RepoJournal
Linux

Linux

The kernel, distros, and the rigs of the moment

Pick a date

  1. Sun 3 may
  2. Mon 4 may
  3. Tue 5 may
  4. Wed 6 may
  5. Thu 7 may
  6. Fri 8 may
  7. Sat 9 may
  8. Sun 10 may
  9. Mon 11 may
  10. Tue 12 may
  11. Wed 13 may
  12. Thu 14 may
  13. Fri 15 may
  14. Sat 16 may
  15. Sun 17 may
  16. Tue 19 may
  17. Wed 20 may
  18. Thu 21 may
  19. Sat 23 may
  20. Sun 24 may
  21. Mon 25 may
  22. Tue 26 may
  23. Wed 27 may
  24. Fri 29 may
  25. Sat 30 may
  26. Sun 31 may
  27. Mon 1 jun
  28. Mon 8 jun
  29. Tue 9 jun
  30. Today 10 jun

The Wire · Showcase

KERNEL PATCHES PACKET LOOPS THAT CRASH NETWORK STACKS

By RepoJournal · Filed · About Linux

Linux is shipping a fundamental fix to how the kernel tracks packets bouncing through the network scheduler, closing a hole that let mirred and netem create infinite loops across CPUs.

The kernel team merged a critical patchset [1] that adds a 2-bit per-skb tc_depth field to track packet loops in the traffic control layer. The old per-CPU loop counters [2] broke in two scenarios: when packets queued through the backlog lost state, and when XPS/RPS moved packets between CPUs mid-processing. A per-skb field solves both. This lands alongside a fix to act_mirred [3] that was returning the wrong error code when redirect failed, preventing the kernel from properly accounting dropped packets. Separately, the keyring subsystem [4] picked up a critical RCU locking fix where find_key_to_update() was calling assoc_array_find() without holding the read lock, a lifetime bug that persistent key handling exposed. The nvme-tcp driver [5] also standardized its errno handling in the TLS handshake path, storing negative values consistently across all branches.

Action items

References

  1. [1] Merge branch 'net-sched-fix-packet-loops-in-mirred-and-netem' torvalds/linux
  2. [2] net: Introduce skb tc depth field to track packet loops torvalds/linux
  3. [3] net/sched: act_mirred: Fix return code in early mirred redirect error paths torvalds/linux
  4. [4] security/keys: fix missed RCU read section on lookup torvalds/linux
  5. [5] nvme-tcp: store negative errno in queue->tls_err torvalds/linux

FAQ

What changed in Linux on May 29, 2026?
Linux is shipping a fundamental fix to how the kernel tracks packets bouncing through the network scheduler, closing a hole that let mirred and netem create infinite loops across CPUs.
What should Linux teams do about it?
Pull the net-sched mirred and netem loop fixes into your next merge window • Backport the keyring RCU fix to stable kernels if you expose persistent keys • Monitor nvme-tcp systems for TLS handshake error handling consistency
Which Linux repositories shipped on May 29, 2026?
torvalds/linux

Related across the cluster

For your repos

The showcase is a teaser.
Your wire is the product.

Same engine. Different stack. Below: what changes when the wire is yours.

Showcase wire

  • •14 famous open source orgs
  • •One wire per day
  • •Public, generic
  • •Read on the web, when you remember

Your wire

  • →Up to 1,500 of your repos - orgs, deps, vendors
  • →Morning and evening briefs
  • →Action items routed to your team
  • →Slack delivery, email, breaking-news CVE alerts

Want a hands-on demo first? Ask a current user for an invite link.