diff --git a/read_buffer/src/chunk.rs b/read_buffer/src/chunk.rs
index 1ca84a1789..01a8571d55 100644
--- a/read_buffer/src/chunk.rs
+++ b/read_buffer/src/chunk.rs
@@ -258,7 +258,7 @@ impl Chunk {
     }
 
     /// Returns the distinct set of column names that contain data matching the
-    /// provided optional predicate.
+    /// provided predicate, which may be empty.
     ///
     /// `dst` is a buffer that will be populated with results. `column_names` is
     /// smart enough to short-circuit processing on row groups when it
diff --git a/read_buffer/src/column/bool.rs b/read_buffer/src/column/bool.rs
index a71a16bb64..704e535d09 100644
--- a/read_buffer/src/column/bool.rs
+++ b/read_buffer/src/column/bool.rs
@@ -307,11 +307,7 @@ impl Bool {
     /// Returns true if the column contains any non-null values at the rows
     /// provided.
     pub fn has_non_null_value(&self, row_ids: &[u32]) -> bool {
-        if !self.contains_null() {
-            return true;
-        }
-
-        row_ids.iter().any(|id| !self.arr.is_null(*id as usize))
+        !self.contains_null() || row_ids.iter().any(|id| !self.arr.is_null(*id as usize))
     }
 }
 
diff --git a/read_buffer/src/column/dictionary/plain.rs b/read_buffer/src/column/dictionary/plain.rs
index 84b66015bb..b589a98c10 100644
--- a/read_buffer/src/column/dictionary/plain.rs
+++ b/read_buffer/src/column/dictionary/plain.rs
@@ -670,7 +670,7 @@ impl Plain {
             return true;
         }
 
-        // If there are more than one dictionary entries then there is a non-null
+        // If there is more than one dictionary entry then there is a non-null
         // value in the column.
         self.entries.len() > 1
     }
diff --git a/read_buffer/src/column/dictionary/rle.rs b/read_buffer/src/column/dictionary/rle.rs
index 879bbdae92..2fc29e37d1 100644
--- a/read_buffer/src/column/dictionary/rle.rs
+++ b/read_buffer/src/column/dictionary/rle.rs
@@ -801,7 +801,7 @@ impl RLE {
             return true;
         }
 
-        // If there are any non-null rows then there will entries in the
+        // If there are any non-null rows then there are entries in the
         // dictionary.
         !self.entry_index.is_empty()
     }
diff --git a/read_buffer/src/lib.rs b/read_buffer/src/lib.rs
index ba1a68f522..7e006e304b 100644
--- a/read_buffer/src/lib.rs
+++ b/read_buffer/src/lib.rs
@@ -488,7 +488,7 @@ impl Database {
             })?;
 
         let chunk_data = partition.data.read().unwrap();
-        let mut filtered_chunks = vec![];
+        let mut filtered_chunks = Vec::with_capacity(chunk_ids.len());
         for id in chunk_ids {
             filtered_chunks.push(
                 chunk_data
diff --git a/read_buffer/src/row_group.rs b/read_buffer/src/row_group.rs
index 75c9a4741d..8e8a5c5629 100644
--- a/read_buffer/src/row_group.rs
+++ b/read_buffer/src/row_group.rs
@@ -923,10 +923,10 @@ impl RowGroup {
         dst.aggregates.push(AggregateResults(aggregate_row)); // write the row
     }
 
-    /// Given the optional predicate, determines a set of rows contained in this
-    /// row group that satisfy it. Any column that contains a non-null value at
-    /// any of these row positions is then included in the results, which are
-    /// added to `dst`.
+    /// Given the predicate (which may be empty), determine a set of rows
+    /// contained in this row group that satisfy it. Any column that contains a
+    /// non-null value at any of these row positions is then included in the
+    /// results, which are added to `dst`.
     ///
     /// As an optimisation, the contents of `dst` are checked before execution
     /// and any columns already existing in the set are not interrogated.