mirrors/Nim
1
0
Fork 0
mirror of https://github.com/nim-lang/Nim.git synced 2026-04-28 02:03:59 +00:00
Code Issues Packages Projects Releases Wiki Activity
22,936 Commits 86 Branches 71 Tags
797b05eda68ecbfa3dbff29866c8d2ed9bfd39c4
Commit Graph

706 Commits

Author SHA1 Message Date
ringabout
56ed4e0bb9 fixes #23759; rework move for refc (#23764) 
fixes #23759
 2024-06-29 10:43:41 +02:00
ringabout
948bb38335 ref #20653; fixes chronos empty case branches (#23706) 
ref #20653

```nim
  Error* = object
    case kind*: ErrorType
    of ErrorA:
      discard
    of ErrorB:
      discard
```
For an object variants without fields, it shouldn't generate empty
brackets for default values since there are no fields at all in case
branches.
 2024-06-14 15:55:08 +02:00
ringabout
309f97af4c fixes #23627; Simple destructor code gives invalid C (#23631) 
fixes #23627

```nim
type
  TestObj = object of RootObj

  TestTestObj = object of RootObj
    testo: TestObj

proc `=destroy`(x: TestTestObj) =
  echo "Destructor for TestTestObj"

proc testCaseT() =
  echo "\nTest Case T"
  let tt1 {.used.} = TestTestObj(testo: TestObj())
```

When generating const object fields, it's likely that
we need to generate type infos for the object, which may be an object
with
custom hooks. We need to generate potential consts in the hooks first.

https://github.com/nim-lang/Nim/pull/20433 changed the semantics of
initialization. It should evaluate`BracedInit` first.
 2024-05-21 14:53:08 +02:00
ringabout
b87732b5f1 fixes #16671; openarray conversion for object construction (#23618) 
fixes #16671

related to https://github.com/nim-lang/Nim/pull/18911
 2024-05-16 23:27:08 +02:00
ringabout
e662043fd1 rework wasMoved, move on the JS backend (#23577) 
`reset`, `wasMoved` and `move` doesn't support primitive types, which
generate `null` for these types. It is now produce `x = default(...)` in
the backend. Ideally it should be done by ast2ir in the future
 2024-05-08 09:11:46 -06:00
ringabout
72d0ba2df5 remove unused magics: mIntToStr, mInt64ToStr, mFloatToStr (#23486) 
mIntToStr, mInt64ToStr, mFloatToStr,
 2024-04-09 14:39:14 +02:00
ringabout
4b6a9e4add fixes #23422; card regression (#23437) 
fixes #23422

ref https://github.com/nim-lang/Nim/issues/20997
https://github.com/nim-lang/Nim/pull/21165

The function `cardSet` is used for large sets that are stored in the
form of arrays. It shouldn't be passed as a pointer
 2024-03-28 11:04:47 +01:00
ringabout
7cd3d60683 fixes #12703; nim cpp rejects valid code would lose const qualifier for cstring to string via cstrToNimstr (#23371) 
fixes #12703
ref #19588
 2024-03-05 18:06:58 +01:00
ringabout
4b67cccf50 fixes regression #23280; Operations on inline toOpenArray len return a wrong result (#23285) 
fixes #23280
 2024-02-06 06:24:02 +01:00
ringabout
d44b0b1869 fixes #22597; avoid side effects for call returning openArray types (#23257) 
fixes #22597

```nim
proc autoToOpenArray*[T](s: Slice[T]): openArray[T] =
  echo "here twice"
  result = toOpenArray(s.p, s.first, s.last)
```
For functions returning openarray types, `fixupCall` creates a temporary
variable to store the return value: `let tmp = autoToOpenArray()`. But
`genOpenArrayConv` cannot handle openarray assignements with side
effects. It should have stored the right part of the assignment first
instead of calling the right part twice.
 2024-01-26 06:06:08 +01:00
ringabout
29ac3c9986 fixes #22923; fixes =dup issues (#23182) 
fixes #22923
 2024-01-11 11:23:42 +01:00
Andreas Rumpf
91ad6a740b type refactor: part 4 (#23077) 2023-12-15 10:20:57 +01:00
ringabout
cca5684a17 fixes yet another strictdefs bug (#23069) 2023-12-15 08:13:25 +01:00
Andreas Rumpf
e51e98997b type refactoring: part 2 (#23059) 2023-12-13 10:29:58 +01:00
Andreas Rumpf
db603237c6 Types: Refactorings; step 1 (#23055) 2023-12-12 16:54:50 +01:00
Juan M Gómez
ca577dbab1 C++: ptr fields now pulls the whole type if it's a member in nkDotExpr (#22855) 2023-10-23 08:59:14 +02:00
Andreas Rumpf
8990626ca9 NIR: progress (#22817) 
Done:

- [x] Implement conversions to openArray/varargs.
- [x] Implement index/range checking.
 2023-10-12 23:33:38 +02:00
Andreas Rumpf
816589b667 NIR: Nim intermediate representation (#22777) 
Theoretical Benefits / Plans: 

- Typed assembler-like language.
- Allows for a CPS transformation.
- Can replace the existing C backend by a new C backend.
- Can replace the VM.
- Can do more effective "not nil" checking and static array bounds
checking.
- Can be used instead of the DFA.
- Easily translatable to LLVM.
- Reasonably easy to produce native code from.
- Tiny memory consumption. No pointers, no cry.

**In very early stages of development.**

Todo:
- [x] Map Nim types to IR types.
- [ ] Map Nim AST to IR instructions:
  - [x] Map bitsets to bitops.
  - [ ] Implement string cases.
  - [ ] Implement range and index checks.
  - [x] Implement `default(T)` builtin.
  - [x] Implement multi string concat.
- [ ] Write some analysis passes.
- [ ] Write a backend.
- [x] Integrate into the compilation pipeline.
 2023-10-11 17:44:14 +02:00
Andreas Rumpf
8f5b90f886 produce better code for object constructions and 'result' [backport] (#22668) 2023-09-11 18:48:20 +02:00
Juan M Gómez
8032f252b2 fixes #22669 constructor pragma doesnt init Nim default fields (#22670) 
fixes #22669 constructor pragma doesnt init Nim default fields

---------

Co-authored-by: Andreas Rumpf <rumpf_a@web.de>
 2023-09-10 12:45:36 +02:00
Amjad Ben Hedhili
8853fb0775 Make newSeqOfCap not initialize memory. (#21842) 
It's used in `newSeqUninitialized`.

---------

Co-authored-by: ringabout <43030857+ringabout@users.noreply.github.com>
 2023-09-09 21:11:45 +02:00
ringabout
5bd1afc3f9 fixes #17197; fixes #22560; fixes the dest of newSeqOfCap in refc (#22594) 2023-08-31 19:04:32 +08:00
Juan M Gómez
d7634c1bd4 fixes an issue where sometimes wasMoved produced bad codegen for cpp (#22587) 2023-08-30 07:22:36 +02:00
ringabout
469c9cfab4 unpublic the sons field of PType; the precursor to PType refactorings (#22446) 
* unpublic the sons field of PType

* tiny fixes

* fixes an omittance

* fixes IC

* fixes
 2023-08-11 22:18:24 +08:00
ringabout
faf1c91e6a fixes move sideeffects issues [backport] (#22439) 
* fixes move sideeffects issues [backport]

* fix openarray

* fixes openarray
 2023-08-10 18:04:29 +02:00
ringabout
8523b543d6 getTemp and friends now return TLoc as requested (#22440) 
getTemp and friends now return `TLoc`
 2023-08-10 14:17:15 +02:00
ringabout
a6610745d8 initLocExpr and friends now return TLoc (#22434) 
`initLocExpr` and friends now return TLoc
 2023-08-10 07:57:34 +02:00
ringabout
91c3221855 simplify isAtom condition (#22430) 2023-08-09 20:57:13 +02:00
ringabout
10a6e4c236 clean up gc:arc or gc:orc in docs and in error messages (#22408) 
* clean up gc:arc/orc in docs

* in error messages
 2023-08-08 05:55:18 -04:00
ringabout
260b4236fc use out parameters for getTemp (#22399) 2023-08-07 10:11:59 +02:00
ringabout
93ced31353 use strictdefs for compiler (#22365) 
* wip; use strictdefs for compiler

* checkpoint

* complete the chores

* more fixes

* first phase cleanup

* Update compiler/bitsets.nim

* cleanup
 2023-08-06 14:26:21 +02:00
Bung
6b913b4741 Revert "fix #22173 sink paramers not moved into closure (refc) (#22… (#22376) 
Revert "fix #22173 `sink` paramers not moved into closure (refc) (#22359)"

This reverts commit b40da812f7.
 2023-08-03 19:56:05 +02:00
Bung
b40da812f7 fix #22173 sink paramers not moved into closure (refc) (#22359) 
* use genRefAssign when assign to sink string

* add test case
 2023-08-02 14:08:51 +02:00
ringabout
f0f3904ff0 implement ensureMove (#22339) 
* implement `ensureMove`

* use an additional flag

* improve some logics

* progress: fixes discard ensureMove

* forbids nested expressions

* improve error messages

* checkpoint

* fixes cursor

* ADD MORE TESTS

* fixes cursorinference again

* tiny cleanup

* improve error messages

* fixes docs

* implement comments add more tests

* fixes js
 2023-07-29 10:57:03 +02:00
ringabout
7d9fe106ec fixes #22268; fixes move codegen (#22288) 2023-07-19 12:57:58 +02:00
Andreas Rumpf
a15db5d60b fixes #22175 (#22229) 2023-07-06 15:15:50 +02:00
ringabout
dfa0d2569e alternative to #22219; adds a pointer wrapper for T destructor (#22225) 
* alternative to #22219; adds a pointer wrapper for T destructor

* clean up and add comments

* Update compiler/ccgtypes.nim

* tidy up

* fixes comments

* fixes cpp

* fixes cpp
 2023-07-06 11:14:42 +02:00
ringabout
d139d99946 fixes #19101; zero initialization union casts (#22185) 
* zero initialization union casts

* cleans up and adds a test case for #19101

* uses nimZeroMem
 2023-06-29 10:51:18 +02:00
Andreas Rumpf
e8d0f1c3ae fixes #16331; aliasing of tuple construction within a single assignme… (#22113) 
* fixes #16331; aliasing of tuple construction within a single assignment, great coding style

* added test case
 2023-06-16 20:59:59 +02:00
ringabout
edb64bcff4 fixes explicit deref block (#22093) 
fixes explicit deref
 2023-06-15 09:56:58 +02:00
ringabout
64b27edd3a make move use =wasMoved internally (#22032) 
* make `move` use `=wasMoved` internally

* fixes tests

* fixes spawn finally

* fixes views

* rename to internalMove

* add a test case
 2023-06-09 15:53:12 +02:00
ringabout
1133f20fe2 lift the =dup hook (#21903) 
* fixes tests again
* remove helper functions
* fixes closures, owned refs
* final cleanup
 2023-06-02 16:03:32 +02:00
ringabout
108410ac34 fixes fieldDefect loses enum type info in ORC; consistent with VM and refc (#21954) 
fixes fieldDefect loses enum type info in ORC
 2023-05-29 14:59:59 +02:00
heterodoxic
af3fd5a010 fixes #15428 by updating deep open array copy codegen (#21935) 
* fix #15428

* add test
 2023-05-27 15:27:42 +02:00
heterodoxic
6128ef53c5 fix #10964 by honoring pointer deref syntax if a reified openarray is used to get an array's length (#21925) 
* fix #10964

* add test
 2023-05-27 06:54:41 +02:00
Juan M Gómez
44f059c75e implements allow byref to work in params #21873 (#21875) 2023-05-21 00:19:09 +02:00
Juan M Gómez
a852b2e9cf refactor gettypedesc so it accepts its own kind instead of symkind (#21867) 2023-05-19 21:31:57 +02:00
heterodoxic
cf1ab2a30b fix #21848 (#21852) 2023-05-15 21:16:49 +02:00
ringabout
71dc929ad7 bring #21802 back; fixes #21753 [backport] (#21815) 
* bring #21802 back; fixes #21753 [backport]

* adds tests and multiple fixes

* add test cases

* refactor and remove startId

* fixes custom hooks and adds tests

* handle tyUncheckedArray better
 2023-05-11 10:29:11 +02:00
ringabout
4533e894ad adds an experimental mm:atomicArc switch (#21798) 2023-05-08 16:25:47 +02:00