9536d65230
main: Don't pass --seed for zig run.
2024-07-25 19:30:13 -07:00
a3c74aca99
add --debug-rt CLI arg to the compiler + bonus edits
...
The flag makes compiler_rt and libfuzzer be in debug mode.
Also:
* fuzzer: override debug logs and disable debug logs for frequently
called functions
* std.Build.Fuzz: fix bug of rerunning the old unit test binary
* report errors from rebuilding the unit tests better
* link.Elf: additionally add tsan lib and fuzzer lib to the hash
2024-07-25 18:52:21 -07:00
3256df2ff8
llvm: always include instrumentation function attributes
2024-07-25 18:52:20 -07:00
1c35e73b61
llvm: Don't emit safety memset() for stores of undef in Debug with safety off.
...
Before, this code:
@setRuntimeSafety(false);
var arr: [38]elf.Addr = undefined;
would emit a call to memset() in the output code in Debug mode, while in all the
release modes, LLVM optimized the memset() out as expected. Emitting the call in
Debug mode is problematic in some contexts, e.g. in std.os.linux.start_pie where
we are not yet ready to correctly perform calls because relocations haven't been
applied yet, or in the early stages of a dynamic linker, etc.
2024-07-25 18:04:50 -07:00
b22357b88b
Merge pull request #20783 from ziglang/cache-fix
...
build compiler_rt and fuzzer in parallel; fix false positive cache hits
2024-07-25 07:12:39 -07:00
18685e928d
main: avoid sending an empty file system inputs message
...
When all compiler inputs are invalid paths, there could be errors yet no
valid file system inputs.
Closes #20713
2024-07-25 03:59:30 -07:00
94f4f9c4ef
add sub-compilation cache inputs to parents in whole mode
...
closes #20782
2024-07-24 19:40:54 -07:00
4135cc9d0b
CLI: better error message when build runner cannot be spawned
2024-07-24 19:40:38 -07:00
3844b91db7
Compilation: build compiler_rt and fuzzer in parallel
...
With the rest of the pipeline.
Tracked by #9188
2024-07-24 18:24:00 -07:00
f4f5b2bc41
llvm: force TargetOptions.UseInitArray to true
...
The `TargetOptions` default constructor initializes all `bool`s to
`false`, yet clang defaults to setting this option to `true`. Since
recent glibc versions on linux do not appear to support this being set
to `false`, just changing the default for now unless a use case for
making it configurable is found.
2024-07-24 15:24:17 -07:00
75cf7fca90
Merge pull request #20767 from ziglang/emit-options
...
add "trace pc guard" as optional additional coverage instrumentation
2024-07-24 07:36:21 -07:00
fa95c89a71
Merge pull request #20758 from pavelverigo/stage2-wasm-compiler-rt-test-pass
...
stage2-wasm: pass compiler_rt test suite
2024-07-23 20:42:25 -07:00
26d2a7960e
default "trace pc guard" coverage off
...
* Add -f(no-)sanitize-coverage-trace-pc-guard CLI flag which defaults to
off. This value lowers to TracePCGuard = true (LLVM backend) and -Xclang
-fsanitize-coverage-trace-pc-guard. These settings are not
automatically included with -ffuzz.
* Add `Build.Step.Compile` flag for sanitize_coverage_trace_pc_guard
with appropriate documentation.
* Add `zig cc` integration for the respective flags.
* Avoid crashing in ELF linker code when -ffuzz -femit-llvm-ir used
together.
2024-07-23 17:30:53 -07:00
33d4742456
LLVM: more fine-grained sancov emit options
...
Exposes sanitizer coverage flags to the target machine emit function.
Makes it easier to change sancov options without rebuilding the C++
files.
This also enables PCTable = true for sancov which is needed by AFL, and
adds the corresponding Clang flag.
2024-07-23 16:04:58 -07:00
7aaebd1774
Merge pull request #20751 from Rexicon226/riscv-eflags
...
elf: add riscv eflag collisions
2024-07-23 22:29:57 +02:00
ba8522e6c7
stage2-wasm: fix bigint div and trunc
2024-07-23 21:17:04 +02:00
718f8d5314
Merge pull request #20706 from alexrp/sigaction-nosys
...
`std.posix`: Make `sigaction()` infallible
2024-07-23 11:49:43 -07:00
33c7984183
add std.testing.random_seed
...
closes #17609
2024-07-23 11:43:12 -07:00
6f3e9939d0
Merge pull request #20725 from ziglang/fuzz
...
initial support for integrated fuzzing
2024-07-23 11:39:19 -07:00
d71312d104
stage2-wasm: mul_sat 32 bits <=, i64, i128
2024-07-23 17:06:18 +02:00
f1af53f68e
macho: use pread syscall when loading tapi file
...
This avoids mixing preads with reads which do not mix well especially
on Windows.
2024-07-23 10:33:02 +02:00
782a9d16c7
elf: add riscv eflag collisions
2024-07-22 18:01:26 -07:00
61ad1be6bd
fix macho linker integration with libfuzzer
2024-07-22 14:26:17 -07:00
1cb994899d
Compilation: fix regression in addCCArgs
...
`-fno-sanitize=function` must come after `-fsanitize=undefined` or it
has no effect.
2024-07-22 14:26:17 -07:00
105b91d30f
ZigLLVMTargetMachineEmitToFile: make sancov enable the options
2024-07-22 14:26:17 -07:00
25198810c8
add new builtin: @disableInstrumentation
...
This is needed to ensure that start code does not try to access thread
local storage before it has set up thread local storage.
2024-07-22 13:07:02 -07:00
b9225aea78
add libfuzzer to linking
2024-07-22 13:07:02 -07:00
54b7e144b1
initial support for integrated fuzzing
...
* Add the `-ffuzz` and `-fno-fuzz` CLI arguments.
* Detect fuzz testing flags from zig cc.
* Set the correct clang flags when fuzz testing is requested. It can be
combined with TSAN and UBSAN.
* Compilation: build fuzzer library when needed which is currently an
empty zig file.
* Add optforfuzzing to every function in the llvm backend for modules
that have requested fuzzing.
* In ZigLLVMTargetMachineEmitToFile, add the optimization passes for
sanitizer coverage.
* std.mem.eql uses a naive implementation optimized for fuzzing when
builtin.fuzz is true.
Tracked by #20702
2024-07-22 13:07:02 -07:00
7c37c55161
macho: fix 32bit compilation issues
2024-07-22 17:32:43 +02:00
fcff82feb2
dev: add riscv64-linux support
2024-07-22 05:50:48 -07:00
8541119e9b
macho: handle empty string in ZigObject.getString
2024-07-22 13:35:55 +02:00
96fa29b90f
macho: write sections in parallel in -r mode
2024-07-22 12:06:02 +02:00
7e5a2673ac
macho: calc section sizes in parallel in -r mode
2024-07-22 12:06:02 +02:00
06a0da3e8a
macho: cache string len
2024-07-22 12:06:02 +02:00
79fefec599
add more tracy hooks
2024-07-22 12:06:02 +02:00
2b84592858
macho: run more things in parallel
2024-07-22 12:06:02 +02:00
5d9fd5bcde
macho: remove stale code
2024-07-22 12:06:02 +02:00
617f89f082
macho: parse input files in parallel
2024-07-22 12:06:02 +02:00
cba3389d90
macho: redo input file parsing in prep for multithreading
2024-07-22 12:05:56 +02:00
179a6e61e8
Merge pull request #20708 from alexrp/target-cleanup-2
...
`std.Target`: Remove more dead architecture tags.
2024-07-21 19:24:04 -07:00
42d9017feb
Sema: fix OOB access in coerceTupleToStruct ( #19620 )
...
Co-authored-by: Veikka Tuominen <git@vexu.eu >
2024-07-21 23:56:04 +00:00
be9841335e
std.Target.Os: Rename lv2 to ps3.
...
It is very non-obvious that this is what lv2 refers to, and we already use ps4
and ps5 to refer to the later models, so let's just be consistent.
2024-07-21 22:38:30 +02:00
db8bc4770c
std.Target: Remove the tce/tcele arch tags.
...
There is no obvious reason why this would be relevant for Zig to target. I
rather question the value of LLVM even having target triple code for this, too.
See: https://blog.llvm.org/2010/06/tce-project-co-design-of-application.html
See: https://github.com/cpc/llvmtce
2024-07-21 22:38:30 +02:00
21cc5a2044
std.Target: Remove the shave arch tag.
...
This was added as an architecture to LLVM's target triple parser and the Clang
driver in 2015. No backend ever materialized as far as I can see (same for GCC).
In 2016, other code referring to it started using "Myriad" instead. Ultimately,
all code related to it that isn't in the target triple parser was removed. It
seems to be a real product, just... literally no one seems to know anything
about the ISA. I figure after almost a decade with no public ISA documentation
to speak of, and no LLVM backend to reference, it's probably safe to assume that
we're not going to learn much about this ISA, making it useless for Zig.
See: https://github.com/llvm/llvm-project/commit/1b5767f72b5a037ca8f1802d737de97f8d92263d
See: https://github.com/llvm/llvm-project/commit/84a7564b28360843ee9afec5d3823c89623eb6a5
See: https://github.com/llvm/llvm-project/commit/8cfe9d8f2ad3a52ba7fd5841d3939aa810536e16
2024-07-21 22:38:30 +02:00
f1e0c35db4
std.Target: Remove hsail/hsail64 arch tags.
...
This seems to just be dead.
See: https://github.com/search?q=repo%3Allvm%2Fllvm-project%20hsail&type=code
See: https://github.com/HSAFoundation/HSAIL-Tools/commits/master
2024-07-21 22:38:30 +02:00
67a052df81
std.Target: Remove amdil/amdil64 arch tags.
...
This is really obscure and no one is 100% sure what it is. It seems to be old
and unused. My suspicion is that it's just an old term for "AMDGPU" before it
was upstreamed to LLVM.
See: https://github.com/search?q=repo%3Allvm%2Fllvm-project+amdil&type=code
2024-07-21 22:38:30 +02:00
c825b567b2
std.Target: Remove the r600 arch tag.
...
These are quite old GPUs, and it is unlikely that Zig will ever be able to
target them.
See: https://en.wikipedia.org/wiki/Radeon_HD_2000_series
2024-07-21 22:38:30 +02:00
9848623e62
std.Target: Remove the renderscript32/renderscript64 arch tags.
...
It's dead: https://developer.android.com/guide/topics/renderscript/migrate
2024-07-21 22:38:30 +02:00
2cced8903e
std.posix: Consider invalid signal numbers to sigaction() to be programmer error.
...
The set of signals that cannot have their action changed is documented in POSIX,
and any additional, non-standard signals are documented by the specific OS. I
see no valid reason why EINVAL should be considered an unpredictable error here.
2024-07-21 22:38:09 +02:00
f303c3943f
Revert "Merge pull request #20380 from tau-dev/master"
...
This reverts commit 397be0c9cc18d412ab2f
2024-07-21 02:44:58 -07:00
Alex Rønne Petersen