From 1b8e968d886205d09f88f4ef566ef6b88aa6e6ce Mon Sep 17 00:00:00 2001
From: itsjunetime <junewelker@gmail.com>
Date: Fri, 8 Nov 2024 16:01:01 -0700
Subject: [PATCH 1/7] Fix Buffer::bit_slice losing length with byte-aligned
 offsets

---
 Cargo.toml                           |  2 +-
 arrow-buffer/src/buffer/immutable.rs | 35 +++++++++++++++++++++++++++-
 arrow-flight/src/encode.rs           |  2 +-
 3 files changed, 36 insertions(+), 3 deletions(-)

diff --git a/Cargo.toml b/Cargo.toml
index f210ae210012..c5527fc41cca 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -74,7 +74,7 @@ include = [
     "Cargo.toml",
 ]
 edition = "2021"
-rust-version = "1.62"
+rust-version = "1.75"
 
 [workspace.dependencies]
 arrow = { version = "53.2.0", path = "./arrow", default-features = false }
diff --git a/arrow-buffer/src/buffer/immutable.rs b/arrow-buffer/src/buffer/immutable.rs
index 8d1a46583fca..4c100ab07bd6 100644
--- a/arrow-buffer/src/buffer/immutable.rs
+++ b/arrow-buffer/src/buffer/immutable.rs
@@ -265,7 +265,7 @@ impl Buffer {
     /// otherwise a new buffer is allocated and filled with a copy of the bits in the range.
     pub fn bit_slice(&self, offset: usize, len: usize) -> Self {
         if offset % 8 == 0 {
-            return self.slice(offset / 8);
+            return self.slice_with_length(offset / 8, len.div_ceil(8));
         }
 
         bitwise_unary_op_helper(self, offset, len, |a| a)
@@ -860,4 +860,37 @@ mod tests {
         let iter_len = usize::MAX / std::mem::size_of::<u64>() + 1;
         let _ = Buffer::from_iter(std::iter::repeat(0_u64).take(iter_len));
     }
+
+    #[test]
+    fn bit_slice_length_preserved() {
+        // Create a boring buffer
+        let buf = Buffer::from_iter(std::iter::repeat(true).take(64));
+
+        let assert_preserved = |offset: usize, len: usize| {
+            let new_buf = buf.bit_slice(offset, len);
+            assert_eq!(new_buf.len(), len.div_ceil(8));
+
+            // if the offset is not byte-aligned, we have to create a deep copy to a new buffer
+            // (since the `offset` value inside a Buffer is byte-granular, not bit-granular), so
+            // checking the offset should always return 0 if so. If the offset IS byte-aligned, we
+            // want to make sure it doesn't unnecessarily create a deep copy.
+            if offset % 8 == 0 {
+                assert_eq!(new_buf.ptr_offset(), offset / 8);
+            } else {
+                assert_eq!(new_buf.ptr_offset(), 0);
+            }
+        };
+
+        // go through every available value for offset
+        for o in 0..=64 {
+            // and go through every length that could accompany that offset - we can't have a
+            // situation where offset + len > 64, because that would go past the end of the buffer,
+            // so we use the map to ensure it's in range.
+            for l in (o..=64).map(|l| l - o) {
+                // and we just want to make sure every one of these keeps its offset and length
+                // when neeeded
+                assert_preserved(o, l);
+            }
+        }
+    }
 }
diff --git a/arrow-flight/src/encode.rs b/arrow-flight/src/encode.rs
index ae3475c7c7d7..e65295e8750f 100644
--- a/arrow-flight/src/encode.rs
+++ b/arrow-flight/src/encode.rs
@@ -1741,7 +1741,7 @@ mod tests {
 
         let batch = RecordBatch::try_from_iter(vec![("a1", Arc::new(array) as _)]).unwrap();
 
-        verify_encoded_split(batch, 160).await;
+        verify_encoded_split(batch, 48).await;
     }
 
     #[tokio::test]

From 0db95dd6f78ec3670c5ea9b97b13af3a714239a7 Mon Sep 17 00:00:00 2001
From: itsjunetime <junewelker@gmail.com>
Date: Fri, 8 Nov 2024 16:48:13 -0700
Subject: [PATCH 2/7] Fix msrv confusion and bit_util::ceil usage

---
 Cargo.toml                                   | 2 +-
 arrow-buffer/src/buffer/immutable.rs         | 6 +++---
 arrow-pyarrow-integration-testing/Cargo.toml | 2 +-
 arrow/Cargo.toml                             | 2 +-
 object_store/Cargo.toml                      | 2 +-
 5 files changed, 7 insertions(+), 7 deletions(-)

diff --git a/Cargo.toml b/Cargo.toml
index c5527fc41cca..bf2274a40306 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -74,7 +74,7 @@ include = [
     "Cargo.toml",
 ]
 edition = "2021"
-rust-version = "1.75"
+rust-version = "1.70"
 
 [workspace.dependencies]
 arrow = { version = "53.2.0", path = "./arrow", default-features = false }
diff --git a/arrow-buffer/src/buffer/immutable.rs b/arrow-buffer/src/buffer/immutable.rs
index 4c100ab07bd6..7254313b7da5 100644
--- a/arrow-buffer/src/buffer/immutable.rs
+++ b/arrow-buffer/src/buffer/immutable.rs
@@ -23,7 +23,7 @@ use std::sync::Arc;
 use crate::alloc::{Allocation, Deallocation, ALIGNMENT};
 use crate::util::bit_chunk_iterator::{BitChunks, UnalignedBitChunk};
 use crate::BufferBuilder;
-use crate::{bytes::Bytes, native::ArrowNativeType};
+use crate::{bit_util, bytes::Bytes, native::ArrowNativeType};
 
 use super::ops::bitwise_unary_op_helper;
 use super::{MutableBuffer, ScalarBuffer};
@@ -265,7 +265,7 @@ impl Buffer {
     /// otherwise a new buffer is allocated and filled with a copy of the bits in the range.
     pub fn bit_slice(&self, offset: usize, len: usize) -> Self {
         if offset % 8 == 0 {
-            return self.slice_with_length(offset / 8, len.div_ceil(8));
+            return self.slice_with_length(offset / 8, bit_util::ceil(len, 8));
         }
 
         bitwise_unary_op_helper(self, offset, len, |a| a)
@@ -868,7 +868,7 @@ mod tests {
 
         let assert_preserved = |offset: usize, len: usize| {
             let new_buf = buf.bit_slice(offset, len);
-            assert_eq!(new_buf.len(), len.div_ceil(8));
+            assert_eq!(new_buf.len(), bit_util::ceil(len, 8));
 
             // if the offset is not byte-aligned, we have to create a deep copy to a new buffer
             // (since the `offset` value inside a Buffer is byte-granular, not bit-granular), so
diff --git a/arrow-pyarrow-integration-testing/Cargo.toml b/arrow-pyarrow-integration-testing/Cargo.toml
index 0834f2d13384..630c4218352c 100644
--- a/arrow-pyarrow-integration-testing/Cargo.toml
+++ b/arrow-pyarrow-integration-testing/Cargo.toml
@@ -25,7 +25,7 @@ authors = ["Apache Arrow <dev@arrow.apache.org>"]
 license = "Apache-2.0"
 keywords = [ "arrow" ]
 edition = "2021"
-rust-version = "1.62"
+rust-version = { workspace = true }
 publish = false
 
 [lib]
diff --git a/arrow/Cargo.toml b/arrow/Cargo.toml
index a0fd96415a1d..c8e162ff9b02 100644
--- a/arrow/Cargo.toml
+++ b/arrow/Cargo.toml
@@ -31,7 +31,7 @@ include = [
     "Cargo.toml",
 ]
 edition = { workspace = true }
-rust-version = "1.70.0"
+rust-version = { workspace = true }
 
 [lib]
 name = "arrow"
diff --git a/object_store/Cargo.toml b/object_store/Cargo.toml
index 86d1392ebf61..c79e42a167d8 100644
--- a/object_store/Cargo.toml
+++ b/object_store/Cargo.toml
@@ -24,7 +24,7 @@ readme = "README.md"
 description = "A generic object store interface for uniformly interacting with AWS S3, Google Cloud Storage, Azure Blob Storage and local files."
 keywords = ["object", "storage", "cloud"]
 repository = "https://github.com/apache/arrow-rs/tree/master/object_store"
-rust-version = "1.64.0"
+rust-version = "1.70.0"
 
 [package.metadata.docs.rs]
 all-features = true

From c0a9689fb1ee051f158332970c704ae2d852ffbe Mon Sep 17 00:00:00 2001
From: itsjunetime <junewelker@gmail.com>
Date: Tue, 12 Nov 2024 11:35:02 -0700
Subject: [PATCH 3/7] Remove workspace reference in non-workspace crate

---
 arrow-pyarrow-integration-testing/Cargo.toml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arrow-pyarrow-integration-testing/Cargo.toml b/arrow-pyarrow-integration-testing/Cargo.toml
index 630c4218352c..5f8d48dc58d5 100644
--- a/arrow-pyarrow-integration-testing/Cargo.toml
+++ b/arrow-pyarrow-integration-testing/Cargo.toml
@@ -25,7 +25,7 @@ authors = ["Apache Arrow <dev@arrow.apache.org>"]
 license = "Apache-2.0"
 keywords = [ "arrow" ]
 edition = "2021"
-rust-version = { workspace = true }
+rust-version = "1.70"
 publish = false
 
 [lib]

From 1957c4595a73875ee73dd98e13e9545a29fcabb7 Mon Sep 17 00:00:00 2001
From: itsjunetime <junewelker@gmail.com>
Date: Tue, 12 Nov 2024 11:55:02 -0700
Subject: [PATCH 4/7] Fix more msrvs to be correct

---
 arrow-flight/gen/Cargo.toml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arrow-flight/gen/Cargo.toml b/arrow-flight/gen/Cargo.toml
index c7fe89beb93a..53b275b52396 100644
--- a/arrow-flight/gen/Cargo.toml
+++ b/arrow-flight/gen/Cargo.toml
@@ -20,7 +20,7 @@ name = "gen"
 description = "Code generation for arrow-flight"
 version = "0.1.0"
 edition = { workspace = true }
-rust-version = { workspace = true }
+rust-version = "1.71.1"
 authors = { workspace = true }
 homepage = { workspace = true }
 repository = { workspace = true }

From c37f8dad68fd901499b03966f2082926a623162c Mon Sep 17 00:00:00 2001
From: itsjunetime <junewelker@gmail.com>
Date: Tue, 12 Nov 2024 11:55:22 -0700
Subject: [PATCH 5/7] Fix even more msrvs to be accurate

---
 arrow-avro/Cargo.toml                | 1 -
 arrow-integration-testing/Cargo.toml | 2 +-
 arrow-json/Cargo.toml                | 1 -
 arrow-row/Cargo.toml                 | 1 -
 arrow-schema/Cargo.toml              | 2 +-
 parquet/Cargo.toml                   | 2 +-
 6 files changed, 3 insertions(+), 6 deletions(-)

diff --git a/arrow-avro/Cargo.toml b/arrow-avro/Cargo.toml
index d2436f0c15de..93e602320a1d 100644
--- a/arrow-avro/Cargo.toml
+++ b/arrow-avro/Cargo.toml
@@ -53,4 +53,3 @@ crc = { version = "3.0", optional = true }
 
 
 [dev-dependencies]
-
diff --git a/arrow-integration-testing/Cargo.toml b/arrow-integration-testing/Cargo.toml
index 7be56d919852..c3cd6346c1f9 100644
--- a/arrow-integration-testing/Cargo.toml
+++ b/arrow-integration-testing/Cargo.toml
@@ -25,7 +25,7 @@ authors = { workspace = true }
 license = { workspace = true }
 edition = { workspace = true }
 publish = false
-rust-version = { workspace = true }
+rust-version = "1.75.0"
 
 [lib]
 crate-type = ["lib", "cdylib"]
diff --git a/arrow-json/Cargo.toml b/arrow-json/Cargo.toml
index 517bb03d2064..5a3990efdfd6 100644
--- a/arrow-json/Cargo.toml
+++ b/arrow-json/Cargo.toml
@@ -60,4 +60,3 @@ rand = { version = "0.8", default-features = false, features = ["std", "std_rng"
 [[bench]]
 name = "serde"
 harness = false
-
diff --git a/arrow-row/Cargo.toml b/arrow-row/Cargo.toml
index 3754afb4dbc6..c04c84a55e1b 100644
--- a/arrow-row/Cargo.toml
+++ b/arrow-row/Cargo.toml
@@ -53,4 +53,3 @@ arrow-ord = { workspace = true }
 rand = { version = "0.8", default-features = false, features = ["std", "std_rng"] }
 
 [features]
-
diff --git a/arrow-schema/Cargo.toml b/arrow-schema/Cargo.toml
index 628d4a683cac..945c6f2ed464 100644
--- a/arrow-schema/Cargo.toml
+++ b/arrow-schema/Cargo.toml
@@ -26,7 +26,7 @@ license = { workspace = true }
 keywords = { workspace = true }
 include = { workspace = true }
 edition = { workspace = true }
-rust-version = { workspace = true }
+rust-version = "1.64"
 
 [lib]
 name = "arrow_schema"
diff --git a/parquet/Cargo.toml b/parquet/Cargo.toml
index 4064baba0947..5e2c13bf9c4d 100644
--- a/parquet/Cargo.toml
+++ b/parquet/Cargo.toml
@@ -26,7 +26,7 @@ authors = { workspace = true }
 keywords = ["arrow", "parquet", "hadoop"]
 readme = "README.md"
 edition = { workspace = true }
-rust-version = "1.70.0"
+rust-version = { workspace = true }
 
 [target.'cfg(target_arch = "wasm32")'.dependencies]
 ahash = { version = "0.8", default-features = false, features = ["compile-time-rng"] }

From 8f56f5d56540ef174aee5a28d91e8d4407d73431 Mon Sep 17 00:00:00 2001
From: itsjunetime <junewelker@gmail.com>
Date: Tue, 12 Nov 2024 12:05:37 -0700
Subject: [PATCH 6/7] Revert object_store msrv change and remove unnecessary
 ahash downgrade

---
 .github/workflows/rust.yml | 2 --
 object_store/Cargo.toml    | 2 +-
 2 files changed, 1 insertion(+), 3 deletions(-)

diff --git a/.github/workflows/rust.yml b/.github/workflows/rust.yml
index 1b65c5057de1..bdc4d0186f52 100644
--- a/.github/workflows/rust.yml
+++ b/.github/workflows/rust.yml
@@ -122,8 +122,6 @@ jobs:
         uses: ./.github/actions/setup-builder
       - name: Install cargo-msrv
         run: cargo install cargo-msrv
-      - name: Downgrade arrow dependencies
-        run: cargo update -p ahash --precise 0.8.7
       - name: Check arrow
         working-directory: arrow
         run: cargo msrv --log-target stdout verify
diff --git a/object_store/Cargo.toml b/object_store/Cargo.toml
index c79e42a167d8..86d1392ebf61 100644
--- a/object_store/Cargo.toml
+++ b/object_store/Cargo.toml
@@ -24,7 +24,7 @@ readme = "README.md"
 description = "A generic object store interface for uniformly interacting with AWS S3, Google Cloud Storage, Azure Blob Storage and local files."
 keywords = ["object", "storage", "cloud"]
 repository = "https://github.com/apache/arrow-rs/tree/master/object_store"
-rust-version = "1.70.0"
+rust-version = "1.64.0"
 
 [package.metadata.docs.rs]
 all-features = true

From 5c49b6f312ce070ae57559dfa08ea0622ad8e7a1 Mon Sep 17 00:00:00 2001
From: itsjunetime <junewelker@gmail.com>
Date: Sat, 16 Nov 2024 17:19:01 -0700
Subject: [PATCH 7/7] Remove MSRV changes

---
 .github/workflows/rust.yml                   | 2 ++
 Cargo.toml                                   | 2 +-
 arrow-avro/Cargo.toml                        | 1 +
 arrow-flight/gen/Cargo.toml                  | 2 +-
 arrow-integration-testing/Cargo.toml         | 2 +-
 arrow-json/Cargo.toml                        | 1 +
 arrow-pyarrow-integration-testing/Cargo.toml | 2 +-
 arrow-row/Cargo.toml                         | 1 +
 arrow-schema/Cargo.toml                      | 2 +-
 arrow/Cargo.toml                             | 2 +-
 parquet/Cargo.toml                           | 2 +-
 11 files changed, 12 insertions(+), 7 deletions(-)

diff --git a/.github/workflows/rust.yml b/.github/workflows/rust.yml
index bdc4d0186f52..1b65c5057de1 100644
--- a/.github/workflows/rust.yml
+++ b/.github/workflows/rust.yml
@@ -122,6 +122,8 @@ jobs:
         uses: ./.github/actions/setup-builder
       - name: Install cargo-msrv
         run: cargo install cargo-msrv
+      - name: Downgrade arrow dependencies
+        run: cargo update -p ahash --precise 0.8.7
       - name: Check arrow
         working-directory: arrow
         run: cargo msrv --log-target stdout verify
diff --git a/Cargo.toml b/Cargo.toml
index bf2274a40306..f210ae210012 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -74,7 +74,7 @@ include = [
     "Cargo.toml",
 ]
 edition = "2021"
-rust-version = "1.70"
+rust-version = "1.62"
 
 [workspace.dependencies]
 arrow = { version = "53.2.0", path = "./arrow", default-features = false }
diff --git a/arrow-avro/Cargo.toml b/arrow-avro/Cargo.toml
index 93e602320a1d..d2436f0c15de 100644
--- a/arrow-avro/Cargo.toml
+++ b/arrow-avro/Cargo.toml
@@ -53,3 +53,4 @@ crc = { version = "3.0", optional = true }
 
 
 [dev-dependencies]
+
diff --git a/arrow-flight/gen/Cargo.toml b/arrow-flight/gen/Cargo.toml
index 53b275b52396..c7fe89beb93a 100644
--- a/arrow-flight/gen/Cargo.toml
+++ b/arrow-flight/gen/Cargo.toml
@@ -20,7 +20,7 @@ name = "gen"
 description = "Code generation for arrow-flight"
 version = "0.1.0"
 edition = { workspace = true }
-rust-version = "1.71.1"
+rust-version = { workspace = true }
 authors = { workspace = true }
 homepage = { workspace = true }
 repository = { workspace = true }
diff --git a/arrow-integration-testing/Cargo.toml b/arrow-integration-testing/Cargo.toml
index c3cd6346c1f9..7be56d919852 100644
--- a/arrow-integration-testing/Cargo.toml
+++ b/arrow-integration-testing/Cargo.toml
@@ -25,7 +25,7 @@ authors = { workspace = true }
 license = { workspace = true }
 edition = { workspace = true }
 publish = false
-rust-version = "1.75.0"
+rust-version = { workspace = true }
 
 [lib]
 crate-type = ["lib", "cdylib"]
diff --git a/arrow-json/Cargo.toml b/arrow-json/Cargo.toml
index 5a3990efdfd6..517bb03d2064 100644
--- a/arrow-json/Cargo.toml
+++ b/arrow-json/Cargo.toml
@@ -60,3 +60,4 @@ rand = { version = "0.8", default-features = false, features = ["std", "std_rng"
 [[bench]]
 name = "serde"
 harness = false
+
diff --git a/arrow-pyarrow-integration-testing/Cargo.toml b/arrow-pyarrow-integration-testing/Cargo.toml
index 5f8d48dc58d5..0834f2d13384 100644
--- a/arrow-pyarrow-integration-testing/Cargo.toml
+++ b/arrow-pyarrow-integration-testing/Cargo.toml
@@ -25,7 +25,7 @@ authors = ["Apache Arrow <dev@arrow.apache.org>"]
 license = "Apache-2.0"
 keywords = [ "arrow" ]
 edition = "2021"
-rust-version = "1.70"
+rust-version = "1.62"
 publish = false
 
 [lib]
diff --git a/arrow-row/Cargo.toml b/arrow-row/Cargo.toml
index c04c84a55e1b..3754afb4dbc6 100644
--- a/arrow-row/Cargo.toml
+++ b/arrow-row/Cargo.toml
@@ -53,3 +53,4 @@ arrow-ord = { workspace = true }
 rand = { version = "0.8", default-features = false, features = ["std", "std_rng"] }
 
 [features]
+
diff --git a/arrow-schema/Cargo.toml b/arrow-schema/Cargo.toml
index 945c6f2ed464..628d4a683cac 100644
--- a/arrow-schema/Cargo.toml
+++ b/arrow-schema/Cargo.toml
@@ -26,7 +26,7 @@ license = { workspace = true }
 keywords = { workspace = true }
 include = { workspace = true }
 edition = { workspace = true }
-rust-version = "1.64"
+rust-version = { workspace = true }
 
 [lib]
 name = "arrow_schema"
diff --git a/arrow/Cargo.toml b/arrow/Cargo.toml
index c8e162ff9b02..a0fd96415a1d 100644
--- a/arrow/Cargo.toml
+++ b/arrow/Cargo.toml
@@ -31,7 +31,7 @@ include = [
     "Cargo.toml",
 ]
 edition = { workspace = true }
-rust-version = { workspace = true }
+rust-version = "1.70.0"
 
 [lib]
 name = "arrow"
diff --git a/parquet/Cargo.toml b/parquet/Cargo.toml
index 5e2c13bf9c4d..4064baba0947 100644
--- a/parquet/Cargo.toml
+++ b/parquet/Cargo.toml
@@ -26,7 +26,7 @@ authors = { workspace = true }
 keywords = ["arrow", "parquet", "hadoop"]
 readme = "README.md"
 edition = { workspace = true }
-rust-version = { workspace = true }
+rust-version = "1.70.0"
 
 [target.'cfg(target_arch = "wasm32")'.dependencies]
 ahash = { version = "0.8", default-features = false, features = ["compile-time-rng"] }