hexagon: optimization for HMX mat_mul (#21554)

* hexagon: add async HMX worker

Introduce hmx-worker (dedicated thread for HMX compute) to overlap HMX
matmul with HVX dequant/DMA stages in the pipeline path, replacing the
previous synchronous HMX calls that blocked the main thread.

* hexagon: cost-based VTCM chunk search for out-stationary matmul

* hexagon: fix futex race in hmx_worker_drain
Store the boolean to local variable avoid atomic load twice

* hex-mm: hmx optimize scatter/transpose and use HMX intrinsics

* hex-vmem: drop vmem limit a touch under 3GB on v73

* hexagon: add fwd declaration of htp_context

* hex-hmx: replace hmx-worker with hmx-queue that mimics dma-queue interface

Simplifies the overall implemantion, reduces thread wakeup roundtrips.

* hex-mm: add debug log to hmx work func called from hmx-queue

* Update hmx-queue.h

Co-authored-by: Max Krasnyansky <max.krasnyansky@gmail.com>

---------

Co-authored-by: Kim-Chyan Gan <kgan@qti.qualcomm.com>
Co-authored-by: Max Krasnyansky <maxk@qti.qualcomm.com>
Co-authored-by: Max Krasnyansky <max.krasnyansky@gmail.com>

ggml/src/ggml-hexagon/htp/hvx-base.h

@@ -116,9 +116,14 @@ static inline HVX_VectorPred hvx_vec_is_nan_f16(HVX_Vector v) {
 }
 
 static inline HVX_Vector hvx_vec_f32_to_f16_shuff(HVX_Vector v0, HVX_Vector v1) {
+#if __HVX_ARCH__ >= 81
+    HVX_Vector q0 = Q6_Vqf32_equals_Vsf(v0);
+    HVX_Vector q1 = Q6_Vqf32_equals_Vsf(v1);
+#else
     const HVX_Vector zero = Q6_V_vzero();
     HVX_Vector q0 = Q6_Vqf32_vadd_VsfVsf(v0, zero);
     HVX_Vector q1 = Q6_Vqf32_vadd_VsfVsf(v1, zero);
+#endif
     return Q6_Vhf_equals_Wqf32(Q6_W_vcombine_VV(q1, q0));
 }