fix: null pointer crash in increment_step_counters + wire PER rng_step

Browse Source

Root cause: rng_step was passed as 0u64 (null) to increment_step_counters
kernel, causing atomicAdd on address 0 → CUDA_ERROR_ILLEGAL_ADDRESS →
cascading CUBLAS_STATUS_EXECUTION_FAILED on all subsequent operations.

Fixes:
- Replace all atomicAdd with plain writes (single-thread kernel)
- Add null guards for optional pointers (iqn_t, attn_t, rng_step)
- Allocate pinned device-mapped rng_step in GpuReplayBuffer
- Wire rng_step_dev_ptr from replay buffer to FusedTrainingCtx
- Remove stale atomicAdd comment

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

...

This commit is contained in:

jgrusewski

2026-04-19 01:36:37 +02:00

parent 1d3a533f99

commit 56c5458008

5 changed files with 43 additions and 4755 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download Patch File Download Diff File Expand all files Collapse all files

4742

pal_precommit.changeset

File diff suppressed because it is too large Load Diff