Mitchell Hashimoto
015cc41e50
stage2: zirReify for enums
2022-03-11 17:44:52 -08:00
Andrew Kelley
6f986298c6
Merge pull request #11125 from jmc-88/cbe
...
CBE: promote an already passing test, and add implementation for clz, ctz for integers
2022-03-11 20:15:13 -05:00
Daniele Cocca
226fcd7c70
CBE: implement clz, ctz for ints <= 128 bits
2022-03-11 23:12:15 +00:00
Mitchell Hashimoto
fca51c81bc
stage2: is_non_error always returns comptime true for empty error set
2022-03-11 17:16:34 -05:00
Mitchell Hashimoto
797f4db227
stage2: Array len field should be a usize not comptime_int
2022-03-11 17:15:36 -05:00
Daniele Cocca
6dcfbfbfb2
CBE: mark union_with_members as passing
2022-03-11 20:53:14 +00:00
Andrew Kelley
79b169c5a5
Merge pull request #11122 from mitchellh/anon-has-field
...
stage2: @hasField for anon structs
2022-03-11 15:01:51 -05:00
Andrew Kelley
a2517117e7
Sema: make @hasField support tuples too
2022-03-11 12:59:20 -07:00
Mitchell Hashimoto
6a9c9afbae
stage2: @hasField for anon structs
2022-03-11 12:51:21 -07:00
Andrew Kelley
86a98b172b
std.os: disable failing fnctl file locking test
...
See #11074
2022-03-11 12:49:53 -07:00
Andrew Kelley
4c1cc4d8d9
Merge pull request #11120 from Vexu/stage2
...
Stage2: make std.rand tests pass
2022-03-11 13:48:28 -05:00
Joachim Schmidt
5fbae9cd6f
Merge pull request #11121 from joachimschmidt557/stage2-arm
...
stage2 ARM: implement caller-preserved registers
2022-03-11 17:43:36 +01:00
Veikka Tuominen
03b8206f27
Sema: make check for whether call should be memoized more thorough
2022-03-11 16:56:33 +02:00
joachimschmidt557
4590e980f7
stage2 ARM: implement caller-saved registers
2022-03-11 14:12:11 +01:00
joachimschmidt557
06058ed6f3
stage2 regalloc: replace Register.allocIndex with generic indexOfReg
...
* callee_preserved_regs and other ABI-specific information have been
moved to the respective abi.zig files
2022-03-11 13:29:16 +01:00
Veikka Tuominen
01cd4119b0
Sema: implement @shuffle at comptime and for differing lengths
2022-03-11 14:25:15 +02:00
Veikka Tuominen
cba68090a6
stage2: implement @shuffle at runtime
2022-03-11 13:12:32 +02:00
Veikka Tuominen
db42624170
Sema: enable shl and bitwise for vectors at runtime
2022-03-11 13:06:22 +02:00
Veikka Tuominen
98a5998d83
Sema: improve detection of generic parameters
2022-03-11 11:25:45 +02:00
Andrew Kelley
078037ab9b
stage2: passing threadlocal tests for x86_64-linux
...
* use the real start code for LLVM backend with x86_64-linux
- there is still a check for zig_backend after initializing the TLS
area to skip some stuff.
* introduce new AIR instructions and implement them for the LLVM
backend. They are the same as `call` except with a modifier.
- call_always_tail
- call_never_tail
- call_never_inline
* LLVM backend calls hasRuntimeBitsIgnoringComptime in more places to
avoid unnecessarily depending on comptimeOnly being resolved for some
types.
* LLVM backend: remove duplicate code for setting linkage and value
name. The canonical place for this is in `updateDeclExports`.
* LLVM backend: do some assembly template massaging to make `%%`
rendered as `%`. More hacks will be needed to make inline assembly
catch up with stage1.
2022-03-11 00:04:42 -07:00
Andrew Kelley
b28b3f6f7b
stage2: fix comptime element load of undef array
2022-03-10 22:57:13 -07:00
Andrew Kelley
273da9efd9
AstGen: structInitExpr and arrayInitExpr avoid crash
...
when an inferred alloc is passed as the result pointer of a block.
2022-03-10 17:52:18 -07:00
Andrew Kelley
a30d283981
AstGen: lower anon struct inits differently
...
This is a companion commit to f2a5d0bf94897554e25e889dc1c6c4c7fc6c1217.
What that one did for tuples, this one does for anonymous structs.
2022-03-10 17:52:18 -07:00
Andrew Kelley
b642fa24a6
stage2: implement integer pointer constants
2022-03-10 17:52:18 -07:00
Jakub Konka
b0dc61fae2
macos: add mach_* syscalls for process mgmt
2022-03-10 22:39:04 +01:00
Andrew Kelley
96f614d374
Merge pull request #11098 from mitchellh/error-merge-eql
...
stage2: error set type equality, error and error union value equality
2022-03-10 16:22:20 -05:00
Andrew Kelley
9f163310f2
stage2: improve Type.eql and Type.hash for error sets
...
* Reduce branching in Type.eql and Type.hash for error sets.
* `Type.eql` uses element-wise bytes comparison since it can rely on
the error sets being pre-sorted.
* Avoid unnecessarily skipping tests that are passing.
2022-03-10 14:21:37 -07:00
Mitchell Hashimoto
569870ca41
stage2: error_set_merged type equality
...
This implements type equality for error sets. This is done
through element-wise error set comparison.
Inferred error sets are always distinct types and other error sets are
always sorted. See #11022 .
2022-03-10 14:20:16 -07:00
Andrew Kelley
0b82c02945
Merge pull request #11113 from Vexu/stage2
...
stage2: if generic function evaluates to another generic function call it inline
2022-03-10 15:10:51 -05:00
Marc Tiehuis
8bab1b405f
math: fix big.int div, gcd and bitAnd edge-cases
...
Fixes #10932 .
2022-03-10 13:56:07 -05:00
Andrew Kelley
6e49ba77f3
std: add sort method to ArrayHashMap and MultiArrayList
...
This also adds `std.sort.sortContext` and
`std.sort.insertionSortContext` which are more advanced methods that
allow overriding the `swap` method. The former calls the latter for now
because reworking the main sort implementation is a big task that can be
done later without any changes to the API.
2022-03-10 13:13:17 -05:00
Veikka Tuominen
f9e4344bb5
Sema: implement zirStructInit is_ref=true union
2022-03-10 20:00:07 +02:00
Veikka Tuominen
673fafc231
stage2 llvm: implement lowerParentPtr for int_{u,i}64
2022-03-10 19:52:03 +02:00
Veikka Tuominen
e0fb0770d1
Sema: if generic function evaluates to another generic function call it inline
...
```zig
fn foo(a: anytype, b: @TypeOf(a)) void { _ = b; }
test {
// foo evaluates to `fn (type) void` and must be called inline
foo(u32, u32);
}
```
2022-03-10 13:04:55 +02:00
Veikka Tuominen
b9f521b402
Sema: add coercion from [:x]T to [*:x]T
2022-03-10 12:02:31 +02:00
Andrew Kelley
f736cde397
Sema: implement pointer to tuple to pointer to array coercion
...
This involved an LLVM backend fix for the aggregate_init instruction.
2022-03-09 18:49:37 -07:00
Meghan
017d3864de
std: fix false positive for zig.isValidId with empty string ( #11104 )
...
* std: fix false positive for `zig.isValidId` with empty string, fixes #11099
* std: add zig.isValidId tests
2022-03-09 19:38:47 -05:00
Andrew Kelley
f32a77b30d
Sema: implement pointer-to-tuple coercion to slice and struct
2022-03-09 17:33:01 -07:00
Andrew Kelley
3b6e8fa59e
Sema: fix crash with @sizeOf on unions
2022-03-09 16:02:42 -07:00
Andrew Kelley
fd85cfe154
std.mem: remove redundant namespaces in test names
...
related: #7923
2022-03-09 15:44:08 -07:00
Andrew Kelley
83bb3d1ad6
Sema: fix generic fn instantiation with anytype
...
When the anytype parameter had only one-possible-value
(e.g. `void`), it would create a mismatch in the function type and the
function call. Now the function type and the callsite both omit the
one-possible-value anytype parameter in instantiated generic functions.
2022-03-09 15:35:13 -07:00
Andrew Kelley
2aa4a32097
Merge pull request #11105 from Luukdegram/wasm-misc
...
stage2: wasm - miscellaneous improvements
2022-03-09 16:03:20 -05:00
Luuk de Gram
90f08a69aa
wasm: Enable passing behavior tests
...
This also adds some float-related instructions to MIR/Emit
2022-03-09 13:53:20 -07:00
Luuk de Gram
3ea113e008
wasm: Implement field_parent_ptr
2022-03-09 13:53:10 -07:00
Luuk de Gram
f931c0638d
wasm: Implement errunion_payload_ptr_set
...
This also fixes `ret_ptr` where the pointee type is zero-sized.
2022-03-09 13:53:10 -07:00
Luuk de Gram
d01bfa032d
wasm: Preliminary fptrunc support
...
This implements the initial fptrunc instruction. For all other floating-point truncating,
a call to compiler-rt is required. (This also updates fpext to emit the same error).
2022-03-09 13:53:10 -07:00
Luuk de Gram
684b81f366
wasm: Implement fpext
...
This implements initial support for floating-point promotion for bitsizes <= 64
2022-03-09 13:53:10 -07:00
Luuk de Gram
557f396f61
wasm: Improve switch implementation
...
- Implement switching over booleans and pointers.
- Fix sparse-detection where the lowest value was never truly set
as it started at a non-zero number and the case was > 50.
- Fix indexing the jump table by ensuring it starts indexing from 0.
2022-03-09 13:53:10 -07:00
Andrew Kelley
b936fe0a58
Merge pull request #11101 from Vexu/stage2
...
Sema: handle noreturn result in condbr_inline
2022-03-09 15:07:00 -05:00
Veikka Tuominen
01b454f851
AstGen: ensure lableld block implicitly ends in a break
2022-03-09 17:12:40 +02:00