5647dd84c1
Revert d714cf3340 ( #21273 )
...
Turns out this was already fixed in #21964 .
I have no idea why GitHub showed an incorrect diff in #21273 , or how applying the diff to master was even possible, but here we are.
2025-01-26 20:52:29 -08:00
7599216341
std.zip: Include the last paragraph in top-level doc comments
...
Co-authored-by: Alex Rønne Petersen <alex@alexrp.com >
2025-01-27 12:24:36 +09:00
ae5b1a9ca2
langref: fixes incorrect description of cmpxchg functions
2025-01-26 23:22:04 +01:00
658a4a527c
Fix incompatibility with multi-config cmake generator
...
Avoid hard coding the path of zig2 relative to PROJECT_BINARY_DIR
2025-01-26 21:06:09 +01:00
3f245616a5
link to Tuples section in Hello World docs, and give an example of an empty tuple
2025-01-26 20:53:25 +01:00
9c45fb4b09
langref: correct return type of comptime known bitcount ops
2025-01-26 20:30:06 +01:00
d714cf3340
std.c.darwin: Fix EXC.MASK compile error ( #21273 )
2025-01-26 19:56:14 +01:00
6255ee3d2f
fix typo in @typeName langref
2025-01-26 19:40:57 +01:00
672bc8141f
fix: Only suggest try on destructure of error union if payload type can be destructured ( #21510 )
2025-01-26 19:38:07 +01:00
9c6d728b0c
std.meta: handle comptime fields in hasUniqueRepresentation ( #22132 )
2025-01-26 18:23:34 +00:00
3179961854
std.Build: add more explicit error sets to some functions
...
Signed-off-by: Eric Joldasov <bratishkaerik@landless-city.net >
2025-01-26 19:03:33 +01:00
738d813619
Add zig std to langref.html ( #22360 )
2025-01-26 18:58:30 +01:00
5b3eaff80d
std: check max depth for vector type in formatType
2025-01-26 18:53:11 +01:00
24965af295
std.os.linux: Adding sigdelset ( #22406 )
2025-01-26 18:49:25 +01:00
df9fdb1861
std.net.listen: no REUSEPORT with unix socket
...
On Linux, set REUSEPORT option on an unix socket returns a EOPNOTSUPP error.
See https://github.com/torvalds/linux/commit/5b0af621c3f6ef9261cf6067812f2fd9943acb4b
2025-01-26 18:40:34 +01:00
df1cd62720
compiler-rt: ensure memcpy Element has no unused bytes
2025-01-26 07:00:24 -08:00
654da648b3
x86_64: rewrite @min/@max for float vectors
2025-01-26 09:51:07 -05:00
0c890bb9a4
x86_64: rewrite @min/@max for scalar floats
2025-01-26 06:58:37 -05:00
92b20e4216
stf: do not call xcode-select or xcrun by absolute path in darwin sdk detection
2025-01-26 09:07:39 +01:00
6a4986b2d6
Update LDR_DATA_TABLE_ENTRY struct
2025-01-26 03:40:17 +01:00
3767b08039
Merge pull request #22602 from mlugg/incr-embedfile
...
incremental: handle `@embedFile`
2025-01-26 01:41:56 +00:00
8fa47bb904
Merge pull request #22230 from alexrp/lto-stuff
...
Disable LTO by default + some LTO fixes
2025-01-25 13:17:25 -08:00
015a5e1982
std.Build: Make ofmt part of standardTargetOptions().
2025-01-25 20:39:10 +01:00
d0d5ca2b6c
Merge pull request #22581 from jacobly0/x86_64-rewrite
...
x86_64: rewrite `@abs` on floats
2025-01-25 07:30:45 -08:00
b60e39fe8f
Compilation: Disable LTO by default.
...
LLD has too many LTO bugs, and we're dropping the LLD dependency soon anyway.
2025-01-25 14:59:02 +01:00
9f116d38b8
libtsan: Build with unwind tables like upstream.
2025-01-25 14:58:55 +01:00
46a1a15861
Compilation: Remove the _tls_index hack for MinGW.
...
No longer needed since we disable LTO for mingw32.
2025-01-25 14:58:44 +01:00
1e8b82f6b4
compiler: Rework LTO settings for some Zig-provided libraries.
...
* compiler-rt and mingw32 have both run into LLD bugs, and LLVM disables LTO for
its compiler-rt, so disable LTO for these.
* While we haven't run into any bugs in it, LLVM disables LTO for its libtsan,
so follow suit just to be safe.
* Allow LTO for libfuzzer as LLVM does.
2025-01-25 14:57:33 +01:00
afe2fed34d
link: Set machine and float ABI when invoking ld.lld and lld-link.
...
If this isn't done, LTO can completely miscompile the input bitcode modules for
certain targets where we need to explicitly set these ABIs (because LLVM's
defaults are bad).
2025-01-25 14:57:33 +01:00
cd8c92fc40
compiler: Explicitly specify ABI for arm, mips, and powerpc when talking to LLVM.
...
See 652c5151429e279f70396ee601416c87a70c1bec. Better to avoid relying on default
LLVM behavior going forward.
2025-01-25 14:56:59 +01:00
b95081209a
compiler: Explicitly specify loongarch ABI when talking to LLVM.
...
Necessary because of: https://github.com/llvm/llvm-project/commit/dc665fa5f5b8b572479ceac6bf32e0174de65f1e
2025-01-25 14:56:59 +01:00
673544b783
compiler-rt: Pass called functions into inline asm so the compiler sees the usage.
2025-01-25 14:54:02 +01:00
35da1e1e8f
compiler-rt: Don't export __a(u)ll(div,rem) if linking libc.
...
libc provides these when targeting MSVC.
2025-01-25 13:49:21 +01:00
05a877a9a1
compiler-rt: More accurately export Windows Arm functions.
2025-01-25 13:49:21 +01:00
fcf8d5ada2
incr-check: check compile errors against expected
...
Also modifies all incremental cases using `#expect_error` to include the
errors and notes which are expected.
2025-01-25 09:32:40 +00:00
7ef345f342
incr-check: deal with Windows stupidity
...
The real problem here is that Git for Windows has horrendous defaults
which convert LF to CRLF. However, rather than changing this
configuration on the CI runners, it's worth supporting inexplicable CRLF
in these files so that anyone else cloning Zig on Windows doesn't get
unexpected test failures.
2025-01-25 08:57:20 +00:00
f47b8de2ad
incremental: handle @embedFile
...
Uses of `@embedFile` register dependencies on the corresponding
`Zcu.EmbedFile`. At the start of every update, we iterate all embedded
files and update them if necessary, and invalidate the dependencies if
they changed.
In order to properly integrate with the lazy analysis model, failed
embed files are now reported by the `AnalUnit` which actually used
`@embedFile`; the filesystem error is stored in the `Zcu.EmbedFile`.
An incremental test is added covering incremental updates to embedded
files, and I have verified locally that dependency invalidation is
working correctly.
2025-01-25 06:07:08 +00:00
5202c977d9
std: add fs.path.fmtJoin
...
This allows joining paths without allocating using a `Writer`.
2025-01-25 04:48:00 +00:00
8ba3812eee
Merge pull request #22594 from mlugg/panic-stuff
...
compiler: yet more panic handler changes
2025-01-25 04:10:55 +00:00
c7433212d1
x86_64: rewrite scalar and vector int @min and @max
2025-01-24 21:02:32 -05:00
7701cfa032
x86_64: mitigate miscomp during switch dispatch
2025-01-24 20:56:11 -05:00
ba82d6e83e
x86_64: fix typo and lower optimized insts
2025-01-24 20:56:11 -05:00
b1fa89439a
x86_64: rewrite float vector @abs and equality comparisons
2025-01-24 20:56:11 -05:00
ae3d95fc8d
x86_64: rewrite scalar float equality comparisons
2025-01-24 20:56:11 -05:00
921725427e
Merge pull request #22596 from ziglang/pipeline
...
Compilation pipeline: repeat failed prelink tasks
2025-01-24 19:52:09 -05:00
b6726913d3
Zcu: remove null_stack_trace
...
The new simplifications to the panic handler have eliminated the need
for this piece of memoized state.
2025-01-24 22:33:23 +00:00
3232e59ab9
std: remove old declarations now a zig1 update has happened
2025-01-24 22:33:23 +00:00
38b47034f7
compiler_rt: upgrade to new panic interface
2025-01-24 22:33:23 +00:00
5a6666db55
all: update for panic.unwrapError and panic.call signature changes
2025-01-24 22:33:23 +00:00
dd334d5ee5
stage1: update zig1.wasm
...
Oh dear, not again! Sorry, I changed the panic handler...
Signed-off-by: mlugg <mlugg@mlugg.co.uk >
2025-01-24 21:07:14 +00:00
Alex Rønne Petersen