(arrow) branch main updated: GH-38479: [C++] Avoid passing null pointer to LZ4 frame decompressor (#39125)

[felipecrv]

This is an automated email from the ASF dual-hosted git repository.

felipecrv pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/arrow.git


The following commit(s) were added to refs/heads/main by this push:
     new b75755a2c0 GH-38479: [C++] Avoid passing null pointer to LZ4 frame 
decompressor (#39125)
b75755a2c0 is described below

commit b75755a2c06419abda8859e56f3bcc64f148d681
Author: Antoine Pitrou <anto...@python.org>
AuthorDate: Fri Dec 8 20:04:03 2023 +0100

    GH-38479: [C++] Avoid passing null pointer to LZ4 frame decompressor 
(#39125)
    
    ### Rationale for this change
    
    Avoid undefined behavior in LZ4 when adding an offset to a null pointer.
    
    ### Are these changes tested?
    
    Yes.
    
    ### Are there any user-facing changes?
    
    No.
    * Closes: #38479
---
 cpp/src/arrow/io/compressed.cc        | 4 +++-
 cpp/src/arrow/util/compression_lz4.cc | 1 +
 2 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/cpp/src/arrow/io/compressed.cc b/cpp/src/arrow/io/compressed.cc
index 72977f0f29..6c484242a4 100644
--- a/cpp/src/arrow/io/compressed.cc
+++ b/cpp/src/arrow/io/compressed.cc
@@ -279,6 +279,8 @@ class CompressedInputStream::Impl {
   // Decompress some data from the compressed_ buffer.
   // Call this function only if the decompressed_ buffer is empty.
   Status DecompressData() {
+    DCHECK_NE(compressed_->data(), nullptr);
+
     int64_t decompress_size = kDecompressSize;
 
     while (true) {
@@ -329,7 +331,7 @@ class CompressedInputStream::Impl {
   // Try to feed more data into the decompressed_ buffer.
   Status RefillDecompressed(bool* has_data) {
     // First try to read data from the decompressor
-    if (compressed_) {
+    if (compressed_ && compressed_->size() != 0) {
       if (decompressor_->IsFinished()) {
         // We just went over the end of a previous compressed stream.
         RETURN_NOT_OK(decompressor_->Reset());
diff --git a/cpp/src/arrow/util/compression_lz4.cc 
b/cpp/src/arrow/util/compression_lz4.cc
index 17e013c13e..be957afab3 100644
--- a/cpp/src/arrow/util/compression_lz4.cc
+++ b/cpp/src/arrow/util/compression_lz4.cc
@@ -109,6 +109,7 @@ class LZ4Decompressor : public Decompressor {
     auto dst_capacity = static_cast<size_t>(output_len);
     size_t ret;
 
+    DCHECK_NE(src, nullptr);
     ret =
         LZ4F_decompress(ctx_, dst, &dst_capacity, src, &src_size, nullptr /* 
options */);
     if (LZ4F_isError(ret)) {