| Age | Commit message (Collapse) | Author |
|---|
|
stage2: change logic for detecting whether the main package is inside
the std package. Previously it relied on realpath() which is not portable.
This uses resolve() which is how imports already work.
* stage2: fix cleanup bug when creating Module
* flatten lib/std/special/* to lib/*
- this was motivated by making main_pkg_is_inside_std false for
compiler_rt & friends.
* rename "mini libc" to "universal libc"
|
|
This change:
- Adds generic implementation of the float -> integer conversion
functions floatXiYf, including support for f80
- Updates the existing implementation of integer -> float conversion
fixXiYf to support f16 and f80
- Fixes the handling of the explicit integer bit in `__trunctfxf2`
- Combines the test cases for fixXfYi/floatXiYf into a single file
- Renames `fmodl` to `fmodq`, since it operates on 128-bit floats
The new implementation for floatXiYf has been benchmarked, and generally
provides equal or better performance versus the current implementations:
Throughput (MiB/s) - Before
| u32 | i32 | u64 | i64 | u128 | i128 |
-----|----------|----------|----------|----------|----------|----------|
f16 | none | none | none | none | none | none |
f32 | 2231.67 | 2001.19 | 1745.66 | 1405.77 | 2173.99 | 1874.63 |
f64 | 1407.17 | 1055.83 | 2911.68 | 2437.21 | 1676.05 | 1476.67 |
f80 | none | none | none | none | none | none |
f128 | 327.56 | 321.25 | 645.92 | 654.52 | 1153.56 | 1096.27 |
Throughput (MiB/s) - After
| u32 | i32 | u64 | i64 | u128 | i128 |
-----|----------|----------|----------|----------|----------|----------|
f16 | 1407.61 | 1637.25 | 3555.03 | 2594.56 | 3680.60 | 3063.34 |
f32 | 2101.36 | 2122.62 | 3225.46 | 3123.86 | 2860.05 | 1985.21 |
f64 | 1395.57 | 1314.87 | 2409.24 | 2196.30 | 2384.95 | 1908.15 |
f80 | 475.53 | 457.92 | 884.50 | 812.12 | 1475.27 | 1382.16 |
f128 | 359.60 | 350.91 | 723.08 | 706.80 | 1296.42 | 1198.87 |
|
|
* goals
- zig as linker for object files generated by other compilers
- zig-specific runtime features for eventual standardisation
* changes
- missing routines are marked with `missing`
- structure inspired by libgcc docs, but improved order and wording
- rename misspelled functions
- reorder and rephrase compiler_rt.zig to reflect documentation
- potential decimal float or fixed-point arithmetic support:
* 'Decimal float library routines' ca. 120 functions
* 'Fixed-point fractional library routines' ca. 300 functions
thanks to @Vexu for multiple reviews and @scheibo for review
|
|
Get rid of `std.math.F80Repr`. Instead of trying to match the memory
layout of f80, we treat it as a value, same as the other floating point
types. The functions `make_f80` and `break_f80` are introduced to
compose an f80 value out of its parts, and the inverse operation.
stage2 LLVM backend: fix pointer to zero length array tripping LLVM
assertion. It now checks for when the element type is a zero-bit type
and lowers such thing the same way that pointers to other zero-bit types
are lowered.
Both stage1 and stage2 LLVM backends are adjusted so that f80 is lowered
as x86_fp80 on x86_64 and i386 architectures, and identical to a u80 on
others. LLVM constants are lowered in a less hacky way now that #10860
is fixed, by using the expression `(exp << 64) | fraction` using llvm
constants.
Sema is improved to handle c_longdouble by recursively handling it
correctly for whatever the float bit width is. In both stage1 and
stage2.
|
|
|
