Lumpiasty/llama.cpp
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

hexagon: HMX quantized matmul rework (#23368)

Browse Source 
* hmx-mm: update debug logging in hmx-mm

* hmx-mm: update dequant logic to use HVX_vector_x2/4

* hmx-mm: remove non-pipelined version of the quantize matmul

It seems that we don't reall need non-pipelined version

* hmx-mm: use activation depth mode and update naming

Co-authored-by: Kim-Chyan Gan <kgan@qti.qualcomm.com>

* hex-mm: minor hmx matmul naming updates

* hmx-mm: remove unused vars

* snapdragon: scripts bump default ubatch-size to 1K

* hexagon: combine HMX and power and clock settings into a single set_power call

* hmx-mm: remove leftover of the scale repl helper

* hexagon: fix editconf error

---------

Co-authored-by: Kim-Chyan Gan <kgan@qti.qualcomm.com>
This commit is contained in:
Max Krasnyansky
2026-05-20 07:39:01 -07:00
committed by GitHub
co-authored by Kim-Chyan Gan
parent e947228222
commit c9872a2575
12 changed files with 212 additions and 523 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-hexagon/htp/main.c
+22 -20
View File
@@ -87,6 +87,27 @@ AEEResult htp_iface_open(const char * uri, remote_handle64 * handle) {
}
}
#if __HVX_ARCH__ >= 75
{
// Power on HMX and set HMX clock
HAP_power_request_t request;
memset(&request, 0, sizeof(HAP_power_request_t));
request.type = HAP_power_set_HMX_v2;
request.hmx_v2.set_power = TRUE;
request.hmx_v2.power_up = TRUE;
request.hmx_v2.set_clock = TRUE;
request.hmx_v2.target_corner = HAP_DCVS_EXP_VCORNER_MAX;
request.hmx_v2.min_corner = HAP_DCVS_EXP_VCORNER_MAX;
request.hmx_v2.max_corner = HAP_DCVS_EXP_VCORNER_MAX;
request.hmx_v2.perf_mode = HAP_CLK_PERF_HIGH;
FARF(ALWAYS, "Setting HMX clock\n");
err = HAP_power_set((void *) ctx, &request);
if (err != AEE_SUCCESS) {
FARF(ERROR, "Error setting HMX clock.");
return err;
}
}
#else
{
// Power on HMX
HAP_power_request_t request;
@@ -94,31 +115,12 @@ AEEResult htp_iface_open(const char * uri, remote_handle64 * handle) {
request.type = HAP_power_set_HMX;
request.hmx.power_up = TRUE;
FARF(ALWAYS, "Powering HMX on\n");
err = HAP_power_set((void *) &ctx, &request);
err = HAP_power_set((void *) ctx, &request);
if (err != AEE_SUCCESS) {
FARF(ERROR, "Error powering on HMX.");
return err;
}
}
#if __HVX_ARCH__ >= 75
{
// Set HMX clock
HAP_power_request_t request;
memset(&request, 0, sizeof(HAP_power_request_t));
request.type = HAP_power_set_HMX_v2;
request.hmx_v2.set_clock = TRUE;
request.hmx_v2.target_corner = HAP_DCVS_EXP_VCORNER_MAX;
request.hmx_v2.min_corner = HAP_DCVS_EXP_VCORNER_MAX;
request.hmx_v2.max_corner = HAP_DCVS_EXP_VCORNER_MAX;
request.hmx_v2.perf_mode = HAP_CLK_PERF_HIGH;
FARF(ALWAYS, "Setting HMX clock\n");
err = HAP_power_set((void *) &ctx, &request);
if (err != AEE_SUCCESS) {
FARF(ERROR, "Error setting HMX clock.");
return err;
}
}
#endif
return AEE_SUCCESS;