Request for RISC-V Architecture Support in Elasticsearch

[gong-flying]

Description

Description: As RISC-V architecture gains momentum in cloud-native and edge computing scenarios, we propose adding official RISC-V support to Elasticsearch. This would enable deployments on RISC-V based servers/IoT devices and align with the open-source philosophy shared by both projects.

Technical Considerations:

1. JDK Compatibility: Requires OpenJDK RISC-V port (currently experimental in JDK 21+). Track progress via OpenJDK Portola Project
2. Native Libraries:
   Rebuild JNI components (e.g., Lucene native modules, Netty transport) for RISC-V
   Verify dependencies like zlib/jemalloc compatibility
3. CI/CD Pipeline:
   Add RISC-V build targets (QEMU emulation or physical hardware)
   Cross-compilation support in Gradle build system
4. Performance Testing:
   Benchmark search/ingestion performance on RISC-V vs x86/ARM
   Memory management validation for 64-bit RV64GC

Community Impact:

1. Attract developers in RISC-V ecosystem (China/EU markets show strong adoption)
2. Enable edge computing use cases with low-power devices
3. Strengthen open-source infrastructure diversity

Proposed Milestones:
Phase 1: CI pipeline setup with RISC-V emulation
Phase 2: Core functionality validation
Phase 3: Performance optimization

Additional Context: Recent RISC-V developments show promising server-grade implementations like Alibaba Xuantie and Ventana Veyron. Reference implementations exist in Python and MySQL.

Would appreciate feedback from maintainers on feasibility and potential collaboration avenues.