cbfa87cbea
Merge pull request #25889 from mlugg/incremental-llvm-warn
...
compiler: warn when using -fincremental with LLVM backend
2025-11-10 21:10:31 +00:00
ca34abbcde
ci: remove TERM=dumb workaround on Forgejo Actions
...
https://codeberg.org/forgejo/forgejo/issues/9244
2025-11-10 17:02:27 +01:00
d182c7e3bc
Merge pull request #25886 from alexrp/kvx
...
beginnings of KVX target support (via CBE)
2025-11-10 16:38:23 +01:00
966809862f
Merge pull request #25839 from Rexicon226/socket-bind-fix
...
Io.net: fix compile error in `receive` and `receiveTimeout`
2025-11-10 14:38:54 +00:00
fb914a9a10
sema: print @panic message at comptime
2025-11-10 14:21:26 +00:00
d942f693c5
std.zig.AstGen: properly handle grouped_expression
...
This fixes an endless loop in the compiler.
2025-11-10 14:02:37 +00:00
ce355e0ba5
Add ML-DSA post-quantum signatures ( #25862 )
...
ML-DSA is a post-quantum signature scheme that was recently
standardized by NIST.
Keys and signatures are pretty large, not making it a drop-in
replacement for classical signature schemes.
But if you are shipping keys that may still be used in 10 years
or whenever large quantum computers able to break ECC arrive,
it that ever happens, and you don't have the ability to replace
these keys, ML-DSA is for you.
Performance is great, verification is faster than Ed25519 / ECDSA.
I tried manual vectorization, but it wasn't worth it, the compiler
does at good job at auto-vectorization already.
2025-11-10 14:11:30 +01:00
13993c4f62
compiler: warn when using -fincremental with LLVM backend
...
This configuration hasn't had much work put into it yet, so is all but
guaranteed to miscompile or crash. Since users are starting to try out
`-fincremental`, and LLVM is still the default backend in many cases,
it's worth having this warning to avoid bug reports like
https://github.com/ziglang/zig/issues/25873 .
2025-11-10 12:56:35 +00:00
6d280dc1b0
compiler: update logFn to use color
...
To match the new default implementation. In fact, I implemented this by
simply dispatching *to* the default implementation after the debug log
guard; no need to complicate things!
2025-11-10 12:55:39 +00:00
cde865e06a
llvm: set sub-arch for spirv 1.6
2025-11-10 12:54:04 +01:00
f34b5ce288
std.Target.DynamicLinker: define standard paths for alpha, microblaze, sh on netbsd and openbsd
2025-11-10 09:40:44 +01:00
255fc44036
std.Target: fix toElfMachine() for arc/arceb
2025-11-10 09:40:44 +01:00
a382b7bfc5
std.simd: suggest 1024-bit vectors for kvx
2025-11-10 09:40:44 +01:00
39b5c8ee8a
std.debug.Dwarf: add kvx ip/fp/sp register mappings
2025-11-10 09:40:44 +01:00
94b6c7ce5f
std.debug.cpu_context: add kvx context implementation
2025-11-10 09:40:44 +01:00
80e567b6bd
std.pie: add kvx support
2025-11-10 09:40:44 +01:00
8fae6630c4
std.start: add kvx support
2025-11-10 09:40:44 +01:00
f5d31cff71
zig.h: add kvx support
2025-11-10 09:40:44 +01:00
4ff6e290dd
cbe: kvx uses $-prefixed registers
2025-11-10 09:40:44 +01:00
350eaa5bac
std.builtin.assembly: add Clobbers for kvx
2025-11-10 09:40:42 +01:00
1e2c216156
compiler: kvx requires 8-byte aligned functions
2025-11-10 08:20:24 +01:00
378eec95b8
std.Target: kvx requires 32-byte aligned stacks
2025-11-10 08:20:24 +01:00
2c470d24b3
std.Target: add Arch tag and info for kvx
2025-11-10 08:20:21 +01:00
c58687225b
Merge pull request #25875 from Zirunis/patch-2
...
Update Style Guide to suggest TitleCase for Type aliases as is the reality in the standard library
2025-11-09 22:40:24 -08:00
f3309a96a7
Merge pull request #25819 from jacobly0/elfv2-emit-obj
2025-11-10 00:10:17 -05:00
e1f12124fd
Remove StringAlias because it implies the existence of a String type and is a redundant example anyway
...
Co-authored-by: Ryan Liptak <squeek502@hotmail.com >
2025-11-09 14:05:09 +01:00
bdbfc7de3f
std.zig.Zir: remove ref_start_index from enum fields of Index and OptionalIndex
...
This change removes the ref_start_index from the possible enum values of
Index and OptionalIndex. It is not really a index, but a constant that
tells the offset of static Refs, so lets move it where such constant
belongs i.e. to the Ref.
2025-11-09 10:45:14 +00:00
57889cae80
posix: reduce the number of assumptions made by dl_iterate_phdr
...
Not yet fully compatible with the new linker, but still progress.
Closes #25786
2025-11-09 03:31:26 -05:00
d33c00cad0
MappedFile: avoid allocating file space with inconsistent state
2025-11-09 03:31:26 -05:00
02a241f472
Elf2: incrementally update object relocs
2025-11-09 03:31:26 -05:00
b1d46339b5
Elf2: implement object relocs
2025-11-09 03:31:26 -05:00
80c961159b
Elf2: elide unused nodes when emitting objects
2025-11-09 03:31:26 -05:00
7bcefe5a22
Elf2: emit object files that are slightly more object-like
2025-11-09 03:31:26 -05:00
5358af7ba4
incr-check: Kill child process on error
...
Since the child process is spawned with the tmp directory as its CWD, the child process opens it without DELETE access. On error, the child process would still be alive while the tmp directory is attempting to be deleted, so it would fail with `.SHARING_VIOLATION => return error.FileBusy`.
Fixes arguably the least important part of #22510 , since it's only the directory itself that would fail to get deleted, all the files inside would get deleted just fine.
2025-11-09 07:45:31 +00:00
4dcc8a80a9
Simplified and unified sentence structure of the naming convention logic
2025-11-09 04:44:05 +01:00
07f50c0351
Update Style Guide to suggest TitleCase for Type aliases
2025-11-09 04:35:01 +01:00
b31a03f134
Let CRT take care of the entry point for wWinMain if libc is linked
...
Fixes #7852
Before, the modified test would fail with:
```
error: lld-link: undefined symbol: wWinMain
note: referenced by C:\Users\Ryan\Programming\Zig\zig-x86_64-windows-0.15.1\lib\libc\mingw\crt\crtexewin.c:66
note: libmingw32.lib(ucrtexewin.obj):(wmain)
```
2025-11-08 17:11:12 -08:00
be4eaed7c4
Merge pull request #25860 from squeek502/coalesce-to-std-zig
...
Move/coalesce `CompressDebugSections` and `RcIncludes` enums to `std.zig`
2025-11-08 02:34:44 -08:00
43eb9b52cc
link.Elf: Ensure archive header fields are not left blank
...
ld.lld fails with "truncated or malformed archive" errors when reading
archive header field values that are not valid numbers.
2025-11-08 02:11:52 -08:00
38d44404a5
Fix param name in doc comment for std.Io.Reader.peek()
...
The old doc comment mentioned a parameter `len` three times, but the
function does not accept such a parameter - it is actually called `n`.
2025-11-08 00:34:41 -08:00
da77d306b6
Move/coalesce RcIncludes enum to std.zig.RcIncludes
2025-11-07 19:16:52 -08:00
f587209e04
Move/coalesce CompressDebugSections enum to std.zig.CompressDebugSections
2025-11-07 19:15:55 -08:00
74d2536715
Merge pull request #25158 from castholm/subsystem
...
Misc. Windows subsystem refactorings
2025-11-07 18:56:52 -08:00
bf15c791fa
Merge pull request #25820 from GiuseppeCesarano/process
...
Child.start_suspended ported to posix
2025-11-07 23:17:55 +01:00
852a1f718a
Fix kqueue definitions on NetBSD
...
EVFILT_USER and NOTE_TRIGGER were wrong.
Added missing ones along the way.
2025-11-07 22:23:46 +01:00
92f64899c1
sema: disallow slices of opaque types
2025-11-07 12:12:32 +00:00
19af9fa488
Merge pull request #25848 from Rexicon226/llvm-valgrind-clobber
...
llvm: fix up clobbers for valgrind requests
2025-11-07 11:55:09 +01:00
f4159eff92
std.Child: start_suspended ported to posix
2025-11-07 08:53:43 +01:00
5c0309a9e5
std.posix: implemented getpid and getppid
2025-11-07 08:52:35 +01:00
4b593a6c24
std.crypto: improve KT documentation, use key_length for B3 key length ( #25807 )
...
It was not obvious that the KT128/KT256 customization string can be
used to set a key, or what it was designed to be used for at all.
Also properly use key_length and not digest_length for the BLAKE3
key length (no practical changes as they are both 32, but that was
confusing).
Remove unneeded simd_degree copies by the way, and that doesn't need
to be in the public interface.
2025-11-07 08:20:04 +01:00
Matthew Lugg