UPSTREAM PR #1173: feat: add sd-server API support for steps, sampler and scheduler

[loci-dev]

Mirrored from leejet/stable-diffusion.cpp#1173

This is on top of #1166 just to make it easier to test.

The step numbers should probably be limited in some way?

[loci-review]

Explore the complete analysis inside the Version Insights

I've successfully generated a comprehensive summary report for your project. The report reveals significant performance regressions in Pull Request #10 for the stable-diffusion.cpp repository.

Key Highlights:

⚠️ Critical Issues Found:

• Multiple functions showing 200%+ performance degradation
• STL container operations (begin, cbegin, back) are severely impacted
• Response times have increased dramatically (e.g., vector::cbegin went from 83.63μs to 264.43μs)

Most Affected Areas:

1. Red-black tree iterators (+220% response time)
2. Vector const iterators (+216% response time)
3. Vector back element access (+73% response time)

Recommendation: The report suggests immediate investigation is needed before merging PR #10, as these regressions indicate a systemic issue affecting core STL operations.

Would you like me to provide more details on any specific aspect of this performance analysis?

[loci-review]

Explore the complete analysis inside the Version Insights

I've successfully retrieved the summary report for your project. The analysis shows performance changes in PR #10 for the stable-diffusion.cpp repository, with 3 key functions affected:

Key Highlights:

1. _S_max_size (memory allocator) - Shows significant performance degradation with increases of 131-173%
2. end (vector iterator) - Shows significant improvement with decreases of 69-75%
3. _M_range_initialize (vector initialization) - Shows minor improvement with 3-17% decrease

The report indicates that all affected functions are related to C++ STL vector operations, suggesting the PR involves changes to data structure usage or memory management. The main concern is the regression in the memory allocation function, which may need further investigation.

Would you like more details on any specific aspect of this performance analysis?