hexagon: add support for basic and extended Op profiling (#22269)

* hexagon: restore HTP_OPMASK_QUEUE

* hexagon: honor OPMASK_SKIP_COMPUTE in hmx-matmul

* hex-prof: restore op profiling

* hex-prof: enable PMU

* hexagon: simplify and improve op-queuing with full profiling support

Add separate profile descriptors.

* hexagon: remove opsync and rename opmask into opstage

opsync is no longer needed since the profiler is fully async now.
opmask name was confusing and opstage is more accurate.

* hexagon: refactor opbatch queue handling

* hexagon: add iface hooks for enabling profiler from the host

Also move all the PMU setup stuff out of the hex-utils since it's not inteded for normal use.

* hexagon: make profiler mode configurable

On older devices getting PMU counters is expensive so it's now optional.

* hexagon: add support for setting profiler pmu events from env

* hexagon: simplify profiler output (no need to print buffs, etc)

* hexagon: simplify pmu counter formating

* hexagon: add a simple profile post-proc tool

* hex-prof: add support for reading logs from stdin

* hexagon: document GGML_HEXAGON_PROFILE

* hex-prof: update default width for dims field

* hex-prof: fix linter warnings and errors

* Update ggml/src/ggml-hexagon/htp/htp-ops.h

Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>

* Update scripts/snapdragon/ggml-hexagon-profile.py

Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>

---------

Co-authored-by: Trivikram Reddy <tamarnat@qti.qualcomm.com>
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>
This commit is contained in:
Max Krasnyansky
2026-04-23 14:17:21 -07:00
committed by GitHub
co-authored by Sigbjørn Skjæret Trivikram Reddy
parent 187a456370
commit 5d2b52d80d
19 changed files with 671 additions and 245 deletions
+27 -9
View File
@@ -42,9 +42,9 @@ enum htp_data_type {
// Mask to enable various stages of the Ops.
// Used for debugging and profiling.
enum htp_op_mask {
HTP_OPMASK_QUEUE = (1 << 0), // Enable Queueing (ie calls into the DSP)
HTP_OPMASK_COMPUTE = (1 << 1), // Enable Compute
enum htp_op_stage {
HTP_OPSTAGE_QUEUE = (1 << 0), // Enable Queueing (ie calls into NPU)
HTP_OPSTAGE_COMPUTE = (1 << 1), // Enable Compute
};
// Do not reorder first 4 (used as an index)
@@ -137,27 +137,45 @@ struct htp_op_desc {
int32_t params[HTP_OP_MAX_PARAMS]; // Params for the op, e.g. epsilon of RMS norm
uint16_t src[HTP_OP_MAX_INPUTS]; // Input tensors indices
uint16_t dst; // Output tensor index
};
// the rest is filled in-place by the NPU
uint32_t prof_usecs; // Number of usec per request
uint32_t prof_cycles; // Number of cycles per request
uint32_t prof_pkts; // Number of instruction packets per request
uint32_t unused;
enum htp_profiler_mode {
HTP_PROF_DISABLED = 0,
HTP_PROF_BASIC = 1,
HTP_PROF_PMU = 2,
};
#define HTP_PROF_PMU_NCNT 8
// Profile descriptor
struct htp_prof_desc {
uint32_t opcode; // GGML/HTP Op
uint32_t usecs; // Number of usec
uint32_t cycles; // Number of cycles
uint32_t pad; // Unused
uint32_t pmu[HTP_PROF_PMU_NCNT]; // PMU counters
};
struct htp_opbatch_req {
uint32_t id; // Batch id
uint32_t n_bufs; // Number of buffers
uint32_t n_tensors; // Number of tensors
uint32_t n_ops; // Number of ops
uint32_t flags; // unused
uint32_t pad; // unused
// struct htp_buf_desc bufs[]; -- dspqueue buf 0
// struct htp_tensor tensors[]; -- dspqueue buf 0
// struct htp_op_desc ops[]; -- dspqueue buf 0
};
struct htp_opbatch_rsp {
uint32_t id; // Batch id
uint32_t status; // HTP_STATUS_...
// struct htp_op_req ops[]; -- dspqueue buf 0
uint32_t n_bufs; // Number of buffers
uint32_t n_tensors; // Number of tensors
uint32_t n_ops; // Number of op profile descriptors
uint32_t pad; // unused
// struct htp_prof_desc profs[]; -- dspqueue buf 0
};
#endif /* HTP_OPS_H */