Unfortunately, due to the Windows equivalent of executable permissions
being a bit tricky, there is follow-up work to be done.
What is done in this commit is the hash modifications. At the fetch
layer, executable bits inside packages are ignored. In the hash
computation layer, executable bit is implemented for POSIX but not yet
for Windows. This means that the hash will not break again in the future
for packages that do not have any executable files, but it will break
for packages that do.
This is a hash-breaking change.
Closes#14308
PR #10610 addressed most of the points from #1854. This
additional paragraph and examples covers the OMISSIONS section
clarifying issues about indexing into non-ASCII strings (whether valid
UTF-8 or not). I think this finally closes#1854.
Several enhancements to the build system. Many breaking changes to the API.
* combine `std.build` and `std.build.Builder` into `std.Build`
* eliminate `setTarget` and `setBuildMode`; use an options struct for `b.addExecutable` and friends
* implement passing options to dependency packages. closes#14285
* rename `LibExeObjStep` to `CompileStep`
* move src.type.CType to std lib, use it from std.Build, this helps with populating config.h files.
Usage of `catch unreachable` in build scripts is completely harmless
because build scripts are always run in Debug mode, however, it sets a
poor example for beginners to learn from.
These declarations are now aliases of their new APIs and marked as
deprecated via doc comments:
* std.build.Builder
* std.build
* std.Build.LibExeObjStep
This is a breaking change that makes the API for creating build
artifacts no longer have any period of time where the target and
optimization mode are not set.
* introduce the concept of maps to user input options, but don't
implement it for command line arg parsing yet.
* remove setPreferredReleaseMode and standardReleaseOptions in favor of
standardOptimizeOption which has a future-proof options parameter.
These functions are currently footgunny when working with pointers to
arrays and slices. They just return the stated length of the array/slice
without iterating and looking for the first sentinel, even if the
array/slice is a sentinel terminated type.
From looking at the quite small list of places in the standard
library/compiler that this change breaks existing code, the new code
looks to be more readable in all cases.
The usage of std.mem.span/len was totally unneeded in most of the cases
affected by this breaking change.
We could remove these functions entirely in favor of other existing
functions in std.mem such as std.mem.sliceTo(), but that would be a
somewhat nasty breaking change as std.mem.span() is very widely used for
converting sentinel terminated pointers to slices. It is however not at
all widely used for anything else.
Therefore I think it is better to break these few non-standard and
potentially incorrect usages of these functions now and at some later
time, if deemed worthwhile, finally remove these functions.
If we wait for at least a full release cycle so that everyone adapts to
this change first, updating for the removal could be a simple find and
replace without needing to worry about the semantics.
- cbe: Implement linksection support, to support TLS when not linking libc
- cbe: Support under-aligned variables / struct fields
- cbe: Support packed structs (in the C definition of packed)
- windows: Fix regression with x86 _tls_array
- compiler_rt: Add 128-bit atomics to compiler_rt
- tests: Re-enable threadlocal tests on cbe+windows, and llvm+x86
- tests: Re-enable f80 tests that now pass
- ci: change windows ci to run the CBE behaviour tests with -lc, to match how the compiler is bootstrapped
- update zig1.wasm
This enhances the debugging experience as upon encountering a
breakpoint in a function, all arguments passed as registers have
already been moved to the stack, ready to be inspected by the
debugger.
The system linker shows warning when `msync` is linked instead of `__msync13`:
"warning: reference to compatibility msync(); include <sys/mman.h> for correct reference"
closes#14422
