Add a case for shadows when blur_radius = 0

[iamnbutler]

Closes #22433

Before/After (macOS):

For some reason the non-blurred one seems much lower quality, so we may need to tinker with the samples, or something else.

I'm unsure if this is a problem on Linux/in the Blade renderer, but since no changes were made outside of the medal shaders we can probably take this macOS-specific win for now.

Release Notes:

• gpui: Fixed an issue where shadows with a blur_radius of 0 would not render.

[jansol]

Does metal have something like the GLSL smoothstep()? With appropriate threshold values using that instead of the hard conditional here should fix the aliased edges nicely.

[iamnbutler]

oh, I'll take a look!

[jansol]

EDIT: nvm, I was reading the way distance is calculated wrong.
Alternatively a simple alpha = (1.0 - min(1.0, distance * distance)) should look good too.

[jansol]

Blade shaders are directly transcribed from the metal ones so they'll need the same fix.

[iamnbutler]

Success! We have a 0-blur, non-jagged shadow:

[iamnbutler]

I think I feel good about where the metal shader change is now. Going to see if I can get the blade change done next!

[iamnbutler]

For future reference:

It looks like the blade renderer (at least when run on macOS using cargo run --example shadow -p gpui --features "gpui/macos-blade" has a number of shadow issues that don't match the metal one.

For now I'll work on fixing blur_radius = 0 but we should try to track these differences!

Edit: #22711

[iamnbutler]

Since this PR makes no structural changes, and the current blade rendering is unchanged, I'll merge this one and open a new PR with the blade changes!