[linux] saturate disk (3-10gb/s depending on hardware), become compute bound.

.gitignore

@@ -67,3 +67,4 @@ c/bench/
 c/tests/test_decode_batch
 c/tests/test_i4_acc512
 c/tests/test_idot
+c/tests/test_uring

README.md

@@ -468,7 +468,7 @@ works against the colibrì OpenAI-compatible server (in review, #21) or any othe
 compatible endpoint. Nothing leaves the endpoint you configure. The terminal
 `coli chat` remains the first-class interface.
 
-Useful knobs (env or flags): `--temp T` token sampling temperature (default 0.7 + nucleus 0.90 — tuned for int4; 0 = greedy), `--topp 0.7` adaptive expert top-p (30–40% less disk), `--ngen N` max tokens per answer (`:more` in chat continues a truncated one), `--repin N` adapt RAM/VRAM hot experts every N emitted tokens, `AUTOPIN=0` disable the learning cache's auto-pin, `THINK=1` enable GLM-5.2's reasoning block, `DRAFT=n` MTP draft depth, `GRAMMAR=g.gbnf` grammar-forced drafts for constrained JSON/NDJSON output (`GRAMMAR_DRAFT=n` caps the forced span), `TF=1` teacher-forcing validation, `PILOT=1` router-lookahead disk prefetch (experimental — see below), `CAP_RAISE=0` don't auto-grow the expert cache, `PIPE=0` disable the async expert-load pool (**default ON on Windows** since `windows-optimizations` — overlaps expert `pread` with the matmul so the CPU isn't idle waiting on the SSD; measured −18% disk service time), `RAM_GB=<n>` claim more RAM for the expert cache than the conservative auto-detect (e.g. `RAM_GB=31` on a 32 GB host raises the cache cap and hit rate measurably).
+Useful knobs (env or flags): `--temp T` token sampling temperature (default 0.7 + nucleus 0.90 — tuned for int4; 0 = greedy), `--topp 0.7` adaptive expert top-p (30–40% less disk), `--ngen N` max tokens per answer (`:more` in chat continues a truncated one), `--repin N` adapt RAM/VRAM hot experts every N emitted tokens, `AUTOPIN=0` disable the learning cache's auto-pin, `THINK=1` enable GLM-5.2's reasoning block, `DRAFT=n` MTP draft depth, `GRAMMAR=g.gbnf` grammar-forced drafts for constrained JSON/NDJSON output (`GRAMMAR_DRAFT=n` caps the forced span), `TF=1` teacher-forcing validation, `PILOT=1` router-lookahead disk prefetch (experimental — see below), `URING=1` Linux-only batched expert I/O (implies `PIPE=1`; also batches `PILOT_REAL`), `PIPE=0` disable the async expert-load pool (**default ON on Windows** — overlaps expert `pread` with the matmul so the CPU isn't idle waiting on the SSD; measured −18% disk service time), `RAM_GB=<n>` claim more RAM for the expert cache than the conservative auto-detect (e.g. `RAM_GB=31` on a 32 GB host raises the cache cap and hit rate measurably), `CAP_RAISE=0` don't auto-grow the expert cache.
 
 ### Resource policy
 

c/Makefile

@@ -15,6 +15,7 @@ MINGW  := $(findstring mingw,$(TRIPLET))
 CYGWIN := $(findstring cygwin,$(TRIPLET))
 DARWIN := $(findstring darwin,$(TRIPLET))
 PPC64  := $(findstring powerpc64,$(TRIPLET))
+LINUX  := $(findstring linux,$(TRIPLET))
 IS_WIN := $(MINGW)$(CYGWIN)
 
 # Fallbacks for the rare toolchain that does not answer -dumpmachine: keep the
@@ -148,6 +149,9 @@ PYTHON    ?= python3
 endif
 CUDA_OBJ  =
 TEST_BINS = tests/test_json$(EXE) tests/test_st$(EXE) tests/test_tier$(EXE) tests/test_grammar$(EXE) tests/test_schema_gbnf$(EXE) tests/test_decode_batch$(EXE) tests/test_idot$(EXE) tests/test_kv_alloc$(EXE) tests/test_i4_acc512$(EXE) tests/test_compat_direct$(EXE)
+ifneq (,$(LINUX))
+TEST_BINS += tests/test_uring$(EXE)
+endif
 
 # Windows CUDA DLL path: host links the loader, NOT cudart.
 ifneq ($(IS_WIN),)
@@ -190,7 +194,7 @@ all: glm$(EXE)
 # phony 'glm' → 'glm.exe' on Windows (so 'make glm' and 'coli build' work on every platform)
 glm: glm$(EXE)
 
-glm$(EXE): glm.c st.h json.h tok.h tok_unicode.h compat.h grammar.h $(CUDA_OBJ) $(METAL_OBJ)
+glm$(EXE): glm.c st.h uring.h json.h tok.h tok_unicode.h compat.h grammar.h $(CUDA_OBJ) $(METAL_OBJ)
 	$(CC) $(CFLAGS) glm.c $(CUDA_OBJ) $(METAL_OBJ) -o glm$(EXE) $(LDFLAGS)
 
 # Windows runtime loader object: resolves coli_cuda_* from coli_cuda.dll.
@@ -258,7 +262,7 @@ tests/test_schema_gbnf$(EXE): tests/test_schema_gbnf.c schema_gbnf.h grammar.h j
 tests/test_decode_batch$(EXE): tests/test_decode_batch.c decode_batch.h
 	$(CC) $(CFLAGS) $< -o $@ $(LDFLAGS)
 
-tests/test_idot$(EXE): tests/test_idot.c glm.c st.h json.h tok.h tok_unicode.h compat.h grammar.h tier.h
+tests/test_idot$(EXE): tests/test_idot.c glm.c st.h uring.h json.h tok.h tok_unicode.h compat.h grammar.h tier.h
 	$(CC) $(CFLAGS) $< -o $@ $(LDFLAGS)
 
 tests/test_kv_alloc$(EXE): tests/test_kv_alloc.c glm.c st.h json.h tok.h tok_unicode.h compat.h grammar.h tier.h
@@ -270,6 +274,9 @@ tests/test_i4_acc512$(EXE): tests/test_i4_acc512.c
 tests/test_compat_direct$(EXE): tests/test_compat_direct.c compat.h
 	$(CC) $(CFLAGS) $< -o $@ $(LDFLAGS)
 
+tests/test_uring$(EXE): tests/test_uring.c glm.c st.h uring.h json.h tok.h tok_unicode.h compat.h grammar.h tier.h
+	$(CC) $(CFLAGS) $< -o $@ $(LDFLAGS)
+
 test-c: $(TEST_BINS)
 	$(PYTHON) tools/run_tests.py $(TEST_BINS)
 
@@ -303,4 +310,4 @@ clean:
 
 bench: iobench$(EXE)
 	@if [ -n "$(ARGS)" ]; then ./iobench$(EXE) $(ARGS); else echo "built iobench$(EXE) — run: ./iobench$(EXE) <file> <MB> <iters> <threads> <direct 0|1>"; fi
-.PHONY: all glm cuda-test cuda-bench cuda-dll portable test-c test-python test check clean install uninstall bench
+.PHONY: all glm cuda-test cuda-bench cuda-dll portable test-c test-python test check clean install uninstall bench