Skip to content

MERGE_LOG.md

Latest commit

 

History

History
134 lines (106 loc) · 7.33 KB

MERGE_LOG.md

File metadata and controls

134 lines (106 loc) · 7.33 KB
Raw

Upstream merge log

Record of syncs of this fork (andreaborio/hebrus) with upstream antirez/ds4, and how the fork's engine additions were preserved across upstream refactors. See FORK_NOTES.md for per-change upstreamability and README.md ("This fork") for what each addition does.

2026-06-24 — publish sync with upstream main @ 80ebbc3

Result: ✅ landed on origin/main. sync-fix still contains upstream main plus the fork's five mainline additions. The README now also points to the separate experimental GLM-5.2 branch wip/glm52-metal64-strict-probe, clearly scoped as Apple Silicon / Metal short-context bring-up rather than a mainline feature.

No new upstream commits were present after 80ebbc3 during the 2026-06-24 refresh; the work here is publishing the already-validated sync and bringing the fork README up to date before pushing the repository now named andreaborio/hebrus.

2026-06-24 verification on M5 Pro / 64 GB

The available full-size DeepSeek V4 Flash GGUF was larger than resident memory, so model tests were run through Metal SSD streaming instead of default full residency. A global DS4_TEST_SSD_STREAMING=1 make test is not a valid local substitute for upstream's normal full-residency suite: it over-applies streaming to tests that open multiple engines and, with a 16 GiB cache under macOS lock pressure, reduced the runtime cache to pathological sizes. The accepted local matrix was therefore split by test surface:

check local command/config result
build make clean && make && make ds4_test ✅ OK
CPU portability make cpu ✅ OK (CPU-only unused warnings)
eval extractor self-test ./ds4-eval --self-test-extractors ✅ OK
agent unit tests ./ds4_agent_test ✅ OK
Q4_K dot unit tests make q4k-dot-test ✅ OK
server unit tests ./ds4_test --server ✅ OK
Metal kernel unit tests ./ds4_test --metal-kernels ✅ OK
official logprob vectors SSD streaming, requested 40 GiB cache, capped to 36 GiB ✅ OK
streaming decode/prefill correctness SSD streaming, 11 GiB cache ✅ OK
Metal tensor equivalence SSD streaming, 11 GiB cache ✅ OK
local golden vectors SSD streaming, 11 GiB cache ✅ OK
Metal short prefill SSD streaming, 11 GiB cache ✅ OK
long-context recall SSD streaming, 11 GiB cache ✅ OK
Metal SSD cache-pressure repro test's built-in 16 GiB cache ✅ OK
MTP verify depth no local DS4_TEST_MTP GGUF ✅ OK / self-skipped

Two model-behavior tests were not green in SSD streaming on this 64 GB Mac:

  • ./ds4_test --tool-call-quality: the exact path tried to access dense Metal model ranges that are not mapped by this SSD-streaming configuration.
  • ./ds4_test --think-tool-recovery: recovery triggered and generated the expected-looking DSML stanza, but the parser saw calls=0.

Those two should be re-run on a machine/configuration that can execute the normal full-residency upstream suite. They are documented here so the fork sync does not claim a stronger test result than the local hardware can support.

2026-06-18 — sync with upstream main @ 80ebbc3

Result: ✅ synced and re-validated. Branch sync-fix = upstream main + the fork's non-streaming features (imatrix-collection API, expert prune mask, PROFILE_FULL). The fork's streaming feature (#4) was dropped because it converged upstream. Passes the same ds4_test correctness suite as pre-merge main.

Divergence

  • Fork main (d2101a5) was 11 commits ahead, 18 behind upstream main (merge-base 8384adf).
  • The fork's ds4.c carried +203 / −19 lines across 17 hunks — four interwoven features in exactly the areas upstream refactored (streaming, session, prefill, imatrix):
    1. DS4_EXPERT_PROFILE_FULL toggle
    2. expert prune mask (DS4_EXPERT_PRUNE_MASK)
    3. streaming "experts-uniform" detection + span-decode helpers
    4. imatrix-collection API (ds4_session_imatrix_*, collector, prefill hooks)

What went wrong with the naive merge (and how it was caught)

A plain git merge upstream/main produced one textual conflict (weights_streaming_layer_experts_uniform); the other 16 hunks auto-merged and it built. But the CONTRIBUTING acceptance suite caught a regression:

test naive merge (720aad1) pre-merge main (d2101a5)
ds4_test --streaming-decode-prefill-correctness ❌ 65 failures, garbage decode logits (cand≈1.9e37) ✅ OK (cand==ref, rms=0)

→ a merge-introduced regression in streaming decode. Root cause: the streaming feature (#4) had fully converged upstream — antirez independently added model_map_span_vec_include_layer_decode, weights_streaming_layer_experts_uniform, streaming_layer_routed_expert_bytes and refactored ~10 streaming functions. The auto-merge textually combined our streaming edits with his in the same functions (non-overlapping lines, so no conflict flagged) → incoherent streaming → corrupt expert reads. (A first try keeping the whole fork ds4.c instead failed to build: upstream changed function signatures, error: too many arguments … expected 3, have 11.)

Fix

Take upstream's ds4.c (its streaming) and re-apply only the 13 non-streaming fork hunks (profile + prune + imatrix API) — they land in functions upstream didn't touch, so they graft cleanly (git apply --3way, 0 conflicts) and link. The 5 streaming hunks are dropped (now upstream). Commits: merge 5800f15 (take-upstream-ds4.c) → 8236528 (re-graft non-streaming).

Verification (CONTRIBUTING acceptance) — on the fix

check result
make (Metal, M5 Pro) ✅ ds4 + ds4-server built
make cpu (CPU portability) ✅ rc=0
ds4_test --server ✅ OK
ds4_test --streaming-decode-prefill-correctness ✅ OK — cand==ref, rms=0, cold_warm_neq=0 (matches main)
ds4_test --metal-ssd-streaming-cache-pressure ✅ OK (was in the 65 failures; now passes)
imatrix collection (ds4 --imatrix-dataset … --imatrix-out) ✅ valid 430 MB .dat, 387k routed observations
expert prune mask ✅ expert prune mask ACTIVE (430 experts pruned), generation coherent
ds4-bench (speed, default path) ✅ runs, sane t/s (ctx2048: prefill 153.7, gen 6.57). No regression expected: the decode path is now upstream's (the streaming test gives bit-identical logits to main), and the fork features are off by default.

imatrix speed: exempt from the no-speed-regression rule by project decision (on-edge imatrix is inherently slower when active; default path unaffected).

Still recommended before landing on main: make test (full upstream suite) and, if you want a formal speed number, a ds4-bench before/after sweep vs main. The smoke + streaming-equivalence tests above already establish the default path is unchanged.

To reproduce

git checkout -b sync-fix main
git merge upstream/main
git checkout --theirs ds4.c && git add ds4.c && git commit      # take upstream streaming
# re-apply only the non-streaming fork hunks (profile + prune + imatrix), NOT streaming:
git diff 8384adf main -- ds4.c | <keep non-stream hunks> | git apply --3way
make && DS4_TEST_SSD_STREAMING=1 ./ds4_test --streaming-decode-prefill-correctness   # must == main