The gateway's tool-calling path had unit coverage (parse_tool_calls,
render_chat) but nothing exercised the real subprocess wire protocol or the
HTTP surface a coding client actually hits. #401 reports plain-text replies
where tool_calls were expected; every documented path checks out, so pin the
whole path down with a mock engine speaking SUBMIT/DATA/DONE and assert:
- non-stream: tool_calls populated, finish_reason tool_calls, no raw markers
- stream: markers suppressed across 20-way chunk splits, tool_calls delta
- tool-result round trip: <|observation|><tool_response> rendering, text reply
- no tools: plain text untouched
Also emit a stderr diagnosis when tools are declared and tool-call markers
are present in the reply but the strict parse matches nothing (typically
quantization-mangled output) pointing at COLI_TOOL_SALVAGE=1 -- the likely
field condition behind #401.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A single NaN or +Inf logit silently broke the default sampling path. +Inf
became `mx`, then `expf((Inf-mx))`/`expf((NaN-mx))` is NaN, the softmax sum went
NaN, every probability went NaN — and dist_sample's fallback loop
`if(g_pbuf[i]>0)` is false for NaN at every index, so it returned 0. Every
subsequent token: 0. No error, no warning. @KingIcyCreamProjects found it.
dist_build now takes `mx` over finite logits only, gives a non-finite logit
probability 0, and when the distribution is unusable (no finite logit, or a
non-finite/zero sum) collapses to a delta on the finite argmax and warns ONCE
on stderr — degraded, but a valid token and a visible cause, never a silent
stream of zeros. The finite argmax uses the index found during the mx pass
(robust even when lo[0] itself is NaN, where argmax_v would wrongly return 0).
tests/test_sample_nan.c: healthy logits still sample correctly; NaN/+Inf
injected at lo[0], the middle, and the last position all pick the finite
argmax; an all-non-finite vocab leaves no NaN in the buffer and doesn't crash.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
@KingIcyCreamProjects caught a real artifact on the 9950X3D (#357 thread): the
nk=8192 cell reported ~75x, far above the real ~13-40x algorithm. Root cause was
two compounding bench bugs, both verified against the code:
1. ONE frozen input per cell. partial_select_desc's median-of-three pivot is
fully deterministic (no RNG), and bench_dsa_select froze the input then ran
2000 reps on that identical array -- so all 2000 reps hit the exact same
pivot sequence. A single lucky input spiked one nk row (stable across re-runs
because it's deterministic, not because it's real).
2. brng was a static global, initialized once and NEVER reset between cells.
So each (shape,nk) cell's input depended on every prior cell's brand() draws
-- reordering nks[] or adding a shape silently shifted all later inputs.
Fix:
- brng_seed() reseeds per (shape, nk, seed) so every cell is reproducible and
independent of cell ordering.
- Report the MEDIAN of N_SEEDS=11 independent inputs, each itself a median over
N_REPEAT=2000 timing reps. A lucky pivot now moves one of 11 samples, not the
reported number.
Measured on Intel Core Ultra 9 185H (this fix, median of 11 seeds):
realistic@8192 10.97x (was 13.57x single-seed on this box)
uniform@8192 9.39x (was 7.48x)
plateau@8192 16.11x (plateau ignores the RNG, unchanged as expected)
band: 6-26x across all cells, no anomalous spikes.
Correctness is unaffected (test_dsa_select, 129 cases, unchanged). This is a bench
fidelity fix only -- no engine code touched.
Per review: the collapse starts one line before the sum — seeding
mx=lo[0] means a NaN at index 0 makes mx NaN, every (lo[i]-mx) NaN, and
the softmax is doomed at the max-finding, not the normalize. Seed mx
from -INFINITY and skip NaNs (x==x), mirroring the argmax_v change; if
nothing finite survives, fall back to mx=0 and let the post-sum guard
decide. The isfinite(s) guard is now the second line of defense rather
than the only one.
Clean logits take a byte-identical path (the extra x==x compare is
noise next to V expf calls). test_logit_nan gains the NaN-at-index-0
and all-NaN dist_build cases; test_topp's 123-case sweep still passes
on this tree, confirming no interaction with the #354 heap select.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
On the default serve path (TEMP>0, 0<NUCLEUS<1) a single NaN or +Inf in
the logits — a bad streamed expert tile, or an fp overflow in the matmul
at a low-RAM eviction boundary — poisoned softmax: g_pbuf became all-NaN,
dist_sample never satisfied cum>=u, and the fallback returned token 0. The
engine then emitted an unbroken run of token 0 with NO error. The greedy
path was equally blind: argmax_v started bv=lo[0] and `lo[i]>NaN` is always
false, so a NaN at index 0 pinned the argmax to 0.
- argmax_v: skip NaN (x==x) and seed from -inf, so it returns the max
finite/+Inf entry instead of being NaN-pinned to 0. Covers greedy decode
and the speculative-verify argmax path.
- dist_build: after the softmax sum, if s is non-finite or <=0, collapse
g_pbuf to a one-hot over the finite argmax and warn once, instead of
dividing every entry into NaN. Covers the nucleus and verify paths.
Both are O(1)/free on the happy path (one branch after the existing loop;
one extra comparison inside the existing argmax loop). Degrade + diagnose,
never silently corrupt.
test_logit_nan (wired into TEST_BINS): asserts argmax_v skips NaN/picks
+Inf, dist_build yields a finite normalized one-hot on the max finite
logit, dist_sample emits that token (not 0), and clean logits still give a
valid distribution. Fails on stock dev, passes with this change.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The DSA lightning indexer selects the top-index_topk (2048) context keys to
attend to by finding the threshold = keep-th largest attention score. It
previously full-qsorted all nk scores per layer per token (O(nk log nk)) just
to read one pivot value, then scanned the original array in position order to
build the kept set.
Replace the qsort with partial_select_desc (Hoare quickselect, median-of-three,
descending): O(nk) average to partition the keep largest into a[0..k), then the
threshold is min of that block. The two position-order scans (>thr then ==thr)
are UNCHANGED, so the kept-position set is bit-identical -- a stronger contract
than #335's sampling heap (which was multiset-only because the heap was unstable
and changed accumulation order). The quickselect pivot IS by definition the
keep-th largest, so the new threshold equals old tmp[keep-1] exactly.
Measured (bench_dsa_select, keep=2048, median of 2000 reps):
nk=2049: 119us -> 5.7us (21x)
nk=8192: 626us -> 43us (15x)
nk=32768: 2.8ms -> 0.28ms (10x)
nk=65536: 6.6ms -> 0.47ms (14x)
The gap widens with context (linear vs n-log-n). DSA only fires past index_topk,
so this is precisely the long-conversation regime where decode latency matters.
Adds test_dsa_select (in TEST_BINS): 129 cases asserting element-wise identical
kept-set vs an independent qsort reference across shapes (random, peaked,
sorted, reverse-sorted, tie-plateau, all-equal) and edges (keep==1, keep==nk).
Also directly checks the partition invariant.
Adds bench_dsa_select (on-demand, NOT a gate): reproduces the table above.
test_topp proves correctness; bench_topp measures the latency claim. It re-implements
the OLD dist_build (full-vocab qsort) inline on a private buffer and times it against
the REAL new dist_build over identical inputs in one process: V=151936, temp=0.7,
3 shapes (realistic / uniform / plateau) x 4 nuc values (0.5/0.9/0.95/0.99), 2000
timed reps each, median reported. Deliberately NOT in TEST_BINS -- it's a microbench,
not a gate. Build on demand: make tests/bench_topp && ./tests/bench_topp
dist_build() sorted the entire 151936-entry vocab by probability (qsort) on every
sampled token whenever 0 < g_nuc < 1 — the serve default — and again per draft
position under rejection sampling. Measured cost: 5.6-8.0 ms/call; the actual work
is finding the few-hundred-token head whose cumulative mass reaches g_nuc.
Replace the full qsort + linear scan with a max-heap partial select:
- Floyd heapify g_pidx over V by descending g_pbuf prob (O(V), cache-friendly)
- pop winners to the array's high end until cum >= g_nuc (k * O(log V))
- the remaining heap prefix IS the tail -> zero it, renormalize the head
Winners land in g_pidx[out..V-1] in descending order, so s2 accumulates in the
same order as before -> head is unchanged on tie-free shapes (ties were already
unspecified under the unstable qsort). All four dist_build/dist_sample contract
properties hold: g_pbuf stays id-indexed, g_pidx stays internal, the tail is
fully zeroed, the head renormalizes to 1.
No API change, no caller change, no new globals.
c/tests/test_topp.c (new): drives the REAL dist_build via the include-glm.c
pattern against an independent double-precision reimplementation of the OLD
algorithm. 123 cases: 6 sizes (1..1519) x 5 shapes (uniform/peaked/geometric/
plateau/sharptail) x 4 nuc values, plus the g_nuc>=1 guard-off paths and V=1.
Tie-free shapes compare head values within 1e-6 rel (float vs double renorm
noise); tie shapes compare value-multisets. No scratch files -> builds clean on
Windows MinGW without the unmerged mkdtemp shim (#352).
Same fix pattern as test_stops: mkdtemp with a CWD-relative template
(MinGW resolves Windows paths; /tmp is not one), and the fork/pipe/
truncate-based truncation subtest compiles only where those exist —
Windows still runs the cross-platform chunk-loop content check.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Two latent bugs in every st.h reader, both hit in the field:
- a single pread caps at ~2^31 bytes on Linux, so any tensor past
2.1 GB (bf16 embed/unembed tensors of large models qualify) came
back silently truncated with perror printing '... : Success'
(errno untouched by a short read) — the same misleading-error
symptom glm.c fixed for its own reads in #236;
- no EINTR retry.
st_pread_full loops in ST_PREAD_CHUNK pieces (1 GB default, override
for tests), retries EINTR, and reports offset + progress on failure.
All five read sites converted; behavior on well-formed files is
byte-identical (GLM oracle re-verified on this branch: 32/32).
tests/test_st_pread builds with -DST_PREAD_CHUNK=7 so a 96-byte tensor
exercises the multi-chunk loop, and forks a child against a shard
truncated after st_init (init's static bounds check means the pread
path only fires when a file shrinks under a live handle) asserting
exit(1) with a 'short read' message and no 'Success'.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
My test_stops.c called mkdtemp("/tmp/coli_stops_XXXXXX"). These binaries are
built by MinGW into NATIVE Windows .exe files, which resolve Windows paths —
"/tmp" is not one, so mkdtemp failed ENOENT and `make check` went red on the
windows job the moment #143 gave us cross-platform CI.
Now a relative template in the CWD, which is what test_compat_direct.c already
does (`#define TMPF "test_direct.tmp"`). test_uring.c uses /tmp but is
Linux-only by construction (Makefile guards it behind $(LINUX)); I copied the
wrong neighbour.
Worth being precise about what this was, because the red job looked scarier
than it was: Windows is fine. `make check` built the engine, ran the whole C
suite, and passed everything else — test_i4_grouped, kv_alloc, compat_direct —
before tripping on my temp path. The CI caught a bug in the test, not in the
port. That is #143 earning its keep on day one.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The engine armed its stop tokens from config.json's eos_token_id and nothing
else. That trusts metadata written by third-party conversion tooling, which is
a thing we already know goes wrong: the README documents a mirror shipping
int4 MTP heads that silently give 0% draft acceptance. GLM-5.2 declares THREE
eos ids (<|endoftext|>, <|user|>, <|observation|>); a converter that rewrites
config.json with a reduced list leaves the engine stopping on fewer tokens
than the model emits, and the missed ones get detokenized and printed into the
chat as literal text while generation runs past the end of the turn.
Two independent defenses:
- eos_token_id is now unioned with generation_config.json, which is
HuggingFace's authority for generation (config.json often carries a
partial legacy copy). An extra stop is harmless; a missing one is not.
- every added-token the TOKENIZER marks "special":true is armed as a stop,
whatever the configs say. Those are control tokens (<|user|>, <|assistant|>,
<sop>, [gMASK], the image/video/audio markers) and are never legitimate
content in a reply -- GLM itself lists three of them as official eos.
<think>/<tool_call>/<arg_key> are "special":false and are deliberately NOT
swept up: they are real output. tok.h was parsing added_tokens but throwing
the "special" flag away, so the distinction wasn't available to anyone.
On the real per-row checkpoint this takes the armed set from 3 to 18:
[stop] 18 stop tokens: 154820 154827 154829 154821 ... (15 from the
tokenizer's special set)
Honesty about scope: this is hygiene for a class of bug, NOT a fix for the
trailing-junk report on #298 that prompted it. I hypothesised @woolcoxm's g64
checkpoint had lost eos ids in conversion; he checked, and it hadn't -- his
config arms all three correctly. The emit path is also innocent: is_stop() is
checked BEFORE emit() at every one of the four call sites (4215, 4256, 4908,
4987), so a correctly-armed stop cannot be printed. His trailing junk is still
unexplained and is more likely quantization noise. What this commit buys is
that a checkpoint we don't control cannot leak control tokens into a reply,
which was true before and is not now.
tests/test_stops.c covers both defenses: the union, a missing
generation_config.json, BOTH configs mutilated (the tokenizer still stops all
five control tokens while leaving <think> alone), and T=NULL (the validation
path keeps config-only behaviour).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
@woolcoxm's matmul_i4_grouped_pair reads x once instead of twice for the
gate+up pair. Verified here against his branch (86e91b1 merged onto dev):
correct to ~2e-8 relative vs the double reference, and BIT-EXACT against two
separate matmul_i4_grouped calls on aligned shapes -- which is the shape the
real g64 checkpoints have (I = 2048 / 6144, gs = 64). His kernel is good.
Guarded behind COLI_HAVE_GROUPED_PAIR since the function only exists on that
branch; add -DCOLI_HAVE_GROUPED_PAIR to the test's Makefile rule when #298
lands and the pair cases activate.
The checks are deliberately asymmetric, and the reason is worth recording.
Bit-exactness is asserted ONLY when I % gs == 0: there every group is covered
by the AVX2 body, whose accumulation order matches the unfused kernel, so any
difference is a real bug. With a partial last group the tail falls to scalar
code and the compiler may contract/reassociate the fused body differently,
producing ~1e-7 differences -- rounding, not logic. My first version demanded
bit-exactness everywhere and duly "found" a bug in his kernel that did not
exist; the tell was that only `up` differed and never `gate`, which is FP luck
rather than a code path. Correctness is checked everywhere against the double
reference; identity only where identity is actually implied.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
matmul_i4_grouped is the reference the CUDA fmt=4 port (#298) is expected to
reproduce, and it had no test of its own. @woolcoxm is currently debugging a
CUDA backend against an oracle nobody had verified, which is two moving
targets at once -- and he can't cross-check on CPU, since a 5-prompt run
takes 8 hours on the 744B model.
This checks matmul_i4_grouped against a plain-C reference that dequantizes
nibble -> (v-8)*scale[i/gs] and accumulates in double, over 11 shapes: I a
clean multiple of gs, a partial last group (the glen clamp), odd I (the
scalar nibble tail), gs > I, gs=16/64/128, S>1, and the nibble extremes
0x00/0xFF -- which decode to -8/+7 because the format is offset-encoded, not
two's complement. Reading that backwards turns 15 into -1 and looks like
data-dependent noise rather than a bug.
All 11 shapes match to ~1e-8 relative, so the CPU kernel is exact and can be
trusted as the reference.
One note on the tolerance, because the first draft of this test got it wrong
and "found" a bug that wasn't there: the error is compared against the sum of
|terms|, not against |result|. A dot product of signed terms can land near
zero through cancellation, and then a 1e-6 absolute error -- ordinary f32
accumulator precision -- reads as a 1e-3 relative one. A wrong scale index or
a wrong group boundary shifts the result by a fraction of the terms, so it is
still caught at 1e-6.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The engine already tracks where each turn's wall time goes (expert-disk
service, I/O wait, expert matmul, attention, lm_head) — it just only spoke
at exit or under PROF=1. Stream it instead:
- glm.c: mux serve emits a per-turn "PROF" protocol line next to TIERS/HITS
(window deltas per request, same convention as the STAT hit%); the phase
window base is now always captured (a few loads per request).
- openai_server.py: parses PROF into a 120-turn rolling window and serves it
at /profile (read-only, same trust level as /health).
- web: new Profiling tab — stat tiles (tok/s, wall, tokens/forward, disk
service), wall-time composition bars for the last turn and the window,
per-turn throughput and stacked phase columns with hover readouts, and a
table of recent turns. Disk service is shown apart from the stack: it
overlaps with compute, so only the I/O wait the compute thread felt counts
inside wall time. Phase colours are a CVD-validated set with gaps + legend
+ table so identity never rides on colour alone.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WhTmF8yvEBgSkUKSVfZF7P
Select a portable architecture from the compiler target instead of forcing x86-64-v3 on every platform. On macOS, only enable Homebrew OpenMP when its header and library actually exist, preserving the dependency-free fallback.
On Windows the engine self-exec OMP tuning never runs (Linux/FreeBSD-only)
and posix_fadvise readahead is a compat.h no-op, so a stock Windows run
leaves large measured wins on the table. The launcher now setdefaults, on
win32 only, each independently overridable by setting the variable:
- OMP_WAIT_POLICY=active, GOMP_SPINCOUNT=200000, OMP_DYNAMIC=FALSE,
OMP_NUM_THREADS=<physical cores> (parity with the glm.c self-exec block;
COLI_NO_OMP_TUNE disables exactly this block, presence-based like the
engine). OMP_PROC_BIND/OMP_PLACES deliberately omitted and also removed
from environment_for_plan on win32: MinGW libgomp has no affinity support
("Affinity not supported on this configuration").
- DIRECT=1: unbuffered expert reads. Measured on a 9950X3D + Samsung 9100
PRO Gen5 + Win11: iobench 10.68 GB/s O_DIRECT vs 9.03 buffered (warm);
end-to-end REPLAY 0.48 -> 1.02 tok/s. Matches #162 (1.47x same class).
- PIPE=1: load/matmul overlap, byte-identical output; +8% on top of DIRECT
(PIPE_WORKERS untouched at 8 - 4/8/16 swept flat on Gen5).
- PILOT_REAL=1: real cross-layer prefetch, the only working prefetch on
Windows; +11% and expert hit rate +19 points.
Full ladder methodology and numbers: 96-token greedy REPLAY, one lever per
step, medians of 3-4 runs (see the fork tuning doc referenced in the PR).
tests/test_env_defaults.py covers the defaults, explicit-override-wins,
the kill-switch scope, and the non-win32 no-op.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
vmmlaq_s32 computes a 2x2 int32 tile (2 weight rows x 2 activation
rows) per instruction on 8-deep segments. Tile o and s in pairs,
halving weight traffic and doubling per-instruction work at S>=2.
Four independent accumulators over a 64-deep unroll keep the loop
throughput-bound (a single chained accumulator measures no better
than SDOT: latency-bound). S=1 and all tails (odd o, odd s, I not a
multiple of 16/32) keep the existing SDOT/scalar code, and scales
apply in the same order, so results are bit-identical.
Compile-time gated on __ARM_FEATURE_MATMUL_INT8. The default Darwin
build passes no -mcpu and is byte-identical (still SDOT, IDOT_KERNEL
"neon"). Opt in with ARCH=native (new Darwin Makefile knob, appends
-mcpu=<arch>), which reports IDOT_KERNEL "neon-i8mm". The same gate
lights up on any aarch64 with i8mm (Graviton3+, Grace).
test_idot grows a driver-level exactness check through matmul_qt_ex:
fmt 1 and 2, S in {2,3,4,5,8}, O in {1,2,3,64,65}, I in {16,17,100,
1408}, bitwise float equality against a plain-C reference. Green on
both build flavors.
Measured on an M5 Pro (18 threads, matmul_qt_ex microbenchmark at
GLM-5.2 expert shapes, best of 3 process runs, vs the SDOT baseline):
gateup int4 S=8 499.7 -> 1076.6 GF/s (+115%)
gateup int8 S=8 512.9 -> 1166.3 GF/s (+127%)
down int4 S=8 696.9 -> 1186.0 GF/s (+70%)
S=1 decode rows unchanged (SDOT path untouched)
opencode / the ai-sdk OpenAI-compatible client sends a large default max_tokens
(> the server's --max-tokens cap, 1024 by default), and generation_options
returned 400 "must be an integer between 1 and 1024" — even for a trivial
"hello". OpenAI-compatible servers clamp to their own ceiling rather than
reject. Now max_tokens > limit is clamped to limit; only non-int / non-positive
values are a hard error. Test updated to assert the clamp + keep the <1 reject.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Cut Windows decode disk I/O from 2.06s/tok to 1.70s/tok (budget=4), meeting the
<=2s/tok target. Two changes on the pread expert-load path:
1. compat.h: replace the posix_fadvise no-op with a real WILLNEED cache-warmer
(overlapped ReadFile into a scratch buffer -> populates the standby page cache
so the later synchronous pread faults from RAM). Re-arms the existing
expert_prefetch/PILOT/next-block prefetch chain on Windows. DONTNEED stays a
no-op (matches macOS; Windows standby-list trimming self-regulates).
Measured: hit rate 16.4% -> 27.6%.
2. glm.c: flip PIPE (async expert-load thread pool) from default OFF to default ON
on Windows. Dispatches expert pread onto worker threads so loads overlap the
matmul, instead of blocking serial load-then-compute. PIPE=0 opts out.
Measured: expert-disk 65.9s -> 54.3s (-18%).
Also adds compat_fadvise assertions to tests/test_compat_direct.c (data integrity
after cache-warmer, safe no-op on bad fd / non-WILLNEED).
mmap (CreateFileMapping/MapViewOfFile) was implemented and tested at length but
reverted: it regressed on Windows (RSS bloat from touched mapped pages collapsed
the expert cache via ullAvailPhys — a fundamental Windows-vs-Linux difference).
Full findings + the dead-end analysis recorded in issue_diskio.md.
c/tests/test_schema_gbnf and c/tests/test_compat_direct were committed
as Linux x86-64 ELF executables (slipped in via #111). On any other
platform make test-c considers them up to date and execs them, failing
with OSError: [Errno 8] Exec format error. Remove them and add the two
names to .gitignore alongside the other test binaries already listed,
so each platform rebuilds its own.
kv_alloc guards every KVState free with if(k->Lc) precisely so it can be
called again on the same KVState (context resize, slot re-init). Exercise
that path: allocate, touch the cache, allocate again at a larger size.
Fails at this commit with a double-free (fixed in the next commit):
malloc: *** error for object 0x7: pointer being freed was not allocated
* Fuse CUDA expert MLP execution
* Group CUDA expert transfers by device
* Instrument grouped CUDA expert execution
* Bound grouped CUDA decode scratch
* Execute expert groups across GPUs in parallel
* Release host backing for multi-GPU experts
* Define quality-preserving memory policies
* Overlap cold expert loading with resident compute
* Adapt expert placement with session LFRU
* Fuse q4 expert gate and up dispatch
* Plan CPU work on physical cores
* Batch grouped expert CUDA kernels
* Separate VRAM and RAM expert placement
* Add ragged multi-sequence decode forward
* feat(runtime): add continuous decode scheduler
* Route concurrent API requests through batch scheduler
* Harden multiplex request lifecycle and framing
* Cancel disconnected multiplex requests
* Bind API port before starting the engine
* fix automatic KV slot allocation
* add native int4 Tensor Core grouped GEMM
* add Tensor Core throughput benchmark
* optimize packed int4 low-row kernels
* add asynchronous CUDA staging streams
* document validated six-GPU dense acceleration
* tune six-GPU expert hot set
* raise validated expert hot-set target
* add CUDA MLA absorption core
* fuse grouped expert gate and up projections
* Warn for explicit lossy routing flags
* Add full-resident expert placement mode
* Adapt VRAM expert slots to live routes
* Accelerate int4 matvec on AVX-512
* Reduce AVX-512 and RoPE decode overhead
* Seed every GPU expert layer after prefill
* Limit live GPU swaps during decode
* CUDA batch MLA attention, kv_b head-sharding, fused o_proj, expert-group dispatch, W4A16 kernels
Lab-qualified on the 6x RTX 5090 machine (914-token request benchmark):
- batch MLA absorption kernel (COLI_CUDA_ATTN=1): whole-batch attention on
device, 154.8s -> 102.4s
- attention -> o_proj fusion on the layer device: -> 97.4s
- kv_b head-sharding across cards (COLI_CUDA_ATTN_SHARD=1), no weight
duplication: -> 94.05s
- per-device expert-group dispatch with pinned-buffer async transfers,
W4A16 tensor-core kernels for the shared expert, OMP hot-thread tuning
Negative results (reverted, kept out): GPU-side weighted scatter-add
(atomics + per-layer D2H lose 43.8%), shared-expert fused small-batch
kernel (-38.8%), W4A4 grouped tensor cores (int4 activations corrupt
output). Details in the lab research log.
* GPU resident pipeline: device-resident prefill attention chain, GPU expert groups in prefill, batched router, W4A16 mixed dispatch
COLI_CUDA_PIPE=1 keeps the prefill data plane on the layer home device;
control flow (routing, cache/pin management) stays on CPU. Any CUDA
failure falls back to the unchanged CPU path.
- Device primitives + unit tests (tests/test_pipe_cuda.cu): rmsnorm
(strided), interleaved RoPE, silu-mul, residual add, fixed-order row
merge (no atomics), device-input GEMM, persistent per-device scratch.
All verified against the engine's CPU math on SM120 (worst 1.2e-5).
- attn_pipe_prefill: q_a -> norm -> q_b -> rope -> kv_a -> norm -> rope ->
batch attention -> o_proj in one device chain (q_a/q_b/kv_a colocated
with kv_b); only the final [S,D] and the new KV rows return to host.
Attention 41.2s -> 30.8s on the 1571-token benchmark.
- Prefill batch-union now uses the GPU expert groups (previously gated to
S<=64, leaving all VRAM-resident experts idle during prefill - measured
21ms of GPU expert time in a 148s prefill). Expert phase 78.9s -> 69.0s.
- Router computed as one batched matmul instead of S sequential rows
(bit-identical math).
- W4A16 tensor-core path for expert groups (COLI_CUDA_TC_W4A16=1) with
row-count mixed dispatch: >=16 rows per expert use tensor cores, smaller
batches keep the naive kernel (tensor cores measured negative below
~16 rows). Expert phase 69.0s -> 64.3s, decode unaffected.
Net on the 1571-token prefill benchmark: 148.8s -> 114.3-126.8s
(component timings stable across runs; wall drifts +-3-5s because
.coli_usage placement learning shifts the expert tiers between runs).
PROFILO now also prints the prefill-phase breakdown.
* Skip OMP hot-thread tuning when CUDA is enabled
The active-spin worker team measured 66.9s->20.9s on the CPU-only Zen5
build, but on the six-GPU full-residency workload the spinning workers
contend with the CUDA dispatch threads: ~4x slower prefill with the
process stuck near 1.8 cores. Gate the tuning on COLI_CUDA so each
configuration keeps the behavior it was measured to prefer.
* Inc.2a: sparse layers fully resident on the layer device, residual hops cards at layer boundaries
COLI_CUDA_PIPE=2 keeps the residual stream on the layer home device for
consecutive sparse layers (cudaMemcpyPeer at boundaries): in/post norms,
attention chain, both residual adds and the shared-expert MLP run on
device. Per layer only the post-norm activations (router + CPU-tier
experts + group gather), the new KV rows and, on DSA indexer layers, the
pre-attention norm leave the card. Per-layer transfers drop from ~130MB
to ~70MB. A device-side snapshot at layer entry makes any mid-layer CUDA
failure fall back to the unchanged CPU path idempotently.
1571-token prefill: 127.1s (PIPE=1 control) -> 117.6/118.9s, components
attention 30.8->26.1, other 31.8->22.5-24.5; output verified coherent
against the control.
* Head-sharded attention inside the pipe: negative on PCIe star topology, gated opt-in
Slicing q per card from the home device and collecting ctx back
serializes ~95MB/layer through the home card's PCIe link: attention
26.1s -> 41.4/44.4s on the 1571-token benchmark (two repeats), wall
117.6 -> 135-138s. The standalone host-path sharding won because six
cards uploaded from host RAM in parallel; a home-device star has no
such parallelism without NVLink. Kept behind COLI_CUDA_PIPE_SHARD=1
for interconnects where peer bandwidth does not share one root port.
* Inc.3: device-resident KV shadow for decode attention
Decode re-uploaded the whole latent+rope window per layer per token
(~300MB/token at 1571 context). Each layer now keeps a device shadow of
the compressed KV on its kv_b card, bulk-synced when behind and appended
incrementally; the host cache stays canonical. Invalidation on kv_bind
(slot switch), kv_alloc (resize) and on any overwrite of mirrored rows,
with the legacy full-upload path as fallback.
Measured (COLI_CUDA_PIPE gate): short-context decode 5.48 -> 5.59/5.87
tok/s, 1571-context decode 4.14 -> 4.22 tok/s. Decode remains CPU-expert
bound; the shadow removes the transfer tax, not the compute.
* tools: unified user-experience benchmark (bench_ux.sh)
Two fixed scenarios (short chat, long-document QA), TTFT + decode tok/s
+ first-line drift check, TEMP=0 DRAFT=0 enforced, medians over REPS
runs. Encodes the measurement discipline from the lab record: same
binary per comparison, judge medians because .coli_usage placement
learning drifts wall times between runs.
* tools: bench_ux.sh executable bit
* gitignore compiled test binaries
* tools: expert_atlas.py — measure per-expert topic affinity (#175)
Diffs .coli_usage across 10 themed probe batches (code/math/chinese/
prose/science/law/poetry/structured/translation/casual, 3 prompts each)
driven through a running API server — one engine load total. Every
touched expert gets a topic-affinity vector, entropy, and a specialist/
generalist label; output experts.json feeds the Brain page hover.
* serve: persist .coli_usage after every turn in mux mode, not only at exit
run_serve_mux saved the learning cache once at shutdown; a crash lost
the whole session's routing history, and live consumers of the file
(expert_atlas.py diffs it between probe batches) saw a frozen snapshot.
Now saved per turn like the interactive path (165KB write, negligible).
* web: Brain hover shows measured expert atlas when published
If /experts.json (from tools/expert_atlas.py, #175) is served next to
the app, the tooltip upgrades from the depth heuristic to measured
data: specialist/generalist label, entropy, and the top-3 topic
affinities. Row index maps to real layer (row+3, last row = MTP 78).
Falls back to the heuristic when no atlas is published.
---------
Co-authored-by: JustVugg <JustVugg@users.noreply.github.com>
* SCHEMA=<file.json>: JSON-Schema -> GBNF compiler for grammar-forced drafts (#48/#70 follow-up)
schema_gbnf.h compiles a practical JSON-Schema subset (strict objects, string/
number/integer/boolean/null, enum/const, arrays with items, nesting) into the
byte-level GBNF subset grammar.h parses, so structured-output workloads get
grammar-forced drafts without hand-writing GBNF. Unsupported keywords fail soft:
the engine runs without a grammar and output is unchanged (drafts are verified,
never constraints - a wrong compile can only cost acceptance, not correctness).
grammar_setup: GRAMMAR= (raw GBNF) keeps precedence; SCHEMA= feeds the compiler
into the same gr_parse path. 8 test groups in tests/test_schema_gbnf.c walk
compiled grammars end-to-end through the PDA (forced spans, enum disambiguation,
nested instances, escapes, leading-zero rejection, fail-closed fallbacks).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* schema_gbnf: whitespace-tolerant emission (jws at separators)
Measured on GLM-5.2 current main (#146): the greedy continuation writes sloppy
JSON (spaces after colons, fences, long free text) and a compact-only grammar
desyncs at the first stray space, forfeiting every span after it. jws points are
not forced themselves (two legal bytes) but the multi-byte spans around them
keep drafting and the walker survives non-compact output - strictly
acceptance-positive for a verified draft source. Tests re-derived for the new
span boundaries + a sloppy-instance walk.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Co-authored-by: JustVugg <JustVugg@users.noreply.github.com>
* cuda-dll: fix Windows build — MSVC host flags, CUDA_PATH default, POSIX setenv shim in the kernel test (#157)
First hardware validation of the #131 CUDA_DLL path (RTX PRO 6000 Blackwell
sm_120, MSVC 14.44 + CUDA 13.2, MSYS2 UCRT64 host build) found three blockers
that made 'make cuda-dll' unbuildable as shipped:
- NVCCFLAGS passed GCC-style -Xcompiler=-Wall,-Wextra to the MSVC host
compiler (hard error D8021). Use -Xcompiler=-W3 on Windows — dash form,
since MSYS make mangles /W3 into a filesystem path.
- NVCC defaulted to $(CUDA_HOME)/bin/nvcc with CUDA_HOME=/usr/local/cuda;
on Windows default CUDA_HOME from the installer's CUDA_PATH and NVCC to
plain 'nvcc' from PATH (CUDA_PATH contains spaces, which the unquoted
recipe checks cannot survive; an MSVC PATH environment is already required).
- tests/test_backend_cuda.cu used POSIX setenv/unsetenv (undefined under
MSVC); add a two-line _putenv_s shim.
Also corrects the stale '11 API symbols' comment (the header exports 15 and
backend_loader.c resolves all 15).
Validated: make cuda-dll (stock flags) + make glm CUDA_DLL=1 ARCH=native →
[CUDA] device init on sm_120, tiny oracle TF 32/32 + greedy 20/20, kernel
suite 'q8/q4/q2/f32 correctness ok', graceful no-dll fallback, plain build
byte-identical CPU behavior.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
* cuda: fix heap corruption in expert_host_release on Windows (CUDA_RELEASE_HOST=1)
expert_host_release() freed the expert slab with plain free(), but the slab
is posix_memalign'd — which compat.h maps to _aligned_malloc on Windows, so
free() corrupts the CRT heap: instant 0xC0000374 crash on the first released
expert. This is the exact pattern the compat.h audit fixed at the original
expert_load site ("l'unico sito che libera memoria aligned e' free(s->slab)");
this call site was added later and reintroduced it. compat_aligned_free is
plain free on POSIX, so non-Windows behavior is unchanged. fslab stays plain
free (malloc/falloc on the CPU path).
Found running the VRAM expert tier on real hardware (80 GB resident on an
RTX PRO 6000, CUDA_RELEASE_HOST=1 to avoid 80 GB of host double-residency —
reproducible crash before, clean generation after).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
---------
Co-authored-by: lEWFkRAD <186512915+lEWFkRAD@users.noreply.github.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: JustVugg <JustVugg@users.noreply.github.com>
* win: direct I/O via FILE_FLAG_NO_BUFFERING + compat_fsize + VirtualLock primitives
compat_open_direct() gives Windows the O_DIRECT twin fd st.h already uses
on Linux/macOS: FILE_FLAG_NO_BUFFERING, same 4K-alignment contract as
O_DIRECT (the engine's DIRECT=1 path already aligns offset/len and slabs
are posix_memalign'd).
Measured on GLM-5.2 744B int4, Ryzen 9 9950X3D / 126 GB / PCIe4 NVMe
(5.8 GB/s at the engine's 19MBx8T pattern), Windows 11, MinGW GCC 16.1,
32-token greedy runs at --topp 0.7, 40 GB pin, current dev HEAD:
buffered: 0.38 tok/s (expert-disk dominates)
DIRECT=1: 0.56 tok/s (1.47x) — byte-identical greedy output vs buffered
compat_fsize() (GetFileSizeEx): CRT lseek(SEEK_END) returns -1 on
NO_BUFFERING fds (measured on UCRT); iobench uses it and gains a
NO_BUFFERING branch so disk numbers are comparable across platforms.
compat_mlock/compat_munlock: VirtualLock with working-set growth (bare
VirtualLock caps at the default working-set minimum, a few hundred KB).
Wired into the engine in the next commit.
tests/test_compat_direct.c covers the alignment contract, data integrity,
fsize on both fd kinds; skips cleanly off Windows.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* win: wire VirtualLock into mem_wire, munlock pairing in expert_host_release
MLOCK=1 was a silent no-op on Windows: pinned experts could be paged out
by working-set trimming under memory pressure. mem_wire now uses
compat_mlock (VirtualLock + working-set growth); expert_host_release
unlocks before freeing, mirroring the POSIX branch.
Validated: 39.6 GB pin wired in 17s on a 126 GB machine, zero failures;
TF oracle 32/32 with MLOCK=1.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* cuda: thread-local current-device cache in select_ctx
cudaSetDevice on every call is expensive when the serial expert loop
alternates devices. Measured on RTX 5090 + RTX 4090 (Windows, DLL
backend, pre-#68 dispatch): expert-matmul 14.3s -> 25.4s per 32 tokens
going from 1 to 2 devices, entirely per-call context switching. The
current device is per-thread in the CUDA runtime, so a thread_local
cache skips redundant switches; multi-GPU expert serving becomes
positive-scaling instead of negative.
Kernel suite passes on sm_120 + sm_89; TF oracle 32/32 dual-GPU.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---------
Co-authored-by: olorin <io@zyphyr.co>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
JustVugg