fix(stress): stagger john GPU launches to prevent GWS tuning contention
When 8 john processes start simultaneously they race for GPU memory during OpenCL GWS auto-tuning. Slower devices settle on a smaller work size (~594MiB vs 762MiB) and run at 40% instead of 100% load. Add 3s sleep between launches so each instance finishes memory allocation before the next one starts. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
@@ -191,7 +191,10 @@ CHOSEN_FORMAT=$(choose_format) || {
|
|||||||
|
|
||||||
echo "format=${CHOSEN_FORMAT}"
|
echo "format=${CHOSEN_FORMAT}"
|
||||||
PIDS=""
|
PIDS=""
|
||||||
|
_first=1
|
||||||
for opencl_id in $(echo "${JOHN_DEVICES}" | tr ',' ' '); do
|
for opencl_id in $(echo "${JOHN_DEVICES}" | tr ',' ' '); do
|
||||||
|
[ "${_first}" = "1" ] || sleep 3
|
||||||
|
_first=0
|
||||||
./john --test="${SECONDS}" --format="${CHOSEN_FORMAT}" --devices="${opencl_id}" &
|
./john --test="${SECONDS}" --format="${CHOSEN_FORMAT}" --devices="${opencl_id}" &
|
||||||
PIDS="${PIDS} $!"
|
PIDS="${PIDS} $!"
|
||||||
done
|
done
|
||||||
|
|||||||
Reference in New Issue
Block a user