Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
mstdn.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
A general-purpose Mastodon server with a 500 character limit. All languages are welcome.

Administered by:

Server stats:

13K
active users

mstdn.social: AboutStatusProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.8

#bittwiddling

0 posts0 participants0 posts today
Public
Shane Celis

I'm sure this has been done many times before, but I'm still happy to be bit-twiddling in #shaders. The star is this line:

(vertex_index & 3u) + (vertex_index >> 2u)

"Give me the first two bits but add one if the third bit is set.”

The shader generates a UV-mapped quad (right-triangle pair) from the 0 through 5 vertex indices. The screenshot code is abridged; full code is at the link.

github.com/shanecelis/bevy/blo #gamedev #bevyengine #bittwiddling #wgsl

// Gives viewport in clip space `(min_x, min_y, max_x, max_y)` @group(1) @binding(0) var<uniform> viewport: vec4<f32>; // This vertex shader produces the following UVs, when drawn using indices 0..6: // // 0 | 0-----------2 // | | . ´4 // | | a .´ | // | | .´ b | // 1 | 1´5_________3 // V +--------------- // U 0 1 // // The UV vectors can be converted clip-space vertices to cover the entire // screen with the following code: // // ```wgsl // let clip_position = vec4<f32>(uv * vec2<f32>(2.0, -2.0) + vec2<f32>(-1.0, 1.0), 0.0, 1.0); // ``` @vertex fn viewport_vertex_shader(@builtin(vertex_index) vertex_index: u32) -> ViewportVertexOutput { let index = (vertex_index & 3u) + (vertex_index >> 2u); var uv = vec2<f32>( f32(index >> 1u), // X is second bit, 0 or 1. f32(index & 1u) // Y is first bit, 0 or 1. ); let size = viewport.zw - viewport.xy; let clip_position = vec4<f32>(uv * vec2<f32>(size.x, -size.y) + vec2<f32>(viewport.x, viewport.z), 0.0, 1.0); return ViewportVertexOutput(clip_position, uv); }
Public
Kernel Bob

In 2013, past me wrote some mysterious code. Effectively, it was this.

uint8_t decay(uint8_t x) { return (x >> 1) | (x >> 2); }

Sort of a low level multiply by 0.75, but with an | instead of a +. The target was an 8 bit AVR. Is there any good reason for using | here? I vaguely remember a bug where certain values didn't decay, but testing shows all 255 nonzero values of x do decay.

Any thoughts?

#CProgramming#BitTwiddling
ExploreLive feeds
About