rizin/subprojects/softfloat.wrap

[wrap-git]
url = https://github.com/rizinorg/softfloat
revision = 537d18e71a51aea70f6b54334854f7014c6458c7
directory = softfloat
depth = 1
patch_directory = softfloat

[provide]
softfloat=softfloat_dep
Use SoftFloat 3e for implementing arithmetic operations in RzFloat (#4535) * Add a new test for checking 80-bit floating point operations * New test `f80_ieee_div_test` tests the division of two 80-bit floats * Add SoftFloat 2c as a meson subproject * Add softfloat code to make the failing test case pass * Update the hash for the latest softfloat revision * Implement `rz_float_sqrt` using SoftFloat * Run the `f80_ieee_div_test` only for x86 * Replace SoftFloat version 2c with 3e * 3e has less bugs and more features * Modify the implementation in accordance * Update SoftFloat revision and add a guard around the 80-bit div test * Use SoftFloat for add, sub, mul operations as well * Make rem and mod also use SoftFloat functions * Also add test for mod and rem, and fix behavior of rem * Add comment about behavior of mod and rem * Add comments for tests which have different results for mod and rem * Simplify usage of loop variable as suggested in review * Remove unused macro from float.c * Implement `FMA` and `ROUND` using SoftFloat API * Add more tests for 80-bit floats * Change remote to a repository under rizinorg * Add info about the rounding mode in the Doxygen for rem and mod * Add comments in tests for rem and mod in `test_float.c` * Use bitvectors to initialize 80-bit soft floats * This makes the tests more portable and hence they can be run on any platform * Remove guards for f80 tests since they are portable now 2024-06-25 22:06:24 +05:30			`[wrap-git]`
			`url = https://github.com/rizinorg/softfloat`
Add wrap for native SoftFloat subproject 2024-09-04 03:18:21 +02:00			`revision = 537d18e71a51aea70f6b54334854f7014c6458c7`
Use SoftFloat 3e for implementing arithmetic operations in RzFloat (#4535) * Add a new test for checking 80-bit floating point operations * New test `f80_ieee_div_test` tests the division of two 80-bit floats * Add SoftFloat 2c as a meson subproject * Add softfloat code to make the failing test case pass * Update the hash for the latest softfloat revision * Implement `rz_float_sqrt` using SoftFloat * Run the `f80_ieee_div_test` only for x86 * Replace SoftFloat version 2c with 3e * 3e has less bugs and more features * Modify the implementation in accordance * Update SoftFloat revision and add a guard around the 80-bit div test * Use SoftFloat for add, sub, mul operations as well * Make rem and mod also use SoftFloat functions * Also add test for mod and rem, and fix behavior of rem * Add comment about behavior of mod and rem * Add comments for tests which have different results for mod and rem * Simplify usage of loop variable as suggested in review * Remove unused macro from float.c * Implement `FMA` and `ROUND` using SoftFloat API * Add more tests for 80-bit floats * Change remote to a repository under rizinorg * Add info about the rounding mode in the Doxygen for rem and mod * Add comments in tests for rem and mod in `test_float.c` * Use bitvectors to initialize 80-bit soft floats * This makes the tests more portable and hence they can be run on any platform * Remove guards for f80 tests since they are portable now 2024-06-25 22:06:24 +05:30			`directory = softfloat`
			`depth = 1`
Improve RzIL floating-point support (#6626) The `RzFloat` changes fix or improve: - binary80 explicit-integer-bit, pseudo-value, infinity, and NaN handling; - binary16 conversions; - gradual underflow and directed rounding; - overflow, underflow, invalid-operation, and inexact exception reporting; - exception propagation through nested conversions and arithmetic operations; - binary80 fused multiply-add rounding, including reduced-precision and double-rounding edge cases; - thread-local SoftFloat state, preventing rounding state from leaking between threads. The `RzIL` changes add scoped binary80 precision support through `RzFloatRPrecision` and `FWITH_RPREC`. The supported precisions are 32, 64, and 80. Precision scopes restore the previous thread-local SoftFloat state after successful evaluation and evaluation failures. Runtime rounding modes are represented explicitly by dedicated pure opcodes: - `FCONVERT_WITH_RMODE` - `FROUND_WITH_RMODE` - `FSQRT_WITH_RMODE` - `FADD_WITH_RMODE` - `FSUB_WITH_RMODE` - `FMUL_WITH_RMODE` - `FDIV_WITH_RMODE` - `FMOD_WITH_RMODE` Their rounding-mode operand is a 32-bit IL bitvector whose values correspond to `RzFloatRMode`: RNE, RNA, RTP, RTN, and RTZ. Invalid operand widths are rejected by validation, while invalid runtime values cause evaluation to fail with an error. Dedicated opcodes keep runtime-controlled floating-point expressions compact. This is useful for architectures whose rounding mode is selected from register state and avoids the expression duplication caused by expanding every operation into nested `ITE` branches. The new operations are supported by: - construction, duplication, and destruction; - type and operand validation; - VM evaluation; - plain, Unicode, and JSON exporters; - graph output and opcode stringification. `FEXCEPT` now emits a VM event only when the queried exception is present, while preserving exceptions raised by nested conversions and arithmetic operations. 2026-08-10 01:06:17 +08:00			`patch_directory = softfloat`
Add wrap for native SoftFloat subproject 2024-09-04 03:18:21 +02:00
			`[provide]`
			`softfloat=softfloat_dep`