sleepy/llama.cpp
1
0
Fork 0
Code Issues 10 Pull Requests 1 Actions 140 Packages Projects Releases Wiki Activity

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

Browse Source 
* 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
parent 187a456370
commit 5d2b52d80d
19 changed files with 671 additions and 245 deletions
Download Patch File Download Diff File Expand all files Collapse all files
scripts/snapdragon/windows/run-bench.ps1
+3 -3
View File
@@ -21,11 +21,11 @@ if ($null -ne $env:V) {
}
if ($null -ne $env:PROF) {
$env:GGML_HEXAGON_PROFILE=$env:PROF; $env:GGML_HEXAGON_OPSYNC=1
$env:GGML_HEXAGON_PROFILE=$env:PROF
}
if ($null -ne $env:OPMASK) {
$env:GGML_HEXAGON_OPMASK=$env:OPMASK
if ($null -ne $env:OPSTAGE) {
$env:GGML_HEXAGON_OPSTAGE=$env:OPSTAGE
}
if ($null -ne $env:NHVX) {
scripts/snapdragon/windows/run-cli.ps1
+3 -3
View File
@@ -25,11 +25,11 @@ if ($null -ne $env:SCHED) {
}
if ($null -ne $env:PROF) {
$env:GGML_HEXAGON_PROFILE=$env:PROF; $env:GGML_HEXAGON_OPSYNC=1
$env:GGML_HEXAGON_PROFILE=$env:PROF
}
if ($null -ne $env:OPMASK) {
$env:GGML_HEXAGON_OPMASK=$env:OPMASK
if ($null -ne $env:OPSTAGE) {
$env:GGML_HEXAGON_OPSTAGE=$env:OPSTAGE
}
if ($null -ne $env:NHVX) {
scripts/snapdragon/windows/run-completion.ps1
+3 -3
View File
@@ -25,11 +25,11 @@ if ($null -ne $env:SCHED) {
}
if ($null -ne $env:PROF) {
$env:GGML_HEXAGON_PROFILE=$env:PROF; $env:GGML_HEXAGON_OPSYNC=1
$env:GGML_HEXAGON_PROFILE=$env:PROF
}
if ($null -ne $env:OPMASK) {
$env:GGML_HEXAGON_OPMASK=$env:OPMASK
if ($null -ne $env:OPSTAGE) {
$env:GGML_HEXAGON_OPSTAGE=$env:OPSTAGE
}
if ($null -ne $env:NHVX) {
scripts/snapdragon/windows/run-mtmd.ps1
+3 -3
View File
@@ -34,11 +34,11 @@ if ($null -ne $env:SCHED) {
}
if ($null -ne $env:PROF) {
$env:GGML_HEXAGON_PROFILE=$env:PROF; $env:GGML_HEXAGON_OPSYNC=1
$env:GGML_HEXAGON_PROFILE=$env:PROF
}
if ($null -ne $env:OPMASK) {
$env:GGML_HEXAGON_OPMASK=$env:OPMASK
if ($null -ne $env:OPSTAGE) {
$env:GGML_HEXAGON_OPSTAGE=$env:OPSTAGE
}
if ($null -ne $env:NHVX) {
scripts/snapdragon/windows/run-tool.ps1
+3 -3
View File
@@ -31,11 +31,11 @@ if ($null -ne $env:SCHED) {
}
if ($null -ne $env:PROF) {
$env:GGML_HEXAGON_PROFILE=$env:PROF; $env:GGML_HEXAGON_OPSYNC=1
$env:GGML_HEXAGON_PROFILE=$env:PROF
}
if ($null -ne $env:OPMASK) {
$env:GGML_HEXAGON_OPMASK=$env:OPMASK
if ($null -ne $env:OPSTAGE) {
$env:GGML_HEXAGON_OPSTAGE=$env:OPSTAGE
}
if ($null -ne $env:NHVX) {