llvm-project.git/libc/utils/MPFRWrapper/MPFRUtils.cpp, branch users/vitalybuka/spr/main.clangcodegen-remove-simplifycfgpass-preceding-removetrapspass Unnamed repository; edit this file 'description' to name the repository. [libc][math][c23] Add exp2m1f C23 math function (#86996) 2024-04-04T12:22:45+00:00 OverMighty its.overmighty@gmail.com 2024-04-04T12:22:45+00:00 a8c59750d911eb30d5664696db19af445dd770f8 Fixes #86502. cc @lntue 
Fixes #86502.

cc @lntue
 [libc][math] Implement atan2f correctly rounded to all rounding modes. (#86716) 2024-04-01T17:31:07+00:00 lntue 35648136+lntue@users.noreply.github.com 2024-04-01T17:31:07+00:00 2be722587f5987891ed8b2904a03f983e987f226 We compute atan2f(y, x) in 2 stages: - Fast step: perform computations in double precision , with relative errors < 2^-50 - Accurate step: if the result from the Fast step fails Ziv's rounding test, then we perform computations in double-double precision, with relative errors < 2^-100. On Ryzen 5900X, worst-case latency is ~ 200 clocks, compared to average latency ~ 60 clocks, and average reciprocal throughput ~ 20 clocks. 
We compute atan2f(y, x) in 2 stages:
- Fast step: perform computations in double precision , with relative
errors < 2^-50
- Accurate step: if the result from the Fast step fails Ziv's rounding
test, then we perform computations in double-double precision, with
relative errors < 2^-100.

On Ryzen 5900X, worst-case latency is ~ 200 clocks, compared to average
latency ~ 60 clocks, and average reciprocal throughput ~ 20 clocks.
 [libc] Remove direct math.h includes (#85324) 2024-03-18T21:19:33+00:00 Michael Jones michaelrj@google.com 2024-03-18T21:19:33+00:00 5d56b34807e0f6e7a6684e57bec7c1751778862c Reland of #84991 A downstream overlay mode user ran into issues with the isnan macro not working in our sources with a specific libc configuration. This patch replaces the last direct includes of math.h with our internal math_macros.h, along with the necessary build system changes. 
Reland of #84991

A downstream overlay mode user ran into issues with the isnan macro not
working in our sources with a specific libc configuration. This patch
replaces the last direct includes of math.h with our internal
math_macros.h, along with the necessary build system changes.
 [libc][NFC] Remove `FPBits` cast operator (#79142) 2024-01-23T16:30:19+00:00 Guillaume Chatelet gchatelet@google.com 2024-01-23T16:30:19+00:00 2856db0d3b691907e055265c10f3ccc9b04f594e The semantics for casting can range from "bitcast" (same representation) to "different representation", to "type promotion". Here we remove the cast operator and force usage of `get_val` as the only function to get the floating point value, making the intent clearer and more consistent. 
The semantics for casting can range from "bitcast" (same representation)
to "different representation", to "type promotion". Here we remove the
cast operator and force usage of `get_val` as the only function to get
the floating point value, making the intent clearer and more consistent.
 [reland][libc] Remove unnecessary `FPBits` functions and properties (#79128) 2024-01-23T12:48:03+00:00 Guillaume Chatelet gchatelet@google.com 2024-01-23T12:48:03+00:00 6b02d2f86389d68a0cf2162377c5dda05bd4b68a - reland #79113 - Fix aarch64 RISC-V build 
- reland #79113
- Fix aarch64 RISC-V build
 Revert "[libc] Remove unnecessary `FPBits` functions and properties" (#79118) 2024-01-23T10:51:18+00:00 Guillaume Chatelet gchatelet@google.com 2024-01-23T10:51:18+00:00 b524eed9259e26b5812e2b55cc05e42fda217486 Reverts llvm/llvm-project#79113 It broke aarch64 build bot machines. 
Reverts llvm/llvm-project#79113
It broke aarch64 build bot machines.
 [libc] Remove unnecessary `FPBits` functions and properties (#79113) 2024-01-23T10:48:28+00:00 Guillaume Chatelet gchatelet@google.com 2024-01-23T10:48:28+00:00 3bc86bf3bf742506818cf4d94c9227e4afed6f19 This patch reduces the surface of `FPBits`. 
This patch reduces the surface of `FPBits`.
 [libc] Fix is_subnormal for Intel Extended Precision (#78592) 2024-01-19T08:36:03+00:00 Guillaume Chatelet gchatelet@google.com 2024-01-19T08:36:03+00:00 14f0c06f48aca3e6f015944c88097b404005645e Also turn a set of `get_biased_exponent() == 0` into `is_subnormal()` which is clearer. 
Also turn a set of `get_biased_exponent() == 0` into `is_subnormal()`
which is clearer.
 [libc][NFC] Remove `FloatProperties` (#76508) 2024-01-03T08:51:58+00:00 Guillaume Chatelet gchatelet@google.com 2024-01-03T08:51:58+00:00 c09e6905567a6b546bb2fd9e863511a8fb939b19 Access is now done through `FPBits` exclusively. This patch also renames a few internal structs and uses `T` instead of `FP` as a template parameter. 
Access is now done through `FPBits` exclusively.
This patch also renames a few internal structs and uses `T` instead of
`FP` as a template parameter.
 [libc][NFC] Integrate `FloatProperties` into `FPBits` (#76506) 2023-12-28T14:42:47+00:00 Guillaume Chatelet gchatelet@google.com 2023-12-28T14:42:47+00:00 c23991478aa79ea1040531b3077976ae56415b24 `FloatProperties` is always included when `FPBits` is. This will help further refactoring. 
`FloatProperties` is always included when `FPBits` is. This will help
further refactoring.