UPSTREAM PR #18640: memory: add is_iswa for memory_hybrid

[loci-dev]

Mirrored from ggml-org/llama.cpp#18640

Alternative to ggml-org/llama.cpp#18601 , reuse llama_memory_hybrid instead of duplicating the class

(I haven't tested it yet, just pushing this PR for discussion)

[loci-review]

Explore the complete analysis inside the Version Insights

I've successfully retrieved the summary report for your project. The report shows performance analysis for pull request #834 in the auroralabs-loci/llama.cpp repository.

Key Highlights:

• 10 functions analyzed with the most significant response time changes
• Major performance regressions detected across STL container operations
• The worst regression is in std::vector::begin() with a 68.34% increase in response time
• Memory management functions also show notable degradation (31.62% increase)

The report includes detailed metrics for each function including response times, throughput changes, and specific code locations. It also provides recommendations for investigating the performance issues, focusing on STL usage, memory operations, and compiler optimization settings.

Would you like me to provide more details about any specific aspect of this report?