Skip to content

Users/gilbelee/fix ibv probe crash - #364

Merged
AtlantaPepsi merged 3 commits into
developfrom
users/gilbelee/fix-ibv-probe-crash
Oct 22, 2026
Merged

AtlantaPepsi merged 3 commits into
developfrom
users/gilbelee/fix-ibv-probe-crash

Conversation

@gilbertlee-amd

Copy link
Copy Markdown
Collaborator
  • Isolate IBV device probing in a forked child to survive buggy ibverbs providers
  • Default A2A_DIRECT=0 on MI4XX for the a2a preset

gilbertlee-amd and others added 2 commits October 21, 2026 12:02
… providers

TransferBench::GetIbvDeviceList() calls ibv_open_device()/ibv_query_device()/
ibv_query_port() on every listed RDMA device during startup topology
detection, even with no args. On a host with a buggy vendor ibverbs
provider plugin, this crashed with SIGSEGV inside the provider's
device-open path before TransferBench printed anything.

Each device is now probed in a forked child process; if the child
crashes or exits non-zero, the parent reports that device as inactive
with a warning instead of dying itself. Other devices are unaffected.

Bump version to 1.71.01 and update CHANGELOG.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
On MI4XX (gfx1250) hardware, full all-to-all lets DPX route transfers
more intuitively than restricting to directly-connected XGMI links, so
default A2A_DIRECT=0 there (still 1 on other hardware, overridable via
the env var). Applies to both the a2a and a2asweep presets, computed
once in EnvVars alongside the existing per-arch GFX_UNROLL default.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Keeps the version fallback in sync with the v1.71.01 CHANGELOG entry.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
@AtlantaPepsi
AtlantaPepsi merged commit f8a8d83 into develop Oct 22, 2026
12 of 15 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants