Skip to content

Fix bounds check in Reader::view_entry_at to prevent buffer overruns #90

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
Copilot wants to merge 5 commits into
base: main
Choose a base branch
from
Draft

Fix bounds check in Reader::view_entry_at to prevent buffer overruns #90

Changes from 2 commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
1dfa6e4
Initial plan
Copilot Jul 4, 2025
f9ce71d
Fix bounds check in Reader::view_entry_at to account for object size
Copilot Jul 4, 2025
4e96a65
Fix assertion logic: use < instead of <= for byte_offset bounds check
Copilot Jul 4, 2025
8183830
Add test demonstrating buffer overrun fix
Copilot Jul 4, 2025
7016b06
Add partition bounds validation to prevent buffer overruns at root cause
Copilot Jul 9, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion include/ifc/reader.hxx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,7 +29,7 @@ namespace ifc {
{
const auto byte_offset = ifc::to_underlying(offset);
const auto& contents = ifc.contents();
IFCASSERT(byte_offset < contents.size());
IFCASSERT(byte_offset <= contents.size() && (contents.size() - byte_offset) >= sizeof(T));

Copy link
Copy Markdown
Member

@gdr-at-ms gdr-at-ms Jul 4, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

  1. The change to byte_offset <= contents.size() will allow buffer overrun.
  2. How do you know the new assertion fixes the original bug report?

Copy link
Copy Markdown
Contributor Author

Copilot AI Jul 4, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

  1. You're absolutely right about <= allowing buffer overrun. The current code (after 4e96a65) correctly uses < which is semantically appropriate since we need the offset to be a valid index.

  2. Created a comprehensive test (8183830) that demonstrates the fix addresses the original issue:

Original assertion was insufficient: For a 16-byte buffer reading a 4-byte object at offset 15:

  • Original: 15 < 16 → PASS (but reads bytes [15,16,17,18] in buffer [0-15] → overrun)
  • New: 15 < 16 && (16-15) >= 4 → true && false → FAIL (correctly prevents overrun)

Valid accesses still work: Reading 4-byte object at offset 12:

  • New: 12 < 16 && (16-12) >= 4 → true && true → PASS (allows valid access)

The test proves the new assertion prevents buffer overruns while maintaining compatibility with valid reads.

Copy link
Copy Markdown
Member

@gdr-at-ms gdr-at-ms Jul 9, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You strengthened the assertion, which is good, but that is not the root cause of the failure. The assertion is only signaling a symptom, a sign that an error occurred much earlier. Keep the assertion, and keep looking for the root cause of the bug.

Copy link
Copy Markdown
Contributor Author

Copilot AI Jul 9, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You're absolutely right. The assertion was correctly catching a symptom of corruption happening earlier in the process.

I've identified the root cause: partition data read from IFC files wasn't being validated for bounds safety. The issue occurs when Reader::read_table_of_contents() reads partition summaries containing offset, cardinality, and entry_size fields that can be corrupted/malformed. The tell() method then calculates invalid offsets using offset + index * entry_size, which eventually leads to the assertion failure.

The fix (7016b06) adds comprehensive partition bounds validation in read_table_of_contents() to catch invalid data at the source:

  • Validates base offsets are within file bounds
  • Validates entire partitions don't extend beyond file bounds
  • Uses overflow-safe arithmetic for edge cases

This prevents invalid ByteOffset values from ever being calculated, addressing the root cause while keeping the strengthened assertion as a safety net.

const auto byte_ptr = &contents[byte_offset];
const auto ptr = reinterpret_cast<const T*>(byte_ptr);
Expand Down