PR #13101 recently renamed the "i386" architecture to "x86", and it
seems the specific CPU model got swept up in that. "x86" is an umbrella
term that describes a family of CPUs, and the "i386" is the oldest
supported model under that umbrella.
This function is redundant with CType.sizeInBits(), and until the
previous commit they disagreed about the correct long double type
for several targets. Although they're all synced up now, it's much
simpler just to have a single source of truth.
These updates were made by testing against the `sizeof/_Alignof` reported
by Clang for all supported arch-OS-ABI combinations and correcting any
discrepancies.
This is bound to have a few errors (the recent long double fix for i386
Android is one example), but Clang is certainly not a bad place to start,
especially for our most popular targets.
This makes the following changes for i386:
long long and unsigned long long have 4 byte alignment on non-Windows
f64 (double) has 4-byte alignment on non-Windows
long double is 80 bits and has 4 byte alignment on mingw
long double on android is 64 bits, not 80: https://www.uclibc.org/docs/psABI-i386.pdfFixes#12453Fixes#12987
* riscv64: adjust alignment and size of 128-bit integers.
* take ofmt=c into account for ABI alignment of 128-bit integers and
structs.
* Type: make packed struct support intInfo
* fix f80 alignment for i386-windows-msvc
Notable changes:
`_i386`, `_i486`, and `_i686` are renamed to `i386`, `i486`,
and `i686` respectively. `std.zig.fmtId` is enhanced to support
formatting `i386` as `@"i386"`.
Some CPU features which are actually CPU models have been
properly flattened, such as `apple_a12`, `apple_a13`, `apple_a7`,
`cortex_a78c`, `exynos_m4`, `neoverse_e1`, `neoverse_n1`,
`neoverse_n2`, `neoverse_v1`.
Some CPU features have been added and some have been removed, following
LLVM's lead.
CSky CPU features support is added.
Prior to this change we would assume the ABI for Apple targets to
be GNU which could result in subtle errors in LLVM emitting calls
to non-existent system libc provided functions such as `_sincosf`
which is a GNU extension and as such is not provided by macOS for example.
This would result in linker errors where the linker would not be
able to find the said symbol in `libSystem.tbd`.
With this change, we now correctly identify macOS (and other Apple
platforms) as having ABI `unknown` which translates to unspecified
in LLVM under-the-hood:
```
// main.ll
target triple = "aarch64-unknown-macos-unknown"
```
Note however that we never suffix the target OS with target version
such as `macos11` or `macos12` which means we fail to instruct LLVM
of potential optimisations provided by the OS such as the availability
of function `___sincosf_stret`. I suggest we investigate that in a
follow-up commit.
Rename all references of sparcv9 to sparc64, to make Zig align more with
other projects. Also, added new function to convert glibc arch name to Zig
arch name, since it refers to the architecture as sparcv9.
This is based on the suggestion by @kubkon in PR 11847.
(https://github.com/ziglang/zig/pull/11487#pullrequestreview-963761757)
These targets now have a similar disagreement with LLVM about the
alignment of 128-bit integers as x86_64:
* riscv64
* powerpc64
* powerpc64le
* mips64
* mips64el
* sparcv9
See #2987
For x86_64, LLVMABIAlignmentOfType(i128) reports 8. However I think 16
is a better number for two reasons:
1. Better machine code when loading into SIMD register.
2. The C ABI wants 16 for extern structs.
Prior to this commit, the logic for ABI size and ABI alignment for
integers was naive and incorrect. This results in wasted hardware as
well as undefined behavior in the LLVM backend when we memset an
incorrect number of bytes to 0xaa due to disagreeing with LLVM about the
ABI size of integers.
This commit introduces a "max int align" value which is different per
Target. This value is used to derive the ABI size and alignment of all
integers.
This commit makes an interesting change from stage1, which treats
128-bit integers as 16-bytes aligned for x86_64-linux. stage1 is
incorrect. The maximum integer alignment on this system is only 8 bytes.
This change breaks the behavior test called "128-bit cmpxchg" because on
that target, 128-bit cmpxchg does require a 16-bytes aligned pointer to
a 128 bit integer. However, this alignment property does not belong on
*all* 128 bit integers - only on the pointer type in the `@cmpxchg`
builtin function prototype. The user can then use an alignment override
annotation on a 128-bit integer variable or struct field to obtain such
a pointer.
According to Apple docs, the long double type is a double precision
IEEE754 binary floating-point type, which makes it identical to the
double type. This behavior contrasts to the standard specification,
in which a long double is a quad-precision, IEEE754 binary,
floating-point type.
Thus, we need to take this into account when using the compiler
intrinsics so that we select the correct function version for
FloatMulAdd.
* mul_add AIR instruction: use `pl_op` instead of `ty_pl`. The type is
always the same as the operand; no need to waste bytes redundantly
storing the type.
* AstGen: use coerced_ty for all the operands except for one which we
use to communicate the type.
* Sema: use the correct source location for requireRuntimeBlock in
handling of `@mulAdd`.
* native backends: handle liveness even for the functions that are
TODO.
* C backend: implement `@mulAdd`. It lowers to libc calls.
* LLVM backend: make `@mulAdd` handle all float types.
- improved fptrunc and fpext to handle f80 with compiler-rt calls.
* Value.mulAdd: handle all float types and use the `@mulAdd` builtin.
* behavior tests: revert the changes to testing `@mulAdd`. These
changes broke the test coverage, making it only tested at
compile-time.
Improved f80 support:
* std.math.fma handles f80
* move fma functions from freestanding libc to compiler-rt
- add __fmax and fmal
- make __fmax and fmaq only exported when they don't alias fmal.
- make their linkage weak just like the rest of compiler-rt symbols.
* removed `longDoubleIsF128` and replaced it with `longDoubleIs` which
takes a type as a parameter. The implementation is now more accurate
and handles more targets. Similarly, in stage2 the function
CTypes.sizeInBits is more accurate for long double for more targets.
* zig_clang is fully updated
* zig_llvm is fully updated
Some initial work on codegen.cpp is in place for upgrading to LLVM's
new opaque pointers. However there is much more to be done.
A few of zig llvm bindings for deprecated functions have been updated;
more need to be updated.
This number tracks the glibc version in the oldest still-active LTS
version of Debian, which is Jessie, extended LTS expiring in June 2022,
at which point this number can be bumped again.
`getExternalExecutor` is moved from `std.zig.CrossTarget` to
`std.zig.system.NativeTargetInfo.getExternalExecutor`.
The function also now communicates a bit more information about *why*
the host is unable to execute a binary. The CLI is updated to report
this information in a useful manner.
`getExternalExecutor` is also improved to detect such patterns as:
* x86_64 is able to execute x86 binaries
* aarch64 is able to execute arm binaries
* etc.
Added qemu-hexagon support to `getExternalExecutor`.
`std.Target.canExecBinaries` of is removed; callers should use the more
powerful `getExternalExecutor` instead.
Now that `zig test` tries to run the resulting binary no matter what,
this commit has a follow-up change to the build system and docgen to
utilize the `getExternalExecutor` function and pass `--test-no-exec`
in some cases to avoid getting the error.
Additionally:
* refactor: extract NativePaths and NativeTargetInfo into their own
files named after the structs.
* small improvement to langref to reduce the complexity of the `callconv`
expression in a couple examples.
Due to a deficiency in LLD, we need to special-case BPF to a simple
file copy when generating relocatables. Normally, we would expect
`lld -r` to work. However, because LLD wants to resolve BPF relocations
which it shouldn't, it fails before even generating the relocatable.
Co-authored-by: Matthew Knight <mattnite@protonmail.com>
This is mainly because arm64 macOS doesn't support all
versions supported by x86_64 macOS. This is just a temporary
thing until both architectures support the same set of OSes.
