⚠ This page is served via a proxy. Original site: https://github.com
This service does not collect credentials or authentication data.
Skip to content

Conversation

@masaori335
Copy link
Contributor

I found that if OpenDir has own reader-writer lock, it doesn't need StripeSM mutex. This means we can avoid the StripeSM mutex lock contention issue for reader-while-writer cases.

Benchmarking RWW is a bit tricky but one of benchmark says max rps is improved 9.9% in below conditions.

Conditions:

  • 10 urls
  • plaintext http
  • response body size: 256 bytes
  • origin returns Cache-Control: public, max-age=0 ///< some requests goes RWW path
  • 63 exec_thread
  • 40 stripes (8 disks x 5 volumes)

Result:

  • vanilla: 58,220.9 req/s
  • patch: 63,999.7 req/s

part of #12788

Copy link
Contributor

Copilot AI left a comment

Choose a reason for hiding this comment

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

Pull request overview

This PR optimizes cache read performance by introducing a dedicated reader-writer lock (ts::bravo::shared_mutex) in the OpenDir class, reducing mutex contention for reader-while-writer (RWW) scenarios. The change allows concurrent read operations on OpenDir without requiring the StripeSM mutex, resulting in a reported 9.9% improvement in maximum requests per second under specific test conditions.

Changes:

  • Added ts::bravo::shared_mutex to OpenDir for fine-grained locking instead of relying on StripeSM mutex
  • Refactored open_read to work without holding the stripe mutex, enabling lock-free RWW path
  • Introduced new_CacheVC_for_read helper function to reduce code duplication in Cache.cc
  • Added CACHE_EVENT_OPEN_DIR_RETRY event type for handling retries in the new locking scheme
  • Made OpenDir members private and reorganized API documentation in StripeSM.h

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 6 comments.

Show a summary per file
File Description
src/iocore/cache/StripeSM.h Added API grouping comments to clarify OpenDir and PreservationTable methods
src/iocore/cache/P_CacheDir.h Changed OpenDir to class with private members, added _shared_mutex for concurrency control
src/iocore/cache/CacheDir.cc Implemented shared/exclusive locking in open_write, close_write, and open_read; updated signal_readers for new locking model
src/iocore/cache/Cache.cc Refactored open_read to check OpenDir without stripe mutex first, added helper function for CacheVC creation
include/iocore/cache/CacheDefs.h Added CACHE_EVENT_OPEN_DIR_RETRY event type for retry handling
Comments suppressed due to low confidence (1)

src/iocore/cache/CacheDir.cc:193

  • Potential use-after-free race condition: The open_read() method returns a raw pointer to an OpenDirEntry while holding a shared lock, but the lock is released when the function returns (line 183 creates a scoped lock). The caller then uses this pointer without any lock protection. Meanwhile, close_write() can acquire the writer lock and free the OpenDirEntry (line 174: THREAD_FREE). This creates a window where the returned OpenDirEntry pointer could be freed while the caller is still using it, leading to a use-after-free. The old code avoided this because both operations held the stripe mutex. Consider using reference counting for OpenDirEntry or ensuring the caller holds some protection until done with the entry.
OpenDirEntry *
OpenDir::open_read(const CryptoHash *key) const
{
  ts::bravo::shared_lock lock(_shared_mutex);

  unsigned int h = key->slice32(0);
  int          b = h % OPEN_DIR_BUCKETS;
  for (OpenDirEntry *d = _bucket[b].head; d; d = d->link.next) {
    if (d->writers.head->first_key == *key) {
      return d;
    }
  }
  return nullptr;
}

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

*/
int
OpenDir::signal_readers(int /* event ATS_UNUSED */, Event * /* e ATS_UNUSED */)
OpenDir::signal_readers(int event, Event * /* ATS UNUSED*/)
Copy link

Copilot AI Jan 12, 2026

Choose a reason for hiding this comment

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

The comment formatting uses "ATS UNUSED" with a space, but elsewhere in the codebase it's typically "ATS_UNUSED" with an underscore. Consider maintaining consistency with the existing pattern.

Suggested change
OpenDir::signal_readers(int event, Event * /* ATS UNUSED*/)
OpenDir::signal_readers(int event, Event * /* ATS_UNUSED */)

Copilot uses AI. Check for mistakes.
Comment on lines +568 to +570
// Read-While-Writer
// This OpenDirEntry lookup doesn't need stripe mutex lock because OpenDir has own reader-writer lock
od = stripe->open_read(key);
Copy link

Copilot AI Jan 12, 2026

Choose a reason for hiding this comment

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

Potential lock ordering issue: This code calls stripe->open_read() (which now acquires _shared_mutex as a reader) while holding stripe->mutex. In other places like CacheRead.cc line 212, the same pattern exists. However, in Cache.cc line 570, open_read() is called without holding stripe->mutex. While this change is intentional to reduce contention, ensure that this mixed lock ordering (sometimes stripe->mutex then _shared_mutex, sometimes just _shared_mutex) doesn't introduce potential deadlock scenarios. The Bravo reader-writer lock should handle this safely since multiple readers can coexist, but this should be verified through testing under high concurrency.

Copilot uses AI. Check for mistakes.
Copy link
Contributor Author

Choose a reason for hiding this comment

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

I guess this is fine, because prior to the change, when we jump to the CacheVC::openReadFromWriter from the Lwriter label, it's out of scope of StripeSM mutex lock. CacheVC::openReadFromWriter has another check of OpenDir.

Copy link
Contributor Author

Choose a reason for hiding this comment

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

od = stripe->open_read(&first_key); // recheck in case the lock failed

Copy link
Contributor Author

Choose a reason for hiding this comment

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

Ah, the reader lock of OpenDir might be need to be hold while this OpenDirEntry is used in the CacheVC::openReadFromWriter. Let me consider in case.

Comment on lines +229 to 246
class OpenDir : public Continuation
{
public:
OpenDir();

int open_write(CacheVC *c, int allow_if_writers, int max_writers);
int close_write(CacheVC *c);
OpenDirEntry *open_read(const CryptoHash *key) const;
int signal_readers(int event, Event *e);

OpenDir();
// event handler
int signal_readers(int event, Event *e);

private:
mutable ts::bravo::shared_mutex _shared_mutex;

Queue<CacheVC, Link_CacheVC_opendir_link> _delayed_readers;
DLL<OpenDirEntry> _bucket[OPEN_DIR_BUCKETS];
};
Copy link

Copilot AI Jan 12, 2026

Choose a reason for hiding this comment

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

Missing documentation: The class documentation should be updated to explain the new locking mechanism. Specifically, document that OpenDir now uses its own reader-writer lock (_shared_mutex) instead of relying on the StripeSM mutex, which allows concurrent reads while writes are in progress. This is a significant architectural change that should be clearly documented for maintainability.

Copilot uses AI. Check for mistakes.
Comment on lines +104 to +112
// OpenDir API
int open_write(CacheVC *cont, int allow_if_writers, int max_writers);
int open_write_lock(CacheVC *cont, int allow_if_writers, int max_writers);
int close_write(CacheVC *cont);
OpenDirEntry *open_read(const CryptoHash *key) const;
int close_read(CacheVC *cont) const;

// PreservationTable API
int begin_read(CacheVC *cont) const;
int close_read(CacheVC *cont) const;
Copy link

Copilot AI Jan 12, 2026

Choose a reason for hiding this comment

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

Inconsistent formatting: The comment "// OpenDir API" and "// PreservationTable API" use double slashes, but there are also methods without such grouping comments. Consider whether this level of API grouping is helpful, and if so, ensure it's applied consistently throughout the class. For example, other method groups could benefit from similar API comments.

Copilot uses AI. Check for mistakes.
#ifdef DEBUG

DbgCtl dbg_ctl_cache_stats{"cache_stats"};
DbgCtl dbg_ctl_cache_open_dir{"cache_open_dir"};
Copy link

Copilot AI Jan 12, 2026

Choose a reason for hiding this comment

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

Unused debug control: The debug control variable dbg_ctl_cache_open_dir is defined but never used in the code. If this was added for future debugging purposes, consider adding a comment explaining its intended use. Otherwise, it should be removed to avoid clutter.

Suggested change
DbgCtl dbg_ctl_cache_open_dir{"cache_open_dir"};

Copilot uses AI. Check for mistakes.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant