #expr)
{
@@ -146,7 +148,7 @@ fn bool HashMap.set(&map, Key key, Value value) @operator([]=)
// If the map isn't initialized, use the defaults to initialize it.
if (!map.allocator)
{
map.init_new();
map.new_init();
}
uint hash = rehash(key.hash());
uint index = index_for(hash, map.table.len);
@@ -197,14 +199,14 @@ fn void HashMap.free(&map)
fn Key[] HashMap.key_tlist(&map)
{
return map.key_new_list(mem::temp()) @inline;
return map.key_new_list(allocator::temp()) @inline;
}
fn Key[] HashMap.key_new_list(&map, Allocator* allocator = mem::heap())
fn Key[] HashMap.key_new_list(&map, Allocator allocator = allocator::heap())
{
if (!map.count) return {};
Key[] list = allocator.new_array(Key, map.count);
Key[] list = allocator::alloc_array(allocator, Key, map.count);
usz index = 0;
foreach (Entry* entry : map.table)
{
@@ -241,13 +243,13 @@ macro HashMap.@each_entry(map; @body(entry))
fn Value[] HashMap.value_tlist(&map)
{
return map.value_new_list(mem::temp()) @inline;
return map.value_new_list(allocator::temp()) @inline;
}
fn Value[] HashMap.value_new_list(&map, Allocator* allocator = mem::heap())
fn Value[] HashMap.value_new_list(&map, Allocator allocator = allocator::heap())
{
if (!map.count) return {};
Value[] list = allocator.new_array(Value, map.count);
Value[] list = allocator::alloc_array(allocator, Value, map.count);
usz index = 0;
foreach (Entry* entry : map.table)
{
@@ -278,11 +280,10 @@ fn bool HashMap.has_value(&map, Value v) @if(VALUE_IS_EQUATABLE)
fn void HashMap.add_entry(&map, uint hash, Key key, Value value, uint bucket_index) @private
{
Entry* entry = map.allocator.new(Entry);
$if COPY_KEYS:
key = key.copy(map.allocator);
$endif
*entry = { .hash = hash, .key = key, .value = value, .next = map.table[bucket_index] };
Entry* entry = allocator::new(map.allocator, Entry, { .hash = hash, .key = key, .value = value, .next = map.table[bucket_index] });
map.table[bucket_index] = entry;
if (map.count++ >= map.threshold)
{
@@ -299,7 +300,7 @@ fn void HashMap.resize(&map, uint new_capacity) @private
map.threshold = uint.max;
return;
}
Entry*[] new_table = map.allocator.new_zero_array(Entry*, new_capacity);
Entry*[] new_table = allocator::new_array(map.allocator, Entry*, new_capacity);
map.transfer(new_table);
map.table = new_table;
map.free_internal(old_table.ptr);
@@ -363,7 +364,7 @@ fn void HashMap.put_for_create(&map, Key key, Value value) @private
fn void HashMap.free_internal(&map, void* ptr) @inline @private
{
map.allocator.free(ptr);
allocator::free(map.allocator, ptr);
}
fn bool HashMap.remove_entry_for_key(&map, Key key) @private
@@ -398,11 +399,10 @@ fn bool HashMap.remove_entry_for_key(&map, Key key) @private
fn void HashMap.create_entry(&map, uint hash, Key key, Value value, int bucket_index) @private
{
Entry *e = map.table[bucket_index];
Entry* entry = map.allocator.new(Entry);
$if COPY_KEYS:
key = key.copy(map.allocator);
$endif
*entry = { .hash = hash, .key = key, .value = value, .next = map.table[bucket_index] };
Entry* entry = allocator::new(map.allocator, Entry, { .hash = hash, .key = key, .value = value, .next = map.table[bucket_index] });
map.table[bucket_index] = entry;
map.count++;
}
@@ -410,7 +410,7 @@ fn void HashMap.create_entry(&map, uint hash, Key key, Value value, int bucket_i
fn void HashMap.free_entry(&self, Entry *entry) @local
{
$if COPY_KEYS:
self.allocator.free(entry.key);
allocator::free(self.allocator, entry.key);
$endif
self.free_internal(entry);
}

View File

@@ -2,9 +2,7 @@
// Use of this source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
module std::collections::object;
import std::collections::map;
import std::collections::list;
import std::io;
import std::collections::map, std::collections::list, std::io;
const Object TRUE_OBJECT = { .b = true, .type = bool.typeid };
const Object FALSE_OBJECT = { .b = false, .type = bool.typeid };
@@ -13,7 +11,7 @@ const Object NULL_OBJECT = { .type = void*.typeid };
struct Object (Printable)
{
typeid type;
Allocator* allocator;
Allocator allocator;
union
{
uint128 i;
@@ -78,11 +76,9 @@ fn usz! Object.to_format(&self, Formatter* formatter) @dynamic
}
}
fn Object* new_obj(Allocator* allocator)
fn Object* new_obj(Allocator allocator)
{
Object* o = allocator.new(Object);
*o = { .allocator = allocator, .type = void.typeid };
return o;
return allocator::new(allocator, Object, { .allocator = allocator, .type = void.typeid });
}
fn Object* new_null()
@@ -90,32 +86,24 @@ fn Object* new_null()
return &NULL_OBJECT;
}
fn Object* new_int(int128 i, Allocator* allocator)
fn Object* new_int(int128 i, Allocator allocator)
{
Object* o = allocator.new(Object);
*o = { .i = i, .allocator = allocator, .type = int128.typeid };
return o;
return allocator::new(allocator, Object, { .i = i, .allocator = allocator, .type = int128.typeid });
}
macro Object* new_enum(e, Allocator* allocator)
macro Object* new_enum(e, Allocator allocator)
{
Object* o = allocator.new(Object);
*o = { .i = (int128)e, .allocator = allocator, .type = @typeid(e) };
return o;
return allocator::new(allocator, Object, { .i = (int128)e, .allocator = allocator, .type = @typeid(e) });
}
fn Object* new_float(double f, Allocator* allocator)
fn Object* new_float(double f, Allocator allocator)
{
Object* o = allocator.new(Object);
*o = { .f = f, .allocator = allocator, .type = double.typeid };
return o;
return allocator::new(allocator, Object, { .f = f, .allocator = allocator, .type = double.typeid });
}
fn Object* new_string(String s, Allocator* allocator)
fn Object* new_string(String s, Allocator allocator)
{
Object* o = allocator.new(Object);
*o = { .s = s.copy(allocator), .allocator = allocator, .type = String.typeid };
return o;
return allocator::new(allocator, Object, { .s = s.copy(allocator), .allocator = allocator, .type = String.typeid });
}
@@ -131,7 +119,7 @@ fn void Object.free(&self)
case void:
break;
case String:
self.allocator.free(self.s);
allocator::free(self.allocator, self.s);
case ObjectInternalList:
foreach (ol : self.array)
{
@@ -140,13 +128,13 @@ fn void Object.free(&self)
self.array.free();
case ObjectInternalMap:
self.map.@each_entry(; ObjectInternalMapEntry* entry) {
self.allocator.free(entry.key);
allocator::free(self.allocator, entry.key);
entry.value.free();
};
default:
break;
}
if (self.allocator) self.allocator.free(self);
if (self.allocator) allocator::free(self.allocator, self);
}
fn bool Object.is_null(&self) @inline => self == &NULL_OBJECT;
@@ -168,7 +156,7 @@ fn void Object.init_map_if_needed(&self) @private
if (self.is_empty())
{
self.type = ObjectInternalMap.typeid;
self.map.init_new(.allocator = self.allocator);
self.map.new_init(.allocator = self.allocator);
}
}
@@ -180,7 +168,7 @@ fn void Object.init_array_if_needed(&self) @private
if (self.is_empty())
{
self.type = ObjectInternalList.typeid;
self.array.init_new(.allocator = self.allocator);
self.array.new_init(.allocator = self.allocator);
}
}
@@ -193,7 +181,7 @@ fn void Object.set_object(&self, String key, Object* new_object) @private
ObjectInternalMapEntry*! entry = self.map.get_entry(key);
defer
{
(void)self.allocator.free(entry.key);
(void)allocator::free(self.allocator, entry.key);
(void)entry.value.free();
}
self.map.set(key.copy(self.map.allocator), new_object);
@@ -246,10 +234,10 @@ macro Object* Object.set_at(&self, usz index, String key, value)
* @require self.is_indexable()
* @ensure return != null
**/
macro Object* Object.append(&self, value)
macro Object* Object.push(&self, value)
{
Object* val = self.object_from_value(value);
self.append_object(val);
self.push_object(val);
return val;
}
@@ -280,10 +268,10 @@ fn usz Object.get_len(&self)
/**
* @require self.is_indexable()
**/
fn void Object.append_object(&self, Object* to_append)
fn void Object.push_object(&self, Object* to_append)
{
self.init_array_if_needed();
self.array.append(to_append);
self.array.push(to_append);
}
/**
@@ -294,11 +282,11 @@ fn void Object.set_object_at(&self, usz index, Object* to_set)
self.init_array_if_needed();
while (self.array.len() < index)
{
self.array.append(&NULL_OBJECT);
self.array.push(&NULL_OBJECT);
}
if (self.array.len() == index)
{
self.array.append(to_set);
self.array.push(to_set);
return;
}
self.array.get(index).free();

View File

@@ -27,7 +27,7 @@ distinct PriorityQueue = inline PrivatePriorityQueue(<Type, false>);
distinct PriorityQueueMax = inline PrivatePriorityQueue(<Type, true>);
module std::collections::priorityqueue::private(<Type, MAX>);
import std::collections::list;
import std::collections::list, std::io;
def Heap = List(<Type>);
@@ -36,14 +36,14 @@ struct PrivatePriorityQueue (Printable)
Heap heap;
}
fn void PrivatePriorityQueue.init_new(&self, usz initial_capacity = 16, Allocator* allocator = mem::heap()) @inline
fn void PrivatePriorityQueue.new_init(&self, usz initial_capacity = 16, Allocator allocator = allocator::heap()) @inline
{
self.heap.init_new(initial_capacity, allocator);
self.heap.new_init(initial_capacity, allocator);
}
fn void PrivatePriorityQueue.init_temp(&self, usz initial_capacity = 16) @inline
fn void PrivatePriorityQueue.temp_init(&self, usz initial_capacity = 16) @inline
{
self.heap.init_new(initial_capacity, mem::temp()) @inline;
self.heap.new_init(initial_capacity, allocator::temp()) @inline;
}
fn void PrivatePriorityQueue.push(&self, Type element)
@@ -74,29 +74,32 @@ fn Type! PrivatePriorityQueue.pop(&self)
usz i = 0;
usz len = self.heap.len();
if (!len) return IteratorResult.NO_MORE_ELEMENT?;
usz newCount = len - 1;
self.heap.swap(0, newCount);
while ((2 * i + 1) < newCount)
usz new_count = len - 1;
self.heap.swap(0, new_count);
while OUTER: ((2 * i + 1) < new_count)
{
usz j = 2 * i + 1;
Type itemj = self.heap[j];
if ((j + 1) < newCount)
{
Type nextj = self.heap[j + 1];
$if MAX:
bool ok = greater(nextj, itemj);
$else
bool ok = less(nextj, itemj);
$endif
if (ok) j++;
}
Type left = self.heap[j];
Type item = self.heap[i];
$if MAX:
bool ok = less(item, itemj);
$else
bool ok = greater(item, itemj);
$endif
if (!ok) break;
switch
{
case j + 1 < new_count:
Type right = self.heap[j + 1];
$if MAX:
if (!greater(right, left)) nextcase;
if (!greater(right, item)) break OUTER;
$else
if (!greater(left, right)) nextcase;
if (!greater(item, right)) break OUTER;
$endif
j++;
default:
$if MAX:
if (!greater(left, item)) break OUTER;
$else
if (!greater(item, left)) break OUTER;
$endif
}
self.heap.swap(i, j);
i = j;
}
@@ -104,7 +107,7 @@ fn Type! PrivatePriorityQueue.pop(&self)
return self.heap.pop();
}
fn Type! PrivatePriorityQueue.peek(&self)
fn Type! PrivatePriorityQueue.first(&self)
{
if (!self.len()) return IteratorResult.NO_MORE_ELEMENT?;
return self.heap.get(0);
@@ -128,7 +131,7 @@ fn bool PrivatePriorityQueue.is_empty(&self)
/**
* @require index < self.len()
*/
fn Type PrivatePriorityQueue.peek_at(&self, usz index) @operator([])
fn Type PrivatePriorityQueue.get(&self, usz index) @operator([])
{
return self.heap[index];
}
@@ -138,7 +141,7 @@ fn usz! PrivatePriorityQueue.to_format(&self, Formatter* formatter) @dynamic
return self.heap.to_format(formatter);
}
fn String PrivatePriorityQueue.to_new_string(&self, Allocator* allocator = mem::heap()) @dynamic
fn String PrivatePriorityQueue.to_new_string(&self, Allocator allocator = allocator::heap()) @dynamic
{
return self.heap.to_new_string(allocator);
}

View File

@@ -2,6 +2,7 @@
* @require Type.is_ordered : "The type must be ordered"
**/
module std::collections::range(<Type>);
import std::io;
struct Range (Printable)
{
@@ -28,14 +29,14 @@ fn Type Range.get(&self, usz index) @operator([])
return (Type)(self.start + (usz)index);
}
fn String Range.to_new_string(&self, Allocator* allocator = mem::heap()) @dynamic
fn String Range.to_new_string(&self, Allocator allocator = allocator::heap()) @dynamic
{
return string::new_format("[%s..%s]", self.start, self.end, .allocator = allocator);
}
fn String Range.to_tstring(&self)
{
return self.to_new_string(mem::temp());
return self.to_new_string(allocator::temp());
}
fn usz! Range.to_format(&self, Formatter* formatter) @dynamic
@@ -65,14 +66,14 @@ fn usz! ExclusiveRange.to_format(&self, Formatter* formatter) @dynamic
return formatter.printf("[%s..<%s]", self.start, self.end)!;
}
fn String ExclusiveRange.to_new_string(&self, Allocator* allocator = mem::heap()) @dynamic
fn String ExclusiveRange.to_new_string(&self, Allocator allocator = allocator::heap()) @dynamic
{
return string::new_format("[%s..<%s]", self.start, self.end, .allocator = allocator);
}
fn String ExclusiveRange.to_tstring(&self)
{
return self.to_new_string(mem::temp());
return self.to_new_string(allocator::temp());
}
/**

View File

@@ -12,7 +12,7 @@ fn void RingBuffer.init(&self) @inline
*self = {};
}
fn void RingBuffer.putc(&self, Type c)
fn void RingBuffer.push(&self, Type c)
{
if (self.written < SIZE)
{
@@ -26,7 +26,7 @@ fn void RingBuffer.putc(&self, Type c)
}
}
fn Type RingBuffer.getc(&self, usz index)
fn Type RingBuffer.get(&self, usz index) @operator([])
{
index %= SIZE;
usz avail = SIZE - self.head;
@@ -37,7 +37,7 @@ fn Type RingBuffer.getc(&self, usz index)
return self.buf[index - avail];
}
fn Type! RingBuffer.popc(&self)
fn Type! RingBuffer.pop(&self)
{
switch
{
@@ -52,7 +52,7 @@ fn Type! RingBuffer.popc(&self)
}
}
fn usz RingBuffer.get(&self, usz index, Type[] buffer)
fn usz RingBuffer.read(&self, usz index, Type[] buffer)
{
index %= SIZE;
if (self.written < SIZE)
@@ -87,7 +87,7 @@ fn usz RingBuffer.get(&self, usz index, Type[] buffer)
return n1 + n2;
}
fn void RingBuffer.push(&self, Type[] buffer)
fn void RingBuffer.write(&self, Type[] buffer)
{
usz i;
while (self.written < SIZE && i < buffer.len)

View File

@@ -2,6 +2,7 @@
// Use of this source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
module std::core::mem::allocator;
import std::math;
struct ArenaAllocator (Allocator)
{
@@ -29,9 +30,11 @@ struct ArenaAllocatorHeader @local
char[*] data;
}
/*
* @require ptr != null
**/
fn void ArenaAllocator.release(&self, void* ptr, bool) @dynamic
{
if (!ptr) return;
assert((uptr)ptr >= (uptr)self.data.ptr, "Pointer originates from a different allocator.");
ArenaAllocatorHeader* header = ptr - ArenaAllocatorHeader.sizeof;
// Reclaim memory if it's the last element.
@@ -40,53 +43,40 @@ fn void ArenaAllocator.release(&self, void* ptr, bool) @dynamic
self.used -= header.size + ArenaAllocatorHeader.sizeof;
}
}
fn usz ArenaAllocator.mark(&self) @dynamic => self.used;
fn void ArenaAllocator.reset(&self, usz mark) @dynamic => self.used = mark;
/**
* @require !alignment || math::is_power_of_2(alignment)
* @require alignment <= mem::MAX_MEMORY_ALIGNMENT `alignment too big`
* @require offset <= mem::MAX_MEMORY_ALIGNMENT `offset too big`
* @require offset <= size && offset >= 0
* @require mem::aligned_offset(offset, ArenaAllocatorHeader.alignof) == offset
* @require size > 0
**/
fn void*! ArenaAllocator.acquire(&self, usz size, bool clear, usz alignment, usz offset) @dynamic
fn void*! ArenaAllocator.acquire(&self, usz size, AllocInitType init_type, usz alignment) @dynamic
{
if (!size) return null;
alignment = alignment_for_allocation(alignment);
usz total_len = self.data.len;
if (size > total_len) return AllocationFailure.CHUNK_TOO_LARGE?;
void* start_mem = self.data.ptr;
void* unaligned_pointer_to_offset = start_mem + self.used + ArenaAllocatorHeader.sizeof + offset;
void* aligned_pointer_to_offset = mem::aligned_pointer(unaligned_pointer_to_offset, alignment);
usz end = (usz)(aligned_pointer_to_offset - self.data.ptr) + size - offset;
void* unaligned_pointer_to_offset = start_mem + self.used + ArenaAllocatorHeader.sizeof;
void* mem = mem::aligned_pointer(unaligned_pointer_to_offset, alignment);
usz end = (usz)(mem - self.data.ptr) + size;
if (end > total_len) return AllocationFailure.OUT_OF_MEMORY?;
self.used = end;
void* mem = aligned_pointer_to_offset - offset;
ArenaAllocatorHeader* header = mem - ArenaAllocatorHeader.sizeof;
header.size = size;
if (clear) mem::clear(mem, size, mem::DEFAULT_MEM_ALIGNMENT);
if (init_type == ZERO) mem::clear(mem, size, mem::DEFAULT_MEM_ALIGNMENT);
return mem;
}
/**
* @require !alignment || math::is_power_of_2(alignment)
* @require alignment <= mem::MAX_MEMORY_ALIGNMENT `alignment too big`
* @require offset <= mem::MAX_MEMORY_ALIGNMENT `offset too big`
* @require offset <= size && offset >= 0
* @require mem::aligned_offset(offset, ArenaAllocatorHeader.alignof) == offset
* @require old_pointer != null
* @require size > 0
**/
fn void*! ArenaAllocator.resize(&self, void *old_pointer, usz size, usz alignment, usz offset) @dynamic
fn void*! ArenaAllocator.resize(&self, void *old_pointer, usz size, usz alignment) @dynamic
{
if (!size)
{
self.release(old_pointer, alignment > 0);
return null;
}
if (!old_pointer)
{
return self.acquire(size, true, alignment, offset);
}
alignment = alignment_for_allocation(alignment);
assert(old_pointer >= self.data.ptr, "Pointer originates from a different allocator.");
usz total_len = self.data.len;
@@ -94,7 +84,7 @@ fn void*! ArenaAllocator.resize(&self, void *old_pointer, usz size, usz alignmen
ArenaAllocatorHeader* header = old_pointer - ArenaAllocatorHeader.sizeof;
usz old_size = header.size;
// Do last allocation and alignment match?
if (&self.data[self.used] == old_pointer + old_size && mem::ptr_is_aligned(old_pointer + offset, alignment))
if (&self.data[self.used] == old_pointer + old_size && mem::ptr_is_aligned(old_pointer, alignment))
{
if (old_size >= size)
{
@@ -110,7 +100,7 @@ fn void*! ArenaAllocator.resize(&self, void *old_pointer, usz size, usz alignmen
return old_pointer;
}
// Otherwise just allocate new memory.
void* mem = self.acquire(size, false, alignment, offset)!;
void* mem = self.acquire(size, NO_ZERO, alignment)!;
mem::copy(mem, old_pointer, old_size, mem::DEFAULT_MEM_ALIGNMENT, mem::DEFAULT_MEM_ALIGNMENT);
return mem;
}

View File

@@ -1,11 +1,12 @@
// Copyright (c) 2021 Christoffer Lerno. All rights reserved.
// Copyright (c) 2021-2024 Christoffer Lerno. All rights reserved.
// Use of this source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
module std::core::mem::allocator;
import std::math;
struct DynamicArenaAllocator (Allocator)
{
Allocator* backing_allocator;
Allocator backing_allocator;
DynamicArenaPage* page;
DynamicArenaPage* unused_page;
usz page_size;
@@ -15,7 +16,7 @@ struct DynamicArenaAllocator (Allocator)
* @param [&inout] allocator
* @require page_size >= 128
**/
fn void DynamicArenaAllocator.init(&self, usz page_size, Allocator* allocator)
fn void DynamicArenaAllocator.init(&self, usz page_size, Allocator allocator)
{
self.page = null;
self.unused_page = null;
@@ -29,14 +30,14 @@ fn void DynamicArenaAllocator.free(&self)
while (page)
{
DynamicArenaPage* next_page = page.prev_arena;
self.backing_allocator.free(page);
allocator::free(self.backing_allocator, page);
page = next_page;
}
page = self.unused_page;
while (page)
{
DynamicArenaPage* next_page = page.prev_arena;
self.backing_allocator.free(page);
allocator::free(self.backing_allocator, page);
page = next_page;
}
self.page = null;
@@ -58,11 +59,11 @@ struct DynamicArenaChunk @local
}
/**
* @require ptr
* @require self.page `tried to free pointer on invalid allocator`
*/
fn void DynamicArenaAllocator.release(&self, void* ptr, bool) @dynamic
{
if (!ptr) return;
DynamicArenaPage* current_page = self.page;
if (ptr == current_page.current_stack_ptr)
{
@@ -72,19 +73,12 @@ fn void DynamicArenaAllocator.release(&self, void* ptr, bool) @dynamic
}
/**
* @require size > 0 `Resize doesn't support zeroing`
* @require old_pointer != null `Resize doesn't handle null pointers`
* @require self.page `tried to realloc pointer on invalid allocator`
*/
fn void*! DynamicArenaAllocator.resize(&self, void* old_pointer, usz size, usz alignment, usz offset) @dynamic
fn void*! DynamicArenaAllocator.resize(&self, void* old_pointer, usz size, usz alignment) @dynamic
{
if (!size)
{
self.release(old_pointer, alignment > 0);
return null;
}
if (!old_pointer)
{
return self.acquire(size, true, alignment, offset);
}
DynamicArenaPage* current_page = self.page;
alignment = alignment_for_allocation(alignment);
usz* old_size_ptr = old_pointer - DEFAULT_SIZE_PREFIX;
@@ -108,7 +102,7 @@ fn void*! DynamicArenaAllocator.resize(&self, void* old_pointer, usz size, usz a
current_page.used += add_size;
return old_pointer;
}
void* new_mem = self.acquire(size, false, alignment, offset)!;
void* new_mem = self.acquire(size, NO_ZERO, alignment)!;
mem::copy(new_mem, old_pointer, old_size, mem::DEFAULT_MEM_ALIGNMENT);
return new_mem;
}
@@ -134,21 +128,21 @@ fn void DynamicArenaAllocator.reset(&self, usz mark = 0) @dynamic
* @require math::is_power_of_2(alignment)
* @require size > 0
*/
fn void*! DynamicArenaAllocator._alloc_new(&self, usz size, usz alignment, usz offset) @local
fn void*! DynamicArenaAllocator._alloc_new(&self, usz size, usz alignment) @local
{
// First, make sure that we can align it, extending the page size if needed.
usz page_size = max(self.page_size, mem::aligned_offset(size + DynamicArenaChunk.sizeof + offset, alignment) - offset);
usz page_size = max(self.page_size, mem::aligned_offset(size + DynamicArenaChunk.sizeof, alignment));
// Grab the page without alignment (we do it ourselves)
void* mem = self.backing_allocator.alloc_checked(page_size)!;
DynamicArenaPage*! page = self.backing_allocator.new(DynamicArenaPage);
void* mem = allocator::malloc_try(self.backing_allocator, page_size)!;
DynamicArenaPage*! page = allocator::new_try(self.backing_allocator, DynamicArenaPage);
if (catch err = page)
{
self.backing_allocator.free(mem);
allocator::free(self.backing_allocator, mem);
return err?;
}
page.memory = mem;
void* mem_start = mem::aligned_pointer(mem + offset + DynamicArenaChunk.sizeof, alignment) - offset;
void* mem_start = mem::aligned_pointer(mem + DynamicArenaChunk.sizeof, alignment);
assert(mem_start + size < mem + page_size);
DynamicArenaChunk* chunk = (DynamicArenaChunk*)mem_start - 1;
chunk.size = size;
@@ -161,11 +155,11 @@ fn void*! DynamicArenaAllocator._alloc_new(&self, usz size, usz alignment, usz o
}
/**
* @require size > 0 `acquire expects size > 0`
* @require !alignment || math::is_power_of_2(alignment)
*/
fn void*! DynamicArenaAllocator.acquire(&self, usz size, bool clear, usz alignment, usz offset) @dynamic
fn void*! DynamicArenaAllocator.acquire(&self, usz size, AllocInitType init_type, usz alignment) @dynamic
{
if (!size) return null;
alignment = alignment_for_allocation(alignment);
DynamicArenaPage* page = self.page;
void* ptr = {|
@@ -175,14 +169,14 @@ fn void*! DynamicArenaAllocator.acquire(&self, usz size, bool clear, usz alignme
self.unused_page = page.prev_arena;
page.prev_arena = null;
}
if (!page) return self._alloc_new(size, alignment, offset);
void* start = mem::aligned_pointer(page.memory + page.used + DynamicArenaChunk.sizeof + offset, alignment) - offset;
if (!page) return self._alloc_new(size, alignment);
void* start = mem::aligned_pointer(page.memory + page.used + DynamicArenaChunk.sizeof, alignment);
usz new_used = start - page.memory + size;
if ALLOCATE_NEW: (new_used > page.total)
{
if ((page = self.unused_page))
{
start = mem::aligned_pointer(page.memory + page.used + DynamicArenaChunk.sizeof + offset, alignment) - offset;
start = mem::aligned_pointer(page.memory + page.used + DynamicArenaChunk.sizeof, alignment);
new_used = start + size - page.memory;
if (page.total >= new_used)
{
@@ -192,7 +186,7 @@ fn void*! DynamicArenaAllocator.acquire(&self, usz size, bool clear, usz alignme
break ALLOCATE_NEW;
}
}
return self._alloc_new(size, alignment, offset);
return self._alloc_new(size, alignment);
}
page.used = new_used;
assert(start + size == page.memory + page.used);
@@ -201,6 +195,6 @@ fn void*! DynamicArenaAllocator.acquire(&self, usz size, bool clear, usz alignme
chunk.size = size;
return mem;
|}!;
if (clear) mem::clear(ptr, size, mem::DEFAULT_MEM_ALIGNMENT);
if (init_type == ZERO) mem::clear(ptr, size, mem::DEFAULT_MEM_ALIGNMENT);
return ptr;
}

View File

@@ -1,9 +1,9 @@
// Copyright (c) 2021-2023 Christoffer Lerno. All rights reserved.
// Copyright (c) 2021-2024 Christoffer Lerno. All rights reserved.
// Use of this source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
module std::core::mem::allocator;
import std::math;
struct SimpleHeapAllocator (Allocator)
{
@@ -21,29 +21,19 @@ fn void SimpleHeapAllocator.init(&self, MemoryAllocFn allocator)
self.free_list = null;
}
fn void*! SimpleHeapAllocator.acquire(&self, usz size, bool clear, usz alignment, usz offset) @dynamic
fn void*! SimpleHeapAllocator.acquire(&self, usz size, AllocInitType init_type, usz alignment) @dynamic
{
if (!size) return null;
if (clear)
if (init_type == ZERO)
{
return alignment > 0 ? @aligned_calloc(self._calloc, size, alignment, offset) : self._calloc(size);
return alignment > 0 ? @aligned_alloc(self._calloc, size, alignment) : self._calloc(size);
}
return alignment > 0 ? @aligned_alloc(self._alloc, size, alignment, offset) : self._alloc(size);
return alignment > 0 ? @aligned_alloc(self._alloc, size, alignment) : self._alloc(size);
}
fn void*! SimpleHeapAllocator.resize(&self, void* old_pointer, usz size, usz alignment, usz offset) @dynamic
fn void*! SimpleHeapAllocator.resize(&self, void* old_pointer, usz size, usz alignment) @dynamic
{
if (!size)
{
self.release(old_pointer, alignment > 0);
return null;
}
if (!old_pointer)
{
return self.acquire(size, true, alignment, offset);
}
return alignment > 0
? @aligned_realloc(self._calloc, self._free, old_pointer, size, alignment, offset)
? @aligned_realloc(self._calloc, self._free, old_pointer, size, alignment)
: self._realloc(old_pointer, size);
}

View File

@@ -1,4 +1,4 @@
// Copyright (c) 2021 Christoffer Lerno. All rights reserved.
// Copyright (c) 2021-2024 Christoffer Lerno. All rights reserved.
// Use of this source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
@@ -6,21 +6,123 @@ module std::core::mem::allocator;
import libc;
const LibcAllocator LIBC_ALLOCATOR = {};
distinct LibcAllocator (Allocator) = uptr;
fn void*! LibcAllocator.acquire(&self, usz bytes, bool clear, usz alignment, usz offset) @dynamic
module std::core::mem::allocator @if(env::POSIX);
import std::os;
import libc;
fn void*! LibcAllocator.acquire(&self, usz bytes, AllocInitType init_type, usz alignment) @dynamic
{
assert(alignment != 0 || offset == 0);
if (clear)
if (init_type == ZERO)
{
void* data = alignment ? @aligned_calloc(fn void*(usz bytes) => libc::calloc(bytes, 1), bytes, alignment, offset)!! : libc::calloc(bytes, 1);
void* data @noinit;
if (alignment > mem::DEFAULT_MEM_ALIGNMENT)
{
if (posix::posix_memalign(&data, alignment, bytes)) return AllocationFailure.OUT_OF_MEMORY?;
mem::clear(data, bytes, mem::DEFAULT_MEM_ALIGNMENT);
return data;
}
return libc::calloc(1, bytes) ?: AllocationFailure.OUT_OF_MEMORY?;
}
else
{
void* data @noinit;
if (alignment > mem::DEFAULT_MEM_ALIGNMENT)
{
if (posix::posix_memalign(&data, alignment, bytes)) return AllocationFailure.OUT_OF_MEMORY?;
}
else
{
if (!(data = libc::malloc(bytes))) return AllocationFailure.OUT_OF_MEMORY?;
}
$if env::TESTING:
for (usz i = 0; i < bytes; i++) ((char*)data)[i] = 0xAA;
$endif
return data;
}
}
fn void*! LibcAllocator.resize(&self, void* old_ptr, usz new_bytes, usz alignment) @dynamic
{
if (alignment <= mem::DEFAULT_MEM_ALIGNMENT) return libc::realloc(old_ptr, new_bytes) ?: AllocationFailure.OUT_OF_MEMORY?;
void* new_ptr;
if (posix::posix_memalign(&new_ptr, alignment, new_bytes)) return AllocationFailure.OUT_OF_MEMORY?;
$switch
$case env::DARWIN:
usz old_usable_size = darwin::malloc_size(old_ptr);
$case env::LINUX:
usz old_usable_size = linux::malloc_usable_size(old_ptr);
$default:
usz old_usable_size = new_bytes;
$endswitch
usz copy_size = new_bytes < old_usable_size ? new_bytes : old_usable_size;
mem::copy(new_ptr, old_ptr, copy_size, mem::DEFAULT_MEM_ALIGNMENT, mem::DEFAULT_MEM_ALIGNMENT);
libc::free(old_ptr);
return new_ptr;
}
fn void LibcAllocator.release(&self, void* old_ptr, bool aligned) @dynamic
{
libc::free(old_ptr);
}
module std::core::mem::allocator @if(env::WIN32);
import std::os::win32;
import libc;
fn void*! LibcAllocator.acquire(&self, usz bytes, AllocInitType init_type, usz alignment) @dynamic
{
if (init_type == ZERO)
{
if (alignment > 0)
{
return win32::_aligned_recalloc(null, bytes, alignment) ?: AllocationFailure.OUT_OF_MEMORY?;
}
return libc::calloc(1, bytes) ?: AllocationFailure.OUT_OF_MEMORY?;
}
void* data = alignment > 0 ? win32::_aligned_malloc(bytes, alignment) : libc::malloc(bytes);
if (!data) return AllocationFailure.OUT_OF_MEMORY?;
$if env::TESTING:
for (usz i = 0; i < bytes; i++) ((char*)data)[i] = 0xAA;
$endif
return data;
}
fn void*! LibcAllocator.resize(&self, void* old_ptr, usz new_bytes, usz alignment) @dynamic
{
if (alignment)
{
return win32::_aligned_realloc(old_ptr, new_bytes, alignment) ?: AllocationFailure.OUT_OF_MEMORY?;
}
return libc::realloc(old_ptr, new_bytes) ?: AllocationFailure.OUT_OF_MEMORY?;
}
fn void LibcAllocator.release(&self, void* old_ptr, bool aligned) @dynamic
{
if (aligned)
{
win32::_aligned_free(old_ptr);
return;
}
libc::free(old_ptr);
}
module std::core::mem::allocator @if(!env::WIN32 && !env::POSIX);
import libc;
fn void*! LibcAllocator.acquire(&self, usz bytes, AllocInitType init_type, usz alignment) @dynamic
{
if (init_type == ZERO)
{
void* data = alignment ? @aligned_alloc(fn void*(usz bytes) => libc::calloc(bytes, 1), bytes, alignment)!! : libc::calloc(bytes, 1);
return data ?: AllocationFailure.OUT_OF_MEMORY?;
}
else
{
void* data = alignment ? @aligned_alloc(libc::malloc, bytes, alignment, offset)!! : libc::malloc(bytes);
void* data = alignment ? @aligned_alloc(libc::malloc, bytes, alignment)!! : libc::malloc(bytes);
if (!data) return AllocationFailure.OUT_OF_MEMORY?;
$if env::TESTING:
for (usz i = 0; i < bytes; i++) ((char*)data)[i] = 0xAA;
@@ -29,21 +131,12 @@ fn void*! LibcAllocator.acquire(&self, usz bytes, bool clear, usz alignment, usz
}
}
fn void*! LibcAllocator.resize(&self, void* old_ptr, usz new_bytes, usz alignment, usz offset) @dynamic
fn void*! LibcAllocator.resize(&self, void* old_ptr, usz new_bytes, usz alignment) @dynamic
{
assert(alignment != 0 || offset == 0);
if (!new_bytes)
{
self.release(old_ptr, alignment > 0);
return null;
}
if (!old_ptr)
{
return self.acquire(new_bytes, true, alignment, offset);
}
if (alignment)
{
void* data = @aligned_realloc(fn void*(usz bytes) => libc::calloc(bytes, 1), libc::free, old_ptr, new_bytes, alignment, offset)!!;
void* data = @aligned_realloc(fn void*(usz bytes) => libc::malloc(bytes), libc::free, old_ptr, new_bytes, alignment)!!;
return data ?: AllocationFailure.OUT_OF_MEMORY?;
}
return libc::realloc(old_ptr, new_bytes) ?: AllocationFailure.OUT_OF_MEMORY?;

View File

@@ -2,7 +2,7 @@ module std::core::mem::allocator;
struct OnStackAllocator (Allocator)
{
Allocator* backing_allocator;
Allocator backing_allocator;
char[] data;
usz used;
OnStackAllocatorExtraChunk* chunk;
@@ -20,7 +20,7 @@ struct OnStackAllocatorExtraChunk @local
* @param [&inout] allocator
* Initialize a memory arena for use using the provided bytes.
**/
fn void OnStackAllocator.init(&self, char[] data, Allocator* allocator)
fn void OnStackAllocator.init(&self, char[] data, Allocator allocator)
{
self.data = data;
self.backing_allocator = allocator;
@@ -34,15 +34,15 @@ fn void OnStackAllocator.free(&self)
{
if (chunk.is_aligned)
{
self.backing_allocator.free_aligned(chunk.data);
allocator::free_aligned(self.backing_allocator, chunk.data);
}
else
{
self.backing_allocator.free(chunk.data);
allocator::free(self.backing_allocator, chunk.data);
}
void* old = chunk;
chunk = chunk.prev;
self.backing_allocator.free(old);
allocator::free(self.backing_allocator, old);
}
self.chunk = null;
self.used = 0;
@@ -54,9 +54,11 @@ struct OnStackAllocatorHeader
char[*] data;
}
/**
* @require old_pointer
**/
fn void OnStackAllocator.release(&self, void* old_pointer, bool aligned) @dynamic
{
if (!old_pointer) return;
if (allocation_in_stack_mem(self, old_pointer)) return;
on_stack_allocator_remove_chunk(self, old_pointer);
self.release(old_pointer, aligned);
@@ -76,7 +78,7 @@ fn void on_stack_allocator_remove_chunk(OnStackAllocator* a, void* ptr) @local
if (chunk.data == ptr)
{
*addr = chunk.prev;
a.backing_allocator.free(chunk);
allocator::free(a.backing_allocator, chunk);
return;
}
addr = &chunk.prev;
@@ -98,56 +100,49 @@ fn OnStackAllocatorExtraChunk* on_stack_allocator_find_chunk(OnStackAllocator* a
/**
* @require size > 0
* @require old_pointer != null
* @require alignment <= mem::MAX_MEMORY_ALIGNMENT `alignment too big`
* @require offset <= mem::MAX_MEMORY_ALIGNMENT `offset too big`
* @require offset <= size && offset >= 0
* @require mem::aligned_offset(offset, OnStackAllocatorExtraChunk.alignof) == offset
**/
fn void*! OnStackAllocator.resize(&self, void* old_pointer, usz size, usz alignment, usz offset) @dynamic
fn void*! OnStackAllocator.resize(&self, void* old_pointer, usz size, usz alignment) @dynamic
{
if (!allocation_in_stack_mem(self, old_pointer))
{
OnStackAllocatorExtraChunk* chunk = on_stack_allocator_find_chunk(self, old_pointer);
assert(chunk, "Tried to realloc pointer not belonging to the allocator");
return chunk.data = self.backing_allocator.resize(old_pointer, size, alignment, offset)!;
return chunk.data = self.backing_allocator.resize(old_pointer, size, alignment)!;
}
OnStackAllocatorHeader* header = old_pointer - OnStackAllocatorHeader.sizeof;
usz old_size = header.size;
void* mem = self.acquire(size, true, alignment, offset)!;
void* mem = self.acquire(size, NO_ZERO, alignment)!;
mem::copy(mem, old_pointer, old_size, mem::DEFAULT_MEM_ALIGNMENT, mem::DEFAULT_MEM_ALIGNMENT);
return mem;
}
/**
* @require alignment <= mem::MAX_MEMORY_ALIGNMENT `alignment too big`
* @require offset <= mem::MAX_MEMORY_ALIGNMENT `offset too big`
* @require offset <= size && offset >= 0
* @require offset == 0 || alignment > 0
* @require mem::aligned_offset(offset, OnStackAllocatorHeader.alignof) == offset
* @require size > 0
**/
fn void*! OnStackAllocator.acquire(&self, usz size, bool clear, usz alignment, usz offset) @dynamic
fn void*! OnStackAllocator.acquire(&self, usz size, AllocInitType init_type, usz alignment) @dynamic
{
if (size == 0) return null;
bool aligned = alignment > 0;
alignment = alignment_for_allocation(alignment);
usz total_len = self.data.len;
void* start_mem = self.data.ptr;
void* unaligned_pointer_to_offset = start_mem + self.used + OnStackAllocatorHeader.sizeof + offset;
void* aligned_pointer_to_offset = mem::aligned_pointer(unaligned_pointer_to_offset, alignment);
usz end = (usz)(aligned_pointer_to_offset - self.data.ptr) + size - offset;
Allocator* backing_allocator = self.backing_allocator;
void* unaligned_pointer_to_offset = start_mem + self.used + OnStackAllocatorHeader.sizeof ;
void* mem = mem::aligned_pointer(unaligned_pointer_to_offset, alignment);
usz end = (usz)(mem - self.data.ptr) + size;
Allocator backing_allocator = self.backing_allocator;
if (end > total_len)
{
OnStackAllocatorExtraChunk* chunk = backing_allocator.alloc_checked(OnStackAllocatorExtraChunk.sizeof)!;
defer catch backing_allocator.free(chunk);
OnStackAllocatorExtraChunk* chunk = allocator::alloc_try(backing_allocator, OnStackAllocatorExtraChunk)!;
defer catch allocator::free(backing_allocator, chunk);
defer try self.chunk = chunk;
*chunk = { .prev = self.chunk, .is_aligned = aligned };
return chunk.data = backing_allocator.acquire(size, clear, aligned ? alignment : 0, offset)!;
return chunk.data = backing_allocator.acquire(size, init_type, aligned ? alignment : 0)!;
}
self.used = end;
void *mem = aligned_pointer_to_offset - offset;
OnStackAllocatorHeader* header = mem - OnStackAllocatorHeader.sizeof;
header.size = size;
return mem;

View File

@@ -1,5 +1,5 @@
module std::core::mem::allocator;
import std::io;
import std::io, std::math;
struct TempAllocatorChunk @local
{
@@ -9,7 +9,7 @@ struct TempAllocatorChunk @local
struct TempAllocator (Allocator)
{
Allocator* backing_allocator;
Allocator backing_allocator;
TempAllocatorPage* last_page;
usz used;
usz capacity;
@@ -35,9 +35,9 @@ macro bool TempAllocatorPage.is_aligned(&self) => self.size & PAGE_IS_ALIGNED ==
/**
* @require size >= 16
**/
fn TempAllocator*! new_temp(usz size, Allocator* allocator)
fn TempAllocator*! new_temp_allocator(usz size, Allocator allocator)
{
TempAllocator* temp = allocator.alloc_checked(TempAllocator.sizeof + size)!;
TempAllocator* temp = allocator::alloc_with_padding(allocator, TempAllocator, size)!;
temp.last_page = null;
temp.backing_allocator = allocator;
temp.used = 0;
@@ -71,11 +71,10 @@ fn void TempAllocator.reset(&self, usz mark) @dynamic
fn void! TempAllocator._free_page(&self, TempAllocatorPage* page) @inline @local
{
void* mem = page.start;
if (page.is_aligned()) return self.backing_allocator.free_aligned(mem);
return self.backing_allocator.free(mem);
return self.backing_allocator.release(mem, page.is_aligned());
}
fn void*! TempAllocator._realloc_page(&self, TempAllocatorPage* page, usz size, usz alignment, usz offset) @inline @local
fn void*! TempAllocator._realloc_page(&self, TempAllocatorPage* page, usz size, usz alignment) @inline @local
{
// Then the actual start pointer:
void* real_pointer = page.start;
@@ -90,53 +89,36 @@ fn void*! TempAllocator._realloc_page(&self, TempAllocatorPage* page, usz size,
*pointer_to_prev = page.prev_page;
usz page_size = page.pagesize();
// Clear on size > original size.
void* data = self.acquire(size, size > page_size, alignment, offset)!;
void* data = self.acquire(size, NO_ZERO, alignment)!;
mem::copy(data, &page.data[0], page_size, mem::DEFAULT_MEM_ALIGNMENT, mem::DEFAULT_MEM_ALIGNMENT);
if (page.is_aligned())
{
self.backing_allocator.free_aligned(real_pointer);
}
else
{
self.backing_allocator.free(real_pointer);
}
self.backing_allocator.release(real_pointer, page.is_aligned());
return data;
}
fn void*! TempAllocator.resize(&self, void* pointer, usz size, usz alignment, usz offset) @dynamic
fn void*! TempAllocator.resize(&self, void* pointer, usz size, usz alignment) @dynamic
{
if (!size)
{
self.release(pointer, alignment > 0);
return null;
}
if (!pointer)
{
return self.acquire(size, true, alignment, offset);
}
TempAllocatorChunk *chunk = pointer - TempAllocatorChunk.sizeof;
if (chunk.size == (usz)-1)
{
assert(self.last_page, "Realloc of non temp pointer");
// First grab the page
TempAllocatorPage *page = pointer - TempAllocatorPage.sizeof;
return self._realloc_page(page, size, alignment, offset);
return self._realloc_page(page, size, alignment);
}
// TODO optimize last allocation
TempAllocatorChunk* data = self.acquire(size, size > chunk.size, alignment, offset)!;
TempAllocatorChunk* data = self.acquire(size, NO_ZERO, alignment)!;
mem::copy(data, pointer, chunk.size, mem::DEFAULT_MEM_ALIGNMENT, mem::DEFAULT_MEM_ALIGNMENT);
return data;
}
/**
* @require size > 0
* @require !alignment || math::is_power_of_2(alignment)
* @require alignment <= mem::MAX_MEMORY_ALIGNMENT `alignment too big`
**/
fn void*! TempAllocator.acquire(&self, usz size, bool clear, usz alignment, usz offset) @dynamic
fn void*! TempAllocator.acquire(&self, usz size, AllocInitType init_type, usz alignment) @dynamic
{
if (!size) return null;
alignment = alignment_for_allocation(alignment);
void* start_mem = &self.data;
void* starting_ptr = start_mem + self.used;
@@ -144,7 +126,7 @@ fn void*! TempAllocator.acquire(&self, usz size, bool clear, usz alignment, usz
void* mem = aligned_header_start + TempAllocatorChunk.sizeof;
if (alignment > TempAllocatorChunk.alignof)
{
mem = mem::aligned_pointer(mem + offset, alignment) - offset;
mem = mem::aligned_pointer(mem, alignment);
}
usz new_usage = (usz)(mem - start_mem) + size;
@@ -154,7 +136,7 @@ fn void*! TempAllocator.acquire(&self, usz size, bool clear, usz alignment, usz
TempAllocatorChunk* chunk_start = mem - TempAllocatorChunk.sizeof;
chunk_start.size = size;
self.used = new_usage;
if (clear) mem::clear(mem, size, mem::DEFAULT_MEM_ALIGNMENT);
if (init_type == ZERO) mem::clear(mem, size, mem::DEFAULT_MEM_ALIGNMENT);
return mem;
}
@@ -162,19 +144,20 @@ fn void*! TempAllocator.acquire(&self, usz size, bool clear, usz alignment, usz
TempAllocatorPage* page;
// We have something we need to align.
if (alignment > mem::DEFAULT_MEM_ALIGNMENT || offset)
if (alignment > mem::DEFAULT_MEM_ALIGNMENT)
{
// This is actually simpler, since it will create the offset for us.
usz total_alloc_size = TempAllocatorPage.sizeof + size;
if (clear)
usz total_alloc_size = mem::aligned_offset(TempAllocatorPage.sizeof + size, alignment);
if (init_type == ZERO)
{
page = self.backing_allocator.calloc_aligned(total_alloc_size, alignment, TempAllocatorPage.sizeof + offset)!;
mem = allocator::calloc_aligned(self.backing_allocator, total_alloc_size, alignment)!;
}
else
{
page = self.backing_allocator.alloc_aligned(total_alloc_size, alignment, TempAllocatorPage.sizeof + offset)!;
mem = allocator::malloc_aligned(self.backing_allocator, total_alloc_size, alignment)!;
}
page.start = page;
page = (TempAllocatorPage*)mem - 1;
page.start = mem;
page.size = size | PAGE_IS_ALIGNED;
}
else
@@ -182,7 +165,7 @@ fn void*! TempAllocator.acquire(&self, usz size, bool clear, usz alignment, usz
// Here we might need to pad
usz padded_header_size = mem::aligned_offset(TempAllocatorPage.sizeof, mem::DEFAULT_MEM_ALIGNMENT);
usz total_alloc_size = padded_header_size + size;
void* alloc = self.backing_allocator.acquire(total_alloc_size, clear, 0, 0)!;
void* alloc = self.backing_allocator.acquire(total_alloc_size, init_type, 0)!;
// Find the page.
page = alloc + padded_header_size - TempAllocatorPage.sizeof;

View File

@@ -1,10 +1,9 @@
// Copyright (c) 2021 Christoffer Lerno. All rights reserved.
// Copyright (c) 2021-2024 Christoffer Lerno. All rights reserved.
// Use of this source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
module std::core::mem::allocator;
import std::collections::map;
import std::collections::list;
import std::collections, std::io, std::os::backtrace;
const MAX_BACKTRACE = 16;
struct Allocation
@@ -21,7 +20,7 @@ def AllocMap = HashMap(<uptr, Allocation>);
// is not compatible with allocators that uses mark()
struct TrackingAllocator (Allocator)
{
Allocator* inner_allocator;
Allocator inner_allocator;
AllocMap map;
usz mem_total;
usz allocs_total;
@@ -32,10 +31,10 @@ struct TrackingAllocator (Allocator)
*
* @param [&inout] allocator "The allocator to track"
**/
fn void TrackingAllocator.init(&self, Allocator* allocator)
fn void TrackingAllocator.init(&self, Allocator allocator)
{
*self = { .inner_allocator = allocator };
self.map.init_new(.allocator = allocator);
self.map.new_init(.allocator = allocator);
}
/**
@@ -70,7 +69,7 @@ fn usz TrackingAllocator.total_allocated(&self) => self.mem_total;
**/
fn usz TrackingAllocator.total_allocation_count(&self) => self.allocs_total;
fn Allocation[] TrackingAllocator.allocations_tlist(&self, Allocator* allocator)
fn Allocation[] TrackingAllocator.allocations_tlist(&self, Allocator allocator)
{
return self.map.value_tlist();
}
@@ -80,48 +79,36 @@ fn Allocation[] TrackingAllocator.allocations_tlist(&self, Allocator* allocator)
**/
fn usz TrackingAllocator.allocation_count(&self) => self.map.count;
fn void*! TrackingAllocator.acquire(&self, usz size, bool clear, usz alignment, usz offset) @dynamic
fn void*! TrackingAllocator.acquire(&self, usz size, AllocInitType init_type, usz alignment) @dynamic
{
void* data = self.inner_allocator.acquire(size, clear, alignment, offset)!;
void* data = self.inner_allocator.acquire(size, init_type, alignment)!;
self.allocs_total++;
void*[MAX_BACKTRACE] bt;
backtrace::capture_current(&bt);
self.map.set((uptr)data, { data, size, bt });
self.mem_total += size;
self.allocs_total++;
if (data)
{
void*[MAX_BACKTRACE] bt;
backtrace::capture_current(&bt);
self.map.set((uptr)data, { data, size, bt });
self.mem_total += size;
self.allocs_total++;
}
return data;
}
fn void*! TrackingAllocator.resize(&self, void* old_pointer, usz size, usz alignment, usz offset) @dynamic
fn void*! TrackingAllocator.resize(&self, void* old_pointer, usz size, usz alignment) @dynamic
{
void* data = self.inner_allocator.resize(old_pointer, size, alignment, offset)!;
if (old_pointer)
{
self.map.remove((uptr)old_pointer);
}
if (data)
{
void*[MAX_BACKTRACE] bt;
backtrace::capture_current(&bt);
self.map.set((uptr)data, { data, size, bt });
self.mem_total += size;
self.allocs_total++;
}
void* data = self.inner_allocator.resize(old_pointer, size, alignment)!;
self.map.remove((uptr)old_pointer);
void*[MAX_BACKTRACE] bt;
backtrace::capture_current(&bt);
self.map.set((uptr)data, { data, size, bt });
self.mem_total += size;
self.allocs_total++;
return data;
}
fn void TrackingAllocator.release(&self, void* old_pointer, bool is_aligned) @dynamic
{
if (old_pointer)
{
if (catch self.map.remove((uptr)old_pointer))
{
assert(false, "Attempt to release untracked pointer %p, this is likely a bug.", old_pointer);
}
}
if (catch self.map.remove((uptr)old_pointer))
{
unreachable("Attempt to release untracked pointer %p, this is likely a bug.", old_pointer);
}
self.inner_allocator.release(old_pointer, is_aligned);
}
@@ -132,7 +119,7 @@ fn void TrackingAllocator.clear(&self)
fn void TrackingAllocator.print_report(&self) => self.fprint_report(io::stdout())!!;
fn void! TrackingAllocator.fprint_report(&self, OutStream* out)
fn void! TrackingAllocator.fprint_report(&self, OutStream out)
{
usz total = 0;
@@ -168,7 +155,7 @@ fn void! TrackingAllocator.fprint_report(&self, OutStream* out)
Backtrace trace = backtrace::BACKTRACE_UNKNOWN;
if (allocation.backtrace[3])
{
trace = backtrace::symbolize_backtrace(allocation.backtrace[3:1], mem::temp()).get(0) ?? backtrace::BACKTRACE_UNKNOWN;
trace = backtrace::symbolize_backtrace(allocation.backtrace[3:1], allocator::temp()).get(0) ?? backtrace::BACKTRACE_UNKNOWN;
}
if (trace.function.len) leaks = true;
io::fprintfn(out, "%13s %p %s:%d", allocation.size,
@@ -207,7 +194,7 @@ fn void! TrackingAllocator.fprint_report(&self, OutStream* out)
break;
}
}
BacktraceList list = backtrace::symbolize_backtrace(allocation.backtrace[3..(end - 1)], mem::temp())!;
BacktraceList list = backtrace::symbolize_backtrace(allocation.backtrace[3..(end - 1)], allocator::temp())!;
io::fprintfn(out, "Allocation %d (%d bytes): ", i + 1, allocation.size)!;
foreach (trace : list)
{

View File

@@ -1,4 +1,5 @@
module std::core::array;
import std::core::array::slice;
/**
* @param [in] array
@@ -15,6 +16,19 @@ macro index_of(array, element)
return SearchResult.MISSING?;
}
/**
* @require @typekind(array) == VECTOR || @typekind(array) == ARRAY
* @require @typekind(array[0]) == VECTOR || @typekind(array[0]) == ARRAY
**/
macro slice2d(array, x = 0, xlen = 0, y = 0, ylen = 0)
{
if (xlen < 1) xlen = $typeof(array[0]).len + xlen;
if (ylen < 1) ylen = $typeof(array).len + ylen;
var $ElementType = $typeof(array[0][0]);
return Slice2d(<$ElementType>) { ($ElementType*)&array, $typeof(array[0]).len, y, ylen, x, xlen };
}
/**
* @param [in] array
* @param [in] element
@@ -31,20 +45,20 @@ macro rindex_of(array, element)
}
/**
* Concatenate two arrays or subarrays, returning a subarray containing the concatenation of them.
* Concatenate two arrays or slices, returning a slice containing the concatenation of them.
*
* @param [in] arr1
* @param [in] arr2
* @param [&inout] allocator "The allocator to use, default is the heap allocator"
* @require @typekind(arr1) == SUBARRAY || @typekind(arr1) == ARRAY
* @require @typekind(arr2) == SUBARRAY || @typekind(arr2) == ARRAY
* @require @typekind(arr1) == SLICE || @typekind(arr1) == ARRAY
* @require @typekind(arr2) == SLICE || @typekind(arr2) == ARRAY
* @require @typeis(arr1[0], $typeof(arr2[0])) "Arrays must have the same type"
* @ensure result.len == arr1.len + arr2.len
**/
macro concat_new(arr1, arr2, Allocator* allocator = mem::heap())
macro concat_new(arr1, arr2, Allocator allocator = allocator::heap())
{
var $Type = $typeof(arr1[0]);
$Type[] result = allocator.new_array($Type, arr1.len + arr2.len);
$Type[] result = allocator::alloc_array(allocator, $Type, arr1.len + arr2.len);
if (arr1.len > 0)
{
mem::copy(result.ptr, &arr1[0], arr1.len * $Type.sizeof, $Type.alignof, $Type.alignof);
@@ -57,14 +71,77 @@ macro concat_new(arr1, arr2, Allocator* allocator = mem::heap())
}
/**
* Concatenate two arrays or subarrays, returning a subarray containing the concatenation of them,
* Concatenate two arrays or slices, returning a slice containing the concatenation of them,
* allocated using the temp allocator.
*
* @param [in] arr1
* @param [in] arr2
* @require @typekind(arr1) == SUBARRAY || @typekind(arr1) == ARRAY
* @require @typekind(arr2) == SUBARRAY || @typekind(arr2) == ARRAY
* @require @typekind(arr1) == SLICE || @typekind(arr1) == ARRAY
* @require @typekind(arr2) == SLICE || @typekind(arr2) == ARRAY
* @require @typeis(arr1[0], $typeof(arr2[0])) "Arrays must have the same type"
* @ensure result.len == arr1.len + arr2.len
**/
macro tconcat(arr1, arr2) => concat(arr1, arr2, mem::temp());
macro tconcat(arr1, arr2) => concat(arr1, arr2, allocator::temp());
module std::core::array::slice(<Type>);
struct Slice2d
{
Type* ptr;
usz inner_len;
usz ystart;
usz ylen;
usz xstart;
usz xlen;
}
fn usz Slice2d.len(&self) @operator(len)
{
return self.ylen;
}
fn usz Slice2d.count(&self)
{
return self.ylen * self.xlen;
}
macro void Slice2d.@each(&self; @body(usz[<2>], Type))
{
foreach (y, line : *self)
{
foreach (x, val : line)
{
@body({ x, y }, val);
}
}
}
macro void Slice2d.@each_ref(&self; @body(usz[<2>], Type*))
{
foreach (y, line : *self)
{
foreach (x, &val : line)
{
@body({ x, y }, val);
}
}
}
/**
* @require idy >= 0 && idy < self.ylen
**/
macro Type[] Slice2d.get(self, usz idy) @operator([])
{
return (self.ptr + self.inner_len * (idy + self.ystart))[self.xstart:self.xlen];
}
/**
* @require y >= 0 && y < self.ylen
* @require x >= 0 && x < self.xlen
**/
fn Slice2d Slice2d.slice(&self, isz x = 0, isz xlen = 0, isz y = 0, isz ylen = 0)
{
if (xlen < 1) xlen = self.xlen + xlen;
if (ylen < 1) ylen = self.ylen + ylen;
return { self.ptr, self.inner_len, y + self.ystart, ylen, x + self.xstart, xlen };
}

View File

@@ -88,7 +88,7 @@ bitstruct UInt128LE : uint128 @littleendian
}
/**
* @require is_array_or_sub_of_char(bytes) "argument must be an array, a pointer to an array or a subarray of char"
* @require is_array_or_slice_of_char(bytes) "argument must be an array, a pointer to an array or a slice of char"
* @require is_bitorder($Type) "type must be a bitorder integer"
**/
macro read(bytes, $Type)
@@ -104,7 +104,7 @@ macro read(bytes, $Type)
}
/**
* @require is_arrayptr_or_sub_of_char(bytes) "argument must be a pointer to an array or a subarray of char"
* @require is_arrayptr_or_slice_of_char(bytes) "argument must be a pointer to an array or a slice of char"
* @require is_bitorder($Type) "type must be a bitorder integer"
**/
macro write(x, bytes, $Type)
@@ -144,7 +144,7 @@ macro is_bitorder($Type)
$endswitch
}
macro bool is_array_or_sub_of_char(bytes)
macro bool is_array_or_slice_of_char(bytes)
{
$switch (@typekind(bytes))
$case POINTER:
@@ -154,7 +154,7 @@ macro bool is_array_or_sub_of_char(bytes)
return $Inner2.typeid == char.typeid;
$endif
$case ARRAY:
$case SUBARRAY:
$case SLICE:
var $Inner = $typefrom($typeof(bytes).inner);
return $Inner.typeid == char.typeid;
$default:
@@ -162,7 +162,7 @@ macro bool is_array_or_sub_of_char(bytes)
$endswitch
}
macro bool is_arrayptr_or_sub_of_char(bytes)
macro bool is_arrayptr_or_slice_of_char(bytes)
{
$switch (@typekind(bytes))
$case POINTER:
@@ -171,7 +171,7 @@ macro bool is_arrayptr_or_sub_of_char(bytes)
var $Inner2 = $typefrom($Inner.inner);
return $Inner2.typeid == char.typeid;
$endif
$case SUBARRAY:
$case SLICE:
var $Inner = $typefrom($typeof(bytes).inner);
return $Inner.typeid == char.typeid;
$default:

View File

@@ -1,9 +1,8 @@
// Copyright (c) 2021-2022 Christoffer Lerno and contributors. All rights reserved.
// Copyright (c) 2021-2024 Christoffer Lerno and contributors. All rights reserved.
// Use of this source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
module std::core::builtin;
import libc;
import std::hash;
import libc, std::hash, std::io, std::os::backtrace;
/**
* Use `IteratorResult` when reading the end of an iterator, or accessing a result out of bounds.
@@ -62,7 +61,7 @@ macro void @swap(&a, &b) @builtin
* @ensure @typeis(return, $Type*)
* @return! CastResult.TYPE_MISMATCH
**/
macro anycast(any* v, $Type) @builtin
macro anycast(any v, $Type) @builtin
{
if (v.type != $Type.typeid) return CastResult.TYPE_MISMATCH?;
return ($Type*)v.ptr;
@@ -75,7 +74,7 @@ fn bool print_backtrace(String message, int backtraces_to_ignore) @if(env::NATIV
void*[256] buffer;
void*[] backtraces = backtrace::capture_current(&buffer);
backtraces_to_ignore++;
BacktraceList! backtrace = backtrace::symbolize_backtrace(backtraces, mem::temp());
BacktraceList! backtrace = backtrace::symbolize_backtrace(backtraces, allocator::temp());
if (catch backtrace) return false;
if (backtrace.len() <= backtraces_to_ignore) return false;
io::eprint("\nERROR: '");
@@ -84,17 +83,18 @@ fn bool print_backtrace(String message, int backtraces_to_ignore) @if(env::NATIV
foreach (i, &trace : backtrace)
{
if (i < backtraces_to_ignore) continue;
String inline_suffix = trace.is_inline ? " [inline]" : "";
if (trace.is_unknown())
{
io::eprintn(" in ???");
io::eprintfn(" in ???%s", inline_suffix);
continue;
}
if (trace.has_file())
{
io::eprintfn(" in %s (%s:%d) [%s]", trace.function, trace.file, trace.line, trace.object_file);
io::eprintfn(" in %s (%s:%d) [%s]%s", trace.function, trace.file, trace.line, trace.object_file, inline_suffix);
continue;
}
io::eprintfn(" in %s (source unavailable) [%s]", trace.function, trace.object_file);
io::eprintfn(" in %s (source unavailable) [%s]%s", trace.function, trace.object_file, inline_suffix);
}
return true;
};
@@ -127,10 +127,10 @@ PanicFn panic = &default_panic;
fn void panicf(String fmt, String file, String function, uint line, args...)
{
@stack_mem(512; Allocator* allocator)
@stack_mem(512; Allocator allocator)
{
DString s;
s.init_new(.allocator = allocator);
s.new_init(.allocator = allocator);
s.appendf(fmt, ...args);
panic(s.str_view(), file, function, line);
};
@@ -615,7 +615,7 @@ macro void* get_returnaddress(int n)
}
module std::core::builtin @if((env::LINUX || env::DARWIN) && env::COMPILER_SAFE_MODE && env::DEBUG_SYMBOLS);
import libc;
import libc, std::io;
fn void sig_panic(String message)
{

View File

@@ -1,4 +1,4 @@
// Copyright (c) 2021-2022 Christoffer Lerno and contributors. All rights reserved.
// Copyright (c) 2021-2024 Christoffer Lerno and contributors. All rights reserved.
// Use of this source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
module std::core::builtin;

View File

@@ -1,4 +1,4 @@
// Copyright (c) 2021 Christoffer Lerno. All rights reserved.
// Copyright (c) 2021-2024 Christoffer Lerno. All rights reserved.
// Use of this source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
module std::core::cinterop;

View File

@@ -8,10 +8,10 @@ const usz MIN_CAPACITY @private = 16;
/**
* @require !self.data() "String already initialized"
**/
fn DString DString.init_new(&self, usz capacity = MIN_CAPACITY, Allocator* allocator = mem::heap())
fn DString DString.new_init(&self, usz capacity = MIN_CAPACITY, Allocator allocator = allocator::heap())
{
if (capacity < MIN_CAPACITY) capacity = MIN_CAPACITY;
StringData* data = allocator.new(StringData, .end_padding = capacity);
StringData* data = allocator::alloc_with_padding(allocator, StringData, capacity)!!;
data.allocator = allocator;
data.len = 0;
data.capacity = capacity;
@@ -21,20 +21,20 @@ fn DString DString.init_new(&self, usz capacity = MIN_CAPACITY, Allocator* alloc
/**
* @require !self.data() "String already initialized"
**/
fn DString DString.init_temp(&self, usz capacity = MIN_CAPACITY)
fn DString DString.temp_init(&self, usz capacity = MIN_CAPACITY)
{
self.init_new(capacity, mem::temp()) @inline;
self.new_init(capacity, allocator::temp()) @inline;
return *self;
}
fn DString new_with_capacity(usz capacity, Allocator* allocator = mem::heap())
fn DString new_with_capacity(usz capacity, Allocator allocator = allocator::heap())
{
return DString{}.init_new(capacity, allocator);
return DString{}.new_init(capacity, allocator);
}
fn DString temp_with_capacity(usz capacity) => new_with_capacity(capacity, mem::temp()) @inline;
fn DString temp_with_capacity(usz capacity) => new_with_capacity(capacity, allocator::temp()) @inline;
fn DString new(String c = "", Allocator* allocator = mem::heap())
fn DString new(String c = "", Allocator allocator = allocator::heap())
{
usz len = c.len;
StringData* data = (StringData*)new_with_capacity(len, allocator);
@@ -46,18 +46,18 @@ fn DString new(String c = "", Allocator* allocator = mem::heap())
return (DString)data;
}
fn DString temp_new(String s = "") => new(s, mem::temp()) @inline;
fn DString temp_new(String s = "") => new(s, allocator::temp()) @inline;
fn DString DString.new_concat(self, DString b, Allocator* allocator = mem::heap())
fn DString DString.new_concat(self, DString b, Allocator allocator = allocator::heap())
{
DString string;
string.init_new(self.len() + b.len(), allocator);
string.new_init(self.len() + b.len(), allocator);
string.append(self);
string.append(b);
return string;
}
fn DString DString.new_tconcat(self, DString b) => self.new_concat(b, mem::temp());
fn DString DString.temp_concat(self, DString b) => self.new_concat(b, allocator::temp());
fn ZString DString.zstr_view(&self)
{
@@ -146,9 +146,9 @@ fn void DString.append_char32(&self, Char32 c)
data.len += n;
}
fn DString DString.tcopy(&self) => self.copy(mem::temp());
fn DString DString.tcopy(&self) => self.copy(allocator::temp());
fn DString DString.copy(self, Allocator* allocator = null)
fn DString DString.copy(self, Allocator allocator = null)
{
if (!self)
{
@@ -156,32 +156,32 @@ fn DString DString.copy(self, Allocator* allocator = null)
return (DString)null;
}
StringData* data = self.data();
if (!allocator) allocator = mem::heap();
if (!allocator) allocator = allocator::heap();
DString new_string = new_with_capacity(data.capacity, allocator);
mem::copy((char*)new_string.data(), (char*)data, StringData.sizeof + data.len);
return new_string;
}
fn ZString DString.copy_zstr(self, Allocator* allocator = mem::heap())
fn ZString DString.copy_zstr(self, Allocator allocator = allocator::heap())
{
usz str_len = self.len();
if (!str_len)
{
return (ZString)allocator.calloc(1);
return (ZString)allocator::calloc(allocator, 1);
}
char* zstr = allocator.alloc(str_len + 1);
char* zstr = allocator::malloc(allocator, str_len + 1);
StringData* data = self.data();
mem::copy(zstr, &data.chars, str_len);
zstr[str_len] = 0;
return (ZString)zstr;
}
fn String DString.copy_str(self, Allocator* allocator = mem::heap())
fn String DString.copy_str(self, Allocator allocator = allocator::heap())
{
return (String)self.copy_zstr(allocator)[:self.len()];
}
fn String DString.tcopy_str(self) => self.copy_str(mem::temp()) @inline;
fn String DString.tcopy_str(self) => self.copy_str(allocator::temp()) @inline;
fn bool DString.equals(self, DString other_string)
{
@@ -204,7 +204,7 @@ fn void DString.free(&self)
if (!*self) return;
StringData* data = self.data();
if (!data) return;
data.allocator.free(data);
allocator::free(data.allocator, data);
*self = (DString)null;
}
@@ -240,7 +240,7 @@ fn void DString.append_chars(&self, String str)
data.len += other_len;
}
fn Char32[] DString.copy_utf32(&self, Allocator* allocator = mem::heap())
fn Char32[] DString.copy_utf32(&self, Allocator allocator = allocator::heap())
{
return self.str_view().to_new_utf32(allocator) @inline!!;
}
@@ -385,7 +385,7 @@ fn usz! DString.appendfn(&self, String format, args...) @maydiscard
return len + 1;
}
fn DString new_join(String[] s, String joiner, Allocator* allocator = mem::heap())
fn DString new_join(String[] s, String joiner, Allocator allocator = allocator::heap())
{
if (!s.len) return (DString)null;
usz total_size = joiner.len * s.len;
@@ -429,10 +429,10 @@ fn void DString.reserve(&self, usz addition)
if (new_capacity < MIN_CAPACITY) new_capacity = MIN_CAPACITY;
while (new_capacity < len) new_capacity *= 2;
data.capacity = new_capacity;
*self = (DString)data.allocator.realloc(data, StringData.sizeof + new_capacity);
*self = (DString)allocator::realloc(data.allocator, data, StringData.sizeof + new_capacity);
}
fn usz! DString.read_from_stream(&self, InStream* reader)
fn usz! DString.read_from_stream(&self, InStream reader)
{
if (&reader.available)
{
@@ -464,7 +464,7 @@ fn usz! DString.read_from_stream(&self, InStream* reader)
struct StringData @private
{
Allocator* allocator;
Allocator allocator;
usz len;
usz capacity;
char[*] chars;

View File

@@ -1,4 +1,4 @@
// Copyright (c) 2021 Christoffer Lerno. All rights reserved.
// Copyright (c) 2021-2024 Christoffer Lerno. All rights reserved.
// Use of this source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
module std::core::env;
@@ -123,6 +123,7 @@ const bool BIG_ENDIAN = $$PLATFORM_BIG_ENDIAN;
const bool I128_NATIVE_SUPPORT = $$PLATFORM_I128_SUPPORTED;
const bool F16_SUPPORT = $$PLATFORM_F16_SUPPORTED;
const bool F128_SUPPORT = $$PLATFORM_F128_SUPPORTED;
const REGISTER_SIZE = $$REGISTER_SIZE;
const bool COMPILER_SAFE_MODE = $$COMPILER_SAFE_MODE;
const bool DEBUG_SYMBOLS = $$DEBUG_SYMBOLS;
const usz LLVM_VERSION = $$LLVM_VERSION;
@@ -131,6 +132,7 @@ const bool TESTING = $$TESTING;
const MemoryEnvironment MEMORY_ENV = (MemoryEnvironment)$$MEMORY_ENVIRONMENT;
const bool TRACK_MEMORY = DEBUG_SYMBOLS && (COMPILER_SAFE_MODE || TESTING);
const bool X86_64 = ARCH_TYPE == X86_64;
const bool X86 = ARCH_TYPE == X86;
const bool AARCH64 = ARCH_TYPE == AARCH64;
const bool NATIVE_STACKTRACE = LINUX || DARWIN || WIN32;
const bool LINUX = LIBC && OS_TYPE == LINUX;

View File

@@ -2,6 +2,8 @@
// Use of this source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
module std::core::mem;
import std::core::mem::allocator @public;
import std::math;
const MAX_MEMORY_ALIGNMENT = 0x1000_0000;
const DEFAULT_MEM_ALIGNMENT = (void*.alignof) * 2;
@@ -251,11 +253,12 @@ fn bool ptr_is_aligned(void* ptr, usz alignment) @inline
macro void clear(void* dst, usz len, usz $dst_align = 0, bool $is_volatile = false, bool $inlined = false)
{
$if $inlined:
$$memset_inline(dst, (char)0, len, $is_volatile, $dst_align);
$else
$$memset(dst, (char)0, len, $is_volatile, $dst_align);
$endif
$$memset(dst, (char)0, len, $is_volatile, $dst_align);
}
macro void clear_inline(void* dst, usz $len, usz $dst_align = 0, bool $is_volatile = false)
{
$$memset_inline(dst, (char)0, $len, $is_volatile, $dst_align);
}
/**
@@ -267,17 +270,30 @@ macro void clear(void* dst, usz len, usz $dst_align = 0, bool $is_volatile = fal
* @param $dst_align "the alignment of the destination if different from the default, 0 assumes the default"
* @param $src_align "the alignment of the destination if different from the default, 0 assumes the default"
* @param $is_volatile "True if this copy should be treated as volatile, i.e. it can't be optimized away."
* @param $inlined "True if this copy should never call the OS memcpy."
*
* @require len == 0 || dst + len <= src || src + len <= dst : "Ranges may not overlap"
**/
macro void copy(void* dst, void* src, usz len, usz $dst_align = 0, usz $src_align = 0, bool $is_volatile = false, bool $inlined = false)
{
$if $inlined:
$$memcpy_inline(dst, src, len, $is_volatile, $dst_align, $src_align);
$else
$$memcpy(dst, src, len, $is_volatile, $dst_align, $src_align);
$endif
$$memcpy(dst, src, len, $is_volatile, $dst_align, $src_align);
}
/**
* Copy memory from src to dst efficiently, assuming the memory ranges do not overlap, it
* will always be inlined and never call memcopy
*
* @param [&out] dst "The destination to copy to"
* @param [&in] src "The source to copy from"
* @param $len "The number of bytes to copy"
* @param $dst_align "the alignment of the destination if different from the default, 0 assumes the default"
* @param $src_align "the alignment of the destination if different from the default, 0 assumes the default"
* @param $is_volatile "True if this copy should be treated as volatile, i.e. it can't be optimized away."
*
* @require $len == 0 || dst + $len <= src || src + $len <= dst : "Ranges may not overlap"
**/
macro void copy_inline(void* dst, void* src, usz $len, usz $dst_align = 0, usz $src_align = 0, bool $is_volatile = false)
{
$$memcpy_inline(dst, src, $len, $is_volatile, $dst_align, $src_align);
}
/**
@@ -303,23 +319,33 @@ macro void move(void* dst, void* src, usz len, usz $dst_align = 0, usz $src_alig
* @param len "The number of bytes to copy"
* @param $dst_align "the alignment of the destination if different from the default, 0 assumes the default"
* @param $is_volatile "True if this copy should be treated as volatile, i.e. it can't be optimized away."
* @param $inlined "True if this copy should never call the OS memset."
*
* @ensure !len || (dst[0] == val && dst[len - 1] == val)
**/
macro void set(void* dst, char val, usz len, usz $dst_align = 0, bool $is_volatile = false, bool $inlined = false)
macro void set(void* dst, char val, usz len, usz $dst_align = 0, bool $is_volatile = false)
{
$if $inlined:
$$memset_inline(dst, val, len, $is_volatile, $dst_align);
$else
$$memset(dst, val, len, $is_volatile, $dst_align);
$endif
$$memset(dst, val, len, $is_volatile, $dst_align);
}
/**
* @require values::@inner_kind(a) == TypeKind.SUBARRAY || values::@inner_kind(a) == TypeKind.POINTER
* @require values::@inner_kind(b) == TypeKind.SUBARRAY || values::@inner_kind(b) == TypeKind.POINTER
* @require values::@inner_kind(a) != TypeKind.SUBARRAY || len == -1
* Sets all memory in a region to that of the provided byte. Never calls OS memset.
*
* @param [&out] dst "The destination to copy to"
* @param val "The value to copy into memory"
* @param $len "The number of bytes to copy"
* @param $dst_align "the alignment of the destination if different from the default, 0 assumes the default"
* @param $is_volatile "True if this copy should be treated as volatile, i.e. it can't be optimized away."
*
* @ensure !$len || (dst[0] == val && dst[$len - 1] == val)
**/
macro void set_inline(void* dst, char val, usz $len, usz $dst_align = 0, bool $is_volatile = false)
{
$$memset_inline(dst, val, $len, $is_volatile, $dst_align);
}
/**
* @require values::@inner_kind(a) == TypeKind.SLICE || values::@inner_kind(a) == TypeKind.POINTER
* @require values::@inner_kind(b) == TypeKind.SLICE || values::@inner_kind(b) == TypeKind.POINTER
* @require values::@inner_kind(a) != TypeKind.SLICE || len == -1
* @require values::@inner_kind(a) != TypeKind.POINTER || len > -1
* @require values::@assign_to(a, b) && values::@assign_to(b, a)
**/
@@ -330,7 +356,7 @@ macro bool equals(a, b, isz len = -1, usz $align = 0)
$endif
void* x @noinit;
void* y @noinit;
$if values::@inner_kind(a) == TypeKind.SUBARRAY:
$if values::@inner_kind(a) == TypeKind.SLICE:
len = a.len;
if (len != b.len) return false;
x = a.ptr;
@@ -377,34 +403,34 @@ macro type_alloc_must_be_aligned($Type)
/**
* Run with a specific allocator inside of the macro body.
**/
macro void @scoped(Allocator* allocator; @body())
macro void @scoped(Allocator allocator; @body())
{
Allocator* old_allocator = thread_allocator;
thread_allocator = allocator;
defer thread_allocator = old_allocator;
Allocator old_allocator = allocator::thread_allocator;
allocator::thread_allocator = allocator;
defer allocator::thread_allocator = old_allocator;
@body();
}
macro void @report_heap_allocs_in_scope(;@body())
{
TrackingAllocator tracker;
tracker.init(thread_allocator);
Allocator* old_allocator = thread_allocator;
thread_allocator = &tracker;
tracker.init(allocator::thread_allocator);
Allocator old_allocator = allocator::thread_allocator;
allocator::thread_allocator = &tracker;
defer
{
thread_allocator = old_allocator;
allocator::thread_allocator = old_allocator;
tracker.print_report();
tracker.free();
}
@body();
}
macro void @stack_mem(usz $size; @body(Allocator* mem)) @builtin
macro void @stack_mem(usz $size; @body(Allocator mem)) @builtin
{
char[$size] buffer;
OnStackAllocator allocator;
allocator.init(&buffer, mem::heap());
allocator.init(&buffer, allocator::heap());
defer allocator.free();
@body(&allocator);
}
@@ -413,7 +439,7 @@ macro void @stack_pool(usz $size; @body) @builtin
{
char[$size] buffer;
OnStackAllocator allocator;
allocator.init(&buffer, mem::heap());
allocator.init(&buffer, allocator::heap());
defer allocator.free();
mem::@scoped(&allocator)
{
@@ -421,76 +447,62 @@ macro void @stack_pool(usz $size; @body) @builtin
};
}
struct TempState
{
TempAllocator* old;
TempAllocator* current;
usz mark;
}
/**
* Push the current temp allocator. A push must always be balanced with a pop using the current state.
**/
fn TempState temp_push(TempAllocator* other = null)
{
TempAllocator* current = allocator::temp();
TempAllocator* old = current;
if (other == current)
{
current = allocator::temp_allocator_next();
}
return { old, current, current.used };
}
/**
* Pop the current temp allocator. A pop must always be balanced with a push.
**/
fn void temp_pop(TempState old_state)
{
assert(allocator::thread_temp_allocator == old_state.current, "Tried to pop temp allocators out of order.");
assert(old_state.current.used >= old_state.mark, "Tried to pop temp allocators out of order.");
old_state.current.reset(old_state.mark);
allocator::thread_temp_allocator = old_state.old;
}
macro void @pool(TempAllocator* #other_temp = null; @body) @builtin
{
TempAllocator* current = temp();
TempAllocator* current = allocator::temp();
var $has_arg = !$is_const(#other_temp);
$if $has_arg:
TempAllocator* original = current;
if (current == (void*)#other_temp) current = temp_allocator_next();
if (current == (void*)#other_temp) current = allocator::temp_allocator_next();
$endif
usz mark = current.used;
defer
{
current.reset(mark);
$if $has_arg:
thread_temp_allocator = original;
allocator::thread_temp_allocator = original;
$endif;
}
@body();
}
tlocal Allocator* thread_allocator @private = &allocator::LIBC_ALLOCATOR;
tlocal TempAllocator* thread_temp_allocator @private = null;
tlocal TempAllocator*[2] temp_allocator_pair @private;
Allocator* temp_base_allocator @private = &allocator::LIBC_ALLOCATOR;
macro TempAllocator* create_default_sized_temp_allocator(Allocator* allocator) @local
{
$switch (env::MEMORY_ENV)
$case NORMAL:
return allocator::new_temp(1024 * 256, allocator)!!;
$case SMALL:
return allocator::new_temp(1024 * 16, allocator)!!;
$case TINY:
return allocator::new_temp(1024 * 2, allocator)!!;
$case NONE:
unreachable("Temp allocator must explicitly created when memory-env is set to 'none'.");
$endswitch
}
fn TempAllocator *temp_allocator_next() @private
{
if (!thread_temp_allocator)
{
init_default_temp_allocators();
return thread_temp_allocator;
}
usz index = thread_temp_allocator == temp_allocator_pair[0] ? 1 : 0;
return thread_temp_allocator = temp_allocator_pair[index];
}
import libc;
fn void init_default_temp_allocators() @private
{
temp_allocator_pair[0] = create_default_sized_temp_allocator(temp_base_allocator);
temp_allocator_pair[1] = create_default_sized_temp_allocator(temp_base_allocator);
thread_temp_allocator = temp_allocator_pair[0];
}
macro TempAllocator* temp()
{
if (!thread_temp_allocator)
{
init_default_temp_allocators();
}
return thread_temp_allocator;
}
macro Allocator* current_allocator() => thread_allocator;
macro Allocator* heap() => thread_allocator;
module std::core::mem @if(WASM_NOLIBC);
@@ -505,7 +517,7 @@ fn void initialize_wasm_mem() @init(1) @private
if (start > mem::DEFAULT_MEM_ALIGNMENT) allocator::wasm_memory.use = start;
wasm_allocator.init(fn (x) => allocator::wasm_memory.allocate_block(x));
temp_base_allocator = &wasm_allocator;
thread_allocator = &wasm_allocator;
allocator::thread_allocator = &wasm_allocator;
}
module std::core::mem;
@@ -520,88 +532,182 @@ macro TrackingEnv* get_tracking_env()
$endif
}
macro @clone(value) @builtin
macro @clone(value) @builtin @nodiscard
{
return mem::heap().clone(value);
return allocator::clone(allocator::heap(), value);
}
macro @tclone(value) @builtin
macro @tclone(value) @builtin @nodiscard
{
return mem::temp().clone(value);
return temp_new($typeof(value), value);
}
fn void* malloc(usz size) @builtin @inline
fn void* malloc(usz size) @builtin @inline @nodiscard
{
return mem::heap().alloc(size);
return allocator::malloc(allocator::heap(), size);
}
fn void* tmalloc(usz size, usz alignment = 0, usz offset = 0) @builtin @inline
fn void* tmalloc(usz size, usz alignment = 0) @builtin @inline @nodiscard
{
return temp().acquire(size, false, alignment, offset)!!;
if (!size) return null;
return allocator::temp().acquire(size, NO_ZERO, alignment)!!;
}
macro new($Type)
/**
* @require $vacount < 2 : "Too many arguments."
* @require $or($vacount == 0, $assignable($vaexpr(0), $Type)) : "The second argument must be an initializer for the type"
* @require $Type.alignof <= DEFAULT_MEM_ALIGNMENT : "Types with alignment exceeding the default must use 'alloc_aligned' instead"
**/
macro new($Type, ...) @nodiscard
{
return heap().new($Type);
$if $vacount == 0:
return ($Type*)calloc($Type.sizeof);
$else
$Type* val = malloc($Type.sizeof);
*val = $vaexpr(0);
return val;
$endif
}
macro new_clear($Type)
/**
* Allocate using an aligned allocation. This is necessary for types with a default memory alignment
* exceeding DEFAULT_MEM_ALIGNMENT. IMPORTANT! It must be freed using free_aligned.
* @require $vacount < 2 : "Too many arguments."
* @require $or($vacount == 0, $assignable($vaexpr(0), $Type)) : "The second argument must be an initializer for the type"
**/
macro new_aligned($Type, ...) @nodiscard
{
return heap().new_clear($Type);
$if $vacount == 0:
return ($Type*)calloc_aligned($Type.sizeof, $Type.alignof);
$else
$Type* val = malloc_aligned($Type.sizeof, $Type.alignof);
*val = $vaexpr(0);
return val;
$endif
}
macro new_temp($Type)
/**
* @require $Type.alignof <= DEFAULT_MEM_ALIGNMENT : "Types with alignment exceeding the default must use 'alloc_aligned' instead"
**/
macro alloc($Type) @nodiscard
{
return ($Type*)malloc($Type.sizeof);
}
/**
* Allocate using an aligned allocation. This is necessary for types with a default memory alignment
* exceeding DEFAULT_MEM_ALIGNMENT. IMPORTANT! It must be freed using free_aligned.
**/
macro alloc_aligned($Type) @nodiscard
{
return ($Type*)malloc_aligned($Type.sizeof, $Type.alignof);
}
/**
* @require $vacount < 2 : "Too many arguments."
* @require $or($vacount == 0, $assignable($vaexpr(0), $Type)) : "The second argument must be an initializer for the type"
**/
macro temp_new($Type, ...) @nodiscard
{
$if $vacount == 0:
return ($Type*)tcalloc($Type.sizeof) @inline;
$else
$Type* val = tmalloc($Type.sizeof) @inline;
*val = $vaexpr(0);
return val;
$endif
}
macro temp_alloc($Type) @nodiscard
{
return tmalloc($Type.sizeof);
}
macro new_temp_clear($Type)
/**
* @require $Type.alignof <= DEFAULT_MEM_ALIGNMENT : "Types with alignment exceeding the default must use 'new_array_aligned' instead"
**/
macro new_array($Type, usz elements) @nodiscard
{
return tcalloc($Type.sizeof);
return allocator::new_array(allocator::heap(), $Type, elements);
}
macro new_array($Type, usz elements)
/**
* Allocate using an aligned allocation. This is necessary for types with a default memory alignment
* exceeding DEFAULT_MEM_ALIGNMENT. IMPORTANT! It must be freed using free_aligned.
**/
macro new_array_aligned($Type, usz elements) @nodiscard
{
return heap().new_array($Type, elements);
return allocator::new_array_aligned(allocator::heap(), $Type, elements);
}
macro temp_array($Type, usz elements)
/**
* @require $Type.alignof <= DEFAULT_MEM_ALIGNMENT : "Types with alignment exceeding the default must use 'alloc_array_aligned' instead"
**/
macro alloc_array($Type, usz elements) @nodiscard
{
return allocator::alloc_array(allocator::heap(), $Type, elements);
}
/**
* Allocate using an aligned allocation. This is necessary for types with a default memory alignment
* exceeding DEFAULT_MEM_ALIGNMENT. IMPORTANT! It must be freed using free_aligned.
**/
macro alloc_array_aligned($Type, usz elements) @nodiscard
{
return allocator::alloc_array(allocator::heap(), $Type, elements);
}
macro temp_alloc_array($Type, usz elements) @nodiscard
{
return (($Type*)tmalloc($Type.sizeof * elements, $Type.alignof))[:elements];
}
macro new_zero_array($Type, usz elements)
{
return heap().new_zero_array($Type, elements);
}
macro temp_zero_array($Type, usz elements)
macro temp_new_array($Type, usz elements) @nodiscard
{
return (($Type*)tcalloc($Type.sizeof * elements, $Type.alignof))[:elements];
}
fn void* calloc(usz size) @builtin @inline
fn void* calloc(usz size) @builtin @inline @nodiscard
{
return heap().calloc(size);
return allocator::calloc(allocator::heap(), size);
}
fn void* tcalloc(usz size, usz alignment = 0, usz offset = 0) @builtin @inline
fn void* calloc_aligned(usz size, usz alignment) @builtin @inline @nodiscard
{
return temp().acquire(size, false, alignment, offset)!!;
return allocator::calloc_aligned(allocator::heap(), size, alignment)!!;
}
fn void* realloc(void *ptr, usz new_size) @builtin @inline
fn void* tcalloc(usz size, usz alignment = 0) @builtin @inline @nodiscard
{
return heap().realloc(ptr, new_size);
if (!size) return null;
return allocator::temp().acquire(size, ZERO, alignment)!!;
}
fn void* realloc(void *ptr, usz new_size) @builtin @inline @nodiscard
{
return allocator::realloc(allocator::heap(), ptr, new_size);
}
fn void* realloc_aligned(void *ptr, usz new_size, usz alignment) @builtin @inline @nodiscard
{
return allocator::realloc_aligned(allocator::heap(), ptr, new_size, alignment)!!;
}
fn void free(void* ptr) @builtin @inline
{
heap().free(ptr);
return allocator::free(allocator::heap(), ptr);
}
fn void* trealloc(void* ptr, usz size, usz alignment = mem::DEFAULT_MEM_ALIGNMENT) @builtin @inline
fn void free_aligned(void* ptr) @builtin @inline
{
return temp().resize(ptr, size, alignment, 0)!!;
return allocator::free_aligned(allocator::heap(), ptr);
}
fn void* trealloc(void* ptr, usz size, usz alignment = mem::DEFAULT_MEM_ALIGNMENT) @builtin @inline @nodiscard
{
if (!size) return null;
if (!ptr) return tmalloc(size, alignment);
return allocator::temp().resize(ptr, size, alignment)!!;
}

View File

@@ -10,88 +10,35 @@ struct TrackingEnv
uint line;
}
enum AllocInitType
{
NO_ZERO,
ZERO
}
interface Allocator
{
fn void reset(usz mark) @optional;
fn usz mark() @optional;
fn void*! acquire(usz size, bool clear, usz alignment, usz offset);
fn void*! resize(void* ptr, usz new_size, usz alignment, usz offset);
/**
* @require !alignment || math::is_power_of_2(alignment)
* @require alignment <= mem::MAX_MEMORY_ALIGNMENT `alignment too big`
* @require size > 0
**/
fn void*! acquire(usz size, AllocInitType init_type, usz alignment = 0);
/**
* @require !alignment || math::is_power_of_2(alignment)
* @require alignment <= mem::MAX_MEMORY_ALIGNMENT `alignment too big`
* @require ptr != null
* @require new_size > 0
**/
fn void*! resize(void* ptr, usz new_size, usz alignment = 0);
/**
* @require ptr != null
**/
fn void release(void* ptr, bool aligned);
}
struct AlignedBlock
{
usz len;
void* start;
}
/**
* @require bytes > 0
* @require alignment > 0
**/
macro void*! @aligned_alloc(#alloc_fn, usz bytes, usz alignment, usz offset)
{
usz header = mem::aligned_offset(AlignedBlock.sizeof + offset, alignment) - offset;
$if @typekind(#alloc_fn(bytes)) == OPTIONAL:
void* data = #alloc_fn(header + bytes)!;
$else
void* data = #alloc_fn(header + bytes);
$endif
void* mem = mem::aligned_pointer(data + header + offset, alignment) - offset;
assert(mem > data);
AlignedBlock* desc = (AlignedBlock*)mem - 1;
*desc = { bytes, data };
return mem;
}
/**
* @require bytes > 0
* @require alignment > 0
**/
macro void*! @aligned_calloc(#calloc_fn, usz bytes, usz alignment, usz offset)
{
usz header = mem::aligned_offset(AlignedBlock.sizeof + offset, alignment) - offset;
$if @typekind(#calloc_fn(bytes)) == OPTIONAL:
void* data = #calloc_fn(header + bytes)!;
$else
void* data = #calloc_fn(header + bytes);
$endif
void* mem = mem::aligned_pointer(data + header + offset, alignment) - offset;
AlignedBlock* desc = (AlignedBlock*)mem - 1;
assert(mem > data);
*desc = { bytes, data };
return mem;
}
/**
* @require bytes > 0
* @require alignment > 0
**/
macro void*! @aligned_realloc(#calloc_fn, #free_fn, void* old_pointer, usz bytes, usz alignment, usz offset)
{
AlignedBlock* desc = (AlignedBlock*)old_pointer - 1;
void* data_start = desc.start;
void* new_data = @aligned_calloc(#calloc_fn, bytes, alignment, offset)!;
mem::copy(new_data, old_pointer, desc.len < bytes ? desc.len : bytes, mem::DEFAULT_MEM_ALIGNMENT, mem::DEFAULT_MEM_ALIGNMENT);
$if @typekind(#free_fn(data_start)) == OPTIONAL:
#free_fn(data_start)!;
$else
#free_fn(data_start);
$endif
return new_data;
}
macro void! @aligned_free(#free_fn, void* old_pointer)
{
AlignedBlock* desc = (AlignedBlock*)old_pointer - 1;
$if @typekind(#free_fn(desc.start)) == OPTIONAL:
#free_fn(desc.start)!;
$else
#free_fn(desc.start);
$endif
}
def MemoryAllocFn = fn char[]!(usz);
fault AllocationFailure
@@ -105,120 +52,296 @@ fn usz alignment_for_allocation(usz alignment) @inline @private
return alignment < mem::DEFAULT_MEM_ALIGNMENT ? alignment = mem::DEFAULT_MEM_ALIGNMENT : alignment;
}
// Allocator "functions"
macro void*! Allocator.alloc_checked(&self, usz size)
macro void* malloc(Allocator allocator, usz size) @nodiscard
{
return malloc_try(allocator, size)!!;
}
macro void*! malloc_try(Allocator allocator, usz size) @nodiscard
{
if (!size) return null;
$if env::TESTING:
char* data = self.acquire(size, false, 0, 0)!;
char* data = allocator.acquire(size, NO_ZERO)!;
mem::set(data, 0xAA, size, mem::DEFAULT_MEM_ALIGNMENT);
return data;
$else
return self.acquire(size, false, 0, 0);
return allocator.acquire(size, NO_ZERO);
$endif
}
macro void*! Allocator.calloc_checked(&self, usz size)
macro void* calloc(Allocator allocator, usz size) @nodiscard
{
return self.acquire(size, true, 0, 0);
}
macro void*! Allocator.realloc_checked(&self, void* ptr, usz new_size)
{
return self.resize(ptr, new_size, 0, 0);
return calloc_try(allocator, size)!!;
}
macro Allocator.new_array(&self, $Type, usz size, usz end_padding = 0)
macro void*! calloc_try(Allocator allocator, usz size) @nodiscard
{
return (($Type*)self.alloc_checked($Type.sizeof * size + end_padding))[:size]!!;
if (!size) return null;
return allocator.acquire(size, ZERO);
}
macro Allocator.new_array_checked(&self, $Type, usz size, usz end_padding = 0)
macro void* realloc(Allocator allocator, void* ptr, usz new_size) @nodiscard
{
return (($Type*)self.alloc_checked($Type.sizeof * size + end_padding))[:size];
return realloc_try(allocator, ptr, new_size)!!;
}
macro Allocator.new_zero_array(&self, $Type, usz size, usz end_padding = 0)
macro void*! realloc_try(Allocator allocator, void* ptr, usz new_size) @nodiscard
{
return (($Type*)self.calloc_checked($Type.sizeof * size + end_padding))[:size]!!;
if (!new_size)
{
free(allocator, ptr);
return null;
}
if (!ptr) return allocator.acquire(new_size, NO_ZERO);
return allocator.resize(ptr, new_size);
}
macro Allocator.new_zero_array_checked(&self, $Type, usz size, usz end_padding = 0)
{
return (($Type*)self.calloc_checked($Type.sizeof * size + end_padding))[:size];
}
macro Allocator.new(&self, $Type, usz end_padding = 0) @nodiscard
{
return ($Type*)self.alloc_checked($Type.sizeof + end_padding)!!;
}
macro Allocator.new_checked(&self, $Type, usz end_padding = 0) @nodiscard
{
return ($Type*)self.alloc_checked($Type.sizeof + end_padding);
}
macro Allocator.new_clear(&self, $Type, usz end_padding = 0) @nodiscard
{
return ($Type*)self.calloc_checked($Type.sizeof + end_padding)!!;
}
macro Allocator.new_clear_checked(&self, $Type, usz end_padding = 0) @nodiscard
{
return ($Type*)self.calloc_checked($Type.sizeof + end_padding);
}
macro Allocator.clone(&self, value)
{
var x = self.alloc($typeof(value));
*x = value;
return x;
}
macro void* Allocator.alloc(&self, usz size) @nodiscard
{
return self.alloc_checked(size)!!;
}
macro void* Allocator.calloc(&self, usz size) @nodiscard
{
return self.acquire(size, true, 0, 0)!!;
}
macro void* Allocator.realloc(&self, void* ptr, usz new_size) @nodiscard
{
return self.resize(ptr, new_size, 0, 0)!!;
}
macro void*! Allocator.alloc_aligned(&self, usz size, usz alignment, usz offset = 0)
macro void free(Allocator allocator, void* ptr)
{
if (!ptr) return;
$if env::TESTING:
char* data = self.acquire(size, false, alignment, offset)!;
((char*)ptr)[0] = 0xBA;
$endif
allocator.release(ptr, false);
}
macro void*! malloc_aligned(Allocator allocator, usz size, usz alignment) @nodiscard
{
if (!size) return null;
$if env::TESTING:
char* data = allocator.acquire(size, NO_ZERO, alignment)!;
mem::set(data, 0xAA, size, mem::DEFAULT_MEM_ALIGNMENT);
return data;
$else
return self.acquire(size, false, alignment, offset);
return allocator.acquire(size, NO_ZERO, alignment);
$endif
}
macro void*! Allocator.calloc_aligned(&self, usz size, usz alignment, usz offset = 0)
macro void*! calloc_aligned(Allocator allocator, usz size, usz alignment) @nodiscard
{
return self.acquire(size, true, alignment, offset);
}
macro void*! Allocator.realloc_aligned(&self, void* ptr, usz new_size, usz alignment = 0, usz offset = 0)
{
return self.resize(ptr, new_size, alignment, offset);
if (!size) return null;
return allocator.acquire(size, ZERO, alignment);
}
macro void Allocator.free(&self, void* ptr)
macro void*! realloc_aligned(Allocator allocator, void* ptr, usz new_size, usz alignment) @nodiscard
{
if (!new_size)
{
free_aligned(allocator, ptr);
return null;
}
if (!ptr)
{
return malloc_aligned(allocator, new_size, alignment);
}
return allocator.resize(ptr, new_size, alignment);
}
macro void free_aligned(Allocator allocator, void* ptr)
{
if (!ptr) return;
$if env::TESTING:
if (ptr) ((char*)ptr)[0] = 0xBA;
((char*)ptr)[0] = 0xBA;
$endif
self.release(ptr, false);
allocator.release(ptr, .aligned = true);
}
macro void Allocator.free_aligned(&self, void* ptr)
/**
* @require $vacount < 2 : "Too many arguments."
* @require $or($vacount == 0, $assignable($vaexpr(0), $Type)) : "The second argument must be an initializer for the type"
**/
macro new(Allocator allocator, $Type, ...) @nodiscard
{
$if env::TESTING:
if (ptr) ((char*)ptr)[0] = 0xBA;
$if $vacount == 0:
return ($Type*)calloc(allocator, $Type.sizeof);
$else
$Type* val = malloc(allocator, $Type.sizeof);
*val = $vaexpr(0);
return val;
$endif
self.release(ptr, true);
}
/**
* @require $vacount < 2 : "Too many arguments."
* @require $or($vacount == 0, $assignable($vaexpr(0), $Type)) : "The second argument must be an initializer for the type"
**/
macro new_try(Allocator allocator, $Type, ...) @nodiscard
{
$if $vacount == 0:
return ($Type*)calloc_try(allocator, $Type.sizeof);
$else
$Type* val = malloc_try(allocator, $Type.sizeof)!;
*val = $vaexpr(0);
return val;
$endif
}
macro new_with_padding(Allocator allocator, $Type, usz padding) @nodiscard
{
return ($Type*)calloc_try(allocator, $Type.sizeof + padding);
}
macro alloc(Allocator allocator, $Type) @nodiscard
{
return ($Type*)malloc(allocator, $Type.sizeof);
}
macro alloc_try(Allocator allocator, $Type) @nodiscard
{
return ($Type*)malloc_try(allocator, $Type.sizeof);
}
macro alloc_with_padding(Allocator allocator, $Type, usz padding) @nodiscard
{
return ($Type*)malloc_try(allocator, $Type.sizeof + padding);
}
macro new_array(Allocator allocator, $Type, usz elements) @nodiscard
{
return new_array_try(allocator, $Type, elements)!!;
}
macro new_array_try(Allocator allocator, $Type, usz elements) @nodiscard
{
return (($Type*)calloc_try(allocator, $Type.sizeof * elements))[:elements];
}
macro new_array_aligned(Allocator allocator, $Type, usz elements) @nodiscard
{
return ((Type*)calloc_aligned(allocator, $Type.sizeof * elements, $Type.alignof))[:elements]!!;
}
macro alloc_array(Allocator allocator, $Type, usz elements) @nodiscard
{
return alloc_array_try(allocator, $Type, elements)!!;
}
macro alloc_array_aligned(Allocator allocator, $Type, usz elements) @nodiscard
{
return ((Type*)malloc_aligned(allocator, $Type.sizeof * elements, $Type.alignof))[:elements]!!;
}
macro alloc_array_try(Allocator allocator, $Type, usz elements) @nodiscard
{
return (($Type*)malloc_try(allocator, $Type.sizeof * elements))[:elements];
}
macro clone(Allocator allocator, value) @nodiscard
{
return new(allocator, $typeof(value), value);
}
fn any clone_any(Allocator allocator, any value) @nodiscard
{
usz size = value.type.sizeof;
void* data = malloc(allocator, size);
mem::copy(data, value.ptr, size);
return any_make(data, value.type);
}
/**
* @require bytes > 0
* @require alignment > 0
* @require bytes <= isz.max
**/
macro void*! @aligned_alloc(#alloc_fn, usz bytes, usz alignment)
{
if (alignment < void*.alignof) alignment = void*.alignof;
usz header = AlignedBlock.sizeof + alignment;
usz alignsize = bytes + header;
$if @typekind(#alloc_fn(bytes)) == OPTIONAL:
void* data = #alloc_fn(alignsize)!;
$else
void* data = #alloc_fn(alignsize);
$endif
void* mem = mem::aligned_pointer(data + AlignedBlock.sizeof, alignment);
AlignedBlock* desc = (AlignedBlock*)mem - 1;
assert(mem > data);
*desc = { bytes, data };
return mem;
}
struct AlignedBlock
{
usz len;
void* start;
}
macro void! @aligned_free(#free_fn, void* old_pointer)
{
AlignedBlock* desc = (AlignedBlock*)old_pointer - 1;
$if @typekind(#free_fn(desc.start)) == OPTIONAL:
#free_fn(desc.start)!;
$else
#free_fn(desc.start);
$endif
}
/**
* @require bytes > 0
* @require alignment > 0
**/
macro void*! @aligned_realloc(#calloc_fn, #free_fn, void* old_pointer, usz bytes, usz alignment)
{
AlignedBlock* desc = (AlignedBlock*)old_pointer - 1;
void* data_start = desc.start;
void* new_data = @aligned_alloc(#calloc_fn, bytes, alignment)!;
mem::copy(new_data, old_pointer, desc.len < bytes ? desc.len : bytes, 1, 1);
$if @typekind(#free_fn(data_start)) == OPTIONAL:
#free_fn(data_start)!;
$else
#free_fn(data_start);
$endif
return new_data;
}
// All allocators
tlocal Allocator thread_allocator @private = &allocator::LIBC_ALLOCATOR;
tlocal TempAllocator* thread_temp_allocator @private = null;
tlocal TempAllocator*[2] temp_allocator_pair @private;
Allocator temp_base_allocator @private = &allocator::LIBC_ALLOCATOR;
macro TempAllocator* create_default_sized_temp_allocator(Allocator allocator) @local
{
$switch (env::MEMORY_ENV)
$case NORMAL:
return new_temp_allocator(1024 * 256, allocator)!!;
$case SMALL:
return new_temp_allocator(1024 * 16, allocator)!!;
$case TINY:
return new_temp_allocator(1024 * 2, allocator)!!;
$case NONE:
unreachable("Temp allocator must explicitly created when memory-env is set to 'none'.");
$endswitch
}
macro Allocator heap() => thread_allocator;
macro TempAllocator* temp()
{
if (!thread_temp_allocator)
{
init_default_temp_allocators();
}
return thread_temp_allocator;
}
fn void init_default_temp_allocators() @private
{
temp_allocator_pair[0] = create_default_sized_temp_allocator(temp_base_allocator);
temp_allocator_pair[1] = create_default_sized_temp_allocator(temp_base_allocator);
thread_temp_allocator = temp_allocator_pair[0];
}
fn TempAllocator* temp_allocator_next() @private
{
if (!thread_temp_allocator)
{
init_default_temp_allocators();
return thread_temp_allocator;
}
usz index = thread_temp_allocator == temp_allocator_pair[0] ? 1 : 0;
return thread_temp_allocator = temp_allocator_pair[index];
}

View File

@@ -0,0 +1,258 @@
module std::core::cpudetect @if(env::X86 || env::X86_64);
struct CpuId
{
uint eax, ebx, ecx, edx;
}
fn CpuId x86_cpuid(uint eax, uint ecx = 0)
{
int edx;
int ebx;
asm
{
movl $eax, eax;
movl $ecx, ecx;
cpuid;
movl eax, $eax;
movl ebx, $ebx;
movl ecx, $ecx;
movl edx, $edx;
}
return { eax, ebx, ecx, edx };
}
enum X86Feature
{
ADX,
AES,
AMX_BF16,
AMX_COMPLEX,
AMX_FP16,
AMX_INT8,
AMX_TILE,
AVX,
AVX10_1_256,
AVX10_1_512,
AVX2,
AVX5124FMAPS,
AVX5124VNNIW,
AVX512BF16,
AVX512BITALG,
AVX512BW,
AVX512CD,
AVX512DQ,
AVX512ER,
AVX512F,
AVX512FP16,
AVX512IFMA,
AVX512PF,
AVX512VBMI,
AVX512VBMI2,
AVX512VL,
AVX512VNNI,
AVX512VP2INTERSECT,
AVX512VPOPCNTDQ,
AVXIFMA,
AVXNECONVERT,
AVXVNNI,
AVXVNNIINT16,
AVXVNNIINT8,
BMI,
BMI2,
CLDEMOTE,
CLFLUSHOPT,
CLWB,
CLZERO,
CMOV,
CMPCCXADD,
CMPXCHG16B,
CX8,
ENQCMD,
F16C,
FMA,
FMA4,
FSGSBASE,
FXSR,
GFNI,
HRESET,
INVPCID,
KL,
LWP,
LZCNT,
MMX,
MOVBE,
MOVDIR64B,
MOVDIRI,
MWAITX,
PCLMUL,
PCONFIG,
PKU,
POPCNT,
PREFETCHI,
PREFETCHWT1,
PRFCHW,
PTWRITE,
RAOINT,
RDPID,
RDPRU,
RDRND,
RDSEED,
RTM,
SAHF,
SERIALIZE,
SGX,
SHA,
SHA512,
SHSTK,
SM3,
SM4,
SSE,
SSE2,
SSE3,
SSE4_1,
SSE4_2,
SSE4_A,
SSSE3,
TBM,
TSXLDTRK,
UINTR,
USERMSR,
VAES,
VPCLMULQDQ,
WAITPKG,
WBNOINVD,
WIDEKL,
X87,
XOP,
XSAVE,
XSAVEC,
XSAVEOPT,
XSAVES,
}
uint128 x86_features;
fn void add_feature_if_bit(X86Feature feature, uint register, int bit)
{
if (register & 1U << bit) x86_features |= 1u128 << feature.ordinal;
}
fn void x86_initialize_cpu_features()
{
uint max_level = x86_cpuid(0).eax;
CpuId feat = x86_cpuid(1);
CpuId leaf7 = max_level >= 8 ? x86_cpuid(7) : CpuId {};
CpuId leaf7s1 = leaf7.eax >= 1 ? x86_cpuid(7, 1) : CpuId {};
CpuId ext1 = x86_cpuid(0x80000000).eax >= 0x80000001 ? x86_cpuid(0x80000001) : CpuId {};
CpuId ext8 = x86_cpuid(0x80000000).eax >= 0x80000008 ? x86_cpuid(0x80000008) : CpuId {};
CpuId leaf_d = max_level >= 0xd ? x86_cpuid(0xd, 0x1) : CpuId {};
CpuId leaf_14 = max_level >= 0x14 ? x86_cpuid(0x14) : CpuId {};
CpuId leaf_19 = max_level >= 0x19 ? x86_cpuid(0x19) : CpuId {};
CpuId leaf_24 = max_level >= 0x24 ? x86_cpuid(0x24) : CpuId {};
add_feature_if_bit(ADX, leaf7.ebx, 19);
add_feature_if_bit(AES, feat.ecx, 25);
add_feature_if_bit(AMX_BF16, leaf7.edx, 22);
add_feature_if_bit(AMX_COMPLEX, leaf7s1.edx, 8);
add_feature_if_bit(AMX_FP16, leaf7s1.eax, 21);
add_feature_if_bit(AMX_INT8, leaf7.edx, 25);
add_feature_if_bit(AMX_TILE, leaf7.edx, 24);
add_feature_if_bit(AVX, feat.ecx, 28);
add_feature_if_bit(AVX10_1_256, leaf7s1.edx, 19);
add_feature_if_bit(AVX10_1_512, leaf_24.ebx, 18);
add_feature_if_bit(AVX2, leaf7.ebx, 5);
add_feature_if_bit(AVX5124FMAPS, leaf7.edx, 3);
add_feature_if_bit(AVX5124VNNIW, leaf7.edx, 2);
add_feature_if_bit(AVX512BF16, leaf7s1.eax, 5);
add_feature_if_bit(AVX512BITALG, leaf7.ecx, 12);
add_feature_if_bit(AVX512BW, leaf7.ebx, 30);
add_feature_if_bit(AVX512CD, leaf7.ebx, 28);
add_feature_if_bit(AVX512DQ, leaf7.ebx, 17);
add_feature_if_bit(AVX512ER, leaf7.ebx, 27);
add_feature_if_bit(AVX512F, leaf7.ebx, 16);
add_feature_if_bit(AVX512FP16, leaf7.edx, 23);
add_feature_if_bit(AVX512IFMA, leaf7.ebx, 21);
add_feature_if_bit(AVX512PF, leaf7.ebx, 26);
add_feature_if_bit(AVX512VBMI, leaf7.ecx, 1);
add_feature_if_bit(AVX512VBMI2, leaf7.ecx, 6);
add_feature_if_bit(AVX512VL, leaf7.ebx, 31);
add_feature_if_bit(AVX512VNNI, leaf7.ecx, 11);
add_feature_if_bit(AVX512VP2INTERSECT, leaf7.edx, 8);
add_feature_if_bit(AVX512VPOPCNTDQ, leaf7.ecx, 14);
add_feature_if_bit(AVXIFMA, leaf7s1.eax, 23);
add_feature_if_bit(AVXNECONVERT, leaf7s1.edx, 5);
add_feature_if_bit(AVXVNNI, leaf7s1.eax, 4);
add_feature_if_bit(AVXVNNIINT16, leaf7s1.edx, 10);
add_feature_if_bit(AVXVNNIINT8, leaf7s1.edx, 4);
add_feature_if_bit(BMI, leaf7.ebx, 3);
add_feature_if_bit(BMI2, leaf7.ebx, 8);
add_feature_if_bit(CLDEMOTE, leaf7.ecx, 25);
add_feature_if_bit(CLFLUSHOPT, leaf7.ebx, 23);
add_feature_if_bit(CLWB, leaf7.ebx, 24);
add_feature_if_bit(CLZERO, ext8.ecx, 0);
add_feature_if_bit(CMOV, feat.edx, 15);
add_feature_if_bit(CMPCCXADD, leaf7s1.eax, 7);
add_feature_if_bit(CMPXCHG16B, feat.ecx, 12);
add_feature_if_bit(CX8, feat.edx, 8);
add_feature_if_bit(ENQCMD, leaf7.ecx, 29);
add_feature_if_bit(F16C, feat.ecx, 29);
add_feature_if_bit(FMA, feat.ecx, 12);
add_feature_if_bit(FMA4, ext1.ecx, 16);
add_feature_if_bit(FSGSBASE, leaf7.ebx, 0);
add_feature_if_bit(FXSR, feat.edx, 24);
add_feature_if_bit(GFNI, leaf7.ecx, 8);
add_feature_if_bit(HRESET, leaf7s1.eax, 22);
add_feature_if_bit(INVPCID, leaf7.ebx, 10);
add_feature_if_bit(KL, leaf7.ecx, 23);
add_feature_if_bit(LWP, ext1.ecx, 15);
add_feature_if_bit(LZCNT, ext1.ecx, 5);
add_feature_if_bit(MMX, feat.edx, 23);
add_feature_if_bit(MOVBE, feat.ecx, 22);
add_feature_if_bit(MOVDIR64B, leaf7.ecx, 28);
add_feature_if_bit(MOVDIRI, leaf7.ecx, 27);
add_feature_if_bit(MWAITX, ext1.ecx, 29);
add_feature_if_bit(PCLMUL, feat.ecx, 1);
add_feature_if_bit(PCONFIG, leaf7.edx, 18);
add_feature_if_bit(PKU, leaf7.ecx, 4);
add_feature_if_bit(POPCNT, feat.ecx, 23);
add_feature_if_bit(PREFETCHI, leaf7s1.edx, 14);
add_feature_if_bit(PREFETCHWT1, leaf7.ecx, 0);
add_feature_if_bit(PRFCHW, ext1.ecx, 8);
add_feature_if_bit(PTWRITE, leaf_14.ebx, 4);
add_feature_if_bit(RAOINT, leaf7s1.eax, 3);
add_feature_if_bit(RDPID, leaf7.ecx, 22);
add_feature_if_bit(RDPRU, ext8.ecx, 4);
add_feature_if_bit(RDRND, feat.ecx, 30);
add_feature_if_bit(RDSEED, leaf7.ebx, 18);
add_feature_if_bit(RTM, leaf7.ebx, 11);
add_feature_if_bit(SAHF, ext1.ecx, 0);
add_feature_if_bit(SERIALIZE, leaf7.edx, 14);
add_feature_if_bit(SGX, leaf7.ebx, 2);
add_feature_if_bit(SHA, leaf7.ebx, 29);
add_feature_if_bit(SHA512, leaf7s1.eax, 0);
add_feature_if_bit(SHSTK, leaf7.ecx, 7);
add_feature_if_bit(SM3, leaf7s1.eax, 1);
add_feature_if_bit(SM4, leaf7s1.eax, 2);
add_feature_if_bit(SSE, feat.edx, 25);
add_feature_if_bit(SSE2, feat.edx, 26);
add_feature_if_bit(SSE3, feat.ecx, 0);
add_feature_if_bit(SSE4_1, feat.ecx, 19);
add_feature_if_bit(SSE4_2, feat.ecx, 20);
add_feature_if_bit(SSE4_A, ext1.ecx, 6);
add_feature_if_bit(SSSE3, feat.ecx, 9);
add_feature_if_bit(TBM, ext1.ecx, 21);
add_feature_if_bit(TSXLDTRK, leaf7.edx, 16);
add_feature_if_bit(UINTR, leaf7.edx, 5);
add_feature_if_bit(USERMSR, leaf7s1.edx, 15);
add_feature_if_bit(VAES, leaf7.ecx, 9);
add_feature_if_bit(VPCLMULQDQ, leaf7.ecx, 10);
add_feature_if_bit(WAITPKG, leaf7.ecx, 5);
add_feature_if_bit(WBNOINVD, ext8.ecx, 9);
add_feature_if_bit(WIDEKL, leaf_19.ebx, 2);
add_feature_if_bit(X87, feat.edx, 0);
add_feature_if_bit(XOP, ext1.ecx, 11);
add_feature_if_bit(XSAVE, feat.ecx, 26);
add_feature_if_bit(XSAVEC, leaf_d.eax, 1);
add_feature_if_bit(XSAVEOPT, leaf_d.eax, 0);
add_feature_if_bit(XSAVES, leaf_d.eax, 3);
}

View File

@@ -0,0 +1,254 @@
module std::core::machoruntime @if(env::DARWIN) @private;
struct SegmentCommand64
{
uint cmd;
uint cmdsize;
char[16] segname;
ulong vmaddr;
ulong vmsize;
ulong fileoff;
ulong filesize;
uint maxprot;
uint initprot;
uint nsects;
uint flags;
}
struct LoadCommand
{
uint cmd;
uint cmdsize;
}
struct Section64
{
char[16] sectname;
char[16] segname;
ulong addr;
ulong size;
uint offset;
uint align;
uint reloff;
uint nreloc;
uint flags;
uint reserved1;
uint reserved2;
uint reserved3;
}
struct MachHeader
{
uint magic;
uint cputype;
uint cpusubtype;
uint filetype;
uint ncmds;
uint sizeofcmds;
uint flags;
}
struct MachHeader64
{
inline MachHeader header;
uint reserved;
}
const LC_SEGMENT_64 = 0x19;
fault MachoSearch
{
NOT_FOUND
}
fn bool name_cmp(char* a, char[16]* b)
{
for (usz i = 0; i < 16; i++)
{
if (a[i] != (*b)[i]) return false;
if (a[i] == '\0') return true;
}
return false;
}
fn SegmentCommand64*! find_segment(MachHeader* header, char* segname)
{
LoadCommand* command = (void*)header + MachHeader64.sizeof;
for (uint i = 0; i < header.ncmds; i++)
{
if (command.cmd == LC_SEGMENT_64)
{
SegmentCommand64* segment = (SegmentCommand64*)command;
if (name_cmp(segname, &segment.segname)) return segment;
}
command = (void*)command + command.cmdsize;
}
return MachoSearch.NOT_FOUND?;
}
fn Section64*! find_section(SegmentCommand64* command, char* sectname)
{
Section64* section = (void*)command + SegmentCommand64.sizeof;
for (uint i = 0; i < command.nsects; i++)
{
if (name_cmp(sectname, &section.sectname)) return section;
section++;
}
return MachoSearch.NOT_FOUND?;
}
macro find_segment_section_body(MachHeader* header, char* segname, char* sectname, $Type)
{
Section64*! section = find_section(find_segment(header, segname), sectname);
if (catch section)
{
return $Type[] {};
}
$Type* ptr = (void*)header + section.offset;
return ptr[:section.size / $Type.sizeof];
}
def DyldCallback = fn void (MachHeader* mh, isz vmaddr_slide);
extern fn void _dyld_register_func_for_add_image(DyldCallback);
struct DlInfo
{
char* dli_fname;
void* dli_fbase;
char* dli_sname;
void* dli_saddr;
}
extern fn void printf(char*, ...);
extern fn int dladdr(MachHeader* mh, DlInfo* dlinfo);
extern fn void* realloc(void* ptr, usz size);
extern fn void* malloc(usz size);
extern fn void free(void* ptr);
def CallbackFn = fn void();
struct Callback
{
uint priority;
CallbackFn xtor;
Callback* next;
}
struct DynamicMethod
{
void* fn_ptr;
char* sel;
union
{
DynamicMethod* next;
TypeId* type;
}
}
enum StartupState
{
NOT_STARTED,
INIT,
RUN_CTORS,
READ_DYLIB,
RUN_DYLIB_CTORS,
RUN_DTORS,
SHUTDOWN
}
StartupState runtime_state = NOT_STARTED;
Callback* ctor_first;
Callback* dtor_first;
fn void runtime_startup() @public @export("__c3_runtime_startup")
{
if (runtime_state != NOT_STARTED) return;
runtime_state = INIT;
_dyld_register_func_for_add_image(&dl_reg_callback);
assert(runtime_state == INIT);
runtime_state = RUN_CTORS;
Callback* ctor = ctor_first;
while (ctor)
{
ctor.xtor();
ctor = ctor.next;
}
assert(runtime_state == RUN_CTORS);
runtime_state = READ_DYLIB;
ctor = null;
}
fn void runtime_finalize() @public @export("__c3_runtime_finalize")
{
if (runtime_state != READ_DYLIB) return;
runtime_state = RUN_DTORS;
Callback* dtor = dtor_first;
while (dtor)
{
dtor.xtor();
dtor = dtor.next;
}
assert(runtime_state == RUN_DTORS);
runtime_state = SHUTDOWN;
}
fn void append_xxlizer(Callback** ref, Callback* cb)
{
while (Callback* current = *ref, current)
{
if (current.priority > cb.priority)
{
cb.next = current;
break;
}
ref = &current.next;
}
*ref = cb;
}
struct TypeId
{
char type;
TypeId* parentof;
DynamicMethod* dtable;
usz sizeof;
TypeId* inner;
usz len;
typeid[*] additional;
}
fn void dl_reg_callback(MachHeader* mh, isz vmaddr_slide)
{
usz size = 0;
assert(runtime_state == INIT || runtime_state == READ_DYLIB, "State was %s", runtime_state);
foreach (&dm : find_segment_section_body(mh, "__DATA", "__c3_dynamic", DynamicMethod))
{
TypeId* type = dm.type;
dm.next = type.dtable;
type.dtable = dm;
DynamicMethod* m = dm;
while (m)
{
m = m.next;
}
}
foreach (&cb : find_segment_section_body(mh, "__DATA", "__c3dtor", Callback))
{
append_xxlizer(&dtor_first, cb);
}
foreach (&cb : find_segment_section_body(mh, "__DATA", "__c3ctor", Callback))
{
append_xxlizer(&ctor_first, cb);
}
if (runtime_state != READ_DYLIB) return;
runtime_state = RUN_DYLIB_CTORS;
Callback* ctor = ctor_first;
ctor_first = null;
while (ctor)
{
ctor.xtor();
ctor = ctor.next;
}
assert(runtime_state == RUN_DYLIB_CTORS);
runtime_state = READ_DYLIB;
}

View File

@@ -21,7 +21,7 @@ macro int @main_to_void_main(#m, int, char**)
macro String[] args_to_strings(int argc, char** argv) @private
{
String[] list = mem::new_array(String, argc);
String[] list = mem::alloc_array(String, argc);
for (int i = 0; i < argc; i++)
{
char* arg = argv[i];
@@ -68,7 +68,7 @@ macro String[] win_command_line_to_strings(ushort* cmd_line) @private
macro String[] wargs_strings(int argc, Char16** argv) @private
{
String[] list = mem::new_array(String, argc);
String[] list = mem::alloc_array(String, argc);
for (int i = 0; i < argc; i++)
{
Char16* arg = argv[i];

View File

@@ -2,15 +2,15 @@
// Use of this source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
module std::core::runtime;
import libc;
import libc, std::time, std::io, std::sort;
struct AnyStruct
struct AnyRaw
{
void* ptr;
typeid type;
}
struct SubArrayStruct
struct SliceRaw
{
void* ptr;
usz len;
@@ -24,11 +24,11 @@ struct BenchmarkUnit
BenchmarkFn func;
}
fn BenchmarkUnit[] benchmark_collection_create(Allocator* allocator = mem::heap())
fn BenchmarkUnit[] benchmark_collection_create(Allocator allocator = allocator::heap())
{
BenchmarkFn[] fns = $$BENCHMARK_FNS;
String[] names = $$BENCHMARK_NAMES;
BenchmarkUnit[] benchmarks = allocator.new_array(BenchmarkUnit, names.len);
BenchmarkUnit[] benchmarks = allocator::alloc_array(allocator, BenchmarkUnit, names.len);
foreach (i, benchmark : fns)
{
benchmarks[i] = { names[i], fns[i] };
@@ -130,7 +130,7 @@ fn bool default_benchmark_runner()
{
@pool()
{
return run_benchmarks(benchmark_collection_create(mem::temp()));
return run_benchmarks(benchmark_collection_create(allocator::temp()));
};
}
@@ -142,11 +142,11 @@ struct TestUnit
TestFn func;
}
fn TestUnit[] test_collection_create(Allocator* allocator = mem::heap())
fn TestUnit[] test_collection_create(Allocator allocator = allocator::heap())
{
TestFn[] fns = $$TEST_FNS;
String[] names = $$TEST_NAMES;
TestUnit[] tests = allocator.new_array(TestUnit, names.len);
TestUnit[] tests = allocator::alloc_array(allocator, TestUnit, names.len);
foreach (i, test : fns)
{
tests[i] = { names[i], fns[i] };
@@ -236,7 +236,7 @@ fn bool default_test_runner()
{
@pool()
{
return run_tests(test_collection_create(mem::temp()));
return run_tests(test_collection_create(allocator::temp()));
};
}

View File

@@ -31,6 +31,11 @@ fault NumberConversion
FLOAT_OUT_OF_RANGE,
}
/**
* Return a temporary String created using the formatting function.
*
* @param [in] fmt `The formatting string`
**/
macro String tformat(String fmt, ...)
{
DString str = dstring::temp_with_capacity(fmt.len + $vacount * 8);
@@ -38,7 +43,25 @@ macro String tformat(String fmt, ...)
return str.str_view();
}
macro String new_format(String fmt, ..., Allocator* allocator = mem::heap())
/**
* Return a temporary ZString created using the formatting function.
*
* @param [in] fmt `The formatting string`
**/
macro ZString tformat_zstr(String fmt, ...)
{
DString str = dstring::temp_with_capacity(fmt.len + $vacount * 8);
str.appendf(fmt, $vasplat());
return str.zstr_view();
}
/**
* Return a new String created using the formatting function.
*
* @param [in] fmt `The formatting string`
* @param [inout] allocator `The allocator to use`
**/
macro String new_format(String fmt, ..., Allocator allocator = allocator::heap())
{
@pool(allocator)
{
@@ -48,18 +71,41 @@ macro String new_format(String fmt, ..., Allocator* allocator = mem::heap())
};
}
/**
* Return a new ZString created using the formatting function.
*
* @param [in] fmt `The formatting string`
* @param [inout] allocator `The allocator to use`
**/
macro ZString new_format_zstr(String fmt, ..., Allocator allocator = allocator::heap())
{
@pool(allocator)
{
DString str = dstring::temp_with_capacity(fmt.len + $vacount * 8);
str.appendf(fmt, $vasplat());
return str.copy_zstr(allocator);
};
}
/**
* Check if a character is in a set.
*
* @param c `the character to check`
* @param [in] set `The formatting string`
* @pure
* @return `True if a character is in the set`
**/
macro bool char_in_set(char c, String set)
{
foreach (ch : set) if (ch == c) return true;
return false;
}
fn String join_new(String[] s, String joiner, Allocator* allocator = mem::heap())
fn String join_new(String[] s, String joiner, Allocator allocator = allocator::heap())
{
if (!s)
{
return (String)allocator.new_zero_array(char, 2)[:0];
return (String)allocator::new_array(allocator, char, 2)[:0];
}
usz total_size = joiner.len * s.len;
@@ -81,8 +127,12 @@ fn String join_new(String[] s, String joiner, Allocator* allocator = mem::heap()
}
/**
* @param [in] string
* @param [in] to_trim
* Remove characters from the front and end of a string.
*
* @param [in] string `The string to trim`
* @param [in] to_trim `The set of characters to trim, defaults to whitespace`
* @pure
* @return `a substring of the string passed in`
**/
fn String String.trim(string, String to_trim = "\t\n\r ")
{
@@ -96,8 +146,12 @@ fn String String.trim(string, String to_trim = "\t\n\r ")
}
/**
* Check if the String starts with the needle.
*
* @param [in] string
* @param [in] needle
* @pure
* @return `'true' if the string starts with the needle`
**/
fn bool String.starts_with(string, String needle)
{
@@ -107,8 +161,12 @@ fn bool String.starts_with(string, String needle)
}
/**
* Check if the String ends with the needle.
*
* @param [in] string
* @param [in] needle
* @pure
* @return `'true' if the string ends with the needle`
**/
fn bool String.ends_with(string, String needle)
{
@@ -122,6 +180,8 @@ fn bool String.ends_with(string, String needle)
*
* @param [in] string
* @param [in] needle
* @pure
* @return `the substring with the prefix removed`
**/
fn String String.strip(string, String needle)
{
@@ -134,6 +194,8 @@ fn String String.strip(string, String needle)
*
* @param [in] string
* @param [in] needle
* @pure
* @return `the substring with the suffix removed`
**/
fn String String.strip_end(string, String needle)
{
@@ -153,11 +215,11 @@ fn String String.strip_end(string, String needle)
* @require needle.len > 0 "The needle must be at least 1 character long"
* @ensure return.len > 0
**/
fn String[] String.split(s, String needle, usz max = 0, Allocator* allocator = mem::heap())
fn String[] String.split(s, String needle, usz max = 0, Allocator allocator = allocator::heap())
{
usz capacity = 16;
usz i = 0;
String* holder = allocator.new_array(String, capacity);
String* holder = allocator::alloc_array(allocator, String, capacity);
bool no_more = false;
while (!no_more)
{
@@ -176,7 +238,7 @@ fn String[] String.split(s, String needle, usz max = 0, Allocator* allocator = m
if (i == capacity)
{
capacity *= 2;
holder = allocator.realloc(holder, String.sizeof * capacity);
holder = allocator::realloc(allocator, holder, String.sizeof * capacity);
}
holder[i++] = res;
}
@@ -193,9 +255,17 @@ fn String[] String.split(s, String needle, usz max = 0, Allocator* allocator = m
**/
fn String[] String.tsplit(s, String needle, usz max = 0)
{
return s.split(needle, max, mem::temp()) @inline;
return s.split(needle, max, allocator::temp()) @inline;
}
/**
* Check if a substring is found in the string.
* @param [in] s
* @param [in] needle "The string to look for."
* @pure
* @return "true if the string contains the substring, false otherwise"
**/
fn bool String.contains(s, String needle)
{
return @ok(s.index_of(needle));
@@ -205,6 +275,7 @@ fn bool String.contains(s, String needle)
* Find the index of the first incidence of a string.
*
* @param [in] s
* @param needle "The character to look for"
* @pure
* @ensure return < s.len
* @return "the index of the needle"
@@ -220,9 +291,32 @@ fn usz! String.index_of_char(s, char needle)
}
/**
* Find the index of the first incidence of a string.
* Find the index of the first incidence of a character.
*
* @param [in] s
* @param needle "The character to look for"
* @param start_index "The index to start with, may exceed max index."
* @pure
* @ensure return < s.len
* @return "the index of the needle"
* @return! SearchResult.MISSING "if the needle cannot be found starting from the start_index"
**/
fn usz! String.index_of_char_from(s, char needle, usz start_index)
{
usz len = s.len;
if (len <= start_index) return SearchResult.MISSING?;
for (usz i = start_index; i < len; i++)
{
if (s[i] == needle) return i;
}
return SearchResult.MISSING?;
}
/**
* Find the index of the first incidence of a character starting from the end.
*
* @param [in] s
* @param needle "the character to find"
* @pure
* @ensure return < s.len
* @return "the index of the needle"
@@ -312,19 +406,19 @@ fn usz ZString.len(str)
}
fn ZString String.zstr_copy(s, Allocator* allocator = mem::heap())
fn ZString String.zstr_copy(s, Allocator allocator = allocator::heap())
{
usz len = s.len;
char* str = allocator.alloc(len + 1);
char* str = allocator::malloc(allocator, len + 1);
mem::copy(str, s.ptr, len);
str[len] = 0;
return (ZString)str;
}
fn String String.concat(s1, String s2, Allocator* allocator = mem::heap())
fn String String.concat(s1, String s2, Allocator allocator = allocator::heap())
{
usz full_len = s1.len + s2.len;
char* str = allocator.alloc(full_len + 1);
char* str = allocator::malloc(allocator, full_len + 1);
usz s1_len = s1.len;
mem::copy(str, s1.ptr, s1_len);
mem::copy(str + s1_len, s2.ptr, s2.len);
@@ -332,37 +426,37 @@ fn String String.concat(s1, String s2, Allocator* allocator = mem::heap())
return (String)str[:full_len];
}
fn String String.tconcat(s1, String s2) => s1.concat(s2, mem::temp());
fn String String.tconcat(s1, String s2) => s1.concat(s2, allocator::temp());
fn ZString String.zstr_tcopy(s) => s.zstr_copy(mem::temp()) @inline;
fn ZString String.zstr_tcopy(s) => s.zstr_copy(allocator::temp()) @inline;
fn String String.copy(s, Allocator* allocator = mem::heap())
fn String String.copy(s, Allocator allocator = allocator::heap())
{
usz len = s.len;
char* str = allocator.alloc(len + 1);
char* str = allocator::malloc(allocator, len + 1);
mem::copy(str, s.ptr, len);
str[len] = 0;
return (String)str[:len];
}
fn void String.free(&s, Allocator* allocator = mem::heap())
fn void String.free(&s, Allocator allocator = allocator::heap())
{
if (!s.len) return;
allocator.free(s.ptr);
allocator::free(allocator, s.ptr);
*s = "";
}
fn String String.tcopy(s) => s.copy(mem::temp()) @inline;
fn String String.tcopy(s) => s.copy(allocator::temp()) @inline;
fn String ZString.copy(z, Allocator* allocator = mem::temp())
fn String ZString.copy(z, Allocator allocator = allocator::temp())
{
return z.str_view().copy(allocator) @inline;
}
fn String ZString.tcopy(z)
{
return z.str_view().copy(mem::temp()) @inline;
return z.str_view().copy(allocator::temp()) @inline;
}
/**
@@ -371,10 +465,10 @@ fn String ZString.tcopy(z)
* @return! UnicodeResult.INVALID_UTF8 "If the string contained an invalid UTF-8 sequence"
* @return! AllocationFailure "If allocation of the string fails"
**/
fn Char16[]! String.to_new_utf16(s, Allocator* allocator = mem::heap())
fn Char16[]! String.to_new_utf16(s, Allocator allocator = allocator::heap())
{
usz len16 = conv::utf16len_for_utf8(s);
Char16* data = allocator.new_array_checked(Char16, len16 + 1)!;
Char16* data = allocator::alloc_array_try(allocator, Char16, len16 + 1)!;
conv::utf8to16_unsafe(s, data)!;
data[len16] = 0;
return data[:len16];
@@ -388,10 +482,10 @@ fn Char16[]! String.to_new_utf16(s, Allocator* allocator = mem::heap())
**/
fn Char16[]! String.to_temp_utf16(s)
{
return s.to_new_utf16(mem::temp());
return s.to_new_utf16(allocator::temp());
}
fn WString! String.to_new_wstring(s, Allocator* allocator = mem::heap())
fn WString! String.to_new_wstring(s, Allocator allocator = allocator::heap())
{
return (WString)s.to_new_utf16(allocator).ptr;
}
@@ -401,10 +495,10 @@ fn WString! String.to_temp_wstring(s)
return (WString)s.to_temp_utf16().ptr;
}
fn Char32[]! String.to_new_utf32(s, Allocator* allocator = mem::heap())
fn Char32[]! String.to_new_utf32(s, Allocator allocator = allocator::heap())
{
usz codepoints = conv::utf8_codepoints(s);
Char32* data = allocator.new_array(Char32, codepoints + 1);
Char32* data = allocator::alloc_array_try(allocator, Char32, codepoints + 1)!;
conv::utf8to32_unsafe(s, data)!;
data[codepoints] = 0;
return data[:codepoints];
@@ -412,32 +506,52 @@ fn Char32[]! String.to_new_utf32(s, Allocator* allocator = mem::heap())
fn Char32[]! String.to_temp_utf32(s)
{
return s.to_new_utf32(mem::temp());
return s.to_new_utf32(allocator::temp());
}
/**
* Convert a string to ASCII lower case.
*
* @param [inout] s
* @pure
**/
fn void String.convert_ascii_to_lower(s)
{
foreach (&c : s) if (c.is_upper()) *c += 'a' - 'A';
foreach (&c : s) if (c.is_upper() @pure) *c += 'a' - 'A';
}
fn String String.new_ascii_to_lower(s, Allocator* allocator = mem::heap())
fn String String.new_ascii_to_lower(s, Allocator allocator = allocator::heap())
{
String copy = s.copy(allocator);
copy.convert_ascii_to_lower();
return copy;
}
fn String String.temp_ascii_to_lower(s, Allocator* allocator = mem::heap())
fn String String.temp_ascii_to_lower(s, Allocator allocator = allocator::heap())
{
return s.new_ascii_to_lower(mem::temp());
return s.new_ascii_to_lower(allocator::temp());
}
/**
* Convert a string to ASCII upper case.
*
* @param [inout] s
* @pure
**/
fn void String.convert_ascii_to_upper(s)
{
foreach (&c : s) if (c.is_lower()) *c -= 'a' - 'A';
foreach (&c : s) if (c.is_lower() @pure) *c -= 'a' - 'A';
}
fn String String.new_ascii_to_upper(s, Allocator* allocator = mem::heap())
/**
* Returns a string converted to ASCII upper case.
*
* @param [in] s
* @param [inout] allocator
*
* @return `a new String converted to ASCII upper case.`
**/
fn String String.new_ascii_to_upper(s, Allocator allocator = allocator::heap())
{
String copy = s.copy(allocator);
copy.convert_ascii_to_upper();
@@ -449,32 +563,36 @@ fn StringIterator String.iterator(s)
return { s, 0 };
}
/**
* @param [in] s
* @return `a temporary String converted to ASCII upper case.`
**/
fn String String.temp_ascii_to_upper(s)
{
return s.new_ascii_to_upper(mem::temp());
return s.new_ascii_to_upper(allocator::temp());
}
fn String! new_from_utf32(Char32[] utf32, Allocator* allocator = mem::heap())
fn String! new_from_utf32(Char32[] utf32, Allocator allocator = allocator::heap())
{
usz len = conv::utf8len_for_utf32(utf32);
char* data = allocator.alloc_checked(len + 1)!;
defer catch allocator.free(data);
char* data = allocator::malloc_try(allocator, len + 1)!;
defer catch allocator::free(allocator, data);
conv::utf32to8_unsafe(utf32, data);
data[len] = 0;
return (String)data[:len];
}
fn String! new_from_utf16(Char16[] utf16, Allocator* allocator = mem::heap())
fn String! new_from_utf16(Char16[] utf16, Allocator allocator = allocator::heap())
{
usz len = conv::utf8len_for_utf16(utf16);
char* data = allocator.alloc_checked(len + 1)!;
defer catch allocator.free(data);
char* data = allocator::malloc_try(allocator, len + 1)!;
defer catch allocator::free(allocator, data);
conv::utf16to8_unsafe(utf16, data)!;
data[len] = 0;
return (String)data[:len];
}
fn String! new_from_wstring(WString wstring, Allocator* allocator = mem::heap())
fn String! new_from_wstring(WString wstring, Allocator allocator = allocator::heap())
{
usz utf16_len;
while (wstring[utf16_len] != 0) utf16_len++;
@@ -482,8 +600,8 @@ fn String! new_from_wstring(WString wstring, Allocator* allocator = mem::heap())
return new_from_utf16(utf16, allocator);
}
fn String! temp_from_wstring(WString wstring) => new_from_wstring(wstring, mem::temp()) @inline;
fn String! temp_from_utf16(Char16[] utf16) => new_from_utf16(utf16, mem::temp()) @inline;
fn String! temp_from_wstring(WString wstring) => new_from_wstring(wstring, allocator::temp()) @inline;
fn String! temp_from_utf16(Char16[] utf16) => new_from_utf16(utf16, allocator::temp()) @inline;
fn usz String.utf8_codepoints(s)
{
@@ -495,7 +613,11 @@ fn usz String.utf8_codepoints(s)
return len;
}
macro String.to_integer(string, $Type)
/**
* @require (base <= 10 && base > 1) || base == 16 : "Unsupported base"
**/
macro String.to_integer(string, $Type, int base = 10)
{
usz len = string.len;
usz index = 0;
@@ -515,8 +637,8 @@ macro String.to_integer(string, $Type)
break;
}
if (len == index) return NumberConversion.MALFORMED_INTEGER?;
$Type base = 10;
if (string[index] == '0')
$Type base_used = ($Type)base;
if (string[index] == '0' && base == 10)
{
index++;
if (index == len) return ($Type)0;
@@ -524,15 +646,15 @@ macro String.to_integer(string, $Type)
{
case 'x':
case 'X':
base = 16;
base_used = 16;
index++;
case 'b':
case 'B':
base = 2;
base_used = 2;
index++;
case 'o':
case 'O':
base = 8;
base_used = 8;
index++;
default:
break;
@@ -544,21 +666,21 @@ macro String.to_integer(string, $Type)
{
char c = {|
char ch = string[index++];
if (base != 16 || ch < 'A') return (char)(ch - '0');
if (ch <= 'F') return (char)(ch - 'A');
if (base_used != 16 || ch < 'A') return (char)(ch - '0');
if (ch <= 'F') return (char)(ch - 'A' + 10);
if (ch < 'a') return NumberConversion.MALFORMED_INTEGER?;
if (ch > 'f') return NumberConversion.MALFORMED_INTEGER?;
return (char)(ch - 'a');
return (char)(ch - 'a' + 10);
|}!;
if (c >= base) return NumberConversion.MALFORMED_INTEGER?;
if (c >= base_used) return NumberConversion.MALFORMED_INTEGER?;
value = {|
if (is_negative)
{
$Type new_value = value * base - c;
$Type new_value = value * base_used - c;
if (new_value > value) return NumberConversion.INTEGER_OVERFLOW?;
return new_value;
}
$Type new_value = value * base + c;
$Type new_value = value * base_used + c;
if (new_value < value) return NumberConversion.INTEGER_OVERFLOW?;
return new_value;
|}!;
@@ -566,17 +688,17 @@ macro String.to_integer(string, $Type)
return value;
}
fn int128! String.to_int128(s) => s.to_integer(int128);
fn long! String.to_long(s) => s.to_integer(long);
fn int! String.to_int(s) => s.to_integer(int);
fn short! String.to_short(s) => s.to_integer(short);
fn ichar! String.to_ichar(s) => s.to_integer(ichar);
fn int128! String.to_int128(s, int base = 10) => s.to_integer(int128, base);
fn long! String.to_long(s, int base = 10) => s.to_integer(long, base);
fn int! String.to_int(s, int base = 10) => s.to_integer(int, base);
fn short! String.to_short(s, int base = 10) => s.to_integer(short, base);
fn ichar! String.to_ichar(s, int base = 10) => s.to_integer(ichar, base);
fn uint128! String.to_uint128(s) => s.to_integer(uint128);
fn ulong! String.to_ulong(s) => s.to_integer(ulong);
fn uint! String.to_uint(s) => s.to_integer(uint);
fn ushort! String.to_ushort(s) => s.to_integer(ushort);
fn char! String.to_uchar(s) => s.to_integer(char);
fn uint128! String.to_uint128(s, int base = 10) => s.to_integer(uint128, base);
fn ulong! String.to_ulong(s, int base = 10) => s.to_integer(ulong, base);
fn uint! String.to_uint(s, int base = 10) => s.to_integer(uint, base);
fn ushort! String.to_ushort(s, int base = 10) => s.to_integer(ushort, base);
fn char! String.to_uchar(s, int base = 10) => s.to_integer(char, base);
fn double! String.to_double(s) => s.to_real(double);
fn float! String.to_float(s) => s.to_real(float);

View File

@@ -11,7 +11,7 @@ fault ConversionResult
/**
* @require $Type.kindof.is_int() || $Type.kindof == TypeKind.ENUM "Argument was not an integer"
**/
macro any_to_int(any* v, $Type)
macro any_to_int(any v, $Type)
{
typeid any_type = v.type;
TypeKind kind = any_type.kindof;
@@ -108,10 +108,10 @@ fn bool TypeKind.is_int(kind) @inline
return kind == TypeKind.SIGNED_INT || kind == TypeKind.UNSIGNED_INT;
}
macro bool is_subarray_convertable($Type)
macro bool is_slice_convertable($Type)
{
$switch ($Type.kindof)
$case SUBARRAY:
$case SLICE:
return true;
$case POINTER:
return $Type.inner.kindof == TypeKind.ARRAY;
@@ -298,10 +298,11 @@ enum TypeKind : char
FUNC,
OPTIONAL,
ARRAY,
SUBARRAY,
SLICE,
VECTOR,
DISTINCT,
POINTER,
INTERFACE,
}
struct TypeEnum

View File

@@ -3,22 +3,22 @@ import std::io;
struct CsvReader
{
InStream* stream;
InStream stream;
String separator;
}
fn void CsvReader.init(&self, InStream* stream, String separator = ",")
fn void CsvReader.init(&self, InStream stream, String separator = ",")
{
self.stream = stream;
self.separator = separator;
}
fn String[]! CsvReader.read_new_row(self, Allocator* allocator = mem::heap())
fn String[]! CsvReader.read_new_row(self, Allocator allocator = allocator::heap())
{
return self.read_new_row_with_allocator(mem::temp()) @inline;
return self.read_new_row_with_allocator(allocator::temp()) @inline;
}
fn String[]! CsvReader.read_new_row_with_allocator(self, Allocator* allocator = mem::heap())
fn String[]! CsvReader.read_new_row_with_allocator(self, Allocator allocator = allocator::heap())
{
@pool(allocator)
{
@@ -28,7 +28,7 @@ fn String[]! CsvReader.read_new_row_with_allocator(self, Allocator* allocator =
fn String[]! CsvReader.read_temp_row(self)
{
return self.read_new_row_with_allocator(mem::temp()) @inline;
return self.read_new_row_with_allocator(allocator::temp()) @inline;
}
fn void! CsvReader.skip_row(self) @maydiscard
@@ -45,7 +45,7 @@ macro CsvReader.@each_row(self, int rows = int.max; @body(String[] row))
String sep = self.separator;
while (rows--)
{
@stack_mem(512; Allocator* mem)
@stack_mem(512; Allocator mem)
{
String[] parts;
@pool()

View File

@@ -15,7 +15,7 @@ fault JsonParsingError
INVALID_NUMBER,
}
fn Object*! parse(InStream* s, Allocator* allocator = mem::heap())
fn Object*! parse(InStream s, Allocator allocator = allocator::heap())
{
JsonContext context = { .last_string = dstring::new_with_capacity(64, allocator), .stream = s, .allocator = allocator };
defer context.last_string.free();
@@ -44,8 +44,8 @@ enum JsonTokenType @local
struct JsonContext @local
{
uint line;
InStream* stream;
Allocator* allocator;
InStream stream;
Allocator allocator;
JsonTokenType token;
DString last_string;
double last_number;
@@ -170,7 +170,7 @@ fn Object*! parse_array(JsonContext* context) @local
while (token != JsonTokenType.RBRACKET)
{
Object* element = parse_from_token(context, token)!;
list.append(element);
list.push(element);
token = advance(context)!;
if (token == JsonTokenType.COMMA)
{

View File

@@ -2,12 +2,12 @@ module std::io;
struct BitReader
{
InStream* reader;
InStream reader;
uint bits;
uint len;
}
fn void BitReader.init(&self, InStream* byte_reader)
fn void BitReader.init(&self, InStream byte_reader)
{
*self = { .reader = byte_reader };
}
@@ -40,12 +40,12 @@ fn char! BitReader.read_bits(&self, uint nbits)
struct BitWriter
{
OutStream* writer;
OutStream writer;
uint bits;
uint len;
}
fn void BitWriter.init(&self, OutStream* byte_writer)
fn void BitWriter.init(&self, OutStream byte_writer)
{
*self = { .writer = byte_writer };
}

View File

@@ -7,7 +7,7 @@ struct File (InStream, OutStream)
}
module std::io::file;
import libc;
import libc, std::io::path, std::io::os;
fn File! open(String filename, String mode)
{
@@ -161,14 +161,14 @@ fn char[]! load_buffer(String filename, char[] buffer)
}
fn char[]! load_new(String filename, Allocator* allocator = mem::heap())
fn char[]! load_new(String filename, Allocator allocator = allocator::heap())
{
File file = open(filename, "rb")!;
defer (void)file.close();
usz len = file.seek(0, END)!;
file.seek(0, SET)!;
char* data = allocator.alloc_checked(len)!;
defer catch allocator.free(data);
char* data = allocator::malloc_try(allocator, len)!;
defer catch allocator::free(allocator, data);
usz read = 0;
while (read < len)
{
@@ -179,7 +179,7 @@ fn char[]! load_new(String filename, Allocator* allocator = mem::heap())
fn char[]! load_temp(String filename)
{
return load_new(filename, mem::temp());
return load_new(filename, allocator::temp());
}
/**

View File

@@ -6,7 +6,7 @@ const int PRINTF_NTOA_BUFFER_SIZE = 256;
interface Printable
{
fn String to_new_string(Allocator *allocator) @optional;
fn String to_new_string(Allocator allocator) @optional;
fn usz! to_format(Formatter* formatter) @optional;
}
@@ -72,7 +72,7 @@ fn usz! Formatter.out(&self, char c) @private
return 1;
}
fn usz! Formatter.print_with_function(&self, Printable* arg)
fn usz! Formatter.print_with_function(&self, Printable arg)
{
if (&arg.to_format)
{
@@ -98,16 +98,16 @@ fn usz! Formatter.print_with_function(&self, Printable* arg)
self.width = old_width;
self.prec = old_prec;
}
@pool()
@stack_mem(1024; Allocator mem)
{
return self.out_substr(arg.to_new_string(mem::temp()));
return self.out_substr(arg.to_new_string(mem));
};
}
return SearchResult.MISSING?;
}
fn usz! Formatter.out_str(&self, any* arg) @private
fn usz! Formatter.out_str(&self, any arg) @private
{
switch (arg.type.kindof)
{
@@ -119,7 +119,7 @@ fn usz! Formatter.out_str(&self, any* arg) @private
case FAULT:
return self.out_substr((*(anyfault*)arg.ptr).nameof);
case ANY:
return self.out_str(*(any**)arg);
return self.out_str(*(any*)arg);
case OPTIONAL:
unreachable();
case SIGNED_INT:
@@ -149,7 +149,7 @@ fn usz! Formatter.out_str(&self, any* arg) @private
return self.out_substr(*(bool*)arg.ptr ? "true" : "false");
default:
}
usz! n = self.print_with_function((Printable*)arg);
usz! n = self.print_with_function((Printable)arg);
if (catch err = n)
{
case SearchResult.MISSING:
@@ -246,7 +246,7 @@ fn usz! Formatter.out_str(&self, any* arg) @private
}
len += self.out_substr(">]")!;
return len;
case SUBARRAY:
case SLICE:
// this is SomeType[] so grab the "SomeType"
typeid inner = arg.type.inner;
if (inner == char.typeid)
@@ -291,7 +291,7 @@ fn void! out_null_fn(void* data @unused, char c @unused) @private
fn usz! Formatter.vprintf(&self, String format, any*[] anys)
fn usz! Formatter.vprintf(&self, String format, any[] anys)
{
if (!self.out_fn)
{
@@ -358,7 +358,7 @@ fn usz! Formatter.vprintf(&self, String format, any*[] anys)
// evaluate specifier
uint base = 0;
if (variant_index >= anys.len) return PrintFault.MISSING_ARG?;
any* current = anys[variant_index++];
any current = anys[variant_index++];
switch (c)
{
case 'd':

View File

@@ -1,4 +1,5 @@
module std::io;
import std::math;
const char[16] XDIGITS_H = "0123456789ABCDEF";
const char[16] XDIGITS_L = "0123456789abcdef";
@@ -9,7 +10,7 @@ fn usz! Formatter.adjust(&self, usz len) @local
return self.pad(' ', self.width, len);
}
fn uint128! int_from_any(any* arg, bool *is_neg) @private
fn uint128! int_from_any(any arg, bool *is_neg) @private
{
switch (arg.type.kindof)
{
@@ -62,14 +63,11 @@ fn uint128! int_from_any(any* arg, bool *is_neg) @private
}
}
fn FloatType! float_from_any(any* arg) @private
fn FloatType! float_from_any(any arg) @private
{
$if env::F128_SUPPORT:
if (arg.type == float128.typeid) return (FloatType)*((float128*)arg.ptr);
$endif
$if env::F16_SUPPORT:
if (arg.type == float16.typeid) return *((float16*)arg.ptr);
$endif
if (arg.type.kindof == TypeKind.DISTINCT)
{
return float_from_any(arg.as_inner());
@@ -587,14 +585,14 @@ fn usz! Formatter.ntoa_format(&self, String buf, usz len, bool negative, uint ba
}
fn usz! Formatter.ntoa_any(&self, any* arg, uint base) @private
fn usz! Formatter.ntoa_any(&self, any arg, uint base) @private
{
bool is_neg;
uint128 val = int_from_any(arg, &is_neg)!!;
return self.ntoa(val, is_neg, base) @inline;
}
fn usz! Formatter.out_char(&self, any* arg) @private
fn usz! Formatter.out_char(&self, any arg) @private
{
usz len = 1;
uint l = 1;
@@ -648,21 +646,21 @@ fn void! printf_advance_format(usz format_len, usz *index_ptr) @inline @private
if (val >= format_len) return FormattingFault.UNTERMINATED_FORMAT?;
}
fn any*! next_any(any** args_ptr, usz args_len, usz* arg_index_ptr) @inline @private
fn any! next_any(any* args_ptr, usz args_len, usz* arg_index_ptr) @inline @private
{
if (*arg_index_ptr >= args_len) return FormattingFault.MISSING_ARG?;
return args_ptr[(*arg_index_ptr)++];
}
fn int! printf_parse_format_field(
any** args_ptr, usz args_len, usz* args_index_ptr,
any* args_ptr, usz args_len, usz* args_index_ptr,
char* format_ptr, usz format_len, usz* index_ptr) @inline @private
{
char c = format_ptr[*index_ptr];
if (c.is_digit()) return simple_atoi(format_ptr, format_len, index_ptr);
if (c != '*') return 0;
printf_advance_format(format_len, index_ptr)!;
any* val = next_any(args_ptr, args_len, args_index_ptr)!;
any val = next_any(args_ptr, args_len, args_index_ptr)!;
if (!val.type.kindof.is_int()) return FormattingFault.INVALID_WIDTH_ARG?;
uint! intval = types::any_to_int(val, int);
return intval ?? FormattingFault.INVALID_WIDTH_ARG?;

View File

@@ -49,9 +49,9 @@ fault IoError
* @param stream
* @require @is_instream(stream)
**/
macro String! readline(stream = io::stdin(), Allocator* allocator = mem::heap())
macro String! readline(stream = io::stdin(), Allocator allocator = allocator::heap())
{
bool $is_stream = @typeid(stream) == InStream*.typeid;
bool $is_stream = @typeid(stream) == InStream.typeid;
$if $is_stream:
$typeof(&stream.read_byte) func = &stream.read_byte;
char val = func((void*)stream)!;
@@ -84,7 +84,7 @@ macro String! readline(stream = io::stdin(), Allocator* allocator = mem::heap())
};
}
macro String! treadline(stream = io::stdin()) => readline(stream, mem::temp()) @inline;
macro String! treadline(stream = io::stdin()) => readline(stream, allocator::temp()) @inline;
/**
* @require @is_outstream(out) "The output must implement OutStream"
@@ -108,14 +108,14 @@ macro usz! fprint(out, x)
$endswitch
}
fn usz! fprintf(OutStream* out, String format, args...)
fn usz! fprintf(OutStream out, String format, args...)
{
Formatter formatter;
formatter.init(&out_putstream_fn, &out);
return formatter.vprintf(format, args);
}
fn usz! fprintfn(OutStream* out, String format, args...) @maydiscard
fn usz! fprintfn(OutStream out, String format, args...) @maydiscard
{
Formatter formatter;
formatter.init(&out_putstream_fn, &out);
@@ -133,7 +133,7 @@ macro usz! fprintn(out, x = "")
usz len = fprint(out, x)!;
out.write_byte('\n')!;
$switch
$case @typeid(out) == OutStream*.typeid:
$case @typeid(out) == OutStream.typeid:
if (&out.flush) out.flush()!;
$case $defined(out.flush):
out.flush()!;
@@ -164,7 +164,7 @@ macro void eprintn(x)
fn void! out_putstream_fn(void* data, char c) @private
{
OutStream** stream = data;
OutStream* stream = data;
return (*stream).write_byte(c);
}
@@ -193,7 +193,7 @@ fn usz! printfn(String format, args...) @maydiscard
fn usz! eprintf(String format, args...) @maydiscard
{
Formatter formatter;
OutStream* stream = stderr();
OutStream stream = stderr();
formatter.init(&out_putstream_fn, &stream);
return formatter.vprintf(format, args);
}
@@ -202,7 +202,7 @@ fn usz! eprintf(String format, args...) @maydiscard
fn usz! eprintfn(String format, args...) @maydiscard
{
Formatter formatter;
OutStream* stream = stderr();
OutStream stream = stderr();
formatter.init(&out_putstream_fn, &stream);
usz len = formatter.vprintf(format, args)! + 1;
stderr().write_byte('\n')!;

View File

@@ -1,5 +1,5 @@
module std::io::os;
import libc;
import std::io::path, libc, std::os;
macro void! native_chdir(Path path)
{

View File

@@ -1,5 +1,5 @@
module std::io::os;
import libc;
import libc, std::os, std::io;
fn void! native_stat(Stat* stat, String path) @if(env::DARWIN || env::LINUX)
{

View File

@@ -1,7 +1,7 @@
module std::io::os;
import libc;
import libc, std::os;
macro String! getcwd(Allocator* allocator = mem::heap())
macro String! getcwd(Allocator allocator = allocator::heap())
{
$switch
$case env::WIN32:

View File

@@ -1,9 +1,10 @@
module std::io::file::os @if(env::POSIX);
module std::io::os @if(env::POSIX);
import std::io, std::os;
fn PathList! native_ls(Path dir, bool no_dirs, bool no_symlinks, String mask, Allocator* allocator)
fn PathList! native_ls(Path dir, bool no_dirs, bool no_symlinks, String mask, Allocator allocator)
{
PathList list;
list.init_new(.allocator = allocator);
list.new_init(.allocator = allocator);
DIRPtr directory = posix::opendir(dir.str_view() ? dir.as_zstr() : (ZString)".");
defer if (directory) posix::closedir(directory);
if (!directory) return (path::is_dir(dir) ? IoError.CANNOT_READ_DIR : IoError.FILE_NOT_DIR)?;
@@ -15,17 +16,18 @@ fn PathList! native_ls(Path dir, bool no_dirs, bool no_symlinks, String mask, Al
if (entry.d_type == posix::DT_LNK && no_symlinks) continue;
if (entry.d_type == posix::DT_DIR && no_dirs) continue;
Path path = path::new(name, allocator)!!;
list.append(path);
list.push(path);
}
return list;
}
module std::io::os @if(env::WIN32);
import std::time, std::os, std::io;
fn PathList! native_ls(Path dir, bool no_dirs, bool no_symlinks, String mask, Allocator* allocator)
fn PathList! native_ls(Path dir, bool no_dirs, bool no_symlinks, String mask, Allocator allocator)
{
PathList list;
list.init_new(.allocator = allocator);
list.new_init(.allocator = allocator);
@pool(allocator)
{
@@ -41,7 +43,7 @@ fn PathList! native_ls(Path dir, bool no_dirs, bool no_symlinks, String mask, Al
{
String filename = string::temp_from_wstring((WString)&find_data.cFileName)!;
if (filename == ".." || filename == ".") continue;
list.append(path::new(filename, allocator)!);
list.push(path::new(filename, allocator)!);
};
} while (win32::findNextFileW(find, &find_data));
return list;

View File

@@ -1,5 +1,5 @@
module std::io::file::os @if(env::POSIX);
import libc;
module std::io::os @if(env::POSIX);
import std::io, std::os, libc;
/**
* @require dir.str_view()
@@ -33,6 +33,7 @@ fn void! native_rmtree(Path dir)
}
module std::io::os @if(env::WIN32);
import std::io, std::time, std::os;
fn void! native_rmtree(Path path)
{
@@ -46,7 +47,7 @@ fn void! native_rmtree(Path path)
{
@pool()
{
String filename = string::new_from_wstring((WString)&find_data.cFileName, mem::temp())!;
String filename = string::new_from_wstring((WString)&find_data.cFileName, allocator::temp())!;
if (filename == "." || filename == "..") continue;
Path file_path = path.tappend(filename)!;
if (find_data.dwFileAttributes & win32::FILE_ATTRIBUTE_DIRECTORY)

View File

@@ -1,6 +1,7 @@
module std::io::os @if(env::LIBC);
import std::io::path, std::os;
fn Path! native_temp_directory(Allocator* allocator = mem::heap()) @if(!env::WIN32)
fn Path! native_temp_directory(Allocator allocator = allocator::heap()) @if(!env::WIN32)
{
foreach (String env : { "TMPDIR", "TMP", "TEMP", "TEMPDIR" })
{
@@ -10,13 +11,13 @@ fn Path! native_temp_directory(Allocator* allocator = mem::heap()) @if(!env::WIN
return path::new("/tmp", allocator);
}
fn Path! native_temp_directory(Allocator* allocator = mem::heap()) @if(env::WIN32)
fn Path! native_temp_directory(Allocator allocator = allocator::heap()) @if(env::WIN32)
{
@pool(allocator)
{
Win32_DWORD len = win32::getTempPathW(0, null);
if (!len) return IoError.GENERAL_ERROR?;
Char16[] buff = mem::temp_array(Char16, len + (usz)1);
Char16[] buff = mem::temp_alloc_array(Char16, len + (usz)1);
if (!win32::getTempPathW(len, buff)) return IoError.GENERAL_ERROR?;
return path::new(string::temp_from_utf16(buff[:len]), allocator);
};
@@ -24,7 +25,7 @@ fn Path! native_temp_directory(Allocator* allocator = mem::heap()) @if(env::WIN3
module std::io::os @if(env::NO_LIBC);
macro Path! native_temp_directory(Allocator* allocator = mem::heap())
macro Path! native_temp_directory(Allocator allocator = allocator::heap())
{
return IoError.UNSUPPORTED_OPERATION?;
}

View File

@@ -1,5 +1,6 @@
module std::io::path;
import std::collections::list;
import std::collections::list, std::io::os;
import std::os::win32;
const PathEnv DEFAULT_PATH_ENV = env::WIN32 ? PathEnv.WIN32 : PathEnv.POSIX;
const char PREFERRED_SEPARATOR_WIN32 = '\\';
@@ -26,11 +27,11 @@ enum PathEnv
POSIX
}
fn Path! getcwd(Allocator* allocator = mem::heap())
fn Path! getcwd(Allocator allocator = allocator::heap())
{
@pool(allocator)
{
return new(os::getcwd(mem::temp()), allocator);
return new(os::getcwd(allocator::temp()), allocator);
};
}
@@ -38,9 +39,9 @@ fn bool is_dir(Path path) => os::native_is_dir(path.str_view());
fn bool is_file(Path path) => os::native_is_file(path.str_view());
fn usz! file_size(Path path) => os::native_file_size(path.str_view());
fn bool exists(Path path) => os::native_file_or_dir_exists(path.str_view());
fn Path! tgetcwd() => getcwd(mem::temp()) @inline;
fn Path! tgetcwd() => getcwd(allocator::temp()) @inline;
fn void! chdir(Path path) => os::native_chdir(path) @inline;
fn Path! temp_directory(Allocator* allocator = mem::heap()) => os::native_temp_directory(allocator);
fn Path! temp_directory(Allocator allocator = allocator::heap()) => os::native_temp_directory(allocator);
fn void! delete(Path path) => os::native_remove(path.str_view()) @inline;
macro bool is_separator(char c, PathEnv path_env = DEFAULT_PATH_ENV)
@@ -58,7 +59,7 @@ macro bool is_win32_separator(char c)
return c == '/' || c == '\\';
}
fn PathList! ls(Path dir, bool no_dirs = false, bool no_symlinks = false, String mask = "", Allocator* allocator = mem::heap())
fn PathList! ls(Path dir, bool no_dirs = false, bool no_symlinks = false, String mask = "", Allocator allocator = allocator::heap())
{
$if $defined(os::native_ls):
return os::native_ls(dir, no_dirs, no_symlinks, mask, allocator);
@@ -105,17 +106,17 @@ fn void! rmtree(Path path)
$endif
}
fn Path! new(String path, Allocator* allocator = mem::heap(), PathEnv path_env = DEFAULT_PATH_ENV)
fn Path! new(String path, Allocator allocator = allocator::heap(), PathEnv path_env = DEFAULT_PATH_ENV)
{
return { normalize(path.copy(allocator), path_env), path_env };
}
fn Path! temp_new(String path, PathEnv path_env = DEFAULT_PATH_ENV)
{
return new(path, mem::temp(), path_env);
return new(path, allocator::temp(), path_env);
}
fn Path! new_win32_wstring(WString path, Allocator* allocator = mem::heap())
fn Path! new_win32_wstring(WString path, Allocator allocator = allocator::heap())
{
@pool(allocator)
{
@@ -123,12 +124,12 @@ fn Path! new_win32_wstring(WString path, Allocator* allocator = mem::heap())
};
}
fn Path! new_windows(String path, Allocator* allocator = mem::heap())
fn Path! new_windows(String path, Allocator allocator = allocator::heap())
{
return new(path, allocator, WIN32);
}
fn Path! new_posix(String path, Allocator* allocator = mem::heap())
fn Path! new_posix(String path, Allocator allocator = allocator::heap())
{
return new(path, allocator, POSIX);
}
@@ -143,7 +144,7 @@ fn bool Path.equals(self, Path p2)
*
* @param [in] filename
**/
fn Path! Path.append(self, String filename, Allocator* allocator = mem::heap())
fn Path! Path.append(self, String filename, Allocator allocator = allocator::heap())
{
if (!self.path_string.len) return new(filename, allocator, self.env)!;
assert(!is_separator(self.path_string[^1], self.env));
@@ -158,7 +159,7 @@ fn Path! Path.append(self, String filename, Allocator* allocator = mem::heap())
};
}
fn Path! Path.tappend(self, String filename) => self.append(filename, mem::temp());
fn Path! Path.tappend(self, String filename) => self.append(filename, allocator::temp());
fn usz Path.start_of_base_name(self) @local
{
@@ -166,7 +167,19 @@ fn usz Path.start_of_base_name(self) @local
if (!path_str.len) return 0;
if (self.env == PathEnv.WIN32)
{
return path_str.rindex_of_char('\\') + 1 ?? volume_name_len(path_str, self.env)!!;
if (try index = path_str.rindex_of_char('\\'))
{
// c:\ style path, we're done!
if (path_str[0] != '\\') return index + 1;
// Handle \\server\foo
// Find the \ before "foo"
usz last_index = 2 + path_str[2..].index_of_char('\\')!!;
// If they don't match, we're done
assert(last_index <= index, "Invalid normalized, path %d vs %s in %s", last_index, index, path_str);
if (last_index != index) return index + 1;
// Otherwise just default to the volume length.
}
return volume_name_len(path_str, self.env)!!;
}
return path_str.rindex_of_char('/') + 1 ?? 0;
}
@@ -176,28 +189,39 @@ fn bool! Path.is_absolute(self)
String path_str = self.str_view();
if (!path_str.len) return false;
usz path_start = volume_name_len(path_str, self.env)!;
if (path_start > 0 && path_str[0] == '\\') return true;
return path_start < path_str.len && is_separator(path_str[path_start], self.env);
}
fn Path! Path.absolute(self, Allocator* allocator = mem::heap())
/**
* @require self.env == DEFAULT_PATH_ENV : "This method is only available on native paths"
**/
fn Path! Path.absolute(self, Allocator allocator = allocator::heap())
{
String path_str = self.str_view();
if (!path_str.len) path_str = ".";
if (!path_str.len) return PathResult.INVALID_PATH?;
if (self.is_absolute()!) return new(path_str, allocator, self.env);
if (path_str == ".")
{
String cwd = os::getcwd(mem::temp())!;
return new(cwd, allocator, self.env);
@pool(allocator)
{
String cwd = os::getcwd(allocator::temp())!;
return new(cwd, allocator, self.env);
};
}
switch (self.env)
{
case WIN32:
usz path_start = volume_name_len(path_str, self.env)!;
if (path_start > 0) return self;
case POSIX:
if (path_str[0] == PREFERRED_SEPARATOR_POSIX) return self;
}
String cwd = os::getcwd(mem::temp())!;
return Path{ cwd, self.env }.append(path_str, allocator)!;
$if DEFAULT_PATH_ENV == WIN32:
@pool(allocator)
{
const usz BUFFER_LEN = 4096;
WString buffer = (WString)mem::temp_alloc_array(Char16, BUFFER_LEN);
buffer = win32::_wfullpath(buffer, path_str.to_temp_wstring()!, BUFFER_LEN);
if (!buffer) return PathResult.INVALID_PATH?;
return { string::new_from_wstring(buffer, allocator), WIN32 };
};
$else
String cwd = os::getcwd(allocator::temp())!;
return Path { cwd, self.env }.append(path_str, allocator)!;
$endif
}
fn String Path.basename(self)
@@ -208,13 +232,21 @@ fn String Path.basename(self)
return path_str[basename_start..];
}
fn String Path.dirname(self)
{
usz basename_start = self.start_of_base_name();
String path_str = self.path_string;
if (basename_start == 0) return "";
if (basename_start == 0) return ".";
usz start = volume_name_len(path_str, self.env)!!;
if (basename_start <= start + 1) return path_str[:basename_start];
if (basename_start <= start + 1)
{
if (self.env == WIN32 && basename_start > start && path_str[0..1] == `\\`)
{
return path_str[:basename_start - 1];
}
return path_str[:basename_start];
}
return path_str[:basename_start - 1];
}
@@ -248,13 +280,20 @@ fn usz! volume_name_len(String path, PathEnv path_env) @local
while (count < len && path[count] == '\\') count++;
// Not 2 => folded paths
if (count != 2) return 0;
// Check that we have a name followed by '/'
// Check that we have a name followed by '\'
isz base_found = 0;
for (usz i = 2; i < len; i++)
{
char c = path[i];
if (is_win32_separator(c)) return i;
if (is_win32_separator(c))
{
if (base_found) return i;
base_found = i;
continue;
}
if (is_reserved_win32_path_char(c)) return PathResult.INVALID_PATH?;
}
if (base_found > 0 && base_found + 1 < len) return len;
return PathResult.INVALID_PATH?;
case 'A'..'Z':
case 'a'..'z':
@@ -281,6 +320,10 @@ fn String! normalize(String path_str, PathEnv path_env = DEFAULT_PATH_ENV)
{
if (!path_str.len) return "";
usz path_start = volume_name_len(path_str, path_env)!;
if (path_start > 0 && path_env == PathEnv.WIN32)
{
for (usz i = 0; i < path_start; i++) if (path_str[i] == '/') path_str[i] = '\\';
}
usz path_len = path_str.len;
if (path_start == path_len) return path_str;
char path_separator = path_env == PathEnv.WIN32 ? PREFERRED_SEPARATOR_WIN32 : PREFERRED_SEPARATOR_POSIX;
@@ -384,7 +427,9 @@ fn String! normalize(String path_str, PathEnv path_env = DEFAULT_PATH_ENV)
len++;
}
if (len > path_start + 1 && is_separator(path_str[len - 1], path_env)) len--;
path_str.ptr[len] = 0;
if (path_str.len > len) path_str.ptr[len] = 0;
// Empty path after normalization -> "."
if (!len) return ".";
return path_str[:len];
}
@@ -395,6 +440,7 @@ fn String Path.root_directory(self)
String path_str = self.str_view();
usz len = path_str.len;
if (!len) return "";
if (path_str == ".") return ".";
if (self.env == PathEnv.WIN32)
{
usz root_len = volume_name_len(path_str, self.env)!!;
@@ -417,11 +463,12 @@ def PathWalker = fn bool! (Path, bool is_dir, void*);
/*
* Walk the path recursively. PathWalker is run on every file and
* directory found. Return true to abort the walk.
* @require self.env == DEFAULT_PATH_ENV : "This method is only available on native paths"
*/
fn bool! Path.walk(self, PathWalker w, void* data)
{
const PATH_MAX = 512;
@stack_mem(PATH_MAX; Allocator* allocator)
@stack_mem(PATH_MAX; Allocator allocator)
{
Path abs = self.absolute(allocator)!;
PathList files = ls(abs, .allocator = allocator)!;
@@ -448,6 +495,11 @@ fn bool Path.has_suffix(self, String str)
return self.str_view().ends_with(str);
}
fn void Path.free_with_allocator(self, Allocator allocator)
{
allocator::free(allocator, self.path_string.ptr);
}
fn void Path.free(self)
{
free(self.path_string.ptr);
@@ -459,7 +511,7 @@ fn usz! Path.to_format(&self, Formatter* formatter) @dynamic
return formatter.print(self.str_view());
}
fn String Path.to_new_string(&self, Allocator* allocator = mem::heap()) @dynamic
fn String Path.to_new_string(&self, Allocator allocator = allocator::heap()) @dynamic
{
return self.str_view().copy(allocator);
}

View File

@@ -1,4 +1,5 @@
module std::io;
import std::math;
interface InStream
{
@@ -8,7 +9,7 @@ interface InStream
fn usz! available() @optional;
fn usz! read(char[] buffer);
fn char! read_byte();
fn usz! write_to(OutStream* out) @optional;
fn usz! write_to(OutStream out) @optional;
fn void! pushback_byte() @optional;
}
@@ -20,10 +21,10 @@ interface OutStream
fn void! flush() @optional;
fn usz! write(char[] bytes);
fn void! write_byte(char c);
fn usz! read_to(InStream* in) @optional;
fn usz! read_to(InStream in) @optional;
}
fn usz! available(InStream* s)
fn usz! available(InStream s)
{
if (&s.available) return s.available();
if (&s.seek)
@@ -38,19 +39,19 @@ fn usz! available(InStream* s)
macro bool @is_instream(#expr)
{
return $assignable(#expr, InStream*);
return $assignable(#expr, InStream);
}
macro bool @is_outstream(#expr)
{
return $assignable(#expr, OutStream*);
return $assignable(#expr, OutStream);
}
/**
* @param [&out] ref
* @require @is_instream(stream)
**/
macro usz! read_any(stream, any* ref)
macro usz! read_any(stream, any ref)
{
return read_all(stream, ((char*)ref)[:ref.type.sizeof]);
}
@@ -60,7 +61,7 @@ macro usz! read_any(stream, any* ref)
* @require @is_outstream(stream)
* @ensure return == ref.type.sizeof
*/
macro usz! write_any(stream, any* ref)
macro usz! write_any(stream, any ref)
{
return write_all(stream, ((char*)ref)[:ref.type.sizeof]);
}
@@ -133,7 +134,7 @@ macro void! @pushback_using_seek(&s)
s.seek(-1, CURSOR)!;
}
fn usz! copy_to(InStream* in, OutStream* dst, char[] buffer = {})
fn usz! copy_to(InStream in, OutStream dst, char[] buffer = {})
{
if (buffer.len) return copy_through_buffer(in, dst, buffer);
if (&in.write_to) return in.write_to(dst);
@@ -142,12 +143,12 @@ fn usz! copy_to(InStream* in, OutStream* dst, char[] buffer = {})
$case NORMAL:
@pool()
{
return copy_through_buffer(in, dst, mem::temp_array(char, 4096));
return copy_through_buffer(in, dst, mem::temp_alloc_array(char, 4096));
};
$case SMALL:
@pool()
{
return copy_through_buffer(in, dst, mem::temp_array(char, 1024));
return copy_through_buffer(in, dst, mem::temp_alloc_array(char, 1024));
};
$case TINY:
$case NONE:
@@ -155,7 +156,7 @@ fn usz! copy_to(InStream* in, OutStream* dst, char[] buffer = {})
$endswitch
}
macro usz! copy_through_buffer(InStream *in, OutStream* dst, char[] buffer) @local
macro usz! copy_through_buffer(InStream in, OutStream dst, char[] buffer) @local
{
usz total_copied;
while (true)

View File

@@ -2,7 +2,7 @@ module std::io;
struct ReadBuffer (InStream)
{
InStream* wrapped_stream;
InStream wrapped_stream;
char[] bytes;
usz read_idx;
usz write_idx;
@@ -14,7 +14,7 @@ struct ReadBuffer (InStream)
* @require bytes.len > 0
* @require self.bytes.len == 0 "Init may not run on already initialized data"
**/
fn ReadBuffer* ReadBuffer.init(&self, InStream* wrapped_stream, char[] bytes)
fn ReadBuffer* ReadBuffer.init(&self, InStream wrapped_stream, char[] bytes)
{
*self = { .wrapped_stream = wrapped_stream, .bytes = bytes };
return self;
@@ -63,7 +63,7 @@ fn void! ReadBuffer.refill(&self) @local @inline
struct WriteBuffer (OutStream)
{
OutStream* wrapped_stream;
OutStream wrapped_stream;
char[] bytes;
usz index;
}
@@ -74,7 +74,7 @@ struct WriteBuffer (OutStream)
* @require bytes.len > 0 "Non-empty buffer required"
* @require self.bytes.len == 0 "Init may not run on already initialized data"
**/
fn WriteBuffer* WriteBuffer.init(&self, OutStream* wrapped_stream, char[] bytes)
fn WriteBuffer* WriteBuffer.init(&self, OutStream wrapped_stream, char[] bytes)
{
*self = { .wrapped_stream = wrapped_stream, .bytes = bytes };
return self;

View File

@@ -3,7 +3,7 @@ import std::math;
struct ByteBuffer (InStream, OutStream)
{
Allocator* allocator;
Allocator allocator;
usz max_read;
char[] bytes;
usz read_idx;
@@ -16,7 +16,7 @@ struct ByteBuffer (InStream, OutStream)
* max_read defines how many bytes might be kept before its internal buffer is shrinked.
* @require self.bytes.len == 0 "Buffer already initialized."
**/
fn ByteBuffer*! ByteBuffer.init_new(&self, usz max_read, usz initial_capacity = 16, Allocator* allocator = mem::heap())
fn ByteBuffer*! ByteBuffer.new_init(&self, usz max_read, usz initial_capacity = 16, Allocator allocator = allocator::heap())
{
*self = { .allocator = allocator, .max_read = max_read };
initial_capacity = max(initial_capacity, 16);
@@ -24,9 +24,9 @@ fn ByteBuffer*! ByteBuffer.init_new(&self, usz max_read, usz initial_capacity =
return self;
}
fn ByteBuffer*! ByteBuffer.init_temp(&self, usz max_read, usz initial_capacity = 16)
fn ByteBuffer*! ByteBuffer.temp_init(&self, usz max_read, usz initial_capacity = 16)
{
return self.init_new(max_read, initial_capacity, mem::temp());
return self.new_init(max_read, initial_capacity, allocator::temp());
}
/**
@@ -41,7 +41,7 @@ fn ByteBuffer*! ByteBuffer.init_with_buffer(&self, char[] buf)
fn void ByteBuffer.free(&self)
{
if (self.allocator) self.allocator.free(self.bytes);
if (self.allocator) allocator::free(self.allocator, self.bytes);
*self = {};
}
@@ -131,7 +131,7 @@ fn usz! ByteBuffer.available(&self) @inline @dynamic
fn void! ByteBuffer.grow(&self, usz n)
{
n = math::next_power_of_2(n);
char* p = self.allocator.realloc_aligned(self.bytes, n, .alignment = char.alignof)!;
char* p = allocator::realloc_aligned(self.allocator, self.bytes, n, .alignment = char.alignof)!;
self.bytes = p[:n];
}

View File

@@ -53,7 +53,7 @@ fn usz! ByteReader.seek(&self, isz offset, Seek seek) @dynamic
return new_index;
}
fn usz! ByteReader.write_to(&self, OutStream* writer) @dynamic
fn usz! ByteReader.write_to(&self, OutStream writer) @dynamic
{
if (self.index >= self.bytes.len) return 0;
usz written = writer.write(self.bytes[self.index..])!;

View File

@@ -1,10 +1,11 @@
module std::io;
import std::math;
struct ByteWriter (OutStream)
{
char[] bytes;
usz index;
Allocator* allocator;
Allocator allocator;
}
/**
@@ -13,7 +14,7 @@ struct ByteWriter (OutStream)
* @require self.bytes.len == 0 "Init may not run on on already initialized data"
* @ensure (bool)allocator, self.index == 0
**/
fn ByteWriter* ByteWriter.init_new(&self, Allocator* allocator = mem::heap())
fn ByteWriter* ByteWriter.new_init(&self, Allocator allocator = allocator::heap())
{
*self = { .bytes = {}, .allocator = allocator };
return self;
@@ -24,9 +25,9 @@ fn ByteWriter* ByteWriter.init_new(&self, Allocator* allocator = mem::heap())
* @require self.bytes.len == 0 "Init may not run on on already initialized data"
* @ensure self.index == 0
**/
fn ByteWriter* ByteWriter.init_temp(&self)
fn ByteWriter* ByteWriter.temp_init(&self)
{
return self.init_new(mem::temp());
return self.new_init(allocator::temp()) @inline;
}
fn ByteWriter* ByteWriter.init_with_buffer(&self, char[] data)
@@ -38,7 +39,7 @@ fn ByteWriter* ByteWriter.init_with_buffer(&self, char[] data)
fn void! ByteWriter.destroy(&self) @dynamic
{
if (!self.allocator) return;
if (void* ptr = self.bytes.ptr) self.allocator.free(ptr);
if (void* ptr = self.bytes.ptr) allocator::free(self.allocator, ptr);
*self = { };
}
@@ -53,7 +54,7 @@ fn void! ByteWriter.ensure_capacity(&self, usz len) @inline
if (!self.allocator) return IoError.OUT_OF_SPACE?;
if (len < 16) len = 16;
usz new_capacity = math::next_power_of_2(len);
char* new_ptr = self.allocator.realloc_checked(self.bytes.ptr, new_capacity)!;
char* new_ptr = allocator::realloc_try(self.allocator, self.bytes.ptr, new_capacity)!;
self.bytes = new_ptr[:new_capacity];
}
@@ -75,7 +76,7 @@ fn void! ByteWriter.write_byte(&self, char c) @dynamic
* @param [&inout] self
* @param reader
**/
fn usz! ByteWriter.read_from(&self, InStream* reader) @dynamic
fn usz! ByteWriter.read_from(&self, InStream reader) @dynamic
{
usz start_index = self.index;
if (&reader.available)

View File

@@ -2,7 +2,7 @@ module std::io;
struct LimitReader (InStream)
{
InStream* wrapped_stream;
InStream wrapped_stream;
usz limit;
}
@@ -10,7 +10,7 @@ struct LimitReader (InStream)
* @param [&inout] wrapped_stream "The stream to read from"
* @param limit "The max limit to read"
**/
fn LimitReader* LimitReader.init(&self, InStream* wrapped_stream, usz limit)
fn LimitReader* LimitReader.init(&self, InStream wrapped_stream, usz limit)
{
*self = { .wrapped_stream = wrapped_stream, .limit = limit };
return self;

View File

@@ -2,7 +2,7 @@ module std::io;
struct Scanner (InStream)
{
InStream* wrapped_stream;
InStream wrapped_stream;
char[] buf;
usz pattern_idx;
usz read_idx;
@@ -16,7 +16,7 @@ struct Scanner (InStream)
* @param [&in] stream "The stream to read data from."
* @require buffer.len > 0 "Non-empty buffer required."
**/
fn void Scanner.init(&self, InStream* stream, char[] buffer)
fn void Scanner.init(&self, InStream stream, char[] buffer)
{
*self = { .wrapped_stream = stream, .buf = buffer };
}

View File

@@ -92,13 +92,13 @@ fault MatrixError
def Complexf = Complex(<float>);
def Complex = Complex(<double>);
def complexf_identity = complex::identity(<float>);
def complex_identity = complex::identity(<double>);
def COMPLEX_IDENTITY = complex::IDENTITY(<double>);
def COMPLEXF_IDENTITY = complex::IDENTITY(<float>);
def Quaternionf = Quaternion(<float>);
def Quaternion = Quaternion(<double>);
def quaternionf_identity = quaternion::identity(<float>);
def quaternion_identity = quaternion::identity(<double>);
def QUATERNION_IDENTITY = quaternion::IDENTITY(<double>);
def QUATERNIONF_IDENTITY = quaternion::IDENTITY(<float>);
def Matrix2f = Matrix2x2(<float>);
def Matrix2 = Matrix2x2(<double>);
@@ -118,6 +118,14 @@ def MATRIX3F_IDENTITY = matrix::IDENTITY3(<float>);
def MATRIX4_IDENTITY = matrix::IDENTITY4(<double>);
def MATRIX4F_IDENTITY = matrix::IDENTITY4(<float>);
/**
* @require types::is_numerical($typeof(x)) `The input must be a numerical value or numerical vector`
**/
macro deg_to_rad(x) {
return x * PI / 180;
}
/**
* @require types::is_numerical($typeof(x)) `The input must be a numerical value or numerical vector`
**/
@@ -390,14 +398,14 @@ macro nearbyint(x) => $$nearbyint(x);
/**
* @require values::@is_promotable_to_floatlike(x) `The input must be a number or a float vector`
* @require $assignable(exp, $typeof(x)) || values::@is_int(exp) `The input must be an integer, castable to the type of x`
* @require $assignable(exp, $typeof(values::promote_int(x))) || values::@is_int(exp) `The input must be an integer, castable to the type of x`
**/
macro pow(x, exp)
{
$if types::is_floatlike($typeof(exp)):
return $$pow(x, ($typeof(x))exp);
return $$pow(values::promote_int(x), ($typeof(values::promote_int(x)))exp);
$else
return $$pow_int(x, exp);
return $$pow_int(values::promote_int(x), exp);
$endif
}
@@ -439,6 +447,16 @@ macro rint(x) => $$rint(x);
**/
macro round(x) => $$round(x);
/**
* @require values::@is_floatlike(x) `The input must be a floating point value or float vector`
**/
macro round_to_decimals(x, int decimal_places)
{
var div = $$pow_int(($typeof(x))10, decimal_places);
return round(div * x) / div;
}
/**
* @require values::@is_floatlike(x) `The input must be a floating point value or float vector`
**/

View File

@@ -10,7 +10,7 @@ union Complex
}
macro Complex identity() => { 1, 0 };
const Complex IDENTITY = { 1, 0 };
macro Complex Complex.add(self, Complex b) => Complex { .v = self.v + b.v };
macro Complex Complex.add_each(self, Real b) => Complex { .v = self.v + b };
macro Complex Complex.sub(self, Complex b) => Complex { .v = self.v - b.v };

View File

@@ -401,13 +401,13 @@ fn Matrix4x4 ortho(Real left, Real right, Real top, Real bottom, Real near, Real
fn Matrix4x4 perspective(Real fov, Real aspect_ratio, Real near, Real far)
{
Real f = (Real)math::tan(math::PI * 0.5 - 0.5 * fov);
Real rangeInv = (Real)1.0 / (near - far);
Real range_inv = (Real)1.0 / (near - far);
return {
f / aspect_ratio, 0, 0, 0,
0, f, 0, 0,
0, 0, (near + far) * rangeInv, -1,
0, 0, near * far * rangeInv * 2, 0,
0, 0, (near + far) * range_inv, near * far * range_inv * 2,
0, 0, -1, 0,
};
}

View File

@@ -9,7 +9,8 @@ union Quaternion
Real[<4>] v;
}
macro Quaternion identity() => { 0, 0, 0, 1 };
const Quaternion IDENTITY = { 0, 0, 0, 1 };
macro Quaternion Quaternion.add(Quaternion a, Quaternion b) => Quaternion { .v = a.v + b.v };
macro Quaternion Quaternion.add_each(Quaternion a, Real b) => Quaternion { .v = a.v + b };
macro Quaternion Quaternion.sub(Quaternion a, Quaternion b) => Quaternion { .v = a.v - b.v };
@@ -69,30 +70,16 @@ fn Quaternion Quaternion.mul(a, Quaternion b)
macro into_matrix(Quaternion* q, $Type) @private
{
$Type result = { .m = { [0] = 1, [5] = 1, [10] = 1, [15] = 1 } };
Quaternion norm = q.normalize();
Quaternion rotation = q.normalize();
var x = rotation.i;
var y = rotation.j;
var z = rotation.k;
var w = rotation.l;
var ii = norm.i*norm.i;
var jj = norm.j*norm.j;
var kk = norm.k*norm.k;
var ik = norm.i*norm.k;
var ij = norm.i*q.j;
var jk = norm.j*norm.k;
var li = norm.l*norm.i;
var lj = norm.l*norm.j;
var lk = norm.l*norm.k;
result.m00 = 1 - 2*(jj + kk);
result.m01 = 2*(ik + lk);
result.m02 = 2*(ij - lj);
result.m10 = 2*(ij - lk);
result.m11 = 1 - 2*(ii + kk);
result.m12 = 2*(jk + li);
result.m20 = 2*(ik + lj);
result.m21 = 2*(jk - li);
result.m22 = 1 - 2*(ii + jj);
return result;
return $Type {
1 - 2*y*y - 2*z*z, 2*x*y - 2*z*w, 2*x*z + 2*y*w, 0,
2*x*y + 2*z*w, 1 - 2*x*x - 2*z*z, 2*y*z - 2*x*w, 0,
2*x*z - 2*y*w, 2*y*z + 2*x*w , 1 - 2*x*x - 2*y*y, 0,
0.0, 0.0, 0.0, 1.0,
};
}

View File

@@ -48,7 +48,7 @@ fn int rand(int max) @builtin
/**
* @require is_random(random)
**/
macro void next_bool(random)
macro bool next_bool(random)
{
return random.next_byte() & 1;
}
@@ -94,4 +94,4 @@ macro @random_value_to_bytes(#function, char[] bytes)
bytes = bytes[$byte_size..];
}
unreachable();
}
}

View File

@@ -203,10 +203,10 @@ macro matrix_look_at($Type, eye, target, up) @private
var vy = vz.cross(vx);
return $Type {
vx[0], vy[0], vz[0], 0,
vx[1], vy[1], vz[1], 0,
vx[2], vy[2], vz[2], 0,
- vx.dot(eye), - vy.dot(eye), - vz.dot(eye), 1
vx[0], vx[1], vx[2], - vx.dot(eye),
vy[0], vy[1], vy[2], - vy.dot(eye),
vz[0], vz[1], vz[2], - vz.dot(eye),
0.0, 0.0, 0.0, 1
};
}

View File

@@ -1,5 +1,5 @@
module std::math::random;
import std::hash::fnv32a;
import std::hash::fnv32a, std::time;
const ODD_PHI64 @local = 0x9e3779b97f4a7c15;
const MUL_MCG64 @local = 0xf1357aea2e62a9c5;
@@ -26,7 +26,7 @@ fn void seeder(char[] input, char[] out_buffer)
usz out_chars = out_buffer.len;
@pool()
{
ulong[] words = mem::temp_array(ulong, (out_chars + 7) / 8);
ulong[] words = mem::temp_alloc_array(ulong, (out_chars + 7) / 8);
words[..] = ODD_PHI64;
usz words_len_2 = words.len * 2;
@@ -85,8 +85,8 @@ fn char[8 * 4] entropy()
hash(&entropy),
random_int,
hash(clock::now()),
hash(&DString.init_new),
hash(mem::heap())
hash(&DString.new_init),
hash(allocator::heap())
};
return bitcast(entropy_data, char[8 * 4]);
}

View File

@@ -4,9 +4,9 @@ import std::ascii;
enum IpProtocol : char (AIFamily ai_family)
{
UNSPECIFIED (os::AF_UNSPEC),
IPV4 (os::AF_INET),
IPV6 (os::AF_INET6),
UNSPECIFIED = os::AF_UNSPEC,
IPV4 = os::AF_INET,
IPV6 = os::AF_INET6,
}
struct InetAddress (Printable)
@@ -56,7 +56,7 @@ fn usz! InetAddress.to_format(InetAddress* addr, Formatter* formatter) @dynamic
return formatter.printf("%d.%d.%d.%d", addr.ipv4.a, addr.ipv4.b, addr.ipv4.c, addr.ipv4.d)!;
}
fn String InetAddress.to_new_string(InetAddress* addr, Allocator* allocator = mem::heap()) @dynamic
fn String InetAddress.to_new_string(InetAddress* addr, Allocator allocator = allocator::heap()) @dynamic
{
if (addr.is_ipv6)
{

View File

@@ -1,4 +1,5 @@
module std::net;
import std::io;
fault NetError
{
@@ -56,7 +57,7 @@ fn uint! ipv4toint(String s)
return out;
}
fn String! int_to_new_ipv4(uint val, Allocator* allocator = mem::heap())
fn String! int_to_new_ipv4(uint val, Allocator allocator = allocator::heap())
{
char[3 * 4 + 3 + 1] buffer;
String res = (String)io::bprintf(&buffer, "%d.%d.%d.%d", val >> 24, (val >> 16) & 0xFF, (val >> 8) & 0xFF, val & 0xFF)!;
@@ -65,5 +66,5 @@ fn String! int_to_new_ipv4(uint val, Allocator* allocator = mem::heap())
fn String! int_to_temp_ipv4(uint val)
{
return int_to_new_ipv4(val, mem::temp());
return int_to_new_ipv4(val, allocator::temp());
}

View File

@@ -1,5 +1,5 @@
module std::net::os @if(env::POSIX && SUPPORTS_INET);
import libc;
import std::io, libc;
const int F_GETFL = 3;
const int F_SETFL = 4;

View File

@@ -1,6 +1,5 @@
module std::net::os @if(env::WIN32);
import std::os::win32;
import libc;
import std::os, std::io, libc;
const AIFamily PLATFORM_AF_IPX = 6;
const AIFamily PLATFORM_AF_APPLETALK = 16;

View File

@@ -1,6 +1,5 @@
module std::net @if(os::SUPPORTS_INET);
import std::io;
import libc;
import std::io, std::os, std::time, libc;
struct Socket (InStream, OutStream)
{
@@ -84,12 +83,12 @@ macro Socket new_socket(fd, ai)
enum SocketOption : char (CInt value)
{
REUSEADDR (os::SO_REUSEADDR),
REUSEPORT (os::SO_REUSEPORT) @if(!env::WIN32),
KEEPALIVE (os::SO_KEEPALIVE),
BROADCAST (os::SO_BROADCAST),
OOBINLINE (os::SO_OOBINLINE),
DONTROUTE (os::SO_DONTROUTE),
REUSEADDR = os::SO_REUSEADDR,
REUSEPORT @if(!env::WIN32) = os::SO_REUSEPORT,
KEEPALIVE = os::SO_KEEPALIVE,
BROADCAST = os::SO_BROADCAST,
OOBINLINE = os::SO_OOBINLINE,
DONTROUTE = os::SO_DONTROUTE,
}
fn bool! Socket.get_broadcast(&self) => self.get_option(BROADCAST);

View File

@@ -1,5 +1,5 @@
module std::net @if(os::SUPPORTS_INET);
import libc;
import std::time, libc, std::os;
macro apply_sockoptions(sockfd, options) @private
{

View File

@@ -1,6 +1,6 @@
module std::net::tcp @if(os::SUPPORTS_INET);
import std::net @public;
import libc;
import std::time, libc;
distinct TcpSocket = inline Socket;
distinct TcpServerSocket = inline Socket;

View File

@@ -1,5 +1,5 @@
module std::os::backtrace;
import std::collections::list;
import std::collections::list, std::os, std::io;
fault BacktraceFault
{
@@ -10,7 +10,7 @@ fault BacktraceFault
RESOLUTION_FAILED,
}
const Backtrace BACKTRACE_UNKNOWN = { 0, "", "", "", 0, null };
const Backtrace BACKTRACE_UNKNOWN = { 0, "", "", "", 0, null, false };
struct Backtrace (Printable)
{
@@ -19,7 +19,8 @@ struct Backtrace (Printable)
String object_file;
String file;
uint line;
Allocator* allocator;
Allocator allocator;
bool is_inline;
}
@@ -35,25 +36,26 @@ fn bool Backtrace.is_unknown(&self)
fn usz! Backtrace.to_format(&self, Formatter* formatter) @dynamic
{
String inline_suffix = self.is_inline ? " [inline]" : "";
if (self.has_file())
{
return formatter.printf("%s (in %s) (%s:%d)", self.function, self.object_file, self.file, self.line);
return formatter.printf("%s (in %s) (%s:%d)%s", self.function, self.object_file, self.file, self.line, inline_suffix);
}
if (self.is_unknown())
{
return formatter.printf("??? (in unknown)");
return formatter.printf("??? (in unknown)%s", inline_suffix);
}
return formatter.printf("%s (in %s) (source unavailable)", self.function, self.object_file);
return formatter.printf("%s (in %s) (source unavailable)%s", self.function, self.object_file, inline_suffix);
}
fn void Backtrace.free(&self)
{
if (!self.allocator) return;
self.allocator.free(self.function);
self.allocator.free(self.object_file);
self.allocator.free(self.file);
allocator::free(self.allocator, self.function);
allocator::free(self.allocator, self.object_file);
allocator::free(self.allocator, self.file);
}
fn Backtrace* Backtrace.init(&self, uptr offset, String function, String object_file, String file = "", uint line = 0, Allocator* allocator)
fn Backtrace* Backtrace.init(&self, uptr offset, String function, String object_file, String file = "", uint line = 0, Allocator allocator)
{
if (!allocator)
{
@@ -95,7 +97,7 @@ def symbolize_backtrace = linux::symbolize_backtrace @if(env::LINUX);
def symbolize_backtrace = win32::symbolize_backtrace @if(env::WIN32);
def symbolize_backtrace = darwin::symbolize_backtrace @if(env::DARWIN);
fn BacktraceList! symbolize_backtrace(void*[] backtrace, Allocator* allocator) @if(!env::NATIVE_STACKTRACE)
fn BacktraceList! symbolize_backtrace(void*[] backtrace, Allocator allocator) @if(!env::NATIVE_STACKTRACE)
{
return {};
}

View File

@@ -2,14 +2,14 @@
// Use of this source code is governed by the MIT license
// a copy of which can be found in the LICENSE_STDLIB file.
module std::os::env;
import libc;
import std::io::path, libc, std::os;
/**
* @param [in] name
* @require name.len > 0
* @return! SearchResult.MISSING
**/
fn String! get_var(String name, Allocator* allocator = mem::heap())
fn String! get_var(String name, Allocator allocator = allocator::heap())
{
@pool(allocator)
{
@@ -38,7 +38,7 @@ fn String! get_var(String name, Allocator* allocator = mem::heap())
fn String! get_var_temp(String name)
{
return get_var(name, mem::temp());
return get_var(name, allocator::temp());
}
/**
@@ -72,7 +72,7 @@ fn bool set_var(String name, String value, bool overwrite = true)
/**
* Returns the current user's home directory.
**/
fn String! get_home_dir(Allocator* using = mem::heap())
fn String! get_home_dir(Allocator using = allocator::heap())
{
String home;
$if !env::WIN32:
@@ -86,7 +86,7 @@ fn String! get_home_dir(Allocator* using = mem::heap())
/**
* Returns the current user's config directory.
**/
fn Path! get_config_dir(Allocator* allocator = mem::heap())
fn Path! get_config_dir(Allocator allocator = allocator::heap())
{
@pool(allocator)
{
@@ -126,7 +126,7 @@ fn bool clear_var(String name)
};
}
fn String! executable_path(Allocator *allocator = mem::heap())
fn String! executable_path(Allocator allocator = allocator::heap())
{
$if env::DARWIN:
return darwin::executable_path(allocator);

2
lib/std/os/linux/heap.c3 Normal file
View File

@@ -0,0 +1,2 @@
module std::os::linux @if(env::LINUX);
extern fn usz malloc_usable_size(void* ptr);

View File

@@ -1,8 +1,5 @@
module std::os::linux @if(env::LINUX);
import libc;
import std::os::posix;
import std::io;
import std::collections::list;
import libc, std::os, std::io, std::collections::list;
extern fn isz readlink(ZString path, char* buf, usz bufsize);
@@ -131,48 +128,37 @@ fn ulong! elf_module_image_base(String path) @local
return 0;
}
fn Backtrace! backtrace_load_from_exec(void* addr, Allocator* allocator) @local
fn void! backtrace_add_from_exec(BacktraceList* list, void* addr, Allocator allocator) @local
{
char[] buf = mem::temp_array(char, 1024);
char[] buf = mem::temp_alloc_array(char, 1024);
String exec_path = process::execute_stdout_to_buffer(buf, {"realpath", "-e", string::tformat("/proc/%d/exe", posix::getpid())})!;
String obj_name = exec_path.copy(allocator);
String addr2line = process::execute_stdout_to_buffer(buf, {"addr2line", "-p", "-i", "-C", "-f", "-e", exec_path, string::tformat("0x%x", addr)})!;
return backtrace_from_addr2line(addr, addr2line, obj_name, "???", allocator);
return backtrace_add_addr2line(list, addr, addr2line, obj_name, "???", allocator);
}
fn Backtrace! backtrace_load_from_dlinfo(void* addr, Linux_Dl_info* info, Allocator* allocator) @local
fn void! backtrace_add_from_dlinfo(BacktraceList* list, void* addr, Linux_Dl_info* info, Allocator allocator) @local
{
char[] buf = mem::temp_array(char, 1024);
char[] buf = mem::temp_alloc_array(char, 1024);
void* obj_addr = addr - (uptr)info.dli_fbase + (uptr)elf_module_image_base(info.dli_fname.str_view())!;
ZString obj_path = info.dli_fname;
String sname = info.dli_sname ? info.dli_sname.str_view() : "???";
String addr2line = process::execute_stdout_to_buffer(buf, {"addr2line", "-p", "-i", "-C", "-f", "-e", obj_path.str_view(), string::tformat("0x%x", obj_addr - 1)})!;
return backtrace_from_addr2line(addr, addr2line, info.dli_fname.str_view(), sname, allocator);
return backtrace_add_addr2line(list, addr, addr2line, info.dli_fname.str_view(), sname, allocator);
}
fn Backtrace! backtrace_from_addr2line(void* addr, String addr2line, String obj_name, String func_name, Allocator* allocator) @local
fn Backtrace! backtrace_line_parse(String string, String obj_name, String func_name, bool is_inlined, Allocator allocator)
{
String[] parts = addr2line.tsplit(" at ");
if (parts.len != 2)
{
return {
.function = func_name.copy(allocator),
.object_file = obj_name.copy(allocator),
.offset = (uptr)addr,
.file = "".copy(allocator),
.line = 0,
.allocator = allocator
};
}
String[] parts = string.trim().tsplit(" at ");
if (parts.len != 2) return SearchResult.MISSING?;
uint line = 0;
uint line = 0;
String source = "";
if (!parts[1].contains("?") && parts[1].contains(":"))
{
usz index = parts[1].rindex_of_char(':')!;
source = parts[1][:index];
line = parts[1][index + 1..].to_uint()!;
}
@@ -182,28 +168,56 @@ fn Backtrace! backtrace_from_addr2line(void* addr, String addr2line, String obj_
.file = source.copy(allocator),
.line = line,
.allocator = allocator,
.is_inline = is_inlined
};
}
fn Backtrace! backtrace_load_element(void* addr, Allocator* allocator = mem::heap()) @local
fn void! backtrace_add_addr2line(BacktraceList* list, void* addr, String addr2line, String obj_name, String func_name, Allocator allocator) @local
{
if (!addr) return backtrace::BACKTRACE_UNKNOWN;
String[] inline_parts = addr2line.tsplit("(inlined by)");
usz last = inline_parts.len - 1;
foreach (i, part : inline_parts)
{
bool is_inline = i != last;
Backtrace! trace = backtrace_line_parse(part, obj_name, func_name, is_inline, allocator);
if (catch trace)
{
list.push({
.function = func_name.copy(allocator),
.object_file = obj_name.copy(allocator),
.offset = (uptr)addr,
.file = "".copy(allocator),
.line = 0,
.allocator = allocator,
.is_inline = is_inline
});
continue;
}
list.push(trace);
}
}
fn void! backtrace_add_element(BacktraceList *list, void* addr, Allocator allocator = allocator::heap()) @local
{
if (!addr)
{
list.push(backtrace::BACKTRACE_UNKNOWN);
return;
}
@pool(allocator) {
Linux_Dl_info info;
if (dladdr(addr, &info) == 0)
{
return backtrace_load_from_exec(addr, allocator);
return backtrace_add_from_exec(list, addr, allocator);
}
return backtrace_load_from_dlinfo(addr, &info, allocator);
return backtrace_add_from_dlinfo(list, addr, &info, allocator);
};
}
fn BacktraceList! symbolize_backtrace(void*[] backtrace, Allocator* allocator)
fn BacktraceList! symbolize_backtrace(void*[] backtrace, Allocator allocator)
{
BacktraceList list;
list.init_new(backtrace.len, allocator);
list.new_init(backtrace.len, allocator);
defer catch
{
foreach (trace : list)
@@ -216,8 +230,7 @@ fn BacktraceList! symbolize_backtrace(void*[] backtrace, Allocator* allocator)
{
foreach (addr : backtrace)
{
Backtrace trace = backtrace_load_element(addr, allocator)!;
list.append(trace);
backtrace_add_element(&list, addr, allocator)!;
}
};
return list;

View File

@@ -1,18 +1,18 @@
module std::os::macos::cf @if(env::DARWIN);
module std::os::macos::cf @if(env::DARWIN) @link(env::DARWIN, "CoreFoundation.framework");
distinct CFAllocatorRef = void*;
distinct CFAllocatorContextRef = void*;
def CFOptionFlags = usz;
macro CFAllocatorRef default_allocator() => _macos_CFAllocatorGetDefault();
macro void CFAllocatorRef.dealloc(CFAllocatorRef allocator, void* ptr) => _macos_CFAllocatorDeallocate(allocator, ptr);
macro void* CFAllocatorRef.alloc(CFAllocatorRef allocator, usz size) => _macos_CFAllocatorAllocate(allocator, size, 0);
macro usz CFAllocatorRef.get_preferred_size(CFAllocatorRef allocator, usz req_size) => _macos_CFAllocatorGetPreferredSizeForSize(allocator, req_size, 0);
macro void CFAllocatorRef.set_default(CFAllocatorRef allocator) => _macos_CFAllocatorSetDefault(allocator);
macro CFAllocatorRef default_allocator() => macos_CFAllocatorGetDefault();
macro void CFAllocatorRef.dealloc(CFAllocatorRef allocator, void* ptr) => macos_CFAllocatorDeallocate(allocator, ptr);
macro void* CFAllocatorRef.alloc(CFAllocatorRef allocator, usz size) => macos_CFAllocatorAllocate(allocator, size, 0);
macro usz CFAllocatorRef.get_preferred_size(CFAllocatorRef allocator, usz req_size) => macos_CFAllocatorGetPreferredSizeForSize(allocator, req_size, 0);
macro void CFAllocatorRef.set_default(CFAllocatorRef allocator) => macos_CFAllocatorSetDefault(allocator);
extern fn CFAllocatorRef _macos_CFAllocatorCreate(CFAllocatorRef allocator, CFAllocatorContextRef context) @extern("CFAllocatorCreate");
extern fn void _macos_CFAllocatorDeallocate(CFAllocatorRef allocator, void* ptr) @extern("CFAllocatorDeallocate");
extern fn CFAllocatorRef _macos_CFAllocatorGetDefault() @extern("CFAllocatorGetDefault");
extern fn void _macos_CFAllocatorSetDefault(CFAllocatorRef allocator) @extern("CFAllocatorSetDefault");
extern fn void* _macos_CFAllocatorAllocate(CFAllocatorRef allocator, CFIndex size, CFOptionFlags hint) @extern("CFAllocatorAllocate");
extern fn CFIndex _macos_CFAllocatorGetPreferredSizeForSize(CFAllocatorRef allocator, CFIndex size, CFOptionFlags hint) @extern("CFAllocatorGetPreferredSizeForSize");
extern fn CFAllocatorRef macos_CFAllocatorCreate(CFAllocatorRef allocator, CFAllocatorContextRef context) @extern("CFAllocatorCreate") @builtin;
extern fn void macos_CFAllocatorDeallocate(CFAllocatorRef allocator, void* ptr) @extern("CFAllocatorDeallocate") @builtin;
extern fn CFAllocatorRef macos_CFAllocatorGetDefault() @extern("CFAllocatorGetDefault") @builtin;
extern fn void macos_CFAllocatorSetDefault(CFAllocatorRef allocator) @extern("CFAllocatorSetDefault") @builtin;
extern fn void* macos_CFAllocatorAllocate(CFAllocatorRef allocator, CFIndex size, CFOptionFlags hint) @extern("CFAllocatorAllocate") @builtin;
extern fn CFIndex macos_CFAllocatorGetPreferredSizeForSize(CFAllocatorRef allocator, CFIndex size, CFOptionFlags hint) @extern("CFAllocatorGetPreferredSizeForSize") @builtin;

View File

@@ -1,11 +1,11 @@
module std::os::macos::cf @if(env::DARWIN);
module std::os::macos::cf @if(env::DARWIN) @link(env::DARWIN, "CoreFoundation.framework");
distinct CFArrayRef = void*;
distinct CFArrayCallBacksRef = void*;
distinct CFMutableArrayRef = void*;
extern fn CFArrayRef _macos_CFArrayCreate(CFAllocatorRef allocator, void** values, CFIndex num_values, CFArrayCallBacksRef callBacks) @extern("CFArrayCreate");
extern fn CFArrayRef _macos_CFArrayCopy(CFAllocatorRef allocator, CFArrayRef array) @extern("CFArrayCopy");
extern fn CFIndex _macos_CFArrayGetCount(CFArrayRef array) @extern("CFArrayGetCount");
extern fn void _macos_CFArrayAppendArray(CFMutableArrayRef theArray, CFArrayRef otherArray, CFRange otherRange) @extern("CFArrayAppendArray");
extern fn CFMutableArrayRef _macos_CFArrayCreateMutable(CFAllocatorRef allocator, CFIndex capacity, CFArrayCallBacksRef callBacks) @extern("CFArrayCreateMutable");
extern fn void _macos_CFArrayAppendValue(CFMutableArrayRef theArray, void *value) @extern("CFArrayAppendValue");
extern fn CFArrayRef macos_CFArrayCreate(CFAllocatorRef allocator, void** values, CFIndex num_values, CFArrayCallBacksRef callBacks) @extern("CFArrayCreate") @builtin;
extern fn CFArrayRef macos_CFArrayCopy(CFAllocatorRef allocator, CFArrayRef array) @extern("CFArrayCopy") @builtin;
extern fn CFIndex macos_CFArrayGetCount(CFArrayRef array) @extern("CFArrayGetCount") @builtin;
extern fn void macos_CFArrayAppendArray(CFMutableArrayRef theArray, CFArrayRef otherArray, CFRange otherRange) @extern("CFArrayAppendArray") @builtin;
extern fn CFMutableArrayRef macos_CFArrayCreateMutable(CFAllocatorRef allocator, CFIndex capacity, CFArrayCallBacksRef callBacks) @extern("CFArrayCreateMutable") @builtin;
extern fn void macos_CFArrayAppendValue(CFMutableArrayRef theArray, void *value) @extern("CFArrayAppendValue") @builtin;

View File

@@ -1,4 +1,4 @@
module std::os::macos::cf @if(env::DARWIN);
module std::os::macos::cf @if(env::DARWIN) @link(env::DARWIN, "CoreFoundation.framework");
distinct CFTypeRef = void*;
def CFIndex = isz;
@@ -8,5 +8,5 @@ struct CFRange
CFIndex length;
}
extern fn CFTypeRef _macos_CFRetain(CFTypeRef cf) @extern("CFRetain");
extern fn void _macos_CFRelease(CFTypeRef cf) @extern("CFRelease");
extern fn CFTypeRef macos_CFRetain(CFTypeRef cf) @extern("CFRetain") @builtin;
extern fn void macos_CFRelease(CFTypeRef cf) @extern("CFRelease") @builtin;

View File

@@ -1,5 +1,5 @@
module std::os::darwin @if(env::DARWIN);
import std::collections::list;
import std::collections::list, std::os;
const CTL_UNSPEC = 0; /* unused */
const CTL_KERN = 1; /* "high kernel": proc, limits */
@@ -68,7 +68,7 @@ struct Darwin_segment_command_64
}
fn String! executable_path(Allocator *allocator)
fn String! executable_path(Allocator allocator)
{
char[4096] path;
uint len = path.len;
@@ -80,7 +80,7 @@ fn uptr! load_address() @local
{
Darwin_segment_command_64* cmd = darwin::getsegbyname("__TEXT");
if (!cmd) return BacktraceFault.SEGMENT_NOT_FOUND?;
String path = env::executable_path(mem::temp()) ?? BacktraceFault.EXECUTABLE_PATH_NOT_FOUND?!;
String path = env::executable_path(allocator::temp()) ?? BacktraceFault.EXECUTABLE_PATH_NOT_FOUND?!;
uint dyld_count = darwin::_dyld_image_count();
for (uint i = 0; i < dyld_count; i++)
{
@@ -93,7 +93,7 @@ fn uptr! load_address() @local
}
fn Backtrace! backtrace_load_element(String execpath, void* buffer, void* load_address, Allocator* allocator = mem::heap()) @local
fn Backtrace! backtrace_load_element(String execpath, void* buffer, void* load_address, Allocator allocator = allocator::heap()) @local
{
@pool(allocator)
{
@@ -132,11 +132,11 @@ fn Backtrace! backtrace_load_element(String execpath, void* buffer, void* load_a
};
}
fn BacktraceList! symbolize_backtrace(void*[] backtrace, Allocator* allocator)
fn BacktraceList! symbolize_backtrace(void*[] backtrace, Allocator allocator)
{
void *load_addr = (void *)load_address()!;
BacktraceList list;
list.init_new(backtrace.len, allocator);
list.new_init(backtrace.len, allocator);
defer catch
{
foreach (trace : list)
@@ -147,10 +147,10 @@ fn BacktraceList! symbolize_backtrace(void*[] backtrace, Allocator* allocator)
}
@pool(allocator)
{
String execpath = executable_path(mem::temp())!;
String execpath = executable_path(allocator::temp())!;
foreach (addr : backtrace)
{
list.append(backtrace_load_element(execpath, addr, load_addr, allocator) ?? backtrace::BACKTRACE_UNKNOWN);
list.push(backtrace_load_element(execpath, addr, load_addr, allocator) ?? backtrace::BACKTRACE_UNKNOWN);
}
};
return list;

3
lib/std/os/macos/heap.c3 Normal file
View File

@@ -0,0 +1,3 @@
module std::os::darwin @if(env::DARWIN);
extern fn usz malloc_size(void* ptr);

View File

@@ -1,4 +1,4 @@
module std::os::macos::objc @if(env::DARWIN);
module std::os::macos::objc @if(env::DARWIN) @link(env::DARWIN, "CoreFoundation.framework");
distinct Class = void*;
distinct Method = void*;
@@ -10,38 +10,36 @@ fault ObjcFailure
CLASS_NOT_FOUND
}
macro char* Class.name(Class cls) => _macos_class_getName(cls);
macro Class Class.superclass(Class cls) => _macos_class_getSuperclass(cls);
macro bool Class.responds_to(Class cls, Selector sel) => _macos_class_respondsToSelector(cls, sel);
macro Method Class.method(Class cls, Selector name) => _macos_class_getClassMethod(cls, name);
macro ZString Class.name(Class cls) => macos_class_getName(cls);
macro Class Class.superclass(Class cls) => macos_class_getSuperclass(cls);
macro bool Class.responds_to(Class cls, Selector sel) => macos_class_respondsToSelector(cls, sel);
macro Method Class.method(Class cls, Selector name) => macos_class_getClassMethod(cls, name);
macro bool Selector.equals(Selector a, Selector b) => a == b;
macro bool Class.equals(Class a, Class b) => a == b;
macro Selector selector_register(char* c) => _macos_sel_registerName(c);
macro Class! class_by_name(char* c)
macro Class! class_by_name(ZString c)
{
Class cls = _macos_objc_lookUpClass(c);
if (!cls) return ObjcFailure.CLASS_NOT_FOUND?;
return cls;
Class cls = macos_objc_lookUpClass(c);
return cls ?: ObjcFailure.CLASS_NOT_FOUND?;
}
macro Class[] class_get_list(Allocator *allocator = mem::heap())
macro Class[] class_get_list(Allocator allocator = allocator::heap())
{
int num_classes = _macos_objc_getClassList(null, 0);
int num_classes = macos_objc_getClassList(null, 0);
if (!num_classes) return {};
Class[] entries = allocator.new_array(Class, num_classes);
_macos_objc_getClassList(entries.ptr, entries.len);
macos_objc_getClassList(entries.ptr, entries.len);
return entries;
}
extern fn Class _macos_objc_getClass(char* name) @extern("objc_getClass");
extern fn int _macos_objc_getClassList(Class* buffer, int buffer_count) @extern("objc_getClassList");
extern fn char* _macos_class_getName(Class cls) @extern("class_getName");
extern fn Class _macos_class_getSuperclass(Class cls) @extern("class_getSuperclass");
extern fn Method _macos_class_getClassMethod(Class cls, Selector name) @extern("class_getClassMethod");
extern fn bool _macos_class_respondsToSelector(Class cls, Selector name) @extern("class_respondsToSelector");
extern fn Selector _macos_sel_registerName(char* str) @extern("sel_registerName");
extern fn Class _macos_objc_lookUpClass(char* name) @extern("objc_lookUpClass");
extern fn Class macos_objc_getClass(ZString name) @extern("objc_getClass") @builtin;
extern fn int macos_objc_getClassList(Class* buffer, int buffer_count) @extern("objc_getClassList") @builtin;
extern fn ZString macos_class_getName(Class cls) @extern("class_getName") @builtin;
extern fn Class macos_class_getSuperclass(Class cls) @extern("class_getSuperclass") @builtin;
extern fn Method macos_class_getClassMethod(Class cls, Selector name) @extern("class_getClassMethod") @builtin;
extern fn bool macos_class_respondsToSelector(Class cls, Selector name) @extern("class_respondsToSelector") @builtin;
extern fn Selector macos_sel_registerName(ZString str) @extern("sel_registerName") @builtin;
extern fn Class macos_objc_lookUpClass(ZString name) @extern("objc_lookUpClass") @builtin;

3
lib/std/os/posix/heap.c3 Normal file
View File

@@ -0,0 +1,3 @@
module std::os::posix @if(env::POSIX);
extern fn CInt posix_memalign(void **memptr, usz alignment, usz size);

View File

@@ -1,4 +1,5 @@
module std::os::posix @if(env::POSIX);
import std::thread;
import libc;
const PTHREAD_MUTEX_NORMAL = 0;

View File

@@ -1,6 +1,5 @@
module std::os::process @if(env::WIN32 || env::POSIX);
import std::io::file;
import libc;
import std::io, libc, std::os;
// This code is based on https://github.com/sheredom/subprocess.h
@@ -247,7 +246,7 @@ fn SubProcess! create(String[] command_line, SubProcessOptions options = {}, Str
**/
fn ZString* tcopy_command_line(String[] command_line) @local @inline @if(env::POSIX)
{
ZString* copy = mem::temp_array(ZString, command_line.len + 1);
ZString* copy = mem::temp_alloc_array(ZString, command_line.len + 1);
foreach (i, str : command_line)
{
copy[i] = str.zstr_tcopy();
@@ -260,7 +259,7 @@ const ZString[1] EMPTY_ENVIRONMENT @if(env::POSIX) = { null };
fn ZString* tcopy_env(String[] environment) @local @inline @if(env::POSIX)
{
if (!environment.len) return &EMPTY_ENVIRONMENT;
ZString* copy = mem::temp_array(ZString, environment.len + 1);
ZString* copy = mem::temp_alloc_array(ZString, environment.len + 1);
copy[environment.len] = null;
foreach (i, str : environment)
{

View File

@@ -99,6 +99,8 @@ extern fn CFile _fdopen(int fd, ZString mode);
extern fn CInt _access(ZString path, CInt mode);
extern fn CInt _waccess(WString path, CInt mode);
extern fn WString _wfullpath(WString absPath, WString relPath, usz maxLength);
/*
extern ulong _win32_GetCurrentDirectoryW(ulong, Char16* buffer) @extern("GetCurrentDirectoryW");
extern bool _win32_CreateSymbolicLinkW(WString symlink_file, WString target_file, ulong flags) @extern("CreateSymbolicLinkW");

11
lib/std/os/win32/heap.c3 Normal file
View File

@@ -0,0 +1,11 @@
module std::os::win32 @if(env::WIN32);
extern fn void* _aligned_malloc(usz size, usz alignment);
extern fn void* _aligned_realloc(void* memblock, usz size, usz alignment);
extern fn void* _aligned_recalloc(void* memblock, usz size, usz alignment);
extern fn void _aligned_free(void* memblock);
extern fn void _aligned_msize(void* memblock, usz alignment, usz offset);
extern fn void* _aligned_offset_malloc(usz size, usz alignment, usz offset);
extern fn void* _aligned_offset_realloc(void* memblock, usz size, usz alignment, usz offset);
extern fn void* _aligned_offset_recalloc(void* memblock, usz size, usz alignment, usz offset);
extern fn usz _msize(void* memblock);

View File

@@ -1,4 +1,5 @@
module std::os::win32 @if(env::WIN32);
import std::thread, std::os::backtrace;
const Win32_DWORD STARTF_USESTDHANDLES = 0x00000100;
const Win32_DWORD CREATE_NO_WINDOW = 0x08000000;
@@ -47,14 +48,14 @@ const UNDNAME_COMPLETE = 0x0000;
extern fn void initializeCriticalSection(Win32_CRITICAL_SECTION* section) @extern("InitializeCriticalSection");
extern fn void deleteCriticalSection(Win32_CRITICAL_SECTION* section) @extern("DeleteCriticalSection");
extern fn Win32_HANDLE createMutex(void*, bool, void*) @extern("CreateMutexA");
extern fn Win32_HANDLE createMutex(void*, Win32_BOOL, void*) @extern("CreateMutexA");
extern fn Win32_BOOL releaseMutex(Win32_HANDLE) @extern("ReleaseMutex");
extern fn void enterCriticalSection(Win32_CRITICAL_SECTION* section) @extern("EnterCriticalSection");
extern fn void leaveCriticalSection(Win32_CRITICAL_SECTION* section) @extern("LeaveCriticalSection");
extern fn Win32_BOOL tryEnterCriticalSection(Win32_CRITICAL_SECTION* section) @extern("TryEnterCriticalSection");
extern fn uint waitForSingleObject(Win32_HANDLE, uint milliseconds) @extern("WaitForSingleObject");
extern fn void sleep(uint ms) @extern("Sleep");
extern fn uint waitForMultipleObjects(uint count, Win32_HANDLE* handles, bool wait_all, uint ms) @extern("WaitForMultipleObjects");
extern fn uint waitForMultipleObjects(uint count, Win32_HANDLE* handles, Win32_BOOL wait_all, uint ms) @extern("WaitForMultipleObjects");
extern fn Win32_BOOL resetEvent(Win32_HANDLE event) @extern("ResetEvent");
extern fn Win32_BOOL setEvent(Win32_HANDLE handle) @extern("SetEvent");
extern fn long interlockedCompareExchange(int* dest, int exchange, int comperand) @extern("InterlockedCompareExchange");
@@ -102,6 +103,8 @@ extern fn Win32_BOOL symFromAddr(Win32_HANDLE hProcess, Win32_DWORD64 address, W
extern fn Win32_BOOL symGetLineFromAddr64(Win32_HANDLE hProcess, Win32_DWORD64 dwAddr, Win32_PDWORD pdwDisplacement, Win32_PIMAGEHLP_LINE64 line) @extern("SymGetLineFromAddr64");
extern fn Win32_WORD rtlCaptureStackBackTrace(Win32_DWORD framesToSkip, Win32_DWORD framesToCapture, Win32_PVOID *backTrace, Win32_PDWORD backTraceHash) @extern("RtlCaptureStackBackTrace");
extern fn Win32_BOOL symGetModuleInfo64(Win32_HANDLE hProcess, Win32_DWORD64 qwAddr, Win32_PIMAGEHLP_MODULE64 moduleInfo) @extern("SymGetModuleInfo64");
extern fn Win32_HANDLE getModuleHandleA(Win32_LPCSTR lpModuleName) @extern("GetModuleHandleA");
extern fn Win32_HANDLE getModuleHandleW(Win32_LPCWSTR lpModuleName) @extern("GetModuleHandleW");
fn Win32_DWORD! load_modules()
{
@@ -149,21 +152,21 @@ struct Symbol
Win32_DWORD64 displacement;
fn BacktraceList! symbolize_backtrace(void*[] backtrace, Allocator* allocator)
fn BacktraceList! symbolize_backtrace(void*[] backtrace, Allocator allocator)
{
BacktraceList list;
list.init_new(backtrace.len, allocator);
list.new_init(backtrace.len, allocator);
Win32_HANDLE process = getCurrentProcess();
symInitialize(process, null, 1);
defer symCleanup(process);
foreach (addr : backtrace)
{
list.append(resolve_backtrace(addr, process, allocator) ?? backtrace::BACKTRACE_UNKNOWN);
list.push(resolve_backtrace(addr, process, allocator) ?? backtrace::BACKTRACE_UNKNOWN);
}
return list;
}
fn Backtrace! resolve_backtrace(void* addr, Win32_HANDLE process, Allocator* allocator)
fn Backtrace! resolve_backtrace(void* addr, Win32_HANDLE process, Allocator allocator)
{
Symbol symbol;
//Win32_DWORD image_type = load_modules()!;

View File

@@ -1,6 +1,5 @@
module std::threads::os @if(env::POSIX);
import std::os::posix;
import libc;
module std::thread::os @if(env::POSIX);
import std::os::posix, std::time, libc;
struct NativeMutex
{
@@ -151,8 +150,8 @@ fn void* callback(void* arg) @private
fn void! NativeThread.create(&thread, ThreadFn thread_fn, void* arg)
{
PosixThreadData *thread_data = mem::new(PosixThreadData);
*thread_data = { .thread_fn = thread_fn, .arg = arg };
PosixThreadData *thread_data = mem::new(PosixThreadData, { .thread_fn = thread_fn, .arg = arg });
if (posix::pthread_create(thread, null, &callback, thread_data) != 0)
{
*thread = null;

View File

@@ -1,5 +1,5 @@
module std::thread::os @if(env::WIN32);
import std::os::win32;
import std::os::win32, std::time;
distinct NativeThread = inline Win32_HANDLE;
@@ -48,7 +48,7 @@ fn void! NativeMutex.init(&mtx, MutexType type)
win32::initializeCriticalSection(&(mtx.critical_section));
return;
}
if (!(mtx.handle = win32::createMutex(null, false, null))) return ThreadFault.INIT_FAILED?;
if (!(mtx.handle = win32::createMutex(null, 0, null))) return ThreadFault.INIT_FAILED?;
}
fn void! NativeMutex.destroy(&mtx)
@@ -208,7 +208,7 @@ fn void! timedwait(NativeConditionVariable* cond, NativeMutex* mtx, uint timeout
mtx.unlock()!;
uint result = win32::waitForMultipleObjects(2, &cond.events, false, timeout);
uint result = win32::waitForMultipleObjects(2, &cond.events, 0, timeout);
switch (result)
{
case win32::WAIT_TIMEOUT:

View File

@@ -1,5 +1,6 @@
module std::thread;
import std::thread::os;
import std::time;
distinct MutexType = int;

View File

@@ -1,11 +1,12 @@
module std::time::clock;
import std::time::os;
fn Clock now()
{
$if $defined(native_clock):
$if $defined(os::native_clock):
return os::native_clock();
$else
return 0;
unreachable("Clock unsupported");
$endif
}

View File

@@ -1,4 +1,5 @@
module std::time;
import std::io, std::time::os;
distinct Time = long;
distinct Duration = long;
@@ -77,7 +78,7 @@ fn Time now()
$endif
}
fn Time Time.add_seconds(time, long seconds) => time + (Time)(seconds * (long)MS);
fn Time Time.add_seconds(time, long seconds) => time + (Time)(seconds * (long)SEC);
fn Time Time.add_minutes(time, long minutes) => time + (Time)(minutes * (long)MIN);
fn Time Time.add_hours(time, long hours) => time + (Time)(hours * (long)HOUR);
fn Time Time.add_days(time, long days) => time + (Time)(days * (long)DAY);

View File

@@ -6,8 +6,6 @@
import platform
import io
import os
import sys
import json
import shutil
import hashlib
@@ -16,7 +14,6 @@ import tempfile
import argparse
import subprocess
import urllib.request
import re
from pathlib import Path
OUTPUT = Path("msvc_temp") # output folder
@@ -176,10 +173,10 @@ for pkg in msvc_packages:
sdk_packages = [
# Windows SDK libs
f"Windows SDK for Windows Store Apps Libs-x86_en-us.msi",
f"Windows SDK Desktop Libs x64-x86_en-us.msi",
"Windows SDK for Windows Store Apps Libs-x86_en-us.msi",
"Windows SDK Desktop Libs x64-x86_en-us.msi",
# CRT headers & libs
f"Universal CRT Headers Libraries and Sources-x86_en-us.msi",
"Universal CRT Headers Libraries and Sources-x86_en-us.msi",
]
with tempfile.TemporaryDirectory() as d:

View File

@@ -1,5 +1,197 @@
# C3C Release Notes
## 0.6.0 Change list
### Changes / improvements
- `@default` implementations for interfaces removed.
- `any*` => `any`, same for interfaces.
- Private / local globals now have `internal` visibility in LLVM.
- Updated enum syntax.
- 'rgba' also available for swizzling.
- The name "subarray" has been replaced by the more well known name "slice' across the codebase.
- Improved alignment handling.
- Add `--output-dir` to command line. #1155
- Allow making distinct types out of "void", "typeid", "anyfault" and faults.
- Removed `--system-linker` setting.
- "Try" expressions may not be any binary or unary expressions. So for example `try foo() + 1` is disallowed.
- Added `$$REGISTER_SIZE` for int register size.
- `assert(false)` only allowed in unused branches or in tests. Compile time failed asserts is a compile time error.
- Require expression blocks returning values to have the value used.
- Detect "unsigned >= 0" as errors.
- Improve callstack debug information #1184.
- Request jump table using @jump for switches.
- Improved inline debug information.
- Improved error messages on inlined macros.
- Introduce MSVC compatible SIMD ABI.
- `$foreach` doesn't create an implicit syntactic scope.
- Error of `@if` depends on `@if`
### Fixes
- Fixed issue in safe mode when converting enums.
- Better checking of operator methods.
- Bug when assigning an optional from an optional.
- Lambdas were not type checked thoroughly #1185.
- Fix problems using reflection on interface types #1203.
- `@param` with unnamed macro varargs could crash the compiler.
- Compiler crash using enum nameof from different module #1205.
### Stdlib changes
- "init_new/init_temp" removed.
- LinkedList API rewritten.
- List "pop" and "remove" function now return Optionals.
- RingBuffer API rewritten. Allocator interface changed.
- Deprecated Allocator, DString and mem functions removed.
- "identity" functions are now constants for Matrix and Complex numbers.
- Removed 'append' from Object and List, replaced by 'push'.
- `GenericList` renamed `AnyList`.
- Proper handling of '.' and Win32 '//server' paths.
- Path normalization - fix possible null terminator out of bounds.
## 0.5.6 Change list
### Changes / improvements
- Support `defer (catch err)`
- Added `print-input` command argument to print all files used for compilation
- Allow recursive function definitions as long as they are pointers. #1182
- Default CPU to native if less than AVX, otherwise use AVX.
- Bounds checking on length for `foo[1:2]` slicing #1191.
- Foreach uses non-wrapping add/dec.
### Fixes
- Incorrect length passed to scratch buffer printf.
- Casting to a bitstruct would be allowed even if the type was the wrong size.
- Generic modules parameterized with constants would sometimes get the wrong parameterized module name causing conversion errors #1192.
- Duplicate emit of expressions on negation would incorrectly compile negated macros.
- Casting a slice address to its pointer type should not compile #1193.
- Union is not properly zero-initialized with designated initializer #1194.
- Compile time fmod evaluates to 0 #1195.
- Assertion failed when casting (unsigned) argument to enum #1196
- Correct debug info on parameters without storage.
- Fix location on foreach debug output.
- Compiler crash on designated initializer for structs with bitstruct.
### Stdlib changes
- Add 'zstr' variants for `string::new_format` / `string::tformat`.
- Fix mutex and wait signatures for Win32.
## 0.5.5 Change list
### Changes / improvements
- Disallow multiple `_` in a row in digits, e.g. `1__000`.
- Added `@link` attribute.
- New 'linker' build option.
- "linker" project setting updated, "system-linker" removed.
### Fixes
- Struct/union members now correctly rejects members without storage size #1147.
- `math::pow` will now correctly promote integer arguments.
- Pointer difference would fail where alignment != size (structs etc) #1150
- Fixed array calculation for npot2 vectors.
- $$memcpy_inline and $$memset_inline fixed.
- `.$Type = ...` and `.$foo = ...` now works #1156.
- `int.min` incorrect behaviour #1154.
- Bitstruct cast to other bitstruct by way of underlying type would fail #1159.
- Bug in `time.add_seconds` #1162.
- Remove initial './' in Win32 and convert '/' to '\' for paths when running a binary.
- 'output' directory for projects was incorrect in templates.
- Regression: no stacktrace.
- For MacOS, running with higher optimization would crash as initializers were removed.
- `compile-run` and `run` now returns the proper return code.
- Allow String constants -> ichar*, and allow integer pointers to explicitly convert between unsigned signed.
- Bug in unaligned return value lowering for Aarch64.
### Stdlib changes
- Added `new_aligned` and `alloc_aligned` functions to prevent accidental under-alignment when allocating simd.
- Fixes to realloc of aligned allocations
- Use native Windows calls on aligned allocations on Windows.
- mem::copy_inline, mem::clear_inline and mem::set_inline added.
- mem::copy / clear / set no longer has an `$inline` attribute.
- Native aligned libc malloc on Windows & POSIX.
- Simplification of the allocator interface.
- CoreFoundation only linked on MacOS when used.
## 0.5.4 Change list
### Changes / improvements
- Hash variables may now take a designated initializer.
- Added @safemacro to override the `@` requirement for non-function-like macros.
- More information available with debug log in non debug builds.
- Removed install_win_reqs.bat which didn't work well.
- Support `**` to mean `./**`
- MacOS init/finalizer now respects priority.
- Bitstructs supports `!=` and `==`.
- Support Windows `.def` files using `--windef`.
- Bitstructs now fold compile time constant bit ops.
- Fix issue where in some cases a constant global with a string wasn't folded (e.g. in asm stmts)
- Lateral implicit imports removed.
- Default to '.' if no libdir is specified.
- Improved error messages for `--lib`.
- Added `--linker` to set the linker #1067.
### Fixes
- Fixes to macro context evaluation with macro varargs.
- Dynamic methods registered before init functions on MacOS.
- Fixed clobber on x86 `cpuid` instruction.
- Removed invalid syntax from grammar.y.
- `output` project setting now respected.
- Aliased declarations caused errors when used in initializers.
- Aliased consts used as constant initializers caused errors.
- Exported module names replace `::` by `_`.
- Const ternary would evaluate incorrectly for ?:
- `$$MODULE` would report the incorrect module name in macros.
- Fixed debug info for globals and for/switch scopes.
- `out` now correctly detects subscript[] use.
- Ambiguous recursive imports are now correctly detected.
- Overzealous local escape check corrected #1127.
- Fixes to the matrix functions #1130.
### Stdlib changes
- Deprecated `Allocator` helper functions.
- Added `mem::allocator` functions corresponding to removed allocator functions.
- Changed `mem::new` / `mem::temp_new` to accept an optional initializer, and will clear by default.
- Mem `_clear` and `_zero` variants deprecated. "new_*" functions will clear by default.
- Mem "alloc_*" functions replace old "new_*" behaviour.
- Fixed temp memory issue with formatter.
- Added temp_push and temp_pop for pushing / popping the temp allocator manually (or from C).
- Added byte_size to `List`
- Added `GenericList`.
## 0.5.3 Change list
### Changes / improvements
- Migrate from using actual type with GEP, use i8 or i8 array instead.
- Optimize foreach for single element arrays.
- Move all calls to panic due to checks to the end of the function.
### Fixes
- Single module command line option was not respected.
- Fixed issue with compile time defined types (String in this case), which would crash the compiler in certain cases.
- Projects now correctly respect optimization directives.
- Generic modules now correctly follow the implicit import rules of regular modules.
- Passing an untyped list to a macro and then using it as a vaarg would crash the compiler.
- Extern const globals now work correctly.
### Stdlib changes
- init_new/init_temp deprecated, replaced by new_init and temp_init.
## 0.5.2 Change list
### Changes / improvements
- Allow trailing comma in calls and parameters #1092.
### Fixes
- Fixes issue where single character filenames like 'a.c3' would be rejected.
- Better errors when index type doesn't match len() when doing user defined foreach.
- Fixes to `to_int` for hexadecimal strings.
- Fixed issue when using a generic type from a generic type.
- Bug with vector parameters when the size > 2 and modified.
- Missing error on assigning to in-parameters through subscripting.
- Inference of a vector on the lhs of a binary expression would cause a crash.
- Fixes to PriorityQueue
### Stdlib changes
- Allow `to_int` family functions take a base, parsing base 2-10 and 16.
## 0.5.1 Change list
### Changes / improvements
@@ -7,7 +199,7 @@
- Do not link with debug libraries unless using static libraries.
- Add 'print-linking' build option.
- System linker may be used even if the target arch is different from current.
- Slice -> array/vector works for constant slice lenghts.
- Slice -> array/vector works for constant slice lengths.
### Fixes
- On Aarch64 use the correct frame pointer type.
@@ -193,7 +385,7 @@
- Added `saturated` math.
- Added `@expect`, `@unlikely` and `@likely` macros.
- Temp allocator uses memory-env to determine starting size.
- Temp allocator is now accessed using `mem::temp()`, heap allocator using `mem::heap()`.
- Temp allocator is now accessed using `mem::temp()`, heap allocator using `allocator::heap()`.
- Float parsing added.
- Additions to std::net, ipv4/ipv6 parsing.
- Stream api.
@@ -243,13 +435,13 @@
- Fixed errors on flexible array slices.
- Fix of `readdir` issues on macOS.
- Fix to slice assignment of distinct types.
- Fix of issue casting subarrays to distinct types.
- Fix of issue casting slices to distinct types.
- Fixes to `split`, `rindex_of`.
- List no longer uses the temp allocator by default.
- Remove test global when not in test mode.
- Fix sum/product on floats.
- Fix error on void! return of macros.
- Removed too permissive casts on subarrays.
- Removed too permissive casts on slices.
- Using C files correctly places objects in the build folder.
- Fix of overaligned deref.
- Fix negating a float vector.
@@ -352,7 +544,7 @@
- Added type.inner and type.len reflection.
- Support float mod operations.
- Add float.max/min.
- Allow [in] contract to be used on subarray types.
- Allow [in] contract to be used on slices.
- Add linker and linked dir arguments to build files.
- Auto-import std::core.
- LLVM 15 support.

View File

@@ -18,7 +18,7 @@ Some short names:
| int | cond | sw/sn | always | explptr | no | expand | explbase | edist | no | no | no | no | no | edist | no |
| float | cond | expl | sw/sn | no | no | expand | no | edist | no | no | no | no | no | no | no |
| pointer | cond | explptr | no | ptrconv | arve | expand | no | edist | no | no | no | yes | expl | no | expl |
| subarray | cond | no | no | no | saconv | no | no | edist | no? | no | no | no | no | no | no |
| slice | cond | no | no | no | saconv | no | no | edist | no? | no | no | no | no | no | no |
| vec | cond | no | no | no | no | as base | no | edist | expl | no | no | no | no | no | no |
| bits | no | explbase | no | no | no | no | no? | edist | explbase | no | no | no | no | no | no |
| distc | edist | edist | edist | edist | edist | edist | edist | edist | edist | edist | edist | edist | edist | edist | edist |

View File

@@ -9,7 +9,7 @@ fault InterpretError
INTEPRET_FAILED
}
fn void! print_error(usz pos, String err)
fn void! print_error_type_at(usz pos, String err)
{
io::printfn("Error at %s: %s", pos, err);
return InterpretError.INTEPRET_FAILED?;
@@ -25,10 +25,10 @@ fn void! brainf(String program)
switch (c)
{
case '<':
if (!mem) return print_error(sp, "Memory underflow");
if (!mem) return print_error_type_at(sp, "Memory underflow");
mem--;
case '>':
if (mem == BF_MEM - 1) return print_error(sp, "Memory overflow");
if (mem == BF_MEM - 1) return print_error_type_at(sp, "Memory overflow");
mem++;
case '+':
memory[mem]++;
@@ -45,7 +45,7 @@ fn void! brainf(String program)
usz start = sp - 1;
while (indent)
{
if (sp == program.len) return print_error(start, "No matching ']'");
if (sp == program.len) return print_error_type_at(start, "No matching ']'");
switch (program[sp++])
{
case ']': indent--;
@@ -59,7 +59,7 @@ fn void! brainf(String program)
usz indent = 1;
while (indent)
{
if (!sp) return print_error(start, "No matching '['");
if (!sp) return print_error_type_at(start, "No matching '['");
switch (program[--sp])
{
case ']': indent++;

Some files were not shown because too many files have changed in this diff Show More