Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦
C
fix(xdna): skip the lane inside an OpenMP team
Kenneth-Javier Ortigozae80291908d23
diff --git a/c/colibri.c b/c/colibri.cindex 45bd6250..92e7582f 100644--- a/c/colibri.c+++ b/c/colibri.c@@ -6418,23 +6418,29 @@ static void moe(Model *m, Layer *l, int layer, float *x, int S, float *out, int* M/K/N, so an unrelated operation of the same shape cannot inherit this* one's qualification. sh_down is deliberately absent below: its* orientation (I=sI, O=D) is not what the qualified F3 artifact- * computes. Inert without the internal test control; there is no public- * switch and no automatic policy in this slice. */+ * computes. Reached only on an explicit request (coli --xdna / COLI_XDNA);+ * there is no automatic policy.+ *+ * omp_in_parallel() skips the lane, exactly as the GPU dispatches above do:+ * backend_xdna.c keeps process-wide staging buffers and a wrapped-identity+ * pair with no locking, so it must never be entered from inside an OpenMP+ * team. Today this tail runs after the parallel region closes; the guard+ * keeps that true if it ever changes. Inside a team, matmul_qt runs. */#ifdef COLI_XDNA- if(!coli_xdna_try_matmul(COLI_XDNA_FAMILY_MOE_SHARED_GATE_UP, &l->sh_gate.xdna,- l->sh_gate.fmt, l->sh_gate.q4, l->sh_gate.s,- l->sh_gate.I, l->sh_gate.O, l->sh_gate.gs, l->sh_gate.planar,- sg, x, S))+ if(omp_in_parallel() || !coli_xdna_try_matmul(COLI_XDNA_FAMILY_MOE_SHARED_GATE_UP, &l->sh_gate.xdna,+ l->sh_gate.fmt, l->sh_gate.q4, l->sh_gate.s,+ l->sh_gate.I, l->sh_gate.O, l->sh_gate.gs, l->sh_gate.planar,+ sg, x, S))#endif#ifdef COLI_VULKANif(!vk_matmul_qt(&l->sh_gate, sg, x, S))#endifmatmul_qt(sg, x, &l->sh_gate, S);#ifdef COLI_XDNA- if(!coli_xdna_try_matmul(COLI_XDNA_FAMILY_MOE_SHARED_GATE_UP, &l->sh_up.xdna,- l->sh_up.fmt, l->sh_up.q4, l->sh_up.s,- l->sh_up.I, l->sh_up.O, l->sh_up.gs, l->sh_up.planar,- su, x, S))+ if(omp_in_parallel() || !coli_xdna_try_matmul(COLI_XDNA_FAMILY_MOE_SHARED_GATE_UP, &l->sh_up.xdna,+ l->sh_up.fmt, l->sh_up.q4, l->sh_up.s,+ l->sh_up.I, l->sh_up.O, l->sh_up.gs, l->sh_up.planar,+ su, x, S))#endif#ifdef COLI_VULKANif(!vk_matmul_qt(&l->sh_up, su, x, S))