Jungfraujoch

Author	SHA1	Message	Date
leonarski_fandClaude Opus 4.8	cc925b2668	Remove NUMAHWPolicy and the libnuma dependency Build Packages / build:rpm (ubuntu2404_nocuda) (push) Successful in 8m30s Details Build Packages / build:rpm (rocky8_nocuda) (push) Successful in 10m13s Details Build Packages / build:rpm (ubuntu2204_nocuda) (push) Successful in 9m45s Details Build Packages / build:rpm (rocky9_nocuda) (push) Successful in 11m13s Details Build Packages / build:rpm (rocky8_sls9) (push) Successful in 9m51s Details Build Packages / build:rpm (rocky8) (push) Successful in 8m29s Details Build Packages / build:rpm (rocky9_sls9) (push) Successful in 9m31s Details Build Packages / build:rpm (rocky9) (push) Successful in 9m42s Details Build Packages / build:rpm (ubuntu2204) (push) Successful in 8m47s Details Build Packages / build:rpm (ubuntu2404) (push) Successful in 8m23s Details Build Packages / Generate python client (push) Successful in 19s Details Build Packages / Build documentation (push) Successful in 38s Details Build Packages / Create release (push) Skipped Details Build Packages / XDS test (durin plugin) (push) Successful in 6m18s Details Build Packages / XDS test (neggia plugin) (push) Successful in 6m4s Details Build Packages / XDS test (JFJoch plugin) (push) Successful in 6m35s Details Build Packages / DIALS test (push) Successful in 10m39s Details Build Packages / Unit tests (push) Successful in 1h24m58s Details NUMA CPU/memory pinning is no longer worthwhile: the FPGA DMA buffers are placed device-local by the kernel (dma_alloc_coherent), the big RAM ring buffer is random-access (first-touch handles placement), and GPU work is already spread across all visible devices. So drop the pinning entirely and with it libnuma. - Delete NUMAHWPolicy; the only concern worth keeping - GPU selection - is done directly via pin_gpu() (round-robin over visible GPUs) in the indexer pool and the Lite analysis threads. CPU-only threads (FPGA acquire/pedestal/summation/frame-transform) no longer bind anything. - Drop get_gpu_numa_node() (sysfs lookup) - only SelectGPUAndItsNUMA used it. - numa_policy broker setting is deprecated and ignored (kept in the API for backward compatibility; warns once on startup). - Remove NUMA_LIBRARY / numa.h / numaif.h detection from CMake. - Docs: drop the NUMA dependency, remove the numa_policy config example, and document running multiple brokers on disjoint GPUs via CUDA_VISIBLE_DEVICES. - Remove NUMA_GPU_REVIEW.md (the planning note; this work is now done). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>	2026-06-17 20:25:36 +02:00
leonarski_f	c44a07aac2	v1.0.0-rc.90	2025-10-02 15:55:39 +02:00
leonarski_f	040cf08386	v1.0.0-rc.36	2025-05-05 19:32:22 +02:00
leonarski_f	ddf4c75645	v1.0.0-rc.31	2025-03-02 13:15:28 +01:00
leonarski_f	71290f374a	version 1.0.0-rc.27	2024-12-02 21:17:14 +01:00
leonarski_f	28d224afab	version 1.0.0-rc.25	2024-11-22 21:25:20 +01:00
leonarski_f	adc13ff33e	version 1.0.0-rc.24	2024-11-17 14:55:09 +01:00