Skip to content

precise_math attribute on functions #2080

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
kvark wants to merge 3 commits into
base: main
Choose a base branch
from
Open

precise_math attribute on functions #2080

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
c2e4178
precise_math attribute on functions
kvark Sep 1, 2021
9c8c768
Apply David's suggestions
kvark Sep 1, 2021
09e6e9c
Remove the nan/zero handling
kvark Sep 8, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
16 changes: 16 additions & 0 deletions wgsl/index.bs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -535,6 +535,18 @@ literal_or_ident
The third parameter, if provided, specifies the z dimension, otherwise is assumed to be 1.
Each dimension must be at least 1 and at most an upper bound specified by the WebGPU API.

<tr><td><dfn noexport dfn-for="attribute">`precise_math`</dfn>
<td>*None*

Indicates that the arithmetic computations in the function should be performed
under stricter [[#floating-point-evaluation]]:
* without [=Reassociation|reassociating=] subexpressions

Note: this translates to `NoContraction` decoration in SPIR-V, `precise` qualifier in HLSL,
and a subset of `"-fno-fast-math" group of compile options in MSL.

This attribute doesn't recursively affect other functions called from the current function.

</table>


Expand Down Expand Up @@ -5378,6 +5390,7 @@ param
[SHORTNAME] defines the following attributes that can be applied to function declarations:
* [=attribute/stage=]
* [=attribute/workgroup_size=]
* [=attribute/precise_math=]

[SHORTNAME] defines the following attributes that can be applied to function
parameters and return types:
Expand Down Expand Up @@ -6238,6 +6251,9 @@ the following exceptions:
* Other operations are required to preserve denormalized numbers.
* The accuracy of operations is given in [[#floating-point-accuracy]].

[SHORTNAME] attemps to reduce the scope of applied exceptions to floating-point operations
in functions with [=attribute/precise_math=] attributes.

### Floating Point Accuracy ### {#floating-point-accuracy}

<div algorithm="correctly rounded">
Expand Down