Skip to content

ncclCommSplit Segfault with Non-Blocking Init

A dangling group-job pointer causes a segfault during ncclCommSplit when using threads with non-blocking communicator init; fixed in NCCL 2.26.2.

Quick answer

A dangling group-job pointer causes a segfault during ncclCommSplit when using threads with non-blocking communicator init.

Communication#nccl#commsplit#segfault#non-blocking#threads#2.26.2

What this failure is

ncclCommSplit Segfault with Non-Blocking Init is a Communication failure seen during ML training runs. A dangling group-job pointer causes a segfault during ncclCommSplit when using threads with non-blocking communicator init; fixed in NCCL 2.26.2. Common tags: Nccl, Commsplit, Segfault, Non Blocking.

Live diagnosis, no signup

Is this what broke your run? Paste your log.

You're reading about ncclCommSplit Segfault with Non-Blocking Init. Paste your own crash log or traceback below and get the real root cause for YOUR run, not this generic entry. No account, no card. Logs are masked at ingress and never saved to account history.

training_logs.txt
No log to hand? Try one:

3 free diagnoses/day

Want 14 days on the Scale plan?

Request an evaluation code. A verified workplace organization activates up to 50 diagnoses a day, alerts, history, and follow-up questions. No credit card or automatic subscription.

Evaluate one incident

Why it happens (the mechanism)

NcclCommEnsureReady prematurely completed a groupJob. This happened before ncclCommGetAsyncError was called. Leaving a dangling pointer in the group chain. Taken together, these mechanisms explain why the failure is reproducible, why it tends to surface on specific workloads or scales, and why generic mitigation attempts often fall short without addressing the underlying cause.

What you'll observe

  • Segfault during AllGather after ncclCommSplit
  • Only with non-blocking communicator init + threads
  • Setting blocking mode avoids it

Common symptoms and what they mean

SymptomWhy it happens
SIGSEGV in ncclGroupCommJoin at include/group.hncclCommEnsureReady prematurely completed a groupJob
GDB shows *pp = 0x1 (dangling pointer) in group traversalThis happened before ncclCommGetAsyncError was called
Crash after ncclCommSplit + ncclAllGatherLeaving a dangling pointer in the group chain

Which systems are affected

  • NCCL 2.21.5 / 2.25.1 with non-blocking init
  • Multi-threaded communicator creation
  • IB clusters using SHARP

How to confirm this is the problem

Use this checklist to test the hypothesis against a small reproduction. No single line proves the root cause, so preserve the preceding events and compare one variable at a time.

  • ✓Reproduce the failure from a clean checkpoint/seed: the symptom must appear without warm-up state from a previous run.
  • ✓Verified signal present: SIGSEGV in ncclGroupCommJoin at include/group.h
  • ✓Verified signal present: GDB shows *pp = 0x1 (dangling pointer) in group traversal
  • ✓Verified signal present: Crash after ncclCommSplit + ncclAllGather
  • ✓A targeted fix from the "How to fix it" section eliminates or substantially reduces the symptom within one validation pass.

The fix and the prevention pattern

The root cause is on this page and stays free. A free account adds the exact remediation steps, saved history, and the fix on every entry in the encyclopedia.

Sign up free. Unlock the full analysis

No credit card. Daily allowance follows verified trust tier. Instant access.

Diagnose this failure in VS Code

Select the traceback or open the failed terminal, then run Denpex locally to see the initiating rank, collateral failures, exact fix, and verification command without uploading the log.

Install the free VS Code extension

NCCL errors in context

NCCL is where a distributed job reports failure, which is not the same as where it failed. The hub lists every common NCCL error next to what it actually indicates, and the environment variables that tell them apart.

Compare every nccl error side by side

Related failures to investigate next

Nccl Cuda Failure often co-occurs with this failure because both share underlying assumptions about communication; if you fix this entry but still see drift, validate the related primary cause next. nccl cuda failure →
Nccl Bucket Size Mismatch often co-occurs with this failure because both share underlying assumptions about communication; if you fix this entry but still see drift, validate the related neighboring cause next. nccl bucket size mismatch →
Nccl Nvls Dualport Corruption often co-occurs with this failure because both share underlying assumptions about communication; if you fix this entry but still see drift, validate the related neighboring cause next. nccl nvls dualport corruption →

Root cause

  • ncclCommEnsureReady prematurely completed a groupJob
  • This happened before ncclCommGetAsyncError was called
  • Leaving a dangling pointer in the group chain

The fix and how to prevent it

Evaluate Denpex on your own logs

Request a Scale evaluation code. A verified workplace organization activates 14 days with up to 50 diagnoses a day. Every account keeps its current diagnosis allowance and gets a verification path. No card or automatic subscription.

We send a single-use code tied to that address. Static provider and TLD rules do not reject valid addresses. Account trust determines the benefit after signup.

References

Don't just read the fix, diagnose your run

The encyclopedia tells you what went wrong. Denpex tells you what went wrong in YOUR training run. With your logs, your config, and your stack.

Diagnose your logs freeBrowse all failures

Related Communication errors

NCCL CUDA Failure

Communication · critical

NCCL P2P Disabled

Communication · medium

NCCL Hang Detection

Communication · high

NCCL Wrong Rank Configuration

Communication · critical