[ENH] Use AVX in distance calculations (#5258)

## Description of changes

This PR updates the distance crate to prioritize AVX over SSE, and
forces builds with AVX flags when ENABLE_AVX512 is set in the Dockerfile

## Test plan

_How are these changes tested?_

- [ ] Tests pass locally with `pytest` for python, `yarn test` for js,
`cargo test` for rust

## Migration plan

_Are there any migrations, or any forwards/backwards compatibility
changes needed in order to make sure this change deploys reliably?_

## Observability plan

_What is the plan to instrument and monitor this change?_

## Documentation Changes

_Are all docstrings for user-facing APIs updated if required? Do we need
to make documentation changes in the [docs
section](https://github.com/chroma-core/chroma/tree/main/docs/docs.trychroma.com)?_

Author: jairad26

rust/distance/src/types.rs

@@ -243,15 +243,6 @@ impl DistanceFunction {
                         return unsafe { crate::distance_neon::euclidean_distance(a, b) };
                     }
                 }
-                #[cfg(all(
-                    any(target_arch = "x86_64", target_arch = "x86"),
-                    target_feature = "sse"
-                ))]
-                {
-                    if std::arch::is_x86_feature_detected!("sse") {
-                        return unsafe { crate::distance_sse::euclidean_distance(a, b) };
-                    }
-                }
                 #[cfg(all(
                     target_arch = "x86_64",
                     all(target_feature = "avx", target_feature = "fma")
@@ -263,6 +254,15 @@ impl DistanceFunction {
                         return unsafe { crate::distance_avx::euclidean_distance(a, b) };
                     }
                 }
+                #[cfg(all(
+                    any(target_arch = "x86_64", target_arch = "x86"),
+                    target_feature = "sse"
+                ))]
+                {
+                    if std::arch::is_x86_feature_detected!("sse") {
+                        return unsafe { crate::distance_sse::euclidean_distance(a, b) };
+                    }
+                }
                 let mut sum = 0.0;
                 for i in 0..a.len() {
                     sum += (a[i] - b[i]).powi(2);
@@ -276,15 +276,6 @@ impl DistanceFunction {
                         return unsafe { crate::distance_neon::cosine_distance(a, b) };
                     }
                 }
-                #[cfg(all(
-                    any(target_arch = "x86_64", target_arch = "x86"),
-                    target_feature = "sse"
-                ))]
-                {
-                    if std::arch::is_x86_feature_detected!("sse") {
-                        return unsafe { crate::distance_sse::cosine_distance(a, b) };
-                    }
-                }
                 #[cfg(all(
                     target_arch = "x86_64",
                     all(target_feature = "avx", target_feature = "fma")
@@ -296,6 +287,15 @@ impl DistanceFunction {
                         return unsafe { crate::distance_avx::cosine_distance(a, b) };
                     }
                 }
+                #[cfg(all(
+                    any(target_arch = "x86_64", target_arch = "x86"),
+                    target_feature = "sse"
+                ))]
+                {
+                    if std::arch::is_x86_feature_detected!("sse") {
+                        return unsafe { crate::distance_sse::cosine_distance(a, b) };
+                    }
+                }
                 // For cosine we just assume the vectors have been normalized, since that
                 // is what our indices expect.
                 let mut sum = 0.0;
@@ -311,15 +311,6 @@ impl DistanceFunction {
                         return unsafe { crate::distance_neon::inner_product(a, b) };
                     }
                 }
-                #[cfg(all(
-                    any(target_arch = "x86_64", target_arch = "x86"),
-                    target_feature = "sse"
-                ))]
-                {
-                    if std::arch::is_x86_feature_detected!("sse") {
-                        return unsafe { crate::distance_sse::inner_product(a, b) };
-                    }
-                }
                 #[cfg(all(
                     target_arch = "x86_64",
                     all(target_feature = "avx", target_feature = "fma")
@@ -331,6 +322,15 @@ impl DistanceFunction {
                         return unsafe { crate::distance_avx::inner_product(a, b) };
                     }
                 }
+                #[cfg(all(
+                    any(target_arch = "x86_64", target_arch = "x86"),
+                    target_feature = "sse"
+                ))]
+                {
+                    if std::arch::is_x86_feature_detected!("sse") {
+                        return unsafe { crate::distance_sse::inner_product(a, b) };
+                    }
+                }
                 let mut sum = 0.0;
                 for i in 0..a.len() {
                     sum += a[i] * b[i];