Update the genToc funtion to make the name required in the Code node,
and add an additional optional field for the expected error, to use with
test_err, test_safety and obj_err.
Update langref.html.in to ensure all code blocks have a name that
- is unique, so that a doctest can be identified by it
- is descriptive
For test, test_err and test_safefy, ensure that the doctest name starts
with "test_", excluding doctests in the "Zig Test" section and doctests
that are imported by other doctests.
Ensure that the indentation of code_begin and code_end blocks are
consistent.
Fix a typo in pointer_arthemtic.
This fixes a bug in std.net caused during the introduction of
meta.assumeSentinel due to the unfortunate semantics of mem.span()
This leaves only 3 remaining uses of meta.assumeSentinel() in the
standard library, each of which could be a simple @ptrCast([*:0]T, foo)
instead. I think this function should likely be removed.
This avoids a warning that sometimes occurs when providing both a
-target argument that contains a version as well as the
-mmacosx-version-min argument. Zig provides the correct value in both
places, so it doesn't matter which one gets overridden.
Previously, if a source file was referenced from multiple packages, it
just became owned by the first one AstGen happened to reach; this was a
problem, because it could lead to inconsistent behaviour in the compiler
based on a race condition. This could be fixed by just analyzing such
files multiple times - however, it was pointed out by Andrew that it
might make more sense to enforce files being part of at most a single
package. Having a file in multiple packages would not only impact
compile times (due to Sema having to run multiple times on potentially a
lot of code) but is also a confusing anti-pattern which more often than
not is a mistake on the part of the user.
Resolves: #13662
By @Vexu's suggestion, since fetching the name from the parent package
is error-prone and complex, and optimising Package for size isn't really
a priority.
From RFC 1952:
> If FHCRC is set, a CRC16 for the gzip header is present,
> immediately before the compressed data. The CRC16 consists
> of the two least significant bytes of the CRC32 for all
> bytes of the gzip header up to and not including the CRC16.
* no longer repeat -lc on the linker line redundantly
* when using linkLibrary() with a static library, it will now also put
the static library's static library dependencies on the linker line,
recursively.
* refactor out a common pattern to an addFlag function
Changes only pertain arm64 as for x86_64 turns out we are
missing one additional feature which is synthesising unwind info
from only eh frame records, or at least this is what it looks like
when I analysed the output of Apple's ld.
