Panama FP32 matmul: first call in a JVM differs by an ULP from later calls (JIT-dependent reduceLanes order)
维护者通常 1 天内回复
还没有人认领这个 Issue。
评估
- 难度
- 4/5
- 预计耗时
- 3-5 天
- 新手友好度
- 45/100
- Issue 类型
- 缺陷
- 描述清晰度
- 基本清楚
- 活跃度
- 冷清
- 技术栈
- kotlin
- 领域
- performance
调研方向
Start with PanamaVectorMatmulKernel and inspect its FloatVector.reduceLanes(ADD) reduction, then read DirectCpuExecutionContextScheduleTest for the existing warm-up behavior. Reproduce the first-call versus later-call difference on the same context with identical inputs. Done when repeated calls produce consistent results regardless of whether the JVM has compiled the intrinsic.
由索引模型根据 Issue 内容生成。
描述
Summary
On a fresh DirectCpuExecutionContext() the first dense FP32 matmul in a JVM process produces results that differ by 1 ULP (in ~37k of 153k elements for m=300, k=64, n=512) from every later call on the same context with the same inputs. After that first call the results are stable and independent of threading or schedule.
Observed while writing SKEEP-005 parity tests (DirectCpuExecutionContextScheduleTest): probe output
same-ctx twice diff=37355 (first call vs second call)
seq twice diff=0
par-vs-seq diff=0
Likely cause
PanamaVectorMatmulKernel reduces with FloatVector.reduceLanes(ADD), whose lane-reduction order is unspecified and differs between the interpreter/C1 fallback and the C2 intrinsic. The first call runs before the intrinsic is compiled.
Why it matters
- Golden-token parity gates (
*GoldenTokenParityTestin SKaiNET-transformers) assert exact llama.cpp text equality; a warm-up-dependent ULP can flip a greedy argmax. - Any test comparing two contexts must warm up first (the SKEEP-005 tests now do).
Options
- Replace
reduceLaneswith an explicit fixed-order tree reduction (reduceLanesToLong-style) or accumulate into a vector and reduce with a deterministic scalar loop. - Document the property and add a warm-up helper for parity tests.
Not fixed in feature/skeep-005-schedules; the schedule tests warm up before comparing.
- 主要语言
- Kotlin
- 星标
- 52
- 派生
- 15
- 平均合并
- 1 天 15 小时
- 30 天内合并 PR
- 36
环境准备
- 没有 Dockerfile 或 Docker Compose 文件
- 没有 Pull Request 模板
- 阅读贡献指南
从这里开始
- 先读完整个 Issue，再读项目的贡献指南。
- 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
- Fork 仓库，在一个分支上完成修改。
- 提交 Pull Request，并在描述里引用这个 Issue 编号。
SKaiNET-developers/SKaiNET 的其他 Issue
-
coding good first issue size:xs skill:kotlin-core sub-issue
难度 2/5 1-3 小时 新手友好度 88/100
SKaiNET-developers/SKaiNET#1323 ·
维护者通常 1 天内回复
-
coding good first issue platform size:xs skill:js sub-issue
难度 2/5 1-3 小时 新手友好度 72/100
SKaiNET-developers/SKaiNET#1232 ·
维护者通常 1 天内回复
-
skeep tracking
难度 5/5 一周以上 新手友好度 15/100
SKaiNET-developers/SKaiNET#1331 ·
维护者通常 1 天内回复
-
assessment size:s skill:review sub-issue
难度 4/5 1-2 天 新手友好度 18/100
SKaiNET-developers/SKaiNET#1330 ·
维护者通常 1 天内回复
-
documentation good first issue size:s skill:docs sub-issue
难度 3/5 1-2 天 新手友好度 78/100
SKaiNET-developers/SKaiNET#1329 ·
维护者通常 1 天内回复
查看 SKaiNET-developers/SKaiNET 的全部 Issue
相似的 Issue
-
难度 2/5 1-3 小时 新手友好度 83/100
Kotlin/kotlin-lsp#288 ·
-
Cassiopeia drops the `präsentiert von` credit printed in the event body可能已有人在做 关联的 PR 仍在进行中或已合并。 未关闭afk-ok area:data-quality importer size:S
难度 2/5 1-3 小时 新手友好度 77/100
enorm-labs/event-junkie#3166 ·
维护者通常 1 天内回复
-
enhancement good first issue
难度 2/5 1-3 小时 新手友好度 84/100
rajpreetcodes/Skills_MCP_Server#25 ·
维护者通常 1 天内回复
-
bugbane/qf good first issue
难度 2/5 1-3 小时 新手友好度 64/100
osservatorionessuno/bugbane#173 ·
维护者通常 1 天内回复
-
Android: build fails with Android Gradle Plugin 9 (kotlinOptions, lintOptions, proguard-android.txt)未关闭
难度 2/5 1-3 小时 新手友好度 78/100
capacitor-community/stripe#544 ·
维护者通常 1 天内回复