0
Fork 0
mirror of https://codeberg.org/ziglang/zig synced 2026-11-18 07:05:05 +00:00
Commit graph

218 commits

Author SHA1 Message Date
Brian Ferri
25cafe0b34 spirv: fix vector casts, packed field pointers and debug names of comptime-known values (#37040)
I'm writing a small numeric library whose kernels run on the CPU, on CUDA and on Vulkan through the self-hosted SPIR-V backend, and I was making them generic over the float type. `f64` needs its own `exp` and `log` on the GPU, so I wrote them as polynomial series, and the `f64` Vulkan module stopped validating. This fixes the three backend bugs underneath that, and a crash I ran into after merging master.

The series take floats apart with `@bitCast` on vectors. Between vectors of the same length this reaches the Vulkan/OpenGL fallback in `bitCast`, which stores the source into a function variable of the destination type:

```zig
export fn main() callconv(.{ .spirv_kernel = .{ .x = 1, .y = 1, .z = 1 } }) void {
    var x: @Vector(2, f32) = .{ 1, 2 };
    _ = &x;
    const bits: @Vector(2, i32) = @bitCast(x);
    _ = bits;
}
```

```
$ zig build-exe -fno-llvm -target spirv64-vulkan -mcpu vulkan_v1_2 bitcast.zig -femit-bin=bitcast.spv
$ spirv-val --target-env vulkan1.2 bitcast.spv
error: line 45: OpStore Pointer <id> '36[%36]'s type does not match Object <id> '29[%29]'s type.
```

An elementwise vector bitcast is now a scalar `bitCast` per lane, the way `buildConvert` and friends already unroll vectors. Going through the scalar path also normalizes lanes of strange integers: a single `OpBitcast` on the whole vector validates for `@Vector(n, u3)` to `@Vector(n, i3)`, but leaves lanes that don't compare equal to the negative values they should hold.

The series also convert between integer and float lanes. A vector of one lane, or of more than four without `vector16`, is an array in SPIR-V, and `airFloatFromInt` and `airIntFromFloat` emitted a single `OpConvert*` with that array as its result type:

```zig
export fn main() callconv(.{ .spirv_kernel = .{ .x = 1, .y = 1, .z = 1 } }) void {
    var x: @Vector(1, i32) = .{1};
    _ = &x;
    const y: @Vector(1, f32) = @floatFromInt(x);
    _ = y;
}
```

```
$ spirv-val --target-env vulkan1.2 convert.spv
error: line 40: Expected float scalar or vector type as Result Type: ConvertSToF
  %y = OpConvertSToF %_arr_f32_u32_1 %26
```

`@floatCast` and `@intCast` already go through `buildConvert`, which unrolls the vector; these two now do as well, with `buildConvert` picking the int/float conversion opcodes.

The coefficients sit in a comptime array that the series take as a comptime slice, and that alone fails to compile, although nothing reads the slice at runtime:

```zig
const weights = [_]f32{ 1, 2, 3 };

inline fn first(comptime w: []const f32) f32 {
    return w[0];
}

export fn main() callconv(.{ .spirv_kernel = .{ .x = 1, .y = 1, .z = 1 } }) void {
    var x: f32 = first(&weights);
    _ = &x;
}
```

```
$ zig build-exe -fno-llvm -target spirv64-vulkan -mcpu vulkan_v1_2 dbgvar.zig -femit-bin=dbgvar.spv
dbgvar.zig:7:8: error: TODO (SPIR-V): pointer to constant 'dbgvar.weights'
```

It appears `airDbgVar` lowers the operand only to attach an `OpName` to it, and a pointer to a non-extern constant is still a TODO. A comptime-known value has no runtime variable of its own to name, so `airDbgVar` now leaves interned operands alone. The same thing held back 58 of the behavior tests skipped on SPIR-V, and the test commit un-skips them.

After merging master, four of those tests crashed the compiler in `structFieldPtr`, which has `unreachable` for packed containers since ebb8683ef1 ([#36996](https://codeberg.org/ziglang/zig/pulls/36996)):

```zig
const Flags = packed struct { a: bool, b: bool };

export fn main() callconv(.{ .spirv_kernel = .{ .x = 1, .y = 1, .z = 1 } }) void {
    var flags: Flags = .{ .a = true, .b = false };
    const b = &flags.b;
    b.* = true;
}
```

```
thread 3961624 panic: reached unreachable code
src/codegen/spirv/CodeGen.zig:6906:27: 0x4767157 in structFieldPtr (main.zig)
            .@"packed" => unreachable,
```

My reading is that the `unreachable` holds for the pointer derivations it was written for, since `constantPtr` now turns packed pointers into pointers to their host integer first, but `airStructFieldPtr` and `airStructFieldPtrIndex` still pass packed structs through. A pointer to a packed field is a bit-pointer into the struct's own host integer, so it is the struct pointer: that is what the old code computed, and what the C backend now does with `.begin`. Packed unions get back the TODO error they had before. I may be missing a plan for packed field pointers here, and I'm happy to change it if so. The commit also un-skips three packed tests that crash on master the same way. I suppose nothing caught this because those tests are skipped on SPIR-V and the `spirv64-vulkan` behavior target is commented out in `test/tests.zig`.

I ran the behavior tests through `tools/unskip_spirv_tests.zig` and [zig-spirv-test-executor](https://github.com/Snektron/zig-spirv-test-executor) on a Radeon 680M, and the new and un-skipped tests on an RTX 4060 as well. Every fix has tests that fail on master and pass with it; the vector ones compare lane by lane, since `expectEqual` on vectors doesn't compile for SPIR-V yet. Running each behavior file whole, nothing that passes on master fails. Five more skips already pass on master without these changes, so I left them out, but I can add them. The new vector tests copy the aarch64, arm, sparc64 and riscv64 skips of the tests around them, which I couldn't check.

Reviewed-on: https://codeberg.org/ziglang/zig/pulls/37040
Reviewed-by: Ali Cheraghi <alichraghi@noreply.codeberg.org>
2026-10-04 11:50:49 +02:00
Ali Cheraghi
65b95b0af4 add unskip_spirv_tests.zig tool 2026-09-18 16:36:24 +03:30
Pavel Verigo
e045b77d99 stage2-wasm: linker emit obj + fixes/hacks 2026-07-28 22:42:11 +02:00
Jacob Young
3b8913a321 llvm: enable passing tests
Closes #10875
Closes #12396
Closes #20680
Closes #25734
Closes #30854
Closes #35519
2026-07-27 14:44:41 -04:00
Jacob Young
865f329e4f cbe: update for compiler_rt abi changes
* `tools/generate_c_size_and_align_checks.zig`: superceded by any cbe
   output artifact
 * `cTypePreferredAlignment`: no uses, if brought back, cbe should be
   updated to verify `__alignof`

Closes #19991
Closes #20881
Closes #35523
2026-07-27 14:35:12 -04:00
Ali Cheraghi
65e74dfbda @extern: add flat decoration 2026-06-18 13:38:58 +02:00
Ali Cheraghi
508cbec694 spirv: implement tuple types 2026-06-18 13:38:58 +02:00
Ali Cheraghi
4653794852 spirv: implement switch with ranges and loop_switch_br switch_dispatch 2026-06-18 13:38:58 +02:00
Ali Cheraghi
a5fbbb8305 spirv: composite integers
composite integers (wider than usize) are represented in as `[N]u32` arrays.
however, no operations were implemented before.
following operations are implemented:

- bitwise (and, or, xor, not)
- comparison (eq/neq)
- add/sub/mul/shl/shr
- `@abs`, `@intCast`

this removes 36 SPIR-V behavior test skips.

also fixed `derivePtr` to use `OpAccessChain` instead of `OpBitcast`
for array pointer casts (e.g. `*[100]u8` to `*u8`) in logical addressing
mode, where pointer bitcasts are not valid.
2026-06-18 13:38:58 +02:00
Alex Rønne Petersen
6aaea5af41
std.Target: add abin32 and x32 Abi tags
We only had gnu/musl variants of these previously.
2026-06-18 02:37:52 +02:00
Ali Cheraghi
9115f88386 test: skip failing/crashing behavior tests for spirv backend
tested with `-target spirv64-vulkan` and `-mcpu vulkan_v1_2+float16+variable_pointers+float64+int64`
2026-06-14 09:11:59 +02:00
Justus Klausecker
ffac200e66 AstGen: move check for primitive integer type with leading zero
Identifiers like `u0_` should be legal but were previously rejected because
they were spuriously interpreted as primitive integer type identifiers.
Now the integer is first parsed and only if that succeeds it's checked for
a leading zero.
2026-06-11 12:08:45 +02:00
Krzysztof Wolicki
9e80795623
all: update to use new std.lang.Type definitions 2026-05-27 10:03:51 +01:00
Matthew Lugg
fdac89d6cd
remove uses of array multiplication
In preparation for its removal as accepted in
https://github.com/ziglang/zig/issues/24738.
2026-04-30 08:57:51 +01:00
Matthew Lugg
106850fd4c
tests: remove uses of void{}
In preparation for this syntax to be removed from the language per
https://github.com/ziglang/zig/issues/15213.
2026-04-29 23:27:58 +01:00
Alex Rønne Petersen
1af476d5d8
behavior: re-enable a bunch of behavior tests that now pass
closes https://github.com/ziglang/zig/issues/9660
closes https://github.com/ziglang/zig/issues/21050
closes https://github.com/ziglang/zig/issues/21090
closes https://github.com/ziglang/zig/issues/21091
2026-04-25 21:54:48 +02:00
Pavel Verigo
22945fbbdc stage2-wasm: vector, std tests 2026-04-22 00:19:46 +02:00
Pavel Verigo
557caecaaa stage2-wasm: start big int support 2026-04-02 15:48:43 +02:00
Ali Cheraghi
dec1163fbb
all: replace all @Type usages
Co-authored-by: Matthew Lugg <mlugg@mlugg.co.uk>
2025-11-22 22:42:38 +00:00
Alex Rønne Petersen
9ab7eec23e represent Mac Catalyst as aarch64-maccatalyst-none rather than aarch64-ios-macabi
Apple's own headers and tbd files prefer to think of Mac Catalyst as a distinct
OS target. Earlier, when DriverKit support was added to LLVM, it was represented
a distinct OS. So why Apple decided to only represent Mac Catalyst as an ABI in
the target triple is beyond me. But this isn't the first time they've ignored
established target triple norms (see: armv7k and aarch64_32) and it probably
won't be the last.

While doing this, I also audited all Darwin OS prongs throughout the codebase
and made sure they cover all the tags.
2025-11-14 11:33:35 +01:00
Jacob Young
869ef00602 aarch64: more progress
- factor out `loadReg`
 - support all general system control registers in inline asm
 - fix asserts after iterating field offsets
 - fix typo in `slice_elem_val`
 - fix translation of argument locations
2025-07-25 14:20:23 -04:00
Jacob Young
5060ab99c9 aarch64: add new from scratch self-hosted backend 2025-07-22 19:43:47 -07:00
Andrew Kelley
34f64432b0 remove usingnamespace from the language
closes #20663
2025-07-07 13:39:48 -07:00
Jacob Young
1f98c98fff x86_64: increase passing test coverage on windows
Now that codegen has no references to linker state this is much easier.

Closes #24153
2025-06-19 18:41:12 -04:00
Ali Cheraghi
872f68c9cb
rename spirv backend name
`stage2_spirv64` -> `stage2_spirv`
2025-06-16 13:22:19 +03:30
Jacob Young
b483defc5a Legalize: implement scalarization of binary operations 2025-05-31 18:54:28 -04:00
Alex Rønne Petersen
b3f52ae339
test: Disable arrays and vectors with big integers on MIPS N32.
https://github.com/ziglang/zig/issues/23805
2025-05-08 21:32:22 +02:00
Ali Cheraghi
aec0f9b3e7
test: skip failing tests with spirv-vulkan 2025-02-24 19:39:42 +01:00
Jacob Young
c3d33440f0 x86_64: pass more behavior tests 2025-01-16 20:47:30 -05:00
Jacob Young
666d76d85c x86_64: implement load and store 2025-01-16 20:47:30 -05:00
kcbanner
2d888a8e63 tests: re-enable tests that now pass on aarch64-windows 2024-10-22 11:39:29 +02:00
Robin Voetter
86b88ea7da
spirv: skip range switch tests
This is not yet implemented
2024-10-13 01:58:11 +02:00
Meghan Denny
5e4da1ff30
std: add arch bits for s390x-linux (#21342)
see #21402
2024-09-24 13:35:12 -07:00
Alex Rønne Petersen
5cb9668632 test: Re-enable a bunch of behavior tests with LLVM.
Closes #10627.
Closes #12013.
Closes #18034.
2024-09-12 07:28:37 +02:00
mlugg
0fe3fd01dd
std: update std.builtin.Type fields to follow naming conventions
The compiler actually doesn't need any functional changes for this: Sema
does reification based on the tag indices of `std.builtin.Type` already!
So, no zig1.wasm update is necessary.

This change is necessary to disallow name clashes between fields and
decls on a type, which is a prerequisite of #9938.
2024-08-28 08:39:59 +01:00
mlugg
f47f6d766e
behavior,cases: add @branchHint test coverage 2024-08-27 00:44:35 +01:00
mlugg
6808ce27bd
compiler,lib,test,langref: migrate @setCold to @branchHint 2024-08-27 00:44:35 +01:00
mlugg
a3a737e9a6
lib,test,tools,doc: update usages of @export 2024-08-27 00:44:35 +01:00
David Rubin
2e8351cc9e elf: fix up riscv for .got.zig rewrite 2024-08-15 08:53:41 +02:00
David Rubin
8da212c11b
riscv: update tests and fix reuse bug 2024-07-26 04:19:58 -07:00
David Rubin
a1f6a8ef90
riscv: airAsm rewrite
with this rewrite we can call functions inside of
inline assembly, enabling us to use the default start.zig logic

all that's left is to implement lr/sc loops for atomically manipulating
1 and 2 byte values, after which we can use the segfault handler logic.
2024-07-26 04:19:55 -07:00
David Rubin
b533e848a2
riscv: enable passing tests 2024-07-26 04:19:17 -07:00
David Rubin
81ca3a1d59
riscv: fix logic bug in ptr_elem_ptr
I was doing duplicate work with `elemOffset` multiplying by the abi size and then the `ptr_add` `genBinOp` also multiplying.

This led to having writes happening in the wrong place.
2024-07-26 04:05:41 -07:00
David Rubin
8d30fc45c4
riscv: implement more operators
we can run `std.debug.print` now, with both run-time strings and integers!
2024-07-26 04:05:39 -07:00
Pavel Verigo
d1bd9518f9 stage2-wasm: fix big int comparison
Unexpected to be found only now
2024-07-18 17:18:17 +02:00
David Rubin
4fd8900337
riscv: rewrite "binOp"
Reorganize how the binOp and genBinOp functions work.

I've spent quite a while here reading exactly through the spec and so many
tests are enabled because of several critical issues the old design had.

There are some regressions that will take a long time to figure out individually
so I will ignore them for now, and pray they get fixed by themselves. When
we're closer to 100% passing is when I will start diving into them one-by-one.
2024-06-13 02:24:39 -07:00
David Rubin
a270c6f8c8
riscv: implement optional logic 2024-06-13 02:22:33 -07:00
David Rubin
b67995689d
riscv: add airAggregateInit for arrays 2024-06-13 02:22:04 -07:00
David Rubin
05de6c279b
riscv: std.fmt.format running
- implements `airSlice`, `airBitAnd`, `airBitOr`, `airShr`.

- got a basic design going for the `airErrorName` but for some reason it simply returns
empty bytes. will investigate further.

- only generating `.got.zig` entries when not compiling an object or shared library

- reduced the total amount of ops a mnemonic can have to 3, simplifying the logic
2024-06-13 02:20:47 -07:00
David Rubin
ffb63a05a3 riscv: finally fix bug + airAggregateInit
i just hadn't realized that I placed the `riscv_start` branch in the non-simplified
starts
2024-05-11 02:17:24 -07:00